Index: projects/bsd_rdma_4_9_stable_11/sys/dev/mlx5/mlx5_ib/mlx5_ib.h =================================================================== --- projects/bsd_rdma_4_9_stable_11/sys/dev/mlx5/mlx5_ib/mlx5_ib.h (revision 331011) +++ projects/bsd_rdma_4_9_stable_11/sys/dev/mlx5/mlx5_ib/mlx5_ib.h (revision 331012) @@ -1,993 +1,998 @@ /*- * Copyright (c) 2013-2015, Mellanox Technologies, Ltd. All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * * THIS SOFTWARE IS PROVIDED BY AUTHOR AND CONTRIBUTORS `AS IS' AND * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL AUTHOR OR CONTRIBUTORS BE LIABLE * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF * SUCH DAMAGE. * * $FreeBSD$ */ #ifndef MLX5_IB_H #define MLX5_IB_H #include #include #include #include #include #include #include #include #include #include #include #include #define mlx5_ib_dbg(dev, format, arg...) \ pr_debug("%s:%s:%d:(pid %d): " format, (dev)->ib_dev.name, __func__, \ __LINE__, current->pid, ##arg) #define mlx5_ib_err(dev, format, arg...) \ pr_err("%s:%s:%d:(pid %d): " format, (dev)->ib_dev.name, __func__, \ __LINE__, current->pid, ##arg) #define mlx5_ib_warn(dev, format, arg...) \ pr_warn("%s:%s:%d:(pid %d): " format, (dev)->ib_dev.name, __func__, \ __LINE__, current->pid, ##arg) #define field_avail(type, fld, sz) (offsetof(type, fld) + \ sizeof(((type *)0)->fld) <= (sz)) #define MLX5_IB_DEFAULT_UIDX 0xffffff #define MLX5_USER_ASSIGNED_UIDX_MASK __mlx5_mask(qpc, user_index) enum { MLX5_IB_MMAP_CMD_SHIFT = 8, MLX5_IB_MMAP_CMD_MASK = 0xff, }; enum mlx5_ib_mmap_cmd { MLX5_IB_MMAP_REGULAR_PAGE = 0, MLX5_IB_MMAP_GET_CONTIGUOUS_PAGES = 1, MLX5_IB_MMAP_WC_PAGE = 2, MLX5_IB_MMAP_NC_PAGE = 3, /* 5 is chosen in order to be compatible with old versions of libmlx5 */ MLX5_IB_MMAP_CORE_CLOCK = 5, }; enum { MLX5_RES_SCAT_DATA32_CQE = 0x1, MLX5_RES_SCAT_DATA64_CQE = 0x2, MLX5_REQ_SCAT_DATA32_CQE = 0x11, MLX5_REQ_SCAT_DATA64_CQE = 0x22, }; enum mlx5_ib_latency_class { MLX5_IB_LATENCY_CLASS_LOW, MLX5_IB_LATENCY_CLASS_MEDIUM, MLX5_IB_LATENCY_CLASS_HIGH, MLX5_IB_LATENCY_CLASS_FAST_PATH }; enum mlx5_ib_mad_ifc_flags { MLX5_MAD_IFC_IGNORE_MKEY = 1, MLX5_MAD_IFC_IGNORE_BKEY = 2, MLX5_MAD_IFC_NET_VIEW = 4, }; enum { MLX5_CROSS_CHANNEL_UUAR = 0, }; enum { MLX5_CQE_VERSION_V0, MLX5_CQE_VERSION_V1, }; struct mlx5_ib_vma_private_data { struct list_head list; struct vm_area_struct *vma; }; struct mlx5_ib_ucontext { struct ib_ucontext ibucontext; struct list_head db_page_list; /* protect doorbell record alloc/free */ struct mutex db_page_mutex; struct mlx5_uuar_info uuari; u8 cqe_version; /* Transport Domain number */ u32 tdn; struct list_head vma_private_list; }; static inline struct mlx5_ib_ucontext *to_mucontext(struct ib_ucontext *ibucontext) { return container_of(ibucontext, struct mlx5_ib_ucontext, ibucontext); } struct mlx5_ib_pd { struct ib_pd ibpd; u32 pdn; }; #define MLX5_IB_FLOW_MCAST_PRIO (MLX5_BY_PASS_NUM_PRIOS - 1) #define MLX5_IB_FLOW_LAST_PRIO (MLX5_BY_PASS_NUM_REGULAR_PRIOS - 1) #if (MLX5_IB_FLOW_LAST_PRIO <= 0) #error "Invalid number of bypass priorities" #endif #define MLX5_IB_FLOW_LEFTOVERS_PRIO (MLX5_IB_FLOW_MCAST_PRIO + 1) #define MLX5_IB_NUM_FLOW_FT (MLX5_IB_FLOW_LEFTOVERS_PRIO + 1) #define MLX5_IB_NUM_SNIFFER_FTS 2 struct mlx5_ib_flow_prio { struct mlx5_flow_table *flow_table; unsigned int refcount; }; struct mlx5_ib_flow_handler { struct list_head list; struct ib_flow ibflow; struct mlx5_ib_flow_prio *prio; struct mlx5_flow_rule *rule; }; struct mlx5_ib_flow_db { struct mlx5_ib_flow_prio prios[MLX5_IB_NUM_FLOW_FT]; struct mlx5_ib_flow_prio sniffer[MLX5_IB_NUM_SNIFFER_FTS]; struct mlx5_flow_table *lag_demux_ft; /* Protect flow steering bypass flow tables * when add/del flow rules. * only single add/removal of flow steering rule could be done * simultaneously. */ struct mutex lock; }; /* Use macros here so that don't have to duplicate * enum ib_send_flags and enum ib_qp_type for low-level driver */ #define MLX5_IB_SEND_UMR_UNREG IB_SEND_RESERVED_START #define MLX5_IB_SEND_UMR_FAIL_IF_FREE (IB_SEND_RESERVED_START << 1) #define MLX5_IB_SEND_UMR_UPDATE_MTT (IB_SEND_RESERVED_START << 2) #define MLX5_IB_SEND_UMR_UPDATE_TRANSLATION (IB_SEND_RESERVED_START << 3) #define MLX5_IB_SEND_UMR_UPDATE_PD (IB_SEND_RESERVED_START << 4) #define MLX5_IB_SEND_UMR_UPDATE_ACCESS IB_SEND_RESERVED_END #define MLX5_IB_QPT_REG_UMR IB_QPT_RESERVED1 /* * IB_QPT_GSI creates the software wrapper around GSI, and MLX5_IB_QPT_HW_GSI * creates the actual hardware QP. */ #define MLX5_IB_QPT_HW_GSI IB_QPT_RESERVED2 #define MLX5_IB_WR_UMR IB_WR_RESERVED1 /* Private QP creation flags to be passed in ib_qp_init_attr.create_flags. * * These flags are intended for internal use by the mlx5_ib driver, and they * rely on the range reserved for that use in the ib_qp_create_flags enum. */ /* Create a UD QP whose source QP number is 1 */ static inline enum ib_qp_create_flags mlx5_ib_create_qp_sqpn_qp1(void) { return IB_QP_CREATE_RESERVED_START; } struct wr_list { u16 opcode; u16 next; }; struct mlx5_ib_wq { u64 *wrid; u32 *wr_data; struct wr_list *w_list; unsigned *wqe_head; u16 unsig_count; /* serialize post to the work queue */ spinlock_t lock; int wqe_cnt; int max_post; int max_gs; int offset; int wqe_shift; unsigned head; unsigned tail; u16 cur_post; u16 last_poll; void *qend; }; struct mlx5_ib_rwq { struct ib_wq ibwq; struct mlx5_core_qp core_qp; u32 rq_num_pas; u32 log_rq_stride; u32 log_rq_size; u32 rq_page_offset; u32 log_page_size; struct ib_umem *umem; size_t buf_size; unsigned int page_shift; int create_type; struct mlx5_db db; u32 user_index; u32 wqe_count; u32 wqe_shift; int wq_sig; }; enum { MLX5_QP_USER, MLX5_QP_KERNEL, MLX5_QP_EMPTY }; enum { MLX5_WQ_USER, MLX5_WQ_KERNEL }; struct mlx5_ib_rwq_ind_table { struct ib_rwq_ind_table ib_rwq_ind_tbl; u32 rqtn; }; /* * Connect-IB can trigger up to four concurrent pagefaults * per-QP. */ enum mlx5_ib_pagefault_context { MLX5_IB_PAGEFAULT_RESPONDER_READ, MLX5_IB_PAGEFAULT_REQUESTOR_READ, MLX5_IB_PAGEFAULT_RESPONDER_WRITE, MLX5_IB_PAGEFAULT_REQUESTOR_WRITE, MLX5_IB_PAGEFAULT_CONTEXTS }; static inline enum mlx5_ib_pagefault_context mlx5_ib_get_pagefault_context(struct mlx5_pagefault *pagefault) { return pagefault->flags & (MLX5_PFAULT_REQUESTOR | MLX5_PFAULT_WRITE); } struct mlx5_ib_pfault { struct work_struct work; struct mlx5_pagefault mpfault; }; struct mlx5_ib_ubuffer { struct ib_umem *umem; int buf_size; u64 buf_addr; }; struct mlx5_ib_qp_base { struct mlx5_ib_qp *container_mibqp; struct mlx5_core_qp mqp; struct mlx5_ib_ubuffer ubuffer; }; struct mlx5_ib_qp_trans { struct mlx5_ib_qp_base base; u16 xrcdn; u8 alt_port; u8 atomic_rd_en; u8 resp_depth; }; struct mlx5_ib_rss_qp { u32 tirn; }; struct mlx5_ib_rq { struct mlx5_ib_qp_base base; struct mlx5_ib_wq *rq; struct mlx5_ib_ubuffer ubuffer; struct mlx5_db *doorbell; u32 tirn; u8 state; }; struct mlx5_ib_sq { struct mlx5_ib_qp_base base; struct mlx5_ib_wq *sq; struct mlx5_ib_ubuffer ubuffer; struct mlx5_db *doorbell; u32 tisn; u8 state; }; struct mlx5_ib_raw_packet_qp { struct mlx5_ib_sq sq; struct mlx5_ib_rq rq; }; struct mlx5_ib_qp { struct ib_qp ibqp; union { struct mlx5_ib_qp_trans trans_qp; struct mlx5_ib_raw_packet_qp raw_packet_qp; struct mlx5_ib_rss_qp rss_qp; }; struct mlx5_buf buf; struct mlx5_db db; struct mlx5_ib_wq rq; u8 sq_signal_bits; u8 fm_cache; struct mlx5_ib_wq sq; /* serialize qp state modifications */ struct mutex mutex; u32 flags; u8 port; u8 state; int wq_sig; int scat_cqe; int max_inline_data; struct mlx5_bf *bf; int has_rq; /* only for user space QPs. For kernel * we have it from the bf object */ int uuarn; int create_type; /* Store signature errors */ bool signature_en; #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING /* * A flag that is true for QP's that are in a state that doesn't * allow page faults, and shouldn't schedule any more faults. */ int disable_page_faults; /* * The disable_page_faults_lock protects a QP's disable_page_faults * field, allowing for a thread to atomically check whether the QP * allows page faults, and if so schedule a page fault. */ spinlock_t disable_page_faults_lock; struct mlx5_ib_pfault pagefaults[MLX5_IB_PAGEFAULT_CONTEXTS]; #endif struct list_head qps_list; struct list_head cq_recv_list; struct list_head cq_send_list; }; struct mlx5_ib_cq_buf { struct mlx5_buf buf; struct ib_umem *umem; int cqe_size; int nent; }; enum mlx5_ib_qp_flags { MLX5_IB_QP_LSO = IB_QP_CREATE_IPOIB_UD_LSO, MLX5_IB_QP_BLOCK_MULTICAST_LOOPBACK = IB_QP_CREATE_BLOCK_MULTICAST_LOOPBACK, MLX5_IB_QP_CROSS_CHANNEL = IB_QP_CREATE_CROSS_CHANNEL, MLX5_IB_QP_MANAGED_SEND = IB_QP_CREATE_MANAGED_SEND, MLX5_IB_QP_MANAGED_RECV = IB_QP_CREATE_MANAGED_RECV, MLX5_IB_QP_SIGNATURE_HANDLING = 1 << 5, /* QP uses 1 as its source QP number */ MLX5_IB_QP_SQPN_QP1 = 1 << 6, MLX5_IB_QP_CAP_SCATTER_FCS = 1 << 7, MLX5_IB_QP_RSS = 1 << 8, }; struct mlx5_umr_wr { struct ib_send_wr wr; union { u64 virt_addr; u64 offset; } target; struct ib_pd *pd; unsigned int page_shift; unsigned int npages; u32 length; int access_flags; u32 mkey; }; static inline struct mlx5_umr_wr *umr_wr(struct ib_send_wr *wr) { return container_of(wr, struct mlx5_umr_wr, wr); } struct mlx5_shared_mr_info { int mr_id; struct ib_umem *umem; }; struct mlx5_ib_cq { struct ib_cq ibcq; struct mlx5_core_cq mcq; struct mlx5_ib_cq_buf buf; struct mlx5_db db; /* serialize access to the CQ */ spinlock_t lock; /* protect resize cq */ struct mutex resize_mutex; struct mlx5_ib_cq_buf *resize_buf; struct ib_umem *resize_umem; int cqe_size; struct list_head list_send_qp; struct list_head list_recv_qp; u32 create_flags; struct list_head wc_list; enum ib_cq_notify_flags notify_flags; struct work_struct notify_work; }; struct mlx5_ib_wc { struct ib_wc wc; struct list_head list; }; struct mlx5_ib_srq { struct ib_srq ibsrq; struct mlx5_core_srq msrq; struct mlx5_buf buf; struct mlx5_db db; u64 *wrid; /* protect SRQ hanlding */ spinlock_t lock; int head; int tail; u16 wqe_ctr; struct ib_umem *umem; /* serialize arming a SRQ */ struct mutex mutex; int wq_sig; }; struct mlx5_ib_xrcd { struct ib_xrcd ibxrcd; u32 xrcdn; }; enum mlx5_ib_mtt_access_flags { MLX5_IB_MTT_READ = (1 << 0), MLX5_IB_MTT_WRITE = (1 << 1), }; #define MLX5_IB_MTT_PRESENT (MLX5_IB_MTT_READ | MLX5_IB_MTT_WRITE) struct mlx5_ib_mr { struct ib_mr ibmr; void *descs; dma_addr_t desc_map; int ndescs; int max_descs; int desc_size; int access_mode; struct mlx5_core_mr mmkey; struct ib_umem *umem; struct mlx5_shared_mr_info *smr_info; struct list_head list; int order; int umred; int npages; struct mlx5_ib_dev *dev; u32 out[MLX5_ST_SZ_DW(create_mkey_out)]; struct mlx5_core_sig_ctx *sig; int live; void *descs_alloc; int access_flags; /* Needed for rereg MR */ }; struct mlx5_ib_mw { struct ib_mw ibmw; struct mlx5_core_mr mmkey; }; struct mlx5_ib_umr_context { struct ib_cqe cqe; enum ib_wc_status status; struct completion done; }; struct umr_common { struct ib_pd *pd; struct ib_cq *cq; struct ib_qp *qp; /* control access to UMR QP */ struct semaphore sem; }; enum { MLX5_FMR_INVALID, MLX5_FMR_VALID, MLX5_FMR_BUSY, }; struct mlx5_cache_ent { struct list_head head; /* sync access to the cahce entry */ spinlock_t lock; struct dentry *dir; char name[4]; u32 order; u32 size; u32 cur; u32 miss; u32 limit; struct dentry *fsize; struct dentry *fcur; struct dentry *fmiss; struct dentry *flimit; struct mlx5_ib_dev *dev; struct work_struct work; struct delayed_work dwork; int pending; }; struct mlx5_mr_cache { struct workqueue_struct *wq; struct mlx5_cache_ent ent[MAX_MR_CACHE_ENTRIES]; int stopped; struct dentry *root; unsigned long last_add; }; struct mlx5_ib_gsi_qp; struct mlx5_ib_port_resources { struct mlx5_ib_resources *devr; struct mlx5_ib_gsi_qp *gsi; struct work_struct pkey_change_work; }; struct mlx5_ib_resources { struct ib_cq *c0; struct ib_xrcd *x0; struct ib_xrcd *x1; struct ib_pd *p0; struct ib_srq *s0; struct ib_srq *s1; struct mlx5_ib_port_resources ports[2]; /* Protects changes to the port resources */ struct mutex mutex; }; struct mlx5_ib_port { u16 q_cnt_id; }; struct mlx5_roce { /* Protect mlx5_ib_get_netdev from invoking dev_hold() with a NULL * netdev pointer */ rwlock_t netdev_lock; struct net_device *netdev; struct notifier_block nb; atomic_t next_port; }; struct mlx5_ib_dev { struct ib_device ib_dev; struct mlx5_core_dev *mdev; struct mlx5_roce roce; MLX5_DECLARE_DOORBELL_LOCK(uar_lock); int num_ports; /* serialize update of capability mask */ struct mutex cap_mask_mutex; bool ib_active; struct umr_common umrc; /* sync used page count stats */ struct mlx5_ib_resources devr; struct mlx5_mr_cache cache; struct timer_list delay_timer; /* Prevents soft lock on massive reg MRs */ struct mutex slow_path_mutex; int fill_delay; #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING struct ib_odp_caps odp_caps; /* * Sleepable RCU that prevents destruction of MRs while they are still * being used by a page fault handler. */ struct srcu_struct mr_srcu; #endif struct mlx5_ib_flow_db flow_db; /* protect resources needed as part of reset flow */ spinlock_t reset_flow_resource_lock; struct list_head qp_list; /* Array with num_ports elements */ struct mlx5_ib_port *port; }; static inline struct mlx5_ib_cq *to_mibcq(struct mlx5_core_cq *mcq) { return container_of(mcq, struct mlx5_ib_cq, mcq); } static inline struct mlx5_ib_xrcd *to_mxrcd(struct ib_xrcd *ibxrcd) { return container_of(ibxrcd, struct mlx5_ib_xrcd, ibxrcd); } static inline struct mlx5_ib_dev *to_mdev(struct ib_device *ibdev) { return container_of(ibdev, struct mlx5_ib_dev, ib_dev); } static inline struct mlx5_ib_cq *to_mcq(struct ib_cq *ibcq) { return container_of(ibcq, struct mlx5_ib_cq, ibcq); } static inline struct mlx5_ib_qp *to_mibqp(struct mlx5_core_qp *mqp) { return container_of(mqp, struct mlx5_ib_qp_base, mqp)->container_mibqp; } static inline struct mlx5_ib_rwq *to_mibrwq(struct mlx5_core_qp *core_qp) { return container_of(core_qp, struct mlx5_ib_rwq, core_qp); } static inline struct mlx5_ib_mr *to_mibmr(struct mlx5_core_mr *mmkey) { return container_of(mmkey, struct mlx5_ib_mr, mmkey); } static inline struct mlx5_ib_pd *to_mpd(struct ib_pd *ibpd) { return container_of(ibpd, struct mlx5_ib_pd, ibpd); } static inline struct mlx5_ib_srq *to_msrq(struct ib_srq *ibsrq) { return container_of(ibsrq, struct mlx5_ib_srq, ibsrq); } static inline struct mlx5_ib_qp *to_mqp(struct ib_qp *ibqp) { return container_of(ibqp, struct mlx5_ib_qp, ibqp); } static inline struct mlx5_ib_rwq *to_mrwq(struct ib_wq *ibwq) { return container_of(ibwq, struct mlx5_ib_rwq, ibwq); } static inline struct mlx5_ib_rwq_ind_table *to_mrwq_ind_table(struct ib_rwq_ind_table *ib_rwq_ind_tbl) { return container_of(ib_rwq_ind_tbl, struct mlx5_ib_rwq_ind_table, ib_rwq_ind_tbl); } static inline struct mlx5_ib_srq *to_mibsrq(struct mlx5_core_srq *msrq) { return container_of(msrq, struct mlx5_ib_srq, msrq); } static inline struct mlx5_ib_mr *to_mmr(struct ib_mr *ibmr) { return container_of(ibmr, struct mlx5_ib_mr, ibmr); } static inline struct mlx5_ib_mw *to_mmw(struct ib_mw *ibmw) { return container_of(ibmw, struct mlx5_ib_mw, ibmw); } struct mlx5_ib_ah { struct ib_ah ibah; struct mlx5_av av; }; static inline struct mlx5_ib_ah *to_mah(struct ib_ah *ibah) { return container_of(ibah, struct mlx5_ib_ah, ibah); } int mlx5_ib_db_map_user(struct mlx5_ib_ucontext *context, unsigned long virt, struct mlx5_db *db); void mlx5_ib_db_unmap_user(struct mlx5_ib_ucontext *context, struct mlx5_db *db); void __mlx5_ib_cq_clean(struct mlx5_ib_cq *cq, u32 qpn, struct mlx5_ib_srq *srq); void mlx5_ib_cq_clean(struct mlx5_ib_cq *cq, u32 qpn, struct mlx5_ib_srq *srq); void mlx5_ib_free_srq_wqe(struct mlx5_ib_srq *srq, int wqe_index); int mlx5_MAD_IFC(struct mlx5_ib_dev *dev, int ignore_mkey, int ignore_bkey, u8 port, const struct ib_wc *in_wc, const struct ib_grh *in_grh, const void *in_mad, void *response_mad); struct ib_ah *mlx5_ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr); int mlx5_ib_query_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr); int mlx5_ib_destroy_ah(struct ib_ah *ah); struct ib_srq *mlx5_ib_create_srq(struct ib_pd *pd, struct ib_srq_init_attr *init_attr, struct ib_udata *udata); int mlx5_ib_modify_srq(struct ib_srq *ibsrq, struct ib_srq_attr *attr, enum ib_srq_attr_mask attr_mask, struct ib_udata *udata); int mlx5_ib_query_srq(struct ib_srq *ibsrq, struct ib_srq_attr *srq_attr); int mlx5_ib_destroy_srq(struct ib_srq *srq); int mlx5_ib_post_srq_recv(struct ib_srq *ibsrq, struct ib_recv_wr *wr, struct ib_recv_wr **bad_wr); struct ib_qp *mlx5_ib_create_qp(struct ib_pd *pd, struct ib_qp_init_attr *init_attr, struct ib_udata *udata); int mlx5_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, int attr_mask, struct ib_udata *udata); int mlx5_ib_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *qp_attr, int qp_attr_mask, struct ib_qp_init_attr *qp_init_attr); int mlx5_ib_destroy_qp(struct ib_qp *qp); int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, struct ib_send_wr **bad_wr); int mlx5_ib_post_recv(struct ib_qp *ibqp, struct ib_recv_wr *wr, struct ib_recv_wr **bad_wr); void *mlx5_get_send_wqe(struct mlx5_ib_qp *qp, int n); int mlx5_ib_read_user_wqe(struct mlx5_ib_qp *qp, int send, int wqe_index, void *buffer, u32 length, struct mlx5_ib_qp_base *base); struct ib_cq *mlx5_ib_create_cq(struct ib_device *ibdev, const struct ib_cq_init_attr *attr, struct ib_ucontext *context, struct ib_udata *udata); int mlx5_ib_destroy_cq(struct ib_cq *cq); int mlx5_ib_poll_cq(struct ib_cq *ibcq, int num_entries, struct ib_wc *wc); int mlx5_ib_arm_cq(struct ib_cq *ibcq, enum ib_cq_notify_flags flags); int mlx5_ib_modify_cq(struct ib_cq *cq, u16 cq_count, u16 cq_period); int mlx5_ib_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata); struct ib_mr *mlx5_ib_get_dma_mr(struct ib_pd *pd, int acc); struct ib_mr *mlx5_ib_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, u64 virt_addr, int access_flags, struct ib_udata *udata); struct ib_mw *mlx5_ib_alloc_mw(struct ib_pd *pd, enum ib_mw_type type, struct ib_udata *udata); int mlx5_ib_dealloc_mw(struct ib_mw *mw); int mlx5_ib_update_mtt(struct mlx5_ib_mr *mr, u64 start_page_index, int npages, int zap); int mlx5_ib_rereg_user_mr(struct ib_mr *ib_mr, int flags, u64 start, u64 length, u64 virt_addr, int access_flags, struct ib_pd *pd, struct ib_udata *udata); +struct ib_mr *mlx5_ib_reg_phys_mr(struct ib_pd *pd, + struct ib_phys_buf *buffer_list, + int num_phys_buf, + int access_flags, + u64 *virt_addr); int mlx5_ib_dereg_mr(struct ib_mr *ibmr); struct ib_mr *mlx5_ib_alloc_mr(struct ib_pd *pd, enum ib_mr_type mr_type, u32 max_num_sg); int mlx5_ib_map_mr_sg(struct ib_mr *ibmr, struct scatterlist *sg, int sg_nents, unsigned int *sg_offset); int mlx5_ib_process_mad(struct ib_device *ibdev, int mad_flags, u8 port_num, const struct ib_wc *in_wc, const struct ib_grh *in_grh, const struct ib_mad_hdr *in, size_t in_mad_size, struct ib_mad_hdr *out, size_t *out_mad_size, u16 *out_mad_pkey_index); struct ib_xrcd *mlx5_ib_alloc_xrcd(struct ib_device *ibdev, struct ib_ucontext *context, struct ib_udata *udata); int mlx5_ib_dealloc_xrcd(struct ib_xrcd *xrcd); int mlx5_ib_get_buf_offset(u64 addr, int page_shift, u32 *offset); int mlx5_query_ext_port_caps(struct mlx5_ib_dev *dev, u8 port); int mlx5_query_mad_ifc_smp_attr_node_info(struct ib_device *ibdev, struct ib_smp *out_mad); int mlx5_query_mad_ifc_system_image_guid(struct ib_device *ibdev, __be64 *sys_image_guid); int mlx5_query_mad_ifc_max_pkeys(struct ib_device *ibdev, u16 *max_pkeys); int mlx5_query_mad_ifc_vendor_id(struct ib_device *ibdev, u32 *vendor_id); int mlx5_query_mad_ifc_node_desc(struct mlx5_ib_dev *dev, char *node_desc); int mlx5_query_mad_ifc_node_guid(struct mlx5_ib_dev *dev, __be64 *node_guid); int mlx5_query_mad_ifc_pkey(struct ib_device *ibdev, u8 port, u16 index, u16 *pkey); int mlx5_query_mad_ifc_gids(struct ib_device *ibdev, u8 port, int index, union ib_gid *gid); int mlx5_query_mad_ifc_port(struct ib_device *ibdev, u8 port, struct ib_port_attr *props); int mlx5_ib_query_port(struct ib_device *ibdev, u8 port, struct ib_port_attr *props); int mlx5_ib_init_fmr(struct mlx5_ib_dev *dev); void mlx5_ib_cleanup_fmr(struct mlx5_ib_dev *dev); void mlx5_ib_cont_pages(struct ib_umem *umem, u64 addr, int *count, int *shift, int *ncont, int *order); void __mlx5_ib_populate_pas(struct mlx5_ib_dev *dev, struct ib_umem *umem, int page_shift, size_t offset, size_t num_pages, __be64 *pas, int access_flags); void mlx5_ib_populate_pas(struct mlx5_ib_dev *dev, struct ib_umem *umem, int page_shift, __be64 *pas, int access_flags); void mlx5_ib_copy_pas(u64 *old, u64 *new, int step, int num); int mlx5_ib_get_cqe_size(struct mlx5_ib_dev *dev, struct ib_cq *ibcq); int mlx5_mr_cache_init(struct mlx5_ib_dev *dev); int mlx5_mr_cache_cleanup(struct mlx5_ib_dev *dev); int mlx5_mr_ib_cont_pages(struct ib_umem *umem, u64 addr, int *count, int *shift); int mlx5_ib_check_mr_status(struct ib_mr *ibmr, u32 check_mask, struct ib_mr_status *mr_status); struct ib_wq *mlx5_ib_create_wq(struct ib_pd *pd, struct ib_wq_init_attr *init_attr, struct ib_udata *udata); int mlx5_ib_destroy_wq(struct ib_wq *wq); int mlx5_ib_modify_wq(struct ib_wq *wq, struct ib_wq_attr *wq_attr, u32 wq_attr_mask, struct ib_udata *udata); struct ib_rwq_ind_table *mlx5_ib_create_rwq_ind_table(struct ib_device *device, struct ib_rwq_ind_table_init_attr *init_attr, struct ib_udata *udata); int mlx5_ib_destroy_rwq_ind_table(struct ib_rwq_ind_table *wq_ind_table); #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING extern struct workqueue_struct *mlx5_ib_page_fault_wq; void mlx5_ib_internal_fill_odp_caps(struct mlx5_ib_dev *dev); void mlx5_ib_mr_pfault_handler(struct mlx5_ib_qp *qp, struct mlx5_ib_pfault *pfault); void mlx5_ib_odp_create_qp(struct mlx5_ib_qp *qp); int mlx5_ib_odp_init_one(struct mlx5_ib_dev *ibdev); void mlx5_ib_odp_remove_one(struct mlx5_ib_dev *ibdev); int __init mlx5_ib_odp_init(void); void mlx5_ib_odp_cleanup(void); void mlx5_ib_qp_disable_pagefaults(struct mlx5_ib_qp *qp); void mlx5_ib_qp_enable_pagefaults(struct mlx5_ib_qp *qp); void mlx5_ib_invalidate_range(struct ib_umem *umem, unsigned long start, unsigned long end); #else /* CONFIG_INFINIBAND_ON_DEMAND_PAGING */ static inline void mlx5_ib_internal_fill_odp_caps(struct mlx5_ib_dev *dev) { return; } static inline void mlx5_ib_odp_create_qp(struct mlx5_ib_qp *qp) {} static inline int mlx5_ib_odp_init_one(struct mlx5_ib_dev *ibdev) { return 0; } static inline void mlx5_ib_odp_remove_one(struct mlx5_ib_dev *ibdev) {} static inline int mlx5_ib_odp_init(void) { return 0; } static inline void mlx5_ib_odp_cleanup(void) {} static inline void mlx5_ib_qp_disable_pagefaults(struct mlx5_ib_qp *qp) {} static inline void mlx5_ib_qp_enable_pagefaults(struct mlx5_ib_qp *qp) {} #endif /* CONFIG_INFINIBAND_ON_DEMAND_PAGING */ int mlx5_ib_get_vf_config(struct ib_device *device, int vf, u8 port, struct ifla_vf_info *info); int mlx5_ib_set_vf_link_state(struct ib_device *device, int vf, u8 port, int state); int mlx5_ib_get_vf_stats(struct ib_device *device, int vf, u8 port, struct ifla_vf_stats *stats); int mlx5_ib_set_vf_guid(struct ib_device *device, int vf, u8 port, u64 guid, int type); __be16 mlx5_get_roce_udp_sport(struct mlx5_ib_dev *dev, u8 port_num, int index); /* GSI QP helper functions */ struct ib_qp *mlx5_ib_gsi_create_qp(struct ib_pd *pd, struct ib_qp_init_attr *init_attr); int mlx5_ib_gsi_destroy_qp(struct ib_qp *qp); int mlx5_ib_gsi_modify_qp(struct ib_qp *qp, struct ib_qp_attr *attr, int attr_mask); int mlx5_ib_gsi_query_qp(struct ib_qp *qp, struct ib_qp_attr *qp_attr, int qp_attr_mask, struct ib_qp_init_attr *qp_init_attr); int mlx5_ib_gsi_post_send(struct ib_qp *qp, struct ib_send_wr *wr, struct ib_send_wr **bad_wr); int mlx5_ib_gsi_post_recv(struct ib_qp *qp, struct ib_recv_wr *wr, struct ib_recv_wr **bad_wr); void mlx5_ib_gsi_pkey_change(struct mlx5_ib_gsi_qp *gsi); int mlx5_ib_generate_wc(struct ib_cq *ibcq, struct ib_wc *wc); static inline void init_query_mad(struct ib_smp *mad) { mad->base_version = 1; mad->mgmt_class = IB_MGMT_CLASS_SUBN_LID_ROUTED; mad->class_version = 1; mad->method = IB_MGMT_METHOD_GET; } static inline u8 convert_access(int acc) { return (acc & IB_ACCESS_REMOTE_ATOMIC ? MLX5_PERM_ATOMIC : 0) | (acc & IB_ACCESS_REMOTE_WRITE ? MLX5_PERM_REMOTE_WRITE : 0) | (acc & IB_ACCESS_REMOTE_READ ? MLX5_PERM_REMOTE_READ : 0) | (acc & IB_ACCESS_LOCAL_WRITE ? MLX5_PERM_LOCAL_WRITE : 0) | MLX5_PERM_LOCAL_READ; } static inline int is_qp1(enum ib_qp_type qp_type) { return qp_type == MLX5_IB_QPT_HW_GSI; } #define MLX5_MAX_UMR_SHIFT 16 #define MLX5_MAX_UMR_PAGES (1 << MLX5_MAX_UMR_SHIFT) static inline u32 check_cq_create_flags(u32 flags) { /* * It returns non-zero value for unsupported CQ * create flags, otherwise it returns zero. */ return (flags & ~(IB_CQ_FLAGS_IGNORE_OVERRUN | IB_CQ_FLAGS_TIMESTAMP_COMPLETION)); } static inline int verify_assign_uidx(u8 cqe_version, u32 cmd_uidx, u32 *user_index) { if (cqe_version) { if ((cmd_uidx == MLX5_IB_DEFAULT_UIDX) || (cmd_uidx & ~MLX5_USER_ASSIGNED_UIDX_MASK)) return -EINVAL; *user_index = cmd_uidx; } else { *user_index = MLX5_IB_DEFAULT_UIDX; } return 0; } static inline int get_qp_user_index(struct mlx5_ib_ucontext *ucontext, struct mlx5_ib_create_qp *ucmd, int inlen, u32 *user_index) { u8 cqe_version = ucontext->cqe_version; if (field_avail(struct mlx5_ib_create_qp, uidx, inlen) && !cqe_version && (ucmd->uidx == MLX5_IB_DEFAULT_UIDX)) return 0; if (!!(field_avail(struct mlx5_ib_create_qp, uidx, inlen) != !!cqe_version)) return -EINVAL; return verify_assign_uidx(cqe_version, ucmd->uidx, user_index); } static inline int get_srq_user_index(struct mlx5_ib_ucontext *ucontext, struct mlx5_ib_create_srq *ucmd, int inlen, u32 *user_index) { u8 cqe_version = ucontext->cqe_version; if (field_avail(struct mlx5_ib_create_srq, uidx, inlen) && !cqe_version && (ucmd->uidx == MLX5_IB_DEFAULT_UIDX)) return 0; if (!!(field_avail(struct mlx5_ib_create_srq, uidx, inlen) != !!cqe_version)) return -EINVAL; return verify_assign_uidx(cqe_version, ucmd->uidx, user_index); } #endif /* MLX5_IB_H */ Index: projects/bsd_rdma_4_9_stable_11/sys/dev/mlx5/mlx5_ib/mlx5_ib_main.c =================================================================== --- projects/bsd_rdma_4_9_stable_11/sys/dev/mlx5/mlx5_ib/mlx5_ib_main.c (revision 331011) +++ projects/bsd_rdma_4_9_stable_11/sys/dev/mlx5/mlx5_ib/mlx5_ib_main.c (revision 331012) @@ -1,3229 +1,3230 @@ /*- * Copyright (c) 2013-2015, Mellanox Technologies, Ltd. All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * * THIS SOFTWARE IS PROVIDED BY AUTHOR AND CONTRIBUTORS `AS IS' AND * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL AUTHOR OR CONTRIBUTORS BE LIABLE * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF * SUCH DAMAGE. * * $FreeBSD$ */ #include #include #include #include #include #if defined(CONFIG_X86) #include #endif #include #include #include #undef inode #include #include #include #include #include #include #include #include #include #include #include "mlx5_ib.h" #define DRIVER_NAME "mlx5_ib" #define DRIVER_VERSION "3.4.1-BETA" #define DRIVER_RELDATE "October 2017" MODULE_DESCRIPTION("Mellanox Connect-IB HCA IB driver"); MODULE_LICENSE("Dual BSD/GPL"); MODULE_DEPEND(mlx5ib, linuxkpi, 1, 1, 1); MODULE_DEPEND(mlx5ib, mlx5, 1, 1, 1); MODULE_DEPEND(mlx5ib, ibcore, 1, 1, 1); MODULE_VERSION(mlx5ib, 1); static int deprecated_prof_sel = 2; module_param_named(prof_sel, deprecated_prof_sel, int, 0444); MODULE_PARM_DESC(prof_sel, "profile selector. Deprecated here. Moved to module mlx5_core"); static char mlx5_version[] = DRIVER_NAME ": Mellanox Connect-IB Infiniband driver v" DRIVER_VERSION " (" DRIVER_RELDATE ")\n"; enum { MLX5_ATOMIC_SIZE_QP_8BYTES = 1 << 3, }; static enum rdma_link_layer mlx5_port_type_cap_to_rdma_ll(int port_type_cap) { switch (port_type_cap) { case MLX5_CAP_PORT_TYPE_IB: return IB_LINK_LAYER_INFINIBAND; case MLX5_CAP_PORT_TYPE_ETH: return IB_LINK_LAYER_ETHERNET; default: return IB_LINK_LAYER_UNSPECIFIED; } } static enum rdma_link_layer mlx5_ib_port_link_layer(struct ib_device *device, u8 port_num) { struct mlx5_ib_dev *dev = to_mdev(device); int port_type_cap = MLX5_CAP_GEN(dev->mdev, port_type); return mlx5_port_type_cap_to_rdma_ll(port_type_cap); } static bool mlx5_netdev_match(struct net_device *ndev, struct mlx5_core_dev *mdev, const char *dname) { return ndev->if_type == IFT_ETHER && ndev->if_dname != NULL && strcmp(ndev->if_dname, dname) == 0 && ndev->if_softc != NULL && *(struct mlx5_core_dev **)ndev->if_softc == mdev; } static int mlx5_netdev_event(struct notifier_block *this, unsigned long event, void *ptr) { struct net_device *ndev = netdev_notifier_info_to_dev(ptr); struct mlx5_ib_dev *ibdev = container_of(this, struct mlx5_ib_dev, roce.nb); switch (event) { case NETDEV_REGISTER: case NETDEV_UNREGISTER: write_lock(&ibdev->roce.netdev_lock); /* check if network interface belongs to mlx5en */ if (mlx5_netdev_match(ndev, ibdev->mdev, "mce")) ibdev->roce.netdev = (event == NETDEV_UNREGISTER) ? NULL : ndev; write_unlock(&ibdev->roce.netdev_lock); break; case NETDEV_UP: case NETDEV_DOWN: { struct net_device *upper = NULL; if ((upper == ndev || (!upper && ndev == ibdev->roce.netdev)) && ibdev->ib_active) { struct ib_event ibev = {0}; ibev.device = &ibdev->ib_dev; ibev.event = (event == NETDEV_UP) ? IB_EVENT_PORT_ACTIVE : IB_EVENT_PORT_ERR; ibev.element.port_num = 1; ib_dispatch_event(&ibev); } break; } default: break; } return NOTIFY_DONE; } static struct net_device *mlx5_ib_get_netdev(struct ib_device *device, u8 port_num) { struct mlx5_ib_dev *ibdev = to_mdev(device); struct net_device *ndev; /* Ensure ndev does not disappear before we invoke dev_hold() */ read_lock(&ibdev->roce.netdev_lock); ndev = ibdev->roce.netdev; if (ndev) dev_hold(ndev); read_unlock(&ibdev->roce.netdev_lock); return ndev; } static int mlx5_query_port_roce(struct ib_device *device, u8 port_num, struct ib_port_attr *props) { struct mlx5_ib_dev *dev = to_mdev(device); struct net_device *ndev; enum ib_mtu ndev_ib_mtu; u16 qkey_viol_cntr; memset(props, 0, sizeof(*props)); props->port_cap_flags |= IB_PORT_CM_SUP; props->port_cap_flags |= IB_PORT_IP_BASED_GIDS; props->gid_tbl_len = MLX5_CAP_ROCE(dev->mdev, roce_address_table_size); props->max_mtu = IB_MTU_4096; props->max_msg_sz = 1 << MLX5_CAP_GEN(dev->mdev, log_max_msg); props->pkey_tbl_len = 1; props->state = IB_PORT_DOWN; props->phys_state = 3; mlx5_query_nic_vport_qkey_viol_cntr(dev->mdev, &qkey_viol_cntr); props->qkey_viol_cntr = qkey_viol_cntr; ndev = mlx5_ib_get_netdev(device, port_num); if (!ndev) return 0; if (netif_running(ndev) && netif_carrier_ok(ndev)) { props->state = IB_PORT_ACTIVE; props->phys_state = 5; } ndev_ib_mtu = iboe_get_mtu(ndev->if_mtu); dev_put(ndev); props->active_mtu = min(props->max_mtu, ndev_ib_mtu); props->active_width = IB_WIDTH_4X; /* TODO */ props->active_speed = IB_SPEED_QDR; /* TODO */ return 0; } static void ib_gid_to_mlx5_roce_addr(const union ib_gid *gid, const struct ib_gid_attr *attr, void *mlx5_addr) { #define MLX5_SET_RA(p, f, v) MLX5_SET(roce_addr_layout, p, f, v) char *mlx5_addr_l3_addr = MLX5_ADDR_OF(roce_addr_layout, mlx5_addr, source_l3_address); void *mlx5_addr_mac = MLX5_ADDR_OF(roce_addr_layout, mlx5_addr, source_mac_47_32); if (!gid) return; ether_addr_copy(mlx5_addr_mac, IF_LLADDR(attr->ndev)); if (is_vlan_dev(attr->ndev)) { MLX5_SET_RA(mlx5_addr, vlan_valid, 1); MLX5_SET_RA(mlx5_addr, vlan_id, vlan_dev_vlan_id(attr->ndev)); } switch (attr->gid_type) { case IB_GID_TYPE_IB: MLX5_SET_RA(mlx5_addr, roce_version, MLX5_ROCE_VERSION_1); break; case IB_GID_TYPE_ROCE_UDP_ENCAP: MLX5_SET_RA(mlx5_addr, roce_version, MLX5_ROCE_VERSION_2); break; default: WARN_ON(true); } if (attr->gid_type != IB_GID_TYPE_IB) { if (ipv6_addr_v4mapped((void *)gid)) MLX5_SET_RA(mlx5_addr, roce_l3_type, MLX5_ROCE_L3_TYPE_IPV4); else MLX5_SET_RA(mlx5_addr, roce_l3_type, MLX5_ROCE_L3_TYPE_IPV6); } if ((attr->gid_type == IB_GID_TYPE_IB) || !ipv6_addr_v4mapped((void *)gid)) memcpy(mlx5_addr_l3_addr, gid, sizeof(*gid)); else memcpy(&mlx5_addr_l3_addr[12], &gid->raw[12], 4); } static int set_roce_addr(struct ib_device *device, u8 port_num, unsigned int index, const union ib_gid *gid, const struct ib_gid_attr *attr) { struct mlx5_ib_dev *dev = to_mdev(device); u32 in[MLX5_ST_SZ_DW(set_roce_address_in)] = {0}; u32 out[MLX5_ST_SZ_DW(set_roce_address_out)] = {0}; void *in_addr = MLX5_ADDR_OF(set_roce_address_in, in, roce_address); enum rdma_link_layer ll = mlx5_ib_port_link_layer(device, port_num); if (ll != IB_LINK_LAYER_ETHERNET) return -EINVAL; ib_gid_to_mlx5_roce_addr(gid, attr, in_addr); MLX5_SET(set_roce_address_in, in, roce_address_index, index); MLX5_SET(set_roce_address_in, in, opcode, MLX5_CMD_OP_SET_ROCE_ADDRESS); return mlx5_cmd_exec(dev->mdev, in, sizeof(in), out, sizeof(out)); } static int mlx5_ib_add_gid(struct ib_device *device, u8 port_num, unsigned int index, const union ib_gid *gid, const struct ib_gid_attr *attr, __always_unused void **context) { return set_roce_addr(device, port_num, index, gid, attr); } static int mlx5_ib_del_gid(struct ib_device *device, u8 port_num, unsigned int index, __always_unused void **context) { return set_roce_addr(device, port_num, index, NULL, NULL); } __be16 mlx5_get_roce_udp_sport(struct mlx5_ib_dev *dev, u8 port_num, int index) { struct ib_gid_attr attr; union ib_gid gid; if (ib_get_cached_gid(&dev->ib_dev, port_num, index, &gid, &attr)) return 0; if (!attr.ndev) return 0; dev_put(attr.ndev); if (attr.gid_type != IB_GID_TYPE_ROCE_UDP_ENCAP) return 0; return cpu_to_be16(MLX5_CAP_ROCE(dev->mdev, r_roce_min_src_udp_port)); } static int mlx5_use_mad_ifc(struct mlx5_ib_dev *dev) { if (MLX5_CAP_GEN(dev->mdev, port_type) == MLX5_CAP_PORT_TYPE_IB) return !MLX5_CAP_GEN(dev->mdev, ib_virt); return 0; } enum { MLX5_VPORT_ACCESS_METHOD_MAD, MLX5_VPORT_ACCESS_METHOD_HCA, MLX5_VPORT_ACCESS_METHOD_NIC, }; static int mlx5_get_vport_access_method(struct ib_device *ibdev) { if (mlx5_use_mad_ifc(to_mdev(ibdev))) return MLX5_VPORT_ACCESS_METHOD_MAD; if (mlx5_ib_port_link_layer(ibdev, 1) == IB_LINK_LAYER_ETHERNET) return MLX5_VPORT_ACCESS_METHOD_NIC; return MLX5_VPORT_ACCESS_METHOD_HCA; } static void get_atomic_caps(struct mlx5_ib_dev *dev, struct ib_device_attr *props) { u8 tmp; u8 atomic_operations = MLX5_CAP_ATOMIC(dev->mdev, atomic_operations); u8 atomic_size_qp = MLX5_CAP_ATOMIC(dev->mdev, atomic_size_qp); u8 atomic_req_8B_endianness_mode = MLX5_CAP_ATOMIC(dev->mdev, atomic_req_8B_endianess_mode); /* Check if HW supports 8 bytes standard atomic operations and capable * of host endianness respond */ tmp = MLX5_ATOMIC_OPS_CMP_SWAP | MLX5_ATOMIC_OPS_FETCH_ADD; if (((atomic_operations & tmp) == tmp) && (atomic_size_qp & MLX5_ATOMIC_SIZE_QP_8BYTES) && (atomic_req_8B_endianness_mode)) { props->atomic_cap = IB_ATOMIC_HCA; } else { props->atomic_cap = IB_ATOMIC_NONE; } } static int mlx5_query_system_image_guid(struct ib_device *ibdev, __be64 *sys_image_guid) { struct mlx5_ib_dev *dev = to_mdev(ibdev); struct mlx5_core_dev *mdev = dev->mdev; u64 tmp; int err; switch (mlx5_get_vport_access_method(ibdev)) { case MLX5_VPORT_ACCESS_METHOD_MAD: return mlx5_query_mad_ifc_system_image_guid(ibdev, sys_image_guid); case MLX5_VPORT_ACCESS_METHOD_HCA: err = mlx5_query_hca_vport_system_image_guid(mdev, &tmp); break; case MLX5_VPORT_ACCESS_METHOD_NIC: err = mlx5_query_nic_vport_system_image_guid(mdev, &tmp); break; default: return -EINVAL; } if (!err) *sys_image_guid = cpu_to_be64(tmp); return err; } static int mlx5_query_max_pkeys(struct ib_device *ibdev, u16 *max_pkeys) { struct mlx5_ib_dev *dev = to_mdev(ibdev); struct mlx5_core_dev *mdev = dev->mdev; switch (mlx5_get_vport_access_method(ibdev)) { case MLX5_VPORT_ACCESS_METHOD_MAD: return mlx5_query_mad_ifc_max_pkeys(ibdev, max_pkeys); case MLX5_VPORT_ACCESS_METHOD_HCA: case MLX5_VPORT_ACCESS_METHOD_NIC: *max_pkeys = mlx5_to_sw_pkey_sz(MLX5_CAP_GEN(mdev, pkey_table_size)); return 0; default: return -EINVAL; } } static int mlx5_query_vendor_id(struct ib_device *ibdev, u32 *vendor_id) { struct mlx5_ib_dev *dev = to_mdev(ibdev); switch (mlx5_get_vport_access_method(ibdev)) { case MLX5_VPORT_ACCESS_METHOD_MAD: return mlx5_query_mad_ifc_vendor_id(ibdev, vendor_id); case MLX5_VPORT_ACCESS_METHOD_HCA: case MLX5_VPORT_ACCESS_METHOD_NIC: return mlx5_core_query_vendor_id(dev->mdev, vendor_id); default: return -EINVAL; } } static int mlx5_query_node_guid(struct mlx5_ib_dev *dev, __be64 *node_guid) { u64 tmp; int err; switch (mlx5_get_vport_access_method(&dev->ib_dev)) { case MLX5_VPORT_ACCESS_METHOD_MAD: return mlx5_query_mad_ifc_node_guid(dev, node_guid); case MLX5_VPORT_ACCESS_METHOD_HCA: err = mlx5_query_hca_vport_node_guid(dev->mdev, &tmp); break; case MLX5_VPORT_ACCESS_METHOD_NIC: err = mlx5_query_nic_vport_node_guid(dev->mdev, &tmp); break; default: return -EINVAL; } if (!err) *node_guid = cpu_to_be64(tmp); return err; } struct mlx5_reg_node_desc { u8 desc[IB_DEVICE_NODE_DESC_MAX]; }; static int mlx5_query_node_desc(struct mlx5_ib_dev *dev, char *node_desc) { struct mlx5_reg_node_desc in; if (mlx5_use_mad_ifc(dev)) return mlx5_query_mad_ifc_node_desc(dev, node_desc); memset(&in, 0, sizeof(in)); return mlx5_core_access_reg(dev->mdev, &in, sizeof(in), node_desc, sizeof(struct mlx5_reg_node_desc), MLX5_REG_NODE_DESC, 0, 0); } static int mlx5_ib_query_device(struct ib_device *ibdev, struct ib_device_attr *props, struct ib_udata *uhw) { struct mlx5_ib_dev *dev = to_mdev(ibdev); struct mlx5_core_dev *mdev = dev->mdev; int err = -ENOMEM; int max_rq_sg; int max_sq_sg; u64 min_page_size = 1ull << MLX5_CAP_GEN(mdev, log_pg_sz); struct mlx5_ib_query_device_resp resp = {}; size_t resp_len; u64 max_tso; resp_len = sizeof(resp.comp_mask) + sizeof(resp.response_length); if (uhw->outlen && uhw->outlen < resp_len) return -EINVAL; else resp.response_length = resp_len; if (uhw->inlen && !ib_is_udata_cleared(uhw, 0, uhw->inlen)) return -EINVAL; memset(props, 0, sizeof(*props)); err = mlx5_query_system_image_guid(ibdev, &props->sys_image_guid); if (err) return err; err = mlx5_query_max_pkeys(ibdev, &props->max_pkeys); if (err) return err; err = mlx5_query_vendor_id(ibdev, &props->vendor_id); if (err) return err; props->fw_ver = ((u64)fw_rev_maj(dev->mdev) << 32) | (fw_rev_min(dev->mdev) << 16) | fw_rev_sub(dev->mdev); props->device_cap_flags = IB_DEVICE_CHANGE_PHY_PORT | IB_DEVICE_PORT_ACTIVE_EVENT | IB_DEVICE_SYS_IMAGE_GUID | IB_DEVICE_RC_RNR_NAK_GEN; if (MLX5_CAP_GEN(mdev, pkv)) props->device_cap_flags |= IB_DEVICE_BAD_PKEY_CNTR; if (MLX5_CAP_GEN(mdev, qkv)) props->device_cap_flags |= IB_DEVICE_BAD_QKEY_CNTR; if (MLX5_CAP_GEN(mdev, apm)) props->device_cap_flags |= IB_DEVICE_AUTO_PATH_MIG; if (MLX5_CAP_GEN(mdev, xrc)) props->device_cap_flags |= IB_DEVICE_XRC; if (MLX5_CAP_GEN(mdev, imaicl)) { props->device_cap_flags |= IB_DEVICE_MEM_WINDOW | IB_DEVICE_MEM_WINDOW_TYPE_2B; props->max_mw = 1 << MLX5_CAP_GEN(mdev, log_max_mkey); /* We support 'Gappy' memory registration too */ props->device_cap_flags |= IB_DEVICE_SG_GAPS_REG; } props->device_cap_flags |= IB_DEVICE_MEM_MGT_EXTENSIONS; if (MLX5_CAP_GEN(mdev, sho)) { props->device_cap_flags |= IB_DEVICE_SIGNATURE_HANDOVER; /* At this stage no support for signature handover */ props->sig_prot_cap = IB_PROT_T10DIF_TYPE_1 | IB_PROT_T10DIF_TYPE_2 | IB_PROT_T10DIF_TYPE_3; props->sig_guard_cap = IB_GUARD_T10DIF_CRC | IB_GUARD_T10DIF_CSUM; } if (MLX5_CAP_GEN(mdev, block_lb_mc)) props->device_cap_flags |= IB_DEVICE_BLOCK_MULTICAST_LOOPBACK; if (MLX5_CAP_GEN(dev->mdev, eth_net_offloads)) { if (MLX5_CAP_ETH(mdev, csum_cap)) props->device_cap_flags |= IB_DEVICE_RAW_IP_CSUM; if (field_avail(typeof(resp), tso_caps, uhw->outlen)) { max_tso = MLX5_CAP_ETH(mdev, max_lso_cap); if (max_tso) { resp.tso_caps.max_tso = 1 << max_tso; resp.tso_caps.supported_qpts |= 1 << IB_QPT_RAW_PACKET; resp.response_length += sizeof(resp.tso_caps); } } if (field_avail(typeof(resp), rss_caps, uhw->outlen)) { resp.rss_caps.rx_hash_function = MLX5_RX_HASH_FUNC_TOEPLITZ; resp.rss_caps.rx_hash_fields_mask = MLX5_RX_HASH_SRC_IPV4 | MLX5_RX_HASH_DST_IPV4 | MLX5_RX_HASH_SRC_IPV6 | MLX5_RX_HASH_DST_IPV6 | MLX5_RX_HASH_SRC_PORT_TCP | MLX5_RX_HASH_DST_PORT_TCP | MLX5_RX_HASH_SRC_PORT_UDP | MLX5_RX_HASH_DST_PORT_UDP; resp.response_length += sizeof(resp.rss_caps); } } else { if (field_avail(typeof(resp), tso_caps, uhw->outlen)) resp.response_length += sizeof(resp.tso_caps); if (field_avail(typeof(resp), rss_caps, uhw->outlen)) resp.response_length += sizeof(resp.rss_caps); } if (MLX5_CAP_GEN(mdev, ipoib_ipoib_offloads)) { props->device_cap_flags |= IB_DEVICE_UD_IP_CSUM; props->device_cap_flags |= IB_DEVICE_UD_TSO; } if (MLX5_CAP_GEN(dev->mdev, eth_net_offloads) && MLX5_CAP_ETH(dev->mdev, scatter_fcs)) props->device_cap_flags |= IB_DEVICE_RAW_SCATTER_FCS; if (mlx5_get_flow_namespace(dev->mdev, MLX5_FLOW_NAMESPACE_BYPASS)) props->device_cap_flags |= IB_DEVICE_MANAGED_FLOW_STEERING; props->vendor_part_id = mdev->pdev->device; props->hw_ver = mdev->pdev->revision; props->max_mr_size = ~0ull; props->page_size_cap = ~(min_page_size - 1); props->max_qp = 1 << MLX5_CAP_GEN(mdev, log_max_qp); props->max_qp_wr = 1 << MLX5_CAP_GEN(mdev, log_max_qp_sz); max_rq_sg = MLX5_CAP_GEN(mdev, max_wqe_sz_rq) / sizeof(struct mlx5_wqe_data_seg); max_sq_sg = (MLX5_CAP_GEN(mdev, max_wqe_sz_sq) - sizeof(struct mlx5_wqe_ctrl_seg)) / sizeof(struct mlx5_wqe_data_seg); props->max_sge = min(max_rq_sg, max_sq_sg); props->max_sge_rd = MLX5_MAX_SGE_RD; props->max_cq = 1 << MLX5_CAP_GEN(mdev, log_max_cq); props->max_cqe = (1 << MLX5_CAP_GEN(mdev, log_max_cq_sz)) - 1; props->max_mr = 1 << MLX5_CAP_GEN(mdev, log_max_mkey); props->max_pd = 1 << MLX5_CAP_GEN(mdev, log_max_pd); props->max_qp_rd_atom = 1 << MLX5_CAP_GEN(mdev, log_max_ra_req_qp); props->max_qp_init_rd_atom = 1 << MLX5_CAP_GEN(mdev, log_max_ra_res_qp); props->max_srq = 1 << MLX5_CAP_GEN(mdev, log_max_srq); props->max_srq_wr = (1 << MLX5_CAP_GEN(mdev, log_max_srq_sz)) - 1; props->local_ca_ack_delay = MLX5_CAP_GEN(mdev, local_ca_ack_delay); props->max_res_rd_atom = props->max_qp_rd_atom * props->max_qp; props->max_srq_sge = max_rq_sg - 1; props->max_fast_reg_page_list_len = 1 << MLX5_CAP_GEN(mdev, log_max_klm_list_size); get_atomic_caps(dev, props); props->masked_atomic_cap = IB_ATOMIC_NONE; props->max_mcast_grp = 1 << MLX5_CAP_GEN(mdev, log_max_mcg); props->max_mcast_qp_attach = MLX5_CAP_GEN(mdev, max_qp_mcg); props->max_total_mcast_qp_attach = props->max_mcast_qp_attach * props->max_mcast_grp; props->max_map_per_fmr = INT_MAX; /* no limit in ConnectIB */ props->hca_core_clock = MLX5_CAP_GEN(mdev, device_frequency_khz); props->timestamp_mask = 0x7FFFFFFFFFFFFFFFULL; #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING if (MLX5_CAP_GEN(mdev, pg)) props->device_cap_flags |= IB_DEVICE_ON_DEMAND_PAGING; props->odp_caps = dev->odp_caps; #endif if (MLX5_CAP_GEN(mdev, cd)) props->device_cap_flags |= IB_DEVICE_CROSS_CHANNEL; if (!mlx5_core_is_pf(mdev)) props->device_cap_flags |= IB_DEVICE_VIRTUAL_FUNCTION; if (mlx5_ib_port_link_layer(ibdev, 1) == IB_LINK_LAYER_ETHERNET) { props->rss_caps.max_rwq_indirection_tables = 1 << MLX5_CAP_GEN(dev->mdev, log_max_rqt); props->rss_caps.max_rwq_indirection_table_size = 1 << MLX5_CAP_GEN(dev->mdev, log_max_rqt_size); props->rss_caps.supported_qpts = 1 << IB_QPT_RAW_PACKET; props->max_wq_type_rq = 1 << MLX5_CAP_GEN(dev->mdev, log_max_rq); } if (uhw->outlen) { err = ib_copy_to_udata(uhw, &resp, resp.response_length); if (err) return err; } return 0; } enum mlx5_ib_width { MLX5_IB_WIDTH_1X = 1 << 0, MLX5_IB_WIDTH_2X = 1 << 1, MLX5_IB_WIDTH_4X = 1 << 2, MLX5_IB_WIDTH_8X = 1 << 3, MLX5_IB_WIDTH_12X = 1 << 4 }; static int translate_active_width(struct ib_device *ibdev, u8 active_width, u8 *ib_width) { struct mlx5_ib_dev *dev = to_mdev(ibdev); int err = 0; if (active_width & MLX5_IB_WIDTH_1X) { *ib_width = IB_WIDTH_1X; } else if (active_width & MLX5_IB_WIDTH_2X) { mlx5_ib_dbg(dev, "active_width %d is not supported by IB spec\n", (int)active_width); err = -EINVAL; } else if (active_width & MLX5_IB_WIDTH_4X) { *ib_width = IB_WIDTH_4X; } else if (active_width & MLX5_IB_WIDTH_8X) { *ib_width = IB_WIDTH_8X; } else if (active_width & MLX5_IB_WIDTH_12X) { *ib_width = IB_WIDTH_12X; } else { mlx5_ib_dbg(dev, "Invalid active_width %d\n", (int)active_width); err = -EINVAL; } return err; } enum ib_max_vl_num { __IB_MAX_VL_0 = 1, __IB_MAX_VL_0_1 = 2, __IB_MAX_VL_0_3 = 3, __IB_MAX_VL_0_7 = 4, __IB_MAX_VL_0_14 = 5, }; enum mlx5_vl_hw_cap { MLX5_VL_HW_0 = 1, MLX5_VL_HW_0_1 = 2, MLX5_VL_HW_0_2 = 3, MLX5_VL_HW_0_3 = 4, MLX5_VL_HW_0_4 = 5, MLX5_VL_HW_0_5 = 6, MLX5_VL_HW_0_6 = 7, MLX5_VL_HW_0_7 = 8, MLX5_VL_HW_0_14 = 15 }; static int translate_max_vl_num(struct ib_device *ibdev, u8 vl_hw_cap, u8 *max_vl_num) { switch (vl_hw_cap) { case MLX5_VL_HW_0: *max_vl_num = __IB_MAX_VL_0; break; case MLX5_VL_HW_0_1: *max_vl_num = __IB_MAX_VL_0_1; break; case MLX5_VL_HW_0_3: *max_vl_num = __IB_MAX_VL_0_3; break; case MLX5_VL_HW_0_7: *max_vl_num = __IB_MAX_VL_0_7; break; case MLX5_VL_HW_0_14: *max_vl_num = __IB_MAX_VL_0_14; break; default: return -EINVAL; } return 0; } static int mlx5_query_hca_port(struct ib_device *ibdev, u8 port, struct ib_port_attr *props) { struct mlx5_ib_dev *dev = to_mdev(ibdev); struct mlx5_core_dev *mdev = dev->mdev; u32 *rep; int replen = MLX5_ST_SZ_BYTES(query_hca_vport_context_out); struct mlx5_ptys_reg *ptys; struct mlx5_pmtu_reg *pmtu; struct mlx5_pvlc_reg pvlc; void *ctx; int err; rep = mlx5_vzalloc(replen); ptys = kzalloc(sizeof(*ptys), GFP_KERNEL); pmtu = kzalloc(sizeof(*pmtu), GFP_KERNEL); if (!rep || !ptys || !pmtu) { err = -ENOMEM; goto out; } memset(props, 0, sizeof(*props)); err = mlx5_query_hca_vport_context(mdev, port, 0, rep, replen); if (err) goto out; ctx = MLX5_ADDR_OF(query_hca_vport_context_out, rep, hca_vport_context); props->lid = MLX5_GET(hca_vport_context, ctx, lid); props->lmc = MLX5_GET(hca_vport_context, ctx, lmc); props->sm_lid = MLX5_GET(hca_vport_context, ctx, sm_lid); props->sm_sl = MLX5_GET(hca_vport_context, ctx, sm_sl); props->state = MLX5_GET(hca_vport_context, ctx, vport_state); props->phys_state = MLX5_GET(hca_vport_context, ctx, port_physical_state); props->port_cap_flags = MLX5_GET(hca_vport_context, ctx, cap_mask1); props->gid_tbl_len = mlx5_get_gid_table_len(MLX5_CAP_GEN(mdev, gid_table_size)); props->max_msg_sz = 1 << MLX5_CAP_GEN(mdev, log_max_msg); props->pkey_tbl_len = mlx5_to_sw_pkey_sz(MLX5_CAP_GEN(mdev, pkey_table_size)); props->bad_pkey_cntr = MLX5_GET(hca_vport_context, ctx, pkey_violation_counter); props->qkey_viol_cntr = MLX5_GET(hca_vport_context, ctx, qkey_violation_counter); props->subnet_timeout = MLX5_GET(hca_vport_context, ctx, subnet_timeout); props->init_type_reply = MLX5_GET(hca_vport_context, ctx, init_type_reply); props->grh_required = MLX5_GET(hca_vport_context, ctx, grh_required); ptys->proto_mask |= MLX5_PTYS_IB; ptys->local_port = port; err = mlx5_core_access_ptys(mdev, ptys, 0); if (err) goto out; err = translate_active_width(ibdev, ptys->ib_link_width_oper, &props->active_width); if (err) goto out; props->active_speed = (u8)ptys->ib_proto_oper; pmtu->local_port = port; err = mlx5_core_access_pmtu(mdev, pmtu, 0); if (err) goto out; props->max_mtu = pmtu->max_mtu; props->active_mtu = pmtu->oper_mtu; memset(&pvlc, 0, sizeof(pvlc)); pvlc.local_port = port; err = mlx5_core_access_pvlc(mdev, &pvlc, 0); if (err) goto out; err = translate_max_vl_num(ibdev, pvlc.vl_hw_cap, &props->max_vl_num); out: kvfree(rep); kfree(ptys); kfree(pmtu); return err; } int mlx5_ib_query_port(struct ib_device *ibdev, u8 port, struct ib_port_attr *props) { switch (mlx5_get_vport_access_method(ibdev)) { case MLX5_VPORT_ACCESS_METHOD_MAD: return mlx5_query_mad_ifc_port(ibdev, port, props); case MLX5_VPORT_ACCESS_METHOD_HCA: return mlx5_query_hca_port(ibdev, port, props); case MLX5_VPORT_ACCESS_METHOD_NIC: return mlx5_query_port_roce(ibdev, port, props); default: return -EINVAL; } } static int mlx5_ib_query_gid(struct ib_device *ibdev, u8 port, int index, union ib_gid *gid) { struct mlx5_ib_dev *dev = to_mdev(ibdev); struct mlx5_core_dev *mdev = dev->mdev; switch (mlx5_get_vport_access_method(ibdev)) { case MLX5_VPORT_ACCESS_METHOD_MAD: return mlx5_query_mad_ifc_gids(ibdev, port, index, gid); case MLX5_VPORT_ACCESS_METHOD_HCA: return mlx5_query_hca_vport_gid(mdev, port, 0, index, gid); default: return -EINVAL; } } static int mlx5_ib_query_pkey(struct ib_device *ibdev, u8 port, u16 index, u16 *pkey) { struct mlx5_ib_dev *dev = to_mdev(ibdev); struct mlx5_core_dev *mdev = dev->mdev; switch (mlx5_get_vport_access_method(ibdev)) { case MLX5_VPORT_ACCESS_METHOD_MAD: return mlx5_query_mad_ifc_pkey(ibdev, port, index, pkey); case MLX5_VPORT_ACCESS_METHOD_HCA: case MLX5_VPORT_ACCESS_METHOD_NIC: return mlx5_query_hca_vport_pkey(mdev, 0, port, 0, index, pkey); default: return -EINVAL; } } static int mlx5_ib_modify_device(struct ib_device *ibdev, int mask, struct ib_device_modify *props) { struct mlx5_ib_dev *dev = to_mdev(ibdev); struct mlx5_reg_node_desc in; struct mlx5_reg_node_desc out; int err; if (mask & ~IB_DEVICE_MODIFY_NODE_DESC) return -EOPNOTSUPP; if (!(mask & IB_DEVICE_MODIFY_NODE_DESC)) return 0; /* * If possible, pass node desc to FW, so it can generate * a 144 trap. If cmd fails, just ignore. */ memcpy(&in, props->node_desc, IB_DEVICE_NODE_DESC_MAX); err = mlx5_core_access_reg(dev->mdev, &in, sizeof(in), &out, sizeof(out), MLX5_REG_NODE_DESC, 0, 1); if (err) return err; memcpy(ibdev->node_desc, props->node_desc, IB_DEVICE_NODE_DESC_MAX); return err; } static int mlx5_ib_modify_port(struct ib_device *ibdev, u8 port, int mask, struct ib_port_modify *props) { struct mlx5_ib_dev *dev = to_mdev(ibdev); struct ib_port_attr attr; u32 tmp; int err; mutex_lock(&dev->cap_mask_mutex); err = mlx5_ib_query_port(ibdev, port, &attr); if (err) goto out; tmp = (attr.port_cap_flags | props->set_port_cap_mask) & ~props->clr_port_cap_mask; err = mlx5_set_port_caps(dev->mdev, port, tmp); out: mutex_unlock(&dev->cap_mask_mutex); return err; } static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev, struct ib_udata *udata) { struct mlx5_ib_dev *dev = to_mdev(ibdev); struct mlx5_ib_alloc_ucontext_req_v2 req = {}; struct mlx5_ib_alloc_ucontext_resp resp = {}; struct mlx5_ib_ucontext *context; struct mlx5_uuar_info *uuari; struct mlx5_uar *uars; int gross_uuars; int num_uars; int ver; int uuarn; int err; int i; size_t reqlen; size_t min_req_v2 = offsetof(struct mlx5_ib_alloc_ucontext_req_v2, max_cqe_version); if (!dev->ib_active) return ERR_PTR(-EAGAIN); if (udata->inlen < sizeof(struct ib_uverbs_cmd_hdr)) return ERR_PTR(-EINVAL); reqlen = udata->inlen - sizeof(struct ib_uverbs_cmd_hdr); if (reqlen == sizeof(struct mlx5_ib_alloc_ucontext_req)) ver = 0; else if (reqlen >= min_req_v2) ver = 2; else return ERR_PTR(-EINVAL); err = ib_copy_from_udata(&req, udata, min(reqlen, sizeof(req))); if (err) return ERR_PTR(err); if (req.flags) return ERR_PTR(-EINVAL); if (req.total_num_uuars > MLX5_MAX_UUARS) return ERR_PTR(-ENOMEM); if (req.total_num_uuars == 0) return ERR_PTR(-EINVAL); if (req.comp_mask || req.reserved0 || req.reserved1 || req.reserved2) return ERR_PTR(-EOPNOTSUPP); if (reqlen > sizeof(req) && !ib_is_udata_cleared(udata, sizeof(req), reqlen - sizeof(req))) return ERR_PTR(-EOPNOTSUPP); req.total_num_uuars = ALIGN(req.total_num_uuars, MLX5_NON_FP_BF_REGS_PER_PAGE); if (req.num_low_latency_uuars > req.total_num_uuars - 1) return ERR_PTR(-EINVAL); num_uars = req.total_num_uuars / MLX5_NON_FP_BF_REGS_PER_PAGE; gross_uuars = num_uars * MLX5_BF_REGS_PER_PAGE; resp.qp_tab_size = 1 << MLX5_CAP_GEN(dev->mdev, log_max_qp); if (mlx5_core_is_pf(dev->mdev) && MLX5_CAP_GEN(dev->mdev, bf)) resp.bf_reg_size = 1 << MLX5_CAP_GEN(dev->mdev, log_bf_reg_size); resp.cache_line_size = cache_line_size(); resp.max_sq_desc_sz = MLX5_CAP_GEN(dev->mdev, max_wqe_sz_sq); resp.max_rq_desc_sz = MLX5_CAP_GEN(dev->mdev, max_wqe_sz_rq); resp.max_send_wqebb = 1 << MLX5_CAP_GEN(dev->mdev, log_max_qp_sz); resp.max_recv_wr = 1 << MLX5_CAP_GEN(dev->mdev, log_max_qp_sz); resp.max_srq_recv_wr = 1 << MLX5_CAP_GEN(dev->mdev, log_max_srq_sz); resp.cqe_version = min_t(__u8, (__u8)MLX5_CAP_GEN(dev->mdev, cqe_version), req.max_cqe_version); resp.response_length = min(offsetof(typeof(resp), response_length) + sizeof(resp.response_length), udata->outlen); context = kzalloc(sizeof(*context), GFP_KERNEL); if (!context) return ERR_PTR(-ENOMEM); uuari = &context->uuari; mutex_init(&uuari->lock); uars = kcalloc(num_uars, sizeof(*uars), GFP_KERNEL); if (!uars) { err = -ENOMEM; goto out_ctx; } uuari->bitmap = kcalloc(BITS_TO_LONGS(gross_uuars), sizeof(*uuari->bitmap), GFP_KERNEL); if (!uuari->bitmap) { err = -ENOMEM; goto out_uar_ctx; } /* * clear all fast path uuars */ for (i = 0; i < gross_uuars; i++) { uuarn = i & 3; if (uuarn == 2 || uuarn == 3) set_bit(i, uuari->bitmap); } uuari->count = kcalloc(gross_uuars, sizeof(*uuari->count), GFP_KERNEL); if (!uuari->count) { err = -ENOMEM; goto out_bitmap; } for (i = 0; i < num_uars; i++) { err = mlx5_cmd_alloc_uar(dev->mdev, &uars[i].index); if (err) goto out_count; } #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING context->ibucontext.invalidate_range = &mlx5_ib_invalidate_range; #endif if (MLX5_CAP_GEN(dev->mdev, log_max_transport_domain)) { err = mlx5_alloc_transport_domain(dev->mdev, &context->tdn); if (err) goto out_uars; } INIT_LIST_HEAD(&context->vma_private_list); INIT_LIST_HEAD(&context->db_page_list); mutex_init(&context->db_page_mutex); resp.tot_uuars = req.total_num_uuars; resp.num_ports = MLX5_CAP_GEN(dev->mdev, num_ports); if (field_avail(typeof(resp), cqe_version, udata->outlen)) resp.response_length += sizeof(resp.cqe_version); if (field_avail(typeof(resp), cmds_supp_uhw, udata->outlen)) { resp.cmds_supp_uhw |= MLX5_USER_CMDS_SUPP_UHW_QUERY_DEVICE; resp.response_length += sizeof(resp.cmds_supp_uhw); } /* * We don't want to expose information from the PCI bar that is located * after 4096 bytes, so if the arch only supports larger pages, let's * pretend we don't support reading the HCA's core clock. This is also * forced by mmap function. */ if (PAGE_SIZE <= 4096 && field_avail(typeof(resp), hca_core_clock_offset, udata->outlen)) { resp.comp_mask |= MLX5_IB_ALLOC_UCONTEXT_RESP_MASK_CORE_CLOCK_OFFSET; resp.hca_core_clock_offset = offsetof(struct mlx5_init_seg, internal_timer_h) % PAGE_SIZE; resp.response_length += sizeof(resp.hca_core_clock_offset) + sizeof(resp.reserved2); } err = ib_copy_to_udata(udata, &resp, resp.response_length); if (err) goto out_td; uuari->ver = ver; uuari->num_low_latency_uuars = req.num_low_latency_uuars; uuari->uars = uars; uuari->num_uars = num_uars; context->cqe_version = resp.cqe_version; return &context->ibucontext; out_td: if (MLX5_CAP_GEN(dev->mdev, log_max_transport_domain)) mlx5_dealloc_transport_domain(dev->mdev, context->tdn); out_uars: for (i--; i >= 0; i--) mlx5_cmd_free_uar(dev->mdev, uars[i].index); out_count: kfree(uuari->count); out_bitmap: kfree(uuari->bitmap); out_uar_ctx: kfree(uars); out_ctx: kfree(context); return ERR_PTR(err); } static int mlx5_ib_dealloc_ucontext(struct ib_ucontext *ibcontext) { struct mlx5_ib_ucontext *context = to_mucontext(ibcontext); struct mlx5_ib_dev *dev = to_mdev(ibcontext->device); struct mlx5_uuar_info *uuari = &context->uuari; int i; if (MLX5_CAP_GEN(dev->mdev, log_max_transport_domain)) mlx5_dealloc_transport_domain(dev->mdev, context->tdn); for (i = 0; i < uuari->num_uars; i++) { if (mlx5_cmd_free_uar(dev->mdev, uuari->uars[i].index)) mlx5_ib_warn(dev, "failed to free UAR 0x%x\n", uuari->uars[i].index); } kfree(uuari->count); kfree(uuari->bitmap); kfree(uuari->uars); kfree(context); return 0; } static phys_addr_t uar_index2pfn(struct mlx5_ib_dev *dev, int index) { return (pci_resource_start(dev->mdev->pdev, 0) >> PAGE_SHIFT) + index; } static int get_command(unsigned long offset) { return (offset >> MLX5_IB_MMAP_CMD_SHIFT) & MLX5_IB_MMAP_CMD_MASK; } static int get_arg(unsigned long offset) { return offset & ((1 << MLX5_IB_MMAP_CMD_SHIFT) - 1); } static int get_index(unsigned long offset) { return get_arg(offset); } static void mlx5_ib_vma_open(struct vm_area_struct *area) { /* vma_open is called when a new VMA is created on top of our VMA. This * is done through either mremap flow or split_vma (usually due to * mlock, madvise, munmap, etc.) We do not support a clone of the VMA, * as this VMA is strongly hardware related. Therefore we set the * vm_ops of the newly created/cloned VMA to NULL, to prevent it from * calling us again and trying to do incorrect actions. We assume that * the original VMA size is exactly a single page, and therefore all * "splitting" operation will not happen to it. */ area->vm_ops = NULL; } static void mlx5_ib_vma_close(struct vm_area_struct *area) { struct mlx5_ib_vma_private_data *mlx5_ib_vma_priv_data; /* It's guaranteed that all VMAs opened on a FD are closed before the * file itself is closed, therefore no sync is needed with the regular * closing flow. (e.g. mlx5 ib_dealloc_ucontext) * However need a sync with accessing the vma as part of * mlx5_ib_disassociate_ucontext. * The close operation is usually called under mm->mmap_sem except when * process is exiting. * The exiting case is handled explicitly as part of * mlx5_ib_disassociate_ucontext. */ mlx5_ib_vma_priv_data = (struct mlx5_ib_vma_private_data *)area->vm_private_data; /* setting the vma context pointer to null in the mlx5_ib driver's * private data, to protect a race condition in * mlx5_ib_disassociate_ucontext(). */ mlx5_ib_vma_priv_data->vma = NULL; list_del(&mlx5_ib_vma_priv_data->list); kfree(mlx5_ib_vma_priv_data); } static const struct vm_operations_struct mlx5_ib_vm_ops = { .open = mlx5_ib_vma_open, .close = mlx5_ib_vma_close }; static int mlx5_ib_set_vma_data(struct vm_area_struct *vma, struct mlx5_ib_ucontext *ctx) { struct mlx5_ib_vma_private_data *vma_prv; struct list_head *vma_head = &ctx->vma_private_list; vma_prv = kzalloc(sizeof(*vma_prv), GFP_KERNEL); if (!vma_prv) return -ENOMEM; vma_prv->vma = vma; vma->vm_private_data = vma_prv; vma->vm_ops = &mlx5_ib_vm_ops; list_add(&vma_prv->list, vma_head); return 0; } static void mlx5_ib_disassociate_ucontext(struct ib_ucontext *ibcontext) { int ret; struct vm_area_struct *vma; struct mlx5_ib_vma_private_data *vma_private, *n; struct mlx5_ib_ucontext *context = to_mucontext(ibcontext); struct task_struct *owning_process = NULL; struct mm_struct *owning_mm = NULL; owning_process = get_pid_task(ibcontext->tgid, PIDTYPE_PID); if (!owning_process) return; owning_mm = get_task_mm(owning_process); if (!owning_mm) { pr_info("no mm, disassociate ucontext is pending task termination\n"); while (1) { put_task_struct(owning_process); usleep_range(1000, 2000); owning_process = get_pid_task(ibcontext->tgid, PIDTYPE_PID); if (!owning_process /* || owning_process->state == TASK_DEAD */) { pr_info("disassociate ucontext done, task was terminated\n"); /* in case task was dead need to release the * task struct. */ if (owning_process) put_task_struct(owning_process); return; } } } /* need to protect from a race on closing the vma as part of * mlx5_ib_vma_close. */ down_read(&owning_mm->mmap_sem); list_for_each_entry_safe(vma_private, n, &context->vma_private_list, list) { vma = vma_private->vma; ret = zap_vma_ptes(vma, vma->vm_start, PAGE_SIZE); WARN_ONCE(ret, "%s: zap_vma_ptes failed", __func__); /* context going to be destroyed, should * not access ops any more. */ vma->vm_ops = NULL; list_del(&vma_private->list); kfree(vma_private); } up_read(&owning_mm->mmap_sem); mmput(owning_mm); put_task_struct(owning_process); } static inline char *mmap_cmd2str(enum mlx5_ib_mmap_cmd cmd) { switch (cmd) { case MLX5_IB_MMAP_WC_PAGE: return "WC"; case MLX5_IB_MMAP_REGULAR_PAGE: return "best effort WC"; case MLX5_IB_MMAP_NC_PAGE: return "NC"; default: return NULL; } } static int uar_mmap(struct mlx5_ib_dev *dev, enum mlx5_ib_mmap_cmd cmd, struct vm_area_struct *vma, struct mlx5_ib_ucontext *context) { struct mlx5_uuar_info *uuari = &context->uuari; int err; unsigned long idx; phys_addr_t pfn, pa; pgprot_t prot; switch (cmd) { case MLX5_IB_MMAP_WC_PAGE: /* Some architectures don't support WC memory */ #if defined(CONFIG_X86) if (!pat_enabled()) return -EPERM; #elif !(defined(CONFIG_PPC) || (defined(CONFIG_ARM) && defined(CONFIG_MMU))) return -EPERM; #endif /* fall through */ case MLX5_IB_MMAP_REGULAR_PAGE: /* For MLX5_IB_MMAP_REGULAR_PAGE do the best effort to get WC */ prot = pgprot_writecombine(vma->vm_page_prot); break; case MLX5_IB_MMAP_NC_PAGE: prot = pgprot_noncached(vma->vm_page_prot); break; default: return -EINVAL; } if (vma->vm_end - vma->vm_start != PAGE_SIZE) return -EINVAL; idx = get_index(vma->vm_pgoff); if (idx >= uuari->num_uars) return -EINVAL; pfn = uar_index2pfn(dev, uuari->uars[idx].index); mlx5_ib_dbg(dev, "uar idx 0x%lx, pfn %pa\n", idx, &pfn); vma->vm_page_prot = prot; err = io_remap_pfn_range(vma, vma->vm_start, pfn, PAGE_SIZE, vma->vm_page_prot); if (err) { mlx5_ib_err(dev, "io_remap_pfn_range failed with error=%d, vm_start=0x%llx, pfn=%pa, mmap_cmd=%s\n", err, (unsigned long long)vma->vm_start, &pfn, mmap_cmd2str(cmd)); return -EAGAIN; } pa = pfn << PAGE_SHIFT; mlx5_ib_dbg(dev, "mapped %s at 0x%llx, PA %pa\n", mmap_cmd2str(cmd), (unsigned long long)vma->vm_start, &pa); return mlx5_ib_set_vma_data(vma, context); } static int mlx5_ib_mmap(struct ib_ucontext *ibcontext, struct vm_area_struct *vma) { struct mlx5_ib_ucontext *context = to_mucontext(ibcontext); struct mlx5_ib_dev *dev = to_mdev(ibcontext->device); unsigned long command; phys_addr_t pfn; command = get_command(vma->vm_pgoff); switch (command) { case MLX5_IB_MMAP_WC_PAGE: case MLX5_IB_MMAP_NC_PAGE: case MLX5_IB_MMAP_REGULAR_PAGE: return uar_mmap(dev, command, vma, context); case MLX5_IB_MMAP_GET_CONTIGUOUS_PAGES: return -ENOSYS; case MLX5_IB_MMAP_CORE_CLOCK: if (vma->vm_end - vma->vm_start != PAGE_SIZE) return -EINVAL; if (vma->vm_flags & VM_WRITE) return -EPERM; /* Don't expose to user-space information it shouldn't have */ if (PAGE_SIZE > 4096) return -EOPNOTSUPP; vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot); pfn = (dev->mdev->iseg_base + offsetof(struct mlx5_init_seg, internal_timer_h)) >> PAGE_SHIFT; if (io_remap_pfn_range(vma, vma->vm_start, pfn, PAGE_SIZE, vma->vm_page_prot)) return -EAGAIN; mlx5_ib_dbg(dev, "mapped internal timer at 0x%llx, PA 0x%llx\n", (unsigned long long)vma->vm_start, (unsigned long long)pfn << PAGE_SHIFT); break; default: return -EINVAL; } return 0; } static struct ib_pd *mlx5_ib_alloc_pd(struct ib_device *ibdev, struct ib_ucontext *context, struct ib_udata *udata) { struct mlx5_ib_alloc_pd_resp resp; struct mlx5_ib_pd *pd; int err; pd = kmalloc(sizeof(*pd), GFP_KERNEL); if (!pd) return ERR_PTR(-ENOMEM); err = mlx5_core_alloc_pd(to_mdev(ibdev)->mdev, &pd->pdn); if (err) { kfree(pd); return ERR_PTR(err); } if (context) { resp.pdn = pd->pdn; if (ib_copy_to_udata(udata, &resp, sizeof(resp))) { mlx5_core_dealloc_pd(to_mdev(ibdev)->mdev, pd->pdn); kfree(pd); return ERR_PTR(-EFAULT); } } return &pd->ibpd; } static int mlx5_ib_dealloc_pd(struct ib_pd *pd) { struct mlx5_ib_dev *mdev = to_mdev(pd->device); struct mlx5_ib_pd *mpd = to_mpd(pd); mlx5_core_dealloc_pd(mdev->mdev, mpd->pdn); kfree(mpd); return 0; } enum { MATCH_CRITERIA_ENABLE_OUTER_BIT, MATCH_CRITERIA_ENABLE_MISC_BIT, MATCH_CRITERIA_ENABLE_INNER_BIT }; #define HEADER_IS_ZERO(match_criteria, headers) \ !(memchr_inv(MLX5_ADDR_OF(fte_match_param, match_criteria, headers), \ 0, MLX5_FLD_SZ_BYTES(fte_match_param, headers))) \ static u8 get_match_criteria_enable(u32 *match_criteria) { u8 match_criteria_enable; match_criteria_enable = (!HEADER_IS_ZERO(match_criteria, outer_headers)) << MATCH_CRITERIA_ENABLE_OUTER_BIT; match_criteria_enable |= (!HEADER_IS_ZERO(match_criteria, misc_parameters)) << MATCH_CRITERIA_ENABLE_MISC_BIT; match_criteria_enable |= (!HEADER_IS_ZERO(match_criteria, inner_headers)) << MATCH_CRITERIA_ENABLE_INNER_BIT; return match_criteria_enable; } static void set_proto(void *outer_c, void *outer_v, u8 mask, u8 val) { MLX5_SET(fte_match_set_lyr_2_4, outer_c, ip_protocol, mask); MLX5_SET(fte_match_set_lyr_2_4, outer_v, ip_protocol, val); } static void set_tos(void *outer_c, void *outer_v, u8 mask, u8 val) { MLX5_SET(fte_match_set_lyr_2_4, outer_c, ip_ecn, mask); MLX5_SET(fte_match_set_lyr_2_4, outer_v, ip_ecn, val); MLX5_SET(fte_match_set_lyr_2_4, outer_c, ip_dscp, mask >> 2); MLX5_SET(fte_match_set_lyr_2_4, outer_v, ip_dscp, val >> 2); } #define LAST_ETH_FIELD vlan_tag #define LAST_IB_FIELD sl #define LAST_IPV4_FIELD tos #define LAST_IPV6_FIELD traffic_class #define LAST_TCP_UDP_FIELD src_port /* Field is the last supported field */ #define FIELDS_NOT_SUPPORTED(filter, field)\ memchr_inv((void *)&filter.field +\ sizeof(filter.field), 0,\ sizeof(filter) -\ offsetof(typeof(filter), field) -\ sizeof(filter.field)) static int parse_flow_attr(u32 *match_c, u32 *match_v, const union ib_flow_spec *ib_spec) { void *outer_headers_c = MLX5_ADDR_OF(fte_match_param, match_c, outer_headers); void *outer_headers_v = MLX5_ADDR_OF(fte_match_param, match_v, outer_headers); void *misc_params_c = MLX5_ADDR_OF(fte_match_param, match_c, misc_parameters); void *misc_params_v = MLX5_ADDR_OF(fte_match_param, match_v, misc_parameters); switch (ib_spec->type) { case IB_FLOW_SPEC_ETH: if (FIELDS_NOT_SUPPORTED(ib_spec->eth.mask, LAST_ETH_FIELD)) return -ENOTSUPP; ether_addr_copy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_c, dmac_47_16), ib_spec->eth.mask.dst_mac); ether_addr_copy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_v, dmac_47_16), ib_spec->eth.val.dst_mac); ether_addr_copy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_c, smac_47_16), ib_spec->eth.mask.src_mac); ether_addr_copy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_v, smac_47_16), ib_spec->eth.val.src_mac); if (ib_spec->eth.mask.vlan_tag) { MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, cvlan_tag, 1); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, cvlan_tag, 1); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, first_vid, ntohs(ib_spec->eth.mask.vlan_tag)); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, first_vid, ntohs(ib_spec->eth.val.vlan_tag)); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, first_cfi, ntohs(ib_spec->eth.mask.vlan_tag) >> 12); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, first_cfi, ntohs(ib_spec->eth.val.vlan_tag) >> 12); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, first_prio, ntohs(ib_spec->eth.mask.vlan_tag) >> 13); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, first_prio, ntohs(ib_spec->eth.val.vlan_tag) >> 13); } MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, ethertype, ntohs(ib_spec->eth.mask.ether_type)); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, ethertype, ntohs(ib_spec->eth.val.ether_type)); break; case IB_FLOW_SPEC_IPV4: if (FIELDS_NOT_SUPPORTED(ib_spec->ipv4.mask, LAST_IPV4_FIELD)) return -ENOTSUPP; MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, ethertype, 0xffff); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, ethertype, ETH_P_IP); memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_c, src_ipv4_src_ipv6.ipv4_layout.ipv4), &ib_spec->ipv4.mask.src_ip, sizeof(ib_spec->ipv4.mask.src_ip)); memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_v, src_ipv4_src_ipv6.ipv4_layout.ipv4), &ib_spec->ipv4.val.src_ip, sizeof(ib_spec->ipv4.val.src_ip)); memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_c, dst_ipv4_dst_ipv6.ipv4_layout.ipv4), &ib_spec->ipv4.mask.dst_ip, sizeof(ib_spec->ipv4.mask.dst_ip)); memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_v, dst_ipv4_dst_ipv6.ipv4_layout.ipv4), &ib_spec->ipv4.val.dst_ip, sizeof(ib_spec->ipv4.val.dst_ip)); set_tos(outer_headers_c, outer_headers_v, ib_spec->ipv4.mask.tos, ib_spec->ipv4.val.tos); set_proto(outer_headers_c, outer_headers_v, ib_spec->ipv4.mask.proto, ib_spec->ipv4.val.proto); break; case IB_FLOW_SPEC_IPV6: if (FIELDS_NOT_SUPPORTED(ib_spec->ipv6.mask, LAST_IPV6_FIELD)) return -ENOTSUPP; MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, ethertype, 0xffff); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, ethertype, IPPROTO_IPV6); memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_c, src_ipv4_src_ipv6.ipv6_layout.ipv6), &ib_spec->ipv6.mask.src_ip, sizeof(ib_spec->ipv6.mask.src_ip)); memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_v, src_ipv4_src_ipv6.ipv6_layout.ipv6), &ib_spec->ipv6.val.src_ip, sizeof(ib_spec->ipv6.val.src_ip)); memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_c, dst_ipv4_dst_ipv6.ipv6_layout.ipv6), &ib_spec->ipv6.mask.dst_ip, sizeof(ib_spec->ipv6.mask.dst_ip)); memcpy(MLX5_ADDR_OF(fte_match_set_lyr_2_4, outer_headers_v, dst_ipv4_dst_ipv6.ipv6_layout.ipv6), &ib_spec->ipv6.val.dst_ip, sizeof(ib_spec->ipv6.val.dst_ip)); set_tos(outer_headers_c, outer_headers_v, ib_spec->ipv6.mask.traffic_class, ib_spec->ipv6.val.traffic_class); set_proto(outer_headers_c, outer_headers_v, ib_spec->ipv6.mask.next_hdr, ib_spec->ipv6.val.next_hdr); MLX5_SET(fte_match_set_misc, misc_params_c, outer_ipv6_flow_label, ntohl(ib_spec->ipv6.mask.flow_label)); MLX5_SET(fte_match_set_misc, misc_params_v, outer_ipv6_flow_label, ntohl(ib_spec->ipv6.val.flow_label)); break; case IB_FLOW_SPEC_TCP: if (FIELDS_NOT_SUPPORTED(ib_spec->tcp_udp.mask, LAST_TCP_UDP_FIELD)) return -ENOTSUPP; MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, ip_protocol, 0xff); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, ip_protocol, IPPROTO_TCP); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, tcp_sport, ntohs(ib_spec->tcp_udp.mask.src_port)); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, tcp_sport, ntohs(ib_spec->tcp_udp.val.src_port)); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, tcp_dport, ntohs(ib_spec->tcp_udp.mask.dst_port)); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, tcp_dport, ntohs(ib_spec->tcp_udp.val.dst_port)); break; case IB_FLOW_SPEC_UDP: if (FIELDS_NOT_SUPPORTED(ib_spec->tcp_udp.mask, LAST_TCP_UDP_FIELD)) return -ENOTSUPP; MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, ip_protocol, 0xff); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, ip_protocol, IPPROTO_UDP); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, udp_sport, ntohs(ib_spec->tcp_udp.mask.src_port)); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, udp_sport, ntohs(ib_spec->tcp_udp.val.src_port)); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_c, udp_dport, ntohs(ib_spec->tcp_udp.mask.dst_port)); MLX5_SET(fte_match_set_lyr_2_4, outer_headers_v, udp_dport, ntohs(ib_spec->tcp_udp.val.dst_port)); break; default: return -EINVAL; } return 0; } /* If a flow could catch both multicast and unicast packets, * it won't fall into the multicast flow steering table and this rule * could steal other multicast packets. */ static bool flow_is_multicast_only(struct ib_flow_attr *ib_attr) { struct ib_flow_spec_eth *eth_spec; if (ib_attr->type != IB_FLOW_ATTR_NORMAL || ib_attr->size < sizeof(struct ib_flow_attr) + sizeof(struct ib_flow_spec_eth) || ib_attr->num_of_specs < 1) return false; eth_spec = (struct ib_flow_spec_eth *)(ib_attr + 1); if (eth_spec->type != IB_FLOW_SPEC_ETH || eth_spec->size != sizeof(*eth_spec)) return false; return is_multicast_ether_addr(eth_spec->mask.dst_mac) && is_multicast_ether_addr(eth_spec->val.dst_mac); } static bool is_valid_attr(const struct ib_flow_attr *flow_attr) { union ib_flow_spec *ib_spec = (union ib_flow_spec *)(flow_attr + 1); bool has_ipv4_spec = false; bool eth_type_ipv4 = true; unsigned int spec_index; /* Validate that ethertype is correct */ for (spec_index = 0; spec_index < flow_attr->num_of_specs; spec_index++) { if (ib_spec->type == IB_FLOW_SPEC_ETH && ib_spec->eth.mask.ether_type) { if (!((ib_spec->eth.mask.ether_type == htons(0xffff)) && ib_spec->eth.val.ether_type == htons(ETH_P_IP))) eth_type_ipv4 = false; } else if (ib_spec->type == IB_FLOW_SPEC_IPV4) { has_ipv4_spec = true; } ib_spec = (void *)ib_spec + ib_spec->size; } return !has_ipv4_spec || eth_type_ipv4; } static void put_flow_table(struct mlx5_ib_dev *dev, struct mlx5_ib_flow_prio *prio, bool ft_added) { prio->refcount -= !!ft_added; if (!prio->refcount) { mlx5_destroy_flow_table(prio->flow_table); prio->flow_table = NULL; } } static int mlx5_ib_destroy_flow(struct ib_flow *flow_id) { struct mlx5_ib_dev *dev = to_mdev(flow_id->qp->device); struct mlx5_ib_flow_handler *handler = container_of(flow_id, struct mlx5_ib_flow_handler, ibflow); struct mlx5_ib_flow_handler *iter, *tmp; mutex_lock(&dev->flow_db.lock); list_for_each_entry_safe(iter, tmp, &handler->list, list) { mlx5_del_flow_rule(iter->rule); put_flow_table(dev, iter->prio, true); list_del(&iter->list); kfree(iter); } mlx5_del_flow_rule(handler->rule); put_flow_table(dev, handler->prio, true); mutex_unlock(&dev->flow_db.lock); kfree(handler); return 0; } static int ib_prio_to_core_prio(unsigned int priority, bool dont_trap) { priority *= 2; if (!dont_trap) priority++; return priority; } enum flow_table_type { MLX5_IB_FT_RX, MLX5_IB_FT_TX }; #define MLX5_FS_MAX_TYPES 10 #define MLX5_FS_MAX_ENTRIES 32000UL static struct mlx5_ib_flow_prio *get_flow_table(struct mlx5_ib_dev *dev, struct ib_flow_attr *flow_attr, enum flow_table_type ft_type) { bool dont_trap = flow_attr->flags & IB_FLOW_ATTR_FLAGS_DONT_TRAP; struct mlx5_flow_namespace *ns = NULL; struct mlx5_ib_flow_prio *prio; struct mlx5_flow_table *ft; int num_entries; int num_groups; int priority; int err = 0; if (flow_attr->type == IB_FLOW_ATTR_NORMAL) { if (flow_is_multicast_only(flow_attr) && !dont_trap) priority = MLX5_IB_FLOW_MCAST_PRIO; else priority = ib_prio_to_core_prio(flow_attr->priority, dont_trap); ns = mlx5_get_flow_namespace(dev->mdev, MLX5_FLOW_NAMESPACE_BYPASS); num_entries = MLX5_FS_MAX_ENTRIES; num_groups = MLX5_FS_MAX_TYPES; prio = &dev->flow_db.prios[priority]; } else if (flow_attr->type == IB_FLOW_ATTR_ALL_DEFAULT || flow_attr->type == IB_FLOW_ATTR_MC_DEFAULT) { ns = mlx5_get_flow_namespace(dev->mdev, MLX5_FLOW_NAMESPACE_LEFTOVERS); build_leftovers_ft_param("bypass", &priority, &num_entries, &num_groups); prio = &dev->flow_db.prios[MLX5_IB_FLOW_LEFTOVERS_PRIO]; } else if (flow_attr->type == IB_FLOW_ATTR_SNIFFER) { if (!MLX5_CAP_FLOWTABLE(dev->mdev, allow_sniffer_and_nic_rx_shared_tir)) return ERR_PTR(-ENOTSUPP); ns = mlx5_get_flow_namespace(dev->mdev, ft_type == MLX5_IB_FT_RX ? MLX5_FLOW_NAMESPACE_SNIFFER_RX : MLX5_FLOW_NAMESPACE_SNIFFER_TX); prio = &dev->flow_db.sniffer[ft_type]; priority = 0; num_entries = 1; num_groups = 1; } if (!ns) return ERR_PTR(-ENOTSUPP); ft = prio->flow_table; if (!ft) { ft = mlx5_create_auto_grouped_flow_table(ns, priority, "bypass", num_entries, num_groups); if (!IS_ERR(ft)) { prio->refcount = 0; prio->flow_table = ft; } else { err = PTR_ERR(ft); } } return err ? ERR_PTR(err) : prio; } static struct mlx5_ib_flow_handler *create_flow_rule(struct mlx5_ib_dev *dev, struct mlx5_ib_flow_prio *ft_prio, const struct ib_flow_attr *flow_attr, struct mlx5_flow_destination *dst) { struct mlx5_flow_table *ft = ft_prio->flow_table; struct mlx5_ib_flow_handler *handler; struct mlx5_flow_spec *spec; const void *ib_flow = (const void *)flow_attr + sizeof(*flow_attr); unsigned int spec_index; u32 action; int err = 0; if (!is_valid_attr(flow_attr)) return ERR_PTR(-EINVAL); spec = mlx5_vzalloc(sizeof(*spec)); handler = kzalloc(sizeof(*handler), GFP_KERNEL); if (!handler || !spec) { err = -ENOMEM; goto free; } INIT_LIST_HEAD(&handler->list); for (spec_index = 0; spec_index < flow_attr->num_of_specs; spec_index++) { err = parse_flow_attr(spec->match_criteria, spec->match_value, ib_flow); if (err < 0) goto free; ib_flow += ((union ib_flow_spec *)ib_flow)->size; } spec->match_criteria_enable = get_match_criteria_enable(spec->match_criteria); action = dst ? MLX5_FLOW_CONTEXT_ACTION_FWD_DEST : MLX5_FLOW_CONTEXT_ACTION_FWD_NEXT_PRIO; handler->rule = mlx5_add_flow_rule(ft, spec->match_criteria_enable, spec->match_criteria, spec->match_value, action, MLX5_FS_DEFAULT_FLOW_TAG, dst); if (IS_ERR(handler->rule)) { err = PTR_ERR(handler->rule); goto free; } ft_prio->refcount++; handler->prio = ft_prio; ft_prio->flow_table = ft; free: if (err) kfree(handler); kvfree(spec); return err ? ERR_PTR(err) : handler; } static struct mlx5_ib_flow_handler *create_dont_trap_rule(struct mlx5_ib_dev *dev, struct mlx5_ib_flow_prio *ft_prio, struct ib_flow_attr *flow_attr, struct mlx5_flow_destination *dst) { struct mlx5_ib_flow_handler *handler_dst = NULL; struct mlx5_ib_flow_handler *handler = NULL; handler = create_flow_rule(dev, ft_prio, flow_attr, NULL); if (!IS_ERR(handler)) { handler_dst = create_flow_rule(dev, ft_prio, flow_attr, dst); if (IS_ERR(handler_dst)) { mlx5_del_flow_rule(handler->rule); ft_prio->refcount--; kfree(handler); handler = handler_dst; } else { list_add(&handler_dst->list, &handler->list); } } return handler; } enum { LEFTOVERS_MC, LEFTOVERS_UC, }; static struct mlx5_ib_flow_handler *create_leftovers_rule(struct mlx5_ib_dev *dev, struct mlx5_ib_flow_prio *ft_prio, struct ib_flow_attr *flow_attr, struct mlx5_flow_destination *dst) { struct mlx5_ib_flow_handler *handler_ucast = NULL; struct mlx5_ib_flow_handler *handler = NULL; static struct { struct ib_flow_attr flow_attr; struct ib_flow_spec_eth eth_flow; } leftovers_specs[] = { [LEFTOVERS_MC] = { .flow_attr = { .num_of_specs = 1, .size = sizeof(leftovers_specs[0]) }, .eth_flow = { .type = IB_FLOW_SPEC_ETH, .size = sizeof(struct ib_flow_spec_eth), .mask = {.dst_mac = {0x1} }, .val = {.dst_mac = {0x1} } } }, [LEFTOVERS_UC] = { .flow_attr = { .num_of_specs = 1, .size = sizeof(leftovers_specs[0]) }, .eth_flow = { .type = IB_FLOW_SPEC_ETH, .size = sizeof(struct ib_flow_spec_eth), .mask = {.dst_mac = {0x1} }, .val = {.dst_mac = {} } } } }; handler = create_flow_rule(dev, ft_prio, &leftovers_specs[LEFTOVERS_MC].flow_attr, dst); if (!IS_ERR(handler) && flow_attr->type == IB_FLOW_ATTR_ALL_DEFAULT) { handler_ucast = create_flow_rule(dev, ft_prio, &leftovers_specs[LEFTOVERS_UC].flow_attr, dst); if (IS_ERR(handler_ucast)) { mlx5_del_flow_rule(handler->rule); ft_prio->refcount--; kfree(handler); handler = handler_ucast; } else { list_add(&handler_ucast->list, &handler->list); } } return handler; } static struct mlx5_ib_flow_handler *create_sniffer_rule(struct mlx5_ib_dev *dev, struct mlx5_ib_flow_prio *ft_rx, struct mlx5_ib_flow_prio *ft_tx, struct mlx5_flow_destination *dst) { struct mlx5_ib_flow_handler *handler_rx; struct mlx5_ib_flow_handler *handler_tx; int err; static const struct ib_flow_attr flow_attr = { .num_of_specs = 0, .size = sizeof(flow_attr) }; handler_rx = create_flow_rule(dev, ft_rx, &flow_attr, dst); if (IS_ERR(handler_rx)) { err = PTR_ERR(handler_rx); goto err; } handler_tx = create_flow_rule(dev, ft_tx, &flow_attr, dst); if (IS_ERR(handler_tx)) { err = PTR_ERR(handler_tx); goto err_tx; } list_add(&handler_tx->list, &handler_rx->list); return handler_rx; err_tx: mlx5_del_flow_rule(handler_rx->rule); ft_rx->refcount--; kfree(handler_rx); err: return ERR_PTR(err); } static struct ib_flow *mlx5_ib_create_flow(struct ib_qp *qp, struct ib_flow_attr *flow_attr, int domain) { struct mlx5_ib_dev *dev = to_mdev(qp->device); struct mlx5_ib_qp *mqp = to_mqp(qp); struct mlx5_ib_flow_handler *handler = NULL; struct mlx5_flow_destination *dst = NULL; struct mlx5_ib_flow_prio *ft_prio_tx = NULL; struct mlx5_ib_flow_prio *ft_prio; int err; if (flow_attr->priority > MLX5_IB_FLOW_LAST_PRIO) return ERR_PTR(-ENOSPC); if (domain != IB_FLOW_DOMAIN_USER || flow_attr->port > MLX5_CAP_GEN(dev->mdev, num_ports) || (flow_attr->flags & ~IB_FLOW_ATTR_FLAGS_DONT_TRAP)) return ERR_PTR(-EINVAL); dst = kzalloc(sizeof(*dst), GFP_KERNEL); if (!dst) return ERR_PTR(-ENOMEM); mutex_lock(&dev->flow_db.lock); ft_prio = get_flow_table(dev, flow_attr, MLX5_IB_FT_RX); if (IS_ERR(ft_prio)) { err = PTR_ERR(ft_prio); goto unlock; } if (flow_attr->type == IB_FLOW_ATTR_SNIFFER) { ft_prio_tx = get_flow_table(dev, flow_attr, MLX5_IB_FT_TX); if (IS_ERR(ft_prio_tx)) { err = PTR_ERR(ft_prio_tx); ft_prio_tx = NULL; goto destroy_ft; } } dst->type = MLX5_FLOW_DESTINATION_TYPE_TIR; if (mqp->flags & MLX5_IB_QP_RSS) dst->tir_num = mqp->rss_qp.tirn; else dst->tir_num = mqp->raw_packet_qp.rq.tirn; if (flow_attr->type == IB_FLOW_ATTR_NORMAL) { if (flow_attr->flags & IB_FLOW_ATTR_FLAGS_DONT_TRAP) { handler = create_dont_trap_rule(dev, ft_prio, flow_attr, dst); } else { handler = create_flow_rule(dev, ft_prio, flow_attr, dst); } } else if (flow_attr->type == IB_FLOW_ATTR_ALL_DEFAULT || flow_attr->type == IB_FLOW_ATTR_MC_DEFAULT) { handler = create_leftovers_rule(dev, ft_prio, flow_attr, dst); } else if (flow_attr->type == IB_FLOW_ATTR_SNIFFER) { handler = create_sniffer_rule(dev, ft_prio, ft_prio_tx, dst); } else { err = -EINVAL; goto destroy_ft; } if (IS_ERR(handler)) { err = PTR_ERR(handler); handler = NULL; goto destroy_ft; } mutex_unlock(&dev->flow_db.lock); kfree(dst); return &handler->ibflow; destroy_ft: put_flow_table(dev, ft_prio, false); if (ft_prio_tx) put_flow_table(dev, ft_prio_tx, false); unlock: mutex_unlock(&dev->flow_db.lock); kfree(dst); kfree(handler); return ERR_PTR(err); } static int mlx5_ib_mcg_attach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid) { struct mlx5_ib_dev *dev = to_mdev(ibqp->device); int err; err = mlx5_core_attach_mcg(dev->mdev, gid, ibqp->qp_num); if (err) mlx5_ib_warn(dev, "failed attaching QPN 0x%x, MGID %pI6\n", ibqp->qp_num, gid->raw); return err; } static int mlx5_ib_mcg_detach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid) { struct mlx5_ib_dev *dev = to_mdev(ibqp->device); int err; err = mlx5_core_detach_mcg(dev->mdev, gid, ibqp->qp_num); if (err) mlx5_ib_warn(dev, "failed detaching QPN 0x%x, MGID %pI6\n", ibqp->qp_num, gid->raw); return err; } static int init_node_data(struct mlx5_ib_dev *dev) { int err; err = mlx5_query_node_desc(dev, dev->ib_dev.node_desc); if (err) return err; return mlx5_query_node_guid(dev, &dev->ib_dev.node_guid); } static ssize_t show_fw_pages(struct device *device, struct device_attribute *attr, char *buf) { struct mlx5_ib_dev *dev = container_of(device, struct mlx5_ib_dev, ib_dev.dev); return sprintf(buf, "%lld\n", (long long)dev->mdev->priv.fw_pages); } static ssize_t show_reg_pages(struct device *device, struct device_attribute *attr, char *buf) { struct mlx5_ib_dev *dev = container_of(device, struct mlx5_ib_dev, ib_dev.dev); return sprintf(buf, "%d\n", atomic_read(&dev->mdev->priv.reg_pages)); } static ssize_t show_hca(struct device *device, struct device_attribute *attr, char *buf) { struct mlx5_ib_dev *dev = container_of(device, struct mlx5_ib_dev, ib_dev.dev); return sprintf(buf, "MT%d\n", dev->mdev->pdev->device); } static ssize_t show_rev(struct device *device, struct device_attribute *attr, char *buf) { struct mlx5_ib_dev *dev = container_of(device, struct mlx5_ib_dev, ib_dev.dev); return sprintf(buf, "%x\n", dev->mdev->pdev->revision); } static ssize_t show_board(struct device *device, struct device_attribute *attr, char *buf) { struct mlx5_ib_dev *dev = container_of(device, struct mlx5_ib_dev, ib_dev.dev); return sprintf(buf, "%.*s\n", MLX5_BOARD_ID_LEN, dev->mdev->board_id); } static DEVICE_ATTR(hw_rev, S_IRUGO, show_rev, NULL); static DEVICE_ATTR(hca_type, S_IRUGO, show_hca, NULL); static DEVICE_ATTR(board_id, S_IRUGO, show_board, NULL); static DEVICE_ATTR(fw_pages, S_IRUGO, show_fw_pages, NULL); static DEVICE_ATTR(reg_pages, S_IRUGO, show_reg_pages, NULL); static struct device_attribute *mlx5_class_attributes[] = { &dev_attr_hw_rev, &dev_attr_hca_type, &dev_attr_board_id, &dev_attr_fw_pages, &dev_attr_reg_pages, }; static void pkey_change_handler(struct work_struct *work) { struct mlx5_ib_port_resources *ports = container_of(work, struct mlx5_ib_port_resources, pkey_change_work); mutex_lock(&ports->devr->mutex); mlx5_ib_gsi_pkey_change(ports->gsi); mutex_unlock(&ports->devr->mutex); } static void mlx5_ib_handle_internal_error(struct mlx5_ib_dev *ibdev) { struct mlx5_ib_qp *mqp; struct mlx5_ib_cq *send_mcq, *recv_mcq; struct mlx5_core_cq *mcq; struct list_head cq_armed_list; unsigned long flags_qp; unsigned long flags_cq; unsigned long flags; INIT_LIST_HEAD(&cq_armed_list); /* Go over qp list reside on that ibdev, sync with create/destroy qp.*/ spin_lock_irqsave(&ibdev->reset_flow_resource_lock, flags); list_for_each_entry(mqp, &ibdev->qp_list, qps_list) { spin_lock_irqsave(&mqp->sq.lock, flags_qp); if (mqp->sq.tail != mqp->sq.head) { send_mcq = to_mcq(mqp->ibqp.send_cq); spin_lock_irqsave(&send_mcq->lock, flags_cq); if (send_mcq->mcq.comp && mqp->ibqp.send_cq->comp_handler) { if (!send_mcq->mcq.reset_notify_added) { send_mcq->mcq.reset_notify_added = 1; list_add_tail(&send_mcq->mcq.reset_notify, &cq_armed_list); } } spin_unlock_irqrestore(&send_mcq->lock, flags_cq); } spin_unlock_irqrestore(&mqp->sq.lock, flags_qp); spin_lock_irqsave(&mqp->rq.lock, flags_qp); /* no handling is needed for SRQ */ if (!mqp->ibqp.srq) { if (mqp->rq.tail != mqp->rq.head) { recv_mcq = to_mcq(mqp->ibqp.recv_cq); spin_lock_irqsave(&recv_mcq->lock, flags_cq); if (recv_mcq->mcq.comp && mqp->ibqp.recv_cq->comp_handler) { if (!recv_mcq->mcq.reset_notify_added) { recv_mcq->mcq.reset_notify_added = 1; list_add_tail(&recv_mcq->mcq.reset_notify, &cq_armed_list); } } spin_unlock_irqrestore(&recv_mcq->lock, flags_cq); } } spin_unlock_irqrestore(&mqp->rq.lock, flags_qp); } /*At that point all inflight post send were put to be executed as of we * lock/unlock above locks Now need to arm all involved CQs. */ list_for_each_entry(mcq, &cq_armed_list, reset_notify) { mcq->comp(mcq); } spin_unlock_irqrestore(&ibdev->reset_flow_resource_lock, flags); } static void mlx5_ib_event(struct mlx5_core_dev *dev, void *context, enum mlx5_dev_event event, unsigned long param) { struct mlx5_ib_dev *ibdev = (struct mlx5_ib_dev *)context; struct ib_event ibev; bool fatal = false; u8 port = 0; switch (event) { case MLX5_DEV_EVENT_SYS_ERROR: ibev.event = IB_EVENT_DEVICE_FATAL; mlx5_ib_handle_internal_error(ibdev); fatal = true; break; case MLX5_DEV_EVENT_PORT_UP: case MLX5_DEV_EVENT_PORT_DOWN: case MLX5_DEV_EVENT_PORT_INITIALIZED: port = (u8)param; /* In RoCE, port up/down events are handled in * mlx5_netdev_event(). */ if (mlx5_ib_port_link_layer(&ibdev->ib_dev, port) == IB_LINK_LAYER_ETHERNET) return; ibev.event = (event == MLX5_DEV_EVENT_PORT_UP) ? IB_EVENT_PORT_ACTIVE : IB_EVENT_PORT_ERR; break; case MLX5_DEV_EVENT_LID_CHANGE: ibev.event = IB_EVENT_LID_CHANGE; port = (u8)param; break; case MLX5_DEV_EVENT_PKEY_CHANGE: ibev.event = IB_EVENT_PKEY_CHANGE; port = (u8)param; schedule_work(&ibdev->devr.ports[port - 1].pkey_change_work); break; case MLX5_DEV_EVENT_GUID_CHANGE: ibev.event = IB_EVENT_GID_CHANGE; port = (u8)param; break; case MLX5_DEV_EVENT_CLIENT_REREG: ibev.event = IB_EVENT_CLIENT_REREGISTER; port = (u8)param; break; default: break; } ibev.device = &ibdev->ib_dev; ibev.element.port_num = port; if (port < 1 || port > ibdev->num_ports) { mlx5_ib_warn(ibdev, "warning: event on port %d\n", port); return; } if (ibdev->ib_active) ib_dispatch_event(&ibev); if (fatal) ibdev->ib_active = false; } static void get_ext_port_caps(struct mlx5_ib_dev *dev) { int port; for (port = 1; port <= MLX5_CAP_GEN(dev->mdev, num_ports); port++) mlx5_query_ext_port_caps(dev, port); } static int get_port_caps(struct mlx5_ib_dev *dev) { struct ib_device_attr *dprops = NULL; struct ib_port_attr *pprops = NULL; int err = -ENOMEM; int port; struct ib_udata uhw = {.inlen = 0, .outlen = 0}; pprops = kmalloc(sizeof(*pprops), GFP_KERNEL); if (!pprops) goto out; dprops = kmalloc(sizeof(*dprops), GFP_KERNEL); if (!dprops) goto out; err = mlx5_ib_query_device(&dev->ib_dev, dprops, &uhw); if (err) { mlx5_ib_warn(dev, "query_device failed %d\n", err); goto out; } for (port = 1; port <= MLX5_CAP_GEN(dev->mdev, num_ports); port++) { err = mlx5_ib_query_port(&dev->ib_dev, port, pprops); if (err) { mlx5_ib_warn(dev, "query_port %d failed %d\n", port, err); break; } dev->mdev->port_caps[port - 1].pkey_table_len = dprops->max_pkeys; dev->mdev->port_caps[port - 1].gid_table_len = pprops->gid_tbl_len; mlx5_ib_dbg(dev, "pkey_table_len %d, gid_table_len %d\n", dprops->max_pkeys, pprops->gid_tbl_len); } out: kfree(pprops); kfree(dprops); return err; } static void destroy_umrc_res(struct mlx5_ib_dev *dev) { int err; err = mlx5_mr_cache_cleanup(dev); if (err) mlx5_ib_warn(dev, "mr cache cleanup failed\n"); mlx5_ib_destroy_qp(dev->umrc.qp); ib_free_cq(dev->umrc.cq); ib_dealloc_pd(dev->umrc.pd); } enum { MAX_UMR_WR = 128, }; static int create_umr_res(struct mlx5_ib_dev *dev) { struct ib_qp_init_attr *init_attr = NULL; struct ib_qp_attr *attr = NULL; struct ib_pd *pd; struct ib_cq *cq; struct ib_qp *qp; int ret; attr = kzalloc(sizeof(*attr), GFP_KERNEL); init_attr = kzalloc(sizeof(*init_attr), GFP_KERNEL); if (!attr || !init_attr) { ret = -ENOMEM; goto error_0; } pd = ib_alloc_pd(&dev->ib_dev, 0); if (IS_ERR(pd)) { mlx5_ib_dbg(dev, "Couldn't create PD for sync UMR QP\n"); ret = PTR_ERR(pd); goto error_0; } cq = ib_alloc_cq(&dev->ib_dev, NULL, 128, 0, IB_POLL_SOFTIRQ); if (IS_ERR(cq)) { mlx5_ib_dbg(dev, "Couldn't create CQ for sync UMR QP\n"); ret = PTR_ERR(cq); goto error_2; } init_attr->send_cq = cq; init_attr->recv_cq = cq; init_attr->sq_sig_type = IB_SIGNAL_ALL_WR; init_attr->cap.max_send_wr = MAX_UMR_WR; init_attr->cap.max_send_sge = 1; init_attr->qp_type = MLX5_IB_QPT_REG_UMR; init_attr->port_num = 1; qp = mlx5_ib_create_qp(pd, init_attr, NULL); if (IS_ERR(qp)) { mlx5_ib_dbg(dev, "Couldn't create sync UMR QP\n"); ret = PTR_ERR(qp); goto error_3; } qp->device = &dev->ib_dev; qp->real_qp = qp; qp->uobject = NULL; qp->qp_type = MLX5_IB_QPT_REG_UMR; attr->qp_state = IB_QPS_INIT; attr->port_num = 1; ret = mlx5_ib_modify_qp(qp, attr, IB_QP_STATE | IB_QP_PKEY_INDEX | IB_QP_PORT, NULL); if (ret) { mlx5_ib_dbg(dev, "Couldn't modify UMR QP\n"); goto error_4; } memset(attr, 0, sizeof(*attr)); attr->qp_state = IB_QPS_RTR; attr->path_mtu = IB_MTU_256; ret = mlx5_ib_modify_qp(qp, attr, IB_QP_STATE, NULL); if (ret) { mlx5_ib_dbg(dev, "Couldn't modify umr QP to rtr\n"); goto error_4; } memset(attr, 0, sizeof(*attr)); attr->qp_state = IB_QPS_RTS; ret = mlx5_ib_modify_qp(qp, attr, IB_QP_STATE, NULL); if (ret) { mlx5_ib_dbg(dev, "Couldn't modify umr QP to rts\n"); goto error_4; } dev->umrc.qp = qp; dev->umrc.cq = cq; dev->umrc.pd = pd; sema_init(&dev->umrc.sem, MAX_UMR_WR); ret = mlx5_mr_cache_init(dev); if (ret) { mlx5_ib_warn(dev, "mr cache init failed %d\n", ret); goto error_4; } kfree(attr); kfree(init_attr); return 0; error_4: mlx5_ib_destroy_qp(qp); error_3: ib_free_cq(cq); error_2: ib_dealloc_pd(pd); error_0: kfree(attr); kfree(init_attr); return ret; } static int create_dev_resources(struct mlx5_ib_resources *devr) { struct ib_srq_init_attr attr; struct mlx5_ib_dev *dev; struct ib_cq_init_attr cq_attr = {.cqe = 1}; int port; int ret = 0; dev = container_of(devr, struct mlx5_ib_dev, devr); mutex_init(&devr->mutex); devr->p0 = mlx5_ib_alloc_pd(&dev->ib_dev, NULL, NULL); if (IS_ERR(devr->p0)) { ret = PTR_ERR(devr->p0); goto error0; } devr->p0->device = &dev->ib_dev; devr->p0->uobject = NULL; atomic_set(&devr->p0->usecnt, 0); devr->c0 = mlx5_ib_create_cq(&dev->ib_dev, &cq_attr, NULL, NULL); if (IS_ERR(devr->c0)) { ret = PTR_ERR(devr->c0); goto error1; } devr->c0->device = &dev->ib_dev; devr->c0->uobject = NULL; devr->c0->comp_handler = NULL; devr->c0->event_handler = NULL; devr->c0->cq_context = NULL; atomic_set(&devr->c0->usecnt, 0); devr->x0 = mlx5_ib_alloc_xrcd(&dev->ib_dev, NULL, NULL); if (IS_ERR(devr->x0)) { ret = PTR_ERR(devr->x0); goto error2; } devr->x0->device = &dev->ib_dev; devr->x0->inode = NULL; atomic_set(&devr->x0->usecnt, 0); mutex_init(&devr->x0->tgt_qp_mutex); INIT_LIST_HEAD(&devr->x0->tgt_qp_list); devr->x1 = mlx5_ib_alloc_xrcd(&dev->ib_dev, NULL, NULL); if (IS_ERR(devr->x1)) { ret = PTR_ERR(devr->x1); goto error3; } devr->x1->device = &dev->ib_dev; devr->x1->inode = NULL; atomic_set(&devr->x1->usecnt, 0); mutex_init(&devr->x1->tgt_qp_mutex); INIT_LIST_HEAD(&devr->x1->tgt_qp_list); memset(&attr, 0, sizeof(attr)); attr.attr.max_sge = 1; attr.attr.max_wr = 1; attr.srq_type = IB_SRQT_XRC; attr.ext.xrc.cq = devr->c0; attr.ext.xrc.xrcd = devr->x0; devr->s0 = mlx5_ib_create_srq(devr->p0, &attr, NULL); if (IS_ERR(devr->s0)) { ret = PTR_ERR(devr->s0); goto error4; } devr->s0->device = &dev->ib_dev; devr->s0->pd = devr->p0; devr->s0->uobject = NULL; devr->s0->event_handler = NULL; devr->s0->srq_context = NULL; devr->s0->srq_type = IB_SRQT_XRC; devr->s0->ext.xrc.xrcd = devr->x0; devr->s0->ext.xrc.cq = devr->c0; atomic_inc(&devr->s0->ext.xrc.xrcd->usecnt); atomic_inc(&devr->s0->ext.xrc.cq->usecnt); atomic_inc(&devr->p0->usecnt); atomic_set(&devr->s0->usecnt, 0); memset(&attr, 0, sizeof(attr)); attr.attr.max_sge = 1; attr.attr.max_wr = 1; attr.srq_type = IB_SRQT_BASIC; devr->s1 = mlx5_ib_create_srq(devr->p0, &attr, NULL); if (IS_ERR(devr->s1)) { ret = PTR_ERR(devr->s1); goto error5; } devr->s1->device = &dev->ib_dev; devr->s1->pd = devr->p0; devr->s1->uobject = NULL; devr->s1->event_handler = NULL; devr->s1->srq_context = NULL; devr->s1->srq_type = IB_SRQT_BASIC; devr->s1->ext.xrc.cq = devr->c0; atomic_inc(&devr->p0->usecnt); atomic_set(&devr->s0->usecnt, 0); for (port = 0; port < ARRAY_SIZE(devr->ports); ++port) { INIT_WORK(&devr->ports[port].pkey_change_work, pkey_change_handler); devr->ports[port].devr = devr; } return 0; error5: mlx5_ib_destroy_srq(devr->s0); error4: mlx5_ib_dealloc_xrcd(devr->x1); error3: mlx5_ib_dealloc_xrcd(devr->x0); error2: mlx5_ib_destroy_cq(devr->c0); error1: mlx5_ib_dealloc_pd(devr->p0); error0: return ret; } static void destroy_dev_resources(struct mlx5_ib_resources *devr) { struct mlx5_ib_dev *dev = container_of(devr, struct mlx5_ib_dev, devr); int port; mlx5_ib_destroy_srq(devr->s1); mlx5_ib_destroy_srq(devr->s0); mlx5_ib_dealloc_xrcd(devr->x0); mlx5_ib_dealloc_xrcd(devr->x1); mlx5_ib_destroy_cq(devr->c0); mlx5_ib_dealloc_pd(devr->p0); /* Make sure no change P_Key work items are still executing */ for (port = 0; port < dev->num_ports; ++port) cancel_work_sync(&devr->ports[port].pkey_change_work); } static u32 get_core_cap_flags(struct ib_device *ibdev) { struct mlx5_ib_dev *dev = to_mdev(ibdev); enum rdma_link_layer ll = mlx5_ib_port_link_layer(ibdev, 1); u8 l3_type_cap = MLX5_CAP_ROCE(dev->mdev, l3_type); u8 roce_version_cap = MLX5_CAP_ROCE(dev->mdev, roce_version); u32 ret = 0; if (ll == IB_LINK_LAYER_INFINIBAND) return RDMA_CORE_PORT_IBA_IB; if (!(l3_type_cap & MLX5_ROCE_L3_TYPE_IPV4_CAP)) return 0; if (!(l3_type_cap & MLX5_ROCE_L3_TYPE_IPV6_CAP)) return 0; if (roce_version_cap & MLX5_ROCE_VERSION_1_CAP) ret |= RDMA_CORE_PORT_IBA_ROCE; if (roce_version_cap & MLX5_ROCE_VERSION_2_CAP) ret |= RDMA_CORE_PORT_IBA_ROCE_UDP_ENCAP; return ret; } static int mlx5_port_immutable(struct ib_device *ibdev, u8 port_num, struct ib_port_immutable *immutable) { struct ib_port_attr attr; int err; err = mlx5_ib_query_port(ibdev, port_num, &attr); if (err) return err; immutable->pkey_tbl_len = attr.pkey_tbl_len; immutable->gid_tbl_len = attr.gid_tbl_len; immutable->core_cap_flags = get_core_cap_flags(ibdev); immutable->max_mad_size = IB_MGMT_MAD_SIZE; return 0; } static void get_dev_fw_str(struct ib_device *ibdev, char *str, size_t str_len) { struct mlx5_ib_dev *dev = container_of(ibdev, struct mlx5_ib_dev, ib_dev); snprintf(str, str_len, "%d.%d.%04d", fw_rev_maj(dev->mdev), fw_rev_min(dev->mdev), fw_rev_sub(dev->mdev)); } static int mlx5_roce_lag_init(struct mlx5_ib_dev *dev) { return 0; } static void mlx5_roce_lag_cleanup(struct mlx5_ib_dev *dev) { } static void mlx5_remove_roce_notifier(struct mlx5_ib_dev *dev) { if (dev->roce.nb.notifier_call) { unregister_netdevice_notifier(&dev->roce.nb); dev->roce.nb.notifier_call = NULL; } } static int mlx5_enable_roce(struct mlx5_ib_dev *dev) { VNET_ITERATOR_DECL(vnet_iter); struct net_device *idev; int err; /* Check if mlx5en net device already exists */ VNET_LIST_RLOCK(); VNET_FOREACH(vnet_iter) { IFNET_RLOCK(); CURVNET_SET_QUIET(vnet_iter); TAILQ_FOREACH(idev, &V_ifnet, if_link) { /* check if network interface belongs to mlx5en */ if (!mlx5_netdev_match(idev, dev->mdev, "mce")) continue; write_lock(&dev->roce.netdev_lock); dev->roce.netdev = idev; write_unlock(&dev->roce.netdev_lock); } CURVNET_RESTORE(); IFNET_RUNLOCK(); } VNET_LIST_RUNLOCK(); dev->roce.nb.notifier_call = mlx5_netdev_event; err = register_netdevice_notifier(&dev->roce.nb); if (err) { dev->roce.nb.notifier_call = NULL; return err; } err = mlx5_nic_vport_enable_roce(dev->mdev); if (err) goto err_unregister_netdevice_notifier; err = mlx5_roce_lag_init(dev); if (err) goto err_disable_roce; return 0; err_disable_roce: mlx5_nic_vport_disable_roce(dev->mdev); err_unregister_netdevice_notifier: mlx5_remove_roce_notifier(dev); return err; } static void mlx5_disable_roce(struct mlx5_ib_dev *dev) { mlx5_roce_lag_cleanup(dev); mlx5_nic_vport_disable_roce(dev->mdev); } static void mlx5_ib_dealloc_q_port_counter(struct mlx5_ib_dev *dev, u8 port_num) { mlx5_vport_dealloc_q_counter(dev->mdev, MLX5_INTERFACE_PROTOCOL_IB, dev->port[port_num].q_cnt_id); dev->port[port_num].q_cnt_id = 0; } static void mlx5_ib_dealloc_q_counters(struct mlx5_ib_dev *dev) { unsigned int i; for (i = 0; i < dev->num_ports; i++) mlx5_ib_dealloc_q_port_counter(dev, i); } static int mlx5_ib_alloc_q_counters(struct mlx5_ib_dev *dev) { int i; int ret; for (i = 0; i < dev->num_ports; i++) { ret = mlx5_vport_alloc_q_counter(dev->mdev, MLX5_INTERFACE_PROTOCOL_IB, &dev->port[i].q_cnt_id); if (ret) { mlx5_ib_warn(dev, "couldn't allocate queue counter for port %d, err %d\n", i + 1, ret); goto dealloc_counters; } } return 0; dealloc_counters: while (--i >= 0) mlx5_ib_dealloc_q_port_counter(dev, i); return ret; } static const char * const names[] = { "rx_write_requests", "rx_read_requests", "rx_atomic_requests", "out_of_buffer", "out_of_sequence", "duplicate_request", "rnr_nak_retry_err", "packet_seq_err", "implied_nak_seq_err", "local_ack_timeout_err", }; static const size_t stats_offsets[] = { MLX5_BYTE_OFF(query_q_counter_out, rx_write_requests), MLX5_BYTE_OFF(query_q_counter_out, rx_read_requests), MLX5_BYTE_OFF(query_q_counter_out, rx_atomic_requests), MLX5_BYTE_OFF(query_q_counter_out, out_of_buffer), MLX5_BYTE_OFF(query_q_counter_out, out_of_sequence), MLX5_BYTE_OFF(query_q_counter_out, duplicate_request), MLX5_BYTE_OFF(query_q_counter_out, rnr_nak_retry_err), MLX5_BYTE_OFF(query_q_counter_out, packet_seq_err), MLX5_BYTE_OFF(query_q_counter_out, implied_nak_seq_err), MLX5_BYTE_OFF(query_q_counter_out, local_ack_timeout_err), }; static struct rdma_hw_stats *mlx5_ib_alloc_hw_stats(struct ib_device *ibdev, u8 port_num) { BUILD_BUG_ON(ARRAY_SIZE(names) != ARRAY_SIZE(stats_offsets)); /* We support only per port stats */ if (port_num == 0) return NULL; return rdma_alloc_hw_stats_struct(names, ARRAY_SIZE(names), RDMA_HW_STATS_DEFAULT_LIFESPAN); } static int mlx5_ib_get_hw_stats(struct ib_device *ibdev, struct rdma_hw_stats *stats, u8 port, int index) { struct mlx5_ib_dev *dev = to_mdev(ibdev); int outlen = MLX5_ST_SZ_BYTES(query_q_counter_out); void *out; __be32 val; int ret; int i; if (!port || !stats) return -ENOSYS; out = mlx5_vzalloc(outlen); if (!out) return -ENOMEM; ret = mlx5_vport_query_q_counter(dev->mdev, dev->port[port - 1].q_cnt_id, 0, out, outlen); if (ret) goto free; for (i = 0; i < ARRAY_SIZE(names); i++) { val = *(__be32 *)(out + stats_offsets[i]); stats->value[i] = (u64)be32_to_cpu(val); } free: kvfree(out); return ARRAY_SIZE(names); } static void *mlx5_ib_add(struct mlx5_core_dev *mdev) { struct mlx5_ib_dev *dev; enum rdma_link_layer ll; int port_type_cap; const char *name; int err; int i; port_type_cap = MLX5_CAP_GEN(mdev, port_type); ll = mlx5_port_type_cap_to_rdma_ll(port_type_cap); if ((ll == IB_LINK_LAYER_ETHERNET) && !MLX5_CAP_GEN(mdev, roce)) return NULL; printk_once(KERN_INFO "%s", mlx5_version); dev = (struct mlx5_ib_dev *)ib_alloc_device(sizeof(*dev)); if (!dev) return NULL; dev->mdev = mdev; dev->port = kcalloc(MLX5_CAP_GEN(mdev, num_ports), sizeof(*dev->port), GFP_KERNEL); if (!dev->port) goto err_dealloc; rwlock_init(&dev->roce.netdev_lock); err = get_port_caps(dev); if (err) goto err_free_port; if (mlx5_use_mad_ifc(dev)) get_ext_port_caps(dev); MLX5_INIT_DOORBELL_LOCK(&dev->uar_lock); name = "mlx5_%d"; strlcpy(dev->ib_dev.name, name, IB_DEVICE_NAME_MAX); dev->ib_dev.owner = THIS_MODULE; dev->ib_dev.node_type = RDMA_NODE_IB_CA; dev->ib_dev.local_dma_lkey = 0 /* not supported for now */; dev->num_ports = MLX5_CAP_GEN(mdev, num_ports); dev->ib_dev.phys_port_cnt = dev->num_ports; dev->ib_dev.num_comp_vectors = dev->mdev->priv.eq_table.num_comp_vectors; dev->ib_dev.dma_device = &mdev->pdev->dev; dev->ib_dev.uverbs_abi_ver = MLX5_IB_UVERBS_ABI_VERSION; dev->ib_dev.uverbs_cmd_mask = (1ull << IB_USER_VERBS_CMD_GET_CONTEXT) | (1ull << IB_USER_VERBS_CMD_QUERY_DEVICE) | (1ull << IB_USER_VERBS_CMD_QUERY_PORT) | (1ull << IB_USER_VERBS_CMD_ALLOC_PD) | (1ull << IB_USER_VERBS_CMD_DEALLOC_PD) | (1ull << IB_USER_VERBS_CMD_REG_MR) | (1ull << IB_USER_VERBS_CMD_REREG_MR) | (1ull << IB_USER_VERBS_CMD_DEREG_MR) | (1ull << IB_USER_VERBS_CMD_CREATE_COMP_CHANNEL) | (1ull << IB_USER_VERBS_CMD_CREATE_CQ) | (1ull << IB_USER_VERBS_CMD_RESIZE_CQ) | (1ull << IB_USER_VERBS_CMD_DESTROY_CQ) | (1ull << IB_USER_VERBS_CMD_CREATE_QP) | (1ull << IB_USER_VERBS_CMD_MODIFY_QP) | (1ull << IB_USER_VERBS_CMD_QUERY_QP) | (1ull << IB_USER_VERBS_CMD_DESTROY_QP) | (1ull << IB_USER_VERBS_CMD_ATTACH_MCAST) | (1ull << IB_USER_VERBS_CMD_DETACH_MCAST) | (1ull << IB_USER_VERBS_CMD_CREATE_SRQ) | (1ull << IB_USER_VERBS_CMD_MODIFY_SRQ) | (1ull << IB_USER_VERBS_CMD_QUERY_SRQ) | (1ull << IB_USER_VERBS_CMD_DESTROY_SRQ) | (1ull << IB_USER_VERBS_CMD_CREATE_XSRQ) | (1ull << IB_USER_VERBS_CMD_OPEN_QP); dev->ib_dev.uverbs_ex_cmd_mask = (1ull << IB_USER_VERBS_EX_CMD_QUERY_DEVICE) | (1ull << IB_USER_VERBS_EX_CMD_CREATE_CQ) | (1ull << IB_USER_VERBS_EX_CMD_CREATE_QP); dev->ib_dev.query_device = mlx5_ib_query_device; dev->ib_dev.query_port = mlx5_ib_query_port; dev->ib_dev.get_link_layer = mlx5_ib_port_link_layer; if (ll == IB_LINK_LAYER_ETHERNET) dev->ib_dev.get_netdev = mlx5_ib_get_netdev; dev->ib_dev.query_gid = mlx5_ib_query_gid; dev->ib_dev.add_gid = mlx5_ib_add_gid; dev->ib_dev.del_gid = mlx5_ib_del_gid; dev->ib_dev.query_pkey = mlx5_ib_query_pkey; dev->ib_dev.modify_device = mlx5_ib_modify_device; dev->ib_dev.modify_port = mlx5_ib_modify_port; dev->ib_dev.alloc_ucontext = mlx5_ib_alloc_ucontext; dev->ib_dev.dealloc_ucontext = mlx5_ib_dealloc_ucontext; dev->ib_dev.mmap = mlx5_ib_mmap; dev->ib_dev.alloc_pd = mlx5_ib_alloc_pd; dev->ib_dev.dealloc_pd = mlx5_ib_dealloc_pd; dev->ib_dev.create_ah = mlx5_ib_create_ah; dev->ib_dev.query_ah = mlx5_ib_query_ah; dev->ib_dev.destroy_ah = mlx5_ib_destroy_ah; dev->ib_dev.create_srq = mlx5_ib_create_srq; dev->ib_dev.modify_srq = mlx5_ib_modify_srq; dev->ib_dev.query_srq = mlx5_ib_query_srq; dev->ib_dev.destroy_srq = mlx5_ib_destroy_srq; dev->ib_dev.post_srq_recv = mlx5_ib_post_srq_recv; dev->ib_dev.create_qp = mlx5_ib_create_qp; dev->ib_dev.modify_qp = mlx5_ib_modify_qp; dev->ib_dev.query_qp = mlx5_ib_query_qp; dev->ib_dev.destroy_qp = mlx5_ib_destroy_qp; dev->ib_dev.post_send = mlx5_ib_post_send; dev->ib_dev.post_recv = mlx5_ib_post_recv; dev->ib_dev.create_cq = mlx5_ib_create_cq; dev->ib_dev.modify_cq = mlx5_ib_modify_cq; dev->ib_dev.resize_cq = mlx5_ib_resize_cq; dev->ib_dev.destroy_cq = mlx5_ib_destroy_cq; dev->ib_dev.poll_cq = mlx5_ib_poll_cq; dev->ib_dev.req_notify_cq = mlx5_ib_arm_cq; dev->ib_dev.get_dma_mr = mlx5_ib_get_dma_mr; dev->ib_dev.reg_user_mr = mlx5_ib_reg_user_mr; dev->ib_dev.rereg_user_mr = mlx5_ib_rereg_user_mr; + dev->ib_dev.reg_phys_mr = mlx5_ib_reg_phys_mr; dev->ib_dev.dereg_mr = mlx5_ib_dereg_mr; dev->ib_dev.attach_mcast = mlx5_ib_mcg_attach; dev->ib_dev.detach_mcast = mlx5_ib_mcg_detach; dev->ib_dev.process_mad = mlx5_ib_process_mad; dev->ib_dev.alloc_mr = mlx5_ib_alloc_mr; dev->ib_dev.map_mr_sg = mlx5_ib_map_mr_sg; dev->ib_dev.check_mr_status = mlx5_ib_check_mr_status; dev->ib_dev.get_port_immutable = mlx5_port_immutable; dev->ib_dev.get_dev_fw_str = get_dev_fw_str; if (mlx5_core_is_pf(mdev)) { dev->ib_dev.get_vf_config = mlx5_ib_get_vf_config; dev->ib_dev.set_vf_link_state = mlx5_ib_set_vf_link_state; dev->ib_dev.get_vf_stats = mlx5_ib_get_vf_stats; dev->ib_dev.set_vf_guid = mlx5_ib_set_vf_guid; } dev->ib_dev.disassociate_ucontext = mlx5_ib_disassociate_ucontext; mlx5_ib_internal_fill_odp_caps(dev); if (MLX5_CAP_GEN(mdev, imaicl)) { dev->ib_dev.alloc_mw = mlx5_ib_alloc_mw; dev->ib_dev.dealloc_mw = mlx5_ib_dealloc_mw; dev->ib_dev.uverbs_cmd_mask |= (1ull << IB_USER_VERBS_CMD_ALLOC_MW) | (1ull << IB_USER_VERBS_CMD_DEALLOC_MW); } if (MLX5_CAP_GEN(dev->mdev, out_of_seq_cnt) && MLX5_CAP_GEN(dev->mdev, retransmission_q_counters)) { dev->ib_dev.get_hw_stats = mlx5_ib_get_hw_stats; dev->ib_dev.alloc_hw_stats = mlx5_ib_alloc_hw_stats; } if (MLX5_CAP_GEN(mdev, xrc)) { dev->ib_dev.alloc_xrcd = mlx5_ib_alloc_xrcd; dev->ib_dev.dealloc_xrcd = mlx5_ib_dealloc_xrcd; dev->ib_dev.uverbs_cmd_mask |= (1ull << IB_USER_VERBS_CMD_OPEN_XRCD) | (1ull << IB_USER_VERBS_CMD_CLOSE_XRCD); } if (mlx5_ib_port_link_layer(&dev->ib_dev, 1) == IB_LINK_LAYER_ETHERNET) { dev->ib_dev.create_flow = mlx5_ib_create_flow; dev->ib_dev.destroy_flow = mlx5_ib_destroy_flow; dev->ib_dev.create_wq = mlx5_ib_create_wq; dev->ib_dev.modify_wq = mlx5_ib_modify_wq; dev->ib_dev.destroy_wq = mlx5_ib_destroy_wq; dev->ib_dev.create_rwq_ind_table = mlx5_ib_create_rwq_ind_table; dev->ib_dev.destroy_rwq_ind_table = mlx5_ib_destroy_rwq_ind_table; dev->ib_dev.uverbs_ex_cmd_mask |= (1ull << IB_USER_VERBS_EX_CMD_CREATE_FLOW) | (1ull << IB_USER_VERBS_EX_CMD_DESTROY_FLOW) | (1ull << IB_USER_VERBS_EX_CMD_CREATE_WQ) | (1ull << IB_USER_VERBS_EX_CMD_MODIFY_WQ) | (1ull << IB_USER_VERBS_EX_CMD_DESTROY_WQ) | (1ull << IB_USER_VERBS_EX_CMD_CREATE_RWQ_IND_TBL) | (1ull << IB_USER_VERBS_EX_CMD_DESTROY_RWQ_IND_TBL); } err = init_node_data(dev); if (err) goto err_free_port; mutex_init(&dev->flow_db.lock); mutex_init(&dev->cap_mask_mutex); INIT_LIST_HEAD(&dev->qp_list); spin_lock_init(&dev->reset_flow_resource_lock); if (ll == IB_LINK_LAYER_ETHERNET) { err = mlx5_enable_roce(dev); if (err) goto err_free_port; } err = create_dev_resources(&dev->devr); if (err) goto err_disable_roce; err = mlx5_ib_odp_init_one(dev); if (err) goto err_rsrc; err = mlx5_ib_alloc_q_counters(dev); if (err) goto err_odp; err = ib_register_device(&dev->ib_dev, NULL); if (err) goto err_q_cnt; err = create_umr_res(dev); if (err) goto err_dev; for (i = 0; i < ARRAY_SIZE(mlx5_class_attributes); i++) { err = device_create_file(&dev->ib_dev.dev, mlx5_class_attributes[i]); if (err) goto err_umrc; } dev->ib_active = true; return dev; err_umrc: destroy_umrc_res(dev); err_dev: ib_unregister_device(&dev->ib_dev); err_q_cnt: mlx5_ib_dealloc_q_counters(dev); err_odp: mlx5_ib_odp_remove_one(dev); err_rsrc: destroy_dev_resources(&dev->devr); err_disable_roce: if (ll == IB_LINK_LAYER_ETHERNET) { mlx5_disable_roce(dev); mlx5_remove_roce_notifier(dev); } err_free_port: kfree(dev->port); err_dealloc: ib_dealloc_device((struct ib_device *)dev); return NULL; } static void mlx5_ib_remove(struct mlx5_core_dev *mdev, void *context) { struct mlx5_ib_dev *dev = context; enum rdma_link_layer ll = mlx5_ib_port_link_layer(&dev->ib_dev, 1); mlx5_remove_roce_notifier(dev); ib_unregister_device(&dev->ib_dev); mlx5_ib_dealloc_q_counters(dev); destroy_umrc_res(dev); mlx5_ib_odp_remove_one(dev); destroy_dev_resources(&dev->devr); if (ll == IB_LINK_LAYER_ETHERNET) mlx5_disable_roce(dev); kfree(dev->port); ib_dealloc_device(&dev->ib_dev); } static struct mlx5_interface mlx5_ib_interface = { .add = mlx5_ib_add, .remove = mlx5_ib_remove, .event = mlx5_ib_event, .protocol = MLX5_INTERFACE_PROTOCOL_IB, }; static int __init mlx5_ib_init(void) { int err; if (deprecated_prof_sel != 2) pr_warn("prof_sel is deprecated for mlx5_ib, set it for mlx5_core\n"); err = mlx5_ib_odp_init(); if (err) return err; err = mlx5_register_interface(&mlx5_ib_interface); if (err) goto clean_odp; return err; clean_odp: mlx5_ib_odp_cleanup(); return err; } static void __exit mlx5_ib_cleanup(void) { mlx5_unregister_interface(&mlx5_ib_interface); mlx5_ib_odp_cleanup(); } module_init_order(mlx5_ib_init, SI_ORDER_THIRD); module_exit_order(mlx5_ib_cleanup, SI_ORDER_THIRD); Index: projects/bsd_rdma_4_9_stable_11/sys/dev/mlx5/mlx5_ib/mlx5_ib_mr.c =================================================================== --- projects/bsd_rdma_4_9_stable_11/sys/dev/mlx5/mlx5_ib/mlx5_ib_mr.c (revision 331011) +++ projects/bsd_rdma_4_9_stable_11/sys/dev/mlx5/mlx5_ib/mlx5_ib_mr.c (revision 331012) @@ -1,1664 +1,1787 @@ /*- * Copyright (c) 2013-2015, Mellanox Technologies, Ltd. All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * * THIS SOFTWARE IS PROVIDED BY AUTHOR AND CONTRIBUTORS `AS IS' AND * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL AUTHOR OR CONTRIBUTORS BE LIABLE * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF * SUCH DAMAGE. * * $FreeBSD$ */ #include #include #include #include #include #include #include #include "mlx5_ib.h" enum { MAX_PENDING_REG_MR = 8, }; #define MLX5_UMR_ALIGN 2048 #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING static __be64 mlx5_ib_update_mtt_emergency_buffer[ MLX5_UMR_MTT_MIN_CHUNK_SIZE/sizeof(__be64)] __aligned(MLX5_UMR_ALIGN); static DEFINE_MUTEX(mlx5_ib_update_mtt_emergency_buffer_mutex); #endif static int clean_mr(struct mlx5_ib_mr *mr); static int destroy_mkey(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr) { int err = mlx5_core_destroy_mkey(dev->mdev, &mr->mmkey); #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING /* Wait until all page fault handlers using the mr complete. */ synchronize_srcu(&dev->mr_srcu); #endif return err; } static int order2idx(struct mlx5_ib_dev *dev, int order) { struct mlx5_mr_cache *cache = &dev->cache; if (order < cache->ent[0].order) return 0; else return order - cache->ent[0].order; } static bool use_umr_mtt_update(struct mlx5_ib_mr *mr, u64 start, u64 length) { return ((u64)1 << mr->order) * MLX5_ADAPTER_PAGE_SIZE >= length + (start & (MLX5_ADAPTER_PAGE_SIZE - 1)); } #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING static void update_odp_mr(struct mlx5_ib_mr *mr) { if (mr->umem->odp_data) { /* * This barrier prevents the compiler from moving the * setting of umem->odp_data->private to point to our * MR, before reg_umr finished, to ensure that the MR * initialization have finished before starting to * handle invalidations. */ smp_wmb(); mr->umem->odp_data->private = mr; /* * Make sure we will see the new * umem->odp_data->private value in the invalidation * routines, before we can get page faults on the * MR. Page faults can happen once we put the MR in * the tree, below this line. Without the barrier, * there can be a fault handling and an invalidation * before umem->odp_data->private == mr is visible to * the invalidation handler. */ smp_wmb(); } } #endif static void reg_mr_callback(int status, void *context) { struct mlx5_ib_mr *mr = context; struct mlx5_ib_dev *dev = mr->dev; struct mlx5_mr_cache *cache = &dev->cache; int c = order2idx(dev, mr->order); struct mlx5_cache_ent *ent = &cache->ent[c]; u8 key; unsigned long flags; struct mlx5_mr_table *table = &dev->mdev->priv.mr_table; int err; spin_lock_irqsave(&ent->lock, flags); ent->pending--; spin_unlock_irqrestore(&ent->lock, flags); if (status) { mlx5_ib_warn(dev, "async reg mr failed. status %d\n", status); kfree(mr); dev->fill_delay = 1; mod_timer(&dev->delay_timer, jiffies + HZ); return; } spin_lock_irqsave(&dev->mdev->priv.mkey_lock, flags); key = dev->mdev->priv.mkey_key++; spin_unlock_irqrestore(&dev->mdev->priv.mkey_lock, flags); mr->mmkey.key = mlx5_idx_to_mkey(MLX5_GET(create_mkey_out, mr->out, mkey_index)) | key; cache->last_add = jiffies; spin_lock_irqsave(&ent->lock, flags); list_add_tail(&mr->list, &ent->head); ent->cur++; ent->size++; spin_unlock_irqrestore(&ent->lock, flags); spin_lock_irqsave(&table->lock, flags); err = radix_tree_insert(&table->tree, mlx5_mkey_to_idx(mr->mmkey.key), &mr->mmkey); if (err) pr_err("Error inserting to mkey tree. 0x%x\n", -err); spin_unlock_irqrestore(&table->lock, flags); } static int add_keys(struct mlx5_ib_dev *dev, int c, int num) { struct mlx5_mr_cache *cache = &dev->cache; struct mlx5_cache_ent *ent = &cache->ent[c]; int inlen = MLX5_ST_SZ_BYTES(create_mkey_in); struct mlx5_ib_mr *mr; int npages = 1 << ent->order; void *mkc; u32 *in; int err = 0; int i; in = kzalloc(inlen, GFP_KERNEL); if (!in) return -ENOMEM; mkc = MLX5_ADDR_OF(create_mkey_in, in, memory_key_mkey_entry); for (i = 0; i < num; i++) { if (ent->pending >= MAX_PENDING_REG_MR) { err = -EAGAIN; break; } mr = kzalloc(sizeof(*mr), GFP_KERNEL); if (!mr) { err = -ENOMEM; break; } mr->order = ent->order; mr->umred = 1; mr->dev = dev; MLX5_SET(mkc, mkc, free, 1); MLX5_SET(mkc, mkc, umr_en, 1); MLX5_SET(mkc, mkc, access_mode, MLX5_ACCESS_MODE_MTT); MLX5_SET(mkc, mkc, qpn, 0xffffff); MLX5_SET(mkc, mkc, translations_octword_size, (npages + 1) / 2); MLX5_SET(mkc, mkc, log_page_size, 12); spin_lock_irq(&ent->lock); ent->pending++; spin_unlock_irq(&ent->lock); err = mlx5_core_create_mkey(dev->mdev, &mr->mmkey, (struct mlx5_create_mkey_mbox_in *)in, inlen, reg_mr_callback, mr, (struct mlx5_create_mkey_mbox_out *)mr->out); if (err) { spin_lock_irq(&ent->lock); ent->pending--; spin_unlock_irq(&ent->lock); mlx5_ib_warn(dev, "create mkey failed %d\n", err); kfree(mr); break; } } kfree(in); return err; } static void remove_keys(struct mlx5_ib_dev *dev, int c, int num) { struct mlx5_mr_cache *cache = &dev->cache; struct mlx5_cache_ent *ent = &cache->ent[c]; struct mlx5_ib_mr *mr; int err; int i; for (i = 0; i < num; i++) { spin_lock_irq(&ent->lock); if (list_empty(&ent->head)) { spin_unlock_irq(&ent->lock); return; } mr = list_first_entry(&ent->head, struct mlx5_ib_mr, list); list_del(&mr->list); ent->cur--; ent->size--; spin_unlock_irq(&ent->lock); err = destroy_mkey(dev, mr); if (err) mlx5_ib_warn(dev, "failed destroy mkey\n"); else kfree(mr); } } static int someone_adding(struct mlx5_mr_cache *cache) { int i; for (i = 0; i < MAX_MR_CACHE_ENTRIES; i++) { if (cache->ent[i].cur < cache->ent[i].limit) return 1; } return 0; } static void __cache_work_func(struct mlx5_cache_ent *ent) { struct mlx5_ib_dev *dev = ent->dev; struct mlx5_mr_cache *cache = &dev->cache; int i = order2idx(dev, ent->order); int err; if (cache->stopped) return; ent = &dev->cache.ent[i]; if (ent->cur < 2 * ent->limit && !dev->fill_delay) { err = add_keys(dev, i, 1); if (ent->cur < 2 * ent->limit) { if (err == -EAGAIN) { mlx5_ib_dbg(dev, "returned eagain, order %d\n", i + 2); queue_delayed_work(cache->wq, &ent->dwork, msecs_to_jiffies(3)); } else if (err) { mlx5_ib_warn(dev, "command failed order %d, err %d\n", i + 2, err); queue_delayed_work(cache->wq, &ent->dwork, msecs_to_jiffies(1000)); } else { queue_work(cache->wq, &ent->work); } } } else if (ent->cur > 2 * ent->limit) { /* * The remove_keys() logic is performed as garbage collection * task. Such task is intended to be run when no other active * processes are running. * * The need_resched() will return TRUE if there are user tasks * to be activated in near future. * * In such case, we don't execute remove_keys() and postpone * the garbage collection work to try to run in next cycle, * in order to free CPU resources to other tasks. */ if (!need_resched() && !someone_adding(cache) && time_after(jiffies, cache->last_add + 300 * HZ)) { remove_keys(dev, i, 1); if (ent->cur > ent->limit) queue_work(cache->wq, &ent->work); } else { queue_delayed_work(cache->wq, &ent->dwork, 300 * HZ); } } } static void delayed_cache_work_func(struct work_struct *work) { struct mlx5_cache_ent *ent; ent = container_of(work, struct mlx5_cache_ent, dwork.work); __cache_work_func(ent); } static void cache_work_func(struct work_struct *work) { struct mlx5_cache_ent *ent; ent = container_of(work, struct mlx5_cache_ent, work); __cache_work_func(ent); } static struct mlx5_ib_mr *alloc_cached_mr(struct mlx5_ib_dev *dev, int order) { struct mlx5_mr_cache *cache = &dev->cache; struct mlx5_ib_mr *mr = NULL; struct mlx5_cache_ent *ent; int c; int i; c = order2idx(dev, order); if (c < 0 || c >= MAX_MR_CACHE_ENTRIES) { mlx5_ib_warn(dev, "order %d, cache index %d\n", order, c); return NULL; } for (i = c; i < MAX_MR_CACHE_ENTRIES; i++) { ent = &cache->ent[i]; mlx5_ib_dbg(dev, "order %d, cache index %d\n", ent->order, i); spin_lock_irq(&ent->lock); if (!list_empty(&ent->head)) { mr = list_first_entry(&ent->head, struct mlx5_ib_mr, list); list_del(&mr->list); ent->cur--; spin_unlock_irq(&ent->lock); if (ent->cur < ent->limit) queue_work(cache->wq, &ent->work); break; } spin_unlock_irq(&ent->lock); queue_work(cache->wq, &ent->work); } if (!mr) cache->ent[c].miss++; return mr; } static void free_cached_mr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr) { struct mlx5_mr_cache *cache = &dev->cache; struct mlx5_cache_ent *ent; int shrink = 0; int c; c = order2idx(dev, mr->order); if (c < 0 || c >= MAX_MR_CACHE_ENTRIES) { mlx5_ib_warn(dev, "order %d, cache index %d\n", mr->order, c); return; } ent = &cache->ent[c]; spin_lock_irq(&ent->lock); list_add_tail(&mr->list, &ent->head); ent->cur++; if (ent->cur > 2 * ent->limit) shrink = 1; spin_unlock_irq(&ent->lock); if (shrink) queue_work(cache->wq, &ent->work); } static void clean_keys(struct mlx5_ib_dev *dev, int c) { struct mlx5_mr_cache *cache = &dev->cache; struct mlx5_cache_ent *ent = &cache->ent[c]; struct mlx5_ib_mr *mr; int err; cancel_delayed_work(&ent->dwork); while (1) { spin_lock_irq(&ent->lock); if (list_empty(&ent->head)) { spin_unlock_irq(&ent->lock); return; } mr = list_first_entry(&ent->head, struct mlx5_ib_mr, list); list_del(&mr->list); ent->cur--; ent->size--; spin_unlock_irq(&ent->lock); err = destroy_mkey(dev, mr); if (err) mlx5_ib_warn(dev, "failed destroy mkey\n"); else kfree(mr); } } static void delay_time_func(unsigned long ctx) { struct mlx5_ib_dev *dev = (struct mlx5_ib_dev *)ctx; dev->fill_delay = 0; } int mlx5_mr_cache_init(struct mlx5_ib_dev *dev) { struct mlx5_mr_cache *cache = &dev->cache; struct mlx5_cache_ent *ent; int limit; int i; mutex_init(&dev->slow_path_mutex); cache->wq = alloc_ordered_workqueue("mkey_cache", WQ_MEM_RECLAIM); if (!cache->wq) { mlx5_ib_warn(dev, "failed to create work queue\n"); return -ENOMEM; } setup_timer(&dev->delay_timer, delay_time_func, (unsigned long)dev); for (i = 0; i < MAX_MR_CACHE_ENTRIES; i++) { INIT_LIST_HEAD(&cache->ent[i].head); spin_lock_init(&cache->ent[i].lock); ent = &cache->ent[i]; INIT_LIST_HEAD(&ent->head); spin_lock_init(&ent->lock); ent->order = i + 2; ent->dev = dev; if (dev->mdev->profile->mask & MLX5_PROF_MASK_MR_CACHE) limit = dev->mdev->profile->mr_cache[i].limit; else limit = 0; INIT_WORK(&ent->work, cache_work_func); INIT_DELAYED_WORK(&ent->dwork, delayed_cache_work_func); ent->limit = limit; queue_work(cache->wq, &ent->work); } return 0; } int mlx5_mr_cache_cleanup(struct mlx5_ib_dev *dev) { int i; dev->cache.stopped = 1; flush_workqueue(dev->cache.wq); for (i = 0; i < MAX_MR_CACHE_ENTRIES; i++) clean_keys(dev, i); destroy_workqueue(dev->cache.wq); del_timer_sync(&dev->delay_timer); return 0; } struct ib_mr *mlx5_ib_get_dma_mr(struct ib_pd *pd, int acc) { struct mlx5_ib_dev *dev = to_mdev(pd->device); int inlen = MLX5_ST_SZ_BYTES(create_mkey_in); struct mlx5_core_dev *mdev = dev->mdev; struct mlx5_ib_mr *mr; void *mkc; u32 *in; int err; mr = kzalloc(sizeof(*mr), GFP_KERNEL); if (!mr) return ERR_PTR(-ENOMEM); in = kzalloc(inlen, GFP_KERNEL); if (!in) { err = -ENOMEM; goto err_free; } mkc = MLX5_ADDR_OF(create_mkey_in, in, memory_key_mkey_entry); MLX5_SET(mkc, mkc, access_mode, MLX5_ACCESS_MODE_PA); MLX5_SET(mkc, mkc, a, !!(acc & IB_ACCESS_REMOTE_ATOMIC)); MLX5_SET(mkc, mkc, rw, !!(acc & IB_ACCESS_REMOTE_WRITE)); MLX5_SET(mkc, mkc, rr, !!(acc & IB_ACCESS_REMOTE_READ)); MLX5_SET(mkc, mkc, lw, !!(acc & IB_ACCESS_LOCAL_WRITE)); MLX5_SET(mkc, mkc, lr, 1); MLX5_SET(mkc, mkc, length64, 1); MLX5_SET(mkc, mkc, pd, to_mpd(pd)->pdn); MLX5_SET(mkc, mkc, qpn, 0xffffff); MLX5_SET64(mkc, mkc, start_addr, 0); err = mlx5_core_create_mkey(mdev, &mr->mmkey, (struct mlx5_create_mkey_mbox_in *)in, inlen, NULL, NULL, NULL); if (err) goto err_in; kfree(in); mr->ibmr.lkey = mr->mmkey.key; mr->ibmr.rkey = mr->mmkey.key; mr->umem = NULL; return &mr->ibmr; err_in: kfree(in); err_free: kfree(mr); return ERR_PTR(err); } static int get_octo_len(u64 addr, u64 len, int page_size) { u64 offset; int npages; offset = addr & (page_size - 1); npages = ALIGN(len + offset, page_size) >> ilog2(page_size); return (npages + 1) / 2; } static int use_umr(int order) { return order <= MLX5_MAX_UMR_SHIFT; } static int dma_map_mr_pas(struct mlx5_ib_dev *dev, struct ib_umem *umem, int npages, int page_shift, int *size, __be64 **mr_pas, dma_addr_t *dma) { __be64 *pas; struct device *ddev = dev->ib_dev.dma_device; /* * UMR copies MTTs in units of MLX5_UMR_MTT_ALIGNMENT bytes. * To avoid copying garbage after the pas array, we allocate * a little more. */ *size = ALIGN(sizeof(u64) * npages, MLX5_UMR_MTT_ALIGNMENT); *mr_pas = kmalloc(*size + MLX5_UMR_ALIGN - 1, GFP_KERNEL); if (!(*mr_pas)) return -ENOMEM; pas = PTR_ALIGN(*mr_pas, MLX5_UMR_ALIGN); mlx5_ib_populate_pas(dev, umem, page_shift, pas, MLX5_IB_MTT_PRESENT); /* Clear padding after the actual pages. */ memset(pas + npages, 0, *size - npages * sizeof(u64)); *dma = dma_map_single(ddev, pas, *size, DMA_TO_DEVICE); if (dma_mapping_error(ddev, *dma)) { kfree(*mr_pas); return -ENOMEM; } return 0; } static void prep_umr_wqe_common(struct ib_pd *pd, struct ib_send_wr *wr, struct ib_sge *sg, u64 dma, int n, u32 key, int page_shift) { struct mlx5_ib_dev *dev = to_mdev(pd->device); struct mlx5_umr_wr *umrwr = umr_wr(wr); sg->addr = dma; sg->length = ALIGN(sizeof(u64) * n, 64); sg->lkey = dev->umrc.pd->local_dma_lkey; wr->next = NULL; wr->sg_list = sg; if (n) wr->num_sge = 1; else wr->num_sge = 0; wr->opcode = MLX5_IB_WR_UMR; umrwr->npages = n; umrwr->page_shift = page_shift; umrwr->mkey = key; } static void prep_umr_reg_wqe(struct ib_pd *pd, struct ib_send_wr *wr, struct ib_sge *sg, u64 dma, int n, u32 key, int page_shift, u64 virt_addr, u64 len, int access_flags) { struct mlx5_umr_wr *umrwr = umr_wr(wr); prep_umr_wqe_common(pd, wr, sg, dma, n, key, page_shift); wr->send_flags = 0; umrwr->target.virt_addr = virt_addr; umrwr->length = len; umrwr->access_flags = access_flags; umrwr->pd = pd; } static void prep_umr_unreg_wqe(struct mlx5_ib_dev *dev, struct ib_send_wr *wr, u32 key) { struct mlx5_umr_wr *umrwr = umr_wr(wr); wr->send_flags = MLX5_IB_SEND_UMR_UNREG | MLX5_IB_SEND_UMR_FAIL_IF_FREE; wr->opcode = MLX5_IB_WR_UMR; umrwr->mkey = key; } static struct ib_umem *mr_umem_get(struct ib_pd *pd, u64 start, u64 length, int access_flags, int *npages, int *page_shift, int *ncont, int *order) { struct mlx5_ib_dev *dev = to_mdev(pd->device); struct ib_umem *umem = ib_umem_get(pd->uobject->context, start, length, access_flags, 0); if (IS_ERR(umem)) { mlx5_ib_err(dev, "umem get failed (%ld)\n", PTR_ERR(umem)); return (void *)umem; } mlx5_ib_cont_pages(umem, start, npages, page_shift, ncont, order); if (!*npages) { mlx5_ib_warn(dev, "avoid zero region\n"); ib_umem_release(umem); return ERR_PTR(-EINVAL); } mlx5_ib_dbg(dev, "npages %d, ncont %d, order %d, page_shift %d\n", *npages, *ncont, *order, *page_shift); return umem; } static void mlx5_ib_umr_done(struct ib_cq *cq, struct ib_wc *wc) { struct mlx5_ib_umr_context *context = container_of(wc->wr_cqe, struct mlx5_ib_umr_context, cqe); context->status = wc->status; complete(&context->done); } static inline void mlx5_ib_init_umr_context(struct mlx5_ib_umr_context *context) { context->cqe.done = mlx5_ib_umr_done; context->status = -1; init_completion(&context->done); } static struct mlx5_ib_mr *reg_umr(struct ib_pd *pd, struct ib_umem *umem, u64 virt_addr, u64 len, int npages, int page_shift, int order, int access_flags) { struct mlx5_ib_dev *dev = to_mdev(pd->device); struct device *ddev = dev->ib_dev.dma_device; struct umr_common *umrc = &dev->umrc; struct mlx5_ib_umr_context umr_context; struct mlx5_umr_wr umrwr = {}; struct ib_send_wr *bad; struct mlx5_ib_mr *mr; struct ib_sge sg; int size; __be64 *mr_pas; dma_addr_t dma; int err = 0; int i; for (i = 0; i < 1; i++) { mr = alloc_cached_mr(dev, order); if (mr) break; err = add_keys(dev, order2idx(dev, order), 1); if (err && err != -EAGAIN) { mlx5_ib_warn(dev, "add_keys failed, err %d\n", err); break; } } if (!mr) return ERR_PTR(-EAGAIN); err = dma_map_mr_pas(dev, umem, npages, page_shift, &size, &mr_pas, &dma); if (err) goto free_mr; mlx5_ib_init_umr_context(&umr_context); umrwr.wr.wr_cqe = &umr_context.cqe; prep_umr_reg_wqe(pd, &umrwr.wr, &sg, dma, npages, mr->mmkey.key, page_shift, virt_addr, len, access_flags); down(&umrc->sem); err = ib_post_send(umrc->qp, &umrwr.wr, &bad); if (err) { mlx5_ib_warn(dev, "post send failed, err %d\n", err); goto unmap_dma; } else { wait_for_completion(&umr_context.done); if (umr_context.status != IB_WC_SUCCESS) { mlx5_ib_warn(dev, "reg umr failed\n"); err = -EFAULT; } } mr->mmkey.iova = virt_addr; mr->mmkey.size = len; mr->mmkey.pd = to_mpd(pd)->pdn; mr->live = 1; unmap_dma: up(&umrc->sem); dma_unmap_single(ddev, dma, size, DMA_TO_DEVICE); kfree(mr_pas); free_mr: if (err) { free_cached_mr(dev, mr); return ERR_PTR(err); } return mr; } #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING int mlx5_ib_update_mtt(struct mlx5_ib_mr *mr, u64 start_page_index, int npages, int zap) { struct mlx5_ib_dev *dev = mr->dev; struct device *ddev = dev->ib_dev.dma_device; struct umr_common *umrc = &dev->umrc; struct mlx5_ib_umr_context umr_context; struct ib_umem *umem = mr->umem; int size; __be64 *pas; dma_addr_t dma; struct ib_send_wr *bad; struct mlx5_umr_wr wr; struct ib_sge sg; int err = 0; const int page_index_alignment = MLX5_UMR_MTT_ALIGNMENT / sizeof(u64); const int page_index_mask = page_index_alignment - 1; size_t pages_mapped = 0; size_t pages_to_map = 0; size_t pages_iter = 0; int use_emergency_buf = 0; /* UMR copies MTTs in units of MLX5_UMR_MTT_ALIGNMENT bytes, * so we need to align the offset and length accordingly */ if (start_page_index & page_index_mask) { npages += start_page_index & page_index_mask; start_page_index &= ~page_index_mask; } pages_to_map = ALIGN(npages, page_index_alignment); if (start_page_index + pages_to_map > MLX5_MAX_UMR_PAGES) return -EINVAL; size = sizeof(u64) * pages_to_map; size = min_t(int, PAGE_SIZE, size); /* We allocate with GFP_ATOMIC to avoid recursion into page-reclaim * code, when we are called from an invalidation. The pas buffer must * be 2k-aligned for Connect-IB. */ pas = (__be64 *)get_zeroed_page(GFP_ATOMIC); if (!pas) { mlx5_ib_warn(dev, "unable to allocate memory during MTT update, falling back to slower chunked mechanism.\n"); pas = mlx5_ib_update_mtt_emergency_buffer; size = MLX5_UMR_MTT_MIN_CHUNK_SIZE; use_emergency_buf = 1; mutex_lock(&mlx5_ib_update_mtt_emergency_buffer_mutex); memset(pas, 0, size); } pages_iter = size / sizeof(u64); dma = dma_map_single(ddev, pas, size, DMA_TO_DEVICE); if (dma_mapping_error(ddev, dma)) { mlx5_ib_err(dev, "unable to map DMA during MTT update.\n"); err = -ENOMEM; goto free_pas; } for (pages_mapped = 0; pages_mapped < pages_to_map && !err; pages_mapped += pages_iter, start_page_index += pages_iter) { dma_sync_single_for_cpu(ddev, dma, size, DMA_TO_DEVICE); npages = min_t(size_t, pages_iter, ib_umem_num_pages(umem) - start_page_index); if (!zap) { __mlx5_ib_populate_pas(dev, umem, PAGE_SHIFT, start_page_index, npages, pas, MLX5_IB_MTT_PRESENT); /* Clear padding after the pages brought from the * umem. */ memset(pas + npages, 0, size - npages * sizeof(u64)); } dma_sync_single_for_device(ddev, dma, size, DMA_TO_DEVICE); mlx5_ib_init_umr_context(&umr_context); memset(&wr, 0, sizeof(wr)); wr.wr.wr_cqe = &umr_context.cqe; sg.addr = dma; sg.length = ALIGN(npages * sizeof(u64), MLX5_UMR_MTT_ALIGNMENT); sg.lkey = dev->umrc.pd->local_dma_lkey; wr.wr.send_flags = MLX5_IB_SEND_UMR_FAIL_IF_FREE | MLX5_IB_SEND_UMR_UPDATE_MTT; wr.wr.sg_list = &sg; wr.wr.num_sge = 1; wr.wr.opcode = MLX5_IB_WR_UMR; wr.npages = sg.length / sizeof(u64); wr.page_shift = PAGE_SHIFT; wr.mkey = mr->mmkey.key; wr.target.offset = start_page_index; down(&umrc->sem); err = ib_post_send(umrc->qp, &wr.wr, &bad); if (err) { mlx5_ib_err(dev, "UMR post send failed, err %d\n", err); } else { wait_for_completion(&umr_context.done); if (umr_context.status != IB_WC_SUCCESS) { mlx5_ib_err(dev, "UMR completion failed, code %d\n", umr_context.status); err = -EFAULT; } } up(&umrc->sem); } dma_unmap_single(ddev, dma, size, DMA_TO_DEVICE); free_pas: if (!use_emergency_buf) free_page((unsigned long)pas); else mutex_unlock(&mlx5_ib_update_mtt_emergency_buffer_mutex); return err; } #endif /* * If ibmr is NULL it will be allocated by reg_create. * Else, the given ibmr will be used. */ static struct mlx5_ib_mr *reg_create(struct ib_mr *ibmr, struct ib_pd *pd, u64 virt_addr, u64 length, struct ib_umem *umem, int npages, int page_shift, int access_flags) { struct mlx5_ib_dev *dev = to_mdev(pd->device); struct mlx5_ib_mr *mr; __be64 *pas; void *mkc; int inlen; u32 *in; int err; bool pg_cap = !!(MLX5_CAP_GEN(dev->mdev, pg)); mr = ibmr ? to_mmr(ibmr) : kzalloc(sizeof(*mr), GFP_KERNEL); if (!mr) return ERR_PTR(-ENOMEM); inlen = MLX5_ST_SZ_BYTES(create_mkey_in) + sizeof(*pas) * ((npages + 1) / 2) * 2; in = mlx5_vzalloc(inlen); if (!in) { err = -ENOMEM; goto err_1; } pas = (__be64 *)MLX5_ADDR_OF(create_mkey_in, in, klm_pas_mtt); mlx5_ib_populate_pas(dev, umem, page_shift, pas, pg_cap ? MLX5_IB_MTT_PRESENT : 0); /* The pg_access bit allows setting the access flags * in the page list submitted with the command. */ MLX5_SET(create_mkey_in, in, pg_access, !!(pg_cap)); mkc = MLX5_ADDR_OF(create_mkey_in, in, memory_key_mkey_entry); MLX5_SET(mkc, mkc, access_mode, MLX5_ACCESS_MODE_MTT); MLX5_SET(mkc, mkc, a, !!(access_flags & IB_ACCESS_REMOTE_ATOMIC)); MLX5_SET(mkc, mkc, rw, !!(access_flags & IB_ACCESS_REMOTE_WRITE)); MLX5_SET(mkc, mkc, rr, !!(access_flags & IB_ACCESS_REMOTE_READ)); MLX5_SET(mkc, mkc, lw, !!(access_flags & IB_ACCESS_LOCAL_WRITE)); MLX5_SET(mkc, mkc, lr, 1); MLX5_SET64(mkc, mkc, start_addr, virt_addr); MLX5_SET64(mkc, mkc, len, length); MLX5_SET(mkc, mkc, pd, to_mpd(pd)->pdn); MLX5_SET(mkc, mkc, bsf_octword_size, 0); MLX5_SET(mkc, mkc, translations_octword_size, get_octo_len(virt_addr, length, 1 << page_shift)); MLX5_SET(mkc, mkc, log_page_size, page_shift); MLX5_SET(mkc, mkc, qpn, 0xffffff); MLX5_SET(create_mkey_in, in, translations_octword_actual_size, get_octo_len(virt_addr, length, 1 << page_shift)); err = mlx5_core_create_mkey(dev->mdev, &mr->mmkey, (struct mlx5_create_mkey_mbox_in *)in, inlen, NULL, NULL, NULL); if (err) { mlx5_ib_warn(dev, "create mkey failed\n"); goto err_2; } mr->umem = umem; mr->dev = dev; mr->live = 1; kvfree(in); mlx5_ib_dbg(dev, "mkey = 0x%x\n", mr->mmkey.key); return mr; err_2: kvfree(in); err_1: if (!ibmr) kfree(mr); return ERR_PTR(err); } static void set_mr_fileds(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr, int npages, u64 length, int access_flags) { mr->npages = npages; atomic_add(npages, &dev->mdev->priv.reg_pages); mr->ibmr.lkey = mr->mmkey.key; mr->ibmr.rkey = mr->mmkey.key; mr->ibmr.length = length; mr->access_flags = access_flags; } struct ib_mr *mlx5_ib_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, u64 virt_addr, int access_flags, struct ib_udata *udata) { struct mlx5_ib_dev *dev = to_mdev(pd->device); struct mlx5_ib_mr *mr = NULL; struct ib_umem *umem; int page_shift; int npages; int ncont; int order; int err; mlx5_ib_dbg(dev, "start 0x%llx, virt_addr 0x%llx, length 0x%llx, access_flags 0x%x\n", (long long)start, (long long)virt_addr, (long long)length, access_flags); umem = mr_umem_get(pd, start, length, access_flags, &npages, &page_shift, &ncont, &order); if (IS_ERR(umem)) return (void *)umem; if (use_umr(order)) { mr = reg_umr(pd, umem, virt_addr, length, ncont, page_shift, order, access_flags); if (PTR_ERR(mr) == -EAGAIN) { mlx5_ib_dbg(dev, "cache empty for order %d", order); mr = NULL; } } else if (access_flags & IB_ACCESS_ON_DEMAND) { err = -EINVAL; pr_err("Got MR registration for ODP MR > 512MB, not supported for Connect-IB"); goto error; } if (!mr) { mutex_lock(&dev->slow_path_mutex); mr = reg_create(NULL, pd, virt_addr, length, umem, ncont, page_shift, access_flags); mutex_unlock(&dev->slow_path_mutex); } if (IS_ERR(mr)) { err = PTR_ERR(mr); goto error; } mlx5_ib_dbg(dev, "mkey 0x%x\n", mr->mmkey.key); mr->umem = umem; set_mr_fileds(dev, mr, npages, length, access_flags); #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING update_odp_mr(mr); #endif return &mr->ibmr; error: ib_umem_release(umem); return ERR_PTR(err); } static int unreg_umr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr) { struct mlx5_core_dev *mdev = dev->mdev; struct umr_common *umrc = &dev->umrc; struct mlx5_ib_umr_context umr_context; struct mlx5_umr_wr umrwr = {}; struct ib_send_wr *bad; int err; if (mdev->state == MLX5_DEVICE_STATE_INTERNAL_ERROR) return 0; mlx5_ib_init_umr_context(&umr_context); umrwr.wr.wr_cqe = &umr_context.cqe; prep_umr_unreg_wqe(dev, &umrwr.wr, mr->mmkey.key); down(&umrc->sem); err = ib_post_send(umrc->qp, &umrwr.wr, &bad); if (err) { up(&umrc->sem); mlx5_ib_dbg(dev, "err %d\n", err); goto error; } else { wait_for_completion(&umr_context.done); up(&umrc->sem); } if (umr_context.status != IB_WC_SUCCESS) { mlx5_ib_warn(dev, "unreg umr failed\n"); err = -EFAULT; goto error; } return 0; error: return err; } static int rereg_umr(struct ib_pd *pd, struct mlx5_ib_mr *mr, u64 virt_addr, u64 length, int npages, int page_shift, int order, int access_flags, int flags) { struct mlx5_ib_dev *dev = to_mdev(pd->device); struct device *ddev = dev->ib_dev.dma_device; struct mlx5_ib_umr_context umr_context; struct ib_send_wr *bad; struct mlx5_umr_wr umrwr = {}; struct ib_sge sg; struct umr_common *umrc = &dev->umrc; dma_addr_t dma = 0; __be64 *mr_pas = NULL; int size; int err; mlx5_ib_init_umr_context(&umr_context); umrwr.wr.wr_cqe = &umr_context.cqe; umrwr.wr.send_flags = MLX5_IB_SEND_UMR_FAIL_IF_FREE; if (flags & IB_MR_REREG_TRANS) { err = dma_map_mr_pas(dev, mr->umem, npages, page_shift, &size, &mr_pas, &dma); if (err) return err; umrwr.target.virt_addr = virt_addr; umrwr.length = length; umrwr.wr.send_flags |= MLX5_IB_SEND_UMR_UPDATE_TRANSLATION; } prep_umr_wqe_common(pd, &umrwr.wr, &sg, dma, npages, mr->mmkey.key, page_shift); if (flags & IB_MR_REREG_PD) { umrwr.pd = pd; umrwr.wr.send_flags |= MLX5_IB_SEND_UMR_UPDATE_PD; } if (flags & IB_MR_REREG_ACCESS) { umrwr.access_flags = access_flags; umrwr.wr.send_flags |= MLX5_IB_SEND_UMR_UPDATE_ACCESS; } /* post send request to UMR QP */ down(&umrc->sem); err = ib_post_send(umrc->qp, &umrwr.wr, &bad); if (err) { mlx5_ib_warn(dev, "post send failed, err %d\n", err); } else { wait_for_completion(&umr_context.done); if (umr_context.status != IB_WC_SUCCESS) { mlx5_ib_warn(dev, "reg umr failed (%u)\n", umr_context.status); err = -EFAULT; } } up(&umrc->sem); if (flags & IB_MR_REREG_TRANS) { dma_unmap_single(ddev, dma, size, DMA_TO_DEVICE); kfree(mr_pas); } return err; } int mlx5_ib_rereg_user_mr(struct ib_mr *ib_mr, int flags, u64 start, u64 length, u64 virt_addr, int new_access_flags, struct ib_pd *new_pd, struct ib_udata *udata) { struct mlx5_ib_dev *dev = to_mdev(ib_mr->device); struct mlx5_ib_mr *mr = to_mmr(ib_mr); struct ib_pd *pd = (flags & IB_MR_REREG_PD) ? new_pd : ib_mr->pd; int access_flags = flags & IB_MR_REREG_ACCESS ? new_access_flags : mr->access_flags; u64 addr = (flags & IB_MR_REREG_TRANS) ? virt_addr : mr->umem->address; u64 len = (flags & IB_MR_REREG_TRANS) ? length : mr->umem->length; int page_shift = 0; int npages = 0; int ncont = 0; int order = 0; int err; mlx5_ib_dbg(dev, "start 0x%llx, virt_addr 0x%llx, length 0x%llx, access_flags 0x%x\n", (long long)start, (long long)virt_addr, (long long)length, access_flags); if (flags != IB_MR_REREG_PD) { /* * Replace umem. This needs to be done whether or not UMR is * used. */ flags |= IB_MR_REREG_TRANS; ib_umem_release(mr->umem); mr->umem = mr_umem_get(pd, addr, len, access_flags, &npages, &page_shift, &ncont, &order); if (IS_ERR(mr->umem)) { err = PTR_ERR(mr->umem); mr->umem = NULL; return err; } } if (flags & IB_MR_REREG_TRANS && !use_umr_mtt_update(mr, addr, len)) { /* * UMR can't be used - MKey needs to be replaced. */ if (mr->umred) { err = unreg_umr(dev, mr); if (err) mlx5_ib_warn(dev, "Failed to unregister MR\n"); } else { err = destroy_mkey(dev, mr); if (err) mlx5_ib_warn(dev, "Failed to destroy MKey\n"); } if (err) return err; mr = reg_create(ib_mr, pd, addr, len, mr->umem, ncont, page_shift, access_flags); if (IS_ERR(mr)) return PTR_ERR(mr); mr->umred = 0; } else { /* * Send a UMR WQE */ err = rereg_umr(pd, mr, addr, len, npages, page_shift, order, access_flags, flags); if (err) { mlx5_ib_warn(dev, "Failed to rereg UMR\n"); return err; } } if (flags & IB_MR_REREG_PD) { ib_mr->pd = pd; mr->mmkey.pd = to_mpd(pd)->pdn; } if (flags & IB_MR_REREG_ACCESS) mr->access_flags = access_flags; if (flags & IB_MR_REREG_TRANS) { atomic_sub(mr->npages, &dev->mdev->priv.reg_pages); set_mr_fileds(dev, mr, npages, len, access_flags); mr->mmkey.iova = addr; mr->mmkey.size = len; } #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING update_odp_mr(mr); #endif return 0; } static int mlx5_alloc_priv_descs(struct ib_device *device, struct mlx5_ib_mr *mr, int ndescs, int desc_size) { int size = ndescs * desc_size; int add_size; int ret; add_size = max_t(int, MLX5_UMR_ALIGN - 1, 0); mr->descs_alloc = kzalloc(size + add_size, GFP_KERNEL); if (!mr->descs_alloc) return -ENOMEM; mr->descs = PTR_ALIGN(mr->descs_alloc, MLX5_UMR_ALIGN); mr->desc_map = dma_map_single(device->dma_device, mr->descs, size, DMA_TO_DEVICE); if (dma_mapping_error(device->dma_device, mr->desc_map)) { ret = -ENOMEM; goto err; } return 0; err: kfree(mr->descs_alloc); return ret; } static void mlx5_free_priv_descs(struct mlx5_ib_mr *mr) { if (mr->descs) { struct ib_device *device = mr->ibmr.device; int size = mr->max_descs * mr->desc_size; dma_unmap_single(device->dma_device, mr->desc_map, size, DMA_TO_DEVICE); kfree(mr->descs_alloc); mr->descs = NULL; } } static int clean_mr(struct mlx5_ib_mr *mr) { struct mlx5_ib_dev *dev = to_mdev(mr->ibmr.device); int umred = mr->umred; int err; if (mr->sig) { if (mlx5_core_destroy_psv(dev->mdev, mr->sig->psv_memory.psv_idx)) mlx5_ib_warn(dev, "failed to destroy mem psv %d\n", mr->sig->psv_memory.psv_idx); if (mlx5_core_destroy_psv(dev->mdev, mr->sig->psv_wire.psv_idx)) mlx5_ib_warn(dev, "failed to destroy wire psv %d\n", mr->sig->psv_wire.psv_idx); kfree(mr->sig); mr->sig = NULL; } mlx5_free_priv_descs(mr); if (!umred) { err = destroy_mkey(dev, mr); if (err) { mlx5_ib_warn(dev, "failed to destroy mkey 0x%x (%d)\n", mr->mmkey.key, err); return err; } } else { err = unreg_umr(dev, mr); if (err) { mlx5_ib_warn(dev, "failed unregister\n"); return err; } free_cached_mr(dev, mr); } if (!umred) kfree(mr); return 0; } +CTASSERT(sizeof(((struct ib_phys_buf *)0)->size) == 8); + +struct ib_mr * +mlx5_ib_reg_phys_mr(struct ib_pd *pd, + struct ib_phys_buf *buffer_list, + int num_phys_buf, + int access_flags, + u64 *virt_addr) +{ + struct mlx5_ib_dev *dev = to_mdev(pd->device); + struct mlx5_ib_mr *mr; + __be64 *pas; + void *mkc; + u32 *in; + u64 total_size; + u32 octo_len; + bool pg_cap = !!(MLX5_CAP_GEN(dev->mdev, pg)); + unsigned long mask; + int shift; + int npages; + int inlen; + int err; + int i, j, n; + + mask = buffer_list[0].addr ^ *virt_addr; + total_size = 0; + for (i = 0; i < num_phys_buf; ++i) { + if (i != 0) + mask |= buffer_list[i].addr; + if (i != num_phys_buf - 1) + mask |= buffer_list[i].addr + buffer_list[i].size; + + total_size += buffer_list[i].size; + } + + if (mask & ~PAGE_MASK) + return ERR_PTR(-EINVAL); + + shift = __ffs(mask | 1 << 31); + + buffer_list[0].size += buffer_list[0].addr & ((1ULL << shift) - 1); + buffer_list[0].addr &= ~0ULL << shift; + + npages = 0; + for (i = 0; i < num_phys_buf; ++i) + npages += (buffer_list[i].size + (1ULL << shift) - 1) >> shift; + + if (!npages) { + mlx5_ib_warn(dev, "avoid zero region\n"); + return ERR_PTR(-EINVAL); + } + + mr = kzalloc(sizeof *mr, GFP_KERNEL); + if (!mr) + return ERR_PTR(-ENOMEM); + + octo_len = get_octo_len(*virt_addr, total_size, 1ULL << shift); + octo_len = ALIGN(octo_len, 4); + + inlen = MLX5_ST_SZ_BYTES(create_mkey_in) + (octo_len * 16); + in = mlx5_vzalloc(inlen); + if (!in) { + kfree(mr); + return ERR_PTR(-ENOMEM); + } + pas = (__be64 *)MLX5_ADDR_OF(create_mkey_in, in, klm_pas_mtt); + + n = 0; + for (i = 0; i < num_phys_buf; ++i) { + for (j = 0; + j < (buffer_list[i].size + (1ULL << shift) - 1) >> shift; + ++j) { + u64 temp = buffer_list[i].addr + ((u64) j << shift); + if (pg_cap) + temp |= MLX5_IB_MTT_PRESENT; + pas[n++] = cpu_to_be64(temp); + } + } + + /* + * The MLX5_MKEY_INBOX_PG_ACCESS bit allows setting the access + * flags in the page list submitted with the command: + */ + MLX5_SET(create_mkey_in, in, pg_access, !!(pg_cap)); + + mkc = MLX5_ADDR_OF(create_mkey_in, in, memory_key_mkey_entry); + MLX5_SET(mkc, mkc, access_mode, MLX5_ACCESS_MODE_MTT); + MLX5_SET(mkc, mkc, a, !!(access_flags & IB_ACCESS_REMOTE_ATOMIC)); + MLX5_SET(mkc, mkc, rw, !!(access_flags & IB_ACCESS_REMOTE_WRITE)); + MLX5_SET(mkc, mkc, rr, !!(access_flags & IB_ACCESS_REMOTE_READ)); + MLX5_SET(mkc, mkc, lw, !!(access_flags & IB_ACCESS_LOCAL_WRITE)); + MLX5_SET(mkc, mkc, lr, 1); + + MLX5_SET64(mkc, mkc, start_addr, *virt_addr); + MLX5_SET64(mkc, mkc, len, total_size); + MLX5_SET(mkc, mkc, pd, to_mpd(pd)->pdn); + MLX5_SET(mkc, mkc, bsf_octword_size, 0); + MLX5_SET(mkc, mkc, translations_octword_size, octo_len); + MLX5_SET(mkc, mkc, log_page_size, shift); + MLX5_SET(mkc, mkc, qpn, 0xffffff); + MLX5_SET(create_mkey_in, in, translations_octword_actual_size, octo_len); + + err = mlx5_core_create_mkey(dev->mdev, &mr->mmkey, + (struct mlx5_create_mkey_mbox_in *)in, inlen, + NULL, NULL, NULL); + mr->umem = NULL; + mr->dev = dev; + mr->live = 1; + mr->npages = npages; + mr->ibmr.lkey = mr->mmkey.key; + mr->ibmr.rkey = mr->mmkey.key; + mr->ibmr.length = total_size; + mr->access_flags = access_flags; + + kvfree(in); + + if (err) { + kfree(mr); + return ERR_PTR(err); + } + return &mr->ibmr; +} + int mlx5_ib_dereg_mr(struct ib_mr *ibmr) { struct mlx5_ib_dev *dev = to_mdev(ibmr->device); struct mlx5_ib_mr *mr = to_mmr(ibmr); int npages = mr->npages; struct ib_umem *umem = mr->umem; #ifdef CONFIG_INFINIBAND_ON_DEMAND_PAGING if (umem && umem->odp_data) { /* Prevent new page faults from succeeding */ mr->live = 0; /* Wait for all running page-fault handlers to finish. */ synchronize_srcu(&dev->mr_srcu); /* Destroy all page mappings */ mlx5_ib_invalidate_range(umem, ib_umem_start(umem), ib_umem_end(umem)); /* * We kill the umem before the MR for ODP, * so that there will not be any invalidations in * flight, looking at the *mr struct. */ ib_umem_release(umem); atomic_sub(npages, &dev->mdev->priv.reg_pages); /* Avoid double-freeing the umem. */ umem = NULL; } #endif clean_mr(mr); if (umem) { ib_umem_release(umem); atomic_sub(npages, &dev->mdev->priv.reg_pages); } return 0; } struct ib_mr *mlx5_ib_alloc_mr(struct ib_pd *pd, enum ib_mr_type mr_type, u32 max_num_sg) { struct mlx5_ib_dev *dev = to_mdev(pd->device); int inlen = MLX5_ST_SZ_BYTES(create_mkey_in); int ndescs = ALIGN(max_num_sg, 4); struct mlx5_ib_mr *mr; void *mkc; u32 *in; int err; mr = kzalloc(sizeof(*mr), GFP_KERNEL); if (!mr) return ERR_PTR(-ENOMEM); in = kzalloc(inlen, GFP_KERNEL); if (!in) { err = -ENOMEM; goto err_free; } mkc = MLX5_ADDR_OF(create_mkey_in, in, memory_key_mkey_entry); MLX5_SET(mkc, mkc, free, 1); MLX5_SET(mkc, mkc, translations_octword_size, ndescs); MLX5_SET(mkc, mkc, qpn, 0xffffff); MLX5_SET(mkc, mkc, pd, to_mpd(pd)->pdn); if (mr_type == IB_MR_TYPE_MEM_REG) { mr->access_mode = MLX5_ACCESS_MODE_MTT; MLX5_SET(mkc, mkc, log_page_size, PAGE_SHIFT); err = mlx5_alloc_priv_descs(pd->device, mr, ndescs, sizeof(u64)); if (err) goto err_free_in; mr->desc_size = sizeof(u64); mr->max_descs = ndescs; } else if (mr_type == IB_MR_TYPE_SG_GAPS) { mr->access_mode = MLX5_ACCESS_MODE_KLM; err = mlx5_alloc_priv_descs(pd->device, mr, ndescs, sizeof(struct mlx5_klm)); if (err) goto err_free_in; mr->desc_size = sizeof(struct mlx5_klm); mr->max_descs = ndescs; } else if (mr_type == IB_MR_TYPE_SIGNATURE) { u32 psv_index[2]; MLX5_SET(mkc, mkc, bsf_en, 1); MLX5_SET(mkc, mkc, bsf_octword_size, MLX5_MKEY_BSF_OCTO_SIZE); mr->sig = kzalloc(sizeof(*mr->sig), GFP_KERNEL); if (!mr->sig) { err = -ENOMEM; goto err_free_in; } /* create mem & wire PSVs */ err = mlx5_core_create_psv(dev->mdev, to_mpd(pd)->pdn, 2, psv_index); if (err) goto err_free_sig; mr->access_mode = MLX5_ACCESS_MODE_KLM; mr->sig->psv_memory.psv_idx = psv_index[0]; mr->sig->psv_wire.psv_idx = psv_index[1]; mr->sig->sig_status_checked = true; mr->sig->sig_err_exists = false; /* Next UMR, Arm SIGERR */ ++mr->sig->sigerr_count; } else { mlx5_ib_warn(dev, "Invalid mr type %d\n", mr_type); err = -EINVAL; goto err_free_in; } MLX5_SET(mkc, mkc, access_mode, mr->access_mode); MLX5_SET(mkc, mkc, umr_en, 1); err = mlx5_core_create_mkey(dev->mdev, &mr->mmkey, (struct mlx5_create_mkey_mbox_in *)in, inlen, NULL, NULL, NULL); if (err) goto err_destroy_psv; mr->ibmr.lkey = mr->mmkey.key; mr->ibmr.rkey = mr->mmkey.key; mr->umem = NULL; kfree(in); return &mr->ibmr; err_destroy_psv: if (mr->sig) { if (mlx5_core_destroy_psv(dev->mdev, mr->sig->psv_memory.psv_idx)) mlx5_ib_warn(dev, "failed to destroy mem psv %d\n", mr->sig->psv_memory.psv_idx); if (mlx5_core_destroy_psv(dev->mdev, mr->sig->psv_wire.psv_idx)) mlx5_ib_warn(dev, "failed to destroy wire psv %d\n", mr->sig->psv_wire.psv_idx); } mlx5_free_priv_descs(mr); err_free_sig: kfree(mr->sig); err_free_in: kfree(in); err_free: kfree(mr); return ERR_PTR(err); } struct ib_mw *mlx5_ib_alloc_mw(struct ib_pd *pd, enum ib_mw_type type, struct ib_udata *udata) { struct mlx5_ib_dev *dev = to_mdev(pd->device); int inlen = MLX5_ST_SZ_BYTES(create_mkey_in); struct mlx5_ib_mw *mw = NULL; u32 *in = NULL; void *mkc; int ndescs; int err; struct mlx5_ib_alloc_mw req = {}; struct { __u32 comp_mask; __u32 response_length; } resp = {}; err = ib_copy_from_udata(&req, udata, min(udata->inlen, sizeof(req))); if (err) return ERR_PTR(err); if (req.comp_mask || req.reserved1 || req.reserved2) return ERR_PTR(-EOPNOTSUPP); if (udata->inlen > sizeof(req) && !ib_is_udata_cleared(udata, sizeof(req), udata->inlen - sizeof(req))) return ERR_PTR(-EOPNOTSUPP); ndescs = req.num_klms ? roundup(req.num_klms, 4) : roundup(1, 4); mw = kzalloc(sizeof(*mw), GFP_KERNEL); in = kzalloc(inlen, GFP_KERNEL); if (!mw || !in) { err = -ENOMEM; goto free; } mkc = MLX5_ADDR_OF(create_mkey_in, in, memory_key_mkey_entry); MLX5_SET(mkc, mkc, free, 1); MLX5_SET(mkc, mkc, translations_octword_size, ndescs); MLX5_SET(mkc, mkc, pd, to_mpd(pd)->pdn); MLX5_SET(mkc, mkc, umr_en, 1); MLX5_SET(mkc, mkc, lr, 1); MLX5_SET(mkc, mkc, access_mode, MLX5_ACCESS_MODE_KLM); MLX5_SET(mkc, mkc, en_rinval, !!((type == IB_MW_TYPE_2))); MLX5_SET(mkc, mkc, qpn, 0xffffff); err = mlx5_core_create_mkey(dev->mdev, &mw->mmkey, (struct mlx5_create_mkey_mbox_in *)in, inlen, NULL, NULL, NULL); if (err) goto free; mw->ibmw.rkey = mw->mmkey.key; resp.response_length = min(offsetof(typeof(resp), response_length) + sizeof(resp.response_length), udata->outlen); if (resp.response_length) { err = ib_copy_to_udata(udata, &resp, resp.response_length); if (err) { mlx5_core_destroy_mkey(dev->mdev, &mw->mmkey); goto free; } } kfree(in); return &mw->ibmw; free: kfree(mw); kfree(in); return ERR_PTR(err); } int mlx5_ib_dealloc_mw(struct ib_mw *mw) { struct mlx5_ib_mw *mmw = to_mmw(mw); int err; err = mlx5_core_destroy_mkey((to_mdev(mw->device))->mdev, &mmw->mmkey); if (!err) kfree(mmw); return err; } int mlx5_ib_check_mr_status(struct ib_mr *ibmr, u32 check_mask, struct ib_mr_status *mr_status) { struct mlx5_ib_mr *mmr = to_mmr(ibmr); int ret = 0; if (check_mask & ~IB_MR_CHECK_SIG_STATUS) { pr_err("Invalid status check mask\n"); ret = -EINVAL; goto done; } mr_status->fail_status = 0; if (check_mask & IB_MR_CHECK_SIG_STATUS) { if (!mmr->sig) { ret = -EINVAL; pr_err("signature status check requested on a non-signature enabled MR\n"); goto done; } mmr->sig->sig_status_checked = true; if (!mmr->sig->sig_err_exists) goto done; if (ibmr->lkey == mmr->sig->err_item.key) memcpy(&mr_status->sig_err, &mmr->sig->err_item, sizeof(mr_status->sig_err)); else { mr_status->sig_err.err_type = IB_SIG_BAD_GUARD; mr_status->sig_err.sig_err_offset = 0; mr_status->sig_err.key = mmr->sig->err_item.key; } mmr->sig->sig_err_exists = false; mr_status->fail_status |= IB_MR_CHECK_SIG_STATUS; } done: return ret; } static int mlx5_ib_sg_to_klms(struct mlx5_ib_mr *mr, struct scatterlist *sgl, unsigned short sg_nents, unsigned int *sg_offset_p) { struct scatterlist *sg = sgl; struct mlx5_klm *klms = mr->descs; unsigned int sg_offset = sg_offset_p ? *sg_offset_p : 0; u32 lkey = mr->ibmr.pd->local_dma_lkey; int i; mr->ibmr.iova = sg_dma_address(sg) + sg_offset; mr->ibmr.length = 0; mr->ndescs = sg_nents; for_each_sg(sgl, sg, sg_nents, i) { if (unlikely(i > mr->max_descs)) break; klms[i].va = cpu_to_be64(sg_dma_address(sg) + sg_offset); klms[i].bcount = cpu_to_be32(sg_dma_len(sg) - sg_offset); klms[i].key = cpu_to_be32(lkey); mr->ibmr.length += sg_dma_len(sg); sg_offset = 0; } if (sg_offset_p) *sg_offset_p = sg_offset; return i; } static int mlx5_set_page(struct ib_mr *ibmr, u64 addr) { struct mlx5_ib_mr *mr = to_mmr(ibmr); __be64 *descs; if (unlikely(mr->ndescs == mr->max_descs)) return -ENOMEM; descs = mr->descs; descs[mr->ndescs++] = cpu_to_be64(addr | MLX5_EN_RD | MLX5_EN_WR); return 0; } int mlx5_ib_map_mr_sg(struct ib_mr *ibmr, struct scatterlist *sg, int sg_nents, unsigned int *sg_offset) { struct mlx5_ib_mr *mr = to_mmr(ibmr); int n; mr->ndescs = 0; ib_dma_sync_single_for_cpu(ibmr->device, mr->desc_map, mr->desc_size * mr->max_descs, DMA_TO_DEVICE); if (mr->access_mode == MLX5_ACCESS_MODE_KLM) n = mlx5_ib_sg_to_klms(mr, sg, sg_nents, sg_offset); else n = ib_sg_to_pages(ibmr, sg, sg_nents, sg_offset, mlx5_set_page); ib_dma_sync_single_for_device(ibmr->device, mr->desc_map, mr->desc_size * mr->max_descs, DMA_TO_DEVICE); return n; }