From: Eric Dumazet <dada1@cosmosbay.com>
To: Andi Kleen <ak@suse.de>
Cc: Kirill Korotaev <dev@sw.ru>,
linux-kernel@vger.kernel.org, Linus Torvalds <torvalds@osdl.org>,
Andrew Morton <akpm@osdl.org>,
xemul@sw.ru, Andrey Savochkin <saw@sawoct.com>,
st@sw.ru, discuss@x86-64.org
Subject: Re: SMP syncronization on AMD processors (broken?)
Date: Thu, 06 Oct 2005 15:50:36 +0200 [thread overview]
Message-ID: <43452BAC.3000306@cosmosbay.com> (raw)
In-Reply-To: <p73hdbuzs7l.fsf@verdi.suse.de>
Andi Kleen a écrit :
> Kirill Korotaev <dev@sw.ru> writes:
>
>
>>Please help with a not simple question about spin_lock/spin_unlock on
>>SMP archs. The question is whether concurrent spin_lock()'s should
>>acquire it in more or less "fair" fashinon or one of CPUs can starve
>>any arbitrary time while others do reacquire it in a loop.
>
>
> They are not fully fair because of the NUMAness of the system.
> Same on many other NUMA systems.
>
> We considered long ago to use queued locks to avoid this, but
> they are quite costly for the uncongested case and never seemed worth it.
>
> So live with it.
Unrelated, but that reminds me that current spinlock implementation on x86
imply that NR_CPUS should be < 128.
Maybe we should reflect this in Kconfig ?
config NR_CPUS
range 2 128
Or use a plain int for spinlock, instead of a signed char.
Eric
next prev parent reply other threads:[~2005-10-06 13:51 UTC|newest]
Thread overview: 31+ messages / expand[flat|nested] mbox.gz Atom feed top
2005-10-06 13:05 SMP syncronization on AMD processors (broken?) Kirill Korotaev
2005-10-06 13:14 ` linux-os (Dick Johnson)
2005-10-06 13:19 ` Arjan van de Ven
2005-10-06 13:32 ` Andrey Savochkin
2005-10-06 14:22 ` Arjan van de Ven
2005-10-06 13:32 ` Andi Kleen
2005-10-06 13:46 ` Andrey Savochkin
2005-10-06 14:02 ` [discuss] " Andi Kleen
2005-10-06 14:52 ` Linus Torvalds
2005-10-06 15:21 ` Andrey Savochkin
2005-10-06 15:46 ` Linus Torvalds
2005-10-11 0:59 ` Andrew Morton
2005-10-11 1:20 ` Andi Kleen
2005-10-11 3:20 ` Joe Seigh
2005-10-06 13:50 ` Eric Dumazet [this message]
2005-10-06 13:56 ` [discuss] " Andi Kleen
2005-10-06 14:10 ` Eric Dumazet
2005-10-06 14:11 ` Andi Kleen
2005-10-06 14:45 ` Linus Torvalds
2005-10-06 15:34 ` Hugh Dickins
2005-10-06 15:53 ` Eric Dumazet
2005-10-06 16:01 ` Linus Torvalds
2005-10-07 20:38 ` Joe Seigh
2005-10-07 20:57 ` Stephen Hemminger
2005-10-13 18:24 ` Joe Seigh
-- strict thread matches above, loose matches on Subject: below --
2005-10-08 9:31 Chuck Ebbert
2005-10-11 23:50 linux
2005-10-12 2:12 ` Christopher Friesen
2005-10-12 2:39 ` linux
2005-10-12 3:27 ` Kyle Moffett
2005-10-13 12:25 ` Kirill Korotaev
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=43452BAC.3000306@cosmosbay.com \
--to=dada1@cosmosbay.com \
--cc=ak@suse.de \
--cc=akpm@osdl.org \
--cc=dev@sw.ru \
--cc=discuss@x86-64.org \
--cc=linux-kernel@vger.kernel.org \
--cc=saw@sawoct.com \
--cc=st@sw.ru \
--cc=torvalds@osdl.org \
--cc=xemul@sw.ru \
/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