From 39289bfc221423b27570e7c9157b690828e786cb Mon Sep 17 00:00:00 2001 From: Jason Gunthorpe Date: Mon, 22 Jul 2019 17:01:30 +0000 Subject: RDMA: Make most headers compile stand alone So that rdma can work with CONFIG_KERNEL_HEADER_TEST and CONFIG_HEADERS_CHECK. Link: https://lore.kernel.org/r/20190722170126.GA16453@ziepe.ca Signed-off-by: Jason Gunthorpe --- include/Kbuild | 6 ------ include/rdma/ib.h | 2 ++ include/rdma/iw_portmap.h | 3 +++ include/rdma/opa_port_info.h | 2 ++ include/rdma/rdmavt_cq.h | 1 + include/rdma/signature.h | 2 ++ 6 files changed, 10 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/include/Kbuild b/include/Kbuild index c38f0d46b267..fc2aa4e20658 100644 --- a/include/Kbuild +++ b/include/Kbuild @@ -945,12 +945,6 @@ header-test- += net/xdp.h header-test- += net/xdp_priv.h header-test- += pcmcia/cistpl.h header-test- += pcmcia/ds.h -header-test- += rdma/ib.h -header-test- += rdma/iw_portmap.h -header-test- += rdma/opa_port_info.h -header-test- += rdma/rdmavt_cq.h -header-test- += rdma/restrack.h -header-test- += rdma/signature.h header-test- += rdma/tid_rdma_defs.h header-test- += scsi/fc/fc_encaps.h header-test- += scsi/fc/fc_fc2.h diff --git a/include/rdma/ib.h b/include/rdma/ib.h index 4f385ec54f80..fe2fc9e91588 100644 --- a/include/rdma/ib.h +++ b/include/rdma/ib.h @@ -36,6 +36,8 @@ #include #include #include +#include +#include struct ib_addr { union { diff --git a/include/rdma/iw_portmap.h b/include/rdma/iw_portmap.h index b9fee7feeeb5..c89535047c42 100644 --- a/include/rdma/iw_portmap.h +++ b/include/rdma/iw_portmap.h @@ -33,6 +33,9 @@ #ifndef _IW_PORTMAP_H #define _IW_PORTMAP_H +#include +#include + #define IWPM_ULIBNAME_SIZE 32 #define IWPM_DEVNAME_SIZE 32 #define IWPM_IFNAME_SIZE 16 diff --git a/include/rdma/opa_port_info.h b/include/rdma/opa_port_info.h index 7147a9263011..bdbfe25d3854 100644 --- a/include/rdma/opa_port_info.h +++ b/include/rdma/opa_port_info.h @@ -33,6 +33,8 @@ #if !defined(OPA_PORT_INFO_H) #define OPA_PORT_INFO_H +#include + #define OPA_PORT_LINK_MODE_NOP 0 /* No change */ #define OPA_PORT_LINK_MODE_OPA 4 /* Port mode is OPA */ diff --git a/include/rdma/rdmavt_cq.h b/include/rdma/rdmavt_cq.h index 04c519ef6d71..574eb7278f46 100644 --- a/include/rdma/rdmavt_cq.h +++ b/include/rdma/rdmavt_cq.h @@ -53,6 +53,7 @@ #include #include +#include /* * Define an ib_cq_notify value that is not valid so we know when CQ diff --git a/include/rdma/signature.h b/include/rdma/signature.h index f24cc2a1d3c5..d16b0fcc8344 100644 --- a/include/rdma/signature.h +++ b/include/rdma/signature.h @@ -6,6 +6,8 @@ #ifndef _RDMA_SIGNATURE_H_ #define _RDMA_SIGNATURE_H_ +#include + enum ib_signature_prot_cap { IB_PROT_T10DIF_TYPE_1 = 1, IB_PROT_T10DIF_TYPE_2 = 1 << 1, -- cgit v1.2.1 From 1d2fedd8561dc469a7503855ee602f4bb3eccfa7 Mon Sep 17 00:00:00 2001 From: Parav Pandit Date: Tue, 23 Jul 2019 10:02:05 +0300 Subject: RDMA/core: Support netlink commands in non init_net net namespaces Now that IB core supports RDMA device binding with specific net namespace, enable IB core to accept netlink commands in non init_net namespaces. This is done by having per net namespace netlink socket. At present only netlink device handling client RDMA_NL_NLDEV supports device handling in multiple net namespaces. Hence do not accept netlink messages for other clients in non init_net net namespaces. Link: https://lore.kernel.org/r/20190723070205.6247-1-leon@kernel.org Signed-off-by: Parav Pandit Signed-off-by: Leon Romanovsky Reviewed-by: Jason Gunthorpe Signed-off-by: Jason Gunthorpe --- include/rdma/rdma_netlink.h | 10 +++++++--- 1 file changed, 7 insertions(+), 3 deletions(-) (limited to 'include') diff --git a/include/rdma/rdma_netlink.h b/include/rdma/rdma_netlink.h index 6631624e4d7c..ab22759de7ea 100644 --- a/include/rdma/rdma_netlink.h +++ b/include/rdma/rdma_netlink.h @@ -76,28 +76,32 @@ int ibnl_put_attr(struct sk_buff *skb, struct nlmsghdr *nlh, /** * Send the supplied skb to a specific userspace PID. + * @net: Net namespace in which to send the skb * @skb: The netlink skb * @pid: Userspace netlink process ID * Returns 0 on success or a negative error code. */ -int rdma_nl_unicast(struct sk_buff *skb, u32 pid); +int rdma_nl_unicast(struct net *net, struct sk_buff *skb, u32 pid); /** * Send, with wait/1 retry, the supplied skb to a specific userspace PID. + * @net: Net namespace in which to send the skb * @skb: The netlink skb * @pid: Userspace netlink process ID * Returns 0 on success or a negative error code. */ -int rdma_nl_unicast_wait(struct sk_buff *skb, __u32 pid); +int rdma_nl_unicast_wait(struct net *net, struct sk_buff *skb, __u32 pid); /** * Send the supplied skb to a netlink group. + * @net: Net namespace in which to send the skb * @skb: The netlink skb * @group: Netlink group ID * @flags: allocation flags * Returns 0 on success or a negative error code. */ -int rdma_nl_multicast(struct sk_buff *skb, unsigned int group, gfp_t flags); +int rdma_nl_multicast(struct net *net, struct sk_buff *skb, + unsigned int group, gfp_t flags); /** * Check if there are any listeners to the netlink group -- cgit v1.2.1 From 0058eb589881056b49a4ba15dfa3f1b8db53991c Mon Sep 17 00:00:00 2001 From: Michal Kalderon Date: Tue, 9 Jul 2019 17:17:33 +0300 Subject: qed*: Change dpi_addr to be denoted with __iomem Several casts were required around dpi_addr parameter in qed_rdma_if.h This is an address on the doorbell bar and should therefore be marked with __iomem. Link: https://lore.kernel.org/r/20190709141735.19193-5-michal.kalderon@marvell.com Reported-by: Jason Gunthorpe Signed-off-by: Ariel Elior Signed-off-by: Michal Kalderon Signed-off-by: Jason Gunthorpe --- include/linux/qed/qed_rdma_if.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/linux/qed/qed_rdma_if.h b/include/linux/qed/qed_rdma_if.h index 898f595ea3d6..74efca15fde7 100644 --- a/include/linux/qed/qed_rdma_if.h +++ b/include/linux/qed/qed_rdma_if.h @@ -225,7 +225,7 @@ struct qed_rdma_start_in_params { struct qed_rdma_add_user_out_params { u16 dpi; - u64 dpi_addr; + void __iomem *dpi_addr; u64 dpi_phys_addr; u32 dpi_size; u16 wid_count; -- cgit v1.2.1 From 20cf4e026730104892fa1268de0371a631cee294 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 29 Jul 2019 13:22:09 -0400 Subject: rdma: Enable ib_alloc_cq to spread work over a device's comp_vectors MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Send and Receive completion is handled on a single CPU selected at the time each Completion Queue is allocated. Typically this is when an initiator instantiates an RDMA transport, or when a target accepts an RDMA connection. Some ULPs cannot open a connection per CPU to spread completion workload across available CPUs and MSI vectors. For such ULPs, provide an API that allows the RDMA core to select a completion vector based on the device's complement of available comp_vecs. ULPs that invoke ib_alloc_cq() with only comp_vector 0 are converted to use the new API so that their completion workloads interfere less with each other. Suggested-by: Håkon Bugge Signed-off-by: Chuck Lever Reviewed-by: Leon Romanovsky Cc: Cc: Link: https://lore.kernel.org/r/20190729171923.13428.52555.stgit@manet.1015granger.net Signed-off-by: Doug Ledford --- include/rdma/ib_verbs.h | 19 +++++++++++++++++++ 1 file changed, 19 insertions(+) (limited to 'include') diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index c5f8a9f17063..2a1523ccd7ab 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -3711,6 +3711,25 @@ static inline struct ib_cq *ib_alloc_cq(struct ib_device *dev, void *private, NULL); } +struct ib_cq *__ib_alloc_cq_any(struct ib_device *dev, void *private, + int nr_cqe, enum ib_poll_context poll_ctx, + const char *caller); + +/** + * ib_alloc_cq_any: Allocate kernel CQ + * @dev: The IB device + * @private: Private data attached to the CQE + * @nr_cqe: Number of CQEs in the CQ + * @poll_ctx: Context used for polling the CQ + */ +static inline struct ib_cq *ib_alloc_cq_any(struct ib_device *dev, + void *private, int nr_cqe, + enum ib_poll_context poll_ctx) +{ + return __ib_alloc_cq_any(dev, private, nr_cqe, poll_ctx, + KBUILD_MODNAME); +} + /** * ib_free_cq_user - Free kernel/user CQ * @cq: The CQ to free -- cgit v1.2.1 From 05bb411ada9508b48044ef5d84dd8bc46cece607 Mon Sep 17 00:00:00 2001 From: Gal Pressman Date: Thu, 1 Aug 2019 20:14:46 +0300 Subject: RDMA/core: Introduce ratelimited ibdev printk functions Add ratelimited helpers to the ibdev_* printk functions. Implementation inspired by counterpart dev_*_ratelimited functions. Signed-off-by: Gal Pressman Reviewed-by: Leon Romanovsky Link: https://lore.kernel.org/r/20190801171447.54440-2-galpress@amazon.com Signed-off-by: Doug Ledford --- include/rdma/ib_verbs.h | 42 ++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 42 insertions(+) (limited to 'include') diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index 2a1523ccd7ab..0ecda7d15df2 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -107,6 +107,48 @@ static inline void ibdev_dbg(const struct ib_device *ibdev, const char *format, ...) {} #endif +#define ibdev_level_ratelimited(ibdev_level, ibdev, fmt, ...) \ +do { \ + static DEFINE_RATELIMIT_STATE(_rs, \ + DEFAULT_RATELIMIT_INTERVAL, \ + DEFAULT_RATELIMIT_BURST); \ + if (__ratelimit(&_rs)) \ + ibdev_level(ibdev, fmt, ##__VA_ARGS__); \ +} while (0) + +#define ibdev_emerg_ratelimited(ibdev, fmt, ...) \ + ibdev_level_ratelimited(ibdev_emerg, ibdev, fmt, ##__VA_ARGS__) +#define ibdev_alert_ratelimited(ibdev, fmt, ...) \ + ibdev_level_ratelimited(ibdev_alert, ibdev, fmt, ##__VA_ARGS__) +#define ibdev_crit_ratelimited(ibdev, fmt, ...) \ + ibdev_level_ratelimited(ibdev_crit, ibdev, fmt, ##__VA_ARGS__) +#define ibdev_err_ratelimited(ibdev, fmt, ...) \ + ibdev_level_ratelimited(ibdev_err, ibdev, fmt, ##__VA_ARGS__) +#define ibdev_warn_ratelimited(ibdev, fmt, ...) \ + ibdev_level_ratelimited(ibdev_warn, ibdev, fmt, ##__VA_ARGS__) +#define ibdev_notice_ratelimited(ibdev, fmt, ...) \ + ibdev_level_ratelimited(ibdev_notice, ibdev, fmt, ##__VA_ARGS__) +#define ibdev_info_ratelimited(ibdev, fmt, ...) \ + ibdev_level_ratelimited(ibdev_info, ibdev, fmt, ##__VA_ARGS__) + +#if defined(CONFIG_DYNAMIC_DEBUG) +/* descriptor check is first to prevent flooding with "callbacks suppressed" */ +#define ibdev_dbg_ratelimited(ibdev, fmt, ...) \ +do { \ + static DEFINE_RATELIMIT_STATE(_rs, \ + DEFAULT_RATELIMIT_INTERVAL, \ + DEFAULT_RATELIMIT_BURST); \ + DEFINE_DYNAMIC_DEBUG_METADATA(descriptor, fmt); \ + if (DYNAMIC_DEBUG_BRANCH(descriptor) && __ratelimit(&_rs)) \ + __dynamic_ibdev_dbg(&descriptor, ibdev, fmt, \ + ##__VA_ARGS__); \ +} while (0) +#else +__printf(2, 3) __cold +static inline +void ibdev_dbg_ratelimited(const struct ib_device *ibdev, const char *format, ...) {} +#endif + union ib_gid { u8 raw[16]; struct { -- cgit v1.2.1 From 72a7720fca37fec0daf295923f17ac5d88a613e1 Mon Sep 17 00:00:00 2001 From: Kamal Heib Date: Wed, 7 Aug 2019 13:31:35 +0300 Subject: RDMA: Introduce ib_port_phys_state enum In order to improve readability, add ib_port_phys_state enum to replace the use of magic numbers. Signed-off-by: Kamal Heib Reviewed-by: Andrew Boyer Acked-by: Michal Kalderon Acked-by: Bernard Metzler Link: https://lore.kernel.org/r/20190807103138.17219-2-kamalheib1@gmail.com Signed-off-by: Doug Ledford --- include/rdma/ib_verbs.h | 10 ++++++++++ 1 file changed, 10 insertions(+) (limited to 'include') diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index 0ecda7d15df2..391499008a22 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -493,6 +493,16 @@ enum ib_port_state { IB_PORT_ACTIVE_DEFER = 5 }; +enum ib_port_phys_state { + IB_PORT_PHYS_STATE_SLEEP = 1, + IB_PORT_PHYS_STATE_POLLING = 2, + IB_PORT_PHYS_STATE_DISABLED = 3, + IB_PORT_PHYS_STATE_PORT_CONFIGURATION_TRAINING = 4, + IB_PORT_PHYS_STATE_LINK_UP = 5, + IB_PORT_PHYS_STATE_LINK_ERROR_RECOVERY = 6, + IB_PORT_PHYS_STATE_PHY_TEST = 7, +}; + enum ib_port_width { IB_WIDTH_1X = 1, IB_WIDTH_2X = 16, -- cgit v1.2.1 From 972d7560ee37d48a9121ca1b7796d107a16784e7 Mon Sep 17 00:00:00 2001 From: Yishai Hadas Date: Thu, 8 Aug 2019 11:43:57 +0300 Subject: IB/mlx5: Add legacy events to DEVX list Add two events that were defined in the device specification but were not exposed in the driver list. Post this patch those events can be read over the DEVX events interface once be reported by the firmware. Signed-off-by: Yishai Hadas Reviewed-by: Edward Srouji Signed-off-by: Leon Romanovsky Link: https://lore.kernel.org/r/20190808084358.29517-4-leon@kernel.org Signed-off-by: Doug Ledford --- include/linux/mlx5/device.h | 9 +++++++++ 1 file changed, 9 insertions(+) (limited to 'include') diff --git a/include/linux/mlx5/device.h b/include/linux/mlx5/device.h index ce9839c8bc1a..e427af260ebe 100644 --- a/include/linux/mlx5/device.h +++ b/include/linux/mlx5/device.h @@ -328,6 +328,7 @@ enum mlx5_event { MLX5_EVENT_TYPE_GPIO_EVENT = 0x15, MLX5_EVENT_TYPE_PORT_MODULE_EVENT = 0x16, MLX5_EVENT_TYPE_TEMP_WARN_EVENT = 0x17, + MLX5_EVENT_TYPE_XRQ_ERROR = 0x18, MLX5_EVENT_TYPE_REMOTE_CONFIG = 0x19, MLX5_EVENT_TYPE_GENERAL_EVENT = 0x22, MLX5_EVENT_TYPE_MONITOR_COUNTER = 0x24, @@ -345,6 +346,7 @@ enum mlx5_event { MLX5_EVENT_TYPE_ESW_FUNCTIONS_CHANGED = 0xe, MLX5_EVENT_TYPE_DCT_DRAINED = 0x1c, + MLX5_EVENT_TYPE_DCT_KEY_VIOLATION = 0x1d, MLX5_EVENT_TYPE_FPGA_ERROR = 0x20, MLX5_EVENT_TYPE_FPGA_QP_ERROR = 0x21, @@ -584,6 +586,12 @@ struct mlx5_eqe_cq_err { u8 syndrome; }; +struct mlx5_eqe_xrq_err { + __be32 reserved1[5]; + __be32 type_xrqn; + __be32 reserved2; +}; + struct mlx5_eqe_port_state { u8 reserved0[8]; u8 port; @@ -698,6 +706,7 @@ union ev_data { struct mlx5_eqe_pps pps; struct mlx5_eqe_dct dct; struct mlx5_eqe_temp_warning temp_warning; + struct mlx5_eqe_xrq_err xrq_err; } __packed; struct mlx5_eqe { -- cgit v1.2.1 From b2299e83815c59ab59c4ee4fb4842b3b28e5072f Mon Sep 17 00:00:00 2001 From: Leon Romanovsky Date: Mon, 19 Aug 2019 14:45:47 +0300 Subject: RDMA: Delete DEBUG code There is no need to keep DEBUG defines for out-of-the tree testing. Signed-off-by: Leon Romanovsky Link: https://lore.kernel.org/r/20190819114547.20704-1-leon@kernel.org Signed-off-by: Doug Ledford --- include/rdma/ib_verbs.h | 3 --- 1 file changed, 3 deletions(-) (limited to 'include') diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index 391499008a22..08e966c8081a 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -98,9 +98,6 @@ void ibdev_info(const struct ib_device *ibdev, const char *format, ...); #if defined(CONFIG_DYNAMIC_DEBUG) #define ibdev_dbg(__dev, format, args...) \ dynamic_ibdev_dbg(__dev, format, ##args) -#elif defined(DEBUG) -#define ibdev_dbg(__dev, format, args...) \ - ibdev_printk(KERN_DEBUG, __dev, format, ##args) #else __printf(2, 3) __cold static inline -- cgit v1.2.1 From d8abe88450beb96a66e434323eb6ab737654b840 Mon Sep 17 00:00:00 2001 From: Mark Zhang Date: Mon, 19 Aug 2019 14:36:26 +0300 Subject: RDMA/mlx5: RDMA_RX flow type support for user applications Currently user applications can only steer TCP/IP(NIC RX/RX) traffic. This patch adds RDMA_RX as a new flow type to allow the user to insert steering rules to control RDMA traffic. Two destinations are supported(but not set at the same time): devx flow table object and QP. Signed-off-by: Mark Zhang Signed-off-by: Leon Romanovsky Link: https://lore.kernel.org/r/20190819113626.20284-4-leon@kernel.org Signed-off-by: Doug Ledford --- include/uapi/rdma/mlx5_user_ioctl_verbs.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include') diff --git a/include/uapi/rdma/mlx5_user_ioctl_verbs.h b/include/uapi/rdma/mlx5_user_ioctl_verbs.h index 7e9900b0e746..88b6ca70c2fe 100644 --- a/include/uapi/rdma/mlx5_user_ioctl_verbs.h +++ b/include/uapi/rdma/mlx5_user_ioctl_verbs.h @@ -43,6 +43,7 @@ enum mlx5_ib_uapi_flow_table_type { MLX5_IB_UAPI_FLOW_TABLE_TYPE_NIC_RX = 0x0, MLX5_IB_UAPI_FLOW_TABLE_TYPE_NIC_TX = 0x1, MLX5_IB_UAPI_FLOW_TABLE_TYPE_FDB = 0x2, + MLX5_IB_UAPI_FLOW_TABLE_TYPE_RDMA_RX = 0x3, }; enum mlx5_ib_uapi_flow_action_packet_reformat_type { -- cgit v1.2.1 From 7199435414868bd656284349edc1a1f528fe3662 Mon Sep 17 00:00:00 2001 From: Kaike Wan Date: Wed, 11 Sep 2019 07:30:47 -0400 Subject: IB/{rdmavt, hfi1, qib}: Add a counter for credit waits This patch adds a counter for credit waits to assist field debugging. Link: https://lore.kernel.org/r/20190911113047.126040.10857.stgit@awfm-01.aw.intel.com Reviewed-by: Mike Marciniszyn Signed-off-by: Kaike Wan Signed-off-by: Mike Marciniszyn Signed-off-by: Dennis Dalessandro Signed-off-by: Jason Gunthorpe --- include/rdma/rdma_vt.h | 1 + include/rdma/rdmavt_qp.h | 35 +++++++++++++++++++++++++++++++++++ 2 files changed, 36 insertions(+) (limited to 'include') diff --git a/include/rdma/rdma_vt.h b/include/rdma/rdma_vt.h index 525848e227dc..ac5a9430abb6 100644 --- a/include/rdma/rdma_vt.h +++ b/include/rdma/rdma_vt.h @@ -116,6 +116,7 @@ struct rvt_ibport { u64 n_unaligned; u64 n_rc_dupreq; u64 n_rc_seqnak; + u64 n_rc_crwaits; u16 pkey_violations; u16 qkey_violations; u16 mkey_violations; diff --git a/include/rdma/rdmavt_qp.h b/include/rdma/rdmavt_qp.h index e06c77d76463..b550ae89bf85 100644 --- a/include/rdma/rdmavt_qp.h +++ b/include/rdma/rdmavt_qp.h @@ -973,6 +973,41 @@ static inline void rvt_free_rq(struct rvt_rq *rq) rq->wq = NULL; } +/** + * rvt_to_iport - Get the ibport pointer + * @qp: the qp pointer + * + * This function returns the ibport pointer from the qp pointer. + */ +static inline struct rvt_ibport *rvt_to_iport(struct rvt_qp *qp) +{ + struct rvt_dev_info *rdi = ib_to_rvt(qp->ibqp.device); + + return rdi->ports[qp->port_num - 1]; +} + +/** + * rvt_rc_credit_avail - Check if there are enough RC credits for the request + * @qp: the qp + * @wqe: the request + * + * This function returns false when there are not enough credits for the given + * request and true otherwise. + */ +static inline bool rvt_rc_credit_avail(struct rvt_qp *qp, struct rvt_swqe *wqe) +{ + lockdep_assert_held(&qp->s_lock); + if (!(qp->s_flags & RVT_S_UNLIMITED_CREDIT) && + rvt_cmp_msn(wqe->ssn, qp->s_lsn + 1) > 0) { + struct rvt_ibport *rvp = rvt_to_iport(qp); + + qp->s_flags |= RVT_S_WAIT_SSN_CREDIT; + rvp->n_rc_crwaits++; + return false; + } + return true; +} + struct rvt_qp_iter *rvt_qp_iter_init(struct rvt_dev_info *rdi, u64 v, void (*cb)(struct rvt_qp *qp, u64 v)); -- cgit v1.2.1