From 470a55358186d0bb93558a87d13159dfbc989351 Mon Sep 17 00:00:00 2001 From: Bodong Wang Date: Tue, 22 Sep 2015 23:18:10 +0300 Subject: IB/core: Add support of checksum capability reporting for RC and RAW Two enum members IB_DEVICE_RC_IP_CSUM and IB_DEVICE_RAW_IP_CSUM are added to ib_device_cap_flags. Device should set these two flags if they support insertion of UDP and TCP checksum on outgoing IPv4 messages and can verify the validity of checksum for incoming IPv4 messages, for RC IPoIB and RAW over Ethernet respectively. They are similar to IB_DEVICE_UD_IP_CSUM. Signed-off-by: Bodong Wang Signed-off-by: Doug Ledford --- include/rdma/ib_verbs.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) (limited to 'include') diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index 7845fae6f2df..e1f65e204d37 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -137,6 +137,8 @@ enum ib_device_cap_flags { IB_DEVICE_BLOCK_MULTICAST_LOOPBACK = (1<<22), IB_DEVICE_MEM_WINDOW_TYPE_2A = (1<<23), IB_DEVICE_MEM_WINDOW_TYPE_2B = (1<<24), + IB_DEVICE_RC_IP_CSUM = (1<<25), + IB_DEVICE_RAW_IP_CSUM = (1<<26), IB_DEVICE_MANAGED_FLOW_STEERING = (1<<29), IB_DEVICE_SIGNATURE_HANDOVER = (1<<30), IB_DEVICE_ON_DEMAND_PAGING = (1<<31), @@ -873,7 +875,6 @@ enum ib_qp_create_flags { IB_QP_CREATE_RESERVED_END = 1 << 31, }; - /* * Note: users may not call ib_close_qp or ib_destroy_qp from the event_handler * callback to destroy the passed in QP. -- cgit v1.2.3 From e622f2f4ad2142d2a613a57fb85f8cf737935ef5 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 8 Oct 2015 09:16:33 +0100 Subject: IB: split struct ib_send_wr This patch split up struct ib_send_wr so that all non-trivial verbs use their own structure which embedds struct ib_send_wr. This dramaticly shrinks the size of a WR for most common operations: sizeof(struct ib_send_wr) (old): 96 sizeof(struct ib_send_wr): 48 sizeof(struct ib_rdma_wr): 64 sizeof(struct ib_atomic_wr): 96 sizeof(struct ib_ud_wr): 88 sizeof(struct ib_fast_reg_wr): 88 sizeof(struct ib_bind_mw_wr): 96 sizeof(struct ib_sig_handover_wr): 80 And with Sagi's pending MR rework the fast registration WR will also be down to a reasonable size: sizeof(struct ib_fastreg_wr): 64 Signed-off-by: Christoph Hellwig Reviewed-by: Bart Van Assche [srp, srpt] Reviewed-by: Chuck Lever [sunrpc] Tested-by: Haggai Eran Tested-by: Sagi Grimberg Tested-by: Steve Wise --- drivers/infiniband/core/agent.c | 2 +- drivers/infiniband/core/mad.c | 40 ++--- drivers/infiniband/core/mad_priv.h | 2 +- drivers/infiniband/core/uverbs_cmd.c | 143 ++++++++++------- drivers/infiniband/hw/cxgb3/iwch_qp.c | 36 ++--- drivers/infiniband/hw/cxgb4/qp.c | 46 +++--- drivers/infiniband/hw/mlx4/mad.c | 54 ++++--- drivers/infiniband/hw/mlx4/mr.c | 20 +-- drivers/infiniband/hw/mlx4/qp.c | 178 +++++++++++---------- drivers/infiniband/hw/mlx5/mlx5_ib.h | 6 + drivers/infiniband/hw/mlx5/mr.c | 52 +++--- drivers/infiniband/hw/mlx5/qp.c | 114 ++++++------- drivers/infiniband/hw/mthca/mthca_qp.c | 84 +++++----- drivers/infiniband/hw/nes/nes_verbs.c | 41 ++--- drivers/infiniband/hw/ocrdma/ocrdma_verbs.c | 53 +++--- drivers/infiniband/hw/qib/qib_keys.c | 23 +-- drivers/infiniband/hw/qib/qib_qp.c | 2 +- drivers/infiniband/hw/qib/qib_rc.c | 38 ++--- drivers/infiniband/hw/qib/qib_ruc.c | 20 +-- drivers/infiniband/hw/qib/qib_uc.c | 4 +- drivers/infiniband/hw/qib/qib_ud.c | 20 +-- drivers/infiniband/hw/qib/qib_verbs.c | 22 ++- drivers/infiniband/hw/qib/qib_verbs.h | 8 +- drivers/infiniband/ulp/ipoib/ipoib.h | 4 +- drivers/infiniband/ulp/ipoib/ipoib_cm.c | 4 +- drivers/infiniband/ulp/ipoib/ipoib_ib.c | 22 +-- drivers/infiniband/ulp/ipoib/ipoib_main.c | 2 +- drivers/infiniband/ulp/ipoib/ipoib_multicast.c | 2 +- drivers/infiniband/ulp/ipoib/ipoib_verbs.c | 6 +- drivers/infiniband/ulp/iser/iscsi_iser.h | 10 +- drivers/infiniband/ulp/iser/iser_memory.c | 72 ++++----- drivers/infiniband/ulp/iser/iser_verbs.c | 2 +- drivers/infiniband/ulp/isert/ib_isert.c | 156 +++++++++--------- drivers/infiniband/ulp/isert/ib_isert.h | 6 +- drivers/infiniband/ulp/srp/ib_srp.c | 26 +-- drivers/infiniband/ulp/srpt/ib_srpt.c | 32 ++-- .../staging/lustre/lnet/klnds/o2iblnd/o2iblnd.h | 2 +- .../staging/lustre/lnet/klnds/o2iblnd/o2iblnd_cb.c | 35 ++-- drivers/staging/rdma/amso1100/c2_qp.c | 8 +- drivers/staging/rdma/ehca/ehca_reqs.c | 53 +++--- drivers/staging/rdma/hfi1/keys.c | 22 +-- drivers/staging/rdma/hfi1/qp.c | 2 +- drivers/staging/rdma/hfi1/rc.c | 24 +-- drivers/staging/rdma/hfi1/ruc.c | 18 +-- drivers/staging/rdma/hfi1/uc.c | 4 +- drivers/staging/rdma/hfi1/ud.c | 20 +-- drivers/staging/rdma/hfi1/verbs.c | 23 ++- drivers/staging/rdma/hfi1/verbs.h | 10 +- drivers/staging/rdma/ipath/ipath_rc.c | 24 +-- drivers/staging/rdma/ipath/ipath_ruc.c | 16 +- drivers/staging/rdma/ipath/ipath_uc.c | 4 +- drivers/staging/rdma/ipath/ipath_ud.c | 26 +-- drivers/staging/rdma/ipath/ipath_verbs.c | 20 ++- drivers/staging/rdma/ipath/ipath_verbs.h | 9 +- include/rdma/ib_verbs.h | 130 +++++++++------ net/rds/ib.h | 6 +- net/rds/ib_send.c | 71 ++++---- net/rds/iw.h | 6 +- net/rds/iw_rdma.c | 29 ++-- net/rds/iw_send.c | 113 ++++++------- net/sunrpc/xprtrdma/frwr_ops.c | 23 +-- net/sunrpc/xprtrdma/svc_rdma_recvfrom.c | 70 ++++---- net/sunrpc/xprtrdma/svc_rdma_sendto.c | 18 +-- 63 files changed, 1152 insertions(+), 986 deletions(-) (limited to 'include') diff --git a/drivers/infiniband/core/agent.c b/drivers/infiniband/core/agent.c index 0429040304fd..4fa524dfb6cf 100644 --- a/drivers/infiniband/core/agent.c +++ b/drivers/infiniband/core/agent.c @@ -126,7 +126,7 @@ void agent_send_response(const struct ib_mad_hdr *mad_hdr, const struct ib_grh * mad_send_wr = container_of(send_buf, struct ib_mad_send_wr_private, send_buf); - mad_send_wr->send_wr.wr.ud.port_num = port_num; + mad_send_wr->send_wr.port_num = port_num; } if (ib_post_send_mad(send_buf, NULL)) { diff --git a/drivers/infiniband/core/mad.c b/drivers/infiniband/core/mad.c index 4b5c72311deb..844d9bb22700 100644 --- a/drivers/infiniband/core/mad.c +++ b/drivers/infiniband/core/mad.c @@ -752,7 +752,7 @@ static int handle_outgoing_dr_smp(struct ib_mad_agent_private *mad_agent_priv, struct ib_device *device = mad_agent_priv->agent.device; u8 port_num; struct ib_wc mad_wc; - struct ib_send_wr *send_wr = &mad_send_wr->send_wr; + struct ib_ud_wr *send_wr = &mad_send_wr->send_wr; size_t mad_size = port_mad_size(mad_agent_priv->qp_info->port_priv); u16 out_mad_pkey_index = 0; u16 drslid; @@ -761,7 +761,7 @@ static int handle_outgoing_dr_smp(struct ib_mad_agent_private *mad_agent_priv, if (rdma_cap_ib_switch(device) && smp->mgmt_class == IB_MGMT_CLASS_SUBN_DIRECTED_ROUTE) - port_num = send_wr->wr.ud.port_num; + port_num = send_wr->port_num; else port_num = mad_agent_priv->agent.port_num; @@ -832,9 +832,9 @@ static int handle_outgoing_dr_smp(struct ib_mad_agent_private *mad_agent_priv, } build_smp_wc(mad_agent_priv->agent.qp, - send_wr->wr_id, drslid, - send_wr->wr.ud.pkey_index, - send_wr->wr.ud.port_num, &mad_wc); + send_wr->wr.wr_id, drslid, + send_wr->pkey_index, + send_wr->port_num, &mad_wc); if (opa && smp->base_version == OPA_MGMT_BASE_VERSION) { mad_wc.byte_len = mad_send_wr->send_buf.hdr_len @@ -894,7 +894,7 @@ static int handle_outgoing_dr_smp(struct ib_mad_agent_private *mad_agent_priv, local->mad_send_wr = mad_send_wr; if (opa) { - local->mad_send_wr->send_wr.wr.ud.pkey_index = out_mad_pkey_index; + local->mad_send_wr->send_wr.pkey_index = out_mad_pkey_index; local->return_wc_byte_len = mad_size; } /* Reference MAD agent until send side of local completion handled */ @@ -1039,14 +1039,14 @@ struct ib_mad_send_buf * ib_create_send_mad(struct ib_mad_agent *mad_agent, mad_send_wr->sg_list[1].lkey = mad_agent->qp->pd->local_dma_lkey; - mad_send_wr->send_wr.wr_id = (unsigned long) mad_send_wr; - mad_send_wr->send_wr.sg_list = mad_send_wr->sg_list; - mad_send_wr->send_wr.num_sge = 2; - mad_send_wr->send_wr.opcode = IB_WR_SEND; - mad_send_wr->send_wr.send_flags = IB_SEND_SIGNALED; - mad_send_wr->send_wr.wr.ud.remote_qpn = remote_qpn; - mad_send_wr->send_wr.wr.ud.remote_qkey = IB_QP_SET_QKEY; - mad_send_wr->send_wr.wr.ud.pkey_index = pkey_index; + mad_send_wr->send_wr.wr.wr_id = (unsigned long) mad_send_wr; + mad_send_wr->send_wr.wr.sg_list = mad_send_wr->sg_list; + mad_send_wr->send_wr.wr.num_sge = 2; + mad_send_wr->send_wr.wr.opcode = IB_WR_SEND; + mad_send_wr->send_wr.wr.send_flags = IB_SEND_SIGNALED; + mad_send_wr->send_wr.remote_qpn = remote_qpn; + mad_send_wr->send_wr.remote_qkey = IB_QP_SET_QKEY; + mad_send_wr->send_wr.pkey_index = pkey_index; if (rmpp_active) { ret = alloc_send_rmpp_list(mad_send_wr, mad_size, gfp_mask); @@ -1151,7 +1151,7 @@ int ib_send_mad(struct ib_mad_send_wr_private *mad_send_wr) /* Set WR ID to find mad_send_wr upon completion */ qp_info = mad_send_wr->mad_agent_priv->qp_info; - mad_send_wr->send_wr.wr_id = (unsigned long)&mad_send_wr->mad_list; + mad_send_wr->send_wr.wr.wr_id = (unsigned long)&mad_send_wr->mad_list; mad_send_wr->mad_list.mad_queue = &qp_info->send_queue; mad_agent = mad_send_wr->send_buf.mad_agent; @@ -1179,7 +1179,7 @@ int ib_send_mad(struct ib_mad_send_wr_private *mad_send_wr) spin_lock_irqsave(&qp_info->send_queue.lock, flags); if (qp_info->send_queue.count < qp_info->send_queue.max_active) { - ret = ib_post_send(mad_agent->qp, &mad_send_wr->send_wr, + ret = ib_post_send(mad_agent->qp, &mad_send_wr->send_wr.wr, &bad_send_wr); list = &qp_info->send_queue.list; } else { @@ -1244,7 +1244,7 @@ int ib_post_send_mad(struct ib_mad_send_buf *send_buf, * request associated with the completion */ next_send_buf = send_buf->next; - mad_send_wr->send_wr.wr.ud.ah = send_buf->ah; + mad_send_wr->send_wr.ah = send_buf->ah; if (((struct ib_mad_hdr *) send_buf->mad)->mgmt_class == IB_MGMT_CLASS_SUBN_DIRECTED_ROUTE) { @@ -2457,7 +2457,7 @@ retry: ib_mad_complete_send_wr(mad_send_wr, &mad_send_wc); if (queued_send_wr) { - ret = ib_post_send(qp_info->qp, &queued_send_wr->send_wr, + ret = ib_post_send(qp_info->qp, &queued_send_wr->send_wr.wr, &bad_send_wr); if (ret) { dev_err(&port_priv->device->dev, @@ -2515,7 +2515,7 @@ static void mad_error_handler(struct ib_mad_port_private *port_priv, struct ib_send_wr *bad_send_wr; mad_send_wr->retry = 0; - ret = ib_post_send(qp_info->qp, &mad_send_wr->send_wr, + ret = ib_post_send(qp_info->qp, &mad_send_wr->send_wr.wr, &bad_send_wr); if (ret) ib_mad_send_done_handler(port_priv, wc); @@ -2713,7 +2713,7 @@ static void local_completions(struct work_struct *work) build_smp_wc(recv_mad_agent->agent.qp, (unsigned long) local->mad_send_wr, be16_to_cpu(IB_LID_PERMISSIVE), - local->mad_send_wr->send_wr.wr.ud.pkey_index, + local->mad_send_wr->send_wr.pkey_index, recv_mad_agent->agent.port_num, &wc); local->mad_priv->header.recv_wc.wc = &wc; diff --git a/drivers/infiniband/core/mad_priv.h b/drivers/infiniband/core/mad_priv.h index 4a4f7aad0978..990698a6ab4b 100644 --- a/drivers/infiniband/core/mad_priv.h +++ b/drivers/infiniband/core/mad_priv.h @@ -123,7 +123,7 @@ struct ib_mad_send_wr_private { struct ib_mad_send_buf send_buf; u64 header_mapping; u64 payload_mapping; - struct ib_send_wr send_wr; + struct ib_ud_wr send_wr; struct ib_sge sg_list[IB_MAD_SEND_REQ_MAX_SG]; __be64 tid; unsigned long timeout; diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c index be4cb9f04be3..8adb71fd6a3a 100644 --- a/drivers/infiniband/core/uverbs_cmd.c +++ b/drivers/infiniband/core/uverbs_cmd.c @@ -2303,6 +2303,12 @@ ssize_t ib_uverbs_destroy_qp(struct ib_uverbs_file *file, return in_len; } +static void *alloc_wr(size_t wr_size, __u32 num_sge) +{ + return kmalloc(ALIGN(wr_size, sizeof (struct ib_sge)) + + num_sge * sizeof (struct ib_sge), GFP_KERNEL); +}; + ssize_t ib_uverbs_post_send(struct ib_uverbs_file *file, struct ib_device *ib_dev, const char __user *buf, int in_len, @@ -2351,14 +2357,83 @@ ssize_t ib_uverbs_post_send(struct ib_uverbs_file *file, goto out_put; } - next = kmalloc(ALIGN(sizeof *next, sizeof (struct ib_sge)) + - user_wr->num_sge * sizeof (struct ib_sge), - GFP_KERNEL); - if (!next) { - ret = -ENOMEM; + if (is_ud) { + struct ib_ud_wr *ud; + + if (user_wr->opcode != IB_WR_SEND && + user_wr->opcode != IB_WR_SEND_WITH_IMM) { + ret = -EINVAL; + goto out_put; + } + + ud = alloc_wr(sizeof(*ud), user_wr->num_sge); + if (!ud) { + ret = -ENOMEM; + goto out_put; + } + + ud->ah = idr_read_ah(user_wr->wr.ud.ah, file->ucontext); + if (!ud->ah) { + kfree(ud); + ret = -EINVAL; + goto out_put; + } + ud->remote_qpn = user_wr->wr.ud.remote_qpn; + ud->remote_qkey = user_wr->wr.ud.remote_qkey; + + next = &ud->wr; + } else if (user_wr->opcode == IB_WR_RDMA_WRITE_WITH_IMM || + user_wr->opcode == IB_WR_RDMA_WRITE || + user_wr->opcode == IB_WR_RDMA_READ) { + struct ib_rdma_wr *rdma; + + rdma = alloc_wr(sizeof(*rdma), user_wr->num_sge); + if (!rdma) { + ret = -ENOMEM; + goto out_put; + } + + rdma->remote_addr = user_wr->wr.rdma.remote_addr; + rdma->rkey = user_wr->wr.rdma.rkey; + + next = &rdma->wr; + } else if (user_wr->opcode == IB_WR_ATOMIC_CMP_AND_SWP || + user_wr->opcode == IB_WR_ATOMIC_FETCH_AND_ADD) { + struct ib_atomic_wr *atomic; + + atomic = alloc_wr(sizeof(*atomic), user_wr->num_sge); + if (!atomic) { + ret = -ENOMEM; + goto out_put; + } + + atomic->remote_addr = user_wr->wr.atomic.remote_addr; + atomic->compare_add = user_wr->wr.atomic.compare_add; + atomic->swap = user_wr->wr.atomic.swap; + atomic->rkey = user_wr->wr.atomic.rkey; + + next = &atomic->wr; + } else if (user_wr->opcode == IB_WR_SEND || + user_wr->opcode == IB_WR_SEND_WITH_IMM || + user_wr->opcode == IB_WR_SEND_WITH_INV) { + next = alloc_wr(sizeof(*next), user_wr->num_sge); + if (!next) { + ret = -ENOMEM; + goto out_put; + } + } else { + ret = -EINVAL; goto out_put; } + if (user_wr->opcode == IB_WR_SEND_WITH_IMM || + user_wr->opcode == IB_WR_RDMA_WRITE_WITH_IMM) { + next->ex.imm_data = + (__be32 __force) user_wr->ex.imm_data; + } else if (user_wr->opcode == IB_WR_SEND_WITH_INV) { + next->ex.invalidate_rkey = user_wr->ex.invalidate_rkey; + } + if (!last) wr = next; else @@ -2371,60 +2446,6 @@ ssize_t ib_uverbs_post_send(struct ib_uverbs_file *file, next->opcode = user_wr->opcode; next->send_flags = user_wr->send_flags; - if (is_ud) { - if (next->opcode != IB_WR_SEND && - next->opcode != IB_WR_SEND_WITH_IMM) { - ret = -EINVAL; - goto out_put; - } - - next->wr.ud.ah = idr_read_ah(user_wr->wr.ud.ah, - file->ucontext); - if (!next->wr.ud.ah) { - ret = -EINVAL; - goto out_put; - } - next->wr.ud.remote_qpn = user_wr->wr.ud.remote_qpn; - next->wr.ud.remote_qkey = user_wr->wr.ud.remote_qkey; - if (next->opcode == IB_WR_SEND_WITH_IMM) - next->ex.imm_data = - (__be32 __force) user_wr->ex.imm_data; - } else { - switch (next->opcode) { - case IB_WR_RDMA_WRITE_WITH_IMM: - next->ex.imm_data = - (__be32 __force) user_wr->ex.imm_data; - case IB_WR_RDMA_WRITE: - case IB_WR_RDMA_READ: - next->wr.rdma.remote_addr = - user_wr->wr.rdma.remote_addr; - next->wr.rdma.rkey = - user_wr->wr.rdma.rkey; - break; - case IB_WR_SEND_WITH_IMM: - next->ex.imm_data = - (__be32 __force) user_wr->ex.imm_data; - break; - case IB_WR_SEND_WITH_INV: - next->ex.invalidate_rkey = - user_wr->ex.invalidate_rkey; - break; - case IB_WR_ATOMIC_CMP_AND_SWP: - case IB_WR_ATOMIC_FETCH_AND_ADD: - next->wr.atomic.remote_addr = - user_wr->wr.atomic.remote_addr; - next->wr.atomic.compare_add = - user_wr->wr.atomic.compare_add; - next->wr.atomic.swap = user_wr->wr.atomic.swap; - next->wr.atomic.rkey = user_wr->wr.atomic.rkey; - case IB_WR_SEND: - break; - default: - ret = -EINVAL; - goto out_put; - } - } - if (next->num_sge) { next->sg_list = (void *) next + ALIGN(sizeof *next, sizeof (struct ib_sge)); @@ -2458,8 +2479,8 @@ out_put: put_qp_read(qp); while (wr) { - if (is_ud && wr->wr.ud.ah) - put_ah_read(wr->wr.ud.ah); + if (is_ud && ud_wr(wr)->ah) + put_ah_read(ud_wr(wr)->ah); next = wr->next; kfree(wr); wr = next; diff --git a/drivers/infiniband/hw/cxgb3/iwch_qp.c b/drivers/infiniband/hw/cxgb3/iwch_qp.c index b57c0befd962..bac0508fedd9 100644 --- a/drivers/infiniband/hw/cxgb3/iwch_qp.c +++ b/drivers/infiniband/hw/cxgb3/iwch_qp.c @@ -95,8 +95,8 @@ static int build_rdma_write(union t3_wr *wqe, struct ib_send_wr *wr, wqe->write.reserved[0] = 0; wqe->write.reserved[1] = 0; wqe->write.reserved[2] = 0; - wqe->write.stag_sink = cpu_to_be32(wr->wr.rdma.rkey); - wqe->write.to_sink = cpu_to_be64(wr->wr.rdma.remote_addr); + wqe->write.stag_sink = cpu_to_be32(rdma_wr(wr)->rkey); + wqe->write.to_sink = cpu_to_be64(rdma_wr(wr)->remote_addr); if (wr->opcode == IB_WR_RDMA_WRITE_WITH_IMM) { plen = 4; @@ -137,8 +137,8 @@ static int build_rdma_read(union t3_wr *wqe, struct ib_send_wr *wr, wqe->read.local_inv = 0; wqe->read.reserved[0] = 0; wqe->read.reserved[1] = 0; - wqe->read.rem_stag = cpu_to_be32(wr->wr.rdma.rkey); - wqe->read.rem_to = cpu_to_be64(wr->wr.rdma.remote_addr); + wqe->read.rem_stag = cpu_to_be32(rdma_wr(wr)->rkey); + wqe->read.rem_to = cpu_to_be64(rdma_wr(wr)->remote_addr); wqe->read.local_stag = cpu_to_be32(wr->sg_list[0].lkey); wqe->read.local_len = cpu_to_be32(wr->sg_list[0].length); wqe->read.local_to = cpu_to_be64(wr->sg_list[0].addr); @@ -146,27 +146,27 @@ static int build_rdma_read(union t3_wr *wqe, struct ib_send_wr *wr, return 0; } -static int build_fastreg(union t3_wr *wqe, struct ib_send_wr *wr, +static int build_fastreg(union t3_wr *wqe, struct ib_send_wr *send_wr, u8 *flit_cnt, int *wr_cnt, struct t3_wq *wq) { + struct ib_fast_reg_wr *wr = fast_reg_wr(send_wr); int i; __be64 *p; - if (wr->wr.fast_reg.page_list_len > T3_MAX_FASTREG_DEPTH) + if (wr->page_list_len > T3_MAX_FASTREG_DEPTH) return -EINVAL; *wr_cnt = 1; - wqe->fastreg.stag = cpu_to_be32(wr->wr.fast_reg.rkey); - wqe->fastreg.len = cpu_to_be32(wr->wr.fast_reg.length); - wqe->fastreg.va_base_hi = cpu_to_be32(wr->wr.fast_reg.iova_start >> 32); - wqe->fastreg.va_base_lo_fbo = - cpu_to_be32(wr->wr.fast_reg.iova_start & 0xffffffff); + wqe->fastreg.stag = cpu_to_be32(wr->rkey); + wqe->fastreg.len = cpu_to_be32(wr->length); + wqe->fastreg.va_base_hi = cpu_to_be32(wr->iova_start >> 32); + wqe->fastreg.va_base_lo_fbo = cpu_to_be32(wr->iova_start & 0xffffffff); wqe->fastreg.page_type_perms = cpu_to_be32( - V_FR_PAGE_COUNT(wr->wr.fast_reg.page_list_len) | - V_FR_PAGE_SIZE(wr->wr.fast_reg.page_shift-12) | + V_FR_PAGE_COUNT(wr->page_list_len) | + V_FR_PAGE_SIZE(wr->page_shift-12) | V_FR_TYPE(TPT_VATO) | - V_FR_PERMS(iwch_ib_to_tpt_access(wr->wr.fast_reg.access_flags))); + V_FR_PERMS(iwch_ib_to_tpt_access(wr->access_flags))); p = &wqe->fastreg.pbl_addrs[0]; - for (i = 0; i < wr->wr.fast_reg.page_list_len; i++, p++) { + for (i = 0; i < wr->page_list_len; i++, p++) { /* If we need a 2nd WR, then set it up */ if (i == T3_MAX_FASTREG_FRAG) { @@ -175,14 +175,14 @@ static int build_fastreg(union t3_wr *wqe, struct ib_send_wr *wr, Q_PTR2IDX((wq->wptr+1), wq->size_log2)); build_fw_riwrh((void *)wqe, T3_WR_FASTREG, 0, Q_GENBIT(wq->wptr + 1, wq->size_log2), - 0, 1 + wr->wr.fast_reg.page_list_len - T3_MAX_FASTREG_FRAG, + 0, 1 + wr->page_list_len - T3_MAX_FASTREG_FRAG, T3_EOP); p = &wqe->pbl_frag.pbl_addrs[0]; } - *p = cpu_to_be64((u64)wr->wr.fast_reg.page_list->page_list[i]); + *p = cpu_to_be64((u64)wr->page_list->page_list[i]); } - *flit_cnt = 5 + wr->wr.fast_reg.page_list_len; + *flit_cnt = 5 + wr->page_list_len; if (*flit_cnt > 15) *flit_cnt = 15; return 0; diff --git a/drivers/infiniband/hw/cxgb4/qp.c b/drivers/infiniband/hw/cxgb4/qp.c index 6517e1208ccb..b60498fff99a 100644 --- a/drivers/infiniband/hw/cxgb4/qp.c +++ b/drivers/infiniband/hw/cxgb4/qp.c @@ -528,8 +528,8 @@ static int build_rdma_write(struct t4_sq *sq, union t4_wr *wqe, if (wr->num_sge > T4_MAX_SEND_SGE) return -EINVAL; wqe->write.r2 = 0; - wqe->write.stag_sink = cpu_to_be32(wr->wr.rdma.rkey); - wqe->write.to_sink = cpu_to_be64(wr->wr.rdma.remote_addr); + wqe->write.stag_sink = cpu_to_be32(rdma_wr(wr)->rkey); + wqe->write.to_sink = cpu_to_be64(rdma_wr(wr)->remote_addr); if (wr->num_sge) { if (wr->send_flags & IB_SEND_INLINE) { ret = build_immd(sq, wqe->write.u.immd_src, wr, @@ -566,10 +566,10 @@ static int build_rdma_read(union t4_wr *wqe, struct ib_send_wr *wr, u8 *len16) if (wr->num_sge > 1) return -EINVAL; if (wr->num_sge) { - wqe->read.stag_src = cpu_to_be32(wr->wr.rdma.rkey); - wqe->read.to_src_hi = cpu_to_be32((u32)(wr->wr.rdma.remote_addr + wqe->read.stag_src = cpu_to_be32(rdma_wr(wr)->rkey); + wqe->read.to_src_hi = cpu_to_be32((u32)(rdma_wr(wr)->remote_addr >> 32)); - wqe->read.to_src_lo = cpu_to_be32((u32)wr->wr.rdma.remote_addr); + wqe->read.to_src_lo = cpu_to_be32((u32)rdma_wr(wr)->remote_addr); wqe->read.stag_sink = cpu_to_be32(wr->sg_list[0].lkey); wqe->read.plen = cpu_to_be32(wr->sg_list[0].length); wqe->read.to_sink_hi = cpu_to_be32((u32)(wr->sg_list[0].addr @@ -606,39 +606,36 @@ static int build_rdma_recv(struct c4iw_qp *qhp, union t4_recv_wr *wqe, } static int build_fastreg(struct t4_sq *sq, union t4_wr *wqe, - struct ib_send_wr *wr, u8 *len16, u8 t5dev) + struct ib_send_wr *send_wr, u8 *len16, u8 t5dev) { - + struct ib_fast_reg_wr *wr = fast_reg_wr(send_wr); struct fw_ri_immd *imdp; __be64 *p; int i; - int pbllen = roundup(wr->wr.fast_reg.page_list_len * sizeof(u64), 32); + int pbllen = roundup(wr->page_list_len * sizeof(u64), 32); int rem; - if (wr->wr.fast_reg.page_list_len > - t4_max_fr_depth(use_dsgl)) + if (wr->page_list_len > t4_max_fr_depth(use_dsgl)) return -EINVAL; wqe->fr.qpbinde_to_dcacpu = 0; - wqe->fr.pgsz_shift = wr->wr.fast_reg.page_shift - 12; + wqe->fr.pgsz_shift = wr->page_shift - 12; wqe->fr.addr_type = FW_RI_VA_BASED_TO; - wqe->fr.mem_perms = c4iw_ib_to_tpt_access(wr->wr.fast_reg.access_flags); + wqe->fr.mem_perms = c4iw_ib_to_tpt_access(wr->access_flags); wqe->fr.len_hi = 0; - wqe->fr.len_lo = cpu_to_be32(wr->wr.fast_reg.length); - wqe->fr.stag = cpu_to_be32(wr->wr.fast_reg.rkey); - wqe->fr.va_hi = cpu_to_be32(wr->wr.fast_reg.iova_start >> 32); - wqe->fr.va_lo_fbo = cpu_to_be32(wr->wr.fast_reg.iova_start & - 0xffffffff); + wqe->fr.len_lo = cpu_to_be32(wr->length); + wqe->fr.stag = cpu_to_be32(wr->rkey); + wqe->fr.va_hi = cpu_to_be32(wr->iova_start >> 32); + wqe->fr.va_lo_fbo = cpu_to_be32(wr->iova_start & 0xffffffff); if (t5dev && use_dsgl && (pbllen > max_fr_immd)) { struct c4iw_fr_page_list *c4pl = - to_c4iw_fr_page_list(wr->wr.fast_reg.page_list); + to_c4iw_fr_page_list(wr->page_list); struct fw_ri_dsgl *sglp; - for (i = 0; i < wr->wr.fast_reg.page_list_len; i++) { - wr->wr.fast_reg.page_list->page_list[i] = (__force u64) - cpu_to_be64((u64) - wr->wr.fast_reg.page_list->page_list[i]); + for (i = 0; i < wr->page_list_len; i++) { + wr->page_list->page_list[i] = (__force u64) + cpu_to_be64((u64)wr->page_list->page_list[i]); } sglp = (struct fw_ri_dsgl *)(&wqe->fr + 1); @@ -657,9 +654,8 @@ static int build_fastreg(struct t4_sq *sq, union t4_wr *wqe, imdp->immdlen = cpu_to_be32(pbllen); p = (__be64 *)(imdp + 1); rem = pbllen; - for (i = 0; i < wr->wr.fast_reg.page_list_len; i++) { - *p = cpu_to_be64( - (u64)wr->wr.fast_reg.page_list->page_list[i]); + for (i = 0; i < wr->page_list_len; i++) { + *p = cpu_to_be64((u64)wr->page_list->page_list[i]); rem -= sizeof(*p); if (++p == (__be64 *)&sq->queue[sq->size]) p = (__be64 *)sq->queue; diff --git a/drivers/infiniband/hw/mlx4/mad.c b/drivers/infiniband/hw/mlx4/mad.c index 1cd75ff02251..5f2de2ed5598 100644 --- a/drivers/infiniband/hw/mlx4/mad.c +++ b/drivers/infiniband/hw/mlx4/mad.c @@ -457,7 +457,8 @@ int mlx4_ib_send_to_slave(struct mlx4_ib_dev *dev, int slave, u8 port, struct ib_grh *grh, struct ib_mad *mad) { struct ib_sge list; - struct ib_send_wr wr, *bad_wr; + struct ib_ud_wr wr; + struct ib_send_wr *bad_wr; struct mlx4_ib_demux_pv_ctx *tun_ctx; struct mlx4_ib_demux_pv_qp *tun_qp; struct mlx4_rcv_tunnel_mad *tun_mad; @@ -582,18 +583,18 @@ int mlx4_ib_send_to_slave(struct mlx4_ib_dev *dev, int slave, u8 port, list.length = sizeof (struct mlx4_rcv_tunnel_mad); list.lkey = tun_ctx->pd->local_dma_lkey; - wr.wr.ud.ah = ah; - wr.wr.ud.port_num = port; - wr.wr.ud.remote_qkey = IB_QP_SET_QKEY; - wr.wr.ud.remote_qpn = dqpn; - wr.next = NULL; - wr.wr_id = ((u64) tun_tx_ix) | MLX4_TUN_SET_WRID_QPN(dest_qpt); - wr.sg_list = &list; - wr.num_sge = 1; - wr.opcode = IB_WR_SEND; - wr.send_flags = IB_SEND_SIGNALED; - - ret = ib_post_send(src_qp, &wr, &bad_wr); + wr.ah = ah; + wr.port_num = port; + wr.remote_qkey = IB_QP_SET_QKEY; + wr.remote_qpn = dqpn; + wr.wr.next = NULL; + wr.wr.wr_id = ((u64) tun_tx_ix) | MLX4_TUN_SET_WRID_QPN(dest_qpt); + wr.wr.sg_list = &list; + wr.wr.num_sge = 1; + wr.wr.opcode = IB_WR_SEND; + wr.wr.send_flags = IB_SEND_SIGNALED; + + ret = ib_post_send(src_qp, &wr.wr, &bad_wr); out: if (ret) ib_destroy_ah(ah); @@ -1175,7 +1176,8 @@ int mlx4_ib_send_to_wire(struct mlx4_ib_dev *dev, int slave, u8 port, u8 *s_mac, struct ib_mad *mad) { struct ib_sge list; - struct ib_send_wr wr, *bad_wr; + struct ib_ud_wr wr; + struct ib_send_wr *bad_wr; struct mlx4_ib_demux_pv_ctx *sqp_ctx; struct mlx4_ib_demux_pv_qp *sqp; struct mlx4_mad_snd_buf *sqp_mad; @@ -1246,22 +1248,22 @@ int mlx4_ib_send_to_wire(struct mlx4_ib_dev *dev, int slave, u8 port, list.length = sizeof (struct mlx4_mad_snd_buf); list.lkey = sqp_ctx->pd->local_dma_lkey; - wr.wr.ud.ah = ah; - wr.wr.ud.port_num = port; - wr.wr.ud.pkey_index = wire_pkey_ix; - wr.wr.ud.remote_qkey = qkey; - wr.wr.ud.remote_qpn = remote_qpn; - wr.next = NULL; - wr.wr_id = ((u64) wire_tx_ix) | MLX4_TUN_SET_WRID_QPN(src_qpnum); - wr.sg_list = &list; - wr.num_sge = 1; - wr.opcode = IB_WR_SEND; - wr.send_flags = IB_SEND_SIGNALED; + wr.ah = ah; + wr.port_num = port; + wr.pkey_index = wire_pkey_ix; + wr.remote_qkey = qkey; + wr.remote_qpn = remote_qpn; + wr.wr.next = NULL; + wr.wr.wr_id = ((u64) wire_tx_ix) | MLX4_TUN_SET_WRID_QPN(src_qpnum); + wr.wr.sg_list = &list; + wr.wr.num_sge = 1; + wr.wr.opcode = IB_WR_SEND; + wr.wr.send_flags = IB_SEND_SIGNALED; if (s_mac) memcpy(to_mah(ah)->av.eth.s_mac, s_mac, 6); - ret = ib_post_send(send_qp, &wr, &bad_wr); + ret = ib_post_send(send_qp, &wr.wr, &bad_wr); out: if (ret) ib_destroy_ah(ah); diff --git a/drivers/infiniband/hw/mlx4/mr.c b/drivers/infiniband/hw/mlx4/mr.c index 2542fd3c1a49..5bba176e9dfa 100644 --- a/drivers/infiniband/hw/mlx4/mr.c +++ b/drivers/infiniband/hw/mlx4/mr.c @@ -321,21 +321,21 @@ err_free: int mlx4_ib_bind_mw(struct ib_qp *qp, struct ib_mw *mw, struct ib_mw_bind *mw_bind) { - struct ib_send_wr wr; + struct ib_bind_mw_wr wr; struct ib_send_wr *bad_wr; int ret; memset(&wr, 0, sizeof(wr)); - wr.opcode = IB_WR_BIND_MW; - wr.wr_id = mw_bind->wr_id; - wr.send_flags = mw_bind->send_flags; - wr.wr.bind_mw.mw = mw; - wr.wr.bind_mw.bind_info = mw_bind->bind_info; - wr.wr.bind_mw.rkey = ib_inc_rkey(mw->rkey); - - ret = mlx4_ib_post_send(qp, &wr, &bad_wr); + wr.wr.opcode = IB_WR_BIND_MW; + wr.wr.wr_id = mw_bind->wr_id; + wr.wr.send_flags = mw_bind->send_flags; + wr.mw = mw; + wr.bind_info = mw_bind->bind_info; + wr.rkey = ib_inc_rkey(mw->rkey); + + ret = mlx4_ib_post_send(qp, &wr.wr, &bad_wr); if (!ret) - mw->rkey = wr.wr.bind_mw.rkey; + mw->rkey = wr.rkey; return ret; } diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c index 4ad9be3ad61c..3831cddb551f 100644 --- a/drivers/infiniband/hw/mlx4/qp.c +++ b/drivers/infiniband/hw/mlx4/qp.c @@ -2036,14 +2036,14 @@ static int vf_get_qp0_qkey(struct mlx4_dev *dev, int qpn, u32 *qkey) } static int build_sriov_qp0_header(struct mlx4_ib_sqp *sqp, - struct ib_send_wr *wr, + struct ib_ud_wr *wr, void *wqe, unsigned *mlx_seg_len) { struct mlx4_ib_dev *mdev = to_mdev(sqp->qp.ibqp.device); struct ib_device *ib_dev = &mdev->ib_dev; struct mlx4_wqe_mlx_seg *mlx = wqe; struct mlx4_wqe_inline_seg *inl = wqe + sizeof *mlx; - struct mlx4_ib_ah *ah = to_mah(wr->wr.ud.ah); + struct mlx4_ib_ah *ah = to_mah(wr->ah); u16 pkey; u32 qkey; int send_size; @@ -2051,13 +2051,13 @@ static int build_sriov_qp0_header(struct mlx4_ib_sqp *sqp, int spc; int i; - if (wr->opcode != IB_WR_SEND) + if (wr->wr.opcode != IB_WR_SEND) return -EINVAL; send_size = 0; - for (i = 0; i < wr->num_sge; ++i) - send_size += wr->sg_list[i].length; + for (i = 0; i < wr->wr.num_sge; ++i) + send_size += wr->wr.sg_list[i].length; /* for proxy-qp0 sends, need to add in size of tunnel header */ /* for tunnel-qp0 sends, tunnel header is already in s/g list */ @@ -2082,11 +2082,11 @@ static int build_sriov_qp0_header(struct mlx4_ib_sqp *sqp, mlx->rlid = sqp->ud_header.lrh.destination_lid; sqp->ud_header.lrh.virtual_lane = 0; - sqp->ud_header.bth.solicited_event = !!(wr->send_flags & IB_SEND_SOLICITED); + sqp->ud_header.bth.solicited_event = !!(wr->wr.send_flags & IB_SEND_SOLICITED); ib_get_cached_pkey(ib_dev, sqp->qp.port, 0, &pkey); sqp->ud_header.bth.pkey = cpu_to_be16(pkey); if (sqp->qp.mlx4_ib_qp_type == MLX4_IB_QPT_TUN_SMI_OWNER) - sqp->ud_header.bth.destination_qpn = cpu_to_be32(wr->wr.ud.remote_qpn); + sqp->ud_header.bth.destination_qpn = cpu_to_be32(wr->remote_qpn); else sqp->ud_header.bth.destination_qpn = cpu_to_be32(mdev->dev->caps.qp0_tunnel[sqp->qp.port - 1]); @@ -2158,14 +2158,14 @@ static void mlx4_u64_to_smac(u8 *dst_mac, u64 src_mac) } } -static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_send_wr *wr, +static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_ud_wr *wr, void *wqe, unsigned *mlx_seg_len) { struct ib_device *ib_dev = sqp->qp.ibqp.device; struct mlx4_wqe_mlx_seg *mlx = wqe; struct mlx4_wqe_ctrl_seg *ctrl = wqe; struct mlx4_wqe_inline_seg *inl = wqe + sizeof *mlx; - struct mlx4_ib_ah *ah = to_mah(wr->wr.ud.ah); + struct mlx4_ib_ah *ah = to_mah(wr->ah); union ib_gid sgid; u16 pkey; int send_size; @@ -2179,8 +2179,8 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_send_wr *wr, bool is_grh; send_size = 0; - for (i = 0; i < wr->num_sge; ++i) - send_size += wr->sg_list[i].length; + for (i = 0; i < wr->wr.num_sge; ++i) + send_size += wr->wr.sg_list[i].length; is_eth = rdma_port_get_link_layer(sqp->qp.ibqp.device, sqp->qp.port) == IB_LINK_LAYER_ETHERNET; is_grh = mlx4_ib_ah_grh_present(ah); @@ -2257,7 +2257,7 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_send_wr *wr, mlx->rlid = sqp->ud_header.lrh.destination_lid; } - switch (wr->opcode) { + switch (wr->wr.opcode) { case IB_WR_SEND: sqp->ud_header.bth.opcode = IB_OPCODE_UD_SEND_ONLY; sqp->ud_header.immediate_present = 0; @@ -2265,7 +2265,7 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_send_wr *wr, case IB_WR_SEND_WITH_IMM: sqp->ud_header.bth.opcode = IB_OPCODE_UD_SEND_ONLY_WITH_IMMEDIATE; sqp->ud_header.immediate_present = 1; - sqp->ud_header.immediate_data = wr->ex.imm_data; + sqp->ud_header.immediate_data = wr->wr.ex.imm_data; break; default: return -EINVAL; @@ -2308,16 +2308,16 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_send_wr *wr, if (sqp->ud_header.lrh.destination_lid == IB_LID_PERMISSIVE) sqp->ud_header.lrh.source_lid = IB_LID_PERMISSIVE; } - sqp->ud_header.bth.solicited_event = !!(wr->send_flags & IB_SEND_SOLICITED); + sqp->ud_header.bth.solicited_event = !!(wr->wr.send_flags & IB_SEND_SOLICITED); if (!sqp->qp.ibqp.qp_num) ib_get_cached_pkey(ib_dev, sqp->qp.port, sqp->pkey_index, &pkey); else - ib_get_cached_pkey(ib_dev, sqp->qp.port, wr->wr.ud.pkey_index, &pkey); + ib_get_cached_pkey(ib_dev, sqp->qp.port, wr->pkey_index, &pkey); sqp->ud_header.bth.pkey = cpu_to_be16(pkey); - sqp->ud_header.bth.destination_qpn = cpu_to_be32(wr->wr.ud.remote_qpn); + sqp->ud_header.bth.destination_qpn = cpu_to_be32(wr->remote_qpn); sqp->ud_header.bth.psn = cpu_to_be32((sqp->send_psn++) & ((1 << 24) - 1)); - sqp->ud_header.deth.qkey = cpu_to_be32(wr->wr.ud.remote_qkey & 0x80000000 ? - sqp->qkey : wr->wr.ud.remote_qkey); + sqp->ud_header.deth.qkey = cpu_to_be32(wr->remote_qkey & 0x80000000 ? + sqp->qkey : wr->remote_qkey); sqp->ud_header.deth.source_qpn = cpu_to_be32(sqp->qp.ibqp.qp_num); header_size = ib_ud_header_pack(&sqp->ud_header, sqp->header_buf); @@ -2405,43 +2405,45 @@ static __be32 convert_access(int acc) cpu_to_be32(MLX4_WQE_FMR_PERM_LOCAL_READ); } -static void set_fmr_seg(struct mlx4_wqe_fmr_seg *fseg, struct ib_send_wr *wr) +static void set_fmr_seg(struct mlx4_wqe_fmr_seg *fseg, + struct ib_fast_reg_wr *wr) { - struct mlx4_ib_fast_reg_page_list *mfrpl = to_mfrpl(wr->wr.fast_reg.page_list); + struct mlx4_ib_fast_reg_page_list *mfrpl = to_mfrpl(wr->page_list); int i; - for (i = 0; i < wr->wr.fast_reg.page_list_len; ++i) + for (i = 0; i < wr->page_list_len; ++i) mfrpl->mapped_page_list[i] = - cpu_to_be64(wr->wr.fast_reg.page_list->page_list[i] | + cpu_to_be64(wr->page_list->page_list[i] | MLX4_MTT_FLAG_PRESENT); - fseg->flags = convert_access(wr->wr.fast_reg.access_flags); - fseg->mem_key = cpu_to_be32(wr->wr.fast_reg.rkey); + fseg->flags = convert_access(wr->access_flags); + fseg->mem_key = cpu_to_be32(wr->rkey); fseg->buf_list = cpu_to_be64(mfrpl->map); - fseg->start_addr = cpu_to_be64(wr->wr.fast_reg.iova_start); - fseg->reg_len = cpu_to_be64(wr->wr.fast_reg.length); + fseg->start_addr = cpu_to_be64(wr->iova_start); + fseg->reg_len = cpu_to_be64(wr->length); fseg->offset = 0; /* XXX -- is this just for ZBVA? */ - fseg->page_size = cpu_to_be32(wr->wr.fast_reg.page_shift); + fseg->page_size = cpu_to_be32(wr->page_shift); fseg->reserved[0] = 0; fseg->reserved[1] = 0; } -static void set_bind_seg(struct mlx4_wqe_bind_seg *bseg, struct ib_send_wr *wr) +static void set_bind_seg(struct mlx4_wqe_bind_seg *bseg, + struct ib_bind_mw_wr *wr) { bseg->flags1 = - convert_access(wr->wr.bind_mw.bind_info.mw_access_flags) & + convert_access(wr->bind_info.mw_access_flags) & cpu_to_be32(MLX4_WQE_FMR_AND_BIND_PERM_REMOTE_READ | MLX4_WQE_FMR_AND_BIND_PERM_REMOTE_WRITE | MLX4_WQE_FMR_AND_BIND_PERM_ATOMIC); bseg->flags2 = 0; - if (wr->wr.bind_mw.mw->type == IB_MW_TYPE_2) + if (wr->mw->type == IB_MW_TYPE_2) bseg->flags2 |= cpu_to_be32(MLX4_WQE_BIND_TYPE_2); - if (wr->wr.bind_mw.bind_info.mw_access_flags & IB_ZERO_BASED) + if (wr->bind_info.mw_access_flags & IB_ZERO_BASED) bseg->flags2 |= cpu_to_be32(MLX4_WQE_BIND_ZERO_BASED); - bseg->new_rkey = cpu_to_be32(wr->wr.bind_mw.rkey); - bseg->lkey = cpu_to_be32(wr->wr.bind_mw.bind_info.mr->lkey); - bseg->addr = cpu_to_be64(wr->wr.bind_mw.bind_info.addr); - bseg->length = cpu_to_be64(wr->wr.bind_mw.bind_info.length); + bseg->new_rkey = cpu_to_be32(wr->rkey); + bseg->lkey = cpu_to_be32(wr->bind_info.mr->lkey); + bseg->addr = cpu_to_be64(wr->bind_info.addr); + bseg->length = cpu_to_be64(wr->bind_info.length); } static void set_local_inv_seg(struct mlx4_wqe_local_inval_seg *iseg, u32 rkey) @@ -2458,46 +2460,47 @@ static __always_inline void set_raddr_seg(struct mlx4_wqe_raddr_seg *rseg, rseg->reserved = 0; } -static void set_atomic_seg(struct mlx4_wqe_atomic_seg *aseg, struct ib_send_wr *wr) +static void set_atomic_seg(struct mlx4_wqe_atomic_seg *aseg, + struct ib_atomic_wr *wr) { - if (wr->opcode == IB_WR_ATOMIC_CMP_AND_SWP) { - aseg->swap_add = cpu_to_be64(wr->wr.atomic.swap); - aseg->compare = cpu_to_be64(wr->wr.atomic.compare_add); - } else if (wr->opcode == IB_WR_MASKED_ATOMIC_FETCH_AND_ADD) { - aseg->swap_add = cpu_to_be64(wr->wr.atomic.compare_add); - aseg->compare = cpu_to_be64(wr->wr.atomic.compare_add_mask); + if (wr->wr.opcode == IB_WR_ATOMIC_CMP_AND_SWP) { + aseg->swap_add = cpu_to_be64(wr->swap); + aseg->compare = cpu_to_be64(wr->compare_add); + } else if (wr->wr.opcode == IB_WR_MASKED_ATOMIC_FETCH_AND_ADD) { + aseg->swap_add = cpu_to_be64(wr->compare_add); + aseg->compare = cpu_to_be64(wr->compare_add_mask); } else { - aseg->swap_add = cpu_to_be64(wr->wr.atomic.compare_add); + aseg->swap_add = cpu_to_be64(wr->compare_add); aseg->compare = 0; } } static void set_masked_atomic_seg(struct mlx4_wqe_masked_atomic_seg *aseg, - struct ib_send_wr *wr) + struct ib_atomic_wr *wr) { - aseg->swap_add = cpu_to_be64(wr->wr.atomic.swap); - aseg->swap_add_mask = cpu_to_be64(wr->wr.atomic.swap_mask); - aseg->compare = cpu_to_be64(wr->wr.atomic.compare_add); - aseg->compare_mask = cpu_to_be64(wr->wr.atomic.compare_add_mask); + aseg->swap_add = cpu_to_be64(wr->swap); + aseg->swap_add_mask = cpu_to_be64(wr->swap_mask); + aseg->compare = cpu_to_be64(wr->compare_add); + aseg->compare_mask = cpu_to_be64(wr->compare_add_mask); } static void set_datagram_seg(struct mlx4_wqe_datagram_seg *dseg, - struct ib_send_wr *wr) + struct ib_ud_wr *wr) { - memcpy(dseg->av, &to_mah(wr->wr.ud.ah)->av, sizeof (struct mlx4_av)); - dseg->dqpn = cpu_to_be32(wr->wr.ud.remote_qpn); - dseg->qkey = cpu_to_be32(wr->wr.ud.remote_qkey); - dseg->vlan = to_mah(wr->wr.ud.ah)->av.eth.vlan; - memcpy(dseg->mac, to_mah(wr->wr.ud.ah)->av.eth.mac, 6); + memcpy(dseg->av, &to_mah(wr->ah)->av, sizeof (struct mlx4_av)); + dseg->dqpn = cpu_to_be32(wr->remote_qpn); + dseg->qkey = cpu_to_be32(wr->remote_qkey); + dseg->vlan = to_mah(wr->ah)->av.eth.vlan; + memcpy(dseg->mac, to_mah(wr->ah)->av.eth.mac, 6); } static void set_tunnel_datagram_seg(struct mlx4_ib_dev *dev, struct mlx4_wqe_datagram_seg *dseg, - struct ib_send_wr *wr, + struct ib_ud_wr *wr, enum mlx4_ib_qp_type qpt) { - union mlx4_ext_av *av = &to_mah(wr->wr.ud.ah)->av; + union mlx4_ext_av *av = &to_mah(wr->ah)->av; struct mlx4_av sqp_av = {0}; int port = *((u8 *) &av->ib.port_pd) & 0x3; @@ -2516,18 +2519,18 @@ static void set_tunnel_datagram_seg(struct mlx4_ib_dev *dev, dseg->qkey = cpu_to_be32(IB_QP_SET_QKEY); } -static void build_tunnel_header(struct ib_send_wr *wr, void *wqe, unsigned *mlx_seg_len) +static void build_tunnel_header(struct ib_ud_wr *wr, void *wqe, unsigned *mlx_seg_len) { struct mlx4_wqe_inline_seg *inl = wqe; struct mlx4_ib_tunnel_header hdr; - struct mlx4_ib_ah *ah = to_mah(wr->wr.ud.ah); + struct mlx4_ib_ah *ah = to_mah(wr->ah); int spc; int i; memcpy(&hdr.av, &ah->av, sizeof hdr.av); - hdr.remote_qpn = cpu_to_be32(wr->wr.ud.remote_qpn); - hdr.pkey_index = cpu_to_be16(wr->wr.ud.pkey_index); - hdr.qkey = cpu_to_be32(wr->wr.ud.remote_qkey); + hdr.remote_qpn = cpu_to_be32(wr->remote_qpn); + hdr.pkey_index = cpu_to_be16(wr->pkey_index); + hdr.qkey = cpu_to_be32(wr->remote_qkey); memcpy(hdr.mac, ah->av.eth.mac, 6); hdr.vlan = ah->av.eth.vlan; @@ -2599,22 +2602,22 @@ static void __set_data_seg(struct mlx4_wqe_data_seg *dseg, struct ib_sge *sg) dseg->addr = cpu_to_be64(sg->addr); } -static int build_lso_seg(struct mlx4_wqe_lso_seg *wqe, struct ib_send_wr *wr, +static int build_lso_seg(struct mlx4_wqe_lso_seg *wqe, struct ib_ud_wr *wr, struct mlx4_ib_qp *qp, unsigned *lso_seg_len, __be32 *lso_hdr_sz, __be32 *blh) { - unsigned halign = ALIGN(sizeof *wqe + wr->wr.ud.hlen, 16); + unsigned halign = ALIGN(sizeof *wqe + wr->hlen, 16); if (unlikely(halign > MLX4_IB_CACHE_LINE_SIZE)) *blh = cpu_to_be32(1 << 6); if (unlikely(!(qp->flags & MLX4_IB_QP_LSO) && - wr->num_sge > qp->sq.max_gs - (halign >> 4))) + wr->wr.num_sge > qp->sq.max_gs - (halign >> 4))) return -EINVAL; - memcpy(wqe->header, wr->wr.ud.header, wr->wr.ud.hlen); + memcpy(wqe->header, wr->header, wr->hlen); - *lso_hdr_sz = cpu_to_be32(wr->wr.ud.mss << 16 | wr->wr.ud.hlen); + *lso_hdr_sz = cpu_to_be32(wr->mss << 16 | wr->hlen); *lso_seg_len = halign; return 0; } @@ -2713,11 +2716,11 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, case IB_WR_ATOMIC_CMP_AND_SWP: case IB_WR_ATOMIC_FETCH_AND_ADD: case IB_WR_MASKED_ATOMIC_FETCH_AND_ADD: - set_raddr_seg(wqe, wr->wr.atomic.remote_addr, - wr->wr.atomic.rkey); + set_raddr_seg(wqe, atomic_wr(wr)->remote_addr, + atomic_wr(wr)->rkey); wqe += sizeof (struct mlx4_wqe_raddr_seg); - set_atomic_seg(wqe, wr); + set_atomic_seg(wqe, atomic_wr(wr)); wqe += sizeof (struct mlx4_wqe_atomic_seg); size += (sizeof (struct mlx4_wqe_raddr_seg) + @@ -2726,11 +2729,11 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, break; case IB_WR_MASKED_ATOMIC_CMP_AND_SWP: - set_raddr_seg(wqe, wr->wr.atomic.remote_addr, - wr->wr.atomic.rkey); + set_raddr_seg(wqe, atomic_wr(wr)->remote_addr, + atomic_wr(wr)->rkey); wqe += sizeof (struct mlx4_wqe_raddr_seg); - set_masked_atomic_seg(wqe, wr); + set_masked_atomic_seg(wqe, atomic_wr(wr)); wqe += sizeof (struct mlx4_wqe_masked_atomic_seg); size += (sizeof (struct mlx4_wqe_raddr_seg) + @@ -2741,8 +2744,8 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, case IB_WR_RDMA_READ: case IB_WR_RDMA_WRITE: case IB_WR_RDMA_WRITE_WITH_IMM: - set_raddr_seg(wqe, wr->wr.rdma.remote_addr, - wr->wr.rdma.rkey); + set_raddr_seg(wqe, rdma_wr(wr)->remote_addr, + rdma_wr(wr)->rkey); wqe += sizeof (struct mlx4_wqe_raddr_seg); size += sizeof (struct mlx4_wqe_raddr_seg) / 16; break; @@ -2758,7 +2761,7 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, case IB_WR_FAST_REG_MR: ctrl->srcrb_flags |= cpu_to_be32(MLX4_WQE_CTRL_STRONG_ORDER); - set_fmr_seg(wqe, wr); + set_fmr_seg(wqe, fast_reg_wr(wr)); wqe += sizeof (struct mlx4_wqe_fmr_seg); size += sizeof (struct mlx4_wqe_fmr_seg) / 16; break; @@ -2766,7 +2769,7 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, case IB_WR_BIND_MW: ctrl->srcrb_flags |= cpu_to_be32(MLX4_WQE_CTRL_STRONG_ORDER); - set_bind_seg(wqe, wr); + set_bind_seg(wqe, bind_mw_wr(wr)); wqe += sizeof(struct mlx4_wqe_bind_seg); size += sizeof(struct mlx4_wqe_bind_seg) / 16; break; @@ -2777,7 +2780,8 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, break; case MLX4_IB_QPT_TUN_SMI_OWNER: - err = build_sriov_qp0_header(to_msqp(qp), wr, ctrl, &seglen); + err = build_sriov_qp0_header(to_msqp(qp), ud_wr(wr), + ctrl, &seglen); if (unlikely(err)) { *bad_wr = wr; goto out; @@ -2788,19 +2792,20 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, case MLX4_IB_QPT_TUN_SMI: case MLX4_IB_QPT_TUN_GSI: /* this is a UD qp used in MAD responses to slaves. */ - set_datagram_seg(wqe, wr); + set_datagram_seg(wqe, ud_wr(wr)); /* set the forced-loopback bit in the data seg av */ *(__be32 *) wqe |= cpu_to_be32(0x80000000); wqe += sizeof (struct mlx4_wqe_datagram_seg); size += sizeof (struct mlx4_wqe_datagram_seg) / 16; break; case MLX4_IB_QPT_UD: - set_datagram_seg(wqe, wr); + set_datagram_seg(wqe, ud_wr(wr)); wqe += sizeof (struct mlx4_wqe_datagram_seg); size += sizeof (struct mlx4_wqe_datagram_seg) / 16; if (wr->opcode == IB_WR_LSO) { - err = build_lso_seg(wqe, wr, qp, &seglen, &lso_hdr_sz, &blh); + err = build_lso_seg(wqe, ud_wr(wr), qp, &seglen, + &lso_hdr_sz, &blh); if (unlikely(err)) { *bad_wr = wr; goto out; @@ -2812,7 +2817,8 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, break; case MLX4_IB_QPT_PROXY_SMI_OWNER: - err = build_sriov_qp0_header(to_msqp(qp), wr, ctrl, &seglen); + err = build_sriov_qp0_header(to_msqp(qp), ud_wr(wr), + ctrl, &seglen); if (unlikely(err)) { *bad_wr = wr; goto out; @@ -2823,7 +2829,7 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, add_zero_len_inline(wqe); wqe += 16; size++; - build_tunnel_header(wr, wqe, &seglen); + build_tunnel_header(ud_wr(wr), wqe, &seglen); wqe += seglen; size += seglen / 16; break; @@ -2833,18 +2839,20 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, * In this case we first add a UD segment targeting * the tunnel qp, and then add a header with address * information */ - set_tunnel_datagram_seg(to_mdev(ibqp->device), wqe, wr, + set_tunnel_datagram_seg(to_mdev(ibqp->device), wqe, + ud_wr(wr), qp->mlx4_ib_qp_type); wqe += sizeof (struct mlx4_wqe_datagram_seg); size += sizeof (struct mlx4_wqe_datagram_seg) / 16; - build_tunnel_header(wr, wqe, &seglen); + build_tunnel_header(ud_wr(wr), wqe, &seglen); wqe += seglen; size += seglen / 16; break; case MLX4_IB_QPT_SMI: case MLX4_IB_QPT_GSI: - err = build_mlx_header(to_msqp(qp), wr, ctrl, &seglen); + err = build_mlx_header(to_msqp(qp), ud_wr(wr), ctrl, + &seglen); if (unlikely(err)) { *bad_wr = wr; goto out; diff --git a/drivers/infiniband/hw/mlx5/mlx5_ib.h b/drivers/infiniband/hw/mlx5/mlx5_ib.h index 22123b79d550..29f3ecdbe790 100644 --- a/drivers/infiniband/hw/mlx5/mlx5_ib.h +++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h @@ -245,6 +245,7 @@ enum mlx5_ib_qp_flags { }; struct mlx5_umr_wr { + struct ib_send_wr wr; union { u64 virt_addr; u64 offset; @@ -257,6 +258,11 @@ struct mlx5_umr_wr { u32 mkey; }; +static inline struct mlx5_umr_wr *umr_wr(struct ib_send_wr *wr) +{ + return container_of(wr, struct mlx5_umr_wr, wr); +} + struct mlx5_shared_mr_info { int mr_id; struct ib_umem *umem; diff --git a/drivers/infiniband/hw/mlx5/mr.c b/drivers/infiniband/hw/mlx5/mr.c index 54a15b5d336d..b30d4ae0fb61 100644 --- a/drivers/infiniband/hw/mlx5/mr.c +++ b/drivers/infiniband/hw/mlx5/mr.c @@ -687,7 +687,7 @@ static void prep_umr_reg_wqe(struct ib_pd *pd, struct ib_send_wr *wr, int access_flags) { struct mlx5_ib_dev *dev = to_mdev(pd->device); - struct mlx5_umr_wr *umrwr = (struct mlx5_umr_wr *)&wr->wr.fast_reg; + struct mlx5_umr_wr *umrwr = umr_wr(wr); sg->addr = dma; sg->length = ALIGN(sizeof(u64) * n, 64); @@ -715,7 +715,7 @@ static void prep_umr_reg_wqe(struct ib_pd *pd, struct ib_send_wr *wr, static void prep_umr_unreg_wqe(struct mlx5_ib_dev *dev, struct ib_send_wr *wr, u32 key) { - struct mlx5_umr_wr *umrwr = (struct mlx5_umr_wr *)&wr->wr.fast_reg; + struct mlx5_umr_wr *umrwr = umr_wr(wr); wr->send_flags = MLX5_IB_SEND_UMR_UNREG | MLX5_IB_SEND_UMR_FAIL_IF_FREE; wr->opcode = MLX5_IB_WR_UMR; @@ -752,7 +752,8 @@ static struct mlx5_ib_mr *reg_umr(struct ib_pd *pd, struct ib_umem *umem, struct device *ddev = dev->ib_dev.dma_device; struct umr_common *umrc = &dev->umrc; struct mlx5_ib_umr_context umr_context; - struct ib_send_wr wr, *bad; + struct mlx5_umr_wr umrwr; + struct ib_send_wr *bad; struct mlx5_ib_mr *mr; struct ib_sge sg; int size; @@ -798,14 +799,14 @@ static struct mlx5_ib_mr *reg_umr(struct ib_pd *pd, struct ib_umem *umem, goto free_pas; } - memset(&wr, 0, sizeof(wr)); - wr.wr_id = (u64)(unsigned long)&umr_context; - prep_umr_reg_wqe(pd, &wr, &sg, dma, npages, mr->mmr.key, page_shift, - virt_addr, len, access_flags); + memset(&umrwr, 0, sizeof(umrwr)); + umrwr.wr.wr_id = (u64)(unsigned long)&umr_context; + prep_umr_reg_wqe(pd, &umrwr.wr, &sg, dma, npages, mr->mmr.key, + page_shift, virt_addr, len, access_flags); mlx5_ib_init_umr_context(&umr_context); down(&umrc->sem); - err = ib_post_send(umrc->qp, &wr, &bad); + err = ib_post_send(umrc->qp, &umrwr.wr, &bad); if (err) { mlx5_ib_warn(dev, "post send failed, err %d\n", err); goto unmap_dma; @@ -851,8 +852,8 @@ int mlx5_ib_update_mtt(struct mlx5_ib_mr *mr, u64 start_page_index, int npages, int size; __be64 *pas; dma_addr_t dma; - struct ib_send_wr wr, *bad; - struct mlx5_umr_wr *umrwr = (struct mlx5_umr_wr *)&wr.wr.fast_reg; + struct ib_send_wr *bad; + struct mlx5_umr_wr wr; struct ib_sge sg; int err = 0; const int page_index_alignment = MLX5_UMR_MTT_ALIGNMENT / sizeof(u64); @@ -917,26 +918,26 @@ int mlx5_ib_update_mtt(struct mlx5_ib_mr *mr, u64 start_page_index, int npages, dma_sync_single_for_device(ddev, dma, size, DMA_TO_DEVICE); memset(&wr, 0, sizeof(wr)); - wr.wr_id = (u64)(unsigned long)&umr_context; + wr.wr.wr_id = (u64)(unsigned long)&umr_context; sg.addr = dma; sg.length = ALIGN(npages * sizeof(u64), MLX5_UMR_MTT_ALIGNMENT); sg.lkey = dev->umrc.pd->local_dma_lkey; - wr.send_flags = MLX5_IB_SEND_UMR_FAIL_IF_FREE | + wr.wr.send_flags = MLX5_IB_SEND_UMR_FAIL_IF_FREE | MLX5_IB_SEND_UMR_UPDATE_MTT; - wr.sg_list = &sg; - wr.num_sge = 1; - wr.opcode = MLX5_IB_WR_UMR; - umrwr->npages = sg.length / sizeof(u64); - umrwr->page_shift = PAGE_SHIFT; - umrwr->mkey = mr->mmr.key; - umrwr->target.offset = start_page_index; + wr.wr.sg_list = &sg; + wr.wr.num_sge = 1; + wr.wr.opcode = MLX5_IB_WR_UMR; + wr.npages = sg.length / sizeof(u64); + wr.page_shift = PAGE_SHIFT; + wr.mkey = mr->mmr.key; + wr.target.offset = start_page_index; mlx5_ib_init_umr_context(&umr_context); down(&umrc->sem); - err = ib_post_send(umrc->qp, &wr, &bad); + err = ib_post_send(umrc->qp, &wr.wr, &bad); if (err) { mlx5_ib_err(dev, "UMR post send failed, err %d\n", err); } else { @@ -1122,16 +1123,17 @@ static int unreg_umr(struct mlx5_ib_dev *dev, struct mlx5_ib_mr *mr) { struct umr_common *umrc = &dev->umrc; struct mlx5_ib_umr_context umr_context; - struct ib_send_wr wr, *bad; + struct mlx5_umr_wr umrwr; + struct ib_send_wr *bad; int err; - memset(&wr, 0, sizeof(wr)); - wr.wr_id = (u64)(unsigned long)&umr_context; - prep_umr_unreg_wqe(dev, &wr, mr->mmr.key); + memset(&umrwr.wr, 0, sizeof(umrwr)); + umrwr.wr.wr_id = (u64)(unsigned long)&umr_context; + prep_umr_unreg_wqe(dev, &umrwr.wr, mr->mmr.key); mlx5_ib_init_umr_context(&umr_context); down(&umrc->sem); - err = ib_post_send(umrc->qp, &wr, &bad); + err = ib_post_send(umrc->qp, &umrwr.wr, &bad); if (err) { up(&umrc->sem); mlx5_ib_dbg(dev, "err %d\n", err); diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c index 6f521a3418e8..d4c36af4270f 100644 --- a/drivers/infiniband/hw/mlx5/qp.c +++ b/drivers/infiniband/hw/mlx5/qp.c @@ -1838,9 +1838,9 @@ static __always_inline void set_raddr_seg(struct mlx5_wqe_raddr_seg *rseg, static void set_datagram_seg(struct mlx5_wqe_datagram_seg *dseg, struct ib_send_wr *wr) { - memcpy(&dseg->av, &to_mah(wr->wr.ud.ah)->av, sizeof(struct mlx5_av)); - dseg->av.dqp_dct = cpu_to_be32(wr->wr.ud.remote_qpn | MLX5_EXTENDED_UD_AV); - dseg->av.key.qkey.qkey = cpu_to_be32(wr->wr.ud.remote_qkey); + memcpy(&dseg->av, &to_mah(ud_wr(wr)->ah)->av, sizeof(struct mlx5_av)); + dseg->av.dqp_dct = cpu_to_be32(ud_wr(wr)->remote_qpn | MLX5_EXTENDED_UD_AV); + dseg->av.key.qkey.qkey = cpu_to_be32(ud_wr(wr)->remote_qkey); } static void set_data_ptr_seg(struct mlx5_wqe_data_seg *dseg, struct ib_sge *sg) @@ -1908,7 +1908,7 @@ static void set_frwr_umr_segment(struct mlx5_wqe_umr_ctrl_seg *umr, } umr->flags = (1 << 5); /* fail if not free */ - umr->klm_octowords = get_klm_octo(wr->wr.fast_reg.page_list_len); + umr->klm_octowords = get_klm_octo(fast_reg_wr(wr)->page_list_len); umr->mkey_mask = frwr_mkey_mask(); } @@ -1952,7 +1952,7 @@ static __be64 get_umr_update_mtt_mask(void) static void set_reg_umr_segment(struct mlx5_wqe_umr_ctrl_seg *umr, struct ib_send_wr *wr) { - struct mlx5_umr_wr *umrwr = (struct mlx5_umr_wr *)&wr->wr.fast_reg; + struct mlx5_umr_wr *umrwr = umr_wr(wr); memset(umr, 0, sizeof(*umr)); @@ -1996,20 +1996,20 @@ static void set_mkey_segment(struct mlx5_mkey_seg *seg, struct ib_send_wr *wr, return; } - seg->flags = get_umr_flags(wr->wr.fast_reg.access_flags) | + seg->flags = get_umr_flags(fast_reg_wr(wr)->access_flags) | MLX5_ACCESS_MODE_MTT; *writ = seg->flags & (MLX5_PERM_LOCAL_WRITE | IB_ACCESS_REMOTE_WRITE); - seg->qpn_mkey7_0 = cpu_to_be32((wr->wr.fast_reg.rkey & 0xff) | 0xffffff00); + seg->qpn_mkey7_0 = cpu_to_be32((fast_reg_wr(wr)->rkey & 0xff) | 0xffffff00); seg->flags_pd = cpu_to_be32(MLX5_MKEY_REMOTE_INVAL); - seg->start_addr = cpu_to_be64(wr->wr.fast_reg.iova_start); - seg->len = cpu_to_be64(wr->wr.fast_reg.length); - seg->xlt_oct_size = cpu_to_be32((wr->wr.fast_reg.page_list_len + 1) / 2); - seg->log2_page_size = wr->wr.fast_reg.page_shift; + seg->start_addr = cpu_to_be64(fast_reg_wr(wr)->iova_start); + seg->len = cpu_to_be64(fast_reg_wr(wr)->length); + seg->xlt_oct_size = cpu_to_be32((fast_reg_wr(wr)->page_list_len + 1) / 2); + seg->log2_page_size = fast_reg_wr(wr)->page_shift; } static void set_reg_mkey_segment(struct mlx5_mkey_seg *seg, struct ib_send_wr *wr) { - struct mlx5_umr_wr *umrwr = (struct mlx5_umr_wr *)&wr->wr.fast_reg; + struct mlx5_umr_wr *umrwr = umr_wr(wr); memset(seg, 0, sizeof(*seg)); if (wr->send_flags & MLX5_IB_SEND_UMR_UNREG) { @@ -2034,15 +2034,15 @@ static void set_frwr_pages(struct mlx5_wqe_data_seg *dseg, struct mlx5_ib_pd *pd, int writ) { - struct mlx5_ib_fast_reg_page_list *mfrpl = to_mfrpl(wr->wr.fast_reg.page_list); - u64 *page_list = wr->wr.fast_reg.page_list->page_list; + struct mlx5_ib_fast_reg_page_list *mfrpl = to_mfrpl(fast_reg_wr(wr)->page_list); + u64 *page_list = fast_reg_wr(wr)->page_list->page_list; u64 perm = MLX5_EN_RD | (writ ? MLX5_EN_WR : 0); int i; - for (i = 0; i < wr->wr.fast_reg.page_list_len; i++) + for (i = 0; i < fast_reg_wr(wr)->page_list_len; i++) mfrpl->mapped_page_list[i] = cpu_to_be64(page_list[i] | perm); dseg->addr = cpu_to_be64(mfrpl->map); - dseg->byte_count = cpu_to_be32(ALIGN(sizeof(u64) * wr->wr.fast_reg.page_list_len, 64)); + dseg->byte_count = cpu_to_be32(ALIGN(sizeof(u64) * fast_reg_wr(wr)->page_list_len, 64)); dseg->lkey = cpu_to_be32(pd->ibpd.local_dma_lkey); } @@ -2224,22 +2224,22 @@ static int mlx5_set_bsf(struct ib_mr *sig_mr, return 0; } -static int set_sig_data_segment(struct ib_send_wr *wr, struct mlx5_ib_qp *qp, - void **seg, int *size) +static int set_sig_data_segment(struct ib_sig_handover_wr *wr, + struct mlx5_ib_qp *qp, void **seg, int *size) { - struct ib_sig_attrs *sig_attrs = wr->wr.sig_handover.sig_attrs; - struct ib_mr *sig_mr = wr->wr.sig_handover.sig_mr; + struct ib_sig_attrs *sig_attrs = wr->sig_attrs; + struct ib_mr *sig_mr = wr->sig_mr; struct mlx5_bsf *bsf; - u32 data_len = wr->sg_list->length; - u32 data_key = wr->sg_list->lkey; - u64 data_va = wr->sg_list->addr; + u32 data_len = wr->wr.sg_list->length; + u32 data_key = wr->wr.sg_list->lkey; + u64 data_va = wr->wr.sg_list->addr; int ret; int wqe_size; - if (!wr->wr.sig_handover.prot || - (data_key == wr->wr.sig_handover.prot->lkey && - data_va == wr->wr.sig_handover.prot->addr && - data_len == wr->wr.sig_handover.prot->length)) { + if (!wr->prot || + (data_key == wr->prot->lkey && + data_va == wr->prot->addr && + data_len == wr->prot->length)) { /** * Source domain doesn't contain signature information * or data and protection are interleaved in memory. @@ -2273,8 +2273,8 @@ static int set_sig_data_segment(struct ib_send_wr *wr, struct mlx5_ib_qp *qp, struct mlx5_stride_block_ctrl_seg *sblock_ctrl; struct mlx5_stride_block_entry *data_sentry; struct mlx5_stride_block_entry *prot_sentry; - u32 prot_key = wr->wr.sig_handover.prot->lkey; - u64 prot_va = wr->wr.sig_handover.prot->addr; + u32 prot_key = wr->prot->lkey; + u64 prot_va = wr->prot->addr; u16 block_size = sig_attrs->mem.sig.dif.pi_interval; int prot_size; @@ -2326,16 +2326,16 @@ static int set_sig_data_segment(struct ib_send_wr *wr, struct mlx5_ib_qp *qp, } static void set_sig_mkey_segment(struct mlx5_mkey_seg *seg, - struct ib_send_wr *wr, u32 nelements, + struct ib_sig_handover_wr *wr, u32 nelements, u32 length, u32 pdn) { - struct ib_mr *sig_mr = wr->wr.sig_handover.sig_mr; + struct ib_mr *sig_mr = wr->sig_mr; u32 sig_key = sig_mr->rkey; u8 sigerr = to_mmr(sig_mr)->sig->sigerr_count & 1; memset(seg, 0, sizeof(*seg)); - seg->flags = get_umr_flags(wr->wr.sig_handover.access_flags) | + seg->flags = get_umr_flags(wr->access_flags) | MLX5_ACCESS_MODE_KLM; seg->qpn_mkey7_0 = cpu_to_be32((sig_key & 0xff) | 0xffffff00); seg->flags_pd = cpu_to_be32(MLX5_MKEY_REMOTE_INVAL | sigerr << 26 | @@ -2346,7 +2346,7 @@ static void set_sig_mkey_segment(struct mlx5_mkey_seg *seg, } static void set_sig_umr_segment(struct mlx5_wqe_umr_ctrl_seg *umr, - struct ib_send_wr *wr, u32 nelements) + u32 nelements) { memset(umr, 0, sizeof(*umr)); @@ -2357,37 +2357,37 @@ static void set_sig_umr_segment(struct mlx5_wqe_umr_ctrl_seg *umr, } -static int set_sig_umr_wr(struct ib_send_wr *wr, struct mlx5_ib_qp *qp, +static int set_sig_umr_wr(struct ib_send_wr *send_wr, struct mlx5_ib_qp *qp, void **seg, int *size) { - struct mlx5_ib_mr *sig_mr = to_mmr(wr->wr.sig_handover.sig_mr); + struct ib_sig_handover_wr *wr = sig_handover_wr(send_wr); + struct mlx5_ib_mr *sig_mr = to_mmr(wr->sig_mr); u32 pdn = get_pd(qp)->pdn; u32 klm_oct_size; int region_len, ret; - if (unlikely(wr->num_sge != 1) || - unlikely(wr->wr.sig_handover.access_flags & - IB_ACCESS_REMOTE_ATOMIC) || + if (unlikely(wr->wr.num_sge != 1) || + unlikely(wr->access_flags & IB_ACCESS_REMOTE_ATOMIC) || unlikely(!sig_mr->sig) || unlikely(!qp->signature_en) || unlikely(!sig_mr->sig->sig_status_checked)) return -EINVAL; /* length of the protected region, data + protection */ - region_len = wr->sg_list->length; - if (wr->wr.sig_handover.prot && - (wr->wr.sig_handover.prot->lkey != wr->sg_list->lkey || - wr->wr.sig_handover.prot->addr != wr->sg_list->addr || - wr->wr.sig_handover.prot->length != wr->sg_list->length)) - region_len += wr->wr.sig_handover.prot->length; + region_len = wr->wr.sg_list->length; + if (wr->prot && + (wr->prot->lkey != wr->wr.sg_list->lkey || + wr->prot->addr != wr->wr.sg_list->addr || + wr->prot->length != wr->wr.sg_list->length)) + region_len += wr->prot->length; /** * KLM octoword size - if protection was provided * then we use strided block format (3 octowords), * else we use single KLM (1 octoword) **/ - klm_oct_size = wr->wr.sig_handover.prot ? 3 : 1; + klm_oct_size = wr->prot ? 3 : 1; - set_sig_umr_segment(*seg, wr, klm_oct_size); + set_sig_umr_segment(*seg, klm_oct_size); *seg += sizeof(struct mlx5_wqe_umr_ctrl_seg); *size += sizeof(struct mlx5_wqe_umr_ctrl_seg) / 16; if (unlikely((*seg == qp->sq.qend))) @@ -2454,8 +2454,8 @@ static int set_frwr_li_wr(void **seg, struct ib_send_wr *wr, int *size, if (unlikely((*seg == qp->sq.qend))) *seg = mlx5_get_send_wqe(qp, 0); if (!li) { - if (unlikely(wr->wr.fast_reg.page_list_len > - wr->wr.fast_reg.page_list->max_page_list_len)) + if (unlikely(fast_reg_wr(wr)->page_list_len > + fast_reg_wr(wr)->page_list->max_page_list_len)) return -ENOMEM; set_frwr_pages(*seg, wr, mdev, pd, writ); @@ -2636,8 +2636,8 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, case IB_WR_RDMA_READ: case IB_WR_RDMA_WRITE: case IB_WR_RDMA_WRITE_WITH_IMM: - set_raddr_seg(seg, wr->wr.rdma.remote_addr, - wr->wr.rdma.rkey); + set_raddr_seg(seg, rdma_wr(wr)->remote_addr, + rdma_wr(wr)->rkey); seg += sizeof(struct mlx5_wqe_raddr_seg); size += sizeof(struct mlx5_wqe_raddr_seg) / 16; break; @@ -2666,7 +2666,7 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, case IB_WR_FAST_REG_MR: next_fence = MLX5_FENCE_MODE_INITIATOR_SMALL; qp->sq.wr_data[idx] = IB_WR_FAST_REG_MR; - ctrl->imm = cpu_to_be32(wr->wr.fast_reg.rkey); + ctrl->imm = cpu_to_be32(fast_reg_wr(wr)->rkey); err = set_frwr_li_wr(&seg, wr, &size, mdev, to_mpd(ibqp->pd), qp); if (err) { mlx5_ib_warn(dev, "\n"); @@ -2678,7 +2678,7 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, case IB_WR_REG_SIG_MR: qp->sq.wr_data[idx] = IB_WR_REG_SIG_MR; - mr = to_mmr(wr->wr.sig_handover.sig_mr); + mr = to_mmr(sig_handover_wr(wr)->sig_mr); ctrl->imm = cpu_to_be32(mr->ibmr.rkey); err = set_sig_umr_wr(wr, qp, &seg, &size); @@ -2706,7 +2706,7 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, goto out; } - err = set_psv_wr(&wr->wr.sig_handover.sig_attrs->mem, + err = set_psv_wr(&sig_handover_wr(wr)->sig_attrs->mem, mr->sig->psv_memory.psv_idx, &seg, &size); if (err) { @@ -2728,7 +2728,7 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, } next_fence = MLX5_FENCE_MODE_INITIATOR_SMALL; - err = set_psv_wr(&wr->wr.sig_handover.sig_attrs->wire, + err = set_psv_wr(&sig_handover_wr(wr)->sig_attrs->wire, mr->sig->psv_wire.psv_idx, &seg, &size); if (err) { @@ -2752,8 +2752,8 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, switch (wr->opcode) { case IB_WR_RDMA_WRITE: case IB_WR_RDMA_WRITE_WITH_IMM: - set_raddr_seg(seg, wr->wr.rdma.remote_addr, - wr->wr.rdma.rkey); + set_raddr_seg(seg, rdma_wr(wr)->remote_addr, + rdma_wr(wr)->rkey); seg += sizeof(struct mlx5_wqe_raddr_seg); size += sizeof(struct mlx5_wqe_raddr_seg) / 16; break; @@ -2780,7 +2780,7 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, goto out; } qp->sq.wr_data[idx] = MLX5_IB_WR_UMR; - ctrl->imm = cpu_to_be32(wr->wr.fast_reg.rkey); + ctrl->imm = cpu_to_be32(umr_wr(wr)->mkey); set_reg_umr_segment(seg, wr); seg += sizeof(struct mlx5_wqe_umr_ctrl_seg); size += sizeof(struct mlx5_wqe_umr_ctrl_seg) / 16; diff --git a/drivers/infiniband/hw/mthca/mthca_qp.c b/drivers/infiniband/hw/mthca/mthca_qp.c index e354b2f04ad9..35fe506e2cfa 100644 --- a/drivers/infiniband/hw/mthca/mthca_qp.c +++ b/drivers/infiniband/hw/mthca/mthca_qp.c @@ -1476,7 +1476,7 @@ void mthca_free_qp(struct mthca_dev *dev, /* Create UD header for an MLX send and build a data segment for it */ static int build_mlx_header(struct mthca_dev *dev, struct mthca_sqp *sqp, - int ind, struct ib_send_wr *wr, + int ind, struct ib_ud_wr *wr, struct mthca_mlx_seg *mlx, struct mthca_data_seg *data) { @@ -1485,10 +1485,10 @@ static int build_mlx_header(struct mthca_dev *dev, struct mthca_sqp *sqp, u16 pkey; ib_ud_header_init(256, /* assume a MAD */ 1, 0, 0, - mthca_ah_grh_present(to_mah(wr->wr.ud.ah)), 0, + mthca_ah_grh_present(to_mah(wr->ah)), 0, &sqp->ud_header); - err = mthca_read_ah(dev, to_mah(wr->wr.ud.ah), &sqp->ud_header); + err = mthca_read_ah(dev, to_mah(wr->ah), &sqp->ud_header); if (err) return err; mlx->flags &= ~cpu_to_be32(MTHCA_NEXT_SOLICIT | 1); @@ -1499,7 +1499,7 @@ static int build_mlx_header(struct mthca_dev *dev, struct mthca_sqp *sqp, mlx->rlid = sqp->ud_header.lrh.destination_lid; mlx->vcrc = 0; - switch (wr->opcode) { + switch (wr->wr.opcode) { case IB_WR_SEND: sqp->ud_header.bth.opcode = IB_OPCODE_UD_SEND_ONLY; sqp->ud_header.immediate_present = 0; @@ -1507,7 +1507,7 @@ static int build_mlx_header(struct mthca_dev *dev, struct mthca_sqp *sqp, case IB_WR_SEND_WITH_IMM: sqp->ud_header.bth.opcode = IB_OPCODE_UD_SEND_ONLY_WITH_IMMEDIATE; sqp->ud_header.immediate_present = 1; - sqp->ud_header.immediate_data = wr->ex.imm_data; + sqp->ud_header.immediate_data = wr->wr.ex.imm_data; break; default: return -EINVAL; @@ -1516,18 +1516,18 @@ static int build_mlx_header(struct mthca_dev *dev, struct mthca_sqp *sqp, sqp->ud_header.lrh.virtual_lane = !sqp->qp.ibqp.qp_num ? 15 : 0; if (sqp->ud_header.lrh.destination_lid == IB_LID_PERMISSIVE) sqp->ud_header.lrh.source_lid = IB_LID_PERMISSIVE; - sqp->ud_header.bth.solicited_event = !!(wr->send_flags & IB_SEND_SOLICITED); + sqp->ud_header.bth.solicited_event = !!(wr->wr.send_flags & IB_SEND_SOLICITED); if (!sqp->qp.ibqp.qp_num) ib_get_cached_pkey(&dev->ib_dev, sqp->qp.port, sqp->pkey_index, &pkey); else ib_get_cached_pkey(&dev->ib_dev, sqp->qp.port, - wr->wr.ud.pkey_index, &pkey); + wr->pkey_index, &pkey); sqp->ud_header.bth.pkey = cpu_to_be16(pkey); - sqp->ud_header.bth.destination_qpn = cpu_to_be32(wr->wr.ud.remote_qpn); + sqp->ud_header.bth.destination_qpn = cpu_to_be32(wr->remote_qpn); sqp->ud_header.bth.psn = cpu_to_be32((sqp->send_psn++) & ((1 << 24) - 1)); - sqp->ud_header.deth.qkey = cpu_to_be32(wr->wr.ud.remote_qkey & 0x80000000 ? - sqp->qkey : wr->wr.ud.remote_qkey); + sqp->ud_header.deth.qkey = cpu_to_be32(wr->remote_qkey & 0x80000000 ? + sqp->qkey : wr->remote_qkey); sqp->ud_header.deth.source_qpn = cpu_to_be32(sqp->qp.ibqp.qp_num); header_size = ib_ud_header_pack(&sqp->ud_header, @@ -1569,34 +1569,34 @@ static __always_inline void set_raddr_seg(struct mthca_raddr_seg *rseg, } static __always_inline void set_atomic_seg(struct mthca_atomic_seg *aseg, - struct ib_send_wr *wr) + struct ib_atomic_wr *wr) { - if (wr->opcode == IB_WR_ATOMIC_CMP_AND_SWP) { - aseg->swap_add = cpu_to_be64(wr->wr.atomic.swap); - aseg->compare = cpu_to_be64(wr->wr.atomic.compare_add); + if (wr->wr.opcode == IB_WR_ATOMIC_CMP_AND_SWP) { + aseg->swap_add = cpu_to_be64(wr->swap); + aseg->compare = cpu_to_be64(wr->compare_add); } else { - aseg->swap_add = cpu_to_be64(wr->wr.atomic.compare_add); + aseg->swap_add = cpu_to_be64(wr->compare_add); aseg->compare = 0; } } static void set_tavor_ud_seg(struct mthca_tavor_ud_seg *useg, - struct ib_send_wr *wr) + struct ib_ud_wr *wr) { - useg->lkey = cpu_to_be32(to_mah(wr->wr.ud.ah)->key); - useg->av_addr = cpu_to_be64(to_mah(wr->wr.ud.ah)->avdma); - useg->dqpn = cpu_to_be32(wr->wr.ud.remote_qpn); - useg->qkey = cpu_to_be32(wr->wr.ud.remote_qkey); + useg->lkey = cpu_to_be32(to_mah(wr->ah)->key); + useg->av_addr = cpu_to_be64(to_mah(wr->ah)->avdma); + useg->dqpn = cpu_to_be32(wr->remote_qpn); + useg->qkey = cpu_to_be32(wr->remote_qkey); } static void set_arbel_ud_seg(struct mthca_arbel_ud_seg *useg, - struct ib_send_wr *wr) + struct ib_ud_wr *wr) { - memcpy(useg->av, to_mah(wr->wr.ud.ah)->av, MTHCA_AV_SIZE); - useg->dqpn = cpu_to_be32(wr->wr.ud.remote_qpn); - useg->qkey = cpu_to_be32(wr->wr.ud.remote_qkey); + memcpy(useg->av, to_mah(wr->ah)->av, MTHCA_AV_SIZE); + useg->dqpn = cpu_to_be32(wr->remote_qpn); + useg->qkey = cpu_to_be32(wr->remote_qkey); } int mthca_tavor_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, @@ -1664,11 +1664,11 @@ int mthca_tavor_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, switch (wr->opcode) { case IB_WR_ATOMIC_CMP_AND_SWP: case IB_WR_ATOMIC_FETCH_AND_ADD: - set_raddr_seg(wqe, wr->wr.atomic.remote_addr, - wr->wr.atomic.rkey); + set_raddr_seg(wqe, atomic_wr(wr)->remote_addr, + atomic_wr(wr)->rkey); wqe += sizeof (struct mthca_raddr_seg); - set_atomic_seg(wqe, wr); + set_atomic_seg(wqe, atomic_wr(wr)); wqe += sizeof (struct mthca_atomic_seg); size += (sizeof (struct mthca_raddr_seg) + sizeof (struct mthca_atomic_seg)) / 16; @@ -1677,8 +1677,8 @@ int mthca_tavor_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, case IB_WR_RDMA_WRITE: case IB_WR_RDMA_WRITE_WITH_IMM: case IB_WR_RDMA_READ: - set_raddr_seg(wqe, wr->wr.rdma.remote_addr, - wr->wr.rdma.rkey); + set_raddr_seg(wqe, rdma_wr(wr)->remote_addr, + rdma_wr(wr)->rkey); wqe += sizeof (struct mthca_raddr_seg); size += sizeof (struct mthca_raddr_seg) / 16; break; @@ -1694,8 +1694,8 @@ int mthca_tavor_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, switch (wr->opcode) { case IB_WR_RDMA_WRITE: case IB_WR_RDMA_WRITE_WITH_IMM: - set_raddr_seg(wqe, wr->wr.rdma.remote_addr, - wr->wr.rdma.rkey); + set_raddr_seg(wqe, rdma_wr(wr)->remote_addr, + rdma_wr(wr)->rkey); wqe += sizeof (struct mthca_raddr_seg); size += sizeof (struct mthca_raddr_seg) / 16; break; @@ -1708,13 +1708,13 @@ int mthca_tavor_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, break; case UD: - set_tavor_ud_seg(wqe, wr); + set_tavor_ud_seg(wqe, ud_wr(wr)); wqe += sizeof (struct mthca_tavor_ud_seg); size += sizeof (struct mthca_tavor_ud_seg) / 16; break; case MLX: - err = build_mlx_header(dev, to_msqp(qp), ind, wr, + err = build_mlx_header(dev, to_msqp(qp), ind, ud_wr(wr), wqe - sizeof (struct mthca_next_seg), wqe); if (err) { @@ -2005,11 +2005,11 @@ int mthca_arbel_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, switch (wr->opcode) { case IB_WR_ATOMIC_CMP_AND_SWP: case IB_WR_ATOMIC_FETCH_AND_ADD: - set_raddr_seg(wqe, wr->wr.atomic.remote_addr, - wr->wr.atomic.rkey); + set_raddr_seg(wqe, atomic_wr(wr)->remote_addr, + atomic_wr(wr)->rkey); wqe += sizeof (struct mthca_raddr_seg); - set_atomic_seg(wqe, wr); + set_atomic_seg(wqe, atomic_wr(wr)); wqe += sizeof (struct mthca_atomic_seg); size += (sizeof (struct mthca_raddr_seg) + sizeof (struct mthca_atomic_seg)) / 16; @@ -2018,8 +2018,8 @@ int mthca_arbel_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, case IB_WR_RDMA_READ: case IB_WR_RDMA_WRITE: case IB_WR_RDMA_WRITE_WITH_IMM: - set_raddr_seg(wqe, wr->wr.rdma.remote_addr, - wr->wr.rdma.rkey); + set_raddr_seg(wqe, rdma_wr(wr)->remote_addr, + rdma_wr(wr)->rkey); wqe += sizeof (struct mthca_raddr_seg); size += sizeof (struct mthca_raddr_seg) / 16; break; @@ -2035,8 +2035,8 @@ int mthca_arbel_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, switch (wr->opcode) { case IB_WR_RDMA_WRITE: case IB_WR_RDMA_WRITE_WITH_IMM: - set_raddr_seg(wqe, wr->wr.rdma.remote_addr, - wr->wr.rdma.rkey); + set_raddr_seg(wqe, rdma_wr(wr)->remote_addr, + rdma_wr(wr)->rkey); wqe += sizeof (struct mthca_raddr_seg); size += sizeof (struct mthca_raddr_seg) / 16; break; @@ -2049,13 +2049,13 @@ int mthca_arbel_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, break; case UD: - set_arbel_ud_seg(wqe, wr); + set_arbel_ud_seg(wqe, ud_wr(wr)); wqe += sizeof (struct mthca_arbel_ud_seg); size += sizeof (struct mthca_arbel_ud_seg) / 16; break; case MLX: - err = build_mlx_header(dev, to_msqp(qp), ind, wr, + err = build_mlx_header(dev, to_msqp(qp), ind, ud_wr(wr), wqe - sizeof (struct mthca_next_seg), wqe); if (err) { diff --git a/drivers/infiniband/hw/nes/nes_verbs.c b/drivers/infiniband/hw/nes/nes_verbs.c index 44cb513f9a87..f71b37b75f82 100644 --- a/drivers/infiniband/hw/nes/nes_verbs.c +++ b/drivers/infiniband/hw/nes/nes_verbs.c @@ -3372,9 +3372,9 @@ static int nes_post_send(struct ib_qp *ibqp, struct ib_send_wr *ib_wr, wqe_misc |= NES_IWARP_SQ_WQE_LOCAL_FENCE; set_wqe_32bit_value(wqe->wqe_words, NES_IWARP_SQ_WQE_RDMA_STAG_IDX, - ib_wr->wr.rdma.rkey); + rdma_wr(ib_wr)->rkey); set_wqe_64bit_value(wqe->wqe_words, NES_IWARP_SQ_WQE_RDMA_TO_LOW_IDX, - ib_wr->wr.rdma.remote_addr); + rdma_wr(ib_wr)->remote_addr); if ((ib_wr->send_flags & IB_SEND_INLINE) && ((nes_drv_opt & NES_DRV_OPT_NO_INLINE_DATA) == 0) && @@ -3409,9 +3409,9 @@ static int nes_post_send(struct ib_qp *ibqp, struct ib_send_wr *ib_wr, } set_wqe_64bit_value(wqe->wqe_words, NES_IWARP_SQ_WQE_RDMA_TO_LOW_IDX, - ib_wr->wr.rdma.remote_addr); + rdma_wr(ib_wr)->remote_addr); set_wqe_32bit_value(wqe->wqe_words, NES_IWARP_SQ_WQE_RDMA_STAG_IDX, - ib_wr->wr.rdma.rkey); + rdma_wr(ib_wr)->rkey); set_wqe_32bit_value(wqe->wqe_words, NES_IWARP_SQ_WQE_RDMA_LENGTH_IDX, ib_wr->sg_list->length); set_wqe_64bit_value(wqe->wqe_words, NES_IWARP_SQ_WQE_FRAG0_LOW_IDX, @@ -3428,15 +3428,16 @@ static int nes_post_send(struct ib_qp *ibqp, struct ib_send_wr *ib_wr, case IB_WR_FAST_REG_MR: { int i; - int flags = ib_wr->wr.fast_reg.access_flags; + struct ib_fast_reg_wr *fwr = fast_reg_wr(ib_wr); + int flags = fwr->access_flags; struct nes_ib_fast_reg_page_list *pnesfrpl = - container_of(ib_wr->wr.fast_reg.page_list, + container_of(fwr->page_list, struct nes_ib_fast_reg_page_list, ibfrpl); u64 *src_page_list = pnesfrpl->ibfrpl.page_list; u64 *dst_page_list = pnesfrpl->nes_wqe_pbl.kva; - if (ib_wr->wr.fast_reg.page_list_len > + if (fwr->page_list_len > (NES_4K_PBL_CHUNK_SIZE / sizeof(u64))) { nes_debug(NES_DBG_IW_TX, "SQ_FMR: bad page_list_len\n"); err = -EINVAL; @@ -3445,19 +3446,19 @@ static int nes_post_send(struct ib_qp *ibqp, struct ib_send_wr *ib_wr, wqe_misc = NES_IWARP_SQ_OP_FAST_REG; set_wqe_64bit_value(wqe->wqe_words, NES_IWARP_SQ_FMR_WQE_VA_FBO_LOW_IDX, - ib_wr->wr.fast_reg.iova_start); + fwr->iova_start); set_wqe_32bit_value(wqe->wqe_words, NES_IWARP_SQ_FMR_WQE_LENGTH_LOW_IDX, - ib_wr->wr.fast_reg.length); + fwr->length); set_wqe_32bit_value(wqe->wqe_words, NES_IWARP_SQ_FMR_WQE_LENGTH_HIGH_IDX, 0); set_wqe_32bit_value(wqe->wqe_words, NES_IWARP_SQ_FMR_WQE_MR_STAG_IDX, - ib_wr->wr.fast_reg.rkey); + fwr->rkey); /* Set page size: */ - if (ib_wr->wr.fast_reg.page_shift == 12) { + if (fwr->page_shift == 12) { wqe_misc |= NES_IWARP_SQ_FMR_WQE_PAGE_SIZE_4K; - } else if (ib_wr->wr.fast_reg.page_shift == 21) { + } else if (fwr->page_shift == 21) { wqe_misc |= NES_IWARP_SQ_FMR_WQE_PAGE_SIZE_2M; } else { nes_debug(NES_DBG_IW_TX, "Invalid page shift," @@ -3480,11 +3481,11 @@ static int nes_post_send(struct ib_qp *ibqp, struct ib_send_wr *ib_wr, wqe_misc |= NES_IWARP_SQ_FMR_WQE_RIGHTS_ENABLE_WINDOW_BIND; /* Fill in PBL info: */ - if (ib_wr->wr.fast_reg.page_list_len > + if (fwr->page_list_len > pnesfrpl->ibfrpl.max_page_list_len) { nes_debug(NES_DBG_IW_TX, "Invalid page list length," " ib_wr=%p, value=%u, max=%u\n", - ib_wr, ib_wr->wr.fast_reg.page_list_len, + ib_wr, fwr->page_list_len, pnesfrpl->ibfrpl.max_page_list_len); err = -EINVAL; break; @@ -3496,19 +3497,19 @@ static int nes_post_send(struct ib_qp *ibqp, struct ib_send_wr *ib_wr, set_wqe_32bit_value(wqe->wqe_words, NES_IWARP_SQ_FMR_WQE_PBL_LENGTH_IDX, - ib_wr->wr.fast_reg.page_list_len * 8); + fwr->page_list_len * 8); - for (i = 0; i < ib_wr->wr.fast_reg.page_list_len; i++) + for (i = 0; i < fwr->page_list_len; i++) dst_page_list[i] = cpu_to_le64(src_page_list[i]); nes_debug(NES_DBG_IW_TX, "SQ_FMR: iova_start: %llx, " "length: %d, rkey: %0x, pgl_paddr: %llx, " "page_list_len: %u, wqe_misc: %x\n", - (unsigned long long) ib_wr->wr.fast_reg.iova_start, - ib_wr->wr.fast_reg.length, - ib_wr->wr.fast_reg.rkey, + (unsigned long long) fwr->iova_start, + fwr->length, + fwr->rkey, (unsigned long long) pnesfrpl->nes_wqe_pbl.paddr, - ib_wr->wr.fast_reg.page_list_len, + fwr->page_list_len, wqe_misc); break; } diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c b/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c index 1f3affb6a477..eb09e224acb9 100644 --- a/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c +++ b/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c @@ -1997,13 +1997,13 @@ static void ocrdma_build_ud_hdr(struct ocrdma_qp *qp, { struct ocrdma_ewqe_ud_hdr *ud_hdr = (struct ocrdma_ewqe_ud_hdr *)(hdr + 1); - struct ocrdma_ah *ah = get_ocrdma_ah(wr->wr.ud.ah); + struct ocrdma_ah *ah = get_ocrdma_ah(ud_wr(wr)->ah); - ud_hdr->rsvd_dest_qpn = wr->wr.ud.remote_qpn; + ud_hdr->rsvd_dest_qpn = ud_wr(wr)->remote_qpn; if (qp->qp_type == IB_QPT_GSI) ud_hdr->qkey = qp->qkey; else - ud_hdr->qkey = wr->wr.ud.remote_qkey; + ud_hdr->qkey = ud_wr(wr)->remote_qkey; ud_hdr->rsvd_ahid = ah->id; if (ah->av->valid & OCRDMA_AV_VLAN_VALID) hdr->cw |= (OCRDMA_FLAG_AH_VLAN_PR << OCRDMA_WQE_FLAGS_SHIFT); @@ -2106,9 +2106,9 @@ static int ocrdma_build_write(struct ocrdma_qp *qp, struct ocrdma_hdr_wqe *hdr, status = ocrdma_build_inline_sges(qp, hdr, sge, wr, wqe_size); if (status) return status; - ext_rw->addr_lo = wr->wr.rdma.remote_addr; - ext_rw->addr_hi = upper_32_bits(wr->wr.rdma.remote_addr); - ext_rw->lrkey = wr->wr.rdma.rkey; + ext_rw->addr_lo = rdma_wr(wr)->remote_addr; + ext_rw->addr_hi = upper_32_bits(rdma_wr(wr)->remote_addr); + ext_rw->lrkey = rdma_wr(wr)->rkey; ext_rw->len = hdr->total_len; return 0; } @@ -2126,13 +2126,14 @@ static void ocrdma_build_read(struct ocrdma_qp *qp, struct ocrdma_hdr_wqe *hdr, hdr->cw |= (OCRDMA_READ << OCRDMA_WQE_OPCODE_SHIFT); hdr->cw |= (OCRDMA_TYPE_LKEY << OCRDMA_WQE_TYPE_SHIFT); - ext_rw->addr_lo = wr->wr.rdma.remote_addr; - ext_rw->addr_hi = upper_32_bits(wr->wr.rdma.remote_addr); - ext_rw->lrkey = wr->wr.rdma.rkey; + ext_rw->addr_lo = rdma_wr(wr)->remote_addr; + ext_rw->addr_hi = upper_32_bits(rdma_wr(wr)->remote_addr); + ext_rw->lrkey = rdma_wr(wr)->rkey; ext_rw->len = hdr->total_len; } -static void build_frmr_pbes(struct ib_send_wr *wr, struct ocrdma_pbl *pbl_tbl, +static void build_frmr_pbes(struct ib_fast_reg_wr *wr, + struct ocrdma_pbl *pbl_tbl, struct ocrdma_hw_mr *hwmr) { int i; @@ -2144,12 +2145,12 @@ static void build_frmr_pbes(struct ib_send_wr *wr, struct ocrdma_pbl *pbl_tbl, num_pbes = 0; /* go through the OS phy regions & fill hw pbe entries into pbls. */ - for (i = 0; i < wr->wr.fast_reg.page_list_len; i++) { + for (i = 0; i < wr->page_list_len; i++) { /* number of pbes can be more for one OS buf, when * buffers are of different sizes. * split the ib_buf to one or more pbes. */ - buf_addr = wr->wr.fast_reg.page_list->page_list[i]; + buf_addr = wr->page_list->page_list[i]; pbe->pa_lo = cpu_to_le32((u32) (buf_addr & PAGE_MASK)); pbe->pa_hi = cpu_to_le32((u32) upper_32_bits(buf_addr)); num_pbes += 1; @@ -2178,9 +2179,10 @@ static int get_encoded_page_size(int pg_sz) static int ocrdma_build_fr(struct ocrdma_qp *qp, struct ocrdma_hdr_wqe *hdr, - struct ib_send_wr *wr) + struct ib_send_wr *send_wr) { u64 fbo; + struct ib_fast_reg_wr *wr = fast_reg_wr(send_wr); struct ocrdma_ewqe_fr *fast_reg = (struct ocrdma_ewqe_fr *)(hdr + 1); struct ocrdma_mr *mr; struct ocrdma_dev *dev = get_ocrdma_dev(qp->ibqp.device); @@ -2188,33 +2190,32 @@ static int ocrdma_build_fr(struct ocrdma_qp *qp, struct ocrdma_hdr_wqe *hdr, wqe_size = roundup(wqe_size, OCRDMA_WQE_ALIGN_BYTES); - if (wr->wr.fast_reg.page_list_len > dev->attr.max_pages_per_frmr) + if (wr->page_list_len > dev->attr.max_pages_per_frmr) return -EINVAL; hdr->cw |= (OCRDMA_FR_MR << OCRDMA_WQE_OPCODE_SHIFT); hdr->cw |= ((wqe_size / OCRDMA_WQE_STRIDE) << OCRDMA_WQE_SIZE_SHIFT); - if (wr->wr.fast_reg.page_list_len == 0) + if (wr->page_list_len == 0) BUG(); - if (wr->wr.fast_reg.access_flags & IB_ACCESS_LOCAL_WRITE) + if (wr->access_flags & IB_ACCESS_LOCAL_WRITE) hdr->rsvd_lkey_flags |= OCRDMA_LKEY_FLAG_LOCAL_WR; - if (wr->wr.fast_reg.access_flags & IB_ACCESS_REMOTE_WRITE) + if (wr->access_flags & IB_ACCESS_REMOTE_WRITE) hdr->rsvd_lkey_flags |= OCRDMA_LKEY_FLAG_REMOTE_WR; - if (wr->wr.fast_reg.access_flags & IB_ACCESS_REMOTE_READ) + if (wr->access_flags & IB_ACCESS_REMOTE_READ) hdr->rsvd_lkey_flags |= OCRDMA_LKEY_FLAG_REMOTE_RD; - hdr->lkey = wr->wr.fast_reg.rkey; - hdr->total_len = wr->wr.fast_reg.length; + hdr->lkey = wr->rkey; + hdr->total_len = wr->length; - fbo = wr->wr.fast_reg.iova_start - - (wr->wr.fast_reg.page_list->page_list[0] & PAGE_MASK); + fbo = wr->iova_start - (wr->page_list->page_list[0] & PAGE_MASK); - fast_reg->va_hi = upper_32_bits(wr->wr.fast_reg.iova_start); - fast_reg->va_lo = (u32) (wr->wr.fast_reg.iova_start & 0xffffffff); + fast_reg->va_hi = upper_32_bits(wr->iova_start); + fast_reg->va_lo = (u32) (wr->iova_start & 0xffffffff); fast_reg->fbo_hi = upper_32_bits(fbo); fast_reg->fbo_lo = (u32) fbo & 0xffffffff; - fast_reg->num_sges = wr->wr.fast_reg.page_list_len; + fast_reg->num_sges = wr->page_list_len; fast_reg->size_sge = - get_encoded_page_size(1 << wr->wr.fast_reg.page_shift); + get_encoded_page_size(1 << wr->page_shift); mr = (struct ocrdma_mr *) (unsigned long) dev->stag_arr[(hdr->lkey >> 8) & (OCRDMA_MAX_STAG - 1)]; build_frmr_pbes(wr, mr->hwmr.pbl_table, &mr->hwmr); diff --git a/drivers/infiniband/hw/qib/qib_keys.c b/drivers/infiniband/hw/qib/qib_keys.c index 5afaa218508d..eaf139a33b2e 100644 --- a/drivers/infiniband/hw/qib/qib_keys.c +++ b/drivers/infiniband/hw/qib/qib_keys.c @@ -338,12 +338,13 @@ bail: /* * Initialize the memory region specified by the work reqeust. */ -int qib_fast_reg_mr(struct qib_qp *qp, struct ib_send_wr *wr) +int qib_fast_reg_mr(struct qib_qp *qp, struct ib_send_wr *send_wr) { + struct ib_fast_reg_wr *wr = fast_reg_wr(send_wr); struct qib_lkey_table *rkt = &to_idev(qp->ibqp.device)->lk_table; struct qib_pd *pd = to_ipd(qp->ibqp.pd); struct qib_mregion *mr; - u32 rkey = wr->wr.fast_reg.rkey; + u32 rkey = wr->rkey; unsigned i, n, m; int ret = -EINVAL; unsigned long flags; @@ -360,22 +361,22 @@ int qib_fast_reg_mr(struct qib_qp *qp, struct ib_send_wr *wr) if (unlikely(mr == NULL || qp->ibqp.pd != mr->pd)) goto bail; - if (wr->wr.fast_reg.page_list_len > mr->max_segs) + if (wr->page_list_len > mr->max_segs) goto bail; - ps = 1UL << wr->wr.fast_reg.page_shift; - if (wr->wr.fast_reg.length > ps * wr->wr.fast_reg.page_list_len) + ps = 1UL << wr->page_shift; + if (wr->length > ps * wr->page_list_len) goto bail; - mr->user_base = wr->wr.fast_reg.iova_start; - mr->iova = wr->wr.fast_reg.iova_start; + mr->user_base = wr->iova_start; + mr->iova = wr->iova_start; mr->lkey = rkey; - mr->length = wr->wr.fast_reg.length; - mr->access_flags = wr->wr.fast_reg.access_flags; - page_list = wr->wr.fast_reg.page_list->page_list; + mr->length = wr->length; + mr->access_flags = wr->access_flags; + page_list = wr->page_list->page_list; m = 0; n = 0; - for (i = 0; i < wr->wr.fast_reg.page_list_len; i++) { + for (i = 0; i < wr->page_list_len; i++) { mr->map[m]->segs[n].vaddr = (void *) page_list[i]; mr->map[m]->segs[n].length = ps; if (++n == QIB_SEGSZ) { diff --git a/drivers/infiniband/hw/qib/qib_qp.c b/drivers/infiniband/hw/qib/qib_qp.c index 4fa88ba2963e..40f85bb3e0d3 100644 --- a/drivers/infiniband/hw/qib/qib_qp.c +++ b/drivers/infiniband/hw/qib/qib_qp.c @@ -436,7 +436,7 @@ static void clear_mr_refs(struct qib_qp *qp, int clr_sends) if (qp->ibqp.qp_type == IB_QPT_UD || qp->ibqp.qp_type == IB_QPT_SMI || qp->ibqp.qp_type == IB_QPT_GSI) - atomic_dec(&to_iah(wqe->wr.wr.ud.ah)->refcount); + atomic_dec(&to_iah(wqe->ud_wr.ah)->refcount); if (++qp->s_last >= qp->s_size) qp->s_last = 0; } diff --git a/drivers/infiniband/hw/qib/qib_rc.c b/drivers/infiniband/hw/qib/qib_rc.c index 4544d6f88ad7..e6b7556d5221 100644 --- a/drivers/infiniband/hw/qib/qib_rc.c +++ b/drivers/infiniband/hw/qib/qib_rc.c @@ -373,10 +373,11 @@ int qib_make_rc_req(struct qib_qp *qp) qp->s_flags |= QIB_S_WAIT_SSN_CREDIT; goto bail; } + ohdr->u.rc.reth.vaddr = - cpu_to_be64(wqe->wr.wr.rdma.remote_addr); + cpu_to_be64(wqe->rdma_wr.remote_addr); ohdr->u.rc.reth.rkey = - cpu_to_be32(wqe->wr.wr.rdma.rkey); + cpu_to_be32(wqe->rdma_wr.rkey); ohdr->u.rc.reth.length = cpu_to_be32(len); hwords += sizeof(struct ib_reth) / sizeof(u32); wqe->lpsn = wqe->psn; @@ -386,15 +387,15 @@ int qib_make_rc_req(struct qib_qp *qp) len = pmtu; break; } - if (wqe->wr.opcode == IB_WR_RDMA_WRITE) + if (wqe->rdma_wr.wr.opcode == IB_WR_RDMA_WRITE) qp->s_state = OP(RDMA_WRITE_ONLY); else { - qp->s_state = - OP(RDMA_WRITE_ONLY_WITH_IMMEDIATE); + qp->s_state = OP(RDMA_WRITE_ONLY_WITH_IMMEDIATE); /* Immediate data comes after RETH */ - ohdr->u.rc.imm_data = wqe->wr.ex.imm_data; + ohdr->u.rc.imm_data = + wqe->rdma_wr.wr.ex.imm_data; hwords += 1; - if (wqe->wr.send_flags & IB_SEND_SOLICITED) + if (wqe->rdma_wr.wr.send_flags & IB_SEND_SOLICITED) bth0 |= IB_BTH_SOLICITED; } bth2 |= IB_BTH_REQ_ACK; @@ -424,10 +425,11 @@ int qib_make_rc_req(struct qib_qp *qp) qp->s_next_psn += (len - 1) / pmtu; wqe->lpsn = qp->s_next_psn++; } + ohdr->u.rc.reth.vaddr = - cpu_to_be64(wqe->wr.wr.rdma.remote_addr); + cpu_to_be64(wqe->rdma_wr.remote_addr); ohdr->u.rc.reth.rkey = - cpu_to_be32(wqe->wr.wr.rdma.rkey); + cpu_to_be32(wqe->rdma_wr.rkey); ohdr->u.rc.reth.length = cpu_to_be32(len); qp->s_state = OP(RDMA_READ_REQUEST); hwords += sizeof(ohdr->u.rc.reth) / sizeof(u32); @@ -455,24 +457,24 @@ int qib_make_rc_req(struct qib_qp *qp) qp->s_lsn++; wqe->lpsn = wqe->psn; } - if (wqe->wr.opcode == IB_WR_ATOMIC_CMP_AND_SWP) { + if (wqe->atomic_wr.wr.opcode == IB_WR_ATOMIC_CMP_AND_SWP) { qp->s_state = OP(COMPARE_SWAP); ohdr->u.atomic_eth.swap_data = cpu_to_be64( - wqe->wr.wr.atomic.swap); + wqe->atomic_wr.swap); ohdr->u.atomic_eth.compare_data = cpu_to_be64( - wqe->wr.wr.atomic.compare_add); + wqe->atomic_wr.compare_add); } else { qp->s_state = OP(FETCH_ADD); ohdr->u.atomic_eth.swap_data = cpu_to_be64( - wqe->wr.wr.atomic.compare_add); + wqe->atomic_wr.compare_add); ohdr->u.atomic_eth.compare_data = 0; } ohdr->u.atomic_eth.vaddr[0] = cpu_to_be32( - wqe->wr.wr.atomic.remote_addr >> 32); + wqe->atomic_wr.remote_addr >> 32); ohdr->u.atomic_eth.vaddr[1] = cpu_to_be32( - wqe->wr.wr.atomic.remote_addr); + wqe->atomic_wr.remote_addr); ohdr->u.atomic_eth.rkey = cpu_to_be32( - wqe->wr.wr.atomic.rkey); + wqe->atomic_wr.rkey); hwords += sizeof(struct ib_atomic_eth) / sizeof(u32); ss = NULL; len = 0; @@ -597,9 +599,9 @@ int qib_make_rc_req(struct qib_qp *qp) */ len = ((qp->s_psn - wqe->psn) & QIB_PSN_MASK) * pmtu; ohdr->u.rc.reth.vaddr = - cpu_to_be64(wqe->wr.wr.rdma.remote_addr + len); + cpu_to_be64(wqe->rdma_wr.remote_addr + len); ohdr->u.rc.reth.rkey = - cpu_to_be32(wqe->wr.wr.rdma.rkey); + cpu_to_be32(wqe->rdma_wr.rkey); ohdr->u.rc.reth.length = cpu_to_be32(wqe->length - len); qp->s_state = OP(RDMA_READ_REQUEST); hwords += sizeof(ohdr->u.rc.reth) / sizeof(u32); diff --git a/drivers/infiniband/hw/qib/qib_ruc.c b/drivers/infiniband/hw/qib/qib_ruc.c index 22e356ca8058..b1aa21bdd484 100644 --- a/drivers/infiniband/hw/qib/qib_ruc.c +++ b/drivers/infiniband/hw/qib/qib_ruc.c @@ -459,8 +459,8 @@ again: if (wqe->length == 0) break; if (unlikely(!qib_rkey_ok(qp, &qp->r_sge.sge, wqe->length, - wqe->wr.wr.rdma.remote_addr, - wqe->wr.wr.rdma.rkey, + wqe->rdma_wr.remote_addr, + wqe->rdma_wr.rkey, IB_ACCESS_REMOTE_WRITE))) goto acc_err; qp->r_sge.sg_list = NULL; @@ -472,8 +472,8 @@ again: if (unlikely(!(qp->qp_access_flags & IB_ACCESS_REMOTE_READ))) goto inv_err; if (unlikely(!qib_rkey_ok(qp, &sqp->s_sge.sge, wqe->length, - wqe->wr.wr.rdma.remote_addr, - wqe->wr.wr.rdma.rkey, + wqe->rdma_wr.remote_addr, + wqe->rdma_wr.rkey, IB_ACCESS_REMOTE_READ))) goto acc_err; release = 0; @@ -490,18 +490,18 @@ again: if (unlikely(!(qp->qp_access_flags & IB_ACCESS_REMOTE_ATOMIC))) goto inv_err; if (unlikely(!qib_rkey_ok(qp, &qp->r_sge.sge, sizeof(u64), - wqe->wr.wr.atomic.remote_addr, - wqe->wr.wr.atomic.rkey, + wqe->atomic_wr.remote_addr, + wqe->atomic_wr.rkey, IB_ACCESS_REMOTE_ATOMIC))) goto acc_err; /* Perform atomic OP and save result. */ maddr = (atomic64_t *) qp->r_sge.sge.vaddr; - sdata = wqe->wr.wr.atomic.compare_add; + sdata = wqe->atomic_wr.compare_add; *(u64 *) sqp->s_sge.sge.vaddr = - (wqe->wr.opcode == IB_WR_ATOMIC_FETCH_AND_ADD) ? + (wqe->atomic_wr.wr.opcode == IB_WR_ATOMIC_FETCH_AND_ADD) ? (u64) atomic64_add_return(sdata, maddr) - sdata : (u64) cmpxchg((u64 *) qp->r_sge.sge.vaddr, - sdata, wqe->wr.wr.atomic.swap); + sdata, wqe->atomic_wr.swap); qib_put_mr(qp->r_sge.sge.mr); qp->r_sge.num_sge = 0; goto send_comp; @@ -785,7 +785,7 @@ void qib_send_complete(struct qib_qp *qp, struct qib_swqe *wqe, if (qp->ibqp.qp_type == IB_QPT_UD || qp->ibqp.qp_type == IB_QPT_SMI || qp->ibqp.qp_type == IB_QPT_GSI) - atomic_dec(&to_iah(wqe->wr.wr.ud.ah)->refcount); + atomic_dec(&to_iah(wqe->ud_wr.ah)->refcount); /* See ch. 11.2.4.1 and 10.7.3.1 */ if (!(qp->s_flags & QIB_S_SIGNAL_REQ_WR) || diff --git a/drivers/infiniband/hw/qib/qib_uc.c b/drivers/infiniband/hw/qib/qib_uc.c index aa3a8035bb68..06a564589c35 100644 --- a/drivers/infiniband/hw/qib/qib_uc.c +++ b/drivers/infiniband/hw/qib/qib_uc.c @@ -129,9 +129,9 @@ int qib_make_uc_req(struct qib_qp *qp) case IB_WR_RDMA_WRITE: case IB_WR_RDMA_WRITE_WITH_IMM: ohdr->u.rc.reth.vaddr = - cpu_to_be64(wqe->wr.wr.rdma.remote_addr); + cpu_to_be64(wqe->rdma_wr.remote_addr); ohdr->u.rc.reth.rkey = - cpu_to_be32(wqe->wr.wr.rdma.rkey); + cpu_to_be32(wqe->rdma_wr.rkey); ohdr->u.rc.reth.length = cpu_to_be32(len); hwords += sizeof(struct ib_reth) / 4; if (len > pmtu) { diff --git a/drivers/infiniband/hw/qib/qib_ud.c b/drivers/infiniband/hw/qib/qib_ud.c index 26243b722b5e..59193f67ea78 100644 --- a/drivers/infiniband/hw/qib/qib_ud.c +++ b/drivers/infiniband/hw/qib/qib_ud.c @@ -59,7 +59,7 @@ static void qib_ud_loopback(struct qib_qp *sqp, struct qib_swqe *swqe) u32 length; enum ib_qp_type sqptype, dqptype; - qp = qib_lookup_qpn(ibp, swqe->wr.wr.ud.remote_qpn); + qp = qib_lookup_qpn(ibp, swqe->ud_wr.remote_qpn); if (!qp) { ibp->n_pkt_drops++; return; @@ -76,7 +76,7 @@ static void qib_ud_loopback(struct qib_qp *sqp, struct qib_swqe *swqe) goto drop; } - ah_attr = &to_iah(swqe->wr.wr.ud.ah)->attr; + ah_attr = &to_iah(swqe->ud_wr.ah)->attr; ppd = ppd_from_ibp(ibp); if (qp->ibqp.qp_num > 1) { @@ -106,8 +106,8 @@ static void qib_ud_loopback(struct qib_qp *sqp, struct qib_swqe *swqe) if (qp->ibqp.qp_num) { u32 qkey; - qkey = (int)swqe->wr.wr.ud.remote_qkey < 0 ? - sqp->qkey : swqe->wr.wr.ud.remote_qkey; + qkey = (int)swqe->ud_wr.remote_qkey < 0 ? + sqp->qkey : swqe->ud_wr.remote_qkey; if (unlikely(qkey != qp->qkey)) { u16 lid; @@ -210,7 +210,7 @@ static void qib_ud_loopback(struct qib_qp *sqp, struct qib_swqe *swqe) wc.qp = &qp->ibqp; wc.src_qp = sqp->ibqp.qp_num; wc.pkey_index = qp->ibqp.qp_type == IB_QPT_GSI ? - swqe->wr.wr.ud.pkey_index : 0; + swqe->ud_wr.pkey_index : 0; wc.slid = ppd->lid | (ah_attr->src_path_bits & ((1 << ppd->lmc) - 1)); wc.sl = ah_attr->sl; wc.dlid_path_bits = ah_attr->dlid & ((1 << ppd->lmc) - 1); @@ -277,7 +277,7 @@ int qib_make_ud_req(struct qib_qp *qp) /* Construct the header. */ ibp = to_iport(qp->ibqp.device, qp->port_num); ppd = ppd_from_ibp(ibp); - ah_attr = &to_iah(wqe->wr.wr.ud.ah)->attr; + ah_attr = &to_iah(wqe->ud_wr.ah)->attr; if (ah_attr->dlid >= QIB_MULTICAST_LID_BASE) { if (ah_attr->dlid != QIB_PERMISSIVE_LID) this_cpu_inc(ibp->pmastats->n_multicast_xmit); @@ -363,7 +363,7 @@ int qib_make_ud_req(struct qib_qp *qp) bth0 |= extra_bytes << 20; bth0 |= qp->ibqp.qp_type == IB_QPT_SMI ? QIB_DEFAULT_P_KEY : qib_get_pkey(ibp, qp->ibqp.qp_type == IB_QPT_GSI ? - wqe->wr.wr.ud.pkey_index : qp->s_pkey_index); + wqe->ud_wr.pkey_index : qp->s_pkey_index); ohdr->bth[0] = cpu_to_be32(bth0); /* * Use the multicast QP if the destination LID is a multicast LID. @@ -371,14 +371,14 @@ int qib_make_ud_req(struct qib_qp *qp) ohdr->bth[1] = ah_attr->dlid >= QIB_MULTICAST_LID_BASE && ah_attr->dlid != QIB_PERMISSIVE_LID ? cpu_to_be32(QIB_MULTICAST_QPN) : - cpu_to_be32(wqe->wr.wr.ud.remote_qpn); + cpu_to_be32(wqe->ud_wr.remote_qpn); ohdr->bth[2] = cpu_to_be32(qp->s_next_psn++ & QIB_PSN_MASK); /* * Qkeys with the high order bit set mean use the * qkey from the QP context instead of the WR (see 10.2.5). */ - ohdr->u.ud.deth[0] = cpu_to_be32((int)wqe->wr.wr.ud.remote_qkey < 0 ? - qp->qkey : wqe->wr.wr.ud.remote_qkey); + ohdr->u.ud.deth[0] = cpu_to_be32((int)wqe->ud_wr.remote_qkey < 0 ? + qp->qkey : wqe->ud_wr.remote_qkey); ohdr->u.ud.deth[1] = cpu_to_be32(qp->ibqp.qp_num); done: diff --git a/drivers/infiniband/hw/qib/qib_verbs.c b/drivers/infiniband/hw/qib/qib_verbs.c index 3dcc4985b60f..a6b0b098ff30 100644 --- a/drivers/infiniband/hw/qib/qib_verbs.c +++ b/drivers/infiniband/hw/qib/qib_verbs.c @@ -374,7 +374,7 @@ static int qib_post_one_send(struct qib_qp *qp, struct ib_send_wr *wr, wr->opcode != IB_WR_SEND_WITH_IMM) goto bail_inval; /* Check UD destination address PD */ - if (qp->ibqp.pd != wr->wr.ud.ah->pd) + if (qp->ibqp.pd != ud_wr(wr)->ah->pd) goto bail_inval; } else if ((unsigned) wr->opcode > IB_WR_ATOMIC_FETCH_AND_ADD) goto bail_inval; @@ -397,7 +397,23 @@ static int qib_post_one_send(struct qib_qp *qp, struct ib_send_wr *wr, rkt = &to_idev(qp->ibqp.device)->lk_table; pd = to_ipd(qp->ibqp.pd); wqe = get_swqe_ptr(qp, qp->s_head); - wqe->wr = *wr; + + if (qp->ibqp.qp_type != IB_QPT_UC && + qp->ibqp.qp_type != IB_QPT_RC) + memcpy(&wqe->ud_wr, ud_wr(wr), sizeof(wqe->ud_wr)); + else if (wr->opcode == IB_WR_FAST_REG_MR) + memcpy(&wqe->fast_reg_wr, fast_reg_wr(wr), + sizeof(wqe->fast_reg_wr)); + else if (wr->opcode == IB_WR_RDMA_WRITE_WITH_IMM || + wr->opcode == IB_WR_RDMA_WRITE || + wr->opcode == IB_WR_RDMA_READ) + memcpy(&wqe->rdma_wr, rdma_wr(wr), sizeof(wqe->rdma_wr)); + else if (wr->opcode == IB_WR_ATOMIC_CMP_AND_SWP || + wr->opcode == IB_WR_ATOMIC_FETCH_AND_ADD) + memcpy(&wqe->atomic_wr, atomic_wr(wr), sizeof(wqe->atomic_wr)); + else + memcpy(&wqe->wr, wr, sizeof(wqe->wr)); + wqe->length = 0; j = 0; if (wr->num_sge) { @@ -426,7 +442,7 @@ static int qib_post_one_send(struct qib_qp *qp, struct ib_send_wr *wr, qp->port_num - 1)->ibmtu) goto bail_inval_free; else - atomic_inc(&to_iah(wr->wr.ud.ah)->refcount); + atomic_inc(&to_iah(ud_wr(wr)->ah)->refcount); wqe->ssn = qp->s_ssn++; qp->s_head = next; diff --git a/drivers/infiniband/hw/qib/qib_verbs.h b/drivers/infiniband/hw/qib/qib_verbs.h index a08df70e8503..8aa16851a5e6 100644 --- a/drivers/infiniband/hw/qib/qib_verbs.h +++ b/drivers/infiniband/hw/qib/qib_verbs.h @@ -338,7 +338,13 @@ struct qib_mr { * in qp->s_max_sge. */ struct qib_swqe { - struct ib_send_wr wr; /* don't use wr.sg_list */ + union { + struct ib_send_wr wr; /* don't use wr.sg_list */ + struct ib_ud_wr ud_wr; + struct ib_fast_reg_wr fast_reg_wr; + struct ib_rdma_wr rdma_wr; + struct ib_atomic_wr atomic_wr; + }; u32 psn; /* first packet sequence number */ u32 lpsn; /* last packet sequence number */ u32 ssn; /* send sequence number */ diff --git a/drivers/infiniband/ulp/ipoib/ipoib.h b/drivers/infiniband/ulp/ipoib/ipoib.h index 4cd5428a2399..453860ade65e 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib.h +++ b/drivers/infiniband/ulp/ipoib/ipoib.h @@ -360,7 +360,7 @@ struct ipoib_dev_priv { unsigned tx_head; unsigned tx_tail; struct ib_sge tx_sge[MAX_SKB_FRAGS + 1]; - struct ib_send_wr tx_wr; + struct ib_ud_wr tx_wr; unsigned tx_outstanding; struct ib_wc send_wc[MAX_SEND_CQE]; @@ -527,7 +527,7 @@ static inline void ipoib_build_sge(struct ipoib_dev_priv *priv, priv->tx_sge[i + off].addr = mapping[i + off]; priv->tx_sge[i + off].length = skb_frag_size(&frags[i]); } - priv->tx_wr.num_sge = nr_frags + off; + priv->tx_wr.wr.num_sge = nr_frags + off; } #ifdef CONFIG_INFINIBAND_IPOIB_DEBUG diff --git a/drivers/infiniband/ulp/ipoib/ipoib_cm.c b/drivers/infiniband/ulp/ipoib/ipoib_cm.c index c78dc1638030..3ae9726efb98 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_cm.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_cm.c @@ -700,9 +700,9 @@ static inline int post_send(struct ipoib_dev_priv *priv, ipoib_build_sge(priv, tx_req); - priv->tx_wr.wr_id = wr_id | IPOIB_OP_CM; + priv->tx_wr.wr.wr_id = wr_id | IPOIB_OP_CM; - return ib_post_send(tx->qp, &priv->tx_wr, &bad_wr); + return ib_post_send(tx->qp, &priv->tx_wr.wr, &bad_wr); } void ipoib_cm_send(struct net_device *dev, struct sk_buff *skb, struct ipoib_cm_tx *tx) diff --git a/drivers/infiniband/ulp/ipoib/ipoib_ib.c b/drivers/infiniband/ulp/ipoib/ipoib_ib.c index d266667ca9b8..5ea0c14070d1 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_ib.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_ib.c @@ -518,19 +518,19 @@ static inline int post_send(struct ipoib_dev_priv *priv, ipoib_build_sge(priv, tx_req); - priv->tx_wr.wr_id = wr_id; - priv->tx_wr.wr.ud.remote_qpn = qpn; - priv->tx_wr.wr.ud.ah = address; + priv->tx_wr.wr.wr_id = wr_id; + priv->tx_wr.remote_qpn = qpn; + priv->tx_wr.ah = address; if (head) { - priv->tx_wr.wr.ud.mss = skb_shinfo(skb)->gso_size; - priv->tx_wr.wr.ud.header = head; - priv->tx_wr.wr.ud.hlen = hlen; - priv->tx_wr.opcode = IB_WR_LSO; + priv->tx_wr.mss = skb_shinfo(skb)->gso_size; + priv->tx_wr.header = head; + priv->tx_wr.hlen = hlen; + priv->tx_wr.wr.opcode = IB_WR_LSO; } else - priv->tx_wr.opcode = IB_WR_SEND; + priv->tx_wr.wr.opcode = IB_WR_SEND; - return ib_post_send(priv->qp, &priv->tx_wr, &bad_wr); + return ib_post_send(priv->qp, &priv->tx_wr.wr, &bad_wr); } void ipoib_send(struct net_device *dev, struct sk_buff *skb, @@ -583,9 +583,9 @@ void ipoib_send(struct net_device *dev, struct sk_buff *skb, } if (skb->ip_summed == CHECKSUM_PARTIAL) - priv->tx_wr.send_flags |= IB_SEND_IP_CSUM; + priv->tx_wr.wr.send_flags |= IB_SEND_IP_CSUM; else - priv->tx_wr.send_flags &= ~IB_SEND_IP_CSUM; + priv->tx_wr.wr.send_flags &= ~IB_SEND_IP_CSUM; if (++priv->tx_outstanding == ipoib_sendq_size) { ipoib_dbg(priv, "TX ring full, stopping kernel net queue\n"); diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c index f74316e679d2..65d916cc70c7 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_main.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c @@ -461,7 +461,7 @@ int ipoib_set_mode(struct net_device *dev, const char *buf) netdev_update_features(dev); dev_set_mtu(dev, ipoib_cm_max_mtu(dev)); rtnl_unlock(); - priv->tx_wr.send_flags &= ~IB_SEND_IP_CSUM; + priv->tx_wr.wr.send_flags &= ~IB_SEND_IP_CSUM; ipoib_flush_paths(dev); rtnl_lock(); diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c index 136cbefe00f8..029589b21fe9 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c @@ -245,7 +245,7 @@ static int ipoib_mcast_join_finish(struct ipoib_mcast *mcast, priv->qkey = be32_to_cpu(priv->broadcast->mcmember.qkey); spin_unlock_irq(&priv->lock); - priv->tx_wr.wr.ud.remote_qkey = priv->qkey; + priv->tx_wr.remote_qkey = priv->qkey; set_qkey = 1; } diff --git a/drivers/infiniband/ulp/ipoib/ipoib_verbs.c b/drivers/infiniband/ulp/ipoib/ipoib_verbs.c index 78845b6e8b81..d48c5bae7877 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_verbs.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_verbs.c @@ -221,9 +221,9 @@ int ipoib_transport_dev_init(struct net_device *dev, struct ib_device *ca) for (i = 0; i < MAX_SKB_FRAGS + 1; ++i) priv->tx_sge[i].lkey = priv->pd->local_dma_lkey; - priv->tx_wr.opcode = IB_WR_SEND; - priv->tx_wr.sg_list = priv->tx_sge; - priv->tx_wr.send_flags = IB_SEND_SIGNALED; + priv->tx_wr.wr.opcode = IB_WR_SEND; + priv->tx_wr.wr.sg_list = priv->tx_sge; + priv->tx_wr.wr.send_flags = IB_SEND_SIGNALED; priv->rx_sge[0].lkey = priv->pd->local_dma_lkey; diff --git a/drivers/infiniband/ulp/iser/iscsi_iser.h b/drivers/infiniband/ulp/iser/iscsi_iser.h index a5edd6ede692..2fab519dbd86 100644 --- a/drivers/infiniband/ulp/iser/iscsi_iser.h +++ b/drivers/infiniband/ulp/iser/iscsi_iser.h @@ -300,7 +300,11 @@ struct iser_tx_desc { int num_sge; bool mapped; u8 wr_idx; - struct ib_send_wr wrs[ISER_MAX_WRS]; + union iser_wr { + struct ib_send_wr send; + struct ib_fast_reg_wr fast_reg; + struct ib_sig_handover_wr sig; + } wrs[ISER_MAX_WRS]; struct iser_mem_reg data_reg; struct iser_mem_reg prot_reg; struct ib_sig_attrs sig_attrs; @@ -712,11 +716,11 @@ iser_reg_desc_put_fmr(struct ib_conn *ib_conn, static inline struct ib_send_wr * iser_tx_next_wr(struct iser_tx_desc *tx_desc) { - struct ib_send_wr *cur_wr = &tx_desc->wrs[tx_desc->wr_idx]; + struct ib_send_wr *cur_wr = &tx_desc->wrs[tx_desc->wr_idx].send; struct ib_send_wr *last_wr; if (tx_desc->wr_idx) { - last_wr = &tx_desc->wrs[tx_desc->wr_idx - 1]; + last_wr = &tx_desc->wrs[tx_desc->wr_idx - 1].send; last_wr->next = cur_wr; } tx_desc->wr_idx++; diff --git a/drivers/infiniband/ulp/iser/iser_memory.c b/drivers/infiniband/ulp/iser/iser_memory.c index 4c46d67d37a1..f45e6a352173 100644 --- a/drivers/infiniband/ulp/iser/iser_memory.c +++ b/drivers/infiniband/ulp/iser/iser_memory.c @@ -683,7 +683,7 @@ iser_reg_sig_mr(struct iscsi_iser_task *iser_task, { struct iser_tx_desc *tx_desc = &iser_task->desc; struct ib_sig_attrs *sig_attrs = &tx_desc->sig_attrs; - struct ib_send_wr *wr; + struct ib_sig_handover_wr *wr; int ret; memset(sig_attrs, 0, sizeof(*sig_attrs)); @@ -693,26 +693,24 @@ iser_reg_sig_mr(struct iscsi_iser_task *iser_task, iser_set_prot_checks(iser_task->sc, &sig_attrs->check_mask); - if (!pi_ctx->sig_mr_valid) { - wr = iser_tx_next_wr(tx_desc); - iser_inv_rkey(wr, pi_ctx->sig_mr); - } - - wr = iser_tx_next_wr(tx_desc); - wr->opcode = IB_WR_REG_SIG_MR; - wr->wr_id = ISER_FASTREG_LI_WRID; - wr->sg_list = &data_reg->sge; - wr->num_sge = 1; - wr->send_flags = 0; - wr->wr.sig_handover.sig_attrs = sig_attrs; - wr->wr.sig_handover.sig_mr = pi_ctx->sig_mr; + if (!pi_ctx->sig_mr_valid) + iser_inv_rkey(iser_tx_next_wr(tx_desc), pi_ctx->sig_mr); + + wr = sig_handover_wr(iser_tx_next_wr(tx_desc)); + wr->wr.opcode = IB_WR_REG_SIG_MR; + wr->wr.wr_id = ISER_FASTREG_LI_WRID; + wr->wr.sg_list = &data_reg->sge; + wr->wr.num_sge = 1; + wr->wr.send_flags = 0; + wr->sig_attrs = sig_attrs; + wr->sig_mr = pi_ctx->sig_mr; if (scsi_prot_sg_count(iser_task->sc)) - wr->wr.sig_handover.prot = &prot_reg->sge; + wr->prot = &prot_reg->sge; else - wr->wr.sig_handover.prot = NULL; - wr->wr.sig_handover.access_flags = IB_ACCESS_LOCAL_WRITE | - IB_ACCESS_REMOTE_READ | - IB_ACCESS_REMOTE_WRITE; + wr->prot = NULL; + wr->access_flags = IB_ACCESS_LOCAL_WRITE | + IB_ACCESS_REMOTE_READ | + IB_ACCESS_REMOTE_WRITE; pi_ctx->sig_mr_valid = 0; sig_reg->sge.lkey = pi_ctx->sig_mr->lkey; @@ -737,7 +735,7 @@ static int iser_fast_reg_mr(struct iscsi_iser_task *iser_task, struct ib_mr *mr = rsc->mr; struct ib_fast_reg_page_list *frpl = rsc->frpl; struct iser_tx_desc *tx_desc = &iser_task->desc; - struct ib_send_wr *wr; + struct ib_fast_reg_wr *wr; int offset, size, plen; plen = iser_sg_to_page_vec(mem, device->ib_device, frpl->page_list, @@ -747,24 +745,22 @@ static int iser_fast_reg_mr(struct iscsi_iser_task *iser_task, return -EINVAL; } - if (!rsc->mr_valid) { - wr = iser_tx_next_wr(tx_desc); - iser_inv_rkey(wr, mr); - } - - wr = iser_tx_next_wr(tx_desc); - wr->opcode = IB_WR_FAST_REG_MR; - wr->wr_id = ISER_FASTREG_LI_WRID; - wr->send_flags = 0; - wr->wr.fast_reg.iova_start = frpl->page_list[0] + offset; - wr->wr.fast_reg.page_list = frpl; - wr->wr.fast_reg.page_list_len = plen; - wr->wr.fast_reg.page_shift = SHIFT_4K; - wr->wr.fast_reg.length = size; - wr->wr.fast_reg.rkey = mr->rkey; - wr->wr.fast_reg.access_flags = (IB_ACCESS_LOCAL_WRITE | - IB_ACCESS_REMOTE_WRITE | - IB_ACCESS_REMOTE_READ); + if (!rsc->mr_valid) + iser_inv_rkey(iser_tx_next_wr(tx_desc), mr); + + wr = fast_reg_wr(iser_tx_next_wr(tx_desc)); + wr->wr.opcode = IB_WR_FAST_REG_MR; + wr->wr.wr_id = ISER_FASTREG_LI_WRID; + wr->wr.send_flags = 0; + wr->iova_start = frpl->page_list[0] + offset; + wr->page_list = frpl; + wr->page_list_len = plen; + wr->page_shift = SHIFT_4K; + wr->length = size; + wr->rkey = mr->rkey; + wr->access_flags = (IB_ACCESS_LOCAL_WRITE | + IB_ACCESS_REMOTE_WRITE | + IB_ACCESS_REMOTE_READ); rsc->mr_valid = 0; reg->sge.lkey = mr->lkey; diff --git a/drivers/infiniband/ulp/iser/iser_verbs.c b/drivers/infiniband/ulp/iser/iser_verbs.c index 85132d867bc8..b26022e30af1 100644 --- a/drivers/infiniband/ulp/iser/iser_verbs.c +++ b/drivers/infiniband/ulp/iser/iser_verbs.c @@ -1135,7 +1135,7 @@ int iser_post_send(struct ib_conn *ib_conn, struct iser_tx_desc *tx_desc, wr->opcode = IB_WR_SEND; wr->send_flags = signal ? IB_SEND_SIGNALED : 0; - ib_ret = ib_post_send(ib_conn->qp, &tx_desc->wrs[0], &bad_wr); + ib_ret = ib_post_send(ib_conn->qp, &tx_desc->wrs[0].send, &bad_wr); if (ib_ret) iser_err("ib_post_send failed, ret:%d opcode:%d\n", ib_ret, bad_wr->opcode); diff --git a/drivers/infiniband/ulp/isert/ib_isert.c b/drivers/infiniband/ulp/isert/ib_isert.c index 403bd29443b8..02c4c0b4569d 100644 --- a/drivers/infiniband/ulp/isert/ib_isert.c +++ b/drivers/infiniband/ulp/isert/ib_isert.c @@ -1703,10 +1703,10 @@ isert_unmap_cmd(struct isert_cmd *isert_cmd, struct isert_conn *isert_conn) isert_unmap_data_buf(isert_conn, &wr->data); } - if (wr->send_wr) { + if (wr->rdma_wr) { isert_dbg("Cmd %p free send_wr\n", isert_cmd); - kfree(wr->send_wr); - wr->send_wr = NULL; + kfree(wr->rdma_wr); + wr->rdma_wr = NULL; } if (wr->ib_sge) { @@ -1741,7 +1741,7 @@ isert_unreg_rdma(struct isert_cmd *isert_cmd, struct isert_conn *isert_conn) } wr->ib_sge = NULL; - wr->send_wr = NULL; + wr->rdma_wr = NULL; } static void @@ -1910,7 +1910,7 @@ isert_completion_rdma_write(struct iser_tx_desc *tx_desc, } device->unreg_rdma_mem(isert_cmd, isert_conn); - wr->send_wr_num = 0; + wr->rdma_wr_num = 0; if (ret) transport_send_check_condition_and_sense(se_cmd, se_cmd->pi_err, 0); @@ -1938,7 +1938,7 @@ isert_completion_rdma_read(struct iser_tx_desc *tx_desc, iscsit_stop_dataout_timer(cmd); device->unreg_rdma_mem(isert_cmd, isert_conn); cmd->write_data_done = wr->data.len; - wr->send_wr_num = 0; + wr->rdma_wr_num = 0; isert_dbg("Cmd: %p RDMA_READ comp calling execute_cmd\n", isert_cmd); spin_lock_bh(&cmd->istate_lock); @@ -2384,7 +2384,7 @@ isert_put_text_rsp(struct iscsi_cmd *cmd, struct iscsi_conn *conn) static int isert_build_rdma_wr(struct isert_conn *isert_conn, struct isert_cmd *isert_cmd, - struct ib_sge *ib_sge, struct ib_send_wr *send_wr, + struct ib_sge *ib_sge, struct ib_rdma_wr *rdma_wr, u32 data_left, u32 offset) { struct iscsi_cmd *cmd = isert_cmd->iscsi_cmd; @@ -2399,8 +2399,8 @@ isert_build_rdma_wr(struct isert_conn *isert_conn, struct isert_cmd *isert_cmd, sg_nents = min(cmd->se_cmd.t_data_nents - sg_off, isert_conn->max_sge); page_off = offset % PAGE_SIZE; - send_wr->sg_list = ib_sge; - send_wr->wr_id = (uintptr_t)&isert_cmd->tx_desc; + rdma_wr->wr.sg_list = ib_sge; + rdma_wr->wr.wr_id = (uintptr_t)&isert_cmd->tx_desc; /* * Perform mapping of TCM scatterlist memory ib_sge dma_addr. */ @@ -2425,11 +2425,11 @@ isert_build_rdma_wr(struct isert_conn *isert_conn, struct isert_cmd *isert_cmd, isert_dbg("Incrementing ib_sge pointer to %p\n", ib_sge); } - send_wr->num_sge = ++i; + rdma_wr->wr.num_sge = ++i; isert_dbg("Set outgoing sg_list: %p num_sg: %u from TCM SGLs\n", - send_wr->sg_list, send_wr->num_sge); + rdma_wr->wr.sg_list, rdma_wr->wr.num_sge); - return send_wr->num_sge; + return rdma_wr->wr.num_sge; } static int @@ -2440,7 +2440,7 @@ isert_map_rdma(struct iscsi_conn *conn, struct iscsi_cmd *cmd, struct isert_cmd *isert_cmd = iscsit_priv_cmd(cmd); struct isert_conn *isert_conn = conn->context; struct isert_data_buf *data = &wr->data; - struct ib_send_wr *send_wr; + struct ib_rdma_wr *rdma_wr; struct ib_sge *ib_sge; u32 offset, data_len, data_left, rdma_write_max, va_offset = 0; int ret = 0, i, ib_sge_cnt; @@ -2465,11 +2465,11 @@ isert_map_rdma(struct iscsi_conn *conn, struct iscsi_cmd *cmd, } wr->ib_sge = ib_sge; - wr->send_wr_num = DIV_ROUND_UP(data->nents, isert_conn->max_sge); - wr->send_wr = kzalloc(sizeof(struct ib_send_wr) * wr->send_wr_num, + wr->rdma_wr_num = DIV_ROUND_UP(data->nents, isert_conn->max_sge); + wr->rdma_wr = kzalloc(sizeof(struct ib_rdma_wr) * wr->rdma_wr_num, GFP_KERNEL); - if (!wr->send_wr) { - isert_dbg("Unable to allocate wr->send_wr\n"); + if (!wr->rdma_wr) { + isert_dbg("Unable to allocate wr->rdma_wr\n"); ret = -ENOMEM; goto unmap_cmd; } @@ -2477,31 +2477,31 @@ isert_map_rdma(struct iscsi_conn *conn, struct iscsi_cmd *cmd, wr->isert_cmd = isert_cmd; rdma_write_max = isert_conn->max_sge * PAGE_SIZE; - for (i = 0; i < wr->send_wr_num; i++) { - send_wr = &isert_cmd->rdma_wr.send_wr[i]; + for (i = 0; i < wr->rdma_wr_num; i++) { + rdma_wr = &isert_cmd->rdma_wr.rdma_wr[i]; data_len = min(data_left, rdma_write_max); - send_wr->send_flags = 0; + rdma_wr->wr.send_flags = 0; if (wr->iser_ib_op == ISER_IB_RDMA_WRITE) { - send_wr->opcode = IB_WR_RDMA_WRITE; - send_wr->wr.rdma.remote_addr = isert_cmd->read_va + offset; - send_wr->wr.rdma.rkey = isert_cmd->read_stag; - if (i + 1 == wr->send_wr_num) - send_wr->next = &isert_cmd->tx_desc.send_wr; + rdma_wr->wr.opcode = IB_WR_RDMA_WRITE; + rdma_wr->remote_addr = isert_cmd->read_va + offset; + rdma_wr->rkey = isert_cmd->read_stag; + if (i + 1 == wr->rdma_wr_num) + rdma_wr->wr.next = &isert_cmd->tx_desc.send_wr; else - send_wr->next = &wr->send_wr[i + 1]; + rdma_wr->wr.next = &wr->rdma_wr[i + 1].wr; } else { - send_wr->opcode = IB_WR_RDMA_READ; - send_wr->wr.rdma.remote_addr = isert_cmd->write_va + va_offset; - send_wr->wr.rdma.rkey = isert_cmd->write_stag; - if (i + 1 == wr->send_wr_num) - send_wr->send_flags = IB_SEND_SIGNALED; + rdma_wr->wr.opcode = IB_WR_RDMA_READ; + rdma_wr->remote_addr = isert_cmd->write_va + va_offset; + rdma_wr->rkey = isert_cmd->write_stag; + if (i + 1 == wr->rdma_wr_num) + rdma_wr->wr.send_flags = IB_SEND_SIGNALED; else - send_wr->next = &wr->send_wr[i + 1]; + rdma_wr->wr.next = &wr->rdma_wr[i + 1].wr; } ib_sge_cnt = isert_build_rdma_wr(isert_conn, isert_cmd, ib_sge, - send_wr, data_len, offset); + rdma_wr, data_len, offset); ib_sge += ib_sge_cnt; offset += data_len; @@ -2581,8 +2581,8 @@ isert_fast_reg_mr(struct isert_conn *isert_conn, struct ib_device *ib_dev = device->ib_device; struct ib_mr *mr; struct ib_fast_reg_page_list *frpl; - struct ib_send_wr fr_wr, inv_wr; - struct ib_send_wr *bad_wr, *wr = NULL; + struct ib_fast_reg_wr fr_wr; + struct ib_send_wr inv_wr, *bad_wr, *wr = NULL; int ret, pagelist_len; u32 page_off; @@ -2620,20 +2620,20 @@ isert_fast_reg_mr(struct isert_conn *isert_conn, /* Prepare FASTREG WR */ memset(&fr_wr, 0, sizeof(fr_wr)); - fr_wr.wr_id = ISER_FASTREG_LI_WRID; - fr_wr.opcode = IB_WR_FAST_REG_MR; - fr_wr.wr.fast_reg.iova_start = frpl->page_list[0] + page_off; - fr_wr.wr.fast_reg.page_list = frpl; - fr_wr.wr.fast_reg.page_list_len = pagelist_len; - fr_wr.wr.fast_reg.page_shift = PAGE_SHIFT; - fr_wr.wr.fast_reg.length = mem->len; - fr_wr.wr.fast_reg.rkey = mr->rkey; - fr_wr.wr.fast_reg.access_flags = IB_ACCESS_LOCAL_WRITE; + fr_wr.wr.wr_id = ISER_FASTREG_LI_WRID; + fr_wr.wr.opcode = IB_WR_FAST_REG_MR; + fr_wr.iova_start = frpl->page_list[0] + page_off; + fr_wr.page_list = frpl; + fr_wr.page_list_len = pagelist_len; + fr_wr.page_shift = PAGE_SHIFT; + fr_wr.length = mem->len; + fr_wr.rkey = mr->rkey; + fr_wr.access_flags = IB_ACCESS_LOCAL_WRITE; if (!wr) - wr = &fr_wr; + wr = &fr_wr.wr; else - wr->next = &fr_wr; + wr->next = &fr_wr.wr; ret = ib_post_send(isert_conn->qp, wr, &bad_wr); if (ret) { @@ -2714,8 +2714,8 @@ isert_reg_sig_mr(struct isert_conn *isert_conn, struct isert_rdma_wr *rdma_wr, struct fast_reg_descriptor *fr_desc) { - struct ib_send_wr sig_wr, inv_wr; - struct ib_send_wr *bad_wr, *wr = NULL; + struct ib_sig_handover_wr sig_wr; + struct ib_send_wr inv_wr, *bad_wr, *wr = NULL; struct pi_context *pi_ctx = fr_desc->pi_ctx; struct ib_sig_attrs sig_attrs; int ret; @@ -2733,20 +2733,20 @@ isert_reg_sig_mr(struct isert_conn *isert_conn, } memset(&sig_wr, 0, sizeof(sig_wr)); - sig_wr.opcode = IB_WR_REG_SIG_MR; - sig_wr.wr_id = ISER_FASTREG_LI_WRID; - sig_wr.sg_list = &rdma_wr->ib_sg[DATA]; - sig_wr.num_sge = 1; - sig_wr.wr.sig_handover.access_flags = IB_ACCESS_LOCAL_WRITE; - sig_wr.wr.sig_handover.sig_attrs = &sig_attrs; - sig_wr.wr.sig_handover.sig_mr = pi_ctx->sig_mr; + sig_wr.wr.opcode = IB_WR_REG_SIG_MR; + sig_wr.wr.wr_id = ISER_FASTREG_LI_WRID; + sig_wr.wr.sg_list = &rdma_wr->ib_sg[DATA]; + sig_wr.wr.num_sge = 1; + sig_wr.access_flags = IB_ACCESS_LOCAL_WRITE; + sig_wr.sig_attrs = &sig_attrs; + sig_wr.sig_mr = pi_ctx->sig_mr; if (se_cmd->t_prot_sg) - sig_wr.wr.sig_handover.prot = &rdma_wr->ib_sg[PROT]; + sig_wr.prot = &rdma_wr->ib_sg[PROT]; if (!wr) - wr = &sig_wr; + wr = &sig_wr.wr; else - wr->next = &sig_wr; + wr->next = &sig_wr.wr; ret = ib_post_send(isert_conn->qp, wr, &bad_wr); if (ret) { @@ -2840,7 +2840,7 @@ isert_reg_rdma(struct iscsi_conn *conn, struct iscsi_cmd *cmd, struct isert_cmd *isert_cmd = iscsit_priv_cmd(cmd); struct isert_conn *isert_conn = conn->context; struct fast_reg_descriptor *fr_desc = NULL; - struct ib_send_wr *send_wr; + struct ib_rdma_wr *rdma_wr; struct ib_sge *ib_sg; u32 offset; int ret = 0; @@ -2881,26 +2881,26 @@ isert_reg_rdma(struct iscsi_conn *conn, struct iscsi_cmd *cmd, memcpy(&wr->s_ib_sge, ib_sg, sizeof(*ib_sg)); wr->ib_sge = &wr->s_ib_sge; - wr->send_wr_num = 1; - memset(&wr->s_send_wr, 0, sizeof(*send_wr)); - wr->send_wr = &wr->s_send_wr; + wr->rdma_wr_num = 1; + memset(&wr->s_rdma_wr, 0, sizeof(wr->s_rdma_wr)); + wr->rdma_wr = &wr->s_rdma_wr; wr->isert_cmd = isert_cmd; - send_wr = &isert_cmd->rdma_wr.s_send_wr; - send_wr->sg_list = &wr->s_ib_sge; - send_wr->num_sge = 1; - send_wr->wr_id = (uintptr_t)&isert_cmd->tx_desc; + rdma_wr = &isert_cmd->rdma_wr.s_rdma_wr; + rdma_wr->wr.sg_list = &wr->s_ib_sge; + rdma_wr->wr.num_sge = 1; + rdma_wr->wr.wr_id = (uintptr_t)&isert_cmd->tx_desc; if (wr->iser_ib_op == ISER_IB_RDMA_WRITE) { - send_wr->opcode = IB_WR_RDMA_WRITE; - send_wr->wr.rdma.remote_addr = isert_cmd->read_va; - send_wr->wr.rdma.rkey = isert_cmd->read_stag; - send_wr->send_flags = !isert_prot_cmd(isert_conn, se_cmd) ? + rdma_wr->wr.opcode = IB_WR_RDMA_WRITE; + rdma_wr->remote_addr = isert_cmd->read_va; + rdma_wr->rkey = isert_cmd->read_stag; + rdma_wr->wr.send_flags = !isert_prot_cmd(isert_conn, se_cmd) ? 0 : IB_SEND_SIGNALED; } else { - send_wr->opcode = IB_WR_RDMA_READ; - send_wr->wr.rdma.remote_addr = isert_cmd->write_va; - send_wr->wr.rdma.rkey = isert_cmd->write_stag; - send_wr->send_flags = IB_SEND_SIGNALED; + rdma_wr->wr.opcode = IB_WR_RDMA_READ; + rdma_wr->remote_addr = isert_cmd->write_va; + rdma_wr->rkey = isert_cmd->write_stag; + rdma_wr->wr.send_flags = IB_SEND_SIGNALED; } return 0; @@ -2948,11 +2948,11 @@ isert_put_datain(struct iscsi_conn *conn, struct iscsi_cmd *cmd) isert_init_tx_hdrs(isert_conn, &isert_cmd->tx_desc); isert_init_send_wr(isert_conn, isert_cmd, &isert_cmd->tx_desc.send_wr); - isert_cmd->rdma_wr.s_send_wr.next = &isert_cmd->tx_desc.send_wr; - wr->send_wr_num += 1; + isert_cmd->rdma_wr.s_rdma_wr.wr.next = &isert_cmd->tx_desc.send_wr; + wr->rdma_wr_num += 1; } - rc = ib_post_send(isert_conn->qp, wr->send_wr, &wr_failed); + rc = ib_post_send(isert_conn->qp, &wr->rdma_wr->wr, &wr_failed); if (rc) isert_warn("ib_post_send() failed for IB_WR_RDMA_WRITE\n"); @@ -2986,7 +2986,7 @@ isert_get_dataout(struct iscsi_conn *conn, struct iscsi_cmd *cmd, bool recovery) return rc; } - rc = ib_post_send(isert_conn->qp, wr->send_wr, &wr_failed); + rc = ib_post_send(isert_conn->qp, &wr->rdma_wr->wr, &wr_failed); if (rc) isert_warn("ib_post_send() failed for IB_WR_RDMA_READ\n"); diff --git a/drivers/infiniband/ulp/isert/ib_isert.h b/drivers/infiniband/ulp/isert/ib_isert.h index 6a04ba3c0f72..0a4a7861cce9 100644 --- a/drivers/infiniband/ulp/isert/ib_isert.h +++ b/drivers/infiniband/ulp/isert/ib_isert.h @@ -118,9 +118,9 @@ struct isert_rdma_wr { enum iser_ib_op_code iser_ib_op; struct ib_sge *ib_sge; struct ib_sge s_ib_sge; - int send_wr_num; - struct ib_send_wr *send_wr; - struct ib_send_wr s_send_wr; + int rdma_wr_num; + struct ib_rdma_wr *rdma_wr; + struct ib_rdma_wr s_rdma_wr; struct ib_sge ib_sg[3]; struct isert_data_buf data; struct isert_data_buf prot; diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c index b481490ad257..1390f99ca76b 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.c +++ b/drivers/infiniband/ulp/srp/ib_srp.c @@ -1306,7 +1306,7 @@ static int srp_map_finish_fr(struct srp_map_state *state, struct srp_target_port *target = ch->target; struct srp_device *dev = target->srp_host->srp_dev; struct ib_send_wr *bad_wr; - struct ib_send_wr wr; + struct ib_fast_reg_wr wr; struct srp_fr_desc *desc; u32 rkey; @@ -1324,17 +1324,17 @@ static int srp_map_finish_fr(struct srp_map_state *state, sizeof(state->pages[0]) * state->npages); memset(&wr, 0, sizeof(wr)); - wr.opcode = IB_WR_FAST_REG_MR; - wr.wr_id = FAST_REG_WR_ID_MASK; - wr.wr.fast_reg.iova_start = state->base_dma_addr; - wr.wr.fast_reg.page_list = desc->frpl; - wr.wr.fast_reg.page_list_len = state->npages; - wr.wr.fast_reg.page_shift = ilog2(dev->mr_page_size); - wr.wr.fast_reg.length = state->dma_len; - wr.wr.fast_reg.access_flags = (IB_ACCESS_LOCAL_WRITE | - IB_ACCESS_REMOTE_READ | - IB_ACCESS_REMOTE_WRITE); - wr.wr.fast_reg.rkey = desc->mr->lkey; + wr.wr.opcode = IB_WR_FAST_REG_MR; + wr.wr.wr_id = FAST_REG_WR_ID_MASK; + wr.iova_start = state->base_dma_addr; + wr.page_list = desc->frpl; + wr.page_list_len = state->npages; + wr.page_shift = ilog2(dev->mr_page_size); + wr.length = state->dma_len; + wr.access_flags = (IB_ACCESS_LOCAL_WRITE | + IB_ACCESS_REMOTE_READ | + IB_ACCESS_REMOTE_WRITE); + wr.rkey = desc->mr->lkey; *state->fr.next++ = desc; state->nmdesc++; @@ -1342,7 +1342,7 @@ static int srp_map_finish_fr(struct srp_map_state *state, srp_map_desc(state, state->base_dma_addr, state->dma_len, desc->mr->rkey); - return ib_post_send(ch->qp, &wr, &bad_wr); + return ib_post_send(ch->qp, &wr.wr, &bad_wr); } static int srp_finish_mapping(struct srp_map_state *state, diff --git a/drivers/infiniband/ulp/srpt/ib_srpt.c b/drivers/infiniband/ulp/srpt/ib_srpt.c index f6fe0414139b..d65533e3a5eb 100644 --- a/drivers/infiniband/ulp/srpt/ib_srpt.c +++ b/drivers/infiniband/ulp/srpt/ib_srpt.c @@ -2822,7 +2822,7 @@ static int srpt_cm_handler(struct ib_cm_id *cm_id, struct ib_cm_event *event) static int srpt_perform_rdmas(struct srpt_rdma_ch *ch, struct srpt_send_ioctx *ioctx) { - struct ib_send_wr wr; + struct ib_rdma_wr wr; struct ib_send_wr *bad_wr; struct rdma_iu *riu; int i; @@ -2850,29 +2850,29 @@ static int srpt_perform_rdmas(struct srpt_rdma_ch *ch, for (i = 0; i < n_rdma; ++i, ++riu) { if (dir == DMA_FROM_DEVICE) { - wr.opcode = IB_WR_RDMA_WRITE; - wr.wr_id = encode_wr_id(i == n_rdma - 1 ? + wr.wr.opcode = IB_WR_RDMA_WRITE; + wr.wr.wr_id = encode_wr_id(i == n_rdma - 1 ? SRPT_RDMA_WRITE_LAST : SRPT_RDMA_MID, ioctx->ioctx.index); } else { - wr.opcode = IB_WR_RDMA_READ; - wr.wr_id = encode_wr_id(i == n_rdma - 1 ? + wr.wr.opcode = IB_WR_RDMA_READ; + wr.wr.wr_id = encode_wr_id(i == n_rdma - 1 ? SRPT_RDMA_READ_LAST : SRPT_RDMA_MID, ioctx->ioctx.index); } - wr.next = NULL; - wr.wr.rdma.remote_addr = riu->raddr; - wr.wr.rdma.rkey = riu->rkey; - wr.num_sge = riu->sge_cnt; - wr.sg_list = riu->sge; + wr.wr.next = NULL; + wr.remote_addr = riu->raddr; + wr.rkey = riu->rkey; + wr.wr.num_sge = riu->sge_cnt; + wr.wr.sg_list = riu->sge; /* only get completion event for the last rdma write */ if (i == (n_rdma - 1) && dir == DMA_TO_DEVICE) - wr.send_flags = IB_SEND_SIGNALED; + wr.wr.send_flags = IB_SEND_SIGNALED; - ret = ib_post_send(ch->qp, &wr, &bad_wr); + ret = ib_post_send(ch->qp, &wr.wr, &bad_wr); if (ret) break; } @@ -2881,11 +2881,11 @@ static int srpt_perform_rdmas(struct srpt_rdma_ch *ch, pr_err("%s[%d]: ib_post_send() returned %d for %d/%d\n", __func__, __LINE__, ret, i, n_rdma); if (ret && i > 0) { - wr.num_sge = 0; - wr.wr_id = encode_wr_id(SRPT_RDMA_ABORT, ioctx->ioctx.index); - wr.send_flags = IB_SEND_SIGNALED; + wr.wr.num_sge = 0; + wr.wr.wr_id = encode_wr_id(SRPT_RDMA_ABORT, ioctx->ioctx.index); + wr.wr.send_flags = IB_SEND_SIGNALED; while (ch->state == CH_LIVE && - ib_post_send(ch->qp, &wr, &bad_wr) != 0) { + ib_post_send(ch->qp, &wr.wr, &bad_wr) != 0) { pr_info("Trying to abort failed RDMA transfer [%d]\n", ioctx->ioctx.index); msleep(1000); diff --git a/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd.h b/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd.h index f4b6c33ac318..144449122778 100644 --- a/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd.h +++ b/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd.h @@ -525,7 +525,7 @@ typedef struct kib_tx /* transmit message */ __u64 tx_msgaddr; /* message buffer (I/O addr) */ DECLARE_PCI_UNMAP_ADDR(tx_msgunmap); /* for dma_unmap_single() */ int tx_nwrq; /* # send work items */ - struct ib_send_wr *tx_wrq; /* send work items... */ + struct ib_rdma_wr *tx_wrq; /* send work items... */ struct ib_sge *tx_sge; /* ...and their memory */ kib_rdma_desc_t *tx_rd; /* rdma descriptor */ int tx_nfrags; /* # entries in... */ diff --git a/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd_cb.c b/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd_cb.c index a23a6d956a4d..a34f1707c167 100644 --- a/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd_cb.c +++ b/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd_cb.c @@ -834,7 +834,7 @@ kiblnd_post_tx_locked(kib_conn_t *conn, kib_tx_t *tx, int credit) /* close_conn will launch failover */ rc = -ENETDOWN; } else { - rc = ib_post_send(conn->ibc_cmid->qp, tx->tx_wrq, &bad_wrq); + rc = ib_post_send(conn->ibc_cmid->qp, &tx->tx_wrq->wr, &bad_wrq); } conn->ibc_last_send = jiffies; @@ -1008,7 +1008,7 @@ kiblnd_init_tx_msg(lnet_ni_t *ni, kib_tx_t *tx, int type, int body_nob) { kib_hca_dev_t *hdev = tx->tx_pool->tpo_hdev; struct ib_sge *sge = &tx->tx_sge[tx->tx_nwrq]; - struct ib_send_wr *wrq = &tx->tx_wrq[tx->tx_nwrq]; + struct ib_rdma_wr *wrq = &tx->tx_wrq[tx->tx_nwrq]; int nob = offsetof(kib_msg_t, ibm_u) + body_nob; struct ib_mr *mr; @@ -1027,12 +1027,12 @@ kiblnd_init_tx_msg(lnet_ni_t *ni, kib_tx_t *tx, int type, int body_nob) memset(wrq, 0, sizeof(*wrq)); - wrq->next = NULL; - wrq->wr_id = kiblnd_ptr2wreqid(tx, IBLND_WID_TX); - wrq->sg_list = sge; - wrq->num_sge = 1; - wrq->opcode = IB_WR_SEND; - wrq->send_flags = IB_SEND_SIGNALED; + wrq->wr.next = NULL; + wrq->wr.wr_id = kiblnd_ptr2wreqid(tx, IBLND_WID_TX); + wrq->wr.sg_list = sge; + wrq->wr.num_sge = 1; + wrq->wr.opcode = IB_WR_SEND; + wrq->wr.send_flags = IB_SEND_SIGNALED; tx->tx_nwrq++; } @@ -1044,7 +1044,7 @@ kiblnd_init_rdma(kib_conn_t *conn, kib_tx_t *tx, int type, kib_msg_t *ibmsg = tx->tx_msg; kib_rdma_desc_t *srcrd = tx->tx_rd; struct ib_sge *sge = &tx->tx_sge[0]; - struct ib_send_wr *wrq = &tx->tx_wrq[0]; + struct ib_rdma_wr *wrq = &tx->tx_wrq[0], *next; int rc = resid; int srcidx; int dstidx; @@ -1090,16 +1090,17 @@ kiblnd_init_rdma(kib_conn_t *conn, kib_tx_t *tx, int type, sge->length = wrknob; wrq = &tx->tx_wrq[tx->tx_nwrq]; + next = wrq + 1; - wrq->next = wrq + 1; - wrq->wr_id = kiblnd_ptr2wreqid(tx, IBLND_WID_RDMA); - wrq->sg_list = sge; - wrq->num_sge = 1; - wrq->opcode = IB_WR_RDMA_WRITE; - wrq->send_flags = 0; + wrq->wr.next = &next->wr; + wrq->wr.wr_id = kiblnd_ptr2wreqid(tx, IBLND_WID_RDMA); + wrq->wr.sg_list = sge; + wrq->wr.num_sge = 1; + wrq->wr.opcode = IB_WR_RDMA_WRITE; + wrq->wr.send_flags = 0; - wrq->wr.rdma.remote_addr = kiblnd_rd_frag_addr(dstrd, dstidx); - wrq->wr.rdma.rkey = kiblnd_rd_frag_key(dstrd, dstidx); + wrq->remote_addr = kiblnd_rd_frag_addr(dstrd, dstidx); + wrq->rkey = kiblnd_rd_frag_key(dstrd, dstidx); srcidx = kiblnd_rd_consume_frag(srcrd, srcidx, wrknob); dstidx = kiblnd_rd_consume_frag(dstrd, dstidx, wrknob); diff --git a/drivers/staging/rdma/amso1100/c2_qp.c b/drivers/staging/rdma/amso1100/c2_qp.c index 86708dee58b1..4c43ca935cc7 100644 --- a/drivers/staging/rdma/amso1100/c2_qp.c +++ b/drivers/staging/rdma/amso1100/c2_qp.c @@ -860,9 +860,9 @@ int c2_post_send(struct ib_qp *ibqp, struct ib_send_wr *ib_wr, flags |= SQ_READ_FENCE; } wr.sqwr.rdma_write.remote_stag = - cpu_to_be32(ib_wr->wr.rdma.rkey); + cpu_to_be32(rdma_wr(ib_wr)->rkey); wr.sqwr.rdma_write.remote_to = - cpu_to_be64(ib_wr->wr.rdma.remote_addr); + cpu_to_be64(rdma_wr(ib_wr)->remote_addr); err = move_sgl((struct c2_data_addr *) & (wr.sqwr.rdma_write.data), ib_wr->sg_list, @@ -889,9 +889,9 @@ int c2_post_send(struct ib_qp *ibqp, struct ib_send_wr *ib_wr, wr.sqwr.rdma_read.local_to = cpu_to_be64(ib_wr->sg_list->addr); wr.sqwr.rdma_read.remote_stag = - cpu_to_be32(ib_wr->wr.rdma.rkey); + cpu_to_be32(rdma_wr(ib_wr)->rkey); wr.sqwr.rdma_read.remote_to = - cpu_to_be64(ib_wr->wr.rdma.remote_addr); + cpu_to_be64(rdma_wr(ib_wr)->remote_addr); wr.sqwr.rdma_read.length = cpu_to_be32(ib_wr->sg_list->length); break; diff --git a/drivers/staging/rdma/ehca/ehca_reqs.c b/drivers/staging/rdma/ehca/ehca_reqs.c index 47f94984353d..10e2074384f5 100644 --- a/drivers/staging/rdma/ehca/ehca_reqs.c +++ b/drivers/staging/rdma/ehca/ehca_reqs.c @@ -110,19 +110,19 @@ static inline int ehca_write_rwqe(struct ipz_queue *ipz_rqueue, /* need ib_mad struct */ #include -static void trace_send_wr_ud(const struct ib_send_wr *send_wr) +static void trace_ud_wr(const struct ib_ud_wr *ud_wr) { int idx; int j; - while (send_wr) { - struct ib_mad_hdr *mad_hdr = send_wr->wr.ud.mad_hdr; - struct ib_sge *sge = send_wr->sg_list; - ehca_gen_dbg("send_wr#%x wr_id=%lx num_sge=%x " - "send_flags=%x opcode=%x", idx, send_wr->wr_id, - send_wr->num_sge, send_wr->send_flags, - send_wr->opcode); + while (ud_wr) { + struct ib_mad_hdr *mad_hdr = ud_wrmad_hdr; + struct ib_sge *sge = ud_wr->wr.sg_list; + ehca_gen_dbg("ud_wr#%x wr_id=%lx num_sge=%x " + "send_flags=%x opcode=%x", idx, ud_wr->wr.wr_id, + ud_wr->wr.num_sge, ud_wr->wr.send_flags, + ud_wr->.wr.opcode); if (mad_hdr) { - ehca_gen_dbg("send_wr#%x mad_hdr base_version=%x " + ehca_gen_dbg("ud_wr#%x mad_hdr base_version=%x " "mgmt_class=%x class_version=%x method=%x " "status=%x class_specific=%x tid=%lx " "attr_id=%x resv=%x attr_mod=%x", @@ -134,33 +134,33 @@ static void trace_send_wr_ud(const struct ib_send_wr *send_wr) mad_hdr->resv, mad_hdr->attr_mod); } - for (j = 0; j < send_wr->num_sge; j++) { + for (j = 0; j < ud_wr->wr.num_sge; j++) { u8 *data = __va(sge->addr); - ehca_gen_dbg("send_wr#%x sge#%x addr=%p length=%x " + ehca_gen_dbg("ud_wr#%x sge#%x addr=%p length=%x " "lkey=%x", idx, j, data, sge->length, sge->lkey); /* assume length is n*16 */ - ehca_dmp(data, sge->length, "send_wr#%x sge#%x", + ehca_dmp(data, sge->length, "ud_wr#%x sge#%x", idx, j); sge++; } /* eof for j */ idx++; - send_wr = send_wr->next; - } /* eof while send_wr */ + ud_wr = ud_wr(ud_wr->wr.next); + } /* eof while ud_wr */ } #endif /* DEBUG_GSI_SEND_WR */ static inline int ehca_write_swqe(struct ehca_qp *qp, struct ehca_wqe *wqe_p, - const struct ib_send_wr *send_wr, + struct ib_send_wr *send_wr, u32 sq_map_idx, int hidden) { u32 idx; u64 dma_length; struct ehca_av *my_av; - u32 remote_qkey = send_wr->wr.ud.remote_qkey; + u32 remote_qkey; struct ehca_qmap_entry *qmap_entry = &qp->sq_map.map[sq_map_idx]; if (unlikely((send_wr->num_sge < 0) || @@ -223,20 +223,21 @@ static inline int ehca_write_swqe(struct ehca_qp *qp, /* no break is intential here */ case IB_QPT_UD: /* IB 1.2 spec C10-15 compliance */ - if (send_wr->wr.ud.remote_qkey & 0x80000000) + remote_qkey = ud_wr(send_wr)->remote_qkey; + if (remote_qkey & 0x80000000) remote_qkey = qp->qkey; - wqe_p->destination_qp_number = send_wr->wr.ud.remote_qpn << 8; + wqe_p->destination_qp_number = ud_wr(send_wr)->remote_qpn << 8; wqe_p->local_ee_context_qkey = remote_qkey; - if (unlikely(!send_wr->wr.ud.ah)) { - ehca_gen_err("wr.ud.ah is NULL. qp=%p", qp); + if (unlikely(!ud_wr(send_wr)->ah)) { + ehca_gen_err("ud_wr(send_wr) is NULL. qp=%p", qp); return -EINVAL; } - if (unlikely(send_wr->wr.ud.remote_qpn == 0)) { + if (unlikely(ud_wr(send_wr)->remote_qpn == 0)) { ehca_gen_err("dest QP# is 0. qp=%x", qp->real_qp_num); return -EINVAL; } - my_av = container_of(send_wr->wr.ud.ah, struct ehca_av, ib_ah); + my_av = container_of(ud_wr(send_wr)->ah, struct ehca_av, ib_ah); wqe_p->u.ud_av.ud_av = my_av->av; /* @@ -255,9 +256,9 @@ static inline int ehca_write_swqe(struct ehca_qp *qp, qp->qp_type == IB_QPT_GSI) wqe_p->u.ud_av.ud_av.pmtu = 1; if (qp->qp_type == IB_QPT_GSI) { - wqe_p->pkeyi = send_wr->wr.ud.pkey_index; + wqe_p->pkeyi = ud_wr(send_wr)->pkey_index; #ifdef DEBUG_GSI_SEND_WR - trace_send_wr_ud(send_wr); + trace_ud_wr(ud_wr(send_wr)); #endif /* DEBUG_GSI_SEND_WR */ } break; @@ -269,8 +270,8 @@ static inline int ehca_write_swqe(struct ehca_qp *qp, case IB_QPT_RC: /* TODO: atomic not implemented */ wqe_p->u.nud.remote_virtual_address = - send_wr->wr.rdma.remote_addr; - wqe_p->u.nud.rkey = send_wr->wr.rdma.rkey; + rdma_wr(send_wr)->remote_addr; + wqe_p->u.nud.rkey = rdma_wr(send_wr)->rkey; /* * omitted checking of IB_SEND_INLINE diff --git a/drivers/staging/rdma/hfi1/keys.c b/drivers/staging/rdma/hfi1/keys.c index f6eff177ace1..82c21b1c0263 100644 --- a/drivers/staging/rdma/hfi1/keys.c +++ b/drivers/staging/rdma/hfi1/keys.c @@ -358,12 +358,12 @@ bail: /* * Initialize the memory region specified by the work request. */ -int hfi1_fast_reg_mr(struct hfi1_qp *qp, struct ib_send_wr *wr) +int hfi1_fast_reg_mr(struct hfi1_qp *qp, struct ib_fast_reg_wr *wr) { struct hfi1_lkey_table *rkt = &to_idev(qp->ibqp.device)->lk_table; struct hfi1_pd *pd = to_ipd(qp->ibqp.pd); struct hfi1_mregion *mr; - u32 rkey = wr->wr.fast_reg.rkey; + u32 rkey = wr->rkey; unsigned i, n, m; int ret = -EINVAL; unsigned long flags; @@ -380,22 +380,22 @@ int hfi1_fast_reg_mr(struct hfi1_qp *qp, struct ib_send_wr *wr) if (unlikely(mr == NULL || qp->ibqp.pd != mr->pd)) goto bail; - if (wr->wr.fast_reg.page_list_len > mr->max_segs) + if (wr->page_list_len > mr->max_segs) goto bail; - ps = 1UL << wr->wr.fast_reg.page_shift; - if (wr->wr.fast_reg.length > ps * wr->wr.fast_reg.page_list_len) + ps = 1UL << wr->page_shift; + if (wr->length > ps * wr->page_list_len) goto bail; - mr->user_base = wr->wr.fast_reg.iova_start; - mr->iova = wr->wr.fast_reg.iova_start; + mr->user_base = wr->iova_start; + mr->iova = wr->iova_start; mr->lkey = rkey; - mr->length = wr->wr.fast_reg.length; - mr->access_flags = wr->wr.fast_reg.access_flags; - page_list = wr->wr.fast_reg.page_list->page_list; + mr->length = wr->length; + mr->access_flags = wr->access_flags; + page_list = wr->page_list->page_list; m = 0; n = 0; - for (i = 0; i < wr->wr.fast_reg.page_list_len; i++) { + for (i = 0; i < wr->page_list_len; i++) { mr->map[m]->segs[n].vaddr = (void *) page_list[i]; mr->map[m]->segs[n].length = ps; if (++n == HFI1_SEGSZ) { diff --git a/drivers/staging/rdma/hfi1/qp.c b/drivers/staging/rdma/hfi1/qp.c index df1fa56eaf85..f8c36166962f 100644 --- a/drivers/staging/rdma/hfi1/qp.c +++ b/drivers/staging/rdma/hfi1/qp.c @@ -422,7 +422,7 @@ static void clear_mr_refs(struct hfi1_qp *qp, int clr_sends) if (qp->ibqp.qp_type == IB_QPT_UD || qp->ibqp.qp_type == IB_QPT_SMI || qp->ibqp.qp_type == IB_QPT_GSI) - atomic_dec(&to_iah(wqe->wr.wr.ud.ah)->refcount); + atomic_dec(&to_iah(wqe->ud_wr.ah)->refcount); if (++qp->s_last >= qp->s_size) qp->s_last = 0; } diff --git a/drivers/staging/rdma/hfi1/rc.c b/drivers/staging/rdma/hfi1/rc.c index 632dd5ba7dfd..fd0ac608c62d 100644 --- a/drivers/staging/rdma/hfi1/rc.c +++ b/drivers/staging/rdma/hfi1/rc.c @@ -404,9 +404,9 @@ int hfi1_make_rc_req(struct hfi1_qp *qp) goto bail; } ohdr->u.rc.reth.vaddr = - cpu_to_be64(wqe->wr.wr.rdma.remote_addr); + cpu_to_be64(wqe->rdma_wr.remote_addr); ohdr->u.rc.reth.rkey = - cpu_to_be32(wqe->wr.wr.rdma.rkey); + cpu_to_be32(wqe->rdma_wr.rkey); ohdr->u.rc.reth.length = cpu_to_be32(len); hwords += sizeof(struct ib_reth) / sizeof(u32); wqe->lpsn = wqe->psn; @@ -455,9 +455,9 @@ int hfi1_make_rc_req(struct hfi1_qp *qp) wqe->lpsn = qp->s_next_psn++; } ohdr->u.rc.reth.vaddr = - cpu_to_be64(wqe->wr.wr.rdma.remote_addr); + cpu_to_be64(wqe->rdma_wr.remote_addr); ohdr->u.rc.reth.rkey = - cpu_to_be32(wqe->wr.wr.rdma.rkey); + cpu_to_be32(wqe->rdma_wr.rkey); ohdr->u.rc.reth.length = cpu_to_be32(len); qp->s_state = OP(RDMA_READ_REQUEST); hwords += sizeof(ohdr->u.rc.reth) / sizeof(u32); @@ -488,21 +488,21 @@ int hfi1_make_rc_req(struct hfi1_qp *qp) if (wqe->wr.opcode == IB_WR_ATOMIC_CMP_AND_SWP) { qp->s_state = OP(COMPARE_SWAP); ohdr->u.atomic_eth.swap_data = cpu_to_be64( - wqe->wr.wr.atomic.swap); + wqe->atomic_wr.swap); ohdr->u.atomic_eth.compare_data = cpu_to_be64( - wqe->wr.wr.atomic.compare_add); + wqe->atomic_wr.compare_add); } else { qp->s_state = OP(FETCH_ADD); ohdr->u.atomic_eth.swap_data = cpu_to_be64( - wqe->wr.wr.atomic.compare_add); + wqe->atomic_wr.compare_add); ohdr->u.atomic_eth.compare_data = 0; } ohdr->u.atomic_eth.vaddr[0] = cpu_to_be32( - wqe->wr.wr.atomic.remote_addr >> 32); + wqe->atomic_wr.remote_addr >> 32); ohdr->u.atomic_eth.vaddr[1] = cpu_to_be32( - wqe->wr.wr.atomic.remote_addr); + wqe->atomic_wr.remote_addr); ohdr->u.atomic_eth.rkey = cpu_to_be32( - wqe->wr.wr.atomic.rkey); + wqe->atomic_wr.rkey); hwords += sizeof(struct ib_atomic_eth) / sizeof(u32); ss = NULL; len = 0; @@ -629,9 +629,9 @@ int hfi1_make_rc_req(struct hfi1_qp *qp) */ len = (delta_psn(qp->s_psn, wqe->psn)) * pmtu; ohdr->u.rc.reth.vaddr = - cpu_to_be64(wqe->wr.wr.rdma.remote_addr + len); + cpu_to_be64(wqe->rdma_wr.remote_addr + len); ohdr->u.rc.reth.rkey = - cpu_to_be32(wqe->wr.wr.rdma.rkey); + cpu_to_be32(wqe->rdma_wr.rkey); ohdr->u.rc.reth.length = cpu_to_be32(wqe->length - len); qp->s_state = OP(RDMA_READ_REQUEST); hwords += sizeof(ohdr->u.rc.reth) / sizeof(u32); diff --git a/drivers/staging/rdma/hfi1/ruc.c b/drivers/staging/rdma/hfi1/ruc.c index a4115288db66..d614474770b3 100644 --- a/drivers/staging/rdma/hfi1/ruc.c +++ b/drivers/staging/rdma/hfi1/ruc.c @@ -481,8 +481,8 @@ again: if (wqe->length == 0) break; if (unlikely(!hfi1_rkey_ok(qp, &qp->r_sge.sge, wqe->length, - wqe->wr.wr.rdma.remote_addr, - wqe->wr.wr.rdma.rkey, + wqe->rdma_wr.remote_addr, + wqe->rdma_wr.rkey, IB_ACCESS_REMOTE_WRITE))) goto acc_err; qp->r_sge.sg_list = NULL; @@ -494,8 +494,8 @@ again: if (unlikely(!(qp->qp_access_flags & IB_ACCESS_REMOTE_READ))) goto inv_err; if (unlikely(!hfi1_rkey_ok(qp, &sqp->s_sge.sge, wqe->length, - wqe->wr.wr.rdma.remote_addr, - wqe->wr.wr.rdma.rkey, + wqe->rdma_wr.remote_addr, + wqe->rdma_wr.rkey, IB_ACCESS_REMOTE_READ))) goto acc_err; release = 0; @@ -512,18 +512,18 @@ again: if (unlikely(!(qp->qp_access_flags & IB_ACCESS_REMOTE_ATOMIC))) goto inv_err; if (unlikely(!hfi1_rkey_ok(qp, &qp->r_sge.sge, sizeof(u64), - wqe->wr.wr.atomic.remote_addr, - wqe->wr.wr.atomic.rkey, + wqe->atomic_wr.remote_addr, + wqe->atomic_wr.rkey, IB_ACCESS_REMOTE_ATOMIC))) goto acc_err; /* Perform atomic OP and save result. */ maddr = (atomic64_t *) qp->r_sge.sge.vaddr; - sdata = wqe->wr.wr.atomic.compare_add; + sdata = wqe->atomic_wr.compare_add; *(u64 *) sqp->s_sge.sge.vaddr = (wqe->wr.opcode == IB_WR_ATOMIC_FETCH_AND_ADD) ? (u64) atomic64_add_return(sdata, maddr) - sdata : (u64) cmpxchg((u64 *) qp->r_sge.sge.vaddr, - sdata, wqe->wr.wr.atomic.swap); + sdata, wqe->atomic_wr.swap); hfi1_put_mr(qp->r_sge.sge.mr); qp->r_sge.num_sge = 0; goto send_comp; @@ -913,7 +913,7 @@ void hfi1_send_complete(struct hfi1_qp *qp, struct hfi1_swqe *wqe, if (qp->ibqp.qp_type == IB_QPT_UD || qp->ibqp.qp_type == IB_QPT_SMI || qp->ibqp.qp_type == IB_QPT_GSI) - atomic_dec(&to_iah(wqe->wr.wr.ud.ah)->refcount); + atomic_dec(&to_iah(wqe->ud_wr.ah)->refcount); /* See ch. 11.2.4.1 and 10.7.3.1 */ if (!(qp->s_flags & HFI1_S_SIGNAL_REQ_WR) || diff --git a/drivers/staging/rdma/hfi1/uc.c b/drivers/staging/rdma/hfi1/uc.c index b536f397737c..6095039c4485 100644 --- a/drivers/staging/rdma/hfi1/uc.c +++ b/drivers/staging/rdma/hfi1/uc.c @@ -147,9 +147,9 @@ int hfi1_make_uc_req(struct hfi1_qp *qp) case IB_WR_RDMA_WRITE: case IB_WR_RDMA_WRITE_WITH_IMM: ohdr->u.rc.reth.vaddr = - cpu_to_be64(wqe->wr.wr.rdma.remote_addr); + cpu_to_be64(wqe->rdma_wr.remote_addr); ohdr->u.rc.reth.rkey = - cpu_to_be32(wqe->wr.wr.rdma.rkey); + cpu_to_be32(wqe->rdma_wr.rkey); ohdr->u.rc.reth.length = cpu_to_be32(len); hwords += sizeof(struct ib_reth) / 4; if (len > pmtu) { diff --git a/drivers/staging/rdma/hfi1/ud.c b/drivers/staging/rdma/hfi1/ud.c index d40d1a1e10aa..5a9c784bec04 100644 --- a/drivers/staging/rdma/hfi1/ud.c +++ b/drivers/staging/rdma/hfi1/ud.c @@ -80,7 +80,7 @@ static void ud_loopback(struct hfi1_qp *sqp, struct hfi1_swqe *swqe) rcu_read_lock(); - qp = hfi1_lookup_qpn(ibp, swqe->wr.wr.ud.remote_qpn); + qp = hfi1_lookup_qpn(ibp, swqe->ud_wr.remote_qpn); if (!qp) { ibp->n_pkt_drops++; rcu_read_unlock(); @@ -98,7 +98,7 @@ static void ud_loopback(struct hfi1_qp *sqp, struct hfi1_swqe *swqe) goto drop; } - ah_attr = &to_iah(swqe->wr.wr.ud.ah)->attr; + ah_attr = &to_iah(swqe->ud_wr.ah)->attr; ppd = ppd_from_ibp(ibp); if (qp->ibqp.qp_num > 1) { @@ -128,8 +128,8 @@ static void ud_loopback(struct hfi1_qp *sqp, struct hfi1_swqe *swqe) if (qp->ibqp.qp_num) { u32 qkey; - qkey = (int)swqe->wr.wr.ud.remote_qkey < 0 ? - sqp->qkey : swqe->wr.wr.ud.remote_qkey; + qkey = (int)swqe->ud_wr.remote_qkey < 0 ? + sqp->qkey : swqe->ud_wr.remote_qkey; if (unlikely(qkey != qp->qkey)) { u16 lid; @@ -234,7 +234,7 @@ static void ud_loopback(struct hfi1_qp *sqp, struct hfi1_swqe *swqe) if (qp->ibqp.qp_type == IB_QPT_GSI || qp->ibqp.qp_type == IB_QPT_SMI) { if (sqp->ibqp.qp_type == IB_QPT_GSI || sqp->ibqp.qp_type == IB_QPT_SMI) - wc.pkey_index = swqe->wr.wr.ud.pkey_index; + wc.pkey_index = swqe->ud_wr.pkey_index; else wc.pkey_index = sqp->s_pkey_index; } else { @@ -309,7 +309,7 @@ int hfi1_make_ud_req(struct hfi1_qp *qp) /* Construct the header. */ ibp = to_iport(qp->ibqp.device, qp->port_num); ppd = ppd_from_ibp(ibp); - ah_attr = &to_iah(wqe->wr.wr.ud.ah)->attr; + ah_attr = &to_iah(wqe->ud_wr.ah)->attr; if (ah_attr->dlid < HFI1_MULTICAST_LID_BASE || ah_attr->dlid == HFI1_PERMISSIVE_LID) { lid = ah_attr->dlid & ~((1 << ppd->lmc) - 1); @@ -401,18 +401,18 @@ int hfi1_make_ud_req(struct hfi1_qp *qp) bth0 |= IB_BTH_SOLICITED; bth0 |= extra_bytes << 20; if (qp->ibqp.qp_type == IB_QPT_GSI || qp->ibqp.qp_type == IB_QPT_SMI) - bth0 |= hfi1_get_pkey(ibp, wqe->wr.wr.ud.pkey_index); + bth0 |= hfi1_get_pkey(ibp, wqe->ud_wr.pkey_index); else bth0 |= hfi1_get_pkey(ibp, qp->s_pkey_index); ohdr->bth[0] = cpu_to_be32(bth0); - ohdr->bth[1] = cpu_to_be32(wqe->wr.wr.ud.remote_qpn); + ohdr->bth[1] = cpu_to_be32(wqe->ud_wr.remote_qpn); ohdr->bth[2] = cpu_to_be32(mask_psn(qp->s_next_psn++)); /* * Qkeys with the high order bit set mean use the * qkey from the QP context instead of the WR (see 10.2.5). */ - ohdr->u.ud.deth[0] = cpu_to_be32((int)wqe->wr.wr.ud.remote_qkey < 0 ? - qp->qkey : wqe->wr.wr.ud.remote_qkey); + ohdr->u.ud.deth[0] = cpu_to_be32((int)wqe->ud_wr.remote_qkey < 0 ? + qp->qkey : wqe->ud_wr.remote_qkey); ohdr->u.ud.deth[1] = cpu_to_be32(qp->ibqp.qp_num); /* disarm any ahg */ qp->s_hdr->ahgcount = 0; diff --git a/drivers/staging/rdma/hfi1/verbs.c b/drivers/staging/rdma/hfi1/verbs.c index 41bb59eb001c..981e6c1b79a3 100644 --- a/drivers/staging/rdma/hfi1/verbs.c +++ b/drivers/staging/rdma/hfi1/verbs.c @@ -391,7 +391,7 @@ static int post_one_send(struct hfi1_qp *qp, struct ib_send_wr *wr) wr->opcode != IB_WR_SEND_WITH_IMM) return -EINVAL; /* Check UD destination address PD */ - if (qp->ibqp.pd != wr->wr.ud.ah->pd) + if (qp->ibqp.pd != ud_wr(wr)->ah->pd) return -EINVAL; } else if ((unsigned) wr->opcode > IB_WR_ATOMIC_FETCH_AND_ADD) return -EINVAL; @@ -412,7 +412,24 @@ static int post_one_send(struct hfi1_qp *qp, struct ib_send_wr *wr) rkt = &to_idev(qp->ibqp.device)->lk_table; pd = to_ipd(qp->ibqp.pd); wqe = get_swqe_ptr(qp, qp->s_head); - wqe->wr = *wr; + + + if (qp->ibqp.qp_type != IB_QPT_UC && + qp->ibqp.qp_type != IB_QPT_RC) + memcpy(&wqe->ud_wr, ud_wr(wr), sizeof(wqe->ud_wr)); + else if (wr->opcode == IB_WR_FAST_REG_MR) + memcpy(&wqe->fast_reg_wr, fast_reg_wr(wr), + sizeof(wqe->fast_reg_wr)); + else if (wr->opcode == IB_WR_RDMA_WRITE_WITH_IMM || + wr->opcode == IB_WR_RDMA_WRITE || + wr->opcode == IB_WR_RDMA_READ) + memcpy(&wqe->rdma_wr, rdma_wr(wr), sizeof(wqe->rdma_wr)); + else if (wr->opcode == IB_WR_ATOMIC_CMP_AND_SWP || + wr->opcode == IB_WR_ATOMIC_FETCH_AND_ADD) + memcpy(&wqe->atomic_wr, atomic_wr(wr), sizeof(wqe->atomic_wr)); + else + memcpy(&wqe->wr, wr, sizeof(wqe->wr)); + wqe->length = 0; j = 0; if (wr->num_sge) { @@ -438,7 +455,7 @@ static int post_one_send(struct hfi1_qp *qp, struct ib_send_wr *wr) if (wqe->length > 0x80000000U) goto bail_inval_free; } else { - struct hfi1_ah *ah = to_iah(wr->wr.ud.ah); + struct hfi1_ah *ah = to_iah(ud_wr(wr)->ah); atomic_inc(&ah->refcount); } diff --git a/drivers/staging/rdma/hfi1/verbs.h b/drivers/staging/rdma/hfi1/verbs.h index ed903a93baf7..cf5a3c956284 100644 --- a/drivers/staging/rdma/hfi1/verbs.h +++ b/drivers/staging/rdma/hfi1/verbs.h @@ -348,7 +348,13 @@ struct hfi1_mr { * in qp->s_max_sge. */ struct hfi1_swqe { - struct ib_send_wr wr; /* don't use wr.sg_list */ + union { + struct ib_send_wr wr; /* don't use wr.sg_list */ + struct ib_rdma_wr rdma_wr; + struct ib_atomic_wr atomic_wr; + struct ib_ud_wr ud_wr; + struct ib_fast_reg_wr fast_reg_wr; + }; u32 psn; /* first packet sequence number */ u32 lpsn; /* last packet sequence number */ u32 ssn; /* send sequence number */ @@ -1025,7 +1031,7 @@ struct ib_fast_reg_page_list *hfi1_alloc_fast_reg_page_list( void hfi1_free_fast_reg_page_list(struct ib_fast_reg_page_list *pl); -int hfi1_fast_reg_mr(struct hfi1_qp *qp, struct ib_send_wr *wr); +int hfi1_fast_reg_mr(struct hfi1_qp *qp, struct ib_fast_reg_wr *wr); struct ib_fmr *hfi1_alloc_fmr(struct ib_pd *pd, int mr_access_flags, struct ib_fmr_attr *fmr_attr); diff --git a/drivers/staging/rdma/ipath/ipath_rc.c b/drivers/staging/rdma/ipath/ipath_rc.c index 79b3dbc97179..d4aa53574e57 100644 --- a/drivers/staging/rdma/ipath/ipath_rc.c +++ b/drivers/staging/rdma/ipath/ipath_rc.c @@ -350,9 +350,9 @@ int ipath_make_rc_req(struct ipath_qp *qp) goto bail; } ohdr->u.rc.reth.vaddr = - cpu_to_be64(wqe->wr.wr.rdma.remote_addr); + cpu_to_be64(wqe->rdma_wr.remote_addr); ohdr->u.rc.reth.rkey = - cpu_to_be32(wqe->wr.wr.rdma.rkey); + cpu_to_be32(wqe->rdma_wr.rkey); ohdr->u.rc.reth.length = cpu_to_be32(len); hwords += sizeof(struct ib_reth) / sizeof(u32); wqe->lpsn = wqe->psn; @@ -401,9 +401,9 @@ int ipath_make_rc_req(struct ipath_qp *qp) wqe->lpsn = qp->s_next_psn++; } ohdr->u.rc.reth.vaddr = - cpu_to_be64(wqe->wr.wr.rdma.remote_addr); + cpu_to_be64(wqe->rdma_wr.remote_addr); ohdr->u.rc.reth.rkey = - cpu_to_be32(wqe->wr.wr.rdma.rkey); + cpu_to_be32(wqe->rdma_wr.rkey); ohdr->u.rc.reth.length = cpu_to_be32(len); qp->s_state = OP(RDMA_READ_REQUEST); hwords += sizeof(ohdr->u.rc.reth) / sizeof(u32); @@ -433,21 +433,21 @@ int ipath_make_rc_req(struct ipath_qp *qp) if (wqe->wr.opcode == IB_WR_ATOMIC_CMP_AND_SWP) { qp->s_state = OP(COMPARE_SWAP); ohdr->u.atomic_eth.swap_data = cpu_to_be64( - wqe->wr.wr.atomic.swap); + wqe->atomic_wr.swap); ohdr->u.atomic_eth.compare_data = cpu_to_be64( - wqe->wr.wr.atomic.compare_add); + wqe->atomic_wr.compare_add); } else { qp->s_state = OP(FETCH_ADD); ohdr->u.atomic_eth.swap_data = cpu_to_be64( - wqe->wr.wr.atomic.compare_add); + wqe->atomic_wr.compare_add); ohdr->u.atomic_eth.compare_data = 0; } ohdr->u.atomic_eth.vaddr[0] = cpu_to_be32( - wqe->wr.wr.atomic.remote_addr >> 32); + wqe->atomic_wr.remote_addr >> 32); ohdr->u.atomic_eth.vaddr[1] = cpu_to_be32( - wqe->wr.wr.atomic.remote_addr); + wqe->atomic_wr.remote_addr); ohdr->u.atomic_eth.rkey = cpu_to_be32( - wqe->wr.wr.atomic.rkey); + wqe->atomic_wr.rkey); hwords += sizeof(struct ib_atomic_eth) / sizeof(u32); ss = NULL; len = 0; @@ -567,9 +567,9 @@ int ipath_make_rc_req(struct ipath_qp *qp) ipath_init_restart(qp, wqe); len = ((qp->s_psn - wqe->psn) & IPATH_PSN_MASK) * pmtu; ohdr->u.rc.reth.vaddr = - cpu_to_be64(wqe->wr.wr.rdma.remote_addr + len); + cpu_to_be64(wqe->rdma_wr.remote_addr + len); ohdr->u.rc.reth.rkey = - cpu_to_be32(wqe->wr.wr.rdma.rkey); + cpu_to_be32(wqe->rdma_wr.rkey); ohdr->u.rc.reth.length = cpu_to_be32(qp->s_len); qp->s_state = OP(RDMA_READ_REQUEST); hwords += sizeof(ohdr->u.rc.reth) / sizeof(u32); diff --git a/drivers/staging/rdma/ipath/ipath_ruc.c b/drivers/staging/rdma/ipath/ipath_ruc.c index 1f95bbaf7602..46af8b03d3d4 100644 --- a/drivers/staging/rdma/ipath/ipath_ruc.c +++ b/drivers/staging/rdma/ipath/ipath_ruc.c @@ -353,8 +353,8 @@ again: if (wqe->length == 0) break; if (unlikely(!ipath_rkey_ok(qp, &qp->r_sge, wqe->length, - wqe->wr.wr.rdma.remote_addr, - wqe->wr.wr.rdma.rkey, + wqe->rdma_wr.remote_addr, + wqe->rdma_wr.rkey, IB_ACCESS_REMOTE_WRITE))) goto acc_err; break; @@ -363,8 +363,8 @@ again: if (unlikely(!(qp->qp_access_flags & IB_ACCESS_REMOTE_READ))) goto inv_err; if (unlikely(!ipath_rkey_ok(qp, &sqp->s_sge, wqe->length, - wqe->wr.wr.rdma.remote_addr, - wqe->wr.wr.rdma.rkey, + wqe->rdma_wr.remote_addr, + wqe->rdma_wr.rkey, IB_ACCESS_REMOTE_READ))) goto acc_err; qp->r_sge.sge = wqe->sg_list[0]; @@ -377,18 +377,18 @@ again: if (unlikely(!(qp->qp_access_flags & IB_ACCESS_REMOTE_ATOMIC))) goto inv_err; if (unlikely(!ipath_rkey_ok(qp, &qp->r_sge, sizeof(u64), - wqe->wr.wr.atomic.remote_addr, - wqe->wr.wr.atomic.rkey, + wqe->atomic_wr.remote_addr, + wqe->atomic_wr.rkey, IB_ACCESS_REMOTE_ATOMIC))) goto acc_err; /* Perform atomic OP and save result. */ maddr = (atomic64_t *) qp->r_sge.sge.vaddr; - sdata = wqe->wr.wr.atomic.compare_add; + sdata = wqe->atomic_wr.compare_add; *(u64 *) sqp->s_sge.sge.vaddr = (wqe->wr.opcode == IB_WR_ATOMIC_FETCH_AND_ADD) ? (u64) atomic64_add_return(sdata, maddr) - sdata : (u64) cmpxchg((u64 *) qp->r_sge.sge.vaddr, - sdata, wqe->wr.wr.atomic.swap); + sdata, wqe->atomic_wr.swap); goto send_comp; default: diff --git a/drivers/staging/rdma/ipath/ipath_uc.c b/drivers/staging/rdma/ipath/ipath_uc.c index 22e60998f1a7..0246b30280b9 100644 --- a/drivers/staging/rdma/ipath/ipath_uc.c +++ b/drivers/staging/rdma/ipath/ipath_uc.c @@ -126,9 +126,9 @@ int ipath_make_uc_req(struct ipath_qp *qp) case IB_WR_RDMA_WRITE: case IB_WR_RDMA_WRITE_WITH_IMM: ohdr->u.rc.reth.vaddr = - cpu_to_be64(wqe->wr.wr.rdma.remote_addr); + cpu_to_be64(wqe->rdma_wr.remote_addr); ohdr->u.rc.reth.rkey = - cpu_to_be32(wqe->wr.wr.rdma.rkey); + cpu_to_be32(wqe->rdma_wr.rkey); ohdr->u.rc.reth.length = cpu_to_be32(len); hwords += sizeof(struct ib_reth) / 4; if (len > pmtu) { diff --git a/drivers/staging/rdma/ipath/ipath_ud.c b/drivers/staging/rdma/ipath/ipath_ud.c index e8a2a915251e..3ffc1565d03d 100644 --- a/drivers/staging/rdma/ipath/ipath_ud.c +++ b/drivers/staging/rdma/ipath/ipath_ud.c @@ -65,7 +65,7 @@ static void ipath_ud_loopback(struct ipath_qp *sqp, struct ipath_swqe *swqe) u32 rlen; u32 length; - qp = ipath_lookup_qpn(&dev->qp_table, swqe->wr.wr.ud.remote_qpn); + qp = ipath_lookup_qpn(&dev->qp_table, swqe->ud_wr.remote_qpn); if (!qp || !(ib_ipath_state_ops[qp->state] & IPATH_PROCESS_RECV_OK)) { dev->n_pkt_drops++; goto done; @@ -77,8 +77,8 @@ static void ipath_ud_loopback(struct ipath_qp *sqp, struct ipath_swqe *swqe) * qkey from the QP context instead of the WR (see 10.2.5). */ if (unlikely(qp->ibqp.qp_num && - ((int) swqe->wr.wr.ud.remote_qkey < 0 ? - sqp->qkey : swqe->wr.wr.ud.remote_qkey) != qp->qkey)) { + ((int) swqe->ud_wr.remote_qkey < 0 ? + sqp->qkey : swqe->ud_wr.remote_qkey) != qp->qkey)) { /* XXX OK to lose a count once in a while. */ dev->qkey_violations++; dev->n_pkt_drops++; @@ -175,7 +175,7 @@ static void ipath_ud_loopback(struct ipath_qp *sqp, struct ipath_swqe *swqe) } else spin_unlock_irqrestore(&rq->lock, flags); - ah_attr = &to_iah(swqe->wr.wr.ud.ah)->attr; + ah_attr = &to_iah(swqe->ud_wr.ah)->attr; if (ah_attr->ah_flags & IB_AH_GRH) { ipath_copy_sge(&rsge, &ah_attr->grh, sizeof(struct ib_grh)); wc.wc_flags |= IB_WC_GRH; @@ -225,7 +225,7 @@ static void ipath_ud_loopback(struct ipath_qp *sqp, struct ipath_swqe *swqe) wc.port_num = 1; /* Signal completion event if the solicited bit is set. */ ipath_cq_enter(to_icq(qp->ibqp.recv_cq), &wc, - swqe->wr.send_flags & IB_SEND_SOLICITED); + swqe->ud_wr.wr.send_flags & IB_SEND_SOLICITED); drop: if (atomic_dec_and_test(&qp->refcount)) wake_up(&qp->wait); @@ -280,7 +280,7 @@ int ipath_make_ud_req(struct ipath_qp *qp) next_cur = 0; /* Construct the header. */ - ah_attr = &to_iah(wqe->wr.wr.ud.ah)->attr; + ah_attr = &to_iah(wqe->ud_wr.ah)->attr; if (ah_attr->dlid >= IPATH_MULTICAST_LID_BASE) { if (ah_attr->dlid != IPATH_PERMISSIVE_LID) dev->n_multicast_xmit++; @@ -322,7 +322,7 @@ int ipath_make_ud_req(struct ipath_qp *qp) qp->s_wqe = wqe; qp->s_sge.sge = wqe->sg_list[0]; qp->s_sge.sg_list = wqe->sg_list + 1; - qp->s_sge.num_sge = wqe->wr.num_sge; + qp->s_sge.num_sge = wqe->ud_wr.wr.num_sge; if (ah_attr->ah_flags & IB_AH_GRH) { /* Header size in 32-bit words. */ @@ -340,9 +340,9 @@ int ipath_make_ud_req(struct ipath_qp *qp) lrh0 = IPATH_LRH_BTH; ohdr = &qp->s_hdr.u.oth; } - if (wqe->wr.opcode == IB_WR_SEND_WITH_IMM) { + if (wqe->ud_wr.wr.opcode == IB_WR_SEND_WITH_IMM) { qp->s_hdrwords++; - ohdr->u.ud.imm_data = wqe->wr.ex.imm_data; + ohdr->u.ud.imm_data = wqe->ud_wr.wr.ex.imm_data; bth0 = IB_OPCODE_UD_SEND_ONLY_WITH_IMMEDIATE << 24; } else bth0 = IB_OPCODE_UD_SEND_ONLY << 24; @@ -360,7 +360,7 @@ int ipath_make_ud_req(struct ipath_qp *qp) qp->s_hdr.lrh[3] = cpu_to_be16(lid); } else qp->s_hdr.lrh[3] = IB_LID_PERMISSIVE; - if (wqe->wr.send_flags & IB_SEND_SOLICITED) + if (wqe->ud_wr.wr.send_flags & IB_SEND_SOLICITED) bth0 |= 1 << 23; bth0 |= extra_bytes << 20; bth0 |= qp->ibqp.qp_type == IB_QPT_SMI ? IPATH_DEFAULT_P_KEY : @@ -372,14 +372,14 @@ int ipath_make_ud_req(struct ipath_qp *qp) ohdr->bth[1] = ah_attr->dlid >= IPATH_MULTICAST_LID_BASE && ah_attr->dlid != IPATH_PERMISSIVE_LID ? cpu_to_be32(IPATH_MULTICAST_QPN) : - cpu_to_be32(wqe->wr.wr.ud.remote_qpn); + cpu_to_be32(wqe->ud_wr.remote_qpn); ohdr->bth[2] = cpu_to_be32(qp->s_next_psn++ & IPATH_PSN_MASK); /* * Qkeys with the high order bit set mean use the * qkey from the QP context instead of the WR (see 10.2.5). */ - ohdr->u.ud.deth[0] = cpu_to_be32((int)wqe->wr.wr.ud.remote_qkey < 0 ? - qp->qkey : wqe->wr.wr.ud.remote_qkey); + ohdr->u.ud.deth[0] = cpu_to_be32((int)wqe->ud_wr.remote_qkey < 0 ? + qp->qkey : wqe->ud_wr.remote_qkey); ohdr->u.ud.deth[1] = cpu_to_be32(qp->ibqp.qp_num); done: diff --git a/drivers/staging/rdma/ipath/ipath_verbs.c b/drivers/staging/rdma/ipath/ipath_verbs.c index ed2bbc2f7eae..15633ec1843f 100644 --- a/drivers/staging/rdma/ipath/ipath_verbs.c +++ b/drivers/staging/rdma/ipath/ipath_verbs.c @@ -374,7 +374,7 @@ static int ipath_post_one_send(struct ipath_qp *qp, struct ib_send_wr *wr) wr->opcode != IB_WR_SEND_WITH_IMM) goto bail_inval; /* Check UD destination address PD */ - if (qp->ibqp.pd != wr->wr.ud.ah->pd) + if (qp->ibqp.pd != ud_wr(wr)->ah->pd) goto bail_inval; } else if ((unsigned) wr->opcode > IB_WR_ATOMIC_FETCH_AND_ADD) goto bail_inval; @@ -395,7 +395,23 @@ static int ipath_post_one_send(struct ipath_qp *qp, struct ib_send_wr *wr) } wqe = get_swqe_ptr(qp, qp->s_head); - wqe->wr = *wr; + + if (qp->ibqp.qp_type != IB_QPT_UC && + qp->ibqp.qp_type != IB_QPT_RC) + memcpy(&wqe->ud_wr, ud_wr(wr), sizeof(wqe->ud_wr)); + else if (wr->opcode == IB_WR_FAST_REG_MR) + memcpy(&wqe->fast_reg_wr, fast_reg_wr(wr), + sizeof(wqe->fast_reg_wr)); + else if (wr->opcode == IB_WR_RDMA_WRITE_WITH_IMM || + wr->opcode == IB_WR_RDMA_WRITE || + wr->opcode == IB_WR_RDMA_READ) + memcpy(&wqe->rdma_wr, rdma_wr(wr), sizeof(wqe->rdma_wr)); + else if (wr->opcode == IB_WR_ATOMIC_CMP_AND_SWP || + wr->opcode == IB_WR_ATOMIC_FETCH_AND_ADD) + memcpy(&wqe->atomic_wr, atomic_wr(wr), sizeof(wqe->atomic_wr)); + else + memcpy(&wqe->wr, wr, sizeof(wqe->wr)); + wqe->length = 0; if (wr->num_sge) { acc = wr->opcode >= IB_WR_RDMA_READ ? diff --git a/drivers/staging/rdma/ipath/ipath_verbs.h b/drivers/staging/rdma/ipath/ipath_verbs.h index ec167e545e15..ed102a26ec08 100644 --- a/drivers/staging/rdma/ipath/ipath_verbs.h +++ b/drivers/staging/rdma/ipath/ipath_verbs.h @@ -277,7 +277,14 @@ struct ipath_mr { * in qp->s_max_sge. */ struct ipath_swqe { - struct ib_send_wr wr; /* don't use wr.sg_list */ + union { + struct ib_send_wr wr; /* don't use wr.sg_list */ + struct ib_ud_wr ud_wr; + struct ib_fast_reg_wr fast_reg_wr; + struct ib_rdma_wr rdma_wr; + struct ib_atomic_wr atomic_wr; + }; + u32 psn; /* first packet sequence number */ u32 lpsn; /* last packet sequence number */ u32 ssn; /* send sequence number */ diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index 7845fae6f2df..25f022c9aaac 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -1100,54 +1100,94 @@ struct ib_send_wr { __be32 imm_data; u32 invalidate_rkey; } ex; - union { - struct { - u64 remote_addr; - u32 rkey; - } rdma; - struct { - u64 remote_addr; - u64 compare_add; - u64 swap; - u64 compare_add_mask; - u64 swap_mask; - u32 rkey; - } atomic; - struct { - struct ib_ah *ah; - void *header; - int hlen; - int mss; - u32 remote_qpn; - u32 remote_qkey; - u16 pkey_index; /* valid for GSI only */ - u8 port_num; /* valid for DR SMPs on switch only */ - } ud; - struct { - u64 iova_start; - struct ib_fast_reg_page_list *page_list; - unsigned int page_shift; - unsigned int page_list_len; - u32 length; - int access_flags; - u32 rkey; - } fast_reg; - struct { - struct ib_mw *mw; - /* The new rkey for the memory window. */ - u32 rkey; - struct ib_mw_bind_info bind_info; - } bind_mw; - struct { - struct ib_sig_attrs *sig_attrs; - struct ib_mr *sig_mr; - int access_flags; - struct ib_sge *prot; - } sig_handover; - } wr; u32 xrc_remote_srq_num; /* XRC TGT QPs only */ }; +struct ib_rdma_wr { + struct ib_send_wr wr; + u64 remote_addr; + u32 rkey; +}; + +static inline struct ib_rdma_wr *rdma_wr(struct ib_send_wr *wr) +{ + return container_of(wr, struct ib_rdma_wr, wr); +} + +struct ib_atomic_wr { + struct ib_send_wr wr; + u64 remote_addr; + u64 compare_add; + u64 swap; + u64 compare_add_mask; + u64 swap_mask; + u32 rkey; +}; + +static inline struct ib_atomic_wr *atomic_wr(struct ib_send_wr *wr) +{ + return container_of(wr, struct ib_atomic_wr, wr); +} + +struct ib_ud_wr { + struct ib_send_wr wr; + struct ib_ah *ah; + void *header; + int hlen; + int mss; + u32 remote_qpn; + u32 remote_qkey; + u16 pkey_index; /* valid for GSI only */ + u8 port_num; /* valid for DR SMPs on switch only */ +}; + +static inline struct ib_ud_wr *ud_wr(struct ib_send_wr *wr) +{ + return container_of(wr, struct ib_ud_wr, wr); +} + +struct ib_fast_reg_wr { + struct ib_send_wr wr; + u64 iova_start; + struct ib_fast_reg_page_list *page_list; + unsigned int page_shift; + unsigned int page_list_len; + u32 length; + int access_flags; + u32 rkey; +}; + +static inline struct ib_fast_reg_wr *fast_reg_wr(struct ib_send_wr *wr) +{ + return container_of(wr, struct ib_fast_reg_wr, wr); +} + +struct ib_bind_mw_wr { + struct ib_send_wr wr; + struct ib_mw *mw; + /* The new rkey for the memory window. */ + u32 rkey; + struct ib_mw_bind_info bind_info; +}; + +static inline struct ib_bind_mw_wr *bind_mw_wr(struct ib_send_wr *wr) +{ + return container_of(wr, struct ib_bind_mw_wr, wr); +} + +struct ib_sig_handover_wr { + struct ib_send_wr wr; + struct ib_sig_attrs *sig_attrs; + struct ib_mr *sig_mr; + int access_flags; + struct ib_sge *prot; +}; + +static inline struct ib_sig_handover_wr *sig_handover_wr(struct ib_send_wr *wr) +{ + return container_of(wr, struct ib_sig_handover_wr, wr); +} + struct ib_recv_wr { struct ib_recv_wr *next; u64 wr_id; diff --git a/net/rds/ib.h b/net/rds/ib.h index aae60fda77f6..301c48385166 100644 --- a/net/rds/ib.h +++ b/net/rds/ib.h @@ -69,7 +69,11 @@ struct rds_ib_connect_private { struct rds_ib_send_work { void *s_op; - struct ib_send_wr s_wr; + union { + struct ib_send_wr s_wr; + struct ib_rdma_wr s_rdma_wr; + struct ib_atomic_wr s_atomic_wr; + }; struct ib_sge s_sge[RDS_IB_MAX_SGE]; unsigned long s_queued; }; diff --git a/net/rds/ib_send.c b/net/rds/ib_send.c index 4e88047086b6..987386e9931b 100644 --- a/net/rds/ib_send.c +++ b/net/rds/ib_send.c @@ -785,23 +785,23 @@ int rds_ib_xmit_atomic(struct rds_connection *conn, struct rm_atomic_op *op) send->s_queued = jiffies; if (op->op_type == RDS_ATOMIC_TYPE_CSWP) { - send->s_wr.opcode = IB_WR_MASKED_ATOMIC_CMP_AND_SWP; - send->s_wr.wr.atomic.compare_add = op->op_m_cswp.compare; - send->s_wr.wr.atomic.swap = op->op_m_cswp.swap; - send->s_wr.wr.atomic.compare_add_mask = op->op_m_cswp.compare_mask; - send->s_wr.wr.atomic.swap_mask = op->op_m_cswp.swap_mask; + send->s_atomic_wr.wr.opcode = IB_WR_MASKED_ATOMIC_CMP_AND_SWP; + send->s_atomic_wr.compare_add = op->op_m_cswp.compare; + send->s_atomic_wr.swap = op->op_m_cswp.swap; + send->s_atomic_wr.compare_add_mask = op->op_m_cswp.compare_mask; + send->s_atomic_wr.swap_mask = op->op_m_cswp.swap_mask; } else { /* FADD */ - send->s_wr.opcode = IB_WR_MASKED_ATOMIC_FETCH_AND_ADD; - send->s_wr.wr.atomic.compare_add = op->op_m_fadd.add; - send->s_wr.wr.atomic.swap = 0; - send->s_wr.wr.atomic.compare_add_mask = op->op_m_fadd.nocarry_mask; - send->s_wr.wr.atomic.swap_mask = 0; + send->s_atomic_wr.wr.opcode = IB_WR_MASKED_ATOMIC_FETCH_AND_ADD; + send->s_atomic_wr.compare_add = op->op_m_fadd.add; + send->s_atomic_wr.swap = 0; + send->s_atomic_wr.compare_add_mask = op->op_m_fadd.nocarry_mask; + send->s_atomic_wr.swap_mask = 0; } nr_sig = rds_ib_set_wr_signal_state(ic, send, op->op_notify); - send->s_wr.num_sge = 1; - send->s_wr.next = NULL; - send->s_wr.wr.atomic.remote_addr = op->op_remote_addr; - send->s_wr.wr.atomic.rkey = op->op_rkey; + send->s_atomic_wr.wr.num_sge = 1; + send->s_atomic_wr.wr.next = NULL; + send->s_atomic_wr.remote_addr = op->op_remote_addr; + send->s_atomic_wr.rkey = op->op_rkey; send->s_op = op; rds_message_addref(container_of(send->s_op, struct rds_message, atomic)); @@ -826,11 +826,11 @@ int rds_ib_xmit_atomic(struct rds_connection *conn, struct rm_atomic_op *op) if (nr_sig) atomic_add(nr_sig, &ic->i_signaled_sends); - failed_wr = &send->s_wr; - ret = ib_post_send(ic->i_cm_id->qp, &send->s_wr, &failed_wr); + failed_wr = &send->s_atomic_wr.wr; + ret = ib_post_send(ic->i_cm_id->qp, &send->s_atomic_wr.wr, &failed_wr); rdsdebug("ic %p send %p (wr %p) ret %d wr %p\n", ic, - send, &send->s_wr, ret, failed_wr); - BUG_ON(failed_wr != &send->s_wr); + send, &send->s_atomic_wr, ret, failed_wr); + BUG_ON(failed_wr != &send->s_atomic_wr.wr); if (ret) { printk(KERN_WARNING "RDS/IB: atomic ib_post_send to %pI4 " "returned %d\n", &conn->c_faddr, ret); @@ -839,9 +839,9 @@ int rds_ib_xmit_atomic(struct rds_connection *conn, struct rm_atomic_op *op) goto out; } - if (unlikely(failed_wr != &send->s_wr)) { + if (unlikely(failed_wr != &send->s_atomic_wr.wr)) { printk(KERN_WARNING "RDS/IB: atomic ib_post_send() rc=%d, but failed_wqe updated!\n", ret); - BUG_ON(failed_wr != &send->s_wr); + BUG_ON(failed_wr != &send->s_atomic_wr.wr); } out: @@ -912,22 +912,23 @@ int rds_ib_xmit_rdma(struct rds_connection *conn, struct rm_rdma_op *op) nr_sig += rds_ib_set_wr_signal_state(ic, send, op->op_notify); send->s_wr.opcode = op->op_write ? IB_WR_RDMA_WRITE : IB_WR_RDMA_READ; - send->s_wr.wr.rdma.remote_addr = remote_addr; - send->s_wr.wr.rdma.rkey = op->op_rkey; + send->s_rdma_wr.remote_addr = remote_addr; + send->s_rdma_wr.rkey = op->op_rkey; if (num_sge > max_sge) { - send->s_wr.num_sge = max_sge; + send->s_rdma_wr.wr.num_sge = max_sge; num_sge -= max_sge; } else { - send->s_wr.num_sge = num_sge; + send->s_rdma_wr.wr.num_sge = num_sge; } - send->s_wr.next = NULL; + send->s_rdma_wr.wr.next = NULL; if (prev) - prev->s_wr.next = &send->s_wr; + prev->s_rdma_wr.wr.next = &send->s_rdma_wr.wr; - for (j = 0; j < send->s_wr.num_sge && scat != &op->op_sg[op->op_count]; j++) { + for (j = 0; j < send->s_rdma_wr.wr.num_sge && + scat != &op->op_sg[op->op_count]; j++) { len = ib_sg_dma_len(ic->i_cm_id->device, scat); send->s_sge[j].addr = ib_sg_dma_address(ic->i_cm_id->device, scat); @@ -942,7 +943,9 @@ int rds_ib_xmit_rdma(struct rds_connection *conn, struct rm_rdma_op *op) } rdsdebug("send %p wr %p num_sge %u next %p\n", send, - &send->s_wr, send->s_wr.num_sge, send->s_wr.next); + &send->s_rdma_wr.wr, + send->s_rdma_wr.wr.num_sge, + send->s_rdma_wr.wr.next); prev = send; if (++send == &ic->i_sends[ic->i_send_ring.w_nr]) @@ -963,11 +966,11 @@ int rds_ib_xmit_rdma(struct rds_connection *conn, struct rm_rdma_op *op) if (nr_sig) atomic_add(nr_sig, &ic->i_signaled_sends); - failed_wr = &first->s_wr; - ret = ib_post_send(ic->i_cm_id->qp, &first->s_wr, &failed_wr); + failed_wr = &first->s_rdma_wr.wr; + ret = ib_post_send(ic->i_cm_id->qp, &first->s_rdma_wr.wr, &failed_wr); rdsdebug("ic %p first %p (wr %p) ret %d wr %p\n", ic, - first, &first->s_wr, ret, failed_wr); - BUG_ON(failed_wr != &first->s_wr); + first, &first->s_rdma_wr.wr, ret, failed_wr); + BUG_ON(failed_wr != &first->s_rdma_wr.wr); if (ret) { printk(KERN_WARNING "RDS/IB: rdma ib_post_send to %pI4 " "returned %d\n", &conn->c_faddr, ret); @@ -976,9 +979,9 @@ int rds_ib_xmit_rdma(struct rds_connection *conn, struct rm_rdma_op *op) goto out; } - if (unlikely(failed_wr != &first->s_wr)) { + if (unlikely(failed_wr != &first->s_rdma_wr.wr)) { printk(KERN_WARNING "RDS/IB: ib_post_send() rc=%d, but failed_wqe updated!\n", ret); - BUG_ON(failed_wr != &first->s_wr); + BUG_ON(failed_wr != &first->s_rdma_wr.wr); } diff --git a/net/rds/iw.h b/net/rds/iw.h index cbe6674e31ee..fe858e5dd8d1 100644 --- a/net/rds/iw.h +++ b/net/rds/iw.h @@ -77,7 +77,11 @@ struct rds_iw_send_work { struct ib_fast_reg_page_list *s_page_list; unsigned char s_remap_count; - struct ib_send_wr s_wr; + union { + struct ib_send_wr s_send_wr; + struct ib_rdma_wr s_rdma_wr; + struct ib_fast_reg_wr s_fast_reg_wr; + }; struct ib_sge s_sge[RDS_IW_MAX_SGE]; unsigned long s_queued; }; diff --git a/net/rds/iw_rdma.c b/net/rds/iw_rdma.c index 6a8fbd6e69e7..f8a612cc69e6 100644 --- a/net/rds/iw_rdma.c +++ b/net/rds/iw_rdma.c @@ -696,7 +696,8 @@ static int rds_iw_init_fastreg(struct rds_iw_mr_pool *pool, static int rds_iw_rdma_build_fastreg(struct rds_iw_mapping *mapping) { struct rds_iw_mr *ibmr = mapping->m_mr; - struct ib_send_wr f_wr, *failed_wr; + struct ib_fast_reg_wr f_wr; + struct ib_send_wr *failed_wr; int ret; /* @@ -709,22 +710,22 @@ static int rds_iw_rdma_build_fastreg(struct rds_iw_mapping *mapping) mapping->m_rkey = ibmr->mr->rkey; memset(&f_wr, 0, sizeof(f_wr)); - f_wr.wr_id = RDS_IW_FAST_REG_WR_ID; - f_wr.opcode = IB_WR_FAST_REG_MR; - f_wr.wr.fast_reg.length = mapping->m_sg.bytes; - f_wr.wr.fast_reg.rkey = mapping->m_rkey; - f_wr.wr.fast_reg.page_list = ibmr->page_list; - f_wr.wr.fast_reg.page_list_len = mapping->m_sg.dma_len; - f_wr.wr.fast_reg.page_shift = PAGE_SHIFT; - f_wr.wr.fast_reg.access_flags = IB_ACCESS_LOCAL_WRITE | + f_wr.wr.wr_id = RDS_IW_FAST_REG_WR_ID; + f_wr.wr.opcode = IB_WR_FAST_REG_MR; + f_wr.length = mapping->m_sg.bytes; + f_wr.rkey = mapping->m_rkey; + f_wr.page_list = ibmr->page_list; + f_wr.page_list_len = mapping->m_sg.dma_len; + f_wr.page_shift = PAGE_SHIFT; + f_wr.access_flags = IB_ACCESS_LOCAL_WRITE | IB_ACCESS_REMOTE_READ | IB_ACCESS_REMOTE_WRITE; - f_wr.wr.fast_reg.iova_start = 0; - f_wr.send_flags = IB_SEND_SIGNALED; + f_wr.iova_start = 0; + f_wr.wr.send_flags = IB_SEND_SIGNALED; - failed_wr = &f_wr; - ret = ib_post_send(ibmr->cm_id->qp, &f_wr, &failed_wr); - BUG_ON(failed_wr != &f_wr); + failed_wr = &f_wr.wr; + ret = ib_post_send(ibmr->cm_id->qp, &f_wr.wr, &failed_wr); + BUG_ON(failed_wr != &f_wr.wr); if (ret) printk_ratelimited(KERN_WARNING "RDS/IW: %s:%d ib_post_send returned %d\n", __func__, __LINE__, ret); diff --git a/net/rds/iw_send.c b/net/rds/iw_send.c index 86152ec3b887..f6e23c515b44 100644 --- a/net/rds/iw_send.c +++ b/net/rds/iw_send.c @@ -137,13 +137,13 @@ void rds_iw_send_init_ring(struct rds_iw_connection *ic) send->s_op = NULL; send->s_mapping = NULL; - send->s_wr.next = NULL; - send->s_wr.wr_id = i; - send->s_wr.sg_list = send->s_sge; - send->s_wr.num_sge = 1; - send->s_wr.opcode = IB_WR_SEND; - send->s_wr.send_flags = 0; - send->s_wr.ex.imm_data = 0; + send->s_send_wr.next = NULL; + send->s_send_wr.wr_id = i; + send->s_send_wr.sg_list = send->s_sge; + send->s_send_wr.num_sge = 1; + send->s_send_wr.opcode = IB_WR_SEND; + send->s_send_wr.send_flags = 0; + send->s_send_wr.ex.imm_data = 0; sge = rds_iw_data_sge(ic, send->s_sge); sge->lkey = 0; @@ -179,7 +179,7 @@ void rds_iw_send_clear_ring(struct rds_iw_connection *ic) ib_dereg_mr(send->s_mr); BUG_ON(!send->s_page_list); ib_free_fast_reg_page_list(send->s_page_list); - if (send->s_wr.opcode == 0xdead) + if (send->s_send_wr.opcode == 0xdead) continue; if (send->s_rm) rds_iw_send_unmap_rm(ic, send, IB_WC_WR_FLUSH_ERR); @@ -247,7 +247,7 @@ void rds_iw_send_cq_comp_handler(struct ib_cq *cq, void *context) send = &ic->i_sends[oldest]; /* In the error case, wc.opcode sometimes contains garbage */ - switch (send->s_wr.opcode) { + switch (send->s_send_wr.opcode) { case IB_WR_SEND: if (send->s_rm) rds_iw_send_unmap_rm(ic, send, wc.status); @@ -262,12 +262,12 @@ void rds_iw_send_cq_comp_handler(struct ib_cq *cq, void *context) default: printk_ratelimited(KERN_NOTICE "RDS/IW: %s: unexpected opcode 0x%x in WR!\n", - __func__, send->s_wr.opcode); + __func__, send->s_send_wr.opcode); break; } - send->s_wr.opcode = 0xdead; - send->s_wr.num_sge = 1; + send->s_send_wr.opcode = 0xdead; + send->s_send_wr.num_sge = 1; if (time_after(jiffies, send->s_queued + HZ/2)) rds_iw_stats_inc(s_iw_tx_stalled); @@ -455,10 +455,10 @@ rds_iw_xmit_populate_wr(struct rds_iw_connection *ic, WARN_ON(pos != send - ic->i_sends); - send->s_wr.send_flags = send_flags; - send->s_wr.opcode = IB_WR_SEND; - send->s_wr.num_sge = 2; - send->s_wr.next = NULL; + send->s_send_wr.send_flags = send_flags; + send->s_send_wr.opcode = IB_WR_SEND; + send->s_send_wr.num_sge = 2; + send->s_send_wr.next = NULL; send->s_queued = jiffies; send->s_op = NULL; @@ -472,7 +472,7 @@ rds_iw_xmit_populate_wr(struct rds_iw_connection *ic, } else { /* We're sending a packet with no payload. There is only * one SGE */ - send->s_wr.num_sge = 1; + send->s_send_wr.num_sge = 1; sge = &send->s_sge[0]; } @@ -672,23 +672,23 @@ int rds_iw_xmit(struct rds_connection *conn, struct rds_message *rm, */ if (ic->i_unsignaled_wrs-- == 0) { ic->i_unsignaled_wrs = rds_iw_sysctl_max_unsig_wrs; - send->s_wr.send_flags |= IB_SEND_SIGNALED | IB_SEND_SOLICITED; + send->s_send_wr.send_flags |= IB_SEND_SIGNALED | IB_SEND_SOLICITED; } ic->i_unsignaled_bytes -= len; if (ic->i_unsignaled_bytes <= 0) { ic->i_unsignaled_bytes = rds_iw_sysctl_max_unsig_bytes; - send->s_wr.send_flags |= IB_SEND_SIGNALED | IB_SEND_SOLICITED; + send->s_send_wr.send_flags |= IB_SEND_SIGNALED | IB_SEND_SOLICITED; } /* * Always signal the last one if we're stopping due to flow control. */ if (flow_controlled && i == (work_alloc-1)) - send->s_wr.send_flags |= IB_SEND_SIGNALED | IB_SEND_SOLICITED; + send->s_send_wr.send_flags |= IB_SEND_SIGNALED | IB_SEND_SOLICITED; rdsdebug("send %p wr %p num_sge %u next %p\n", send, - &send->s_wr, send->s_wr.num_sge, send->s_wr.next); + &send->s_send_wr, send->s_send_wr.num_sge, send->s_send_wr.next); sent += len; rm->data.op_dmaoff += len; @@ -722,7 +722,7 @@ add_header: } if (prev) - prev->s_wr.next = &send->s_wr; + prev->s_send_wr.next = &send->s_send_wr; prev = send; pos = (pos + 1) % ic->i_send_ring.w_nr; @@ -736,7 +736,7 @@ add_header: /* if we finished the message then send completion owns it */ if (scat == &rm->data.op_sg[rm->data.op_count]) { prev->s_rm = ic->i_rm; - prev->s_wr.send_flags |= IB_SEND_SIGNALED | IB_SEND_SOLICITED; + prev->s_send_wr.send_flags |= IB_SEND_SIGNALED | IB_SEND_SOLICITED; ic->i_rm = NULL; } @@ -748,11 +748,11 @@ add_header: rds_iw_send_add_credits(conn, credit_alloc - i); /* XXX need to worry about failed_wr and partial sends. */ - failed_wr = &first->s_wr; - ret = ib_post_send(ic->i_cm_id->qp, &first->s_wr, &failed_wr); + failed_wr = &first->s_send_wr; + ret = ib_post_send(ic->i_cm_id->qp, &first->s_send_wr, &failed_wr); rdsdebug("ic %p first %p (wr %p) ret %d wr %p\n", ic, - first, &first->s_wr, ret, failed_wr); - BUG_ON(failed_wr != &first->s_wr); + first, &first->s_send_wr, ret, failed_wr); + BUG_ON(failed_wr != &first->s_send_wr); if (ret) { printk(KERN_WARNING "RDS/IW: ib_post_send to %pI4 " "returned %d\n", &conn->c_faddr, ret); @@ -778,14 +778,14 @@ static void rds_iw_build_send_fastreg(struct rds_iw_device *rds_iwdev, struct rd * in the sg list is added to the fast reg page list and placed * inside the fast_reg_mr WR. */ - send->s_wr.opcode = IB_WR_FAST_REG_MR; - send->s_wr.wr.fast_reg.length = len; - send->s_wr.wr.fast_reg.rkey = send->s_mr->rkey; - send->s_wr.wr.fast_reg.page_list = send->s_page_list; - send->s_wr.wr.fast_reg.page_list_len = nent; - send->s_wr.wr.fast_reg.page_shift = PAGE_SHIFT; - send->s_wr.wr.fast_reg.access_flags = IB_ACCESS_REMOTE_WRITE; - send->s_wr.wr.fast_reg.iova_start = sg_addr; + send->s_fast_reg_wr.wr.opcode = IB_WR_FAST_REG_MR; + send->s_fast_reg_wr.length = len; + send->s_fast_reg_wr.rkey = send->s_mr->rkey; + send->s_fast_reg_wr.page_list = send->s_page_list; + send->s_fast_reg_wr.page_list_len = nent; + send->s_fast_reg_wr.page_shift = PAGE_SHIFT; + send->s_fast_reg_wr.access_flags = IB_ACCESS_REMOTE_WRITE; + send->s_fast_reg_wr.iova_start = sg_addr; ib_update_fast_reg_key(send->s_mr, send->s_remap_count++); } @@ -863,7 +863,7 @@ int rds_iw_xmit_rdma(struct rds_connection *conn, struct rm_rdma_op *op) num_sge = op->op_count; for (i = 0; i < work_alloc && scat != &op->op_sg[op->op_count]; i++) { - send->s_wr.send_flags = 0; + send->s_rdma_wr.wr.send_flags = 0; send->s_queued = jiffies; /* @@ -872,7 +872,7 @@ int rds_iw_xmit_rdma(struct rds_connection *conn, struct rm_rdma_op *op) */ if (ic->i_unsignaled_wrs-- == 0) { ic->i_unsignaled_wrs = rds_iw_sysctl_max_unsig_wrs; - send->s_wr.send_flags = IB_SEND_SIGNALED; + send->s_rdma_wr.wr.send_flags = IB_SEND_SIGNALED; } /* To avoid the need to have the plumbing to invalidate the fastreg_mr used @@ -880,29 +880,30 @@ int rds_iw_xmit_rdma(struct rds_connection *conn, struct rm_rdma_op *op) * IB_WR_RDMA_READ_WITH_INV will invalidate it after the read has completed. */ if (op->op_write) - send->s_wr.opcode = IB_WR_RDMA_WRITE; + send->s_rdma_wr.wr.opcode = IB_WR_RDMA_WRITE; else - send->s_wr.opcode = IB_WR_RDMA_READ_WITH_INV; + send->s_rdma_wr.wr.opcode = IB_WR_RDMA_READ_WITH_INV; - send->s_wr.wr.rdma.remote_addr = remote_addr; - send->s_wr.wr.rdma.rkey = op->op_rkey; + send->s_rdma_wr.remote_addr = remote_addr; + send->s_rdma_wr.rkey = op->op_rkey; send->s_op = op; if (num_sge > rds_iwdev->max_sge) { - send->s_wr.num_sge = rds_iwdev->max_sge; + send->s_rdma_wr.wr.num_sge = rds_iwdev->max_sge; num_sge -= rds_iwdev->max_sge; } else - send->s_wr.num_sge = num_sge; + send->s_rdma_wr.wr.num_sge = num_sge; - send->s_wr.next = NULL; + send->s_rdma_wr.wr.next = NULL; if (prev) - prev->s_wr.next = &send->s_wr; + prev->s_send_wr.next = &send->s_rdma_wr.wr; - for (j = 0; j < send->s_wr.num_sge && scat != &op->op_sg[op->op_count]; j++) { + for (j = 0; j < send->s_rdma_wr.wr.num_sge && + scat != &op->op_sg[op->op_count]; j++) { len = ib_sg_dma_len(ic->i_cm_id->device, scat); - if (send->s_wr.opcode == IB_WR_RDMA_READ_WITH_INV) + if (send->s_rdma_wr.wr.opcode == IB_WR_RDMA_READ_WITH_INV) send->s_page_list->page_list[j] = ib_sg_dma_address(ic->i_cm_id->device, scat); else { send->s_sge[j].addr = ib_sg_dma_address(ic->i_cm_id->device, scat); @@ -917,15 +918,17 @@ int rds_iw_xmit_rdma(struct rds_connection *conn, struct rm_rdma_op *op) scat++; } - if (send->s_wr.opcode == IB_WR_RDMA_READ_WITH_INV) { - send->s_wr.num_sge = 1; + if (send->s_rdma_wr.wr.opcode == IB_WR_RDMA_READ_WITH_INV) { + send->s_rdma_wr.wr.num_sge = 1; send->s_sge[0].addr = conn->c_xmit_rm->m_rs->rs_user_addr; send->s_sge[0].length = conn->c_xmit_rm->m_rs->rs_user_bytes; send->s_sge[0].lkey = ic->i_sends[fr_pos].s_mr->lkey; } rdsdebug("send %p wr %p num_sge %u next %p\n", send, - &send->s_wr, send->s_wr.num_sge, send->s_wr.next); + &send->s_rdma_wr, + send->s_rdma_wr.wr.num_sge, + send->s_rdma_wr.wr.next); prev = send; if (++send == &ic->i_sends[ic->i_send_ring.w_nr]) @@ -934,7 +937,7 @@ int rds_iw_xmit_rdma(struct rds_connection *conn, struct rm_rdma_op *op) /* if we finished the message then send completion owns it */ if (scat == &op->op_sg[op->op_count]) - first->s_wr.send_flags = IB_SEND_SIGNALED; + first->s_rdma_wr.wr.send_flags = IB_SEND_SIGNALED; if (i < work_alloc) { rds_iw_ring_unalloc(&ic->i_send_ring, work_alloc - i); @@ -953,11 +956,11 @@ int rds_iw_xmit_rdma(struct rds_connection *conn, struct rm_rdma_op *op) work_alloc++; } - failed_wr = &first->s_wr; - ret = ib_post_send(ic->i_cm_id->qp, &first->s_wr, &failed_wr); + failed_wr = &first->s_rdma_wr.wr; + ret = ib_post_send(ic->i_cm_id->qp, &first->s_rdma_wr.wr, &failed_wr); rdsdebug("ic %p first %p (wr %p) ret %d wr %p\n", ic, - first, &first->s_wr, ret, failed_wr); - BUG_ON(failed_wr != &first->s_wr); + first, &first->s_rdma_wr, ret, failed_wr); + BUG_ON(failed_wr != &first->s_rdma_wr.wr); if (ret) { printk(KERN_WARNING "RDS/IW: rdma ib_post_send to %pI4 " "returned %d\n", &conn->c_faddr, ret); diff --git a/net/sunrpc/xprtrdma/frwr_ops.c b/net/sunrpc/xprtrdma/frwr_ops.c index 5318951b3b53..0d2f46f600b6 100644 --- a/net/sunrpc/xprtrdma/frwr_ops.c +++ b/net/sunrpc/xprtrdma/frwr_ops.c @@ -312,7 +312,8 @@ frwr_op_map(struct rpcrdma_xprt *r_xprt, struct rpcrdma_mr_seg *seg, struct rpcrdma_mw *mw; struct rpcrdma_frmr *frmr; struct ib_mr *mr; - struct ib_send_wr fastreg_wr, *bad_wr; + struct ib_fast_reg_wr fastreg_wr; + struct ib_send_wr *bad_wr; u8 key; int len, pageoff; int i, rc; @@ -358,23 +359,23 @@ frwr_op_map(struct rpcrdma_xprt *r_xprt, struct rpcrdma_mr_seg *seg, __func__, mw, i, len); memset(&fastreg_wr, 0, sizeof(fastreg_wr)); - fastreg_wr.wr_id = (unsigned long)(void *)mw; - fastreg_wr.opcode = IB_WR_FAST_REG_MR; - fastreg_wr.wr.fast_reg.iova_start = seg1->mr_dma + pageoff; - fastreg_wr.wr.fast_reg.page_list = frmr->fr_pgl; - fastreg_wr.wr.fast_reg.page_shift = PAGE_SHIFT; - fastreg_wr.wr.fast_reg.page_list_len = page_no; - fastreg_wr.wr.fast_reg.length = len; - fastreg_wr.wr.fast_reg.access_flags = writing ? + fastreg_wr.wr.wr_id = (unsigned long)(void *)mw; + fastreg_wr.wr.opcode = IB_WR_FAST_REG_MR; + fastreg_wr.iova_start = seg1->mr_dma + pageoff; + fastreg_wr.page_list = frmr->fr_pgl; + fastreg_wr.page_shift = PAGE_SHIFT; + fastreg_wr.page_list_len = page_no; + fastreg_wr.length = len; + fastreg_wr.access_flags = writing ? IB_ACCESS_REMOTE_WRITE | IB_ACCESS_LOCAL_WRITE : IB_ACCESS_REMOTE_READ; mr = frmr->fr_mr; key = (u8)(mr->rkey & 0x000000FF); ib_update_fast_reg_key(mr, ++key); - fastreg_wr.wr.fast_reg.rkey = mr->rkey; + fastreg_wr.rkey = mr->rkey; DECR_CQCOUNT(&r_xprt->rx_ep); - rc = ib_post_send(ia->ri_id->qp, &fastreg_wr, &bad_wr); + rc = ib_post_send(ia->ri_id->qp, &fastreg_wr.wr, &bad_wr); if (rc) goto out_senderr; diff --git a/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c b/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c index cb5174284074..7be42d0da19e 100644 --- a/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c +++ b/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c @@ -126,7 +126,7 @@ int rdma_read_chunk_lcl(struct svcxprt_rdma *xprt, u64 rs_offset, bool last) { - struct ib_send_wr read_wr; + struct ib_rdma_wr read_wr; int pages_needed = PAGE_ALIGN(*page_offset + rs_length) >> PAGE_SHIFT; struct svc_rdma_op_ctxt *ctxt = svc_rdma_get_context(xprt); int ret, read, pno; @@ -179,16 +179,16 @@ int rdma_read_chunk_lcl(struct svcxprt_rdma *xprt, clear_bit(RDMACTXT_F_LAST_CTXT, &ctxt->flags); memset(&read_wr, 0, sizeof(read_wr)); - read_wr.wr_id = (unsigned long)ctxt; - read_wr.opcode = IB_WR_RDMA_READ; - ctxt->wr_op = read_wr.opcode; - read_wr.send_flags = IB_SEND_SIGNALED; - read_wr.wr.rdma.rkey = rs_handle; - read_wr.wr.rdma.remote_addr = rs_offset; - read_wr.sg_list = ctxt->sge; - read_wr.num_sge = pages_needed; - - ret = svc_rdma_send(xprt, &read_wr); + read_wr.wr.wr_id = (unsigned long)ctxt; + read_wr.wr.opcode = IB_WR_RDMA_READ; + ctxt->wr_op = read_wr.wr.opcode; + read_wr.wr.send_flags = IB_SEND_SIGNALED; + read_wr.rkey = rs_handle; + read_wr.remote_addr = rs_offset; + read_wr.wr.sg_list = ctxt->sge; + read_wr.wr.num_sge = pages_needed; + + ret = svc_rdma_send(xprt, &read_wr.wr); if (ret) { pr_err("svcrdma: Error %d posting RDMA_READ\n", ret); set_bit(XPT_CLOSE, &xprt->sc_xprt.xpt_flags); @@ -218,9 +218,9 @@ int rdma_read_chunk_frmr(struct svcxprt_rdma *xprt, u64 rs_offset, bool last) { - struct ib_send_wr read_wr; + struct ib_rdma_wr read_wr; struct ib_send_wr inv_wr; - struct ib_send_wr fastreg_wr; + struct ib_fast_reg_wr fastreg_wr; u8 key; int pages_needed = PAGE_ALIGN(*page_offset + rs_length) >> PAGE_SHIFT; struct svc_rdma_op_ctxt *ctxt = svc_rdma_get_context(xprt); @@ -289,31 +289,31 @@ int rdma_read_chunk_frmr(struct svcxprt_rdma *xprt, /* Prepare FASTREG WR */ memset(&fastreg_wr, 0, sizeof(fastreg_wr)); - fastreg_wr.opcode = IB_WR_FAST_REG_MR; - fastreg_wr.send_flags = IB_SEND_SIGNALED; - fastreg_wr.wr.fast_reg.iova_start = (unsigned long)frmr->kva; - fastreg_wr.wr.fast_reg.page_list = frmr->page_list; - fastreg_wr.wr.fast_reg.page_list_len = frmr->page_list_len; - fastreg_wr.wr.fast_reg.page_shift = PAGE_SHIFT; - fastreg_wr.wr.fast_reg.length = frmr->map_len; - fastreg_wr.wr.fast_reg.access_flags = frmr->access_flags; - fastreg_wr.wr.fast_reg.rkey = frmr->mr->lkey; - fastreg_wr.next = &read_wr; + fastreg_wr.wr.opcode = IB_WR_FAST_REG_MR; + fastreg_wr.wr.send_flags = IB_SEND_SIGNALED; + fastreg_wr.iova_start = (unsigned long)frmr->kva; + fastreg_wr.page_list = frmr->page_list; + fastreg_wr.page_list_len = frmr->page_list_len; + fastreg_wr.page_shift = PAGE_SHIFT; + fastreg_wr.length = frmr->map_len; + fastreg_wr.access_flags = frmr->access_flags; + fastreg_wr.rkey = frmr->mr->lkey; + fastreg_wr.wr.next = &read_wr.wr; /* Prepare RDMA_READ */ memset(&read_wr, 0, sizeof(read_wr)); - read_wr.send_flags = IB_SEND_SIGNALED; - read_wr.wr.rdma.rkey = rs_handle; - read_wr.wr.rdma.remote_addr = rs_offset; - read_wr.sg_list = ctxt->sge; - read_wr.num_sge = 1; + read_wr.wr.send_flags = IB_SEND_SIGNALED; + read_wr.rkey = rs_handle; + read_wr.remote_addr = rs_offset; + read_wr.wr.sg_list = ctxt->sge; + read_wr.wr.num_sge = 1; if (xprt->sc_dev_caps & SVCRDMA_DEVCAP_READ_W_INV) { - read_wr.opcode = IB_WR_RDMA_READ_WITH_INV; - read_wr.wr_id = (unsigned long)ctxt; - read_wr.ex.invalidate_rkey = ctxt->frmr->mr->lkey; + read_wr.wr.opcode = IB_WR_RDMA_READ_WITH_INV; + read_wr.wr.wr_id = (unsigned long)ctxt; + read_wr.wr.ex.invalidate_rkey = ctxt->frmr->mr->lkey; } else { - read_wr.opcode = IB_WR_RDMA_READ; - read_wr.next = &inv_wr; + read_wr.wr.opcode = IB_WR_RDMA_READ; + read_wr.wr.next = &inv_wr; /* Prepare invalidate */ memset(&inv_wr, 0, sizeof(inv_wr)); inv_wr.wr_id = (unsigned long)ctxt; @@ -321,10 +321,10 @@ int rdma_read_chunk_frmr(struct svcxprt_rdma *xprt, inv_wr.send_flags = IB_SEND_SIGNALED | IB_SEND_FENCE; inv_wr.ex.invalidate_rkey = frmr->mr->lkey; } - ctxt->wr_op = read_wr.opcode; + ctxt->wr_op = read_wr.wr.opcode; /* Post the chain */ - ret = svc_rdma_send(xprt, &fastreg_wr); + ret = svc_rdma_send(xprt, &fastreg_wr.wr); if (ret) { pr_err("svcrdma: Error %d posting RDMA_READ\n", ret); set_bit(XPT_CLOSE, &xprt->sc_xprt.xpt_flags); diff --git a/net/sunrpc/xprtrdma/svc_rdma_sendto.c b/net/sunrpc/xprtrdma/svc_rdma_sendto.c index 1dfae8317065..969a1ab75fc3 100644 --- a/net/sunrpc/xprtrdma/svc_rdma_sendto.c +++ b/net/sunrpc/xprtrdma/svc_rdma_sendto.c @@ -217,7 +217,7 @@ static int send_write(struct svcxprt_rdma *xprt, struct svc_rqst *rqstp, u32 xdr_off, int write_len, struct svc_rdma_req_map *vec) { - struct ib_send_wr write_wr; + struct ib_rdma_wr write_wr; struct ib_sge *sge; int xdr_sge_no; int sge_no; @@ -282,17 +282,17 @@ static int send_write(struct svcxprt_rdma *xprt, struct svc_rqst *rqstp, /* Prepare WRITE WR */ memset(&write_wr, 0, sizeof write_wr); ctxt->wr_op = IB_WR_RDMA_WRITE; - write_wr.wr_id = (unsigned long)ctxt; - write_wr.sg_list = &sge[0]; - write_wr.num_sge = sge_no; - write_wr.opcode = IB_WR_RDMA_WRITE; - write_wr.send_flags = IB_SEND_SIGNALED; - write_wr.wr.rdma.rkey = rmr; - write_wr.wr.rdma.remote_addr = to; + write_wr.wr.wr_id = (unsigned long)ctxt; + write_wr.wr.sg_list = &sge[0]; + write_wr.wr.num_sge = sge_no; + write_wr.wr.opcode = IB_WR_RDMA_WRITE; + write_wr.wr.send_flags = IB_SEND_SIGNALED; + write_wr.rkey = rmr; + write_wr.remote_addr = to; /* Post It */ atomic_inc(&rdma_stat_write); - if (svc_rdma_send(xprt, &write_wr)) + if (svc_rdma_send(xprt, &write_wr.wr)) goto err; return write_len - bc; err: -- cgit v1.2.3 From 25556ae6b965321c7e7469faa06ddbeae50dac91 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Wed, 19 Aug 2015 14:58:43 +0200 Subject: IB: remove xrc_remote_srq_num from struct ib_send_wr The field is only initialized in mlx, but never used. If we want to add proper XRC support it should be done with a new struct ib_xrc_wr. This shrinks the various WR structures by another 4 bytes. Signed-off-by: Christoph Hellwig Reviewed-by: Sagi Grimberg Reviewed-by: Jason Gunthorpe Tested-by: Haggai Eran --- drivers/infiniband/hw/mlx5/qp.c | 1 - include/rdma/ib_verbs.h | 1 - 2 files changed, 2 deletions(-) (limited to 'include') diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c index d4c36af4270f..9bad68820061 100644 --- a/drivers/infiniband/hw/mlx5/qp.c +++ b/drivers/infiniband/hw/mlx5/qp.c @@ -2627,7 +2627,6 @@ int mlx5_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr, switch (ibqp->qp_type) { case IB_QPT_XRC_INI: xrc = seg; - xrc->xrc_srqn = htonl(wr->xrc_remote_srq_num); seg += sizeof(*xrc); size += sizeof(*xrc) / 16; /* fall through */ diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index 25f022c9aaac..edf02908a0fd 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -1100,7 +1100,6 @@ struct ib_send_wr { __be32 imm_data; u32 invalidate_rkey; } ex; - u32 xrc_remote_srq_num; /* XRC TGT QPs only */ }; struct ib_rdma_wr { -- cgit v1.2.3 From 931cf9a3e55c7a4c0d7340140643ce6fdf7b6e96 Mon Sep 17 00:00:00 2001 From: Hal Rosenstock Date: Wed, 30 Sep 2015 15:04:29 -0400 Subject: ib_pack.h: Fix commentary IBA reference for CNP in IB opcode enum IBA spec is now 1.3 not 3.1 and vol 1 should be mentioned as there is also vol 2. Signed-off-by: Hal Rosenstock Signed-off-by: Doug Ledford --- include/rdma/ib_pack.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include') diff --git a/include/rdma/ib_pack.h b/include/rdma/ib_pack.h index 709a5331e6b9..e99d8f9a4551 100644 --- a/include/rdma/ib_pack.h +++ b/include/rdma/ib_pack.h @@ -76,7 +76,7 @@ enum { IB_OPCODE_UC = 0x20, IB_OPCODE_RD = 0x40, IB_OPCODE_UD = 0x60, - /* per IBTA 3.1 Table 38, A10.3.2 */ + /* per IBTA 1.3 vol 1 Table 38, A10.3.2 */ IB_OPCODE_CNP = 0x80, /* operations -- just used to define real constants */ -- cgit v1.2.3 From 6d8a74972b7115b41d4d17a4444c026755ca24be Mon Sep 17 00:00:00 2001 From: Eran Ben Elisha Date: Wed, 21 Oct 2015 17:00:42 +0300 Subject: IB/core: Extend ib_uverbs_create_qp ib_uverbs_ex_create_qp follows the extension verbs mechanism. New features (for example, QP creation flags field which is added in a downstream patch) could used via user-space libraries without breaking the ABI. Signed-off-by: Eran Ben Elisha Reviewed-by: Haggai Eran Signed-off-by: Doug Ledford --- drivers/infiniband/core/uverbs.h | 1 + drivers/infiniband/core/uverbs_cmd.c | 254 +++++++++++++++++++++++++--------- drivers/infiniband/core/uverbs_main.c | 1 + include/uapi/rdma/ib_user_verbs.h | 26 ++++ 4 files changed, 217 insertions(+), 65 deletions(-) (limited to 'include') diff --git a/drivers/infiniband/core/uverbs.h b/drivers/infiniband/core/uverbs.h index 3863d33c243d..94bbd8c155fc 100644 --- a/drivers/infiniband/core/uverbs.h +++ b/drivers/infiniband/core/uverbs.h @@ -272,5 +272,6 @@ IB_UVERBS_DECLARE_EX_CMD(create_flow); IB_UVERBS_DECLARE_EX_CMD(destroy_flow); IB_UVERBS_DECLARE_EX_CMD(query_device); IB_UVERBS_DECLARE_EX_CMD(create_cq); +IB_UVERBS_DECLARE_EX_CMD(create_qp); #endif /* UVERBS_H */ diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c index 88b3b78340f2..b2a37d545e0b 100644 --- a/drivers/infiniband/core/uverbs_cmd.c +++ b/drivers/infiniband/core/uverbs_cmd.c @@ -1741,66 +1741,65 @@ ssize_t ib_uverbs_destroy_cq(struct ib_uverbs_file *file, return in_len; } -ssize_t ib_uverbs_create_qp(struct ib_uverbs_file *file, - struct ib_device *ib_dev, - const char __user *buf, int in_len, - int out_len) -{ - struct ib_uverbs_create_qp cmd; - struct ib_uverbs_create_qp_resp resp; - struct ib_udata udata; - struct ib_uqp_object *obj; - struct ib_device *device; - struct ib_pd *pd = NULL; - struct ib_xrcd *xrcd = NULL; - struct ib_uobject *uninitialized_var(xrcd_uobj); - struct ib_cq *scq = NULL, *rcq = NULL; - struct ib_srq *srq = NULL; - struct ib_qp *qp; - struct ib_qp_init_attr attr; - int ret; - - if (out_len < sizeof resp) - return -ENOSPC; - - if (copy_from_user(&cmd, buf, sizeof cmd)) - return -EFAULT; +static int create_qp(struct ib_uverbs_file *file, + struct ib_udata *ucore, + struct ib_udata *uhw, + struct ib_uverbs_ex_create_qp *cmd, + size_t cmd_sz, + int (*cb)(struct ib_uverbs_file *file, + struct ib_uverbs_ex_create_qp_resp *resp, + struct ib_udata *udata), + void *context) +{ + struct ib_uqp_object *obj; + struct ib_device *device; + struct ib_pd *pd = NULL; + struct ib_xrcd *xrcd = NULL; + struct ib_uobject *uninitialized_var(xrcd_uobj); + struct ib_cq *scq = NULL, *rcq = NULL; + struct ib_srq *srq = NULL; + struct ib_qp *qp; + char *buf; + struct ib_qp_init_attr attr; + struct ib_uverbs_ex_create_qp_resp resp; + int ret; - if (cmd.qp_type == IB_QPT_RAW_PACKET && !capable(CAP_NET_RAW)) + if (cmd->qp_type == IB_QPT_RAW_PACKET && !capable(CAP_NET_RAW)) return -EPERM; - INIT_UDATA(&udata, buf + sizeof cmd, - (unsigned long) cmd.response + sizeof resp, - in_len - sizeof cmd, out_len - sizeof resp); - obj = kzalloc(sizeof *obj, GFP_KERNEL); if (!obj) return -ENOMEM; - init_uobj(&obj->uevent.uobject, cmd.user_handle, file->ucontext, &qp_lock_class); + init_uobj(&obj->uevent.uobject, cmd->user_handle, file->ucontext, + &qp_lock_class); down_write(&obj->uevent.uobject.mutex); - if (cmd.qp_type == IB_QPT_XRC_TGT) { - xrcd = idr_read_xrcd(cmd.pd_handle, file->ucontext, &xrcd_uobj); + if (cmd->qp_type == IB_QPT_XRC_TGT) { + xrcd = idr_read_xrcd(cmd->pd_handle, file->ucontext, + &xrcd_uobj); if (!xrcd) { ret = -EINVAL; goto err_put; } device = xrcd->device; } else { - if (cmd.qp_type == IB_QPT_XRC_INI) { - cmd.max_recv_wr = cmd.max_recv_sge = 0; + if (cmd->qp_type == IB_QPT_XRC_INI) { + cmd->max_recv_wr = 0; + cmd->max_recv_sge = 0; } else { - if (cmd.is_srq) { - srq = idr_read_srq(cmd.srq_handle, file->ucontext); + if (cmd->is_srq) { + srq = idr_read_srq(cmd->srq_handle, + file->ucontext); if (!srq || srq->srq_type != IB_SRQT_BASIC) { ret = -EINVAL; goto err_put; } } - if (cmd.recv_cq_handle != cmd.send_cq_handle) { - rcq = idr_read_cq(cmd.recv_cq_handle, file->ucontext, 0); + if (cmd->recv_cq_handle != cmd->send_cq_handle) { + rcq = idr_read_cq(cmd->recv_cq_handle, + file->ucontext, 0); if (!rcq) { ret = -EINVAL; goto err_put; @@ -1808,9 +1807,9 @@ ssize_t ib_uverbs_create_qp(struct ib_uverbs_file *file, } } - scq = idr_read_cq(cmd.send_cq_handle, file->ucontext, !!rcq); + scq = idr_read_cq(cmd->send_cq_handle, file->ucontext, !!rcq); rcq = rcq ?: scq; - pd = idr_read_pd(cmd.pd_handle, file->ucontext); + pd = idr_read_pd(cmd->pd_handle, file->ucontext); if (!pd || !scq) { ret = -EINVAL; goto err_put; @@ -1825,31 +1824,49 @@ ssize_t ib_uverbs_create_qp(struct ib_uverbs_file *file, attr.recv_cq = rcq; attr.srq = srq; attr.xrcd = xrcd; - attr.sq_sig_type = cmd.sq_sig_all ? IB_SIGNAL_ALL_WR : IB_SIGNAL_REQ_WR; - attr.qp_type = cmd.qp_type; + attr.sq_sig_type = cmd->sq_sig_all ? IB_SIGNAL_ALL_WR : + IB_SIGNAL_REQ_WR; + attr.qp_type = cmd->qp_type; attr.create_flags = 0; - attr.cap.max_send_wr = cmd.max_send_wr; - attr.cap.max_recv_wr = cmd.max_recv_wr; - attr.cap.max_send_sge = cmd.max_send_sge; - attr.cap.max_recv_sge = cmd.max_recv_sge; - attr.cap.max_inline_data = cmd.max_inline_data; + attr.cap.max_send_wr = cmd->max_send_wr; + attr.cap.max_recv_wr = cmd->max_recv_wr; + attr.cap.max_send_sge = cmd->max_send_sge; + attr.cap.max_recv_sge = cmd->max_recv_sge; + attr.cap.max_inline_data = cmd->max_inline_data; obj->uevent.events_reported = 0; INIT_LIST_HEAD(&obj->uevent.event_list); INIT_LIST_HEAD(&obj->mcast_list); - if (cmd.qp_type == IB_QPT_XRC_TGT) + if (cmd_sz >= offsetof(typeof(*cmd), create_flags) + + sizeof(cmd->create_flags)) + attr.create_flags = cmd->create_flags; + + if (attr.create_flags) { + ret = -EINVAL; + goto err_put; + } + + buf = (void *)cmd + sizeof(*cmd); + if (cmd_sz > sizeof(*cmd)) + if (!(buf[0] == 0 && !memcmp(buf, buf + 1, + cmd_sz - sizeof(*cmd) - 1))) { + ret = -EINVAL; + goto err_put; + } + + if (cmd->qp_type == IB_QPT_XRC_TGT) qp = ib_create_qp(pd, &attr); else - qp = device->create_qp(pd, &attr, &udata); + qp = device->create_qp(pd, &attr, uhw); if (IS_ERR(qp)) { ret = PTR_ERR(qp); goto err_put; } - if (cmd.qp_type != IB_QPT_XRC_TGT) { + if (cmd->qp_type != IB_QPT_XRC_TGT) { qp->real_qp = qp; qp->device = device; qp->pd = pd; @@ -1875,19 +1892,20 @@ ssize_t ib_uverbs_create_qp(struct ib_uverbs_file *file, goto err_destroy; memset(&resp, 0, sizeof resp); - resp.qpn = qp->qp_num; - resp.qp_handle = obj->uevent.uobject.id; - resp.max_recv_sge = attr.cap.max_recv_sge; - resp.max_send_sge = attr.cap.max_send_sge; - resp.max_recv_wr = attr.cap.max_recv_wr; - resp.max_send_wr = attr.cap.max_send_wr; - resp.max_inline_data = attr.cap.max_inline_data; + resp.base.qpn = qp->qp_num; + resp.base.qp_handle = obj->uevent.uobject.id; + resp.base.max_recv_sge = attr.cap.max_recv_sge; + resp.base.max_send_sge = attr.cap.max_send_sge; + resp.base.max_recv_wr = attr.cap.max_recv_wr; + resp.base.max_send_wr = attr.cap.max_send_wr; + resp.base.max_inline_data = attr.cap.max_inline_data; - if (copy_to_user((void __user *) (unsigned long) cmd.response, - &resp, sizeof resp)) { - ret = -EFAULT; - goto err_copy; - } + resp.response_length = offsetof(typeof(resp), response_length) + + sizeof(resp.response_length); + + ret = cb(file, &resp, ucore); + if (ret) + goto err_cb; if (xrcd) { obj->uxrcd = container_of(xrcd_uobj, struct ib_uxrcd_object, @@ -1913,9 +1931,8 @@ ssize_t ib_uverbs_create_qp(struct ib_uverbs_file *file, up_write(&obj->uevent.uobject.mutex); - return in_len; - -err_copy: + return 0; +err_cb: idr_remove_uobj(&ib_uverbs_qp_idr, &obj->uevent.uobject); err_destroy: @@ -1937,6 +1954,113 @@ err_put: return ret; } +static int ib_uverbs_create_qp_cb(struct ib_uverbs_file *file, + struct ib_uverbs_ex_create_qp_resp *resp, + struct ib_udata *ucore) +{ + if (ib_copy_to_udata(ucore, &resp->base, sizeof(resp->base))) + return -EFAULT; + + return 0; +} + +ssize_t ib_uverbs_create_qp(struct ib_uverbs_file *file, + struct ib_device *ib_dev, + const char __user *buf, int in_len, + int out_len) +{ + struct ib_uverbs_create_qp cmd; + struct ib_uverbs_ex_create_qp cmd_ex; + struct ib_udata ucore; + struct ib_udata uhw; + ssize_t resp_size = sizeof(struct ib_uverbs_create_qp_resp); + int err; + + if (out_len < resp_size) + return -ENOSPC; + + if (copy_from_user(&cmd, buf, sizeof(cmd))) + return -EFAULT; + + INIT_UDATA(&ucore, buf, (unsigned long)cmd.response, sizeof(cmd), + resp_size); + INIT_UDATA(&uhw, buf + sizeof(cmd), + (unsigned long)cmd.response + resp_size, + in_len - sizeof(cmd), out_len - resp_size); + + memset(&cmd_ex, 0, sizeof(cmd_ex)); + cmd_ex.user_handle = cmd.user_handle; + cmd_ex.pd_handle = cmd.pd_handle; + cmd_ex.send_cq_handle = cmd.send_cq_handle; + cmd_ex.recv_cq_handle = cmd.recv_cq_handle; + cmd_ex.srq_handle = cmd.srq_handle; + cmd_ex.max_send_wr = cmd.max_send_wr; + cmd_ex.max_recv_wr = cmd.max_recv_wr; + cmd_ex.max_send_sge = cmd.max_send_sge; + cmd_ex.max_recv_sge = cmd.max_recv_sge; + cmd_ex.max_inline_data = cmd.max_inline_data; + cmd_ex.sq_sig_all = cmd.sq_sig_all; + cmd_ex.qp_type = cmd.qp_type; + cmd_ex.is_srq = cmd.is_srq; + + err = create_qp(file, &ucore, &uhw, &cmd_ex, + offsetof(typeof(cmd_ex), is_srq) + + sizeof(cmd.is_srq), ib_uverbs_create_qp_cb, + NULL); + + if (err) + return err; + + return in_len; +} + +static int ib_uverbs_ex_create_qp_cb(struct ib_uverbs_file *file, + struct ib_uverbs_ex_create_qp_resp *resp, + struct ib_udata *ucore) +{ + if (ib_copy_to_udata(ucore, resp, resp->response_length)) + return -EFAULT; + + return 0; +} + +int ib_uverbs_ex_create_qp(struct ib_uverbs_file *file, + struct ib_device *ib_dev, + struct ib_udata *ucore, + struct ib_udata *uhw) +{ + struct ib_uverbs_ex_create_qp_resp resp; + struct ib_uverbs_ex_create_qp cmd = {0}; + int err; + + if (ucore->inlen < (offsetof(typeof(cmd), comp_mask) + + sizeof(cmd.comp_mask))) + return -EINVAL; + + err = ib_copy_from_udata(&cmd, ucore, min(sizeof(cmd), ucore->inlen)); + if (err) + return err; + + if (cmd.comp_mask) + return -EINVAL; + + if (cmd.reserved) + return -EINVAL; + + if (ucore->outlen < (offsetof(typeof(resp), response_length) + + sizeof(resp.response_length))) + return -ENOSPC; + + err = create_qp(file, ucore, uhw, &cmd, + min(ucore->inlen, sizeof(cmd)), + ib_uverbs_ex_create_qp_cb, NULL); + + if (err) + return err; + + return 0; +} + ssize_t ib_uverbs_open_qp(struct ib_uverbs_file *file, struct ib_device *ib_dev, const char __user *buf, int in_len, int out_len) diff --git a/drivers/infiniband/core/uverbs_main.c b/drivers/infiniband/core/uverbs_main.c index c29a660c72fe..e3ef28861be6 100644 --- a/drivers/infiniband/core/uverbs_main.c +++ b/drivers/infiniband/core/uverbs_main.c @@ -127,6 +127,7 @@ static int (*uverbs_ex_cmd_table[])(struct ib_uverbs_file *file, [IB_USER_VERBS_EX_CMD_DESTROY_FLOW] = ib_uverbs_ex_destroy_flow, [IB_USER_VERBS_EX_CMD_QUERY_DEVICE] = ib_uverbs_ex_query_device, [IB_USER_VERBS_EX_CMD_CREATE_CQ] = ib_uverbs_ex_create_cq, + [IB_USER_VERBS_EX_CMD_CREATE_QP] = ib_uverbs_ex_create_qp, }; static void ib_uverbs_add_one(struct ib_device *device); diff --git a/include/uapi/rdma/ib_user_verbs.h b/include/uapi/rdma/ib_user_verbs.h index 978841eeaff1..8126c143a519 100644 --- a/include/uapi/rdma/ib_user_verbs.h +++ b/include/uapi/rdma/ib_user_verbs.h @@ -92,6 +92,7 @@ enum { enum { IB_USER_VERBS_EX_CMD_QUERY_DEVICE = IB_USER_VERBS_CMD_QUERY_DEVICE, IB_USER_VERBS_EX_CMD_CREATE_CQ = IB_USER_VERBS_CMD_CREATE_CQ, + IB_USER_VERBS_EX_CMD_CREATE_QP = IB_USER_VERBS_CMD_CREATE_QP, IB_USER_VERBS_EX_CMD_CREATE_FLOW = IB_USER_VERBS_CMD_THRESHOLD, IB_USER_VERBS_EX_CMD_DESTROY_FLOW, }; @@ -516,6 +517,25 @@ struct ib_uverbs_create_qp { __u64 driver_data[0]; }; +struct ib_uverbs_ex_create_qp { + __u64 user_handle; + __u32 pd_handle; + __u32 send_cq_handle; + __u32 recv_cq_handle; + __u32 srq_handle; + __u32 max_send_wr; + __u32 max_recv_wr; + __u32 max_send_sge; + __u32 max_recv_sge; + __u32 max_inline_data; + __u8 sq_sig_all; + __u8 qp_type; + __u8 is_srq; + __u8 reserved; + __u32 comp_mask; + __u32 create_flags; +}; + struct ib_uverbs_open_qp { __u64 response; __u64 user_handle; @@ -538,6 +558,12 @@ struct ib_uverbs_create_qp_resp { __u32 reserved; }; +struct ib_uverbs_ex_create_qp_resp { + struct ib_uverbs_create_qp_resp base; + __u32 comp_mask; + __u32 response_length; +}; + /* * This struct needs to remain a multiple of 8 bytes to keep the * alignment of the modify QP parameters. -- cgit v1.2.3 From 9a8928359736ab170303ee8a2cc15db54e3a4a8f Mon Sep 17 00:00:00 2001 From: Maor Gottlieb Date: Thu, 15 Oct 2015 14:44:38 +0300 Subject: net/mlx4_core: Add support for filtering multicast loopback Update device capabilities regarding HW filtering multicast loopback support. Add MLX4_UPDATE_QP_ETH_SRC_CHECK_MC_LB attribute to mlx4_update_qp to enable changing QP context to support filtering incoming multicast loopback traffic according the sender's counter index. Set the corresponding bits in QP context to force the loopback source checks if attribute is given and HW supports it. Signed-off-by: Maor Gottlieb Signed-off-by: Eran Ben Elisha Signed-off-by: Doug Ledford --- drivers/net/ethernet/mellanox/mlx4/fw.c | 6 +++++ drivers/net/ethernet/mellanox/mlx4/qp.c | 19 +++++++++++++- .../net/ethernet/mellanox/mlx4/resource_tracker.c | 30 +++++++++++++++++----- include/linux/mlx4/device.h | 2 ++ include/linux/mlx4/qp.h | 24 +++++++++++++---- 5 files changed, 68 insertions(+), 13 deletions(-) (limited to 'include') diff --git a/drivers/net/ethernet/mellanox/mlx4/fw.c b/drivers/net/ethernet/mellanox/mlx4/fw.c index e8ec1dec5789..b3be3a060311 100644 --- a/drivers/net/ethernet/mellanox/mlx4/fw.c +++ b/drivers/net/ethernet/mellanox/mlx4/fw.c @@ -155,6 +155,8 @@ static void dump_dev_cap_flags2(struct mlx4_dev *dev, u64 flags) [27] = "Port beacon support", [28] = "RX-ALL support", [29] = "802.1ad offload support", + [31] = "Modifying loopback source checks using UPDATE_QP support", + [32] = "Loopback source checks support", }; int i; @@ -964,6 +966,10 @@ int mlx4_QUERY_DEV_CAP(struct mlx4_dev *dev, struct mlx4_dev_cap *dev_cap) MLX4_GET(field32, outbox, QUERY_DEV_CAP_EXT_2_FLAGS_OFFSET); if (field32 & (1 << 16)) dev_cap->flags2 |= MLX4_DEV_CAP_FLAG2_UPDATE_QP; + if (field32 & (1 << 18)) + dev_cap->flags2 |= MLX4_DEV_CAP_FLAG2_UPDATE_QP_SRC_CHECK_LB; + if (field32 & (1 << 19)) + dev_cap->flags2 |= MLX4_DEV_CAP_FLAG2_LB_SRC_CHK; if (field32 & (1 << 26)) dev_cap->flags2 |= MLX4_DEV_CAP_FLAG2_VLAN_CONTROL; if (field32 & (1 << 20)) diff --git a/drivers/net/ethernet/mellanox/mlx4/qp.c b/drivers/net/ethernet/mellanox/mlx4/qp.c index 20268634a9ab..b16249577aa2 100644 --- a/drivers/net/ethernet/mellanox/mlx4/qp.c +++ b/drivers/net/ethernet/mellanox/mlx4/qp.c @@ -436,6 +436,23 @@ int mlx4_update_qp(struct mlx4_dev *dev, u32 qpn, cmd->qp_context.pri_path.grh_mylmc = params->smac_index; } + if (attr & MLX4_UPDATE_QP_ETH_SRC_CHECK_MC_LB) { + if (!(dev->caps.flags2 + & MLX4_DEV_CAP_FLAG2_UPDATE_QP_SRC_CHECK_LB)) { + mlx4_warn(dev, + "Trying to set src check LB, but it isn't supported\n"); + err = -ENOTSUPP; + goto out; + } + pri_addr_path_mask |= + 1ULL << MLX4_UPD_QP_PATH_MASK_ETH_SRC_CHECK_MC_LB; + if (params->flags & + MLX4_UPDATE_QP_PARAMS_FLAGS_ETH_CHECK_MC_LB) { + cmd->qp_context.pri_path.fl |= + MLX4_FL_ETH_SRC_CHECK_MC_LB; + } + } + if (attr & MLX4_UPDATE_QP_VSD) { qp_mask |= 1ULL << MLX4_UPD_QP_MASK_VSD; if (params->flags & MLX4_UPDATE_QP_PARAMS_FLAGS_VSD_ENABLE) @@ -458,7 +475,7 @@ int mlx4_update_qp(struct mlx4_dev *dev, u32 qpn, err = mlx4_cmd(dev, mailbox->dma, qpn & 0xffffff, 0, MLX4_CMD_UPDATE_QP, MLX4_CMD_TIME_CLASS_A, MLX4_CMD_NATIVE); - +out: mlx4_free_cmd_mailbox(dev, mailbox); return err; } diff --git a/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c b/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c index 731423ca575d..502f3350088e 100644 --- a/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c +++ b/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c @@ -770,9 +770,12 @@ static int update_vport_qp_param(struct mlx4_dev *dev, } } + /* preserve IF_COUNTER flag */ + qpc->pri_path.vlan_control &= + MLX4_CTRL_ETH_SRC_CHECK_IF_COUNTER; if (vp_oper->state.link_state == IFLA_VF_LINK_STATE_DISABLE && dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_UPDATE_QP) { - qpc->pri_path.vlan_control = + qpc->pri_path.vlan_control |= MLX4_VLAN_CTRL_ETH_TX_BLOCK_TAGGED | MLX4_VLAN_CTRL_ETH_TX_BLOCK_PRIO_TAGGED | MLX4_VLAN_CTRL_ETH_TX_BLOCK_UNTAGGED | @@ -780,12 +783,12 @@ static int update_vport_qp_param(struct mlx4_dev *dev, MLX4_VLAN_CTRL_ETH_RX_BLOCK_UNTAGGED | MLX4_VLAN_CTRL_ETH_RX_BLOCK_TAGGED; } else if (0 != vp_oper->state.default_vlan) { - qpc->pri_path.vlan_control = + qpc->pri_path.vlan_control |= MLX4_VLAN_CTRL_ETH_TX_BLOCK_TAGGED | MLX4_VLAN_CTRL_ETH_RX_BLOCK_PRIO_TAGGED | MLX4_VLAN_CTRL_ETH_RX_BLOCK_UNTAGGED; } else { /* priority tagged */ - qpc->pri_path.vlan_control = + qpc->pri_path.vlan_control |= MLX4_VLAN_CTRL_ETH_TX_BLOCK_TAGGED | MLX4_VLAN_CTRL_ETH_RX_BLOCK_TAGGED; } @@ -3762,9 +3765,6 @@ int mlx4_INIT2RTR_QP_wrapper(struct mlx4_dev *dev, int slave, update_gid(dev, inbox, (u8)slave); adjust_proxy_tun_qkey(dev, vhcr, qpc); orig_sched_queue = qpc->pri_path.sched_queue; - err = update_vport_qp_param(dev, inbox, slave, qpn); - if (err) - return err; err = get_res(dev, slave, qpn, RES_QP, &qp); if (err) @@ -3774,6 +3774,10 @@ int mlx4_INIT2RTR_QP_wrapper(struct mlx4_dev *dev, int slave, goto out; } + err = update_vport_qp_param(dev, inbox, slave, qpn); + if (err) + goto out; + err = mlx4_DMA_wrapper(dev, slave, vhcr, inbox, outbox, cmd); out: /* if no error, save sched queue value passed in by VF. This is @@ -4208,7 +4212,9 @@ static int add_eth_header(struct mlx4_dev *dev, int slave, } -#define MLX4_UPD_QP_PATH_MASK_SUPPORTED (1ULL << MLX4_UPD_QP_PATH_MASK_MAC_INDEX) +#define MLX4_UPD_QP_PATH_MASK_SUPPORTED ( \ + 1ULL << MLX4_UPD_QP_PATH_MASK_MAC_INDEX |\ + 1ULL << MLX4_UPD_QP_PATH_MASK_ETH_SRC_CHECK_MC_LB) int mlx4_UPDATE_QP_wrapper(struct mlx4_dev *dev, int slave, struct mlx4_vhcr *vhcr, struct mlx4_cmd_mailbox *inbox, @@ -4231,6 +4237,16 @@ int mlx4_UPDATE_QP_wrapper(struct mlx4_dev *dev, int slave, (pri_addr_path_mask & ~MLX4_UPD_QP_PATH_MASK_SUPPORTED)) return -EPERM; + if ((pri_addr_path_mask & + (1ULL << MLX4_UPD_QP_PATH_MASK_ETH_SRC_CHECK_MC_LB)) && + !(dev->caps.flags2 & + MLX4_DEV_CAP_FLAG2_UPDATE_QP_SRC_CHECK_LB)) { + mlx4_warn(dev, + "Src check LB for slave %d isn't supported\n", + slave); + return -ENOTSUPP; + } + /* Just change the smac for the QP */ err = get_res(dev, slave, qpn, RES_QP, &rqp); if (err) { diff --git a/include/linux/mlx4/device.h b/include/linux/mlx4/device.h index baad4cb8e9b0..dac6872dbaea 100644 --- a/include/linux/mlx4/device.h +++ b/include/linux/mlx4/device.h @@ -214,6 +214,8 @@ enum { MLX4_DEV_CAP_FLAG2_IGNORE_FCS = 1LL << 28, MLX4_DEV_CAP_FLAG2_PHV_EN = 1LL << 29, MLX4_DEV_CAP_FLAG2_SKIP_OUTER_VLAN = 1LL << 30, + MLX4_DEV_CAP_FLAG2_UPDATE_QP_SRC_CHECK_LB = 1ULL << 31, + MLX4_DEV_CAP_FLAG2_LB_SRC_CHK = 1ULL << 32, }; enum { diff --git a/include/linux/mlx4/qp.h b/include/linux/mlx4/qp.h index de45a51b3f04..fe052e234906 100644 --- a/include/linux/mlx4/qp.h +++ b/include/linux/mlx4/qp.h @@ -135,7 +135,10 @@ struct mlx4_rss_context { struct mlx4_qp_path { u8 fl; - u8 vlan_control; + union { + u8 vlan_control; + u8 control; + }; u8 disable_pkey_check; u8 pkey_index; u8 counter_index; @@ -156,9 +159,16 @@ struct mlx4_qp_path { }; enum { /* fl */ - MLX4_FL_CV = 1 << 6, - MLX4_FL_ETH_HIDE_CQE_VLAN = 1 << 2 + MLX4_FL_CV = 1 << 6, + MLX4_FL_ETH_HIDE_CQE_VLAN = 1 << 2, + MLX4_FL_ETH_SRC_CHECK_MC_LB = 1 << 1, + MLX4_FL_ETH_SRC_CHECK_UC_LB = 1 << 0, }; + +enum { /* control */ + MLX4_CTRL_ETH_SRC_CHECK_IF_COUNTER = 1 << 7, +}; + enum { /* vlan_control */ MLX4_VLAN_CTRL_ETH_TX_BLOCK_TAGGED = 1 << 6, MLX4_VLAN_CTRL_ETH_TX_BLOCK_PRIO_TAGGED = 1 << 5, /* 802.1p priority tag */ @@ -254,6 +264,8 @@ enum { MLX4_UPD_QP_PATH_MASK_SCHED_QUEUE = 14 + 32, MLX4_UPD_QP_PATH_MASK_IF_COUNTER_INDEX = 15 + 32, MLX4_UPD_QP_PATH_MASK_FVL_RX = 16 + 32, + MLX4_UPD_QP_PATH_MASK_ETH_SRC_CHECK_UC_LB = 18 + 32, + MLX4_UPD_QP_PATH_MASK_ETH_SRC_CHECK_MC_LB = 19 + 32, }; enum { /* param3 */ @@ -436,11 +448,13 @@ enum mlx4_update_qp_attr { MLX4_UPDATE_QP_VSD = 1 << 1, MLX4_UPDATE_QP_RATE_LIMIT = 1 << 2, MLX4_UPDATE_QP_QOS_VPORT = 1 << 3, - MLX4_UPDATE_QP_SUPPORTED_ATTRS = (1 << 4) - 1 + MLX4_UPDATE_QP_ETH_SRC_CHECK_MC_LB = 1 << 4, + MLX4_UPDATE_QP_SUPPORTED_ATTRS = (1 << 5) - 1 }; enum mlx4_update_qp_params_flags { - MLX4_UPDATE_QP_PARAMS_FLAGS_VSD_ENABLE = 1 << 0, + MLX4_UPDATE_QP_PARAMS_FLAGS_ETH_CHECK_MC_LB = 1 << 0, + MLX4_UPDATE_QP_PARAMS_FLAGS_VSD_ENABLE = 1 << 1, }; struct mlx4_update_qp_params { -- cgit v1.2.3 From 55ee3ab2e49a9ead850722ef47698243dd226d16 Mon Sep 17 00:00:00 2001 From: Matan Barak Date: Thu, 15 Oct 2015 18:38:45 +0300 Subject: IB/core: Add netdev and gid attributes paramteres to cache Adding an ability to query the IB cache by a netdev and get the attributes of a GID. These parameters are necessary in order to successfully resolve the required GID (when the netdevice is known) and get the Ethernet L2 attributes from a GID. Signed-off-by: Matan Barak Reviewed-By: Devesh Sharma Signed-off-by: Doug Ledford --- drivers/infiniband/core/cache.c | 10 ++++++---- drivers/infiniband/core/cm.c | 5 +++-- drivers/infiniband/core/cma.c | 10 ++++++---- drivers/infiniband/core/device.c | 17 ++++++++++++----- drivers/infiniband/core/mad.c | 2 +- drivers/infiniband/core/multicast.c | 3 ++- drivers/infiniband/core/sa_query.c | 2 +- drivers/infiniband/core/sysfs.c | 2 +- drivers/infiniband/core/verbs.c | 7 ++++--- drivers/infiniband/hw/mlx4/main.c | 4 ++-- drivers/infiniband/hw/mlx4/qp.c | 5 +++-- drivers/infiniband/hw/mthca/mthca_av.c | 2 +- drivers/infiniband/hw/ocrdma/ocrdma_verbs.c | 2 +- drivers/infiniband/ulp/ipoib/ipoib_main.c | 2 +- drivers/infiniband/ulp/ipoib/ipoib_multicast.c | 2 +- drivers/infiniband/ulp/srp/ib_srp.c | 2 +- drivers/infiniband/ulp/srpt/ib_srpt.c | 3 ++- include/rdma/ib_cache.h | 13 +++++++++---- include/rdma/ib_verbs.h | 5 +++-- 19 files changed, 60 insertions(+), 38 deletions(-) (limited to 'include') diff --git a/drivers/infiniband/core/cache.c b/drivers/infiniband/core/cache.c index 87471ef37198..5c054072ef20 100644 --- a/drivers/infiniband/core/cache.c +++ b/drivers/infiniband/core/cache.c @@ -649,21 +649,23 @@ static int gid_table_setup_one(struct ib_device *ib_dev) int ib_get_cached_gid(struct ib_device *device, u8 port_num, int index, - union ib_gid *gid) + union ib_gid *gid, + struct ib_gid_attr *gid_attr) { if (port_num < rdma_start_port(device) || port_num > rdma_end_port(device)) return -EINVAL; - return __ib_cache_gid_get(device, port_num, index, gid, NULL); + return __ib_cache_gid_get(device, port_num, index, gid, gid_attr); } EXPORT_SYMBOL(ib_get_cached_gid); int ib_find_cached_gid(struct ib_device *device, const union ib_gid *gid, + struct net_device *ndev, u8 *port_num, u16 *index) { - return ib_cache_gid_find(device, gid, NULL, port_num, index); + return ib_cache_gid_find(device, gid, ndev, port_num, index); } EXPORT_SYMBOL(ib_find_cached_gid); @@ -845,7 +847,7 @@ static void ib_cache_update(struct ib_device *device, if (!use_roce_gid_table) { for (i = 0; i < gid_cache->table_len; ++i) { ret = ib_query_gid(device, port, i, - gid_cache->table + i); + gid_cache->table + i, NULL); if (ret) { printk(KERN_WARNING "ib_query_gid failed (%d) for %s (index %d)\n", ret, device->name, i); diff --git a/drivers/infiniband/core/cm.c b/drivers/infiniband/core/cm.c index 4f918b929eca..2d8a0e4c42d6 100644 --- a/drivers/infiniband/core/cm.c +++ b/drivers/infiniband/core/cm.c @@ -365,7 +365,7 @@ static int cm_init_av_by_path(struct ib_sa_path_rec *path, struct cm_av *av) read_lock_irqsave(&cm.device_lock, flags); list_for_each_entry(cm_dev, &cm.device_list, list) { if (!ib_find_cached_gid(cm_dev->ib_device, &path->sgid, - &p, NULL)) { + NULL, &p, NULL)) { port = cm_dev->port[p-1]; break; } @@ -1643,7 +1643,8 @@ static int cm_req_handler(struct cm_work *work) ret = cm_init_av_by_path(&work->path[0], &cm_id_priv->av); if (ret) { ib_get_cached_gid(work->port->cm_dev->ib_device, - work->port->port_num, 0, &work->path[0].sgid); + work->port->port_num, 0, &work->path[0].sgid, + NULL); ib_send_cm_rej(cm_id, IB_CM_REJ_INVALID_GID, &work->path[0].sgid, sizeof work->path[0].sgid, NULL, 0); diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index 36b12d560e17..e4d9d75ffc08 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -438,7 +438,7 @@ static inline int cma_validate_port(struct ib_device *device, u8 port, if ((dev_type != ARPHRD_INFINIBAND) && rdma_protocol_ib(device, port)) return ret; - ret = ib_find_cached_gid(device, gid, &found_port, NULL); + ret = ib_find_cached_gid(device, gid, NULL, &found_port, NULL); if (port != found_port) return -ENODEV; @@ -531,7 +531,9 @@ static int cma_resolve_ib_dev(struct rdma_id_private *id_priv) if (ib_find_cached_pkey(cur_dev->device, p, pkey, &index)) continue; - for (i = 0; !ib_get_cached_gid(cur_dev->device, p, i, &gid); i++) { + for (i = 0; !ib_get_cached_gid(cur_dev->device, p, i, + &gid, NULL); + i++) { if (!memcmp(&gid, dgid, sizeof(gid))) { cma_dev = cur_dev; sgid = gid; @@ -718,7 +720,7 @@ static int cma_modify_qp_rtr(struct rdma_id_private *id_priv, goto out; ret = ib_query_gid(id_priv->id.device, id_priv->id.port_num, - qp_attr.ah_attr.grh.sgid_index, &sgid); + qp_attr.ah_attr.grh.sgid_index, &sgid, NULL); if (ret) goto out; @@ -2426,7 +2428,7 @@ static int cma_bind_loopback(struct rdma_id_private *id_priv) p = 1; port_found: - ret = ib_get_cached_gid(cma_dev->device, p, 0, &gid); + ret = ib_get_cached_gid(cma_dev->device, p, 0, &gid, NULL); if (ret) goto out; diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c index 17639117afc6..f22ce487fd3a 100644 --- a/drivers/infiniband/core/device.c +++ b/drivers/infiniband/core/device.c @@ -672,14 +672,20 @@ EXPORT_SYMBOL(ib_query_port); * @port_num:Port number to query * @index:GID table index to query * @gid:Returned GID + * @attr: Returned GID attributes related to this GID index (only in RoCE). + * NULL means ignore. * * ib_query_gid() fetches the specified GID table entry. */ int ib_query_gid(struct ib_device *device, - u8 port_num, int index, union ib_gid *gid) + u8 port_num, int index, union ib_gid *gid, + struct ib_gid_attr *attr) { if (rdma_cap_roce_gid_table(device, port_num)) - return ib_get_cached_gid(device, port_num, index, gid); + return ib_get_cached_gid(device, port_num, index, gid, attr); + + if (attr) + return -EINVAL; return device->query_gid(device, port_num, index, gid); } @@ -819,12 +825,13 @@ EXPORT_SYMBOL(ib_modify_port); * a specified GID value occurs. * @device: The device to query. * @gid: The GID value to search for. + * @ndev: The ndev related to the GID to search for. * @port_num: The port number of the device where the GID value was found. * @index: The index into the GID table where the GID was found. This * parameter may be NULL. */ int ib_find_gid(struct ib_device *device, union ib_gid *gid, - u8 *port_num, u16 *index) + struct net_device *ndev, u8 *port_num, u16 *index) { union ib_gid tmp_gid; int ret, port, i; @@ -832,14 +839,14 @@ int ib_find_gid(struct ib_device *device, union ib_gid *gid, for (port = rdma_start_port(device); port <= rdma_end_port(device); ++port) { if (rdma_cap_roce_gid_table(device, port)) { if (!ib_cache_gid_find_by_port(device, gid, port, - NULL, index)) { + ndev, index)) { *port_num = port; return 0; } } for (i = 0; i < device->port_immutable[port].gid_tbl_len; ++i) { - ret = ib_query_gid(device, port, i, &tmp_gid); + ret = ib_query_gid(device, port, i, &tmp_gid, NULL); if (ret) return ret; if (!memcmp(&tmp_gid, gid, sizeof *gid)) { diff --git a/drivers/infiniband/core/mad.c b/drivers/infiniband/core/mad.c index 4b5c72311deb..fa63b89e15aa 100644 --- a/drivers/infiniband/core/mad.c +++ b/drivers/infiniband/core/mad.c @@ -1877,7 +1877,7 @@ static inline int rcv_has_same_gid(const struct ib_mad_agent_private *mad_agent_ ((1 << lmc) - 1))); } else { if (ib_get_cached_gid(device, port_num, - attr.grh.sgid_index, &sgid)) + attr.grh.sgid_index, &sgid, NULL)) return 0; return !memcmp(sgid.raw, rwc->recv_buf.grh->dgid.raw, 16); diff --git a/drivers/infiniband/core/multicast.c b/drivers/infiniband/core/multicast.c index d38d8b2b2979..bb6685fb08c6 100644 --- a/drivers/infiniband/core/multicast.c +++ b/drivers/infiniband/core/multicast.c @@ -729,7 +729,8 @@ int ib_init_ah_from_mcmember(struct ib_device *device, u8 port_num, u16 gid_index; u8 p; - ret = ib_find_cached_gid(device, &rec->port_gid, &p, &gid_index); + ret = ib_find_cached_gid(device, &rec->port_gid, + NULL, &p, &gid_index); if (ret) return ret; diff --git a/drivers/infiniband/core/sa_query.c b/drivers/infiniband/core/sa_query.c index 8c014b33d8e0..9a4e7891ada2 100644 --- a/drivers/infiniband/core/sa_query.c +++ b/drivers/infiniband/core/sa_query.c @@ -1010,7 +1010,7 @@ int ib_init_ah_from_path(struct ib_device *device, u8 port_num, ah_attr->ah_flags = IB_AH_GRH; ah_attr->grh.dgid = rec->dgid; - ret = ib_find_cached_gid(device, &rec->sgid, &port_num, + ret = ib_find_cached_gid(device, &rec->sgid, NULL, &port_num, &gid_index); if (ret) return ret; diff --git a/drivers/infiniband/core/sysfs.c b/drivers/infiniband/core/sysfs.c index 34cdd74b0a17..b1f37d4095fa 100644 --- a/drivers/infiniband/core/sysfs.c +++ b/drivers/infiniband/core/sysfs.c @@ -289,7 +289,7 @@ static ssize_t show_port_gid(struct ib_port *p, struct port_attribute *attr, union ib_gid gid; ssize_t ret; - ret = ib_query_gid(p->ibdev, p->port_num, tab_attr->index, &gid); + ret = ib_query_gid(p->ibdev, p->port_num, tab_attr->index, &gid, NULL); if (ret) return ret; diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c index e1f2c9887f3f..6012f5e5b97c 100644 --- a/drivers/infiniband/core/verbs.c +++ b/drivers/infiniband/core/verbs.c @@ -344,8 +344,8 @@ int ib_init_ah_from_wc(struct ib_device *device, u8 port_num, ah_attr->ah_flags = IB_AH_GRH; ah_attr->grh.dgid = grh->sgid; - ret = ib_find_cached_gid(device, &grh->dgid, &port_num, - &gid_index); + ret = ib_find_cached_gid(device, &grh->dgid, + NULL, &port_num, &gid_index); if (ret) return ret; @@ -988,7 +988,8 @@ int ib_resolve_eth_l2_attrs(struct ib_qp *qp, if ((*qp_attr_mask & IB_QP_AV) && (rdma_cap_eth_ah(qp->device, qp_attr->ah_attr.port_num))) { ret = ib_query_gid(qp->device, qp_attr->ah_attr.port_num, - qp_attr->ah_attr.grh.sgid_index, &sgid); + qp_attr->ah_attr.grh.sgid_index, &sgid, + NULL); if (ret) goto out; if (rdma_link_local_addr((struct in6_addr *)qp_attr->ah_attr.grh.dgid.raw)) { diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c index 8779d26d1e61..f63d5427bfc0 100644 --- a/drivers/infiniband/hw/mlx4/main.c +++ b/drivers/infiniband/hw/mlx4/main.c @@ -335,7 +335,7 @@ int mlx4_ib_gid_index_to_real_index(struct mlx4_ib_dev *ibdev, if (!rdma_cap_roce_gid_table(&ibdev->ib_dev, port_num)) return index; - ret = ib_get_cached_gid(&ibdev->ib_dev, port_num, index, &gid); + ret = ib_get_cached_gid(&ibdev->ib_dev, port_num, index, &gid, NULL); if (ret) return ret; @@ -756,7 +756,7 @@ static int mlx4_ib_query_gid(struct ib_device *ibdev, u8 port, int index, if (!rdma_cap_roce_gid_table(ibdev, port)) return -ENODEV; - ret = ib_get_cached_gid(ibdev, port, index, gid); + ret = ib_get_cached_gid(ibdev, port, index, gid, NULL); if (ret == -EAGAIN) { memcpy(gid, &zgid, sizeof(*gid)); return 0; diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c index ba25a1bfa52f..b52c9e1cdf1a 100644 --- a/drivers/infiniband/hw/mlx4/qp.c +++ b/drivers/infiniband/hw/mlx4/qp.c @@ -2269,7 +2269,8 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_send_wr *wr, } else { err = ib_get_cached_gid(ib_dev, be32_to_cpu(ah->av.ib.port_pd) >> 24, - ah->av.ib.gid_index, &sgid); + ah->av.ib.gid_index, &sgid, + NULL); if (err) return err; } @@ -2311,7 +2312,7 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_send_wr *wr, ib_get_cached_gid(ib_dev, be32_to_cpu(ah->av.ib.port_pd) >> 24, ah->av.ib.gid_index, - &sqp->ud_header.grh.source_gid); + &sqp->ud_header.grh.source_gid, NULL); } memcpy(sqp->ud_header.grh.destination_gid.raw, ah->av.ib.dgid, 16); diff --git a/drivers/infiniband/hw/mthca/mthca_av.c b/drivers/infiniband/hw/mthca/mthca_av.c index 32f6c6315454..bcac294042f5 100644 --- a/drivers/infiniband/hw/mthca/mthca_av.c +++ b/drivers/infiniband/hw/mthca/mthca_av.c @@ -281,7 +281,7 @@ int mthca_read_ah(struct mthca_dev *dev, struct mthca_ah *ah, ib_get_cached_gid(&dev->ib_dev, be32_to_cpu(ah->av->port_pd) >> 24, ah->av->gid_index % dev->limits.gid_table_len, - &header->grh.source_gid); + &header->grh.source_gid, NULL); memcpy(header->grh.destination_gid.raw, ah->av->dgid, 16); } diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c b/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c index 1c4e83d5d153..9bb710a402cd 100644 --- a/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c +++ b/drivers/infiniband/hw/ocrdma/ocrdma_verbs.c @@ -73,7 +73,7 @@ int ocrdma_query_gid(struct ib_device *ibdev, u8 port, if (index >= OCRDMA_MAX_SGID) return -EINVAL; - ret = ib_get_cached_gid(ibdev, port, index, sgid); + ret = ib_get_cached_gid(ibdev, port, index, sgid, NULL); if (ret == -EAGAIN) { memcpy(sgid, &zgid, sizeof(*sgid)); return 0; diff --git a/drivers/infiniband/ulp/ipoib/ipoib_main.c b/drivers/infiniband/ulp/ipoib/ipoib_main.c index babba05d7a0e..cbb6721d0a65 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_main.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_main.c @@ -1860,7 +1860,7 @@ static struct net_device *ipoib_add_port(const char *format, priv->dev->broadcast[8] = priv->pkey >> 8; priv->dev->broadcast[9] = priv->pkey & 0xff; - result = ib_query_gid(hca, port, 0, &priv->local_gid); + result = ib_query_gid(hca, port, 0, &priv->local_gid, NULL); if (result) { printk(KERN_WARNING "%s: ib_query_gid port %d failed (ret = %d)\n", hca->name, port, result); diff --git a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c index d750a86042f3..d4b97614196c 100644 --- a/drivers/infiniband/ulp/ipoib/ipoib_multicast.c +++ b/drivers/infiniband/ulp/ipoib/ipoib_multicast.c @@ -561,7 +561,7 @@ void ipoib_mcast_join_task(struct work_struct *work) } priv->local_lid = port_attr.lid; - if (ib_query_gid(priv->ca, priv->port, 0, &priv->local_gid)) + if (ib_query_gid(priv->ca, priv->port, 0, &priv->local_gid, NULL)) ipoib_warn(priv, "ib_query_gid() failed\n"); else memcpy(priv->dev->dev_addr + 4, priv->local_gid.raw, sizeof (union ib_gid)); diff --git a/drivers/infiniband/ulp/srp/ib_srp.c b/drivers/infiniband/ulp/srp/ib_srp.c index b481490ad257..8ba887824d05 100644 --- a/drivers/infiniband/ulp/srp/ib_srp.c +++ b/drivers/infiniband/ulp/srp/ib_srp.c @@ -3213,7 +3213,7 @@ static ssize_t srp_create_target(struct device *dev, INIT_WORK(&target->tl_err_work, srp_tl_err_work); INIT_WORK(&target->remove_work, srp_remove_work); spin_lock_init(&target->lock); - ret = ib_query_gid(ibdev, host->port, 0, &target->sgid); + ret = ib_query_gid(ibdev, host->port, 0, &target->sgid, NULL); if (ret) goto out; diff --git a/drivers/infiniband/ulp/srpt/ib_srpt.c b/drivers/infiniband/ulp/srpt/ib_srpt.c index f6fe0414139b..a7ac77a02593 100644 --- a/drivers/infiniband/ulp/srpt/ib_srpt.c +++ b/drivers/infiniband/ulp/srpt/ib_srpt.c @@ -546,7 +546,8 @@ static int srpt_refresh_port(struct srpt_port *sport) sport->sm_lid = port_attr.sm_lid; sport->lid = port_attr.lid; - ret = ib_query_gid(sport->sdev->device, sport->port, 0, &sport->gid); + ret = ib_query_gid(sport->sdev->device, sport->port, 0, &sport->gid, + NULL); if (ret) goto err_query_port; diff --git a/include/rdma/ib_cache.h b/include/rdma/ib_cache.h index bd92130f4ac5..dcc9bed9b69a 100644 --- a/include/rdma/ib_cache.h +++ b/include/rdma/ib_cache.h @@ -43,6 +43,8 @@ * @port_num: The port number of the device to query. * @index: The index into the cached GID table to query. * @gid: The GID value found at the specified index. + * @attr: The GID attribute found at the specified index (only in RoCE). + * NULL means ignore (output parameter). * * ib_get_cached_gid() fetches the specified GID table entry stored in * the local software cache. @@ -50,13 +52,15 @@ int ib_get_cached_gid(struct ib_device *device, u8 port_num, int index, - union ib_gid *gid); + union ib_gid *gid, + struct ib_gid_attr *attr); /** * ib_find_cached_gid - Returns the port number and GID table index where * a specified GID value occurs. * @device: The device to query. * @gid: The GID value to search for. + * @ndev: In RoCE, the net device of the device. NULL means ignore. * @port_num: The port number of the device where the GID value was found. * @index: The index into the cached GID table where the GID was found. This * parameter may be NULL. @@ -64,10 +68,11 @@ int ib_get_cached_gid(struct ib_device *device, * ib_find_cached_gid() searches for the specified GID value in * the local software cache. */ -int ib_find_cached_gid(struct ib_device *device, +int ib_find_cached_gid(struct ib_device *device, const union ib_gid *gid, - u8 *port_num, - u16 *index); + struct net_device *ndev, + u8 *port_num, + u16 *index); /** * ib_get_cached_pkey - Returns a cached PKey table entry diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index e1f65e204d37..98ded0b749cd 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -2177,7 +2177,8 @@ static inline bool rdma_cap_roce_gid_table(const struct ib_device *device, } int ib_query_gid(struct ib_device *device, - u8 port_num, int index, union ib_gid *gid); + u8 port_num, int index, union ib_gid *gid, + struct ib_gid_attr *attr); int ib_query_pkey(struct ib_device *device, u8 port_num, u16 index, u16 *pkey); @@ -2191,7 +2192,7 @@ int ib_modify_port(struct ib_device *device, struct ib_port_modify *port_modify); int ib_find_gid(struct ib_device *device, union ib_gid *gid, - u8 *port_num, u16 *index); + struct net_device *ndev, u8 *port_num, u16 *index); int ib_find_pkey(struct ib_device *device, u8 port_num, u16 pkey, u16 *index); -- cgit v1.2.3 From d300ec528b799ca87935b3667f5563f397f00f85 Mon Sep 17 00:00:00 2001 From: Matan Barak Date: Thu, 15 Oct 2015 18:38:46 +0300 Subject: IB/core: Expose and rename ib_find_cached_gid_by_port cache API Sometime consumers might want to search for a GID in a specific port. For example, when a WC arrives and we want to search the GID that matches that port - it's better to search only the relevant port. Exposing and renaming ib_cache_gid_find_by_port in order to match the naming convention of the module. Signed-off-by: Matan Barak Signed-off-by: Doug Ledford --- drivers/infiniband/core/cache.c | 9 +++++---- drivers/infiniband/core/core_priv.h | 5 ----- drivers/infiniband/core/device.c | 4 ++-- include/rdma/ib_cache.h | 19 +++++++++++++++++++ 4 files changed, 26 insertions(+), 11 deletions(-) (limited to 'include') diff --git a/drivers/infiniband/core/cache.c b/drivers/infiniband/core/cache.c index 5c054072ef20..639a7266bfaf 100644 --- a/drivers/infiniband/core/cache.c +++ b/drivers/infiniband/core/cache.c @@ -409,10 +409,10 @@ static int ib_cache_gid_find(struct ib_device *ib_dev, mask, port, index); } -int ib_cache_gid_find_by_port(struct ib_device *ib_dev, - const union ib_gid *gid, - u8 port, struct net_device *ndev, - u16 *index) +int ib_find_cached_gid_by_port(struct ib_device *ib_dev, + const union ib_gid *gid, + u8 port, struct net_device *ndev, + u16 *index) { int local_index; struct ib_gid_table **ports_table = ib_dev->cache.gid_cache; @@ -438,6 +438,7 @@ int ib_cache_gid_find_by_port(struct ib_device *ib_dev, return -ENOENT; } +EXPORT_SYMBOL(ib_find_cached_gid_by_port); static struct ib_gid_table *alloc_gid_table(int sz) { diff --git a/drivers/infiniband/core/core_priv.h b/drivers/infiniband/core/core_priv.h index 70bb36ebb03b..0df82b1f84c1 100644 --- a/drivers/infiniband/core/core_priv.h +++ b/drivers/infiniband/core/core_priv.h @@ -65,11 +65,6 @@ void ib_enum_all_roce_netdevs(roce_netdev_filter filter, roce_netdev_callback cb, void *cookie); -int ib_cache_gid_find_by_port(struct ib_device *ib_dev, - const union ib_gid *gid, - u8 port, struct net_device *ndev, - u16 *index); - enum ib_cache_gid_default_mode { IB_CACHE_GID_DEFAULT_MODE_SET, IB_CACHE_GID_DEFAULT_MODE_DELETE diff --git a/drivers/infiniband/core/device.c b/drivers/infiniband/core/device.c index f22ce487fd3a..179e8134d57f 100644 --- a/drivers/infiniband/core/device.c +++ b/drivers/infiniband/core/device.c @@ -838,8 +838,8 @@ int ib_find_gid(struct ib_device *device, union ib_gid *gid, for (port = rdma_start_port(device); port <= rdma_end_port(device); ++port) { if (rdma_cap_roce_gid_table(device, port)) { - if (!ib_cache_gid_find_by_port(device, gid, port, - ndev, index)) { + if (!ib_find_cached_gid_by_port(device, gid, port, + ndev, index)) { *port_num = port; return 0; } diff --git a/include/rdma/ib_cache.h b/include/rdma/ib_cache.h index dcc9bed9b69a..679d7ca6a3ee 100644 --- a/include/rdma/ib_cache.h +++ b/include/rdma/ib_cache.h @@ -74,6 +74,25 @@ int ib_find_cached_gid(struct ib_device *device, u8 *port_num, u16 *index); +/** + * ib_find_cached_gid_by_port - Returns the GID table index where a specified + * GID value occurs + * @device: The device to query. + * @gid: The GID value to search for. + * @port_num: The port number of the device where the GID value sould be + * searched. + * @ndev: In RoCE, the net device of the device. Null means ignore. + * @index: The index into the cached GID table where the GID was found. This + * parameter may be NULL. + * + * ib_find_cached_gid() searches for the specified GID value in + * the local software cache. + */ +int ib_find_cached_gid_by_port(struct ib_device *device, + const union ib_gid *gid, + u8 port_num, + struct net_device *ndev, + u16 *index); /** * ib_get_cached_pkey - Returns a cached PKey table entry * @device: The device to query. -- cgit v1.2.3 From ba36e37fd3ca3dc8f215b14bcfdccf9f41b65767 Mon Sep 17 00:00:00 2001 From: Matan Barak Date: Thu, 15 Oct 2015 18:38:47 +0300 Subject: IB/core: Add netdev to path record In order to find the sgid_index, one could just query the IB cache with the correct GID and netdevice. Therefore, instead of storing the L2 attributes directly in the path, we only store the ifindex and net and use them later to get the sgid_index. The vlan_id and smac L2 attributes are removed in a later patch. Signed-off-by: Matan Barak Reviewed-By: Devesh Sharma Signed-off-by: Doug Ledford --- drivers/infiniband/core/sa_query.c | 13 +++++++++++-- drivers/infiniband/core/uverbs_marshall.c | 2 ++ include/rdma/ib_sa.h | 10 ++++++++++ 3 files changed, 23 insertions(+), 2 deletions(-) (limited to 'include') diff --git a/drivers/infiniband/core/sa_query.c b/drivers/infiniband/core/sa_query.c index 9a4e7891ada2..c9d9d7afcdb4 100644 --- a/drivers/infiniband/core/sa_query.c +++ b/drivers/infiniband/core/sa_query.c @@ -1007,18 +1007,25 @@ int ib_init_ah_from_path(struct ib_device *device, u8 port_num, force_grh = rdma_cap_eth_ah(device, port_num); if (rec->hop_limit > 1 || force_grh) { + struct net_device *ndev = ib_get_ndev_from_path(rec); + ah_attr->ah_flags = IB_AH_GRH; ah_attr->grh.dgid = rec->dgid; - ret = ib_find_cached_gid(device, &rec->sgid, NULL, &port_num, + ret = ib_find_cached_gid(device, &rec->sgid, ndev, &port_num, &gid_index); - if (ret) + if (ret) { + if (ndev) + dev_put(ndev); return ret; + } ah_attr->grh.sgid_index = gid_index; ah_attr->grh.flow_label = be32_to_cpu(rec->flow_label); ah_attr->grh.hop_limit = rec->hop_limit; ah_attr->grh.traffic_class = rec->traffic_class; + if (ndev) + dev_put(ndev); } if (force_grh) { memcpy(ah_attr->dmac, rec->dmac, ETH_ALEN); @@ -1151,6 +1158,8 @@ static void ib_sa_path_rec_callback(struct ib_sa_query *sa_query, ib_unpack(path_rec_table, ARRAY_SIZE(path_rec_table), mad->data, &rec); rec.vlan_id = 0xffff; + rec.net = NULL; + rec.ifindex = 0; memset(rec.dmac, 0, ETH_ALEN); memset(rec.smac, 0, ETH_ALEN); query->callback(status, &rec, query->context); diff --git a/drivers/infiniband/core/uverbs_marshall.c b/drivers/infiniband/core/uverbs_marshall.c index abd97247443e..484698c94d7f 100644 --- a/drivers/infiniband/core/uverbs_marshall.c +++ b/drivers/infiniband/core/uverbs_marshall.c @@ -144,5 +144,7 @@ void ib_copy_path_rec_from_user(struct ib_sa_path_rec *dst, memset(dst->smac, 0, sizeof(dst->smac)); memset(dst->dmac, 0, sizeof(dst->dmac)); dst->vlan_id = 0xffff; + dst->net = NULL; + dst->ifindex = 0; } EXPORT_SYMBOL(ib_copy_path_rec_from_user); diff --git a/include/rdma/ib_sa.h b/include/rdma/ib_sa.h index 7e071a6abb34..406ecf177f21 100644 --- a/include/rdma/ib_sa.h +++ b/include/rdma/ib_sa.h @@ -39,6 +39,7 @@ #include #include +#include #include #include @@ -157,8 +158,17 @@ struct ib_sa_path_rec { u8 smac[ETH_ALEN]; u8 dmac[ETH_ALEN]; u16 vlan_id; + /* ignored in IB */ + int ifindex; + /* ignored in IB */ + struct net *net; }; +static inline struct net_device *ib_get_ndev_from_path(struct ib_sa_path_rec *rec) +{ + return rec->net ? dev_get_by_index(rec->net, rec->ifindex) : NULL; +} + #define IB_SA_MCMEMBER_REC_MGID IB_SA_COMP_MASK( 0) #define IB_SA_MCMEMBER_REC_PORT_GID IB_SA_COMP_MASK( 1) #define IB_SA_MCMEMBER_REC_QKEY IB_SA_COMP_MASK( 2) -- cgit v1.2.3 From 99b27e3b5da0871cb43980960fb14ff625adffad Mon Sep 17 00:00:00 2001 From: Matan Barak Date: Thu, 15 Oct 2015 18:38:50 +0300 Subject: IB/cache: Add ib_find_gid_by_filter cache API GID cache API users might want to search for GIDs with specific attributes rather than just specifying GID, net device and port. This is used in a later patch, where we find the sgid index by L2 Ethernet attributes. Signed-off-by: Matan Barak Reviewed-By: Devesh Sharma Signed-off-by: Doug Ledford --- drivers/infiniband/core/cache.c | 93 +++++++++++++++++++++++++++++++++++++++++ include/rdma/ib_cache.h | 8 ++++ 2 files changed, 101 insertions(+) (limited to 'include') diff --git a/drivers/infiniband/core/cache.c b/drivers/infiniband/core/cache.c index 639a7266bfaf..89bebeada38b 100644 --- a/drivers/infiniband/core/cache.c +++ b/drivers/infiniband/core/cache.c @@ -440,6 +440,81 @@ int ib_find_cached_gid_by_port(struct ib_device *ib_dev, } EXPORT_SYMBOL(ib_find_cached_gid_by_port); +/** + * ib_find_gid_by_filter - Returns the GID table index where a specified + * GID value occurs + * @device: The device to query. + * @gid: The GID value to search for. + * @port_num: The port number of the device where the GID value could be + * searched. + * @filter: The filter function is executed on any matching GID in the table. + * If the filter function returns true, the corresponding index is returned, + * otherwise, we continue searching the GID table. It's guaranteed that + * while filter is executed, ndev field is valid and the structure won't + * change. filter is executed in an atomic context. filter must not be NULL. + * @index: The index into the cached GID table where the GID was found. This + * parameter may be NULL. + * + * ib_cache_gid_find_by_filter() searches for the specified GID value + * of which the filter function returns true in the port's GID table. + * This function is only supported on RoCE ports. + * + */ +static int ib_cache_gid_find_by_filter(struct ib_device *ib_dev, + const union ib_gid *gid, + u8 port, + bool (*filter)(const union ib_gid *, + const struct ib_gid_attr *, + void *), + void *context, + u16 *index) +{ + struct ib_gid_table **ports_table = ib_dev->cache.gid_cache; + struct ib_gid_table *table; + unsigned int i; + bool found = false; + + if (!ports_table) + return -EOPNOTSUPP; + + if (port < rdma_start_port(ib_dev) || + port > rdma_end_port(ib_dev) || + !rdma_protocol_roce(ib_dev, port)) + return -EPROTONOSUPPORT; + + table = ports_table[port - rdma_start_port(ib_dev)]; + + for (i = 0; i < table->sz; i++) { + struct ib_gid_attr attr; + unsigned long flags; + + read_lock_irqsave(&table->data_vec[i].lock, flags); + if (table->data_vec[i].props & GID_TABLE_ENTRY_INVALID) + goto next; + + if (memcmp(gid, &table->data_vec[i].gid, sizeof(*gid))) + goto next; + + memcpy(&attr, &table->data_vec[i].attr, sizeof(attr)); + + if (filter(gid, &attr, context)) + found = true; + +next: + read_unlock_irqrestore(&table->data_vec[i].lock, flags); + + if (found) + break; + } + + if (!found) + return -ENOENT; + + if (index) + *index = i; + return 0; +} + static struct ib_gid_table *alloc_gid_table(int sz) { unsigned int i; @@ -670,6 +745,24 @@ int ib_find_cached_gid(struct ib_device *device, } EXPORT_SYMBOL(ib_find_cached_gid); +int ib_find_gid_by_filter(struct ib_device *device, + const union ib_gid *gid, + u8 port_num, + bool (*filter)(const union ib_gid *gid, + const struct ib_gid_attr *, + void *), + void *context, u16 *index) +{ + /* Only RoCE GID table supports filter function */ + if (!rdma_cap_roce_gid_table(device, port_num) && filter) + return -EPROTONOSUPPORT; + + return ib_cache_gid_find_by_filter(device, gid, + port_num, filter, + context, index); +} +EXPORT_SYMBOL(ib_find_gid_by_filter); + int ib_get_cached_pkey(struct ib_device *device, u8 port_num, int index, diff --git a/include/rdma/ib_cache.h b/include/rdma/ib_cache.h index 679d7ca6a3ee..269a27cf0a46 100644 --- a/include/rdma/ib_cache.h +++ b/include/rdma/ib_cache.h @@ -93,6 +93,14 @@ int ib_find_cached_gid_by_port(struct ib_device *device, u8 port_num, struct net_device *ndev, u16 *index); + +int ib_find_gid_by_filter(struct ib_device *device, + const union ib_gid *gid, + u8 port_num, + bool (*filter)(const union ib_gid *gid, + const struct ib_gid_attr *, + void *), + void *context, u16 *index); /** * ib_get_cached_pkey - Returns a cached PKey table entry * @device: The device to query. -- cgit v1.2.3 From dbf727de7440f73c4b92be4b958cbc24977e8ca2 Mon Sep 17 00:00:00 2001 From: Matan Barak Date: Thu, 15 Oct 2015 18:38:51 +0300 Subject: IB/core: Use GID table in AH creation and dmac resolution Previously, vlan id and source MAC were used from QP attributes. Since the net device is now stored in the GID attributes, they could be used instead of getting this information from the QP attributes. IB_QP_SMAC, IB_QP_ALT_SMAC, IB_QP_VID and IB_QP_ALT_VID were removed because there is no known libibverbs that uses them. This commit also modifies the vendors (mlx4, ocrdma) drivers in order to use the new approach. ocrdma driver changes were done by Somnath Kotur Signed-off-by: Matan Barak Signed-off-by: Doug Ledford --- drivers/infiniband/core/addr.c | 3 +- drivers/infiniband/core/core_priv.h | 4 +- drivers/infiniband/core/uverbs_cmd.c | 2 +- drivers/infiniband/core/verbs.c | 160 ++++++++++++++++++------------- drivers/infiniband/hw/mlx4/ah.c | 17 +++- drivers/infiniband/hw/mlx4/mad.c | 12 ++- drivers/infiniband/hw/mlx4/mcg.c | 2 +- drivers/infiniband/hw/mlx4/mlx4_ib.h | 2 +- drivers/infiniband/hw/mlx4/qp.c | 47 +++++++-- drivers/infiniband/hw/ocrdma/ocrdma_ah.c | 22 +++-- drivers/infiniband/hw/ocrdma/ocrdma_hw.c | 30 +++--- include/rdma/ib_addr.h | 2 +- 12 files changed, 193 insertions(+), 110 deletions(-) (limited to 'include') diff --git a/drivers/infiniband/core/addr.c b/drivers/infiniband/core/addr.c index 746cdf56bc76..d3c42b3c1b51 100644 --- a/drivers/infiniband/core/addr.c +++ b/drivers/infiniband/core/addr.c @@ -458,7 +458,7 @@ static void resolve_cb(int status, struct sockaddr *src_addr, } int rdma_addr_find_dmac_by_grh(const union ib_gid *sgid, const union ib_gid *dgid, - u8 *dmac, u16 *vlan_id) + u8 *dmac, u16 *vlan_id, int if_index) { int ret = 0; struct rdma_dev_addr dev_addr; @@ -476,6 +476,7 @@ int rdma_addr_find_dmac_by_grh(const union ib_gid *sgid, const union ib_gid *dgi rdma_gid2ip(&dgid_addr._sockaddr, dgid); memset(&dev_addr, 0, sizeof(dev_addr)); + dev_addr.bound_dev_if = if_index; ctx.addr = &dev_addr; init_completion(&ctx.comp); diff --git a/drivers/infiniband/core/core_priv.h b/drivers/infiniband/core/core_priv.h index 0df82b1f84c1..5cf6eb716f00 100644 --- a/drivers/infiniband/core/core_priv.h +++ b/drivers/infiniband/core/core_priv.h @@ -46,8 +46,8 @@ void ib_device_unregister_sysfs(struct ib_device *device); void ib_cache_setup(void); void ib_cache_cleanup(void); -int ib_resolve_eth_l2_attrs(struct ib_qp *qp, - struct ib_qp_attr *qp_attr, int *qp_attr_mask); +int ib_resolve_eth_dmac(struct ib_qp *qp, + struct ib_qp_attr *qp_attr, int *qp_attr_mask); typedef void (*roce_netdev_callback)(struct ib_device *device, u8 port, struct net_device *idev, void *cookie); diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c index ece9f4c5e4a4..74dbfd782e19 100644 --- a/drivers/infiniband/core/uverbs_cmd.c +++ b/drivers/infiniband/core/uverbs_cmd.c @@ -2345,7 +2345,7 @@ ssize_t ib_uverbs_modify_qp(struct ib_uverbs_file *file, attr->alt_ah_attr.port_num = cmd.alt_dest.port_num; if (qp->real_qp == qp) { - ret = ib_resolve_eth_l2_attrs(qp, attr, &cmd.attr_mask); + ret = ib_resolve_eth_dmac(qp, attr, &cmd.attr_mask); if (ret) goto release_qp; ret = qp->device->modify_qp(qp, attr, diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c index 6012f5e5b97c..46d97f09fbae 100644 --- a/drivers/infiniband/core/verbs.c +++ b/drivers/infiniband/core/verbs.c @@ -41,6 +41,9 @@ #include #include #include +#include +#include +#include #include #include @@ -308,6 +311,35 @@ struct ib_ah *ib_create_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr) } EXPORT_SYMBOL(ib_create_ah); +struct find_gid_index_context { + u16 vlan_id; +}; + +static bool find_gid_index(const union ib_gid *gid, + const struct ib_gid_attr *gid_attr, + void *context) +{ + struct find_gid_index_context *ctx = + (struct find_gid_index_context *)context; + + if ((!!(ctx->vlan_id != 0xffff) == !is_vlan_dev(gid_attr->ndev)) || + (is_vlan_dev(gid_attr->ndev) && + vlan_dev_vlan_id(gid_attr->ndev) != ctx->vlan_id)) + return false; + + return true; +} + +static int get_sgid_index_from_eth(struct ib_device *device, u8 port_num, + u16 vlan_id, const union ib_gid *sgid, + u16 *gid_index) +{ + struct find_gid_index_context context = {.vlan_id = vlan_id}; + + return ib_find_gid_by_filter(device, sgid, port_num, find_gid_index, + &context, gid_index); +} + int ib_init_ah_from_wc(struct ib_device *device, u8 port_num, const struct ib_wc *wc, const struct ib_grh *grh, struct ib_ah_attr *ah_attr) @@ -318,21 +350,30 @@ int ib_init_ah_from_wc(struct ib_device *device, u8 port_num, memset(ah_attr, 0, sizeof *ah_attr); if (rdma_cap_eth_ah(device, port_num)) { + u16 vlan_id = wc->wc_flags & IB_WC_WITH_VLAN ? + wc->vlan_id : 0xffff; + if (!(wc->wc_flags & IB_WC_GRH)) return -EPROTOTYPE; - if (wc->wc_flags & IB_WC_WITH_SMAC && - wc->wc_flags & IB_WC_WITH_VLAN) { - memcpy(ah_attr->dmac, wc->smac, ETH_ALEN); - ah_attr->vlan_id = wc->vlan_id; - } else { + if (!(wc->wc_flags & IB_WC_WITH_SMAC) || + !(wc->wc_flags & IB_WC_WITH_VLAN)) { ret = rdma_addr_find_dmac_by_grh(&grh->dgid, &grh->sgid, - ah_attr->dmac, &ah_attr->vlan_id); + ah_attr->dmac, + wc->wc_flags & IB_WC_WITH_VLAN ? + NULL : &vlan_id, + 0); if (ret) return ret; } - } else { - ah_attr->vlan_id = 0xffff; + + ret = get_sgid_index_from_eth(device, port_num, vlan_id, + &grh->dgid, &gid_index); + if (ret) + return ret; + + if (wc->wc_flags & IB_WC_WITH_SMAC) + memcpy(ah_attr->dmac, wc->smac, ETH_ALEN); } ah_attr->dlid = wc->slid; @@ -344,10 +385,13 @@ int ib_init_ah_from_wc(struct ib_device *device, u8 port_num, ah_attr->ah_flags = IB_AH_GRH; ah_attr->grh.dgid = grh->sgid; - ret = ib_find_cached_gid(device, &grh->dgid, - NULL, &port_num, &gid_index); - if (ret) - return ret; + if (!rdma_cap_eth_ah(device, port_num)) { + ret = ib_find_cached_gid_by_port(device, &grh->dgid, + port_num, NULL, + &gid_index); + if (ret) + return ret; + } ah_attr->grh.sgid_index = (u8) gid_index; flow_class = be32_to_cpu(grh->version_tclass_flow); @@ -617,9 +661,7 @@ EXPORT_SYMBOL(ib_create_qp); static const struct { int valid; enum ib_qp_attr_mask req_param[IB_QPT_MAX]; - enum ib_qp_attr_mask req_param_add_eth[IB_QPT_MAX]; enum ib_qp_attr_mask opt_param[IB_QPT_MAX]; - enum ib_qp_attr_mask opt_param_add_eth[IB_QPT_MAX]; } qp_state_table[IB_QPS_ERR + 1][IB_QPS_ERR + 1] = { [IB_QPS_RESET] = { [IB_QPS_RESET] = { .valid = 1 }, @@ -700,12 +742,6 @@ static const struct { IB_QP_MAX_DEST_RD_ATOMIC | IB_QP_MIN_RNR_TIMER), }, - .req_param_add_eth = { - [IB_QPT_RC] = (IB_QP_SMAC), - [IB_QPT_UC] = (IB_QP_SMAC), - [IB_QPT_XRC_INI] = (IB_QP_SMAC), - [IB_QPT_XRC_TGT] = (IB_QP_SMAC) - }, .opt_param = { [IB_QPT_UD] = (IB_QP_PKEY_INDEX | IB_QP_QKEY), @@ -726,21 +762,7 @@ static const struct { [IB_QPT_GSI] = (IB_QP_PKEY_INDEX | IB_QP_QKEY), }, - .opt_param_add_eth = { - [IB_QPT_RC] = (IB_QP_ALT_SMAC | - IB_QP_VID | - IB_QP_ALT_VID), - [IB_QPT_UC] = (IB_QP_ALT_SMAC | - IB_QP_VID | - IB_QP_ALT_VID), - [IB_QPT_XRC_INI] = (IB_QP_ALT_SMAC | - IB_QP_VID | - IB_QP_ALT_VID), - [IB_QPT_XRC_TGT] = (IB_QP_ALT_SMAC | - IB_QP_VID | - IB_QP_ALT_VID) - } - } + }, }, [IB_QPS_RTR] = { [IB_QPS_RESET] = { .valid = 1 }, @@ -962,13 +984,6 @@ int ib_modify_qp_is_ok(enum ib_qp_state cur_state, enum ib_qp_state next_state, req_param = qp_state_table[cur_state][next_state].req_param[type]; opt_param = qp_state_table[cur_state][next_state].opt_param[type]; - if (ll == IB_LINK_LAYER_ETHERNET) { - req_param |= qp_state_table[cur_state][next_state]. - req_param_add_eth[type]; - opt_param |= qp_state_table[cur_state][next_state]. - opt_param_add_eth[type]; - } - if ((mask & req_param) != req_param) return 0; @@ -979,41 +994,52 @@ int ib_modify_qp_is_ok(enum ib_qp_state cur_state, enum ib_qp_state next_state, } EXPORT_SYMBOL(ib_modify_qp_is_ok); -int ib_resolve_eth_l2_attrs(struct ib_qp *qp, - struct ib_qp_attr *qp_attr, int *qp_attr_mask) +int ib_resolve_eth_dmac(struct ib_qp *qp, + struct ib_qp_attr *qp_attr, int *qp_attr_mask) { int ret = 0; - union ib_gid sgid; - if ((*qp_attr_mask & IB_QP_AV) && - (rdma_cap_eth_ah(qp->device, qp_attr->ah_attr.port_num))) { - ret = ib_query_gid(qp->device, qp_attr->ah_attr.port_num, - qp_attr->ah_attr.grh.sgid_index, &sgid, - NULL); - if (ret) - goto out; + if (*qp_attr_mask & IB_QP_AV) { + if (qp_attr->ah_attr.port_num < rdma_start_port(qp->device) || + qp_attr->ah_attr.port_num > rdma_end_port(qp->device)) + return -EINVAL; + + if (!rdma_cap_eth_ah(qp->device, qp_attr->ah_attr.port_num)) + return 0; + if (rdma_link_local_addr((struct in6_addr *)qp_attr->ah_attr.grh.dgid.raw)) { - rdma_get_ll_mac((struct in6_addr *)qp_attr->ah_attr.grh.dgid.raw, qp_attr->ah_attr.dmac); - rdma_get_ll_mac((struct in6_addr *)sgid.raw, qp_attr->smac); - if (!(*qp_attr_mask & IB_QP_VID)) - qp_attr->vlan_id = rdma_get_vlan_id(&sgid); + rdma_get_ll_mac((struct in6_addr *)qp_attr->ah_attr.grh.dgid.raw, + qp_attr->ah_attr.dmac); } else { - ret = rdma_addr_find_dmac_by_grh(&sgid, &qp_attr->ah_attr.grh.dgid, - qp_attr->ah_attr.dmac, &qp_attr->vlan_id); - if (ret) - goto out; - ret = rdma_addr_find_smac_by_sgid(&sgid, qp_attr->smac, NULL); - if (ret) + union ib_gid sgid; + struct ib_gid_attr sgid_attr; + int ifindex; + + ret = ib_query_gid(qp->device, + qp_attr->ah_attr.port_num, + qp_attr->ah_attr.grh.sgid_index, + &sgid, &sgid_attr); + + if (ret || !sgid_attr.ndev) { + if (!ret) + ret = -ENXIO; goto out; + } + + ifindex = sgid_attr.ndev->ifindex; + + ret = rdma_addr_find_dmac_by_grh(&sgid, + &qp_attr->ah_attr.grh.dgid, + qp_attr->ah_attr.dmac, + NULL, ifindex); + + dev_put(sgid_attr.ndev); } - *qp_attr_mask |= IB_QP_SMAC; - if (qp_attr->vlan_id < 0xFFFF) - *qp_attr_mask |= IB_QP_VID; } out: return ret; } -EXPORT_SYMBOL(ib_resolve_eth_l2_attrs); +EXPORT_SYMBOL(ib_resolve_eth_dmac); int ib_modify_qp(struct ib_qp *qp, @@ -1022,7 +1048,7 @@ int ib_modify_qp(struct ib_qp *qp, { int ret; - ret = ib_resolve_eth_l2_attrs(qp, qp_attr, &qp_attr_mask); + ret = ib_resolve_eth_dmac(qp, qp_attr, &qp_attr_mask); if (ret) return ret; diff --git a/drivers/infiniband/hw/mlx4/ah.c b/drivers/infiniband/hw/mlx4/ah.c index 1688a17de4fe..86af71351d9a 100644 --- a/drivers/infiniband/hw/mlx4/ah.c +++ b/drivers/infiniband/hw/mlx4/ah.c @@ -76,7 +76,10 @@ static struct ib_ah *create_iboe_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr struct mlx4_dev *dev = ibdev->dev; int is_mcast = 0; struct in6_addr in6; - u16 vlan_tag; + u16 vlan_tag = 0xffff; + union ib_gid sgid; + struct ib_gid_attr gid_attr; + int ret; memcpy(&in6, ah_attr->grh.dgid.raw, sizeof(in6)); if (rdma_is_multicast_addr(&in6)) { @@ -85,7 +88,17 @@ static struct ib_ah *create_iboe_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr } else { memcpy(ah->av.eth.mac, ah_attr->dmac, ETH_ALEN); } - vlan_tag = ah_attr->vlan_id; + ret = ib_get_cached_gid(pd->device, ah_attr->port_num, + ah_attr->grh.sgid_index, &sgid, &gid_attr); + if (ret) + return ERR_PTR(ret); + memset(ah->av.eth.s_mac, 0, ETH_ALEN); + if (gid_attr.ndev) { + if (is_vlan_dev(gid_attr.ndev)) + vlan_tag = vlan_dev_vlan_id(gid_attr.ndev); + memcpy(ah->av.eth.s_mac, gid_attr.ndev->dev_addr, ETH_ALEN); + dev_put(gid_attr.ndev); + } if (vlan_tag < 0x1000) vlan_tag |= (ah_attr->sl & 7) << 13; ah->av.eth.port_pd = cpu_to_be32(to_mpd(pd)->pdn | (ah_attr->port_num << 24)); diff --git a/drivers/infiniband/hw/mlx4/mad.c b/drivers/infiniband/hw/mlx4/mad.c index 68f256716785..67c7120f0d52 100644 --- a/drivers/infiniband/hw/mlx4/mad.c +++ b/drivers/infiniband/hw/mlx4/mad.c @@ -1183,7 +1183,7 @@ static int is_proxy_qp0(struct mlx4_ib_dev *dev, int qpn, int slave) int mlx4_ib_send_to_wire(struct mlx4_ib_dev *dev, int slave, u8 port, enum ib_qp_type dest_qpt, u16 pkey_index, u32 remote_qpn, u32 qkey, struct ib_ah_attr *attr, - u8 *s_mac, struct ib_mad *mad) + u8 *s_mac, u16 vlan_id, struct ib_mad *mad) { struct ib_sge list; struct ib_send_wr wr, *bad_wr; @@ -1270,6 +1270,9 @@ int mlx4_ib_send_to_wire(struct mlx4_ib_dev *dev, int slave, u8 port, wr.send_flags = IB_SEND_SIGNALED; if (s_mac) memcpy(to_mah(ah)->av.eth.s_mac, s_mac, 6); + if (vlan_id < 0x1000) + vlan_id |= (attr->sl & 7) << 13; + to_mah(ah)->av.eth.vlan = cpu_to_be16(vlan_id); ret = ib_post_send(send_qp, &wr, &bad_wr); @@ -1306,6 +1309,7 @@ static void mlx4_ib_multiplex_mad(struct mlx4_ib_demux_pv_ctx *ctx, struct ib_wc u8 *slave_id; int slave; int port; + u16 vlan_id; /* Get slave that sent this packet */ if (wc->src_qp < dev->dev->phys_caps.base_proxy_sqpn || @@ -1394,10 +1398,10 @@ static void mlx4_ib_multiplex_mad(struct mlx4_ib_demux_pv_ctx *ctx, struct ib_wc fill_in_real_sgid_index(dev, slave, ctx->port, &ah_attr); memcpy(ah_attr.dmac, tunnel->hdr.mac, 6); - ah_attr.vlan_id = be16_to_cpu(tunnel->hdr.vlan); + vlan_id = be16_to_cpu(tunnel->hdr.vlan); /* if slave have default vlan use it */ mlx4_get_slave_default_vlan(dev->dev, ctx->port, slave, - &ah_attr.vlan_id, &ah_attr.sl); + &vlan_id, &ah_attr.sl); mlx4_ib_send_to_wire(dev, slave, ctx->port, is_proxy_qp0(dev, wc->src_qp, slave) ? @@ -1405,7 +1409,7 @@ static void mlx4_ib_multiplex_mad(struct mlx4_ib_demux_pv_ctx *ctx, struct ib_wc be16_to_cpu(tunnel->hdr.pkey_index), be32_to_cpu(tunnel->hdr.remote_qpn), be32_to_cpu(tunnel->hdr.qkey), - &ah_attr, wc->smac, &tunnel->mad); + &ah_attr, wc->smac, vlan_id, &tunnel->mad); } static int mlx4_ib_alloc_pv_bufs(struct mlx4_ib_demux_pv_ctx *ctx, diff --git a/drivers/infiniband/hw/mlx4/mcg.c b/drivers/infiniband/hw/mlx4/mcg.c index 2d5bccd71fc6..99451d887266 100644 --- a/drivers/infiniband/hw/mlx4/mcg.c +++ b/drivers/infiniband/hw/mlx4/mcg.c @@ -222,7 +222,7 @@ static int send_mad_to_wire(struct mlx4_ib_demux_ctx *ctx, struct ib_mad *mad) spin_unlock_irqrestore(&dev->sm_lock, flags); return mlx4_ib_send_to_wire(dev, mlx4_master_func_num(dev->dev), ctx->port, IB_QPT_GSI, 0, 1, IB_QP1_QKEY, - &ah_attr, NULL, mad); + &ah_attr, NULL, 0xffff, mad); } static int send_mad_to_slave(int slave, struct mlx4_ib_demux_ctx *ctx, diff --git a/drivers/infiniband/hw/mlx4/mlx4_ib.h b/drivers/infiniband/hw/mlx4/mlx4_ib.h index 086416a863ae..28a99e08030e 100644 --- a/drivers/infiniband/hw/mlx4/mlx4_ib.h +++ b/drivers/infiniband/hw/mlx4/mlx4_ib.h @@ -821,7 +821,7 @@ int mlx4_ib_send_to_slave(struct mlx4_ib_dev *dev, int slave, u8 port, int mlx4_ib_send_to_wire(struct mlx4_ib_dev *dev, int slave, u8 port, enum ib_qp_type dest_qpt, u16 pkey_index, u32 remote_qpn, u32 qkey, struct ib_ah_attr *attr, u8 *s_mac, - struct ib_mad *mad); + u16 vlan_id, struct ib_mad *mad); __be64 mlx4_ib_get_new_demux_tid(struct mlx4_ib_demux_ctx *ctx); diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c index b52c9e1cdf1a..42a051b088e4 100644 --- a/drivers/infiniband/hw/mlx4/qp.c +++ b/drivers/infiniband/hw/mlx4/qp.c @@ -1409,11 +1409,12 @@ static int _mlx4_set_path(struct mlx4_ib_dev *dev, const struct ib_ah_attr *ah, static int mlx4_set_path(struct mlx4_ib_dev *dev, const struct ib_qp_attr *qp, enum ib_qp_attr_mask qp_attr_mask, struct mlx4_ib_qp *mqp, - struct mlx4_qp_path *path, u8 port) + struct mlx4_qp_path *path, u8 port, + u16 vlan_id, u8 *smac) { return _mlx4_set_path(dev, &qp->ah_attr, - mlx4_mac_to_u64((u8 *)qp->smac), - (qp_attr_mask & IB_QP_VID) ? qp->vlan_id : 0xffff, + mlx4_mac_to_u64(smac), + vlan_id, path, &mqp->pri, port); } @@ -1424,9 +1425,8 @@ static int mlx4_set_alt_path(struct mlx4_ib_dev *dev, struct mlx4_qp_path *path, u8 port) { return _mlx4_set_path(dev, &qp->alt_ah_attr, - mlx4_mac_to_u64((u8 *)qp->alt_smac), - (qp_attr_mask & IB_QP_ALT_VID) ? - qp->alt_vlan_id : 0xffff, + 0, + 0xffff, path, &mqp->alt, port); } @@ -1442,7 +1442,8 @@ static void update_mcg_macs(struct mlx4_ib_dev *dev, struct mlx4_ib_qp *qp) } } -static int handle_eth_ud_smac_index(struct mlx4_ib_dev *dev, struct mlx4_ib_qp *qp, u8 *smac, +static int handle_eth_ud_smac_index(struct mlx4_ib_dev *dev, + struct mlx4_ib_qp *qp, struct mlx4_qp_context *context) { u64 u64_mac; @@ -1635,9 +1636,33 @@ static int __mlx4_ib_modify_qp(struct ib_qp *ibqp, } if (attr_mask & IB_QP_AV) { + u8 port_num = mlx4_is_bonded(to_mdev(ibqp->device)->dev) ? 1 : + attr_mask & IB_QP_PORT ? attr->port_num : qp->port; + union ib_gid gid; + struct ib_gid_attr gid_attr; + u16 vlan = 0xffff; + u8 smac[ETH_ALEN]; + int status = 0; + + if (rdma_cap_eth_ah(&dev->ib_dev, port_num) && + attr->ah_attr.ah_flags & IB_AH_GRH) { + int index = attr->ah_attr.grh.sgid_index; + + status = ib_get_cached_gid(ibqp->device, port_num, + index, &gid, &gid_attr); + if (!status && !memcmp(&gid, &zgid, sizeof(gid))) + status = -ENOENT; + if (!status && gid_attr.ndev) { + vlan = rdma_vlan_dev_vlan_id(gid_attr.ndev); + memcpy(smac, gid_attr.ndev->dev_addr, ETH_ALEN); + dev_put(gid_attr.ndev); + } + } + if (status) + goto out; + if (mlx4_set_path(dev, attr, attr_mask, qp, &context->pri_path, - attr_mask & IB_QP_PORT ? - attr->port_num : qp->port)) + port_num, vlan, smac)) goto out; optpar |= (MLX4_QP_OPTPAR_PRIMARY_ADDR_PATH | @@ -1774,7 +1799,7 @@ static int __mlx4_ib_modify_qp(struct ib_qp *ibqp, if (qp->mlx4_ib_qp_type == MLX4_IB_QPT_UD || qp->mlx4_ib_qp_type == MLX4_IB_QPT_PROXY_GSI || qp->mlx4_ib_qp_type == MLX4_IB_QPT_TUN_GSI) { - err = handle_eth_ud_smac_index(dev, qp, (u8 *)attr->smac, context); + err = handle_eth_ud_smac_index(dev, qp, context); if (err) { err = -EINVAL; goto out; @@ -2271,6 +2296,8 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_send_wr *wr, be32_to_cpu(ah->av.ib.port_pd) >> 24, ah->av.ib.gid_index, &sgid, NULL); + if (!err && !memcmp(&sgid, &zgid, sizeof(sgid))) + err = -ENOENT; if (err) return err; } diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_ah.c b/drivers/infiniband/hw/ocrdma/ocrdma_ah.c index 44766fee1f4e..9820074be59d 100644 --- a/drivers/infiniband/hw/ocrdma/ocrdma_ah.c +++ b/drivers/infiniband/hw/ocrdma/ocrdma_ah.c @@ -45,6 +45,7 @@ #include #include +#include #include "ocrdma.h" #include "ocrdma_verbs.h" @@ -56,10 +57,9 @@ static inline int set_av_attr(struct ocrdma_dev *dev, struct ocrdma_ah *ah, struct ib_ah_attr *attr, union ib_gid *sgid, - int pdid, bool *isvlan) + int pdid, bool *isvlan, u16 vlan_tag) { int status = 0; - u16 vlan_tag; struct ocrdma_eth_vlan eth; struct ocrdma_grh grh; int eth_sz; @@ -68,7 +68,6 @@ static inline int set_av_attr(struct ocrdma_dev *dev, struct ocrdma_ah *ah, memset(&grh, 0, sizeof(grh)); /* VLAN */ - vlan_tag = attr->vlan_id; if (!vlan_tag || (vlan_tag > 0xFFF)) vlan_tag = dev->pvid; if (vlan_tag || dev->pfc_state) { @@ -115,9 +114,11 @@ static inline int set_av_attr(struct ocrdma_dev *dev, struct ocrdma_ah *ah, struct ib_ah *ocrdma_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr) { u32 *ahid_addr; - bool isvlan = false; int status; struct ocrdma_ah *ah; + bool isvlan = false; + u16 vlan_tag = 0xffff; + struct ib_gid_attr sgid_attr; struct ocrdma_pd *pd = get_ocrdma_pd(ibpd); struct ocrdma_dev *dev = get_ocrdma_dev(ibpd->device); union ib_gid sgid; @@ -135,18 +136,25 @@ struct ib_ah *ocrdma_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr) if (status) goto av_err; - status = ocrdma_query_gid(&dev->ibdev, 1, attr->grh.sgid_index, &sgid); + status = ib_get_cached_gid(&dev->ibdev, 1, attr->grh.sgid_index, &sgid, + &sgid_attr); if (status) { pr_err("%s(): Failed to query sgid, status = %d\n", __func__, status); goto av_conf_err; } + if (sgid_attr.ndev) { + if (is_vlan_dev(sgid_attr.ndev)) + vlan_tag = vlan_dev_vlan_id(sgid_attr.ndev); + dev_put(sgid_attr.ndev); + } if ((pd->uctx) && (!rdma_is_multicast_addr((struct in6_addr *)attr->grh.dgid.raw)) && (!rdma_link_local_addr((struct in6_addr *)attr->grh.dgid.raw))) { status = rdma_addr_find_dmac_by_grh(&sgid, &attr->grh.dgid, - attr->dmac, &attr->vlan_id); + attr->dmac, &vlan_tag, + sgid_attr.ndev->ifindex); if (status) { pr_err("%s(): Failed to resolve dmac from gid." "status = %d\n", __func__, status); @@ -154,7 +162,7 @@ struct ib_ah *ocrdma_create_ah(struct ib_pd *ibpd, struct ib_ah_attr *attr) } } - status = set_av_attr(dev, ah, attr, &sgid, pd->id, &isvlan); + status = set_av_attr(dev, ah, attr, &sgid, pd->id, &isvlan, vlan_tag); if (status) goto av_conf_err; diff --git a/drivers/infiniband/hw/ocrdma/ocrdma_hw.c b/drivers/infiniband/hw/ocrdma/ocrdma_hw.c index 9d9914299866..30f67bebffa3 100644 --- a/drivers/infiniband/hw/ocrdma/ocrdma_hw.c +++ b/drivers/infiniband/hw/ocrdma/ocrdma_hw.c @@ -47,6 +47,7 @@ #include #include +#include #include "ocrdma.h" #include "ocrdma_hw.h" @@ -2470,6 +2471,7 @@ static int ocrdma_set_av_params(struct ocrdma_qp *qp, int status; struct ib_ah_attr *ah_attr = &attrs->ah_attr; union ib_gid sgid, zgid; + struct ib_gid_attr sgid_attr; u32 vlan_id = 0xFFFF; u8 mac_addr[6]; struct ocrdma_dev *dev = get_ocrdma_dev(qp->ibqp.device); @@ -2488,10 +2490,14 @@ static int ocrdma_set_av_params(struct ocrdma_qp *qp, cmd->flags |= OCRDMA_QP_PARA_FLOW_LBL_VALID; memcpy(&cmd->params.dgid[0], &ah_attr->grh.dgid.raw[0], sizeof(cmd->params.dgid)); - status = ocrdma_query_gid(&dev->ibdev, 1, - ah_attr->grh.sgid_index, &sgid); - if (status) - return status; + + status = ib_get_cached_gid(&dev->ibdev, 1, ah_attr->grh.sgid_index, + &sgid, &sgid_attr); + if (!status && sgid_attr.ndev) { + vlan_id = rdma_vlan_dev_vlan_id(sgid_attr.ndev); + memcpy(mac_addr, sgid_attr.ndev->dev_addr, ETH_ALEN); + dev_put(sgid_attr.ndev); + } memset(&zgid, 0, sizeof(zgid)); if (!memcmp(&sgid, &zgid, sizeof(zgid))) @@ -2508,17 +2514,15 @@ static int ocrdma_set_av_params(struct ocrdma_qp *qp, ocrdma_cpu_to_le32(&cmd->params.dgid[0], sizeof(cmd->params.dgid)); ocrdma_cpu_to_le32(&cmd->params.sgid[0], sizeof(cmd->params.sgid)); cmd->params.vlan_dmac_b4_to_b5 = mac_addr[4] | (mac_addr[5] << 8); - if (attr_mask & IB_QP_VID) { - vlan_id = attrs->vlan_id; - } else if (dev->pfc_state) { - vlan_id = 0; - pr_err("ocrdma%d:Using VLAN with PFC is recommended\n", - dev->id); - pr_err("ocrdma%d:Using VLAN 0 for this connection\n", - dev->id); - } if (vlan_id < 0x1000) { + if (dev->pfc_state) { + vlan_id = 0; + pr_err("ocrdma%d:Using VLAN with PFC is recommended\n", + dev->id); + pr_err("ocrdma%d:Using VLAN 0 for this connection\n", + dev->id); + } cmd->params.vlan_dmac_b4_to_b5 |= vlan_id << OCRDMA_QP_PARAMS_VLAN_SHIFT; cmd->flags |= OCRDMA_QP_PARA_VLAN_EN_VALID; diff --git a/include/rdma/ib_addr.h b/include/rdma/ib_addr.h index fde33ac6b58a..17e4a8bdecab 100644 --- a/include/rdma/ib_addr.h +++ b/include/rdma/ib_addr.h @@ -112,7 +112,7 @@ int rdma_addr_size(struct sockaddr *addr); int rdma_addr_find_smac_by_sgid(union ib_gid *sgid, u8 *smac, u16 *vlan_id); int rdma_addr_find_dmac_by_grh(const union ib_gid *sgid, const union ib_gid *dgid, - u8 *smac, u16 *vlan_id); + u8 *smac, u16 *vlan_id, int if_index); static inline u16 ib_addr_get_pkey(struct rdma_dev_addr *dev_addr) { -- cgit v1.2.3 From aa744cc01fe0f21dfbe2744d3fd5f2fb3244c9b3 Mon Sep 17 00:00:00 2001 From: Matan Barak Date: Thu, 15 Oct 2015 18:38:53 +0300 Subject: IB/core: Remove smac and vlan id from qp_attr and ah_attr Smac and vlan id could be resolved from the GID attribute, and thus these attributes aren't needed anymore. Removing them. Signed-off-by: Matan Barak Reviewed-By: Devesh Sharma Signed-off-by: Doug Ledford --- drivers/infiniband/core/sa_query.c | 4 ---- drivers/infiniband/core/ucma.c | 1 - drivers/infiniband/core/uverbs_cmd.c | 1 - include/rdma/ib_verbs.h | 13 ++++--------- 4 files changed, 4 insertions(+), 15 deletions(-) (limited to 'include') diff --git a/drivers/infiniband/core/sa_query.c b/drivers/infiniband/core/sa_query.c index c9d9d7afcdb4..77f5afc297c6 100644 --- a/drivers/infiniband/core/sa_query.c +++ b/drivers/infiniband/core/sa_query.c @@ -1029,11 +1029,7 @@ int ib_init_ah_from_path(struct ib_device *device, u8 port_num, } if (force_grh) { memcpy(ah_attr->dmac, rec->dmac, ETH_ALEN); - ah_attr->vlan_id = rec->vlan_id; - } else { - ah_attr->vlan_id = 0xffff; } - return 0; } EXPORT_SYMBOL(ib_init_ah_from_path); diff --git a/drivers/infiniband/core/ucma.c b/drivers/infiniband/core/ucma.c index 30467d10df91..3a342dfaac15 100644 --- a/drivers/infiniband/core/ucma.c +++ b/drivers/infiniband/core/ucma.c @@ -1211,7 +1211,6 @@ static int ucma_set_ib_path(struct ucma_context *ctx, return -EINVAL; memset(&sa_path, 0, sizeof(sa_path)); - sa_path.vlan_id = 0xffff; ib_sa_unpack_path(path_data->path_rec, &sa_path); ret = rdma_set_ib_paths(ctx->cm_id, &sa_path, 1); diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c index 74dbfd782e19..8fd081ae9aa9 100644 --- a/drivers/infiniband/core/uverbs_cmd.c +++ b/drivers/infiniband/core/uverbs_cmd.c @@ -2822,7 +2822,6 @@ ssize_t ib_uverbs_create_ah(struct ib_uverbs_file *file, attr.grh.sgid_index = cmd.attr.grh.sgid_index; attr.grh.hop_limit = cmd.attr.grh.hop_limit; attr.grh.traffic_class = cmd.attr.grh.traffic_class; - attr.vlan_id = 0; memset(&attr.dmac, 0, sizeof(attr.dmac)); memcpy(attr.grh.dgid.raw, cmd.attr.grh.dgid, 16); diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index 98ded0b749cd..e4cc389c43cb 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -699,7 +699,6 @@ struct ib_ah_attr { u8 ah_flags; u8 port_num; u8 dmac[ETH_ALEN]; - u16 vlan_id; }; enum ib_wc_status { @@ -958,10 +957,10 @@ enum ib_qp_attr_mask { IB_QP_PATH_MIG_STATE = (1<<18), IB_QP_CAP = (1<<19), IB_QP_DEST_QPN = (1<<20), - IB_QP_SMAC = (1<<21), - IB_QP_ALT_SMAC = (1<<22), - IB_QP_VID = (1<<23), - IB_QP_ALT_VID = (1<<24), + IB_QP_RESERVED1 = (1<<21), + IB_QP_RESERVED2 = (1<<22), + IB_QP_RESERVED3 = (1<<23), + IB_QP_RESERVED4 = (1<<24), }; enum ib_qp_state { @@ -1011,10 +1010,6 @@ struct ib_qp_attr { u8 rnr_retry; u8 alt_port_num; u8 alt_timeout; - u8 smac[ETH_ALEN]; - u8 alt_smac[ETH_ALEN]; - u16 vlan_id; - u16 alt_vlan_id; }; enum ib_wr_opcode { -- cgit v1.2.3 From 10e07f13c06690488087f5d3f2c59a9728def339 Mon Sep 17 00:00:00 2001 From: Matan Barak Date: Thu, 15 Oct 2015 18:38:54 +0300 Subject: IB/core: Remove smac and vlan id from path record The GID cache accompanies every GID with attributes. The GID attributes link the GID with its netdevice, which could be resolved to smac and vlan id easily. Since we've added the netdevice (ifindex and net) to the path record, storing the L2 attributes is duplicated data and hence these attributes are removed. Signed-off-by: Matan Barak Reviewed-By: Devesh Sharma Signed-off-by: Doug Ledford --- drivers/infiniband/core/cma.c | 2 -- drivers/infiniband/core/sa_query.c | 2 -- drivers/infiniband/core/uverbs_marshall.c | 2 -- include/rdma/ib_sa.h | 2 -- 4 files changed, 8 deletions(-) (limited to 'include') diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index 09ddff8cc2ac..be41784a62e4 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -2307,9 +2307,7 @@ static int cma_resolve_iboe_route(struct rdma_id_private *id_priv) goto err2; } - route->path_rec->vlan_id = rdma_vlan_dev_vlan_id(ndev); memcpy(route->path_rec->dmac, addr->dev_addr.dst_dev_addr, ETH_ALEN); - memcpy(route->path_rec->smac, ndev->dev_addr, ndev->addr_len); rdma_ip2gid((struct sockaddr *)&id_priv->id.route.addr.src_addr, &route->path_rec->sgid); diff --git a/drivers/infiniband/core/sa_query.c b/drivers/infiniband/core/sa_query.c index 77f5afc297c6..dcdaa79e3f0f 100644 --- a/drivers/infiniband/core/sa_query.c +++ b/drivers/infiniband/core/sa_query.c @@ -1153,11 +1153,9 @@ static void ib_sa_path_rec_callback(struct ib_sa_query *sa_query, ib_unpack(path_rec_table, ARRAY_SIZE(path_rec_table), mad->data, &rec); - rec.vlan_id = 0xffff; rec.net = NULL; rec.ifindex = 0; memset(rec.dmac, 0, ETH_ALEN); - memset(rec.smac, 0, ETH_ALEN); query->callback(status, &rec, query->context); } else query->callback(status, NULL, query->context); diff --git a/drivers/infiniband/core/uverbs_marshall.c b/drivers/infiniband/core/uverbs_marshall.c index 484698c94d7f..7d2f14c9bbef 100644 --- a/drivers/infiniband/core/uverbs_marshall.c +++ b/drivers/infiniband/core/uverbs_marshall.c @@ -141,9 +141,7 @@ void ib_copy_path_rec_from_user(struct ib_sa_path_rec *dst, dst->preference = src->preference; dst->packet_life_time_selector = src->packet_life_time_selector; - memset(dst->smac, 0, sizeof(dst->smac)); memset(dst->dmac, 0, sizeof(dst->dmac)); - dst->vlan_id = 0xffff; dst->net = NULL; dst->ifindex = 0; } diff --git a/include/rdma/ib_sa.h b/include/rdma/ib_sa.h index 406ecf177f21..301969552d0a 100644 --- a/include/rdma/ib_sa.h +++ b/include/rdma/ib_sa.h @@ -155,9 +155,7 @@ struct ib_sa_path_rec { u8 packet_life_time_selector; u8 packet_life_time; u8 preference; - u8 smac[ETH_ALEN]; u8 dmac[ETH_ALEN]; - u16 vlan_id; /* ignored in IB */ int ifindex; /* ignored in IB */ -- cgit v1.2.3 From 565edd1d555513ab5d67a847d50d7c14c82ef6c3 Mon Sep 17 00:00:00 2001 From: Guy Shapiro Date: Thu, 22 Oct 2015 15:20:08 +0300 Subject: IB/addr: Pass network namespace as a parameter Add network namespace support to the ib_addr module. For that, all the address resolution and matching should be done using the appropriate namespace instead of init_net. This is achieved by: 1. Adding an explicit network namespace argument to exported function that require a namespace. 2. Saving the namespace in the rdma_addr_client structure. 3. Using it when calling networking functions. In order to preserve the behavior of calling modules, &init_net is passed as the parameter in calls from other modules. This is modified as namespace support is added on more levels. Signed-off-by: Haggai Eran Signed-off-by: Yotam Kenneth Signed-off-by: Shachar Raindel Signed-off-by: Guy Shapiro Signed-off-by: Doug Ledford --- drivers/infiniband/core/addr.c | 17 +++++++++-------- drivers/infiniband/core/cma.c | 1 + include/rdma/ib_addr.h | 16 +++++++++++++++- 3 files changed, 25 insertions(+), 9 deletions(-) (limited to 'include') diff --git a/drivers/infiniband/core/addr.c b/drivers/infiniband/core/addr.c index d3c42b3c1b51..34b1adad07aa 100644 --- a/drivers/infiniband/core/addr.c +++ b/drivers/infiniband/core/addr.c @@ -128,7 +128,7 @@ int rdma_translate_ip(struct sockaddr *addr, struct rdma_dev_addr *dev_addr, int ret = -EADDRNOTAVAIL; if (dev_addr->bound_dev_if) { - dev = dev_get_by_index(&init_net, dev_addr->bound_dev_if); + dev = dev_get_by_index(dev_addr->net, dev_addr->bound_dev_if); if (!dev) return -ENODEV; ret = rdma_copy_addr(dev_addr, dev, NULL); @@ -138,7 +138,7 @@ int rdma_translate_ip(struct sockaddr *addr, struct rdma_dev_addr *dev_addr, switch (addr->sa_family) { case AF_INET: - dev = ip_dev_find(&init_net, + dev = ip_dev_find(dev_addr->net, ((struct sockaddr_in *) addr)->sin_addr.s_addr); if (!dev) @@ -149,12 +149,11 @@ int rdma_translate_ip(struct sockaddr *addr, struct rdma_dev_addr *dev_addr, *vlan_id = rdma_vlan_dev_vlan_id(dev); dev_put(dev); break; - #if IS_ENABLED(CONFIG_IPV6) case AF_INET6: rcu_read_lock(); - for_each_netdev_rcu(&init_net, dev) { - if (ipv6_chk_addr(&init_net, + for_each_netdev_rcu(dev_addr->net, dev) { + if (ipv6_chk_addr(dev_addr->net, &((struct sockaddr_in6 *) addr)->sin6_addr, dev, 1)) { ret = rdma_copy_addr(dev_addr, dev, NULL); @@ -236,7 +235,7 @@ static int addr4_resolve(struct sockaddr_in *src_in, fl4.daddr = dst_ip; fl4.saddr = src_ip; fl4.flowi4_oif = addr->bound_dev_if; - rt = ip_route_output_key(&init_net, &fl4); + rt = ip_route_output_key(addr->net, &fl4); if (IS_ERR(rt)) { ret = PTR_ERR(rt); goto out; @@ -278,12 +277,12 @@ static int addr6_resolve(struct sockaddr_in6 *src_in, fl6.saddr = src_in->sin6_addr; fl6.flowi6_oif = addr->bound_dev_if; - dst = ip6_route_output(&init_net, NULL, &fl6); + dst = ip6_route_output(addr->net, NULL, &fl6); if ((ret = dst->error)) goto put; if (ipv6_addr_any(&fl6.saddr)) { - ret = ipv6_dev_get_saddr(&init_net, ip6_dst_idev(dst)->dev, + ret = ipv6_dev_get_saddr(addr->net, ip6_dst_idev(dst)->dev, &fl6.daddr, 0, &fl6.saddr); if (ret) goto put; @@ -477,6 +476,7 @@ int rdma_addr_find_dmac_by_grh(const union ib_gid *sgid, const union ib_gid *dgi memset(&dev_addr, 0, sizeof(dev_addr)); dev_addr.bound_dev_if = if_index; + dev_addr.net = &init_net; ctx.addr = &dev_addr; init_completion(&ctx.comp); @@ -511,6 +511,7 @@ int rdma_addr_find_smac_by_sgid(union ib_gid *sgid, u8 *smac, u16 *vlan_id) rdma_gid2ip(&gid_addr._sockaddr, sgid); memset(&dev_addr, 0, sizeof(dev_addr)); + dev_addr.net = &init_net; ret = rdma_translate_ip(&gid_addr._sockaddr, &dev_addr, vlan_id); if (ret) return ret; diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index be41784a62e4..80a8b469a9a8 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -610,6 +610,7 @@ struct rdma_cm_id *rdma_create_id(rdma_cm_event_handler event_handler, INIT_LIST_HEAD(&id_priv->listen_list); INIT_LIST_HEAD(&id_priv->mc_list); get_random_bytes(&id_priv->seq_num, sizeof id_priv->seq_num); + id_priv->id.route.addr.dev_addr.net = &init_net; return &id_priv->id; } diff --git a/include/rdma/ib_addr.h b/include/rdma/ib_addr.h index 17e4a8bdecab..11528591d0d7 100644 --- a/include/rdma/ib_addr.h +++ b/include/rdma/ib_addr.h @@ -47,6 +47,7 @@ #include #include #include +#include struct rdma_addr_client { atomic_t refcount; @@ -64,6 +65,16 @@ void rdma_addr_register_client(struct rdma_addr_client *client); */ void rdma_addr_unregister_client(struct rdma_addr_client *client); +/** + * struct rdma_dev_addr - Contains resolved RDMA hardware addresses + * @src_dev_addr: Source MAC address. + * @dst_dev_addr: Destination MAC address. + * @broadcast: Broadcast address of the device. + * @dev_type: The interface hardware type of the device. + * @bound_dev_if: An optional device interface index. + * @transport: The transport type used. + * @net: Network namespace containing the bound_dev_if net_dev. + */ struct rdma_dev_addr { unsigned char src_dev_addr[MAX_ADDR_LEN]; unsigned char dst_dev_addr[MAX_ADDR_LEN]; @@ -71,11 +82,14 @@ struct rdma_dev_addr { unsigned short dev_type; int bound_dev_if; enum rdma_transport_type transport; + struct net *net; }; /** * rdma_translate_ip - Translate a local IP address to an RDMA hardware * address. + * + * The dev_addr->net field must be initialized. */ int rdma_translate_ip(struct sockaddr *addr, struct rdma_dev_addr *dev_addr, u16 *vlan_id); @@ -90,7 +104,7 @@ int rdma_translate_ip(struct sockaddr *addr, struct rdma_dev_addr *dev_addr, * @dst_addr: The destination address to resolve. * @addr: A reference to a data location that will receive the resolved * addresses. The data location must remain valid until the callback has - * been invoked. + * been invoked. The net field of the addr struct must be valid. * @timeout_ms: Amount of time to wait for the address resolution to complete. * @callback: Call invoked once address resolution has completed, timed out, * or been canceled. A status of 0 indicates success. -- cgit v1.2.3 From fa20105e09e97e81aadf02f722c31195e4a75c84 Mon Sep 17 00:00:00 2001 From: Guy Shapiro Date: Thu, 22 Oct 2015 15:20:10 +0300 Subject: IB/cma: Add support for network namespaces Add support for network namespaces in the ib_cma module. This is accomplished by: 1. Adding network namespace parameter for rdma_create_id. This parameter is used to populate the network namespace field in rdma_id_private. rdma_create_id keeps a reference on the network namespace. 2. Using the network namespace from the rdma_id instead of init_net inside of ib_cma, when listening on an ID and when looking for an ID for an incoming request. 3. Decrementing the reference count for the appropriate network namespace when calling rdma_destroy_id. In order to preserve the current behavior init_net is passed when calling from other modules. Signed-off-by: Guy Shapiro Signed-off-by: Haggai Eran Signed-off-by: Yotam Kenneth Signed-off-by: Shachar Raindel Signed-off-by: Doug Ledford --- drivers/infiniband/core/cma.c | 46 +++++++++++++--------- drivers/infiniband/core/ucma.c | 3 +- drivers/infiniband/ulp/iser/iser_verbs.c | 2 +- drivers/infiniband/ulp/isert/ib_isert.c | 2 +- .../staging/lustre/lnet/klnds/o2iblnd/o2iblnd.h | 4 +- include/rdma/rdma_cm.h | 6 ++- net/9p/trans_rdma.c | 4 +- net/rds/ib.c | 2 +- net/rds/ib_cm.c | 2 +- net/rds/iw.c | 2 +- net/rds/iw_cm.c | 2 +- net/rds/rdma_transport.c | 4 +- net/sunrpc/xprtrdma/svc_rdma_transport.c | 4 +- net/sunrpc/xprtrdma/verbs.c | 3 +- 14 files changed, 52 insertions(+), 34 deletions(-) (limited to 'include') diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index ac03c32ca7f1..7e93eb1f33eb 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -600,7 +600,8 @@ static int cma_disable_callback(struct rdma_id_private *id_priv, return 0; } -struct rdma_cm_id *rdma_create_id(rdma_cm_event_handler event_handler, +struct rdma_cm_id *rdma_create_id(struct net *net, + rdma_cm_event_handler event_handler, void *context, enum rdma_port_space ps, enum ib_qp_type qp_type) { @@ -624,7 +625,7 @@ struct rdma_cm_id *rdma_create_id(rdma_cm_event_handler event_handler, INIT_LIST_HEAD(&id_priv->listen_list); INIT_LIST_HEAD(&id_priv->mc_list); get_random_bytes(&id_priv->seq_num, sizeof id_priv->seq_num); - id_priv->id.route.addr.dev_addr.net = &init_net; + id_priv->id.route.addr.dev_addr.net = get_net(net); return &id_priv->id; } @@ -1278,7 +1279,7 @@ static bool cma_match_net_dev(const struct rdma_id_private *id_priv, cma_protocol_roce(&id_priv->id); return !addr->dev_addr.bound_dev_if || - (net_eq(dev_net(net_dev), &init_net) && + (net_eq(dev_net(net_dev), addr->dev_addr.net) && addr->dev_addr.bound_dev_if == net_dev->ifindex); } @@ -1339,7 +1340,7 @@ static struct rdma_id_private *cma_id_from_event(struct ib_cm_id *cm_id, } } - bind_list = cma_ps_find(&init_net, + bind_list = cma_ps_find(*net_dev ? dev_net(*net_dev) : &init_net, rdma_ps_from_service_id(req.service_id), cma_port_from_service_id(req.service_id)); id_priv = cma_find_listener(bind_list, cm_id, ib_event, &req, *net_dev); @@ -1411,6 +1412,7 @@ static void cma_cancel_operation(struct rdma_id_private *id_priv, static void cma_release_port(struct rdma_id_private *id_priv) { struct rdma_bind_list *bind_list = id_priv->bind_list; + struct net *net = id_priv->id.route.addr.dev_addr.net; if (!bind_list) return; @@ -1418,7 +1420,7 @@ static void cma_release_port(struct rdma_id_private *id_priv) mutex_lock(&lock); hlist_del(&id_priv->node); if (hlist_empty(&bind_list->owners)) { - cma_ps_remove(&init_net, bind_list->ps, bind_list->port); + cma_ps_remove(net, bind_list->ps, bind_list->port); kfree(bind_list); } mutex_unlock(&lock); @@ -1477,6 +1479,7 @@ void rdma_destroy_id(struct rdma_cm_id *id) cma_deref_id(id_priv->id.context); kfree(id_priv->id.route.path_rec); + put_net(id_priv->id.route.addr.dev_addr.net); kfree(id_priv); } EXPORT_SYMBOL(rdma_destroy_id); @@ -1607,7 +1610,8 @@ static struct rdma_id_private *cma_new_conn_id(struct rdma_cm_id *listen_id, ib_event->param.req_rcvd.primary_path->service_id; int ret; - id = rdma_create_id(listen_id->event_handler, listen_id->context, + id = rdma_create_id(listen_id->route.addr.dev_addr.net, + listen_id->event_handler, listen_id->context, listen_id->ps, ib_event->param.req_rcvd.qp_type); if (IS_ERR(id)) return NULL; @@ -1662,9 +1666,10 @@ static struct rdma_id_private *cma_new_udp_id(struct rdma_cm_id *listen_id, struct rdma_id_private *id_priv; struct rdma_cm_id *id; const sa_family_t ss_family = listen_id->route.addr.src_addr.ss_family; + struct net *net = listen_id->route.addr.dev_addr.net; int ret; - id = rdma_create_id(listen_id->event_handler, listen_id->context, + id = rdma_create_id(net, listen_id->event_handler, listen_id->context, listen_id->ps, IB_QPT_UD); if (IS_ERR(id)) return NULL; @@ -1901,7 +1906,8 @@ static int iw_conn_req_handler(struct iw_cm_id *cm_id, return -ECONNABORTED; /* Create a new RDMA id for the new IW CM ID */ - new_cm_id = rdma_create_id(listen_id->id.event_handler, + new_cm_id = rdma_create_id(listen_id->id.route.addr.dev_addr.net, + listen_id->id.event_handler, listen_id->id.context, RDMA_PS_TCP, IB_QPT_RC); if (IS_ERR(new_cm_id)) { @@ -2029,12 +2035,13 @@ static void cma_listen_on_dev(struct rdma_id_private *id_priv, { struct rdma_id_private *dev_id_priv; struct rdma_cm_id *id; + struct net *net = id_priv->id.route.addr.dev_addr.net; int ret; if (cma_family(id_priv) == AF_IB && !rdma_cap_ib_cm(cma_dev->device, 1)) return; - id = rdma_create_id(cma_listen_handler, id_priv, id_priv->id.ps, + id = rdma_create_id(net, cma_listen_handler, id_priv, id_priv->id.ps, id_priv->id.qp_type); if (IS_ERR(id)) return; @@ -2708,7 +2715,8 @@ static int cma_alloc_port(enum rdma_port_space ps, if (!bind_list) return -ENOMEM; - ret = cma_ps_alloc(&init_net, ps, bind_list, snum); + ret = cma_ps_alloc(id_priv->id.route.addr.dev_addr.net, ps, bind_list, + snum); if (ret < 0) goto err; @@ -2727,13 +2735,14 @@ static int cma_alloc_any_port(enum rdma_port_space ps, static unsigned int last_used_port; int low, high, remaining; unsigned int rover; + struct net *net = id_priv->id.route.addr.dev_addr.net; - inet_get_local_port_range(&init_net, &low, &high); + inet_get_local_port_range(net, &low, &high); remaining = (high - low) + 1; rover = prandom_u32() % remaining + low; retry: if (last_used_port != rover && - !cma_ps_find(&init_net, ps, (unsigned short)rover)) { + !cma_ps_find(net, ps, (unsigned short)rover)) { int ret = cma_alloc_port(ps, id_priv, rover); /* * Remember previously used port number in order to avoid @@ -2799,7 +2808,7 @@ static int cma_use_port(enum rdma_port_space ps, if (snum < PROT_SOCK && !capable(CAP_NET_BIND_SERVICE)) return -EACCES; - bind_list = cma_ps_find(&init_net, ps, snum); + bind_list = cma_ps_find(id_priv->id.route.addr.dev_addr.net, ps, snum); if (!bind_list) { ret = cma_alloc_port(ps, id_priv, snum); } else { @@ -2991,8 +3000,11 @@ int rdma_bind_addr(struct rdma_cm_id *id, struct sockaddr *addr) if (addr->sa_family == AF_INET) id_priv->afonly = 1; #if IS_ENABLED(CONFIG_IPV6) - else if (addr->sa_family == AF_INET6) - id_priv->afonly = init_net.ipv6.sysctl.bindv6only; + else if (addr->sa_family == AF_INET6) { + struct net *net = id_priv->id.route.addr.dev_addr.net; + + id_priv->afonly = net->ipv6.sysctl.bindv6only; + } #endif } ret = cma_get_port(id_priv); @@ -3797,6 +3809,7 @@ static int cma_netdev_change(struct net_device *ndev, struct rdma_id_private *id dev_addr = &id_priv->id.route.addr.dev_addr; if ((dev_addr->bound_dev_if == ndev->ifindex) && + (net_eq(dev_net(ndev), dev_addr->net)) && memcmp(dev_addr->src_dev_addr, ndev->dev_addr, ndev->addr_len)) { printk(KERN_INFO "RDMA CM addr change for ndev %s used by id %p\n", ndev->name, &id_priv->id); @@ -3822,9 +3835,6 @@ static int cma_netdev_callback(struct notifier_block *self, unsigned long event, struct rdma_id_private *id_priv; int ret = NOTIFY_DONE; - if (dev_net(ndev) != &init_net) - return NOTIFY_DONE; - if (event != NETDEV_BONDING_FAILOVER) return NOTIFY_DONE; diff --git a/drivers/infiniband/core/ucma.c b/drivers/infiniband/core/ucma.c index 3a342dfaac15..e80c107450ab 100644 --- a/drivers/infiniband/core/ucma.c +++ b/drivers/infiniband/core/ucma.c @@ -472,7 +472,8 @@ static ssize_t ucma_create_id(struct ucma_file *file, const char __user *inbuf, return -ENOMEM; ctx->uid = cmd.uid; - ctx->cm_id = rdma_create_id(ucma_event_handler, ctx, cmd.ps, qp_type); + ctx->cm_id = rdma_create_id(&init_net, ucma_event_handler, ctx, cmd.ps, + qp_type); if (IS_ERR(ctx->cm_id)) { ret = PTR_ERR(ctx->cm_id); goto err1; diff --git a/drivers/infiniband/ulp/iser/iser_verbs.c b/drivers/infiniband/ulp/iser/iser_verbs.c index 85132d867bc8..f3122372d49f 100644 --- a/drivers/infiniband/ulp/iser/iser_verbs.c +++ b/drivers/infiniband/ulp/iser/iser_verbs.c @@ -1017,7 +1017,7 @@ int iser_connect(struct iser_conn *iser_conn, ib_conn->beacon.wr_id = ISER_BEACON_WRID; ib_conn->beacon.opcode = IB_WR_SEND; - ib_conn->cma_id = rdma_create_id(iser_cma_handler, + ib_conn->cma_id = rdma_create_id(&init_net, iser_cma_handler, (void *)iser_conn, RDMA_PS_TCP, IB_QPT_RC); if (IS_ERR(ib_conn->cma_id)) { diff --git a/drivers/infiniband/ulp/isert/ib_isert.c b/drivers/infiniband/ulp/isert/ib_isert.c index b58ac25ce19d..7d41c9d6d6bf 100644 --- a/drivers/infiniband/ulp/isert/ib_isert.c +++ b/drivers/infiniband/ulp/isert/ib_isert.c @@ -3096,7 +3096,7 @@ isert_setup_id(struct isert_np *isert_np) sa = (struct sockaddr *)&np->np_sockaddr; isert_dbg("ksockaddr: %p, sa: %p\n", &np->np_sockaddr, sa); - id = rdma_create_id(isert_cma_handler, isert_np, + id = rdma_create_id(&init_net, isert_cma_handler, isert_np, RDMA_PS_TCP, IB_QPT_RC); if (IS_ERR(id)) { isert_err("rdma_create_id() failed: %ld\n", PTR_ERR(id)); diff --git a/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd.h b/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd.h index f4b6c33ac318..5015a9e830bd 100644 --- a/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd.h +++ b/drivers/staging/lustre/lnet/klnds/o2iblnd/o2iblnd.h @@ -128,7 +128,9 @@ extern kib_tunables_t kiblnd_tunables; IBLND_CREDIT_HIGHWATER_V1 : \ *kiblnd_tunables.kib_peercredits_hiw) /* when eagerly to return credits */ -#define kiblnd_rdma_create_id(cb, dev, ps, qpt) rdma_create_id(cb, dev, ps, qpt) +#define kiblnd_rdma_create_id(cb, dev, ps, qpt) rdma_create_id(&init_net, \ + cb, dev, \ + ps, qpt) static inline int kiblnd_concurrent_sends_v1(void) diff --git a/include/rdma/rdma_cm.h b/include/rdma/rdma_cm.h index c92522c192d2..463ec0ccbe51 100644 --- a/include/rdma/rdma_cm.h +++ b/include/rdma/rdma_cm.h @@ -160,13 +160,17 @@ struct rdma_cm_id { /** * rdma_create_id - Create an RDMA identifier. * + * @net: The network namespace in which to create the new id. * @event_handler: User callback invoked to report events associated with the * returned rdma_id. * @context: User specified context associated with the id. * @ps: RDMA port space. * @qp_type: type of queue pair associated with the id. + * + * The id holds a reference on the network namespace until it is destroyed. */ -struct rdma_cm_id *rdma_create_id(rdma_cm_event_handler event_handler, +struct rdma_cm_id *rdma_create_id(struct net *net, + rdma_cm_event_handler event_handler, void *context, enum rdma_port_space ps, enum ib_qp_type qp_type); diff --git a/net/9p/trans_rdma.c b/net/9p/trans_rdma.c index ba1210253f5e..52b4a2f993f2 100644 --- a/net/9p/trans_rdma.c +++ b/net/9p/trans_rdma.c @@ -655,8 +655,8 @@ rdma_create_trans(struct p9_client *client, const char *addr, char *args) return -ENOMEM; /* Create the RDMA CM ID */ - rdma->cm_id = rdma_create_id(p9_cm_event_handler, client, RDMA_PS_TCP, - IB_QPT_RC); + rdma->cm_id = rdma_create_id(&init_net, p9_cm_event_handler, client, + RDMA_PS_TCP, IB_QPT_RC); if (IS_ERR(rdma->cm_id)) goto error; diff --git a/net/rds/ib.c b/net/rds/ib.c index 2d3f2ab475df..cd64ef9a4748 100644 --- a/net/rds/ib.c +++ b/net/rds/ib.c @@ -317,7 +317,7 @@ static int rds_ib_laddr_check(struct net *net, __be32 addr) /* Create a CMA ID and try to bind it. This catches both * IB and iWARP capable NICs. */ - cm_id = rdma_create_id(NULL, NULL, RDMA_PS_TCP, IB_QPT_RC); + cm_id = rdma_create_id(&init_net, NULL, NULL, RDMA_PS_TCP, IB_QPT_RC); if (IS_ERR(cm_id)) return PTR_ERR(cm_id); diff --git a/net/rds/ib_cm.c b/net/rds/ib_cm.c index 9043f5c04787..f5a98068faf0 100644 --- a/net/rds/ib_cm.c +++ b/net/rds/ib_cm.c @@ -565,7 +565,7 @@ int rds_ib_conn_connect(struct rds_connection *conn) /* XXX I wonder what affect the port space has */ /* delegate cm event handler to rdma_transport */ - ic->i_cm_id = rdma_create_id(rds_rdma_cm_event_handler, conn, + ic->i_cm_id = rdma_create_id(&init_net, rds_rdma_cm_event_handler, conn, RDMA_PS_TCP, IB_QPT_RC); if (IS_ERR(ic->i_cm_id)) { ret = PTR_ERR(ic->i_cm_id); diff --git a/net/rds/iw.c b/net/rds/iw.c index 3df0295c6659..576f1825fc55 100644 --- a/net/rds/iw.c +++ b/net/rds/iw.c @@ -223,7 +223,7 @@ static int rds_iw_laddr_check(struct net *net, __be32 addr) /* Create a CMA ID and try to bind it. This catches both * IB and iWARP capable NICs. */ - cm_id = rdma_create_id(NULL, NULL, RDMA_PS_TCP, IB_QPT_RC); + cm_id = rdma_create_id(&init_net, NULL, NULL, RDMA_PS_TCP, IB_QPT_RC); if (IS_ERR(cm_id)) return PTR_ERR(cm_id); diff --git a/net/rds/iw_cm.c b/net/rds/iw_cm.c index a6553a6fb2bc..aea4c911bc76 100644 --- a/net/rds/iw_cm.c +++ b/net/rds/iw_cm.c @@ -524,7 +524,7 @@ int rds_iw_conn_connect(struct rds_connection *conn) /* XXX I wonder what affect the port space has */ /* delegate cm event handler to rdma_transport */ - ic->i_cm_id = rdma_create_id(rds_rdma_cm_event_handler, conn, + ic->i_cm_id = rdma_create_id(&init_net, rds_rdma_cm_event_handler, conn, RDMA_PS_TCP, IB_QPT_RC); if (IS_ERR(ic->i_cm_id)) { ret = PTR_ERR(ic->i_cm_id); diff --git a/net/rds/rdma_transport.c b/net/rds/rdma_transport.c index b9b40af5345b..9c1fed81bf0f 100644 --- a/net/rds/rdma_transport.c +++ b/net/rds/rdma_transport.c @@ -142,8 +142,8 @@ static int rds_rdma_listen_init(void) struct rdma_cm_id *cm_id; int ret; - cm_id = rdma_create_id(rds_rdma_cm_event_handler, NULL, RDMA_PS_TCP, - IB_QPT_RC); + cm_id = rdma_create_id(&init_net, rds_rdma_cm_event_handler, NULL, + RDMA_PS_TCP, IB_QPT_RC); if (IS_ERR(cm_id)) { ret = PTR_ERR(cm_id); printk(KERN_ERR "RDS/RDMA: failed to setup listener, " diff --git a/net/sunrpc/xprtrdma/svc_rdma_transport.c b/net/sunrpc/xprtrdma/svc_rdma_transport.c index fcc3eb80c265..4a41122d586f 100644 --- a/net/sunrpc/xprtrdma/svc_rdma_transport.c +++ b/net/sunrpc/xprtrdma/svc_rdma_transport.c @@ -692,8 +692,8 @@ static struct svc_xprt *svc_rdma_create(struct svc_serv *serv, if (!cma_xprt) return ERR_PTR(-ENOMEM); - listen_id = rdma_create_id(rdma_listen_handler, cma_xprt, RDMA_PS_TCP, - IB_QPT_RC); + listen_id = rdma_create_id(&init_net, rdma_listen_handler, cma_xprt, + RDMA_PS_TCP, IB_QPT_RC); if (IS_ERR(listen_id)) { ret = PTR_ERR(listen_id); dprintk("svcrdma: rdma_create_id failed = %d\n", ret); diff --git a/net/sunrpc/xprtrdma/verbs.c b/net/sunrpc/xprtrdma/verbs.c index 81e8d31f3abc..6c06ba088fea 100644 --- a/net/sunrpc/xprtrdma/verbs.c +++ b/net/sunrpc/xprtrdma/verbs.c @@ -432,7 +432,8 @@ rpcrdma_create_id(struct rpcrdma_xprt *xprt, init_completion(&ia->ri_done); - id = rdma_create_id(rpcrdma_conn_upcall, xprt, RDMA_PS_TCP, IB_QPT_RC); + id = rdma_create_id(&init_net, rpcrdma_conn_upcall, xprt, RDMA_PS_TCP, + IB_QPT_RC); if (IS_ERR(id)) { rc = PTR_ERR(id); dprintk("RPC: %s: rdma_create_id() failed %i\n", -- cgit v1.2.3 From 4c67e2bfc8b7121d51434362fa7c2d012f8bcf1b Mon Sep 17 00:00:00 2001 From: Sagi Grimberg Date: Tue, 13 Oct 2015 19:11:24 +0300 Subject: IB/core: Introduce new fast registration API The new fast registration verb ib_map_mr_sg receives a scatterlist and converts it to a page list under the verbs API thus hiding the specific HW mapping details away from the consumer. The provider drivers are provided with a generic helper ib_sg_to_pages that converts a scatterlist into a vector of page addresses. The drivers can still perform any HW specific page address setting by passing a set_page function pointer which will be invoked for each page address. This allows drivers to avoid keeping a shadow page vectors and convert them to HW specific translations by doing extra copies. This API will allow ULPs to remove the duplicated code of constructing a page vector from a given sg list. The send work request ib_reg_wr also shrinks as it will contain only mr, key and access flags in addition. Signed-off-by: Sagi Grimberg Tested-by: Christoph Hellwig Signed-off-by: Doug Ledford --- drivers/infiniband/core/verbs.c | 107 ++++++++++++++++++++++++++++++++++++++++ include/rdma/ib_verbs.h | 44 +++++++++++++++++ 2 files changed, 151 insertions(+) (limited to 'include') diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c index 46d97f09fbae..5673df061213 100644 --- a/drivers/infiniband/core/verbs.c +++ b/drivers/infiniband/core/verbs.c @@ -1496,3 +1496,110 @@ int ib_check_mr_status(struct ib_mr *mr, u32 check_mask, mr->device->check_mr_status(mr, check_mask, mr_status) : -ENOSYS; } EXPORT_SYMBOL(ib_check_mr_status); + +/** + * ib_map_mr_sg() - Map the largest prefix of a dma mapped SG list + * and set it the memory region. + * @mr: memory region + * @sg: dma mapped scatterlist + * @sg_nents: number of entries in sg + * @page_size: page vector desired page size + * + * Constraints: + * - The first sg element is allowed to have an offset. + * - Each sg element must be aligned to page_size (or physically + * contiguous to the previous element). In case an sg element has a + * non contiguous offset, the mapping prefix will not include it. + * - The last sg element is allowed to have length less than page_size. + * - If sg_nents total byte length exceeds the mr max_num_sge * page_size + * then only max_num_sg entries will be mapped. + * + * Returns the number of sg elements that were mapped to the memory region. + * + * After this completes successfully, the memory region + * is ready for registration. + */ +int ib_map_mr_sg(struct ib_mr *mr, + struct scatterlist *sg, + int sg_nents, + unsigned int page_size) +{ + if (unlikely(!mr->device->map_mr_sg)) + return -ENOSYS; + + mr->page_size = page_size; + + return mr->device->map_mr_sg(mr, sg, sg_nents); +} +EXPORT_SYMBOL(ib_map_mr_sg); + +/** + * ib_sg_to_pages() - Convert the largest prefix of a sg list + * to a page vector + * @mr: memory region + * @sgl: dma mapped scatterlist + * @sg_nents: number of entries in sg + * @set_page: driver page assignment function pointer + * + * Core service helper for drivers to covert the largest + * prefix of given sg list to a page vector. The sg list + * prefix converted is the prefix that meet the requirements + * of ib_map_mr_sg. + * + * Returns the number of sg elements that were assigned to + * a page vector. + */ +int ib_sg_to_pages(struct ib_mr *mr, + struct scatterlist *sgl, + int sg_nents, + int (*set_page)(struct ib_mr *, u64)) +{ + struct scatterlist *sg; + u64 last_end_dma_addr = 0, last_page_addr = 0; + unsigned int last_page_off = 0; + u64 page_mask = ~((u64)mr->page_size - 1); + int i; + + mr->iova = sg_dma_address(&sgl[0]); + mr->length = 0; + + for_each_sg(sgl, sg, sg_nents, i) { + u64 dma_addr = sg_dma_address(sg); + unsigned int dma_len = sg_dma_len(sg); + u64 end_dma_addr = dma_addr + dma_len; + u64 page_addr = dma_addr & page_mask; + + if (i && page_addr != dma_addr) { + if (last_end_dma_addr != dma_addr) { + /* gap */ + goto done; + + } else if (last_page_off + dma_len <= mr->page_size) { + /* chunk this fragment with the last */ + mr->length += dma_len; + last_end_dma_addr += dma_len; + last_page_off += dma_len; + continue; + } else { + /* map starting from the next page */ + page_addr = last_page_addr + mr->page_size; + dma_len -= mr->page_size - last_page_off; + } + } + + do { + if (unlikely(set_page(mr, page_addr))) + goto done; + page_addr += mr->page_size; + } while (page_addr < end_dma_addr); + + mr->length += dma_len; + last_end_dma_addr = end_dma_addr; + last_page_addr = end_dma_addr & page_mask; + last_page_off = end_dma_addr & ~page_mask; + } + +done: + return i; +} +EXPORT_SYMBOL(ib_sg_to_pages); diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index 85103aff909b..065d37c13aa6 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -738,6 +738,7 @@ enum ib_wc_opcode { IB_WC_LSO, IB_WC_LOCAL_INV, IB_WC_FAST_REG_MR, + IB_WC_REG_MR, IB_WC_MASKED_COMP_SWAP, IB_WC_MASKED_FETCH_ADD, /* @@ -1025,6 +1026,7 @@ enum ib_wr_opcode { IB_WR_RDMA_READ_WITH_INV, IB_WR_LOCAL_INV, IB_WR_FAST_REG_MR, + IB_WR_REG_MR, IB_WR_MASKED_ATOMIC_CMP_AND_SWP, IB_WR_MASKED_ATOMIC_FETCH_AND_ADD, IB_WR_BIND_MW, @@ -1157,6 +1159,18 @@ static inline struct ib_fast_reg_wr *fast_reg_wr(struct ib_send_wr *wr) return container_of(wr, struct ib_fast_reg_wr, wr); } +struct ib_reg_wr { + struct ib_send_wr wr; + struct ib_mr *mr; + u32 key; + int access; +}; + +static inline struct ib_reg_wr *reg_wr(struct ib_send_wr *wr) +{ + return container_of(wr, struct ib_reg_wr, wr); +} + struct ib_bind_mw_wr { struct ib_send_wr wr; struct ib_mw *mw; @@ -1369,6 +1383,9 @@ struct ib_mr { struct ib_uobject *uobject; u32 lkey; u32 rkey; + u64 iova; + u32 length; + unsigned int page_size; atomic_t usecnt; /* count number of MWs */ }; @@ -1753,6 +1770,9 @@ struct ib_device { struct ib_mr * (*alloc_mr)(struct ib_pd *pd, enum ib_mr_type mr_type, u32 max_num_sg); + int (*map_mr_sg)(struct ib_mr *mr, + struct scatterlist *sg, + int sg_nents); struct ib_fast_reg_page_list * (*alloc_fast_reg_page_list)(struct ib_device *device, int page_list_len); void (*free_fast_reg_page_list)(struct ib_fast_reg_page_list *page_list); @@ -3059,4 +3079,28 @@ struct net_device *ib_get_net_dev_by_params(struct ib_device *dev, u8 port, u16 pkey, const union ib_gid *gid, const struct sockaddr *addr); +int ib_map_mr_sg(struct ib_mr *mr, + struct scatterlist *sg, + int sg_nents, + unsigned int page_size); + +static inline int +ib_map_mr_sg_zbva(struct ib_mr *mr, + struct scatterlist *sg, + int sg_nents, + unsigned int page_size) +{ + int n; + + n = ib_map_mr_sg(mr, sg, sg_nents, page_size); + mr->iova = 0; + + return n; +} + +int ib_sg_to_pages(struct ib_mr *mr, + struct scatterlist *sgl, + int sg_nents, + int (*set_page)(struct ib_mr *, u64)); + #endif /* IB_VERBS_H */ -- cgit v1.2.3 From 412a15c0fe537c59c794d4e8134580b9cb984a0c Mon Sep 17 00:00:00 2001 From: Sagi Grimberg Date: Tue, 13 Oct 2015 19:11:36 +0300 Subject: svcrdma: Port to new memory registration API Instead of maintaining a fastreg page list, keep an sg table and convert an array of pages to a sg list. Then call ib_map_mr_sg and construct ib_reg_wr. Signed-off-by: Sagi Grimberg Acked-by: Christoph Hellwig Tested-by: Steve Wise Tested-by: Selvin Xavier Signed-off-by: Doug Ledford --- include/linux/sunrpc/svc_rdma.h | 6 +-- net/sunrpc/xprtrdma/svc_rdma_recvfrom.c | 76 ++++++++++++++++++-------------- net/sunrpc/xprtrdma/svc_rdma_transport.c | 34 +++++--------- 3 files changed, 55 insertions(+), 61 deletions(-) (limited to 'include') diff --git a/include/linux/sunrpc/svc_rdma.h b/include/linux/sunrpc/svc_rdma.h index 7ccc961f33e9..1e4438ea2380 100644 --- a/include/linux/sunrpc/svc_rdma.h +++ b/include/linux/sunrpc/svc_rdma.h @@ -105,11 +105,9 @@ struct svc_rdma_chunk_sge { }; struct svc_rdma_fastreg_mr { struct ib_mr *mr; - void *kva; - struct ib_fast_reg_page_list *page_list; - int page_list_len; + struct scatterlist *sg; + int sg_nents; unsigned long access_flags; - unsigned long map_len; enum dma_data_direction direction; struct list_head frmr_list; }; diff --git a/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c b/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c index 7be42d0da19e..cb0991345816 100644 --- a/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c +++ b/net/sunrpc/xprtrdma/svc_rdma_recvfrom.c @@ -220,12 +220,12 @@ int rdma_read_chunk_frmr(struct svcxprt_rdma *xprt, { struct ib_rdma_wr read_wr; struct ib_send_wr inv_wr; - struct ib_fast_reg_wr fastreg_wr; + struct ib_reg_wr reg_wr; u8 key; - int pages_needed = PAGE_ALIGN(*page_offset + rs_length) >> PAGE_SHIFT; + int nents = PAGE_ALIGN(*page_offset + rs_length) >> PAGE_SHIFT; struct svc_rdma_op_ctxt *ctxt = svc_rdma_get_context(xprt); struct svc_rdma_fastreg_mr *frmr = svc_rdma_get_frmr(xprt); - int ret, read, pno; + int ret, read, pno, dma_nents, n; u32 pg_off = *page_offset; u32 pg_no = *page_no; @@ -234,16 +234,14 @@ int rdma_read_chunk_frmr(struct svcxprt_rdma *xprt, ctxt->direction = DMA_FROM_DEVICE; ctxt->frmr = frmr; - pages_needed = min_t(int, pages_needed, xprt->sc_frmr_pg_list_len); - read = min_t(int, pages_needed << PAGE_SHIFT, rs_length); + nents = min_t(unsigned int, nents, xprt->sc_frmr_pg_list_len); + read = min_t(int, nents << PAGE_SHIFT, rs_length); - frmr->kva = page_address(rqstp->rq_arg.pages[pg_no]); frmr->direction = DMA_FROM_DEVICE; frmr->access_flags = (IB_ACCESS_LOCAL_WRITE|IB_ACCESS_REMOTE_WRITE); - frmr->map_len = pages_needed << PAGE_SHIFT; - frmr->page_list_len = pages_needed; + frmr->sg_nents = nents; - for (pno = 0; pno < pages_needed; pno++) { + for (pno = 0; pno < nents; pno++) { int len = min_t(int, rs_length, PAGE_SIZE - pg_off); head->arg.pages[pg_no] = rqstp->rq_arg.pages[pg_no]; @@ -251,17 +249,12 @@ int rdma_read_chunk_frmr(struct svcxprt_rdma *xprt, head->arg.len += len; if (!pg_off) head->count++; + + sg_set_page(&frmr->sg[pno], rqstp->rq_arg.pages[pg_no], + len, pg_off); + rqstp->rq_respages = &rqstp->rq_arg.pages[pg_no+1]; rqstp->rq_next_page = rqstp->rq_respages + 1; - frmr->page_list->page_list[pno] = - ib_dma_map_page(xprt->sc_cm_id->device, - head->arg.pages[pg_no], 0, - PAGE_SIZE, DMA_FROM_DEVICE); - ret = ib_dma_mapping_error(xprt->sc_cm_id->device, - frmr->page_list->page_list[pno]); - if (ret) - goto err; - atomic_inc(&xprt->sc_dma_used); /* adjust offset and wrap to next page if needed */ pg_off += len; @@ -277,28 +270,42 @@ int rdma_read_chunk_frmr(struct svcxprt_rdma *xprt, else clear_bit(RDMACTXT_F_LAST_CTXT, &ctxt->flags); + dma_nents = ib_dma_map_sg(xprt->sc_cm_id->device, + frmr->sg, frmr->sg_nents, + frmr->direction); + if (!dma_nents) { + pr_err("svcrdma: failed to dma map sg %p\n", + frmr->sg); + return -ENOMEM; + } + atomic_inc(&xprt->sc_dma_used); + + n = ib_map_mr_sg(frmr->mr, frmr->sg, frmr->sg_nents, PAGE_SIZE); + if (unlikely(n != frmr->sg_nents)) { + pr_err("svcrdma: failed to map mr %p (%d/%d elements)\n", + frmr->mr, n, frmr->sg_nents); + return n < 0 ? n : -EINVAL; + } + /* Bump the key */ key = (u8)(frmr->mr->lkey & 0x000000FF); ib_update_fast_reg_key(frmr->mr, ++key); - ctxt->sge[0].addr = (unsigned long)frmr->kva + *page_offset; + ctxt->sge[0].addr = frmr->mr->iova; ctxt->sge[0].lkey = frmr->mr->lkey; - ctxt->sge[0].length = read; + ctxt->sge[0].length = frmr->mr->length; ctxt->count = 1; ctxt->read_hdr = head; - /* Prepare FASTREG WR */ - memset(&fastreg_wr, 0, sizeof(fastreg_wr)); - fastreg_wr.wr.opcode = IB_WR_FAST_REG_MR; - fastreg_wr.wr.send_flags = IB_SEND_SIGNALED; - fastreg_wr.iova_start = (unsigned long)frmr->kva; - fastreg_wr.page_list = frmr->page_list; - fastreg_wr.page_list_len = frmr->page_list_len; - fastreg_wr.page_shift = PAGE_SHIFT; - fastreg_wr.length = frmr->map_len; - fastreg_wr.access_flags = frmr->access_flags; - fastreg_wr.rkey = frmr->mr->lkey; - fastreg_wr.wr.next = &read_wr.wr; + /* Prepare REG WR */ + reg_wr.wr.opcode = IB_WR_REG_MR; + reg_wr.wr.wr_id = 0; + reg_wr.wr.send_flags = IB_SEND_SIGNALED; + reg_wr.wr.num_sge = 0; + reg_wr.mr = frmr->mr; + reg_wr.key = frmr->mr->lkey; + reg_wr.access = frmr->access_flags; + reg_wr.wr.next = &read_wr.wr; /* Prepare RDMA_READ */ memset(&read_wr, 0, sizeof(read_wr)); @@ -324,7 +331,7 @@ int rdma_read_chunk_frmr(struct svcxprt_rdma *xprt, ctxt->wr_op = read_wr.wr.opcode; /* Post the chain */ - ret = svc_rdma_send(xprt, &fastreg_wr.wr); + ret = svc_rdma_send(xprt, ®_wr.wr); if (ret) { pr_err("svcrdma: Error %d posting RDMA_READ\n", ret); set_bit(XPT_CLOSE, &xprt->sc_xprt.xpt_flags); @@ -338,7 +345,8 @@ int rdma_read_chunk_frmr(struct svcxprt_rdma *xprt, atomic_inc(&rdma_stat_read); return ret; err: - svc_rdma_unmap_dma(ctxt); + ib_dma_unmap_sg(xprt->sc_cm_id->device, + frmr->sg, frmr->sg_nents, frmr->direction); svc_rdma_put_context(ctxt, 0); svc_rdma_put_frmr(xprt, frmr); return ret; diff --git a/net/sunrpc/xprtrdma/svc_rdma_transport.c b/net/sunrpc/xprtrdma/svc_rdma_transport.c index 4a41122d586f..a266e870d870 100644 --- a/net/sunrpc/xprtrdma/svc_rdma_transport.c +++ b/net/sunrpc/xprtrdma/svc_rdma_transport.c @@ -732,7 +732,7 @@ static struct svc_xprt *svc_rdma_create(struct svc_serv *serv, static struct svc_rdma_fastreg_mr *rdma_alloc_frmr(struct svcxprt_rdma *xprt) { struct ib_mr *mr; - struct ib_fast_reg_page_list *pl; + struct scatterlist *sg; struct svc_rdma_fastreg_mr *frmr; u32 num_sg; @@ -745,13 +745,14 @@ static struct svc_rdma_fastreg_mr *rdma_alloc_frmr(struct svcxprt_rdma *xprt) if (IS_ERR(mr)) goto err_free_frmr; - pl = ib_alloc_fast_reg_page_list(xprt->sc_cm_id->device, - num_sg); - if (IS_ERR(pl)) + sg = kcalloc(RPCSVC_MAXPAGES, sizeof(*sg), GFP_KERNEL); + if (!sg) goto err_free_mr; + sg_init_table(sg, RPCSVC_MAXPAGES); + frmr->mr = mr; - frmr->page_list = pl; + frmr->sg = sg; INIT_LIST_HEAD(&frmr->frmr_list); return frmr; @@ -771,8 +772,8 @@ static void rdma_dealloc_frmr_q(struct svcxprt_rdma *xprt) frmr = list_entry(xprt->sc_frmr_q.next, struct svc_rdma_fastreg_mr, frmr_list); list_del_init(&frmr->frmr_list); + kfree(frmr->sg); ib_dereg_mr(frmr->mr); - ib_free_fast_reg_page_list(frmr->page_list); kfree(frmr); } } @@ -786,8 +787,7 @@ struct svc_rdma_fastreg_mr *svc_rdma_get_frmr(struct svcxprt_rdma *rdma) frmr = list_entry(rdma->sc_frmr_q.next, struct svc_rdma_fastreg_mr, frmr_list); list_del_init(&frmr->frmr_list); - frmr->map_len = 0; - frmr->page_list_len = 0; + frmr->sg_nents = 0; } spin_unlock_bh(&rdma->sc_frmr_q_lock); if (frmr) @@ -796,25 +796,13 @@ struct svc_rdma_fastreg_mr *svc_rdma_get_frmr(struct svcxprt_rdma *rdma) return rdma_alloc_frmr(rdma); } -static void frmr_unmap_dma(struct svcxprt_rdma *xprt, - struct svc_rdma_fastreg_mr *frmr) -{ - int page_no; - for (page_no = 0; page_no < frmr->page_list_len; page_no++) { - dma_addr_t addr = frmr->page_list->page_list[page_no]; - if (ib_dma_mapping_error(frmr->mr->device, addr)) - continue; - atomic_dec(&xprt->sc_dma_used); - ib_dma_unmap_page(frmr->mr->device, addr, PAGE_SIZE, - frmr->direction); - } -} - void svc_rdma_put_frmr(struct svcxprt_rdma *rdma, struct svc_rdma_fastreg_mr *frmr) { if (frmr) { - frmr_unmap_dma(rdma, frmr); + ib_dma_unmap_sg(rdma->sc_cm_id->device, + frmr->sg, frmr->sg_nents, frmr->direction); + atomic_dec(&rdma->sc_dma_used); spin_lock_bh(&rdma->sc_frmr_q_lock); WARN_ON_ONCE(!list_empty(&frmr->frmr_list)); list_add(&frmr->frmr_list, &rdma->sc_frmr_q); -- cgit v1.2.3 From 39bfc271bd687be2c8e396e976c0fb9a97963400 Mon Sep 17 00:00:00 2001 From: Sagi Grimberg Date: Tue, 13 Oct 2015 19:11:49 +0300 Subject: IB/core: Remove old fast registration API No callers and no providers left, go ahead and remove it. Signed-off-by: Sagi Grimberg Acked-by: Christoph Hellwig Signed-off-by: Doug Ledford --- drivers/infiniband/core/verbs.c | 25 ------------------- include/rdma/ib_verbs.h | 54 ----------------------------------------- 2 files changed, 79 deletions(-) (limited to 'include') diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c index 5673df061213..e2e53f9d7a22 100644 --- a/drivers/infiniband/core/verbs.c +++ b/drivers/infiniband/core/verbs.c @@ -1280,31 +1280,6 @@ struct ib_mr *ib_alloc_mr(struct ib_pd *pd, } EXPORT_SYMBOL(ib_alloc_mr); -struct ib_fast_reg_page_list *ib_alloc_fast_reg_page_list(struct ib_device *device, - int max_page_list_len) -{ - struct ib_fast_reg_page_list *page_list; - - if (!device->alloc_fast_reg_page_list) - return ERR_PTR(-ENOSYS); - - page_list = device->alloc_fast_reg_page_list(device, max_page_list_len); - - if (!IS_ERR(page_list)) { - page_list->device = device; - page_list->max_page_list_len = max_page_list_len; - } - - return page_list; -} -EXPORT_SYMBOL(ib_alloc_fast_reg_page_list); - -void ib_free_fast_reg_page_list(struct ib_fast_reg_page_list *page_list) -{ - page_list->device->free_fast_reg_page_list(page_list); -} -EXPORT_SYMBOL(ib_free_fast_reg_page_list); - /* Memory windows */ struct ib_mw *ib_alloc_mw(struct ib_pd *pd, enum ib_mw_type type) diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index 065d37c13aa6..324e9bf8e66c 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -737,7 +737,6 @@ enum ib_wc_opcode { IB_WC_BIND_MW, IB_WC_LSO, IB_WC_LOCAL_INV, - IB_WC_FAST_REG_MR, IB_WC_REG_MR, IB_WC_MASKED_COMP_SWAP, IB_WC_MASKED_FETCH_ADD, @@ -1025,7 +1024,6 @@ enum ib_wr_opcode { IB_WR_SEND_WITH_INV, IB_WR_RDMA_READ_WITH_INV, IB_WR_LOCAL_INV, - IB_WR_FAST_REG_MR, IB_WR_REG_MR, IB_WR_MASKED_ATOMIC_CMP_AND_SWP, IB_WR_MASKED_ATOMIC_FETCH_AND_ADD, @@ -1064,12 +1062,6 @@ struct ib_sge { u32 lkey; }; -struct ib_fast_reg_page_list { - struct ib_device *device; - u64 *page_list; - unsigned int max_page_list_len; -}; - /** * struct ib_mw_bind_info - Parameters for a memory window bind operation. * @mr: A memory region to bind the memory window to. @@ -1143,22 +1135,6 @@ static inline struct ib_ud_wr *ud_wr(struct ib_send_wr *wr) return container_of(wr, struct ib_ud_wr, wr); } -struct ib_fast_reg_wr { - struct ib_send_wr wr; - u64 iova_start; - struct ib_fast_reg_page_list *page_list; - unsigned int page_shift; - unsigned int page_list_len; - u32 length; - int access_flags; - u32 rkey; -}; - -static inline struct ib_fast_reg_wr *fast_reg_wr(struct ib_send_wr *wr) -{ - return container_of(wr, struct ib_fast_reg_wr, wr); -} - struct ib_reg_wr { struct ib_send_wr wr; struct ib_mr *mr; @@ -1773,9 +1749,6 @@ struct ib_device { int (*map_mr_sg)(struct ib_mr *mr, struct scatterlist *sg, int sg_nents); - struct ib_fast_reg_page_list * (*alloc_fast_reg_page_list)(struct ib_device *device, - int page_list_len); - void (*free_fast_reg_page_list)(struct ib_fast_reg_page_list *page_list); int (*rereg_phys_mr)(struct ib_mr *mr, int mr_rereg_mask, struct ib_pd *pd, @@ -2884,33 +2857,6 @@ struct ib_mr *ib_alloc_mr(struct ib_pd *pd, enum ib_mr_type mr_type, u32 max_num_sg); -/** - * ib_alloc_fast_reg_page_list - Allocates a page list array - * @device - ib device pointer. - * @page_list_len - size of the page list array to be allocated. - * - * This allocates and returns a struct ib_fast_reg_page_list * and a - * page_list array that is at least page_list_len in size. The actual - * size is returned in max_page_list_len. The caller is responsible - * for initializing the contents of the page_list array before posting - * a send work request with the IB_WC_FAST_REG_MR opcode. - * - * The page_list array entries must be translated using one of the - * ib_dma_*() functions just like the addresses passed to - * ib_map_phys_fmr(). Once the ib_post_send() is issued, the struct - * ib_fast_reg_page_list must not be modified by the caller until the - * IB_WC_FAST_REG_MR work request completes. - */ -struct ib_fast_reg_page_list *ib_alloc_fast_reg_page_list( - struct ib_device *device, int page_list_len); - -/** - * ib_free_fast_reg_page_list - Deallocates a previously allocated - * page list array. - * @page_list - struct ib_fast_reg_page_list pointer to be deallocated. - */ -void ib_free_fast_reg_page_list(struct ib_fast_reg_page_list *page_list); - /** * ib_update_fast_reg_key - updates the key portion of the fast_reg MR * R_Key and L_Key. -- cgit v1.2.3 From db7489e07669073970358b6cacf6a9dd8dc9275e Mon Sep 17 00:00:00 2001 From: Bart Van Assche Date: Mon, 3 Aug 2015 10:01:52 -0700 Subject: IB/core, cma: Make __attribute_const__ declarations sparse-friendly Move the __attribute_const__ declarations such that sparse understands that these apply to the function itself and not to the return type. This avoids that sparse reports error messages like the following: drivers/infiniband/core/verbs.c:73:12: error: symbol 'ib_event_msg' redeclared with different type (originally declared at include/rdma/ib_verbs.h:470) - different modifiers Fixes: 2b1b5b601230 ("IB/core, cma: Nice log-friendly string helpers") Signed-off-by: Bart Van Assche Cc: Sagi Grimberg Reviewed-by: Sagi Grimberg Signed-off-by: Doug Ledford --- drivers/infiniband/core/cma.c | 2 +- drivers/infiniband/core/verbs.c | 4 ++-- include/rdma/ib_verbs.h | 4 ++-- include/rdma/rdma_cm.h | 2 +- 4 files changed, 6 insertions(+), 6 deletions(-) (limited to 'include') diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c index 7e93eb1f33eb..944cd90417bc 100644 --- a/drivers/infiniband/core/cma.c +++ b/drivers/infiniband/core/cma.c @@ -88,7 +88,7 @@ static const char * const cma_events[] = { [RDMA_CM_EVENT_TIMEWAIT_EXIT] = "timewait exit", }; -const char *rdma_event_msg(enum rdma_cm_event_type event) +const char *__attribute_const__ rdma_event_msg(enum rdma_cm_event_type event) { size_t index = event; diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c index e2e53f9d7a22..043a60ee6836 100644 --- a/drivers/infiniband/core/verbs.c +++ b/drivers/infiniband/core/verbs.c @@ -73,7 +73,7 @@ static const char * const ib_events[] = { [IB_EVENT_GID_CHANGE] = "GID changed", }; -const char *ib_event_msg(enum ib_event_type event) +const char *__attribute_const__ ib_event_msg(enum ib_event_type event) { size_t index = event; @@ -107,7 +107,7 @@ static const char * const wc_statuses[] = { [IB_WC_GENERAL_ERR] = "general error", }; -const char *ib_wc_status_msg(enum ib_wc_status status) +const char *__attribute_const__ ib_wc_status_msg(enum ib_wc_status status) { size_t index = status; diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h index 324e9bf8e66c..9a68a19532ba 100644 --- a/include/rdma/ib_verbs.h +++ b/include/rdma/ib_verbs.h @@ -476,7 +476,7 @@ enum ib_event_type { IB_EVENT_GID_CHANGE, }; -__attribute_const__ const char *ib_event_msg(enum ib_event_type event); +const char *__attribute_const__ ib_event_msg(enum ib_event_type event); struct ib_event { struct ib_device *device; @@ -726,7 +726,7 @@ enum ib_wc_status { IB_WC_GENERAL_ERR }; -__attribute_const__ const char *ib_wc_status_msg(enum ib_wc_status status); +const char *__attribute_const__ ib_wc_status_msg(enum ib_wc_status status); enum ib_wc_opcode { IB_WC_SEND, diff --git a/include/rdma/rdma_cm.h b/include/rdma/rdma_cm.h index 463ec0ccbe51..afe44fde72a5 100644 --- a/include/rdma/rdma_cm.h +++ b/include/rdma/rdma_cm.h @@ -62,7 +62,7 @@ enum rdma_cm_event_type { RDMA_CM_EVENT_TIMEWAIT_EXIT }; -__attribute_const__ const char *rdma_event_msg(enum rdma_cm_event_type event); +const char *__attribute_const__ rdma_event_msg(enum rdma_cm_event_type event); enum rdma_port_space { RDMA_PS_SDP = 0x0001, -- cgit v1.2.3