aboutsummaryrefslogtreecommitdiffstats
path: root/drivers/nvme/target
diff options
context:
space:
mode:
authorSteve Wise <swise@opengridcomputing.com>2018-06-18 08:05:26 -0700
committerJason Gunthorpe <jgg@mellanox.com>2018-06-18 13:17:28 -0600
commit33023fb85a42b53bf778bc025f9667b582282be4 (patch)
tree66d55701f7cc41636dcfc4353e38b4262791b47f /drivers/nvme/target
parentIB/rxe: avoid unnecessary NULL check (diff)
downloadlinux-dev-33023fb85a42b53bf778bc025f9667b582282be4.tar.xz
linux-dev-33023fb85a42b53bf778bc025f9667b582282be4.zip
IB/core: add max_send_sge and max_recv_sge attributes
This patch replaces the ib_device_attr.max_sge with max_send_sge and max_recv_sge. It allows ulps to take advantage of devices that have very different send and recv sge depths. For example cxgb4 has a max_recv_sge of 4, yet a max_send_sge of 16. Splitting out these attributes allows much more efficient use of the SQ for cxgb4 with ulps that use the RDMA_RW API. Consider a large RDMA WRITE that has 16 scattergather entries. With max_sge of 4, the ulp would send 4 WRITE WRs, but with max_sge of 16, it can be done with 1 WRITE WR. Acked-by: Sagi Grimberg <sagi@grimberg.me> Acked-by: Christoph Hellwig <hch@lst.de> Acked-by: Selvin Xavier <selvin.xavier@broadcom.com> Acked-by: Shiraz Saleem <shiraz.saleem@intel.com> Acked-by: Dennis Dalessandro <dennis.dalessandro@intel.com> Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Diffstat (limited to 'drivers/nvme/target')
-rw-r--r--drivers/nvme/target/rdma.c2
1 files changed, 1 insertions, 1 deletions
diff --git a/drivers/nvme/target/rdma.c b/drivers/nvme/target/rdma.c
index 52e0c5d579a7..0d7f3d603f1d 100644
--- a/drivers/nvme/target/rdma.c
+++ b/drivers/nvme/target/rdma.c
@@ -874,7 +874,7 @@ static int nvmet_rdma_create_queue_ib(struct nvmet_rdma_queue *queue)
qp_attr.cap.max_send_wr = queue->send_queue_size + 1;
qp_attr.cap.max_rdma_ctxs = queue->send_queue_size;
qp_attr.cap.max_send_sge = max(ndev->device->attrs.max_sge_rd,
- ndev->device->attrs.max_sge);
+ ndev->device->attrs.max_send_sge);
if (ndev->srq) {
qp_attr.srq = ndev->srq;