From: Richard Sandiford <rdsandiford@googlemail.com>
To: Maxim Kuvyrkov <maxim.kuvyrkov@linaro.org>
Cc: Sebastian Pop <sebpop@gmail.com>,
GCC Patches <gcc-patches@gcc.gnu.org>,
Vladimir Makarov <vmakarov@redhat.com>,
Ramana Radhakrishnan <ramana.radhakrishnan@arm.com>,
Marcus Shawcroft <marcus.shawcroft@arm.com>,
Richard Earnshaw <rearnsha@arm.com>
Subject: Re: [PATCH] Account for prologue spills in reg_pressure scheduling
Date: Mon, 20 Oct 2014 21:21:00 -0000 [thread overview]
Message-ID: <87wq7u30g6.fsf@googlemail.com> (raw)
In-Reply-To: <183D9AE9-ADFE-4E20-A8FF-B323A631FCC4@linaro.org> (Maxim Kuvyrkov's message of "Tue, 21 Oct 2014 09:56:34 +1300")
Maxim Kuvyrkov <maxim.kuvyrkov@linaro.org> writes:
> [Adding ARM maintainers to CC]
>
> On Oct 21, 2014, at 9:44 AM, Sebastian Pop <sebpop@gmail.com> wrote:
>
>> Hi Maxim,
>>
>> Maxim Kuvyrkov wrote:
>>> Thanks, benchmarking results are welcome! AArch64 doesn't use reg_pressure
>>> scheduling by default. Use "-fsched-pressure
>>> --param=sched-pressure-algorithm=2" to enable same thing as on ARM. I would
>>> imagine C++ and Fortran floating-point code to be most affected.
>>
>> On aarch64 I only see perf improvements with your patch: no perf
>> degradations on
>> all the tests that I have run.
>>
>> base0: r216447, -O3
>> base1: r216447, -O3 -fsched-pressure --param=sched-pressure-algorithm=2
>> patch: r216447 + your patch, -O3 -fsched-pressure
>> --param=sched-pressure-algorithm=2
>>
>> patch vs. base1 is only an improvement.
>>
>> base1 vs. base0 has a few good improvements, and some small degradations: your
>> patch improves the perf for one of the degradations to the point it is better
>> now with -fsched-pressure --param=sched-pressure-algorithm=2 than at -O3.
>>
>> Could we turn on "-fsched-pressure --param=sched-pressure-algorithm=2" by
>> default for aarch64?
>
> These are great results, yay!
+1. Thanks for running these tests. If you have time, it'd also be
interesting to try the same thing with --param=sched-pressure-algorithm=1
(which should be equivalent to not having the --param, but better safe
than sorry). Is algorithm 1 or algorithm 2 better for aarch64?
Thanks,
Richard
next prev parent reply other threads:[~2014-10-20 21:17 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-10-20 7:03 Maxim Kuvyrkov
2014-10-20 19:13 ` Sebastian Pop
2014-10-20 19:23 ` Maxim Kuvyrkov
2014-10-20 20:44 ` Sebastian Pop
2014-10-20 20:59 ` Maxim Kuvyrkov
2014-10-20 21:21 ` Sebastian Pop
2014-10-20 21:21 ` Richard Sandiford [this message]
2014-10-20 21:57 ` Ramana Radhakrishnan
2014-10-20 22:27 ` Maxim Kuvyrkov
2014-10-21 0:01 ` Sebastian Pop
2014-10-20 22:13 ` Evandro Menezes
2014-10-21 15:27 ` Vladimir Makarov
2014-10-22 7:45 ` Maxim Kuvyrkov
2014-10-22 12:51 ` Richard Sandiford
2014-10-22 14:47 ` Vladimir Makarov
2014-10-23 3:19 ` Maxim Kuvyrkov
2014-10-23 7:25 ` Richard Sandiford
2014-10-23 7:28 ` Maxim Kuvyrkov
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=87wq7u30g6.fsf@googlemail.com \
--to=rdsandiford@googlemail.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=marcus.shawcroft@arm.com \
--cc=maxim.kuvyrkov@linaro.org \
--cc=ramana.radhakrishnan@arm.com \
--cc=rearnsha@arm.com \
--cc=sebpop@gmail.com \
--cc=vmakarov@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;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).