public inbox for gcc-prs@sourceware.org
help / color / mirror / Atom feed
* Re: optimization/10774: Problem with ``-march=pentium4 -O''
@ 2003-05-15 3:06 Dara Hazeghi
0 siblings, 0 replies; 6+ messages in thread
From: Dara Hazeghi @ 2003-05-15 3:06 UTC (permalink / raw)
To: nobody; +Cc: gcc-prs
The following reply was made to PR optimization/10774; it has been noted by GNATS.
From: Dara Hazeghi <dhazeghi@yahoo.com>
To: Mikhail Teterin <mi+kde@aldan.algebra.com>
Cc: gcc-gnats@gcc.gnu.org
Subject: Re: optimization/10774: Problem with ``-march=pentium4 -O''
Date: Wed, 14 May 2003 20:02:59 -0700 (PDT)
> = By "works," you mean the testcase exhibits the
> correct behavior?
> =
> = Dara
>
> Yes. I also tested the original problem I had --
> FreeBSD pow(3) was
> producing incorrect results, when the libm was
> compiled for P4, which,
> for me, led to the failure of the lcms (see
> littlecms.com) selftest.
> After recompiling the libm with gcc33, I can get
> through the lcms'
> selftest. I wonder, if the fixes can be backported
> to the 3.2.x branch,
> so the OS vendors can incorporate them without
> jumping to the 3.3 right
> away...
They can be backported. However, there is nobody
maintaining the gcc 3.2 branch, so these fixes will
probably be localized to the various distributors.
Dara
__________________________________
Do you Yahoo!?
The New Yahoo! Search - Faster. Easier. Bingo.
http://search.yahoo.com
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: optimization/10774: Problem with ``-march=pentium4 -O''
@ 2003-05-15 14:26 Mikhail Teterin
0 siblings, 0 replies; 6+ messages in thread
From: Mikhail Teterin @ 2003-05-15 14:26 UTC (permalink / raw)
To: nobody; +Cc: gcc-prs
The following reply was made to PR optimization/10774; it has been noted by GNATS.
From: Mikhail Teterin <mi+kde@aldan.algebra.com>
To: gcc-prs@gcc.gnu.org, das@FreeBSD.org, gcc-bugs@gcc.gnu.org,
mi+gcc@aldan.algebra.com, gcc-gnats@gcc.gnu.org, nobody@gcc.gnu.org
Cc: freebsd-current@FreeBSD.org
Subject: Re: optimization/10774: Problem with ``-march=pentium4 -O''
Date: Thu, 15 May 2003 10:18:10 -0400
http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&pr=10774
This was a duplicate of the earlier
http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&pr=10189
submitted on my behalf (?) by till@f111.hadiko.de...
The duplicate contained a smaller test-case, however, and, indeed,
both test-cases generate correct code with the gcc-3.3
-mi
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: optimization/10774: Problem with ``-march=pentium4 -O''
@ 2003-05-15 13:55 bangerth
0 siblings, 0 replies; 6+ messages in thread
From: bangerth @ 2003-05-15 13:55 UTC (permalink / raw)
To: das, gcc-bugs, gcc-prs, mi+gcc, nobody
Synopsis: Problem with ``-march=pentium4 -O''
State-Changed-From-To: open->closed
State-Changed-By: bangerth
State-Changed-When: Thu May 15 13:55:55 2003
State-Changed-Why:
Fixed
http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=10774
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: optimization/10774: Problem with ``-march=pentium4 -O''
@ 2003-05-15 1:06 Mikhail Teterin
0 siblings, 0 replies; 6+ messages in thread
From: Mikhail Teterin @ 2003-05-15 1:06 UTC (permalink / raw)
To: nobody; +Cc: gcc-prs
The following reply was made to PR optimization/10774; it has been noted by GNATS.
From: Mikhail Teterin <Mikhail.Teterin@murex.com>
To: Dara Hazeghi <dhazeghi@yahoo.com>, das@FreeBSD.org,
mi+gcc@aldan.algebra.com, gcc-gnats@gcc.gnu.org, nobody@gcc.gnu.org
Cc: obrien@FreeBSD.org
Subject: Re: optimization/10774: Problem with ``-march=pentium4 -O''
Date: Wed, 14 May 2003 21:03:42 -0400
On Wednesday 14 May 2003 01:57 pm, Dara Hazeghi wrote:
=
http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-trail&database=gcc&pr=10774
=
= Hello,
=
= -march=pentium4 used to have problems. A good number were fixed in
= gcc 3.3. Can you by any chance check that the problems still occur
= when compiling with gcc 3.3 (I'd help, but don't have access to a
= P-IV!)?
Yes, the
gcc version 3.3 20030509 (prerelease) [FreeBSD]
and, presumably, the just released (3 hours after I finished building
the prerelease) gcc-3.3 works too. I hope, FreeBSD updates its compiler
the soonest -- prior to the 5.1 release... Thanks!
-mi
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: optimization/10774: Problem with ``-march=pentium4 -O''
@ 2003-05-14 18:06 Dara Hazeghi
0 siblings, 0 replies; 6+ messages in thread
From: Dara Hazeghi @ 2003-05-14 18:06 UTC (permalink / raw)
To: nobody; +Cc: gcc-prs
The following reply was made to PR optimization/10774; it has been noted by GNATS.
From: Dara Hazeghi <dhazeghi@yahoo.com>
To: das@FreeBSD.org, mi+gcc@aldan.algebra.com, gcc-gnats@gcc.gnu.org,
nobody@gcc.gnu.org
Cc:
Subject: Re: optimization/10774: Problem with ``-march=pentium4 -O''
Date: Wed, 14 May 2003 10:57:48 -0700
http://gcc.gnu.org/cgi-bin/gnatsweb.pl?cmd=view%20audit-
trail&database=gcc&pr=10774
Hello,
-march=pentium4 used to have problems. A good number were fixed in gcc
3.3. Can you by any chance check that the problems still occur when
compiling with gcc 3.3 (I'd help, but don't have access to a P-IV!)?
Thanks,
Dara
^ permalink raw reply [flat|nested] 6+ messages in thread
* optimization/10774: Problem with ``-march=pentium4 -O''
@ 2003-05-14 0:26 mi+gcc
0 siblings, 0 replies; 6+ messages in thread
From: mi+gcc @ 2003-05-14 0:26 UTC (permalink / raw)
To: gcc-gnats; +Cc: das
>Number: 10774
>Category: optimization
>Synopsis: Problem with ``-march=pentium4 -O''
>Confidential: no
>Severity: serious
>Priority: medium
>Responsible: unassigned
>State: open
>Class: wrong-code
>Submitter-Id: net
>Arrival-Date: Wed May 14 00:26:00 UTC 2003
>Closed-Date:
>Last-Modified:
>Originator: Mikhail Teterin
>Release: gcc version 3.2.2
>Organization:
>Environment:
Reproduceable with:
gcc version 3.2.2 [FreeBSD] 20030205 (release)
gcc version 3.2 20020903 (Red Hat Linux 8.0 3.2-7)
>Description:
The attached file (you can remove the ``#include <stdio.h>'' line if you
insist) gets miscompiled with ``-march=pentium4 -O''. Removing the -O
or specifying another architecture (such as pentium3) fixes the problem.
The expected result is 0.5. But an optimized for pentium4 version
produces 5.29463e-315 (same value on both Linux and FreeBSD).
According to the debugger, the 22-nd line is skipped in the P4 case.
See also the FreeBSD problem reports:
http://www.FreeBSD.org/cgi/query-pr.cgi?pr=bin/43299
http://www.FreeBSD.org/cgi/query-pr.cgi?pr=i386/51210
>How-To-Repeat:
% cc -O1 -march=pentium4 simple_pow.c -Wall && ./a.out
7.000000 ^ 2.100000
7.000000 ^ 2.100000
ax is 5.29463e-315
result is 5.29463e-315
% cc -O1 -march=pentium3 simple_pow.c -Wall && ./a.out
7.000000 ^ 2.100000
7.000000 ^ 2.100000
ax is 0.5
result is 0.5
>Fix:
>Release-Note:
>Audit-Trail:
>Unformatted:
----gnatsweb-attachment----
Content-Type: text/x-csrc; name="simple_pow.c"
Content-Transfer-Encoding: base64
Content-Disposition: attachment; filename="simple_pow.c"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^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2003-05-15 14:26 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2003-05-15 3:06 optimization/10774: Problem with ``-march=pentium4 -O'' Dara Hazeghi
-- strict thread matches above, loose matches on Subject: below --
2003-05-15 14:26 Mikhail Teterin
2003-05-15 13:55 bangerth
2003-05-15 1:06 Mikhail Teterin
2003-05-14 18:06 Dara Hazeghi
2003-05-14 0:26 mi+gcc
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).