ivampiresp/Spdk
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

nvmf/vfio-user: don't check controller ready state in group poll context

Browse Source 
The group poll context is for queue pair state, so we don't need to
check controller state here, and for the disconnect case below, the ADMIN
queue pair will be removed from group poll.

Also add spdk_unlikely in the poll context.

Change-Id: I5ef32ef3cf41ad757a7cb167e1e1fa32c52a84d6
Signed-off-by: Changpeng Liu <changpeng.liu@intel.com>
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/6227
Community-CI: Broadcom CI
Community-CI: Mellanox Build Bot
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This commit is contained in:
Changpeng Liu 2021-02-02 22:40:19 +08:00 committed by Tomasz Zawadzki
parent b05b3d0a12
commit 0f30fea99b
1 changed files with 4 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
lib/nvmf/vfio_user.c
View File

@ -1164,7 +1164,6 @@ nvmf_vfio_user_prop_req_rsp(struct nvmf_vfio_user_req *req, void *cb_arg)
}
}
qpair->ctrlr->ready = true;
return 0;
}
@ -1277,9 +1276,6 @@ access_bar0_fn(vfu_ctx_t *vfu_ctx, char *buf, size_t count, loff_t pos,
req->req.length = count;
req->req.data = buf;
/* Mark the controller as busy to limit the queue depth for fabric get/set to 1 */
ctrlr->ready = false;
spdk_nvmf_request_exec_fabrics(&req->req);
return count;
@ -2266,15 +2262,14 @@ nvmf_vfio_user_poll_group_poll(struct spdk_nvmf_transport_poll_group *group)
TAILQ_FOREACH_SAFE(vu_qpair, &vu_group->qps, link, tmp) {
ctrlr = vu_qpair->ctrlr;
if (!ctrlr->ready) {
continue;
}
assert(ctrlr != NULL);
if (nvmf_qpair_is_admin_queue(&vu_qpair->qpair)) {
if (spdk_unlikely(nvmf_qpair_is_admin_queue(&vu_qpair->qpair))) {
int err;
err = nvmf_vfio_user_ctrlr_poll(ctrlr);
if (spdk_unlikely(err) != 0) {
/* initiator shutdown or reset, waiting for another re-connect */
if (err == -ENOTCONN) {
TAILQ_REMOVE(&vu_group->qps, vu_qpair, link);
ctrlr->ready = false;
@ -2286,7 +2281,7 @@ nvmf_vfio_user_poll_group_poll(struct spdk_nvmf_transport_poll_group *group)
}
}
if (vu_qpair->state != VFIO_USER_QPAIR_ACTIVE || !vu_qpair->sq.size) {
if (spdk_unlikely(vu_qpair->state != VFIO_USER_QPAIR_ACTIVE || !vu_qpair->sq.size)) {
continue;
}