From cd3bafc73d11eb51cb2d3691629718431e1768ce Mon Sep 17 00:00:00 2001 From: Hajime Tazaki Date: Wed, 4 Feb 2015 23:31:10 +0900 Subject: xfrm6: Fix a offset value for network header in _decode_session6 When a network-layer header has multiple IPv6 extension headers, then offset for mobility header goes wrong. This regression breaks an xfrm policy lookup for a particular receive packet. Binding update packets of Mobile IPv6 are all discarded without this fix. Fixes: de3b7a06dfe1 ("xfrm6: Fix transport header offset in _decode_session6.") Signed-off-by: Hajime Tazaki Signed-off-by: Steffen Klassert --- net/ipv6/xfrm6_policy.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/ipv6/xfrm6_policy.c b/net/ipv6/xfrm6_policy.c index 48bf5a06847b..8d2d01b4800a 100644 --- a/net/ipv6/xfrm6_policy.c +++ b/net/ipv6/xfrm6_policy.c @@ -200,6 +200,7 @@ _decode_session6(struct sk_buff *skb, struct flowi *fl, int reverse) #if IS_ENABLED(CONFIG_IPV6_MIP6) case IPPROTO_MH: + offset += ipv6_optlen(exthdr); if (!onlyproto && pskb_may_pull(skb, nh + offset + 3 - skb->data)) { struct ip6_mh *mh; -- cgit v1.2.1 From 044a832a7779c0638bea2d0fea901c055b995f4a Mon Sep 17 00:00:00 2001 From: Steffen Klassert Date: Mon, 12 Jan 2015 13:38:49 +0100 Subject: xfrm: Fix local error reporting crash with interfamily tunnels We set the outer mode protocol too early. As a result, the local error handler might dispatch to the wrong address family and report the error to a wrong socket type. We fix this by setting the outer protocol to the skb after we accessed the inner mode for the last time, right before we do the atcual encapsulation where we switch finally to the outer mode. Reported-by: Chris Ruehl Tested-by: Chris Ruehl Signed-off-by: Steffen Klassert --- net/ipv4/xfrm4_output.c | 2 +- net/ipv6/xfrm6_output.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/net/ipv4/xfrm4_output.c b/net/ipv4/xfrm4_output.c index d5f6bd9a210a..dab73813cb92 100644 --- a/net/ipv4/xfrm4_output.c +++ b/net/ipv4/xfrm4_output.c @@ -63,6 +63,7 @@ int xfrm4_prepare_output(struct xfrm_state *x, struct sk_buff *skb) return err; IPCB(skb)->flags |= IPSKB_XFRM_TUNNEL_SIZE; + skb->protocol = htons(ETH_P_IP); return x->outer_mode->output2(x, skb); } @@ -71,7 +72,6 @@ EXPORT_SYMBOL(xfrm4_prepare_output); int xfrm4_output_finish(struct sk_buff *skb) { memset(IPCB(skb), 0, sizeof(*IPCB(skb))); - skb->protocol = htons(ETH_P_IP); #ifdef CONFIG_NETFILTER IPCB(skb)->flags |= IPSKB_XFRM_TRANSFORMED; diff --git a/net/ipv6/xfrm6_output.c b/net/ipv6/xfrm6_output.c index ca3f29b98ae5..010f8bd2d577 100644 --- a/net/ipv6/xfrm6_output.c +++ b/net/ipv6/xfrm6_output.c @@ -114,6 +114,7 @@ int xfrm6_prepare_output(struct xfrm_state *x, struct sk_buff *skb) return err; skb->ignore_df = 1; + skb->protocol = htons(ETH_P_IPV6); return x->outer_mode->output2(x, skb); } @@ -122,7 +123,6 @@ EXPORT_SYMBOL(xfrm6_prepare_output); int xfrm6_output_finish(struct sk_buff *skb) { memset(IP6CB(skb), 0, sizeof(*IP6CB(skb))); - skb->protocol = htons(ETH_P_IPV6); #ifdef CONFIG_NETFILTER IP6CB(skb)->flags |= IP6SKB_XFRM_TRANSFORMED; -- cgit v1.2.1 From ac37e2515c1a89c477459a2020b6bfdedabdb91b Mon Sep 17 00:00:00 2001 From: huaibin Wang Date: Wed, 11 Feb 2015 18:10:36 +0100 Subject: xfrm: release dst_orig in case of error in xfrm_lookup() dst_orig should be released on error. Function like __xfrm_route_forward() expects that behavior. Since a recent commit, xfrm_lookup() may also be called by xfrm_lookup_route(), which expects the opposite. Let's introduce a new flag (XFRM_LOOKUP_KEEP_DST_REF) to tell what should be done in case of error. Fixes: f92ee61982d("xfrm: Generate blackhole routes only from route lookup functions") Signed-off-by: huaibin Wang Signed-off-by: Nicolas Dichtel Signed-off-by: Steffen Klassert --- include/net/dst.h | 1 + net/xfrm/xfrm_policy.c | 12 ++++++------ 2 files changed, 7 insertions(+), 6 deletions(-) diff --git a/include/net/dst.h b/include/net/dst.h index a8ae4e760778..0fb99a26e973 100644 --- a/include/net/dst.h +++ b/include/net/dst.h @@ -481,6 +481,7 @@ void dst_init(void); enum { XFRM_LOOKUP_ICMP = 1 << 0, XFRM_LOOKUP_QUEUE = 1 << 1, + XFRM_LOOKUP_KEEP_DST_REF = 1 << 2, }; struct flowi; diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c index cee479bc655c..638af0655aaf 100644 --- a/net/xfrm/xfrm_policy.c +++ b/net/xfrm/xfrm_policy.c @@ -2269,11 +2269,9 @@ struct dst_entry *xfrm_lookup(struct net *net, struct dst_entry *dst_orig, * have the xfrm_state's. We need to wait for KM to * negotiate new SA's or bail out with error.*/ if (net->xfrm.sysctl_larval_drop) { - dst_release(dst); - xfrm_pols_put(pols, drop_pols); XFRM_INC_STATS(net, LINUX_MIB_XFRMOUTNOSTATES); - - return ERR_PTR(-EREMOTE); + err = -EREMOTE; + goto error; } err = -EAGAIN; @@ -2324,7 +2322,8 @@ nopol: error: dst_release(dst); dropdst: - dst_release(dst_orig); + if (!(flags & XFRM_LOOKUP_KEEP_DST_REF)) + dst_release(dst_orig); xfrm_pols_put(pols, drop_pols); return ERR_PTR(err); } @@ -2338,7 +2337,8 @@ struct dst_entry *xfrm_lookup_route(struct net *net, struct dst_entry *dst_orig, struct sock *sk, int flags) { struct dst_entry *dst = xfrm_lookup(net, dst_orig, fl, sk, - flags | XFRM_LOOKUP_QUEUE); + flags | XFRM_LOOKUP_QUEUE | + XFRM_LOOKUP_KEEP_DST_REF); if (IS_ERR(dst) && PTR_ERR(dst) == -EREMOTE) return make_blackhole(net, dst_orig->ops->family, dst_orig); -- cgit v1.2.1 From 09cadf6e088b59c335116bf0e2667486bc126c6a Mon Sep 17 00:00:00 2001 From: Valentin Rothberg Date: Wed, 11 Feb 2015 16:37:57 +0100 Subject: regmap-irq: set IRQF_ONESHOT flag to ensure IRQ request Since commit 1c6c69525b40eb76de8adf039409722015927dc3 ("genirq: Reject bogus threaded irq requests") threaded IRQs without a primary handler need to be requested with IRQF_ONESHOT, otherwise the request will fail. The %irq_flags flag is used to request the threaded IRQ and is also a parameter of the caller. Hence, we cannot be sure that IRQF_ONESHOT is set. This change avoids the potentially missing flag by setting IRQF_ONESHOT when requesting the threaded IRQ. Generated by: scripts/coccinelle/misc/irqf_oneshot.cocci Signed-off-by: Valentin Rothberg Signed-off-by: Mark Brown --- drivers/base/regmap/regmap-irq.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/base/regmap/regmap-irq.c b/drivers/base/regmap/regmap-irq.c index 6299a50a5960..a6c3f75b4b01 100644 --- a/drivers/base/regmap/regmap-irq.c +++ b/drivers/base/regmap/regmap-irq.c @@ -499,7 +499,8 @@ int regmap_add_irq_chip(struct regmap *map, int irq, int irq_flags, goto err_alloc; } - ret = request_threaded_irq(irq, NULL, regmap_irq_thread, irq_flags, + ret = request_threaded_irq(irq, NULL, regmap_irq_thread, + irq_flags | IRQF_ONESHOT, chip->name, d); if (ret != 0) { dev_err(map->dev, "Failed to request IRQ %d for %s: %d\n", -- cgit v1.2.1 From fdc0074c5fc8c7adb8186cbb123fe2082d9bd05f Mon Sep 17 00:00:00 2001 From: Chen-Yu Tsai Date: Mon, 9 Feb 2015 18:23:20 +0800 Subject: ARM: sunxi: Have ARCH_SUNXI select RESET_CONTROLLER for clock driver usage As the sunxi usb clocks all contain a reset controller, it is not possible to build the sunxi clock driver without RESET_CONTROLLER enabled. Doing so results in an undefined symbol error: drivers/built-in.o: In function `sunxi_gates_clk_setup': linux/drivers/clk/sunxi/clk-sunxi.c:1071: undefined reference to `reset_controller_register' This is possible if building a minimal kernel without PHY_SUN4I_USB. The dependency issue is made visible at compile time instead of link time by the new A80 mmc clocks, which also use a reset control itself. This patch makes ARCH_SUNXI select ARCH_HAS_RESET_CONTROLLER and RESET_CONTROLLER. Fixes: 559482d1f950 ARM: sunxi: Split the various SoCs support in Kconfig Cc: # 3.16+ Reported-by: Lourens Rozema Acked-by: Arnd Bergmann Signed-off-by: Chen-Yu Tsai Signed-off-by: Maxime Ripard --- arch/arm/mach-sunxi/Kconfig | 8 ++------ 1 file changed, 2 insertions(+), 6 deletions(-) diff --git a/arch/arm/mach-sunxi/Kconfig b/arch/arm/mach-sunxi/Kconfig index a77604fbaf25..81502b90dd91 100644 --- a/arch/arm/mach-sunxi/Kconfig +++ b/arch/arm/mach-sunxi/Kconfig @@ -1,10 +1,12 @@ menuconfig ARCH_SUNXI bool "Allwinner SoCs" if ARCH_MULTI_V7 select ARCH_REQUIRE_GPIOLIB + select ARCH_HAS_RESET_CONTROLLER select CLKSRC_MMIO select GENERIC_IRQ_CHIP select PINCTRL select SUN4I_TIMER + select RESET_CONTROLLER if ARCH_SUNXI @@ -20,10 +22,8 @@ config MACH_SUN5I config MACH_SUN6I bool "Allwinner A31 (sun6i) SoCs support" default ARCH_SUNXI - select ARCH_HAS_RESET_CONTROLLER select ARM_GIC select MFD_SUN6I_PRCM - select RESET_CONTROLLER select SUN5I_HSTIMER config MACH_SUN7I @@ -37,16 +37,12 @@ config MACH_SUN7I config MACH_SUN8I bool "Allwinner A23 (sun8i) SoCs support" default ARCH_SUNXI - select ARCH_HAS_RESET_CONTROLLER select ARM_GIC select MFD_SUN6I_PRCM - select RESET_CONTROLLER config MACH_SUN9I bool "Allwinner (sun9i) SoCs support" default ARCH_SUNXI - select ARCH_HAS_RESET_CONTROLLER select ARM_GIC - select RESET_CONTROLLER endif -- cgit v1.2.1 From 54331db99a2a7b621a83865608b2c59913291517 Mon Sep 17 00:00:00 2001 From: Luciano Coelho Date: Thu, 5 Feb 2015 00:39:22 +0200 Subject: iwlwifi: mvm: call ieee80211_scan_completed() even if scan abort fails A scan abort command failure is not that unusual, since we may try to send it after the scan has actually completed but before we received the completed notification from the firmware. The scan abort can also fail for other reasons, such as a timeout. In such cases, we should clear things up so the next scans will work again. To do so, don't return immediately in case of failures, but call ieee80211_scan_completed() and clear the scan_status flags. Signed-off-by: Luciano Coelho Reviewed-by: Johannes Berg Signed-off-by: Emmanuel Grumbach --- drivers/net/wireless/iwlwifi/mvm/scan.c | 13 ++++++------- 1 file changed, 6 insertions(+), 7 deletions(-) diff --git a/drivers/net/wireless/iwlwifi/mvm/scan.c b/drivers/net/wireless/iwlwifi/mvm/scan.c index 7e9aa3cb3254..c47c8051da77 100644 --- a/drivers/net/wireless/iwlwifi/mvm/scan.c +++ b/drivers/net/wireless/iwlwifi/mvm/scan.c @@ -1128,8 +1128,10 @@ int iwl_mvm_scan_offload_stop(struct iwl_mvm *mvm, bool notify) if (mvm->scan_status == IWL_MVM_SCAN_NONE) return 0; - if (iwl_mvm_is_radio_killed(mvm)) + if (iwl_mvm_is_radio_killed(mvm)) { + ret = 0; goto out; + } if (mvm->scan_status != IWL_MVM_SCAN_SCHED && (!(mvm->fw->ucode_capa.api[0] & IWL_UCODE_TLV_API_LMAC_SCAN) || @@ -1148,16 +1150,14 @@ int iwl_mvm_scan_offload_stop(struct iwl_mvm *mvm, bool notify) IWL_DEBUG_SCAN(mvm, "Send stop %sscan failed %d\n", sched ? "offloaded " : "", ret); iwl_remove_notification(&mvm->notif_wait, &wait_scan_done); - return ret; + goto out; } IWL_DEBUG_SCAN(mvm, "Successfully sent stop %sscan\n", sched ? "offloaded " : ""); ret = iwl_wait_notification(&mvm->notif_wait, &wait_scan_done, 1 * HZ); - if (ret) - return ret; - +out: /* * Clear the scan status so the next scan requests will succeed. This * also ensures the Rx handler doesn't do anything, as the scan was @@ -1167,7 +1167,6 @@ int iwl_mvm_scan_offload_stop(struct iwl_mvm *mvm, bool notify) if (mvm->scan_status == IWL_MVM_SCAN_OS) iwl_mvm_unref(mvm, IWL_MVM_REF_SCAN); -out: mvm->scan_status = IWL_MVM_SCAN_NONE; if (notify) { @@ -1177,7 +1176,7 @@ out: ieee80211_scan_completed(mvm->hw, true); } - return 0; + return ret; } static void iwl_mvm_unified_scan_fill_tx_cmd(struct iwl_mvm *mvm, -- cgit v1.2.1 From 57bff1485096c53f943e26b1c5847f2a9dfe84db Mon Sep 17 00:00:00 2001 From: Eyal Shapira Date: Mon, 2 Feb 2015 15:21:27 +0200 Subject: iwlwifi: mvm: rs: fix BT Coex check to look at the correct ant The check to avoid the shared antenna was passed the wrong antenna parameter. It should have checked whether the antenna of the next column we're considering is allowed and instead it was passed the current antenna. This could lead to a wrong choice of the next column in the rs algorithm and non optimal performance. Fixes: commit 219fb66b49fac64bb ("iwlwifi: mvm: rs - don't use the shared antenna when BT load is high") CC: [3.19] Signed-off-by: Eyal Shapira Reviewed-by: Johannes Berg Signed-off-by: Emmanuel Grumbach --- drivers/net/wireless/iwlwifi/mvm/rs.c | 21 ++++++++++++++------- 1 file changed, 14 insertions(+), 7 deletions(-) diff --git a/drivers/net/wireless/iwlwifi/mvm/rs.c b/drivers/net/wireless/iwlwifi/mvm/rs.c index 194bd1f939ca..efa9688a4cf1 100644 --- a/drivers/net/wireless/iwlwifi/mvm/rs.c +++ b/drivers/net/wireless/iwlwifi/mvm/rs.c @@ -134,9 +134,12 @@ enum rs_column_mode { #define MAX_NEXT_COLUMNS 7 #define MAX_COLUMN_CHECKS 3 +struct rs_tx_column; + typedef bool (*allow_column_func_t) (struct iwl_mvm *mvm, struct ieee80211_sta *sta, - struct iwl_scale_tbl_info *tbl); + struct iwl_scale_tbl_info *tbl, + const struct rs_tx_column *next_col); struct rs_tx_column { enum rs_column_mode mode; @@ -147,13 +150,15 @@ struct rs_tx_column { }; static bool rs_ant_allow(struct iwl_mvm *mvm, struct ieee80211_sta *sta, - struct iwl_scale_tbl_info *tbl) + struct iwl_scale_tbl_info *tbl, + const struct rs_tx_column *next_col) { - return iwl_mvm_bt_coex_is_ant_avail(mvm, tbl->rate.ant); + return iwl_mvm_bt_coex_is_ant_avail(mvm, next_col->ant); } static bool rs_mimo_allow(struct iwl_mvm *mvm, struct ieee80211_sta *sta, - struct iwl_scale_tbl_info *tbl) + struct iwl_scale_tbl_info *tbl, + const struct rs_tx_column *next_col) { if (!sta->ht_cap.ht_supported) return false; @@ -171,7 +176,8 @@ static bool rs_mimo_allow(struct iwl_mvm *mvm, struct ieee80211_sta *sta, } static bool rs_siso_allow(struct iwl_mvm *mvm, struct ieee80211_sta *sta, - struct iwl_scale_tbl_info *tbl) + struct iwl_scale_tbl_info *tbl, + const struct rs_tx_column *next_col) { if (!sta->ht_cap.ht_supported) return false; @@ -180,7 +186,8 @@ static bool rs_siso_allow(struct iwl_mvm *mvm, struct ieee80211_sta *sta, } static bool rs_sgi_allow(struct iwl_mvm *mvm, struct ieee80211_sta *sta, - struct iwl_scale_tbl_info *tbl) + struct iwl_scale_tbl_info *tbl, + const struct rs_tx_column *next_col) { struct rs_rate *rate = &tbl->rate; struct ieee80211_sta_ht_cap *ht_cap = &sta->ht_cap; @@ -1590,7 +1597,7 @@ static enum rs_column rs_get_next_column(struct iwl_mvm *mvm, for (j = 0; j < MAX_COLUMN_CHECKS; j++) { allow_func = next_col->checks[j]; - if (allow_func && !allow_func(mvm, sta, tbl)) + if (allow_func && !allow_func(mvm, sta, tbl, next_col)) break; } -- cgit v1.2.1 From e7d3abab81bfb9cda80b7f5aa027fbf9e62a37fd Mon Sep 17 00:00:00 2001 From: Luciano Coelho Date: Fri, 6 Feb 2015 10:19:05 +0200 Subject: iwlwifi: mvm: don't try to stop scans that are not running anymore In certain conditions, mac80211 may ask us to stop a scan (scheduled or normal) that is not running anymore. This can also happen when we are doing a different type of scan, for instance, mac80211 can ask us to stop a scheduled scan when we are running a normal scan, due to some race conditions. In this case, we would stop the wrong type of scan and leave everything everything in a wrong state. To fix this, simply ignore scan stop requests for scans types that are not running. Signed-off-by: Luciano Coelho Reviewed-by: Johannes Berg Signed-off-by: Emmanuel Grumbach --- drivers/net/wireless/iwlwifi/mvm/mac80211.c | 33 +++++++++++++++++++++++++++-- 1 file changed, 31 insertions(+), 2 deletions(-) diff --git a/drivers/net/wireless/iwlwifi/mvm/mac80211.c b/drivers/net/wireless/iwlwifi/mvm/mac80211.c index 1ff7ec08532d..35feebfd048f 100644 --- a/drivers/net/wireless/iwlwifi/mvm/mac80211.c +++ b/drivers/net/wireless/iwlwifi/mvm/mac80211.c @@ -2215,7 +2215,19 @@ static void iwl_mvm_mac_cancel_hw_scan(struct ieee80211_hw *hw, mutex_lock(&mvm->mutex); - iwl_mvm_cancel_scan(mvm); + /* Due to a race condition, it's possible that mac80211 asks + * us to stop a hw_scan when it's already stopped. This can + * happen, for instance, if we stopped the scan ourselves, + * called ieee80211_scan_completed() and the userspace called + * cancel scan scan before ieee80211_scan_work() could run. + * To handle that, simply return if the scan is not running. + */ + /* FIXME: for now, we ignore this race for UMAC scans, since + * they don't set the scan_status. + */ + if ((mvm->scan_status == IWL_MVM_SCAN_OS) || + (mvm->fw->ucode_capa.capa[0] & IWL_UCODE_TLV_CAPA_UMAC_SCAN)) + iwl_mvm_cancel_scan(mvm); mutex_unlock(&mvm->mutex); } @@ -2559,12 +2571,29 @@ static int iwl_mvm_mac_sched_scan_stop(struct ieee80211_hw *hw, int ret; mutex_lock(&mvm->mutex); + + /* Due to a race condition, it's possible that mac80211 asks + * us to stop a sched_scan when it's already stopped. This + * can happen, for instance, if we stopped the scan ourselves, + * called ieee80211_sched_scan_stopped() and the userspace called + * stop sched scan scan before ieee80211_sched_scan_stopped_work() + * could run. To handle this, simply return if the scan is + * not running. + */ + /* FIXME: for now, we ignore this race for UMAC scans, since + * they don't set the scan_status. + */ + if (mvm->scan_status != IWL_MVM_SCAN_SCHED && + !(mvm->fw->ucode_capa.capa[0] & IWL_UCODE_TLV_CAPA_UMAC_SCAN)) { + mutex_unlock(&mvm->mutex); + return 0; + } + ret = iwl_mvm_scan_offload_stop(mvm, false); mutex_unlock(&mvm->mutex); iwl_mvm_wait_for_async_handlers(mvm); return ret; - } static int iwl_mvm_mac_set_key(struct ieee80211_hw *hw, -- cgit v1.2.1 From 833d9b9785b3eedfaf2c869a6a63deba88058599 Mon Sep 17 00:00:00 2001 From: Andrei Otcheretianski Date: Sun, 15 Feb 2015 18:33:23 +0200 Subject: iwlwifi: mvm: Fix ROC removal iwl_mvm_stop_roc removes TE only if running flag is set. This is not correct since this flag is only set when the TE is started. This resulted in a TE not being removed, when mac80211 believes that there are no active ROCs. Fixes: bf5da87f60a9 ("iwlwifi: mvm: add remove flow for AUX ROC time events") Signed-off-by: Andrei Otcheretianski Reviewed-by: Matti Gottlieb Signed-off-by: Emmanuel Grumbach --- drivers/net/wireless/iwlwifi/mvm/time-event.c | 9 +++------ 1 file changed, 3 insertions(+), 6 deletions(-) diff --git a/drivers/net/wireless/iwlwifi/mvm/time-event.c b/drivers/net/wireless/iwlwifi/mvm/time-event.c index 54fafbf9a711..f8d6f306dd76 100644 --- a/drivers/net/wireless/iwlwifi/mvm/time-event.c +++ b/drivers/net/wireless/iwlwifi/mvm/time-event.c @@ -750,8 +750,7 @@ void iwl_mvm_stop_roc(struct iwl_mvm *mvm) * request */ list_for_each_entry(te_data, &mvm->time_event_list, list) { - if (te_data->vif->type == NL80211_IFTYPE_P2P_DEVICE && - te_data->running) { + if (te_data->vif->type == NL80211_IFTYPE_P2P_DEVICE) { mvmvif = iwl_mvm_vif_from_mac80211(te_data->vif); is_p2p = true; goto remove_te; @@ -766,10 +765,8 @@ void iwl_mvm_stop_roc(struct iwl_mvm *mvm) * request */ list_for_each_entry(te_data, &mvm->aux_roc_te_list, list) { - if (te_data->running) { - mvmvif = iwl_mvm_vif_from_mac80211(te_data->vif); - goto remove_te; - } + mvmvif = iwl_mvm_vif_from_mac80211(te_data->vif); + goto remove_te; } remove_te: -- cgit v1.2.1 From c9faccc9d2febc106994fc605c3957214973cc0f Mon Sep 17 00:00:00 2001 From: Eyal Shapira Date: Mon, 23 Feb 2015 10:02:23 +0200 Subject: iwlwifi: mvm: disable beamformer unless FW supports it Current FW is declaring support for BFER in ucode_capa.capa but it doesn't really support it unless the new LQ_SS_PARAMS API is supported as well. Avoid publishing BFER in our VHT caps if FW doesn't support. Signed-off-by: Eyal Shapira Signed-off-by: Emmanuel Grumbach --- drivers/net/wireless/iwlwifi/mvm/mac80211.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/net/wireless/iwlwifi/mvm/mac80211.c b/drivers/net/wireless/iwlwifi/mvm/mac80211.c index 35feebfd048f..09654e73a533 100644 --- a/drivers/net/wireless/iwlwifi/mvm/mac80211.c +++ b/drivers/net/wireless/iwlwifi/mvm/mac80211.c @@ -405,7 +405,10 @@ int iwl_mvm_mac_setup_register(struct iwl_mvm *mvm) hw->wiphy->bands[IEEE80211_BAND_5GHZ] = &mvm->nvm_data->bands[IEEE80211_BAND_5GHZ]; - if (mvm->fw->ucode_capa.capa[0] & IWL_UCODE_TLV_CAPA_BEAMFORMER) + if ((mvm->fw->ucode_capa.capa[0] & + IWL_UCODE_TLV_CAPA_BEAMFORMER) && + (mvm->fw->ucode_capa.api[0] & + IWL_UCODE_TLV_API_LQ_SS_PARAMS)) hw->wiphy->bands[IEEE80211_BAND_5GHZ]->vht_cap.cap |= IEEE80211_VHT_CAP_SU_BEAMFORMER_CAPABLE; } -- cgit v1.2.1 From b388e6a7a6ba988998ddd83919ae8d3debf1a13d Mon Sep 17 00:00:00 2001 From: Brian Norris Date: Mon, 23 Feb 2015 13:07:22 -0800 Subject: UBI: fix missing brace control flow commit 0e707ae79ba3 ("UBI: do propagate positive error codes up") seems to have produced an unintended change in the control flow here. Completely untested, but it looks obvious. Caught by Coverity, which didn't like the indentation. CID 1271184. Signed-off-by: Brian Norris Cc: Dan Carpenter Signed-off-by: Richard Weinberger --- drivers/mtd/ubi/eba.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/mtd/ubi/eba.c b/drivers/mtd/ubi/eba.c index da4c79259f67..16e34b37d134 100644 --- a/drivers/mtd/ubi/eba.c +++ b/drivers/mtd/ubi/eba.c @@ -425,9 +425,10 @@ retry: ubi_warn(ubi, "corrupted VID header at PEB %d, LEB %d:%d", pnum, vol_id, lnum); err = -EBADMSG; - } else + } else { err = -EINVAL; ubi_ro_mode(ubi); + } } goto out_free; } else if (err == UBI_IO_BITFLIPS) -- cgit v1.2.1 From 5c0c75d33d45a8dc7a2af815834812d41f5361e8 Mon Sep 17 00:00:00 2001 From: Masanari Iida Date: Sat, 31 Jan 2015 14:49:38 +0900 Subject: ARM: pxa: Fix typo in zeus.c This patch fix a typo in struct platform_device can_regulator_device. Signed-off-by: Masanari Iida Acked-by: Daniel Mack Signed-off-by: Robert Jarzmik --- arch/arm/mach-pxa/zeus.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm/mach-pxa/zeus.c b/arch/arm/mach-pxa/zeus.c index 205f9bf3821e..ac2ae5c71ab4 100644 --- a/arch/arm/mach-pxa/zeus.c +++ b/arch/arm/mach-pxa/zeus.c @@ -412,7 +412,7 @@ static struct fixed_voltage_config can_regulator_pdata = { }; static struct platform_device can_regulator_device = { - .name = "reg-fixed-volage", + .name = "reg-fixed-voltage", .id = 0, .dev = { .platform_data = &can_regulator_pdata, -- cgit v1.2.1 From d6cf30ca716b347587b35923eda400ad2d9e8832 Mon Sep 17 00:00:00 2001 From: Robert Jarzmik Date: Sat, 14 Feb 2015 22:41:56 +0100 Subject: ARM: pxa: fix pxa interrupts handling in DT The commit "ARM: pxa: arbitrarily set first interrupt number" changed the first pxa interrupt to 16. As a consequence, device-tree builds got broken, because : - pxa_mask_irq() and pxa_unmask_irq() are using IRQ_BIT() - IRQ_BIT(x) calculates the interrupts as : x - PXA_IRQ(0) Before the commit, the first interrupt shift, PXA_IRQ(0) was 0, therefore IRQ_BIT(x) was x. After the change, it is necessary that the same shift of 16 is applied between the virtual interrupt number and the hardware irq number. This situation comes from the common irq_chip shared between legacy platform builds and device-tree builds. Fix the broken interrupts in DT case by adding this shift in the DT case too. As a consequence of the IRQ_BIT() is removed alltogether from interrupts handling, even in the platform data types of platforms : - a legacy irq domain is used - the irq_chip handles hardware interrupts - the virtual to hardware interrupt conversion is fully handled by irq domain mechanics Signed-off-by: Robert Jarzmik --- arch/arm/Kconfig | 1 + arch/arm/mach-pxa/irq.c | 111 +++++++++++++++++++++--------------------------- 2 files changed, 49 insertions(+), 63 deletions(-) diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index 9f1f09a2bc9b..cf4c0c99aa25 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -619,6 +619,7 @@ config ARCH_PXA select GENERIC_CLOCKEVENTS select GPIO_PXA select HAVE_IDE + select IRQ_DOMAIN select MULTI_IRQ_HANDLER select PLAT_PXA select SPARSE_IRQ diff --git a/arch/arm/mach-pxa/irq.c b/arch/arm/mach-pxa/irq.c index 0eecd83c624e..89a7c06570d3 100644 --- a/arch/arm/mach-pxa/irq.c +++ b/arch/arm/mach-pxa/irq.c @@ -11,6 +11,7 @@ * it under the terms of the GNU General Public License version 2 as * published by the Free Software Foundation. */ +#include #include #include #include @@ -40,7 +41,6 @@ #define ICHP_VAL_IRQ (1 << 31) #define ICHP_IRQ(i) (((i) >> 16) & 0x7fff) #define IPR_VALID (1 << 31) -#define IRQ_BIT(n) (((n) - PXA_IRQ(0)) & 0x1f) #define MAX_INTERNAL_IRQS 128 @@ -51,6 +51,7 @@ static void __iomem *pxa_irq_base; static int pxa_internal_irq_nr; static bool cpu_has_ipr; +static struct irq_domain *pxa_irq_domain; static inline void __iomem *irq_base(int i) { @@ -66,18 +67,20 @@ static inline void __iomem *irq_base(int i) void pxa_mask_irq(struct irq_data *d) { void __iomem *base = irq_data_get_irq_chip_data(d); + irq_hw_number_t irq = irqd_to_hwirq(d); uint32_t icmr = __raw_readl(base + ICMR); - icmr &= ~(1 << IRQ_BIT(d->irq)); + icmr &= ~BIT(irq & 0x1f); __raw_writel(icmr, base + ICMR); } void pxa_unmask_irq(struct irq_data *d) { void __iomem *base = irq_data_get_irq_chip_data(d); + irq_hw_number_t irq = irqd_to_hwirq(d); uint32_t icmr = __raw_readl(base + ICMR); - icmr |= 1 << IRQ_BIT(d->irq); + icmr |= BIT(irq & 0x1f); __raw_writel(icmr, base + ICMR); } @@ -118,40 +121,63 @@ asmlinkage void __exception_irq_entry ichp_handle_irq(struct pt_regs *regs) } while (1); } -void __init pxa_init_irq(int irq_nr, int (*fn)(struct irq_data *, unsigned int)) +static int pxa_irq_map(struct irq_domain *h, unsigned int virq, + irq_hw_number_t hw) { - int irq, i, n; + void __iomem *base = irq_base(hw / 32); - BUG_ON(irq_nr > MAX_INTERNAL_IRQS); + /* initialize interrupt priority */ + if (cpu_has_ipr) + __raw_writel(hw | IPR_VALID, pxa_irq_base + IPR(hw)); + + irq_set_chip_and_handler(virq, &pxa_internal_irq_chip, + handle_level_irq); + irq_set_chip_data(virq, base); + set_irq_flags(virq, IRQF_VALID); + + return 0; +} + +static struct irq_domain_ops pxa_irq_ops = { + .map = pxa_irq_map, + .xlate = irq_domain_xlate_onecell, +}; + +static __init void +pxa_init_irq_common(struct device_node *node, int irq_nr, + int (*fn)(struct irq_data *, unsigned int)) +{ + int n; pxa_internal_irq_nr = irq_nr; - cpu_has_ipr = !cpu_is_pxa25x(); - pxa_irq_base = io_p2v(0x40d00000); + pxa_irq_domain = irq_domain_add_legacy(node, irq_nr, + PXA_IRQ(0), 0, + &pxa_irq_ops, NULL); + if (!pxa_irq_domain) + panic("Unable to add PXA IRQ domain\n"); + irq_set_default_host(pxa_irq_domain); for (n = 0; n < irq_nr; n += 32) { void __iomem *base = irq_base(n >> 5); __raw_writel(0, base + ICMR); /* disable all IRQs */ __raw_writel(0, base + ICLR); /* all IRQs are IRQ, not FIQ */ - for (i = n; (i < (n + 32)) && (i < irq_nr); i++) { - /* initialize interrupt priority */ - if (cpu_has_ipr) - __raw_writel(i | IPR_VALID, pxa_irq_base + IPR(i)); - - irq = PXA_IRQ(i); - irq_set_chip_and_handler(irq, &pxa_internal_irq_chip, - handle_level_irq); - irq_set_chip_data(irq, base); - set_irq_flags(irq, IRQF_VALID); - } } - /* only unmasked interrupts kick us out of idle */ __raw_writel(1, irq_base(0) + ICCR); pxa_internal_irq_chip.irq_set_wake = fn; } +void __init pxa_init_irq(int irq_nr, int (*fn)(struct irq_data *, unsigned int)) +{ + BUG_ON(irq_nr > MAX_INTERNAL_IRQS); + + pxa_irq_base = io_p2v(0x40d00000); + cpu_has_ipr = !cpu_is_pxa25x(); + pxa_init_irq_common(NULL, irq_nr, fn); +} + #ifdef CONFIG_PM static unsigned long saved_icmr[MAX_INTERNAL_IRQS/32]; static unsigned long saved_ipr[MAX_INTERNAL_IRQS]; @@ -203,30 +229,6 @@ struct syscore_ops pxa_irq_syscore_ops = { }; #ifdef CONFIG_OF -static struct irq_domain *pxa_irq_domain; - -static int pxa_irq_map(struct irq_domain *h, unsigned int virq, - irq_hw_number_t hw) -{ - void __iomem *base = irq_base(hw / 32); - - /* initialize interrupt priority */ - if (cpu_has_ipr) - __raw_writel(hw | IPR_VALID, pxa_irq_base + IPR(hw)); - - irq_set_chip_and_handler(hw, &pxa_internal_irq_chip, - handle_level_irq); - irq_set_chip_data(hw, base); - set_irq_flags(hw, IRQF_VALID); - - return 0; -} - -static struct irq_domain_ops pxa_irq_ops = { - .map = pxa_irq_map, - .xlate = irq_domain_xlate_onecell, -}; - static const struct of_device_id intc_ids[] __initconst = { { .compatible = "marvell,pxa-intc", }, {} @@ -236,7 +238,7 @@ void __init pxa_dt_irq_init(int (*fn)(struct irq_data *, unsigned int)) { struct device_node *node; struct resource res; - int n, ret; + int ret; node = of_find_matching_node(NULL, intc_ids); if (!node) { @@ -267,23 +269,6 @@ void __init pxa_dt_irq_init(int (*fn)(struct irq_data *, unsigned int)) return; } - pxa_irq_domain = irq_domain_add_legacy(node, pxa_internal_irq_nr, 0, 0, - &pxa_irq_ops, NULL); - if (!pxa_irq_domain) - panic("Unable to add PXA IRQ domain\n"); - - irq_set_default_host(pxa_irq_domain); - - for (n = 0; n < pxa_internal_irq_nr; n += 32) { - void __iomem *base = irq_base(n >> 5); - - __raw_writel(0, base + ICMR); /* disable all IRQs */ - __raw_writel(0, base + ICLR); /* all IRQs are IRQ, not FIQ */ - } - - /* only unmasked interrupts kick us out of idle */ - __raw_writel(1, irq_base(0) + ICCR); - - pxa_internal_irq_chip.irq_set_wake = fn; + pxa_init_irq_common(node, pxa_internal_irq_nr, fn); } #endif /* CONFIG_OF */ -- cgit v1.2.1 From 540623caa6c769d9d19e6044949f5fa2fe1a33a6 Mon Sep 17 00:00:00 2001 From: Emmanuel Grumbach Date: Mon, 23 Feb 2015 02:40:07 +0200 Subject: iwlwifi: fix max_ht_ampdu_exponent for older devices The commit below didn't update the max_ht_ampdu_exponent for the devices listed in iwl-[1-6]000.c which, in result, became 0 instead of 8K. This reduced the size of the Rx AMPDU from 64K to 8K which had an impact in the Rx throughput. One user reported that because of this, his downstream throughput droppped by a half. CC: [3.19] Fixes: c064ddf318aa ("iwlwifi: change max HT and VHT A-MPDU exponent") Reported-and-tested-by: Valentin Manea Signed-off-by: Emmanuel Grumbach --- drivers/net/wireless/iwlwifi/iwl-1000.c | 6 ++++-- drivers/net/wireless/iwlwifi/iwl-2000.c | 13 +++++++++---- drivers/net/wireless/iwlwifi/iwl-5000.c | 6 ++++-- drivers/net/wireless/iwlwifi/iwl-6000.c | 18 ++++++++++++------ 4 files changed, 29 insertions(+), 14 deletions(-) diff --git a/drivers/net/wireless/iwlwifi/iwl-1000.c b/drivers/net/wireless/iwlwifi/iwl-1000.c index c3817fae16c0..06f6cc08f451 100644 --- a/drivers/net/wireless/iwlwifi/iwl-1000.c +++ b/drivers/net/wireless/iwlwifi/iwl-1000.c @@ -95,7 +95,8 @@ static const struct iwl_eeprom_params iwl1000_eeprom_params = { .nvm_calib_ver = EEPROM_1000_TX_POWER_VERSION, \ .base_params = &iwl1000_base_params, \ .eeprom_params = &iwl1000_eeprom_params, \ - .led_mode = IWL_LED_BLINK + .led_mode = IWL_LED_BLINK, \ + .max_ht_ampdu_exponent = IEEE80211_HT_MAX_AMPDU_64K const struct iwl_cfg iwl1000_bgn_cfg = { .name = "Intel(R) Centrino(R) Wireless-N 1000 BGN", @@ -121,7 +122,8 @@ const struct iwl_cfg iwl1000_bg_cfg = { .base_params = &iwl1000_base_params, \ .eeprom_params = &iwl1000_eeprom_params, \ .led_mode = IWL_LED_RF_STATE, \ - .rx_with_siso_diversity = true + .rx_with_siso_diversity = true, \ + .max_ht_ampdu_exponent = IEEE80211_HT_MAX_AMPDU_64K const struct iwl_cfg iwl100_bgn_cfg = { .name = "Intel(R) Centrino(R) Wireless-N 100 BGN", diff --git a/drivers/net/wireless/iwlwifi/iwl-2000.c b/drivers/net/wireless/iwlwifi/iwl-2000.c index 21e5d0843a62..890b95f497d6 100644 --- a/drivers/net/wireless/iwlwifi/iwl-2000.c +++ b/drivers/net/wireless/iwlwifi/iwl-2000.c @@ -123,7 +123,9 @@ static const struct iwl_eeprom_params iwl20x0_eeprom_params = { .nvm_calib_ver = EEPROM_2000_TX_POWER_VERSION, \ .base_params = &iwl2000_base_params, \ .eeprom_params = &iwl20x0_eeprom_params, \ - .led_mode = IWL_LED_RF_STATE + .led_mode = IWL_LED_RF_STATE, \ + .max_ht_ampdu_exponent = IEEE80211_HT_MAX_AMPDU_64K + const struct iwl_cfg iwl2000_2bgn_cfg = { .name = "Intel(R) Centrino(R) Wireless-N 2200 BGN", @@ -149,7 +151,8 @@ const struct iwl_cfg iwl2000_2bgn_d_cfg = { .nvm_calib_ver = EEPROM_2000_TX_POWER_VERSION, \ .base_params = &iwl2030_base_params, \ .eeprom_params = &iwl20x0_eeprom_params, \ - .led_mode = IWL_LED_RF_STATE + .led_mode = IWL_LED_RF_STATE, \ + .max_ht_ampdu_exponent = IEEE80211_HT_MAX_AMPDU_64K const struct iwl_cfg iwl2030_2bgn_cfg = { .name = "Intel(R) Centrino(R) Wireless-N 2230 BGN", @@ -170,7 +173,8 @@ const struct iwl_cfg iwl2030_2bgn_cfg = { .base_params = &iwl2000_base_params, \ .eeprom_params = &iwl20x0_eeprom_params, \ .led_mode = IWL_LED_RF_STATE, \ - .rx_with_siso_diversity = true + .rx_with_siso_diversity = true, \ + .max_ht_ampdu_exponent = IEEE80211_HT_MAX_AMPDU_64K const struct iwl_cfg iwl105_bgn_cfg = { .name = "Intel(R) Centrino(R) Wireless-N 105 BGN", @@ -197,7 +201,8 @@ const struct iwl_cfg iwl105_bgn_d_cfg = { .base_params = &iwl2030_base_params, \ .eeprom_params = &iwl20x0_eeprom_params, \ .led_mode = IWL_LED_RF_STATE, \ - .rx_with_siso_diversity = true + .rx_with_siso_diversity = true, \ + .max_ht_ampdu_exponent = IEEE80211_HT_MAX_AMPDU_64K const struct iwl_cfg iwl135_bgn_cfg = { .name = "Intel(R) Centrino(R) Wireless-N 135 BGN", diff --git a/drivers/net/wireless/iwlwifi/iwl-5000.c b/drivers/net/wireless/iwlwifi/iwl-5000.c index 332bbede39e5..724194e23414 100644 --- a/drivers/net/wireless/iwlwifi/iwl-5000.c +++ b/drivers/net/wireless/iwlwifi/iwl-5000.c @@ -93,7 +93,8 @@ static const struct iwl_eeprom_params iwl5000_eeprom_params = { .nvm_calib_ver = EEPROM_5000_TX_POWER_VERSION, \ .base_params = &iwl5000_base_params, \ .eeprom_params = &iwl5000_eeprom_params, \ - .led_mode = IWL_LED_BLINK + .led_mode = IWL_LED_BLINK, \ + .max_ht_ampdu_exponent = IEEE80211_HT_MAX_AMPDU_64K const struct iwl_cfg iwl5300_agn_cfg = { .name = "Intel(R) Ultimate N WiFi Link 5300 AGN", @@ -158,7 +159,8 @@ const struct iwl_cfg iwl5350_agn_cfg = { .base_params = &iwl5000_base_params, \ .eeprom_params = &iwl5000_eeprom_params, \ .led_mode = IWL_LED_BLINK, \ - .internal_wimax_coex = true + .internal_wimax_coex = true, \ + .max_ht_ampdu_exponent = IEEE80211_HT_MAX_AMPDU_64K const struct iwl_cfg iwl5150_agn_cfg = { .name = "Intel(R) WiMAX/WiFi Link 5150 AGN", diff --git a/drivers/net/wireless/iwlwifi/iwl-6000.c b/drivers/net/wireless/iwlwifi/iwl-6000.c index 8f2c3c8c6b84..21b2630763dc 100644 --- a/drivers/net/wireless/iwlwifi/iwl-6000.c +++ b/drivers/net/wireless/iwlwifi/iwl-6000.c @@ -145,7 +145,8 @@ static const struct iwl_eeprom_params iwl6000_eeprom_params = { .nvm_calib_ver = EEPROM_6005_TX_POWER_VERSION, \ .base_params = &iwl6000_g2_base_params, \ .eeprom_params = &iwl6000_eeprom_params, \ - .led_mode = IWL_LED_RF_STATE + .led_mode = IWL_LED_RF_STATE, \ + .max_ht_ampdu_exponent = IEEE80211_HT_MAX_AMPDU_64K const struct iwl_cfg iwl6005_2agn_cfg = { .name = "Intel(R) Centrino(R) Advanced-N 6205 AGN", @@ -199,7 +200,8 @@ const struct iwl_cfg iwl6005_2agn_mow2_cfg = { .nvm_calib_ver = EEPROM_6030_TX_POWER_VERSION, \ .base_params = &iwl6000_g2_base_params, \ .eeprom_params = &iwl6000_eeprom_params, \ - .led_mode = IWL_LED_RF_STATE + .led_mode = IWL_LED_RF_STATE, \ + .max_ht_ampdu_exponent = IEEE80211_HT_MAX_AMPDU_64K const struct iwl_cfg iwl6030_2agn_cfg = { .name = "Intel(R) Centrino(R) Advanced-N 6230 AGN", @@ -235,7 +237,8 @@ const struct iwl_cfg iwl6030_2bg_cfg = { .nvm_calib_ver = EEPROM_6030_TX_POWER_VERSION, \ .base_params = &iwl6000_g2_base_params, \ .eeprom_params = &iwl6000_eeprom_params, \ - .led_mode = IWL_LED_RF_STATE + .led_mode = IWL_LED_RF_STATE, \ + .max_ht_ampdu_exponent = IEEE80211_HT_MAX_AMPDU_64K const struct iwl_cfg iwl6035_2agn_cfg = { .name = "Intel(R) Centrino(R) Advanced-N 6235 AGN", @@ -290,7 +293,8 @@ const struct iwl_cfg iwl130_bg_cfg = { .nvm_calib_ver = EEPROM_6000_TX_POWER_VERSION, \ .base_params = &iwl6000_base_params, \ .eeprom_params = &iwl6000_eeprom_params, \ - .led_mode = IWL_LED_BLINK + .led_mode = IWL_LED_BLINK, \ + .max_ht_ampdu_exponent = IEEE80211_HT_MAX_AMPDU_64K const struct iwl_cfg iwl6000i_2agn_cfg = { .name = "Intel(R) Centrino(R) Advanced-N 6200 AGN", @@ -322,7 +326,8 @@ const struct iwl_cfg iwl6000i_2bg_cfg = { .base_params = &iwl6050_base_params, \ .eeprom_params = &iwl6000_eeprom_params, \ .led_mode = IWL_LED_BLINK, \ - .internal_wimax_coex = true + .internal_wimax_coex = true, \ + .max_ht_ampdu_exponent = IEEE80211_HT_MAX_AMPDU_64K const struct iwl_cfg iwl6050_2agn_cfg = { .name = "Intel(R) Centrino(R) Advanced-N + WiMAX 6250 AGN", @@ -347,7 +352,8 @@ const struct iwl_cfg iwl6050_2abg_cfg = { .base_params = &iwl6050_base_params, \ .eeprom_params = &iwl6000_eeprom_params, \ .led_mode = IWL_LED_BLINK, \ - .internal_wimax_coex = true + .internal_wimax_coex = true, \ + .max_ht_ampdu_exponent = IEEE80211_HT_MAX_AMPDU_64K const struct iwl_cfg iwl6150_bgn_cfg = { .name = "Intel(R) Centrino(R) Wireless-N + WiMAX 6150 BGN", -- cgit v1.2.1 From 0d2783626a53d4c922f82d51fa675cb5d13f0d36 Mon Sep 17 00:00:00 2001 From: Miklos Szeredi Date: Thu, 26 Feb 2015 11:45:47 +0100 Subject: fuse: notify: don't move pages fuse_try_move_page() is not prepared for replacing pages that have already been read. Reported-by: Al Viro Signed-off-by: Miklos Szeredi Cc: stable@vger.kernel.org --- fs/fuse/dev.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index ed19a7d622fa..b0d7e13fae3d 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -1797,6 +1797,9 @@ copy_finish: static int fuse_notify(struct fuse_conn *fc, enum fuse_notify_code code, unsigned int size, struct fuse_copy_state *cs) { + /* Don't try to move pages (yet) */ + cs->move_pages = 0; + switch (code) { case FUSE_NOTIFY_POLL: return fuse_notify_poll(fc, size, cs); -- cgit v1.2.1 From aa991b3b267e24f578bac7b09cc57579b660304b Mon Sep 17 00:00:00 2001 From: Miklos Szeredi Date: Thu, 26 Feb 2015 11:45:47 +0100 Subject: fuse: set stolen page uptodate Regular pipe buffers' ->steal method (generic_pipe_buf_steal()) doesn't set PG_uptodate. Don't warn on this condition, just set the uptodate flag. Signed-off-by: Miklos Szeredi Cc: stable@vger.kernel.org --- fs/fuse/dev.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index b0d7e13fae3d..71c4619af333 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -890,8 +890,8 @@ static int fuse_try_move_page(struct fuse_copy_state *cs, struct page **pagep) newpage = buf->page; - if (WARN_ON(!PageUptodate(newpage))) - return -EIO; + if (!PageUptodate(newpage)) + SetPageUptodate(newpage); ClearPageMappedToDisk(newpage); -- cgit v1.2.1 From 5f027a3bf184d1d36e68745f7cd3718a8b879cc0 Mon Sep 17 00:00:00 2001 From: Joe Thornber Date: Fri, 27 Feb 2015 14:09:12 +0000 Subject: dm thin: fix to consistently zero-fill reads to unprovisioned blocks It was always intended that a read to an unprovisioned block will return zeroes regardless of whether the pool is in read-only or read-write mode. thin_bio_map() was inconsistent with its handling of such reads when the pool is in read-only mode, it now properly zero-fills the bios it returns in response to unprovisioned block reads. Eliminate thin_bio_map()'s special read-only mode handling of -ENODATA and just allow the IO to be deferred to the worker which will result in pool->process_bio() handling the IO (which already properly zero-fills reads to unprovisioned blocks). Reported-by: Eric Sandeen Signed-off-by: Joe Thornber Signed-off-by: Mike Snitzer Cc: stable@vger.kernel.org --- drivers/md/dm-thin.c | 11 ----------- 1 file changed, 11 deletions(-) diff --git a/drivers/md/dm-thin.c b/drivers/md/dm-thin.c index 654773cb1eee..921aafd12aee 100644 --- a/drivers/md/dm-thin.c +++ b/drivers/md/dm-thin.c @@ -2358,17 +2358,6 @@ static int thin_bio_map(struct dm_target *ti, struct bio *bio) return DM_MAPIO_REMAPPED; case -ENODATA: - if (get_pool_mode(tc->pool) == PM_READ_ONLY) { - /* - * This block isn't provisioned, and we have no way - * of doing so. - */ - handle_unserviceable_bio(tc->pool, bio); - cell_defer_no_holder(tc, virt_cell); - return DM_MAPIO_SUBMITTED; - } - /* fall through */ - case -EWOULDBLOCK: thin_defer_cell(tc, virt_cell); return DM_MAPIO_SUBMITTED; -- cgit v1.2.1 From ab7c7bb6f4ab95dbca96fcfc4463cd69843e3e24 Mon Sep 17 00:00:00 2001 From: Mikulas Patocka Date: Fri, 27 Feb 2015 14:04:27 -0500 Subject: dm: hold suspend_lock while suspending device during device deletion __dm_destroy() must take the suspend_lock so that its presuspend and postsuspend calls do not race with an internal suspend. Signed-off-by: Mikulas Patocka Signed-off-by: Mike Snitzer Cc: stable@vger.kernel.org --- drivers/md/dm.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/md/dm.c b/drivers/md/dm.c index 73f28802dc7a..e79d5ccfda64 100644 --- a/drivers/md/dm.c +++ b/drivers/md/dm.c @@ -2638,10 +2638,16 @@ static void __dm_destroy(struct mapped_device *md, bool wait) if (dm_request_based(md)) flush_kthread_worker(&md->kworker); + /* + * Take suspend_lock so that presuspend and postsuspend methods + * do not race with internal suspend. + */ + mutex_lock(&md->suspend_lock); if (!dm_suspended_md(md)) { dm_table_presuspend_targets(map); dm_table_postsuspend_targets(map); } + mutex_unlock(&md->suspend_lock); /* dm_put_live_table must be before msleep, otherwise deadlock is possible */ dm_put_live_table(md, srcu_idx); -- cgit v1.2.1 From b735fede8d957d9d255e9c5cf3964cfa59799637 Mon Sep 17 00:00:00 2001 From: Mikulas Patocka Date: Thu, 26 Feb 2015 11:40:35 -0500 Subject: dm snapshot: suspend origin when doing exception handover In the function snapshot_resume we perform exception store handover. If there is another active snapshot target, the exception store is moved from this target to the target that is being resumed. The problem is that if there is some pending exception, it will point to an incorrect exception store after that handover, causing a crash due to dm-snap-persistent.c:get_exception()'s BUG_ON. This bug can be triggered by repeatedly changing snapshot permissions with "lvchange -p r" and "lvchange -p rw" while there are writes on the associated origin device. To fix this bug, we must suspend the origin device when doing the exception store handover to make sure that there are no pending exceptions: - introduce _origin_hash that keeps track of dm_origin structures. - introduce functions __lookup_dm_origin, __insert_dm_origin and __remove_dm_origin that manipulate the origin hash. - modify snapshot_resume so that it calls dm_internal_suspend_fast() and dm_internal_resume_fast() on the origin device. NOTE to stable@ people: When backporting to kernels 3.12-3.18, use dm_internal_suspend and dm_internal_resume instead of dm_internal_suspend_fast and dm_internal_resume_fast. When backporting to kernels older than 3.12, you need to pick functions dm_internal_suspend and dm_internal_resume from the commit fd2ed4d252701d3bbed4cd3e3d267ad469bb832a. Signed-off-by: Mikulas Patocka Signed-off-by: Mike Snitzer Cc: stable@vger.kernel.org --- drivers/md/dm-snap.c | 93 +++++++++++++++++++++++++++++++++++++++++++++++----- drivers/md/dm.c | 2 ++ 2 files changed, 86 insertions(+), 9 deletions(-) diff --git a/drivers/md/dm-snap.c b/drivers/md/dm-snap.c index 8b204ae216ab..c2bf822bad6f 100644 --- a/drivers/md/dm-snap.c +++ b/drivers/md/dm-snap.c @@ -20,6 +20,8 @@ #include #include +#include "dm.h" + #include "dm-exception-store.h" #define DM_MSG_PREFIX "snapshots" @@ -290,6 +292,16 @@ struct origin { struct list_head snapshots; }; +/* + * This structure is allocated for each origin target + */ +struct dm_origin { + struct dm_dev *dev; + struct dm_target *ti; + unsigned split_boundary; + struct list_head hash_list; +}; + /* * Size of the hash table for origin volumes. If we make this * the size of the minors list then it should be nearly perfect @@ -297,6 +309,7 @@ struct origin { #define ORIGIN_HASH_SIZE 256 #define ORIGIN_MASK 0xFF static struct list_head *_origins; +static struct list_head *_dm_origins; static struct rw_semaphore _origins_lock; static DECLARE_WAIT_QUEUE_HEAD(_pending_exceptions_done); @@ -310,12 +323,22 @@ static int init_origin_hash(void) _origins = kmalloc(ORIGIN_HASH_SIZE * sizeof(struct list_head), GFP_KERNEL); if (!_origins) { - DMERR("unable to allocate memory"); + DMERR("unable to allocate memory for _origins"); return -ENOMEM; } - for (i = 0; i < ORIGIN_HASH_SIZE; i++) INIT_LIST_HEAD(_origins + i); + + _dm_origins = kmalloc(ORIGIN_HASH_SIZE * sizeof(struct list_head), + GFP_KERNEL); + if (!_dm_origins) { + DMERR("unable to allocate memory for _dm_origins"); + kfree(_origins); + return -ENOMEM; + } + for (i = 0; i < ORIGIN_HASH_SIZE; i++) + INIT_LIST_HEAD(_dm_origins + i); + init_rwsem(&_origins_lock); return 0; @@ -324,6 +347,7 @@ static int init_origin_hash(void) static void exit_origin_hash(void) { kfree(_origins); + kfree(_dm_origins); } static unsigned origin_hash(struct block_device *bdev) @@ -350,6 +374,30 @@ static void __insert_origin(struct origin *o) list_add_tail(&o->hash_list, sl); } +static struct dm_origin *__lookup_dm_origin(struct block_device *origin) +{ + struct list_head *ol; + struct dm_origin *o; + + ol = &_dm_origins[origin_hash(origin)]; + list_for_each_entry (o, ol, hash_list) + if (bdev_equal(o->dev->bdev, origin)) + return o; + + return NULL; +} + +static void __insert_dm_origin(struct dm_origin *o) +{ + struct list_head *sl = &_dm_origins[origin_hash(o->dev->bdev)]; + list_add_tail(&o->hash_list, sl); +} + +static void __remove_dm_origin(struct dm_origin *o) +{ + list_del(&o->hash_list); +} + /* * _origins_lock must be held when calling this function. * Returns number of snapshots registered using the supplied cow device, plus: @@ -1841,8 +1889,20 @@ static void snapshot_resume(struct dm_target *ti) { struct dm_snapshot *s = ti->private; struct dm_snapshot *snap_src = NULL, *snap_dest = NULL; + struct dm_origin *o; + struct mapped_device *origin_md = NULL; down_read(&_origins_lock); + + o = __lookup_dm_origin(s->origin->bdev); + if (o) + origin_md = dm_table_get_md(o->ti->table); + if (origin_md == dm_table_get_md(ti->table)) + origin_md = NULL; + + if (origin_md) + dm_internal_suspend_fast(origin_md); + (void) __find_snapshots_sharing_cow(s, &snap_src, &snap_dest, NULL); if (snap_src && snap_dest) { down_write(&snap_src->lock); @@ -1851,6 +1911,10 @@ static void snapshot_resume(struct dm_target *ti) up_write(&snap_dest->lock); up_write(&snap_src->lock); } + + if (origin_md) + dm_internal_resume_fast(origin_md); + up_read(&_origins_lock); /* Now we have correct chunk size, reregister */ @@ -2133,11 +2197,6 @@ static int origin_write_extent(struct dm_snapshot *merging_snap, * Origin: maps a linear range of a device, with hooks for snapshotting. */ -struct dm_origin { - struct dm_dev *dev; - unsigned split_boundary; -}; - /* * Construct an origin mapping: * The context for an origin is merely a 'struct dm_dev *' @@ -2166,6 +2225,7 @@ static int origin_ctr(struct dm_target *ti, unsigned int argc, char **argv) goto bad_open; } + o->ti = ti; ti->private = o; ti->num_flush_bios = 1; @@ -2180,6 +2240,7 @@ bad_alloc: static void origin_dtr(struct dm_target *ti) { struct dm_origin *o = ti->private; + dm_put_device(ti, o->dev); kfree(o); } @@ -2216,6 +2277,19 @@ static void origin_resume(struct dm_target *ti) struct dm_origin *o = ti->private; o->split_boundary = get_origin_minimum_chunksize(o->dev->bdev); + + down_write(&_origins_lock); + __insert_dm_origin(o); + up_write(&_origins_lock); +} + +static void origin_postsuspend(struct dm_target *ti) +{ + struct dm_origin *o = ti->private; + + down_write(&_origins_lock); + __remove_dm_origin(o); + up_write(&_origins_lock); } static void origin_status(struct dm_target *ti, status_type_t type, @@ -2258,12 +2332,13 @@ static int origin_iterate_devices(struct dm_target *ti, static struct target_type origin_target = { .name = "snapshot-origin", - .version = {1, 8, 1}, + .version = {1, 9, 0}, .module = THIS_MODULE, .ctr = origin_ctr, .dtr = origin_dtr, .map = origin_map, .resume = origin_resume, + .postsuspend = origin_postsuspend, .status = origin_status, .merge = origin_merge, .iterate_devices = origin_iterate_devices, @@ -2271,7 +2346,7 @@ static struct target_type origin_target = { static struct target_type snapshot_target = { .name = "snapshot", - .version = {1, 12, 0}, + .version = {1, 13, 0}, .module = THIS_MODULE, .ctr = snapshot_ctr, .dtr = snapshot_dtr, diff --git a/drivers/md/dm.c b/drivers/md/dm.c index e79d5ccfda64..6e2b2e97abe9 100644 --- a/drivers/md/dm.c +++ b/drivers/md/dm.c @@ -3121,6 +3121,7 @@ void dm_internal_suspend_fast(struct mapped_device *md) flush_workqueue(md->wq); dm_wait_for_completion(md, TASK_UNINTERRUPTIBLE); } +EXPORT_SYMBOL_GPL(dm_internal_suspend_fast); void dm_internal_resume_fast(struct mapped_device *md) { @@ -3132,6 +3133,7 @@ void dm_internal_resume_fast(struct mapped_device *md) done: mutex_unlock(&md->suspend_lock); } +EXPORT_SYMBOL_GPL(dm_internal_resume_fast); /*----------------------------------------------------------------- * Event notification. -- cgit v1.2.1 From 09ee96b21456883e108c3b00597bb37ec512151b Mon Sep 17 00:00:00 2001 From: Mikulas Patocka Date: Thu, 26 Feb 2015 11:41:28 -0500 Subject: dm snapshot: suspend merging snapshot when doing exception handover The "dm snapshot: suspend origin when doing exception handover" commit fixed a exception store handover bug associated with pending exceptions to the "snapshot-origin" target. However, a similar problem exists in snapshot merging. When snapshot merging is in progress, we use the target "snapshot-merge" instead of "snapshot-origin". Consequently, during exception store handover, we must find the snapshot-merge target and suspend its associated mapped_device. To avoid lockdep warnings, the target must be suspended and resumed without holding _origins_lock. Introduce a dm_hold() function that grabs a reference on a mapped_device, but unlike dm_get(), it doesn't crash if the device has the DMF_FREEING flag set, it returns an error in this case. In snapshot_resume() we grab the reference to the origin device using dm_hold() while holding _origins_lock (_origins_lock guarantees that the device won't disappear). Then we release _origins_lock, suspend the device and grab _origins_lock again. NOTE to stable@ people: When backporting to kernels 3.18 and older, use dm_internal_suspend and dm_internal_resume instead of dm_internal_suspend_fast and dm_internal_resume_fast. Signed-off-by: Mikulas Patocka Signed-off-by: Mike Snitzer Cc: stable@vger.kernel.org --- drivers/md/dm-snap.c | 35 +++++++++++++++++++++++++++++------ drivers/md/dm.c | 13 +++++++++++++ include/linux/device-mapper.h | 1 + 3 files changed, 43 insertions(+), 6 deletions(-) diff --git a/drivers/md/dm-snap.c b/drivers/md/dm-snap.c index c2bf822bad6f..f83a0f3fc365 100644 --- a/drivers/md/dm-snap.c +++ b/drivers/md/dm-snap.c @@ -1888,20 +1888,39 @@ static int snapshot_preresume(struct dm_target *ti) static void snapshot_resume(struct dm_target *ti) { struct dm_snapshot *s = ti->private; - struct dm_snapshot *snap_src = NULL, *snap_dest = NULL; + struct dm_snapshot *snap_src = NULL, *snap_dest = NULL, *snap_merging = NULL; struct dm_origin *o; struct mapped_device *origin_md = NULL; + bool must_restart_merging = false; down_read(&_origins_lock); o = __lookup_dm_origin(s->origin->bdev); if (o) origin_md = dm_table_get_md(o->ti->table); + if (!origin_md) { + (void) __find_snapshots_sharing_cow(s, NULL, NULL, &snap_merging); + if (snap_merging) + origin_md = dm_table_get_md(snap_merging->ti->table); + } if (origin_md == dm_table_get_md(ti->table)) origin_md = NULL; + if (origin_md) { + if (dm_hold(origin_md)) + origin_md = NULL; + } - if (origin_md) + up_read(&_origins_lock); + + if (origin_md) { dm_internal_suspend_fast(origin_md); + if (snap_merging && test_bit(RUNNING_MERGE, &snap_merging->state_bits)) { + must_restart_merging = true; + stop_merge(snap_merging); + } + } + + down_read(&_origins_lock); (void) __find_snapshots_sharing_cow(s, &snap_src, &snap_dest, NULL); if (snap_src && snap_dest) { @@ -1912,11 +1931,15 @@ static void snapshot_resume(struct dm_target *ti) up_write(&snap_src->lock); } - if (origin_md) - dm_internal_resume_fast(origin_md); - up_read(&_origins_lock); + if (origin_md) { + if (must_restart_merging) + start_merge(snap_merging); + dm_internal_resume_fast(origin_md); + dm_put(origin_md); + } + /* Now we have correct chunk size, reregister */ reregister_snapshot(s); @@ -2360,7 +2383,7 @@ static struct target_type snapshot_target = { static struct target_type merge_target = { .name = dm_snapshot_merge_target_name, - .version = {1, 2, 0}, + .version = {1, 3, 0}, .module = THIS_MODULE, .ctr = snapshot_ctr, .dtr = snapshot_dtr, diff --git a/drivers/md/dm.c b/drivers/md/dm.c index 6e2b2e97abe9..9b641b38b857 100644 --- a/drivers/md/dm.c +++ b/drivers/md/dm.c @@ -2616,6 +2616,19 @@ void dm_get(struct mapped_device *md) BUG_ON(test_bit(DMF_FREEING, &md->flags)); } +int dm_hold(struct mapped_device *md) +{ + spin_lock(&_minor_lock); + if (test_bit(DMF_FREEING, &md->flags)) { + spin_unlock(&_minor_lock); + return -EBUSY; + } + dm_get(md); + spin_unlock(&_minor_lock); + return 0; +} +EXPORT_SYMBOL_GPL(dm_hold); + const char *dm_device_name(struct mapped_device *md) { return md->name; diff --git a/include/linux/device-mapper.h b/include/linux/device-mapper.h index 2646aed1d3fe..fd23978d93fe 100644 --- a/include/linux/device-mapper.h +++ b/include/linux/device-mapper.h @@ -375,6 +375,7 @@ int dm_create(int minor, struct mapped_device **md); */ struct mapped_device *dm_get_md(dev_t dev); void dm_get(struct mapped_device *md); +int dm_hold(struct mapped_device *md); void dm_put(struct mapped_device *md); /* -- cgit v1.2.1 From e5db29806b99ce2b2640d2e4d4fcb983cea115c5 Mon Sep 17 00:00:00 2001 From: "Darrick J. Wong" Date: Fri, 27 Feb 2015 10:44:38 -0800 Subject: dm io: deal with wandering queue limits when handling REQ_DISCARD and REQ_WRITE_SAME Since it's possible for the discard and write same queue limits to change while the upper level command is being sliced and diced, fix up both of them (a) to reject IO if the special command is unsupported at the start of the function and (b) read the limits once and let the commands error out on their own if the status happens to change. Signed-off-by: Darrick J. Wong Signed-off-by: Mikulas Patocka Signed-off-by: Mike Snitzer Cc: stable@vger.kernel.org --- drivers/md/dm-io.c | 15 +++++++++++---- 1 file changed, 11 insertions(+), 4 deletions(-) diff --git a/drivers/md/dm-io.c b/drivers/md/dm-io.c index 37de0173b6d2..74adcd2c967e 100644 --- a/drivers/md/dm-io.c +++ b/drivers/md/dm-io.c @@ -289,9 +289,16 @@ static void do_region(int rw, unsigned region, struct dm_io_region *where, struct request_queue *q = bdev_get_queue(where->bdev); unsigned short logical_block_size = queue_logical_block_size(q); sector_t num_sectors; + unsigned int uninitialized_var(special_cmd_max_sectors); - /* Reject unsupported discard requests */ - if ((rw & REQ_DISCARD) && !blk_queue_discard(q)) { + /* + * Reject unsupported discard and write same requests. + */ + if (rw & REQ_DISCARD) + special_cmd_max_sectors = q->limits.max_discard_sectors; + else if (rw & REQ_WRITE_SAME) + special_cmd_max_sectors = q->limits.max_write_same_sectors; + if ((rw & (REQ_DISCARD | REQ_WRITE_SAME)) && special_cmd_max_sectors == 0) { dec_count(io, region, -EOPNOTSUPP); return; } @@ -317,7 +324,7 @@ static void do_region(int rw, unsigned region, struct dm_io_region *where, store_io_and_region_in_bio(bio, io, region); if (rw & REQ_DISCARD) { - num_sectors = min_t(sector_t, q->limits.max_discard_sectors, remaining); + num_sectors = min_t(sector_t, special_cmd_max_sectors, remaining); bio->bi_iter.bi_size = num_sectors << SECTOR_SHIFT; remaining -= num_sectors; } else if (rw & REQ_WRITE_SAME) { @@ -326,7 +333,7 @@ static void do_region(int rw, unsigned region, struct dm_io_region *where, */ dp->get_page(dp, &page, &len, &offset); bio_add_page(bio, page, logical_block_size, offset); - num_sectors = min_t(sector_t, q->limits.max_write_same_sectors, remaining); + num_sectors = min_t(sector_t, special_cmd_max_sectors, remaining); bio->bi_iter.bi_size = num_sectors << SECTOR_SHIFT; offset = 0; -- cgit v1.2.1 From cb26285df11fc6c6c2110abc8fcb8d02f6f96764 Mon Sep 17 00:00:00 2001 From: Rickard Strandqvist Date: Sat, 20 Dec 2014 13:33:54 +0100 Subject: arch: sparc: kernel: traps_64.c: Remove some unused functions Removes some functions that are not used anywhere: do_fpdis_tl1() do_iae_tl1() do_dae_tl1() do_cee_tl1() This was partially found by using a static code analysis program called cppcheck. Signed-off-by: Rickard Strandqvist Signed-off-by: David S. Miller --- arch/sparc/kernel/entry.h | 4 ---- arch/sparc/kernel/traps_64.c | 28 ---------------------------- 2 files changed, 32 deletions(-) diff --git a/arch/sparc/kernel/entry.h b/arch/sparc/kernel/entry.h index 88d322b67fac..07cc49e541f4 100644 --- a/arch/sparc/kernel/entry.h +++ b/arch/sparc/kernel/entry.h @@ -98,11 +98,7 @@ void sun4v_do_mna(struct pt_regs *regs, void do_privop(struct pt_regs *regs); void do_privact(struct pt_regs *regs); void do_cee(struct pt_regs *regs); -void do_cee_tl1(struct pt_regs *regs); -void do_dae_tl1(struct pt_regs *regs); -void do_iae_tl1(struct pt_regs *regs); void do_div0_tl1(struct pt_regs *regs); -void do_fpdis_tl1(struct pt_regs *regs); void do_fpieee_tl1(struct pt_regs *regs); void do_fpother_tl1(struct pt_regs *regs); void do_ill_tl1(struct pt_regs *regs); diff --git a/arch/sparc/kernel/traps_64.c b/arch/sparc/kernel/traps_64.c index a27651e866e7..1ef1af4cf96b 100644 --- a/arch/sparc/kernel/traps_64.c +++ b/arch/sparc/kernel/traps_64.c @@ -2564,27 +2564,6 @@ void do_cee(struct pt_regs *regs) die_if_kernel("TL0: Cache Error Exception", regs); } -void do_cee_tl1(struct pt_regs *regs) -{ - exception_enter(); - dump_tl1_traplog((struct tl1_traplog *)(regs + 1)); - die_if_kernel("TL1: Cache Error Exception", regs); -} - -void do_dae_tl1(struct pt_regs *regs) -{ - exception_enter(); - dump_tl1_traplog((struct tl1_traplog *)(regs + 1)); - die_if_kernel("TL1: Data Access Exception", regs); -} - -void do_iae_tl1(struct pt_regs *regs) -{ - exception_enter(); - dump_tl1_traplog((struct tl1_traplog *)(regs + 1)); - die_if_kernel("TL1: Instruction Access Exception", regs); -} - void do_div0_tl1(struct pt_regs *regs) { exception_enter(); @@ -2592,13 +2571,6 @@ void do_div0_tl1(struct pt_regs *regs) die_if_kernel("TL1: DIV0 Exception", regs); } -void do_fpdis_tl1(struct pt_regs *regs) -{ - exception_enter(); - dump_tl1_traplog((struct tl1_traplog *)(regs + 1)); - die_if_kernel("TL1: FPU Disabled", regs); -} - void do_fpieee_tl1(struct pt_regs *regs) { exception_enter(); -- cgit v1.2.1 From 8f765b84918de82789c1f7650490e15208cb1eb3 Mon Sep 17 00:00:00 2001 From: Rickard Strandqvist Date: Sat, 20 Dec 2014 13:42:47 +0100 Subject: arch: sparc: kernel: starfire.c: Remove unused function Remove the function starfire_hard_smp_processor_id() that is not used anywhere. This was partially found by using a static code analysis program called cppcheck. Signed-off-by: Rickard Strandqvist Signed-off-by: David S. Miller --- arch/sparc/include/asm/starfire.h | 1 - arch/sparc/kernel/starfire.c | 5 ----- 2 files changed, 6 deletions(-) diff --git a/arch/sparc/include/asm/starfire.h b/arch/sparc/include/asm/starfire.h index c100dc27a0a9..176fa0ad19f1 100644 --- a/arch/sparc/include/asm/starfire.h +++ b/arch/sparc/include/asm/starfire.h @@ -12,7 +12,6 @@ extern int this_is_starfire; void check_if_starfire(void); -int starfire_hard_smp_processor_id(void); void starfire_hookup(int); unsigned int starfire_translate(unsigned long imap, unsigned int upaid); diff --git a/arch/sparc/kernel/starfire.c b/arch/sparc/kernel/starfire.c index 82281a566bb8..167fdfd9c837 100644 --- a/arch/sparc/kernel/starfire.c +++ b/arch/sparc/kernel/starfire.c @@ -28,11 +28,6 @@ void check_if_starfire(void) this_is_starfire = 1; } -int starfire_hard_smp_processor_id(void) -{ - return upa_readl(0x1fff40000d0UL); -} - /* * Each Starfire board has 32 registers which perform translation * and delivery of traditional interrupt packets into the extended -- cgit v1.2.1 From 94ab5990760a54bb1f0fca99e0d374260cae3b8b Mon Sep 17 00:00:00 2001 From: Dave Kleikamp Date: Tue, 6 Jan 2015 18:31:39 -0600 Subject: sparc64: fatal trap should stop all cpus "echo c > /proc/sysrq-trigger" does not result in a system crash. There are two problems. One is that the trap handler ignores the global variable, panic_on_oops. The other is that smp_send_stop() is a no-op which leaves the other cpus running normally when one cpu panics. Signed-off-by: Dave Kleikamp Signed-off-by: David S. Miller --- arch/sparc/kernel/smp_64.c | 27 ++++++++++++++++++++++++--- arch/sparc/kernel/traps_64.c | 2 ++ 2 files changed, 26 insertions(+), 3 deletions(-) diff --git a/arch/sparc/kernel/smp_64.c b/arch/sparc/kernel/smp_64.c index da6f1a7fc4db..61139d9924ca 100644 --- a/arch/sparc/kernel/smp_64.c +++ b/arch/sparc/kernel/smp_64.c @@ -1406,11 +1406,32 @@ void __irq_entry smp_receive_signal_client(int irq, struct pt_regs *regs) scheduler_ipi(); } -/* This is a nop because we capture all other cpus - * anyways when making the PROM active. - */ +static void stop_this_cpu(void *dummy) +{ + prom_stopself(); +} + void smp_send_stop(void) { + int cpu; + + if (tlb_type == hypervisor) { + for_each_online_cpu(cpu) { + if (cpu == smp_processor_id()) + continue; +#ifdef CONFIG_SUN_LDOMS + if (ldom_domaining_enabled) { + unsigned long hv_err; + hv_err = sun4v_cpu_stop(cpu); + if (hv_err) + printk(KERN_ERR "sun4v_cpu_stop() " + "failed err=%lu\n", hv_err); + } else +#endif + prom_stopcpu_cpuid(cpu); + } + } else + smp_call_function(stop_this_cpu, NULL, 0); } /** diff --git a/arch/sparc/kernel/traps_64.c b/arch/sparc/kernel/traps_64.c index 1ef1af4cf96b..0e699745d643 100644 --- a/arch/sparc/kernel/traps_64.c +++ b/arch/sparc/kernel/traps_64.c @@ -2427,6 +2427,8 @@ void __noreturn die_if_kernel(char *str, struct pt_regs *regs) } user_instruction_dump ((unsigned int __user *) regs->tpc); } + if (panic_on_oops) + panic("Fatal exception"); if (regs->tstate & TSTATE_PRIV) do_exit(SIGKILL); do_exit(SIGSEGV); -- cgit v1.2.1 From 9555b47fab149ee23bddc842c264dd6f3b51f52d Mon Sep 17 00:00:00 2001 From: Ricardo Ribalda Date: Thu, 29 Jan 2015 15:52:02 +0100 Subject: sparc: io_64.h: Replace io function-link macros Function like macros cannot be assigned to function pointers. This patch convert the function-like macros into object-macros, that the precompiler will replace with the name of the final function. With this patch this kind of code will work: if (priv->mode_big_endian) priv.read = ioread32be; else priv.read = ioread32; Same approach has been taken on asm-generic/io.h Reported-by: kbuild test robot Fixes: 99082eab63449f9d spi/xilinx: Remove iowrite/ioread wrappers Signed-off-by: Ricardo Ribalda Delgado Acked-by: David S. Miller Signed-off-by: David S. Miller --- arch/sparc/include/asm/io_64.h | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-) diff --git a/arch/sparc/include/asm/io_64.h b/arch/sparc/include/asm/io_64.h index 9b672be70dda..50d4840d9aeb 100644 --- a/arch/sparc/include/asm/io_64.h +++ b/arch/sparc/include/asm/io_64.h @@ -407,16 +407,16 @@ static inline void iounmap(volatile void __iomem *addr) { } -#define ioread8(X) readb(X) -#define ioread16(X) readw(X) -#define ioread16be(X) __raw_readw(X) -#define ioread32(X) readl(X) -#define ioread32be(X) __raw_readl(X) -#define iowrite8(val,X) writeb(val,X) -#define iowrite16(val,X) writew(val,X) -#define iowrite16be(val,X) __raw_writew(val,X) -#define iowrite32(val,X) writel(val,X) -#define iowrite32be(val,X) __raw_writel(val,X) +#define ioread8 readb +#define ioread16 readw +#define ioread16be __raw_readw +#define ioread32 readl +#define ioread32be __raw_readl +#define iowrite8 writeb +#define iowrite16 writew +#define iowrite16be __raw_writew +#define iowrite32 writel +#define iowrite32be __raw_writel /* Create a virtual mapping cookie for an IO port range */ void __iomem *ioport_map(unsigned long port, unsigned int nr); -- cgit v1.2.1 From 001eabfd54c0cbf9d7d16264ddc8cc0bee67e3ed Mon Sep 17 00:00:00 2001 From: Ard Biesheuvel Date: Thu, 26 Feb 2015 07:22:05 +0000 Subject: crypto: arm/aes update NEON AES module to latest OpenSSL version This updates the bit sliced AES module to the latest version in the upstream OpenSSL repository (e620e5ae37bc). This is needed to fix a bug in the XTS decryption path, where data chunked in a certain way could trigger the ciphertext stealing code, which is not supposed to be active in the kernel build (The kernel implementation of XTS only supports round multiples of the AES block size of 16 bytes, whereas the conformant OpenSSL implementation of XTS supports inputs of arbitrary size by applying ciphertext stealing). This is fixed in the upstream version by adding the missing #ifndef XTS_CHAIN_TWEAK around the offending instructions. The upstream code also contains the change applied by Russell to build the code unconditionally, i.e., even if __LINUX_ARM_ARCH__ < 7, but implemented slightly differently. Cc: stable@vger.kernel.org Fixes: e4e7f10bfc40 ("ARM: add support for bit sliced AES using NEON instructions") Reported-by: Adrian Kotelba Signed-off-by: Ard Biesheuvel Tested-by: Milan Broz Signed-off-by: Herbert Xu --- arch/arm/crypto/aesbs-core.S_shipped | 12 ++++++++---- arch/arm/crypto/bsaes-armv7.pl | 12 ++++++++---- 2 files changed, 16 insertions(+), 8 deletions(-) diff --git a/arch/arm/crypto/aesbs-core.S_shipped b/arch/arm/crypto/aesbs-core.S_shipped index 71e5fc7cfb18..1d1800f71c5b 100644 --- a/arch/arm/crypto/aesbs-core.S_shipped +++ b/arch/arm/crypto/aesbs-core.S_shipped @@ -58,14 +58,18 @@ # define VFP_ABI_FRAME 0 # define BSAES_ASM_EXTENDED_KEY # define XTS_CHAIN_TWEAK -# define __ARM_ARCH__ 7 +# define __ARM_ARCH__ __LINUX_ARM_ARCH__ +# define __ARM_MAX_ARCH__ 7 #endif #ifdef __thumb__ # define adrl adr #endif -#if __ARM_ARCH__>=7 +#if __ARM_MAX_ARCH__>=7 +.arch armv7-a +.fpu neon + .text .syntax unified @ ARMv7-capable assembler is expected to handle this #ifdef __thumb2__ @@ -74,8 +78,6 @@ .code 32 #endif -.fpu neon - .type _bsaes_decrypt8,%function .align 4 _bsaes_decrypt8: @@ -2095,9 +2097,11 @@ bsaes_xts_decrypt: vld1.8 {q8}, [r0] @ initial tweak adr r2, .Lxts_magic +#ifndef XTS_CHAIN_TWEAK tst r9, #0xf @ if not multiple of 16 it ne @ Thumb2 thing, sanity check in ARM subne r9, #0x10 @ subtract another 16 bytes +#endif subs r9, #0x80 blo .Lxts_dec_short diff --git a/arch/arm/crypto/bsaes-armv7.pl b/arch/arm/crypto/bsaes-armv7.pl index be068db960ee..a4d3856e7d24 100644 --- a/arch/arm/crypto/bsaes-armv7.pl +++ b/arch/arm/crypto/bsaes-armv7.pl @@ -701,14 +701,18 @@ $code.=<<___; # define VFP_ABI_FRAME 0 # define BSAES_ASM_EXTENDED_KEY # define XTS_CHAIN_TWEAK -# define __ARM_ARCH__ 7 +# define __ARM_ARCH__ __LINUX_ARM_ARCH__ +# define __ARM_MAX_ARCH__ 7 #endif #ifdef __thumb__ # define adrl adr #endif -#if __ARM_ARCH__>=7 +#if __ARM_MAX_ARCH__>=7 +.arch armv7-a +.fpu neon + .text .syntax unified @ ARMv7-capable assembler is expected to handle this #ifdef __thumb2__ @@ -717,8 +721,6 @@ $code.=<<___; .code 32 #endif -.fpu neon - .type _bsaes_decrypt8,%function .align 4 _bsaes_decrypt8: @@ -2076,9 +2078,11 @@ bsaes_xts_decrypt: vld1.8 {@XMM[8]}, [r0] @ initial tweak adr $magic, .Lxts_magic +#ifndef XTS_CHAIN_TWEAK tst $len, #0xf @ if not multiple of 16 it ne @ Thumb2 thing, sanity check in ARM subne $len, #0x10 @ subtract another 16 bytes +#endif subs $len, #0x80 blo .Lxts_dec_short -- cgit v1.2.1 From 486b908d4412510d66ee348ba765de8d93441345 Mon Sep 17 00:00:00 2001 From: Ping Cheng Date: Fri, 20 Feb 2015 14:25:58 -0800 Subject: HID: wacom: do not send pen events before touch is up/forced out If pen comes in proximity while touch is down, we force touch up before sending pen events. Otherwise, there can be unfinished touch events compete with pen events. This idea has been fully implemented for Tablet PCs. But other tablets that support both pen and touch are not fully considered. Signed-off-by: Ping Cheng Signed-off-by: Jiri Kosina --- drivers/hid/wacom_wac.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/hid/wacom_wac.c b/drivers/hid/wacom_wac.c index 046351cf17f3..69827c928e50 100644 --- a/drivers/hid/wacom_wac.c +++ b/drivers/hid/wacom_wac.c @@ -554,6 +554,9 @@ static int wacom_intuos_inout(struct wacom_wac *wacom) if (features->quirks & WACOM_QUIRK_MULTI_INPUT) wacom->shared->stylus_in_proximity = true; + if (wacom->shared->touch_down) + return 1; + /* in Range while exiting */ if (((data[1] & 0xfe) == 0x20) && wacom->reporting_data) { input_report_key(input, BTN_TOUCH, 0); @@ -1759,6 +1762,9 @@ static int wacom_bpt_pen(struct wacom_wac *wacom) return 0; } + if (wacom->shared->touch_down) + return 0; + prox = (data[1] & 0x20) == 0x20; /* -- cgit v1.2.1 From e0d41fd435ad71b86380f27195aa117400439f37 Mon Sep 17 00:00:00 2001 From: Ping Cheng Date: Fri, 20 Feb 2015 14:27:30 -0800 Subject: HID: wacom: rely on actual touch down count to decide touch_down touch_down is a flag to indicate if there are touches on tablet or not. Since one set of touch events may be posted over more than one data packet/touch frame, and pen may come in proximity while touch events are partially sent, counting all touch events for the set reflects the actual status of touch_down. Signed-off-by: Ping Cheng Signed-off-by: Jiri Kosina --- drivers/hid/wacom_wac.c | 75 ++++++++++++++++++++++++++++--------------------- 1 file changed, 43 insertions(+), 32 deletions(-) diff --git a/drivers/hid/wacom_wac.c b/drivers/hid/wacom_wac.c index 69827c928e50..cf767419cdc4 100644 --- a/drivers/hid/wacom_wac.c +++ b/drivers/hid/wacom_wac.c @@ -1046,27 +1046,28 @@ static int wacom_24hdt_irq(struct wacom_wac *wacom) struct input_dev *input = wacom->input; unsigned char *data = wacom->data; int i; - int current_num_contacts = 0; + int current_num_contacts = data[61]; int contacts_to_send = 0; int num_contacts_left = 4; /* maximum contacts per packet */ int byte_per_packet = WACOM_BYTES_PER_24HDT_PACKET; int y_offset = 2; + static int contact_with_no_pen_down_count = 0; if (wacom->features.type == WACOM_27QHDT) { current_num_contacts = data[63]; num_contacts_left = 10; byte_per_packet = WACOM_BYTES_PER_QHDTHID_PACKET; y_offset = 0; - } else { - current_num_contacts = data[61]; } /* * First packet resets the counter since only the first * packet in series will have non-zero current_num_contacts. */ - if (current_num_contacts) + if (current_num_contacts) { wacom->num_contacts_left = current_num_contacts; + contact_with_no_pen_down_count = 0; + } contacts_to_send = min(num_contacts_left, wacom->num_contacts_left); @@ -1099,15 +1100,16 @@ static int wacom_24hdt_irq(struct wacom_wac *wacom) input_report_abs(input, ABS_MT_WIDTH_MINOR, min(w, h)); input_report_abs(input, ABS_MT_ORIENTATION, w > h); } + contact_with_no_pen_down_count++; } } input_mt_report_pointer_emulation(input, true); wacom->num_contacts_left -= contacts_to_send; - if (wacom->num_contacts_left <= 0) + if (wacom->num_contacts_left <= 0) { wacom->num_contacts_left = 0; - - wacom->shared->touch_down = (wacom->num_contacts_left > 0); + wacom->shared->touch_down = (contact_with_no_pen_down_count > 0); + } return 1; } @@ -1119,6 +1121,7 @@ static int wacom_mt_touch(struct wacom_wac *wacom) int current_num_contacts = data[2]; int contacts_to_send = 0; int x_offset = 0; + static int contact_with_no_pen_down_count = 0; /* MTTPC does not support Height and Width */ if (wacom->features.type == MTTPC || wacom->features.type == MTTPC_B) @@ -1128,8 +1131,10 @@ static int wacom_mt_touch(struct wacom_wac *wacom) * First packet resets the counter since only the first * packet in series will have non-zero current_num_contacts. */ - if (current_num_contacts) + if (current_num_contacts) { wacom->num_contacts_left = current_num_contacts; + contact_with_no_pen_down_count = 0; + } /* There are at most 5 contacts per packet */ contacts_to_send = min(5, wacom->num_contacts_left); @@ -1150,15 +1155,16 @@ static int wacom_mt_touch(struct wacom_wac *wacom) int y = get_unaligned_le16(&data[offset + x_offset + 9]); input_report_abs(input, ABS_MT_POSITION_X, x); input_report_abs(input, ABS_MT_POSITION_Y, y); + contact_with_no_pen_down_count++; } } input_mt_report_pointer_emulation(input, true); wacom->num_contacts_left -= contacts_to_send; - if (wacom->num_contacts_left < 0) + if (wacom->num_contacts_left <= 0) { wacom->num_contacts_left = 0; - - wacom->shared->touch_down = (wacom->num_contacts_left > 0); + wacom->shared->touch_down = (contact_with_no_pen_down_count > 0); + } return 1; } @@ -1196,29 +1202,25 @@ static int wacom_tpc_single_touch(struct wacom_wac *wacom, size_t len) { unsigned char *data = wacom->data; struct input_dev *input = wacom->input; - bool prox; + bool prox = !wacom->shared->stylus_in_proximity; int x = 0, y = 0; if (wacom->features.touch_max > 1 || len > WACOM_PKGLEN_TPC2FG) return 0; - if (!wacom->shared->stylus_in_proximity) { - if (len == WACOM_PKGLEN_TPC1FG) { - prox = data[0] & 0x01; - x = get_unaligned_le16(&data[1]); - y = get_unaligned_le16(&data[3]); - } else if (len == WACOM_PKGLEN_TPC1FG_B) { - prox = data[2] & 0x01; - x = get_unaligned_le16(&data[3]); - y = get_unaligned_le16(&data[5]); - } else { - prox = data[1] & 0x01; - x = le16_to_cpup((__le16 *)&data[2]); - y = le16_to_cpup((__le16 *)&data[4]); - } - } else - /* force touch out when pen is in prox */ - prox = 0; + if (len == WACOM_PKGLEN_TPC1FG) { + prox = prox && (data[0] & 0x01); + x = get_unaligned_le16(&data[1]); + y = get_unaligned_le16(&data[3]); + } else if (len == WACOM_PKGLEN_TPC1FG_B) { + prox = prox && (data[2] & 0x01); + x = get_unaligned_le16(&data[3]); + y = get_unaligned_le16(&data[5]); + } else { + prox = prox && (data[1] & 0x01); + x = le16_to_cpup((__le16 *)&data[2]); + y = le16_to_cpup((__le16 *)&data[4]); + } if (prox) { input_report_abs(input, ABS_X, x); @@ -1616,6 +1618,7 @@ static int wacom_bpt_touch(struct wacom_wac *wacom) struct input_dev *pad_input = wacom->pad_input; unsigned char *data = wacom->data; int i; + int contact_with_no_pen_down_count = 0; if (data[0] != 0x02) return 0; @@ -1643,6 +1646,7 @@ static int wacom_bpt_touch(struct wacom_wac *wacom) } input_report_abs(input, ABS_MT_POSITION_X, x); input_report_abs(input, ABS_MT_POSITION_Y, y); + contact_with_no_pen_down_count++; } } @@ -1652,11 +1656,12 @@ static int wacom_bpt_touch(struct wacom_wac *wacom) input_report_key(pad_input, BTN_FORWARD, (data[1] & 0x04) != 0); input_report_key(pad_input, BTN_BACK, (data[1] & 0x02) != 0); input_report_key(pad_input, BTN_RIGHT, (data[1] & 0x01) != 0); + wacom->shared->touch_down = (contact_with_no_pen_down_count > 0); return 1; } -static void wacom_bpt3_touch_msg(struct wacom_wac *wacom, unsigned char *data) +static int wacom_bpt3_touch_msg(struct wacom_wac *wacom, unsigned char *data, int last_touch_count) { struct wacom_features *features = &wacom->features; struct input_dev *input = wacom->input; @@ -1664,7 +1669,7 @@ static void wacom_bpt3_touch_msg(struct wacom_wac *wacom, unsigned char *data) int slot = input_mt_get_slot_by_key(input, data[0]); if (slot < 0) - return; + return 0; touch = touch && !wacom->shared->stylus_in_proximity; @@ -1696,7 +1701,9 @@ static void wacom_bpt3_touch_msg(struct wacom_wac *wacom, unsigned char *data) input_report_abs(input, ABS_MT_POSITION_Y, y); input_report_abs(input, ABS_MT_TOUCH_MAJOR, width); input_report_abs(input, ABS_MT_TOUCH_MINOR, height); + last_touch_count++; } + return last_touch_count; } static void wacom_bpt3_button_msg(struct wacom_wac *wacom, unsigned char *data) @@ -1721,6 +1728,7 @@ static int wacom_bpt3_touch(struct wacom_wac *wacom) unsigned char *data = wacom->data; int count = data[1] & 0x07; int i; + int contact_with_no_pen_down_count = 0; if (data[0] != 0x02) return 0; @@ -1731,12 +1739,15 @@ static int wacom_bpt3_touch(struct wacom_wac *wacom) int msg_id = data[offset]; if (msg_id >= 2 && msg_id <= 17) - wacom_bpt3_touch_msg(wacom, data + offset); + contact_with_no_pen_down_count = + wacom_bpt3_touch_msg(wacom, data + offset, + contact_with_no_pen_down_count); else if (msg_id == 128) wacom_bpt3_button_msg(wacom, data + offset); } input_mt_report_pointer_emulation(input, true); + wacom->shared->touch_down = (contact_with_no_pen_down_count > 0); return 1; } -- cgit v1.2.1 From d0c22119f574b851e63360c6b8660fe9593bbc3c Mon Sep 17 00:00:00 2001 From: Bob Copeland Date: Mon, 2 Mar 2015 14:28:52 -0500 Subject: mac80211: drop unencrypted frames in mesh fwding The mesh forwarding path was not checking that data frames were protected when running an encrypted network; add the necessary check. Cc: stable@vger.kernel.org Reported-by: Johannes Berg Signed-off-by: Bob Copeland Signed-off-by: Johannes Berg --- net/mac80211/rx.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index 1101563357ea..944bdc04e913 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -2214,6 +2214,9 @@ ieee80211_rx_h_mesh_fwding(struct ieee80211_rx_data *rx) hdr = (struct ieee80211_hdr *) skb->data; mesh_hdr = (struct ieee80211s_hdr *) (skb->data + hdrlen); + if (ieee80211_drop_unencrypted(rx, hdr->frame_control)) + return RX_DROP_MONITOR; + /* frame is in RMC, don't forward */ if (ieee80211_is_data(hdr->frame_control) && is_multicast_ether_addr(hdr->addr1) && -- cgit v1.2.1 From aa75ebc275b2a91b193654a177daf900ad6703f0 Mon Sep 17 00:00:00 2001 From: Michal Kazior Date: Tue, 10 Feb 2015 12:48:44 +0100 Subject: mac80211: disable u-APSD queues by default Some APs experience problems when working with U-APSD. Decreasing the probability of that happening by using legacy mode for all ACs but VO isn't enough. Cisco 4410N originally forced us to enable VO by default only because it treated non-VO ACs as legacy. However some APs (notably Netgear R7000) silently reclassify packets to different ACs. Since u-APSD ACs require trigger frames for frame retrieval clients would never see some frames (e.g. ARP responses) or would fetch them accidentally after a long time. It makes little sense to enable u-APSD queues by default because it needs userspace applications to be aware of it to actually take advantage of the possible additional powersavings. Implicitly depending on driver autotrigger frame support doesn't make much sense. Cc: stable@vger.kernel.org Signed-off-by: Michal Kazior Signed-off-by: Johannes Berg --- net/mac80211/ieee80211_i.h | 23 +++++++++++++++++------ 1 file changed, 17 insertions(+), 6 deletions(-) diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index 3afe36824703..c0e089c194f1 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -58,13 +58,24 @@ struct ieee80211_local; #define IEEE80211_UNSET_POWER_LEVEL INT_MIN /* - * Some APs experience problems when working with U-APSD. Decrease the - * probability of that happening by using legacy mode for all ACs but VO. - * The AP that caused us trouble was a Cisco 4410N. It ignores our - * setting, and always treats non-VO ACs as legacy. + * Some APs experience problems when working with U-APSD. Decreasing the + * probability of that happening by using legacy mode for all ACs but VO isn't + * enough. + * + * Cisco 4410N originally forced us to enable VO by default only because it + * treated non-VO ACs as legacy. + * + * However some APs (notably Netgear R7000) silently reclassify packets to + * different ACs. Since u-APSD ACs require trigger frames for frame retrieval + * clients would never see some frames (e.g. ARP responses) or would fetch them + * accidentally after a long time. + * + * It makes little sense to enable u-APSD queues by default because it needs + * userspace applications to be aware of it to actually take advantage of the + * possible additional powersavings. Implicitly depending on driver autotrigger + * frame support doesn't make much sense. */ -#define IEEE80211_DEFAULT_UAPSD_QUEUES \ - IEEE80211_WMM_IE_STA_QOSINFO_AC_VO +#define IEEE80211_DEFAULT_UAPSD_QUEUES 0 #define IEEE80211_DEFAULT_MAX_SP_LEN \ IEEE80211_WMM_IE_STA_QOSINFO_SP_ALL -- cgit v1.2.1 From 0ff66cffde47de51c155ebdd2356403276c04cc4 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Rafa=C5=82=20Mi=C5=82ecki?= Date: Mon, 2 Mar 2015 17:18:55 +0100 Subject: b43: fix support for 5 GHz only BCM43228 model MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit It was incorrectly detected as 2 GHz device. Signed-off-by: Rafał Miłecki Cc: stable@vger.kernel.org # 3.17+ Signed-off-by: Kalle Valo --- drivers/net/wireless/b43/main.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/net/wireless/b43/main.c b/drivers/net/wireless/b43/main.c index ccbdb05b28cd..75345c1e8c34 100644 --- a/drivers/net/wireless/b43/main.c +++ b/drivers/net/wireless/b43/main.c @@ -5370,6 +5370,7 @@ static void b43_supported_bands(struct b43_wldev *dev, bool *have_2ghz_phy, case 0x432a: /* BCM4321 */ case 0x432d: /* BCM4322 */ case 0x4352: /* BCM43222 */ + case 0x435a: /* BCM43228 */ case 0x4333: /* BCM4331 */ case 0x43a2: /* BCM4360 */ case 0x43b3: /* BCM4352 */ -- cgit v1.2.1 From c8f0345586694a33f828bc6b177fb21eb1702325 Mon Sep 17 00:00:00 2001 From: Larry Finger Date: Tue, 24 Feb 2015 09:23:01 -0600 Subject: rtlwifi: Improve handling of IPv6 packets Routine rtl_is_special_data() is supposed to identify packets that need to use a low bit rate so that the probability of successful transmission is high. The current version has a bug that causes all IPv6 packets to be labelled as special, with a corresponding low rate of transmission. A complete fix will be quite intrusive, but until that is available, all IPv6 packets are identified as regular. This patch also removes a magic number. Reported-and-tested-by: Alan Fisher Signed-off-by: Larry Finger Cc: Stable [3.18+] Cc: Alan Fisher Signed-off-by: Kalle Valo --- drivers/net/wireless/rtlwifi/base.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/drivers/net/wireless/rtlwifi/base.c b/drivers/net/wireless/rtlwifi/base.c index 1d4677460711..074f716020aa 100644 --- a/drivers/net/wireless/rtlwifi/base.c +++ b/drivers/net/wireless/rtlwifi/base.c @@ -1386,8 +1386,11 @@ u8 rtl_is_special_data(struct ieee80211_hw *hw, struct sk_buff *skb, u8 is_tx) } return true; - } else if (0x86DD == ether_type) { - return true; + } else if (ETH_P_IPV6 == ether_type) { + /* TODO: Handle any IPv6 cases that need special handling. + * For now, always return false + */ + goto end; } end: -- cgit v1.2.1 From 53eb2516972b8c4628651dfcb926cb9ef8b2864a Mon Sep 17 00:00:00 2001 From: Rob Gardner Date: Mon, 2 Mar 2015 23:16:55 -0700 Subject: sparc: semtimedop() unreachable due to comparison error A bug was reported that the semtimedop() system call was always failing eith ENOSYS. Since SEMCTL is defined as 3, and SEMTIMEDOP is defined as 4, the comparison "call <= SEMCTL" will always prevent SEMTIMEDOP from getting through to the semaphore ops switch statement. This is corrected by changing the comparison to "call <= SEMTIMEDOP". Orabug: 20633375 Signed-off-by: Rob Gardner Signed-off-by: David S. Miller --- arch/sparc/kernel/sys_sparc_64.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/sparc/kernel/sys_sparc_64.c b/arch/sparc/kernel/sys_sparc_64.c index c85403d0496c..30e7ddb27a3a 100644 --- a/arch/sparc/kernel/sys_sparc_64.c +++ b/arch/sparc/kernel/sys_sparc_64.c @@ -333,7 +333,7 @@ SYSCALL_DEFINE6(sparc_ipc, unsigned int, call, int, first, unsigned long, second long err; /* No need for backward compatibility. We can start fresh... */ - if (call <= SEMCTL) { + if (call <= SEMTIMEDOP) { switch (call) { case SEMOP: err = sys_semtimedop(first, ptr, -- cgit v1.2.1 From 21647f73835efdd0cc71b899668a8848ad9bd8cf Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Tue, 24 Feb 2015 19:52:47 +0800 Subject: phy: miphy28lp: Avoid calling of_get_child_count() multiple times Currently, of_get_child_count() is called in each iteration of the for loop in miphy28lp_xlate(). This patch stores the return value of of_get_child_count() in miphy_dev->nphys and call of_get_child_count() once in miphy28lp_probe(). Signed-off-by: Axel Lin Signed-off-by: Kishon Vijay Abraham I --- drivers/phy/phy-miphy28lp.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/drivers/phy/phy-miphy28lp.c b/drivers/phy/phy-miphy28lp.c index 9b2848e6115d..d44493230d0c 100644 --- a/drivers/phy/phy-miphy28lp.c +++ b/drivers/phy/phy-miphy28lp.c @@ -228,6 +228,7 @@ struct miphy28lp_dev { struct regmap *regmap; struct mutex miphy_mutex; struct miphy28lp_phy **phys; + int nphys; }; struct miphy_initval { @@ -1116,7 +1117,7 @@ static struct phy *miphy28lp_xlate(struct device *dev, return ERR_PTR(-EINVAL); } - for (index = 0; index < of_get_child_count(dev->of_node); index++) + for (index = 0; index < miphy_dev->nphys; index++) if (phynode == miphy_dev->phys[index]->phy->dev.of_node) { miphy_phy = miphy_dev->phys[index]; break; @@ -1200,15 +1201,15 @@ static int miphy28lp_probe(struct platform_device *pdev) struct miphy28lp_dev *miphy_dev; struct phy_provider *provider; struct phy *phy; - int chancount, port = 0; - int ret; + int ret, port = 0; miphy_dev = devm_kzalloc(&pdev->dev, sizeof(*miphy_dev), GFP_KERNEL); if (!miphy_dev) return -ENOMEM; - chancount = of_get_child_count(np); - miphy_dev->phys = devm_kzalloc(&pdev->dev, sizeof(phy) * chancount, + miphy_dev->nphys = of_get_child_count(np); + miphy_dev->phys = devm_kzalloc(&pdev->dev, + sizeof(phy) * miphy_dev->nphys, GFP_KERNEL); if (!miphy_dev->phys) return -ENOMEM; -- cgit v1.2.1 From 5bd568f5d2c95dbfeda4425c0d355093e29ae7f2 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Tue, 24 Feb 2015 19:53:45 +0800 Subject: phy: miphy365x: Avoid calling of_get_child_count() multiple times Currently, of_get_child_count() is called in each iteration of the for loop in miphy365x_xlate(). This patch stores the return value of of_get_child_count() in miphy_dev->nphys and call of_get_child_count() once in miphy365x_probe(). Signed-off-by: Axel Lin Signed-off-by: Kishon Vijay Abraham I --- drivers/phy/phy-miphy365x.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/drivers/phy/phy-miphy365x.c b/drivers/phy/phy-miphy365x.c index 6c80154e8bff..61177a6c465a 100644 --- a/drivers/phy/phy-miphy365x.c +++ b/drivers/phy/phy-miphy365x.c @@ -150,6 +150,7 @@ struct miphy365x_dev { struct regmap *regmap; struct mutex miphy_mutex; struct miphy365x_phy **phys; + int nphys; }; /* @@ -485,7 +486,7 @@ static struct phy *miphy365x_xlate(struct device *dev, return ERR_PTR(-EINVAL); } - for (index = 0; index < of_get_child_count(dev->of_node); index++) + for (index = 0; index < miphy_dev->nphys; index++) if (phynode == miphy_dev->phys[index]->phy->dev.of_node) { miphy_phy = miphy_dev->phys[index]; break; @@ -541,15 +542,15 @@ static int miphy365x_probe(struct platform_device *pdev) struct miphy365x_dev *miphy_dev; struct phy_provider *provider; struct phy *phy; - int chancount, port = 0; - int ret; + int ret, port = 0; miphy_dev = devm_kzalloc(&pdev->dev, sizeof(*miphy_dev), GFP_KERNEL); if (!miphy_dev) return -ENOMEM; - chancount = of_get_child_count(np); - miphy_dev->phys = devm_kzalloc(&pdev->dev, sizeof(phy) * chancount, + miphy_dev->nphys = of_get_child_count(np); + miphy_dev->phys = devm_kzalloc(&pdev->dev, + sizeof(phy) * miphy_dev->nphys, GFP_KERNEL); if (!miphy_dev->phys) return -ENOMEM; -- cgit v1.2.1 From 7a83b145b5891f445009b341361e2e458bd13d2b Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Thu, 26 Feb 2015 07:24:49 +0800 Subject: phy: armada375-usb2: Set drvdata for phy and use it At the context we have pointer to struct phy, it's useful to call phy_get_drvdata() to get the address of cluster_phy. This has slightly better readability than calling dev_get_drvdata(phy->dev.parent). Signed-off-by: Axel Lin Signed-off-by: Kishon Vijay Abraham I --- drivers/phy/phy-armada375-usb2.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/phy/phy-armada375-usb2.c b/drivers/phy/phy-armada375-usb2.c index 7c99ca256f05..8ccc3952c13d 100644 --- a/drivers/phy/phy-armada375-usb2.c +++ b/drivers/phy/phy-armada375-usb2.c @@ -37,7 +37,7 @@ static int armada375_usb_phy_init(struct phy *phy) struct armada375_cluster_phy *cluster_phy; u32 reg; - cluster_phy = dev_get_drvdata(phy->dev.parent); + cluster_phy = phy_get_drvdata(phy); if (!cluster_phy) return -ENODEV; @@ -131,6 +131,7 @@ static int armada375_usb_phy_probe(struct platform_device *pdev) cluster_phy->reg = usb_cluster_base; dev_set_drvdata(dev, cluster_phy); + phy_set_drvdata(phy, cluster_phy); phy_provider = devm_of_phy_provider_register(&pdev->dev, armada375_usb_phy_xlate); -- cgit v1.2.1 From 991e45f8f8ff079a04caa710be417e8e713e092c Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Wed, 25 Feb 2015 22:10:53 +0800 Subject: phy: xgene: Remove duplicate code to set ctx->dev Set it once is enough and it's done after devm_kzalloc(). Signed-off-by: Axel Lin Signed-off-by: Kishon Vijay Abraham I --- drivers/phy/phy-xgene.c | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/phy/phy-xgene.c b/drivers/phy/phy-xgene.c index 29214a36ea28..2263cd010032 100644 --- a/drivers/phy/phy-xgene.c +++ b/drivers/phy/phy-xgene.c @@ -1704,7 +1704,6 @@ static int xgene_phy_probe(struct platform_device *pdev) for (i = 0; i < MAX_LANE; i++) ctx->sata_param.speed[i] = 2; /* Default to Gen3 */ - ctx->dev = &pdev->dev; platform_set_drvdata(pdev, ctx); ctx->phy = devm_phy_create(ctx->dev, NULL, &xgene_phy_ops); -- cgit v1.2.1 From 235b633eb513f8471b9f23635345ede6e49371ab Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Wed, 25 Feb 2015 22:12:53 +0800 Subject: phy: miphy28lp: Add missing .owner field in miphy28lp_ops Add missing .owner field in miphy28lp_ops, which is used for refcounting. Signed-off-by: Axel Lin Signed-off-by: Kishon Vijay Abraham I --- drivers/phy/phy-miphy28lp.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/phy/phy-miphy28lp.c b/drivers/phy/phy-miphy28lp.c index d44493230d0c..4fe1755e3aa8 100644 --- a/drivers/phy/phy-miphy28lp.c +++ b/drivers/phy/phy-miphy28lp.c @@ -1139,6 +1139,7 @@ static struct phy *miphy28lp_xlate(struct device *dev, static struct phy_ops miphy28lp_ops = { .init = miphy28lp_init, + .owner = THIS_MODULE, }; static int miphy28lp_probe_resets(struct device_node *node, -- cgit v1.2.1 From cfd565d1e102941ec61b2a33c3c474961300a6fb Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Thu, 26 Feb 2015 11:48:08 +0800 Subject: phy: exynos-mipi-video: Fixup the test for state->regmap syscon_regmap_lookup_by_phandle() returns ERR_PTR on error. Thus don't use null test against state->regmap. Signed-off-by: Axel Lin Signed-off-by: Kishon Vijay Abraham I --- drivers/phy/phy-exynos-mipi-video.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/phy/phy-exynos-mipi-video.c b/drivers/phy/phy-exynos-mipi-video.c index f017b2f2a54e..d19649328d05 100644 --- a/drivers/phy/phy-exynos-mipi-video.c +++ b/drivers/phy/phy-exynos-mipi-video.c @@ -59,7 +59,7 @@ static int __set_phy_state(struct exynos_mipi_video_phy *state, else reset = EXYNOS4_MIPI_PHY_SRESETN; - if (state->regmap) { + if (!IS_ERR(state->regmap)) { mutex_lock(&state->mutex); regmap_read(state->regmap, offset, &val); if (on) -- cgit v1.2.1 From 7d70e15480c0450d2bfafaad338a32e884fc215e Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Wed, 4 Mar 2015 10:37:43 -0500 Subject: writeback: add missing INITIAL_JIFFIES init in global_update_bandwidth() global_update_bandwidth() uses static variable update_time as the timestamp for the last update but forgets to initialize it to INITIALIZE_JIFFIES. This means that global_dirty_limit will be 5 mins into the future on 32bit and some large amount jiffies into the past on 64bit. This isn't critical as the only effect is that global_dirty_limit won't be updated for the first 5 mins after booting on 32bit machines, especially given the auxiliary nature of global_dirty_limit's role - protecting against global dirty threshold's sudden dips; however, it does lead to unintended suboptimal behavior. Fix it. Fixes: c42843f2f0bb ("writeback: introduce smoothed global dirty limit") Signed-off-by: Tejun Heo Acked-by: Jan Kara Cc: Wu Fengguang Cc: Jens Axboe Cc: stable@vger.kernel.org Signed-off-by: Jens Axboe --- mm/page-writeback.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/mm/page-writeback.c b/mm/page-writeback.c index 45e187b2d971..b4fd980a93eb 100644 --- a/mm/page-writeback.c +++ b/mm/page-writeback.c @@ -922,7 +922,7 @@ static void global_update_bandwidth(unsigned long thresh, unsigned long now) { static DEFINE_SPINLOCK(dirty_lock); - static unsigned long update_time; + static unsigned long update_time = INITIAL_JIFFIES; /* * check locklessly first to optimize away locking for the most time -- cgit v1.2.1 From 4ceba98d3fe204c59e5f63c4d834b45dcfe789f0 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Wed, 4 Mar 2015 15:29:17 +0100 Subject: regmap: Skip read-only registers in regcache_sync() regcache_sync() spews warnings when a value was cached for a read-only register as it tries to write all registers no matter whether they are writable or not. This patch adds regmap_wrtieable() checks for avoiding it in regcache_sync_block_single() and regcache_block_raw(). Signed-off-by: Takashi Iwai Signed-off-by: Mark Brown --- drivers/base/regmap/regcache.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/base/regmap/regcache.c b/drivers/base/regmap/regcache.c index f373c35f9e1d..da84f544c544 100644 --- a/drivers/base/regmap/regcache.c +++ b/drivers/base/regmap/regcache.c @@ -608,7 +608,8 @@ static int regcache_sync_block_single(struct regmap *map, void *block, for (i = start; i < end; i++) { regtmp = block_base + (i * map->reg_stride); - if (!regcache_reg_present(cache_present, i)) + if (!regcache_reg_present(cache_present, i) || + !regmap_writeable(map, regtmp)) continue; val = regcache_get_val(map, block, i); @@ -677,7 +678,8 @@ static int regcache_sync_block_raw(struct regmap *map, void *block, for (i = start; i < end; i++) { regtmp = block_base + (i * map->reg_stride); - if (!regcache_reg_present(cache_present, i)) { + if (!regcache_reg_present(cache_present, i) || + !regmap_writeable(map, regtmp)) { ret = regcache_sync_block_raw_flush(map, &data, base, regtmp); if (ret != 0) -- cgit v1.2.1 From ce9594c6b332fd6fe464e22a83b0e6e0a287aac6 Mon Sep 17 00:00:00 2001 From: Lars-Peter Clausen Date: Fri, 27 Feb 2015 14:13:02 +0100 Subject: ASoC: ak4671: Fix control-less DAPM routes Routes without a control must use NULL for the control name. The ak4671 driver uses "NULL" instead in a few places. Previous to commit 5fe5b767dc6f ("ASoC: dapm: Do not pretend to support controls for non mixer/mux widgets") the DAPM core silently ignored non-NULL controls on non-mixer and non-mux routes. But starting with that commit it will complain and not add the route breaking the ak4671 driver in the process. This patch replaces the incorrect "NULL" control name with NULL to fix the issue. Fixes: 5fe5b767dc6f ("ASoC: dapm: Do not pretend to support controls for non mixer/mux widgets") Signed-off-by: Lars-Peter Clausen Signed-off-by: Mark Brown Cc: stable@vger.kernel.org --- sound/soc/codecs/ak4671.c | 44 ++++++++++++++++++++++---------------------- 1 file changed, 22 insertions(+), 22 deletions(-) diff --git a/sound/soc/codecs/ak4671.c b/sound/soc/codecs/ak4671.c index 632e89f793a7..2a58b1dccd2f 100644 --- a/sound/soc/codecs/ak4671.c +++ b/sound/soc/codecs/ak4671.c @@ -343,25 +343,25 @@ static const struct snd_soc_dapm_widget ak4671_dapm_widgets[] = { }; static const struct snd_soc_dapm_route ak4671_intercon[] = { - {"DAC Left", "NULL", "PMPLL"}, - {"DAC Right", "NULL", "PMPLL"}, - {"ADC Left", "NULL", "PMPLL"}, - {"ADC Right", "NULL", "PMPLL"}, + {"DAC Left", NULL, "PMPLL"}, + {"DAC Right", NULL, "PMPLL"}, + {"ADC Left", NULL, "PMPLL"}, + {"ADC Right", NULL, "PMPLL"}, /* Outputs */ - {"LOUT1", "NULL", "LOUT1 Mixer"}, - {"ROUT1", "NULL", "ROUT1 Mixer"}, - {"LOUT2", "NULL", "LOUT2 Mix Amp"}, - {"ROUT2", "NULL", "ROUT2 Mix Amp"}, - {"LOUT3", "NULL", "LOUT3 Mixer"}, - {"ROUT3", "NULL", "ROUT3 Mixer"}, + {"LOUT1", NULL, "LOUT1 Mixer"}, + {"ROUT1", NULL, "ROUT1 Mixer"}, + {"LOUT2", NULL, "LOUT2 Mix Amp"}, + {"ROUT2", NULL, "ROUT2 Mix Amp"}, + {"LOUT3", NULL, "LOUT3 Mixer"}, + {"ROUT3", NULL, "ROUT3 Mixer"}, {"LOUT1 Mixer", "DACL", "DAC Left"}, {"ROUT1 Mixer", "DACR", "DAC Right"}, {"LOUT2 Mixer", "DACHL", "DAC Left"}, {"ROUT2 Mixer", "DACHR", "DAC Right"}, - {"LOUT2 Mix Amp", "NULL", "LOUT2 Mixer"}, - {"ROUT2 Mix Amp", "NULL", "ROUT2 Mixer"}, + {"LOUT2 Mix Amp", NULL, "LOUT2 Mixer"}, + {"ROUT2 Mix Amp", NULL, "ROUT2 Mixer"}, {"LOUT3 Mixer", "DACSL", "DAC Left"}, {"ROUT3 Mixer", "DACSR", "DAC Right"}, @@ -381,18 +381,18 @@ static const struct snd_soc_dapm_route ak4671_intercon[] = { {"LIN2", NULL, "Mic Bias"}, {"RIN2", NULL, "Mic Bias"}, - {"ADC Left", "NULL", "LIN MUX"}, - {"ADC Right", "NULL", "RIN MUX"}, + {"ADC Left", NULL, "LIN MUX"}, + {"ADC Right", NULL, "RIN MUX"}, /* Analog Loops */ - {"LIN1 Mixing Circuit", "NULL", "LIN1"}, - {"RIN1 Mixing Circuit", "NULL", "RIN1"}, - {"LIN2 Mixing Circuit", "NULL", "LIN2"}, - {"RIN2 Mixing Circuit", "NULL", "RIN2"}, - {"LIN3 Mixing Circuit", "NULL", "LIN3"}, - {"RIN3 Mixing Circuit", "NULL", "RIN3"}, - {"LIN4 Mixing Circuit", "NULL", "LIN4"}, - {"RIN4 Mixing Circuit", "NULL", "RIN4"}, + {"LIN1 Mixing Circuit", NULL, "LIN1"}, + {"RIN1 Mixing Circuit", NULL, "RIN1"}, + {"LIN2 Mixing Circuit", NULL, "LIN2"}, + {"RIN2 Mixing Circuit", NULL, "RIN2"}, + {"LIN3 Mixing Circuit", NULL, "LIN3"}, + {"RIN3 Mixing Circuit", NULL, "RIN3"}, + {"LIN4 Mixing Circuit", NULL, "LIN4"}, + {"RIN4 Mixing Circuit", NULL, "RIN4"}, {"LOUT1 Mixer", "LINL1", "LIN1 Mixing Circuit"}, {"ROUT1 Mixer", "RINR1", "RIN1 Mixing Circuit"}, -- cgit v1.2.1 From 8e6a75c102f8e232b599a06e06731d8c5d5f2c5d Mon Sep 17 00:00:00 2001 From: Lars-Peter Clausen Date: Fri, 27 Feb 2015 14:13:03 +0100 Subject: ASoC: da732x: Fix control-less DAPM routes Routes without a control must use NULL for the control name. The da732x driver uses "NULL" instead in a few places. Previous to commit 5fe5b767dc6f ("ASoC: dapm: Do not pretend to support controls for non mixer/mux widgets") the DAPM core silently ignored non-NULL controls on non-mixer and non-mux routes. But starting with that commit it will complain and not add the route breaking the da732x driver in the process. This patch replaces the incorrect "NULL" control name with NULL to fix the issue. Fixes: 5fe5b767dc6f ("ASoC: dapm: Do not pretend to support controls for non mixer/mux widgets") Signed-off-by: Lars-Peter Clausen Acked-by: Adam Thomson Signed-off-by: Mark Brown Cc: stable@vger.kernel.org --- sound/soc/codecs/da732x.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/sound/soc/codecs/da732x.c b/sound/soc/codecs/da732x.c index ffe96175a8a5..911c26c705fc 100644 --- a/sound/soc/codecs/da732x.c +++ b/sound/soc/codecs/da732x.c @@ -876,11 +876,11 @@ static const struct snd_soc_dapm_widget da732x_dapm_widgets[] = { static const struct snd_soc_dapm_route da732x_dapm_routes[] = { /* Inputs */ - {"AUX1L PGA", "NULL", "AUX1L"}, - {"AUX1R PGA", "NULL", "AUX1R"}, + {"AUX1L PGA", NULL, "AUX1L"}, + {"AUX1R PGA", NULL, "AUX1R"}, {"MIC1 PGA", NULL, "MIC1"}, - {"MIC2 PGA", "NULL", "MIC2"}, - {"MIC3 PGA", "NULL", "MIC3"}, + {"MIC2 PGA", NULL, "MIC2"}, + {"MIC3 PGA", NULL, "MIC3"}, /* Capture Path */ {"ADC1 Left MUX", "MIC1", "MIC1 PGA"}, -- cgit v1.2.1 From cdd3d2a93f08823a0b9802147dc28c99029dfdfd Mon Sep 17 00:00:00 2001 From: Lars-Peter Clausen Date: Fri, 27 Feb 2015 14:13:04 +0100 Subject: ASoC: sn95031: Fix control-less DAPM routes Routes without a control must use NULL for the control name. The sn95031 driver uses "NULL" instead in a few places. Previous to commit 5fe5b767dc6f ("ASoC: dapm: Do not pretend to support controls for non mixer/mux widgets") the DAPM core silently ignored non-NULL controls on non-mixer and non-mux routes. But starting with that commit it will complain and not add the route breaking the sn95031 driver in the process. This patch replaces the incorrect "NULL" control name with NULL to fix the issue. Fixes: 5fe5b767dc6f ("ASoC: dapm: Do not pretend to support controls for non mixer/mux widgets") Signed-off-by: Lars-Peter Clausen Acked-by: Vinod Koul Signed-off-by: Mark Brown Cc: stable@vger.kernel.org --- sound/soc/codecs/sn95031.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/sound/soc/codecs/sn95031.c b/sound/soc/codecs/sn95031.c index 47b257e41809..82095d6cd070 100644 --- a/sound/soc/codecs/sn95031.c +++ b/sound/soc/codecs/sn95031.c @@ -538,8 +538,8 @@ static const struct snd_soc_dapm_route sn95031_audio_map[] = { /* speaker map */ { "IHFOUTL", NULL, "Speaker Rail"}, { "IHFOUTR", NULL, "Speaker Rail"}, - { "IHFOUTL", "NULL", "Speaker Left Playback"}, - { "IHFOUTR", "NULL", "Speaker Right Playback"}, + { "IHFOUTL", NULL, "Speaker Left Playback"}, + { "IHFOUTR", NULL, "Speaker Right Playback"}, { "Speaker Left Playback", NULL, "Speaker Left Filter"}, { "Speaker Right Playback", NULL, "Speaker Right Filter"}, { "Speaker Left Filter", NULL, "IHFDAC Left"}, -- cgit v1.2.1 From 90aff15b3e0858eaefdcd390e64849542845d489 Mon Sep 17 00:00:00 2001 From: "Maciej S. Szmigiero" Date: Wed, 4 Mar 2015 22:48:30 +0100 Subject: fsl_ssi: fix of_property_read_u32_array return value check of_property_read_u32_array returns 0 on success, so the return value shouldn't be inverted twice, first on assignment then in condition expression. Signed-off-by: Maciej Szmigiero Signed-off-by: Mark Brown --- sound/soc/fsl/fsl_ssi.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/fsl/fsl_ssi.c b/sound/soc/fsl/fsl_ssi.c index d7365c5d7ec0..134388f7d1b8 100644 --- a/sound/soc/fsl/fsl_ssi.c +++ b/sound/soc/fsl/fsl_ssi.c @@ -1227,7 +1227,7 @@ static int fsl_ssi_imx_probe(struct platform_device *pdev, ssi_private->dma_params_tx.addr = ssi_private->ssi_phys + CCSR_SSI_STX0; ssi_private->dma_params_rx.addr = ssi_private->ssi_phys + CCSR_SSI_SRX0; - ret = !of_property_read_u32_array(np, "dmas", dmas, 4); + ret = of_property_read_u32_array(np, "dmas", dmas, 4); if (ssi_private->use_dma && !ret && dmas[2] == IMX_DMATYPE_SSI_DUAL) { ssi_private->use_dual_fifo = true; /* When using dual fifo mode, we need to keep watermark -- cgit v1.2.1 From 4f6e24ed9de8634d6471ef86b382cba6d4e57ca8 Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Thu, 5 Mar 2015 10:45:30 +1030 Subject: virtio_console: init work unconditionally when multiport is off, we don't initialize config work, but we then cancel uninitialized control_work on freeze. Signed-off-by: Michael S. Tsirkin Reviewed-by: Amit Shah Signed-off-by: Rusty Russell Cc: stable@kernel.org --- drivers/char/virtio_console.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/char/virtio_console.c b/drivers/char/virtio_console.c index fae2dbbf5745..def736ddfc0e 100644 --- a/drivers/char/virtio_console.c +++ b/drivers/char/virtio_console.c @@ -2040,12 +2040,13 @@ static int virtcons_probe(struct virtio_device *vdev) virtio_device_ready(portdev->vdev); + INIT_WORK(&portdev->control_work, &control_work_handler); + if (multiport) { unsigned int nr_added_bufs; spin_lock_init(&portdev->c_ivq_lock); spin_lock_init(&portdev->c_ovq_lock); - INIT_WORK(&portdev->control_work, &control_work_handler); nr_added_bufs = fill_queue(portdev->c_ivq, &portdev->c_ivq_lock); -- cgit v1.2.1 From eeb8a7e8bb123e84daeef84f5a2eab99ad2839a2 Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Thu, 5 Mar 2015 10:45:49 +1030 Subject: virtio_console: avoid config access from irq when multiport is off, virtio console invokes config access from irq context, config access is blocking on s390. Fix this up by scheduling work from config irq - similar to what we do for multiport configs. Signed-off-by: Michael S. Tsirkin Reviewed-by: Amit Shah Signed-off-by: Rusty Russell Cc: stable@kernel.org --- drivers/char/virtio_console.c | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) diff --git a/drivers/char/virtio_console.c b/drivers/char/virtio_console.c index def736ddfc0e..72d7028f779b 100644 --- a/drivers/char/virtio_console.c +++ b/drivers/char/virtio_console.c @@ -142,6 +142,7 @@ struct ports_device { * notification */ struct work_struct control_work; + struct work_struct config_work; struct list_head ports; @@ -1837,10 +1838,21 @@ static void config_intr(struct virtio_device *vdev) portdev = vdev->priv; + if (!use_multiport(portdev)) + schedule_work(&portdev->config_work); +} + +static void config_work_handler(struct work_struct *work) +{ + struct ports_device *portdev; + + portdev = container_of(work, struct ports_device, control_work); if (!use_multiport(portdev)) { + struct virtio_device *vdev; struct port *port; u16 rows, cols; + vdev = portdev->vdev; virtio_cread(vdev, struct virtio_console_config, cols, &cols); virtio_cread(vdev, struct virtio_console_config, rows, &rows); @@ -2040,6 +2052,7 @@ static int virtcons_probe(struct virtio_device *vdev) virtio_device_ready(portdev->vdev); + INIT_WORK(&portdev->config_work, &config_work_handler); INIT_WORK(&portdev->control_work, &control_work_handler); if (multiport) { @@ -2114,6 +2127,8 @@ static void virtcons_remove(struct virtio_device *vdev) /* Finish up work that's lined up */ if (use_multiport(portdev)) cancel_work_sync(&portdev->control_work); + else + cancel_work_sync(&portdev->config_work); list_for_each_entry_safe(port, port2, &portdev->ports, list) unplug_port(port); @@ -2165,6 +2180,7 @@ static int virtcons_freeze(struct virtio_device *vdev) virtqueue_disable_cb(portdev->c_ivq); cancel_work_sync(&portdev->control_work); + cancel_work_sync(&portdev->config_work); /* * Once more: if control_work_handler() was running, it would * enable the cb as the last step. -- cgit v1.2.1 From ff6b8090e26ef7649ef0cc6b42389141ef48b0cf Mon Sep 17 00:00:00 2001 From: Sudip Mukherjee Date: Tue, 27 Jan 2015 18:08:22 +0530 Subject: nbd: fix possible memory leak we have already allocated memory for nbd_dev, but we were not releasing that memory and just returning the error value. Signed-off-by: Sudip Mukherjee Acked-by: Paul Clements Cc: Signed-off-by: Markus Pargmann --- drivers/block/nbd.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/block/nbd.c b/drivers/block/nbd.c index 4bc2a5cb9935..a98c41f72c63 100644 --- a/drivers/block/nbd.c +++ b/drivers/block/nbd.c @@ -803,10 +803,6 @@ static int __init nbd_init(void) return -EINVAL; } - nbd_dev = kcalloc(nbds_max, sizeof(*nbd_dev), GFP_KERNEL); - if (!nbd_dev) - return -ENOMEM; - part_shift = 0; if (max_part > 0) { part_shift = fls(max_part); @@ -828,6 +824,10 @@ static int __init nbd_init(void) if (nbds_max > 1UL << (MINORBITS - part_shift)) return -EINVAL; + nbd_dev = kcalloc(nbds_max, sizeof(*nbd_dev), GFP_KERNEL); + if (!nbd_dev) + return -ENOMEM; + for (i = 0; i < nbds_max; i++) { struct gendisk *disk = alloc_disk(1 << part_shift); if (!disk) -- cgit v1.2.1 From 40eeb111d7c88bfbc38e1dfe330bc4cec05e0806 Mon Sep 17 00:00:00 2001 From: Linus Walleij Date: Thu, 5 Mar 2015 10:08:14 +0100 Subject: Revert "pinctrl: consumer: use correct retval for placeholder functions" This reverts commit 5a7d2efdd93f6c4bb6cd3d5df3d2f5611c9b87ac. As per discussion on the mailing list, this is not the right thing to do. NULL cookies are valid in the stubs. Reported-by: Wolfram Sang Signed-off-by: Linus Walleij --- include/linux/pinctrl/consumer.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/include/linux/pinctrl/consumer.h b/include/linux/pinctrl/consumer.h index 72c0415d6c21..18eccefea06e 100644 --- a/include/linux/pinctrl/consumer.h +++ b/include/linux/pinctrl/consumer.h @@ -82,7 +82,7 @@ static inline int pinctrl_gpio_direction_output(unsigned gpio) static inline struct pinctrl * __must_check pinctrl_get(struct device *dev) { - return ERR_PTR(-ENOSYS); + return NULL; } static inline void pinctrl_put(struct pinctrl *p) @@ -93,7 +93,7 @@ static inline struct pinctrl_state * __must_check pinctrl_lookup_state( struct pinctrl *p, const char *name) { - return ERR_PTR(-ENOSYS); + return NULL; } static inline int pinctrl_select_state(struct pinctrl *p, @@ -104,7 +104,7 @@ static inline int pinctrl_select_state(struct pinctrl *p, static inline struct pinctrl * __must_check devm_pinctrl_get(struct device *dev) { - return ERR_PTR(-ENOSYS); + return NULL; } static inline void devm_pinctrl_put(struct pinctrl *p) -- cgit v1.2.1 From 4cd4b50cc2429294c23a1998c33fdfd804db0f37 Mon Sep 17 00:00:00 2001 From: Emmanuel Grumbach Date: Thu, 5 Mar 2015 13:43:15 +0200 Subject: iwlwifi: mvm: BT Coex - fix a NULL pointer exception The commit below introduced an unsafe dereference of mvmvif->phy_ctxt. It can be NULL even if we hold the mutex. We can be handling a BT Coex notification while the vif has already been unassigned. This can happen since the BT Coex notification is hanled asynchronuously: we can have started to handle the BT Coex notification trying to acquire the mutex while the unassign flow already got it. The BT Coex notification handling will wait for the mutext. I'll get it later, but then mvmvif->phy_ctxt will be NULL. Panic log: BUG: unable to handle kernel NULL pointer dereference at (null) IP: [] iwl_mvm_bt_notif_iterator+0x9d/0x340 [iwlmvm] *pdpt = 0000000000000000 *pde = f000eef300000007 Oops: 0000 [#1] SMP Workqueue: events iwl_mvm_async_handlers_wk [iwlmvm] task: ed719b20 ti: ec03e000 task.ti: ec03e000 EIP: 0060:[] EFLAGS: 00010202 CPU: 2 EIP is at iwl_mvm_bt_notif_iterator+0x9d/0x340 [iwlmvm] EAX: 00000000 EBX: f6d3cb70 ECX: f6d3cb70 EDX: 00000000 ESI: ec03fe40 EDI: efeb8810 EBP: ec03fdf0 ESP: ec03fdac DS: 007b ES: 007b FS: 00d8 GS: 0000 SS: 0068 CR0: 80050033 CR2: 00000000 CR3: 01a1a000 CR4: 001407f0 Stack: f743ca80 f744a404 ec03fdcc c10e3952 00003aba f743ca80 00000246 f743ca80 00000246 00000000 00000001 00000000 ebd45ff6 ebd458a4 f6d3c500 ebd45578 ebd44b01 ec03fe18 f99e1bc2 00000002 ebd44bc0 f9851770 00000000 f6d3c500 Call Trace: [] ? ring_buffer_unlock_commit+0xa2/0xd0 [] __iterate_interfaces+0x82/0x110 [mac80211] [] ? iwl_mvm_bt_coex_reduced_txp+0x140/0x140 [iwlmvm] [] ieee80211_iterate_active_interfaces_atomic+0x1a/0x20 [mac80211] [] iwl_mvm_bt_coex_notif_handle+0x77/0x280 [iwlmvm] [] iwl_mvm_rx_bt_coex_notif_old+0x211/0x220 [iwlmvm] [] iwl_mvm_rx_bt_coex_notif+0x19b/0x1b0 [iwlmvm] [] iwl_mvm_async_handlers_wk+0x7f/0xe0 [iwlmvm] CC: [3.19+] Fixes: 123f515635b1 ("iwlwifi: mvm: BT Coex - add support for TTC / RRC") Signed-off-by: Emmanuel Grumbach --- drivers/net/wireless/iwlwifi/mvm/coex.c | 3 ++- drivers/net/wireless/iwlwifi/mvm/coex_legacy.c | 3 ++- 2 files changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/net/wireless/iwlwifi/mvm/coex.c b/drivers/net/wireless/iwlwifi/mvm/coex.c index 1ec4d55155f7..7810c41cf9a7 100644 --- a/drivers/net/wireless/iwlwifi/mvm/coex.c +++ b/drivers/net/wireless/iwlwifi/mvm/coex.c @@ -793,7 +793,8 @@ static void iwl_mvm_bt_notif_iterator(void *_data, u8 *mac, if (!vif->bss_conf.assoc) smps_mode = IEEE80211_SMPS_AUTOMATIC; - if (IWL_COEX_IS_RRC_ON(mvm->last_bt_notif.ttc_rrc_status, + if (mvmvif->phy_ctxt && + IWL_COEX_IS_RRC_ON(mvm->last_bt_notif.ttc_rrc_status, mvmvif->phy_ctxt->id)) smps_mode = IEEE80211_SMPS_AUTOMATIC; diff --git a/drivers/net/wireless/iwlwifi/mvm/coex_legacy.c b/drivers/net/wireless/iwlwifi/mvm/coex_legacy.c index d530ef3da107..542ee74f290a 100644 --- a/drivers/net/wireless/iwlwifi/mvm/coex_legacy.c +++ b/drivers/net/wireless/iwlwifi/mvm/coex_legacy.c @@ -832,7 +832,8 @@ static void iwl_mvm_bt_notif_iterator(void *_data, u8 *mac, if (!vif->bss_conf.assoc) smps_mode = IEEE80211_SMPS_AUTOMATIC; - if (data->notif->rrc_enabled & BIT(mvmvif->phy_ctxt->id)) + if (mvmvif->phy_ctxt && + data->notif->rrc_enabled & BIT(mvmvif->phy_ctxt->id)) smps_mode = IEEE80211_SMPS_AUTOMATIC; IWL_DEBUG_COEX(data->mvm, -- cgit v1.2.1 From 6c8ca30eec7b6f8eb09c957e8dcced89e5f100c7 Mon Sep 17 00:00:00 2001 From: Nicolin Chen Date: Wed, 4 Mar 2015 21:05:04 -0800 Subject: ASoC: fsl_ssi: Don't try to round-up for PM divisor calculation According to i.MX6 Series Reference Manual, the formula to calculate the sys clock is sysclk rate = bclk rate * (div2 + 1) * (7 * psr + 1) * (pm + 1) * 2 Commit aafa85e71a75 ("ASoC: fsl_ssi: Add DAI master mode support for SSI on i.MX series") added the divisor calculation which relies on the clk_round_rate(). However, at that time, clk_round_rate() didn't provide closest clock rates for some cases because it might not use a correct rounding policy. So using the original formula (pm + 1) for PM divisor was not able to give us a desired clock rate. And then we used (pm + 2) to do the trick. However, the clk-divider driver has been refined a lot since commit b11d282dbea2 ("clk: divider: fix rate calculation for fractional rates") Now using (pm + 2) trick would result an incorrect clock rate. So this patch fixes the problem by removing the useless trick. Reported-by: Stephane Cerveau Signed-off-by: Nicolin Chen Signed-off-by: Mark Brown --- sound/soc/fsl/fsl_ssi.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/fsl/fsl_ssi.c b/sound/soc/fsl/fsl_ssi.c index 134388f7d1b8..7eebc0889c9d 100644 --- a/sound/soc/fsl/fsl_ssi.c +++ b/sound/soc/fsl/fsl_ssi.c @@ -603,7 +603,7 @@ static int fsl_ssi_set_bclk(struct snd_pcm_substream *substream, factor = (div2 + 1) * (7 * psr + 1) * 2; for (i = 0; i < 255; i++) { - tmprate = freq * factor * (i + 2); + tmprate = freq * factor * (i + 1); if (baudclk_is_used) clkrate = clk_get_rate(ssi_private->baudclk); -- cgit v1.2.1 From 7b8f10da3bf1056546133c9f54f49ce389fd95ab Mon Sep 17 00:00:00 2001 From: Yongbae Park Date: Tue, 3 Mar 2015 19:46:49 +0900 Subject: clocksource: efm32: Fix a NULL pointer dereference MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The initialisation of the efm32 clocksource first sets up the irq and only after that initialises the data needed for irq handling. In case this initialisation is delayed the irq handler would dereference a NULL pointer. I'm not aware of anything that could delay the process in such a way, but it's better to be safe than sorry, so setup the irq only when the clock event device is ready. Cc: stable@vger.kernel.org Acked-by: Uwe Kleine-König Signed-off-by: Yongbae Park Signed-off-by: Daniel Lezcano --- drivers/clocksource/time-efm32.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/clocksource/time-efm32.c b/drivers/clocksource/time-efm32.c index bba62f9deefb..ec57ba2bbd87 100644 --- a/drivers/clocksource/time-efm32.c +++ b/drivers/clocksource/time-efm32.c @@ -225,12 +225,12 @@ static int __init efm32_clockevent_init(struct device_node *np) clock_event_ddata.base = base; clock_event_ddata.periodic_top = DIV_ROUND_CLOSEST(rate, 1024 * HZ); - setup_irq(irq, &efm32_clock_event_irq); - clockevents_config_and_register(&clock_event_ddata.evtdev, DIV_ROUND_CLOSEST(rate, 1024), 0xf, 0xffff); + setup_irq(irq, &efm32_clock_event_irq); + return 0; err_get_irq: -- cgit v1.2.1 From 1096be084ac59927158ce80ff1d31c33eed0e565 Mon Sep 17 00:00:00 2001 From: Yongbae Park Date: Tue, 3 Mar 2015 13:05:48 +0900 Subject: clockevents: sun5i: Fix setup_irq init sequence The interrupt is enabled before the handler is set. Even this bug did not appear, it is potentially dangerous as it can lead to a NULL pointer dereference. Fix the error by enabling the interrupt after clockevents_config_and_register() is called. Cc: stable@vger.kernel.org Signed-off-by: Yongbae Park Signed-off-by: Daniel Lezcano --- drivers/clocksource/timer-sun5i.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/clocksource/timer-sun5i.c b/drivers/clocksource/timer-sun5i.c index 02268448dc85..5dcbf90b8015 100644 --- a/drivers/clocksource/timer-sun5i.c +++ b/drivers/clocksource/timer-sun5i.c @@ -178,10 +178,6 @@ static void __init sun5i_timer_init(struct device_node *node) ticks_per_jiffy = DIV_ROUND_UP(rate, HZ); - ret = setup_irq(irq, &sun5i_timer_irq); - if (ret) - pr_warn("failed to setup irq %d\n", irq); - /* Enable timer0 interrupt */ val = readl(timer_base + TIMER_IRQ_EN_REG); writel(val | TIMER_IRQ_EN(0), timer_base + TIMER_IRQ_EN_REG); @@ -191,6 +187,10 @@ static void __init sun5i_timer_init(struct device_node *node) clockevents_config_and_register(&sun5i_clockevent, rate, TIMER_SYNC_TICKS, 0xffffffff); + + ret = setup_irq(irq, &sun5i_timer_irq); + if (ret) + pr_warn("failed to setup irq %d\n", irq); } CLOCKSOURCE_OF_DECLARE(sun5i_a13, "allwinner,sun5i-a13-hstimer", sun5i_timer_init); -- cgit v1.2.1 From 3995614d9b0320e10ce202836c8477e1bcf1a2d4 Mon Sep 17 00:00:00 2001 From: Arnaldo Carvalho de Melo Date: Thu, 5 Mar 2015 15:27:28 -0300 Subject: perf annotate: Fix fallback to unparsed disassembler line When annotating source/disasm lines the perf tools parse the output of objdump, trying to provide augmented output that allows navigating jumps, calls, etc. But when a line output by objdump can't be parsed the annotation code falls back to just presenting the unparsed line. When fixing a leak in the 0fb9f2aab738 commit ("perf annotate: Fix memory leaks in LOCK handling") we failed to take that into account and instead tried to free one of the data structures that should be freed only when successfully allocated, oops, segfault. There was a change in the way the objdump output for lock prefixed instructions is formatted that lead the relevant parser to fail to grok it. At least RHEL7 works ok, but Fedora 20 segfaults. Fix it by making the ins__delete() destructor work like the most basic destructor: free(). Namely make it accept a NULL pointer and when handling it just do nothing. Further investigation is needed to figure out the nature of the objdump output change so as to make the parser grok it. Reported-by: Linus Torvalds Cc: David Ahern Cc: Jiri Olsa Cc: Namhyung Kim Cc: Paul Mackerras Cc: Peter Zijlstra Cc: Rabin Vincent Link: http://lkml.kernel.org/n/tip-7wsy0zo292pif0yjoqpfryrz@git.kernel.org Signed-off-by: Arnaldo Carvalho de Melo --- tools/perf/util/annotate.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/tools/perf/util/annotate.c b/tools/perf/util/annotate.c index 61bf9128e1f2..9d9db3b296dd 100644 --- a/tools/perf/util/annotate.c +++ b/tools/perf/util/annotate.c @@ -30,6 +30,8 @@ static int disasm_line__parse(char *line, char **namep, char **rawp); static void ins__delete(struct ins_operands *ops) { + if (ops == NULL) + return; zfree(&ops->source.raw); zfree(&ops->source.name); zfree(&ops->target.raw); -- cgit v1.2.1 From e893286918d2cde3a94850d8f7101cd1039e0c62 Mon Sep 17 00:00:00 2001 From: Jiri Slaby Date: Thu, 5 Mar 2015 09:13:31 +0100 Subject: x86/vdso: Fix the build on GCC5 On gcc5 the kernel does not link: ld: .eh_frame_hdr table[4] FDE at 0000000000000648 overlaps table[5] FDE at 0000000000000670. Because prior GCC versions always emitted NOPs on ALIGN directives, but gcc5 started omitting them. .LSTARTFDEDLSI1 says: /* HACK: The dwarf2 unwind routines will subtract 1 from the return address to get an address in the middle of the presumed call instruction. Since we didn't get here via a call, we need to include the nop before the real start to make up for it. */ .long .LSTART_sigreturn-1-. /* PC-relative start address */ But commit 69d0627a7f6e ("x86 vDSO: reorder vdso32 code") from 2.6.25 replaced .org __kernel_vsyscall+32,0x90 by ALIGN right before __kernel_sigreturn. Of course, ALIGN need not generate any NOP in there. Esp. gcc5 collapses vclock_gettime.o and int80.o together with no generated NOPs as "ALIGN". So fix this by adding to that point at least a single NOP and make the function ALIGN possibly with more NOPs then. Kudos for reporting and diagnosing should go to Richard. Reported-by: Richard Biener Signed-off-by: Jiri Slaby Acked-by: Andy Lutomirski Cc: Cc: Borislav Petkov Cc: H. Peter Anvin Cc: Linus Torvalds Cc: Thomas Gleixner Link: http://lkml.kernel.org/r/1425543211-12542-1-git-send-email-jslaby@suse.cz Signed-off-by: Ingo Molnar --- arch/x86/vdso/vdso32/sigreturn.S | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/x86/vdso/vdso32/sigreturn.S b/arch/x86/vdso/vdso32/sigreturn.S index 31776d0efc8c..d7ec4e251c0a 100644 --- a/arch/x86/vdso/vdso32/sigreturn.S +++ b/arch/x86/vdso/vdso32/sigreturn.S @@ -17,6 +17,7 @@ .text .globl __kernel_sigreturn .type __kernel_sigreturn,@function + nop /* this guy is needed for .LSTARTFDEDLSI1 below (watch for HACK) */ ALIGN __kernel_sigreturn: .LSTART_sigreturn: -- cgit v1.2.1 From f8323b6bb2cc7d26941d4838dd4375952980a88a Mon Sep 17 00:00:00 2001 From: Mika Westerberg Date: Mon, 23 Feb 2015 14:53:10 +0200 Subject: pinctrl: baytrail: Relax GPIO request rules Zotac ZBOX PI320, a Baytrail based mini-PC, has power button connected to a GPIO pin and it is exposed to the operating system as Windows 8 button array. This is implemented in Linux as a driver using gpio_keys. However, BIOS on this particula machine forgot to mux the pin to be a GPIO instead of native function, which results following message to be seen on the console: byt_gpio INT33FC:02: pin 16 cannot be used as GPIO. This causes power button to not work as the driver was not able to request the GPIO it needs. So instead of completely preventing this we allow turning the pin as GPIO but issue warning that something might be wrong. Reported-by: Benjamin Adler Signed-off-by: Mika Westerberg Signed-off-by: Linus Walleij --- drivers/pinctrl/intel/pinctrl-baytrail.c | 35 ++++++++++++++++++++------------ 1 file changed, 22 insertions(+), 13 deletions(-) diff --git a/drivers/pinctrl/intel/pinctrl-baytrail.c b/drivers/pinctrl/intel/pinctrl-baytrail.c index 5afe03e28b91..e44f2fd6753f 100644 --- a/drivers/pinctrl/intel/pinctrl-baytrail.c +++ b/drivers/pinctrl/intel/pinctrl-baytrail.c @@ -158,40 +158,49 @@ static void __iomem *byt_gpio_reg(struct gpio_chip *chip, unsigned offset, return vg->reg_base + reg_offset + reg; } -static bool is_special_pin(struct byt_gpio *vg, unsigned offset) +static u32 byt_get_gpio_mux(struct byt_gpio *vg, unsigned offset) { /* SCORE pin 92-93 */ if (!strcmp(vg->range->name, BYT_SCORE_ACPI_UID) && offset >= 92 && offset <= 93) - return true; + return 1; /* SUS pin 11-21 */ if (!strcmp(vg->range->name, BYT_SUS_ACPI_UID) && offset >= 11 && offset <= 21) - return true; + return 1; - return false; + return 0; } static int byt_gpio_request(struct gpio_chip *chip, unsigned offset) { struct byt_gpio *vg = to_byt_gpio(chip); void __iomem *reg = byt_gpio_reg(chip, offset, BYT_CONF0_REG); - u32 value; - bool special; + u32 value, gpio_mux; /* * In most cases, func pin mux 000 means GPIO function. * But, some pins may have func pin mux 001 represents - * GPIO function. Only allow user to export pin with - * func pin mux preset as GPIO function by BIOS/FW. + * GPIO function. + * + * Because there are devices out there where some pins were not + * configured correctly we allow changing the mux value from + * request (but print out warning about that). */ value = readl(reg) & BYT_PIN_MUX; - special = is_special_pin(vg, offset); - if ((special && value != 1) || (!special && value)) { - dev_err(&vg->pdev->dev, - "pin %u cannot be used as GPIO.\n", offset); - return -EINVAL; + gpio_mux = byt_get_gpio_mux(vg, offset); + if (WARN_ON(gpio_mux != value)) { + unsigned long flags; + + spin_lock_irqsave(&vg->lock, flags); + value = readl(reg) & ~BYT_PIN_MUX; + value |= gpio_mux; + writel(value, reg); + spin_unlock_irqrestore(&vg->lock, flags); + + dev_warn(&vg->pdev->dev, + "pin %u forcibly re-configured as GPIO\n", offset); } pm_runtime_get(&vg->pdev->dev); -- cgit v1.2.1 From 95f0972c7e4cbf3fc68160131c5ac2f033481d00 Mon Sep 17 00:00:00 2001 From: Mika Westerberg Date: Mon, 23 Feb 2015 14:53:11 +0200 Subject: pinctrl: baytrail: Clear interrupt triggering from pins that are in GPIO mode If the pin is already configured as GPIO and it has any of the triggering flags set, we may get spurious interrupts depending on the state of the pin. Prevent this by clearing the triggering flags on such pins. However, if the pin is also configured as "direct IRQ" we leave the flags as is. Otherwise it will prevent interrupts that are routed directly to IO-APIC. Signed-off-by: Mika Westerberg Signed-off-by: Linus Walleij --- drivers/pinctrl/intel/pinctrl-baytrail.c | 36 +++++++++++++++++++++++++------- 1 file changed, 29 insertions(+), 7 deletions(-) diff --git a/drivers/pinctrl/intel/pinctrl-baytrail.c b/drivers/pinctrl/intel/pinctrl-baytrail.c index e44f2fd6753f..d264b099182d 100644 --- a/drivers/pinctrl/intel/pinctrl-baytrail.c +++ b/drivers/pinctrl/intel/pinctrl-baytrail.c @@ -158,6 +158,19 @@ static void __iomem *byt_gpio_reg(struct gpio_chip *chip, unsigned offset, return vg->reg_base + reg_offset + reg; } +static void byt_gpio_clear_triggering(struct byt_gpio *vg, unsigned offset) +{ + void __iomem *reg = byt_gpio_reg(&vg->chip, offset, BYT_CONF0_REG); + unsigned long flags; + u32 value; + + spin_lock_irqsave(&vg->lock, flags); + value = readl(reg); + value &= ~(BYT_TRIG_POS | BYT_TRIG_NEG | BYT_TRIG_LVL); + writel(value, reg); + spin_unlock_irqrestore(&vg->lock, flags); +} + static u32 byt_get_gpio_mux(struct byt_gpio *vg, unsigned offset) { /* SCORE pin 92-93 */ @@ -211,14 +224,8 @@ static int byt_gpio_request(struct gpio_chip *chip, unsigned offset) static void byt_gpio_free(struct gpio_chip *chip, unsigned offset) { struct byt_gpio *vg = to_byt_gpio(chip); - void __iomem *reg = byt_gpio_reg(&vg->chip, offset, BYT_CONF0_REG); - u32 value; - - /* clear interrupt triggering */ - value = readl(reg); - value &= ~(BYT_TRIG_POS | BYT_TRIG_NEG | BYT_TRIG_LVL); - writel(value, reg); + byt_gpio_clear_triggering(vg, offset); pm_runtime_put(&vg->pdev->dev); } @@ -481,6 +488,21 @@ static void byt_gpio_irq_init_hw(struct byt_gpio *vg) { void __iomem *reg; u32 base, value; + int i; + + /* + * Clear interrupt triggers for all pins that are GPIOs and + * do not use direct IRQ mode. This will prevent spurious + * interrupts from misconfigured pins. + */ + for (i = 0; i < vg->chip.ngpio; i++) { + value = readl(byt_gpio_reg(&vg->chip, i, BYT_CONF0_REG)); + if ((value & BYT_PIN_MUX) == byt_get_gpio_mux(vg, i) && + !(value & BYT_DIRECT_IRQ_EN)) { + byt_gpio_clear_triggering(vg, i); + dev_dbg(&vg->pdev->dev, "disabling GPIO %d\n", i); + } + } /* clear interrupt status trigger registers */ for (base = 0; base < vg->chip.ngpio; base += 32) { -- cgit v1.2.1 From 31e4329f99062a06dca5a493bb4495a63b2dc6ba Mon Sep 17 00:00:00 2001 From: Mika Westerberg Date: Mon, 23 Feb 2015 14:53:12 +0200 Subject: pinctrl: baytrail: Rework interrupt handling Instead of handling everything in the driver's first level interrupt handler, we can take advantage of already existing flow handlers that are provided by the IRQ core. This changes the functionality a bit also. Previously the driver looped over pending interrupts in a single loop, restarting the loop if some interrupt changed state. This caused problem with Lenovo Thinkpad 10 digitizer that it was not able to deassert the interrupt before the driver disabled the interrupt for good (looplimit was exhausted). Rework the interrupt handling logic a bit so that we provide proper mask, ack and unmask operations in terms of Baytrail GPIO hardware and loop over pending interrupts only once. If the interrupt remains asserted the first level handler will be re-triggered automatically. Signed-off-by: Mika Westerberg Signed-off-by: Linus Walleij --- drivers/pinctrl/intel/pinctrl-baytrail.c | 100 +++++++++++++++++-------------- 1 file changed, 56 insertions(+), 44 deletions(-) diff --git a/drivers/pinctrl/intel/pinctrl-baytrail.c b/drivers/pinctrl/intel/pinctrl-baytrail.c index d264b099182d..2318057a309b 100644 --- a/drivers/pinctrl/intel/pinctrl-baytrail.c +++ b/drivers/pinctrl/intel/pinctrl-baytrail.c @@ -252,23 +252,13 @@ static int byt_irq_type(struct irq_data *d, unsigned type) value &= ~(BYT_DIRECT_IRQ_EN | BYT_TRIG_POS | BYT_TRIG_NEG | BYT_TRIG_LVL); - switch (type) { - case IRQ_TYPE_LEVEL_HIGH: - value |= BYT_TRIG_LVL; - case IRQ_TYPE_EDGE_RISING: - value |= BYT_TRIG_POS; - break; - case IRQ_TYPE_LEVEL_LOW: - value |= BYT_TRIG_LVL; - case IRQ_TYPE_EDGE_FALLING: - value |= BYT_TRIG_NEG; - break; - case IRQ_TYPE_EDGE_BOTH: - value |= (BYT_TRIG_NEG | BYT_TRIG_POS); - break; - } writel(value, reg); + if (type & IRQ_TYPE_EDGE_BOTH) + __irq_set_handler_locked(d->irq, handle_edge_irq); + else if (type & IRQ_TYPE_LEVEL_MASK) + __irq_set_handler_locked(d->irq, handle_level_irq); + spin_unlock_irqrestore(&vg->lock, flags); return 0; @@ -426,58 +416,80 @@ static void byt_gpio_irq_handler(unsigned irq, struct irq_desc *desc) struct irq_data *data = irq_desc_get_irq_data(desc); struct byt_gpio *vg = to_byt_gpio(irq_desc_get_handler_data(desc)); struct irq_chip *chip = irq_data_get_irq_chip(data); - u32 base, pin, mask; + u32 base, pin; void __iomem *reg; - u32 pending; + unsigned long pending; unsigned virq; - int looplimit = 0; /* check from GPIO controller which pin triggered the interrupt */ for (base = 0; base < vg->chip.ngpio; base += 32) { - reg = byt_gpio_reg(&vg->chip, base, BYT_INT_STAT_REG); - - while ((pending = readl(reg))) { - pin = __ffs(pending); - mask = BIT(pin); - /* Clear before handling so we can't lose an edge */ - writel(mask, reg); - + pending = readl(reg); + for_each_set_bit(pin, &pending, 32) { virq = irq_find_mapping(vg->chip.irqdomain, base + pin); generic_handle_irq(virq); - - /* In case bios or user sets triggering incorretly a pin - * might remain in "interrupt triggered" state. - */ - if (looplimit++ > 32) { - dev_err(&vg->pdev->dev, - "Gpio %d interrupt flood, disabling\n", - base + pin); - - reg = byt_gpio_reg(&vg->chip, base + pin, - BYT_CONF0_REG); - mask = readl(reg); - mask &= ~(BYT_TRIG_NEG | BYT_TRIG_POS | - BYT_TRIG_LVL); - writel(mask, reg); - mask = readl(reg); /* flush */ - break; - } } } chip->irq_eoi(data); } +static void byt_irq_ack(struct irq_data *d) +{ + struct gpio_chip *gc = irq_data_get_irq_chip_data(d); + struct byt_gpio *vg = to_byt_gpio(gc); + unsigned offset = irqd_to_hwirq(d); + void __iomem *reg; + + reg = byt_gpio_reg(&vg->chip, offset, BYT_INT_STAT_REG); + writel(BIT(offset % 32), reg); +} + static void byt_irq_unmask(struct irq_data *d) { + struct gpio_chip *gc = irq_data_get_irq_chip_data(d); + struct byt_gpio *vg = to_byt_gpio(gc); + unsigned offset = irqd_to_hwirq(d); + unsigned long flags; + void __iomem *reg; + u32 value; + + spin_lock_irqsave(&vg->lock, flags); + + reg = byt_gpio_reg(&vg->chip, offset, BYT_CONF0_REG); + value = readl(reg); + + switch (irqd_get_trigger_type(d)) { + case IRQ_TYPE_LEVEL_HIGH: + value |= BYT_TRIG_LVL; + case IRQ_TYPE_EDGE_RISING: + value |= BYT_TRIG_POS; + break; + case IRQ_TYPE_LEVEL_LOW: + value |= BYT_TRIG_LVL; + case IRQ_TYPE_EDGE_FALLING: + value |= BYT_TRIG_NEG; + break; + case IRQ_TYPE_EDGE_BOTH: + value |= (BYT_TRIG_NEG | BYT_TRIG_POS); + break; + } + + writel(value, reg); + + spin_unlock_irqrestore(&vg->lock, flags); } static void byt_irq_mask(struct irq_data *d) { + struct gpio_chip *gc = irq_data_get_irq_chip_data(d); + struct byt_gpio *vg = to_byt_gpio(gc); + + byt_gpio_clear_triggering(vg, irqd_to_hwirq(d)); } static struct irq_chip byt_irqchip = { .name = "BYT-GPIO", + .irq_ack = byt_irq_ack, .irq_mask = byt_irq_mask, .irq_unmask = byt_irq_unmask, .irq_set_type = byt_irq_type, -- cgit v1.2.1 From fcc18deb7682dafcf6176b4af81d1554ffabd8b1 Mon Sep 17 00:00:00 2001 From: Mika Westerberg Date: Mon, 23 Feb 2015 14:53:13 +0200 Subject: pinctrl: baytrail: Save pin context over system sleep The BIOS might reconfigure pins as it needs when S3 is entered. This might cause drivers using the GPIOs to fail because the state was wrong or interrupts stopped working. Fix this by saving and restoring enough pin context over system sleep. Reported-by: Hans Holmberg Signed-off-by: Mika Westerberg Signed-off-by: Linus Walleij --- drivers/pinctrl/intel/pinctrl-baytrail.c | 83 +++++++++++++++++++++++++++++++- 1 file changed, 81 insertions(+), 2 deletions(-) diff --git a/drivers/pinctrl/intel/pinctrl-baytrail.c b/drivers/pinctrl/intel/pinctrl-baytrail.c index 2318057a309b..2062c224e32f 100644 --- a/drivers/pinctrl/intel/pinctrl-baytrail.c +++ b/drivers/pinctrl/intel/pinctrl-baytrail.c @@ -66,6 +66,10 @@ #define BYT_DIR_MASK (BIT(1) | BIT(2)) #define BYT_TRIG_MASK (BIT(26) | BIT(25) | BIT(24)) +#define BYT_CONF0_RESTORE_MASK (BYT_DIRECT_IRQ_EN | BYT_TRIG_MASK | \ + BYT_PIN_MUX) +#define BYT_VAL_RESTORE_MASK (BYT_DIR_MASK | BYT_LEVEL) + #define BYT_NGPIO_SCORE 102 #define BYT_NGPIO_NCORE 28 #define BYT_NGPIO_SUS 44 @@ -134,12 +138,18 @@ static struct pinctrl_gpio_range byt_ranges[] = { }, }; +struct byt_gpio_pin_context { + u32 conf0; + u32 val; +}; + struct byt_gpio { struct gpio_chip chip; struct platform_device *pdev; spinlock_t lock; void __iomem *reg_base; struct pinctrl_gpio_range *range; + struct byt_gpio_pin_context *saved_context; }; #define to_byt_gpio(c) container_of(c, struct byt_gpio, chip) @@ -584,6 +594,11 @@ static int byt_gpio_probe(struct platform_device *pdev) gc->can_sleep = false; gc->dev = dev; +#ifdef CONFIG_PM_SLEEP + vg->saved_context = devm_kcalloc(&pdev->dev, gc->ngpio, + sizeof(*vg->saved_context), GFP_KERNEL); +#endif + ret = gpiochip_add(gc); if (ret) { dev_err(&pdev->dev, "failed adding byt-gpio chip\n"); @@ -612,6 +627,69 @@ static int byt_gpio_probe(struct platform_device *pdev) return 0; } +#ifdef CONFIG_PM_SLEEP +static int byt_gpio_suspend(struct device *dev) +{ + struct platform_device *pdev = to_platform_device(dev); + struct byt_gpio *vg = platform_get_drvdata(pdev); + int i; + + for (i = 0; i < vg->chip.ngpio; i++) { + void __iomem *reg; + u32 value; + + reg = byt_gpio_reg(&vg->chip, i, BYT_CONF0_REG); + value = readl(reg) & BYT_CONF0_RESTORE_MASK; + vg->saved_context[i].conf0 = value; + + reg = byt_gpio_reg(&vg->chip, i, BYT_VAL_REG); + value = readl(reg) & BYT_VAL_RESTORE_MASK; + vg->saved_context[i].val = value; + } + + return 0; +} + +static int byt_gpio_resume(struct device *dev) +{ + struct platform_device *pdev = to_platform_device(dev); + struct byt_gpio *vg = platform_get_drvdata(pdev); + int i; + + for (i = 0; i < vg->chip.ngpio; i++) { + void __iomem *reg; + u32 value; + + reg = byt_gpio_reg(&vg->chip, i, BYT_CONF0_REG); + value = readl(reg); + if ((value & BYT_CONF0_RESTORE_MASK) != + vg->saved_context[i].conf0) { + value &= ~BYT_CONF0_RESTORE_MASK; + value |= vg->saved_context[i].conf0; + writel(value, reg); + dev_info(dev, "restored pin %d conf0 %#08x", i, value); + } + + reg = byt_gpio_reg(&vg->chip, i, BYT_VAL_REG); + value = readl(reg); + if ((value & BYT_VAL_RESTORE_MASK) != + vg->saved_context[i].val) { + u32 v; + + v = value & ~BYT_VAL_RESTORE_MASK; + v |= vg->saved_context[i].val; + if (v != value) { + writel(v, reg); + dev_dbg(dev, "restored pin %d val %#08x\n", + i, v); + } + } + } + + return 0; +} +#endif + static int byt_gpio_runtime_suspend(struct device *dev) { return 0; @@ -623,8 +701,9 @@ static int byt_gpio_runtime_resume(struct device *dev) } static const struct dev_pm_ops byt_gpio_pm_ops = { - .runtime_suspend = byt_gpio_runtime_suspend, - .runtime_resume = byt_gpio_runtime_resume, + SET_LATE_SYSTEM_SLEEP_PM_OPS(byt_gpio_suspend, byt_gpio_resume) + SET_RUNTIME_PM_OPS(byt_gpio_runtime_suspend, byt_gpio_runtime_resume, + NULL) }; static const struct acpi_device_id byt_gpio_acpi_match[] = { -- cgit v1.2.1 From c7d910b87d3c8e9fcf4077089ca4327c12eee099 Mon Sep 17 00:00:00 2001 From: Eric Nelson Date: Fri, 27 Feb 2015 08:06:45 -0700 Subject: ASoC: sgtl5000: remove useless register write clearing CHRGPUMP_POWERUP The SGTL5000_CHIP_ANA_POWER register is cached. Update the cached value instead of writing it directly. Patch inspired by Russell King's more colorful remarks in this patch: https://github.com/SolidRun/linux-imx6-3.14/commit/dd4bf6a Signed-off-by: Eric Nelson Signed-off-by: Mark Brown Cc: stable@vger.kernel.org --- sound/soc/codecs/sgtl5000.c | 8 +------- 1 file changed, 1 insertion(+), 7 deletions(-) diff --git a/sound/soc/codecs/sgtl5000.c b/sound/soc/codecs/sgtl5000.c index e182e6569bbd..3593a1496056 100644 --- a/sound/soc/codecs/sgtl5000.c +++ b/sound/soc/codecs/sgtl5000.c @@ -1151,13 +1151,7 @@ static int sgtl5000_set_power_regs(struct snd_soc_codec *codec) /* Enable VDDC charge pump */ ana_pwr |= SGTL5000_VDDC_CHRGPMP_POWERUP; } else if (vddio >= 3100 && vdda >= 3100) { - /* - * if vddio and vddd > 3.1v, - * charge pump should be clean before set ana_pwr - */ - snd_soc_update_bits(codec, SGTL5000_CHIP_ANA_POWER, - SGTL5000_VDDC_CHRGPMP_POWERUP, 0); - + ana_pwr &= ~SGTL5000_VDDC_CHRGPMP_POWERUP; /* VDDC use VDDIO rail */ lreg_ctrl |= SGTL5000_VDDC_ASSN_OVRD; lreg_ctrl |= SGTL5000_VDDC_MAN_ASSN_VDDIO << -- cgit v1.2.1 From a4ee556137a5bb4b542c5023e6fead4b7cf33495 Mon Sep 17 00:00:00 2001 From: Bard Liao Date: Fri, 6 Mar 2015 10:12:58 +0800 Subject: ASoC: rt286: Change the DMI mapping for Dino The board ID will be changed between revisions. So, it is better to map it by project name. Signed-off-by: Bard Liao Signed-off-by: Mark Brown --- sound/soc/codecs/rt286.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/codecs/rt286.c b/sound/soc/codecs/rt286.c index f374840a5a7c..9b541e52da8c 100644 --- a/sound/soc/codecs/rt286.c +++ b/sound/soc/codecs/rt286.c @@ -1198,7 +1198,7 @@ static struct dmi_system_id dmi_dell_dino[] = { .ident = "Dell Dino", .matches = { DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."), - DMI_MATCH(DMI_BOARD_NAME, "0144P8") + DMI_MATCH(DMI_PRODUCT_NAME, "XPS 13 9343") } }, { } -- cgit v1.2.1 From 3fe0607a04ed7deea7c048052fd63b8670e7a176 Mon Sep 17 00:00:00 2001 From: "Lu, Han" Date: Wed, 25 Feb 2015 08:26:21 +0800 Subject: ASoC: Intel: remove conflicts when load/unload multiple firmware images Details: 1. Unload all modules on fw_list of dsp when suspend, and reload all modules on fw_list when resume. 2. A DSP expects only one scratch, but hsw_parse_fw_image() allocates scratch blocks for each firmware image it parses. Move the allocate function sst_block_alloc_scratch() out of hsw_parse_fw_image() to make sure a scratch be allocated only after all firmware images be parsed. Signed-off-by: Lu, Han Signed-off-by: Mark Brown --- sound/soc/intel/sst-haswell-dsp.c | 3 --- sound/soc/intel/sst-haswell-ipc.c | 32 ++++++++++++++++++++++++-------- 2 files changed, 24 insertions(+), 11 deletions(-) diff --git a/sound/soc/intel/sst-haswell-dsp.c b/sound/soc/intel/sst-haswell-dsp.c index 57039b00efc2..f6e1e6b2b18e 100644 --- a/sound/soc/intel/sst-haswell-dsp.c +++ b/sound/soc/intel/sst-haswell-dsp.c @@ -207,9 +207,6 @@ static int hsw_parse_fw_image(struct sst_fw *sst_fw) module = (void *)module + sizeof(*module) + module->mod_size; } - /* allocate scratch mem regions */ - sst_block_alloc_scratch(dsp); - return 0; } diff --git a/sound/soc/intel/sst-haswell-ipc.c b/sound/soc/intel/sst-haswell-ipc.c index 8156cc1accb7..6c7052a40e36 100644 --- a/sound/soc/intel/sst-haswell-ipc.c +++ b/sound/soc/intel/sst-haswell-ipc.c @@ -1870,6 +1870,7 @@ static void sst_hsw_drop_all(struct sst_hsw *hsw) int sst_hsw_dsp_load(struct sst_hsw *hsw) { struct sst_dsp *dsp = hsw->dsp; + struct sst_fw *sst_fw, *t; int ret; dev_dbg(hsw->dev, "loading audio DSP...."); @@ -1886,12 +1887,17 @@ int sst_hsw_dsp_load(struct sst_hsw *hsw) return ret; } - ret = sst_fw_reload(hsw->sst_fw); - if (ret < 0) { - dev_err(hsw->dev, "error: SST FW reload failed\n"); - sst_dsp_dma_put_channel(dsp); - return -ENOMEM; + list_for_each_entry_safe_reverse(sst_fw, t, &dsp->fw_list, list) { + ret = sst_fw_reload(sst_fw); + if (ret < 0) { + dev_err(hsw->dev, "error: SST FW reload failed\n"); + sst_dsp_dma_put_channel(dsp); + return -ENOMEM; + } } + ret = sst_block_alloc_scratch(hsw->dsp); + if (ret < 0) + return -EINVAL; sst_dsp_dma_put_channel(dsp); return 0; @@ -1947,12 +1953,17 @@ int sst_hsw_dsp_runtime_suspend(struct sst_hsw *hsw) int sst_hsw_dsp_runtime_sleep(struct sst_hsw *hsw) { - sst_fw_unload(hsw->sst_fw); - sst_block_free_scratch(hsw->dsp); + struct sst_fw *sst_fw, *t; + struct sst_dsp *dsp = hsw->dsp; + + list_for_each_entry_safe(sst_fw, t, &dsp->fw_list, list) { + sst_fw_unload(sst_fw); + } + sst_block_free_scratch(dsp); hsw->boot_complete = false; - sst_dsp_sleep(hsw->dsp); + sst_dsp_sleep(dsp); return 0; } @@ -2081,6 +2092,11 @@ int sst_hsw_dsp_init(struct device *dev, struct sst_pdata *pdata) goto fw_err; } + /* allocate scratch mem regions */ + ret = sst_block_alloc_scratch(hsw->dsp); + if (ret < 0) + goto boot_err; + /* wait for DSP boot completion */ sst_dsp_boot(hsw->dsp); ret = wait_event_timeout(hsw->boot_wait, hsw->boot_complete, -- cgit v1.2.1 From 3f1615340acea54e21f4b9d4d65921540dca84b2 Mon Sep 17 00:00:00 2001 From: Pontus Fuchs Date: Fri, 6 Mar 2015 16:18:41 +0100 Subject: brcmfmac: Perform bound checking on vendor command buffer A short or malformed vendor command buffer could cause reads outside the command buffer. Cc: stable@vger.kernel.org # v3.19 Signed-off-by: Pontus Fuchs [arend@broadcom.com: slightly modified debug trace output] Signed-off-by: Arend van Spriel Signed-off-by: Kalle Valo --- drivers/net/wireless/brcm80211/brcmfmac/vendor.c | 15 ++++++++++++--- 1 file changed, 12 insertions(+), 3 deletions(-) diff --git a/drivers/net/wireless/brcm80211/brcmfmac/vendor.c b/drivers/net/wireless/brcm80211/brcmfmac/vendor.c index 50cdf7090198..8eff2753abad 100644 --- a/drivers/net/wireless/brcm80211/brcmfmac/vendor.c +++ b/drivers/net/wireless/brcm80211/brcmfmac/vendor.c @@ -39,13 +39,22 @@ static int brcmf_cfg80211_vndr_cmds_dcmd_handler(struct wiphy *wiphy, void *dcmd_buf = NULL, *wr_pointer; u16 msglen, maxmsglen = PAGE_SIZE - 0x100; - brcmf_dbg(TRACE, "cmd %x set %d len %d\n", cmdhdr->cmd, cmdhdr->set, - cmdhdr->len); + if (len < sizeof(*cmdhdr)) { + brcmf_err("vendor command too short: %d\n", len); + return -EINVAL; + } vif = container_of(wdev, struct brcmf_cfg80211_vif, wdev); ifp = vif->ifp; - len -= sizeof(struct brcmf_vndr_dcmd_hdr); + brcmf_dbg(TRACE, "ifidx=%d, cmd=%d\n", ifp->ifidx, cmdhdr->cmd); + + if (cmdhdr->offset > len) { + brcmf_err("bad buffer offset %d > %d\n", cmdhdr->offset, len); + return -EINVAL; + } + + len -= cmdhdr->offset; ret_len = cmdhdr->len; if (ret_len > 0 || len > 0) { if (len > BRCMF_DCMD_MAXLEN) { -- cgit v1.2.1 From 12cb89e37a0c25fae7a0f1d2e4985558db9d0b13 Mon Sep 17 00:00:00 2001 From: "Ivan T. Ivanov" Date: Fri, 6 Mar 2015 17:26:17 +0200 Subject: spi: qup: Fix cs-num DT property parsing num-cs is 32 bit property, don't read just upper 16 bits. Fixes: 4a8573abe965 (spi: qup: Remove chip select function) Signed-off-by: Ivan T. Ivanov Signed-off-by: Mark Brown Cc: stable@vger.kernel.org --- drivers/spi/spi-qup.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/spi/spi-qup.c b/drivers/spi/spi-qup.c index ff9cdbdb6672..2b2c359f5a50 100644 --- a/drivers/spi/spi-qup.c +++ b/drivers/spi/spi-qup.c @@ -498,7 +498,7 @@ static int spi_qup_probe(struct platform_device *pdev) struct resource *res; struct device *dev; void __iomem *base; - u32 max_freq, iomode; + u32 max_freq, iomode, num_cs; int ret, irq, size; dev = &pdev->dev; @@ -550,10 +550,11 @@ static int spi_qup_probe(struct platform_device *pdev) } /* use num-cs unless not present or out of range */ - if (of_property_read_u16(dev->of_node, "num-cs", - &master->num_chipselect) || - (master->num_chipselect > SPI_NUM_CHIPSELECTS)) + if (of_property_read_u32(dev->of_node, "num-cs", &num_cs) || + num_cs > SPI_NUM_CHIPSELECTS) master->num_chipselect = SPI_NUM_CHIPSELECTS; + else + master->num_chipselect = num_cs; master->bus_num = pdev->id; master->mode_bits = SPI_CPOL | SPI_CPHA | SPI_CS_HIGH | SPI_LOOP; -- cgit v1.2.1 From 854d2f241d71f6ca08ccde30e6c7c2e403363e52 Mon Sep 17 00:00:00 2001 From: Andy Shevchenko Date: Fri, 6 Mar 2015 14:42:01 +0200 Subject: spi: dw-mid: clear BUSY flag fist and test other one The logic of DMA completion is broken now since test_and_clear_bit() never returns the other bit is set. It means condition are always false and we have spi_finalize_current_transfer() called per each DMA completion which is wrong. The patch fixes logic by clearing BUSY bit first and then check for the other one. Fixes: 30c8eb52cc4a (spi: dw-mid: split rx and tx callbacks when DMA) Signed-off-by: Andy Shevchenko Signed-off-by: Mark Brown Cc: stable@vger.kernel.org --- drivers/spi/spi-dw-mid.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/spi/spi-dw-mid.c b/drivers/spi/spi-dw-mid.c index 3ce39d10fafb..4f8c798e0633 100644 --- a/drivers/spi/spi-dw-mid.c +++ b/drivers/spi/spi-dw-mid.c @@ -108,7 +108,8 @@ static void dw_spi_dma_tx_done(void *arg) { struct dw_spi *dws = arg; - if (test_and_clear_bit(TX_BUSY, &dws->dma_chan_busy) & BIT(RX_BUSY)) + clear_bit(TX_BUSY, &dws->dma_chan_busy); + if (test_bit(RX_BUSY, &dws->dma_chan_busy)) return; dw_spi_xfer_done(dws); } @@ -156,7 +157,8 @@ static void dw_spi_dma_rx_done(void *arg) { struct dw_spi *dws = arg; - if (test_and_clear_bit(RX_BUSY, &dws->dma_chan_busy) & BIT(TX_BUSY)) + clear_bit(RX_BUSY, &dws->dma_chan_busy); + if (test_bit(TX_BUSY, &dws->dma_chan_busy)) return; dw_spi_xfer_done(dws); } -- cgit v1.2.1 From 328f494d95aac8bd4896aea2328bc281053bcb71 Mon Sep 17 00:00:00 2001 From: Lars-Peter Clausen Date: Sat, 7 Mar 2015 17:10:01 +0100 Subject: regmap: regcache-rbtree: Fix present bitmap resize When inserting a new register into a block at the lower end the present bitmap is currently shifted into the wrong direction. The effect of this is that the bitmap becomes corrupted and registers which are present might be reported as not present and vice versa. Fix this by shifting left rather than right. Fixes: 472fdec7380c("regmap: rbtree: Reduce number of nodes, take 2") Reported-by: Daniel Baluta Signed-off-by: Lars-Peter Clausen Signed-off-by: Mark Brown Cc: stable@vger.kernel.org --- drivers/base/regmap/regcache-rbtree.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/base/regmap/regcache-rbtree.c b/drivers/base/regmap/regcache-rbtree.c index d453a2c98ad0..81751a49d8bf 100644 --- a/drivers/base/regmap/regcache-rbtree.c +++ b/drivers/base/regmap/regcache-rbtree.c @@ -307,7 +307,7 @@ static int regcache_rbtree_insert_to_block(struct regmap *map, if (pos == 0) { memmove(blk + offset * map->cache_word_size, blk, rbnode->blklen * map->cache_word_size); - bitmap_shift_right(present, present, offset, blklen); + bitmap_shift_left(present, present, offset, blklen); } /* update the rbnode block, its size and the base register */ -- cgit v1.2.1 From 0548bf4f5ad6fc3bd93c4940fa48078b34609682 Mon Sep 17 00:00:00 2001 From: Javier Martinez Canillas Date: Mon, 2 Mar 2015 21:40:39 +0100 Subject: regulator: Only enable disabled regulators on resume The _regulator_do_enable() call ought to be a no-op when called on an already-enabled regulator. However, as an optimization _regulator_enable() doesn't call _regulator_do_enable() on an already enabled regulator. That means we never test the case of calling _regulator_do_enable() during normal usage and there may be hidden bugs or warnings. We have seen warnings issued by the tps65090 driver and bugs when using the GPIO enable pin. Let's match the same optimization that _regulator_enable() in regulator_suspend_finish(). That may speed up suspend/resume and also avoids exposing hidden bugs. [Use much clearer commit message from Doug Anderson] Signed-off-by: Javier Martinez Canillas Signed-off-by: Mark Brown Cc: stable@vger.kernel.org --- drivers/regulator/core.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/drivers/regulator/core.c b/drivers/regulator/core.c index b899947d839d..0e271e57504a 100644 --- a/drivers/regulator/core.c +++ b/drivers/regulator/core.c @@ -3807,9 +3807,11 @@ int regulator_suspend_finish(void) list_for_each_entry(rdev, ®ulator_list, list) { mutex_lock(&rdev->mutex); if (rdev->use_count > 0 || rdev->constraints->always_on) { - error = _regulator_do_enable(rdev); - if (error) - ret = error; + if (!_regulator_is_enabled(rdev)) { + error = _regulator_do_enable(rdev); + if (error) + ret = error; + } } else { if (!have_full_constraints()) goto unlock; -- cgit v1.2.1 From 29d62ec5f87fbeec8413e2215ddad12e7f972e4c Mon Sep 17 00:00:00 2001 From: Doug Anderson Date: Tue, 3 Mar 2015 15:20:47 -0800 Subject: regulator: core: Fix enable GPIO reference counting Normally _regulator_do_enable() isn't called on an already-enabled rdev. That's because the main caller, _regulator_enable() always calls _regulator_is_enabled() and only calls _regulator_do_enable() if the rdev was not already enabled. However, there is one caller of _regulator_do_enable() that doesn't check: regulator_suspend_finish(). While we might want to make regulator_suspend_finish() behave more like _regulator_enable(), it's probably also a good idea to make _regulator_do_enable() robust if it is called on an already enabled rdev. At the moment, _regulator_do_enable() is _not_ robust for already enabled rdevs if we're using an ena_pin. Each time _regulator_do_enable() is called for an rdev using an ena_pin the reference count of the ena_pin is incremented even if the rdev was already enabled. This is not as intended because the ena_pin is for something else: for keeping track of how many active rdevs there are sharing the same ena_pin. Here's how the reference counting works here: * Each time _regulator_enable() is called we increment rdev->use_count, so _regulator_enable() calls need to be balanced with _regulator_disable() calls. * There is no explicit reference counting in _regulator_do_enable() which is normally just a warapper around rdev->desc->ops->enable() with code for supporting delays. It's not expected that the "ops->enable()" call do reference counting. * Since regulator_ena_gpio_ctrl() does have reference counting (handling the sharing of the pin amongst multiple rdevs), we shouldn't call it if the current rdev is already enabled. Note that as part of this we cleanup (remove) the initting of ena_gpio_state in regulator_register(). In _regulator_do_enable(), _regulator_do_disable() and _regulator_is_enabled() is is clear that ena_gpio_state should be the state of whether this particular rdev has requested the GPIO be enabled. regulator_register() was initting it as the actual state of the pin. Fixes: 967cfb18c0e3 ("regulator: core: manage enable GPIO list") Signed-off-by: Doug Anderson Signed-off-by: Mark Brown Cc: stable@vger.kernel.org --- drivers/regulator/core.c | 26 ++++++++++++-------------- 1 file changed, 12 insertions(+), 14 deletions(-) diff --git a/drivers/regulator/core.c b/drivers/regulator/core.c index 0e271e57504a..fafeb32427c1 100644 --- a/drivers/regulator/core.c +++ b/drivers/regulator/core.c @@ -1839,10 +1839,12 @@ static int _regulator_do_enable(struct regulator_dev *rdev) } if (rdev->ena_pin) { - ret = regulator_ena_gpio_ctrl(rdev, true); - if (ret < 0) - return ret; - rdev->ena_gpio_state = 1; + if (!rdev->ena_gpio_state) { + ret = regulator_ena_gpio_ctrl(rdev, true); + if (ret < 0) + return ret; + rdev->ena_gpio_state = 1; + } } else if (rdev->desc->ops->enable) { ret = rdev->desc->ops->enable(rdev); if (ret < 0) @@ -1939,10 +1941,12 @@ static int _regulator_do_disable(struct regulator_dev *rdev) trace_regulator_disable(rdev_get_name(rdev)); if (rdev->ena_pin) { - ret = regulator_ena_gpio_ctrl(rdev, false); - if (ret < 0) - return ret; - rdev->ena_gpio_state = 0; + if (rdev->ena_gpio_state) { + ret = regulator_ena_gpio_ctrl(rdev, false); + if (ret < 0) + return ret; + rdev->ena_gpio_state = 0; + } } else if (rdev->desc->ops->disable) { ret = rdev->desc->ops->disable(rdev); @@ -3633,12 +3637,6 @@ regulator_register(const struct regulator_desc *regulator_desc, config->ena_gpio, ret); goto wash; } - - if (config->ena_gpio_flags & GPIOF_OUT_INIT_HIGH) - rdev->ena_gpio_state = 1; - - if (config->ena_gpio_invert) - rdev->ena_gpio_state = !rdev->ena_gpio_state; } /* set regulator constraints */ -- cgit v1.2.1 From 34e81ab4556f3b1371763861e74e3600818924b5 Mon Sep 17 00:00:00 2001 From: Lars-Peter Clausen Date: Sat, 7 Mar 2015 19:34:03 +0100 Subject: ASoC: Fix component lists locking Any access to the component_list, codec_list and platform_list needs to be properly locked by the client_mutex. Otherwise undefined behavior can occur if the list is modified in one thread and concurrently accessed from another thread. This patch adds the missing locking to the debugfs file handlers that display the registered components, as well as the various components unregister functions. Furthermore the client_lock is now held for the whole snd_soc_instantiate_card() sequence to make sure that component removal does not race against the card registration. Reported-by: Takashi Iwai Signed-off-by: Lars-Peter Clausen Signed-off-by: Mark Brown --- sound/soc/soc-core.c | 41 ++++++++++++++++++++++++++++++----------- 1 file changed, 30 insertions(+), 11 deletions(-) diff --git a/sound/soc/soc-core.c b/sound/soc/soc-core.c index 30579ca5bacb..e5c990889dcc 100644 --- a/sound/soc/soc-core.c +++ b/sound/soc/soc-core.c @@ -347,6 +347,8 @@ static ssize_t codec_list_read_file(struct file *file, char __user *user_buf, if (!buf) return -ENOMEM; + mutex_lock(&client_mutex); + list_for_each_entry(codec, &codec_list, list) { len = snprintf(buf + ret, PAGE_SIZE - ret, "%s\n", codec->component.name); @@ -358,6 +360,8 @@ static ssize_t codec_list_read_file(struct file *file, char __user *user_buf, } } + mutex_unlock(&client_mutex); + if (ret >= 0) ret = simple_read_from_buffer(user_buf, count, ppos, buf, ret); @@ -382,6 +386,8 @@ static ssize_t dai_list_read_file(struct file *file, char __user *user_buf, if (!buf) return -ENOMEM; + mutex_lock(&client_mutex); + list_for_each_entry(component, &component_list, list) { list_for_each_entry(dai, &component->dai_list, list) { len = snprintf(buf + ret, PAGE_SIZE - ret, "%s\n", @@ -395,6 +401,8 @@ static ssize_t dai_list_read_file(struct file *file, char __user *user_buf, } } + mutex_unlock(&client_mutex); + ret = simple_read_from_buffer(user_buf, count, ppos, buf, ret); kfree(buf); @@ -418,6 +426,8 @@ static ssize_t platform_list_read_file(struct file *file, if (!buf) return -ENOMEM; + mutex_lock(&client_mutex); + list_for_each_entry(platform, &platform_list, list) { len = snprintf(buf + ret, PAGE_SIZE - ret, "%s\n", platform->component.name); @@ -429,6 +439,8 @@ static ssize_t platform_list_read_file(struct file *file, } } + mutex_unlock(&client_mutex); + ret = simple_read_from_buffer(user_buf, count, ppos, buf, ret); kfree(buf); @@ -836,6 +848,8 @@ static struct snd_soc_component *soc_find_component( { struct snd_soc_component *component; + lockdep_assert_held(&client_mutex); + list_for_each_entry(component, &component_list, list) { if (of_node) { if (component->dev->of_node == of_node) @@ -854,6 +868,8 @@ static struct snd_soc_dai *snd_soc_find_dai( struct snd_soc_component *component; struct snd_soc_dai *dai; + lockdep_assert_held(&client_mutex); + /* Find CPU DAI from registered DAIs*/ list_for_each_entry(component, &component_list, list) { if (dlc->of_node && component->dev->of_node != dlc->of_node) @@ -1508,6 +1524,7 @@ static int snd_soc_instantiate_card(struct snd_soc_card *card) struct snd_soc_codec *codec; int ret, i, order; + mutex_lock(&client_mutex); mutex_lock_nested(&card->mutex, SND_SOC_CARD_CLASS_INIT); /* bind DAIs */ @@ -1662,6 +1679,7 @@ static int snd_soc_instantiate_card(struct snd_soc_card *card) card->instantiated = 1; snd_soc_dapm_sync(&card->dapm); mutex_unlock(&card->mutex); + mutex_unlock(&client_mutex); return 0; @@ -1680,6 +1698,7 @@ card_probe_error: base_error: mutex_unlock(&card->mutex); + mutex_unlock(&client_mutex); return ret; } @@ -2713,13 +2732,6 @@ static void snd_soc_component_del_unlocked(struct snd_soc_component *component) list_del(&component->list); } -static void snd_soc_component_del(struct snd_soc_component *component) -{ - mutex_lock(&client_mutex); - snd_soc_component_del_unlocked(component); - mutex_unlock(&client_mutex); -} - int snd_soc_register_component(struct device *dev, const struct snd_soc_component_driver *cmpnt_drv, struct snd_soc_dai_driver *dai_drv, @@ -2767,14 +2779,17 @@ void snd_soc_unregister_component(struct device *dev) { struct snd_soc_component *cmpnt; + mutex_lock(&client_mutex); list_for_each_entry(cmpnt, &component_list, list) { if (dev == cmpnt->dev && cmpnt->registered_as_component) goto found; } + mutex_unlock(&client_mutex); return; found: - snd_soc_component_del(cmpnt); + snd_soc_component_del_unlocked(cmpnt); + mutex_unlock(&client_mutex); snd_soc_component_cleanup(cmpnt); kfree(cmpnt); } @@ -2882,10 +2897,14 @@ struct snd_soc_platform *snd_soc_lookup_platform(struct device *dev) { struct snd_soc_platform *platform; + mutex_lock(&client_mutex); list_for_each_entry(platform, &platform_list, list) { - if (dev == platform->dev) + if (dev == platform->dev) { + mutex_unlock(&client_mutex); return platform; + } } + mutex_unlock(&client_mutex); return NULL; } @@ -3090,15 +3109,15 @@ void snd_soc_unregister_codec(struct device *dev) { struct snd_soc_codec *codec; + mutex_lock(&client_mutex); list_for_each_entry(codec, &codec_list, list) { if (dev == codec->dev) goto found; } + mutex_unlock(&client_mutex); return; found: - - mutex_lock(&client_mutex); list_del(&codec->list); snd_soc_component_del_unlocked(&codec->component); mutex_unlock(&client_mutex); -- cgit v1.2.1 From 8b04baba10b007f8b6c245a50be73cf09cc3a414 Mon Sep 17 00:00:00 2001 From: Daniel Martin Date: Sun, 8 Mar 2015 22:27:37 -0700 Subject: Input: synaptics - split synaptics_resolution(), query first Split the function synaptics_resolution() into synaptics_resolution() and synaptics_quirks(). synaptics_resolution() will be called before synaptics_quirks() to query dimensions and resolutions before overwriting them with quirks. Cc: stable@vger.kernel.org Signed-off-by: Daniel Martin Acked-by: Hans de Goede Signed-off-by: Dmitry Torokhov --- drivers/input/mouse/synaptics.c | 35 +++++++++++++++++++++++------------ 1 file changed, 23 insertions(+), 12 deletions(-) diff --git a/drivers/input/mouse/synaptics.c b/drivers/input/mouse/synaptics.c index 23e26e0768b5..b501dda75dcb 100644 --- a/drivers/input/mouse/synaptics.c +++ b/drivers/input/mouse/synaptics.c @@ -343,7 +343,6 @@ static int synaptics_resolution(struct psmouse *psmouse) { struct synaptics_data *priv = psmouse->private; unsigned char resp[3]; - int i; if (SYN_ID_MAJOR(priv->identity) < 4) return 0; @@ -355,17 +354,6 @@ static int synaptics_resolution(struct psmouse *psmouse) } } - for (i = 0; min_max_pnpid_table[i].pnp_ids; i++) { - if (psmouse_matches_pnp_id(psmouse, - min_max_pnpid_table[i].pnp_ids)) { - priv->x_min = min_max_pnpid_table[i].x_min; - priv->x_max = min_max_pnpid_table[i].x_max; - priv->y_min = min_max_pnpid_table[i].y_min; - priv->y_max = min_max_pnpid_table[i].y_max; - return 0; - } - } - if (SYN_EXT_CAP_REQUESTS(priv->capabilities) >= 5 && SYN_CAP_MAX_DIMENSIONS(priv->ext_cap_0c)) { if (synaptics_send_cmd(psmouse, SYN_QUE_EXT_MAX_COORDS, resp)) { @@ -391,6 +379,27 @@ static int synaptics_resolution(struct psmouse *psmouse) return 0; } +/* + * Apply quirk(s) if the hardware matches + */ + +static void synaptics_apply_quirks(struct psmouse *psmouse) +{ + struct synaptics_data *priv = psmouse->private; + int i; + + for (i = 0; min_max_pnpid_table[i].pnp_ids; i++) { + if (psmouse_matches_pnp_id(psmouse, + min_max_pnpid_table[i].pnp_ids)) { + priv->x_min = min_max_pnpid_table[i].x_min; + priv->x_max = min_max_pnpid_table[i].x_max; + priv->y_min = min_max_pnpid_table[i].y_min; + priv->y_max = min_max_pnpid_table[i].y_max; + break; + } + } +} + static int synaptics_query_hardware(struct psmouse *psmouse) { if (synaptics_identify(psmouse)) @@ -406,6 +415,8 @@ static int synaptics_query_hardware(struct psmouse *psmouse) if (synaptics_resolution(psmouse)) return -1; + synaptics_apply_quirks(psmouse); + return 0; } -- cgit v1.2.1 From 9aff65982d0f58a78a27769fba7e97bc937b2593 Mon Sep 17 00:00:00 2001 From: Daniel Martin Date: Sun, 8 Mar 2015 22:28:29 -0700 Subject: Input: synaptics - log queried and quirked dimension values Logging the dimension values we queried and the values we use from a quirk to overwrite can be helpful for debugging. This partly relates to bug: https://bugzilla.kernel.org/show_bug.cgi?id=91541 Cc: stable@vger.kernel.org Signed-off-by: Daniel Martin Acked-by: Hans de Goede Signed-off-by: Dmitry Torokhov --- drivers/input/mouse/synaptics.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/drivers/input/mouse/synaptics.c b/drivers/input/mouse/synaptics.c index b501dda75dcb..47c5dca20a60 100644 --- a/drivers/input/mouse/synaptics.c +++ b/drivers/input/mouse/synaptics.c @@ -362,6 +362,9 @@ static int synaptics_resolution(struct psmouse *psmouse) } else { priv->x_max = (resp[0] << 5) | ((resp[1] & 0x0f) << 1); priv->y_max = (resp[2] << 5) | ((resp[1] & 0xf0) >> 3); + psmouse_info(psmouse, + "queried max coordinates: x [..%d], y [..%d]\n", + priv->x_max, priv->y_max); } } @@ -373,6 +376,9 @@ static int synaptics_resolution(struct psmouse *psmouse) } else { priv->x_min = (resp[0] << 5) | ((resp[1] & 0x0f) << 1); priv->y_min = (resp[2] << 5) | ((resp[1] & 0xf0) >> 3); + psmouse_info(psmouse, + "queried min coordinates: x [%d..], y [%d..]\n", + priv->x_min, priv->y_min); } } @@ -395,6 +401,10 @@ static void synaptics_apply_quirks(struct psmouse *psmouse) priv->x_max = min_max_pnpid_table[i].x_max; priv->y_min = min_max_pnpid_table[i].y_min; priv->y_max = min_max_pnpid_table[i].y_max; + psmouse_info(psmouse, + "quirked min/max coordinates: x [%d..%d], y [%d..%d]\n", + priv->x_min, priv->x_max, + priv->y_min, priv->y_max); break; } } -- cgit v1.2.1 From ac097930f0730a9b777737de2b51e0fc49d2be7a Mon Sep 17 00:00:00 2001 From: Daniel Martin Date: Sun, 8 Mar 2015 22:28:40 -0700 Subject: Input: synaptics - query min dimensions for fw v8.1 Query the min dimensions even if the check SYN_EXT_CAP_REQUESTS(priv->capabilities) >= 7 fails, but we know that the firmware version 8.1 is safe. With that we don't need quirks for post-2013 models anymore as they expose correct min and max dimensions. Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=91541 Cc: stable@vger.kernel.org Signed-off-by: Daniel Martin re-order the tests to check SYN_CAP_MIN_DIMENSIONS even on FW 8.1 Signed-off-by: Benjamin Tissoires Acked-by: Hans de Goede Signed-off-by: Dmitry Torokhov --- drivers/input/mouse/synaptics.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/drivers/input/mouse/synaptics.c b/drivers/input/mouse/synaptics.c index 47c5dca20a60..87c37f745b92 100644 --- a/drivers/input/mouse/synaptics.c +++ b/drivers/input/mouse/synaptics.c @@ -368,8 +368,14 @@ static int synaptics_resolution(struct psmouse *psmouse) } } - if (SYN_EXT_CAP_REQUESTS(priv->capabilities) >= 7 && - SYN_CAP_MIN_DIMENSIONS(priv->ext_cap_0c)) { + if (SYN_CAP_MIN_DIMENSIONS(priv->ext_cap_0c) && + (SYN_EXT_CAP_REQUESTS(priv->capabilities) >= 7 || + /* + * Firmware v8.1 does not report proper number of extended + * capabilities, but has been proven to report correct min + * coordinates. + */ + SYN_ID_FULL(priv->identity) == 0x801)) { if (synaptics_send_cmd(psmouse, SYN_QUE_EXT_MIN_COORDS, resp)) { psmouse_warn(psmouse, "device claims to have min coordinates query, but I'm not able to read it.\n"); -- cgit v1.2.1 From b05f4d1c332a22f98c037fa64f249aa30877adaf Mon Sep 17 00:00:00 2001 From: Daniel Martin Date: Sun, 8 Mar 2015 22:29:07 -0700 Subject: Input: synaptics - remove obsolete min/max quirk for X240 The firmware of the X240 (LEN0035, 2013/12) exposes the same values x [1232..5710], y [1156..4696] as the quirk applies. Cc: stable@vger.kernel.org Signed-off-by: Daniel Martin Acked-by: Hans de Goede Signed-off-by: Dmitry Torokhov --- drivers/input/mouse/synaptics.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/input/mouse/synaptics.c b/drivers/input/mouse/synaptics.c index 87c37f745b92..af686a82b02b 100644 --- a/drivers/input/mouse/synaptics.c +++ b/drivers/input/mouse/synaptics.c @@ -131,7 +131,7 @@ static const struct min_max_quirk min_max_pnpid_table[] = { 1024, 5052, 2258, 4832 }, { - (const char * const []){"LEN0035", "LEN0042", NULL}, + (const char * const []){"LEN0042", NULL}, 1232, 5710, 1156, 4696 }, { -- cgit v1.2.1 From 5b3089ddb540401c1ad2e385a03d7e89ff954585 Mon Sep 17 00:00:00 2001 From: Daniel Martin Date: Sun, 8 Mar 2015 22:29:15 -0700 Subject: Input: synaptics - support min/max board id in min_max_pnpid_table Add a min/max range for board ids to the min/max coordinates quirk. This makes it possible to restrict quirks to specific models based upon their board id. The define ANY_BOARD_ID (0) serves as a wild card. Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=91541 Cc: stable@vger.kernel.org Signed-off-by: Daniel Martin Acked-by: Hans de Goede Signed-off-by: Dmitry Torokhov --- drivers/input/mouse/synaptics.c | 42 +++++++++++++++++++++++++++++------------ 1 file changed, 30 insertions(+), 12 deletions(-) diff --git a/drivers/input/mouse/synaptics.c b/drivers/input/mouse/synaptics.c index af686a82b02b..a900a385e5c3 100644 --- a/drivers/input/mouse/synaptics.c +++ b/drivers/input/mouse/synaptics.c @@ -120,32 +120,41 @@ void synaptics_reset(struct psmouse *psmouse) static bool cr48_profile_sensor; +#define ANY_BOARD_ID 0 struct min_max_quirk { const char * const *pnp_ids; + struct { + unsigned long int min, max; + } board_id; int x_min, x_max, y_min, y_max; }; static const struct min_max_quirk min_max_pnpid_table[] = { { (const char * const []){"LEN0033", NULL}, + {ANY_BOARD_ID, ANY_BOARD_ID}, 1024, 5052, 2258, 4832 }, { (const char * const []){"LEN0042", NULL}, + {ANY_BOARD_ID, ANY_BOARD_ID}, 1232, 5710, 1156, 4696 }, { (const char * const []){"LEN0034", "LEN0036", "LEN0037", "LEN0039", "LEN2002", "LEN2004", NULL}, + {ANY_BOARD_ID, ANY_BOARD_ID}, 1024, 5112, 2024, 4832 }, { (const char * const []){"LEN2001", NULL}, + {ANY_BOARD_ID, ANY_BOARD_ID}, 1024, 5022, 2508, 4832 }, { (const char * const []){"LEN2006", NULL}, + {ANY_BOARD_ID, ANY_BOARD_ID}, 1264, 5675, 1171, 4688 }, { } @@ -401,18 +410,27 @@ static void synaptics_apply_quirks(struct psmouse *psmouse) int i; for (i = 0; min_max_pnpid_table[i].pnp_ids; i++) { - if (psmouse_matches_pnp_id(psmouse, - min_max_pnpid_table[i].pnp_ids)) { - priv->x_min = min_max_pnpid_table[i].x_min; - priv->x_max = min_max_pnpid_table[i].x_max; - priv->y_min = min_max_pnpid_table[i].y_min; - priv->y_max = min_max_pnpid_table[i].y_max; - psmouse_info(psmouse, - "quirked min/max coordinates: x [%d..%d], y [%d..%d]\n", - priv->x_min, priv->x_max, - priv->y_min, priv->y_max); - break; - } + if (!psmouse_matches_pnp_id(psmouse, + min_max_pnpid_table[i].pnp_ids)) + continue; + + if (min_max_pnpid_table[i].board_id.min != ANY_BOARD_ID && + priv->board_id < min_max_pnpid_table[i].board_id.min) + continue; + + if (min_max_pnpid_table[i].board_id.max != ANY_BOARD_ID && + priv->board_id > min_max_pnpid_table[i].board_id.max) + continue; + + priv->x_min = min_max_pnpid_table[i].x_min; + priv->x_max = min_max_pnpid_table[i].x_max; + priv->y_min = min_max_pnpid_table[i].y_min; + priv->y_max = min_max_pnpid_table[i].y_max; + psmouse_info(psmouse, + "quirked min/max coordinates: x [%d..%d], y [%d..%d]\n", + priv->x_min, priv->x_max, + priv->y_min, priv->y_max); + break; } } -- cgit v1.2.1 From 02e07492cdfae9c86e3bd21c0beec88dbcc1e9e8 Mon Sep 17 00:00:00 2001 From: Benjamin Tissoires Date: Sun, 8 Mar 2015 22:29:25 -0700 Subject: Input: synaptics - skip quirks when post-2013 dimensions Post-2013 Lenovo laptops provide correct min/max dimensions, which are different with the ones currently quirked. According to https://bugzilla.kernel.org/show_bug.cgi?id=91541 the following board ids are assigned in the post-2013 touchpads: t440p/t440s: LEN0036 -> 2964/2962 t540p: LEN0034 -> 2964 Using 2961 as the common minimum makes these 3 laptops OK. We may need to update those values later if other pnp_ids has a lower board_id. Cc: stable@vger.kernel.org Signed-off-by: Benjamin Tissoires Acked-by: Hans de Goede Signed-off-by: Dmitry Torokhov --- drivers/input/mouse/synaptics.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/input/mouse/synaptics.c b/drivers/input/mouse/synaptics.c index a900a385e5c3..9567a708aa64 100644 --- a/drivers/input/mouse/synaptics.c +++ b/drivers/input/mouse/synaptics.c @@ -144,7 +144,7 @@ static const struct min_max_quirk min_max_pnpid_table[] = { (const char * const []){"LEN0034", "LEN0036", "LEN0037", "LEN0039", "LEN2002", "LEN2004", NULL}, - {ANY_BOARD_ID, ANY_BOARD_ID}, + {ANY_BOARD_ID, 2961}, 1024, 5112, 2024, 4832 }, { -- cgit v1.2.1 From dc5465dc8a6d5cae8a0e1d8826bdcb2e4cb261ab Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Sun, 8 Mar 2015 22:30:43 -0700 Subject: Input: synaptics - fix middle button on Lenovo 2015 products On the X1 Carbon 3rd gen (with a 2015 broadwell cpu), the physical middle button of the trackstick (attached to the touchpad serio device, of course) seems to get lost. Actually, the touchpads reports 3 extra buttons, which falls in the switch below to the '2' case. Let's handle the case of odd numbers also, so that the middle button finds its way back. Cc: stable@vger.kernel.org Signed-off-by: Benjamin Tissoires Acked-by: Hans de Goede Signed-off-by: Dmitry Torokhov --- drivers/input/mouse/synaptics.c | 44 ++++++++++++++++++++--------------------- 1 file changed, 21 insertions(+), 23 deletions(-) diff --git a/drivers/input/mouse/synaptics.c b/drivers/input/mouse/synaptics.c index 9567a708aa64..e78cc5578527 100644 --- a/drivers/input/mouse/synaptics.c +++ b/drivers/input/mouse/synaptics.c @@ -658,6 +658,18 @@ static void synaptics_parse_agm(const unsigned char buf[], priv->agm_pending = true; } +static void synaptics_parse_ext_buttons(const unsigned char buf[], + struct synaptics_data *priv, + struct synaptics_hw_state *hw) +{ + unsigned int ext_bits = + (SYN_CAP_MULTI_BUTTON_NO(priv->ext_cap) + 1) >> 1; + unsigned int ext_mask = GENMASK(ext_bits - 1, 0); + + hw->ext_buttons = buf[4] & ext_mask; + hw->ext_buttons |= (buf[5] & ext_mask) << ext_bits; +} + static bool is_forcepad; static int synaptics_parse_hw_state(const unsigned char buf[], @@ -744,28 +756,9 @@ static int synaptics_parse_hw_state(const unsigned char buf[], hw->down = ((buf[0] ^ buf[3]) & 0x02) ? 1 : 0; } - if (SYN_CAP_MULTI_BUTTON_NO(priv->ext_cap) && + if (SYN_CAP_MULTI_BUTTON_NO(priv->ext_cap) > 0 && ((buf[0] ^ buf[3]) & 0x02)) { - switch (SYN_CAP_MULTI_BUTTON_NO(priv->ext_cap) & ~0x01) { - default: - /* - * if nExtBtn is greater than 8 it should be - * considered invalid and treated as 0 - */ - break; - case 8: - hw->ext_buttons |= ((buf[5] & 0x08)) ? 0x80 : 0; - hw->ext_buttons |= ((buf[4] & 0x08)) ? 0x40 : 0; - case 6: - hw->ext_buttons |= ((buf[5] & 0x04)) ? 0x20 : 0; - hw->ext_buttons |= ((buf[4] & 0x04)) ? 0x10 : 0; - case 4: - hw->ext_buttons |= ((buf[5] & 0x02)) ? 0x08 : 0; - hw->ext_buttons |= ((buf[4] & 0x02)) ? 0x04 : 0; - case 2: - hw->ext_buttons |= ((buf[5] & 0x01)) ? 0x02 : 0; - hw->ext_buttons |= ((buf[4] & 0x01)) ? 0x01 : 0; - } + synaptics_parse_ext_buttons(buf, priv, hw); } } else { hw->x = (((buf[1] & 0x1f) << 8) | buf[2]); @@ -832,6 +825,7 @@ static void synaptics_report_buttons(struct psmouse *psmouse, { struct input_dev *dev = psmouse->dev; struct synaptics_data *priv = psmouse->private; + int ext_bits = (SYN_CAP_MULTI_BUTTON_NO(priv->ext_cap) + 1) >> 1; int i; input_report_key(dev, BTN_LEFT, hw->left); @@ -845,8 +839,12 @@ static void synaptics_report_buttons(struct psmouse *psmouse, input_report_key(dev, BTN_BACK, hw->down); } - for (i = 0; i < SYN_CAP_MULTI_BUTTON_NO(priv->ext_cap); i++) - input_report_key(dev, BTN_0 + i, hw->ext_buttons & (1 << i)); + for (i = 0; i < ext_bits; i++) { + input_report_key(dev, BTN_0 + 2 * i, + hw->ext_buttons & (1 << i)); + input_report_key(dev, BTN_1 + 2 * i, + hw->ext_buttons & (1 << (i + ext_bits))); + } } static void synaptics_report_slot(struct input_dev *dev, int slot, -- cgit v1.2.1 From ebc80840b850db72f7ae84fbcf77630ae5409629 Mon Sep 17 00:00:00 2001 From: Benjamin Tissoires Date: Sun, 8 Mar 2015 22:32:43 -0700 Subject: Input: synaptics - handle spurious release of trackstick buttons The Fimware 8.1 has a bug in which the extra buttons are only sent when the ExtBit is 1. This should be fixed in a future FW update which should have a bump of the minor version. Cc: stable@vger.kernel.org Signed-off-by: Benjamin Tissoires Acked-by: Hans de Goede Signed-off-by: Dmitry Torokhov --- drivers/input/mouse/synaptics.c | 33 +++++++++++++++++++++++++-------- 1 file changed, 25 insertions(+), 8 deletions(-) diff --git a/drivers/input/mouse/synaptics.c b/drivers/input/mouse/synaptics.c index e78cc5578527..2f42a712f3e0 100644 --- a/drivers/input/mouse/synaptics.c +++ b/drivers/input/mouse/synaptics.c @@ -820,14 +820,36 @@ static void synaptics_report_semi_mt_data(struct input_dev *dev, } } -static void synaptics_report_buttons(struct psmouse *psmouse, - const struct synaptics_hw_state *hw) +static void synaptics_report_ext_buttons(struct psmouse *psmouse, + const struct synaptics_hw_state *hw) { struct input_dev *dev = psmouse->dev; struct synaptics_data *priv = psmouse->private; int ext_bits = (SYN_CAP_MULTI_BUTTON_NO(priv->ext_cap) + 1) >> 1; int i; + if (!SYN_CAP_MULTI_BUTTON_NO(priv->ext_cap)) + return; + + /* Bug in FW 8.1, buttons are reported only when ExtBit is 1 */ + if (SYN_ID_FULL(priv->identity) == 0x801 && + !((psmouse->packet[0] ^ psmouse->packet[3]) & 0x02)) + return; + + for (i = 0; i < ext_bits; i++) { + input_report_key(dev, BTN_0 + 2 * i, + hw->ext_buttons & (1 << i)); + input_report_key(dev, BTN_1 + 2 * i, + hw->ext_buttons & (1 << (i + ext_bits))); + } +} + +static void synaptics_report_buttons(struct psmouse *psmouse, + const struct synaptics_hw_state *hw) +{ + struct input_dev *dev = psmouse->dev; + struct synaptics_data *priv = psmouse->private; + input_report_key(dev, BTN_LEFT, hw->left); input_report_key(dev, BTN_RIGHT, hw->right); @@ -839,12 +861,7 @@ static void synaptics_report_buttons(struct psmouse *psmouse, input_report_key(dev, BTN_BACK, hw->down); } - for (i = 0; i < ext_bits; i++) { - input_report_key(dev, BTN_0 + 2 * i, - hw->ext_buttons & (1 << i)); - input_report_key(dev, BTN_1 + 2 * i, - hw->ext_buttons & (1 << (i + ext_bits))); - } + synaptics_report_ext_buttons(psmouse, hw); } static void synaptics_report_slot(struct input_dev *dev, int slot, -- cgit v1.2.1 From b57a7128be24062b5b5b26032b7cd58f1651547e Mon Sep 17 00:00:00 2001 From: Benjamin Tissoires Date: Sun, 8 Mar 2015 22:33:36 -0700 Subject: Input: synaptics - do not retrieve the board id on old firmwares The board id capability has been added in firmware 7.5. Cc: stable@vger.kernel.org Signed-off-by: Benjamin Tissoires Acked-by: Hans de Goede Signed-off-by: Dmitry Torokhov --- drivers/input/mouse/synaptics.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/input/mouse/synaptics.c b/drivers/input/mouse/synaptics.c index 2f42a712f3e0..2176874a41b1 100644 --- a/drivers/input/mouse/synaptics.c +++ b/drivers/input/mouse/synaptics.c @@ -250,6 +250,10 @@ static int synaptics_board_id(struct psmouse *psmouse) struct synaptics_data *priv = psmouse->private; unsigned char bid[3]; + /* firmwares prior 7.5 have no board_id encoded */ + if (SYN_ID_FULL(priv->identity) < 0x705) + return 0; + if (synaptics_send_cmd(psmouse, SYN_QUE_MODES, bid)) return -1; priv->board_id = ((bid[0] & 0xfc) << 6) | bid[1]; -- cgit v1.2.1 From 06aa374bc70468b517dd36b95c48c8f391c08a27 Mon Sep 17 00:00:00 2001 From: Benjamin Tissoires Date: Sun, 8 Mar 2015 22:34:03 -0700 Subject: Input: synaptics - retrieve the extended capabilities in query $10 Newer Synaptics touchpads need to get information from the query $10. Retrieve it if available. Signed-off-by: Benjamin Tissoires Acked-by: Hans de Goede Signed-off-by: Dmitry Torokhov --- drivers/input/mouse/synaptics.c | 23 ++++++++++++++++++++--- drivers/input/mouse/synaptics.h | 23 +++++++++++++++++++++++ 2 files changed, 43 insertions(+), 3 deletions(-) diff --git a/drivers/input/mouse/synaptics.c b/drivers/input/mouse/synaptics.c index 2176874a41b1..8f6a153677b9 100644 --- a/drivers/input/mouse/synaptics.c +++ b/drivers/input/mouse/synaptics.c @@ -241,11 +241,24 @@ static int synaptics_model_id(struct psmouse *psmouse) return 0; } +static int synaptics_more_extended_queries(struct psmouse *psmouse) +{ + struct synaptics_data *priv = psmouse->private; + unsigned char buf[3]; + + if (synaptics_send_cmd(psmouse, SYN_QUE_MEXT_CAPAB_10, buf)) + return -1; + + priv->ext_cap_10 = (buf[0]<<16) | (buf[1]<<8) | buf[2]; + + return 0; +} + /* - * Read the board id from the touchpad + * Read the board id and the "More Extended Queries" from the touchpad * The board id is encoded in the "QUERY MODES" response */ -static int synaptics_board_id(struct psmouse *psmouse) +static int synaptics_query_modes(struct psmouse *psmouse) { struct synaptics_data *priv = psmouse->private; unsigned char bid[3]; @@ -257,6 +270,10 @@ static int synaptics_board_id(struct psmouse *psmouse) if (synaptics_send_cmd(psmouse, SYN_QUE_MODES, bid)) return -1; priv->board_id = ((bid[0] & 0xfc) << 6) | bid[1]; + + if (SYN_MEXT_CAP_BIT(bid[0])) + return synaptics_more_extended_queries(psmouse); + return 0; } @@ -446,7 +463,7 @@ static int synaptics_query_hardware(struct psmouse *psmouse) return -1; if (synaptics_firmware_id(psmouse)) return -1; - if (synaptics_board_id(psmouse)) + if (synaptics_query_modes(psmouse)) return -1; if (synaptics_capability(psmouse)) return -1; diff --git a/drivers/input/mouse/synaptics.h b/drivers/input/mouse/synaptics.h index 1bd01f21783b..8d3761ce8f54 100644 --- a/drivers/input/mouse/synaptics.h +++ b/drivers/input/mouse/synaptics.h @@ -22,6 +22,7 @@ #define SYN_QUE_EXT_CAPAB_0C 0x0c #define SYN_QUE_EXT_MAX_COORDS 0x0d #define SYN_QUE_EXT_MIN_COORDS 0x0f +#define SYN_QUE_MEXT_CAPAB_10 0x10 /* synatics modes */ #define SYN_BIT_ABSOLUTE_MODE (1 << 7) @@ -53,6 +54,7 @@ #define SYN_EXT_CAP_REQUESTS(c) (((c) & 0x700000) >> 20) #define SYN_CAP_MULTI_BUTTON_NO(ec) (((ec) & 0x00f000) >> 12) #define SYN_CAP_PRODUCT_ID(ec) (((ec) & 0xff0000) >> 16) +#define SYN_MEXT_CAP_BIT(m) ((m) & (1 << 1)) /* * The following describes response for the 0x0c query. @@ -89,6 +91,26 @@ #define SYN_CAP_REDUCED_FILTERING(ex0c) ((ex0c) & 0x000400) #define SYN_CAP_IMAGE_SENSOR(ex0c) ((ex0c) & 0x000800) +/* + * The following descibes response for the 0x10 query. + * + * byte mask name meaning + * ---- ---- ------- ------------ + * 1 0x01 ext buttons are stick buttons exported in the extended + * capability are actually meant to be used + * by the tracktick (pass-through). + * 1 0x02 SecurePad the touchpad is a SecurePad, so it + * contains a built-in fingerprint reader. + * 1 0xe0 more ext count how many more extented queries are + * available after this one. + * 2 0xff SecurePad width the width of the SecurePad fingerprint + * reader. + * 3 0xff SecurePad height the height of the SecurePad fingerprint + * reader. + */ +#define SYN_CAP_EXT_BUTTONS_STICK(ex10) ((ex10) & 0x010000) +#define SYN_CAP_SECUREPAD(ex10) ((ex10) & 0x020000) + /* synaptics modes query bits */ #define SYN_MODE_ABSOLUTE(m) ((m) & (1 << 7)) #define SYN_MODE_RATE(m) ((m) & (1 << 6)) @@ -156,6 +178,7 @@ struct synaptics_data { unsigned long int capabilities; /* Capabilities */ unsigned long int ext_cap; /* Extended Capabilities */ unsigned long int ext_cap_0c; /* Ext Caps from 0x0c query */ + unsigned long int ext_cap_10; /* Ext Caps from 0x10 query */ unsigned long int identity; /* Identification */ unsigned int x_res, y_res; /* X/Y resolution in units/mm */ unsigned int x_max, y_max; /* Max coordinates (from FW) */ -- cgit v1.2.1 From 3adde1f59195df2965f632e22b31f97fb371612f Mon Sep 17 00:00:00 2001 From: Benjamin Tissoires Date: Sun, 8 Mar 2015 22:34:50 -0700 Subject: Input: synaptics - remove TOPBUTTONPAD property for Lenovos 2015 The 2015 series of the Lenovo thinkpads added back the hardware buttons on top of the touchpad for the trackstick. Unfortunately, Lenovo used the PNPIDs that are supposed to be "5 buttons" touchpads, so the new laptops also have the INPUT_PROP_TOPBUTTONPAD. Yay! Instead of manually removing each of the new ones, or hoping that we know all the current ones, we can consider that the PNPIDs list that were given contains touchpads that have the trackstick buttons, either physically wired to them, or emulated with the top software button property. Thanks to the extra buttons capability in query $10, we can reliably detect the physical buttons from the software ones, and so we can remove the TOPBUTTONPAD property even if it was declared as such. Signed-off-by: Benjamin Tissoires Acked-by: Hans de Goede Signed-off-by: Dmitry Torokhov --- drivers/input/mouse/synaptics.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/input/mouse/synaptics.c b/drivers/input/mouse/synaptics.c index 8f6a153677b9..9d599eb79f17 100644 --- a/drivers/input/mouse/synaptics.c +++ b/drivers/input/mouse/synaptics.c @@ -1570,7 +1570,8 @@ static void set_input_params(struct psmouse *psmouse, if (SYN_CAP_CLICKPAD(priv->ext_cap_0c)) { __set_bit(INPUT_PROP_BUTTONPAD, dev->propbit); - if (psmouse_matches_pnp_id(psmouse, topbuttonpad_pnp_ids)) + if (psmouse_matches_pnp_id(psmouse, topbuttonpad_pnp_ids) && + !SYN_CAP_EXT_BUTTONS_STICK(priv->ext_cap_10)) __set_bit(INPUT_PROP_TOPBUTTONPAD, dev->propbit); /* Clickpads report only left button */ __clear_bit(BTN_RIGHT, dev->keybit); -- cgit v1.2.1 From cdd9dc195916ef5644cfac079094c3c1d1616e4c Mon Sep 17 00:00:00 2001 From: Benjamin Tissoires Date: Sun, 8 Mar 2015 22:35:41 -0700 Subject: Input: synaptics - re-route tracksticks buttons on the Lenovo 2015 series The 2015 series of the Lenovo thinkpads added back the hardware buttons on top of the touchpad for the trackstick. Unfortunately, they are wired to the touchpad, and not the trackstick. Thus, they are seen as extra buttons from the kernel point of view. This leads to a problem in user space because extra buttons on synaptics devices used to be used as scroll up/down buttons. So in the end, the experience for the user is scroll events for buttons left and right when using the trackstick. Yay! Fortunately, the firmware advertises such behavior in the extended capability $10, and so we can re-route the buttons through the pass-through interface. Hallelujah-expressed-by: Peter Hutterer Signed-off-by: Benjamin Tissoires Acked-by: Hans de Goede Signed-off-by: Dmitry Torokhov --- drivers/input/mouse/synaptics.c | 47 +++++++++++++++++++++++++++++++---------- drivers/input/mouse/synaptics.h | 5 +++++ 2 files changed, 41 insertions(+), 11 deletions(-) diff --git a/drivers/input/mouse/synaptics.c b/drivers/input/mouse/synaptics.c index 9d599eb79f17..ecc7811cbd46 100644 --- a/drivers/input/mouse/synaptics.c +++ b/drivers/input/mouse/synaptics.c @@ -579,18 +579,22 @@ static int synaptics_is_pt_packet(unsigned char *buf) return (buf[0] & 0xFC) == 0x84 && (buf[3] & 0xCC) == 0xC4; } -static void synaptics_pass_pt_packet(struct serio *ptport, unsigned char *packet) +static void synaptics_pass_pt_packet(struct psmouse *psmouse, + struct serio *ptport, + unsigned char *packet) { + struct synaptics_data *priv = psmouse->private; struct psmouse *child = serio_get_drvdata(ptport); if (child && child->state == PSMOUSE_ACTIVATED) { - serio_interrupt(ptport, packet[1], 0); + serio_interrupt(ptport, packet[1] | priv->pt_buttons, 0); serio_interrupt(ptport, packet[4], 0); serio_interrupt(ptport, packet[5], 0); if (child->pktsize == 4) serio_interrupt(ptport, packet[2], 0); - } else + } else { serio_interrupt(ptport, packet[1], 0); + } } static void synaptics_pt_activate(struct psmouse *psmouse) @@ -847,6 +851,7 @@ static void synaptics_report_ext_buttons(struct psmouse *psmouse, struct input_dev *dev = psmouse->dev; struct synaptics_data *priv = psmouse->private; int ext_bits = (SYN_CAP_MULTI_BUTTON_NO(priv->ext_cap) + 1) >> 1; + char buf[6] = { 0x00, 0x00, 0x00, 0x00, 0x00, 0x00}; int i; if (!SYN_CAP_MULTI_BUTTON_NO(priv->ext_cap)) @@ -857,12 +862,30 @@ static void synaptics_report_ext_buttons(struct psmouse *psmouse, !((psmouse->packet[0] ^ psmouse->packet[3]) & 0x02)) return; - for (i = 0; i < ext_bits; i++) { - input_report_key(dev, BTN_0 + 2 * i, - hw->ext_buttons & (1 << i)); - input_report_key(dev, BTN_1 + 2 * i, - hw->ext_buttons & (1 << (i + ext_bits))); + if (!SYN_CAP_EXT_BUTTONS_STICK(priv->ext_cap_10)) { + for (i = 0; i < ext_bits; i++) { + input_report_key(dev, BTN_0 + 2 * i, + hw->ext_buttons & (1 << i)); + input_report_key(dev, BTN_1 + 2 * i, + hw->ext_buttons & (1 << (i + ext_bits))); + } + return; } + + /* + * This generation of touchpads has the trackstick buttons + * physically wired to the touchpad. Re-route them through + * the pass-through interface. + */ + if (!priv->pt_port) + return; + + /* The trackstick expects at most 3 buttons */ + priv->pt_buttons = SYN_CAP_EXT_BUTTON_STICK_L(hw->ext_buttons) | + SYN_CAP_EXT_BUTTON_STICK_R(hw->ext_buttons) << 1 | + SYN_CAP_EXT_BUTTON_STICK_M(hw->ext_buttons) << 2; + + synaptics_pass_pt_packet(psmouse, priv->pt_port, buf); } static void synaptics_report_buttons(struct psmouse *psmouse, @@ -1459,7 +1482,8 @@ static psmouse_ret_t synaptics_process_byte(struct psmouse *psmouse) if (SYN_CAP_PASS_THROUGH(priv->capabilities) && synaptics_is_pt_packet(psmouse->packet)) { if (priv->pt_port) - synaptics_pass_pt_packet(priv->pt_port, psmouse->packet); + synaptics_pass_pt_packet(psmouse, priv->pt_port, + psmouse->packet); } else synaptics_process_packet(psmouse); @@ -1561,8 +1585,9 @@ static void set_input_params(struct psmouse *psmouse, __set_bit(BTN_BACK, dev->keybit); } - for (i = 0; i < SYN_CAP_MULTI_BUTTON_NO(priv->ext_cap); i++) - __set_bit(BTN_0 + i, dev->keybit); + if (!SYN_CAP_EXT_BUTTONS_STICK(priv->ext_cap_10)) + for (i = 0; i < SYN_CAP_MULTI_BUTTON_NO(priv->ext_cap); i++) + __set_bit(BTN_0 + i, dev->keybit); __clear_bit(EV_REL, dev->evbit); __clear_bit(REL_X, dev->relbit); diff --git a/drivers/input/mouse/synaptics.h b/drivers/input/mouse/synaptics.h index 8d3761ce8f54..f39539c70219 100644 --- a/drivers/input/mouse/synaptics.h +++ b/drivers/input/mouse/synaptics.h @@ -111,6 +111,10 @@ #define SYN_CAP_EXT_BUTTONS_STICK(ex10) ((ex10) & 0x010000) #define SYN_CAP_SECUREPAD(ex10) ((ex10) & 0x020000) +#define SYN_CAP_EXT_BUTTON_STICK_L(eb) (!!((eb) & 0x01)) +#define SYN_CAP_EXT_BUTTON_STICK_M(eb) (!!((eb) & 0x02)) +#define SYN_CAP_EXT_BUTTON_STICK_R(eb) (!!((eb) & 0x04)) + /* synaptics modes query bits */ #define SYN_MODE_ABSOLUTE(m) ((m) & (1 << 7)) #define SYN_MODE_RATE(m) ((m) & (1 << 6)) @@ -192,6 +196,7 @@ struct synaptics_data { bool disable_gesture; /* disable gestures */ struct serio *pt_port; /* Pass-through serio port */ + unsigned char pt_buttons; /* Pass-through buttons */ struct synaptics_mt_state mt_state; /* Current mt finger state */ bool mt_state_lost; /* mt_state may be incorrect */ -- cgit v1.2.1 From 860e6f7fcbe5653ec4e394f9ee335f2032398beb Mon Sep 17 00:00:00 2001 From: Benjamin Tissoires Date: Sun, 8 Mar 2015 22:38:55 -0700 Subject: Input: synaptics - remove X1 Carbon 3rd gen from the topbuttonpad list Lenovo decided to switch back to physical buttons for the trackstick on their latest series. The PNPId list was provided before they reverted back to physical buttons, so it contains the new models too. We can know from the touchpad capabilities that the touchpad has physical buttons, so removing the ids from the list is not mandatory. It is still nicer to remove the wrong ids, so start by removing the X1 Carbon 3rd gen, with the PNPId of LEN0048. Signed-off-by: Benjamin Tissoires Acked-by: Hans de Goede Signed-off-by: Dmitry Torokhov --- drivers/input/mouse/synaptics.c | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/input/mouse/synaptics.c b/drivers/input/mouse/synaptics.c index ecc7811cbd46..160def02cde2 100644 --- a/drivers/input/mouse/synaptics.c +++ b/drivers/input/mouse/synaptics.c @@ -183,7 +183,6 @@ static const char * const topbuttonpad_pnp_ids[] = { "LEN0045", "LEN0046", "LEN0047", - "LEN0048", "LEN0049", "LEN2000", "LEN2001", /* Edge E431 */ -- cgit v1.2.1 From 8f004f3f4daf5dc98dc78f8e62497ad834053855 Mon Sep 17 00:00:00 2001 From: Benjamin Tissoires Date: Sun, 8 Mar 2015 22:39:17 -0700 Subject: Input: synaptics - remove X250 from the topbuttonpad list Lenovo X250 has a PnpID of LEN0046, but it does not have the top software button requirement. For the record, Lenovo T450s and W541 have a PnpID of LEN200f and LEN004a, so they are not on the top software button list. Signed-off-by: Benjamin Tissoires Reviewed-by: Daniel Martin Signed-off-by: Dmitry Torokhov --- drivers/input/mouse/synaptics.c | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/input/mouse/synaptics.c b/drivers/input/mouse/synaptics.c index 160def02cde2..ca7ca8d4eb33 100644 --- a/drivers/input/mouse/synaptics.c +++ b/drivers/input/mouse/synaptics.c @@ -181,7 +181,6 @@ static const char * const topbuttonpad_pnp_ids[] = { "LEN0041", "LEN0042", /* Yoga */ "LEN0045", - "LEN0046", "LEN0047", "LEN0049", "LEN2000", -- cgit v1.2.1 From c312530589ed9524fc7cc921105dc9b67ea32d6a Mon Sep 17 00:00:00 2001 From: Malcolm Priestley Date: Mon, 9 Feb 2015 19:11:33 +0000 Subject: staging: vt6655: vnt_tx_packet fix dma_idx selection. There is still a problem that dma_idx is causing packets to go onto the wrong tx path. Protect dma_idx fully with the present first lock and use pTDInfo->byFlags TD_FLAGS_NETIF_SKB to set MACvTransmit. Signed-off-by: Malcolm Priestley Signed-off-by: Greg Kroah-Hartman --- drivers/staging/vt6655/device_main.c | 16 +++++++++------- 1 file changed, 9 insertions(+), 7 deletions(-) diff --git a/drivers/staging/vt6655/device_main.c b/drivers/staging/vt6655/device_main.c index 4324282afe49..f5c5872b587e 100644 --- a/drivers/staging/vt6655/device_main.c +++ b/drivers/staging/vt6655/device_main.c @@ -1187,12 +1187,14 @@ static int vnt_tx_packet(struct vnt_private *priv, struct sk_buff *skb) { struct ieee80211_hdr *hdr = (struct ieee80211_hdr *)skb->data; PSTxDesc head_td; - u32 dma_idx = TYPE_AC0DMA; + u32 dma_idx; unsigned long flags; spin_lock_irqsave(&priv->lock, flags); - if (!ieee80211_is_data(hdr->frame_control)) + if (ieee80211_is_data(hdr->frame_control)) + dma_idx = TYPE_AC0DMA; + else dma_idx = TYPE_TXDMA0; if (AVAIL_TD(priv, dma_idx) < 1) { @@ -1206,6 +1208,9 @@ static int vnt_tx_packet(struct vnt_private *priv, struct sk_buff *skb) head_td->pTDInfo->skb = skb; + if (dma_idx == TYPE_AC0DMA) + head_td->pTDInfo->byFlags = TD_FLAGS_NETIF_SKB; + priv->iTDUsed[dma_idx]++; /* Take ownership */ @@ -1234,13 +1239,10 @@ static int vnt_tx_packet(struct vnt_private *priv, struct sk_buff *skb) head_td->buff_addr = cpu_to_le32(head_td->pTDInfo->skb_dma); - if (dma_idx == TYPE_AC0DMA) { - head_td->pTDInfo->byFlags = TD_FLAGS_NETIF_SKB; - + if (head_td->pTDInfo->byFlags & TD_FLAGS_NETIF_SKB) MACvTransmitAC0(priv->PortOffset); - } else { + else MACvTransmit0(priv->PortOffset); - } spin_unlock_irqrestore(&priv->lock, flags); -- cgit v1.2.1 From 163fe301b9f78b6de57d0014eafe504fd20c0cd4 Mon Sep 17 00:00:00 2001 From: Malcolm Priestley Date: Sat, 7 Mar 2015 16:36:37 +0000 Subject: staging: vt6656: vnt_rf_setpower: fix missing rate RATE_12M When the driver sets this rate a power of zero value is set causing data flow stoppage until another rate is tried. Signed-off-by: Malcolm Priestley Cc: # v3.17+ Signed-off-by: Greg Kroah-Hartman --- drivers/staging/vt6656/rf.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/staging/vt6656/rf.c b/drivers/staging/vt6656/rf.c index c42cde59f598..c4286ccac320 100644 --- a/drivers/staging/vt6656/rf.c +++ b/drivers/staging/vt6656/rf.c @@ -640,6 +640,7 @@ int vnt_rf_setpower(struct vnt_private *priv, u32 rate, u32 channel) break; case RATE_6M: case RATE_9M: + case RATE_12M: case RATE_18M: case RATE_24M: case RATE_36M: -- cgit v1.2.1 From 40c8790bcb7ac74f3038153cd09310e220c6a1df Mon Sep 17 00:00:00 2001 From: Malcolm Priestley Date: Sat, 7 Mar 2015 17:04:54 +0000 Subject: vt6655: RFbSetPower fix missing rate RATE_12M When the driver sets this rate a power of zero value is set causing data flow stoppage until another rate is tried. Signed-off-by: Malcolm Priestley Cc: Signed-off-by: Greg Kroah-Hartman --- drivers/staging/vt6655/rf.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/staging/vt6655/rf.c b/drivers/staging/vt6655/rf.c index 941b2adca95a..7626f635f160 100644 --- a/drivers/staging/vt6655/rf.c +++ b/drivers/staging/vt6655/rf.c @@ -794,6 +794,7 @@ bool RFbSetPower( break; case RATE_6M: case RATE_9M: + case RATE_12M: case RATE_18M: byPwr = priv->abyOFDMPwrTbl[uCH]; if (priv->byRFType == RF_UW2452) -- cgit v1.2.1 From 1f51d5801859e0b382dcc8f06875811d63ec8953 Mon Sep 17 00:00:00 2001 From: Malcolm Priestley Date: Sat, 7 Mar 2015 17:04:55 +0000 Subject: vt6655: Fix late setting of byRFType. byRFType is not set prior to registration of mac80211 causing unpredictable operation after channel scans. With byRFType unset all channels are enabled this causes tx power to be set to values not present its eeprom. Move setting of this variable to vt6655_probe. byRFType must have a mask set. byRevId not used by driver and is removed. Signed-off-by: Malcolm Priestley Cc: # v3.19+ Signed-off-by: Greg Kroah-Hartman --- drivers/staging/vt6655/device_main.c | 16 ++++++---------- 1 file changed, 6 insertions(+), 10 deletions(-) diff --git a/drivers/staging/vt6655/device_main.c b/drivers/staging/vt6655/device_main.c index f5c5872b587e..03b2a90b9ac0 100644 --- a/drivers/staging/vt6655/device_main.c +++ b/drivers/staging/vt6655/device_main.c @@ -330,16 +330,6 @@ static void device_init_registers(struct vnt_private *pDevice) /* zonetype initial */ pDevice->byOriginalZonetype = pDevice->abyEEPROM[EEP_OFS_ZONETYPE]; - /* Get RFType */ - pDevice->byRFType = SROMbyReadEmbedded(pDevice->PortOffset, EEP_OFS_RFTYPE); - - /* force change RevID for VT3253 emu */ - if ((pDevice->byRFType & RF_EMU) != 0) - pDevice->byRevId = 0x80; - - pDevice->byRFType &= RF_MASK; - pr_debug("pDevice->byRFType = %x\n", pDevice->byRFType); - if (!pDevice->bZoneRegExist) pDevice->byZoneType = pDevice->abyEEPROM[EEP_OFS_ZONETYPE]; @@ -1780,6 +1770,12 @@ vt6655_probe(struct pci_dev *pcid, const struct pci_device_id *ent) MACvInitialize(priv->PortOffset); MACvReadEtherAddress(priv->PortOffset, priv->abyCurrentNetAddr); + /* Get RFType */ + priv->byRFType = SROMbyReadEmbedded(priv->PortOffset, EEP_OFS_RFTYPE); + priv->byRFType &= RF_MASK; + + dev_dbg(&pcid->dev, "RF Type = %x\n", priv->byRFType); + device_get_options(priv); device_set_options(priv); /* Mask out the options cannot be set to the chip */ -- cgit v1.2.1 From 798523973dcc93c2440932dc4dfe76fbf571f668 Mon Sep 17 00:00:00 2001 From: Sudeep Holla Date: Wed, 4 Mar 2015 17:07:57 +0000 Subject: usb: isp1760: fix possible deadlock in isp1760_udc_irq Use spin_{un,}lock_irq{save,restore} in isp1760_udc_{start,stop} to prevent following potentially deadlock scenario between isp1760_udc_{start,stop} and isp1760_udc_irq : ================================= [ INFO: inconsistent lock state ] 4.0.0-rc2-00004-gf7bb2ef60173 #51 Not tainted --------------------------------- inconsistent {HARDIRQ-ON-W} -> {IN-HARDIRQ-W} usage. in:imklog/2118 [HC1[1]:SC0[0]:HE0:SE1] takes: (&(&udc->lock)->rlock){?.+...}, at: [] isp1760_udc_irq+0x367/0x9dc {HARDIRQ-ON-W} state was registered at: [] _raw_spin_lock+0x23/0x30 [] isp1760_udc_start+0x23/0xf8 [] udc_bind_to_driver+0x71/0xb0 [] usb_gadget_probe_driver+0x53/0x9c [] usb_composite_probe+0x8a/0xa4 [libcomposite] [] 0xbf8311a7 [] do_one_initcall+0x8d/0x17c [] do_init_module+0x49/0x148 [] load_module+0xb7f/0xbc4 [] SyS_finit_module+0x51/0x74 [] ret_fast_syscall+0x1/0x68 irq event stamp: 4966 hardirqs last enabled at (4965): [] _raw_spin_unlock_irq+0x1f/0x24 hardirqs last disabled at (4966): [] __irq_svc+0x33/0x64 softirqs last enabled at (4458): [] __do_softirq+0x23d/0x2d0 softirqs last disabled at (4389): [] irq_exit+0xef/0x15c other info that might help us debug this: Possible unsafe locking scenario: CPU0 ---- lock(&(&udc->lock)->rlock); lock(&(&udc->lock)->rlock); *** DEADLOCK *** 1 lock held by in:imklog/2118: #0: (&f->f_pos_lock){+.+.+.}, at: [] __fdget_pos+0x31/0x34 Signed-off-by: Sudeep Holla Cc: Greg Kroah-Hartman Acked-by: Laurent Pinchart Signed-off-by: Felipe Balbi --- drivers/usb/isp1760/isp1760-udc.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/drivers/usb/isp1760/isp1760-udc.c b/drivers/usb/isp1760/isp1760-udc.c index 9612d7990565..19e6a172ff82 100644 --- a/drivers/usb/isp1760/isp1760-udc.c +++ b/drivers/usb/isp1760/isp1760-udc.c @@ -1191,6 +1191,7 @@ static int isp1760_udc_start(struct usb_gadget *gadget, struct usb_gadget_driver *driver) { struct isp1760_udc *udc = gadget_to_udc(gadget); + unsigned long flags; /* The hardware doesn't support low speed. */ if (driver->max_speed < USB_SPEED_FULL) { @@ -1198,7 +1199,7 @@ static int isp1760_udc_start(struct usb_gadget *gadget, return -EINVAL; } - spin_lock(&udc->lock); + spin_lock_irqsave(&udc->lock, flags); if (udc->driver) { dev_err(udc->isp->dev, "UDC already has a gadget driver\n"); @@ -1208,7 +1209,7 @@ static int isp1760_udc_start(struct usb_gadget *gadget, udc->driver = driver; - spin_unlock(&udc->lock); + spin_unlock_irqrestore(&udc->lock, flags); dev_dbg(udc->isp->dev, "starting UDC with driver %s\n", driver->function); @@ -1232,6 +1233,7 @@ static int isp1760_udc_start(struct usb_gadget *gadget, static int isp1760_udc_stop(struct usb_gadget *gadget) { struct isp1760_udc *udc = gadget_to_udc(gadget); + unsigned long flags; dev_dbg(udc->isp->dev, "%s\n", __func__); @@ -1239,9 +1241,9 @@ static int isp1760_udc_stop(struct usb_gadget *gadget) isp1760_udc_write(udc, DC_MODE, 0); - spin_lock(&udc->lock); + spin_lock_irqsave(&udc->lock, flags); udc->driver = NULL; - spin_unlock(&udc->lock); + spin_unlock_irqrestore(&udc->lock, flags); return 0; } -- cgit v1.2.1 From 1c390eb360c3f6bc9a06d2260eccad195c505de5 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Sat, 28 Feb 2015 00:19:41 +0100 Subject: usb: musb: fix Kconfig regression A recent bug fix I did that was marked for stable backports introduced a slightly wrong dependency on CONFIG_OMAP_CONTROL_PHY. I was missing the fact that the PHY driver already stubs out the omap_control_usb_set_mode, and we only need to add a dependency to prevent the musb-omap2430 driver from being built-in when the phy driver is a loadable module, but we should not prevent it from being built altogether when the phy driver is disabled. Signed-off-by: Arnd Bergmann Fixes: ca784be36cc725 ("usb: start using the control module driver") Cc: # v3.9+ Acked-by: Acked-by: Pavel Machek Tested-by: Aaro Koskinen Signed-off-by: Felipe Balbi --- drivers/usb/musb/Kconfig | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/usb/musb/Kconfig b/drivers/usb/musb/Kconfig index 14e1628483d9..39db8b603627 100644 --- a/drivers/usb/musb/Kconfig +++ b/drivers/usb/musb/Kconfig @@ -79,7 +79,8 @@ config USB_MUSB_TUSB6010 config USB_MUSB_OMAP2PLUS tristate "OMAP2430 and onwards" - depends on ARCH_OMAP2PLUS && USB && OMAP_CONTROL_PHY + depends on ARCH_OMAP2PLUS && USB + depends on OMAP_CONTROL_PHY || !OMAP_CONTROL_PHY select GENERIC_PHY config USB_MUSB_AM35X -- cgit v1.2.1 From 80b4a0f8feeb6ee7fa4430a2b4ae1155ed923bd2 Mon Sep 17 00:00:00 2001 From: Valentin Rothberg Date: Sun, 1 Mar 2015 16:54:32 +0100 Subject: usb: isp1760: set IRQ flags properly The IRQF_DISABLED is a NOOP and scheduled to be removed. According to commit e58aa3d2d0cc ("genirq: Run irq handlers with interrupts disabled") running IRQ handlers with interrupts enabled can cause stack overflows when the interrupt line of the issuing device is still active. This patch removes using this deprecated flag and additionally removes redundantly setting IRQF_SHARED for isp1760_udc_register(). Signed-off-by: Valentin Rothberg Acked-by: Laurent Pinchart Signed-off-by: Felipe Balbi --- drivers/usb/isp1760/isp1760-core.c | 3 +-- drivers/usb/isp1760/isp1760-udc.c | 4 ++-- 2 files changed, 3 insertions(+), 4 deletions(-) diff --git a/drivers/usb/isp1760/isp1760-core.c b/drivers/usb/isp1760/isp1760-core.c index b9827556455f..bfa402cf3a27 100644 --- a/drivers/usb/isp1760/isp1760-core.c +++ b/drivers/usb/isp1760/isp1760-core.c @@ -151,8 +151,7 @@ int isp1760_register(struct resource *mem, int irq, unsigned long irqflags, } if (IS_ENABLED(CONFIG_USB_ISP1761_UDC) && !udc_disabled) { - ret = isp1760_udc_register(isp, irq, irqflags | IRQF_SHARED | - IRQF_DISABLED); + ret = isp1760_udc_register(isp, irq, irqflags); if (ret < 0) { isp1760_hcd_unregister(&isp->hcd); return ret; diff --git a/drivers/usb/isp1760/isp1760-udc.c b/drivers/usb/isp1760/isp1760-udc.c index 19e6a172ff82..47674f9c6df2 100644 --- a/drivers/usb/isp1760/isp1760-udc.c +++ b/drivers/usb/isp1760/isp1760-udc.c @@ -1453,8 +1453,8 @@ int isp1760_udc_register(struct isp1760_device *isp, int irq, sprintf(udc->irqname, "%s (udc)", devname); - ret = request_irq(irq, isp1760_udc_irq, IRQF_SHARED | IRQF_DISABLED | - irqflags, udc->irqname, udc); + ret = request_irq(irq, isp1760_udc_irq, IRQF_SHARED | irqflags, + udc->irqname, udc); if (ret < 0) goto error; -- cgit v1.2.1 From 1998adab1c188076eaf356a8ae28217856f0ee92 Mon Sep 17 00:00:00 2001 From: Sudeep Holla Date: Thu, 26 Feb 2015 11:47:57 +0000 Subject: usb: isp1760: add peripheral/device controller chip id As per the SAF1761 data sheet[0], the DcChipID register represents the hardware version number (0001h) and the chip ID (1582h) for the Peripheral Controller. However as per the ISP1761 data sheet[1], the DcChipID register represents the hardware version number (0015h) and the chip ID (8210h) for the Peripheral Controller. This patch adds support for both the chip ID values. [0] http://www.nxp.com/documents/data_sheet/SAF1761.pdf [1] http://pdf.datasheetcatalog.com/datasheets2/74/742102_1.pdf Acked-by: Laurent Pinchart Signed-off-by: Sudeep Holla Signed-off-by: Felipe Balbi --- drivers/usb/isp1760/isp1760-udc.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/usb/isp1760/isp1760-udc.c b/drivers/usb/isp1760/isp1760-udc.c index 47674f9c6df2..f32c292cc868 100644 --- a/drivers/usb/isp1760/isp1760-udc.c +++ b/drivers/usb/isp1760/isp1760-udc.c @@ -1413,7 +1413,7 @@ static int isp1760_udc_init(struct isp1760_udc *udc) return -ENODEV; } - if (chipid != 0x00011582) { + if (chipid != 0x00011582 && chipid != 0x00158210) { dev_err(udc->isp->dev, "udc: invalid chip ID 0x%08x\n", chipid); return -ENODEV; } -- cgit v1.2.1 From 88660f7fb94cda1f8f63ee92bfcd0db39a6361e2 Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Thu, 5 Mar 2015 13:24:41 +1030 Subject: virtio_balloon: set DRIVER_OK before using device virtio spec requires that all drivers set DRIVER_OK before using devices. While balloon isn't yet included in the virtio 1 spec, previous spec versions also required this. virtio balloon might violate this rule: probe calls kthread_run before setting DRIVER_OK, which might run immediately and cause balloon to inflate/deflate. To fix, call virtio_device_ready before running the kthread. Signed-off-by: Michael S. Tsirkin Signed-off-by: Rusty Russell Cc: stable@kernel.org --- drivers/virtio/virtio_balloon.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/virtio/virtio_balloon.c b/drivers/virtio/virtio_balloon.c index 0413157f3b49..b36fe56677d5 100644 --- a/drivers/virtio/virtio_balloon.c +++ b/drivers/virtio/virtio_balloon.c @@ -499,6 +499,8 @@ static int virtballoon_probe(struct virtio_device *vdev) if (err < 0) goto out_oom_notify; + virtio_device_ready(vdev); + vb->thread = kthread_run(balloon, vb, "vballoon"); if (IS_ERR(vb->thread)) { err = PTR_ERR(vb->thread); -- cgit v1.2.1 From 7e41a9def062167b5405711a42c9ecfd163e31a9 Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Fri, 6 Mar 2015 12:50:03 +1030 Subject: virtio_blk: typo fix Now that QEmu reuses linux virtio headers, we noticed a typo in the exported virtio block header. Fix it up. Signed-off-by: Michael S. Tsirkin Signed-off-by: Rusty Russell --- include/uapi/linux/virtio_blk.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/uapi/linux/virtio_blk.h b/include/uapi/linux/virtio_blk.h index 3c53eec4ae22..b695ba959186 100644 --- a/include/uapi/linux/virtio_blk.h +++ b/include/uapi/linux/virtio_blk.h @@ -60,7 +60,7 @@ struct virtio_blk_config { __u32 size_max; /* The maximum number of segments (if VIRTIO_BLK_F_SEG_MAX) */ __u32 seg_max; - /* geometry the device (if VIRTIO_BLK_F_GEOMETRY) */ + /* geometry of the device (if VIRTIO_BLK_F_GEOMETRY) */ struct virtio_blk_geometry { __u16 cylinders; __u8 heads; -- cgit v1.2.1 From 0fa2a56437d0b7ef5d86eef2778ad3469ca72d5a Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Fri, 6 Mar 2015 12:50:03 +1030 Subject: virtio_blk: fix comment for virtio 1.0 Fix up comment to match virtio 1.0 logic: virtio_blk_outhdr isn't the first elements anymore, the only requirement is that it comes first in the s/g list. Signed-off-by: Michael S. Tsirkin Signed-off-by: Rusty Russell --- include/uapi/linux/virtio_blk.h | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/include/uapi/linux/virtio_blk.h b/include/uapi/linux/virtio_blk.h index b695ba959186..19c66fcbab8a 100644 --- a/include/uapi/linux/virtio_blk.h +++ b/include/uapi/linux/virtio_blk.h @@ -119,7 +119,11 @@ struct virtio_blk_config { #define VIRTIO_BLK_T_BARRIER 0x80000000 #endif /* !VIRTIO_BLK_NO_LEGACY */ -/* This is the first element of the read scatter-gather list. */ +/* + * This comes first in the read scatter-gather list. + * For legacy virtio, if VIRTIO_F_ANY_LAYOUT is not negotiated, + * this is the first element of the read scatter-gather list. + */ struct virtio_blk_outhdr { /* VIRTIO_BLK_T* */ __virtio32 type; -- cgit v1.2.1 From 3d2a3774c1b046f548ebea0391a602fd5685a307 Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Tue, 10 Mar 2015 11:55:08 +1030 Subject: virtio-balloon: do not call blocking ops when !TASK_RUNNING virtio balloon has this code: wait_event_interruptible(vb->config_change, (diff = towards_target(vb)) != 0 || vb->need_stats_update || kthread_should_stop() || freezing(current)); Which is a problem because towards_target() call might block after wait_event_interruptible sets task state to TAST_INTERRUPTIBLE, causing the task_struct::state collision typical of nesting of sleeping primitives See also http://lwn.net/Articles/628628/ or Thomas's bug report http://article.gmane.org/gmane.linux.kernel.virtualization/24846 for a fuller explanation. To fix, rewrite using wait_woken. Cc: stable@vger.kernel.org Reported-by: Thomas Huth Signed-off-by: Michael S. Tsirkin Tested-by: Thomas Huth Reviewed-by: Cornelia Huck Signed-off-by: Rusty Russell --- drivers/virtio/virtio_balloon.c | 19 ++++++++++++++----- 1 file changed, 14 insertions(+), 5 deletions(-) diff --git a/drivers/virtio/virtio_balloon.c b/drivers/virtio/virtio_balloon.c index b36fe56677d5..6a356e344f82 100644 --- a/drivers/virtio/virtio_balloon.c +++ b/drivers/virtio/virtio_balloon.c @@ -29,6 +29,7 @@ #include #include #include +#include /* * Balloon device works in 4K page units. So each page is pointed to by @@ -334,17 +335,25 @@ static int virtballoon_oom_notify(struct notifier_block *self, static int balloon(void *_vballoon) { struct virtio_balloon *vb = _vballoon; + DEFINE_WAIT_FUNC(wait, woken_wake_function); set_freezable(); while (!kthread_should_stop()) { s64 diff; try_to_freeze(); - wait_event_interruptible(vb->config_change, - (diff = towards_target(vb)) != 0 - || vb->need_stats_update - || kthread_should_stop() - || freezing(current)); + + add_wait_queue(&vb->config_change, &wait); + for (;;) { + if ((diff = towards_target(vb)) != 0 || + vb->need_stats_update || + kthread_should_stop() || + freezing(current)) + break; + wait_woken(&wait, TASK_INTERRUPTIBLE, MAX_SCHEDULE_TIMEOUT); + } + remove_wait_queue(&vb->config_change, &wait); + if (vb->need_stats_update) stats_handle_request(vb); if (diff > 0) -- cgit v1.2.1 From 394838c96013ba414a24ffe7a2a593a9154daadf Mon Sep 17 00:00:00 2001 From: Andy Lutomirski Date: Mon, 9 Mar 2015 17:42:31 -0700 Subject: x86/asm/entry/32: Fix user_mode() misuses The one in do_debug() is probably harmless, but better safe than sorry. Signed-off-by: Andy Lutomirski Cc: Cc: Borislav Petkov Cc: Dave Hansen Cc: H. Peter Anvin Cc: Linus Torvalds Cc: Thomas Gleixner Link: http://lkml.kernel.org/r/d67deaa9df5458363623001f252d1aee3215d014.1425948056.git.luto@amacapital.net Signed-off-by: Ingo Molnar --- arch/x86/kernel/traps.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/x86/kernel/traps.c b/arch/x86/kernel/traps.c index 9d2073e2ecc9..4ff5d162ff9f 100644 --- a/arch/x86/kernel/traps.c +++ b/arch/x86/kernel/traps.c @@ -384,7 +384,7 @@ dotraplinkage void do_bounds(struct pt_regs *regs, long error_code) goto exit; conditional_sti(regs); - if (!user_mode(regs)) + if (!user_mode_vm(regs)) die("bounds", regs, error_code); if (!cpu_feature_enabled(X86_FEATURE_MPX)) { @@ -637,7 +637,7 @@ dotraplinkage void do_debug(struct pt_regs *regs, long error_code) * then it's very likely the result of an icebp/int01 trap. * User wants a sigtrap for that. */ - if (!dr6 && user_mode(regs)) + if (!dr6 && user_mode_vm(regs)) user_icebp = 1; /* Catch kmemcheck conditions first of all! */ -- cgit v1.2.1 From 5778d39d070b4ac5f889928175b7f2d53ae7504e Mon Sep 17 00:00:00 2001 From: WANG Cong Date: Mon, 9 Mar 2015 17:03:40 -0700 Subject: net_sched: fix struct tc_u_hnode layout in u32 We dynamically allocate divisor+1 entries for ->ht[] in tc_u_hnode: ht = kzalloc(sizeof(*ht) + divisor*sizeof(void *), GFP_KERNEL); So ->ht is supposed to be the last field of this struct, however this is broken, since an rcu head is appended after it. Fixes: 1ce87720d456 ("net: sched: make cls_u32 lockless") Cc: Jamal Hadi Salim Cc: John Fastabend Signed-off-by: Cong Wang Acked-by: Eric Dumazet Signed-off-by: David S. Miller --- net/sched/cls_u32.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/net/sched/cls_u32.c b/net/sched/cls_u32.c index 09487afbfd51..95fdf4e40051 100644 --- a/net/sched/cls_u32.c +++ b/net/sched/cls_u32.c @@ -78,8 +78,11 @@ struct tc_u_hnode { struct tc_u_common *tp_c; int refcnt; unsigned int divisor; - struct tc_u_knode __rcu *ht[1]; struct rcu_head rcu; + /* The 'ht' field MUST be the last field in structure to allow for + * more entries allocated at end of structure. + */ + struct tc_u_knode __rcu *ht[1]; }; struct tc_u_common { -- cgit v1.2.1 From 4736edc764b5464d625385ef89ed0c3c88b09897 Mon Sep 17 00:00:00 2001 From: Yongbae Park Date: Tue, 10 Mar 2015 11:15:39 +0900 Subject: ibmveth: enable interrupts after napi_complete() The interrupt is enabled before napi_complete(). A network timeout occurs if the interrupt handler is called before napi_complete(). Fix the bug by enabling the interrupt after napi_complete(). Signed-off-by: Yongbae Park Signed-off-by: David S. Miller --- drivers/net/ethernet/ibm/ibmveth.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/ibm/ibmveth.c b/drivers/net/ethernet/ibm/ibmveth.c index 072426a72745..cd7675ac5bf9 100644 --- a/drivers/net/ethernet/ibm/ibmveth.c +++ b/drivers/net/ethernet/ibm/ibmveth.c @@ -1136,6 +1136,8 @@ restart_poll: ibmveth_replenish_task(adapter); if (frames_processed < budget) { + napi_complete(napi); + /* We think we are done - reenable interrupts, * then check once more to make sure we are done. */ @@ -1144,8 +1146,6 @@ restart_poll: BUG_ON(lpar_rc != H_SUCCESS); - napi_complete(napi); - if (ibmveth_rxq_pending_buffer(adapter) && napi_reschedule(napi)) { lpar_rc = h_vio_signal(adapter->vdev->unit_address, -- cgit v1.2.1 From 5a3dba7a5fcc02b78d92c35e2ca53f21ae3402c9 Mon Sep 17 00:00:00 2001 From: Yongbae Park Date: Tue, 10 Mar 2015 11:35:07 +0900 Subject: net: WIZnet drivers: enable interrupts after napi_complete() The interrupt is enabled before napi_complete(). A network timeout occurs if the interrupt handler is called before napi_complete(). Fix the bug by enabling the interrupt after napi_complete(). Signed-off-by: Yongbae Park Signed-off-by: David S. Miller --- drivers/net/ethernet/wiznet/w5100.c | 2 +- drivers/net/ethernet/wiznet/w5300.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/wiznet/w5100.c b/drivers/net/ethernet/wiznet/w5100.c index a495931a66a1..0e0fbb5842b3 100644 --- a/drivers/net/ethernet/wiznet/w5100.c +++ b/drivers/net/ethernet/wiznet/w5100.c @@ -498,9 +498,9 @@ static int w5100_napi_poll(struct napi_struct *napi, int budget) } if (rx_count < budget) { + napi_complete(napi); w5100_write(priv, W5100_IMR, IR_S0); mmiowb(); - napi_complete(napi); } return rx_count; diff --git a/drivers/net/ethernet/wiznet/w5300.c b/drivers/net/ethernet/wiznet/w5300.c index 09322d9db578..4b310002258d 100644 --- a/drivers/net/ethernet/wiznet/w5300.c +++ b/drivers/net/ethernet/wiznet/w5300.c @@ -418,9 +418,9 @@ static int w5300_napi_poll(struct napi_struct *napi, int budget) } if (rx_count < budget) { + napi_complete(napi); w5300_write(priv, W5300_IMR, IR_S0); mmiowb(); - napi_complete(napi); } return rx_count; -- cgit v1.2.1 From 549e783f6a1504fcd24576302bc3818538b677f0 Mon Sep 17 00:00:00 2001 From: "qipeng.zha" Date: Tue, 3 Mar 2015 18:13:22 +0800 Subject: pinctrl: update direction_output function of cherryview driver From the comments of gpiod_direction_output(), need to set @value as initial output, so update the lowlevel routine to make it work. Signed-off-by: jason.cj.chen Signed-off-by: qipeng.zha Acked-by: Mika Westerberg Signed-off-by: Linus Walleij --- drivers/pinctrl/intel/pinctrl-cherryview.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/pinctrl/intel/pinctrl-cherryview.c b/drivers/pinctrl/intel/pinctrl-cherryview.c index 3034fd03bced..82f691eeeec4 100644 --- a/drivers/pinctrl/intel/pinctrl-cherryview.c +++ b/drivers/pinctrl/intel/pinctrl-cherryview.c @@ -1226,6 +1226,7 @@ static int chv_gpio_direction_input(struct gpio_chip *chip, unsigned offset) static int chv_gpio_direction_output(struct gpio_chip *chip, unsigned offset, int value) { + chv_gpio_set(chip, offset, value); return pinctrl_gpio_direction_output(chip->base + offset); } -- cgit v1.2.1 From af5cbc9822f6bbe399925760a4d5ee82c21f258c Mon Sep 17 00:00:00 2001 From: Nimrod Andy Date: Tue, 10 Mar 2015 19:09:41 +0800 Subject: net: fec: fix receive VLAN CTAG HW acceleration issue The current driver support receive VLAN CTAG HW acceleration feature (NETIF_F_HW_VLAN_CTAG_RX) through software simulation. There calls the api .skb_copy_to_linear_data_offset() to skip the VLAN tag, but there have overlap between the two memory data point range. The patch just fix the issue. V2: Michael Grzeschik suggest to use memmove() instead of skb_copy_to_linear_data_offset(). Reported-by: Michael Grzeschik Fixes: 1b7bde6d659d ("net: fec: implement rx_copybreak to improve rx performance") Signed-off-by: Fugang Duan Signed-off-by: David S. Miller --- drivers/net/ethernet/freescale/fec_main.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/drivers/net/ethernet/freescale/fec_main.c b/drivers/net/ethernet/freescale/fec_main.c index 99492b7e3713..787db5026191 100644 --- a/drivers/net/ethernet/freescale/fec_main.c +++ b/drivers/net/ethernet/freescale/fec_main.c @@ -1479,8 +1479,7 @@ fec_enet_rx_queue(struct net_device *ndev, int budget, u16 queue_id) vlan_packet_rcvd = true; - skb_copy_to_linear_data_offset(skb, VLAN_HLEN, - data, (2 * ETH_ALEN)); + memmove(skb->data + VLAN_HLEN, data, ETH_ALEN * 2); skb_pull(skb, VLAN_HLEN); } -- cgit v1.2.1 From e3d50738e59af9e58f569e54ff8af1840bea906c Mon Sep 17 00:00:00 2001 From: Hariprasad Shenai Date: Tue, 10 Mar 2015 17:44:52 +0530 Subject: cxgb4: fix coccinelle warnings Commit 16e47624e76b43db ("cxgb4: Add new scheme to update T4/T5 firmware") introduced below coccinelle warning. >> drivers/net/ethernet/chelsio/cxgb4/t4_hw.c:994:2-8: Replace memcpy with struct assignment Reported-by: Fengguang Wu Signed-off-by: Hariprasad Shenai Signed-off-by: David S. Miller --- drivers/net/ethernet/chelsio/cxgb4/t4_hw.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/chelsio/cxgb4/t4_hw.c b/drivers/net/ethernet/chelsio/cxgb4/t4_hw.c index 853c38997c82..1abdfa123c6c 100644 --- a/drivers/net/ethernet/chelsio/cxgb4/t4_hw.c +++ b/drivers/net/ethernet/chelsio/cxgb4/t4_hw.c @@ -1120,7 +1120,7 @@ int t4_prep_fw(struct adapter *adap, struct fw_info *fw_info, } /* Installed successfully, update the cached header too. */ - memcpy(card_fw, fs_fw, sizeof(*card_fw)); + *card_fw = *fs_fw; card_fw_usable = 1; *reset = 0; /* already reset as part of load_fw */ } -- cgit v1.2.1 From 8bf1268f48ad9bf5d6401b4db913e6d85b0863f6 Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 10 Mar 2015 16:41:35 +0000 Subject: ARM: dma-api: fix off-by-one error in __dma_supported() When validating the mask against the amount of memory we have available (so that we can trap 32-bit DMA addresses with >32-bits memory), we had not taken account of the fact that max_pfn is the maximum PFN number plus one that would be in the system. There are several references in the code which bear this out: mm/page_owner.c: for (; pfn < max_pfn; pfn++) { } arch/x86/kernel/setup.c: high_memory = (void *)__va(max_pfn * PAGE_SIZE - 1) Signed-off-by: Russell King --- arch/arm/mm/dma-mapping.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm/mm/dma-mapping.c b/arch/arm/mm/dma-mapping.c index 170a116d1b29..c27447653903 100644 --- a/arch/arm/mm/dma-mapping.c +++ b/arch/arm/mm/dma-mapping.c @@ -171,7 +171,7 @@ static int __dma_supported(struct device *dev, u64 mask, bool warn) */ if (sizeof(mask) != sizeof(dma_addr_t) && mask > (dma_addr_t)~0 && - dma_to_pfn(dev, ~0) < max_pfn) { + dma_to_pfn(dev, ~0) < max_pfn - 1) { if (warn) { dev_warn(dev, "Coherent DMA mask %#llx is larger than dma_addr_t allows\n", mask); -- cgit v1.2.1 From 6d021b724481fbb908eb29384898deb9f00dfe70 Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 10 Mar 2015 19:40:55 +0000 Subject: ARM: dump pgd, pmd and pte states on unhandled data abort faults It can be useful to dump the page table entries when an unhandled data abort fault occurs. This can aid debugging of these situations, for example, a STREX instruction causing an external abort on non-linefetch fault, as has been reported recently. Signed-off-by: Russell King --- arch/arm/mm/fault.c | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/arm/mm/fault.c b/arch/arm/mm/fault.c index a982dc3190df..6333d9c17875 100644 --- a/arch/arm/mm/fault.c +++ b/arch/arm/mm/fault.c @@ -552,6 +552,7 @@ do_DataAbort(unsigned long addr, unsigned int fsr, struct pt_regs *regs) pr_alert("Unhandled fault: %s (0x%03x) at 0x%08lx\n", inf->name, fsr, addr); + show_pte(current->mm, addr); info.si_signo = inf->sig; info.si_errno = 0; -- cgit v1.2.1 From 7768eed8bf1d2e5eefa38c573f15f737a9824052 Mon Sep 17 00:00:00 2001 From: Oliver Hartkopp Date: Tue, 10 Mar 2015 19:03:46 +0100 Subject: net: add comment for sock_efree() usage Signed-off-by: Oliver Hartkopp Acked-by: Alexander Duyck Signed-off-by: David S. Miller --- net/core/sock.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/net/core/sock.c b/net/core/sock.c index 93c8b20c91e4..78e89eb7eb70 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -1655,6 +1655,10 @@ void sock_rfree(struct sk_buff *skb) } EXPORT_SYMBOL(sock_rfree); +/* + * Buffer destructor for skbs that are not used directly in read or write + * path, e.g. for error handler skbs. Automatically called from kfree_skb. + */ void sock_efree(struct sk_buff *skb) { sock_put(skb->sk); -- cgit v1.2.1 From 2bf4c1d483d911cda5dd385527194d23e5cea73d Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 10 Mar 2015 12:39:03 +0100 Subject: ASoC: adav80x: Fix wrong value references for boolean kctl The correct values referred by a boolean control are value.integer.value[], not value.enumerated.item[]. The former is long while the latter is int, so it's even incompatible on 64bit architectures. Signed-off-by: Takashi Iwai Acked-by: Lars-Peter Clausen Signed-off-by: Mark Brown Cc: --- sound/soc/codecs/adav80x.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/sound/soc/codecs/adav80x.c b/sound/soc/codecs/adav80x.c index b67480f1b1aa..4373ada95648 100644 --- a/sound/soc/codecs/adav80x.c +++ b/sound/soc/codecs/adav80x.c @@ -317,7 +317,7 @@ static int adav80x_put_deemph(struct snd_kcontrol *kcontrol, { struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct adav80x *adav80x = snd_soc_codec_get_drvdata(codec); - unsigned int deemph = ucontrol->value.enumerated.item[0]; + unsigned int deemph = ucontrol->value.integer.value[0]; if (deemph > 1) return -EINVAL; @@ -333,7 +333,7 @@ static int adav80x_get_deemph(struct snd_kcontrol *kcontrol, struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct adav80x *adav80x = snd_soc_codec_get_drvdata(codec); - ucontrol->value.enumerated.item[0] = adav80x->deemph; + ucontrol->value.integer.value[0] = adav80x->deemph; return 0; }; -- cgit v1.2.1 From 08641d9b7bf915144a57a736b42642e13eb1167f Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 10 Mar 2015 12:39:04 +0100 Subject: ASoC: ak4641: Fix wrong value references for boolean kctl The correct values referred by a boolean control are value.integer.value[], not value.enumerated.item[]. The former is long while the latter is int, so it's even incompatible on 64bit architectures. Signed-off-by: Takashi Iwai Signed-off-by: Mark Brown Cc: --- sound/soc/codecs/ak4641.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/sound/soc/codecs/ak4641.c b/sound/soc/codecs/ak4641.c index 70861c7b1631..81b54a270bd8 100644 --- a/sound/soc/codecs/ak4641.c +++ b/sound/soc/codecs/ak4641.c @@ -76,7 +76,7 @@ static int ak4641_put_deemph(struct snd_kcontrol *kcontrol, { struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct ak4641_priv *ak4641 = snd_soc_codec_get_drvdata(codec); - int deemph = ucontrol->value.enumerated.item[0]; + int deemph = ucontrol->value.integer.value[0]; if (deemph > 1) return -EINVAL; @@ -92,7 +92,7 @@ static int ak4641_get_deemph(struct snd_kcontrol *kcontrol, struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct ak4641_priv *ak4641 = snd_soc_codec_get_drvdata(codec); - ucontrol->value.enumerated.item[0] = ak4641->deemph; + ucontrol->value.integer.value[0] = ak4641->deemph; return 0; }; -- cgit v1.2.1 From e8371aa0fecb73fb8a4b2e0296b025b11e7d6229 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 10 Mar 2015 12:39:05 +0100 Subject: ASoC: cs4271: Fix wrong value references for boolean kctl The correct values referred by a boolean control are value.integer.value[], not value.enumerated.item[]. The former is long while the latter is int, so it's even incompatible on 64bit architectures. Signed-off-by: Takashi Iwai Acked-by: Paul Handrigan Signed-off-by: Mark Brown Cc: --- sound/soc/codecs/cs4271.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/sound/soc/codecs/cs4271.c b/sound/soc/codecs/cs4271.c index 79a4efcb894c..7d3a6accaf9a 100644 --- a/sound/soc/codecs/cs4271.c +++ b/sound/soc/codecs/cs4271.c @@ -286,7 +286,7 @@ static int cs4271_get_deemph(struct snd_kcontrol *kcontrol, struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct cs4271_private *cs4271 = snd_soc_codec_get_drvdata(codec); - ucontrol->value.enumerated.item[0] = cs4271->deemph; + ucontrol->value.integer.value[0] = cs4271->deemph; return 0; } @@ -296,7 +296,7 @@ static int cs4271_put_deemph(struct snd_kcontrol *kcontrol, struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct cs4271_private *cs4271 = snd_soc_codec_get_drvdata(codec); - cs4271->deemph = ucontrol->value.enumerated.item[0]; + cs4271->deemph = ucontrol->value.integer.value[0]; return cs4271_set_deemph(codec); } -- cgit v1.2.1 From d223b0e7fcfecc23380e7de45eb6a0e7b328c17c Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 10 Mar 2015 12:39:06 +0100 Subject: ASoC: es8238: Fix wrong value references for boolean kctl The correct values referred by a boolean control are value.integer.value[], not value.enumerated.item[]. The former is long while the latter is int, so it's even incompatible on 64bit architectures. Signed-off-by: Takashi Iwai Signed-off-by: Mark Brown Cc: --- sound/soc/codecs/es8328.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/sound/soc/codecs/es8328.c b/sound/soc/codecs/es8328.c index f27325155ace..c5f35a07e8e4 100644 --- a/sound/soc/codecs/es8328.c +++ b/sound/soc/codecs/es8328.c @@ -120,7 +120,7 @@ static int es8328_get_deemph(struct snd_kcontrol *kcontrol, struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct es8328_priv *es8328 = snd_soc_codec_get_drvdata(codec); - ucontrol->value.enumerated.item[0] = es8328->deemph; + ucontrol->value.integer.value[0] = es8328->deemph; return 0; } @@ -129,7 +129,7 @@ static int es8328_put_deemph(struct snd_kcontrol *kcontrol, { struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct es8328_priv *es8328 = snd_soc_codec_get_drvdata(codec); - int deemph = ucontrol->value.enumerated.item[0]; + int deemph = ucontrol->value.integer.value[0]; int ret; if (deemph > 1) -- cgit v1.2.1 From d7f58db49d9ad92bdb12d21fdc2308b76bc2ed38 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 10 Mar 2015 12:39:07 +0100 Subject: ASoC: pcm1681: Fix wrong value references for boolean kctl The correct values referred by a boolean control are value.integer.value[], not value.enumerated.item[]. The former is long while the latter is int, so it's even incompatible on 64bit architectures. Signed-off-by: Takashi Iwai Signed-off-by: Mark Brown Cc: --- sound/soc/codecs/pcm1681.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/sound/soc/codecs/pcm1681.c b/sound/soc/codecs/pcm1681.c index a722a023c262..477e13d30971 100644 --- a/sound/soc/codecs/pcm1681.c +++ b/sound/soc/codecs/pcm1681.c @@ -118,7 +118,7 @@ static int pcm1681_get_deemph(struct snd_kcontrol *kcontrol, struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct pcm1681_private *priv = snd_soc_codec_get_drvdata(codec); - ucontrol->value.enumerated.item[0] = priv->deemph; + ucontrol->value.integer.value[0] = priv->deemph; return 0; } @@ -129,7 +129,7 @@ static int pcm1681_put_deemph(struct snd_kcontrol *kcontrol, struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct pcm1681_private *priv = snd_soc_codec_get_drvdata(codec); - priv->deemph = ucontrol->value.enumerated.item[0]; + priv->deemph = ucontrol->value.integer.value[0]; return pcm1681_set_deemph(codec); } -- cgit v1.2.1 From 4c523ef61160b7d478371ddc9f48c8ce0a00d675 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 10 Mar 2015 12:39:08 +0100 Subject: ASoC: tas5086: Fix wrong value references for boolean kctl The correct values referred by a boolean control are value.integer.value[], not value.enumerated.item[]. The former is long while the latter is int, so it's even incompatible on 64bit architectures. Signed-off-by: Takashi Iwai Signed-off-by: Mark Brown Cc: --- sound/soc/codecs/tas5086.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/sound/soc/codecs/tas5086.c b/sound/soc/codecs/tas5086.c index 249ef5c4c762..32942bed34b1 100644 --- a/sound/soc/codecs/tas5086.c +++ b/sound/soc/codecs/tas5086.c @@ -281,7 +281,7 @@ static int tas5086_get_deemph(struct snd_kcontrol *kcontrol, struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct tas5086_private *priv = snd_soc_codec_get_drvdata(codec); - ucontrol->value.enumerated.item[0] = priv->deemph; + ucontrol->value.integer.value[0] = priv->deemph; return 0; } @@ -292,7 +292,7 @@ static int tas5086_put_deemph(struct snd_kcontrol *kcontrol, struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct tas5086_private *priv = snd_soc_codec_get_drvdata(codec); - priv->deemph = ucontrol->value.enumerated.item[0]; + priv->deemph = ucontrol->value.integer.value[0]; return tas5086_set_deemph(codec); } -- cgit v1.2.1 From 00a14c2968e3d55817e0fa35c78106ca840537bf Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 10 Mar 2015 12:39:09 +0100 Subject: ASoC: wm2000: Fix wrong value references for boolean kctl The correct values referred by a boolean control are value.integer.value[], not value.enumerated.item[]. The former is long while the latter is int, so it's even incompatible on 64bit architectures. Signed-off-by: Takashi Iwai Acked-by: Charles Keepax Signed-off-by: Mark Brown Cc: --- sound/soc/codecs/wm2000.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/sound/soc/codecs/wm2000.c b/sound/soc/codecs/wm2000.c index 8d9de49a5052..21d5402e343f 100644 --- a/sound/soc/codecs/wm2000.c +++ b/sound/soc/codecs/wm2000.c @@ -610,7 +610,7 @@ static int wm2000_anc_mode_get(struct snd_kcontrol *kcontrol, struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct wm2000_priv *wm2000 = dev_get_drvdata(codec->dev); - ucontrol->value.enumerated.item[0] = wm2000->anc_active; + ucontrol->value.integer.value[0] = wm2000->anc_active; return 0; } @@ -620,7 +620,7 @@ static int wm2000_anc_mode_put(struct snd_kcontrol *kcontrol, { struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct wm2000_priv *wm2000 = dev_get_drvdata(codec->dev); - int anc_active = ucontrol->value.enumerated.item[0]; + int anc_active = ucontrol->value.integer.value[0]; int ret; if (anc_active > 1) @@ -643,7 +643,7 @@ static int wm2000_speaker_get(struct snd_kcontrol *kcontrol, struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct wm2000_priv *wm2000 = dev_get_drvdata(codec->dev); - ucontrol->value.enumerated.item[0] = wm2000->spk_ena; + ucontrol->value.integer.value[0] = wm2000->spk_ena; return 0; } @@ -653,7 +653,7 @@ static int wm2000_speaker_put(struct snd_kcontrol *kcontrol, { struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct wm2000_priv *wm2000 = dev_get_drvdata(codec->dev); - int val = ucontrol->value.enumerated.item[0]; + int val = ucontrol->value.integer.value[0]; int ret; if (val > 1) -- cgit v1.2.1 From bd14016fbf31aa199026f1e2358eab695f374eb1 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 10 Mar 2015 12:39:10 +0100 Subject: ASoC: wm8731: Fix wrong value references for boolean kctl The correct values referred by a boolean control are value.integer.value[], not value.enumerated.item[]. The former is long while the latter is int, so it's even incompatible on 64bit architectures. Signed-off-by: Takashi Iwai Acked-by: Charles Keepax Signed-off-by: Mark Brown Cc: --- sound/soc/codecs/wm8731.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/sound/soc/codecs/wm8731.c b/sound/soc/codecs/wm8731.c index 098c143f44d6..c6d10533e2bd 100644 --- a/sound/soc/codecs/wm8731.c +++ b/sound/soc/codecs/wm8731.c @@ -125,7 +125,7 @@ static int wm8731_get_deemph(struct snd_kcontrol *kcontrol, struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct wm8731_priv *wm8731 = snd_soc_codec_get_drvdata(codec); - ucontrol->value.enumerated.item[0] = wm8731->deemph; + ucontrol->value.integer.value[0] = wm8731->deemph; return 0; } @@ -135,7 +135,7 @@ static int wm8731_put_deemph(struct snd_kcontrol *kcontrol, { struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct wm8731_priv *wm8731 = snd_soc_codec_get_drvdata(codec); - int deemph = ucontrol->value.enumerated.item[0]; + int deemph = ucontrol->value.integer.value[0]; int ret = 0; if (deemph > 1) -- cgit v1.2.1 From 24cc883c1fd16df34211ae41624aa6d3cd906693 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 10 Mar 2015 12:39:11 +0100 Subject: ASoC: wm8903: Fix wrong value references for boolean kctl The correct values referred by a boolean control are value.integer.value[], not value.enumerated.item[]. The former is long while the latter is int, so it's even incompatible on 64bit architectures. Signed-off-by: Takashi Iwai Acked-by: Charles Keepax Signed-off-by: Mark Brown Cc: --- sound/soc/codecs/wm8903.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/sound/soc/codecs/wm8903.c b/sound/soc/codecs/wm8903.c index dde462c082be..04b04f8e147c 100644 --- a/sound/soc/codecs/wm8903.c +++ b/sound/soc/codecs/wm8903.c @@ -442,7 +442,7 @@ static int wm8903_get_deemph(struct snd_kcontrol *kcontrol, struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct wm8903_priv *wm8903 = snd_soc_codec_get_drvdata(codec); - ucontrol->value.enumerated.item[0] = wm8903->deemph; + ucontrol->value.integer.value[0] = wm8903->deemph; return 0; } @@ -452,7 +452,7 @@ static int wm8903_put_deemph(struct snd_kcontrol *kcontrol, { struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct wm8903_priv *wm8903 = snd_soc_codec_get_drvdata(codec); - int deemph = ucontrol->value.enumerated.item[0]; + int deemph = ucontrol->value.integer.value[0]; int ret = 0; if (deemph > 1) -- cgit v1.2.1 From eaddf6fd959074f6a6e71deffe079c71eef35da6 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 10 Mar 2015 12:39:12 +0100 Subject: ASoC: wm8904: Fix wrong value references for boolean kctl The correct values referred by a boolean control are value.integer.value[], not value.enumerated.item[]. The former is long while the latter is int, so it's even incompatible on 64bit architectures. Signed-off-by: Takashi Iwai Acked-by: Charles Keepax Signed-off-by: Mark Brown Cc: --- sound/soc/codecs/wm8904.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/sound/soc/codecs/wm8904.c b/sound/soc/codecs/wm8904.c index d3b3f57668cc..215e93c1ddf0 100644 --- a/sound/soc/codecs/wm8904.c +++ b/sound/soc/codecs/wm8904.c @@ -525,7 +525,7 @@ static int wm8904_get_deemph(struct snd_kcontrol *kcontrol, struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct wm8904_priv *wm8904 = snd_soc_codec_get_drvdata(codec); - ucontrol->value.enumerated.item[0] = wm8904->deemph; + ucontrol->value.integer.value[0] = wm8904->deemph; return 0; } @@ -534,7 +534,7 @@ static int wm8904_put_deemph(struct snd_kcontrol *kcontrol, { struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct wm8904_priv *wm8904 = snd_soc_codec_get_drvdata(codec); - int deemph = ucontrol->value.enumerated.item[0]; + int deemph = ucontrol->value.integer.value[0]; if (deemph > 1) return -EINVAL; -- cgit v1.2.1 From 07892b10356f17717abdc578acbef72db86c880e Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 10 Mar 2015 12:39:13 +0100 Subject: ASoC: wm8955: Fix wrong value references for boolean kctl The correct values referred by a boolean control are value.integer.value[], not value.enumerated.item[]. The former is long while the latter is int, so it's even incompatible on 64bit architectures. Signed-off-by: Takashi Iwai Acked-by: Charles Keepax Signed-off-by: Mark Brown Cc: --- sound/soc/codecs/wm8955.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/sound/soc/codecs/wm8955.c b/sound/soc/codecs/wm8955.c index 1ab2d462afad..00bec915d652 100644 --- a/sound/soc/codecs/wm8955.c +++ b/sound/soc/codecs/wm8955.c @@ -393,7 +393,7 @@ static int wm8955_get_deemph(struct snd_kcontrol *kcontrol, struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct wm8955_priv *wm8955 = snd_soc_codec_get_drvdata(codec); - ucontrol->value.enumerated.item[0] = wm8955->deemph; + ucontrol->value.integer.value[0] = wm8955->deemph; return 0; } @@ -402,7 +402,7 @@ static int wm8955_put_deemph(struct snd_kcontrol *kcontrol, { struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct wm8955_priv *wm8955 = snd_soc_codec_get_drvdata(codec); - int deemph = ucontrol->value.enumerated.item[0]; + int deemph = ucontrol->value.integer.value[0]; if (deemph > 1) return -EINVAL; -- cgit v1.2.1 From b4a18c8b1af15ebfa9054a3d2aef7b0a7e6f2a05 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 10 Mar 2015 12:39:14 +0100 Subject: ASoC: wm8960: Fix wrong value references for boolean kctl The correct values referred by a boolean control are value.integer.value[], not value.enumerated.item[]. The former is long while the latter is int, so it's even incompatible on 64bit architectures. Signed-off-by: Takashi Iwai Acked-by: Charles Keepax Signed-off-by: Mark Brown Cc: --- sound/soc/codecs/wm8960.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/sound/soc/codecs/wm8960.c b/sound/soc/codecs/wm8960.c index cf8fecf97f2c..3035d9856415 100644 --- a/sound/soc/codecs/wm8960.c +++ b/sound/soc/codecs/wm8960.c @@ -184,7 +184,7 @@ static int wm8960_get_deemph(struct snd_kcontrol *kcontrol, struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct wm8960_priv *wm8960 = snd_soc_codec_get_drvdata(codec); - ucontrol->value.enumerated.item[0] = wm8960->deemph; + ucontrol->value.integer.value[0] = wm8960->deemph; return 0; } @@ -193,7 +193,7 @@ static int wm8960_put_deemph(struct snd_kcontrol *kcontrol, { struct snd_soc_codec *codec = snd_soc_kcontrol_codec(kcontrol); struct wm8960_priv *wm8960 = snd_soc_codec_get_drvdata(codec); - int deemph = ucontrol->value.enumerated.item[0]; + int deemph = ucontrol->value.integer.value[0]; if (deemph > 1) return -EINVAL; -- cgit v1.2.1 From 4b0b669b86a963f71feaa1a694e881832fdf4f86 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 10 Mar 2015 12:39:15 +0100 Subject: ASoC: wm9712: Fix wrong value references for boolean kctl The correct values referred by a boolean control are value.integer.value[], not value.enumerated.item[]. The former is long while the latter is int, so it's even incompatible on 64bit architectures. Signed-off-by: Takashi Iwai Acked-by: Charles Keepax Signed-off-by: Mark Brown Cc: --- sound/soc/codecs/wm9712.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/sound/soc/codecs/wm9712.c b/sound/soc/codecs/wm9712.c index 9517571e820d..98c9525bd751 100644 --- a/sound/soc/codecs/wm9712.c +++ b/sound/soc/codecs/wm9712.c @@ -180,7 +180,7 @@ static int wm9712_hp_mixer_put(struct snd_kcontrol *kcontrol, struct snd_soc_dapm_context *dapm = snd_soc_dapm_kcontrol_dapm(kcontrol); struct snd_soc_codec *codec = snd_soc_dapm_to_codec(dapm); struct wm9712_priv *wm9712 = snd_soc_codec_get_drvdata(codec); - unsigned int val = ucontrol->value.enumerated.item[0]; + unsigned int val = ucontrol->value.integer.value[0]; struct soc_mixer_control *mc = (struct soc_mixer_control *)kcontrol->private_value; unsigned int mixer, mask, shift, old; @@ -193,7 +193,7 @@ static int wm9712_hp_mixer_put(struct snd_kcontrol *kcontrol, mutex_lock(&wm9712->lock); old = wm9712->hp_mixer[mixer]; - if (ucontrol->value.enumerated.item[0]) + if (ucontrol->value.integer.value[0]) wm9712->hp_mixer[mixer] |= mask; else wm9712->hp_mixer[mixer] &= ~mask; @@ -231,7 +231,7 @@ static int wm9712_hp_mixer_get(struct snd_kcontrol *kcontrol, mixer = mc->shift >> 8; shift = mc->shift & 0xff; - ucontrol->value.enumerated.item[0] = + ucontrol->value.integer.value[0] = (wm9712->hp_mixer[mixer] >> shift) & 1; return 0; -- cgit v1.2.1 From 87a8b286e2f63c048a586dc677140d4a5b5808aa Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Tue, 10 Mar 2015 12:39:16 +0100 Subject: ASoC: wm9713: Fix wrong value references for boolean kctl The correct values referred by a boolean control are value.integer.value[], not value.enumerated.item[]. The former is long while the latter is int, so it's even incompatible on 64bit architectures. Signed-off-by: Takashi Iwai Acked-by: Charles Keepax Signed-off-by: Mark Brown Cc: --- sound/soc/codecs/wm9713.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/sound/soc/codecs/wm9713.c b/sound/soc/codecs/wm9713.c index 68222917b396..79552953e1bd 100644 --- a/sound/soc/codecs/wm9713.c +++ b/sound/soc/codecs/wm9713.c @@ -255,7 +255,7 @@ static int wm9713_hp_mixer_put(struct snd_kcontrol *kcontrol, struct snd_soc_dapm_context *dapm = snd_soc_dapm_kcontrol_dapm(kcontrol); struct snd_soc_codec *codec = snd_soc_dapm_to_codec(dapm); struct wm9713_priv *wm9713 = snd_soc_codec_get_drvdata(codec); - unsigned int val = ucontrol->value.enumerated.item[0]; + unsigned int val = ucontrol->value.integer.value[0]; struct soc_mixer_control *mc = (struct soc_mixer_control *)kcontrol->private_value; unsigned int mixer, mask, shift, old; @@ -268,7 +268,7 @@ static int wm9713_hp_mixer_put(struct snd_kcontrol *kcontrol, mutex_lock(&wm9713->lock); old = wm9713->hp_mixer[mixer]; - if (ucontrol->value.enumerated.item[0]) + if (ucontrol->value.integer.value[0]) wm9713->hp_mixer[mixer] |= mask; else wm9713->hp_mixer[mixer] &= ~mask; @@ -306,7 +306,7 @@ static int wm9713_hp_mixer_get(struct snd_kcontrol *kcontrol, mixer = mc->shift >> 8; shift = mc->shift & 0xff; - ucontrol->value.enumerated.item[0] = + ucontrol->value.integer.value[0] = (wm9713->hp_mixer[mixer] >> shift) & 1; return 0; -- cgit v1.2.1 From 509d612b2fc4b66a58f1af762ac69829ed11c0ce Mon Sep 17 00:00:00 2001 From: Yunzhi Li Date: Fri, 6 Mar 2015 15:17:10 +0800 Subject: usb: dwc2: host: fix dwc2 disconnect bug When dwc2 controller detects a disconnect interrupt, dwc2_hcd_disconnect() should be called immediately to do clean-up jobs and set port_connect_status_change flag to notify usb hub driver disconnect status. Tested-by: Vincent Palatin Acked-by: John Youn Signed-off-by: Yunzhi Li Signed-off-by: Felipe Balbi --- drivers/usb/dwc2/core_intr.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/usb/dwc2/core_intr.c b/drivers/usb/dwc2/core_intr.c index 02e3e2d4ea56..6cf047878dba 100644 --- a/drivers/usb/dwc2/core_intr.c +++ b/drivers/usb/dwc2/core_intr.c @@ -377,6 +377,9 @@ static void dwc2_handle_disconnect_intr(struct dwc2_hsotg *hsotg) dwc2_is_host_mode(hsotg) ? "Host" : "Device", dwc2_op_state_str(hsotg)); + if (hsotg->op_state == OTG_STATE_A_HOST) + dwc2_hcd_disconnect(hsotg); + /* Change to L3 (OFF) state */ hsotg->lx_state = DWC2_L3; -- cgit v1.2.1 From d0f347d62814ec0f599a05c61c5619d5e999e4ae Mon Sep 17 00:00:00 2001 From: David Dueck Date: Sun, 8 Feb 2015 16:29:30 +0100 Subject: usb: phy: am335x-control: check return value of bus_find_device This fixes a potential null pointer dereference. Cc: # v3.16+ Fixes: d4332013919a ("driver core: dev_get_drvdata: Don't check for NULL dev") Acked-by: Sebastian Andrzej Siewior Signed-off-by: David Dueck Signed-off-by: Felipe Balbi --- drivers/usb/phy/phy-am335x-control.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/usb/phy/phy-am335x-control.c b/drivers/usb/phy/phy-am335x-control.c index 403fab772724..7b3035ff9434 100644 --- a/drivers/usb/phy/phy-am335x-control.c +++ b/drivers/usb/phy/phy-am335x-control.c @@ -126,6 +126,9 @@ struct phy_control *am335x_get_phy_control(struct device *dev) return NULL; dev = bus_find_device(&platform_bus_type, NULL, node, match); + if (!dev) + return NULL; + ctrl_usb = dev_get_drvdata(dev); if (!ctrl_usb) return NULL; -- cgit v1.2.1 From dc9be0fac70a2ad86e31a81372bb0bdfb6945353 Mon Sep 17 00:00:00 2001 From: Paolo Bonzini Date: Thu, 5 Mar 2015 11:54:46 +0100 Subject: kvm: move advertising of KVM_CAP_IRQFD to common code POWER supports irqfds but forgot to advertise them. Some userspace does not check for the capability, but others check it---thus they work on x86 and s390 but not POWER. To avoid that other architectures in the future make the same mistake, let common code handle KVM_CAP_IRQFD the same way as KVM_CAP_IRQFD_RESAMPLE. Reported-and-tested-by: Greg Kurz Cc: stable@vger.kernel.org Fixes: 297e21053a52f060944e9f0de4c64fad9bcd72fc Signed-off-by: Paolo Bonzini Signed-off-by: Marcelo Tosatti --- arch/s390/kvm/kvm-s390.c | 1 - arch/x86/kvm/x86.c | 1 - virt/kvm/kvm_main.c | 1 + 3 files changed, 1 insertion(+), 2 deletions(-) diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c index f6579cfde2df..19e17bd7aec0 100644 --- a/arch/s390/kvm/kvm-s390.c +++ b/arch/s390/kvm/kvm-s390.c @@ -165,7 +165,6 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext) case KVM_CAP_ONE_REG: case KVM_CAP_ENABLE_CAP: case KVM_CAP_S390_CSS_SUPPORT: - case KVM_CAP_IRQFD: case KVM_CAP_IOEVENTFD: case KVM_CAP_DEVICE_CTRL: case KVM_CAP_ENABLE_CAP_VM: diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c index bd7a70be41b3..32bf19ef3115 100644 --- a/arch/x86/kvm/x86.c +++ b/arch/x86/kvm/x86.c @@ -2744,7 +2744,6 @@ int kvm_vm_ioctl_check_extension(struct kvm *kvm, long ext) case KVM_CAP_USER_NMI: case KVM_CAP_REINJECT_CONTROL: case KVM_CAP_IRQ_INJECT_STATUS: - case KVM_CAP_IRQFD: case KVM_CAP_IOEVENTFD: case KVM_CAP_IOEVENTFD_NO_LENGTH: case KVM_CAP_PIT2: diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index a1093700f3a4..a2214d9609bd 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -2492,6 +2492,7 @@ static long kvm_vm_ioctl_check_extension_generic(struct kvm *kvm, long arg) case KVM_CAP_SIGNAL_MSI: #endif #ifdef CONFIG_HAVE_KVM_IRQFD + case KVM_CAP_IRQFD: case KVM_CAP_IRQFD_RESAMPLE: #endif case KVM_CAP_CHECK_EXTENSION_VM: -- cgit v1.2.1 From 4363890079674db7b00cf1bb0e6fa430e846e86b Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Tue, 10 Mar 2015 21:58:32 -0400 Subject: net: Handle unregister properly when netdev namespace change fails. If rtnl_newlink() fails on it's call to dev_change_net_namespace(), we have to make use of the ->dellink() method, if present, just like we do when rtnl_configure_link() fails. Fixes: 317f4810e45e ("rtnl: allow to create device with IFLA_LINK_NETNSID set") Signed-off-by: David S. Miller --- net/core/rtnetlink.c | 26 +++++++++++++------------- 1 file changed, 13 insertions(+), 13 deletions(-) diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c index 25b4b5d23485..ee0608bb3bc0 100644 --- a/net/core/rtnetlink.c +++ b/net/core/rtnetlink.c @@ -2166,28 +2166,28 @@ replay: } } err = rtnl_configure_link(dev, ifm); - if (err < 0) { - if (ops->newlink) { - LIST_HEAD(list_kill); - - ops->dellink(dev, &list_kill); - unregister_netdevice_many(&list_kill); - } else { - unregister_netdevice(dev); - } - goto out; - } - + if (err < 0) + goto out_unregister; if (link_net) { err = dev_change_net_namespace(dev, dest_net, ifname); if (err < 0) - unregister_netdevice(dev); + goto out_unregister; } out: if (link_net) put_net(link_net); put_net(dest_net); return err; +out_unregister: + if (ops->newlink) { + LIST_HEAD(list_kill); + + ops->dellink(dev, &list_kill); + unregister_netdevice_many(&list_kill); + } else { + unregister_netdevice(dev); + } + goto out; } } -- cgit v1.2.1 From af69decc7ca80fa24c77a5a00234f1cd2a957de8 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Fri, 27 Feb 2015 11:50:57 +0800 Subject: phy: exynos-mipi-video: Use spin_lock to protct state->regmap rmw operations The state->regmap is initialized by devm_regmap_init_mmio(). So it's fine to use spin_lock rather than mutex to protct state->regmap rmw operations. Signed-off-by: Axel Lin Acked-by: Sylwester Nawrocki Tested-by: Sylwester Nawrocki [Julia.Lawall@lip6.fr: Found an issue with the original patch w.r.t unbalanced spin_lock call] Signed-off-by: Julia Lawall Signed-off-by: Kishon Vijay Abraham I --- drivers/phy/phy-exynos-mipi-video.c | 9 +++------ 1 file changed, 3 insertions(+), 6 deletions(-) diff --git a/drivers/phy/phy-exynos-mipi-video.c b/drivers/phy/phy-exynos-mipi-video.c index d19649328d05..df7519a39ba0 100644 --- a/drivers/phy/phy-exynos-mipi-video.c +++ b/drivers/phy/phy-exynos-mipi-video.c @@ -43,7 +43,6 @@ struct exynos_mipi_video_phy { } phys[EXYNOS_MIPI_PHYS_NUM]; spinlock_t slock; void __iomem *regs; - struct mutex mutex; struct regmap *regmap; }; @@ -59,8 +58,9 @@ static int __set_phy_state(struct exynos_mipi_video_phy *state, else reset = EXYNOS4_MIPI_PHY_SRESETN; + spin_lock(&state->slock); + if (!IS_ERR(state->regmap)) { - mutex_lock(&state->mutex); regmap_read(state->regmap, offset, &val); if (on) val |= reset; @@ -72,11 +72,9 @@ static int __set_phy_state(struct exynos_mipi_video_phy *state, else if (!(val & EXYNOS4_MIPI_PHY_RESET_MASK)) val &= ~EXYNOS4_MIPI_PHY_ENABLE; regmap_write(state->regmap, offset, val); - mutex_unlock(&state->mutex); } else { addr = state->regs + EXYNOS_MIPI_PHY_CONTROL(id / 2); - spin_lock(&state->slock); val = readl(addr); if (on) val |= reset; @@ -90,9 +88,9 @@ static int __set_phy_state(struct exynos_mipi_video_phy *state, val &= ~EXYNOS4_MIPI_PHY_ENABLE; writel(val, addr); - spin_unlock(&state->slock); } + spin_unlock(&state->slock); return 0; } @@ -158,7 +156,6 @@ static int exynos_mipi_video_phy_probe(struct platform_device *pdev) dev_set_drvdata(dev, state); spin_lock_init(&state->slock); - mutex_init(&state->mutex); for (i = 0; i < EXYNOS_MIPI_PHYS_NUM; i++) { struct phy *phy = devm_phy_create(dev, NULL, -- cgit v1.2.1 From d0167ad2954ee2d1c70704c454c646086b6653d6 Mon Sep 17 00:00:00 2001 From: Mathias Nyman Date: Tue, 10 Mar 2015 19:49:00 +0200 Subject: Revert "xhci: Clear the host side toggle manually when endpoint is 'soft reset'" This reverts commit 27082e2654dc ("xhci: Clear the host side toggle manually") Turns out this fix to enable soft resetting endpoints wasn't mature enough. It caused regression with some usb DVB-T devices and needs some more tuning to get the endpiont ring pointers set correctly. The original commit was tagged for stable 3.18, and should be reverted from there as well. Cc: stable # v3.18 Signed-off-by: Mathias Nyman Signed-off-by: Greg Kroah-Hartman --- drivers/usb/host/xhci-ring.c | 2 +- drivers/usb/host/xhci.c | 100 ++++--------------------------------------- drivers/usb/host/xhci.h | 2 - 3 files changed, 10 insertions(+), 94 deletions(-) diff --git a/drivers/usb/host/xhci-ring.c b/drivers/usb/host/xhci-ring.c index 5fb66db89e05..73485fa4372f 100644 --- a/drivers/usb/host/xhci-ring.c +++ b/drivers/usb/host/xhci-ring.c @@ -1729,7 +1729,7 @@ static void xhci_cleanup_halted_endpoint(struct xhci_hcd *xhci, if (!command) return; - ep->ep_state |= EP_HALTED | EP_RECENTLY_HALTED; + ep->ep_state |= EP_HALTED; ep->stopped_stream = stream_id; xhci_queue_reset_ep(xhci, command, slot_id, ep_index); diff --git a/drivers/usb/host/xhci.c b/drivers/usb/host/xhci.c index b06d1a53652d..ec8ac1674854 100644 --- a/drivers/usb/host/xhci.c +++ b/drivers/usb/host/xhci.c @@ -1338,12 +1338,6 @@ int xhci_urb_enqueue(struct usb_hcd *hcd, struct urb *urb, gfp_t mem_flags) goto exit; } - /* Reject urb if endpoint is in soft reset, queue must stay empty */ - if (xhci->devs[slot_id]->eps[ep_index].ep_state & EP_CONFIG_PENDING) { - xhci_warn(xhci, "Can't enqueue URB while ep is in soft reset\n"); - ret = -EINVAL; - } - if (usb_endpoint_xfer_isoc(&urb->ep->desc)) size = urb->number_of_packets; else @@ -2954,36 +2948,23 @@ void xhci_cleanup_stalled_ring(struct xhci_hcd *xhci, } } -/* Called after clearing a halted device. USB core should have sent the control +/* Called when clearing halted device. The core should have sent the control * message to clear the device halt condition. The host side of the halt should - * already be cleared with a reset endpoint command issued immediately when the - * STALL tx event was received. + * already be cleared with a reset endpoint command issued when the STALL tx + * event was received. + * + * Context: in_interrupt */ void xhci_endpoint_reset(struct usb_hcd *hcd, struct usb_host_endpoint *ep) { struct xhci_hcd *xhci; - struct usb_device *udev; - struct xhci_virt_device *virt_dev; - struct xhci_virt_ep *virt_ep; - struct xhci_input_control_ctx *ctrl_ctx; - struct xhci_command *command; - unsigned int ep_index, ep_state; - unsigned long flags; - u32 ep_flag; xhci = hcd_to_xhci(hcd); - udev = (struct usb_device *) ep->hcpriv; - if (!ep->hcpriv) - return; - virt_dev = xhci->devs[udev->slot_id]; - ep_index = xhci_get_endpoint_index(&ep->desc); - virt_ep = &virt_dev->eps[ep_index]; - ep_state = virt_ep->ep_state; /* - * Implement the config ep command in xhci 4.6.8 additional note: + * We might need to implement the config ep cmd in xhci 4.8.1 note: * The Reset Endpoint Command may only be issued to endpoints in the * Halted state. If software wishes reset the Data Toggle or Sequence * Number of an endpoint that isn't in the Halted state, then software @@ -2991,72 +2972,9 @@ void xhci_endpoint_reset(struct usb_hcd *hcd, * for the target endpoint. that is in the Stopped state. */ - if (ep_state & SET_DEQ_PENDING || ep_state & EP_RECENTLY_HALTED) { - virt_ep->ep_state &= ~EP_RECENTLY_HALTED; - xhci_dbg(xhci, "ep recently halted, no toggle reset needed\n"); - return; - } - - /* Only interrupt and bulk ep's use Data toggle, USB2 spec 5.5.4-> */ - if (usb_endpoint_xfer_control(&ep->desc) || - usb_endpoint_xfer_isoc(&ep->desc)) - return; - - ep_flag = xhci_get_endpoint_flag(&ep->desc); - - if (ep_flag == SLOT_FLAG || ep_flag == EP0_FLAG) - return; - - command = xhci_alloc_command(xhci, true, true, GFP_NOWAIT); - if (!command) { - xhci_err(xhci, "Could not allocate xHCI command structure.\n"); - return; - } - - spin_lock_irqsave(&xhci->lock, flags); - - /* block ringing ep doorbell */ - virt_ep->ep_state |= EP_CONFIG_PENDING; - - /* - * Make sure endpoint ring is empty before resetting the toggle/seq. - * Driver is required to synchronously cancel all transfer request. - * - * xhci 4.6.6 says we can issue a configure endpoint command on a - * running endpoint ring as long as it's idle (queue empty) - */ - - if (!list_empty(&virt_ep->ring->td_list)) { - dev_err(&udev->dev, "EP not empty, refuse reset\n"); - spin_unlock_irqrestore(&xhci->lock, flags); - goto cleanup; - } - - xhci_dbg(xhci, "Reset toggle/seq for slot %d, ep_index: %d\n", - udev->slot_id, ep_index); - - ctrl_ctx = xhci_get_input_control_ctx(command->in_ctx); - if (!ctrl_ctx) { - xhci_err(xhci, "Could not get input context, bad type. virt_dev: %p, in_ctx %p\n", - virt_dev, virt_dev->in_ctx); - spin_unlock_irqrestore(&xhci->lock, flags); - goto cleanup; - } - xhci_setup_input_ctx_for_config_ep(xhci, command->in_ctx, - virt_dev->out_ctx, ctrl_ctx, - ep_flag, ep_flag); - xhci_endpoint_copy(xhci, command->in_ctx, virt_dev->out_ctx, ep_index); - - xhci_queue_configure_endpoint(xhci, command, command->in_ctx->dma, - udev->slot_id, false); - xhci_ring_cmd_db(xhci); - spin_unlock_irqrestore(&xhci->lock, flags); - - wait_for_completion(command->completion); - -cleanup: - virt_ep->ep_state &= ~EP_CONFIG_PENDING; - xhci_free_command(xhci, command); + /* For now just print debug to follow the situation */ + xhci_dbg(xhci, "Endpoint 0x%x ep reset callback called\n", + ep->desc.bEndpointAddress); } static int xhci_check_streams_endpoint(struct xhci_hcd *xhci, diff --git a/drivers/usb/host/xhci.h b/drivers/usb/host/xhci.h index 265ab1771d24..8e421b89632d 100644 --- a/drivers/usb/host/xhci.h +++ b/drivers/usb/host/xhci.h @@ -865,8 +865,6 @@ struct xhci_virt_ep { #define EP_HAS_STREAMS (1 << 4) /* Transitioning the endpoint to not using streams, don't enqueue URBs */ #define EP_GETTING_NO_STREAMS (1 << 5) -#define EP_RECENTLY_HALTED (1 << 6) -#define EP_CONFIG_PENDING (1 << 7) /* ---- Related to URB cancellation ---- */ struct list_head cancelled_td_list; struct xhci_td *stopped_td; -- cgit v1.2.1 From b2c08ba27fe9940e5338cf1852d54a4588f80370 Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Wed, 11 Mar 2015 14:15:10 +0100 Subject: Revert "pcmcia: add missing include for new pci resource handler" This reverts commit d885d4f3728f386034bb2f7a61b7f2054c49b2d4 as the patch that it fixes is about to be reverted. Reported-by: Alan Cox Cc: Arnd Bergmann Cc: Jim Davis Signed-off-by: Greg Kroah-Hartman --- drivers/pcmcia/rsrc_pci.c | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/pcmcia/rsrc_pci.c b/drivers/pcmcia/rsrc_pci.c index 1f67b3ba70fb..d11b7d408ed6 100644 --- a/drivers/pcmcia/rsrc_pci.c +++ b/drivers/pcmcia/rsrc_pci.c @@ -1,7 +1,6 @@ #include #include #include -#include #include #include -- cgit v1.2.1 From 05c0006776374a1013bc30a7daeee3f8017147f2 Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Wed, 11 Mar 2015 14:20:51 +0100 Subject: Revert "pcmcia: fix incorrect bracketing on a test" This reverts commit c3762b248faf9db2b00b36c0535f79758942069e. The file this fixes is about to be reverted. Reported-by: Alan Cox Signed-off-by: Greg Kroah-Hartman --- drivers/pcmcia/rsrc_pci.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/pcmcia/rsrc_pci.c b/drivers/pcmcia/rsrc_pci.c index d11b7d408ed6..8934d3c01f80 100644 --- a/drivers/pcmcia/rsrc_pci.c +++ b/drivers/pcmcia/rsrc_pci.c @@ -155,7 +155,7 @@ static struct resource *res_pci_find_mem(u_long base, u_long num, static int res_pci_init(struct pcmcia_socket *s) { - if (!s->cb_dev || !(s->features & SS_CAP_PAGE_REGS)) { + if (!s->cb_dev || (!s->features & SS_CAP_PAGE_REGS)) { dev_err(&s->dev, "not supported by res_pci\n"); return -EOPNOTSUPP; } -- cgit v1.2.1 From 3d7a8278fdfdea5be4c647853171a0df5d13c1d3 Mon Sep 17 00:00:00 2001 From: Greg Kroah-Hartman Date: Wed, 11 Mar 2015 14:21:23 +0100 Subject: Revert "pcmcia: add a new resource manager for non ISA systems" This reverts commit 02b03846bb2befc558bfd0665749d6bb26f4c2f1. Alan writes: it seems there is a regression in there for some configuration of I/O based devices. I'll take a look at it over the next couple of kernel releases and see what is up then resubmit it with fixes. Reported-by: Alan Cox Signed-off-by: Greg Kroah-Hartman --- drivers/pcmcia/Kconfig | 12 +--- drivers/pcmcia/Makefile | 1 - drivers/pcmcia/rsrc_pci.c | 172 ---------------------------------------------- 3 files changed, 3 insertions(+), 182 deletions(-) delete mode 100644 drivers/pcmcia/rsrc_pci.c diff --git a/drivers/pcmcia/Kconfig b/drivers/pcmcia/Kconfig index 3bb49252a098..45f67c63d385 100644 --- a/drivers/pcmcia/Kconfig +++ b/drivers/pcmcia/Kconfig @@ -69,8 +69,7 @@ config YENTA tristate "CardBus yenta-compatible bridge support" depends on PCI select CARDBUS if !EXPERT - select PCCARD_NONSTATIC if PCMCIA != n && ISA - select PCCARD_PCI if PCMCIA !=n && !ISA + select PCCARD_NONSTATIC if PCMCIA != n ---help--- This option enables support for CardBus host bridges. Virtually all modern PCMCIA bridges are CardBus compatible. A "bridge" is @@ -110,8 +109,7 @@ config YENTA_TOSHIBA config PD6729 tristate "Cirrus PD6729 compatible bridge support" depends on PCMCIA && PCI - select PCCARD_NONSTATIC if PCMCIA != n && ISA - select PCCARD_PCI if PCMCIA !=n && !ISA + select PCCARD_NONSTATIC help This provides support for the Cirrus PD6729 PCI-to-PCMCIA bridge device, found in some older laptops and PCMCIA card readers. @@ -119,8 +117,7 @@ config PD6729 config I82092 tristate "i82092 compatible bridge support" depends on PCMCIA && PCI - select PCCARD_NONSTATIC if PCMCIA != n && ISA - select PCCARD_PCI if PCMCIA !=n && !ISA + select PCCARD_NONSTATIC help This provides support for the Intel I82092AA PCI-to-PCMCIA bridge device, found in some older laptops and more commonly in evaluation boards for the @@ -291,9 +288,6 @@ config ELECTRA_CF Say Y here to support the CompactFlash controller on the PA Semi Electra eval board. -config PCCARD_PCI - bool - config PCCARD_NONSTATIC bool diff --git a/drivers/pcmcia/Makefile b/drivers/pcmcia/Makefile index f1a7ca04d89e..27e94b30cf96 100644 --- a/drivers/pcmcia/Makefile +++ b/drivers/pcmcia/Makefile @@ -12,7 +12,6 @@ obj-$(CONFIG_PCMCIA) += pcmcia.o pcmcia_rsrc-y += rsrc_mgr.o pcmcia_rsrc-$(CONFIG_PCCARD_NONSTATIC) += rsrc_nonstatic.o pcmcia_rsrc-$(CONFIG_PCCARD_IODYN) += rsrc_iodyn.o -pcmcia_rsrc-$(CONFIG_PCCARD_PCI) += rsrc_pci.o obj-$(CONFIG_PCCARD) += pcmcia_rsrc.o diff --git a/drivers/pcmcia/rsrc_pci.c b/drivers/pcmcia/rsrc_pci.c deleted file mode 100644 index 8934d3c01f80..000000000000 --- a/drivers/pcmcia/rsrc_pci.c +++ /dev/null @@ -1,172 +0,0 @@ -#include -#include -#include - -#include -#include -#include "cs_internal.h" - - -struct pcmcia_align_data { - unsigned long mask; - unsigned long offset; -}; - -static resource_size_t pcmcia_align(void *align_data, - const struct resource *res, - resource_size_t size, resource_size_t align) -{ - struct pcmcia_align_data *data = align_data; - resource_size_t start; - - start = (res->start & ~data->mask) + data->offset; - if (start < res->start) - start += data->mask + 1; - return start; -} - -static struct resource *find_io_region(struct pcmcia_socket *s, - unsigned long base, int num, - unsigned long align) -{ - struct resource *res = pcmcia_make_resource(0, num, IORESOURCE_IO, - dev_name(&s->dev)); - struct pcmcia_align_data data; - int ret; - - data.mask = align - 1; - data.offset = base & data.mask; - - ret = pci_bus_alloc_resource(s->cb_dev->bus, res, num, 1, - base, 0, pcmcia_align, &data); - if (ret != 0) { - kfree(res); - res = NULL; - } - return res; -} - -static int res_pci_find_io(struct pcmcia_socket *s, unsigned int attr, - unsigned int *base, unsigned int num, - unsigned int align, struct resource **parent) -{ - int i, ret = 0; - - /* Check for an already-allocated window that must conflict with - * what was asked for. It is a hack because it does not catch all - * potential conflicts, just the most obvious ones. - */ - for (i = 0; i < MAX_IO_WIN; i++) { - if (!s->io[i].res) - continue; - - if (!*base) - continue; - - if ((s->io[i].res->start & (align-1)) == *base) - return -EBUSY; - } - - for (i = 0; i < MAX_IO_WIN; i++) { - struct resource *res = s->io[i].res; - unsigned int try; - - if (res && (res->flags & IORESOURCE_BITS) != - (attr & IORESOURCE_BITS)) - continue; - - if (!res) { - if (align == 0) - align = 0x10000; - - res = s->io[i].res = find_io_region(s, *base, num, - align); - if (!res) - return -EINVAL; - - *base = res->start; - s->io[i].res->flags = - ((res->flags & ~IORESOURCE_BITS) | - (attr & IORESOURCE_BITS)); - s->io[i].InUse = num; - *parent = res; - return 0; - } - - /* Try to extend top of window */ - try = res->end + 1; - if ((*base == 0) || (*base == try)) { - ret = adjust_resource(s->io[i].res, res->start, - resource_size(res) + num); - if (ret) - continue; - *base = try; - s->io[i].InUse += num; - *parent = res; - return 0; - } - - /* Try to extend bottom of window */ - try = res->start - num; - if ((*base == 0) || (*base == try)) { - ret = adjust_resource(s->io[i].res, - res->start - num, - resource_size(res) + num); - if (ret) - continue; - *base = try; - s->io[i].InUse += num; - *parent = res; - return 0; - } - } - return -EINVAL; -} - -static struct resource *res_pci_find_mem(u_long base, u_long num, - u_long align, int low, struct pcmcia_socket *s) -{ - struct resource *res = pcmcia_make_resource(0, num, IORESOURCE_MEM, - dev_name(&s->dev)); - struct pcmcia_align_data data; - unsigned long min; - int ret; - - if (align < 0x20000) - align = 0x20000; - data.mask = align - 1; - data.offset = base & data.mask; - - min = 0; - if (!low) - min = 0x100000UL; - - ret = pci_bus_alloc_resource(s->cb_dev->bus, - res, num, 1, min, 0, - pcmcia_align, &data); - - if (ret != 0) { - kfree(res); - res = NULL; - } - return res; -} - - -static int res_pci_init(struct pcmcia_socket *s) -{ - if (!s->cb_dev || (!s->features & SS_CAP_PAGE_REGS)) { - dev_err(&s->dev, "not supported by res_pci\n"); - return -EOPNOTSUPP; - } - return 0; -} - -struct pccard_resource_ops pccard_nonstatic_ops = { - .validate_mem = NULL, - .find_io = res_pci_find_io, - .find_mem = res_pci_find_mem, - .init = res_pci_init, - .exit = NULL, -}; -EXPORT_SYMBOL(pccard_nonstatic_ops); -- cgit v1.2.1 From a987370f8e7a1677ae385042644326d9cd145a20 Mon Sep 17 00:00:00 2001 From: Marc Zyngier Date: Tue, 10 Mar 2015 19:06:59 +0000 Subject: arm64: KVM: Fix stage-2 PGD allocation to have per-page refcounting We're using __get_free_pages with to allocate the guest's stage-2 PGD. The standard behaviour of this function is to return a set of pages where only the head page has a valid refcount. This behaviour gets us into trouble when we're trying to increment the refcount on a non-head page: page:ffff7c00cfb693c0 count:0 mapcount:0 mapping: (null) index:0x0 flags: 0x4000000000000000() page dumped because: VM_BUG_ON_PAGE((*({ __attribute__((unused)) typeof((&page->_count)->counter) __var = ( typeof((&page->_count)->counter)) 0; (volatile typeof((&page->_count)->counter) *)&((&page->_count)->counter); })) <= 0) BUG: failure at include/linux/mm.h:548/get_page()! Kernel panic - not syncing: BUG! CPU: 1 PID: 1695 Comm: kvm-vcpu-0 Not tainted 4.0.0-rc1+ #3825 Hardware name: APM X-Gene Mustang board (DT) Call trace: [] dump_backtrace+0x0/0x13c [] show_stack+0x10/0x1c [] dump_stack+0x74/0x94 [] panic+0x100/0x240 [] stage2_get_pmd+0x17c/0x2bc [] kvm_handle_guest_abort+0x4b4/0x6b0 [] handle_exit+0x58/0x180 [] kvm_arch_vcpu_ioctl_run+0x114/0x45c [] kvm_vcpu_ioctl+0x2e0/0x754 [] do_vfs_ioctl+0x424/0x5c8 [] SyS_ioctl+0x40/0x78 CPU0: stopping A possible approach for this is to split the compound page using split_page() at allocation time, and change the teardown path to free one page at a time. It turns out that alloc_pages_exact() and free_pages_exact() does exactly that. While we're at it, the PGD allocation code is reworked to reduce duplication. This has been tested on an X-Gene platform with a 4kB/48bit-VA host kernel, and kvmtool hacked to place memory in the second page of the hardware PGD (PUD for the host kernel). Also regression-tested on a Cubietruck (Cortex-A7). [ Reworked to use alloc_pages_exact() and free_pages_exact() and to return pointers directly instead of by reference as arguments - Christoffer ] Reported-by: Mark Rutland Signed-off-by: Marc Zyngier Signed-off-by: Christoffer Dall --- arch/arm/include/asm/kvm_mmu.h | 10 +++--- arch/arm/kvm/mmu.c | 67 +++++++++++++++++++++++++++++----------- arch/arm64/include/asm/kvm_mmu.h | 46 +++------------------------ 3 files changed, 57 insertions(+), 66 deletions(-) diff --git a/arch/arm/include/asm/kvm_mmu.h b/arch/arm/include/asm/kvm_mmu.h index bf0fe99e8ca9..c57c41dc7e87 100644 --- a/arch/arm/include/asm/kvm_mmu.h +++ b/arch/arm/include/asm/kvm_mmu.h @@ -162,16 +162,14 @@ static inline bool kvm_page_empty(void *ptr) #define KVM_PREALLOC_LEVEL 0 -static inline int kvm_prealloc_hwpgd(struct kvm *kvm, pgd_t *pgd) +static inline void *kvm_get_hwpgd(struct kvm *kvm) { - return 0; + return kvm->arch.pgd; } -static inline void kvm_free_hwpgd(struct kvm *kvm) { } - -static inline void *kvm_get_hwpgd(struct kvm *kvm) +static inline unsigned int kvm_get_hwpgd_size(void) { - return kvm->arch.pgd; + return PTRS_PER_S2_PGD * sizeof(pgd_t); } struct kvm; diff --git a/arch/arm/kvm/mmu.c b/arch/arm/kvm/mmu.c index 3e6859bc3e11..a48a73c6b866 100644 --- a/arch/arm/kvm/mmu.c +++ b/arch/arm/kvm/mmu.c @@ -632,6 +632,20 @@ int create_hyp_io_mappings(void *from, void *to, phys_addr_t phys_addr) __phys_to_pfn(phys_addr), PAGE_HYP_DEVICE); } +/* Free the HW pgd, one page at a time */ +static void kvm_free_hwpgd(void *hwpgd) +{ + free_pages_exact(hwpgd, kvm_get_hwpgd_size()); +} + +/* Allocate the HW PGD, making sure that each page gets its own refcount */ +static void *kvm_alloc_hwpgd(void) +{ + unsigned int size = kvm_get_hwpgd_size(); + + return alloc_pages_exact(size, GFP_KERNEL | __GFP_ZERO); +} + /** * kvm_alloc_stage2_pgd - allocate level-1 table for stage-2 translation. * @kvm: The KVM struct pointer for the VM. @@ -645,15 +659,31 @@ int create_hyp_io_mappings(void *from, void *to, phys_addr_t phys_addr) */ int kvm_alloc_stage2_pgd(struct kvm *kvm) { - int ret; pgd_t *pgd; + void *hwpgd; if (kvm->arch.pgd != NULL) { kvm_err("kvm_arch already initialized?\n"); return -EINVAL; } + hwpgd = kvm_alloc_hwpgd(); + if (!hwpgd) + return -ENOMEM; + + /* When the kernel uses more levels of page tables than the + * guest, we allocate a fake PGD and pre-populate it to point + * to the next-level page table, which will be the real + * initial page table pointed to by the VTTBR. + * + * When KVM_PREALLOC_LEVEL==2, we allocate a single page for + * the PMD and the kernel will use folded pud. + * When KVM_PREALLOC_LEVEL==1, we allocate 2 consecutive PUD + * pages. + */ if (KVM_PREALLOC_LEVEL > 0) { + int i; + /* * Allocate fake pgd for the page table manipulation macros to * work. This is not used by the hardware and we have no @@ -661,30 +691,32 @@ int kvm_alloc_stage2_pgd(struct kvm *kvm) */ pgd = (pgd_t *)kmalloc(PTRS_PER_S2_PGD * sizeof(pgd_t), GFP_KERNEL | __GFP_ZERO); + + if (!pgd) { + kvm_free_hwpgd(hwpgd); + return -ENOMEM; + } + + /* Plug the HW PGD into the fake one. */ + for (i = 0; i < PTRS_PER_S2_PGD; i++) { + if (KVM_PREALLOC_LEVEL == 1) + pgd_populate(NULL, pgd + i, + (pud_t *)hwpgd + i * PTRS_PER_PUD); + else if (KVM_PREALLOC_LEVEL == 2) + pud_populate(NULL, pud_offset(pgd, 0) + i, + (pmd_t *)hwpgd + i * PTRS_PER_PMD); + } } else { /* * Allocate actual first-level Stage-2 page table used by the * hardware for Stage-2 page table walks. */ - pgd = (pgd_t *)__get_free_pages(GFP_KERNEL | __GFP_ZERO, S2_PGD_ORDER); + pgd = (pgd_t *)hwpgd; } - if (!pgd) - return -ENOMEM; - - ret = kvm_prealloc_hwpgd(kvm, pgd); - if (ret) - goto out_err; - kvm_clean_pgd(pgd); kvm->arch.pgd = pgd; return 0; -out_err: - if (KVM_PREALLOC_LEVEL > 0) - kfree(pgd); - else - free_pages((unsigned long)pgd, S2_PGD_ORDER); - return ret; } /** @@ -785,11 +817,10 @@ void kvm_free_stage2_pgd(struct kvm *kvm) return; unmap_stage2_range(kvm, 0, KVM_PHYS_SIZE); - kvm_free_hwpgd(kvm); + kvm_free_hwpgd(kvm_get_hwpgd(kvm)); if (KVM_PREALLOC_LEVEL > 0) kfree(kvm->arch.pgd); - else - free_pages((unsigned long)kvm->arch.pgd, S2_PGD_ORDER); + kvm->arch.pgd = NULL; } diff --git a/arch/arm64/include/asm/kvm_mmu.h b/arch/arm64/include/asm/kvm_mmu.h index 6458b5373142..a099cd9cdef8 100644 --- a/arch/arm64/include/asm/kvm_mmu.h +++ b/arch/arm64/include/asm/kvm_mmu.h @@ -171,43 +171,6 @@ static inline bool kvm_s2pmd_readonly(pmd_t *pmd) #define KVM_PREALLOC_LEVEL (0) #endif -/** - * kvm_prealloc_hwpgd - allocate inital table for VTTBR - * @kvm: The KVM struct pointer for the VM. - * @pgd: The kernel pseudo pgd - * - * When the kernel uses more levels of page tables than the guest, we allocate - * a fake PGD and pre-populate it to point to the next-level page table, which - * will be the real initial page table pointed to by the VTTBR. - * - * When KVM_PREALLOC_LEVEL==2, we allocate a single page for the PMD and - * the kernel will use folded pud. When KVM_PREALLOC_LEVEL==1, we - * allocate 2 consecutive PUD pages. - */ -static inline int kvm_prealloc_hwpgd(struct kvm *kvm, pgd_t *pgd) -{ - unsigned int i; - unsigned long hwpgd; - - if (KVM_PREALLOC_LEVEL == 0) - return 0; - - hwpgd = __get_free_pages(GFP_KERNEL | __GFP_ZERO, PTRS_PER_S2_PGD_SHIFT); - if (!hwpgd) - return -ENOMEM; - - for (i = 0; i < PTRS_PER_S2_PGD; i++) { - if (KVM_PREALLOC_LEVEL == 1) - pgd_populate(NULL, pgd + i, - (pud_t *)hwpgd + i * PTRS_PER_PUD); - else if (KVM_PREALLOC_LEVEL == 2) - pud_populate(NULL, pud_offset(pgd, 0) + i, - (pmd_t *)hwpgd + i * PTRS_PER_PMD); - } - - return 0; -} - static inline void *kvm_get_hwpgd(struct kvm *kvm) { pgd_t *pgd = kvm->arch.pgd; @@ -224,12 +187,11 @@ static inline void *kvm_get_hwpgd(struct kvm *kvm) return pmd_offset(pud, 0); } -static inline void kvm_free_hwpgd(struct kvm *kvm) +static inline unsigned int kvm_get_hwpgd_size(void) { - if (KVM_PREALLOC_LEVEL > 0) { - unsigned long hwpgd = (unsigned long)kvm_get_hwpgd(kvm); - free_pages(hwpgd, PTRS_PER_S2_PGD_SHIFT); - } + if (KVM_PREALLOC_LEVEL > 0) + return PTRS_PER_S2_PGD * PAGE_SIZE; + return PTRS_PER_S2_PGD * sizeof(pgd_t); } static inline bool kvm_page_empty(void *ptr) -- cgit v1.2.1 From 04b8dc85bf4a64517e3cf20e409eeaa503b15cc1 Mon Sep 17 00:00:00 2001 From: Marc Zyngier Date: Tue, 10 Mar 2015 19:07:00 +0000 Subject: arm64: KVM: Do not use pgd_index to index stage-2 pgd The kernel's pgd_index macro is designed to index a normal, page sized array. KVM is a bit diffferent, as we can use concatenated pages to have a bigger address space (for example 40bit IPA with 4kB pages gives us an 8kB PGD. In the above case, the use of pgd_index will always return an index inside the first 4kB, which makes a guest that has memory above 0x8000000000 rather unhappy, as it spins forever in a page fault, whist the host happilly corrupts the lower pgd. The obvious fix is to get our own kvm_pgd_index that does the right thing(tm). Tested on X-Gene with a hacked kvmtool that put memory at a stupidly high address. Reviewed-by: Christoffer Dall Signed-off-by: Marc Zyngier Signed-off-by: Christoffer Dall --- arch/arm/include/asm/kvm_mmu.h | 3 ++- arch/arm/kvm/mmu.c | 8 ++++---- arch/arm64/include/asm/kvm_mmu.h | 2 ++ 3 files changed, 8 insertions(+), 5 deletions(-) diff --git a/arch/arm/include/asm/kvm_mmu.h b/arch/arm/include/asm/kvm_mmu.h index c57c41dc7e87..4cf48c3aca13 100644 --- a/arch/arm/include/asm/kvm_mmu.h +++ b/arch/arm/include/asm/kvm_mmu.h @@ -149,13 +149,14 @@ static inline bool kvm_s2pmd_readonly(pmd_t *pmd) (__boundary - 1 < (end) - 1)? __boundary: (end); \ }) +#define kvm_pgd_index(addr) pgd_index(addr) + static inline bool kvm_page_empty(void *ptr) { struct page *ptr_page = virt_to_page(ptr); return page_count(ptr_page) == 1; } - #define kvm_pte_table_empty(kvm, ptep) kvm_page_empty(ptep) #define kvm_pmd_table_empty(kvm, pmdp) kvm_page_empty(pmdp) #define kvm_pud_table_empty(kvm, pudp) (0) diff --git a/arch/arm/kvm/mmu.c b/arch/arm/kvm/mmu.c index a48a73c6b866..5656d79c5a44 100644 --- a/arch/arm/kvm/mmu.c +++ b/arch/arm/kvm/mmu.c @@ -290,7 +290,7 @@ static void unmap_range(struct kvm *kvm, pgd_t *pgdp, phys_addr_t addr = start, end = start + size; phys_addr_t next; - pgd = pgdp + pgd_index(addr); + pgd = pgdp + kvm_pgd_index(addr); do { next = kvm_pgd_addr_end(addr, end); if (!pgd_none(*pgd)) @@ -355,7 +355,7 @@ static void stage2_flush_memslot(struct kvm *kvm, phys_addr_t next; pgd_t *pgd; - pgd = kvm->arch.pgd + pgd_index(addr); + pgd = kvm->arch.pgd + kvm_pgd_index(addr); do { next = kvm_pgd_addr_end(addr, end); stage2_flush_puds(kvm, pgd, addr, next); @@ -830,7 +830,7 @@ static pud_t *stage2_get_pud(struct kvm *kvm, struct kvm_mmu_memory_cache *cache pgd_t *pgd; pud_t *pud; - pgd = kvm->arch.pgd + pgd_index(addr); + pgd = kvm->arch.pgd + kvm_pgd_index(addr); if (WARN_ON(pgd_none(*pgd))) { if (!cache) return NULL; @@ -1120,7 +1120,7 @@ static void stage2_wp_range(struct kvm *kvm, phys_addr_t addr, phys_addr_t end) pgd_t *pgd; phys_addr_t next; - pgd = kvm->arch.pgd + pgd_index(addr); + pgd = kvm->arch.pgd + kvm_pgd_index(addr); do { /* * Release kvm_mmu_lock periodically if the memory region is diff --git a/arch/arm64/include/asm/kvm_mmu.h b/arch/arm64/include/asm/kvm_mmu.h index a099cd9cdef8..bbfb600fa822 100644 --- a/arch/arm64/include/asm/kvm_mmu.h +++ b/arch/arm64/include/asm/kvm_mmu.h @@ -158,6 +158,8 @@ static inline bool kvm_s2pmd_readonly(pmd_t *pmd) #define PTRS_PER_S2_PGD (1 << PTRS_PER_S2_PGD_SHIFT) #define S2_PGD_ORDER get_order(PTRS_PER_S2_PGD * sizeof(pgd_t)) +#define kvm_pgd_index(addr) (((addr) >> PGDIR_SHIFT) & (PTRS_PER_S2_PGD - 1)) + /* * If we are concatenating first level stage-2 page tables, we would have less * than or equal to 16 pointers in the fake PGD, because that's what the -- cgit v1.2.1 From 84ed7412b5eee1011579b3db7454b9cb6d26fa65 Mon Sep 17 00:00:00 2001 From: Marc Zyngier Date: Tue, 10 Mar 2015 19:07:01 +0000 Subject: arm64: KVM: Fix outdated comment about VTCR_EL2.PS Commit 87366d8cf7b3 ("arm64: Add boot time configuration of Intermediate Physical Address size") removed the hardcoded setting of VTCR_EL2.PS to use ID_AA64MMFR0_EL1.PARange instead, but didn't remove the (now rather misleading) comment. Fix the comments to match reality (at least for the next few minutes). Acked-by: Christoffer Dall Signed-off-by: Marc Zyngier Signed-off-by: Christoffer Dall --- arch/arm64/include/asm/kvm_arm.h | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/arch/arm64/include/asm/kvm_arm.h b/arch/arm64/include/asm/kvm_arm.h index 94674eb7e7bb..54bb4ba97441 100644 --- a/arch/arm64/include/asm/kvm_arm.h +++ b/arch/arm64/include/asm/kvm_arm.h @@ -129,6 +129,9 @@ * 40 bits wide (T0SZ = 24). Systems with a PARange smaller than 40 bits are * not known to exist and will break with this configuration. * + * VTCR_EL2.PS is extracted from ID_AA64MMFR0_EL1.PARange at boot time + * (see hyp-init.S). + * * Note that when using 4K pages, we concatenate two first level page tables * together. * @@ -138,7 +141,6 @@ #ifdef CONFIG_ARM64_64K_PAGES /* * Stage2 translation configuration: - * 40bits output (PS = 2) * 40bits input (T0SZ = 24) * 64kB pages (TG0 = 1) * 2 level page tables (SL = 1) @@ -150,7 +152,6 @@ #else /* * Stage2 translation configuration: - * 40bits output (PS = 2) * 40bits input (T0SZ = 24) * 4kB pages (TG0 = 0) * 3 level page tables (SL = 1) -- cgit v1.2.1 From 2c247804796bbcaa90087f2196f68fdc20a5fe04 Mon Sep 17 00:00:00 2001 From: Felipe Balbi Date: Wed, 11 Mar 2015 10:00:05 -0500 Subject: Revert "usb: gadget: zero: Add support for interrupt EP" This reverts commit ef11982dd7a657512c362242508bb4021e0d67b6. That commit creates a problem for some UDCs (at least musb) where it allocates an endpoints with a 64-byte FIFO, but later tries to use that same FIFO for 1024-byte packets. Before implementing this, composite framework needs to be modified so we only allocate endpoints after we know negotiated speed, however that needs quite a bit of extra work. Cc: # v3.18+ Signed-off-by: Felipe Balbi --- drivers/usb/gadget/function/f_loopback.c | 3 +- drivers/usb/gadget/function/f_sourcesink.c | 511 ++--------------------------- drivers/usb/gadget/function/g_zero.h | 13 +- drivers/usb/gadget/legacy/zero.c | 21 -- 4 files changed, 22 insertions(+), 526 deletions(-) diff --git a/drivers/usb/gadget/function/f_loopback.c b/drivers/usb/gadget/function/f_loopback.c index 298b46112b1a..39f49f1ad22f 100644 --- a/drivers/usb/gadget/function/f_loopback.c +++ b/drivers/usb/gadget/function/f_loopback.c @@ -289,8 +289,7 @@ static void disable_loopback(struct f_loopback *loop) struct usb_composite_dev *cdev; cdev = loop->function.config->cdev; - disable_endpoints(cdev, loop->in_ep, loop->out_ep, NULL, NULL, NULL, - NULL); + disable_endpoints(cdev, loop->in_ep, loop->out_ep, NULL, NULL); VDBG(cdev, "%s disabled\n", loop->function.name); } diff --git a/drivers/usb/gadget/function/f_sourcesink.c b/drivers/usb/gadget/function/f_sourcesink.c index e3dae47baef3..3a5ae9900b1e 100644 --- a/drivers/usb/gadget/function/f_sourcesink.c +++ b/drivers/usb/gadget/function/f_sourcesink.c @@ -23,15 +23,6 @@ #include "gadget_chips.h" #include "u_f.h" -#define USB_MS_TO_SS_INTERVAL(x) USB_MS_TO_HS_INTERVAL(x) - -enum eptype { - EP_CONTROL = 0, - EP_BULK, - EP_ISOC, - EP_INTERRUPT, -}; - /* * SOURCE/SINK FUNCTION ... a primary testing vehicle for USB peripheral * controller drivers. @@ -64,8 +55,6 @@ struct f_sourcesink { struct usb_ep *out_ep; struct usb_ep *iso_in_ep; struct usb_ep *iso_out_ep; - struct usb_ep *int_in_ep; - struct usb_ep *int_out_ep; int cur_alt; }; @@ -79,10 +68,6 @@ static unsigned isoc_interval; static unsigned isoc_maxpacket; static unsigned isoc_mult; static unsigned isoc_maxburst; -static unsigned int_interval; /* In ms */ -static unsigned int_maxpacket; -static unsigned int_mult; -static unsigned int_maxburst; static unsigned buflen; /*-------------------------------------------------------------------------*/ @@ -107,16 +92,6 @@ static struct usb_interface_descriptor source_sink_intf_alt1 = { /* .iInterface = DYNAMIC */ }; -static struct usb_interface_descriptor source_sink_intf_alt2 = { - .bLength = USB_DT_INTERFACE_SIZE, - .bDescriptorType = USB_DT_INTERFACE, - - .bAlternateSetting = 2, - .bNumEndpoints = 2, - .bInterfaceClass = USB_CLASS_VENDOR_SPEC, - /* .iInterface = DYNAMIC */ -}; - /* full speed support: */ static struct usb_endpoint_descriptor fs_source_desc = { @@ -155,26 +130,6 @@ static struct usb_endpoint_descriptor fs_iso_sink_desc = { .bInterval = 4, }; -static struct usb_endpoint_descriptor fs_int_source_desc = { - .bLength = USB_DT_ENDPOINT_SIZE, - .bDescriptorType = USB_DT_ENDPOINT, - - .bEndpointAddress = USB_DIR_IN, - .bmAttributes = USB_ENDPOINT_XFER_INT, - .wMaxPacketSize = cpu_to_le16(64), - .bInterval = GZERO_INT_INTERVAL, -}; - -static struct usb_endpoint_descriptor fs_int_sink_desc = { - .bLength = USB_DT_ENDPOINT_SIZE, - .bDescriptorType = USB_DT_ENDPOINT, - - .bEndpointAddress = USB_DIR_OUT, - .bmAttributes = USB_ENDPOINT_XFER_INT, - .wMaxPacketSize = cpu_to_le16(64), - .bInterval = GZERO_INT_INTERVAL, -}; - static struct usb_descriptor_header *fs_source_sink_descs[] = { (struct usb_descriptor_header *) &source_sink_intf_alt0, (struct usb_descriptor_header *) &fs_sink_desc, @@ -185,10 +140,6 @@ static struct usb_descriptor_header *fs_source_sink_descs[] = { (struct usb_descriptor_header *) &fs_source_desc, (struct usb_descriptor_header *) &fs_iso_sink_desc, (struct usb_descriptor_header *) &fs_iso_source_desc, - (struct usb_descriptor_header *) &source_sink_intf_alt2, -#define FS_ALT_IFC_2_OFFSET 8 - (struct usb_descriptor_header *) &fs_int_sink_desc, - (struct usb_descriptor_header *) &fs_int_source_desc, NULL, }; @@ -228,24 +179,6 @@ static struct usb_endpoint_descriptor hs_iso_sink_desc = { .bInterval = 4, }; -static struct usb_endpoint_descriptor hs_int_source_desc = { - .bLength = USB_DT_ENDPOINT_SIZE, - .bDescriptorType = USB_DT_ENDPOINT, - - .bmAttributes = USB_ENDPOINT_XFER_INT, - .wMaxPacketSize = cpu_to_le16(1024), - .bInterval = USB_MS_TO_HS_INTERVAL(GZERO_INT_INTERVAL), -}; - -static struct usb_endpoint_descriptor hs_int_sink_desc = { - .bLength = USB_DT_ENDPOINT_SIZE, - .bDescriptorType = USB_DT_ENDPOINT, - - .bmAttributes = USB_ENDPOINT_XFER_INT, - .wMaxPacketSize = cpu_to_le16(1024), - .bInterval = USB_MS_TO_HS_INTERVAL(GZERO_INT_INTERVAL), -}; - static struct usb_descriptor_header *hs_source_sink_descs[] = { (struct usb_descriptor_header *) &source_sink_intf_alt0, (struct usb_descriptor_header *) &hs_source_desc, @@ -256,10 +189,6 @@ static struct usb_descriptor_header *hs_source_sink_descs[] = { (struct usb_descriptor_header *) &hs_sink_desc, (struct usb_descriptor_header *) &hs_iso_source_desc, (struct usb_descriptor_header *) &hs_iso_sink_desc, - (struct usb_descriptor_header *) &source_sink_intf_alt2, -#define HS_ALT_IFC_2_OFFSET 8 - (struct usb_descriptor_header *) &hs_int_source_desc, - (struct usb_descriptor_header *) &hs_int_sink_desc, NULL, }; @@ -335,42 +264,6 @@ static struct usb_ss_ep_comp_descriptor ss_iso_sink_comp_desc = { .wBytesPerInterval = cpu_to_le16(1024), }; -static struct usb_endpoint_descriptor ss_int_source_desc = { - .bLength = USB_DT_ENDPOINT_SIZE, - .bDescriptorType = USB_DT_ENDPOINT, - - .bmAttributes = USB_ENDPOINT_XFER_INT, - .wMaxPacketSize = cpu_to_le16(1024), - .bInterval = USB_MS_TO_SS_INTERVAL(GZERO_INT_INTERVAL), -}; - -static struct usb_ss_ep_comp_descriptor ss_int_source_comp_desc = { - .bLength = USB_DT_SS_EP_COMP_SIZE, - .bDescriptorType = USB_DT_SS_ENDPOINT_COMP, - - .bMaxBurst = 0, - .bmAttributes = 0, - .wBytesPerInterval = cpu_to_le16(1024), -}; - -static struct usb_endpoint_descriptor ss_int_sink_desc = { - .bLength = USB_DT_ENDPOINT_SIZE, - .bDescriptorType = USB_DT_ENDPOINT, - - .bmAttributes = USB_ENDPOINT_XFER_INT, - .wMaxPacketSize = cpu_to_le16(1024), - .bInterval = USB_MS_TO_SS_INTERVAL(GZERO_INT_INTERVAL), -}; - -static struct usb_ss_ep_comp_descriptor ss_int_sink_comp_desc = { - .bLength = USB_DT_SS_EP_COMP_SIZE, - .bDescriptorType = USB_DT_SS_ENDPOINT_COMP, - - .bMaxBurst = 0, - .bmAttributes = 0, - .wBytesPerInterval = cpu_to_le16(1024), -}; - static struct usb_descriptor_header *ss_source_sink_descs[] = { (struct usb_descriptor_header *) &source_sink_intf_alt0, (struct usb_descriptor_header *) &ss_source_desc, @@ -387,12 +280,6 @@ static struct usb_descriptor_header *ss_source_sink_descs[] = { (struct usb_descriptor_header *) &ss_iso_source_comp_desc, (struct usb_descriptor_header *) &ss_iso_sink_desc, (struct usb_descriptor_header *) &ss_iso_sink_comp_desc, - (struct usb_descriptor_header *) &source_sink_intf_alt2, -#define SS_ALT_IFC_2_OFFSET 14 - (struct usb_descriptor_header *) &ss_int_source_desc, - (struct usb_descriptor_header *) &ss_int_source_comp_desc, - (struct usb_descriptor_header *) &ss_int_sink_desc, - (struct usb_descriptor_header *) &ss_int_sink_comp_desc, NULL, }; @@ -414,21 +301,6 @@ static struct usb_gadget_strings *sourcesink_strings[] = { }; /*-------------------------------------------------------------------------*/ -static const char *get_ep_string(enum eptype ep_type) -{ - switch (ep_type) { - case EP_ISOC: - return "ISOC-"; - case EP_INTERRUPT: - return "INTERRUPT-"; - case EP_CONTROL: - return "CTRL-"; - case EP_BULK: - return "BULK-"; - default: - return "UNKNOWN-"; - } -} static inline struct usb_request *ss_alloc_ep_req(struct usb_ep *ep, int len) { @@ -456,8 +328,7 @@ static void disable_ep(struct usb_composite_dev *cdev, struct usb_ep *ep) void disable_endpoints(struct usb_composite_dev *cdev, struct usb_ep *in, struct usb_ep *out, - struct usb_ep *iso_in, struct usb_ep *iso_out, - struct usb_ep *int_in, struct usb_ep *int_out) + struct usb_ep *iso_in, struct usb_ep *iso_out) { disable_ep(cdev, in); disable_ep(cdev, out); @@ -465,10 +336,6 @@ void disable_endpoints(struct usb_composite_dev *cdev, disable_ep(cdev, iso_in); if (iso_out) disable_ep(cdev, iso_out); - if (int_in) - disable_ep(cdev, int_in); - if (int_out) - disable_ep(cdev, int_out); } static int @@ -485,7 +352,6 @@ sourcesink_bind(struct usb_configuration *c, struct usb_function *f) return id; source_sink_intf_alt0.bInterfaceNumber = id; source_sink_intf_alt1.bInterfaceNumber = id; - source_sink_intf_alt2.bInterfaceNumber = id; /* allocate bulk endpoints */ ss->in_ep = usb_ep_autoconfig(cdev->gadget, &fs_source_desc); @@ -546,55 +412,14 @@ no_iso: if (isoc_maxpacket > 1024) isoc_maxpacket = 1024; - /* sanity check the interrupt module parameters */ - if (int_interval < 1) - int_interval = 1; - if (int_interval > 4096) - int_interval = 4096; - if (int_mult > 2) - int_mult = 2; - if (int_maxburst > 15) - int_maxburst = 15; - - /* fill in the FS interrupt descriptors from the module parameters */ - fs_int_source_desc.wMaxPacketSize = int_maxpacket > 64 ? - 64 : int_maxpacket; - fs_int_source_desc.bInterval = int_interval > 255 ? - 255 : int_interval; - fs_int_sink_desc.wMaxPacketSize = int_maxpacket > 64 ? - 64 : int_maxpacket; - fs_int_sink_desc.bInterval = int_interval > 255 ? - 255 : int_interval; - - /* allocate int endpoints */ - ss->int_in_ep = usb_ep_autoconfig(cdev->gadget, &fs_int_source_desc); - if (!ss->int_in_ep) - goto no_int; - ss->int_in_ep->driver_data = cdev; /* claim */ - - ss->int_out_ep = usb_ep_autoconfig(cdev->gadget, &fs_int_sink_desc); - if (ss->int_out_ep) { - ss->int_out_ep->driver_data = cdev; /* claim */ - } else { - ss->int_in_ep->driver_data = NULL; - ss->int_in_ep = NULL; -no_int: - fs_source_sink_descs[FS_ALT_IFC_2_OFFSET] = NULL; - hs_source_sink_descs[HS_ALT_IFC_2_OFFSET] = NULL; - ss_source_sink_descs[SS_ALT_IFC_2_OFFSET] = NULL; - } - - if (int_maxpacket > 1024) - int_maxpacket = 1024; - /* support high speed hardware */ hs_source_desc.bEndpointAddress = fs_source_desc.bEndpointAddress; hs_sink_desc.bEndpointAddress = fs_sink_desc.bEndpointAddress; /* - * Fill in the HS isoc and interrupt descriptors from the module - * parameters. We assume that the user knows what they are doing and - * won't give parameters that their UDC doesn't support. + * Fill in the HS isoc descriptors from the module parameters. + * We assume that the user knows what they are doing and won't + * give parameters that their UDC doesn't support. */ hs_iso_source_desc.wMaxPacketSize = isoc_maxpacket; hs_iso_source_desc.wMaxPacketSize |= isoc_mult << 11; @@ -607,17 +432,6 @@ no_int: hs_iso_sink_desc.bInterval = isoc_interval; hs_iso_sink_desc.bEndpointAddress = fs_iso_sink_desc.bEndpointAddress; - hs_int_source_desc.wMaxPacketSize = int_maxpacket; - hs_int_source_desc.wMaxPacketSize |= int_mult << 11; - hs_int_source_desc.bInterval = USB_MS_TO_HS_INTERVAL(int_interval); - hs_int_source_desc.bEndpointAddress = - fs_int_source_desc.bEndpointAddress; - - hs_int_sink_desc.wMaxPacketSize = int_maxpacket; - hs_int_sink_desc.wMaxPacketSize |= int_mult << 11; - hs_int_sink_desc.bInterval = USB_MS_TO_HS_INTERVAL(int_interval); - hs_int_sink_desc.bEndpointAddress = fs_int_sink_desc.bEndpointAddress; - /* support super speed hardware */ ss_source_desc.bEndpointAddress = fs_source_desc.bEndpointAddress; @@ -625,9 +439,9 @@ no_int: fs_sink_desc.bEndpointAddress; /* - * Fill in the SS isoc and interrupt descriptors from the module - * parameters. We assume that the user knows what they are doing and - * won't give parameters that their UDC doesn't support. + * Fill in the SS isoc descriptors from the module parameters. + * We assume that the user knows what they are doing and won't + * give parameters that their UDC doesn't support. */ ss_iso_source_desc.wMaxPacketSize = isoc_maxpacket; ss_iso_source_desc.bInterval = isoc_interval; @@ -646,37 +460,17 @@ no_int: isoc_maxpacket * (isoc_mult + 1) * (isoc_maxburst + 1); ss_iso_sink_desc.bEndpointAddress = fs_iso_sink_desc.bEndpointAddress; - ss_int_source_desc.wMaxPacketSize = int_maxpacket; - ss_int_source_desc.bInterval = USB_MS_TO_SS_INTERVAL(int_interval); - ss_int_source_comp_desc.bmAttributes = int_mult; - ss_int_source_comp_desc.bMaxBurst = int_maxburst; - ss_int_source_comp_desc.wBytesPerInterval = - int_maxpacket * (int_mult + 1) * (int_maxburst + 1); - ss_int_source_desc.bEndpointAddress = - fs_int_source_desc.bEndpointAddress; - - ss_int_sink_desc.wMaxPacketSize = int_maxpacket; - ss_int_sink_desc.bInterval = USB_MS_TO_SS_INTERVAL(int_interval); - ss_int_sink_comp_desc.bmAttributes = int_mult; - ss_int_sink_comp_desc.bMaxBurst = int_maxburst; - ss_int_sink_comp_desc.wBytesPerInterval = - int_maxpacket * (int_mult + 1) * (int_maxburst + 1); - ss_int_sink_desc.bEndpointAddress = fs_int_sink_desc.bEndpointAddress; - ret = usb_assign_descriptors(f, fs_source_sink_descs, hs_source_sink_descs, ss_source_sink_descs); if (ret) return ret; - DBG(cdev, "%s speed %s: IN/%s, OUT/%s, ISO-IN/%s, ISO-OUT/%s, " - "INT-IN/%s, INT-OUT/%s\n", + DBG(cdev, "%s speed %s: IN/%s, OUT/%s, ISO-IN/%s, ISO-OUT/%s\n", (gadget_is_superspeed(c->cdev->gadget) ? "super" : (gadget_is_dualspeed(c->cdev->gadget) ? "dual" : "full")), f->name, ss->in_ep->name, ss->out_ep->name, ss->iso_in_ep ? ss->iso_in_ep->name : "", - ss->iso_out_ep ? ss->iso_out_ep->name : "", - ss->int_in_ep ? ss->int_in_ep->name : "", - ss->int_out_ep ? ss->int_out_ep->name : ""); + ss->iso_out_ep ? ss->iso_out_ep->name : ""); return 0; } @@ -807,15 +601,14 @@ static void source_sink_complete(struct usb_ep *ep, struct usb_request *req) } static int source_sink_start_ep(struct f_sourcesink *ss, bool is_in, - enum eptype ep_type, int speed) + bool is_iso, int speed) { struct usb_ep *ep; struct usb_request *req; int i, size, status; for (i = 0; i < 8; i++) { - switch (ep_type) { - case EP_ISOC: + if (is_iso) { switch (speed) { case USB_SPEED_SUPER: size = isoc_maxpacket * (isoc_mult + 1) * @@ -831,28 +624,9 @@ static int source_sink_start_ep(struct f_sourcesink *ss, bool is_in, } ep = is_in ? ss->iso_in_ep : ss->iso_out_ep; req = ss_alloc_ep_req(ep, size); - break; - case EP_INTERRUPT: - switch (speed) { - case USB_SPEED_SUPER: - size = int_maxpacket * (int_mult + 1) * - (int_maxburst + 1); - break; - case USB_SPEED_HIGH: - size = int_maxpacket * (int_mult + 1); - break; - default: - size = int_maxpacket > 1023 ? - 1023 : int_maxpacket; - break; - } - ep = is_in ? ss->int_in_ep : ss->int_out_ep; - req = ss_alloc_ep_req(ep, size); - break; - default: + } else { ep = is_in ? ss->in_ep : ss->out_ep; req = ss_alloc_ep_req(ep, 0); - break; } if (!req) @@ -870,12 +644,12 @@ static int source_sink_start_ep(struct f_sourcesink *ss, bool is_in, cdev = ss->function.config->cdev; ERROR(cdev, "start %s%s %s --> %d\n", - get_ep_string(ep_type), is_in ? "IN" : "OUT", - ep->name, status); + is_iso ? "ISO-" : "", is_in ? "IN" : "OUT", + ep->name, status); free_ep_req(ep, req); } - if (!(ep_type == EP_ISOC)) + if (!is_iso) break; } @@ -888,7 +662,7 @@ static void disable_source_sink(struct f_sourcesink *ss) cdev = ss->function.config->cdev; disable_endpoints(cdev, ss->in_ep, ss->out_ep, ss->iso_in_ep, - ss->iso_out_ep, ss->int_in_ep, ss->int_out_ep); + ss->iso_out_ep); VDBG(cdev, "%s disabled\n", ss->function.name); } @@ -900,62 +674,6 @@ enable_source_sink(struct usb_composite_dev *cdev, struct f_sourcesink *ss, int speed = cdev->gadget->speed; struct usb_ep *ep; - if (alt == 2) { - /* Configure for periodic interrupt endpoint */ - ep = ss->int_in_ep; - if (ep) { - result = config_ep_by_speed(cdev->gadget, - &(ss->function), ep); - if (result) - return result; - - result = usb_ep_enable(ep); - if (result < 0) - return result; - - ep->driver_data = ss; - result = source_sink_start_ep(ss, true, EP_INTERRUPT, - speed); - if (result < 0) { -fail1: - ep = ss->int_in_ep; - if (ep) { - usb_ep_disable(ep); - ep->driver_data = NULL; - } - return result; - } - } - - /* - * one interrupt endpoint reads (sinks) anything OUT (from the - * host) - */ - ep = ss->int_out_ep; - if (ep) { - result = config_ep_by_speed(cdev->gadget, - &(ss->function), ep); - if (result) - goto fail1; - - result = usb_ep_enable(ep); - if (result < 0) - goto fail1; - - ep->driver_data = ss; - result = source_sink_start_ep(ss, false, EP_INTERRUPT, - speed); - if (result < 0) { - ep = ss->int_out_ep; - usb_ep_disable(ep); - ep->driver_data = NULL; - goto fail1; - } - } - - goto out; - } - /* one bulk endpoint writes (sources) zeroes IN (to the host) */ ep = ss->in_ep; result = config_ep_by_speed(cdev->gadget, &(ss->function), ep); @@ -966,7 +684,7 @@ fail1: return result; ep->driver_data = ss; - result = source_sink_start_ep(ss, true, EP_BULK, speed); + result = source_sink_start_ep(ss, true, false, speed); if (result < 0) { fail: ep = ss->in_ep; @@ -985,7 +703,7 @@ fail: goto fail; ep->driver_data = ss; - result = source_sink_start_ep(ss, false, EP_BULK, speed); + result = source_sink_start_ep(ss, false, false, speed); if (result < 0) { fail2: ep = ss->out_ep; @@ -1008,7 +726,7 @@ fail2: goto fail2; ep->driver_data = ss; - result = source_sink_start_ep(ss, true, EP_ISOC, speed); + result = source_sink_start_ep(ss, true, true, speed); if (result < 0) { fail3: ep = ss->iso_in_ep; @@ -1031,14 +749,13 @@ fail3: goto fail3; ep->driver_data = ss; - result = source_sink_start_ep(ss, false, EP_ISOC, speed); + result = source_sink_start_ep(ss, false, true, speed); if (result < 0) { usb_ep_disable(ep); ep->driver_data = NULL; goto fail3; } } - out: ss->cur_alt = alt; @@ -1054,8 +771,6 @@ static int sourcesink_set_alt(struct usb_function *f, if (ss->in_ep->driver_data) disable_source_sink(ss); - else if (alt == 2 && ss->int_in_ep->driver_data) - disable_source_sink(ss); return enable_source_sink(cdev, ss, alt); } @@ -1168,10 +883,6 @@ static struct usb_function *source_sink_alloc_func( isoc_maxpacket = ss_opts->isoc_maxpacket; isoc_mult = ss_opts->isoc_mult; isoc_maxburst = ss_opts->isoc_maxburst; - int_interval = ss_opts->int_interval; - int_maxpacket = ss_opts->int_maxpacket; - int_mult = ss_opts->int_mult; - int_maxburst = ss_opts->int_maxburst; buflen = ss_opts->bulk_buflen; ss->function.name = "source/sink"; @@ -1468,182 +1179,6 @@ static struct f_ss_opts_attribute f_ss_opts_bulk_buflen = f_ss_opts_bulk_buflen_show, f_ss_opts_bulk_buflen_store); -static ssize_t f_ss_opts_int_interval_show(struct f_ss_opts *opts, char *page) -{ - int result; - - mutex_lock(&opts->lock); - result = sprintf(page, "%u", opts->int_interval); - mutex_unlock(&opts->lock); - - return result; -} - -static ssize_t f_ss_opts_int_interval_store(struct f_ss_opts *opts, - const char *page, size_t len) -{ - int ret; - u32 num; - - mutex_lock(&opts->lock); - if (opts->refcnt) { - ret = -EBUSY; - goto end; - } - - ret = kstrtou32(page, 0, &num); - if (ret) - goto end; - - if (num > 4096) { - ret = -EINVAL; - goto end; - } - - opts->int_interval = num; - ret = len; -end: - mutex_unlock(&opts->lock); - return ret; -} - -static struct f_ss_opts_attribute f_ss_opts_int_interval = - __CONFIGFS_ATTR(int_interval, S_IRUGO | S_IWUSR, - f_ss_opts_int_interval_show, - f_ss_opts_int_interval_store); - -static ssize_t f_ss_opts_int_maxpacket_show(struct f_ss_opts *opts, char *page) -{ - int result; - - mutex_lock(&opts->lock); - result = sprintf(page, "%u", opts->int_maxpacket); - mutex_unlock(&opts->lock); - - return result; -} - -static ssize_t f_ss_opts_int_maxpacket_store(struct f_ss_opts *opts, - const char *page, size_t len) -{ - int ret; - u16 num; - - mutex_lock(&opts->lock); - if (opts->refcnt) { - ret = -EBUSY; - goto end; - } - - ret = kstrtou16(page, 0, &num); - if (ret) - goto end; - - if (num > 1024) { - ret = -EINVAL; - goto end; - } - - opts->int_maxpacket = num; - ret = len; -end: - mutex_unlock(&opts->lock); - return ret; -} - -static struct f_ss_opts_attribute f_ss_opts_int_maxpacket = - __CONFIGFS_ATTR(int_maxpacket, S_IRUGO | S_IWUSR, - f_ss_opts_int_maxpacket_show, - f_ss_opts_int_maxpacket_store); - -static ssize_t f_ss_opts_int_mult_show(struct f_ss_opts *opts, char *page) -{ - int result; - - mutex_lock(&opts->lock); - result = sprintf(page, "%u", opts->int_mult); - mutex_unlock(&opts->lock); - - return result; -} - -static ssize_t f_ss_opts_int_mult_store(struct f_ss_opts *opts, - const char *page, size_t len) -{ - int ret; - u8 num; - - mutex_lock(&opts->lock); - if (opts->refcnt) { - ret = -EBUSY; - goto end; - } - - ret = kstrtou8(page, 0, &num); - if (ret) - goto end; - - if (num > 2) { - ret = -EINVAL; - goto end; - } - - opts->int_mult = num; - ret = len; -end: - mutex_unlock(&opts->lock); - return ret; -} - -static struct f_ss_opts_attribute f_ss_opts_int_mult = - __CONFIGFS_ATTR(int_mult, S_IRUGO | S_IWUSR, - f_ss_opts_int_mult_show, - f_ss_opts_int_mult_store); - -static ssize_t f_ss_opts_int_maxburst_show(struct f_ss_opts *opts, char *page) -{ - int result; - - mutex_lock(&opts->lock); - result = sprintf(page, "%u", opts->int_maxburst); - mutex_unlock(&opts->lock); - - return result; -} - -static ssize_t f_ss_opts_int_maxburst_store(struct f_ss_opts *opts, - const char *page, size_t len) -{ - int ret; - u8 num; - - mutex_lock(&opts->lock); - if (opts->refcnt) { - ret = -EBUSY; - goto end; - } - - ret = kstrtou8(page, 0, &num); - if (ret) - goto end; - - if (num > 15) { - ret = -EINVAL; - goto end; - } - - opts->int_maxburst = num; - ret = len; -end: - mutex_unlock(&opts->lock); - return ret; -} - -static struct f_ss_opts_attribute f_ss_opts_int_maxburst = - __CONFIGFS_ATTR(int_maxburst, S_IRUGO | S_IWUSR, - f_ss_opts_int_maxburst_show, - f_ss_opts_int_maxburst_store); - static struct configfs_attribute *ss_attrs[] = { &f_ss_opts_pattern.attr, &f_ss_opts_isoc_interval.attr, @@ -1651,10 +1186,6 @@ static struct configfs_attribute *ss_attrs[] = { &f_ss_opts_isoc_mult.attr, &f_ss_opts_isoc_maxburst.attr, &f_ss_opts_bulk_buflen.attr, - &f_ss_opts_int_interval.attr, - &f_ss_opts_int_maxpacket.attr, - &f_ss_opts_int_mult.attr, - &f_ss_opts_int_maxburst.attr, NULL, }; @@ -1684,8 +1215,6 @@ static struct usb_function_instance *source_sink_alloc_inst(void) ss_opts->isoc_interval = GZERO_ISOC_INTERVAL; ss_opts->isoc_maxpacket = GZERO_ISOC_MAXPACKET; ss_opts->bulk_buflen = GZERO_BULK_BUFLEN; - ss_opts->int_interval = GZERO_INT_INTERVAL; - ss_opts->int_maxpacket = GZERO_INT_MAXPACKET; config_group_init_type_name(&ss_opts->func_inst.group, "", &ss_func_type); diff --git a/drivers/usb/gadget/function/g_zero.h b/drivers/usb/gadget/function/g_zero.h index 2ce28b9d97cc..15f180904f8a 100644 --- a/drivers/usb/gadget/function/g_zero.h +++ b/drivers/usb/gadget/function/g_zero.h @@ -10,8 +10,6 @@ #define GZERO_QLEN 32 #define GZERO_ISOC_INTERVAL 4 #define GZERO_ISOC_MAXPACKET 1024 -#define GZERO_INT_INTERVAL 1 /* Default interrupt interval = 1 ms */ -#define GZERO_INT_MAXPACKET 1024 struct usb_zero_options { unsigned pattern; @@ -19,10 +17,6 @@ struct usb_zero_options { unsigned isoc_maxpacket; unsigned isoc_mult; unsigned isoc_maxburst; - unsigned int_interval; /* In ms */ - unsigned int_maxpacket; - unsigned int_mult; - unsigned int_maxburst; unsigned bulk_buflen; unsigned qlen; }; @@ -34,10 +28,6 @@ struct f_ss_opts { unsigned isoc_maxpacket; unsigned isoc_mult; unsigned isoc_maxburst; - unsigned int_interval; /* In ms */ - unsigned int_maxpacket; - unsigned int_mult; - unsigned int_maxburst; unsigned bulk_buflen; /* @@ -72,7 +62,6 @@ int lb_modinit(void); void free_ep_req(struct usb_ep *ep, struct usb_request *req); void disable_endpoints(struct usb_composite_dev *cdev, struct usb_ep *in, struct usb_ep *out, - struct usb_ep *iso_in, struct usb_ep *iso_out, - struct usb_ep *int_in, struct usb_ep *int_out); + struct usb_ep *iso_in, struct usb_ep *iso_out); #endif /* __G_ZERO_H */ diff --git a/drivers/usb/gadget/legacy/zero.c b/drivers/usb/gadget/legacy/zero.c index ff97ac93ac03..5ee95152493c 100644 --- a/drivers/usb/gadget/legacy/zero.c +++ b/drivers/usb/gadget/legacy/zero.c @@ -68,8 +68,6 @@ static struct usb_zero_options gzero_options = { .isoc_maxpacket = GZERO_ISOC_MAXPACKET, .bulk_buflen = GZERO_BULK_BUFLEN, .qlen = GZERO_QLEN, - .int_interval = GZERO_INT_INTERVAL, - .int_maxpacket = GZERO_INT_MAXPACKET, }; /*-------------------------------------------------------------------------*/ @@ -268,21 +266,6 @@ module_param_named(isoc_maxburst, gzero_options.isoc_maxburst, uint, S_IRUGO|S_IWUSR); MODULE_PARM_DESC(isoc_maxburst, "0 - 15 (ss only)"); -module_param_named(int_interval, gzero_options.int_interval, uint, - S_IRUGO|S_IWUSR); -MODULE_PARM_DESC(int_interval, "1 - 16"); - -module_param_named(int_maxpacket, gzero_options.int_maxpacket, uint, - S_IRUGO|S_IWUSR); -MODULE_PARM_DESC(int_maxpacket, "0 - 1023 (fs), 0 - 1024 (hs/ss)"); - -module_param_named(int_mult, gzero_options.int_mult, uint, S_IRUGO|S_IWUSR); -MODULE_PARM_DESC(int_mult, "0 - 2 (hs/ss only)"); - -module_param_named(int_maxburst, gzero_options.int_maxburst, uint, - S_IRUGO|S_IWUSR); -MODULE_PARM_DESC(int_maxburst, "0 - 15 (ss only)"); - static struct usb_function *func_lb; static struct usb_function_instance *func_inst_lb; @@ -318,10 +301,6 @@ static int __init zero_bind(struct usb_composite_dev *cdev) ss_opts->isoc_maxpacket = gzero_options.isoc_maxpacket; ss_opts->isoc_mult = gzero_options.isoc_mult; ss_opts->isoc_maxburst = gzero_options.isoc_maxburst; - ss_opts->int_interval = gzero_options.int_interval; - ss_opts->int_maxpacket = gzero_options.int_maxpacket; - ss_opts->int_mult = gzero_options.int_mult; - ss_opts->int_maxburst = gzero_options.int_maxburst; ss_opts->bulk_buflen = gzero_options.bulk_buflen; func_ss = usb_get_function(func_inst_ss); -- cgit v1.2.1 From 7fd6f640f2dd17dac6ddd6702c378cb0bb9cfa11 Mon Sep 17 00:00:00 2001 From: Peter Hurley Date: Wed, 11 Mar 2015 09:19:16 -0400 Subject: serial: 8250_dw: Fix deadlock in LCR workaround Trying to write console output from within the serial console driver while the port->lock is held causes recursive deadlock: CPU 0 spin_lock_irqsave(&port->lock) printk() console_unlock() call_console_drivers() serial8250_console_write() spin_lock_irqsave(&port->lock) ** DEADLOCK ** The 8250_dw i/o accessors try to write a console error message if the LCR workaround was unsuccessful. When the port->lock is already held (eg., when called from serial8250_set_termios()), this deadlocks. Make the error message a FIXME until a general solution is devised. Cc: Tim Kryger Reported-by: Zhang Zhen Signed-off-by: Peter Hurley Signed-off-by: Greg Kroah-Hartman --- drivers/tty/serial/8250/8250_dw.c | 15 ++++++++++++--- 1 file changed, 12 insertions(+), 3 deletions(-) diff --git a/drivers/tty/serial/8250/8250_dw.c b/drivers/tty/serial/8250/8250_dw.c index 2ab229ddee38..6ae5b8560e4d 100644 --- a/drivers/tty/serial/8250/8250_dw.c +++ b/drivers/tty/serial/8250/8250_dw.c @@ -119,7 +119,10 @@ static void dw8250_serial_out(struct uart_port *p, int offset, int value) dw8250_force_idle(p); writeb(value, p->membase + (UART_LCR << p->regshift)); } - dev_err(p->dev, "Couldn't set LCR to %d\n", value); + /* + * FIXME: this deadlocks if port->lock is already held + * dev_err(p->dev, "Couldn't set LCR to %d\n", value); + */ } } @@ -163,7 +166,10 @@ static void dw8250_serial_outq(struct uart_port *p, int offset, int value) __raw_writeq(value & 0xff, p->membase + (UART_LCR << p->regshift)); } - dev_err(p->dev, "Couldn't set LCR to %d\n", value); + /* + * FIXME: this deadlocks if port->lock is already held + * dev_err(p->dev, "Couldn't set LCR to %d\n", value); + */ } } #endif /* CONFIG_64BIT */ @@ -187,7 +193,10 @@ static void dw8250_serial_out32(struct uart_port *p, int offset, int value) dw8250_force_idle(p); writel(value, p->membase + (UART_LCR << p->regshift)); } - dev_err(p->dev, "Couldn't set LCR to %d\n", value); + /* + * FIXME: this deadlocks if port->lock is already held + * dev_err(p->dev, "Couldn't set LCR to %d\n", value); + */ } } -- cgit v1.2.1 From a415457733b5fa40bc996bf1f4df471cd98d3608 Mon Sep 17 00:00:00 2001 From: "oliver@neukum.org" Date: Tue, 10 Mar 2015 16:36:22 +0100 Subject: HID: add ALWAYS_POLL quirk for a Logitech 0xc007 This device disconnects every 60s without X Signed-off-by: Oliver Neukum Signed-off-by: Jiri Kosina --- drivers/hid/hid-ids.h | 1 + drivers/hid/usbhid/hid-quirks.c | 1 + 2 files changed, 2 insertions(+) diff --git a/drivers/hid/hid-ids.h b/drivers/hid/hid-ids.h index 204312bfab2c..8a7c347e8080 100644 --- a/drivers/hid/hid-ids.h +++ b/drivers/hid/hid-ids.h @@ -586,6 +586,7 @@ #define USB_VENDOR_ID_LOGITECH 0x046d #define USB_DEVICE_ID_LOGITECH_AUDIOHUB 0x0a0e #define USB_DEVICE_ID_LOGITECH_T651 0xb00c +#define USB_DEVICE_ID_LOGITECH_C077 0xc007 #define USB_DEVICE_ID_LOGITECH_RECEIVER 0xc101 #define USB_DEVICE_ID_LOGITECH_HARMONY_FIRST 0xc110 #define USB_DEVICE_ID_LOGITECH_HARMONY_LAST 0xc14f diff --git a/drivers/hid/usbhid/hid-quirks.c b/drivers/hid/usbhid/hid-quirks.c index 9be99a67bfe2..a82127753461 100644 --- a/drivers/hid/usbhid/hid-quirks.c +++ b/drivers/hid/usbhid/hid-quirks.c @@ -78,6 +78,7 @@ static const struct hid_blacklist { { USB_VENDOR_ID_ELO, USB_DEVICE_ID_ELO_TS2700, HID_QUIRK_NOGET }, { USB_VENDOR_ID_FORMOSA, USB_DEVICE_ID_FORMOSA_IR_RECEIVER, HID_QUIRK_NO_INIT_REPORTS }, { USB_VENDOR_ID_FREESCALE, USB_DEVICE_ID_FREESCALE_MX28, HID_QUIRK_NOGET }, + { USB_VENDOR_ID_LOGITECH, USB_DEVICE_ID_LOGITECH_C077, HID_QUIRK_ALWAYS_POLL }, { USB_VENDOR_ID_MGE, USB_DEVICE_ID_MGE_UPS, HID_QUIRK_NOGET }, { USB_VENDOR_ID_MICROSOFT, USB_DEVICE_ID_MS_TYPE_COVER_3, HID_QUIRK_NO_INIT_REPORTS }, { USB_VENDOR_ID_MICROSOFT, USB_DEVICE_ID_MS_TYPE_COVER_3_JP, HID_QUIRK_NO_INIT_REPORTS }, -- cgit v1.2.1 From 366c1bd191c49380d81b15b96cf7d4e3528a82a2 Mon Sep 17 00:00:00 2001 From: chas williams - CONTRACTOR Date: Wed, 11 Mar 2015 16:18:01 -0400 Subject: MAINTAINERS: Update my email address Changed to my private email address. Signed-off-by: Chas Williams -- CONTRACTOR Signed-off-by: David S. Miller --- MAINTAINERS | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index 69cc89f7a9c9..80dddbbee5fc 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -1730,7 +1730,7 @@ S: Maintained F: drivers/net/ethernet/atheros/ ATM -M: Chas Williams +M: Chas Williams <3chas3@gmail.com> L: linux-atm-general@lists.sourceforge.net (moderated for non-subscribers) L: netdev@vger.kernel.org W: http://linux-atm.sourceforge.net -- cgit v1.2.1 From 9949afa42be0b76f5832db112ce51bb6b35b2abb Mon Sep 17 00:00:00 2001 From: Neal Cardwell Date: Tue, 10 Mar 2015 17:17:03 -0400 Subject: tcp: fix tcp_cong_avoid_ai() credit accumulation bug with decreases in w The recent change to tcp_cong_avoid_ai() to handle stretch ACKs introduced a bug where snd_cwnd_cnt could accumulate a very large value while w was large, and then if w was reduced snd_cwnd could be incremented by a large delta, leading to a large burst and high packet loss. This was tickled when CUBIC's bictcp_update() sets "ca->cnt = 100 * cwnd". This bug crept in while preparing the upstream version of 814d488c6126. Testing: This patch has been tested in datacenter netperf transfers and live youtube.com and google.com servers. Fixes: 814d488c6126 ("tcp: fix the timid additive increase on stretch ACKs") Signed-off-by: Neal Cardwell Signed-off-by: Yuchung Cheng Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/tcp_cong.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/net/ipv4/tcp_cong.c b/net/ipv4/tcp_cong.c index d694088214cd..62856e185a93 100644 --- a/net/ipv4/tcp_cong.c +++ b/net/ipv4/tcp_cong.c @@ -378,6 +378,12 @@ EXPORT_SYMBOL_GPL(tcp_slow_start); */ void tcp_cong_avoid_ai(struct tcp_sock *tp, u32 w, u32 acked) { + /* If credits accumulated at a higher w, apply them gently now. */ + if (tp->snd_cwnd_cnt >= w) { + tp->snd_cwnd_cnt = 0; + tp->snd_cwnd++; + } + tp->snd_cwnd_cnt += acked; if (tp->snd_cwnd_cnt >= w) { u32 delta = tp->snd_cwnd_cnt / w; -- cgit v1.2.1 From d578e18ce93f5d33a7120fd57c453e22a4c0fc37 Mon Sep 17 00:00:00 2001 From: Neal Cardwell Date: Tue, 10 Mar 2015 17:17:04 -0400 Subject: tcp: restore 1.5x per RTT limit to CUBIC cwnd growth in congestion avoidance Commit 814d488c6126 ("tcp: fix the timid additive increase on stretch ACKs") fixed a bug where tcp_cong_avoid_ai() would either credit a connection with an increase of snd_cwnd_cnt, or increase snd_cwnd, but not both, resulting in cwnd increasing by 1 packet on at most every alternate invocation of tcp_cong_avoid_ai(). Although the commit correctly implemented the CUBIC algorithm, which can increase cwnd by as much as 1 packet per 1 packet ACKed (2x per RTT), in practice that could be too aggressive: in tests on network paths with small buffers, YouTube server retransmission rates nearly doubled. This commit restores CUBIC to a maximum cwnd growth rate of 1 packet per 2 packets ACKed (1.5x per RTT). In YouTube tests this restored retransmit rates to low levels. Testing: This patch has been tested in datacenter netperf transfers and live youtube.com and google.com servers. Fixes: 9cd981dcf174 ("tcp: fix stretch ACK bugs in CUBIC") Signed-off-by: Neal Cardwell Signed-off-by: Yuchung Cheng Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/tcp_cubic.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/net/ipv4/tcp_cubic.c b/net/ipv4/tcp_cubic.c index 4b276d1ed980..06d3d665a9fd 100644 --- a/net/ipv4/tcp_cubic.c +++ b/net/ipv4/tcp_cubic.c @@ -306,8 +306,10 @@ tcp_friendliness: } } - if (ca->cnt == 0) /* cannot be zero */ - ca->cnt = 1; + /* The maximum rate of cwnd increase CUBIC allows is 1 packet per + * 2 packets ACKed, meaning cwnd grows at 1.5x per RTT. + */ + ca->cnt = max(ca->cnt, 2U); } static void bictcp_cong_avoid(struct sock *sk, u32 ack, u32 acked) -- cgit v1.2.1 From b1cb59cf2efe7971d3d72a7b963d09a512d994c9 Mon Sep 17 00:00:00 2001 From: Alexey Kodanev Date: Wed, 11 Mar 2015 14:29:17 +0300 Subject: net: sysctl_net_core: check SNDBUF and RCVBUF for min length sysctl has sysctl.net.core.rmem_*/wmem_* parameters which can be set to incorrect values. Given that 'struct sk_buff' allocates from rcvbuf, incorrectly set buffer length could result to memory allocation failures. For example, set them as follows: # sysctl net.core.rmem_default=64 net.core.wmem_default = 64 # sysctl net.core.wmem_default=64 net.core.wmem_default = 64 # ping localhost -s 1024 -i 0 > /dev/null This could result to the following failure: skbuff: skb_over_panic: text:ffffffff81628db4 len:-32 put:-32 head:ffff88003a1cc200 data:ffff88003a1cc200 tail:0xffffffe0 end:0xc0 dev: kernel BUG at net/core/skbuff.c:102! invalid opcode: 0000 [#1] SMP ... task: ffff88003b7f5550 ti: ffff88003ae88000 task.ti: ffff88003ae88000 RIP: 0010:[] [] skb_put+0xa1/0xb0 RSP: 0018:ffff88003ae8bc68 EFLAGS: 00010296 RAX: 000000000000008d RBX: 00000000ffffffe0 RCX: 0000000000000000 RDX: ffff88003fdcf598 RSI: ffff88003fdcd9c8 RDI: ffff88003fdcd9c8 RBP: ffff88003ae8bc88 R08: 0000000000000001 R09: 0000000000000000 R10: 0000000000000001 R11: 00000000000002b2 R12: 0000000000000000 R13: 0000000000000000 R14: ffff88003d3f7300 R15: ffff88000012a900 FS: 00007fa0e2b4a840(0000) GS:ffff88003fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000000d0f7e0 CR3: 000000003b8fb000 CR4: 00000000000006f0 Stack: ffff88003a1cc200 00000000ffffffe0 00000000000000c0 ffffffff818cab1d ffff88003ae8bd68 ffffffff81628db4 ffff88003ae8bd48 ffff88003b7f5550 ffff880031a09408 ffff88003b7f5550 ffff88000012aa48 ffff88000012ab00 Call Trace: [] unix_stream_sendmsg+0x2c4/0x470 [] sock_write_iter+0x146/0x160 [] new_sync_write+0x92/0xd0 [] vfs_write+0xd6/0x180 [] SyS_write+0x59/0xd0 [] system_call_fastpath+0x12/0x17 Code: 00 00 48 89 44 24 10 8b 87 c8 00 00 00 48 89 44 24 08 48 8b 87 d8 00 00 00 48 c7 c7 30 db 91 81 48 89 04 24 31 c0 e8 4f a8 0e 00 <0f> 0b eb fe 66 66 2e 0f 1f 84 00 00 00 00 00 55 48 89 e5 48 83 RIP [] skb_put+0xa1/0xb0 RSP Kernel panic - not syncing: Fatal exception Moreover, the possible minimum is 1, so we can get another kernel panic: ... BUG: unable to handle kernel paging request at ffff88013caee5c0 IP: [] __alloc_skb+0x12f/0x1f0 ... Signed-off-by: Alexey Kodanev Signed-off-by: David S. Miller --- net/core/sysctl_net_core.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/net/core/sysctl_net_core.c b/net/core/sysctl_net_core.c index 433424804284..8ce351ffceb1 100644 --- a/net/core/sysctl_net_core.c +++ b/net/core/sysctl_net_core.c @@ -25,6 +25,8 @@ static int zero = 0; static int one = 1; static int ushort_max = USHRT_MAX; +static int min_sndbuf = SOCK_MIN_SNDBUF; +static int min_rcvbuf = SOCK_MIN_RCVBUF; static int net_msg_warn; /* Unused, but still a sysctl */ @@ -237,7 +239,7 @@ static struct ctl_table net_core_table[] = { .maxlen = sizeof(int), .mode = 0644, .proc_handler = proc_dointvec_minmax, - .extra1 = &one, + .extra1 = &min_sndbuf, }, { .procname = "rmem_max", @@ -245,7 +247,7 @@ static struct ctl_table net_core_table[] = { .maxlen = sizeof(int), .mode = 0644, .proc_handler = proc_dointvec_minmax, - .extra1 = &one, + .extra1 = &min_rcvbuf, }, { .procname = "wmem_default", @@ -253,7 +255,7 @@ static struct ctl_table net_core_table[] = { .maxlen = sizeof(int), .mode = 0644, .proc_handler = proc_dointvec_minmax, - .extra1 = &one, + .extra1 = &min_sndbuf, }, { .procname = "rmem_default", @@ -261,7 +263,7 @@ static struct ctl_table net_core_table[] = { .maxlen = sizeof(int), .mode = 0644, .proc_handler = proc_dointvec_minmax, - .extra1 = &one, + .extra1 = &min_rcvbuf, }, { .procname = "dev_weight", -- cgit v1.2.1 From c8a4d29988edb0db9ee80669f2e5e21bd9f7e0d0 Mon Sep 17 00:00:00 2001 From: David Vrabel Date: Wed, 11 Mar 2015 15:27:59 +0000 Subject: xen-netback: notify immediately after pushing Tx response. This fixes a performance regression introduced by 7fbb9d8415d4a51cf542e87cf3a717a9f7e6aedc (xen-netback: release pending index before pushing Tx responses) Moving the notify outside of the spin locks means it can be delayed a long time (if the dealloc thread is descheduled or there is an interrupt or softirq). Signed-off-by: David Vrabel Reviewed-by: Zoltan Kiss Acked-by: Wei Liu Signed-off-by: David S. Miller --- drivers/net/xen-netback/netback.c | 23 ++++++++++++----------- 1 file changed, 12 insertions(+), 11 deletions(-) diff --git a/drivers/net/xen-netback/netback.c b/drivers/net/xen-netback/netback.c index cab9f5257f57..997cf0901ac2 100644 --- a/drivers/net/xen-netback/netback.c +++ b/drivers/net/xen-netback/netback.c @@ -96,6 +96,7 @@ static void xenvif_idx_release(struct xenvif_queue *queue, u16 pending_idx, static void make_tx_response(struct xenvif_queue *queue, struct xen_netif_tx_request *txp, s8 st); +static void push_tx_responses(struct xenvif_queue *queue); static inline int tx_work_todo(struct xenvif_queue *queue); @@ -655,15 +656,10 @@ static void xenvif_tx_err(struct xenvif_queue *queue, unsigned long flags; do { - int notify; - spin_lock_irqsave(&queue->response_lock, flags); make_tx_response(queue, txp, XEN_NETIF_RSP_ERROR); - RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&queue->tx, notify); + push_tx_responses(queue); spin_unlock_irqrestore(&queue->response_lock, flags); - if (notify) - notify_remote_via_irq(queue->tx_irq); - if (cons == end) break; txp = RING_GET_REQUEST(&queue->tx, cons++); @@ -1657,7 +1653,6 @@ static void xenvif_idx_release(struct xenvif_queue *queue, u16 pending_idx, { struct pending_tx_info *pending_tx_info; pending_ring_idx_t index; - int notify; unsigned long flags; pending_tx_info = &queue->pending_tx_info[pending_idx]; @@ -1673,12 +1668,9 @@ static void xenvif_idx_release(struct xenvif_queue *queue, u16 pending_idx, index = pending_index(queue->pending_prod++); queue->pending_ring[index] = pending_idx; - RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&queue->tx, notify); + push_tx_responses(queue); spin_unlock_irqrestore(&queue->response_lock, flags); - - if (notify) - notify_remote_via_irq(queue->tx_irq); } @@ -1699,6 +1691,15 @@ static void make_tx_response(struct xenvif_queue *queue, queue->tx.rsp_prod_pvt = ++i; } +static void push_tx_responses(struct xenvif_queue *queue) +{ + int notify; + + RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&queue->tx, notify); + if (notify) + notify_remote_via_irq(queue->tx_irq); +} + static struct xen_netif_rx_response *make_rx_response(struct xenvif_queue *queue, u16 id, s8 st, -- cgit v1.2.1 From c29390c6dfeee0944ac6b5610ebbe403944378fc Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Wed, 11 Mar 2015 18:42:02 -0700 Subject: xps: must clear sender_cpu before forwarding John reported that my previous commit added a regression on his router. This is because sender_cpu & napi_id share a common location, so get_xps_queue() can see garbage and perform an out of bound access. We need to make sure sender_cpu is cleared before doing the transmit, otherwise any NIC busy poll enabled (skb_mark_napi_id()) can trigger this bug. Signed-off-by: Eric Dumazet Reported-by: John Bisected-by: John Fixes: 2bd82484bb4c ("xps: fix xps for stacked devices") Signed-off-by: David S. Miller --- include/linux/skbuff.h | 7 +++++++ net/core/skbuff.c | 2 +- net/ipv4/ip_forward.c | 1 + net/ipv6/ip6_output.c | 1 + 4 files changed, 10 insertions(+), 1 deletion(-) diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h index 30007afe70b3..f54d6659713a 100644 --- a/include/linux/skbuff.h +++ b/include/linux/skbuff.h @@ -948,6 +948,13 @@ static inline void skb_copy_hash(struct sk_buff *to, const struct sk_buff *from) to->l4_hash = from->l4_hash; }; +static inline void skb_sender_cpu_clear(struct sk_buff *skb) +{ +#ifdef CONFIG_XPS + skb->sender_cpu = 0; +#endif +} + #ifdef NET_SKBUFF_DATA_USES_OFFSET static inline unsigned char *skb_end_pointer(const struct sk_buff *skb) { diff --git a/net/core/skbuff.c b/net/core/skbuff.c index f80507823531..434e78e5254d 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -4173,7 +4173,7 @@ void skb_scrub_packet(struct sk_buff *skb, bool xnet) skb->ignore_df = 0; skb_dst_drop(skb); skb->mark = 0; - skb->sender_cpu = 0; + skb_sender_cpu_clear(skb); skb_init_secmark(skb); secpath_reset(skb); nf_reset(skb); diff --git a/net/ipv4/ip_forward.c b/net/ipv4/ip_forward.c index 787b3c294ce6..d9bc28ac5d1b 100644 --- a/net/ipv4/ip_forward.c +++ b/net/ipv4/ip_forward.c @@ -67,6 +67,7 @@ static int ip_forward_finish(struct sk_buff *skb) if (unlikely(opt->optlen)) ip_forward_options(skb); + skb_sender_cpu_clear(skb); return dst_output(skb); } diff --git a/net/ipv6/ip6_output.c b/net/ipv6/ip6_output.c index 0a04a37305d5..7e80b61b51ff 100644 --- a/net/ipv6/ip6_output.c +++ b/net/ipv6/ip6_output.c @@ -318,6 +318,7 @@ static int ip6_forward_proxy_check(struct sk_buff *skb) static inline int ip6_forward_finish(struct sk_buff *skb) { + skb_sender_cpu_clear(skb); return dst_output(skb); } -- cgit v1.2.1 From 3a8dd9711e0792f64394edafadd66c2d1f1904df Mon Sep 17 00:00:00 2001 From: Willem de Bruijn Date: Wed, 11 Mar 2015 15:43:55 -0400 Subject: sock: fix possible NULL sk dereference in __skb_tstamp_tx Test that sk != NULL before reading sk->sk_tsflags. Fixes: 49ca0d8bfaf3 ("net-timestamp: no-payload option") Reported-by: One Thousand Gnomes Signed-off-by: Willem de Bruijn Signed-off-by: David S. Miller --- net/core/skbuff.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/net/core/skbuff.c b/net/core/skbuff.c index 434e78e5254d..8e4ac97c8477 100644 --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -3733,9 +3733,13 @@ void __skb_tstamp_tx(struct sk_buff *orig_skb, struct sock *sk, int tstype) { struct sk_buff *skb; - bool tsonly = sk->sk_tsflags & SOF_TIMESTAMPING_OPT_TSONLY; + bool tsonly; - if (!sk || !skb_may_tx_timestamp(sk, tsonly)) + if (!sk) + return; + + tsonly = sk->sk_tsflags & SOF_TIMESTAMPING_OPT_TSONLY; + if (!skb_may_tx_timestamp(sk, tsonly)) return; if (tsonly) -- cgit v1.2.1 From f862e07cf95d5b62a5fc5e981dd7d0dbaf33a501 Mon Sep 17 00:00:00 2001 From: Arnd Bergmann Date: Wed, 11 Mar 2015 22:46:59 +0100 Subject: rds: avoid potential stack overflow The rds_iw_update_cm_id function stores a large 'struct rds_sock' object on the stack in order to pass a pair of addresses. This happens to just fit withint the 1024 byte stack size warning limit on x86, but just exceed that limit on ARM, which gives us this warning: net/rds/iw_rdma.c:200:1: warning: the frame size of 1056 bytes is larger than 1024 bytes [-Wframe-larger-than=] As the use of this large variable is basically bogus, we can rearrange the code to not do that. Instead of passing an rds socket into rds_iw_get_device, we now just pass the two addresses that we have available in rds_iw_update_cm_id, and we change rds_iw_get_mr accordingly, to create two address structures on the stack there. Signed-off-by: Arnd Bergmann Acked-by: Sowmini Varadhan Signed-off-by: David S. Miller --- net/rds/iw_rdma.c | 40 ++++++++++++++++++++++------------------ 1 file changed, 22 insertions(+), 18 deletions(-) diff --git a/net/rds/iw_rdma.c b/net/rds/iw_rdma.c index a817705ce2d0..dba8d0864f18 100644 --- a/net/rds/iw_rdma.c +++ b/net/rds/iw_rdma.c @@ -88,7 +88,9 @@ static unsigned int rds_iw_unmap_fastreg_list(struct rds_iw_mr_pool *pool, int *unpinned); static void rds_iw_destroy_fastreg(struct rds_iw_mr_pool *pool, struct rds_iw_mr *ibmr); -static int rds_iw_get_device(struct rds_sock *rs, struct rds_iw_device **rds_iwdev, struct rdma_cm_id **cm_id) +static int rds_iw_get_device(struct sockaddr_in *src, struct sockaddr_in *dst, + struct rds_iw_device **rds_iwdev, + struct rdma_cm_id **cm_id) { struct rds_iw_device *iwdev; struct rds_iw_cm_id *i_cm_id; @@ -112,15 +114,15 @@ static int rds_iw_get_device(struct rds_sock *rs, struct rds_iw_device **rds_iwd src_addr->sin_port, dst_addr->sin_addr.s_addr, dst_addr->sin_port, - rs->rs_bound_addr, - rs->rs_bound_port, - rs->rs_conn_addr, - rs->rs_conn_port); + src->sin_addr.s_addr, + src->sin_port, + dst->sin_addr.s_addr, + dst->sin_port); #ifdef WORKING_TUPLE_DETECTION - if (src_addr->sin_addr.s_addr == rs->rs_bound_addr && - src_addr->sin_port == rs->rs_bound_port && - dst_addr->sin_addr.s_addr == rs->rs_conn_addr && - dst_addr->sin_port == rs->rs_conn_port) { + if (src_addr->sin_addr.s_addr == src->sin_addr.s_addr && + src_addr->sin_port == src->sin_port && + dst_addr->sin_addr.s_addr == dst->sin_addr.s_addr && + dst_addr->sin_port == dst->sin_port) { #else /* FIXME - needs to compare the local and remote * ipaddr/port tuple, but the ipaddr is the only @@ -128,7 +130,7 @@ static int rds_iw_get_device(struct rds_sock *rs, struct rds_iw_device **rds_iwd * zero'ed. It doesn't appear to be properly populated * during connection setup... */ - if (src_addr->sin_addr.s_addr == rs->rs_bound_addr) { + if (src_addr->sin_addr.s_addr == src->sin_addr.s_addr) { #endif spin_unlock_irq(&iwdev->spinlock); *rds_iwdev = iwdev; @@ -180,19 +182,13 @@ int rds_iw_update_cm_id(struct rds_iw_device *rds_iwdev, struct rdma_cm_id *cm_i { struct sockaddr_in *src_addr, *dst_addr; struct rds_iw_device *rds_iwdev_old; - struct rds_sock rs; struct rdma_cm_id *pcm_id; int rc; src_addr = (struct sockaddr_in *)&cm_id->route.addr.src_addr; dst_addr = (struct sockaddr_in *)&cm_id->route.addr.dst_addr; - rs.rs_bound_addr = src_addr->sin_addr.s_addr; - rs.rs_bound_port = src_addr->sin_port; - rs.rs_conn_addr = dst_addr->sin_addr.s_addr; - rs.rs_conn_port = dst_addr->sin_port; - - rc = rds_iw_get_device(&rs, &rds_iwdev_old, &pcm_id); + rc = rds_iw_get_device(src_addr, dst_addr, &rds_iwdev_old, &pcm_id); if (rc) rds_iw_remove_cm_id(rds_iwdev, cm_id); @@ -598,9 +594,17 @@ void *rds_iw_get_mr(struct scatterlist *sg, unsigned long nents, struct rds_iw_device *rds_iwdev; struct rds_iw_mr *ibmr = NULL; struct rdma_cm_id *cm_id; + struct sockaddr_in src = { + .sin_addr.s_addr = rs->rs_bound_addr, + .sin_port = rs->rs_bound_port, + }; + struct sockaddr_in dst = { + .sin_addr.s_addr = rs->rs_conn_addr, + .sin_port = rs->rs_conn_port, + }; int ret; - ret = rds_iw_get_device(rs, &rds_iwdev, &cm_id); + ret = rds_iw_get_device(&src, &dst, &rds_iwdev, &cm_id); if (ret || !cm_id) { ret = -ENODEV; goto out; -- cgit v1.2.1 From 0f9722e37f8359e228f4c04e6822ad0bf9a4e083 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Fri, 27 Feb 2015 12:34:10 +0800 Subject: phy: exynos-dp-video: Kill exynos_dp_video_phy_pwr_isol function If IS_ERR(state->regs) the .probe fails. So IS_ERR(state->regs) test in exynos_dp_video_phy_pwr_isol() is not necessary. exynos_dp_video_phy_pwr_isol() simply does a regmap_update_bits() call now, just call regmap_update_bits() instead and return proper return value. Signed-off-by: Axel Lin Reviewed-by: Sylwester Nawrocki Signed-off-by: Kishon Vijay Abraham I --- drivers/phy/phy-exynos-dp-video.c | 24 ++++-------------------- 1 file changed, 4 insertions(+), 20 deletions(-) diff --git a/drivers/phy/phy-exynos-dp-video.c b/drivers/phy/phy-exynos-dp-video.c index f86cbe68ddaf..179cbf9451aa 100644 --- a/drivers/phy/phy-exynos-dp-video.c +++ b/drivers/phy/phy-exynos-dp-video.c @@ -30,28 +30,13 @@ struct exynos_dp_video_phy { const struct exynos_dp_video_phy_drvdata *drvdata; }; -static void exynos_dp_video_phy_pwr_isol(struct exynos_dp_video_phy *state, - unsigned int on) -{ - unsigned int val; - - if (IS_ERR(state->regs)) - return; - - val = on ? 0 : EXYNOS5_PHY_ENABLE; - - regmap_update_bits(state->regs, state->drvdata->phy_ctrl_offset, - EXYNOS5_PHY_ENABLE, val); -} - static int exynos_dp_video_phy_power_on(struct phy *phy) { struct exynos_dp_video_phy *state = phy_get_drvdata(phy); /* Disable power isolation on DP-PHY */ - exynos_dp_video_phy_pwr_isol(state, 0); - - return 0; + return regmap_update_bits(state->regs, state->drvdata->phy_ctrl_offset, + EXYNOS5_PHY_ENABLE, EXYNOS5_PHY_ENABLE); } static int exynos_dp_video_phy_power_off(struct phy *phy) @@ -59,9 +44,8 @@ static int exynos_dp_video_phy_power_off(struct phy *phy) struct exynos_dp_video_phy *state = phy_get_drvdata(phy); /* Enable power isolation on DP-PHY */ - exynos_dp_video_phy_pwr_isol(state, 1); - - return 0; + return regmap_update_bits(state->regs, state->drvdata->phy_ctrl_offset, + EXYNOS5_PHY_ENABLE, 0); } static struct phy_ops exynos_dp_video_phy_ops = { -- cgit v1.2.1 From 1cbdfc48c3d4064e2515e4d837e714fd1aaa2d6e Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Mon, 2 Mar 2015 16:10:54 +0800 Subject: phy: hix5hd2-sata: Check return value of platform_get_resource This prevent NULL pointer dereference if res is NULL. Signed-off-by: Axel Lin Acked-by: Zhangfei Gao Signed-off-by: Kishon Vijay Abraham I --- drivers/phy/phy-hix5hd2-sata.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/phy/phy-hix5hd2-sata.c b/drivers/phy/phy-hix5hd2-sata.c index 34915b4202f1..d6b22659cac1 100644 --- a/drivers/phy/phy-hix5hd2-sata.c +++ b/drivers/phy/phy-hix5hd2-sata.c @@ -147,6 +147,9 @@ static int hix5hd2_sata_phy_probe(struct platform_device *pdev) return -ENOMEM; res = platform_get_resource(pdev, IORESOURCE_MEM, 0); + if (!res) + return -EINVAL; + priv->base = devm_ioremap(dev, res->start, resource_size(res)); if (!priv->base) return -ENOMEM; -- cgit v1.2.1 From bd4abc2f96dd06c0085112240aed74c96538d6e7 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Tue, 3 Mar 2015 09:08:43 +0800 Subject: phy: samsung-usb2: Remove NULL terminating entry from phys array Current code uses num_phys settings to tell the number of entries in phys. Thus remove the NULL terminating entry from phys array which is not necessary. Signed-off-by: Axel Lin Signed-off-by: Kishon Vijay Abraham I --- drivers/phy/phy-exynos4210-usb2.c | 1 - drivers/phy/phy-exynos4x12-usb2.c | 1 - drivers/phy/phy-exynos5250-usb2.c | 1 - 3 files changed, 3 deletions(-) diff --git a/drivers/phy/phy-exynos4210-usb2.c b/drivers/phy/phy-exynos4210-usb2.c index 236a52ad94eb..f30bbb0fb3b2 100644 --- a/drivers/phy/phy-exynos4210-usb2.c +++ b/drivers/phy/phy-exynos4210-usb2.c @@ -250,7 +250,6 @@ static const struct samsung_usb2_common_phy exynos4210_phys[] = { .power_on = exynos4210_power_on, .power_off = exynos4210_power_off, }, - {}, }; const struct samsung_usb2_phy_config exynos4210_usb2_phy_config = { diff --git a/drivers/phy/phy-exynos4x12-usb2.c b/drivers/phy/phy-exynos4x12-usb2.c index 0b9de88579b1..765da90a536f 100644 --- a/drivers/phy/phy-exynos4x12-usb2.c +++ b/drivers/phy/phy-exynos4x12-usb2.c @@ -361,7 +361,6 @@ static const struct samsung_usb2_common_phy exynos4x12_phys[] = { .power_on = exynos4x12_power_on, .power_off = exynos4x12_power_off, }, - {}, }; const struct samsung_usb2_phy_config exynos3250_usb2_phy_config = { diff --git a/drivers/phy/phy-exynos5250-usb2.c b/drivers/phy/phy-exynos5250-usb2.c index 1c139aa0d074..2ed1735a076a 100644 --- a/drivers/phy/phy-exynos5250-usb2.c +++ b/drivers/phy/phy-exynos5250-usb2.c @@ -391,7 +391,6 @@ static const struct samsung_usb2_common_phy exynos5250_phys[] = { .power_on = exynos5250_power_on, .power_off = exynos5250_power_off, }, - {}, }; const struct samsung_usb2_phy_config exynos5250_usb2_phy_config = { -- cgit v1.2.1 From a5e5d3c0b239a67d712995f37140e354b5547003 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Tue, 3 Mar 2015 20:04:55 +0800 Subject: phy: ti-pipe3: Simplify ti_pipe3_dpll_wait_lock implementation Code simplification. No functional change. Signed-off-by: Axel Lin Acked-by: Roger Quadros Signed-off-by: Kishon Vijay Abraham I --- drivers/phy/phy-ti-pipe3.c | 10 +++------- 1 file changed, 3 insertions(+), 7 deletions(-) diff --git a/drivers/phy/phy-ti-pipe3.c b/drivers/phy/phy-ti-pipe3.c index 95c88f929f27..ed72b0d01dde 100644 --- a/drivers/phy/phy-ti-pipe3.c +++ b/drivers/phy/phy-ti-pipe3.c @@ -165,15 +165,11 @@ static int ti_pipe3_dpll_wait_lock(struct ti_pipe3 *phy) cpu_relax(); val = ti_pipe3_readl(phy->pll_ctrl_base, PLL_STATUS); if (val & PLL_LOCK) - break; + return 0; } while (!time_after(jiffies, timeout)); - if (!(val & PLL_LOCK)) { - dev_err(phy->dev, "DPLL failed to lock\n"); - return -EBUSY; - } - - return 0; + dev_err(phy->dev, "DPLL failed to lock\n"); + return -EBUSY; } static int ti_pipe3_dpll_program(struct ti_pipe3 *phy) -- cgit v1.2.1 From 6b08e36ba32b3cb4d44af617c8984f9c2bfdf964 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Wed, 4 Mar 2015 09:33:32 +0800 Subject: phy: rockchip-usb: Fixup rockchip_usb_phy_power_on failure path If rockchip_usb_phy_power() fails, we need to call clk_disable_unprepare() before return. This is to ensure we have balanced clk_enable/disable calls. Also remove unneeded ret checking in rockchip_usb_phy_power_off. Signed-off-by: Axel Lin Signed-off-by: Kishon Vijay Abraham I --- drivers/phy/phy-rockchip-usb.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/phy/phy-rockchip-usb.c b/drivers/phy/phy-rockchip-usb.c index 22011c3b6a4b..7d4c33643768 100644 --- a/drivers/phy/phy-rockchip-usb.c +++ b/drivers/phy/phy-rockchip-usb.c @@ -61,8 +61,6 @@ static int rockchip_usb_phy_power_off(struct phy *_phy) return ret; clk_disable_unprepare(phy->clk); - if (ret) - return ret; return 0; } @@ -78,8 +76,10 @@ static int rockchip_usb_phy_power_on(struct phy *_phy) /* Power up usb phy analog blocks by set siddq 0 */ ret = rockchip_usb_phy_power(phy, 0); - if (ret) + if (ret) { + clk_disable_unprepare(phy->clk); return ret; + } return 0; } -- cgit v1.2.1 From 3a43477fa36eee2997313380cb52f486e2bff316 Mon Sep 17 00:00:00 2001 From: Roger Tseng Date: Fri, 6 Mar 2015 11:36:06 +0800 Subject: mfd: rtsx_usb: Prevent DMA from stack Functions rtsx_usb_ep0_read_register() and rtsx_usb_get_card_status() both use arbitrary buffer addresses from arguments directly for DMA and the buffers could be located in stack. This was caught by DMA-API debug check. Fixes this by using double-buffers via kzalloc in both functions to guarantee the validity of DMA buffer. WARNING: CPU: 1 PID: 25 at lib/dma-debug.c:1166 check_for_stack+0x96/0xe0() ehci-pci 0000:00:1a.0: DMA-API: device driver maps memory from stack [addr=ffff8801199e3cef] Modules linked in: rtsx_usb_ms arc4 memstick intel_rapl iosf_mbi rtl8192ce snd_hda_codec_hdmi snd_hda_codec_realtek snd_hda_codec_generic snd_hda_intel rtl_pci rtl8192c_common snd_hda_controller x86_pkg_temp_thermal snd_hda_codec rtlwifi mac80211 coretemp kvm_intel kvm iTCO_wdt snd_hwdep snd_seq snd_seq_device crct10dif_pclmul iTCO_vendor_support sparse_keymap cfg80211 crc32_pclmul snd_pcm crc32c_intel ghash_clmulni_intel rfkill i2c_i801 snd_timer shpchp snd serio_raw mei_me lpc_ich soundcore mei tpm_tis tpm wmi nfsd auth_rpcgss nfs_acl lockd grace sunrpc i915 rtsx_usb_sdmmc mmc_core 8021q uas garp stp i2c_algo_bit llc mrp drm_kms_helper usb_storage drm rtsx_usb mfd_core r8169 mii video CPU: 1 PID: 25 Comm: kworker/1:2 Not tainted 3.20.0-0.rc0.git7.3.fc22.x86_64 #1 Hardware name: WB WB-B06211/WB-B0621, BIOS EB062IWB V1.0 12/12/2013 Workqueue: events rtsx_usb_ms_handle_req [rtsx_usb_ms] 0000000000000000 000000003d188e66 ffff8801199e3808 ffffffff8187642b 0000000000000000 ffff8801199e3860 ffff8801199e3848 ffffffff810ab39a ffff8801199e3864 ffff8801199e3cef ffff880119b57098 ffff880119b37320 Call Trace: [] dump_stack+0x4c/0x65 [] warn_slowpath_common+0x8a/0xc0 [] warn_slowpath_fmt+0x55/0x70 [] ? _raw_spin_unlock_irqrestore+0x36/0x70 [] check_for_stack+0x96/0xe0 [] debug_dma_map_page+0x104/0x150 [] usb_hcd_map_urb_for_dma+0x646/0x790 [] usb_hcd_submit_urb+0x1d5/0xa90 [] ? mark_held_locks+0x7f/0xc0 [] ? mark_held_locks+0x7f/0xc0 [] ? lockdep_init_map+0x65/0x5d0 [] usb_submit_urb+0x42e/0x5f0 [] usb_start_wait_urb+0x77/0x190 [] ? __kmalloc+0x205/0x2d0 [] usb_control_msg+0xdc/0x130 [] rtsx_usb_ep0_read_register+0x59/0x70 [rtsx_usb] [] ? rtsx_usb_get_rsp+0x41/0x50 [rtsx_usb] [] rtsx_usb_ms_handle_req+0x7ce/0x9c5 [rtsx_usb_ms] Reported-by: Josh Boyer Signed-off-by: Roger Tseng Signed-off-by: Lee Jones --- drivers/mfd/rtsx_usb.c | 30 ++++++++++++++++++++++++------ 1 file changed, 24 insertions(+), 6 deletions(-) diff --git a/drivers/mfd/rtsx_usb.c b/drivers/mfd/rtsx_usb.c index ede50244f265..dbd907d7170e 100644 --- a/drivers/mfd/rtsx_usb.c +++ b/drivers/mfd/rtsx_usb.c @@ -196,18 +196,27 @@ EXPORT_SYMBOL_GPL(rtsx_usb_ep0_write_register); int rtsx_usb_ep0_read_register(struct rtsx_ucr *ucr, u16 addr, u8 *data) { u16 value; + u8 *buf; + int ret; if (!data) return -EINVAL; - *data = 0; + + buf = kzalloc(sizeof(u8), GFP_KERNEL); + if (!buf) + return -ENOMEM; addr |= EP0_READ_REG_CMD << EP0_OP_SHIFT; value = swab16(addr); - return usb_control_msg(ucr->pusb_dev, + ret = usb_control_msg(ucr->pusb_dev, usb_rcvctrlpipe(ucr->pusb_dev, 0), RTSX_USB_REQ_REG_OP, USB_DIR_IN | USB_TYPE_VENDOR | USB_RECIP_DEVICE, - value, 0, data, 1, 100); + value, 0, buf, 1, 100); + *data = *buf; + + kfree(buf); + return ret; } EXPORT_SYMBOL_GPL(rtsx_usb_ep0_read_register); @@ -288,18 +297,27 @@ static int rtsx_usb_get_status_with_bulk(struct rtsx_ucr *ucr, u16 *status) int rtsx_usb_get_card_status(struct rtsx_ucr *ucr, u16 *status) { int ret; + u16 *buf; if (!status) return -EINVAL; - if (polling_pipe == 0) + if (polling_pipe == 0) { + buf = kzalloc(sizeof(u16), GFP_KERNEL); + if (!buf) + return -ENOMEM; + ret = usb_control_msg(ucr->pusb_dev, usb_rcvctrlpipe(ucr->pusb_dev, 0), RTSX_USB_REQ_POLL, USB_DIR_IN | USB_TYPE_VENDOR | USB_RECIP_DEVICE, - 0, 0, status, 2, 100); - else + 0, 0, buf, 2, 100); + *status = *buf; + + kfree(buf); + } else { ret = rtsx_usb_get_status_with_bulk(ucr, status); + } /* usb_control_msg may return positive when success */ if (ret < 0) -- cgit v1.2.1 From c8648508ebfc597058d2cd00b6c539110264a167 Mon Sep 17 00:00:00 2001 From: Ameya Palande <2ameya@gmail.com> Date: Thu, 26 Feb 2015 12:05:51 -0800 Subject: mfd: kempld-core: Fix callback return value check On success, callback function returns 0. So invert the if condition check so that we can break out of loop. Cc: stable@vger.kernel.org Signed-off-by: Ameya Palande <2ameya@gmail.com> Signed-off-by: Lee Jones --- drivers/mfd/kempld-core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/mfd/kempld-core.c b/drivers/mfd/kempld-core.c index f38ec424872e..5615522f8d62 100644 --- a/drivers/mfd/kempld-core.c +++ b/drivers/mfd/kempld-core.c @@ -739,7 +739,7 @@ static int __init kempld_init(void) for (id = kempld_dmi_table; id->matches[0].slot != DMI_NONE; id++) if (strstr(id->ident, force_device_id)) - if (id->callback && id->callback(id)) + if (id->callback && !id->callback(id)) break; if (id->matches[0].slot == DMI_NONE) return -ENODEV; -- cgit v1.2.1 From 7486341a98f26857f383aec88ffa10950087c3a1 Mon Sep 17 00:00:00 2001 From: "Li, Aubrey" Date: Wed, 11 Mar 2015 16:09:00 +0800 Subject: x86/platform, acpi: Bypass legacy PIC and PIT in ACPI hardware reduced mode On a platform in ACPI Hardware-reduced mode, the legacy PIC and PIT may not be initialized even though they may be present in silicon. Touching these legacy components causes unexpected results on the system. On the Bay Trail-T(ASUS-T100) platform, touching these legacy components blocks platform hardware low idle power state(S0ix) during system suspend. So we should bypass them in ACPI hardware reduced mode. Suggested-by: Arjan van de Ven Signed-off-by: Li Aubrey Cc: Cc: Alan Cox Cc: H. Peter Anvin Cc: Len Brown Cc: Rafael J. Wysocki Cc: Rafael J. Wysocki Link: http://lkml.kernel.org/r/54FFF81C.20703@linux.intel.com Signed-off-by: Ingo Molnar --- arch/x86/kernel/acpi/boot.c | 25 +++++++++++++++++++++++++ 1 file changed, 25 insertions(+) diff --git a/arch/x86/kernel/acpi/boot.c b/arch/x86/kernel/acpi/boot.c index 3d525c6124f6..803b684676ff 100644 --- a/arch/x86/kernel/acpi/boot.c +++ b/arch/x86/kernel/acpi/boot.c @@ -1337,6 +1337,26 @@ static int __init dmi_ignore_irq0_timer_override(const struct dmi_system_id *d) return 0; } +/* + * ACPI offers an alternative platform interface model that removes + * ACPI hardware requirements for platforms that do not implement + * the PC Architecture. + * + * We initialize the Hardware-reduced ACPI model here: + */ +static void __init acpi_reduced_hw_init(void) +{ + if (acpi_gbl_reduced_hardware) { + /* + * Override x86_init functions and bypass legacy pic + * in Hardware-reduced ACPI mode + */ + x86_init.timers.timer_init = x86_init_noop; + x86_init.irqs.pre_vector_init = x86_init_noop; + legacy_pic = &null_legacy_pic; + } +} + /* * If your system is blacklisted here, but you find that acpi=force * works for you, please contact linux-acpi@vger.kernel.org @@ -1536,6 +1556,11 @@ int __init early_acpi_boot_init(void) */ early_acpi_process_madt(); + /* + * Hardware-reduced ACPI mode initialization: + */ + acpi_reduced_hw_init(); + return 0; } -- cgit v1.2.1 From 78146572b9cd20452da47951812f35b1ad4906be Mon Sep 17 00:00:00 2001 From: Ian Wilson Date: Thu, 12 Mar 2015 09:37:58 +0000 Subject: netfilter: Zero the tuple in nfnl_cthelper_parse_tuple() nfnl_cthelper_parse_tuple() is called from nfnl_cthelper_new(), nfnl_cthelper_get() and nfnl_cthelper_del(). In each case they pass a pointer to an nf_conntrack_tuple data structure local variable: struct nf_conntrack_tuple tuple; ... ret = nfnl_cthelper_parse_tuple(&tuple, tb[NFCTH_TUPLE]); The problem is that this local variable is not initialized, and nfnl_cthelper_parse_tuple() only initializes two fields: src.l3num and dst.protonum. This leaves all other fields with undefined values based on whatever is on the stack: tuple->src.l3num = ntohs(nla_get_be16(tb[NFCTH_TUPLE_L3PROTONUM])); tuple->dst.protonum = nla_get_u8(tb[NFCTH_TUPLE_L4PROTONUM]); The symptom observed was that when the rpc and tns helpers were added then traffic to port 1536 was being sent to user-space. Signed-off-by: Ian Wilson Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nfnetlink_cthelper.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/netfilter/nfnetlink_cthelper.c b/net/netfilter/nfnetlink_cthelper.c index a5599fc51a6f..54330fb5efaf 100644 --- a/net/netfilter/nfnetlink_cthelper.c +++ b/net/netfilter/nfnetlink_cthelper.c @@ -77,6 +77,9 @@ nfnl_cthelper_parse_tuple(struct nf_conntrack_tuple *tuple, if (!tb[NFCTH_TUPLE_L3PROTONUM] || !tb[NFCTH_TUPLE_L4PROTONUM]) return -EINVAL; + /* Not all fields are initialized so first zero the tuple */ + memset(tuple, 0, sizeof(struct nf_conntrack_tuple)); + tuple->src.l3num = ntohs(nla_get_be16(tb[NFCTH_TUPLE_L3PROTONUM])); tuple->dst.protonum = nla_get_u8(tb[NFCTH_TUPLE_L4PROTONUM]); -- cgit v1.2.1 From c8a470cab030bae8f9e6e5cfff72b047b7c627a7 Mon Sep 17 00:00:00 2001 From: Daniel J Blueman Date: Thu, 12 Mar 2015 16:55:13 +0100 Subject: x86/apic/numachip: Fix sibling map with NumaChip On NumaChip systems, the physical processor ID assignment wasn't accounting for the number of nodes in AMD multi-module processors, giving an incorrect sibling map: $ cd /sys/devices/system/cpu/cpu29/topology $ grep . * core_id:5 core_siblings:00000000,ff000000 core_siblings_list:24-31 physical_package_id:3 thread_siblings:00000000,30000000 thread_siblings_list:28-29 This fixes it: $ cd /sys/devices/system/cpu/cpu29/topology $ grep . * core_id:5 core_siblings:00000000,ffff0000 core_siblings_list:16-31 physical_package_id:1 thread_siblings:00000000,30000000 thread_siblings_list:28-29 Signed-off-by: Daniel J Blueman Signed-off-by: Borislav Petkov Cc: Cc: H. Peter Anvin Cc: Steffen Persvold Cc: Thomas Gleixner Link: http://lkml.kernel.org/r/1426135950-10110-1-git-send-email-daniel@numascale.com Signed-off-by: Ingo Molnar --- arch/x86/kernel/apic/apic_numachip.c | 22 ++++++++++++++++------ 1 file changed, 16 insertions(+), 6 deletions(-) diff --git a/arch/x86/kernel/apic/apic_numachip.c b/arch/x86/kernel/apic/apic_numachip.c index c2fd21fed002..017149cded07 100644 --- a/arch/x86/kernel/apic/apic_numachip.c +++ b/arch/x86/kernel/apic/apic_numachip.c @@ -37,10 +37,12 @@ static const struct apic apic_numachip; static unsigned int get_apic_id(unsigned long x) { unsigned long value; - unsigned int id; + unsigned int id = (x >> 24) & 0xff; - rdmsrl(MSR_FAM10H_NODE_ID, value); - id = ((x >> 24) & 0xffU) | ((value << 2) & 0xff00U); + if (static_cpu_has_safe(X86_FEATURE_NODEID_MSR)) { + rdmsrl(MSR_FAM10H_NODE_ID, value); + id |= (value << 2) & 0xff00; + } return id; } @@ -155,10 +157,18 @@ static int __init numachip_probe(void) static void fixup_cpu_id(struct cpuinfo_x86 *c, int node) { - if (c->phys_proc_id != node) { - c->phys_proc_id = node; - per_cpu(cpu_llc_id, smp_processor_id()) = node; + u64 val; + u32 nodes = 1; + + this_cpu_write(cpu_llc_id, node); + + /* Account for nodes per socket in multi-core-module processors */ + if (static_cpu_has_safe(X86_FEATURE_NODEID_MSR)) { + rdmsrl(MSR_FAM10H_NODE_ID, val); + nodes = ((val >> 3) & 7) + 1; } + + c->phys_proc_id = node / nodes; } static int __init numachip_system_init(void) -- cgit v1.2.1 From ab3971b1e7d72270a2a259a29c1a40351b889740 Mon Sep 17 00:00:00 2001 From: Jason Wang Date: Thu, 12 Mar 2015 13:57:44 +0800 Subject: virtio-net: correctly delete napi hash We don't delete napi from hash list during module exit. This will cause the following panic when doing module load and unload: BUG: unable to handle kernel paging request at 0000004e00000075 IP: [] napi_hash_add+0x6b/0xf0 PGD 3c5d5067 PUD 0 Oops: 0000 [#1] SMP ... Call Trace: [] init_vqs+0x107/0x490 [virtio_net] [] virtnet_probe+0x562/0x791815639d880be [virtio_net] [] virtio_dev_probe+0x137/0x200 [] driver_probe_device+0x7a/0x250 [] __driver_attach+0x93/0xa0 [] ? __device_attach+0x40/0x40 [] bus_for_each_dev+0x63/0xa0 [] driver_attach+0x19/0x20 [] bus_add_driver+0x170/0x220 [] ? 0xffffffffa0a60000 [] driver_register+0x5f/0xf0 [] register_virtio_driver+0x1b/0x30 [] virtio_net_driver_init+0x10/0x12 [virtio_net] This patch fixes this by doing this in virtnet_free_queues(). And also don't delete napi in virtnet_freeze() since it will call virtnet_free_queues() which has already did this. Fixes 91815639d880 ("virtio-net: rx busy polling support") Cc: Rusty Russell Cc: Michael S. Tsirkin Signed-off-by: Jason Wang Acked-by: Michael S. Tsirkin Reviewed-by: Michael S. Tsirkin Signed-off-by: David S. Miller --- drivers/net/virtio_net.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) diff --git a/drivers/net/virtio_net.c b/drivers/net/virtio_net.c index f1ff3666f090..59b0e9754ae3 100644 --- a/drivers/net/virtio_net.c +++ b/drivers/net/virtio_net.c @@ -1448,8 +1448,10 @@ static void virtnet_free_queues(struct virtnet_info *vi) { int i; - for (i = 0; i < vi->max_queue_pairs; i++) + for (i = 0; i < vi->max_queue_pairs; i++) { + napi_hash_del(&vi->rq[i].napi); netif_napi_del(&vi->rq[i].napi); + } kfree(vi->rq); kfree(vi->sq); @@ -1948,11 +1950,8 @@ static int virtnet_freeze(struct virtio_device *vdev) cancel_delayed_work_sync(&vi->refill); if (netif_running(vi->dev)) { - for (i = 0; i < vi->max_queue_pairs; i++) { + for (i = 0; i < vi->max_queue_pairs; i++) napi_disable(&vi->rq[i].napi); - napi_hash_del(&vi->rq[i].napi); - netif_napi_del(&vi->rq[i].napi); - } } remove_vq_common(vi); -- cgit v1.2.1 From c1a6bff28cbff796bf6e7db5cf42ec9244911be2 Mon Sep 17 00:00:00 2001 From: Petr Matousek Date: Wed, 11 Mar 2015 12:16:09 +0100 Subject: kvm: x86: i8259: return initialized data on invalid-size read If data is read from PIC with invalid access size, the return data stays uninitialized even though success is returned. Fix this by always initializing the data. Signed-off-by: Petr Matousek Reported-by: Nadav Amit Signed-off-by: Marcelo Tosatti --- arch/x86/kvm/i8259.c | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/x86/kvm/i8259.c b/arch/x86/kvm/i8259.c index cc31f7c06d3d..9541ba34126b 100644 --- a/arch/x86/kvm/i8259.c +++ b/arch/x86/kvm/i8259.c @@ -507,6 +507,7 @@ static int picdev_read(struct kvm_pic *s, return -EOPNOTSUPP; if (len != 1) { + memset(val, 0, len); pr_pic_unimpl("non byte read\n"); return 0; } -- cgit v1.2.1 From 8051a2a518fcf3827a143470083ad6008697ff17 Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Thu, 12 Mar 2015 11:53:41 +1030 Subject: 9p/trans_virtio: fix hot-unplug On device hot-unplug, 9p/virtio currently will kfree channel while it might still be in use. Of course, it might stay used forever, so it's an extremely ugly hack, but it seems better than use-after-free that we have now. [ Unused variable removed, whitespace cleanup, msg single-lined --RR ] Signed-off-by: Michael S. Tsirkin Signed-off-by: Rusty Russell --- net/9p/trans_virtio.c | 24 ++++++++++++++++++++---- 1 file changed, 20 insertions(+), 4 deletions(-) diff --git a/net/9p/trans_virtio.c b/net/9p/trans_virtio.c index d8e376a5f0f1..36a1a739ad68 100644 --- a/net/9p/trans_virtio.c +++ b/net/9p/trans_virtio.c @@ -658,14 +658,30 @@ p9_virtio_create(struct p9_client *client, const char *devname, char *args) static void p9_virtio_remove(struct virtio_device *vdev) { struct virtio_chan *chan = vdev->priv; - - if (chan->inuse) - p9_virtio_close(chan->client); - vdev->config->del_vqs(vdev); + unsigned long warning_time; mutex_lock(&virtio_9p_lock); + + /* Remove self from list so we don't get new users. */ list_del(&chan->chan_list); + warning_time = jiffies; + + /* Wait for existing users to close. */ + while (chan->inuse) { + mutex_unlock(&virtio_9p_lock); + msleep(250); + if (time_after(jiffies, warning_time + 10 * HZ)) { + dev_emerg(&vdev->dev, + "p9_virtio_remove: waiting for device in use.\n"); + warning_time = jiffies; + } + mutex_lock(&virtio_9p_lock); + } + mutex_unlock(&virtio_9p_lock); + + vdev->config->del_vqs(vdev); + sysfs_remove_file(&(vdev->dev.kobj), &dev_attr_mount_tag.attr); kobject_uevent(&(vdev->dev.kobj), KOBJ_CHANGE); kfree(chan->tag); -- cgit v1.2.1 From 71e4b8bf0482fc7d70e9d4c10b13c207a285d58a Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Thu, 12 Mar 2015 11:54:41 +1030 Subject: virtio_rpmsg: set DRIVER_OK before using device virtio spec requires that all drivers set DRIVER_OK before using devices. While rpmsg isn't yet included in the virtio 1 spec, previous spec versions also required this. virtio rpmsg violates this rule: is calls kick before setting DRIVER_OK. The fix isn't trivial since simply calling virtio_device_ready earlier would mean we might get an interrupt in parallel with adding buffers. Instead, split kick out to prepare+notify calls. prepare before virtio_device_ready - when we know we won't get interrupts. notify right afterwards. Signed-off-by: Michael S. Tsirkin Acked-by: Ohad Ben-Cohen Signed-off-by: Rusty Russell --- drivers/rpmsg/virtio_rpmsg_bus.c | 17 ++++++++++++++++- 1 file changed, 16 insertions(+), 1 deletion(-) diff --git a/drivers/rpmsg/virtio_rpmsg_bus.c b/drivers/rpmsg/virtio_rpmsg_bus.c index 92f6af6da699..73354ee27877 100644 --- a/drivers/rpmsg/virtio_rpmsg_bus.c +++ b/drivers/rpmsg/virtio_rpmsg_bus.c @@ -951,6 +951,7 @@ static int rpmsg_probe(struct virtio_device *vdev) void *bufs_va; int err = 0, i; size_t total_buf_space; + bool notify; vrp = kzalloc(sizeof(*vrp), GFP_KERNEL); if (!vrp) @@ -1030,8 +1031,22 @@ static int rpmsg_probe(struct virtio_device *vdev) } } + /* + * Prepare to kick but don't notify yet - we can't do this before + * device is ready. + */ + notify = virtqueue_kick_prepare(vrp->rvq); + + /* From this point on, we can notify and get callbacks. */ + virtio_device_ready(vdev); + /* tell the remote processor it can start sending messages */ - virtqueue_kick(vrp->rvq); + /* + * this might be concurrent with callbacks, but we are only + * doing notify, not a full kick here, so that's ok. + */ + if (notify) + virtqueue_notify(vrp->rvq); dev_info(&vdev->dev, "rpmsg host is online\n"); -- cgit v1.2.1 From 87e7bf1450c9f6bd0927f63ebc0fe2d12e8bc83d Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Thu, 12 Mar 2015 12:56:43 +1030 Subject: virtio_mmio: generation support virtio_mmio currently lacks generation support which makes multi-byte field access racy. Fix by getting the value at offset 0xfc for version 2 devices. Nothing we can do for version 1, so return generation id 0. Signed-off-by: Michael S. Tsirkin Signed-off-by: Rusty Russell --- drivers/virtio/virtio_mmio.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/drivers/virtio/virtio_mmio.c b/drivers/virtio/virtio_mmio.c index cad569890908..9c877d2375a5 100644 --- a/drivers/virtio/virtio_mmio.c +++ b/drivers/virtio/virtio_mmio.c @@ -174,6 +174,16 @@ static void vm_set(struct virtio_device *vdev, unsigned offset, writeb(ptr[i], vm_dev->base + VIRTIO_MMIO_CONFIG + offset + i); } +static u32 vm_generation(struct virtio_device *vdev) +{ + struct virtio_mmio_device *vm_dev = to_virtio_mmio_device(vdev); + + if (vm_dev->version == 1) + return 0; + else + return readl(vm_dev->base + VIRTIO_MMIO_CONFIG_GENERATION); +} + static u8 vm_get_status(struct virtio_device *vdev) { struct virtio_mmio_device *vm_dev = to_virtio_mmio_device(vdev); @@ -440,6 +450,7 @@ static const char *vm_bus_name(struct virtio_device *vdev) static const struct virtio_config_ops virtio_mmio_config_ops = { .get = vm_get, .set = vm_set, + .generation = vm_generation, .get_status = vm_get_status, .set_status = vm_set_status, .reset = vm_reset, -- cgit v1.2.1 From a4994b810d52ccb26de922c8d231fe05d14610d4 Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Fri, 13 Mar 2015 11:59:11 +1030 Subject: uapi/virtio_scsi: allow overriding CDB/SENSE size QEMU wants to use virtio scsi structures with a different VIRTIO_SCSI_CDB_SIZE/VIRTIO_SCSI_SENSE_SIZE, let's add ifdefs to allow overriding them. Keep the old defines under new names: VIRTIO_SCSI_CDB_DEFAULT_SIZE/VIRTIO_SCSI_SENSE_DEFAULT_SIZE, since that's what these values really are: defaults for cdb/sense size fields. Suggested-by: Paolo Bonzini Signed-off-by: Michael S. Tsirkin Signed-off-by: Rusty Russell --- include/uapi/linux/virtio_scsi.h | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) diff --git a/include/uapi/linux/virtio_scsi.h b/include/uapi/linux/virtio_scsi.h index 42b9370771b0..cc18ef8825c0 100644 --- a/include/uapi/linux/virtio_scsi.h +++ b/include/uapi/linux/virtio_scsi.h @@ -29,8 +29,16 @@ #include -#define VIRTIO_SCSI_CDB_SIZE 32 -#define VIRTIO_SCSI_SENSE_SIZE 96 +/* Default values of the CDB and sense data size configuration fields */ +#define VIRTIO_SCSI_CDB_DEFAULT_SIZE 32 +#define VIRTIO_SCSI_SENSE_DEFAULT_SIZE 96 + +#ifndef VIRTIO_SCSI_CDB_SIZE +#define VIRTIO_SCSI_CDB_SIZE VIRTIO_SCSI_CDB_DEFAULT_SIZE +#endif +#ifndef VIRTIO_SCSI_SENSE_SIZE +#define VIRTIO_SCSI_SENSE_SIZE VIRTIO_SCSI_SENSE_DEFAULT_SIZE +#endif /* SCSI command request, followed by data-out */ struct virtio_scsi_cmd_req { -- cgit v1.2.1 From d415a7f1c1a8406b22d95b943c66a5b73a37bc19 Mon Sep 17 00:00:00 2001 From: Leon Yu Date: Thu, 26 Feb 2015 20:43:33 +0800 Subject: perf: Fix context leak in put_event() Commit: a83fe28e2e45 ("perf: Fix put_event() ctx lock") changed the locking logic in put_event() by replacing mutex_lock_nested() with perf_event_ctx_lock_nested(), but didn't fix the subsequent mutex_unlock() with a correct counterpart, perf_event_ctx_unlock(). Contexts are thus leaked as a result of incremented refcount in perf_event_ctx_lock_nested(). Signed-off-by: Leon Yu Cc: Arnaldo Carvalho de Melo Cc: Paul Mackerras Cc: Peter Zijlstra Fixes: a83fe28e2e45 ("perf: Fix put_event() ctx lock") Link: http://lkml.kernel.org/r/1424954613-5034-1-git-send-email-chianglungyu@gmail.com Signed-off-by: Ingo Molnar --- kernel/events/core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/events/core.c b/kernel/events/core.c index f04daabfd1cf..453ef61311d4 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -3591,7 +3591,7 @@ static void put_event(struct perf_event *event) ctx = perf_event_ctx_lock_nested(event, SINGLE_DEPTH_NESTING); WARN_ON_ONCE(ctx->parent_ctx); perf_remove_from_context(event, true); - mutex_unlock(&ctx->mutex); + perf_event_ctx_unlock(event, ctx); _free_event(event); } -- cgit v1.2.1 From bdf6c79278b3fb6caf1811ae877078c5f424bcb1 Mon Sep 17 00:00:00 2001 From: Torsten Fleischer Date: Mon, 23 Feb 2015 17:54:10 +0100 Subject: dmaengine: at_hdmac: Fix calculation of the residual bytes This patch fixes the following issues regarding to the calculation of the residue: 1. The residue is always calculated for the current transfer even if the cookie is associated to a pending transfer. 2. For scatter/gather DMA the calculation of the residue for the current transfer doesn't include the bytes of the child descriptors that are already transferred. It only calculates the difference between the transfer's total length minus the number of bytes that are already transferred for the current child descriptor. For example: There is a scatter/gather DMA transfer with a total length of 1 MByte. Getting the residue several times while the transfer is running shows something like that: 1: residue = 975584 2: residue = 1002766 3: residue = 992627 4: residue = 983767 5: residue = 985694 6: residue = 1008094 7: residue = 1009741 8: residue = 1011195 3. The driver stores the residue but never resets it when starting a new transfer. For example: If there are two subsequent DMA transfers. The first one with a total length of 1 MByte and the second one with a total length of 1 kByte. Getting the residue for both transfers shows something like that: transfer 1: residue = 975584 transfer 2: residue = 1048380 Changes from V1: * Fixed coding style of the multi-line comments. * Improved accuracy of the residue calculation when the transfer for the first descriptor is active. Changes from V2: * Member 'tx_width' of 'struct at_desc' restored, because the transfer width can't be derived from the source width when using "slave_sg". The transfer width is needed for the calculation of the residue if either the transfer of the first or the last descriptor is in progress. In the case of a "memory_to_memory_sg" transfer (part of this patch series) the transfer width of both descriptors may differ. Thus it is required to additionally set 'tx_width' of the last descriptor. * Added functions for multiply used calculations. Signed-off-by: Torsten Fleischer Acked-by: Nicolas Ferre Signed-off-by: Vinod Koul --- drivers/dma/at_hdmac.c | 184 +++++++++++++++++++++++++++----------------- drivers/dma/at_hdmac_regs.h | 7 +- 2 files changed, 115 insertions(+), 76 deletions(-) diff --git a/drivers/dma/at_hdmac.c b/drivers/dma/at_hdmac.c index 1e1a4c567542..0b4fc6fb48ce 100644 --- a/drivers/dma/at_hdmac.c +++ b/drivers/dma/at_hdmac.c @@ -238,93 +238,126 @@ static void atc_dostart(struct at_dma_chan *atchan, struct at_desc *first) } /* - * atc_get_current_descriptors - - * locate the descriptor which equal to physical address in DSCR - * @atchan: the channel we want to start - * @dscr_addr: physical descriptor address in DSCR + * atc_get_desc_by_cookie - get the descriptor of a cookie + * @atchan: the DMA channel + * @cookie: the cookie to get the descriptor for */ -static struct at_desc *atc_get_current_descriptors(struct at_dma_chan *atchan, - u32 dscr_addr) +static struct at_desc *atc_get_desc_by_cookie(struct at_dma_chan *atchan, + dma_cookie_t cookie) { - struct at_desc *desc, *_desc, *child, *desc_cur = NULL; + struct at_desc *desc, *_desc; - list_for_each_entry_safe(desc, _desc, &atchan->active_list, desc_node) { - if (desc->lli.dscr == dscr_addr) { - desc_cur = desc; - break; - } + list_for_each_entry_safe(desc, _desc, &atchan->queue, desc_node) { + if (desc->txd.cookie == cookie) + return desc; + } - list_for_each_entry(child, &desc->tx_list, desc_node) { - if (child->lli.dscr == dscr_addr) { - desc_cur = child; - break; - } - } + list_for_each_entry_safe(desc, _desc, &atchan->active_list, desc_node) { + if (desc->txd.cookie == cookie) + return desc; } - return desc_cur; + return NULL; } -/* - * atc_get_bytes_left - - * Get the number of bytes residue in dma buffer, - * @chan: the channel we want to start +/** + * atc_calc_bytes_left - calculates the number of bytes left according to the + * value read from CTRLA. + * + * @current_len: the number of bytes left before reading CTRLA + * @ctrla: the value of CTRLA + * @desc: the descriptor containing the transfer width + */ +static inline int atc_calc_bytes_left(int current_len, u32 ctrla, + struct at_desc *desc) +{ + return current_len - ((ctrla & ATC_BTSIZE_MAX) << desc->tx_width); +} + +/** + * atc_calc_bytes_left_from_reg - calculates the number of bytes left according + * to the current value of CTRLA. + * + * @current_len: the number of bytes left before reading CTRLA + * @atchan: the channel to read CTRLA for + * @desc: the descriptor containing the transfer width + */ +static inline int atc_calc_bytes_left_from_reg(int current_len, + struct at_dma_chan *atchan, struct at_desc *desc) +{ + u32 ctrla = channel_readl(atchan, CTRLA); + + return atc_calc_bytes_left(current_len, ctrla, desc); +} + +/** + * atc_get_bytes_left - get the number of bytes residue for a cookie + * @chan: DMA channel + * @cookie: transaction identifier to check status of */ -static int atc_get_bytes_left(struct dma_chan *chan) +static int atc_get_bytes_left(struct dma_chan *chan, dma_cookie_t cookie) { struct at_dma_chan *atchan = to_at_dma_chan(chan); - struct at_dma *atdma = to_at_dma(chan->device); - int chan_id = atchan->chan_common.chan_id; struct at_desc *desc_first = atc_first_active(atchan); - struct at_desc *desc_cur; - int ret = 0, count = 0; + struct at_desc *desc; + int ret; + u32 ctrla, dscr; /* - * Initialize necessary values in the first time. - * remain_desc record remain desc length. + * If the cookie doesn't match to the currently running transfer then + * we can return the total length of the associated DMA transfer, + * because it is still queued. */ - if (atchan->remain_desc == 0) - /* First descriptor embedds the transaction length */ - atchan->remain_desc = desc_first->len; + desc = atc_get_desc_by_cookie(atchan, cookie); + if (desc == NULL) + return -EINVAL; + else if (desc != desc_first) + return desc->total_len; - /* - * This happens when current descriptor transfer complete. - * The residual buffer size should reduce current descriptor length. - */ - if (unlikely(test_bit(ATC_IS_BTC, &atchan->status))) { - clear_bit(ATC_IS_BTC, &atchan->status); - desc_cur = atc_get_current_descriptors(atchan, - channel_readl(atchan, DSCR)); - if (!desc_cur) { - ret = -EINVAL; - goto out; - } + /* cookie matches to the currently running transfer */ + ret = desc_first->total_len; - count = (desc_cur->lli.ctrla & ATC_BTSIZE_MAX) - << desc_first->tx_width; - if (atchan->remain_desc < count) { - ret = -EINVAL; - goto out; + if (desc_first->lli.dscr) { + /* hardware linked list transfer */ + + /* + * Calculate the residue by removing the length of the child + * descriptors already transferred from the total length. + * To get the current child descriptor we can use the value of + * the channel's DSCR register and compare it against the value + * of the hardware linked list structure of each child + * descriptor. + */ + + ctrla = channel_readl(atchan, CTRLA); + rmb(); /* ensure CTRLA is read before DSCR */ + dscr = channel_readl(atchan, DSCR); + + /* for the first descriptor we can be more accurate */ + if (desc_first->lli.dscr == dscr) + return atc_calc_bytes_left(ret, ctrla, desc_first); + + ret -= desc_first->len; + list_for_each_entry(desc, &desc_first->tx_list, desc_node) { + if (desc->lli.dscr == dscr) + break; + + ret -= desc->len; } - atchan->remain_desc -= count; - ret = atchan->remain_desc; - } else { /* - * Get residual bytes when current - * descriptor transfer in progress. + * For the last descriptor in the chain we can calculate + * the remaining bytes using the channel's register. + * Note that the transfer width of the first and last + * descriptor may differ. */ - count = (channel_readl(atchan, CTRLA) & ATC_BTSIZE_MAX) - << (desc_first->tx_width); - ret = atchan->remain_desc - count; + if (!desc->lli.dscr) + ret = atc_calc_bytes_left_from_reg(ret, atchan, desc); + } else { + /* single transfer */ + ret = atc_calc_bytes_left_from_reg(ret, atchan, desc_first); } - /* - * Check fifo empty. - */ - if (!(dma_readl(atdma, CHSR) & AT_DMA_EMPT(chan_id))) - atc_issue_pending(chan); -out: return ret; } @@ -539,8 +572,6 @@ static irqreturn_t at_dma_interrupt(int irq, void *dev_id) /* Give information to tasklet */ set_bit(ATC_IS_ERROR, &atchan->status); } - if (pending & AT_DMA_BTC(i)) - set_bit(ATC_IS_BTC, &atchan->status); tasklet_schedule(&atchan->tasklet); ret = IRQ_HANDLED; } @@ -653,14 +684,18 @@ atc_prep_dma_memcpy(struct dma_chan *chan, dma_addr_t dest, dma_addr_t src, desc->lli.ctrlb = ctrlb; desc->txd.cookie = 0; + desc->len = xfer_count << src_width; atc_desc_chain(&first, &prev, desc); } /* First descriptor of the chain embedds additional information */ first->txd.cookie = -EBUSY; - first->len = len; + first->total_len = len; + + /* set transfer width for the calculation of the residue */ first->tx_width = src_width; + prev->tx_width = src_width; /* set end-of-link to the last link descriptor of list*/ set_desc_eol(desc); @@ -752,6 +787,7 @@ atc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, | ATC_SRC_WIDTH(mem_width) | len >> mem_width; desc->lli.ctrlb = ctrlb; + desc->len = len; atc_desc_chain(&first, &prev, desc); total_len += len; @@ -792,6 +828,7 @@ atc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, | ATC_DST_WIDTH(mem_width) | len >> reg_width; desc->lli.ctrlb = ctrlb; + desc->len = len; atc_desc_chain(&first, &prev, desc); total_len += len; @@ -806,8 +843,11 @@ atc_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl, /* First descriptor of the chain embedds additional information */ first->txd.cookie = -EBUSY; - first->len = total_len; + first->total_len = total_len; + + /* set transfer width for the calculation of the residue */ first->tx_width = reg_width; + prev->tx_width = reg_width; /* first link descriptor of list is responsible of flags */ first->txd.flags = flags; /* client is in control of this ack */ @@ -872,6 +912,7 @@ atc_dma_cyclic_fill_desc(struct dma_chan *chan, struct at_desc *desc, | ATC_FC_MEM2PER | ATC_SIF(atchan->mem_if) | ATC_DIF(atchan->per_if); + desc->len = period_len; break; case DMA_DEV_TO_MEM: @@ -883,6 +924,7 @@ atc_dma_cyclic_fill_desc(struct dma_chan *chan, struct at_desc *desc, | ATC_FC_PER2MEM | ATC_SIF(atchan->per_if) | ATC_DIF(atchan->mem_if); + desc->len = period_len; break; default: @@ -964,7 +1006,7 @@ atc_prep_dma_cyclic(struct dma_chan *chan, dma_addr_t buf_addr, size_t buf_len, /* First descriptor of the chain embedds additional information */ first->txd.cookie = -EBUSY; - first->len = buf_len; + first->total_len = buf_len; first->tx_width = reg_width; return &first->txd; @@ -1118,7 +1160,7 @@ atc_tx_status(struct dma_chan *chan, spin_lock_irqsave(&atchan->lock, flags); /* Get number of bytes left in the active transactions */ - bytes = atc_get_bytes_left(chan); + bytes = atc_get_bytes_left(chan, cookie); spin_unlock_irqrestore(&atchan->lock, flags); @@ -1214,7 +1256,6 @@ static int atc_alloc_chan_resources(struct dma_chan *chan) spin_lock_irqsave(&atchan->lock, flags); atchan->descs_allocated = i; - atchan->remain_desc = 0; list_splice(&tmp_list, &atchan->free_list); dma_cookie_init(chan); spin_unlock_irqrestore(&atchan->lock, flags); @@ -1257,7 +1298,6 @@ static void atc_free_chan_resources(struct dma_chan *chan) list_splice_init(&atchan->free_list, &list); atchan->descs_allocated = 0; atchan->status = 0; - atchan->remain_desc = 0; dev_vdbg(chan2dev(chan), "free_chan_resources: done\n"); } diff --git a/drivers/dma/at_hdmac_regs.h b/drivers/dma/at_hdmac_regs.h index d6bba6c636c2..2727ca560572 100644 --- a/drivers/dma/at_hdmac_regs.h +++ b/drivers/dma/at_hdmac_regs.h @@ -181,8 +181,9 @@ struct at_lli { * @at_lli: hardware lli structure * @txd: support for the async_tx api * @desc_node: node on the channed descriptors list - * @len: total transaction bytecount + * @len: descriptor byte count * @tx_width: transfer width + * @total_len: total transaction byte count */ struct at_desc { /* FIRST values the hardware uses */ @@ -194,6 +195,7 @@ struct at_desc { struct list_head desc_node; size_t len; u32 tx_width; + size_t total_len; }; static inline struct at_desc * @@ -213,7 +215,6 @@ txd_to_at_desc(struct dma_async_tx_descriptor *txd) enum atc_status { ATC_IS_ERROR = 0, ATC_IS_PAUSED = 1, - ATC_IS_BTC = 2, ATC_IS_CYCLIC = 24, }; @@ -231,7 +232,6 @@ enum atc_status { * @save_cfg: configuration register that is saved on suspend/resume cycle * @save_dscr: for cyclic operations, preserve next descriptor address in * the cyclic list on suspend/resume cycle - * @remain_desc: to save remain desc length * @dma_sconfig: configuration for slave transfers, passed via * .device_config * @lock: serializes enqueue/dequeue operations to descriptors lists @@ -251,7 +251,6 @@ struct at_dma_chan { struct tasklet_struct tasklet; u32 save_cfg; u32 save_dscr; - u32 remain_desc; struct dma_slave_config dma_sconfig; spinlock_t lock; -- cgit v1.2.1 From ccfe8c3f7e52ae83155cb038753f4c75b774ca8a Mon Sep 17 00:00:00 2001 From: Stephan Mueller Date: Thu, 12 Mar 2015 09:17:51 +0100 Subject: crypto: aesni - fix memory usage in GCM decryption The kernel crypto API logic requires the caller to provide the length of (ciphertext || authentication tag) as cryptlen for the AEAD decryption operation. Thus, the cipher implementation must calculate the size of the plaintext output itself and cannot simply use cryptlen. The RFC4106 GCM decryption operation tries to overwrite cryptlen memory in req->dst. As the destination buffer for decryption only needs to hold the plaintext memory but cryptlen references the input buffer holding (ciphertext || authentication tag), the assumption of the destination buffer length in RFC4106 GCM operation leads to a too large size. This patch simply uses the already calculated plaintext size. In addition, this patch fixes the offset calculation of the AAD buffer pointer: as mentioned before, cryptlen already includes the size of the tag. Thus, the tag does not need to be added. With the addition, the AAD will be written beyond the already allocated buffer. Note, this fixes a kernel crash that can be triggered from user space via AF_ALG(aead) -- simply use the libkcapi test application from [1] and update it to use rfc4106-gcm-aes. Using [1], the changes were tested using CAVS vectors to demonstrate that the crypto operation still delivers the right results. [1] http://www.chronox.de/libkcapi.html CC: Tadeusz Struk Cc: stable@vger.kernel.org Signed-off-by: Stephan Mueller Signed-off-by: Herbert Xu --- arch/x86/crypto/aesni-intel_glue.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/x86/crypto/aesni-intel_glue.c b/arch/x86/crypto/aesni-intel_glue.c index 947c6bf52c33..54f60ab41c63 100644 --- a/arch/x86/crypto/aesni-intel_glue.c +++ b/arch/x86/crypto/aesni-intel_glue.c @@ -1155,7 +1155,7 @@ static int __driver_rfc4106_decrypt(struct aead_request *req) src = kmalloc(req->cryptlen + req->assoclen, GFP_ATOMIC); if (!src) return -ENOMEM; - assoc = (src + req->cryptlen + auth_tag_len); + assoc = (src + req->cryptlen); scatterwalk_map_and_copy(src, req->src, 0, req->cryptlen, 0); scatterwalk_map_and_copy(assoc, req->assoc, 0, req->assoclen, 0); @@ -1180,7 +1180,7 @@ static int __driver_rfc4106_decrypt(struct aead_request *req) scatterwalk_done(&src_sg_walk, 0, 0); scatterwalk_done(&assoc_sg_walk, 0, 0); } else { - scatterwalk_map_and_copy(dst, req->dst, 0, req->cryptlen, 1); + scatterwalk_map_and_copy(dst, req->dst, 0, tempCipherLen, 1); kfree(src); } return retval; -- cgit v1.2.1 From b52104e509479c4709eb9d81642df77c5ef2716b Mon Sep 17 00:00:00 2001 From: Wei Yongjun Date: Fri, 27 Feb 2015 19:41:45 +0800 Subject: arm/arm64: KVM: fix missing unlock on error in kvm_vgic_create() Add the missing unlock before return from function kvm_vgic_create() in the error handling case. Signed-off-by: Wei Yongjun Signed-off-by: Christoffer Dall --- virt/kvm/arm/vgic.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/virt/kvm/arm/vgic.c b/virt/kvm/arm/vgic.c index 0cc6ab6005a0..4b2c2e7856a3 100644 --- a/virt/kvm/arm/vgic.c +++ b/virt/kvm/arm/vgic.c @@ -1583,8 +1583,10 @@ int kvm_vgic_create(struct kvm *kvm, u32 type) * emulation. So check this here again. KVM_CREATE_DEVICE does * the proper checks already. */ - if (type == KVM_DEV_TYPE_ARM_VGIC_V2 && !vgic->can_emulate_gicv2) - return -ENODEV; + if (type == KVM_DEV_TYPE_ARM_VGIC_V2 && !vgic->can_emulate_gicv2) { + ret = -ENODEV; + goto out; + } /* * Any time a vcpu is run, vcpu_load is called which tries to grab the -- cgit v1.2.1 From d8bdff59cea141d2e5f7e98c1b11d3e0271640bd Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Fri, 13 Mar 2015 10:52:14 +1100 Subject: netfilter: Fix potential crash in nft_hash walker When we get back an EAGAIN from rhashtable_walk_next we were treating it as a valid object which obviously doesn't work too well. Luckily this is hard to trigger so it seems nobody has run into it yet. This patch fixes it by redoing the next call when we get an EAGAIN. Signed-off-by: Herbert Xu Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nft_hash.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/net/netfilter/nft_hash.c b/net/netfilter/nft_hash.c index c82df0a48fcd..37c15e674884 100644 --- a/net/netfilter/nft_hash.c +++ b/net/netfilter/nft_hash.c @@ -153,6 +153,8 @@ static void nft_hash_walk(const struct nft_ctx *ctx, const struct nft_set *set, iter->err = err; goto out; } + + continue; } if (iter->count < iter->skip) -- cgit v1.2.1 From 0a64815091bd0ad6c6cdfaac2fae55b0f3ecf974 Mon Sep 17 00:00:00 2001 From: Hendrik Brueckner Date: Mon, 9 Mar 2015 17:34:18 +0100 Subject: s390/cpum_sf: add diagnostic sampling event only if it is authorized The SF_CYCLES_BASIC_DIAG is always registered even if it is turned of in the current hardware configuration. Because diagnostic-sampling is typically not turned on in the hardware configuration, do not register this perf event by default. Enable it only if the diagnostic-sampling function is authorized. Signed-off-by: Hendrik Brueckner Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/perf_cpum_sf.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/arch/s390/kernel/perf_cpum_sf.c b/arch/s390/kernel/perf_cpum_sf.c index c3f8d157cb0d..e6a1578fc000 100644 --- a/arch/s390/kernel/perf_cpum_sf.c +++ b/arch/s390/kernel/perf_cpum_sf.c @@ -1415,7 +1415,7 @@ CPUMF_EVENT_ATTR(SF, SF_CYCLES_BASIC_DIAG, PERF_EVENT_CPUM_SF_DIAG); static struct attribute *cpumsf_pmu_events_attr[] = { CPUMF_EVENT_PTR(SF, SF_CYCLES_BASIC), - CPUMF_EVENT_PTR(SF, SF_CYCLES_BASIC_DIAG), + NULL, NULL, }; @@ -1606,8 +1606,11 @@ static int __init init_cpum_sampling_pmu(void) return -EINVAL; } - if (si.ad) + if (si.ad) { sfb_set_limits(CPUM_SF_MIN_SDB, CPUM_SF_MAX_SDB); + cpumsf_pmu_events_attr[1] = + CPUMF_EVENT_PTR(SF, SF_CYCLES_BASIC_DIAG); + } sfdbg = debug_register(KMSG_COMPONENT, 2, 1, 80); if (!sfdbg) -- cgit v1.2.1 From 20e76ee184a7cea155268377c4e414eea1dab6fd Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Fri, 13 Mar 2015 10:31:42 +0100 Subject: s390/ftrace: fix compile error if CONFIG_KPROBES is disabled Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/ftrace.c | 61 ++++++++++++++++++++++++++++++++++------------- 1 file changed, 45 insertions(+), 16 deletions(-) diff --git a/arch/s390/kernel/ftrace.c b/arch/s390/kernel/ftrace.c index 82c19899574f..6c79f1b44fe7 100644 --- a/arch/s390/kernel/ftrace.c +++ b/arch/s390/kernel/ftrace.c @@ -57,6 +57,44 @@ unsigned long ftrace_plt; +static inline void ftrace_generate_orig_insn(struct ftrace_insn *insn) +{ +#ifdef CC_USING_HOTPATCH + /* brcl 0,0 */ + insn->opc = 0xc004; + insn->disp = 0; +#else + /* stg r14,8(r15) */ + insn->opc = 0xe3e0; + insn->disp = 0xf0080024; +#endif +} + +static inline int is_kprobe_on_ftrace(struct ftrace_insn *insn) +{ +#ifdef CONFIG_KPROBES + if (insn->opc == BREAKPOINT_INSTRUCTION) + return 1; +#endif + return 0; +} + +static inline void ftrace_generate_kprobe_nop_insn(struct ftrace_insn *insn) +{ +#ifdef CONFIG_KPROBES + insn->opc = BREAKPOINT_INSTRUCTION; + insn->disp = KPROBE_ON_FTRACE_NOP; +#endif +} + +static inline void ftrace_generate_kprobe_call_insn(struct ftrace_insn *insn) +{ +#ifdef CONFIG_KPROBES + insn->opc = BREAKPOINT_INSTRUCTION; + insn->disp = KPROBE_ON_FTRACE_CALL; +#endif +} + int ftrace_modify_call(struct dyn_ftrace *rec, unsigned long old_addr, unsigned long addr) { @@ -72,16 +110,9 @@ int ftrace_make_nop(struct module *mod, struct dyn_ftrace *rec, return -EFAULT; if (addr == MCOUNT_ADDR) { /* Initial code replacement */ -#ifdef CC_USING_HOTPATCH - /* We expect to see brcl 0,0 */ - ftrace_generate_nop_insn(&orig); -#else - /* We expect to see stg r14,8(r15) */ - orig.opc = 0xe3e0; - orig.disp = 0xf0080024; -#endif + ftrace_generate_orig_insn(&orig); ftrace_generate_nop_insn(&new); - } else if (old.opc == BREAKPOINT_INSTRUCTION) { + } else if (is_kprobe_on_ftrace(&old)) { /* * If we find a breakpoint instruction, a kprobe has been * placed at the beginning of the function. We write the @@ -89,9 +120,8 @@ int ftrace_make_nop(struct module *mod, struct dyn_ftrace *rec, * bytes of the original instruction so that the kprobes * handler can execute a nop, if it reaches this breakpoint. */ - new.opc = orig.opc = BREAKPOINT_INSTRUCTION; - orig.disp = KPROBE_ON_FTRACE_CALL; - new.disp = KPROBE_ON_FTRACE_NOP; + ftrace_generate_kprobe_call_insn(&orig); + ftrace_generate_kprobe_nop_insn(&new); } else { /* Replace ftrace call with a nop. */ ftrace_generate_call_insn(&orig, rec->ip); @@ -111,7 +141,7 @@ int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr) if (probe_kernel_read(&old, (void *) rec->ip, sizeof(old))) return -EFAULT; - if (old.opc == BREAKPOINT_INSTRUCTION) { + if (is_kprobe_on_ftrace(&old)) { /* * If we find a breakpoint instruction, a kprobe has been * placed at the beginning of the function. We write the @@ -119,9 +149,8 @@ int ftrace_make_call(struct dyn_ftrace *rec, unsigned long addr) * bytes of the original instruction so that the kprobes * handler can execute a brasl if it reaches this breakpoint. */ - new.opc = orig.opc = BREAKPOINT_INSTRUCTION; - orig.disp = KPROBE_ON_FTRACE_NOP; - new.disp = KPROBE_ON_FTRACE_CALL; + ftrace_generate_kprobe_nop_insn(&orig); + ftrace_generate_kprobe_call_insn(&new); } else { /* Replace nop with an ftrace call. */ ftrace_generate_nop_insn(&orig); -- cgit v1.2.1 From e143fa93c28669a7f0851e6850b1da5b1945fd53 Mon Sep 17 00:00:00 2001 From: Martin Schwidefsky Date: Fri, 13 Mar 2015 11:19:09 +0100 Subject: s390/mm: limit STACK_RND_MASK for compat tasks For compat tasks the mmap randomization does not use the maximum randomization value from mmap_rnd_mask but the fixed value of 0x7ff. This needs to be respected in the definition of STACK_RND_MASK as well. Signed-off-by: Martin Schwidefsky --- arch/s390/include/asm/elf.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/s390/include/asm/elf.h b/arch/s390/include/asm/elf.h index c9df40b5c0ac..c9c875d9ed31 100644 --- a/arch/s390/include/asm/elf.h +++ b/arch/s390/include/asm/elf.h @@ -211,7 +211,7 @@ do { \ extern unsigned long mmap_rnd_mask; -#define STACK_RND_MASK (mmap_rnd_mask) +#define STACK_RND_MASK (test_thread_flag(TIF_31BIT) ? 0x7ff : mmap_rnd_mask) #define ARCH_DLINFO \ do { \ -- cgit v1.2.1 From 2f1bce487cd0a02623cff3d877940f9a2026341c Mon Sep 17 00:00:00 2001 From: Thierry Reding Date: Wed, 25 Feb 2015 16:16:29 +0100 Subject: phy: Find the right match in devm_phy_destroy() devm_phy_create() stores the pointer to the new PHY at the address returned by devres_alloc(). The res parameter passed to devm_phy_match() is therefore the location where the pointer to the PHY is stored, hence it needs to be dereferenced before comparing to the match data in order to find the correct match. Cc: # v3.13+ Signed-off-by: Thierry Reding Signed-off-by: Kishon Vijay Abraham I --- drivers/phy/phy-core.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/drivers/phy/phy-core.c b/drivers/phy/phy-core.c index a12d35338313..04fc84f2b289 100644 --- a/drivers/phy/phy-core.c +++ b/drivers/phy/phy-core.c @@ -52,7 +52,9 @@ static void devm_phy_consume(struct device *dev, void *res) static int devm_phy_match(struct device *dev, void *res, void *match_data) { - return res == match_data; + struct phy **phy = res; + + return *phy == match_data; } /** -- cgit v1.2.1 From a7c80ebcac3068b1c3cb27d538d29558c30010c8 Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Fri, 13 Mar 2015 09:53:09 +0100 Subject: x86/fpu: Avoid math_state_restore() without used_math() in __restore_xstate_sig() math_state_restore() assumes it is called with irqs disabled, but this is not true if the caller is __restore_xstate_sig(). This means that if ia32_fxstate == T and __copy_from_user() fails, __restore_xstate_sig() returns with irqs disabled too. This triggers: BUG: sleeping function called from invalid context at kernel/locking/rwsem.c:41 dump_stack ___might_sleep ? _raw_spin_unlock_irqrestore __might_sleep down_read ? _raw_spin_unlock_irqrestore print_vma_addr signal_fault sys32_rt_sigreturn Change __restore_xstate_sig() to call set_used_math() unconditionally. This avoids enabling and disabling interrupts in math_state_restore(). If copy_from_user() fails, we can simply do fpu_finit() by hand. [ Note: this is only the first step. math_state_restore() should not check used_math(), it should set this flag. While init_fpu() should simply die. ] Signed-off-by: Oleg Nesterov Signed-off-by: Borislav Petkov Cc: Cc: Andy Lutomirski Cc: Borislav Petkov Cc: Dave Hansen Cc: Fenghua Yu Cc: H. Peter Anvin Cc: Linus Torvalds Cc: Pekka Riikonen Cc: Quentin Casasnovas Cc: Rik van Riel Cc: Suresh Siddha Cc: Thomas Gleixner Link: http://lkml.kernel.org/r/20150307153844.GB25954@redhat.com Signed-off-by: Ingo Molnar --- arch/x86/kernel/xsave.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/arch/x86/kernel/xsave.c b/arch/x86/kernel/xsave.c index 34f66e58a896..cdc6cf903078 100644 --- a/arch/x86/kernel/xsave.c +++ b/arch/x86/kernel/xsave.c @@ -379,7 +379,7 @@ int __restore_xstate_sig(void __user *buf, void __user *buf_fx, int size) * thread's fpu state, reconstruct fxstate from the fsave * header. Sanitize the copied state etc. */ - struct xsave_struct *xsave = &tsk->thread.fpu.state->xsave; + struct fpu *fpu = &tsk->thread.fpu; struct user_i387_ia32_struct env; int err = 0; @@ -393,14 +393,15 @@ int __restore_xstate_sig(void __user *buf, void __user *buf_fx, int size) */ drop_fpu(tsk); - if (__copy_from_user(xsave, buf_fx, state_size) || + if (__copy_from_user(&fpu->state->xsave, buf_fx, state_size) || __copy_from_user(&env, buf, sizeof(env))) { + fpu_finit(fpu); err = -1; } else { sanitize_restored_xstate(tsk, &env, xstate_bv, fx_only); - set_used_math(); } + set_used_math(); if (use_eager_fpu()) { preempt_disable(); math_state_restore(); -- cgit v1.2.1 From f4c3686386393c120710dd34df2a74183ab805fd Mon Sep 17 00:00:00 2001 From: Oleg Nesterov Date: Fri, 13 Mar 2015 09:53:10 +0100 Subject: x86/fpu: Drop_fpu() should not assume that tsk equals current drop_fpu() does clear_used_math() and usually this is correct because tsk == current. However switch_fpu_finish()->restore_fpu_checking() is called before __switch_to() updates the "current_task" variable. If it fails, we will wrongly clear the PF_USED_MATH flag of the previous task. So use clear_stopped_child_used_math() instead. Signed-off-by: Oleg Nesterov Signed-off-by: Borislav Petkov Reviewed-by: Rik van Riel Cc: Cc: Andy Lutomirski Cc: Borislav Petkov Cc: Dave Hansen Cc: Fenghua Yu Cc: H. Peter Anvin Cc: Linus Torvalds Cc: Pekka Riikonen Cc: Quentin Casasnovas Cc: Suresh Siddha Cc: Thomas Gleixner Link: http://lkml.kernel.org/r/20150309171041.GB11388@redhat.com Signed-off-by: Ingo Molnar --- arch/x86/include/asm/fpu-internal.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/x86/include/asm/fpu-internal.h b/arch/x86/include/asm/fpu-internal.h index 0dbc08282291..72ba21a8b5fc 100644 --- a/arch/x86/include/asm/fpu-internal.h +++ b/arch/x86/include/asm/fpu-internal.h @@ -370,7 +370,7 @@ static inline void drop_fpu(struct task_struct *tsk) preempt_disable(); tsk->thread.fpu_counter = 0; __drop_fpu(tsk); - clear_used_math(); + clear_stopped_child_used_math(tsk); preempt_enable(); } -- cgit v1.2.1 From ecd5fb026d460bf8fb883254fb9bd01a1085c185 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Fri, 27 Feb 2015 23:44:24 +0800 Subject: phy: exynos5-usbdrd: Fix off-by-one valid value checking for args->args[0] Current code uses args->args[0] as array subscript of phy_drd->phys[]. So the valid value range for args->args[0] is 0 ... EXYNOS5_DRDPHYS_NUM - 1. Signed-off-by: Axel Lin Reviewed by: Vivek Gautam Signed-off-by: Kishon Vijay Abraham I --- drivers/phy/phy-exynos5-usbdrd.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/phy/phy-exynos5-usbdrd.c b/drivers/phy/phy-exynos5-usbdrd.c index 04374018425f..e2a0be750ad9 100644 --- a/drivers/phy/phy-exynos5-usbdrd.c +++ b/drivers/phy/phy-exynos5-usbdrd.c @@ -531,7 +531,7 @@ static struct phy *exynos5_usbdrd_phy_xlate(struct device *dev, { struct exynos5_usbdrd_phy *phy_drd = dev_get_drvdata(dev); - if (WARN_ON(args->args[0] > EXYNOS5_DRDPHYS_NUM)) + if (WARN_ON(args->args[0] >= EXYNOS5_DRDPHYS_NUM)) return ERR_PTR(-ENODEV); return phy_drd->phys[args->args[0]].phy; -- cgit v1.2.1 From 8f27f167de5cd8260ad1ad3bb8166363de6f2620 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Sun, 1 Mar 2015 22:28:23 +0800 Subject: phy: twl4030-usb: Remove redundant assignment for twl->linkstat It's pointless to set twl->linkstat twice. Signed-off-by: Axel Lin Signed-off-by: Kishon Vijay Abraham I --- drivers/phy/phy-twl4030-usb.c | 1 - 1 file changed, 1 deletion(-) diff --git a/drivers/phy/phy-twl4030-usb.c b/drivers/phy/phy-twl4030-usb.c index 8e87f54671f3..bc42d6a8939f 100644 --- a/drivers/phy/phy-twl4030-usb.c +++ b/drivers/phy/phy-twl4030-usb.c @@ -666,7 +666,6 @@ static int twl4030_usb_probe(struct platform_device *pdev) twl->dev = &pdev->dev; twl->irq = platform_get_irq(pdev, 0); twl->vbus_supplied = false; - twl->linkstat = -EINVAL; twl->linkstat = OMAP_MUSB_UNKNOWN; twl->phy.dev = twl->dev; -- cgit v1.2.1 From d8d52948a0240d7d2d217d15a673364d990bf9e8 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Thu, 5 Mar 2015 09:52:59 +0800 Subject: phy: miphy365x: Convert to devm_kcalloc and fix wrong sizeof Prefer devm_kcalloc over devm_kzalloc with multiply. In additional, use sizeof(phy) is incorrect, fix it. Signed-off-by: Axel Lin Acked-by: Lee Jones Signed-off-by: Kishon Vijay Abraham I --- drivers/phy/phy-miphy365x.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/phy/phy-miphy365x.c b/drivers/phy/phy-miphy365x.c index 61177a6c465a..51b459db9137 100644 --- a/drivers/phy/phy-miphy365x.c +++ b/drivers/phy/phy-miphy365x.c @@ -549,9 +549,8 @@ static int miphy365x_probe(struct platform_device *pdev) return -ENOMEM; miphy_dev->nphys = of_get_child_count(np); - miphy_dev->phys = devm_kzalloc(&pdev->dev, - sizeof(phy) * miphy_dev->nphys, - GFP_KERNEL); + miphy_dev->phys = devm_kcalloc(&pdev->dev, miphy_dev->nphys, + sizeof(*miphy_dev->phys), GFP_KERNEL); if (!miphy_dev->phys) return -ENOMEM; -- cgit v1.2.1 From 018e6ff3c09f32b8743e06870c4dd6ed9a0b7ff5 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Thu, 5 Mar 2015 09:43:09 +0800 Subject: phy: miphy28lp: Convert to devm_kcalloc and fix wrong sizof Prefer devm_kcalloc over devm_kzalloc with multiply. In additional, use sizeof(phy) is incorrect, fix it. Signed-off-by: Axel Lin Acked-by: Gabriel Fernandez Signed-off-by: Kishon Vijay Abraham I --- drivers/phy/phy-miphy28lp.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/phy/phy-miphy28lp.c b/drivers/phy/phy-miphy28lp.c index 4fe1755e3aa8..933435214acc 100644 --- a/drivers/phy/phy-miphy28lp.c +++ b/drivers/phy/phy-miphy28lp.c @@ -1209,9 +1209,8 @@ static int miphy28lp_probe(struct platform_device *pdev) return -ENOMEM; miphy_dev->nphys = of_get_child_count(np); - miphy_dev->phys = devm_kzalloc(&pdev->dev, - sizeof(phy) * miphy_dev->nphys, - GFP_KERNEL); + miphy_dev->phys = devm_kcalloc(&pdev->dev, miphy_dev->nphys, + sizeof(*miphy_dev->phys), GFP_KERNEL); if (!miphy_dev->phys) return -ENOMEM; -- cgit v1.2.1 From 736b67a32062240592aad49033859f9712dd18ca Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Fri, 6 Mar 2015 15:55:10 +0800 Subject: phy: core: Fixup return value of phy_exit when !pm_runtime_enabled When phy_pm_runtime_get_sync() returns -ENOTSUPP, phy_exit() also returns -ENOTSUPP if !phy->ops->exit. Fix it. Also move the code to override ret close to the code we got ret. I think it is less error prone this way. Signed-off-by: Axel Lin Acked-by: Roger Quadros Signed-off-by: Kishon Vijay Abraham I --- drivers/phy/phy-core.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/drivers/phy/phy-core.c b/drivers/phy/phy-core.c index 04fc84f2b289..3791838f4bd4 100644 --- a/drivers/phy/phy-core.c +++ b/drivers/phy/phy-core.c @@ -225,6 +225,7 @@ int phy_init(struct phy *phy) ret = phy_pm_runtime_get_sync(phy); if (ret < 0 && ret != -ENOTSUPP) return ret; + ret = 0; /* Override possible ret == -ENOTSUPP */ mutex_lock(&phy->mutex); if (phy->init_count == 0 && phy->ops->init) { @@ -233,8 +234,6 @@ int phy_init(struct phy *phy) dev_err(&phy->dev, "phy init failed --> %d\n", ret); goto out; } - } else { - ret = 0; /* Override possible ret == -ENOTSUPP */ } ++phy->init_count; @@ -255,6 +254,7 @@ int phy_exit(struct phy *phy) ret = phy_pm_runtime_get_sync(phy); if (ret < 0 && ret != -ENOTSUPP) return ret; + ret = 0; /* Override possible ret == -ENOTSUPP */ mutex_lock(&phy->mutex); if (phy->init_count == 1 && phy->ops->exit) { @@ -289,6 +289,7 @@ int phy_power_on(struct phy *phy) ret = phy_pm_runtime_get_sync(phy); if (ret < 0 && ret != -ENOTSUPP) return ret; + ret = 0; /* Override possible ret == -ENOTSUPP */ mutex_lock(&phy->mutex); if (phy->power_count == 0 && phy->ops->power_on) { @@ -297,8 +298,6 @@ int phy_power_on(struct phy *phy) dev_err(&phy->dev, "phy poweron failed --> %d\n", ret); goto out; } - } else { - ret = 0; /* Override possible ret == -ENOTSUPP */ } ++phy->power_count; mutex_unlock(&phy->mutex); -- cgit v1.2.1 From dd64ad387cc0528416037982a60f0b90acccce42 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Sat, 7 Mar 2015 00:01:21 +0800 Subject: phy: ti/omap: Fix modalias Remove extra space in MODULE_ALIAS. Signed-off-by: Axel Lin Acked-by: Roger Quadros Signed-off-by: Kishon Vijay Abraham I --- drivers/phy/phy-omap-control.c | 2 +- drivers/phy/phy-omap-usb2.c | 2 +- drivers/phy/phy-ti-pipe3.c | 2 +- 3 files changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/phy/phy-omap-control.c b/drivers/phy/phy-omap-control.c index efe724f97e02..93252e053a31 100644 --- a/drivers/phy/phy-omap-control.c +++ b/drivers/phy/phy-omap-control.c @@ -360,7 +360,7 @@ static void __exit omap_control_phy_exit(void) } module_exit(omap_control_phy_exit); -MODULE_ALIAS("platform: omap_control_phy"); +MODULE_ALIAS("platform:omap_control_phy"); MODULE_AUTHOR("Texas Instruments Inc."); MODULE_DESCRIPTION("OMAP Control Module PHY Driver"); MODULE_LICENSE("GPL v2"); diff --git a/drivers/phy/phy-omap-usb2.c b/drivers/phy/phy-omap-usb2.c index 6f4aef3db248..c4917b2cf14c 100644 --- a/drivers/phy/phy-omap-usb2.c +++ b/drivers/phy/phy-omap-usb2.c @@ -383,7 +383,7 @@ static struct platform_driver omap_usb2_driver = { module_platform_driver(omap_usb2_driver); -MODULE_ALIAS("platform: omap_usb2"); +MODULE_ALIAS("platform:omap_usb2"); MODULE_AUTHOR("Texas Instruments Inc."); MODULE_DESCRIPTION("OMAP USB2 phy driver"); MODULE_LICENSE("GPL v2"); diff --git a/drivers/phy/phy-ti-pipe3.c b/drivers/phy/phy-ti-pipe3.c index ed72b0d01dde..2ba610b72ca2 100644 --- a/drivers/phy/phy-ti-pipe3.c +++ b/drivers/phy/phy-ti-pipe3.c @@ -604,7 +604,7 @@ static struct platform_driver ti_pipe3_driver = { module_platform_driver(ti_pipe3_driver); -MODULE_ALIAS("platform: ti_pipe3"); +MODULE_ALIAS("platform:ti_pipe3"); MODULE_AUTHOR("Texas Instruments Inc."); MODULE_DESCRIPTION("TI PIPE3 phy driver"); MODULE_LICENSE("GPL v2"); -- cgit v1.2.1 From b1ff3231b2d4197ef5024f9c57ffc6cfa562590c Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Wed, 11 Mar 2015 09:41:34 +0800 Subject: phy: omap-usb2: Fix missing clk_prepare call when using old dt name Current code does not call clk_prepare(phy->optclk) when using the old usb_otg_ss_refclk960m name. Fix it. Signed-off-by: Axel Lin Signed-off-by: Kishon Vijay Abraham I --- drivers/phy/phy-omap-usb2.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/phy/phy-omap-usb2.c b/drivers/phy/phy-omap-usb2.c index c4917b2cf14c..4757e765696a 100644 --- a/drivers/phy/phy-omap-usb2.c +++ b/drivers/phy/phy-omap-usb2.c @@ -296,10 +296,11 @@ static int omap_usb2_probe(struct platform_device *pdev) dev_warn(&pdev->dev, "found usb_otg_ss_refclk960m, please fix DTS\n"); } - } else { - clk_prepare(phy->optclk); } + if (!IS_ERR(phy->optclk)) + clk_prepare(phy->optclk); + usb_add_phy_dev(&phy->phy); return 0; -- cgit v1.2.1 From 670125bda1d86edfadf81dc56a87582ac7fbd47b Mon Sep 17 00:00:00 2001 From: Wincy Van Date: Wed, 4 Mar 2015 14:31:56 +0800 Subject: KVM: VMX: Set msr bitmap correctly if vcpu is in guest mode In commit 3af18d9c5fe9 ("KVM: nVMX: Prepare for using hardware MSR bitmap"), we are setting MSR_BITMAP in prepare_vmcs02 if we should use hardware. This is not enough since the field will be modified by following vmx_set_efer. Fix this by setting vmx_msr_bitmap_nested in vmx_set_msr_bitmap if vcpu is in guest mode. Signed-off-by: Wincy Van Signed-off-by: Marcelo Tosatti --- arch/x86/kvm/vmx.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-) diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c index f7b20b417a3a..10a481b7674d 100644 --- a/arch/x86/kvm/vmx.c +++ b/arch/x86/kvm/vmx.c @@ -2168,7 +2168,10 @@ static void vmx_set_msr_bitmap(struct kvm_vcpu *vcpu) { unsigned long *msr_bitmap; - if (irqchip_in_kernel(vcpu->kvm) && apic_x2apic_mode(vcpu->arch.apic)) { + if (is_guest_mode(vcpu)) + msr_bitmap = vmx_msr_bitmap_nested; + else if (irqchip_in_kernel(vcpu->kvm) && + apic_x2apic_mode(vcpu->arch.apic)) { if (is_long_mode(vcpu)) msr_bitmap = vmx_msr_bitmap_longmode_x2apic; else @@ -9218,9 +9221,9 @@ static void prepare_vmcs02(struct kvm_vcpu *vcpu, struct vmcs12 *vmcs12) } if (cpu_has_vmx_msr_bitmap() && - exec_control & CPU_BASED_USE_MSR_BITMAPS && - nested_vmx_merge_msr_bitmap(vcpu, vmcs12)) { - vmcs_write64(MSR_BITMAP, __pa(vmx_msr_bitmap_nested)); + exec_control & CPU_BASED_USE_MSR_BITMAPS) { + nested_vmx_merge_msr_bitmap(vcpu, vmcs12); + /* MSR_BITMAP will be set by following vmx_set_efer. */ } else exec_control &= ~CPU_BASED_USE_MSR_BITMAPS; -- cgit v1.2.1 From 9a30b096b543932de218dd3501b5562e00a8792d Mon Sep 17 00:00:00 2001 From: Mike Snitzer Date: Thu, 12 Mar 2015 23:53:26 -0400 Subject: blk-mq: fix use of incorrect goto label in blk_mq_init_queue error path If percpu_ref_init() fails the allocated q and hctxs must get cleaned up; using 'err_map' doesn't allow that to happen. Signed-off-by: Mike Snitzer Reviewed-by: Ming Lei Cc: stable@kernel.org Signed-off-by: Jens Axboe --- block/blk-mq.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/block/blk-mq.c b/block/blk-mq.c index 4f4bea21052e..b7b8933ec241 100644 --- a/block/blk-mq.c +++ b/block/blk-mq.c @@ -1938,7 +1938,7 @@ struct request_queue *blk_mq_init_queue(struct blk_mq_tag_set *set) */ if (percpu_ref_init(&q->mq_usage_counter, blk_mq_usage_counter_release, PERCPU_REF_INIT_ATOMIC, GFP_KERNEL)) - goto err_map; + goto err_mq_usage; setup_timer(&q->timeout, blk_mq_rq_timer, (unsigned long) q); blk_queue_rq_timeout(q, 30000); @@ -1981,7 +1981,7 @@ struct request_queue *blk_mq_init_queue(struct blk_mq_tag_set *set) blk_mq_init_cpu_queues(q, set->nr_hw_queues); if (blk_mq_init_hw_queues(q, set)) - goto err_hw; + goto err_mq_usage; mutex_lock(&all_q_mutex); list_add_tail(&q->all_q_node, &all_q_list); @@ -1993,7 +1993,7 @@ struct request_queue *blk_mq_init_queue(struct blk_mq_tag_set *set) return q; -err_hw: +err_mq_usage: blk_cleanup_queue(q); err_hctxs: kfree(map); -- cgit v1.2.1 From b57578b3d5f53016c18a9ae5365cc6e05cd70c7a Mon Sep 17 00:00:00 2001 From: Ameen Ali Date: Fri, 13 Mar 2015 16:15:52 +0200 Subject: tulip_core.c : out-of-bounds check. Array index 'j' is used before limits check. Suggest put limit check before index use. Signed-off-by : Signed-off-by: David S. Miller --- drivers/net/ethernet/dec/tulip/tulip_core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/dec/tulip/tulip_core.c b/drivers/net/ethernet/dec/tulip/tulip_core.c index 3b42556f7f8d..ed41559bae77 100644 --- a/drivers/net/ethernet/dec/tulip/tulip_core.c +++ b/drivers/net/ethernet/dec/tulip/tulip_core.c @@ -589,7 +589,7 @@ static void tulip_tx_timeout(struct net_device *dev) (unsigned int)tp->rx_ring[i].buffer1, (unsigned int)tp->rx_ring[i].buffer2, buf[0], buf[1], buf[2]); - for (j = 0; buf[j] != 0xee && j < 1600; j++) + for (j = 0; ((j < 1600) && buf[j] != 0xee); j++) if (j < 100) pr_cont(" %02x", buf[j]); pr_cont(" j=%d\n", j); -- cgit v1.2.1 From 40fb70f3aa0a67d28a30c854d4e7aa10b0511db9 Mon Sep 17 00:00:00 2001 From: Alexey Kodanev Date: Fri, 13 Mar 2015 19:13:53 +0300 Subject: vxlan: fix wrong usage of VXLAN_VID_MASK commit dfd8645ea1bd9127 wrongly assumes that VXLAN_VDI_MASK includes eight lower order reserved bits of VNI field that are using for remote checksum offload. Right now, when VNI number greater then 0xffff, vxlan_udp_encap_recv() will always return with 'bad_flag' error, reducing the usable vni range from 0..16777215 to 0..65535. Also, it doesn't really check whether RCO bits processed or not. Fix it by adding new VNI mask which has all 32 bits of VNI field: 24 bits for id and 8 bits for other usage. Signed-off-by: Alexey Kodanev Signed-off-by: David S. Miller --- drivers/net/vxlan.c | 4 ++-- include/net/vxlan.h | 1 + 2 files changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/net/vxlan.c b/drivers/net/vxlan.c index 1e0a775ea882..f8528a4cf54f 100644 --- a/drivers/net/vxlan.c +++ b/drivers/net/vxlan.c @@ -1218,7 +1218,7 @@ static int vxlan_udp_encap_recv(struct sock *sk, struct sk_buff *skb) goto drop; flags &= ~VXLAN_HF_RCO; - vni &= VXLAN_VID_MASK; + vni &= VXLAN_VNI_MASK; } /* For backwards compatibility, only allow reserved fields to be @@ -1239,7 +1239,7 @@ static int vxlan_udp_encap_recv(struct sock *sk, struct sk_buff *skb) flags &= ~VXLAN_GBP_USED_BITS; } - if (flags || (vni & ~VXLAN_VID_MASK)) { + if (flags || vni & ~VXLAN_VNI_MASK) { /* If there are any unprocessed flags remaining treat * this as a malformed packet. This behavior diverges from * VXLAN RFC (RFC7348) which stipulates that bits in reserved diff --git a/include/net/vxlan.h b/include/net/vxlan.h index eabd3a038674..c73e7abbbaa5 100644 --- a/include/net/vxlan.h +++ b/include/net/vxlan.h @@ -91,6 +91,7 @@ struct vxlanhdr { #define VXLAN_N_VID (1u << 24) #define VXLAN_VID_MASK (VXLAN_N_VID - 1) +#define VXLAN_VNI_MASK (VXLAN_VID_MASK << 8) #define VXLAN_HLEN (sizeof(struct udphdr) + sizeof(struct vxlanhdr)) struct vxlan_metadata { -- cgit v1.2.1 From a2fe37b69d4fe369c284d50927193fed81c238a0 Mon Sep 17 00:00:00 2001 From: Fabio Estevam Date: Fri, 13 Mar 2015 14:07:54 -0300 Subject: Revert "net: fec: fix the warning found by dma debug" MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This reverts commit 2b995f63987013bacde99168218f9c7b252bdcf1. Панов Андрей reported the following regression: "Commit 2b995f63987013bacde99168218f9c7b252bdcf1 in 4.0.0-rc3 introduces a nasty bug in transmit, corrupting packets. To reproduce: $ dd if=/dev/zero of=zeros bs=1M count=20 $ md5sum -b zeros 8f4e33f3dc3e414ff94e5fb6905cba8c *zeros This checksum is correct. Copy file "zeros" to another host with NFS, and it gets corrupted, checksum is changed. File should be big, small amounts of transmit isn't affected. I use an i.MX6 Quad board. If this commit is reverted, all works fine." Reported-by: Панов Андрей Signed-off-by: Fabio Estevam Signed-off-by: David S. Miller --- drivers/net/ethernet/freescale/fec_main.c | 34 ++++++++++--------------------- 1 file changed, 11 insertions(+), 23 deletions(-) diff --git a/drivers/net/ethernet/freescale/fec_main.c b/drivers/net/ethernet/freescale/fec_main.c index 787db5026191..78e1ce09b1ab 100644 --- a/drivers/net/ethernet/freescale/fec_main.c +++ b/drivers/net/ethernet/freescale/fec_main.c @@ -1189,13 +1189,12 @@ static void fec_enet_tx_queue(struct net_device *ndev, u16 queue_id) { struct fec_enet_private *fep; - struct bufdesc *bdp, *bdp_t; + struct bufdesc *bdp; unsigned short status; struct sk_buff *skb; struct fec_enet_priv_tx_q *txq; struct netdev_queue *nq; int index = 0; - int i, bdnum; int entries_free; fep = netdev_priv(ndev); @@ -1216,29 +1215,18 @@ fec_enet_tx_queue(struct net_device *ndev, u16 queue_id) if (bdp == txq->cur_tx) break; - bdp_t = bdp; - bdnum = 1; - index = fec_enet_get_bd_index(txq->tx_bd_base, bdp_t, fep); - skb = txq->tx_skbuff[index]; - while (!skb) { - bdp_t = fec_enet_get_nextdesc(bdp_t, fep, queue_id); - index = fec_enet_get_bd_index(txq->tx_bd_base, bdp_t, fep); - skb = txq->tx_skbuff[index]; - bdnum++; - } - if (skb_shinfo(skb)->nr_frags && - (status = bdp_t->cbd_sc) & BD_ENET_TX_READY) - break; + index = fec_enet_get_bd_index(txq->tx_bd_base, bdp, fep); - for (i = 0; i < bdnum; i++) { - if (!IS_TSO_HEADER(txq, bdp->cbd_bufaddr)) - dma_unmap_single(&fep->pdev->dev, bdp->cbd_bufaddr, - bdp->cbd_datlen, DMA_TO_DEVICE); - bdp->cbd_bufaddr = 0; - if (i < bdnum - 1) - bdp = fec_enet_get_nextdesc(bdp, fep, queue_id); - } + skb = txq->tx_skbuff[index]; txq->tx_skbuff[index] = NULL; + if (!IS_TSO_HEADER(txq, bdp->cbd_bufaddr)) + dma_unmap_single(&fep->pdev->dev, bdp->cbd_bufaddr, + bdp->cbd_datlen, DMA_TO_DEVICE); + bdp->cbd_bufaddr = 0; + if (!skb) { + bdp = fec_enet_get_nextdesc(bdp, fep, queue_id); + continue; + } /* Check for errors. */ if (status & (BD_ENET_TX_HB | BD_ENET_TX_LC | -- cgit v1.2.1 From 43b68879de27b1993518687fbc6013da80cdcbfe Mon Sep 17 00:00:00 2001 From: Gregory CLEMENT Date: Thu, 26 Feb 2015 18:20:48 +0100 Subject: cpuidle: mvebu: Fix the CPU PM notifier usage As stated in kernel/cpu_pm.c, "Platform is responsible for ensuring that cpu_pm_enter is not called twice on the same CPU before cpu_pm_exit is called.". In the current code in case of failure when calling mvebu_v7_cpu_suspend, the function cpu_pm_exit() is never called whereas cpu_pm_enter() was called just before. This patch moves the cpu_pm_exit() in order to balance the cpu_pm_enter() calls. Cc: stable@vger.kernel.org Reported-by: Fulvio Benini Signed-off-by: Gregory CLEMENT Signed-off-by: Daniel Lezcano --- drivers/cpuidle/cpuidle-mvebu-v7.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/cpuidle/cpuidle-mvebu-v7.c b/drivers/cpuidle/cpuidle-mvebu-v7.c index 38e68618513a..cefa07438ae1 100644 --- a/drivers/cpuidle/cpuidle-mvebu-v7.c +++ b/drivers/cpuidle/cpuidle-mvebu-v7.c @@ -37,11 +37,11 @@ static int mvebu_v7_enter_idle(struct cpuidle_device *dev, deepidle = true; ret = mvebu_v7_cpu_suspend(deepidle); + cpu_pm_exit(); + if (ret) return ret; - cpu_pm_exit(); - return index; } -- cgit v1.2.1 From ce6031c89a35cffd5a5992b08377b77f49a004b9 Mon Sep 17 00:00:00 2001 From: Sebastien Rannou Date: Fri, 13 Feb 2015 15:55:03 +0100 Subject: cpuidle: mvebu: Update cpuidle thresholds for Armada XP SOCs Originally, the thresholds used in the cpuidle driver for Armada SOCs were temporarily chosen, leaving room for improvements. This commit updates the thresholds for the Armada XP SOCs with values that positively impact performances: without patch with patch vendor kernel - iperf localhost (gbit/sec) ~3.7 ~6.4 ~5.4 - ioping tmpfs (iops) ~163k ~206k ~179k - ioping tmpfs (mib/s) ~636 ~805 ~699 The idle power consumption is negatively impacted (proportionally less than the performance gain), and we are still performing better than the vendor kernel here: without patch with patch vendor kernel - power consumption idle (W) ~2.4 ~3.2 ~4.4 - power consumption busy (W) ~8.6 ~8.3 ~8.6 There is still room for improvement regarding the value of these thresholds, they were chosen to mimic the vendor kernel. This patch only impacts Armada XP SOCs and was tested on Online Labs C1 boards. A similar approach can be taken to improve the performances of the Armada 370 and Armada 38x SOCs. Thanks a lot to Thomas Petazzoni, Gregory Clement and Willy Tarreau for the discussions and tips around this topic. Signed-off-by: Sebastien Rannou Signed-off-by: Daniel Lezcano Acked-by: Gregory CLEMENT --- drivers/cpuidle/cpuidle-mvebu-v7.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/cpuidle/cpuidle-mvebu-v7.c b/drivers/cpuidle/cpuidle-mvebu-v7.c index cefa07438ae1..980151f34707 100644 --- a/drivers/cpuidle/cpuidle-mvebu-v7.c +++ b/drivers/cpuidle/cpuidle-mvebu-v7.c @@ -50,17 +50,17 @@ static struct cpuidle_driver armadaxp_idle_driver = { .states[0] = ARM_CPUIDLE_WFI_STATE, .states[1] = { .enter = mvebu_v7_enter_idle, - .exit_latency = 10, + .exit_latency = 100, .power_usage = 50, - .target_residency = 100, + .target_residency = 1000, .name = "MV CPU IDLE", .desc = "CPU power down", }, .states[2] = { .enter = mvebu_v7_enter_idle, - .exit_latency = 100, + .exit_latency = 1000, .power_usage = 5, - .target_residency = 1000, + .target_residency = 10000, .flags = MVEBU_V7_FLAG_DEEP_IDLE, .name = "MV CPU DEEP IDLE", .desc = "CPU and L2 Fabric power down", -- cgit v1.2.1 From c8e2c80d7ec00d020320f905822bf49c5ad85250 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Fri, 13 Mar 2015 09:49:59 -0700 Subject: inet_diag: fix possible overflow in inet_diag_dump_one_icsk() inet_diag_dump_one_icsk() allocates too small skb. Add inet_sk_attr_size() helper right before inet_sk_diag_fill() so that it can be updated if/when new attributes are added. iproute2/ss currently does not use this dump_one() interface, this might explain nobody noticed this problem yet. Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/inet_diag.c | 18 +++++++++++++++--- 1 file changed, 15 insertions(+), 3 deletions(-) diff --git a/net/ipv4/inet_diag.c b/net/ipv4/inet_diag.c index 81751f12645f..592aff37366b 100644 --- a/net/ipv4/inet_diag.c +++ b/net/ipv4/inet_diag.c @@ -71,6 +71,20 @@ static inline void inet_diag_unlock_handler( mutex_unlock(&inet_diag_table_mutex); } +static size_t inet_sk_attr_size(void) +{ + return nla_total_size(sizeof(struct tcp_info)) + + nla_total_size(1) /* INET_DIAG_SHUTDOWN */ + + nla_total_size(1) /* INET_DIAG_TOS */ + + nla_total_size(1) /* INET_DIAG_TCLASS */ + + nla_total_size(sizeof(struct inet_diag_meminfo)) + + nla_total_size(sizeof(struct inet_diag_msg)) + + nla_total_size(SK_MEMINFO_VARS * sizeof(u32)) + + nla_total_size(TCP_CA_NAME_MAX) + + nla_total_size(sizeof(struct tcpvegas_info)) + + 64; +} + int inet_sk_diag_fill(struct sock *sk, struct inet_connection_sock *icsk, struct sk_buff *skb, struct inet_diag_req_v2 *req, struct user_namespace *user_ns, @@ -326,9 +340,7 @@ int inet_diag_dump_one_icsk(struct inet_hashinfo *hashinfo, struct sk_buff *in_s if (err) goto out; - rep = nlmsg_new(sizeof(struct inet_diag_msg) + - sizeof(struct inet_diag_meminfo) + - sizeof(struct tcp_info) + 64, GFP_KERNEL); + rep = nlmsg_new(inet_sk_attr_size(), GFP_KERNEL); if (!rep) { err = -ENOMEM; goto out; -- cgit v1.2.1 From d22071293f17eedc96df25093d8e99d09cd76463 Mon Sep 17 00:00:00 2001 From: Fabian Frederick Date: Sat, 14 Feb 2015 13:10:22 +0100 Subject: btrfs: fix sizeof format specifier in btrfs_check_super_valid() This patch fixes mips compilation warning: fs/btrfs/disk-io.c: In function 'btrfs_check_super_valid': fs/btrfs/disk-io.c:3927:21: warning: format '%lu' expects argument of type 'long unsigned int', but argument 3 has type 'unsigned int' [-Wformat] Signed-off-by: Fabian Frederick Acked-by: Geert Uytterhoeven Signed-off-by: Chris Mason --- fs/btrfs/disk-io.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 41b320e235d7..1577b91940dd 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -3923,7 +3923,7 @@ static int btrfs_check_super_valid(struct btrfs_fs_info *fs_info, } if (btrfs_super_sys_array_size(sb) < sizeof(struct btrfs_disk_key) + sizeof(struct btrfs_chunk)) { - printk(KERN_ERR "BTRFS: system chunk array too small %u < %lu\n", + printk(KERN_ERR "BTRFS: system chunk array too small %u < %zu\n", btrfs_super_sys_array_size(sb), sizeof(struct btrfs_disk_key) + sizeof(struct btrfs_chunk)); -- cgit v1.2.1 From b4924a0fa18d7f69bde3a84521258e7a55828186 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Fri, 6 Mar 2015 20:23:44 +0800 Subject: Btrfs: catch transaction abortion after waiting for it This problem is uncovered by a test case: http://patchwork.ozlabs.org/patch/244297. Fsync() can report success when it actually doesn't. When we have several threads running fsync() at the same tiem and in one fsync() we get a transaction abortion due to some problems(in the test case it's disk failures), and other fsync()s may return successfully which makes userspace programs think that data is now safely flushed into disk. It's because that after fsyncs() fail btrfs_sync_log() due to disk failures, they get to try btrfs_commit_transaction() where it finds that there is already a transaction being committed, and they'll just call wait_for_commit() and return. Note that we actually check "trans->aborted" in btrfs_end_transaction, but it's likely that the error message is still not yet throwed out and only after wait_for_commit() we're sure whether the transaction is committed successfully. This add the necessary check and it now passes the test. Signed-off-by: Liu Bo Signed-off-by: Chris Mason --- fs/btrfs/transaction.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 323c6541d3dc..07b985f2a814 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -1811,6 +1811,9 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, wait_for_commit(root, cur_trans); + if (unlikely(cur_trans->aborted)) + ret = cur_trans->aborted; + btrfs_put_transaction(cur_trans); return ret; -- cgit v1.2.1 From 48da5f0a4cb2b6b44579f5737e8be888c0d02526 Mon Sep 17 00:00:00 2001 From: Liu Bo Date: Fri, 13 Mar 2015 14:24:38 +0800 Subject: Btrfs: fix comp_oper to get right order Case (oper1->seq > oper2->seq) should differ with case (oper1->seq < oper2->seq). Signed-off-by: Liu Bo Reviewed-by: David Sterba Reviewed-by: Filipe Manana Signed-off-by: Chris Mason --- fs/btrfs/qgroup.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/btrfs/qgroup.c b/fs/btrfs/qgroup.c index 97159a8e91d4..058c79eecbfb 100644 --- a/fs/btrfs/qgroup.c +++ b/fs/btrfs/qgroup.c @@ -1259,7 +1259,7 @@ static int comp_oper(struct btrfs_qgroup_operation *oper1, if (oper1->seq < oper2->seq) return -1; if (oper1->seq > oper2->seq) - return -1; + return 1; if (oper1->ref_root < oper2->ref_root) return -1; if (oper1->ref_root > oper2->ref_root) -- cgit v1.2.1 From 8461a3de770477a9a7b8eeaebcc4804dbc26ca38 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 13 Mar 2015 15:12:08 -0400 Subject: Btrfs: fix merge delalloc logic My patch to properly count outstanding extents wrt MAX_EXTENT_SIZE introduced a regression when re-dirtying already dirty areas. We have logic in split to make sure we are taking the largest space into account but didn't have it for merge, so it was sometimes making us think we were turning a tiny extent into a huge extent, when in reality we already had a huge extent and needed to use the other side in our logic. This fixes the regression that was reported by a user on list. Thanks, Reported-by: Markus Trippelsdorf Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/inode.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 91a87f53be3c..97b601bec326 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -1592,7 +1592,12 @@ static void btrfs_merge_extent_hook(struct inode *inode, return; old_size = other->end - other->start + 1; - new_size = old_size + (new->end - new->start + 1); + if (old_size < (new->end - new->start + 1)) + old_size = (new->end - new->start + 1); + if (new->start > other->start) + new_size = new->end - other->start + 1; + else + new_size = other->end - new->start + 1; /* we're not bigger than the max, unreserve the space and go */ if (new_size <= BTRFS_MAX_EXTENT_SIZE) { -- cgit v1.2.1 From 6a41dd0922e3c63e677c2d8f7906ce6a3e097af1 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 13 Mar 2015 15:12:23 -0400 Subject: Btrfs: account for the correct number of extents for delalloc reservations Direct IO can easily pass in an buffer that is greater than BTRFS_MAX_EXTENT_SIZE, so take this into account when reserving extents in the delalloc reservation code. Thanks, Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/extent-tree.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 92146a5afdc1..96c613bfe157 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -5110,7 +5110,11 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes) num_bytes = ALIGN(num_bytes, root->sectorsize); spin_lock(&BTRFS_I(inode)->lock); - BTRFS_I(inode)->outstanding_extents++; + nr_extents = (unsigned)div64_u64(num_bytes + + BTRFS_MAX_EXTENT_SIZE - 1, + BTRFS_MAX_EXTENT_SIZE); + BTRFS_I(inode)->outstanding_extents += nr_extents; + nr_extents = 0; if (BTRFS_I(inode)->outstanding_extents > BTRFS_I(inode)->reserved_extents) -- cgit v1.2.1 From ea526d18990018f224e5734748975bea1824545f Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 13 Mar 2015 16:40:45 -0400 Subject: Btrfs: fix ASSERT(list_empty(&cur_trans->dirty_bgs_list) Dave could hit this assert consistently running btrfs/078. This is because when we update the block groups we could truncate the free space, which would try to delete the csums for that range and dirty the csum root. For this to happen we have to have already written out the csum root so it's kind of hard to hit this case. This patch fixes this by changing the logic to only write the dirty block groups if the dirty_cowonly_roots list is empty. This will get us the same effect as before since we add the extent root last, and will cover the case that we dirty some other root again but not the extent root. Thanks, Reported-by: David Sterba Signed-off-by: Josef Bacik Signed-off-by: Chris Mason --- fs/btrfs/transaction.c | 31 ++++++++++++++++++------------- 1 file changed, 18 insertions(+), 13 deletions(-) diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 07b985f2a814..2fe3ef5e9de3 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -1023,7 +1023,6 @@ static int update_cowonly_root(struct btrfs_trans_handle *trans, u64 old_root_bytenr; u64 old_root_used; struct btrfs_root *tree_root = root->fs_info->tree_root; - bool extent_root = (root->objectid == BTRFS_EXTENT_TREE_OBJECTID); old_root_used = btrfs_root_used(&root->root_item); btrfs_write_dirty_block_groups(trans, root); @@ -1031,9 +1030,7 @@ static int update_cowonly_root(struct btrfs_trans_handle *trans, while (1) { old_root_bytenr = btrfs_root_bytenr(&root->root_item); if (old_root_bytenr == root->node->start && - old_root_used == btrfs_root_used(&root->root_item) && - (!extent_root || - list_empty(&trans->transaction->dirty_bgs))) + old_root_used == btrfs_root_used(&root->root_item)) break; btrfs_set_root_node(&root->root_item, root->node); @@ -1044,14 +1041,6 @@ static int update_cowonly_root(struct btrfs_trans_handle *trans, return ret; old_root_used = btrfs_root_used(&root->root_item); - if (extent_root) { - ret = btrfs_write_dirty_block_groups(trans, root); - if (ret) - return ret; - } - ret = btrfs_run_delayed_refs(trans, root, (unsigned long)-1); - if (ret) - return ret; } return 0; @@ -1068,6 +1057,7 @@ static noinline int commit_cowonly_roots(struct btrfs_trans_handle *trans, struct btrfs_root *root) { struct btrfs_fs_info *fs_info = root->fs_info; + struct list_head *dirty_bgs = &trans->transaction->dirty_bgs; struct list_head *next; struct extent_buffer *eb; int ret; @@ -1099,7 +1089,7 @@ static noinline int commit_cowonly_roots(struct btrfs_trans_handle *trans, ret = btrfs_run_delayed_refs(trans, root, (unsigned long)-1); if (ret) return ret; - +again: while (!list_empty(&fs_info->dirty_cowonly_roots)) { next = fs_info->dirty_cowonly_roots.next; list_del_init(next); @@ -1112,8 +1102,23 @@ static noinline int commit_cowonly_roots(struct btrfs_trans_handle *trans, ret = update_cowonly_root(trans, root); if (ret) return ret; + ret = btrfs_run_delayed_refs(trans, root, (unsigned long)-1); + if (ret) + return ret; } + while (!list_empty(dirty_bgs)) { + ret = btrfs_write_dirty_block_groups(trans, root); + if (ret) + return ret; + ret = btrfs_run_delayed_refs(trans, root, (unsigned long)-1); + if (ret) + return ret; + } + + if (!list_empty(&fs_info->dirty_cowonly_roots)) + goto again; + list_add_tail(&fs_info->extent_root->dirty_list, &trans->transaction->switch_commits); btrfs_after_dev_replace_commit(fs_info); -- cgit v1.2.1 From d474a4d365aaa5c7aabcf11a74ea43aa23f6f2e9 Mon Sep 17 00:00:00 2001 From: Jacob Pan Date: Fri, 13 Mar 2015 03:48:56 -0700 Subject: powercap / RAPL: handle domains with different energy units The current driver assumes all RAPL domains within a CPU package have the same energy unit. This is no longer true for HSW server CPUs since DRAM domain has is own fixed energy unit which can be different than the package energy unit enumerated by package power MSR. In fact, the default HSW EP package power unit is 61uJ whereas DRAM domain unit is 15.3uJ. The result is that DRAM power consumption is counted 4x more than real power reported by energy counters, similarly for max_energy_range_uj of DRAM domain. This patch adds domain specific energy unit per cpu type, it allows domain energy unit to override package energy unit if non zero. Please see this document for details. "Intel Xeon Processor E5-1600 and E5-2600 v3 Product Families, Volume 2 of 2. Datasheet, September 2014, Reference Number: 330784-001 " Signed-off-by: Jacob Pan Cc: 3.10+ # 3.10+ Signed-off-by: Rafael J. Wysocki --- drivers/powercap/intel_rapl.c | 54 +++++++++++++++++++++++++++++++------------ 1 file changed, 39 insertions(+), 15 deletions(-) diff --git a/drivers/powercap/intel_rapl.c b/drivers/powercap/intel_rapl.c index 97b5e4ee1ca4..63d4033eb683 100644 --- a/drivers/powercap/intel_rapl.c +++ b/drivers/powercap/intel_rapl.c @@ -73,7 +73,7 @@ #define TIME_WINDOW_MAX_MSEC 40000 #define TIME_WINDOW_MIN_MSEC 250 - +#define ENERGY_UNIT_SCALE 1000 /* scale from driver unit to powercap unit */ enum unit_type { ARBITRARY_UNIT, /* no translation */ POWER_UNIT, @@ -158,6 +158,7 @@ struct rapl_domain { struct rapl_power_limit rpl[NR_POWER_LIMITS]; u64 attr_map; /* track capabilities */ unsigned int state; + unsigned int domain_energy_unit; int package_id; }; #define power_zone_to_rapl_domain(_zone) \ @@ -190,6 +191,7 @@ struct rapl_defaults { void (*set_floor_freq)(struct rapl_domain *rd, bool mode); u64 (*compute_time_window)(struct rapl_package *rp, u64 val, bool to_raw); + unsigned int dram_domain_energy_unit; }; static struct rapl_defaults *rapl_defaults; @@ -227,7 +229,8 @@ static int rapl_read_data_raw(struct rapl_domain *rd, static int rapl_write_data_raw(struct rapl_domain *rd, enum rapl_primitives prim, unsigned long long value); -static u64 rapl_unit_xlate(int package, enum unit_type type, u64 value, +static u64 rapl_unit_xlate(struct rapl_domain *rd, int package, + enum unit_type type, u64 value, int to_raw); static void package_power_limit_irq_save(int package_id); @@ -305,7 +308,9 @@ static int get_energy_counter(struct powercap_zone *power_zone, u64 *energy_raw) static int get_max_energy_counter(struct powercap_zone *pcd_dev, u64 *energy) { - *energy = rapl_unit_xlate(0, ENERGY_UNIT, ENERGY_STATUS_MASK, 0); + struct rapl_domain *rd = power_zone_to_rapl_domain(pcd_dev); + + *energy = rapl_unit_xlate(rd, 0, ENERGY_UNIT, ENERGY_STATUS_MASK, 0); return 0; } @@ -639,6 +644,11 @@ static void rapl_init_domains(struct rapl_package *rp) rd->msrs[4] = MSR_DRAM_POWER_INFO; rd->rpl[0].prim_id = PL1_ENABLE; rd->rpl[0].name = pl1_name; + rd->domain_energy_unit = + rapl_defaults->dram_domain_energy_unit; + if (rd->domain_energy_unit) + pr_info("DRAM domain energy unit %dpj\n", + rd->domain_energy_unit); break; } if (mask) { @@ -648,11 +658,13 @@ static void rapl_init_domains(struct rapl_package *rp) } } -static u64 rapl_unit_xlate(int package, enum unit_type type, u64 value, +static u64 rapl_unit_xlate(struct rapl_domain *rd, int package, + enum unit_type type, u64 value, int to_raw) { u64 units = 1; struct rapl_package *rp; + u64 scale = 1; rp = find_package_by_id(package); if (!rp) @@ -663,7 +675,12 @@ static u64 rapl_unit_xlate(int package, enum unit_type type, u64 value, units = rp->power_unit; break; case ENERGY_UNIT: - units = rp->energy_unit; + scale = ENERGY_UNIT_SCALE; + /* per domain unit takes precedence */ + if (rd && rd->domain_energy_unit) + units = rd->domain_energy_unit; + else + units = rp->energy_unit; break; case TIME_UNIT: return rapl_defaults->compute_time_window(rp, value, to_raw); @@ -673,11 +690,11 @@ static u64 rapl_unit_xlate(int package, enum unit_type type, u64 value, }; if (to_raw) - return div64_u64(value, units); + return div64_u64(value, units) * scale; value *= units; - return value; + return div64_u64(value, scale); } /* in the order of enum rapl_primitives */ @@ -773,7 +790,7 @@ static int rapl_read_data_raw(struct rapl_domain *rd, final = value & rp->mask; final = final >> rp->shift; if (xlate) - *data = rapl_unit_xlate(rd->package_id, rp->unit, final, 0); + *data = rapl_unit_xlate(rd, rd->package_id, rp->unit, final, 0); else *data = final; @@ -799,7 +816,7 @@ static int rapl_write_data_raw(struct rapl_domain *rd, "failed to read msr 0x%x on cpu %d\n", msr, cpu); return -EIO; } - value = rapl_unit_xlate(rd->package_id, rp->unit, value, 1); + value = rapl_unit_xlate(rd, rd->package_id, rp->unit, value, 1); msr_val &= ~rp->mask; msr_val |= value << rp->shift; if (wrmsrl_safe_on_cpu(cpu, msr, msr_val)) { @@ -818,7 +835,7 @@ static int rapl_write_data_raw(struct rapl_domain *rd, * calculate units differ on different CPUs. * We convert the units to below format based on CPUs. * i.e. - * energy unit: microJoules : Represented in microJoules by default + * energy unit: picoJoules : Represented in picoJoules by default * power unit : microWatts : Represented in milliWatts by default * time unit : microseconds: Represented in seconds by default */ @@ -834,7 +851,7 @@ static int rapl_check_unit_core(struct rapl_package *rp, int cpu) } value = (msr_val & ENERGY_UNIT_MASK) >> ENERGY_UNIT_OFFSET; - rp->energy_unit = 1000000 / (1 << value); + rp->energy_unit = ENERGY_UNIT_SCALE * 1000000 / (1 << value); value = (msr_val & POWER_UNIT_MASK) >> POWER_UNIT_OFFSET; rp->power_unit = 1000000 / (1 << value); @@ -842,7 +859,7 @@ static int rapl_check_unit_core(struct rapl_package *rp, int cpu) value = (msr_val & TIME_UNIT_MASK) >> TIME_UNIT_OFFSET; rp->time_unit = 1000000 / (1 << value); - pr_debug("Core CPU package %d energy=%duJ, time=%dus, power=%duW\n", + pr_debug("Core CPU package %d energy=%dpJ, time=%dus, power=%duW\n", rp->id, rp->energy_unit, rp->time_unit, rp->power_unit); return 0; @@ -859,7 +876,7 @@ static int rapl_check_unit_atom(struct rapl_package *rp, int cpu) return -ENODEV; } value = (msr_val & ENERGY_UNIT_MASK) >> ENERGY_UNIT_OFFSET; - rp->energy_unit = 1 << value; + rp->energy_unit = ENERGY_UNIT_SCALE * 1 << value; value = (msr_val & POWER_UNIT_MASK) >> POWER_UNIT_OFFSET; rp->power_unit = (1 << value) * 1000; @@ -867,7 +884,7 @@ static int rapl_check_unit_atom(struct rapl_package *rp, int cpu) value = (msr_val & TIME_UNIT_MASK) >> TIME_UNIT_OFFSET; rp->time_unit = 1000000 / (1 << value); - pr_debug("Atom package %d energy=%duJ, time=%dus, power=%duW\n", + pr_debug("Atom package %d energy=%dpJ, time=%dus, power=%duW\n", rp->id, rp->energy_unit, rp->time_unit, rp->power_unit); return 0; @@ -1017,6 +1034,13 @@ static const struct rapl_defaults rapl_defaults_core = { .compute_time_window = rapl_compute_time_window_core, }; +static const struct rapl_defaults rapl_defaults_hsw_server = { + .check_unit = rapl_check_unit_core, + .set_floor_freq = set_floor_freq_default, + .compute_time_window = rapl_compute_time_window_core, + .dram_domain_energy_unit = 15300, +}; + static const struct rapl_defaults rapl_defaults_atom = { .check_unit = rapl_check_unit_atom, .set_floor_freq = set_floor_freq_atom, @@ -1037,7 +1061,7 @@ static const struct x86_cpu_id rapl_ids[] = { RAPL_CPU(0x3a, rapl_defaults_core),/* Ivy Bridge */ RAPL_CPU(0x3c, rapl_defaults_core),/* Haswell */ RAPL_CPU(0x3d, rapl_defaults_core),/* Broadwell */ - RAPL_CPU(0x3f, rapl_defaults_core),/* Haswell */ + RAPL_CPU(0x3f, rapl_defaults_hsw_server),/* Haswell servers */ RAPL_CPU(0x45, rapl_defaults_core),/* Haswell ULT */ RAPL_CPU(0x4C, rapl_defaults_atom),/* Braswell */ RAPL_CPU(0x4A, rapl_defaults_atom),/* Tangier */ -- cgit v1.2.1 From 963a822b6d5fece27c88522ac34dd48928571c8b Mon Sep 17 00:00:00 2001 From: Michal Simek Date: Mon, 9 Mar 2015 15:03:32 +0100 Subject: net: can: Enable xilinx driver for ARM64 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Enable the xilinx driver for ARM64. Signed-off-by: Michal Simek Acked-by: Sören Brinkmann Signed-off-by: Marc Kleine-Budde --- drivers/net/can/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/can/Kconfig b/drivers/net/can/Kconfig index 98d73aab52fe..58808f651452 100644 --- a/drivers/net/can/Kconfig +++ b/drivers/net/can/Kconfig @@ -131,7 +131,7 @@ config CAN_RCAR config CAN_XILINXCAN tristate "Xilinx CAN" - depends on ARCH_ZYNQ || MICROBLAZE || COMPILE_TEST + depends on ARCH_ZYNQ || ARM64 || MICROBLAZE || COMPILE_TEST depends on COMMON_CLK && HAS_IOMEM ---help--- Xilinx CAN driver. This driver supports both soft AXI CAN IP and -- cgit v1.2.1 From a9dc960c37b0d4eb192598dc4c94276270454514 Mon Sep 17 00:00:00 2001 From: "Ahmed S. Darwish" Date: Sat, 14 Mar 2015 09:02:49 -0400 Subject: can: kvaser_usb: Fix tx queue start/stop race conditions A number of tx queue wake-up events went missing due to the outlined scenario below. Start state is a pool of 16 tx URBs, active tx_urbs count = 15, with the netdev tx queue open. CPU #1 [softirq] CPU #2 [softirq] start_xmit() tx_acknowledge() ................ ................ atomic_inc(&tx_urbs); if (atomic_read(&tx_urbs) >= 16) { --> atomic_dec(&tx_urbs); netif_wake_queue(); return; <-- netif_stop_queue(); } At the end, the correct state expected is a 15 tx_urbs count value with the tx queue state _open_. Due to the race, we get the same tx_urbs value but with the tx queue state _stopped_. The wake-up event is completely lost. Thus avoid hand-rolled concurrency mechanisms and use a proper lock for contexts and tx queue protection. Signed-off-by: Ahmed S. Darwish Cc: linux-stable Signed-off-by: Marc Kleine-Budde --- drivers/net/can/usb/kvaser_usb.c | 83 ++++++++++++++++++++++++---------------- 1 file changed, 51 insertions(+), 32 deletions(-) diff --git a/drivers/net/can/usb/kvaser_usb.c b/drivers/net/can/usb/kvaser_usb.c index a316fa4b91ab..e97a08ce0b90 100644 --- a/drivers/net/can/usb/kvaser_usb.c +++ b/drivers/net/can/usb/kvaser_usb.c @@ -14,6 +14,7 @@ * Copyright (C) 2015 Valeo S.A. */ +#include #include #include #include @@ -467,10 +468,11 @@ struct kvaser_usb { struct kvaser_usb_net_priv { struct can_priv can; - atomic_t active_tx_urbs; - struct usb_anchor tx_submitted; + spinlock_t tx_contexts_lock; + int active_tx_contexts; struct kvaser_usb_tx_urb_context tx_contexts[MAX_TX_URBS]; + struct usb_anchor tx_submitted; struct completion start_comp, stop_comp; struct kvaser_usb *dev; @@ -694,6 +696,7 @@ static void kvaser_usb_tx_acknowledge(const struct kvaser_usb *dev, struct kvaser_usb_net_priv *priv; struct sk_buff *skb; struct can_frame *cf; + unsigned long flags; u8 channel, tid; channel = msg->u.tx_acknowledge_header.channel; @@ -737,12 +740,15 @@ static void kvaser_usb_tx_acknowledge(const struct kvaser_usb *dev, stats->tx_packets++; stats->tx_bytes += context->dlc; - can_get_echo_skb(priv->netdev, context->echo_index); - context->echo_index = MAX_TX_URBS; - atomic_dec(&priv->active_tx_urbs); + spin_lock_irqsave(&priv->tx_contexts_lock, flags); + can_get_echo_skb(priv->netdev, context->echo_index); + context->echo_index = MAX_TX_URBS; + --priv->active_tx_contexts; netif_wake_queue(priv->netdev); + + spin_unlock_irqrestore(&priv->tx_contexts_lock, flags); } static void kvaser_usb_simple_msg_callback(struct urb *urb) @@ -803,17 +809,6 @@ static int kvaser_usb_simple_msg_async(struct kvaser_usb_net_priv *priv, return 0; } -static void kvaser_usb_unlink_tx_urbs(struct kvaser_usb_net_priv *priv) -{ - int i; - - usb_kill_anchored_urbs(&priv->tx_submitted); - atomic_set(&priv->active_tx_urbs, 0); - - for (i = 0; i < MAX_TX_URBS; i++) - priv->tx_contexts[i].echo_index = MAX_TX_URBS; -} - static void kvaser_usb_rx_error_update_can_state(struct kvaser_usb_net_priv *priv, const struct kvaser_usb_error_summary *es, struct can_frame *cf) @@ -1515,6 +1510,24 @@ error: return err; } +static void kvaser_usb_reset_tx_urb_contexts(struct kvaser_usb_net_priv *priv) +{ + int i; + + priv->active_tx_contexts = 0; + for (i = 0; i < MAX_TX_URBS; i++) + priv->tx_contexts[i].echo_index = MAX_TX_URBS; +} + +/* This method might sleep. Do not call it in the atomic context + * of URB completions. + */ +static void kvaser_usb_unlink_tx_urbs(struct kvaser_usb_net_priv *priv) +{ + usb_kill_anchored_urbs(&priv->tx_submitted); + kvaser_usb_reset_tx_urb_contexts(priv); +} + static void kvaser_usb_unlink_all_urbs(struct kvaser_usb *dev) { int i; @@ -1634,6 +1647,7 @@ static netdev_tx_t kvaser_usb_start_xmit(struct sk_buff *skb, struct kvaser_msg *msg; int i, err, ret = NETDEV_TX_OK; u8 *msg_tx_can_flags = NULL; /* GCC */ + unsigned long flags; if (can_dropped_invalid_skb(netdev, skb)) return NETDEV_TX_OK; @@ -1687,12 +1701,21 @@ static netdev_tx_t kvaser_usb_start_xmit(struct sk_buff *skb, if (cf->can_id & CAN_RTR_FLAG) *msg_tx_can_flags |= MSG_FLAG_REMOTE_FRAME; + spin_lock_irqsave(&priv->tx_contexts_lock, flags); for (i = 0; i < ARRAY_SIZE(priv->tx_contexts); i++) { if (priv->tx_contexts[i].echo_index == MAX_TX_URBS) { context = &priv->tx_contexts[i]; + + context->echo_index = i; + can_put_echo_skb(skb, netdev, context->echo_index); + ++priv->active_tx_contexts; + if (priv->active_tx_contexts >= MAX_TX_URBS) + netif_stop_queue(netdev); + break; } } + spin_unlock_irqrestore(&priv->tx_contexts_lock, flags); /* This should never happen; it implies a flow control bug */ if (!context) { @@ -1704,7 +1727,6 @@ static netdev_tx_t kvaser_usb_start_xmit(struct sk_buff *skb, } context->priv = priv; - context->echo_index = i; context->dlc = cf->can_dlc; msg->u.tx_can.tid = context->echo_index; @@ -1716,18 +1738,17 @@ static netdev_tx_t kvaser_usb_start_xmit(struct sk_buff *skb, kvaser_usb_write_bulk_callback, context); usb_anchor_urb(urb, &priv->tx_submitted); - can_put_echo_skb(skb, netdev, context->echo_index); - - atomic_inc(&priv->active_tx_urbs); - - if (atomic_read(&priv->active_tx_urbs) >= MAX_TX_URBS) - netif_stop_queue(netdev); - err = usb_submit_urb(urb, GFP_ATOMIC); if (unlikely(err)) { + spin_lock_irqsave(&priv->tx_contexts_lock, flags); + can_free_echo_skb(netdev, context->echo_index); + context->echo_index = MAX_TX_URBS; + --priv->active_tx_contexts; + netif_wake_queue(netdev); + + spin_unlock_irqrestore(&priv->tx_contexts_lock, flags); - atomic_dec(&priv->active_tx_urbs); usb_unanchor_urb(urb); stats->tx_dropped++; @@ -1854,7 +1875,7 @@ static int kvaser_usb_init_one(struct usb_interface *intf, struct kvaser_usb *dev = usb_get_intfdata(intf); struct net_device *netdev; struct kvaser_usb_net_priv *priv; - int i, err; + int err; err = kvaser_usb_send_simple_msg(dev, CMD_RESET_CHIP, channel); if (err) @@ -1868,19 +1889,17 @@ static int kvaser_usb_init_one(struct usb_interface *intf, priv = netdev_priv(netdev); + init_usb_anchor(&priv->tx_submitted); init_completion(&priv->start_comp); init_completion(&priv->stop_comp); - init_usb_anchor(&priv->tx_submitted); - atomic_set(&priv->active_tx_urbs, 0); - - for (i = 0; i < ARRAY_SIZE(priv->tx_contexts); i++) - priv->tx_contexts[i].echo_index = MAX_TX_URBS; - priv->dev = dev; priv->netdev = netdev; priv->channel = channel; + spin_lock_init(&priv->tx_contexts_lock); + kvaser_usb_reset_tx_urb_contexts(priv); + priv->can.state = CAN_STATE_STOPPED; priv->can.clock.freq = CAN_USB_CLOCK; priv->can.bittiming_const = &kvaser_usb_bittiming_const; -- cgit v1.2.1 From ae705930fca6322600690df9dc1c7d0516145a93 Mon Sep 17 00:00:00 2001 From: Christoffer Dall Date: Fri, 13 Mar 2015 17:02:56 +0000 Subject: arm/arm64: KVM: Keep elrsr/aisr in sync with software model MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit There is an interesting bug in the vgic code, which manifests itself when the KVM run loop has a signal pending or needs a vmid generation rollover after having disabled interrupts but before actually switching to the guest. In this case, we flush the vgic as usual, but we sync back the vgic state and exit to userspace before entering the guest. The consequence is that we will be syncing the list registers back to the software model using the GICH_ELRSR and GICH_EISR from the last execution of the guest, potentially overwriting a list register containing an interrupt. This showed up during migration testing where we would capture a state where the VM has masked the arch timer but there were no interrupts, resulting in a hung test. Cc: Marc Zyngier Reported-by: Alex Bennee Signed-off-by: Christoffer Dall Signed-off-by: Alex Bennée Acked-by: Marc Zyngier Signed-off-by: Christoffer Dall --- include/kvm/arm_vgic.h | 1 + virt/kvm/arm/vgic-v2.c | 8 ++++++++ virt/kvm/arm/vgic-v3.c | 8 ++++++++ virt/kvm/arm/vgic.c | 16 ++++++++++++++++ 4 files changed, 33 insertions(+) diff --git a/include/kvm/arm_vgic.h b/include/kvm/arm_vgic.h index 7c55dd5dd2c9..66203b268984 100644 --- a/include/kvm/arm_vgic.h +++ b/include/kvm/arm_vgic.h @@ -114,6 +114,7 @@ struct vgic_ops { void (*sync_lr_elrsr)(struct kvm_vcpu *, int, struct vgic_lr); u64 (*get_elrsr)(const struct kvm_vcpu *vcpu); u64 (*get_eisr)(const struct kvm_vcpu *vcpu); + void (*clear_eisr)(struct kvm_vcpu *vcpu); u32 (*get_interrupt_status)(const struct kvm_vcpu *vcpu); void (*enable_underflow)(struct kvm_vcpu *vcpu); void (*disable_underflow)(struct kvm_vcpu *vcpu); diff --git a/virt/kvm/arm/vgic-v2.c b/virt/kvm/arm/vgic-v2.c index a0a7b5d1a070..f9b9c7c51372 100644 --- a/virt/kvm/arm/vgic-v2.c +++ b/virt/kvm/arm/vgic-v2.c @@ -72,6 +72,8 @@ static void vgic_v2_sync_lr_elrsr(struct kvm_vcpu *vcpu, int lr, { if (!(lr_desc.state & LR_STATE_MASK)) vcpu->arch.vgic_cpu.vgic_v2.vgic_elrsr |= (1ULL << lr); + else + vcpu->arch.vgic_cpu.vgic_v2.vgic_elrsr &= ~(1ULL << lr); } static u64 vgic_v2_get_elrsr(const struct kvm_vcpu *vcpu) @@ -84,6 +86,11 @@ static u64 vgic_v2_get_eisr(const struct kvm_vcpu *vcpu) return vcpu->arch.vgic_cpu.vgic_v2.vgic_eisr; } +static void vgic_v2_clear_eisr(struct kvm_vcpu *vcpu) +{ + vcpu->arch.vgic_cpu.vgic_v2.vgic_eisr = 0; +} + static u32 vgic_v2_get_interrupt_status(const struct kvm_vcpu *vcpu) { u32 misr = vcpu->arch.vgic_cpu.vgic_v2.vgic_misr; @@ -148,6 +155,7 @@ static const struct vgic_ops vgic_v2_ops = { .sync_lr_elrsr = vgic_v2_sync_lr_elrsr, .get_elrsr = vgic_v2_get_elrsr, .get_eisr = vgic_v2_get_eisr, + .clear_eisr = vgic_v2_clear_eisr, .get_interrupt_status = vgic_v2_get_interrupt_status, .enable_underflow = vgic_v2_enable_underflow, .disable_underflow = vgic_v2_disable_underflow, diff --git a/virt/kvm/arm/vgic-v3.c b/virt/kvm/arm/vgic-v3.c index 3a62d8a9a2c6..dff06021e748 100644 --- a/virt/kvm/arm/vgic-v3.c +++ b/virt/kvm/arm/vgic-v3.c @@ -104,6 +104,8 @@ static void vgic_v3_sync_lr_elrsr(struct kvm_vcpu *vcpu, int lr, { if (!(lr_desc.state & LR_STATE_MASK)) vcpu->arch.vgic_cpu.vgic_v3.vgic_elrsr |= (1U << lr); + else + vcpu->arch.vgic_cpu.vgic_v3.vgic_elrsr &= ~(1U << lr); } static u64 vgic_v3_get_elrsr(const struct kvm_vcpu *vcpu) @@ -116,6 +118,11 @@ static u64 vgic_v3_get_eisr(const struct kvm_vcpu *vcpu) return vcpu->arch.vgic_cpu.vgic_v3.vgic_eisr; } +static void vgic_v3_clear_eisr(struct kvm_vcpu *vcpu) +{ + vcpu->arch.vgic_cpu.vgic_v3.vgic_eisr = 0; +} + static u32 vgic_v3_get_interrupt_status(const struct kvm_vcpu *vcpu) { u32 misr = vcpu->arch.vgic_cpu.vgic_v3.vgic_misr; @@ -192,6 +199,7 @@ static const struct vgic_ops vgic_v3_ops = { .sync_lr_elrsr = vgic_v3_sync_lr_elrsr, .get_elrsr = vgic_v3_get_elrsr, .get_eisr = vgic_v3_get_eisr, + .clear_eisr = vgic_v3_clear_eisr, .get_interrupt_status = vgic_v3_get_interrupt_status, .enable_underflow = vgic_v3_enable_underflow, .disable_underflow = vgic_v3_disable_underflow, diff --git a/virt/kvm/arm/vgic.c b/virt/kvm/arm/vgic.c index 4b2c2e7856a3..c9f60f524588 100644 --- a/virt/kvm/arm/vgic.c +++ b/virt/kvm/arm/vgic.c @@ -883,6 +883,11 @@ static inline u64 vgic_get_eisr(struct kvm_vcpu *vcpu) return vgic_ops->get_eisr(vcpu); } +static inline void vgic_clear_eisr(struct kvm_vcpu *vcpu) +{ + vgic_ops->clear_eisr(vcpu); +} + static inline u32 vgic_get_interrupt_status(struct kvm_vcpu *vcpu) { return vgic_ops->get_interrupt_status(vcpu); @@ -922,6 +927,7 @@ static void vgic_retire_lr(int lr_nr, int irq, struct kvm_vcpu *vcpu) vgic_set_lr(vcpu, lr_nr, vlr); clear_bit(lr_nr, vgic_cpu->lr_used); vgic_cpu->vgic_irq_lr_map[irq] = LR_EMPTY; + vgic_sync_lr_elrsr(vcpu, lr_nr, vlr); } /* @@ -978,6 +984,7 @@ bool vgic_queue_irq(struct kvm_vcpu *vcpu, u8 sgi_source_id, int irq) BUG_ON(!test_bit(lr, vgic_cpu->lr_used)); vlr.state |= LR_STATE_PENDING; vgic_set_lr(vcpu, lr, vlr); + vgic_sync_lr_elrsr(vcpu, lr, vlr); return true; } } @@ -999,6 +1006,7 @@ bool vgic_queue_irq(struct kvm_vcpu *vcpu, u8 sgi_source_id, int irq) vlr.state |= LR_EOI_INT; vgic_set_lr(vcpu, lr, vlr); + vgic_sync_lr_elrsr(vcpu, lr, vlr); return true; } @@ -1136,6 +1144,14 @@ static bool vgic_process_maintenance(struct kvm_vcpu *vcpu) if (status & INT_STATUS_UNDERFLOW) vgic_disable_underflow(vcpu); + /* + * In the next iterations of the vcpu loop, if we sync the vgic state + * after flushing it, but before entering the guest (this happens for + * pending signals and vmid rollovers), then make sure we don't pick + * up any old maintenance interrupts here. + */ + vgic_clear_eisr(vcpu); + return level_pending; } -- cgit v1.2.1 From b4331b433a2860389c9e35a523f0127788248500 Mon Sep 17 00:00:00 2001 From: Heiko Stuebner Date: Sat, 14 Mar 2015 19:32:06 +0100 Subject: MAINTAINERS: add rockchip regexp to the ARM/Rockchip entry The regexp option is a nice way to catch even weirder paths like the current drivers/gpu/drm/rockchip/* or others in the future. Signed-off-by: Heiko Stuebner --- MAINTAINERS | 1 + 1 file changed, 1 insertion(+) diff --git a/MAINTAINERS b/MAINTAINERS index ddc5a8cf9a8a..0d273bd264ac 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -1351,6 +1351,7 @@ F: drivers/i2c/busses/i2c-rk3x.c F: drivers/*/*rockchip* F: drivers/*/*/*rockchip* F: sound/soc/rockchip/ +N: rockchip ARM/SAMSUNG EXYNOS ARM ARCHITECTURES M: Kukjin Kim -- cgit v1.2.1 From 54b0bc602541fcc2dd9f2480623c00552e0b220e Mon Sep 17 00:00:00 2001 From: Alexandru M Stan Date: Fri, 13 Mar 2015 17:55:32 -0700 Subject: ARM: dts: rockchip: disable gmac by default in rk3288.dtsi This block should not be enabled by default or else if the kconfig is set, it will try to load/probe even if there's no phy connected. Signed-off-by: Alexandru M Stan Signed-off-by: Heiko Stuebner --- arch/arm/boot/dts/rk3288.dtsi | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/arm/boot/dts/rk3288.dtsi b/arch/arm/boot/dts/rk3288.dtsi index d771f687a13b..eccc78d3220b 100644 --- a/arch/arm/boot/dts/rk3288.dtsi +++ b/arch/arm/boot/dts/rk3288.dtsi @@ -411,6 +411,7 @@ "mac_clk_rx", "mac_clk_tx", "clk_mac_ref", "clk_mac_refout", "aclk_mac", "pclk_mac"; + status = "disabled"; }; usb_host0_ehci: usb@ff500000 { -- cgit v1.2.1 From 4c906c279886550d2aaac6facf71d709158e4e3c Mon Sep 17 00:00:00 2001 From: Venkat Venkatsubra Date: Fri, 13 Mar 2015 07:08:22 -0700 Subject: bridge: reset bridge mtu after deleting an interface On adding an interface br_add_if() sets the MTU to the min of all the interfaces. Do the same thing on removing an interface too in br_del_if. Signed-off-by: Venkat Venkatsubra Acked-by: Roopa Prabhu Signed-off-by: David S. Miller --- net/bridge/br_if.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/net/bridge/br_if.c b/net/bridge/br_if.c index b087d278c679..1849d96b3c91 100644 --- a/net/bridge/br_if.c +++ b/net/bridge/br_if.c @@ -563,6 +563,8 @@ int br_del_if(struct net_bridge *br, struct net_device *dev) */ del_nbp(p); + dev_set_mtu(br->dev, br_min_mtu(br)); + spin_lock_bh(&br->lock); changed_addr = br_stp_recalculate_bridge_id(br); spin_unlock_bh(&br->lock); -- cgit v1.2.1 From d20f7807996c69537e07443ef8dec4e01a28b099 Mon Sep 17 00:00:00 2001 From: Li Jun Date: Sun, 8 Mar 2015 16:05:01 +0800 Subject: usb: chipidea: otg: add a_alt_hnp_support response for B device This patch adds response to a_alt_hnp_support set feature request from legacy A device, that is, B-device can provide a message to the user indicating that the user needs to connect the B-device to an alternate port on the A-device. A device sets this feature indicates to the B-device that it is connected to an A-device port that is not capable of HNP, but that the A-device does have an alternate port that is capable of HNP. [Peter] Without this patch, the OTG B device can't be enumerated on non-HNP port at A device, see below log: [ 2.287464] usb 1-1: Dual-Role OTG device on non-HNP port [ 2.293105] usb 1-1: can't set HNP mode: -32 [ 2.417422] usb 1-1: new high-speed USB device number 4 using ci_hdrc [ 2.460635] usb 1-1: Dual-Role OTG device on non-HNP port [ 2.466424] usb 1-1: can't set HNP mode: -32 [ 2.587464] usb 1-1: new high-speed USB device number 5 using ci_hdrc [ 2.630649] usb 1-1: Dual-Role OTG device on non-HNP port [ 2.636436] usb 1-1: can't set HNP mode: -32 [ 2.641003] usb usb1-port1: unable to enumerate USB device Cc: stable Acked-by: Peter Chen Signed-off-by: Li Jun Signed-off-by: Peter Chen Signed-off-by: Greg Kroah-Hartman --- drivers/usb/chipidea/udc.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/drivers/usb/chipidea/udc.c b/drivers/usb/chipidea/udc.c index ff451048c1ac..4bfb7ac0239f 100644 --- a/drivers/usb/chipidea/udc.c +++ b/drivers/usb/chipidea/udc.c @@ -929,6 +929,13 @@ __acquires(hwep->lock) return retval; } +static int otg_a_alt_hnp_support(struct ci_hdrc *ci) +{ + dev_warn(&ci->gadget.dev, + "connect the device to an alternate port if you want HNP\n"); + return isr_setup_status_phase(ci); +} + /** * isr_setup_packet_handler: setup packet handler * @ci: UDC descriptor @@ -1061,6 +1068,10 @@ __acquires(ci->lock) ci); } break; + case USB_DEVICE_A_ALT_HNP_SUPPORT: + if (ci_otg_is_fsm_mode(ci)) + err = otg_a_alt_hnp_support(ci); + break; default: goto delegate; } -- cgit v1.2.1 From 963ffa3e97d4011cbc545d4236179f227e199f3f Mon Sep 17 00:00:00 2001 From: Peter Chen Date: Sun, 15 Feb 2015 12:22:59 +0800 Subject: MAINTAINERS: add entry for USB OTG FSM Add MAINTAINER entry for USB OTG Finite State Machine Cc: Felipe Balbi Signed-off-by: Peter Chen Signed-off-by: Greg Kroah-Hartman --- MAINTAINERS | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/MAINTAINERS b/MAINTAINERS index 6239a305dff0..43391d668bf5 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -10196,6 +10196,13 @@ S: Maintained F: Documentation/usb/ohci.txt F: drivers/usb/host/ohci* +USB OTG FSM (Finite State Machine) +M: Peter Chen +T: git git://github.com/hzpeterchen/linux-usb.git +L: linux-usb@vger.kernel.org +S: Maintained +F: drivers/usb/common/usb-otg-fsm.c + USB OVER IP DRIVER M: Valentina Manea M: Shuah Khan -- cgit v1.2.1 From 9b028649b9d0ae72090904629dad06b022f4ddc7 Mon Sep 17 00:00:00 2001 From: Forest Wilkinson Date: Thu, 12 Mar 2015 23:58:16 -0700 Subject: HID: tivo: enable all buttons on the TiVo Slide Pro remote The linux kernel has supported the TiVo Slide remote control for some time, but does not recognize the USB ID of the newer Slide Pro. This patch adds the missing data structures so the newer remote will be recognized by the driver, thereby allowing the TiVo, LiveTV, and Thumbs Up/Down buttons to be mapped with a hwdb file. Signed-off-by: Forest Wilkinson Signed-off-by: Jiri Kosina --- drivers/hid/hid-core.c | 1 + drivers/hid/hid-ids.h | 1 + drivers/hid/hid-tivo.c | 1 + 3 files changed, 3 insertions(+) diff --git a/drivers/hid/hid-core.c b/drivers/hid/hid-core.c index 7c669c328c4c..56ce8c2b5530 100644 --- a/drivers/hid/hid-core.c +++ b/drivers/hid/hid-core.c @@ -1959,6 +1959,7 @@ static const struct hid_device_id hid_have_special_driver[] = { { HID_USB_DEVICE(USB_VENDOR_ID_THRUSTMASTER, 0xb65a) }, { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_TIVO, USB_DEVICE_ID_TIVO_SLIDE_BT) }, { HID_USB_DEVICE(USB_VENDOR_ID_TIVO, USB_DEVICE_ID_TIVO_SLIDE) }, + { HID_USB_DEVICE(USB_VENDOR_ID_TIVO, USB_DEVICE_ID_TIVO_SLIDE_PRO) }, { HID_USB_DEVICE(USB_VENDOR_ID_TOPSEED, USB_DEVICE_ID_TOPSEED_CYBERLINK) }, { HID_USB_DEVICE(USB_VENDOR_ID_TOPSEED2, USB_DEVICE_ID_TOPSEED2_RF_COMBO) }, { HID_USB_DEVICE(USB_VENDOR_ID_TWINHAN, USB_DEVICE_ID_TWINHAN_IR_REMOTE) }, diff --git a/drivers/hid/hid-ids.h b/drivers/hid/hid-ids.h index 8a7c347e8080..9c4786759f16 100644 --- a/drivers/hid/hid-ids.h +++ b/drivers/hid/hid-ids.h @@ -899,6 +899,7 @@ #define USB_VENDOR_ID_TIVO 0x150a #define USB_DEVICE_ID_TIVO_SLIDE_BT 0x1200 #define USB_DEVICE_ID_TIVO_SLIDE 0x1201 +#define USB_DEVICE_ID_TIVO_SLIDE_PRO 0x1203 #define USB_VENDOR_ID_TOPSEED 0x0766 #define USB_DEVICE_ID_TOPSEED_CYBERLINK 0x0204 diff --git a/drivers/hid/hid-tivo.c b/drivers/hid/hid-tivo.c index d790d8d71f7f..d98696927453 100644 --- a/drivers/hid/hid-tivo.c +++ b/drivers/hid/hid-tivo.c @@ -64,6 +64,7 @@ static const struct hid_device_id tivo_devices[] = { /* TiVo Slide Bluetooth remote, pairs with a Broadcom dongle */ { HID_BLUETOOTH_DEVICE(USB_VENDOR_ID_TIVO, USB_DEVICE_ID_TIVO_SLIDE_BT) }, { HID_USB_DEVICE(USB_VENDOR_ID_TIVO, USB_DEVICE_ID_TIVO_SLIDE) }, + { HID_USB_DEVICE(USB_VENDOR_ID_TIVO, USB_DEVICE_ID_TIVO_SLIDE_PRO) }, { } }; MODULE_DEVICE_TABLE(hid, tivo_devices); -- cgit v1.2.1 From 3eeff778e00c956875c70b145c52638c313dfb23 Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 14 Mar 2015 05:22:21 +0000 Subject: caif: fix MSG_OOB test in caif_seqpkt_recvmsg() It should be checking flags, not msg->msg_flags. It's ->sendmsg() instances that need to look for that in ->msg_flags, ->recvmsg() ones (including the other ->recvmsg() instance in that file, as well as unix_dgram_recvmsg() this one claims to be imitating) check in flags. Braino had been introduced in commit dcda13 ("caif: Bugfix - use MSG_TRUNC in receive") back in 2010, so it goes quite a while back. Signed-off-by: Al Viro Signed-off-by: David S. Miller --- net/caif/caif_socket.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/caif/caif_socket.c b/net/caif/caif_socket.c index 769b185fefbd..a6e2da0bc718 100644 --- a/net/caif/caif_socket.c +++ b/net/caif/caif_socket.c @@ -281,7 +281,7 @@ static int caif_seqpkt_recvmsg(struct kiocb *iocb, struct socket *sock, int copylen; ret = -EOPNOTSUPP; - if (m->msg_flags&MSG_OOB) + if (flags & MSG_OOB) goto read_error; skb = skb_recv_datagram(sk, flags, 0 , &ret); -- cgit v1.2.1 From 7d985ed1dca5c90535d67ce92ef6ca520302340a Mon Sep 17 00:00:00 2001 From: Al Viro Date: Sat, 14 Mar 2015 05:34:56 +0000 Subject: rxrpc: bogus MSG_PEEK test in rxrpc_recvmsg() [I would really like an ACK on that one from dhowells; it appears to be quite straightforward, but...] MSG_PEEK isn't passed to ->recvmsg() via msg->msg_flags; as the matter of fact, neither the kernel users of rxrpc, nor the syscalls ever set that bit in there. It gets passed via flags; in fact, another such check in the same function is done correctly - as flags & MSG_PEEK. It had been that way (effectively disabled) for 8 years, though, so the patch needs beating up - that case had never been tested. If it is correct, it's -stable fodder. Signed-off-by: Al Viro Signed-off-by: David S. Miller --- net/rxrpc/ar-recvmsg.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/rxrpc/ar-recvmsg.c b/net/rxrpc/ar-recvmsg.c index 4575485ad1b4..19a560626dc4 100644 --- a/net/rxrpc/ar-recvmsg.c +++ b/net/rxrpc/ar-recvmsg.c @@ -87,7 +87,7 @@ int rxrpc_recvmsg(struct kiocb *iocb, struct socket *sock, if (!skb) { /* nothing remains on the queue */ if (copied && - (msg->msg_flags & MSG_PEEK || timeo == 0)) + (flags & MSG_PEEK || timeo == 0)) goto out; /* wait for a message to turn up */ -- cgit v1.2.1 From c105e86ace5a32ee4760a502bc45dcd26fed2375 Mon Sep 17 00:00:00 2001 From: Tobias Klauser Date: Fri, 13 Mar 2015 16:40:06 +0800 Subject: nios2: Remove ucontext.h from exported arch headers Commit 92d5dd8cd6e2 ("nios2: update pt_regs") removed the nios2 specific ucontext.h, replacing it with the version from asm-generic. Thus it's no longer necessary to include ucontext.h in exported headers. Cc: Chung-Ling Tang Signed-off-by: Tobias Klauser Acked-by: Ley Foon Tan --- arch/nios2/include/uapi/asm/Kbuild | 1 - 1 file changed, 1 deletion(-) diff --git a/arch/nios2/include/uapi/asm/Kbuild b/arch/nios2/include/uapi/asm/Kbuild index 376131194cc3..e0bb972a50d7 100644 --- a/arch/nios2/include/uapi/asm/Kbuild +++ b/arch/nios2/include/uapi/asm/Kbuild @@ -1,6 +1,5 @@ include include/uapi/asm-generic/Kbuild.asm header-y += elf.h -header-y += ucontext.h generic-y += ucontext.h -- cgit v1.2.1 From 10640d34552ccd8fabe7b15b0c4e3a102247952d Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Sun, 15 Mar 2015 13:48:03 +0300 Subject: isdn: icn: use strlcpy() when parsing setup options If you pass an invalid string here then you probably deserve the memory corruption, but it annoys static analysis tools so lets fix it. Signed-off-by: Dan Carpenter Signed-off-by: David S. Miller --- drivers/isdn/icn/icn.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/isdn/icn/icn.c b/drivers/isdn/icn/icn.c index 6a7447c304ac..358a574d9e8b 100644 --- a/drivers/isdn/icn/icn.c +++ b/drivers/isdn/icn/icn.c @@ -1609,7 +1609,7 @@ icn_setup(char *line) if (ints[0] > 1) membase = (unsigned long)ints[2]; if (str && *str) { - strcpy(sid, str); + strlcpy(sid, str, sizeof(sid)); icn_id = sid; if ((p = strchr(sid, ','))) { *p++ = 0; -- cgit v1.2.1 From 6347e2a10f7031dc3725e6f4519089517c0ca521 Mon Sep 17 00:00:00 2001 From: Ley Foon Tan Date: Mon, 16 Mar 2015 15:35:25 +0800 Subject: nios2: mm: do not invoke OOM killer on kernel fault OOM Follow commit 871341023c771ad. Kernel faults are expected to handle OOM conditions gracefully (gup, uaccess etc.), so they should never invoke the OOM killer. Reserve this for faults triggered in user context when it is the only option. Signed-off-by: Ley Foon Tan --- arch/nios2/mm/fault.c | 6 ------ 1 file changed, 6 deletions(-) diff --git a/arch/nios2/mm/fault.c b/arch/nios2/mm/fault.c index 0d231adfe576..0c9b6afe69e9 100644 --- a/arch/nios2/mm/fault.c +++ b/arch/nios2/mm/fault.c @@ -126,7 +126,6 @@ good_area: break; } -survive: /* * If for any reason at all we couldn't handle the fault, * make sure we exit gracefully rather than endlessly redo @@ -220,11 +219,6 @@ no_context: */ out_of_memory: up_read(&mm->mmap_sem); - if (is_global_init(tsk)) { - yield(); - down_read(&mm->mmap_sem); - goto survive; - } if (!user_mode(regs)) goto no_context; pagefault_out_of_memory(); -- cgit v1.2.1 From 0f611d28fc2e13cfec64e1c544c16a086886805a Mon Sep 17 00:00:00 2001 From: Andrei Otcheretianski Date: Thu, 12 Mar 2015 08:53:30 +0200 Subject: mac80211: count interfaces correctly for combination checks Since moving the interface combination checks to mac80211, it's broken because it now only considers interfaces with an assigned channel context, so for example any interface that isn't active can still be up, which is clearly an issue; also, in particular P2P-Device wdevs are an issue since they never have a chanctx. Fix this by counting running interfaces instead the ones with a channel context assigned. Cc: stable@vger.kernel.org [3.16+] Fixes: 73de86a38962b ("cfg80211/mac80211: move interface counting for combination check to mac80211") Signed-off-by: Andrei Otcheretianski Signed-off-by: Emmanuel Grumbach [rewrite commit message, dig out the commit it fixes] Signed-off-by: Johannes Berg --- net/mac80211/util.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/mac80211/util.c b/net/mac80211/util.c index 8428f4a95479..747bdcf72e92 100644 --- a/net/mac80211/util.c +++ b/net/mac80211/util.c @@ -3178,7 +3178,7 @@ int ieee80211_check_combinations(struct ieee80211_sub_if_data *sdata, wdev_iter = &sdata_iter->wdev; if (sdata_iter == sdata || - rcu_access_pointer(sdata_iter->vif.chanctx_conf) == NULL || + !ieee80211_sdata_running(sdata_iter) || local->hw.wiphy->software_iftypes & BIT(wdev_iter->iftype)) continue; -- cgit v1.2.1 From 70a3fd6c61c46c07c63cab935dca9a17d8de1709 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 12 Mar 2015 08:53:29 +0200 Subject: mac80211: ask for ECSA IE to be considered for beacon parse CRC When a beacon from the AP contains only the ECSA IE, and not a CSA IE as well, this ECSA IE is not considered for calculating the CRC and the beacon might be dropped as not being interesting. This is clearly wrong, it should be handled and the channel switch should be executed. Fix this by including the ECSA IE ID in the bitmap of interesting IEs. Reported-by: Gil Tribush Reviewed-by: Luciano Coelho Signed-off-by: Johannes Berg --- net/mac80211/mlme.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index 10ac6324c1d0..cde8cd3d6595 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -3204,7 +3204,8 @@ static const u64 care_about_ies = (1ULL << WLAN_EID_CHANNEL_SWITCH) | (1ULL << WLAN_EID_PWR_CONSTRAINT) | (1ULL << WLAN_EID_HT_CAPABILITY) | - (1ULL << WLAN_EID_HT_OPERATION); + (1ULL << WLAN_EID_HT_OPERATION) | + (1ULL << WLAN_EID_EXT_CHANSWITCH_ANN); static void ieee80211_rx_mgmt_beacon(struct ieee80211_sub_if_data *sdata, struct ieee80211_mgmt *mgmt, size_t len, -- cgit v1.2.1 From 496fcc294daab18799e190c0264863d653588d1f Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 12 Mar 2015 08:53:27 +0200 Subject: nl80211: ignore HT/VHT capabilities without QoS/WMM As HT/VHT depend heavily on QoS/WMM, it's not a good idea to let userspace add clients that have HT/VHT but not QoS/WMM. Since it does so in certain cases we've observed (client is using HT IEs but not QoS/WMM) just ignore the HT/VHT info at this point and don't pass it down to the drivers which might unconditionally use it. Cc: stable@vger.kernel.org Signed-off-by: Johannes Berg --- net/wireless/nl80211.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index be2501538011..b6f84f6a2a09 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -4400,6 +4400,16 @@ static int nl80211_new_station(struct sk_buff *skb, struct genl_info *info) if (parse_station_flags(info, dev->ieee80211_ptr->iftype, ¶ms)) return -EINVAL; + /* HT/VHT requires QoS, but if we don't have that just ignore HT/VHT + * as userspace might just pass through the capabilities from the IEs + * directly, rather than enforcing this restriction and returning an + * error in this case. + */ + if (!(params.sta_flags_set & BIT(NL80211_STA_FLAG_WME))) { + params.ht_capa = NULL; + params.vht_capa = NULL; + } + /* When you run into this, adjust the code below for the new flag */ BUILD_BUG_ON(NL80211_STA_FLAG_MAX != 7); -- cgit v1.2.1 From f84eaa1068315409ffbef57e6fea312180787db3 Mon Sep 17 00:00:00 2001 From: Johannes Berg Date: Thu, 12 Mar 2015 08:53:26 +0200 Subject: mac80211: ignore CSA to same channel If the AP is confused and starts doing a CSA to the same channel, just ignore that request instead of trying to act it out since it was likely sent in error anyway. In the case of the bug I was investigating the GO was misbehaving and sending out a beacon with CSA IEs still included after having actually done the channel switch. Signed-off-by: Johannes Berg --- net/mac80211/ieee80211_i.h | 1 + net/mac80211/mlme.c | 13 +++++++++++++ 2 files changed, 14 insertions(+) diff --git a/net/mac80211/ieee80211_i.h b/net/mac80211/ieee80211_i.h index c0e089c194f1..8d53d65bd2ab 100644 --- a/net/mac80211/ieee80211_i.h +++ b/net/mac80211/ieee80211_i.h @@ -464,6 +464,7 @@ struct ieee80211_if_managed { unsigned int flags; bool csa_waiting_bcn; + bool csa_ignored_same_chan; bool beacon_crc_valid; u32 beacon_crc; diff --git a/net/mac80211/mlme.c b/net/mac80211/mlme.c index cde8cd3d6595..142f66aece18 100644 --- a/net/mac80211/mlme.c +++ b/net/mac80211/mlme.c @@ -1150,6 +1150,17 @@ ieee80211_sta_process_chanswitch(struct ieee80211_sub_if_data *sdata, return; } + if (cfg80211_chandef_identical(&csa_ie.chandef, + &sdata->vif.bss_conf.chandef)) { + if (ifmgd->csa_ignored_same_chan) + return; + sdata_info(sdata, + "AP %pM tries to chanswitch to same channel, ignore\n", + ifmgd->associated->bssid); + ifmgd->csa_ignored_same_chan = true; + return; + } + mutex_lock(&local->mtx); mutex_lock(&local->chanctx_mtx); conf = rcu_dereference_protected(sdata->vif.chanctx_conf, @@ -1210,6 +1221,7 @@ ieee80211_sta_process_chanswitch(struct ieee80211_sub_if_data *sdata, sdata->vif.csa_active = true; sdata->csa_chandef = csa_ie.chandef; sdata->csa_block_tx = csa_ie.mode; + ifmgd->csa_ignored_same_chan = false; if (sdata->csa_block_tx) ieee80211_stop_vif_queues(local, sdata, @@ -2090,6 +2102,7 @@ static void ieee80211_set_disassoc(struct ieee80211_sub_if_data *sdata, sdata->vif.csa_active = false; ifmgd->csa_waiting_bcn = false; + ifmgd->csa_ignored_same_chan = false; if (sdata->csa_block_tx) { ieee80211_wake_vif_queues(local, sdata, IEEE80211_QUEUE_STOP_REASON_CSA); -- cgit v1.2.1 From 69797dafe35541bfff1989c0b37c66ed785faf0e Mon Sep 17 00:00:00 2001 From: Borislav Petkov Date: Mon, 16 Mar 2015 11:06:28 +0100 Subject: Revert "x86/mm/ASLR: Propagate base load address calculation" This reverts commit: f47233c2d34f ("x86/mm/ASLR: Propagate base load address calculation") The main reason for the revert is that the new boot flag does not work at all currently, and in order to make this work, we need non-trivial changes to the x86 boot code which we didn't manage to get done in time for merging. And even if we did, they would've been too risky so instead of rushing things and break booting 4.1 on boxes left and right, we will be very strict and conservative and will take our time with this to fix and test it properly. Reported-by: Yinghai Lu Signed-off-by: Borislav Petkov Cc: Ard Biesheuvel Cc: Baoquan He Cc: H. Peter Anvin Cc: Josh Triplett Cc: Junjie Mao Cc: Kees Cook Cc: Linus Torvalds Cc: Matt Fleming Link: http://lkml.kernel.org/r/20150316100628.GD22995@pd.tnic Signed-off-by: Ingo Molnar --- arch/x86/boot/compressed/aslr.c | 34 +--------------------------------- arch/x86/boot/compressed/misc.c | 3 +-- arch/x86/boot/compressed/misc.h | 6 ++---- arch/x86/include/asm/page_types.h | 2 -- arch/x86/include/uapi/asm/bootparam.h | 1 - arch/x86/kernel/module.c | 10 +++++++++- arch/x86/kernel/setup.c | 22 ++++------------------ 7 files changed, 17 insertions(+), 61 deletions(-) diff --git a/arch/x86/boot/compressed/aslr.c b/arch/x86/boot/compressed/aslr.c index 7083c16cccba..bb1376381985 100644 --- a/arch/x86/boot/compressed/aslr.c +++ b/arch/x86/boot/compressed/aslr.c @@ -14,13 +14,6 @@ static const char build_str[] = UTS_RELEASE " (" LINUX_COMPILE_BY "@" LINUX_COMPILE_HOST ") (" LINUX_COMPILER ") " UTS_VERSION; -struct kaslr_setup_data { - __u64 next; - __u32 type; - __u32 len; - __u8 data[1]; -} kaslr_setup_data; - #define I8254_PORT_CONTROL 0x43 #define I8254_PORT_COUNTER0 0x40 #define I8254_CMD_READBACK 0xC0 @@ -302,29 +295,7 @@ static unsigned long find_random_addr(unsigned long minimum, return slots_fetch_random(); } -static void add_kaslr_setup_data(struct boot_params *params, __u8 enabled) -{ - struct setup_data *data; - - kaslr_setup_data.type = SETUP_KASLR; - kaslr_setup_data.len = 1; - kaslr_setup_data.next = 0; - kaslr_setup_data.data[0] = enabled; - - data = (struct setup_data *)(unsigned long)params->hdr.setup_data; - - while (data && data->next) - data = (struct setup_data *)(unsigned long)data->next; - - if (data) - data->next = (unsigned long)&kaslr_setup_data; - else - params->hdr.setup_data = (unsigned long)&kaslr_setup_data; - -} - -unsigned char *choose_kernel_location(struct boot_params *params, - unsigned char *input, +unsigned char *choose_kernel_location(unsigned char *input, unsigned long input_size, unsigned char *output, unsigned long output_size) @@ -335,17 +306,14 @@ unsigned char *choose_kernel_location(struct boot_params *params, #ifdef CONFIG_HIBERNATION if (!cmdline_find_option_bool("kaslr")) { debug_putstr("KASLR disabled by default...\n"); - add_kaslr_setup_data(params, 0); goto out; } #else if (cmdline_find_option_bool("nokaslr")) { debug_putstr("KASLR disabled by cmdline...\n"); - add_kaslr_setup_data(params, 0); goto out; } #endif - add_kaslr_setup_data(params, 1); /* Record the various known unsafe memory ranges. */ mem_avoid_init((unsigned long)input, input_size, diff --git a/arch/x86/boot/compressed/misc.c b/arch/x86/boot/compressed/misc.c index 5903089c818f..a950864a64da 100644 --- a/arch/x86/boot/compressed/misc.c +++ b/arch/x86/boot/compressed/misc.c @@ -401,8 +401,7 @@ asmlinkage __visible void *decompress_kernel(void *rmode, memptr heap, * the entire decompressed kernel plus relocation table, or the * entire decompressed kernel plus .bss and .brk sections. */ - output = choose_kernel_location(real_mode, input_data, input_len, - output, + output = choose_kernel_location(input_data, input_len, output, output_len > run_size ? output_len : run_size); diff --git a/arch/x86/boot/compressed/misc.h b/arch/x86/boot/compressed/misc.h index ee3576b2666b..04477d68403f 100644 --- a/arch/x86/boot/compressed/misc.h +++ b/arch/x86/boot/compressed/misc.h @@ -57,8 +57,7 @@ int cmdline_find_option_bool(const char *option); #if CONFIG_RANDOMIZE_BASE /* aslr.c */ -unsigned char *choose_kernel_location(struct boot_params *params, - unsigned char *input, +unsigned char *choose_kernel_location(unsigned char *input, unsigned long input_size, unsigned char *output, unsigned long output_size); @@ -66,8 +65,7 @@ unsigned char *choose_kernel_location(struct boot_params *params, bool has_cpuflag(int flag); #else static inline -unsigned char *choose_kernel_location(struct boot_params *params, - unsigned char *input, +unsigned char *choose_kernel_location(unsigned char *input, unsigned long input_size, unsigned char *output, unsigned long output_size) diff --git a/arch/x86/include/asm/page_types.h b/arch/x86/include/asm/page_types.h index 95e11f79f123..f97fbe3abb67 100644 --- a/arch/x86/include/asm/page_types.h +++ b/arch/x86/include/asm/page_types.h @@ -51,8 +51,6 @@ extern int devmem_is_allowed(unsigned long pagenr); extern unsigned long max_low_pfn_mapped; extern unsigned long max_pfn_mapped; -extern bool kaslr_enabled; - static inline phys_addr_t get_max_mapped(void) { return (phys_addr_t)max_pfn_mapped << PAGE_SHIFT; diff --git a/arch/x86/include/uapi/asm/bootparam.h b/arch/x86/include/uapi/asm/bootparam.h index 44e6dd7e36a2..225b0988043a 100644 --- a/arch/x86/include/uapi/asm/bootparam.h +++ b/arch/x86/include/uapi/asm/bootparam.h @@ -7,7 +7,6 @@ #define SETUP_DTB 2 #define SETUP_PCI 3 #define SETUP_EFI 4 -#define SETUP_KASLR 5 /* ram_size flags */ #define RAMDISK_IMAGE_START_MASK 0x07FF diff --git a/arch/x86/kernel/module.c b/arch/x86/kernel/module.c index 9bbb9b35c144..d1ac80b72c72 100644 --- a/arch/x86/kernel/module.c +++ b/arch/x86/kernel/module.c @@ -47,13 +47,21 @@ do { \ #ifdef CONFIG_RANDOMIZE_BASE static unsigned long module_load_offset; +static int randomize_modules = 1; /* Mutex protects the module_load_offset. */ static DEFINE_MUTEX(module_kaslr_mutex); +static int __init parse_nokaslr(char *p) +{ + randomize_modules = 0; + return 0; +} +early_param("nokaslr", parse_nokaslr); + static unsigned long int get_module_load_offset(void) { - if (kaslr_enabled) { + if (randomize_modules) { mutex_lock(&module_kaslr_mutex); /* * Calculate the module_load_offset the first time this diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c index 98dc9317286e..0a2421cca01f 100644 --- a/arch/x86/kernel/setup.c +++ b/arch/x86/kernel/setup.c @@ -122,8 +122,6 @@ unsigned long max_low_pfn_mapped; unsigned long max_pfn_mapped; -bool __read_mostly kaslr_enabled = false; - #ifdef CONFIG_DMI RESERVE_BRK(dmi_alloc, 65536); #endif @@ -427,11 +425,6 @@ static void __init reserve_initrd(void) } #endif /* CONFIG_BLK_DEV_INITRD */ -static void __init parse_kaslr_setup(u64 pa_data, u32 data_len) -{ - kaslr_enabled = (bool)(pa_data + sizeof(struct setup_data)); -} - static void __init parse_setup_data(void) { struct setup_data *data; @@ -457,9 +450,6 @@ static void __init parse_setup_data(void) case SETUP_EFI: parse_efi_setup(pa_data, data_len); break; - case SETUP_KASLR: - parse_kaslr_setup(pa_data, data_len); - break; default: break; } @@ -842,14 +832,10 @@ static void __init trim_low_memory_range(void) static int dump_kernel_offset(struct notifier_block *self, unsigned long v, void *p) { - if (kaslr_enabled) - pr_emerg("Kernel Offset: 0x%lx from 0x%lx (relocation range: 0x%lx-0x%lx)\n", - (unsigned long)&_text - __START_KERNEL, - __START_KERNEL, - __START_KERNEL_map, - MODULES_VADDR-1); - else - pr_emerg("Kernel Offset: disabled\n"); + pr_emerg("Kernel Offset: 0x%lx from 0x%lx " + "(relocation range: 0x%lx-0x%lx)\n", + (unsigned long)&_text - __START_KERNEL, __START_KERNEL, + __START_KERNEL_map, MODULES_VADDR-1); return 0; } -- cgit v1.2.1 From 855832e47c1e51db701786ed76f8a9fec323aad6 Mon Sep 17 00:00:00 2001 From: Robin Gong Date: Sun, 15 Feb 2015 10:00:35 +0800 Subject: dmaengine: imx-sdma: switch to dynamic context mode after script loaded Below comments got from Page4724 of Reference Manual of i.mx6q: http://cache.freescale.com/files/32bit/doc/ref_manual/IMX6DQRM.pdf --"Static context mode should be used for the first channel called after reset to ensure that the all context RAM for that channel is initialized during the context SAVE phase when the channel is done or yields. Subsequent calls to the same channel or different channels may use any of the dynamic context modes. This will ensure that all context locations for the bootload channel are initialized, and prevent undefined values in context RAM from being loaded during the context restore if the channel is re-started later" Unfortunately, the rule was broken by commit(5b28aa319bba96987316425a1131813d87cbab35) .This patch just take them back. Signed-off-by: Robin Gong Signed-off-by: Vinod Koul --- drivers/dma/imx-sdma.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/drivers/dma/imx-sdma.c b/drivers/dma/imx-sdma.c index 18c0a131e4e4..66a0efb9651d 100644 --- a/drivers/dma/imx-sdma.c +++ b/drivers/dma/imx-sdma.c @@ -531,6 +531,10 @@ static int sdma_run_channel0(struct sdma_engine *sdma) dev_err(sdma->dev, "Timeout waiting for CH0 ready\n"); } + /* Set bits of CONFIG register with dynamic context switching */ + if (readl(sdma->regs + SDMA_H_CONFIG) == 0) + writel_relaxed(SDMA_H_CONFIG_CSM, sdma->regs + SDMA_H_CONFIG); + return ret ? 0 : -ETIMEDOUT; } @@ -1394,9 +1398,6 @@ static int sdma_init(struct sdma_engine *sdma) writel_relaxed(ccb_phys, sdma->regs + SDMA_H_C0PTR); - /* Set bits of CONFIG register with given context switching mode */ - writel_relaxed(SDMA_H_CONFIG_CSM, sdma->regs + SDMA_H_CONFIG); - /* Initializes channel's priorities */ sdma_set_channel_priority(&sdma->channel[0], 7); -- cgit v1.2.1 From d16da513c9c8f394216b8dd7c258e667b2c43c74 Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Sun, 15 Mar 2015 14:03:50 +0100 Subject: regulator: tps65910: Add missing #include MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit drivers/regulator/tps65910-regulator.c: In function ‘tps65910_parse_dt_reg_data’: drivers/regulator/tps65910-regulator.c:1018: error: implicit declaration of function ‘of_get_child_by_name’ drivers/regulator/tps65910-regulator.c:1018: warning: assignment makes pointer from integer without a cast drivers/regulator/tps65910-regulator.c:1034: error: implicit declaration of function ‘of_node_put’ drivers/regulator/tps65910-regulator.c:1056: error: implicit declaration of function ‘of_property_read_u32’ Signed-off-by: Geert Uytterhoeven Signed-off-by: Mark Brown --- drivers/regulator/tps65910-regulator.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/regulator/tps65910-regulator.c b/drivers/regulator/tps65910-regulator.c index e2cffe01b807..fb991ec76423 100644 --- a/drivers/regulator/tps65910-regulator.c +++ b/drivers/regulator/tps65910-regulator.c @@ -17,6 +17,7 @@ #include #include #include +#include #include #include #include -- cgit v1.2.1 From cc261738add93947d138d2fabad9f4dbed4e5c00 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Mon, 16 Mar 2015 10:18:08 +0100 Subject: ALSA: hda - Treat stereo-to-mono mix properly The commit [ef403edb7558: ALSA: hda - Don't access stereo amps for mono channel widgets] fixed the handling of mono widgets in general, but it still misses an exceptional case: namely, a mono mixer widget taking a single stereo input. In this case, it has stereo volumes although it's a mono widget, and thus we have to take care of both left and right input channels, as stated in HD-audio spec ("7.1.3 Widget Interconnection Rules"). This patch covers this missing piece by adding proper checks of stereo amps in both the generic parser and the proc output codes. Reported-by: Raymond Yau Cc: Signed-off-by: Takashi Iwai --- sound/pci/hda/hda_generic.c | 21 +++++++++++++++++++-- sound/pci/hda/hda_proc.c | 38 ++++++++++++++++++++++++++++++-------- 2 files changed, 49 insertions(+), 10 deletions(-) diff --git a/sound/pci/hda/hda_generic.c b/sound/pci/hda/hda_generic.c index fe18071bf93a..8ec5289f8e05 100644 --- a/sound/pci/hda/hda_generic.c +++ b/sound/pci/hda/hda_generic.c @@ -687,13 +687,30 @@ static int get_amp_val_to_activate(struct hda_codec *codec, hda_nid_t nid, return val; } +/* is this a stereo widget or a stereo-to-mono mix? */ +static bool is_stereo_amps(struct hda_codec *codec, hda_nid_t nid, int dir) +{ + unsigned int wcaps = get_wcaps(codec, nid); + hda_nid_t conn; + + if (wcaps & AC_WCAP_STEREO) + return true; + if (dir != HDA_INPUT || get_wcaps_type(wcaps) != AC_WID_AUD_MIX) + return false; + if (snd_hda_get_num_conns(codec, nid) != 1) + return false; + if (snd_hda_get_connections(codec, nid, &conn, 1) < 0) + return false; + return !!(get_wcaps(codec, conn) & AC_WCAP_STEREO); +} + /* initialize the amp value (only at the first time) */ static void init_amp(struct hda_codec *codec, hda_nid_t nid, int dir, int idx) { unsigned int caps = query_amp_caps(codec, nid, dir); int val = get_amp_val_to_activate(codec, nid, dir, caps, false); - if (get_wcaps(codec, nid) & AC_WCAP_STEREO) + if (is_stereo_amps(codec, nid, dir)) snd_hda_codec_amp_init_stereo(codec, nid, dir, idx, 0xff, val); else snd_hda_codec_amp_init(codec, nid, 0, dir, idx, 0xff, val); @@ -703,7 +720,7 @@ static void init_amp(struct hda_codec *codec, hda_nid_t nid, int dir, int idx) static int update_amp(struct hda_codec *codec, hda_nid_t nid, int dir, int idx, unsigned int mask, unsigned int val) { - if (get_wcaps(codec, nid) & AC_WCAP_STEREO) + if (is_stereo_amps(codec, nid, dir)) return snd_hda_codec_amp_stereo(codec, nid, dir, idx, mask, val); else diff --git a/sound/pci/hda/hda_proc.c b/sound/pci/hda/hda_proc.c index ce5a6da83419..05e19f78b4cb 100644 --- a/sound/pci/hda/hda_proc.c +++ b/sound/pci/hda/hda_proc.c @@ -134,13 +134,38 @@ static void print_amp_caps(struct snd_info_buffer *buffer, (caps & AC_AMPCAP_MUTE) >> AC_AMPCAP_MUTE_SHIFT); } +/* is this a stereo widget or a stereo-to-mono mix? */ +static bool is_stereo_amps(struct hda_codec *codec, hda_nid_t nid, + int dir, unsigned int wcaps, int indices) +{ + hda_nid_t conn; + + if (wcaps & AC_WCAP_STEREO) + return true; + /* check for a stereo-to-mono mix; it must be: + * only a single connection, only for input, and only a mixer widget + */ + if (indices != 1 || dir != HDA_INPUT || + get_wcaps_type(wcaps) != AC_WID_AUD_MIX) + return false; + + if (snd_hda_get_raw_connections(codec, nid, &conn, 1) < 0) + return false; + /* the connection source is a stereo? */ + wcaps = snd_hda_param_read(codec, conn, AC_PAR_AUDIO_WIDGET_CAP); + return !!(wcaps & AC_WCAP_STEREO); +} + static void print_amp_vals(struct snd_info_buffer *buffer, struct hda_codec *codec, hda_nid_t nid, - int dir, int stereo, int indices) + int dir, unsigned int wcaps, int indices) { unsigned int val; + bool stereo; int i; + stereo = is_stereo_amps(codec, nid, dir, wcaps, indices); + dir = dir == HDA_OUTPUT ? AC_AMP_GET_OUTPUT : AC_AMP_GET_INPUT; for (i = 0; i < indices; i++) { snd_iprintf(buffer, " ["); @@ -757,12 +782,10 @@ static void print_codec_info(struct snd_info_entry *entry, (codec->single_adc_amp && wid_type == AC_WID_AUD_IN)) print_amp_vals(buffer, codec, nid, HDA_INPUT, - wid_caps & AC_WCAP_STEREO, - 1); + wid_caps, 1); else print_amp_vals(buffer, codec, nid, HDA_INPUT, - wid_caps & AC_WCAP_STEREO, - conn_len); + wid_caps, conn_len); } if (wid_caps & AC_WCAP_OUT_AMP) { snd_iprintf(buffer, " Amp-Out caps: "); @@ -771,11 +794,10 @@ static void print_codec_info(struct snd_info_entry *entry, if (wid_type == AC_WID_PIN && codec->pin_amp_workaround) print_amp_vals(buffer, codec, nid, HDA_OUTPUT, - wid_caps & AC_WCAP_STEREO, - conn_len); + wid_caps, conn_len); else print_amp_vals(buffer, codec, nid, HDA_OUTPUT, - wid_caps & AC_WCAP_STEREO, 1); + wid_caps, 1); } switch (wid_type) { -- cgit v1.2.1 From 09d042a2eb90ee2c86d80c48ad096ae3f5776cef Mon Sep 17 00:00:00 2001 From: Dmitry Torokhov Date: Mon, 16 Mar 2015 09:17:16 -0700 Subject: Revert "Input: synaptics - use dmax in input_mt_assign_slots" This reverts commit 6ab17a8484f03c188a93713369912f1545eb26e9 since it, according to Benjamin, causes issues with slot assignment: E: 15.669119 0000 0000 0000 # ------------ SYN_REPORT (0) ---------- E: 15.954242 0003 002f 0000 # EV_ABS / ABS_MT_SLOT 0 E: 15.954242 0003 0039 0505 # EV_ABS / ABS_MT_TRACKING_ID 505 E: 15.954242 0003 0035 3851 # EV_ABS / ABS_MT_POSITION_X 3851 E: 15.954242 0003 0036 4076 # EV_ABS / ABS_MT_POSITION_Y 4076 E: 15.954242 0003 003a 0034 # EV_ABS / ABS_MT_PRESSURE 34 E: 15.954242 0001 014a 0001 # EV_KEY / BTN_TOUCH 1 E: 15.954242 0003 0000 3851 # EV_ABS / ABS_X 3851 E: 15.954242 0003 0001 4076 # EV_ABS / ABS_Y 4076 E: 15.954242 0003 0018 0034 # EV_ABS / ABS_PRESSURE 34 E: 15.954242 0001 0145 0001 # EV_KEY / BTN_TOOL_FINGER 1 E: 15.954242 0000 0000 0000 # ------------ SYN_REPORT (0) ---------- ... (bunch of regular events)... E: 16.020614 0000 0000 0000 # ------------ SYN_REPORT (0) ---------- E: 16.043601 0003 0035 3873 # EV_ABS / ABS_MT_POSITION_X 3873 E: 16.043601 0003 0036 3903 # EV_ABS / ABS_MT_POSITION_Y 3903 E: 16.043601 0003 003a 0050 # EV_ABS / ABS_MT_PRESSURE 50 E: 16.043601 0003 0035 3032 # EV_ABS / ABS_MT_POSITION_X 3032 E: 16.043601 0003 0036 3832 # EV_ABS / ABS_MT_POSITION_Y 3832 E: 16.043601 0003 003a 0044 # EV_ABS / ABS_MT_PRESSURE 44 E: 16.043601 0003 0000 3032 # EV_ABS / ABS_X 3032 E: 16.043601 0003 0001 3832 # EV_ABS / ABS_Y 3832 E: 16.043601 0003 0018 0044 # EV_ABS / ABS_PRESSURE 44 E: 16.043601 0001 0145 0000 # EV_KEY / BTN_TOOL_FINGER 0 E: 16.043601 0001 014d 0001 # EV_KEY / BTN_TOOL_DOUBLETAP 1 E: 16.043601 0000 0000 0000 # ------------ SYN_REPORT (0) ---------- E: 16.068837 0003 002f 0001 # EV_ABS / ABS_MT_SLOT 1 E: 16.068837 0003 0039 0506 # EV_ABS / ABS_MT_TRACKING_ID 506 E: 16.068837 0003 0035 3912 # EV_ABS / ABS_MT_POSITION_X 3912 E: 16.068837 0003 0036 3743 # EV_ABS / ABS_MT_POSITION_Y 3743 E: 16.068837 0003 003a 0056 # EV_ABS / ABS_MT_PRESSURE 56 E: 16.068837 0003 002f 0000 # EV_ABS / ABS_MT_SLOT 0 E: 16.068837 0003 0035 3026 # EV_ABS / ABS_MT_POSITION_X 3026 E: 16.068837 0003 0036 3708 # EV_ABS / ABS_MT_POSITION_Y 3708 E: 16.068837 0003 003a 0052 # EV_ABS / ABS_MT_PRESSURE 52 E: 16.068837 0003 0000 3026 # EV_ABS / ABS_X 3026 E: 16.068837 0003 0001 3708 # EV_ABS / ABS_Y 3708 E: 16.068837 0003 0018 0052 # EV_ABS / ABS_PRESSURE 52 E: 16.068837 0000 0000 0000 # ------------ SYN_REPORT (0) ---------- Slot 0 and 1 gets inverted in the second report above, which introduces a cursor jump. The problem is that this cursor jump is often enough to leave the current widget, and X sends the scrolling events to whoever is now under the cursor. Reported-by: Benjamin Tissoires Reported-by: Hans de Goede --- drivers/input/mouse/synaptics.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/drivers/input/mouse/synaptics.c b/drivers/input/mouse/synaptics.c index c74bfa1c05e3..dda605836546 100644 --- a/drivers/input/mouse/synaptics.c +++ b/drivers/input/mouse/synaptics.c @@ -67,9 +67,6 @@ #define X_MAX_POSITIVE 8176 #define Y_MAX_POSITIVE 8176 -/* maximum ABS_MT_POSITION displacement (in mm) */ -#define DMAX 10 - /***************************************************************************** * Stuff we need even when we do not want native Synaptics support ****************************************************************************/ @@ -915,7 +912,7 @@ static void synaptics_report_mt_data(struct psmouse *psmouse, pos[i].y = synaptics_invert_y(hw[i]->y); } - input_mt_assign_slots(dev, slot, pos, nsemi, DMAX * priv->x_res); + input_mt_assign_slots(dev, slot, pos, nsemi, 0); for (i = 0; i < nsemi; i++) { input_mt_slot(dev, slot[i]); -- cgit v1.2.1 From a104a45ba7a51b5b4c5e8437020d9d48edf22f89 Mon Sep 17 00:00:00 2001 From: Andy Shevchenko Date: Mon, 9 Mar 2015 12:16:42 +0200 Subject: dmaengine: dw: append MODULE_ALIAS for platform driver The commit 9cade1a46c77 (dma: dw: split driver to library part and platform code) introduced a separate platform driver but missed to add a MODULE_ALIAS("platform:dw_dmac"); to that module. The patch adds this to get driver loaded automatically if platform device is registered. Reported-by: "Blin, Jerome" Fixes: 9cade1a46c77 (dma: dw: split driver to library part and platform code) Signed-off-by: Andy Shevchenko Signed-off-by: Vinod Koul --- drivers/dma/dw/platform.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/dma/dw/platform.c b/drivers/dma/dw/platform.c index 6565a361e7e5..b2c3ae071429 100644 --- a/drivers/dma/dw/platform.c +++ b/drivers/dma/dw/platform.c @@ -26,6 +26,8 @@ #include "internal.h" +#define DRV_NAME "dw_dmac" + static struct dma_chan *dw_dma_of_xlate(struct of_phandle_args *dma_spec, struct of_dma *ofdma) { @@ -284,7 +286,7 @@ static struct platform_driver dw_driver = { .remove = dw_remove, .shutdown = dw_shutdown, .driver = { - .name = "dw_dmac", + .name = DRV_NAME, .pm = &dw_dev_pm_ops, .of_match_table = of_match_ptr(dw_dma_of_id_table), .acpi_match_table = ACPI_PTR(dw_dma_acpi_id_table), @@ -305,3 +307,4 @@ module_exit(dw_exit); MODULE_LICENSE("GPL v2"); MODULE_DESCRIPTION("Synopsys DesignWare DMA Controller platform driver"); +MODULE_ALIAS("platform:" DRV_NAME); -- cgit v1.2.1 From 7cff4b1836a9d3f18aadd6e88fd43055e2ff4132 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Mon, 16 Mar 2015 10:44:52 +1100 Subject: kernfs: handle poll correctly on 'direct_read' files. Kernfs supports two styles of read: direct_read and seqfile_read. The latter supports 'poll' correctly thanks to the update of '->event' in kernfs_seq_show. The former does not as '->event' is never updated on a read. So add an appropriate update in kernfs_file_direct_read(). This was noticed because some 'md' sysfs attributes were recently changed to use direct reads. Reported-by: Prakash Punnoor Reported-by: Torsten Kaiser Fixes: 750f199ee8b578062341e6ddfe36c59ac8ff2dcb Signed-off-by: NeilBrown Acked-by: Tejun Heo Signed-off-by: Greg Kroah-Hartman --- fs/kernfs/file.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/kernfs/file.c b/fs/kernfs/file.c index b684e8a132e6..2bacb9988566 100644 --- a/fs/kernfs/file.c +++ b/fs/kernfs/file.c @@ -207,6 +207,7 @@ static ssize_t kernfs_file_direct_read(struct kernfs_open_file *of, goto out_free; } + of->event = atomic_read(&of->kn->attr.open->event); ops = kernfs_ops(of->kn); if (ops->read) len = ops->read(of, buf, len, *ppos); -- cgit v1.2.1 From d5e7cafd69da24e6d6cc988fab6ea313a2577efc Mon Sep 17 00:00:00 2001 From: JeHyeon Yeon Date: Mon, 16 Mar 2015 01:03:19 +0000 Subject: LZ4 : fix the data abort issue If the part of the compression data are corrupted, or the compression data is totally fake, the memory access over the limit is possible. This is the log from my system usning lz4 decompression. [6502]data abort, halting [6503]r0 0x00000000 r1 0x00000000 r2 0xdcea0ffc r3 0xdcea0ffc [6509]r4 0xb9ab0bfd r5 0xdcea0ffc r6 0xdcea0ff8 r7 0xdce80000 [6515]r8 0x00000000 r9 0x00000000 r10 0x00000000 r11 0xb9a98000 [6522]r12 0xdcea1000 usp 0x00000000 ulr 0x00000000 pc 0x820149bc [6528]spsr 0x400001f3 and the memory addresses of some variables at the moment are ref:0xdcea0ffc, op:0xdcea0ffc, oend:0xdcea1000 As you can see, COPYLENGH is 8bytes, so @ref and @op can access the momory over @oend. Signed-off-by: JeHyeon Yeon Reviewed-by: David Sterba Signed-off-by: Greg Kroah-Hartman --- lib/lz4/lz4_decompress.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/lib/lz4/lz4_decompress.c b/lib/lz4/lz4_decompress.c index 7a85967060a5..f0f5c5c3de12 100644 --- a/lib/lz4/lz4_decompress.c +++ b/lib/lz4/lz4_decompress.c @@ -139,6 +139,9 @@ static int lz4_uncompress(const char *source, char *dest, int osize) /* Error: request to write beyond destination buffer */ if (cpy > oend) goto _output_error; + if ((ref + COPYLENGTH) > oend || + (op + COPYLENGTH) > oend) + goto _output_error; LZ4_SECURECOPY(ref, op, (oend - COPYLENGTH)); while (op < cpy) *op++ = *ref++; -- cgit v1.2.1 From a8e0c246dacfb0558e801ab81af3f670056fd1b2 Mon Sep 17 00:00:00 2001 From: Michal Schmidt Date: Mon, 16 Mar 2015 16:15:59 +0100 Subject: bnx2x: fix encapsulation features on 57710/57711 E1x chips (57710, 57711(E)) have no support for encapsulation offload. bnx2x incorrectly advertises the support as available. Setting of those features is conditional on "!CHIP_IS_E1x(bp)", but the bp struct is not initialized yet at this point and consequently any chip passes the check. The check must use the "chip_is_e1x" local variable instead to work correctly. Signed-off-by: Michal Schmidt Signed-off-by: David S. Miller --- drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c index bef750a09027..996e215fc324 100644 --- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c +++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c @@ -12769,7 +12769,7 @@ static int bnx2x_init_dev(struct bnx2x *bp, struct pci_dev *pdev, NETIF_F_TSO | NETIF_F_TSO_ECN | NETIF_F_TSO6 | NETIF_F_RXCSUM | NETIF_F_LRO | NETIF_F_GRO | NETIF_F_RXHASH | NETIF_F_HW_VLAN_CTAG_TX; - if (!CHIP_IS_E1x(bp)) { + if (!chip_is_e1x) { dev->hw_features |= NETIF_F_GSO_GRE | NETIF_F_GSO_UDP_TUNNEL | NETIF_F_GSO_IPIP | NETIF_F_GSO_SIT; dev->hw_enc_features = -- cgit v1.2.1 From aaad2d8c7b62489f03ceac4fa3c9ebb17ccc7860 Mon Sep 17 00:00:00 2001 From: Ben Goz Date: Sun, 8 Mar 2015 14:17:02 +0200 Subject: drm/amdkfd: destroy mqd when destroying kernel queue This patch adds a missing destruction of mqd, when destroying a kernel queue. Without the destruction, there is a memory leakage when repeatedly creating and destroying kernel queues. Signed-off-by: Ben Goz Signed-off-by: Oded Gabbay Reviewed-by: Alex Deucher Cc: stable@vger.kernel.org --- drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c | 22 +++++++++++++--------- 1 file changed, 13 insertions(+), 9 deletions(-) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c index e415a2a9207e..c7d298e62c96 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_kernel_queue.c @@ -44,7 +44,7 @@ static bool initialize(struct kernel_queue *kq, struct kfd_dev *dev, BUG_ON(!kq || !dev); BUG_ON(type != KFD_QUEUE_TYPE_DIQ && type != KFD_QUEUE_TYPE_HIQ); - pr_debug("kfd: In func %s initializing queue type %d size %d\n", + pr_debug("amdkfd: In func %s initializing queue type %d size %d\n", __func__, KFD_QUEUE_TYPE_HIQ, queue_size); nop.opcode = IT_NOP; @@ -69,12 +69,16 @@ static bool initialize(struct kernel_queue *kq, struct kfd_dev *dev, prop.doorbell_ptr = kfd_get_kernel_doorbell(dev, &prop.doorbell_off); - if (prop.doorbell_ptr == NULL) + if (prop.doorbell_ptr == NULL) { + pr_err("amdkfd: error init doorbell"); goto err_get_kernel_doorbell; + } retval = kfd_gtt_sa_allocate(dev, queue_size, &kq->pq); - if (retval != 0) + if (retval != 0) { + pr_err("amdkfd: error init pq queues size (%d)\n", queue_size); goto err_pq_allocate_vidmem; + } kq->pq_kernel_addr = kq->pq->cpu_ptr; kq->pq_gpu_addr = kq->pq->gpu_addr; @@ -165,10 +169,8 @@ err_rptr_allocate_vidmem: err_eop_allocate_vidmem: kfd_gtt_sa_free(dev, kq->pq); err_pq_allocate_vidmem: - pr_err("kfd: error init pq\n"); kfd_release_kernel_doorbell(dev, prop.doorbell_ptr); err_get_kernel_doorbell: - pr_err("kfd: error init doorbell"); return false; } @@ -187,6 +189,8 @@ static void uninitialize(struct kernel_queue *kq) else if (kq->queue->properties.type == KFD_QUEUE_TYPE_DIQ) kfd_gtt_sa_free(kq->dev, kq->fence_mem_obj); + kq->mqd->uninit_mqd(kq->mqd, kq->queue->mqd, kq->queue->mqd_mem_obj); + kfd_gtt_sa_free(kq->dev, kq->rptr_mem); kfd_gtt_sa_free(kq->dev, kq->wptr_mem); kq->ops_asic_specific.uninitialize(kq); @@ -211,7 +215,7 @@ static int acquire_packet_buffer(struct kernel_queue *kq, queue_address = (unsigned int *)kq->pq_kernel_addr; queue_size_dwords = kq->queue->properties.queue_size / sizeof(uint32_t); - pr_debug("kfd: In func %s\nrptr: %d\nwptr: %d\nqueue_address 0x%p\n", + pr_debug("amdkfd: In func %s\nrptr: %d\nwptr: %d\nqueue_address 0x%p\n", __func__, rptr, wptr, queue_address); available_size = (rptr - 1 - wptr + queue_size_dwords) % @@ -296,7 +300,7 @@ struct kernel_queue *kernel_queue_init(struct kfd_dev *dev, } if (kq->ops.initialize(kq, dev, type, KFD_KERNEL_QUEUE_SIZE) == false) { - pr_err("kfd: failed to init kernel queue\n"); + pr_err("amdkfd: failed to init kernel queue\n"); kfree(kq); return NULL; } @@ -319,7 +323,7 @@ static __attribute__((unused)) void test_kq(struct kfd_dev *dev) BUG_ON(!dev); - pr_err("kfd: starting kernel queue test\n"); + pr_err("amdkfd: starting kernel queue test\n"); kq = kernel_queue_init(dev, KFD_QUEUE_TYPE_HIQ); BUG_ON(!kq); @@ -330,7 +334,7 @@ static __attribute__((unused)) void test_kq(struct kfd_dev *dev) buffer[i] = kq->nop_packet; kq->ops.submit_packet(kq); - pr_err("kfd: ending kernel queue test\n"); + pr_err("amdkfd: ending kernel queue test\n"); } -- cgit v1.2.1 From 4fadf6b6570edc40c01eb7760055f9adc19971a5 Mon Sep 17 00:00:00 2001 From: Ben Goz Date: Tue, 10 Mar 2015 14:02:25 +0200 Subject: drm/amdkfd: Fix SDMA queue init. in non-HWS mode This patch fixes the SDMA queue initialization, when running in non-HWS mode. The first fix is to move the initialization of SDMA VM parameters before the initialization of the SDMA MQD. The second fix is to load the MQD to an HQD after the initialization of the MQD. Signed-off-by: Ben Goz Signed-off-by: Oded Gabbay Reviewed-by: Alex Deucher --- drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c index 910ff8ab9c9c..d8135adb2238 100644 --- a/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device_queue_manager.c @@ -645,6 +645,7 @@ static int create_sdma_queue_nocpsch(struct device_queue_manager *dqm, pr_debug(" sdma queue id: %d\n", q->properties.sdma_queue_id); pr_debug(" sdma engine id: %d\n", q->properties.sdma_engine_id); + init_sdma_vm(dqm, q, qpd); retval = mqd->init_mqd(mqd, &q->mqd, &q->mqd_mem_obj, &q->gart_mqd_addr, &q->properties); if (retval != 0) { @@ -652,7 +653,14 @@ static int create_sdma_queue_nocpsch(struct device_queue_manager *dqm, return retval; } - init_sdma_vm(dqm, q, qpd); + retval = mqd->load_mqd(mqd, q->mqd, 0, + 0, NULL); + if (retval != 0) { + deallocate_sdma_queue(dqm, q->sdma_id); + mqd->uninit_mqd(mqd, q->mqd, q->mqd_mem_obj); + return retval; + } + return 0; } -- cgit v1.2.1 From e405ca3a1bf166f741506c07c2a277b5d48af8f7 Mon Sep 17 00:00:00 2001 From: Ben Goz Date: Sun, 8 Mar 2015 14:15:16 +0200 Subject: drm/radeon: Changing number of compute pipe lines The current CP firmware can handle Usermode Queues only on MEC1. To reflect this firmware change, this commit reduces number of compute pipelines to 4 - 1, from 8 - 1 (the first pipeline is allocated for kgd). Signed-off-by: Ben Goz Signed-off-by: Oded Gabbay Cc: stable@vger.kernel.org --- drivers/gpu/drm/radeon/radeon_kfd.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/radeon/radeon_kfd.c b/drivers/gpu/drm/radeon/radeon_kfd.c index 061eaa9c19c7..122eb5693ba1 100644 --- a/drivers/gpu/drm/radeon/radeon_kfd.c +++ b/drivers/gpu/drm/radeon/radeon_kfd.c @@ -153,7 +153,7 @@ void radeon_kfd_device_init(struct radeon_device *rdev) .compute_vmid_bitmap = 0xFF00, .first_compute_pipe = 1, - .compute_pipe_count = 8 - 1, + .compute_pipe_count = 4 - 1, }; radeon_doorbell_get_kfd_info(rdev, -- cgit v1.2.1 From d6e5b7cc9819f9a108294f256dd80939e91a0a1f Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Pali=20Roh=C3=A1r?= Date: Thu, 26 Feb 2015 14:49:56 +0100 Subject: ARM: dts: omap3: Add missing dmas for crypto MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This patch adds missing dma DTS definitions for omap aes and sham drivers. Without it kernel drivers do not work for device tree based booting while it works for legacy booting on general purpose SoCs. Note that further changes are still needed for high secure SoCs. But since that never worked in legacy boot mode either, those will be sent separately. Signed-off-by: Pali Rohár Acked-by: Pavel Machek [tony@atomide.com: updated comments] Signed-off-by: Tony Lindgren --- arch/arm/boot/dts/omap3.dtsi | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/arch/arm/boot/dts/omap3.dtsi b/arch/arm/boot/dts/omap3.dtsi index f4f78c40b564..3fdc84fddb70 100644 --- a/arch/arm/boot/dts/omap3.dtsi +++ b/arch/arm/boot/dts/omap3.dtsi @@ -92,6 +92,8 @@ ti,hwmods = "aes"; reg = <0x480c5000 0x50>; interrupts = <0>; + dmas = <&sdma 65 &sdma 66>; + dma-names = "tx", "rx"; }; prm: prm@48306000 { @@ -550,6 +552,8 @@ ti,hwmods = "sham"; reg = <0x480c3000 0x64>; interrupts = <49>; + dmas = <&sdma 69>; + dma-names = "rx"; }; smartreflex_core: smartreflex@480cb000 { -- cgit v1.2.1 From e5ed5b60272871786b1c5434079925bc60d771b7 Mon Sep 17 00:00:00 2001 From: Suman Anna Date: Wed, 11 Mar 2015 18:38:38 -0500 Subject: ARM: OMAP2+: Fix socbus family info for AM33xx devices The family information in the soc-bus data is currently not classified properly for AM33xx devices, and a read of /sys/bus/soc/devices/soc0/family currently shows "Unknown". Fix the same. Signed-off-by: Suman Anna Signed-off-by: Tony Lindgren --- arch/arm/mach-omap2/id.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/arm/mach-omap2/id.c b/arch/arm/mach-omap2/id.c index 2a2f4d56e4c8..25f1beea453e 100644 --- a/arch/arm/mach-omap2/id.c +++ b/arch/arm/mach-omap2/id.c @@ -720,6 +720,8 @@ static const char * __init omap_get_family(void) return kasprintf(GFP_KERNEL, "OMAP4"); else if (soc_is_omap54xx()) return kasprintf(GFP_KERNEL, "OMAP5"); + else if (soc_is_am33xx() || soc_is_am335x()) + return kasprintf(GFP_KERNEL, "AM33xx"); else if (soc_is_am43xx()) return kasprintf(GFP_KERNEL, "AM43xx"); else if (soc_is_dra7xx()) -- cgit v1.2.1 From adc346b133c952ec6988d90f6fa79cbe0a3eb7ef Mon Sep 17 00:00:00 2001 From: Ben Skeggs Date: Tue, 27 Jan 2015 15:09:39 +1000 Subject: drm/nouveau/fifo/nv04: remove the loop from the interrupt handler Complete bong hit (and not the last...), the hardware will reassert the interrupt to PMC if it's necessary. Also potentially harmful in the face of interrupts such as the non-stall interrupt, which remain active in NV_PFIFO_INTR even when we don't care about servicing it. It appears (hopefully, fdo#87244), that under certain loads, the methods may pass quickly enough to hit the "100 spins and kill PFIFO" thing that we had going on. Not ideal ;) Signed-off-by: Ben Skeggs --- drivers/gpu/drm/nouveau/nvkm/engine/fifo/nv04.c | 85 ++++++++++--------------- 1 file changed, 35 insertions(+), 50 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/nv04.c b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/nv04.c index b038b6eb51db..043e4296084c 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/fifo/nv04.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/fifo/nv04.c @@ -502,72 +502,57 @@ nv04_fifo_intr(struct nvkm_subdev *subdev) { struct nvkm_device *device = nv_device(subdev); struct nv04_fifo_priv *priv = (void *)subdev; - uint32_t status, reassign; - int cnt = 0; + u32 mask = nv_rd32(priv, NV03_PFIFO_INTR_EN_0); + u32 stat = nv_rd32(priv, NV03_PFIFO_INTR_0) & mask; + u32 reassign, chid, get, sem; reassign = nv_rd32(priv, NV03_PFIFO_CACHES) & 1; - while ((status = nv_rd32(priv, NV03_PFIFO_INTR_0)) && (cnt++ < 100)) { - uint32_t chid, get; - - nv_wr32(priv, NV03_PFIFO_CACHES, 0); - - chid = nv_rd32(priv, NV03_PFIFO_CACHE1_PUSH1) & priv->base.max; - get = nv_rd32(priv, NV03_PFIFO_CACHE1_GET); + nv_wr32(priv, NV03_PFIFO_CACHES, 0); - if (status & NV_PFIFO_INTR_CACHE_ERROR) { - nv04_fifo_cache_error(device, priv, chid, get); - status &= ~NV_PFIFO_INTR_CACHE_ERROR; - } + chid = nv_rd32(priv, NV03_PFIFO_CACHE1_PUSH1) & priv->base.max; + get = nv_rd32(priv, NV03_PFIFO_CACHE1_GET); - if (status & NV_PFIFO_INTR_DMA_PUSHER) { - nv04_fifo_dma_pusher(device, priv, chid); - status &= ~NV_PFIFO_INTR_DMA_PUSHER; - } + if (stat & NV_PFIFO_INTR_CACHE_ERROR) { + nv04_fifo_cache_error(device, priv, chid, get); + stat &= ~NV_PFIFO_INTR_CACHE_ERROR; + } - if (status & NV_PFIFO_INTR_SEMAPHORE) { - uint32_t sem; + if (stat & NV_PFIFO_INTR_DMA_PUSHER) { + nv04_fifo_dma_pusher(device, priv, chid); + stat &= ~NV_PFIFO_INTR_DMA_PUSHER; + } - status &= ~NV_PFIFO_INTR_SEMAPHORE; - nv_wr32(priv, NV03_PFIFO_INTR_0, - NV_PFIFO_INTR_SEMAPHORE); + if (stat & NV_PFIFO_INTR_SEMAPHORE) { + stat &= ~NV_PFIFO_INTR_SEMAPHORE; + nv_wr32(priv, NV03_PFIFO_INTR_0, NV_PFIFO_INTR_SEMAPHORE); - sem = nv_rd32(priv, NV10_PFIFO_CACHE1_SEMAPHORE); - nv_wr32(priv, NV10_PFIFO_CACHE1_SEMAPHORE, sem | 0x1); + sem = nv_rd32(priv, NV10_PFIFO_CACHE1_SEMAPHORE); + nv_wr32(priv, NV10_PFIFO_CACHE1_SEMAPHORE, sem | 0x1); - nv_wr32(priv, NV03_PFIFO_CACHE1_GET, get + 4); - nv_wr32(priv, NV04_PFIFO_CACHE1_PULL0, 1); - } + nv_wr32(priv, NV03_PFIFO_CACHE1_GET, get + 4); + nv_wr32(priv, NV04_PFIFO_CACHE1_PULL0, 1); + } - if (device->card_type == NV_50) { - if (status & 0x00000010) { - status &= ~0x00000010; - nv_wr32(priv, 0x002100, 0x00000010); - } - - if (status & 0x40000000) { - nv_wr32(priv, 0x002100, 0x40000000); - nvkm_fifo_uevent(&priv->base); - status &= ~0x40000000; - } + if (device->card_type == NV_50) { + if (stat & 0x00000010) { + stat &= ~0x00000010; + nv_wr32(priv, 0x002100, 0x00000010); } - if (status) { - nv_warn(priv, "unknown intr 0x%08x, ch %d\n", - status, chid); - nv_wr32(priv, NV03_PFIFO_INTR_0, status); - status = 0; + if (stat & 0x40000000) { + nv_wr32(priv, 0x002100, 0x40000000); + nvkm_fifo_uevent(&priv->base); + stat &= ~0x40000000; } - - nv_wr32(priv, NV03_PFIFO_CACHES, reassign); } - if (status) { - nv_error(priv, "still angry after %d spins, halt\n", cnt); - nv_wr32(priv, 0x002140, 0); - nv_wr32(priv, 0x000140, 0); + if (stat) { + nv_warn(priv, "unknown intr 0x%08x\n", stat); + nv_mask(priv, NV03_PFIFO_INTR_EN_0, stat, 0x00000000); + nv_wr32(priv, NV03_PFIFO_INTR_0, stat); } - nv_wr32(priv, 0x000100, 0x00000100); + nv_wr32(priv, NV03_PFIFO_CACHES, reassign); } static int -- cgit v1.2.1 From 404ba3f79089a01c1ebacccafa08a5db4a4cd2af Mon Sep 17 00:00:00 2001 From: Ben Skeggs Date: Tue, 27 Jan 2015 15:44:06 +1000 Subject: drm/nouveau/gr/gf100: fix some accidental or'ing of buffer addresses fdo#83992 Signed-off-by: Ben Skeggs --- drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c | 4 ++-- drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c | 4 ++-- drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c | 4 ++-- 3 files changed, 6 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c index 2e7ec389eea7..57e2c5b13123 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgf100.c @@ -1032,9 +1032,9 @@ gf100_grctx_generate_bundle(struct gf100_grctx *info) const int s = 8; const int b = mmio_vram(info, impl->bundle_size, (1 << s), access); mmio_refn(info, 0x408004, 0x00000000, s, b); - mmio_refn(info, 0x408008, 0x80000000 | (impl->bundle_size >> s), 0, b); + mmio_wr32(info, 0x408008, 0x80000000 | (impl->bundle_size >> s)); mmio_refn(info, 0x418808, 0x00000000, s, b); - mmio_refn(info, 0x41880c, 0x80000000 | (impl->bundle_size >> s), 0, b); + mmio_wr32(info, 0x41880c, 0x80000000 | (impl->bundle_size >> s)); } void diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c index b52300d8861a..5e9454ba158f 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgk104.c @@ -851,9 +851,9 @@ gk104_grctx_generate_bundle(struct gf100_grctx *info) const int s = 8; const int b = mmio_vram(info, impl->bundle_size, (1 << s), access); mmio_refn(info, 0x408004, 0x00000000, s, b); - mmio_refn(info, 0x408008, 0x80000000 | (impl->bundle_size >> s), 0, b); + mmio_wr32(info, 0x408008, 0x80000000 | (impl->bundle_size >> s)); mmio_refn(info, 0x418808, 0x00000000, s, b); - mmio_refn(info, 0x41880c, 0x80000000 | (impl->bundle_size >> s), 0, b); + mmio_wr32(info, 0x41880c, 0x80000000 | (impl->bundle_size >> s)); mmio_wr32(info, 0x4064c8, (state_limit << 16) | token_limit); } diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c index 956f4dce960c..b2fae6e389e2 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/gr/ctxgm107.c @@ -871,9 +871,9 @@ gm107_grctx_generate_bundle(struct gf100_grctx *info) const int s = 8; const int b = mmio_vram(info, impl->bundle_size, (1 << s), access); mmio_refn(info, 0x408004, 0x00000000, s, b); - mmio_refn(info, 0x408008, 0x80000000 | (impl->bundle_size >> s), 0, b); + mmio_wr32(info, 0x408008, 0x80000000 | (impl->bundle_size >> s)); mmio_refn(info, 0x418e24, 0x00000000, s, b); - mmio_refn(info, 0x418e28, 0x80000000 | (impl->bundle_size >> s), 0, b); + mmio_wr32(info, 0x418e28, 0x80000000 | (impl->bundle_size >> s)); mmio_wr32(info, 0x4064c8, (state_limit << 16) | token_limit); } -- cgit v1.2.1 From 9fcaa149e7d264822f7ef748ca4a7af643d88ec9 Mon Sep 17 00:00:00 2001 From: Ben Skeggs Date: Mon, 2 Feb 2015 09:08:14 +1000 Subject: drm/nouveau/device: post write to NV_PMC_BOOT_1 when flipping endian switch fdo#88868 Signed-off-by: Ben Skeggs --- drivers/gpu/drm/nouveau/nvkm/engine/device/base.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c index 29bd539af183..6efa8f38ff54 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/base.c @@ -340,11 +340,13 @@ nvkm_devobj_ctor(struct nvkm_object *parent, struct nvkm_object *engine, /* switch mmio to cpu's native endianness */ #ifndef __BIG_ENDIAN - if (ioread32_native(map + 0x000004) != 0x00000000) + if (ioread32_native(map + 0x000004) != 0x00000000) { #else - if (ioread32_native(map + 0x000004) == 0x00000000) + if (ioread32_native(map + 0x000004) == 0x00000000) { #endif iowrite32_native(0x01000001, map + 0x000004); + ioread32_native(map); + } /* read boot0 and strapping information */ boot0 = ioread32_native(map + 0x000000); -- cgit v1.2.1 From 7e547adcea7b9d927009717e1f3303879d5f2687 Mon Sep 17 00:00:00 2001 From: Stefan Huehner Date: Sat, 21 Feb 2015 22:23:56 +0100 Subject: drm/nouveau/device/gm100: Basic GM206 bring up (as copy of GM204) Enough to get VGA monitor on DVI-I output have output. HDMI output not yet working Signed-off-by: Ben Skeggs --- drivers/gpu/drm/nouveau/nvkm/engine/device/gm100.c | 43 ++++++++++++++++++++++ 1 file changed, 43 insertions(+) diff --git a/drivers/gpu/drm/nouveau/nvkm/engine/device/gm100.c b/drivers/gpu/drm/nouveau/nvkm/engine/device/gm100.c index 539561ed3281..108d048da764 100644 --- a/drivers/gpu/drm/nouveau/nvkm/engine/device/gm100.c +++ b/drivers/gpu/drm/nouveau/nvkm/engine/device/gm100.c @@ -140,6 +140,49 @@ gm100_identify(struct nvkm_device *device) device->oclass[NVDEV_ENGINE_MSVLD ] = &gk104_msvld_oclass; device->oclass[NVDEV_ENGINE_MSPDEC ] = &gk104_mspdec_oclass; device->oclass[NVDEV_ENGINE_MSPPP ] = &gf100_msppp_oclass; +#endif + break; + case 0x126: + device->cname = "GM206"; + device->oclass[NVDEV_SUBDEV_VBIOS ] = &nvkm_bios_oclass; + device->oclass[NVDEV_SUBDEV_GPIO ] = gk104_gpio_oclass; + device->oclass[NVDEV_SUBDEV_I2C ] = gm204_i2c_oclass; + device->oclass[NVDEV_SUBDEV_FUSE ] = &gm107_fuse_oclass; +#if 0 + /* looks to be some non-trivial changes */ + device->oclass[NVDEV_SUBDEV_CLK ] = &gk104_clk_oclass; + /* priv ring says no to 0x10eb14 writes */ + device->oclass[NVDEV_SUBDEV_THERM ] = &gm107_therm_oclass; +#endif + device->oclass[NVDEV_SUBDEV_MXM ] = &nv50_mxm_oclass; + device->oclass[NVDEV_SUBDEV_DEVINIT] = gm204_devinit_oclass; + device->oclass[NVDEV_SUBDEV_MC ] = gk20a_mc_oclass; + device->oclass[NVDEV_SUBDEV_BUS ] = gf100_bus_oclass; + device->oclass[NVDEV_SUBDEV_TIMER ] = &gk20a_timer_oclass; + device->oclass[NVDEV_SUBDEV_FB ] = gm107_fb_oclass; + device->oclass[NVDEV_SUBDEV_LTC ] = gm107_ltc_oclass; + device->oclass[NVDEV_SUBDEV_IBUS ] = &gk104_ibus_oclass; + device->oclass[NVDEV_SUBDEV_INSTMEM] = nv50_instmem_oclass; + device->oclass[NVDEV_SUBDEV_MMU ] = &gf100_mmu_oclass; + device->oclass[NVDEV_SUBDEV_BAR ] = &gf100_bar_oclass; + device->oclass[NVDEV_SUBDEV_PMU ] = gk208_pmu_oclass; +#if 0 + device->oclass[NVDEV_SUBDEV_VOLT ] = &nv40_volt_oclass; +#endif + device->oclass[NVDEV_ENGINE_DMAOBJ ] = gf110_dmaeng_oclass; +#if 0 + device->oclass[NVDEV_ENGINE_FIFO ] = gk208_fifo_oclass; + device->oclass[NVDEV_ENGINE_SW ] = gf100_sw_oclass; + device->oclass[NVDEV_ENGINE_GR ] = gm107_gr_oclass; +#endif + device->oclass[NVDEV_ENGINE_DISP ] = gm204_disp_oclass; +#if 0 + device->oclass[NVDEV_ENGINE_CE0 ] = &gm204_ce0_oclass; + device->oclass[NVDEV_ENGINE_CE1 ] = &gm204_ce1_oclass; + device->oclass[NVDEV_ENGINE_CE2 ] = &gm204_ce2_oclass; + device->oclass[NVDEV_ENGINE_MSVLD ] = &gk104_msvld_oclass; + device->oclass[NVDEV_ENGINE_MSPDEC ] = &gk104_mspdec_oclass; + device->oclass[NVDEV_ENGINE_MSPPP ] = &gf100_msppp_oclass; #endif break; default: -- cgit v1.2.1 From 5a6f690ca565f536ab0fc05e267b08d5a7c36b46 Mon Sep 17 00:00:00 2001 From: Stefan Huehner Date: Sun, 22 Feb 2015 15:46:36 +0100 Subject: drm/nouveau/bios: fix i2c table parsing for dcb 4.1 Code before looked only at bit 31 to decide if a port is unused. However dcb 4.1 spec says 0x1F in bits 31-27 and 26-22 means unused. This fixed hdmi monitor detection on GM206. Signed-off-by: Ben Skeggs --- drivers/gpu/drm/nouveau/nvkm/subdev/bios/i2c.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/i2c.c b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/i2c.c index d1a89b2bd5c1..c4e1f085ee10 100644 --- a/drivers/gpu/drm/nouveau/nvkm/subdev/bios/i2c.c +++ b/drivers/gpu/drm/nouveau/nvkm/subdev/bios/i2c.c @@ -74,7 +74,11 @@ dcb_i2c_parse(struct nvkm_bios *bios, u8 idx, struct dcb_i2c_entry *info) u16 ent = dcb_i2c_entry(bios, idx, &ver, &len); if (ent) { if (ver >= 0x41) { - if (!(nv_ro32(bios, ent) & 0x80000000)) + u32 ent_value = nv_ro32(bios, ent); + u8 i2c_port = (ent_value >> 27) & 0x1f; + u8 dpaux_port = (ent_value >> 22) & 0x1f; + /* value 0x1f means unused according to DCB 4.x spec */ + if (i2c_port == 0x1f && dpaux_port == 0x1f) info->type = DCB_I2C_UNUSED; else info->type = DCB_I2C_PMGR; -- cgit v1.2.1 From 704a0b5f234db26de5203740999e39523cfa4e3a Mon Sep 17 00:00:00 2001 From: "Michael S. Tsirkin" Date: Tue, 17 Mar 2015 12:11:30 +1030 Subject: virtio_mmio: fix access width for mmio Going over the virtio mmio code, I noticed that it doesn't correctly access modern device config values using "natural" accessors: it uses readb to get/set them byte by byte, while the virtio 1.0 spec explicitly states: 4.2.2.2 Driver Requirements: MMIO Device Register Layout ... The driver MUST only use 32 bit wide and aligned reads and writes to access the control registers described in table 4.1. For the device-specific configuration space, the driver MUST use 8 bit wide accesses for 8 bit wide fields, 16 bit wide and aligned accesses for 16 bit wide fields and 32 bit wide and aligned accesses for 32 and 64 bit wide fields. Borrow code from virtio_pci_modern to do this correctly. Signed-off-by: Michael S. Tsirkin Signed-off-by: Rusty Russell --- drivers/virtio/virtio_mmio.c | 79 +++++++++++++++++++++++++++++++++++++++----- 1 file changed, 71 insertions(+), 8 deletions(-) diff --git a/drivers/virtio/virtio_mmio.c b/drivers/virtio/virtio_mmio.c index 9c877d2375a5..6010d7ec0a0f 100644 --- a/drivers/virtio/virtio_mmio.c +++ b/drivers/virtio/virtio_mmio.c @@ -156,22 +156,85 @@ static void vm_get(struct virtio_device *vdev, unsigned offset, void *buf, unsigned len) { struct virtio_mmio_device *vm_dev = to_virtio_mmio_device(vdev); - u8 *ptr = buf; - int i; + void __iomem *base = vm_dev->base + VIRTIO_MMIO_CONFIG; + u8 b; + __le16 w; + __le32 l; - for (i = 0; i < len; i++) - ptr[i] = readb(vm_dev->base + VIRTIO_MMIO_CONFIG + offset + i); + if (vm_dev->version == 1) { + u8 *ptr = buf; + int i; + + for (i = 0; i < len; i++) + ptr[i] = readb(base + offset + i); + return; + } + + switch (len) { + case 1: + b = readb(base + offset); + memcpy(buf, &b, sizeof b); + break; + case 2: + w = cpu_to_le16(readw(base + offset)); + memcpy(buf, &w, sizeof w); + break; + case 4: + l = cpu_to_le32(readl(base + offset)); + memcpy(buf, &l, sizeof l); + break; + case 8: + l = cpu_to_le32(readl(base + offset)); + memcpy(buf, &l, sizeof l); + l = cpu_to_le32(ioread32(base + offset + sizeof l)); + memcpy(buf + sizeof l, &l, sizeof l); + break; + default: + BUG(); + } } static void vm_set(struct virtio_device *vdev, unsigned offset, const void *buf, unsigned len) { struct virtio_mmio_device *vm_dev = to_virtio_mmio_device(vdev); - const u8 *ptr = buf; - int i; + void __iomem *base = vm_dev->base + VIRTIO_MMIO_CONFIG; + u8 b; + __le16 w; + __le32 l; - for (i = 0; i < len; i++) - writeb(ptr[i], vm_dev->base + VIRTIO_MMIO_CONFIG + offset + i); + if (vm_dev->version == 1) { + const u8 *ptr = buf; + int i; + + for (i = 0; i < len; i++) + writeb(ptr[i], base + offset + i); + + return; + } + + switch (len) { + case 1: + memcpy(&b, buf, sizeof b); + writeb(b, base + offset); + break; + case 2: + memcpy(&w, buf, sizeof w); + writew(le16_to_cpu(w), base + offset); + break; + case 4: + memcpy(&l, buf, sizeof l); + writel(le32_to_cpu(l), base + offset); + break; + case 8: + memcpy(&l, buf, sizeof l); + writel(le32_to_cpu(l), base + offset); + memcpy(&l, buf + sizeof l, sizeof l); + writel(le32_to_cpu(l), base + offset + sizeof l); + break; + default: + BUG(); + } } static u32 vm_generation(struct virtio_device *vdev) -- cgit v1.2.1 From 8cb2c2dc472775479a1a7e78180955f6f1cb0b0a Mon Sep 17 00:00:00 2001 From: Petr Mladek Date: Thu, 12 Mar 2015 12:55:13 +0100 Subject: livepatch: Fix subtle race with coming and going modules There is a notifier that handles live patches for coming and going modules. It takes klp_mutex lock to avoid races with coming and going patches but it does not keep the lock all the time. Therefore the following races are possible: 1. The notifier is called sometime in STATE_MODULE_COMING. The module is visible by find_module() in this state all the time. It means that new patch can be registered and enabled even before the notifier is called. It might create wrong order of stacked patches, see below for an example. 2. New patch could still see the module in the GOING state even after the notifier has been called. It will try to initialize the related object structures but the module could disappear at any time. There will stay mess in the structures. It might even cause an invalid memory access. This patch solves the problem by adding a boolean variable into struct module. The value is true after the coming and before the going handler is called. New patches need to be applied when the value is true and they need to ignore the module when the value is false. Note that we need to know state of all modules on the system. The races are related to new patches. Therefore we do not know what modules will get patched. Also note that we could not simply ignore going modules. The code from the module could be called even in the GOING state until mod->exit() finishes. If we start supporting patches with semantic changes between function calls, we need to apply new patches to any still usable code. See below for an example. Finally note that the patch solves only the situation when a new patch is registered. There are no such problems when the patch is being removed. It does not matter who disable the patch first, whether the normal disable_patch() or the module notifier. There is nothing to do once the patch is disabled. Alternative solutions: ====================== + reject new patches when a patched module is coming or going; this is ugly + wait with adding new patch until the module leaves the COMING and GOING states; this might be dangerous and complicated; we would need to release kgr_lock in the middle of the patch registration to avoid a deadlock with the coming and going handlers; also we might need a waitqueue for each module which seems to be even bigger overhead than the boolean + stop modules from entering COMING and GOING states; wait until modules leave these states when they are already there; looks complicated; we would need to ignore the module that asked to stop the others to avoid a deadlock; also it is unclear what to do when two modules asked to stop others and both are in COMING state (situation when two new patches are applied) + always register/enable new patches and fix up the potential mess (registered patches order) in klp_module_init(); this is nasty and prone to regressions in the future development + add another MODULE_STATE where the kallsyms are visible but the module is not used yet; this looks too complex; the module states are checked on "many" locations Example of patch stacking breakage: =================================== The notifier could _not_ _simply_ ignore already initialized module objects. For example, let's have three patches (P1, P2, P3) for functions a() and b() where a() is from vmcore and b() is from a module M. Something like: a() b() P1 a1() b1() P2 a2() b2() P3 a3() b3(3) If you load the module M after all patches are registered and enabled. The ftrace ops for function a() and b() has listed the functions in this order: ops_a->func_stack -> list(a3,a2,a1) ops_b->func_stack -> list(b3,b2,b1) , so the pointer to b3() is the first and will be used. Then you might have the following scenario. Let's start with state when patches P1 and P2 are registered and enabled but the module M is not loaded. Then ftrace ops for b() does not exist. Then we get into the following race: CPU0 CPU1 load_module(M) complete_formation() mod->state = MODULE_STATE_COMING; mutex_unlock(&module_mutex); klp_register_patch(P3); klp_enable_patch(P3); # STATE 1 klp_module_notify(M) klp_module_notify_coming(P1); klp_module_notify_coming(P2); klp_module_notify_coming(P3); # STATE 2 The ftrace ops for a() and b() then looks: STATE1: ops_a->func_stack -> list(a3,a2,a1); ops_b->func_stack -> list(b3); STATE2: ops_a->func_stack -> list(a3,a2,a1); ops_b->func_stack -> list(b2,b1,b3); therefore, b2() is used for the module but a3() is used for vmcore because they were the last added. Example of the race with going modules: ======================================= CPU0 CPU1 delete_module() #SYSCALL try_stop_module() mod->state = MODULE_STATE_GOING; mutex_unlock(&module_mutex); klp_register_patch() klp_enable_patch() #save place to switch universe b() # from module that is going a() # from core (patched) mod->exit(); Note that the function b() can be called until we call mod->exit(). If we do not apply patch against b() because it is in MODULE_STATE_GOING, it will call patched a() with modified semantic and things might get wrong. [jpoimboe@redhat.com: use one boolean instead of two] Signed-off-by: Petr Mladek Acked-by: Josh Poimboeuf Acked-by: Rusty Russell Signed-off-by: Jiri Kosina --- include/linux/module.h | 4 ++++ kernel/livepatch/core.c | 30 ++++++++++++++++++++++++++---- 2 files changed, 30 insertions(+), 4 deletions(-) diff --git a/include/linux/module.h b/include/linux/module.h index b653d7c0a05a..7232fde6a991 100644 --- a/include/linux/module.h +++ b/include/linux/module.h @@ -344,6 +344,10 @@ struct module { unsigned long *ftrace_callsites; #endif +#ifdef CONFIG_LIVEPATCH + bool klp_alive; +#endif + #ifdef CONFIG_MODULE_UNLOAD /* What modules depend on me? */ struct list_head source_list; diff --git a/kernel/livepatch/core.c b/kernel/livepatch/core.c index 01ca08804f51..3f9f1d6b4c2e 100644 --- a/kernel/livepatch/core.c +++ b/kernel/livepatch/core.c @@ -89,16 +89,28 @@ static bool klp_is_object_loaded(struct klp_object *obj) /* sets obj->mod if object is not vmlinux and module is found */ static void klp_find_object_module(struct klp_object *obj) { + struct module *mod; + if (!klp_is_module(obj)) return; mutex_lock(&module_mutex); /* - * We don't need to take a reference on the module here because we have - * the klp_mutex, which is also taken by the module notifier. This - * prevents any module from unloading until we release the klp_mutex. + * We do not want to block removal of patched modules and therefore + * we do not take a reference here. The patches are removed by + * a going module handler instead. + */ + mod = find_module(obj->name); + /* + * Do not mess work of the module coming and going notifiers. + * Note that the patch might still be needed before the going handler + * is called. Module functions can be called even in the GOING state + * until mod->exit() finishes. This is especially important for + * patches that modify semantic of the functions. */ - obj->mod = find_module(obj->name); + if (mod && mod->klp_alive) + obj->mod = mod; + mutex_unlock(&module_mutex); } @@ -767,6 +779,7 @@ static int klp_init_object(struct klp_patch *patch, struct klp_object *obj) return -EINVAL; obj->state = KLP_DISABLED; + obj->mod = NULL; klp_find_object_module(obj); @@ -961,6 +974,15 @@ static int klp_module_notify(struct notifier_block *nb, unsigned long action, mutex_lock(&klp_mutex); + /* + * Each module has to know that the notifier has been called. + * We never know what module will get patched by a new patch. + */ + if (action == MODULE_STATE_COMING) + mod->klp_alive = true; + else /* MODULE_STATE_GOING */ + mod->klp_alive = false; + list_for_each_entry(patch, &klp_patches, list) { for (obj = patch->objs; obj->funcs; obj++) { if (!klp_is_module(obj) || strcmp(obj->name, mod->name)) -- cgit v1.2.1 From e03826d5045e81a66a4fad7be9a8ecdaeb7911cf Mon Sep 17 00:00:00 2001 From: Keerthy Date: Tue, 17 Mar 2015 15:56:04 +0530 Subject: regulator: palmas: Correct TPS659038 register definition for REGEN2 The register offset for REGEN2_CTRL in different for TPS659038 chip as when compared with other Palmas family PMICs. In the case of TPS659038 the wrong offset pointed to PLLEN_CTRL and was causing a hang. Correcting the same. Signed-off-by: Keerthy Signed-off-by: Mark Brown Cc: stable@vger.kernel.org --- drivers/regulator/palmas-regulator.c | 4 ++++ include/linux/mfd/palmas.h | 3 +++ 2 files changed, 7 insertions(+) diff --git a/drivers/regulator/palmas-regulator.c b/drivers/regulator/palmas-regulator.c index 9205f433573c..18198316b6cf 100644 --- a/drivers/regulator/palmas-regulator.c +++ b/drivers/regulator/palmas-regulator.c @@ -1572,6 +1572,10 @@ static int palmas_regulators_probe(struct platform_device *pdev) if (!pmic) return -ENOMEM; + if (of_device_is_compatible(node, "ti,tps659038-pmic")) + palmas_generic_regs_info[PALMAS_REG_REGEN2].ctrl_addr = + TPS659038_REGEN2_CTRL; + pmic->dev = &pdev->dev; pmic->palmas = palmas; palmas->pmic = pmic; diff --git a/include/linux/mfd/palmas.h b/include/linux/mfd/palmas.h index fb0390a1a498..ee7b1ce7a6f8 100644 --- a/include/linux/mfd/palmas.h +++ b/include/linux/mfd/palmas.h @@ -2999,6 +2999,9 @@ enum usb_irq_events { #define PALMAS_GPADC_TRIM15 0x0E #define PALMAS_GPADC_TRIM16 0x0F +/* TPS659038 regen2_ctrl offset iss different from palmas */ +#define TPS659038_REGEN2_CTRL 0x12 + /* TPS65917 Interrupt registers */ /* Registers for function INTERRUPT */ -- cgit v1.2.1 From d6b6cb1d3e6f78d55c2d4043d77d0d8def3f3b99 Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Tue, 17 Mar 2015 13:21:42 +0100 Subject: netfilter: nf_tables: allow to change chain policy without hook if it exists If there's an existing base chain, we have to allow to change the default policy without indicating the hook information. However, if the chain doesn't exists, we have to enforce the presence of the hook attribute. Signed-off-by: Pablo Neira Ayuso --- net/netfilter/nf_tables_api.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/net/netfilter/nf_tables_api.c b/net/netfilter/nf_tables_api.c index 6ab777912237..ac1a9528dbf2 100644 --- a/net/netfilter/nf_tables_api.c +++ b/net/netfilter/nf_tables_api.c @@ -1225,7 +1225,10 @@ static int nf_tables_newchain(struct sock *nlsk, struct sk_buff *skb, if (nla[NFTA_CHAIN_POLICY]) { if ((chain != NULL && - !(chain->flags & NFT_BASE_CHAIN)) || + !(chain->flags & NFT_BASE_CHAIN))) + return -EOPNOTSUPP; + + if (chain == NULL && nla[NFTA_CHAIN_HOOK] == NULL) return -EOPNOTSUPP; -- cgit v1.2.1 From dcdf7f6ddba006f3482ebee73dfa6b75aec5f07b Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Mon, 2 Mar 2015 16:37:31 -0500 Subject: Btrfs: prepare block group cache before writing Writing the block group cache will modify the extent tree quite a bit because it truncates the old space cache and pre-allocates new stuff. To try and cut down on the churn lets do the setup dance first, then later on hopefully we can avoid looping with newly dirtied roots. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 2 ++ fs/btrfs/extent-tree.c | 26 ++++++++++++++++++++++++++ fs/btrfs/transaction.c | 5 ++++- 3 files changed, 32 insertions(+), 1 deletion(-) diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index b3dd55f52f71..a0c90a324f53 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -3386,6 +3386,8 @@ int btrfs_inc_extent_ref(struct btrfs_trans_handle *trans, int btrfs_write_dirty_block_groups(struct btrfs_trans_handle *trans, struct btrfs_root *root); +int btrfs_setup_space_cache(struct btrfs_trans_handle *trans, + struct btrfs_root *root); int btrfs_extent_readonly(struct btrfs_root *root, u64 bytenr); int btrfs_free_block_groups(struct btrfs_fs_info *info); int btrfs_read_block_groups(struct btrfs_root *root); diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 96c613bfe157..3ac3fefb1625 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3325,6 +3325,32 @@ out: return ret; } +int btrfs_setup_space_cache(struct btrfs_trans_handle *trans, + struct btrfs_root *root) +{ + struct btrfs_block_group_cache *cache, *tmp; + struct btrfs_transaction *cur_trans = trans->transaction; + struct btrfs_path *path; + + if (list_empty(&cur_trans->dirty_bgs) || + !btrfs_test_opt(root, SPACE_CACHE)) + return 0; + + path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; + + /* Could add new block groups, use _safe just in case */ + list_for_each_entry_safe(cache, tmp, &cur_trans->dirty_bgs, + dirty_list) { + if (cache->disk_cache_state == BTRFS_DC_CLEAR) + cache_save_setup(cache, trans, path); + } + + btrfs_free_path(path); + return 0; +} + int btrfs_write_dirty_block_groups(struct btrfs_trans_handle *trans, struct btrfs_root *root) { diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 2fe3ef5e9de3..932709af5163 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -1025,7 +1025,6 @@ static int update_cowonly_root(struct btrfs_trans_handle *trans, struct btrfs_root *tree_root = root->fs_info->tree_root; old_root_used = btrfs_root_used(&root->root_item); - btrfs_write_dirty_block_groups(trans, root); while (1) { old_root_bytenr = btrfs_root_bytenr(&root->root_item); @@ -1085,6 +1084,10 @@ static noinline int commit_cowonly_roots(struct btrfs_trans_handle *trans, if (ret) return ret; + ret = btrfs_setup_space_cache(trans, root); + if (ret) + return ret; + /* run_qgroups might have added some more refs */ ret = btrfs_run_delayed_refs(trans, root, (unsigned long)-1); if (ret) -- cgit v1.2.1 From ab676b7d6fbf4b294bf198fb27ade5b0e865c7ce Mon Sep 17 00:00:00 2001 From: "Kirill A. Shutemov" Date: Mon, 9 Mar 2015 23:11:12 +0200 Subject: pagemap: do not leak physical addresses to non-privileged userspace As pointed by recent post[1] on exploiting DRAM physical imperfection, /proc/PID/pagemap exposes sensitive information which can be used to do attacks. This disallows anybody without CAP_SYS_ADMIN to read the pagemap. [1] http://googleprojectzero.blogspot.com/2015/03/exploiting-dram-rowhammer-bug-to-gain.html [ Eventually we might want to do anything more finegrained, but for now this is the simple model. - Linus ] Signed-off-by: Kirill A. Shutemov Acked-by: Konstantin Khlebnikov Acked-by: Andy Lutomirski Cc: Pavel Emelyanov Cc: Andrew Morton Cc: Mark Seaborn Cc: stable@vger.kernel.org Signed-off-by: Linus Torvalds --- fs/proc/task_mmu.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index 956b75d61809..6dee68d013ff 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -1325,6 +1325,9 @@ out: static int pagemap_open(struct inode *inode, struct file *file) { + /* do not disclose physical addresses: attack vector */ + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; pr_warn_once("Bits 55-60 of /proc/PID/pagemap entries are about " "to stop being page-shift some time soon. See the " "linux/Documentation/vm/pagemap.txt for details.\n"); -- cgit v1.2.1 From a76fc9dda87b51010e4bc60b5e0065a70180b465 Mon Sep 17 00:00:00 2001 From: Suman Anna Date: Mon, 16 Mar 2015 20:14:02 -0500 Subject: ARM: OMAP: dmtimer: check for pm_runtime_get_sync() failure The current OMAP dmtimer probe does not check for the return status of pm_runtime_get_sync() before initializing the timer registers. Any timer with missing hwmod data would return a failure here, and the access of registers without enabling the clocks for the timer would trigger a l3_noc interrupt and a kernel boot hang. Add proper checking so that the probe would return a failure graciously without hanging the kernel boot. Signed-off-by: Suman Anna Signed-off-by: Tony Lindgren --- arch/arm/plat-omap/dmtimer.c | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-) diff --git a/arch/arm/plat-omap/dmtimer.c b/arch/arm/plat-omap/dmtimer.c index db10169a08de..f32c74c0e1de 100644 --- a/arch/arm/plat-omap/dmtimer.c +++ b/arch/arm/plat-omap/dmtimer.c @@ -799,6 +799,7 @@ static int omap_dm_timer_probe(struct platform_device *pdev) struct device *dev = &pdev->dev; const struct of_device_id *match; const struct dmtimer_platform_data *pdata; + int ret; match = of_match_device(of_match_ptr(omap_timer_match), dev); pdata = match ? match->data : dev->platform_data; @@ -860,7 +861,12 @@ static int omap_dm_timer_probe(struct platform_device *pdev) } if (!timer->reserved) { - pm_runtime_get_sync(dev); + ret = pm_runtime_get_sync(dev); + if (ret < 0) { + dev_err(dev, "%s: pm_runtime_get_sync failed!\n", + __func__); + goto err_get_sync; + } __omap_dm_timer_init_regs(timer); pm_runtime_put(dev); } @@ -873,6 +879,11 @@ static int omap_dm_timer_probe(struct platform_device *pdev) dev_dbg(dev, "Device Probed.\n"); return 0; + +err_get_sync: + pm_runtime_put_noidle(dev); + pm_runtime_disable(dev); + return ret; } /** -- cgit v1.2.1 From 51b7e5728ebcded3f2ced9cd3ff71076c91e85de Mon Sep 17 00:00:00 2001 From: Suman Anna Date: Mon, 16 Mar 2015 20:14:03 -0500 Subject: ARM: OMAP: dmtimer: disable pm runtime on remove Disable the pm_runtime of the device upon remove. This is added to balance the pm_runtime_enable() invoked in the probe. Signed-off-by: Suman Anna Signed-off-by: Tony Lindgren --- arch/arm/plat-omap/dmtimer.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/arm/plat-omap/dmtimer.c b/arch/arm/plat-omap/dmtimer.c index f32c74c0e1de..8ca94d379bc3 100644 --- a/arch/arm/plat-omap/dmtimer.c +++ b/arch/arm/plat-omap/dmtimer.c @@ -910,6 +910,8 @@ static int omap_dm_timer_remove(struct platform_device *pdev) } spin_unlock_irqrestore(&dm_timer_lock, flags); + pm_runtime_disable(&pdev->dev); + return ret; } -- cgit v1.2.1 From ad41faa88e39af451427c921a0f8b441e104b6fa Mon Sep 17 00:00:00 2001 From: Nicolas Dichtel Date: Tue, 17 Mar 2015 11:16:00 +0100 Subject: netdevice.h: fix ndo_bridge_* comments The argument 'flags' was missing in ndo_bridge_setlink(). ndo_bridge_dellink() was missing. Fixes: 407af3299ef1 ("bridge: Add netlink interface to configure vlans on bridge ports") Fixes: add511b38266 ("bridge: add flags argument to ndo_bridge_setlink and ndo_bridge_dellink") CC: Vlad Yasevich CC: Roopa Prabhu Signed-off-by: Nicolas Dichtel Signed-off-by: David S. Miller --- include/linux/netdevice.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index 429d1790a27e..dcf6ec27739b 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -965,9 +965,12 @@ typedef u16 (*select_queue_fallback_t)(struct net_device *dev, * Used to add FDB entries to dump requests. Implementers should add * entries to skb and update idx with the number of entries. * - * int (*ndo_bridge_setlink)(struct net_device *dev, struct nlmsghdr *nlh) + * int (*ndo_bridge_setlink)(struct net_device *dev, struct nlmsghdr *nlh, + * u16 flags) * int (*ndo_bridge_getlink)(struct sk_buff *skb, u32 pid, u32 seq, * struct net_device *dev, u32 filter_mask) + * int (*ndo_bridge_dellink)(struct net_device *dev, struct nlmsghdr *nlh, + * u16 flags); * * int (*ndo_change_carrier)(struct net_device *dev, bool new_carrier); * Called to change device carrier. Soft-devices (like dummy, team, etc) -- cgit v1.2.1 From 37355565ba57fd45f78f0934305be2761b641f8f Mon Sep 17 00:00:00 2001 From: Nicolas Dichtel Date: Mon, 16 Mar 2015 15:56:05 +0100 Subject: ip6_tunnel: fix error code when tunnel exists After commit 2b0bb01b6edb, the kernel returns -ENOBUFS when user tries to add an existing tunnel with ioctl API: $ ip -6 tunnel add ip6tnl1 mode ip6ip6 dev eth1 add tunnel "ip6tnl0" failed: No buffer space available It's confusing, the right error is EEXIST. This patch also change a bit the code returned: - ENOBUFS -> ENOMEM - ENOENT -> ENODEV Fixes: 2b0bb01b6edb ("ip6_tunnel: Return an error when adding an existing tunnel.") CC: Steffen Klassert Reported-by: Pierre Cheynier Signed-off-by: Nicolas Dichtel Signed-off-by: David S. Miller --- net/ipv6/ip6_tunnel.c | 33 +++++++++++++++++---------------- 1 file changed, 17 insertions(+), 16 deletions(-) diff --git a/net/ipv6/ip6_tunnel.c b/net/ipv6/ip6_tunnel.c index 266a264ec212..ddd94eca19b3 100644 --- a/net/ipv6/ip6_tunnel.c +++ b/net/ipv6/ip6_tunnel.c @@ -314,7 +314,7 @@ out: * Create tunnel matching given parameters. * * Return: - * created tunnel or NULL + * created tunnel or error pointer **/ static struct ip6_tnl *ip6_tnl_create(struct net *net, struct __ip6_tnl_parm *p) @@ -322,7 +322,7 @@ static struct ip6_tnl *ip6_tnl_create(struct net *net, struct __ip6_tnl_parm *p) struct net_device *dev; struct ip6_tnl *t; char name[IFNAMSIZ]; - int err; + int err = -ENOMEM; if (p->name[0]) strlcpy(name, p->name, IFNAMSIZ); @@ -348,7 +348,7 @@ static struct ip6_tnl *ip6_tnl_create(struct net *net, struct __ip6_tnl_parm *p) failed_free: ip6_dev_free(dev); failed: - return NULL; + return ERR_PTR(err); } /** @@ -362,7 +362,7 @@ failed: * tunnel device is created and registered for use. * * Return: - * matching tunnel or NULL + * matching tunnel or error pointer **/ static struct ip6_tnl *ip6_tnl_locate(struct net *net, @@ -380,13 +380,13 @@ static struct ip6_tnl *ip6_tnl_locate(struct net *net, if (ipv6_addr_equal(local, &t->parms.laddr) && ipv6_addr_equal(remote, &t->parms.raddr)) { if (create) - return NULL; + return ERR_PTR(-EEXIST); return t; } } if (!create) - return NULL; + return ERR_PTR(-ENODEV); return ip6_tnl_create(net, p); } @@ -1420,7 +1420,7 @@ ip6_tnl_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) } ip6_tnl_parm_from_user(&p1, &p); t = ip6_tnl_locate(net, &p1, 0); - if (t == NULL) + if (IS_ERR(t)) t = netdev_priv(dev); } else { memset(&p, 0, sizeof(p)); @@ -1445,7 +1445,7 @@ ip6_tnl_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) ip6_tnl_parm_from_user(&p1, &p); t = ip6_tnl_locate(net, &p1, cmd == SIOCADDTUNNEL); if (cmd == SIOCCHGTUNNEL) { - if (t != NULL) { + if (!IS_ERR(t)) { if (t->dev != dev) { err = -EEXIST; break; @@ -1457,14 +1457,15 @@ ip6_tnl_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) else err = ip6_tnl_update(t, &p1); } - if (t) { + if (!IS_ERR(t)) { err = 0; ip6_tnl_parm_to_user(&p, &t->parms); if (copy_to_user(ifr->ifr_ifru.ifru_data, &p, sizeof(p))) err = -EFAULT; - } else - err = (cmd == SIOCADDTUNNEL ? -ENOBUFS : -ENOENT); + } else { + err = PTR_ERR(t); + } break; case SIOCDELTUNNEL: err = -EPERM; @@ -1478,7 +1479,7 @@ ip6_tnl_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd) err = -ENOENT; ip6_tnl_parm_from_user(&p1, &p); t = ip6_tnl_locate(net, &p1, 0); - if (t == NULL) + if (IS_ERR(t)) break; err = -EPERM; if (t->dev == ip6n->fb_tnl_dev) @@ -1672,12 +1673,13 @@ static int ip6_tnl_newlink(struct net *src_net, struct net_device *dev, struct nlattr *tb[], struct nlattr *data[]) { struct net *net = dev_net(dev); - struct ip6_tnl *nt; + struct ip6_tnl *nt, *t; nt = netdev_priv(dev); ip6_tnl_netlink_parms(data, &nt->parms); - if (ip6_tnl_locate(net, &nt->parms, 0)) + t = ip6_tnl_locate(net, &nt->parms, 0); + if (!IS_ERR(t)) return -EEXIST; return ip6_tnl_create2(dev); @@ -1697,8 +1699,7 @@ static int ip6_tnl_changelink(struct net_device *dev, struct nlattr *tb[], ip6_tnl_netlink_parms(data, &p); t = ip6_tnl_locate(net, &p, 0); - - if (t) { + if (!IS_ERR(t)) { if (t->dev != dev) return -EEXIST; } else -- cgit v1.2.1 From cb7cf8a33ff73cf638481d1edf883d8968f934f8 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Mon, 16 Mar 2015 12:19:24 -0700 Subject: inet: Clean up inet_csk_wait_for_connect() vs. might_sleep() I got the following trace with current net-next kernel : [14723.885290] WARNING: CPU: 26 PID: 22658 at kernel/sched/core.c:7285 __might_sleep+0x89/0xa0() [14723.885325] do not call blocking ops when !TASK_RUNNING; state=1 set at [] prepare_to_wait_exclusive+0x34/0xa0 [14723.885355] CPU: 26 PID: 22658 Comm: netserver Not tainted 4.0.0-dbg-DEV #1379 [14723.885359] ffffffff81a223a8 ffff881fae9e7ca8 ffffffff81650b5d 0000000000000001 [14723.885364] ffff881fae9e7cf8 ffff881fae9e7ce8 ffffffff810a72e7 0000000000000000 [14723.885367] ffffffff81a57620 000000000000093a 0000000000000000 ffff881fae9e7e64 [14723.885371] Call Trace: [14723.885377] [] dump_stack+0x4c/0x65 [14723.885382] [] warn_slowpath_common+0x97/0xe0 [14723.885386] [] warn_slowpath_fmt+0x46/0x50 [14723.885390] [] ? trace_hardirqs_on_caller+0x10d/0x1d0 [14723.885393] [] ? prepare_to_wait_exclusive+0x34/0xa0 [14723.885396] [] ? prepare_to_wait_exclusive+0x34/0xa0 [14723.885399] [] __might_sleep+0x89/0xa0 [14723.885403] [] lock_sock_nested+0x36/0xb0 [14723.885406] [] ? release_sock+0x173/0x1c0 [14723.885411] [] inet_csk_accept+0x157/0x2a0 [14723.885415] [] ? abort_exclusive_wait+0xc0/0xc0 [14723.885419] [] inet_accept+0x2d/0x150 [14723.885424] [] SYSC_accept4+0xff/0x210 [14723.885428] [] ? retint_swapgs+0xe/0x44 [14723.885431] [] ? trace_hardirqs_on_caller+0x10d/0x1d0 [14723.885437] [] ? trace_hardirqs_on_thunk+0x3a/0x3f [14723.885441] [] SyS_accept+0x10/0x20 [14723.885444] [] system_call_fastpath+0x12/0x17 [14723.885447] ---[ end trace ff74cd83355b1873 ]--- In commit 26cabd31259ba43f68026ce3f62b78094124333f Peter added a sched_annotate_sleep() in sk_wait_event() Is the following patch needed as well ? Alternative would be to use sk_wait_event() from inet_csk_wait_for_connect() Signed-off-by: Eric Dumazet Acked-by: Peter Zijlstra (Intel) Signed-off-by: David S. Miller --- net/ipv4/inet_connection_sock.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/ipv4/inet_connection_sock.c b/net/ipv4/inet_connection_sock.c index 14d02ea905b6..3e44b9b0b78e 100644 --- a/net/ipv4/inet_connection_sock.c +++ b/net/ipv4/inet_connection_sock.c @@ -268,6 +268,7 @@ static int inet_csk_wait_for_connect(struct sock *sk, long timeo) release_sock(sk); if (reqsk_queue_empty(&icsk->icsk_accept_queue)) timeo = schedule_timeout(timeo); + sched_annotate_sleep(); lock_sock(sk); err = 0; if (!reqsk_queue_empty(&icsk->icsk_accept_queue)) -- cgit v1.2.1 From 8d7d9cca4390062ccd09ffd9fdb37d1c4eeea9ac Mon Sep 17 00:00:00 2001 From: Robert Jarzmik Date: Thu, 12 Feb 2015 17:59:10 +0100 Subject: Revert "smc91x: retrieve IRQ and trigger flags in a modern way" The commit breaks the legacy platforms, ie. these not using device-tree, and setting up the interrupt resources with a flag to activate edge detection. The issue was found on the zylonite platform. The reason is that zylonite uses platform resources to pass the interrupt number and the irq flags (here IORESOURCE_IRQ_HIGHEDGE). It expects the driver to request the irq with these flags, which in turn setups the irq as high edge triggered. After the patch, this was supposed to be taken care of with : irq_resflags = irqd_get_trigger_type(irq_get_irq_data(ndev->irq)); But irq_resflags is 0 for legacy platforms, while for example in arch/arm/mach-pxa/zylonite.c, in struct resource smc91x_resources[] the irq flag is specified. This breaks zylonite because the interrupt is not setup as triggered, and hardware doesn't provide interrupts. Signed-off-by: Robert Jarzmik Signed-off-by: David S. Miller --- drivers/net/ethernet/smsc/smc91x.c | 20 ++++++++------------ 1 file changed, 8 insertions(+), 12 deletions(-) diff --git a/drivers/net/ethernet/smsc/smc91x.c b/drivers/net/ethernet/smsc/smc91x.c index 5d093dc0f5f5..8678e39aba08 100644 --- a/drivers/net/ethernet/smsc/smc91x.c +++ b/drivers/net/ethernet/smsc/smc91x.c @@ -2248,10 +2248,9 @@ static int smc_drv_probe(struct platform_device *pdev) const struct of_device_id *match = NULL; struct smc_local *lp; struct net_device *ndev; - struct resource *res; + struct resource *res, *ires; unsigned int __iomem *addr; unsigned long irq_flags = SMC_IRQ_FLAGS; - unsigned long irq_resflags; int ret; ndev = alloc_etherdev(sizeof(struct smc_local)); @@ -2343,19 +2342,16 @@ static int smc_drv_probe(struct platform_device *pdev) goto out_free_netdev; } - ndev->irq = platform_get_irq(pdev, 0); - if (ndev->irq <= 0) { + ires = platform_get_resource(pdev, IORESOURCE_IRQ, 0); + if (!ires) { ret = -ENODEV; goto out_release_io; } - /* - * If this platform does not specify any special irqflags, or if - * the resource supplies a trigger, override the irqflags with - * the trigger flags from the resource. - */ - irq_resflags = irqd_get_trigger_type(irq_get_irq_data(ndev->irq)); - if (irq_flags == -1 || irq_resflags & IRQF_TRIGGER_MASK) - irq_flags = irq_resflags & IRQF_TRIGGER_MASK; + + ndev->irq = ires->start; + + if (irq_flags == -1 || ires->flags & IRQF_TRIGGER_MASK) + irq_flags = ires->flags & IRQF_TRIGGER_MASK; ret = smc_request_attrib(pdev, ndev); if (ret) -- cgit v1.2.1 From e2c7d8877e5caa2356b5bc8207535e83b126f653 Mon Sep 17 00:00:00 2001 From: Benjamin Tissoires Date: Thu, 5 Mar 2015 17:36:35 -0500 Subject: HID: wacom: check for wacom->shared before following the pointer 486b908 (HID: wacom: do not send pen events before touch is up/forced out) introduces a kernel oops when plugging a tablet without touch. wacom->shared is null for these devices so this leads to a null pointer exception. Change the condition to make it clear that what we need is wacom->shared not NULL. Signed-off-by: Benjamin Tissoires Signed-off-by: Jiri Kosina --- drivers/hid/wacom_wac.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/drivers/hid/wacom_wac.c b/drivers/hid/wacom_wac.c index cf767419cdc4..bbe32d66e500 100644 --- a/drivers/hid/wacom_wac.c +++ b/drivers/hid/wacom_wac.c @@ -551,11 +551,12 @@ static int wacom_intuos_inout(struct wacom_wac *wacom) (features->type == CINTIQ && !(data[1] & 0x40))) return 1; - if (features->quirks & WACOM_QUIRK_MULTI_INPUT) + if (wacom->shared) { wacom->shared->stylus_in_proximity = true; - if (wacom->shared->touch_down) - return 1; + if (wacom->shared->touch_down) + return 1; + } /* in Range while exiting */ if (((data[1] & 0xfe) == 0x20) && wacom->reporting_data) { -- cgit v1.2.1 From ba117213554bc747561c5b7bf274d60ac93b8598 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Fri, 13 Mar 2015 15:01:24 -0400 Subject: Btrfs: account merges/splits properly My fix Btrfs: fix merge delalloc logic only fixed half of the problems, it didn't fix the case where we have two large extents on either side and then join them together with a new small extent. We need to instead keep track of how many extents we have accounted for with each side of the new extent, and then see how many extents we need for the new large extent. If they match then we know we need to keep our reservation, otherwise we need to drop our reservation. This shows up with a case like this [BTRFS_MAX_EXTENT_SIZE+4K][4K HOLE][BTRFS_MAX_EXTENT_SIZE+4K] Previously the logic would have said that the number extents required for the new size (3) is larger than the number of extents required for the largest side (2) therefore we need to keep our reservation. But this isn't the case, since both sides require a reservation of 2 which leads to 4 for the whole range currently reserved, but we only need 3, so we need to drop one of the reservations. The same problem existed for splits, we'd think we only need 3 extents when creating the hole but in reality we need 4. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/inode.c | 57 +++++++++++++++++++++++++++++--------------------------- 1 file changed, 30 insertions(+), 27 deletions(-) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 97b601bec326..bb74a4181075 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -1542,30 +1542,17 @@ static void btrfs_split_extent_hook(struct inode *inode, u64 new_size; /* - * We need the largest size of the remaining extent to see if we - * need to add a new outstanding extent. Think of the following - * case - * - * [MEAX_EXTENT_SIZEx2 - 4k][4k] - * - * The new_size would just be 4k and we'd think we had enough - * outstanding extents for this if we only took one side of the - * split, same goes for the other direction. We need to see if - * the larger size still is the same amount of extents as the - * original size, because if it is we need to add a new - * outstanding extent. But if we split up and the larger size - * is less than the original then we are good to go since we've - * already accounted for the extra extent in our original - * accounting. + * See the explanation in btrfs_merge_extent_hook, the same + * applies here, just in reverse. */ new_size = orig->end - split + 1; - if ((split - orig->start) > new_size) - new_size = split - orig->start; - - num_extents = div64_u64(size + BTRFS_MAX_EXTENT_SIZE - 1, + num_extents = div64_u64(new_size + BTRFS_MAX_EXTENT_SIZE - 1, BTRFS_MAX_EXTENT_SIZE); - if (div64_u64(new_size + BTRFS_MAX_EXTENT_SIZE - 1, - BTRFS_MAX_EXTENT_SIZE) < num_extents) + new_size = split - orig->start; + num_extents += div64_u64(new_size + BTRFS_MAX_EXTENT_SIZE - 1, + BTRFS_MAX_EXTENT_SIZE); + if (div64_u64(size + BTRFS_MAX_EXTENT_SIZE - 1, + BTRFS_MAX_EXTENT_SIZE) >= num_extents) return; } @@ -1591,9 +1578,6 @@ static void btrfs_merge_extent_hook(struct inode *inode, if (!(other->state & EXTENT_DELALLOC)) return; - old_size = other->end - other->start + 1; - if (old_size < (new->end - new->start + 1)) - old_size = (new->end - new->start + 1); if (new->start > other->start) new_size = new->end - other->start + 1; else @@ -1608,13 +1592,32 @@ static void btrfs_merge_extent_hook(struct inode *inode, } /* - * If we grew by another max_extent, just return, we want to keep that - * reserved amount. + * We have to add up either side to figure out how many extents were + * accounted for before we merged into one big extent. If the number of + * extents we accounted for is <= the amount we need for the new range + * then we can return, otherwise drop. Think of it like this + * + * [ 4k][MAX_SIZE] + * + * So we've grown the extent by a MAX_SIZE extent, this would mean we + * need 2 outstanding extents, on one side we have 1 and the other side + * we have 1 so they are == and we can return. But in this case + * + * [MAX_SIZE+4k][MAX_SIZE+4k] + * + * Each range on their own accounts for 2 extents, but merged together + * they are only 3 extents worth of accounting, so we need to drop in + * this case. */ + old_size = other->end - other->start + 1; num_extents = div64_u64(old_size + BTRFS_MAX_EXTENT_SIZE - 1, BTRFS_MAX_EXTENT_SIZE); + old_size = new->end - new->start + 1; + num_extents += div64_u64(old_size + BTRFS_MAX_EXTENT_SIZE - 1, + BTRFS_MAX_EXTENT_SIZE); + if (div64_u64(new_size + BTRFS_MAX_EXTENT_SIZE - 1, - BTRFS_MAX_EXTENT_SIZE) > num_extents) + BTRFS_MAX_EXTENT_SIZE) >= num_extents) return; spin_lock(&BTRFS_I(inode)->lock); -- cgit v1.2.1 From bcb7e449ec6350121ac7ca138c0b050ba7caca47 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Mon, 16 Mar 2015 17:38:02 -0400 Subject: Btrfs: just free dummy extent buffers If we fail during our sanity tests we could get NULL deref's because we unload the module before the dummy extent buffers are free'd via RCU. So check for this case and just free the things directly. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/extent_io.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 29850d4a3827..d13ceadcbf18 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -4968,6 +4968,12 @@ static int release_extent_buffer(struct extent_buffer *eb) /* Should be safe to release our pages at this point */ btrfs_release_extent_buffer_page(eb); +#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS + if (unlikely(test_bit(EXTENT_BUFFER_DUMMY, &eb->bflags))) { + __free_extent_buffer(eb); + return 1; + } +#endif call_rcu(&eb->rcu_head, btrfs_release_extent_buffer_rcu); return 1; } -- cgit v1.2.1 From 6a3891c551268dd4ff0969b883c4c8b8d974db8f Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Mon, 16 Mar 2015 17:38:52 -0400 Subject: Btrfs: add sanity test for outstanding_extents accounting I introduced a regression wrt outstanding_extents accounting. These are tricky areas that aren't easily covered by xfstests as we could change MAX_EXTENT_SIZE at any time. So add sanity tests to cover the various conditions that are tricky in order to make sure we don't introduce regressions in the future. Thanks, Signed-off-by: Josef Bacik --- fs/btrfs/ctree.h | 3 + fs/btrfs/extent-tree.c | 3 + fs/btrfs/inode.c | 15 ++++ fs/btrfs/tests/inode-tests.c | 197 ++++++++++++++++++++++++++++++++++++++++++- 4 files changed, 217 insertions(+), 1 deletion(-) diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index a0c90a324f53..5bd721a1516e 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -3910,6 +3910,9 @@ int btrfs_prealloc_file_range_trans(struct inode *inode, loff_t actual_len, u64 *alloc_hint); int btrfs_inode_check_errors(struct inode *inode); extern const struct dentry_operations btrfs_dentry_operations; +#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS +void btrfs_test_inode_set_ops(struct inode *inode); +#endif /* ioctl.c */ long btrfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg); diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 3ac3fefb1625..d1ca86ea9993 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -5285,6 +5285,9 @@ void btrfs_delalloc_release_metadata(struct inode *inode, u64 num_bytes) if (dropped > 0) to_free += btrfs_calc_trans_metadata_size(root, dropped); + if (btrfs_test_is_dummy_root(root)) + return; + trace_btrfs_space_reservation(root->fs_info, "delalloc", btrfs_ino(inode), to_free, 0); if (root->fs_info->quota_enabled) { diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index bb74a4181075..3717b3d4c2ce 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -108,6 +108,13 @@ static struct extent_map *create_pinned_em(struct inode *inode, u64 start, static int btrfs_dirty_inode(struct inode *inode); +#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS +void btrfs_test_inode_set_ops(struct inode *inode) +{ + BTRFS_I(inode)->io_tree.ops = &btrfs_extent_io_ops; +} +#endif + static int btrfs_init_inode_security(struct btrfs_trans_handle *trans, struct inode *inode, struct inode *dir, const struct qstr *qstr) @@ -1694,6 +1701,10 @@ static void btrfs_set_bit_hook(struct inode *inode, spin_unlock(&BTRFS_I(inode)->lock); } + /* For sanity tests */ + if (btrfs_test_is_dummy_root(root)) + return; + __percpu_counter_add(&root->fs_info->delalloc_bytes, len, root->fs_info->delalloc_batch); spin_lock(&BTRFS_I(inode)->lock); @@ -1749,6 +1760,10 @@ static void btrfs_clear_bit_hook(struct inode *inode, root != root->fs_info->tree_root) btrfs_delalloc_release_metadata(inode, len); + /* For sanity tests. */ + if (btrfs_test_is_dummy_root(root)) + return; + if (root->root_key.objectid != BTRFS_DATA_RELOC_TREE_OBJECTID && do_list && !(state->state & EXTENT_NORESERVE)) btrfs_free_reserved_data_space(inode, len); diff --git a/fs/btrfs/tests/inode-tests.c b/fs/btrfs/tests/inode-tests.c index a116b55ce788..054fc0d97131 100644 --- a/fs/btrfs/tests/inode-tests.c +++ b/fs/btrfs/tests/inode-tests.c @@ -911,6 +911,197 @@ out: return ret; } +static int test_extent_accounting(void) +{ + struct inode *inode = NULL; + struct btrfs_root *root = NULL; + int ret = -ENOMEM; + + inode = btrfs_new_test_inode(); + if (!inode) { + test_msg("Couldn't allocate inode\n"); + return ret; + } + + root = btrfs_alloc_dummy_root(); + if (IS_ERR(root)) { + test_msg("Couldn't allocate root\n"); + goto out; + } + + root->fs_info = btrfs_alloc_dummy_fs_info(); + if (!root->fs_info) { + test_msg("Couldn't allocate dummy fs info\n"); + goto out; + } + + BTRFS_I(inode)->root = root; + btrfs_test_inode_set_ops(inode); + + /* [BTRFS_MAX_EXTENT_SIZE] */ + BTRFS_I(inode)->outstanding_extents++; + ret = btrfs_set_extent_delalloc(inode, 0, BTRFS_MAX_EXTENT_SIZE - 1, + NULL); + if (ret) { + test_msg("btrfs_set_extent_delalloc returned %d\n", ret); + goto out; + } + if (BTRFS_I(inode)->outstanding_extents != 1) { + ret = -EINVAL; + test_msg("Miscount, wanted 1, got %u\n", + BTRFS_I(inode)->outstanding_extents); + goto out; + } + + /* [BTRFS_MAX_EXTENT_SIZE][4k] */ + BTRFS_I(inode)->outstanding_extents++; + ret = btrfs_set_extent_delalloc(inode, BTRFS_MAX_EXTENT_SIZE, + BTRFS_MAX_EXTENT_SIZE + 4095, NULL); + if (ret) { + test_msg("btrfs_set_extent_delalloc returned %d\n", ret); + goto out; + } + if (BTRFS_I(inode)->outstanding_extents != 2) { + ret = -EINVAL; + test_msg("Miscount, wanted 2, got %u\n", + BTRFS_I(inode)->outstanding_extents); + goto out; + } + + /* [BTRFS_MAX_EXTENT_SIZE/2][4K HOLE][the rest] */ + ret = clear_extent_bit(&BTRFS_I(inode)->io_tree, + BTRFS_MAX_EXTENT_SIZE >> 1, + (BTRFS_MAX_EXTENT_SIZE >> 1) + 4095, + EXTENT_DELALLOC | EXTENT_DIRTY | + EXTENT_UPTODATE | EXTENT_DO_ACCOUNTING, 0, 0, + NULL, GFP_NOFS); + if (ret) { + test_msg("clear_extent_bit returned %d\n", ret); + goto out; + } + if (BTRFS_I(inode)->outstanding_extents != 2) { + ret = -EINVAL; + test_msg("Miscount, wanted 2, got %u\n", + BTRFS_I(inode)->outstanding_extents); + goto out; + } + + /* [BTRFS_MAX_EXTENT_SIZE][4K] */ + BTRFS_I(inode)->outstanding_extents++; + ret = btrfs_set_extent_delalloc(inode, BTRFS_MAX_EXTENT_SIZE >> 1, + (BTRFS_MAX_EXTENT_SIZE >> 1) + 4095, + NULL); + if (ret) { + test_msg("btrfs_set_extent_delalloc returned %d\n", ret); + goto out; + } + if (BTRFS_I(inode)->outstanding_extents != 2) { + ret = -EINVAL; + test_msg("Miscount, wanted 2, got %u\n", + BTRFS_I(inode)->outstanding_extents); + goto out; + } + + /* + * [BTRFS_MAX_EXTENT_SIZE+4K][4K HOLE][BTRFS_MAX_EXTENT_SIZE+4K] + * + * I'm artificially adding 2 to outstanding_extents because in the + * buffered IO case we'd add things up as we go, but I don't feel like + * doing that here, this isn't the interesting case we want to test. + */ + BTRFS_I(inode)->outstanding_extents += 2; + ret = btrfs_set_extent_delalloc(inode, BTRFS_MAX_EXTENT_SIZE + 8192, + (BTRFS_MAX_EXTENT_SIZE << 1) + 12287, + NULL); + if (ret) { + test_msg("btrfs_set_extent_delalloc returned %d\n", ret); + goto out; + } + if (BTRFS_I(inode)->outstanding_extents != 4) { + ret = -EINVAL; + test_msg("Miscount, wanted 4, got %u\n", + BTRFS_I(inode)->outstanding_extents); + goto out; + } + + /* [BTRFS_MAX_EXTENT_SIZE+4k][4k][BTRFS_MAX_EXTENT_SIZE+4k] */ + BTRFS_I(inode)->outstanding_extents++; + ret = btrfs_set_extent_delalloc(inode, BTRFS_MAX_EXTENT_SIZE+4096, + BTRFS_MAX_EXTENT_SIZE+8191, NULL); + if (ret) { + test_msg("btrfs_set_extent_delalloc returned %d\n", ret); + goto out; + } + if (BTRFS_I(inode)->outstanding_extents != 3) { + ret = -EINVAL; + test_msg("Miscount, wanted 3, got %u\n", + BTRFS_I(inode)->outstanding_extents); + goto out; + } + + /* [BTRFS_MAX_EXTENT_SIZE+4k][4K HOLE][BTRFS_MAX_EXTENT_SIZE+4k] */ + ret = clear_extent_bit(&BTRFS_I(inode)->io_tree, + BTRFS_MAX_EXTENT_SIZE+4096, + BTRFS_MAX_EXTENT_SIZE+8191, + EXTENT_DIRTY | EXTENT_DELALLOC | + EXTENT_DO_ACCOUNTING | EXTENT_UPTODATE, 0, 0, + NULL, GFP_NOFS); + if (ret) { + test_msg("clear_extent_bit returned %d\n", ret); + goto out; + } + if (BTRFS_I(inode)->outstanding_extents != 4) { + ret = -EINVAL; + test_msg("Miscount, wanted 4, got %u\n", + BTRFS_I(inode)->outstanding_extents); + goto out; + } + + /* + * Refill the hole again just for good measure, because I thought it + * might fail and I'd rather satisfy my paranoia at this point. + */ + BTRFS_I(inode)->outstanding_extents++; + ret = btrfs_set_extent_delalloc(inode, BTRFS_MAX_EXTENT_SIZE+4096, + BTRFS_MAX_EXTENT_SIZE+8191, NULL); + if (ret) { + test_msg("btrfs_set_extent_delalloc returned %d\n", ret); + goto out; + } + if (BTRFS_I(inode)->outstanding_extents != 3) { + ret = -EINVAL; + test_msg("Miscount, wanted 3, got %u\n", + BTRFS_I(inode)->outstanding_extents); + goto out; + } + + /* Empty */ + ret = clear_extent_bit(&BTRFS_I(inode)->io_tree, 0, (u64)-1, + EXTENT_DIRTY | EXTENT_DELALLOC | + EXTENT_DO_ACCOUNTING | EXTENT_UPTODATE, 0, 0, + NULL, GFP_NOFS); + if (ret) { + test_msg("clear_extent_bit returned %d\n", ret); + goto out; + } + if (BTRFS_I(inode)->outstanding_extents) { + ret = -EINVAL; + test_msg("Miscount, wanted 0, got %u\n", + BTRFS_I(inode)->outstanding_extents); + goto out; + } + ret = 0; +out: + if (ret) + clear_extent_bit(&BTRFS_I(inode)->io_tree, 0, (u64)-1, + EXTENT_DIRTY | EXTENT_DELALLOC | + EXTENT_DO_ACCOUNTING | EXTENT_UPTODATE, 0, 0, + NULL, GFP_NOFS); + iput(inode); + btrfs_free_dummy_root(root); + return ret; +} + int btrfs_test_inodes(void) { int ret; @@ -924,5 +1115,9 @@ int btrfs_test_inodes(void) if (ret) return ret; test_msg("Running hole first btrfs_get_extent test\n"); - return test_hole_first(); + ret = test_hole_first(); + if (ret) + return ret; + test_msg("Running outstanding_extents tests\n"); + return test_extent_accounting(); } -- cgit v1.2.1 From e1cbbfa5f5aaf40a1fe70856fac4dfcc33e0e651 Mon Sep 17 00:00:00 2001 From: Josef Bacik Date: Tue, 17 Mar 2015 10:52:28 -0400 Subject: Btrfs: fix outstanding_extents accounting in DIO We are keeping track of how many extents we need to reserve properly based on the amount we want to write, but we were still incrementing outstanding_extents if we wrote less than what we requested. This isn't quite right since we will be limited to our max extent size. So instead lets do something horrible! Keep track of how many outstanding_extents we reserved, and decrement each time we allocate an extent. If we use our entire reserve make sure to jack up outstanding_extents on the inode so the accounting works out properly. Thanks, Reported-by: Filipe Manana Signed-off-by: Josef Bacik --- fs/btrfs/inode.c | 37 +++++++++++++++++++++++++++++++++++-- 1 file changed, 35 insertions(+), 2 deletions(-) diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 3717b3d4c2ce..aa1fb534f69f 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -7239,7 +7239,7 @@ static int btrfs_get_blocks_direct(struct inode *inode, sector_t iblock, u64 start = iblock << inode->i_blkbits; u64 lockstart, lockend; u64 len = bh_result->b_size; - u64 orig_len = len; + u64 *outstanding_extents = NULL; int unlock_bits = EXTENT_LOCKED; int ret = 0; @@ -7251,6 +7251,16 @@ static int btrfs_get_blocks_direct(struct inode *inode, sector_t iblock, lockstart = start; lockend = start + len - 1; + if (current->journal_info) { + /* + * Need to pull our outstanding extents and set journal_info to NULL so + * that anything that needs to check if there's a transction doesn't get + * confused. + */ + outstanding_extents = current->journal_info; + current->journal_info = NULL; + } + /* * If this errors out it's because we couldn't invalidate pagecache for * this range and we need to fallback to buffered. @@ -7374,11 +7384,20 @@ unlock: if (start + len > i_size_read(inode)) i_size_write(inode, start + len); - if (len < orig_len) { + /* + * If we have an outstanding_extents count still set then we're + * within our reservation, otherwise we need to adjust our inode + * counter appropriately. + */ + if (*outstanding_extents) { + (*outstanding_extents)--; + } else { spin_lock(&BTRFS_I(inode)->lock); BTRFS_I(inode)->outstanding_extents++; spin_unlock(&BTRFS_I(inode)->lock); } + + current->journal_info = outstanding_extents; btrfs_free_reserved_data_space(inode, len); } @@ -7402,6 +7421,8 @@ unlock: unlock_err: clear_extent_bit(&BTRFS_I(inode)->io_tree, lockstart, lockend, unlock_bits, 1, 0, &cached_state, GFP_NOFS); + if (outstanding_extents) + current->journal_info = outstanding_extents; return ret; } @@ -8101,6 +8122,7 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, { struct file *file = iocb->ki_filp; struct inode *inode = file->f_mapping->host; + u64 outstanding_extents = 0; size_t count = 0; int flags = 0; bool wakeup = true; @@ -8138,6 +8160,16 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, ret = btrfs_delalloc_reserve_space(inode, count); if (ret) goto out; + outstanding_extents = div64_u64(count + + BTRFS_MAX_EXTENT_SIZE - 1, + BTRFS_MAX_EXTENT_SIZE); + + /* + * We need to know how many extents we reserved so that we can + * do the accounting properly if we go over the number we + * originally calculated. Abuse current->journal_info for this. + */ + current->journal_info = &outstanding_extents; } else if (test_bit(BTRFS_INODE_READDIO_NEED_LOCK, &BTRFS_I(inode)->runtime_flags)) { inode_dio_done(inode); @@ -8150,6 +8182,7 @@ static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb, iter, offset, btrfs_get_blocks_direct, NULL, btrfs_submit_direct, flags); if (rw & WRITE) { + current->journal_info = NULL; if (ret < 0 && ret != -EIOCBQUEUED) btrfs_delalloc_release_space(inode, count); else if (ret >= 0 && (size_t)ret < count) -- cgit v1.2.1 From 886016835a87757839baced91a7ff09e9636b6b1 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Tue, 17 Mar 2015 16:38:10 +0100 Subject: rtc: at91rm9200: double locking bug in at91_rtc_interrupt() There is a typo here so we deadlock. Fixes: dd1f1f391dd7 ('rtc: at91rm9200: rework wakeup and interrupt handling') Signed-off-by: Dan Carpenter Acked-by: Boris Brezillon Reported-by: David Dueck Signed-off-by: Nicolas Ferre Signed-off-by: Rafael J. Wysocki --- drivers/rtc/rtc-at91rm9200.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/rtc/rtc-at91rm9200.c b/drivers/rtc/rtc-at91rm9200.c index b4f7744f6751..b283a1a573b3 100644 --- a/drivers/rtc/rtc-at91rm9200.c +++ b/drivers/rtc/rtc-at91rm9200.c @@ -324,7 +324,7 @@ static irqreturn_t at91_rtc_interrupt(int irq, void *dev_id) ret = IRQ_HANDLED; } - spin_lock(&suspended_lock); + spin_unlock(&suspended_lock); return ret; } -- cgit v1.2.1 From bd8733738c5af6114dd15d340b3f8713e9b624c2 Mon Sep 17 00:00:00 2001 From: Boris Brezillon Date: Sun, 8 Feb 2015 19:23:42 +0100 Subject: pinctrl: at91: move lock/unlock_as_irq calls into request/release The gpiochip_lock_as_irq call can fail and return an error, while the irq_startup is not expected to fail (returns an unsigned int which is not checked by irq core code). irq_request/release_resources functions have been created to address this problem. Move gpiochip_lock/unlock_as_irq calls into irq_request/release_resources functions to prevent using a gpio as an irq if the gpiochip_lock_as_irq call failed. Signed-off-by: Boris Brezillon Acked-by: Ludovic Desroches Acked-by: Nicolas Ferre Acked-by: Jean-Christophe PLAGNIOL-VILLARD Signed-off-by: Linus Walleij --- drivers/pinctrl/pinctrl-at91.c | 17 +++++++---------- 1 file changed, 7 insertions(+), 10 deletions(-) diff --git a/drivers/pinctrl/pinctrl-at91.c b/drivers/pinctrl/pinctrl-at91.c index f4cd0b9b2438..a4814066ea08 100644 --- a/drivers/pinctrl/pinctrl-at91.c +++ b/drivers/pinctrl/pinctrl-at91.c @@ -1477,28 +1477,25 @@ static void gpio_irq_ack(struct irq_data *d) /* the interrupt is already cleared before by reading ISR */ } -static unsigned int gpio_irq_startup(struct irq_data *d) +static int gpio_irq_request_res(struct irq_data *d) { struct at91_gpio_chip *at91_gpio = irq_data_get_irq_chip_data(d); unsigned pin = d->hwirq; int ret; ret = gpiochip_lock_as_irq(&at91_gpio->chip, pin); - if (ret) { + if (ret) dev_err(at91_gpio->chip.dev, "unable to lock pind %lu IRQ\n", d->hwirq); - return ret; - } - gpio_irq_unmask(d); - return 0; + + return ret; } -static void gpio_irq_shutdown(struct irq_data *d) +static void gpio_irq_release_res(struct irq_data *d) { struct at91_gpio_chip *at91_gpio = irq_data_get_irq_chip_data(d); unsigned pin = d->hwirq; - gpio_irq_mask(d); gpiochip_unlock_as_irq(&at91_gpio->chip, pin); } @@ -1577,8 +1574,8 @@ void at91_pinctrl_gpio_resume(void) static struct irq_chip gpio_irqchip = { .name = "GPIO", .irq_ack = gpio_irq_ack, - .irq_startup = gpio_irq_startup, - .irq_shutdown = gpio_irq_shutdown, + .irq_request_resources = gpio_irq_request_res, + .irq_release_resources = gpio_irq_release_res, .irq_disable = gpio_irq_mask, .irq_mask = gpio_irq_mask, .irq_unmask = gpio_irq_unmask, -- cgit v1.2.1 From 0790ec172de1bd2e23f1dbd4925426b6cc3c1b72 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Radim=20Kr=C4=8Dm=C3=A1=C5=99?= Date: Tue, 17 Mar 2015 14:02:32 +0100 Subject: KVM: nVMX: mask unrestricted_guest if disabled on L0 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit If EPT was enabled, unrestricted_guest was allowed in L1 regardless of L0. L1 triple faulted when running L2 guest that required emulation. Another side effect was 'WARN_ON_ONCE(vmx->nested.nested_run_pending)' in L0's dmesg: WARNING: CPU: 0 PID: 0 at arch/x86/kvm/vmx.c:9190 nested_vmx_vmexit+0x96e/0xb00 [kvm_intel] () Prevent this scenario by masking SECONDARY_EXEC_UNRESTRICTED_GUEST when the host doesn't have it enabled. Fixes: 78051e3b7e35 ("KVM: nVMX: Disable unrestricted mode if ept=0") Cc: stable@vger.kernel.org Tested-By: Kashyap Chamarthy Signed-off-by: Radim Krčmář Signed-off-by: Marcelo Tosatti --- arch/x86/kvm/vmx.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/arch/x86/kvm/vmx.c b/arch/x86/kvm/vmx.c index 10a481b7674d..ae4f6d35d19c 100644 --- a/arch/x86/kvm/vmx.c +++ b/arch/x86/kvm/vmx.c @@ -2479,8 +2479,7 @@ static void nested_vmx_setup_ctls_msrs(struct vcpu_vmx *vmx) if (enable_ept) { /* nested EPT: emulate EPT also to L1 */ vmx->nested.nested_vmx_secondary_ctls_high |= - SECONDARY_EXEC_ENABLE_EPT | - SECONDARY_EXEC_UNRESTRICTED_GUEST; + SECONDARY_EXEC_ENABLE_EPT; vmx->nested.nested_vmx_ept_caps = VMX_EPT_PAGE_WALK_4_BIT | VMX_EPTP_WB_BIT | VMX_EPT_2MB_PAGE_BIT | VMX_EPT_INVEPT_BIT; @@ -2494,6 +2493,10 @@ static void nested_vmx_setup_ctls_msrs(struct vcpu_vmx *vmx) } else vmx->nested.nested_vmx_ept_caps = 0; + if (enable_unrestricted_guest) + vmx->nested.nested_vmx_secondary_ctls_high |= + SECONDARY_EXEC_UNRESTRICTED_GUEST; + /* miscellaneous data */ rdmsr(MSR_IA32_VMX_MISC, vmx->nested.nested_vmx_misc_low, -- cgit v1.2.1 From ced585c83b27deca427c606a34dd3eaa6b96d82b Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Tue, 17 Mar 2015 20:25:57 +0100 Subject: act_bpf: allow non-default TC_ACT opcodes as BPF exec outcome Revisiting commit d23b8ad8ab23 ("tc: add BPF based action") with regards to eBPF support, I was thinking that it might be better to improve return semantics from a BPF program invoked through BPF_PROG_RUN(). Currently, in case filter_res is 0, we overwrite the default action opcode with TC_ACT_SHOT. A default action opcode configured through tc's m_bpf can be: TC_ACT_RECLASSIFY, TC_ACT_PIPE, TC_ACT_SHOT, TC_ACT_UNSPEC, TC_ACT_OK. In cls_bpf, we have the possibility to overwrite the default class associated with the classifier in case filter_res is _not_ 0xffffffff (-1). That allows us to fold multiple [e]BPF programs into a single one, where they would otherwise need to be defined as a separate classifier with its own classid, needlessly redoing parsing work, etc. Similarly, we could do better in act_bpf: Since above TC_ACT* opcodes are exported to UAPI anyway, we reuse them for return-code-to-tc-opcode mapping, where we would allow above possibilities. Thus, like in cls_bpf, a filter_res of 0xffffffff (-1) means that the configured _default_ action is used. Any unkown return code from the BPF program would fail in tcf_bpf() with TC_ACT_UNSPEC. Should we one day want to make use of TC_ACT_STOLEN or TC_ACT_QUEUED, which both have the same semantics, we have the option to either use that as a default action (filter_res of 0xffffffff) or non-default BPF return code. All that will allow us to transparently use tcf_bpf() for both BPF flavours. Signed-off-by: Daniel Borkmann Cc: Jiri Pirko Cc: Alexei Starovoitov Cc: Jamal Hadi Salim Acked-by: Jiri Pirko Signed-off-by: David S. Miller --- net/sched/act_bpf.c | 36 ++++++++++++++++++++++++++++-------- 1 file changed, 28 insertions(+), 8 deletions(-) diff --git a/net/sched/act_bpf.c b/net/sched/act_bpf.c index 82c5d7fc1988..5f6288fa3f12 100644 --- a/net/sched/act_bpf.c +++ b/net/sched/act_bpf.c @@ -25,21 +25,41 @@ static int tcf_bpf(struct sk_buff *skb, const struct tc_action *a, struct tcf_result *res) { struct tcf_bpf *b = a->priv; - int action; - int filter_res; + int action, filter_res; spin_lock(&b->tcf_lock); + b->tcf_tm.lastuse = jiffies; bstats_update(&b->tcf_bstats, skb); - action = b->tcf_action; filter_res = BPF_PROG_RUN(b->filter, skb); - if (filter_res == 0) { - /* Return code 0 from the BPF program - * is being interpreted as a drop here. - */ - action = TC_ACT_SHOT; + + /* A BPF program may overwrite the default action opcode. + * Similarly as in cls_bpf, if filter_res == -1 we use the + * default action specified from tc. + * + * In case a different well-known TC_ACT opcode has been + * returned, it will overwrite the default one. + * + * For everything else that is unkown, TC_ACT_UNSPEC is + * returned. + */ + switch (filter_res) { + case TC_ACT_PIPE: + case TC_ACT_RECLASSIFY: + case TC_ACT_OK: + action = filter_res; + break; + case TC_ACT_SHOT: + action = filter_res; b->tcf_qstats.drops++; + break; + case TC_ACT_UNSPEC: + action = b->tcf_action; + break; + default: + action = TC_ACT_UNSPEC; + break; } spin_unlock(&b->tcf_lock); -- cgit v1.2.1 From bead55ef775f6e25a8d286c0d47030580f577bec Mon Sep 17 00:00:00 2001 From: hujianyang Date: Thu, 15 Jan 2015 13:17:36 +0800 Subject: ovl: print error message for invalid mount options Overlayfs should print an error message if an incorrect mount option is caught like other filesystems. After this patch, improper option input could be clearly known. Reported-by: Fabian Sturm Signed-off-by: hujianyang Signed-off-by: Miklos Szeredi --- fs/overlayfs/super.c | 1 + 1 file changed, 1 insertion(+) diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c index b90952f528b1..ab3c8cb8e52d 100644 --- a/fs/overlayfs/super.c +++ b/fs/overlayfs/super.c @@ -615,6 +615,7 @@ static int ovl_parse_opt(char *opt, struct ovl_config *config) break; default: + pr_err("overlayfs: unrecognized mount option \"%s\" or missing value\n", p); return -EINVAL; } } -- cgit v1.2.1 From 6be4506e34cf6075a1307b646e0a6c46c1c9010d Mon Sep 17 00:00:00 2001 From: hujianyang Date: Thu, 15 Jan 2015 13:19:21 +0800 Subject: ovl: check lowerdir amount for non-upper mount Recently multi-lower layer mount support allow upperdir and workdir to be omitted, then cause overlayfs can be mount with only one lowerdir directory. This action make no sense and have potential risk. This patch check the total number of lower directories to prevent mounting overlayfs with only one directory. Also, an error message is added to indicate lower directories exceed OVL_MAX_STACK limit. Signed-off-by: hujianyang Signed-off-by: Miklos Szeredi --- fs/overlayfs/super.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c index ab3c8cb8e52d..edbb3ebcdaad 100644 --- a/fs/overlayfs/super.c +++ b/fs/overlayfs/super.c @@ -870,8 +870,14 @@ static int ovl_fill_super(struct super_block *sb, void *data, int silent) err = -EINVAL; stacklen = ovl_split_lowerdirs(lowertmp); - if (stacklen > OVL_MAX_STACK) + if (stacklen > OVL_MAX_STACK) { + pr_err("overlayfs: too many lower directries, limit is %d\n", + OVL_MAX_STACK); goto out_free_lowertmp; + } else if (!ufs->config.upperdir && stacklen == 1) { + pr_err("overlayfs: at least 2 lowerdir are needed while upperdir nonexistent\n"); + goto out_free_lowertmp; + } stack = kcalloc(stacklen, sizeof(struct path), GFP_KERNEL); if (!stack) -- cgit v1.2.1 From 71cbad7e694ee81233b3be3a38b81c3d5872cc6f Mon Sep 17 00:00:00 2001 From: hujianyang Date: Thu, 15 Jan 2015 13:20:57 +0800 Subject: ovl: upper fs should not be R/O After importing multi-lower layer support, users could mount a r/o partition as the left most lowerdir instead of using it as upperdir. And a r/o upperdir may cause an error like overlayfs: failed to create directory ./workdir/work during mount. This patch check the *s_flags* of upper fs and return an error if it is a r/o partition. The checking of *upper_mnt->mnt_sb->s_flags* can be removed now. This patch also remove /* FIXME: workdir is not needed for a R/O mount */ from ovl_fill_super() because: 1) for upper fs r/o case Setting a r/o partition as upper is prevented, no need to care about workdir in this case. 2) for "mount overlay -o ro" with a r/w upper fs case Users could remount overlayfs to r/w in this case, so workdir should not be omitted. Signed-off-by: hujianyang Signed-off-by: Miklos Szeredi --- fs/overlayfs/super.c | 24 +++++++++++++++++++----- 1 file changed, 19 insertions(+), 5 deletions(-) diff --git a/fs/overlayfs/super.c b/fs/overlayfs/super.c index edbb3ebcdaad..5f0d1993e6e3 100644 --- a/fs/overlayfs/super.c +++ b/fs/overlayfs/super.c @@ -529,8 +529,7 @@ static int ovl_remount(struct super_block *sb, int *flags, char *data) { struct ovl_fs *ufs = sb->s_fs_info; - if (!(*flags & MS_RDONLY) && - (!ufs->upper_mnt || (ufs->upper_mnt->mnt_sb->s_flags & MS_RDONLY))) + if (!(*flags & MS_RDONLY) && !ufs->upper_mnt) return -EROFS; return 0; @@ -619,6 +618,15 @@ static int ovl_parse_opt(char *opt, struct ovl_config *config) return -EINVAL; } } + + /* Workdir is useless in non-upper mount */ + if (!config->upperdir && config->workdir) { + pr_info("overlayfs: option \"workdir=%s\" is useless in a non-upper mount, ignore\n", + config->workdir); + kfree(config->workdir); + config->workdir = NULL; + } + return 0; } @@ -838,7 +846,6 @@ static int ovl_fill_super(struct super_block *sb, void *data, int silent) sb->s_stack_depth = 0; if (ufs->config.upperdir) { - /* FIXME: workdir is not needed for a R/O mount */ if (!ufs->config.workdir) { pr_err("overlayfs: missing 'workdir'\n"); goto out_free_config; @@ -848,6 +855,13 @@ static int ovl_fill_super(struct super_block *sb, void *data, int silent) if (err) goto out_free_config; + /* Upper fs should not be r/o */ + if (upperpath.mnt->mnt_sb->s_flags & MS_RDONLY) { + pr_err("overlayfs: upper fs is r/o, try multi-lower layers mount\n"); + err = -EINVAL; + goto out_put_upperpath; + } + err = ovl_mount_dir(ufs->config.workdir, &workpath); if (err) goto out_put_upperpath; @@ -939,8 +953,8 @@ static int ovl_fill_super(struct super_block *sb, void *data, int silent) ufs->numlower++; } - /* If the upper fs is r/o or nonexistent, we mark overlayfs r/o too */ - if (!ufs->upper_mnt || (ufs->upper_mnt->mnt_sb->s_flags & MS_RDONLY)) + /* If the upper fs is nonexistent, we mark overlayfs r/o too */ + if (!ufs->upper_mnt) sb->s_flags |= MS_RDONLY; sb->s_d_op = &ovl_dentry_operations; -- cgit v1.2.1 From ef6d24cc7f5b2b5c4184eddb039e2add6231a122 Mon Sep 17 00:00:00 2001 From: Hans de Goede Date: Sun, 8 Mar 2015 22:13:57 +0100 Subject: pinctrl: sun4i: GPIOs configured as irq must be set to input before reading On sun4i-a10, when GPIOs are configured as external interrupt the value for them in the data register does not seem to get updated, so set their mux to input (and restore afterwards) when reading the pin. Missed edges seem to be buffered, so this does not introduce a race condition. I've also tested this on sun5i-a13 and sun7i-a20 and those do not seem to be affected, the input value representation in the data register does seem to correctly get updated to the actual pin value while in irq mode there. Signed-off-by: Hans de Goede Acked-by: Maxime Ripard Signed-off-by: Linus Walleij --- drivers/pinctrl/sunxi/pinctrl-sun4i-a10.c | 1 + drivers/pinctrl/sunxi/pinctrl-sunxi.c | 14 ++++++++++++-- drivers/pinctrl/sunxi/pinctrl-sunxi.h | 4 ++++ 3 files changed, 17 insertions(+), 2 deletions(-) diff --git a/drivers/pinctrl/sunxi/pinctrl-sun4i-a10.c b/drivers/pinctrl/sunxi/pinctrl-sun4i-a10.c index 24c5d88f943f..3c68a8e5e0dd 100644 --- a/drivers/pinctrl/sunxi/pinctrl-sun4i-a10.c +++ b/drivers/pinctrl/sunxi/pinctrl-sun4i-a10.c @@ -1011,6 +1011,7 @@ static const struct sunxi_pinctrl_desc sun4i_a10_pinctrl_data = { .pins = sun4i_a10_pins, .npins = ARRAY_SIZE(sun4i_a10_pins), .irq_banks = 1, + .irq_read_needs_mux = true, }; static int sun4i_a10_pinctrl_probe(struct platform_device *pdev) diff --git a/drivers/pinctrl/sunxi/pinctrl-sunxi.c b/drivers/pinctrl/sunxi/pinctrl-sunxi.c index 3d0744337736..f8e171b76693 100644 --- a/drivers/pinctrl/sunxi/pinctrl-sunxi.c +++ b/drivers/pinctrl/sunxi/pinctrl-sunxi.c @@ -29,6 +29,7 @@ #include #include "../core.h" +#include "../../gpio/gpiolib.h" #include "pinctrl-sunxi.h" static struct irq_chip sunxi_pinctrl_edge_irq_chip; @@ -464,10 +465,19 @@ static int sunxi_pinctrl_gpio_direction_input(struct gpio_chip *chip, static int sunxi_pinctrl_gpio_get(struct gpio_chip *chip, unsigned offset) { struct sunxi_pinctrl *pctl = dev_get_drvdata(chip->dev); - u32 reg = sunxi_data_reg(offset); u8 index = sunxi_data_offset(offset); - u32 val = (readl(pctl->membase + reg) >> index) & DATA_PINS_MASK; + u32 set_mux = pctl->desc->irq_read_needs_mux && + test_bit(FLAG_USED_AS_IRQ, &chip->desc[offset].flags); + u32 val; + + if (set_mux) + sunxi_pmx_set(pctl->pctl_dev, offset, SUN4I_FUNC_INPUT); + + val = (readl(pctl->membase + reg) >> index) & DATA_PINS_MASK; + + if (set_mux) + sunxi_pmx_set(pctl->pctl_dev, offset, SUN4I_FUNC_IRQ); return val; } diff --git a/drivers/pinctrl/sunxi/pinctrl-sunxi.h b/drivers/pinctrl/sunxi/pinctrl-sunxi.h index 5a51523a3459..e248e81a0f9e 100644 --- a/drivers/pinctrl/sunxi/pinctrl-sunxi.h +++ b/drivers/pinctrl/sunxi/pinctrl-sunxi.h @@ -77,6 +77,9 @@ #define IRQ_LEVEL_LOW 0x03 #define IRQ_EDGE_BOTH 0x04 +#define SUN4I_FUNC_INPUT 0 +#define SUN4I_FUNC_IRQ 6 + struct sunxi_desc_function { const char *name; u8 muxval; @@ -94,6 +97,7 @@ struct sunxi_pinctrl_desc { int npins; unsigned pin_base; unsigned irq_banks; + bool irq_read_needs_mux; }; struct sunxi_pinctrl_function { -- cgit v1.2.1 From 5c95ed47f1777e9e9b1eb29e48f34e9af3139f29 Mon Sep 17 00:00:00 2001 From: Fabrice Gasnier Date: Thu, 12 Mar 2015 14:04:42 +0100 Subject: ARM: 8310/1: l2c: Fix prefetch settings dt parsing Allow prefetch settings overriding by device tree, in case l2x0_cache_size_of_parse() returns value, prefetch tuning properties are silently ignored. E.g. arm,double-linefill* and arm,prefetch*. This happens for example, when "cache-size" or "cache-sets" properties haven't been filled in l2c dt node. Comments from Fabrice Gasnier: Allow device tree to override the L2C prefetch settings, even when l2x0_cache_size_of_parse() fails to parse the cache geometry due to (eg) missing "cache-size" or "cache-sets" properties. Signed-off-by: Fabrice Gasnier Reviewed-by: Tomasz Figa Signed-off-by: Russell King --- arch/arm/mm/cache-l2x0.c | 33 ++++++++++++++++----------------- 1 file changed, 16 insertions(+), 17 deletions(-) diff --git a/arch/arm/mm/cache-l2x0.c b/arch/arm/mm/cache-l2x0.c index c6c7696b8db9..8f15f70622a6 100644 --- a/arch/arm/mm/cache-l2x0.c +++ b/arch/arm/mm/cache-l2x0.c @@ -1131,23 +1131,22 @@ static void __init l2c310_of_parse(const struct device_node *np, } ret = l2x0_cache_size_of_parse(np, aux_val, aux_mask, &assoc, SZ_512K); - if (ret) - return; - - switch (assoc) { - case 16: - *aux_val &= ~L2X0_AUX_CTRL_ASSOC_MASK; - *aux_val |= L310_AUX_CTRL_ASSOCIATIVITY_16; - *aux_mask &= ~L2X0_AUX_CTRL_ASSOC_MASK; - break; - case 8: - *aux_val &= ~L2X0_AUX_CTRL_ASSOC_MASK; - *aux_mask &= ~L2X0_AUX_CTRL_ASSOC_MASK; - break; - default: - pr_err("L2C-310 OF cache associativity %d invalid, only 8 or 16 permitted\n", - assoc); - break; + if (!ret) { + switch (assoc) { + case 16: + *aux_val &= ~L2X0_AUX_CTRL_ASSOC_MASK; + *aux_val |= L310_AUX_CTRL_ASSOCIATIVITY_16; + *aux_mask &= ~L2X0_AUX_CTRL_ASSOC_MASK; + break; + case 8: + *aux_val &= ~L2X0_AUX_CTRL_ASSOC_MASK; + *aux_mask &= ~L2X0_AUX_CTRL_ASSOC_MASK; + break; + default: + pr_err("L2C-310 OF cache associativity %d invalid, only 8 or 16 permitted\n", + assoc); + break; + } } prefetch = l2x0_saved_regs.prefetch_ctrl; -- cgit v1.2.1 From f2ca09f381a59e1eddb89aa70207740c2ee0fe94 Mon Sep 17 00:00:00 2001 From: Laura Abbott Date: Fri, 13 Mar 2015 21:41:45 +0100 Subject: ARM: 8311/1: Don't use is_module_addr in setting page attributes The set_memory_* functions currently only support module addresses. The addresses are validated using is_module_addr. That function is special though and relies on internal state in the module subsystem to work properly. At the time of module initialization and calling set_memory_*, it's too early for is_module_addr to work properly so it always returns false. Rather than be subject to the whims of the module state, just bounds check against the module virtual address range. Signed-off-by: Laura Abbott Signed-off-by: Russell King --- arch/arm/mm/pageattr.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/arch/arm/mm/pageattr.c b/arch/arm/mm/pageattr.c index 004e35cdcfff..cf30daff8932 100644 --- a/arch/arm/mm/pageattr.c +++ b/arch/arm/mm/pageattr.c @@ -49,7 +49,10 @@ static int change_memory_common(unsigned long addr, int numpages, WARN_ON_ONCE(1); } - if (!is_module_address(start) || !is_module_address(end - 1)) + if (start < MODULES_VADDR || start >= MODULES_END) + return -EINVAL; + + if (end < MODULES_VADDR || start >= MODULES_END) return -EINVAL; data.set_mask = set_mask; -- cgit v1.2.1 From 526299ce4eab2e35ba733b03771d112147676b12 Mon Sep 17 00:00:00 2001 From: Mason Date: Tue, 17 Mar 2015 21:37:25 +0100 Subject: ARM: 8313/1: Use read_cpuid_ext() macro instead of inline asm Replace inline asm statement in __get_cpu_architecture() with equivalent macro invocation, i.e. read_cpuid_ext(CPUID_EXT_MMFR0); As an added bonus, this squashes a potential bug, described by Paul Walmsley in commit 067e710b9a98 ("ARM: 7801/1: prevent gcc 4.5 from reordering extended CP15 reads above is_smp() test"). Signed-off-by: Marc Gonzalez Signed-off-by: Russell King --- arch/arm/kernel/setup.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/arch/arm/kernel/setup.c b/arch/arm/kernel/setup.c index e55408e96559..1d60bebea4b8 100644 --- a/arch/arm/kernel/setup.c +++ b/arch/arm/kernel/setup.c @@ -246,12 +246,9 @@ static int __get_cpu_architecture(void) if (cpu_arch) cpu_arch += CPU_ARCH_ARMv3; } else if ((read_cpuid_id() & 0x000f0000) == 0x000f0000) { - unsigned int mmfr0; - /* Revised CPUID format. Read the Memory Model Feature * Register 0 and check for VMSAv7 or PMSAv7 */ - asm("mrc p15, 0, %0, c0, c1, 4" - : "=r" (mmfr0)); + unsigned int mmfr0 = read_cpuid_ext(CPUID_EXT_MMFR0); if ((mmfr0 & 0x0000000f) >= 0x00000003 || (mmfr0 & 0x000000f0) >= 0x00000030) cpu_arch = CPU_ARCH_ARMv7; -- cgit v1.2.1 From 391949b6f02121371e3d7d9082c6d17fd9853034 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= Date: Wed, 18 Mar 2015 11:27:28 +0100 Subject: spi: trigger trace event for message-done before mesg->complete MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit With spidev the mesg->complete callback points to spidev_complete. Calling this unblocks spidev_sync and so spidev_sync_write finishes. As the struct spi_message just read is a local variable in spidev_sync_write and recording the trace event accesses this message the recording is better done first. The same can happen for spidev_sync_read. This fixes an oops observed on a 3.14-rt system with spidev activity after echo 1 > /sys/kernel/debug/tracing/events/spi/enable . Signed-off-by: Uwe Kleine-König Signed-off-by: Mark Brown Cc: stable@vger.kernel.org --- drivers/spi/spi.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/spi/spi.c b/drivers/spi/spi.c index c64a3e59fce3..57a195041dc7 100644 --- a/drivers/spi/spi.c +++ b/drivers/spi/spi.c @@ -1105,13 +1105,14 @@ void spi_finalize_current_message(struct spi_master *master) "failed to unprepare message: %d\n", ret); } } + + trace_spi_message_done(mesg); + master->cur_msg_prepared = false; mesg->state = NULL; if (mesg->complete) mesg->complete(mesg->context); - - trace_spi_message_done(mesg); } EXPORT_SYMBOL_GPL(spi_finalize_current_message); -- cgit v1.2.1 From 5fcc3c88f9a7944b655856e78341289705cf4cde Mon Sep 17 00:00:00 2001 From: Andrzej Hajda Date: Wed, 18 Feb 2015 12:17:07 +0100 Subject: drm/exynos: remove unused files These files are not used anymore. Signed-off-by: Andrzej Hajda Signed-off-by: Inki Dae --- drivers/gpu/drm/exynos/exynos_drm_connector.c | 245 -------------------------- drivers/gpu/drm/exynos/exynos_drm_connector.h | 20 --- 2 files changed, 265 deletions(-) delete mode 100644 drivers/gpu/drm/exynos/exynos_drm_connector.c delete mode 100644 drivers/gpu/drm/exynos/exynos_drm_connector.h diff --git a/drivers/gpu/drm/exynos/exynos_drm_connector.c b/drivers/gpu/drm/exynos/exynos_drm_connector.c deleted file mode 100644 index ba9b3d5ed672..000000000000 --- a/drivers/gpu/drm/exynos/exynos_drm_connector.c +++ /dev/null @@ -1,245 +0,0 @@ -/* - * Copyright (c) 2011 Samsung Electronics Co., Ltd. - * Authors: - * Inki Dae - * Joonyoung Shim - * Seung-Woo Kim - * - * This program is free software; you can redistribute it and/or modify it - * under the terms of the GNU General Public License as published by the - * Free Software Foundation; either version 2 of the License, or (at your - * option) any later version. - */ - -#include -#include - -#include -#include "exynos_drm_drv.h" -#include "exynos_drm_encoder.h" -#include "exynos_drm_connector.h" - -#define to_exynos_connector(x) container_of(x, struct exynos_drm_connector,\ - drm_connector) - -struct exynos_drm_connector { - struct drm_connector drm_connector; - uint32_t encoder_id; - struct exynos_drm_display *display; -}; - -static int exynos_drm_connector_get_modes(struct drm_connector *connector) -{ - struct exynos_drm_connector *exynos_connector = - to_exynos_connector(connector); - struct exynos_drm_display *display = exynos_connector->display; - struct edid *edid = NULL; - unsigned int count = 0; - int ret; - - /* - * if get_edid() exists then get_edid() callback of hdmi side - * is called to get edid data through i2c interface else - * get timing from the FIMD driver(display controller). - * - * P.S. in case of lcd panel, count is always 1 if success - * because lcd panel has only one mode. - */ - if (display->ops->get_edid) { - edid = display->ops->get_edid(display, connector); - if (IS_ERR_OR_NULL(edid)) { - ret = PTR_ERR(edid); - edid = NULL; - DRM_ERROR("Panel operation get_edid failed %d\n", ret); - goto out; - } - - count = drm_add_edid_modes(connector, edid); - if (!count) { - DRM_ERROR("Add edid modes failed %d\n", count); - goto out; - } - - drm_mode_connector_update_edid_property(connector, edid); - } else { - struct exynos_drm_panel_info *panel; - struct drm_display_mode *mode = drm_mode_create(connector->dev); - if (!mode) { - DRM_ERROR("failed to create a new display mode.\n"); - return 0; - } - - if (display->ops->get_panel) - panel = display->ops->get_panel(display); - else { - drm_mode_destroy(connector->dev, mode); - return 0; - } - - drm_display_mode_from_videomode(&panel->vm, mode); - mode->width_mm = panel->width_mm; - mode->height_mm = panel->height_mm; - connector->display_info.width_mm = mode->width_mm; - connector->display_info.height_mm = mode->height_mm; - - mode->type = DRM_MODE_TYPE_DRIVER | DRM_MODE_TYPE_PREFERRED; - drm_mode_set_name(mode); - drm_mode_probed_add(connector, mode); - - count = 1; - } - -out: - kfree(edid); - return count; -} - -static int exynos_drm_connector_mode_valid(struct drm_connector *connector, - struct drm_display_mode *mode) -{ - struct exynos_drm_connector *exynos_connector = - to_exynos_connector(connector); - struct exynos_drm_display *display = exynos_connector->display; - int ret = MODE_BAD; - - DRM_DEBUG_KMS("%s\n", __FILE__); - - if (display->ops->check_mode) - if (!display->ops->check_mode(display, mode)) - ret = MODE_OK; - - return ret; -} - -static struct drm_encoder *exynos_drm_best_encoder( - struct drm_connector *connector) -{ - struct drm_device *dev = connector->dev; - struct exynos_drm_connector *exynos_connector = - to_exynos_connector(connector); - return drm_encoder_find(dev, exynos_connector->encoder_id); -} - -static struct drm_connector_helper_funcs exynos_connector_helper_funcs = { - .get_modes = exynos_drm_connector_get_modes, - .mode_valid = exynos_drm_connector_mode_valid, - .best_encoder = exynos_drm_best_encoder, -}; - -static int exynos_drm_connector_fill_modes(struct drm_connector *connector, - unsigned int max_width, unsigned int max_height) -{ - struct exynos_drm_connector *exynos_connector = - to_exynos_connector(connector); - struct exynos_drm_display *display = exynos_connector->display; - unsigned int width, height; - - width = max_width; - height = max_height; - - /* - * if specific driver want to find desired_mode using maxmum - * resolution then get max width and height from that driver. - */ - if (display->ops->get_max_resol) - display->ops->get_max_resol(display, &width, &height); - - return drm_helper_probe_single_connector_modes(connector, width, - height); -} - -/* get detection status of display device. */ -static enum drm_connector_status -exynos_drm_connector_detect(struct drm_connector *connector, bool force) -{ - struct exynos_drm_connector *exynos_connector = - to_exynos_connector(connector); - struct exynos_drm_display *display = exynos_connector->display; - enum drm_connector_status status = connector_status_disconnected; - - if (display->ops->is_connected) { - if (display->ops->is_connected(display)) - status = connector_status_connected; - else - status = connector_status_disconnected; - } - - return status; -} - -static void exynos_drm_connector_destroy(struct drm_connector *connector) -{ - struct exynos_drm_connector *exynos_connector = - to_exynos_connector(connector); - - drm_connector_unregister(connector); - drm_connector_cleanup(connector); - kfree(exynos_connector); -} - -static struct drm_connector_funcs exynos_connector_funcs = { - .dpms = drm_helper_connector_dpms, - .fill_modes = exynos_drm_connector_fill_modes, - .detect = exynos_drm_connector_detect, - .destroy = exynos_drm_connector_destroy, -}; - -struct drm_connector *exynos_drm_connector_create(struct drm_device *dev, - struct drm_encoder *encoder) -{ - struct exynos_drm_connector *exynos_connector; - struct exynos_drm_display *display = exynos_drm_get_display(encoder); - struct drm_connector *connector; - int type; - int err; - - exynos_connector = kzalloc(sizeof(*exynos_connector), GFP_KERNEL); - if (!exynos_connector) - return NULL; - - connector = &exynos_connector->drm_connector; - - switch (display->type) { - case EXYNOS_DISPLAY_TYPE_HDMI: - type = DRM_MODE_CONNECTOR_HDMIA; - connector->interlace_allowed = true; - connector->polled = DRM_CONNECTOR_POLL_HPD; - break; - case EXYNOS_DISPLAY_TYPE_VIDI: - type = DRM_MODE_CONNECTOR_VIRTUAL; - connector->polled = DRM_CONNECTOR_POLL_HPD; - break; - default: - type = DRM_MODE_CONNECTOR_Unknown; - break; - } - - drm_connector_init(dev, connector, &exynos_connector_funcs, type); - drm_connector_helper_add(connector, &exynos_connector_helper_funcs); - - err = drm_connector_register(connector); - if (err) - goto err_connector; - - exynos_connector->encoder_id = encoder->base.id; - exynos_connector->display = display; - connector->dpms = DRM_MODE_DPMS_OFF; - connector->encoder = encoder; - - err = drm_mode_connector_attach_encoder(connector, encoder); - if (err) { - DRM_ERROR("failed to attach a connector to a encoder\n"); - goto err_sysfs; - } - - DRM_DEBUG_KMS("connector has been created\n"); - - return connector; - -err_sysfs: - drm_connector_unregister(connector); -err_connector: - drm_connector_cleanup(connector); - kfree(exynos_connector); - return NULL; -} diff --git a/drivers/gpu/drm/exynos/exynos_drm_connector.h b/drivers/gpu/drm/exynos/exynos_drm_connector.h deleted file mode 100644 index 4eb20d78379a..000000000000 --- a/drivers/gpu/drm/exynos/exynos_drm_connector.h +++ /dev/null @@ -1,20 +0,0 @@ -/* - * Copyright (c) 2011 Samsung Electronics Co., Ltd. - * Authors: - * Inki Dae - * Joonyoung Shim - * Seung-Woo Kim - * - * This program is free software; you can redistribute it and/or modify it - * under the terms of the GNU General Public License as published by the - * Free Software Foundation; either version 2 of the License, or (at your - * option) any later version. - */ - -#ifndef _EXYNOS_DRM_CONNECTOR_H_ -#define _EXYNOS_DRM_CONNECTOR_H_ - -struct drm_connector *exynos_drm_connector_create(struct drm_device *dev, - struct drm_encoder *encoder); - -#endif -- cgit v1.2.1 From aed45ab4b07472920ed22ad43b8ffa123c590d57 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Fri, 20 Feb 2015 13:54:43 +0300 Subject: drm/exynos: IS_ERR() vs NULL bug of_iomap() doesn't return error pointers, it returns NULL on error. Signed-off-by: Dan Carpenter Reviewed-by: Gustavo Padovan Signed-off-by: Inki Dae --- drivers/gpu/drm/exynos/exynos7_drm_decon.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/exynos/exynos7_drm_decon.c b/drivers/gpu/drm/exynos/exynos7_drm_decon.c index 63f02e2380ae..970046199608 100644 --- a/drivers/gpu/drm/exynos/exynos7_drm_decon.c +++ b/drivers/gpu/drm/exynos/exynos7_drm_decon.c @@ -888,8 +888,8 @@ static int decon_probe(struct platform_device *pdev) of_node_put(i80_if_timings); ctx->regs = of_iomap(dev->of_node, 0); - if (IS_ERR(ctx->regs)) { - ret = PTR_ERR(ctx->regs); + if (!ctx->regs) { + ret = -ENOMEM; goto err_del_component; } -- cgit v1.2.1 From 995fdfb9c8e9cf9707966c6936eb6ea1a8b68194 Mon Sep 17 00:00:00 2001 From: Charles Keepax Date: Tue, 17 Feb 2015 17:14:41 +0000 Subject: drm/exynos: Check for NULL dereference of crtc The commit "drm/exynos: remove exynos_plane_dpms" (d9ea6256) removed the use of the enabled flag, which means that the code may attempt to call win_enable on a NULL crtc. This results in the following oops on Arndale: [ 1.673479] Unable to handle kernel NULL pointer dereference at virtual address 00000368 [ 1.681500] pgd = c0004000 [ 1.684154] [00000368] *pgd=00000000 [ 1.687713] Internal error: Oops: 5 [#1] PREEMPT SMP ARM [ 1.693012] Modules linked in: [ 1.696045] CPU: 1 PID: 1 Comm: swapper/0 Not tainted 3.19.0-07545-g57485fa #1907 [ 1.703524] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree) (....) [ 2.014803] [] (exynos_plane_destroy) from [] (drm_mode_config_cleanup+0x168/0x20c) [ 2.024178] [] (drm_mode_config_cleanup) from [] (exynos_drm_load+0xac/0x12c) This patch adds in a check to ensure exynos_crtc is not NULL before it is dereferenced. Signed-off-by: Charles Keepax Signed-off-by: Inki Dae --- drivers/gpu/drm/exynos/exynos_drm_plane.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/exynos/exynos_drm_plane.c b/drivers/gpu/drm/exynos/exynos_drm_plane.c index a5616872eee7..8ad5b7294eb4 100644 --- a/drivers/gpu/drm/exynos/exynos_drm_plane.c +++ b/drivers/gpu/drm/exynos/exynos_drm_plane.c @@ -175,7 +175,7 @@ static int exynos_disable_plane(struct drm_plane *plane) struct exynos_drm_plane *exynos_plane = to_exynos_plane(plane); struct exynos_drm_crtc *exynos_crtc = to_exynos_crtc(plane->crtc); - if (exynos_crtc->ops->win_disable) + if (exynos_crtc && exynos_crtc->ops->win_disable) exynos_crtc->ops->win_disable(exynos_crtc, exynos_plane->zpos); -- cgit v1.2.1 From 3da6acfc895601739d64a89891f576f5012c6c51 Mon Sep 17 00:00:00 2001 From: Inki Dae Date: Fri, 6 Mar 2015 22:40:22 +0900 Subject: drm/exynos: fix typo config name correctly. This patch fixes DRM_EXYNOS7DECON to DRM_EXYNOS7_DECON. Signed-off-by: Inki Dae --- drivers/gpu/drm/exynos/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/exynos/Kconfig b/drivers/gpu/drm/exynos/Kconfig index a5e74612100e..0a6780367d28 100644 --- a/drivers/gpu/drm/exynos/Kconfig +++ b/drivers/gpu/drm/exynos/Kconfig @@ -50,7 +50,7 @@ config DRM_EXYNOS_DSI config DRM_EXYNOS_DP bool "EXYNOS DRM DP driver support" - depends on (DRM_EXYNOS_FIMD || DRM_EXYNOS7DECON) && ARCH_EXYNOS && (DRM_PTN3460=n || DRM_PTN3460=y || DRM_PTN3460=DRM_EXYNOS) + depends on (DRM_EXYNOS_FIMD || DRM_EXYNOS7_DECON) && ARCH_EXYNOS && (DRM_PTN3460=n || DRM_PTN3460=y || DRM_PTN3460=DRM_EXYNOS) default DRM_EXYNOS select DRM_PANEL help -- cgit v1.2.1 From cdbfca890714c14cafb6f65cab89b3e3ffad876f Mon Sep 17 00:00:00 2001 From: Hyungwon Hwang Date: Thu, 12 Mar 2015 13:36:02 +0900 Subject: drm/exynos: fix the initialization order in FIMD Since commit 0f04cf8df0b20a97369cb634663fef0578cbf273 ("drm/exynos: fix wrong pipe calculation for crtc"), fimd_clear_channel() can be called when is_drm_iommu_supported() returns true. In this case, the kernel is going to be panicked because crtc is not set yet. [ 1.211156] [drm] Initialized drm 1.1.0 20060810 [ 1.216785] Unable to handle kernel NULL pointer dereference at virtual address 00000350 [ 1.223415] pgd = c0004000 [ 1.226086] [00000350] *pgd=00000000 [ 1.229649] Internal error: Oops: 5 [#1] PREEMPT SMP ARM [ 1.234940] Modules linked in: [ 1.237982] CPU: 2 PID: 1 Comm: swapper/0 Not tainted 4.0.0-rc1-00062-g7a7cc79-dirty #123 [ 1.246136] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree) [ 1.252214] task: ee8c8000 ti: ee8d0000 task.ti: ee8d0000 [ 1.257606] PC is at fimd_wait_for_vblank+0x8/0xc8 [ 1.262370] LR is at fimd_bind+0x138/0x1a8 [ 1.266450] pc : [] lr : [] psr: 20000113 [ 1.266450] sp : ee8d1d28 ip : 00000000 fp : 00000000 [ 1.277906] r10: 00000001 r9 : c09d693c r8 : c0a2d6a8 [ 1.283114] r7 : 00000034 r6 : 00000001 r5 : ee0bb400 r4 : ee244c10 [ 1.289624] r3 : 00000000 r2 : 00000000 r1 : 00000001 r0 : 00000000 [ 1.296135] Flags: nzCv IRQs on FIQs on Mode SVC_32 ISA ARM Segment kernel [ 1.303426] Control: 10c5387d Table: 4000404a DAC: 00000015 [ 1.309154] Process swapper/0 (pid: 1, stack limit = 0xee8d0210) [ 1.315143] Stack: (0xee8d1d28 to 0xee8d2000) [ 1.319486] 1d20: 00000000 c0113d18 ee0bb400 ee0bb400 ee245c30 eebbe210 [ 1.327645] 1d40: ee008a40 ee244c10 ee0bb400 00000001 00000034 c02fb834 00000000 c030a858 [ 1.335804] 1d60: ee244a10 eeb60780 ee008a40 eeb60740 ee0bb400 c03030d0 00000000 00000000 [ 1.343963] 1d80: ee244a10 ee0bb400 00000000 eeb60740 eeb60810 00000000 00000000 c02f6ba4 [ 1.352123] 1da0: ee0bb400 00000000 00000000 c02e0500 ee244a00 c0a04a14 ee0bb400 c02e1de4 [ 1.360282] 1dc0: 00000000 c030a858 00000002 eeb60820 eeb60820 00000002 eeb60780 c03033d4 [ 1.368441] 1de0: c06e9cec 00000000 ee244a10 eeb60780 c0a056f8 c03035fc c0a04b24 c0a04b24 [ 1.376600] 1e00: ee244a10 00000001 c0a049d0 c02f6d34 c0ad462c eeba0790 00000000 ee244a10 [ 1.384759] 1e20: ffffffed c0a049d0 00000000 c03090b0 ee244a10 c0ad462c c0a2d840 c03077a0 [ 1.392919] 1e40: eeb5e880 c024b738 000008db ee244a10 c0a049d0 ee244a44 00000000 c09e71d8 [ 1.401078] 1e60: 000000c6 c0307a6c c0a049d0 00000000 c03079e0 c0305ea8 ee826e5c ee1dc7b4 [ 1.409237] 1e80: c0a049d0 eeb5e880 c0a058a8 c0306e2c c0896204 c0a049d0 c06e9d10 c0a049d0 [ 1.417396] 1ea0: c06e9d10 c0ad4600 00000000 c0308360 00000000 00000003 c06e9d10 c02f6e14 [ 1.425555] 1ec0: 00000000 c0896204 ffffffff 00000000 00000000 00000000 00000000 00000000 [ 1.433714] 1ee0: 00000000 00000000 c02f6d5c c02f6d5c 00000000 eeb5d740 c09e71d8 c0008a30 [ 1.441874] 1f00: ef7fca5e 00000000 00000000 00000066 00000000 ee8d1f28 c003ff1c c02514e8 [ 1.450033] 1f20: 60000113 ffffffff c093906c ef7fca5e 000000c6 c004018c 00000000 c093906c [ 1.458192] 1f40: c08a9690 c093840c 00000006 00000006 c09eb2ac c09c0d74 00000006 c09c0d54 [ 1.466351] 1f60: c0a3d680 c09745a0 c09d693c 000000c6 00000000 c0974db4 00000006 00000006 [ 1.474510] 1f80: c09745a0 ffffffff 00000000 c0692e00 00000000 00000000 00000000 00000000 [ 1.482669] 1fa0: 00000000 c0692e08 00000000 c000f040 00000000 00000000 00000000 00000000 [ 1.490828] 1fc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 [ 1.498988] 1fe0: 00000000 00000000 00000000 00000000 00000013 00000000 ffffffff ffffffff [ 1.507159] [] (fimd_wait_for_vblank) from [] (fimd_bind+0x138/0x1a8) [ 1.515313] [] (fimd_bind) from [] (component_bind_all+0xc4/0x20c) [ 1.523209] [] (component_bind_all) from [] (exynos_drm_load+0xa0/0x140) [ 1.531632] [] (exynos_drm_load) from [] (drm_dev_register+0xa0/0xf4) [ 1.539788] [] (drm_dev_register) from [] (drm_platform_init+0x44/0xcc) [ 1.548121] [] (drm_platform_init) from [] (try_to_bring_up_master.part.1+0xc8/0x104) [ 1.557668] [] (try_to_bring_up_master.part.1) from [] (component_master_add_with_match+0xd0/0x118) [ 1.568431] [] (component_master_add_with_match) from [] (exynos_drm_platform_probe+0xf0/0x118) [ 1.578847] [] (exynos_drm_platform_probe) from [] (platform_drv_probe+0x48/0x98) [ 1.588052] [] (platform_drv_probe) from [] (driver_probe_device+0x140/0x380) [ 1.596902] [] (driver_probe_device) from [] (__driver_attach+0x8c/0x90) [ 1.605321] [] (__driver_attach) from [] (bus_for_each_dev+0x54/0x88) [ 1.613480] [] (bus_for_each_dev) from [] (bus_add_driver+0xec/0x200) [ 1.621640] [] (bus_add_driver) from [] (driver_register+0x78/0xf4) [ 1.629625] [] (driver_register) from [] (exynos_drm_init+0xb8/0x11c) [ 1.637785] [] (exynos_drm_init) from [] (do_one_initcall+0xac/0x1ec) [ 1.645950] [] (do_one_initcall) from [] (kernel_init_freeable+0x194/0x268) [ 1.654626] [] (kernel_init_freeable) from [] (kernel_init+0x8/0xe4) [ 1.662699] [] (kernel_init) from [] (ret_from_fork+0x14/0x34) [ 1.670246] Code: eaffffd5 c09df884 e92d40f0 e24dd01c (e5905350) [ 1.676408] ---[ end trace 804468492f306a6f ]--- [ 1.680948] Kernel panic - not syncing: Attempted to kill init! exitcode=0x0000000b [ 1.680948] [ 1.690035] CPU1: stopping [ 1.692727] CPU: 1 PID: 0 Comm: swapper/1 Tainted: G D 4.0.0-rc1-00062-g7a7cc79-dirty #123 [ 1.702097] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree) [ 1.708192] [] (unwind_backtrace) from [] (show_stack+0x10/0x14) [ 1.715908] [] (show_stack) from [] (dump_stack+0x78/0xc8) [ 1.723108] [] (dump_stack) from [] (handle_IPI+0x16c/0x2b4) [ 1.730485] [] (handle_IPI) from [] (gic_handle_irq+0x64/0x6c) [ 1.738036] [] (gic_handle_irq) from [] (__irq_svc+0x40/0x74) [ 1.745498] Exception stack(0xee8fdf98 to 0xee8fdfe0) [ 1.750533] df80: 00000000 00000000 [ 1.758695] dfa0: ee8fdfe8 c0021780 c09df938 00000015 10c0387d c0a3d988 4000406a c09df8d4 [ 1.766853] dfc0: c0a27a74 c09df940 01000000 ee8fdfe0 c00101c0 c00101c4 60000113 ffffffff [ 1.775015] [] (__irq_svc) from [] (arch_cpu_idle+0x30/0x3c) [ 1.782397] [] (arch_cpu_idle) from [] (cpu_startup_entry+0x180/0x324) [ 1.790639] [] (cpu_startup_entry) from [<40008764>] (0x40008764) [ 1.797579] CPU0: stopping [ 1.800272] CPU: 0 PID: 0 Comm: swapper/0 Tainted: G D 4.0.0-rc1-00062-g7a7cc79-dirty #123 [ 1.809642] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree) [ 1.815730] [] (unwind_backtrace) from [] (show_stack+0x10/0x14) [ 1.823450] [] (show_stack) from [] (dump_stack+0x78/0xc8) [ 1.830653] [] (dump_stack) from [] (handle_IPI+0x16c/0x2b4) [ 1.838030] [] (handle_IPI) from [] (gic_handle_irq+0x64/0x6c) [ 1.845581] [] (gic_handle_irq) from [] (__irq_svc+0x40/0x74) [ 1.853043] Exception stack(0xc09ddf60 to 0xc09ddfa8) [ 1.858081] df60: 00000000 00000000 c09ddfb0 c0021780 c09df938 00000001 ffffffff c0a3d680 [ 1.866239] df80: c09c0dec c09df8d4 c0a27a74 c09df940 01000000 c09ddfa8 c00101c0 c00101c4 [ 1.874396] dfa0: 60000113 ffffffff [ 1.877872] [] (__irq_svc) from [] (arch_cpu_idle+0x30/0x3c) [ 1.885251] [] (arch_cpu_idle) from [] (cpu_startup_entry+0x180/0x324) [ 1.893499] [] (cpu_startup_entry) from [] (start_kernel+0x324/0x37c) [ 1.901655] [] (start_kernel) from [<40008074>] (0x40008074) [ 1.908161] CPU3: stopping [ 1.910855] CPU: 3 PID: 0 Comm: swapper/3 Tainted: G D 4.0.0-rc1-00062-g7a7cc79-dirty #123 [ 1.920225] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree) [ 1.926313] [] (unwind_backtrace) from [] (show_stack+0x10/0x14) [ 1.934034] [] (show_stack) from [] (dump_stack+0x78/0xc8) [ 1.941237] [] (dump_stack) from [] (handle_IPI+0x16c/0x2b4) [ 1.948613] [] (handle_IPI) from [] (gic_handle_irq+0x64/0x6c) [ 1.956165] [] (gic_handle_irq) from [] (__irq_svc+0x40/0x74) [ 1.963626] Exception stack(0xee901f98 to 0xee901fe0) [ 1.968661] 1f80: 00000000 00000000 [ 1.976823] 1fa0: ee901fe8 c0021780 c09df938 00000015 10c0387d c0a3d988 4000406a c09df8d4 [ 1.984982] 1fc0: c0a27a74 c09df940 01000000 ee901fe0 c00101c0 c00101c4 60000113 ffffffff [ 1.993143] [] (__irq_svc) from [] (arch_cpu_idle+0x30/0x3c) [ 2.000522] [] (arch_cpu_idle) from [] (cpu_startup_entry+0x180/0x324) [ 2.008765] [] (cpu_startup_entry) from [<40008764>] (0x40008764) [ 2.015710] ---[ end Kernel panic - not syncing: Attempted to kill init! exitcode=0x0000000b Signed-off-by: Hyungwon Hwang Signed-off-by: Inki Dae --- drivers/gpu/drm/exynos/exynos_drm_fimd.c | 29 +++++++++++------------------ 1 file changed, 11 insertions(+), 18 deletions(-) diff --git a/drivers/gpu/drm/exynos/exynos_drm_fimd.c b/drivers/gpu/drm/exynos/exynos_drm_fimd.c index 925fc69af1a0..c300e22da8ac 100644 --- a/drivers/gpu/drm/exynos/exynos_drm_fimd.c +++ b/drivers/gpu/drm/exynos/exynos_drm_fimd.c @@ -284,14 +284,9 @@ static void fimd_clear_channel(struct fimd_context *ctx) } } -static int fimd_ctx_initialize(struct fimd_context *ctx, +static int fimd_iommu_attach_devices(struct fimd_context *ctx, struct drm_device *drm_dev) { - struct exynos_drm_private *priv; - priv = drm_dev->dev_private; - - ctx->drm_dev = drm_dev; - ctx->pipe = priv->pipe++; /* attach this sub driver to iommu mapping if supported. */ if (is_drm_iommu_supported(ctx->drm_dev)) { @@ -313,7 +308,7 @@ static int fimd_ctx_initialize(struct fimd_context *ctx, return 0; } -static void fimd_ctx_remove(struct fimd_context *ctx) +static void fimd_iommu_detach_devices(struct fimd_context *ctx) { /* detach this sub driver from iommu mapping if supported. */ if (is_drm_iommu_supported(ctx->drm_dev)) @@ -1056,25 +1051,23 @@ static int fimd_bind(struct device *dev, struct device *master, void *data) { struct fimd_context *ctx = dev_get_drvdata(dev); struct drm_device *drm_dev = data; + struct exynos_drm_private *priv = drm_dev->dev_private; int ret; - ret = fimd_ctx_initialize(ctx, drm_dev); - if (ret) { - DRM_ERROR("fimd_ctx_initialize failed.\n"); - return ret; - } + ctx->drm_dev = drm_dev; + ctx->pipe = priv->pipe++; ctx->crtc = exynos_drm_crtc_create(drm_dev, ctx->pipe, EXYNOS_DISPLAY_TYPE_LCD, &fimd_crtc_ops, ctx); - if (IS_ERR(ctx->crtc)) { - fimd_ctx_remove(ctx); - return PTR_ERR(ctx->crtc); - } if (ctx->display) exynos_drm_create_enc_conn(drm_dev, ctx->display); + ret = fimd_iommu_attach_devices(ctx, drm_dev); + if (ret) + return ret; + return 0; } @@ -1086,10 +1079,10 @@ static void fimd_unbind(struct device *dev, struct device *master, fimd_dpms(ctx->crtc, DRM_MODE_DPMS_OFF); + fimd_iommu_detach_devices(ctx); + if (ctx->display) exynos_dpi_remove(ctx->display); - - fimd_ctx_remove(ctx); } static const struct component_ops fimd_component_ops = { -- cgit v1.2.1 From cf39284d41f67964cf42b21bb386c012cf5b7f65 Mon Sep 17 00:00:00 2001 From: Axel Lin Date: Wed, 18 Mar 2015 08:57:41 +0800 Subject: regulator: Fix documentation for regmap in the config dev_get_regulator() does not exist, fix the typo. Signed-off-by: Axel Lin Signed-off-by: Mark Brown --- include/linux/regulator/driver.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/linux/regulator/driver.h b/include/linux/regulator/driver.h index d4ad5b5a02bb..045f709cb89b 100644 --- a/include/linux/regulator/driver.h +++ b/include/linux/regulator/driver.h @@ -316,7 +316,7 @@ struct regulator_desc { * @driver_data: private regulator data * @of_node: OpenFirmware node to parse for device tree bindings (may be * NULL). - * @regmap: regmap to use for core regmap helpers if dev_get_regulator() is + * @regmap: regmap to use for core regmap helpers if dev_get_regmap() is * insufficient. * @ena_gpio_initialized: GPIO controlling regulator enable was properly * initialized, meaning that >= 0 is a valid gpio -- cgit v1.2.1 From e32643a7443e2a601f86cb1be4a0c080949e007f Mon Sep 17 00:00:00 2001 From: Boris Brezillon Date: Tue, 17 Mar 2015 17:15:46 +0100 Subject: USB: ehci-atmel: rework clk handling The EHCI IP only needs the UTMI/UPLL (uclk) and the peripheral (iclk) clocks to work properly. Remove the useless system clock (fclk). Avoid calling set_rate on the fixed rate UTMI/IPLL clock and remove useless IS_ENABLED(CONFIG_COMMON_CLK) tests (all at91 platforms have been moved to the CCF). This patch also fixes a bug introduced by 3440ef1 (ARM: at91/dt: fix USB high-speed clock to select UTMI), which was leaving the usb clock uninitialized and preventing the OHCI driver from setting the usb clock rate to 48MHz. This bug was caused by several things: 1/ usb clock drivers set the CLK_SET_RATE_GATE flag, which means the rate cannot be changed once the clock is prepared 2/ The EHCI driver was retrieving and preparing/enabling the uhpck clock which was in turn preparing its parent clock (the usb clock), thus preventing any rate change because of 1/ Fixes: 3440ef169100 ("ARM: at91/dt: fix USB high-speed clock to select UTMI") Signed-off-by: Boris Brezillon Acked-by: Alan Stern Acked-by: Nicolas Ferre Signed-off-by: Greg Kroah-Hartman --- drivers/usb/host/ehci-atmel.c | 30 +++++++++--------------------- 1 file changed, 9 insertions(+), 21 deletions(-) diff --git a/drivers/usb/host/ehci-atmel.c b/drivers/usb/host/ehci-atmel.c index 663f7908b15c..be0964a801e8 100644 --- a/drivers/usb/host/ehci-atmel.c +++ b/drivers/usb/host/ehci-atmel.c @@ -34,7 +34,6 @@ static const char hcd_name[] = "ehci-atmel"; struct atmel_ehci_priv { struct clk *iclk; - struct clk *fclk; struct clk *uclk; bool clocked; }; @@ -51,12 +50,9 @@ static void atmel_start_clock(struct atmel_ehci_priv *atmel_ehci) { if (atmel_ehci->clocked) return; - if (IS_ENABLED(CONFIG_COMMON_CLK)) { - clk_set_rate(atmel_ehci->uclk, 48000000); - clk_prepare_enable(atmel_ehci->uclk); - } + + clk_prepare_enable(atmel_ehci->uclk); clk_prepare_enable(atmel_ehci->iclk); - clk_prepare_enable(atmel_ehci->fclk); atmel_ehci->clocked = true; } @@ -64,10 +60,9 @@ static void atmel_stop_clock(struct atmel_ehci_priv *atmel_ehci) { if (!atmel_ehci->clocked) return; - clk_disable_unprepare(atmel_ehci->fclk); + clk_disable_unprepare(atmel_ehci->iclk); - if (IS_ENABLED(CONFIG_COMMON_CLK)) - clk_disable_unprepare(atmel_ehci->uclk); + clk_disable_unprepare(atmel_ehci->uclk); atmel_ehci->clocked = false; } @@ -146,20 +141,13 @@ static int ehci_atmel_drv_probe(struct platform_device *pdev) retval = -ENOENT; goto fail_request_resource; } - atmel_ehci->fclk = devm_clk_get(&pdev->dev, "uhpck"); - if (IS_ERR(atmel_ehci->fclk)) { - dev_err(&pdev->dev, "Error getting function clock\n"); - retval = -ENOENT; + + atmel_ehci->uclk = devm_clk_get(&pdev->dev, "usb_clk"); + if (IS_ERR(atmel_ehci->uclk)) { + dev_err(&pdev->dev, "failed to get uclk\n"); + retval = PTR_ERR(atmel_ehci->uclk); goto fail_request_resource; } - if (IS_ENABLED(CONFIG_COMMON_CLK)) { - atmel_ehci->uclk = devm_clk_get(&pdev->dev, "usb_clk"); - if (IS_ERR(atmel_ehci->uclk)) { - dev_err(&pdev->dev, "failed to get uclk\n"); - retval = PTR_ERR(atmel_ehci->uclk); - goto fail_request_resource; - } - } ehci = hcd_to_ehci(hcd); /* registers start at offset 0x0 */ -- cgit v1.2.1 From a239118a24b3bf9089751068e431dfb63dc4168b Mon Sep 17 00:00:00 2001 From: Alex Deucher Date: Tue, 17 Mar 2015 11:53:33 -0400 Subject: drm/radeon: drop ttm two ended allocation MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit radeon_bo_create() calls radeon_ttm_placement_from_domain() before ttm_bo_init() is called. radeon_ttm_placement_from_domain() uses the ttm bo size to determine when to select top down allocation but since the ttm bo is not initialized yet the check is always false. It only took effect when buffers were validated later. It also seemed to regress suspend and resume on some systems possibly due to it not taking effect in radeon_bo_create(). radeon_bo_create() and radeon_ttm_placement_from_domain() need to be reworked substantially for this to be optimally effective. Re-enable it at that point. Noticed-by: Oded Gabbay Reviewed-by: Christian König Signed-off-by: Alex Deucher Cc: stable@vger.kernel.org --- drivers/gpu/drm/radeon/radeon_object.c | 11 ----------- 1 file changed, 11 deletions(-) diff --git a/drivers/gpu/drm/radeon/radeon_object.c b/drivers/gpu/drm/radeon/radeon_object.c index 43e09942823e..318165d4855c 100644 --- a/drivers/gpu/drm/radeon/radeon_object.c +++ b/drivers/gpu/drm/radeon/radeon_object.c @@ -173,17 +173,6 @@ void radeon_ttm_placement_from_domain(struct radeon_bo *rbo, u32 domain) else rbo->placements[i].lpfn = 0; } - - /* - * Use two-ended allocation depending on the buffer size to - * improve fragmentation quality. - * 512kb was measured as the most optimal number. - */ - if (rbo->tbo.mem.size > 512 * 1024) { - for (i = 0; i < c; i++) { - rbo->placements[i].flags |= TTM_PL_FLAG_TOPDOWN; - } - } } int radeon_bo_create(struct radeon_device *rdev, -- cgit v1.2.1 From bda13e35d584dabf52c9f77e0fe62683ac4d9f86 Mon Sep 17 00:00:00 2001 From: Hans de Goede Date: Mon, 16 Mar 2015 15:18:13 +0100 Subject: uas: Add US_FL_NO_ATA_1X for Initio Corporation controllers / devices A new uas compatible controller has shown up in some people's devices from the manufacturer Initio Corporation, this controller needs the US_FL_NO_ATA_1X quirk to work properly with uas, so add it to the uas quirks table. Reported-and-tested-by: Benjamin Tissoires Cc: Benjamin Tissoires Cc: stable@vger.kernel.org # 3.16 Signed-off-by: Hans de Goede Signed-off-by: Greg Kroah-Hartman --- drivers/usb/storage/unusual_uas.h | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/usb/storage/unusual_uas.h b/drivers/usb/storage/unusual_uas.h index 82570425fdfe..c85ea530085f 100644 --- a/drivers/usb/storage/unusual_uas.h +++ b/drivers/usb/storage/unusual_uas.h @@ -113,6 +113,13 @@ UNUSUAL_DEV(0x0bc2, 0xab2a, 0x0000, 0x9999, USB_SC_DEVICE, USB_PR_DEVICE, NULL, US_FL_NO_ATA_1X), +/* Reported-by: Benjamin Tissoires */ +UNUSUAL_DEV(0x13fd, 0x3940, 0x0000, 0x9999, + "Initio Corporation", + "", + USB_SC_DEVICE, USB_PR_DEVICE, NULL, + US_FL_NO_ATA_1X), + /* Reported-by: Tom Arild Naess */ UNUSUAL_DEV(0x152d, 0x0539, 0x0000, 0x9999, "JMicron", -- cgit v1.2.1 From a886bd92267c9e3d5c912860c6fb5a68479a7643 Mon Sep 17 00:00:00 2001 From: Peter Chen Date: Thu, 12 Mar 2015 09:47:53 +0800 Subject: usb: common: otg-fsm: only signal connect after switching to peripheral We should signal connect (pull up dp) after we have already at peripheral mode, otherwise, the dp may be toggled due to we reset controller or do disconnect during the initialization for peripheral, then, the host may be confused during the enumeration, eg, it finds the reset can't succeed, but the device is still there, see below error message. hub 1-0:1.0: USB hub found hub 1-0:1.0: 1 port detected hub 1-0:1.0: cannot reset port 1 (err = -32) hub 1-0:1.0: cannot reset port 1 (err = -32) hub 1-0:1.0: cannot reset port 1 (err = -32) hub 1-0:1.0: cannot reset port 1 (err = -32) hub 1-0:1.0: cannot reset port 1 (err = -32) hub 1-0:1.0: Cannot enable port 1. Maybe the USB cable is bad? hub 1-0:1.0: cannot reset port 1 (err = -32) hub 1-0:1.0: cannot reset port 1 (err = -32) hub 1-0:1.0: cannot reset port 1 (err = -32) hub 1-0:1.0: cannot reset port 1 (err = -32) hub 1-0:1.0: cannot reset port 1 (err = -32) hub 1-0:1.0: Cannot enable port 1. Maybe the USB cable is bad? hub 1-0:1.0: cannot reset port 1 (err = -32) hub 1-0:1.0: cannot reset port 1 (err = -32) hub 1-0:1.0: cannot reset port 1 (err = -32) hub 1-0:1.0: cannot reset port 1 (err = -32) hub 1-0:1.0: cannot reset port 1 (err = -32) hub 1-0:1.0: Cannot enable port 1. Maybe the USB cable is bad? hub 1-0:1.0: cannot reset port 1 (err = -32) hub 1-0:1.0: cannot reset port 1 (err = -32) hub 1-0:1.0: cannot reset port 1 (err = -32) hub 1-0:1.0: cannot reset port 1 (err = -32) hub 1-0:1.0: cannot reset port 1 (err = -32) hub 1-0:1.0: Cannot enable port 1. Maybe the USB cable is bad? hub 1-0:1.0: unable to enumerate USB device on port 1 Fixes: the issue existed when the otg fsm code was added. Cc: # v3.16+ Signed-off-by: Peter Chen Signed-off-by: Greg Kroah-Hartman --- drivers/usb/common/usb-otg-fsm.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/usb/common/usb-otg-fsm.c b/drivers/usb/common/usb-otg-fsm.c index c6b35b77dab7..61d538aa2346 100644 --- a/drivers/usb/common/usb-otg-fsm.c +++ b/drivers/usb/common/usb-otg-fsm.c @@ -150,9 +150,9 @@ static int otg_set_state(struct otg_fsm *fsm, enum usb_otg_state new_state) break; case OTG_STATE_B_PERIPHERAL: otg_chrg_vbus(fsm, 0); - otg_loc_conn(fsm, 1); otg_loc_sof(fsm, 0); otg_set_protocol(fsm, PROTO_GADGET); + otg_loc_conn(fsm, 1); break; case OTG_STATE_B_WAIT_ACON: otg_chrg_vbus(fsm, 0); @@ -213,10 +213,10 @@ static int otg_set_state(struct otg_fsm *fsm, enum usb_otg_state new_state) break; case OTG_STATE_A_PERIPHERAL: - otg_loc_conn(fsm, 1); otg_loc_sof(fsm, 0); otg_set_protocol(fsm, PROTO_GADGET); otg_drv_vbus(fsm, 1); + otg_loc_conn(fsm, 1); otg_add_timer(fsm, A_BIDL_ADIS); break; case OTG_STATE_A_WAIT_VFALL: -- cgit v1.2.1 From ea524c7e3d3c031cf095c04bc93af42fa3d308fd Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Tue, 17 Mar 2015 23:25:36 +0000 Subject: dmaengine: pl08x: Define capabilities for generic capabilities reporting Ensure that clients can automatically configure themselves and avoid a nasty warning at boot by providing capability information. Signed-off-by: Mark Brown Signed-off-by: Vinod Koul --- drivers/dma/amba-pl08x.c | 14 ++++++++++++++ 1 file changed, 14 insertions(+) diff --git a/drivers/dma/amba-pl08x.c b/drivers/dma/amba-pl08x.c index 4a5fd245014e..83aa55d6fa5d 100644 --- a/drivers/dma/amba-pl08x.c +++ b/drivers/dma/amba-pl08x.c @@ -97,6 +97,12 @@ #define DRIVER_NAME "pl08xdmac" +#define PL80X_DMA_BUSWIDTHS \ + BIT(DMA_SLAVE_BUSWIDTH_UNDEFINED) | \ + BIT(DMA_SLAVE_BUSWIDTH_1_BYTE) | \ + BIT(DMA_SLAVE_BUSWIDTH_2_BYTES) | \ + BIT(DMA_SLAVE_BUSWIDTH_4_BYTES) + static struct amba_driver pl08x_amba_driver; struct pl08x_driver_data; @@ -2070,6 +2076,10 @@ static int pl08x_probe(struct amba_device *adev, const struct amba_id *id) pl08x->memcpy.device_pause = pl08x_pause; pl08x->memcpy.device_resume = pl08x_resume; pl08x->memcpy.device_terminate_all = pl08x_terminate_all; + pl08x->memcpy.src_addr_widths = PL80X_DMA_BUSWIDTHS; + pl08x->memcpy.dst_addr_widths = PL80X_DMA_BUSWIDTHS; + pl08x->memcpy.directions = BIT(DMA_MEM_TO_MEM); + pl08x->memcpy.residue_granularity = DMA_RESIDUE_GRANULARITY_SEGMENT; /* Initialize slave engine */ dma_cap_set(DMA_SLAVE, pl08x->slave.cap_mask); @@ -2086,6 +2096,10 @@ static int pl08x_probe(struct amba_device *adev, const struct amba_id *id) pl08x->slave.device_pause = pl08x_pause; pl08x->slave.device_resume = pl08x_resume; pl08x->slave.device_terminate_all = pl08x_terminate_all; + pl08x->slave.src_addr_widths = PL80X_DMA_BUSWIDTHS; + pl08x->slave.dst_addr_widths = PL80X_DMA_BUSWIDTHS; + pl08x->slave.directions = BIT(DMA_DEV_TO_MEM) | BIT(DMA_MEM_TO_DEV); + pl08x->slave.residue_granularity = DMA_RESIDUE_GRANULARITY_SEGMENT; /* Get the platform data */ pl08x->pd = dev_get_platdata(&adev->dev); -- cgit v1.2.1 From 217e8b16a43a1780e77607dc019c5f3b26fab48a Mon Sep 17 00:00:00 2001 From: Moni Shoua Date: Wed, 18 Mar 2015 16:51:35 +0200 Subject: IB/mlx4: Verify net device validity on port change event Processing an event is done in a different context from the one when the event was dispatched. This requires a check that the slave net device is still valid when the event is being processed. The check is done under the iboe lock which ensure correctness. Fixes: a57500903093 ('IB/mlx4: Add port aggregation support') Signed-off-by: Moni Shoua Signed-off-by: Or Gerlitz Signed-off-by: David S. Miller --- drivers/infiniband/hw/mlx4/main.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c index ac6e2b710ea6..b972c0b41799 100644 --- a/drivers/infiniband/hw/mlx4/main.c +++ b/drivers/infiniband/hw/mlx4/main.c @@ -2697,8 +2697,12 @@ static void handle_bonded_port_state_event(struct work_struct *work) spin_lock_bh(&ibdev->iboe.lock); for (i = 0; i < MLX4_MAX_PORTS; ++i) { struct net_device *curr_netdev = ibdev->iboe.netdevs[i]; + enum ib_port_state curr_port_state; - enum ib_port_state curr_port_state = + if (!curr_netdev) + continue; + + curr_port_state = (netif_running(curr_netdev) && netif_carrier_ok(curr_netdev)) ? IB_PORT_ACTIVE : IB_PORT_DOWN; -- cgit v1.2.1 From a16f3565703cfc3094938fb3c979cbb90f6d9eb4 Mon Sep 17 00:00:00 2001 From: Eran Ben Elisha Date: Wed, 18 Mar 2015 16:51:36 +0200 Subject: net/mlx4_en: Fix off-by-one in ethtool statistics display NUM_PORT_STATS was 9 instead of 10, which caused off-by-one bug when displaying the statistics starting from tx_chksum_offload in ethtool. Fixes: f8c6455bb04b ('net/mlx4_en: Extend checksum offloading by CHECKSUM COMPLETE') Signed-off-by: Eran Ben Elisha Signed-off-by: Hadar Hen Zion Signed-off-by: Or Gerlitz Signed-off-by: David S. Miller --- drivers/net/ethernet/mellanox/mlx4/mlx4_en.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h index 2a8268e6be15..ebbe244e80dd 100644 --- a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h +++ b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h @@ -453,7 +453,7 @@ struct mlx4_en_port_stats { unsigned long rx_chksum_none; unsigned long rx_chksum_complete; unsigned long tx_chksum_offload; -#define NUM_PORT_STATS 9 +#define NUM_PORT_STATS 10 }; struct mlx4_en_perf_stats { -- cgit v1.2.1 From 61a3855bb726cbb062ef02a31a832dea455456e0 Mon Sep 17 00:00:00 2001 From: Majd Dibbiny Date: Wed, 18 Mar 2015 16:51:37 +0200 Subject: IB/mlx4: Saturate RoCE port PMA counters in case of overflow For RoCE ports, we set the u32 PMA values based on u64 HCA counters. In case of overflow, according to the IB spec, we have to saturate a counter to its max value, do that. Fixes: c37791349cc7 ('IB/mlx4: Support PMA counters for IBoE') Signed-off-by: Majd Dibbiny Signed-off-by: Eran Ben Elisha Signed-off-by: Hadar Hen Zion Signed-off-by: Or Gerlitz Signed-off-by: David S. Miller --- drivers/infiniband/hw/mlx4/mad.c | 20 ++++++++++++++++---- 1 file changed, 16 insertions(+), 4 deletions(-) diff --git a/drivers/infiniband/hw/mlx4/mad.c b/drivers/infiniband/hw/mlx4/mad.c index c7619716c31d..59040265e361 100644 --- a/drivers/infiniband/hw/mlx4/mad.c +++ b/drivers/infiniband/hw/mlx4/mad.c @@ -64,6 +64,14 @@ enum { #define GUID_TBL_BLK_NUM_ENTRIES 8 #define GUID_TBL_BLK_SIZE (GUID_TBL_ENTRY_SIZE * GUID_TBL_BLK_NUM_ENTRIES) +/* Counters should be saturate once they reach their maximum value */ +#define ASSIGN_32BIT_COUNTER(counter, value) do {\ + if ((value) > U32_MAX) \ + counter = cpu_to_be32(U32_MAX); \ + else \ + counter = cpu_to_be32(value); \ +} while (0) + struct mlx4_mad_rcv_buf { struct ib_grh grh; u8 payload[256]; @@ -806,10 +814,14 @@ static int ib_process_mad(struct ib_device *ibdev, int mad_flags, u8 port_num, static void edit_counter(struct mlx4_counter *cnt, struct ib_pma_portcounters *pma_cnt) { - pma_cnt->port_xmit_data = cpu_to_be32((be64_to_cpu(cnt->tx_bytes)>>2)); - pma_cnt->port_rcv_data = cpu_to_be32((be64_to_cpu(cnt->rx_bytes)>>2)); - pma_cnt->port_xmit_packets = cpu_to_be32(be64_to_cpu(cnt->tx_frames)); - pma_cnt->port_rcv_packets = cpu_to_be32(be64_to_cpu(cnt->rx_frames)); + ASSIGN_32BIT_COUNTER(pma_cnt->port_xmit_data, + (be64_to_cpu(cnt->tx_bytes) >> 2)); + ASSIGN_32BIT_COUNTER(pma_cnt->port_rcv_data, + (be64_to_cpu(cnt->rx_bytes) >> 2)); + ASSIGN_32BIT_COUNTER(pma_cnt->port_xmit_packets, + be64_to_cpu(cnt->tx_frames)); + ASSIGN_32BIT_COUNTER(pma_cnt->port_rcv_packets, + be64_to_cpu(cnt->rx_frames)); } static int iboe_process_mad(struct ib_device *ibdev, int mad_flags, u8 port_num, -- cgit v1.2.1 From 39de961a4a3317741a7ac0cb9607593f9ffec779 Mon Sep 17 00:00:00 2001 From: Eran Ben Elisha Date: Wed, 18 Mar 2015 16:51:38 +0200 Subject: net/mlx4_en: Set statistics bitmap at port init Port statistics bitmap will now be initialized at port init. Even before starting the port, statistics are visible to the user and must be properly masked. Signed-off-by: Eran Ben Elisha Signed-off-by: Hadar Hen Zion Signed-off-by: Or Gerlitz Signed-off-by: David S. Miller --- drivers/net/ethernet/mellanox/mlx4/en_netdev.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c index 2a210c4efb89..ebce5bb24df9 100644 --- a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c +++ b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c @@ -1698,8 +1698,6 @@ int mlx4_en_start_port(struct net_device *dev) /* Schedule multicast task to populate multicast list */ queue_work(mdev->workqueue, &priv->rx_mode_task); - mlx4_set_stats_bitmap(mdev->dev, &priv->stats_bitmap); - #ifdef CONFIG_MLX4_EN_VXLAN if (priv->mdev->dev->caps.tunnel_offload_mode == MLX4_TUNNEL_OFFLOAD_MODE_VXLAN) vxlan_get_rx_port(dev); @@ -2853,6 +2851,8 @@ int mlx4_en_init_netdev(struct mlx4_en_dev *mdev, int port, queue_delayed_work(mdev->workqueue, &priv->service_task, SERVICE_TASK_DELAY); + mlx4_set_stats_bitmap(mdev->dev, &priv->stats_bitmap); + return 0; out: -- cgit v1.2.1 From 077155332265f1d64c57bd6c49748a8b7e72a3f9 Mon Sep 17 00:00:00 2001 From: Kishon Vijay Abraham I Date: Fri, 20 Feb 2015 14:21:14 +0530 Subject: ARM: dts: dra7: remove ti,hwmod property from pcie phy Now that we don't have hwmod entry for pcie PHY remove the ti,hwmod property from PCIE PHY's. Otherwise we will get: platform 4a094000.pciephy: Cannot lookup hwmod 'pcie1-phy' Signed-off-by: Kishon Vijay Abraham I [tony@atomide.com: updated comments] Signed-off-by: Tony Lindgren --- arch/arm/boot/dts/dra7.dtsi | 2 -- 1 file changed, 2 deletions(-) diff --git a/arch/arm/boot/dts/dra7.dtsi b/arch/arm/boot/dts/dra7.dtsi index 127608d79033..c4659a979c41 100644 --- a/arch/arm/boot/dts/dra7.dtsi +++ b/arch/arm/boot/dts/dra7.dtsi @@ -1111,7 +1111,6 @@ "wkupclk", "refclk", "div-clk", "phy-div"; #phy-cells = <0>; - ti,hwmods = "pcie1-phy"; }; pcie2_phy: pciephy@4a095000 { @@ -1130,7 +1129,6 @@ "wkupclk", "refclk", "div-clk", "phy-div"; #phy-cells = <0>; - ti,hwmods = "pcie2-phy"; status = "disabled"; }; }; -- cgit v1.2.1 From 599c376c49323127c9bdbb0fa61a3d4743819bc2 Mon Sep 17 00:00:00 2001 From: Tony Lindgren Date: Wed, 18 Mar 2015 13:41:34 -0700 Subject: ARM: dts: Fix gpio interrupts for dm816x Commit 7800064ba507 ("ARM: dts: Add basic dm816x device tree configuration") added basic devices for dm816x, but I was not able to test the GPIO interrupts earlier until I found some suitable pins to test with. We can mux the MMC card detect and write protect pins from SD_SDCD and SD_SDWP mode to use a normal GPIO interrupts that are also suitable for the MMC subsystem. This turned out several issues that need to be fixed: - I set the GPIO type wrong to be compatible with omap3 instead of omap4. The GPIO controller on dm816x has EOI interrupt register like omap4 and am335x. - I got the GPIO interrupt numbers wrong as each bank has two and we only use one. They need to be set up the same way as on am335x. - The gpio banks are missing interrupt controller related properties. With these changes the GPIO interrupts can be used with the MMC card detect pin, so let's wire that up. Let's also mux all the MMC lines for completeness while at it. For the first GPIO bank I tested using GPMC lines temporarily muxed to GPIOs on the dip switch 10. Cc: Brian Hutchinson Cc: Matthijs van Duin Signed-off-by: Tony Lindgren --- arch/arm/boot/dts/dm8168-evm.dts | 19 +++++++++++++++++++ arch/arm/boot/dts/dm816x.dtsi | 18 ++++++++++++++---- 2 files changed, 33 insertions(+), 4 deletions(-) diff --git a/arch/arm/boot/dts/dm8168-evm.dts b/arch/arm/boot/dts/dm8168-evm.dts index d3a29c1b8417..afe678f6d2e9 100644 --- a/arch/arm/boot/dts/dm8168-evm.dts +++ b/arch/arm/boot/dts/dm8168-evm.dts @@ -36,6 +36,20 @@ >; }; + mmc_pins: pinmux_mmc_pins { + pinctrl-single,pins = < + DM816X_IOPAD(0x0a70, MUX_MODE0) /* SD_POW */ + DM816X_IOPAD(0x0a74, MUX_MODE0) /* SD_CLK */ + DM816X_IOPAD(0x0a78, MUX_MODE0) /* SD_CMD */ + DM816X_IOPAD(0x0a7C, MUX_MODE0) /* SD_DAT0 */ + DM816X_IOPAD(0x0a80, MUX_MODE0) /* SD_DAT1 */ + DM816X_IOPAD(0x0a84, MUX_MODE0) /* SD_DAT2 */ + DM816X_IOPAD(0x0a88, MUX_MODE0) /* SD_DAT2 */ + DM816X_IOPAD(0x0a8c, MUX_MODE2) /* GP1[7] */ + DM816X_IOPAD(0x0a90, MUX_MODE2) /* GP1[8] */ + >; + }; + usb0_pins: pinmux_usb0_pins { pinctrl-single,pins = < DM816X_IOPAD(0x0d00, MUX_MODE0) /* USB0_DRVVBUS */ @@ -137,7 +151,12 @@ }; &mmc1 { + pinctrl-names = "default"; + pinctrl-0 = <&mmc_pins>; vmmc-supply = <&vmmcsd_fixed>; + bus-width = <4>; + cd-gpios = <&gpio2 7 GPIO_ACTIVE_LOW>; + wp-gpios = <&gpio2 8 GPIO_ACTIVE_LOW>; }; /* At least dm8168-evm rev c won't support multipoint, later may */ diff --git a/arch/arm/boot/dts/dm816x.dtsi b/arch/arm/boot/dts/dm816x.dtsi index 3c97b5f2addc..f35715bc6992 100644 --- a/arch/arm/boot/dts/dm816x.dtsi +++ b/arch/arm/boot/dts/dm816x.dtsi @@ -150,17 +150,27 @@ }; gpio1: gpio@48032000 { - compatible = "ti,omap3-gpio"; + compatible = "ti,omap4-gpio"; ti,hwmods = "gpio1"; + ti,gpio-always-on; reg = <0x48032000 0x1000>; - interrupts = <97>; + interrupts = <96>; + gpio-controller; + #gpio-cells = <2>; + interrupt-controller; + #interrupt-cells = <2>; }; gpio2: gpio@4804c000 { - compatible = "ti,omap3-gpio"; + compatible = "ti,omap4-gpio"; ti,hwmods = "gpio2"; + ti,gpio-always-on; reg = <0x4804c000 0x1000>; - interrupts = <99>; + interrupts = <98>; + gpio-controller; + #gpio-cells = <2>; + interrupt-controller; + #interrupt-cells = <2>; }; gpmc: gpmc@50000000 { -- cgit v1.2.1 From bc188d818edf325ae38cfa43254a0b10a4defd65 Mon Sep 17 00:00:00 2001 From: Sam Bradshaw Date: Wed, 18 Mar 2015 17:06:18 -0600 Subject: blkmq: Fix NULL pointer deref when all reserved tags in When allocating from the reserved tags pool, bt_get() is called with a NULL hctx. If all tags are in use, the hw queue is kicked to push out any pending IO, potentially freeing tags, and tag allocation is retried. The problem is that blk_mq_run_hw_queue() doesn't check for a NULL hctx. So we avoid it with a simple NULL hctx test. Tested by hammering mtip32xx with concurrent smartctl/hdparm. Signed-off-by: Sam Bradshaw Signed-off-by: Selvan Mani Fixes: b32232073e80 ("blk-mq: fix hang in bt_get()") Cc: stable@kernel.org Added appropriate comment. Signed-off-by: Jens Axboe --- block/blk-mq-tag.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/block/blk-mq-tag.c b/block/blk-mq-tag.c index d53a764b05ea..be3290cc0644 100644 --- a/block/blk-mq-tag.c +++ b/block/blk-mq-tag.c @@ -278,9 +278,11 @@ static int bt_get(struct blk_mq_alloc_data *data, /* * We're out of tags on this hardware queue, kick any * pending IO submits before going to sleep waiting for - * some to complete. + * some to complete. Note that hctx can be NULL here for + * reserved tag allocation. */ - blk_mq_run_hw_queue(hctx, false); + if (hctx) + blk_mq_run_hw_queue(hctx, false); /* * Retry tag allocation after running the hardware queue, -- cgit v1.2.1 From 3c08158e0ef5d6a2d4ae21d9eda218c468bc774f Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Wed, 18 Mar 2015 19:15:28 -0700 Subject: sparc: Fix /proc/kcore /proc/kcore investigates the "System RAM" elements in /proc/iomem to initialize it's memory tables. Therefore we have to register them before it tries to do so. kcore uses device_initcall() so let's use arch_initcall() for the registry. Also we need ARCH_PROC_KCORE_TEXT to get the virtual addresses of the kernel image correct. Reported-by: David Ahern Signed-off-by: David S. Miller --- arch/sparc/Kconfig | 3 +++ arch/sparc/mm/init_64.c | 2 +- 2 files changed, 4 insertions(+), 1 deletion(-) diff --git a/arch/sparc/Kconfig b/arch/sparc/Kconfig index 96ac69c5eba0..efb00ec75805 100644 --- a/arch/sparc/Kconfig +++ b/arch/sparc/Kconfig @@ -86,6 +86,9 @@ config ARCH_DEFCONFIG default "arch/sparc/configs/sparc32_defconfig" if SPARC32 default "arch/sparc/configs/sparc64_defconfig" if SPARC64 +config ARCH_PROC_KCORE_TEXT + def_bool y + config IOMMU_HELPER bool default y if SPARC64 diff --git a/arch/sparc/mm/init_64.c b/arch/sparc/mm/init_64.c index 3ea267c53320..4ca0d6ba5ec8 100644 --- a/arch/sparc/mm/init_64.c +++ b/arch/sparc/mm/init_64.c @@ -2820,7 +2820,7 @@ static int __init report_memory(void) return 0; } -device_initcall(report_memory); +arch_initcall(report_memory); #ifdef CONFIG_SMP #define do_flush_tlb_kernel_range smp_flush_tlb_kernel_range -- cgit v1.2.1 From 8d006e0105978619fb472e150c88b0d49337fe2b Mon Sep 17 00:00:00 2001 From: Ondrej Zary Date: Wed, 18 Mar 2015 23:01:01 +0100 Subject: Revert "net: cx82310_eth: use common match macro" This reverts commit 11ad714b98f6d9ca0067568442afe3e70eb94845 because it breaks cx82310_eth. The custom USB_DEVICE_CLASS macro matches bDeviceClass, bDeviceSubClass and bDeviceProtocol but the common USB_DEVICE_AND_INTERFACE_INFO matches bInterfaceClass, bInterfaceSubClass and bInterfaceProtocol instead, which are not specified. Signed-off-by: Ondrej Zary Signed-off-by: David S. Miller --- drivers/net/usb/cx82310_eth.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) diff --git a/drivers/net/usb/cx82310_eth.c b/drivers/net/usb/cx82310_eth.c index 3eed708a6182..fe48f4c51373 100644 --- a/drivers/net/usb/cx82310_eth.c +++ b/drivers/net/usb/cx82310_eth.c @@ -300,9 +300,18 @@ static const struct driver_info cx82310_info = { .tx_fixup = cx82310_tx_fixup, }; +#define USB_DEVICE_CLASS(vend, prod, cl, sc, pr) \ + .match_flags = USB_DEVICE_ID_MATCH_DEVICE | \ + USB_DEVICE_ID_MATCH_DEV_INFO, \ + .idVendor = (vend), \ + .idProduct = (prod), \ + .bDeviceClass = (cl), \ + .bDeviceSubClass = (sc), \ + .bDeviceProtocol = (pr) + static const struct usb_device_id products[] = { { - USB_DEVICE_AND_INTERFACE_INFO(0x0572, 0xcb01, 0xff, 0, 0), + USB_DEVICE_CLASS(0x0572, 0xcb01, 0xff, 0, 0), .driver_info = (unsigned long) &cx82310_info }, { }, -- cgit v1.2.1 From 842159640782539a80153c040d6fc2b80756aa3a Mon Sep 17 00:00:00 2001 From: Nicholas Mc Guire Date: Tue, 3 Mar 2015 05:52:51 -0500 Subject: ide_tape: convert jiffies with jiffies_to_msecs Use jiffies_to_msecs for converting jiffies as it handles all of the corner cases reliably and also helps readability. The printk format is fixed up as jiffies_to_msecs returns unsigned int not unsigned long. Signed-off-by: Nicholas Mc Guire Signed-off-by: David S. Miller --- drivers/ide/ide-tape.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/ide/ide-tape.c b/drivers/ide/ide-tape.c index 1793aea4a7d2..6eb738ca6d2f 100644 --- a/drivers/ide/ide-tape.c +++ b/drivers/ide/ide-tape.c @@ -1793,11 +1793,11 @@ static void idetape_setup(ide_drive_t *drive, idetape_tape_t *tape, int minor) tape->best_dsc_rw_freq = clamp_t(unsigned long, t, IDETAPE_DSC_RW_MIN, IDETAPE_DSC_RW_MAX); printk(KERN_INFO "ide-tape: %s <-> %s: %dKBps, %d*%dkB buffer, " - "%lums tDSC%s\n", + "%ums tDSC%s\n", drive->name, tape->name, *(u16 *)&tape->caps[14], (*(u16 *)&tape->caps[16] * 512) / tape->buffer_size, tape->buffer_size / 1024, - tape->best_dsc_rw_freq * 1000 / HZ, + jiffies_to_msecs(tape->best_dsc_rw_freq), (drive->dev_flags & IDE_DFLAG_USING_DMA) ? ", DMA" : ""); ide_proc_register_driver(drive, tape->driver); -- cgit v1.2.1 From 4017a7ee693d1cae6735c0dac21594a7c6416c4c Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Mon, 2 Mar 2015 01:10:28 +0100 Subject: netfilter: restore rule tracing via nfnetlink_log MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Since fab4085 ("netfilter: log: nf_log_packet() as real unified interface"), the loginfo structure that is passed to nf_log_packet() is used to explicitly indicate the logger type you want to use. This is a problem for people tracing rules through nfnetlink_log since packets are always routed to the NF_LOG_TYPE logger after the aforementioned patch. We can fix this by removing the trace loginfo structures, but that still changes the log level from 4 to 5 for tracing messages and there may be someone relying on this outthere. So let's just introduce a new nf_log_trace() function that restores the former behaviour. Reported-by: Markus Kötter Signed-off-by: Pablo Neira Ayuso --- include/net/netfilter/nf_log.h | 10 ++++++++++ net/ipv4/netfilter/ip_tables.c | 6 +++--- net/ipv6/netfilter/ip6_tables.c | 6 +++--- net/netfilter/nf_log.c | 24 ++++++++++++++++++++++++ net/netfilter/nf_tables_core.c | 8 ++++---- 5 files changed, 44 insertions(+), 10 deletions(-) diff --git a/include/net/netfilter/nf_log.h b/include/net/netfilter/nf_log.h index 534e1f2ac4fc..57639fca223a 100644 --- a/include/net/netfilter/nf_log.h +++ b/include/net/netfilter/nf_log.h @@ -79,6 +79,16 @@ void nf_log_packet(struct net *net, const struct nf_loginfo *li, const char *fmt, ...); +__printf(8, 9) +void nf_log_trace(struct net *net, + u_int8_t pf, + unsigned int hooknum, + const struct sk_buff *skb, + const struct net_device *in, + const struct net_device *out, + const struct nf_loginfo *li, + const char *fmt, ...); + struct nf_log_buf; struct nf_log_buf *nf_log_buf_open(void); diff --git a/net/ipv4/netfilter/ip_tables.c b/net/ipv4/netfilter/ip_tables.c index 99e810f84671..cf5e82f39d3b 100644 --- a/net/ipv4/netfilter/ip_tables.c +++ b/net/ipv4/netfilter/ip_tables.c @@ -272,9 +272,9 @@ static void trace_packet(const struct sk_buff *skb, &chainname, &comment, &rulenum) != 0) break; - nf_log_packet(net, AF_INET, hook, skb, in, out, &trace_loginfo, - "TRACE: %s:%s:%s:%u ", - tablename, chainname, comment, rulenum); + nf_log_trace(net, AF_INET, hook, skb, in, out, &trace_loginfo, + "TRACE: %s:%s:%s:%u ", + tablename, chainname, comment, rulenum); } #endif diff --git a/net/ipv6/netfilter/ip6_tables.c b/net/ipv6/netfilter/ip6_tables.c index e080fbbbc0e5..bb00c6f2a885 100644 --- a/net/ipv6/netfilter/ip6_tables.c +++ b/net/ipv6/netfilter/ip6_tables.c @@ -298,9 +298,9 @@ static void trace_packet(const struct sk_buff *skb, &chainname, &comment, &rulenum) != 0) break; - nf_log_packet(net, AF_INET6, hook, skb, in, out, &trace_loginfo, - "TRACE: %s:%s:%s:%u ", - tablename, chainname, comment, rulenum); + nf_log_trace(net, AF_INET6, hook, skb, in, out, &trace_loginfo, + "TRACE: %s:%s:%s:%u ", + tablename, chainname, comment, rulenum); } #endif diff --git a/net/netfilter/nf_log.c b/net/netfilter/nf_log.c index 0d8448f19dfe..675d12c69e32 100644 --- a/net/netfilter/nf_log.c +++ b/net/netfilter/nf_log.c @@ -212,6 +212,30 @@ void nf_log_packet(struct net *net, } EXPORT_SYMBOL(nf_log_packet); +void nf_log_trace(struct net *net, + u_int8_t pf, + unsigned int hooknum, + const struct sk_buff *skb, + const struct net_device *in, + const struct net_device *out, + const struct nf_loginfo *loginfo, const char *fmt, ...) +{ + va_list args; + char prefix[NF_LOG_PREFIXLEN]; + const struct nf_logger *logger; + + rcu_read_lock(); + logger = rcu_dereference(net->nf.nf_loggers[pf]); + if (logger) { + va_start(args, fmt); + vsnprintf(prefix, sizeof(prefix), fmt, args); + va_end(args); + logger->logfn(net, pf, hooknum, skb, in, out, loginfo, prefix); + } + rcu_read_unlock(); +} +EXPORT_SYMBOL(nf_log_trace); + #define S_SIZE (1024 - (sizeof(unsigned int) + 1)) struct nf_log_buf { diff --git a/net/netfilter/nf_tables_core.c b/net/netfilter/nf_tables_core.c index 3b90eb2b2c55..2d298dccb6dd 100644 --- a/net/netfilter/nf_tables_core.c +++ b/net/netfilter/nf_tables_core.c @@ -94,10 +94,10 @@ static void nft_trace_packet(const struct nft_pktinfo *pkt, { struct net *net = dev_net(pkt->in ? pkt->in : pkt->out); - nf_log_packet(net, pkt->xt.family, pkt->ops->hooknum, pkt->skb, pkt->in, - pkt->out, &trace_loginfo, "TRACE: %s:%s:%s:%u ", - chain->table->name, chain->name, comments[type], - rulenum); + nf_log_trace(net, pkt->xt.family, pkt->ops->hooknum, pkt->skb, pkt->in, + pkt->out, &trace_loginfo, "TRACE: %s:%s:%s:%u ", + chain->table->name, chain->name, comments[type], + rulenum); } unsigned int -- cgit v1.2.1 From 6b7a783ebd2181aa2e0e6f9f5509da8466e321e3 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Sat, 21 Feb 2015 15:15:16 +1100 Subject: mmc: pwrseq_simple: fix error path in mmc_pwrseq_simple_alloc The current error-path code (when gpiod_get_index() reports an error) can never free pwrseq->reset_gpios[0], but might try to tree pwrseq->reset_gpios[-1], which has unfortunate consequences. Signed-off-by: NeilBrown Fixes: 934f1f48330ed695927a51fa068dc5d673f2da19 Acked-by: Javier Martinez Canillas Signed-off-by: Ulf Hansson Reported-by: Srinivas Kandagatla --- drivers/mmc/core/pwrseq_simple.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/mmc/core/pwrseq_simple.c b/drivers/mmc/core/pwrseq_simple.c index e9f1d8d84613..c53f14a7ce54 100644 --- a/drivers/mmc/core/pwrseq_simple.c +++ b/drivers/mmc/core/pwrseq_simple.c @@ -124,7 +124,7 @@ int mmc_pwrseq_simple_alloc(struct mmc_host *host, struct device *dev) PTR_ERR(pwrseq->reset_gpios[i]) != -ENOSYS) { ret = PTR_ERR(pwrseq->reset_gpios[i]); - while (--i) + while (i--) gpiod_put(pwrseq->reset_gpios[i]); goto clk_put; -- cgit v1.2.1 From d7c146053dd195b90c79b9b8131431f44541d015 Mon Sep 17 00:00:00 2001 From: Laurent Pinchart Date: Wed, 18 Mar 2015 00:21:32 +0200 Subject: of/irq: Fix of_irq_parse_one() returned error codes The error code paths that require cleanup use a goto to jump to the cleanup code and return an error code. However, the error code variable res, which is initialized to -EINVAL when declared, is then overwritten with the return value of of_parse_phandle_with_args(), and reused as the return code from of_irq_parse_one(). This leads to an undetermined error being returned instead of the expected -EINVAL value. Fix it. Signed-off-by: Laurent Pinchart Cc: stable@vger.kernel.org # 3.13+ Signed-off-by: Rob Herring --- drivers/of/irq.c | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) diff --git a/drivers/of/irq.c b/drivers/of/irq.c index 0d7765807f49..1a7980692f25 100644 --- a/drivers/of/irq.c +++ b/drivers/of/irq.c @@ -290,7 +290,7 @@ int of_irq_parse_one(struct device_node *device, int index, struct of_phandle_ar struct device_node *p; const __be32 *intspec, *tmp, *addr; u32 intsize, intlen; - int i, res = -EINVAL; + int i, res; pr_debug("of_irq_parse_one: dev=%s, index=%d\n", of_node_full_name(device), index); @@ -323,15 +323,19 @@ int of_irq_parse_one(struct device_node *device, int index, struct of_phandle_ar /* Get size of interrupt specifier */ tmp = of_get_property(p, "#interrupt-cells", NULL); - if (tmp == NULL) + if (tmp == NULL) { + res = -EINVAL; goto out; + } intsize = be32_to_cpu(*tmp); pr_debug(" intsize=%d intlen=%d\n", intsize, intlen); /* Check index */ - if ((index + 1) * intsize > intlen) + if ((index + 1) * intsize > intlen) { + res = -EINVAL; goto out; + } /* Copy intspec into irq structure */ intspec += index * intsize; -- cgit v1.2.1 From 5ca1b0dd016701f67994414a2af50dec6efcf103 Mon Sep 17 00:00:00 2001 From: Brian Norris Date: Tue, 17 Mar 2015 12:30:32 -0700 Subject: of: unittest: Add option string test case with longer path There were regressions seen with commit 106937e8ccdc ("of: fix handling of '/' in options for of_find_node_by_path()"), where we couldn't handle extra '/' before the ':'. Let's test for this now. Confirmed that this test fails without the previous patch and passes when patched. All other tests pass. Signed-off-by: Brian Norris Acked-by: Leif Lindholm Signed-off-by: Rob Herring --- drivers/of/unittest.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/drivers/of/unittest.c b/drivers/of/unittest.c index aba8946cac46..52c45c7df07f 100644 --- a/drivers/of/unittest.c +++ b/drivers/of/unittest.c @@ -97,6 +97,11 @@ static void __init of_selftest_find_node_by_name(void) "option path test, subcase #1 failed\n"); of_node_put(np); + np = of_find_node_opts_by_path("/testcase-data/testcase-device1:test/option", &options); + selftest(np && !strcmp("test/option", options), + "option path test, subcase #2 failed\n"); + of_node_put(np); + np = of_find_node_opts_by_path("/testcase-data:testoption", NULL); selftest(np, "NULL option path test failed\n"); of_node_put(np); -- cgit v1.2.1 From 721a09e95c786346b4188863a1cfa3909c76f690 Mon Sep 17 00:00:00 2001 From: Brian Norris Date: Tue, 17 Mar 2015 12:30:31 -0700 Subject: of: handle both '/' and ':' in path strings Commit 106937e8ccdc ("of: fix handling of '/' in options for of_find_node_by_path()") caused a regression in OF handling of stdout-path. While it fixes some cases which have '/' after the ':', it breaks cases where there is more than one '/' *before* the ':'. For example, it breaks this boot string stdout-path = "/rdb/serial@f040ab00:115200"; So rather than doing sequentialized checks (first for '/', then for ':'; or vice versa), to get the correct behavior we need to check for the first occurrence of either one of them. It so happens that the handy strcspn() helper can do just that. Fixes: 106937e8ccdc ("of: fix handling of '/' in options for of_find_node_by_path()") Signed-off-by: Brian Norris Cc: stable@vger.kernel.org # 3.19 Acked-by: Leif Lindholm Signed-off-by: Rob Herring --- drivers/of/base.c | 7 +------ 1 file changed, 1 insertion(+), 6 deletions(-) diff --git a/drivers/of/base.c b/drivers/of/base.c index adb8764861c0..966d6fdcf427 100644 --- a/drivers/of/base.c +++ b/drivers/of/base.c @@ -715,13 +715,8 @@ static struct device_node *__of_find_node_by_path(struct device_node *parent, { struct device_node *child; int len; - const char *end; - end = strchr(path, ':'); - if (!end) - end = strchrnul(path, '/'); - - len = end - path; + len = strcspn(path, "/:"); if (!len) return NULL; -- cgit v1.2.1 From f64255b5072d9c46cef8655d51cf7e10285abed7 Mon Sep 17 00:00:00 2001 From: Peter Hurley Date: Tue, 17 Mar 2015 16:46:33 -0400 Subject: Revert "of: Fix premature bootconsole disable with 'stdout-path'" This reverts commit 2fa645cb2703d9b3786d850db815414dfeefa51d. The assumption that at least 1 preferred console will be registered when the stdout-path property is set is invalid, which can result in _no_ consoles. Signed-off-by: Peter Hurley Signed-off-by: Rob Herring --- drivers/of/base.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/drivers/of/base.c b/drivers/of/base.c index 966d6fdcf427..8f165b112e03 100644 --- a/drivers/of/base.c +++ b/drivers/of/base.c @@ -1888,10 +1888,8 @@ void of_alias_scan(void * (*dt_alloc)(u64 size, u64 align)) name = of_get_property(of_chosen, "linux,stdout-path", NULL); if (IS_ENABLED(CONFIG_PPC) && !name) name = of_get_property(of_aliases, "stdout", NULL); - if (name) { + if (name) of_stdout = of_find_node_opts_by_path(name, &of_stdout_options); - add_preferred_console("stdout-path", 0, NULL); - } } if (!of_aliases) -- cgit v1.2.1 From 94e4fe2cab3d43b3ba7c3f721743006a8c9d913a Mon Sep 17 00:00:00 2001 From: Tom Van Braeckel Date: Mon, 12 Jan 2015 05:22:16 +0100 Subject: fuse: explicitly set /dev/fuse file's private_data The misc subsystem (which is used for /dev/fuse) initializes private_data to point to the misc device when a driver has registered a custom open file operation, and initializes it to NULL when a custom open file operation has *not* been provided. This subtle quirk is confusing, to the point where kernel code registers *empty* file open operations to have private_data point to the misc device structure. And it leads to bugs, where the addition or removal of a custom open file operation surprisingly changes the initial contents of a file's private_data structure. So to simplify things in the misc subsystem, a patch [1] has been proposed to *always* set the private_data to point to the misc device, instead of only doing this when a custom open file operation has been registered. But before this patch can be applied we need to modify drivers that make the assumption that a misc device file's private_data is initialized to NULL because they didn't register a custom open file operation, so they don't rely on this assumption anymore. FUSE uses private_data to store the fuse_conn and errors out if this is not initialized to NULL at mount time. Hence, we now set a file's private_data to NULL explicitly, to be independent of whatever value the misc subsystem initializes it to by default. [1] https://lkml.org/lkml/2014/12/4/939 Reported-by: Giedrius Statkevicius Reported-by: Thierry Reding Signed-off-by: Tom Van Braeckel Signed-off-by: Miklos Szeredi --- fs/fuse/dev.c | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index 71c4619af333..39706c57ad3c 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -1353,6 +1353,17 @@ static ssize_t fuse_dev_do_read(struct fuse_conn *fc, struct file *file, return err; } +static int fuse_dev_open(struct inode *inode, struct file *file) +{ + /* + * The fuse device's file's private_data is used to hold + * the fuse_conn(ection) when it is mounted, and is used to + * keep track of whether the file has been mounted already. + */ + file->private_data = NULL; + return 0; +} + static ssize_t fuse_dev_read(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos) { @@ -2220,6 +2231,7 @@ static int fuse_dev_fasync(int fd, struct file *file, int on) const struct file_operations fuse_dev_operations = { .owner = THIS_MODULE, + .open = fuse_dev_open, .llseek = no_llseek, .read = do_sync_read, .aio_read = fuse_dev_read, -- cgit v1.2.1 From 1ac31de744202a3a14601170a57f155b4a8d2c21 Mon Sep 17 00:00:00 2001 From: Mark James Date: Tue, 17 Mar 2015 21:35:23 +0000 Subject: ARM: socfpga: dts: fix spi1 interrupt The socfpga.dtsi currently has the wrong interrupt number set for SPI master 1 Trying to use the master without this change results in the kernel boot process waiting forever for an interrupt that will never occur while attempting to probe any slave devices configured in the device tree as being under SPI master 1. The change works for the Cyclone V, and according to the Arria 5 handbook should be good there too. Signed-off-by: Mark James Acked-by: Steffen Trumtrar Signed-off-by: Dinh Nguyen --- arch/arm/boot/dts/socfpga.dtsi | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm/boot/dts/socfpga.dtsi b/arch/arm/boot/dts/socfpga.dtsi index 9d8760956752..d9176e606173 100644 --- a/arch/arm/boot/dts/socfpga.dtsi +++ b/arch/arm/boot/dts/socfpga.dtsi @@ -660,7 +660,7 @@ #address-cells = <1>; #size-cells = <0>; reg = <0xfff01000 0x1000>; - interrupts = <0 156 4>; + interrupts = <0 155 4>; num-cs = <4>; clocks = <&spi_m_clk>; status = "disabled"; -- cgit v1.2.1 From 67d8712dcc70aa16d8e14a52eb73870e3cbddfc2 Mon Sep 17 00:00:00 2001 From: Shuah Khan Date: Wed, 18 Mar 2015 11:57:39 -0600 Subject: selftests: Fix build failures when invoked from kselftest target Several tests that rely on implicit build rules fail to build, when invoked from the main Makefile kselftest target. These failures are due to --no-builtin-rules and --no-builtin-variables options set in the inherited MAKEFLAGS. --no-builtin-rules eliminates the use of built-in implicit rules and --no-builtin-variables is for not defining built-in variables. These two options override the use of implicit rules resulting in build failures. In addition, inherited LDFLAGS result in build failures and there is no need to define LDFLAGS. Clear LDFLAGS and MAKEFLAG when make is invoked from the main Makefile kselftest target. Fixing this at selftests Makefile avoids changing the main Makefile and keeps this change self contained at selftests level. Signed-off-by: Shuah Khan Acked-by: Michael Ellerman --- tools/testing/selftests/Makefile | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/tools/testing/selftests/Makefile b/tools/testing/selftests/Makefile index 4e511221a0c1..0db571340edb 100644 --- a/tools/testing/selftests/Makefile +++ b/tools/testing/selftests/Makefile @@ -22,6 +22,14 @@ TARGETS += vm TARGETS_HOTPLUG = cpu-hotplug TARGETS_HOTPLUG += memory-hotplug +# Clear LDFLAGS and MAKEFLAGS if called from main +# Makefile to avoid test build failures when test +# Makefile doesn't have explicit build rules. +ifeq (1,$(MAKELEVEL)) +undefine LDFLAGS +override MAKEFLAGS = +endif + all: for TARGET in $(TARGETS); do \ make -C $$TARGET; \ -- cgit v1.2.1 From 5067c0469c643512f24786990e315f9c15cc7d24 Mon Sep 17 00:00:00 2001 From: Shaohua Li Date: Thu, 12 Mar 2015 10:32:18 -0700 Subject: ata: Add a new flag to destinguish sas controller SAS controller has its own tag allocation, which doesn't directly match to ATA tag, so SAS and SATA have different code path for ata tags. Originally we use port->scsi_host (98bd4be1) to destinguish SAS controller, but libsas set ->scsi_host too, so we can't use it for the destinguish, we add a new flag for this purpose. Without this patch, the following oops can happen because scsi-mq uses a host-wide tag map shared among all devices with some integer tag values >= ATA_MAX_QUEUE. These unexpectedly high tag values cause __ata_qc_from_tag() to return NULL, which is then dereferenced in ata_qc_new_init(). BUG: unable to handle kernel NULL pointer dereference at 0000000000000058 IP: [] ata_qc_new_init+0x3e/0x120 PGD 32adf0067 PUD 32adf1067 PMD 0 Oops: 0002 [#1] SMP DEBUG_PAGEALLOC Modules linked in: iscsi_tcp libiscsi_tcp libiscsi scsi_transport_iscsi igb i2c_algo_bit ptp pps_core pm80xx libsas scsi_transport_sas sg coretemp eeprom w83795 i2c_i801 CPU: 4 PID: 1450 Comm: cydiskbench Not tainted 4.0.0-rc3 #1 Hardware name: Supermicro X8DTH-i/6/iF/6F/X8DTH, BIOS 2.1b 05/04/12 task: ffff8800ba86d500 ti: ffff88032a064000 task.ti: ffff88032a064000 RIP: 0010:[] [] ata_qc_new_init+0x3e/0x120 RSP: 0018:ffff88032a067858 EFLAGS: 00010046 RAX: 0000000000000000 RBX: ffff8800ba0d2230 RCX: 000000000000002a RDX: ffffffff80505ae0 RSI: 0000000000000020 RDI: ffff8800ba0d2230 RBP: ffff88032a067868 R08: 0000000000000201 R09: 0000000000000001 R10: 0000000000000000 R11: 0000000000000000 R12: ffff8800ba0d0000 R13: ffff8800ba0d2230 R14: ffffffff80505ae0 R15: ffff8800ba0d0000 FS: 0000000041223950(0063) GS:ffff88033e480000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: 0000000000000058 CR3: 000000032a0a3000 CR4: 00000000000006e0 Stack: ffff880329eee758 ffff880329eee758 ffff88032a0678a8 ffffffff80502dad ffff8800ba167978 ffff880329eee758 ffff88032bf9c520 ffff8800ba167978 ffff88032bf9c520 ffff88032bf9a290 ffff88032a0678b8 ffffffff80506909 Call Trace: [] ata_scsi_translate+0x3d/0x1b0 [] ata_sas_queuecmd+0x149/0x2a0 [] sas_queuecommand+0xa0/0x1f0 [libsas] [] scsi_dispatch_cmd+0xd4/0x1a0 [] scsi_queue_rq+0x66f/0x7f0 [] __blk_mq_run_hw_queue+0x208/0x3f0 [] blk_mq_run_hw_queue+0x88/0xc0 [] blk_mq_insert_request+0xc4/0x130 [] blk_execute_rq_nowait+0x73/0x160 [] sg_common_write+0x3da/0x720 [sg] [] sg_new_write+0x250/0x360 [sg] [] sg_write+0x13b/0x450 [sg] [] vfs_write+0xd1/0x1b0 [] SyS_write+0x54/0xc0 [] system_call_fastpath+0x12/0x17 tj: updated description. Fixes: 12cb5ce101ab ("libata: use blk taging") Reported-and-tested-by: Tony Battersby Signed-off-by: Shaohua Li Signed-off-by: Tejun Heo --- drivers/ata/libata-core.c | 4 ++-- drivers/scsi/ipr.c | 3 ++- drivers/scsi/libsas/sas_ata.c | 3 ++- include/linux/libata.h | 1 + 4 files changed, 7 insertions(+), 4 deletions(-) diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c index 4c35f0822d06..ef150ebb4c30 100644 --- a/drivers/ata/libata-core.c +++ b/drivers/ata/libata-core.c @@ -4737,7 +4737,7 @@ struct ata_queued_cmd *ata_qc_new_init(struct ata_device *dev, int tag) return NULL; /* libsas case */ - if (!ap->scsi_host) { + if (ap->flags & ATA_FLAG_SAS_HOST) { tag = ata_sas_allocate_tag(ap); if (tag < 0) return NULL; @@ -4776,7 +4776,7 @@ void ata_qc_free(struct ata_queued_cmd *qc) tag = qc->tag; if (likely(ata_tag_valid(tag))) { qc->tag = ATA_TAG_POISON; - if (!ap->scsi_host) + if (ap->flags & ATA_FLAG_SAS_HOST) ata_sas_free_tag(tag, ap); } } diff --git a/drivers/scsi/ipr.c b/drivers/scsi/ipr.c index 9219953ee949..d9afc51af7d3 100644 --- a/drivers/scsi/ipr.c +++ b/drivers/scsi/ipr.c @@ -6815,7 +6815,8 @@ static struct ata_port_operations ipr_sata_ops = { }; static struct ata_port_info sata_port_info = { - .flags = ATA_FLAG_SATA | ATA_FLAG_PIO_DMA, + .flags = ATA_FLAG_SATA | ATA_FLAG_PIO_DMA | + ATA_FLAG_SAS_HOST, .pio_mask = ATA_PIO4_ONLY, .mwdma_mask = ATA_MWDMA2, .udma_mask = ATA_UDMA6, diff --git a/drivers/scsi/libsas/sas_ata.c b/drivers/scsi/libsas/sas_ata.c index 932d9cc98d2f..9c706d8c1441 100644 --- a/drivers/scsi/libsas/sas_ata.c +++ b/drivers/scsi/libsas/sas_ata.c @@ -547,7 +547,8 @@ static struct ata_port_operations sas_sata_ops = { }; static struct ata_port_info sata_port_info = { - .flags = ATA_FLAG_SATA | ATA_FLAG_PIO_DMA | ATA_FLAG_NCQ, + .flags = ATA_FLAG_SATA | ATA_FLAG_PIO_DMA | ATA_FLAG_NCQ | + ATA_FLAG_SAS_HOST, .pio_mask = ATA_PIO4, .mwdma_mask = ATA_MWDMA2, .udma_mask = ATA_UDMA6, diff --git a/include/linux/libata.h b/include/linux/libata.h index fc03efa64ffe..6b08cc106c21 100644 --- a/include/linux/libata.h +++ b/include/linux/libata.h @@ -232,6 +232,7 @@ enum { * led */ ATA_FLAG_NO_DIPM = (1 << 23), /* host not happy with DIPM */ ATA_FLAG_LOWTAG = (1 << 24), /* host wants lowest available tag */ + ATA_FLAG_SAS_HOST = (1 << 25), /* SAS host */ /* bits 24:31 of ap->flags are reserved for LLD specific flags */ -- cgit v1.2.1 From 133d558216d9db3617a9fdeebd1bce9afff5f973 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 5 Mar 2015 14:17:31 +0100 Subject: Subject: nfsd: don't recursively call nfsd4_cb_layout_fail Due to a merge error when creating c5c707f9 ("nfsd: implement pNFS layout recalls"), we recursively call nfsd4_cb_layout_fail from itself, leading to stack overflows. Signed-off-by: Christoph Hellwig Fixes: c5c707f9 ("nfsd: implement pNFS layout recalls") Signed-off-by: J. Bruce Fields --- fs/nfsd/nfs4layouts.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/fs/nfsd/nfs4layouts.c b/fs/nfsd/nfs4layouts.c index 3c1bfa1..1028a06 100644 --- a/fs/nfsd/nfs4layouts.c +++ b/fs/nfsd/nfs4layouts.c @@ -587,8 +587,6 @@ nfsd4_cb_layout_fail(struct nfs4_layout_stateid *ls) rpc_ntop((struct sockaddr *)&clp->cl_addr, addr_str, sizeof(addr_str)); - nfsd4_cb_layout_fail(ls); - printk(KERN_WARNING "nfsd: client %s failed to respond to layout recall. " " Fencing..\n", addr_str); -- 1.9.1 --- fs/nfsd/nfs4layouts.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/fs/nfsd/nfs4layouts.c b/fs/nfsd/nfs4layouts.c index 3c1bfa155571..1028a0629543 100644 --- a/fs/nfsd/nfs4layouts.c +++ b/fs/nfsd/nfs4layouts.c @@ -587,8 +587,6 @@ nfsd4_cb_layout_fail(struct nfs4_layout_stateid *ls) rpc_ntop((struct sockaddr *)&clp->cl_addr, addr_str, sizeof(addr_str)); - nfsd4_cb_layout_fail(ls); - printk(KERN_WARNING "nfsd: client %s failed to respond to layout recall. " " Fencing..\n", addr_str); -- cgit v1.2.1 From c6b570d97c0e77f570bb6b2ed30d372b2b1e9aae Mon Sep 17 00:00:00 2001 From: Philipp Zabel Date: Mon, 9 Mar 2015 12:20:13 +0100 Subject: regmap: introduce regmap_name to fix syscon regmap trace events This patch fixes a NULL pointer dereference when enabling regmap event tracing in the presence of a syscon regmap, introduced by commit bdb0066df96e ("mfd: syscon: Decouple syscon interface from platform devices"). That patch introduced syscon regmaps that have their dev field set to NULL. The regmap trace events expect it to point to a valid struct device and feed it to dev_name(): $ echo 1 > /sys/kernel/debug/tracing/events/regmap/enable Unable to handle kernel NULL pointer dereference at virtual address 0000002c pgd = 80004000 [0000002c] *pgd=00000000 Internal error: Oops: 17 [#1] SMP ARM Modules linked in: coda videobuf2_vmalloc CPU: 0 PID: 304 Comm: kworker/0:2 Not tainted 4.0.0-rc2+ #9197 Hardware name: Freescale i.MX6 Quad/DualLite (Device Tree) Workqueue: events_freezable thermal_zone_device_check task: 9f25a200 ti: 9f1ee000 task.ti: 9f1ee000 PC is at ftrace_raw_event_regmap_block+0x3c/0xe4 LR is at _regmap_raw_read+0x1bc/0x1cc pc : [<803636e8>] lr : [<80365f2c>] psr: 600f0093 sp : 9f1efd78 ip : 9f1efdb8 fp : 9f1efdb4 r10: 00000004 r9 : 00000001 r8 : 00000001 r7 : 00000180 r6 : 00000000 r5 : 9f00e3c0 r4 : 00000003 r3 : 00000001 r2 : 00000180 r1 : 00000000 r0 : 9f00e3c0 Flags: nZCv IRQs off FIQs on Mode SVC_32 ISA ARM Segment kernel Control: 10c5387d Table: 2d91004a DAC: 00000015 Process kworker/0:2 (pid: 304, stack limit = 0x9f1ee210) Stack: (0x9f1efd78 to 0x9f1f0000) fd60: 9f1efda4 9f1efd88 fd80: 800708c0 805f9510 80927140 800f0013 9f1fc800 9eb2f490 00000000 00000180 fda0: 808e3840 00000001 9f1efdfc 9f1efdb8 80365f2c 803636b8 805f8958 800708e0 fdc0: a00f0013 803636ac 9f16de00 00000180 80927140 9f1fc800 9f1fc800 9f1efe6c fde0: 9f1efe6c 9f732400 00000000 00000000 9f1efe1c 9f1efe00 80365f70 80365d7c fe00: 80365f3c 9f1fc800 9f1fc800 00000180 9f1efe44 9f1efe20 803656a4 80365f48 fe20: 9f1fc800 00000180 9f1efe6c 9f1efe6c 9f732400 00000000 9f1efe64 9f1efe48 fe40: 803657bc 80365634 00000001 9e95f910 9f1fc800 9f1efeb4 9f1efe8c 9f1efe68 fe60: 80452ac0 80365778 9f1efe8c 9f1efe78 9e93d400 9e93d5e8 9f1efeb4 9f72ef40 fe80: 9f1efeac 9f1efe90 8044e11c 80452998 8045298c 9e93d608 9e93d400 808e1978 fea0: 9f1efecc 9f1efeb0 8044fd14 8044e0d0 ffffffff 9f25a200 9e93d608 9e481380 fec0: 9f1efedc 9f1efed0 8044fde8 8044fcec 9f1eff1c 9f1efee0 80038d50 8044fdd8 fee0: 9f1ee020 9f72ef40 9e481398 00000000 00000008 9f72ef54 9f1ee020 9f72ef40 ff00: 9e481398 9e481380 00000008 9f72ef40 9f1eff5c 9f1eff20 80039754 80038bfc ff20: 00000000 9e481380 80894100 808e1662 00000000 9e4f2ec0 00000000 9e481380 ff40: 800396f8 00000000 00000000 00000000 9f1effac 9f1eff60 8003e020 80039704 ff60: ffffffff 00000000 ffffffff 9e481380 00000000 00000000 9f1eff78 9f1eff78 ff80: 00000000 00000000 9f1eff88 9f1eff88 9e4f2ec0 8003df30 00000000 00000000 ffa0: 00000000 9f1effb0 8000eb60 8003df3c 00000000 00000000 00000000 00000000 ffc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 ffe0: 00000000 00000000 00000000 00000000 00000013 00000000 ffffffff ffffffff Backtrace: [<803636ac>] (ftrace_raw_event_regmap_block) from [<80365f2c>] (_regmap_raw_read+0x1bc/0x1cc) r9:00000001 r8:808e3840 r7:00000180 r6:00000000 r5:9eb2f490 r4:9f1fc800 [<80365d70>] (_regmap_raw_read) from [<80365f70>] (_regmap_bus_read+0x34/0x6c) r10:00000000 r9:00000000 r8:9f732400 r7:9f1efe6c r6:9f1efe6c r5:9f1fc800 r4:9f1fc800 [<80365f3c>] (_regmap_bus_read) from [<803656a4>] (_regmap_read+0x7c/0x144) r6:00000180 r5:9f1fc800 r4:9f1fc800 r3:80365f3c [<80365628>] (_regmap_read) from [<803657bc>] (regmap_read+0x50/0x70) r9:00000000 r8:9f732400 r7:9f1efe6c r6:9f1efe6c r5:00000180 r4:9f1fc800 [<8036576c>] (regmap_read) from [<80452ac0>] (imx_get_temp+0x134/0x1a4) r6:9f1efeb4 r5:9f1fc800 r4:9e95f910 r3:00000001 [<8045298c>] (imx_get_temp) from [<8044e11c>] (thermal_zone_get_temp+0x58/0x74) r7:9f72ef40 r6:9f1efeb4 r5:9e93d5e8 r4:9e93d400 [<8044e0c4>] (thermal_zone_get_temp) from [<8044fd14>] (thermal_zone_device_update+0x34/0xec) r6:808e1978 r5:9e93d400 r4:9e93d608 r3:8045298c [<8044fce0>] (thermal_zone_device_update) from [<8044fde8>] (thermal_zone_device_check+0x1c/0x20) r5:9e481380 r4:9e93d608 [<8044fdcc>] (thermal_zone_device_check) from [<80038d50>] (process_one_work+0x160/0x3d4) [<80038bf0>] (process_one_work) from [<80039754>] (worker_thread+0x5c/0x4f4) r10:9f72ef40 r9:00000008 r8:9e481380 r7:9e481398 r6:9f72ef40 r5:9f1ee020 r4:9f72ef54 [<800396f8>] (worker_thread) from [<8003e020>] (kthread+0xf0/0x108) r10:00000000 r9:00000000 r8:00000000 r7:800396f8 r6:9e481380 r5:00000000 r4:9e4f2ec0 [<8003df30>] (kthread) from [<8000eb60>] (ret_from_fork+0x14/0x34) r7:00000000 r6:00000000 r5:8003df30 r4:9e4f2ec0 Code: e3140040 1a00001a e3140020 1a000016 (e596002c) ---[ end trace 193c15c2494ec960 ]--- Fixes: bdb0066df96e (mfd: syscon: Decouple syscon interface from platform devices) Signed-off-by: Philipp Zabel Signed-off-by: Mark Brown Cc: stable@vger.kernel.org --- drivers/base/regmap/internal.h | 8 +++ drivers/base/regmap/regcache.c | 16 +++--- drivers/base/regmap/regmap.c | 32 +++++------ include/trace/events/regmap.h | 123 ++++++++++++++++++++--------------------- 4 files changed, 91 insertions(+), 88 deletions(-) diff --git a/drivers/base/regmap/internal.h b/drivers/base/regmap/internal.h index beb8b27d4621..a13587b5c2be 100644 --- a/drivers/base/regmap/internal.h +++ b/drivers/base/regmap/internal.h @@ -243,4 +243,12 @@ extern struct regcache_ops regcache_rbtree_ops; extern struct regcache_ops regcache_lzo_ops; extern struct regcache_ops regcache_flat_ops; +static inline const char *regmap_name(const struct regmap *map) +{ + if (map->dev) + return dev_name(map->dev); + + return map->name; +} + #endif diff --git a/drivers/base/regmap/regcache.c b/drivers/base/regmap/regcache.c index f373c35f9e1d..f5db662e951e 100644 --- a/drivers/base/regmap/regcache.c +++ b/drivers/base/regmap/regcache.c @@ -218,7 +218,7 @@ int regcache_read(struct regmap *map, ret = map->cache_ops->read(map, reg, value); if (ret == 0) - trace_regmap_reg_read_cache(map->dev, reg, *value); + trace_regmap_reg_read_cache(map, reg, *value); return ret; } @@ -311,7 +311,7 @@ int regcache_sync(struct regmap *map) dev_dbg(map->dev, "Syncing %s cache\n", map->cache_ops->name); name = map->cache_ops->name; - trace_regcache_sync(map->dev, name, "start"); + trace_regcache_sync(map, name, "start"); if (!map->cache_dirty) goto out; @@ -346,7 +346,7 @@ out: regmap_async_complete(map); - trace_regcache_sync(map->dev, name, "stop"); + trace_regcache_sync(map, name, "stop"); return ret; } @@ -381,7 +381,7 @@ int regcache_sync_region(struct regmap *map, unsigned int min, name = map->cache_ops->name; dev_dbg(map->dev, "Syncing %s cache from %d-%d\n", name, min, max); - trace_regcache_sync(map->dev, name, "start region"); + trace_regcache_sync(map, name, "start region"); if (!map->cache_dirty) goto out; @@ -401,7 +401,7 @@ out: regmap_async_complete(map); - trace_regcache_sync(map->dev, name, "stop region"); + trace_regcache_sync(map, name, "stop region"); return ret; } @@ -428,7 +428,7 @@ int regcache_drop_region(struct regmap *map, unsigned int min, map->lock(map->lock_arg); - trace_regcache_drop_region(map->dev, min, max); + trace_regcache_drop_region(map, min, max); ret = map->cache_ops->drop(map, min, max); @@ -455,7 +455,7 @@ void regcache_cache_only(struct regmap *map, bool enable) map->lock(map->lock_arg); WARN_ON(map->cache_bypass && enable); map->cache_only = enable; - trace_regmap_cache_only(map->dev, enable); + trace_regmap_cache_only(map, enable); map->unlock(map->lock_arg); } EXPORT_SYMBOL_GPL(regcache_cache_only); @@ -493,7 +493,7 @@ void regcache_cache_bypass(struct regmap *map, bool enable) map->lock(map->lock_arg); WARN_ON(map->cache_only && enable); map->cache_bypass = enable; - trace_regmap_cache_bypass(map->dev, enable); + trace_regmap_cache_bypass(map, enable); map->unlock(map->lock_arg); } EXPORT_SYMBOL_GPL(regcache_cache_bypass); diff --git a/drivers/base/regmap/regmap.c b/drivers/base/regmap/regmap.c index f99b098ddabf..dbfe6a69c3da 100644 --- a/drivers/base/regmap/regmap.c +++ b/drivers/base/regmap/regmap.c @@ -1281,7 +1281,7 @@ int _regmap_raw_write(struct regmap *map, unsigned int reg, if (map->async && map->bus->async_write) { struct regmap_async *async; - trace_regmap_async_write_start(map->dev, reg, val_len); + trace_regmap_async_write_start(map, reg, val_len); spin_lock_irqsave(&map->async_lock, flags); async = list_first_entry_or_null(&map->async_free, @@ -1339,8 +1339,7 @@ int _regmap_raw_write(struct regmap *map, unsigned int reg, return ret; } - trace_regmap_hw_write_start(map->dev, reg, - val_len / map->format.val_bytes); + trace_regmap_hw_write_start(map, reg, val_len / map->format.val_bytes); /* If we're doing a single register write we can probably just * send the work_buf directly, otherwise try to do a gather @@ -1372,8 +1371,7 @@ int _regmap_raw_write(struct regmap *map, unsigned int reg, kfree(buf); } - trace_regmap_hw_write_done(map->dev, reg, - val_len / map->format.val_bytes); + trace_regmap_hw_write_done(map, reg, val_len / map->format.val_bytes); return ret; } @@ -1407,12 +1405,12 @@ static int _regmap_bus_formatted_write(void *context, unsigned int reg, map->format.format_write(map, reg, val); - trace_regmap_hw_write_start(map->dev, reg, 1); + trace_regmap_hw_write_start(map, reg, 1); ret = map->bus->write(map->bus_context, map->work_buf, map->format.buf_size); - trace_regmap_hw_write_done(map->dev, reg, 1); + trace_regmap_hw_write_done(map, reg, 1); return ret; } @@ -1470,7 +1468,7 @@ int _regmap_write(struct regmap *map, unsigned int reg, dev_info(map->dev, "%x <= %x\n", reg, val); #endif - trace_regmap_reg_write(map->dev, reg, val); + trace_regmap_reg_write(map, reg, val); return map->reg_write(context, reg, val); } @@ -1773,7 +1771,7 @@ static int _regmap_raw_multi_reg_write(struct regmap *map, for (i = 0; i < num_regs; i++) { int reg = regs[i].reg; int val = regs[i].def; - trace_regmap_hw_write_start(map->dev, reg, 1); + trace_regmap_hw_write_start(map, reg, 1); map->format.format_reg(u8, reg, map->reg_shift); u8 += reg_bytes + pad_bytes; map->format.format_val(u8, val, 0); @@ -1788,7 +1786,7 @@ static int _regmap_raw_multi_reg_write(struct regmap *map, for (i = 0; i < num_regs; i++) { int reg = regs[i].reg; - trace_regmap_hw_write_done(map->dev, reg, 1); + trace_regmap_hw_write_done(map, reg, 1); } return ret; } @@ -2059,15 +2057,13 @@ static int _regmap_raw_read(struct regmap *map, unsigned int reg, void *val, */ u8[0] |= map->read_flag_mask; - trace_regmap_hw_read_start(map->dev, reg, - val_len / map->format.val_bytes); + trace_regmap_hw_read_start(map, reg, val_len / map->format.val_bytes); ret = map->bus->read(map->bus_context, map->work_buf, map->format.reg_bytes + map->format.pad_bytes, val, val_len); - trace_regmap_hw_read_done(map->dev, reg, - val_len / map->format.val_bytes); + trace_regmap_hw_read_done(map, reg, val_len / map->format.val_bytes); return ret; } @@ -2123,7 +2119,7 @@ static int _regmap_read(struct regmap *map, unsigned int reg, dev_info(map->dev, "%x => %x\n", reg, *val); #endif - trace_regmap_reg_read(map->dev, reg, *val); + trace_regmap_reg_read(map, reg, *val); if (!map->cache_bypass) regcache_write(map, reg, *val); @@ -2480,7 +2476,7 @@ void regmap_async_complete_cb(struct regmap_async *async, int ret) struct regmap *map = async->map; bool wake; - trace_regmap_async_io_complete(map->dev); + trace_regmap_async_io_complete(map); spin_lock(&map->async_lock); list_move(&async->list, &map->async_free); @@ -2525,7 +2521,7 @@ int regmap_async_complete(struct regmap *map) if (!map->bus || !map->bus->async_write) return 0; - trace_regmap_async_complete_start(map->dev); + trace_regmap_async_complete_start(map); wait_event(map->async_waitq, regmap_async_is_done(map)); @@ -2534,7 +2530,7 @@ int regmap_async_complete(struct regmap *map) map->async_ret = 0; spin_unlock_irqrestore(&map->async_lock, flags); - trace_regmap_async_complete_done(map->dev); + trace_regmap_async_complete_done(map); return ret; } diff --git a/include/trace/events/regmap.h b/include/trace/events/regmap.h index 23d561512f64..22317d2b52ab 100644 --- a/include/trace/events/regmap.h +++ b/include/trace/events/regmap.h @@ -7,27 +7,26 @@ #include #include -struct device; -struct regmap; +#include "../../../drivers/base/regmap/internal.h" /* * Log register events */ DECLARE_EVENT_CLASS(regmap_reg, - TP_PROTO(struct device *dev, unsigned int reg, + TP_PROTO(struct regmap *map, unsigned int reg, unsigned int val), - TP_ARGS(dev, reg, val), + TP_ARGS(map, reg, val), TP_STRUCT__entry( - __string( name, dev_name(dev) ) - __field( unsigned int, reg ) - __field( unsigned int, val ) + __string( name, regmap_name(map) ) + __field( unsigned int, reg ) + __field( unsigned int, val ) ), TP_fast_assign( - __assign_str(name, dev_name(dev)); + __assign_str(name, regmap_name(map)); __entry->reg = reg; __entry->val = val; ), @@ -39,45 +38,45 @@ DECLARE_EVENT_CLASS(regmap_reg, DEFINE_EVENT(regmap_reg, regmap_reg_write, - TP_PROTO(struct device *dev, unsigned int reg, + TP_PROTO(struct regmap *map, unsigned int reg, unsigned int val), - TP_ARGS(dev, reg, val) + TP_ARGS(map, reg, val) ); DEFINE_EVENT(regmap_reg, regmap_reg_read, - TP_PROTO(struct device *dev, unsigned int reg, + TP_PROTO(struct regmap *map, unsigned int reg, unsigned int val), - TP_ARGS(dev, reg, val) + TP_ARGS(map, reg, val) ); DEFINE_EVENT(regmap_reg, regmap_reg_read_cache, - TP_PROTO(struct device *dev, unsigned int reg, + TP_PROTO(struct regmap *map, unsigned int reg, unsigned int val), - TP_ARGS(dev, reg, val) + TP_ARGS(map, reg, val) ); DECLARE_EVENT_CLASS(regmap_block, - TP_PROTO(struct device *dev, unsigned int reg, int count), + TP_PROTO(struct regmap *map, unsigned int reg, int count), - TP_ARGS(dev, reg, count), + TP_ARGS(map, reg, count), TP_STRUCT__entry( - __string( name, dev_name(dev) ) - __field( unsigned int, reg ) - __field( int, count ) + __string( name, regmap_name(map) ) + __field( unsigned int, reg ) + __field( int, count ) ), TP_fast_assign( - __assign_str(name, dev_name(dev)); + __assign_str(name, regmap_name(map)); __entry->reg = reg; __entry->count = count; ), @@ -89,48 +88,48 @@ DECLARE_EVENT_CLASS(regmap_block, DEFINE_EVENT(regmap_block, regmap_hw_read_start, - TP_PROTO(struct device *dev, unsigned int reg, int count), + TP_PROTO(struct regmap *map, unsigned int reg, int count), - TP_ARGS(dev, reg, count) + TP_ARGS(map, reg, count) ); DEFINE_EVENT(regmap_block, regmap_hw_read_done, - TP_PROTO(struct device *dev, unsigned int reg, int count), + TP_PROTO(struct regmap *map, unsigned int reg, int count), - TP_ARGS(dev, reg, count) + TP_ARGS(map, reg, count) ); DEFINE_EVENT(regmap_block, regmap_hw_write_start, - TP_PROTO(struct device *dev, unsigned int reg, int count), + TP_PROTO(struct regmap *map, unsigned int reg, int count), - TP_ARGS(dev, reg, count) + TP_ARGS(map, reg, count) ); DEFINE_EVENT(regmap_block, regmap_hw_write_done, - TP_PROTO(struct device *dev, unsigned int reg, int count), + TP_PROTO(struct regmap *map, unsigned int reg, int count), - TP_ARGS(dev, reg, count) + TP_ARGS(map, reg, count) ); TRACE_EVENT(regcache_sync, - TP_PROTO(struct device *dev, const char *type, + TP_PROTO(struct regmap *map, const char *type, const char *status), - TP_ARGS(dev, type, status), + TP_ARGS(map, type, status), TP_STRUCT__entry( - __string( name, dev_name(dev) ) - __string( status, status ) - __string( type, type ) - __field( int, type ) + __string( name, regmap_name(map) ) + __string( status, status ) + __string( type, type ) + __field( int, type ) ), TP_fast_assign( - __assign_str(name, dev_name(dev)); + __assign_str(name, regmap_name(map)); __assign_str(status, status); __assign_str(type, type); ), @@ -141,17 +140,17 @@ TRACE_EVENT(regcache_sync, DECLARE_EVENT_CLASS(regmap_bool, - TP_PROTO(struct device *dev, bool flag), + TP_PROTO(struct regmap *map, bool flag), - TP_ARGS(dev, flag), + TP_ARGS(map, flag), TP_STRUCT__entry( - __string( name, dev_name(dev) ) - __field( int, flag ) + __string( name, regmap_name(map) ) + __field( int, flag ) ), TP_fast_assign( - __assign_str(name, dev_name(dev)); + __assign_str(name, regmap_name(map)); __entry->flag = flag; ), @@ -161,32 +160,32 @@ DECLARE_EVENT_CLASS(regmap_bool, DEFINE_EVENT(regmap_bool, regmap_cache_only, - TP_PROTO(struct device *dev, bool flag), + TP_PROTO(struct regmap *map, bool flag), - TP_ARGS(dev, flag) + TP_ARGS(map, flag) ); DEFINE_EVENT(regmap_bool, regmap_cache_bypass, - TP_PROTO(struct device *dev, bool flag), + TP_PROTO(struct regmap *map, bool flag), - TP_ARGS(dev, flag) + TP_ARGS(map, flag) ); DECLARE_EVENT_CLASS(regmap_async, - TP_PROTO(struct device *dev), + TP_PROTO(struct regmap *map), - TP_ARGS(dev), + TP_ARGS(map), TP_STRUCT__entry( - __string( name, dev_name(dev) ) + __string( name, regmap_name(map) ) ), TP_fast_assign( - __assign_str(name, dev_name(dev)); + __assign_str(name, regmap_name(map)); ), TP_printk("%s", __get_str(name)) @@ -194,50 +193,50 @@ DECLARE_EVENT_CLASS(regmap_async, DEFINE_EVENT(regmap_block, regmap_async_write_start, - TP_PROTO(struct device *dev, unsigned int reg, int count), + TP_PROTO(struct regmap *map, unsigned int reg, int count), - TP_ARGS(dev, reg, count) + TP_ARGS(map, reg, count) ); DEFINE_EVENT(regmap_async, regmap_async_io_complete, - TP_PROTO(struct device *dev), + TP_PROTO(struct regmap *map), - TP_ARGS(dev) + TP_ARGS(map) ); DEFINE_EVENT(regmap_async, regmap_async_complete_start, - TP_PROTO(struct device *dev), + TP_PROTO(struct regmap *map), - TP_ARGS(dev) + TP_ARGS(map) ); DEFINE_EVENT(regmap_async, regmap_async_complete_done, - TP_PROTO(struct device *dev), + TP_PROTO(struct regmap *map), - TP_ARGS(dev) + TP_ARGS(map) ); TRACE_EVENT(regcache_drop_region, - TP_PROTO(struct device *dev, unsigned int from, + TP_PROTO(struct regmap *map, unsigned int from, unsigned int to), - TP_ARGS(dev, from, to), + TP_ARGS(map, from, to), TP_STRUCT__entry( - __string( name, dev_name(dev) ) - __field( unsigned int, from ) - __field( unsigned int, to ) + __string( name, regmap_name(map) ) + __field( unsigned int, from ) + __field( unsigned int, to ) ), TP_fast_assign( - __assign_str(name, dev_name(dev)); + __assign_str(name, regmap_name(map)); __entry->from = from; __entry->to = to; ), -- cgit v1.2.1 From 5b0d4b5514bbcce69b516d0742f2cfc84ebd6db3 Mon Sep 17 00:00:00 2001 From: David Ahern Date: Thu, 19 Mar 2015 16:05:57 -0400 Subject: sparc: perf: Remove redundant perf_pmu_{en|dis}able calls perf_pmu_disable is called by core perf code before pmu->del and the enable function is called by core perf code afterwards. No need to call again within sparc_pmu_del. Ditto for pmu->add and sparc_pmu_add. Signed-off-by: David Ahern Acked-by: Bob Picco Signed-off-by: David S. Miller --- arch/sparc/kernel/perf_event.c | 4 ---- 1 file changed, 4 deletions(-) diff --git a/arch/sparc/kernel/perf_event.c b/arch/sparc/kernel/perf_event.c index 46a5e4508752..6dc4e793df4c 100644 --- a/arch/sparc/kernel/perf_event.c +++ b/arch/sparc/kernel/perf_event.c @@ -1101,7 +1101,6 @@ static void sparc_pmu_del(struct perf_event *event, int _flags) int i; local_irq_save(flags); - perf_pmu_disable(event->pmu); for (i = 0; i < cpuc->n_events; i++) { if (event == cpuc->event[i]) { @@ -1127,7 +1126,6 @@ static void sparc_pmu_del(struct perf_event *event, int _flags) } } - perf_pmu_enable(event->pmu); local_irq_restore(flags); } @@ -1361,7 +1359,6 @@ static int sparc_pmu_add(struct perf_event *event, int ef_flags) unsigned long flags; local_irq_save(flags); - perf_pmu_disable(event->pmu); n0 = cpuc->n_events; if (n0 >= sparc_pmu->max_hw_events) @@ -1394,7 +1391,6 @@ nocheck: ret = 0; out: - perf_pmu_enable(event->pmu); local_irq_restore(flags); return ret; } -- cgit v1.2.1 From d51291cb8f32bfae6b331e1838651f3ddefa73a5 Mon Sep 17 00:00:00 2001 From: David Ahern Date: Thu, 19 Mar 2015 16:06:17 -0400 Subject: sparc: perf: Make counting mode actually work Currently perf-stat (aka, counting mode) does not work: $ perf stat ls ... Performance counter stats for 'ls': 1.585665 task-clock (msec) # 0.580 CPUs utilized 24 context-switches # 0.015 M/sec 0 cpu-migrations # 0.000 K/sec 86 page-faults # 0.054 M/sec cycles stalled-cycles-frontend stalled-cycles-backend instructions branches branch-misses 0.002735100 seconds time elapsed The reason is that state is never reset (stays with PERF_HES_UPTODATE set). Add a call to sparc_pmu_enable_event during the added_event handling. Clean up the encoding since pmu_start calls sparc_pmu_enable_event which does the same. Passing PERF_EF_RELOAD to sparc_pmu_start means the call to sparc_perf_event_set_period can be removed as well. With this patch: $ perf stat ls ... Performance counter stats for 'ls': 1.552890 task-clock (msec) # 0.552 CPUs utilized 24 context-switches # 0.015 M/sec 0 cpu-migrations # 0.000 K/sec 86 page-faults # 0.055 M/sec 5,748,997 cycles # 3.702 GHz stalled-cycles-frontend:HG stalled-cycles-backend:HG 1,684,362 instructions:HG # 0.29 insns per cycle 295,133 branches:HG # 190.054 M/sec 28,007 branch-misses:HG # 9.49% of all branches 0.002815665 seconds time elapsed Signed-off-by: David Ahern Acked-by: Bob Picco Signed-off-by: David S. Miller --- arch/sparc/kernel/perf_event.c | 11 +++-------- 1 file changed, 3 insertions(+), 8 deletions(-) diff --git a/arch/sparc/kernel/perf_event.c b/arch/sparc/kernel/perf_event.c index 6dc4e793df4c..af53c25da2e7 100644 --- a/arch/sparc/kernel/perf_event.c +++ b/arch/sparc/kernel/perf_event.c @@ -960,6 +960,8 @@ out: cpuc->pcr[0] |= cpuc->event[0]->hw.config_base; } +static void sparc_pmu_start(struct perf_event *event, int flags); + /* On this PMU each PIC has it's own PCR control register. */ static void calculate_multiple_pcrs(struct cpu_hw_events *cpuc) { @@ -972,20 +974,13 @@ static void calculate_multiple_pcrs(struct cpu_hw_events *cpuc) struct perf_event *cp = cpuc->event[i]; struct hw_perf_event *hwc = &cp->hw; int idx = hwc->idx; - u64 enc; if (cpuc->current_idx[i] != PIC_NO_INDEX) continue; - sparc_perf_event_set_period(cp, hwc, idx); cpuc->current_idx[i] = idx; - enc = perf_event_get_enc(cpuc->events[i]); - cpuc->pcr[idx] &= ~mask_for_index(idx); - if (hwc->state & PERF_HES_STOPPED) - cpuc->pcr[idx] |= nop_for_index(idx); - else - cpuc->pcr[idx] |= event_encoding(enc, idx); + sparc_pmu_start(cp, PERF_EF_RELOAD); } out: for (i = 0; i < cpuc->n_events; i++) { -- cgit v1.2.1 From b5aff55d89c27aedcae9521155b81b6aebb6c5d8 Mon Sep 17 00:00:00 2001 From: David Ahern Date: Thu, 19 Mar 2015 16:06:37 -0400 Subject: sparc: perf: Add support M7 processor The M7 processor has a different hypervisor group id and different PCR fast trap values. PIC read/write functions and PCR bit fields are the same as the T4 so those are reused. Signed-off-by: David Ahern Acked-by: Bob Picco Signed-off-by: David S. Miller --- arch/sparc/include/asm/hypervisor.h | 12 +++++++++++ arch/sparc/kernel/hvapi.c | 1 + arch/sparc/kernel/hvcalls.S | 16 +++++++++++++++ arch/sparc/kernel/pcr.c | 33 ++++++++++++++++++++++++++++++ arch/sparc/kernel/perf_event.c | 40 +++++++++++++++++++++++++++++++++++++ 5 files changed, 102 insertions(+) diff --git a/arch/sparc/include/asm/hypervisor.h b/arch/sparc/include/asm/hypervisor.h index 4f6725ff4c33..f5b6537306f0 100644 --- a/arch/sparc/include/asm/hypervisor.h +++ b/arch/sparc/include/asm/hypervisor.h @@ -2957,6 +2957,17 @@ unsigned long sun4v_t5_set_perfreg(unsigned long reg_num, unsigned long reg_val); #endif + +#define HV_FAST_M7_GET_PERFREG 0x43 +#define HV_FAST_M7_SET_PERFREG 0x44 + +#ifndef __ASSEMBLY__ +unsigned long sun4v_m7_get_perfreg(unsigned long reg_num, + unsigned long *reg_val); +unsigned long sun4v_m7_set_perfreg(unsigned long reg_num, + unsigned long reg_val); +#endif + /* Function numbers for HV_CORE_TRAP. */ #define HV_CORE_SET_VER 0x00 #define HV_CORE_PUTCHAR 0x01 @@ -2981,6 +2992,7 @@ unsigned long sun4v_t5_set_perfreg(unsigned long reg_num, #define HV_GRP_SDIO 0x0108 #define HV_GRP_SDIO_ERR 0x0109 #define HV_GRP_REBOOT_DATA 0x0110 +#define HV_GRP_M7_PERF 0x0114 #define HV_GRP_NIAG_PERF 0x0200 #define HV_GRP_FIRE_PERF 0x0201 #define HV_GRP_N2_CPU 0x0202 diff --git a/arch/sparc/kernel/hvapi.c b/arch/sparc/kernel/hvapi.c index 5c55145bfbf0..662500fa555f 100644 --- a/arch/sparc/kernel/hvapi.c +++ b/arch/sparc/kernel/hvapi.c @@ -48,6 +48,7 @@ static struct api_info api_table[] = { { .group = HV_GRP_VT_CPU, }, { .group = HV_GRP_T5_CPU, }, { .group = HV_GRP_DIAG, .flags = FLAG_PRE_API }, + { .group = HV_GRP_M7_PERF, }, }; static DEFINE_SPINLOCK(hvapi_lock); diff --git a/arch/sparc/kernel/hvcalls.S b/arch/sparc/kernel/hvcalls.S index caedf8320416..afbaba52d2f1 100644 --- a/arch/sparc/kernel/hvcalls.S +++ b/arch/sparc/kernel/hvcalls.S @@ -837,3 +837,19 @@ ENTRY(sun4v_t5_set_perfreg) retl nop ENDPROC(sun4v_t5_set_perfreg) + +ENTRY(sun4v_m7_get_perfreg) + mov %o1, %o4 + mov HV_FAST_M7_GET_PERFREG, %o5 + ta HV_FAST_TRAP + stx %o1, [%o4] + retl + nop +ENDPROC(sun4v_m7_get_perfreg) + +ENTRY(sun4v_m7_set_perfreg) + mov HV_FAST_M7_SET_PERFREG, %o5 + ta HV_FAST_TRAP + retl + nop +ENDPROC(sun4v_m7_set_perfreg) diff --git a/arch/sparc/kernel/pcr.c b/arch/sparc/kernel/pcr.c index 7e967c8018c8..eb978c77c76a 100644 --- a/arch/sparc/kernel/pcr.c +++ b/arch/sparc/kernel/pcr.c @@ -217,6 +217,31 @@ static const struct pcr_ops n5_pcr_ops = { .pcr_nmi_disable = PCR_N4_PICNPT, }; +static u64 m7_pcr_read(unsigned long reg_num) +{ + unsigned long val; + + (void) sun4v_m7_get_perfreg(reg_num, &val); + + return val; +} + +static void m7_pcr_write(unsigned long reg_num, u64 val) +{ + (void) sun4v_m7_set_perfreg(reg_num, val); +} + +static const struct pcr_ops m7_pcr_ops = { + .read_pcr = m7_pcr_read, + .write_pcr = m7_pcr_write, + .read_pic = n4_pic_read, + .write_pic = n4_pic_write, + .nmi_picl_value = n4_picl_value, + .pcr_nmi_enable = (PCR_N4_PICNPT | PCR_N4_STRACE | + PCR_N4_UTRACE | PCR_N4_TOE | + (26 << PCR_N4_SL_SHIFT)), + .pcr_nmi_disable = PCR_N4_PICNPT, +}; static unsigned long perf_hsvc_group; static unsigned long perf_hsvc_major; @@ -248,6 +273,10 @@ static int __init register_perf_hsvc(void) perf_hsvc_group = HV_GRP_T5_CPU; break; + case SUN4V_CHIP_SPARC_M7: + perf_hsvc_group = HV_GRP_M7_PERF; + break; + default: return -ENODEV; } @@ -293,6 +322,10 @@ static int __init setup_sun4v_pcr_ops(void) pcr_ops = &n5_pcr_ops; break; + case SUN4V_CHIP_SPARC_M7: + pcr_ops = &m7_pcr_ops; + break; + default: ret = -ENODEV; break; diff --git a/arch/sparc/kernel/perf_event.c b/arch/sparc/kernel/perf_event.c index af53c25da2e7..86eebfa3b158 100644 --- a/arch/sparc/kernel/perf_event.c +++ b/arch/sparc/kernel/perf_event.c @@ -792,6 +792,42 @@ static const struct sparc_pmu niagara4_pmu = { .num_pic_regs = 4, }; +static void sparc_m7_write_pmc(int idx, u64 val) +{ + u64 pcr; + + pcr = pcr_ops->read_pcr(idx); + /* ensure ov and ntc are reset */ + pcr &= ~(PCR_N4_OV | PCR_N4_NTC); + + pcr_ops->write_pic(idx, val & 0xffffffff); + + pcr_ops->write_pcr(idx, pcr); +} + +static const struct sparc_pmu sparc_m7_pmu = { + .event_map = niagara4_event_map, + .cache_map = &niagara4_cache_map, + .max_events = ARRAY_SIZE(niagara4_perfmon_event_map), + .read_pmc = sparc_vt_read_pmc, + .write_pmc = sparc_m7_write_pmc, + .upper_shift = 5, + .lower_shift = 5, + .event_mask = 0x7ff, + .user_bit = PCR_N4_UTRACE, + .priv_bit = PCR_N4_STRACE, + + /* We explicitly don't support hypervisor tracing. */ + .hv_bit = 0, + + .irq_bit = PCR_N4_TOE, + .upper_nop = 0, + .lower_nop = 0, + .flags = 0, + .max_hw_events = 4, + .num_pcrs = 4, + .num_pic_regs = 4, +}; static const struct sparc_pmu *sparc_pmu __read_mostly; static u64 event_encoding(u64 event_id, int idx) @@ -1658,6 +1694,10 @@ static bool __init supported_pmu(void) sparc_pmu = &niagara4_pmu; return true; } + if (!strcmp(sparc_pmu_type, "sparc-m7")) { + sparc_pmu = &sparc_m7_pmu; + return true; + } return false; } -- cgit v1.2.1 From 31aaa98c248da766ece922bbbe8cc78cfd0bc920 Mon Sep 17 00:00:00 2001 From: David Ahern Date: Thu, 19 Mar 2015 16:06:53 -0400 Subject: sparc: Touch NMI watchdog when walking cpus and calling printk With the increase in number of CPUs calls to functions that dump output to console (e.g., arch_trigger_all_cpu_backtrace) can take a long time to complete. If IRQs are disabled eventually the NMI watchdog kicks in and creates more havoc. Avoid by telling the NMI watchdog everything is ok. Signed-off-by: David Ahern Signed-off-by: David S. Miller --- arch/sparc/kernel/process_64.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/arch/sparc/kernel/process_64.c b/arch/sparc/kernel/process_64.c index 0be7bf978cb1..46a59643bb1c 100644 --- a/arch/sparc/kernel/process_64.c +++ b/arch/sparc/kernel/process_64.c @@ -287,6 +287,8 @@ void arch_trigger_all_cpu_backtrace(bool include_self) printk(" TPC[%lx] O7[%lx] I7[%lx] RPC[%lx]\n", gp->tpc, gp->o7, gp->i7, gp->rpc); } + + touch_nmi_watchdog(); } memset(global_cpu_snapshot, 0, sizeof(global_cpu_snapshot)); @@ -362,6 +364,8 @@ static void pmu_snapshot_all_cpus(void) (cpu == this_cpu ? '*' : ' '), cpu, pp->pcr[0], pp->pcr[1], pp->pcr[2], pp->pcr[3], pp->pic[0], pp->pic[1], pp->pic[2], pp->pic[3]); + + touch_nmi_watchdog(); } memset(global_cpu_snapshot, 0, sizeof(global_cpu_snapshot)); -- cgit v1.2.1 From 755563bc79c764c90b9f44db5e4fe6c556d3440c Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Thu, 19 Mar 2015 19:29:01 +1100 Subject: powerpc/powernv: Fixes for hypervisor doorbell handling Since we can now use hypervisor doorbells for host IPIs, this makes sure we clear the host IPI flag when taking a doorbell interrupt, and clears any pending doorbell IPI in pnv_smp_cpu_kill_self() (as we already do for IPIs sent via the XICS interrupt controller). Otherwise if there did happen to be a leftover pending doorbell interrupt for an offline CPU thread for any reason, it would prevent that thread from going into a power-saving mode; it would instead keep waking up because of the interrupt. Signed-off-by: Paul Mackerras Signed-off-by: Michael Ellerman --- arch/powerpc/include/asm/ppc-opcode.h | 3 +++ arch/powerpc/include/asm/reg.h | 3 +++ arch/powerpc/kernel/dbell.c | 2 ++ arch/powerpc/platforms/powernv/smp.c | 14 ++++++++++++-- 4 files changed, 20 insertions(+), 2 deletions(-) diff --git a/arch/powerpc/include/asm/ppc-opcode.h b/arch/powerpc/include/asm/ppc-opcode.h index 03cd858a401c..4cbe23af400a 100644 --- a/arch/powerpc/include/asm/ppc-opcode.h +++ b/arch/powerpc/include/asm/ppc-opcode.h @@ -153,6 +153,7 @@ #define PPC_INST_MFSPR_PVR_MASK 0xfc1fffff #define PPC_INST_MFTMR 0x7c0002dc #define PPC_INST_MSGSND 0x7c00019c +#define PPC_INST_MSGCLR 0x7c0001dc #define PPC_INST_MSGSNDP 0x7c00011c #define PPC_INST_MTTMR 0x7c0003dc #define PPC_INST_NOP 0x60000000 @@ -309,6 +310,8 @@ ___PPC_RB(b) | __PPC_EH(eh)) #define PPC_MSGSND(b) stringify_in_c(.long PPC_INST_MSGSND | \ ___PPC_RB(b)) +#define PPC_MSGCLR(b) stringify_in_c(.long PPC_INST_MSGCLR | \ + ___PPC_RB(b)) #define PPC_MSGSNDP(b) stringify_in_c(.long PPC_INST_MSGSNDP | \ ___PPC_RB(b)) #define PPC_POPCNTB(a, s) stringify_in_c(.long PPC_INST_POPCNTB | \ diff --git a/arch/powerpc/include/asm/reg.h b/arch/powerpc/include/asm/reg.h index 1c874fb533bb..af56b5c6c81a 100644 --- a/arch/powerpc/include/asm/reg.h +++ b/arch/powerpc/include/asm/reg.h @@ -608,13 +608,16 @@ #define SRR1_ISI_N_OR_G 0x10000000 /* ISI: Access is no-exec or G */ #define SRR1_ISI_PROT 0x08000000 /* ISI: Other protection fault */ #define SRR1_WAKEMASK 0x00380000 /* reason for wakeup */ +#define SRR1_WAKEMASK_P8 0x003c0000 /* reason for wakeup on POWER8 */ #define SRR1_WAKESYSERR 0x00300000 /* System error */ #define SRR1_WAKEEE 0x00200000 /* External interrupt */ #define SRR1_WAKEMT 0x00280000 /* mtctrl */ #define SRR1_WAKEHMI 0x00280000 /* Hypervisor maintenance */ #define SRR1_WAKEDEC 0x00180000 /* Decrementer interrupt */ +#define SRR1_WAKEDBELL 0x00140000 /* Privileged doorbell on P8 */ #define SRR1_WAKETHERM 0x00100000 /* Thermal management interrupt */ #define SRR1_WAKERESET 0x00100000 /* System reset */ +#define SRR1_WAKEHDBELL 0x000c0000 /* Hypervisor doorbell on P8 */ #define SRR1_WAKESTATE 0x00030000 /* Powersave exit mask [46:47] */ #define SRR1_WS_DEEPEST 0x00030000 /* Some resources not maintained, * may not be recoverable */ diff --git a/arch/powerpc/kernel/dbell.c b/arch/powerpc/kernel/dbell.c index f4217819cc31..2128f3a96c32 100644 --- a/arch/powerpc/kernel/dbell.c +++ b/arch/powerpc/kernel/dbell.c @@ -17,6 +17,7 @@ #include #include +#include #ifdef CONFIG_SMP void doorbell_setup_this_cpu(void) @@ -41,6 +42,7 @@ void doorbell_exception(struct pt_regs *regs) may_hard_irq_enable(); + kvmppc_set_host_ipi(smp_processor_id(), 0); __this_cpu_inc(irq_stat.doorbell_irqs); smp_ipi_demux(); diff --git a/arch/powerpc/platforms/powernv/smp.c b/arch/powerpc/platforms/powernv/smp.c index fc34025ef822..38a45088f633 100644 --- a/arch/powerpc/platforms/powernv/smp.c +++ b/arch/powerpc/platforms/powernv/smp.c @@ -33,6 +33,8 @@ #include #include #include +#include +#include #include "powernv.h" @@ -149,7 +151,7 @@ static int pnv_smp_cpu_disable(void) static void pnv_smp_cpu_kill_self(void) { unsigned int cpu; - unsigned long srr1; + unsigned long srr1, wmask; u32 idle_states; /* Standard hot unplug procedure */ @@ -161,6 +163,10 @@ static void pnv_smp_cpu_kill_self(void) generic_set_cpu_dead(cpu); smp_wmb(); + wmask = SRR1_WAKEMASK; + if (cpu_has_feature(CPU_FTR_ARCH_207S)) + wmask = SRR1_WAKEMASK_P8; + idle_states = pnv_get_supported_cpuidle_states(); /* We don't want to take decrementer interrupts while we are offline, * so clear LPCR:PECE1. We keep PECE2 enabled. @@ -191,10 +197,14 @@ static void pnv_smp_cpu_kill_self(void) * having finished executing in a KVM guest, then srr1 * contains 0. */ - if ((srr1 & SRR1_WAKEMASK) == SRR1_WAKEEE) { + if ((srr1 & wmask) == SRR1_WAKEEE) { icp_native_flush_interrupt(); local_paca->irq_happened &= PACA_IRQ_HARD_DIS; smp_mb(); + } else if ((srr1 & wmask) == SRR1_WAKEHDBELL) { + unsigned long msg = PPC_DBELL_TYPE(PPC_DBELL_SERVER); + asm volatile(PPC_MSGCLR(%0) : : "r" (msg)); + kvmppc_set_host_ipi(cpu, 0); } if (cpu_core_split_required()) -- cgit v1.2.1 From ddee09c099c35074e50aaf9157efd22429d3acdf Mon Sep 17 00:00:00 2001 From: Benjamin Herrenschmidt Date: Thu, 19 Mar 2015 14:12:06 +1100 Subject: powerpc: Add PVR for POWER8NVL processor There's a new variant of POWER8 coming called "POWER8 with NVLink". The core is identical to POWER8 but unfortunately they strapped it with a different PVR, so we need to add an explicit entry for it. Signed-off-by: Benjamin Herrenschmidt Signed-off-by: Michael Ellerman --- arch/powerpc/kernel/cputable.c | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+) diff --git a/arch/powerpc/kernel/cputable.c b/arch/powerpc/kernel/cputable.c index f337666768a7..f83046878336 100644 --- a/arch/powerpc/kernel/cputable.c +++ b/arch/powerpc/kernel/cputable.c @@ -437,6 +437,26 @@ static struct cpu_spec __initdata cpu_specs[] = { .machine_check_early = __machine_check_early_realmode_p8, .platform = "power8", }, + { /* Power8NVL */ + .pvr_mask = 0xffff0000, + .pvr_value = 0x004c0000, + .cpu_name = "POWER8NVL (raw)", + .cpu_features = CPU_FTRS_POWER8, + .cpu_user_features = COMMON_USER_POWER8, + .cpu_user_features2 = COMMON_USER2_POWER8, + .mmu_features = MMU_FTRS_POWER8, + .icache_bsize = 128, + .dcache_bsize = 128, + .num_pmcs = 6, + .pmc_type = PPC_PMC_IBM, + .oprofile_cpu_type = "ppc64/power8", + .oprofile_type = PPC_OPROFILE_INVALID, + .cpu_setup = __setup_cpu_power8, + .cpu_restore = __restore_cpu_power8, + .flush_tlb = __flush_tlb_power8, + .machine_check_early = __machine_check_early_realmode_p8, + .platform = "power8", + }, { /* Power8 DD1: Does not support doorbell IPIs */ .pvr_mask = 0xffffff00, .pvr_value = 0x004d0100, -- cgit v1.2.1 From f6ff04149637723261aa4738958b0098b929ee9e Mon Sep 17 00:00:00 2001 From: Tyrel Datwyler Date: Wed, 4 Mar 2015 11:59:33 -0800 Subject: powerpc/pseries: Little endian fixes for post mobility device tree update We currently use the device tree update code in the kernel after resuming from a suspend operation to re-sync the kernels view of the device tree with that of the hypervisor. The code as it stands is not endian safe as it relies on parsing buffers returned by RTAS calls that thusly contains data in big endian format. This patch annotates variables and structure members with __be types as well as performing necessary byte swaps to cpu endian for data that needs to be parsed. Signed-off-by: Tyrel Datwyler Cc: Nathan Fontenot Cc: Cyril Bur Cc: stable@vger.kernel.org Signed-off-by: Michael Ellerman --- arch/powerpc/platforms/pseries/mobility.c | 44 ++++++++++++++++--------------- 1 file changed, 23 insertions(+), 21 deletions(-) diff --git a/arch/powerpc/platforms/pseries/mobility.c b/arch/powerpc/platforms/pseries/mobility.c index 90cf3dcbd9f2..8f35d525cede 100644 --- a/arch/powerpc/platforms/pseries/mobility.c +++ b/arch/powerpc/platforms/pseries/mobility.c @@ -25,10 +25,10 @@ static struct kobject *mobility_kobj; struct update_props_workarea { - u32 phandle; - u32 state; - u64 reserved; - u32 nprops; + __be32 phandle; + __be32 state; + __be64 reserved; + __be32 nprops; } __packed; #define NODE_ACTION_MASK 0xff000000 @@ -54,11 +54,11 @@ static int mobility_rtas_call(int token, char *buf, s32 scope) return rc; } -static int delete_dt_node(u32 phandle) +static int delete_dt_node(__be32 phandle) { struct device_node *dn; - dn = of_find_node_by_phandle(phandle); + dn = of_find_node_by_phandle(be32_to_cpu(phandle)); if (!dn) return -ENOENT; @@ -127,7 +127,7 @@ static int update_dt_property(struct device_node *dn, struct property **prop, return 0; } -static int update_dt_node(u32 phandle, s32 scope) +static int update_dt_node(__be32 phandle, s32 scope) { struct update_props_workarea *upwa; struct device_node *dn; @@ -136,6 +136,7 @@ static int update_dt_node(u32 phandle, s32 scope) char *prop_data; char *rtas_buf; int update_properties_token; + u32 nprops; u32 vd; update_properties_token = rtas_token("ibm,update-properties"); @@ -146,7 +147,7 @@ static int update_dt_node(u32 phandle, s32 scope) if (!rtas_buf) return -ENOMEM; - dn = of_find_node_by_phandle(phandle); + dn = of_find_node_by_phandle(be32_to_cpu(phandle)); if (!dn) { kfree(rtas_buf); return -ENOENT; @@ -162,6 +163,7 @@ static int update_dt_node(u32 phandle, s32 scope) break; prop_data = rtas_buf + sizeof(*upwa); + nprops = be32_to_cpu(upwa->nprops); /* On the first call to ibm,update-properties for a node the * the first property value descriptor contains an empty @@ -170,17 +172,17 @@ static int update_dt_node(u32 phandle, s32 scope) */ if (*prop_data == 0) { prop_data++; - vd = *(u32 *)prop_data; + vd = be32_to_cpu(*(__be32 *)prop_data); prop_data += vd + sizeof(vd); - upwa->nprops--; + nprops--; } - for (i = 0; i < upwa->nprops; i++) { + for (i = 0; i < nprops; i++) { char *prop_name; prop_name = prop_data; prop_data += strlen(prop_name) + 1; - vd = *(u32 *)prop_data; + vd = be32_to_cpu(*(__be32 *)prop_data); prop_data += sizeof(vd); switch (vd) { @@ -212,13 +214,13 @@ static int update_dt_node(u32 phandle, s32 scope) return 0; } -static int add_dt_node(u32 parent_phandle, u32 drc_index) +static int add_dt_node(__be32 parent_phandle, __be32 drc_index) { struct device_node *dn; struct device_node *parent_dn; int rc; - parent_dn = of_find_node_by_phandle(parent_phandle); + parent_dn = of_find_node_by_phandle(be32_to_cpu(parent_phandle)); if (!parent_dn) return -ENOENT; @@ -237,7 +239,7 @@ static int add_dt_node(u32 parent_phandle, u32 drc_index) int pseries_devicetree_update(s32 scope) { char *rtas_buf; - u32 *data; + __be32 *data; int update_nodes_token; int rc; @@ -254,17 +256,17 @@ int pseries_devicetree_update(s32 scope) if (rc && rc != 1) break; - data = (u32 *)rtas_buf + 4; - while (*data & NODE_ACTION_MASK) { + data = (__be32 *)rtas_buf + 4; + while (be32_to_cpu(*data) & NODE_ACTION_MASK) { int i; - u32 action = *data & NODE_ACTION_MASK; - int node_count = *data & NODE_COUNT_MASK; + u32 action = be32_to_cpu(*data) & NODE_ACTION_MASK; + u32 node_count = be32_to_cpu(*data) & NODE_COUNT_MASK; data++; for (i = 0; i < node_count; i++) { - u32 phandle = *data++; - u32 drc_index; + __be32 phandle = *data++; + __be32 drc_index; switch (action) { case DELETE_DT_NODE: -- cgit v1.2.1 From 4b36b68ca8bab6edf7e99f859c42a91f3ad1846e Mon Sep 17 00:00:00 2001 From: Nicholas Bellinger Date: Sun, 22 Feb 2015 22:17:13 -0800 Subject: target: Disallow changing of WRITE cache/FUA attrs after export Now that incoming FUA=1 bit check is enforced for backends with FUA or WCE disabled, go ahead and disallow the changing of related backend attributes when active fabric exports exist. This is required to avoid potential failures with existing initiator LUN registrations that have been previously created with FUA=1. Reported-by: Christoph Hellwig Cc: Doug Gilbert Cc: James Bottomley Cc: Ronnie Sahlberg Signed-off-by: Nicholas Bellinger --- drivers/target/target_core_device.c | 16 +++++++++++++++- 1 file changed, 15 insertions(+), 1 deletion(-) diff --git a/drivers/target/target_core_device.c b/drivers/target/target_core_device.c index 58f49ff69b14..37449bdd62f6 100644 --- a/drivers/target/target_core_device.c +++ b/drivers/target/target_core_device.c @@ -767,6 +767,16 @@ int se_dev_set_emulate_fua_write(struct se_device *dev, int flag) pr_err("Illegal value %d\n", flag); return -EINVAL; } + if (flag && + dev->transport->get_write_cache) { + pr_err("emulate_fua_write not supported for this device\n"); + return -EINVAL; + } + if (dev->export_count) { + pr_err("emulate_fua_write cannot be changed with active" + " exports: %d\n", dev->export_count); + return -EINVAL; + } dev->dev_attrib.emulate_fua_write = flag; pr_debug("dev[%p]: SE Device Forced Unit Access WRITEs: %d\n", dev, dev->dev_attrib.emulate_fua_write); @@ -801,7 +811,11 @@ int se_dev_set_emulate_write_cache(struct se_device *dev, int flag) pr_err("emulate_write_cache not supported for this device\n"); return -EINVAL; } - + if (dev->export_count) { + pr_err("emulate_write_cache cannot be changed with active" + " exports: %d\n", dev->export_count); + return -EINVAL; + } dev->dev_attrib.emulate_write_cache = flag; pr_debug("dev[%p]: SE Device WRITE_CACHE_EMULATION flag: %d\n", dev, dev->dev_attrib.emulate_write_cache); -- cgit v1.2.1 From 2a03ee8c5677acd7ab61b0be7b880b3d3de1bcf2 Mon Sep 17 00:00:00 2001 From: Nicholas Bellinger Date: Wed, 25 Feb 2015 22:56:37 -0800 Subject: Revert "iscsi-target: Avoid IN_LOGOUT failure case for iser-target" This reverts commit 72859d91d93319c00a18c29f577e56bf73a8654a. The original patch was wrong, iscsit_close_connection() still needs to release iscsi_conn during both normal + exception IN_LOGOUT status with ib_isert enabled. The original OOPs is due to completing conn_logout_comp early within iscsit_close_connection(), causing isert_wait4logout() to complete instead of waiting for iscsit_logout_post_handler_*() to be called. Reported-by: Sagi Grimberg Cc: Sagi Grimberg Cc: Slava Shwartsman Signed-off-by: Nicholas Bellinger --- drivers/target/iscsi/iscsi_target_erl0.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/drivers/target/iscsi/iscsi_target_erl0.c b/drivers/target/iscsi/iscsi_target_erl0.c index 1c197bad6132..bdd8731a4daa 100644 --- a/drivers/target/iscsi/iscsi_target_erl0.c +++ b/drivers/target/iscsi/iscsi_target_erl0.c @@ -22,7 +22,6 @@ #include #include -#include #include "iscsi_target_seq_pdu_list.h" #include "iscsi_target_tq.h" #include "iscsi_target_erl0.h" @@ -940,8 +939,7 @@ void iscsit_take_action_for_connection_exit(struct iscsi_conn *conn) if (conn->conn_state == TARG_CONN_STATE_IN_LOGOUT) { spin_unlock_bh(&conn->state_lock); - if (conn->conn_transport->transport_type == ISCSI_TCP) - iscsit_close_connection(conn); + iscsit_close_connection(conn); return; } -- cgit v1.2.1 From f068fbc82e7696d67b1bb8189306865bedf368b6 Mon Sep 17 00:00:00 2001 From: Nicholas Bellinger Date: Mon, 23 Feb 2015 00:57:51 -0800 Subject: iscsi-target: Avoid early conn_logout_comp for iser connections This patch fixes a iser specific logout bug where early complete() of conn->conn_logout_comp in iscsit_close_connection() was causing isert_wait4logout() to complete too soon, triggering a use after free NULL pointer dereference of iscsi_conn memory. The complete() was originally added for traditional iscsi-target when a ISCSI_LOGOUT_OP failed in iscsi_target_rx_opcode(), but given iser-target does not wait in logout failure, this special case needs to be avoided. Reported-by: Sagi Grimberg Cc: Sagi Grimberg Cc: Slava Shwartsman Cc: # v3.10+ Signed-off-by: Nicholas Bellinger --- drivers/target/iscsi/iscsi_target.c | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 deletions(-) diff --git a/drivers/target/iscsi/iscsi_target.c b/drivers/target/iscsi/iscsi_target.c index 50bad55a0c42..2accb6e47beb 100644 --- a/drivers/target/iscsi/iscsi_target.c +++ b/drivers/target/iscsi/iscsi_target.c @@ -4256,11 +4256,17 @@ int iscsit_close_connection( pr_debug("Closing iSCSI connection CID %hu on SID:" " %u\n", conn->cid, sess->sid); /* - * Always up conn_logout_comp just in case the RX Thread is sleeping - * and the logout response never got sent because the connection - * failed. + * Always up conn_logout_comp for the traditional TCP case just in case + * the RX Thread in iscsi_target_rx_opcode() is sleeping and the logout + * response never got sent because the connection failed. + * + * However for iser-target, isert_wait4logout() is using conn_logout_comp + * to signal logout response TX interrupt completion. Go ahead and skip + * this for iser since isert_rx_opcode() does not wait on logout failure, + * and to avoid iscsi_conn pointer dereference in iser-target code. */ - complete(&conn->conn_logout_comp); + if (conn->conn_transport->transport_type == ISCSI_TCP) + complete(&conn->conn_logout_comp); iscsi_release_thread_set(conn); -- cgit v1.2.1 From 75c3d0bf9caebb502e96683b2bc37f9692437e68 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Thu, 19 Mar 2015 22:25:16 -0700 Subject: tcm_qla2xxx: Fix incorrect use of __transport_register_session This patch fixes the incorrect use of __transport_register_session() in tcm_qla2xxx_check_initiator_node_acl() code, that does not perform explicit se_tpg->session_lock when accessing se_tpg->tpg_sess_list to add new se_sess nodes. Given that tcm_qla2xxx_check_initiator_node_acl() is not called with qla_hw->hardware_lock held for all accesses of ->tpg_sess_list, the code should be using transport_register_session() instead. Signed-off-by: Bart Van Assche Cc: Giridhar Malavali Cc: Quinn Tran Cc: # 3.5+ Signed-off-by: Nicholas Bellinger --- drivers/scsi/qla2xxx/tcm_qla2xxx.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/scsi/qla2xxx/tcm_qla2xxx.c b/drivers/scsi/qla2xxx/tcm_qla2xxx.c index 99f43b7fc9ab..ab4879e12ea7 100644 --- a/drivers/scsi/qla2xxx/tcm_qla2xxx.c +++ b/drivers/scsi/qla2xxx/tcm_qla2xxx.c @@ -1596,7 +1596,7 @@ static int tcm_qla2xxx_check_initiator_node_acl( /* * Finally register the new FC Nexus with TCM */ - __transport_register_session(se_nacl->se_tpg, se_nacl, se_sess, sess); + transport_register_session(se_nacl->se_tpg, se_nacl, se_sess, sess); return 0; } -- cgit v1.2.1 From 2f450cc1fbe9713f79b217e61ab204e263723ead Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Thu, 12 Feb 2015 11:48:49 +0100 Subject: loop/usb/vhost-scsi/xen-scsiback: Fix use of __transport_register_session This patch changes loopback, usb-gadget, vhost-scsi and xen-scsiback fabric code to invoke transport_register_session() instead of the unprotected flavour, to ensure se_tpg->session_lock is taken when adding new session list nodes to se_tpg->tpg_sess_list. Note that since these four fabric drivers already hold their own internal TPG mutexes when accessing se_tpg->tpg_sess_list, and consist of a single se_session created through configfs attribute access, no list corruption can currently occur. So for correctness sake, go ahead and use the se_tpg->session_lock protected version for these four fabric drivers. Signed-off-by: Bart Van Assche Signed-off-by: Nicholas Bellinger --- drivers/target/loopback/tcm_loop.c | 7 ++----- drivers/usb/gadget/legacy/tcm_usb_gadget.c | 5 ++--- drivers/vhost/scsi.c | 5 ++--- drivers/xen/xen-scsiback.c | 7 ++----- 4 files changed, 8 insertions(+), 16 deletions(-) diff --git a/drivers/target/loopback/tcm_loop.c b/drivers/target/loopback/tcm_loop.c index 6b3c32954689..c36bd7c29136 100644 --- a/drivers/target/loopback/tcm_loop.c +++ b/drivers/target/loopback/tcm_loop.c @@ -953,11 +953,8 @@ static int tcm_loop_make_nexus( transport_free_session(tl_nexus->se_sess); goto out; } - /* - * Now, register the SAS I_T Nexus as active with the call to - * transport_register_session() - */ - __transport_register_session(se_tpg, tl_nexus->se_sess->se_node_acl, + /* Now, register the SAS I_T Nexus as active. */ + transport_register_session(se_tpg, tl_nexus->se_sess->se_node_acl, tl_nexus->se_sess, tl_nexus); tl_tpg->tl_nexus = tl_nexus; pr_debug("TCM_Loop_ConfigFS: Established I_T Nexus to emulated" diff --git a/drivers/usb/gadget/legacy/tcm_usb_gadget.c b/drivers/usb/gadget/legacy/tcm_usb_gadget.c index 3a494168661e..6e0a019aad54 100644 --- a/drivers/usb/gadget/legacy/tcm_usb_gadget.c +++ b/drivers/usb/gadget/legacy/tcm_usb_gadget.c @@ -1740,10 +1740,9 @@ static int tcm_usbg_make_nexus(struct usbg_tpg *tpg, char *name) goto err_session; } /* - * Now register the TCM vHost virtual I_T Nexus as active with the - * call to __transport_register_session() + * Now register the TCM vHost virtual I_T Nexus as active. */ - __transport_register_session(se_tpg, tv_nexus->tvn_se_sess->se_node_acl, + transport_register_session(se_tpg, tv_nexus->tvn_se_sess->se_node_acl, tv_nexus->tvn_se_sess, tv_nexus); tpg->tpg_nexus = tv_nexus; mutex_unlock(&tpg->tpg_mutex); diff --git a/drivers/vhost/scsi.c b/drivers/vhost/scsi.c index 8d4f3f1ff799..71df240a467a 100644 --- a/drivers/vhost/scsi.c +++ b/drivers/vhost/scsi.c @@ -1956,10 +1956,9 @@ static int vhost_scsi_make_nexus(struct vhost_scsi_tpg *tpg, goto out; } /* - * Now register the TCM vhost virtual I_T Nexus as active with the - * call to __transport_register_session() + * Now register the TCM vhost virtual I_T Nexus as active. */ - __transport_register_session(se_tpg, tv_nexus->tvn_se_sess->se_node_acl, + transport_register_session(se_tpg, tv_nexus->tvn_se_sess->se_node_acl, tv_nexus->tvn_se_sess, tv_nexus); tpg->tpg_nexus = tv_nexus; diff --git a/drivers/xen/xen-scsiback.c b/drivers/xen/xen-scsiback.c index 61653a03a8f5..8511277bf09c 100644 --- a/drivers/xen/xen-scsiback.c +++ b/drivers/xen/xen-scsiback.c @@ -1661,11 +1661,8 @@ static int scsiback_make_nexus(struct scsiback_tpg *tpg, name); goto out; } - /* - * Now register the TCM pvscsi virtual I_T Nexus as active with the - * call to __transport_register_session() - */ - __transport_register_session(se_tpg, tv_nexus->tvn_se_sess->se_node_acl, + /* Now register the TCM pvscsi virtual I_T Nexus as active. */ + transport_register_session(se_tpg, tv_nexus->tvn_se_sess->se_node_acl, tv_nexus->tvn_se_sess, tv_nexus); tpg->tpg_nexus = tv_nexus; -- cgit v1.2.1 From 7544e597343e2166daba3f32e4708533aa53c233 Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Wed, 18 Feb 2015 15:33:58 +0100 Subject: target: Fix reference leak in target_get_sess_cmd() error path This patch fixes a se_cmd->cmd_kref leak buf when se_sess->sess_tearing_down is true within target_get_sess_cmd() submission path code. This se_cmd reference leak can occur during active session shutdown when ack_kref=1 is passed by target_submit_cmd_[map_sgls,tmr]() callers. Signed-off-by: Bart Van Assche Cc: # 3.6+ Signed-off-by: Nicholas Bellinger --- drivers/target/target_core_transport.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c index 0adc0f650213..ac3cbabdbdf0 100644 --- a/drivers/target/target_core_transport.c +++ b/drivers/target/target_core_transport.c @@ -2389,6 +2389,10 @@ int target_get_sess_cmd(struct se_session *se_sess, struct se_cmd *se_cmd, list_add_tail(&se_cmd->se_cmd_list, &se_sess->sess_cmd_list); out: spin_unlock_irqrestore(&se_sess->sess_cmd_lock, flags); + + if (ret && ack_kref) + target_put_sess_cmd(se_sess, se_cmd); + return ret; } EXPORT_SYMBOL(target_get_sess_cmd); -- cgit v1.2.1 From d556546e7ecd9fca199df4698943024d40044f8e Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Wed, 25 Feb 2015 16:21:03 +0300 Subject: tcm_fc: missing curly braces in ft_invl_hw_context() This patch adds a missing set of conditional check braces in ft_invl_hw_context() originally introduced by commit dcd998ccd when handling DDP failures in ft_recv_write_data() code. commit dcd998ccdbf74a7d8fe0f0a44e85da1ed5975946 Author: Kiran Patil Date: Wed Aug 3 09:20:01 2011 +0000 tcm_fc: Handle DDP/SW fc_frame_payload_get failures in ft_recv_write_data Signed-off-by: Dan Carpenter Cc: Kiran Patil Cc: # 3.1+ Signed-off-by: Nicholas Bellinger --- drivers/target/tcm_fc/tfc_io.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/target/tcm_fc/tfc_io.c b/drivers/target/tcm_fc/tfc_io.c index 97b486c3dda1..583e755d8091 100644 --- a/drivers/target/tcm_fc/tfc_io.c +++ b/drivers/target/tcm_fc/tfc_io.c @@ -359,7 +359,7 @@ void ft_invl_hw_context(struct ft_cmd *cmd) ep = fc_seq_exch(seq); if (ep) { lport = ep->lp; - if (lport && (ep->xid <= lport->lro_xid)) + if (lport && (ep->xid <= lport->lro_xid)) { /* * "ddp_done" trigger invalidation of HW * specific DDP context @@ -374,6 +374,7 @@ void ft_invl_hw_context(struct ft_cmd *cmd) * identified using ep->xid) */ cmd->was_ddp_setup = 0; + } } } } -- cgit v1.2.1 From 215a8fe4198f607f34ecdbc9969dae783d8b5a61 Mon Sep 17 00:00:00 2001 From: Nicholas Bellinger Date: Fri, 27 Feb 2015 03:54:13 -0800 Subject: target/pscsi: Fix NULL pointer dereference in get_device_type This patch fixes a NULL pointer dereference OOPs with pSCSI backends within target_core_stat.c code. The bug is caused by a configfs attr read if no pscsi_dev_virt->pdv_sd has been configured. Reported-by: Olaf Hering Cc: Signed-off-by: Nicholas Bellinger --- drivers/target/target_core_pscsi.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/target/target_core_pscsi.c b/drivers/target/target_core_pscsi.c index 1045dcd7bf65..f6c954c4635f 100644 --- a/drivers/target/target_core_pscsi.c +++ b/drivers/target/target_core_pscsi.c @@ -1121,7 +1121,7 @@ static u32 pscsi_get_device_type(struct se_device *dev) struct pscsi_dev_virt *pdv = PSCSI_DEV(dev); struct scsi_device *sd = pdv->pdv_sd; - return sd->type; + return (sd) ? sd->type : TYPE_NO_LUN; } static sector_t pscsi_get_blocks(struct se_device *dev) -- cgit v1.2.1 From 5f7da044f8bc1cfb21c962edf34bd5699a76e7ae Mon Sep 17 00:00:00 2001 From: Nicholas Bellinger Date: Thu, 5 Mar 2015 03:28:24 +0000 Subject: target: Fix virtual LUN=0 target_configure_device failure OOPs This patch fixes a NULL pointer dereference triggered by a late target_configure_device() -> alloc_workqueue() failure that results in target_free_device() being called with DF_CONFIGURED already set, which subsequently OOPses in destroy_workqueue() code. Currently this only happens at modprobe target_core_mod time when core_dev_setup_virtual_lun0() -> target_configure_device() fails, and the explicit target_free_device() gets called. To address this bug originally introduced by commit 0fd97ccf45, go ahead and move DF_CONFIGURED to end of target_configure_device() code to handle this special failure case. Reported-by: Claudio Fleiner Cc: Claudio Fleiner Cc: Christoph Hellwig Cc: # v3.7+ Signed-off-by: Nicholas Bellinger --- drivers/target/target_core_device.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/target/target_core_device.c b/drivers/target/target_core_device.c index 37449bdd62f6..7fc5eae875de 100644 --- a/drivers/target/target_core_device.c +++ b/drivers/target/target_core_device.c @@ -1548,8 +1548,6 @@ int target_configure_device(struct se_device *dev) ret = dev->transport->configure_device(dev); if (ret) goto out; - dev->dev_flags |= DF_CONFIGURED; - /* * XXX: there is not much point to have two different values here.. */ @@ -1611,6 +1609,8 @@ int target_configure_device(struct se_device *dev) list_add_tail(&dev->g_dev_node, &g_device_list); mutex_unlock(&g_device_mutex); + dev->dev_flags |= DF_CONFIGURED; + return 0; out_free_alua: -- cgit v1.2.1 From 9bc6548f372d8c829235095d91de99d8df79db6e Mon Sep 17 00:00:00 2001 From: Christophe Vu-Brugier Date: Thu, 19 Mar 2015 14:30:13 +0100 Subject: target: do not reject FUA CDBs when write cache is enabled but emulate_write_cache is 0 A check that rejects a CDB with FUA bit set if no write cache is emulated was added by the following commit: fde9f50 target: Add sanity checks for DPO/FUA bit usage The condition is as follows: if (!dev->dev_attrib.emulate_fua_write || !dev->dev_attrib.emulate_write_cache) However, this check is wrong if the backend device supports WCE but "emulate_write_cache" is disabled. This patch uses se_dev_check_wce() (previously named spc_check_dev_wce) to invoke transport->get_write_cache() if the device has a write cache or check the "emulate_write_cache" attribute otherwise. Reported-by: Christoph Hellwig Signed-off-by: Christophe Vu-Brugier Signed-off-by: Nicholas Bellinger --- drivers/target/target_core_device.c | 12 ++++++++++++ drivers/target/target_core_sbc.c | 3 +-- drivers/target/target_core_spc.c | 19 +++---------------- include/target/target_core_backend.h | 1 + 4 files changed, 17 insertions(+), 18 deletions(-) diff --git a/drivers/target/target_core_device.c b/drivers/target/target_core_device.c index 7fc5eae875de..79b4ec3ca2db 100644 --- a/drivers/target/target_core_device.c +++ b/drivers/target/target_core_device.c @@ -650,6 +650,18 @@ static u32 se_dev_align_max_sectors(u32 max_sectors, u32 block_size) return aligned_max_sectors; } +bool se_dev_check_wce(struct se_device *dev) +{ + bool wce = false; + + if (dev->transport->get_write_cache) + wce = dev->transport->get_write_cache(dev); + else if (dev->dev_attrib.emulate_write_cache > 0) + wce = true; + + return wce; +} + int se_dev_set_max_unmap_lba_count( struct se_device *dev, u32 max_unmap_lba_count) diff --git a/drivers/target/target_core_sbc.c b/drivers/target/target_core_sbc.c index 9a2f9d3a6e70..3e7297411110 100644 --- a/drivers/target/target_core_sbc.c +++ b/drivers/target/target_core_sbc.c @@ -708,8 +708,7 @@ sbc_check_dpofua(struct se_device *dev, struct se_cmd *cmd, unsigned char *cdb) } } if (cdb[1] & 0x8) { - if (!dev->dev_attrib.emulate_fua_write || - !dev->dev_attrib.emulate_write_cache) { + if (!dev->dev_attrib.emulate_fua_write || !se_dev_check_wce(dev)) { pr_err("Got CDB: 0x%02x with FUA bit set, but device" " does not advertise support for FUA write\n", cdb[0]); diff --git a/drivers/target/target_core_spc.c b/drivers/target/target_core_spc.c index 460e93109473..6c8bd6bc175c 100644 --- a/drivers/target/target_core_spc.c +++ b/drivers/target/target_core_spc.c @@ -454,19 +454,6 @@ check_scsi_name: } EXPORT_SYMBOL(spc_emulate_evpd_83); -static bool -spc_check_dev_wce(struct se_device *dev) -{ - bool wce = false; - - if (dev->transport->get_write_cache) - wce = dev->transport->get_write_cache(dev); - else if (dev->dev_attrib.emulate_write_cache > 0) - wce = true; - - return wce; -} - /* Extended INQUIRY Data VPD Page */ static sense_reason_t spc_emulate_evpd_86(struct se_cmd *cmd, unsigned char *buf) @@ -490,7 +477,7 @@ spc_emulate_evpd_86(struct se_cmd *cmd, unsigned char *buf) buf[5] = 0x07; /* If WriteCache emulation is enabled, set V_SUP */ - if (spc_check_dev_wce(dev)) + if (se_dev_check_wce(dev)) buf[6] = 0x01; /* If an LBA map is present set R_SUP */ spin_lock(&cmd->se_dev->t10_alua.lba_map_lock); @@ -897,7 +884,7 @@ static int spc_modesense_caching(struct se_cmd *cmd, u8 pc, u8 *p) if (pc == 1) goto out; - if (spc_check_dev_wce(dev)) + if (se_dev_check_wce(dev)) p[2] = 0x04; /* Write Cache Enable */ p[12] = 0x20; /* Disabled Read Ahead */ @@ -1009,7 +996,7 @@ static sense_reason_t spc_emulate_modesense(struct se_cmd *cmd) (cmd->se_deve->lun_flags & TRANSPORT_LUNFLAGS_READ_ONLY))) spc_modesense_write_protect(&buf[length], type); - if ((spc_check_dev_wce(dev)) && + if ((se_dev_check_wce(dev)) && (dev->dev_attrib.emulate_fua_write > 0)) spc_modesense_dpofua(&buf[length], type); diff --git a/include/target/target_core_backend.h b/include/target/target_core_backend.h index db81c65b8f48..d61be7297b2c 100644 --- a/include/target/target_core_backend.h +++ b/include/target/target_core_backend.h @@ -111,6 +111,7 @@ void array_free(void *array, int n); void target_core_setup_sub_cits(struct se_subsystem_api *); /* attribute helpers from target_core_device.c for backend drivers */ +bool se_dev_check_wce(struct se_device *); int se_dev_set_max_unmap_lba_count(struct se_device *, u32); int se_dev_set_max_unmap_block_desc_count(struct se_device *, u32); int se_dev_set_unmap_granularity(struct se_device *, u32); -- cgit v1.2.1 From 8f902b005ece690f0f50b217975601b804905dc8 Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Fri, 20 Mar 2015 20:39:38 +1100 Subject: KVM: PPC: Book3S HV: Fix spinlock/mutex ordering issue in kvmppc_set_lpcr() Currently, kvmppc_set_lpcr() has a spinlock around the whole function, and inside that does mutex_lock(&kvm->lock). It is not permitted to take a mutex while holding a spinlock, because the mutex_lock might call schedule(). In addition, this causes lockdep to warn about a lock ordering issue: ====================================================== [ INFO: possible circular locking dependency detected ] 3.18.0-kvm-04645-gdfea862-dirty #131 Not tainted ------------------------------------------------------- qemu-system-ppc/8179 is trying to acquire lock: (&kvm->lock){+.+.+.}, at: [] .kvmppc_set_lpcr+0xf4/0x1c0 [kvm_hv] but task is already holding lock: (&(&vcore->lock)->rlock){+.+...}, at: [] .kvmppc_set_lpcr+0x40/0x1c0 [kvm_hv] which lock already depends on the new lock. the existing dependency chain (in reverse order) is: -> #1 (&(&vcore->lock)->rlock){+.+...}: [] .mutex_lock_nested+0x80/0x570 [] .kvmppc_vcpu_run_hv+0xc4/0xe40 [kvm_hv] [] .kvmppc_vcpu_run+0x2c/0x40 [kvm] [] .kvm_arch_vcpu_ioctl_run+0x54/0x160 [kvm] [] .kvm_vcpu_ioctl+0x4a8/0x7b0 [kvm] [] .do_vfs_ioctl+0x444/0x770 [] .SyS_ioctl+0xc4/0xe0 [] syscall_exit+0x0/0x98 -> #0 (&kvm->lock){+.+.+.}: [] .lock_acquire+0xcc/0x1a0 [] .mutex_lock_nested+0x80/0x570 [] .kvmppc_set_lpcr+0xf4/0x1c0 [kvm_hv] [] .kvmppc_set_one_reg_hv+0x4dc/0x990 [kvm_hv] [] .kvmppc_set_one_reg+0x44/0x330 [kvm] [] .kvm_vcpu_ioctl_set_one_reg+0x5c/0x150 [kvm] [] .kvm_arch_vcpu_ioctl+0x214/0x2c0 [kvm] [] .kvm_vcpu_ioctl+0xe0/0x7b0 [kvm] [] .do_vfs_ioctl+0x444/0x770 [] .SyS_ioctl+0xc4/0xe0 [] syscall_exit+0x0/0x98 other info that might help us debug this: Possible unsafe locking scenario: CPU0 CPU1 ---- ---- lock(&(&vcore->lock)->rlock); lock(&kvm->lock); lock(&(&vcore->lock)->rlock); lock(&kvm->lock); *** DEADLOCK *** 2 locks held by qemu-system-ppc/8179: #0: (&vcpu->mutex){+.+.+.}, at: [] .vcpu_load+0x28/0x90 [kvm] #1: (&(&vcore->lock)->rlock){+.+...}, at: [] .kvmppc_set_lpcr+0x40/0x1c0 [kvm_hv] stack backtrace: CPU: 4 PID: 8179 Comm: qemu-system-ppc Not tainted 3.18.0-kvm-04645-gdfea862-dirty #131 Call Trace: [c000001a66c0f310] [c000000000b486ac] .dump_stack+0x88/0xb4 (unreliable) [c000001a66c0f390] [c0000000000f8bec] .print_circular_bug+0x27c/0x3d0 [c000001a66c0f440] [c0000000000fe9e8] .__lock_acquire+0x2028/0x2190 [c000001a66c0f5d0] [c0000000000ff28c] .lock_acquire+0xcc/0x1a0 [c000001a66c0f6a0] [c000000000b3c120] .mutex_lock_nested+0x80/0x570 [c000001a66c0f7c0] [d00000000ecc1f54] .kvmppc_set_lpcr+0xf4/0x1c0 [kvm_hv] [c000001a66c0f860] [d00000000ecc510c] .kvmppc_set_one_reg_hv+0x4dc/0x990 [kvm_hv] [c000001a66c0f8d0] [d00000000eb9f234] .kvmppc_set_one_reg+0x44/0x330 [kvm] [c000001a66c0f960] [d00000000eb9c9dc] .kvm_vcpu_ioctl_set_one_reg+0x5c/0x150 [kvm] [c000001a66c0f9f0] [d00000000eb9ced4] .kvm_arch_vcpu_ioctl+0x214/0x2c0 [kvm] [c000001a66c0faf0] [d00000000eb940b0] .kvm_vcpu_ioctl+0xe0/0x7b0 [kvm] [c000001a66c0fcb0] [c00000000026cbb4] .do_vfs_ioctl+0x444/0x770 [c000001a66c0fd90] [c00000000026cfa4] .SyS_ioctl+0xc4/0xe0 [c000001a66c0fe30] [c000000000009264] syscall_exit+0x0/0x98 This fixes it by moving the mutex_lock()/mutex_unlock() pair outside the spin-locked region. Signed-off-by: Paul Mackerras Signed-off-by: Alexander Graf --- arch/powerpc/kvm/book3s_hv.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/arch/powerpc/kvm/book3s_hv.c b/arch/powerpc/kvm/book3s_hv.c index de4018a1bc4b..b2731933f807 100644 --- a/arch/powerpc/kvm/book3s_hv.c +++ b/arch/powerpc/kvm/book3s_hv.c @@ -942,20 +942,20 @@ static int kvm_arch_vcpu_ioctl_set_sregs_hv(struct kvm_vcpu *vcpu, static void kvmppc_set_lpcr(struct kvm_vcpu *vcpu, u64 new_lpcr, bool preserve_top32) { + struct kvm *kvm = vcpu->kvm; struct kvmppc_vcore *vc = vcpu->arch.vcore; u64 mask; + mutex_lock(&kvm->lock); spin_lock(&vc->lock); /* * If ILE (interrupt little-endian) has changed, update the * MSR_LE bit in the intr_msr for each vcpu in this vcore. */ if ((new_lpcr & LPCR_ILE) != (vc->lpcr & LPCR_ILE)) { - struct kvm *kvm = vcpu->kvm; struct kvm_vcpu *vcpu; int i; - mutex_lock(&kvm->lock); kvm_for_each_vcpu(i, vcpu, kvm) { if (vcpu->arch.vcore != vc) continue; @@ -964,7 +964,6 @@ static void kvmppc_set_lpcr(struct kvm_vcpu *vcpu, u64 new_lpcr, else vcpu->arch.intr_msr &= ~MSR_LE; } - mutex_unlock(&kvm->lock); } /* @@ -981,6 +980,7 @@ static void kvmppc_set_lpcr(struct kvm_vcpu *vcpu, u64 new_lpcr, mask &= 0xFFFFFFFF; vc->lpcr = (vc->lpcr & ~mask) | (new_lpcr & mask); spin_unlock(&vc->lock); + mutex_unlock(&kvm->lock); } static int kvmppc_get_one_reg_hv(struct kvm_vcpu *vcpu, u64 id, -- cgit v1.2.1 From ecb6d6185b3ae40067330eb889977bf2a51f7429 Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Fri, 20 Mar 2015 20:39:39 +1100 Subject: KVM: PPC: Book3S HV: Endian fix for accessing VPA yield count The VPA (virtual processor area) is defined by PAPR and is therefore big-endian, so we need a be32_to_cpu when reading it in kvmppc_get_yield_count(). Without this, H_CONFER always fails on a little-endian host, causing SMP guests to waste time spinning on spinlocks. Signed-off-by: Paul Mackerras Signed-off-by: Alexander Graf --- arch/powerpc/kvm/book3s_hv.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/powerpc/kvm/book3s_hv.c b/arch/powerpc/kvm/book3s_hv.c index b2731933f807..de747563d29d 100644 --- a/arch/powerpc/kvm/book3s_hv.c +++ b/arch/powerpc/kvm/book3s_hv.c @@ -636,7 +636,7 @@ static int kvmppc_get_yield_count(struct kvm_vcpu *vcpu) spin_lock(&vcpu->arch.vpa_update_lock); lppaca = (struct lppaca *)vcpu->arch.vpa.pinned_addr; if (lppaca) - yield_count = lppaca->yield_count; + yield_count = be32_to_cpu(lppaca->yield_count); spin_unlock(&vcpu->arch.vpa_update_lock); return yield_count; } -- cgit v1.2.1 From 2bf27601c7b50b6ced72f27304109dc52eb52919 Mon Sep 17 00:00:00 2001 From: Paul Mackerras Date: Fri, 20 Mar 2015 20:39:40 +1100 Subject: KVM: PPC: Book3S HV: Fix instruction emulation Commit 4a157d61b48c ("KVM: PPC: Book3S HV: Fix endianness of instruction obtained from HEIR register") had the side effect that we no longer reset vcpu->arch.last_inst to -1 on guest exit in the cases where the instruction is not fetched from the guest. This means that if instruction emulation turns out to be required in those cases, the host will emulate the wrong instruction, since vcpu->arch.last_inst will contain the last instruction that was emulated. This fixes it by making sure that vcpu->arch.last_inst is reset to -1 in those cases. Signed-off-by: Paul Mackerras Signed-off-by: Alexander Graf --- arch/powerpc/kvm/book3s_hv_rmhandlers.S | 1 + 1 file changed, 1 insertion(+) diff --git a/arch/powerpc/kvm/book3s_hv_rmhandlers.S b/arch/powerpc/kvm/book3s_hv_rmhandlers.S index bb94e6f20c81..6cbf1630cb70 100644 --- a/arch/powerpc/kvm/book3s_hv_rmhandlers.S +++ b/arch/powerpc/kvm/book3s_hv_rmhandlers.S @@ -1005,6 +1005,7 @@ END_FTR_SECTION_IFSET(CPU_FTR_HAS_PPR) /* Save HEIR (HV emulation assist reg) in emul_inst if this is an HEI (HV emulation interrupt, e40) */ li r3,KVM_INST_FETCH_FAILED + stw r3,VCPU_LAST_INST(r9) cmpwi r12,BOOK3S_INTERRUPT_H_EMUL_ASSIST bne 11f mfspr r3,SPRN_HEIR -- cgit v1.2.1 From aca5e361e84d20ac68eedb37b7c4561a33f56198 Mon Sep 17 00:00:00 2001 From: Rodrigo Vivi Date: Fri, 13 Mar 2015 16:13:59 -0700 Subject: drm/i915: Fix sink crc connector iteration Regressed by this commit: commit 3455454e18ca3f92c565700539e744c620d8276b Author: Ander Conselvan de Oliveira Date: Tue Mar 3 15:21:56 2015 +0200 drm/i915: Add a for_each_intel_connector macro Cc: Ander Conselvan de Oliveira Signed-off-by: Rodrigo Vivi Reviewed-by: Ander Conselvan de Oliveira Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_debugfs.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c index 1a52d6ab0f80..aaf756047a20 100644 --- a/drivers/gpu/drm/i915/i915_debugfs.c +++ b/drivers/gpu/drm/i915/i915_debugfs.c @@ -2308,7 +2308,7 @@ static int i915_sink_crc(struct seq_file *m, void *data) u8 crc[6]; drm_modeset_lock_all(dev); - for_each_intel_encoder(dev, connector) { + for_each_intel_connector(dev, connector) { if (connector->base.dpms != DRM_MODE_DPMS_ON) continue; -- cgit v1.2.1 From ec7adb6ee79c8c9fe64d63ad638a31cd62e55515 Mon Sep 17 00:00:00 2001 From: Joonas Lahtinen Date: Mon, 16 Mar 2015 14:11:13 +0200 Subject: drm/i915: Do not use ggtt_view with (aliasing) PPGTT GGTT views are only applicable when dealing with GGTT. Change the code to reject ggtt_view where it should not be used and require it when it should be. v2: - Dropped _ppgtt_ infixes, allow both types to be passed - Disregard other but normal views when no view is specified - More checks that valid parameters are passed - More readable error checking v3: - Prefer WARN_ONCE over BUG_ON when there is code path for failure Signed-off-by: Joonas Lahtinen Reviewed-by: Tvrtko Ursulin [danvet: Drop unecessary forward decl from earlier patch iterations.] [danvet: Remove unused variable spotted by Tvrtko.] Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_drv.h | 100 ++++++++------------- drivers/gpu/drm/i915/i915_gem.c | 174 ++++++++++++++++++++++++++++-------- drivers/gpu/drm/i915/i915_gem_gtt.c | 71 +++++++++++---- 3 files changed, 226 insertions(+), 119 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 8ba7e1b7b733..81f60b48def2 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -2623,20 +2623,16 @@ void i915_gem_vma_destroy(struct i915_vma *vma); #define PIN_GLOBAL 0x4 #define PIN_OFFSET_BIAS 0x8 #define PIN_OFFSET_MASK (~4095) -int __must_check i915_gem_object_pin_view(struct drm_i915_gem_object *obj, - struct i915_address_space *vm, - uint32_t alignment, - uint64_t flags, - const struct i915_ggtt_view *view); -static inline -int __must_check i915_gem_object_pin(struct drm_i915_gem_object *obj, - struct i915_address_space *vm, - uint32_t alignment, - uint64_t flags) -{ - return i915_gem_object_pin_view(obj, vm, alignment, flags, - &i915_ggtt_view_normal); -} +int __must_check +i915_gem_object_pin(struct drm_i915_gem_object *obj, + struct i915_address_space *vm, + uint32_t alignment, + uint64_t flags); +int __must_check +i915_gem_object_ggtt_pin(struct drm_i915_gem_object *obj, + const struct i915_ggtt_view *view, + uint32_t alignment, + uint64_t flags); int i915_vma_bind(struct i915_vma *vma, enum i915_cache_level cache_level, u32 flags); @@ -2800,60 +2796,46 @@ struct dma_buf *i915_gem_prime_export(struct drm_device *dev, void i915_gem_restore_fences(struct drm_device *dev); -unsigned long i915_gem_obj_offset_view(struct drm_i915_gem_object *o, - struct i915_address_space *vm, - enum i915_ggtt_view_type view); -static inline -unsigned long i915_gem_obj_offset(struct drm_i915_gem_object *o, - struct i915_address_space *vm) +unsigned long +i915_gem_obj_ggtt_offset_view(struct drm_i915_gem_object *o, + enum i915_ggtt_view_type view); +unsigned long +i915_gem_obj_offset(struct drm_i915_gem_object *o, + struct i915_address_space *vm); +static inline unsigned long +i915_gem_obj_ggtt_offset(struct drm_i915_gem_object *o) { - return i915_gem_obj_offset_view(o, vm, I915_GGTT_VIEW_NORMAL); + return i915_gem_obj_ggtt_offset_view(o, I915_GGTT_VIEW_NORMAL); } + bool i915_gem_obj_bound_any(struct drm_i915_gem_object *o); -bool i915_gem_obj_bound_view(struct drm_i915_gem_object *o, - struct i915_address_space *vm, - enum i915_ggtt_view_type view); -static inline +bool i915_gem_obj_ggtt_bound_view(struct drm_i915_gem_object *o, + enum i915_ggtt_view_type view); bool i915_gem_obj_bound(struct drm_i915_gem_object *o, - struct i915_address_space *vm) -{ - return i915_gem_obj_bound_view(o, vm, I915_GGTT_VIEW_NORMAL); -} + struct i915_address_space *vm); unsigned long i915_gem_obj_size(struct drm_i915_gem_object *o, struct i915_address_space *vm); -struct i915_vma *i915_gem_obj_to_vma_view(struct drm_i915_gem_object *obj, - struct i915_address_space *vm, - const struct i915_ggtt_view *view); -static inline -struct i915_vma *i915_gem_obj_to_vma(struct drm_i915_gem_object *obj, - struct i915_address_space *vm) -{ - return i915_gem_obj_to_vma_view(obj, vm, &i915_ggtt_view_normal); -} - struct i915_vma * -i915_gem_obj_lookup_or_create_vma_view(struct drm_i915_gem_object *obj, - struct i915_address_space *vm, - const struct i915_ggtt_view *view); +i915_gem_obj_to_vma(struct drm_i915_gem_object *obj, + struct i915_address_space *vm); +struct i915_vma * +i915_gem_obj_to_ggtt_view(struct drm_i915_gem_object *obj, + const struct i915_ggtt_view *view); -static inline struct i915_vma * i915_gem_obj_lookup_or_create_vma(struct drm_i915_gem_object *obj, - struct i915_address_space *vm) -{ - return i915_gem_obj_lookup_or_create_vma_view(obj, vm, - &i915_ggtt_view_normal); -} + struct i915_address_space *vm); +struct i915_vma * +i915_gem_obj_lookup_or_create_ggtt_vma(struct drm_i915_gem_object *obj, + const struct i915_ggtt_view *view); -struct i915_vma *i915_gem_obj_to_ggtt(struct drm_i915_gem_object *obj); -static inline bool i915_gem_obj_is_pinned(struct drm_i915_gem_object *obj) { - struct i915_vma *vma; - list_for_each_entry(vma, &obj->vma_list, vma_link) - if (vma->pin_count > 0) - return true; - return false; +static inline struct i915_vma * +i915_gem_obj_to_ggtt(struct drm_i915_gem_object *obj) +{ + return i915_gem_obj_to_ggtt_view(obj, &i915_ggtt_view_normal); } +bool i915_gem_obj_is_pinned(struct drm_i915_gem_object *obj); /* Some GGTT VM helpers */ #define i915_obj_to_ggtt(obj) \ @@ -2876,13 +2858,7 @@ i915_vm_to_ppgtt(struct i915_address_space *vm) static inline bool i915_gem_obj_ggtt_bound(struct drm_i915_gem_object *obj) { - return i915_gem_obj_bound(obj, i915_obj_to_ggtt(obj)); -} - -static inline unsigned long -i915_gem_obj_ggtt_offset(struct drm_i915_gem_object *obj) -{ - return i915_gem_obj_offset(obj, i915_obj_to_ggtt(obj)); + return i915_gem_obj_ggtt_bound_view(obj, I915_GGTT_VIEW_NORMAL); } static inline unsigned long diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 0fe313d0f609..cc8672a16da4 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -3518,9 +3518,9 @@ static bool i915_gem_valid_gtt_space(struct i915_vma *vma, static struct i915_vma * i915_gem_object_bind_to_vm(struct drm_i915_gem_object *obj, struct i915_address_space *vm, + const struct i915_ggtt_view *ggtt_view, unsigned alignment, - uint64_t flags, - const struct i915_ggtt_view *view) + uint64_t flags) { struct drm_device *dev = obj->base.dev; struct drm_i915_private *dev_priv = dev->dev_private; @@ -3532,6 +3532,9 @@ i915_gem_object_bind_to_vm(struct drm_i915_gem_object *obj, struct i915_vma *vma; int ret; + if(WARN_ON(i915_is_ggtt(vm) != !!ggtt_view)) + return ERR_PTR(-EINVAL); + fence_size = i915_gem_get_gtt_size(dev, obj->base.size, obj->tiling_mode); @@ -3570,7 +3573,9 @@ i915_gem_object_bind_to_vm(struct drm_i915_gem_object *obj, i915_gem_object_pin_pages(obj); - vma = i915_gem_obj_lookup_or_create_vma_view(obj, vm, view); + vma = ggtt_view ? i915_gem_obj_lookup_or_create_ggtt_vma(obj, ggtt_view) : + i915_gem_obj_lookup_or_create_vma(obj, vm); + if (IS_ERR(vma)) goto err_unpin; @@ -4167,12 +4172,12 @@ i915_vma_misplaced(struct i915_vma *vma, uint32_t alignment, uint64_t flags) return false; } -int -i915_gem_object_pin_view(struct drm_i915_gem_object *obj, - struct i915_address_space *vm, - uint32_t alignment, - uint64_t flags, - const struct i915_ggtt_view *view) +static int +i915_gem_object_do_pin(struct drm_i915_gem_object *obj, + struct i915_address_space *vm, + const struct i915_ggtt_view *ggtt_view, + uint32_t alignment, + uint64_t flags) { struct drm_i915_private *dev_priv = obj->base.dev->dev_private; struct i915_vma *vma; @@ -4188,17 +4193,29 @@ i915_gem_object_pin_view(struct drm_i915_gem_object *obj, if (WARN_ON((flags & (PIN_MAPPABLE | PIN_GLOBAL)) == PIN_MAPPABLE)) return -EINVAL; - vma = i915_gem_obj_to_vma_view(obj, vm, view); + if (WARN_ON(i915_is_ggtt(vm) != !!ggtt_view)) + return -EINVAL; + + vma = ggtt_view ? i915_gem_obj_to_ggtt_view(obj, ggtt_view) : + i915_gem_obj_to_vma(obj, vm); + + if (IS_ERR(vma)) + return PTR_ERR(vma); + if (vma) { if (WARN_ON(vma->pin_count == DRM_I915_GEM_OBJECT_MAX_PIN_COUNT)) return -EBUSY; if (i915_vma_misplaced(vma, alignment, flags)) { + unsigned long offset; + offset = ggtt_view ? i915_gem_obj_ggtt_offset_view(obj, ggtt_view->type) : + i915_gem_obj_offset(obj, vm); WARN(vma->pin_count, - "bo is already pinned with incorrect alignment:" + "bo is already pinned in %s with incorrect alignment:" " offset=%lx, req.alignment=%x, req.map_and_fenceable=%d," " obj->map_and_fenceable=%d\n", - i915_gem_obj_offset_view(obj, vm, view->type), + ggtt_view ? "ggtt" : "ppgtt", + offset, alignment, !!(flags & PIN_MAPPABLE), obj->map_and_fenceable); @@ -4212,8 +4229,8 @@ i915_gem_object_pin_view(struct drm_i915_gem_object *obj, bound = vma ? vma->bound : 0; if (vma == NULL || !drm_mm_node_allocated(&vma->node)) { - vma = i915_gem_object_bind_to_vm(obj, vm, alignment, - flags, view); + vma = i915_gem_object_bind_to_vm(obj, vm, ggtt_view, alignment, + flags); if (IS_ERR(vma)) return PTR_ERR(vma); } @@ -4254,6 +4271,30 @@ i915_gem_object_pin_view(struct drm_i915_gem_object *obj, return 0; } +int +i915_gem_object_pin(struct drm_i915_gem_object *obj, + struct i915_address_space *vm, + uint32_t alignment, + uint64_t flags) +{ + return i915_gem_object_do_pin(obj, vm, + i915_is_ggtt(vm) ? &i915_ggtt_view_normal : NULL, + alignment, flags); +} + +int +i915_gem_object_ggtt_pin(struct drm_i915_gem_object *obj, + const struct i915_ggtt_view *view, + uint32_t alignment, + uint64_t flags) +{ + if (WARN_ONCE(!view, "no view specified")) + return -EINVAL; + + return i915_gem_object_do_pin(obj, i915_obj_to_ggtt(obj), view, + alignment, flags); +} + void i915_gem_object_ggtt_unpin(struct drm_i915_gem_object *obj) { @@ -4559,15 +4600,32 @@ void i915_gem_free_object(struct drm_gem_object *gem_obj) intel_runtime_pm_put(dev_priv); } -struct i915_vma *i915_gem_obj_to_vma_view(struct drm_i915_gem_object *obj, - struct i915_address_space *vm, - const struct i915_ggtt_view *view) +struct i915_vma *i915_gem_obj_to_vma(struct drm_i915_gem_object *obj, + struct i915_address_space *vm) { struct i915_vma *vma; - list_for_each_entry(vma, &obj->vma_list, vma_link) - if (vma->vm == vm && vma->ggtt_view.type == view->type) + list_for_each_entry(vma, &obj->vma_list, vma_link) { + if (i915_is_ggtt(vma->vm) && + vma->ggtt_view.type != I915_GGTT_VIEW_NORMAL) + continue; + if (vma->vm == vm) return vma; + } + return NULL; +} + +struct i915_vma *i915_gem_obj_to_ggtt_view(struct drm_i915_gem_object *obj, + const struct i915_ggtt_view *view) +{ + struct i915_address_space *ggtt = i915_obj_to_ggtt(obj); + struct i915_vma *vma; + if (WARN_ONCE(!view, "no view specified")) + return ERR_PTR(-EINVAL); + + list_for_each_entry(vma, &obj->vma_list, vma_link) + if (vma->vm == ggtt && vma->ggtt_view.type == view->type) + return vma; return NULL; } @@ -5176,9 +5234,9 @@ i915_gem_shrinker_count(struct shrinker *shrinker, struct shrink_control *sc) } /* All the new VM stuff */ -unsigned long i915_gem_obj_offset_view(struct drm_i915_gem_object *o, - struct i915_address_space *vm, - enum i915_ggtt_view_type view) +unsigned long +i915_gem_obj_offset(struct drm_i915_gem_object *o, + struct i915_address_space *vm) { struct drm_i915_private *dev_priv = o->base.dev->dev_private; struct i915_vma *vma; @@ -5186,23 +5244,57 @@ unsigned long i915_gem_obj_offset_view(struct drm_i915_gem_object *o, WARN_ON(vm == &dev_priv->mm.aliasing_ppgtt->base); list_for_each_entry(vma, &o->vma_list, vma_link) { - if (vma->vm == vm && vma->ggtt_view.type == view) + if (i915_is_ggtt(vma->vm) && + vma->ggtt_view.type != I915_GGTT_VIEW_NORMAL) + continue; + if (vma->vm == vm) return vma->node.start; - } + WARN(1, "%s vma for this object not found.\n", i915_is_ggtt(vm) ? "global" : "ppgtt"); return -1; } -bool i915_gem_obj_bound_view(struct drm_i915_gem_object *o, - struct i915_address_space *vm, - enum i915_ggtt_view_type view) +unsigned long +i915_gem_obj_ggtt_offset_view(struct drm_i915_gem_object *o, + enum i915_ggtt_view_type view) { + struct i915_address_space *ggtt = i915_obj_to_ggtt(o); struct i915_vma *vma; list_for_each_entry(vma, &o->vma_list, vma_link) - if (vma->vm == vm && + if (vma->vm == ggtt && vma->ggtt_view.type == view) + return vma->node.start; + + WARN(1, "global vma for this object not found.\n"); + return -1; +} + +bool i915_gem_obj_bound(struct drm_i915_gem_object *o, + struct i915_address_space *vm) +{ + struct i915_vma *vma; + + list_for_each_entry(vma, &o->vma_list, vma_link) { + if (i915_is_ggtt(vma->vm) && + vma->ggtt_view.type != I915_GGTT_VIEW_NORMAL) + continue; + if (vma->vm == vm && drm_mm_node_allocated(&vma->node)) + return true; + } + + return false; +} + +bool i915_gem_obj_ggtt_bound_view(struct drm_i915_gem_object *o, + enum i915_ggtt_view_type view) +{ + struct i915_address_space *ggtt = i915_obj_to_ggtt(o); + struct i915_vma *vma; + + list_for_each_entry(vma, &o->vma_list, vma_link) + if (vma->vm == ggtt && vma->ggtt_view.type == view && drm_mm_node_allocated(&vma->node)) return true; @@ -5231,10 +5323,13 @@ unsigned long i915_gem_obj_size(struct drm_i915_gem_object *o, BUG_ON(list_empty(&o->vma_list)); - list_for_each_entry(vma, &o->vma_list, vma_link) + list_for_each_entry(vma, &o->vma_list, vma_link) { + if (i915_is_ggtt(vma->vm) && + vma->ggtt_view.type != I915_GGTT_VIEW_NORMAL) + continue; if (vma->vm == vm) return vma->node.size; - + } return 0; } @@ -5334,15 +5429,16 @@ i915_gem_shrinker_oom(struct notifier_block *nb, unsigned long event, void *ptr) return NOTIFY_DONE; } -struct i915_vma *i915_gem_obj_to_ggtt(struct drm_i915_gem_object *obj) +bool i915_gem_obj_is_pinned(struct drm_i915_gem_object *obj) { - struct i915_address_space *ggtt = i915_obj_to_ggtt(obj); struct i915_vma *vma; - - list_for_each_entry(vma, &obj->vma_list, vma_link) - if (vma->vm == ggtt && - vma->ggtt_view.type == I915_GGTT_VIEW_NORMAL) - return vma; - - return NULL; + list_for_each_entry(vma, &obj->vma_list, vma_link) { + if (i915_is_ggtt(vma->vm) && + vma->ggtt_view.type != I915_GGTT_VIEW_NORMAL) + continue; + if (vma->pin_count > 0) + return true; + } + return false; } + diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c index 2034f7cf238b..f1b9ea61de4b 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c @@ -67,8 +67,9 @@ * i915_ggtt_view_type and struct i915_ggtt_view. * * A new flavour of core GEM functions which work with GGTT bound objects were - * added with the _view suffix. They take the struct i915_ggtt_view parameter - * encapsulating all metadata required to implement a view. + * added with the _ggtt_ infix, and sometimes with _view postfix to avoid + * renaming in large amounts of code. They take the struct i915_ggtt_view + * parameter encapsulating all metadata required to implement a view. * * As a helper for callers which are only interested in the normal view, * globally const i915_ggtt_view_normal singleton instance exists. All old core @@ -1726,11 +1727,15 @@ static void ggtt_bind_vma(struct i915_vma *vma, struct drm_device *dev = vma->vm->dev; struct drm_i915_private *dev_priv = dev->dev_private; struct drm_i915_gem_object *obj = vma->obj; + struct sg_table *pages = obj->pages; /* Currently applicable only to VLV */ if (obj->gt_ro) flags |= PTE_READ_ONLY; + if (i915_is_ggtt(vma->vm)) + pages = vma->ggtt_view.pages; + /* If there is no aliasing PPGTT, or the caller needs a global mapping, * or we have a global mapping already but the cacheability flags have * changed, set the global PTEs. @@ -1745,7 +1750,7 @@ static void ggtt_bind_vma(struct i915_vma *vma, if (!dev_priv->mm.aliasing_ppgtt || flags & GLOBAL_BIND) { if (!(vma->bound & GLOBAL_BIND) || (cache_level != obj->cache_level)) { - vma->vm->insert_entries(vma->vm, vma->ggtt_view.pages, + vma->vm->insert_entries(vma->vm, pages, vma->node.start, cache_level, flags); vma->bound |= GLOBAL_BIND; @@ -1756,8 +1761,7 @@ static void ggtt_bind_vma(struct i915_vma *vma, (!(vma->bound & LOCAL_BIND) || (cache_level != obj->cache_level))) { struct i915_hw_ppgtt *appgtt = dev_priv->mm.aliasing_ppgtt; - appgtt->base.insert_entries(&appgtt->base, - vma->ggtt_view.pages, + appgtt->base.insert_entries(&appgtt->base, pages, vma->node.start, cache_level, flags); vma->bound |= LOCAL_BIND; @@ -2331,23 +2335,28 @@ int i915_gem_gtt_init(struct drm_device *dev) return 0; } -static struct i915_vma *__i915_gem_vma_create(struct drm_i915_gem_object *obj, - struct i915_address_space *vm, - const struct i915_ggtt_view *view) +static struct i915_vma * +__i915_gem_vma_create(struct drm_i915_gem_object *obj, + struct i915_address_space *vm, + const struct i915_ggtt_view *ggtt_view) { struct i915_vma *vma = kzalloc(sizeof(*vma), GFP_KERNEL); if (vma == NULL) return ERR_PTR(-ENOMEM); + if (WARN_ON(i915_is_ggtt(vm) != !!ggtt_view)) + return ERR_PTR(-EINVAL); + INIT_LIST_HEAD(&vma->vma_link); INIT_LIST_HEAD(&vma->mm_list); INIT_LIST_HEAD(&vma->exec_list); vma->vm = vm; vma->obj = obj; - vma->ggtt_view = *view; if (INTEL_INFO(vm->dev)->gen >= 6) { if (i915_is_ggtt(vm)) { + vma->ggtt_view = *ggtt_view; + vma->unbind_vma = ggtt_unbind_vma; vma->bind_vma = ggtt_bind_vma; } else { @@ -2356,6 +2365,7 @@ static struct i915_vma *__i915_gem_vma_create(struct drm_i915_gem_object *obj, } } else { BUG_ON(!i915_is_ggtt(vm)); + vma->ggtt_view = *ggtt_view; vma->unbind_vma = i915_ggtt_unbind_vma; vma->bind_vma = i915_ggtt_bind_vma; } @@ -2368,21 +2378,44 @@ static struct i915_vma *__i915_gem_vma_create(struct drm_i915_gem_object *obj, } struct i915_vma * -i915_gem_obj_lookup_or_create_vma_view(struct drm_i915_gem_object *obj, - struct i915_address_space *vm, +i915_gem_obj_lookup_or_create_vma(struct drm_i915_gem_object *obj, + struct i915_address_space *vm) +{ + struct i915_vma *vma; + + vma = i915_gem_obj_to_vma(obj, vm); + if (!vma) + vma = __i915_gem_vma_create(obj, vm, + i915_is_ggtt(vm) ? &i915_ggtt_view_normal : NULL); + + return vma; +} + +struct i915_vma * +i915_gem_obj_lookup_or_create_ggtt_vma(struct drm_i915_gem_object *obj, const struct i915_ggtt_view *view) { + struct i915_address_space *ggtt = i915_obj_to_ggtt(obj); struct i915_vma *vma; - vma = i915_gem_obj_to_vma_view(obj, vm, view); + if (WARN_ON(!view)) + return ERR_PTR(-EINVAL); + + vma = i915_gem_obj_to_ggtt_view(obj, view); + + if (IS_ERR(vma)) + return vma; + if (!vma) - vma = __i915_gem_vma_create(obj, vm, view); + vma = __i915_gem_vma_create(obj, ggtt, view); return vma; + } + static inline -int i915_get_vma_pages(struct i915_vma *vma) +int i915_get_ggtt_vma_pages(struct i915_vma *vma) { if (vma->ggtt_view.pages) return 0; @@ -2394,7 +2427,7 @@ int i915_get_vma_pages(struct i915_vma *vma) vma->ggtt_view.type); if (!vma->ggtt_view.pages) { - DRM_ERROR("Failed to get pages for VMA view type %u!\n", + DRM_ERROR("Failed to get pages for GGTT view type %u!\n", vma->ggtt_view.type); return -EINVAL; } @@ -2415,10 +2448,12 @@ int i915_get_vma_pages(struct i915_vma *vma) int i915_vma_bind(struct i915_vma *vma, enum i915_cache_level cache_level, u32 flags) { - int ret = i915_get_vma_pages(vma); + if (i915_is_ggtt(vma->vm)) { + int ret = i915_get_ggtt_vma_pages(vma); - if (ret) - return ret; + if (ret) + return ret; + } vma->bind_vma(vma, cache_level, flags); -- cgit v1.2.1 From f4896f1529c43fb0688164ed33aebdf525a34fb7 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 12 Mar 2015 17:10:27 +0200 Subject: drm/i915: Make the DP rates int instead of uint32_t MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit No point in using uint32_t here, just plain old int will do. Signed-off-by: Ville Syrjälä Reviewed-by: Sonika Jindal Reviewed-by: Todd Previte Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_dp.c | 24 ++++++++++++------------ 1 file changed, 12 insertions(+), 12 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index ca60060710d2..a9b984734b1a 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -85,10 +85,9 @@ static const struct dp_link_dpll chv_dpll[] = { { .p1 = 2, .p2 = 1, .n = 1, .m1 = 2, .m2 = 0x6c00000 } } }; /* Skylake supports following rates */ -static const uint32_t gen9_rates[] = { 162000, 216000, 270000, 324000, - 432000, 540000 }; - -static const uint32_t default_rates[] = { 162000, 270000, 540000 }; +static const int gen9_rates[] = { 162000, 216000, 270000, + 324000, 432000, 540000 }; +static const int default_rates[] = { 162000, 270000, 540000 }; /** * is_edp - is the given port attached to an eDP panel (either CPU or PCH) @@ -1142,7 +1141,7 @@ hsw_dp_set_ddi_pll_sel(struct intel_crtc_state *pipe_config, int link_bw) } static int -intel_read_sink_rates(struct intel_dp *intel_dp, uint32_t *sink_rates) +intel_read_sink_rates(struct intel_dp *intel_dp, int *sink_rates) { struct drm_device *dev = intel_dp_to_dev(intel_dp); int i = 0; @@ -1169,7 +1168,7 @@ intel_read_sink_rates(struct intel_dp *intel_dp, uint32_t *sink_rates) } static int -intel_read_source_rates(struct intel_dp *intel_dp, uint32_t *source_rates) +intel_read_source_rates(struct intel_dp *intel_dp, int *source_rates) { struct drm_device *dev = intel_dp_to_dev(intel_dp); int i; @@ -1220,8 +1219,9 @@ intel_dp_set_clock(struct intel_encoder *encoder, } } -static int intel_supported_rates(const uint32_t *source_rates, int source_len, -const uint32_t *sink_rates, int sink_len, uint32_t *supported_rates) +static int intel_supported_rates(const int *source_rates, int source_len, + const int *sink_rates, int sink_len, + int *supported_rates) { int i = 0, j = 0, k = 0; @@ -1248,7 +1248,7 @@ const uint32_t *sink_rates, int sink_len, uint32_t *supported_rates) return k; } -static int rate_to_index(uint32_t find, const uint32_t *rates) +static int rate_to_index(int find, const int *rates) { int i = 0; @@ -1278,9 +1278,9 @@ intel_dp_compute_config(struct intel_encoder *encoder, int max_clock; int bpp, mode_rate; int link_avail, link_clock; - uint32_t sink_rates[8]; - uint32_t supported_rates[8] = {0}; - uint32_t source_rates[8]; + int sink_rates[8]; + int supported_rates[8] = {0}; + int source_rates[8]; int source_len, sink_len, supported_len; sink_len = intel_read_sink_rates(intel_dp, sink_rates); -- cgit v1.2.1 From ea2d8a427fa99b70457518c8d3516f46e572b95d Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 12 Mar 2015 17:10:28 +0200 Subject: drm/i915: Store the converted link rates in intel_dp->supported_rates[] MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit No point in converting from hardware format every single time, just store the rates in the final format under intel_dp. Signed-off-by: Ville Syrjälä Reviewed-by: Sonika Jindal Reviewed-by: Todd Previte Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_dp.c | 33 +++++++++++++++++++-------------- drivers/gpu/drm/i915/intel_drv.h | 3 ++- 2 files changed, 21 insertions(+), 15 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index a9b984734b1a..f71ede776f2b 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -1144,8 +1144,6 @@ static int intel_read_sink_rates(struct intel_dp *intel_dp, int *sink_rates) { struct drm_device *dev = intel_dp_to_dev(intel_dp); - int i = 0; - uint16_t val; if (INTEL_INFO(dev)->gen >= 9 && intel_dp->supported_rates[0]) { /* @@ -1153,18 +1151,12 @@ intel_read_sink_rates(struct intel_dp *intel_dp, int *sink_rates) * link rate table method, so read link rates from * supported_link_rates */ - for (i = 0; i < DP_MAX_SUPPORTED_RATES; ++i) { - val = le16_to_cpu(intel_dp->supported_rates[i]); - if (val == 0) - break; - - sink_rates[i] = val * 200; - } + memcpy(sink_rates, intel_dp->supported_rates, + sizeof(intel_dp->supported_rates)); - if (i <= 0) - DRM_ERROR("No rates in SUPPORTED_LINK_RATES"); + return intel_dp->num_supported_rates; } - return i; + return 0; } static int @@ -3754,10 +3746,23 @@ intel_dp_get_dpcd(struct intel_dp *intel_dp) (intel_dp->dpcd[DP_EDP_CONFIGURATION_CAP] & DP_DPCD_DISPLAY_CONTROL_CAPABLE) && (intel_dp_dpcd_read_wake(&intel_dp->aux, DP_EDP_DPCD_REV, &rev, 1) == 1) && (rev >= 0x03)) { /* eDp v1.4 or higher */ + __le16 supported_rates[DP_MAX_SUPPORTED_RATES]; + int i; + intel_dp_dpcd_read_wake(&intel_dp->aux, DP_SUPPORTED_LINK_RATES, - intel_dp->supported_rates, - sizeof(intel_dp->supported_rates)); + supported_rates, + sizeof(supported_rates)); + + for (i = 0; i < ARRAY_SIZE(supported_rates); i++) { + int val = le16_to_cpu(supported_rates[i]); + + if (val == 0) + break; + + intel_dp->supported_rates[i] = val * 200; + } + intel_dp->num_supported_rates = i; } if (!(intel_dp->dpcd[DP_DOWNSTREAMPORT_PRESENT] & DP_DWN_STRM_PORT_PRESENT)) diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index c77128c67cf8..69c8437be611 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -627,7 +627,8 @@ struct intel_dp { uint8_t dpcd[DP_RECEIVER_CAP_SIZE]; uint8_t psr_dpcd[EDP_PSR_RECEIVER_CAP_SIZE]; uint8_t downstream_ports[DP_MAX_DOWNSTREAM_PORTS]; - __le16 supported_rates[DP_MAX_SUPPORTED_RATES]; + uint8_t num_supported_rates; + int supported_rates[DP_MAX_SUPPORTED_RATES]; struct drm_dp_aux aux; uint8_t train_set[4]; int panel_power_up_delay; -- cgit v1.2.1 From 636280ba55fd95aaf14ecbfb2c32cb60a9e74e58 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 12 Mar 2015 17:10:29 +0200 Subject: drm/i915: Don't copy the DP source rates arrays MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit The source rates don't change, so we can just point the caller at the const arrays. Signed-off-by: Ville Syrjälä Reviewed-by: Sonika Jindal Reviewed-by: Todd Previte Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_dp.c | 24 ++++++++++-------------- 1 file changed, 10 insertions(+), 14 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index f71ede776f2b..12dd56b99b4d 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -1160,22 +1160,18 @@ intel_read_sink_rates(struct intel_dp *intel_dp, int *sink_rates) } static int -intel_read_source_rates(struct intel_dp *intel_dp, int *source_rates) +intel_dp_source_rates(struct intel_dp *intel_dp, const int **source_rates) { struct drm_device *dev = intel_dp_to_dev(intel_dp); - int i; - int max_default_rate; - if (INTEL_INFO(dev)->gen >= 9 && intel_dp->supported_rates[0]) { - for (i = 0; i < ARRAY_SIZE(gen9_rates); ++i) - source_rates[i] = gen9_rates[i]; - } else { - /* Index of the max_link_bw supported + 1 */ - max_default_rate = (intel_dp_max_link_bw(intel_dp) >> 3) + 1; - for (i = 0; i < max_default_rate; ++i) - source_rates[i] = default_rates[i]; + if (INTEL_INFO(dev)->gen >= 9) { + *source_rates = gen9_rates; + return ARRAY_SIZE(gen9_rates); } - return i; + + *source_rates = default_rates; + + return (intel_dp_max_link_bw(intel_dp) >> 3) + 1; } static void @@ -1272,12 +1268,12 @@ intel_dp_compute_config(struct intel_encoder *encoder, int link_avail, link_clock; int sink_rates[8]; int supported_rates[8] = {0}; - int source_rates[8]; + const int *source_rates; int source_len, sink_len, supported_len; sink_len = intel_read_sink_rates(intel_dp, sink_rates); - source_len = intel_read_source_rates(intel_dp, source_rates); + source_len = intel_dp_source_rates(intel_dp, &source_rates); supported_len = intel_supported_rates(source_rates, source_len, sink_rates, sink_len, supported_rates); -- cgit v1.2.1 From 12f6a2e21be1ec4ad3dd6e2657b675f0779ee907 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 12 Mar 2015 17:10:30 +0200 Subject: drm/i915: Don't copy sink rates either MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Once we've read the rates from the sink we don't have to mess with them, so the caller can just look at the stored rates without doing extra copies. If the sink doesn't support the new link rate stuff, we just point the caller at the default_rates[] array. Signed-off-by: Ville Syrjälä Reviewed-by: Sonika Jindal Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_dp.c | 24 +++++++++--------------- 1 file changed, 9 insertions(+), 15 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 12dd56b99b4d..756917512cb9 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -1141,22 +1141,16 @@ hsw_dp_set_ddi_pll_sel(struct intel_crtc_state *pipe_config, int link_bw) } static int -intel_read_sink_rates(struct intel_dp *intel_dp, int *sink_rates) +intel_dp_sink_rates(struct intel_dp *intel_dp, const int **sink_rates) { - struct drm_device *dev = intel_dp_to_dev(intel_dp); - - if (INTEL_INFO(dev)->gen >= 9 && intel_dp->supported_rates[0]) { - /* - * Receiver supports only main-link rate selection by - * link rate table method, so read link rates from - * supported_link_rates - */ - memcpy(sink_rates, intel_dp->supported_rates, - sizeof(intel_dp->supported_rates)); - + if (intel_dp->num_supported_rates) { + *sink_rates = intel_dp->supported_rates; return intel_dp->num_supported_rates; } - return 0; + + *sink_rates = default_rates; + + return (intel_dp_max_link_bw(intel_dp) >> 3) + 1; } static int @@ -1266,12 +1260,12 @@ intel_dp_compute_config(struct intel_encoder *encoder, int max_clock; int bpp, mode_rate; int link_avail, link_clock; - int sink_rates[8]; + const int *sink_rates; int supported_rates[8] = {0}; const int *source_rates; int source_len, sink_len, supported_len; - sink_len = intel_read_sink_rates(intel_dp, sink_rates); + sink_len = intel_dp_sink_rates(intel_dp, &sink_rates); source_len = intel_dp_source_rates(intel_dp, &source_rates); -- cgit v1.2.1 From d098a5054373d860f062b42466fb965841c161fe Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 12 Mar 2015 17:10:31 +0200 Subject: drm/i915: Remove special case from intel_supported_rates() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Now that both source and sink rates are always filled in there's no need for any special cases in intel_supported_rates(). Signed-off-by: Ville Syrjälä Reviewed-by: Sonika Jindal Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_dp.c | 8 -------- 1 file changed, 8 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 756917512cb9..4cbb956f0fe3 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -1207,14 +1207,6 @@ static int intel_supported_rates(const int *source_rates, int source_len, { int i = 0, j = 0, k = 0; - /* For panels with edp version less than 1.4 */ - if (sink_len == 0) { - for (i = 0; i < source_len; ++i) - supported_rates[i] = source_rates[i]; - return source_len; - } - - /* For edp1.4 panels, find the common rates between source and sink */ while (i < source_len && j < sink_len) { if (source_rates[i] == sink_rates[j]) { supported_rates[k] = source_rates[i]; -- cgit v1.2.1 From 1db10e28b25687fea1d11a94196b0477145a24a8 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 12 Mar 2015 17:10:32 +0200 Subject: drm/i915: Fully separate source vs. sink rates MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Remove the sink vs. source limit mess from intel_dp_max_link_bw() and just move the source restriction checks to intel_dp_source_rates(). Signed-off-by: Ville Syrjälä Reviewed-by: Sonika Jindal [danvet: Resolve conflict with WaDisableHBR2:skl patch.] Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_dp.c | 28 +++++++++++----------------- 1 file changed, 11 insertions(+), 17 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 4cbb956f0fe3..7553f08d87e1 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -126,22 +126,11 @@ int intel_dp_max_link_bw(struct intel_dp *intel_dp) { int max_link_bw = intel_dp->dpcd[DP_MAX_LINK_RATE]; - struct drm_device *dev = intel_dp->attached_connector->base.dev; switch (max_link_bw) { case DP_LINK_BW_1_62: case DP_LINK_BW_2_7: - break; - case DP_LINK_BW_5_4: /* 1.2 capable displays may advertise higher bw */ - if (IS_SKYLAKE(dev) && INTEL_REVID(dev) <= SKL_REVID_B0) - /* WaDisableHBR2:skl */ - max_link_bw = DP_LINK_BW_2_7; - else if (((IS_HASWELL(dev) && !IS_HSW_ULX(dev)) || - INTEL_INFO(dev)->gen >= 8) && - intel_dp->dpcd[DP_DPCD_REV] >= 0x12) - max_link_bw = DP_LINK_BW_5_4; - else - max_link_bw = DP_LINK_BW_2_7; + case DP_LINK_BW_5_4: break; default: WARN(1, "invalid max DP link bw val %x, using 1.62Gbps\n", @@ -1154,10 +1143,8 @@ intel_dp_sink_rates(struct intel_dp *intel_dp, const int **sink_rates) } static int -intel_dp_source_rates(struct intel_dp *intel_dp, const int **source_rates) +intel_dp_source_rates(struct drm_device *dev, const int **source_rates) { - struct drm_device *dev = intel_dp_to_dev(intel_dp); - if (INTEL_INFO(dev)->gen >= 9) { *source_rates = gen9_rates; return ARRAY_SIZE(gen9_rates); @@ -1165,7 +1152,14 @@ intel_dp_source_rates(struct intel_dp *intel_dp, const int **source_rates) *source_rates = default_rates; - return (intel_dp_max_link_bw(intel_dp) >> 3) + 1; + if (IS_SKYLAKE(dev) && INTEL_REVID(dev) <= SKL_REVID_B0) + /* WaDisableHBR2:skl */ + return (DP_LINK_BW_2_7 >> 3) + 1; + else if (INTEL_INFO(dev)->gen >= 8 || + (IS_HASWELL(dev) && !IS_HSW_ULX(dev))) + return (DP_LINK_BW_5_4 >> 3) + 1; + else + return (DP_LINK_BW_2_7 >> 3) + 1; } static void @@ -1259,7 +1253,7 @@ intel_dp_compute_config(struct intel_encoder *encoder, sink_len = intel_dp_sink_rates(intel_dp, &sink_rates); - source_len = intel_dp_source_rates(intel_dp, &source_rates); + source_len = intel_dp_source_rates(dev, &source_rates); supported_len = intel_supported_rates(source_rates, source_len, sink_rates, sink_len, supported_rates); -- cgit v1.2.1 From 2ecae76ad6c6da042dbbcce0084bd6b6c5700845 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 12 Mar 2015 17:10:33 +0200 Subject: drm/i915: Hide the source vs. sink rate handling from intel_dp_compute_config() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit intel_dp_compute_config() only really needs to know the rates supported by both source and sink, so hide the raw source and sink arrays from it. Signed-off-by: Ville Syrjälä Reviewed-by: Sonika Jindal Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_dp.c | 37 +++++++++++++++++++++++-------------- 1 file changed, 23 insertions(+), 14 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 7553f08d87e1..45c35681eb48 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -1195,9 +1195,9 @@ intel_dp_set_clock(struct intel_encoder *encoder, } } -static int intel_supported_rates(const int *source_rates, int source_len, - const int *sink_rates, int sink_len, - int *supported_rates) +static int intersect_rates(const int *source_rates, int source_len, + const int *sink_rates, int sink_len, + int *supported_rates) { int i = 0, j = 0, k = 0; @@ -1216,6 +1216,21 @@ static int intel_supported_rates(const int *source_rates, int source_len, return k; } +static int intel_supported_rates(struct intel_dp *intel_dp, + int *supported_rates) +{ + struct drm_device *dev = intel_dp_to_dev(intel_dp); + const int *source_rates, *sink_rates; + int source_len, sink_len; + + sink_len = intel_dp_sink_rates(intel_dp, &sink_rates); + source_len = intel_dp_source_rates(dev, &source_rates); + + return intersect_rates(source_rates, source_len, + sink_rates, sink_len, + supported_rates); +} + static int rate_to_index(int find, const int *rates) { int i = 0; @@ -1246,17 +1261,10 @@ intel_dp_compute_config(struct intel_encoder *encoder, int max_clock; int bpp, mode_rate; int link_avail, link_clock; - const int *sink_rates; - int supported_rates[8] = {0}; - const int *source_rates; - int source_len, sink_len, supported_len; - - sink_len = intel_dp_sink_rates(intel_dp, &sink_rates); - - source_len = intel_dp_source_rates(dev, &source_rates); + int supported_rates[DP_MAX_SUPPORTED_RATES] = {}; + int supported_len; - supported_len = intel_supported_rates(source_rates, source_len, - sink_rates, sink_len, supported_rates); + supported_len = intel_supported_rates(intel_dp, supported_rates); /* No common link rates between source and sink */ WARN_ON(supported_len <= 0); @@ -1355,7 +1363,8 @@ found: if (INTEL_INFO(dev)->gen >= 9 && intel_dp->supported_rates[0]) { intel_dp->rate_select = - rate_to_index(supported_rates[clock], sink_rates); + rate_to_index(supported_rates[clock], + intel_dp->supported_rates); intel_dp->link_bw = 0; } -- cgit v1.2.1 From 50fec21a7db84a5c40e27c83bd55a5c9b435efbf Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 12 Mar 2015 17:10:34 +0200 Subject: drm/i915: Fix max link rate in intel_dp_mode_valid() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Consider the link rates reported by the sink via DP_SUPPORTED_LINK_RATES when checking modes against the max link rate. Signed-off-by: Ville Syrjälä Reviewed-by: Sonika Jindal Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_dp.c | 15 ++++++++++++++- drivers/gpu/drm/i915/intel_drv.h | 1 + 2 files changed, 15 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 45c35681eb48..5d8ec3a33502 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -206,7 +206,7 @@ intel_dp_mode_valid(struct drm_connector *connector, target_clock = fixed_mode->clock; } - max_link_clock = drm_dp_bw_code_to_link_rate(intel_dp_max_link_bw(intel_dp)); + max_link_clock = intel_dp_max_link_rate(intel_dp); max_lanes = intel_dp_max_lane_count(intel_dp); max_rate = intel_dp_max_data_rate(max_link_clock, max_lanes); @@ -1242,6 +1242,19 @@ static int rate_to_index(int find, const int *rates) return i; } +int +intel_dp_max_link_rate(struct intel_dp *intel_dp) +{ + int rates[DP_MAX_SUPPORTED_RATES] = {}; + int len; + + len = intel_supported_rates(intel_dp, rates); + if (WARN_ON(len <= 0)) + return 162000; + + return rates[rate_to_index(0, rates) - 1]; +} + bool intel_dp_compute_config(struct intel_encoder *encoder, struct intel_crtc_state *pipe_config) diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index 69c8437be611..d5a1f1f46ce5 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -1062,6 +1062,7 @@ void intel_dp_add_properties(struct intel_dp *intel_dp, struct drm_connector *co void intel_dp_mst_suspend(struct drm_device *dev); void intel_dp_mst_resume(struct drm_device *dev); int intel_dp_max_link_bw(struct intel_dp *intel_dp); +int intel_dp_max_link_rate(struct intel_dp *intel_dp); void intel_dp_hot_plug(struct intel_encoder *intel_encoder); void vlv_power_sequencer_reset(struct drm_i915_private *dev_priv); uint32_t intel_dp_pack_aux(const uint8_t *src, int src_bytes); -- cgit v1.2.1 From bc27b7d3f096e88f9b6571965452925fc93f2ca3 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 12 Mar 2015 17:10:35 +0200 Subject: drm/i915: Use DP_LINK_RATE_SET whenever possible MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Drop the gen9 checks from the code and issue DP_LINK_RATE_SET whenever the sink reports to support it. Signed-off-by: Ville Syrjälä Reviewed-by: Sonika Jindal Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_dp.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 5d8ec3a33502..aa3c6781b65c 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -1371,14 +1371,15 @@ found: intel_dp->lane_count = lane_count; - intel_dp->link_bw = - drm_dp_link_rate_to_bw_code(supported_rates[clock]); - - if (INTEL_INFO(dev)->gen >= 9 && intel_dp->supported_rates[0]) { + if (intel_dp->num_supported_rates) { + intel_dp->link_bw = 0; intel_dp->rate_select = rate_to_index(supported_rates[clock], intel_dp->supported_rates); - intel_dp->link_bw = 0; + } else { + intel_dp->link_bw = + drm_dp_link_rate_to_bw_code(supported_rates[clock]); + intel_dp->rate_select = 0; } pipe_config->pipe_bpp = bpp; @@ -3492,7 +3493,7 @@ intel_dp_start_link_train(struct intel_dp *intel_dp) if (drm_dp_enhanced_frame_cap(intel_dp->dpcd)) link_config[1] |= DP_LANE_COUNT_ENHANCED_FRAME_EN; drm_dp_dpcd_write(&intel_dp->aux, DP_LINK_BW_SET, link_config, 2); - if (INTEL_INFO(dev)->gen >= 9 && intel_dp->supported_rates[0]) + if (intel_dp->num_supported_rates) drm_dp_dpcd_write(&intel_dp->aux, DP_LINK_RATE_SET, &intel_dp->rate_select, 1); -- cgit v1.2.1 From ed4e9c1d46980e76254573ccd59efb3e8aec77d2 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 12 Mar 2015 17:10:36 +0200 Subject: drm/i915: Fix MST link rate handling MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Now that intel_dp_max_link_bw() no longer considers the source restrictions we may try to enable MST with 5.4GHz even when the source doesn't support it. To fix that switch the code over to handle the link rate in the same way as the SST code handles it. Signed-off-by: Ville Syrjälä Reviewed-by: Sonika Jindal Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_dp.c | 12 ++++++++---- drivers/gpu/drm/i915/intel_dp_mst.c | 16 +++++++++++++--- drivers/gpu/drm/i915/intel_drv.h | 2 +- 3 files changed, 22 insertions(+), 8 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index aa3c6781b65c..c02e022ace70 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -122,8 +122,8 @@ static void vlv_init_panel_power_sequencer(struct intel_dp *intel_dp); static void vlv_steal_power_sequencer(struct drm_device *dev, enum pipe pipe); -int -intel_dp_max_link_bw(struct intel_dp *intel_dp) +static int +intel_dp_max_link_bw(struct intel_dp *intel_dp) { int max_link_bw = intel_dp->dpcd[DP_MAX_LINK_RATE]; @@ -1255,6 +1255,11 @@ intel_dp_max_link_rate(struct intel_dp *intel_dp) return rates[rate_to_index(0, rates) - 1]; } +int intel_dp_rate_select(struct intel_dp *intel_dp, int rate) +{ + return rate_to_index(rate, intel_dp->supported_rates); +} + bool intel_dp_compute_config(struct intel_encoder *encoder, struct intel_crtc_state *pipe_config) @@ -1374,8 +1379,7 @@ found: if (intel_dp->num_supported_rates) { intel_dp->link_bw = 0; intel_dp->rate_select = - rate_to_index(supported_rates[clock], - intel_dp->supported_rates); + intel_dp_rate_select(intel_dp, supported_rates[clock]); } else { intel_dp->link_bw = drm_dp_link_rate_to_bw_code(supported_rates[clock]); diff --git a/drivers/gpu/drm/i915/intel_dp_mst.c b/drivers/gpu/drm/i915/intel_dp_mst.c index be124928ca14..7e6f12597a6c 100644 --- a/drivers/gpu/drm/i915/intel_dp_mst.c +++ b/drivers/gpu/drm/i915/intel_dp_mst.c @@ -38,7 +38,7 @@ static bool intel_dp_mst_compute_config(struct intel_encoder *encoder, struct intel_dp *intel_dp = &intel_dig_port->dp; struct drm_device *dev = encoder->base.dev; int bpp; - int lane_count, slots; + int lane_count, slots, rate; struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode; struct intel_connector *found = NULL, *intel_connector; int mst_pbn; @@ -52,11 +52,21 @@ static bool intel_dp_mst_compute_config(struct intel_encoder *encoder, * seem to suggest we should do otherwise. */ lane_count = drm_dp_max_lane_count(intel_dp->dpcd); - intel_dp->link_bw = intel_dp_max_link_bw(intel_dp); + + rate = intel_dp_max_link_rate(intel_dp); + + if (intel_dp->num_supported_rates) { + intel_dp->link_bw = 0; + intel_dp->rate_select = intel_dp_rate_select(intel_dp, rate); + } else { + intel_dp->link_bw = drm_dp_link_rate_to_bw_code(rate); + intel_dp->rate_select = 0; + } + intel_dp->lane_count = lane_count; pipe_config->pipe_bpp = 24; - pipe_config->port_clock = drm_dp_bw_code_to_link_rate(intel_dp->link_bw); + pipe_config->port_clock = rate; for_each_intel_connector(dev, intel_connector) { if (intel_connector->new_encoder == encoder) { diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index d5a1f1f46ce5..417419887b54 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -1061,8 +1061,8 @@ void intel_edp_panel_off(struct intel_dp *intel_dp); void intel_dp_add_properties(struct intel_dp *intel_dp, struct drm_connector *connector); void intel_dp_mst_suspend(struct drm_device *dev); void intel_dp_mst_resume(struct drm_device *dev); -int intel_dp_max_link_bw(struct intel_dp *intel_dp); int intel_dp_max_link_rate(struct intel_dp *intel_dp); +int intel_dp_rate_select(struct intel_dp *intel_dp, int rate); void intel_dp_hot_plug(struct intel_encoder *intel_encoder); void vlv_power_sequencer_reset(struct drm_i915_private *dev_priv); uint32_t intel_dp_pack_aux(const uint8_t *src, int src_bytes); -- cgit v1.2.1 From e6bda3e4cb439679946c9378628d906ab32bf590 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 12 Mar 2015 17:10:37 +0200 Subject: drm/i915: Avoid overflowing the DP link rate arrays MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Complain loudly if we ever attempt to overflow the the supported_rates[] array. This should never happen since the sink_rates[] array will always be smaller or of equal size. But should someone change that we want to catch it without scribblign over the stack. Signed-off-by: Ville Syrjälä Reviewed-by: Sonika Jindal Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_dp.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index c02e022ace70..ff0aa6f87679 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -1203,6 +1203,8 @@ static int intersect_rates(const int *source_rates, int source_len, while (i < source_len && j < sink_len) { if (source_rates[i] == sink_rates[j]) { + if (WARN_ON(k >= DP_MAX_SUPPORTED_RATES)) + return k; supported_rates[k] = source_rates[i]; ++k; ++i; -- cgit v1.2.1 From fe51bfb95c996733150c44d21e1c9f4b6322a326 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 12 Mar 2015 17:10:38 +0200 Subject: drm/i915: Add eDP intermediate frequencies for CHV MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit "P1273_DPLL_Programming Spreadsheet.xlsm" lists a boatload of frequencies for eDP. Try to use them all. For now I've decided not to add hardcoded DPLL dividers for these cases since chv_find_best_dpll() works just fine. I've not actually tested any of these since I don't have an eDP 1.4 panel. Signed-off-by: Ville Syrjälä Reviewed-by: Sonika Jindal Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_dp.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index ff0aa6f87679..ce0bdec4b6f9 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -87,6 +87,9 @@ static const struct dp_link_dpll chv_dpll[] = { /* Skylake supports following rates */ static const int gen9_rates[] = { 162000, 216000, 270000, 324000, 432000, 540000 }; +static const int chv_rates[] = { 162000, 202500, 210000, 216000, + 243000, 270000, 324000, 405000, + 420000, 432000, 540000 }; static const int default_rates[] = { 162000, 270000, 540000 }; /** @@ -1148,6 +1151,9 @@ intel_dp_source_rates(struct drm_device *dev, const int **source_rates) if (INTEL_INFO(dev)->gen >= 9) { *source_rates = gen9_rates; return ARRAY_SIZE(gen9_rates); + } else if (IS_CHERRYVIEW(dev)) { + *source_rates = chv_rates; + return ARRAY_SIZE(chv_rates); } *source_rates = default_rates; -- cgit v1.2.1 From 0336400ebeeabb7c7187f1ed1bca04ff2191adf8 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 12 Mar 2015 17:10:39 +0200 Subject: drm/i915: Include the sink/source/supported rates in debug output MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit TODO: Is there an actually nice way to print an array of ints? Signed-off-by: Ville Syrjälä Reviewed-by: Sonika Jindal Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_dp.c | 43 +++++++++++++++++++++++++++++++++++++++++ 1 file changed, 43 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index ce0bdec4b6f9..e2c54912cf11 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -1239,6 +1239,46 @@ static int intel_supported_rates(struct intel_dp *intel_dp, supported_rates); } +static void snprintf_int_array(char *str, size_t len, + const int *array, int nelem) +{ + int i; + + str[0] = '\0'; + + for (i = 0; i < nelem; i++) { + int r = snprintf(str, len, "%d,", array[i]); + if (r >= len) + return; + str += r; + len -= r; + } +} + +static void intel_dp_print_rates(struct intel_dp *intel_dp) +{ + struct drm_device *dev = intel_dp_to_dev(intel_dp); + const int *source_rates, *sink_rates; + int source_len, sink_len, supported_len; + int supported_rates[DP_MAX_SUPPORTED_RATES]; + char str[128]; /* FIXME: too big for stack? */ + + if ((drm_debug & DRM_UT_KMS) == 0) + return; + + source_len = intel_dp_source_rates(dev, &source_rates); + snprintf_int_array(str, sizeof(str), source_rates, source_len); + DRM_DEBUG_KMS("source rates: %s\n", str); + + sink_len = intel_dp_sink_rates(intel_dp, &sink_rates); + snprintf_int_array(str, sizeof(str), sink_rates, sink_len); + DRM_DEBUG_KMS("sink rates: %s\n", str); + + supported_len = intel_supported_rates(intel_dp, supported_rates); + snprintf_int_array(str, sizeof(str), supported_rates, supported_len); + DRM_DEBUG_KMS("supported rates: %s\n", str); +} + static int rate_to_index(int find, const int *rates) { int i = 0; @@ -3775,6 +3815,9 @@ intel_dp_get_dpcd(struct intel_dp *intel_dp) } intel_dp->num_supported_rates = i; } + + intel_dp_print_rates(intel_dp); + if (!(intel_dp->dpcd[DP_DOWNSTREAMPORT_PRESENT] & DP_DWN_STRM_PORT_PRESENT)) return true; /* native DP sink */ -- cgit v1.2.1 From 94ca719ee47f287b11942c899e6b81045311b786 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Fri, 13 Mar 2015 19:40:31 +0200 Subject: drm/i915: Unconfuse DP link rate array names MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit To keep things clear rename the intel_dp->supported_rates[] to intel_dp->sink_rates[], and rename the supported_rates[] name we used elsewhere for the intersection of source and sink rates to common_rates[]. Cc: Sonika Jindal Signed-off-by: Ville Syrjälä Reviewed-by: Sonika Jindal Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_dp.c | 70 ++++++++++++++++++------------------- drivers/gpu/drm/i915/intel_dp_mst.c | 2 +- drivers/gpu/drm/i915/intel_drv.h | 5 +-- 3 files changed, 39 insertions(+), 38 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index e2c54912cf11..5256c064da05 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -1135,9 +1135,9 @@ hsw_dp_set_ddi_pll_sel(struct intel_crtc_state *pipe_config, int link_bw) static int intel_dp_sink_rates(struct intel_dp *intel_dp, const int **sink_rates) { - if (intel_dp->num_supported_rates) { - *sink_rates = intel_dp->supported_rates; - return intel_dp->num_supported_rates; + if (intel_dp->num_sink_rates) { + *sink_rates = intel_dp->sink_rates; + return intel_dp->num_sink_rates; } *sink_rates = default_rates; @@ -1203,7 +1203,7 @@ intel_dp_set_clock(struct intel_encoder *encoder, static int intersect_rates(const int *source_rates, int source_len, const int *sink_rates, int sink_len, - int *supported_rates) + int *common_rates) { int i = 0, j = 0, k = 0; @@ -1211,7 +1211,7 @@ static int intersect_rates(const int *source_rates, int source_len, if (source_rates[i] == sink_rates[j]) { if (WARN_ON(k >= DP_MAX_SUPPORTED_RATES)) return k; - supported_rates[k] = source_rates[i]; + common_rates[k] = source_rates[i]; ++k; ++i; ++j; @@ -1224,8 +1224,8 @@ static int intersect_rates(const int *source_rates, int source_len, return k; } -static int intel_supported_rates(struct intel_dp *intel_dp, - int *supported_rates) +static int intel_dp_common_rates(struct intel_dp *intel_dp, + int *common_rates) { struct drm_device *dev = intel_dp_to_dev(intel_dp); const int *source_rates, *sink_rates; @@ -1236,7 +1236,7 @@ static int intel_supported_rates(struct intel_dp *intel_dp, return intersect_rates(source_rates, source_len, sink_rates, sink_len, - supported_rates); + common_rates); } static void snprintf_int_array(char *str, size_t len, @@ -1259,8 +1259,8 @@ static void intel_dp_print_rates(struct intel_dp *intel_dp) { struct drm_device *dev = intel_dp_to_dev(intel_dp); const int *source_rates, *sink_rates; - int source_len, sink_len, supported_len; - int supported_rates[DP_MAX_SUPPORTED_RATES]; + int source_len, sink_len, common_len; + int common_rates[DP_MAX_SUPPORTED_RATES]; char str[128]; /* FIXME: too big for stack? */ if ((drm_debug & DRM_UT_KMS) == 0) @@ -1274,9 +1274,9 @@ static void intel_dp_print_rates(struct intel_dp *intel_dp) snprintf_int_array(str, sizeof(str), sink_rates, sink_len); DRM_DEBUG_KMS("sink rates: %s\n", str); - supported_len = intel_supported_rates(intel_dp, supported_rates); - snprintf_int_array(str, sizeof(str), supported_rates, supported_len); - DRM_DEBUG_KMS("supported rates: %s\n", str); + common_len = intel_dp_common_rates(intel_dp, common_rates); + snprintf_int_array(str, sizeof(str), common_rates, common_len); + DRM_DEBUG_KMS("common rates: %s\n", str); } static int rate_to_index(int find, const int *rates) @@ -1296,7 +1296,7 @@ intel_dp_max_link_rate(struct intel_dp *intel_dp) int rates[DP_MAX_SUPPORTED_RATES] = {}; int len; - len = intel_supported_rates(intel_dp, rates); + len = intel_dp_common_rates(intel_dp, rates); if (WARN_ON(len <= 0)) return 162000; @@ -1305,7 +1305,7 @@ intel_dp_max_link_rate(struct intel_dp *intel_dp) int intel_dp_rate_select(struct intel_dp *intel_dp, int rate) { - return rate_to_index(rate, intel_dp->supported_rates); + return rate_to_index(rate, intel_dp->sink_rates); } bool @@ -1327,15 +1327,15 @@ intel_dp_compute_config(struct intel_encoder *encoder, int max_clock; int bpp, mode_rate; int link_avail, link_clock; - int supported_rates[DP_MAX_SUPPORTED_RATES] = {}; - int supported_len; + int common_rates[DP_MAX_SUPPORTED_RATES] = {}; + int common_len; - supported_len = intel_supported_rates(intel_dp, supported_rates); + common_len = intel_dp_common_rates(intel_dp, common_rates); /* No common link rates between source and sink */ - WARN_ON(supported_len <= 0); + WARN_ON(common_len <= 0); - max_clock = supported_len - 1; + max_clock = common_len - 1; if (HAS_PCH_SPLIT(dev) && !HAS_DDI(dev) && port != PORT_A) pipe_config->has_pch_encoder = true; @@ -1360,7 +1360,7 @@ intel_dp_compute_config(struct intel_encoder *encoder, DRM_DEBUG_KMS("DP link computation with max lane count %i " "max bw %d pixel clock %iKHz\n", - max_lane_count, supported_rates[max_clock], + max_lane_count, common_rates[max_clock], adjusted_mode->crtc_clock); /* Walk through all bpp values. Luckily they're all nicely spaced with 2 @@ -1393,7 +1393,7 @@ intel_dp_compute_config(struct intel_encoder *encoder, lane_count <= max_lane_count; lane_count <<= 1) { - link_clock = supported_rates[clock]; + link_clock = common_rates[clock]; link_avail = intel_dp_max_data_rate(link_clock, lane_count); @@ -1424,18 +1424,18 @@ found: intel_dp->lane_count = lane_count; - if (intel_dp->num_supported_rates) { + if (intel_dp->num_sink_rates) { intel_dp->link_bw = 0; intel_dp->rate_select = - intel_dp_rate_select(intel_dp, supported_rates[clock]); + intel_dp_rate_select(intel_dp, common_rates[clock]); } else { intel_dp->link_bw = - drm_dp_link_rate_to_bw_code(supported_rates[clock]); + drm_dp_link_rate_to_bw_code(common_rates[clock]); intel_dp->rate_select = 0; } pipe_config->pipe_bpp = bpp; - pipe_config->port_clock = supported_rates[clock]; + pipe_config->port_clock = common_rates[clock]; DRM_DEBUG_KMS("DP link bw %02x lane count %d clock %d bpp %d\n", intel_dp->link_bw, intel_dp->lane_count, @@ -1458,7 +1458,7 @@ found: } if (IS_SKYLAKE(dev) && is_edp(intel_dp)) - skl_edp_set_pll_config(pipe_config, supported_rates[clock]); + skl_edp_set_pll_config(pipe_config, common_rates[clock]); else if (IS_HASWELL(dev) || IS_BROADWELL(dev)) hsw_dp_set_ddi_pll_sel(pipe_config, intel_dp->link_bw); else @@ -3545,7 +3545,7 @@ intel_dp_start_link_train(struct intel_dp *intel_dp) if (drm_dp_enhanced_frame_cap(intel_dp->dpcd)) link_config[1] |= DP_LANE_COUNT_ENHANCED_FRAME_EN; drm_dp_dpcd_write(&intel_dp->aux, DP_LINK_BW_SET, link_config, 2); - if (intel_dp->num_supported_rates) + if (intel_dp->num_sink_rates) drm_dp_dpcd_write(&intel_dp->aux, DP_LINK_RATE_SET, &intel_dp->rate_select, 1); @@ -3797,23 +3797,23 @@ intel_dp_get_dpcd(struct intel_dp *intel_dp) (intel_dp->dpcd[DP_EDP_CONFIGURATION_CAP] & DP_DPCD_DISPLAY_CONTROL_CAPABLE) && (intel_dp_dpcd_read_wake(&intel_dp->aux, DP_EDP_DPCD_REV, &rev, 1) == 1) && (rev >= 0x03)) { /* eDp v1.4 or higher */ - __le16 supported_rates[DP_MAX_SUPPORTED_RATES]; + __le16 sink_rates[DP_MAX_SUPPORTED_RATES]; int i; intel_dp_dpcd_read_wake(&intel_dp->aux, DP_SUPPORTED_LINK_RATES, - supported_rates, - sizeof(supported_rates)); + sink_rates, + sizeof(sink_rates)); - for (i = 0; i < ARRAY_SIZE(supported_rates); i++) { - int val = le16_to_cpu(supported_rates[i]); + for (i = 0; i < ARRAY_SIZE(sink_rates); i++) { + int val = le16_to_cpu(sink_rates[i]); if (val == 0) break; - intel_dp->supported_rates[i] = val * 200; + intel_dp->sink_rates[i] = val * 200; } - intel_dp->num_supported_rates = i; + intel_dp->num_sink_rates = i; } intel_dp_print_rates(intel_dp); diff --git a/drivers/gpu/drm/i915/intel_dp_mst.c b/drivers/gpu/drm/i915/intel_dp_mst.c index 7e6f12597a6c..971163e53e59 100644 --- a/drivers/gpu/drm/i915/intel_dp_mst.c +++ b/drivers/gpu/drm/i915/intel_dp_mst.c @@ -55,7 +55,7 @@ static bool intel_dp_mst_compute_config(struct intel_encoder *encoder, rate = intel_dp_max_link_rate(intel_dp); - if (intel_dp->num_supported_rates) { + if (intel_dp->num_sink_rates) { intel_dp->link_bw = 0; intel_dp->rate_select = intel_dp_rate_select(intel_dp, rate); } else { diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index 417419887b54..a1baaa188b0a 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -627,8 +627,9 @@ struct intel_dp { uint8_t dpcd[DP_RECEIVER_CAP_SIZE]; uint8_t psr_dpcd[EDP_PSR_RECEIVER_CAP_SIZE]; uint8_t downstream_ports[DP_MAX_DOWNSTREAM_PORTS]; - uint8_t num_supported_rates; - int supported_rates[DP_MAX_SUPPORTED_RATES]; + /* sink rates as reported by DP_SUPPORTED_LINK_RATES */ + uint8_t num_sink_rates; + int sink_rates[DP_MAX_SUPPORTED_RATES]; struct drm_dp_aux aux; uint8_t train_set[4]; int panel_power_up_delay; -- cgit v1.2.1 From a1ddefd8f3963cd9edfda86a0affff8e53d4a762 Mon Sep 17 00:00:00 2001 From: Jani Nikula Date: Tue, 17 Mar 2015 17:18:54 +0200 Subject: drm/i915/dp: return number of bytes written for short aux/i2c writes MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Allow for a larger receive data size, and check if the receiver returned the number of bytes written. Without this, we've basically skipped all the unwritten bytes for short writes. Signed-off-by: Jani Nikula Reviewed-by: Ville Syrjälä Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_dp.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 5256c064da05..3967af10f53c 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -951,7 +951,7 @@ intel_dp_aux_transfer(struct drm_dp_aux *aux, struct drm_dp_aux_msg *msg) case DP_AUX_NATIVE_WRITE: case DP_AUX_I2C_WRITE: txsize = msg->size ? HEADER_SIZE + msg->size : BARE_ADDRESS_SIZE; - rxsize = 1; + rxsize = 2; /* 0 or 1 data bytes */ if (WARN_ON(txsize > 20)) return -E2BIG; @@ -962,8 +962,13 @@ intel_dp_aux_transfer(struct drm_dp_aux *aux, struct drm_dp_aux_msg *msg) if (ret > 0) { msg->reply = rxbuf[0] >> 4; - /* Return payload size. */ - ret = msg->size; + if (ret > 1) { + /* Number of bytes written in a short write. */ + ret = clamp_t(int, rxbuf[1], 0, msg->size); + } else { + /* Return payload size. */ + ret = msg->size; + } } break; -- cgit v1.2.1 From dabde5c7320059af62ce0a706ad11c793bad1345 Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Wed, 18 Mar 2015 11:21:58 +0300 Subject: drm/i915: memory leak in __i915_gem_vma_create() In the original code then if WARN_ON(i915_is_ggtt(vm) != !!ggtt_view) was true then we leak "vma". Presumably that doesn't happen often but static checkers complain and this bug is easy to fix. Fixes: c3bbb6f2825d ('drm/i915: Do not use ggtt_view with (aliasing) PPGTT') Signed-off-by: Dan Carpenter Reviewed-by: Jani Nikula Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_gem_gtt.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c index f1b9ea61de4b..cbf013fd6b98 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c @@ -2340,12 +2340,13 @@ __i915_gem_vma_create(struct drm_i915_gem_object *obj, struct i915_address_space *vm, const struct i915_ggtt_view *ggtt_view) { - struct i915_vma *vma = kzalloc(sizeof(*vma), GFP_KERNEL); - if (vma == NULL) - return ERR_PTR(-ENOMEM); + struct i915_vma *vma; if (WARN_ON(i915_is_ggtt(vm) != !!ggtt_view)) return ERR_PTR(-EINVAL); + vma = kzalloc(sizeof(*vma), GFP_KERNEL); + if (vma == NULL) + return ERR_PTR(-ENOMEM); INIT_LIST_HEAD(&vma->vma_link); INIT_LIST_HEAD(&vma->mm_list); -- cgit v1.2.1 From 6fafab76d5189312bf45e71254b5aba5faf7a31b Mon Sep 17 00:00:00 2001 From: Tvrtko Ursulin Date: Tue, 17 Mar 2015 15:36:51 +0000 Subject: drm/i915: Turn on PIN_GLOBAL in i915_gem_object_ggtt_pin This makes the interface consistent to old i915_gem_obj_ggtt_pin. Signed-off-by: Tvrtko Ursulin Cc: Joonas Lahtinen Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_gem.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index cc8672a16da4..979dc42b72f0 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -4292,7 +4292,7 @@ i915_gem_object_ggtt_pin(struct drm_i915_gem_object *obj, return -EINVAL; return i915_gem_object_do_pin(obj, i915_obj_to_ggtt(obj), view, - alignment, flags); + alignment, flags | PIN_GLOBAL); } void -- cgit v1.2.1 From edf4427b8055dc93eb5222d8174b07a75ba24fb5 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Wed, 14 Jan 2015 11:20:56 +0000 Subject: drm/i915: Fallback to using CPU relocations for large batch buffers If the batch buffer is too large to fit into the aperture and we need a GTT mapping for relocations, we currently fail. This only applies to a subset of machines for a subset of environments, quite undesirable. We can simply check after failing to insert the batch into the GTT as to whether we only need a mappable binding for relocation and, if so, we can revert to using a non-mappable binding and an alternate relocation method. However, using relocate_entry_cpu() is excruciatingly slow for large buffers on non-LLC as the entire buffer requires clflushing before and after the relocation handling. Alternatively, we can implement a third relocation method that only clflushes around the relocation entry. This is still slower than updating through the GTT, so we prefer using the GTT where possible, but is orders of magnitude faster as we typically do not have to then clflush the entire buffer. An alternative idea of using a temporary WC mapping of the backing store is promising (it should be faster than using the GTT itself), but requires fairly extensive arch/x86 support - along the lines of kmap_atomic_prof_pfn() (which is not universally implemented even for x86). Testcase: igt/gem_exec_big #pnv,byt Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=88392 Signed-off-by: Chris Wilson [danvet: Add a WARN_ONCE for the impossible reloc case and explain in a short comment why we want to avoid ping-pong.] Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_gem_execbuffer.c | 86 ++++++++++++++++++++++++++---- 1 file changed, 75 insertions(+), 11 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c index dc10bc43864e..9838d11e43ee 100644 --- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c +++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c @@ -251,7 +251,6 @@ static inline int use_cpu_reloc(struct drm_i915_gem_object *obj) { return (HAS_LLC(obj->base.dev) || obj->base.write_domain == I915_GEM_DOMAIN_CPU || - !obj->map_and_fenceable || obj->cache_level != I915_CACHE_NONE); } @@ -337,6 +336,51 @@ relocate_entry_gtt(struct drm_i915_gem_object *obj, return 0; } +static void +clflush_write32(void *addr, uint32_t value) +{ + /* This is not a fast path, so KISS. */ + drm_clflush_virt_range(addr, sizeof(uint32_t)); + *(uint32_t *)addr = value; + drm_clflush_virt_range(addr, sizeof(uint32_t)); +} + +static int +relocate_entry_clflush(struct drm_i915_gem_object *obj, + struct drm_i915_gem_relocation_entry *reloc, + uint64_t target_offset) +{ + struct drm_device *dev = obj->base.dev; + uint32_t page_offset = offset_in_page(reloc->offset); + uint64_t delta = (int)reloc->delta + target_offset; + char *vaddr; + int ret; + + ret = i915_gem_object_set_to_gtt_domain(obj, true); + if (ret) + return ret; + + vaddr = kmap_atomic(i915_gem_object_get_page(obj, + reloc->offset >> PAGE_SHIFT)); + clflush_write32(vaddr + page_offset, lower_32_bits(delta)); + + if (INTEL_INFO(dev)->gen >= 8) { + page_offset = offset_in_page(page_offset + sizeof(uint32_t)); + + if (page_offset == 0) { + kunmap_atomic(vaddr); + vaddr = kmap_atomic(i915_gem_object_get_page(obj, + (reloc->offset + sizeof(uint32_t)) >> PAGE_SHIFT)); + } + + clflush_write32(vaddr + page_offset, upper_32_bits(delta)); + } + + kunmap_atomic(vaddr); + + return 0; +} + static int i915_gem_execbuffer_relocate_entry(struct drm_i915_gem_object *obj, struct eb_vmas *eb, @@ -426,8 +470,14 @@ i915_gem_execbuffer_relocate_entry(struct drm_i915_gem_object *obj, if (use_cpu_reloc(obj)) ret = relocate_entry_cpu(obj, reloc, target_offset); - else + else if (obj->map_and_fenceable) ret = relocate_entry_gtt(obj, reloc, target_offset); + else if (cpu_has_clflush) + ret = relocate_entry_clflush(obj, reloc, target_offset); + else { + WARN_ONCE(1, "Impossible case in relocation handling\n"); + ret = -ENODEV; + } if (ret) return ret; @@ -525,6 +575,12 @@ i915_gem_execbuffer_relocate(struct eb_vmas *eb) return ret; } +static bool only_mappable_for_reloc(unsigned int flags) +{ + return (flags & (EXEC_OBJECT_NEEDS_FENCE | __EXEC_OBJECT_NEEDS_MAP)) == + __EXEC_OBJECT_NEEDS_MAP; +} + static int i915_gem_execbuffer_reserve_vma(struct i915_vma *vma, struct intel_engine_cs *ring, @@ -536,14 +592,21 @@ i915_gem_execbuffer_reserve_vma(struct i915_vma *vma, int ret; flags = 0; - if (entry->flags & __EXEC_OBJECT_NEEDS_MAP) - flags |= PIN_GLOBAL | PIN_MAPPABLE; - if (entry->flags & EXEC_OBJECT_NEEDS_GTT) - flags |= PIN_GLOBAL; - if (entry->flags & __EXEC_OBJECT_NEEDS_BIAS) - flags |= BATCH_OFFSET_BIAS | PIN_OFFSET_BIAS; + if (!drm_mm_node_allocated(&vma->node)) { + if (entry->flags & __EXEC_OBJECT_NEEDS_MAP) + flags |= PIN_GLOBAL | PIN_MAPPABLE; + if (entry->flags & EXEC_OBJECT_NEEDS_GTT) + flags |= PIN_GLOBAL; + if (entry->flags & __EXEC_OBJECT_NEEDS_BIAS) + flags |= BATCH_OFFSET_BIAS | PIN_OFFSET_BIAS; + } ret = i915_gem_object_pin(obj, vma->vm, entry->alignment, flags); + if ((ret == -ENOSPC || ret == -E2BIG) && + only_mappable_for_reloc(entry->flags)) + ret = i915_gem_object_pin(obj, vma->vm, + entry->alignment, + flags & ~(PIN_GLOBAL | PIN_MAPPABLE)); if (ret) return ret; @@ -605,13 +668,14 @@ eb_vma_misplaced(struct i915_vma *vma) vma->node.start & (entry->alignment - 1)) return true; - if (entry->flags & __EXEC_OBJECT_NEEDS_MAP && !obj->map_and_fenceable) - return true; - if (entry->flags & __EXEC_OBJECT_NEEDS_BIAS && vma->node.start < BATCH_OFFSET_BIAS) return true; + /* avoid costly ping-pong once a batch bo ended up non-mappable */ + if (entry->flags & __EXEC_OBJECT_NEEDS_MAP && !obj->map_and_fenceable) + return !only_mappable_for_reloc(entry->flags); + return false; } -- cgit v1.2.1 From aed242ff7ebb697e4dff912bd4dc7ec7192f7581 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Wed, 18 Mar 2015 09:48:21 +0000 Subject: drm/i915: Relax RPS contraints to allows setting minfreq on idle When we idle, we set the GPU frequency to the hardware minimum (not user minimum). We introduce a new variable to distinguish between the different roles, and to allow easy tuning of the idle frequency without impacting over aspects of RPS. Setting the minimum frequency should be a safety blanket as the pcu on the GPU should be power gating itself anyway. However, in order for us to do set the absolute minimum frequency, we need to relax a few of our assertions that we do not exceed the user limits. v2: Add idle_freq v3: Init idle_freq for vlv and add a bunch of WARNs Signed-off-by: Chris Wilson Cc: Deepak S Reviewed-by: Deepak S Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_debugfs.c | 6 +++++ drivers/gpu/drm/i915/i915_drv.h | 1 + drivers/gpu/drm/i915/intel_pm.c | 44 +++++++++++++++++++++++-------------- 3 files changed, 35 insertions(+), 16 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_debugfs.c b/drivers/gpu/drm/i915/i915_debugfs.c index aaf756047a20..007c7d7d8295 100644 --- a/drivers/gpu/drm/i915/i915_debugfs.c +++ b/drivers/gpu/drm/i915/i915_debugfs.c @@ -1200,6 +1200,9 @@ static int i915_frequency_info(struct seq_file *m, void *unused) seq_printf(m, "Max overclocked frequency: %dMHz\n", intel_gpu_freq(dev_priv, dev_priv->rps.max_freq)); + + seq_printf(m, "Idle freq: %d MHz\n", + intel_gpu_freq(dev_priv, dev_priv->rps.idle_freq)); } else if (IS_VALLEYVIEW(dev)) { u32 freq_sts; @@ -1214,6 +1217,9 @@ static int i915_frequency_info(struct seq_file *m, void *unused) seq_printf(m, "min GPU freq: %d MHz\n", intel_gpu_freq(dev_priv, dev_priv->rps.min_freq)); + seq_printf(m, "idle GPU freq: %d MHz\n", + intel_gpu_freq(dev_priv, dev_priv->rps.idle_freq)); + seq_printf(m, "efficient (RPe) frequency: %d MHz\n", intel_gpu_freq(dev_priv, dev_priv->rps.efficient_freq)); diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 81f60b48def2..a06536cfce6d 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -1025,6 +1025,7 @@ struct intel_gen6_power_mgmt { u8 max_freq_softlimit; /* Max frequency permitted by the driver */ u8 max_freq; /* Maximum frequency, RP0 if not overclocking */ u8 min_freq; /* AKA RPn. Minimum frequency */ + u8 idle_freq; /* Frequency to request when we are idle */ u8 efficient_freq; /* AKA RPe. Pre-determined balanced frequency */ u8 rp1_freq; /* "less than" RP0 power/freqency */ u8 rp0_freq; /* Non-overclocked max frequency. */ diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 288c9d24098e..beab305e320d 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -3855,9 +3855,9 @@ static void gen6_set_rps_thresholds(struct drm_i915_private *dev_priv, u8 val) break; } /* Max/min bins are special */ - if (val == dev_priv->rps.min_freq_softlimit) + if (val <= dev_priv->rps.min_freq_softlimit) new_power = LOW_POWER; - if (val == dev_priv->rps.max_freq_softlimit) + if (val >= dev_priv->rps.max_freq_softlimit) new_power = HIGH_POWER; if (new_power == dev_priv->rps.power) return; @@ -3940,8 +3940,8 @@ static void gen6_set_rps(struct drm_device *dev, u8 val) struct drm_i915_private *dev_priv = dev->dev_private; WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock)); - WARN_ON(val > dev_priv->rps.max_freq_softlimit); - WARN_ON(val < dev_priv->rps.min_freq_softlimit); + WARN_ON(val > dev_priv->rps.max_freq); + WARN_ON(val < dev_priv->rps.min_freq); /* min/max delay may still have been modified so be sure to * write the limits value. @@ -3979,8 +3979,8 @@ static void valleyview_set_rps(struct drm_device *dev, u8 val) struct drm_i915_private *dev_priv = dev->dev_private; WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock)); - WARN_ON(val > dev_priv->rps.max_freq_softlimit); - WARN_ON(val < dev_priv->rps.min_freq_softlimit); + WARN_ON(val > dev_priv->rps.max_freq); + WARN_ON(val < dev_priv->rps.min_freq); if (WARN_ONCE(IS_CHERRYVIEW(dev) && (val & 1), "Odd GPU freq value\n")) @@ -4007,10 +4007,11 @@ static void valleyview_set_rps(struct drm_device *dev, u8 val) static void vlv_set_rps_idle(struct drm_i915_private *dev_priv) { struct drm_device *dev = dev_priv->dev; + u32 val = dev_priv->rps.idle_freq; /* CHV and latest VLV don't need to force the gfx clock */ if (IS_CHERRYVIEW(dev) || dev->pdev->revision >= 0xd) { - valleyview_set_rps(dev_priv->dev, dev_priv->rps.min_freq_softlimit); + valleyview_set_rps(dev_priv->dev, val); return; } @@ -4018,7 +4019,7 @@ static void vlv_set_rps_idle(struct drm_i915_private *dev_priv) * When we are idle. Drop to min voltage state. */ - if (dev_priv->rps.cur_freq <= dev_priv->rps.min_freq_softlimit) + if (dev_priv->rps.cur_freq <= val) return; /* Mask turbo interrupt so that they will not come in between */ @@ -4027,10 +4028,9 @@ static void vlv_set_rps_idle(struct drm_i915_private *dev_priv) vlv_force_gfx_clock(dev_priv, true); - dev_priv->rps.cur_freq = dev_priv->rps.min_freq_softlimit; + dev_priv->rps.cur_freq = val; - vlv_punit_write(dev_priv, PUNIT_REG_GPU_FREQ_REQ, - dev_priv->rps.min_freq_softlimit); + vlv_punit_write(dev_priv, PUNIT_REG_GPU_FREQ_REQ, val); if (wait_for(((vlv_punit_read(dev_priv, PUNIT_REG_GPU_FREQ_STS)) & GENFREQSTATUS) == 0, 100)) @@ -4038,8 +4038,7 @@ static void vlv_set_rps_idle(struct drm_i915_private *dev_priv) vlv_force_gfx_clock(dev_priv, false); - I915_WRITE(GEN6_PMINTRMSK, - gen6_rps_pm_mask(dev_priv, dev_priv->rps.cur_freq)); + I915_WRITE(GEN6_PMINTRMSK, gen6_rps_pm_mask(dev_priv, val)); } void gen6_rps_idle(struct drm_i915_private *dev_priv) @@ -4051,7 +4050,7 @@ void gen6_rps_idle(struct drm_i915_private *dev_priv) if (IS_VALLEYVIEW(dev)) vlv_set_rps_idle(dev_priv); else - gen6_set_rps(dev_priv->dev, dev_priv->rps.min_freq_softlimit); + gen6_set_rps(dev_priv->dev, dev_priv->rps.idle_freq); dev_priv->rps.last_adj = 0; } mutex_unlock(&dev_priv->rps.hw_lock); @@ -4209,6 +4208,8 @@ static void gen6_init_rps_frequencies(struct drm_device *dev) dev_priv->rps.max_freq); } + dev_priv->rps.idle_freq = dev_priv->rps.min_freq; + /* Preserve min/max settings in case of re-init */ if (dev_priv->rps.max_freq_softlimit == 0) dev_priv->rps.max_freq_softlimit = dev_priv->rps.max_freq; @@ -4375,7 +4376,7 @@ static void gen8_enable_rps(struct drm_device *dev) /* 6: Ring frequency + overclocking (our driver does this later */ dev_priv->rps.power = HIGH_POWER; /* force a reset */ - gen6_set_rps(dev_priv->dev, dev_priv->rps.min_freq_softlimit); + gen6_set_rps(dev_priv->dev, dev_priv->rps.idle_freq); intel_uncore_forcewake_put(dev_priv, FORCEWAKE_ALL); } @@ -4469,7 +4470,7 @@ static void gen6_enable_rps(struct drm_device *dev) } dev_priv->rps.power = HIGH_POWER; /* force a reset */ - gen6_set_rps(dev_priv->dev, dev_priv->rps.min_freq_softlimit); + gen6_set_rps(dev_priv->dev, dev_priv->rps.idle_freq); rc6vids = 0; ret = sandybridge_pcode_read(dev_priv, GEN6_PCODE_READ_RC6VIDS, &rc6vids); @@ -4834,6 +4835,8 @@ static void valleyview_init_gt_powersave(struct drm_device *dev) intel_gpu_freq(dev_priv, dev_priv->rps.min_freq), dev_priv->rps.min_freq); + dev_priv->rps.idle_freq = dev_priv->rps.min_freq; + /* Preserve min/max settings in case of re-init */ if (dev_priv->rps.max_freq_softlimit == 0) dev_priv->rps.max_freq_softlimit = dev_priv->rps.max_freq; @@ -4909,6 +4912,8 @@ static void cherryview_init_gt_powersave(struct drm_device *dev) dev_priv->rps.min_freq) & 1, "Odd GPU freq values\n"); + dev_priv->rps.idle_freq = dev_priv->rps.min_freq; + /* Preserve min/max settings in case of re-init */ if (dev_priv->rps.max_freq_softlimit == 0) dev_priv->rps.max_freq_softlimit = dev_priv->rps.max_freq; @@ -5686,6 +5691,13 @@ static void intel_gen6_powersave_work(struct work_struct *work) gen6_enable_rps(dev); __gen6_update_ring_freq(dev); } + + WARN_ON(dev_priv->rps.max_freq < dev_priv->rps.min_freq); + WARN_ON(dev_priv->rps.idle_freq > dev_priv->rps.max_freq); + + WARN_ON(dev_priv->rps.efficient_freq < dev_priv->rps.min_freq); + WARN_ON(dev_priv->rps.efficient_freq > dev_priv->rps.max_freq); + dev_priv->rps.enabled = true; gen6_enable_rps_interrupts(dev); -- cgit v1.2.1 From 43cf3bf084ba097463d67e756ff821505bdaa69d Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Wed, 18 Mar 2015 09:48:22 +0000 Subject: drm/i915: Improved w/a for rps on Baytrail MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Rewrite commit 31685c258e0b0ad6aa486c5ec001382cf8a64212 Author: Deepak S Date: Thu Jul 3 17:33:01 2014 -0400 drm/i915/vlv: WA for Turbo and RC6 to work together. Other than code clarity, the major improvement is to disable the extra interrupts generated when idle. However, the reclocking remains rather slow under the new manual regime, in particular it fails to downclock as quickly as desired. The second major improvement is that for certain workloads, like games, we need to combine render+media activity counters as the work of displaying the frame is split across the engines and both need to be taken into account when deciding the global GPU frequency as memory cycles are shared. Signed-off-by: Chris Wilson Cc: Deepak S Cc: Ville Syrjälä Cc: Rodrigo Vivi Cc: Daniel Vetter Reviewed-by: Deepak S Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_irq.c | 155 +++++++++++++---------------------- drivers/gpu/drm/i915/i915_reg.h | 4 +- drivers/gpu/drm/i915/intel_display.c | 2 + drivers/gpu/drm/i915/intel_drv.h | 2 + drivers/gpu/drm/i915/intel_pm.c | 22 ++++- 5 files changed, 81 insertions(+), 104 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c index 49ad5fb82ace..8d8d33d068dd 100644 --- a/drivers/gpu/drm/i915/i915_irq.c +++ b/drivers/gpu/drm/i915/i915_irq.c @@ -997,129 +997,84 @@ static void notify_ring(struct drm_device *dev, wake_up_all(&ring->irq_queue); } -static u32 vlv_c0_residency(struct drm_i915_private *dev_priv, - struct intel_rps_ei *rps_ei) +static void vlv_c0_read(struct drm_i915_private *dev_priv, + struct intel_rps_ei *ei) { - u32 cz_ts, cz_freq_khz; - u32 render_count, media_count; - u32 elapsed_render, elapsed_media, elapsed_time; - u32 residency = 0; - - cz_ts = vlv_punit_read(dev_priv, PUNIT_REG_CZ_TIMESTAMP); - cz_freq_khz = DIV_ROUND_CLOSEST(dev_priv->mem_freq * 1000, 4); - - render_count = I915_READ(VLV_RENDER_C0_COUNT_REG); - media_count = I915_READ(VLV_MEDIA_C0_COUNT_REG); - - if (rps_ei->cz_clock == 0) { - rps_ei->cz_clock = cz_ts; - rps_ei->render_c0 = render_count; - rps_ei->media_c0 = media_count; - - return dev_priv->rps.cur_freq; - } - - elapsed_time = cz_ts - rps_ei->cz_clock; - rps_ei->cz_clock = cz_ts; + ei->cz_clock = vlv_punit_read(dev_priv, PUNIT_REG_CZ_TIMESTAMP); + ei->render_c0 = I915_READ(VLV_RENDER_C0_COUNT); + ei->media_c0 = I915_READ(VLV_MEDIA_C0_COUNT); +} - elapsed_render = render_count - rps_ei->render_c0; - rps_ei->render_c0 = render_count; +static bool vlv_c0_above(struct drm_i915_private *dev_priv, + const struct intel_rps_ei *old, + const struct intel_rps_ei *now, + int threshold) +{ + u64 time, c0; - elapsed_media = media_count - rps_ei->media_c0; - rps_ei->media_c0 = media_count; + if (old->cz_clock == 0) + return false; - /* Convert all the counters into common unit of milli sec */ - elapsed_time /= VLV_CZ_CLOCK_TO_MILLI_SEC; - elapsed_render /= cz_freq_khz; - elapsed_media /= cz_freq_khz; + time = now->cz_clock - old->cz_clock; + time *= threshold * dev_priv->mem_freq; - /* - * Calculate overall C0 residency percentage - * only if elapsed time is non zero + /* Workload can be split between render + media, e.g. SwapBuffers + * being blitted in X after being rendered in mesa. To account for + * this we need to combine both engines into our activity counter. */ - if (elapsed_time) { - residency = - ((max(elapsed_render, elapsed_media) * 100) - / elapsed_time); - } + c0 = now->render_c0 - old->render_c0; + c0 += now->media_c0 - old->media_c0; + c0 *= 100 * VLV_CZ_CLOCK_TO_MILLI_SEC * 4 / 1000; - return residency; + return c0 >= time; } -/** - * vlv_calc_delay_from_C0_counters - Increase/Decrease freq based on GPU - * busy-ness calculated from C0 counters of render & media power wells - * @dev_priv: DRM device private - * - */ -static int vlv_calc_delay_from_C0_counters(struct drm_i915_private *dev_priv) +void gen6_rps_reset_ei(struct drm_i915_private *dev_priv) { - u32 residency_C0_up = 0, residency_C0_down = 0; - int new_delay, adj; - - dev_priv->rps.ei_interrupt_count++; - - WARN_ON(!mutex_is_locked(&dev_priv->rps.hw_lock)); + vlv_c0_read(dev_priv, &dev_priv->rps.down_ei); + dev_priv->rps.up_ei = dev_priv->rps.down_ei; + dev_priv->rps.ei_interrupt_count = 0; +} +static u32 vlv_wa_c0_ei(struct drm_i915_private *dev_priv, u32 pm_iir) +{ + struct intel_rps_ei now; + u32 events = 0; - if (dev_priv->rps.up_ei.cz_clock == 0) { - vlv_c0_residency(dev_priv, &dev_priv->rps.up_ei); - vlv_c0_residency(dev_priv, &dev_priv->rps.down_ei); - return dev_priv->rps.cur_freq; - } + if ((pm_iir & GEN6_PM_RP_UP_EI_EXPIRED) == 0) + return 0; + vlv_c0_read(dev_priv, &now); + if (now.cz_clock == 0) + return 0; /* * To down throttle, C0 residency should be less than down threshold * for continous EI intervals. So calculate down EI counters * once in VLV_INT_COUNT_FOR_DOWN_EI */ - if (dev_priv->rps.ei_interrupt_count == VLV_INT_COUNT_FOR_DOWN_EI) { - + if (++dev_priv->rps.ei_interrupt_count >= VLV_INT_COUNT_FOR_DOWN_EI) { + pm_iir |= GEN6_PM_RP_DOWN_EI_EXPIRED; dev_priv->rps.ei_interrupt_count = 0; - - residency_C0_down = vlv_c0_residency(dev_priv, - &dev_priv->rps.down_ei); - } else { - residency_C0_up = vlv_c0_residency(dev_priv, - &dev_priv->rps.up_ei); } - new_delay = dev_priv->rps.cur_freq; - - adj = dev_priv->rps.last_adj; - /* C0 residency is greater than UP threshold. Increase Frequency */ - if (residency_C0_up >= VLV_RP_UP_EI_THRESHOLD) { - if (adj > 0) - adj *= 2; - else - adj = 1; - - if (dev_priv->rps.cur_freq < dev_priv->rps.max_freq_softlimit) - new_delay = dev_priv->rps.cur_freq + adj; - - /* - * For better performance, jump directly - * to RPe if we're below it. - */ - if (new_delay < dev_priv->rps.efficient_freq) - new_delay = dev_priv->rps.efficient_freq; + if (pm_iir & GEN6_PM_RP_DOWN_EI_EXPIRED) { + if (!vlv_c0_above(dev_priv, + &dev_priv->rps.down_ei, &now, + VLV_RP_DOWN_EI_THRESHOLD)) + events |= GEN6_PM_RP_DOWN_THRESHOLD; + dev_priv->rps.down_ei = now; + } - } else if (!dev_priv->rps.ei_interrupt_count && - (residency_C0_down < VLV_RP_DOWN_EI_THRESHOLD)) { - if (adj < 0) - adj *= 2; - else - adj = -1; - /* - * This means, C0 residency is less than down threshold over - * a period of VLV_INT_COUNT_FOR_DOWN_EI. So, reduce the freq - */ - if (dev_priv->rps.cur_freq > dev_priv->rps.min_freq_softlimit) - new_delay = dev_priv->rps.cur_freq + adj; + if (pm_iir & GEN6_PM_RP_UP_EI_EXPIRED) { + if (vlv_c0_above(dev_priv, + &dev_priv->rps.up_ei, &now, + VLV_RP_UP_EI_THRESHOLD)) + events |= GEN6_PM_RP_UP_THRESHOLD; + dev_priv->rps.up_ei = now; } - return new_delay; + return events; } static void gen6_pm_rps_work(struct work_struct *work) @@ -1149,6 +1104,8 @@ static void gen6_pm_rps_work(struct work_struct *work) mutex_lock(&dev_priv->rps.hw_lock); + pm_iir |= vlv_wa_c0_ei(dev_priv, pm_iir); + adj = dev_priv->rps.last_adj; if (pm_iir & GEN6_PM_RP_UP_THRESHOLD) { if (adj > 0) @@ -1171,8 +1128,6 @@ static void gen6_pm_rps_work(struct work_struct *work) else new_delay = dev_priv->rps.min_freq_softlimit; adj = 0; - } else if (pm_iir & GEN6_PM_RP_UP_EI_EXPIRED) { - new_delay = vlv_calc_delay_from_C0_counters(dev_priv); } else if (pm_iir & GEN6_PM_RP_DOWN_THRESHOLD) { if (adj < 0) adj *= 2; diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index cc8ebabc488d..2d76c566d843 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -6220,8 +6220,8 @@ enum skl_disp_power_wells { #define GEN6_GT_GFX_RC6p 0x13810C #define GEN6_GT_GFX_RC6pp 0x138110 -#define VLV_RENDER_C0_COUNT_REG 0x138118 -#define VLV_MEDIA_C0_COUNT_REG 0x13811C +#define VLV_RENDER_C0_COUNT 0x138118 +#define VLV_MEDIA_C0_COUNT 0x13811C #define GEN6_PCODE_MAILBOX 0x138124 #define GEN6_PCODE_READY (1<<31) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 90b460cf2b57..f1c0295f69e5 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -9201,6 +9201,8 @@ void intel_mark_busy(struct drm_device *dev) intel_runtime_pm_get(dev_priv); i915_update_gfx_val(dev_priv); + if (INTEL_INFO(dev)->gen >= 6) + gen6_rps_busy(dev_priv); dev_priv->mm.busy = true; } diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index a1baaa188b0a..8bb18e507f5f 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -1242,6 +1242,8 @@ void intel_disable_gt_powersave(struct drm_device *dev); void intel_suspend_gt_powersave(struct drm_device *dev); void intel_reset_gt_powersave(struct drm_device *dev); void gen6_update_ring_freq(struct drm_device *dev); +void gen6_rps_busy(struct drm_i915_private *dev_priv); +void gen6_rps_reset_ei(struct drm_i915_private *dev_priv); void gen6_rps_idle(struct drm_i915_private *dev_priv); void gen6_rps_boost(struct drm_i915_private *dev_priv); void ilk_wm_get_hw_state(struct drm_device *dev); diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index beab305e320d..68c9cc252d36 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -4041,6 +4041,18 @@ static void vlv_set_rps_idle(struct drm_i915_private *dev_priv) I915_WRITE(GEN6_PMINTRMSK, gen6_rps_pm_mask(dev_priv, val)); } +void gen6_rps_busy(struct drm_i915_private *dev_priv) +{ + mutex_lock(&dev_priv->rps.hw_lock); + if (dev_priv->rps.enabled) { + if (dev_priv->pm_rps_events & (GEN6_PM_RP_DOWN_EI_EXPIRED | GEN6_PM_RP_UP_EI_EXPIRED)) + gen6_rps_reset_ei(dev_priv); + I915_WRITE(GEN6_PMINTRMSK, + gen6_rps_pm_mask(dev_priv, dev_priv->rps.cur_freq)); + } + mutex_unlock(&dev_priv->rps.hw_lock); +} + void gen6_rps_idle(struct drm_i915_private *dev_priv) { struct drm_device *dev = dev_priv->dev; @@ -4052,15 +4064,21 @@ void gen6_rps_idle(struct drm_i915_private *dev_priv) else gen6_set_rps(dev_priv->dev, dev_priv->rps.idle_freq); dev_priv->rps.last_adj = 0; + I915_WRITE(GEN6_PMINTRMSK, 0xffffffff); } mutex_unlock(&dev_priv->rps.hw_lock); } void gen6_rps_boost(struct drm_i915_private *dev_priv) { + u32 val; + mutex_lock(&dev_priv->rps.hw_lock); - if (dev_priv->rps.enabled) { - intel_set_rps(dev_priv->dev, dev_priv->rps.max_freq_softlimit); + val = dev_priv->rps.max_freq_softlimit; + if (dev_priv->rps.enabled && + dev_priv->mm.busy && + dev_priv->rps.cur_freq < val) { + intel_set_rps(dev_priv->dev, val); dev_priv->rps.last_adj = 0; } mutex_unlock(&dev_priv->rps.hw_lock); -- cgit v1.2.1 From 6f4b12f89c78fd0030aa51ead17eaf234108f60d Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Wed, 18 Mar 2015 09:48:23 +0000 Subject: drm/i915: Use down ei for manual Baytrail RPS calculations Use both up/down manual ei calcuations for symmetry and greater flexibility for reclocking, instead of faking the down interrupt based on a fixed integer number of up interrupts. Signed-off-by: Chris Wilson Reviewed-by: Deepak S Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_drv.h | 2 -- drivers/gpu/drm/i915/i915_irq.c | 15 ++------------- drivers/gpu/drm/i915/i915_reg.h | 1 - drivers/gpu/drm/i915/intel_pm.c | 5 ++--- 4 files changed, 4 insertions(+), 19 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index a06536cfce6d..b156bc30c9c9 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -1031,8 +1031,6 @@ struct intel_gen6_power_mgmt { u8 rp0_freq; /* Non-overclocked max frequency. */ u32 cz_freq; - u32 ei_interrupt_count; - int last_adj; enum { LOW_POWER, BETWEEN, HIGH_POWER } power; diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c index 8d8d33d068dd..6d8340d5a111 100644 --- a/drivers/gpu/drm/i915/i915_irq.c +++ b/drivers/gpu/drm/i915/i915_irq.c @@ -1033,7 +1033,6 @@ void gen6_rps_reset_ei(struct drm_i915_private *dev_priv) { vlv_c0_read(dev_priv, &dev_priv->rps.down_ei); dev_priv->rps.up_ei = dev_priv->rps.down_ei; - dev_priv->rps.ei_interrupt_count = 0; } static u32 vlv_wa_c0_ei(struct drm_i915_private *dev_priv, u32 pm_iir) @@ -1041,23 +1040,13 @@ static u32 vlv_wa_c0_ei(struct drm_i915_private *dev_priv, u32 pm_iir) struct intel_rps_ei now; u32 events = 0; - if ((pm_iir & GEN6_PM_RP_UP_EI_EXPIRED) == 0) + if ((pm_iir & (GEN6_PM_RP_DOWN_EI_EXPIRED | GEN6_PM_RP_UP_EI_EXPIRED)) == 0) return 0; vlv_c0_read(dev_priv, &now); if (now.cz_clock == 0) return 0; - /* - * To down throttle, C0 residency should be less than down threshold - * for continous EI intervals. So calculate down EI counters - * once in VLV_INT_COUNT_FOR_DOWN_EI - */ - if (++dev_priv->rps.ei_interrupt_count >= VLV_INT_COUNT_FOR_DOWN_EI) { - pm_iir |= GEN6_PM_RP_DOWN_EI_EXPIRED; - dev_priv->rps.ei_interrupt_count = 0; - } - if (pm_iir & GEN6_PM_RP_DOWN_EI_EXPIRED) { if (!vlv_c0_above(dev_priv, &dev_priv->rps.down_ei, &now, @@ -4254,7 +4243,7 @@ void intel_irq_init(struct drm_i915_private *dev_priv) /* Let's track the enabled rps events */ if (IS_VALLEYVIEW(dev_priv) && !IS_CHERRYVIEW(dev_priv)) /* WaGsvRC0ResidencyMethod:vlv */ - dev_priv->pm_rps_events = GEN6_PM_RP_UP_EI_EXPIRED; + dev_priv->pm_rps_events = GEN6_PM_RP_DOWN_EI_EXPIRED | GEN6_PM_RP_UP_EI_EXPIRED; else dev_priv->pm_rps_events = GEN6_PM_RPS_EVENTS; diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index 2d76c566d843..5b84ee686f99 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -673,7 +673,6 @@ enum skl_disp_power_wells { #define VLV_CZ_CLOCK_TO_MILLI_SEC 100000 #define VLV_RP_UP_EI_THRESHOLD 90 #define VLV_RP_DOWN_EI_THRESHOLD 70 -#define VLV_INT_COUNT_FOR_DOWN_EI 5 /* vlv2 north clock has */ #define CCK_FUSE_REG 0x8 diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 68c9cc252d36..e18f0fd22cf2 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -3922,11 +3922,10 @@ static u32 gen6_rps_pm_mask(struct drm_i915_private *dev_priv, u8 val) u32 mask = 0; if (val > dev_priv->rps.min_freq_softlimit) - mask |= GEN6_PM_RP_DOWN_THRESHOLD | GEN6_PM_RP_DOWN_TIMEOUT; + mask |= GEN6_PM_RP_DOWN_EI_EXPIRED | GEN6_PM_RP_DOWN_THRESHOLD | GEN6_PM_RP_DOWN_TIMEOUT; if (val < dev_priv->rps.max_freq_softlimit) - mask |= GEN6_PM_RP_UP_THRESHOLD; + mask |= GEN6_PM_RP_UP_EI_EXPIRED | GEN6_PM_RP_UP_THRESHOLD; - mask |= dev_priv->pm_rps_events & (GEN6_PM_RP_DOWN_EI_EXPIRED | GEN6_PM_RP_UP_EI_EXPIRED); mask &= dev_priv->pm_rps_events; return gen6_sanitize_rps_pm_mask(dev_priv, ~mask); -- cgit v1.2.1 From be6a03769504753696033b7595d4dcbc4a16d088 Mon Sep 17 00:00:00 2001 From: Daniel Vetter Date: Wed, 18 Mar 2015 10:46:04 +0100 Subject: drm/i915: Extract i915_gem_shrinker.c Two code changes: - Extract i915_gem_shrinker_init. - Inline i915_gem_object_is_purgeable since we open-code it everywhere else too. This already has the benefit of pulling all the shrinker code together, next patch adds a bit of kerneldoc. Signed-off-by: Daniel Vetter Reviewed-by: Chris Wilson Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/Makefile | 1 + drivers/gpu/drm/i915/i915_drv.h | 17 +- drivers/gpu/drm/i915/i915_gem.c | 274 +---------------------------- drivers/gpu/drm/i915/i915_gem_shrinker.c | 291 +++++++++++++++++++++++++++++++ 4 files changed, 306 insertions(+), 277 deletions(-) create mode 100644 drivers/gpu/drm/i915/i915_gem_shrinker.c diff --git a/drivers/gpu/drm/i915/Makefile b/drivers/gpu/drm/i915/Makefile index d3ebaf204408..a69002e2257d 100644 --- a/drivers/gpu/drm/i915/Makefile +++ b/drivers/gpu/drm/i915/Makefile @@ -28,6 +28,7 @@ i915-y += i915_cmd_parser.o \ i915_gem_execbuffer.o \ i915_gem_gtt.o \ i915_gem.o \ + i915_gem_shrinker.o \ i915_gem_stolen.o \ i915_gem_tiling.o \ i915_gem_userptr.o \ diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index b156bc30c9c9..eb38cd1ef216 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -2600,12 +2600,6 @@ int i915_gem_get_aperture_ioctl(struct drm_device *dev, void *data, int i915_gem_wait_ioctl(struct drm_device *dev, void *data, struct drm_file *file_priv); void i915_gem_load(struct drm_device *dev); -unsigned long i915_gem_shrink(struct drm_i915_private *dev_priv, - long target, - unsigned flags); -#define I915_SHRINK_PURGEABLE 0x1 -#define I915_SHRINK_UNBOUND 0x2 -#define I915_SHRINK_BOUND 0x4 void *i915_gem_object_alloc(struct drm_device *dev); void i915_gem_object_free(struct drm_i915_gem_object *obj); void i915_gem_object_init(struct drm_i915_gem_object *obj, @@ -2953,6 +2947,17 @@ i915_gem_object_create_stolen_for_preallocated(struct drm_device *dev, u32 gtt_offset, u32 size); +/* i915_gem_shrinker.c */ +unsigned long i915_gem_shrink(struct drm_i915_private *dev_priv, + long target, + unsigned flags); +#define I915_SHRINK_PURGEABLE 0x1 +#define I915_SHRINK_UNBOUND 0x2 +#define I915_SHRINK_BOUND 0x4 +unsigned long i915_gem_shrink_all(struct drm_i915_private *dev_priv); +void i915_gem_shrinker_init(struct drm_i915_private *dev_priv); + + /* i915_gem_tiling.c */ static inline bool i915_gem_object_needs_bit17_swizzle(struct drm_i915_gem_object *obj) { diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 979dc42b72f0..092f25cfb8d5 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -1,5 +1,5 @@ /* - * Copyright © 2008 Intel Corporation + * Copyright © 2008-2015 Intel Corporation * * Permission is hereby granted, free of charge, to any person obtaining a * copy of this software and associated documentation files (the "Software"), @@ -32,7 +32,6 @@ #include "i915_vgpu.h" #include "i915_trace.h" #include "intel_drv.h" -#include #include #include #include @@ -53,15 +52,6 @@ static void i915_gem_object_update_fence(struct drm_i915_gem_object *obj, struct drm_i915_fence_reg *fence, bool enable); -static unsigned long i915_gem_shrinker_count(struct shrinker *shrinker, - struct shrink_control *sc); -static unsigned long i915_gem_shrinker_scan(struct shrinker *shrinker, - struct shrink_control *sc); -static int i915_gem_shrinker_oom(struct notifier_block *nb, - unsigned long event, - void *ptr); -static unsigned long i915_gem_shrink_all(struct drm_i915_private *dev_priv); - static bool cpu_cache_is_coherent(struct drm_device *dev, enum i915_cache_level level) { @@ -1936,12 +1926,6 @@ i915_gem_mmap_gtt_ioctl(struct drm_device *dev, void *data, return i915_gem_mmap_gtt(file, dev, args->handle, &args->offset); } -static inline int -i915_gem_object_is_purgeable(struct drm_i915_gem_object *obj) -{ - return obj->madv == I915_MADV_DONTNEED; -} - /* Immediately discard the backing storage */ static void i915_gem_object_truncate(struct drm_i915_gem_object *obj) @@ -2047,85 +2031,6 @@ i915_gem_object_put_pages(struct drm_i915_gem_object *obj) return 0; } -unsigned long -i915_gem_shrink(struct drm_i915_private *dev_priv, - long target, unsigned flags) -{ - const struct { - struct list_head *list; - unsigned int bit; - } phases[] = { - { &dev_priv->mm.unbound_list, I915_SHRINK_UNBOUND }, - { &dev_priv->mm.bound_list, I915_SHRINK_BOUND }, - { NULL, 0 }, - }, *phase; - unsigned long count = 0; - - /* - * As we may completely rewrite the (un)bound list whilst unbinding - * (due to retiring requests) we have to strictly process only - * one element of the list at the time, and recheck the list - * on every iteration. - * - * In particular, we must hold a reference whilst removing the - * object as we may end up waiting for and/or retiring the objects. - * This might release the final reference (held by the active list) - * and result in the object being freed from under us. This is - * similar to the precautions the eviction code must take whilst - * removing objects. - * - * Also note that although these lists do not hold a reference to - * the object we can safely grab one here: The final object - * unreferencing and the bound_list are both protected by the - * dev->struct_mutex and so we won't ever be able to observe an - * object on the bound_list with a reference count equals 0. - */ - for (phase = phases; phase->list; phase++) { - struct list_head still_in_list; - - if ((flags & phase->bit) == 0) - continue; - - INIT_LIST_HEAD(&still_in_list); - while (count < target && !list_empty(phase->list)) { - struct drm_i915_gem_object *obj; - struct i915_vma *vma, *v; - - obj = list_first_entry(phase->list, - typeof(*obj), global_list); - list_move_tail(&obj->global_list, &still_in_list); - - if (flags & I915_SHRINK_PURGEABLE && - !i915_gem_object_is_purgeable(obj)) - continue; - - drm_gem_object_reference(&obj->base); - - /* For the unbound phase, this should be a no-op! */ - list_for_each_entry_safe(vma, v, - &obj->vma_list, vma_link) - if (i915_vma_unbind(vma)) - break; - - if (i915_gem_object_put_pages(obj) == 0) - count += obj->base.size >> PAGE_SHIFT; - - drm_gem_object_unreference(&obj->base); - } - list_splice(&still_in_list, phase->list); - } - - return count; -} - -static unsigned long -i915_gem_shrink_all(struct drm_i915_private *dev_priv) -{ - i915_gem_evict_everything(dev_priv->dev); - return i915_gem_shrink(dev_priv, LONG_MAX, - I915_SHRINK_BOUND | I915_SHRINK_UNBOUND); -} - static int i915_gem_object_get_pages_gtt(struct drm_i915_gem_object *obj) { @@ -4425,7 +4330,7 @@ i915_gem_madvise_ioctl(struct drm_device *dev, void *data, obj->madv = args->madv; /* if the object is no longer attached, discard its backing storage */ - if (i915_gem_object_is_purgeable(obj) && obj->pages == NULL) + if (obj->madv == I915_MADV_DONTNEED && obj->pages == NULL) i915_gem_object_truncate(obj); args->retained = obj->madv != __I915_MADV_PURGED; @@ -5064,13 +4969,7 @@ i915_gem_load(struct drm_device *dev) dev_priv->mm.interruptible = true; - dev_priv->mm.shrinker.scan_objects = i915_gem_shrinker_scan; - dev_priv->mm.shrinker.count_objects = i915_gem_shrinker_count; - dev_priv->mm.shrinker.seeks = DEFAULT_SEEKS; - register_shrinker(&dev_priv->mm.shrinker); - - dev_priv->mm.oom_notifier.notifier_call = i915_gem_shrinker_oom; - register_oom_notifier(&dev_priv->mm.oom_notifier); + i915_gem_shrinker_init(dev_priv); i915_gem_batch_pool_init(dev, &dev_priv->mm.batch_pool); @@ -5162,77 +5061,6 @@ void i915_gem_track_fb(struct drm_i915_gem_object *old, } } -static bool mutex_is_locked_by(struct mutex *mutex, struct task_struct *task) -{ - if (!mutex_is_locked(mutex)) - return false; - -#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_MUTEXES) - return mutex->owner == task; -#else - /* Since UP may be pre-empted, we cannot assume that we own the lock */ - return false; -#endif -} - -static bool i915_gem_shrinker_lock(struct drm_device *dev, bool *unlock) -{ - if (!mutex_trylock(&dev->struct_mutex)) { - if (!mutex_is_locked_by(&dev->struct_mutex, current)) - return false; - - if (to_i915(dev)->mm.shrinker_no_lock_stealing) - return false; - - *unlock = false; - } else - *unlock = true; - - return true; -} - -static int num_vma_bound(struct drm_i915_gem_object *obj) -{ - struct i915_vma *vma; - int count = 0; - - list_for_each_entry(vma, &obj->vma_list, vma_link) - if (drm_mm_node_allocated(&vma->node)) - count++; - - return count; -} - -static unsigned long -i915_gem_shrinker_count(struct shrinker *shrinker, struct shrink_control *sc) -{ - struct drm_i915_private *dev_priv = - container_of(shrinker, struct drm_i915_private, mm.shrinker); - struct drm_device *dev = dev_priv->dev; - struct drm_i915_gem_object *obj; - unsigned long count; - bool unlock; - - if (!i915_gem_shrinker_lock(dev, &unlock)) - return 0; - - count = 0; - list_for_each_entry(obj, &dev_priv->mm.unbound_list, global_list) - if (obj->pages_pin_count == 0) - count += obj->base.size >> PAGE_SHIFT; - - list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list) { - if (!i915_gem_obj_is_pinned(obj) && - obj->pages_pin_count == num_vma_bound(obj)) - count += obj->base.size >> PAGE_SHIFT; - } - - if (unlock) - mutex_unlock(&dev->struct_mutex); - - return count; -} - /* All the new VM stuff */ unsigned long i915_gem_obj_offset(struct drm_i915_gem_object *o, @@ -5333,102 +5161,6 @@ unsigned long i915_gem_obj_size(struct drm_i915_gem_object *o, return 0; } -static unsigned long -i915_gem_shrinker_scan(struct shrinker *shrinker, struct shrink_control *sc) -{ - struct drm_i915_private *dev_priv = - container_of(shrinker, struct drm_i915_private, mm.shrinker); - struct drm_device *dev = dev_priv->dev; - unsigned long freed; - bool unlock; - - if (!i915_gem_shrinker_lock(dev, &unlock)) - return SHRINK_STOP; - - freed = i915_gem_shrink(dev_priv, - sc->nr_to_scan, - I915_SHRINK_BOUND | - I915_SHRINK_UNBOUND | - I915_SHRINK_PURGEABLE); - if (freed < sc->nr_to_scan) - freed += i915_gem_shrink(dev_priv, - sc->nr_to_scan - freed, - I915_SHRINK_BOUND | - I915_SHRINK_UNBOUND); - if (unlock) - mutex_unlock(&dev->struct_mutex); - - return freed; -} - -static int -i915_gem_shrinker_oom(struct notifier_block *nb, unsigned long event, void *ptr) -{ - struct drm_i915_private *dev_priv = - container_of(nb, struct drm_i915_private, mm.oom_notifier); - struct drm_device *dev = dev_priv->dev; - struct drm_i915_gem_object *obj; - unsigned long timeout = msecs_to_jiffies(5000) + 1; - unsigned long pinned, bound, unbound, freed_pages; - bool was_interruptible; - bool unlock; - - while (!i915_gem_shrinker_lock(dev, &unlock) && --timeout) { - schedule_timeout_killable(1); - if (fatal_signal_pending(current)) - return NOTIFY_DONE; - } - if (timeout == 0) { - pr_err("Unable to purge GPU memory due lock contention.\n"); - return NOTIFY_DONE; - } - - was_interruptible = dev_priv->mm.interruptible; - dev_priv->mm.interruptible = false; - - freed_pages = i915_gem_shrink_all(dev_priv); - - dev_priv->mm.interruptible = was_interruptible; - - /* Because we may be allocating inside our own driver, we cannot - * assert that there are no objects with pinned pages that are not - * being pointed to by hardware. - */ - unbound = bound = pinned = 0; - list_for_each_entry(obj, &dev_priv->mm.unbound_list, global_list) { - if (!obj->base.filp) /* not backed by a freeable object */ - continue; - - if (obj->pages_pin_count) - pinned += obj->base.size; - else - unbound += obj->base.size; - } - list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list) { - if (!obj->base.filp) - continue; - - if (obj->pages_pin_count) - pinned += obj->base.size; - else - bound += obj->base.size; - } - - if (unlock) - mutex_unlock(&dev->struct_mutex); - - if (freed_pages || unbound || bound) - pr_info("Purging GPU memory, %lu bytes freed, %lu bytes still pinned.\n", - freed_pages << PAGE_SHIFT, pinned); - if (unbound || bound) - pr_err("%lu and %lu bytes still available in the " - "bound and unbound GPU page lists.\n", - bound, unbound); - - *(unsigned long *)ptr += freed_pages; - return NOTIFY_DONE; -} - bool i915_gem_obj_is_pinned(struct drm_i915_gem_object *obj) { struct i915_vma *vma; diff --git a/drivers/gpu/drm/i915/i915_gem_shrinker.c b/drivers/gpu/drm/i915/i915_gem_shrinker.c new file mode 100644 index 000000000000..9ac78b3d6899 --- /dev/null +++ b/drivers/gpu/drm/i915/i915_gem_shrinker.c @@ -0,0 +1,291 @@ +/* + * Copyright © 2008-2015 Intel Corporation + * + * Permission is hereby granted, free of charge, to any person obtaining a + * copy of this software and associated documentation files (the "Software"), + * to deal in the Software without restriction, including without limitation + * the rights to use, copy, modify, merge, publish, distribute, sublicense, + * and/or sell copies of the Software, and to permit persons to whom the + * Software is furnished to do so, subject to the following conditions: + * + * The above copyright notice and this permission notice (including the next + * paragraph) shall be included in all copies or substantial portions of the + * Software. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR + * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, + * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL + * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER + * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING + * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS + * IN THE SOFTWARE. + * + */ + +#include +#include +#include +#include +#include +#include +#include +#include + +#include "i915_drv.h" +#include "i915_trace.h" + +static bool mutex_is_locked_by(struct mutex *mutex, struct task_struct *task) +{ + if (!mutex_is_locked(mutex)) + return false; + +#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_MUTEXES) + return mutex->owner == task; +#else + /* Since UP may be pre-empted, we cannot assume that we own the lock */ + return false; +#endif +} + +unsigned long +i915_gem_shrink(struct drm_i915_private *dev_priv, + long target, unsigned flags) +{ + const struct { + struct list_head *list; + unsigned int bit; + } phases[] = { + { &dev_priv->mm.unbound_list, I915_SHRINK_UNBOUND }, + { &dev_priv->mm.bound_list, I915_SHRINK_BOUND }, + { NULL, 0 }, + }, *phase; + unsigned long count = 0; + + /* + * As we may completely rewrite the (un)bound list whilst unbinding + * (due to retiring requests) we have to strictly process only + * one element of the list at the time, and recheck the list + * on every iteration. + * + * In particular, we must hold a reference whilst removing the + * object as we may end up waiting for and/or retiring the objects. + * This might release the final reference (held by the active list) + * and result in the object being freed from under us. This is + * similar to the precautions the eviction code must take whilst + * removing objects. + * + * Also note that although these lists do not hold a reference to + * the object we can safely grab one here: The final object + * unreferencing and the bound_list are both protected by the + * dev->struct_mutex and so we won't ever be able to observe an + * object on the bound_list with a reference count equals 0. + */ + for (phase = phases; phase->list; phase++) { + struct list_head still_in_list; + + if ((flags & phase->bit) == 0) + continue; + + INIT_LIST_HEAD(&still_in_list); + while (count < target && !list_empty(phase->list)) { + struct drm_i915_gem_object *obj; + struct i915_vma *vma, *v; + + obj = list_first_entry(phase->list, + typeof(*obj), global_list); + list_move_tail(&obj->global_list, &still_in_list); + + if (flags & I915_SHRINK_PURGEABLE && + obj->madv != I915_MADV_DONTNEED) + continue; + + drm_gem_object_reference(&obj->base); + + /* For the unbound phase, this should be a no-op! */ + list_for_each_entry_safe(vma, v, + &obj->vma_list, vma_link) + if (i915_vma_unbind(vma)) + break; + + if (i915_gem_object_put_pages(obj) == 0) + count += obj->base.size >> PAGE_SHIFT; + + drm_gem_object_unreference(&obj->base); + } + list_splice(&still_in_list, phase->list); + } + + return count; +} + +unsigned long i915_gem_shrink_all(struct drm_i915_private *dev_priv) +{ + i915_gem_evict_everything(dev_priv->dev); + return i915_gem_shrink(dev_priv, LONG_MAX, + I915_SHRINK_BOUND | I915_SHRINK_UNBOUND); +} + +static bool i915_gem_shrinker_lock(struct drm_device *dev, bool *unlock) +{ + if (!mutex_trylock(&dev->struct_mutex)) { + if (!mutex_is_locked_by(&dev->struct_mutex, current)) + return false; + + if (to_i915(dev)->mm.shrinker_no_lock_stealing) + return false; + + *unlock = false; + } else + *unlock = true; + + return true; +} + +static int num_vma_bound(struct drm_i915_gem_object *obj) +{ + struct i915_vma *vma; + int count = 0; + + list_for_each_entry(vma, &obj->vma_list, vma_link) + if (drm_mm_node_allocated(&vma->node)) + count++; + + return count; +} + +static unsigned long +i915_gem_shrinker_count(struct shrinker *shrinker, struct shrink_control *sc) +{ + struct drm_i915_private *dev_priv = + container_of(shrinker, struct drm_i915_private, mm.shrinker); + struct drm_device *dev = dev_priv->dev; + struct drm_i915_gem_object *obj; + unsigned long count; + bool unlock; + + if (!i915_gem_shrinker_lock(dev, &unlock)) + return 0; + + count = 0; + list_for_each_entry(obj, &dev_priv->mm.unbound_list, global_list) + if (obj->pages_pin_count == 0) + count += obj->base.size >> PAGE_SHIFT; + + list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list) { + if (!i915_gem_obj_is_pinned(obj) && + obj->pages_pin_count == num_vma_bound(obj)) + count += obj->base.size >> PAGE_SHIFT; + } + + if (unlock) + mutex_unlock(&dev->struct_mutex); + + return count; +} + +static unsigned long +i915_gem_shrinker_scan(struct shrinker *shrinker, struct shrink_control *sc) +{ + struct drm_i915_private *dev_priv = + container_of(shrinker, struct drm_i915_private, mm.shrinker); + struct drm_device *dev = dev_priv->dev; + unsigned long freed; + bool unlock; + + if (!i915_gem_shrinker_lock(dev, &unlock)) + return SHRINK_STOP; + + freed = i915_gem_shrink(dev_priv, + sc->nr_to_scan, + I915_SHRINK_BOUND | + I915_SHRINK_UNBOUND | + I915_SHRINK_PURGEABLE); + if (freed < sc->nr_to_scan) + freed += i915_gem_shrink(dev_priv, + sc->nr_to_scan - freed, + I915_SHRINK_BOUND | + I915_SHRINK_UNBOUND); + if (unlock) + mutex_unlock(&dev->struct_mutex); + + return freed; +} + +static int +i915_gem_shrinker_oom(struct notifier_block *nb, unsigned long event, void *ptr) +{ + struct drm_i915_private *dev_priv = + container_of(nb, struct drm_i915_private, mm.oom_notifier); + struct drm_device *dev = dev_priv->dev; + struct drm_i915_gem_object *obj; + unsigned long timeout = msecs_to_jiffies(5000) + 1; + unsigned long pinned, bound, unbound, freed_pages; + bool was_interruptible; + bool unlock; + + while (!i915_gem_shrinker_lock(dev, &unlock) && --timeout) { + schedule_timeout_killable(1); + if (fatal_signal_pending(current)) + return NOTIFY_DONE; + } + if (timeout == 0) { + pr_err("Unable to purge GPU memory due lock contention.\n"); + return NOTIFY_DONE; + } + + was_interruptible = dev_priv->mm.interruptible; + dev_priv->mm.interruptible = false; + + freed_pages = i915_gem_shrink_all(dev_priv); + + dev_priv->mm.interruptible = was_interruptible; + + /* Because we may be allocating inside our own driver, we cannot + * assert that there are no objects with pinned pages that are not + * being pointed to by hardware. + */ + unbound = bound = pinned = 0; + list_for_each_entry(obj, &dev_priv->mm.unbound_list, global_list) { + if (!obj->base.filp) /* not backed by a freeable object */ + continue; + + if (obj->pages_pin_count) + pinned += obj->base.size; + else + unbound += obj->base.size; + } + list_for_each_entry(obj, &dev_priv->mm.bound_list, global_list) { + if (!obj->base.filp) + continue; + + if (obj->pages_pin_count) + pinned += obj->base.size; + else + bound += obj->base.size; + } + + if (unlock) + mutex_unlock(&dev->struct_mutex); + + if (freed_pages || unbound || bound) + pr_info("Purging GPU memory, %lu bytes freed, %lu bytes still pinned.\n", + freed_pages << PAGE_SHIFT, pinned); + if (unbound || bound) + pr_err("%lu and %lu bytes still available in the " + "bound and unbound GPU page lists.\n", + bound, unbound); + + *(unsigned long *)ptr += freed_pages; + return NOTIFY_DONE; +} + +void i915_gem_shrinker_init(struct drm_i915_private *dev_priv) +{ + dev_priv->mm.shrinker.scan_objects = i915_gem_shrinker_scan; + dev_priv->mm.shrinker.count_objects = i915_gem_shrinker_count; + dev_priv->mm.shrinker.seeks = DEFAULT_SEEKS; + register_shrinker(&dev_priv->mm.shrinker); + + dev_priv->mm.oom_notifier.notifier_call = i915_gem_shrinker_oom; + register_oom_notifier(&dev_priv->mm.oom_notifier); +} -- cgit v1.2.1 From eb0b44adc08c0be01a027eb009e9cdadc31e65a2 Mon Sep 17 00:00:00 2001 From: Daniel Vetter Date: Wed, 18 Mar 2015 14:47:59 +0100 Subject: drm/i915: kerneldoc for i915_gem_shrinker.c And remove one bogus * from i915_gem_gtt.c since that's not a kerneldoc there. v2: Review from Chris: - Clarify memory space to better distinguish from address space. - Add note that shrink doesn't guarantee the freed memory and that users must fall back to shrink_all. - Explain how pinning ties in with eviction/shrinker. Cc: Chris Wilson Signed-off-by: Daniel Vetter Reviewed-by: Chris Wilson Signed-off-by: Daniel Vetter --- Documentation/DocBook/drm.tmpl | 13 +++++++++- drivers/gpu/drm/i915/i915_gem_evict.c | 4 +++ drivers/gpu/drm/i915/i915_gem_gtt.c | 2 +- drivers/gpu/drm/i915/i915_gem_shrinker.c | 44 ++++++++++++++++++++++++++++++++ 4 files changed, 61 insertions(+), 2 deletions(-) diff --git a/Documentation/DocBook/drm.tmpl b/Documentation/DocBook/drm.tmpl index 7a45775518f6..f4976cd7b32b 100644 --- a/Documentation/DocBook/drm.tmpl +++ b/Documentation/DocBook/drm.tmpl @@ -4184,7 +4184,7 @@ int num_ioctls; Buffer Object Eviction - This section documents the interface function for evicting buffer + This section documents the interface functions for evicting buffer objects to make space available in the virtual gpu address spaces. Note that this is mostly orthogonal to shrinking buffer objects caches, which has the goal to make main memory (shared with the gpu @@ -4192,6 +4192,17 @@ int num_ioctls; !Idrivers/gpu/drm/i915/i915_gem_evict.c + + Buffer Object Memory Shrinking + + This section documents the interface function for shrinking memory + usage of buffer object caches. Shrinking is used to make main memory + available. Note that this is mostly orthogonal to evicting buffer + objects, which has the goal to make space in gpu virtual address + spaces. + +!Idrivers/gpu/drm/i915/i915_gem_shrinker.c + diff --git a/drivers/gpu/drm/i915/i915_gem_evict.c b/drivers/gpu/drm/i915/i915_gem_evict.c index e3a49d94da3a..d09e35ed9c9a 100644 --- a/drivers/gpu/drm/i915/i915_gem_evict.c +++ b/drivers/gpu/drm/i915/i915_gem_evict.c @@ -63,6 +63,10 @@ mark_free(struct i915_vma *vma, struct list_head *unwind) * * This function is used by the object/vma binding code. * + * Since this function is only used to free up virtual address space it only + * ignores pinned vmas, and not object where the backing storage itself is + * pinned. Hence obj->pages_pin_count does not protect against eviction. + * * To clarify: This is for freeing up virtual address space, not for freeing * memory in e.g. the shrinker. */ diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c index cbf013fd6b98..d8ff1a8e9d43 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c @@ -698,7 +698,7 @@ static int gen8_ppgtt_setup_page_tables(struct i915_hw_ppgtt *ppgtt, return 0; } -/** +/* * GEN8 legacy ppgtt programming is accomplished through a max 4 PDP registers * with a net effect resembling a 2-level page table in normal x86 terms. Each * PDP represents 1GB of memory 4 * 512 * 512 * 4096 = 4GB legacy 32b address diff --git a/drivers/gpu/drm/i915/i915_gem_shrinker.c b/drivers/gpu/drm/i915/i915_gem_shrinker.c index 9ac78b3d6899..f7929e769250 100644 --- a/drivers/gpu/drm/i915/i915_gem_shrinker.c +++ b/drivers/gpu/drm/i915/i915_gem_shrinker.c @@ -47,6 +47,30 @@ static bool mutex_is_locked_by(struct mutex *mutex, struct task_struct *task) #endif } +/** + * i915_gem_shrink - Shrink buffer object caches + * @dev_priv: i915 device + * @target: amount of memory to make available, in pages + * @flags: control flags for selecting cache types + * + * This function is the main interface to the shrinker. It will try to release + * up to @target pages of main memory backing storage from buffer objects. + * Selection of the specific caches can be done with @flags. This is e.g. useful + * when purgeable objects should be removed from caches preferentially. + * + * Note that it's not guaranteed that released amount is actually available as + * free system memory - the pages might still be in-used to due to other reasons + * (like cpu mmaps) or the mm core has reused them before we could grab them. + * Therefore code that needs to explicitly shrink buffer objects caches (e.g. to + * avoid deadlocks in memory reclaim) must fall back to i915_gem_shrink_all(). + * + * Also note that any kind of pinning (both per-vma address space pins and + * backing storage pins at the buffer object level) result in the shrinker code + * having to skip the object. + * + * Returns: + * The number of pages of backing storage actually released. + */ unsigned long i915_gem_shrink(struct drm_i915_private *dev_priv, long target, unsigned flags) @@ -118,6 +142,20 @@ i915_gem_shrink(struct drm_i915_private *dev_priv, return count; } +/** + * i915_gem_shrink - Shrink buffer object caches completely + * @dev_priv: i915 device + * + * This is a simple wraper around i915_gem_shrink() to aggressively shrink all + * caches completely. It also first waits for and retires all outstanding + * requests to also be able to release backing storage for active objects. + * + * This should only be used in code to intentionally quiescent the gpu or as a + * last-ditch effort when memory seems to have run out. + * + * Returns: + * The number of pages of backing storage actually released. + */ unsigned long i915_gem_shrink_all(struct drm_i915_private *dev_priv) { i915_gem_evict_everything(dev_priv->dev); @@ -279,6 +317,12 @@ i915_gem_shrinker_oom(struct notifier_block *nb, unsigned long event, void *ptr) return NOTIFY_DONE; } +/** + * i915_gem_shrinker_init - Initialize i915 shrinker + * @dev_priv: i915 device + * + * This function registers and sets up the i915 shrinker and OOM handler. + */ void i915_gem_shrinker_init(struct drm_i915_private *dev_priv) { dev_priv->mm.shrinker.scan_objects = i915_gem_shrinker_scan; -- cgit v1.2.1 From d2d9cbbd224f5fc9254c9952a78bd8cfed3b96f9 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 19 Mar 2015 11:44:06 +0200 Subject: drm/i915: Send out the full AUX address MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit AUX addresses are 20 bits long. Send out the entire address instead of just the low 16 bits. Signed-off-by: Ville Syrjälä Reviewed-by: Jani Nikula Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_dp.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 3967af10f53c..637dd5376c82 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -942,8 +942,9 @@ intel_dp_aux_transfer(struct drm_dp_aux *aux, struct drm_dp_aux_msg *msg) size_t txsize, rxsize; int ret; - txbuf[0] = msg->request << 4; - txbuf[1] = msg->address >> 8; + txbuf[0] = (msg->request << 4) | + ((msg->address >> 16) & 0xf); + txbuf[1] = (msg->address >> 8) & 0xff; txbuf[2] = msg->address & 0xff; txbuf[3] = msg->size - 1; -- cgit v1.2.1 From 07749ef32c4fd60334c2451739460dd1cf600281 Mon Sep 17 00:00:00 2001 From: Michel Thierry Date: Mon, 16 Mar 2015 16:00:54 +0000 Subject: drm/i915: page table generalizations No functional changes, but will improve code clarity and removed some duplicated defines. Signed-off-by: Michel Thierry Reviewed-by: Mika Kuoppala Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_gem_gtt.c | 160 ++++++++++++++++++------------------ drivers/gpu/drm/i915/i915_gem_gtt.h | 28 ++++--- 2 files changed, 96 insertions(+), 92 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c index d8ff1a8e9d43..6f5a0cf77fad 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c @@ -148,11 +148,11 @@ static void ppgtt_bind_vma(struct i915_vma *vma, u32 flags); static void ppgtt_unbind_vma(struct i915_vma *vma); -static inline gen8_gtt_pte_t gen8_pte_encode(dma_addr_t addr, - enum i915_cache_level level, - bool valid) +static inline gen8_pte_t gen8_pte_encode(dma_addr_t addr, + enum i915_cache_level level, + bool valid) { - gen8_gtt_pte_t pte = valid ? _PAGE_PRESENT | _PAGE_RW : 0; + gen8_pte_t pte = valid ? _PAGE_PRESENT | _PAGE_RW : 0; pte |= addr; switch (level) { @@ -170,11 +170,11 @@ static inline gen8_gtt_pte_t gen8_pte_encode(dma_addr_t addr, return pte; } -static inline gen8_ppgtt_pde_t gen8_pde_encode(struct drm_device *dev, - dma_addr_t addr, - enum i915_cache_level level) +static inline gen8_pde_t gen8_pde_encode(struct drm_device *dev, + dma_addr_t addr, + enum i915_cache_level level) { - gen8_ppgtt_pde_t pde = _PAGE_PRESENT | _PAGE_RW; + gen8_pde_t pde = _PAGE_PRESENT | _PAGE_RW; pde |= addr; if (level != I915_CACHE_NONE) pde |= PPAT_CACHED_PDE_INDEX; @@ -183,11 +183,11 @@ static inline gen8_ppgtt_pde_t gen8_pde_encode(struct drm_device *dev, return pde; } -static gen6_gtt_pte_t snb_pte_encode(dma_addr_t addr, - enum i915_cache_level level, - bool valid, u32 unused) +static gen6_pte_t snb_pte_encode(dma_addr_t addr, + enum i915_cache_level level, + bool valid, u32 unused) { - gen6_gtt_pte_t pte = valid ? GEN6_PTE_VALID : 0; + gen6_pte_t pte = valid ? GEN6_PTE_VALID : 0; pte |= GEN6_PTE_ADDR_ENCODE(addr); switch (level) { @@ -205,11 +205,11 @@ static gen6_gtt_pte_t snb_pte_encode(dma_addr_t addr, return pte; } -static gen6_gtt_pte_t ivb_pte_encode(dma_addr_t addr, - enum i915_cache_level level, - bool valid, u32 unused) +static gen6_pte_t ivb_pte_encode(dma_addr_t addr, + enum i915_cache_level level, + bool valid, u32 unused) { - gen6_gtt_pte_t pte = valid ? GEN6_PTE_VALID : 0; + gen6_pte_t pte = valid ? GEN6_PTE_VALID : 0; pte |= GEN6_PTE_ADDR_ENCODE(addr); switch (level) { @@ -229,11 +229,11 @@ static gen6_gtt_pte_t ivb_pte_encode(dma_addr_t addr, return pte; } -static gen6_gtt_pte_t byt_pte_encode(dma_addr_t addr, - enum i915_cache_level level, - bool valid, u32 flags) +static gen6_pte_t byt_pte_encode(dma_addr_t addr, + enum i915_cache_level level, + bool valid, u32 flags) { - gen6_gtt_pte_t pte = valid ? GEN6_PTE_VALID : 0; + gen6_pte_t pte = valid ? GEN6_PTE_VALID : 0; pte |= GEN6_PTE_ADDR_ENCODE(addr); if (!(flags & PTE_READ_ONLY)) @@ -245,11 +245,11 @@ static gen6_gtt_pte_t byt_pte_encode(dma_addr_t addr, return pte; } -static gen6_gtt_pte_t hsw_pte_encode(dma_addr_t addr, - enum i915_cache_level level, - bool valid, u32 unused) +static gen6_pte_t hsw_pte_encode(dma_addr_t addr, + enum i915_cache_level level, + bool valid, u32 unused) { - gen6_gtt_pte_t pte = valid ? GEN6_PTE_VALID : 0; + gen6_pte_t pte = valid ? GEN6_PTE_VALID : 0; pte |= HSW_PTE_ADDR_ENCODE(addr); if (level != I915_CACHE_NONE) @@ -258,11 +258,11 @@ static gen6_gtt_pte_t hsw_pte_encode(dma_addr_t addr, return pte; } -static gen6_gtt_pte_t iris_pte_encode(dma_addr_t addr, - enum i915_cache_level level, - bool valid, u32 unused) +static gen6_pte_t iris_pte_encode(dma_addr_t addr, + enum i915_cache_level level, + bool valid, u32 unused) { - gen6_gtt_pte_t pte = valid ? GEN6_PTE_VALID : 0; + gen6_pte_t pte = valid ? GEN6_PTE_VALID : 0; pte |= HSW_PTE_ADDR_ENCODE(addr); switch (level) { @@ -324,7 +324,7 @@ static int alloc_pt_range(struct i915_page_directory_entry *pd, uint16_t pde, si int i, ret; /* 512 is the max page tables per page_directory on any platform. */ - if (WARN_ON(pde + count > GEN6_PPGTT_PD_ENTRIES)) + if (WARN_ON(pde + count > I915_PDES)) return -EINVAL; for (i = pde; i < pde + count; i++) { @@ -402,7 +402,7 @@ static int gen8_mm_switch(struct i915_hw_ppgtt *ppgtt, int i, ret; /* bit of a hack to find the actual last used pd */ - int used_pd = ppgtt->num_pd_entries / GEN8_PDES_PER_PAGE; + int used_pd = ppgtt->num_pd_entries / I915_PDES; for (i = used_pd - 1; i >= 0; i--) { dma_addr_t addr = ppgtt->pdp.page_directory[i]->daddr; @@ -421,7 +421,7 @@ static void gen8_ppgtt_clear_range(struct i915_address_space *vm, { struct i915_hw_ppgtt *ppgtt = container_of(vm, struct i915_hw_ppgtt, base); - gen8_gtt_pte_t *pt_vaddr, scratch_pte; + gen8_pte_t *pt_vaddr, scratch_pte; unsigned pdpe = start >> GEN8_PDPE_SHIFT & GEN8_PDPE_MASK; unsigned pde = start >> GEN8_PDE_SHIFT & GEN8_PDE_MASK; unsigned pte = start >> GEN8_PTE_SHIFT & GEN8_PTE_MASK; @@ -452,8 +452,8 @@ static void gen8_ppgtt_clear_range(struct i915_address_space *vm, page_table = pt->page; last_pte = pte + num_entries; - if (last_pte > GEN8_PTES_PER_PAGE) - last_pte = GEN8_PTES_PER_PAGE; + if (last_pte > GEN8_PTES) + last_pte = GEN8_PTES; pt_vaddr = kmap_atomic(page_table); @@ -467,7 +467,7 @@ static void gen8_ppgtt_clear_range(struct i915_address_space *vm, kunmap_atomic(pt_vaddr); pte = 0; - if (++pde == GEN8_PDES_PER_PAGE) { + if (++pde == I915_PDES) { pdpe++; pde = 0; } @@ -481,7 +481,7 @@ static void gen8_ppgtt_insert_entries(struct i915_address_space *vm, { struct i915_hw_ppgtt *ppgtt = container_of(vm, struct i915_hw_ppgtt, base); - gen8_gtt_pte_t *pt_vaddr; + gen8_pte_t *pt_vaddr; unsigned pdpe = start >> GEN8_PDPE_SHIFT & GEN8_PDPE_MASK; unsigned pde = start >> GEN8_PDE_SHIFT & GEN8_PDE_MASK; unsigned pte = start >> GEN8_PTE_SHIFT & GEN8_PTE_MASK; @@ -504,12 +504,12 @@ static void gen8_ppgtt_insert_entries(struct i915_address_space *vm, pt_vaddr[pte] = gen8_pte_encode(sg_page_iter_dma_address(&sg_iter), cache_level, true); - if (++pte == GEN8_PTES_PER_PAGE) { + if (++pte == GEN8_PTES) { if (!HAS_LLC(ppgtt->base.dev)) drm_clflush_virt_range(pt_vaddr, PAGE_SIZE); kunmap_atomic(pt_vaddr); pt_vaddr = NULL; - if (++pde == GEN8_PDES_PER_PAGE) { + if (++pde == I915_PDES) { pdpe++; pde = 0; } @@ -530,7 +530,7 @@ static void gen8_free_page_tables(struct i915_page_directory_entry *pd, struct d if (!pd->page) return; - for (i = 0; i < GEN8_PDES_PER_PAGE; i++) { + for (i = 0; i < I915_PDES; i++) { if (WARN_ON(!pd->page_table[i])) continue; @@ -566,7 +566,7 @@ static void gen8_ppgtt_unmap_pages(struct i915_hw_ppgtt *ppgtt) pci_unmap_page(hwdev, ppgtt->pdp.page_directory[i]->daddr, PAGE_SIZE, PCI_DMA_BIDIRECTIONAL); - for (j = 0; j < GEN8_PDES_PER_PAGE; j++) { + for (j = 0; j < I915_PDES; j++) { struct i915_page_directory_entry *pd = ppgtt->pdp.page_directory[i]; struct i915_page_table_entry *pt; dma_addr_t addr; @@ -599,7 +599,7 @@ static int gen8_ppgtt_allocate_page_tables(struct i915_hw_ppgtt *ppgtt) for (i = 0; i < ppgtt->num_pd_pages; i++) { ret = alloc_pt_range(ppgtt->pdp.page_directory[i], - 0, GEN8_PDES_PER_PAGE, ppgtt->base.dev); + 0, I915_PDES, ppgtt->base.dev); if (ret) goto unwind_out; } @@ -649,7 +649,7 @@ static int gen8_ppgtt_alloc(struct i915_hw_ppgtt *ppgtt, if (ret) goto err_out; - ppgtt->num_pd_entries = max_pdp * GEN8_PDES_PER_PAGE; + ppgtt->num_pd_entries = max_pdp * I915_PDES; return 0; @@ -711,7 +711,7 @@ static int gen8_ppgtt_setup_page_tables(struct i915_hw_ppgtt *ppgtt, static int gen8_ppgtt_init(struct i915_hw_ppgtt *ppgtt, uint64_t size) { const int max_pdp = DIV_ROUND_UP(size, 1 << 30); - const int min_pt_pages = GEN8_PDES_PER_PAGE * max_pdp; + const int min_pt_pages = I915_PDES * max_pdp; int i, j, ret; if (size % (1<<30)) @@ -734,7 +734,7 @@ static int gen8_ppgtt_init(struct i915_hw_ppgtt *ppgtt, uint64_t size) if (ret) goto bail; - for (j = 0; j < GEN8_PDES_PER_PAGE; j++) { + for (j = 0; j < I915_PDES; j++) { ret = gen8_ppgtt_setup_page_tables(ppgtt, i, j); if (ret) goto bail; @@ -751,9 +751,9 @@ static int gen8_ppgtt_init(struct i915_hw_ppgtt *ppgtt, uint64_t size) */ for (i = 0; i < GEN8_LEGACY_PDPES; i++) { struct i915_page_directory_entry *pd = ppgtt->pdp.page_directory[i]; - gen8_ppgtt_pde_t *pd_vaddr; + gen8_pde_t *pd_vaddr; pd_vaddr = kmap_atomic(ppgtt->pdp.page_directory[i]->page); - for (j = 0; j < GEN8_PDES_PER_PAGE; j++) { + for (j = 0; j < I915_PDES; j++) { struct i915_page_table_entry *pt = pd->page_table[j]; dma_addr_t addr = pt->daddr; pd_vaddr[j] = gen8_pde_encode(ppgtt->base.dev, addr, @@ -771,11 +771,11 @@ static int gen8_ppgtt_init(struct i915_hw_ppgtt *ppgtt, uint64_t size) ppgtt->base.start = 0; /* This is the area that we advertise as usable for the caller */ - ppgtt->base.total = max_pdp * GEN8_PDES_PER_PAGE * GEN8_PTES_PER_PAGE * PAGE_SIZE; + ppgtt->base.total = max_pdp * I915_PDES * GEN8_PTES * PAGE_SIZE; /* Set all ptes to a valid scratch page. Also above requested space */ ppgtt->base.clear_range(&ppgtt->base, 0, - ppgtt->num_pd_pages * GEN8_PTES_PER_PAGE * PAGE_SIZE, + ppgtt->num_pd_pages * GEN8_PTES * PAGE_SIZE, true); DRM_DEBUG_DRIVER("Allocated %d pages for page directories (%d wasted)\n", @@ -795,22 +795,22 @@ static void gen6_dump_ppgtt(struct i915_hw_ppgtt *ppgtt, struct seq_file *m) { struct drm_i915_private *dev_priv = ppgtt->base.dev->dev_private; struct i915_address_space *vm = &ppgtt->base; - gen6_gtt_pte_t __iomem *pd_addr; - gen6_gtt_pte_t scratch_pte; + gen6_pte_t __iomem *pd_addr; + gen6_pte_t scratch_pte; uint32_t pd_entry; int pte, pde; scratch_pte = vm->pte_encode(vm->scratch.addr, I915_CACHE_LLC, true, 0); - pd_addr = (gen6_gtt_pte_t __iomem *)dev_priv->gtt.gsm + - ppgtt->pd.pd_offset / sizeof(gen6_gtt_pte_t); + pd_addr = (gen6_pte_t __iomem *)dev_priv->gtt.gsm + + ppgtt->pd.pd_offset / sizeof(gen6_pte_t); seq_printf(m, " VM %p (pd_offset %x-%x):\n", vm, ppgtt->pd.pd_offset, ppgtt->pd.pd_offset + ppgtt->num_pd_entries); for (pde = 0; pde < ppgtt->num_pd_entries; pde++) { u32 expected; - gen6_gtt_pte_t *pt_vaddr; + gen6_pte_t *pt_vaddr; dma_addr_t pt_addr = ppgtt->pd.page_table[pde]->daddr; pd_entry = readl(pd_addr + pde); expected = (GEN6_PDE_ADDR_ENCODE(pt_addr) | GEN6_PDE_VALID); @@ -823,9 +823,9 @@ static void gen6_dump_ppgtt(struct i915_hw_ppgtt *ppgtt, struct seq_file *m) seq_printf(m, "\tPDE: %x\n", pd_entry); pt_vaddr = kmap_atomic(ppgtt->pd.page_table[pde]->page); - for (pte = 0; pte < I915_PPGTT_PT_ENTRIES; pte+=4) { + for (pte = 0; pte < GEN6_PTES; pte+=4) { unsigned long va = - (pde * PAGE_SIZE * I915_PPGTT_PT_ENTRIES) + + (pde * PAGE_SIZE * GEN6_PTES) + (pte * PAGE_SIZE); int i; bool found = false; @@ -851,13 +851,13 @@ static void gen6_dump_ppgtt(struct i915_hw_ppgtt *ppgtt, struct seq_file *m) static void gen6_write_pdes(struct i915_hw_ppgtt *ppgtt) { struct drm_i915_private *dev_priv = ppgtt->base.dev->dev_private; - gen6_gtt_pte_t __iomem *pd_addr; + gen6_pte_t __iomem *pd_addr; uint32_t pd_entry; int i; WARN_ON(ppgtt->pd.pd_offset & 0x3f); - pd_addr = (gen6_gtt_pte_t __iomem*)dev_priv->gtt.gsm + - ppgtt->pd.pd_offset / sizeof(gen6_gtt_pte_t); + pd_addr = (gen6_pte_t __iomem*)dev_priv->gtt.gsm + + ppgtt->pd.pd_offset / sizeof(gen6_pte_t); for (i = 0; i < ppgtt->num_pd_entries; i++) { dma_addr_t pt_addr; @@ -1023,19 +1023,19 @@ static void gen6_ppgtt_clear_range(struct i915_address_space *vm, { struct i915_hw_ppgtt *ppgtt = container_of(vm, struct i915_hw_ppgtt, base); - gen6_gtt_pte_t *pt_vaddr, scratch_pte; + gen6_pte_t *pt_vaddr, scratch_pte; unsigned first_entry = start >> PAGE_SHIFT; unsigned num_entries = length >> PAGE_SHIFT; - unsigned act_pt = first_entry / I915_PPGTT_PT_ENTRIES; - unsigned first_pte = first_entry % I915_PPGTT_PT_ENTRIES; + unsigned act_pt = first_entry / GEN6_PTES; + unsigned first_pte = first_entry % GEN6_PTES; unsigned last_pte, i; scratch_pte = vm->pte_encode(vm->scratch.addr, I915_CACHE_LLC, true, 0); while (num_entries) { last_pte = first_pte + num_entries; - if (last_pte > I915_PPGTT_PT_ENTRIES) - last_pte = I915_PPGTT_PT_ENTRIES; + if (last_pte > GEN6_PTES) + last_pte = GEN6_PTES; pt_vaddr = kmap_atomic(ppgtt->pd.page_table[act_pt]->page); @@ -1057,10 +1057,10 @@ static void gen6_ppgtt_insert_entries(struct i915_address_space *vm, { struct i915_hw_ppgtt *ppgtt = container_of(vm, struct i915_hw_ppgtt, base); - gen6_gtt_pte_t *pt_vaddr; + gen6_pte_t *pt_vaddr; unsigned first_entry = start >> PAGE_SHIFT; - unsigned act_pt = first_entry / I915_PPGTT_PT_ENTRIES; - unsigned act_pte = first_entry % I915_PPGTT_PT_ENTRIES; + unsigned act_pt = first_entry / GEN6_PTES; + unsigned act_pte = first_entry % GEN6_PTES; struct sg_page_iter sg_iter; pt_vaddr = NULL; @@ -1072,7 +1072,7 @@ static void gen6_ppgtt_insert_entries(struct i915_address_space *vm, vm->pte_encode(sg_page_iter_dma_address(&sg_iter), cache_level, true, flags); - if (++act_pte == I915_PPGTT_PT_ENTRIES) { + if (++act_pte == GEN6_PTES) { kunmap_atomic(pt_vaddr); pt_vaddr = NULL; act_pt++; @@ -1151,7 +1151,7 @@ alloc: if (ppgtt->node.start < dev_priv->gtt.mappable_end) DRM_DEBUG("Forced to use aperture for PDEs\n"); - ppgtt->num_pd_entries = GEN6_PPGTT_PD_ENTRIES; + ppgtt->num_pd_entries = I915_PDES; return 0; } @@ -1231,11 +1231,11 @@ static int gen6_ppgtt_init(struct i915_hw_ppgtt *ppgtt) ppgtt->base.insert_entries = gen6_ppgtt_insert_entries; ppgtt->base.cleanup = gen6_ppgtt_cleanup; ppgtt->base.start = 0; - ppgtt->base.total = ppgtt->num_pd_entries * I915_PPGTT_PT_ENTRIES * PAGE_SIZE; + ppgtt->base.total = ppgtt->num_pd_entries * GEN6_PTES * PAGE_SIZE; ppgtt->debug_dump = gen6_dump_ppgtt; ppgtt->pd.pd_offset = - ppgtt->node.start / PAGE_SIZE * sizeof(gen6_gtt_pte_t); + ppgtt->node.start / PAGE_SIZE * sizeof(gen6_pte_t); ppgtt->base.clear_range(&ppgtt->base, 0, ppgtt->base.total, true); @@ -1541,7 +1541,7 @@ int i915_gem_gtt_prepare_object(struct drm_i915_gem_object *obj) return 0; } -static inline void gen8_set_pte(void __iomem *addr, gen8_gtt_pte_t pte) +static inline void gen8_set_pte(void __iomem *addr, gen8_pte_t pte) { #ifdef writeq writeq(pte, addr); @@ -1558,8 +1558,8 @@ static void gen8_ggtt_insert_entries(struct i915_address_space *vm, { struct drm_i915_private *dev_priv = vm->dev->dev_private; unsigned first_entry = start >> PAGE_SHIFT; - gen8_gtt_pte_t __iomem *gtt_entries = - (gen8_gtt_pte_t __iomem *)dev_priv->gtt.gsm + first_entry; + gen8_pte_t __iomem *gtt_entries = + (gen8_pte_t __iomem *)dev_priv->gtt.gsm + first_entry; int i = 0; struct sg_page_iter sg_iter; dma_addr_t addr = 0; /* shut up gcc */ @@ -1604,8 +1604,8 @@ static void gen6_ggtt_insert_entries(struct i915_address_space *vm, { struct drm_i915_private *dev_priv = vm->dev->dev_private; unsigned first_entry = start >> PAGE_SHIFT; - gen6_gtt_pte_t __iomem *gtt_entries = - (gen6_gtt_pte_t __iomem *)dev_priv->gtt.gsm + first_entry; + gen6_pte_t __iomem *gtt_entries = + (gen6_pte_t __iomem *)dev_priv->gtt.gsm + first_entry; int i = 0; struct sg_page_iter sg_iter; dma_addr_t addr = 0; @@ -1643,8 +1643,8 @@ static void gen8_ggtt_clear_range(struct i915_address_space *vm, struct drm_i915_private *dev_priv = vm->dev->dev_private; unsigned first_entry = start >> PAGE_SHIFT; unsigned num_entries = length >> PAGE_SHIFT; - gen8_gtt_pte_t scratch_pte, __iomem *gtt_base = - (gen8_gtt_pte_t __iomem *) dev_priv->gtt.gsm + first_entry; + gen8_pte_t scratch_pte, __iomem *gtt_base = + (gen8_pte_t __iomem *) dev_priv->gtt.gsm + first_entry; const int max_entries = gtt_total_entries(dev_priv->gtt) - first_entry; int i; @@ -1669,8 +1669,8 @@ static void gen6_ggtt_clear_range(struct i915_address_space *vm, struct drm_i915_private *dev_priv = vm->dev->dev_private; unsigned first_entry = start >> PAGE_SHIFT; unsigned num_entries = length >> PAGE_SHIFT; - gen6_gtt_pte_t scratch_pte, __iomem *gtt_base = - (gen6_gtt_pte_t __iomem *) dev_priv->gtt.gsm + first_entry; + gen6_pte_t scratch_pte, __iomem *gtt_base = + (gen6_pte_t __iomem *) dev_priv->gtt.gsm + first_entry; const int max_entries = gtt_total_entries(dev_priv->gtt) - first_entry; int i; @@ -2185,7 +2185,7 @@ static int gen8_gmch_probe(struct drm_device *dev, gtt_size = gen8_get_total_gtt_size(snb_gmch_ctl); } - *gtt_total = (gtt_size / sizeof(gen8_gtt_pte_t)) << PAGE_SHIFT; + *gtt_total = (gtt_size / sizeof(gen8_pte_t)) << PAGE_SHIFT; if (IS_CHERRYVIEW(dev)) chv_setup_private_ppat(dev_priv); @@ -2230,7 +2230,7 @@ static int gen6_gmch_probe(struct drm_device *dev, *stolen = gen6_get_stolen_size(snb_gmch_ctl); gtt_size = gen6_get_total_gtt_size(snb_gmch_ctl); - *gtt_total = (gtt_size / sizeof(gen6_gtt_pte_t)) << PAGE_SHIFT; + *gtt_total = (gtt_size / sizeof(gen6_pte_t)) << PAGE_SHIFT; ret = ggtt_probe_common(dev, gtt_size); diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.h b/drivers/gpu/drm/i915/i915_gem_gtt.h index c9e93f5070bc..5ca7c5eff88b 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.h +++ b/drivers/gpu/drm/i915/i915_gem_gtt.h @@ -36,13 +36,13 @@ struct drm_i915_file_private; -typedef uint32_t gen6_gtt_pte_t; -typedef uint64_t gen8_gtt_pte_t; -typedef gen8_gtt_pte_t gen8_ppgtt_pde_t; +typedef uint32_t gen6_pte_t; +typedef uint64_t gen8_pte_t; +typedef uint64_t gen8_pde_t; #define gtt_total_entries(gtt) ((gtt).base.total >> PAGE_SHIFT) -#define I915_PPGTT_PT_ENTRIES (PAGE_SIZE / sizeof(gen6_gtt_pte_t)) + /* gen6-hsw has bit 11-4 for physical addr bit 39-32 */ #define GEN6_GTT_ADDR_ENCODE(addr) ((addr) | (((addr) >> 28) & 0xff0)) #define GEN6_PTE_ADDR_ENCODE(addr) GEN6_GTT_ADDR_ENCODE(addr) @@ -51,8 +51,13 @@ typedef gen8_gtt_pte_t gen8_ppgtt_pde_t; #define GEN6_PTE_UNCACHED (1 << 1) #define GEN6_PTE_VALID (1 << 0) -#define GEN6_PPGTT_PD_ENTRIES 512 -#define GEN6_PD_SIZE (GEN6_PPGTT_PD_ENTRIES * PAGE_SIZE) +#define I915_PTES(pte_len) (PAGE_SIZE / (pte_len)) +#define I915_PTE_MASK(pte_len) (I915_PTES(pte_len) - 1) +#define I915_PDES 512 +#define I915_PDE_MASK (I915_PDES - 1) + +#define GEN6_PTES I915_PTES(sizeof(gen6_pte_t)) +#define GEN6_PD_SIZE (I915_PDES * PAGE_SIZE) #define GEN6_PD_ALIGN (PAGE_SIZE * 16) #define GEN6_PDE_VALID (1 << 0) @@ -89,8 +94,7 @@ typedef gen8_gtt_pte_t gen8_ppgtt_pde_t; #define GEN8_PTE_SHIFT 12 #define GEN8_PTE_MASK 0x1ff #define GEN8_LEGACY_PDPES 4 -#define GEN8_PTES_PER_PAGE (PAGE_SIZE / sizeof(gen8_gtt_pte_t)) -#define GEN8_PDES_PER_PAGE (PAGE_SIZE / sizeof(gen8_ppgtt_pde_t)) +#define GEN8_PTES I915_PTES(sizeof(gen8_pte_t)) #define PPAT_UNCACHED_INDEX (_PAGE_PWT | _PAGE_PCD) #define PPAT_CACHED_PDE_INDEX 0 /* WB LLC */ @@ -199,7 +203,7 @@ struct i915_page_directory_entry { dma_addr_t daddr; }; - struct i915_page_table_entry *page_table[GEN6_PPGTT_PD_ENTRIES]; /* PDEs */ + struct i915_page_table_entry *page_table[I915_PDES]; /* PDEs */ }; struct i915_page_directory_pointer_entry { @@ -243,9 +247,9 @@ struct i915_address_space { struct list_head inactive_list; /* FIXME: Need a more generic return type */ - gen6_gtt_pte_t (*pte_encode)(dma_addr_t addr, - enum i915_cache_level level, - bool valid, u32 flags); /* Create a valid PTE */ + gen6_pte_t (*pte_encode)(dma_addr_t addr, + enum i915_cache_level level, + bool valid, u32 flags); /* Create a valid PTE */ void (*clear_range)(struct i915_address_space *vm, uint64_t start, uint64_t length, -- cgit v1.2.1 From 317b4e903636305cfe702ab3e5b3d68547a69e72 Mon Sep 17 00:00:00 2001 From: Ben Widawsky Date: Mon, 16 Mar 2015 16:00:55 +0000 Subject: drm/i915: Extract context switch skip and add pd load logic In Gen8, PDPs are saved and restored with legacy contexts (legacy contexts only exist on the render ring). So change the ordering of LRI vs MI_SET_CONTEXT for the initialization of the context. Also the only cases in which we need to manually update the PDPs are when MI_RESTORE_INHIBIT has been set in MI_SET_CONTEXT (i.e. when the context is not yet initialized or it is the default context). Legacy submission is not available post GEN8, so it isn't necessary to add extra checks for newer generations. v2: Use new functions to replace the logic right away (Daniel) v3: Add missing pd load logic. v4: Add warning in case pd_load_pre & pd_load_post are true, and add missing trace_switch_mm. Cleaned up pd_load conditions. Add more information about when is pd_load_post needed. (Mika) Signed-off-by: Ben Widawsky Signed-off-by: Michel Thierry (v2+) Reviewed-by: Mika Kuoppala Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_gem_context.c | 75 ++++++++++++++++++++++++++++++++- 1 file changed, 73 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c index 70346b0028f9..b6ea85dcf9e6 100644 --- a/drivers/gpu/drm/i915/i915_gem_context.c +++ b/drivers/gpu/drm/i915/i915_gem_context.c @@ -569,6 +569,56 @@ mi_set_context(struct intel_engine_cs *ring, return ret; } +static inline bool should_skip_switch(struct intel_engine_cs *ring, + struct intel_context *from, + struct intel_context *to) +{ + if (from == to && !to->remap_slice) + return true; + + return false; +} + +static bool +needs_pd_load_pre(struct intel_engine_cs *ring, struct intel_context *to) +{ + struct drm_i915_private *dev_priv = ring->dev->dev_private; + + if (!to->ppgtt) + return false; + + if (INTEL_INFO(ring->dev)->gen < 8) + return true; + + if (ring != &dev_priv->ring[RCS]) + return true; + + return false; +} + +static bool +needs_pd_load_post(struct intel_engine_cs *ring, struct intel_context *to) +{ + struct drm_i915_private *dev_priv = ring->dev->dev_private; + + if (!to->ppgtt) + return false; + + if (!IS_GEN8(ring->dev)) + return false; + + if (ring != &dev_priv->ring[RCS]) + return false; + + if (!to->legacy_hw_ctx.initialized) + return true; + + if (i915_gem_context_is_default(to)) + return true; + + return false; +} + static int do_switch(struct intel_engine_cs *ring, struct intel_context *to) { @@ -584,7 +634,7 @@ static int do_switch(struct intel_engine_cs *ring, BUG_ON(!i915_gem_obj_is_pinned(from->legacy_hw_ctx.rcs_state)); } - if (from == to && !to->remap_slice) + if (should_skip_switch(ring, from, to)) return 0; /* Trying to pin first makes error handling easier. */ @@ -602,7 +652,14 @@ static int do_switch(struct intel_engine_cs *ring, */ from = ring->last_context; - if (to->ppgtt) { + /* We should never emit switch_mm more than once */ + WARN_ON(needs_pd_load_pre(ring, to) && needs_pd_load_post(ring, to)); + + if (needs_pd_load_pre(ring, to)) { + /* Older GENs and non render rings still want the load first, + * "PP_DCLV followed by PP_DIR_BASE register through Load + * Register Immediate commands in Ring Buffer before submitting + * a context."*/ trace_switch_mm(ring, to); ret = to->ppgtt->switch_mm(to->ppgtt, ring); if (ret) @@ -644,6 +701,20 @@ static int do_switch(struct intel_engine_cs *ring, if (ret) goto unpin_out; + if (needs_pd_load_post(ring, to)) { + trace_switch_mm(ring, to); + ret = to->ppgtt->switch_mm(to->ppgtt, ring); + /* The hardware context switch is emitted, but we haven't + * actually changed the state - so it's probably safe to bail + * here. Still, let the user know something dangerous has + * happened. + */ + if (ret) { + DRM_ERROR("Failed to change address space on context switch\n"); + goto unpin_out; + } + } + for (i = 0; i < MAX_L3_SLICES; i++) { if (!(to->remap_slice & (1< Date: Mon, 16 Mar 2015 16:00:56 +0000 Subject: drm/i915: Track GEN6 page table usage Instead of implementing the full tracking + dynamic allocation, this patch does a bit less than half of the work, by tracking and warning on unexpected conditions. The tracking itself follows which PTEs within a page table are currently being used for objects. The next patch will modify this to actually allocate the page tables only when necessary. With the current patch there isn't much in the way of making a gen agnostic range allocation function. However, in the next patch we'll add more specificity which makes having separate functions a bit easier to manage. One important change introduced here is that DMA mappings are created/destroyed at the same page directories/tables are allocated/deallocated. Notice that aliasing PPGTT is not managed here. The patch which actually begins dynamic allocation/teardown explains the reasoning for this. v2: s/pdp.page_directory/pdp.page_directories Make a scratch page allocation helper v3: Rebase and expand commit message. v4: Allocate required pagetables only when it is needed, _bind_to_vm instead of bind_vma (Daniel). v5: Rebased to remove the unnecessary noise in the diff, also: - PDE mask is GEN agnostic, renamed GEN6_PDE_MASK to I915_PDE_MASK. - Removed unnecessary checks in gen6_alloc_va_range. - Changed map/unmap_px_single macros to use dma functions directly and be part of a static inline function instead. - Moved drm_device plumbing through page tables operation to its own patch. - Moved allocate/teardown_va_range calls until they are fully implemented (in subsequent patch). - Merged pt and scratch_pt unmap_and_free path. - Moved scratch page allocator helper to the patch that will use it. v6: Reduce complexity by not tearing down pagetables dynamically, the same can be achieved while freeing empty vms. (Daniel) v7: s/i915_dma_map_px_single/i915_dma_map_single s/gen6_write_pdes/gen6_write_pde Prevent a NULL case when only GGTT is available. (Mika) v8: Rebased after s/page_tables/page_table/. v9: Reworked i915_pte_index and i915_pte_count. Also exercise bitmap allocation here (gen6_alloc_va_range) and fix incorrect write_page_range in i915_gem_restore_gtt_mappings (Mika). Cc: Daniel Vetter Cc: Mika Kuoppala Signed-off-by: Ben Widawsky Signed-off-by: Michel Thierry (v3+) Reviewed-by: Mika Kuoppala Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_gem.c | 9 ++ drivers/gpu/drm/i915/i915_gem_gtt.c | 199 +++++++++++++++++++++++++----------- drivers/gpu/drm/i915/i915_gem_gtt.h | 73 +++++++++++++ 3 files changed, 219 insertions(+), 62 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 092f25cfb8d5..69190af87a42 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -3510,6 +3510,15 @@ search_free: if (ret) goto err_remove_node; + /* allocate before insert / bind */ + if (vma->vm->allocate_va_range) { + ret = vma->vm->allocate_va_range(vma->vm, + vma->node.start, + vma->node.size); + if (ret) + goto err_remove_node; + } + trace_i915_vma_bind(vma, flags); ret = i915_vma_bind(vma, obj->cache_level, flags & PIN_GLOBAL ? GLOBAL_BIND : 0); diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c index 6f5a0cf77fad..7d206b59eb3c 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c @@ -279,29 +279,88 @@ static gen6_pte_t iris_pte_encode(dma_addr_t addr, return pte; } -static void unmap_and_free_pt(struct i915_page_table_entry *pt, struct drm_device *dev) +#define i915_dma_unmap_single(px, dev) \ + __i915_dma_unmap_single((px)->daddr, dev) + +static inline void __i915_dma_unmap_single(dma_addr_t daddr, + struct drm_device *dev) +{ + struct device *device = &dev->pdev->dev; + + dma_unmap_page(device, daddr, 4096, PCI_DMA_BIDIRECTIONAL); +} + +/** + * i915_dma_map_single() - Create a dma mapping for a page table/dir/etc. + * @px: Page table/dir/etc to get a DMA map for + * @dev: drm device + * + * Page table allocations are unified across all gens. They always require a + * single 4k allocation, as well as a DMA mapping. If we keep the structs + * symmetric here, the simple macro covers us for every page table type. + * + * Return: 0 if success. + */ +#define i915_dma_map_single(px, dev) \ + i915_dma_map_page_single((px)->page, (dev), &(px)->daddr) + +static inline int i915_dma_map_page_single(struct page *page, + struct drm_device *dev, + dma_addr_t *daddr) +{ + struct device *device = &dev->pdev->dev; + + *daddr = dma_map_page(device, page, 0, 4096, PCI_DMA_BIDIRECTIONAL); + return dma_mapping_error(device, *daddr); +} + +static void unmap_and_free_pt(struct i915_page_table_entry *pt, + struct drm_device *dev) { if (WARN_ON(!pt->page)) return; + + i915_dma_unmap_single(pt, dev); __free_page(pt->page); + kfree(pt->used_ptes); kfree(pt); } static struct i915_page_table_entry *alloc_pt_single(struct drm_device *dev) { struct i915_page_table_entry *pt; + const size_t count = INTEL_INFO(dev)->gen >= 8 ? + GEN8_PTES : GEN6_PTES; + int ret = -ENOMEM; pt = kzalloc(sizeof(*pt), GFP_KERNEL); if (!pt) return ERR_PTR(-ENOMEM); + pt->used_ptes = kcalloc(BITS_TO_LONGS(count), sizeof(*pt->used_ptes), + GFP_KERNEL); + + if (!pt->used_ptes) + goto fail_bitmap; + pt->page = alloc_page(GFP_KERNEL | __GFP_ZERO); - if (!pt->page) { - kfree(pt); - return ERR_PTR(-ENOMEM); - } + if (!pt->page) + goto fail_page; + + ret = i915_dma_map_single(pt, dev); + if (ret) + goto fail_dma; return pt; + +fail_dma: + __free_page(pt->page); +fail_page: + kfree(pt->used_ptes); +fail_bitmap: + kfree(pt); + + return ERR_PTR(ret); } /** @@ -848,26 +907,36 @@ static void gen6_dump_ppgtt(struct i915_hw_ppgtt *ppgtt, struct seq_file *m) } } -static void gen6_write_pdes(struct i915_hw_ppgtt *ppgtt) +/* Write pde (index) from the page directory @pd to the page table @pt */ +static void gen6_write_pde(struct i915_page_directory_entry *pd, + const int pde, struct i915_page_table_entry *pt) { - struct drm_i915_private *dev_priv = ppgtt->base.dev->dev_private; - gen6_pte_t __iomem *pd_addr; - uint32_t pd_entry; - int i; + /* Caller needs to make sure the write completes if necessary */ + struct i915_hw_ppgtt *ppgtt = + container_of(pd, struct i915_hw_ppgtt, pd); + u32 pd_entry; - WARN_ON(ppgtt->pd.pd_offset & 0x3f); - pd_addr = (gen6_pte_t __iomem*)dev_priv->gtt.gsm + - ppgtt->pd.pd_offset / sizeof(gen6_pte_t); - for (i = 0; i < ppgtt->num_pd_entries; i++) { - dma_addr_t pt_addr; + pd_entry = GEN6_PDE_ADDR_ENCODE(pt->daddr); + pd_entry |= GEN6_PDE_VALID; - pt_addr = ppgtt->pd.page_table[i]->daddr; - pd_entry = GEN6_PDE_ADDR_ENCODE(pt_addr); - pd_entry |= GEN6_PDE_VALID; + writel(pd_entry, ppgtt->pd_addr + pde); +} - writel(pd_entry, pd_addr + i); - } - readl(pd_addr); +/* Write all the page tables found in the ppgtt structure to incrementing page + * directories. */ +static void gen6_write_page_range(struct drm_i915_private *dev_priv, + struct i915_page_directory_entry *pd, + uint32_t start, uint32_t length) +{ + struct i915_page_table_entry *pt; + uint32_t pde, temp; + + gen6_for_each_pde(pt, pd, start, length, temp, pde) + gen6_write_pde(pd, pde, pt); + + /* Make sure write is complete before other code can use this page + * table. Also require for WC mapped PTEs */ + readl(dev_priv->gtt.gsm); } static uint32_t get_pd_offset(struct i915_hw_ppgtt *ppgtt) @@ -1093,6 +1162,28 @@ static void gen6_ppgtt_unmap_pages(struct i915_hw_ppgtt *ppgtt) 4096, PCI_DMA_BIDIRECTIONAL); } +static int gen6_alloc_va_range(struct i915_address_space *vm, + uint64_t start, uint64_t length) +{ + struct i915_hw_ppgtt *ppgtt = + container_of(vm, struct i915_hw_ppgtt, base); + struct i915_page_table_entry *pt; + uint32_t pde, temp; + + gen6_for_each_pde(pt, &ppgtt->pd, start, length, temp, pde) { + DECLARE_BITMAP(tmp_bitmap, GEN6_PTES); + + bitmap_zero(tmp_bitmap, GEN6_PTES); + bitmap_set(tmp_bitmap, gen6_pte_index(start), + gen6_pte_count(start, length)); + + bitmap_or(pt->used_ptes, pt->used_ptes, tmp_bitmap, + GEN6_PTES); + } + + return 0; +} + static void gen6_ppgtt_free(struct i915_hw_ppgtt *ppgtt) { int i; @@ -1139,20 +1230,24 @@ alloc: 0, dev_priv->gtt.base.total, 0); if (ret) - return ret; + goto err_out; retried = true; goto alloc; } if (ret) - return ret; + goto err_out; + if (ppgtt->node.start < dev_priv->gtt.mappable_end) DRM_DEBUG("Forced to use aperture for PDEs\n"); ppgtt->num_pd_entries = I915_PDES; return 0; + +err_out: + return ret; } static int gen6_ppgtt_alloc(struct i915_hw_ppgtt *ppgtt) @@ -1174,30 +1269,6 @@ static int gen6_ppgtt_alloc(struct i915_hw_ppgtt *ppgtt) return 0; } -static int gen6_ppgtt_setup_page_tables(struct i915_hw_ppgtt *ppgtt) -{ - struct drm_device *dev = ppgtt->base.dev; - int i; - - for (i = 0; i < ppgtt->num_pd_entries; i++) { - struct page *page; - dma_addr_t pt_addr; - - page = ppgtt->pd.page_table[i]->page; - pt_addr = pci_map_page(dev->pdev, page, 0, 4096, - PCI_DMA_BIDIRECTIONAL); - - if (pci_dma_mapping_error(dev->pdev, pt_addr)) { - gen6_ppgtt_unmap_pages(ppgtt); - return -EIO; - } - - ppgtt->pd.page_table[i]->daddr = pt_addr; - } - - return 0; -} - static int gen6_ppgtt_init(struct i915_hw_ppgtt *ppgtt) { struct drm_device *dev = ppgtt->base.dev; @@ -1221,12 +1292,7 @@ static int gen6_ppgtt_init(struct i915_hw_ppgtt *ppgtt) if (ret) return ret; - ret = gen6_ppgtt_setup_page_tables(ppgtt); - if (ret) { - gen6_ppgtt_free(ppgtt); - return ret; - } - + ppgtt->base.allocate_va_range = gen6_alloc_va_range; ppgtt->base.clear_range = gen6_ppgtt_clear_range; ppgtt->base.insert_entries = gen6_ppgtt_insert_entries; ppgtt->base.cleanup = gen6_ppgtt_cleanup; @@ -1237,13 +1303,17 @@ static int gen6_ppgtt_init(struct i915_hw_ppgtt *ppgtt) ppgtt->pd.pd_offset = ppgtt->node.start / PAGE_SIZE * sizeof(gen6_pte_t); + ppgtt->pd_addr = (gen6_pte_t __iomem *)dev_priv->gtt.gsm + + ppgtt->pd.pd_offset / sizeof(gen6_pte_t); + ppgtt->base.clear_range(&ppgtt->base, 0, ppgtt->base.total, true); + gen6_write_page_range(dev_priv, &ppgtt->pd, 0, ppgtt->base.total); + DRM_DEBUG_DRIVER("Allocated pde space (%lldM) at GTT entry: %llx\n", ppgtt->node.size >> 20, ppgtt->node.start / PAGE_SIZE); - gen6_write_pdes(ppgtt); DRM_DEBUG("Adding PPGTT at offset %x\n", ppgtt->pd.pd_offset << 10); @@ -1514,15 +1584,20 @@ void i915_gem_restore_gtt_mappings(struct drm_device *dev) return; } - list_for_each_entry(vm, &dev_priv->vm_list, global_link) { - /* TODO: Perhaps it shouldn't be gen6 specific */ - if (i915_is_ggtt(vm)) { - if (dev_priv->mm.aliasing_ppgtt) - gen6_write_pdes(dev_priv->mm.aliasing_ppgtt); - continue; - } + if (USES_PPGTT(dev)) { + list_for_each_entry(vm, &dev_priv->vm_list, global_link) { + /* TODO: Perhaps it shouldn't be gen6 specific */ + + struct i915_hw_ppgtt *ppgtt = + container_of(vm, struct i915_hw_ppgtt, + base); - gen6_write_pdes(container_of(vm, struct i915_hw_ppgtt, base)); + if (i915_is_ggtt(vm)) + ppgtt = dev_priv->mm.aliasing_ppgtt; + + gen6_write_page_range(dev_priv, &ppgtt->pd, + 0, ppgtt->base.total); + } } i915_ggtt_flush(dev_priv); diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.h b/drivers/gpu/drm/i915/i915_gem_gtt.h index 5ca7c5eff88b..7472312deff9 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.h +++ b/drivers/gpu/drm/i915/i915_gem_gtt.h @@ -55,10 +55,12 @@ typedef uint64_t gen8_pde_t; #define I915_PTE_MASK(pte_len) (I915_PTES(pte_len) - 1) #define I915_PDES 512 #define I915_PDE_MASK (I915_PDES - 1) +#define NUM_PTE(pde_shift) (1 << (pde_shift - PAGE_SHIFT)) #define GEN6_PTES I915_PTES(sizeof(gen6_pte_t)) #define GEN6_PD_SIZE (I915_PDES * PAGE_SIZE) #define GEN6_PD_ALIGN (PAGE_SIZE * 16) +#define GEN6_PDE_SHIFT 22 #define GEN6_PDE_VALID (1 << 0) #define GEN7_PTE_CACHE_L3_LLC (3 << 1) @@ -194,6 +196,8 @@ struct i915_vma { struct i915_page_table_entry { struct page *page; dma_addr_t daddr; + + unsigned long *used_ptes; }; struct i915_page_directory_entry { @@ -250,6 +254,9 @@ struct i915_address_space { gen6_pte_t (*pte_encode)(dma_addr_t addr, enum i915_cache_level level, bool valid, u32 flags); /* Create a valid PTE */ + int (*allocate_va_range)(struct i915_address_space *vm, + uint64_t start, + uint64_t length); void (*clear_range)(struct i915_address_space *vm, uint64_t start, uint64_t length, @@ -302,12 +309,78 @@ struct i915_hw_ppgtt { struct drm_i915_file_private *file_priv; + gen6_pte_t __iomem *pd_addr; + int (*enable)(struct i915_hw_ppgtt *ppgtt); int (*switch_mm)(struct i915_hw_ppgtt *ppgtt, struct intel_engine_cs *ring); void (*debug_dump)(struct i915_hw_ppgtt *ppgtt, struct seq_file *m); }; +/* For each pde iterates over every pde between from start until start + length. + * If start, and start+length are not perfectly divisible, the macro will round + * down, and up as needed. The macro modifies pde, start, and length. Dev is + * only used to differentiate shift values. Temp is temp. On gen6/7, start = 0, + * and length = 2G effectively iterates over every PDE in the system. + * + * XXX: temp is not actually needed, but it saves doing the ALIGN operation. + */ +#define gen6_for_each_pde(pt, pd, start, length, temp, iter) \ + for (iter = gen6_pde_index(start), pt = (pd)->page_table[iter]; \ + length > 0 && iter < I915_PDES; \ + pt = (pd)->page_table[++iter], \ + temp = ALIGN(start+1, 1 << GEN6_PDE_SHIFT) - start, \ + temp = min_t(unsigned, temp, length), \ + start += temp, length -= temp) + +static inline uint32_t i915_pte_index(uint64_t address, uint32_t pde_shift) +{ + const uint32_t mask = NUM_PTE(pde_shift) - 1; + + return (address >> PAGE_SHIFT) & mask; +} + +/* Helper to counts the number of PTEs within the given length. This count + * does not cross a page table boundary, so the max value would be + * GEN6_PTES for GEN6, and GEN8_PTES for GEN8. +*/ +static inline uint32_t i915_pte_count(uint64_t addr, size_t length, + uint32_t pde_shift) +{ + const uint64_t mask = ~((1 << pde_shift) - 1); + uint64_t end; + + WARN_ON(length == 0); + WARN_ON(offset_in_page(addr|length)); + + end = addr + length; + + if ((addr & mask) != (end & mask)) + return NUM_PTE(pde_shift) - i915_pte_index(addr, pde_shift); + + return i915_pte_index(end, pde_shift) - i915_pte_index(addr, pde_shift); +} + +static inline uint32_t i915_pde_index(uint64_t addr, uint32_t shift) +{ + return (addr >> shift) & I915_PDE_MASK; +} + +static inline uint32_t gen6_pte_index(uint32_t addr) +{ + return i915_pte_index(addr, GEN6_PDE_SHIFT); +} + +static inline size_t gen6_pte_count(uint32_t addr, uint32_t length) +{ + return i915_pte_count(addr, length, GEN6_PDE_SHIFT); +} + +static inline uint32_t gen6_pde_index(uint32_t addr) +{ + return i915_pde_index(addr, GEN6_PDE_SHIFT); +} + int i915_gem_gtt_init(struct drm_device *dev); void i915_gem_init_global_gtt(struct drm_device *dev); void i915_global_gtt_cleanup(struct drm_device *dev); -- cgit v1.2.1 From 563222a745012cc2bb20c5d5cbff1c1ec4832c05 Mon Sep 17 00:00:00 2001 From: Ben Widawsky Date: Thu, 19 Mar 2015 12:53:28 +0000 Subject: drm/i915: Track page table reload need This patch was formerly known as, "Force pd restore when PDEs change, gen6-7." I had to change the name because it is needed for GEN8 too. The real issue this is trying to solve is when a new object is mapped into the current address space. The GPU does not snoop the new mapping so we must do the gen specific action to reload the page tables. GEN8 and GEN7 do differ in the way they load page tables for the RCS. GEN8 does so with the context restore, while GEN7 requires the proper load commands in the command streamer. Non-render is similar for both. Caveat for GEN7 The docs say you cannot change the PDEs of a currently running context. We never map new PDEs of a running context, and expect them to be present - so I think this is okay. (We can unmap, but this should also be okay since we only unmap unreferenced objects that the GPU shouldn't be tryingto va->pa xlate.) The MI_SET_CONTEXT command does have a flag to signal that even if the context is the same, force a reload. It's unclear exactly what this does, but I have a hunch it's the right thing to do. The logic assumes that we always emit a context switch after mapping new PDEs, and before we submit a batch. This is the case today, and has been the case since the inception of hardware contexts. A note in the comment let's the user know. It's not just for gen8. If the current context has mappings change, we need a context reload to switch v2: Rebased after ppgtt clean up patches. Split the warning for aliasing and true ppgtt options. And do not break aliasing ppgtt, where to->ppgtt is always null. v3: Invalidate PPGTT TLBs inside alloc_va_range. v4: Rename ppgtt_invalidate_tlbs to mark_tlbs_dirty and move pd_dirty_rings from i915_address_space to i915_hw_ppgtt. Fixes when neither ctx->ppgtt and aliasing_ppgtt exist. v5: Removed references to teardown_va_range. v6: Updated needs_pd_load_pre/post. v7: Fix pd_dirty_rings check in needs_pd_load_post, and update/move comment about updated PDEs to object_pin/bind (Mika). Cc: Mika Kuoppala Signed-off-by: Ben Widawsky Signed-off-by: Michel Thierry (v2+) Reviewed-by: Mika Kuoppala Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_gem.c | 4 ++++ drivers/gpu/drm/i915/i915_gem_context.c | 26 ++++++++++++++++++++------ drivers/gpu/drm/i915/i915_gem_execbuffer.c | 7 +++++++ drivers/gpu/drm/i915/i915_gem_gtt.c | 11 +++++++++++ drivers/gpu/drm/i915/i915_gem_gtt.h | 1 + 5 files changed, 43 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 69190af87a42..84e2a231b03c 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -4143,6 +4143,10 @@ i915_gem_object_do_pin(struct drm_i915_gem_object *obj, bound = vma ? vma->bound : 0; if (vma == NULL || !drm_mm_node_allocated(&vma->node)) { + /* In true PPGTT, bind has possibly changed PDEs, which + * means we must do a context switch before the GPU can + * accurately read some of the VMAs. + */ vma = i915_gem_object_bind_to_vm(obj, vm, ggtt_view, alignment, flags); if (IS_ERR(vma)) diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c index b6ea85dcf9e6..dd9ab36a039b 100644 --- a/drivers/gpu/drm/i915/i915_gem_context.c +++ b/drivers/gpu/drm/i915/i915_gem_context.c @@ -573,8 +573,20 @@ static inline bool should_skip_switch(struct intel_engine_cs *ring, struct intel_context *from, struct intel_context *to) { - if (from == to && !to->remap_slice) - return true; + struct drm_i915_private *dev_priv = ring->dev->dev_private; + + if (to->remap_slice) + return false; + + if (to->ppgtt) { + if (from == to && !test_bit(ring->id, + &to->ppgtt->pd_dirty_rings)) + return true; + } else if (dev_priv->mm.aliasing_ppgtt) { + if (from == to && !test_bit(ring->id, + &dev_priv->mm.aliasing_ppgtt->pd_dirty_rings)) + return true; + } return false; } @@ -610,10 +622,7 @@ needs_pd_load_post(struct intel_engine_cs *ring, struct intel_context *to) if (ring != &dev_priv->ring[RCS]) return false; - if (!to->legacy_hw_ctx.initialized) - return true; - - if (i915_gem_context_is_default(to)) + if (to->ppgtt->pd_dirty_rings) return true; return false; @@ -664,6 +673,9 @@ static int do_switch(struct intel_engine_cs *ring, ret = to->ppgtt->switch_mm(to->ppgtt, ring); if (ret) goto unpin_out; + + /* Doing a PD load always reloads the page dirs */ + clear_bit(ring->id, &to->ppgtt->pd_dirty_rings); } if (ring != &dev_priv->ring[RCS]) { @@ -696,6 +708,8 @@ static int do_switch(struct intel_engine_cs *ring, if (!to->legacy_hw_ctx.initialized || i915_gem_context_is_default(to)) hw_flags |= MI_RESTORE_INHIBIT; + else if (to->ppgtt && test_and_clear_bit(ring->id, &to->ppgtt->pd_dirty_rings)) + hw_flags |= MI_FORCE_RESTORE; ret = mi_set_context(ring, to, hw_flags); if (ret) diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c index 9838d11e43ee..43335de0f713 100644 --- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c +++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c @@ -1251,6 +1251,13 @@ i915_gem_ringbuffer_submission(struct drm_device *dev, struct drm_file *file, if (ret) goto error; + if (ctx->ppgtt) + WARN(ctx->ppgtt->pd_dirty_rings & (1<id), + "%s didn't clear reload\n", ring->name); + else if (dev_priv->mm.aliasing_ppgtt) + WARN(dev_priv->mm.aliasing_ppgtt->pd_dirty_rings & + (1<id), "%s didn't clear reload\n", ring->name); + instp_mode = args->flags & I915_EXEC_CONSTANTS_MASK; instp_mask = I915_EXEC_CONSTANTS_MASK; switch (instp_mode) { diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c index 7d206b59eb3c..645c3636c571 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c @@ -1162,6 +1162,16 @@ static void gen6_ppgtt_unmap_pages(struct i915_hw_ppgtt *ppgtt) 4096, PCI_DMA_BIDIRECTIONAL); } +/* PDE TLBs are a pain invalidate pre GEN8. It requires a context reload. If we + * are switching between contexts with the same LRCA, we also must do a force + * restore. + */ +static inline void mark_tlbs_dirty(struct i915_hw_ppgtt *ppgtt) +{ + /* If current vm != vm, */ + ppgtt->pd_dirty_rings = INTEL_INFO(ppgtt->base.dev)->ring_mask; +} + static int gen6_alloc_va_range(struct i915_address_space *vm, uint64_t start, uint64_t length) { @@ -1181,6 +1191,7 @@ static int gen6_alloc_va_range(struct i915_address_space *vm, GEN6_PTES); } + mark_tlbs_dirty(ppgtt); return 0; } diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.h b/drivers/gpu/drm/i915/i915_gem_gtt.h index 7472312deff9..75e29f738a1b 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.h +++ b/drivers/gpu/drm/i915/i915_gem_gtt.h @@ -300,6 +300,7 @@ struct i915_hw_ppgtt { struct i915_address_space base; struct kref ref; struct drm_mm_node node; + unsigned long pd_dirty_rings; unsigned num_pd_entries; unsigned num_pd_pages; /* gen8+ */ union { -- cgit v1.2.1 From 6702cf16e0ba8b0129f5aa1b6609d4e9c70bc13b Mon Sep 17 00:00:00 2001 From: Ben Widawsky Date: Mon, 16 Mar 2015 16:00:58 +0000 Subject: drm/i915: Initialize all contexts The problem is we're going to switch to a new context, which could be the default context. The plan was to use restore inhibit, which would be fine, except if we are using dynamic page tables (which we will). If we use dynamic page tables and we don't load new page tables, the previous page tables might go away, and future operations will fault. CTXA runs. switch to default, restore inhibit CTXA dies and has its address space taken away. Run CTXB, tries to save using the context A's address space - this fails. The general solution is to make sure every context has it's own state, and its own address space. For cases when we must restore inhibit, first thing we do is load a valid address space. I thought this would be enough, but apparently there are references within the context itself which will refer to the old address space - therefore, we also must reinitialize. v2: to->ppgtt is only valid in full ppgtt. v3: Rebased. v4: Make post PDP update clearer. Signed-off-by: Ben Widawsky Signed-off-by: Michel Thierry (v2+) Reviewed-by: Mika Kuoppala Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_gem_context.c | 28 +++++++++++++++++++--------- 1 file changed, 19 insertions(+), 9 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_gem_context.c b/drivers/gpu/drm/i915/i915_gem_context.c index dd9ab36a039b..f3e84c44d009 100644 --- a/drivers/gpu/drm/i915/i915_gem_context.c +++ b/drivers/gpu/drm/i915/i915_gem_context.c @@ -609,7 +609,8 @@ needs_pd_load_pre(struct intel_engine_cs *ring, struct intel_context *to) } static bool -needs_pd_load_post(struct intel_engine_cs *ring, struct intel_context *to) +needs_pd_load_post(struct intel_engine_cs *ring, struct intel_context *to, + u32 hw_flags) { struct drm_i915_private *dev_priv = ring->dev->dev_private; @@ -622,7 +623,7 @@ needs_pd_load_post(struct intel_engine_cs *ring, struct intel_context *to) if (ring != &dev_priv->ring[RCS]) return false; - if (to->ppgtt->pd_dirty_rings) + if (hw_flags & MI_RESTORE_INHIBIT) return true; return false; @@ -661,9 +662,6 @@ static int do_switch(struct intel_engine_cs *ring, */ from = ring->last_context; - /* We should never emit switch_mm more than once */ - WARN_ON(needs_pd_load_pre(ring, to) && needs_pd_load_post(ring, to)); - if (needs_pd_load_pre(ring, to)) { /* Older GENs and non render rings still want the load first, * "PP_DCLV followed by PP_DIR_BASE register through Load @@ -706,16 +704,28 @@ static int do_switch(struct intel_engine_cs *ring, goto unpin_out; } - if (!to->legacy_hw_ctx.initialized || i915_gem_context_is_default(to)) + if (!to->legacy_hw_ctx.initialized) { hw_flags |= MI_RESTORE_INHIBIT; - else if (to->ppgtt && test_and_clear_bit(ring->id, &to->ppgtt->pd_dirty_rings)) + /* NB: If we inhibit the restore, the context is not allowed to + * die because future work may end up depending on valid address + * space. This means we must enforce that a page table load + * occur when this occurs. */ + } else if (to->ppgtt && + test_and_clear_bit(ring->id, &to->ppgtt->pd_dirty_rings)) hw_flags |= MI_FORCE_RESTORE; + /* We should never emit switch_mm more than once */ + WARN_ON(needs_pd_load_pre(ring, to) && + needs_pd_load_post(ring, to, hw_flags)); + ret = mi_set_context(ring, to, hw_flags); if (ret) goto unpin_out; - if (needs_pd_load_post(ring, to)) { + /* GEN8 does *not* require an explicit reload if the PDPs have been + * setup, and we do not wish to move them. + */ + if (needs_pd_load_post(ring, to, hw_flags)) { trace_switch_mm(ring, to); ret = to->ppgtt->switch_mm(to->ppgtt, ring); /* The hardware context switch is emitted, but we haven't @@ -766,7 +776,7 @@ static int do_switch(struct intel_engine_cs *ring, i915_gem_context_unreference(from); } - uninitialized = !to->legacy_hw_ctx.initialized && from == NULL; + uninitialized = !to->legacy_hw_ctx.initialized; to->legacy_hw_ctx.initialized = true; done: -- cgit v1.2.1 From bdd7554d568fa165b0e86fc32b1cde3c895ff774 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 19 Mar 2015 17:57:11 +0200 Subject: drm/i915: Kill intel_plane->obj MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit intel_plane->obj is not used anymore so kill it. Also don't pass both the fb and obj to the sprite .update_plane() hook, as just passing the fb is enough. Signed-off-by: Ville Syrjälä Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_display.c | 7 ------- drivers/gpu/drm/i915/intel_drv.h | 2 -- drivers/gpu/drm/i915/intel_sprite.c | 18 ++++++++++-------- 3 files changed, 10 insertions(+), 17 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index f1c0295f69e5..3c91bd19987b 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -12091,8 +12091,6 @@ intel_commit_primary_plane(struct drm_plane *plane, struct drm_device *dev = plane->dev; struct drm_i915_private *dev_priv = dev->dev_private; struct intel_crtc *intel_crtc; - struct drm_i915_gem_object *obj = intel_fb_obj(fb); - struct intel_plane *intel_plane = to_intel_plane(plane); struct drm_rect *src = &state->src; crtc = crtc ? crtc : plane->crtc; @@ -12102,8 +12100,6 @@ intel_commit_primary_plane(struct drm_plane *plane, crtc->x = src->x1 >> 16; crtc->y = src->y1 >> 16; - intel_plane->obj = obj; - if (intel_crtc->active) { if (state->visible) { /* FIXME: kill this fastboot hack */ @@ -12367,7 +12363,6 @@ intel_commit_cursor_plane(struct drm_plane *plane, struct drm_crtc *crtc = state->base.crtc; struct drm_device *dev = plane->dev; struct intel_crtc *intel_crtc; - struct intel_plane *intel_plane = to_intel_plane(plane); struct drm_i915_gem_object *obj = intel_fb_obj(state->base.fb); uint32_t addr; @@ -12378,8 +12373,6 @@ intel_commit_cursor_plane(struct drm_plane *plane, crtc->cursor_x = state->base.crtc_x; crtc->cursor_y = state->base.crtc_y; - intel_plane->obj = obj; - if (intel_crtc->cursor_bo == obj) goto update; diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index 8bb18e507f5f..c06854df466d 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -507,7 +507,6 @@ struct intel_plane { struct drm_plane base; int plane; enum pipe pipe; - struct drm_i915_gem_object *obj; bool can_scale; int max_downscale; @@ -527,7 +526,6 @@ struct intel_plane { void (*update_plane)(struct drm_plane *plane, struct drm_crtc *crtc, struct drm_framebuffer *fb, - struct drm_i915_gem_object *obj, int crtc_x, int crtc_y, unsigned int crtc_w, unsigned int crtc_h, uint32_t x, uint32_t y, diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c index a82873631851..03c2d4c11518 100644 --- a/drivers/gpu/drm/i915/intel_sprite.c +++ b/drivers/gpu/drm/i915/intel_sprite.c @@ -179,7 +179,7 @@ static void intel_update_primary_plane(struct intel_crtc *crtc) static void skl_update_plane(struct drm_plane *drm_plane, struct drm_crtc *crtc, struct drm_framebuffer *fb, - struct drm_i915_gem_object *obj, int crtc_x, int crtc_y, + int crtc_x, int crtc_y, unsigned int crtc_w, unsigned int crtc_h, uint32_t x, uint32_t y, uint32_t src_w, uint32_t src_h) @@ -187,6 +187,7 @@ skl_update_plane(struct drm_plane *drm_plane, struct drm_crtc *crtc, struct drm_device *dev = drm_plane->dev; struct drm_i915_private *dev_priv = dev->dev_private; struct intel_plane *intel_plane = to_intel_plane(drm_plane); + struct drm_i915_gem_object *obj = intel_fb_obj(fb); const int pipe = intel_plane->pipe; const int plane = intel_plane->plane + 1; u32 plane_ctl, stride_div; @@ -407,7 +408,7 @@ chv_update_csc(struct intel_plane *intel_plane, uint32_t format) static void vlv_update_plane(struct drm_plane *dplane, struct drm_crtc *crtc, struct drm_framebuffer *fb, - struct drm_i915_gem_object *obj, int crtc_x, int crtc_y, + int crtc_x, int crtc_y, unsigned int crtc_w, unsigned int crtc_h, uint32_t x, uint32_t y, uint32_t src_w, uint32_t src_h) @@ -416,6 +417,7 @@ vlv_update_plane(struct drm_plane *dplane, struct drm_crtc *crtc, struct drm_i915_private *dev_priv = dev->dev_private; struct intel_plane *intel_plane = to_intel_plane(dplane); struct intel_crtc *intel_crtc = to_intel_crtc(crtc); + struct drm_i915_gem_object *obj = intel_fb_obj(fb); int pipe = intel_plane->pipe; int plane = intel_plane->plane; u32 sprctl; @@ -608,7 +610,7 @@ vlv_get_colorkey(struct drm_plane *dplane, static void ivb_update_plane(struct drm_plane *plane, struct drm_crtc *crtc, struct drm_framebuffer *fb, - struct drm_i915_gem_object *obj, int crtc_x, int crtc_y, + int crtc_x, int crtc_y, unsigned int crtc_w, unsigned int crtc_h, uint32_t x, uint32_t y, uint32_t src_w, uint32_t src_h) @@ -617,6 +619,7 @@ ivb_update_plane(struct drm_plane *plane, struct drm_crtc *crtc, struct drm_i915_private *dev_priv = dev->dev_private; struct intel_plane *intel_plane = to_intel_plane(plane); struct intel_crtc *intel_crtc = to_intel_crtc(crtc); + struct drm_i915_gem_object *obj = intel_fb_obj(fb); int pipe = intel_plane->pipe; u32 sprctl, sprscale = 0; unsigned long sprsurf_offset, linear_offset; @@ -813,7 +816,7 @@ ivb_get_colorkey(struct drm_plane *plane, struct drm_intel_sprite_colorkey *key) static void ilk_update_plane(struct drm_plane *plane, struct drm_crtc *crtc, struct drm_framebuffer *fb, - struct drm_i915_gem_object *obj, int crtc_x, int crtc_y, + int crtc_x, int crtc_y, unsigned int crtc_w, unsigned int crtc_h, uint32_t x, uint32_t y, uint32_t src_w, uint32_t src_h) @@ -822,6 +825,7 @@ ilk_update_plane(struct drm_plane *plane, struct drm_crtc *crtc, struct drm_i915_private *dev_priv = dev->dev_private; struct intel_plane *intel_plane = to_intel_plane(plane); struct intel_crtc *intel_crtc = to_intel_crtc(crtc); + struct drm_i915_gem_object *obj = intel_fb_obj(fb); int pipe = intel_plane->pipe; unsigned long dvssurf_offset, linear_offset; u32 dvscntr, dvsscale; @@ -1275,7 +1279,6 @@ intel_commit_sprite_plane(struct drm_plane *plane, struct intel_crtc *intel_crtc; struct intel_plane *intel_plane = to_intel_plane(plane); struct drm_framebuffer *fb = state->base.fb; - struct drm_i915_gem_object *obj = intel_fb_obj(fb); int crtc_x, crtc_y; unsigned int crtc_w, crtc_h; uint32_t src_x, src_y, src_w, src_h; @@ -1283,8 +1286,7 @@ intel_commit_sprite_plane(struct drm_plane *plane, crtc = crtc ? crtc : plane->crtc; intel_crtc = to_intel_crtc(crtc); - plane->fb = state->base.fb; - intel_plane->obj = obj; + plane->fb = fb; if (intel_crtc->active) { intel_crtc->primary_enabled = !state->hides_primary; @@ -1298,7 +1300,7 @@ intel_commit_sprite_plane(struct drm_plane *plane, src_y = state->src.y1; src_w = drm_rect_width(&state->src); src_h = drm_rect_height(&state->src); - intel_plane->update_plane(plane, crtc, fb, obj, + intel_plane->update_plane(plane, crtc, fb, crtc_x, crtc_y, crtc_w, crtc_h, src_x, src_y, src_w, src_h); } else { -- cgit v1.2.1 From d5dd62bd9a19bcea4bcf20f66de56591fdd0d8d2 Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Tue, 17 Mar 2015 11:40:03 +0200 Subject: drm/i915: factor out vlv_PLL_is_optimal Factor out the logic to decide whether the newly calculated dividers are better than the best found so far. Do this for clarity and to prepare for the upcoming BXT helper needing the same. No functional change. Signed-off-by: Imre Deak Reviewed-by: Jesse Barnes Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_display.c | 50 ++++++++++++++++++++++++++---------- 1 file changed, 36 insertions(+), 14 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 3c91bd19987b..9dc9a85d0008 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -776,6 +776,33 @@ g4x_find_best_dpll(const intel_limit_t *limit, struct intel_crtc *crtc, return found; } +/* + * Check if the calculated PLL configuration is more optimal compared to the + * best configuration and error found so far. Return the calculated error. + */ +static bool vlv_PLL_is_optimal(struct drm_device *dev, int target_freq, + const intel_clock_t *calculated_clock, + const intel_clock_t *best_clock, + unsigned int best_error_ppm, + unsigned int *error_ppm) +{ + *error_ppm = div_u64(1000000ULL * + abs(target_freq - calculated_clock->dot), + target_freq); + /* + * Prefer a better P value over a better (smaller) error if the error + * is small. Ensure this preference for future configurations too by + * setting the error to 0. + */ + if (*error_ppm < 100 && calculated_clock->p > best_clock->p) { + *error_ppm = 0; + + return true; + } + + return *error_ppm + 10 < best_error_ppm; +} + static bool vlv_find_best_dpll(const intel_limit_t *limit, struct intel_crtc *crtc, int target, int refclk, intel_clock_t *match_clock, @@ -800,7 +827,7 @@ vlv_find_best_dpll(const intel_limit_t *limit, struct intel_crtc *crtc, clock.p = clock.p1 * clock.p2; /* based on hardware requirement, prefer bigger m1,m2 values */ for (clock.m1 = limit->m1.min; clock.m1 <= limit->m1.max; clock.m1++) { - unsigned int ppm, diff; + unsigned int ppm; clock.m2 = DIV_ROUND_CLOSEST(target * clock.p * clock.n, refclk * clock.m1); @@ -811,20 +838,15 @@ vlv_find_best_dpll(const intel_limit_t *limit, struct intel_crtc *crtc, &clock)) continue; - diff = abs(clock.dot - target); - ppm = div_u64(1000000ULL * diff, target); - - if (ppm < 100 && clock.p > best_clock->p) { - bestppm = 0; - *best_clock = clock; - found = true; - } + if (!vlv_PLL_is_optimal(dev, target, + &clock, + best_clock, + bestppm, &ppm)) + continue; - if (bestppm >= 10 && ppm < bestppm - 10) { - bestppm = ppm; - *best_clock = clock; - found = true; - } + *best_clock = clock; + bestppm = ppm; + found = true; } } } -- cgit v1.2.1 From 24be4e4650de2fc323ee0737f945ad366b6c11cc Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Tue, 17 Mar 2015 11:40:04 +0200 Subject: drm/i915: check for div-by-zero in vlv_PLL_is_optimal Signed-off-by: Imre Deak Reviewed-by: Jesse Barnes Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_display.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 9dc9a85d0008..b0aa6a4784c4 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -786,6 +786,9 @@ static bool vlv_PLL_is_optimal(struct drm_device *dev, int target_freq, unsigned int best_error_ppm, unsigned int *error_ppm) { + if (WARN_ON_ONCE(!target_freq)) + return false; + *error_ppm = div_u64(1000000ULL * abs(target_freq - calculated_clock->dot), target_freq); -- cgit v1.2.1 From 9ca3ba011b9f1bf7fc73a4e7dc0c9edec2aeb149 Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Tue, 17 Mar 2015 11:40:05 +0200 Subject: drm/i915/chv: use vlv_PLL_is_optimal in chv_find_best_dpll Prepare chv_find_best_dpll to be used for BXT too, where we want to consider the error between target and calculated frequency too when choosing a better PLL configuration. No functional change. Signed-off-by: Imre Deak Reviewed-by: Jesse Barnes Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_display.c | 26 ++++++++++++++++++++------ 1 file changed, 20 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index b0aa6a4784c4..60230dcf532f 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -786,6 +786,16 @@ static bool vlv_PLL_is_optimal(struct drm_device *dev, int target_freq, unsigned int best_error_ppm, unsigned int *error_ppm) { + /* + * For CHV ignore the error and consider only the P value. + * Prefer a bigger P value based on HW requirements. + */ + if (IS_CHERRYVIEW(dev)) { + *error_ppm = 0; + + return calculated_clock->p > best_clock->p; + } + if (WARN_ON_ONCE(!target_freq)) return false; @@ -864,11 +874,13 @@ chv_find_best_dpll(const intel_limit_t *limit, struct intel_crtc *crtc, intel_clock_t *best_clock) { struct drm_device *dev = crtc->base.dev; + unsigned int best_error_ppm; intel_clock_t clock; uint64_t m2; int found = false; memset(best_clock, 0, sizeof(*best_clock)); + best_error_ppm = 1000000; /* * Based on hardware doc, the n always set to 1, and m1 always @@ -882,6 +894,7 @@ chv_find_best_dpll(const intel_limit_t *limit, struct intel_crtc *crtc, for (clock.p2 = limit->p2.p2_fast; clock.p2 >= limit->p2.p2_slow; clock.p2 -= clock.p2 > 10 ? 2 : 1) { + unsigned int error_ppm; clock.p = clock.p1 * clock.p2; @@ -898,12 +911,13 @@ chv_find_best_dpll(const intel_limit_t *limit, struct intel_crtc *crtc, if (!intel_PLL_is_valid(dev, limit, &clock)) continue; - /* based on hardware requirement, prefer bigger p - */ - if (clock.p > best_clock->p) { - *best_clock = clock; - found = true; - } + if (!vlv_PLL_is_optimal(dev, target, &clock, best_clock, + best_error_ppm, &error_ppm)) + continue; + + *best_clock = clock; + best_error_ppm = error_ppm; + found = true; } } -- cgit v1.2.1 From 08fd59fcc8c0f9732214fb93ad5cb09f67cb1124 Mon Sep 17 00:00:00 2001 From: Matt Roper Date: Wed, 18 Mar 2015 15:04:47 -0700 Subject: drm/i915: Move vblank wait determination to 'check' phase Determining whether we'll need to wait for vblanks is something we should determine during the atomic 'check' phase, not the 'commit' phase. Note that we only set these bits in the branch of 'check' where intel_crtc->active is true so that we don't try to wait on a disabled CRTC. The whole 'wait for vblank after update' flag should go away in the future, once we start handling watermarks in a proper atomic manner. This regression has been introduced in commit 2fdd7def16dd7580f297827930126c16b152ec11 Author: Matt Roper Date: Wed Mar 4 10:49:04 2015 -0800 drm/i915: Don't clobber plane state on internal disables Cc: Paulo Zanoni Root-cause-analysis-by: Paulo Zanoni Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=89550 Testcase: igt/pm_rpm/legacy-planes Testcase: igt/pm_rpm/legacy-planes-dpms Testcase: igt/pm_rpm/universal-planes Testcase: igt/pm_rpm/universal-planes-dpms Signed-off-by: Matt Roper Tested-by: Paulo Zanoni Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_sprite.c | 24 ++++++++++-------------- 1 file changed, 10 insertions(+), 14 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c index 03c2d4c11518..cd531fa997de 100644 --- a/drivers/gpu/drm/i915/intel_sprite.c +++ b/drivers/gpu/drm/i915/intel_sprite.c @@ -750,13 +750,6 @@ ivb_disable_plane(struct drm_plane *plane, struct drm_crtc *crtc) I915_WRITE(SPRSURF(pipe), 0); intel_flush_primary_plane(dev_priv, intel_crtc->plane); - - /* - * Avoid underruns when disabling the sprite. - * FIXME remove once watermark updates are done properly. - */ - intel_crtc->atomic.wait_vblank = true; - intel_crtc->atomic.update_sprite_watermarks |= (1 << drm_plane_index(plane)); } static int @@ -941,13 +934,6 @@ ilk_disable_plane(struct drm_plane *plane, struct drm_crtc *crtc) I915_WRITE(DVSSURF(pipe), 0); intel_flush_primary_plane(dev_priv, intel_crtc->plane); - - /* - * Avoid underruns when disabling the sprite. - * FIXME remove once watermark updates are done properly. - */ - intel_crtc->atomic.wait_vblank = true; - intel_crtc->atomic.update_sprite_watermarks |= (1 << drm_plane_index(plane)); } /** @@ -1266,6 +1252,16 @@ finish: plane->state->fb->modifier[0] != state->base.fb->modifier[0]) intel_crtc->atomic.update_wm = true; + + if (!state->visible) { + /* + * Avoid underruns when disabling the sprite. + * FIXME remove once watermark updates are done properly. + */ + intel_crtc->atomic.wait_vblank = true; + intel_crtc->atomic.update_sprite_watermarks |= + (1 << drm_plane_index(plane)); + } } return 0; -- cgit v1.2.1 From 47ecbb208bf44be0e798495d08837fe7c4138201 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 19 Mar 2015 21:18:57 +0200 Subject: drm/i915: Eliminate the RMW sprite colorkey management MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Store the colorkey in intel_plane and kill off all the RMW stuff handling it. This is just an intermediate step and eventually the colorkey needs to be converted into some properties. v2: Actually update the hardware state in the set_colorkey ioctl (Daniel) Signed-off-by: Ville Syrjälä Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_drv.h | 7 +- drivers/gpu/drm/i915/intel_sprite.c | 297 ++++++++---------------------------- 2 files changed, 69 insertions(+), 235 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index c06854df466d..6f20f3a01398 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -510,6 +510,9 @@ struct intel_plane { bool can_scale; int max_downscale; + /* FIXME convert to properties */ + struct drm_intel_sprite_colorkey ckey; + /* Since we need to change the watermarks before/after * enabling/disabling the planes, we need to store the parameters here * as the other pieces of the struct may not reflect the values we want @@ -536,10 +539,6 @@ struct intel_plane { struct intel_plane_state *state); void (*commit_plane)(struct drm_plane *plane, struct intel_plane_state *state); - int (*update_colorkey)(struct drm_plane *plane, - struct drm_intel_sprite_colorkey *key); - void (*get_colorkey)(struct drm_plane *plane, - struct drm_intel_sprite_colorkey *key); }; struct intel_watermark_params { diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c index cd531fa997de..d61e35d32e33 100644 --- a/drivers/gpu/drm/i915/intel_sprite.c +++ b/drivers/gpu/drm/i915/intel_sprite.c @@ -192,6 +192,7 @@ skl_update_plane(struct drm_plane *drm_plane, struct drm_crtc *crtc, const int plane = intel_plane->plane + 1; u32 plane_ctl, stride_div; int pixel_size = drm_format_plane_cpp(fb->pixel_format, 0); + const struct drm_intel_sprite_colorkey *key = &intel_plane->ckey; plane_ctl = I915_READ(PLANE_CTL(pipe, plane)); @@ -202,6 +203,7 @@ skl_update_plane(struct drm_plane *drm_plane, struct drm_crtc *crtc, plane_ctl &= ~PLANE_CTL_TILED_MASK; plane_ctl &= ~PLANE_CTL_ALPHA_MASK; plane_ctl &= ~PLANE_CTL_ROTATE_MASK; + plane_ctl &= ~PLANE_CTL_KEY_ENABLE_MASK; /* Trickle feed has to be enabled */ plane_ctl &= ~PLANE_CTL_TRICKLE_FEED_DISABLE; @@ -281,6 +283,17 @@ skl_update_plane(struct drm_plane *drm_plane, struct drm_crtc *crtc, crtc_w--; crtc_h--; + if (key->flags) { + I915_WRITE(PLANE_KEYVAL(pipe, plane), key->min_value); + I915_WRITE(PLANE_KEYMAX(pipe, plane), key->max_value); + I915_WRITE(PLANE_KEYMSK(pipe, plane), key->channel_mask); + } + + if (key->flags & I915_SET_COLORKEY_DESTINATION) + plane_ctl |= PLANE_CTL_KEY_ENABLE_DESTINATION; + else if (key->flags & I915_SET_COLORKEY_SOURCE) + plane_ctl |= PLANE_CTL_KEY_ENABLE_SOURCE; + I915_WRITE(PLANE_OFFSET(pipe, plane), (y << 16) | x); I915_WRITE(PLANE_STRIDE(pipe, plane), fb->pitches[0] / stride_div); I915_WRITE(PLANE_POS(pipe, plane), (crtc_y << 16) | crtc_x); @@ -309,63 +322,6 @@ skl_disable_plane(struct drm_plane *drm_plane, struct drm_crtc *crtc) intel_update_sprite_watermarks(drm_plane, crtc, 0, 0, 0, false, false); } -static int -skl_update_colorkey(struct drm_plane *drm_plane, - struct drm_intel_sprite_colorkey *key) -{ - struct drm_device *dev = drm_plane->dev; - struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_plane *intel_plane = to_intel_plane(drm_plane); - const int pipe = intel_plane->pipe; - const int plane = intel_plane->plane; - u32 plane_ctl; - - I915_WRITE(PLANE_KEYVAL(pipe, plane), key->min_value); - I915_WRITE(PLANE_KEYMAX(pipe, plane), key->max_value); - I915_WRITE(PLANE_KEYMSK(pipe, plane), key->channel_mask); - - plane_ctl = I915_READ(PLANE_CTL(pipe, plane)); - plane_ctl &= ~PLANE_CTL_KEY_ENABLE_MASK; - if (key->flags & I915_SET_COLORKEY_DESTINATION) - plane_ctl |= PLANE_CTL_KEY_ENABLE_DESTINATION; - else if (key->flags & I915_SET_COLORKEY_SOURCE) - plane_ctl |= PLANE_CTL_KEY_ENABLE_SOURCE; - I915_WRITE(PLANE_CTL(pipe, plane), plane_ctl); - - POSTING_READ(PLANE_CTL(pipe, plane)); - - return 0; -} - -static void -skl_get_colorkey(struct drm_plane *drm_plane, - struct drm_intel_sprite_colorkey *key) -{ - struct drm_device *dev = drm_plane->dev; - struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_plane *intel_plane = to_intel_plane(drm_plane); - const int pipe = intel_plane->pipe; - const int plane = intel_plane->plane; - u32 plane_ctl; - - key->min_value = I915_READ(PLANE_KEYVAL(pipe, plane)); - key->max_value = I915_READ(PLANE_KEYMAX(pipe, plane)); - key->channel_mask = I915_READ(PLANE_KEYMSK(pipe, plane)); - - plane_ctl = I915_READ(PLANE_CTL(pipe, plane)); - - switch (plane_ctl & PLANE_CTL_KEY_ENABLE_MASK) { - case PLANE_CTL_KEY_ENABLE_DESTINATION: - key->flags = I915_SET_COLORKEY_DESTINATION; - break; - case PLANE_CTL_KEY_ENABLE_SOURCE: - key->flags = I915_SET_COLORKEY_SOURCE; - break; - default: - key->flags = I915_SET_COLORKEY_NONE; - } -} - static void chv_update_csc(struct intel_plane *intel_plane, uint32_t format) { @@ -423,6 +379,7 @@ vlv_update_plane(struct drm_plane *dplane, struct drm_crtc *crtc, u32 sprctl; unsigned long sprsurf_offset, linear_offset; int pixel_size = drm_format_plane_cpp(fb->pixel_format, 0); + const struct drm_intel_sprite_colorkey *key = &intel_plane->ckey; sprctl = I915_READ(SPCNTR(pipe, plane)); @@ -431,6 +388,7 @@ vlv_update_plane(struct drm_plane *dplane, struct drm_crtc *crtc, sprctl &= ~SP_YUV_BYTE_ORDER_MASK; sprctl &= ~SP_TILED; sprctl &= ~SP_ROTATE_180; + sprctl &= ~SP_SOURCE_KEY; switch (fb->pixel_format) { case DRM_FORMAT_YUYV: @@ -513,6 +471,15 @@ vlv_update_plane(struct drm_plane *dplane, struct drm_crtc *crtc, intel_update_primary_plane(intel_crtc); + if (key->flags) { + I915_WRITE(SPKEYMINVAL(pipe, plane), key->min_value); + I915_WRITE(SPKEYMAXVAL(pipe, plane), key->max_value); + I915_WRITE(SPKEYMSK(pipe, plane), key->channel_mask); + } + + if (key->flags & I915_SET_COLORKEY_SOURCE) + sprctl |= SP_SOURCE_KEY; + if (IS_CHERRYVIEW(dev) && pipe == PIPE_B) chv_update_csc(intel_plane, fb->pixel_format); @@ -556,56 +523,6 @@ vlv_disable_plane(struct drm_plane *dplane, struct drm_crtc *crtc) intel_update_sprite_watermarks(dplane, crtc, 0, 0, 0, false, false); } -static int -vlv_update_colorkey(struct drm_plane *dplane, - struct drm_intel_sprite_colorkey *key) -{ - struct drm_device *dev = dplane->dev; - struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_plane *intel_plane = to_intel_plane(dplane); - int pipe = intel_plane->pipe; - int plane = intel_plane->plane; - u32 sprctl; - - if (key->flags & I915_SET_COLORKEY_DESTINATION) - return -EINVAL; - - I915_WRITE(SPKEYMINVAL(pipe, plane), key->min_value); - I915_WRITE(SPKEYMAXVAL(pipe, plane), key->max_value); - I915_WRITE(SPKEYMSK(pipe, plane), key->channel_mask); - - sprctl = I915_READ(SPCNTR(pipe, plane)); - sprctl &= ~SP_SOURCE_KEY; - if (key->flags & I915_SET_COLORKEY_SOURCE) - sprctl |= SP_SOURCE_KEY; - I915_WRITE(SPCNTR(pipe, plane), sprctl); - - POSTING_READ(SPKEYMSK(pipe, plane)); - - return 0; -} - -static void -vlv_get_colorkey(struct drm_plane *dplane, - struct drm_intel_sprite_colorkey *key) -{ - struct drm_device *dev = dplane->dev; - struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_plane *intel_plane = to_intel_plane(dplane); - int pipe = intel_plane->pipe; - int plane = intel_plane->plane; - u32 sprctl; - - key->min_value = I915_READ(SPKEYMINVAL(pipe, plane)); - key->max_value = I915_READ(SPKEYMAXVAL(pipe, plane)); - key->channel_mask = I915_READ(SPKEYMSK(pipe, plane)); - - sprctl = I915_READ(SPCNTR(pipe, plane)); - if (sprctl & SP_SOURCE_KEY) - key->flags = I915_SET_COLORKEY_SOURCE; - else - key->flags = I915_SET_COLORKEY_NONE; -} static void ivb_update_plane(struct drm_plane *plane, struct drm_crtc *crtc, @@ -620,10 +537,11 @@ ivb_update_plane(struct drm_plane *plane, struct drm_crtc *crtc, struct intel_plane *intel_plane = to_intel_plane(plane); struct intel_crtc *intel_crtc = to_intel_crtc(crtc); struct drm_i915_gem_object *obj = intel_fb_obj(fb); - int pipe = intel_plane->pipe; + enum pipe pipe = intel_plane->pipe; u32 sprctl, sprscale = 0; unsigned long sprsurf_offset, linear_offset; int pixel_size = drm_format_plane_cpp(fb->pixel_format, 0); + const struct drm_intel_sprite_colorkey *key = &intel_plane->ckey; sprctl = I915_READ(SPRCTL(pipe)); @@ -633,6 +551,7 @@ ivb_update_plane(struct drm_plane *plane, struct drm_crtc *crtc, sprctl &= ~SPRITE_YUV_BYTE_ORDER_MASK; sprctl &= ~SPRITE_TILED; sprctl &= ~SPRITE_ROTATE_180; + sprctl &= ~(SPRITE_SOURCE_KEY | SPRITE_DEST_KEY); switch (fb->pixel_format) { case DRM_FORMAT_XBGR8888: @@ -709,6 +628,17 @@ ivb_update_plane(struct drm_plane *plane, struct drm_crtc *crtc, intel_update_primary_plane(intel_crtc); + if (key->flags) { + I915_WRITE(SPRKEYVAL(pipe), key->min_value); + I915_WRITE(SPRKEYMAX(pipe), key->max_value); + I915_WRITE(SPRKEYMSK(pipe), key->channel_mask); + } + + if (key->flags & I915_SET_COLORKEY_DESTINATION) + sprctl |= SPRITE_DEST_KEY; + else if (key->flags & I915_SET_COLORKEY_SOURCE) + sprctl |= SPRITE_SOURCE_KEY; + I915_WRITE(SPRSTRIDE(pipe), fb->pitches[0]); I915_WRITE(SPRPOS(pipe), (crtc_y << 16) | crtc_x); @@ -752,60 +682,6 @@ ivb_disable_plane(struct drm_plane *plane, struct drm_crtc *crtc) intel_flush_primary_plane(dev_priv, intel_crtc->plane); } -static int -ivb_update_colorkey(struct drm_plane *plane, - struct drm_intel_sprite_colorkey *key) -{ - struct drm_device *dev = plane->dev; - struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_plane *intel_plane; - u32 sprctl; - int ret = 0; - - intel_plane = to_intel_plane(plane); - - I915_WRITE(SPRKEYVAL(intel_plane->pipe), key->min_value); - I915_WRITE(SPRKEYMAX(intel_plane->pipe), key->max_value); - I915_WRITE(SPRKEYMSK(intel_plane->pipe), key->channel_mask); - - sprctl = I915_READ(SPRCTL(intel_plane->pipe)); - sprctl &= ~(SPRITE_SOURCE_KEY | SPRITE_DEST_KEY); - if (key->flags & I915_SET_COLORKEY_DESTINATION) - sprctl |= SPRITE_DEST_KEY; - else if (key->flags & I915_SET_COLORKEY_SOURCE) - sprctl |= SPRITE_SOURCE_KEY; - I915_WRITE(SPRCTL(intel_plane->pipe), sprctl); - - POSTING_READ(SPRKEYMSK(intel_plane->pipe)); - - return ret; -} - -static void -ivb_get_colorkey(struct drm_plane *plane, struct drm_intel_sprite_colorkey *key) -{ - struct drm_device *dev = plane->dev; - struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_plane *intel_plane; - u32 sprctl; - - intel_plane = to_intel_plane(plane); - - key->min_value = I915_READ(SPRKEYVAL(intel_plane->pipe)); - key->max_value = I915_READ(SPRKEYMAX(intel_plane->pipe)); - key->channel_mask = I915_READ(SPRKEYMSK(intel_plane->pipe)); - key->flags = 0; - - sprctl = I915_READ(SPRCTL(intel_plane->pipe)); - - if (sprctl & SPRITE_DEST_KEY) - key->flags = I915_SET_COLORKEY_DESTINATION; - else if (sprctl & SPRITE_SOURCE_KEY) - key->flags = I915_SET_COLORKEY_SOURCE; - else - key->flags = I915_SET_COLORKEY_NONE; -} - static void ilk_update_plane(struct drm_plane *plane, struct drm_crtc *crtc, struct drm_framebuffer *fb, @@ -823,6 +699,7 @@ ilk_update_plane(struct drm_plane *plane, struct drm_crtc *crtc, unsigned long dvssurf_offset, linear_offset; u32 dvscntr, dvsscale; int pixel_size = drm_format_plane_cpp(fb->pixel_format, 0); + const struct drm_intel_sprite_colorkey *key = &intel_plane->ckey; dvscntr = I915_READ(DVSCNTR(pipe)); @@ -832,6 +709,7 @@ ilk_update_plane(struct drm_plane *plane, struct drm_crtc *crtc, dvscntr &= ~DVS_YUV_BYTE_ORDER_MASK; dvscntr &= ~DVS_TILED; dvscntr &= ~DVS_ROTATE_180; + dvscntr &= ~(DVS_SOURCE_KEY | DVS_DEST_KEY); switch (fb->pixel_format) { case DRM_FORMAT_XBGR8888: @@ -899,6 +777,17 @@ ilk_update_plane(struct drm_plane *plane, struct drm_crtc *crtc, intel_update_primary_plane(intel_crtc); + if (key->flags) { + I915_WRITE(DVSKEYVAL(pipe), key->min_value); + I915_WRITE(DVSKEYMAX(pipe), key->max_value); + I915_WRITE(DVSKEYMSK(pipe), key->channel_mask); + } + + if (key->flags & I915_SET_COLORKEY_DESTINATION) + dvscntr |= DVS_DEST_KEY; + else if (key->flags & I915_SET_COLORKEY_SOURCE) + dvscntr |= DVS_SOURCE_KEY; + I915_WRITE(DVSSTRIDE(pipe), fb->pitches[0]); I915_WRITE(DVSPOS(pipe), (crtc_y << 16) | crtc_x); @@ -1004,67 +893,9 @@ intel_pre_disable_primary(struct drm_crtc *crtc) hsw_disable_ips(intel_crtc); } -static int -ilk_update_colorkey(struct drm_plane *plane, - struct drm_intel_sprite_colorkey *key) -{ - struct drm_device *dev = plane->dev; - struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_plane *intel_plane; - u32 dvscntr; - int ret = 0; - - intel_plane = to_intel_plane(plane); - - I915_WRITE(DVSKEYVAL(intel_plane->pipe), key->min_value); - I915_WRITE(DVSKEYMAX(intel_plane->pipe), key->max_value); - I915_WRITE(DVSKEYMSK(intel_plane->pipe), key->channel_mask); - - dvscntr = I915_READ(DVSCNTR(intel_plane->pipe)); - dvscntr &= ~(DVS_SOURCE_KEY | DVS_DEST_KEY); - if (key->flags & I915_SET_COLORKEY_DESTINATION) - dvscntr |= DVS_DEST_KEY; - else if (key->flags & I915_SET_COLORKEY_SOURCE) - dvscntr |= DVS_SOURCE_KEY; - I915_WRITE(DVSCNTR(intel_plane->pipe), dvscntr); - - POSTING_READ(DVSKEYMSK(intel_plane->pipe)); - - return ret; -} - -static void -ilk_get_colorkey(struct drm_plane *plane, struct drm_intel_sprite_colorkey *key) -{ - struct drm_device *dev = plane->dev; - struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_plane *intel_plane; - u32 dvscntr; - - intel_plane = to_intel_plane(plane); - - key->min_value = I915_READ(DVSKEYVAL(intel_plane->pipe)); - key->max_value = I915_READ(DVSKEYMAX(intel_plane->pipe)); - key->channel_mask = I915_READ(DVSKEYMSK(intel_plane->pipe)); - key->flags = 0; - - dvscntr = I915_READ(DVSCNTR(intel_plane->pipe)); - - if (dvscntr & DVS_DEST_KEY) - key->flags = I915_SET_COLORKEY_DESTINATION; - else if (dvscntr & DVS_SOURCE_KEY) - key->flags = I915_SET_COLORKEY_SOURCE; - else - key->flags = I915_SET_COLORKEY_NONE; -} - static bool colorkey_enabled(struct intel_plane *intel_plane) { - struct drm_intel_sprite_colorkey key; - - intel_plane->get_colorkey(&intel_plane->base, &key); - - return key.flags != I915_SET_COLORKEY_NONE; + return intel_plane->ckey.flags != I915_SET_COLORKEY_NONE; } static int @@ -1317,6 +1148,10 @@ int intel_sprite_set_colorkey(struct drm_device *dev, void *data, if ((set->flags & (I915_SET_COLORKEY_DESTINATION | I915_SET_COLORKEY_SOURCE)) == (I915_SET_COLORKEY_DESTINATION | I915_SET_COLORKEY_SOURCE)) return -EINVAL; + if (IS_VALLEYVIEW(dev) && + set->flags & I915_SET_COLORKEY_DESTINATION) + return -EINVAL; + drm_modeset_lock_all(dev); plane = drm_plane_find(dev, set->plane_id); @@ -1326,7 +1161,15 @@ int intel_sprite_set_colorkey(struct drm_device *dev, void *data, } intel_plane = to_intel_plane(plane); - ret = intel_plane->update_colorkey(plane, set); + intel_plane->ckey = *set; + + /* + * The only way this could fail would be due to + * the current plane state being unsupportable already, + * and we dont't consider that an error for the + * colorkey ioctl. So just ignore any error. + */ + intel_plane_restore(plane); out_unlock: drm_modeset_unlock_all(dev); @@ -1350,7 +1193,7 @@ int intel_sprite_get_colorkey(struct drm_device *dev, void *data, } intel_plane = to_intel_plane(plane); - intel_plane->get_colorkey(plane, get); + *get = intel_plane->ckey; out_unlock: drm_modeset_unlock_all(dev); @@ -1443,8 +1286,6 @@ intel_plane_init(struct drm_device *dev, enum pipe pipe, int plane) intel_plane->max_downscale = 16; intel_plane->update_plane = ilk_update_plane; intel_plane->disable_plane = ilk_disable_plane; - intel_plane->update_colorkey = ilk_update_colorkey; - intel_plane->get_colorkey = ilk_get_colorkey; if (IS_GEN6(dev)) { plane_formats = snb_plane_formats; @@ -1468,16 +1309,12 @@ intel_plane_init(struct drm_device *dev, enum pipe pipe, int plane) if (IS_VALLEYVIEW(dev)) { intel_plane->update_plane = vlv_update_plane; intel_plane->disable_plane = vlv_disable_plane; - intel_plane->update_colorkey = vlv_update_colorkey; - intel_plane->get_colorkey = vlv_get_colorkey; plane_formats = vlv_plane_formats; num_plane_formats = ARRAY_SIZE(vlv_plane_formats); } else { intel_plane->update_plane = ivb_update_plane; intel_plane->disable_plane = ivb_disable_plane; - intel_plane->update_colorkey = ivb_update_colorkey; - intel_plane->get_colorkey = ivb_get_colorkey; plane_formats = snb_plane_formats; num_plane_formats = ARRAY_SIZE(snb_plane_formats); @@ -1492,8 +1329,6 @@ intel_plane_init(struct drm_device *dev, enum pipe pipe, int plane) intel_plane->max_downscale = 1; intel_plane->update_plane = skl_update_plane; intel_plane->disable_plane = skl_disable_plane; - intel_plane->update_colorkey = skl_update_colorkey; - intel_plane->get_colorkey = skl_get_colorkey; plane_formats = skl_plane_formats; num_plane_formats = ARRAY_SIZE(skl_plane_formats); -- cgit v1.2.1 From 48fe4691ae639e60fda37faf06dccdff60245149 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 19 Mar 2015 17:57:13 +0200 Subject: drm/i915: Eliminate plane control register RMW from sprite code MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Replace the RMW access with explicit initialization of the entire plane control register, as was done for primary planes in: commit f45651bae2ee73ae551699d481f76aa6ad92138f Author: Ville Syrjälä Date: Fri Aug 8 21:51:10 2014 +0300 drm/i915: Eliminate rmw from .update_primary_plane() The automagic primary plane disable is still doing RMWs, but that will require more work to untangle, so leave it alone for now. Signed-off-by: Ville Syrjälä Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_sprite.c | 62 ++++++------------------------------- 1 file changed, 10 insertions(+), 52 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c index d61e35d32e33..a4c9c66d7e79 100644 --- a/drivers/gpu/drm/i915/intel_sprite.c +++ b/drivers/gpu/drm/i915/intel_sprite.c @@ -194,19 +194,8 @@ skl_update_plane(struct drm_plane *drm_plane, struct drm_crtc *crtc, int pixel_size = drm_format_plane_cpp(fb->pixel_format, 0); const struct drm_intel_sprite_colorkey *key = &intel_plane->ckey; - plane_ctl = I915_READ(PLANE_CTL(pipe, plane)); - - /* Mask out pixel format bits in case we change it */ - plane_ctl &= ~PLANE_CTL_FORMAT_MASK; - plane_ctl &= ~PLANE_CTL_ORDER_RGBX; - plane_ctl &= ~PLANE_CTL_YUV422_ORDER_MASK; - plane_ctl &= ~PLANE_CTL_TILED_MASK; - plane_ctl &= ~PLANE_CTL_ALPHA_MASK; - plane_ctl &= ~PLANE_CTL_ROTATE_MASK; - plane_ctl &= ~PLANE_CTL_KEY_ENABLE_MASK; - - /* Trickle feed has to be enabled */ - plane_ctl &= ~PLANE_CTL_TRICKLE_FEED_DISABLE; + plane_ctl = PLANE_CTL_ENABLE | + PLANE_CTL_PIPE_CSC_ENABLE; switch (fb->pixel_format) { case DRM_FORMAT_RGB565: @@ -267,9 +256,6 @@ skl_update_plane(struct drm_plane *drm_plane, struct drm_crtc *crtc, if (drm_plane->state->rotation == BIT(DRM_ROTATE_180)) plane_ctl |= PLANE_CTL_ROTATE_180; - plane_ctl |= PLANE_CTL_ENABLE; - plane_ctl |= PLANE_CTL_PIPE_CSC_ENABLE; - intel_update_sprite_watermarks(drm_plane, crtc, src_w, src_h, pixel_size, true, src_w != crtc_w || src_h != crtc_h); @@ -312,8 +298,7 @@ skl_disable_plane(struct drm_plane *drm_plane, struct drm_crtc *crtc) const int pipe = intel_plane->pipe; const int plane = intel_plane->plane + 1; - I915_WRITE(PLANE_CTL(pipe, plane), - I915_READ(PLANE_CTL(pipe, plane)) & ~PLANE_CTL_ENABLE); + I915_WRITE(PLANE_CTL(pipe, plane), 0); /* Activate double buffered register update */ I915_WRITE(PLANE_CTL(pipe, plane), 0); @@ -381,14 +366,7 @@ vlv_update_plane(struct drm_plane *dplane, struct drm_crtc *crtc, int pixel_size = drm_format_plane_cpp(fb->pixel_format, 0); const struct drm_intel_sprite_colorkey *key = &intel_plane->ckey; - sprctl = I915_READ(SPCNTR(pipe, plane)); - - /* Mask out pixel format bits in case we change it */ - sprctl &= ~SP_PIXFORMAT_MASK; - sprctl &= ~SP_YUV_BYTE_ORDER_MASK; - sprctl &= ~SP_TILED; - sprctl &= ~SP_ROTATE_180; - sprctl &= ~SP_SOURCE_KEY; + sprctl = SP_ENABLE; switch (fb->pixel_format) { case DRM_FORMAT_YUYV: @@ -442,8 +420,6 @@ vlv_update_plane(struct drm_plane *dplane, struct drm_crtc *crtc, if (obj->tiling_mode != I915_TILING_NONE) sprctl |= SP_TILED; - sprctl |= SP_ENABLE; - intel_update_sprite_watermarks(dplane, crtc, src_w, src_h, pixel_size, true, src_w != crtc_w || src_h != crtc_h); @@ -513,8 +489,8 @@ vlv_disable_plane(struct drm_plane *dplane, struct drm_crtc *crtc) intel_update_primary_plane(intel_crtc); - I915_WRITE(SPCNTR(pipe, plane), I915_READ(SPCNTR(pipe, plane)) & - ~SP_ENABLE); + I915_WRITE(SPCNTR(pipe, plane), 0); + /* Activate double buffered register update */ I915_WRITE(SPSURF(pipe, plane), 0); @@ -543,15 +519,7 @@ ivb_update_plane(struct drm_plane *plane, struct drm_crtc *crtc, int pixel_size = drm_format_plane_cpp(fb->pixel_format, 0); const struct drm_intel_sprite_colorkey *key = &intel_plane->ckey; - sprctl = I915_READ(SPRCTL(pipe)); - - /* Mask out pixel format bits in case we change it */ - sprctl &= ~SPRITE_PIXFORMAT_MASK; - sprctl &= ~SPRITE_RGB_ORDER_RGBX; - sprctl &= ~SPRITE_YUV_BYTE_ORDER_MASK; - sprctl &= ~SPRITE_TILED; - sprctl &= ~SPRITE_ROTATE_180; - sprctl &= ~(SPRITE_SOURCE_KEY | SPRITE_DEST_KEY); + sprctl = SPRITE_ENABLE; switch (fb->pixel_format) { case DRM_FORMAT_XBGR8888: @@ -590,8 +558,6 @@ ivb_update_plane(struct drm_plane *plane, struct drm_crtc *crtc, else sprctl |= SPRITE_TRICKLE_FEED_DISABLE; - sprctl |= SPRITE_ENABLE; - if (IS_HASWELL(dev) || IS_BROADWELL(dev)) sprctl |= SPRITE_PIPE_CSC_ENABLE; @@ -701,15 +667,7 @@ ilk_update_plane(struct drm_plane *plane, struct drm_crtc *crtc, int pixel_size = drm_format_plane_cpp(fb->pixel_format, 0); const struct drm_intel_sprite_colorkey *key = &intel_plane->ckey; - dvscntr = I915_READ(DVSCNTR(pipe)); - - /* Mask out pixel format bits in case we change it */ - dvscntr &= ~DVS_PIXFORMAT_MASK; - dvscntr &= ~DVS_RGB_ORDER_XBGR; - dvscntr &= ~DVS_YUV_BYTE_ORDER_MASK; - dvscntr &= ~DVS_TILED; - dvscntr &= ~DVS_ROTATE_180; - dvscntr &= ~(DVS_SOURCE_KEY | DVS_DEST_KEY); + dvscntr = DVS_ENABLE; switch (fb->pixel_format) { case DRM_FORMAT_XBGR8888: @@ -745,7 +703,6 @@ ilk_update_plane(struct drm_plane *plane, struct drm_crtc *crtc, if (IS_GEN6(dev)) dvscntr |= DVS_TRICKLE_FEED_DISABLE; /* must disable */ - dvscntr |= DVS_ENABLE; intel_update_sprite_watermarks(plane, crtc, src_w, src_h, pixel_size, true, @@ -816,9 +773,10 @@ ilk_disable_plane(struct drm_plane *plane, struct drm_crtc *crtc) intel_update_primary_plane(intel_crtc); - I915_WRITE(DVSCNTR(pipe), I915_READ(DVSCNTR(pipe)) & ~DVS_ENABLE); + I915_WRITE(DVSCNTR(pipe), 0); /* Disable the scaler */ I915_WRITE(DVSSCALE(pipe), 0); + /* Flush double buffered register updates */ I915_WRITE(DVSSURF(pipe), 0); -- cgit v1.2.1 From 2ddc1dad393d39ee6218e545e37e57f3a9dc050f Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Thu, 19 Mar 2015 17:57:14 +0200 Subject: drm/i915: Fix SKL sprite disable double buffer register update MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Write the PLANE_SURF register instead of PLANE_CTL to arm the double buffer regisrter update. Signed-off-by: Ville Syrjälä Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_sprite.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c index a4c9c66d7e79..0f00209bb220 100644 --- a/drivers/gpu/drm/i915/intel_sprite.c +++ b/drivers/gpu/drm/i915/intel_sprite.c @@ -301,8 +301,8 @@ skl_disable_plane(struct drm_plane *drm_plane, struct drm_crtc *crtc) I915_WRITE(PLANE_CTL(pipe, plane), 0); /* Activate double buffered register update */ - I915_WRITE(PLANE_CTL(pipe, plane), 0); - POSTING_READ(PLANE_CTL(pipe, plane)); + I915_WRITE(PLANE_SURF(pipe, plane), 0); + POSTING_READ(PLANE_SURF(pipe, plane)); intel_update_sprite_watermarks(drm_plane, crtc, 0, 0, 0, false, false); } -- cgit v1.2.1 From 3d8c6dce53a349df8878d078e56bf429bad572f9 Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Fri, 20 Mar 2015 13:56:06 +0100 Subject: netfilter: xt_TPROXY: fix invflags check in tproxy_tg6_check() We have to check for IP6T_INV_PROTO in invflags, instead of flags. Signed-off-by: Pablo Neira Ayuso Acked-by: Balazs Scheidler --- net/netfilter/xt_TPROXY.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/net/netfilter/xt_TPROXY.c b/net/netfilter/xt_TPROXY.c index ef8a926752a9..50e1e5aaf4ce 100644 --- a/net/netfilter/xt_TPROXY.c +++ b/net/netfilter/xt_TPROXY.c @@ -513,8 +513,8 @@ static int tproxy_tg6_check(const struct xt_tgchk_param *par) { const struct ip6t_ip6 *i = par->entryinfo; - if ((i->proto == IPPROTO_TCP || i->proto == IPPROTO_UDP) - && !(i->flags & IP6T_INV_PROTO)) + if ((i->proto == IPPROTO_TCP || i->proto == IPPROTO_UDP) && + !(i->invflags & IP6T_INV_PROTO)) return 0; pr_info("Can be used only in combination with " -- cgit v1.2.1 From 9e8ce4b96b781b003e3174fbbc62e1d4388c8b8f Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Fri, 20 Mar 2015 14:56:19 +0100 Subject: Revert "x86/PCI: Refine the way to release PCI IRQ resources" Commit b4b55cda5874 (Refine the way to release PCI IRQ resources) introduced a regression in the PCI IRQ resource management by causing the IRQ resource of a device, established when pci_enabled_device() is called on a fully disabled device, to be released when the driver is unbound from the device, regardless of the enable_cnt. This leads to the situation that an ill-behaved driver can now make a device unusable to subsequent drivers by an imbalance in their use of pci_enable/disable_device(). That is a serious problem for secondary drivers like vfio-pci, which are innocent of the transgressions of the previous driver. Since the solution of this problem is not immediate and requires further discussion, revert commit b4b55cda5874 and the issue it was supposed to address (a bug related to xen-pciback) will be taken care of in a different way going forward. Reported-by: Alex Williamson Signed-off-by: Rafael J. Wysocki --- arch/x86/include/asm/pci_x86.h | 2 ++ arch/x86/pci/common.c | 34 ++++++---------------------------- arch/x86/pci/intel_mid_pci.c | 4 ++-- arch/x86/pci/irq.c | 15 ++++++++++++++- drivers/acpi/pci_irq.c | 9 ++++++++- 5 files changed, 32 insertions(+), 32 deletions(-) diff --git a/arch/x86/include/asm/pci_x86.h b/arch/x86/include/asm/pci_x86.h index fa1195dae425..164e3f8d3c3d 100644 --- a/arch/x86/include/asm/pci_x86.h +++ b/arch/x86/include/asm/pci_x86.h @@ -93,6 +93,8 @@ extern raw_spinlock_t pci_config_lock; extern int (*pcibios_enable_irq)(struct pci_dev *dev); extern void (*pcibios_disable_irq)(struct pci_dev *dev); +extern bool mp_should_keep_irq(struct device *dev); + struct pci_raw_ops { int (*read)(unsigned int domain, unsigned int bus, unsigned int devfn, int reg, int len, u32 *val); diff --git a/arch/x86/pci/common.c b/arch/x86/pci/common.c index 3d2612b68694..2fb384724ebb 100644 --- a/arch/x86/pci/common.c +++ b/arch/x86/pci/common.c @@ -513,31 +513,6 @@ void __init pcibios_set_cache_line_size(void) } } -/* - * Some device drivers assume dev->irq won't change after calling - * pci_disable_device(). So delay releasing of IRQ resource to driver - * unbinding time. Otherwise it will break PM subsystem and drivers - * like xen-pciback etc. - */ -static int pci_irq_notifier(struct notifier_block *nb, unsigned long action, - void *data) -{ - struct pci_dev *dev = to_pci_dev(data); - - if (action != BUS_NOTIFY_UNBOUND_DRIVER) - return NOTIFY_DONE; - - if (pcibios_disable_irq) - pcibios_disable_irq(dev); - - return NOTIFY_OK; -} - -static struct notifier_block pci_irq_nb = { - .notifier_call = pci_irq_notifier, - .priority = INT_MIN, -}; - int __init pcibios_init(void) { if (!raw_pci_ops) { @@ -550,9 +525,6 @@ int __init pcibios_init(void) if (pci_bf_sort >= pci_force_bf) pci_sort_breadthfirst(); - - bus_register_notifier(&pci_bus_type, &pci_irq_nb); - return 0; } @@ -711,6 +683,12 @@ int pcibios_enable_device(struct pci_dev *dev, int mask) return 0; } +void pcibios_disable_device (struct pci_dev *dev) +{ + if (!pci_dev_msi_enabled(dev) && pcibios_disable_irq) + pcibios_disable_irq(dev); +} + int pci_ext_cfg_avail(void) { if (raw_pci_ext_ops) diff --git a/arch/x86/pci/intel_mid_pci.c b/arch/x86/pci/intel_mid_pci.c index efb849323c74..852aa4c92da0 100644 --- a/arch/x86/pci/intel_mid_pci.c +++ b/arch/x86/pci/intel_mid_pci.c @@ -234,10 +234,10 @@ static int intel_mid_pci_irq_enable(struct pci_dev *dev) static void intel_mid_pci_irq_disable(struct pci_dev *dev) { - if (dev->irq_managed && dev->irq > 0) { + if (!mp_should_keep_irq(&dev->dev) && dev->irq_managed && + dev->irq > 0) { mp_unmap_irq(dev->irq); dev->irq_managed = 0; - dev->irq = 0; } } diff --git a/arch/x86/pci/irq.c b/arch/x86/pci/irq.c index e71b3dbd87b8..5dc6ca5e1741 100644 --- a/arch/x86/pci/irq.c +++ b/arch/x86/pci/irq.c @@ -1256,9 +1256,22 @@ static int pirq_enable_irq(struct pci_dev *dev) return 0; } +bool mp_should_keep_irq(struct device *dev) +{ + if (dev->power.is_prepared) + return true; +#ifdef CONFIG_PM + if (dev->power.runtime_status == RPM_SUSPENDING) + return true; +#endif + + return false; +} + static void pirq_disable_irq(struct pci_dev *dev) { - if (io_apic_assign_pci_irqs && dev->irq_managed && dev->irq) { + if (io_apic_assign_pci_irqs && !mp_should_keep_irq(&dev->dev) && + dev->irq_managed && dev->irq) { mp_unmap_irq(dev->irq); dev->irq = 0; dev->irq_managed = 0; diff --git a/drivers/acpi/pci_irq.c b/drivers/acpi/pci_irq.c index e7f718d6918a..b1def411c0b8 100644 --- a/drivers/acpi/pci_irq.c +++ b/drivers/acpi/pci_irq.c @@ -485,6 +485,14 @@ void acpi_pci_irq_disable(struct pci_dev *dev) if (!pin || !dev->irq_managed || dev->irq <= 0) return; + /* Keep IOAPIC pin configuration when suspending */ + if (dev->dev.power.is_prepared) + return; +#ifdef CONFIG_PM + if (dev->dev.power.runtime_status == RPM_SUSPENDING) + return; +#endif + entry = acpi_pci_irq_lookup(dev, pin); if (!entry) return; @@ -505,6 +513,5 @@ void acpi_pci_irq_disable(struct pci_dev *dev) if (gsi >= 0) { acpi_unregister_gsi(gsi); dev->irq_managed = 0; - dev->irq = 0; } } -- cgit v1.2.1 From 0b37a9a9ebdf47dc9d5290a1c69bd944dde5cc15 Mon Sep 17 00:00:00 2001 From: Michel Thierry Date: Fri, 20 Mar 2015 09:41:03 +0000 Subject: drm/i915: Do not leak objects after capturing error state While running kmemleak chasing a different memleak, I saw that the capture_error_state function was leaking some objects, for example: unreferenced object 0xffff8800a9b72148 (size 8192): comm "kworker/u16:0", pid 1499, jiffies 4295201243 (age 990.096s) hex dump (first 32 bytes): 00 00 04 00 00 00 00 00 5d f4 ff ff 00 00 00 00 ........]....... 00 30 b0 01 00 00 00 00 37 00 00 00 00 00 00 00 .0......7....... backtrace: [] create_object+0x104/0x2c0 [] kmemleak_alloc+0x7a/0xc0 [] __kmalloc+0xeb/0x220 [] kcalloc.constprop.12+0x2d/0x2f [i915] [] i915_capture_error_state+0x3f4/0x1660 [i915] [] i915_handle_error+0x7f/0x660 [i915] [] i915_hangcheck_elapsed+0x2e7/0x470 [i915] [] process_one_work+0x144/0x490 [] worker_thread+0x11d/0x530 [] kthread+0xc9/0xe0 [] ret_from_fork+0x58/0x90 [] 0xffffffffffffffff The following objects are allocated in i915_gem_capture_buffers, but not released in i915_error_state_free: - error->active_bo_count - error->pinned_bo - error->pinned_bo_count - error->active_bo[vm_count] (allocated in i915_gem_capture_vm). The leaks were introduced by commit 95f5301dd880da2dea2c9a9c29750064536d426a Author: Ben Widawsky Date: Wed Jul 31 17:00:15 2013 -0700 drm/i915: Update error capture for VMs v2: Reuse iterator and add culprit commit details (Chris) Cc: Chris Wilson Signed-off-by: Michel Thierry Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_gpu_error.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c index a982849a5edd..2f7cbd3d5524 100644 --- a/drivers/gpu/drm/i915/i915_gpu_error.c +++ b/drivers/gpu/drm/i915/i915_gpu_error.c @@ -555,7 +555,14 @@ static void i915_error_state_free(struct kref *error_ref) } i915_error_object_free(error->semaphore_obj); + + for (i = 0; i < error->vm_count; i++) + kfree(error->active_bo[i]); + kfree(error->active_bo); + kfree(error->active_bo_count); + kfree(error->pinned_bo); + kfree(error->pinned_bo_count); kfree(error->overlay); kfree(error->display); kfree(error); -- cgit v1.2.1 From 7ee8e4f3983c4ff700958a6099c8fd212ea67b94 Mon Sep 17 00:00:00 2001 From: Wenbo Wang Date: Fri, 20 Mar 2015 01:04:54 -0400 Subject: Fix bug in blk_rq_merge_ok Use the right array index to reference the last element of rq->biotail->bi_io_vec[] Signed-off-by: Wenbo Wang Reviewed-by: Chong Yuan Fixes: 66cb45aa41315 ("block: add support for limiting gaps in SG lists") Cc: stable@kernel.org Signed-off-by: Jens Axboe --- block/blk-merge.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/block/blk-merge.c b/block/blk-merge.c index fc1ff3b1ea1f..fd3fee81c23c 100644 --- a/block/blk-merge.c +++ b/block/blk-merge.c @@ -592,7 +592,7 @@ bool blk_rq_merge_ok(struct request *rq, struct bio *bio) if (q->queue_flags & (1 << QUEUE_FLAG_SG_GAPS)) { struct bio_vec *bprev; - bprev = &rq->biotail->bi_io_vec[bio->bi_vcnt - 1]; + bprev = &rq->biotail->bi_io_vec[rq->biotail->bi_vcnt - 1]; if (bvec_gap_to_prev(bprev, bio->bi_io_vec[0].bv_offset)) return false; } -- cgit v1.2.1 From 8e199dfd82ee097b522b00344af6448715d8ee0c Mon Sep 17 00:00:00 2001 From: Sabrina Dubroca Date: Thu, 19 Mar 2015 11:22:32 +0100 Subject: ipv6: call ipv6_proxy_select_ident instead of ipv6_select_ident in udp6_ufo_fragment Matt Grant reported frequent crashes in ipv6_select_ident when udp6_ufo_fragment is called from openvswitch on a skb that doesn't have a dst_entry set. ipv6_proxy_select_ident generates the frag_id without using the dst associated with the skb. This approach was suggested by Vladislav Yasevich. Fixes: 0508c07f5e0c ("ipv6: Select fragment id during UFO segmentation if not set.") Cc: Vladislav Yasevich Reported-by: Matt Grant Tested-by: Matt Grant Signed-off-by: Sabrina Dubroca Acked-by: Vladislav Yasevich Signed-off-by: David S. Miller --- net/ipv6/udp_offload.c | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) diff --git a/net/ipv6/udp_offload.c b/net/ipv6/udp_offload.c index ab889bb16b3c..be2c0ba82c85 100644 --- a/net/ipv6/udp_offload.c +++ b/net/ipv6/udp_offload.c @@ -112,11 +112,9 @@ static struct sk_buff *udp6_ufo_fragment(struct sk_buff *skb, fptr = (struct frag_hdr *)(skb_network_header(skb) + unfrag_ip6hlen); fptr->nexthdr = nexthdr; fptr->reserved = 0; - if (skb_shinfo(skb)->ip6_frag_id) - fptr->identification = skb_shinfo(skb)->ip6_frag_id; - else - ipv6_select_ident(fptr, - (struct rt6_info *)skb_dst(skb)); + if (!skb_shinfo(skb)->ip6_frag_id) + ipv6_proxy_select_ident(skb); + fptr->identification = skb_shinfo(skb)->ip6_frag_id; /* Fragment the skb. ipv6 header and the remaining fields of the * fragment header are updated in ipv6_gso_segment() -- cgit v1.2.1 From 87f966d97b89774162df04d2106c6350c8fe4cb3 Mon Sep 17 00:00:00 2001 From: Markos Chandras Date: Thu, 19 Mar 2015 10:28:14 +0000 Subject: net: ethernet: pcnet32: Setup the SRAM and NOUFLO on Am79C97{3, 5} On a MIPS Malta board, tons of fifo underflow errors have been observed when using u-boot as bootloader instead of YAMON. The reason for that is that YAMON used to set the pcnet device to SRAM mode but u-boot does not. As a result, the default Tx threshold (64 bytes) is now too small to keep the fifo relatively used and it can result to Tx fifo underflow errors. As a result of which, it's best to setup the SRAM on supported controllers so we can always use the NOUFLO bit. Cc: Cc: Cc: Cc: Don Fry Signed-off-by: Markos Chandras Signed-off-by: David S. Miller --- drivers/net/ethernet/amd/pcnet32.c | 31 +++++++++++++++++++++++++++++-- 1 file changed, 29 insertions(+), 2 deletions(-) diff --git a/drivers/net/ethernet/amd/pcnet32.c b/drivers/net/ethernet/amd/pcnet32.c index 11d6e6561df1..15a8190a6f75 100644 --- a/drivers/net/ethernet/amd/pcnet32.c +++ b/drivers/net/ethernet/amd/pcnet32.c @@ -1543,7 +1543,7 @@ pcnet32_probe1(unsigned long ioaddr, int shared, struct pci_dev *pdev) { struct pcnet32_private *lp; int i, media; - int fdx, mii, fset, dxsuflo; + int fdx, mii, fset, dxsuflo, sram; int chip_version; char *chipname; struct net_device *dev; @@ -1580,7 +1580,7 @@ pcnet32_probe1(unsigned long ioaddr, int shared, struct pci_dev *pdev) } /* initialize variables */ - fdx = mii = fset = dxsuflo = 0; + fdx = mii = fset = dxsuflo = sram = 0; chip_version = (chip_version >> 12) & 0xffff; switch (chip_version) { @@ -1613,6 +1613,7 @@ pcnet32_probe1(unsigned long ioaddr, int shared, struct pci_dev *pdev) chipname = "PCnet/FAST III 79C973"; /* PCI */ fdx = 1; mii = 1; + sram = 1; break; case 0x2626: chipname = "PCnet/Home 79C978"; /* PCI */ @@ -1636,6 +1637,7 @@ pcnet32_probe1(unsigned long ioaddr, int shared, struct pci_dev *pdev) chipname = "PCnet/FAST III 79C975"; /* PCI */ fdx = 1; mii = 1; + sram = 1; break; case 0x2628: chipname = "PCnet/PRO 79C976"; @@ -1664,6 +1666,31 @@ pcnet32_probe1(unsigned long ioaddr, int shared, struct pci_dev *pdev) dxsuflo = 1; } + /* + * The Am79C973/Am79C975 controllers come with 12K of SRAM + * which we can use for the Tx/Rx buffers but most importantly, + * the use of SRAM allow us to use the BCR18:NOUFLO bit to avoid + * Tx fifo underflows. + */ + if (sram) { + /* + * The SRAM is being configured in two steps. First we + * set the SRAM size in the BCR25:SRAM_SIZE bits. According + * to the datasheet, each bit corresponds to a 512-byte + * page so we can have at most 24 pages. The SRAM_SIZE + * holds the value of the upper 8 bits of the 16-bit SRAM size. + * The low 8-bits start at 0x00 and end at 0xff. So the + * address range is from 0x0000 up to 0x17ff. Therefore, + * the SRAM_SIZE is set to 0x17. The next step is to set + * the BCR26:SRAM_BND midway through so the Tx and Rx + * buffers can share the SRAM equally. + */ + a->write_bcr(ioaddr, 25, 0x17); + a->write_bcr(ioaddr, 26, 0xc); + /* And finally enable the NOUFLO bit */ + a->write_bcr(ioaddr, 18, a->read_bcr(ioaddr, 18) | (1 << 11)); + } + dev = alloc_etherdev(sizeof(*lp)); if (!dev) { ret = -ENOMEM; -- cgit v1.2.1 From 73ba57bfae4a1914f6a6dac71e3168dd900e00af Mon Sep 17 00:00:00 2001 From: Steven Barth Date: Thu, 19 Mar 2015 16:16:04 +0100 Subject: ipv6: fix backtracking for throw routes for throw routes to trigger evaluation of other policy rules EAGAIN needs to be propagated up to fib_rules_lookup similar to how its done for IPv4 A simple testcase for verification is: ip -6 rule add lookup 33333 priority 33333 ip -6 route add throw 2001:db8::1 ip -6 route add 2001:db8::1 via fe80::1 dev wlan0 table 33333 ip route get 2001:db8::1 Signed-off-by: Steven Barth Signed-off-by: David S. Miller --- net/ipv6/fib6_rules.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/ipv6/fib6_rules.c b/net/ipv6/fib6_rules.c index b4d5e1d97c1b..27ca79682efb 100644 --- a/net/ipv6/fib6_rules.c +++ b/net/ipv6/fib6_rules.c @@ -104,6 +104,7 @@ static int fib6_rule_action(struct fib_rule *rule, struct flowi *flp, goto again; flp6->saddr = saddr; } + err = rt->dst.error; goto out; } again: -- cgit v1.2.1 From 130c93fd10c4d150e39d8879420c1351aa207fa9 Mon Sep 17 00:00:00 2001 From: Will Deacon Date: Thu, 19 Mar 2015 15:43:00 +0000 Subject: arm64: efi: don't restore TTBR0 if active_mm points at init_mm init_mm isn't a normal mm: it has swapper_pg_dir as its pgd (which contains kernel mappings) and is used as the active_mm for the idle thread. When restoring the pgd after an EFI call, we write current->active_mm into TTBR0. If the current task is actually the idle thread (e.g. when initialising the EFI RTC before entering userspace), then the TLB can erroneously populate itself with junk global entries as a result of speculative table walks. When we do eventually return to userspace, the task can end up hitting these junk mappings leading to lockups, corruption or crashes. This patch fixes the problem in the same way as the CPU suspend code by ensuring that we never switch to the init_mm in efi_set_pgd and instead point TTBR0 at the zero page. A check is also added to cpu_switch_mm to BUG if we get passed swapper_pg_dir. Reviewed-by: Ard Biesheuvel Fixes: f3cdfd239da5 ("arm64/efi: move SetVirtualAddressMap() to UEFI stub") Signed-off-by: Will Deacon Signed-off-by: Catalin Marinas --- arch/arm64/include/asm/proc-fns.h | 6 +++++- arch/arm64/kernel/efi.c | 6 +++++- 2 files changed, 10 insertions(+), 2 deletions(-) diff --git a/arch/arm64/include/asm/proc-fns.h b/arch/arm64/include/asm/proc-fns.h index 9a8fd84f8fb2..941c375616e2 100644 --- a/arch/arm64/include/asm/proc-fns.h +++ b/arch/arm64/include/asm/proc-fns.h @@ -39,7 +39,11 @@ extern u64 cpu_do_resume(phys_addr_t ptr, u64 idmap_ttbr); #include -#define cpu_switch_mm(pgd,mm) cpu_do_switch_mm(virt_to_phys(pgd),mm) +#define cpu_switch_mm(pgd,mm) \ +do { \ + BUG_ON(pgd == swapper_pg_dir); \ + cpu_do_switch_mm(virt_to_phys(pgd),mm); \ +} while (0) #define cpu_get_pgd() \ ({ \ diff --git a/arch/arm64/kernel/efi.c b/arch/arm64/kernel/efi.c index 2b8d70164428..ab21e0d58278 100644 --- a/arch/arm64/kernel/efi.c +++ b/arch/arm64/kernel/efi.c @@ -337,7 +337,11 @@ core_initcall(arm64_dmi_init); static void efi_set_pgd(struct mm_struct *mm) { - cpu_switch_mm(mm->pgd, mm); + if (mm == &init_mm) + cpu_set_reserved_ttbr0(); + else + cpu_switch_mm(mm->pgd, mm); + flush_tlb_all(); if (icache_is_aivivt()) __flush_icache_all(); -- cgit v1.2.1 From d22e1537181188e5dc8cbc51451832625035bdc2 Mon Sep 17 00:00:00 2001 From: Josh Hunt Date: Thu, 19 Mar 2015 19:19:30 -0400 Subject: tcp: fix tcp fin memory accounting tcp_send_fin() does not account for the memory it allocates properly, so sk_forward_alloc can be negative in cases where we've sent a FIN: ss example output (ss -amn | grep -B1 f4294): tcp FIN-WAIT-1 0 1 192.168.0.1:45520 192.0.2.1:8080 skmem:(r0,rb87380,t0,tb87380,f4294966016,w1280,o0,bl0) Acked-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/tcp_output.c | 6 +----- 1 file changed, 1 insertion(+), 5 deletions(-) diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index a2a796c5536b..1db253e36045 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c @@ -2773,15 +2773,11 @@ void tcp_send_fin(struct sock *sk) } else { /* Socket is locked, keep trying until memory is available. */ for (;;) { - skb = alloc_skb_fclone(MAX_TCP_HEADER, - sk->sk_allocation); + skb = sk_stream_alloc_skb(sk, 0, sk->sk_allocation); if (skb) break; yield(); } - - /* Reserve space for headers and prepare control bits. */ - skb_reserve(skb, MAX_TCP_HEADER); /* FIN eats a sequence byte, write_seq advanced by tcp_queue_skb(). */ tcp_init_nondata_skb(skb, tp->write_seq, TCPHDR_ACK | TCPHDR_FIN); -- cgit v1.2.1 From 435452aa88474fae5a31fd14fca88f0802e66f53 Mon Sep 17 00:00:00 2001 From: Vasundhara Volam Date: Fri, 20 Mar 2015 06:28:23 -0400 Subject: be2net: Prevent VFs from enabling VLAN promiscuous mode Currently, a PF does not restrict its VF interface from enabling vlan promiscuous mode. This breaks vlan isolation when a vlan (transparent tagging) is configured on a VF. This patch fixes this problem by disabling the vlan promisc capability for VFs. Reported-by: Yoann Juet Signed-off-by: Vasundhara Volam Signed-off-by: Sathya Perla Signed-off-by: David S. Miller --- drivers/net/ethernet/emulex/benet/be.h | 1 + drivers/net/ethernet/emulex/benet/be_cmds.c | 3 +- drivers/net/ethernet/emulex/benet/be_cmds.h | 2 +- drivers/net/ethernet/emulex/benet/be_main.c | 98 ++++++++++++++++++++++------- 4 files changed, 80 insertions(+), 24 deletions(-) diff --git a/drivers/net/ethernet/emulex/benet/be.h b/drivers/net/ethernet/emulex/benet/be.h index 27de37aa90af..92eb0c82bb04 100644 --- a/drivers/net/ethernet/emulex/benet/be.h +++ b/drivers/net/ethernet/emulex/benet/be.h @@ -354,6 +354,7 @@ struct be_vf_cfg { u16 vlan_tag; u32 tx_rate; u32 plink_tracking; + u32 privileges; }; enum vf_state { diff --git a/drivers/net/ethernet/emulex/benet/be_cmds.c b/drivers/net/ethernet/emulex/benet/be_cmds.c index 36916cfa70f9..3e894f449cc1 100644 --- a/drivers/net/ethernet/emulex/benet/be_cmds.c +++ b/drivers/net/ethernet/emulex/benet/be_cmds.c @@ -1918,7 +1918,7 @@ int be_cmd_modify_eqd(struct be_adapter *adapter, struct be_set_eqd *set_eqd, /* Uses sycnhronous mcc */ int be_cmd_vlan_config(struct be_adapter *adapter, u32 if_id, u16 *vtag_array, - u32 num) + u32 num, u32 domain) { struct be_mcc_wrb *wrb; struct be_cmd_req_vlan_config *req; @@ -1936,6 +1936,7 @@ int be_cmd_vlan_config(struct be_adapter *adapter, u32 if_id, u16 *vtag_array, be_wrb_cmd_hdr_prepare(&req->hdr, CMD_SUBSYSTEM_COMMON, OPCODE_COMMON_NTWK_VLAN_CONFIG, sizeof(*req), wrb, NULL); + req->hdr.domain = domain; req->interface_id = if_id; req->untagged = BE_IF_FLAGS_UNTAGGED & be_if_cap_flags(adapter) ? 1 : 0; diff --git a/drivers/net/ethernet/emulex/benet/be_cmds.h b/drivers/net/ethernet/emulex/benet/be_cmds.h index db761e8e42a3..a7634a3f052a 100644 --- a/drivers/net/ethernet/emulex/benet/be_cmds.h +++ b/drivers/net/ethernet/emulex/benet/be_cmds.h @@ -2256,7 +2256,7 @@ int lancer_cmd_get_pport_stats(struct be_adapter *adapter, int be_cmd_get_fw_ver(struct be_adapter *adapter); int be_cmd_modify_eqd(struct be_adapter *adapter, struct be_set_eqd *, int num); int be_cmd_vlan_config(struct be_adapter *adapter, u32 if_id, u16 *vtag_array, - u32 num); + u32 num, u32 domain); int be_cmd_rx_filter(struct be_adapter *adapter, u32 flags, u32 status); int be_cmd_set_flow_control(struct be_adapter *adapter, u32 tx_fc, u32 rx_fc); int be_cmd_get_flow_control(struct be_adapter *adapter, u32 *tx_fc, u32 *rx_fc); diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c index 0a816859aca5..eb2bed92bf60 100644 --- a/drivers/net/ethernet/emulex/benet/be_main.c +++ b/drivers/net/ethernet/emulex/benet/be_main.c @@ -1171,7 +1171,7 @@ static int be_vid_config(struct be_adapter *adapter) for_each_set_bit(i, adapter->vids, VLAN_N_VID) vids[num++] = cpu_to_le16(i); - status = be_cmd_vlan_config(adapter, adapter->if_handle, vids, num); + status = be_cmd_vlan_config(adapter, adapter->if_handle, vids, num, 0); if (status) { dev_err(dev, "Setting HW VLAN filtering failed\n"); /* Set to VLAN promisc mode as setting VLAN filter failed */ @@ -1380,11 +1380,67 @@ static int be_get_vf_config(struct net_device *netdev, int vf, return 0; } +static int be_set_vf_tvt(struct be_adapter *adapter, int vf, u16 vlan) +{ + struct be_vf_cfg *vf_cfg = &adapter->vf_cfg[vf]; + u16 vids[BE_NUM_VLANS_SUPPORTED]; + int vf_if_id = vf_cfg->if_handle; + int status; + + /* Enable Transparent VLAN Tagging */ + status = be_cmd_set_hsw_config(adapter, vlan, vf + 1, vf_if_id, 0); + if (status) + return status; + + /* Clear pre-programmed VLAN filters on VF if any, if TVT is enabled */ + vids[0] = 0; + status = be_cmd_vlan_config(adapter, vf_if_id, vids, 1, vf + 1); + if (!status) + dev_info(&adapter->pdev->dev, + "Cleared guest VLANs on VF%d", vf); + + /* After TVT is enabled, disallow VFs to program VLAN filters */ + if (vf_cfg->privileges & BE_PRIV_FILTMGMT) { + status = be_cmd_set_fn_privileges(adapter, vf_cfg->privileges & + ~BE_PRIV_FILTMGMT, vf + 1); + if (!status) + vf_cfg->privileges &= ~BE_PRIV_FILTMGMT; + } + return 0; +} + +static int be_clear_vf_tvt(struct be_adapter *adapter, int vf) +{ + struct be_vf_cfg *vf_cfg = &adapter->vf_cfg[vf]; + struct device *dev = &adapter->pdev->dev; + int status; + + /* Reset Transparent VLAN Tagging. */ + status = be_cmd_set_hsw_config(adapter, BE_RESET_VLAN_TAG_ID, vf + 1, + vf_cfg->if_handle, 0); + if (status) + return status; + + /* Allow VFs to program VLAN filtering */ + if (!(vf_cfg->privileges & BE_PRIV_FILTMGMT)) { + status = be_cmd_set_fn_privileges(adapter, vf_cfg->privileges | + BE_PRIV_FILTMGMT, vf + 1); + if (!status) { + vf_cfg->privileges |= BE_PRIV_FILTMGMT; + dev_info(dev, "VF%d: FILTMGMT priv enabled", vf); + } + } + + dev_info(dev, + "Disable/re-enable i/f in VM to clear Transparent VLAN tag"); + return 0; +} + static int be_set_vf_vlan(struct net_device *netdev, int vf, u16 vlan, u8 qos) { struct be_adapter *adapter = netdev_priv(netdev); struct be_vf_cfg *vf_cfg = &adapter->vf_cfg[vf]; - int status = 0; + int status; if (!sriov_enabled(adapter)) return -EPERM; @@ -1394,24 +1450,19 @@ static int be_set_vf_vlan(struct net_device *netdev, int vf, u16 vlan, u8 qos) if (vlan || qos) { vlan |= qos << VLAN_PRIO_SHIFT; - if (vf_cfg->vlan_tag != vlan) - status = be_cmd_set_hsw_config(adapter, vlan, vf + 1, - vf_cfg->if_handle, 0); + status = be_set_vf_tvt(adapter, vf, vlan); } else { - /* Reset Transparent Vlan Tagging. */ - status = be_cmd_set_hsw_config(adapter, BE_RESET_VLAN_TAG_ID, - vf + 1, vf_cfg->if_handle, 0); + status = be_clear_vf_tvt(adapter, vf); } if (status) { dev_err(&adapter->pdev->dev, - "VLAN %d config on VF %d failed : %#x\n", vlan, - vf, status); + "VLAN %d config on VF %d failed : %#x\n", vlan, vf, + status); return be_cmd_status(status); } vf_cfg->vlan_tag = vlan; - return 0; } @@ -3339,7 +3390,6 @@ static int be_if_create(struct be_adapter *adapter, u32 *if_handle, u32 cap_flags, u32 vf) { u32 en_flags; - int status; en_flags = BE_IF_FLAGS_UNTAGGED | BE_IF_FLAGS_BROADCAST | BE_IF_FLAGS_MULTICAST | BE_IF_FLAGS_PASS_L3L4_ERRORS | @@ -3347,10 +3397,7 @@ static int be_if_create(struct be_adapter *adapter, u32 *if_handle, en_flags &= cap_flags; - status = be_cmd_if_create(adapter, cap_flags, en_flags, - if_handle, vf); - - return status; + return be_cmd_if_create(adapter, cap_flags, en_flags, if_handle, vf); } static int be_vfs_if_create(struct be_adapter *adapter) @@ -3368,8 +3415,13 @@ static int be_vfs_if_create(struct be_adapter *adapter) if (!BE3_chip(adapter)) { status = be_cmd_get_profile_config(adapter, &res, vf + 1); - if (!status) + if (!status) { cap_flags = res.if_cap_flags; + /* Prevent VFs from enabling VLAN promiscuous + * mode + */ + cap_flags &= ~BE_IF_FLAGS_VLAN_PROMISCUOUS; + } } status = be_if_create(adapter, &vf_cfg->if_handle, @@ -3403,7 +3455,6 @@ static int be_vf_setup(struct be_adapter *adapter) struct device *dev = &adapter->pdev->dev; struct be_vf_cfg *vf_cfg; int status, old_vfs, vf; - u32 privileges; old_vfs = pci_num_vf(adapter->pdev); @@ -3433,15 +3484,18 @@ static int be_vf_setup(struct be_adapter *adapter) for_all_vfs(adapter, vf_cfg, vf) { /* Allow VFs to programs MAC/VLAN filters */ - status = be_cmd_get_fn_privileges(adapter, &privileges, vf + 1); - if (!status && !(privileges & BE_PRIV_FILTMGMT)) { + status = be_cmd_get_fn_privileges(adapter, &vf_cfg->privileges, + vf + 1); + if (!status && !(vf_cfg->privileges & BE_PRIV_FILTMGMT)) { status = be_cmd_set_fn_privileges(adapter, - privileges | + vf_cfg->privileges | BE_PRIV_FILTMGMT, vf + 1); - if (!status) + if (!status) { + vf_cfg->privileges |= BE_PRIV_FILTMGMT; dev_info(dev, "VF%d has FILTMGMT privilege\n", vf); + } } /* Allow full available bandwidth */ -- cgit v1.2.1 From c8ba4ad0b59c511578f8f706aae711f01c990794 Mon Sep 17 00:00:00 2001 From: Suresh Reddy Date: Fri, 20 Mar 2015 06:28:24 -0400 Subject: be2net: restrict MODIFY_EQ_DELAY cmd to a max of 8 EQs Issuing this cmd for more than 8 EQs does not have the intended effect even on BEx and Skyhawk-R. This patch fixes this by issuing this cmd for upto 8 EQs at a time. Signed-off-by: Suresh Reddy Signed-off-by: Sathya Perla Signed-off-by: David S. Miller --- drivers/net/ethernet/emulex/benet/be_cmds.c | 14 +++++--------- 1 file changed, 5 insertions(+), 9 deletions(-) diff --git a/drivers/net/ethernet/emulex/benet/be_cmds.c b/drivers/net/ethernet/emulex/benet/be_cmds.c index 3e894f449cc1..7f05f309e935 100644 --- a/drivers/net/ethernet/emulex/benet/be_cmds.c +++ b/drivers/net/ethernet/emulex/benet/be_cmds.c @@ -1902,15 +1902,11 @@ int be_cmd_modify_eqd(struct be_adapter *adapter, struct be_set_eqd *set_eqd, { int num_eqs, i = 0; - if (lancer_chip(adapter) && num > 8) { - while (num) { - num_eqs = min(num, 8); - __be_cmd_modify_eqd(adapter, &set_eqd[i], num_eqs); - i += num_eqs; - num -= num_eqs; - } - } else { - __be_cmd_modify_eqd(adapter, set_eqd, num); + while (num) { + num_eqs = min(num, 8); + __be_cmd_modify_eqd(adapter, &set_eqd[i], num_eqs); + i += num_eqs; + num -= num_eqs; } return 0; -- cgit v1.2.1 From 25848c9015964d9d97dffd48bbb88b75a25d0a1b Mon Sep 17 00:00:00 2001 From: Suresh Reddy Date: Fri, 20 Mar 2015 06:28:25 -0400 Subject: be2net: use PCI MMIO read instead of config read for errors When an EEH error occurs, the device/slot is disconnected. This condition is more reliably detected (i.e., returns all ones) with an MMIO read rather than a config read -- especially on power platforms. Hence, this patch fixes EEH error detection by replacing config reads with MMIO reads for reading the error registers. The error registers in Skyhawk-R/BE2/BE3 are accessible both via the config space and the PCICFG (BAR0) memory space. Reported-by: Gavin Shan Signed-off-by: Suresh Reddy Signed-off-by: Sathya Perla Signed-off-by: David S. Miller --- drivers/net/ethernet/emulex/benet/be.h | 1 + drivers/net/ethernet/emulex/benet/be_main.c | 33 +++++++++++++++++++---------- 2 files changed, 23 insertions(+), 11 deletions(-) diff --git a/drivers/net/ethernet/emulex/benet/be.h b/drivers/net/ethernet/emulex/benet/be.h index 92eb0c82bb04..27b9fe99a9bd 100644 --- a/drivers/net/ethernet/emulex/benet/be.h +++ b/drivers/net/ethernet/emulex/benet/be.h @@ -424,6 +424,7 @@ struct be_adapter { u8 __iomem *csr; /* CSR BAR used only for BE2/3 */ u8 __iomem *db; /* Door Bell */ + u8 __iomem *pcicfg; /* On SH,BEx only. Shadow of PCI config space */ struct mutex mbox_lock; /* For serializing mbox cmds to BE card */ struct be_dma_mem mbox_mem; diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c index eb2bed92bf60..e6b790f0d9dc 100644 --- a/drivers/net/ethernet/emulex/benet/be_main.c +++ b/drivers/net/ethernet/emulex/benet/be_main.c @@ -2823,14 +2823,12 @@ void be_detect_error(struct be_adapter *adapter) } } } else { - pci_read_config_dword(adapter->pdev, - PCICFG_UE_STATUS_LOW, &ue_lo); - pci_read_config_dword(adapter->pdev, - PCICFG_UE_STATUS_HIGH, &ue_hi); - pci_read_config_dword(adapter->pdev, - PCICFG_UE_STATUS_LOW_MASK, &ue_lo_mask); - pci_read_config_dword(adapter->pdev, - PCICFG_UE_STATUS_HI_MASK, &ue_hi_mask); + ue_lo = ioread32(adapter->pcicfg + PCICFG_UE_STATUS_LOW); + ue_hi = ioread32(adapter->pcicfg + PCICFG_UE_STATUS_HIGH); + ue_lo_mask = ioread32(adapter->pcicfg + + PCICFG_UE_STATUS_LOW_MASK); + ue_hi_mask = ioread32(adapter->pcicfg + + PCICFG_UE_STATUS_HI_MASK); ue_lo = (ue_lo & ~ue_lo_mask); ue_hi = (ue_hi & ~ue_hi_mask); @@ -4874,24 +4872,37 @@ static int be_roce_map_pci_bars(struct be_adapter *adapter) static int be_map_pci_bars(struct be_adapter *adapter) { + struct pci_dev *pdev = adapter->pdev; u8 __iomem *addr; if (BEx_chip(adapter) && be_physfn(adapter)) { - adapter->csr = pci_iomap(adapter->pdev, 2, 0); + adapter->csr = pci_iomap(pdev, 2, 0); if (!adapter->csr) return -ENOMEM; } - addr = pci_iomap(adapter->pdev, db_bar(adapter), 0); + addr = pci_iomap(pdev, db_bar(adapter), 0); if (!addr) goto pci_map_err; adapter->db = addr; + if (skyhawk_chip(adapter) || BEx_chip(adapter)) { + if (be_physfn(adapter)) { + /* PCICFG is the 2nd BAR in BE2 */ + addr = pci_iomap(pdev, BE2_chip(adapter) ? 1 : 0, 0); + if (!addr) + goto pci_map_err; + adapter->pcicfg = addr; + } else { + adapter->pcicfg = adapter->db + SRIOV_VF_PCICFG_OFFSET; + } + } + be_roce_map_pci_bars(adapter); return 0; pci_map_err: - dev_err(&adapter->pdev->dev, "Error in mapping PCI BARs\n"); + dev_err(&pdev->dev, "Error in mapping PCI BARs\n"); be_unmap_pci_bars(adapter); return -ENOMEM; } -- cgit v1.2.1 From 7132813c384515c9dede1ae20e56f3895feb7f1e Mon Sep 17 00:00:00 2001 From: "Suzuki K. Poulose" Date: Thu, 19 Mar 2015 18:17:09 +0000 Subject: arm64: Honor __GFP_ZERO in dma allocations Current implementation doesn't zero out the pages allocated. Honor the __GFP_ZERO flag and zero out if set. Cc: # v3.14+ Acked-by: Will Deacon Signed-off-by: Suzuki K. Poulose Signed-off-by: Catalin Marinas --- arch/arm64/mm/dma-mapping.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) diff --git a/arch/arm64/mm/dma-mapping.c b/arch/arm64/mm/dma-mapping.c index 58e0c2bdde04..ef7d112f5ce0 100644 --- a/arch/arm64/mm/dma-mapping.c +++ b/arch/arm64/mm/dma-mapping.c @@ -51,7 +51,7 @@ static int __init early_coherent_pool(char *p) } early_param("coherent_pool", early_coherent_pool); -static void *__alloc_from_pool(size_t size, struct page **ret_page) +static void *__alloc_from_pool(size_t size, struct page **ret_page, gfp_t flags) { unsigned long val; void *ptr = NULL; @@ -67,6 +67,8 @@ static void *__alloc_from_pool(size_t size, struct page **ret_page) *ret_page = phys_to_page(phys); ptr = (void *)val; + if (flags & __GFP_ZERO) + memset(ptr, 0, size); } return ptr; @@ -101,6 +103,7 @@ static void *__dma_alloc_coherent(struct device *dev, size_t size, flags |= GFP_DMA; if (IS_ENABLED(CONFIG_DMA_CMA) && (flags & __GFP_WAIT)) { struct page *page; + void *addr; size = PAGE_ALIGN(size); page = dma_alloc_from_contiguous(dev, size >> PAGE_SHIFT, @@ -109,7 +112,10 @@ static void *__dma_alloc_coherent(struct device *dev, size_t size, return NULL; *dma_handle = phys_to_dma(dev, page_to_phys(page)); - return page_address(page); + addr = page_address(page); + if (flags & __GFP_ZERO) + memset(addr, 0, size); + return addr; } else { return swiotlb_alloc_coherent(dev, size, dma_handle, flags); } @@ -146,7 +152,7 @@ static void *__dma_alloc(struct device *dev, size_t size, if (!coherent && !(flags & __GFP_WAIT)) { struct page *page = NULL; - void *addr = __alloc_from_pool(size, &page); + void *addr = __alloc_from_pool(size, &page, flags); if (addr) *dma_handle = phys_to_dma(dev, page_to_phys(page)); -- cgit v1.2.1 From 91edd096e224941131f896b86838b1e59553696a Mon Sep 17 00:00:00 2001 From: Catalin Marinas Date: Fri, 20 Mar 2015 16:48:13 +0000 Subject: net: compat: Update get_compat_msghdr() to match copy_msghdr_from_user() behaviour Commit db31c55a6fb2 (net: clamp ->msg_namelen instead of returning an error) introduced the clamping of msg_namelen when the unsigned value was larger than sizeof(struct sockaddr_storage). This caused a msg_namelen of -1 to be valid. The native code was subsequently fixed by commit dbb490b96584 (net: socket: error on a negative msg_namelen). In addition, the native code sets msg_namelen to 0 when msg_name is NULL. This was done in commit (6a2a2b3ae075 net:socket: set msg_namelen to 0 if msg_name is passed as NULL in msghdr struct from userland) and subsequently updated by 08adb7dabd48 (fold verify_iovec() into copy_msghdr_from_user()). This patch brings the get_compat_msghdr() in line with copy_msghdr_from_user(). Fixes: db31c55a6fb2 (net: clamp ->msg_namelen instead of returning an error) Cc: David S. Miller Cc: Dan Carpenter Signed-off-by: Catalin Marinas Signed-off-by: David S. Miller --- net/compat.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/net/compat.c b/net/compat.c index 94d3d5e97883..f7bd286a8280 100644 --- a/net/compat.c +++ b/net/compat.c @@ -49,6 +49,13 @@ ssize_t get_compat_msghdr(struct msghdr *kmsg, __get_user(kmsg->msg_controllen, &umsg->msg_controllen) || __get_user(kmsg->msg_flags, &umsg->msg_flags)) return -EFAULT; + + if (!uaddr) + kmsg->msg_namelen = 0; + + if (kmsg->msg_namelen < 0) + return -EINVAL; + if (kmsg->msg_namelen > sizeof(struct sockaddr_storage)) kmsg->msg_namelen = sizeof(struct sockaddr_storage); kmsg->msg_control = compat_ptr(tmp3); -- cgit v1.2.1 From 4de930efc23b92ddf88ce91c405ee645fe6e27ea Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 20 Mar 2015 17:41:43 +0000 Subject: net: validate the range we feed to iov_iter_init() in sys_sendto/sys_recvfrom Cc: stable@vger.kernel.org # v3.19 Signed-off-by: Al Viro Signed-off-by: David S. Miller --- net/socket.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/net/socket.c b/net/socket.c index bbedbfcb42c2..245330ca0015 100644 --- a/net/socket.c +++ b/net/socket.c @@ -1702,6 +1702,8 @@ SYSCALL_DEFINE6(sendto, int, fd, void __user *, buff, size_t, len, if (len > INT_MAX) len = INT_MAX; + if (unlikely(!access_ok(VERIFY_READ, buff, len))) + return -EFAULT; sock = sockfd_lookup_light(fd, &err, &fput_needed); if (!sock) goto out; @@ -1760,6 +1762,8 @@ SYSCALL_DEFINE6(recvfrom, int, fd, void __user *, ubuf, size_t, size, if (size > INT_MAX) size = INT_MAX; + if (unlikely(!access_ok(VERIFY_WRITE, ubuf, size))) + return -EFAULT; sock = sockfd_lookup_light(fd, &err, &fput_needed); if (!sock) goto out; -- cgit v1.2.1 From 0c35bd4723e4a39ba2da4c13a22cb97986ee10c8 Mon Sep 17 00:00:00 2001 From: NeilBrown Date: Fri, 13 Mar 2015 11:51:18 +1100 Subject: md: fix problems with freeing private data after ->run failure. If ->run() fails, it can either free the data structures it allocated, or leave that task to ->free() which will be called on failures. However: md.c calls ->free() even if ->private_data is NULL, which causes problems in some personalities. raid0.c frees the data, but doesn't clear ->private_data, which will become a problem when we fix md.c So better fix both these issues at once. Reported-by: Richard W.M. Jones Fixes: 5aa61f427e4979be733e4847b9199ff9cc48a47e URL: https://bugzilla.kernel.org/show_bug.cgi?id=94381 Signed-off-by: NeilBrown --- drivers/md/md.c | 3 ++- drivers/md/raid0.c | 2 -- 2 files changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/md/md.c b/drivers/md/md.c index cadf9cc02b25..717daad71fb1 100644 --- a/drivers/md/md.c +++ b/drivers/md/md.c @@ -5080,7 +5080,8 @@ int md_run(struct mddev *mddev) } if (err) { mddev_detach(mddev); - pers->free(mddev, mddev->private); + if (mddev->private) + pers->free(mddev, mddev->private); module_put(pers->owner); bitmap_destroy(mddev); return err; diff --git a/drivers/md/raid0.c b/drivers/md/raid0.c index a13f738a7b39..3ed9f42ddca6 100644 --- a/drivers/md/raid0.c +++ b/drivers/md/raid0.c @@ -467,8 +467,6 @@ static int raid0_run(struct mddev *mddev) dump_zones(mddev); ret = md_integrity_register(mddev); - if (ret) - raid0_free(mddev, conf); return ret; } -- cgit v1.2.1 From f40bff4239d45ac061044a8a79cf6868c62df345 Mon Sep 17 00:00:00 2001 From: Ondrej Zary Date: Sat, 21 Mar 2015 11:29:37 +0100 Subject: cx82310_eth: wait for firmware to become ready When the device is powered up, some (older) firmware versions fail to work properly if we send commands before the boot is complete (everything is OK when the device is hot-plugged). The firmware indicates its ready status by putting the link up. Newer firmwares delay the first command so they don't suffer from this problem. They also report the link being always up. Wait for firmware to become ready (link up) before sending any commands and/or data. This also allows lowering CMD_TIMEOUT value to a reasonable time. Tested with 4.1.0.9 (old) and 4.1.0.30 (new) firmware versions. Signed-off-by: Ondrej Zary Signed-off-by: David S. Miller --- drivers/net/usb/cx82310_eth.c | 30 ++++++++++++++++++++++++------ 1 file changed, 24 insertions(+), 6 deletions(-) diff --git a/drivers/net/usb/cx82310_eth.c b/drivers/net/usb/cx82310_eth.c index fe48f4c51373..1762ad3910b2 100644 --- a/drivers/net/usb/cx82310_eth.c +++ b/drivers/net/usb/cx82310_eth.c @@ -46,8 +46,7 @@ enum cx82310_status { }; #define CMD_PACKET_SIZE 64 -/* first command after power on can take around 8 seconds */ -#define CMD_TIMEOUT 15000 +#define CMD_TIMEOUT 100 #define CMD_REPLY_RETRY 5 #define CX82310_MTU 1514 @@ -78,8 +77,9 @@ static int cx82310_cmd(struct usbnet *dev, enum cx82310_cmd cmd, bool reply, ret = usb_bulk_msg(udev, usb_sndbulkpipe(udev, CMD_EP), buf, CMD_PACKET_SIZE, &actual_len, CMD_TIMEOUT); if (ret < 0) { - dev_err(&dev->udev->dev, "send command %#x: error %d\n", - cmd, ret); + if (cmd != CMD_GET_LINK_STATUS) + dev_err(&dev->udev->dev, "send command %#x: error %d\n", + cmd, ret); goto end; } @@ -90,8 +90,10 @@ static int cx82310_cmd(struct usbnet *dev, enum cx82310_cmd cmd, bool reply, buf, CMD_PACKET_SIZE, &actual_len, CMD_TIMEOUT); if (ret < 0) { - dev_err(&dev->udev->dev, - "reply receive error %d\n", ret); + if (cmd != CMD_GET_LINK_STATUS) + dev_err(&dev->udev->dev, + "reply receive error %d\n", + ret); goto end; } if (actual_len > 0) @@ -134,6 +136,8 @@ static int cx82310_bind(struct usbnet *dev, struct usb_interface *intf) int ret; char buf[15]; struct usb_device *udev = dev->udev; + u8 link[3]; + int timeout = 50; /* avoid ADSL modems - continue only if iProduct is "USB NET CARD" */ if (usb_string(udev, udev->descriptor.iProduct, buf, sizeof(buf)) > 0 @@ -160,6 +164,20 @@ static int cx82310_bind(struct usbnet *dev, struct usb_interface *intf) if (!dev->partial_data) return -ENOMEM; + /* wait for firmware to become ready (indicated by the link being up) */ + while (--timeout) { + ret = cx82310_cmd(dev, CMD_GET_LINK_STATUS, true, NULL, 0, + link, sizeof(link)); + /* the command can time out during boot - it's not an error */ + if (!ret && link[0] == 1 && link[2] == 1) + break; + msleep(500); + }; + if (!timeout) { + dev_err(&udev->dev, "firmware not ready in time\n"); + return -ETIMEDOUT; + } + /* enable ethernet mode (?) */ ret = cx82310_cmd(dev, CMD_ETHERNET_MODE, true, "\x01", 1, NULL, 0); if (ret) { -- cgit v1.2.1 From 749177ccc74f9c6d0f51bd78a15c652a2134aa11 Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Sat, 21 Mar 2015 19:25:05 +0100 Subject: netfilter: nft_compat: set IP6T_F_PROTO flag if protocol is set ip6tables extensions check for this flag to restrict match/target to a given protocol. Without this flag set, SYNPROXY6 returns an error. Signed-off-by: Pablo Neira Ayuso Acked-by: Patrick McHardy --- net/netfilter/nft_compat.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/net/netfilter/nft_compat.c b/net/netfilter/nft_compat.c index 213584cf04b3..65f3e2b6be44 100644 --- a/net/netfilter/nft_compat.c +++ b/net/netfilter/nft_compat.c @@ -133,6 +133,9 @@ nft_target_set_tgchk_param(struct xt_tgchk_param *par, entry->e4.ip.invflags = inv ? IPT_INV_PROTO : 0; break; case AF_INET6: + if (proto) + entry->e6.ipv6.flags |= IP6T_F_PROTO; + entry->e6.ipv6.proto = proto; entry->e6.ipv6.invflags = inv ? IP6T_INV_PROTO : 0; break; @@ -344,6 +347,9 @@ nft_match_set_mtchk_param(struct xt_mtchk_param *par, const struct nft_ctx *ctx, entry->e4.ip.invflags = inv ? IPT_INV_PROTO : 0; break; case AF_INET6: + if (proto) + entry->e6.ipv6.flags |= IP6T_F_PROTO; + entry->e6.ipv6.proto = proto; entry->e6.ipv6.invflags = inv ? IP6T_INV_PROTO : 0; break; -- cgit v1.2.1 From bc465aa9d045feb0e13b4a8f32cc33c1943f62d6 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sun, 22 Mar 2015 16:50:21 -0700 Subject: Linux 4.0-rc5 --- Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Makefile b/Makefile index e734965b1604..14c722f96877 100644 --- a/Makefile +++ b/Makefile @@ -1,7 +1,7 @@ VERSION = 4 PATCHLEVEL = 0 SUBLEVEL = 0 -EXTRAVERSION = -rc4 +EXTRAVERSION = -rc5 NAME = Hurr durr I'ma sheep # *DOCUMENTATION* -- cgit v1.2.1 From 44d5f6f5901e996744858c175baee320ccf1eda3 Mon Sep 17 00:00:00 2001 From: Mahesh Salgaonkar Date: Tue, 17 Mar 2015 16:14:41 +0530 Subject: powerpc/book3s: Fix the MCE code to use CONFIG_KVM_BOOK3S_64_HANDLER commit id 2ba9f0d has changed CONFIG_KVM_BOOK3S_64_HV to tristate to allow HV/PR bits to be built as modules. But the MCE code still depends on CONFIG_KVM_BOOK3S_64_HV which is wrong. When user selects CONFIG_KVM_BOOK3S_64_HV=m to build HV/PR bits as a separate module the relevant MCE code gets excluded. This patch fixes the MCE code to use CONFIG_KVM_BOOK3S_64_HANDLER. This makes sure that the relevant MCE code is included when HV/PR bits are built as a separate modules. Fixes: 2ba9f0d88750 ("kvm: powerpc: book3s: Support building HV and PR KVM as module") Cc: stable@vger.kernel.org # v3.14+ Signed-off-by: Mahesh Salgaonkar Acked-by: Paul Mackerras Signed-off-by: Michael Ellerman --- arch/powerpc/kernel/exceptions-64s.S | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/powerpc/kernel/exceptions-64s.S b/arch/powerpc/kernel/exceptions-64s.S index c2df8150bd7a..9519e6bdc6d7 100644 --- a/arch/powerpc/kernel/exceptions-64s.S +++ b/arch/powerpc/kernel/exceptions-64s.S @@ -1408,7 +1408,7 @@ machine_check_handle_early: bne 9f /* continue in V mode if we are. */ 5: -#ifdef CONFIG_KVM_BOOK3S_64_HV +#ifdef CONFIG_KVM_BOOK3S_64_HANDLER /* * We are coming from kernel context. Check if we are coming from * guest. if yes, then we can continue. We will fall through -- cgit v1.2.1 From 9848de082ff400f8b88e6d550bfce551838af69c Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Ville=20Syrj=C3=A4l=C3=A4?= Date: Fri, 20 Mar 2015 21:28:08 +0200 Subject: drm/i915: Use usleep_range() in wait_for() MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit msleep() can sleep for way too long, so switch wait_for() to use usleep_range() instead. Following a totally unscientific method I just picked the range as W-2W. This cuts the i915 init time on my BSW to almost half: - initcall i915_init+0x0/0xa8 [i915] returned 0 after 419977 usecs + initcall i915_init+0x0/0xa8 [i915] returned 0 after 238419 usecs Note that I didn't perform any other benchmarks on this so far. Cc: Jesse Barnes Signed-off-by: Ville Syrjälä Reviewed-by: Jesse Barnes Reviewed-by: Chris Wilson Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_drv.h | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index 6f20f3a01398..d2a4de0e4f4a 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -56,8 +56,8 @@ ret__ = -ETIMEDOUT; \ break; \ } \ - if (W && drm_can_sleep()) { \ - msleep(W); \ + if ((W) && drm_can_sleep()) { \ + usleep_range((W)*1000, (W)*2000); \ } else { \ cpu_relax(); \ } \ -- cgit v1.2.1 From d525211f9d1be8b523ec7633f080f2116f5ea536 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 19 Feb 2015 18:03:11 +0100 Subject: perf: Fix irq_work 'tail' recursion Vince reported a watchdog lockup like: [] perf_tp_event+0xc4/0x210 [] perf_trace_lock+0x12a/0x160 [] lock_release+0x130/0x260 [] _raw_spin_unlock_irqrestore+0x24/0x40 [] do_send_sig_info+0x5d/0x80 [] send_sigio_to_task+0x12f/0x1a0 [] send_sigio+0xae/0x100 [] kill_fasync+0x97/0xf0 [] perf_event_wakeup+0xd4/0xf0 [] perf_pending_event+0x33/0x60 [] irq_work_run_list+0x4c/0x80 [] irq_work_run+0x18/0x40 [] smp_trace_irq_work_interrupt+0x3f/0xc0 [] trace_irq_work_interrupt+0x6d/0x80 Which is caused by an irq_work generating new irq_work and therefore not allowing forward progress. This happens because processing the perf irq_work triggers another perf event (tracepoint stuff) which in turn generates an irq_work ad infinitum. Avoid this by raising the recursion counter in the irq_work -- which effectively disables all software events (including tracepoints) from actually triggering again. Reported-by: Vince Weaver Tested-by: Vince Weaver Signed-off-by: Peter Zijlstra (Intel) Cc: Arnaldo Carvalho de Melo Cc: Jiri Olsa Cc: Paul Mackerras Cc: Steven Rostedt Cc: Link: http://lkml.kernel.org/r/20150219170311.GH21418@twins.programming.kicks-ass.net Signed-off-by: Ingo Molnar --- kernel/events/core.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/kernel/events/core.c b/kernel/events/core.c index 453ef61311d4..2fabc0627165 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -4574,6 +4574,13 @@ static void perf_pending_event(struct irq_work *entry) { struct perf_event *event = container_of(entry, struct perf_event, pending); + int rctx; + + rctx = perf_swevent_get_recursion_context(); + /* + * If we 'fail' here, that's OK, it means recursion is already disabled + * and we won't recurse 'further'. + */ if (event->pending_disable) { event->pending_disable = 0; @@ -4584,6 +4591,9 @@ static void perf_pending_event(struct irq_work *entry) event->pending_wakeup = 0; perf_event_wakeup(event); } + + if (rctx >= 0) + perf_swevent_put_recursion_context(rctx); } /* -- cgit v1.2.1 From 746db9443ea57fd9c059f62c4bfbf41cf224fe13 Mon Sep 17 00:00:00 2001 From: Brian Silverman Date: Wed, 18 Feb 2015 16:23:56 -0800 Subject: sched: Fix RLIMIT_RTTIME when PI-boosting to RT When non-realtime tasks get priority-inheritance boosted to a realtime scheduling class, RLIMIT_RTTIME starts to apply to them. However, the counter used for checking this (the same one used for SCHED_RR timeslices) was not getting reset. This meant that tasks running with a non-realtime scheduling class which are repeatedly boosted to a realtime one, but never block while they are running realtime, eventually hit the timeout without ever running for a time over the limit. This patch resets the realtime timeslice counter when un-PI-boosting from an RT to a non-RT scheduling class. I have some test code with two threads and a shared PTHREAD_PRIO_INHERIT mutex which induces priority boosting and spins while boosted that gets killed by a SIGXCPU on non-fixed kernels but doesn't with this patch applied. It happens much faster with a CONFIG_PREEMPT_RT kernel, and does happen eventually with PREEMPT_VOLUNTARY kernels. Signed-off-by: Brian Silverman Signed-off-by: Peter Zijlstra (Intel) Cc: austin@peloton-tech.com Cc: Link: http://lkml.kernel.org/r/1424305436-6716-1-git-send-email-brian@peloton-tech.com Signed-off-by: Ingo Molnar --- kernel/sched/core.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/kernel/sched/core.c b/kernel/sched/core.c index f0f831e8a345..62671f53202a 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -3034,6 +3034,8 @@ void rt_mutex_setprio(struct task_struct *p, int prio) } else { if (dl_prio(oldprio)) p->dl.dl_boosted = 0; + if (rt_prio(oldprio)) + p->rt.timeout = 0; p->sched_class = &fair_sched_class; } -- cgit v1.2.1 From 35a9393c95b31870a74f51a3e7455f33f5657b6f Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 26 Feb 2015 16:23:11 +0100 Subject: lockdep: Fix the module unload key range freeing logic Module unload calls lockdep_free_key_range(), which removes entries from the data structures. Most of the lockdep code OTOH assumes the data structures are append only; in specific see the comments in add_lock_to_list() and look_up_lock_class(). Clearly this has only worked by accident; make it work proper. The actual scenario to make it go boom would involve the memory freed by the module unlock being re-allocated and re-used for a lock inside of a rcu-sched grace period. This is a very unlikely scenario, still better plug the hole. Use RCU list iteration in all places and ammend the comments. Change lockdep_free_key_range() to issue a sync_sched() between removal from the lists and returning -- which results in the memory being freed. Further ensure the callers are placed correctly and comment the requirements. Signed-off-by: Peter Zijlstra (Intel) Cc: Andrew Morton Cc: Andrey Tsyvarev Cc: Linus Torvalds Cc: Paul E. McKenney Cc: Peter Zijlstra Cc: Rusty Russell Cc: Thomas Gleixner Signed-off-by: Ingo Molnar --- kernel/locking/lockdep.c | 81 ++++++++++++++++++++++++++++++++---------------- kernel/module.c | 8 ++--- 2 files changed, 59 insertions(+), 30 deletions(-) diff --git a/kernel/locking/lockdep.c b/kernel/locking/lockdep.c index 88d0d4420ad2..ba77ab5f64dd 100644 --- a/kernel/locking/lockdep.c +++ b/kernel/locking/lockdep.c @@ -633,7 +633,7 @@ static int count_matching_names(struct lock_class *new_class) if (!new_class->name) return 0; - list_for_each_entry(class, &all_lock_classes, lock_entry) { + list_for_each_entry_rcu(class, &all_lock_classes, lock_entry) { if (new_class->key - new_class->subclass == class->key) return class->name_version; if (class->name && !strcmp(class->name, new_class->name)) @@ -700,10 +700,12 @@ look_up_lock_class(struct lockdep_map *lock, unsigned int subclass) hash_head = classhashentry(key); /* - * We can walk the hash lockfree, because the hash only - * grows, and we are careful when adding entries to the end: + * We do an RCU walk of the hash, see lockdep_free_key_range(). */ - list_for_each_entry(class, hash_head, hash_entry) { + if (DEBUG_LOCKS_WARN_ON(!irqs_disabled())) + return NULL; + + list_for_each_entry_rcu(class, hash_head, hash_entry) { if (class->key == key) { /* * Huh! same key, different name? Did someone trample @@ -728,7 +730,8 @@ register_lock_class(struct lockdep_map *lock, unsigned int subclass, int force) struct lockdep_subclass_key *key; struct list_head *hash_head; struct lock_class *class; - unsigned long flags; + + DEBUG_LOCKS_WARN_ON(!irqs_disabled()); class = look_up_lock_class(lock, subclass); if (likely(class)) @@ -750,28 +753,26 @@ register_lock_class(struct lockdep_map *lock, unsigned int subclass, int force) key = lock->key->subkeys + subclass; hash_head = classhashentry(key); - raw_local_irq_save(flags); if (!graph_lock()) { - raw_local_irq_restore(flags); return NULL; } /* * We have to do the hash-walk again, to avoid races * with another CPU: */ - list_for_each_entry(class, hash_head, hash_entry) + list_for_each_entry_rcu(class, hash_head, hash_entry) { if (class->key == key) goto out_unlock_set; + } + /* * Allocate a new key from the static array, and add it to * the hash: */ if (nr_lock_classes >= MAX_LOCKDEP_KEYS) { if (!debug_locks_off_graph_unlock()) { - raw_local_irq_restore(flags); return NULL; } - raw_local_irq_restore(flags); print_lockdep_off("BUG: MAX_LOCKDEP_KEYS too low!"); dump_stack(); @@ -798,7 +799,6 @@ register_lock_class(struct lockdep_map *lock, unsigned int subclass, int force) if (verbose(class)) { graph_unlock(); - raw_local_irq_restore(flags); printk("\nnew class %p: %s", class->key, class->name); if (class->name_version > 1) @@ -806,15 +806,12 @@ register_lock_class(struct lockdep_map *lock, unsigned int subclass, int force) printk("\n"); dump_stack(); - raw_local_irq_save(flags); if (!graph_lock()) { - raw_local_irq_restore(flags); return NULL; } } out_unlock_set: graph_unlock(); - raw_local_irq_restore(flags); out_set_class_cache: if (!subclass || force) @@ -870,11 +867,9 @@ static int add_lock_to_list(struct lock_class *class, struct lock_class *this, entry->distance = distance; entry->trace = *trace; /* - * Since we never remove from the dependency list, the list can - * be walked lockless by other CPUs, it's only allocation - * that must be protected by the spinlock. But this also means - * we must make new entries visible only once writes to the - * entry become visible - hence the RCU op: + * Both allocation and removal are done under the graph lock; but + * iteration is under RCU-sched; see look_up_lock_class() and + * lockdep_free_key_range(). */ list_add_tail_rcu(&entry->entry, head); @@ -1025,7 +1020,9 @@ static int __bfs(struct lock_list *source_entry, else head = &lock->class->locks_before; - list_for_each_entry(entry, head, entry) { + DEBUG_LOCKS_WARN_ON(!irqs_disabled()); + + list_for_each_entry_rcu(entry, head, entry) { if (!lock_accessed(entry)) { unsigned int cq_depth; mark_lock_accessed(entry, lock); @@ -2022,7 +2019,7 @@ static inline int lookup_chain_cache(struct task_struct *curr, * We can walk it lock-free, because entries only get added * to the hash: */ - list_for_each_entry(chain, hash_head, entry) { + list_for_each_entry_rcu(chain, hash_head, entry) { if (chain->chain_key == chain_key) { cache_hit: debug_atomic_inc(chain_lookup_hits); @@ -2996,8 +2993,18 @@ void lockdep_init_map(struct lockdep_map *lock, const char *name, if (unlikely(!debug_locks)) return; - if (subclass) + if (subclass) { + unsigned long flags; + + if (DEBUG_LOCKS_WARN_ON(current->lockdep_recursion)) + return; + + raw_local_irq_save(flags); + current->lockdep_recursion = 1; register_lock_class(lock, subclass, 1); + current->lockdep_recursion = 0; + raw_local_irq_restore(flags); + } } EXPORT_SYMBOL_GPL(lockdep_init_map); @@ -3887,9 +3894,17 @@ static inline int within(const void *addr, void *start, unsigned long size) return addr >= start && addr < start + size; } +/* + * Used in module.c to remove lock classes from memory that is going to be + * freed; and possibly re-used by other modules. + * + * We will have had one sync_sched() before getting here, so we're guaranteed + * nobody will look up these exact classes -- they're properly dead but still + * allocated. + */ void lockdep_free_key_range(void *start, unsigned long size) { - struct lock_class *class, *next; + struct lock_class *class; struct list_head *head; unsigned long flags; int i; @@ -3905,7 +3920,7 @@ void lockdep_free_key_range(void *start, unsigned long size) head = classhash_table + i; if (list_empty(head)) continue; - list_for_each_entry_safe(class, next, head, hash_entry) { + list_for_each_entry_rcu(class, head, hash_entry) { if (within(class->key, start, size)) zap_class(class); else if (within(class->name, start, size)) @@ -3916,11 +3931,25 @@ void lockdep_free_key_range(void *start, unsigned long size) if (locked) graph_unlock(); raw_local_irq_restore(flags); + + /* + * Wait for any possible iterators from look_up_lock_class() to pass + * before continuing to free the memory they refer to. + * + * sync_sched() is sufficient because the read-side is IRQ disable. + */ + synchronize_sched(); + + /* + * XXX at this point we could return the resources to the pool; + * instead we leak them. We would need to change to bitmap allocators + * instead of the linear allocators we have now. + */ } void lockdep_reset_lock(struct lockdep_map *lock) { - struct lock_class *class, *next; + struct lock_class *class; struct list_head *head; unsigned long flags; int i, j; @@ -3948,7 +3977,7 @@ void lockdep_reset_lock(struct lockdep_map *lock) head = classhash_table + i; if (list_empty(head)) continue; - list_for_each_entry_safe(class, next, head, hash_entry) { + list_for_each_entry_rcu(class, head, hash_entry) { int match = 0; for (j = 0; j < NR_LOCKDEP_CACHING_CLASSES; j++) diff --git a/kernel/module.c b/kernel/module.c index b3d634ed06c9..99fdf94efce8 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -1865,7 +1865,7 @@ static void free_module(struct module *mod) kfree(mod->args); percpu_modfree(mod); - /* Free lock-classes: */ + /* Free lock-classes; relies on the preceding sync_rcu(). */ lockdep_free_key_range(mod->module_core, mod->core_size); /* Finally, free the core (containing the module structure) */ @@ -3349,9 +3349,6 @@ static int load_module(struct load_info *info, const char __user *uargs, module_bug_cleanup(mod); mutex_unlock(&module_mutex); - /* Free lock-classes: */ - lockdep_free_key_range(mod->module_core, mod->core_size); - /* we can't deallocate the module until we clear memory protection */ unset_module_init_ro_nx(mod); unset_module_core_ro_nx(mod); @@ -3375,6 +3372,9 @@ static int load_module(struct load_info *info, const char __user *uargs, synchronize_rcu(); mutex_unlock(&module_mutex); free_module: + /* Free lock-classes; relies on the preceding sync_rcu() */ + lockdep_free_key_range(mod->module_core, mod->core_size); + module_deallocate(mod, info); free_copy: free_copy(info); -- cgit v1.2.1 From a127d2bcf1fbc8c8e0b5cf0dab54f7d3ff50ce47 Mon Sep 17 00:00:00 2001 From: Preeti U Murthy Date: Wed, 18 Mar 2015 16:19:27 +0530 Subject: timers/tick/broadcast-hrtimer: Fix suspicious RCU usage in idle loop The hrtimer mode of broadcast queues hrtimers in the idle entry path so as to wakeup cpus in deep idle states. The associated call graph is : cpuidle_idle_call() |____ clockevents_notify(CLOCK_EVT_NOTIFY_BROADCAST_ENTER, ....)) |_____tick_broadcast_set_event() |____clockevents_program_event() |____bc_set_next() The hrtimer_{start/cancel} functions call into tracing which uses RCU. But it is not legal to call into RCU in cpuidle because it is one of the quiescent states. Hence protect this region with RCU_NONIDLE which informs RCU that the cpu is momentarily non-idle. As an aside it is helpful to point out that the clock event device that is programmed here is not a per-cpu clock device; it is a pseudo clock device, used by the broadcast framework alone. The per-cpu clock device programming never goes through bc_set_next(). Signed-off-by: Preeti U Murthy Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Paul E. McKenney Cc: linuxppc-dev@ozlabs.org Cc: mpe@ellerman.id.au Cc: tglx@linutronix.de Link: http://lkml.kernel.org/r/20150318104705.17763.56668.stgit@preeti.in.ibm.com Signed-off-by: Ingo Molnar --- kernel/time/tick-broadcast-hrtimer.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) diff --git a/kernel/time/tick-broadcast-hrtimer.c b/kernel/time/tick-broadcast-hrtimer.c index eb682d5c697c..6aac4beedbbe 100644 --- a/kernel/time/tick-broadcast-hrtimer.c +++ b/kernel/time/tick-broadcast-hrtimer.c @@ -49,6 +49,7 @@ static void bc_set_mode(enum clock_event_mode mode, */ static int bc_set_next(ktime_t expires, struct clock_event_device *bc) { + int bc_moved; /* * We try to cancel the timer first. If the callback is on * flight on some other cpu then we let it handle it. If we @@ -60,9 +61,15 @@ static int bc_set_next(ktime_t expires, struct clock_event_device *bc) * restart the timer because we are in the callback, but we * can set the expiry time and let the callback return * HRTIMER_RESTART. + * + * Since we are in the idle loop at this point and because + * hrtimer_{start/cancel} functions call into tracing, + * calls to these functions must be bound within RCU_NONIDLE. */ - if (hrtimer_try_to_cancel(&bctimer) >= 0) { - hrtimer_start(&bctimer, expires, HRTIMER_MODE_ABS_PINNED); + RCU_NONIDLE(bc_moved = (hrtimer_try_to_cancel(&bctimer) >= 0) ? + !hrtimer_start(&bctimer, expires, HRTIMER_MODE_ABS_PINNED) : + 0); + if (bc_moved) { /* Bind the "device" to the cpu */ bc->bound_on = smp_processor_id(); } else if (bc->bound_on == smp_processor_id()) { -- cgit v1.2.1 From 47514da3ac20150cdf764466fbc2010c0fca0163 Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Sun, 15 Mar 2015 21:32:12 +0100 Subject: parisc: Add compile-time check when adding new syscalls Signed-off-by: Helge Deller --- arch/parisc/kernel/syscall_table.S | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/arch/parisc/kernel/syscall_table.S b/arch/parisc/kernel/syscall_table.S index 5a8997d63899..8eefb12d1d33 100644 --- a/arch/parisc/kernel/syscall_table.S +++ b/arch/parisc/kernel/syscall_table.S @@ -55,8 +55,8 @@ #define ENTRY_COMP(_name_) .word sys_##_name_ #endif - ENTRY_SAME(restart_syscall) /* 0 */ - ENTRY_SAME(exit) +90: ENTRY_SAME(restart_syscall) /* 0 */ +91: ENTRY_SAME(exit) ENTRY_SAME(fork_wrapper) ENTRY_SAME(read) ENTRY_SAME(write) @@ -439,7 +439,10 @@ ENTRY_SAME(bpf) ENTRY_COMP(execveat) - /* Nothing yet */ + +.ifne (. - 90b) - (__NR_Linux_syscalls * (91b - 90b)) +.error "size of syscall table does not fit value of __NR_Linux_syscalls" +.endif #undef ENTRY_SAME #undef ENTRY_DIFF -- cgit v1.2.1 From 0e0da48dee8dfbcc0df4b8e2ff4efc7a2c89ba6b Mon Sep 17 00:00:00 2001 From: Mikulas Patocka Date: Wed, 18 Mar 2015 13:42:38 -0400 Subject: parisc: mm: don't count preallocated pmds The patch dc6c9a35b66b520cf67e05d8ca60ebecad3b0479 that counts pmds allocated for a process introduced a bug on 64-bit PA-RISC kernels. The PA-RISC architecture preallocates one pmd with each pgd. This preallocated pmd can never be freed - pmd_free does nothing when it is called with this pmd. When the kernel attempts to free this preallocated pmd, it decreases the count of allocated pmds. The result is that the counter underflows and this error is reported. This patch fixes the bug by artifically incrementing the counter in pmd_free when the kernel tries to free the preallocated pmd. Signed-off-by: Mikulas Patocka Acked-by: Kirill A. Shutemov Signed-off-by: Helge Deller --- arch/parisc/include/asm/pgalloc.h | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/arch/parisc/include/asm/pgalloc.h b/arch/parisc/include/asm/pgalloc.h index f213f5b4c423..63e9ecae1310 100644 --- a/arch/parisc/include/asm/pgalloc.h +++ b/arch/parisc/include/asm/pgalloc.h @@ -74,8 +74,13 @@ static inline void pmd_free(struct mm_struct *mm, pmd_t *pmd) { #ifdef CONFIG_64BIT if(pmd_flag(*pmd) & PxD_FLAG_ATTACHED) - /* This is the permanent pmd attached to the pgd; - * cannot free it */ + /* + * This is the permanent pmd attached to the pgd; + * cannot free it. + * Increment the counter to compensate for the decrement + * done by generic mm code. + */ + mm_inc_nr_pmds(mm); return; #endif free_pages((unsigned long)pmd, PMD_ORDER); -- cgit v1.2.1 From 2e3f0ab2bb4853694570b9610b1fcfbfa8fd295b Mon Sep 17 00:00:00 2001 From: Helge Deller Date: Mon, 16 Mar 2015 21:17:50 +0100 Subject: parisc: Fix pmd code to depend on PT_NLEVELS value, not on CONFIG_64BIT Make the code which sets up the pmd depend on PT_NLEVELS == 3, not on CONFIG_64BIT. The reason is, that a 64bit kernel with a page size greater than 4k doesn't need the pmd and thus has PT_NLEVELS = 2. Signed-off-by: Helge Deller --- arch/parisc/include/asm/pgalloc.h | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) diff --git a/arch/parisc/include/asm/pgalloc.h b/arch/parisc/include/asm/pgalloc.h index 63e9ecae1310..d17437238a2c 100644 --- a/arch/parisc/include/asm/pgalloc.h +++ b/arch/parisc/include/asm/pgalloc.h @@ -26,7 +26,7 @@ static inline pgd_t *pgd_alloc(struct mm_struct *mm) if (likely(pgd != NULL)) { memset(pgd, 0, PAGE_SIZE< Date: Mon, 23 Mar 2015 11:17:56 +0000 Subject: metag: Fix ioremap_wc/ioremap_cached build errors MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit When ioremap_wc() or ioremap_cached() are used without first including asm/pgtable.h, the _PAGE_CACHEABLE or _PAGE_WR_COMBINE definitions aren't found, resulting in build errors like the following (in next-20150323 due to "lib: devres: add a helper function for ioremap_wc"): lib/devres.c: In function ‘devm_ioremap_wc’: lib/devres.c:91: error: ‘_PAGE_WR_COMBINE’ undeclared We can't easily include asm/pgtable.h in asm/io.h due to dependency problems, so split out the _PAGE_* definitions from asm/pgtable.h into a separate asm/pgtable-bits.h header (as a couple of other architectures already do), and include that in io.h instead. Signed-off-by: James Hogan Cc: linux-metag@vger.kernel.org Cc: Abhilash Kesavan Cc: Greg Kroah-Hartman --- arch/metag/include/asm/io.h | 1 + arch/metag/include/asm/pgtable-bits.h | 104 ++++++++++++++++++++++++++++++++++ arch/metag/include/asm/pgtable.h | 95 +------------------------------ 3 files changed, 106 insertions(+), 94 deletions(-) create mode 100644 arch/metag/include/asm/pgtable-bits.h diff --git a/arch/metag/include/asm/io.h b/arch/metag/include/asm/io.h index 9359e5048442..d5779b0ec573 100644 --- a/arch/metag/include/asm/io.h +++ b/arch/metag/include/asm/io.h @@ -2,6 +2,7 @@ #define _ASM_METAG_IO_H #include +#include #define IO_SPACE_LIMIT 0 diff --git a/arch/metag/include/asm/pgtable-bits.h b/arch/metag/include/asm/pgtable-bits.h new file mode 100644 index 000000000000..25ba6729f496 --- /dev/null +++ b/arch/metag/include/asm/pgtable-bits.h @@ -0,0 +1,104 @@ +/* + * Meta page table definitions. + */ + +#ifndef _METAG_PGTABLE_BITS_H +#define _METAG_PGTABLE_BITS_H + +#include + +/* + * Definitions for MMU descriptors + * + * These are the hardware bits in the MMCU pte entries. + * Derived from the Meta toolkit headers. + */ +#define _PAGE_PRESENT MMCU_ENTRY_VAL_BIT +#define _PAGE_WRITE MMCU_ENTRY_WR_BIT +#define _PAGE_PRIV MMCU_ENTRY_PRIV_BIT +/* Write combine bit - this can cause writes to occur out of order */ +#define _PAGE_WR_COMBINE MMCU_ENTRY_WRC_BIT +/* Sys coherent bit - this bit is never used by Linux */ +#define _PAGE_SYS_COHERENT MMCU_ENTRY_SYS_BIT +#define _PAGE_ALWAYS_ZERO_1 0x020 +#define _PAGE_CACHE_CTRL0 0x040 +#define _PAGE_CACHE_CTRL1 0x080 +#define _PAGE_ALWAYS_ZERO_2 0x100 +#define _PAGE_ALWAYS_ZERO_3 0x200 +#define _PAGE_ALWAYS_ZERO_4 0x400 +#define _PAGE_ALWAYS_ZERO_5 0x800 + +/* These are software bits that we stuff into the gaps in the hardware + * pte entries that are not used. Note, these DO get stored in the actual + * hardware, but the hardware just does not use them. + */ +#define _PAGE_ACCESSED _PAGE_ALWAYS_ZERO_1 +#define _PAGE_DIRTY _PAGE_ALWAYS_ZERO_2 + +/* Pages owned, and protected by, the kernel. */ +#define _PAGE_KERNEL _PAGE_PRIV + +/* No cacheing of this page */ +#define _PAGE_CACHE_WIN0 (MMCU_CWIN_UNCACHED << MMCU_ENTRY_CWIN_S) +/* burst cacheing - good for data streaming */ +#define _PAGE_CACHE_WIN1 (MMCU_CWIN_BURST << MMCU_ENTRY_CWIN_S) +/* One cache way per thread */ +#define _PAGE_CACHE_WIN2 (MMCU_CWIN_C1SET << MMCU_ENTRY_CWIN_S) +/* Full on cacheing */ +#define _PAGE_CACHE_WIN3 (MMCU_CWIN_CACHED << MMCU_ENTRY_CWIN_S) + +#define _PAGE_CACHEABLE (_PAGE_CACHE_WIN3 | _PAGE_WR_COMBINE) + +/* which bits are used for cache control ... */ +#define _PAGE_CACHE_MASK (_PAGE_CACHE_CTRL0 | _PAGE_CACHE_CTRL1 | \ + _PAGE_WR_COMBINE) + +/* This is a mask of the bits that pte_modify is allowed to change. */ +#define _PAGE_CHG_MASK (PAGE_MASK) + +#define _PAGE_SZ_SHIFT 1 +#define _PAGE_SZ_4K (0x0) +#define _PAGE_SZ_8K (0x1 << _PAGE_SZ_SHIFT) +#define _PAGE_SZ_16K (0x2 << _PAGE_SZ_SHIFT) +#define _PAGE_SZ_32K (0x3 << _PAGE_SZ_SHIFT) +#define _PAGE_SZ_64K (0x4 << _PAGE_SZ_SHIFT) +#define _PAGE_SZ_128K (0x5 << _PAGE_SZ_SHIFT) +#define _PAGE_SZ_256K (0x6 << _PAGE_SZ_SHIFT) +#define _PAGE_SZ_512K (0x7 << _PAGE_SZ_SHIFT) +#define _PAGE_SZ_1M (0x8 << _PAGE_SZ_SHIFT) +#define _PAGE_SZ_2M (0x9 << _PAGE_SZ_SHIFT) +#define _PAGE_SZ_4M (0xa << _PAGE_SZ_SHIFT) +#define _PAGE_SZ_MASK (0xf << _PAGE_SZ_SHIFT) + +#if defined(CONFIG_PAGE_SIZE_4K) +#define _PAGE_SZ (_PAGE_SZ_4K) +#elif defined(CONFIG_PAGE_SIZE_8K) +#define _PAGE_SZ (_PAGE_SZ_8K) +#elif defined(CONFIG_PAGE_SIZE_16K) +#define _PAGE_SZ (_PAGE_SZ_16K) +#endif +#define _PAGE_TABLE (_PAGE_SZ | _PAGE_PRESENT) + +#if defined(CONFIG_HUGETLB_PAGE_SIZE_8K) +# define _PAGE_SZHUGE (_PAGE_SZ_8K) +#elif defined(CONFIG_HUGETLB_PAGE_SIZE_16K) +# define _PAGE_SZHUGE (_PAGE_SZ_16K) +#elif defined(CONFIG_HUGETLB_PAGE_SIZE_32K) +# define _PAGE_SZHUGE (_PAGE_SZ_32K) +#elif defined(CONFIG_HUGETLB_PAGE_SIZE_64K) +# define _PAGE_SZHUGE (_PAGE_SZ_64K) +#elif defined(CONFIG_HUGETLB_PAGE_SIZE_128K) +# define _PAGE_SZHUGE (_PAGE_SZ_128K) +#elif defined(CONFIG_HUGETLB_PAGE_SIZE_256K) +# define _PAGE_SZHUGE (_PAGE_SZ_256K) +#elif defined(CONFIG_HUGETLB_PAGE_SIZE_512K) +# define _PAGE_SZHUGE (_PAGE_SZ_512K) +#elif defined(CONFIG_HUGETLB_PAGE_SIZE_1M) +# define _PAGE_SZHUGE (_PAGE_SZ_1M) +#elif defined(CONFIG_HUGETLB_PAGE_SIZE_2M) +# define _PAGE_SZHUGE (_PAGE_SZ_2M) +#elif defined(CONFIG_HUGETLB_PAGE_SIZE_4M) +# define _PAGE_SZHUGE (_PAGE_SZ_4M) +#endif + +#endif /* _METAG_PGTABLE_BITS_H */ diff --git a/arch/metag/include/asm/pgtable.h b/arch/metag/include/asm/pgtable.h index d0604c0a8702..ffa3a3a2ecad 100644 --- a/arch/metag/include/asm/pgtable.h +++ b/arch/metag/include/asm/pgtable.h @@ -5,6 +5,7 @@ #ifndef _METAG_PGTABLE_H #define _METAG_PGTABLE_H +#include #include /* Invalid regions on Meta: 0x00000000-0x001FFFFF and 0xFFFF0000-0xFFFFFFFF */ @@ -20,100 +21,6 @@ #define VMALLOC_END 0x7FFFFFFF #endif -/* - * Definitions for MMU descriptors - * - * These are the hardware bits in the MMCU pte entries. - * Derived from the Meta toolkit headers. - */ -#define _PAGE_PRESENT MMCU_ENTRY_VAL_BIT -#define _PAGE_WRITE MMCU_ENTRY_WR_BIT -#define _PAGE_PRIV MMCU_ENTRY_PRIV_BIT -/* Write combine bit - this can cause writes to occur out of order */ -#define _PAGE_WR_COMBINE MMCU_ENTRY_WRC_BIT -/* Sys coherent bit - this bit is never used by Linux */ -#define _PAGE_SYS_COHERENT MMCU_ENTRY_SYS_BIT -#define _PAGE_ALWAYS_ZERO_1 0x020 -#define _PAGE_CACHE_CTRL0 0x040 -#define _PAGE_CACHE_CTRL1 0x080 -#define _PAGE_ALWAYS_ZERO_2 0x100 -#define _PAGE_ALWAYS_ZERO_3 0x200 -#define _PAGE_ALWAYS_ZERO_4 0x400 -#define _PAGE_ALWAYS_ZERO_5 0x800 - -/* These are software bits that we stuff into the gaps in the hardware - * pte entries that are not used. Note, these DO get stored in the actual - * hardware, but the hardware just does not use them. - */ -#define _PAGE_ACCESSED _PAGE_ALWAYS_ZERO_1 -#define _PAGE_DIRTY _PAGE_ALWAYS_ZERO_2 - -/* Pages owned, and protected by, the kernel. */ -#define _PAGE_KERNEL _PAGE_PRIV - -/* No cacheing of this page */ -#define _PAGE_CACHE_WIN0 (MMCU_CWIN_UNCACHED << MMCU_ENTRY_CWIN_S) -/* burst cacheing - good for data streaming */ -#define _PAGE_CACHE_WIN1 (MMCU_CWIN_BURST << MMCU_ENTRY_CWIN_S) -/* One cache way per thread */ -#define _PAGE_CACHE_WIN2 (MMCU_CWIN_C1SET << MMCU_ENTRY_CWIN_S) -/* Full on cacheing */ -#define _PAGE_CACHE_WIN3 (MMCU_CWIN_CACHED << MMCU_ENTRY_CWIN_S) - -#define _PAGE_CACHEABLE (_PAGE_CACHE_WIN3 | _PAGE_WR_COMBINE) - -/* which bits are used for cache control ... */ -#define _PAGE_CACHE_MASK (_PAGE_CACHE_CTRL0 | _PAGE_CACHE_CTRL1 | \ - _PAGE_WR_COMBINE) - -/* This is a mask of the bits that pte_modify is allowed to change. */ -#define _PAGE_CHG_MASK (PAGE_MASK) - -#define _PAGE_SZ_SHIFT 1 -#define _PAGE_SZ_4K (0x0) -#define _PAGE_SZ_8K (0x1 << _PAGE_SZ_SHIFT) -#define _PAGE_SZ_16K (0x2 << _PAGE_SZ_SHIFT) -#define _PAGE_SZ_32K (0x3 << _PAGE_SZ_SHIFT) -#define _PAGE_SZ_64K (0x4 << _PAGE_SZ_SHIFT) -#define _PAGE_SZ_128K (0x5 << _PAGE_SZ_SHIFT) -#define _PAGE_SZ_256K (0x6 << _PAGE_SZ_SHIFT) -#define _PAGE_SZ_512K (0x7 << _PAGE_SZ_SHIFT) -#define _PAGE_SZ_1M (0x8 << _PAGE_SZ_SHIFT) -#define _PAGE_SZ_2M (0x9 << _PAGE_SZ_SHIFT) -#define _PAGE_SZ_4M (0xa << _PAGE_SZ_SHIFT) -#define _PAGE_SZ_MASK (0xf << _PAGE_SZ_SHIFT) - -#if defined(CONFIG_PAGE_SIZE_4K) -#define _PAGE_SZ (_PAGE_SZ_4K) -#elif defined(CONFIG_PAGE_SIZE_8K) -#define _PAGE_SZ (_PAGE_SZ_8K) -#elif defined(CONFIG_PAGE_SIZE_16K) -#define _PAGE_SZ (_PAGE_SZ_16K) -#endif -#define _PAGE_TABLE (_PAGE_SZ | _PAGE_PRESENT) - -#if defined(CONFIG_HUGETLB_PAGE_SIZE_8K) -# define _PAGE_SZHUGE (_PAGE_SZ_8K) -#elif defined(CONFIG_HUGETLB_PAGE_SIZE_16K) -# define _PAGE_SZHUGE (_PAGE_SZ_16K) -#elif defined(CONFIG_HUGETLB_PAGE_SIZE_32K) -# define _PAGE_SZHUGE (_PAGE_SZ_32K) -#elif defined(CONFIG_HUGETLB_PAGE_SIZE_64K) -# define _PAGE_SZHUGE (_PAGE_SZ_64K) -#elif defined(CONFIG_HUGETLB_PAGE_SIZE_128K) -# define _PAGE_SZHUGE (_PAGE_SZ_128K) -#elif defined(CONFIG_HUGETLB_PAGE_SIZE_256K) -# define _PAGE_SZHUGE (_PAGE_SZ_256K) -#elif defined(CONFIG_HUGETLB_PAGE_SIZE_512K) -# define _PAGE_SZHUGE (_PAGE_SZ_512K) -#elif defined(CONFIG_HUGETLB_PAGE_SIZE_1M) -# define _PAGE_SZHUGE (_PAGE_SZ_1M) -#elif defined(CONFIG_HUGETLB_PAGE_SIZE_2M) -# define _PAGE_SZHUGE (_PAGE_SZ_2M) -#elif defined(CONFIG_HUGETLB_PAGE_SIZE_4M) -# define _PAGE_SZHUGE (_PAGE_SZ_4M) -#endif - /* * The Linux memory management assumes a three-level page table setup. On * Meta, we use that, but "fold" the mid level into the top-level page -- cgit v1.2.1 From 6761dd3185a76c5d9ae0141e08f123d97d7f303b Mon Sep 17 00:00:00 2001 From: Tvrtko Ursulin Date: Mon, 23 Mar 2015 11:10:32 +0000 Subject: drm/i915/skl: Extract tile height code into a helper function It will be used in a later patch and also convert all height parameters from int to unsigned int. v2: Rebased for fb modifiers. v3: Fixed v2 rebase. v4: * Height should be unsigned int. * Make it take pixel_format for consistency and simplicity. Signed-off-by: Tvrtko Ursulin Reviewed-by: Michel Thierry (v1) Reviewed-by: Joonas Lahtinen (v4) Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_display.c | 43 +++++++++++++++++++++--------------- drivers/gpu/drm/i915/intel_drv.h | 7 +++--- 2 files changed, 29 insertions(+), 21 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 60230dcf532f..3b9ce89d2a98 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2233,13 +2233,12 @@ static bool need_vtd_wa(struct drm_device *dev) return false; } -int -intel_fb_align_height(struct drm_device *dev, int height, - uint32_t pixel_format, - uint64_t fb_format_modifier) +static unsigned int +intel_tile_height(struct drm_device *dev, uint32_t pixel_format, + uint64_t fb_format_modifier) { - int tile_height; - uint32_t bits_per_pixel; + unsigned int tile_height; + uint32_t pixel_bytes; switch (fb_format_modifier) { case DRM_FORMAT_MOD_NONE: @@ -2252,20 +2251,20 @@ intel_fb_align_height(struct drm_device *dev, int height, tile_height = 32; break; case I915_FORMAT_MOD_Yf_TILED: - bits_per_pixel = drm_format_plane_cpp(pixel_format, 0) * 8; - switch (bits_per_pixel) { + pixel_bytes = drm_format_plane_cpp(pixel_format, 0); + switch (pixel_bytes) { default: - case 8: + case 1: tile_height = 64; break; - case 16: - case 32: + case 2: + case 4: tile_height = 32; break; - case 64: + case 8: tile_height = 16; break; - case 128: + case 16: WARN_ONCE(1, "128-bit pixels are not supported for display!"); tile_height = 16; @@ -2278,7 +2277,15 @@ intel_fb_align_height(struct drm_device *dev, int height, break; } - return ALIGN(height, tile_height); + return tile_height; +} + +unsigned int +intel_fb_align_height(struct drm_device *dev, unsigned int height, + uint32_t pixel_format, uint64_t fb_format_modifier) +{ + return ALIGN(height, intel_tile_height(dev, pixel_format, + fb_format_modifier)); } int @@ -6811,7 +6818,7 @@ i9xx_get_initial_plane_config(struct intel_crtc *crtc, u32 val, base, offset; int pipe = crtc->pipe, plane = crtc->plane; int fourcc, pixel_format; - int aligned_height; + unsigned int aligned_height; struct drm_framebuffer *fb; struct intel_framebuffer *intel_fb; @@ -7849,7 +7856,7 @@ skylake_get_initial_plane_config(struct intel_crtc *crtc, u32 val, base, offset, stride_mult, tiling; int pipe = crtc->pipe; int fourcc, pixel_format; - int aligned_height; + unsigned int aligned_height; struct drm_framebuffer *fb; struct intel_framebuffer *intel_fb; @@ -7957,7 +7964,7 @@ ironlake_get_initial_plane_config(struct intel_crtc *crtc, u32 val, base, offset; int pipe = crtc->pipe; int fourcc, pixel_format; - int aligned_height; + unsigned int aligned_height; struct drm_framebuffer *fb; struct intel_framebuffer *intel_fb; @@ -12883,7 +12890,7 @@ static int intel_framebuffer_init(struct drm_device *dev, struct drm_mode_fb_cmd2 *mode_cmd, struct drm_i915_gem_object *obj) { - int aligned_height; + unsigned int aligned_height; int ret; u32 pitch_limit, stride_alignment; diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index d2a4de0e4f4a..e974dd6fe25f 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -901,9 +901,10 @@ void intel_frontbuffer_flip(struct drm_device *dev, intel_frontbuffer_flush(dev, frontbuffer_bits); } -int intel_fb_align_height(struct drm_device *dev, int height, - uint32_t pixel_format, - uint64_t fb_format_modifier); +unsigned int intel_fb_align_height(struct drm_device *dev, + unsigned int height, + uint32_t pixel_format, + uint64_t fb_format_modifier); void intel_fb_obj_flush(struct drm_i915_gem_object *obj, bool retire); u32 intel_fb_stride_alignment(struct drm_device *dev, uint64_t fb_modifier, -- cgit v1.2.1 From e6617330920fdc705220f57a7dd6f8c7da4ae21d Mon Sep 17 00:00:00 2001 From: Tvrtko Ursulin Date: Mon, 23 Mar 2015 11:10:33 +0000 Subject: drm/i915: Use GGTT view when (un)pinning objects to planes To support frame buffer rotation we need to be able to pass on the information on what kind of GGTT view is required for display. This patch just adds the parameter and makes all the callers default to the normal view. v2: Rebased for ggtt view changes. v3: Don't limit PIN_MAPPABLE to normal views just yet. (Joonas Lahtinen) Signed-off-by: Tvrtko Ursulin Reviewed-by: Joonas Lahtinen (v3) [danvet: s/BUG/WARN/ in the patch hunk because. At least where the BUG_ON isn't fatal right away.] Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_drv.h | 14 +++++++++++--- drivers/gpu/drm/i915/i915_gem.c | 22 +++++++++++++--------- drivers/gpu/drm/i915/intel_display.c | 7 ++++--- drivers/gpu/drm/i915/intel_overlay.c | 3 ++- 4 files changed, 30 insertions(+), 16 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index eb38cd1ef216..e7ed5b3d133f 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -2765,8 +2765,10 @@ i915_gem_object_set_to_cpu_domain(struct drm_i915_gem_object *obj, bool write); int __must_check i915_gem_object_pin_to_display_plane(struct drm_i915_gem_object *obj, u32 alignment, - struct intel_engine_cs *pipelined); -void i915_gem_object_unpin_from_display_plane(struct drm_i915_gem_object *obj); + struct intel_engine_cs *pipelined, + const struct i915_ggtt_view *view); +void i915_gem_object_unpin_from_display_plane(struct drm_i915_gem_object *obj, + const struct i915_ggtt_view *view); int i915_gem_object_attach_phys(struct drm_i915_gem_object *obj, int align); int i915_gem_open(struct drm_device *dev, struct drm_file *file); @@ -2875,7 +2877,13 @@ i915_gem_object_ggtt_unbind(struct drm_i915_gem_object *obj) return i915_vma_unbind(i915_gem_obj_to_ggtt(obj)); } -void i915_gem_object_ggtt_unpin(struct drm_i915_gem_object *obj); +void i915_gem_object_ggtt_unpin_view(struct drm_i915_gem_object *obj, + const struct i915_ggtt_view *view); +static inline void +i915_gem_object_ggtt_unpin(struct drm_i915_gem_object *obj) +{ + i915_gem_object_ggtt_unpin_view(obj, &i915_ggtt_view_normal); +} /* i915_gem_context.c */ int __must_check i915_gem_context_init(struct drm_device *dev); diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 84e2a231b03c..31978b07b6a6 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -3871,7 +3871,8 @@ static bool is_pin_display(struct drm_i915_gem_object *obj) int i915_gem_object_pin_to_display_plane(struct drm_i915_gem_object *obj, u32 alignment, - struct intel_engine_cs *pipelined) + struct intel_engine_cs *pipelined, + const struct i915_ggtt_view *view) { u32 old_read_domains, old_write_domain; bool was_pin_display; @@ -3907,7 +3908,7 @@ i915_gem_object_pin_to_display_plane(struct drm_i915_gem_object *obj, * (e.g. libkms for the bootup splash), we have to ensure that we * always use map_and_fenceable for all scanout buffers. */ - ret = i915_gem_obj_ggtt_pin(obj, alignment, PIN_MAPPABLE); + ret = i915_gem_object_ggtt_pin(obj, view, alignment, PIN_MAPPABLE); if (ret) goto err_unpin_display; @@ -3935,9 +3936,11 @@ err_unpin_display: } void -i915_gem_object_unpin_from_display_plane(struct drm_i915_gem_object *obj) +i915_gem_object_unpin_from_display_plane(struct drm_i915_gem_object *obj, + const struct i915_ggtt_view *view) { - i915_gem_object_ggtt_unpin(obj); + i915_gem_object_ggtt_unpin_view(obj, view); + obj->pin_display = is_pin_display(obj); } @@ -4214,15 +4217,16 @@ i915_gem_object_ggtt_pin(struct drm_i915_gem_object *obj, } void -i915_gem_object_ggtt_unpin(struct drm_i915_gem_object *obj) +i915_gem_object_ggtt_unpin_view(struct drm_i915_gem_object *obj, + const struct i915_ggtt_view *view) { - struct i915_vma *vma = i915_gem_obj_to_ggtt(obj); + struct i915_vma *vma = i915_gem_obj_to_ggtt_view(obj, view); BUG_ON(!vma); - BUG_ON(vma->pin_count == 0); - BUG_ON(!i915_gem_obj_ggtt_bound(obj)); + WARN_ON(vma->pin_count == 0); + WARN_ON(!i915_gem_obj_ggtt_bound_view(obj, view->type)); - if (--vma->pin_count == 0) + if (--vma->pin_count == 0 && view->type == I915_GGTT_VIEW_NORMAL) obj->pin_mappable = false; } diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 3b9ce89d2a98..39380f414aff 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2350,7 +2350,8 @@ intel_pin_and_fence_fb_obj(struct drm_plane *plane, intel_runtime_pm_get(dev_priv); dev_priv->mm.interruptible = false; - ret = i915_gem_object_pin_to_display_plane(obj, alignment, pipelined); + ret = i915_gem_object_pin_to_display_plane(obj, alignment, pipelined, + &i915_ggtt_view_normal); if (ret) goto err_interruptible; @@ -2370,7 +2371,7 @@ intel_pin_and_fence_fb_obj(struct drm_plane *plane, return 0; err_unpin: - i915_gem_object_unpin_from_display_plane(obj); + i915_gem_object_unpin_from_display_plane(obj, &i915_ggtt_view_normal); err_interruptible: dev_priv->mm.interruptible = true; intel_runtime_pm_put(dev_priv); @@ -2382,7 +2383,7 @@ static void intel_unpin_fb_obj(struct drm_i915_gem_object *obj) WARN_ON(!mutex_is_locked(&obj->base.dev->struct_mutex)); i915_gem_object_unpin_fence(obj); - i915_gem_object_unpin_from_display_plane(obj); + i915_gem_object_unpin_from_display_plane(obj, &i915_ggtt_view_normal); } /* Computes the linear offset to the base tile and adjusts x, y. bytes per pixel diff --git a/drivers/gpu/drm/i915/intel_overlay.c b/drivers/gpu/drm/i915/intel_overlay.c index 823d1d97a000..dd92122ed95c 100644 --- a/drivers/gpu/drm/i915/intel_overlay.c +++ b/drivers/gpu/drm/i915/intel_overlay.c @@ -720,7 +720,8 @@ static int intel_overlay_do_put_image(struct intel_overlay *overlay, if (ret != 0) return ret; - ret = i915_gem_object_pin_to_display_plane(new_bo, 0, NULL); + ret = i915_gem_object_pin_to_display_plane(new_bo, 0, NULL, + &i915_ggtt_view_normal); if (ret != 0) return ret; -- cgit v1.2.1 From 82bc3b2daa78d99b5c93ef17e8b660e56447c935 Mon Sep 17 00:00:00 2001 From: Tvrtko Ursulin Date: Mon, 23 Mar 2015 11:10:34 +0000 Subject: drm/i915: Pass in plane state when (un)pinning frame buffers Plane state carries the rotation information which is needed for determining the appropriate GGTT view type. This just adds the parameter with the actual usage coming in future patches. Signed-off-by: Tvrtko Ursulin Reviewed-by: Joonas Lahtinen Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_display.c | 18 ++++++++++++------ drivers/gpu/drm/i915/intel_drv.h | 1 + drivers/gpu/drm/i915/intel_fbdev.c | 2 +- 3 files changed, 14 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 39380f414aff..4d8a397893d1 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2291,6 +2291,7 @@ intel_fb_align_height(struct drm_device *dev, unsigned int height, int intel_pin_and_fence_fb_obj(struct drm_plane *plane, struct drm_framebuffer *fb, + const struct drm_plane_state *plane_state, struct intel_engine_cs *pipelined) { struct drm_device *dev = fb->dev; @@ -2378,8 +2379,11 @@ err_interruptible: return ret; } -static void intel_unpin_fb_obj(struct drm_i915_gem_object *obj) +static void intel_unpin_fb_obj(struct drm_framebuffer *fb, + const struct drm_plane_state *plane_state) { + struct drm_i915_gem_object *obj = intel_fb_obj(fb); + WARN_ON(!mutex_is_locked(&obj->base.dev->struct_mutex)); i915_gem_object_unpin_fence(obj); @@ -9318,7 +9322,7 @@ static void intel_unpin_work_fn(struct work_struct *__work) enum pipe pipe = to_intel_crtc(work->crtc)->pipe; mutex_lock(&dev->struct_mutex); - intel_unpin_fb_obj(intel_fb_obj(work->old_fb)); + intel_unpin_fb_obj(work->old_fb, work->crtc->primary->state); drm_gem_object_unreference(&work->pending_flip_obj->base); intel_fbc_update(dev); @@ -10026,7 +10030,8 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc, ring = &dev_priv->ring[RCS]; } - ret = intel_pin_and_fence_fb_obj(crtc->primary, fb, ring); + ret = intel_pin_and_fence_fb_obj(crtc->primary, fb, + crtc->primary->state, ring); if (ret) goto cleanup_pending; @@ -10066,7 +10071,7 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc, return 0; cleanup_unpin: - intel_unpin_fb_obj(obj); + intel_unpin_fb_obj(fb, crtc->primary->state); cleanup_pending: atomic_dec(&intel_crtc->unpin_work_count); mutex_unlock(&dev->struct_mutex); @@ -12022,7 +12027,7 @@ intel_prepare_plane_fb(struct drm_plane *plane, if (ret) DRM_DEBUG_KMS("failed to attach phys object\n"); } else { - ret = intel_pin_and_fence_fb_obj(plane, fb, NULL); + ret = intel_pin_and_fence_fb_obj(plane, fb, new_state, NULL); } if (ret == 0) @@ -12054,7 +12059,7 @@ intel_cleanup_plane_fb(struct drm_plane *plane, if (plane->type != DRM_PLANE_TYPE_CURSOR || !INTEL_INFO(dev)->cursor_needs_physical) { mutex_lock(&dev->struct_mutex); - intel_unpin_fb_obj(obj); + intel_unpin_fb_obj(fb, old_state); mutex_unlock(&dev->struct_mutex); } } @@ -13940,6 +13945,7 @@ void intel_modeset_gem_init(struct drm_device *dev) if (intel_pin_and_fence_fb_obj(c->primary, c->primary->fb, + c->primary->state, NULL)) { DRM_ERROR("failed to pin boot fb on pipe %d\n", to_intel_crtc(c)->pipe); diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index e974dd6fe25f..a0df556ad895 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -959,6 +959,7 @@ void intel_release_load_detect_pipe(struct drm_connector *connector, struct intel_load_detect_pipe *old); int intel_pin_and_fence_fb_obj(struct drm_plane *plane, struct drm_framebuffer *fb, + const struct drm_plane_state *plane_state, struct intel_engine_cs *pipelined); struct drm_framebuffer * __intel_framebuffer_create(struct drm_device *dev, diff --git a/drivers/gpu/drm/i915/intel_fbdev.c b/drivers/gpu/drm/i915/intel_fbdev.c index 757c0d216f80..4e7e7da2e03b 100644 --- a/drivers/gpu/drm/i915/intel_fbdev.c +++ b/drivers/gpu/drm/i915/intel_fbdev.c @@ -151,7 +151,7 @@ static int intelfb_alloc(struct drm_fb_helper *helper, } /* Flush everything out, we'll be doing GTT only from now on */ - ret = intel_pin_and_fence_fb_obj(NULL, fb, NULL); + ret = intel_pin_and_fence_fb_obj(NULL, fb, NULL, NULL); if (ret) { DRM_ERROR("failed to pin obj: %d\n", ret); goto out_fb; -- cgit v1.2.1 From f64b98cd2e40052c17f67f09a081ff292bac0f77 Mon Sep 17 00:00:00 2001 From: Tvrtko Ursulin Date: Mon, 23 Mar 2015 11:10:35 +0000 Subject: drm/i915: Helper function to determine GGTT view from plane state For now only default implementation defaulting to normal view. v2: Some code review cleanups. (Joonas Lahtinen) Signed-off-by: Tvrtko Ursulin Reviewed-by: Joonas Lahtinen (v2) Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_display.c | 25 ++++++++++++++++++++++--- 1 file changed, 22 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 4d8a397893d1..d3fa09bd3e07 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2288,6 +2288,15 @@ intel_fb_align_height(struct drm_device *dev, unsigned int height, fb_format_modifier)); } +static int +intel_fill_fb_ggtt_view(struct i915_ggtt_view *view, struct drm_framebuffer *fb, + const struct drm_plane_state *plane_state) +{ + *view = i915_ggtt_view_normal; + + return 0; +} + int intel_pin_and_fence_fb_obj(struct drm_plane *plane, struct drm_framebuffer *fb, @@ -2297,6 +2306,7 @@ intel_pin_and_fence_fb_obj(struct drm_plane *plane, struct drm_device *dev = fb->dev; struct drm_i915_private *dev_priv = dev->dev_private; struct drm_i915_gem_object *obj = intel_fb_obj(fb); + struct i915_ggtt_view view; u32 alignment; int ret; @@ -2333,6 +2343,10 @@ intel_pin_and_fence_fb_obj(struct drm_plane *plane, return -EINVAL; } + ret = intel_fill_fb_ggtt_view(&view, fb, plane_state); + if (ret) + return ret; + /* Note that the w/a also requires 64 PTE of padding following the * bo. We currently fill all unused PTE with the shadow page and so * we should always have valid PTE following the scanout preventing @@ -2352,7 +2366,7 @@ intel_pin_and_fence_fb_obj(struct drm_plane *plane, dev_priv->mm.interruptible = false; ret = i915_gem_object_pin_to_display_plane(obj, alignment, pipelined, - &i915_ggtt_view_normal); + &view); if (ret) goto err_interruptible; @@ -2372,7 +2386,7 @@ intel_pin_and_fence_fb_obj(struct drm_plane *plane, return 0; err_unpin: - i915_gem_object_unpin_from_display_plane(obj, &i915_ggtt_view_normal); + i915_gem_object_unpin_from_display_plane(obj, &view); err_interruptible: dev_priv->mm.interruptible = true; intel_runtime_pm_put(dev_priv); @@ -2383,11 +2397,16 @@ static void intel_unpin_fb_obj(struct drm_framebuffer *fb, const struct drm_plane_state *plane_state) { struct drm_i915_gem_object *obj = intel_fb_obj(fb); + struct i915_ggtt_view view; + int ret; WARN_ON(!mutex_is_locked(&obj->base.dev->struct_mutex)); + ret = intel_fill_fb_ggtt_view(&view, fb, plane_state); + WARN_ONCE(ret, "Couldn't get view from plane state!"); + i915_gem_object_unpin_fence(obj); - i915_gem_object_unpin_from_display_plane(obj, &i915_ggtt_view_normal); + i915_gem_object_unpin_from_display_plane(obj, &view); } /* Computes the linear offset to the base tile and adjusts x, y. bytes per pixel -- cgit v1.2.1 From 50470bb011c4be278097670bea92462f4e8c8945 Mon Sep 17 00:00:00 2001 From: Tvrtko Ursulin Date: Mon, 23 Mar 2015 11:10:36 +0000 Subject: drm/i915/skl: Support secondary (rotated) frame buffer mapping MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit 90/270 rotated scanout needs a rotated GTT view of the framebuffer. This is put in a separate VMA with a dedicated ggtt view and wired such that it is created when a framebuffer is pinned to a 90/270 rotated plane. Rotation is only possible with Yb/Yf buffers and error is propagated to user space in case of a mismatch. Special rotated page view is constructed at the VMA creation time by borrowing the DMA addresses from obj->pages. v2: * Do not bother with pages for rotated sg list, just populate the DMA addresses. (Daniel Vetter) * Checkpatch cleanup. v3: * Rebased on top of new plane handling (create rotated mapping when setting the rotation property). * Unpin rotated VMA on unpinning from display plane. * Simplify rotation check using bitwise AND. (Chris Wilson) v4: * Fix unpinning of optional rotated mapping so it is really considered to be optional. v5: * Rebased for fb modifier changes. * Rebased for atomic commit. * Only pin needed view for display. (Ville Syrjälä, Daniel Vetter) v6: * Rebased after preparatory work has been extracted out. (Daniel Vetter) v7: * Slightly simplified tiling geometry calculation. * Moved rotated GGTT view implementation into i915_gem_gtt.c (Daniel Vetter) v8: * Do not use i915_gem_obj_size to get object size since that actually returns the size of an VMA which may not exist. * Rebased for ggtt view changes. v9: * Rebased after code review changes on the preceding patches. * Tidy function definitions. (Joonas Lahtinen) For: VIZ-4726 Signed-off-by: Tvrtko Ursulin Reviewed-by: Michel Thierry (v4) Reviewed-by: Joonas Lahtinen Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_gem.c | 4 +- drivers/gpu/drm/i915/i915_gem_gtt.c | 117 +++++++++++++++++++++++++++++++++-- drivers/gpu/drm/i915/i915_gem_gtt.h | 12 ++++ drivers/gpu/drm/i915/intel_display.c | 27 +++++++- drivers/gpu/drm/i915/intel_drv.h | 4 ++ 5 files changed, 158 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 31978b07b6a6..538cae2ca94e 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -3908,7 +3908,9 @@ i915_gem_object_pin_to_display_plane(struct drm_i915_gem_object *obj, * (e.g. libkms for the bootup splash), we have to ensure that we * always use map_and_fenceable for all scanout buffers. */ - ret = i915_gem_object_ggtt_pin(obj, view, alignment, PIN_MAPPABLE); + ret = i915_gem_object_ggtt_pin(obj, view, alignment, + view->type == I915_GGTT_VIEW_NORMAL ? + PIN_MAPPABLE : 0); if (ret) goto err_unpin_display; diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c index 645c3636c571..fc56c112a7de 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c @@ -2500,15 +2500,119 @@ i915_gem_obj_lookup_or_create_ggtt_vma(struct drm_i915_gem_object *obj, } +static void +rotate_pages(dma_addr_t *in, unsigned int width, unsigned int height, + struct sg_table *st) +{ + unsigned int column, row; + unsigned int src_idx; + struct scatterlist *sg = st->sgl; + + st->nents = 0; + + for (column = 0; column < width; column++) { + src_idx = width * (height - 1) + column; + for (row = 0; row < height; row++) { + st->nents++; + /* We don't need the pages, but need to initialize + * the entries so the sg list can be happily traversed. + * The only thing we need are DMA addresses. + */ + sg_set_page(sg, NULL, PAGE_SIZE, 0); + sg_dma_address(sg) = in[src_idx]; + sg_dma_len(sg) = PAGE_SIZE; + sg = sg_next(sg); + src_idx -= width; + } + } +} + +static struct sg_table * +intel_rotate_fb_obj_pages(struct i915_ggtt_view *ggtt_view, + struct drm_i915_gem_object *obj) +{ + struct drm_device *dev = obj->base.dev; + struct intel_rotation_info *rot_info = &ggtt_view->rotation_info; + unsigned long size, pages, rot_pages; + struct sg_page_iter sg_iter; + unsigned long i; + dma_addr_t *page_addr_list; + struct sg_table *st; + unsigned int tile_pitch, tile_height; + unsigned int width_pages, height_pages; + int ret = ENOMEM; + + pages = obj->base.size / PAGE_SIZE; + + /* Calculate tiling geometry. */ + tile_height = intel_tile_height(dev, rot_info->pixel_format, + rot_info->fb_modifier); + tile_pitch = PAGE_SIZE / tile_height; + width_pages = DIV_ROUND_UP(rot_info->pitch, tile_pitch); + height_pages = DIV_ROUND_UP(rot_info->height, tile_height); + rot_pages = width_pages * height_pages; + size = rot_pages * PAGE_SIZE; + + /* Allocate a temporary list of source pages for random access. */ + page_addr_list = drm_malloc_ab(pages, sizeof(dma_addr_t)); + if (!page_addr_list) + return ERR_PTR(ret); + + /* Allocate target SG list. */ + st = kmalloc(sizeof(*st), GFP_KERNEL); + if (!st) + goto err_st_alloc; + + ret = sg_alloc_table(st, rot_pages, GFP_KERNEL); + if (ret) + goto err_sg_alloc; + + /* Populate source page list from the object. */ + i = 0; + for_each_sg_page(obj->pages->sgl, &sg_iter, obj->pages->nents, 0) { + page_addr_list[i] = sg_page_iter_dma_address(&sg_iter); + i++; + } + + /* Rotate the pages. */ + rotate_pages(page_addr_list, width_pages, height_pages, st); + + DRM_DEBUG_KMS( + "Created rotated page mapping for object size %lu (pitch=%u, height=%u, pixel_format=0x%x, %ux%u tiles, %lu pages).\n", + size, rot_info->pitch, rot_info->height, + rot_info->pixel_format, width_pages, height_pages, + rot_pages); + + drm_free_large(page_addr_list); + + return st; + +err_sg_alloc: + kfree(st); +err_st_alloc: + drm_free_large(page_addr_list); + + DRM_DEBUG_KMS( + "Failed to create rotated mapping for object size %lu! (%d) (pitch=%u, height=%u, pixel_format=0x%x, %ux%u tiles, %lu pages)\n", + size, ret, rot_info->pitch, rot_info->height, + rot_info->pixel_format, width_pages, height_pages, + rot_pages); + return ERR_PTR(ret); +} -static inline -int i915_get_ggtt_vma_pages(struct i915_vma *vma) +static inline int +i915_get_ggtt_vma_pages(struct i915_vma *vma) { + int ret = 0; + if (vma->ggtt_view.pages) return 0; if (vma->ggtt_view.type == I915_GGTT_VIEW_NORMAL) vma->ggtt_view.pages = vma->obj->pages; + else if (vma->ggtt_view.type == I915_GGTT_VIEW_ROTATED) + vma->ggtt_view.pages = + intel_rotate_fb_obj_pages(&vma->ggtt_view, vma->obj); else WARN_ONCE(1, "GGTT view %u not implemented!\n", vma->ggtt_view.type); @@ -2516,10 +2620,15 @@ int i915_get_ggtt_vma_pages(struct i915_vma *vma) if (!vma->ggtt_view.pages) { DRM_ERROR("Failed to get pages for GGTT view type %u!\n", vma->ggtt_view.type); - return -EINVAL; + ret = -EINVAL; + } else if (IS_ERR(vma->ggtt_view.pages)) { + ret = PTR_ERR(vma->ggtt_view.pages); + vma->ggtt_view.pages = NULL; + DRM_ERROR("Failed to get pages for VMA view type %u (%d)!\n", + vma->ggtt_view.type, ret); } - return 0; + return ret; } /** diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.h b/drivers/gpu/drm/i915/i915_gem_gtt.h index 75e29f738a1b..0dad42634fd5 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.h +++ b/drivers/gpu/drm/i915/i915_gem_gtt.h @@ -117,12 +117,24 @@ typedef uint64_t gen8_pde_t; enum i915_ggtt_view_type { I915_GGTT_VIEW_NORMAL = 0, + I915_GGTT_VIEW_ROTATED +}; + +struct intel_rotation_info { + unsigned int height; + unsigned int pitch; + uint32_t pixel_format; + uint64_t fb_modifier; }; struct i915_ggtt_view { enum i915_ggtt_view_type type; struct sg_table *pages; + + union { + struct intel_rotation_info rotation_info; + }; }; extern const struct i915_ggtt_view i915_ggtt_view_normal; diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index d3fa09bd3e07..f605b091550f 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2233,7 +2233,7 @@ static bool need_vtd_wa(struct drm_device *dev) return false; } -static unsigned int +unsigned int intel_tile_height(struct drm_device *dev, uint32_t pixel_format, uint64_t fb_format_modifier) { @@ -2292,8 +2292,33 @@ static int intel_fill_fb_ggtt_view(struct i915_ggtt_view *view, struct drm_framebuffer *fb, const struct drm_plane_state *plane_state) { + struct intel_rotation_info *info = &view->rotation_info; + static const struct i915_ggtt_view rotated_view = + { .type = I915_GGTT_VIEW_ROTATED }; + *view = i915_ggtt_view_normal; + if (!plane_state) + return 0; + + if (!(plane_state->rotation & + (BIT(DRM_ROTATE_90) | BIT(DRM_ROTATE_270)))) + return 0; + + *view = rotated_view; + + info->height = fb->height; + info->pixel_format = fb->pixel_format; + info->pitch = fb->pitches[0]; + info->fb_modifier = fb->modifier[0]; + + if (!(info->fb_modifier == I915_FORMAT_MOD_Y_TILED || + info->fb_modifier == I915_FORMAT_MOD_Yf_TILED)) { + DRM_DEBUG_KMS( + "Y or Yf tiling is needed for 90/270 rotation!\n"); + return -EINVAL; + } + return 0; } diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index a0df556ad895..cb00bd09a1c3 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -984,6 +984,10 @@ int intel_plane_atomic_set_property(struct drm_plane *plane, struct drm_property *property, uint64_t val); +unsigned int +intel_tile_height(struct drm_device *dev, uint32_t pixel_format, + uint64_t fb_format_modifier); + /* shared dpll functions */ struct intel_shared_dpll *intel_crtc_to_shared_dpll(struct intel_crtc *crtc); void assert_shared_dpll(struct drm_i915_private *dev_priv, -- cgit v1.2.1 From 121920faf2ccce9aa66a7e2588415c9647b66104 Mon Sep 17 00:00:00 2001 From: Tvrtko Ursulin Date: Mon, 23 Mar 2015 11:10:37 +0000 Subject: drm/i915/skl: Query display address through a wrapper Need to do this in order to support 90/270 rotated display. v2: Pass in drm_plane instead of plane index to intel_obj_display_address. v3: * Renamed intel_obj_display_address to intel_plane_obj_offset. (Chris Wilson) * Simplified rotation check to bitwise AND. (Chris Wilson) v4: * Extracted 90/270 rotation check into a helper function. (Michel Thierry) v5: * Rebased for ggtt view changes. For: VIZ-4545 Signed-off-by: Tvrtko Ursulin Reviewed-by: Michel Thierry Reviewed-by: Joonas Lahtinen Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_display.c | 23 +++++++++++++++++------ drivers/gpu/drm/i915/intel_drv.h | 9 +++++++++ drivers/gpu/drm/i915/intel_sprite.c | 5 ++++- 3 files changed, 30 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index f605b091550f..d1f909900a07 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2301,8 +2301,7 @@ intel_fill_fb_ggtt_view(struct i915_ggtt_view *view, struct drm_framebuffer *fb, if (!plane_state) return 0; - if (!(plane_state->rotation & - (BIT(DRM_ROTATE_90) | BIT(DRM_ROTATE_270)))) + if (!intel_rotation_90_or_270(plane_state->rotation)) return 0; *view = rotated_view; @@ -2900,6 +2899,17 @@ u32 intel_fb_stride_alignment(struct drm_device *dev, uint64_t fb_modifier, } } +unsigned long intel_plane_obj_offset(struct intel_plane *intel_plane, + struct drm_i915_gem_object *obj) +{ + enum i915_ggtt_view_type view = I915_GGTT_VIEW_NORMAL; + + if (intel_rotation_90_or_270(intel_plane->base.state->rotation)) + view = I915_GGTT_VIEW_ROTATED; + + return i915_gem_obj_ggtt_offset_view(obj, view); +} + static void skylake_update_primary_plane(struct drm_crtc *crtc, struct drm_framebuffer *fb, int x, int y) @@ -2910,6 +2920,7 @@ static void skylake_update_primary_plane(struct drm_crtc *crtc, struct drm_i915_gem_object *obj; int pipe = intel_crtc->pipe; u32 plane_ctl, stride_div; + unsigned long surf_addr; if (!intel_crtc->primary_enabled) { I915_WRITE(PLANE_CTL(pipe, 0), 0); @@ -2976,16 +2987,16 @@ static void skylake_update_primary_plane(struct drm_crtc *crtc, obj = intel_fb_obj(fb); stride_div = intel_fb_stride_alignment(dev, fb->modifier[0], fb->pixel_format); + surf_addr = intel_plane_obj_offset(to_intel_plane(crtc->primary), obj); I915_WRITE(PLANE_CTL(pipe, 0), plane_ctl); - I915_WRITE(PLANE_POS(pipe, 0), 0); I915_WRITE(PLANE_OFFSET(pipe, 0), (y << 16) | x); I915_WRITE(PLANE_SIZE(pipe, 0), (intel_crtc->config->pipe_src_h - 1) << 16 | (intel_crtc->config->pipe_src_w - 1)); I915_WRITE(PLANE_STRIDE(pipe, 0), fb->pitches[0] / stride_div); - I915_WRITE(PLANE_SURF(pipe, 0), i915_gem_obj_ggtt_offset(obj)); + I915_WRITE(PLANE_SURF(pipe, 0), surf_addr); POSTING_READ(PLANE_SURF(pipe, 0)); } @@ -10079,8 +10090,8 @@ static int intel_crtc_page_flip(struct drm_crtc *crtc, if (ret) goto cleanup_pending; - work->gtt_offset = - i915_gem_obj_ggtt_offset(obj) + intel_crtc->dspaddr_offset; + work->gtt_offset = intel_plane_obj_offset(to_intel_plane(primary), obj) + + intel_crtc->dspaddr_offset; if (use_mmio_flip(ring, obj)) { ret = intel_queue_mmio_flip(dev, crtc, fb, obj, ring, diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index cb00bd09a1c3..db0e7bfd0baa 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -988,6 +988,12 @@ unsigned int intel_tile_height(struct drm_device *dev, uint32_t pixel_format, uint64_t fb_format_modifier); +static inline bool +intel_rotation_90_or_270(unsigned int rotation) +{ + return rotation & (BIT(DRM_ROTATE_90) | BIT(DRM_ROTATE_270)); +} + /* shared dpll functions */ struct intel_shared_dpll *intel_crtc_to_shared_dpll(struct intel_crtc *crtc); void assert_shared_dpll(struct drm_i915_private *dev_priv, @@ -1042,6 +1048,9 @@ void intel_mode_from_pipe_config(struct drm_display_mode *mode, void intel_crtc_wait_for_pending_flips(struct drm_crtc *crtc); void intel_modeset_preclose(struct drm_device *dev, struct drm_file *file); +unsigned long intel_plane_obj_offset(struct intel_plane *intel_plane, + struct drm_i915_gem_object *obj); + /* intel_dp.c */ void intel_dp_init(struct drm_device *dev, int output_reg, enum port port); bool intel_dp_init_connector(struct intel_digital_port *intel_dig_port, diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c index 0f00209bb220..005a6fd35f6f 100644 --- a/drivers/gpu/drm/i915/intel_sprite.c +++ b/drivers/gpu/drm/i915/intel_sprite.c @@ -193,6 +193,7 @@ skl_update_plane(struct drm_plane *drm_plane, struct drm_crtc *crtc, u32 plane_ctl, stride_div; int pixel_size = drm_format_plane_cpp(fb->pixel_format, 0); const struct drm_intel_sprite_colorkey *key = &intel_plane->ckey; + unsigned long surf_addr; plane_ctl = PLANE_CTL_ENABLE | PLANE_CTL_PIPE_CSC_ENABLE; @@ -280,12 +281,14 @@ skl_update_plane(struct drm_plane *drm_plane, struct drm_crtc *crtc, else if (key->flags & I915_SET_COLORKEY_SOURCE) plane_ctl |= PLANE_CTL_KEY_ENABLE_SOURCE; + surf_addr = intel_plane_obj_offset(intel_plane, obj); + I915_WRITE(PLANE_OFFSET(pipe, plane), (y << 16) | x); I915_WRITE(PLANE_STRIDE(pipe, plane), fb->pitches[0] / stride_div); I915_WRITE(PLANE_POS(pipe, plane), (crtc_y << 16) | crtc_x); I915_WRITE(PLANE_SIZE(pipe, plane), (crtc_h << 16) | crtc_w); I915_WRITE(PLANE_CTL(pipe, plane), plane_ctl); - I915_WRITE(PLANE_SURF(pipe, plane), i915_gem_obj_ggtt_offset(obj)); + I915_WRITE(PLANE_SURF(pipe, plane), surf_addr); POSTING_READ(PLANE_SURF(pipe, plane)); } -- cgit v1.2.1 From 1fc0a8f7c45275c38d3322089313fe2e309c1f17 Mon Sep 17 00:00:00 2001 From: Tvrtko Ursulin Date: Mon, 23 Mar 2015 11:10:38 +0000 Subject: drm/i915/skl: Take 90/270 rotation into account in watermark calculations v2: Pass in rotation info to sprite plane updates as well. v3: Use helper to determine 90/270 rotation. (Michel Thierry) v4: Rebased for fb modifiers and atomic changes. For: VIZ-4546 Signed-off-by: Tvrtko Ursulin Reviewed-by: Michel Thierry (v3) Reviewed-by: Joonas Lahtinen Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_display.c | 27 +++++++++++++++++++++++---- drivers/gpu/drm/i915/intel_drv.h | 4 ++++ drivers/gpu/drm/i915/intel_pm.c | 18 +++++++++++++++++- drivers/gpu/drm/i915/intel_sprite.c | 5 +---- 4 files changed, 45 insertions(+), 9 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index d1f909900a07..35cdb487814a 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -12033,6 +12033,28 @@ static void intel_shared_dpll_init(struct drm_device *dev) BUG_ON(dev_priv->num_shared_dpll > I915_NUM_PLLS); } +/** + * intel_wm_need_update - Check whether watermarks need updating + * @plane: drm plane + * @state: new plane state + * + * Check current plane state versus the new one to determine whether + * watermarks need to be recalculated. + * + * Returns true or false. + */ +bool intel_wm_need_update(struct drm_plane *plane, + struct drm_plane_state *state) +{ + /* Update watermarks on tiling changes. */ + if (!plane->state->fb || !state->fb || + plane->state->fb->modifier[0] != state->fb->modifier[0] || + plane->state->rotation != state->rotation) + return true; + + return false; +} + /** * intel_prepare_plane_fb - Prepare fb for usage on plane * @plane: drm plane to prepare for @@ -12179,10 +12201,7 @@ intel_check_primary_plane(struct drm_plane *plane, intel_crtc->atomic.update_fbc = true; - /* Update watermarks on tiling changes. */ - if (!plane->state->fb || !state->base.fb || - plane->state->fb->modifier[0] != - state->base.fb->modifier[0]) + if (intel_wm_need_update(plane, &state->base)) intel_crtc->atomic.update_wm = true; } diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index db0e7bfd0baa..811a1db8b6ed 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -501,6 +501,7 @@ struct intel_plane_wm_parameters { bool enabled; bool scaled; u64 tiling; + unsigned int rotation; }; struct intel_plane { @@ -994,6 +995,9 @@ intel_rotation_90_or_270(unsigned int rotation) return rotation & (BIT(DRM_ROTATE_90) | BIT(DRM_ROTATE_270)); } +bool intel_wm_need_update(struct drm_plane *plane, + struct drm_plane_state *state); + /* shared dpll functions */ struct intel_shared_dpll *intel_crtc_to_shared_dpll(struct intel_crtc *crtc); void assert_shared_dpll(struct drm_i915_private *dev_priv, diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index e18f0fd22cf2..753a3af07333 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -2840,6 +2840,7 @@ static void skl_compute_wm_pipe_parameters(struct drm_crtc *crtc, } p->plane[0].horiz_pixels = intel_crtc->config->pipe_src_w; p->plane[0].vert_pixels = intel_crtc->config->pipe_src_h; + p->plane[0].rotation = crtc->primary->state->rotation; fb = crtc->cursor->state->fb; if (fb) { @@ -2897,7 +2898,21 @@ static bool skl_compute_plane_wm(const struct drm_i915_private *dev_priv, if (p_params->tiling == I915_FORMAT_MOD_Y_TILED || p_params->tiling == I915_FORMAT_MOD_Yf_TILED) { - uint32_t y_tile_minimum = plane_blocks_per_line * 4; + uint32_t min_scanlines = 4; + uint32_t y_tile_minimum; + if (intel_rotation_90_or_270(p_params->rotation)) { + switch (p_params->bytes_per_pixel) { + case 1: + min_scanlines = 16; + break; + case 2: + min_scanlines = 8; + break; + case 8: + WARN(1, "Unsupported pixel depth for rotation"); + }; + } + y_tile_minimum = plane_blocks_per_line * min_scanlines; selected_result = max(method2, y_tile_minimum); } else { if ((ddb_allocation / plane_blocks_per_line) >= 1) @@ -3357,6 +3372,7 @@ skl_update_sprite_wm(struct drm_plane *plane, struct drm_crtc *crtc, */ if (fb) intel_plane->wm.tiling = fb->modifier[0]; + intel_plane->wm.rotation = plane->state->rotation; skl_update_wm(crtc); } diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c index 005a6fd35f6f..f41e872ad858 100644 --- a/drivers/gpu/drm/i915/intel_sprite.c +++ b/drivers/gpu/drm/i915/intel_sprite.c @@ -1039,10 +1039,7 @@ finish: if (!intel_crtc->primary_enabled && !state->hides_primary) intel_crtc->atomic.post_enable_primary = true; - /* Update watermarks on tiling changes. */ - if (!plane->state->fb || !state->base.fb || - plane->state->fb->modifier[0] != - state->base.fb->modifier[0]) + if (intel_wm_need_update(plane, &state->base)) intel_crtc->atomic.update_wm = true; if (!state->visible) { -- cgit v1.2.1 From e6e96d73a2aaaa54ed2c0f98693f4bf572712f1c Mon Sep 17 00:00:00 2001 From: Keith Busch Date: Mon, 23 Mar 2015 09:32:37 -0600 Subject: NVMe: Initialize device list head before starting Driver recovery requires the device's list node to have been initialized. Fixes: https://lkml.org/lkml/2015/3/22/262 Reported-by: Steven Noonan Signed-off-by: Keith Busch Cc: Matthew Wilcox Cc: Jens Axboe Signed-off-by: Jens Axboe --- drivers/block/nvme-core.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/block/nvme-core.c b/drivers/block/nvme-core.c index ceb32dd52a6c..e23be20a3417 100644 --- a/drivers/block/nvme-core.c +++ b/drivers/block/nvme-core.c @@ -3003,6 +3003,7 @@ static int nvme_probe(struct pci_dev *pdev, const struct pci_device_id *id) } get_device(dev->device); + INIT_LIST_HEAD(&dev->node); INIT_WORK(&dev->probe_work, nvme_async_probe); schedule_work(&dev->probe_work); return 0; -- cgit v1.2.1 From c72efb658f7c8b27ca3d0efb5cfd5ded9fcac89e Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Mon, 23 Mar 2015 00:18:48 -0400 Subject: writeback: fix possible underflow in write bandwidth calculation From 1ebf33901ecc75d9496862dceb1ef0377980587c Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Mon, 23 Mar 2015 00:08:19 -0400 2f800fbd777b ("writeback: fix dirtied pages accounting on redirty") introduced account_page_redirty() which reverts stat updates for a redirtied page, making BDI_DIRTIED no longer monotonically increasing. bdi_update_write_bandwidth() uses the delta in BDI_DIRTIED as the basis for bandwidth calculation. While unlikely, since the above patch, the newer value may be lower than the recorded past value and underflow the bandwidth calculation leading to a wild result. Fix it by subtracing min of the old and new values when calculating delta. AFAIK, there hasn't been any report of it happening but the resulting erratic behavior would be non-critical and temporary, so it's possible that the issue is happening without being reported. The risk of the fix is very low, so tagged for -stable. Signed-off-by: Tejun Heo Cc: Jens Axboe Cc: Jan Kara Cc: Wu Fengguang Cc: Greg Thelen Fixes: 2f800fbd777b ("writeback: fix dirtied pages accounting on redirty") Cc: stable@vger.kernel.org Signed-off-by: Jens Axboe --- mm/page-writeback.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/mm/page-writeback.c b/mm/page-writeback.c index b4fd980a93eb..644bcb665773 100644 --- a/mm/page-writeback.c +++ b/mm/page-writeback.c @@ -857,8 +857,11 @@ static void bdi_update_write_bandwidth(struct backing_dev_info *bdi, * bw * elapsed + write_bandwidth * (period - elapsed) * write_bandwidth = --------------------------------------------------- * period + * + * @written may have decreased due to account_page_redirty(). + * Avoid underflowing @bw calculation. */ - bw = written - bdi->written_stamp; + bw = written - min(written, bdi->written_stamp); bw *= HZ; if (unlikely(elapsed > period)) { do_div(bw, elapsed); -- cgit v1.2.1 From 2077cef4d5c29cf886192ec32066f783d6a80db8 Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Mon, 23 Mar 2015 09:22:10 -0700 Subject: sparc64: Fix several bugs in memmove(). Firstly, handle zero length calls properly. Believe it or not there are a few of these happening during early boot. Next, we can't just drop to a memcpy() call in the forward copy case where dst <= src. The reason is that the cache initializing stores used in the Niagara memcpy() implementations can end up clearing out cache lines before we've sourced their original contents completely. For example, considering NG4memcpy, the main unrolled loop begins like this: load src + 0x00 load src + 0x08 load src + 0x10 load src + 0x18 load src + 0x20 store dst + 0x00 Assume dst is 64 byte aligned and let's say that dst is src - 8 for this memcpy() call. That store at the end there is the one to the first line in the cache line, thus clearing the whole line, which thus clobbers "src + 0x28" before it even gets loaded. To avoid this, just fall through to a simple copy only mildly optimized for the case where src and dst are 8 byte aligned and the length is a multiple of 8 as well. We could get fancy and call GENmemcpy() but this is good enough for how this thing is actually used. Reported-by: David Ahern Reported-by: Bob Picco Signed-off-by: David S. Miller --- arch/sparc/lib/memmove.S | 35 ++++++++++++++++++++++++++++++++--- 1 file changed, 32 insertions(+), 3 deletions(-) diff --git a/arch/sparc/lib/memmove.S b/arch/sparc/lib/memmove.S index b7f6334e159f..857ad4f8905f 100644 --- a/arch/sparc/lib/memmove.S +++ b/arch/sparc/lib/memmove.S @@ -8,9 +8,11 @@ .text ENTRY(memmove) /* o0=dst o1=src o2=len */ - mov %o0, %g1 + brz,pn %o2, 99f + mov %o0, %g1 + cmp %o0, %o1 - bleu,pt %xcc, memcpy + bleu,pt %xcc, 2f add %o1, %o2, %g7 cmp %g7, %o0 bleu,pt %xcc, memcpy @@ -24,7 +26,34 @@ ENTRY(memmove) /* o0=dst o1=src o2=len */ stb %g7, [%o0] bne,pt %icc, 1b sub %o0, 1, %o0 - +99: retl mov %g1, %o0 + + /* We can't just call memcpy for these memmove cases. On some + * chips the memcpy uses cache initializing stores and when dst + * and src are close enough, those can clobber the source data + * before we've loaded it in. + */ +2: or %o0, %o1, %g7 + or %o2, %g7, %g7 + andcc %g7, 0x7, %g0 + bne,pn %xcc, 4f + nop + +3: ldx [%o1], %g7 + add %o1, 8, %o1 + subcc %o2, 8, %o2 + add %o0, 8, %o0 + bne,pt %icc, 3b + stx %g7, [%o0 - 0x8] + ba,a,pt %xcc, 99b + +4: ldub [%o1], %g7 + add %o1, 1, %o1 + subcc %o2, 1, %o2 + add %o0, 1, %o0 + bne,pt %icc, 4b + stb %g7, [%o0 - 0x1] + ba,a,pt %xcc, 99b ENDPROC(memmove) -- cgit v1.2.1 From 80b311d3118842eb681397233faa0d588df13f92 Mon Sep 17 00:00:00 2001 From: Sebastian Wicki Date: Mon, 23 Mar 2015 17:23:11 +0100 Subject: ALSA: hda - Add dock support for Thinkpad T450s (17aa:5036) This model uses the same dock port as the previous generation. Signed-off-by: Sebastian Wicki Cc: Signed-off-by: Takashi Iwai --- sound/pci/hda/patch_realtek.c | 1 + 1 file changed, 1 insertion(+) diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index 526398a4a442..7b0f72c5c6f1 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -5036,6 +5036,7 @@ static const struct snd_pci_quirk alc269_fixup_tbl[] = { SND_PCI_QUIRK(0x17aa, 0x501a, "Thinkpad", ALC283_FIXUP_INT_MIC), SND_PCI_QUIRK(0x17aa, 0x501e, "Thinkpad L440", ALC292_FIXUP_TPT440_DOCK), SND_PCI_QUIRK(0x17aa, 0x5026, "Thinkpad", ALC269_FIXUP_LIMIT_INT_MIC_BOOST), + SND_PCI_QUIRK(0x17aa, 0x5036, "Thinkpad T450s", ALC292_FIXUP_TPT440_DOCK), SND_PCI_QUIRK(0x17aa, 0x5109, "Thinkpad", ALC269_FIXUP_LIMIT_INT_MIC_BOOST), SND_PCI_QUIRK(0x17aa, 0x3bf8, "Quanta FL1", ALC269_FIXUP_PCM_44K), SND_PCI_QUIRK(0x17aa, 0x9e54, "LENOVO NB", ALC269_FIXUP_LENOVO_EAPD), -- cgit v1.2.1 From e53f21bce4d35a93b23d8fa1a840860f6c74f59e Mon Sep 17 00:00:00 2001 From: Catalin Marinas Date: Mon, 23 Mar 2015 15:06:50 +0000 Subject: arm64: Use the reserved TTBR0 if context switching to the init_mm The idle_task_exit() function may call switch_mm() with next == &init_mm. On arm64, init_mm.pgd cannot be used for user mappings, so this patch simply sets the reserved TTBR0. Cc: Reported-by: Jon Medhurst (Tixy) Tested-by: Jon Medhurst (Tixy) Signed-off-by: Catalin Marinas --- arch/arm64/include/asm/mmu_context.h | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/arch/arm64/include/asm/mmu_context.h b/arch/arm64/include/asm/mmu_context.h index a9eee33dfa62..101a42bde728 100644 --- a/arch/arm64/include/asm/mmu_context.h +++ b/arch/arm64/include/asm/mmu_context.h @@ -151,6 +151,15 @@ switch_mm(struct mm_struct *prev, struct mm_struct *next, { unsigned int cpu = smp_processor_id(); + /* + * init_mm.pgd does not contain any user mappings and it is always + * active for kernel addresses in TTBR1. Just set the reserved TTBR0. + */ + if (next == &init_mm) { + cpu_set_reserved_ttbr0(); + return; + } + if (!cpumask_test_and_set_cpu(cpu, mm_cpumask(next)) || prev != next) check_and_switch_context(next, tsk); } -- cgit v1.2.1 From 63a4f065ece613b6d575b538234375b0e9c23bbc Mon Sep 17 00:00:00 2001 From: Mike Snitzer Date: Mon, 23 Mar 2015 17:01:43 -0400 Subject: dm: fix add_disk() NULL pointer due to race with free_dev() Commit c4db59d31e39 ("fs: don't reassign dirty inodes to default_backing_dev_info") exposed DM to a latent race in free_dev() vs add_disk() in relation to management of the device's minor number. Fix this by refactoring free_dev() to match cleanup order of the alloc_dev() error path. Move cleanup of the gendisk, queue, and bdev to _before_ the cleanup of the idr managed minor number. Also, purely due to cleanup that fell out during the free_dev() audit: - adjust dm_blk_close() to access the gendisk's private_data under the _minor_lock spinlock. - move __dm_destroy()'s dm_get_live_table() call out from under the _minor_lock spinlock. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1202449 Reported-by: Zdenek Kabelac Reported-by: Jeff Moyer Signed-off-by: Mike Snitzer --- drivers/md/dm.c | 26 ++++++++++++++++---------- 1 file changed, 16 insertions(+), 10 deletions(-) diff --git a/drivers/md/dm.c b/drivers/md/dm.c index 9b641b38b857..8001fe9e3434 100644 --- a/drivers/md/dm.c +++ b/drivers/md/dm.c @@ -433,7 +433,6 @@ static int dm_blk_open(struct block_device *bdev, fmode_t mode) dm_get(md); atomic_inc(&md->open_count); - out: spin_unlock(&_minor_lock); @@ -442,16 +441,20 @@ out: static void dm_blk_close(struct gendisk *disk, fmode_t mode) { - struct mapped_device *md = disk->private_data; + struct mapped_device *md; spin_lock(&_minor_lock); + md = disk->private_data; + if (WARN_ON(!md)) + goto out; + if (atomic_dec_and_test(&md->open_count) && (test_bit(DMF_DEFERRED_REMOVE, &md->flags))) queue_work(deferred_remove_workqueue, &deferred_remove_work); dm_put(md); - +out: spin_unlock(&_minor_lock); } @@ -2241,7 +2244,6 @@ static void free_dev(struct mapped_device *md) int minor = MINOR(disk_devt(md->disk)); unlock_fs(md); - bdput(md->bdev); destroy_workqueue(md->wq); if (md->kworker_task) @@ -2252,19 +2254,22 @@ static void free_dev(struct mapped_device *md) mempool_destroy(md->rq_pool); if (md->bs) bioset_free(md->bs); - blk_integrity_unregister(md->disk); - del_gendisk(md->disk); + cleanup_srcu_struct(&md->io_barrier); free_table_devices(&md->table_devices); - free_minor(minor); + dm_stats_cleanup(&md->stats); spin_lock(&_minor_lock); md->disk->private_data = NULL; spin_unlock(&_minor_lock); - + if (blk_get_integrity(md->disk)) + blk_integrity_unregister(md->disk); + del_gendisk(md->disk); put_disk(md->disk); blk_cleanup_queue(md->queue); - dm_stats_cleanup(&md->stats); + bdput(md->bdev); + free_minor(minor); + module_put(THIS_MODULE); kfree(md); } @@ -2642,8 +2647,9 @@ static void __dm_destroy(struct mapped_device *md, bool wait) might_sleep(); - spin_lock(&_minor_lock); map = dm_get_live_table(md, &srcu_idx); + + spin_lock(&_minor_lock); idr_replace(&_minor_idr, MINOR_ALLOCED, MINOR(disk_devt(dm_disk(md)))); set_bit(DMF_FREEING, &md->flags); spin_unlock(&_minor_lock); -- cgit v1.2.1 From c806a6ad35bfa6c92249cd0ca4772d5ac3f8cb68 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Radim=20Kr=C4=8Dm=C3=A1=C5=99?= Date: Wed, 18 Mar 2015 19:38:22 +0100 Subject: KVM: x86: call irq notifiers with directed EOI MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit kvm_ioapic_update_eoi() wasn't called if directed EOI was enabled. We need to do that for irq notifiers. (Like with edge interrupts.) Fix it by skipping EOI broadcast only. Bug: https://bugzilla.kernel.org/show_bug.cgi?id=82211 Signed-off-by: Radim Krčmář Reviewed-by: Paolo Bonzini Tested-by: Bandan Das Signed-off-by: Marcelo Tosatti --- arch/x86/kvm/ioapic.c | 4 +++- arch/x86/kvm/lapic.c | 3 +-- 2 files changed, 4 insertions(+), 3 deletions(-) diff --git a/arch/x86/kvm/ioapic.c b/arch/x86/kvm/ioapic.c index b1947e0f3e10..46d4449772bc 100644 --- a/arch/x86/kvm/ioapic.c +++ b/arch/x86/kvm/ioapic.c @@ -422,6 +422,7 @@ static void __kvm_ioapic_update_eoi(struct kvm_vcpu *vcpu, struct kvm_ioapic *ioapic, int vector, int trigger_mode) { int i; + struct kvm_lapic *apic = vcpu->arch.apic; for (i = 0; i < IOAPIC_NUM_PINS; i++) { union kvm_ioapic_redirect_entry *ent = &ioapic->redirtbl[i]; @@ -443,7 +444,8 @@ static void __kvm_ioapic_update_eoi(struct kvm_vcpu *vcpu, kvm_notify_acked_irq(ioapic->kvm, KVM_IRQCHIP_IOAPIC, i); spin_lock(&ioapic->lock); - if (trigger_mode != IOAPIC_LEVEL_TRIG) + if (trigger_mode != IOAPIC_LEVEL_TRIG || + kvm_apic_get_reg(apic, APIC_SPIV) & APIC_SPIV_DIRECTED_EOI) continue; ASSERT(ent->fields.trig_mode == IOAPIC_LEVEL_TRIG); diff --git a/arch/x86/kvm/lapic.c b/arch/x86/kvm/lapic.c index bd4e34de24c7..4ee827d7bf36 100644 --- a/arch/x86/kvm/lapic.c +++ b/arch/x86/kvm/lapic.c @@ -833,8 +833,7 @@ int kvm_apic_compare_prio(struct kvm_vcpu *vcpu1, struct kvm_vcpu *vcpu2) static void kvm_ioapic_send_eoi(struct kvm_lapic *apic, int vector) { - if (!(kvm_apic_get_reg(apic, APIC_SPIV) & APIC_SPIV_DIRECTED_EOI) && - kvm_ioapic_handles_vector(apic->vcpu->kvm, vector)) { + if (kvm_ioapic_handles_vector(apic->vcpu->kvm, vector)) { int trigger_mode; if (apic_test_vector(vector, apic->regs + APIC_TMR)) trigger_mode = IOAPIC_LEVEL_TRIG; -- cgit v1.2.1 From 744961341d472db6272ed9b42319a90f5a2aa7c4 Mon Sep 17 00:00:00 2001 From: Igor Mammedov Date: Fri, 20 Mar 2015 12:21:37 +0000 Subject: kvm: avoid page allocation failure in kvm_set_memory_region() KVM guest can fail to startup with following trace on host: qemu-system-x86: page allocation failure: order:4, mode:0x40d0 Call Trace: dump_stack+0x47/0x67 warn_alloc_failed+0xee/0x150 __alloc_pages_direct_compact+0x14a/0x150 __alloc_pages_nodemask+0x776/0xb80 alloc_kmem_pages+0x3a/0x110 kmalloc_order+0x13/0x50 kmemdup+0x1b/0x40 __kvm_set_memory_region+0x24a/0x9f0 [kvm] kvm_set_ioapic+0x130/0x130 [kvm] kvm_set_memory_region+0x21/0x40 [kvm] kvm_vm_ioctl+0x43f/0x750 [kvm] Failure happens when attempting to allocate pages for 'struct kvm_memslots', however it doesn't have to be present in physically contiguous (kmalloc-ed) address space, change allocation to kvm_kvzalloc() so that it will be vmalloc-ed when its size is more then a page. Signed-off-by: Igor Mammedov Signed-off-by: Marcelo Tosatti --- virt/kvm/kvm_main.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index a2214d9609bd..cc6a25d95fbf 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -471,7 +471,7 @@ static struct kvm *kvm_create_vm(unsigned long type) BUILD_BUG_ON(KVM_MEM_SLOTS_NUM > SHRT_MAX); r = -ENOMEM; - kvm->memslots = kzalloc(sizeof(struct kvm_memslots), GFP_KERNEL); + kvm->memslots = kvm_kvzalloc(sizeof(struct kvm_memslots)); if (!kvm->memslots) goto out_err_no_srcu; @@ -522,7 +522,7 @@ out_err_no_srcu: out_err_no_disable: for (i = 0; i < KVM_NR_BUSES; i++) kfree(kvm->buses[i]); - kfree(kvm->memslots); + kvfree(kvm->memslots); kvm_arch_free_vm(kvm); return ERR_PTR(r); } @@ -578,7 +578,7 @@ static void kvm_free_physmem(struct kvm *kvm) kvm_for_each_memslot(memslot, slots) kvm_free_physmem_slot(kvm, memslot, NULL); - kfree(kvm->memslots); + kvfree(kvm->memslots); } static void kvm_destroy_devices(struct kvm *kvm) @@ -871,10 +871,10 @@ int __kvm_set_memory_region(struct kvm *kvm, goto out_free; } - slots = kmemdup(kvm->memslots, sizeof(struct kvm_memslots), - GFP_KERNEL); + slots = kvm_kvzalloc(sizeof(struct kvm_memslots)); if (!slots) goto out_free; + memcpy(slots, kvm->memslots, sizeof(struct kvm_memslots)); if ((change == KVM_MR_DELETE) || (change == KVM_MR_MOVE)) { slot = id_to_memslot(slots, mem->slot); @@ -917,7 +917,7 @@ int __kvm_set_memory_region(struct kvm *kvm, kvm_arch_commit_memory_region(kvm, mem, &old, change); kvm_free_physmem_slot(kvm, &old, &new); - kfree(old_memslots); + kvfree(old_memslots); /* * IOMMU mapping: New slots need to be mapped. Old slots need to be @@ -936,7 +936,7 @@ int __kvm_set_memory_region(struct kvm *kvm, return 0; out_slots: - kfree(slots); + kvfree(slots); out_free: kvm_free_physmem_slot(kvm, &new, &old); out: -- cgit v1.2.1 From 8218c3f4df3bb1c637c17552405039a6dd3c1ee1 Mon Sep 17 00:00:00 2001 From: Daniel Vetter Date: Fri, 27 Feb 2015 12:58:13 +0100 Subject: drm: Fixup racy refcounting in plane_force_disable Originally it was impossible to be dropping the last refcount in this function since there was always one around still from the idr. But in commit 83f45fc360c8e16a330474860ebda872d1384c8c Author: Daniel Vetter Date: Wed Aug 6 09:10:18 2014 +0200 drm: Don't grab an fb reference for the idr we've switched to weak references, broke that assumption but forgot to fix it up. Since we still force-disable planes it's only possible to hit this when racing multiple rmfb with fbdev restoring or similar evil things. As long as userspace is nice it's impossible to hit the BUG_ON. But the BUG_ON would most likely be hit from fbdev code, which usually invovles the console_lock besides all modeset locks. So very likely we'd never get the bug reports if this was hit in the wild, hence better be safe than sorry and backport. Spotted by Matt Roper while reviewing other patches. [airlied: pull this back into 4.0 - the oops happens there] Cc: stable@vger.kernel.org Cc: Matt Roper Reviewed-by: Matt Roper Signed-off-by: Daniel Vetter Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_crtc.c | 13 +------------ 1 file changed, 1 insertion(+), 12 deletions(-) diff --git a/drivers/gpu/drm/drm_crtc.c b/drivers/gpu/drm/drm_crtc.c index f6d04c7b5115..679b10e34fb5 100644 --- a/drivers/gpu/drm/drm_crtc.c +++ b/drivers/gpu/drm/drm_crtc.c @@ -525,17 +525,6 @@ void drm_framebuffer_reference(struct drm_framebuffer *fb) } EXPORT_SYMBOL(drm_framebuffer_reference); -static void drm_framebuffer_free_bug(struct kref *kref) -{ - BUG(); -} - -static void __drm_framebuffer_unreference(struct drm_framebuffer *fb) -{ - DRM_DEBUG("%p: FB ID: %d (%d)\n", fb, fb->base.id, atomic_read(&fb->refcount.refcount)); - kref_put(&fb->refcount, drm_framebuffer_free_bug); -} - /** * drm_framebuffer_unregister_private - unregister a private fb from the lookup idr * @fb: fb to unregister @@ -1320,7 +1309,7 @@ void drm_plane_force_disable(struct drm_plane *plane) return; } /* disconnect the plane from the fb and crtc: */ - __drm_framebuffer_unreference(plane->old_fb); + drm_framebuffer_unreference(plane->old_fb); plane->old_fb = NULL; plane->fb = NULL; plane->crtc = NULL; -- cgit v1.2.1 From 58072ccbb81c6f2d67c5b4cc7597707c4fb86a5e Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Mon, 23 Mar 2015 19:11:34 +0200 Subject: drm/i915: fix race when clearing RPS IIR bits When disabling RPS interrupts there is a race where we disable RPS inerrupts while the interrupt handler is running and the handler has already latched the pending RPS interrupt from the master IIR register. Afterwards the disabling path clears the PM IIR bits, making the state of pending interrupts inconsistent from the interrupt handler's point of view. This triggers the following warning: "The master control interrupt lied (PM)!". To fix this make sure that any running interrupt handler (which may have already latched the master IIR) finishes before clearing the IIR bits. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=87347 Signed-off-by: Imre Deak Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_irq.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c index 6d8340d5a111..7afbde4c13e9 100644 --- a/drivers/gpu/drm/i915/i915_irq.c +++ b/drivers/gpu/drm/i915/i915_irq.c @@ -330,6 +330,13 @@ void gen6_disable_rps_interrupts(struct drm_device *dev) __gen6_disable_pm_irq(dev_priv, dev_priv->pm_rps_events); I915_WRITE(gen6_pm_ier(dev_priv), I915_READ(gen6_pm_ier(dev_priv)) & ~dev_priv->pm_rps_events); + + spin_unlock_irq(&dev_priv->irq_lock); + + synchronize_irq(dev->irq); + + spin_lock_irq(&dev_priv->irq_lock); + I915_WRITE(gen6_pm_iir(dev_priv), dev_priv->pm_rps_events); I915_WRITE(gen6_pm_iir(dev_priv), dev_priv->pm_rps_events); -- cgit v1.2.1 From 096fad9ebfadb4922332f98895ea948e8d7fe6a7 Mon Sep 17 00:00:00 2001 From: Imre Deak Date: Mon, 23 Mar 2015 19:11:35 +0200 Subject: drm/i915: move clearing of RPS interrupt bits from disable to reset time The logical place for clearing the RPS latched interrupt bits is when resetting the RPS interrupts, so move the corresponding part from the RPS disable function to the reset function. During resetting we already cleared the IIR bits, so the only thing missing there was clearing pm_iir. Note that we call gen6_disable_rps_interrupts() also during driver load and resume time via intel_uncore_sanitize() when i915 interrupts are still not installed. If there are any pending RPS bits at this point (which after this patch wouldn't be cleared) they will be cleared by the reset code via the interrupt preinstall hooks. Signed-off-by: Imre Deak Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_irq.c | 10 +--------- 1 file changed, 1 insertion(+), 9 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c index 7afbde4c13e9..14ecb4d13a1a 100644 --- a/drivers/gpu/drm/i915/i915_irq.c +++ b/drivers/gpu/drm/i915/i915_irq.c @@ -277,6 +277,7 @@ void gen6_reset_rps_interrupts(struct drm_device *dev) I915_WRITE(reg, dev_priv->pm_rps_events); I915_WRITE(reg, dev_priv->pm_rps_events); POSTING_READ(reg); + dev_priv->rps.pm_iir = 0; spin_unlock_irq(&dev_priv->irq_lock); } @@ -334,15 +335,6 @@ void gen6_disable_rps_interrupts(struct drm_device *dev) spin_unlock_irq(&dev_priv->irq_lock); synchronize_irq(dev->irq); - - spin_lock_irq(&dev_priv->irq_lock); - - I915_WRITE(gen6_pm_iir(dev_priv), dev_priv->pm_rps_events); - I915_WRITE(gen6_pm_iir(dev_priv), dev_priv->pm_rps_events); - - dev_priv->rps.pm_iir = 0; - - spin_unlock_irq(&dev_priv->irq_lock); } /** -- cgit v1.2.1 From c2d885c6c90b0ada58355ee3f53c7581d2904603 Mon Sep 17 00:00:00 2001 From: Ramalingam C Date: Mon, 23 Mar 2015 23:12:02 +0530 Subject: drm/i915: Removing the drrs capability enum initialization As part of allocation of the drm_i915_private variable, drrs capability enum is initialized to DRRS_NOT_SUPPORTED. Hence need not initialize at each connector init. Moreover initializing this enum at connector init will reset the successful DRRS initialization of previous connector, as we have the DRRS support for only one panel at a time. Signed-off-by: Ramalingam C Reviewed-by: Sivakumar Thulasimani Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_dp.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 637dd5376c82..0b26df9c78cb 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -5337,8 +5337,6 @@ static bool intel_edp_init_connector(struct intel_dp *intel_dp, struct edid *edid; enum pipe pipe = INVALID_PIPE; - dev_priv->drrs.type = DRRS_NOT_SUPPORTED; - if (!is_edp(intel_dp)) return true; -- cgit v1.2.1 From f3eab7184ddcd4867cf42e3274ba24a66e1e093d Mon Sep 17 00:00:00 2001 From: Steve Capper Date: Sun, 22 Mar 2015 14:51:51 +0000 Subject: arm64: percpu: Make this_cpu accessors pre-empt safe this_cpu operations were implemented for arm64 in: 5284e1b arm64: xchg: Implement cmpxchg_double f97fc81 arm64: percpu: Implement this_cpu operations Unfortunately, it is possible for pre-emption to take place between address generation and data access. This can lead to cases where data is being manipulated by this_cpu for a different CPU than it was called on. Which effectively breaks the spec. This patch disables pre-emption for the this_cpu operations guaranteeing that address generation and data manipulation take place without a pre-emption in-between. Fixes: 5284e1b4bc8a ("arm64: xchg: Implement cmpxchg_double") Fixes: f97fc810798c ("arm64: percpu: Implement this_cpu operations") Reported-by: Mark Rutland Acked-by: Will Deacon Signed-off-by: Steve Capper [catalin.marinas@arm.com: remove space after type cast] Signed-off-by: Catalin Marinas --- arch/arm64/include/asm/cmpxchg.h | 32 +++++++++++++++++++++-------- arch/arm64/include/asm/percpu.h | 44 ++++++++++++++++++++++++++++++---------- 2 files changed, 57 insertions(+), 19 deletions(-) diff --git a/arch/arm64/include/asm/cmpxchg.h b/arch/arm64/include/asm/cmpxchg.h index cb9593079f29..d8c25b7b18fb 100644 --- a/arch/arm64/include/asm/cmpxchg.h +++ b/arch/arm64/include/asm/cmpxchg.h @@ -246,14 +246,30 @@ static inline unsigned long __cmpxchg_mb(volatile void *ptr, unsigned long old, __ret; \ }) -#define this_cpu_cmpxchg_1(ptr, o, n) cmpxchg_local(raw_cpu_ptr(&(ptr)), o, n) -#define this_cpu_cmpxchg_2(ptr, o, n) cmpxchg_local(raw_cpu_ptr(&(ptr)), o, n) -#define this_cpu_cmpxchg_4(ptr, o, n) cmpxchg_local(raw_cpu_ptr(&(ptr)), o, n) -#define this_cpu_cmpxchg_8(ptr, o, n) cmpxchg_local(raw_cpu_ptr(&(ptr)), o, n) - -#define this_cpu_cmpxchg_double_8(ptr1, ptr2, o1, o2, n1, n2) \ - cmpxchg_double_local(raw_cpu_ptr(&(ptr1)), raw_cpu_ptr(&(ptr2)), \ - o1, o2, n1, n2) +#define _protect_cmpxchg_local(pcp, o, n) \ +({ \ + typeof(*raw_cpu_ptr(&(pcp))) __ret; \ + preempt_disable(); \ + __ret = cmpxchg_local(raw_cpu_ptr(&(pcp)), o, n); \ + preempt_enable(); \ + __ret; \ +}) + +#define this_cpu_cmpxchg_1(ptr, o, n) _protect_cmpxchg_local(ptr, o, n) +#define this_cpu_cmpxchg_2(ptr, o, n) _protect_cmpxchg_local(ptr, o, n) +#define this_cpu_cmpxchg_4(ptr, o, n) _protect_cmpxchg_local(ptr, o, n) +#define this_cpu_cmpxchg_8(ptr, o, n) _protect_cmpxchg_local(ptr, o, n) + +#define this_cpu_cmpxchg_double_8(ptr1, ptr2, o1, o2, n1, n2) \ +({ \ + int __ret; \ + preempt_disable(); \ + __ret = cmpxchg_double_local( raw_cpu_ptr(&(ptr1)), \ + raw_cpu_ptr(&(ptr2)), \ + o1, o2, n1, n2); \ + preempt_enable(); \ + __ret; \ +}) #define cmpxchg64(ptr,o,n) cmpxchg((ptr),(o),(n)) #define cmpxchg64_local(ptr,o,n) cmpxchg_local((ptr),(o),(n)) diff --git a/arch/arm64/include/asm/percpu.h b/arch/arm64/include/asm/percpu.h index 09da25bc596f..4fde8c1df97f 100644 --- a/arch/arm64/include/asm/percpu.h +++ b/arch/arm64/include/asm/percpu.h @@ -204,25 +204,47 @@ static inline unsigned long __percpu_xchg(void *ptr, unsigned long val, return ret; } +#define _percpu_read(pcp) \ +({ \ + typeof(pcp) __retval; \ + preempt_disable(); \ + __retval = (typeof(pcp))__percpu_read(raw_cpu_ptr(&(pcp)), \ + sizeof(pcp)); \ + preempt_enable(); \ + __retval; \ +}) + +#define _percpu_write(pcp, val) \ +do { \ + preempt_disable(); \ + __percpu_write(raw_cpu_ptr(&(pcp)), (unsigned long)(val), \ + sizeof(pcp)); \ + preempt_enable(); \ +} while(0) \ + +#define _pcp_protect(operation, pcp, val) \ +({ \ + typeof(pcp) __retval; \ + preempt_disable(); \ + __retval = (typeof(pcp))operation(raw_cpu_ptr(&(pcp)), \ + (val), sizeof(pcp)); \ + preempt_enable(); \ + __retval; \ +}) + #define _percpu_add(pcp, val) \ - __percpu_add(raw_cpu_ptr(&(pcp)), val, sizeof(pcp)) + _pcp_protect(__percpu_add, pcp, val) -#define _percpu_add_return(pcp, val) (typeof(pcp)) (_percpu_add(pcp, val)) +#define _percpu_add_return(pcp, val) _percpu_add(pcp, val) #define _percpu_and(pcp, val) \ - __percpu_and(raw_cpu_ptr(&(pcp)), val, sizeof(pcp)) + _pcp_protect(__percpu_and, pcp, val) #define _percpu_or(pcp, val) \ - __percpu_or(raw_cpu_ptr(&(pcp)), val, sizeof(pcp)) - -#define _percpu_read(pcp) (typeof(pcp)) \ - (__percpu_read(raw_cpu_ptr(&(pcp)), sizeof(pcp))) - -#define _percpu_write(pcp, val) \ - __percpu_write(raw_cpu_ptr(&(pcp)), (unsigned long)(val), sizeof(pcp)) + _pcp_protect(__percpu_or, pcp, val) #define _percpu_xchg(pcp, val) (typeof(pcp)) \ - (__percpu_xchg(raw_cpu_ptr(&(pcp)), (unsigned long)(val), sizeof(pcp))) + _pcp_protect(__percpu_xchg, pcp, (unsigned long)(val)) #define this_cpu_add_1(pcp, val) _percpu_add(pcp, val) #define this_cpu_add_2(pcp, val) _percpu_add(pcp, val) -- cgit v1.2.1 From b3494a4ab20f6bdf74cdf2badf7918bb65ee8a00 Mon Sep 17 00:00:00 2001 From: Andy Lutomirski Date: Mon, 23 Mar 2015 12:32:54 -0700 Subject: x86/asm/entry: Check for syscall exit work with IRQs disabled We currently have a race: if we're preempted during syscall exit, we can fail to process syscall return work that is queued up while we're preempted in ret_from_sys_call after checking ti.flags. Fix it by disabling interrupts before checking ti.flags. Reported-by: Stefan Seyfried Reported-by: Takashi Iwai Signed-off-by: Andy Lutomirski Acked-by: Denys Vlasenko Cc: Jiri Kosina Cc: Tejun Heo Fixes: 96b6352c1271 ("x86_64, entry: Remove the syscall exit audit") Link: http://lkml.kernel.org/r/189320d42b4d671df78c10555976bb10af1ffc75.1427137498.git.luto@kernel.org Signed-off-by: Ingo Molnar --- arch/x86/kernel/entry_64.S | 18 ++++++++++++++---- 1 file changed, 14 insertions(+), 4 deletions(-) diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S index 1d74d161687c..2babb393915e 100644 --- a/arch/x86/kernel/entry_64.S +++ b/arch/x86/kernel/entry_64.S @@ -364,12 +364,21 @@ system_call_fastpath: * Has incomplete stack frame and undefined top of stack. */ ret_from_sys_call: - testl $_TIF_ALLWORK_MASK,TI_flags+THREAD_INFO(%rsp,RIP-ARGOFFSET) - jnz int_ret_from_sys_call_fixup /* Go the the slow path */ - LOCKDEP_SYS_EXIT DISABLE_INTERRUPTS(CLBR_NONE) TRACE_IRQS_OFF + + /* + * We must check ti flags with interrupts (or at least preemption) + * off because we must *never* return to userspace without + * processing exit work that is enqueued if we're preempted here. + * In particular, returning to userspace with any of the one-shot + * flags (TIF_NOTIFY_RESUME, TIF_USER_RETURN_NOTIFY, etc) set is + * very bad. + */ + testl $_TIF_ALLWORK_MASK,TI_flags+THREAD_INFO(%rsp,RIP-ARGOFFSET) + jnz int_ret_from_sys_call_fixup /* Go the the slow path */ + CFI_REMEMBER_STATE /* * sysretq will re-enable interrupts: @@ -386,7 +395,7 @@ ret_from_sys_call: int_ret_from_sys_call_fixup: FIXUP_TOP_OF_STACK %r11, -ARGOFFSET - jmp int_ret_from_sys_call + jmp int_ret_from_sys_call_irqs_off /* Do syscall tracing */ tracesys: @@ -432,6 +441,7 @@ tracesys_phase2: GLOBAL(int_ret_from_sys_call) DISABLE_INTERRUPTS(CLBR_NONE) TRACE_IRQS_OFF +int_ret_from_sys_call_irqs_off: movl $_TIF_ALLWORK_MASK,%edi /* edi: mask to check */ GLOBAL(int_with_check) -- cgit v1.2.1 From 977104e5606a6df8fe22c0dacd3620fc00b58d61 Mon Sep 17 00:00:00 2001 From: Chen-Yu Tsai Date: Wed, 25 Mar 2015 00:53:26 +0800 Subject: ARM: dts: sun4i: a10-lime: Override and remove 1008MHz OPP setting The Olimex A10-Lime is known to be unstable when running at 1008MHz. Signed-off-by: Chen-Yu Tsai Signed-off-by: Maxime Ripard --- arch/arm/boot/dts/sun4i-a10-olinuxino-lime.dts | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) diff --git a/arch/arm/boot/dts/sun4i-a10-olinuxino-lime.dts b/arch/arm/boot/dts/sun4i-a10-olinuxino-lime.dts index ab7891c43231..75742f8f96f3 100644 --- a/arch/arm/boot/dts/sun4i-a10-olinuxino-lime.dts +++ b/arch/arm/boot/dts/sun4i-a10-olinuxino-lime.dts @@ -56,6 +56,22 @@ model = "Olimex A10-OLinuXino-LIME"; compatible = "olimex,a10-olinuxino-lime", "allwinner,sun4i-a10"; + cpus { + cpu0: cpu@0 { + /* + * The A10-Lime is known to be unstable + * when running at 1008 MHz + */ + operating-points = < + /* kHz uV */ + 912000 1350000 + 864000 1300000 + 624000 1250000 + >; + cooling-max-level = <2>; + }; + }; + soc@01c00000 { emac: ethernet@01c0b000 { pinctrl-names = "default"; -- cgit v1.2.1 From 370a9b5fb04a0d5cc7b7699c788616d6976f4476 Mon Sep 17 00:00:00 2001 From: Chen-Yu Tsai Date: Wed, 25 Mar 2015 00:53:27 +0800 Subject: ARM: dts: sunxi: Remove overclocked/overvoltaged OPP Without proper regulator support for individual boards, it is dangerous to have overclocked/overvoltaged OPPs in the list. Cpufreq will increase the frequency without the accompanying voltage increase, resulting in an unstable system. Remove them for now. We can revisit them with the new version of OPP bindings, which support boost settings and frequency ranges, among other things. Signed-off-by: Chen-Yu Tsai Signed-off-by: Maxime Ripard --- arch/arm/boot/dts/sun4i-a10.dtsi | 3 +-- arch/arm/boot/dts/sun5i-a13.dtsi | 3 +-- arch/arm/boot/dts/sun7i-a20.dtsi | 3 +-- 3 files changed, 3 insertions(+), 6 deletions(-) diff --git a/arch/arm/boot/dts/sun4i-a10.dtsi b/arch/arm/boot/dts/sun4i-a10.dtsi index 5c2925831f20..eebb7853e00b 100644 --- a/arch/arm/boot/dts/sun4i-a10.dtsi +++ b/arch/arm/boot/dts/sun4i-a10.dtsi @@ -75,7 +75,6 @@ clock-latency = <244144>; /* 8 32k periods */ operating-points = < /* kHz uV */ - 1056000 1500000 1008000 1400000 912000 1350000 864000 1300000 @@ -83,7 +82,7 @@ >; #cooling-cells = <2>; cooling-min-level = <0>; - cooling-max-level = <4>; + cooling-max-level = <3>; }; }; diff --git a/arch/arm/boot/dts/sun5i-a13.dtsi b/arch/arm/boot/dts/sun5i-a13.dtsi index f8818f1edbbe..883cb4873688 100644 --- a/arch/arm/boot/dts/sun5i-a13.dtsi +++ b/arch/arm/boot/dts/sun5i-a13.dtsi @@ -47,7 +47,6 @@ clock-latency = <244144>; /* 8 32k periods */ operating-points = < /* kHz uV */ - 1104000 1500000 1008000 1400000 912000 1350000 864000 1300000 @@ -57,7 +56,7 @@ >; #cooling-cells = <2>; cooling-min-level = <0>; - cooling-max-level = <6>; + cooling-max-level = <5>; }; }; diff --git a/arch/arm/boot/dts/sun7i-a20.dtsi b/arch/arm/boot/dts/sun7i-a20.dtsi index 3a8530b79f1c..fdd181792b4b 100644 --- a/arch/arm/boot/dts/sun7i-a20.dtsi +++ b/arch/arm/boot/dts/sun7i-a20.dtsi @@ -105,7 +105,6 @@ clock-latency = <244144>; /* 8 32k periods */ operating-points = < /* kHz uV */ - 1008000 1450000 960000 1400000 912000 1400000 864000 1300000 @@ -116,7 +115,7 @@ >; #cooling-cells = <2>; cooling-min-level = <0>; - cooling-max-level = <7>; + cooling-max-level = <6>; }; cpu@1 { -- cgit v1.2.1 From 59a58cb34d3fe73e6c899cc5d9a87428ca662925 Mon Sep 17 00:00:00 2001 From: Damien Lespiau Date: Thu, 5 Feb 2015 18:30:20 +0000 Subject: drm/i915: Don't try to reference the fb in get_initial_plane_config() Tvrtko noticed a new warning on boot: WARNING: CPU: 1 PID: 353 at include/linux/kref.h:47 drm_framebuffer_reference+0x6c/0x80 [drm]() Call Trace: [] dump_stack+0x4f/0x7b [] warn_slowpath_common+0xaa/0xd0 [] warn_slowpath_null+0x1a/0x20 [] drm_framebuffer_reference+0x6c/0x80 [drm] [] update_state_fb.isra.54+0x47/0x50 [i915] [] skylake_get_initial_plane_config+0x93c/0x950 [i915] [] intel_modeset_init+0x1551/0x17c0 [i915] [] i915_driver_load+0xed0/0x11e0 [i915] [] ? _raw_spin_unlock_irqrestore+0x51/0x70 [] drm_dev_register+0x77/0x110 [drm] [] drm_get_pci_dev+0x11b/0x1f0 [drm] [] ? trace_hardirqs_on+0xd/0x10 [] ? _raw_spin_unlock_irqrestore+0x51/0x70 [] i915_pci_probe+0x56/0x60 [i915] [] pci_device_probe+0x7c/0x100 [] driver_probe_device+0x16d/0x380 We cannot take a reference at this point, not before intel_framebuffer_init() and the underlying drm_framebuffer_init(). Introduced in: commit 706dc7b549175e47f23e913b7f1e52874a7d0f56 Author: Matt Roper Date: Tue Feb 3 13:10:04 2015 -0800 drm/i915: Ensure plane->state->fb stays in sync with plane->fb v2: Don't move update_state_fb(). It was moved around because I originally put update_state_fb() in intel_alloc_plane_obj() before finding a better place. (Matt) Reviewed-by: Matt Roper Reported-by: Tvrtko Ursulin Cc: Matt Roper Cc: Tvrtko Ursulin Signed-off-by: Damien Lespiau Signed-off-by: Daniel Vetter From drm-next: (cherry picked from commit f55548b5af87ebfc586ca75748947f1c1b1a4a52) Signed-off-by: Dave Airlie --- drivers/gpu/drm/i915/intel_display.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 6d22128d97b1..1c12262029fb 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2438,8 +2438,10 @@ intel_find_plane_obj(struct intel_crtc *intel_crtc, if (!intel_crtc->base.primary->fb) return; - if (intel_alloc_plane_obj(intel_crtc, plane_config)) + if (intel_alloc_plane_obj(intel_crtc, plane_config)) { + update_state_fb(intel_crtc->base.primary); return; + } kfree(intel_crtc->base.primary->fb); intel_crtc->base.primary->fb = NULL; @@ -6663,7 +6665,6 @@ i9xx_get_initial_plane_config(struct intel_crtc *crtc, plane_config->size); crtc->base.primary->fb = fb; - update_state_fb(crtc->base.primary); } static void chv_crtc_clock_get(struct intel_crtc *crtc, @@ -7704,7 +7705,6 @@ skylake_get_initial_plane_config(struct intel_crtc *crtc, plane_config->size); crtc->base.primary->fb = fb; - update_state_fb(crtc->base.primary); return; error: @@ -7798,7 +7798,6 @@ ironlake_get_initial_plane_config(struct intel_crtc *crtc, plane_config->size); crtc->base.primary->fb = fb; - update_state_fb(crtc->base.primary); } static bool ironlake_get_pipe_config(struct intel_crtc *crtc, -- cgit v1.2.1 From 1833c9f647e9bda1cd24653ff8f9c207b5f5b911 Mon Sep 17 00:00:00 2001 From: Heiko Carstens Date: Sat, 21 Mar 2015 12:43:08 +0100 Subject: s390/smp: reenable smt after resume After a suspend/resume cycle we missed to enable smt again, which leads to all sorts of bugs, since the kernel assumes smt is enabled, while the hardware thinks it is not. Reported-and-tested-by: Sebastian Ott Reported-by: Stefan Haberland Signed-off-by: Heiko Carstens Signed-off-by: Martin Schwidefsky --- arch/s390/kernel/swsusp_asm64.S | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/arch/s390/kernel/swsusp_asm64.S b/arch/s390/kernel/swsusp_asm64.S index 6b09fdffbd2f..ca6294645dd3 100644 --- a/arch/s390/kernel/swsusp_asm64.S +++ b/arch/s390/kernel/swsusp_asm64.S @@ -177,6 +177,17 @@ restart_entry: lhi %r1,1 sigp %r1,%r0,SIGP_SET_ARCHITECTURE sam64 +#ifdef CONFIG_SMP + larl %r1,smp_cpu_mt_shift + icm %r1,15,0(%r1) + jz smt_done + llgfr %r1,%r1 +smt_loop: + sigp %r1,%r0,SIGP_SET_MULTI_THREADING + brc 8,smt_done /* accepted */ + brc 2,smt_loop /* busy, try again */ +smt_done: +#endif larl %r1,.Lnew_pgm_check_psw lpswe 0(%r1) pgm_check_entry: -- cgit v1.2.1 From c0f404284192f2d4a0159a714372a8c8610c1f6d Mon Sep 17 00:00:00 2001 From: Jesse Barnes Date: Mon, 23 Mar 2015 12:43:50 -0700 Subject: drm/i915: make unsupported fb modifier message DRM_DEBUG Or users can just spam the log all they want. Issue introduced in commit 9a8f0a1290993c86c4e35756a2624bfe461f9036 Author: Tvrtko Ursulin Date: Fri Feb 27 11:15:24 2015 +0000 drm/i915/skl: Allow Y (and Yf) frame buffer creation References: https://bugs.freedesktop.org/show_bug.cgi?id=89628 Signed-off-by: Jesse Barnes Reviewed-by: Jani Nikula Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_display.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 35cdb487814a..f9dc5babcf27 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -13006,8 +13006,8 @@ static int intel_framebuffer_init(struct drm_device *dev, case I915_FORMAT_MOD_X_TILED: break; default: - DRM_ERROR("Unsupported fb modifier 0x%llx!\n", - mode_cmd->modifier[0]); + DRM_DEBUG("Unsupported fb modifier 0x%llx!\n", + mode_cmd->modifier[0]); return -EINVAL; } -- cgit v1.2.1 From 1d00dad56b15ed5dab5802143df2bf61d51b6b55 Mon Sep 17 00:00:00 2001 From: Tvrtko Ursulin Date: Wed, 25 Mar 2015 10:15:26 +0000 Subject: drm/i915/skl: Fix up positive error code It should have been negative since it is returned with ERR_PTR(). Introduced in new code commit: commit 50470bb011c4be278097670bea92462f4e8c8945 Author: Tvrtko Ursulin Date: Mon Mar 23 11:10:36 2015 +0000 drm/i915/skl: Support secondary (rotated) frame buffer mapping Signed-off-by: Tvrtko Ursulin Reported-by: Dan Carpenter Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_gem_gtt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c index fc56c112a7de..9903bb0097a6 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c @@ -2540,7 +2540,7 @@ intel_rotate_fb_obj_pages(struct i915_ggtt_view *ggtt_view, struct sg_table *st; unsigned int tile_pitch, tile_height; unsigned int width_pages, height_pages; - int ret = ENOMEM; + int ret = -ENOMEM; pages = obj->base.size / PAGE_SIZE; -- cgit v1.2.1 From 6c826f349587f6c897da9bd224912ca1aee3d9ea Mon Sep 17 00:00:00 2001 From: Mika Kuoppala Date: Tue, 24 Mar 2015 14:54:19 +0200 Subject: drm/i915: Add fault address to error state for gen8 and gen9 The faulting virtual address is >32bits and has been moved to different registers. Add to error state and output upper register first, in the same line for easy reconstruction of the fault address. v2: correct gen masking (Michel) v3: s/TBL/TLB (Ville) Signed-off-by: Mika Kuoppala Reviewed-by: Michel Thierry Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_drv.h | 2 ++ drivers/gpu/drm/i915/i915_gpu_error.c | 10 ++++++++++ drivers/gpu/drm/i915/i915_reg.h | 3 +++ 3 files changed, 15 insertions(+) diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index e7ed5b3d133f..0a563cc65801 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -427,6 +427,8 @@ struct drm_i915_error_state { u32 forcewake; u32 error; /* gen6+ */ u32 err_int; /* gen7 */ + u32 fault_data0; /* gen8, gen9 */ + u32 fault_data1; /* gen8, gen9 */ u32 done_reg; u32 gac_eco; u32 gam_ecochk; diff --git a/drivers/gpu/drm/i915/i915_gpu_error.c b/drivers/gpu/drm/i915/i915_gpu_error.c index 2f7cbd3d5524..1d4e60df8883 100644 --- a/drivers/gpu/drm/i915/i915_gpu_error.c +++ b/drivers/gpu/drm/i915/i915_gpu_error.c @@ -386,6 +386,11 @@ int i915_error_state_to_str(struct drm_i915_error_state_buf *m, if (INTEL_INFO(dev)->gen >= 6) { err_printf(m, "ERROR: 0x%08x\n", error->error); + + if (INTEL_INFO(dev)->gen >= 8) + err_printf(m, "FAULT_TLB_DATA: 0x%08x 0x%08x\n", + error->fault_data1, error->fault_data0); + err_printf(m, "DONE_REG: 0x%08x\n", error->done_reg); } @@ -1171,6 +1176,11 @@ static void i915_capture_reg_state(struct drm_i915_private *dev_priv, if (IS_GEN7(dev)) error->err_int = I915_READ(GEN7_ERR_INT); + if (INTEL_INFO(dev)->gen >= 8) { + error->fault_data0 = I915_READ(GEN8_FAULT_TLB_DATA0); + error->fault_data1 = I915_READ(GEN8_FAULT_TLB_DATA1); + } + if (IS_GEN6(dev)) { error->forcewake = I915_READ(FORCEWAKE); error->gab_ctl = I915_READ(GAB_CTL); diff --git a/drivers/gpu/drm/i915/i915_reg.h b/drivers/gpu/drm/i915/i915_reg.h index 5b84ee686f99..b522eb6e59a4 100644 --- a/drivers/gpu/drm/i915/i915_reg.h +++ b/drivers/gpu/drm/i915/i915_reg.h @@ -1306,6 +1306,9 @@ enum skl_disp_power_wells { #define ERR_INT_FIFO_UNDERRUN_A (1<<0) #define ERR_INT_FIFO_UNDERRUN(pipe) (1<<(pipe*3)) +#define GEN8_FAULT_TLB_DATA0 0x04b10 +#define GEN8_FAULT_TLB_DATA1 0x04b14 + #define FPGA_DBG 0x42300 #define FPGA_DBG_RM_NOCLAIM (1<<31) -- cgit v1.2.1 From 6436a123a147db51a0b06024a8350f4c230e73ff Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Mon, 23 Mar 2015 18:01:35 -0700 Subject: selinux: fix sel_write_enforce broken return value Return a negative error value like the rest of the entries in this function. Cc: Signed-off-by: Joe Perches Acked-by: Stephen Smalley [PM: tweaked subject line] Signed-off-by: Paul Moore --- security/selinux/selinuxfs.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/security/selinux/selinuxfs.c b/security/selinux/selinuxfs.c index 33db1ad4fd10..138949a31eab 100644 --- a/security/selinux/selinuxfs.c +++ b/security/selinux/selinuxfs.c @@ -152,7 +152,7 @@ static ssize_t sel_write_enforce(struct file *file, const char __user *buf, goto out; /* No partial writes. */ - length = EINVAL; + length = -EINVAL; if (*ppos != 0) goto out; -- cgit v1.2.1 From fb903811c4e3bcd93ccc247235834e2632a3508d Mon Sep 17 00:00:00 2001 From: Ed Cashin Date: Wed, 25 Mar 2015 15:55:06 -0700 Subject: aoe: update aoe maintainer information The coraid.com email address is defunct. The old aoe support area hosted at coraid.com is no longer up. These changes update the email and website to current ones. Signed-off-by: Ed Cashin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- MAINTAINERS | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/MAINTAINERS b/MAINTAINERS index 358eb0105e00..c6cd0f60635c 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -1675,8 +1675,8 @@ F: drivers/misc/eeprom/at24.c F: include/linux/platform_data/at24.h ATA OVER ETHERNET (AOE) DRIVER -M: "Ed L. Cashin" -W: http://support.coraid.com/support/linux +M: "Ed L. Cashin" +W: http://www.openaoe.org/ S: Supported F: Documentation/aoe/ F: drivers/block/aoe/ -- cgit v1.2.1 From ddd2a30d41a5bc578ab094f6dbf080697ea1a7dd Mon Sep 17 00:00:00 2001 From: Lars-Peter Clausen Date: Wed, 25 Mar 2015 15:55:09 -0700 Subject: drivers/rtc/rtc-mrst: fix suspend/resume The Moorestown RTC driver implements suspend and resume callbacks and assigns them to the suspend and resume fields of the device_driver struct. These callbacks are never actually called by anything though. Modify the driver to properly use dev_pm_ops so that the suspend and resume functions are actually executed upon suspend/resume. [akpm@linux-foundation.org: device_driver.name is const char *] Signed-off-by: Lars-Peter Clausen Cc: Alessandro Zummo Cc: Feng Tang Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- drivers/rtc/rtc-mrst.c | 17 +++++++++-------- 1 file changed, 9 insertions(+), 8 deletions(-) diff --git a/drivers/rtc/rtc-mrst.c b/drivers/rtc/rtc-mrst.c index e2436d140175..3a6fd3a8a2ec 100644 --- a/drivers/rtc/rtc-mrst.c +++ b/drivers/rtc/rtc-mrst.c @@ -413,8 +413,8 @@ static void rtc_mrst_do_remove(struct device *dev) mrst->dev = NULL; } -#ifdef CONFIG_PM -static int mrst_suspend(struct device *dev, pm_message_t mesg) +#ifdef CONFIG_PM_SLEEP +static int mrst_suspend(struct device *dev) { struct mrst_rtc *mrst = dev_get_drvdata(dev); unsigned char tmp; @@ -453,7 +453,7 @@ static int mrst_suspend(struct device *dev, pm_message_t mesg) */ static inline int mrst_poweroff(struct device *dev) { - return mrst_suspend(dev, PMSG_HIBERNATE); + return mrst_suspend(dev); } static int mrst_resume(struct device *dev) @@ -490,9 +490,11 @@ static int mrst_resume(struct device *dev) return 0; } +static SIMPLE_DEV_PM_OPS(mrst_pm_ops, mrst_suspend, mrst_resume); +#define MRST_PM_OPS (&mrst_pm_ops) + #else -#define mrst_suspend NULL -#define mrst_resume NULL +#define MRST_PM_OPS NULL static inline int mrst_poweroff(struct device *dev) { @@ -529,9 +531,8 @@ static struct platform_driver vrtc_mrst_platform_driver = { .remove = vrtc_mrst_platform_remove, .shutdown = vrtc_mrst_platform_shutdown, .driver = { - .name = (char *) driver_name, - .suspend = mrst_suspend, - .resume = mrst_resume, + .name = driver_name, + .pm = MRST_PM_OPS, } }; -- cgit v1.2.1 From 3fe89b3e2a7bbf3e97657104b9b33a9d81b950b3 Mon Sep 17 00:00:00 2001 From: Leon Yu Date: Wed, 25 Mar 2015 15:55:11 -0700 Subject: mm: fix anon_vma->degree underflow in anon_vma endless growing prevention I have constantly stumbled upon "kernel BUG at mm/rmap.c:399!" after upgrading to 3.19 and had no luck with 4.0-rc1 neither. So, after looking into new logic introduced by commit 7a3ef208e662 ("mm: prevent endless growth of anon_vma hierarchy"), I found chances are that unlink_anon_vmas() is called without incrementing dst->anon_vma->degree in anon_vma_clone() due to allocation failure. If dst->anon_vma is not NULL in error path, its degree will be incorrectly decremented in unlink_anon_vmas() and eventually underflow when exiting as a result of another call to unlink_anon_vmas(). That's how "kernel BUG at mm/rmap.c:399!" is triggered for me. This patch fixes the underflow by dropping dst->anon_vma when allocation fails. It's safe to do so regardless of original value of dst->anon_vma because dst->anon_vma doesn't have valid meaning if anon_vma_clone() fails. Besides, callers don't care dst->anon_vma in such case neither. Also suggested by Michal Hocko, we can clean up vma_adjust() a bit as anon_vma_clone() now does the work. [akpm@linux-foundation.org: tweak comment] Fixes: 7a3ef208e662 ("mm: prevent endless growth of anon_vma hierarchy") Signed-off-by: Leon Yu Signed-off-by: Konstantin Khlebnikov Reviewed-by: Michal Hocko Acked-by: Rik van Riel Acked-by: David Rientjes Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/mmap.c | 4 +--- mm/rmap.c | 7 +++++++ 2 files changed, 8 insertions(+), 3 deletions(-) diff --git a/mm/mmap.c b/mm/mmap.c index da9990acc08b..9ec50a368634 100644 --- a/mm/mmap.c +++ b/mm/mmap.c @@ -774,10 +774,8 @@ again: remove_next = 1 + (end > next->vm_end); importer->anon_vma = exporter->anon_vma; error = anon_vma_clone(importer, exporter); - if (error) { - importer->anon_vma = NULL; + if (error) return error; - } } } diff --git a/mm/rmap.c b/mm/rmap.c index 5e3e09081164..c161a14b6a8f 100644 --- a/mm/rmap.c +++ b/mm/rmap.c @@ -287,6 +287,13 @@ int anon_vma_clone(struct vm_area_struct *dst, struct vm_area_struct *src) return 0; enomem_failure: + /* + * dst->anon_vma is dropped here otherwise its degree can be incorrectly + * decremented in unlink_anon_vmas(). + * We can safely do this because callers of anon_vma_clone() don't care + * about dst->anon_vma if anon_vma_clone() failed. + */ + dst->anon_vma = NULL; unlink_anon_vmas(dst); return -ENOMEM; } -- cgit v1.2.1 From f683739539e819e9b821a197d80e52258510837b Mon Sep 17 00:00:00 2001 From: Naoya Horiguchi Date: Wed, 25 Mar 2015 15:55:14 -0700 Subject: mm/pagewalk.c: prevent positive return value of walk_page_test() from being passed to callers walk_page_test() is purely pagewalk's internal stuff, and its positive return values are not intended to be passed to the callers of pagewalk. However, in the current code if the last vma in the do-while loop in walk_page_range() happens to return a positive value, it leaks outside walk_page_range(). So the user visible effect is invalid/unexpected return value (according to the reporter, mbind() causes it.) This patch fixes it simply by reinitializing the return value after checked. Another exposed interface, walk_page_vma(), already returns 0 for such cases so no problem. Fixes: fafaa4264eba ("pagewalk: improve vma handling") Signed-off-by: Naoya Horiguchi Signed-off-by: Kazutomo Yoshii Reported-by: Kazutomo Yoshii Acked-by: David Rientjes Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/pagewalk.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/mm/pagewalk.c b/mm/pagewalk.c index 75c1f2878519..29f2f8b853ae 100644 --- a/mm/pagewalk.c +++ b/mm/pagewalk.c @@ -265,8 +265,15 @@ int walk_page_range(unsigned long start, unsigned long end, vma = vma->vm_next; err = walk_page_test(start, next, walk); - if (err > 0) + if (err > 0) { + /* + * positive return values are purely for + * controlling the pagewalk, so should never + * be passed to the callers. + */ + err = 0; continue; + } if (err < 0) break; } -- cgit v1.2.1 From 59ec96719f7521002804c0862888971410e288af Mon Sep 17 00:00:00 2001 From: Joe Perches Date: Wed, 25 Mar 2015 15:55:17 -0700 Subject: MAINTAINERS: correct rtc armada38x pattern entry Commit c6a95dbee793 ("MAINTAINERS: add the RTC driver for the Armada38x") typoed the pattern, fix it. Signed-off-by: Joe Perches Acked-by: Gregory CLEMENT Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- MAINTAINERS | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index c6cd0f60635c..12ebef95224d 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -1186,7 +1186,7 @@ M: Sebastian Hesselbarth L: linux-arm-kernel@lists.infradead.org (moderated for non-subscribers) S: Maintained F: arch/arm/mach-mvebu/ -F: drivers/rtc/armada38x-rtc +F: drivers/rtc/rtc-armada38x.c ARM/Marvell Berlin SoC support M: Sebastian Hesselbarth -- cgit v1.2.1 From b0dc3a342af36f95a68fe229b8f0f73552c5ca08 Mon Sep 17 00:00:00 2001 From: Gu Zheng Date: Wed, 25 Mar 2015 15:55:20 -0700 Subject: mm/memory hotplug: postpone the reset of obsolete pgdat Qiu Xishi reported the following BUG when testing hot-add/hot-remove node under stress condition: BUG: unable to handle kernel paging request at 0000000000025f60 IP: next_online_pgdat+0x1/0x50 PGD 0 Oops: 0000 [#1] SMP ACPI: Device does not support D3cold Modules linked in: fuse nls_iso8859_1 nls_cp437 vfat fat loop dm_mod coretemp mperf crc32c_intel ghash_clmulni_intel aesni_intel ablk_helper cryptd lrw gf128mul glue_helper aes_x86_64 pcspkr microcode igb dca i2c_algo_bit ipv6 megaraid_sas iTCO_wdt i2c_i801 i2c_core iTCO_vendor_support tg3 sg hwmon ptp lpc_ich pps_core mfd_core acpi_pad rtc_cmos button ext3 jbd mbcache sd_mod crc_t10dif scsi_dh_alua scsi_dh_rdac scsi_dh_hp_sw scsi_dh_emc scsi_dh ahci libahci libata scsi_mod [last unloaded: rasf] CPU: 23 PID: 238 Comm: kworker/23:1 Tainted: G O 3.10.15-5885-euler0302 #1 Hardware name: HUAWEI TECHNOLOGIES CO.,LTD. Huawei N1/Huawei N1, BIOS V100R001 03/02/2015 Workqueue: events vmstat_update task: ffffa800d32c0000 ti: ffffa800d32ae000 task.ti: ffffa800d32ae000 RIP: 0010: next_online_pgdat+0x1/0x50 RSP: 0018:ffffa800d32afce8 EFLAGS: 00010286 RAX: 0000000000001440 RBX: ffffffff81da53b8 RCX: 0000000000000082 RDX: 0000000000000000 RSI: 0000000000000082 RDI: 0000000000000000 RBP: ffffa800d32afd28 R08: ffffffff81c93bfc R09: ffffffff81cbdc96 R10: 00000000000040ec R11: 00000000000000a0 R12: ffffa800fffb3440 R13: ffffa800d32afd38 R14: 0000000000000017 R15: ffffa800e6616800 FS: 0000000000000000(0000) GS:ffffa800e6600000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000000025f60 CR3: 0000000001a0b000 CR4: 00000000001407e0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: refresh_cpu_vm_stats+0xd0/0x140 vmstat_update+0x11/0x50 process_one_work+0x194/0x3d0 worker_thread+0x12b/0x410 kthread+0xc6/0xd0 ret_from_fork+0x7c/0xb0 The cause is the "memset(pgdat, 0, sizeof(*pgdat))" at the end of try_offline_node, which will reset all the content of pgdat to 0, as the pgdat is accessed lock-free, so that the users still using the pgdat will panic, such as the vmstat_update routine. process A: offline node XX: vmstat_updat() refresh_cpu_vm_stats() for_each_populated_zone() find online node XX cond_resched() offline cpu and memory, then try_offline_node() node_set_offline(nid), and memset(pgdat, 0, sizeof(*pgdat)) zone = next_zone(zone) pg_data_t *pgdat = zone->zone_pgdat; // here pgdat is NULL now next_online_pgdat(pgdat) next_online_node(pgdat->node_id); // NULL pointer access So the solution here is postponing the reset of obsolete pgdat from try_offline_node() to hotadd_new_pgdat(), and just resetting pgdat->nr_zones and pgdat->classzone_idx to be 0 rather than the memset 0 to avoid breaking pointer information in pgdat. Signed-off-by: Gu Zheng Reported-by: Xishi Qiu Suggested-by: KAMEZAWA Hiroyuki Cc: David Rientjes Cc: Yasuaki Ishimatsu Cc: Taku Izumi Cc: Tang Chen Cc: Xie XiuQi Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/memory_hotplug.c | 13 ++++--------- 1 file changed, 4 insertions(+), 9 deletions(-) diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c index 9fab10795bea..65842d688b7c 100644 --- a/mm/memory_hotplug.c +++ b/mm/memory_hotplug.c @@ -1092,6 +1092,10 @@ static pg_data_t __ref *hotadd_new_pgdat(int nid, u64 start) return NULL; arch_refresh_nodedata(nid, pgdat); + } else { + /* Reset the nr_zones and classzone_idx to 0 before reuse */ + pgdat->nr_zones = 0; + pgdat->classzone_idx = 0; } /* we can use NODE_DATA(nid) from here */ @@ -1977,15 +1981,6 @@ void try_offline_node(int nid) if (is_vmalloc_addr(zone->wait_table)) vfree(zone->wait_table); } - - /* - * Since there is no way to guarentee the address of pgdat/zone is not - * on stack of any kernel threads or used by other kernel objects - * without reference counting or other symchronizing method, do not - * reset node_data and free pgdat here. Just reset it to 0 and reuse - * the memory when the node is online again. - */ - memset(pgdat, 0, sizeof(*pgdat)); } EXPORT_SYMBOL(try_offline_node); -- cgit v1.2.1 From 859b7a0e89120505c304d7afbbe90325abaa0a6b Mon Sep 17 00:00:00 2001 From: Mark Rutland Date: Wed, 25 Mar 2015 15:55:23 -0700 Subject: mm/slub: fix lockups on PREEMPT && !SMP kernels Commit 9aabf810a67c ("mm/slub: optimize alloc/free fastpath by removing preemption on/off") introduced an occasional hang for kernels built with CONFIG_PREEMPT && !CONFIG_SMP. The problem is the following loop the patch introduced to slab_alloc_node and slab_free: do { tid = this_cpu_read(s->cpu_slab->tid); c = raw_cpu_ptr(s->cpu_slab); } while (IS_ENABLED(CONFIG_PREEMPT) && unlikely(tid != c->tid)); GCC 4.9 has been observed to hoist the load of c and c->tid above the loop for !SMP kernels (as in this case raw_cpu_ptr(x) is compile-time constant and does not force a reload). On arm64 the generated assembly looks like: ldr x4, [x0,#8] loop: ldr x1, [x0,#8] cmp x1, x4 b.ne loop If the thread is preempted between the load of c->tid (into x1) and tid (into x4), and an allocation or free occurs in another thread (bumping the cpu_slab's tid), the thread will be stuck in the loop until s->cpu_slab->tid wraps, which may be forever in the absence of allocations/frees on the same CPU. This patch changes the loop condition to access c->tid with READ_ONCE. This ensures that the value is reloaded even when the compiler would otherwise assume it could cache the value, and also ensures that the load will not be torn. Signed-off-by: Mark Rutland Cc: Catalin Marinas Acked-by: Christoph Lameter Cc: David Rientjes Cc: Jesper Dangaard Brouer Cc: Joonsoo Kim Cc: Pekka Enberg Cc: Steve Capper Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/slub.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/mm/slub.c b/mm/slub.c index 6832c4eab104..82c473780c91 100644 --- a/mm/slub.c +++ b/mm/slub.c @@ -2449,7 +2449,8 @@ redo: do { tid = this_cpu_read(s->cpu_slab->tid); c = raw_cpu_ptr(s->cpu_slab); - } while (IS_ENABLED(CONFIG_PREEMPT) && unlikely(tid != c->tid)); + } while (IS_ENABLED(CONFIG_PREEMPT) && + unlikely(tid != READ_ONCE(c->tid))); /* * Irqless object alloc/free algorithm used here depends on sequence @@ -2718,7 +2719,8 @@ redo: do { tid = this_cpu_read(s->cpu_slab->tid); c = raw_cpu_ptr(s->cpu_slab); - } while (IS_ENABLED(CONFIG_PREEMPT) && unlikely(tid != c->tid)); + } while (IS_ENABLED(CONFIG_PREEMPT) && + unlikely(tid != READ_ONCE(c->tid))); /* Same with comment on barrier() in slab_alloc_node() */ barrier(); -- cgit v1.2.1 From cfa869438282be84ad4110bba5027ef1fbbe71e4 Mon Sep 17 00:00:00 2001 From: Laura Abbott Date: Wed, 25 Mar 2015 15:55:26 -0700 Subject: mm/page_alloc.c: call kernel_map_pages in unset_migrateype_isolate Commit 3c605096d315 ("mm/page_alloc: restrict max order of merging on isolated pageblock") changed the logic of unset_migratetype_isolate to check the buddy allocator and explicitly call __free_pages to merge. The page that is being freed in this path never had prep_new_page called so set_page_refcounted is called explicitly but there is no call to kernel_map_pages. With the default kernel_map_pages this is mostly harmless but if kernel_map_pages does any manipulation of the page tables (unmapping or setting pages to read only) this may trigger a fault: alloc_contig_range test_pages_isolated(ceb00, ced00) failed Unable to handle kernel paging request at virtual address ffffffc0cec00000 pgd = ffffffc045fc4000 [ffffffc0cec00000] *pgd=0000000000000000 Internal error: Oops: 9600004f [#1] PREEMPT SMP Modules linked in: exfatfs CPU: 1 PID: 23237 Comm: TimedEventQueue Not tainted 3.10.49-gc72ad36-dirty #1 task: ffffffc03de52100 ti: ffffffc015388000 task.ti: ffffffc015388000 PC is at memset+0xc8/0x1c0 LR is at kernel_map_pages+0x1ec/0x244 Fix this by calling kernel_map_pages to ensure the page is set in the page table properly Fixes: 3c605096d315 ("mm/page_alloc: restrict max order of merging on isolated pageblock") Signed-off-by: Laura Abbott Cc: Naoya Horiguchi Cc: Mel Gorman Acked-by: Rik van Riel Cc: Yasuaki Ishimatsu Cc: Zhang Yanfei Cc: Xishi Qiu Cc: Vladimir Davydov Acked-by: Joonsoo Kim Cc: Gioh Kim Cc: Michal Nazarewicz Cc: Marek Szyprowski Cc: Vlastimil Babka Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/page_isolation.c | 1 + 1 file changed, 1 insertion(+) diff --git a/mm/page_isolation.c b/mm/page_isolation.c index 72f5ac381ab3..755a42c76eb4 100644 --- a/mm/page_isolation.c +++ b/mm/page_isolation.c @@ -103,6 +103,7 @@ void unset_migratetype_isolate(struct page *page, unsigned migratetype) if (!is_migrate_isolate_page(buddy)) { __isolate_free_page(page, order); + kernel_map_pages(page, (1 << order), 1); set_page_refcounted(page); isolated_page = page; } -- cgit v1.2.1 From 3d5d472cf55d1be091e8a145f80602bf435ece85 Mon Sep 17 00:00:00 2001 From: Taesoo Kim Date: Wed, 25 Mar 2015 15:55:29 -0700 Subject: fs/affs/file.c: unlock/release page on error When affs_bread_ino() fails, correctly unlock the page and release the page cache with proper error value. All write_end() should unlock/release the page that was locked by write_beg(). Signed-off-by: Taesoo Kim Cc: Fabian Frederick Cc: Al Viro Cc: Geert Uytterhoeven Cc: Jan Kara Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/affs/file.c | 19 ++++++++++++------- 1 file changed, 12 insertions(+), 7 deletions(-) diff --git a/fs/affs/file.c b/fs/affs/file.c index d2468bf95669..a91795e01a7f 100644 --- a/fs/affs/file.c +++ b/fs/affs/file.c @@ -699,8 +699,10 @@ static int affs_write_end_ofs(struct file *file, struct address_space *mapping, boff = tmp % bsize; if (boff) { bh = affs_bread_ino(inode, bidx, 0); - if (IS_ERR(bh)) - return PTR_ERR(bh); + if (IS_ERR(bh)) { + written = PTR_ERR(bh); + goto err_first_bh; + } tmp = min(bsize - boff, to - from); BUG_ON(boff + tmp > bsize || tmp > bsize); memcpy(AFFS_DATA(bh) + boff, data + from, tmp); @@ -712,14 +714,16 @@ static int affs_write_end_ofs(struct file *file, struct address_space *mapping, bidx++; } else if (bidx) { bh = affs_bread_ino(inode, bidx - 1, 0); - if (IS_ERR(bh)) - return PTR_ERR(bh); + if (IS_ERR(bh)) { + written = PTR_ERR(bh); + goto err_first_bh; + } } while (from + bsize <= to) { prev_bh = bh; bh = affs_getemptyblk_ino(inode, bidx); if (IS_ERR(bh)) - goto out; + goto err_bh; memcpy(AFFS_DATA(bh), data + from, bsize); if (buffer_new(bh)) { AFFS_DATA_HEAD(bh)->ptype = cpu_to_be32(T_DATA); @@ -751,7 +755,7 @@ static int affs_write_end_ofs(struct file *file, struct address_space *mapping, prev_bh = bh; bh = affs_bread_ino(inode, bidx, 1); if (IS_ERR(bh)) - goto out; + goto err_bh; tmp = min(bsize, to - from); BUG_ON(tmp > bsize); memcpy(AFFS_DATA(bh), data + from, tmp); @@ -790,12 +794,13 @@ done: if (tmp > inode->i_size) inode->i_size = AFFS_I(inode)->mmu_private = tmp; +err_first_bh: unlock_page(page); page_cache_release(page); return written; -out: +err_bh: bh = prev_bh; if (!written) written = PTR_ERR(bh); -- cgit v1.2.1 From 1f31e1b1963c240ced453489730bdfc9b0110ceb Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Wed, 25 Mar 2015 15:55:31 -0700 Subject: MAINTAINERS: add Jan as DMI/SMBIOS support maintainer I am familiar with these drivers and I care about them so let me add myself as their maintainer. Signed-off-by: Jean Delvare Acked-by: Matt Fleming Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- MAINTAINERS | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/MAINTAINERS b/MAINTAINERS index 12ebef95224d..88c09ca2584f 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -3252,6 +3252,13 @@ S: Maintained F: Documentation/hwmon/dme1737 F: drivers/hwmon/dme1737.c +DMI/SMBIOS SUPPORT +M: Jean Delvare +S: Maintained +F: drivers/firmware/dmi-id.c +F: drivers/firmware/dmi_scan.c +F: include/linux/dmi.h + DOCKING STATION DRIVER M: Shaohua Li L: linux-acpi@vger.kernel.org -- cgit v1.2.1 From 98cf21c61a7f5419d82f847c4d77bf6e96a76f5f Mon Sep 17 00:00:00 2001 From: Sergei Antonov Date: Wed, 25 Mar 2015 15:55:34 -0700 Subject: hfsplus: fix B-tree corruption after insertion at position 0 Fix B-tree corruption when a new record is inserted at position 0 in the node in hfs_brec_insert(). In this case a hfs_brec_update_parent() is called to update the parent index node (if exists) and it is passed hfs_find_data with a search_key containing a newly inserted key instead of the key to be updated. This results in an inconsistent index node. The bug reproduces on my machine after an extents overflow record for the catalog file (CNID=4) is inserted into the extents overflow B-tree. Because of a low (reserved) value of CNID=4, it has to become the first record in the first leaf node. The resulting first leaf node is correct: ---------------------------------------------------- | key0.CNID=4 | key1.CNID=123 | key2.CNID=456, ... | ---------------------------------------------------- But the parent index key0 still contains the previous key CNID=123: ----------------------- | key0.CNID=123 | ... | ----------------------- A change in hfs_brec_insert() makes hfs_brec_update_parent() work correctly by preventing it from getting fd->record=-1 value from __hfs_brec_find(). Along the way, I removed duplicate code with unification of the if condition. The resulting code is equivalent to the original code because node is never 0. Also hfs_brec_update_parent() will now return an error after getting a negative fd->record value. However, the return value of hfs_brec_update_parent() is not checked anywhere in the file and I'm leaving it unchanged by this patch. brec.c lacks error checking after some other calls too, but this issue is of less importance than the one being fixed by this patch. Signed-off-by: Sergei Antonov Cc: Joe Perches Reviewed-by: Vyacheslav Dubeyko Acked-by: Hin-Tak Leung Cc: Anton Altaparmakov Cc: Al Viro Cc: Christoph Hellwig Cc: Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/hfsplus/brec.c | 20 +++++++++++--------- 1 file changed, 11 insertions(+), 9 deletions(-) diff --git a/fs/hfsplus/brec.c b/fs/hfsplus/brec.c index 6e560d56094b..754fdf8c6356 100644 --- a/fs/hfsplus/brec.c +++ b/fs/hfsplus/brec.c @@ -131,13 +131,16 @@ skip: hfs_bnode_write(node, entry, data_off + key_len, entry_len); hfs_bnode_dump(node); - if (new_node) { - /* update parent key if we inserted a key - * at the start of the first node - */ - if (!rec && new_node != node) - hfs_brec_update_parent(fd); + /* + * update parent key if we inserted a key + * at the start of the node and it is not the new node + */ + if (!rec && new_node != node) { + hfs_bnode_read_key(node, fd->search_key, data_off + size); + hfs_brec_update_parent(fd); + } + if (new_node) { hfs_bnode_put(fd->bnode); if (!new_node->parent) { hfs_btree_inc_height(tree); @@ -168,9 +171,6 @@ skip: goto again; } - if (!rec) - hfs_brec_update_parent(fd); - return 0; } @@ -370,6 +370,8 @@ again: if (IS_ERR(parent)) return PTR_ERR(parent); __hfs_brec_find(parent, fd, hfs_find_rec_by_key); + if (fd->record < 0) + return -ENOENT; hfs_bnode_dump(parent); rec = fd->record; -- cgit v1.2.1 From bea66fbd11af1ca98ae26855eea41eda8582923e Mon Sep 17 00:00:00 2001 From: Mel Gorman Date: Wed, 25 Mar 2015 15:55:37 -0700 Subject: mm: numa: group related processes based on VMA flags instead of page table flags These are three follow-on patches based on the xfsrepair workload Dave Chinner reported was problematic in 4.0-rc1 due to changes in page table management -- https://lkml.org/lkml/2015/3/1/226. Much of the problem was reduced by commit 53da3bc2ba9e ("mm: fix up numa read-only thread grouping logic") and commit ba68bc0115eb ("mm: thp: Return the correct value for change_huge_pmd"). It was known that the performance in 3.19 was still better even if is far less safe. This series aims to restore the performance without compromising on safety. For the test of this mail, I'm comparing 3.19 against 4.0-rc4 and the three patches applied on top autonumabench 3.19.0 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 vanilla vanilla vmwrite-v5r8 preserve-v5r8 slowscan-v5r8 Time System-NUMA01 124.00 ( 0.00%) 161.86 (-30.53%) 107.13 ( 13.60%) 103.13 ( 16.83%) 145.01 (-16.94%) Time System-NUMA01_THEADLOCAL 115.54 ( 0.00%) 107.64 ( 6.84%) 131.87 (-14.13%) 83.30 ( 27.90%) 92.35 ( 20.07%) Time System-NUMA02 9.35 ( 0.00%) 10.44 (-11.66%) 8.95 ( 4.28%) 10.72 (-14.65%) 8.16 ( 12.73%) Time System-NUMA02_SMT 3.87 ( 0.00%) 4.63 (-19.64%) 4.57 (-18.09%) 3.99 ( -3.10%) 3.36 ( 13.18%) Time Elapsed-NUMA01 570.06 ( 0.00%) 567.82 ( 0.39%) 515.78 ( 9.52%) 517.26 ( 9.26%) 543.80 ( 4.61%) Time Elapsed-NUMA01_THEADLOCAL 393.69 ( 0.00%) 384.83 ( 2.25%) 384.10 ( 2.44%) 384.31 ( 2.38%) 380.73 ( 3.29%) Time Elapsed-NUMA02 49.09 ( 0.00%) 49.33 ( -0.49%) 48.86 ( 0.47%) 48.78 ( 0.63%) 50.94 ( -3.77%) Time Elapsed-NUMA02_SMT 47.51 ( 0.00%) 47.15 ( 0.76%) 47.98 ( -0.99%) 48.12 ( -1.28%) 49.56 ( -4.31%) 3.19.0 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 vanilla vanillavmwrite-v5r8preserve-v5r8slowscan-v5r8 User 46334.60 46391.94 44383.95 43971.89 44372.12 System 252.84 284.66 252.61 201.24 249.00 Elapsed 1062.14 1050.96 998.68 1000.94 1026.78 Overall the system CPU usage is comparable and the test is naturally a bit variable. The slowing of the scanner hurts numa01 but on this machine it is an adverse workload and patches that dramatically help it often hurt absolutely everything else. Due to patch 2, the fault activity is interesting 3.19.0 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 vanilla vanillavmwrite-v5r8preserve-v5r8slowscan-v5r8 Minor Faults 2097811 2656646 2597249 1981230 1636841 Major Faults 362 450 365 364 365 Note the impact preserving the write bit across protection updates and fault reduces faults. NUMA alloc hit 1229008 1217015 1191660 1178322 1199681 NUMA alloc miss 0 0 0 0 0 NUMA interleave hit 0 0 0 0 0 NUMA alloc local 1228514 1216317 1190871 1177448 1199021 NUMA base PTE updates 245706197 240041607 238195516 244704842 115012800 NUMA huge PMD updates 479530 468448 464868 477573 224487 NUMA page range updates 491225557 479886983 476207932 489222218 229950144 NUMA hint faults 659753 656503 641678 656926 294842 NUMA hint local faults 381604 373963 360478 337585 186249 NUMA hint local percent 57 56 56 51 63 NUMA pages migrated 5412140 6374899 6266530 5277468 5755096 AutoNUMA cost 5121% 5083% 4994% 5097% 2388% Here the impact of slowing the PTE scanner on migratrion failures is obvious as "NUMA base PTE updates" and "NUMA huge PMD updates" are massively reduced even though the headline performance is very similar. As xfsrepair was the reported workload here is the impact of the series on it. xfsrepair 3.19.0 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 vanilla vanilla vmwrite-v5r8 preserve-v5r8 slowscan-v5r8 Min real-fsmark 1183.29 ( 0.00%) 1165.73 ( 1.48%) 1152.78 ( 2.58%) 1153.64 ( 2.51%) 1177.62 ( 0.48%) Min syst-fsmark 4107.85 ( 0.00%) 4027.75 ( 1.95%) 3986.74 ( 2.95%) 3979.16 ( 3.13%) 4048.76 ( 1.44%) Min real-xfsrepair 441.51 ( 0.00%) 463.96 ( -5.08%) 449.50 ( -1.81%) 440.08 ( 0.32%) 439.87 ( 0.37%) Min syst-xfsrepair 195.76 ( 0.00%) 278.47 (-42.25%) 262.34 (-34.01%) 203.70 ( -4.06%) 143.64 ( 26.62%) Amean real-fsmark 1188.30 ( 0.00%) 1177.34 ( 0.92%) 1157.97 ( 2.55%) 1158.21 ( 2.53%) 1182.22 ( 0.51%) Amean syst-fsmark 4111.37 ( 0.00%) 4055.70 ( 1.35%) 3987.19 ( 3.02%) 3998.72 ( 2.74%) 4061.69 ( 1.21%) Amean real-xfsrepair 450.88 ( 0.00%) 468.32 ( -3.87%) 454.14 ( -0.72%) 442.36 ( 1.89%) 440.59 ( 2.28%) Amean syst-xfsrepair 199.66 ( 0.00%) 290.60 (-45.55%) 277.20 (-38.84%) 204.68 ( -2.51%) 150.55 ( 24.60%) Stddev real-fsmark 4.12 ( 0.00%) 10.82 (-162.29%) 4.14 ( -0.28%) 5.98 (-45.05%) 4.60 (-11.53%) Stddev syst-fsmark 2.63 ( 0.00%) 20.32 (-671.82%) 0.37 ( 85.89%) 16.47 (-525.59%) 15.05 (-471.79%) Stddev real-xfsrepair 6.87 ( 0.00%) 4.55 ( 33.75%) 3.46 ( 49.58%) 1.78 ( 74.12%) 0.52 ( 92.50%) Stddev syst-xfsrepair 3.02 ( 0.00%) 10.30 (-241.37%) 13.17 (-336.37%) 0.71 ( 76.63%) 5.00 (-65.61%) CoeffVar real-fsmark 0.35 ( 0.00%) 0.92 (-164.73%) 0.36 ( -2.91%) 0.52 (-48.82%) 0.39 (-12.10%) CoeffVar syst-fsmark 0.06 ( 0.00%) 0.50 (-682.41%) 0.01 ( 85.45%) 0.41 (-543.22%) 0.37 (-478.78%) CoeffVar real-xfsrepair 1.52 ( 0.00%) 0.97 ( 36.21%) 0.76 ( 49.94%) 0.40 ( 73.62%) 0.12 ( 92.33%) CoeffVar syst-xfsrepair 1.51 ( 0.00%) 3.54 (-134.54%) 4.75 (-214.31%) 0.34 ( 77.20%) 3.32 (-119.63%) Max real-fsmark 1193.39 ( 0.00%) 1191.77 ( 0.14%) 1162.90 ( 2.55%) 1166.66 ( 2.24%) 1188.50 ( 0.41%) Max syst-fsmark 4114.18 ( 0.00%) 4075.45 ( 0.94%) 3987.65 ( 3.08%) 4019.45 ( 2.30%) 4082.80 ( 0.76%) Max real-xfsrepair 457.80 ( 0.00%) 474.60 ( -3.67%) 457.82 ( -0.00%) 444.42 ( 2.92%) 441.03 ( 3.66%) Max syst-xfsrepair 203.11 ( 0.00%) 303.65 (-49.50%) 294.35 (-44.92%) 205.33 ( -1.09%) 155.28 ( 23.55%) The really relevant lines as syst-xfsrepair which is the system CPU usage when running xfsrepair. Note that on my machine the overhead was 45% higher on 4.0-rc4 which may be part of what Dave is seeing. Once we preserve the write bit across faults, it's only 2.51% higher on average. With the full series applied, system CPU usage is 24.6% lower on average. Again, the impact of preserving the write bit on minor faults is obvious and the impact of slowing scanning after migration failures is obvious on the PTE updates. Note also that the number of pages migrated is much reduced even though the headline performance is comparable. 3.19.0 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 vanilla vanillavmwrite-v5r8preserve-v5r8slowscan-v5r8 Minor Faults 153466827 254507978 249163829 153501373 105737890 Major Faults 610 702 690 649 724 NUMA base PTE updates 217735049 210756527 217729596 216937111 144344993 NUMA huge PMD updates 129294 85044 106921 127246 79887 NUMA pages migrated 21938995 29705270 28594162 22687324 16258075 3.19.0 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 4.0.0-rc4 vanilla vanillavmwrite-v5r8preserve-v5r8slowscan-v5r8 Mean sdb-avgqusz 13.47 2.54 2.55 2.47 2.49 Mean sdb-avgrqsz 202.32 140.22 139.50 139.02 138.12 Mean sdb-await 25.92 5.09 5.33 5.02 5.22 Mean sdb-r_await 4.71 0.19 0.83 0.51 0.11 Mean sdb-w_await 104.13 5.21 5.38 5.05 5.32 Mean sdb-svctm 0.59 0.13 0.14 0.13 0.14 Mean sdb-rrqm 0.16 0.00 0.00 0.00 0.00 Mean sdb-wrqm 3.59 1799.43 1826.84 1812.21 1785.67 Max sdb-avgqusz 111.06 12.13 14.05 11.66 15.60 Max sdb-avgrqsz 255.60 190.34 190.01 187.33 191.78 Max sdb-await 168.24 39.28 49.22 44.64 65.62 Max sdb-r_await 660.00 52.00 280.00 76.00 12.00 Max sdb-w_await 7804.00 39.28 49.22 44.64 65.62 Max sdb-svctm 4.00 2.82 2.86 1.98 2.84 Max sdb-rrqm 8.30 0.00 0.00 0.00 0.00 Max sdb-wrqm 34.20 5372.80 5278.60 5386.60 5546.15 FWIW, I also checked SPECjbb in different configurations but it's similar observations -- minor faults lower, PTE update activity lower and performance is roughly comparable against 3.19. This patch (of 3): Threads that share writable data within pages are grouped together as related tasks. This decision is based on whether the PTE is marked dirty which is subject to timing races between the PTE scanner update and when the application writes the page. If the page is file-backed, then background flushes and sync also affect placement. This is unpredictable behaviour which is impossible to reason about so this patch makes grouping decisions based on the VMA flags. Signed-off-by: Mel Gorman Reported-by: Dave Chinner Tested-by: Dave Chinner Cc: Ingo Molnar Cc: Aneesh Kumar Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/huge_memory.c | 13 ++----------- mm/memory.c | 19 +++++++++++-------- 2 files changed, 13 insertions(+), 19 deletions(-) diff --git a/mm/huge_memory.c b/mm/huge_memory.c index 626e93db28ba..2f12e9fcf1a2 100644 --- a/mm/huge_memory.c +++ b/mm/huge_memory.c @@ -1291,17 +1291,8 @@ int do_huge_pmd_numa_page(struct mm_struct *mm, struct vm_area_struct *vma, flags |= TNF_FAULT_LOCAL; } - /* - * Avoid grouping on DSO/COW pages in specific and RO pages - * in general, RO pages shouldn't hurt as much anyway since - * they can be in shared cache state. - * - * FIXME! This checks "pmd_dirty()" as an approximation of - * "is this a read-only page", since checking "pmd_write()" - * is even more broken. We haven't actually turned this into - * a writable page, so pmd_write() will always be false. - */ - if (!pmd_dirty(pmd)) + /* See similar comment in do_numa_page for explanation */ + if (!(vma->vm_flags & VM_WRITE)) flags |= TNF_NO_GROUP; /* diff --git a/mm/memory.c b/mm/memory.c index 411144f977b1..20beb6647dba 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -3069,16 +3069,19 @@ static int do_numa_page(struct mm_struct *mm, struct vm_area_struct *vma, } /* - * Avoid grouping on DSO/COW pages in specific and RO pages - * in general, RO pages shouldn't hurt as much anyway since - * they can be in shared cache state. + * Avoid grouping on RO pages in general. RO pages shouldn't hurt as + * much anyway since they can be in shared cache state. This misses + * the case where a mapping is writable but the process never writes + * to it but pte_write gets cleared during protection updates and + * pte_dirty has unpredictable behaviour between PTE scan updates, + * background writeback, dirty balancing and application behaviour. * - * FIXME! This checks "pmd_dirty()" as an approximation of - * "is this a read-only page", since checking "pmd_write()" - * is even more broken. We haven't actually turned this into - * a writable page, so pmd_write() will always be false. + * TODO: Note that the ideal here would be to avoid a situation where a + * NUMA fault is taken immediately followed by a write fault in + * some cases which would have lower overhead overall but would be + * invasive as the fault paths would need to be unified. */ - if (!pte_dirty(pte)) + if (!(vma->vm_flags & VM_WRITE)) flags |= TNF_NO_GROUP; /* -- cgit v1.2.1 From b191f9b106ea1a24a711dbebb2925d3313da5852 Mon Sep 17 00:00:00 2001 From: Mel Gorman Date: Wed, 25 Mar 2015 15:55:40 -0700 Subject: mm: numa: preserve PTE write permissions across a NUMA hinting fault Protecting a PTE to trap a NUMA hinting fault clears the writable bit and further faults are needed after trapping a NUMA hinting fault to set the writable bit again. This patch preserves the writable bit when trapping NUMA hinting faults. The impact is obvious from the number of minor faults trapped during the basis balancing benchmark and the system CPU usage; autonumabench 4.0.0-rc4 4.0.0-rc4 baseline preserve Time System-NUMA01 107.13 ( 0.00%) 103.13 ( 3.73%) Time System-NUMA01_THEADLOCAL 131.87 ( 0.00%) 83.30 ( 36.83%) Time System-NUMA02 8.95 ( 0.00%) 10.72 (-19.78%) Time System-NUMA02_SMT 4.57 ( 0.00%) 3.99 ( 12.69%) Time Elapsed-NUMA01 515.78 ( 0.00%) 517.26 ( -0.29%) Time Elapsed-NUMA01_THEADLOCAL 384.10 ( 0.00%) 384.31 ( -0.05%) Time Elapsed-NUMA02 48.86 ( 0.00%) 48.78 ( 0.16%) Time Elapsed-NUMA02_SMT 47.98 ( 0.00%) 48.12 ( -0.29%) 4.0.0-rc4 4.0.0-rc4 baseline preserve User 44383.95 43971.89 System 252.61 201.24 Elapsed 998.68 1000.94 Minor Faults 2597249 1981230 Major Faults 365 364 There is a similar drop in system CPU usage using Dave Chinner's xfsrepair workload 4.0.0-rc4 4.0.0-rc4 baseline preserve Amean real-xfsrepair 454.14 ( 0.00%) 442.36 ( 2.60%) Amean syst-xfsrepair 277.20 ( 0.00%) 204.68 ( 26.16%) The patch looks hacky but the alternatives looked worse. The tidest was to rewalk the page tables after a hinting fault but it was more complex than this approach and the performance was worse. It's not generally safe to just mark the page writable during the fault if it's a write fault as it may have been read-only for COW so that approach was discarded. Signed-off-by: Mel Gorman Reported-by: Dave Chinner Tested-by: Dave Chinner Cc: Ingo Molnar Cc: Aneesh Kumar Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/huge_memory.c | 9 ++++++++- mm/memory.c | 8 +++----- mm/mprotect.c | 3 +++ 3 files changed, 14 insertions(+), 6 deletions(-) diff --git a/mm/huge_memory.c b/mm/huge_memory.c index 2f12e9fcf1a2..0a42d1521aa4 100644 --- a/mm/huge_memory.c +++ b/mm/huge_memory.c @@ -1260,6 +1260,7 @@ int do_huge_pmd_numa_page(struct mm_struct *mm, struct vm_area_struct *vma, int target_nid, last_cpupid = -1; bool page_locked; bool migrated = false; + bool was_writable; int flags = 0; /* A PROT_NONE fault should not end up here */ @@ -1354,7 +1355,10 @@ int do_huge_pmd_numa_page(struct mm_struct *mm, struct vm_area_struct *vma, goto out; clear_pmdnuma: BUG_ON(!PageLocked(page)); + was_writable = pmd_write(pmd); pmd = pmd_modify(pmd, vma->vm_page_prot); + if (was_writable) + pmd = pmd_mkwrite(pmd); set_pmd_at(mm, haddr, pmdp, pmd); update_mmu_cache_pmd(vma, addr, pmdp); unlock_page(page); @@ -1478,6 +1482,7 @@ int change_huge_pmd(struct vm_area_struct *vma, pmd_t *pmd, if (__pmd_trans_huge_lock(pmd, vma, &ptl) == 1) { pmd_t entry; + bool preserve_write = prot_numa && pmd_write(*pmd); ret = 1; /* @@ -1493,9 +1498,11 @@ int change_huge_pmd(struct vm_area_struct *vma, pmd_t *pmd, if (!prot_numa || !pmd_protnone(*pmd)) { entry = pmdp_get_and_clear_notify(mm, addr, pmd); entry = pmd_modify(entry, newprot); + if (preserve_write) + entry = pmd_mkwrite(entry); ret = HPAGE_PMD_NR; set_pmd_at(mm, addr, pmd, entry); - BUG_ON(pmd_write(entry)); + BUG_ON(!preserve_write && pmd_write(entry)); } spin_unlock(ptl); } diff --git a/mm/memory.c b/mm/memory.c index 20beb6647dba..d20e12da3a3c 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -3035,6 +3035,7 @@ static int do_numa_page(struct mm_struct *mm, struct vm_area_struct *vma, int last_cpupid; int target_nid; bool migrated = false; + bool was_writable = pte_write(pte); int flags = 0; /* A PROT_NONE fault should not end up here */ @@ -3059,6 +3060,8 @@ static int do_numa_page(struct mm_struct *mm, struct vm_area_struct *vma, /* Make it present again */ pte = pte_modify(pte, vma->vm_page_prot); pte = pte_mkyoung(pte); + if (was_writable) + pte = pte_mkwrite(pte); set_pte_at(mm, addr, ptep, pte); update_mmu_cache(vma, addr, ptep); @@ -3075,11 +3078,6 @@ static int do_numa_page(struct mm_struct *mm, struct vm_area_struct *vma, * to it but pte_write gets cleared during protection updates and * pte_dirty has unpredictable behaviour between PTE scan updates, * background writeback, dirty balancing and application behaviour. - * - * TODO: Note that the ideal here would be to avoid a situation where a - * NUMA fault is taken immediately followed by a write fault in - * some cases which would have lower overhead overall but would be - * invasive as the fault paths would need to be unified. */ if (!(vma->vm_flags & VM_WRITE)) flags |= TNF_NO_GROUP; diff --git a/mm/mprotect.c b/mm/mprotect.c index 44727811bf4c..88584838e704 100644 --- a/mm/mprotect.c +++ b/mm/mprotect.c @@ -75,6 +75,7 @@ static unsigned long change_pte_range(struct vm_area_struct *vma, pmd_t *pmd, oldpte = *pte; if (pte_present(oldpte)) { pte_t ptent; + bool preserve_write = prot_numa && pte_write(oldpte); /* * Avoid trapping faults against the zero or KSM @@ -94,6 +95,8 @@ static unsigned long change_pte_range(struct vm_area_struct *vma, pmd_t *pmd, ptent = ptep_modify_prot_start(mm, addr, pte); ptent = pte_modify(ptent, newprot); + if (preserve_write) + ptent = pte_mkwrite(ptent); /* Avoid taking write faults for known dirty pages */ if (dirty_accountable && pte_dirty(ptent) && -- cgit v1.2.1 From 074c238177a75f5e79af3b2cb6a84e54823ef950 Mon Sep 17 00:00:00 2001 From: Mel Gorman Date: Wed, 25 Mar 2015 15:55:42 -0700 Subject: mm: numa: slow PTE scan rate if migration failures occur Dave Chinner reported the following on https://lkml.org/lkml/2015/3/1/226 Across the board the 4.0-rc1 numbers are much slower, and the degradation is far worse when using the large memory footprint configs. Perf points straight at the cause - this is from 4.0-rc1 on the "-o bhash=101073" config: - 56.07% 56.07% [kernel] [k] default_send_IPI_mask_sequence_phys - default_send_IPI_mask_sequence_phys - 99.99% physflat_send_IPI_mask - 99.37% native_send_call_func_ipi smp_call_function_many - native_flush_tlb_others - 99.85% flush_tlb_page ptep_clear_flush try_to_unmap_one rmap_walk try_to_unmap migrate_pages migrate_misplaced_page - handle_mm_fault - 99.73% __do_page_fault trace_do_page_fault do_async_page_fault + async_page_fault 0.63% native_send_call_func_single_ipi generic_exec_single smp_call_function_single This is showing excessive migration activity even though excessive migrations are meant to get throttled. Normally, the scan rate is tuned on a per-task basis depending on the locality of faults. However, if migrations fail for any reason then the PTE scanner may scan faster if the faults continue to be remote. This means there is higher system CPU overhead and fault trapping at exactly the time we know that migrations cannot happen. This patch tracks when migration failures occur and slows the PTE scanner. Signed-off-by: Mel Gorman Reported-by: Dave Chinner Tested-by: Dave Chinner Cc: Ingo Molnar Cc: Aneesh Kumar Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- include/linux/sched.h | 9 +++++---- kernel/sched/fair.c | 8 ++++++-- mm/huge_memory.c | 3 ++- mm/memory.c | 3 ++- 4 files changed, 15 insertions(+), 8 deletions(-) diff --git a/include/linux/sched.h b/include/linux/sched.h index 6d77432e14ff..a419b65770d6 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1625,11 +1625,11 @@ struct task_struct { /* * numa_faults_locality tracks if faults recorded during the last - * scan window were remote/local. The task scan period is adapted - * based on the locality of the faults with different weights - * depending on whether they were shared or private faults + * scan window were remote/local or failed to migrate. The task scan + * period is adapted based on the locality of the faults with different + * weights depending on whether they were shared or private faults */ - unsigned long numa_faults_locality[2]; + unsigned long numa_faults_locality[3]; unsigned long numa_pages_migrated; #endif /* CONFIG_NUMA_BALANCING */ @@ -1719,6 +1719,7 @@ struct task_struct { #define TNF_NO_GROUP 0x02 #define TNF_SHARED 0x04 #define TNF_FAULT_LOCAL 0x08 +#define TNF_MIGRATE_FAIL 0x10 #ifdef CONFIG_NUMA_BALANCING extern void task_numa_fault(int last_node, int node, int pages, int flags); diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index 7ce18f3c097a..bcfe32088b37 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -1609,9 +1609,11 @@ static void update_task_scan_period(struct task_struct *p, /* * If there were no record hinting faults then either the task is * completely idle or all activity is areas that are not of interest - * to automatic numa balancing. Scan slower + * to automatic numa balancing. Related to that, if there were failed + * migration then it implies we are migrating too quickly or the local + * node is overloaded. In either case, scan slower */ - if (local + shared == 0) { + if (local + shared == 0 || p->numa_faults_locality[2]) { p->numa_scan_period = min(p->numa_scan_period_max, p->numa_scan_period << 1); @@ -2080,6 +2082,8 @@ void task_numa_fault(int last_cpupid, int mem_node, int pages, int flags) if (migrated) p->numa_pages_migrated += pages; + if (flags & TNF_MIGRATE_FAIL) + p->numa_faults_locality[2] += pages; p->numa_faults[task_faults_idx(NUMA_MEMBUF, mem_node, priv)] += pages; p->numa_faults[task_faults_idx(NUMA_CPUBUF, cpu_node, priv)] += pages; diff --git a/mm/huge_memory.c b/mm/huge_memory.c index 0a42d1521aa4..51b3e7c64622 100644 --- a/mm/huge_memory.c +++ b/mm/huge_memory.c @@ -1350,7 +1350,8 @@ int do_huge_pmd_numa_page(struct mm_struct *mm, struct vm_area_struct *vma, if (migrated) { flags |= TNF_MIGRATED; page_nid = target_nid; - } + } else + flags |= TNF_MIGRATE_FAIL; goto out; clear_pmdnuma: diff --git a/mm/memory.c b/mm/memory.c index d20e12da3a3c..97839f5c8c30 100644 --- a/mm/memory.c +++ b/mm/memory.c @@ -3103,7 +3103,8 @@ static int do_numa_page(struct mm_struct *mm, struct vm_area_struct *vma, if (migrated) { page_nid = target_nid; flags |= TNF_MIGRATED; - } + } else + flags |= TNF_MIGRATE_FAIL; out: if (page_nid != -1) -- cgit v1.2.1 From b7b04004ecd9e58cdc6c6ff92f251d5ac5c0adb2 Mon Sep 17 00:00:00 2001 From: Mel Gorman Date: Wed, 25 Mar 2015 15:55:45 -0700 Subject: mm: numa: mark huge PTEs young when clearing NUMA hinting faults Base PTEs are marked young when the NUMA hinting information is cleared but the same does not happen for huge pages which this patch addresses. Note that migrated pages are not marked young as the base page migration code does not assume that migrated pages have been referenced. This could be addressed but beyond the scope of this series which is aimed at Dave Chinners shrink workload that is unlikely to be affected by this issue. Signed-off-by: Mel Gorman Cc: Dave Chinner Cc: Ingo Molnar Cc: Aneesh Kumar Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- mm/huge_memory.c | 1 + 1 file changed, 1 insertion(+) diff --git a/mm/huge_memory.c b/mm/huge_memory.c index 51b3e7c64622..6817b0350c71 100644 --- a/mm/huge_memory.c +++ b/mm/huge_memory.c @@ -1358,6 +1358,7 @@ clear_pmdnuma: BUG_ON(!PageLocked(page)); was_writable = pmd_write(pmd); pmd = pmd_modify(pmd, vma->vm_page_prot); + pmd = pmd_mkyoung(pmd); if (was_writable) pmd = pmd_mkwrite(pmd); set_pmd_at(mm, haddr, pmdp, pmd); -- cgit v1.2.1 From 6914e1e3f63caa829431160f0f7093292daef2d5 Mon Sep 17 00:00:00 2001 From: Vineet Gupta Date: Thu, 26 Mar 2015 09:25:44 +0530 Subject: ARC: SA_SIGINFO ucontext regs off-by-one The regfile provided to SA_SIGINFO signal handler as ucontext was off by one due to pt_regs gutter cleanups in 2013. Before handling signal, user pt_regs are copied onto user_regs_struct and copied back later. Both structs are binary compatible. This was all fine until commit 2fa919045b72 (ARC: pt_regs update #2) which removed the empty stack slot at top of pt_regs (corresponding to first pad) and made the corresponding fixup in struct user_regs_struct (the pad in there was moved out of @scratch - not removed altogether as it is part of ptrace ABI) struct user_regs_struct { + long pad; struct { - long pad; long bta, lp_start, lp_end,.... } scratch; ... } This meant that now user_regs_struct was off by 1 reg w.r.t pt_regs and signal code needs to user_regs_struct.scratch to reflect it as pt_regs, which is what this commit does. This problem was hidden for 2 years, because both save/restore, despite using wrong location, were using the same location. Only an interim inspection (reproducer below) exposed the issue. void handle_segv(int signo, siginfo_t *info, void *context) { ucontext_t *uc = context; struct user_regs_struct *regs = &(uc->uc_mcontext.regs); printf("regs %x %x\n", <=== prints 7 8 (vs. 8 9) regs->scratch.r8, regs->scratch.r9); } int main() { struct sigaction sa; sa.sa_sigaction = handle_segv; sa.sa_flags = SA_SIGINFO; sigemptyset(&sa.sa_mask); sigaction(SIGSEGV, &sa, NULL); asm volatile( "mov r7, 7 \n" "mov r8, 8 \n" "mov r9, 9 \n" "mov r10, 10 \n" :::"r7","r8","r9","r10"); *((unsigned int*)0x10) = 0; } Fixes: 2fa919045b72ec892e "ARC: pt_regs update #2: Remove unused gutter at start of pt_regs" CC: Signed-off-by: Vineet Gupta --- arch/arc/kernel/signal.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/arch/arc/kernel/signal.c b/arch/arc/kernel/signal.c index 114234e83caa..fdd3c98dfb0f 100644 --- a/arch/arc/kernel/signal.c +++ b/arch/arc/kernel/signal.c @@ -67,7 +67,7 @@ stash_usr_regs(struct rt_sigframe __user *sf, struct pt_regs *regs, sigset_t *set) { int err; - err = __copy_to_user(&(sf->uc.uc_mcontext.regs), regs, + err = __copy_to_user(&(sf->uc.uc_mcontext.regs.scratch), regs, sizeof(sf->uc.uc_mcontext.regs.scratch)); err |= __copy_to_user(&sf->uc.uc_sigmask, set, sizeof(sigset_t)); @@ -83,7 +83,7 @@ static int restore_usr_regs(struct pt_regs *regs, struct rt_sigframe __user *sf) if (!err) set_current_blocked(&set); - err |= __copy_from_user(regs, &(sf->uc.uc_mcontext.regs), + err |= __copy_from_user(regs, &(sf->uc.uc_mcontext.regs.scratch), sizeof(sf->uc.uc_mcontext.regs.scratch)); return err; -- cgit v1.2.1 From e4140819dadc3624accac8294881bca8a3cba4ed Mon Sep 17 00:00:00 2001 From: Vineet Gupta Date: Thu, 26 Mar 2015 11:14:41 +0530 Subject: ARC: signal handling robustify A malicious signal handler / restorer can DOS the system by fudging the user regs saved on stack, causing weird things such as sigreturn returning to user mode PC but cpu state still being kernel mode.... Ensure that in sigreturn path status32 always has U bit; any other bogosity (gargbage PC etc) will be taken care of by normal user mode exceptions mechanisms. Reproducer signal handler: void handle_sig(int signo, siginfo_t *info, void *context) { ucontext_t *uc = context; struct user_regs_struct *regs = &(uc->uc_mcontext.regs); regs->scratch.status32 = 0; } Before the fix, kernel would go off to weeds like below: --------->8----------- [ARCLinux]$ ./signal-test Path: /signal-test CPU: 0 PID: 61 Comm: signal-test Not tainted 4.0.0-rc5+ #65 task: 8f177880 ti: 5ffe6000 task.ti: 8f15c000 [ECR ]: 0x00220200 => Invalid Write @ 0x00000010 by insn @ 0x00010698 [EFA ]: 0x00000010 [BLINK ]: 0x2007c1ee [ERET ]: 0x10698 [STAT32]: 0x00000000 : <-------- BTA: 0x00010680 SP: 0x5ffe7e48 FP: 0x00000000 LPS: 0x20003c6c LPE: 0x20003c70 LPC: 0x00000000 ... --------->8----------- Reported-by: Alexey Brodkin Cc: Signed-off-by: Vineet Gupta --- arch/arc/kernel/signal.c | 20 ++++++++++++++++---- 1 file changed, 16 insertions(+), 4 deletions(-) diff --git a/arch/arc/kernel/signal.c b/arch/arc/kernel/signal.c index fdd3c98dfb0f..edda76fae83f 100644 --- a/arch/arc/kernel/signal.c +++ b/arch/arc/kernel/signal.c @@ -131,6 +131,15 @@ SYSCALL_DEFINE0(rt_sigreturn) /* Don't restart from sigreturn */ syscall_wont_restart(regs); + /* + * Ensure that sigreturn always returns to user mode (in case the + * regs saved on user stack got fudged between save and sigreturn) + * Otherwise it is easy to panic the kernel with a custom + * signal handler and/or restorer which clobberes the status32/ret + * to return to a bogus location in kernel mode. + */ + regs->status32 |= STATUS_U_MASK; + return regs->r0; badframe: @@ -229,8 +238,11 @@ setup_rt_frame(struct ksignal *ksig, sigset_t *set, struct pt_regs *regs) /* * handler returns using sigreturn stub provided already by userpsace + * If not, nuke the process right away */ - BUG_ON(!(ksig->ka.sa.sa_flags & SA_RESTORER)); + if(!(ksig->ka.sa.sa_flags & SA_RESTORER)) + return 1; + regs->blink = (unsigned long)ksig->ka.sa.sa_restorer; /* User Stack for signal handler will be above the frame just carved */ @@ -296,12 +308,12 @@ static void handle_signal(struct ksignal *ksig, struct pt_regs *regs) { sigset_t *oldset = sigmask_to_save(); - int ret; + int failed; /* Set up the stack frame */ - ret = setup_rt_frame(ksig, oldset, regs); + failed = setup_rt_frame(ksig, oldset, regs); - signal_setup_done(ret, ksig, 0); + signal_setup_done(failed, ksig, 0); } void do_signal(struct pt_regs *regs) -- cgit v1.2.1 From db48abf4367cb1f9e118defee0a37238638c2752 Mon Sep 17 00:00:00 2001 From: Libin Yang Date: Thu, 26 Mar 2015 13:28:39 +0800 Subject: ALSA: hda_intel: apply the Seperate stream_tag for Sunrise Point The total stream number of Sunrise Point's input and output stream exceeds 15, which will cause some streams do not work because of the overflow on SDxCTL.STRM field if using the legacy stream tag allocation method. This patch uses the new stream tag allocation method by add the flag AZX_DCAPS_SEPARATE_STREAM_TAG for Skylake platform. Signed-off-by: Libin Yang Signed-off-by: Takashi Iwai --- sound/pci/hda/hda_intel.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/pci/hda/hda_intel.c b/sound/pci/hda/hda_intel.c index 4ca3d5d02436..a8a1e14272a1 100644 --- a/sound/pci/hda/hda_intel.c +++ b/sound/pci/hda/hda_intel.c @@ -1989,7 +1989,7 @@ static const struct pci_device_id azx_ids[] = { .driver_data = AZX_DRIVER_PCH | AZX_DCAPS_INTEL_PCH }, /* Sunrise Point */ { PCI_DEVICE(0x8086, 0xa170), - .driver_data = AZX_DRIVER_PCH | AZX_DCAPS_INTEL_PCH }, + .driver_data = AZX_DRIVER_PCH | AZX_DCAPS_INTEL_SKYLAKE }, /* Sunrise Point-LP */ { PCI_DEVICE(0x8086, 0x9d70), .driver_data = AZX_DRIVER_PCH | AZX_DCAPS_INTEL_SKYLAKE }, -- cgit v1.2.1 From 832a3aad1e2927b1684e7369c9f36a370e0b95da Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Wed, 18 Mar 2015 18:19:22 +0000 Subject: drm/i915: Keep ring->active_list and ring->requests_list consistent If we retire requests last, we may use a later seqno and so clear the requests lists without clearing the active list, leading to confusion. Hence we should retire requests first for consistency with the early return. The order used to be important as the lifecycle for the object on the active list was determined by request->seqno. However, the requests themselves are now reference counted removing the constraint from the order of retirement. Fixes regression from commit 1b5a433a4dd967b125131da42b89b5cc0d5b1f57 Author: John Harrison Date: Mon Nov 24 18:49:42 2014 +0000 drm/i915: Convert 'i915_seqno_passed' calls into 'i915_gem_request_completed ' and a WARNING: CPU: 0 PID: 1383 at drivers/gpu/drm/i915/i915_gem_evict.c:279 i915_gem_evict_vm+0x10c/0x140() WARN_ON(!list_empty(&vm->active_list)) Identified by updating WATCH_LISTS: [drm:i915_verify_lists] *ERROR* blitter ring: active list not empty, but no requests WARNING: CPU: 0 PID: 681 at drivers/gpu/drm/i915/i915_gem.c:2751 i915_gem_retire_requests_ring+0x149/0x230() WARN_ON(i915_verify_lists(ring->dev)) Note that this is only a problem in evict_vm where the following happens after a retire_request has cleaned out all requests, but not all active bo: - intel_ring_idle called from i915_gpu_idle notices that no requests are outstanding and immediately returns. - i915_gem_retire_requests_ring called from i915_gem_retire_requests also immediately returns when there's no request, still leaving the bo on the active list. - evict_vm hits the WARN_ON(!list_empty(&vm->active_list)) after evicting all active objects that there's still stuff left that shouldn't be there. Signed-off-by: Chris Wilson Cc: John Harrison Cc: Daniel Vetter Reviewed-by: Daniel Vetter Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/i915_gem.c | 38 +++++++++++++++++++++----------------- 1 file changed, 21 insertions(+), 17 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 5b205863b659..27ea6bdebce7 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -2737,24 +2737,11 @@ i915_gem_retire_requests_ring(struct intel_engine_cs *ring) WARN_ON(i915_verify_lists(ring->dev)); - /* Move any buffers on the active list that are no longer referenced - * by the ringbuffer to the flushing/inactive lists as appropriate, - * before we free the context associated with the requests. + /* Retire requests first as we use it above for the early return. + * If we retire requests last, we may use a later seqno and so clear + * the requests lists without clearing the active list, leading to + * confusion. */ - while (!list_empty(&ring->active_list)) { - struct drm_i915_gem_object *obj; - - obj = list_first_entry(&ring->active_list, - struct drm_i915_gem_object, - ring_list); - - if (!i915_gem_request_completed(obj->last_read_req, true)) - break; - - i915_gem_object_move_to_inactive(obj); - } - - while (!list_empty(&ring->request_list)) { struct drm_i915_gem_request *request; struct intel_ringbuffer *ringbuf; @@ -2789,6 +2776,23 @@ i915_gem_retire_requests_ring(struct intel_engine_cs *ring) i915_gem_free_request(request); } + /* Move any buffers on the active list that are no longer referenced + * by the ringbuffer to the flushing/inactive lists as appropriate, + * before we free the context associated with the requests. + */ + while (!list_empty(&ring->active_list)) { + struct drm_i915_gem_object *obj; + + obj = list_first_entry(&ring->active_list, + struct drm_i915_gem_object, + ring_list); + + if (!i915_gem_request_completed(obj->last_read_req, true)) + break; + + i915_gem_object_move_to_inactive(obj); + } + if (unlikely(ring->trace_irq_req && i915_gem_request_completed(ring->trace_irq_req, true))) { ring->irq_put(ring); -- cgit v1.2.1 From 11bc26fe372fa6da81c82c68f755d2795838a640 Mon Sep 17 00:00:00 2001 From: Richard Weinberger Date: Thu, 26 Mar 2015 10:27:06 +0100 Subject: clocksource/drivers: Fix various !CONFIG_HAS_IOMEM build errors MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Fix !CONFIG_HAS_IOMEM related build failures in three clocksource drivers. The build failures have the pattern of: drivers/clocksource/sh_cmt.c: In function ‘sh_cmt_map_memory’: drivers/clocksource/sh_cmt.c:920:2: error: implicit declaration of function ‘ioremap_nocache’ [-Werror=implicit-function-declaration] cmt->mapbase = ioremap_nocache(mem->start, resource_size(mem)); Signed-off-by: Richard Weinberger Signed-off-by: Daniel Lezcano Acked-by: Geert Uytterhoeven Cc: maxime.ripard@free-electrons.com Link: http://lkml.kernel.org/r/1427362029-6511-1-git-send-email-daniel.lezcano@linaro.org Signed-off-by: Ingo Molnar --- drivers/clocksource/Kconfig | 3 +++ 1 file changed, 3 insertions(+) diff --git a/drivers/clocksource/Kconfig b/drivers/clocksource/Kconfig index 68161f7a07d6..a0b036ccb118 100644 --- a/drivers/clocksource/Kconfig +++ b/drivers/clocksource/Kconfig @@ -192,6 +192,7 @@ config SYS_SUPPORTS_EM_STI config SH_TIMER_CMT bool "Renesas CMT timer driver" if COMPILE_TEST depends on GENERIC_CLOCKEVENTS + depends on HAS_IOMEM default SYS_SUPPORTS_SH_CMT help This enables build of a clocksource and clockevent driver for @@ -201,6 +202,7 @@ config SH_TIMER_CMT config SH_TIMER_MTU2 bool "Renesas MTU2 timer driver" if COMPILE_TEST depends on GENERIC_CLOCKEVENTS + depends on HAS_IOMEM default SYS_SUPPORTS_SH_MTU2 help This enables build of a clockevent driver for the Multi-Function @@ -210,6 +212,7 @@ config SH_TIMER_MTU2 config SH_TIMER_TMU bool "Renesas TMU timer driver" if COMPILE_TEST depends on GENERIC_CLOCKEVENTS + depends on HAS_IOMEM default SYS_SUPPORTS_SH_TMU help This enables build of a clocksource and clockevent driver for -- cgit v1.2.1 From 6e206020324c50a95486f6b279a53512febed92d Mon Sep 17 00:00:00 2001 From: Maxime Ripard Date: Thu, 26 Mar 2015 10:27:09 +0100 Subject: clocksource/drivers/sun5i: Fix cpufreq interaction with sched_clock() The sun5i timer is used as the sched-clock on certain systems, and ever since we started using cpufreq, the cpu clock (that is one of the timer's clock indirect parent) now changes as well, along with the actual sched_clock() rate. This is not accurate and not desirable. We can safely remove the sun5i sched-clock on those systems, since we have other reliable sched_clock() sources in the system. Tested-by: Hans de Goede Signed-off-by: Maxime Ripard Signed-off-by: Daniel Lezcano [ Improved the changelog. ] Cc: richard@nod.at Link: http://lkml.kernel.org/r/1427362029-6511-4-git-send-email-daniel.lezcano@linaro.org Signed-off-by: Ingo Molnar --- drivers/clocksource/timer-sun5i.c | 7 ------- 1 file changed, 7 deletions(-) diff --git a/drivers/clocksource/timer-sun5i.c b/drivers/clocksource/timer-sun5i.c index 5dcbf90b8015..58597fbcc046 100644 --- a/drivers/clocksource/timer-sun5i.c +++ b/drivers/clocksource/timer-sun5i.c @@ -17,7 +17,6 @@ #include #include #include -#include #include #include #include @@ -137,11 +136,6 @@ static struct irqaction sun5i_timer_irq = { .dev_id = &sun5i_clockevent, }; -static u64 sun5i_timer_sched_read(void) -{ - return ~readl(timer_base + TIMER_CNTVAL_LO_REG(1)); -} - static void __init sun5i_timer_init(struct device_node *node) { struct reset_control *rstc; @@ -172,7 +166,6 @@ static void __init sun5i_timer_init(struct device_node *node) writel(TIMER_CTL_ENABLE | TIMER_CTL_RELOAD, timer_base + TIMER_CTL_REG(1)); - sched_clock_register(sun5i_timer_sched_read, 32, rate); clocksource_mmio_init(timer_base + TIMER_CNTVAL_LO_REG(1), node->name, rate, 340, 32, clocksource_mmio_readl_down); -- cgit v1.2.1 From af95b41426e0b58279f8ff0ebe420df49a4e96b8 Mon Sep 17 00:00:00 2001 From: Hui Wang Date: Thu, 26 Mar 2015 17:14:55 +0800 Subject: ALSA: hda - Add one more node in the EAPD supporting candidate list We have a HP machine which use the codec node 0x17 connecting the internal speaker, and from the node capability, we saw the EAPD, if we don't set the EAPD on for this node, the internal speaker can't output any sound. Cc: BugLink: https://bugs.launchpad.net/bugs/1436745 Signed-off-by: Hui Wang Signed-off-by: Takashi Iwai --- sound/pci/hda/patch_realtek.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index 7b0f72c5c6f1..74382137b9f5 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -396,7 +396,7 @@ static void alc_auto_setup_eapd(struct hda_codec *codec, bool on) { /* We currently only handle front, HP */ static hda_nid_t pins[] = { - 0x0f, 0x10, 0x14, 0x15, 0 + 0x0f, 0x10, 0x14, 0x15, 0x17, 0 }; hda_nid_t *p; for (p = pins; *p; p++) -- cgit v1.2.1 From 3164a803416832d61268b758112e8dfd7e35cdf7 Mon Sep 17 00:00:00 2001 From: Damien Lespiau Date: Thu, 5 Feb 2015 19:24:25 +0000 Subject: drm/i915: Fix atomic state when reusing the firmware fb Right now, we get a warning when taking over the firmware fb: [drm:drm_atomic_plane_check] FB set but no CRTC with the following backtrace: [] drm_atomic_check_only+0x35d/0x510 [drm] [] drm_atomic_commit+0x17/0x60 [drm] [] drm_atomic_helper_plane_set_property+0x8d/0xd0 [drm_kms_helper] [] drm_mode_plane_set_obj_prop+0x2d/0x90 [drm] [] restore_fbdev_mode+0x6b/0xf0 [drm_kms_helper] [] drm_fb_helper_restore_fbdev_mode_unlocked+0x29/0x80 [drm_kms_helper] [] drm_fb_helper_set_par+0x22/0x50 [drm_kms_helper] [] intel_fbdev_set_par+0x1a/0x60 [i915] [] fbcon_init+0x4f4/0x580 That's because we update the plane state with the fb from the firmware, but we never associate the plane to that CRTC. We don't quite have the full DRM take over from HW state just yet, so fake enough of the plane atomic state to pass the checks. v2: Fix the state on which we set the CRTC in the case we're sharing the initial fb with another pipe. (Matt) Signed-off-by: Damien Lespiau Reviewed-by: Matt Roper Signed-off-by: Daniel Vetter [Jani: backported to drm-intel-fixes for v4.0-rc] Reference: http://mid.gmane.org/CA+5PVA7yXH=U757w8V=Zj2U1URG4nYNav20NpjtQ4svVueyPNw@mail.gmail.com Reference: http://lkml.kernel.org/r/CA+55aFweWR=nDzc2Y=rCtL_H8JfdprQiCimN5dwc+TgyD4Bjsg@mail.gmail.com Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/intel_display.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 1c12262029fb..30faf6c262e5 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2439,7 +2439,11 @@ intel_find_plane_obj(struct intel_crtc *intel_crtc, return; if (intel_alloc_plane_obj(intel_crtc, plane_config)) { - update_state_fb(intel_crtc->base.primary); + struct drm_plane *primary = intel_crtc->base.primary; + + primary->state->crtc = &intel_crtc->base; + update_state_fb(primary); + return; } @@ -2464,11 +2468,14 @@ intel_find_plane_obj(struct intel_crtc *intel_crtc, continue; if (i915_gem_obj_ggtt_offset(obj) == plane_config->base) { + struct drm_plane *primary = intel_crtc->base.primary; + if (obj->tiling_mode != I915_TILING_NONE) dev_priv->preserve_bios_swizzle = true; drm_framebuffer_reference(c->primary->fb); - intel_crtc->base.primary->fb = c->primary->fb; + primary->fb = c->primary->fb; + primary->state->crtc = &intel_crtc->base; obj->frontbuffer_bits |= INTEL_FRONTBUFFER_PRIMARY(intel_crtc->pipe); break; } -- cgit v1.2.1 From 5f407751b0ca9bd876fe8f15ff28153661c6ba0a Mon Sep 17 00:00:00 2001 From: Daniel Vetter Date: Wed, 25 Mar 2015 18:30:38 +0100 Subject: drm/i915: Fixup legacy plane->crtc link for initial fb config This is a very similar bug in the load detect code fixed in commit 9128b040eb774e04bc23777b005ace2b66ab2a85 Author: Daniel Vetter Date: Tue Mar 3 17:31:21 2015 +0100 drm/i915: Fix modeset state confusion in the load detect code But this time around it was the initial fb code that forgot to update the plane->crtc pointer. Otherwise it's the exact same bug, with the exact same restrains (any set_config call/ioctl that doesn't disable the pipe papers over the bug for free, so fairly hard to hit in normal testing). So if you want the full explanation just go read that one over there - it's rather long ... Cc: Matt Roper Cc: Linus Torvalds Cc: Chris Wilson Cc: Josh Boyer Cc: Jani Nikula Reported-and-tested-by: Josh Boyer Signed-off-by: Daniel Vetter [Jani: backported to drm-intel-fixes for v4.0-rc] Reference: http://mid.gmane.org/CA+5PVA7ChbtJrknqws1qvZcbrg1CW2pQAFkSMURWWgyASRyGXg@mail.gmail.com Signed-off-by: Jani Nikula --- drivers/gpu/drm/i915/intel_display.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 30faf6c262e5..f75173c20f47 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2442,6 +2442,7 @@ intel_find_plane_obj(struct intel_crtc *intel_crtc, struct drm_plane *primary = intel_crtc->base.primary; primary->state->crtc = &intel_crtc->base; + primary->crtc = &intel_crtc->base; update_state_fb(primary); return; @@ -2476,6 +2477,7 @@ intel_find_plane_obj(struct intel_crtc *intel_crtc, drm_framebuffer_reference(c->primary->fb); primary->fb = c->primary->fb; primary->state->crtc = &intel_crtc->base; + primary->crtc = &intel_crtc->base; obj->frontbuffer_bits |= INTEL_FRONTBUFFER_PRIMARY(intel_crtc->pipe); break; } -- cgit v1.2.1 From ab585dea120fa20313b1b5a3be2b3d614f094678 Mon Sep 17 00:00:00 2001 From: Rodrigo Vivi Date: Tue, 24 Mar 2015 12:40:09 -0700 Subject: drm/i915: kill i915.powersave This flag was being mostly used as a meta flag in some cases and not covering other cases. One of the risks is that it was masking some frontbuffer trackings without disabling PSR. So, better to kill this at once and avoid umbrella parameters. Signed-off-by: Rodrigo Vivi Acked-by: Chris Wilson [danvet: Drop unused out: label to appease gcc.] Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_drv.h | 1 - drivers/gpu/drm/i915/i915_params.c | 5 ----- drivers/gpu/drm/i915/intel_display.c | 8 ++------ drivers/gpu/drm/i915/intel_fbc.c | 2 +- drivers/gpu/drm/i915/intel_frontbuffer.c | 3 --- 5 files changed, 3 insertions(+), 16 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 0a563cc65801..256369f29975 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -2443,7 +2443,6 @@ extern int i915_resume_legacy(struct drm_device *dev); struct i915_params { int modeset; int panel_ignore_lid; - unsigned int powersave; int semaphores; unsigned int lvds_downclock; int lvds_channel_mode; diff --git a/drivers/gpu/drm/i915/i915_params.c b/drivers/gpu/drm/i915/i915_params.c index e2d20ffe6586..f3b8b62cb74d 100644 --- a/drivers/gpu/drm/i915/i915_params.c +++ b/drivers/gpu/drm/i915/i915_params.c @@ -27,7 +27,6 @@ struct i915_params i915 __read_mostly = { .modeset = -1, .panel_ignore_lid = 1, - .powersave = 1, .semaphores = -1, .lvds_downclock = 0, .lvds_channel_mode = 0, @@ -65,10 +64,6 @@ MODULE_PARM_DESC(panel_ignore_lid, "Override lid status (0=autodetect, 1=autodetect disabled [default], " "-1=force lid closed, -2=force lid open)"); -module_param_named(powersave, i915.powersave, int, 0600); -MODULE_PARM_DESC(powersave, - "Enable powersavings, fbc, downclocking, etc. (default: true)"); - module_param_named_unsafe(semaphores, i915.semaphores, int, 0400); MODULE_PARM_DESC(semaphores, "Use semaphores for inter-ring sync " diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index f9dc5babcf27..d98df2739666 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -6024,7 +6024,7 @@ static void i9xx_update_pll_dividers(struct intel_crtc *crtc, crtc->lowfreq_avail = false; if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS) && - reduced_clock && i915.powersave) { + reduced_clock) { crtc_state->dpll_hw_state.fp1 = fp2; crtc->lowfreq_avail = true; } else { @@ -7807,7 +7807,7 @@ static int ironlake_crtc_compute_clock(struct intel_crtc *crtc, } } - if (is_lvds && has_reduced_clock && i915.powersave) + if (is_lvds && has_reduced_clock) crtc->lowfreq_avail = true; else crtc->lowfreq_avail = false; @@ -9322,9 +9322,6 @@ void intel_mark_idle(struct drm_device *dev) dev_priv->mm.busy = false; - if (!i915.powersave) - goto out; - for_each_crtc(dev, crtc) { if (!crtc->primary->fb) continue; @@ -9335,7 +9332,6 @@ void intel_mark_idle(struct drm_device *dev) if (INTEL_INFO(dev)->gen >= 6) gen6_rps_idle(dev->dev_private); -out: intel_runtime_pm_put(dev_priv); } diff --git a/drivers/gpu/drm/i915/intel_fbc.c b/drivers/gpu/drm/i915/intel_fbc.c index 9fcf446e95f5..4165ce0644f7 100644 --- a/drivers/gpu/drm/i915/intel_fbc.c +++ b/drivers/gpu/drm/i915/intel_fbc.c @@ -521,7 +521,7 @@ void intel_fbc_update(struct drm_device *dev) goto out_disable; } - if (!i915.enable_fbc || !i915.powersave) { + if (!i915.enable_fbc) { if (set_no_fbc_reason(dev_priv, FBC_MODULE_PARAM)) DRM_DEBUG_KMS("fbc disabled per module param\n"); goto out_disable; diff --git a/drivers/gpu/drm/i915/intel_frontbuffer.c b/drivers/gpu/drm/i915/intel_frontbuffer.c index 0a1bac8ac72b..a20cffb78c0f 100644 --- a/drivers/gpu/drm/i915/intel_frontbuffer.c +++ b/drivers/gpu/drm/i915/intel_frontbuffer.c @@ -110,9 +110,6 @@ static void intel_mark_fb_busy(struct drm_device *dev, struct drm_i915_private *dev_priv = dev->dev_private; enum pipe pipe; - if (!i915.powersave) - return; - for_each_pipe(dev_priv, pipe) { if (!(frontbuffer_bits & INTEL_FRONTBUFFER_ALL_MASK(pipe))) continue; -- cgit v1.2.1 From 5097f8c72a38128e99ae2869e397a2f43d0a0b51 Mon Sep 17 00:00:00 2001 From: Daniel Vetter Date: Wed, 25 Mar 2015 18:30:38 +0100 Subject: drm/i915: Fixup legacy plane->crtc link for initial fb config This is a very similar bug in the load detect code fixed in commit 9128b040eb774e04bc23777b005ace2b66ab2a85 Author: Daniel Vetter Date: Tue Mar 3 17:31:21 2015 +0100 drm/i915: Fix modeset state confusion in the load detect code But this time around it was the initial fb code that forgot to update the plane->crtc pointer. Otherwise it's the exact same bug, with the exact same restrains (any set_config call/ioctl that doesn't disable the pipe papers over the bug for free, so fairly hard to hit in normal testing). So if you want the full explanation just go read that one over there - it's rather long ... Cc: Matt Roper Cc: Linus Torvalds Cc: Chris Wilson Cc: Josh Boyer Cc: Jani Nikula Reported-and-tested-by: Josh Boyer Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_display.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index d98df2739666..3b7c3491ea68 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2594,6 +2594,7 @@ intel_find_plane_obj(struct intel_crtc *intel_crtc, primary->fb = &plane_config->fb->base; primary->state->crtc = &intel_crtc->base; + primary->crtc = &intel_crtc->base; update_state_fb(primary); return; @@ -2627,6 +2628,7 @@ intel_find_plane_obj(struct intel_crtc *intel_crtc, drm_framebuffer_reference(c->primary->fb); primary->fb = c->primary->fb; primary->state->crtc = &intel_crtc->base; + primary->crtc = &intel_crtc->base; update_state_fb(intel_crtc->base.primary); obj->frontbuffer_bits |= INTEL_FRONTBUFFER_PRIMARY(intel_crtc->pipe); break; -- cgit v1.2.1 From 3cd919fc010043d67259ccf37ca5c5892582051e Mon Sep 17 00:00:00 2001 From: Rodrigo Vivi Date: Wed, 25 Mar 2015 12:18:23 -0700 Subject: drm/i915: Remove duplicated psr.active unset psr.active is being unset out of the if so this here is useless and duplicated. Signed-off-by: Rodrigo Vivi Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_psr.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_psr.c b/drivers/gpu/drm/i915/intel_psr.c index b9f40c2e0af7..a8f9348259ae 100644 --- a/drivers/gpu/drm/i915/intel_psr.c +++ b/drivers/gpu/drm/i915/intel_psr.c @@ -532,8 +532,6 @@ static void intel_psr_exit(struct drm_device *dev) WARN_ON(!(val & EDP_PSR_ENABLE)); I915_WRITE(EDP_PSR_CTL(dev), val & ~EDP_PSR_ENABLE); - - dev_priv->psr.active = false; } else { val = I915_READ(VLV_PSRCTL(pipe)); -- cgit v1.2.1 From f6936e2902a7a6be52112e735a280aa2155b2c71 Mon Sep 17 00:00:00 2001 From: Daniel Vetter Date: Thu, 26 Mar 2015 12:17:05 +0100 Subject: drm/i915: Add initial_ prefix to bios fb takeover code In spirit with commit 5724dbd1678e2f573b13f0688277941fad66cb88 Author: Damien Lespiau Date: Tue Jan 20 12:51:52 2015 +0000 drm/i915: Rename plane_config to initial_plane_config to make it clear that this code is all special-purpose for the initial plane takeover. Cc: Damien Lespiau Cc: Tvrtko Ursulin Reviewed-by: Damien Lespiau Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_display.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 3b7c3491ea68..1142ffce66ea 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2508,8 +2508,8 @@ static int skl_format_to_fourcc(int format, bool rgb_order, bool alpha) } static bool -intel_alloc_plane_obj(struct intel_crtc *crtc, - struct intel_initial_plane_config *plane_config) +intel_alloc_initial_plane_obj(struct intel_crtc *crtc, + struct intel_initial_plane_config *plane_config) { struct drm_device *dev = crtc->base.dev; struct drm_i915_gem_object *obj = NULL; @@ -2553,7 +2553,7 @@ intel_alloc_plane_obj(struct intel_crtc *crtc, obj->frontbuffer_bits = INTEL_FRONTBUFFER_PRIMARY(crtc->pipe); mutex_unlock(&dev->struct_mutex); - DRM_DEBUG_KMS("plane fb obj %p\n", obj); + DRM_DEBUG_KMS("initial plane fb obj %p\n", obj); return true; out_unref_obj: @@ -2577,8 +2577,8 @@ update_state_fb(struct drm_plane *plane) } static void -intel_find_plane_obj(struct intel_crtc *intel_crtc, - struct intel_initial_plane_config *plane_config) +intel_find_initial_plane_obj(struct intel_crtc *intel_crtc, + struct intel_initial_plane_config *plane_config) { struct drm_device *dev = intel_crtc->base.dev; struct drm_i915_private *dev_priv = dev->dev_private; @@ -2589,7 +2589,7 @@ intel_find_plane_obj(struct intel_crtc *intel_crtc, if (!plane_config->fb) return; - if (intel_alloc_plane_obj(intel_crtc, plane_config)) { + if (intel_alloc_initial_plane_obj(intel_crtc, plane_config)) { struct drm_plane *primary = intel_crtc->base.primary; primary->fb = &plane_config->fb->base; @@ -13572,7 +13572,7 @@ void intel_modeset_init(struct drm_device *dev) * If the fb is shared between multiple heads, we'll * just get the first one. */ - intel_find_plane_obj(crtc, &crtc->plane_config); + intel_find_initial_plane_obj(crtc, &crtc->plane_config); } } } -- cgit v1.2.1 From 88595ac9ad199fc2e09270ec92e1ff0806033b83 Mon Sep 17 00:00:00 2001 From: Daniel Vetter Date: Thu, 26 Mar 2015 12:42:24 +0100 Subject: drm/i915: always preserve bios swizzling MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Currently we only set preserve_bios_swizzling when the initial fb is shared and totally miss the single-screen case. Fix this by consolidating all the logic for both cases. This seems to go back to when swizzle preservation was originally merged in commit d9ceb8163339134bd3ffb9fb87a0db4698283e32 Author: Jesse Barnes Date: Thu Oct 9 12:57:43 2014 -0700 drm/i915: preserve swizzle settings if necessary v4 Cc: Kristian Høgsberg Cc: Jesse Barnes Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_display.c | 47 +++++++++++++++++------------------- 1 file changed, 22 insertions(+), 25 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 1142ffce66ea..9d314f889613 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2543,14 +2543,11 @@ intel_alloc_initial_plane_obj(struct intel_crtc *crtc, mode_cmd.flags = DRM_MODE_FB_MODIFIERS; mutex_lock(&dev->struct_mutex); - if (intel_framebuffer_init(dev, to_intel_framebuffer(fb), &mode_cmd, obj)) { DRM_DEBUG_KMS("intel fb init failed\n"); goto out_unref_obj; } - - obj->frontbuffer_bits = INTEL_FRONTBUFFER_PRIMARY(crtc->pipe); mutex_unlock(&dev->struct_mutex); DRM_DEBUG_KMS("initial plane fb obj %p\n", obj); @@ -2585,19 +2582,15 @@ intel_find_initial_plane_obj(struct intel_crtc *intel_crtc, struct drm_crtc *c; struct intel_crtc *i; struct drm_i915_gem_object *obj; + struct drm_plane *primary = intel_crtc->base.primary; + struct drm_framebuffer *fb; if (!plane_config->fb) return; if (intel_alloc_initial_plane_obj(intel_crtc, plane_config)) { - struct drm_plane *primary = intel_crtc->base.primary; - - primary->fb = &plane_config->fb->base; - primary->state->crtc = &intel_crtc->base; - primary->crtc = &intel_crtc->base; - update_state_fb(primary); - - return; + fb = &plane_config->fb->base; + goto valid_fb; } kfree(plane_config->fb); @@ -2615,25 +2608,29 @@ intel_find_initial_plane_obj(struct intel_crtc *intel_crtc, if (!i->active) continue; - obj = intel_fb_obj(c->primary->fb); - if (obj == NULL) + fb = c->primary->fb; + if (!fb) continue; + obj = intel_fb_obj(fb); if (i915_gem_obj_ggtt_offset(obj) == plane_config->base) { - struct drm_plane *primary = intel_crtc->base.primary; - - if (obj->tiling_mode != I915_TILING_NONE) - dev_priv->preserve_bios_swizzle = true; - - drm_framebuffer_reference(c->primary->fb); - primary->fb = c->primary->fb; - primary->state->crtc = &intel_crtc->base; - primary->crtc = &intel_crtc->base; - update_state_fb(intel_crtc->base.primary); - obj->frontbuffer_bits |= INTEL_FRONTBUFFER_PRIMARY(intel_crtc->pipe); - break; + drm_framebuffer_reference(fb); + goto valid_fb; } } + + return; + +valid_fb: + obj = intel_fb_obj(fb); + if (obj->tiling_mode != I915_TILING_NONE) + dev_priv->preserve_bios_swizzle = true; + + primary->fb = fb; + primary->state->crtc = &intel_crtc->base; + primary->crtc = &intel_crtc->base; + update_state_fb(primary); + obj->frontbuffer_bits |= INTEL_FRONTBUFFER_PRIMARY(intel_crtc->pipe); } static void i9xx_update_primary_plane(struct drm_crtc *crtc, -- cgit v1.2.1 From 10f81c194a78915f47aa5a17756b6492be133251 Mon Sep 17 00:00:00 2001 From: Ander Conselvan de Oliveira Date: Fri, 20 Mar 2015 16:18:01 +0200 Subject: drm/i915: Add intel_atomic_get_crtc_state() helper function The pattern of getting the crtc state with drm_atomic_get_crtc_state() and then converting it to intel_crtc_state will repeat quite often in the following patches, so add a helper function to save some typing. v2: Fix upcasting so that crtc_state base field could be moved. (Daniel) Signed-off-by: Ander Conselvan de Oliveira Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_drv.h | 13 +++++++++++++ 1 file changed, 13 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index 811a1db8b6ed..267bcfc32f13 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -35,6 +35,7 @@ #include #include #include +#include #define DIV_ROUND_CLOSEST_ULL(ll, d) \ ({ unsigned long long _tmp = (ll)+(d)/2; do_div(_tmp, d); _tmp; }) @@ -562,6 +563,7 @@ struct cxsr_latency { }; #define to_intel_crtc(x) container_of(x, struct intel_crtc, base) +#define to_intel_crtc_state(x) container_of(x, struct intel_crtc_state, base) #define to_intel_connector(x) container_of(x, struct intel_connector, base) #define to_intel_encoder(x) container_of(x, struct intel_encoder, base) #define to_intel_framebuffer(x) container_of(x, struct intel_framebuffer, base) @@ -1303,6 +1305,17 @@ int intel_connector_atomic_get_property(struct drm_connector *connector, struct drm_crtc_state *intel_crtc_duplicate_state(struct drm_crtc *crtc); void intel_crtc_destroy_state(struct drm_crtc *crtc, struct drm_crtc_state *state); +static inline struct intel_crtc_state * +intel_atomic_get_crtc_state(struct drm_atomic_state *state, + struct intel_crtc *crtc) +{ + struct drm_crtc_state *crtc_state; + crtc_state = drm_atomic_get_crtc_state(state, &crtc->base); + if (IS_ERR(crtc_state)) + return ERR_PTR(PTR_ERR(crtc_state)); + + return to_intel_crtc_state(crtc_state); +} /* intel_atomic_plane.c */ struct intel_plane_state *intel_create_plane_state(struct drm_plane *plane); -- cgit v1.2.1 From 49172fee6173d213cf711bfad751e1b38e8fdaaf Mon Sep 17 00:00:00 2001 From: Ander Conselvan de Oliveira Date: Fri, 20 Mar 2015 16:18:02 +0200 Subject: drm/i915: Pass acquire ctx also to intel_release_load_detect_pipe() For now this is not necessary since intel_set_mode() doesn't acquire any new locks. However, once that function is converted to atomic, that will change, since we'll pass an atomic state to it, and that needs to have the right acquire context set. Signed-off-by: Ander Conselvan de Oliveira Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_crt.c | 2 +- drivers/gpu/drm/i915/intel_display.c | 5 +++-- drivers/gpu/drm/i915/intel_drv.h | 3 ++- drivers/gpu/drm/i915/intel_tv.c | 2 +- 4 files changed, 7 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_crt.c b/drivers/gpu/drm/i915/intel_crt.c index e66e17af0a56..974534e784ee 100644 --- a/drivers/gpu/drm/i915/intel_crt.c +++ b/drivers/gpu/drm/i915/intel_crt.c @@ -708,7 +708,7 @@ intel_crt_detect(struct drm_connector *connector, bool force) status = connector_status_connected; else status = intel_crt_load_detect(crt); - intel_release_load_detect_pipe(connector, &tmp); + intel_release_load_detect_pipe(connector, &tmp, &ctx); } else status = connector_status_unknown; diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 9d314f889613..bbd09aaa055d 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -9038,7 +9038,8 @@ fail_unlock: } void intel_release_load_detect_pipe(struct drm_connector *connector, - struct intel_load_detect_pipe *old) + struct intel_load_detect_pipe *old, + struct drm_modeset_acquire_ctx *ctx) { struct intel_encoder *intel_encoder = intel_attached_encoder(connector); @@ -13595,7 +13596,7 @@ static void intel_enable_pipe_a(struct drm_device *dev) return; if (intel_get_load_detect_pipe(crt, NULL, &load_detect_temp, ctx)) - intel_release_load_detect_pipe(crt, &load_detect_temp); + intel_release_load_detect_pipe(crt, &load_detect_temp, ctx); } static bool diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index 267bcfc32f13..cb57b9c446f3 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -959,7 +959,8 @@ bool intel_get_load_detect_pipe(struct drm_connector *connector, struct intel_load_detect_pipe *old, struct drm_modeset_acquire_ctx *ctx); void intel_release_load_detect_pipe(struct drm_connector *connector, - struct intel_load_detect_pipe *old); + struct intel_load_detect_pipe *old, + struct drm_modeset_acquire_ctx *ctx); int intel_pin_and_fence_fb_obj(struct drm_plane *plane, struct drm_framebuffer *fb, const struct drm_plane_state *plane_state, diff --git a/drivers/gpu/drm/i915/intel_tv.c b/drivers/gpu/drm/i915/intel_tv.c index 892d23c8479d..4e6684ee8a00 100644 --- a/drivers/gpu/drm/i915/intel_tv.c +++ b/drivers/gpu/drm/i915/intel_tv.c @@ -1332,7 +1332,7 @@ intel_tv_detect(struct drm_connector *connector, bool force) if (intel_get_load_detect_pipe(connector, &mode, &tmp, &ctx)) { type = intel_tv_detect_type(intel_tv, connector); - intel_release_load_detect_pipe(connector, &tmp); + intel_release_load_detect_pipe(connector, &tmp, &ctx); status = type < 0 ? connector_status_disconnected : connector_status_connected; -- cgit v1.2.1 From 83a57153f5de299c7672d0cb241289c349b784d4 Mon Sep 17 00:00:00 2001 From: Ander Conselvan de Oliveira Date: Fri, 20 Mar 2015 16:18:03 +0200 Subject: drm/i915: Allocate a drm_atomic_state for the legacy modeset code For the atomic conversion, the mode set paths need to be changed to rely on an atomic state instead of using the staged config. By using an atomic state for the legacy code, we will be able to convert the code base in small chunks. v2: Squash patch that adds stat argument to intel_set_mode(). (Ander) Make every caller of intel_set_mode() allocate state. (Daniel) Call drm_atomic_state_clear() in set config's error path. (Daniel) v3: Copy staged config to atomic state in force restore path. (Ander) v4: Don't update ->new_config for disabled pipes in __intel_set_mode(), since it is expected to be NULL in that case. (Ander) v5: Don't change return type of intel_modeset_pipe_config(). (Chandra) Signed-off-by: Ander Conselvan de Oliveira [danvet: Remove spurious ret local variable due to changes in v5.] Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_display.c | 185 +++++++++++++++++++++++++++++------ 1 file changed, 155 insertions(+), 30 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index bbd09aaa055d..66cf0996b881 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -83,7 +83,8 @@ static void ironlake_pch_clock_get(struct intel_crtc *crtc, struct intel_crtc_state *pipe_config); static int intel_set_mode(struct drm_crtc *crtc, struct drm_display_mode *mode, - int x, int y, struct drm_framebuffer *old_fb); + int x, int y, struct drm_framebuffer *old_fb, + struct drm_atomic_state *state); static int intel_framebuffer_init(struct drm_device *dev, struct intel_framebuffer *ifb, struct drm_mode_fb_cmd2 *mode_cmd, @@ -8907,6 +8908,7 @@ bool intel_get_load_detect_pipe(struct drm_connector *connector, struct drm_device *dev = encoder->dev; struct drm_framebuffer *fb; struct drm_mode_config *config = &dev->mode_config; + struct drm_atomic_state *state = NULL; int ret, i = -1; DRM_DEBUG_KMS("[CONNECTOR:%d:%s], [ENCODER:%d:%s]\n", @@ -8988,6 +8990,12 @@ retry: old->load_detect_temp = true; old->release_fb = NULL; + state = drm_atomic_state_alloc(dev); + if (!state) + return false; + + state->acquire_ctx = ctx; + if (!mode) mode = &load_detect_mode; @@ -9010,7 +9018,7 @@ retry: goto fail; } - if (intel_set_mode(crtc, mode, 0, 0, fb)) { + if (intel_set_mode(crtc, mode, 0, 0, fb, state)) { DRM_DEBUG_KMS("failed to set mode on load-detect pipe\n"); if (old->release_fb) old->release_fb->funcs->destroy(old->release_fb); @@ -9029,6 +9037,11 @@ retry: else intel_crtc->new_config = NULL; fail_unlock: + if (state) { + drm_atomic_state_free(state); + state = NULL; + } + if (ret == -EDEADLK) { drm_modeset_backoff(ctx); goto retry; @@ -9041,22 +9054,34 @@ void intel_release_load_detect_pipe(struct drm_connector *connector, struct intel_load_detect_pipe *old, struct drm_modeset_acquire_ctx *ctx) { + struct drm_device *dev = connector->dev; struct intel_encoder *intel_encoder = intel_attached_encoder(connector); struct drm_encoder *encoder = &intel_encoder->base; struct drm_crtc *crtc = encoder->crtc; struct intel_crtc *intel_crtc = to_intel_crtc(crtc); + struct drm_atomic_state *state; DRM_DEBUG_KMS("[CONNECTOR:%d:%s], [ENCODER:%d:%s]\n", connector->base.id, connector->name, encoder->base.id, encoder->name); if (old->load_detect_temp) { + state = drm_atomic_state_alloc(dev); + if (!state) { + DRM_DEBUG_KMS("can't release load detect pipe\n"); + return; + } + + state->acquire_ctx = ctx; + to_intel_connector(connector)->new_encoder = NULL; intel_encoder->new_crtc = NULL; intel_crtc->new_enabled = false; intel_crtc->new_config = NULL; - intel_set_mode(crtc, NULL, 0, 0, NULL); + intel_set_mode(crtc, NULL, 0, 0, NULL, state); + + drm_atomic_state_free(state); if (old->release_fb) { drm_framebuffer_unregister_private(old->release_fb); @@ -10449,10 +10474,22 @@ static bool check_digital_port_conflicts(struct drm_device *dev) return true; } +static void +clear_intel_crtc_state(struct intel_crtc_state *crtc_state) +{ + struct drm_crtc_state tmp_state; + + /* Clear only the intel specific part of the crtc state */ + tmp_state = crtc_state->base; + memset(crtc_state, 0, sizeof *crtc_state); + crtc_state->base = tmp_state; +} + static struct intel_crtc_state * intel_modeset_pipe_config(struct drm_crtc *crtc, struct drm_framebuffer *fb, - struct drm_display_mode *mode) + struct drm_display_mode *mode, + struct drm_atomic_state *state) { struct drm_device *dev = crtc->dev; struct intel_encoder *encoder; @@ -10470,9 +10507,11 @@ intel_modeset_pipe_config(struct drm_crtc *crtc, return ERR_PTR(-EINVAL); } - pipe_config = kzalloc(sizeof(*pipe_config), GFP_KERNEL); - if (!pipe_config) - return ERR_PTR(-ENOMEM); + pipe_config = intel_atomic_get_crtc_state(state, to_intel_crtc(crtc)); + if (IS_ERR(pipe_config)) + return pipe_config; + + clear_intel_crtc_state(pipe_config); pipe_config->base.crtc = crtc; drm_mode_copy(&pipe_config->base.adjusted_mode, mode); @@ -10569,7 +10608,6 @@ encoder_retry: return pipe_config; fail: - kfree(pipe_config); return ERR_PTR(ret); } @@ -11248,6 +11286,7 @@ static struct intel_crtc_state * intel_modeset_compute_config(struct drm_crtc *crtc, struct drm_display_mode *mode, struct drm_framebuffer *fb, + struct drm_atomic_state *state, unsigned *modeset_pipes, unsigned *prepare_pipes, unsigned *disable_pipes) @@ -11258,7 +11297,7 @@ intel_modeset_compute_config(struct drm_crtc *crtc, prepare_pipes, disable_pipes); if ((*modeset_pipes) == 0) - goto out; + return NULL; /* * Note this needs changes when we start tracking multiple modes @@ -11266,14 +11305,13 @@ intel_modeset_compute_config(struct drm_crtc *crtc, * (i.e. one pipe_config for each crtc) rather than just the one * for this crtc. */ - pipe_config = intel_modeset_pipe_config(crtc, fb, mode); - if (IS_ERR(pipe_config)) { - goto out; - } + pipe_config = intel_modeset_pipe_config(crtc, fb, mode, state); + if (IS_ERR(pipe_config)) + return pipe_config; + intel_dump_pipe_config(to_intel_crtc(crtc), pipe_config, "[modeset]"); -out: return pipe_config; } @@ -11318,6 +11356,7 @@ static int __intel_set_mode(struct drm_crtc *crtc, struct drm_device *dev = crtc->dev; struct drm_i915_private *dev_priv = dev->dev_private; struct drm_display_mode *saved_mode; + struct intel_crtc_state *crtc_state_copy = NULL; struct intel_crtc *intel_crtc; int ret = 0; @@ -11325,6 +11364,12 @@ static int __intel_set_mode(struct drm_crtc *crtc, if (!saved_mode) return -ENOMEM; + crtc_state_copy = kmalloc(sizeof(*crtc_state_copy), GFP_KERNEL); + if (!crtc_state_copy) { + ret = -ENOMEM; + goto done; + } + *saved_mode = crtc->mode; if (modeset_pipes) @@ -11411,6 +11456,22 @@ done: if (ret && crtc->state->enable) crtc->mode = *saved_mode; + if (ret == 0 && pipe_config) { + struct intel_crtc *intel_crtc = to_intel_crtc(crtc); + + /* The pipe_config will be freed with the atomic state, so + * make a copy. */ + memcpy(crtc_state_copy, intel_crtc->config, + sizeof *crtc_state_copy); + intel_crtc->config = crtc_state_copy; + intel_crtc->base.state = &crtc_state_copy->base; + + if (modeset_pipes) + intel_crtc->new_config = intel_crtc->config; + } else { + kfree(crtc_state_copy); + } + kfree(saved_mode); return ret; } @@ -11436,27 +11497,81 @@ static int intel_set_mode_pipes(struct drm_crtc *crtc, static int intel_set_mode(struct drm_crtc *crtc, struct drm_display_mode *mode, - int x, int y, struct drm_framebuffer *fb) + int x, int y, struct drm_framebuffer *fb, + struct drm_atomic_state *state) { struct intel_crtc_state *pipe_config; unsigned modeset_pipes, prepare_pipes, disable_pipes; + int ret = 0; - pipe_config = intel_modeset_compute_config(crtc, mode, fb, + pipe_config = intel_modeset_compute_config(crtc, mode, fb, state, &modeset_pipes, &prepare_pipes, &disable_pipes); - if (IS_ERR(pipe_config)) - return PTR_ERR(pipe_config); + if (IS_ERR(pipe_config)) { + ret = PTR_ERR(pipe_config); + goto out; + } + + ret = intel_set_mode_pipes(crtc, mode, x, y, fb, pipe_config, + modeset_pipes, prepare_pipes, + disable_pipes); + if (ret) + goto out; - return intel_set_mode_pipes(crtc, mode, x, y, fb, pipe_config, - modeset_pipes, prepare_pipes, - disable_pipes); +out: + return ret; } void intel_crtc_restore_mode(struct drm_crtc *crtc) { - intel_set_mode(crtc, &crtc->mode, crtc->x, crtc->y, crtc->primary->fb); + struct drm_device *dev = crtc->dev; + struct drm_atomic_state *state; + struct intel_encoder *encoder; + struct intel_connector *connector; + struct drm_connector_state *connector_state; + + state = drm_atomic_state_alloc(dev); + if (!state) { + DRM_DEBUG_KMS("[CRTC:%d] mode restore failed, out of memory", + crtc->base.id); + return; + } + + state->acquire_ctx = dev->mode_config.acquire_ctx; + + /* The force restore path in the HW readout code relies on the staged + * config still keeping the user requested config while the actual + * state has been overwritten by the configuration read from HW. We + * need to copy the staged config to the atomic state, otherwise the + * mode set will just reapply the state the HW is already in. */ + for_each_intel_encoder(dev, encoder) { + if (&encoder->new_crtc->base != crtc) + continue; + + for_each_intel_connector(dev, connector) { + if (connector->new_encoder != encoder) + continue; + + connector_state = drm_atomic_get_connector_state(state, &connector->base); + if (IS_ERR(connector_state)) { + DRM_DEBUG_KMS("Failed to add [CONNECTOR:%d:%s] to state: %ld\n", + connector->base.base.id, + connector->base.name, + PTR_ERR(connector_state)); + continue; + } + + connector_state->crtc = crtc; + connector_state->best_encoder = &encoder->base; + } + } + + intel_set_mode(crtc, &crtc->mode, crtc->x, crtc->y, crtc->primary->fb, + state); + + drm_atomic_state_free(state); } #undef for_each_intel_crtc_masked @@ -11781,6 +11896,7 @@ static int intel_crtc_set_config(struct drm_mode_set *set) { struct drm_device *dev; struct drm_mode_set save_set; + struct drm_atomic_state *state = NULL; struct intel_set_config *config; struct intel_crtc_state *pipe_config; unsigned modeset_pipes, prepare_pipes, disable_pipes; @@ -11825,12 +11941,20 @@ static int intel_crtc_set_config(struct drm_mode_set *set) * such cases. */ intel_set_config_compute_mode_changes(set, config); + state = drm_atomic_state_alloc(dev); + if (!state) { + ret = -ENOMEM; + goto out_config; + } + + state->acquire_ctx = dev->mode_config.acquire_ctx; + ret = intel_modeset_stage_output_state(dev, set, config); if (ret) goto fail; pipe_config = intel_modeset_compute_config(set->crtc, set->mode, - set->fb, + set->fb, state, &modeset_pipes, &prepare_pipes, &disable_pipes); @@ -11850,10 +11974,6 @@ static int intel_crtc_set_config(struct drm_mode_set *set) */ } - /* set_mode will free it in the mode_changed case */ - if (!config->mode_changed) - kfree(pipe_config); - intel_update_pipe_size(to_intel_crtc(set->crtc)); if (config->mode_changed) { @@ -11899,6 +12019,8 @@ static int intel_crtc_set_config(struct drm_mode_set *set) fail: intel_set_config_restore_state(dev, config); + drm_atomic_state_clear(state); + /* * HACK: if the pipe was on, but we didn't have a framebuffer, * force the pipe off to avoid oopsing in the modeset code @@ -11911,11 +12033,15 @@ fail: /* Try to restore the config */ if (config->mode_changed && intel_set_mode(save_set.crtc, save_set.mode, - save_set.x, save_set.y, save_set.fb)) + save_set.x, save_set.y, save_set.fb, + state)) DRM_ERROR("failed to restore config after modeset failure\n"); } out_config: + if (state) + drm_atomic_state_free(state); + intel_set_config_free(config); return ret; } @@ -13968,8 +14094,7 @@ void intel_modeset_setup_hw_state(struct drm_device *dev, struct drm_crtc *crtc = dev_priv->pipe_to_crtc_mapping[pipe]; - intel_set_mode(crtc, &crtc->mode, crtc->x, crtc->y, - crtc->primary->fb); + intel_crtc_restore_mode(crtc); } } else { intel_modeset_update_staged_output_state(dev); -- cgit v1.2.1 From db7542dd402ad7be8e337ab483350f8795a3c7b0 Mon Sep 17 00:00:00 2001 From: Ander Conselvan de Oliveira Date: Fri, 20 Mar 2015 16:18:04 +0200 Subject: drm/i915: Allocate a crtc_state also when the crtc is being disabled For consistency, allocate a new crtc_state for a crtc that is being disabled. Previously only the enabled value of the current state would change. v2: Rebase on v5 of previous patch. (Ander) Signed-off-by: Ander Conselvan de Oliveira [danvet: Resolve rebase conflict.] Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_display.c | 30 ++++++++++++++++++++++-------- 1 file changed, 22 insertions(+), 8 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 66cf0996b881..c7e95745ac6b 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -11291,13 +11291,20 @@ intel_modeset_compute_config(struct drm_crtc *crtc, unsigned *prepare_pipes, unsigned *disable_pipes) { + struct drm_device *dev = crtc->dev; struct intel_crtc_state *pipe_config = NULL; + struct intel_crtc *intel_crtc; intel_modeset_affected_pipes(crtc, modeset_pipes, prepare_pipes, disable_pipes); - if ((*modeset_pipes) == 0) - return NULL; + for_each_intel_crtc_masked(dev, *disable_pipes, intel_crtc) { + pipe_config = intel_atomic_get_crtc_state(state, intel_crtc); + if (IS_ERR(pipe_config)) + return pipe_config; + + pipe_config->base.enable = false; + } /* * Note this needs changes when we start tracking multiple modes @@ -11305,14 +11312,21 @@ intel_modeset_compute_config(struct drm_crtc *crtc, * (i.e. one pipe_config for each crtc) rather than just the one * for this crtc. */ - pipe_config = intel_modeset_pipe_config(crtc, fb, mode, state); - if (IS_ERR(pipe_config)) - return pipe_config; + for_each_intel_crtc_masked(dev, *modeset_pipes, intel_crtc) { + /* FIXME: For now we still expect modeset_pipes has at most + * one bit set. */ + if (WARN_ON(&intel_crtc->base != crtc)) + continue; - intel_dump_pipe_config(to_intel_crtc(crtc), pipe_config, - "[modeset]"); + pipe_config = intel_modeset_pipe_config(crtc, fb, mode, state); + if (IS_ERR(pipe_config)) + return pipe_config; - return pipe_config; + intel_dump_pipe_config(to_intel_crtc(crtc), pipe_config, + "[modeset]"); + } + + return intel_atomic_get_crtc_state(state, to_intel_crtc(crtc));; } static int __intel_set_mode_setup_plls(struct drm_device *dev, -- cgit v1.2.1 From 2f0b57901b6cb1a340c16a408d9da968252eed4f Mon Sep 17 00:00:00 2001 From: kbuild test robot Date: Thu, 26 Mar 2015 22:30:21 +0800 Subject: drm/i915/skl: fix semicolon.cocci warnings drivers/gpu/drm/i915/intel_pm.c:2913:4-5: Unneeded semicolon Removes unneeded semicolon. Generated by: scripts/coccinelle/misc/semicolon.cocci CC: Tvrtko Ursulin Signed-off-by: Fengguang Wu Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_pm.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/intel_pm.c b/drivers/gpu/drm/i915/intel_pm.c index 753a3af07333..fa4ccb346389 100644 --- a/drivers/gpu/drm/i915/intel_pm.c +++ b/drivers/gpu/drm/i915/intel_pm.c @@ -2910,7 +2910,7 @@ static bool skl_compute_plane_wm(const struct drm_i915_private *dev_priv, break; case 8: WARN(1, "Unsupported pixel depth for rotation"); - }; + } } y_tile_minimum = plane_blocks_per_line * min_scanlines; selected_result = max(method2, y_tile_minimum); -- cgit v1.2.1 From 989697255d6019df2b8548a072b0a81abd77c6aa Mon Sep 17 00:00:00 2001 From: Ander Conselvan de Oliveira Date: Fri, 20 Mar 2015 16:18:06 +0200 Subject: drm/i915: Implement connector state duplication So that we can add connector states to the drm_atomic_state used in the legacy modeset. Signed-off-by: Ander Conselvan de Oliveira Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_crt.c | 1 + drivers/gpu/drm/i915/intel_dp.c | 1 + drivers/gpu/drm/i915/intel_dp_mst.c | 1 + drivers/gpu/drm/i915/intel_dsi.c | 1 + drivers/gpu/drm/i915/intel_dvo.c | 1 + drivers/gpu/drm/i915/intel_hdmi.c | 1 + drivers/gpu/drm/i915/intel_lvds.c | 1 + drivers/gpu/drm/i915/intel_sdvo.c | 1 + drivers/gpu/drm/i915/intel_tv.c | 1 + 9 files changed, 9 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_crt.c b/drivers/gpu/drm/i915/intel_crt.c index 974534e784ee..573aaff3389a 100644 --- a/drivers/gpu/drm/i915/intel_crt.c +++ b/drivers/gpu/drm/i915/intel_crt.c @@ -794,6 +794,7 @@ static const struct drm_connector_funcs intel_crt_connector_funcs = { .destroy = intel_crt_destroy, .set_property = intel_crt_set_property, .atomic_destroy_state = drm_atomic_helper_connector_destroy_state, + .atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state, .atomic_get_property = intel_connector_atomic_get_property, }; diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 0b26df9c78cb..20088ec5561f 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -4613,6 +4613,7 @@ static const struct drm_connector_funcs intel_dp_connector_funcs = { .atomic_get_property = intel_connector_atomic_get_property, .destroy = intel_dp_connector_destroy, .atomic_destroy_state = drm_atomic_helper_connector_destroy_state, + .atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state, }; static const struct drm_connector_helper_funcs intel_dp_connector_helper_funcs = { diff --git a/drivers/gpu/drm/i915/intel_dp_mst.c b/drivers/gpu/drm/i915/intel_dp_mst.c index 971163e53e59..37c45f8d0135 100644 --- a/drivers/gpu/drm/i915/intel_dp_mst.c +++ b/drivers/gpu/drm/i915/intel_dp_mst.c @@ -327,6 +327,7 @@ static const struct drm_connector_funcs intel_dp_mst_connector_funcs = { .atomic_get_property = intel_connector_atomic_get_property, .destroy = intel_dp_mst_connector_destroy, .atomic_destroy_state = drm_atomic_helper_connector_destroy_state, + .atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state, }; static int intel_dp_mst_get_modes(struct drm_connector *connector) diff --git a/drivers/gpu/drm/i915/intel_dsi.c b/drivers/gpu/drm/i915/intel_dsi.c index c8c8b24e300c..572251e9810b 100644 --- a/drivers/gpu/drm/i915/intel_dsi.c +++ b/drivers/gpu/drm/i915/intel_dsi.c @@ -975,6 +975,7 @@ static const struct drm_connector_funcs intel_dsi_connector_funcs = { .fill_modes = drm_helper_probe_single_connector_modes, .atomic_get_property = intel_connector_atomic_get_property, .atomic_destroy_state = drm_atomic_helper_connector_destroy_state, + .atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state, }; void intel_dsi_init(struct drm_device *dev) diff --git a/drivers/gpu/drm/i915/intel_dvo.c b/drivers/gpu/drm/i915/intel_dvo.c index d8579510beb0..4ccd6c3f133d 100644 --- a/drivers/gpu/drm/i915/intel_dvo.c +++ b/drivers/gpu/drm/i915/intel_dvo.c @@ -393,6 +393,7 @@ static const struct drm_connector_funcs intel_dvo_connector_funcs = { .fill_modes = drm_helper_probe_single_connector_modes, .atomic_get_property = intel_connector_atomic_get_property, .atomic_destroy_state = drm_atomic_helper_connector_destroy_state, + .atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state, }; static const struct drm_connector_helper_funcs intel_dvo_connector_helper_funcs = { diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi.c index 995c5b261f4f..b13a8becaa14 100644 --- a/drivers/gpu/drm/i915/intel_hdmi.c +++ b/drivers/gpu/drm/i915/intel_hdmi.c @@ -1618,6 +1618,7 @@ static const struct drm_connector_funcs intel_hdmi_connector_funcs = { .atomic_get_property = intel_connector_atomic_get_property, .destroy = intel_hdmi_destroy, .atomic_destroy_state = drm_atomic_helper_connector_destroy_state, + .atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state, }; static const struct drm_connector_helper_funcs intel_hdmi_connector_helper_funcs = { diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c index 24e8730dc189..2b008b024891 100644 --- a/drivers/gpu/drm/i915/intel_lvds.c +++ b/drivers/gpu/drm/i915/intel_lvds.c @@ -535,6 +535,7 @@ static const struct drm_connector_funcs intel_lvds_connector_funcs = { .atomic_get_property = intel_connector_atomic_get_property, .destroy = intel_lvds_destroy, .atomic_destroy_state = drm_atomic_helper_connector_destroy_state, + .atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state, }; static const struct drm_encoder_funcs intel_lvds_enc_funcs = { diff --git a/drivers/gpu/drm/i915/intel_sdvo.c b/drivers/gpu/drm/i915/intel_sdvo.c index 9e554c2cfbb4..f5b7e1e7c5e0 100644 --- a/drivers/gpu/drm/i915/intel_sdvo.c +++ b/drivers/gpu/drm/i915/intel_sdvo.c @@ -2194,6 +2194,7 @@ static const struct drm_connector_funcs intel_sdvo_connector_funcs = { .atomic_get_property = intel_connector_atomic_get_property, .destroy = intel_sdvo_destroy, .atomic_destroy_state = drm_atomic_helper_connector_destroy_state, + .atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state, }; static const struct drm_connector_helper_funcs intel_sdvo_connector_helper_funcs = { diff --git a/drivers/gpu/drm/i915/intel_tv.c b/drivers/gpu/drm/i915/intel_tv.c index 4e6684ee8a00..bc1d9d740904 100644 --- a/drivers/gpu/drm/i915/intel_tv.c +++ b/drivers/gpu/drm/i915/intel_tv.c @@ -1516,6 +1516,7 @@ static const struct drm_connector_funcs intel_tv_connector_funcs = { .atomic_get_property = intel_connector_atomic_get_property, .fill_modes = drm_helper_probe_single_connector_modes, .atomic_destroy_state = drm_atomic_helper_connector_destroy_state, + .atomic_duplicate_state = drm_atomic_helper_connector_duplicate_state, }; static const struct drm_connector_helper_funcs intel_tv_connector_helper_funcs = { -- cgit v1.2.1 From d29b2f9dce3e53ed3d5ada9d42f0edcca72fbfcd Mon Sep 17 00:00:00 2001 From: Ander Conselvan de Oliveira Date: Fri, 20 Mar 2015 16:18:05 +0200 Subject: drm/i915: Update dummy connector atomic state with current config Keep that state updated so that we can write code that depends on it on the follow up patches. v2: Fix BUG due to stale connector_state->crtc value. (Chandra) v3: Update comment about dummy state connectors. (Chandra) Signed-off-by: Ander Conselvan de Oliveira Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_display.c | 45 ++++++++++++++++++++++++++++-------- 1 file changed, 36 insertions(+), 9 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index c7e95745ac6b..de2ab2d9574b 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -10217,6 +10217,27 @@ static void intel_modeset_update_staged_output_state(struct drm_device *dev) } } +/* Transitional helper to copy current connector/encoder state to + * connector->state. This is needed so that code that is partially + * converted to atomic does the right thing. + */ +static void intel_modeset_update_connector_atomic_state(struct drm_device *dev) +{ + struct intel_connector *connector; + + for_each_intel_connector(dev, connector) { + if (connector->base.encoder) { + connector->base.state->best_encoder = + connector->base.encoder; + connector->base.state->crtc = + connector->base.encoder->crtc; + } else { + connector->base.state->best_encoder = NULL; + connector->base.state->crtc = NULL; + } + } +} + /** * intel_modeset_commit_output_state * @@ -10240,6 +10261,8 @@ static void intel_modeset_commit_output_state(struct drm_device *dev) crtc->base.state->enable = crtc->new_enabled; crtc->base.enabled = crtc->new_enabled; } + + intel_modeset_update_connector_atomic_state(dev); } static void @@ -13015,19 +13038,21 @@ static void intel_setup_outputs(struct drm_device *dev) * testing/debug of the plane operations (and only when a specific * kernel module option is given), that shouldn't really matter. * + * We are also relying on these states to convert the legacy mode set + * to use a drm_atomic_state struct. The states are kept consistent + * with actual state, so that it is safe to rely on that instead of + * the staged config. + * * Once atomic support for crtc's + connectors lands, this loop should * be removed since we'll be setting up real connector state, which * will contain Intel-specific properties. */ - if (drm_core_check_feature(dev, DRIVER_ATOMIC)) { - list_for_each_entry(connector, - &dev->mode_config.connector_list, - head) { - if (!WARN_ON(connector->state)) { - connector->state = - kzalloc(sizeof(*connector->state), - GFP_KERNEL); - } + list_for_each_entry(connector, + &dev->mode_config.connector_list, + head) { + if (!WARN_ON(connector->state)) { + connector->state = kzalloc(sizeof(*connector->state), + GFP_KERNEL); } } @@ -14080,6 +14105,8 @@ void intel_modeset_setup_hw_state(struct drm_device *dev, "[setup_hw_state]"); } + intel_modeset_update_connector_atomic_state(dev); + for (i = 0; i < dev_priv->num_shared_dpll; i++) { struct intel_shared_dpll *pll = &dev_priv->shared_dplls[i]; -- cgit v1.2.1 From 944b0c76575753da5a332aab0a1d8c6df65a076b Mon Sep 17 00:00:00 2001 From: Ander Conselvan de Oliveira Date: Fri, 20 Mar 2015 16:18:07 +0200 Subject: drm/i915: Copy the staged connector config to the legacy atomic state With this in place, we can start converting pieces of the modeset code to look at the connector atomic state instead of the staged config. v2: Handle the load detect staged config changes too. (Ander) Remove unnecessary blank line. (Daniel) Signed-off-by: Ander Conselvan de Oliveira Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_display.c | 52 +++++++++++++++++++++++++++++++----- 1 file changed, 45 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index de2ab2d9574b..c0fa9df51008 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -8909,6 +8909,7 @@ bool intel_get_load_detect_pipe(struct drm_connector *connector, struct drm_framebuffer *fb; struct drm_mode_config *config = &dev->mode_config; struct drm_atomic_state *state = NULL; + struct drm_connector_state *connector_state; int ret, i = -1; DRM_DEBUG_KMS("[CONNECTOR:%d:%s], [ENCODER:%d:%s]\n", @@ -8996,6 +8997,15 @@ retry: state->acquire_ctx = ctx; + connector_state = drm_atomic_get_connector_state(state, connector); + if (IS_ERR(connector_state)) { + ret = PTR_ERR(connector_state); + goto fail; + } + + connector_state->crtc = crtc; + connector_state->best_encoder = &intel_encoder->base; + if (!mode) mode = &load_detect_mode; @@ -9061,6 +9071,7 @@ void intel_release_load_detect_pipe(struct drm_connector *connector, struct drm_crtc *crtc = encoder->crtc; struct intel_crtc *intel_crtc = to_intel_crtc(crtc); struct drm_atomic_state *state; + struct drm_connector_state *connector_state; DRM_DEBUG_KMS("[CONNECTOR:%d:%s], [ENCODER:%d:%s]\n", connector->base.id, connector->name, @@ -9068,17 +9079,23 @@ void intel_release_load_detect_pipe(struct drm_connector *connector, if (old->load_detect_temp) { state = drm_atomic_state_alloc(dev); - if (!state) { - DRM_DEBUG_KMS("can't release load detect pipe\n"); - return; - } + if (!state) + goto fail; state->acquire_ctx = ctx; + connector_state = drm_atomic_get_connector_state(state, connector); + if (IS_ERR(connector_state)) + goto fail; + to_intel_connector(connector)->new_encoder = NULL; intel_encoder->new_crtc = NULL; intel_crtc->new_enabled = false; intel_crtc->new_config = NULL; + + connector_state->best_encoder = NULL; + connector_state->crtc = NULL; + intel_set_mode(crtc, NULL, 0, 0, NULL, state); drm_atomic_state_free(state); @@ -9094,6 +9111,11 @@ void intel_release_load_detect_pipe(struct drm_connector *connector, /* Switch crtc and encoder back off if necessary */ if (old->dpms_mode != DRM_MODE_DPMS_ON) connector->funcs->dpms(connector, old->dpms_mode); + + return; +fail: + DRM_DEBUG_KMS("Couldn't release load detect pipe.\n"); + drm_atomic_state_free(state); } static int i9xx_pll_refclk(struct drm_device *dev, @@ -11777,9 +11799,11 @@ intel_set_config_compute_mode_changes(struct drm_mode_set *set, static int intel_modeset_stage_output_state(struct drm_device *dev, struct drm_mode_set *set, - struct intel_set_config *config) + struct intel_set_config *config, + struct drm_atomic_state *state) { struct intel_connector *connector; + struct drm_connector_state *connector_state; struct intel_encoder *encoder; struct intel_crtc *crtc; int ro; @@ -11843,6 +11867,14 @@ intel_modeset_stage_output_state(struct drm_device *dev, } connector->new_encoder->new_crtc = to_intel_crtc(new_crtc); + connector_state = + drm_atomic_get_connector_state(state, &connector->base); + if (IS_ERR(connector_state)) + return PTR_ERR(connector_state); + + connector_state->crtc = new_crtc; + connector_state->best_encoder = &connector->new_encoder->base; + DRM_DEBUG_KMS("[CONNECTOR:%d:%s] to [CRTC:%d]\n", connector->base.base.id, connector->base.name, @@ -11875,9 +11907,15 @@ intel_modeset_stage_output_state(struct drm_device *dev, } /* Now we've also updated encoder->new_crtc for all encoders. */ for_each_intel_connector(dev, connector) { - if (connector->new_encoder) + connector_state = + drm_atomic_get_connector_state(state, &connector->base); + + if (connector->new_encoder) { if (connector->new_encoder != connector->encoder) connector->encoder = connector->new_encoder; + } else { + connector_state->crtc = NULL; + } } for_each_intel_crtc(dev, crtc) { crtc->new_enabled = false; @@ -11986,7 +12024,7 @@ static int intel_crtc_set_config(struct drm_mode_set *set) state->acquire_ctx = dev->mode_config.acquire_ctx; - ret = intel_modeset_stage_output_state(dev, set, config); + ret = intel_modeset_stage_output_state(dev, set, config, state); if (ret) goto fail; -- cgit v1.2.1 From 0b9018793939970aad49e4e0ed21f0fd0db74e4a Mon Sep 17 00:00:00 2001 From: Ander Conselvan de Oliveira Date: Fri, 20 Mar 2015 16:18:08 +0200 Subject: drm/i915: Don't use encoder->new_crtc in intel_modeset_pipe_config() Move towards atomic by using the legacy modeset's drm_atomic_state instead. v2: Move call to drm_atomic_add_affected_connectors() to intel_modeset_compute_config(). (Daniel) Signed-off-by: Ander Conselvan de Oliveira [danvet: Resurrect the ret local variable which I've dropped from an earlier patch and which is now needed.] Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_display.c | 18 ++++++++++++++++-- 1 file changed, 16 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index c0fa9df51008..afceb5836af7 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -10538,8 +10538,11 @@ intel_modeset_pipe_config(struct drm_crtc *crtc, { struct drm_device *dev = crtc->dev; struct intel_encoder *encoder; + struct intel_connector *connector; + struct drm_connector_state *connector_state; struct intel_crtc_state *pipe_config; int plane_bpp, ret = -EINVAL; + int i; bool retry = true; if (!check_encoder_cloning(to_intel_crtc(crtc))) { @@ -10613,11 +10616,17 @@ encoder_retry: * adjust it according to limitations or connector properties, and also * a chance to reject the mode entirely. */ - for_each_intel_encoder(dev, encoder) { + for (i = 0; i < state->num_connector; i++) { + connector = to_intel_connector(state->connectors[i]); + if (!connector) + continue; - if (&encoder->new_crtc->base != crtc) + connector_state = state->connector_states[i]; + if (connector_state->crtc != crtc) continue; + encoder = to_intel_encoder(connector_state->best_encoder); + if (!(encoder->compute_config(encoder, pipe_config))) { DRM_DEBUG_KMS("Encoder config failure\n"); goto fail; @@ -11339,6 +11348,11 @@ intel_modeset_compute_config(struct drm_crtc *crtc, struct drm_device *dev = crtc->dev; struct intel_crtc_state *pipe_config = NULL; struct intel_crtc *intel_crtc; + int ret = 0; + + ret = drm_atomic_add_affected_connectors(state, crtc); + if (ret) + return ERR_PTR(ret); intel_modeset_affected_pipes(crtc, modeset_pipes, prepare_pipes, disable_pipes); -- cgit v1.2.1 From 1486017fbfef5441cc61c805ccace4d3dea27c80 Mon Sep 17 00:00:00 2001 From: Ander Conselvan de Oliveira Date: Fri, 20 Mar 2015 16:18:09 +0200 Subject: drm/i915: Don't use encoder->new_crtc in compute_baseline_pipe_bpp() Move towards atomic by using the legacy modeset's drm_atomic_state instead. Signed-off-by: Ander Conselvan de Oliveira [danvet: Keep the if (!connector) continue; separate so that it's easier to eventually extract a for_each_connector_in_state iterator. And because of the upcast it's also safer.] Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_display.c | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index afceb5836af7..034aa7e8e683 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -10319,8 +10319,9 @@ compute_baseline_pipe_bpp(struct intel_crtc *crtc, struct intel_crtc_state *pipe_config) { struct drm_device *dev = crtc->base.dev; + struct drm_atomic_state *state; struct intel_connector *connector; - int bpp; + int bpp, i; switch (fb->pixel_format) { case DRM_FORMAT_C8: @@ -10360,10 +10361,15 @@ compute_baseline_pipe_bpp(struct intel_crtc *crtc, pipe_config->pipe_bpp = bpp; + state = pipe_config->base.state; + /* Clamp display bpp to EDID value */ - for_each_intel_connector(dev, connector) { - if (!connector->new_encoder || - connector->new_encoder->new_crtc != crtc) + for (i = 0; i < state->num_connector; i++) { + if (!state->connectors[i]) + continue; + + connector = to_intel_connector(state->connectors[i]); + if (state->connector_states[i]->crtc != &crtc->base) continue; connected_sink_compute_bpp(connector, pipe_config); -- cgit v1.2.1 From 84556d58efead4cf4e53d807e5c846c7919be4c9 Mon Sep 17 00:00:00 2001 From: Ander Conselvan de Oliveira Date: Fri, 20 Mar 2015 16:18:10 +0200 Subject: drm/i915: Don't depend on encoder->new_crtc in intel_dp_compute_config() Move towards atomic by using the legacy modeset's drm_atomic_state instead. Signed-off-by: Ander Conselvan de Oliveira Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_dp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 20088ec5561f..0286fc6fb89a 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -1323,7 +1323,7 @@ intel_dp_compute_config(struct intel_encoder *encoder, struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode; struct intel_dp *intel_dp = enc_to_intel_dp(&encoder->base); enum port port = dp_to_dig_port(intel_dp)->port; - struct intel_crtc *intel_crtc = encoder->new_crtc; + struct intel_crtc *intel_crtc = to_intel_crtc(pipe_config->base.crtc); struct intel_connector *intel_connector = intel_dp->attached_connector; int lane_count, clock; int min_lane_count = 1; -- cgit v1.2.1 From 77f06c86622e1237fbab3333021211e8b04fe37d Mon Sep 17 00:00:00 2001 From: Ander Conselvan de Oliveira Date: Fri, 20 Mar 2015 16:18:11 +0200 Subject: drm/i915: Don't depend on encoder->new_crtc in intel_hdmi_compute_config Move towards atomic by using the legacy modeset's drm_atomic_state instead. Signed-off-by: Ander Conselvan de Oliveira Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_hdmi.c | 21 ++++++++++++++++----- 1 file changed, 16 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_hdmi.c b/drivers/gpu/drm/i915/intel_hdmi.c index b13a8becaa14..cacbafdad3ab 100644 --- a/drivers/gpu/drm/i915/intel_hdmi.c +++ b/drivers/gpu/drm/i915/intel_hdmi.c @@ -951,19 +951,30 @@ intel_hdmi_mode_valid(struct drm_connector *connector, return MODE_OK; } -static bool hdmi_12bpc_possible(struct intel_crtc *crtc) +static bool hdmi_12bpc_possible(struct intel_crtc_state *crtc_state) { - struct drm_device *dev = crtc->base.dev; + struct drm_device *dev = crtc_state->base.crtc->dev; + struct drm_atomic_state *state; struct intel_encoder *encoder; + struct drm_connector_state *connector_state; int count = 0, count_hdmi = 0; + int i; if (HAS_GMCH_DISPLAY(dev)) return false; - for_each_intel_encoder(dev, encoder) { - if (encoder->new_crtc != crtc) + state = crtc_state->base.state; + + for (i = 0; i < state->num_connector; i++) { + if (!state->connectors[i]) continue; + connector_state = state->connector_states[i]; + if (connector_state->crtc != crtc_state->base.crtc) + continue; + + encoder = to_intel_encoder(connector_state->best_encoder); + count_hdmi += encoder->type == INTEL_OUTPUT_HDMI; count++; } @@ -1020,7 +1031,7 @@ bool intel_hdmi_compute_config(struct intel_encoder *encoder, */ if (pipe_config->pipe_bpp > 8*3 && pipe_config->has_hdmi_sink && clock_12bpc <= portclock_limit && - hdmi_12bpc_possible(encoder->new_crtc)) { + hdmi_12bpc_possible(pipe_config)) { DRM_DEBUG_KMS("picking bpc to 12 for HDMI output\n"); desired_bpp = 12*3; -- cgit v1.2.1 From 3165c074175cddab1dcfd553042ea4f363bc76e7 Mon Sep 17 00:00:00 2001 From: Ander Conselvan de Oliveira Date: Fri, 20 Mar 2015 16:18:12 +0200 Subject: drm/i915: Use atomic state in intel_ddi_crtc_get_new_encoder() Instead of using connector->new_encoder, get the same information from the pipe_config, thus making the function ready for the atomic conversion. Signed-off-by: Ander Conselvan de Oliveira Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_ddi.c | 24 +++++++++++++++--------- 1 file changed, 15 insertions(+), 9 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_ddi.c b/drivers/gpu/drm/i915/intel_ddi.c index 8aee7d77ce9d..47b9307da24b 100644 --- a/drivers/gpu/drm/i915/intel_ddi.c +++ b/drivers/gpu/drm/i915/intel_ddi.c @@ -492,17 +492,23 @@ intel_ddi_get_crtc_encoder(struct drm_crtc *crtc) } static struct intel_encoder * -intel_ddi_get_crtc_new_encoder(struct intel_crtc *crtc) +intel_ddi_get_crtc_new_encoder(struct intel_crtc_state *crtc_state) { - struct drm_device *dev = crtc->base.dev; - struct intel_encoder *intel_encoder, *ret = NULL; + struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc); + struct intel_encoder *ret = NULL; + struct drm_atomic_state *state; int num_encoders = 0; + int i; - for_each_intel_encoder(dev, intel_encoder) { - if (intel_encoder->new_crtc == crtc) { - ret = intel_encoder; - num_encoders++; - } + state = crtc_state->base.state; + + for (i = 0; i < state->num_connector; i++) { + if (!state->connectors[i] || + state->connector_states[i]->crtc != crtc_state->base.crtc) + continue; + + ret = to_intel_encoder(state->connector_states[i]->best_encoder); + num_encoders++; } WARN(num_encoders != 1, "%d encoders on crtc for pipe %c\n", num_encoders, @@ -1216,7 +1222,7 @@ bool intel_ddi_pll_select(struct intel_crtc *intel_crtc, { struct drm_device *dev = intel_crtc->base.dev; struct intel_encoder *intel_encoder = - intel_ddi_get_crtc_new_encoder(intel_crtc); + intel_ddi_get_crtc_new_encoder(crtc_state); int clock = crtc_state->port_clock; if (IS_SKYLAKE(dev)) -- cgit v1.2.1 From e75f4771f81e969d60177cb0c72526f4a05cbb5f Mon Sep 17 00:00:00 2001 From: Ander Conselvan de Oliveira Date: Fri, 20 Mar 2015 16:18:13 +0200 Subject: drm/i915: Don't use staged config in intel_dp_mst_compute_config() Move towards atomic by using the legacy modeset's drm_atomic_state instead. Signed-off-by: Ander Conselvan de Oliveira Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_dp_mst.c | 17 +++++++++++------ 1 file changed, 11 insertions(+), 6 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_dp_mst.c b/drivers/gpu/drm/i915/intel_dp_mst.c index 37c45f8d0135..5329c855acce 100644 --- a/drivers/gpu/drm/i915/intel_dp_mst.c +++ b/drivers/gpu/drm/i915/intel_dp_mst.c @@ -36,11 +36,11 @@ static bool intel_dp_mst_compute_config(struct intel_encoder *encoder, struct intel_dp_mst_encoder *intel_mst = enc_to_mst(&encoder->base); struct intel_digital_port *intel_dig_port = intel_mst->primary; struct intel_dp *intel_dp = &intel_dig_port->dp; - struct drm_device *dev = encoder->base.dev; - int bpp; + struct drm_atomic_state *state; + int bpp, i; int lane_count, slots, rate; struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode; - struct intel_connector *found = NULL, *intel_connector; + struct intel_connector *found = NULL; int mst_pbn; pipe_config->dp_encoder_is_mst = true; @@ -68,9 +68,14 @@ static bool intel_dp_mst_compute_config(struct intel_encoder *encoder, pipe_config->pipe_bpp = 24; pipe_config->port_clock = rate; - for_each_intel_connector(dev, intel_connector) { - if (intel_connector->new_encoder == encoder) { - found = intel_connector; + state = pipe_config->base.state; + + for (i = 0; i < state->num_connector; i++) { + if (!state->connectors[i]) + continue; + + if (state->connector_states[i]->best_encoder == &encoder->base) { + found = to_intel_connector(state->connectors[i]); break; } } -- cgit v1.2.1 From d21bd67bbbc635424b336b7c881dbcdf6098f9f5 Mon Sep 17 00:00:00 2001 From: Ander Conselvan de Oliveira Date: Fri, 20 Mar 2015 16:18:14 +0200 Subject: drm/i915: Don't use encoder->new_crtc in intel_lvds_compute_config() Move towards atomic by using the legacy modeset's drm_atomic_state instead. Signed-off-by: Ander Conselvan de Oliveira Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_lvds.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/intel_lvds.c b/drivers/gpu/drm/i915/intel_lvds.c index 2b008b024891..06d2da336f7c 100644 --- a/drivers/gpu/drm/i915/intel_lvds.c +++ b/drivers/gpu/drm/i915/intel_lvds.c @@ -286,7 +286,7 @@ static bool intel_lvds_compute_config(struct intel_encoder *intel_encoder, struct intel_connector *intel_connector = &lvds_encoder->attached_connector->base; struct drm_display_mode *adjusted_mode = &pipe_config->base.adjusted_mode; - struct intel_crtc *intel_crtc = lvds_encoder->base.new_crtc; + struct intel_crtc *intel_crtc = to_intel_crtc(pipe_config->base.crtc); unsigned int lvds_bpp; /* Should never happen!! */ -- cgit v1.2.1 From 9ffd906d9a6e50c958bd99971d762a426a12a36a Mon Sep 17 00:00:00 2001 From: Dan Carpenter Date: Wed, 11 Feb 2015 13:26:21 +0300 Subject: watchdog: mtk_wdt: signedness bug in mtk_wdt_start() "ret" should be signed for the error handling to work correctly. This doesn't matter much in real life since mtk_wdt_set_timeout() always succeeds. Signed-off-by: Dan Carpenter Reviewed-by: Matthias Brugger Reviewed-by: Guenter Roeck Signed-off-by: Wim Van Sebroeck --- drivers/watchdog/mtk_wdt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/watchdog/mtk_wdt.c b/drivers/watchdog/mtk_wdt.c index a87f6df6e85f..938b987de551 100644 --- a/drivers/watchdog/mtk_wdt.c +++ b/drivers/watchdog/mtk_wdt.c @@ -133,7 +133,7 @@ static int mtk_wdt_start(struct watchdog_device *wdt_dev) u32 reg; struct mtk_wdt_dev *mtk_wdt = watchdog_get_drvdata(wdt_dev); void __iomem *wdt_base = mtk_wdt->wdt_base; - u32 ret; + int ret; ret = mtk_wdt_set_timeout(wdt_dev, wdt_dev->timeout); if (ret < 0) -- cgit v1.2.1 From a629c08fdb98ebb184d745553af9dda4f05941bf Mon Sep 17 00:00:00 2001 From: James Hogan Date: Fri, 20 Feb 2015 23:45:44 +0000 Subject: watchdog: imgpdc: Fix probe NULL pointer dereference The IMG PDC watchdog probe function calls pdc_wdt_stop() prior to watchdog_set_drvdata(), causing a NULL pointer dereference when pdc_wdt_stop() retrieves the struct pdc_wdt_dev pointer using watchdog_get_drvdata() and reads the register base address through it. Fix by moving the watchdog_set_drvdata() call earlier, to where various other pdc_wdt->wdt_dev fields are initialised. Fixes: 93937669e9b5 ("watchdog: ImgTec PDC Watchdog Timer Driver") Signed-off-by: James Hogan Cc: Ezequiel Garcia Cc: Naidu Tellapati Cc: Jude Abraham Cc: linux-watchdog@vger.kernel.org Reviewed-by: Guenter Roeck Signed-off-by: Wim Van Sebroeck --- drivers/watchdog/imgpdc_wdt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/watchdog/imgpdc_wdt.c b/drivers/watchdog/imgpdc_wdt.c index c8def68d9e4c..32c35eb31e65 100644 --- a/drivers/watchdog/imgpdc_wdt.c +++ b/drivers/watchdog/imgpdc_wdt.c @@ -191,6 +191,7 @@ static int pdc_wdt_probe(struct platform_device *pdev) pdc_wdt->wdt_dev.ops = &pdc_wdt_ops; pdc_wdt->wdt_dev.max_timeout = 1 << PDC_WDT_CONFIG_DELAY_MASK; pdc_wdt->wdt_dev.parent = &pdev->dev; + watchdog_set_drvdata(&pdc_wdt->wdt_dev, pdc_wdt); ret = watchdog_init_timeout(&pdc_wdt->wdt_dev, heartbeat, &pdev->dev); if (ret < 0) { @@ -232,7 +233,6 @@ static int pdc_wdt_probe(struct platform_device *pdev) watchdog_set_nowayout(&pdc_wdt->wdt_dev, nowayout); platform_set_drvdata(pdev, pdc_wdt); - watchdog_set_drvdata(&pdc_wdt->wdt_dev, pdc_wdt); ret = watchdog_register_device(&pdc_wdt->wdt_dev); if (ret) -- cgit v1.2.1 From ae6ee2fd47f76db5a1cd02c23378057bd21c2c8d Mon Sep 17 00:00:00 2001 From: James Hogan Date: Fri, 20 Feb 2015 23:45:45 +0000 Subject: watchdog: imgpdc: Fix default heartbeat The IMG PDC watchdog driver heartbeat module parameter has no default so it is initialised to zero. This results in the following warning during probe: imgpdc-wdt 2006000.wdt: Initial timeout out of range! setting max timeout The module parameter description implies that the default value should be PDC_WDT_DEF_TIMEOUT, which isn't yet used, so initialise it to that. Also tweak the heartbeat module parameter description for consistency. Fixes: 93937669e9b5 ("watchdog: ImgTec PDC Watchdog Timer Driver") Signed-off-by: James Hogan Cc: Ezequiel Garcia Cc: Naidu Tellapati Cc: Jude Abraham Cc: linux-watchdog@vger.kernel.org Reviewed-by: Guenter Roeck Signed-off-by: Wim Van Sebroeck --- drivers/watchdog/imgpdc_wdt.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/watchdog/imgpdc_wdt.c b/drivers/watchdog/imgpdc_wdt.c index 32c35eb31e65..0deaa4f971f5 100644 --- a/drivers/watchdog/imgpdc_wdt.c +++ b/drivers/watchdog/imgpdc_wdt.c @@ -42,10 +42,10 @@ #define PDC_WDT_MIN_TIMEOUT 1 #define PDC_WDT_DEF_TIMEOUT 64 -static int heartbeat; +static int heartbeat = PDC_WDT_DEF_TIMEOUT; module_param(heartbeat, int, 0); -MODULE_PARM_DESC(heartbeat, "Watchdog heartbeats in seconds. " - "(default = " __MODULE_STRING(PDC_WDT_DEF_TIMEOUT) ")"); +MODULE_PARM_DESC(heartbeat, "Watchdog heartbeats in seconds " + "(default=" __MODULE_STRING(PDC_WDT_DEF_TIMEOUT) ")"); static bool nowayout = WATCHDOG_NOWAYOUT; module_param(nowayout, bool, 0); -- cgit v1.2.1 From a8265c59e22ae636617019dda602e1913bdaeca7 Mon Sep 17 00:00:00 2001 From: Daniel Vetter Date: Fri, 27 Mar 2015 09:08:04 +0100 Subject: drm/i915: Rip out GET_SPRITE_COLORKEY ioctl It's completely unused and Tommi noticed that the #define is borked since forever. I've done a git search in userspace and only found broken definitions and no users anywhere. Cc: Tommi Rantala Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_dma.c | 2 +- drivers/gpu/drm/i915/intel_drv.h | 2 -- drivers/gpu/drm/i915/intel_sprite.c | 24 ------------------------ 3 files changed, 1 insertion(+), 27 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_dma.c b/drivers/gpu/drm/i915/i915_dma.c index d49ed68f041e..68e0c85a17cf 100644 --- a/drivers/gpu/drm/i915/i915_dma.c +++ b/drivers/gpu/drm/i915/i915_dma.c @@ -1199,7 +1199,7 @@ const struct drm_ioctl_desc i915_ioctls[] = { DRM_IOCTL_DEF_DRV(I915_OVERLAY_PUT_IMAGE, intel_overlay_put_image, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED), DRM_IOCTL_DEF_DRV(I915_OVERLAY_ATTRS, intel_overlay_attrs, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED), DRM_IOCTL_DEF_DRV(I915_SET_SPRITE_COLORKEY, intel_sprite_set_colorkey, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED), - DRM_IOCTL_DEF_DRV(I915_GET_SPRITE_COLORKEY, intel_sprite_get_colorkey, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED), + DRM_IOCTL_DEF_DRV(I915_GET_SPRITE_COLORKEY, drm_noop, DRM_MASTER|DRM_CONTROL_ALLOW|DRM_UNLOCKED), DRM_IOCTL_DEF_DRV(I915_GEM_WAIT, i915_gem_wait_ioctl, DRM_AUTH|DRM_UNLOCKED|DRM_RENDER_ALLOW), DRM_IOCTL_DEF_DRV(I915_GEM_CONTEXT_CREATE, i915_gem_context_create_ioctl, DRM_UNLOCKED|DRM_RENDER_ALLOW), DRM_IOCTL_DEF_DRV(I915_GEM_CONTEXT_DESTROY, i915_gem_context_destroy_ioctl, DRM_UNLOCKED|DRM_RENDER_ALLOW), diff --git a/drivers/gpu/drm/i915/intel_drv.h b/drivers/gpu/drm/i915/intel_drv.h index cb57b9c446f3..6036e3b73b7b 100644 --- a/drivers/gpu/drm/i915/intel_drv.h +++ b/drivers/gpu/drm/i915/intel_drv.h @@ -1282,8 +1282,6 @@ void intel_flush_primary_plane(struct drm_i915_private *dev_priv, int intel_plane_restore(struct drm_plane *plane); int intel_sprite_set_colorkey(struct drm_device *dev, void *data, struct drm_file *file_priv); -int intel_sprite_get_colorkey(struct drm_device *dev, void *data, - struct drm_file *file_priv); bool intel_pipe_update_start(struct intel_crtc *crtc, uint32_t *start_vbl_count); void intel_pipe_update_end(struct intel_crtc *crtc, u32 start_vbl_count); diff --git a/drivers/gpu/drm/i915/intel_sprite.c b/drivers/gpu/drm/i915/intel_sprite.c index f41e872ad858..e9ff6fc61267 100644 --- a/drivers/gpu/drm/i915/intel_sprite.c +++ b/drivers/gpu/drm/i915/intel_sprite.c @@ -1134,30 +1134,6 @@ out_unlock: return ret; } -int intel_sprite_get_colorkey(struct drm_device *dev, void *data, - struct drm_file *file_priv) -{ - struct drm_intel_sprite_colorkey *get = data; - struct drm_plane *plane; - struct intel_plane *intel_plane; - int ret = 0; - - drm_modeset_lock_all(dev); - - plane = drm_plane_find(dev, get->plane_id); - if (!plane) { - ret = -ENOENT; - goto out_unlock; - } - - intel_plane = to_intel_plane(plane); - *get = intel_plane->ckey; - -out_unlock: - drm_modeset_unlock_all(dev); - return ret; -} - int intel_plane_restore(struct drm_plane *plane) { if (!plane->crtc || !plane->state->fb) -- cgit v1.2.1 From 2c60fae1489c70206e66c28d72b69a3e496c313d Mon Sep 17 00:00:00 2001 From: Tommi Rantala Date: Thu, 26 Mar 2015 21:47:16 +0200 Subject: drm/i915: fix definition of the DRM_IOCTL_I915_GET_SPRITE_COLORKEY ioctl Fix definition of the DRM_IOCTL_I915_GET_SPRITE_COLORKEY ioctl, so that it is different from the DRM_IOCTL_I915_SET_SPRITE_COLORKEY ioctl. Note that this is just for accuracy, the ioctl implementation itself is totally unused and already ripped out. Signed-off-by: Tommi Rantala [danvet: Add note that this is a dead ioctl.] Signed-off-by: Daniel Vetter --- include/uapi/drm/i915_drm.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/include/uapi/drm/i915_drm.h b/include/uapi/drm/i915_drm.h index 8d1be9073380..551b6737f5df 100644 --- a/include/uapi/drm/i915_drm.h +++ b/include/uapi/drm/i915_drm.h @@ -270,7 +270,7 @@ typedef struct _drm_i915_sarea { #define DRM_IOCTL_I915_OVERLAY_PUT_IMAGE DRM_IOW(DRM_COMMAND_BASE + DRM_I915_OVERLAY_PUT_IMAGE, struct drm_intel_overlay_put_image) #define DRM_IOCTL_I915_OVERLAY_ATTRS DRM_IOWR(DRM_COMMAND_BASE + DRM_I915_OVERLAY_ATTRS, struct drm_intel_overlay_attrs) #define DRM_IOCTL_I915_SET_SPRITE_COLORKEY DRM_IOWR(DRM_COMMAND_BASE + DRM_I915_SET_SPRITE_COLORKEY, struct drm_intel_sprite_colorkey) -#define DRM_IOCTL_I915_GET_SPRITE_COLORKEY DRM_IOWR(DRM_COMMAND_BASE + DRM_I915_SET_SPRITE_COLORKEY, struct drm_intel_sprite_colorkey) +#define DRM_IOCTL_I915_GET_SPRITE_COLORKEY DRM_IOWR(DRM_COMMAND_BASE + DRM_I915_GET_SPRITE_COLORKEY, struct drm_intel_sprite_colorkey) #define DRM_IOCTL_I915_GEM_WAIT DRM_IOWR(DRM_COMMAND_BASE + DRM_I915_GEM_WAIT, struct drm_i915_gem_wait) #define DRM_IOCTL_I915_GEM_CONTEXT_CREATE DRM_IOWR (DRM_COMMAND_BASE + DRM_I915_GEM_CONTEXT_CREATE, struct drm_i915_gem_context_create) #define DRM_IOCTL_I915_GEM_CONTEXT_DESTROY DRM_IOW (DRM_COMMAND_BASE + DRM_I915_GEM_CONTEXT_DESTROY, struct drm_i915_gem_context_destroy) -- cgit v1.2.1 From fdc454c1484a20e1345cf4e4d7a9feaee814147f Mon Sep 17 00:00:00 2001 From: Michel Thierry Date: Tue, 24 Mar 2015 15:46:19 +0000 Subject: drm/i915: Prevent out of range pt in gen6_for_each_pde Found by static analysis tool, this was harmless as the pt was not used out of scope though. Introduced by commit 678d96fbb3b5995a2fdff2bca5e1ab4a40b7e968 ("drm/i915: Track GEN6 page table usage"). Cc: Mika Kuoppala Signed-off-by: Michel Thierry Reviewed-by: Mika Kuoppala Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_gem_gtt.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.h b/drivers/gpu/drm/i915/i915_gem_gtt.h index 0dad42634fd5..3d873467377e 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.h +++ b/drivers/gpu/drm/i915/i915_gem_gtt.h @@ -339,9 +339,9 @@ struct i915_hw_ppgtt { * XXX: temp is not actually needed, but it saves doing the ALIGN operation. */ #define gen6_for_each_pde(pt, pd, start, length, temp, iter) \ - for (iter = gen6_pde_index(start), pt = (pd)->page_table[iter]; \ - length > 0 && iter < I915_PDES; \ - pt = (pd)->page_table[++iter], \ + for (iter = gen6_pde_index(start); \ + pt = (pd)->page_table[iter], length > 0 && iter < I915_PDES; \ + iter++, \ temp = ALIGN(start+1, 1 << GEN6_PDE_SHIFT) - start, \ temp = min_t(unsigned, temp, length), \ start += temp, length -= temp) -- cgit v1.2.1 From 1266cdb1c236eabd49cbf1e45f401159fb4c21d1 Mon Sep 17 00:00:00 2001 From: Michel Thierry Date: Tue, 24 Mar 2015 17:06:33 +0000 Subject: drm/i915: Fix i915_dma_map_single positive error code i915_dma_map_single relies on dma_mapping_error, which returns positive error codes. Found by static checker. Introduced by commit 678d96fbb3b5995a2fdff2bca5e1ab4a40b7e968 ("drm/i915: Track GEN6 page table usage"). v2: Return negative error code and renamed commit title. (Dan) v3: Missing reported-by tag (Daniel) Reported-by: Dan Carpenter Cc: Dan Carpenter Cc: Mika Kuoppala Cc: Daniel Vetter Signed-off-by: Michel Thierry Reviewed-by: Mika Kuoppala Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_gem_gtt.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c index 9903bb0097a6..9c155d1f852f 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c @@ -311,7 +311,10 @@ static inline int i915_dma_map_page_single(struct page *page, struct device *device = &dev->pdev->dev; *daddr = dma_map_page(device, page, 0, 4096, PCI_DMA_BIDIRECTIONAL); - return dma_mapping_error(device, *daddr); + if (dma_mapping_error(device, *daddr)) + return -ENOMEM; + + return 0; } static void unmap_and_free_pt(struct i915_page_table_entry *pt, -- cgit v1.2.1 From 59568eb59a20c743623de9c3247d4e22af1a3573 Mon Sep 17 00:00:00 2001 From: Michel Thierry Date: Tue, 24 Mar 2015 15:46:21 +0000 Subject: drm/i915: Remove unnecessary gen6_ppgtt_unmap_pages We are already unmapping them in gen6_ppgtt_free. This function became redundant since commit 06fda602dbca9c59d87db7da71192e4b54c9f5ff ("drm/i915: Create page table allocators"). Cc: Mika Kuoppala Signed-off-by: Michel Thierry Reviewed-by: Mika Kuoppala Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_gem_gtt.c | 11 ----------- 1 file changed, 11 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c index 9c155d1f852f..a322e31655b5 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c @@ -1155,16 +1155,6 @@ static void gen6_ppgtt_insert_entries(struct i915_address_space *vm, kunmap_atomic(pt_vaddr); } -static void gen6_ppgtt_unmap_pages(struct i915_hw_ppgtt *ppgtt) -{ - int i; - - for (i = 0; i < ppgtt->num_pd_entries; i++) - pci_unmap_page(ppgtt->base.dev->pdev, - ppgtt->pd.page_table[i]->daddr, - 4096, PCI_DMA_BIDIRECTIONAL); -} - /* PDE TLBs are a pain invalidate pre GEN8. It requires a context reload. If we * are switching between contexts with the same LRCA, we also must do a force * restore. @@ -1215,7 +1205,6 @@ static void gen6_ppgtt_cleanup(struct i915_address_space *vm) drm_mm_remove_node(&ppgtt->node); - gen6_ppgtt_unmap_pages(ppgtt); gen6_ppgtt_free(ppgtt); } -- cgit v1.2.1 From 4933d51955aca6d2b05da5a7afe25f156016dfa8 Mon Sep 17 00:00:00 2001 From: Michel Thierry Date: Tue, 24 Mar 2015 15:46:22 +0000 Subject: drm/i915: Finish gen6/7 dynamic page table allocation This patch continues on the idea from "Track GEN6 page table usage". From here on, in the steady state, PDEs are all pointing to the scratch page table (as recommended in the spec). When an object is allocated in the VA range, the code will determine if we need to allocate a page for the page table. Similarly when the object is destroyed, we will remove, and free the page table pointing the PDE back to the scratch page. Following patches will work to unify the code a bit as we bring in GEN8 support. GEN6 and GEN8 are different enough that I had a hard time to get to this point with as much common code as I do. The aliasing PPGTT must pre-allocate all of the page tables. There are a few reasons for this. Two trivial ones: aliasing ppgtt goes through the ggtt paths, so it's hard to maintain, we currently do not restore the default context (assuming the previous force reload is indeed necessary). Most importantly though, the only way (it seems from empirical evidence) to invalidate the CS TLBs on non-render ring is to either use ring sync (which requires actually stopping the rings in order to synchronize when the sync completes vs. where you are in execution), or to reload DCLV. Since without full PPGTT we do not ever reload the DCLV register, there is no good way to achieve this. The simplest solution is just to not support dynamic page table creation/destruction in the aliasing PPGTT. We could always reload DCLV, but this seems like quite a bit of excess overhead only to save at most 2MB-4k of memory for the aliasing PPGTT page tables. v2: Make the page table bitmap declared inside the function (Chris) Simplify the way scratching address space works. Move the alloc/teardown tracepoints up a level in the call stack so that both all implementations get the trace. v3: Updated trace event to spit out a name v4: Aliasing ppgtt is now initialized differently (in setup global gtt) v5: Rebase to latest code. Also removed unnecessary aliasing ppgtt check for trace, as it is no longer possible after the PPGTT cleanup patch series of a couple of months ago (Daniel). v6: Implement changes from code review (Daniel): - allocate/teardown_va_range calls added. - Add a scratch page allocation helper (only need the address). - Move trace events to a new patch. - Use updated mark_tlbs_dirty. - Moved pt preallocation for aliasing ppgtt into gen6_ppgtt_init. v7: teardown_va_range removed (Daniel). In init, gen6_ppgtt_clear_range call is only needed for aliasing ppgtt. v8: Rebase after s/page_tables/page_table/. v9: Remove unnecessary scratch flag in page_table struct, future patches can just compare against ppgtt->scratch_pt, and alloc_pt_scratch becomes redundant. Initialize scratch_pt and pt. (Mika) v10: Clean up aliasing ppgtt init error path and prevent leaking the ppgtt obj when init fails. (Mika) Updated commit author. (Daniel) Cc: Mika Kuoppala Signed-off-by: Ben Widawsky Signed-off-by: Michel Thierry (v4+) Reviewed-by: Mika Kuoppala Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_gem_gtt.c | 146 +++++++++++++++++++++++++++++++----- drivers/gpu/drm/i915/i915_gem_gtt.h | 2 + 2 files changed, 129 insertions(+), 19 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c index a322e31655b5..641d3f45aa1e 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c @@ -346,7 +346,7 @@ static struct i915_page_table_entry *alloc_pt_single(struct drm_device *dev) if (!pt->used_ptes) goto fail_bitmap; - pt->page = alloc_page(GFP_KERNEL | __GFP_ZERO); + pt->page = alloc_page(GFP_KERNEL); if (!pt->page) goto fail_page; @@ -381,7 +381,7 @@ fail_bitmap: * Return: 0 if allocation succeeded. */ static int alloc_pt_range(struct i915_page_directory_entry *pd, uint16_t pde, size_t count, - struct drm_device *dev) + struct drm_device *dev) { int i, ret; @@ -1165,13 +1165,70 @@ static inline void mark_tlbs_dirty(struct i915_hw_ppgtt *ppgtt) ppgtt->pd_dirty_rings = INTEL_INFO(ppgtt->base.dev)->ring_mask; } +static void gen6_initialize_pt(struct i915_address_space *vm, + struct i915_page_table_entry *pt) +{ + gen6_pte_t *pt_vaddr, scratch_pte; + int i; + + WARN_ON(vm->scratch.addr == 0); + + scratch_pte = vm->pte_encode(vm->scratch.addr, + I915_CACHE_LLC, true, 0); + + pt_vaddr = kmap_atomic(pt->page); + + for (i = 0; i < GEN6_PTES; i++) + pt_vaddr[i] = scratch_pte; + + kunmap_atomic(pt_vaddr); +} + static int gen6_alloc_va_range(struct i915_address_space *vm, uint64_t start, uint64_t length) { + DECLARE_BITMAP(new_page_tables, I915_PDES); + struct drm_device *dev = vm->dev; + struct drm_i915_private *dev_priv = dev->dev_private; struct i915_hw_ppgtt *ppgtt = container_of(vm, struct i915_hw_ppgtt, base); struct i915_page_table_entry *pt; + const uint32_t start_save = start, length_save = length; uint32_t pde, temp; + int ret; + + WARN_ON(upper_32_bits(start)); + + bitmap_zero(new_page_tables, I915_PDES); + + /* The allocation is done in two stages so that we can bail out with + * minimal amount of pain. The first stage finds new page tables that + * need allocation. The second stage marks use ptes within the page + * tables. + */ + gen6_for_each_pde(pt, &ppgtt->pd, start, length, temp, pde) { + if (pt != ppgtt->scratch_pt) { + WARN_ON(bitmap_empty(pt->used_ptes, GEN6_PTES)); + continue; + } + + /* We've already allocated a page table */ + WARN_ON(!bitmap_empty(pt->used_ptes, GEN6_PTES)); + + pt = alloc_pt_single(dev); + if (IS_ERR(pt)) { + ret = PTR_ERR(pt); + goto unwind_out; + } + + gen6_initialize_pt(vm, pt); + + ppgtt->pd.page_table[pde] = pt; + set_bit(pde, new_page_tables); + } + + start = start_save; + length = length_save; gen6_for_each_pde(pt, &ppgtt->pd, start, length, temp, pde) { DECLARE_BITMAP(tmp_bitmap, GEN6_PTES); @@ -1180,21 +1237,46 @@ static int gen6_alloc_va_range(struct i915_address_space *vm, bitmap_set(tmp_bitmap, gen6_pte_index(start), gen6_pte_count(start, length)); - bitmap_or(pt->used_ptes, pt->used_ptes, tmp_bitmap, + if (test_and_clear_bit(pde, new_page_tables)) + gen6_write_pde(&ppgtt->pd, pde, pt); + + bitmap_or(pt->used_ptes, tmp_bitmap, pt->used_ptes, GEN6_PTES); } + WARN_ON(!bitmap_empty(new_page_tables, I915_PDES)); + + /* Make sure write is complete before other code can use this page + * table. Also require for WC mapped PTEs */ + readl(dev_priv->gtt.gsm); + mark_tlbs_dirty(ppgtt); return 0; + +unwind_out: + for_each_set_bit(pde, new_page_tables, I915_PDES) { + struct i915_page_table_entry *pt = ppgtt->pd.page_table[pde]; + + ppgtt->pd.page_table[pde] = ppgtt->scratch_pt; + unmap_and_free_pt(pt, vm->dev); + } + + mark_tlbs_dirty(ppgtt); + return ret; } static void gen6_ppgtt_free(struct i915_hw_ppgtt *ppgtt) { int i; - for (i = 0; i < ppgtt->num_pd_entries; i++) - unmap_and_free_pt(ppgtt->pd.page_table[i], ppgtt->base.dev); + for (i = 0; i < ppgtt->num_pd_entries; i++) { + struct i915_page_table_entry *pt = ppgtt->pd.page_table[i]; + + if (pt != ppgtt->scratch_pt) + unmap_and_free_pt(ppgtt->pd.page_table[i], ppgtt->base.dev); + } + unmap_and_free_pt(ppgtt->scratch_pt, ppgtt->base.dev); unmap_and_free_pd(&ppgtt->pd); } @@ -1220,6 +1302,12 @@ static int gen6_ppgtt_allocate_page_directories(struct i915_hw_ppgtt *ppgtt) * size. We allocate at the top of the GTT to avoid fragmentation. */ BUG_ON(!drm_mm_initialized(&dev_priv->gtt.base.mm)); + ppgtt->scratch_pt = alloc_pt_single(ppgtt->base.dev); + if (IS_ERR(ppgtt->scratch_pt)) + return PTR_ERR(ppgtt->scratch_pt); + + gen6_initialize_pt(&ppgtt->base, ppgtt->scratch_pt); + alloc: ret = drm_mm_insert_node_in_range_generic(&dev_priv->gtt.base.mm, &ppgtt->node, GEN6_PD_SIZE, @@ -1250,6 +1338,7 @@ alloc: return 0; err_out: + unmap_and_free_pt(ppgtt->scratch_pt, ppgtt->base.dev); return ret; } @@ -1261,18 +1350,20 @@ static int gen6_ppgtt_alloc(struct i915_hw_ppgtt *ppgtt) if (ret) return ret; - ret = alloc_pt_range(&ppgtt->pd, 0, ppgtt->num_pd_entries, - ppgtt->base.dev); + return 0; +} - if (ret) { - drm_mm_remove_node(&ppgtt->node); - return ret; - } +static void gen6_scratch_va_range(struct i915_hw_ppgtt *ppgtt, + uint64_t start, uint64_t length) +{ + struct i915_page_table_entry *unused; + uint32_t pde, temp; - return 0; + gen6_for_each_pde(unused, &ppgtt->pd, start, length, temp, pde) + ppgtt->pd.page_table[pde] = ppgtt->scratch_pt; } -static int gen6_ppgtt_init(struct i915_hw_ppgtt *ppgtt) +static int gen6_ppgtt_init(struct i915_hw_ppgtt *ppgtt, bool aliasing) { struct drm_device *dev = ppgtt->base.dev; struct drm_i915_private *dev_priv = dev->dev_private; @@ -1295,6 +1386,17 @@ static int gen6_ppgtt_init(struct i915_hw_ppgtt *ppgtt) if (ret) return ret; + if (aliasing) { + /* preallocate all pts */ + ret = alloc_pt_range(&ppgtt->pd, 0, ppgtt->num_pd_entries, + ppgtt->base.dev); + + if (ret) { + gen6_ppgtt_cleanup(&ppgtt->base); + return ret; + } + } + ppgtt->base.allocate_va_range = gen6_alloc_va_range; ppgtt->base.clear_range = gen6_ppgtt_clear_range; ppgtt->base.insert_entries = gen6_ppgtt_insert_entries; @@ -1309,7 +1411,10 @@ static int gen6_ppgtt_init(struct i915_hw_ppgtt *ppgtt) ppgtt->pd_addr = (gen6_pte_t __iomem *)dev_priv->gtt.gsm + ppgtt->pd.pd_offset / sizeof(gen6_pte_t); - ppgtt->base.clear_range(&ppgtt->base, 0, ppgtt->base.total, true); + if (aliasing) + ppgtt->base.clear_range(&ppgtt->base, 0, ppgtt->base.total, true); + else + gen6_scratch_va_range(ppgtt, 0, ppgtt->base.total); gen6_write_page_range(dev_priv, &ppgtt->pd, 0, ppgtt->base.total); @@ -1323,7 +1428,8 @@ static int gen6_ppgtt_init(struct i915_hw_ppgtt *ppgtt) return 0; } -static int __hw_ppgtt_init(struct drm_device *dev, struct i915_hw_ppgtt *ppgtt) +static int __hw_ppgtt_init(struct drm_device *dev, struct i915_hw_ppgtt *ppgtt, + bool aliasing) { struct drm_i915_private *dev_priv = dev->dev_private; @@ -1331,7 +1437,7 @@ static int __hw_ppgtt_init(struct drm_device *dev, struct i915_hw_ppgtt *ppgtt) ppgtt->base.scratch = dev_priv->gtt.base.scratch; if (INTEL_INFO(dev)->gen < 8) - return gen6_ppgtt_init(ppgtt); + return gen6_ppgtt_init(ppgtt, aliasing); else return gen8_ppgtt_init(ppgtt, dev_priv->gtt.base.total); } @@ -1340,7 +1446,7 @@ int i915_ppgtt_init(struct drm_device *dev, struct i915_hw_ppgtt *ppgtt) struct drm_i915_private *dev_priv = dev->dev_private; int ret = 0; - ret = __hw_ppgtt_init(dev, ppgtt); + ret = __hw_ppgtt_init(dev, ppgtt, false); if (ret == 0) { kref_init(&ppgtt->ref); drm_mm_init(&ppgtt->base.mm, ppgtt->base.start, @@ -1975,9 +2081,11 @@ static int i915_gem_setup_global_gtt(struct drm_device *dev, if (!ppgtt) return -ENOMEM; - ret = __hw_ppgtt_init(dev, ppgtt); - if (ret != 0) + ret = __hw_ppgtt_init(dev, ppgtt, true); + if (ret) { + kfree(ppgtt); return ret; + } dev_priv->mm.aliasing_ppgtt = ppgtt; } diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.h b/drivers/gpu/drm/i915/i915_gem_gtt.h index 3d873467377e..3f0ad9f25441 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.h +++ b/drivers/gpu/drm/i915/i915_gem_gtt.h @@ -320,6 +320,8 @@ struct i915_hw_ppgtt { struct i915_page_directory_entry pd; }; + struct i915_page_table_entry *scratch_pt; + struct drm_i915_file_private *file_priv; gen6_pte_t __iomem *pd_addr; -- cgit v1.2.1 From 72744cb13c9ccc248331462f35468398160338bf Mon Sep 17 00:00:00 2001 From: Michel Thierry Date: Tue, 24 Mar 2015 15:46:23 +0000 Subject: drm/i915: Add dynamic page trace events Traces for page directories and tables allocation and map. v2: Removed references to teardown. v3: bitmap_scnprintf has been deprecated. v4: Replace bitmap_scnprintf with scnprintf correctly, and get right range lengths. (Mika) Cc: Mika Kuoppala Signed-off-by: Michel Thierry Reviewed-by: Mika Kuoppala Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_gem.c | 2 + drivers/gpu/drm/i915/i915_gem_gtt.c | 5 ++ drivers/gpu/drm/i915/i915_trace.h | 99 +++++++++++++++++++++++++++++++++++++ 3 files changed, 106 insertions(+) diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index 538cae2ca94e..c13e65994654 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -3512,6 +3512,8 @@ search_free: /* allocate before insert / bind */ if (vma->vm->allocate_va_range) { + trace_i915_va_alloc(vma->vm, vma->node.start, vma->node.size, + VM_TO_TRACE_NAME(vma->vm)); ret = vma->vm->allocate_va_range(vma->vm, vma->node.start, vma->node.size); diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c index 641d3f45aa1e..e33b1214c4d8 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c @@ -1225,6 +1225,7 @@ static int gen6_alloc_va_range(struct i915_address_space *vm, ppgtt->pd.page_table[pde] = pt; set_bit(pde, new_page_tables); + trace_i915_page_table_entry_alloc(vm, pde, start, GEN6_PDE_SHIFT); } start = start_save; @@ -1240,6 +1241,10 @@ static int gen6_alloc_va_range(struct i915_address_space *vm, if (test_and_clear_bit(pde, new_page_tables)) gen6_write_pde(&ppgtt->pd, pde, pt); + trace_i915_page_table_entry_map(vm, pde, pt, + gen6_pte_index(start), + gen6_pte_count(start, length), + GEN6_PTES); bitmap_or(pt->used_ptes, tmp_bitmap, pt->used_ptes, GEN6_PTES); } diff --git a/drivers/gpu/drm/i915/i915_trace.h b/drivers/gpu/drm/i915/i915_trace.h index f004d3d89b87..b3070a4501ab 100644 --- a/drivers/gpu/drm/i915/i915_trace.h +++ b/drivers/gpu/drm/i915/i915_trace.h @@ -156,6 +156,105 @@ TRACE_EVENT(i915_vma_unbind, __entry->obj, __entry->offset, __entry->size, __entry->vm) ); +#define VM_TO_TRACE_NAME(vm) \ + (i915_is_ggtt(vm) ? "G" : \ + "P") + +DECLARE_EVENT_CLASS(i915_va, + TP_PROTO(struct i915_address_space *vm, u64 start, u64 length, const char *name), + TP_ARGS(vm, start, length, name), + + TP_STRUCT__entry( + __field(struct i915_address_space *, vm) + __field(u64, start) + __field(u64, end) + __string(name, name) + ), + + TP_fast_assign( + __entry->vm = vm; + __entry->start = start; + __entry->end = start + length - 1; + __assign_str(name, name); + ), + + TP_printk("vm=%p (%s), 0x%llx-0x%llx", + __entry->vm, __get_str(name), __entry->start, __entry->end) +); + +DEFINE_EVENT(i915_va, i915_va_alloc, + TP_PROTO(struct i915_address_space *vm, u64 start, u64 length, const char *name), + TP_ARGS(vm, start, length, name) +); + +DECLARE_EVENT_CLASS(i915_page_table_entry, + TP_PROTO(struct i915_address_space *vm, u32 pde, u64 start, u64 pde_shift), + TP_ARGS(vm, pde, start, pde_shift), + + TP_STRUCT__entry( + __field(struct i915_address_space *, vm) + __field(u32, pde) + __field(u64, start) + __field(u64, end) + ), + + TP_fast_assign( + __entry->vm = vm; + __entry->pde = pde; + __entry->start = start; + __entry->end = ((start + (1ULL << pde_shift)) & ~((1ULL << pde_shift)-1)) - 1; + ), + + TP_printk("vm=%p, pde=%d (0x%llx-0x%llx)", + __entry->vm, __entry->pde, __entry->start, __entry->end) +); + +DEFINE_EVENT(i915_page_table_entry, i915_page_table_entry_alloc, + TP_PROTO(struct i915_address_space *vm, u32 pde, u64 start, u64 pde_shift), + TP_ARGS(vm, pde, start, pde_shift) +); + +/* Avoid extra math because we only support two sizes. The format is defined by + * bitmap_scnprintf. Each 32 bits is 8 HEX digits followed by comma */ +#define TRACE_PT_SIZE(bits) \ + ((((bits) == 1024) ? 288 : 144) + 1) + +DECLARE_EVENT_CLASS(i915_page_table_entry_update, + TP_PROTO(struct i915_address_space *vm, u32 pde, + struct i915_page_table_entry *pt, u32 first, u32 count, u32 bits), + TP_ARGS(vm, pde, pt, first, count, bits), + + TP_STRUCT__entry( + __field(struct i915_address_space *, vm) + __field(u32, pde) + __field(u32, first) + __field(u32, last) + __dynamic_array(char, cur_ptes, TRACE_PT_SIZE(bits)) + ), + + TP_fast_assign( + __entry->vm = vm; + __entry->pde = pde; + __entry->first = first; + __entry->last = first + count - 1; + scnprintf(__get_str(cur_ptes), + TRACE_PT_SIZE(bits), + "%*pb", + bits, + pt->used_ptes); + ), + + TP_printk("vm=%p, pde=%d, updating %u:%u\t%s", + __entry->vm, __entry->pde, __entry->last, __entry->first, + __get_str(cur_ptes)) +); + +DEFINE_EVENT(i915_page_table_entry_update, i915_page_table_entry_map, + TP_PROTO(struct i915_address_space *vm, u32 pde, + struct i915_page_table_entry *pt, u32 first, u32 count, u32 bits), + TP_ARGS(vm, pde, pt, first, count, bits) +); + TRACE_EVENT(i915_gem_object_change_domain, TP_PROTO(struct drm_i915_gem_object *obj, u32 old_read, u32 old_write), TP_ARGS(obj, old_read, old_write), -- cgit v1.2.1 From 679dacd430cac3f58fc80db69c1694621ede00fc Mon Sep 17 00:00:00 2001 From: Ander Conselvan de Oliveira Date: Fri, 20 Mar 2015 16:18:15 +0200 Subject: drm/i915: Pass an atomic state to modeset_global_resources() functions Follow up patches will convert some functions called from there to use the atomic state, instead of directly accessing the new or current config. This patch just changes the parameters, but shouldn't have any functional changes. Signed-off-by: Ander Conselvan de Oliveira Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_drv.h | 2 +- drivers/gpu/drm/i915/intel_display.c | 10 ++++++---- 2 files changed, 7 insertions(+), 5 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 256369f29975..f80edab9e6cc 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -555,7 +555,7 @@ struct drm_i915_display_funcs { struct drm_crtc *crtc, uint32_t sprite_width, uint32_t sprite_height, int pixel_size, bool enable, bool scaled); - void (*modeset_global_resources)(struct drm_device *dev); + void (*modeset_global_resources)(struct drm_atomic_state *state); /* Returns the active state of the crtc, and if the crtc is active, * fills out the pipe-config with the hw state. */ bool (*get_pipe_config)(struct intel_crtc *, diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 034aa7e8e683..871c61b1f7af 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -4930,8 +4930,9 @@ static unsigned long get_crtc_power_domains(struct drm_crtc *crtc) return mask; } -static void modeset_update_crtc_power_domains(struct drm_device *dev) +static void modeset_update_crtc_power_domains(struct drm_atomic_state *state) { + struct drm_device *dev = state->dev; struct drm_i915_private *dev_priv = dev->dev_private; unsigned long pipe_domains[I915_MAX_PIPES] = { 0, }; struct intel_crtc *crtc; @@ -4953,7 +4954,7 @@ static void modeset_update_crtc_power_domains(struct drm_device *dev) } if (dev_priv->display.modeset_global_resources) - dev_priv->display.modeset_global_resources(dev); + dev_priv->display.modeset_global_resources(state); for_each_intel_crtc(dev, crtc) { enum intel_display_power_domain domain; @@ -5201,8 +5202,9 @@ static void vlv_program_pfi_credits(struct drm_i915_private *dev_priv) WARN_ON(I915_READ(GCI_CONTROL) & PFI_CREDIT_RESEND); } -static void valleyview_modeset_global_resources(struct drm_device *dev) +static void valleyview_modeset_global_resources(struct drm_atomic_state *state) { + struct drm_device *dev = state->dev; struct drm_i915_private *dev_priv = dev->dev_private; int max_pixclk = intel_mode_max_pixclk(dev_priv); int req_cdclk = valleyview_calc_cdclk(dev_priv, max_pixclk); @@ -11506,7 +11508,7 @@ static int __intel_set_mode(struct drm_crtc *crtc, * update the the output configuration. */ intel_modeset_update_state(dev, prepare_pipes); - modeset_update_crtc_power_domains(dev); + modeset_update_crtc_power_domains(pipe_config->base.state); /* Set up the DPLL and any encoders state that needs to adjust or depend * on the DPLL. -- cgit v1.2.1 From a93e255f819cae49dff7319b6686f718b8069fec Mon Sep 17 00:00:00 2001 From: Ander Conselvan de Oliveira Date: Fri, 20 Mar 2015 16:18:17 +0200 Subject: drm/i915: Convert intel_pipe_will_have_type() to using atomic state Pass a crtc_state to it and find whether the pipe has an encoder of a given type by looking at the drm_atomic_state the crtc_state points to. Until recently i9xx_get_refclk() used to be called indirectly from vlv_force_pll_on() with a dummy crtc_state. That dummy crtc state is not converted to be part of a full drm atomic state, so add a WARN in case someone decides to call that again with a such dummy state. This was removed in commit 9cbe40c15a753e02f5da16f6de901decf3276cf1 Author: Vijay Purushothaman Date: Thu Mar 5 19:33:08 2015 +0530 drm/i915: Update prop, int co-eff and gain threshold for CHV v2: Warn if there is no connectors for a given crtc. (Daniel) Replace comment i9xx_get_refclk() with a WARN_ON(). (Ander) Signed-off-by: Ander Conselvan de Oliveira [danvet: Add commit reference for when i9xx_get_refclk was removed.] Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_drv.h | 2 +- drivers/gpu/drm/i915/intel_display.c | 133 +++++++++++++++++++++-------------- 2 files changed, 83 insertions(+), 52 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index f80edab9e6cc..a4db7a7db444 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -546,7 +546,7 @@ struct drm_i915_display_funcs { * Returns true on success, false on failure. */ bool (*find_dpll)(const struct intel_limit *limit, - struct intel_crtc *crtc, + struct intel_crtc_state *crtc_state, int target, int refclk, struct dpll *match_clock, struct dpll *best_clock); diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 871c61b1f7af..1c4efeee13c8 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -431,25 +431,41 @@ bool intel_pipe_has_type(struct intel_crtc *crtc, enum intel_output_type type) * intel_pipe_has_type() but looking at encoder->new_crtc instead of * encoder->crtc. */ -static bool intel_pipe_will_have_type(struct intel_crtc *crtc, int type) +static bool intel_pipe_will_have_type(const struct intel_crtc_state *crtc_state, + int type) { - struct drm_device *dev = crtc->base.dev; + struct drm_atomic_state *state = crtc_state->base.state; + struct drm_connector_state *connector_state; struct intel_encoder *encoder; + int i, num_connectors = 0; + + for (i = 0; i < state->num_connector; i++) { + if (!state->connectors[i]) + continue; + + connector_state = state->connector_states[i]; + if (connector_state->crtc != crtc_state->base.crtc) + continue; + + num_connectors++; - for_each_intel_encoder(dev, encoder) - if (encoder->new_crtc == crtc && encoder->type == type) + encoder = to_intel_encoder(connector_state->best_encoder); + if (encoder->type == type) return true; + } + + WARN_ON(num_connectors == 0); return false; } -static const intel_limit_t *intel_ironlake_limit(struct intel_crtc *crtc, - int refclk) +static const intel_limit_t * +intel_ironlake_limit(struct intel_crtc_state *crtc_state, int refclk) { - struct drm_device *dev = crtc->base.dev; + struct drm_device *dev = crtc_state->base.crtc->dev; const intel_limit_t *limit; - if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS)) { + if (intel_pipe_will_have_type(crtc_state, INTEL_OUTPUT_LVDS)) { if (intel_is_dual_link_lvds(dev)) { if (refclk == 100000) limit = &intel_limits_ironlake_dual_lvds_100m; @@ -467,20 +483,21 @@ static const intel_limit_t *intel_ironlake_limit(struct intel_crtc *crtc, return limit; } -static const intel_limit_t *intel_g4x_limit(struct intel_crtc *crtc) +static const intel_limit_t * +intel_g4x_limit(struct intel_crtc_state *crtc_state) { - struct drm_device *dev = crtc->base.dev; + struct drm_device *dev = crtc_state->base.crtc->dev; const intel_limit_t *limit; - if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS)) { + if (intel_pipe_will_have_type(crtc_state, INTEL_OUTPUT_LVDS)) { if (intel_is_dual_link_lvds(dev)) limit = &intel_limits_g4x_dual_channel_lvds; else limit = &intel_limits_g4x_single_channel_lvds; - } else if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_HDMI) || - intel_pipe_will_have_type(crtc, INTEL_OUTPUT_ANALOG)) { + } else if (intel_pipe_will_have_type(crtc_state, INTEL_OUTPUT_HDMI) || + intel_pipe_will_have_type(crtc_state, INTEL_OUTPUT_ANALOG)) { limit = &intel_limits_g4x_hdmi; - } else if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_SDVO)) { + } else if (intel_pipe_will_have_type(crtc_state, INTEL_OUTPUT_SDVO)) { limit = &intel_limits_g4x_sdvo; } else /* The option is for other outputs */ limit = &intel_limits_i9xx_sdvo; @@ -488,17 +505,18 @@ static const intel_limit_t *intel_g4x_limit(struct intel_crtc *crtc) return limit; } -static const intel_limit_t *intel_limit(struct intel_crtc *crtc, int refclk) +static const intel_limit_t * +intel_limit(struct intel_crtc_state *crtc_state, int refclk) { - struct drm_device *dev = crtc->base.dev; + struct drm_device *dev = crtc_state->base.crtc->dev; const intel_limit_t *limit; if (HAS_PCH_SPLIT(dev)) - limit = intel_ironlake_limit(crtc, refclk); + limit = intel_ironlake_limit(crtc_state, refclk); else if (IS_G4X(dev)) { - limit = intel_g4x_limit(crtc); + limit = intel_g4x_limit(crtc_state); } else if (IS_PINEVIEW(dev)) { - if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS)) + if (intel_pipe_will_have_type(crtc_state, INTEL_OUTPUT_LVDS)) limit = &intel_limits_pineview_lvds; else limit = &intel_limits_pineview_sdvo; @@ -507,14 +525,14 @@ static const intel_limit_t *intel_limit(struct intel_crtc *crtc, int refclk) } else if (IS_VALLEYVIEW(dev)) { limit = &intel_limits_vlv; } else if (!IS_GEN2(dev)) { - if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS)) + if (intel_pipe_will_have_type(crtc_state, INTEL_OUTPUT_LVDS)) limit = &intel_limits_i9xx_lvds; else limit = &intel_limits_i9xx_sdvo; } else { - if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS)) + if (intel_pipe_will_have_type(crtc_state, INTEL_OUTPUT_LVDS)) limit = &intel_limits_i8xx_lvds; - else if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_DVO)) + else if (intel_pipe_will_have_type(crtc_state, INTEL_OUTPUT_DVO)) limit = &intel_limits_i8xx_dvo; else limit = &intel_limits_i8xx_dac; @@ -601,15 +619,17 @@ static bool intel_PLL_is_valid(struct drm_device *dev, } static bool -i9xx_find_best_dpll(const intel_limit_t *limit, struct intel_crtc *crtc, +i9xx_find_best_dpll(const intel_limit_t *limit, + struct intel_crtc_state *crtc_state, int target, int refclk, intel_clock_t *match_clock, intel_clock_t *best_clock) { + struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc); struct drm_device *dev = crtc->base.dev; intel_clock_t clock; int err = target; - if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS)) { + if (intel_pipe_will_have_type(crtc_state, INTEL_OUTPUT_LVDS)) { /* * For LVDS just rely on its current settings for dual-channel. * We haven't figured out how to reliably set up different @@ -662,15 +682,17 @@ i9xx_find_best_dpll(const intel_limit_t *limit, struct intel_crtc *crtc, } static bool -pnv_find_best_dpll(const intel_limit_t *limit, struct intel_crtc *crtc, +pnv_find_best_dpll(const intel_limit_t *limit, + struct intel_crtc_state *crtc_state, int target, int refclk, intel_clock_t *match_clock, intel_clock_t *best_clock) { + struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc); struct drm_device *dev = crtc->base.dev; intel_clock_t clock; int err = target; - if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS)) { + if (intel_pipe_will_have_type(crtc_state, INTEL_OUTPUT_LVDS)) { /* * For LVDS just rely on its current settings for dual-channel. * We haven't figured out how to reliably set up different @@ -721,10 +743,12 @@ pnv_find_best_dpll(const intel_limit_t *limit, struct intel_crtc *crtc, } static bool -g4x_find_best_dpll(const intel_limit_t *limit, struct intel_crtc *crtc, +g4x_find_best_dpll(const intel_limit_t *limit, + struct intel_crtc_state *crtc_state, int target, int refclk, intel_clock_t *match_clock, intel_clock_t *best_clock) { + struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc); struct drm_device *dev = crtc->base.dev; intel_clock_t clock; int max_n; @@ -733,7 +757,7 @@ g4x_find_best_dpll(const intel_limit_t *limit, struct intel_crtc *crtc, int err_most = (target >> 8) + (target >> 9); found = false; - if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS)) { + if (intel_pipe_will_have_type(crtc_state, INTEL_OUTPUT_LVDS)) { if (intel_is_dual_link_lvds(dev)) clock.p2 = limit->p2.p2_fast; else @@ -818,10 +842,12 @@ static bool vlv_PLL_is_optimal(struct drm_device *dev, int target_freq, } static bool -vlv_find_best_dpll(const intel_limit_t *limit, struct intel_crtc *crtc, +vlv_find_best_dpll(const intel_limit_t *limit, + struct intel_crtc_state *crtc_state, int target, int refclk, intel_clock_t *match_clock, intel_clock_t *best_clock) { + struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc); struct drm_device *dev = crtc->base.dev; intel_clock_t clock; unsigned int bestppm = 1000000; @@ -870,10 +896,12 @@ vlv_find_best_dpll(const intel_limit_t *limit, struct intel_crtc *crtc, } static bool -chv_find_best_dpll(const intel_limit_t *limit, struct intel_crtc *crtc, +chv_find_best_dpll(const intel_limit_t *limit, + struct intel_crtc_state *crtc_state, int target, int refclk, intel_clock_t *match_clock, intel_clock_t *best_clock) { + struct intel_crtc *crtc = to_intel_crtc(crtc_state->base.crtc); struct drm_device *dev = crtc->base.dev; unsigned int best_error_ppm; intel_clock_t clock; @@ -5795,7 +5823,7 @@ static int intel_crtc_compute_config(struct intel_crtc *crtc, * - LVDS dual channel mode * - Double wide pipe */ - if ((intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS) && + if ((intel_pipe_will_have_type(pipe_config, INTEL_OUTPUT_LVDS) && intel_is_dual_link_lvds(dev)) || pipe_config->double_wide) pipe_config->pipe_src_w &= ~1; @@ -5974,15 +6002,18 @@ static inline bool intel_panel_use_ssc(struct drm_i915_private *dev_priv) && !(dev_priv->quirks & QUIRK_LVDS_SSC_DISABLE); } -static int i9xx_get_refclk(struct intel_crtc *crtc, int num_connectors) +static int i9xx_get_refclk(const struct intel_crtc_state *crtc_state, + int num_connectors) { - struct drm_device *dev = crtc->base.dev; + struct drm_device *dev = crtc_state->base.crtc->dev; struct drm_i915_private *dev_priv = dev->dev_private; int refclk; + WARN_ON(!crtc_state->base.state); + if (IS_VALLEYVIEW(dev)) { refclk = 100000; - } else if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS) && + } else if (intel_pipe_will_have_type(crtc_state, INTEL_OUTPUT_LVDS) && intel_panel_use_ssc(dev_priv) && num_connectors < 2) { refclk = dev_priv->vbt.lvds_ssc_freq; DRM_DEBUG_KMS("using SSC reference clock of %d kHz\n", refclk); @@ -6025,7 +6056,7 @@ static void i9xx_update_pll_dividers(struct intel_crtc *crtc, crtc_state->dpll_hw_state.fp0 = fp; crtc->lowfreq_avail = false; - if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS) && + if (intel_pipe_will_have_type(crtc_state, INTEL_OUTPUT_LVDS) && reduced_clock) { crtc_state->dpll_hw_state.fp1 = fp2; crtc->lowfreq_avail = true; @@ -6383,6 +6414,7 @@ void vlv_force_pll_on(struct drm_device *dev, enum pipe pipe, struct intel_crtc *crtc = to_intel_crtc(intel_get_crtc_for_pipe(dev, pipe)); struct intel_crtc_state pipe_config = { + .base.crtc = &crtc->base, .pixel_multiplier = 1, .dpll = *dpll, }; @@ -6427,12 +6459,12 @@ static void i9xx_update_pll(struct intel_crtc *crtc, i9xx_update_pll_dividers(crtc, crtc_state, reduced_clock); - is_sdvo = intel_pipe_will_have_type(crtc, INTEL_OUTPUT_SDVO) || - intel_pipe_will_have_type(crtc, INTEL_OUTPUT_HDMI); + is_sdvo = intel_pipe_will_have_type(crtc_state, INTEL_OUTPUT_SDVO) || + intel_pipe_will_have_type(crtc_state, INTEL_OUTPUT_HDMI); dpll = DPLL_VGA_MODE_DIS; - if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS)) + if (intel_pipe_will_have_type(crtc_state, INTEL_OUTPUT_LVDS)) dpll |= DPLLB_MODE_LVDS; else dpll |= DPLLB_MODE_DAC_SERIAL; @@ -6475,7 +6507,7 @@ static void i9xx_update_pll(struct intel_crtc *crtc, if (crtc_state->sdvo_tv_clock) dpll |= PLL_REF_INPUT_TVCLKINBC; - else if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS) && + else if (intel_pipe_will_have_type(crtc_state, INTEL_OUTPUT_LVDS) && intel_panel_use_ssc(dev_priv) && num_connectors < 2) dpll |= PLLB_REF_INPUT_SPREADSPECTRUMIN; else @@ -6505,7 +6537,7 @@ static void i8xx_update_pll(struct intel_crtc *crtc, dpll = DPLL_VGA_MODE_DIS; - if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS)) { + if (intel_pipe_will_have_type(crtc_state, INTEL_OUTPUT_LVDS)) { dpll |= (1 << (clock->p1 - 1)) << DPLL_FPA01_P1_POST_DIV_SHIFT; } else { if (clock->p1 == 2) @@ -6516,10 +6548,10 @@ static void i8xx_update_pll(struct intel_crtc *crtc, dpll |= PLL_P2_DIVIDE_BY_4; } - if (!IS_I830(dev) && intel_pipe_will_have_type(crtc, INTEL_OUTPUT_DVO)) + if (!IS_I830(dev) && intel_pipe_will_have_type(crtc_state, INTEL_OUTPUT_DVO)) dpll |= DPLL_DVO_2X_MODE; - if (intel_pipe_will_have_type(crtc, INTEL_OUTPUT_LVDS) && + if (intel_pipe_will_have_type(crtc_state, INTEL_OUTPUT_LVDS) && intel_panel_use_ssc(dev_priv) && num_connectors < 2) dpll |= PLLB_REF_INPUT_SPREADSPECTRUMIN; else @@ -6756,7 +6788,7 @@ static int i9xx_crtc_compute_clock(struct intel_crtc *crtc, return 0; if (!crtc_state->clock_set) { - refclk = i9xx_get_refclk(crtc, num_connectors); + refclk = i9xx_get_refclk(crtc_state, num_connectors); /* * Returns a set of divisors for the desired target clock with @@ -6764,8 +6796,8 @@ static int i9xx_crtc_compute_clock(struct intel_crtc *crtc, * the clock equation: reflck * (5 * (m1 + 2) + (m2 + 2)) / (n + * 2) / p1 / p2. */ - limit = intel_limit(crtc, refclk); - ok = dev_priv->display.find_dpll(limit, crtc, + limit = intel_limit(crtc_state, refclk); + ok = dev_priv->display.find_dpll(limit, crtc_state, crtc_state->port_clock, refclk, NULL, &clock); if (!ok) { @@ -6781,7 +6813,7 @@ static int i9xx_crtc_compute_clock(struct intel_crtc *crtc, * we will disable the LVDS downclock feature. */ has_reduced_clock = - dev_priv->display.find_dpll(limit, crtc, + dev_priv->display.find_dpll(limit, crtc_state, dev_priv->lvds_downclock, refclk, &clock, &reduced_clock); @@ -7609,12 +7641,11 @@ static bool ironlake_compute_clocks(struct drm_crtc *crtc, { struct drm_device *dev = crtc->dev; struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_crtc *intel_crtc = to_intel_crtc(crtc); int refclk; const intel_limit_t *limit; bool ret, is_lvds = false; - is_lvds = intel_pipe_will_have_type(intel_crtc, INTEL_OUTPUT_LVDS); + is_lvds = intel_pipe_will_have_type(crtc_state, INTEL_OUTPUT_LVDS); refclk = ironlake_get_refclk(crtc); @@ -7623,8 +7654,8 @@ static bool ironlake_compute_clocks(struct drm_crtc *crtc, * refclk, or FALSE. The returned values represent the clock equation: * reflck * (5 * (m1 + 2) + (m2 + 2)) / (n + 2) / p1 / p2. */ - limit = intel_limit(intel_crtc, refclk); - ret = dev_priv->display.find_dpll(limit, intel_crtc, + limit = intel_limit(crtc_state, refclk); + ret = dev_priv->display.find_dpll(limit, crtc_state, crtc_state->port_clock, refclk, NULL, clock); if (!ret) @@ -7638,7 +7669,7 @@ static bool ironlake_compute_clocks(struct drm_crtc *crtc, * downclock feature. */ *has_reduced_clock = - dev_priv->display.find_dpll(limit, intel_crtc, + dev_priv->display.find_dpll(limit, crtc_state, dev_priv->lvds_downclock, refclk, clock, reduced_clock); -- cgit v1.2.1 From 723f9aab5537eabd88f96fa49a73a5e8df2ebf7b Mon Sep 17 00:00:00 2001 From: Ander Conselvan de Oliveira Date: Fri, 20 Mar 2015 16:18:18 +0200 Subject: drm/i915: Don't look at staged config crtc when changing DRRS state The function intel_dp_set_drrs_state() would decide which pipe to downclock based on the staged config for the given connector. However, the result of that function is immediate, and it uses input values from crtc->config, so it should be looking at the current crtc instead. Signed-off-by: Ander Conselvan de Oliveira Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_dp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/intel_dp.c b/drivers/gpu/drm/i915/intel_dp.c index 0286fc6fb89a..b70e635ccaf4 100644 --- a/drivers/gpu/drm/i915/intel_dp.c +++ b/drivers/gpu/drm/i915/intel_dp.c @@ -5000,7 +5000,7 @@ static void intel_dp_set_drrs_state(struct drm_device *dev, int refresh_rate) dig_port = dp_to_dig_port(intel_dp); encoder = &dig_port->base; - intel_crtc = encoder->new_crtc; + intel_crtc = to_intel_crtc(encoder->base.crtc); if (!intel_crtc) { DRM_DEBUG_KMS("DRRS: intel_crtc not initialized\n"); -- cgit v1.2.1 From 55bb9992db9b2c65cc3dd11cc75785f80e8d7e66 Mon Sep 17 00:00:00 2001 From: Ander Conselvan de Oliveira Date: Fri, 20 Mar 2015 16:18:19 +0200 Subject: drm/i915: Remove usage of encoder->new_crtc from clock computations Some of the crtc_compute_clock() still depended on encoder->new_crtc since they didn't use intel_pipe_will_have_type() and used an open coded version of that function instead. This patch replaces those with the appropriate code that checks the atomic state intead. Signed-off-by: Ander Conselvan de Oliveira [danvet: Separate the if (!connector) continue to facility easier extraction of a loop iterator for all of these (there's lots more in i915 and atomic helpers).] Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_display.c | 51 +++++++++++++++++++++++++++--------- 1 file changed, 38 insertions(+), 13 deletions(-) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 1c4efeee13c8..dc9a5c1d596c 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -6765,11 +6765,20 @@ static int i9xx_crtc_compute_clock(struct intel_crtc *crtc, bool is_lvds = false, is_dsi = false; struct intel_encoder *encoder; const intel_limit_t *limit; + struct drm_atomic_state *state = crtc_state->base.state; + struct drm_connector_state *connector_state; + int i; - for_each_intel_encoder(dev, encoder) { - if (encoder->new_crtc != crtc) + for (i = 0; i < state->num_connector; i++) { + if (!state->connectors[i]) + continue; + + connector_state = state->connector_states[i]; + if (connector_state->crtc != &crtc->base) continue; + encoder = to_intel_encoder(connector_state->best_encoder); + switch (encoder->type) { case INTEL_OUTPUT_LVDS: is_lvds = true; @@ -7443,18 +7452,26 @@ void intel_init_pch_refclk(struct drm_device *dev) lpt_init_pch_refclk(dev); } -static int ironlake_get_refclk(struct drm_crtc *crtc) +static int ironlake_get_refclk(struct intel_crtc_state *crtc_state) { - struct drm_device *dev = crtc->dev; + struct drm_device *dev = crtc_state->base.crtc->dev; struct drm_i915_private *dev_priv = dev->dev_private; + struct drm_atomic_state *state = crtc_state->base.state; + struct drm_connector_state *connector_state; struct intel_encoder *encoder; - int num_connectors = 0; + int num_connectors = 0, i; bool is_lvds = false; - for_each_intel_encoder(dev, encoder) { - if (encoder->new_crtc != to_intel_crtc(crtc)) + for (i = 0; i < state->num_connector; i++) { + if (!state->connectors[i]) continue; + connector_state = state->connector_states[i]; + if (connector_state->crtc != crtc_state->base.crtc) + continue; + + encoder = to_intel_encoder(connector_state->best_encoder); + switch (encoder->type) { case INTEL_OUTPUT_LVDS: is_lvds = true; @@ -7647,7 +7664,7 @@ static bool ironlake_compute_clocks(struct drm_crtc *crtc, is_lvds = intel_pipe_will_have_type(crtc_state, INTEL_OUTPUT_LVDS); - refclk = ironlake_get_refclk(crtc); + refclk = ironlake_get_refclk(crtc_state); /* * Returns a set of divisors for the desired target clock with the given @@ -7702,16 +7719,24 @@ static uint32_t ironlake_compute_dpll(struct intel_crtc *intel_crtc, struct drm_crtc *crtc = &intel_crtc->base; struct drm_device *dev = crtc->dev; struct drm_i915_private *dev_priv = dev->dev_private; - struct intel_encoder *intel_encoder; + struct drm_atomic_state *state = crtc_state->base.state; + struct drm_connector_state *connector_state; + struct intel_encoder *encoder; uint32_t dpll; - int factor, num_connectors = 0; + int factor, num_connectors = 0, i; bool is_lvds = false, is_sdvo = false; - for_each_intel_encoder(dev, intel_encoder) { - if (intel_encoder->new_crtc != to_intel_crtc(crtc)) + for (i = 0; i < state->num_connector; i++) { + if (!state->connectors[i]) continue; - switch (intel_encoder->type) { + connector_state = state->connector_states[i]; + if (connector_state->crtc != crtc_state->base.crtc) + continue; + + encoder = to_intel_encoder(connector_state->best_encoder); + + switch (encoder->type) { case INTEL_OUTPUT_LVDS: is_lvds = true; break; -- cgit v1.2.1 From 5bedeb2de2127b3b80b40d977a17fe4fce83fe25 Mon Sep 17 00:00:00 2001 From: Daniel Vetter Date: Tue, 3 Mar 2015 18:03:47 +0100 Subject: drm/i915: Add module param to test the load detect code This is useful for writing igts to make sure we don't break this, without being forced to own a one of these dinosaurs. Suggested-by: Jesse Barnes Cc: Matt Roper Signed-off-by: Daniel Vetter Reviewed-by: Ander Conselvan de Oliveira Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_drv.h | 1 + drivers/gpu/drm/i915/i915_params.c | 8 +++++++- drivers/gpu/drm/i915/intel_crt.c | 6 ++++-- 3 files changed, 12 insertions(+), 3 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index a4db7a7db444..6ccdf954b1ca 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -2462,6 +2462,7 @@ struct i915_params { bool enable_hangcheck; bool fastboot; bool prefault_disable; + bool load_detect_test; bool reset; bool disable_display; bool disable_vtd_wa; diff --git a/drivers/gpu/drm/i915/i915_params.c b/drivers/gpu/drm/i915/i915_params.c index f3b8b62cb74d..bb64415a1c3e 100644 --- a/drivers/gpu/drm/i915/i915_params.c +++ b/drivers/gpu/drm/i915/i915_params.c @@ -43,6 +43,7 @@ struct i915_params i915 __read_mostly = { .enable_ips = 1, .fastboot = 0, .prefault_disable = 0, + .load_detect_test = 0, .reset = true, .invert_brightness = 0, .disable_display = 0, @@ -139,11 +140,16 @@ module_param_named(fastboot, i915.fastboot, bool, 0600); MODULE_PARM_DESC(fastboot, "Try to skip unnecessary mode sets at boot time (default: false)"); -module_param_named(prefault_disable, i915.prefault_disable, bool, 0600); +module_param_named_unsafe(prefault_disable, i915.prefault_disable, bool, 0600); MODULE_PARM_DESC(prefault_disable, "Disable page prefaulting for pread/pwrite/reloc (default:false). " "For developers only."); +module_param_named_unsafe(load_detect_test, i915.load_detect_test, bool, 0600); +MODULE_PARM_DESC(load_detect_test, + "Force-enable the VGA load detect code for testing (default:false). " + "For developers only."); + module_param_named(invert_brightness, i915.invert_brightness, int, 0600); MODULE_PARM_DESC(invert_brightness, "Invert backlight brightness " diff --git a/drivers/gpu/drm/i915/intel_crt.c b/drivers/gpu/drm/i915/intel_crt.c index 573aaff3389a..6095a998bdac 100644 --- a/drivers/gpu/drm/i915/intel_crt.c +++ b/drivers/gpu/drm/i915/intel_crt.c @@ -690,7 +690,7 @@ intel_crt_detect(struct drm_connector *connector, bool force) * broken monitor (without edid) to work behind a broken kvm (that fails * to have the right resistors for HP detection) needs to fix this up. * For now just bail out. */ - if (I915_HAS_HOTPLUG(dev)) { + if (I915_HAS_HOTPLUG(dev) && !i915.load_detect_test) { status = connector_status_disconnected; goto out; } @@ -706,8 +706,10 @@ intel_crt_detect(struct drm_connector *connector, bool force) if (intel_get_load_detect_pipe(connector, NULL, &tmp, &ctx)) { if (intel_crt_detect_ddc(connector)) status = connector_status_connected; - else + else if (INTEL_INFO(dev)->gen < 4) status = intel_crt_load_detect(crt); + else + status = connector_status_unknown; intel_release_load_detect_pipe(connector, &tmp, &ctx); } else status = connector_status_unknown; -- cgit v1.2.1 From 2f2cf68261acc4c5356dd07e8a8966612ae5a0d4 Mon Sep 17 00:00:00 2001 From: kbuild test robot Date: Fri, 27 Mar 2015 19:26:35 +0800 Subject: drm/i915: fix simple_return.cocci warnings drivers/gpu/drm/i915/i915_gem_gtt.c:1349:1-4: WARNING: end returns can be simpified and declaration on line 1347 can be dropped Simplify a trivial if-return sequence. Possibly combine with a preceding function call. Generated by: scripts/coccinelle/misc/simple_return.cocci CC: Michel Thierry Signed-off-by: Fengguang Wu Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_gem_gtt.c | 8 +------- 1 file changed, 1 insertion(+), 7 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c index e33b1214c4d8..e82cec2c714e 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c @@ -1349,13 +1349,7 @@ err_out: static int gen6_ppgtt_alloc(struct i915_hw_ppgtt *ppgtt) { - int ret; - - ret = gen6_ppgtt_allocate_page_directories(ppgtt); - if (ret) - return ret; - - return 0; + return gen6_ppgtt_allocate_page_directories(ppgtt); } static void gen6_scratch_va_range(struct i915_hw_ppgtt *ppgtt, -- cgit v1.2.1 From 9abc464854cf12273ebe66dc5e4a9d82b6e0c303 Mon Sep 17 00:00:00 2001 From: Joonas Lahtinen Date: Fri, 27 Mar 2015 13:09:22 +0200 Subject: drm/i915: Compare GGTT view structs instead of types To allow for views where the view type is not defined by the view type only, like it is in stereo or rotated 90 degree view, change the semantic to require the whole view structure for comparison when we match a GGTT view. This allows including parameters like offset to be included in the view which is useful for eg. partial views. v3: - Rely on ggtt_view type being 0 for non-GGTT vma's, which equals to I915_GGTT_VIEW_NORMAL. (Daniel Vetter) - Do not use potentially slower comparison when we only want to know if something is or is not a normal view. - Rebase on top of rotated view patches. Add rotated view singleton. - If one view is missing in comparison they're equal only if both are missing. v4: - Use comparison helper in obj_to_ggtt_view too. (Tvrtko Ursulin) - Do WARN_ON if one view is NULL. (Tvrtko Ursulin) Cc: Daniel Vetter Signed-off-by: Joonas Lahtinen Reviewed-by: Tvrtko Ursulin Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_drv.h | 8 ++++---- drivers/gpu/drm/i915/i915_gem.c | 16 +++++++++------- drivers/gpu/drm/i915/i915_gem_gtt.c | 3 +++ drivers/gpu/drm/i915/i915_gem_gtt.h | 11 +++++++++++ drivers/gpu/drm/i915/intel_display.c | 8 +++----- 5 files changed, 30 insertions(+), 16 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 6ccdf954b1ca..970714bb77a2 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -2795,19 +2795,19 @@ void i915_gem_restore_fences(struct drm_device *dev); unsigned long i915_gem_obj_ggtt_offset_view(struct drm_i915_gem_object *o, - enum i915_ggtt_view_type view); + const struct i915_ggtt_view *view); unsigned long i915_gem_obj_offset(struct drm_i915_gem_object *o, struct i915_address_space *vm); static inline unsigned long i915_gem_obj_ggtt_offset(struct drm_i915_gem_object *o) { - return i915_gem_obj_ggtt_offset_view(o, I915_GGTT_VIEW_NORMAL); + return i915_gem_obj_ggtt_offset_view(o, &i915_ggtt_view_normal); } bool i915_gem_obj_bound_any(struct drm_i915_gem_object *o); bool i915_gem_obj_ggtt_bound_view(struct drm_i915_gem_object *o, - enum i915_ggtt_view_type view); + const struct i915_ggtt_view *view); bool i915_gem_obj_bound(struct drm_i915_gem_object *o, struct i915_address_space *vm); @@ -2855,7 +2855,7 @@ i915_vm_to_ppgtt(struct i915_address_space *vm) static inline bool i915_gem_obj_ggtt_bound(struct drm_i915_gem_object *obj) { - return i915_gem_obj_ggtt_bound_view(obj, I915_GGTT_VIEW_NORMAL); + return i915_gem_obj_ggtt_bound_view(obj, &i915_ggtt_view_normal); } static inline unsigned long diff --git a/drivers/gpu/drm/i915/i915_gem.c b/drivers/gpu/drm/i915/i915_gem.c index c13e65994654..4ca3dab77fa4 100644 --- a/drivers/gpu/drm/i915/i915_gem.c +++ b/drivers/gpu/drm/i915/i915_gem.c @@ -4129,7 +4129,7 @@ i915_gem_object_do_pin(struct drm_i915_gem_object *obj, if (i915_vma_misplaced(vma, alignment, flags)) { unsigned long offset; - offset = ggtt_view ? i915_gem_obj_ggtt_offset_view(obj, ggtt_view->type) : + offset = ggtt_view ? i915_gem_obj_ggtt_offset_view(obj, ggtt_view) : i915_gem_obj_offset(obj, vm); WARN(vma->pin_count, "bo is already pinned in %s with incorrect alignment:" @@ -4228,7 +4228,7 @@ i915_gem_object_ggtt_unpin_view(struct drm_i915_gem_object *obj, BUG_ON(!vma); WARN_ON(vma->pin_count == 0); - WARN_ON(!i915_gem_obj_ggtt_bound_view(obj, view->type)); + WARN_ON(!i915_gem_obj_ggtt_bound_view(obj, view)); if (--vma->pin_count == 0 && view->type == I915_GGTT_VIEW_NORMAL) obj->pin_mappable = false; @@ -4550,7 +4550,8 @@ struct i915_vma *i915_gem_obj_to_ggtt_view(struct drm_i915_gem_object *obj, return ERR_PTR(-EINVAL); list_for_each_entry(vma, &obj->vma_list, vma_link) - if (vma->vm == ggtt && vma->ggtt_view.type == view->type) + if (vma->vm == ggtt && + i915_ggtt_view_equal(&vma->ggtt_view, view)) return vma; return NULL; } @@ -5107,13 +5108,14 @@ i915_gem_obj_offset(struct drm_i915_gem_object *o, unsigned long i915_gem_obj_ggtt_offset_view(struct drm_i915_gem_object *o, - enum i915_ggtt_view_type view) + const struct i915_ggtt_view *view) { struct i915_address_space *ggtt = i915_obj_to_ggtt(o); struct i915_vma *vma; list_for_each_entry(vma, &o->vma_list, vma_link) - if (vma->vm == ggtt && vma->ggtt_view.type == view) + if (vma->vm == ggtt && + i915_ggtt_view_equal(&vma->ggtt_view, view)) return vma->node.start; WARN(1, "global vma for this object not found.\n"); @@ -5137,14 +5139,14 @@ bool i915_gem_obj_bound(struct drm_i915_gem_object *o, } bool i915_gem_obj_ggtt_bound_view(struct drm_i915_gem_object *o, - enum i915_ggtt_view_type view) + const struct i915_ggtt_view *view) { struct i915_address_space *ggtt = i915_obj_to_ggtt(o); struct i915_vma *vma; list_for_each_entry(vma, &o->vma_list, vma_link) if (vma->vm == ggtt && - vma->ggtt_view.type == view && + i915_ggtt_view_equal(&vma->ggtt_view, view) && drm_mm_node_allocated(&vma->node)) return true; diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.c b/drivers/gpu/drm/i915/i915_gem_gtt.c index e82cec2c714e..0239fbff7bf7 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.c +++ b/drivers/gpu/drm/i915/i915_gem_gtt.c @@ -93,6 +93,9 @@ */ const struct i915_ggtt_view i915_ggtt_view_normal; +const struct i915_ggtt_view i915_ggtt_view_rotated = { + .type = I915_GGTT_VIEW_ROTATED +}; static void bdw_setup_private_ppat(struct drm_i915_private *dev_priv); static void chv_setup_private_ppat(struct drm_i915_private *dev_priv); diff --git a/drivers/gpu/drm/i915/i915_gem_gtt.h b/drivers/gpu/drm/i915/i915_gem_gtt.h index 3f0ad9f25441..fc03c99317c9 100644 --- a/drivers/gpu/drm/i915/i915_gem_gtt.h +++ b/drivers/gpu/drm/i915/i915_gem_gtt.h @@ -138,6 +138,7 @@ struct i915_ggtt_view { }; extern const struct i915_ggtt_view i915_ggtt_view_normal; +extern const struct i915_ggtt_view i915_ggtt_view_rotated; enum i915_cache_level; @@ -424,4 +425,14 @@ void i915_gem_restore_gtt_mappings(struct drm_device *dev); int __must_check i915_gem_gtt_prepare_object(struct drm_i915_gem_object *obj); void i915_gem_gtt_finish_object(struct drm_i915_gem_object *obj); +static inline bool +i915_ggtt_view_equal(const struct i915_ggtt_view *a, + const struct i915_ggtt_view *b) +{ + if (WARN_ON(!a || !b)) + return false; + + return a->type == b->type; +} + #endif diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index dc9a5c1d596c..3852cbaf9163 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -2322,8 +2322,6 @@ intel_fill_fb_ggtt_view(struct i915_ggtt_view *view, struct drm_framebuffer *fb, const struct drm_plane_state *plane_state) { struct intel_rotation_info *info = &view->rotation_info; - static const struct i915_ggtt_view rotated_view = - { .type = I915_GGTT_VIEW_ROTATED }; *view = i915_ggtt_view_normal; @@ -2333,7 +2331,7 @@ intel_fill_fb_ggtt_view(struct i915_ggtt_view *view, struct drm_framebuffer *fb, if (!intel_rotation_90_or_270(plane_state->rotation)) return 0; - *view = rotated_view; + *view = i915_ggtt_view_rotated; info->height = fb->height; info->pixel_format = fb->pixel_format; @@ -2930,10 +2928,10 @@ u32 intel_fb_stride_alignment(struct drm_device *dev, uint64_t fb_modifier, unsigned long intel_plane_obj_offset(struct intel_plane *intel_plane, struct drm_i915_gem_object *obj) { - enum i915_ggtt_view_type view = I915_GGTT_VIEW_NORMAL; + const struct i915_ggtt_view *view = &i915_ggtt_view_normal; if (intel_rotation_90_or_270(intel_plane->base.state->rotation)) - view = I915_GGTT_VIEW_ROTATED; + view = &i915_ggtt_view_rotated; return i915_gem_obj_ggtt_offset_view(obj, view); } -- cgit v1.2.1 From 9d918c157fc27716ca8d601745b4776a3145cc93 Mon Sep 17 00:00:00 2001 From: Ander Conselvan de Oliveira Date: Fri, 27 Mar 2015 15:33:51 +0200 Subject: drm/i915: Handle error to get connector state when staging config The return value of one of the calls to drm_atomic_get_connector_state() in intel_modeset_stage_output_state() wasn't checked for errors. Reported-by: Dan Carpenter Signed-off-by: Ander Conselvan de Oliveira Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/intel_display.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/gpu/drm/i915/intel_display.c b/drivers/gpu/drm/i915/intel_display.c index 3852cbaf9163..75955fee6d24 100644 --- a/drivers/gpu/drm/i915/intel_display.c +++ b/drivers/gpu/drm/i915/intel_display.c @@ -11985,6 +11985,8 @@ intel_modeset_stage_output_state(struct drm_device *dev, for_each_intel_connector(dev, connector) { connector_state = drm_atomic_get_connector_state(state, &connector->base); + if (IS_ERR(connector_state)) + return PTR_ERR(connector_state); if (connector->new_encoder) { if (connector->new_encoder != connector->encoder) -- cgit v1.2.1 From 743e78c1d726d875b98ff9689cc77c4d3d5d9ae2 Mon Sep 17 00:00:00 2001 From: Chris Wilson Date: Fri, 27 Mar 2015 11:02:10 +0000 Subject: drm/i915: Skip allocating shadow batch for 0-length batches Since commit 17cabf571e50677d980e9ab2a43c5f11213003ae Author: Chris Wilson Date: Wed Jan 14 11:20:57 2015 +0000 drm/i915: Trim the command parser allocations we may then try to allocate a zero-sized object and attempt to extract its pages. Understandably this fails. Testcase: igt/gem_exec_nop #ivb,byt,hsw Signed-off-by: Chris Wilson Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_gem_execbuffer.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/i915_gem_execbuffer.c b/drivers/gpu/drm/i915/i915_gem_execbuffer.c index 43335de0f713..a3190e793ed4 100644 --- a/drivers/gpu/drm/i915/i915_gem_execbuffer.c +++ b/drivers/gpu/drm/i915/i915_gem_execbuffer.c @@ -1547,7 +1547,7 @@ i915_gem_do_execbuffer(struct drm_device *dev, void *data, goto err; } - if (i915_needs_cmd_parser(ring)) { + if (i915_needs_cmd_parser(ring) && args->batch_len) { batch_obj = i915_gem_execbuffer_parse(ring, &shadow_exec_entry, eb, -- cgit v1.2.1 From 1ff27a3443addb19af609245b218fa510c7d37dd Mon Sep 17 00:00:00 2001 From: Daniel Vetter Date: Fri, 27 Mar 2015 20:21:01 +0100 Subject: drm/i915: Update DRIVER_DATE to 20150327 Signed-off-by: Daniel Vetter --- drivers/gpu/drm/i915/i915_drv.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/i915/i915_drv.h b/drivers/gpu/drm/i915/i915_drv.h index 970714bb77a2..e326ac9730cf 100644 --- a/drivers/gpu/drm/i915/i915_drv.h +++ b/drivers/gpu/drm/i915/i915_drv.h @@ -56,7 +56,7 @@ #define DRIVER_NAME "i915" #define DRIVER_DESC "Intel Graphics" -#define DRIVER_DATE "20150313" +#define DRIVER_DATE "20150327" #undef WARN_ON /* Many gcc seem to no see through this and fall over :( */ -- cgit v1.2.1 From 78d84bc3734c2566dbba09baae2414734661ed6a Mon Sep 17 00:00:00 2001 From: Dave Martin Date: Tue, 17 Mar 2015 12:35:41 +0000 Subject: arm64: juno: Fix misleading name of UART reference clock The UART reference clock speed is 7273.8 kHz, not 72738 kHz. Dots aren't usually used in node names even though ePAPR permits them. However, this can easily be avoided by expressing the frequency in Hz, not kHz. This patch changes the name to refclk7273800hz, reflecting the actual clock speed. Signed-off-by: Dave Martin Acked-by: Liviu Dudau Signed-off-by: Olof Johansson --- arch/arm64/boot/dts/arm/juno-clocks.dtsi | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm64/boot/dts/arm/juno-clocks.dtsi b/arch/arm64/boot/dts/arm/juno-clocks.dtsi index ea2b5666a16f..c9b89efe0f56 100644 --- a/arch/arm64/boot/dts/arm/juno-clocks.dtsi +++ b/arch/arm64/boot/dts/arm/juno-clocks.dtsi @@ -8,7 +8,7 @@ */ /* SoC fixed clocks */ - soc_uartclk: refclk72738khz { + soc_uartclk: refclk7273800hz { compatible = "fixed-clock"; #clock-cells = <0>; clock-frequency = <7273800>; -- cgit v1.2.1 From e42391cd048809d903291d07f86ed3934ce138e9 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sun, 29 Mar 2015 15:26:31 -0700 Subject: Linux 4.0-rc6 --- Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Makefile b/Makefile index 14c722f96877..da36a3be7969 100644 --- a/Makefile +++ b/Makefile @@ -1,7 +1,7 @@ VERSION = 4 PATCHLEVEL = 0 SUBLEVEL = 0 -EXTRAVERSION = -rc5 +EXTRAVERSION = -rc6 NAME = Hurr durr I'ma sheep # *DOCUMENTATION* -- cgit v1.2.1