From: Vladimir Makarov <vmakarov@redhat.com>
To: Jeff Law <law@redhat.com>
Cc: Dave Korn <dave.korn.cygwin@googlemail.com>,
Albert Cohen <Albert.Cohen@inria.fr>,
reply@meinersbur.de, gcc@gcc.gnu.org,
Sid Touati <Sid.Touati@inria.fr>,
Frederic Brault <frederic.brault@inria.fr>
Subject: Re: Register Pressure in Instruction Level Parallelism
Date: Thu, 02 Jul 2009 21:53:00 -0000 [thread overview]
Message-ID: <4A4D2C36.5030202@redhat.com> (raw)
In-Reply-To: <4A4AFDB9.1090402@redhat.com>
Jeff Law wrote:
> Vladimir Makarov wrote:
>> Dave Korn wrote:
>>> In a brief exchange I had with Michael off-list, we discussed
>>> that. I
>>> observed that of the things that reload does,
>>> constraint-satisfaction/insn-variant-selection is its primary
>>> purpose, and
>>> spill/reload code generation is something it often does suboptimally
>>> (and
>>> secondary reloads even worse). If a clever pass running before
>>> reload could
>>> insert explicit spill/reload code at well-chosen places (bearing in
>>> mind
>>> class-based register pressure), it could relieve reload of the
>>> necessity to
>>> generate its own spill code most of the time, and let it just do
>>> what it does
>>> best.
>> IRA actually already inserts spill code in most important places (on
>> loop borders). Besides loop regions, IRA could be extended to other
>> regions (and even bb parts to relief pressure inside the blocks). I
>> am going to work on it to evaluate how much it could give.
> I've already got some code to do this -- I've pondered more than once
> pushing it through independently of the other allocation/reload work
> I'm doing. I could probably be convinced to put the block local
> allocation/spilling on hold to focus on benchmarking and tuning my
> bits to generate spill code.
That is great. I look forward to see the code.
next prev parent reply other threads:[~2009-07-02 21:53 UTC|newest]
Thread overview: 20+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-06-28 12:37 Michael Kruse
2009-06-28 19:06 ` Dave Korn
2009-06-28 22:53 ` Albert Cohen
2009-06-28 23:56 ` Dave Korn
2009-06-29 6:13 ` Albert Cohen
2009-06-29 18:04 ` Vladimir Makarov
2009-07-01 6:09 ` Jeff Law
2009-07-02 21:53 ` Vladimir Makarov [this message]
2009-07-01 5:58 ` Jeff Law
2009-06-29 0:30 ` Michael Kruse
[not found] ` <303e1d290906281549t4ebfce81m5152069742fa9a1@mail.gmail.com>
2009-06-29 6:28 ` Fwd: " Kenneth Zadeck
2009-06-28 23:01 ` Michael Kruse
2009-06-29 0:03 ` Dave Korn
2009-06-29 17:05 ` Vladimir Makarov
2009-06-29 16:27 ` Vladimir Makarov
2009-06-29 20:03 ` Michael Kruse
2009-06-29 20:40 ` Vladimir Makarov
2009-06-29 21:04 ` Michael Kruse
2009-06-30 15:01 ` Albert Cohen
2009-07-01 6:06 ` Jeff Law
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=4A4D2C36.5030202@redhat.com \
--to=vmakarov@redhat.com \
--cc=Albert.Cohen@inria.fr \
--cc=Sid.Touati@inria.fr \
--cc=dave.korn.cygwin@googlemail.com \
--cc=frederic.brault@inria.fr \
--cc=gcc@gcc.gnu.org \
--cc=law@redhat.com \
--cc=reply@meinersbur.de \
/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).