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>
diff --git a/drivers/infiniband/hw/cxgb4/qp.c b/drivers/infiniband/hw/cxgb4/qp.c
index 6517e12..b60498f 100644
--- a/drivers/infiniband/hw/cxgb4/qp.c
+++ b/drivers/infiniband/hw/cxgb4/qp.c
@@ -528,8 +528,8 @@
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 @@
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_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 @@
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;