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

nvmf/tcp: 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: I75753511842dff237bb27561e406c43ea68269fe
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/13666
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 09:38:19 -04:00 committed by Ben Walker
parent b17919d8bc
commit c7ac84d1f2
1 changed files with 0 additions and 25 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

25
lib/nvmf/tcp.c
View File

@ -331,7 +331,6 @@ struct spdk_nvmf_tcp_transport {
struct spdk_nvmf_tcp_poll_group *next_pg; struct spdk_nvmf_tcp_poll_group *next_pg;
struct spdk_poller *accept_poller; struct spdk_poller *accept_poller;
pthread_mutex_t lock;
TAILQ_HEAD(, spdk_nvmf_tcp_port) ports; TAILQ_HEAD(, spdk_nvmf_tcp_port) ports;
TAILQ_HEAD(, spdk_nvmf_tcp_poll_group) poll_groups; TAILQ_HEAD(, spdk_nvmf_tcp_poll_group) poll_groups;
@ -572,7 +571,6 @@ nvmf_tcp_destroy(struct spdk_nvmf_transport *transport,
ttransport = SPDK_CONTAINEROF(transport, struct spdk_nvmf_tcp_transport, transport); ttransport = SPDK_CONTAINEROF(transport, struct spdk_nvmf_tcp_transport, transport);
spdk_poller_unregister(&ttransport->accept_poller); spdk_poller_unregister(&ttransport->accept_poller);
pthread_mutex_destroy(&ttransport->lock);
free(ttransport); free(ttransport);
if (cb_fn) { if (cb_fn) {
@ -671,12 +669,9 @@ nvmf_tcp_create(struct spdk_nvmf_transport_opts *opts)
return NULL; return NULL;
} }
pthread_mutex_init(&ttransport->lock, NULL);
ttransport->accept_poller = SPDK_POLLER_REGISTER(nvmf_tcp_accept, &ttransport->transport, ttransport->accept_poller = SPDK_POLLER_REGISTER(nvmf_tcp_accept, &ttransport->transport,
opts->acceptor_poll_rate); opts->acceptor_poll_rate);
if (!ttransport->accept_poller) { if (!ttransport->accept_poller) {
pthread_mutex_destroy(&ttransport->lock);
free(ttransport); free(ttransport);
return NULL; return NULL;
} }
@ -728,8 +723,6 @@ nvmf_tcp_canon_listen_trid(struct spdk_nvme_transport_id *canon_trid,
/** /**
* Find an existing listening port. * Find an existing listening port.
*
* Caller must hold ttransport->lock.
*/ */
static struct spdk_nvmf_tcp_port * static struct spdk_nvmf_tcp_port *
nvmf_tcp_find_port(struct spdk_nvmf_tcp_transport *ttransport, nvmf_tcp_find_port(struct spdk_nvmf_tcp_transport *ttransport,
@ -774,11 +767,9 @@ nvmf_tcp_listen(struct spdk_nvmf_transport *transport, const struct spdk_nvme_tr
return -EINVAL; return -EINVAL;
} }
pthread_mutex_lock(&ttransport->lock);
port = calloc(1, sizeof(*port)); port = calloc(1, sizeof(*port));
if (!port) { if (!port) {
SPDK_ERRLOG("Port allocation failed\n"); SPDK_ERRLOG("Port allocation failed\n");
pthread_mutex_unlock(&ttransport->lock);
return -ENOMEM; return -ENOMEM;
} }
@ -793,7 +784,6 @@ nvmf_tcp_listen(struct spdk_nvmf_transport *transport, const struct spdk_nvme_tr
trid->traddr, trsvcid_int, trid->traddr, trsvcid_int,
spdk_strerror(errno), errno); spdk_strerror(errno), errno);
free(port); free(port);
pthread_mutex_unlock(&ttransport->lock);
return -errno; return -errno;
} }
@ -810,7 +800,6 @@ nvmf_tcp_listen(struct spdk_nvmf_transport *transport, const struct spdk_nvme_tr
SPDK_ERRLOG("Socket address family mismatch\n"); SPDK_ERRLOG("Socket address family mismatch\n");
spdk_sock_close(&port->listen_sock); spdk_sock_close(&port->listen_sock);
free(port); free(port);
pthread_mutex_unlock(&ttransport->lock);
return -EINVAL; return -EINVAL;
} }
@ -818,7 +807,6 @@ nvmf_tcp_listen(struct spdk_nvmf_transport *transport, const struct spdk_nvme_tr
trid->traddr, trid->trsvcid); trid->traddr, trid->trsvcid);
TAILQ_INSERT_TAIL(&ttransport->ports, port, link); TAILQ_INSERT_TAIL(&ttransport->ports, port, link);
pthread_mutex_unlock(&ttransport->lock);
return 0; return 0;
} }
@ -834,15 +822,12 @@ nvmf_tcp_stop_listen(struct spdk_nvmf_transport *transport,
SPDK_DEBUGLOG(nvmf_tcp, "Removing listen address %s port %s\n", SPDK_DEBUGLOG(nvmf_tcp, "Removing listen address %s port %s\n",
trid->traddr, trid->trsvcid); trid->traddr, trid->trsvcid);
pthread_mutex_lock(&ttransport->lock);
port = nvmf_tcp_find_port(ttransport, trid); port = nvmf_tcp_find_port(ttransport, trid);
if (port) { if (port) {
TAILQ_REMOVE(&ttransport->ports, port, link); TAILQ_REMOVE(&ttransport->ports, port, link);
spdk_sock_close(&port->listen_sock); spdk_sock_close(&port->listen_sock);
free(port); free(port);
} }
pthread_mutex_unlock(&ttransport->lock);
} }
static void nvmf_tcp_qpair_set_recv_state(struct spdk_nvmf_tcp_qpair *tqpair, static void nvmf_tcp_qpair_set_recv_state(struct spdk_nvmf_tcp_qpair *tqpair,
@ -1334,12 +1319,10 @@ nvmf_tcp_poll_group_create(struct spdk_nvmf_transport *transport,
goto cleanup; goto cleanup;
} }
pthread_mutex_lock(&ttransport->lock);
TAILQ_INSERT_TAIL(&ttransport->poll_groups, tgroup, link); TAILQ_INSERT_TAIL(&ttransport->poll_groups, tgroup, link);
if (ttransport->next_pg == NULL) { if (ttransport->next_pg == NULL) {
ttransport->next_pg = tgroup; ttransport->next_pg = tgroup;
} }
pthread_mutex_unlock(&ttransport->lock);
return &tgroup->group; return &tgroup->group;
@ -1359,10 +1342,7 @@ nvmf_tcp_get_optimal_poll_group(struct spdk_nvmf_qpair *qpair)
ttransport = SPDK_CONTAINEROF(qpair->transport, struct spdk_nvmf_tcp_transport, transport); ttransport = SPDK_CONTAINEROF(qpair->transport, struct spdk_nvmf_tcp_transport, transport);
pthread_mutex_lock(&ttransport->lock);
if (TAILQ_EMPTY(&ttransport->poll_groups)) { if (TAILQ_EMPTY(&ttransport->poll_groups)) {
pthread_mutex_unlock(&ttransport->lock);
return NULL; return NULL;
} }
@ -1373,11 +1353,9 @@ nvmf_tcp_get_optimal_poll_group(struct spdk_nvmf_qpair *qpair)
tqpair = SPDK_CONTAINEROF(qpair, struct spdk_nvmf_tcp_qpair, qpair); tqpair = SPDK_CONTAINEROF(qpair, struct spdk_nvmf_tcp_qpair, qpair);
rc = spdk_sock_get_optimal_sock_group(tqpair->sock, &group, hint); rc = spdk_sock_get_optimal_sock_group(tqpair->sock, &group, hint);
if (rc != 0) { if (rc != 0) {
pthread_mutex_unlock(&ttransport->lock);
return NULL; return NULL;
} else if (group != NULL) { } else if (group != NULL) {
/* Optimal poll group was found */ /* Optimal poll group was found */
pthread_mutex_unlock(&ttransport->lock);
return spdk_sock_group_get_ctx(group); return spdk_sock_group_get_ctx(group);
} }
@ -1387,7 +1365,6 @@ nvmf_tcp_get_optimal_poll_group(struct spdk_nvmf_qpair *qpair)
*pg = TAILQ_FIRST(&ttransport->poll_groups); *pg = TAILQ_FIRST(&ttransport->poll_groups);
} }
pthread_mutex_unlock(&ttransport->lock);
return spdk_sock_group_get_ctx(hint); return spdk_sock_group_get_ctx(hint);
} }
@ -1409,7 +1386,6 @@ nvmf_tcp_poll_group_destroy(struct spdk_nvmf_transport_poll_group *group)
ttransport = SPDK_CONTAINEROF(tgroup->group.transport, struct spdk_nvmf_tcp_transport, transport); ttransport = SPDK_CONTAINEROF(tgroup->group.transport, struct spdk_nvmf_tcp_transport, transport);
pthread_mutex_lock(&ttransport->lock);
next_tgroup = TAILQ_NEXT(tgroup, link); next_tgroup = TAILQ_NEXT(tgroup, link);
TAILQ_REMOVE(&ttransport->poll_groups, tgroup, link); TAILQ_REMOVE(&ttransport->poll_groups, tgroup, link);
if (next_tgroup == NULL) { if (next_tgroup == NULL) {
@ -1418,7 +1394,6 @@ nvmf_tcp_poll_group_destroy(struct spdk_nvmf_transport_poll_group *group)
if (ttransport->next_pg == tgroup) { if (ttransport->next_pg == tgroup) {
ttransport->next_pg = next_tgroup; ttransport->next_pg = next_tgroup;
} }
pthread_mutex_unlock(&ttransport->lock);
free(tgroup); free(tgroup);
} }