From 25f40220e56b507fce186985c0ed2967b7f04596 Mon Sep 17 00:00:00 2001
From: Moni Shoua <monis@mellanox.com>
Date: Wed, 23 Dec 2015 14:56:56 +0200
Subject: IB/core: Initialize UD header structure with IP and UDP headers

ib_ud_header_init() is used to format InfiniBand headers
in a buffer up to (but not with) BTH. For RoCE UDP ENCAP it is
required that this function would be able to build also IP and UDP
headers.

Signed-off-by: Moni Shoua <monis@mellanox.com>
Signed-off-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/qp.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

(limited to 'drivers/infiniband/hw/mlx4')

diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c
index 13eaaf45288f..6049644da429 100644
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -2168,7 +2168,7 @@ static int build_sriov_qp0_header(struct mlx4_ib_sqp *sqp,
 	if (sqp->qp.mlx4_ib_qp_type == MLX4_IB_QPT_PROXY_SMI_OWNER)
 		send_size += sizeof (struct mlx4_ib_tunnel_header);
 
-	ib_ud_header_init(send_size, 1, 0, 0, 0, 0, &sqp->ud_header);
+	ib_ud_header_init(send_size, 1, 0, 0, 0, 0, 0, 0, &sqp->ud_header);
 
 	if (sqp->qp.mlx4_ib_qp_type == MLX4_IB_QPT_PROXY_SMI_OWNER) {
 		sqp->ud_header.lrh.service_level =
@@ -2314,7 +2314,10 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_ud_wr *wr,
 			is_vlan = 1;
 		}
 	}
-	ib_ud_header_init(send_size, !is_eth, is_eth, is_vlan, is_grh, 0, &sqp->ud_header);
+	err = ib_ud_header_init(send_size, !is_eth, is_eth, is_vlan, is_grh,
+				0, 0, 0, &sqp->ud_header);
+	if (err)
+		return err;
 
 	if (!is_eth) {
 		sqp->ud_header.lrh.service_level =
-- 
cgit v1.2.3


From feb7c1e38bccfd18cc06677cb648ed2340788fe8 Mon Sep 17 00:00:00 2001
From: Christoph Hellwig <hch@lst.de>
Date: Wed, 23 Dec 2015 19:12:48 +0100
Subject: IB: remove in-kernel support for memory windows

Remove the unused ib_allow_mw and ib_bind_mw functions, remove the
unused IB_WR_BIND_MW and IB_WC_BIND_MW opcodes and move ib_dealloc_mw
into the uverbs module.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagig@mellanox.com>
Reviewed-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> [core]
Reviewed-by: Steve Wise <swise@opengridcomputing.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 Documentation/infiniband/core_locking.txt   |  2 -
 drivers/infiniband/core/uverbs.h            |  2 +
 drivers/infiniband/core/uverbs_cmd.c        |  4 +-
 drivers/infiniband/core/uverbs_main.c       | 13 ++++-
 drivers/infiniband/core/verbs.c             | 36 -------------
 drivers/infiniband/hw/cxgb3/iwch_cq.c       |  4 --
 drivers/infiniband/hw/cxgb3/iwch_provider.c |  1 -
 drivers/infiniband/hw/cxgb3/iwch_provider.h |  3 --
 drivers/infiniband/hw/cxgb3/iwch_qp.c       | 82 ----------------------------
 drivers/infiniband/hw/cxgb4/cq.c            |  3 --
 drivers/infiniband/hw/cxgb4/iw_cxgb4.h      |  2 -
 drivers/infiniband/hw/cxgb4/provider.c      |  1 -
 drivers/infiniband/hw/cxgb4/qp.c            |  5 --
 drivers/infiniband/hw/mlx4/cq.c             |  3 --
 drivers/infiniband/hw/mlx4/main.c           |  1 -
 drivers/infiniband/hw/mlx4/mlx4_ib.h        |  2 -
 drivers/infiniband/hw/mlx4/mr.c             | 22 --------
 drivers/infiniband/hw/mlx4/qp.c             | 27 ----------
 drivers/infiniband/hw/mlx5/cq.c             |  3 --
 drivers/infiniband/hw/mthca/mthca_cq.c      |  3 --
 drivers/infiniband/hw/nes/nes_verbs.c       | 75 --------------------------
 drivers/staging/rdma/amso1100/c2_cq.c       |  3 --
 drivers/staging/rdma/ehca/ehca_iverbs.h     |  3 --
 drivers/staging/rdma/ehca/ehca_main.c       |  1 -
 drivers/staging/rdma/ehca/ehca_mrmw.c       | 12 -----
 drivers/staging/rdma/ehca/ehca_reqs.c       |  1 -
 include/rdma/ib_verbs.h                     | 83 -----------------------------
 27 files changed, 16 insertions(+), 381 deletions(-)

(limited to 'drivers/infiniband/hw/mlx4')

diff --git a/Documentation/infiniband/core_locking.txt b/Documentation/infiniband/core_locking.txt
index e1678542279a..4b1f36b6ada0 100644
--- a/Documentation/infiniband/core_locking.txt
+++ b/Documentation/infiniband/core_locking.txt
@@ -15,7 +15,6 @@ Sleeping and interrupt context
     modify_ah
     query_ah
     destroy_ah
-    bind_mw
     post_send
     post_recv
     poll_cq
@@ -31,7 +30,6 @@ Sleeping and interrupt context
     ib_modify_ah
     ib_query_ah
     ib_destroy_ah
-    ib_bind_mw
     ib_post_send
     ib_post_recv
     ib_req_notify_cq
diff --git a/drivers/infiniband/core/uverbs.h b/drivers/infiniband/core/uverbs.h
index 94bbd8c155fc..612ccfd39bf9 100644
--- a/drivers/infiniband/core/uverbs.h
+++ b/drivers/infiniband/core/uverbs.h
@@ -204,6 +204,8 @@ void ib_uverbs_event_handler(struct ib_event_handler *handler,
 			     struct ib_event *event);
 void ib_uverbs_dealloc_xrcd(struct ib_uverbs_device *dev, struct ib_xrcd *xrcd);
 
+int uverbs_dealloc_mw(struct ib_mw *mw);
+
 struct ib_uverbs_flow_spec {
 	union {
 		union {
diff --git a/drivers/infiniband/core/uverbs_cmd.c b/drivers/infiniband/core/uverbs_cmd.c
index 95610565fb46..5428ebee096f 100644
--- a/drivers/infiniband/core/uverbs_cmd.c
+++ b/drivers/infiniband/core/uverbs_cmd.c
@@ -1243,7 +1243,7 @@ err_copy:
 	idr_remove_uobj(&ib_uverbs_mw_idr, uobj);
 
 err_unalloc:
-	ib_dealloc_mw(mw);
+	uverbs_dealloc_mw(mw);
 
 err_put:
 	put_pd_read(pd);
@@ -1272,7 +1272,7 @@ ssize_t ib_uverbs_dealloc_mw(struct ib_uverbs_file *file,
 
 	mw = uobj->object;
 
-	ret = ib_dealloc_mw(mw);
+	ret = uverbs_dealloc_mw(mw);
 	if (!ret)
 		uobj->live = 0;
 
diff --git a/drivers/infiniband/core/uverbs_main.c b/drivers/infiniband/core/uverbs_main.c
index e3ef28861be6..39680aed99dd 100644
--- a/drivers/infiniband/core/uverbs_main.c
+++ b/drivers/infiniband/core/uverbs_main.c
@@ -133,6 +133,17 @@ static int (*uverbs_ex_cmd_table[])(struct ib_uverbs_file *file,
 static void ib_uverbs_add_one(struct ib_device *device);
 static void ib_uverbs_remove_one(struct ib_device *device, void *client_data);
 
+int uverbs_dealloc_mw(struct ib_mw *mw)
+{
+	struct ib_pd *pd = mw->pd;
+	int ret;
+
+	ret = mw->device->dealloc_mw(mw);
+	if (!ret)
+		atomic_dec(&pd->usecnt);
+	return ret;
+}
+
 static void ib_uverbs_release_dev(struct kobject *kobj)
 {
 	struct ib_uverbs_device *dev =
@@ -224,7 +235,7 @@ static int ib_uverbs_cleanup_ucontext(struct ib_uverbs_file *file,
 		struct ib_mw *mw = uobj->object;
 
 		idr_remove_uobj(&ib_uverbs_mw_idr, uobj);
-		ib_dealloc_mw(mw);
+		uverbs_dealloc_mw(mw);
 		kfree(uobj);
 	}
 
diff --git a/drivers/infiniband/core/verbs.c b/drivers/infiniband/core/verbs.c
index 70b1016be1c8..c5e0f07a7f82 100644
--- a/drivers/infiniband/core/verbs.c
+++ b/drivers/infiniband/core/verbs.c
@@ -1403,42 +1403,6 @@ struct ib_mr *ib_alloc_mr(struct ib_pd *pd,
 }
 EXPORT_SYMBOL(ib_alloc_mr);
 
-/* Memory windows */
-
-struct ib_mw *ib_alloc_mw(struct ib_pd *pd, enum ib_mw_type type)
-{
-	struct ib_mw *mw;
-
-	if (!pd->device->alloc_mw)
-		return ERR_PTR(-ENOSYS);
-
-	mw = pd->device->alloc_mw(pd, type);
-	if (!IS_ERR(mw)) {
-		mw->device  = pd->device;
-		mw->pd      = pd;
-		mw->uobject = NULL;
-		mw->type    = type;
-		atomic_inc(&pd->usecnt);
-	}
-
-	return mw;
-}
-EXPORT_SYMBOL(ib_alloc_mw);
-
-int ib_dealloc_mw(struct ib_mw *mw)
-{
-	struct ib_pd *pd;
-	int ret;
-
-	pd = mw->pd;
-	ret = mw->device->dealloc_mw(mw);
-	if (!ret)
-		atomic_dec(&pd->usecnt);
-
-	return ret;
-}
-EXPORT_SYMBOL(ib_dealloc_mw);
-
 /* "Fast" memory regions */
 
 struct ib_fmr *ib_alloc_fmr(struct ib_pd *pd,
diff --git a/drivers/infiniband/hw/cxgb3/iwch_cq.c b/drivers/infiniband/hw/cxgb3/iwch_cq.c
index cfe404925a39..97fbfd2c298e 100644
--- a/drivers/infiniband/hw/cxgb3/iwch_cq.c
+++ b/drivers/infiniband/hw/cxgb3/iwch_cq.c
@@ -115,10 +115,6 @@ static int iwch_poll_cq_one(struct iwch_dev *rhp, struct iwch_cq *chp,
 		case T3_SEND_WITH_SE_INV:
 			wc->opcode = IB_WC_SEND;
 			break;
-		case T3_BIND_MW:
-			wc->opcode = IB_WC_BIND_MW;
-			break;
-
 		case T3_LOCAL_INV:
 			wc->opcode = IB_WC_LOCAL_INV;
 			break;
diff --git a/drivers/infiniband/hw/cxgb3/iwch_provider.c b/drivers/infiniband/hw/cxgb3/iwch_provider.c
index f7aa019a3c1a..384e1d783507 100644
--- a/drivers/infiniband/hw/cxgb3/iwch_provider.c
+++ b/drivers/infiniband/hw/cxgb3/iwch_provider.c
@@ -1388,7 +1388,6 @@ int iwch_register_device(struct iwch_dev *dev)
 	dev->ibdev.reg_user_mr = iwch_reg_user_mr;
 	dev->ibdev.dereg_mr = iwch_dereg_mr;
 	dev->ibdev.alloc_mw = iwch_alloc_mw;
-	dev->ibdev.bind_mw = iwch_bind_mw;
 	dev->ibdev.dealloc_mw = iwch_dealloc_mw;
 	dev->ibdev.alloc_mr = iwch_alloc_mr;
 	dev->ibdev.map_mr_sg = iwch_map_mr_sg;
diff --git a/drivers/infiniband/hw/cxgb3/iwch_provider.h b/drivers/infiniband/hw/cxgb3/iwch_provider.h
index f4fa6d677c0f..f24df44146c6 100644
--- a/drivers/infiniband/hw/cxgb3/iwch_provider.h
+++ b/drivers/infiniband/hw/cxgb3/iwch_provider.h
@@ -330,9 +330,6 @@ int iwch_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 		      struct ib_send_wr **bad_wr);
 int iwch_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *wr,
 		      struct ib_recv_wr **bad_wr);
-int iwch_bind_mw(struct ib_qp *qp,
-			     struct ib_mw *mw,
-			     struct ib_mw_bind *mw_bind);
 int iwch_poll_cq(struct ib_cq *ibcq, int num_entries, struct ib_wc *wc);
 int iwch_post_terminate(struct iwch_qp *qhp, struct respQ_msg_t *rsp_msg);
 int iwch_post_zb_read(struct iwch_ep *ep);
diff --git a/drivers/infiniband/hw/cxgb3/iwch_qp.c b/drivers/infiniband/hw/cxgb3/iwch_qp.c
index d0548fc6395e..d939980a708f 100644
--- a/drivers/infiniband/hw/cxgb3/iwch_qp.c
+++ b/drivers/infiniband/hw/cxgb3/iwch_qp.c
@@ -526,88 +526,6 @@ out:
 	return err;
 }
 
-int iwch_bind_mw(struct ib_qp *qp,
-			     struct ib_mw *mw,
-			     struct ib_mw_bind *mw_bind)
-{
-	struct iwch_dev *rhp;
-	struct iwch_mw *mhp;
-	struct iwch_qp *qhp;
-	union t3_wr *wqe;
-	u32 pbl_addr;
-	u8 page_size;
-	u32 num_wrs;
-	unsigned long flag;
-	struct ib_sge sgl;
-	int err=0;
-	enum t3_wr_flags t3_wr_flags;
-	u32 idx;
-	struct t3_swsq *sqp;
-
-	qhp = to_iwch_qp(qp);
-	mhp = to_iwch_mw(mw);
-	rhp = qhp->rhp;
-
-	spin_lock_irqsave(&qhp->lock, flag);
-	if (qhp->attr.state > IWCH_QP_STATE_RTS) {
-		spin_unlock_irqrestore(&qhp->lock, flag);
-		return -EINVAL;
-	}
-	num_wrs = Q_FREECNT(qhp->wq.sq_rptr, qhp->wq.sq_wptr,
-			    qhp->wq.sq_size_log2);
-	if (num_wrs == 0) {
-		spin_unlock_irqrestore(&qhp->lock, flag);
-		return -ENOMEM;
-	}
-	idx = Q_PTR2IDX(qhp->wq.wptr, qhp->wq.size_log2);
-	PDBG("%s: idx 0x%0x, mw 0x%p, mw_bind 0x%p\n", __func__, idx,
-	     mw, mw_bind);
-	wqe = (union t3_wr *) (qhp->wq.queue + idx);
-
-	t3_wr_flags = 0;
-	if (mw_bind->send_flags & IB_SEND_SIGNALED)
-		t3_wr_flags = T3_COMPLETION_FLAG;
-
-	sgl.addr = mw_bind->bind_info.addr;
-	sgl.lkey = mw_bind->bind_info.mr->lkey;
-	sgl.length = mw_bind->bind_info.length;
-	wqe->bind.reserved = 0;
-	wqe->bind.type = TPT_VATO;
-
-	/* TBD: check perms */
-	wqe->bind.perms = iwch_ib_to_tpt_bind_access(
-		mw_bind->bind_info.mw_access_flags);
-	wqe->bind.mr_stag = cpu_to_be32(mw_bind->bind_info.mr->lkey);
-	wqe->bind.mw_stag = cpu_to_be32(mw->rkey);
-	wqe->bind.mw_len = cpu_to_be32(mw_bind->bind_info.length);
-	wqe->bind.mw_va = cpu_to_be64(mw_bind->bind_info.addr);
-	err = iwch_sgl2pbl_map(rhp, &sgl, 1, &pbl_addr, &page_size);
-	if (err) {
-		spin_unlock_irqrestore(&qhp->lock, flag);
-		return err;
-	}
-	wqe->send.wrid.id0.hi = qhp->wq.sq_wptr;
-	sqp = qhp->wq.sq + Q_PTR2IDX(qhp->wq.sq_wptr, qhp->wq.sq_size_log2);
-	sqp->wr_id = mw_bind->wr_id;
-	sqp->opcode = T3_BIND_MW;
-	sqp->sq_wptr = qhp->wq.sq_wptr;
-	sqp->complete = 0;
-	sqp->signaled = (mw_bind->send_flags & IB_SEND_SIGNALED);
-	wqe->bind.mr_pbl_addr = cpu_to_be32(pbl_addr);
-	wqe->bind.mr_pagesz = page_size;
-	build_fw_riwrh((void *)wqe, T3_WR_BIND, t3_wr_flags,
-		       Q_GENBIT(qhp->wq.wptr, qhp->wq.size_log2), 0,
-		       sizeof(struct t3_bind_mw_wr) >> 3, T3_SOPEOP);
-	++(qhp->wq.wptr);
-	++(qhp->wq.sq_wptr);
-	spin_unlock_irqrestore(&qhp->lock, flag);
-
-	if (cxio_wq_db_enabled(&qhp->wq))
-		ring_doorbell(qhp->wq.doorbell, qhp->wq.qpid);
-
-	return err;
-}
-
 static inline void build_term_codes(struct respQ_msg_t *rsp_msg,
 				    u8 *layer_type, u8 *ecode)
 {
diff --git a/drivers/infiniband/hw/cxgb4/cq.c b/drivers/infiniband/hw/cxgb4/cq.c
index de9cd6901752..cf21df4a8bf5 100644
--- a/drivers/infiniband/hw/cxgb4/cq.c
+++ b/drivers/infiniband/hw/cxgb4/cq.c
@@ -744,9 +744,6 @@ static int c4iw_poll_cq_one(struct c4iw_cq *chp, struct ib_wc *wc)
 		case FW_RI_SEND_WITH_SE:
 			wc->opcode = IB_WC_SEND;
 			break;
-		case FW_RI_BIND_MW:
-			wc->opcode = IB_WC_BIND_MW;
-			break;
 
 		case FW_RI_LOCAL_INV:
 			wc->opcode = IB_WC_LOCAL_INV;
diff --git a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
index dd00cf2d0b65..fb2de75a0392 100644
--- a/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
+++ b/drivers/infiniband/hw/cxgb4/iw_cxgb4.h
@@ -947,8 +947,6 @@ int c4iw_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 		      struct ib_send_wr **bad_wr);
 int c4iw_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *wr,
 		      struct ib_recv_wr **bad_wr);
-int c4iw_bind_mw(struct ib_qp *qp, struct ib_mw *mw,
-		 struct ib_mw_bind *mw_bind);
 int c4iw_connect(struct iw_cm_id *cm_id, struct iw_cm_conn_param *conn_param);
 int c4iw_create_listen(struct iw_cm_id *cm_id, int backlog);
 int c4iw_destroy_listen(struct iw_cm_id *cm_id);
diff --git a/drivers/infiniband/hw/cxgb4/provider.c b/drivers/infiniband/hw/cxgb4/provider.c
index 249ea57e1d99..ec04272fbdc2 100644
--- a/drivers/infiniband/hw/cxgb4/provider.c
+++ b/drivers/infiniband/hw/cxgb4/provider.c
@@ -552,7 +552,6 @@ int c4iw_register_device(struct c4iw_dev *dev)
 	dev->ibdev.reg_user_mr = c4iw_reg_user_mr;
 	dev->ibdev.dereg_mr = c4iw_dereg_mr;
 	dev->ibdev.alloc_mw = c4iw_alloc_mw;
-	dev->ibdev.bind_mw = c4iw_bind_mw;
 	dev->ibdev.dealloc_mw = c4iw_dealloc_mw;
 	dev->ibdev.alloc_mr = c4iw_alloc_mr;
 	dev->ibdev.map_mr_sg = c4iw_map_mr_sg;
diff --git a/drivers/infiniband/hw/cxgb4/qp.c b/drivers/infiniband/hw/cxgb4/qp.c
index aa515afee724..e99345eb875a 100644
--- a/drivers/infiniband/hw/cxgb4/qp.c
+++ b/drivers/infiniband/hw/cxgb4/qp.c
@@ -933,11 +933,6 @@ int c4iw_post_receive(struct ib_qp *ibqp, struct ib_recv_wr *wr,
 	return err;
 }
 
-int c4iw_bind_mw(struct ib_qp *qp, struct ib_mw *mw, struct ib_mw_bind *mw_bind)
-{
-	return -ENOSYS;
-}
-
 static inline void build_term_codes(struct t4_cqe *err_cqe, u8 *layer_type,
 				    u8 *ecode)
 {
diff --git a/drivers/infiniband/hw/mlx4/cq.c b/drivers/infiniband/hw/mlx4/cq.c
index b88fc8f5ab18..9f8b516eb2b0 100644
--- a/drivers/infiniband/hw/mlx4/cq.c
+++ b/drivers/infiniband/hw/mlx4/cq.c
@@ -811,9 +811,6 @@ repoll:
 			wc->opcode    = IB_WC_MASKED_FETCH_ADD;
 			wc->byte_len  = 8;
 			break;
-		case MLX4_OPCODE_BIND_MW:
-			wc->opcode    = IB_WC_BIND_MW;
-			break;
 		case MLX4_OPCODE_LSO:
 			wc->opcode    = IB_WC_LSO;
 			break;
diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index 97d6878f9938..627267f860f9 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -2283,7 +2283,6 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
 	if (dev->caps.flags & MLX4_DEV_CAP_FLAG_MEM_WINDOW ||
 	    dev->caps.bmme_flags & MLX4_BMME_FLAG_TYPE_2_WIN) {
 		ibdev->ib_dev.alloc_mw = mlx4_ib_alloc_mw;
-		ibdev->ib_dev.bind_mw = mlx4_ib_bind_mw;
 		ibdev->ib_dev.dealloc_mw = mlx4_ib_dealloc_mw;
 
 		ibdev->ib_dev.uverbs_cmd_mask |=
diff --git a/drivers/infiniband/hw/mlx4/mlx4_ib.h b/drivers/infiniband/hw/mlx4/mlx4_ib.h
index 1caa11edac03..8916e9b55324 100644
--- a/drivers/infiniband/hw/mlx4/mlx4_ib.h
+++ b/drivers/infiniband/hw/mlx4/mlx4_ib.h
@@ -704,8 +704,6 @@ struct ib_mr *mlx4_ib_reg_user_mr(struct ib_pd *pd, u64 start, u64 length,
 				  struct ib_udata *udata);
 int mlx4_ib_dereg_mr(struct ib_mr *mr);
 struct ib_mw *mlx4_ib_alloc_mw(struct ib_pd *pd, enum ib_mw_type type);
-int mlx4_ib_bind_mw(struct ib_qp *qp, struct ib_mw *mw,
-		    struct ib_mw_bind *mw_bind);
 int mlx4_ib_dealloc_mw(struct ib_mw *mw);
 struct ib_mr *mlx4_ib_alloc_mr(struct ib_pd *pd,
 			       enum ib_mr_type mr_type,
diff --git a/drivers/infiniband/hw/mlx4/mr.c b/drivers/infiniband/hw/mlx4/mr.c
index 4d1e1c632603..242b94ec105b 100644
--- a/drivers/infiniband/hw/mlx4/mr.c
+++ b/drivers/infiniband/hw/mlx4/mr.c
@@ -366,28 +366,6 @@ err_free:
 	return ERR_PTR(err);
 }
 
-int mlx4_ib_bind_mw(struct ib_qp *qp, struct ib_mw *mw,
-		    struct ib_mw_bind *mw_bind)
-{
-	struct ib_bind_mw_wr  wr;
-	struct ib_send_wr *bad_wr;
-	int ret;
-
-	memset(&wr, 0, sizeof(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.rkey;
-
-	return ret;
-}
-
 int mlx4_ib_dealloc_mw(struct ib_mw *ibmw)
 {
 	struct mlx4_ib_mw *mw = to_mmw(ibmw);
diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c
index 6049644da429..3f5d2af5f8e4 100644
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -115,7 +115,6 @@ static const __be32 mlx4_ib_opcode[] = {
 	[IB_WR_REG_MR]				= cpu_to_be32(MLX4_OPCODE_FMR),
 	[IB_WR_MASKED_ATOMIC_CMP_AND_SWP]	= cpu_to_be32(MLX4_OPCODE_MASKED_ATOMIC_CS),
 	[IB_WR_MASKED_ATOMIC_FETCH_AND_ADD]	= cpu_to_be32(MLX4_OPCODE_MASKED_ATOMIC_FA),
-	[IB_WR_BIND_MW]				= cpu_to_be32(MLX4_OPCODE_BIND_MW),
 };
 
 static struct mlx4_ib_sqp *to_msqp(struct mlx4_ib_qp *mqp)
@@ -2531,25 +2530,6 @@ static void set_reg_seg(struct mlx4_wqe_fmr_seg *fseg,
 	fseg->reserved[1]	= 0;
 }
 
-static void set_bind_seg(struct mlx4_wqe_bind_seg *bseg,
-		struct ib_bind_mw_wr *wr)
-{
-	bseg->flags1 =
-		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->mw->type == IB_MW_TYPE_2)
-		bseg->flags2 |= cpu_to_be32(MLX4_WQE_BIND_TYPE_2);
-	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->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)
 {
 	memset(iseg, 0, sizeof(*iseg));
@@ -2870,13 +2850,6 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 				size += sizeof(struct mlx4_wqe_fmr_seg) / 16;
 				break;
 
-			case IB_WR_BIND_MW:
-				ctrl->srcrb_flags |=
-					cpu_to_be32(MLX4_WQE_CTRL_STRONG_ORDER);
-				set_bind_seg(wqe, bind_mw_wr(wr));
-				wqe  += sizeof(struct mlx4_wqe_bind_seg);
-				size += sizeof(struct mlx4_wqe_bind_seg) / 16;
-				break;
 			default:
 				/* No extra segments required for sends */
 				break;
diff --git a/drivers/infiniband/hw/mlx5/cq.c b/drivers/infiniband/hw/mlx5/cq.c
index 3ce5cfa7a4e0..db2270ad21b2 100644
--- a/drivers/infiniband/hw/mlx5/cq.c
+++ b/drivers/infiniband/hw/mlx5/cq.c
@@ -154,9 +154,6 @@ static void handle_good_req(struct ib_wc *wc, struct mlx5_cqe64 *cqe,
 		wc->opcode    = IB_WC_MASKED_FETCH_ADD;
 		wc->byte_len  = 8;
 		break;
-	case MLX5_OPCODE_BIND_MW:
-		wc->opcode    = IB_WC_BIND_MW;
-		break;
 	case MLX5_OPCODE_UMR:
 		wc->opcode = get_umr_comp(wq, idx);
 		break;
diff --git a/drivers/infiniband/hw/mthca/mthca_cq.c b/drivers/infiniband/hw/mthca/mthca_cq.c
index 40ba83338155..a6531ffe29a6 100644
--- a/drivers/infiniband/hw/mthca/mthca_cq.c
+++ b/drivers/infiniband/hw/mthca/mthca_cq.c
@@ -608,9 +608,6 @@ static inline int mthca_poll_one(struct mthca_dev *dev,
 			entry->opcode    = IB_WC_FETCH_ADD;
 			entry->byte_len  = MTHCA_ATOMIC_BYTE_LEN;
 			break;
-		case MTHCA_OPCODE_BIND_MW:
-			entry->opcode    = IB_WC_BIND_MW;
-			break;
 		default:
 			entry->opcode    = MTHCA_OPCODE_INVALID;
 			break;
diff --git a/drivers/infiniband/hw/nes/nes_verbs.c b/drivers/infiniband/hw/nes/nes_verbs.c
index f76358c39ebb..c8c661ebe3db 100644
--- a/drivers/infiniband/hw/nes/nes_verbs.c
+++ b/drivers/infiniband/hw/nes/nes_verbs.c
@@ -206,80 +206,6 @@ static int nes_dealloc_mw(struct ib_mw *ibmw)
 }
 
 
-/**
- * nes_bind_mw
- */
-static int nes_bind_mw(struct ib_qp *ibqp, struct ib_mw *ibmw,
-		struct ib_mw_bind *ibmw_bind)
-{
-	u64 u64temp;
-	struct nes_vnic *nesvnic = to_nesvnic(ibqp->device);
-	struct nes_device *nesdev = nesvnic->nesdev;
-	/* struct nes_mr *nesmr = to_nesmw(ibmw); */
-	struct nes_qp *nesqp = to_nesqp(ibqp);
-	struct nes_hw_qp_wqe *wqe;
-	unsigned long flags = 0;
-	u32 head;
-	u32 wqe_misc = 0;
-	u32 qsize;
-
-	if (nesqp->ibqp_state > IB_QPS_RTS)
-		return -EINVAL;
-
-	spin_lock_irqsave(&nesqp->lock, flags);
-
-	head = nesqp->hwqp.sq_head;
-	qsize = nesqp->hwqp.sq_tail;
-
-	/* Check for SQ overflow */
-	if (((head + (2 * qsize) - nesqp->hwqp.sq_tail) % qsize) == (qsize - 1)) {
-		spin_unlock_irqrestore(&nesqp->lock, flags);
-		return -ENOMEM;
-	}
-
-	wqe = &nesqp->hwqp.sq_vbase[head];
-	/* nes_debug(NES_DBG_MR, "processing sq wqe at %p, head = %u.\n", wqe, head); */
-	nes_fill_init_qp_wqe(wqe, nesqp, head);
-	u64temp = ibmw_bind->wr_id;
-	set_wqe_64bit_value(wqe->wqe_words, NES_IWARP_SQ_WQE_COMP_SCRATCH_LOW_IDX, u64temp);
-	wqe_misc = NES_IWARP_SQ_OP_BIND;
-
-	wqe_misc |= NES_IWARP_SQ_WQE_LOCAL_FENCE;
-
-	if (ibmw_bind->send_flags & IB_SEND_SIGNALED)
-		wqe_misc |= NES_IWARP_SQ_WQE_SIGNALED_COMPL;
-
-	if (ibmw_bind->bind_info.mw_access_flags & IB_ACCESS_REMOTE_WRITE)
-		wqe_misc |= NES_CQP_STAG_RIGHTS_REMOTE_WRITE;
-	if (ibmw_bind->bind_info.mw_access_flags & IB_ACCESS_REMOTE_READ)
-		wqe_misc |= NES_CQP_STAG_RIGHTS_REMOTE_READ;
-
-	set_wqe_32bit_value(wqe->wqe_words, NES_IWARP_SQ_WQE_MISC_IDX, wqe_misc);
-	set_wqe_32bit_value(wqe->wqe_words, NES_IWARP_SQ_BIND_WQE_MR_IDX,
-			    ibmw_bind->bind_info.mr->lkey);
-	set_wqe_32bit_value(wqe->wqe_words, NES_IWARP_SQ_BIND_WQE_MW_IDX, ibmw->rkey);
-	set_wqe_32bit_value(wqe->wqe_words, NES_IWARP_SQ_BIND_WQE_LENGTH_LOW_IDX,
-			ibmw_bind->bind_info.length);
-	wqe->wqe_words[NES_IWARP_SQ_BIND_WQE_LENGTH_HIGH_IDX] = 0;
-	u64temp = (u64)ibmw_bind->bind_info.addr;
-	set_wqe_64bit_value(wqe->wqe_words, NES_IWARP_SQ_BIND_WQE_VA_FBO_LOW_IDX, u64temp);
-
-	head++;
-	if (head >= qsize)
-		head = 0;
-
-	nesqp->hwqp.sq_head = head;
-	barrier();
-
-	nes_write32(nesdev->regs+NES_WQE_ALLOC,
-			(1 << 24) | 0x00800000 | nesqp->hwqp.qp_id);
-
-	spin_unlock_irqrestore(&nesqp->lock, flags);
-
-	return 0;
-}
-
-
 /*
  * nes_alloc_fast_mr
  */
@@ -3892,7 +3818,6 @@ struct nes_ib_device *nes_init_ofa_device(struct net_device *netdev)
 	nesibdev->ibdev.dereg_mr = nes_dereg_mr;
 	nesibdev->ibdev.alloc_mw = nes_alloc_mw;
 	nesibdev->ibdev.dealloc_mw = nes_dealloc_mw;
-	nesibdev->ibdev.bind_mw = nes_bind_mw;
 
 	nesibdev->ibdev.alloc_mr = nes_alloc_mr;
 	nesibdev->ibdev.map_mr_sg = nes_map_mr_sg;
diff --git a/drivers/staging/rdma/amso1100/c2_cq.c b/drivers/staging/rdma/amso1100/c2_cq.c
index 3ef881f2da0f..7ad0c082485a 100644
--- a/drivers/staging/rdma/amso1100/c2_cq.c
+++ b/drivers/staging/rdma/amso1100/c2_cq.c
@@ -173,9 +173,6 @@ static inline int c2_poll_one(struct c2_dev *c2dev,
 	case C2_WR_TYPE_RDMA_READ:
 		entry->opcode = IB_WC_RDMA_READ;
 		break;
-	case C2_WR_TYPE_BIND_MW:
-		entry->opcode = IB_WC_BIND_MW;
-		break;
 	case C2_WR_TYPE_RECV:
 		entry->byte_len = be32_to_cpu(ce->bytes_rcvd);
 		entry->opcode = IB_WC_RECV;
diff --git a/drivers/staging/rdma/ehca/ehca_iverbs.h b/drivers/staging/rdma/ehca/ehca_iverbs.h
index 17ab33b76882..cca5933fcda6 100644
--- a/drivers/staging/rdma/ehca/ehca_iverbs.h
+++ b/drivers/staging/rdma/ehca/ehca_iverbs.h
@@ -88,9 +88,6 @@ int ehca_dereg_mr(struct ib_mr *mr);
 
 struct ib_mw *ehca_alloc_mw(struct ib_pd *pd, enum ib_mw_type type);
 
-int ehca_bind_mw(struct ib_qp *qp, struct ib_mw *mw,
-		 struct ib_mw_bind *mw_bind);
-
 int ehca_dealloc_mw(struct ib_mw *mw);
 
 struct ib_fmr *ehca_alloc_fmr(struct ib_pd *pd,
diff --git a/drivers/staging/rdma/ehca/ehca_main.c b/drivers/staging/rdma/ehca/ehca_main.c
index 8e920a31d5e7..963274ee652a 100644
--- a/drivers/staging/rdma/ehca/ehca_main.c
+++ b/drivers/staging/rdma/ehca/ehca_main.c
@@ -515,7 +515,6 @@ static int ehca_init_device(struct ehca_shca *shca)
 	shca->ib_device.reg_user_mr	    = ehca_reg_user_mr;
 	shca->ib_device.dereg_mr	    = ehca_dereg_mr;
 	shca->ib_device.alloc_mw	    = ehca_alloc_mw;
-	shca->ib_device.bind_mw		    = ehca_bind_mw;
 	shca->ib_device.dealloc_mw	    = ehca_dealloc_mw;
 	shca->ib_device.alloc_fmr	    = ehca_alloc_fmr;
 	shca->ib_device.map_phys_fmr	    = ehca_map_phys_fmr;
diff --git a/drivers/staging/rdma/ehca/ehca_mrmw.c b/drivers/staging/rdma/ehca/ehca_mrmw.c
index 1c1a8dd09ac5..c6e324522566 100644
--- a/drivers/staging/rdma/ehca/ehca_mrmw.c
+++ b/drivers/staging/rdma/ehca/ehca_mrmw.c
@@ -413,18 +413,6 @@ alloc_mw_exit0:
 
 /*----------------------------------------------------------------------*/
 
-int ehca_bind_mw(struct ib_qp *qp,
-		 struct ib_mw *mw,
-		 struct ib_mw_bind *mw_bind)
-{
-	/* TODO: not supported up to now */
-	ehca_gen_err("bind MW currently not supported by HCAD");
-
-	return -EPERM;
-} /* end ehca_bind_mw() */
-
-/*----------------------------------------------------------------------*/
-
 int ehca_dealloc_mw(struct ib_mw *mw)
 {
 	u64 h_ret;
diff --git a/drivers/staging/rdma/ehca/ehca_reqs.c b/drivers/staging/rdma/ehca/ehca_reqs.c
index 10e2074384f5..11813b880e16 100644
--- a/drivers/staging/rdma/ehca/ehca_reqs.c
+++ b/drivers/staging/rdma/ehca/ehca_reqs.c
@@ -614,7 +614,6 @@ int ehca_post_srq_recv(struct ib_srq *srq,
 static const u8 ib_wc_opcode[255] = {
 	[0x01] = IB_WC_RECV+1,
 	[0x02] = IB_WC_RECV_RDMA_WITH_IMM+1,
-	[0x04] = IB_WC_BIND_MW+1,
 	[0x08] = IB_WC_FETCH_ADD+1,
 	[0x10] = IB_WC_COMP_SWAP+1,
 	[0x20] = IB_WC_RDMA_WRITE+1,
diff --git a/include/rdma/ib_verbs.h b/include/rdma/ib_verbs.h
index 86970f3e90b4..177844265c98 100644
--- a/include/rdma/ib_verbs.h
+++ b/include/rdma/ib_verbs.h
@@ -812,7 +812,6 @@ enum ib_wc_opcode {
 	IB_WC_RDMA_READ,
 	IB_WC_COMP_SWAP,
 	IB_WC_FETCH_ADD,
-	IB_WC_BIND_MW,
 	IB_WC_LSO,
 	IB_WC_LOCAL_INV,
 	IB_WC_REG_MR,
@@ -1110,7 +1109,6 @@ enum ib_wr_opcode {
 	IB_WR_REG_MR,
 	IB_WR_MASKED_ATOMIC_CMP_AND_SWP,
 	IB_WR_MASKED_ATOMIC_FETCH_AND_ADD,
-	IB_WR_BIND_MW,
 	IB_WR_REG_SIG_MR,
 	/* reserve values for low level drivers' internal use.
 	 * These values will not be used at all in the ib core layer.
@@ -1145,23 +1143,6 @@ struct ib_sge {
 	u32	lkey;
 };
 
-/**
- * struct ib_mw_bind_info - Parameters for a memory window bind operation.
- * @mr: A memory region to bind the memory window to.
- * @addr: The address where the memory window should begin.
- * @length: The length of the memory window, in bytes.
- * @mw_access_flags: Access flags from enum ib_access_flags for the window.
- *
- * This struct contains the shared parameters for type 1 and type 2
- * memory window bind operations.
- */
-struct ib_mw_bind_info {
-	struct ib_mr   *mr;
-	u64		addr;
-	u64		length;
-	int		mw_access_flags;
-};
-
 struct ib_cqe {
 	void (*done)(struct ib_cq *cq, struct ib_wc *wc);
 };
@@ -1237,19 +1218,6 @@ 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;
-	/* 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;
@@ -1299,18 +1267,6 @@ enum ib_mr_rereg_flags {
 	IB_MR_REREG_SUPPORTED	= ((IB_MR_REREG_ACCESS << 1) - 1)
 };
 
-/**
- * struct ib_mw_bind - Parameters for a type 1 memory window bind operation.
- * @wr_id:      Work request id.
- * @send_flags: Flags from ib_send_flags enum.
- * @bind_info:  More parameters of the bind operation.
- */
-struct ib_mw_bind {
-	u64                    wr_id;
-	int                    send_flags;
-	struct ib_mw_bind_info bind_info;
-};
-
 struct ib_fmr_attr {
 	int	max_pages;
 	int	max_maps;
@@ -1845,9 +1801,6 @@ struct ib_device {
 						int sg_nents);
 	struct ib_mw *             (*alloc_mw)(struct ib_pd *pd,
 					       enum ib_mw_type type);
-	int                        (*bind_mw)(struct ib_qp *qp,
-					      struct ib_mw *mw,
-					      struct ib_mw_bind *mw_bind);
 	int                        (*dealloc_mw)(struct ib_mw *mw);
 	struct ib_fmr *	           (*alloc_fmr)(struct ib_pd *pd,
 						int mr_access_flags,
@@ -2975,42 +2928,6 @@ static inline u32 ib_inc_rkey(u32 rkey)
 	return ((rkey + 1) & mask) | (rkey & ~mask);
 }
 
-/**
- * ib_alloc_mw - Allocates a memory window.
- * @pd: The protection domain associated with the memory window.
- * @type: The type of the memory window (1 or 2).
- */
-struct ib_mw *ib_alloc_mw(struct ib_pd *pd, enum ib_mw_type type);
-
-/**
- * ib_bind_mw - Posts a work request to the send queue of the specified
- *   QP, which binds the memory window to the given address range and
- *   remote access attributes.
- * @qp: QP to post the bind work request on.
- * @mw: The memory window to bind.
- * @mw_bind: Specifies information about the memory window, including
- *   its address range, remote access rights, and associated memory region.
- *
- * If there is no immediate error, the function will update the rkey member
- * of the mw parameter to its new value. The bind operation can still fail
- * asynchronously.
- */
-static inline int ib_bind_mw(struct ib_qp *qp,
-			     struct ib_mw *mw,
-			     struct ib_mw_bind *mw_bind)
-{
-	/* XXX reference counting in corresponding MR? */
-	return mw->device->bind_mw ?
-		mw->device->bind_mw(qp, mw, mw_bind) :
-		-ENOSYS;
-}
-
-/**
- * ib_dealloc_mw - Deallocates a memory window.
- * @mw: The memory window to deallocate.
- */
-int ib_dealloc_mw(struct ib_mw *mw);
-
 /**
  * ib_alloc_fmr - Allocates a unmapped fast memory region.
  * @pd: The protection domain associated with the unmapped region.
-- 
cgit v1.2.3


From 9afc60dcc406b96d22dfd7e5986d2a9ee44812f0 Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leonro@mellanox.com>
Date: Thu, 17 Dec 2015 09:31:52 +0200
Subject: IB/mlx4: Suppress non-fatal memory allocations

Failure in kmalloc memory allocations will throw a warning about it.
Such warnings are not needed anymore, since in commit 0ef2f05c7e02
("IB/mlx4: Use vmalloc for WR buffers when needed"), fallback mechanism
from kmalloc() to __vmalloc() was added.

Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Sagi Grimberg <sagig@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/qp.c  | 6 ++++--
 drivers/infiniband/hw/mlx4/srq.c | 3 ++-
 2 files changed, 6 insertions(+), 3 deletions(-)

(limited to 'drivers/infiniband/hw/mlx4')

diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c
index 3f5d2af5f8e4..9ed7b7fec5da 100644
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -795,11 +795,13 @@ static int create_qp_common(struct mlx4_ib_dev *dev, struct ib_pd *pd,
 		if (err)
 			goto err_mtt;
 
-		qp->sq.wrid = kmalloc(qp->sq.wqe_cnt * sizeof(u64), gfp);
+		qp->sq.wrid = kmalloc(qp->sq.wqe_cnt * sizeof(u64),
+					gfp | __GFP_NOWARN);
 		if (!qp->sq.wrid)
 			qp->sq.wrid = __vmalloc(qp->sq.wqe_cnt * sizeof(u64),
 						gfp, PAGE_KERNEL);
-		qp->rq.wrid = kmalloc(qp->rq.wqe_cnt * sizeof(u64), gfp);
+		qp->rq.wrid = kmalloc(qp->rq.wqe_cnt * sizeof(u64),
+					gfp | __GFP_NOWARN);
 		if (!qp->rq.wrid)
 			qp->rq.wrid = __vmalloc(qp->rq.wqe_cnt * sizeof(u64),
 						gfp, PAGE_KERNEL);
diff --git a/drivers/infiniband/hw/mlx4/srq.c b/drivers/infiniband/hw/mlx4/srq.c
index 8d133c40fa0e..f416c7463827 100644
--- a/drivers/infiniband/hw/mlx4/srq.c
+++ b/drivers/infiniband/hw/mlx4/srq.c
@@ -171,7 +171,8 @@ struct ib_srq *mlx4_ib_create_srq(struct ib_pd *pd,
 		if (err)
 			goto err_mtt;
 
-		srq->wrid = kmalloc(srq->msrq.max * sizeof (u64), GFP_KERNEL);
+		srq->wrid = kmalloc(srq->msrq.max * sizeof(u64),
+					GFP_KERNEL | __GFP_NOWARN);
 		if (!srq->wrid) {
 			srq->wrid = __vmalloc(srq->msrq.max * sizeof(u64),
 					      GFP_KERNEL, PAGE_KERNEL);
-- 
cgit v1.2.3


From ee37095044b85283c29cc4e7b254a7b30d411a9b Mon Sep 17 00:00:00 2001
From: Leon Romanovsky <leonro@mellanox.com>
Date: Thu, 17 Dec 2015 09:31:53 +0200
Subject: IB/mlx4: Convert kmalloc to kmalloc_array for checkpatch

Convert kmalloc to be kmalloc_array to fix warnings below:

WARNING: Prefer kmalloc_array over kmalloc with multiply
+               qp->sq.wrid = kmalloc(qp->sq.wqe_cnt * sizeof(u64),

WARNING: Prefer kmalloc_array over kmalloc with multiply
+               qp->rq.wrid = kmalloc(qp->rq.wqe_cnt * sizeof(u64),

WARNING: Prefer kmalloc_array over kmalloc with multiply
+               srq->wrid = kmalloc(srq->msrq.max * sizeof(u64),

Signed-off-by: Leon Romanovsky <leonro@mellanox.com>
Reviewed-by: Or Gerlitz <ogerlitz@mellanox.com>
Reviewed-by: Sagi Grimberg <sagig@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/qp.c  | 4 ++--
 drivers/infiniband/hw/mlx4/srq.c | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

(limited to 'drivers/infiniband/hw/mlx4')

diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c
index 9ed7b7fec5da..8d280594e24b 100644
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -795,12 +795,12 @@ static int create_qp_common(struct mlx4_ib_dev *dev, struct ib_pd *pd,
 		if (err)
 			goto err_mtt;
 
-		qp->sq.wrid = kmalloc(qp->sq.wqe_cnt * sizeof(u64),
+		qp->sq.wrid = kmalloc_array(qp->sq.wqe_cnt, sizeof(u64),
 					gfp | __GFP_NOWARN);
 		if (!qp->sq.wrid)
 			qp->sq.wrid = __vmalloc(qp->sq.wqe_cnt * sizeof(u64),
 						gfp, PAGE_KERNEL);
-		qp->rq.wrid = kmalloc(qp->rq.wqe_cnt * sizeof(u64),
+		qp->rq.wrid = kmalloc_array(qp->rq.wqe_cnt, sizeof(u64),
 					gfp | __GFP_NOWARN);
 		if (!qp->rq.wrid)
 			qp->rq.wrid = __vmalloc(qp->rq.wqe_cnt * sizeof(u64),
diff --git a/drivers/infiniband/hw/mlx4/srq.c b/drivers/infiniband/hw/mlx4/srq.c
index f416c7463827..68d5a5fda271 100644
--- a/drivers/infiniband/hw/mlx4/srq.c
+++ b/drivers/infiniband/hw/mlx4/srq.c
@@ -171,7 +171,7 @@ struct ib_srq *mlx4_ib_create_srq(struct ib_pd *pd,
 		if (err)
 			goto err_mtt;
 
-		srq->wrid = kmalloc(srq->msrq.max * sizeof(u64),
+		srq->wrid = kmalloc_array(srq->msrq.max, sizeof(u64),
 					GFP_KERNEL | __GFP_NOWARN);
 		if (!srq->wrid) {
 			srq->wrid = __vmalloc(srq->msrq.max * sizeof(u64),
-- 
cgit v1.2.3


From 4e4081673445485aa6bc90383bdb83e7a96cc48a Mon Sep 17 00:00:00 2001
From: Matan Barak <matanb@mellanox.com>
Date: Thu, 14 Jan 2016 17:47:02 +0200
Subject: IB/mlx4: Initialize hop_limit when creating address handle

Hop limit value wasn't copied from attributes  when ah was created.
This may influence packets for unconnected services to get dropped in
routers when endpoints are not in the same subnet.

Fixes: fa417f7b520e ("IB/mlx4: Add support for IBoE")
Signed-off-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/ah.c | 1 +
 1 file changed, 1 insertion(+)

(limited to 'drivers/infiniband/hw/mlx4')

diff --git a/drivers/infiniband/hw/mlx4/ah.c b/drivers/infiniband/hw/mlx4/ah.c
index 86af71351d9a..4fdb2416ee72 100644
--- a/drivers/infiniband/hw/mlx4/ah.c
+++ b/drivers/infiniband/hw/mlx4/ah.c
@@ -104,6 +104,7 @@ static struct ib_ah *create_iboe_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr
 	ah->av.eth.port_pd = cpu_to_be32(to_mpd(pd)->pdn | (ah_attr->port_num << 24));
 	ah->av.eth.gid_index = mlx4_ib_gid_index_to_real_index(ibdev, ah_attr->port_num, ah_attr->grh.sgid_index);
 	ah->av.eth.vlan = cpu_to_be16(vlan_tag);
+	ah->av.eth.hop_limit = ah_attr->grh.hop_limit;
 	if (ah_attr->static_rate) {
 		ah->av.eth.stat_rate = ah_attr->static_rate + MLX4_STAT_RATE_OFFSET;
 		while (ah->av.eth.stat_rate > IB_RATE_2_5_GBPS + MLX4_STAT_RATE_OFFSET &&
-- 
cgit v1.2.3


From 1049f138166ad6ca2e1c8edd2317a71f5b21b9fe Mon Sep 17 00:00:00 2001
From: Moni Shoua <monis@mellanox.com>
Date: Thu, 14 Jan 2016 17:47:38 +0200
Subject: IB/mlx4: Take source mac from AH instead from the port

In commit dbf727de7440 ("IB/core: Use GID table in AH creation and dmac
resolution") we copy source mac to mlx4_ah from the attributes of
gid at ib_ah_attr.grh.sgid_index. Now we can use it.

Signed-off-by: Moni Shoua <monis@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/qp.c | 24 ++----------------------
 1 file changed, 2 insertions(+), 22 deletions(-)

(limited to 'drivers/infiniband/hw/mlx4')

diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c
index 8d280594e24b..89f921f786b3 100644
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -32,6 +32,7 @@
  */
 
 #include <linux/log2.h>
+#include <linux/etherdevice.h>
 #include <linux/slab.h>
 #include <linux/netdevice.h>
 #include <linux/vmalloc.h>
@@ -2253,16 +2254,6 @@ static int build_sriov_qp0_header(struct mlx4_ib_sqp *sqp,
 	return 0;
 }
 
-static void mlx4_u64_to_smac(u8 *dst_mac, u64 src_mac)
-{
-	int i;
-
-	for (i = ETH_ALEN; i; i--) {
-		dst_mac[i - 1] = src_mac & 0xff;
-		src_mac >>= 8;
-	}
-}
-
 static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_ud_wr *wr,
 			    void *wqe, unsigned *mlx_seg_len)
 {
@@ -2384,27 +2375,16 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_ud_wr *wr,
 
 	if (is_eth) {
 		struct in6_addr in6;
-
 		u16 pcp = (be32_to_cpu(ah->av.ib.sl_tclass_flowlabel) >> 29) << 13;
 
 		mlx->sched_prio = cpu_to_be16(pcp);
 
+		ether_addr_copy(sqp->ud_header.eth.smac_h, ah->av.eth.s_mac);
 		memcpy(sqp->ud_header.eth.dmac_h, ah->av.eth.mac, 6);
-		/* FIXME: cache smac value? */
 		memcpy(&ctrl->srcrb_flags16[0], ah->av.eth.mac, 2);
 		memcpy(&ctrl->imm, ah->av.eth.mac + 2, 4);
 		memcpy(&in6, sgid.raw, sizeof(in6));
 
-		if (!mlx4_is_mfunc(to_mdev(ib_dev)->dev)) {
-			u64 mac = atomic64_read(&to_mdev(ib_dev)->iboe.mac[sqp->qp.port - 1]);
-			u8 smac[ETH_ALEN];
-
-			mlx4_u64_to_smac(smac, mac);
-			memcpy(sqp->ud_header.eth.smac_h, smac, ETH_ALEN);
-		} else {
-			/* use the src mac of the tunnel */
-			memcpy(sqp->ud_header.eth.smac_h, ah->av.eth.s_mac, ETH_ALEN);
-		}
 
 		if (!memcmp(sqp->ud_header.eth.smac_h, sqp->ud_header.eth.dmac_h, 6))
 			mlx->flags |= cpu_to_be32(MLX4_WQE_CTRL_FORCE_LOOPBACK);
-- 
cgit v1.2.3


From 39f426553ea30f4449d022577fd6494c265ef3ef Mon Sep 17 00:00:00 2001
From: Lucas Tanure <tanure@linux.com>
Date: Tue, 19 Jan 2016 12:06:30 -0200
Subject: infiniband: Replace memset with eth_zero_addr

Use eth_zero_addr to assign the zero address to the given address
array instead of memset when second argument is address of zero.

Signed-off-by: Lucas Tanure <tanure@linux.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/ah.c       | 2 +-
 drivers/infiniband/hw/nes/nes_utils.c | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

(limited to 'drivers/infiniband/hw/mlx4')

diff --git a/drivers/infiniband/hw/mlx4/ah.c b/drivers/infiniband/hw/mlx4/ah.c
index 4fdb2416ee72..105246fba2e7 100644
--- a/drivers/infiniband/hw/mlx4/ah.c
+++ b/drivers/infiniband/hw/mlx4/ah.c
@@ -92,7 +92,7 @@ static struct ib_ah *create_iboe_ah(struct ib_pd *pd, struct ib_ah_attr *ah_attr
 				ah_attr->grh.sgid_index, &sgid, &gid_attr);
 	if (ret)
 		return ERR_PTR(ret);
-	memset(ah->av.eth.s_mac, 0, ETH_ALEN);
+	eth_zero_addr(ah->av.eth.s_mac);
 	if (gid_attr.ndev) {
 		if (is_vlan_dev(gid_attr.ndev))
 			vlan_tag = vlan_dev_vlan_id(gid_attr.ndev);
diff --git a/drivers/infiniband/hw/nes/nes_utils.c b/drivers/infiniband/hw/nes/nes_utils.c
index 2042c0f29759..6d3a169c049b 100644
--- a/drivers/infiniband/hw/nes/nes_utils.c
+++ b/drivers/infiniband/hw/nes/nes_utils.c
@@ -727,7 +727,7 @@ int nes_arp_table(struct nes_device *nesdev, u32 ip_addr, u8 *mac_addr, u32 acti
 	if (action == NES_ARP_DELETE) {
 		nes_debug(NES_DBG_NETDEV, "DELETE, arp_index=%d\n", arp_index);
 		nesadapter->arp_table[arp_index].ip_addr = 0;
-		memset(nesadapter->arp_table[arp_index].mac_addr, 0x00, ETH_ALEN);
+		eth_zero_addr(nesadapter->arp_table[arp_index].mac_addr);
 		nes_free_resource(nesadapter, nesadapter->allocated_arps, arp_index);
 		return arp_index;
 	}
-- 
cgit v1.2.3


From b699a859d17b3a7a8edc5441ee99d245dfd9f86c Mon Sep 17 00:00:00 2001
From: Moni Shoua <monis@mellanox.com>
Date: Thu, 14 Jan 2016 17:50:33 +0200
Subject: IB/mlx4: Add gid_type to GID properties

IB core driver adds a property of type to struct ib_gid_attr.
The mlx4 driver should take that in consideration when modifying or
querying the hardware gid table.

Signed-off-by: Moni Shoua <monis@mellanox.com>
Signed-off-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/main.c    | 17 +++++++++++++----
 drivers/infiniband/hw/mlx4/mlx4_ib.h |  1 +
 2 files changed, 14 insertions(+), 4 deletions(-)

(limited to 'drivers/infiniband/hw/mlx4')

diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index 627267f860f9..b179909ac875 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -215,7 +215,8 @@ static int mlx4_ib_add_gid(struct ib_device *device,
 	port_gid_table = &iboe->gids[port_num - 1];
 	spin_lock_bh(&iboe->lock);
 	for (i = 0; i < MLX4_MAX_PORT_GIDS; ++i) {
-		if (!memcmp(&port_gid_table->gids[i].gid, gid, sizeof(*gid))) {
+		if (!memcmp(&port_gid_table->gids[i].gid, gid, sizeof(*gid)) &&
+		    (port_gid_table->gids[i].gid_type == attr->gid_type))  {
 			found = i;
 			break;
 		}
@@ -233,6 +234,7 @@ static int mlx4_ib_add_gid(struct ib_device *device,
 			} else {
 				*context = port_gid_table->gids[free].ctx;
 				memcpy(&port_gid_table->gids[free].gid, gid, sizeof(*gid));
+				port_gid_table->gids[free].gid_type = attr->gid_type;
 				port_gid_table->gids[free].ctx->real_index = free;
 				port_gid_table->gids[free].ctx->refcount = 1;
 				hw_update = 1;
@@ -248,8 +250,10 @@ static int mlx4_ib_add_gid(struct ib_device *device,
 		if (!gids) {
 			ret = -ENOMEM;
 		} else {
-			for (i = 0; i < MLX4_MAX_PORT_GIDS; i++)
+			for (i = 0; i < MLX4_MAX_PORT_GIDS; i++) {
 				memcpy(&gids[i].gid, &port_gid_table->gids[i].gid, sizeof(union ib_gid));
+				gids[i].gid_type = port_gid_table->gids[i].gid_type;
+			}
 		}
 	}
 	spin_unlock_bh(&iboe->lock);
@@ -325,6 +329,7 @@ int mlx4_ib_gid_index_to_real_index(struct mlx4_ib_dev *ibdev,
 	int i;
 	int ret;
 	unsigned long flags;
+	struct ib_gid_attr attr;
 
 	if (port_num > MLX4_MAX_PORTS)
 		return -EINVAL;
@@ -335,10 +340,13 @@ 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, NULL);
+	ret = ib_get_cached_gid(&ibdev->ib_dev, port_num, index, &gid, &attr);
 	if (ret)
 		return ret;
 
+	if (attr.ndev)
+		dev_put(attr.ndev);
+
 	if (!memcmp(&gid, &zgid, sizeof(gid)))
 		return -EINVAL;
 
@@ -346,7 +354,8 @@ int mlx4_ib_gid_index_to_real_index(struct mlx4_ib_dev *ibdev,
 	port_gid_table = &iboe->gids[port_num - 1];
 
 	for (i = 0; i < MLX4_MAX_PORT_GIDS; ++i)
-		if (!memcmp(&port_gid_table->gids[i].gid, &gid, sizeof(gid))) {
+		if (!memcmp(&port_gid_table->gids[i].gid, &gid, sizeof(gid)) &&
+		    attr.gid_type == port_gid_table->gids[i].gid_type) {
 			ctx = port_gid_table->gids[i].ctx;
 			break;
 		}
diff --git a/drivers/infiniband/hw/mlx4/mlx4_ib.h b/drivers/infiniband/hw/mlx4/mlx4_ib.h
index 8916e9b55324..7179fb15edac 100644
--- a/drivers/infiniband/hw/mlx4/mlx4_ib.h
+++ b/drivers/infiniband/hw/mlx4/mlx4_ib.h
@@ -478,6 +478,7 @@ struct gid_cache_context {
 
 struct gid_entry {
 	union ib_gid	gid;
+	enum ib_gid_type gid_type;
 	struct gid_cache_context *ctx;
 };
 
-- 
cgit v1.2.3


From 7e57b85c444c3c1bf3550aa6890666fc4353bd33 Mon Sep 17 00:00:00 2001
From: Moni Shoua <monis@mellanox.com>
Date: Thu, 14 Jan 2016 17:50:35 +0200
Subject: IB/mlx4: Add support for setting RoCEv2 gids in hardware

To tell hardware about a gid with type RoCEv2, software needs a new
modifier to the SET_PORT command: MLX4_SET_PORT_ROCE_ADDR. This can
replace the old method, MLX4_SET_PORT_GID_TABLE, for  RoCEv1 gids.

Signed-off-by: Moni Shoua <monis@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/main.c | 63 +++++++++++++++++++++++++++++++++++++--
 include/linux/mlx4/cmd.h          |  3 +-
 2 files changed, 62 insertions(+), 4 deletions(-)

(limited to 'drivers/infiniband/hw/mlx4')

diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index b179909ac875..edf3b1c0523d 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -154,9 +154,9 @@ static struct net_device *mlx4_ib_get_netdev(struct ib_device *device, u8 port_n
 	return dev;
 }
 
-static int mlx4_ib_update_gids(struct gid_entry *gids,
-			       struct mlx4_ib_dev *ibdev,
-			       u8 port_num)
+static int mlx4_ib_update_gids_v1(struct gid_entry *gids,
+				  struct mlx4_ib_dev *ibdev,
+				  u8 port_num)
 {
 	struct mlx4_cmd_mailbox *mailbox;
 	int err;
@@ -187,6 +187,63 @@ static int mlx4_ib_update_gids(struct gid_entry *gids,
 	return err;
 }
 
+static int mlx4_ib_update_gids_v1_v2(struct gid_entry *gids,
+				     struct mlx4_ib_dev *ibdev,
+				     u8 port_num)
+{
+	struct mlx4_cmd_mailbox *mailbox;
+	int err;
+	struct mlx4_dev *dev = ibdev->dev;
+	int i;
+	struct {
+		union ib_gid	gid;
+		__be32		rsrvd1[2];
+		__be16		rsrvd2;
+		u8		type;
+		u8		version;
+		__be32		rsrvd3;
+	} *gid_tbl;
+
+	mailbox = mlx4_alloc_cmd_mailbox(dev);
+	if (IS_ERR(mailbox))
+		return -ENOMEM;
+
+	gid_tbl = mailbox->buf;
+	for (i = 0; i < MLX4_MAX_PORT_GIDS; ++i) {
+		memcpy(&gid_tbl[i].gid, &gids[i].gid, sizeof(union ib_gid));
+		if (gids[i].gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP) {
+			gid_tbl[i].version = 2;
+			if (!ipv6_addr_v4mapped((struct in6_addr *)&gids[i].gid))
+				gid_tbl[i].type = 1;
+			else
+				memset(&gid_tbl[i].gid, 0, 12);
+		}
+	}
+
+	err = mlx4_cmd(dev, mailbox->dma,
+		       MLX4_SET_PORT_ROCE_ADDR << 8 | port_num,
+		       1, MLX4_CMD_SET_PORT, MLX4_CMD_TIME_CLASS_B,
+		       MLX4_CMD_WRAPPED);
+	if (mlx4_is_bonded(dev))
+		err += mlx4_cmd(dev, mailbox->dma,
+				MLX4_SET_PORT_ROCE_ADDR << 8 | 2,
+				1, MLX4_CMD_SET_PORT, MLX4_CMD_TIME_CLASS_B,
+				MLX4_CMD_WRAPPED);
+
+	mlx4_free_cmd_mailbox(dev, mailbox);
+	return err;
+}
+
+static int mlx4_ib_update_gids(struct gid_entry *gids,
+			       struct mlx4_ib_dev *ibdev,
+			       u8 port_num)
+{
+	if (ibdev->dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_ROCE_V1_V2)
+		return mlx4_ib_update_gids_v1_v2(gids, ibdev, port_num);
+
+	return mlx4_ib_update_gids_v1(gids, ibdev, port_num);
+}
+
 static int mlx4_ib_add_gid(struct ib_device *device,
 			   u8 port_num,
 			   unsigned int index,
diff --git a/include/linux/mlx4/cmd.h b/include/linux/mlx4/cmd.h
index 58391f2e0414..116b284bc4ce 100644
--- a/include/linux/mlx4/cmd.h
+++ b/include/linux/mlx4/cmd.h
@@ -206,7 +206,8 @@ enum {
 	MLX4_SET_PORT_GID_TABLE = 0x5,
 	MLX4_SET_PORT_PRIO2TC	= 0x8,
 	MLX4_SET_PORT_SCHEDULER = 0x9,
-	MLX4_SET_PORT_VXLAN	= 0xB
+	MLX4_SET_PORT_VXLAN	= 0xB,
+	MLX4_SET_PORT_ROCE_ADDR	= 0xD
 };
 
 enum {
-- 
cgit v1.2.3


From 3b5daf28ac4bb9354b7d2f10ce5942cad23e979a Mon Sep 17 00:00:00 2001
From: Moni Shoua <monis@mellanox.com>
Date: Thu, 14 Jan 2016 17:50:39 +0200
Subject: IB/mlx4: Support modify_qp for RoCE v2

In order to support modify_qp for RoCE v2, we need to set
the gid_type in the QP context.

Signed-off-by: Moni Shoua <monis@mellanox.com>
Signed-off-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/qp.c | 37 ++++++++++++++++++++++++++++++++++---
 include/linux/mlx4/qp.h         |  2 +-
 2 files changed, 35 insertions(+), 4 deletions(-)

(limited to 'drivers/infiniband/hw/mlx4')

diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c
index 89f921f786b3..a9433a52a6f0 100644
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -1509,6 +1509,24 @@ static int create_qp_lb_counter(struct mlx4_ib_dev *dev, struct mlx4_ib_qp *qp)
 	return 0;
 }
 
+enum {
+	MLX4_QPC_ROCE_MODE_1 = 0,
+	MLX4_QPC_ROCE_MODE_2 = 2,
+	MLX4_QPC_ROCE_MODE_UNDEFINED = 0xff
+};
+
+static u8 gid_type_to_qpc(enum ib_gid_type gid_type)
+{
+	switch (gid_type) {
+	case IB_GID_TYPE_ROCE:
+		return MLX4_QPC_ROCE_MODE_1;
+	case IB_GID_TYPE_ROCE_UDP_ENCAP:
+		return MLX4_QPC_ROCE_MODE_2;
+	default:
+		return MLX4_QPC_ROCE_MODE_UNDEFINED;
+	}
+}
+
 static int __mlx4_ib_modify_qp(struct ib_qp *ibqp,
 			       const struct ib_qp_attr *attr, int attr_mask,
 			       enum ib_qp_state cur_state, enum ib_qp_state new_state)
@@ -1652,9 +1670,10 @@ static int __mlx4_ib_modify_qp(struct ib_qp *ibqp,
 		u16 vlan = 0xffff;
 		u8 smac[ETH_ALEN];
 		int status = 0;
+		int is_eth = rdma_cap_eth_ah(&dev->ib_dev, port_num) &&
+			attr->ah_attr.ah_flags & IB_AH_GRH;
 
-		if (rdma_cap_eth_ah(&dev->ib_dev, port_num) &&
-		    attr->ah_attr.ah_flags & IB_AH_GRH) {
+		if (is_eth) {
 			int index = attr->ah_attr.grh.sgid_index;
 
 			status = ib_get_cached_gid(ibqp->device, port_num,
@@ -1676,6 +1695,18 @@ static int __mlx4_ib_modify_qp(struct ib_qp *ibqp,
 
 		optpar |= (MLX4_QP_OPTPAR_PRIMARY_ADDR_PATH |
 			   MLX4_QP_OPTPAR_SCHED_QUEUE);
+
+		if (is_eth &&
+		    (cur_state == IB_QPS_INIT && new_state == IB_QPS_RTR)) {
+			u8 qpc_roce_mode = gid_type_to_qpc(gid_attr.gid_type);
+
+			if (qpc_roce_mode == MLX4_QPC_ROCE_MODE_UNDEFINED) {
+				err = -EINVAL;
+				goto out;
+			}
+			context->rlkey_roce_mode |= (qpc_roce_mode << 6);
+		}
+
 	}
 
 	if (attr_mask & IB_QP_TIMEOUT) {
@@ -1847,7 +1878,7 @@ static int __mlx4_ib_modify_qp(struct ib_qp *ibqp,
 		sqd_event = 0;
 
 	if (!ibqp->uobject && cur_state == IB_QPS_RESET && new_state == IB_QPS_INIT)
-		context->rlkey |= (1 << 4);
+		context->rlkey_roce_mode |= (1 << 4);
 
 	/*
 	 * Before passing a kernel QP to the HW, make sure that the
diff --git a/include/linux/mlx4/qp.h b/include/linux/mlx4/qp.h
index cdf110d3f260..587cdf943b52 100644
--- a/include/linux/mlx4/qp.h
+++ b/include/linux/mlx4/qp.h
@@ -194,7 +194,7 @@ struct mlx4_qp_context {
 	u8			mtu_msgmax;
 	u8			rq_size_stride;
 	u8			sq_size_stride;
-	u8			rlkey;
+	u8			rlkey_roce_mode;
 	__be32			usr_page;
 	__be32			local_qpn;
 	__be32			remote_qpn;
-- 
cgit v1.2.3


From 71a39bbbfc5931310073ecbbdbe6d0fa92501fe4 Mon Sep 17 00:00:00 2001
From: Moni Shoua <monis@mellanox.com>
Date: Thu, 14 Jan 2016 17:50:40 +0200
Subject: IB/mlx4: Enable RoCE v2 when the IB device is added

If the hardware supports RoCE v2, we configure the hardware UDP
port according to the RoCE v2 Annex when mlx4_ib device is added.

Signed-off-by: Moni Shoua <monis@mellanox.com>
Signed-off-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/main.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

(limited to 'drivers/infiniband/hw/mlx4')

diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index edf3b1c0523d..8bc0d608a9ca 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -2488,7 +2488,8 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
 	if (mlx4_ib_init_sriov(ibdev))
 		goto err_mad;
 
-	if (dev->caps.flags & MLX4_DEV_CAP_FLAG_IBOE) {
+	if (dev->caps.flags & MLX4_DEV_CAP_FLAG_IBOE ||
+	    dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_ROCE_V1_V2) {
 		if (!iboe->nb.notifier_call) {
 			iboe->nb.notifier_call = mlx4_ib_netdev_event;
 			err = register_netdevice_notifier(&iboe->nb);
@@ -2497,6 +2498,12 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
 				goto err_notif;
 			}
 		}
+		if (dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_ROCE_V1_V2) {
+			err = mlx4_config_roce_v2_port(dev, ROCE_V2_UDP_DPORT);
+			if (err) {
+				goto err_notif;
+			}
+		}
 	}
 
 	for (j = 0; j < ARRAY_SIZE(mlx4_class_attributes); ++j) {
-- 
cgit v1.2.3


From 3ef967a4affeef7bb3b7713dcfed6518b99737c6 Mon Sep 17 00:00:00 2001
From: Moni Shoua <monis@mellanox.com>
Date: Thu, 14 Jan 2016 17:50:41 +0200
Subject: IB/mlx4: Enable send of RoCE QP1 packets with IP/UDP headers

RoCEv2 packets are sent over IP/UDP protocols.
The mlx4 driver uses a type of RAW QP to send packets for QP1 and
therefore needs to build the network headers below BTH in software.

This patch adds option to build QP1 packets with IP and UDP headers if
RoCEv2 is requested.

Signed-off-by: Moni Shoua <monis@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/core/ud_header.c |  2 +-
 drivers/infiniband/hw/mlx4/qp.c     | 60 +++++++++++++++++++++++++++++++------
 include/rdma/ib_pack.h              |  4 +--
 3 files changed, 54 insertions(+), 12 deletions(-)

(limited to 'drivers/infiniband/hw/mlx4')

diff --git a/drivers/infiniband/core/ud_header.c b/drivers/infiniband/core/ud_header.c
index 96697e7378e2..19837d270278 100644
--- a/drivers/infiniband/core/ud_header.c
+++ b/drivers/infiniband/core/ud_header.c
@@ -280,7 +280,7 @@ static const struct ib_field deth_table[] = {
 	  .size_bits    = 24 }
 };
 
-__be16 ib_ud_ip4_csum(struct ib_ud_header *header)
+__sum16 ib_ud_ip4_csum(struct ib_ud_header *header)
 {
 	struct iphdr iph;
 
diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c
index a9433a52a6f0..7d1024655305 100644
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -33,6 +33,7 @@
 
 #include <linux/log2.h>
 #include <linux/etherdevice.h>
+#include <net/ip.h>
 #include <linux/slab.h>
 #include <linux/netdevice.h>
 #include <linux/vmalloc.h>
@@ -2285,6 +2286,7 @@ static int build_sriov_qp0_header(struct mlx4_ib_sqp *sqp,
 	return 0;
 }
 
+#define MLX4_ROCEV2_QP1_SPORT 0xC000
 static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_ud_wr *wr,
 			    void *wqe, unsigned *mlx_seg_len)
 {
@@ -2304,6 +2306,8 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_ud_wr *wr,
 	bool is_eth;
 	bool is_vlan = false;
 	bool is_grh;
+	bool is_udp = false;
+	int ip_version = 0;
 
 	send_size = 0;
 	for (i = 0; i < wr->wr.num_sge; ++i)
@@ -2312,6 +2316,8 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_ud_wr *wr,
 	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);
 	if (is_eth) {
+		struct ib_gid_attr gid_attr;
+
 		if (mlx4_is_mfunc(to_mdev(ib_dev)->dev)) {
 			/* When multi-function is enabled, the ib_core gid
 			 * indexes don't necessarily match the hw ones, so
@@ -2325,20 +2331,33 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_ud_wr *wr,
 			err = ib_get_cached_gid(ib_dev,
 						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)
+						&gid_attr);
+			if (!err) {
+				if (gid_attr.ndev)
+					dev_put(gid_attr.ndev);
+				if (!memcmp(&sgid, &zgid, sizeof(sgid)))
+					err = -ENOENT;
+			}
+			if (!err) {
+				is_udp = gid_attr.gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP;
+				if (is_udp) {
+					if (ipv6_addr_v4mapped((struct in6_addr *)&sgid))
+						ip_version = 4;
+					else
+						ip_version = 6;
+					is_grh = false;
+				}
+			} else {
 				return err;
+			}
 		}
-
 		if (ah->av.eth.vlan != cpu_to_be16(0xffff)) {
 			vlan = be16_to_cpu(ah->av.eth.vlan) & 0x0fff;
 			is_vlan = 1;
 		}
 	}
 	err = ib_ud_header_init(send_size, !is_eth, is_eth, is_vlan, is_grh,
-				0, 0, 0, &sqp->ud_header);
+			  ip_version, is_udp, 0, &sqp->ud_header);
 	if (err)
 		return err;
 
@@ -2349,7 +2368,7 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_ud_wr *wr,
 		sqp->ud_header.lrh.source_lid = cpu_to_be16(ah->av.ib.g_slid & 0x7f);
 	}
 
-	if (is_grh) {
+	if (is_grh || (ip_version == 6)) {
 		sqp->ud_header.grh.traffic_class =
 			(be32_to_cpu(ah->av.ib.sl_tclass_flowlabel) >> 20) & 0xff;
 		sqp->ud_header.grh.flow_label    =
@@ -2378,6 +2397,25 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_ud_wr *wr,
 		       ah->av.ib.dgid, 16);
 	}
 
+	if (ip_version == 4) {
+		sqp->ud_header.ip4.tos =
+			(be32_to_cpu(ah->av.ib.sl_tclass_flowlabel) >> 20) & 0xff;
+		sqp->ud_header.ip4.id = 0;
+		sqp->ud_header.ip4.frag_off = htons(IP_DF);
+		sqp->ud_header.ip4.ttl = ah->av.eth.hop_limit;
+
+		memcpy(&sqp->ud_header.ip4.saddr,
+		       sgid.raw + 12, 4);
+		memcpy(&sqp->ud_header.ip4.daddr, ah->av.ib.dgid + 12, 4);
+		sqp->ud_header.ip4.check = ib_ud_ip4_csum(&sqp->ud_header);
+	}
+
+	if (is_udp) {
+		sqp->ud_header.udp.dport = htons(ROCE_V2_UDP_DPORT);
+		sqp->ud_header.udp.sport = htons(MLX4_ROCEV2_QP1_SPORT);
+		sqp->ud_header.udp.csum = 0;
+	}
+
 	mlx->flags &= cpu_to_be32(MLX4_WQE_CTRL_CQ_UPDATE);
 
 	if (!is_eth) {
@@ -2406,8 +2444,12 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_ud_wr *wr,
 
 	if (is_eth) {
 		struct in6_addr in6;
+		u16 ether_type;
 		u16 pcp = (be32_to_cpu(ah->av.ib.sl_tclass_flowlabel) >> 29) << 13;
 
+		ether_type = (!is_udp) ? MLX4_IB_IBOE_ETHERTYPE :
+			(ip_version == 4 ? ETH_P_IP : ETH_P_IPV6);
+
 		mlx->sched_prio = cpu_to_be16(pcp);
 
 		ether_addr_copy(sqp->ud_header.eth.smac_h, ah->av.eth.s_mac);
@@ -2420,9 +2462,9 @@ static int build_mlx_header(struct mlx4_ib_sqp *sqp, struct ib_ud_wr *wr,
 		if (!memcmp(sqp->ud_header.eth.smac_h, sqp->ud_header.eth.dmac_h, 6))
 			mlx->flags |= cpu_to_be32(MLX4_WQE_CTRL_FORCE_LOOPBACK);
 		if (!is_vlan) {
-			sqp->ud_header.eth.type = cpu_to_be16(MLX4_IB_IBOE_ETHERTYPE);
+			sqp->ud_header.eth.type = cpu_to_be16(ether_type);
 		} else {
-			sqp->ud_header.vlan.type = cpu_to_be16(MLX4_IB_IBOE_ETHERTYPE);
+			sqp->ud_header.vlan.type = cpu_to_be16(ether_type);
 			sqp->ud_header.vlan.tag = cpu_to_be16(vlan | pcp);
 		}
 	} else {
diff --git a/include/rdma/ib_pack.h b/include/rdma/ib_pack.h
index a1930819b97e..0f3daae44bf9 100644
--- a/include/rdma/ib_pack.h
+++ b/include/rdma/ib_pack.h
@@ -234,7 +234,7 @@ struct ib_unpacked_ip4 {
 	__be16	frag_off;
 	u8	ttl;
 	u8	protocol;
-	__be16	check;
+	__sum16	check;
 	__be32	saddr;
 	__be32	daddr;
 };
@@ -280,7 +280,7 @@ void ib_unpack(const struct ib_field        *desc,
 	       void                         *buf,
 	       void                         *structure);
 
-__be16 ib_ud_ip4_csum(struct ib_ud_header *header);
+__sum16 ib_ud_ip4_csum(struct ib_ud_header *header);
 
 int ib_ud_header_init(int		    payload_bytes,
 		      int		    lrh_present,
-- 
cgit v1.2.3


From e1b866c677d69334e2981b19696b7e3d88dff707 Mon Sep 17 00:00:00 2001
From: Moni Shoua <monis@mellanox.com>
Date: Thu, 14 Jan 2016 17:50:42 +0200
Subject: IB/mlx4: Create and use another QP1 for RoCEv2

The mlx4 driver uses a special QP to implement the GSI QP. This kind
of QP allows to build the InfiniBand headers in software.
When mlx4 hardware builds the packet, it calculates the ICRC and puts
it at the end of the payload. However, this ICRC calculation depends
on the QP configuration, which is determined when the QP is modified
(roce_mode during INIT->RTR).
When receiving a packet, the ICRC verification doesn't depend on this
configuration.
Therefore, using two GSI QPs for send (one for each RoCE version) and
one GSI QP for receive are required.

Signed-off-by: Moni Shoua <monis@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/mlx4_ib.h |   7 ++
 drivers/infiniband/hw/mlx4/qp.c      | 159 ++++++++++++++++++++++++++++++-----
 2 files changed, 147 insertions(+), 19 deletions(-)

(limited to 'drivers/infiniband/hw/mlx4')

diff --git a/drivers/infiniband/hw/mlx4/mlx4_ib.h b/drivers/infiniband/hw/mlx4/mlx4_ib.h
index 7179fb15edac..52ce7b000044 100644
--- a/drivers/infiniband/hw/mlx4/mlx4_ib.h
+++ b/drivers/infiniband/hw/mlx4/mlx4_ib.h
@@ -177,11 +177,18 @@ struct mlx4_ib_wq {
 	unsigned		tail;
 };
 
+enum {
+	MLX4_IB_QP_CREATE_ROCE_V2_GSI = IB_QP_CREATE_RESERVED_START
+};
+
 enum mlx4_ib_qp_flags {
 	MLX4_IB_QP_LSO = IB_QP_CREATE_IPOIB_UD_LSO,
 	MLX4_IB_QP_BLOCK_MULTICAST_LOOPBACK = IB_QP_CREATE_BLOCK_MULTICAST_LOOPBACK,
 	MLX4_IB_QP_NETIF = IB_QP_CREATE_NETIF_QP,
 	MLX4_IB_QP_CREATE_USE_GFP_NOIO = IB_QP_CREATE_USE_GFP_NOIO,
+
+	/* Mellanox specific flags start from IB_QP_CREATE_RESERVED_START */
+	MLX4_IB_ROCE_V2_GSI_QP = MLX4_IB_QP_CREATE_ROCE_V2_GSI,
 	MLX4_IB_SRIOV_TUNNEL_QP = 1 << 30,
 	MLX4_IB_SRIOV_SQP = 1 << 31,
 };
diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c
index 7d1024655305..bc5536f00b6c 100644
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -87,6 +87,7 @@ struct mlx4_ib_sqp {
 	u32			send_psn;
 	struct ib_ud_header	ud_header;
 	u8			header_buf[MLX4_IB_UD_HEADER_SIZE];
+	struct ib_qp		*roce_v2_gsi;
 };
 
 enum {
@@ -155,7 +156,10 @@ static int is_sqp(struct mlx4_ib_dev *dev, struct mlx4_ib_qp *qp)
 			}
 		}
 	}
-	return proxy_sqp;
+	if (proxy_sqp)
+		return 1;
+
+	return !!(qp->flags & MLX4_IB_ROCE_V2_GSI_QP);
 }
 
 /* used for INIT/CLOSE port logic */
@@ -1102,9 +1106,9 @@ static u32 get_sqp_num(struct mlx4_ib_dev *dev, struct ib_qp_init_attr *attr)
 		return dev->dev->caps.qp1_proxy[attr->port_num - 1];
 }
 
-struct ib_qp *mlx4_ib_create_qp(struct ib_pd *pd,
-				struct ib_qp_init_attr *init_attr,
-				struct ib_udata *udata)
+static struct ib_qp *_mlx4_ib_create_qp(struct ib_pd *pd,
+					struct ib_qp_init_attr *init_attr,
+					struct ib_udata *udata)
 {
 	struct mlx4_ib_qp *qp = NULL;
 	int err;
@@ -1123,6 +1127,7 @@ struct ib_qp *mlx4_ib_create_qp(struct ib_pd *pd,
 					MLX4_IB_SRIOV_TUNNEL_QP |
 					MLX4_IB_SRIOV_SQP |
 					MLX4_IB_QP_NETIF |
+					MLX4_IB_QP_CREATE_ROCE_V2_GSI |
 					MLX4_IB_QP_CREATE_USE_GFP_NOIO))
 		return ERR_PTR(-EINVAL);
 
@@ -1131,15 +1136,21 @@ struct ib_qp *mlx4_ib_create_qp(struct ib_pd *pd,
 			return ERR_PTR(-EINVAL);
 	}
 
-	if (init_attr->create_flags &&
-	    ((udata && init_attr->create_flags & ~(sup_u_create_flags)) ||
-	     ((init_attr->create_flags & ~(MLX4_IB_SRIOV_SQP |
-					   MLX4_IB_QP_CREATE_USE_GFP_NOIO |
-					   MLX4_IB_QP_BLOCK_MULTICAST_LOOPBACK)) &&
-	      init_attr->qp_type != IB_QPT_UD) ||
-	     ((init_attr->create_flags & MLX4_IB_SRIOV_SQP) &&
-	      init_attr->qp_type > IB_QPT_GSI)))
-		return ERR_PTR(-EINVAL);
+	if (init_attr->create_flags) {
+		if (udata && init_attr->create_flags & ~(sup_u_create_flags))
+			return ERR_PTR(-EINVAL);
+
+		if ((init_attr->create_flags & ~(MLX4_IB_SRIOV_SQP |
+						 MLX4_IB_QP_CREATE_USE_GFP_NOIO |
+						 MLX4_IB_QP_CREATE_ROCE_V2_GSI  |
+						 MLX4_IB_QP_BLOCK_MULTICAST_LOOPBACK) &&
+		     init_attr->qp_type != IB_QPT_UD) ||
+		    (init_attr->create_flags & MLX4_IB_SRIOV_SQP &&
+		     init_attr->qp_type > IB_QPT_GSI) ||
+		    (init_attr->create_flags & MLX4_IB_QP_CREATE_ROCE_V2_GSI &&
+		     init_attr->qp_type != IB_QPT_GSI))
+			return ERR_PTR(-EINVAL);
+	}
 
 	switch (init_attr->qp_type) {
 	case IB_QPT_XRC_TGT:
@@ -1176,19 +1187,29 @@ struct ib_qp *mlx4_ib_create_qp(struct ib_pd *pd,
 	case IB_QPT_SMI:
 	case IB_QPT_GSI:
 	{
+		int sqpn;
+
 		/* Userspace is not allowed to create special QPs: */
 		if (udata)
 			return ERR_PTR(-EINVAL);
+		if (init_attr->create_flags & MLX4_IB_QP_CREATE_ROCE_V2_GSI) {
+			int res = mlx4_qp_reserve_range(to_mdev(pd->device)->dev, 1, 1, &sqpn, 0);
+
+			if (res)
+				return ERR_PTR(res);
+		} else {
+			sqpn = get_sqp_num(to_mdev(pd->device), init_attr);
+		}
 
 		err = create_qp_common(to_mdev(pd->device), pd, init_attr, udata,
-				       get_sqp_num(to_mdev(pd->device), init_attr),
+				       sqpn,
 				       &qp, gfp);
 		if (err)
 			return ERR_PTR(err);
 
 		qp->port	= init_attr->port_num;
-		qp->ibqp.qp_num = init_attr->qp_type == IB_QPT_SMI ? 0 : 1;
-
+		qp->ibqp.qp_num = init_attr->qp_type == IB_QPT_SMI ? 0 :
+			init_attr->create_flags & MLX4_IB_QP_CREATE_ROCE_V2_GSI ? sqpn : 1;
 		break;
 	}
 	default:
@@ -1199,7 +1220,41 @@ struct ib_qp *mlx4_ib_create_qp(struct ib_pd *pd,
 	return &qp->ibqp;
 }
 
-int mlx4_ib_destroy_qp(struct ib_qp *qp)
+struct ib_qp *mlx4_ib_create_qp(struct ib_pd *pd,
+				struct ib_qp_init_attr *init_attr,
+				struct ib_udata *udata) {
+	struct ib_device *device = pd ? pd->device : init_attr->xrcd->device;
+	struct ib_qp *ibqp;
+	struct mlx4_ib_dev *dev = to_mdev(device);
+
+	ibqp = _mlx4_ib_create_qp(pd, init_attr, udata);
+
+	if (!IS_ERR(ibqp) &&
+	    (init_attr->qp_type == IB_QPT_GSI) &&
+	    !(init_attr->create_flags & MLX4_IB_QP_CREATE_ROCE_V2_GSI)) {
+		struct mlx4_ib_sqp *sqp = to_msqp((to_mqp(ibqp)));
+		int is_eth = rdma_cap_eth_ah(&dev->ib_dev, init_attr->port_num);
+
+		if (is_eth &&
+		    dev->dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_ROCE_V1_V2) {
+			init_attr->create_flags |= MLX4_IB_QP_CREATE_ROCE_V2_GSI;
+			sqp->roce_v2_gsi = ib_create_qp(pd, init_attr);
+
+			if (IS_ERR(sqp->roce_v2_gsi)) {
+				pr_err("Failed to create GSI QP for RoCEv2 (%ld)\n", PTR_ERR(sqp->roce_v2_gsi));
+				sqp->roce_v2_gsi = NULL;
+			} else {
+				sqp = to_msqp(to_mqp(sqp->roce_v2_gsi));
+				sqp->qp.flags |= MLX4_IB_ROCE_V2_GSI_QP;
+			}
+
+			init_attr->create_flags &= ~MLX4_IB_QP_CREATE_ROCE_V2_GSI;
+		}
+	}
+	return ibqp;
+}
+
+static int _mlx4_ib_destroy_qp(struct ib_qp *qp)
 {
 	struct mlx4_ib_dev *dev = to_mdev(qp->device);
 	struct mlx4_ib_qp *mqp = to_mqp(qp);
@@ -1228,6 +1283,20 @@ int mlx4_ib_destroy_qp(struct ib_qp *qp)
 	return 0;
 }
 
+int mlx4_ib_destroy_qp(struct ib_qp *qp)
+{
+	struct mlx4_ib_qp *mqp = to_mqp(qp);
+
+	if (mqp->mlx4_ib_qp_type == MLX4_IB_QPT_GSI) {
+		struct mlx4_ib_sqp *sqp = to_msqp(mqp);
+
+		if (sqp->roce_v2_gsi)
+			ib_destroy_qp(sqp->roce_v2_gsi);
+	}
+
+	return _mlx4_ib_destroy_qp(qp);
+}
+
 static int to_mlx4_st(struct mlx4_ib_dev *dev, enum mlx4_ib_qp_type type)
 {
 	switch (type) {
@@ -1654,6 +1723,14 @@ static int __mlx4_ib_modify_qp(struct ib_qp *ibqp,
 			mlx4_ib_steer_qp_reg(dev, qp, 1);
 			steer_qp = 1;
 		}
+
+		if (ibqp->qp_type == IB_QPT_GSI) {
+			enum ib_gid_type gid_type = qp->flags & MLX4_IB_ROCE_V2_GSI_QP ?
+				IB_GID_TYPE_ROCE_UDP_ENCAP : IB_GID_TYPE_ROCE;
+			u8 qpc_roce_mode = gid_type_to_qpc(gid_type);
+
+			context->rlkey_roce_mode |= (qpc_roce_mode << 6);
+		}
 	}
 
 	if (attr_mask & IB_QP_PKEY_INDEX) {
@@ -2056,8 +2133,8 @@ out:
 	return err;
 }
 
-int mlx4_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
-		      int attr_mask, struct ib_udata *udata)
+static int _mlx4_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
+			      int attr_mask, struct ib_udata *udata)
 {
 	struct mlx4_ib_dev *dev = to_mdev(ibqp->device);
 	struct mlx4_ib_qp *qp = to_mqp(ibqp);
@@ -2160,6 +2237,27 @@ out:
 	return err;
 }
 
+int mlx4_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
+		      int attr_mask, struct ib_udata *udata)
+{
+	struct mlx4_ib_qp *mqp = to_mqp(ibqp);
+	int ret;
+
+	ret = _mlx4_ib_modify_qp(ibqp, attr, attr_mask, udata);
+
+	if (mqp->mlx4_ib_qp_type == MLX4_IB_QPT_GSI) {
+		struct mlx4_ib_sqp *sqp = to_msqp(mqp);
+		int err = 0;
+
+		if (sqp->roce_v2_gsi)
+			err = ib_modify_qp(sqp->roce_v2_gsi, attr, attr_mask);
+		if (err)
+			pr_err("Failed to modify GSI QP for RoCEv2 (%d)\n",
+			       err);
+	}
+	return ret;
+}
+
 static int vf_get_qp0_qkey(struct mlx4_dev *dev, int qpn, u32 *qkey)
 {
 	int i;
@@ -2804,6 +2902,29 @@ int mlx4_ib_post_send(struct ib_qp *ibqp, struct ib_send_wr *wr,
 	int i;
 	struct mlx4_ib_dev *mdev = to_mdev(ibqp->device);
 
+	if (qp->mlx4_ib_qp_type == MLX4_IB_QPT_GSI) {
+		struct mlx4_ib_sqp *sqp = to_msqp(qp);
+
+		if (sqp->roce_v2_gsi) {
+			struct mlx4_ib_ah *ah = to_mah(ud_wr(wr)->ah);
+			struct ib_gid_attr gid_attr;
+			union ib_gid gid;
+
+			if (!ib_get_cached_gid(ibqp->device,
+					       be32_to_cpu(ah->av.ib.port_pd) >> 24,
+					       ah->av.ib.gid_index, &gid,
+					       &gid_attr)) {
+				if (gid_attr.ndev)
+					dev_put(gid_attr.ndev);
+				qp = (gid_attr.gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP) ?
+					to_mqp(sqp->roce_v2_gsi) : qp;
+			} else {
+				pr_err("Failed to get gid at index %d. RoCEv2 will not work properly\n",
+				       ah->av.ib.gid_index);
+			}
+		}
+	}
+
 	spin_lock_irqsave(&qp->sq.lock, flags);
 	if (mdev->dev->persist->state & MLX4_DEVICE_STATE_INTERNAL_ERROR) {
 		err = -EIO;
-- 
cgit v1.2.3


From 4ed088e6c2aa2b7b0479791c7251696b541bfa39 Mon Sep 17 00:00:00 2001
From: Matan Barak <matanb@mellanox.com>
Date: Thu, 14 Jan 2016 17:50:43 +0200
Subject: IB/mlx4: Advertise RoCE v2 support

Advertise RoCE v2 support in port_immutable attributes according to
the hardware's capabilities. This enables the verbs stack to use
RoCE v2 mode.

Signed-off-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
---
 drivers/infiniband/hw/mlx4/main.c | 12 +++++++++---
 1 file changed, 9 insertions(+), 3 deletions(-)

(limited to 'drivers/infiniband/hw/mlx4')

diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index 8bc0d608a9ca..1c7ab6cabbb8 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -2185,6 +2185,7 @@ static int mlx4_port_immutable(struct ib_device *ibdev, u8 port_num,
 			       struct ib_port_immutable *immutable)
 {
 	struct ib_port_attr attr;
+	struct mlx4_ib_dev *mdev = to_mdev(ibdev);
 	int err;
 
 	err = mlx4_ib_query_port(ibdev, port_num, &attr);
@@ -2194,10 +2195,15 @@ static int mlx4_port_immutable(struct ib_device *ibdev, u8 port_num,
 	immutable->pkey_tbl_len = attr.pkey_tbl_len;
 	immutable->gid_tbl_len = attr.gid_tbl_len;
 
-	if (mlx4_ib_port_link_layer(ibdev, port_num) == IB_LINK_LAYER_INFINIBAND)
+	if (mlx4_ib_port_link_layer(ibdev, port_num) == IB_LINK_LAYER_INFINIBAND) {
 		immutable->core_cap_flags = RDMA_CORE_PORT_IBA_IB;
-	else
-		immutable->core_cap_flags = RDMA_CORE_PORT_IBA_ROCE;
+	} else {
+		if (mdev->dev->caps.flags & MLX4_DEV_CAP_FLAG_IBOE)
+			immutable->core_cap_flags = RDMA_CORE_PORT_IBA_ROCE;
+		if (mdev->dev->caps.flags2 & MLX4_DEV_CAP_FLAG2_ROCE_V1_V2)
+			immutable->core_cap_flags = RDMA_CORE_PORT_IBA_ROCE |
+				RDMA_CORE_PORT_IBA_ROCE_UDP_ENCAP;
+	}
 
 	immutable->max_mad_size = IB_MGMT_MAD_SIZE;
 
-- 
cgit v1.2.3