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 <hch@lst.de>
Reviewed-by: Bart Van Assche <bart.vanassche@sandisk.com> [srp, srpt]
Reviewed-by: Chuck Lever <chuck.lever@oracle.com> [sunrpc]
Tested-by: Haggai Eran <haggaie@mellanox.com>
Tested-by: Sagi Grimberg <sagig@mellanox.com>
Tested-by: Steve Wise <swise@opengridcomputing.com>

6 files changed, 61 insertions, 38 deletions

diff --git a/drivers/staging/rdma/ipath/ipath_rc.c b/drivers/staging/rdma/ipath/ipath_rc.c
index 79b3dbc..d4aa535 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 1f95bba..46af8b0 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 22e6099..0246b30 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 e8a2a91..3ffc156 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 ed2bbc2..15633ec 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 ec167e5..ed102a2 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 */