diff options
Diffstat (limited to 'drivers/infiniband/sw')
30 files changed, 731 insertions, 181 deletions
diff --git a/drivers/infiniband/sw/rdmavt/ah.c b/drivers/infiniband/sw/rdmavt/ah.c index 16c446142c2a..a96d4aa80ae8 100644 --- a/drivers/infiniband/sw/rdmavt/ah.c +++ b/drivers/infiniband/sw/rdmavt/ah.c @@ -60,32 +60,35 @@ * Return: 0 on success */ int rvt_check_ah(struct ib_device *ibdev, - struct ib_ah_attr *ah_attr) + struct rdma_ah_attr *ah_attr) { int err; + int port_num = rdma_ah_get_port_num(ah_attr); struct ib_port_attr port_attr; struct rvt_dev_info *rdi = ib_to_rvt(ibdev); - enum rdma_link_layer link = rdma_port_get_link_layer(ibdev, - ah_attr->port_num); + enum rdma_link_layer link = rdma_port_get_link_layer(ibdev, port_num); + u32 dlid = rdma_ah_get_dlid(ah_attr); + u8 ah_flags = rdma_ah_get_ah_flags(ah_attr); + u8 static_rate = rdma_ah_get_static_rate(ah_attr); - err = ib_query_port(ibdev, ah_attr->port_num, &port_attr); + err = ib_query_port(ibdev, port_num, &port_attr); if (err) return -EINVAL; - if (ah_attr->port_num < 1 || - ah_attr->port_num > ibdev->phys_port_cnt) + if (port_num < 1 || + port_num > ibdev->phys_port_cnt) return -EINVAL; - if (ah_attr->static_rate != IB_RATE_PORT_CURRENT && - ib_rate_to_mbps(ah_attr->static_rate) < 0) + if (static_rate != IB_RATE_PORT_CURRENT && + ib_rate_to_mbps(static_rate) < 0) return -EINVAL; - if ((ah_attr->ah_flags & IB_AH_GRH) && - ah_attr->grh.sgid_index >= port_attr.gid_tbl_len) + if ((ah_flags & IB_AH_GRH) && + rdma_ah_read_grh(ah_attr)->sgid_index >= port_attr.gid_tbl_len) return -EINVAL; if (link != IB_LINK_LAYER_ETHERNET) { - if (ah_attr->dlid == 0) + if (dlid == 0) return -EINVAL; - if (ah_attr->dlid >= be16_to_cpu(IB_MULTICAST_LID_BASE) && - ah_attr->dlid != be16_to_cpu(IB_LID_PERMISSIVE) && - !(ah_attr->ah_flags & IB_AH_GRH)) + if (dlid >= be16_to_cpu(IB_MULTICAST_LID_BASE) && + dlid != be16_to_cpu(IB_LID_PERMISSIVE) && + !(ah_flags & IB_AH_GRH)) return -EINVAL; } if (rdi->driver_f.check_ah) @@ -104,7 +107,7 @@ EXPORT_SYMBOL(rvt_check_ah); * Return: newly allocated ah */ struct ib_ah *rvt_create_ah(struct ib_pd *pd, - struct ib_ah_attr *ah_attr) + struct rdma_ah_attr *ah_attr) { struct rvt_ah *ah; struct rvt_dev_info *dev = ib_to_rvt(pd->device); @@ -119,7 +122,7 @@ struct ib_ah *rvt_create_ah(struct ib_pd *pd, spin_lock_irqsave(&dev->n_ahs_lock, flags); if (dev->n_ahs_allocated == dev->dparms.props.max_ah) { - spin_unlock(&dev->n_ahs_lock); + spin_unlock_irqrestore(&dev->n_ahs_lock, flags); kfree(ah); return ERR_PTR(-ENOMEM); } @@ -167,7 +170,7 @@ int rvt_destroy_ah(struct ib_ah *ibah) * * Return: 0 on success */ -int rvt_modify_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr) +int rvt_modify_ah(struct ib_ah *ibah, struct rdma_ah_attr *ah_attr) { struct rvt_ah *ah = ibah_to_rvtah(ibah); @@ -186,7 +189,7 @@ int rvt_modify_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr) * * Return: always 0 */ -int rvt_query_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr) +int rvt_query_ah(struct ib_ah *ibah, struct rdma_ah_attr *ah_attr) { struct rvt_ah *ah = ibah_to_rvtah(ibah); diff --git a/drivers/infiniband/sw/rdmavt/ah.h b/drivers/infiniband/sw/rdmavt/ah.h index e9c36be87d79..16105af99189 100644 --- a/drivers/infiniband/sw/rdmavt/ah.h +++ b/drivers/infiniband/sw/rdmavt/ah.h @@ -51,9 +51,9 @@ #include <rdma/rdma_vt.h> struct ib_ah *rvt_create_ah(struct ib_pd *pd, - struct ib_ah_attr *ah_attr); + struct rdma_ah_attr *ah_attr); int rvt_destroy_ah(struct ib_ah *ibah); -int rvt_modify_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr); -int rvt_query_ah(struct ib_ah *ibah, struct ib_ah_attr *ah_attr); +int rvt_modify_ah(struct ib_ah *ibah, struct rdma_ah_attr *ah_attr); +int rvt_query_ah(struct ib_ah *ibah, struct rdma_ah_attr *ah_attr); #endif /* DEF_RVTAH_H */ diff --git a/drivers/infiniband/sw/rdmavt/cq.c b/drivers/infiniband/sw/rdmavt/cq.c index 7aa7a4e312f1..0ae2ff8cf81e 100644 --- a/drivers/infiniband/sw/rdmavt/cq.c +++ b/drivers/infiniband/sw/rdmavt/cq.c @@ -50,6 +50,7 @@ #include <linux/kthread.h> #include "cq.h" #include "vt.h" +#include "trace.h" /** * rvt_cq_enter - add a new entry to the completion queue @@ -93,6 +94,7 @@ void rvt_cq_enter(struct rvt_cq *cq, struct ib_wc *entry, bool solicited) } return; } + trace_rvt_cq_enter(cq, entry, head); if (cq->ip) { wc->uqueue[head].wr_id = entry->wr_id; wc->uqueue[head].status = entry->status; @@ -482,6 +484,7 @@ int rvt_poll_cq(struct ib_cq *ibcq, int num_entries, struct ib_wc *entry) if (tail == wc->head) break; /* The kernel doesn't need a RMB since it has the lock. */ + trace_rvt_cq_poll(cq, &wc->kqueue[tail], npolled); *entry = wc->kqueue[tail]; if (tail >= cq->ibcq.cqe) tail = 0; diff --git a/drivers/infiniband/sw/rdmavt/mad.c b/drivers/infiniband/sw/rdmavt/mad.c index bba241faca61..d6981dc04adb 100644 --- a/drivers/infiniband/sw/rdmavt/mad.c +++ b/drivers/infiniband/sw/rdmavt/mad.c @@ -160,7 +160,7 @@ void rvt_free_mad_agents(struct rvt_dev_info *rdi) ib_unregister_mad_agent(agent); } if (rvp->sm_ah) { - ib_destroy_ah(&rvp->sm_ah->ibah); + rdma_destroy_ah(&rvp->sm_ah->ibah); rvp->sm_ah = NULL; } diff --git a/drivers/infiniband/sw/rdmavt/mcast.c b/drivers/infiniband/sw/rdmavt/mcast.c index 05c8c2afb0e3..1f12b69a0d07 100644 --- a/drivers/infiniband/sw/rdmavt/mcast.c +++ b/drivers/infiniband/sw/rdmavt/mcast.c @@ -100,10 +100,11 @@ static void rvt_mcast_qp_free(struct rvt_mcast_qp *mqp) /** * mcast_alloc - allocate the multicast GID structure * @mgid: the multicast GID + * @lid: the muilticast LID (host order) * * A list of QPs will be attached to this structure. */ -static struct rvt_mcast *rvt_mcast_alloc(union ib_gid *mgid) +static struct rvt_mcast *rvt_mcast_alloc(union ib_gid *mgid, u16 lid) { struct rvt_mcast *mcast; @@ -111,7 +112,9 @@ static struct rvt_mcast *rvt_mcast_alloc(union ib_gid *mgid) if (!mcast) goto bail; - mcast->mgid = *mgid; + mcast->mcast_addr.mgid = *mgid; + mcast->mcast_addr.lid = lid; + INIT_LIST_HEAD(&mcast->qp_list); init_waitqueue_head(&mcast->wait); atomic_set(&mcast->refcount, 0); @@ -131,15 +134,19 @@ static void rvt_mcast_free(struct rvt_mcast *mcast) } /** - * rvt_mcast_find - search the global table for the given multicast GID + * rvt_mcast_find - search the global table for the given multicast GID/LID + * NOTE: It is valid to have 1 MLID with multiple MGIDs. It is not valid + * to have 1 MGID with multiple MLIDs. * @ibp: the IB port structure * @mgid: the multicast GID to search for + * @lid: the multicast LID portion of the multicast address (host order) * * The caller is responsible for decrementing the reference count if found. * * Return: NULL if not found. */ -struct rvt_mcast *rvt_mcast_find(struct rvt_ibport *ibp, union ib_gid *mgid) +struct rvt_mcast *rvt_mcast_find(struct rvt_ibport *ibp, union ib_gid *mgid, + u16 lid) { struct rb_node *n; unsigned long flags; @@ -153,15 +160,18 @@ struct rvt_mcast *rvt_mcast_find(struct rvt_ibport *ibp, union ib_gid *mgid) mcast = rb_entry(n, struct rvt_mcast, rb_node); - ret = memcmp(mgid->raw, mcast->mgid.raw, - sizeof(union ib_gid)); + ret = memcmp(mgid->raw, mcast->mcast_addr.mgid.raw, + sizeof(*mgid)); if (ret < 0) { n = n->rb_left; } else if (ret > 0) { n = n->rb_right; } else { - atomic_inc(&mcast->refcount); - found = mcast; + /* MGID/MLID must match */ + if (mcast->mcast_addr.lid == lid) { + atomic_inc(&mcast->refcount); + found = mcast; + } break; } } @@ -177,7 +187,8 @@ EXPORT_SYMBOL(rvt_mcast_find); * * Return: zero if both were added. Return EEXIST if the GID was already in * the table but the QP was added. Return ESRCH if the QP was already - * attached and neither structure was added. + * attached and neither structure was added. Return EINVAL if the MGID was + * found, but the MLID did NOT match. */ static int rvt_mcast_add(struct rvt_dev_info *rdi, struct rvt_ibport *ibp, struct rvt_mcast *mcast, struct rvt_mcast_qp *mqp) @@ -195,8 +206,9 @@ static int rvt_mcast_add(struct rvt_dev_info *rdi, struct rvt_ibport *ibp, pn = *n; tmcast = rb_entry(pn, struct rvt_mcast, rb_node); - ret = memcmp(mcast->mgid.raw, tmcast->mgid.raw, - sizeof(union ib_gid)); + ret = memcmp(mcast->mcast_addr.mgid.raw, + tmcast->mcast_addr.mgid.raw, + sizeof(mcast->mcast_addr.mgid)); if (ret < 0) { n = &pn->rb_left; continue; @@ -206,6 +218,11 @@ static int rvt_mcast_add(struct rvt_dev_info *rdi, struct rvt_ibport *ibp, continue; } + if (tmcast->mcast_addr.lid != mcast->mcast_addr.lid) { + ret = EINVAL; + goto bail; + } + /* Search the QP list to see if this is already there. */ list_for_each_entry_rcu(p, &tmcast->qp_list, list) { if (p->qp == mqp->qp) { @@ -276,7 +293,7 @@ int rvt_attach_mcast(struct ib_qp *ibqp, union ib_gid *gid, u16 lid) * Allocate data structures since its better to do this outside of * spin locks and it will most likely be needed. */ - mcast = rvt_mcast_alloc(gid); + mcast = rvt_mcast_alloc(gid, lid); if (!mcast) return -ENOMEM; @@ -296,6 +313,10 @@ int rvt_attach_mcast(struct ib_qp *ibqp, union ib_gid *gid, u16 lid) /* Exceeded the maximum number of mcast groups. */ ret = -ENOMEM; goto bail_mqp; + case EINVAL: + /* Invalid MGID/MLID pair */ + ret = -EINVAL; + goto bail_mqp; default: break; } @@ -344,14 +365,20 @@ int rvt_detach_mcast(struct ib_qp *ibqp, union ib_gid *gid, u16 lid) } mcast = rb_entry(n, struct rvt_mcast, rb_node); - ret = memcmp(gid->raw, mcast->mgid.raw, - sizeof(union ib_gid)); - if (ret < 0) + ret = memcmp(gid->raw, mcast->mcast_addr.mgid.raw, + sizeof(*gid)); + if (ret < 0) { n = n->rb_left; - else if (ret > 0) + } else if (ret > 0) { n = n->rb_right; - else + } else { + /* MGID/MLID must match */ + if (mcast->mcast_addr.lid != lid) { + spin_unlock_irq(&ibp->lock); + return -EINVAL; + } break; + } } /* Search the QP list. */ diff --git a/drivers/infiniband/sw/rdmavt/mr.c b/drivers/infiniband/sw/rdmavt/mr.c index ae30b6838d79..aa5f9ea318e4 100644 --- a/drivers/infiniband/sw/rdmavt/mr.c +++ b/drivers/infiniband/sw/rdmavt/mr.c @@ -191,8 +191,9 @@ static int rvt_alloc_lkey(struct rvt_mregion *mr, int dma_region) tmr = rcu_access_pointer(dev->dma_mr); if (!tmr) { - rcu_assign_pointer(dev->dma_mr, mr); mr->lkey_published = 1; + /* Insure published written first */ + rcu_assign_pointer(dev->dma_mr, mr); rvt_get_mr(mr); } goto success; @@ -224,8 +225,9 @@ static int rvt_alloc_lkey(struct rvt_mregion *mr, int dma_region) mr->lkey |= 1 << 8; rkt->gen++; } - rcu_assign_pointer(rkt->table[r], mr); mr->lkey_published = 1; + /* Insure published written first */ + rcu_assign_pointer(rkt->table[r], mr); success: spin_unlock_irqrestore(&rkt->lock, flags); out: @@ -253,23 +255,24 @@ static void rvt_free_lkey(struct rvt_mregion *mr) spin_lock_irqsave(&rkt->lock, flags); if (!lkey) { if (mr->lkey_published) { - RCU_INIT_POINTER(dev->dma_mr, NULL); + mr->lkey_published = 0; + /* insure published is written before pointer */ + rcu_assign_pointer(dev->dma_mr, NULL); rvt_put_mr(mr); } } else { if (!mr->lkey_published) goto out; r = lkey >> (32 - dev->dparms.lkey_table_size); - RCU_INIT_POINTER(rkt->table[r], NULL); + mr->lkey_published = 0; + /* insure published is written before pointer */ + rcu_assign_pointer(rkt->table[r], NULL); } - mr->lkey_published = 0; freed++; out: spin_unlock_irqrestore(&rkt->lock, flags); - if (freed) { - synchronize_rcu(); + if (freed) percpu_ref_kill(&mr->refcount); - } } static struct rvt_mr *__rvt_alloc_mr(int count, struct ib_pd *pd) @@ -405,8 +408,7 @@ struct ib_mr *rvt_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, mr->mr.access_flags = mr_access_flags; mr->umem = umem; - if (is_power_of_2(umem->page_size)) - mr->mr.page_shift = ilog2(umem->page_size); + mr->mr.page_shift = umem->page_shift; m = 0; n = 0; for_each_sg(umem->sg_head.sgl, sg, umem->nmap, entry) { @@ -418,8 +420,9 @@ struct ib_mr *rvt_reg_user_mr(struct ib_pd *pd, u64 start, u64 length, goto bail_inval; } mr->mr.map[m]->segs[n].vaddr = vaddr; - mr->mr.map[m]->segs[n].length = umem->page_size; - trace_rvt_mr_user_seg(&mr->mr, m, n, vaddr, umem->page_size); + mr->mr.map[m]->segs[n].length = BIT(umem->page_shift); + trace_rvt_mr_user_seg(&mr->mr, m, n, vaddr, + BIT(umem->page_shift)); n++; if (n == RVT_SEGSZ) { m++; @@ -822,16 +825,21 @@ int rvt_lkey_ok(struct rvt_lkey_table *rkt, struct rvt_pd *pd, goto ok; } mr = rcu_dereference(rkt->table[sge->lkey >> rkt->shift]); - if (unlikely(!mr || atomic_read(&mr->lkey_invalid) || - mr->lkey != sge->lkey || mr->pd != &pd->ibpd)) + if (!mr) goto bail; + rvt_get_mr(mr); + if (!READ_ONCE(mr->lkey_published)) + goto bail_unref; + + if (unlikely(atomic_read(&mr->lkey_invalid) || + mr->lkey != sge->lkey || mr->pd != &pd->ibpd)) + goto bail_unref; off = sge->addr - mr->user_base; if (unlikely(sge->addr < mr->user_base || off + sge->length > mr->length || (mr->access_flags & acc) != acc)) - goto bail; - rvt_get_mr(mr); + goto bail_unref; rcu_read_unlock(); off += mr->offset; @@ -867,6 +875,8 @@ int rvt_lkey_ok(struct rvt_lkey_table *rkt, struct rvt_pd *pd, isge->n = n; ok: return 1; +bail_unref: + rvt_put_mr(mr); bail: rcu_read_unlock(); return 0; @@ -922,15 +932,20 @@ int rvt_rkey_ok(struct rvt_qp *qp, struct rvt_sge *sge, } mr = rcu_dereference(rkt->table[rkey >> rkt->shift]); - if (unlikely(!mr || atomic_read(&mr->lkey_invalid) || - mr->lkey != rkey || qp->ibqp.pd != mr->pd)) + if (!mr) goto bail; + rvt_get_mr(mr); + /* insure mr read is before test */ + if (!READ_ONCE(mr->lkey_published)) + goto bail_unref; + if (unlikely(atomic_read(&mr->lkey_invalid) || + mr->lkey != rkey || qp->ibqp.pd != mr->pd)) + goto bail_unref; off = vaddr - mr->iova; if (unlikely(vaddr < mr->iova || off + len > mr->length || (mr->access_flags & acc) == 0)) - goto bail; - rvt_get_mr(mr); + goto bail_unref; rcu_read_unlock(); off += mr->offset; @@ -966,6 +981,8 @@ int rvt_rkey_ok(struct rvt_qp *qp, struct rvt_sge *sge, sge->n = n; ok: return 1; +bail_unref: + rvt_put_mr(mr); bail: rcu_read_unlock(); return 0; diff --git a/drivers/infiniband/sw/rdmavt/qp.c b/drivers/infiniband/sw/rdmavt/qp.c index f5ad8d4bfb39..727e81cc2c8f 100644 --- a/drivers/infiniband/sw/rdmavt/qp.c +++ b/drivers/infiniband/sw/rdmavt/qp.c @@ -1,5 +1,5 @@ /* - * Copyright(c) 2016 Intel Corporation. + * Copyright(c) 2016, 2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -117,23 +117,6 @@ const int ib_rvt_state_ops[IB_QPS_ERR + 1] = { }; EXPORT_SYMBOL(ib_rvt_state_ops); -/* - * Translate ib_wr_opcode into ib_wc_opcode. - */ -const enum ib_wc_opcode ib_rvt_wc_opcode[] = { - [IB_WR_RDMA_WRITE] = IB_WC_RDMA_WRITE, - [IB_WR_RDMA_WRITE_WITH_IMM] = IB_WC_RDMA_WRITE, - [IB_WR_SEND] = IB_WC_SEND, - [IB_WR_SEND_WITH_IMM] = IB_WC_SEND, - [IB_WR_RDMA_READ] = IB_WC_RDMA_READ, - [IB_WR_ATOMIC_CMP_AND_SWP] = IB_WC_COMP_SWAP, - [IB_WR_ATOMIC_FETCH_AND_ADD] = IB_WC_FETCH_ADD, - [IB_WR_SEND_WITH_INV] = IB_WC_SEND, - [IB_WR_LOCAL_INV] = IB_WC_LOCAL_INV, - [IB_WR_REG_MR] = IB_WC_REG_MR -}; -EXPORT_SYMBOL(ib_rvt_wc_opcode); - static void get_map_page(struct rvt_qpn_table *qpt, struct rvt_qpn_map *map, gfp_t gfp) @@ -1121,14 +1104,15 @@ int rvt_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, goto inval; if (attr_mask & IB_QP_AV) { - if (attr->ah_attr.dlid >= be16_to_cpu(IB_MULTICAST_LID_BASE)) + if (rdma_ah_get_dlid(&attr->ah_attr) >= + be16_to_cpu(IB_MULTICAST_LID_BASE)) goto inval; if (rvt_check_ah(qp->ibqp.device, &attr->ah_attr)) goto inval; } if (attr_mask & IB_QP_ALT_PATH) { - if (attr->alt_ah_attr.dlid >= + if (rdma_ah_get_dlid(&attr->alt_ah_attr) >= be16_to_cpu(IB_MULTICAST_LID_BASE)) goto inval; if (rvt_check_ah(qp->ibqp.device, &attr->alt_ah_attr)) @@ -1257,7 +1241,7 @@ int rvt_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, if (attr_mask & IB_QP_AV) { qp->remote_ah_attr = attr->ah_attr; - qp->s_srate = attr->ah_attr.static_rate; + qp->s_srate = rdma_ah_get_static_rate(&attr->ah_attr); qp->srate_mbps = ib_rate_to_mbps(qp->s_srate); } @@ -1270,7 +1254,7 @@ int rvt_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, qp->s_mig_state = attr->path_mig_state; if (mig) { qp->remote_ah_attr = qp->alt_ah_attr; - qp->port_num = qp->alt_ah_attr.port_num; + qp->port_num = rdma_ah_get_port_num(&qp->alt_ah_attr); qp->s_pkey_index = qp->s_alt_pkey_index; } } @@ -1441,7 +1425,8 @@ int rvt_query_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr, attr->timeout = qp->timeout; attr->retry_cnt = qp->s_retry_cnt; attr->rnr_retry = qp->s_rnr_retry_cnt; - attr->alt_port_num = qp->alt_ah_attr.port_num; + attr->alt_port_num = + rdma_ah_get_port_num(&qp->alt_ah_attr); attr->alt_timeout = qp->alt_timeout; init_attr->event_handler = qp->ibqp.event_handler; @@ -1789,11 +1774,14 @@ static int rvt_post_one_wr(struct rvt_qp *qp, 0); qp->s_next_psn = wqe->lpsn + 1; } - trace_rvt_post_one_wr(qp, wqe); - if (unlikely(reserved_op)) + if (unlikely(reserved_op)) { + wqe->wr.send_flags |= RVT_SEND_RESERVE_USED; rvt_qp_wqe_reserve(qp, wqe); - else + } else { + wqe->wr.send_flags &= ~RVT_SEND_RESERVE_USED; qp->s_avail--; + } + trace_rvt_post_one_wr(qp, wqe); smp_wmb(); /* see request builders */ qp->s_head = next; @@ -2069,8 +2057,12 @@ static void rvt_rc_timeout(unsigned long arg) spin_lock_irqsave(&qp->r_lock, flags); spin_lock(&qp->s_lock); if (qp->s_flags & RVT_S_TIMER) { + struct rvt_ibport *rvp = rdi->ports[qp->port_num - 1]; + qp->s_flags &= ~RVT_S_TIMER; + rvp->n_rc_timeouts++; del_timer(&qp->s_timer); + trace_rvt_rc_timeout(qp, qp->s_last_psn + 1); if (rdi->driver_f.notify_restart_rc) rdi->driver_f.notify_restart_rc(qp, qp->s_last_psn + 1, diff --git a/drivers/infiniband/sw/rdmavt/trace.h b/drivers/infiniband/sw/rdmavt/trace.h index e2d23acb6a7d..bb4b1e710f22 100644 --- a/drivers/infiniband/sw/rdmavt/trace.h +++ b/drivers/infiniband/sw/rdmavt/trace.h @@ -1,5 +1,5 @@ /* - * Copyright(c) 2016 Intel Corporation. + * Copyright(c) 2016, 2017 Intel Corporation. * * This file is provided under a dual BSD/GPLv2 license. When using or * redistributing this file, you may do so under either license. @@ -52,3 +52,5 @@ #include "trace_qp.h" #include "trace_tx.h" #include "trace_mr.h" +#include "trace_cq.h" +#include "trace_rc.h" diff --git a/drivers/infiniband/sw/rdmavt/trace_cq.h b/drivers/infiniband/sw/rdmavt/trace_cq.h new file mode 100644 index 000000000000..a315850aa9bb --- /dev/null +++ b/drivers/infiniband/sw/rdmavt/trace_cq.h @@ -0,0 +1,127 @@ +/* + * Copyright(c) 2016 Intel Corporation. + * + * This file is provided under a dual BSD/GPLv2 license. When using or + * redistributing this file, you may do so under either license. + * + * GPL LICENSE SUMMARY + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * BSD LICENSE + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * - Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * - 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. + * - Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS 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 THE COPYRIGHT + * OWNER 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. + * + */ +#if !defined(__RVT_TRACE_CQ_H) || defined(TRACE_HEADER_MULTI_READ) +#define __RVT_TRACE_CQ_H + +#include <linux/tracepoint.h> +#include <linux/trace_seq.h> + +#include <rdma/ib_verbs.h> +#include <rdma/rdmavt_cq.h> + +#undef TRACE_SYSTEM +#define TRACE_SYSTEM rvt_cq + +#define wc_opcode_name(opcode) { IB_WC_##opcode, #opcode } +#define show_wc_opcode(opcode) \ +__print_symbolic(opcode, \ + wc_opcode_name(SEND), \ + wc_opcode_name(RDMA_WRITE), \ + wc_opcode_name(RDMA_READ), \ + wc_opcode_name(COMP_SWAP), \ + wc_opcode_name(FETCH_ADD), \ + wc_opcode_name(LSO), \ + wc_opcode_name(LOCAL_INV), \ + wc_opcode_name(REG_MR), \ + wc_opcode_name(MASKED_COMP_SWAP), \ + wc_opcode_name(RECV), \ + wc_opcode_name(RECV_RDMA_WITH_IMM)) + +#define CQ_PRN \ +"[%s] idx %u wr_id %llx status %u opcode %u,%s length %u qpn %x" + +DECLARE_EVENT_CLASS( + rvt_cq_entry_template, + TP_PROTO(struct rvt_cq *cq, struct ib_wc *wc, u32 idx), + TP_ARGS(cq, wc, idx), + TP_STRUCT__entry( + RDI_DEV_ENTRY(cq->rdi) + __field(u64, wr_id) + __field(u32, status) + __field(u32, opcode) + __field(u32, qpn) + __field(u32, length) + __field(u32, idx) + ), + TP_fast_assign( + RDI_DEV_ASSIGN(cq->rdi) + __entry->wr_id = wc->wr_id; + __entry->status = wc->status; + __entry->opcode = wc->opcode; + __entry->length = wc->byte_len; + __entry->qpn = wc->qp->qp_num; + __entry->idx = idx; + ), + TP_printk( + CQ_PRN, + __get_str(dev), + __entry->idx, + __entry->wr_id, + __entry->status, + __entry->opcode, show_wc_opcode(__entry->opcode), + __entry->length, + __entry->qpn + ) +); + +DEFINE_EVENT( + rvt_cq_entry_template, rvt_cq_enter, + TP_PROTO(struct rvt_cq *cq, struct ib_wc *wc, u32 idx), + TP_ARGS(cq, wc, idx)); + +DEFINE_EVENT( + rvt_cq_entry_template, rvt_cq_poll, + TP_PROTO(struct rvt_cq *cq, struct ib_wc *wc, u32 idx), + TP_ARGS(cq, wc, idx)); + +#endif /* __RVT_TRACE_CQ_H */ + +#undef TRACE_INCLUDE_PATH +#undef TRACE_INCLUDE_FILE +#define TRACE_INCLUDE_PATH . +#define TRACE_INCLUDE_FILE trace_cq +#include <trace/define_trace.h> diff --git a/drivers/infiniband/sw/rdmavt/trace_rc.h b/drivers/infiniband/sw/rdmavt/trace_rc.h new file mode 100644 index 000000000000..995276933a55 --- /dev/null +++ b/drivers/infiniband/sw/rdmavt/trace_rc.h @@ -0,0 +1,109 @@ +/* + * Copyright(c) 2017 Intel Corporation. + * + * This file is provided under a dual BSD/GPLv2 license. When using or + * redistributing this file, you may do so under either license. + * + * GPL LICENSE SUMMARY + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of version 2 of the GNU General Public License as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, but + * WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * BSD LICENSE + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * + * - Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * - 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. + * - Neither the name of Intel Corporation nor the names of its + * contributors may be used to endorse or promote products derived + * from this software without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS 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 THE COPYRIGHT + * OWNER 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. + * + */ +#if !defined(__RVT_TRACE_RC_H) || defined(TRACE_HEADER_MULTI_READ) +#define __RVT_TRACE_RC_H + +#include <linux/tracepoint.h> +#include <linux/trace_seq.h> + +#include <rdma/ib_verbs.h> +#include <rdma/rdma_vt.h> + +#undef TRACE_SYSTEM +#define TRACE_SYSTEM rvt_rc + +DECLARE_EVENT_CLASS(rvt_rc_template, + TP_PROTO(struct rvt_qp *qp, u32 psn), + TP_ARGS(qp, psn), + TP_STRUCT__entry( + RDI_DEV_ENTRY(ib_to_rvt(qp->ibqp.device)) + __field(u32, qpn) + __field(u32, s_flags) + __field(u32, psn) + __field(u32, s_psn) + __field(u32, s_next_psn) + __field(u32, s_sending_psn) + __field(u32, s_sending_hpsn) + __field(u32, r_psn) + ), + TP_fast_assign( + RDI_DEV_ASSIGN(ib_to_rvt(qp->ibqp.device)) + __entry->qpn = qp->ibqp.qp_num; + __entry->s_flags = qp->s_flags; + __entry->psn = psn; + __entry->s_psn = qp->s_psn; + __entry->s_next_psn = qp->s_next_psn; + __entry->s_sending_psn = qp->s_sending_psn; + __entry->s_sending_hpsn = qp->s_sending_hpsn; + __entry->r_psn = qp->r_psn; + ), + TP_printk( + "[%s] qpn 0x%x s_flags 0x%x psn 0x%x s_psn 0x%x s_next_psn 0x%x s_sending_psn 0x%x sending_hpsn 0x%x r_psn 0x%x", + __get_str(dev), + __entry->qpn, + __entry->s_flags, + __entry->psn, + __entry->s_psn, + __entry->s_next_psn, + __entry->s_sending_psn, + __entry->s_sending_hpsn, + __entry->r_psn + ) +); + +DEFINE_EVENT(rvt_rc_template, rvt_rc_timeout, + TP_PROTO(struct rvt_qp *qp, u32 psn), + TP_ARGS(qp, psn) +); + +#endif /* __RVT_TRACE_RC_H */ + +#undef TRACE_INCLUDE_PATH +#undef TRACE_INCLUDE_FILE +#define TRACE_INCLUDE_PATH . +#define TRACE_INCLUDE_FILE trace_rc +#include <trace/define_trace.h> diff --git a/drivers/infiniband/sw/rdmavt/trace_tx.h b/drivers/infiniband/sw/rdmavt/trace_tx.h index 0e03173662d8..a613a2223751 100644 --- a/drivers/infiniband/sw/rdmavt/trace_tx.h +++ b/drivers/infiniband/sw/rdmavt/trace_tx.h @@ -71,10 +71,20 @@ __print_symbolic(opcode, \ wr_opcode_name(RDMA_READ_WITH_INV), \ wr_opcode_name(LOCAL_INV), \ wr_opcode_name(MASKED_ATOMIC_CMP_AND_SWP), \ - wr_opcode_name(MASKED_ATOMIC_FETCH_AND_ADD)) + wr_opcode_name(MASKED_ATOMIC_FETCH_AND_ADD), \ + wr_opcode_name(RESERVED1), \ + wr_opcode_name(RESERVED2), \ + wr_opcode_name(RESERVED3), \ + wr_opcode_name(RESERVED4), \ + wr_opcode_name(RESERVED5), \ + wr_opcode_name(RESERVED6), \ + wr_opcode_name(RESERVED7), \ + wr_opcode_name(RESERVED8), \ + wr_opcode_name(RESERVED9), \ + wr_opcode_name(RESERVED10)) #define POS_PRN \ -"[%s] wr_id %llx qpn %x psn 0x%x lpsn 0x%x length %u opcode 0x%.2x,%s size %u avail %u head %u last %u" +"[%s] wqe %p wr_id %llx send_flags %x qpn %x qpt %u psn %x lpsn %x ssn %x length %u opcode 0x%.2x,%s size %u avail %u head %u last %u pid %u num_sge %u" TRACE_EVENT( rvt_post_one_wr, @@ -83,7 +93,9 @@ TRACE_EVENT( TP_STRUCT__entry( RDI_DEV_ENTRY(ib_to_rvt(qp->ibqp.device)) __field(u64, wr_id) + __field(struct rvt_swqe *, wqe) __field(u32, qpn) + __field(u32, qpt) __field(u32, psn) __field(u32, lpsn) __field(u32, length) @@ -92,11 +104,17 @@ TRACE_EVENT( __field(u32, avail) __field(u32, head) __field(u32, last) + __field(u32, ssn) + __field(int, send_flags) + __field(pid_t, pid) + __field(int, num_sge) ), TP_fast_assign( RDI_DEV_ASSIGN(ib_to_rvt(qp->ibqp.device)) + __entry->wqe = wqe; __entry->wr_id = wqe->wr.wr_id; __entry->qpn = qp->ibqp.qp_num; + __entry->qpt = qp->ibqp.qp_type; __entry->psn = wqe->psn; __entry->lpsn = wqe->lpsn; __entry->length = wqe->length; @@ -105,20 +123,30 @@ TRACE_EVENT( __entry->avail = qp->s_avail; __entry->head = qp->s_head; __entry->last = qp->s_last; + __entry->pid = qp->pid; + __entry->ssn = wqe->ssn; + __entry->send_flags = wqe->wr.send_flags; + __entry->num_sge = wqe->wr.num_sge; ), TP_printk( POS_PRN, __get_str(dev), + __entry->wqe, __entry->wr_id, + __entry->send_flags, __entry->qpn, + __entry->qpt, __entry->psn, __entry->lpsn, + __entry->ssn, __entry->length, __entry->opcode, show_wr_opcode(__entry->opcode), __entry->size, __entry->avail, __entry->head, - __entry->last + __entry->last, + __entry->pid, + __entry->num_sge ) ); diff --git a/drivers/infiniband/sw/rxe/Kconfig b/drivers/infiniband/sw/rxe/Kconfig index 6332dedc11e8..320bffc980d8 100644 --- a/drivers/infiniband/sw/rxe/Kconfig +++ b/drivers/infiniband/sw/rxe/Kconfig @@ -2,6 +2,7 @@ config RDMA_RXE tristate "Software RDMA over Ethernet (RoCE) driver" depends on INET && PCI && INFINIBAND depends on NET_UDP_TUNNEL + depends on CRYPTO_CRC32 select DMA_VIRT_OPS ---help--- This driver implements the InfiniBand RDMA transport over diff --git a/drivers/infiniband/sw/rxe/Makefile b/drivers/infiniband/sw/rxe/Makefile index ec35ff022a42..3f12beb7076f 100644 --- a/drivers/infiniband/sw/rxe/Makefile +++ b/drivers/infiniband/sw/rxe/Makefile @@ -20,4 +20,5 @@ rdma_rxe-y := \ rxe_mcast.o \ rxe_task.o \ rxe_net.o \ - rxe_sysfs.o + rxe_sysfs.o \ + rxe_hw_counters.o diff --git a/drivers/infiniband/sw/rxe/rxe.c b/drivers/infiniband/sw/rxe/rxe.c index b12dd9b5a89d..c21c913f911a 100644 --- a/drivers/infiniband/sw/rxe/rxe.c +++ b/drivers/infiniband/sw/rxe/rxe.c @@ -31,6 +31,7 @@ * SOFTWARE. */ +#include <net/addrconf.h> #include "rxe.h" #include "rxe_loc.h" @@ -64,6 +65,8 @@ static void rxe_cleanup(struct rxe_dev *rxe) rxe_pool_cleanup(&rxe->mc_elem_pool); rxe_cleanup_ports(rxe); + + crypto_free_shash(rxe->tfm); } /* called when all references have been dropped */ @@ -178,7 +181,8 @@ static int rxe_init_ports(struct rxe_dev *rxe) return -ENOMEM; port->pkey_tbl[0] = 0xffff; - port->port_guid = rxe_port_guid(rxe); + addrconf_addr_eui48((unsigned char *)&port->port_guid, + rxe->ndev->dev_addr); spin_lock_init(&port->port_lock); diff --git a/drivers/infiniband/sw/rxe/rxe.h b/drivers/infiniband/sw/rxe/rxe.h index a696af81e4a5..ecdba2fce083 100644 --- a/drivers/infiniband/sw/rxe/rxe.h +++ b/drivers/infiniband/sw/rxe/rxe.h @@ -50,6 +50,7 @@ #include <rdma/ib_umem.h> #include <rdma/ib_cache.h> #include <rdma/ib_addr.h> +#include <crypto/hash.h> #include "rxe_net.h" #include "rxe_opcode.h" @@ -64,6 +65,25 @@ #define RXE_ROCE_V2_SPORT (0xc000) +static inline u32 rxe_crc32(struct rxe_dev *rxe, + u32 crc, void *next, size_t len) +{ + int err; + + SHASH_DESC_ON_STACK(shash, rxe->tfm); + + shash->tfm = rxe->tfm; + shash->flags = 0; + *(u32 *)shash_desc_ctx(shash) = crc; + err = crypto_shash_update(shash, next, len); + if (unlikely(err)) { + pr_warn_ratelimited("failed crc calculation, err: %d\n", err); + return crc32_le(crc, next, len); + } + + return *(u32 *)shash_desc_ctx(shash); +} + int rxe_set_mtu(struct rxe_dev *rxe, unsigned int dev_mtu); int rxe_add(struct rxe_dev *rxe, unsigned int mtu); diff --git a/drivers/infiniband/sw/rxe/rxe_av.c b/drivers/infiniband/sw/rxe/rxe_av.c index 604f6fee96bd..5bddf469361b 100644 --- a/drivers/infiniband/sw/rxe/rxe_av.c +++ b/drivers/infiniband/sw/rxe/rxe_av.c @@ -34,21 +34,22 @@ #include "rxe.h" #include "rxe_loc.h" -int rxe_av_chk_attr(struct rxe_dev *rxe, struct ib_ah_attr *attr) +int rxe_av_chk_attr(struct rxe_dev *rxe, struct rdma_ah_attr *attr) { struct rxe_port *port; - if (attr->port_num != 1) { - pr_info("invalid port_num = %d\n", attr->port_num); + if (rdma_ah_get_port_num(attr) != 1) { + pr_info("invalid port_num = %d\n", rdma_ah_get_port_num(attr)); return -EINVAL; } port = &rxe->port; - if (attr->ah_flags & IB_AH_GRH) { - if (attr->grh.sgid_index > port->attr.gid_tbl_len) { - pr_info("invalid sgid index = %d\n", - attr->grh.sgid_index); + if (rdma_ah_get_ah_flags(attr) & IB_AH_GRH) { + u8 sgid_index = rdma_ah_read_grh(attr)->sgid_index; + + if (sgid_index > port->attr.gid_tbl_len) { + pr_info("invalid sgid index = %d\n", sgid_index); return -EINVAL; } } @@ -57,30 +58,33 @@ int rxe_av_chk_attr(struct rxe_dev *rxe, struct ib_ah_attr *attr) } int rxe_av_from_attr(struct rxe_dev *rxe, u8 port_num, - struct rxe_av *av, struct ib_ah_attr *attr) + struct rxe_av *av, struct rdma_ah_attr *attr) { memset(av, 0, sizeof(*av)); - memcpy(&av->grh, &attr->grh, sizeof(attr->grh)); + memcpy(&av->grh, rdma_ah_read_grh(attr), + sizeof(*rdma_ah_read_grh(attr))); av->port_num = port_num; return 0; } int rxe_av_to_attr(struct rxe_dev *rxe, struct rxe_av *av, - struct ib_ah_attr *attr) + struct rdma_ah_attr *attr) { - memcpy(&attr->grh, &av->grh, sizeof(av->grh)); - attr->port_num = av->port_num; + attr->type = RDMA_AH_ATTR_TYPE_ROCE; + memcpy(rdma_ah_retrieve_grh(attr), &av->grh, sizeof(av->grh)); + rdma_ah_set_ah_flags(attr, IB_AH_GRH); + rdma_ah_set_port_num(attr, av->port_num); return 0; } int rxe_av_fill_ip_info(struct rxe_dev *rxe, struct rxe_av *av, - struct ib_ah_attr *attr, + struct rdma_ah_attr *attr, struct ib_gid_attr *sgid_attr, union ib_gid *sgid) { rdma_gid2ip(&av->sgid_addr._sockaddr, sgid); - rdma_gid2ip(&av->dgid_addr._sockaddr, &attr->grh.dgid); + rdma_gid2ip(&av->dgid_addr._sockaddr, &rdma_ah_read_grh(attr)->dgid); av->network_type = ib_gid_to_network_type(sgid_attr->gid_type, sgid); return 0; diff --git a/drivers/infiniband/sw/rxe/rxe_comp.c b/drivers/infiniband/sw/rxe/rxe_comp.c index 4cd55d5617f7..9eb12c2e3c74 100644 --- a/drivers/infiniband/sw/rxe/rxe_comp.c +++ b/drivers/infiniband/sw/rxe/rxe_comp.c @@ -154,6 +154,8 @@ void rxe_comp_queue_pkt(struct rxe_dev *rxe, struct rxe_qp *qp, skb_queue_tail(&qp->resp_pkts, skb); must_sched = skb_queue_len(&qp->resp_pkts) > 1; + if (must_sched != 0) + rxe_counter_inc(rxe, RXE_CNT_COMPLETER_SCHED); rxe_run_task(&qp->comp.task, must_sched); } @@ -236,6 +238,7 @@ static inline enum comp_state check_ack(struct rxe_qp *qp, { unsigned int mask = pkt->mask; u8 syn; + struct rxe_dev *rxe = to_rdev(qp->ibqp.device); /* Check the sequence only */ switch (qp->comp.opcode) { @@ -298,6 +301,7 @@ static inline enum comp_state check_ack(struct rxe_qp *qp, return COMPST_WRITE_SEND; case AETH_RNR_NAK: + rxe_counter_inc(rxe, RXE_CNT_RCV_RNR); return COMPST_RNR_RETRY; case AETH_NAK: @@ -307,6 +311,8 @@ static inline enum comp_state check_ack(struct rxe_qp *qp, * before */ if (psn_compare(pkt->psn, qp->comp.psn) > 0) { + rxe_counter_inc(rxe, + RXE_CNT_RCV_SEQ_ERR); qp->comp.psn = pkt->psn; if (qp->req.wait_psn) { qp->req.wait_psn = 0; @@ -534,6 +540,7 @@ static void rxe_drain_resp_pkts(struct rxe_qp *qp, bool notify) int rxe_completer(void *arg) { struct rxe_qp *qp = (struct rxe_qp *)arg; + struct rxe_dev *rxe = to_rdev(qp->ibqp.device); struct rxe_send_wqe *wqe = wqe; struct sk_buff *skb = NULL; struct rxe_pkt_info *pkt = NULL; @@ -683,8 +690,10 @@ int rxe_completer(void *arg) if (psn_compare(qp->req.psn, qp->comp.psn) > 0) { /* tell the requester to retry the - * send send queue next time around + * send queue next time around */ + rxe_counter_inc(rxe, + RXE_CNT_COMP_RETRY); qp->req.need_retry = 1; rxe_run_task(&qp->req.task, 1); } @@ -699,6 +708,7 @@ int rxe_completer(void *arg) goto exit; } else { + rxe_counter_inc(rxe, RXE_CNT_RETRY_EXCEEDED); wqe->status = IB_WC_RETRY_EXC_ERR; state = COMPST_ERROR; } @@ -720,6 +730,8 @@ int rxe_completer(void *arg) skb = NULL; goto exit; } else { + rxe_counter_inc(rxe, + RXE_CNT_RNR_RETRY_EXCEEDED); wqe->status = IB_WC_RNR_RETRY_EXC_ERR; state = COMPST_ERROR; } diff --git a/drivers/infiniband/sw/rxe/rxe_hw_counters.c b/drivers/infiniband/sw/rxe/rxe_hw_counters.c new file mode 100644 index 000000000000..7ef90aad7dfd --- /dev/null +++ b/drivers/infiniband/sw/rxe/rxe_hw_counters.c @@ -0,0 +1,78 @@ +/* + * Copyright (c) 2017 Mellanox Technologies Ltd. All rights reserved. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - 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. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + */ + +#include "rxe.h" +#include "rxe_hw_counters.h" + +const char * const rxe_counter_name[] = { + [RXE_CNT_SENT_PKTS] = "sent_pkts", + [RXE_CNT_RCVD_PKTS] = "rcvd_pkts", + [RXE_CNT_DUP_REQ] = "duplicate_request", + [RXE_CNT_OUT_OF_SEQ_REQ] = "out_of_sequence", + [RXE_CNT_RCV_RNR] = "rcvd_rnr_err", + [RXE_CNT_SND_RNR] = "send_rnr_err", + [RXE_CNT_RCV_SEQ_ERR] = "rcvd_seq_err", + [RXE_CNT_COMPLETER_SCHED] = "ack_deffered", + [RXE_CNT_RETRY_EXCEEDED] = "retry_exceeded_err", + [RXE_CNT_RNR_RETRY_EXCEEDED] = "retry_rnr_exceeded_err", + [RXE_CNT_COMP_RETRY] = "completer_retry_err", + [RXE_CNT_SEND_ERR] = "send_err", +}; + +int rxe_ib_get_hw_stats(struct ib_device *ibdev, + struct rdma_hw_stats *stats, + u8 port, int index) +{ + struct rxe_dev *dev = to_rdev(ibdev); + unsigned int cnt; + + if (!port || !stats) + return -EINVAL; + + for (cnt = 0; cnt < ARRAY_SIZE(rxe_counter_name); cnt++) + stats->value[cnt] = dev->stats_counters[cnt]; + + return ARRAY_SIZE(rxe_counter_name); +} + +struct rdma_hw_stats *rxe_ib_alloc_hw_stats(struct ib_device *ibdev, + u8 port_num) +{ + BUILD_BUG_ON(ARRAY_SIZE(rxe_counter_name) != RXE_NUM_OF_COUNTERS); + /* We support only per port stats */ + if (!port_num) + return NULL; + + return rdma_alloc_hw_stats_struct(rxe_counter_name, + ARRAY_SIZE(rxe_counter_name), + RDMA_HW_STATS_DEFAULT_LIFESPAN); +} diff --git a/drivers/infiniband/sw/rxe/rxe_hw_counters.h b/drivers/infiniband/sw/rxe/rxe_hw_counters.h new file mode 100644 index 000000000000..f44df1b76742 --- /dev/null +++ b/drivers/infiniband/sw/rxe/rxe_hw_counters.h @@ -0,0 +1,61 @@ +/* + * Copyright (c) 2017 Mellanox Technologies Ltd. All rights reserved. + * + * This software is available to you under a choice of one of two + * licenses. You may choose to be licensed under the terms of the GNU + * General Public License (GPL) Version 2, available from the file + * COPYING in the main directory of this source tree, or the + * OpenIB.org BSD license below: + * + * Redistribution and use in source and binary forms, with or + * without modification, are permitted provided that the following + * conditions are met: + * + * - Redistributions of source code must retain the above + * copyright notice, this list of conditions and the following + * disclaimer. + * + * - 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. + * + * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, + * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF + * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND + * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS + * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN + * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN + * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE + * SOFTWARE. + */ + +#ifndef RXE_HW_COUNTERS_H +#define RXE_HW_COUNTERS_H + +/* + * when adding counters to enum also add + * them to rxe_counter_name[] vector. + */ +enum rxe_counters { + RXE_CNT_SENT_PKTS, + RXE_CNT_RCVD_PKTS, + RXE_CNT_DUP_REQ, + RXE_CNT_OUT_OF_SEQ_REQ, + RXE_CNT_RCV_RNR, + RXE_CNT_SND_RNR, + RXE_CNT_RCV_SEQ_ERR, + RXE_CNT_COMPLETER_SCHED, + RXE_CNT_RETRY_EXCEEDED, + RXE_CNT_RNR_RETRY_EXCEEDED, + RXE_CNT_COMP_RETRY, + RXE_CNT_SEND_ERR, + RXE_NUM_OF_COUNTERS +}; + +struct rdma_hw_stats *rxe_ib_alloc_hw_stats(struct ib_device *ibdev, + u8 port_num); +int rxe_ib_get_hw_stats(struct ib_device *ibdev, + struct rdma_hw_stats *stats, + u8 port, int index); +#endif /* RXE_HW_COUNTERS_H */ diff --git a/drivers/infiniband/sw/rxe/rxe_icrc.c b/drivers/infiniband/sw/rxe/rxe_icrc.c index 413b56b23a06..39e0be31aab1 100644 --- a/drivers/infiniband/sw/rxe/rxe_icrc.c +++ b/drivers/infiniband/sw/rxe/rxe_icrc.c @@ -87,10 +87,10 @@ u32 rxe_icrc_hdr(struct rxe_pkt_info *pkt, struct sk_buff *skb) bth->qpn |= cpu_to_be32(~BTH_QPN_MASK); length = hdr_size + RXE_BTH_BYTES; - crc = crc32_le(crc, pshdr, length); + crc = rxe_crc32(pkt->rxe, crc, pshdr, length); /* And finish to compute the CRC on the remainder of the headers. */ - crc = crc32_le(crc, pkt->hdr + RXE_BTH_BYTES, - rxe_opcode[pkt->opcode].length - RXE_BTH_BYTES); + crc = rxe_crc32(pkt->rxe, crc, pkt->hdr + RXE_BTH_BYTES, + rxe_opcode[pkt->opcode].length - RXE_BTH_BYTES); return crc; } diff --git a/drivers/infiniband/sw/rxe/rxe_loc.h b/drivers/infiniband/sw/rxe/rxe_loc.h index 183a9d379b41..d6299edf9a5b 100644 --- a/drivers/infiniband/sw/rxe/rxe_loc.h +++ b/drivers/infiniband/sw/rxe/rxe_loc.h @@ -36,17 +36,17 @@ /* rxe_av.c */ -int rxe_av_chk_attr(struct rxe_dev *rxe, struct ib_ah_attr *attr); +int rxe_av_chk_attr(struct rxe_dev *rxe, struct rdma_ah_attr *attr); int rxe_av_from_attr(struct rxe_dev *rxe, u8 port_num, - struct rxe_av *av, struct ib_ah_attr *attr); + struct rxe_av *av, struct rdma_ah_attr *attr); int rxe_av_to_attr(struct rxe_dev *rxe, struct rxe_av *av, - struct ib_ah_attr *attr); + struct rdma_ah_attr *attr); int rxe_av_fill_ip_info(struct rxe_dev *rxe, struct rxe_av *av, - struct ib_ah_attr *attr, + struct rdma_ah_attr *attr, struct ib_gid_attr *sgid_attr, union ib_gid *sgid); @@ -145,7 +145,6 @@ int advance_dma_data(struct rxe_dma_info *dma, unsigned int length); int rxe_loopback(struct sk_buff *skb); int rxe_send(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, struct sk_buff *skb); -__be64 rxe_port_guid(struct rxe_dev *rxe); struct sk_buff *rxe_init_packet(struct rxe_dev *rxe, struct rxe_av *av, int paylen, struct rxe_pkt_info *pkt); int rxe_prepare(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, @@ -153,7 +152,6 @@ int rxe_prepare(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, enum rdma_link_layer rxe_link_layer(struct rxe_dev *rxe, unsigned int port_num); const char *rxe_parent_name(struct rxe_dev *rxe, unsigned int port_num); struct device *rxe_dma_device(struct rxe_dev *rxe); -__be64 rxe_node_guid(struct rxe_dev *rxe); int rxe_mcast_add(struct rxe_dev *rxe, union ib_gid *mgid); int rxe_mcast_delete(struct rxe_dev *rxe, union ib_gid *mgid); @@ -278,6 +276,7 @@ static inline int rxe_xmit_packet(struct rxe_dev *rxe, struct rxe_qp *qp, if (err) { rxe->xmit_errors++; + rxe_counter_inc(rxe, RXE_CNT_SEND_ERR); return err; } @@ -287,6 +286,7 @@ static inline int rxe_xmit_packet(struct rxe_dev *rxe, struct rxe_qp *qp, rxe_run_task(&qp->comp.task, 1); } + rxe_counter_inc(rxe, RXE_CNT_SENT_PKTS); goto done; drop: diff --git a/drivers/infiniband/sw/rxe/rxe_mr.c b/drivers/infiniband/sw/rxe/rxe_mr.c index 37eea7441ca4..e37cc89987e1 100644 --- a/drivers/infiniband/sw/rxe/rxe_mr.c +++ b/drivers/infiniband/sw/rxe/rxe_mr.c @@ -191,10 +191,8 @@ int rxe_mem_init_user(struct rxe_dev *rxe, struct rxe_pd *pd, u64 start, goto err1; } - WARN_ON_ONCE(!is_power_of_2(umem->page_size)); - - mem->page_shift = ilog2(umem->page_size); - mem->page_mask = umem->page_size - 1; + mem->page_shift = umem->page_shift; + mem->page_mask = BIT(umem->page_shift) - 1; num_buf = 0; map = mem->map; @@ -210,7 +208,7 @@ int rxe_mem_init_user(struct rxe_dev *rxe, struct rxe_pd *pd, u64 start, } buf->addr = (uintptr_t)vaddr; - buf->size = umem->page_size; + buf->size = BIT(umem->page_shift); num_buf++; buf++; @@ -370,7 +368,8 @@ int rxe_mem_copy(struct rxe_mem *mem, u64 iova, void *addr, int length, ((void *)(uintptr_t)iova) : addr; if (crcp) - *crcp = crc32_le(*crcp, src, length); + *crcp = rxe_crc32(to_rdev(mem->pd->ibpd.device), + *crcp, src, length); memcpy(dest, src, length); @@ -403,7 +402,8 @@ int rxe_mem_copy(struct rxe_mem *mem, u64 iova, void *addr, int length, bytes = length; if (crcp) - crc = crc32_le(crc, src, bytes); + crc = rxe_crc32(to_rdev(mem->pd->ibpd.device), + crc, src, bytes); memcpy(dest, src, bytes); diff --git a/drivers/infiniband/sw/rxe/rxe_net.c b/drivers/infiniband/sw/rxe/rxe_net.c index d8610960630a..c3a140ed4df2 100644 --- a/drivers/infiniband/sw/rxe/rxe_net.c +++ b/drivers/infiniband/sw/rxe/rxe_net.c @@ -84,34 +84,6 @@ struct rxe_dev *get_rxe_by_name(const char *name) struct rxe_recv_sockets recv_sockets; -static __be64 rxe_mac_to_eui64(struct net_device *ndev) -{ - unsigned char *mac_addr = ndev->dev_addr; - __be64 eui64; - unsigned char *dst = (unsigned char *)&eui64; - - dst[0] = mac_addr[0] ^ 2; - dst[1] = mac_addr[1]; - dst[2] = mac_addr[2]; - dst[3] = 0xff; - dst[4] = 0xfe; - dst[5] = mac_addr[3]; - dst[6] = mac_addr[4]; - dst[7] = mac_addr[5]; - - return eui64; -} - -__be64 rxe_node_guid(struct rxe_dev *rxe) -{ - return rxe_mac_to_eui64(rxe->ndev); -} - -__be64 rxe_port_guid(struct rxe_dev *rxe) -{ - return rxe_mac_to_eui64(rxe->ndev); -} - struct device *rxe_dma_device(struct rxe_dev *rxe) { struct net_device *ndev; @@ -210,6 +182,39 @@ static struct dst_entry *rxe_find_route6(struct net_device *ndev, #endif +static struct dst_entry *rxe_find_route(struct rxe_dev *rxe, + struct rxe_qp *qp, + struct rxe_av *av) +{ + struct dst_entry *dst = NULL; + + if (qp_type(qp) == IB_QPT_RC) + dst = sk_dst_get(qp->sk->sk); + + if (!dst || !(dst->obsolete && dst->ops->check(dst, 0))) { + if (dst) + dst_release(dst); + + if (av->network_type == RDMA_NETWORK_IPV4) { + struct in_addr *saddr; + struct in_addr *daddr; + + saddr = &av->sgid_addr._sockaddr_in.sin_addr; + daddr = &av->dgid_addr._sockaddr_in.sin_addr; + dst = rxe_find_route4(rxe->ndev, saddr, daddr); + } else if (av->network_type == RDMA_NETWORK_IPV6) { + struct in6_addr *saddr6; + struct in6_addr *daddr6; + + saddr6 = &av->sgid_addr._sockaddr_in6.sin6_addr; + daddr6 = &av->dgid_addr._sockaddr_in6.sin6_addr; + dst = rxe_find_route6(rxe->ndev, saddr6, daddr6); + } + } + + return dst; +} + static int rxe_udp_encap_recv(struct sock *sk, struct sk_buff *skb) { struct udphdr *udph; @@ -301,7 +306,7 @@ static void prepare_ipv4_hdr(struct dst_entry *dst, struct sk_buff *skb, skb_scrub_packet(skb, xnet); skb_clear_hash(skb); - skb_dst_set(skb, dst); + skb_dst_set(skb, dst_clone(dst)); memset(IPCB(skb), 0, sizeof(*IPCB(skb))); skb_push(skb, sizeof(struct iphdr)); @@ -349,13 +354,14 @@ static void prepare_ipv6_hdr(struct dst_entry *dst, struct sk_buff *skb, static int prepare4(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, struct sk_buff *skb, struct rxe_av *av) { + struct rxe_qp *qp = pkt->qp; struct dst_entry *dst; bool xnet = false; __be16 df = htons(IP_DF); struct in_addr *saddr = &av->sgid_addr._sockaddr_in.sin_addr; struct in_addr *daddr = &av->dgid_addr._sockaddr_in.sin_addr; - dst = rxe_find_route4(rxe->ndev, saddr, daddr); + dst = rxe_find_route(rxe, qp, av); if (!dst) { pr_err("Host not reachable\n"); return -EHOSTUNREACH; @@ -369,17 +375,24 @@ static int prepare4(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, prepare_ipv4_hdr(dst, skb, saddr->s_addr, daddr->s_addr, IPPROTO_UDP, av->grh.traffic_class, av->grh.hop_limit, df, xnet); + + if (qp_type(qp) == IB_QPT_RC) + sk_dst_set(qp->sk->sk, dst); + else + dst_release(dst); + return 0; } static int prepare6(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, struct sk_buff *skb, struct rxe_av *av) { - struct dst_entry *dst; + struct rxe_qp *qp = pkt->qp; + struct dst_entry *dst = NULL; struct in6_addr *saddr = &av->sgid_addr._sockaddr_in6.sin6_addr; struct in6_addr *daddr = &av->dgid_addr._sockaddr_in6.sin6_addr; - dst = rxe_find_route6(rxe->ndev, saddr, daddr); + dst = rxe_find_route(rxe, qp, av); if (!dst) { pr_err("Host not reachable\n"); return -EHOSTUNREACH; @@ -394,6 +407,12 @@ static int prepare6(struct rxe_dev *rxe, struct rxe_pkt_info *pkt, prepare_ipv6_hdr(dst, skb, saddr, daddr, IPPROTO_UDP, av->grh.traffic_class, av->grh.hop_limit); + + if (qp_type(qp) == IB_QPT_RC) + sk_dst_set(qp->sk->sk, dst); + else + dst_release(dst); + return 0; } diff --git a/drivers/infiniband/sw/rxe/rxe_param.h b/drivers/infiniband/sw/rxe/rxe_param.h index 13ed2cc6eaa2..1b596fbbe251 100644 --- a/drivers/infiniband/sw/rxe/rxe_param.h +++ b/drivers/infiniband/sw/rxe/rxe_param.h @@ -114,7 +114,6 @@ enum rxe_device_param { RXE_MAX_UCONTEXT = 512, RXE_NUM_PORT = 1, - RXE_NUM_COMP_VECTORS = 1, RXE_MIN_QP_INDEX = 16, RXE_MAX_QP_INDEX = 0x00020000, diff --git a/drivers/infiniband/sw/rxe/rxe_qp.c b/drivers/infiniband/sw/rxe/rxe_qp.c index f98a19e61a3d..80ccc7c7c341 100644 --- a/drivers/infiniband/sw/rxe/rxe_qp.c +++ b/drivers/infiniband/sw/rxe/rxe_qp.c @@ -273,10 +273,11 @@ static int rxe_qp_init_req(struct rxe_dev *rxe, struct rxe_qp *qp, rxe_init_task(rxe, &qp->comp.task, qp, rxe_completer, "comp"); - setup_timer(&qp->rnr_nak_timer, rnr_nak_timer, (unsigned long)qp); - setup_timer(&qp->retrans_timer, retransmit_timer, (unsigned long)qp); qp->qp_timeout_jiffies = 0; /* Can't be set for UD/UC in modify_qp */ - + if (init->qp_type == IB_QPT_RC) { + setup_timer(&qp->rnr_nak_timer, rnr_nak_timer, (unsigned long)qp); + setup_timer(&qp->retrans_timer, retransmit_timer, (unsigned long)qp); + } return 0; } @@ -630,8 +631,8 @@ int rxe_qp_from_attr(struct rxe_qp *qp, struct ib_qp_attr *attr, int mask, if (mask & IB_QP_AV) { ib_get_cached_gid(&rxe->ib_dev, 1, - attr->ah_attr.grh.sgid_index, &sgid, - &sgid_attr); + rdma_ah_read_grh(&attr->ah_attr)->sgid_index, + &sgid, &sgid_attr); rxe_av_from_attr(rxe, attr->port_num, &qp->pri_av, &attr->ah_attr); rxe_av_fill_ip_info(rxe, &qp->pri_av, &attr->ah_attr, @@ -641,9 +642,11 @@ int rxe_qp_from_attr(struct rxe_qp *qp, struct ib_qp_attr *attr, int mask, } if (mask & IB_QP_ALT_PATH) { - ib_get_cached_gid(&rxe->ib_dev, 1, - attr->alt_ah_attr.grh.sgid_index, &sgid, - &sgid_attr); + u8 sgid_index = + rdma_ah_read_grh(&attr->alt_ah_attr)->sgid_index; + + ib_get_cached_gid(&rxe->ib_dev, 1, sgid_index, + &sgid, &sgid_attr); rxe_av_from_attr(rxe, attr->alt_port_num, &qp->alt_av, &attr->alt_ah_attr); @@ -804,8 +807,10 @@ void rxe_qp_destroy(struct rxe_qp *qp) qp->qp_timeout_jiffies = 0; rxe_cleanup_task(&qp->resp.task); - del_timer_sync(&qp->retrans_timer); - del_timer_sync(&qp->rnr_nak_timer); + if (qp_type(qp) == IB_QPT_RC) { + del_timer_sync(&qp->retrans_timer); + del_timer_sync(&qp->rnr_nak_timer); + } rxe_cleanup_task(&qp->req.task); rxe_cleanup_task(&qp->comp.task); @@ -846,6 +851,14 @@ void rxe_qp_cleanup(struct rxe_pool_entry *arg) qp->resp.mr = NULL; } + if (qp_type(qp) == IB_QPT_RC) { + struct dst_entry *dst = NULL; + + dst = sk_dst_get(qp->sk->sk); + if (dst) + dst_release(dst); + } + free_rd_atomic_resources(qp); kernel_sock_shutdown(qp->sk, SHUT_RDWR); diff --git a/drivers/infiniband/sw/rxe/rxe_recv.c b/drivers/infiniband/sw/rxe/rxe_recv.c index 50886031096f..fb8c83e055e1 100644 --- a/drivers/infiniband/sw/rxe/rxe_recv.c +++ b/drivers/infiniband/sw/rxe/rxe_recv.c @@ -387,8 +387,8 @@ int rxe_rcv(struct sk_buff *skb) pack_icrc = be32_to_cpu(*icrcp); calc_icrc = rxe_icrc_hdr(pkt, skb); - calc_icrc = crc32_le(calc_icrc, (u8 *)payload_addr(pkt), - payload_size(pkt)); + calc_icrc = rxe_crc32(rxe, calc_icrc, (u8 *)payload_addr(pkt), + payload_size(pkt)); calc_icrc = (__force u32)cpu_to_be32(~calc_icrc); if (unlikely(calc_icrc != pack_icrc)) { if (skb->protocol == htons(ETH_P_IPV6)) @@ -403,6 +403,8 @@ int rxe_rcv(struct sk_buff *skb) goto drop; } + rxe_counter_inc(rxe, RXE_CNT_RCVD_PKTS); + if (unlikely(bth_qpn(pkt) == IB_MULTICAST_QPN)) rxe_rcv_mcast_pkt(rxe, skb); else @@ -417,4 +419,3 @@ drop: kfree_skb(skb); return 0; } -EXPORT_SYMBOL(rxe_rcv); diff --git a/drivers/infiniband/sw/rxe/rxe_req.c b/drivers/infiniband/sw/rxe/rxe_req.c index 9f95f50b2909..7ee465d1a1e1 100644 --- a/drivers/infiniband/sw/rxe/rxe_req.c +++ b/drivers/infiniband/sw/rxe/rxe_req.c @@ -32,6 +32,7 @@ */ #include <linux/skbuff.h> +#include <crypto/hash.h> #include "rxe.h" #include "rxe_loc.h" @@ -483,8 +484,7 @@ static int fill_packet(struct rxe_qp *qp, struct rxe_send_wqe *wqe, if (wqe->wr.send_flags & IB_SEND_INLINE) { u8 *tmp = &wqe->dma.inline_data[wqe->dma.sge_offset]; - crc = crc32_le(crc, tmp, paylen); - + crc = rxe_crc32(rxe, crc, tmp, paylen); memcpy(payload_addr(pkt), tmp, paylen); wqe->dma.resid -= paylen; diff --git a/drivers/infiniband/sw/rxe/rxe_resp.c b/drivers/infiniband/sw/rxe/rxe_resp.c index c9dd385ce62e..23039768f541 100644 --- a/drivers/infiniband/sw/rxe/rxe_resp.c +++ b/drivers/infiniband/sw/rxe/rxe_resp.c @@ -149,6 +149,7 @@ static enum resp_states check_psn(struct rxe_qp *qp, struct rxe_pkt_info *pkt) { int diff = psn_compare(pkt->psn, qp->resp.psn); + struct rxe_dev *rxe = to_rdev(qp->ibqp.device); switch (qp_type(qp)) { case IB_QPT_RC: @@ -157,9 +158,11 @@ static enum resp_states check_psn(struct rxe_qp *qp, return RESPST_CLEANUP; qp->resp.sent_psn_nak = 1; + rxe_counter_inc(rxe, RXE_CNT_OUT_OF_SEQ_REQ); return RESPST_ERR_PSN_OUT_OF_SEQ; } else if (diff < 0) { + rxe_counter_inc(rxe, RXE_CNT_DUP_REQ); return RESPST_DUPLICATE_REQUEST; } @@ -478,8 +481,6 @@ static enum resp_states check_rkey(struct rxe_qp *qp, state = RESPST_ERR_LENGTH; goto err; } - - qp->resp.resid = mtu; } else { if (pktlen != resid) { state = RESPST_ERR_LENGTH; @@ -1223,6 +1224,7 @@ void rxe_drain_req_pkts(struct rxe_qp *qp, bool notify) int rxe_responder(void *arg) { struct rxe_qp *qp = (struct rxe_qp *)arg; + struct rxe_dev *rxe = to_rdev(qp->ibqp.device); enum resp_states state; struct rxe_pkt_info *pkt = NULL; int ret = 0; @@ -1311,6 +1313,7 @@ int rxe_responder(void *arg) break; case RESPST_ERR_RNR: if (qp_type(qp) == IB_QPT_RC) { + rxe_counter_inc(rxe, RXE_CNT_SND_RNR); /* RC - class B */ send_ack(qp, pkt, AETH_RNR_NAK | (~AETH_TYPE_MASK & diff --git a/drivers/infiniband/sw/rxe/rxe_verbs.c b/drivers/infiniband/sw/rxe/rxe_verbs.c index 5113e502f6f9..83d709e74dfb 100644 --- a/drivers/infiniband/sw/rxe/rxe_verbs.c +++ b/drivers/infiniband/sw/rxe/rxe_verbs.c @@ -32,9 +32,11 @@ */ #include <linux/dma-mapping.h> +#include <net/addrconf.h> #include "rxe.h" #include "rxe_loc.h" #include "rxe_queue.h" +#include "rxe_hw_counters.h" static int rxe_query_device(struct ib_device *dev, struct ib_device_attr *attr, @@ -295,22 +297,22 @@ static int rxe_dealloc_pd(struct ib_pd *ibpd) return 0; } -static int rxe_init_av(struct rxe_dev *rxe, struct ib_ah_attr *attr, +static int rxe_init_av(struct rxe_dev *rxe, struct rdma_ah_attr *attr, struct rxe_av *av) { int err; union ib_gid sgid; struct ib_gid_attr sgid_attr; - err = ib_get_cached_gid(&rxe->ib_dev, attr->port_num, - attr->grh.sgid_index, &sgid, + err = ib_get_cached_gid(&rxe->ib_dev, rdma_ah_get_port_num(attr), + rdma_ah_read_grh(attr)->sgid_index, &sgid, &sgid_attr); if (err) { pr_err("Failed to query sgid. err = %d\n", err); return err; } - err = rxe_av_from_attr(rxe, attr->port_num, av, attr); + err = rxe_av_from_attr(rxe, rdma_ah_get_port_num(attr), av, attr); if (!err) err = rxe_av_fill_ip_info(rxe, av, attr, &sgid_attr, &sgid); @@ -319,7 +321,8 @@ static int rxe_init_av(struct rxe_dev *rxe, struct ib_ah_attr *attr, return err; } -static struct ib_ah *rxe_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr, +static struct ib_ah *rxe_create_ah(struct ib_pd *ibpd, + struct rdma_ah_attr *attr, struct ib_udata *udata) { @@ -354,7 +357,7 @@ err1: return ERR_PTR(err); } -static int rxe_modify_ah(struct ib_ah *ibah, struct ib_ah_attr *attr) +static int rxe_modify_ah(struct ib_ah *ibah, struct rdma_ah_attr *attr) { int err; struct rxe_dev *rxe = to_rdev(ibah->device); @@ -371,11 +374,13 @@ static int rxe_modify_ah(struct ib_ah *ibah, struct ib_ah_attr *attr) return 0; } -static int rxe_query_ah(struct ib_ah *ibah, struct ib_ah_attr *attr) +static int rxe_query_ah(struct ib_ah *ibah, struct rdma_ah_attr *attr) { struct rxe_dev *rxe = to_rdev(ibah->device); struct rxe_ah *ah = to_rah(ibah); + memset(attr, 0, sizeof(*attr)); + attr->type = ibah->type; rxe_av_to_attr(rxe, &ah->av, attr); return 0; } @@ -1234,10 +1239,11 @@ int rxe_register_device(struct rxe_dev *rxe) dev->owner = THIS_MODULE; dev->node_type = RDMA_NODE_IB_CA; dev->phys_port_cnt = 1; - dev->num_comp_vectors = RXE_NUM_COMP_VECTORS; + dev->num_comp_vectors = num_possible_cpus(); dev->dev.parent = rxe_dma_device(rxe); dev->local_dma_lkey = 0; - dev->node_guid = rxe_node_guid(rxe); + addrconf_addr_eui48((unsigned char *)&dev->node_guid, + rxe->ndev->dev_addr); dev->dev.dma_ops = &dma_virt_ops; dev->uverbs_abi_ver = RXE_UVERBS_ABI_VERSION; @@ -1318,6 +1324,15 @@ int rxe_register_device(struct rxe_dev *rxe) dev->map_mr_sg = rxe_map_mr_sg; dev->attach_mcast = rxe_attach_mcast; dev->detach_mcast = rxe_detach_mcast; + dev->get_hw_stats = rxe_ib_get_hw_stats; + dev->alloc_hw_stats = rxe_ib_alloc_hw_stats; + + rxe->tfm = crypto_alloc_shash("crc32", 0, 0); + if (IS_ERR(rxe->tfm)) { + pr_err("failed to allocate crc algorithm err:%ld\n", + PTR_ERR(rxe->tfm)); + return PTR_ERR(rxe->tfm); + } err = ib_register_device(dev, NULL); if (err) { @@ -1339,6 +1354,8 @@ int rxe_register_device(struct rxe_dev *rxe) err2: ib_unregister_device(dev); err1: + crypto_free_shash(rxe->tfm); + return err; } diff --git a/drivers/infiniband/sw/rxe/rxe_verbs.h b/drivers/infiniband/sw/rxe/rxe_verbs.h index e100c500ae85..5a180fbe40d9 100644 --- a/drivers/infiniband/sw/rxe/rxe_verbs.h +++ b/drivers/infiniband/sw/rxe/rxe_verbs.h @@ -38,6 +38,7 @@ #include <rdma/rdma_user_rxe.h> #include "rxe_pool.h" #include "rxe_task.h" +#include "rxe_hw_counters.h" static inline int pkey_match(u16 key1, u16 key2) { @@ -401,10 +402,18 @@ struct rxe_dev { spinlock_t mmap_offset_lock; /* guard mmap_offset */ int mmap_offset; + u64 stats_counters[RXE_NUM_OF_COUNTERS]; + struct rxe_port port; struct list_head list; + struct crypto_shash *tfm; }; +static inline void rxe_counter_inc(struct rxe_dev *rxe, enum rxe_counters cnt) +{ + rxe->stats_counters[cnt]++; +} + static inline struct rxe_dev *to_rdev(struct ib_device *dev) { return dev ? container_of(dev, struct rxe_dev, ib_dev) : NULL; |