From: Mateusz Guzik <mjguzik@gmail.com>
To: ebiederm@xmission.com, oleg@redhat.com
Cc: brauner@kernel.org, akpm@linux-foundation.org,
linux-mm@kvack.org, linux-kernel@vger.kernel.org,
Mateusz Guzik <mjguzik@gmail.com>
Subject: [PATCH v3 0/6] reduce tasklist_lock hold time on exit and do some
Date: Sat, 1 Feb 2025 17:31:00 +0100 [thread overview]
Message-ID: <20250201163106.28912-1-mjguzik@gmail.com> (raw)
The clone side contends against exit side in a way which avoidably
exacerbates the problem by the latter waiting on locks held by the
former while holding the tasklist_lock.
Whacking this for both add_device_randomness and pids allocation gives
me a 15% speed up for thread creation/destruction in a 24-core vm.
The random patch is worth about 4%.
nothing blew up with lockdep, lightly tested so far
Bench (plop into will-it-scale):
$ cat tests/threadspawn1.c
char *testcase_description = "Thread creation and teardown";
static void *worker(void *arg)
{
return (NULL);
}
void testcase(unsigned long long *iterations, unsigned long nr)
{
pthread_t thread;
int error;
while (1) {
error = pthread_create(&thread, NULL, worker, NULL);
assert(error == 0);
error = pthread_join(thread, NULL);
assert(error == 0);
(*iterations)++;
}
}
v3:
- keep procfs flush where it was, instead hoist get_pid outside of the
lock
- make detach_pid et al accept an array argument of pids to populate
- sprinkle asserts
- drop irq trips around pidmap_lock
- move tty unref outside of tasklist_lock
Mateusz Guzik (6):
exit: perform add_device_randomness() without tasklist_lock
exit: hoist get_pid() in release_task() outside of tasklist_lock
exit: postpone tty_kref_put() until after tasklist_lock is dropped
pid: sprinkle tasklist_lock asserts
pid: perform free_pid() calls outside of tasklist_lock
pid: drop irq disablement around pidmap_lock
include/linux/pid.h | 7 ++--
kernel/exit.c | 45 +++++++++++++++----------
kernel/pid.c | 82 +++++++++++++++++++++++++--------------------
kernel/sys.c | 14 +++++---
4 files changed, 86 insertions(+), 62 deletions(-)
--
2.43.0
next reply other threads:[~2025-02-01 16:31 UTC|newest]
Thread overview: 33+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-02-01 16:31 Mateusz Guzik [this message]
2025-02-01 16:31 ` [PATCH v3 1/6] exit: perform add_device_randomness() without tasklist_lock Mateusz Guzik
2025-02-03 17:51 ` Oleg Nesterov
2025-02-03 17:55 ` Mateusz Guzik
2025-02-01 16:31 ` [PATCH v3 2/6] exit: hoist get_pid() in release_task() outside of tasklist_lock Mateusz Guzik
2025-02-03 19:27 ` Liam R. Howlett
2025-02-03 19:35 ` Mateusz Guzik
2025-02-03 20:13 ` Liam R. Howlett
2025-02-03 20:22 ` Mateusz Guzik
2025-02-03 20:28 ` Liam R. Howlett
2025-02-04 1:51 ` Andrew Morton
2025-02-01 16:31 ` [PATCH v3 3/6] exit: postpone tty_kref_put() until after tasklist_lock is dropped Mateusz Guzik
2025-02-03 18:06 ` Oleg Nesterov
2025-02-03 19:33 ` Mateusz Guzik
2025-02-04 11:22 ` Oleg Nesterov
2025-02-04 12:12 ` Mateusz Guzik
2025-02-01 16:31 ` [PATCH v3 4/6] pid: sprinkle tasklist_lock asserts Mateusz Guzik
2025-02-01 16:31 ` [PATCH v3 5/6] pid: perform free_pid() calls outside of tasklist_lock Mateusz Guzik
2025-02-03 18:49 ` Oleg Nesterov
2025-02-03 19:31 ` Mateusz Guzik
2025-02-03 20:02 ` Oleg Nesterov
2025-02-01 16:31 ` [PATCH v3 6/6] pid: drop irq disablement around pidmap_lock Mateusz Guzik
2025-02-01 17:42 ` Oleg Nesterov
2025-02-01 17:45 ` Oleg Nesterov
2025-02-01 18:19 ` David Laight
2025-02-01 18:42 ` Mateusz Guzik
2025-02-01 21:51 ` David Laight
2025-02-01 22:00 ` Matthew Wilcox
2025-02-02 13:55 ` David Laight
2025-02-02 19:34 ` Mateusz Guzik
2025-02-02 20:44 ` David Laight
2025-02-02 22:06 ` Matthew Wilcox
2025-02-03 17:47 ` [PATCH v3 0/6] reduce tasklist_lock hold time on exit and do some Oleg Nesterov
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=20250201163106.28912-1-mjguzik@gmail.com \
--to=mjguzik@gmail.com \
--cc=akpm@linux-foundation.org \
--cc=brauner@kernel.org \
--cc=ebiederm@xmission.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=oleg@redhat.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