LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Pierre Peiffer <pierre.peiffer@bull.net>
To: LKML <linux-kernel@vger.kernel.org>
Cc: Andrew Morton <akpm@osdl.org>,
	Ulrich Drepper <drepper@redhat.com>, Ingo Molnar <mingo@elte.hu>,
	Jakub Jelinek <jakub@redhat.com>
Subject: Re: [PATCH 2.6.20-rc4 0/4] futexes functionalities and improvements
Date: Fri, 12 Jan 2007 08:53:52 +0100	[thread overview]
Message-ID: <45A73E90.7050805@bull.net> (raw)
In-Reply-To: <20070111134615.34902742.akpm@osdl.org>

Andrew Morton a écrit :
 > OK.  Unfortunately patches 2-4 don't apply without #1 present and the fix
 > is not immediately obvious, so we'll need a respin+retest, please.

Ok, I'll provide updated patches for -mm ASAP.

> On Thu, 11 Jan 2007 09:47:28 -0800
> Ulrich Drepper <drepper@redhat.com> wrote:

>> if the patches allow this, I'd like to see parts 2, 3, and 4 to be in
>> -mm ASAP.  Especially the 64-bit variants are urgently needed.  Just
>> hold off adding the plist use, I am still not convinced that
>> unconditional use is a good thing, especially with one single global list.

Just to avoid any misunderstanding (I (really) understand your point about 
performance issue), but:

* the problem I mention about several futexes hashed on the same key, and thus 
with all potential waiters listed on the same list, is _not_ a new problem which 
comes with this patch: it already exists today, with simple list.

* the measures of performance done with pthread_broadcast (and thus with 
futex_requeue) is a good choice (well, may be not realistic, when considering 
real applications (*)) to put in evidence the performance impact, rather than 
threads making FUTEX_WAIT/FUTEX_WAKE: what is expensive with plist is the 
plist_add operation (which occurs in FUTEX_WAIT), not plist_del (which occurs 
during FUTEX_WAKE => thus, no big impact should be noticed here). Any measure 
will be difficult to do with only FUTEX_WAIT/WAKE.

=> futex_requeue does as many plist_del/plist_add operations as the number of 
threads waiting (minus 1), and thus has a direct impact on the time needed to 
wake everybody (or to wake the first thread to be more precise).

(*) I'll try the volano bench, if I have time.


-- 
Pierre

  parent reply	other threads:[~2007-01-12  7:54 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-01-09 16:15 Pierre Peiffer
2007-01-11 17:47 ` Ulrich Drepper
     [not found]   ` <20070111134615.34902742.akpm@osdl.org>
2007-01-12  7:53     ` Pierre Peiffer [this message]
2007-01-12  7:58       ` Ingo Molnar
2007-01-16  8:34         ` Pierre Peiffer
2007-01-16  9:44           ` Ingo Molnar
2007-01-16 15:14           ` Ulrich Drepper
2007-01-16 15:40             ` Ingo Molnar
2007-01-16 17:46               ` Ulrich Drepper
2007-01-16 17:50                 ` Ingo Molnar
2007-01-17  7:50                   ` Pierre Peiffer

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=45A73E90.7050805@bull.net \
    --to=pierre.peiffer@bull.net \
    --cc=akpm@osdl.org \
    --cc=drepper@redhat.com \
    --cc=jakub@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --subject='Re: [PATCH 2.6.20-rc4 0/4] futexes functionalities and improvements' \
    /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

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).