[Bug target/46768] New: [4.6 Regression] FAIL: gcc.target/i386/pr37434-[24].c

[Bug target/46768] New: [4.6 Regression] FAIL: gcc.target/i386/pr37434-[24].c

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46768

           Summary: [4.6 Regression] FAIL: gcc.target/i386/pr37434-[24].c
           Product: gcc
           Version: 4.6.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: target
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: hjl.tools@gmail.com
                CC: mkuvyrkov@gmail.com


On Linux/ia32, revision 167375 gave:

FAIL: gcc.target/i386/pr37434-2.c scan-assembler pinsrw
FAIL: gcc.target/i386/pr37434-4.c scan-assembler pinsrb

[Bug target/46768] [4.6 Regression] FAIL: gcc.target/i386/pr37434-[24].c

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46768

--- Comment #1 from H.J. Lu <hjl.tools at gmail dot com> 2010-12-02 19:11:26 UTC ---
On Intel64, I got

FAIL: gcc.target/i386/pr34256.c scan-assembler-times mov 2
FAIL: gcc.target/i386/pr37434-2.c scan-assembler pinsrw
FAIL: gcc.target/i386/pr37434-2.c scan-assembler pinsrw
FAIL: gcc.target/i386/pr37434-4.c scan-assembler pinsrb
FAIL: gcc.target/i386/pr37434-4.c scan-assembler pinsrb
FAIL: gcc.target/i386/sse2-init-v2di-2.c scan-assembler movq

[Bug target/46768] [4.6 Regression] FAIL: gcc.target/i386/pr37434-[24].c

[hjl at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46768

--- Comment #2 from hjl at gcc dot gnu.org <hjl at gcc dot gnu.org> 2010-12-02 22:50:49 UTC ---
Author: hjl
Date: Thu Dec  2 22:50:44 2010
New Revision: 167398

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=167398
Log:
Turn on X86_TUNE_INTER_UNIT_MOVES for Core 2 and Core i7.

2010-12-02  H.J. Lu  <hongjiu.lu@intel.com>

    PR target/46768
    * config/i386/i386.c (initial_ix86_tune_features): Turn on
    X86_TUNE_INTER_UNIT_MOVES for Core 2 and Core i7.

Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/config/i386/i386.c

[Bug target/46768] [4.6 Regression] FAIL: gcc.target/i386/pr37434-[24].c

[ubizjak at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46768

Uros Bizjak <ubizjak at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |howarth at nitro dot
                   |                            |med.uc.edu

--- Comment #3 from Uros Bizjak <ubizjak at gmail dot com> 2010-12-02 22:54:17 UTC ---
*** Bug 46767 has been marked as a duplicate of this bug. ***

[Bug target/46768] [4.6 Regression] FAIL: gcc.target/i386/pr37434-[24].c

[howarth at nitro dot med.uc.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46768

--- Comment #4 from Jack Howarth <howarth at nitro dot med.uc.edu> 2010-12-03 04:25:53 UTC ---
On x86_64-apple-darwin10 at r167400, we are still failing...

FAIL: gcc.target/i386/sse2-init-v2di-2.c scan-assembler movq

at -m64.

[Bug target/46768] [4.6 Regression] FAIL: gcc.target/i386/pr37434-[24].c

[mkuvyrkov at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46768

Maxim Kuvyrkov <mkuvyrkov at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |mkuvyrkov at gcc dot
                   |                            |gnu.org

--- Comment #5 from Maxim Kuvyrkov <mkuvyrkov at gcc dot gnu.org> 2010-12-03 11:14:05 UTC ---
(In reply to comment #4)
> On x86_64-apple-darwin10 at r167400, we are still failing...
> 
> FAIL: gcc.target/i386/sse2-init-v2di-2.c scan-assembler movq
> 
> at -m64.

IIUC, this only occurs on x86_64-darwin10 target.

Jack,

If you are interested in fixing this, could you post before and after assembler
dumps for x86_64-darwin10 compiler for the testcase.  

Thank you.

[Bug target/46768] [4.6 Regression] FAIL: gcc.target/i386/pr37434-[24].c

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46768

Richard Guenther <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |4.6.0

[Bug target/46768] [4.6 Regression] FAIL: gcc.target/i386/pr37434-[24].c

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46768

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |RESOLVED
         Resolution|                            |FIXED

--- Comment #6 from H.J. Lu <hjl.tools at gmail dot com> 2010-12-03 14:34:35 UTC ---
(In reply to comment #4)
> On x86_64-apple-darwin10 at r167400, we are still failing...
> 
> FAIL: gcc.target/i386/sse2-init-v2di-2.c scan-assembler movq
> 
> at -m64.

Works for me with a x86_64-apple-darwin10 cross compiler with revision 167416:

    .globl _test
_test:
LFB516:
    movq    %rdi, %xmm0
    ret
LFE516:
    .section __TEXT,__eh_frame,coale

[Bug target/46768] [4.6 Regression] FAIL: gcc.target/i386/pr37434-[24].c

[howarth at nitro dot med.uc.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46768

--- Comment #7 from Jack Howarth <howarth at nitro dot med.uc.edu> 2010-12-03 14:49:40 UTC ---
Created attachment 22615
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=22615
assembly file from gcc.target/i386/sse2-init-v2di-2.c on x86_64-apple-darwin10
at r167372

Generated with...

/Users/howarth/darwin_objdir/gcc/xgcc -B/Users/howarth/darwin_objdir/gcc/
/Users/howarth/gcc/gcc/testsuite/gcc.target/i386/sse2-init-v2di-2.c -O2 -msse4
-march=core2 -S -m64 -o sse2-init-v2di-2.s

[Bug target/46768] [4.6 Regression] FAIL: gcc.target/i386/pr37434-[24].c

[howarth at nitro dot med.uc.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46768

--- Comment #8 from Jack Howarth <howarth at nitro dot med.uc.edu> 2010-12-03 14:50:21 UTC ---
Created attachment 22616
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=22616
assembly file from gcc.target/i386/sse2-init-v2di-2.c on x86_64-apple-darwin10
at r167375

[Bug target/46768] [4.6 Regression] FAIL: gcc.target/i386/pr37434-[24].c

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46768

--- Comment #9 from H.J. Lu <hjl.tools at gmail dot com> 2010-12-03 15:06:31 UTC ---
(In reply to comment #8)
> Created attachment 22616 [details]
> assembly file from gcc.target/i386/sse2-init-v2di-2.c on x86_64-apple-darwin10
> at r167375

It was fixed by revision 167398.  Why do you use r167375?

[Bug target/46768] [4.6 Regression] FAIL: gcc.target/i386/pr37434-[24].c

[howarth at nitro dot med.uc.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46768

--- Comment #10 from Jack Howarth <howarth at nitro dot med.uc.edu> 2010-12-03 15:21:12 UTC ---
(In reply to comment #9)
> (In reply to comment #8)
> > Created attachment 22616 [details] [details]
> > assembly file from gcc.target/i386/sse2-init-v2di-2.c on x86_64-apple-darwin10
> > at r167375
> 
> It was fixed by revision 167398.  Why do you use r167375?

Since it still occurs in r167400 on darwin, the difference between r167372  and
r167375 should capture the assembly changes with the absent movq in
sse2-init-v2di-2.s on x86_64-apple-darwin10.

[Bug target/46768] [4.6 Regression] FAIL: gcc.target/i386/pr37434-[24].c

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46768

--- Comment #11 from H.J. Lu <hjl.tools at gmail dot com> 2010-12-03 15:34:10 UTC ---
(In reply to comment #10)
> (In reply to comment #9)
> > (In reply to comment #8)
> > > Created attachment 22616 [details] [details] [details]
> > > assembly file from gcc.target/i386/sse2-init-v2di-2.c on x86_64-apple-darwin10
> > > at r167375
> > 
> > It was fixed by revision 167398.  Why do you use r167375?
> 
> Since it still occurs in r167400 on darwin, the difference between r167372  and
> r167375 should capture the assembly changes with the absent movq in
> sse2-init-v2di-2.s on x86_64-apple-darwin10.

I don't need to see the assembly code to know revision 167375 is bad.  
Please show the output from revision 167398 or above.

[Bug target/46768] [4.6 Regression] FAIL: gcc.target/i386/pr37434-[24].c

[howarth at nitro dot med.uc.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46768

--- Comment #12 from Jack Howarth <howarth at nitro dot med.uc.edu> 2010-12-03 16:20:12 UTC ---
Created attachment 22619
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=22619
assembly file from gcc.target/i386/sse2-init-v2di-2.c on x86_64-apple-darwin10
at r167430

[Bug target/46768] [4.6 Regression] FAIL: gcc.target/i386/pr37434-[24].c

[howarth at nitro dot med.uc.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46768

--- Comment #13 from Jack Howarth <howarth at nitro dot med.uc.edu> 2010-12-03 16:20:59 UTC ---
Created attachment 22620
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=22620
assembly file from gcc.target/i386/sse2-init-v2di-2.c on x86_64-apple-darwin10
at r167430 with darwin core2 patch

[Bug target/46768] [4.6 Regression] FAIL: gcc.target/i386/pr37434-[24].c

[howarth at nitro dot med.uc.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46768

--- Comment #14 from Jack Howarth <howarth at nitro dot med.uc.edu> 2010-12-03 16:22:05 UTC ---
Created attachment 22621
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=22621
darwin core2 patch

[Bug target/46768] [4.6 Regression] FAIL: gcc.target/i386/pr37434-[24].c

[howarth at nitro dot med.uc.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46768

--- Comment #15 from Jack Howarth <howarth at nitro dot med.uc.edu> 2010-12-03 16:27:52 UTC ---
I understand the issue now. Stock r167430 is fine but the addition of the
approved patch to fix the md files and default i386 darwin to -mtune=core2
causes the failure in...

FAIL: gcc.target/i386/sse2-init-v2di-2.c scan-assembler movq

Apparently the darwin core2 patch will also need to correct the
sse2-init-v2di-2.c testcase with...

Index: gcc.target/i386/sse2-init-v2di-2.c
===================================================================
--- gcc.target/i386/sse2-init-v2di-2.c    (revision 167430)
+++ gcc.target/i386/sse2-init-v2di-2.c    (working copy)
@@ -10,4 +10,4 @@
   return _mm_cvtsi64_si128 (b); 
 }

-/* { dg-final { scan-assembler "movq" } } */
+/* { dg-final { scan-assembler "movd" } } */

[Bug target/46768] [4.6 Regression] FAIL: gcc.target/i386/pr37434-[24].c

[howarth at nitro dot med.uc.edu]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46768

--- Comment #16 from Jack Howarth <howarth at nitro dot med.uc.edu> 2010-12-03 16:29:31 UTC ---
Note the original i386 darwin patch with the rational for the movq to movd
assembly changes can be found at
http://gcc.gnu.org/ml/gcc-patches/2010-08/msg01024.html.

[Bug target/46768] [4.6 Regression] FAIL: gcc.target/i386/pr37434-[24].c

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46768

--- Comment #17 from H.J. Lu <hjl.tools at gmail dot com> 2010-12-03 16:40:25 UTC ---
(In reply to comment #15)
> I understand the issue now. Stock r167430 is fine but the addition of the
> approved patch to fix the md files and default i386 darwin to -mtune=core2
> causes the failure in...
> 
> FAIL: gcc.target/i386/sse2-init-v2di-2.c scan-assembler movq
> 
> Apparently the darwin core2 patch will also need to correct the
> sse2-init-v2di-2.c testcase with...
> 
> Index: gcc.target/i386/sse2-init-v2di-2.c
> ===================================================================
> --- gcc.target/i386/sse2-init-v2di-2.c    (revision 167430)
> +++ gcc.target/i386/sse2-init-v2di-2.c    (working copy)
> @@ -10,4 +10,4 @@
>    return _mm_cvtsi64_si128 (b); 
>  }
> 
> -/* { dg-final { scan-assembler "movq" } } */
> +/* { dg-final { scan-assembler "movd" } } */

That is incorrect. I will fix it properly after Dwarn patch is checked in.