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

nvmf/rdma: remove lock on few transport ops

Browse Source 
it simplifies the code and improves readability
sync is done on generic layer

Signed-off-by: Jacek Kalwas <jacek.kalwas@intel.com>
Change-Id: If324039ef2b26fa8ba026b80ec49788a7b2dcaa3
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/13667
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Community-CI: Broadcom CI <spdk-ci.pdl@broadcom.com>
Community-CI: Mellanox Build Bot
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
Reviewed-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
This commit is contained in:
Jacek Kalwas 2022-07-13 09:41:38 -04:00 committed by Ben Walker
parent c7ac84d1f2
commit 8c35e1bd79
2 changed files with 0 additions and 54 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

51
lib/nvmf/rdma.c
View File

@ -469,7 +469,6 @@ struct spdk_nvmf_rdma_transport {
struct spdk_mempool *data_wr_pool;
struct spdk_poller *accept_poller;
pthread_mutex_t lock;
/* fields used to poll RDMA/IB events */
nfds_t npoll_fds;
@ -2398,35 +2397,12 @@ nvmf_rdma_create(struct spdk_nvmf_transport_opts *opts)
uint32_t min_shared_buffers;
uint32_t min_in_capsule_data_size;
int max_device_sge = SPDK_NVMF_MAX_SGL_ENTRIES;
pthread_mutexattr_t attr;
rtransport = calloc(1, sizeof(*rtransport));
if (!rtransport) {
return NULL;
}
if (pthread_mutexattr_init(&attr)) {
SPDK_ERRLOG("pthread_mutexattr_init() failed\n");
free(rtransport);
return NULL;
}
if (pthread_mutexattr_settype(&attr, PTHREAD_MUTEX_RECURSIVE)) {
SPDK_ERRLOG("pthread_mutexattr_settype() failed\n");
pthread_mutexattr_destroy(&attr);
free(rtransport);
return NULL;
}
if (pthread_mutex_init(&rtransport->lock, &attr)) {
SPDK_ERRLOG("pthread_mutex_init() failed\n");
pthread_mutexattr_destroy(&attr);
free(rtransport);
return NULL;
}
pthread_mutexattr_destroy(&attr);
TAILQ_INIT(&rtransport->devices);
TAILQ_INIT(&rtransport->ports);
TAILQ_INIT(&rtransport->poll_groups);
@ -2730,7 +2706,6 @@ nvmf_rdma_destroy(struct spdk_nvmf_transport *transport,
spdk_mempool_free(rtransport->data_wr_pool);
spdk_poller_unregister(&rtransport->accept_poller);
pthread_mutex_destroy(&rtransport->lock);
free(rtransport);
if (cb_fn) {
@ -2763,11 +2738,9 @@ nvmf_rdma_listen(struct spdk_nvmf_transport *transport, const struct spdk_nvme_t
rtransport = SPDK_CONTAINEROF(transport, struct spdk_nvmf_rdma_transport, transport);
assert(rtransport->event_channel != NULL);
pthread_mutex_lock(&rtransport->lock);
port = calloc(1, sizeof(*port));
if (!port) {
SPDK_ERRLOG("Port allocation failed\n");
pthread_mutex_unlock(&rtransport->lock);
return -ENOMEM;
}
@ -2783,7 +2756,6 @@ nvmf_rdma_listen(struct spdk_nvmf_transport *transport, const struct spdk_nvme_t
default:
SPDK_ERRLOG("Unhandled ADRFAM %d\n", trid->adrfam);
free(port);
pthread_mutex_unlock(&rtransport->lock);
return -EINVAL;
}
@ -2797,7 +2769,6 @@ nvmf_rdma_listen(struct spdk_nvmf_transport *transport, const struct spdk_nvme_t
if (rc) {
SPDK_ERRLOG("getaddrinfo failed: %s (%d)\n", gai_strerror(rc), rc);
free(port);
pthread_mutex_unlock(&rtransport->lock);
return -EINVAL;
}
@ -2806,7 +2777,6 @@ nvmf_rdma_listen(struct spdk_nvmf_transport *transport, const struct spdk_nvme_t
SPDK_ERRLOG("rdma_create_id() failed\n");
freeaddrinfo(res);
free(port);
pthread_mutex_unlock(&rtransport->lock);
return rc;
}
@ -2817,7 +2787,6 @@ nvmf_rdma_listen(struct spdk_nvmf_transport *transport, const struct spdk_nvme_t
SPDK_ERRLOG("rdma_bind_addr() failed\n");
rdma_destroy_id(port->id);
free(port);
pthread_mutex_unlock(&rtransport->lock);
return rc;
}
@ -2825,7 +2794,6 @@ nvmf_rdma_listen(struct spdk_nvmf_transport *transport, const struct spdk_nvme_t
SPDK_ERRLOG("ibv_context is null\n");
rdma_destroy_id(port->id);
free(port);
pthread_mutex_unlock(&rtransport->lock);
return -1;
}
@ -2834,7 +2802,6 @@ nvmf_rdma_listen(struct spdk_nvmf_transport *transport, const struct spdk_nvme_t
SPDK_ERRLOG("rdma_listen() failed\n");
rdma_destroy_id(port->id);
free(port);
pthread_mutex_unlock(&rtransport->lock);
return rc;
}
@ -2849,7 +2816,6 @@ nvmf_rdma_listen(struct spdk_nvmf_transport *transport, const struct spdk_nvme_t
port->id->verbs);
rdma_destroy_id(port->id);
free(port);
pthread_mutex_unlock(&rtransport->lock);
return -EINVAL;
}
@ -2857,7 +2823,6 @@ nvmf_rdma_listen(struct spdk_nvmf_transport *transport, const struct spdk_nvme_t
trid->traddr, trid->trsvcid);
TAILQ_INSERT_TAIL(&rtransport->ports, port, link);
pthread_mutex_unlock(&rtransport->lock);
return 0;
}
@ -2870,7 +2835,6 @@ nvmf_rdma_stop_listen(struct spdk_nvmf_transport *transport,
rtransport = SPDK_CONTAINEROF(transport, struct spdk_nvmf_rdma_transport, transport);
pthread_mutex_lock(&rtransport->lock);
TAILQ_FOREACH_SAFE(port, &rtransport->ports, link, tmp) {
if (spdk_nvme_transport_id_compare(port->trid, trid) == 0) {
TAILQ_REMOVE(&rtransport->ports, port, link);
@ -2879,8 +2843,6 @@ nvmf_rdma_stop_listen(struct spdk_nvmf_transport *transport,
break;
}
}
pthread_mutex_unlock(&rtransport->lock);
}
static void
@ -3449,13 +3411,11 @@ nvmf_rdma_poll_group_create(struct spdk_nvmf_transport *transport,
TAILQ_INIT(&rgroup->pollers);
pthread_mutex_lock(&rtransport->lock);
TAILQ_FOREACH(device, &rtransport->devices, link) {
poller = calloc(1, sizeof(*poller));
if (!poller) {
SPDK_ERRLOG("Unable to allocate memory for new RDMA poller\n");
nvmf_rdma_poll_group_destroy(&rgroup->group);
pthread_mutex_unlock(&rtransport->lock);
return NULL;
}
@ -3484,7 +3444,6 @@ nvmf_rdma_poll_group_create(struct spdk_nvmf_transport *transport,
if (!poller->srq) {
SPDK_ERRLOG("Unable to create shared receive queue, errno %d\n", errno);
nvmf_rdma_poll_group_destroy(&rgroup->group);
pthread_mutex_unlock(&rtransport->lock);
return NULL;
}
@ -3499,7 +3458,6 @@ nvmf_rdma_poll_group_create(struct spdk_nvmf_transport *transport,
if (!poller->resources) {
SPDK_ERRLOG("Unable to allocate resources for shared receive queue.\n");
nvmf_rdma_poll_group_destroy(&rgroup->group);
pthread_mutex_unlock(&rtransport->lock);
return NULL;
}
}
@ -3520,7 +3478,6 @@ nvmf_rdma_poll_group_create(struct spdk_nvmf_transport *transport,
if (!poller->cq) {
SPDK_ERRLOG("Unable to create completion queue\n");
nvmf_rdma_poll_group_destroy(&rgroup->group);
pthread_mutex_unlock(&rtransport->lock);
return NULL;
}
poller->num_cqe = num_cqe;
@ -3532,7 +3489,6 @@ nvmf_rdma_poll_group_create(struct spdk_nvmf_transport *transport,
rtransport->conn_sched.next_io_pg = rgroup;
}
pthread_mutex_unlock(&rtransport->lock);
return &rgroup->group;
}
@ -3545,10 +3501,7 @@ nvmf_rdma_get_optimal_poll_group(struct spdk_nvmf_qpair *qpair)
rtransport = SPDK_CONTAINEROF(qpair->transport, struct spdk_nvmf_rdma_transport, transport);
pthread_mutex_lock(&rtransport->lock);
if (TAILQ_EMPTY(&rtransport->poll_groups)) {
pthread_mutex_unlock(&rtransport->lock);
return NULL;
}
@ -3567,8 +3520,6 @@ nvmf_rdma_get_optimal_poll_group(struct spdk_nvmf_qpair *qpair)
*pg = TAILQ_FIRST(&rtransport->poll_groups);
}
pthread_mutex_unlock(&rtransport->lock);
return result;
}
@ -3616,7 +3567,6 @@ nvmf_rdma_poll_group_destroy(struct spdk_nvmf_transport_poll_group *group)
rtransport = SPDK_CONTAINEROF(rgroup->group.transport, struct spdk_nvmf_rdma_transport, transport);
pthread_mutex_lock(&rtransport->lock);
next_rgroup = TAILQ_NEXT(rgroup, link);
TAILQ_REMOVE(&rtransport->poll_groups, rgroup, link);
if (next_rgroup == NULL) {
@ -3628,7 +3578,6 @@ nvmf_rdma_poll_group_destroy(struct spdk_nvmf_transport_poll_group *group)
if (rtransport->conn_sched.next_io_pg == rgroup) {
rtransport->conn_sched.next_io_pg = next_rgroup;
}
pthread_mutex_unlock(&rtransport->lock);
free(rgroup);
}

3
test/unit/lib/nvmf/rdma.c/rdma_ut.c
View File

@ -792,7 +792,6 @@ test_nvmf_rdma_get_optimal_poll_group(void)
uint32_t i;
rqpair.qpair.transport = transport;
pthread_mutex_init(&rtransport.lock, NULL);
TAILQ_INIT(&rtransport.poll_groups);
for (i = 0; i < TEST_GROUPS_COUNT; i++) {
@ -855,8 +854,6 @@ test_nvmf_rdma_get_optimal_poll_group(void)
rqpair.qpair.qid = 1;
result = nvmf_rdma_get_optimal_poll_group(&rqpair.qpair);
CU_ASSERT(result == NULL);
pthread_mutex_destroy(&rtransport.lock);
}
#undef TEST_GROUPS_COUNT