[Bug tree-optimization/23546] New: ICE in for_each_index, at tree-ssa-loop-im.c:202

[Bug tree-optimization/23546] New: ICE in for_each_index, at tree-ssa-loop-im.c:202

[ma1flfs at bath dot ac dot uk]

When compiling xorg-cvs gcc throws the following ICE in 
programs/Xserver/fb/fbmmx.c  
fbmmx.c: In function 'mmxCombineAddU':  
fbmmx.c:564: internal compiler error: in for_each_index, at  
tree-ssa-loop-im.c:202  
  
gcc (GCC) 4.0.2 20050823 (prerelease)  
  
Configured with: ../gcc-4.0-20050824/configure --prefix=/usr  
--libexecdir=/usr/lib --enable-shared --enable-threads=posix  
--enable-__cxa_atexit --enable-clocale=gnu --enable-languages=c,c++  
  
Condensed command line:  
gcc -c -O2 -march=pentium4 fbmmx.i 
 
Compiling the file without -O2 works. 
  
As the preprocessed file is very long I will attach it to this report instead  
of including it in the comment.

-- 
           Summary: ICE in for_each_index, at tree-ssa-loop-im.c:202
           Product: gcc
           Version: 4.0.2
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: tree-optimization
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: ma1flfs at bath dot ac dot uk
                CC: gcc-bugs at gcc dot gnu dot org
  GCC host triplet: i686-pc-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23546

[Bug tree-optimization/23546] ICE in for_each_index, at tree-ssa-loop-im.c:202

[ma1flfs at bath dot ac dot uk]

------- Additional Comments From ma1flfs at bath dot ac dot uk  2005-08-24 11:58 -------
Created an attachment (id=9575)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=9575&action=view)
Preprocessed source file


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23546

[Bug tree-optimization/23546] [4.0/4.1 Regression] ICE in for_each_index, at tree-ssa-loop-im.c:202

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2005-08-24 12:04 -------
Reducing.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |pinskia at gcc dot gnu dot
                   |                            |org
           Keywords|                            |ice-on-valid-code
      Known to fail|                            |4.1.0
      Known to work|                            |3.4.0
            Summary|ICE in for_each_index, at   |[4.0/4.1 Regression] ICE in
                   |tree-ssa-loop-im.c:202      |for_each_index, at tree-ssa-
                   |                            |loop-im.c:202
   Target Milestone|---                         |4.0.2


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23546

[Bug tree-optimization/23546] [4.0/4.1 Regression] ICE in for_each_index, at tree-ssa-loop-im.c:202

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2005-08-24 12:35 -------
This was exposed by:
2005-08-23  Paolo Bonzini  <bonzini@gnu.org>

        PR middle-end/23517
        * fold-const.c (fold_convert): Use VIEW_CONVERT_EXPR to convert
        between vectors.
        * convert.c (convert_to_integer, convert_to_vector): Likewise.

Reduced testcase:
typedef int __m64 __attribute__ ((__vector_size__ (8)));
       typedef char __v8qi __attribute__ ((__vector_size__ (8)));
         static __inline __m64 _mm_cvtsi32_si64 (int __i) {
   }
          static __inline __m64 _mm_unpacklo_pi8 (__m64 __m1, __m64 __m2) {
     return (__m64) __builtin_ia32_punpcklbw ((__v8qi)__m1, (__v8qi)__m2);
   }
         static __inline __m64 _mm_setzero_si64 (void) {
     return (__m64)0LL;
   }
       typedef unsigned long CARD32;
        static __inline__ __m64 load8888 (CARD32 v) {
       return _mm_unpacklo_pi8 (_mm_cvtsi32_si64 (v), _mm_setzero_si64());
   }
          __attribute__((regparm(3))) void mmxCombineMaskU (CARD32 *src, const CARD32 *mask, int 
width) {
       const CARD32 *end = mask + width;
       while (mask < end) {
          __m64 a = load8888(*mask);
          ++mask;
      }
   }

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |rakdver at gcc dot gnu dot
                   |                            |org, bonzini at gcc dot gnu
                   |                            |dot org
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|                            |1
      Known to work|3.4.0                       |3.4.0 4.0.0
   Last reconfirmed|0000-00-00 00:00:00         |2005-08-24 12:35:03
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23546

[Bug tree-optimization/23546] [4.0/4.1 Regression] ICE in for_each_index, at tree-ssa-loop-im.c:202

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2005-08-24 12:40 -------
The loop is:
  # mask_12 = PHI <mask_26(3), mask_6(1)>;
<L0>:;
  __m2_11 = VIEW_CONVERT_EXPR<vector int>(0);
  D.1667_18 = VIEW_CONVERT_EXPR<vector signed char>(__m2_11);
  D.1668_19 = VIEW_CONVERT_EXPR<vector signed char>(D.1659_15);
  D.1669_20 = __builtin_ia32_punpcklbw (D.1668_19, D.1667_18);
  mask_26 = mask_12 + 4B;
  if (end_7 > mask_26) goto <L10>; else goto <L2>;


so tree-ssa-loop-im is not handling VCE at all.

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23546

[Bug tree-optimization/23546] [4.0/4.1 Regression] ICE in for_each_index, at tree-ssa-loop-im.c:202

[bonzini at gcc dot gnu dot org]

------- Additional Comments From bonzini at gcc dot gnu dot org  2005-08-24 13:50 -------
Nicer test case:

typedef int m64 __attribute__ ((__vector_size__ (8)));

void mmxCombineMaskU (m64 * mask, int width)
{
  while (--width >= 0)
    *mask++ = (m64) 0LL;
}

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23546

[Bug tree-optimization/23546] [4.0/4.1 Regression] ICE in for_each_index, at tree-ssa-loop-im.c:202

[bonzini at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|unassigned at gcc dot gnu   |bonzini at gcc dot gnu dot
                   |dot org                     |org
             Status|NEW                         |ASSIGNED
   Last reconfirmed|2005-08-24 12:35:03         |2005-08-24 13:58:14
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23546

[Bug tree-optimization/23546] [4.0/4.1 Regression] ICE in for_each_index, at tree-ssa-loop-im.c:202

[bonzini at gcc dot gnu dot org]

------- Additional Comments From bonzini at gcc dot gnu dot org  2005-08-24 14:08 -------
It's not handling VCE of an integer constant (or a real constant too, for that
matter).

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23546

[Bug tree-optimization/23546] [4.0/4.1 Regression] ICE in for_each_index, at tree-ssa-loop-im.c:202

[cvs-commit at gcc dot gnu dot org]

------- Additional Comments From cvs-commit at gcc dot gnu dot org  2005-08-25 06:40 -------
Subject: Bug 23546

CVSROOT:	/cvs/gcc
Module name:	gcc
Changes by:	bonzini@gcc.gnu.org	2005-08-25 06:40:14

Modified files:
	gcc            : ChangeLog tree-ssa-loop-im.c 

Log message:
	2005-08-24  Paolo Bonzini  <bonzini@gnu.org>
	
	PR tree-optimization/23546
	* tree-ssa-loop-im.c (for_each_index): Handle INTEGER_CST
	and REAL_CST nodes.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/ChangeLog.diff?cvsroot=gcc&r1=2.9821&r2=2.9822
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/tree-ssa-loop-im.c.diff?cvsroot=gcc&r1=2.50&r2=2.51



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23546

[Bug tree-optimization/23546] [4.0/4.1 Regression] ICE in for_each_index, at tree-ssa-loop-im.c:202

[cvs-commit at gcc dot gnu dot org]

------- Additional Comments From cvs-commit at gcc dot gnu dot org  2005-08-25 07:05 -------
Subject: Bug 23546

CVSROOT:	/cvs/gcc
Module name:	gcc
Changes by:	bonzini@gcc.gnu.org	2005-08-25 07:03:00

Modified files:
	gcc/testsuite  : ChangeLog 
Added files:
	gcc/testsuite/gcc.dg/tree-ssa: pr23546.c 

Log message:
	2005-08-24  Paolo Bonzini  <bonzini@gnu.org>
	
	PR tree-optimization/23546
	* gcc.dg/tree-ssa/pr23546.c: New test.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&r1=1.5957&r2=1.5958
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gcc.dg/tree-ssa/pr23546.c.diff?cvsroot=gcc&r1=NONE&r2=1.1



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23546

[Bug tree-optimization/23546] [4.0/4.1 Regression] ICE in for_each_index, at tree-ssa-loop-im.c:202

[cvs-commit at gcc dot gnu dot org]

------- Additional Comments From cvs-commit at gcc dot gnu dot org  2005-08-25 07:38 -------
Subject: Bug 23546

CVSROOT:	/cvs/gcc
Module name:	gcc
Branch: 	gcc-4_0-branch
Changes by:	bonzini@gcc.gnu.org	2005-08-25 07:37:22

Modified files:
	gcc            : ChangeLog tree-ssa-loop-im.c 
	gcc/testsuite  : ChangeLog 
Added files:
	gcc/testsuite/gcc.dg/tree-ssa: pr23546.c 

Log message:
	gcc:
	2005-08-24  Paolo Bonzini  <bonzini@gnu.org>
	
	PR tree-optimization/23546
	* tree-ssa-loop-im.c (for_each_index): Handle INTEGER_CST
	and REAL_CST nodes.
	
	gcc/testsuite:
	2005-08-24  Paolo Bonzini  <bonzini@gnu.org>
	
	PR tree-optimization/23546
	* gcc.dg/tree-ssa/pr23546.c: New.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/ChangeLog.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=2.7592.2.389&r2=2.7592.2.390
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/tree-ssa-loop-im.c.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=2.28.2.2&r2=2.28.2.3
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=1.5084.2.347&r2=1.5084.2.348
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gcc.dg/tree-ssa/pr23546.c.diff?cvsroot=gcc&only_with_tag=gcc-4_0-branch&r1=NONE&r2=1.1.2.1



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23546

[Bug tree-optimization/23546] [4.0/4.1 Regression] ICE in for_each_index, at tree-ssa-loop-im.c:202

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2005-08-25 11:25 -------
Fixed.

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23546

[Bug tree-optimization/23546] [4.0/4.1 Regression] ICE in for_each_index, at tree-ssa-loop-im.c:202

[ma1flfs at bath dot ac dot uk]

------- Additional Comments From ma1flfs at bath dot ac dot uk  2005-08-26 10:27 -------
Thank you very much for fixing this so quickly! 

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23546