public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug target/63826] New: GCC produces haswell instructions that darwin14 assembler does not like
@ 2014-11-11 20:10 fxcoudert at gcc dot gnu.org
  2014-11-11 20:11 ` [Bug target/63826] " fxcoudert at gcc dot gnu.org
                   ` (2 more replies)
  0 siblings, 3 replies; 4+ messages in thread
From: fxcoudert at gcc dot gnu.org @ 2014-11-11 20:10 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63826

            Bug ID: 63826
           Summary: GCC produces haswell instructions that darwin14
                    assembler does not like
           Product: gcc
           Version: 5.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: target
          Assignee: unassigned at gcc dot gnu.org
          Reporter: fxcoudert at gcc dot gnu.org

gcc.target/i386/pr57275.c currently fails on my laptop, which is a
x86_64-apple-darwin14 with quad-core Intel Core i7 2.3 GHz and Xcode 6.1.

The minimal reproducer is this:

void foo (float b[1024])
{
  for (int i = 0; i < 1024; i++)
    b[i] = i;
}

Compiled with -march=haswell (or, in my case, -march=native) I get errors from
the assembler about instructions it does not like:

$ ./bin/gcc pr57275.c -march=haswell -save-temps
pr57275.s:18:no such instruction: `vxorps %xmm0, %xmm0,%xmm0'
pr57275.s:19:no such instruction: `vcvtsi2ss -4(%rbp), %xmm0,%xmm0'
pr57275.s:20:no such instruction: `vmovss %xmm0, (%rax)'

The list of offending instructions depends on optimization level and whether I
compile with -m32 or -m64 (the default):

At -O0, -O1 or -O2: vcvtsi2ss vmovss vxorps
At -Os: vcvtsi2ss vmovss
At -O3: vcvtdq2ps vcvtsi2ss vmovaps vmovd vmovdqa vmovss vpaddd vpbroadcastd
vxorps vzeroupper
At -O3 -m32: vcvtdq2ps vmovaps vmovd vmovdqa vpaddd vpbroadcastd vzeroupper


^ permalink raw reply	[flat|nested] 4+ messages in thread

* [Bug target/63826] GCC produces haswell instructions that darwin14 assembler does not like
  2014-11-11 20:10 [Bug target/63826] New: GCC produces haswell instructions that darwin14 assembler does not like fxcoudert at gcc dot gnu.org
@ 2014-11-11 20:11 ` fxcoudert at gcc dot gnu.org
  2014-11-11 20:23 ` dominiq at lps dot ens.fr
  2020-09-08 12:08 ` fxcoudert at gcc dot gnu.org
  2 siblings, 0 replies; 4+ messages in thread
From: fxcoudert at gcc dot gnu.org @ 2014-11-11 20:11 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63826

Francois-Xavier Coudert <fxcoudert at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Target|                            |x86_64-apple-darwin14
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2014-11-11
                 CC|                            |iains at gcc dot gnu.org,
                   |                            |mikestump at comcast dot net
               Host|                            |x86_64-apple-darwin14
     Ever confirmed|0                           |1
              Build|                            |x86_64-apple-darwin14


^ permalink raw reply	[flat|nested] 4+ messages in thread

* [Bug target/63826] GCC produces haswell instructions that darwin14 assembler does not like
  2014-11-11 20:10 [Bug target/63826] New: GCC produces haswell instructions that darwin14 assembler does not like fxcoudert at gcc dot gnu.org
  2014-11-11 20:11 ` [Bug target/63826] " fxcoudert at gcc dot gnu.org
@ 2014-11-11 20:23 ` dominiq at lps dot ens.fr
  2020-09-08 12:08 ` fxcoudert at gcc dot gnu.org
  2 siblings, 0 replies; 4+ messages in thread
From: dominiq at lps dot ens.fr @ 2014-11-11 20:23 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63826

--- Comment #1 from Dominique d'Humieres <dominiq at lps dot ens.fr> ---
I am sure it is a duplicate. You can compile your code with -Wa,-q.


^ permalink raw reply	[flat|nested] 4+ messages in thread

* [Bug target/63826] GCC produces haswell instructions that darwin14 assembler does not like
  2014-11-11 20:10 [Bug target/63826] New: GCC produces haswell instructions that darwin14 assembler does not like fxcoudert at gcc dot gnu.org
  2014-11-11 20:11 ` [Bug target/63826] " fxcoudert at gcc dot gnu.org
  2014-11-11 20:23 ` dominiq at lps dot ens.fr
@ 2020-09-08 12:08 ` fxcoudert at gcc dot gnu.org
  2 siblings, 0 replies; 4+ messages in thread
From: fxcoudert at gcc dot gnu.org @ 2020-09-08 12:08 UTC (permalink / raw)
  To: gcc-bugs

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63826

Francois-Xavier Coudert <fxcoudert at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|---                         |WORKSFORME

--- Comment #2 from Francois-Xavier Coudert <fxcoudert at gcc dot gnu.org> ---
Old, not happening anymore.

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2020-09-08 12:08 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2014-11-11 20:10 [Bug target/63826] New: GCC produces haswell instructions that darwin14 assembler does not like fxcoudert at gcc dot gnu.org
2014-11-11 20:11 ` [Bug target/63826] " fxcoudert at gcc dot gnu.org
2014-11-11 20:23 ` dominiq at lps dot ens.fr
2020-09-08 12:08 ` fxcoudert at gcc dot gnu.org

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).