From: Bruce Richardson <bruce.richardson@intel.com>
To: olivier.matz@6wind.com
Cc: thomas.monjalon@6wind.com, dev@dpdk.org,
Bruce Richardson <bruce.richardson@intel.com>
Subject: [dpdk-dev] [PATCH v4 05/14] ring: remove the yield when waiting for tail update
Date: Tue, 28 Mar 2017 21:35:57 +0100 [thread overview]
Message-ID: <20170328203606.27457-6-bruce.richardson@intel.com> (raw)
In-Reply-To: <20170328203606.27457-1-bruce.richardson@intel.com>
There was a compile time setting to enable a ring to yield when
it entered a loop in mp or mc rings waiting for the tail pointer update.
Build time settings are not recommended for enabling/disabling features,
and since this was off by default, remove it completely. If needed, a
runtime enabled equivalent can be used.
Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
Acked-by: Olivier Matz <olivier.matz@6wind.com>
---
config/common_base | 1 -
doc/guides/prog_guide/env_abstraction_layer.rst | 5 ----
doc/guides/rel_notes/release_17_05.rst | 1 +
lib/librte_ring/rte_ring.h | 35 +++++--------------------
4 files changed, 7 insertions(+), 35 deletions(-)
diff --git a/config/common_base b/config/common_base
index 69e91ae..2d54ddf 100644
--- a/config/common_base
+++ b/config/common_base
@@ -452,7 +452,6 @@ CONFIG_RTE_LIBRTE_PMD_NULL_CRYPTO=y
# Compile librte_ring
#
CONFIG_RTE_LIBRTE_RING=y
-CONFIG_RTE_RING_PAUSE_REP_COUNT=0
#
# Compile librte_mempool
diff --git a/doc/guides/prog_guide/env_abstraction_layer.rst b/doc/guides/prog_guide/env_abstraction_layer.rst
index 10a10a8..7c39cd2 100644
--- a/doc/guides/prog_guide/env_abstraction_layer.rst
+++ b/doc/guides/prog_guide/env_abstraction_layer.rst
@@ -352,11 +352,6 @@ Known Issues
3. It MUST not be used by multi-producer/consumer pthreads, whose scheduling policies are SCHED_FIFO or SCHED_RR.
- ``RTE_RING_PAUSE_REP_COUNT`` is defined for rte_ring to reduce contention. It's mainly for case 2, a yield is issued after number of times pause repeat.
-
- It adds a sched_yield() syscall if the thread spins for too long while waiting on the other thread to finish its operations on the ring.
- This gives the preempted thread a chance to proceed and finish with the ring enqueue/dequeue operation.
-
+ rte_timer
Running ``rte_timer_manager()`` on a non-EAL pthread is not allowed. However, resetting/stopping the timer from a non-EAL pthread is allowed.
diff --git a/doc/guides/rel_notes/release_17_05.rst b/doc/guides/rel_notes/release_17_05.rst
index 50123c2..25d8549 100644
--- a/doc/guides/rel_notes/release_17_05.rst
+++ b/doc/guides/rel_notes/release_17_05.rst
@@ -134,6 +134,7 @@ API Changes
* removed the build-time setting ``CONFIG_RTE_RING_SPLIT_PROD_CONS``
* removed the build-time setting ``CONFIG_RTE_LIBRTE_RING_DEBUG``
+ * removed the build-time setting ``CONFIG_RTE_RING_PAUSE_REP_COUNT``
ABI Changes
-----------
diff --git a/lib/librte_ring/rte_ring.h b/lib/librte_ring/rte_ring.h
index 2777b41..f8ac7f5 100644
--- a/lib/librte_ring/rte_ring.h
+++ b/lib/librte_ring/rte_ring.h
@@ -114,11 +114,6 @@ enum rte_ring_queue_behavior {
#define RTE_RING_NAMESIZE (RTE_MEMZONE_NAMESIZE - \
sizeof(RTE_RING_MZ_PREFIX) + 1)
-#ifndef RTE_RING_PAUSE_REP_COUNT
-#define RTE_RING_PAUSE_REP_COUNT 0 /**< Yield after pause num of times, no yield
- * if RTE_RING_PAUSE_REP not defined. */
-#endif
-
struct rte_memzone; /* forward declaration, so as not to require memzone.h */
#if RTE_CACHE_LINE_SIZE < 128
@@ -393,7 +388,7 @@ __rte_ring_mp_do_enqueue(struct rte_ring *r, void * const *obj_table,
uint32_t cons_tail, free_entries;
const unsigned max = n;
int success;
- unsigned i, rep = 0;
+ unsigned int i;
uint32_t mask = r->mask;
int ret;
@@ -447,18 +442,9 @@ __rte_ring_mp_do_enqueue(struct rte_ring *r, void * const *obj_table,
* If there are other enqueues in progress that preceded us,
* we need to wait for them to complete
*/
- while (unlikely(r->prod.tail != prod_head)) {
+ while (unlikely(r->prod.tail != prod_head))
rte_pause();
- /* Set RTE_RING_PAUSE_REP_COUNT to avoid spin too long waiting
- * for other thread finish. It gives pre-empted thread a chance
- * to proceed and finish with ring dequeue operation. */
- if (RTE_RING_PAUSE_REP_COUNT &&
- ++rep == RTE_RING_PAUSE_REP_COUNT) {
- rep = 0;
- sched_yield();
- }
- }
r->prod.tail = prod_next;
return ret;
}
@@ -491,7 +477,7 @@ __rte_ring_sp_do_enqueue(struct rte_ring *r, void * const *obj_table,
{
uint32_t prod_head, cons_tail;
uint32_t prod_next, free_entries;
- unsigned i;
+ unsigned int i;
uint32_t mask = r->mask;
int ret;
@@ -568,7 +554,7 @@ __rte_ring_mc_do_dequeue(struct rte_ring *r, void **obj_table,
uint32_t cons_next, entries;
const unsigned max = n;
int success;
- unsigned i, rep = 0;
+ unsigned int i;
uint32_t mask = r->mask;
/* Avoid the unnecessary cmpset operation below, which is also
@@ -613,18 +599,9 @@ __rte_ring_mc_do_dequeue(struct rte_ring *r, void **obj_table,
* If there are other dequeues in progress that preceded us,
* we need to wait for them to complete
*/
- while (unlikely(r->cons.tail != cons_head)) {
+ while (unlikely(r->cons.tail != cons_head))
rte_pause();
- /* Set RTE_RING_PAUSE_REP_COUNT to avoid spin too long waiting
- * for other thread finish. It gives pre-empted thread a chance
- * to proceed and finish with ring dequeue operation. */
- if (RTE_RING_PAUSE_REP_COUNT &&
- ++rep == RTE_RING_PAUSE_REP_COUNT) {
- rep = 0;
- sched_yield();
- }
- }
r->cons.tail = cons_next;
return behavior == RTE_RING_QUEUE_FIXED ? 0 : n;
@@ -659,7 +636,7 @@ __rte_ring_sc_do_dequeue(struct rte_ring *r, void **obj_table,
{
uint32_t cons_head, prod_tail;
uint32_t cons_next, entries;
- unsigned i;
+ unsigned int i;
uint32_t mask = r->mask;
cons_head = r->cons.head;
--
2.9.3
next prev parent reply other threads:[~2017-03-28 20:36 UTC|newest]
Thread overview: 116+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-02-23 17:23 [dpdk-dev] [PATCH v1 00/14] refactor and cleanup of rte_ring Bruce Richardson
2017-02-23 17:23 ` [dpdk-dev] [PATCH v1 01/14] ring: remove split cacheline build setting Bruce Richardson
2017-02-28 11:35 ` Jerin Jacob
2017-02-28 11:57 ` Bruce Richardson
2017-02-28 12:08 ` Jerin Jacob
2017-02-28 13:52 ` Bruce Richardson
2017-02-28 17:54 ` Jerin Jacob
2017-03-01 9:47 ` Bruce Richardson
2017-03-01 10:17 ` Olivier Matz
2017-03-01 10:42 ` Bruce Richardson
2017-03-01 11:06 ` Olivier Matz
2017-03-01 11:19 ` Jerin Jacob
2017-03-01 12:12 ` Bruce Richardson
2017-02-23 17:23 ` [dpdk-dev] [PATCH v1 02/14] ring: create common structure for prod and cons metadata Bruce Richardson
2017-03-01 10:22 ` Olivier Matz
2017-03-01 10:33 ` Bruce Richardson
2017-02-23 17:23 ` [dpdk-dev] [PATCH v1 03/14] ring: eliminate duplication of size and mask fields Bruce Richardson
2017-02-23 17:23 ` [dpdk-dev] [PATCH v1 04/14] ring: remove debug setting Bruce Richardson
2017-02-23 17:23 ` [dpdk-dev] [PATCH v1 05/14] ring: remove the yield when waiting for tail update Bruce Richardson
2017-02-23 17:23 ` [dpdk-dev] [PATCH v1 06/14] ring: remove watermark support Bruce Richardson
2017-03-01 10:34 ` Olivier Matz
2017-03-01 10:43 ` Bruce Richardson
2017-02-23 17:24 ` [dpdk-dev] [PATCH v1 07/14] ring: make bulk and burst fn return vals consistent Bruce Richardson
2017-02-23 17:24 ` [dpdk-dev] [PATCH v1 08/14] ring: allow enqueue fns to return free space value Bruce Richardson
2017-02-23 17:24 ` [dpdk-dev] [PATCH v1 09/14] ring: allow dequeue fns to return remaining entry count Bruce Richardson
2017-02-23 17:24 ` [dpdk-dev] [PATCH v1 10/14] examples/quota_watermark: use ring space for watermarks Bruce Richardson
2017-02-23 17:24 ` [dpdk-dev] [PATCH v1 11/14] ring: reduce scope of local variables Bruce Richardson
2017-02-23 17:24 ` [dpdk-dev] [PATCH v1 12/14] ring: separate out head index manipulation for enq/deq Bruce Richardson
2017-03-08 10:49 ` Olivier MATZ
2017-03-08 12:06 ` Bruce Richardson
2017-03-14 8:56 ` Olivier Matz
2017-02-23 17:24 ` [dpdk-dev] [PATCH v1 13/14] ring: create common function for updating tail idx Bruce Richardson
2017-02-23 17:24 ` [dpdk-dev] [PATCH v1 14/14] ring: make ring struct and enq/deq macros type agnostic Bruce Richardson
2017-03-07 11:32 ` [dpdk-dev] [PATCH v2 00/14] refactor and cleanup of rte_ring Bruce Richardson
2017-03-07 11:32 ` [dpdk-dev] [PATCH v2 01/14] ring: remove split cacheline build setting Bruce Richardson
2017-03-07 11:32 ` [dpdk-dev] [PATCH v2 02/14] ring: create common structure for prod and cons metadata Bruce Richardson
2017-03-15 14:01 ` Thomas Monjalon
2017-03-22 16:38 ` Bruce Richardson
2017-03-24 14:55 ` Bruce Richardson
2017-03-24 16:41 ` Olivier Matz
2017-03-24 16:57 ` Bruce Richardson
2017-03-07 11:32 ` [dpdk-dev] [PATCH v2 03/14] ring: eliminate duplication of size and mask fields Bruce Richardson
2017-03-07 11:32 ` [dpdk-dev] [PATCH v2 04/14] ring: remove debug setting Bruce Richardson
2017-03-07 11:32 ` [dpdk-dev] [PATCH v2 05/14] ring: remove the yield when waiting for tail update Bruce Richardson
2017-03-07 11:32 ` [dpdk-dev] [PATCH v2 06/14] ring: remove watermark support Bruce Richardson
2017-03-07 11:32 ` [dpdk-dev] [PATCH v2 07/14] ring: make bulk and burst fn return vals consistent Bruce Richardson
2017-03-08 10:22 ` Olivier MATZ
2017-03-08 12:08 ` Bruce Richardson
2017-03-14 8:56 ` Olivier Matz
2017-03-07 11:32 ` [dpdk-dev] [PATCH v2 08/14] ring: allow enqueue fns to return free space value Bruce Richardson
2017-03-07 11:32 ` [dpdk-dev] [PATCH v2 09/14] ring: allow dequeue fns to return remaining entry count Bruce Richardson
2017-03-07 11:32 ` [dpdk-dev] [PATCH v2 10/14] examples/quota_watermark: use ring space for watermarks Bruce Richardson
2017-03-07 11:32 ` [dpdk-dev] [PATCH v2 11/14] ring: reduce scope of local variables Bruce Richardson
2017-03-07 11:32 ` [dpdk-dev] [PATCH v2 12/14] ring: separate out head index manipulation for enq/deq Bruce Richardson
2017-03-07 11:32 ` [dpdk-dev] [PATCH v2 13/14] ring: create common function for updating tail idx Bruce Richardson
2017-03-07 11:32 ` [dpdk-dev] [PATCH v2 14/14] ring: make ring struct and enq/deq macros type agnostic Bruce Richardson
2017-03-14 8:59 ` [dpdk-dev] [PATCH v2 00/14] refactor and cleanup of rte_ring Olivier Matz
2017-03-24 17:09 ` [dpdk-dev] [PATCH v3 " Bruce Richardson
2017-03-24 17:09 ` [dpdk-dev] [PATCH v3 01/14] ring: remove split cacheline build setting Bruce Richardson
2017-03-24 17:09 ` [dpdk-dev] [PATCH v3 02/14] ring: create common structure for prod and cons metadata Bruce Richardson
2017-03-27 7:20 ` Olivier Matz
2017-03-24 17:09 ` [dpdk-dev] [PATCH v3 03/14] ring: eliminate duplication of size and mask fields Bruce Richardson
2017-03-27 9:52 ` Thomas Monjalon
2017-03-27 10:13 ` Bruce Richardson
2017-03-27 10:15 ` Bruce Richardson
2017-03-27 13:13 ` Thomas Monjalon
2017-03-27 14:57 ` Bruce Richardson
2017-03-24 17:09 ` [dpdk-dev] [PATCH v3 04/14] ring: remove debug setting Bruce Richardson
2017-03-24 17:09 ` [dpdk-dev] [PATCH v3 05/14] ring: remove the yield when waiting for tail update Bruce Richardson
2017-03-24 17:10 ` [dpdk-dev] [PATCH v3 06/14] ring: remove watermark support Bruce Richardson
2017-03-24 17:10 ` [dpdk-dev] [PATCH v3 07/14] ring: make bulk and burst fn return vals consistent Bruce Richardson
2017-03-24 17:10 ` [dpdk-dev] [PATCH v3 08/14] ring: allow enqueue fns to return free space value Bruce Richardson
2017-03-28 7:12 ` Thomas Monjalon
2017-03-28 8:16 ` Bruce Richardson
2017-03-24 17:10 ` [dpdk-dev] [PATCH v3 09/14] ring: allow dequeue fns to return remaining entry count Bruce Richardson
2017-03-24 17:10 ` [dpdk-dev] [PATCH v3 10/14] examples/quota_watermark: use ring space for watermarks Bruce Richardson
2017-03-24 17:10 ` [dpdk-dev] [PATCH v3 11/14] ring: reduce scope of local variables Bruce Richardson
2017-03-24 17:10 ` [dpdk-dev] [PATCH v3 12/14] ring: separate out head index manipulation for enq/deq Bruce Richardson
2017-03-24 17:10 ` [dpdk-dev] [PATCH v3 13/14] ring: create common function for updating tail idx Bruce Richardson
2017-03-24 17:10 ` [dpdk-dev] [PATCH v3 14/14] ring: make ring struct and enq/deq macros type agnostic Bruce Richardson
2017-03-28 20:35 ` [dpdk-dev] [PATCH v4 00/14] refactor and cleanup of rte_ring Bruce Richardson
2017-03-28 20:35 ` [dpdk-dev] [PATCH v4 01/14] ring: remove split cacheline build setting Bruce Richardson
2017-03-28 20:35 ` [dpdk-dev] [PATCH v4 02/14] ring: create common structure for prod and cons metadata Bruce Richardson
2017-03-28 20:35 ` [dpdk-dev] [PATCH v4 03/14] ring: eliminate duplication of size and mask fields Bruce Richardson
2017-03-28 20:35 ` [dpdk-dev] [PATCH v4 04/14] ring: remove debug setting Bruce Richardson
2017-03-28 20:35 ` Bruce Richardson [this message]
2017-03-28 20:35 ` [dpdk-dev] [PATCH v4 06/14] ring: remove watermark support Bruce Richardson
2017-03-28 20:35 ` [dpdk-dev] [PATCH v4 07/14] ring: make bulk and burst fn return vals consistent Bruce Richardson
2017-03-28 20:36 ` [dpdk-dev] [PATCH v4 08/14] ring: allow enqueue fns to return free space value Bruce Richardson
2017-03-28 20:36 ` [dpdk-dev] [PATCH v4 09/14] ring: allow dequeue fns to return remaining entry count Bruce Richardson
2017-03-28 20:36 ` [dpdk-dev] [PATCH v4 10/14] examples/quota_watermark: use ring space for watermarks Bruce Richardson
2017-03-28 20:36 ` [dpdk-dev] [PATCH v4 11/14] ring: reduce scope of local variables Bruce Richardson
2017-03-28 20:36 ` [dpdk-dev] [PATCH v4 12/14] ring: separate out head index manipulation for enq/deq Bruce Richardson
2017-03-28 20:36 ` [dpdk-dev] [PATCH v4 13/14] ring: create common function for updating tail idx Bruce Richardson
2017-03-28 20:36 ` [dpdk-dev] [PATCH v4 14/14] ring: make ring struct and enq/deq macros type agnostic Bruce Richardson
2017-03-29 2:47 ` [dpdk-dev] [PATCH v4 00/14] refactor and cleanup of rte_ring Yuanhan Liu
2017-03-29 13:09 ` [dpdk-dev] [PATCH v5 " Bruce Richardson
2017-03-29 13:09 ` [dpdk-dev] [PATCH v5 01/14] ring: remove split cacheline build setting Bruce Richardson
2017-03-29 13:09 ` [dpdk-dev] [PATCH v5 02/14] ring: create common structure for prod and cons metadata Bruce Richardson
2017-03-29 13:09 ` [dpdk-dev] [PATCH v5 03/14] ring: eliminate duplication of size and mask fields Bruce Richardson
2017-03-29 13:09 ` [dpdk-dev] [PATCH v5 04/14] ring: remove debug setting Bruce Richardson
2017-03-29 13:09 ` [dpdk-dev] [PATCH v5 05/14] ring: remove the yield when waiting for tail update Bruce Richardson
2017-03-29 13:09 ` [dpdk-dev] [PATCH v5 06/14] ring: remove watermark support Bruce Richardson
2017-03-29 13:09 ` [dpdk-dev] [PATCH v5 07/14] ring: make bulk and burst fn return vals consistent Bruce Richardson
2017-04-13 6:42 ` Wang, Zhihong
2017-04-13 8:33 ` Bruce Richardson
2017-03-29 13:09 ` [dpdk-dev] [PATCH v5 08/14] ring: allow enqueue fns to return free space value Bruce Richardson
2017-03-29 13:09 ` [dpdk-dev] [PATCH v5 09/14] ring: allow dequeue fns to return remaining entry count Bruce Richardson
2017-03-29 13:09 ` [dpdk-dev] [PATCH v5 10/14] examples/quota_watermark: use ring space for watermarks Bruce Richardson
2017-03-29 13:09 ` [dpdk-dev] [PATCH v5 11/14] ring: reduce scope of local variables Bruce Richardson
2017-03-29 13:09 ` [dpdk-dev] [PATCH v5 12/14] ring: separate out head index manipulation for enq/deq Bruce Richardson
2017-03-29 13:09 ` [dpdk-dev] [PATCH v5 13/14] ring: create common function for updating tail idx Bruce Richardson
2017-03-29 13:09 ` [dpdk-dev] [PATCH v5 14/14] ring: make ring struct and enq/deq macros type agnostic Bruce Richardson
2017-03-29 20:33 ` [dpdk-dev] [PATCH v5 00/14] refactor and cleanup of rte_ring Thomas Monjalon
2017-03-31 14:37 ` Ferruh Yigit
2017-04-03 17:55 ` Thomas Monjalon
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20170328203606.27457-6-bruce.richardson@intel.com \
--to=bruce.richardson@intel.com \
--cc=dev@dpdk.org \
--cc=olivier.matz@6wind.com \
--cc=thomas.monjalon@6wind.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).