From: Andi Kleen <andi@firstfloor.org>
To: Ingo Molnar <mingo@elte.hu>
Cc: Andi Kleen <andi@firstfloor.org>,
Peter Zijlstra <peterz@infradead.org>,
Jason Baron <jbaron@redhat.com>,
linux-kernel@vger.kernel.org, mathieu.desnoyers@polymtl.ca,
hpa@zytor.com, tglx@linutronix.de, rostedt@goodmis.org,
roland@redhat.com, rth@redhat.com, mhiramat@redhat.com,
fweisbec@gmail.com, avi@redhat.com, davem@davemloft.net,
vgoyal@redhat.com, sam@ravnborg.org,
Linus Torvalds <torvalds@linux-foundation.org>,
Arjan van de Ven <arjan@infradead.org>
Subject: Re: [PATCH 03/13] jump label v9: x86 support
Date: Thu, 10 Jun 2010 18:24:53 +0200 [thread overview]
Message-ID: <20100610162453.GF19561@basil.fritz.box> (raw)
In-Reply-To: <20100610153742.GA9118@elte.hu>
On Thu, Jun 10, 2010 at 05:37:42PM +0200, Ingo Molnar wrote:
>
> > [...] It costs you in some benchmarks. [...]
>
> Microbenchmarks mostly, see below.
I didn't make these decisions, but I assume who made them had good reasons
and enough data on larger benchmarks too.
> > A much better to get smaller kernel images is to do more __cold annotations
> > for slow paths. Newer gcc will then simply only do -Os for these functions.
>
> That's an opt-in method and we cannot reach the kinds of 30% code size
> reductions that -Os can achieve. Most code in the kernel is not cache-hot,
> even on microbenchmarks.
Maybe, maybe not. But yes it can be approached from both ways.
Personally I would prefer to simply write less bloated code to get
code reductions. Simpler code is often faster too.
>
> A much better model would be to actively mark hot codepaths with a __hot
> attribute instead. Then the code size difference can be considered on a case
> by case basis.
Yes that works too for those who still use -Os.
e.g. marking the scheduler and a few mm hot paths this way would certain make sense.
>
> And where GCC produces indefensibly crap code there GCC needs to be fixed.
> Crap code often increases size so the fix would increase the efficiency of
> -Os.
In some cases agreed, but common cases it's really: you asked for the smallest
you got it, even if it's slow. It's not -Odwim.
One standard example here is a division by constant. The shortest way is
using DIVI/IDIV if it's not 2^n and small enough, but it's really quite slow
in hardware. If you spend a few more bytes you can do much better for a wide
range of constants.
Most likely we would need a new -O flag to avoid such cases.
BTW I experimented with marking a few common cases like this (e.g. time unit
conversion) hot, but gcc currently has trouble with __hot on inlines. So you
would always need to mark the caller.
-Andi
--
ak@linux.intel.com -- Speaking for myself only.
next prev parent reply other threads:[~2010-06-10 16:24 UTC|newest]
Thread overview: 54+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-06-09 21:38 [PATCH 00/13] jump label v9 Jason Baron
2010-06-09 21:38 ` [PATCH 01/13] jump label v9: notifier atomic call chain notrace Jason Baron
2010-06-09 21:58 ` Frederic Weisbecker
2010-06-10 15:34 ` Jason Baron
2010-06-09 21:38 ` [PATCH 02/13] jump label v9: base patch Jason Baron
2010-06-09 22:35 ` Frederic Weisbecker
2010-06-10 15:44 ` Jason Baron
2010-06-10 16:22 ` Ingo Molnar
2010-06-10 17:11 ` Frederic Weisbecker
2010-06-09 22:36 ` Frederic Weisbecker
2010-06-10 12:06 ` Peter Zijlstra
2010-06-10 12:18 ` Peter Zijlstra
2010-06-09 21:39 ` [PATCH 03/13] jump label v9: x86 support Jason Baron
2010-06-10 12:12 ` Peter Zijlstra
2010-06-10 12:14 ` Ingo Molnar
2010-06-10 13:26 ` Andi Kleen
2010-06-10 14:12 ` Peter Zijlstra
2010-06-10 14:28 ` Andi Kleen
2010-06-10 15:37 ` Ingo Molnar
2010-06-10 16:24 ` Andi Kleen [this message]
2010-06-11 8:12 ` Ingo Molnar
2010-06-11 8:30 ` Andi Kleen
2010-06-10 16:29 ` Steven Rostedt
2010-06-10 15:04 ` Jason Baron
2010-06-10 16:13 ` Mathieu Desnoyers
2010-06-11 0:52 ` Jason Baron
2010-06-11 6:18 ` H. Peter Anvin
2010-06-11 7:58 ` Ingo Molnar
2010-06-10 12:15 ` Peter Zijlstra
2010-06-10 12:33 ` Peter Zijlstra
2010-06-09 21:39 ` [PATCH 04/13] jump label v9: tracepoint support Jason Baron
2010-06-10 12:22 ` Peter Zijlstra
2010-06-09 21:39 ` [PATCH 05/13] jump label v9: add module support Jason Baron
2010-06-09 21:39 ` [PATCH 06/13] jump label v9: move ftrace_dyn_arch_init to common code Jason Baron
2010-06-19 3:24 ` Steven Rostedt
2010-06-09 21:39 ` [PATCH 07/13] jump label v9: sort jump table at build-time Jason Baron
2010-06-09 21:39 ` [PATCH 08/13] jump label v9: initialize workqueue tracepoints *before* they are registered Jason Baron
2010-06-09 21:39 ` [PATCH 09/13] jump label v9: jump_label_text_reserved() to reserve our jump points Jason Baron
2010-06-09 21:39 ` [PATCH 10/13] jump label v9: convert jump label to use a key Jason Baron
2010-06-10 12:38 ` Peter Zijlstra
2010-06-10 12:43 ` Peter Zijlstra
2010-06-10 13:57 ` Jason Baron
2010-06-10 14:46 ` Peter Zijlstra
2010-06-09 21:39 ` [PATCH 11/13] jump label v9: convert dynamic debug to use jump labels Jason Baron
2010-06-09 21:39 ` [PATCH 12/13] jump label v9: sparc64 add jump_label support Jason Baron
2010-06-09 21:39 ` [PATCH 13/13] jump label v9: add docs Jason Baron
2010-06-10 12:49 ` Peter Zijlstra
2010-06-15 3:47 ` [PATCH 00/13] jump label v9 David Miller
2010-06-15 14:28 ` Jason Baron
2010-06-15 15:44 ` Mathieu Desnoyers
2010-06-18 3:45 ` Tony Breeds
2010-06-18 15:18 ` Mathieu Desnoyers
2010-06-15 17:13 ` David Miller
2010-06-15 17:28 ` H. Peter Anvin
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=20100610162453.GF19561@basil.fritz.box \
--to=andi@firstfloor.org \
--cc=arjan@infradead.org \
--cc=avi@redhat.com \
--cc=davem@davemloft.net \
--cc=fweisbec@gmail.com \
--cc=hpa@zytor.com \
--cc=jbaron@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mathieu.desnoyers@polymtl.ca \
--cc=mhiramat@redhat.com \
--cc=mingo@elte.hu \
--cc=peterz@infradead.org \
--cc=roland@redhat.com \
--cc=rostedt@goodmis.org \
--cc=rth@redhat.com \
--cc=sam@ravnborg.org \
--cc=tglx@linutronix.de \
--cc=torvalds@linux-foundation.org \
--cc=vgoyal@redhat.com \
/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
Powered by JetHome