From 20cf4e026730104892fa1268de0371a631cee294 Mon Sep 17 00:00:00 2001 From: Chuck Lever Date: Mon, 29 Jul 2019 13:22:09 -0400 Subject: rdma: Enable ib_alloc_cq to spread work over a device's comp_vectors MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Send and Receive completion is handled on a single CPU selected at the time each Completion Queue is allocated. Typically this is when an initiator instantiates an RDMA transport, or when a target accepts an RDMA connection. Some ULPs cannot open a connection per CPU to spread completion workload across available CPUs and MSI vectors. For such ULPs, provide an API that allows the RDMA core to select a completion vector based on the device's complement of available comp_vecs. ULPs that invoke ib_alloc_cq() with only comp_vector 0 are converted to use the new API so that their completion workloads interfere less with each other. Suggested-by: Håkon Bugge Signed-off-by: Chuck Lever Reviewed-by: Leon Romanovsky Cc: Cc: Link: https://lore.kernel.org/r/20190729171923.13428.52555.stgit@manet.1015granger.net Signed-off-by: Doug Ledford --- net/9p/trans_rdma.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) (limited to 'net/9p/trans_rdma.c') diff --git a/net/9p/trans_rdma.c b/net/9p/trans_rdma.c index bac8dad5dd69..b21c3c209815 100644 --- a/net/9p/trans_rdma.c +++ b/net/9p/trans_rdma.c @@ -685,9 +685,9 @@ rdma_create_trans(struct p9_client *client, const char *addr, char *args) goto error; /* Create the Completion Queue */ - rdma->cq = ib_alloc_cq(rdma->cm_id->device, client, - opts.sq_depth + opts.rq_depth + 1, - 0, IB_POLL_SOFTIRQ); + rdma->cq = ib_alloc_cq_any(rdma->cm_id->device, client, + opts.sq_depth + opts.rq_depth + 1, + IB_POLL_SOFTIRQ); if (IS_ERR(rdma->cq)) goto error; -- cgit v1.2.3-59-g8ed1b