[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <ccfef0d5-c419-0a8d-cf75-9642e9d11f94@torproject.org>
Date: Fri, 12 Mar 2021 12:39:12 -0600
From: Jim Newsome <jnewsome@...project.org>
To: Andrew Morton <akpm@...ux-foundation.org>
Cc: Oleg Nesterov <oleg@...hat.com>,
"Eric W . Biederman" <ebiederm@...ssion.com>,
Christian Brauner <christian@...uner.io>,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH v5] do_wait: make PIDTYPE_PID case O(1) instead of O(n)
On 3/12/21 12:22, Andrew Morton wrote:
>
> Could we please see some performance testing results to permit us to
> evaluate the value of this change?
Sure. I've been doing some ad-hoc measurements with the code below. It
forks 8k children and then waits for them in reverse order (forcing a
full list traversal each time). I'll need to reboot a couple times to
get apples-to-apples measurements on bare metal, though. I'll plan to
run with NUMCHILDREN = 0 -> 8000, by 100.
Does this look like it'd be sufficient, or is there more you'd like to
see? The current form doesn't use ptrace, but I expect the results to be
similar; (maybe more pronounced when tracing threaded children, since
every thread is in the tracee list instead of just the group leaders).
#define NUMCHILDREN 8000
void fork_and_wait() {
pid_t children[NUMCHILDREN];
for (int i = 0; i < NUMCHILDREN; ++i) {
pid_t forkrv = fork();
if (forkrv < 0) {
perror("fork");
exit(1);
}
if (forkrv == 0) {
// child
exit(0);
}
// parent
children[i] = forkrv;
}
for (int i = 0; i < NUMCHILDREN; ++i) {
int wstatus;
if (waitpid(children[NUMCHILDREN - i - 1], &wstatus, 0) < 0) {
perror("waitpid");
exit(1);
}
}
}
Powered by blists - more mailing lists