public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
From: "Fang, Changpeng" <Changpeng.Fang@amd.com>
To: Mark Mitchell <mark@codesourcery.com>,
	       Christian Borntraeger	<borntraeger@de.ibm.com>
Cc: "gcc-patches@gcc.gnu.org" <gcc-patches@gcc.gnu.org>,
	       "H.J. Lu"	<hjl.tools@gmail.com>,
	       "rguenther@suse.de" <rguenther@suse.de>,
	       "sebpop@gmail.com" <sebpop@gmail.com>,
	       Zdenek Dvorak	<rakdver@kam.mff.cuni.cz>,
	       Maxim Kuvyrkov <maxim@codesourcery.com>
Subject: RE: [PATCH] Enabling Software Prefetching by Default at -O3
Date: Tue, 22 Jun 2010 23:40:00 -0000	[thread overview]
Message-ID: <D4C76825A6780047854A11E93CDE84D02F774E@SAUSEXMBP01.amd.com> (raw)
In-Reply-To: <4C1D2304.5080007@codesourcery.com>

Hi, Mark:

>> It also might be worth to investigate if overriding the parameters per
>> -mtune=XXX results in an overall win for -fprefetch-loop-arrays. We did
>> that on s390 since the default values were not ideal

>Yes, that might be a good idea for i7.

>But, in the meantime, I think we should get a version of the patch that
>turns on prefetching on AMD CPUs with -O3.  There's no reason to demand
>consistency for all CPUs and it clearly benefits the AMD CPUs.
>Changpeng, would you please submit a patch that activates this
>optimization only with tuning for AMD CPUs?

It seems I have a problem in setting prefetching default at -O3 only for AMD CPUs.
When we know it is AMD CPUS after the command line options have been parsed,
we don't know whether -fno-prefetch-loop-arrays has been specified in the command line
or not. If we turn on prefetching at -O3, then we could not explicitly turn it off if we want.

What I want is something like:
 if  (!OPTION_SET_P (flag_prefetch_loop_arrays))
   flag_prefetch_loop_arrays = 1;

An alternative way is parsing -m options before other options.

What do you think?

Thanks,

Changpeng

 

  parent reply	other threads:[~2010-06-22 22:07 UTC|newest]

Thread overview: 31+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-06-15 22:21 Fang, Changpeng
2010-06-15 22:24 ` H.J. Lu
2010-06-15 22:28   ` Fang, Changpeng
2010-06-15 22:44     ` H.J. Lu
2010-06-15 22:47       ` Fang, Changpeng
2010-06-15 23:42         ` Mark Mitchell
2010-06-16  0:30           ` H.J. Lu
2010-06-16  3:43             ` Mark Mitchell
2010-06-18  8:51               ` Christian Borntraeger
2010-06-18 18:00                 ` Fang, Changpeng
2010-06-19  0:30             ` H.J. Lu
2010-06-20  0:11               ` Christian Borntraeger
2010-06-20  2:34                 ` Mark Mitchell
2010-06-20  4:29                   ` Andi Kleen
2010-06-22 23:40                   ` Fang, Changpeng [this message]
2010-06-23  6:21                     ` Christian Borntraeger
2010-06-23  9:26                       ` Richard Guenther
2010-06-23 12:02                         ` Joseph S. Myers
2010-06-23 12:30                         ` Christian Borntraeger
2010-06-23 13:13                           ` Manuel López-Ibáñez
2010-06-24 23:37                   ` Fang, Changpeng
2010-06-25 11:48                     ` Richard Guenther
2010-06-25 20:16                       ` Sebastian Pop
2010-06-26 17:01                         ` Richard Guenther
2010-06-26 17:30                           ` Richard Guenther
2010-06-26 17:38                           ` Jan Hubicka
2010-06-26 18:23                             ` Jan Hubicka
2010-06-25 12:29                     ` Christian Borntraeger
2010-07-02 12:46                       ` [PATCH] Enabling Software Prefetching by Default@-O3 Ulrich Weigand
2010-06-18  9:13 ` [PATCH] Enabling Software Prefetching by Default at -O3 Andreas Krebbel
2010-06-18  9:26 ` Andreas Krebbel

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=D4C76825A6780047854A11E93CDE84D02F774E@SAUSEXMBP01.amd.com \
    --to=changpeng.fang@amd.com \
    --cc=borntraeger@de.ibm.com \
    --cc=gcc-patches@gcc.gnu.org \
    --cc=hjl.tools@gmail.com \
    --cc=mark@codesourcery.com \
    --cc=maxim@codesourcery.com \
    --cc=rakdver@kam.mff.cuni.cz \
    --cc=rguenther@suse.de \
    --cc=sebpop@gmail.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;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).