diff options
Diffstat (limited to 'drivers/infiniband/hw/bnxt_re/ib_verbs.c')
| -rw-r--r-- | drivers/infiniband/hw/bnxt_re/ib_verbs.c | 1506 |
1 files changed, 1191 insertions, 315 deletions
diff --git a/drivers/infiniband/hw/bnxt_re/ib_verbs.c b/drivers/infiniband/hw/bnxt_re/ib_verbs.c index e86afecfbe46..f19b55c13d58 100644 --- a/drivers/infiniband/hw/bnxt_re/ib_verbs.c +++ b/drivers/infiniband/hw/bnxt_re/ib_verbs.c @@ -49,9 +49,9 @@ #include <rdma/ib_addr.h> #include <rdma/ib_mad.h> #include <rdma/ib_cache.h> +#include <rdma/ib_pma.h> #include <rdma/uverbs_ioctl.h> - -#include "bnxt_ulp.h" +#include <linux/hashtable.h> #include "roce_hsi.h" #include "qplib_res.h" @@ -61,6 +61,16 @@ #include "bnxt_re.h" #include "ib_verbs.h" +#include "debugfs.h" + +#include <rdma/uverbs_types.h> +#include <rdma/uverbs_std_types.h> + +#include <rdma/ib_user_ioctl_cmds.h> + +#define UVERBS_MODULE_NAME bnxt_re +#include <rdma/uverbs_named_ioctl.h> + #include <rdma/bnxt_re-abi.h> static int __from_ib_access_flags(int iflags) @@ -84,9 +94,9 @@ static int __from_ib_access_flags(int iflags) return qflags; }; -static enum ib_access_flags __to_ib_access_flags(int qflags) +static int __to_ib_access_flags(int qflags) { - enum ib_access_flags iflags = 0; + int iflags = 0; if (qflags & BNXT_QPLIB_ACCESS_LOCAL_WRITE) iflags |= IB_ACCESS_LOCAL_WRITE; @@ -103,7 +113,57 @@ static enum ib_access_flags __to_ib_access_flags(int qflags) if (qflags & BNXT_QPLIB_ACCESS_ON_DEMAND) iflags |= IB_ACCESS_ON_DEMAND; return iflags; -}; +} + +static u8 __qp_access_flags_from_ib(struct bnxt_qplib_chip_ctx *cctx, int iflags) +{ + u8 qflags = 0; + + if (!bnxt_qplib_is_chip_gen_p5_p7(cctx)) + /* For Wh+ */ + return (u8)__from_ib_access_flags(iflags); + + /* For P5, P7 and later chips */ + if (iflags & IB_ACCESS_LOCAL_WRITE) + qflags |= CMDQ_MODIFY_QP_ACCESS_LOCAL_WRITE; + if (iflags & IB_ACCESS_REMOTE_WRITE) + qflags |= CMDQ_MODIFY_QP_ACCESS_REMOTE_WRITE; + if (iflags & IB_ACCESS_REMOTE_READ) + qflags |= CMDQ_MODIFY_QP_ACCESS_REMOTE_READ; + if (iflags & IB_ACCESS_REMOTE_ATOMIC) + qflags |= CMDQ_MODIFY_QP_ACCESS_REMOTE_ATOMIC; + + return qflags; +} + +static int __qp_access_flags_to_ib(struct bnxt_qplib_chip_ctx *cctx, u8 qflags) +{ + int iflags = 0; + + if (!bnxt_qplib_is_chip_gen_p5_p7(cctx)) + /* For Wh+ */ + return __to_ib_access_flags(qflags); + + /* For P5, P7 and later chips */ + if (qflags & CMDQ_MODIFY_QP_ACCESS_LOCAL_WRITE) + iflags |= IB_ACCESS_LOCAL_WRITE; + if (qflags & CMDQ_MODIFY_QP_ACCESS_REMOTE_WRITE) + iflags |= IB_ACCESS_REMOTE_WRITE; + if (qflags & CMDQ_MODIFY_QP_ACCESS_REMOTE_READ) + iflags |= IB_ACCESS_REMOTE_READ; + if (qflags & CMDQ_MODIFY_QP_ACCESS_REMOTE_ATOMIC) + iflags |= IB_ACCESS_REMOTE_ATOMIC; + + return iflags; +} + +static void bnxt_re_check_and_set_relaxed_ordering(struct bnxt_re_dev *rdev, + struct bnxt_qplib_mrw *qplib_mr) +{ + if (_is_relaxed_ordering_supported(rdev->dev_attr->dev_cap_flags2) && + pcie_relaxed_ordering_enabled(rdev->en_dev->pdev)) + qplib_mr->flags |= CMDQ_REGISTER_MR_FLAGS_ENABLE_RO; +} static int bnxt_re_build_sgl(struct ib_sge *ib_sg_list, struct bnxt_qplib_sge *sg_list, int num) @@ -125,7 +185,7 @@ int bnxt_re_query_device(struct ib_device *ibdev, struct ib_udata *udata) { struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); - struct bnxt_qplib_dev_attr *dev_attr = &rdev->dev_attr; + struct bnxt_qplib_dev_attr *dev_attr = rdev->dev_attr; memset(ib_attr, 0, sizeof(*ib_attr)); memcpy(&ib_attr->fw_ver, dev_attr->fw_ver, @@ -138,7 +198,7 @@ int bnxt_re_query_device(struct ib_device *ibdev, ib_attr->vendor_id = rdev->en_dev->pdev->vendor; ib_attr->vendor_part_id = rdev->en_dev->pdev->device; - ib_attr->hw_ver = rdev->en_dev->pdev->subsystem_device; + ib_attr->hw_ver = rdev->en_dev->pdev->revision; ib_attr->max_qp = dev_attr->max_qp; ib_attr->max_qp_wr = dev_attr->max_qp_wqes; ib_attr->device_cap_flags = @@ -193,12 +253,29 @@ int bnxt_re_query_device(struct ib_device *ibdev, return 0; } +int bnxt_re_modify_device(struct ib_device *ibdev, + int device_modify_mask, + struct ib_device_modify *device_modify) +{ + ibdev_dbg(ibdev, "Modify device with mask 0x%x", device_modify_mask); + + if (device_modify_mask & ~IB_DEVICE_MODIFY_NODE_DESC) + return -EOPNOTSUPP; + + if (!(device_modify_mask & IB_DEVICE_MODIFY_NODE_DESC)) + return 0; + + memcpy(ibdev->node_desc, device_modify->node_desc, IB_DEVICE_NODE_DESC_MAX); + return 0; +} + /* Port */ int bnxt_re_query_port(struct ib_device *ibdev, u32 port_num, struct ib_port_attr *port_attr) { struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); - struct bnxt_qplib_dev_attr *dev_attr = &rdev->dev_attr; + struct bnxt_qplib_dev_attr *dev_attr = rdev->dev_attr; + int rc; memset(port_attr, 0, sizeof(*port_attr)); @@ -211,7 +288,9 @@ int bnxt_re_query_port(struct ib_device *ibdev, u32 port_num, } port_attr->max_mtu = IB_MTU_4096; port_attr->active_mtu = iboe_get_mtu(rdev->netdev->mtu); - port_attr->gid_tbl_len = dev_attr->max_sgid; + /* One GID is reserved for RawEth QP. Report one less */ + port_attr->gid_tbl_len = (rdev->rcfw.roce_mirror ? (dev_attr->max_sgid - 1) : + dev_attr->max_sgid); port_attr->port_cap_flags = IB_PORT_CM_SUP | IB_PORT_REINIT_SUP | IB_PORT_DEVICE_MGMT_SUP | IB_PORT_VENDOR_CLASS_SUP; @@ -228,10 +307,10 @@ int bnxt_re_query_port(struct ib_device *ibdev, u32 port_num, port_attr->sm_sl = 0; port_attr->subnet_timeout = 0; port_attr->init_type_reply = 0; - port_attr->active_speed = rdev->active_speed; - port_attr->active_width = rdev->active_width; + rc = ib_get_eth_speed(&rdev->ibdev, port_num, &port_attr->active_speed, + &port_attr->active_width); - return 0; + return rc; } int bnxt_re_get_port_immutable(struct ib_device *ibdev, u32 port_num, @@ -255,8 +334,8 @@ void bnxt_re_query_fw_str(struct ib_device *ibdev, char *str) struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); snprintf(str, IB_FW_VERSION_NAME_MAX, "%d.%d.%d.%d", - rdev->dev_attr.fw_ver[0], rdev->dev_attr.fw_ver[1], - rdev->dev_attr.fw_ver[2], rdev->dev_attr.fw_ver[3]); + rdev->dev_attr->fw_ver[0], rdev->dev_attr->fw_ver[1], + rdev->dev_attr->fw_ver[2], rdev->dev_attr->fw_ver[3]); } int bnxt_re_query_pkey(struct ib_device *ibdev, u32 port_num, @@ -274,7 +353,7 @@ int bnxt_re_query_gid(struct ib_device *ibdev, u32 port_num, int index, union ib_gid *gid) { struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); - int rc = 0; + int rc; /* Ignore port_num */ memset(gid, 0, sizeof(*gid)); @@ -298,7 +377,7 @@ int bnxt_re_del_gid(const struct ib_gid_attr *attr, void **context) if (!ctx) return -EINVAL; - if (sgid_tbl && sgid_tbl->active) { + if (sgid_tbl->active) { if (ctx->idx >= sgid_tbl->max) return -EINVAL; gid_to_del = &sgid_tbl->tbl[ctx->idx].gid; @@ -352,7 +431,7 @@ int bnxt_re_add_gid(const struct ib_gid_attr *attr, void **context) rc = bnxt_qplib_add_sgid(sgid_tbl, (struct bnxt_qplib_gid *)&attr->gid, rdev->qplib_res.netdev->dev_addr, - vlan_id, true, &tbl_idx); + vlan_id, true, &tbl_idx, false, 0); if (rc == -EALREADY) { ctx_tbl = sgid_tbl->ctx; ctx_tbl[tbl_idx]->refcnt++; @@ -390,6 +469,10 @@ static void bnxt_re_create_fence_wqe(struct bnxt_re_pd *pd) struct bnxt_re_fence_data *fence = &pd->fence; struct ib_mr *ib_mr = &fence->mr->ib_mr; struct bnxt_qplib_swqe *wqe = &fence->bind_wqe; + struct bnxt_re_dev *rdev = pd->rdev; + + if (bnxt_qplib_is_chip_gen_p5_p7(rdev->chip_ctx)) + return; memset(wqe, 0, sizeof(*wqe)); wqe->type = BNXT_QPLIB_SWQE_TYPE_BIND_MW; @@ -444,6 +527,9 @@ static void bnxt_re_destroy_fence_mr(struct bnxt_re_pd *pd) struct device *dev = &rdev->en_dev->pdev->dev; struct bnxt_re_mr *mr = fence->mr; + if (bnxt_qplib_is_chip_gen_p5_p7(rdev->chip_ctx)) + return; + if (fence->mw) { bnxt_re_dealloc_mw(fence->mw); fence->mw = NULL; @@ -475,6 +561,9 @@ static int bnxt_re_create_fence_mr(struct bnxt_re_pd *pd) struct ib_mw *mw; int rc; + if (bnxt_qplib_is_chip_gen_p5_p7(rdev->chip_ctx)) + return 0; + dma_addr = dma_map_single(dev, fence->va, BNXT_RE_FENCE_BYTES, DMA_BIDIRECTIONAL); rc = dma_mapping_error(dev, dma_addr); @@ -496,19 +585,24 @@ static int bnxt_re_create_fence_mr(struct bnxt_re_pd *pd) mr->rdev = rdev; mr->qplib_mr.pd = &pd->qplib_pd; mr->qplib_mr.type = CMDQ_ALLOCATE_MRW_MRW_FLAGS_PMR; - mr->qplib_mr.flags = __from_ib_access_flags(mr_access_flags); - rc = bnxt_qplib_alloc_mrw(&rdev->qplib_res, &mr->qplib_mr); - if (rc) { - ibdev_err(&rdev->ibdev, "Failed to alloc fence-HW-MR\n"); - goto fail; - } + mr->qplib_mr.access_flags = __from_ib_access_flags(mr_access_flags); + if (!_is_alloc_mr_unified(rdev->dev_attr->dev_cap_flags)) { + rc = bnxt_qplib_alloc_mrw(&rdev->qplib_res, &mr->qplib_mr); + if (rc) { + ibdev_err(&rdev->ibdev, "Failed to alloc fence-HW-MR\n"); + goto fail; + } - /* Register MR */ - mr->ib_mr.lkey = mr->qplib_mr.lkey; + /* Register MR */ + mr->ib_mr.lkey = mr->qplib_mr.lkey; + } else { + mr->qplib_mr.flags = CMDQ_REGISTER_MR_FLAGS_ALLOC_MR; + } mr->qplib_mr.va = (u64)(unsigned long)fence->va; mr->qplib_mr.total_size = BNXT_RE_FENCE_BYTES; rc = bnxt_qplib_reg_mr(&rdev->qplib_res, &mr->qplib_mr, NULL, - BNXT_RE_FENCE_PBL_SIZE, PAGE_SIZE); + BNXT_RE_FENCE_PBL_SIZE, PAGE_SIZE, + _is_alloc_mr_unified(rdev->dev_attr->dev_cap_flags)); if (rc) { ibdev_err(&rdev->ibdev, "Failed to register fence-MR\n"); goto fail; @@ -533,19 +627,67 @@ fail: return rc; } +static struct bnxt_re_user_mmap_entry* +bnxt_re_mmap_entry_insert(struct bnxt_re_ucontext *uctx, u64 mem_offset, + enum bnxt_re_mmap_flag mmap_flag, u64 *offset) +{ + struct bnxt_re_user_mmap_entry *entry; + int ret; + + entry = kzalloc(sizeof(*entry), GFP_KERNEL); + if (!entry) + return NULL; + + entry->mem_offset = mem_offset; + entry->mmap_flag = mmap_flag; + entry->uctx = uctx; + + switch (mmap_flag) { + case BNXT_RE_MMAP_SH_PAGE: + ret = rdma_user_mmap_entry_insert_exact(&uctx->ib_uctx, + &entry->rdma_entry, PAGE_SIZE, 0); + break; + case BNXT_RE_MMAP_UC_DB: + case BNXT_RE_MMAP_WC_DB: + case BNXT_RE_MMAP_DBR_BAR: + case BNXT_RE_MMAP_DBR_PAGE: + case BNXT_RE_MMAP_TOGGLE_PAGE: + ret = rdma_user_mmap_entry_insert(&uctx->ib_uctx, + &entry->rdma_entry, PAGE_SIZE); + break; + default: + ret = -EINVAL; + break; + } + + if (ret) { + kfree(entry); + return NULL; + } + if (offset) + *offset = rdma_user_mmap_get_offset(&entry->rdma_entry); + + return entry; +} + /* Protection Domains */ int bnxt_re_dealloc_pd(struct ib_pd *ib_pd, struct ib_udata *udata) { struct bnxt_re_pd *pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); struct bnxt_re_dev *rdev = pd->rdev; + if (udata) { + rdma_user_mmap_entry_remove(pd->pd_db_mmap); + pd->pd_db_mmap = NULL; + } + bnxt_re_destroy_fence_mr(pd); if (pd->qplib_pd.id) { if (!bnxt_qplib_dealloc_pd(&rdev->qplib_res, &rdev->qplib_res.pd_tbl, &pd->qplib_pd)) - atomic_dec(&rdev->pd_count); + atomic_dec(&rdev->stats.res.pd_count); } return 0; } @@ -557,25 +699,27 @@ int bnxt_re_alloc_pd(struct ib_pd *ibpd, struct ib_udata *udata) struct bnxt_re_ucontext *ucntx = rdma_udata_to_drv_context( udata, struct bnxt_re_ucontext, ib_uctx); struct bnxt_re_pd *pd = container_of(ibpd, struct bnxt_re_pd, ib_pd); - int rc; + struct bnxt_re_user_mmap_entry *entry = NULL; + u32 active_pds; + int rc = 0; pd->rdev = rdev; - if (bnxt_qplib_alloc_pd(&rdev->qplib_res.pd_tbl, &pd->qplib_pd)) { + if (bnxt_qplib_alloc_pd(&rdev->qplib_res, &pd->qplib_pd)) { ibdev_err(&rdev->ibdev, "Failed to allocate HW PD"); rc = -ENOMEM; goto fail; } if (udata) { - struct bnxt_re_pd_resp resp; + struct bnxt_re_pd_resp resp = {}; if (!ucntx->dpi.dbr) { /* Allocate DPI in alloc_pd to avoid failing of * ibv_devinfo and family of application when DPIs * are depleted. */ - if (bnxt_qplib_alloc_dpi(&rdev->qplib_res.dpi_tbl, - &ucntx->dpi, ucntx)) { + if (bnxt_qplib_alloc_dpi(&rdev->qplib_res, + &ucntx->dpi, ucntx, BNXT_QPLIB_DPI_TYPE_UC)) { rc = -ENOMEM; goto dbfail; } @@ -584,12 +728,21 @@ int bnxt_re_alloc_pd(struct ib_pd *ibpd, struct ib_udata *udata) resp.pdid = pd->qplib_pd.id; /* Still allow mapping this DBR to the new user PD. */ resp.dpi = ucntx->dpi.dpi; - resp.dbr = (u64)ucntx->dpi.umdbr; - rc = ib_copy_to_udata(udata, &resp, sizeof(resp)); + entry = bnxt_re_mmap_entry_insert(ucntx, (u64)ucntx->dpi.umdbr, + BNXT_RE_MMAP_UC_DB, &resp.dbr); + + if (!entry) { + rc = -ENOMEM; + goto dbfail; + } + + pd->pd_db_mmap = &entry->rdma_entry; + + rc = ib_copy_to_udata(udata, &resp, min(sizeof(resp), udata->outlen)); if (rc) { - ibdev_err(&rdev->ibdev, - "Failed to copy user response\n"); + rdma_user_mmap_entry_remove(pd->pd_db_mmap); + rc = -EFAULT; goto dbfail; } } @@ -598,7 +751,9 @@ int bnxt_re_alloc_pd(struct ib_pd *ibpd, struct ib_udata *udata) if (bnxt_re_create_fence_mr(pd)) ibdev_warn(&rdev->ibdev, "Failed to create Fence-MR\n"); - atomic_inc(&rdev->pd_count); + active_pds = atomic_inc_return(&rdev->stats.res.pd_count); + if (active_pds > rdev->stats.res.pd_watermark) + rdev->stats.res.pd_watermark = active_pds; return 0; dbfail: @@ -613,12 +768,20 @@ int bnxt_re_destroy_ah(struct ib_ah *ib_ah, u32 flags) { struct bnxt_re_ah *ah = container_of(ib_ah, struct bnxt_re_ah, ib_ah); struct bnxt_re_dev *rdev = ah->rdev; + bool block = true; + int rc; - bnxt_qplib_destroy_ah(&rdev->qplib_res, &ah->qplib_ah, - !(flags & RDMA_DESTROY_AH_SLEEPABLE)); - atomic_dec(&rdev->ah_count); - - return 0; + block = !(flags & RDMA_DESTROY_AH_SLEEPABLE); + rc = bnxt_qplib_destroy_ah(&rdev->qplib_res, &ah->qplib_ah, block); + if (BNXT_RE_CHECK_RC(rc)) { + if (rc == -ETIMEDOUT) + rc = 0; + else + goto fail; + } + atomic_dec(&rdev->stats.res.ah_count); +fail: + return rc; } static u8 bnxt_re_stack_to_dev_nw_type(enum rdma_network_type ntype) @@ -650,6 +813,7 @@ int bnxt_re_create_ah(struct ib_ah *ib_ah, struct rdma_ah_init_attr *init_attr, const struct ib_gid_attr *sgid_attr; struct bnxt_re_gid_ctx *ctx; struct bnxt_re_ah *ah = container_of(ib_ah, struct bnxt_re_ah, ib_ah); + u32 active_ahs; u8 nw_type; int rc; @@ -702,7 +866,9 @@ int bnxt_re_create_ah(struct ib_ah *ib_ah, struct rdma_ah_init_attr *init_attr, wmb(); /* make sure cache is updated. */ spin_unlock_irqrestore(&uctx->sh_lock, flag); } - atomic_inc(&rdev->ah_count); + active_ahs = atomic_inc_return(&rdev->stats.res.ah_count); + if (active_ahs > rdev->stats.res.ah_watermark) + rdev->stats.res.ah_watermark = active_ahs; return 0; } @@ -748,12 +914,12 @@ void bnxt_re_unlock_cqs(struct bnxt_re_qp *qp, spin_unlock_irqrestore(&qp->scq->cq_lock, flags); } -static int bnxt_re_destroy_gsi_sqp(struct bnxt_re_qp *qp) +static void bnxt_re_destroy_gsi_sqp(struct bnxt_re_qp *qp) { struct bnxt_re_qp *gsi_sqp; struct bnxt_re_ah *gsi_sah; struct bnxt_re_dev *rdev; - int rc = 0; + int rc; rdev = qp->rdev; gsi_sqp = rdev->gsi_ctx.gsi_sqp; @@ -763,22 +929,21 @@ static int bnxt_re_destroy_gsi_sqp(struct bnxt_re_qp *qp) bnxt_qplib_destroy_ah(&rdev->qplib_res, &gsi_sah->qplib_ah, true); - atomic_dec(&rdev->ah_count); + atomic_dec(&rdev->stats.res.ah_count); bnxt_qplib_clean_qp(&qp->qplib_qp); ibdev_dbg(&rdev->ibdev, "Destroy the shadow QP\n"); rc = bnxt_qplib_destroy_qp(&rdev->qplib_res, &gsi_sqp->qplib_qp); - if (rc) { + if (rc) ibdev_err(&rdev->ibdev, "Destroy Shadow QP failed"); - goto fail; - } + bnxt_qplib_free_qp_res(&rdev->qplib_res, &gsi_sqp->qplib_qp); /* remove from active qp list */ mutex_lock(&rdev->qp_lock); list_del(&gsi_sqp->list); mutex_unlock(&rdev->qp_lock); - atomic_dec(&rdev->qp_count); + atomic_dec(&rdev->stats.res.qp_count); kfree(rdev->gsi_ctx.sqp_tbl); kfree(gsi_sah); @@ -786,27 +951,40 @@ static int bnxt_re_destroy_gsi_sqp(struct bnxt_re_qp *qp) rdev->gsi_ctx.gsi_sqp = NULL; rdev->gsi_ctx.gsi_sah = NULL; rdev->gsi_ctx.sqp_tbl = NULL; +} - return 0; -fail: - return rc; +static void bnxt_re_del_unique_gid(struct bnxt_re_dev *rdev) +{ + int rc; + + if (!rdev->rcfw.roce_mirror) + return; + + rc = bnxt_qplib_del_sgid(&rdev->qplib_res.sgid_tbl, + (struct bnxt_qplib_gid *)&rdev->ugid, + 0xFFFF, true); + if (rc) + dev_err(rdev_to_dev(rdev), "Failed to delete unique GID, rc: %d\n", rc); } /* Queue Pairs */ int bnxt_re_destroy_qp(struct ib_qp *ib_qp, struct ib_udata *udata) { struct bnxt_re_qp *qp = container_of(ib_qp, struct bnxt_re_qp, ib_qp); + struct bnxt_qplib_qp *qplib_qp = &qp->qplib_qp; struct bnxt_re_dev *rdev = qp->rdev; + struct bnxt_qplib_nq *scq_nq = NULL; + struct bnxt_qplib_nq *rcq_nq = NULL; unsigned int flags; int rc; + bnxt_re_debug_rem_qpinfo(rdev, qp); + bnxt_qplib_flush_cqn_wq(&qp->qplib_qp); rc = bnxt_qplib_destroy_qp(&rdev->qplib_res, &qp->qplib_qp); - if (rc) { + if (rc) ibdev_err(&rdev->ibdev, "Failed to destroy HW QP"); - return rc; - } if (rdma_is_kernel_res(&qp->ib_qp.res)) { flags = bnxt_re_lock_cqs(qp); @@ -816,20 +994,33 @@ int bnxt_re_destroy_qp(struct ib_qp *ib_qp, struct ib_udata *udata) bnxt_qplib_free_qp_res(&rdev->qplib_res, &qp->qplib_qp); - if (ib_qp->qp_type == IB_QPT_GSI && rdev->gsi_ctx.gsi_sqp) { - rc = bnxt_re_destroy_gsi_sqp(qp); - if (rc) - return rc; - } + if (ib_qp->qp_type == IB_QPT_GSI && rdev->gsi_ctx.gsi_sqp) + bnxt_re_destroy_gsi_sqp(qp); mutex_lock(&rdev->qp_lock); list_del(&qp->list); mutex_unlock(&rdev->qp_lock); - atomic_dec(&rdev->qp_count); + atomic_dec(&rdev->stats.res.qp_count); + if (qp->qplib_qp.type == CMDQ_CREATE_QP_TYPE_RC) + atomic_dec(&rdev->stats.res.rc_qp_count); + else if (qp->qplib_qp.type == CMDQ_CREATE_QP_TYPE_UD) + atomic_dec(&rdev->stats.res.ud_qp_count); + + if (qp->qplib_qp.type == CMDQ_CREATE_QP_TYPE_RAW_ETHERTYPE) + bnxt_re_del_unique_gid(rdev); ib_umem_release(qp->rumem); ib_umem_release(qp->sumem); + /* Flush all the entries of notification queue associated with + * given qp. + */ + scq_nq = qplib_qp->scq->nq; + rcq_nq = qplib_qp->rcq->nq; + bnxt_re_synchronize_nq(scq_nq); + if (scq_nq != rcq_nq) + bnxt_re_synchronize_nq(rcq_nq); + return 0; } @@ -842,6 +1033,8 @@ static u8 __from_ib_qp_type(enum ib_qp_type type) return CMDQ_CREATE_QP_TYPE_RC; case IB_QPT_UD: return CMDQ_CREATE_QP_TYPE_UD; + case IB_QPT_RAW_PACKET: + return CMDQ_CREATE_QP_TYPE_RAW_ETHERTYPE; default: return IB_QPT_MAX; } @@ -880,64 +1073,63 @@ static int bnxt_re_setup_swqe_size(struct bnxt_re_qp *qp, rdev = qp->rdev; qplqp = &qp->qplib_qp; sq = &qplqp->sq; - dev_attr = &rdev->dev_attr; + dev_attr = rdev->dev_attr; align = sizeof(struct sq_send_hdr); ilsize = ALIGN(init_attr->cap.max_inline_data, align); - sq->wqe_size = bnxt_re_get_wqe_size(ilsize, sq->max_sge); - if (sq->wqe_size > bnxt_re_get_swqe_size(dev_attr->max_qp_sges)) - return -EINVAL; - /* For gen p4 and gen p5 backward compatibility mode - * wqe size is fixed to 128 bytes + /* For gen p4 and gen p5 fixed wqe compatibility mode + * wqe size is fixed to 128 bytes - ie 6 SGEs */ - if (sq->wqe_size < bnxt_re_get_swqe_size(dev_attr->max_qp_sges) && - qplqp->wqe_mode == BNXT_QPLIB_WQE_MODE_STATIC) - sq->wqe_size = bnxt_re_get_swqe_size(dev_attr->max_qp_sges); + if (qplqp->wqe_mode == BNXT_QPLIB_WQE_MODE_STATIC) { + sq->wqe_size = bnxt_re_get_swqe_size(BNXT_STATIC_MAX_SGE); + sq->max_sge = BNXT_STATIC_MAX_SGE; + } else { + sq->wqe_size = bnxt_re_get_wqe_size(ilsize, sq->max_sge); + if (sq->wqe_size > bnxt_re_get_swqe_size(dev_attr->max_qp_sges)) + return -EINVAL; + } if (init_attr->cap.max_inline_data) { qplqp->max_inline_data = sq->wqe_size - sizeof(struct sq_send_hdr); init_attr->cap.max_inline_data = qplqp->max_inline_data; - if (qplqp->wqe_mode == BNXT_QPLIB_WQE_MODE_STATIC) - sq->max_sge = qplqp->max_inline_data / - sizeof(struct sq_sge); } return 0; } static int bnxt_re_init_user_qp(struct bnxt_re_dev *rdev, struct bnxt_re_pd *pd, - struct bnxt_re_qp *qp, struct ib_udata *udata) + struct bnxt_re_qp *qp, struct bnxt_re_ucontext *cntx, + struct bnxt_re_qp_req *ureq) { struct bnxt_qplib_qp *qplib_qp; - struct bnxt_re_ucontext *cntx; - struct bnxt_re_qp_req ureq; int bytes = 0, psn_sz; struct ib_umem *umem; int psn_nume; qplib_qp = &qp->qplib_qp; - cntx = rdma_udata_to_drv_context(udata, struct bnxt_re_ucontext, - ib_uctx); - if (ib_copy_from_udata(&ureq, udata, sizeof(ureq))) - return -EFAULT; bytes = (qplib_qp->sq.max_wqe * qplib_qp->sq.wqe_size); /* Consider mapping PSN search memory only for RC QPs. */ if (qplib_qp->type == CMDQ_CREATE_QP_TYPE_RC) { - psn_sz = bnxt_qplib_is_chip_gen_p5(rdev->chip_ctx) ? + psn_sz = bnxt_qplib_is_chip_gen_p5_p7(rdev->chip_ctx) ? sizeof(struct sq_psn_search_ext) : sizeof(struct sq_psn_search); - psn_nume = (qplib_qp->wqe_mode == BNXT_QPLIB_WQE_MODE_STATIC) ? - qplib_qp->sq.max_wqe : - ((qplib_qp->sq.max_wqe * qplib_qp->sq.wqe_size) / - sizeof(struct bnxt_qplib_sge)); + if (cntx && bnxt_re_is_var_size_supported(rdev, cntx)) { + psn_nume = ureq->sq_slots; + } else { + psn_nume = (qplib_qp->wqe_mode == BNXT_QPLIB_WQE_MODE_STATIC) ? + qplib_qp->sq.max_wqe : ((qplib_qp->sq.max_wqe * qplib_qp->sq.wqe_size) / + sizeof(struct bnxt_qplib_sge)); + } + if (_is_host_msn_table(rdev->qplib_res.dattr->dev_cap_flags2)) + psn_nume = roundup_pow_of_two(psn_nume); bytes += (psn_nume * psn_sz); } bytes = PAGE_ALIGN(bytes); - umem = ib_umem_get(&rdev->ibdev, ureq.qpsva, bytes, + umem = ib_umem_get(&rdev->ibdev, ureq->qpsva, bytes, IB_ACCESS_LOCAL_WRITE); if (IS_ERR(umem)) return PTR_ERR(umem); @@ -946,12 +1138,12 @@ static int bnxt_re_init_user_qp(struct bnxt_re_dev *rdev, struct bnxt_re_pd *pd, qplib_qp->sq.sg_info.umem = umem; qplib_qp->sq.sg_info.pgsize = PAGE_SIZE; qplib_qp->sq.sg_info.pgshft = PAGE_SHIFT; - qplib_qp->qp_handle = ureq.qp_handle; + qplib_qp->qp_handle = ureq->qp_handle; if (!qp->qplib_qp.srq) { bytes = (qplib_qp->rq.max_wqe * qplib_qp->rq.wqe_size); bytes = PAGE_ALIGN(bytes); - umem = ib_umem_get(&rdev->ibdev, ureq.qprva, bytes, + umem = ib_umem_get(&rdev->ibdev, ureq->qprva, bytes, IB_ACCESS_LOCAL_WRITE); if (IS_ERR(umem)) goto rqfail; @@ -1010,7 +1202,7 @@ static struct bnxt_re_ah *bnxt_re_create_shadow_qp_ah "Failed to allocate HW AH for Shadow QP"); goto fail; } - atomic_inc(&rdev->ah_count); + atomic_inc(&rdev->stats.res.ah_count); return ah; @@ -1047,6 +1239,7 @@ static struct bnxt_re_qp *bnxt_re_create_shadow_qp /* Shadow QP SQ depth should be same as QP1 RQ depth */ qp->qplib_qp.sq.wqe_size = bnxt_re_get_wqe_size(0, 6); qp->qplib_qp.sq.max_wqe = qp1_qp->rq.max_wqe; + qp->qplib_qp.sq.max_sw_wqe = qp1_qp->rq.max_wqe; qp->qplib_qp.sq.max_sge = 2; /* Q full delta can be 1 since it is internal QP */ qp->qplib_qp.sq.q_full_delta = 1; @@ -1058,6 +1251,7 @@ static struct bnxt_re_qp *bnxt_re_create_shadow_qp qp->qplib_qp.rq.wqe_size = bnxt_re_get_rwqe_size(6); qp->qplib_qp.rq.max_wqe = qp1_qp->rq.max_wqe; + qp->qplib_qp.rq.max_sw_wqe = qp1_qp->rq.max_wqe; qp->qplib_qp.rq.max_sge = qp1_qp->rq.max_sge; /* Q full delta can be 1 since it is internal QP */ qp->qplib_qp.rq.q_full_delta = 1; @@ -1078,7 +1272,7 @@ static struct bnxt_re_qp *bnxt_re_create_shadow_qp INIT_LIST_HEAD(&qp->list); mutex_lock(&rdev->qp_lock); list_add_tail(&qp->list, &rdev->qp_list); - atomic_inc(&rdev->qp_count); + atomic_inc(&rdev->stats.res.qp_count); mutex_unlock(&rdev->qp_lock); return qp; fail: @@ -1087,7 +1281,8 @@ fail: } static int bnxt_re_init_rq_attr(struct bnxt_re_qp *qp, - struct ib_qp_init_attr *init_attr) + struct ib_qp_init_attr *init_attr, + struct bnxt_re_ucontext *uctx) { struct bnxt_qplib_dev_attr *dev_attr; struct bnxt_qplib_qp *qplqp; @@ -1098,7 +1293,7 @@ static int bnxt_re_init_rq_attr(struct bnxt_re_qp *qp, rdev = qp->rdev; qplqp = &qp->qplib_qp; rq = &qplqp->rq; - dev_attr = &rdev->dev_attr; + dev_attr = rdev->dev_attr; if (init_attr->srq) { struct bnxt_re_srq *srq; @@ -1116,8 +1311,9 @@ static int bnxt_re_init_rq_attr(struct bnxt_re_qp *qp, /* Allocate 1 more than what's provided so posting max doesn't * mean empty. */ - entries = roundup_pow_of_two(init_attr->cap.max_recv_wr + 1); + entries = bnxt_re_init_depth(init_attr->cap.max_recv_wr + 1, uctx); rq->max_wqe = min_t(u32, entries, dev_attr->max_qp_wqes + 1); + rq->max_sw_wqe = rq->max_wqe; rq->q_full_delta = 0; rq->sg_info.pgsize = PAGE_SIZE; rq->sg_info.pgshft = PAGE_SHIFT; @@ -1134,9 +1330,9 @@ static void bnxt_re_adjust_gsi_rq_attr(struct bnxt_re_qp *qp) rdev = qp->rdev; qplqp = &qp->qplib_qp; - dev_attr = &rdev->dev_attr; + dev_attr = rdev->dev_attr; - if (!bnxt_qplib_is_chip_gen_p5(rdev->chip_ctx)) { + if (!bnxt_qplib_is_chip_gen_p5_p7(rdev->chip_ctx)) { qplqp->rq.max_sge = dev_attr->max_qp_sges; if (qplqp->rq.max_sge > dev_attr->max_qp_sges) qplqp->rq.max_sge = dev_attr->max_qp_sges; @@ -1146,37 +1342,49 @@ static void bnxt_re_adjust_gsi_rq_attr(struct bnxt_re_qp *qp) static int bnxt_re_init_sq_attr(struct bnxt_re_qp *qp, struct ib_qp_init_attr *init_attr, - struct ib_udata *udata) + struct bnxt_re_ucontext *uctx, + struct bnxt_re_qp_req *ureq) { struct bnxt_qplib_dev_attr *dev_attr; struct bnxt_qplib_qp *qplqp; struct bnxt_re_dev *rdev; struct bnxt_qplib_q *sq; + int diff = 0; int entries; - int diff; int rc; rdev = qp->rdev; qplqp = &qp->qplib_qp; sq = &qplqp->sq; - dev_attr = &rdev->dev_attr; + dev_attr = rdev->dev_attr; sq->max_sge = init_attr->cap.max_send_sge; - if (sq->max_sge > dev_attr->max_qp_sges) { - sq->max_sge = dev_attr->max_qp_sges; - init_attr->cap.max_send_sge = sq->max_sge; - } + entries = init_attr->cap.max_send_wr; + if (uctx && qplqp->wqe_mode == BNXT_QPLIB_WQE_MODE_VARIABLE) { + sq->max_wqe = ureq->sq_slots; + sq->max_sw_wqe = ureq->sq_slots; + sq->wqe_size = sizeof(struct sq_sge); + } else { + if (sq->max_sge > dev_attr->max_qp_sges) { + sq->max_sge = dev_attr->max_qp_sges; + init_attr->cap.max_send_sge = sq->max_sge; + } - rc = bnxt_re_setup_swqe_size(qp, init_attr); - if (rc) - return rc; + rc = bnxt_re_setup_swqe_size(qp, init_attr); + if (rc) + return rc; - entries = init_attr->cap.max_send_wr; - /* Allocate 128 + 1 more than what's provided */ - diff = (qplqp->wqe_mode == BNXT_QPLIB_WQE_MODE_VARIABLE) ? - 0 : BNXT_QPLIB_RESERVED_QP_WRS; - entries = roundup_pow_of_two(entries + diff + 1); - sq->max_wqe = min_t(u32, entries, dev_attr->max_qp_wqes + diff + 1); + /* Allocate 128 + 1 more than what's provided */ + diff = (qplqp->wqe_mode == BNXT_QPLIB_WQE_MODE_VARIABLE) ? + 0 : BNXT_QPLIB_RESERVED_QP_WRS; + entries = bnxt_re_init_depth(entries + diff + 1, uctx); + sq->max_wqe = min_t(u32, entries, dev_attr->max_qp_wqes + diff + 1); + if (qplqp->wqe_mode == BNXT_QPLIB_WQE_MODE_VARIABLE) + sq->max_sw_wqe = bnxt_qplib_get_depth(sq, qplqp->wqe_mode, true); + else + sq->max_sw_wqe = sq->max_wqe; + + } sq->q_full_delta = diff + 1; /* * Reserving one slot for Phantom WQE. Application can @@ -1191,7 +1399,8 @@ static int bnxt_re_init_sq_attr(struct bnxt_re_qp *qp, } static void bnxt_re_adjust_gsi_sq_attr(struct bnxt_re_qp *qp, - struct ib_qp_init_attr *init_attr) + struct ib_qp_init_attr *init_attr, + struct bnxt_re_ucontext *uctx) { struct bnxt_qplib_dev_attr *dev_attr; struct bnxt_qplib_qp *qplqp; @@ -1200,10 +1409,10 @@ static void bnxt_re_adjust_gsi_sq_attr(struct bnxt_re_qp *qp, rdev = qp->rdev; qplqp = &qp->qplib_qp; - dev_attr = &rdev->dev_attr; + dev_attr = rdev->dev_attr; - if (!bnxt_qplib_is_chip_gen_p5(rdev->chip_ctx)) { - entries = roundup_pow_of_two(init_attr->cap.max_send_wr + 1); + if (!bnxt_qplib_is_chip_gen_p5_p7(rdev->chip_ctx)) { + entries = bnxt_re_init_depth(init_attr->cap.max_send_wr + 1, uctx); qplqp->sq.max_wqe = min_t(u32, entries, dev_attr->max_qp_wqes + 1); qplqp->sq.q_full_delta = qplqp->sq.max_wqe - @@ -1229,7 +1438,7 @@ static int bnxt_re_init_qp_type(struct bnxt_re_dev *rdev, goto out; } - if (bnxt_qplib_is_chip_gen_p5(chip_ctx) && + if (bnxt_qplib_is_chip_gen_p5_p7(chip_ctx) && init_attr->qp_type == IB_QPT_GSI) qptype = CMDQ_CREATE_QP_TYPE_GSI; out: @@ -1238,7 +1447,8 @@ out: static int bnxt_re_init_qp_attr(struct bnxt_re_qp *qp, struct bnxt_re_pd *pd, struct ib_qp_init_attr *init_attr, - struct ib_udata *udata) + struct bnxt_re_ucontext *uctx, + struct bnxt_re_qp_req *ureq) { struct bnxt_qplib_dev_attr *dev_attr; struct bnxt_qplib_qp *qplqp; @@ -1248,23 +1458,21 @@ static int bnxt_re_init_qp_attr(struct bnxt_re_qp *qp, struct bnxt_re_pd *pd, rdev = qp->rdev; qplqp = &qp->qplib_qp; - dev_attr = &rdev->dev_attr; + dev_attr = rdev->dev_attr; /* Setup misc params */ ether_addr_copy(qplqp->smac, rdev->netdev->dev_addr); qplqp->pd = &pd->qplib_pd; qplqp->qp_handle = (u64)qplqp; qplqp->max_inline_data = init_attr->cap.max_inline_data; - qplqp->sig_type = ((init_attr->sq_sig_type == IB_SIGNAL_ALL_WR) ? - true : false); + qplqp->sig_type = init_attr->sq_sig_type == IB_SIGNAL_ALL_WR; qptype = bnxt_re_init_qp_type(rdev, init_attr); if (qptype < 0) { rc = qptype; goto out; } qplqp->type = (u8)qptype; - qplqp->wqe_mode = rdev->chip_ctx->modes.wqe_mode; - + qplqp->wqe_mode = bnxt_re_is_var_size_supported(rdev, uctx); if (init_attr->qp_type == IB_QPT_RC) { qplqp->max_rd_atomic = dev_attr->max_qp_rd_atom; qplqp->max_dest_rd_atomic = dev_attr->max_qp_init_rd_atom; @@ -1292,21 +1500,21 @@ static int bnxt_re_init_qp_attr(struct bnxt_re_qp *qp, struct bnxt_re_pd *pd, } /* Setup RQ/SRQ */ - rc = bnxt_re_init_rq_attr(qp, init_attr); + rc = bnxt_re_init_rq_attr(qp, init_attr, uctx); if (rc) goto out; if (init_attr->qp_type == IB_QPT_GSI) bnxt_re_adjust_gsi_rq_attr(qp); /* Setup SQ */ - rc = bnxt_re_init_sq_attr(qp, init_attr, udata); + rc = bnxt_re_init_sq_attr(qp, init_attr, uctx, ureq); if (rc) goto out; if (init_attr->qp_type == IB_QPT_GSI) - bnxt_re_adjust_gsi_sq_attr(qp, init_attr); + bnxt_re_adjust_gsi_sq_attr(qp, init_attr, uctx); - if (udata) /* This will update DPI and qp_handle */ - rc = bnxt_re_init_user_qp(rdev, pd, qp, udata); + if (uctx) /* This will update DPI and qp_handle */ + rc = bnxt_re_init_user_qp(rdev, pd, qp, uctx, ureq); out: return rc; } @@ -1361,7 +1569,7 @@ static int bnxt_re_create_gsi_qp(struct bnxt_re_qp *qp, struct bnxt_re_pd *pd, { struct bnxt_re_dev *rdev; struct bnxt_qplib_qp *qplqp; - int rc = 0; + int rc; rdev = qp->rdev; qplqp = &qp->qplib_qp; @@ -1404,16 +1612,53 @@ static bool bnxt_re_test_qp_limits(struct bnxt_re_dev *rdev, return rc; } +static int bnxt_re_add_unique_gid(struct bnxt_re_dev *rdev) +{ + struct bnxt_qplib_ctx *hctx = &rdev->qplib_ctx; + struct bnxt_qplib_res *res = &rdev->qplib_res; + int rc; + + if (!rdev->rcfw.roce_mirror) + return 0; + + rdev->ugid.global.subnet_prefix = cpu_to_be64(0xfe8000000000abcdLL); + addrconf_ifid_eui48(&rdev->ugid.raw[8], rdev->netdev); + + rc = bnxt_qplib_add_sgid(&res->sgid_tbl, + (struct bnxt_qplib_gid *)&rdev->ugid, + rdev->qplib_res.netdev->dev_addr, + 0xFFFF, true, &rdev->ugid_index, true, + hctx->stats3.fw_id); + if (rc) + dev_err(rdev_to_dev(rdev), "Failed to add unique GID. rc = %d\n", rc); + + return rc; +} + int bnxt_re_create_qp(struct ib_qp *ib_qp, struct ib_qp_init_attr *qp_init_attr, struct ib_udata *udata) { - struct ib_pd *ib_pd = ib_qp->pd; - struct bnxt_re_pd *pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); - struct bnxt_re_dev *rdev = pd->rdev; - struct bnxt_qplib_dev_attr *dev_attr = &rdev->dev_attr; - struct bnxt_re_qp *qp = container_of(ib_qp, struct bnxt_re_qp, ib_qp); + struct bnxt_qplib_dev_attr *dev_attr; + struct bnxt_re_ucontext *uctx; + struct bnxt_re_qp_req ureq; + struct bnxt_re_dev *rdev; + struct bnxt_re_pd *pd; + struct bnxt_re_qp *qp; + struct ib_pd *ib_pd; + u32 active_qps; int rc; + ib_pd = ib_qp->pd; + pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); + rdev = pd->rdev; + dev_attr = rdev->dev_attr; + qp = container_of(ib_qp, struct bnxt_re_qp, ib_qp); + + uctx = rdma_udata_to_drv_context(udata, struct bnxt_re_ucontext, ib_uctx); + if (udata) + if (ib_copy_from_udata(&ureq, udata, min(udata->inlen, sizeof(ureq)))) + return -EFAULT; + rc = bnxt_re_test_qp_limits(rdev, qp_init_attr, dev_attr); if (!rc) { rc = -EINVAL; @@ -1421,12 +1666,12 @@ int bnxt_re_create_qp(struct ib_qp *ib_qp, struct ib_qp_init_attr *qp_init_attr, } qp->rdev = rdev; - rc = bnxt_re_init_qp_attr(qp, pd, qp_init_attr, udata); + rc = bnxt_re_init_qp_attr(qp, pd, qp_init_attr, uctx, &ureq); if (rc) goto fail; if (qp_init_attr->qp_type == IB_QPT_GSI && - !(bnxt_qplib_is_chip_gen_p5(rdev->chip_ctx))) { + !(bnxt_qplib_is_chip_gen_p5_p7(rdev->chip_ctx))) { rc = bnxt_re_create_gsi_qp(qp, pd, qp_init_attr); if (rc == -ENODEV) goto qp_destroy; @@ -1451,6 +1696,17 @@ int bnxt_re_create_qp(struct ib_qp *ib_qp, struct ib_qp_init_attr *qp_init_attr, } } + /* Support for RawEth QP is added to capture TCP pkt dump. + * So unique SGID is used to avoid incorrect statistics on per + * function stats_ctx + */ + if (qp->qplib_qp.type == CMDQ_CREATE_QP_TYPE_RAW_ETHERTYPE) { + rc = bnxt_re_add_unique_gid(rdev); + if (rc) + goto qp_destroy; + qp->qplib_qp.ugid_index = rdev->ugid_index; + } + qp->ib_qp.qp_num = qp->qplib_qp.id; if (qp_init_attr->qp_type == IB_QPT_GSI) rdev->gsi_ctx.gsi_qp = qp; @@ -1460,7 +1716,19 @@ int bnxt_re_create_qp(struct ib_qp *ib_qp, struct ib_qp_init_attr *qp_init_attr, mutex_lock(&rdev->qp_lock); list_add_tail(&qp->list, &rdev->qp_list); mutex_unlock(&rdev->qp_lock); - atomic_inc(&rdev->qp_count); + active_qps = atomic_inc_return(&rdev->stats.res.qp_count); + if (active_qps > rdev->stats.res.qp_watermark) + rdev->stats.res.qp_watermark = active_qps; + if (qp_init_attr->qp_type == IB_QPT_RC) { + active_qps = atomic_inc_return(&rdev->stats.res.rc_qp_count); + if (active_qps > rdev->stats.res.rc_qp_watermark) + rdev->stats.res.rc_qp_watermark = active_qps; + } else if (qp_init_attr->qp_type == IB_QPT_UD) { + active_qps = atomic_inc_return(&rdev->stats.res.ud_qp_count); + if (active_qps > rdev->stats.res.ud_qp_watermark) + rdev->stats.res.ud_qp_watermark = active_qps; + } + bnxt_re_debug_add_qpinfo(rdev, qp); return 0; qp_destroy: @@ -1557,15 +1825,14 @@ int bnxt_re_destroy_srq(struct ib_srq *ib_srq, struct ib_udata *udata) ib_srq); struct bnxt_re_dev *rdev = srq->rdev; struct bnxt_qplib_srq *qplib_srq = &srq->qplib_srq; - struct bnxt_qplib_nq *nq = NULL; - if (qplib_srq->cq) - nq = qplib_srq->cq->nq; + if (rdev->chip_ctx->modes.toggle_bits & BNXT_QPLIB_SRQ_TOGGLE_BIT) { + free_page((unsigned long)srq->uctx_srq_page); + hash_del(&srq->hash_entry); + } bnxt_qplib_destroy_srq(&rdev->qplib_res, qplib_srq); ib_umem_release(srq->umem); - atomic_dec(&rdev->srq_count); - if (nq) - nq->budget--; + atomic_dec(&rdev->stats.res.srq_count); return 0; } @@ -1606,17 +1873,18 @@ int bnxt_re_create_srq(struct ib_srq *ib_srq, struct ib_udata *udata) { struct bnxt_qplib_dev_attr *dev_attr; - struct bnxt_qplib_nq *nq = NULL; + struct bnxt_re_ucontext *uctx; struct bnxt_re_dev *rdev; struct bnxt_re_srq *srq; struct bnxt_re_pd *pd; struct ib_pd *ib_pd; + u32 active_srqs; int rc, entries; ib_pd = ib_srq->pd; pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); rdev = pd->rdev; - dev_attr = &rdev->dev_attr; + dev_attr = rdev->dev_attr; srq = container_of(ib_srq, struct bnxt_re_srq, ib_srq); if (srq_init_attr->attr.max_wr >= dev_attr->max_srq_wqes) { @@ -1630,13 +1898,14 @@ int bnxt_re_create_srq(struct ib_srq *ib_srq, goto exit; } + uctx = rdma_udata_to_drv_context(udata, struct bnxt_re_ucontext, ib_uctx); srq->rdev = rdev; srq->qplib_srq.pd = &pd->qplib_pd; srq->qplib_srq.dpi = &rdev->dpi_privileged; /* Allocate 1 more than what's provided so posting max doesn't * mean empty */ - entries = roundup_pow_of_two(srq_init_attr->attr.max_wr + 1); + entries = bnxt_re_init_depth(srq_init_attr->attr.max_wr + 1, uctx); if (entries > dev_attr->max_srq_wqes + 1) entries = dev_attr->max_srq_wqes + 1; srq->qplib_srq.max_wqe = entries; @@ -1646,8 +1915,9 @@ int bnxt_re_create_srq(struct ib_srq *ib_srq, srq->qplib_srq.wqe_size = bnxt_re_get_rwqe_size(dev_attr->max_srq_sges); srq->qplib_srq.threshold = srq_init_attr->attr.srq_limit; srq->srq_limit = srq_init_attr->attr.srq_limit; - srq->qplib_srq.eventq_hw_ring_id = rdev->nq[0].ring_id; - nq = &rdev->nq[0]; + srq->qplib_srq.eventq_hw_ring_id = rdev->nqr->nq[0].ring_id; + srq->qplib_srq.sg_info.pgsize = PAGE_SIZE; + srq->qplib_srq.sg_info.pgshft = PAGE_SHIFT; if (udata) { rc = bnxt_re_init_user_srq(rdev, pd, srq, udata); @@ -1662,9 +1932,18 @@ int bnxt_re_create_srq(struct ib_srq *ib_srq, } if (udata) { - struct bnxt_re_srq_resp resp; + struct bnxt_re_srq_resp resp = {}; resp.srqid = srq->qplib_srq.id; + if (rdev->chip_ctx->modes.toggle_bits & BNXT_QPLIB_SRQ_TOGGLE_BIT) { + hash_add(rdev->srq_hash, &srq->hash_entry, srq->qplib_srq.id); + srq->uctx_srq_page = (void *)get_zeroed_page(GFP_KERNEL); + if (!srq->uctx_srq_page) { + rc = -ENOMEM; + goto fail; + } + resp.comp_mask |= BNXT_RE_SRQ_TOGGLE_PAGE_SUPPORT; + } rc = ib_copy_to_udata(udata, &resp, sizeof(resp)); if (rc) { ibdev_err(&rdev->ibdev, "SRQ copy to udata failed!"); @@ -1673,9 +1952,9 @@ int bnxt_re_create_srq(struct ib_srq *ib_srq, goto fail; } } - if (nq) - nq->budget++; - atomic_inc(&rdev->srq_count); + active_srqs = atomic_inc_return(&rdev->stats.res.srq_count); + if (active_srqs > rdev->stats.res.srq_watermark) + rdev->stats.res.srq_watermark = active_srqs; spin_lock_init(&srq->lock); return 0; @@ -1693,33 +1972,28 @@ int bnxt_re_modify_srq(struct ib_srq *ib_srq, struct ib_srq_attr *srq_attr, struct bnxt_re_srq *srq = container_of(ib_srq, struct bnxt_re_srq, ib_srq); struct bnxt_re_dev *rdev = srq->rdev; - int rc; switch (srq_attr_mask) { case IB_SRQ_MAX_WR: /* SRQ resize is not supported */ - break; + return -EINVAL; case IB_SRQ_LIMIT: /* Change the SRQ threshold */ if (srq_attr->srq_limit > srq->qplib_srq.max_wqe) return -EINVAL; srq->qplib_srq.threshold = srq_attr->srq_limit; - rc = bnxt_qplib_modify_srq(&rdev->qplib_res, &srq->qplib_srq); - if (rc) { - ibdev_err(&rdev->ibdev, "Modify HW SRQ failed!"); - return rc; - } + bnxt_qplib_srq_arm_db(&srq->qplib_srq.dbinfo, srq->qplib_srq.threshold); + /* On success, update the shadow */ srq->srq_limit = srq_attr->srq_limit; /* No need to Build and send response back to udata */ - break; + return 0; default: ibdev_err(&rdev->ibdev, "Unsupported srq_attr_mask 0x%x", srq_attr_mask); return -EINVAL; } - return 0; } int bnxt_re_query_srq(struct ib_srq *ib_srq, struct ib_srq_attr *srq_attr) @@ -1777,7 +2051,7 @@ static int bnxt_re_modify_shadow_qp(struct bnxt_re_dev *rdev, int qp_attr_mask) { struct bnxt_re_qp *qp = rdev->gsi_ctx.gsi_sqp; - int rc = 0; + int rc; if (qp_attr_mask & IB_QP_STATE) { qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_STATE; @@ -1809,7 +2083,7 @@ int bnxt_re_modify_qp(struct ib_qp *ib_qp, struct ib_qp_attr *qp_attr, { struct bnxt_re_qp *qp = container_of(ib_qp, struct bnxt_re_qp, ib_qp); struct bnxt_re_dev *rdev = qp->rdev; - struct bnxt_qplib_dev_attr *dev_attr = &rdev->dev_attr; + struct bnxt_qplib_dev_attr *dev_attr = rdev->dev_attr; enum ib_qp_state curr_qp_state, new_qp_state; int rc, entries; unsigned int flags; @@ -1863,12 +2137,10 @@ int bnxt_re_modify_qp(struct ib_qp *ib_qp, struct ib_qp_attr *qp_attr, if (qp_attr_mask & IB_QP_ACCESS_FLAGS) { qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_ACCESS; qp->qplib_qp.access = - __from_ib_access_flags(qp_attr->qp_access_flags); + __qp_access_flags_from_ib(qp->qplib_qp.cctx, + qp_attr->qp_access_flags); /* LOCAL_WRITE access must be set to allow RC receive */ - qp->qplib_qp.access |= BNXT_QPLIB_ACCESS_LOCAL_WRITE; - /* Temp: Set all params on QP as of now */ - qp->qplib_qp.access |= CMDQ_MODIFY_QP_ACCESS_REMOTE_WRITE; - qp->qplib_qp.access |= CMDQ_MODIFY_QP_ACCESS_REMOTE_READ; + qp->qplib_qp.access |= CMDQ_MODIFY_QP_ACCESS_LOCAL_WRITE; } if (qp_attr_mask & IB_QP_PKEY_INDEX) { qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_PKEY; @@ -1902,7 +2174,7 @@ int bnxt_re_modify_qp(struct ib_qp *ib_qp, struct ib_qp_attr *qp_attr, qp->qplib_qp.ah.sgid_index = ctx->idx; qp->qplib_qp.ah.host_sgid_index = grh->sgid_index; qp->qplib_qp.ah.hop_limit = grh->hop_limit; - qp->qplib_qp.ah.traffic_class = grh->traffic_class; + qp->qplib_qp.ah.traffic_class = grh->traffic_class >> 2; qp->qplib_qp.ah.sl = rdma_ah_get_sl(&qp_attr->ah_attr); ether_addr_copy(qp->qplib_qp.ah.dmac, qp_attr->ah_attr.roce.dmac); @@ -1929,18 +2201,20 @@ int bnxt_re_modify_qp(struct ib_qp *ib_qp, struct ib_qp_attr *qp_attr, } } - if (qp_attr_mask & IB_QP_PATH_MTU) { - qp->qplib_qp.modify_flags |= - CMDQ_MODIFY_QP_MODIFY_MASK_PATH_MTU; - qp->qplib_qp.path_mtu = __from_ib_mtu(qp_attr->path_mtu); - qp->qplib_qp.mtu = ib_mtu_enum_to_int(qp_attr->path_mtu); - } else if (qp_attr->qp_state == IB_QPS_RTR) { - qp->qplib_qp.modify_flags |= - CMDQ_MODIFY_QP_MODIFY_MASK_PATH_MTU; - qp->qplib_qp.path_mtu = - __from_ib_mtu(iboe_get_mtu(rdev->netdev->mtu)); - qp->qplib_qp.mtu = - ib_mtu_enum_to_int(iboe_get_mtu(rdev->netdev->mtu)); + if (qp_attr->qp_state == IB_QPS_RTR) { + enum ib_mtu qpmtu; + + qpmtu = iboe_get_mtu(rdev->netdev->mtu); + if (qp_attr_mask & IB_QP_PATH_MTU) { + if (ib_mtu_enum_to_int(qp_attr->path_mtu) > + ib_mtu_enum_to_int(qpmtu)) + return -EINVAL; + qpmtu = qp_attr->path_mtu; + } + + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_PATH_MTU; + qp->qplib_qp.path_mtu = __from_ib_mtu(qpmtu); + qp->qplib_qp.mtu = ib_mtu_enum_to_int(qpmtu); } if (qp_attr_mask & IB_QP_TIMEOUT) { @@ -1992,6 +2266,9 @@ int bnxt_re_modify_qp(struct ib_qp *ib_qp, struct ib_qp_attr *qp_attr, qp->qplib_qp.max_dest_rd_atomic = qp_attr->max_dest_rd_atomic; } if (qp_attr_mask & IB_QP_CAP) { + struct bnxt_re_ucontext *uctx = + rdma_udata_to_drv_context(udata, struct bnxt_re_ucontext, ib_uctx); + qp->qplib_qp.modify_flags |= CMDQ_MODIFY_QP_MODIFY_MASK_SQ_SIZE | CMDQ_MODIFY_QP_MODIFY_MASK_RQ_SIZE | @@ -2008,7 +2285,7 @@ int bnxt_re_modify_qp(struct ib_qp *ib_qp, struct ib_qp_attr *qp_attr, "Create QP failed - max exceeded"); return -EINVAL; } - entries = roundup_pow_of_two(qp_attr->cap.max_send_wr); + entries = bnxt_re_init_depth(qp_attr->cap.max_send_wr, uctx); qp->qplib_qp.sq.max_wqe = min_t(u32, entries, dev_attr->max_qp_wqes + 1); qp->qplib_qp.sq.q_full_delta = qp->qplib_qp.sq.max_wqe - @@ -2021,9 +2298,10 @@ int bnxt_re_modify_qp(struct ib_qp *ib_qp, struct ib_qp_attr *qp_attr, qp->qplib_qp.sq.q_full_delta -= 1; qp->qplib_qp.sq.max_sge = qp_attr->cap.max_send_sge; if (qp->qplib_qp.rq.max_wqe) { - entries = roundup_pow_of_two(qp_attr->cap.max_recv_wr); + entries = bnxt_re_init_depth(qp_attr->cap.max_recv_wr, uctx); qp->qplib_qp.rq.max_wqe = min_t(u32, entries, dev_attr->max_qp_wqes + 1); + qp->qplib_qp.rq.max_sw_wqe = qp->qplib_qp.rq.max_wqe; qp->qplib_qp.rq.q_full_delta = qp->qplib_qp.rq.max_wqe - qp_attr->cap.max_recv_wr; qp->qplib_qp.rq.max_sge = qp_attr->cap.max_recv_sge; @@ -2069,11 +2347,12 @@ int bnxt_re_query_qp(struct ib_qp *ib_qp, struct ib_qp_attr *qp_attr, qp_attr->qp_state = __to_ib_qp_state(qplib_qp->state); qp_attr->cur_qp_state = __to_ib_qp_state(qplib_qp->cur_qp_state); qp_attr->en_sqd_async_notify = qplib_qp->en_sqd_async_notify ? 1 : 0; - qp_attr->qp_access_flags = __to_ib_access_flags(qplib_qp->access); + qp_attr->qp_access_flags = __qp_access_flags_to_ib(qp->qplib_qp.cctx, + qplib_qp->access); qp_attr->pkey_index = qplib_qp->pkey_index; qp_attr->qkey = qplib_qp->qkey; qp_attr->ah_attr.type = RDMA_AH_ATTR_TYPE_ROCE; - rdma_ah_set_grh(&qp_attr->ah_attr, NULL, qplib_qp->ah.flow_label, + rdma_ah_set_grh(&qp_attr->ah_attr, NULL, qplib_qp->udp_sport, qplib_qp->ah.host_sgid_index, qplib_qp->ah.hop_limit, qplib_qp->ah.traffic_class); @@ -2085,6 +2364,7 @@ int bnxt_re_query_qp(struct ib_qp *ib_qp, struct ib_qp_attr *qp_attr, qp_attr->retry_cnt = qplib_qp->retry_cnt; qp_attr->rnr_retry = qplib_qp->rnr_retry; qp_attr->min_rnr_timer = qplib_qp->min_rnr_timer; + qp_attr->port_num = __to_ib_port_num(qplib_qp->port_id); qp_attr->rq_psn = qplib_qp->rq.psn; qp_attr->max_rd_atomic = qplib_qp->max_rd_atomic; qp_attr->sq_psn = qplib_qp->sq.psn; @@ -2127,7 +2407,7 @@ static int bnxt_re_build_qp1_send_v2(struct bnxt_re_qp *qp, u8 ip_version = 0; u16 vlan_id = 0xFFFF; void *buf; - int i, rc = 0; + int i, rc; memset(&qp->qp1_hdr, 0, sizeof(qp->qp1_hdr)); @@ -2165,7 +2445,7 @@ static int bnxt_re_build_qp1_send_v2(struct bnxt_re_qp *qp, } is_eth = true; - is_vlan = (vlan_id && (vlan_id < 0x1000)) ? true : false; + is_vlan = vlan_id && (vlan_id < 0x1000); ib_ud_header_init(payload_size, !is_eth, is_eth, is_vlan, is_grh, ip_version, is_udp, 0, &qp->qp1_hdr); @@ -2348,7 +2628,7 @@ static int bnxt_re_build_send_wqe(struct bnxt_re_qp *qp, break; case IB_WR_SEND_WITH_IMM: wqe->type = BNXT_QPLIB_SWQE_TYPE_SEND_WITH_IMM; - wqe->send.imm_data = wr->ex.imm_data; + wqe->send.imm_data = be32_to_cpu(wr->ex.imm_data); break; case IB_WR_SEND_WITH_INV: wqe->type = BNXT_QPLIB_SWQE_TYPE_SEND_WITH_INV; @@ -2378,7 +2658,7 @@ static int bnxt_re_build_rdma_wqe(const struct ib_send_wr *wr, break; case IB_WR_RDMA_WRITE_WITH_IMM: wqe->type = BNXT_QPLIB_SWQE_TYPE_RDMA_WRITE_WITH_IMM; - wqe->rdma.imm_data = wr->ex.imm_data; + wqe->rdma.imm_data = be32_to_cpu(wr->ex.imm_data); break; case IB_WR_RDMA_READ: wqe->type = BNXT_QPLIB_SWQE_TYPE_RDMA_READ; @@ -2434,11 +2714,6 @@ static int bnxt_re_build_inv_wqe(const struct ib_send_wr *wr, wqe->type = BNXT_QPLIB_SWQE_TYPE_LOCAL_INV; wqe->local_inv.inv_l_key = wr->ex.invalidate_rkey; - /* Need unconditional fence for local invalidate - * opcode to work as expected. - */ - wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_UC_FENCE; - if (wr->send_flags & IB_SEND_SIGNALED) wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SIGNAL_COMP; if (wr->send_flags & IB_SEND_SOLICITED) @@ -2461,12 +2736,6 @@ static int bnxt_re_build_reg_wqe(const struct ib_reg_wr *wr, wqe->frmr.levels = qplib_frpl->hwq.level; wqe->type = BNXT_QPLIB_SWQE_TYPE_REG_MR; - /* Need unconditional fence for reg_mr - * opcode to function as expected. - */ - - wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_UC_FENCE; - if (wr->wr.send_flags & IB_SEND_SIGNALED) wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_SIGNAL_COMP; @@ -2592,11 +2861,24 @@ bad: wr = wr->next; } bnxt_qplib_post_send_db(&qp->qplib_qp); - bnxt_ud_qp_hw_stall_workaround(qp); + if (!bnxt_qplib_is_chip_gen_p5_p7(qp->rdev->chip_ctx)) + bnxt_ud_qp_hw_stall_workaround(qp); spin_unlock_irqrestore(&qp->sq_lock, flags); return rc; } +static void bnxt_re_legacy_set_uc_fence(struct bnxt_qplib_swqe *wqe) +{ + /* Need unconditional fence for non-wire memory opcode + * to work as expected. + */ + if (wqe->type == BNXT_QPLIB_SWQE_TYPE_LOCAL_INV || + wqe->type == BNXT_QPLIB_SWQE_TYPE_FAST_REG_MR || + wqe->type == BNXT_QPLIB_SWQE_TYPE_REG_MR || + wqe->type == BNXT_QPLIB_SWQE_TYPE_BIND_MW) + wqe->flags |= BNXT_QPLIB_SWQE_FLAGS_UC_FENCE; +} + int bnxt_re_post_send(struct ib_qp *ib_qp, const struct ib_send_wr *wr, const struct ib_send_wr **bad_wr) { @@ -2676,8 +2958,11 @@ int bnxt_re_post_send(struct ib_qp *ib_qp, const struct ib_send_wr *wr, rc = -EINVAL; goto bad; } - if (!rc) + if (!rc) { + if (!bnxt_qplib_is_chip_gen_p5_p7(qp->rdev->chip_ctx)) + bnxt_re_legacy_set_uc_fence(&wqe); rc = bnxt_qplib_post_send(&qp->qplib_qp, &wqe); + } bad: if (rc) { ibdev_err(&qp->rdev->ibdev, @@ -2689,7 +2974,8 @@ bad: wr = wr->next; } bnxt_qplib_post_send_db(&qp->qplib_qp); - bnxt_ud_qp_hw_stall_workaround(qp); + if (!bnxt_qplib_is_chip_gen_p5_p7(qp->rdev->chip_ctx)) + bnxt_ud_qp_hw_stall_workaround(qp); spin_unlock_irqrestore(&qp->sq_lock, flags); return rc; @@ -2702,7 +2988,6 @@ static int bnxt_re_post_recv_shadow_qp(struct bnxt_re_dev *rdev, struct bnxt_qplib_swqe wqe; int rc = 0; - memset(&wqe, 0, sizeof(wqe)); while (wr) { /* House keeping */ memset(&wqe, 0, sizeof(wqe)); @@ -2787,36 +3072,68 @@ int bnxt_re_post_recv(struct ib_qp *ib_qp, const struct ib_recv_wr *wr, return rc; } +static struct bnxt_qplib_nq *bnxt_re_get_nq(struct bnxt_re_dev *rdev) +{ + int min, indx; + + mutex_lock(&rdev->nqr->load_lock); + for (indx = 0, min = 0; indx < (rdev->nqr->num_msix - 1); indx++) { + if (rdev->nqr->nq[min].load > rdev->nqr->nq[indx].load) + min = indx; + } + rdev->nqr->nq[min].load++; + mutex_unlock(&rdev->nqr->load_lock); + + return &rdev->nqr->nq[min]; +} + +static void bnxt_re_put_nq(struct bnxt_re_dev *rdev, struct bnxt_qplib_nq *nq) +{ + mutex_lock(&rdev->nqr->load_lock); + nq->load--; + mutex_unlock(&rdev->nqr->load_lock); +} + /* Completion Queues */ int bnxt_re_destroy_cq(struct ib_cq *ib_cq, struct ib_udata *udata) { - struct bnxt_re_cq *cq; + struct bnxt_qplib_chip_ctx *cctx; struct bnxt_qplib_nq *nq; struct bnxt_re_dev *rdev; + struct bnxt_re_cq *cq; cq = container_of(ib_cq, struct bnxt_re_cq, ib_cq); rdev = cq->rdev; nq = cq->qplib_cq.nq; + cctx = rdev->chip_ctx; + if (cctx->modes.toggle_bits & BNXT_QPLIB_CQ_TOGGLE_BIT) { + free_page((unsigned long)cq->uctx_cq_page); + hash_del(&cq->hash_entry); + } bnxt_qplib_destroy_cq(&rdev->qplib_res, &cq->qplib_cq); + + bnxt_re_put_nq(rdev, nq); ib_umem_release(cq->umem); - atomic_dec(&rdev->cq_count); - nq->budget--; + atomic_dec(&rdev->stats.res.cq_count); kfree(cq->cql); return 0; } int bnxt_re_create_cq(struct ib_cq *ibcq, const struct ib_cq_init_attr *attr, - struct ib_udata *udata) + struct uverbs_attr_bundle *attrs) { - struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibcq->device, ibdev); - struct bnxt_qplib_dev_attr *dev_attr = &rdev->dev_attr; struct bnxt_re_cq *cq = container_of(ibcq, struct bnxt_re_cq, ib_cq); - int rc, entries; + struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibcq->device, ibdev); + struct ib_udata *udata = &attrs->driver_udata; + struct bnxt_re_ucontext *uctx = + rdma_udata_to_drv_context(udata, struct bnxt_re_ucontext, ib_uctx); + struct bnxt_qplib_dev_attr *dev_attr = rdev->dev_attr; + struct bnxt_qplib_chip_ctx *cctx; int cqe = attr->cqe; - struct bnxt_qplib_nq *nq = NULL; - unsigned int nq_alloc_cnt; + int rc, entries; + u32 active_cqs; if (attr->flags) return -EOPNOTSUPP; @@ -2828,9 +3145,10 @@ int bnxt_re_create_cq(struct ib_cq *ibcq, const struct ib_cq_init_attr *attr, } cq->rdev = rdev; + cctx = rdev->chip_ctx; cq->qplib_cq.cq_handle = (u64)(unsigned long)(&cq->qplib_cq); - entries = roundup_pow_of_two(cqe + 1); + entries = bnxt_re_init_depth(cqe + 1, uctx); if (entries > dev_attr->max_cq_wqes + 1) entries = dev_attr->max_cq_wqes + 1; @@ -2838,8 +3156,6 @@ int bnxt_re_create_cq(struct ib_cq *ibcq, const struct ib_cq_init_attr *attr, cq->qplib_cq.sg_info.pgshft = PAGE_SHIFT; if (udata) { struct bnxt_re_cq_req req; - struct bnxt_re_ucontext *uctx = rdma_udata_to_drv_context( - udata, struct bnxt_re_ucontext, ib_uctx); if (ib_copy_from_udata(&req, udata, sizeof(req))) { rc = -EFAULT; goto fail; @@ -2865,15 +3181,10 @@ int bnxt_re_create_cq(struct ib_cq *ibcq, const struct ib_cq_init_attr *attr, cq->qplib_cq.dpi = &rdev->dpi_privileged; } - /* - * Allocating the NQ in a round robin fashion. nq_alloc_cnt is a - * used for getting the NQ index. - */ - nq_alloc_cnt = atomic_inc_return(&rdev->nq_alloc_cnt); - nq = &rdev->nq[nq_alloc_cnt % (rdev->num_msix - 1)]; cq->qplib_cq.max_wqe = entries; - cq->qplib_cq.cnq_hw_ring_id = nq->ring_id; - cq->qplib_cq.nq = nq; + cq->qplib_cq.coalescing = &rdev->cq_coalescing; + cq->qplib_cq.nq = bnxt_re_get_nq(rdev); + cq->qplib_cq.cnq_hw_ring_id = cq->qplib_cq.nq->ring_id; rc = bnxt_qplib_create_cq(&rdev->qplib_res, &cq->qplib_cq); if (rc) { @@ -2883,28 +3194,41 @@ int bnxt_re_create_cq(struct ib_cq *ibcq, const struct ib_cq_init_attr *attr, cq->ib_cq.cqe = entries; cq->cq_period = cq->qplib_cq.period; - nq->budget++; - atomic_inc(&rdev->cq_count); + active_cqs = atomic_inc_return(&rdev->stats.res.cq_count); + if (active_cqs > rdev->stats.res.cq_watermark) + rdev->stats.res.cq_watermark = active_cqs; spin_lock_init(&cq->cq_lock); if (udata) { - struct bnxt_re_cq_resp resp; + struct bnxt_re_cq_resp resp = {}; + if (cctx->modes.toggle_bits & BNXT_QPLIB_CQ_TOGGLE_BIT) { + hash_add(rdev->cq_hash, &cq->hash_entry, cq->qplib_cq.id); + /* Allocate a page */ + cq->uctx_cq_page = (void *)get_zeroed_page(GFP_KERNEL); + if (!cq->uctx_cq_page) { + rc = -ENOMEM; + goto c2fail; + } + resp.comp_mask |= BNXT_RE_CQ_TOGGLE_PAGE_SUPPORT; + } resp.cqid = cq->qplib_cq.id; resp.tail = cq->qplib_cq.hwq.cons; resp.phase = cq->qplib_cq.period; resp.rsvd = 0; - rc = ib_copy_to_udata(udata, &resp, sizeof(resp)); + rc = ib_copy_to_udata(udata, &resp, min(sizeof(resp), udata->outlen)); if (rc) { ibdev_err(&rdev->ibdev, "Failed to copy CQ udata"); bnxt_qplib_destroy_cq(&rdev->qplib_res, &cq->qplib_cq); - goto c2fail; + goto free_mem; } } return 0; +free_mem: + free_page((unsigned long)cq->uctx_cq_page); c2fail: ib_umem_release(cq->umem); fail: @@ -2940,7 +3264,7 @@ int bnxt_re_resize_cq(struct ib_cq *ibcq, int cqe, struct ib_udata *udata) cq = container_of(ibcq, struct bnxt_re_cq, ib_cq); rdev = cq->rdev; - dev_attr = &rdev->dev_attr; + dev_attr = rdev->dev_attr; if (!ibcq->uobject) { ibdev_err(&rdev->ibdev, "Kernel CQ Resize not supported"); return -EOPNOTSUPP; @@ -2959,12 +3283,11 @@ int bnxt_re_resize_cq(struct ib_cq *ibcq, int cqe, struct ib_udata *udata) return -EINVAL; } - entries = roundup_pow_of_two(cqe + 1); + uctx = rdma_udata_to_drv_context(udata, struct bnxt_re_ucontext, ib_uctx); + entries = bnxt_re_init_depth(cqe + 1, uctx); if (entries > dev_attr->max_cq_wqes + 1) entries = dev_attr->max_cq_wqes + 1; - uctx = rdma_udata_to_drv_context(udata, struct bnxt_re_ucontext, - ib_uctx); /* uverbs consumer */ if (ib_copy_from_udata(&req, udata, sizeof(req))) { rc = -EFAULT; @@ -2976,9 +3299,9 @@ int bnxt_re_resize_cq(struct ib_cq *ibcq, int cqe, struct ib_udata *udata) IB_ACCESS_LOCAL_WRITE); if (IS_ERR(cq->resize_umem)) { rc = PTR_ERR(cq->resize_umem); + ibdev_err(&rdev->ibdev, "%s: ib_umem_get failed! rc = %pe\n", + __func__, cq->resize_umem); cq->resize_umem = NULL; - ibdev_err(&rdev->ibdev, "%s: ib_umem_get failed! rc = %d\n", - __func__, rc); goto fail; } cq->resize_cqe = entries; @@ -2998,6 +3321,7 @@ int bnxt_re_resize_cq(struct ib_cq *ibcq, int cqe, struct ib_udata *udata) } cq->ib_cq.cqe = cq->resize_cqe; + atomic_inc(&rdev->stats.res.resize_count); return 0; @@ -3234,26 +3558,21 @@ static int bnxt_re_process_raw_qp_pkt_rx(struct bnxt_re_qp *gsi_qp, struct bnxt_re_dev *rdev = gsi_qp->rdev; struct bnxt_re_sqp_entries *sqp_entry = NULL; struct bnxt_re_qp *gsi_sqp = rdev->gsi_ctx.gsi_sqp; + dma_addr_t shrq_hdr_buf_map; + struct ib_sge s_sge[2] = {}; + struct ib_sge r_sge[2] = {}; struct bnxt_re_ah *gsi_sah; + struct ib_recv_wr rwr = {}; + dma_addr_t rq_hdr_buf_map; + struct ib_ud_wr udwr = {}; struct ib_send_wr *swr; - struct ib_ud_wr udwr; - struct ib_recv_wr rwr; + u32 skip_bytes = 0; int pkt_type = 0; - u32 tbl_idx; void *rq_hdr_buf; - dma_addr_t rq_hdr_buf_map; - dma_addr_t shrq_hdr_buf_map; u32 offset = 0; - u32 skip_bytes = 0; - struct ib_sge s_sge[2]; - struct ib_sge r_sge[2]; + u32 tbl_idx; int rc; - memset(&udwr, 0, sizeof(udwr)); - memset(&rwr, 0, sizeof(rwr)); - memset(&s_sge, 0, sizeof(s_sge)); - memset(&r_sge, 0, sizeof(r_sge)); - swr = &udwr.wr; tbl_idx = cqe->wr_id; @@ -3341,9 +3660,7 @@ static int bnxt_re_process_raw_qp_pkt_rx(struct bnxt_re_qp *gsi_qp, udwr.remote_qkey = gsi_sqp->qplib_qp.qkey; /* post data received in the send queue */ - rc = bnxt_re_post_send_shadow_qp(rdev, gsi_sqp, swr); - - return 0; + return bnxt_re_post_send_shadow_qp(rdev, gsi_sqp, swr); } static void bnxt_re_process_res_rawqp1_wc(struct ib_wc *wc, @@ -3431,7 +3748,7 @@ static void bnxt_re_process_res_shadow_qp_wc(struct bnxt_re_qp *gsi_sqp, wc->byte_len = orig_cqe->length; wc->qp = &gsi_qp->ib_qp; - wc->ex.imm_data = orig_cqe->immdata; + wc->ex.imm_data = cpu_to_be32(orig_cqe->immdata); wc->src_qp = orig_cqe->src_qp; memcpy(wc->smac, orig_cqe->smac, ETH_ALEN); if (bnxt_re_is_vlan_pkt(orig_cqe, &vlan_id, &sl)) { @@ -3495,7 +3812,7 @@ static int send_phantom_wqe(struct bnxt_re_qp *qp) { struct bnxt_qplib_qp *lib_qp = &qp->qplib_qp; unsigned long flags; - int rc = 0; + int rc; spin_lock_irqsave(&qp->sq_lock, flags); @@ -3576,7 +3893,10 @@ int bnxt_re_poll_cq(struct ib_cq *ib_cq, int num_entries, struct ib_wc *wc) (unsigned long)(cqe->qp_handle), struct bnxt_re_qp, qplib_qp); wc->qp = &qp->ib_qp; - wc->ex.imm_data = cqe->immdata; + if (cqe->flags & CQ_RES_RC_FLAGS_IMM) + wc->ex.imm_data = cpu_to_be32(cqe->immdata); + else + wc->ex.invalidate_rkey = cqe->invrkey; wc->src_qp = cqe->src_qp; memcpy(wc->smac, cqe->smac, ETH_ALEN); wc->port_num = 1; @@ -3685,6 +4005,7 @@ struct ib_mr *bnxt_re_get_dma_mr(struct ib_pd *ib_pd, int mr_access_flags) struct bnxt_re_pd *pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); struct bnxt_re_dev *rdev = pd->rdev; struct bnxt_re_mr *mr; + u32 active_mrs; int rc; mr = kzalloc(sizeof(*mr), GFP_KERNEL); @@ -3693,9 +4014,12 @@ struct ib_mr *bnxt_re_get_dma_mr(struct ib_pd *ib_pd, int mr_access_flags) mr->rdev = rdev; mr->qplib_mr.pd = &pd->qplib_pd; - mr->qplib_mr.flags = __from_ib_access_flags(mr_access_flags); + mr->qplib_mr.access_flags = __from_ib_access_flags(mr_access_flags); mr->qplib_mr.type = CMDQ_ALLOCATE_MRW_MRW_FLAGS_PMR; + if (mr_access_flags & IB_ACCESS_RELAXED_ORDERING) + bnxt_re_check_and_set_relaxed_ordering(rdev, &mr->qplib_mr); + /* Allocate and register 0 as the address */ rc = bnxt_qplib_alloc_mrw(&rdev->qplib_res, &mr->qplib_mr); if (rc) @@ -3704,7 +4028,7 @@ struct ib_mr *bnxt_re_get_dma_mr(struct ib_pd *ib_pd, int mr_access_flags) mr->qplib_mr.hwq.level = PBL_LVL_MAX; mr->qplib_mr.total_size = -1; /* Infinte length */ rc = bnxt_qplib_reg_mr(&rdev->qplib_res, &mr->qplib_mr, NULL, 0, - PAGE_SIZE); + PAGE_SIZE, false); if (rc) goto fail_mr; @@ -3712,7 +4036,9 @@ struct ib_mr *bnxt_re_get_dma_mr(struct ib_pd *ib_pd, int mr_access_flags) if (mr_access_flags & (IB_ACCESS_REMOTE_WRITE | IB_ACCESS_REMOTE_READ | IB_ACCESS_REMOTE_ATOMIC)) mr->ib_mr.rkey = mr->ib_mr.lkey; - atomic_inc(&rdev->mr_count); + active_mrs = atomic_inc_return(&rdev->stats.res.mr_count); + if (active_mrs > rdev->stats.res.mr_watermark) + rdev->stats.res.mr_watermark = active_mrs; return &mr->ib_mr; @@ -3745,7 +4071,7 @@ int bnxt_re_dereg_mr(struct ib_mr *ib_mr, struct ib_udata *udata) ib_umem_release(mr->ib_umem); kfree(mr); - atomic_dec(&rdev->mr_count); + atomic_dec(&rdev->stats.res.mr_count); return rc; } @@ -3775,6 +4101,7 @@ struct ib_mr *bnxt_re_alloc_mr(struct ib_pd *ib_pd, enum ib_mr_type type, struct bnxt_re_pd *pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); struct bnxt_re_dev *rdev = pd->rdev; struct bnxt_re_mr *mr = NULL; + u32 active_mrs; int rc; if (type != IB_MR_TYPE_MEM_REG) { @@ -3790,7 +4117,7 @@ struct ib_mr *bnxt_re_alloc_mr(struct ib_pd *ib_pd, enum ib_mr_type type, mr->rdev = rdev; mr->qplib_mr.pd = &pd->qplib_pd; - mr->qplib_mr.flags = BNXT_QPLIB_FR_PMR; + mr->qplib_mr.access_flags = BNXT_QPLIB_FR_PMR; mr->qplib_mr.type = CMDQ_ALLOCATE_MRW_MRW_FLAGS_PMR; rc = bnxt_qplib_alloc_mrw(&rdev->qplib_res, &mr->qplib_mr); @@ -3813,7 +4140,9 @@ struct ib_mr *bnxt_re_alloc_mr(struct ib_pd *ib_pd, enum ib_mr_type type, goto fail_mr; } - atomic_inc(&rdev->mr_count); + active_mrs = atomic_inc_return(&rdev->stats.res.mr_count); + if (active_mrs > rdev->stats.res.mr_watermark) + rdev->stats.res.mr_watermark = active_mrs; return &mr->ib_mr; fail_mr: @@ -3831,6 +4160,7 @@ struct ib_mw *bnxt_re_alloc_mw(struct ib_pd *ib_pd, enum ib_mw_type type, struct bnxt_re_pd *pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); struct bnxt_re_dev *rdev = pd->rdev; struct bnxt_re_mw *mw; + u32 active_mws; int rc; mw = kzalloc(sizeof(*mw), GFP_KERNEL); @@ -3849,7 +4179,9 @@ struct ib_mw *bnxt_re_alloc_mw(struct ib_pd *ib_pd, enum ib_mw_type type, } mw->ib_mw.rkey = mw->qplib_mw.rkey; - atomic_inc(&rdev->mw_count); + active_mws = atomic_inc_return(&rdev->stats.res.mw_count); + if (active_mws > rdev->stats.res.mw_watermark) + rdev->stats.res.mw_watermark = active_mws; return &mw->ib_mw; fail: @@ -3870,21 +4202,19 @@ int bnxt_re_dealloc_mw(struct ib_mw *ib_mw) } kfree(mw); - atomic_dec(&rdev->mw_count); + atomic_dec(&rdev->stats.res.mw_count); return rc; } -/* uverbs */ -struct ib_mr *bnxt_re_reg_user_mr(struct ib_pd *ib_pd, u64 start, u64 length, - u64 virt_addr, int mr_access_flags, - struct ib_udata *udata) +static struct ib_mr *__bnxt_re_user_reg_mr(struct ib_pd *ib_pd, u64 length, u64 virt_addr, + int mr_access_flags, struct ib_umem *umem) { struct bnxt_re_pd *pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); struct bnxt_re_dev *rdev = pd->rdev; - struct bnxt_re_mr *mr; - struct ib_umem *umem; unsigned long page_size; + struct bnxt_re_mr *mr; int umem_pgs, rc; + u32 active_mrs; if (length > BNXT_RE_MAX_MR_SIZE) { ibdev_err(&rdev->ibdev, "MR Size: %lld > Max supported:%lld\n", @@ -3892,56 +4222,58 @@ struct ib_mr *bnxt_re_reg_user_mr(struct ib_pd *ib_pd, u64 start, u64 length, return ERR_PTR(-ENOMEM); } + page_size = ib_umem_find_best_pgsz(umem, BNXT_RE_PAGE_SIZE_SUPPORTED, virt_addr); + if (!page_size) { + ibdev_err(&rdev->ibdev, "umem page size unsupported!"); + return ERR_PTR(-EINVAL); + } + mr = kzalloc(sizeof(*mr), GFP_KERNEL); if (!mr) return ERR_PTR(-ENOMEM); mr->rdev = rdev; mr->qplib_mr.pd = &pd->qplib_pd; - mr->qplib_mr.flags = __from_ib_access_flags(mr_access_flags); + mr->qplib_mr.access_flags = __from_ib_access_flags(mr_access_flags); mr->qplib_mr.type = CMDQ_ALLOCATE_MRW_MRW_FLAGS_MR; - rc = bnxt_qplib_alloc_mrw(&rdev->qplib_res, &mr->qplib_mr); - if (rc) { - ibdev_err(&rdev->ibdev, "Failed to allocate MR"); - goto free_mr; - } - /* The fixed portion of the rkey is the same as the lkey */ - mr->ib_mr.rkey = mr->qplib_mr.rkey; - - umem = ib_umem_get(&rdev->ibdev, start, length, mr_access_flags); - if (IS_ERR(umem)) { - ibdev_err(&rdev->ibdev, "Failed to get umem"); - rc = -EFAULT; - goto free_mrw; + if (!_is_alloc_mr_unified(rdev->dev_attr->dev_cap_flags)) { + rc = bnxt_qplib_alloc_mrw(&rdev->qplib_res, &mr->qplib_mr); + if (rc) { + ibdev_err(&rdev->ibdev, "Failed to allocate MR rc = %d", rc); + rc = -EIO; + goto free_mr; + } + /* The fixed portion of the rkey is the same as the lkey */ + mr->ib_mr.rkey = mr->qplib_mr.rkey; + } else { + mr->qplib_mr.flags = CMDQ_REGISTER_MR_FLAGS_ALLOC_MR; } mr->ib_umem = umem; - mr->qplib_mr.va = virt_addr; - page_size = ib_umem_find_best_pgsz( - umem, BNXT_RE_PAGE_SIZE_SUPPORTED, virt_addr); - if (!page_size) { - ibdev_err(&rdev->ibdev, "umem page size unsupported!"); - rc = -EFAULT; - goto free_umem; - } mr->qplib_mr.total_size = length; + if (mr_access_flags & IB_ACCESS_RELAXED_ORDERING) + bnxt_re_check_and_set_relaxed_ordering(rdev, &mr->qplib_mr); + umem_pgs = ib_umem_num_dma_blocks(umem, page_size); rc = bnxt_qplib_reg_mr(&rdev->qplib_res, &mr->qplib_mr, umem, - umem_pgs, page_size); + umem_pgs, page_size, + _is_alloc_mr_unified(rdev->dev_attr->dev_cap_flags)); if (rc) { - ibdev_err(&rdev->ibdev, "Failed to register user MR"); - goto free_umem; + ibdev_err(&rdev->ibdev, "Failed to register user MR - rc = %d\n", rc); + rc = -EIO; + goto free_mrw; } mr->ib_mr.lkey = mr->qplib_mr.lkey; mr->ib_mr.rkey = mr->qplib_mr.lkey; - atomic_inc(&rdev->mr_count); + active_mrs = atomic_inc_return(&rdev->stats.res.mr_count); + if (active_mrs > rdev->stats.res.mr_watermark) + rdev->stats.res.mr_watermark = active_mrs; return &mr->ib_mr; -free_umem: - ib_umem_release(umem); + free_mrw: bnxt_qplib_free_mrw(&rdev->qplib_res, &mr->qplib_mr); free_mr: @@ -3949,14 +4281,67 @@ free_mr: return ERR_PTR(rc); } +struct ib_mr *bnxt_re_reg_user_mr(struct ib_pd *ib_pd, u64 start, u64 length, + u64 virt_addr, int mr_access_flags, + struct ib_dmah *dmah, + struct ib_udata *udata) +{ + struct bnxt_re_pd *pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); + struct bnxt_re_dev *rdev = pd->rdev; + struct ib_umem *umem; + struct ib_mr *ib_mr; + + if (dmah) + return ERR_PTR(-EOPNOTSUPP); + + umem = ib_umem_get(&rdev->ibdev, start, length, mr_access_flags); + if (IS_ERR(umem)) + return ERR_CAST(umem); + + ib_mr = __bnxt_re_user_reg_mr(ib_pd, length, virt_addr, mr_access_flags, umem); + if (IS_ERR(ib_mr)) + ib_umem_release(umem); + return ib_mr; +} + +struct ib_mr *bnxt_re_reg_user_mr_dmabuf(struct ib_pd *ib_pd, u64 start, + u64 length, u64 virt_addr, int fd, + int mr_access_flags, + struct ib_dmah *dmah, + struct uverbs_attr_bundle *attrs) +{ + struct bnxt_re_pd *pd = container_of(ib_pd, struct bnxt_re_pd, ib_pd); + struct bnxt_re_dev *rdev = pd->rdev; + struct ib_umem_dmabuf *umem_dmabuf; + struct ib_umem *umem; + struct ib_mr *ib_mr; + + if (dmah) + return ERR_PTR(-EOPNOTSUPP); + + umem_dmabuf = ib_umem_dmabuf_get_pinned(&rdev->ibdev, start, length, + fd, mr_access_flags); + if (IS_ERR(umem_dmabuf)) + return ERR_CAST(umem_dmabuf); + + umem = &umem_dmabuf->umem; + + ib_mr = __bnxt_re_user_reg_mr(ib_pd, length, virt_addr, mr_access_flags, umem); + if (IS_ERR(ib_mr)) + ib_umem_release(umem); + return ib_mr; +} + int bnxt_re_alloc_ucontext(struct ib_ucontext *ctx, struct ib_udata *udata) { struct ib_device *ibdev = ctx->device; struct bnxt_re_ucontext *uctx = container_of(ctx, struct bnxt_re_ucontext, ib_uctx); struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); - struct bnxt_qplib_dev_attr *dev_attr = &rdev->dev_attr; + struct bnxt_qplib_dev_attr *dev_attr = rdev->dev_attr; + struct bnxt_re_user_mmap_entry *entry; struct bnxt_re_uctx_resp resp = {}; + struct bnxt_re_uctx_req ureq = {}; u32 chip_met_rev_num = 0; int rc; @@ -3991,8 +4376,36 @@ int bnxt_re_alloc_ucontext(struct ib_ucontext *ctx, struct ib_udata *udata) resp.cqe_sz = sizeof(struct cq_base); resp.max_cqd = dev_attr->max_cq_wqes; - resp.comp_mask |= BNXT_RE_UCNTX_CMASK_HAVE_MODE; - resp.mode = rdev->chip_ctx->modes.wqe_mode; + if (rdev->chip_ctx->modes.db_push) + resp.comp_mask |= BNXT_RE_UCNTX_CMASK_WC_DPI_ENABLED; + + entry = bnxt_re_mmap_entry_insert(uctx, 0, BNXT_RE_MMAP_SH_PAGE, NULL); + if (!entry) { + rc = -ENOMEM; + goto cfail; + } + uctx->shpage_mmap = &entry->rdma_entry; + if (rdev->pacing.dbr_pacing) + resp.comp_mask |= BNXT_RE_UCNTX_CMASK_DBR_PACING_ENABLED; + + if (_is_host_msn_table(rdev->qplib_res.dattr->dev_cap_flags2)) + resp.comp_mask |= BNXT_RE_UCNTX_CMASK_MSN_TABLE_ENABLED; + + if (udata->inlen >= sizeof(ureq)) { + rc = ib_copy_from_udata(&ureq, udata, min(udata->inlen, sizeof(ureq))); + if (rc) + goto cfail; + if (ureq.comp_mask & BNXT_RE_COMP_MASK_REQ_UCNTX_POW2_SUPPORT) { + resp.comp_mask |= BNXT_RE_UCNTX_CMASK_POW2_DISABLED; + uctx->cmask |= BNXT_RE_UCNTX_CAP_POW2_DISABLED; + } + if (ureq.comp_mask & BNXT_RE_COMP_MASK_REQ_UCNTX_VAR_WQE_SUPPORT) { + resp.comp_mask |= BNXT_RE_UCNTX_CMASK_HAVE_MODE; + resp.mode = rdev->chip_ctx->modes.wqe_mode; + if (resp.mode == BNXT_QPLIB_WQE_MODE_VARIABLE) + uctx->cmask |= BNXT_RE_UCNTX_CAP_VAR_WQE_ENABLED; + } + } rc = ib_copy_to_udata(udata, &resp, min(udata->outlen, sizeof(resp))); if (rc) { @@ -4017,6 +4430,8 @@ void bnxt_re_dealloc_ucontext(struct ib_ucontext *ib_uctx) struct bnxt_re_dev *rdev = uctx->rdev; + rdma_user_mmap_entry_remove(uctx->shpage_mmap); + uctx->shpage_mmap = NULL; if (uctx->shpg) free_page((unsigned long)uctx->shpg); @@ -4024,39 +4439,500 @@ void bnxt_re_dealloc_ucontext(struct ib_ucontext *ib_uctx) /* Free DPI only if this is the first PD allocated by the * application and mark the context dpi as NULL */ - bnxt_qplib_dealloc_dpi(&rdev->qplib_res, - &rdev->qplib_res.dpi_tbl, &uctx->dpi); + bnxt_qplib_dealloc_dpi(&rdev->qplib_res, &uctx->dpi); uctx->dpi.dbr = NULL; } } +static int bnxt_re_setup_vnic(struct bnxt_re_dev *rdev, struct bnxt_re_qp *qp) +{ + int rc; + + rc = bnxt_re_hwrm_alloc_vnic(rdev); + if (rc) + return rc; + + rc = bnxt_re_hwrm_cfg_vnic(rdev, qp->qplib_qp.id); + if (rc) + goto out_free_vnic; + + return 0; +out_free_vnic: + bnxt_re_hwrm_free_vnic(rdev); + return rc; +} + +struct ib_flow *bnxt_re_create_flow(struct ib_qp *ib_qp, + struct ib_flow_attr *attr, + struct ib_udata *udata) +{ + struct bnxt_re_qp *qp = container_of(ib_qp, struct bnxt_re_qp, ib_qp); + struct bnxt_re_dev *rdev = qp->rdev; + struct bnxt_re_flow *flow; + int rc; + + if (attr->type != IB_FLOW_ATTR_SNIFFER || + !rdev->rcfw.roce_mirror) + return ERR_PTR(-EOPNOTSUPP); + + mutex_lock(&rdev->qp_lock); + if (rdev->sniffer_flow_created) { + ibdev_err(&rdev->ibdev, "RoCE Mirroring is already Configured\n"); + mutex_unlock(&rdev->qp_lock); + return ERR_PTR(-EBUSY); + } + + flow = kzalloc(sizeof(*flow), GFP_KERNEL); + if (!flow) { + mutex_unlock(&rdev->qp_lock); + return ERR_PTR(-ENOMEM); + } + + flow->rdev = rdev; + + rc = bnxt_re_setup_vnic(rdev, qp); + if (rc) + goto out_free_flow; + + rc = bnxt_qplib_create_flow(&rdev->qplib_res); + if (rc) + goto out_free_vnic; + + rdev->sniffer_flow_created = 1; + mutex_unlock(&rdev->qp_lock); + + return &flow->ib_flow; + +out_free_vnic: + bnxt_re_hwrm_free_vnic(rdev); +out_free_flow: + mutex_unlock(&rdev->qp_lock); + kfree(flow); + return ERR_PTR(rc); +} + +int bnxt_re_destroy_flow(struct ib_flow *flow_id) +{ + struct bnxt_re_flow *flow = + container_of(flow_id, struct bnxt_re_flow, ib_flow); + struct bnxt_re_dev *rdev = flow->rdev; + int rc; + + mutex_lock(&rdev->qp_lock); + rc = bnxt_qplib_destroy_flow(&rdev->qplib_res); + if (rc) + ibdev_dbg(&rdev->ibdev, "failed to destroy_flow rc = %d\n", rc); + rdev->sniffer_flow_created = 0; + + bnxt_re_hwrm_free_vnic(rdev); + mutex_unlock(&rdev->qp_lock); + kfree(flow); + + return rc; +} + +static struct bnxt_re_cq *bnxt_re_search_for_cq(struct bnxt_re_dev *rdev, u32 cq_id) +{ + struct bnxt_re_cq *cq = NULL, *tmp_cq; + + hash_for_each_possible(rdev->cq_hash, tmp_cq, hash_entry, cq_id) { + if (tmp_cq->qplib_cq.id == cq_id) { + cq = tmp_cq; + break; + } + } + return cq; +} + +static struct bnxt_re_srq *bnxt_re_search_for_srq(struct bnxt_re_dev *rdev, u32 srq_id) +{ + struct bnxt_re_srq *srq = NULL, *tmp_srq; + + hash_for_each_possible(rdev->srq_hash, tmp_srq, hash_entry, srq_id) { + if (tmp_srq->qplib_srq.id == srq_id) { + srq = tmp_srq; + break; + } + } + return srq; +} + /* Helper function to mmap the virtual memory from user app */ int bnxt_re_mmap(struct ib_ucontext *ib_uctx, struct vm_area_struct *vma) { struct bnxt_re_ucontext *uctx = container_of(ib_uctx, struct bnxt_re_ucontext, ib_uctx); - struct bnxt_re_dev *rdev = uctx->rdev; + struct bnxt_re_user_mmap_entry *bnxt_entry; + struct rdma_user_mmap_entry *rdma_entry; + int ret = 0; u64 pfn; - if (vma->vm_end - vma->vm_start != PAGE_SIZE) + rdma_entry = rdma_user_mmap_entry_get(&uctx->ib_uctx, vma); + if (!rdma_entry) return -EINVAL; - if (vma->vm_pgoff) { - vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot); - if (io_remap_pfn_range(vma, vma->vm_start, vma->vm_pgoff, - PAGE_SIZE, vma->vm_page_prot)) { - ibdev_err(&rdev->ibdev, "Failed to map DPI"); - return -EAGAIN; + bnxt_entry = container_of(rdma_entry, struct bnxt_re_user_mmap_entry, + rdma_entry); + + switch (bnxt_entry->mmap_flag) { + case BNXT_RE_MMAP_WC_DB: + pfn = bnxt_entry->mem_offset >> PAGE_SHIFT; + ret = rdma_user_mmap_io(ib_uctx, vma, pfn, PAGE_SIZE, + pgprot_writecombine(vma->vm_page_prot), + rdma_entry); + break; + case BNXT_RE_MMAP_UC_DB: + pfn = bnxt_entry->mem_offset >> PAGE_SHIFT; + ret = rdma_user_mmap_io(ib_uctx, vma, pfn, PAGE_SIZE, + pgprot_noncached(vma->vm_page_prot), + rdma_entry); + break; + case BNXT_RE_MMAP_SH_PAGE: + ret = vm_insert_page(vma, vma->vm_start, virt_to_page(uctx->shpg)); + break; + case BNXT_RE_MMAP_DBR_BAR: + pfn = bnxt_entry->mem_offset >> PAGE_SHIFT; + ret = rdma_user_mmap_io(ib_uctx, vma, pfn, PAGE_SIZE, + pgprot_noncached(vma->vm_page_prot), + rdma_entry); + break; + case BNXT_RE_MMAP_DBR_PAGE: + case BNXT_RE_MMAP_TOGGLE_PAGE: + /* Driver doesn't expect write access for user space */ + if (vma->vm_flags & VM_WRITE) + ret = -EFAULT; + else + ret = vm_insert_page(vma, vma->vm_start, + virt_to_page((void *)bnxt_entry->mem_offset)); + break; + default: + ret = -EINVAL; + break; + } + + rdma_user_mmap_entry_put(rdma_entry); + return ret; +} + +void bnxt_re_mmap_free(struct rdma_user_mmap_entry *rdma_entry) +{ + struct bnxt_re_user_mmap_entry *bnxt_entry; + + bnxt_entry = container_of(rdma_entry, struct bnxt_re_user_mmap_entry, + rdma_entry); + + kfree(bnxt_entry); +} + +int bnxt_re_process_mad(struct ib_device *ibdev, int mad_flags, + u32 port_num, const struct ib_wc *in_wc, + const struct ib_grh *in_grh, + const struct ib_mad *in_mad, struct ib_mad *out_mad, + size_t *out_mad_size, u16 *out_mad_pkey_index) +{ + struct bnxt_re_dev *rdev = to_bnxt_re_dev(ibdev, ibdev); + struct ib_class_port_info cpi = {}; + int ret = IB_MAD_RESULT_SUCCESS; + int rc = 0; + + if (in_mad->mad_hdr.mgmt_class != IB_MGMT_CLASS_PERF_MGMT) + return ret; + + switch (in_mad->mad_hdr.attr_id) { + case IB_PMA_CLASS_PORT_INFO: + cpi.capability_mask = IB_PMA_CLASS_CAP_EXT_WIDTH; + memcpy((out_mad->data + 40), &cpi, sizeof(cpi)); + break; + case IB_PMA_PORT_COUNTERS_EXT: + rc = bnxt_re_assign_pma_port_ext_counters(rdev, out_mad); + break; + case IB_PMA_PORT_COUNTERS: + rc = bnxt_re_assign_pma_port_counters(rdev, out_mad); + break; + default: + rc = -EINVAL; + break; + } + if (rc) + return IB_MAD_RESULT_FAILURE; + ret |= IB_MAD_RESULT_REPLY; + return ret; +} + +static int UVERBS_HANDLER(BNXT_RE_METHOD_NOTIFY_DRV)(struct uverbs_attr_bundle *attrs) +{ + struct bnxt_re_ucontext *uctx; + + uctx = container_of(ib_uverbs_get_ucontext(attrs), struct bnxt_re_ucontext, ib_uctx); + bnxt_re_pacing_alert(uctx->rdev); + return 0; +} + +static int UVERBS_HANDLER(BNXT_RE_METHOD_ALLOC_PAGE)(struct uverbs_attr_bundle *attrs) +{ + struct ib_uobject *uobj = uverbs_attr_get_uobject(attrs, BNXT_RE_ALLOC_PAGE_HANDLE); + enum bnxt_re_alloc_page_type alloc_type; + struct bnxt_re_user_mmap_entry *entry; + enum bnxt_re_mmap_flag mmap_flag; + struct bnxt_qplib_chip_ctx *cctx; + struct bnxt_re_ucontext *uctx; + struct bnxt_re_dev *rdev; + u64 mmap_offset; + u32 length; + u32 dpi; + u64 addr; + int err; + + uctx = container_of(ib_uverbs_get_ucontext(attrs), struct bnxt_re_ucontext, ib_uctx); + if (IS_ERR(uctx)) + return PTR_ERR(uctx); + + err = uverbs_get_const(&alloc_type, attrs, BNXT_RE_ALLOC_PAGE_TYPE); + if (err) + return err; + + rdev = uctx->rdev; + cctx = rdev->chip_ctx; + + switch (alloc_type) { + case BNXT_RE_ALLOC_WC_PAGE: + if (cctx->modes.db_push) { + if (bnxt_qplib_alloc_dpi(&rdev->qplib_res, &uctx->wcdpi, + uctx, BNXT_QPLIB_DPI_TYPE_WC)) + return -ENOMEM; + length = PAGE_SIZE; + dpi = uctx->wcdpi.dpi; + addr = (u64)uctx->wcdpi.umdbr; + mmap_flag = BNXT_RE_MMAP_WC_DB; + } else { + return -EINVAL; } - } else { - pfn = virt_to_phys(uctx->shpg) >> PAGE_SHIFT; - if (remap_pfn_range(vma, vma->vm_start, - pfn, PAGE_SIZE, vma->vm_page_prot)) { - ibdev_err(&rdev->ibdev, "Failed to map shared page"); - return -EAGAIN; + + break; + case BNXT_RE_ALLOC_DBR_BAR_PAGE: + length = PAGE_SIZE; + addr = (u64)rdev->pacing.dbr_bar_addr; + mmap_flag = BNXT_RE_MMAP_DBR_BAR; + break; + + case BNXT_RE_ALLOC_DBR_PAGE: + length = PAGE_SIZE; + addr = (u64)rdev->pacing.dbr_page; + mmap_flag = BNXT_RE_MMAP_DBR_PAGE; + break; + + default: + return -EOPNOTSUPP; + } + + entry = bnxt_re_mmap_entry_insert(uctx, addr, mmap_flag, &mmap_offset); + if (!entry) + return -ENOMEM; + + uobj->object = entry; + uverbs_finalize_uobj_create(attrs, BNXT_RE_ALLOC_PAGE_HANDLE); + err = uverbs_copy_to(attrs, BNXT_RE_ALLOC_PAGE_MMAP_OFFSET, + &mmap_offset, sizeof(mmap_offset)); + if (err) + return err; + + err = uverbs_copy_to(attrs, BNXT_RE_ALLOC_PAGE_MMAP_LENGTH, + &length, sizeof(length)); + if (err) + return err; + + err = uverbs_copy_to(attrs, BNXT_RE_ALLOC_PAGE_DPI, + &dpi, sizeof(dpi)); + if (err) + return err; + + return 0; +} + +static int alloc_page_obj_cleanup(struct ib_uobject *uobject, + enum rdma_remove_reason why, + struct uverbs_attr_bundle *attrs) +{ + struct bnxt_re_user_mmap_entry *entry = uobject->object; + struct bnxt_re_ucontext *uctx = entry->uctx; + + switch (entry->mmap_flag) { + case BNXT_RE_MMAP_WC_DB: + if (uctx && uctx->wcdpi.dbr) { + struct bnxt_re_dev *rdev = uctx->rdev; + + bnxt_qplib_dealloc_dpi(&rdev->qplib_res, &uctx->wcdpi); + uctx->wcdpi.dbr = NULL; } + break; + case BNXT_RE_MMAP_DBR_BAR: + case BNXT_RE_MMAP_DBR_PAGE: + break; + default: + goto exit; + } + rdma_user_mmap_entry_remove(&entry->rdma_entry); +exit: + return 0; +} + +DECLARE_UVERBS_NAMED_METHOD(BNXT_RE_METHOD_ALLOC_PAGE, + UVERBS_ATTR_IDR(BNXT_RE_ALLOC_PAGE_HANDLE, + BNXT_RE_OBJECT_ALLOC_PAGE, + UVERBS_ACCESS_NEW, + UA_MANDATORY), + UVERBS_ATTR_CONST_IN(BNXT_RE_ALLOC_PAGE_TYPE, + enum bnxt_re_alloc_page_type, + UA_MANDATORY), + UVERBS_ATTR_PTR_OUT(BNXT_RE_ALLOC_PAGE_MMAP_OFFSET, + UVERBS_ATTR_TYPE(u64), + UA_MANDATORY), + UVERBS_ATTR_PTR_OUT(BNXT_RE_ALLOC_PAGE_MMAP_LENGTH, + UVERBS_ATTR_TYPE(u32), + UA_MANDATORY), + UVERBS_ATTR_PTR_OUT(BNXT_RE_ALLOC_PAGE_DPI, + UVERBS_ATTR_TYPE(u32), + UA_MANDATORY)); + +DECLARE_UVERBS_NAMED_METHOD_DESTROY(BNXT_RE_METHOD_DESTROY_PAGE, + UVERBS_ATTR_IDR(BNXT_RE_DESTROY_PAGE_HANDLE, + BNXT_RE_OBJECT_ALLOC_PAGE, + UVERBS_ACCESS_DESTROY, + UA_MANDATORY)); + +DECLARE_UVERBS_NAMED_OBJECT(BNXT_RE_OBJECT_ALLOC_PAGE, + UVERBS_TYPE_ALLOC_IDR(alloc_page_obj_cleanup), + &UVERBS_METHOD(BNXT_RE_METHOD_ALLOC_PAGE), + &UVERBS_METHOD(BNXT_RE_METHOD_DESTROY_PAGE)); + +DECLARE_UVERBS_NAMED_METHOD(BNXT_RE_METHOD_NOTIFY_DRV); + +DECLARE_UVERBS_GLOBAL_METHODS(BNXT_RE_OBJECT_NOTIFY_DRV, + &UVERBS_METHOD(BNXT_RE_METHOD_NOTIFY_DRV)); + +/* Toggle MEM */ +static int UVERBS_HANDLER(BNXT_RE_METHOD_GET_TOGGLE_MEM)(struct uverbs_attr_bundle *attrs) +{ + struct ib_uobject *uobj = uverbs_attr_get_uobject(attrs, BNXT_RE_TOGGLE_MEM_HANDLE); + enum bnxt_re_mmap_flag mmap_flag = BNXT_RE_MMAP_TOGGLE_PAGE; + enum bnxt_re_get_toggle_mem_type res_type; + struct bnxt_re_user_mmap_entry *entry; + struct bnxt_re_ucontext *uctx; + struct ib_ucontext *ib_uctx; + struct bnxt_re_dev *rdev; + struct bnxt_re_srq *srq; + u32 length = PAGE_SIZE; + struct bnxt_re_cq *cq; + u64 mem_offset; + u32 offset = 0; + u64 addr = 0; + u32 res_id; + int err; + + ib_uctx = ib_uverbs_get_ucontext(attrs); + if (IS_ERR(ib_uctx)) + return PTR_ERR(ib_uctx); + + err = uverbs_get_const(&res_type, attrs, BNXT_RE_TOGGLE_MEM_TYPE); + if (err) + return err; + + uctx = container_of(ib_uctx, struct bnxt_re_ucontext, ib_uctx); + rdev = uctx->rdev; + err = uverbs_copy_from(&res_id, attrs, BNXT_RE_TOGGLE_MEM_RES_ID); + if (err) + return err; + + switch (res_type) { + case BNXT_RE_CQ_TOGGLE_MEM: + cq = bnxt_re_search_for_cq(rdev, res_id); + if (!cq) + return -EINVAL; + + addr = (u64)cq->uctx_cq_page; + break; + case BNXT_RE_SRQ_TOGGLE_MEM: + srq = bnxt_re_search_for_srq(rdev, res_id); + if (!srq) + return -EINVAL; + + addr = (u64)srq->uctx_srq_page; + break; + + default: + return -EOPNOTSUPP; } + entry = bnxt_re_mmap_entry_insert(uctx, addr, mmap_flag, &mem_offset); + if (!entry) + return -ENOMEM; + + uobj->object = entry; + uverbs_finalize_uobj_create(attrs, BNXT_RE_TOGGLE_MEM_HANDLE); + err = uverbs_copy_to(attrs, BNXT_RE_TOGGLE_MEM_MMAP_PAGE, + &mem_offset, sizeof(mem_offset)); + if (err) + return err; + + err = uverbs_copy_to(attrs, BNXT_RE_TOGGLE_MEM_MMAP_LENGTH, + &length, sizeof(length)); + if (err) + return err; + + err = uverbs_copy_to(attrs, BNXT_RE_TOGGLE_MEM_MMAP_OFFSET, + &offset, sizeof(offset)); + if (err) + return err; + return 0; } + +static int get_toggle_mem_obj_cleanup(struct ib_uobject *uobject, + enum rdma_remove_reason why, + struct uverbs_attr_bundle *attrs) +{ + struct bnxt_re_user_mmap_entry *entry = uobject->object; + + rdma_user_mmap_entry_remove(&entry->rdma_entry); + return 0; +} + +DECLARE_UVERBS_NAMED_METHOD(BNXT_RE_METHOD_GET_TOGGLE_MEM, + UVERBS_ATTR_IDR(BNXT_RE_TOGGLE_MEM_HANDLE, + BNXT_RE_OBJECT_GET_TOGGLE_MEM, + UVERBS_ACCESS_NEW, + UA_MANDATORY), + UVERBS_ATTR_CONST_IN(BNXT_RE_TOGGLE_MEM_TYPE, + enum bnxt_re_get_toggle_mem_type, + UA_MANDATORY), + UVERBS_ATTR_PTR_IN(BNXT_RE_TOGGLE_MEM_RES_ID, + UVERBS_ATTR_TYPE(u32), + UA_MANDATORY), + UVERBS_ATTR_PTR_OUT(BNXT_RE_TOGGLE_MEM_MMAP_PAGE, + UVERBS_ATTR_TYPE(u64), + UA_MANDATORY), + UVERBS_ATTR_PTR_OUT(BNXT_RE_TOGGLE_MEM_MMAP_OFFSET, + UVERBS_ATTR_TYPE(u32), + UA_MANDATORY), + UVERBS_ATTR_PTR_OUT(BNXT_RE_TOGGLE_MEM_MMAP_LENGTH, + UVERBS_ATTR_TYPE(u32), + UA_MANDATORY)); + +DECLARE_UVERBS_NAMED_METHOD_DESTROY(BNXT_RE_METHOD_RELEASE_TOGGLE_MEM, + UVERBS_ATTR_IDR(BNXT_RE_RELEASE_TOGGLE_MEM_HANDLE, + BNXT_RE_OBJECT_GET_TOGGLE_MEM, + UVERBS_ACCESS_DESTROY, + UA_MANDATORY)); + +DECLARE_UVERBS_NAMED_OBJECT(BNXT_RE_OBJECT_GET_TOGGLE_MEM, + UVERBS_TYPE_ALLOC_IDR(get_toggle_mem_obj_cleanup), + &UVERBS_METHOD(BNXT_RE_METHOD_GET_TOGGLE_MEM), + &UVERBS_METHOD(BNXT_RE_METHOD_RELEASE_TOGGLE_MEM)); + +const struct uapi_definition bnxt_re_uapi_defs[] = { + UAPI_DEF_CHAIN_OBJ_TREE_NAMED(BNXT_RE_OBJECT_ALLOC_PAGE), + UAPI_DEF_CHAIN_OBJ_TREE_NAMED(BNXT_RE_OBJECT_NOTIFY_DRV), + UAPI_DEF_CHAIN_OBJ_TREE_NAMED(BNXT_RE_OBJECT_GET_TOGGLE_MEM), + {} +}; |