From: Kaisen You <kaisenx.you@intel.com>
To: dev@dpdk.org
Cc: yidingx.zhou@intel.com, thomas@monjalon.net,
david.marchand@redhat.com, olivier.matz@6wind.com,
ferruh.yigit@amd.com, kaisenx.you@intel.com, zhoumin@loongson.cn,
anatoly.burakov@intel.com, stable@dpdk.org
Subject: [PATCH v7] enhance NUMA affinity heuristic
Date: Tue, 23 May 2023 10:50:04 +0800 [thread overview]
Message-ID: <20230523025004.192071-1-kaisenx.you@intel.com> (raw)
In-Reply-To: <20230425051649.1109428-1-kaisenx.you@intel.com>
When a DPDK application is started on only one numa node, memory is
allocated for only one socket. When interrupt threads use memory,
memory may not be found on the socket where the interrupt thread
is currently located, and memory has to be reallocated on the hugepage,
this operation will lead to performance degradation.
Fixes: 705356f0811f ("eal: simplify control thread creation")
Fixes: 770d41bf3309 ("malloc: fix allocation with unknown socket ID")
Cc: stable@dpdk.org
Signed-off-by: Kaisen You <kaisenx.you@intel.com>
---
Changes since v6:
- New explanation for easy understanding,
Changes since v5:
- Add comments to the code,
Changes since v4:
- mod the patch title,
Changes since v3:
- add the assignment of socket_id in thread initialization,
Changes since v2:
- add uncommitted local change and fix compilation,
Changes since v1:
- accomodate for configurations with main lcore running on multiples
physical cores belonging to different numa,
---
lib/eal/common/eal_common_thread.c | 6 ++++++
lib/eal/common/malloc_heap.c | 9 +++++++++
2 files changed, 15 insertions(+)
diff --git a/lib/eal/common/eal_common_thread.c b/lib/eal/common/eal_common_thread.c
index 079a385630..6479b66da1 100644
--- a/lib/eal/common/eal_common_thread.c
+++ b/lib/eal/common/eal_common_thread.c
@@ -252,6 +252,12 @@ static int ctrl_thread_init(void *arg)
struct rte_thread_ctrl_params *params = arg;
__rte_thread_init(rte_lcore_id(), cpuset);
+ /* set the value of the per-core variable _socket_id to SOCKET_ID_ANY.
+ * Satisfy the judgment condition when threads find memory.
+ * If SOCKET_ID_ANY is not specified, the thread may go to a node with
+ * unallocated memory in a subsequent memory search.
+ */
+ RTE_PER_LCORE(_socket_id) = SOCKET_ID_ANY;
params->ret = rte_thread_set_affinity_by_id(rte_thread_self(), cpuset);
if (params->ret != 0) {
__atomic_store_n(¶ms->ctrl_thread_status,
diff --git a/lib/eal/common/malloc_heap.c b/lib/eal/common/malloc_heap.c
index d25bdc98f9..6d37f8afee 100644
--- a/lib/eal/common/malloc_heap.c
+++ b/lib/eal/common/malloc_heap.c
@@ -716,6 +716,15 @@ malloc_get_numa_socket(void)
if (conf->socket_mem[socket_id] != 0)
return socket_id;
}
+ /* Trying to allocate memory on the main lcore numa node.
+ * especially when the DPDK application is started only on one numa node.
+ */
+ socket_id = rte_lcore_to_socket_id(rte_get_main_lcore());
+ /* When the socket_id obtained in the main lcore numa is SOCKET_ID_ANY,
+ * The probability of finding memory on rte_socket_id_by_idx(0) is higher.
+ */
+ if (socket_id != (unsigned int)SOCKET_ID_ANY)
+ return socket_id;
return rte_socket_id_by_idx(0);
}
--
2.25.1
next prev parent reply other threads:[~2023-05-23 3:07 UTC|newest]
Thread overview: 31+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <20221221104858.296530-1-david.marchand@redhat.com>
2023-01-31 15:05 ` [PATCH v4] net/iavf:enhance " Kaisen You
2023-01-31 16:05 ` Thomas Monjalon
2023-02-01 5:32 ` You, KaisenX
2023-02-01 12:20 ` [PATCH v5] enhance " Kaisen You
2023-02-15 14:22 ` Burakov, Anatoly
2023-02-15 14:47 ` Burakov, Anatoly
2023-02-16 2:50 ` You, KaisenX
2023-03-03 14:07 ` Thomas Monjalon
2023-03-09 1:58 ` You, KaisenX
2023-04-13 0:56 ` You, KaisenX
2023-04-19 12:16 ` Thomas Monjalon
2023-04-21 2:34 ` You, KaisenX
2023-04-21 8:12 ` Thomas Monjalon
2023-04-23 6:52 ` You, KaisenX
2023-04-23 8:57 ` You, KaisenX
2023-04-23 13:19 ` Thomas Monjalon
2023-04-25 5:16 ` [PATCH v6] " Kaisen You
2023-04-27 6:57 ` Thomas Monjalon
2023-05-16 5:19 ` You, KaisenX
2023-05-23 2:50 ` Kaisen You [this message]
2023-05-23 10:44 ` [PATCH v7] " Burakov, Anatoly
2023-05-26 6:44 ` You, KaisenX
2023-05-23 12:45 ` Burakov, Anatoly
2023-05-26 6:50 ` [PATCH v8] " Kaisen You
2023-05-26 8:45 ` Kaisen You
2023-05-26 14:44 ` Burakov, Anatoly
2023-05-26 17:50 ` Stephen Hemminger
2023-05-29 10:37 ` Burakov, Anatoly
2023-06-01 14:42 ` David Marchand
2023-06-06 14:04 ` Thomas Monjalon
2023-06-12 9:36 ` Burakov, Anatoly
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=20230523025004.192071-1-kaisenx.you@intel.com \
--to=kaisenx.you@intel.com \
--cc=anatoly.burakov@intel.com \
--cc=david.marchand@redhat.com \
--cc=dev@dpdk.org \
--cc=ferruh.yigit@amd.com \
--cc=olivier.matz@6wind.com \
--cc=stable@dpdk.org \
--cc=thomas@monjalon.net \
--cc=yidingx.zhou@intel.com \
--cc=zhoumin@loongson.cn \
/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).