summaryrefslogtreecommitdiffstats
path: root/net
Commit message (Collapse)AuthorAgeFilesLines
* Merge branch 'for-davem' of ↵David S. Miller2010-12-2331-218/+769
|\ | | | | | | ssh://master.kernel.org/pub/scm/linux/kernel/git/linville/wireless-next-2.6
| * mac80211: check for CONFIG_MAC80211_LEDS in the tpt_led_trigger declarationLuciano Coelho2010-12-221-0/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | If CONFIG_MAC80211_LEDS is not set, ieee80211_i.h was failing to compile, because struct led_trigger is only declared when CONFIG_LEDS_TRIGGERS is set. This patch adds ifdefs around the tpt_led_trigger declaration in ieee80211_i.h to avoid the problem. Signed-off-by: Luciano Coelho <luciano.coelho@nokia.com> Acked-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * mac80211: selective throughput LED trigger activeJohannes Berg2010-12-225-15/+59
| | | | | | | | | | | | | | | | | | | | | | The throughput LED trigger was always active when the radio was enabled. In most cases that's likely the desired behaviour, but iwlwifi requires it to be only active when one of the virtual interfaces is actually "connected" in some way. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * mac80211: add throughput based LED blink triggerJohannes Berg2010-12-227-9/+178
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | iwlwifi and other drivers like to blink their LED based on throughput. Implement this generically in mac80211, based on a throughput table the driver specifies. That way, drivers can set the blink frequencies depending on their desired behaviour and max throughput. All the drivers need to do is provide an LED class device, best with blink hardware offload. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * mac80211: make LED trigger names available earlyJohannes Berg2010-12-223-20/+22
| | | | | | | | | | | | | | | | | | | | | | | | | | The throughput trigger will require doing LED classdev/trigger handling before register_hw(), so drivers should have access to the trigger names before it. If trigger registration fails, this will still make the trigger name available, but that's not a big problem since the default trigger will the simply not be found. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| * Merge branch 'master' of ↵John W. Linville2010-12-2227-190/+524
| |\ | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-next-2.6 into for-davem Conflicts: drivers/net/wireless/iwlwifi/iwl-1000.c drivers/net/wireless/iwlwifi/iwl-6000.c drivers/net/wireless/iwlwifi/iwl-core.h
| | * mac80211: fix potentially redundant skb data copyingFelix Fietkau2010-12-201-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When an skb is shared, it needs to be duplicated, along with its data buffer. If the skb does not have enough headroom, using skb_copy might cause the data buffer to be copied twice (once by skb_copy and once by pskb_expand_head). Fix this by using skb_clone initially and letting ieee80211_skb_resize sort out the rest. Signed-off-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * mac80211: skip unnecessary pskb_expand_head callsFelix Fietkau2010-12-201-1/+3
| | | | | | | | | | | | | | | | | | | | | | | | If the skb is not cloned and we don't need any extra headroom, there is no point in reallocating the skb head. Signed-off-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * mac80211: fix initialization of skb->cb in ieee80211_subif_start_xmitFelix Fietkau2010-12-201-1/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The change 'mac80211: Fix BUG in pskb_expand_head when transmitting shared skbs' added a check for copying the skb if it's shared, however the tx info variable still points at the cb of the old skb Signed-off-by: Felix Fietkau <nbd@openwrt.org> Acked-by: Helmut Schaa <helmut.schaa@googlemail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * mac80211: Remove unused third address from mesh address extension header.Javier Cardona2010-12-203-24/+16
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The Mesh Control header only includes 0, 1 or 2 addresses. If there is one address, it should be interpreted as Address 4. If there are 2, they are interpreted as Addresses 5 and 6 (Address 4 being the 4th address in the 802.11 header). The address extension used to hold up to 3 addresses instead of the current 2. I'm not sure which draft version changed this, but it is very unlikely that it will change again given the state of the approval process of this draft. See section 7.1.3.6.3 in current draft (8.0). Also, note that the extra address that I'm removing was not being used, so this change has no effect on over-the-air frame formats. But I thought I better remove it before someone does start using it. Signed-off-by: Javier Cardona <javier@cozybit.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * nl80211: Export available antennasBruno Randolf2010-12-201-0/+5
| | | | | | | | | | | | | | | | | | | | | | | | Export the information which antennas are available for configuration as TX or RX antennas via nl80211. Signed-off-by: Bruno Randolf <br1@einfach.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * cfg80211: Separate available antennas for RX and TXBruno Randolf2010-12-201-7/+10
| | | | | | | | | | | | | | | | | | | | | | | | | | | As has been pointed out by Daniel Halperin some devices (e.g. Intel IWL5100) can only TX from a subset of RX antennas, so use separate availability masks for RX and TX. Signed-off-by: Bruno Randolf <br1@einfach.org> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * mac80211: Send mesh non-HWMP path selection frames to userspaceJavier Cardona2010-12-206-11/+21
| | | | | | | | | | | | | | | | | | | | | | | | | | | Let path selection frames for protocols other than HWMP be sent to userspace via NL80211_CMD_REGISTER_FRAME. Also allow userspace to send and receive mesh path selection frames. Signed-off-by: Javier Cardona <javier@cozybit.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * mac80211: Let userspace enable and configure vendor specific path selection.Javier Cardona2010-12-209-32/+138
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Userspace will now be allowed to toggle between the default path selection algorithm (HWMP, implemented in the kernel), and a vendor specific alternative. Also in the same patch, allow userspace to add information elements to mesh beacons. This is accordance with the Extensible Path Selection Framework specified in version 7.0 of the 802.11s draft. Signed-off-by: Javier Cardona <javier@cozybit.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * mac80211: Rename mesh_params to mesh_config to prepare for mesh_setupJavier Cardona2010-12-202-24/+24
| | | | | | | | | | | | | | | | | | | | | | | | | | | Mesh parameters can be to setup a mesh or to configure it. This patch renames the ambiguous name mesh_params to mesh_config in preparation for mesh_setup. Signed-off-by: Javier Cardona <javier@cozybit.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * cfg80211: fix null pointer dereference with a custom regulatory requestLuis R. Rodriguez2010-12-161-1/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Once we moved the core regulatory request to the queue and let the scheduler process it last_request will have been left NULL until the schedular decides to process the first request. When this happens and we are loading a driver with a custom regulatory request like all Atheros drivers we end up with a NULL pointer dereference. We fix this by checking if the request was a custom one. BUG: unable to handle kernel NULL pointer dereference at 0000000000000004 IP: [<ffffffffa016de87>] freq_reg_info_regd.clone.2+0x27/0x130 [cfg80211] PGD 71f91067 PUD 712b2067 PMD 0 Oops: 0000 [#1] PREEMPT SMP last sysfs file: /sys/devices/pci0000:00/0000:00:1d.7/usb2/2-1/firmware/2-1/loading CPU 0 Modules linked in: ath9k_htc(+) ath9k_common ath9k_hw ath <etc> Pid: 3094, comm: insmod Tainted: G W 2.6.37-rc5-wl #16 INVALID/28427ZQ RIP: 0010:[<ffffffffa016de87>] [<ffffffffa016de87>] freq_reg_info_regd.clone.2+0x27/0x130 [cfg80211] RSP: 0018:ffff88007045db78 EFLAGS: 00010282 RAX: 0000000000000000 RBX: ffffffffa047d9a0 RCX: ffff88007045dbd0 RDX: 0000000000004e20 RSI: 000000000024cde0 RDI: ffff8800700483e0 RBP: ffff88007045db98 R08: ffffffffa02f5b40 R09: 0000000000000001 R10: 000000000000000e R11: 0000000000000001 R12: 0000000000000000 R13: ffff88007004e3b0 R14: 0000000000000000 R15: ffff880070048340 FS: 00007f635a707700(0000) GS:ffff880077400000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b CR2: 0000000000000004 CR3: 00000000708a9000 CR4: 00000000000006f0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400 Process insmod (pid: 3094, threadinfo ffff88007045c000, task ffff8800713e3ec0) Stack: ffffffffa047d9a0 0000000000000000 ffff88007004e3b0 0000000000000000 ffff88007045dc08 ffffffffa016e147 000000007045dc08 0000000000000002 ffff8800700483e0 ffffffffa02f5b40 ffff88007045dbd8 0000000000000000 Call Trace: [<ffffffffa016e147>] wiphy_apply_custom_regulatory+0x137/0x1d0 [cfg80211] [<ffffffffa047a690>] ? ath9k_reg_notifier+0x0/0x50 [ath9k_htc] [<ffffffffa02f47f7>] ath_regd_init+0x347/0x430 [ath] [<ffffffffa047b1f5>] ath9k_htc_probe_device+0x6c5/0x960 [ath9k_htc] [<ffffffffa0472a2c>] ath9k_htc_hw_init+0xc/0x30 [ath9k_htc] [<ffffffffa04747e6>] ath9k_hif_usb_probe+0x216/0x3b0 [ath9k_htc] [<ffffffffa03bb6bc>] usb_probe_interface+0x10c/0x210 [usbcore] [<ffffffff812aec26>] driver_probe_device+0x96/0x1c0 [<ffffffff812aedf3>] __driver_attach+0xa3/0xb0 [<ffffffff812aed50>] ? __driver_attach+0x0/0xb0 [<ffffffff812adaae>] bus_for_each_dev+0x5e/0x90 [<ffffffff812ae8c9>] driver_attach+0x19/0x20 [<ffffffff812ae438>] bus_add_driver+0x168/0x320 [<ffffffff812af071>] driver_register+0x71/0x140 [<ffffffff811fc4a8>] ? __raw_spin_lock_init+0x38/0x70 [<ffffffffa03ba39c>] usb_register_driver+0xdc/0x190 [usbcore] [<ffffffffa03a2000>] ? ath9k_htc_init+0x0/0x4f [ath9k_htc] [<ffffffffa047499e>] ath9k_hif_usb_init+0x1e/0x20 [ath9k_htc] [<ffffffffa03a202b>] ath9k_htc_init+0x2b/0x4f [ath9k_htc] [<ffffffff8100212f>] do_one_initcall+0x3f/0x180 [<ffffffff8109ef5b>] sys_init_module+0xbb/0x200 [<ffffffff8100bf52>] system_call_fastpath+0x16/0x1b Code: <etc, who cares> RIP [<ffffffffa016de87>] freq_reg_info_regd.clone.2+0x27/0x130 [cfg80211] RSP <ffff88007045db78> CR2: 0000000000000004 ---[ end trace 79e4193601c8b713 ]--- Reported-by: Sujith Manoharan <Sujith.Manoharan@atheros.com> Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * nl80211: Add notification for dropped Deauth/DisassocJouni Malinen2010-12-164-2/+64
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add a new notification to indicate that a received, unprotected Deauthentication or Disassociation frame was dropped due to management frame protection being in use. This notification is needed to allow user space (e.g., wpa_supplicant) to implement SA Query procedure to recover from association state mismatch between an AP and STA. This is needed to avoid getting stuck in non-working state when MFP (IEEE 802.11w) is used and a protected Deauthentication or Disassociation frame is dropped for any reason. After that, the station would silently discard any unprotected Deauthentication or Disassociation frame that could be indicating that the AP does not have association for the STA (when the Reason Code would be 6 or 7). IEEE Std 802.11w-2009, 11.13 describes this recovery mechanism. Signed-off-by: Jouni Malinen <j@w1.fi> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * wireless:mac80211: kill unuse macro MESH_CFG_CMP_LEN in mesh.hShan Wei2010-12-151-11/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Commit 00d3f14c has removed the references of this macro, but left it only. So remove this definition. commit 00d3f14cf9f12c21428121026a5e1d5f65926447 Author: Johannes Berg <johannes@sipsolutions.net> Date: Tue Feb 10 21:26:00 2009 +0100 mac80211: use cfg80211s BSS infrastructure Remove all the code from mac80211 to keep track of BSSes and use the cfg80211-provided code completely. Signed-off-by: Shan Wei <shanwei@cn.fujitsu.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * mac80211: Add timeout to BA session start APISujith Manoharan2010-12-154-4/+9
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Allow drivers or rate control algorithms to specify BlockAck session timeout when initiating an ADDBA transaction. This is useful in cases where maintaining persistent BA sessions does not incur any overhead. The current timeout value of 5000 TUs is retained for all non ath9k/ath9k_htc drivers. Signed-off-by: Sujith Manoharan <Sujith.Manoharan@atheros.com> Reviewed-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * nl80211: advertise maximum remain-on-channel durationJohannes Berg2010-12-152-1/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | With the upcoming hardware offload implementation, some devices will have a different maximum duration for the remain-on-channel command. Advertise the maximum duration in mac80211, and make mac80211 set it. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * Merge branch 'master' of ↵John W. Linville2010-12-153-1/+9
| | |\ | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | git://git.kernel.org/pub/scm/linux/kernel/git/linville/wireless-2.6 Conflicts: drivers/net/wireless/iwlwifi/iwl-1000.c drivers/net/wireless/iwlwifi/iwl-6000.c drivers/net/wireless/iwlwifi/iwl-core.h
| | * | mac80211: support IBSS RSN with SW cryptoJohannes Berg2010-12-131-1/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When software crypto is used, mac80211 will support IBSS RSN, it doesn't depend on the driver in that case. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | mac80211 default tx_last_beacon false (congestion)Tim Harvey2010-12-131-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The 802.11 spec states that the STA that generated the last Beacon frame shall be the STA that response to a probe request. This is important for congestion reduction when a probe request is received - only 1 node in an adhoc BSS will transmit a response. While mac80211 drivers should provide the tx_last_beacon function to report if they transmitted the last beacon many do not. As an attempt to reduce probe response congestion default this to 0 such that a node not implementing this capability does not contribute to unnecessary congestion. In a modern medium sized office environment I see upwards of 100 probe requests per second received at a given node from various hardware/OS/drivers doing zeroconf 'active probing' as opposed to passively listening for beacons. With a modest 10-node adhoc network consisting of drivers that do not implement this tx_last_beacon feature, I have seen this result in the simultaneous xmit of probe responses accumulating to 500 probe responses per second because of collisions which brings the adhoc network to its knees as well as causes needless congestion. Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | mac80211: support separate default keysJohannes Berg2010-12-138-47/+62
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add support for split default keys (unicast and multicast) in mac80211. Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | cfg80211/nl80211: separate unicast/multicast default TX keysJohannes Berg2010-12-134-24/+115
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Allow userspace to specify that a given key is default only for unicast and/or multicast transmissions. Only WEP keys are for both, WPA/RSN keys set here are GTKs for multicast only. For more future flexibility, allow to specify all combiations. Wireless extensions can only set both so use nl80211; WEP keys (connect keys) must be set as default for both (but 802.1X WEP is still possible). Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | mac80211: clean up RX key checksJohannes Berg2010-12-131-2/+21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Using the default key for "any key set" isn't quite what we should do. It works, but with the upcoming changes it makes life unnecessarily complex, so do something better here and really check for "any key". Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | cfg80211: update information elements in cached BSS structSven Neumann2010-12-131-0/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When a cached BSS struct is updated because a new beacon was received, the code replaces the cached information elements by the IEs from the new beacon. However it did not update the pub.information_elements and pub.len_information_elements fields leaving them either pointing to the old beacon IEs or in an inconsistent state where the data is replaced by the new beacon IEs but len_information_elements still has its value from the first beacon. Fix this by updating the information elements fields if they are pointing to beacon IEs. Signed-off-by: Sven Neumann <s.neumann@raumfeld.com> Reviewed-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: John W. Linville <linville@tuxdriver.com>
| | * | cfg80211: Add antenna availability informationBruno Randolf2010-12-131-2/+13
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Add a field to wiphy for the hardware to report the availble antennas for configuration. Only if this is set to something bigger than zero, will the anntenna configuration ops be executed. Allthough this could be a simple number of antennas, I defined it as a bitmap of antennas which are available for configuration, since it's more consistent with the rest of the antenna API and there could be cases where the hardware allows only configuration of certain antennas. As it does not make much of a difference in size or normal usage, I think it's better to be able to support this, in case the need arises. The antenna configuration is now also checked against the availabe antennas and rejected if it does not match. Signed-off-by: Bruno Randolf <br1@einfach.org> -- v3: always apply available antenna mask (for "all" antennas case). v2: reject antenna configurations which don't match the available antennas Signed-off-by: John W. Linville <linville@tuxdriver.com>
* | | | tcp: cleanup of cwnd initialization in tcp_init_metrics()Jiri Kosina2010-12-231-17/+12
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Commit 86bcebafc5e7f5 ("tcp: fix >2 iw selection") fixed a case when congestion window initialization has been mistakenly omitted by introducing cwnd label and putting backwards goto from the end of the function. This makes the code unnecessarily tricky to read and understand on a first sight. Shuffle the code around a little bit to make it more obvious. Signed-off-by: Jiri Kosina <jkosina@suse.cz> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | | sfq: fix sfq class stats handlingEric Dumazet2010-12-221-5/+11
|/ / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | sfq_walk() runs without qdisc lock. By the time it selects a non empty hash slot and sfq_dump_class_stats() is run (with lock held), slot might have been freed : We then access q->slots[SFQ_EMPTY_SLOT], out of bounds, and crash in slot_queue_walk() On previous kernels, bug is here but out of bounds qs[SFQ_DEPTH] and allot[SFQ_DEPTH] are located in struct sfq_sched_data, so no illegal memory access happens, only possibly wrong data reported to user. Also, slot_dequeue_tail() should make sure slot skb chain is correctly terminated, or sfq_dump_class_stats() can access freed skbs. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | filter: optimize accesses to ancillary dataEric Dumazet2010-12-211-28/+44
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We can translate pseudo load instructions at filter check time to dedicated instructions to speed up filtering and avoid one switch(). libpcap currently uses SKF_AD_PROTOCOL, but custom filters probably use other ancillary accesses. Note : I made the assertion that ancillary data was always accessed with BPF_LD|BPF_?|BPF_ABS instructions, not with BPF_LD|BPF_?|BPF_IND ones (offset given by K constant, not by K + X register) On x86_64, this saves a few bytes of text : # size net/core/filter.o.* text data bss dec hex filename 4864 0 0 4864 1300 net/core/filter.o.new 4944 0 0 4944 1350 net/core/filter.o.old Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | net: timestamp cloned packet in dev_queue_xmit_nitEric Dumazet2010-12-211-7/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Le vendredi 17 décembre 2010 à 10:26 +0100, Eric Dumazet a écrit : > > I think we can add this after latest Changli patch : > > He does one skb_clone() before calling the sniffers. > We could set timestamp on this clone, instead of original skb. > > Problem solved. > [PATCH net-next-2.6] net: timestamp cloned packet in dev_queue_xmit_nit Now we do one clone of skb if at least one sniffer might take packet, we also can do the skb timestamping on the clone and let original packet unchanged. This is a generalization of commit 8caf153974f2 (net: sch_netem: Fix an inconsistency in ingress netem timestamps.) This way, we can have a good idea when packets are delivered to our stack (tcpdump -i ifb0), while a tcpdump on original device gives timestamps right before ingressing. This also speedup our stack, avoiding taking timestamps if not needed. Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Cc: Changli Gao <xiaosuo@gmail.com> Cc: Patrick McHardy <kaber@trash.net> Cc: Jarek Poplawski <jarkao2@gmail.com> Acked-by: Changli Gao <xiaosuo@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | TCP: increase default initial receive window.Nandita Dukkipati2010-12-201-3/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch changes the default initial receive window to 10 mss (defined constant). The default window is limited to the maximum of 10*1460 and 2*mss (when mss > 1460). draft-ietf-tcpm-initcwnd-00 is a proposal to the IETF that recommends increasing TCP's initial congestion window to 10 mss or about 15KB. Leading up to this proposal were several large-scale live Internet experiments with an initial congestion window of 10 mss (IW10), where we showed that the average latency of HTTP responses improved by approximately 10%. This was accompanied by a slight increase in retransmission rate (0.5%), most of which is coming from applications opening multiple simultaneous connections. To understand the extreme worst case scenarios, and fairness issues (IW10 versus IW3), we further conducted controlled testbed experiments. We came away finding minimal negative impact even under low link bandwidths (dial-ups) and small buffers. These results are extremely encouraging to adopting IW10. However, an initial congestion window of 10 mss is useless unless a TCP receiver advertises an initial receive window of at least 10 mss. Fortunately, in the large-scale Internet experiments we found that most widely used operating systems advertised large initial receive windows of 64KB, allowing us to experiment with a wide range of initial congestion windows. Linux systems were among the few exceptions that advertised a small receive window of 6KB. The purpose of this patch is to fix this shortcoming. References: 1. A comprehensive list of all IW10 references to date. http://code.google.com/speed/protocols/tcpm-IW10.html 2. Paper describing results from large-scale Internet experiments with IW10. http://ccr.sigcomm.org/drupal/?q=node/621 3. Controlled testbed experiments under worst case scenarios and a fairness study. http://www.ietf.org/proceedings/79/slides/tcpm-0.pdf 4. Raw test data from testbed experiments (Linux senders/receivers) with initial congestion and receive windows of both 10 mss. http://research.csc.ncsu.edu/netsrv/?q=content/iw10 5. Internet-Draft. Increasing TCP's Initial Window. https://datatracker.ietf.org/doc/draft-ietf-tcpm-initcwnd/ Signed-off-by: Nandita Dukkipati <nanditad@google.com> Acked-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | net_sched: sch_sfq: better struct layoutsEric Dumazet2010-12-201-98/+162
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Here is a respin of patch. I'll send a short patch to make SFQ more fair in presence of large packets as well. Thanks [PATCH v3 net-next-2.6] net_sched: sch_sfq: better struct layouts This patch shrinks sizeof(struct sfq_sched_data) from 0x14f8 (or more if spinlocks are bigger) to 0x1180 bytes, and reduce text size as well. text data bss dec hex filename 4821 152 0 4973 136d old/net/sched/sch_sfq.o 4627 136 0 4763 129b new/net/sched/sch_sfq.o All data for a slot/flow is now grouped in a compact and cache friendly structure, instead of being spreaded in many different points. struct sfq_slot { struct sk_buff *skblist_next; struct sk_buff *skblist_prev; sfq_index qlen; /* number of skbs in skblist */ sfq_index next; /* next slot in sfq chain */ struct sfq_head dep; /* anchor in dep[] chains */ unsigned short hash; /* hash value (index in ht[]) */ short allot; /* credit for this slot */ }; Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Cc: Jarek Poplawski <jarkao2@gmail.com> Cc: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | Merge branch 'master' of ↵David S. Miller2010-12-203-23/+25
|\ \ \ | | | | | | | | | | | | master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6
| * | | net_sched: sch_sfq: fix allot handlingEric Dumazet2010-12-201-12/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When deploying SFQ/IFB here at work, I found the allot management was pretty wrong in sfq, even changing allot from short to int... We should init allot for each new flow, not using a previous value found in slot. Before patch, I saw bursts of several packets per flow, apparently denying the default "quantum 1514" limit I had on my SFQ class. class sfq 11:1 parent 11: (dropped 0, overlimits 0 requeues 0) backlog 0b 7p requeues 0 allot 11546 class sfq 11:46 parent 11: (dropped 0, overlimits 0 requeues 0) backlog 0b 1p requeues 0 allot -23873 class sfq 11:78 parent 11: (dropped 0, overlimits 0 requeues 0) backlog 0b 5p requeues 0 allot 11393 After patch, better fairness among each flow, allot limit being respected, allot is positive : class sfq 11:e parent 11: (dropped 0, overlimits 0 requeues 86) backlog 0b 3p requeues 86 allot 596 class sfq 11:94 parent 11: (dropped 0, overlimits 0 requeues 0) backlog 0b 3p requeues 0 allot 1468 class sfq 11:a4 parent 11: (dropped 0, overlimits 0 requeues 0) backlog 0b 4p requeues 0 allot 650 class sfq 11:bb parent 11: (dropped 0, overlimits 0 requeues 0) backlog 0b 3p requeues 0 allot 596 Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
| * | | ipv6: Fragment locally generated tunnel-mode IPSec6 packets as needed.David Stevens2010-12-192-11/+17
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch modifies IPsec6 to fragment IPv6 packets that are locally generated as needed. This version of the patch only fragments in tunnel mode, so that fragment headers will not be obscured by ESP in transport mode. Signed-off-by: David L Stevens <dlstevens@us.ibm.com> Acked-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | | net_sched: sch_sfq: add backlog info in sfq_dump_class_stats()Eric Dumazet2010-12-201-1/+6
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | We currently return for each active SFQ slot the number of packets in queue. We can also give number of bytes accounted for these packets. tc -s class show dev ifb0 Before patch : class sfq 11:3d9 parent 11: (dropped 0, overlimits 0 requeues 0) backlog 0b 3p requeues 0 allot 1266 After patch : class sfq 11:3e4 parent 11: (dropped 0, overlimits 0 requeues 0) backlog 4380b 3p requeues 0 allot 1212 Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | | ipv4: Flush per-ns routing cache more sanely.David S. Miller2010-12-202-41/+29
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Flush the routing cache only of entries that match the network namespace in which the purge event occurred. Signed-off-by: David S. Miller <davem@davemloft.net> Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
* | | | batman-adv: Return hna count on local buffer fillSven Eckelmann2010-12-201-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | hna_local_fill_buffer must return the number of added hna entries and not the last checked hash bucket. Signed-off-by: Sven Eckelmann <sven@narfation.org> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | | pktgen: Remove unnecessary prefix from pr_<level>Joe Perches2010-12-201-2/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Remove "pktgen: " prefix string from one pr_info. pr_fmt adds it, so this is a duplicate. Signed-off-by: Joe Perches <joe@perches.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | | net: kill unused macrosShan Wei2010-12-196-7/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | These macros never be used, so remove them. Signed-off-by: Shan Wei <shanwei@cn.fujitsu.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | | net: increase skb->users instead of skb_clone()Changli Gao2010-12-191-10/+20
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | In dev_queue_xmit_nit(), we have to clone skbs as we need to mangle skbs, however, we don't need to clone skbs for all the packet_types. Except for the first packet_type, we increase skb->users instead of skb_clone(). Signed-off-by: Changli Gao <xiaosuo@gmail.com> Acked-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | | ipv6: remove duplicate neigh_ifdownstephen hemminger2010-12-181-1/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When device is being set to down, neigh_ifdown was being called twice. Once from addrconf notifier and once from ndisc notifier. Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | | ipv6: fib6_ifdown cleanupstephen hemminger2010-12-181-4/+4
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Remove (unnecessary) casts to make code cleaner. Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Signed-off-by: David S. Miller <davem@davemloft.net>
* | | | Merge branch 'master' of ↵David S. Miller2010-12-1719-28/+101
|\ \ \ \ | |/ / / | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | master.kernel.org:/pub/scm/linux/kernel/git/davem/net-2.6 Conflicts: drivers/net/bnx2x/bnx2x.h drivers/net/wireless/iwlwifi/iwl-1000.c drivers/net/wireless/iwlwifi/iwl-6000.c drivers/net/wireless/iwlwifi/iwl-core.h drivers/vhost/vhost.c
| * | | ipv6: don't flush routes when setting loopback downstephen hemminger2010-12-161-1/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | When loopback device is being brought down, then keep the route table entries because they are special. The entries in the local table for linklocal routes and ::1 address should not be purged. This is a sub optimal solution to the problem and should be replaced by a better fix in future. Signed-off-by: Stephen Hemminger <shemminger@vyatta.com> Acked-by: "Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by: David S. Miller <davem@davemloft.net>
| * | | sctp: fix the return value of getting the sctp partial delivery pointWei Yongjun2010-12-161-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Get the sctp partial delivery point using SCTP_PARTIAL_DELIVERY_POINT socket option should return 0 if success, not -ENOTSUPP. Signed-off-by: Wei Yongjun <yjwei@cn.fujitsu.com> Acked-by: Vlad Yasevich <vladislav.yasevich@hp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
| * | | bridge: fix IPv6 queries for bridge multicast snoopingDavid Stevens2010-12-161-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch fixes a missing ntohs() for bridge IPv6 multicast snooping. Signed-off-by: David L Stevens <dlstevens@us.ibm.com> Acked-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
| * | | net: fix nulls list corruptions in sk_prot_allocOctavian Purdila2010-12-165-12/+39
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Special care is taken inside sk_port_alloc to avoid overwriting skc_node/skc_nulls_node. We should also avoid overwriting skc_bind_node/skc_portaddr_node. The patch fixes the following crash: BUG: unable to handle kernel paging request at fffffffffffffff0 IP: [<ffffffff812ec6dd>] udp4_lib_lookup2+0xad/0x370 [<ffffffff812ecc22>] __udp4_lib_lookup+0x282/0x360 [<ffffffff812ed63e>] __udp4_lib_rcv+0x31e/0x700 [<ffffffff812bba45>] ? ip_local_deliver_finish+0x65/0x190 [<ffffffff812bbbf8>] ? ip_local_deliver+0x88/0xa0 [<ffffffff812eda35>] udp_rcv+0x15/0x20 [<ffffffff812bba45>] ip_local_deliver_finish+0x65/0x190 [<ffffffff812bbbf8>] ip_local_deliver+0x88/0xa0 [<ffffffff812bb2cd>] ip_rcv_finish+0x32d/0x6f0 [<ffffffff8128c14c>] ? netif_receive_skb+0x99c/0x11c0 [<ffffffff812bb94b>] ip_rcv+0x2bb/0x350 [<ffffffff8128c14c>] netif_receive_skb+0x99c/0x11c0 Signed-off-by: Leonard Crestez <lcrestez@ixiacom.com> Signed-off-by: Octavian Purdila <opurdila@ixiacom.com> Acked-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
OpenPOWER on IntegriCloud