From: Andi Kleen <andi@firstfloor.org>
To: Andrew Morton <akpm@linux-foundation.org>
Cc: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>,
Ingo Molnar <mingo@elte.hu>,
linux-kernel@vger.kernel.org, Jason Baron <jbaron@redhat.com>,
Rusty Russell <rusty@rustcorp.com.au>,
Adrian Bunk <bunk@stusta.de>, Andi Kleen <andi@firstfloor.org>,
Christoph Hellwig <hch@infradead.org>
Subject: Re: [patch 02/12] Immediate Values - Architecture Independent Code
Date: Mon, 28 Sep 2009 03:23:37 +0200 [thread overview]
Message-ID: <20090928012337.GC1656@one.firstfloor.org> (raw)
In-Reply-To: <20090924212013.d27226c4.akpm@linux-foundation.org>
On Thu, Sep 24, 2009 at 09:20:13PM -0700, Andrew Morton wrote:
> On Thu, 24 Sep 2009 09:26:28 -0400 Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca> wrote:
>
> > Immediate values are used as read mostly variables that are rarely updated. They
> > use code patching to modify the values inscribed in the instruction stream. It
> > provides a way to save precious cache lines that would otherwise have to be used
> > by these variables.
>
> What a hare-brained concept.
The concept makes a lot of sense. Cache misses are extremly costly
on modern CPUs and when the workload has blown the caches away in user space
it can literally be hundreds or even thousands of cycles to fetch
a data cache line.
Similar optimizations are also quite common in compilers (with
profile feedback) and JITs.
> Do you have any data which indicates that this thing is of any real
> benefit to anyone for anything?
There's a lot of data around that the kernel has very little IPC
due to a lot of cache misses in some workloads. In general this applies
to anything that touches a lot of data in user space and blows
the caches away.
-Andi
--
ak@linux.intel.com -- Speaking for myself only.
next prev parent reply other threads:[~2009-09-28 1:23 UTC|newest]
Thread overview: 34+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-09-24 13:26 [patch 00/12] Immediate Values Mathieu Desnoyers
2009-09-24 13:26 ` [patch 01/12] x86: text_poke_early non static Mathieu Desnoyers
2009-09-24 13:26 ` [patch 02/12] Immediate Values - Architecture Independent Code Mathieu Desnoyers
2009-09-25 4:20 ` Andrew Morton
2009-09-27 23:23 ` Mathieu Desnoyers
2009-09-28 1:23 ` Andi Kleen [this message]
2009-09-28 17:46 ` Andrew Morton
2009-09-28 18:03 ` Arjan van de Ven
2009-09-28 18:40 ` Mathieu Desnoyers
2009-09-28 19:54 ` Andi Kleen
2009-09-28 20:37 ` Arjan van de Ven
2009-09-28 21:32 ` H. Peter Anvin
2009-09-28 22:05 ` Mathieu Desnoyers
2009-09-28 20:11 ` Andi Kleen
2009-09-28 21:16 ` Andrew Morton
2009-09-28 22:01 ` Mathieu Desnoyers
2009-09-24 13:26 ` [patch 03/12] Immediate Values - Kconfig menu in EMBEDDED Mathieu Desnoyers
2009-09-24 13:26 ` [patch 04/12] Immediate Values - x86 Optimization Mathieu Desnoyers
2009-09-24 13:26 ` [patch 05/12] Add text_poke and sync_core to powerpc Mathieu Desnoyers
2009-09-24 13:26 ` [patch 06/12] Immediate Values - Powerpc Optimization Mathieu Desnoyers
2009-09-24 13:26 ` [patch 07/12] Sparc create asm.h Mathieu Desnoyers
2009-09-24 21:10 ` David Miller
2009-09-24 13:26 ` [patch 08/12] sparc64: Optimized immediate value implementation Mathieu Desnoyers
2009-09-24 13:26 ` [patch 09/12] Immediate Values - Documentation Mathieu Desnoyers
2009-09-24 13:26 ` [patch 10/12] Immediate Values Support init Mathieu Desnoyers
2009-09-24 15:33 ` [patch 10.1/12] Immediate values fixes for modules Mathieu Desnoyers
2009-09-24 15:35 ` [patch 10.2/12] Fix Immediate Values x86_64 support old gcc Mathieu Desnoyers
2009-09-24 13:26 ` [patch 11/12] Scheduler Profiling - Use Immediate Values Mathieu Desnoyers
2009-09-24 13:26 ` [patch 12/12] Tracepoints - " Mathieu Desnoyers
2009-09-24 14:51 ` Peter Zijlstra
2009-09-24 15:03 ` Mathieu Desnoyers
2009-09-24 15:06 ` Peter Zijlstra
2009-09-24 16:01 ` [RFC patch] Immediate Values - x86 Optimization NMI and MCE support Mathieu Desnoyers
2009-09-24 21:59 ` Masami Hiramatsu
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=20090928012337.GC1656@one.firstfloor.org \
--to=andi@firstfloor.org \
--cc=akpm@linux-foundation.org \
--cc=bunk@stusta.de \
--cc=hch@infradead.org \
--cc=jbaron@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mathieu.desnoyers@polymtl.ca \
--cc=mingo@elte.hu \
--cc=rusty@rustcorp.com.au \
/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