From: Ingo Molnar <mingo@elte.hu>
To: Andrew Morton <akpm@osdl.org>
Cc: rostedt@goodmis.org, rlrevell@joe-job.com, linux-kernel@vger.kernel.org
Subject: Re: [patch 0/3] j_state_lock, j_list_lock, remove-bitlocks
Date: Wed, 16 Mar 2005 11:12:09 +0100 [thread overview]
Message-ID: <20050316101209.GA16893@elte.hu> (raw)
In-Reply-To: <20050316020408.434cc620.akpm@osdl.org>
* Andrew Morton <akpm@osdl.org> wrote:
> Ingo Molnar <mingo@elte.hu> wrote:
> >
> > > There's a little lock ranking diagram in jbd.h which tells us that
> > > these locks nest inside j_list_lock and j_state_lock. So I guess
> > > you'll need to turn those into semaphores.
> >
> > indeed. I did this (see the three followup patches, against BK-curr),
> > and it builds/boots/works just fine on an ext3 box. Do we want to try
> > this in -mm?
>
> ooh, I'd rather not. I spent an intense three days removing all the
> sleeping locks from ext3 (and three months debugging the result).
> Ended up gaining 1000% on 16-way.
>
> Putting them back in will really hurt the SMP performance.
ah. Yeah. Sniff.
if we gain 1000% on a 16-way then there's something really wrong about
semaphores (or scheduling) though. A semaphore is almost a spinlock, in
the uncontended case - and even under contention we really (should) just
spend the cycles that we'd spend spinning. There will be some
intermediate contention level where semaphores hurt, but 1000% sounds
truly excessive.
Ingo
next prev parent reply other threads:[~2005-03-16 10:13 UTC|newest]
Thread overview: 125+ messages / expand[flat|nested] mbox.gz Atom feed top
2005-02-04 10:03 [patch] Real-Time Preemption, -RT-2.6.11-rc3-V0.7.38-01 Ingo Molnar
2005-02-04 15:19 ` Kevin Hilman
2005-02-04 17:30 ` Ingo Molnar
2005-02-04 18:19 ` Tom Rini
2005-02-07 9:03 ` Ingo Molnar
2005-02-07 14:35 ` Tom Rini
2005-02-08 8:27 ` Ingo Molnar
2005-02-06 4:19 ` Valdis.Kletnieks
2005-02-07 9:21 ` Ingo Molnar
2005-02-07 15:08 ` Real-Time Preemption and UML? Esben Nielsen
2005-02-07 18:35 ` Jeff Dike
2005-02-07 23:14 ` Esben Nielsen
2005-02-08 8:39 ` Ingo Molnar
2005-02-08 18:55 ` Jeff Dike
2005-02-08 21:20 ` Esben Nielsen
2005-02-08 21:44 ` Ingo Molnar
2005-02-08 23:02 ` Esben Nielsen
2005-02-08 7:55 ` [patch] Real-Time Preemption, -RT-2.6.11-rc3-V0.7.38-01 Valdis.Kletnieks
2005-02-08 8:45 ` Ingo Molnar
2005-02-08 10:26 ` Valdis.Kletnieks
2005-02-08 21:58 ` William Weston
2005-02-09 11:51 ` Ingo Molnar
2005-02-10 2:13 ` William Weston
2005-02-10 7:52 ` Ingo Molnar
2005-02-10 20:21 ` George Anzinger
2005-02-10 20:40 ` Ingo Molnar
2005-02-10 21:05 ` George Anzinger
2005-02-11 8:34 ` Ingo Molnar
2005-02-11 9:38 ` Sven Dietrich
2005-02-11 9:42 ` Ingo Molnar
2005-02-11 0:09 ` Sven Dietrich
2005-02-11 6:01 ` George Anzinger
2005-02-11 8:28 ` Ingo Molnar
2005-02-11 9:53 ` Sven Dietrich
2005-02-11 10:04 ` Ingo Molnar
2005-02-11 21:49 ` Steven Rostedt
2005-02-13 12:59 ` Ingo Molnar
2005-02-13 15:11 ` Steven Rostedt
2005-03-03 19:36 ` [patch] Real-Time Preemption, deactivate() scheduling issue Eugeny S. Mints
2005-03-03 22:32 ` Esben Nielsen
2005-03-04 11:56 ` Eugeny S. Mints
2005-03-04 15:45 ` George Anzinger
2005-03-29 8:45 ` Ingo Molnar
2005-02-09 12:48 ` [patch] Real-Time Preemption, -RT-2.6.11-rc3-V0.7.38-01 Stephen Smalley
2005-02-10 2:20 ` William Weston
2005-02-19 5:08 ` Lee Revell
2005-02-19 6:47 ` Lee Revell
2005-02-19 9:00 ` Ingo Molnar
2005-02-19 9:03 ` Ingo Molnar
2005-02-19 20:45 ` Lee Revell
2005-02-20 0:19 ` Lee Revell
2005-03-17 16:33 ` Lee Revell
2005-02-23 2:22 ` Lee Revell
2005-03-10 9:37 ` Steven Rostedt
2005-03-10 9:54 ` Steven Rostedt
2005-03-11 9:57 ` Ingo Molnar
2005-03-11 10:15 ` Steven Rostedt
2005-03-11 10:17 ` Ingo Molnar
2005-03-11 10:24 ` Steven Rostedt
2005-03-11 10:43 ` Andrew Morton
2005-03-11 10:53 ` Steven Rostedt
2005-03-11 14:40 ` Steven Rostedt
2005-03-11 15:08 ` Steven Rostedt
2005-03-11 15:30 ` K.R. Foley
2005-03-11 15:38 ` Ingo Molnar
2005-03-11 16:01 ` Steven Rostedt
2005-03-11 20:39 ` Steven Rostedt
2005-03-11 20:46 ` Lee Revell
2005-03-11 22:06 ` Lee Revell
2005-03-14 7:37 ` Steven Rostedt
2005-03-14 9:33 ` Steven Rostedt
2005-03-14 10:10 ` Steven Rostedt
2005-03-14 15:50 ` Steven Rostedt
2005-03-14 19:02 ` Steven Rostedt
2005-03-15 11:44 ` Steven Rostedt
2005-03-15 12:00 ` Ingo Molnar
2005-03-15 13:07 ` Steven Rostedt
2005-03-15 13:35 ` Ingo Molnar
2005-03-15 13:55 ` Steven Rostedt
2005-03-15 19:12 ` Andrew Morton
2005-03-15 18:05 ` Steven Rostedt
2005-03-15 19:09 ` Lee Revell
2005-03-16 7:50 ` Steven Rostedt
2005-03-16 18:21 ` Lee Revell
2005-03-16 7:31 ` Steven Rostedt
2005-03-16 8:50 ` Ingo Molnar
2005-03-16 9:15 ` Andrew Morton
2005-03-16 9:51 ` [patch 0/3] j_state_lock, j_list_lock, remove-bitlocks Ingo Molnar
2005-03-16 9:53 ` [patch 1/3] j_state_lock -> j_state_sem Ingo Molnar
2005-03-16 9:53 ` [patch 2/3] j_list_lock -> j_list_sem Ingo Molnar
2005-03-16 9:57 ` [patch 3/3] remove bitlocks Ingo Molnar
2005-03-16 10:04 ` [patch 0/3] j_state_lock, j_list_lock, remove-bitlocks Andrew Morton
2005-03-16 10:12 ` Ingo Molnar [this message]
2005-03-16 10:23 ` Steven Rostedt
2005-03-16 10:26 ` Ingo Molnar
2005-03-16 10:26 ` Andrew Morton
2005-03-16 10:29 ` Ingo Molnar
2005-03-16 10:41 ` Andrew Morton
2005-03-16 10:34 ` Arjan van de Ven
2005-03-16 10:19 ` Ingo Molnar
2005-03-16 10:40 ` Andrew Morton
2005-03-16 10:51 ` Ingo Molnar
2005-03-16 11:05 ` Steven Rostedt
2005-03-16 11:19 ` Andrew Morton
2005-03-16 14:04 ` Steven Rostedt
2005-03-16 16:47 ` Steven Rostedt
2005-03-16 17:47 ` Steven Rostedt
2005-03-16 19:20 ` Lee Revell
2005-03-17 7:15 ` Steven Rostedt
2005-03-17 15:41 ` Lee Revell
2005-03-17 16:23 ` Steven Rostedt
2005-03-17 16:36 ` Lee Revell
2005-03-18 6:58 ` Steven Rostedt
2005-03-18 18:19 ` Lee Revell
2005-03-16 21:15 ` Andrew Morton
2005-03-17 9:21 ` Steven Rostedt
2005-03-18 9:23 ` [PATCH] remove lame schedule in journal inverted_lock (was: Re: [patch 0/3] j_state_lock, j_list_lock, remove-bitlocks) Steven Rostedt
2005-03-18 9:32 ` Andrew Morton
2005-03-18 10:38 ` Steven Rostedt
2005-03-18 11:07 ` Andrew Morton
2005-03-18 12:10 ` Steven Rostedt
2005-03-17 9:58 ` [patch 0/3] j_state_lock, j_list_lock, remove-bitlocks Steven Rostedt
2005-03-11 9:28 ` [patch] Real-Time Preemption, -RT-2.6.11-final-V0.7.40-00 Ingo Molnar
2005-03-11 12:10 ` Andrew Walrond
2005-03-14 20:19 ` Tom Rini
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=20050316101209.GA16893@elte.hu \
--to=mingo@elte.hu \
--cc=akpm@osdl.org \
--cc=linux-kernel@vger.kernel.org \
--cc=rlrevell@joe-job.com \
--cc=rostedt@goodmis.org \
/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