From: Jan Hubicka <hubicka@ucw.cz>
To: "Martin Liška" <mliska@suse.cz>
Cc: Richard Biener <richard.guenther@gmail.com>,
Xinliang David Li <davidxl@google.com>,
"gcc@gcc.gnu.org" <gcc@gcc.gnu.org>,
Eugene Rozenfeld <Eugene.Rozenfeld@microsoft.com>
Subject: Re: State of AutoFDO in GCC
Date: Fri, 23 Apr 2021 19:20:57 +0200 [thread overview]
Message-ID: <20210423172057.GB82007@kam.mff.cuni.cz> (raw)
In-Reply-To: <deba67cc-f2c1-8573-3b1f-fd7f1bf8b66e@suse.cz>
> Hi.
>
> The current situation is that AutoFDO doesn't work with pretty simple test-cases
> we have in testsuite:
>
> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71672
> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81379
>
> These are ~5 years old and nothing has happened.
>
> I'm pretty sure the current autofdo can't emit a .gcda file format that
> I've changed in the recent years.
I believe that the gcda files consumed by auto-FDO are not really
sharing the normal gcov file format, just the low level i/o.
See read_profile in auto-profile.c.
Honza
>
> >
> > I think if we want to keep the feature it makes sense to provide create_gcov
> > functionality either directly from perf (input data producer) or from gcc
> > (data consumer). Of course I have no idea about its complexity, license
> > or implementation language ...
>
> For me, it's just an i386 feature (maybe aarch64 has perf counters too?), supported
> only by vendor (Intel) and I'm not planning working on that.
> I don't like having a feature that is obviously broken and potential GCC users get
> bad experience every time they try to use it.
>
> Can we at least deprecate the feature for GCC 11? If these is enough interest,
> we can fix it, if not, I would remove it in GCC 13 timeframe.
>
> Thoughts?
> Martin
>
> >
> > Having the tool third-party makes keeping the whole chain working more
> > difficult.
> >
> > Richard.
> >
> >> Thanks,
> >>
> >> David
> >>
> >> On Thu, Apr 22, 2021 at 3:29 PM Jan Hubicka <hubicka@ucw.cz> wrote:
> >>
> >>>> On 4/22/21 9:58 PM, Eugene Rozenfeld via Gcc wrote:
> >>>>> GCC documentation for AutoFDO points to create_gcov tool that converts
> >>> perf.data file into gcov format that can be consumed by gcc with
> >>> -fauto-profile (https://gcc.gnu.org/onlinedocs/gcc/Optimize-Options.html,
> >>> https://gcc.gnu.org/wiki/AutoFDO/Tutorial).
> >>>>>
> >>>>> I noticed that the source code for create_gcov has been deleted from
> >>> https://github.com/google/autofdo on April 7. I asked about that change
> >>> in that repo and got the following reply:
> >>>>>
> >>>>> https://github.com/google/autofdo/pull/107#issuecomment-819108738
> >>>>>
> >>>>> "Actually we didn't use create_gcov and havn't updated create_gcov for
> >>> years, and we also didn't have enough tests to guarantee it works (It was
> >>> gcc-4.8 when we used and verified create_gcov). If you need it, it is
> >>> welcomed to update create_gcov and add it to the respository."
> >>>>>
> >>>>> Does this mean that AutoFDO is currently dead in gcc?
> >>>>
> >>>> Hello.
> >>>>
> >>>> Yes. I know that even basic test cases have been broken for years in the
> >>> GCC.
> >>>> It's new to me that create_gcov was removed.
> >>>>
> >>>> I tend to send patch to GCC that will remove AutoFDO from GCC.
> >>>> I known Bin spent some time working on AutoFDO, has he came up to
> >>> something?
> >>>
> >>> The GCC side of auto-FDO is not that hard. We have most of
> >>> infrastructure in place, but stopping point for me was always difficulty
> >>> to get gcov-tool working. If some maintainer steps up, I think I can
> >>> fix GCC side.
> >>>
> >>> I am bit unsure how important feature it is - we have FDO that works
> >>> quite well for most users but I know there are some users of the LLVM
> >>> implementation and there is potential to tie this with other hardware
> >>> events to asist i.e. if conversion (where one wants to know how well CPU
> >>> predicts the jump rather than just the jump probability) which I always
> >>> found potentially interesting.
> >>>
> >>> Honza
> >>>>
> >>>> Martin
> >>>>
> >>>>>
> >>>>> Thanks,
> >>>>>
> >>>>> Eugene
> >>>>>
> >>>>
> >>>
>
next prev parent reply other threads:[~2021-04-23 17:20 UTC|newest]
Thread overview: 62+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-04-22 19:58 Eugene Rozenfeld
2021-04-22 20:16 ` Martin Liška
2021-04-22 22:29 ` Jan Hubicka
2021-04-23 4:14 ` Xinliang David Li
2021-04-23 7:00 ` Richard Biener
2021-04-23 7:18 ` Martin Liška
2021-04-23 9:32 ` Richard Biener
2021-04-23 16:41 ` Xinliang David Li
2021-04-23 16:54 ` Jan Hubicka
2021-04-23 17:04 ` Xinliang David Li
2021-04-23 17:16 ` Jan Hubicka
2021-04-23 17:27 ` Xinliang David Li
2021-04-23 17:28 ` Xinliang David Li
2021-04-23 19:28 ` Jan Hubicka
2021-04-23 19:58 ` Xinliang David Li
2021-04-25 19:07 ` Jan Hubicka
2021-04-25 23:18 ` Xinliang David Li
2021-04-26 4:22 ` Wei Mi
2021-04-26 15:11 ` Andi Kleen
2021-04-26 16:57 ` Xinliang David Li
2021-04-26 18:00 ` Andi Kleen
2021-04-26 18:05 ` Xinliang David Li
2021-04-26 18:40 ` Hongtao Yu
2021-04-26 19:13 ` Andi Kleen
2021-04-29 5:40 ` Andi Kleen
2021-04-29 14:45 ` 172060045
2021-04-30 21:43 ` Andi Kleen
2021-05-08 11:25 ` 172060045
2021-05-09 16:28 ` Andi Kleen
2021-05-09 17:01 ` Jan Hubicka
2021-05-10 15:36 ` Andi Kleen
2021-05-10 16:55 ` Joseph Myers
2021-05-10 17:21 ` Andi Kleen
2022-07-26 20:12 ` Eugene Rozenfeld
2022-07-26 22:37 ` David Edelsohn
2022-07-27 7:26 ` Jan Hubicka
2022-07-27 18:30 ` [EXTERNAL] " Eugene Rozenfeld
2022-07-27 18:24 ` Eugene Rozenfeld
2022-07-27 1:31 ` Xionghu Luo
2022-07-27 1:41 ` Xionghu Luo
2022-07-27 18:38 ` [EXTERNAL] " Eugene Rozenfeld
2021-05-10 23:46 ` Wei Mi
2021-05-22 1:28 ` [EXTERNAL] " Eugene Rozenfeld
2021-05-22 16:36 ` Wei Mi
2021-05-25 1:39 ` Eugene Rozenfeld
2021-05-25 3:11 ` Wei Mi
2021-05-25 3:33 ` Eugene Rozenfeld
2021-05-25 3:54 ` Wei Mi
2021-05-25 7:01 ` Eugene Rozenfeld
2021-05-25 16:16 ` Wei Mi
2021-05-25 20:49 ` Eugene Rozenfeld
2021-05-26 3:06 ` Wei Mi
2021-05-26 23:39 ` Eugene Rozenfeld
2021-05-27 2:51 ` Wei Mi
2021-06-12 1:14 ` Eugene Rozenfeld
2021-06-14 17:00 ` Wei Mi
2021-04-23 17:20 ` Jan Hubicka [this message]
2021-04-23 16:36 ` Xinliang David Li
2021-04-30 18:48 ` [EXTERNAL] " Eugene Rozenfeld
2021-04-30 21:45 ` Andi Kleen
2021-06-24 21:45 ` Eugene Rozenfeld
2021-04-23 1:46 ` Bin.Cheng
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=20210423172057.GB82007@kam.mff.cuni.cz \
--to=hubicka@ucw.cz \
--cc=Eugene.Rozenfeld@microsoft.com \
--cc=davidxl@google.com \
--cc=gcc@gcc.gnu.org \
--cc=mliska@suse.cz \
--cc=richard.guenther@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).