From: Ramana Radhakrishnan <ramana.gcc@googlemail.com>
To: Wilco Dijkstra <Wilco.Dijkstra@arm.com>
Cc: "bschmidt@redhat.com" <bschmidt@redhat.com>,
"gcc-patches@gcc.gnu.org" <gcc-patches@gcc.gnu.org>,
Ramana Radhakrishnan <Ramana.Radhakrishnan@arm.com>,
nd <nd@arm.com>, Richard Biener <richard.guenther@gmail.com>,
"ebotcazou@adacore.com" <ebotcazou@adacore.com>,
"dje.gcc@gmail.com" <dje.gcc@gmail.com>,
"law@redhat.com" <law@redhat.com>,
"segher@kernel.crashing.org" <segher@kernel.crashing.org>
Subject: Re: Enabling -frename-registers?
Date: Thu, 05 May 2016 07:42:00 -0000 [thread overview]
Message-ID: <CAJA7tRazVJ6NAZ26B5sZshKVbxvPyDG2o3mrUGK1STpx2vdG6A@mail.gmail.com> (raw)
In-Reply-To: <AM3PR08MB00884B492D88A3EC1F636971837B0@AM3PR08MB0088.eurprd08.prod.outlook.com>
On Wed, May 4, 2016 at 4:20 PM, Wilco Dijkstra <Wilco.Dijkstra@arm.com> wrote:
> Bernd Schmidt wrote:
>> On 05/04/2016 03:25 PM, Ramana Radhakrishnan wrote:
>>> On ARM / AArch32 I haven't seen any performance data yet - the one place we are concerned
>>> about the impact is on Thumb2 code size as regrename may end up inadvertently putting more
>>> things in high registers.
>>
>> In theory at least arm_preferred_rename_class is designed to make the opposite happen. Bernd
>
> I do not see that working unfortunately - Thumb-2 codesize increases by a few percent even with -Os.
> This is primarily due to replacing a low register with IP, which often changes a 16-bit instruction like:
>
> movs r2, #8
>
> into a 32-bit one:
>
> mov ip, #8
>
> This will also affect other targets with multiple instruction sizes. So I think it should check the
> size of the new instruction patterns and only accept a rename if it is not larger (certainly with -Os).
Can you file a bugzilla entry with a testcase that folks can look at please ?
Ramana
>
next prev parent reply other threads:[~2016-05-05 7:42 UTC|newest]
Thread overview: 44+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-05-04 15:20 Wilco Dijkstra
2016-05-04 19:06 ` Eric Botcazou
2016-05-04 22:55 ` Pat Haugen
2016-05-05 7:42 ` Ramana Radhakrishnan [this message]
2016-05-05 13:00 ` Wilco Dijkstra
2016-05-05 16:24 ` Mike Stump
-- strict thread matches above, loose matches on Subject: below --
2016-05-02 11:12 Uros Bizjak
2016-05-02 11:15 ` Bernd Schmidt
2016-05-02 11:21 ` Uros Bizjak
2016-05-02 11:58 ` Uros Bizjak
2016-05-02 13:20 ` Bernd Schmidt
2016-05-03 5:30 ` Uros Bizjak
2016-04-29 13:02 David Edelsohn
2016-04-29 13:33 ` Bernd Schmidt
2016-04-29 13:42 ` David Edelsohn
2016-04-29 13:44 ` Bernd Schmidt
2016-04-29 13:48 ` David Edelsohn
2016-04-29 14:21 ` Richard Biener
2016-04-30 7:57 ` Andreas Schwab
2016-05-03 21:26 ` David Edelsohn
2016-05-03 22:53 ` Bernd Schmidt
2016-05-03 23:09 ` David Edelsohn
2016-05-04 8:39 ` Richard Biener
2016-05-04 9:05 ` Alan Modra
2016-05-04 9:08 ` Bernd Schmidt
2016-05-04 9:38 ` Alan Modra
2016-05-04 10:03 ` Eric Botcazou
2016-05-04 10:11 ` Bernd Schmidt
2016-05-04 10:26 ` Eric Botcazou
2016-05-04 13:25 ` Ramana Radhakrishnan
2016-05-04 13:37 ` Bernd Schmidt
2016-05-04 13:39 ` Ramana Radhakrishnan
2016-05-04 12:54 ` David Edelsohn
2016-05-04 17:50 ` Segher Boessenkool
2016-04-30 2:14 ` Segher Boessenkool
2016-05-02 13:32 ` Jeff Law
2016-05-11 4:07 ` Jeff Law
[not found] <56AB9759.407@t-online.de>
2016-01-29 17:34 ` Bernd Schmidt
2016-01-29 18:50 ` Richard Biener
2016-04-17 21:24 ` Jeff Law
2016-04-26 12:44 ` Bernd Schmidt
2016-04-27 7:11 ` Eric Botcazou
2016-04-27 9:58 ` Bernd Schmidt
2016-04-27 10:36 ` Eric Botcazou
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=CAJA7tRazVJ6NAZ26B5sZshKVbxvPyDG2o3mrUGK1STpx2vdG6A@mail.gmail.com \
--to=ramana.gcc@googlemail.com \
--cc=Ramana.Radhakrishnan@arm.com \
--cc=Wilco.Dijkstra@arm.com \
--cc=bschmidt@redhat.com \
--cc=dje.gcc@gmail.com \
--cc=ebotcazou@adacore.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=law@redhat.com \
--cc=nd@arm.com \
--cc=richard.guenther@gmail.com \
--cc=segher@kernel.crashing.org \
/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).