[Bug c/56120] New: built-in SIMD with statement expression causes ICE: in iterative_hash_expr, at tree.c:6990 when optimization is on

[Bug c/56120] New: built-in SIMD with statement expression causes ICE: in iterative_hash_expr, at tree.c:6990 when optimization is on

[cjxgm at 126 dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56120

             Bug #: 56120
           Summary: built-in SIMD with statement expression causes ICE: in
                    iterative_hash_expr, at tree.c:6990 when optimization
                    is on
    Classification: Unclassified
           Product: gcc
           Version: 4.7.2
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: cjxgm@126.com
            Target: i686-pc-linux-gnu
             Build: /build/src/gcc-4.7.2/configure --prefix=/usr
                    --libdir=/usr/lib --libexecdir=/usr/lib
                    --mandir=/usr/share/man --infodir=/usr/share/info
                    --with-bugurl=https://bugs.archlinux.org/
                    --enable-languages=c,c++,ada,fortran,go,lto,objc,obj-c
                    ++ --enable-shared --ena


The following is the code:

typedef float v4 __attribute__((vector_size(sizeof(float)*4)));
v4 v4refract(v4 n, v4 v, float ior)
{
 return n * ({ v4 __t = (n) * (v); __t[0] + __t[1] + __t[2]; });
}

the compiler output:
$ gcc -msse -O3 -c -o v.o v.c
v.c: In function 'v4refract':
v.c:5:11: internal compiler error: in iterative_hash_expr, at tree.c:6990
Please submit a full bug report,
with preprocessed source if appropriate.
See <https://bugs.archlinux.org/> for instructions.

Compilation succeeds when the optimization level is 0
but if the optimization level is above 0, whatever 1, 2 or 3, produces this
ICE.

[Bug middle-end/56120] built-in SIMD with statement expression causes ICE: in iterative_hash_expr, at tree.c:6990 when optimization is on

[pinskia at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56120

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
           Keywords|                            |ice-on-valid-code
   Last reconfirmed|                            |2013-01-27
          Component|c                           |middle-end
     Ever Confirmed|0                           |1
              Build|/build/src/gcc-4.7.2/config |
                   |ure --prefix=/usr           |
                   |--libdir=/usr/lib           |
                   |--libexecdir=/usr/lib       |
                   |--mandir=/usr/share/man     |
                   |--infodir=/usr/share/info   |
                   |--with-bugurl=https://bugs. |
                   |archlinux.org/              |
                   |--enable-languages=c,c++,ad |
                   |a,fortran,go,lto,objc       |
              Build|obj-c++ --enable-shared     |
                   |--ena                       |

--- Comment #1 from Andrew Pinski <pinskia at gcc dot gnu.org> 2013-01-27 07:39:44 UTC ---
Confirmed, not a regression.  Easy workaround is to assign the statement
expression to a variable before use.

[Bug middle-end/56120] built-in SIMD with statement expression causes ICE: in iterative_hash_expr, at tree.c:6990 when optimization is on

[glisse at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=56120

--- Comment #2 from Marc Glisse <glisse at gcc dot gnu.org> 2013-01-27 09:56:12 UTC ---
I see it with 4.7, but not with trunk, it may already be fixed there.

[Bug middle-end/56120] built-in SIMD with statement expression causes ICE: in iterative_hash_expr, at tree.c:6990 when optimization is on

[pinskia at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=56120

Andrew Pinski <pinskia at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         Resolution|---                         |FIXED
             Status|NEW                         |RESOLVED
   Target Milestone|---                         |4.8.0

--- Comment #3 from Andrew Pinski <pinskia at gcc dot gnu.org> ---
Fixed in GCC 4.8.0.