[Bug target/68124] New: [6 Regression] Many i386 test failures

[Bug target/68124] New: [6 Regression] Many i386 test failures

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68124

            Bug ID: 68124
           Summary: [6 Regression] Many i386 test failures
           Product: gcc
           Version: 6.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: target
          Assignee: unassigned at gcc dot gnu.org
          Reporter: hjl.tools at gmail dot com
  Target Milestone: ---

On Linux/x86, r229459 gave

FAIL: gcc.target/i386/avx2-vect-shift.c (internal compiler error)
FAIL: gcc.target/i386/avx2-vect-shift.c (test for excess errors)
FAIL: gcc.target/i386/pr51235.c (internal compiler error)
FAIL: gcc.target/i386/pr51235.c (test for excess errors)
FAIL: gcc.target/i386/pr58137.c (internal compiler error)
FAIL: gcc.target/i386/pr58137.c (test for excess errors)
FAIL: gcc.target/i386/pr64110.c (internal compiler error)
FAIL: gcc.target/i386/pr64110.c (test for excess errors)

r229452 is OK.

[Bug target/68124] [6 Regression] Many i386 test failures

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68124

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

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

--- Comment #1 from H.J. Lu <hjl.tools at gmail dot com> ---
It was cased by r229458:

[hjl@gnu-ivb-1 gcc]$ ./xgcc -B./ -march=corei7 -O3 -mavx2 -S
/export/gnu/import/git/gcc-regression/gcc/gcc/testsuite/gcc.target/i386/pr58137.c
-m32
/export/gnu/import/git/gcc-regression/gcc/gcc/testsuite/gcc.target/i386/pr58137.c:
In function \u2018more_xrv\u2019:
/export/gnu/import/git/gcc-regression/gcc/gcc/testsuite/gcc.target/i386/pr58137.c:33:1:
internal compiler error: in gen_reg_rtx, at emit-rtl.c:1031
 }
 ^
0x99ad69 gen_reg_rtx(machine_mode)
        ../../../../gcc/gcc/emit-rtl.c:1031
0x9bae10 copy_to_reg(rtx_def*)
        ../../../../gcc/gcc/explow.c:570
0xd26e50 gen_lowpart_general(machine_mode, rtx_def*)
        ../../../../gcc/gcc/rtlhooks.c:62
0x13fddbb gen_split_392(rtx_insn*, rtx_def**)
        ../../../../gcc/gcc/config/i386/sse.md:17100
0x15ab55b split_23
        ../../../../gcc/gcc/config/i386/sse.md:17098
0x15ac314 split_24
        ../../../../gcc/gcc/config/i386/i386.md:4803
0x15afe7c split_insns(rtx_def*, rtx_insn*)
        ../../../../gcc/gcc/config/i386/i386.md:15867
0x9a1011 try_split(rtx_def*, rtx_insn*, int)
        ../../../../gcc/gcc/emit-rtl.c:3666
0xcbce89 split_insn
        ../../../../gcc/gcc/recog.c:2874
0xcbd161 split_all_insns()
        ../../../../gcc/gcc/recog.c:2964
0xcbef85 rest_of_handle_split_after_reload
        ../../../../gcc/gcc/recog.c:3902
0xcbefd6 execute
        ../../../../gcc/gcc/recog.c:3931
Please submit a full bug report,
with preprocessed source if appropriate.
Please include the complete backtrace with any bug report.
See <http://gcc.gnu.org/bugs.html> for instructions.
[hjl@gnu-ivb-1 gcc]$

[Bug target/68124] [6 Regression] Many i386 test failures

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68124

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2015-10-27
   Target Milestone|---                         |6.0
     Ever confirmed|0                           |1

--- Comment #2 from H.J. Lu <hjl.tools at gmail dot com> ---
They failed with -mtune=corei7.

[Bug target/68124] [6 Regression] Many i386 test failures

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68124

--- Comment #3 from H.J. Lu <hjl.tools at gmail dot com> ---
Also

FAIL: gcc.dg/gomp/simd-clones-2.c (internal compiler error)
FAIL: gcc.dg/gomp/simd-clones-2.c (test for excess errors)
FAIL: gcc.dg/lto/simd-function c_lto_simd-function_0.o-c_lto_simd-function_0.o
link,  -fopenmp-simd -O3 -ffast-math -mavx2 -flto -flto-partition=max 
(internal compiler error)
FAIL: gcc.dg/vect/pr64421.c -flto -ffat-lto-objects (internal compiler error)
FAIL: gcc.dg/vect/pr64421.c -flto -ffat-lto-objects (test for excess errors)
FAIL: gcc.dg/vect/pr64421.c (internal compiler error)
FAIL: gcc.dg/vect/pr64421.c (test for excess errors)
FAIL: gcc.dg/vect/vect-simd-clone-11.c -flto -ffat-lto-objects (internal
compiler error)
FAIL: gcc.dg/vect/vect-simd-clone-11.c -flto -ffat-lto-objects (test for excess
errors)
FAIL: gcc.dg/vect/vect-simd-clone-11.c (internal compiler error)
FAIL: gcc.dg/vect/vect-simd-clone-11.c (test for excess errors)
FAIL: gcc.dg/vect/vect-simd-clone-13.c -flto -ffat-lto-objects (internal
compiler error)
FAIL: gcc.dg/vect/vect-simd-clone-13.c -flto -ffat-lto-objects (test for excess
errors)
FAIL: gcc.dg/vect/vect-simd-clone-13.c (internal compiler error)
FAIL: gcc.dg/vect/vect-simd-clone-13.c (test for excess errors)
FAIL: gcc.dg/vect/vect-simd-clone-14.c -flto -ffat-lto-objects (internal
compiler error)
FAIL: gcc.dg/vect/vect-simd-clone-14.c -flto -ffat-lto-objects (test for excess
errors)
FAIL: gcc.dg/vect/vect-simd-clone-14.c (internal compiler error)
FAIL: gcc.dg/vect/vect-simd-clone-14.c (test for excess errors)
FAIL: gcc.dg/vect/vect-simd-clone-15.c -flto -ffat-lto-objects (internal
compiler error)
FAIL: gcc.dg/vect/vect-simd-clone-15.c -flto -ffat-lto-objects (test for excess
errors)
FAIL: gcc.dg/vect/vect-simd-clone-15.c (internal compiler error)
FAIL: gcc.dg/vect/vect-simd-clone-15.c (test for excess errors)
FAIL: gcc.dg/vect/vect-simd-clone-1.c -flto -ffat-lto-objects (internal
compiler error)
FAIL: gcc.dg/vect/vect-simd-clone-1.c -flto -ffat-lto-objects (test for excess
errors)
FAIL: gcc.dg/vect/vect-simd-clone-1.c (internal compiler error)
FAIL: gcc.dg/vect/vect-simd-clone-1.c (test for excess errors)
FAIL: gcc.dg/vect/vect-simd-clone-6.c -flto -ffat-lto-objects (internal
compiler error)
FAIL: gcc.dg/vect/vect-simd-clone-6.c -flto -ffat-lto-objects (test for excess
errors)
FAIL: gcc.dg/vect/vect-simd-clone-6.c (internal compiler error)
FAIL: gcc.dg/vect/vect-simd-clone-6.c (test for excess errors)
FAIL: gcc.dg/vect/vect-simd-clone-7.c -flto -ffat-lto-objects (internal
compiler error)
FAIL: gcc.dg/vect/vect-simd-clone-7.c -flto -ffat-lto-objects (test for excess
errors)
FAIL: gcc.dg/vect/vect-simd-clone-7.c (internal compiler error)
FAIL: gcc.dg/vect/vect-simd-clone-7.c (test for excess errors)
FAIL: gcc.target/i386/avx2-gather-1.c (internal compiler error)
FAIL: gcc.target/i386/avx2-gather-1.c (test for excess errors)
FAIL: gcc.target/i386/avx2-gather-2.c (internal compiler error)
FAIL: gcc.target/i386/avx2-gather-2.c scan-tree-dump-times vect "note:
vectorized 1 loops in function" 16
FAIL: gcc.target/i386/avx2-gather-2.c (test for excess errors)
FAIL: gcc.target/i386/avx2-pr63594-1.c (internal compiler error)
FAIL: gcc.target/i386/avx2-pr63594-1.c (test for excess errors)
FAIL: gcc.target/i386/avx2-pr63594-2.c (internal compiler error)
FAIL: gcc.target/i386/avx2-pr63594-2.c (test for excess errors)
FAIL: gcc.target/i386/avx2-vect-shift.c (internal compiler error)
FAIL: gcc.target/i386/avx2-vect-shift.c (internal compiler error)
FAIL: gcc.target/i386/avx2-vect-shift.c (test for excess errors)
FAIL: gcc.target/i386/avx2-vect-shift.c (test for excess errors)
FAIL: gcc.target/i386/avx512f-pr63594-1.c (internal compiler error)
FAIL: gcc.target/i386/avx512f-pr63594-1.c (test for excess errors)
FAIL: gcc.target/i386/avx512f-pr63594-2.c (internal compiler error)
FAIL: gcc.target/i386/avx512f-pr63594-2.c (test for excess errors)
FAIL: gcc.target/i386/avx512f-vec-init.c (internal compiler error)
FAIL: gcc.target/i386/avx512f-vec-init.c (test for excess errors)
FAIL: gcc.target/i386/pr51235.c (internal compiler error)
FAIL: gcc.target/i386/pr51235.c (test for excess errors)
FAIL: gcc.target/i386/pr51236.c (internal compiler error)
FAIL: gcc.target/i386/pr51236.c (test for excess errors)
FAIL: gcc.target/i386/pr58137.c (internal compiler error)
FAIL: gcc.target/i386/pr58137.c (test for excess errors)
FAIL: gcc.target/i386/pr64110.c (internal compiler error)
FAIL: gcc.target/i386/pr64110.c (internal compiler error)
FAIL: gcc.target/i386/pr64110.c (test for excess errors)
FAIL: gcc.target/i386/pr64110.c (test for excess errors)
FAIL: gcc.target/i386/pr67447.c (internal compiler error)
FAIL: gcc.target/i386/pr67447.c (test for excess errors)
FAIL: g++.dg/vect/simd-clone-1.cc  -std=c++11 (internal compiler error)
FAIL: g++.dg/vect/simd-clone-1.cc  -std=c++11 (test for excess errors)
FAIL: g++.dg/vect/simd-clone-1.cc  -std=c++14 (internal compiler error)
FAIL: g++.dg/vect/simd-clone-1.cc  -std=c++14 (test for excess errors)
FAIL: g++.dg/vect/simd-clone-1.cc  -std=c++98 (internal compiler error)
FAIL: g++.dg/vect/simd-clone-1.cc  -std=c++98 (test for excess errors)

on x86-64 with -march=corei7.

[Bug target/68124] [6 Regression] Many i386 test failures

[rth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68124

Richard Henderson <rth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED
           Assignee|unassigned at gcc dot gnu.org      |rth at gcc dot gnu.org

--- Comment #4 from Richard Henderson <rth at gcc dot gnu.org> ---
Mine.

[Bug target/68124] [6 Regression] Many i386 test failures

[rth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68124

Richard Henderson <rth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |WAITING

--- Comment #5 from Richard Henderson <rth at gcc dot gnu.org> ---
Works for me at r229501.

spawn -ignore SIGHUP /home/rth/gcc/bld/gcc/xgcc -B/home/rth/gcc/bld/gcc/
/home/rth/gcc/git-master/gcc/testsuite/gcc.dg/vect/pr64421.c -mtune=corei7
-fno-diagnostics-show-caret -fdiagnostics-color=never -msse2 -ftree-vectorize
-fno-vect-cost-model -fno-common -O2 -fdump-tree-vect-details -fopenmp-simd
-mavx -lm -o ./pr64421.exe
PASS: gcc.dg/vect/pr64421.c (test for excess errors)

[Bug target/68124] [6 Regression] Many i386 test failures

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68124

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|WAITING                     |NEW

--- Comment #6 from H.J. Lu <hjl.tools at gmail dot com> ---
(In reply to Richard Henderson from comment #5)
> Works for me at r229501.
> 
> spawn -ignore SIGHUP /home/rth/gcc/bld/gcc/xgcc -B/home/rth/gcc/bld/gcc/
> /home/rth/gcc/git-master/gcc/testsuite/gcc.dg/vect/pr64421.c -mtune=corei7
> -fno-diagnostics-show-caret -fdiagnostics-color=never -msse2
> -ftree-vectorize -fno-vect-cost-model -fno-common -O2
> -fdump-tree-vect-details -fopenmp-simd -mavx -lm -o ./pr64421.exe
> PASS: gcc.dg/vect/pr64421.c (test for excess errors)

Add -m32.

[Bug target/68124] [6 Regression] Many i386 test failures

[rth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68124

Richard Henderson <rth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |WAITING

--- Comment #7 from Richard Henderson <rth at gcc dot gnu.org> ---
This was in an i386 vm -- no need for -m32.

[Bug target/68124] [6 Regression] Many i386 test failures

[rth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68124

--- Comment #9 from Richard Henderson <rth at gcc dot gnu.org> ---
Created attachment 36611
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=36611&action=edit
proposed patch

Ho hum.  Post-reload register adjustments ought to use something
than gen_lowpart to force a hard register into a new mode.

So let's be a tad more specific in the checks.

[Bug target/68124] [6 Regression] Many i386 test failures

[hjl.tools at gmail dot com]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68124

--- Comment #10 from H.J. Lu <hjl.tools at gmail dot com> ---
(In reply to Richard Henderson from comment #9)
> Created attachment 36611 [details]
> proposed patch
> 
> Ho hum.  Post-reload register adjustments ought to use something
> than gen_lowpart to force a hard register into a new mode.
> 
> So let's be a tad more specific in the checks.

It seems to fix the problem.  Thanks.

[Bug target/68124] [6 Regression] Many i386 test failures

[rth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68124

--- Comment #11 from Richard Henderson <rth at gcc dot gnu.org> ---
Author: rth
Date: Thu Oct 29 18:36:39 2015
New Revision: 229550

URL: https://gcc.gnu.org/viewcvs?rev=229550&root=gcc&view=rev
Log:
Fix target/68124

        PR target/68124
        PR rtl-opt/67609
        * config/i386/i386.c (ix86_cannot_change_mode_class): Tighten
        sse check to the exact conditions of PR 67609.

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

[Bug target/68124] [6 Regression] Many i386 test failures

[rth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68124

Richard Henderson <rth at gcc dot gnu.org> changed:

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

--- Comment #12 from Richard Henderson <rth at gcc dot gnu.org> ---
Fixed.