Re: [PATCH v3] eal/unix: optimize thread creation with glibc

[Stephen Hemminger <stephen@networkplumber.org>]

On Sun,  3 Nov 2024 12:25:19 +0100
David Marchand <david.marchand@redhat.com> wrote:

> Setting the cpu affinity of the child thread from the parent thread is
> racy when using pthread_setaffinity_np, as the child thread may start
> running and initialize before affinity is set.
> 
> On the other hand, setting the cpu affinity from the child thread itself
> may fail, so the parent thread waits for the child thread to report
> whether this call succeeded.
> 
> This synchronisation point resulted in a significant slow down of
> rte_thread_create() (as seen in the lcores_autotest unit tests, in OBS
> for some ARM systems).
> 
> Another option for setting cpu affinity is to use the not portable
> pthread_attr_setaffinity_np, but it is not available with musl.
> 
> Fixes: b28c6196b132 ("eal/unix: fix thread creation")
> Cc: stable@dpdk.org
> 
> Signed-off-by: David Marchand <david.marchand@redhat.com>
> Acked-by: Luca Boccassi <bluca@debian.org>
> ---
> Changes since v2:
> - added pthread_attr_setaffinity_np() detection,
> 
> Changes since v1:
> - fixed build with FreeBSD,


Looks good, reading the glibc source there are internal flags
to deal with all the states new threads have to deal with.
Glibc is complex there, but comments are worth reading.

I wonder if part of the reason this is more of a problem for DPDK
is that by default new thread inherits cpuset of the parent.
So the child thread will get stuck on same CPU as the main thread.

Acked-by: Stephen Hemminger <stephen@networkplumber.org>