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

nvmf/tcp: Use the common buffer cache for each polling group

Browse Source 
Purpose: To avoid the buffer contention among different
polling groups if there are multiple core configurations
for NVMe-oF tcp transport.

Change-Id: I1c1b0126f3aad28f339ec8bf9355282e08cfa8db
Signed-off-by: Ziye Yang <optimistyzy@gmail.com>
Reviewed-on: https://review.gerrithub.io/c/440444
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Seth Howell <seth.howell5141@gmail.com>
Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
This commit is contained in:
Ziye Yang 2018-12-07 05:48:51 +08:00 committed by Jim Harris
parent 3bb815ae9f
commit 3c88819bc0
2 changed files with 61 additions and 22 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

53
lib/nvmf/tcp.c
View File

@ -297,6 +297,10 @@ struct spdk_nvmf_tcp_transport {
struct spdk_nvmf_tcp_mgmt_channel { struct spdk_nvmf_tcp_mgmt_channel {
/* Requests that are waiting to obtain a data buffer */ /* Requests that are waiting to obtain a data buffer */
TAILQ_HEAD(, nvme_tcp_req) pending_data_buf_queue; TAILQ_HEAD(, nvme_tcp_req) pending_data_buf_queue;
/* Point to the transport polling group */
struct spdk_nvmf_tcp_poll_group *tgroup;
}; };
static void spdk_nvmf_tcp_qpair_process_pending(struct spdk_nvmf_tcp_transport *ttransport, static void spdk_nvmf_tcp_qpair_process_pending(struct spdk_nvmf_tcp_transport *ttransport,
@ -1301,6 +1305,7 @@ spdk_nvmf_tcp_poll_group_destroy(struct spdk_nvmf_transport_poll_group *group)
tgroup = SPDK_CONTAINEROF(group, struct spdk_nvmf_tcp_poll_group, group); tgroup = SPDK_CONTAINEROF(group, struct spdk_nvmf_tcp_poll_group, group);
spdk_sock_group_close(&tgroup->sock_group); spdk_sock_group_close(&tgroup->sock_group);
spdk_poller_unregister(&tgroup->timeout_poller); spdk_poller_unregister(&tgroup->timeout_poller);
free(tgroup); free(tgroup);
} }
@ -2150,6 +2155,26 @@ spdk_nvmf_tcp_req_get_xfer(struct nvme_tcp_req *tcp_req) {
return xfer; return xfer;
} }
static void
spdk_nvmf_tcp_request_free_buffers(struct nvme_tcp_req *tcp_req,
struct spdk_nvmf_transport_poll_group *group, struct spdk_nvmf_transport *transport)
{
for (uint32_t i = 0; i < tcp_req->req.iovcnt; i++) {
assert(tcp_req->buffers[i] != NULL);
if (group->buf_cache_count < group->buf_cache_size) {
STAILQ_INSERT_HEAD(&group->buf_cache,
(struct spdk_nvmf_transport_pg_cache_buf *)tcp_req->buffers[i], link);
group->buf_cache_count++;
} else {
spdk_mempool_put(transport->data_buf_pool, tcp_req->buffers[i]);
}
tcp_req->req.iov[i].iov_base = NULL;
tcp_req->buffers[i] = NULL;
tcp_req->req.iov[i].iov_len = 0;
}
tcp_req->data_from_pool = false;
}
static int static int
spdk_nvmf_tcp_req_fill_iovs(struct spdk_nvmf_tcp_transport *ttransport, spdk_nvmf_tcp_req_fill_iovs(struct spdk_nvmf_tcp_transport *ttransport,
struct nvme_tcp_req *tcp_req) struct nvme_tcp_req *tcp_req)
@ -2157,13 +2182,21 @@ spdk_nvmf_tcp_req_fill_iovs(struct spdk_nvmf_tcp_transport *ttransport,
void *buf = NULL; void *buf = NULL;
uint32_t length = tcp_req->req.length; uint32_t length = tcp_req->req.length;
uint32_t i = 0; uint32_t i = 0;
struct nvme_tcp_qpair *tqpair = SPDK_CONTAINEROF(tcp_req->req.qpair, struct nvme_tcp_qpair, qpair);
struct spdk_nvmf_transport_poll_group *group = &tqpair->ch->tgroup->group;
tcp_req->req.iovcnt = 0; tcp_req->req.iovcnt = 0;
while (length) { while (length) {
if (!(STAILQ_EMPTY(&group->buf_cache))) {
group->buf_cache_count--;
buf = STAILQ_FIRST(&group->buf_cache);
STAILQ_REMOVE_HEAD(&group->buf_cache, link);
} else {
buf = spdk_mempool_get(ttransport->transport.data_buf_pool); buf = spdk_mempool_get(ttransport->transport.data_buf_pool);
if (!buf) { if (!buf) {
goto nomem; goto nomem;
} }
}
tcp_req->req.iov[i].iov_base = (void *)((uintptr_t)(buf + NVMF_DATA_BUFFER_MASK) & tcp_req->req.iov[i].iov_base = (void *)((uintptr_t)(buf + NVMF_DATA_BUFFER_MASK) &
~NVMF_DATA_BUFFER_MASK); ~NVMF_DATA_BUFFER_MASK);
@ -2175,17 +2208,10 @@ spdk_nvmf_tcp_req_fill_iovs(struct spdk_nvmf_tcp_transport *ttransport,
} }
tcp_req->data_from_pool = true; tcp_req->data_from_pool = true;
return 0; return 0;
nomem: nomem:
while (i) { spdk_nvmf_tcp_request_free_buffers(tcp_req, group, &ttransport->transport);
i--;
spdk_mempool_put(ttransport->transport.data_buf_pool, tcp_req->buffers[i]);
tcp_req->req.iov[i].iov_base = NULL;
tcp_req->req.iov[i].iov_len = 0;
}
tcp_req->req.iovcnt = 0; tcp_req->req.iovcnt = 0;
return -ENOMEM; return -ENOMEM;
} }
@ -2465,8 +2491,10 @@ spdk_nvmf_tcp_req_process(struct spdk_nvmf_tcp_transport *ttransport,
int rc; int rc;
enum spdk_nvmf_tcp_req_state prev_state; enum spdk_nvmf_tcp_req_state prev_state;
bool progress = false; bool progress = false;
struct spdk_nvmf_transport_poll_group *group;
tqpair = SPDK_CONTAINEROF(tcp_req->req.qpair, struct nvme_tcp_qpair, qpair); tqpair = SPDK_CONTAINEROF(tcp_req->req.qpair, struct nvme_tcp_qpair, qpair);
group = &tqpair->ch->tgroup->group;
assert(tcp_req->state != TCP_REQUEST_STATE_FREE); assert(tcp_req->state != TCP_REQUEST_STATE_FREE);
/* The loop here is to allow for several back-to-back state changes. */ /* The loop here is to allow for several back-to-back state changes. */
@ -2590,13 +2618,7 @@ spdk_nvmf_tcp_req_process(struct spdk_nvmf_tcp_transport *ttransport,
case TCP_REQUEST_STATE_COMPLETED: case TCP_REQUEST_STATE_COMPLETED:
spdk_trace_record(TRACE_TCP_REQUEST_STATE_COMPLETED, 0, 0, (uintptr_t)tcp_req, 0); spdk_trace_record(TRACE_TCP_REQUEST_STATE_COMPLETED, 0, 0, (uintptr_t)tcp_req, 0);
if (tcp_req->data_from_pool) { if (tcp_req->data_from_pool) {
/* Put the buffer/s back in the pool */ spdk_nvmf_tcp_request_free_buffers(tcp_req, group, &ttransport->transport);
for (uint32_t i = 0; i < tcp_req->req.iovcnt; i++) {
spdk_mempool_put(ttransport->transport.data_buf_pool, tcp_req->buffers[i]);
tcp_req->req.iov[i].iov_base = NULL;
tcp_req->buffers[i] = NULL;
}
tcp_req->data_from_pool = false;
} }
tcp_req->req.length = 0; tcp_req->req.length = 0;
tcp_req->req.iovcnt = 0; tcp_req->req.iovcnt = 0;
@ -2697,6 +2719,7 @@ spdk_nvmf_tcp_poll_group_add(struct spdk_nvmf_transport_poll_group *group,
return -1; return -1;
} }
tqpair->ch->tgroup = tgroup;
tqpair->state = NVME_TCP_QPAIR_STATE_INVALID; tqpair->state = NVME_TCP_QPAIR_STATE_INVALID;
tqpair->timeout = SPDK_NVME_TCP_QPAIR_EXIT_TIMEOUT; tqpair->timeout = SPDK_NVME_TCP_QPAIR_EXIT_TIMEOUT;
tqpair->last_pdu_time = spdk_get_ticks(); tqpair->last_pdu_time = spdk_get_ticks();

24
test/unit/lib/nvmf/tcp.c/tcp_ut.c
View File

@ -51,6 +51,7 @@
#define UT_IO_UNIT_SIZE 1024 #define UT_IO_UNIT_SIZE 1024
#define UT_MAX_AQ_DEPTH 64 #define UT_MAX_AQ_DEPTH 64
#define UT_SQ_HEAD_MAX 128 #define UT_SQ_HEAD_MAX 128
#define UT_NUM_SHARED_BUFFERS 128
SPDK_LOG_REGISTER_COMPONENT("nvmf", SPDK_LOG_NVMF) SPDK_LOG_REGISTER_COMPONENT("nvmf", SPDK_LOG_NVMF)
SPDK_LOG_REGISTER_COMPONENT("nvme", SPDK_LOG_NVME) SPDK_LOG_REGISTER_COMPONENT("nvme", SPDK_LOG_NVME)
@ -216,6 +217,7 @@ test_nvmf_tcp_create(void)
opts.max_io_size = UT_MAX_IO_SIZE; opts.max_io_size = UT_MAX_IO_SIZE;
opts.io_unit_size = UT_IO_UNIT_SIZE; opts.io_unit_size = UT_IO_UNIT_SIZE;
opts.max_aq_depth = UT_MAX_AQ_DEPTH; opts.max_aq_depth = UT_MAX_AQ_DEPTH;
opts.num_shared_buffers = UT_NUM_SHARED_BUFFERS;
/* expect success */ /* expect success */
transport = spdk_nvmf_tcp_create(&opts); transport = spdk_nvmf_tcp_create(&opts);
CU_ASSERT_PTR_NOT_NULL(transport); CU_ASSERT_PTR_NOT_NULL(transport);
@ -239,6 +241,7 @@ test_nvmf_tcp_create(void)
opts.max_io_size = UT_MAX_IO_SIZE; opts.max_io_size = UT_MAX_IO_SIZE;
opts.io_unit_size = UT_MAX_IO_SIZE + 1; opts.io_unit_size = UT_MAX_IO_SIZE + 1;
opts.max_aq_depth = UT_MAX_AQ_DEPTH; opts.max_aq_depth = UT_MAX_AQ_DEPTH;
opts.num_shared_buffers = UT_NUM_SHARED_BUFFERS;
/* expect success */ /* expect success */
transport = spdk_nvmf_tcp_create(&opts); transport = spdk_nvmf_tcp_create(&opts);
CU_ASSERT_PTR_NOT_NULL(transport); CU_ASSERT_PTR_NOT_NULL(transport);
@ -288,6 +291,7 @@ test_nvmf_tcp_destroy(void)
opts.max_io_size = UT_MAX_IO_SIZE; opts.max_io_size = UT_MAX_IO_SIZE;
opts.io_unit_size = UT_IO_UNIT_SIZE; opts.io_unit_size = UT_IO_UNIT_SIZE;
opts.max_aq_depth = UT_MAX_AQ_DEPTH; opts.max_aq_depth = UT_MAX_AQ_DEPTH;
opts.num_shared_buffers = UT_NUM_SHARED_BUFFERS;
transport = spdk_nvmf_tcp_create(&opts); transport = spdk_nvmf_tcp_create(&opts);
CU_ASSERT_PTR_NOT_NULL(transport); CU_ASSERT_PTR_NOT_NULL(transport);
transport->opts = opts; transport->opts = opts;
@ -300,17 +304,29 @@ test_nvmf_tcp_destroy(void)
static void static void
test_nvmf_tcp_poll_group_create(void) test_nvmf_tcp_poll_group_create(void)
{ {
struct spdk_nvmf_tcp_transport ttransport; struct spdk_nvmf_transport *transport;
struct spdk_nvmf_transport_poll_group *group; struct spdk_nvmf_transport_poll_group *group;
struct spdk_thread *thread; struct spdk_thread *thread;
struct spdk_nvmf_transport_opts opts;
thread = spdk_thread_create(NULL); thread = spdk_thread_create(NULL);
SPDK_CU_ASSERT_FATAL(thread != NULL); SPDK_CU_ASSERT_FATAL(thread != NULL);
spdk_set_thread(thread); spdk_set_thread(thread);
memset(&ttransport, 0, sizeof(ttransport)); memset(&opts, 0, sizeof(opts));
group = spdk_nvmf_tcp_poll_group_create(&ttransport.transport); opts.max_queue_depth = UT_MAX_QUEUE_DEPTH;
CU_ASSERT_PTR_NOT_NULL(group); opts.max_qpairs_per_ctrlr = UT_MAX_QPAIRS_PER_CTRLR;
opts.in_capsule_data_size = UT_IN_CAPSULE_DATA_SIZE;
opts.max_io_size = UT_MAX_IO_SIZE;
opts.io_unit_size = UT_IO_UNIT_SIZE;
opts.max_aq_depth = UT_MAX_AQ_DEPTH;
opts.num_shared_buffers = UT_NUM_SHARED_BUFFERS;
transport = spdk_nvmf_tcp_create(&opts);
CU_ASSERT_PTR_NOT_NULL(transport);
transport->opts = opts;
group = spdk_nvmf_tcp_poll_group_create(transport);
SPDK_CU_ASSERT_FATAL(group);
group->transport = transport;
spdk_nvmf_tcp_poll_group_destroy(group); spdk_nvmf_tcp_poll_group_destroy(group);
spdk_thread_exit(thread); spdk_thread_exit(thread);