From: Richard Biener <richard.guenther@gmail.com>
To: Sergei Trofimovich <slyfox@gentoo.org>
Cc: GCC Patches <gcc-patches@gcc.gnu.org>,
Martin Liska <mliska@suse.cz>,
Sergei Trofimovich <siarheit@google.com>
Subject: Re: [PATCH] -fprofile-reproducible: fix option value handling
Date: Tue, 15 Sep 2020 08:40:14 +0200 [thread overview]
Message-ID: <CAFiYyc0qQgCdDTU0i2LdU=ug2QAphTk_MFVha_pMgmPNDJ=GnQ@mail.gmail.com> (raw)
In-Reply-To: <20200914235703.4f247e61@sf>
On Tue, Sep 15, 2020 at 12:57 AM Sergei Trofimovich <slyfox@gentoo.org> wrote:
>
> On Mon, 14 Sep 2020 09:34:08 +0200
> Richard Biener <richard.guenther@gmail.com> wrote:
>
> > On Fri, Sep 11, 2020 at 11:56 PM Sergei Trofimovich via Gcc-patches
> > <gcc-patches@gcc.gnu.org> wrote:
> > >
> > > From: Sergei Trofimovich <siarheit@google.com>
> > >
> > > Before the change option handling did not accept an argument:
> > > xgcc: error: unknown profile reproducibility method '=serial'
> > > xgcc: note: valid arguments to '-fprofile-reproducible' are: multithreaded parallel-runs serial; did you mean 'serial'?
> > >
> > > The change also includes trailing '=' as part of option prefix.
> >
> > Does it still work without an option then?
>
> '-fprofile-reproducible' seems to be unacceptable value.
>
> Initially when I sent the patch I though there was no way to pass
> the option to gcc. But now I understand how to do it (case 4):
>
> Before:
> 1 $ gcc-11.0.0 -c -fprofile-reproducible a.c -o a
> gcc-11.0.0: error: missing argument to '-fprofile-reproducible'
> 2 $ gcc-11.0.0 -c -fprofile-reproducible= a.c -o a
> gcc-11.0.0: error: unknown profile reproducibility method '='
> gcc-11.0.0: note: valid arguments to '-fprofile-reproducible' are: multithreaded parallel-runs serial
> 3 $ gcc-11.0.0 -c -fprofile-reproducible=serial a.c -o a
> gcc-11.0.0: error: unknown profile reproducibility method '=serial'
> gcc-11.0.0: note: valid arguments to '-fprofile-reproducible' are: multithreaded parallel-runs serial; did you mean 'serial'?
> 4 $ gcc-11.0.0 -c -fprofile-reproducibleserial a.c -o a
> # ok
>
> Note: case 4 was a way to pass the option.
>
> After:
> 1 $ ./xgcc -B. -c -fprofile-reproducible a.c -o a.o
> xgcc: error: unrecognized command-line option '-fprofile-reproducible'; did you mean '-fprofile-reproducible='?
> 2 $ ./xgcc -B. -c -fprofile-reproducible= a.c -o a.o
> xgcc: error: missing argument to '-fprofile-reproducible='
> 3 $ ./xgcc -B. -c -fprofile-reproducible=serial a.c -o a.o
> # ok
> 4 $ ./xgcc -B. -c -fprofile-reproducibleserial a.c -o a.o
> xgcc: error: unrecognized command-line option '-fprofile-reproducibleserial'; did you mean '-fprofile-reproducible=serial'?
>
> Note: two problems here:
> a) case 2 got worse diagnostic
> b) case 4 broke something that worked before
>
> I'll look at "a)" to check if it can be easily fixed. Is "b)" worth handling as well?
> I'll need a hint or example how to handle an alias like that.
I think you want sth like
fdiagnostics-color
Common Alias(fdiagnostics-color=,always,never)
;
fdiagnostics-color=
Driver Common Joined RejectNegative Var(flag_diagnostics_show_color)
Enum(diagnostic_color_rule) Init(DIAGNOSTICS_COLOR_NO)
-fdiagnostics-color=[never|always|auto] Colorize diagnostics.
thus have -fprofile-reproducible an alias to
fprofile-reproducible=,<default>, (not sure if it works to omit
the -fno-profile-reproducible case). Note I don't know what <default> should be
(parallel-runs or multithreaded) but since 'multithreaded' isn't even
documented (eh), that should
be probably the default.
Maybe Martin has sth to say here. The question is what
-fprofile-reproducible did up until now...
Richard.
> > OK if so.
> >
> > > gcc/ChangeLog:
> > >
> > > * common.opt: Fix handling of '-fprofile-reproducible' option.
> > > ---
> > > gcc/common.opt | 2 +-
> > > 1 file changed, 1 insertion(+), 1 deletion(-)
> > >
> > > diff --git a/gcc/common.opt b/gcc/common.opt
> > > index dd68c61ae1d..84bf521128d 100644
> > > --- a/gcc/common.opt
> > > +++ b/gcc/common.opt
> > > @@ -2228,7 +2228,7 @@ Enum(profile_reproducibility) String(parallel-runs) Value(PROFILE_REPRODUCIBILIT
> > > EnumValue
> > > Enum(profile_reproducibility) String(multithreaded) Value(PROFILE_REPRODUCIBILITY_MULTITHREADED)
> > >
> > > -fprofile-reproducible
> > > +fprofile-reproducible=
> > > Common Joined RejectNegative Var(flag_profile_reproducible) Enum(profile_reproducibility) Init(PROFILE_REPRODUCIBILITY_SERIAL)
> > > -fprofile-reproducible=[serial|parallel-runs|multithreaded] Control level of reproducibility of profile gathered by -fprofile-generate.
> > >
> > > --
> > > 2.28.0
> > >
>
>
> --
>
> Sergei
prev parent reply other threads:[~2020-09-15 6:40 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-09-11 21:55 Sergei Trofimovich
2020-09-14 7:34 ` Richard Biener
2020-09-14 22:57 ` Sergei Trofimovich
2020-09-15 6:40 ` Richard Biener [this message]
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='CAFiYyc0qQgCdDTU0i2LdU=ug2QAphTk_MFVha_pMgmPNDJ=GnQ@mail.gmail.com' \
--to=richard.guenther@gmail.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=mliska@suse.cz \
--cc=siarheit@google.com \
--cc=slyfox@gentoo.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).