[Bug tree-optimization/48499] New: [4.7 regression] ICE compiling 64-bit gcc.dg/vect/{slp-21.c, no-vfa-pr29145.c} on SPARC

[Bug tree-optimization/48499] New: [4.7 regression] ICE compiling 64-bit gcc.dg/vect/{slp-21.c, no-vfa-pr29145.c} on SPARC

[ro at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48499

           Summary: [4.7 regression] ICE compiling 64-bit
                    gcc.dg/vect/{slp-21.c, no-vfa-pr29145.c} on SPARC
           Product: gcc
           Version: 4.7.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: tree-optimization
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: ro@gcc.gnu.org
                CC: ebotcazou@gcc.gnu.org
              Host: sparc*-*-*
            Target: sparc*-*-*
             Build: sparc*-*-*


Since at least 20110331, two 64-bit gcc.dg/vect tests FAIL with an ICE:

$ /var/gcc/regression/trunk/11-gcc/build/gcc/xgcc
-B/var/gcc/regression/trunk/11-gcc/build/gcc/
/vol/gcc/src/hg/trunk/local/gcc/testsuite/gcc.dg/vect/slp-21.c -mcpu=ultrasparc
-mvis -ftree-vectorize -fno-vect-cost-model -O2 -fdump-tree-vect-details -lm
-m64 -o ./slp-21.exe
/vol/gcc/src/hg/trunk/local/gcc/testsuite/gcc.dg/vect/slp-21.c: In function
'main1':
/vol/gcc/src/hg/trunk/local/gcc/testsuite/gcc.dg/vect/slp-21.c:191:1: error:
unrecognizable insn:
(insn 637 123 634 2 (set (reg:V4HI 4 %g4)
        (mem/u/c/i:V4HI (symbol_ref/u:DI ("*.LLC0") [flags 0x2]) [2 S8 A64]))
/vol/gcc/src/hg/trunk/local/gcc/testsuite/gcc.dg/vect/slp-21.c:9 -1
     (nil))
/vol/gcc/src/hg/trunk/local/gcc/testsuite/gcc.dg/vect/slp-21.c:191:1: internal
compiler error: in extract_insn, at recog.c:2109

$ /var/gcc/regression/trunk/11-gcc/build/gcc/xgcc
-B/var/gcc/regression/trunk/11-gcc/build/gcc/
/vol/gcc/src/hg/trunk/local/gcc/testsuite/gcc.dg/vect/no-vfa-pr29145.c
-mcpu=ultrasparc -mvis -ftree-vectorize -fno-vect-cost-model -O2
-fdump-tree-vect-details --param vect-max-version-for-alias-checks=0 -lm -m64
-o ./no-vfa-pr29145.exe
/vol/gcc/src/hg/trunk/local/gcc/testsuite/gcc.dg/vect/no-vfa-pr29145.c: In
function 'main':
/vol/gcc/src/hg/trunk/local/gcc/testsuite/gcc.dg/vect/no-vfa-pr29145.c:49:1:
error: unrecognizable insn:
(insn 135 45 132 2 (set (reg:V2SI 3 %g3)
        (mem/u/c/i:V2SI (symbol_ref/u:DI ("*.LLC0") [flags 0x2]) [2 S8 A64]))
/vol/gcc/src/hg/trunk/local/gcc/testsuite/gcc.dg/vect/no-vfa-pr29145.c:35 -1
     (nil))
/vol/gcc/src/hg/trunk/local/gcc/testsuite/gcc.dg/vect/no-vfa-pr29145.c:49:1:
internal compiler error: in extract_insn, at recog.c:2109

This is a regression from 4.6.

[Bug tree-optimization/48499] [4.7 regression] ICE compiling 64-bit gcc.dg/vect/{slp-21.c, no-vfa-pr29145.c} on SPARC

[ebotcazou at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48499

Eric Botcazou <ebotcazou at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2011.04.07 17:52:14
     Ever Confirmed|0                           |1

--- Comment #1 from Eric Botcazou <ebotcazou at gcc dot gnu.org> 2011-04-07 17:52:14 UTC ---
Vector support is quite damaged on the SPARC because of the recent IRA changes.

[Bug tree-optimization/48499] [4.7 regression] ICE compiling 64-bit gcc.dg/vect/{slp-21.c, no-vfa-pr29145.c} on SPARC

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48499

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |4.7.0

[Bug tree-optimization/48499] [4.7 regression] ICE compiling 64-bit gcc.dg/vect/{slp-21.c, no-vfa-pr29145.c} on SPARC

[ebotcazou at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48499

Eric Botcazou <ebotcazou at gcc dot gnu.org> changed:

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

--- Comment #2 from Eric Botcazou <ebotcazou at gcc dot gnu.org> 2011-06-27 09:10:11 UTC ---
No more ICEs by me.