public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug c/39588]  New: internal compiler error: in trunc_int_for_mode, at explow.c:55
@ 2009-03-30 13:29 olivier dot roussel at cril dot univ-artois dot fr
  2009-03-30 13:48 ` [Bug c/39588] " olivier dot roussel at cril dot univ-artois dot fr
                   ` (7 more replies)
  0 siblings, 8 replies; 9+ messages in thread
From: olivier dot roussel at cril dot univ-artois dot fr @ 2009-03-30 13:29 UTC (permalink / raw)
  To: gcc-bugs

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1448 bytes --]

# gcc -v
Using built-in specs.
Target: x86_64-unknown-linux-gnu
Configured with: ../gcc-4.3.3/configure --prefix=/tmp/gcc433
--enable-languages=c,c++
Thread model: posix
gcc version 4.3.3 (GCC)

with the following, self-contained file buc.c

#define lit_from_int(in) ((in<0)?(((-in)<<1)|1):(in<<1))
void init_clause(int* literals,int size, int *lits)
{
  int i;
  for(i=0; i < size; i++) {
    lits[i] = lit_from_int(literals[i]);
  }
}

On a x86_64 host, both gcc 4.3.1 (handmade compilation), gcc 4.3.2 (Fedora Core
10 package) and gcc 4.3.3 (handmade compilation) crash with -O3 optimization

# gcc -O3 -c bug.c 
bug.c: In function ‘init_clause’:
bug.c:10: internal compiler error: in trunc_int_for_mode, at explow.c:55

This bug doesn't occur on a i686 host (gcc 4.3.2). On a x86_64 host, it doesn't
 occur when -O2 is used with gcc 4.3.3 or when gcc 4.2.4 is used.


-- 
           Summary: internal compiler error: in trunc_int_for_mode, at
                    explow.c:55
           Product: gcc
           Version: 4.3.3
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: olivier dot roussel at cril dot univ-artois dot fr
 GCC build triplet: x86_64-unknown-linux-gnu
  GCC host triplet: x86_64-unknown-linux-gnu
GCC target triplet: x86_64-unknown-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39588


^ permalink raw reply	[flat|nested] 9+ messages in thread

* [Bug c/39588] internal compiler error: in trunc_int_for_mode, at explow.c:55
  2009-03-30 13:29 [Bug c/39588] New: internal compiler error: in trunc_int_for_mode, at explow.c:55 olivier dot roussel at cril dot univ-artois dot fr
@ 2009-03-30 13:48 ` olivier dot roussel at cril dot univ-artois dot fr
  2009-03-30 14:24 ` [Bug rtl-optimization/39588] [4.3/4.4/4.5 Regression] " rguenth at gcc dot gnu dot org
                   ` (6 subsequent siblings)
  7 siblings, 0 replies; 9+ messages in thread
From: olivier dot roussel at cril dot univ-artois dot fr @ 2009-03-30 13:48 UTC (permalink / raw)
  To: gcc-bugs



------- Comment #1 from olivier dot roussel at cril dot univ-artois dot fr  2009-03-30 13:48 -------
Created an attachment (id=17558)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=17558&action=view)
test source file


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39588


^ permalink raw reply	[flat|nested] 9+ messages in thread

* [Bug rtl-optimization/39588] [4.3/4.4/4.5 Regression] internal compiler error: in trunc_int_for_mode, at explow.c:55
  2009-03-30 13:29 [Bug c/39588] New: internal compiler error: in trunc_int_for_mode, at explow.c:55 olivier dot roussel at cril dot univ-artois dot fr
  2009-03-30 13:48 ` [Bug c/39588] " olivier dot roussel at cril dot univ-artois dot fr
@ 2009-03-30 14:24 ` rguenth at gcc dot gnu dot org
  2009-03-30 14:56 ` rguenth at gcc dot gnu dot org
                   ` (5 subsequent siblings)
  7 siblings, 0 replies; 9+ messages in thread
From: rguenth at gcc dot gnu dot org @ 2009-03-30 14:24 UTC (permalink / raw)
  To: gcc-bugs



------- Comment #2 from rguenth at gcc dot gnu dot org  2009-03-30 14:24 -------
Confirmed.  -ftree-vectorize is the problem or maybe triggers a target problem.

#1  0x0000000000605517 in trunc_int_for_mode (c=0, mode=V4SImode)
    at /space/rguenther/src/svn/trunk/gcc/explow.c:56
56        gcc_assert (SCALAR_INT_MODE_P (mode));

#1  0x0000000000605517 in trunc_int_for_mode (c=0, mode=V4SImode)
    at /space/rguenther/src/svn/trunk/gcc/explow.c:56
#2  0x0000000000e481a3 in merge_outer_ops (pop0=0x7fffffffd228, 
    pconst0=0x7fffffffd220, op1=NEG, const1=0, mode=V4SImode, 
    pcomp_p=0x7fffffffd21c)
    at /space/rguenther/src/svn/trunk/gcc/combine.c:8962
#3  0x0000000000e494f5 in simplify_shift_const_1 (code=ASHIFT, 
    result_mode=V4SImode, varop=0x7ffff61a9850, orig_count=1)
    at /space/rguenther/src/svn/trunk/gcc/combine.c:9486
#4  0x0000000000e49dfb in simplify_shift_const (x=0x7ffff61a25c0, code=ASHIFT, 
    result_mode=V4SImode, varop=0x7ffff61a9850, count=1)
    at /space/rguenther/src/svn/trunk/gcc/combine.c:9723
#5  0x0000000000e3e277 in combine_simplify_rtx (x=0x7ffff61a25c0, 
    op0_mode=V4SImode, in_dest=0)
    at /space/rguenther/src/svn/trunk/gcc/combine.c:5225
#6  0x0000000000e3c4e1 in subst (x=0x7ffff61a25c0, from=0x7ffff61a24e0, 
    to=0x7ffff61a9850, in_dest=0, unique_copy=0)
    at /space/rguenther/src/svn/trunk/gcc/combine.c:4543
#7  0x0000000000e3c2e7 in subst (x=0x7ffff61a25e0, from=0x7ffff61a24e0, 
    to=0x7ffff61a9850, in_dest=0, unique_copy=0)
    at /space/rguenther/src/svn/trunk/gcc/combine.c:4481
#8  0x0000000000e370bf in try_combine (i3=0x7ffff61a0b40, i2=0x7ffff61a0af0, 
    i1=0x0, new_direct_jump_p=0x7fffffffda34)
    at /space/rguenther/src/svn/trunk/gcc/combine.c:2726
#9  0x0000000000e336bc in combine_instructions (f=0x7ffff61a8080, nregs=111)
    at /space/rguenther/src/svn/trunk/gcc/combine.c:1256
#10 0x0000000000e5197b in rest_of_handle_combine ()
    at /space/rguenther/src/svn/trunk/gcc/combine.c:12999


-- 

rguenth at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
          Component|c                           |rtl-optimization
     Ever Confirmed|0                           |1
           Keywords|                            |ice-on-valid-code
   Last reconfirmed|0000-00-00 00:00:00         |2009-03-30 14:24:15
               date|                            |
            Summary|internal compiler error: in |[4.3/4.4/4.5 Regression]
                   |trunc_int_for_mode, at      |internal compiler error: in
                   |explow.c:55                 |trunc_int_for_mode, at
                   |                            |explow.c:55
   Target Milestone|---                         |4.3.4


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39588


^ permalink raw reply	[flat|nested] 9+ messages in thread

* [Bug rtl-optimization/39588] [4.3/4.4/4.5 Regression] internal compiler error: in trunc_int_for_mode, at explow.c:55
  2009-03-30 13:29 [Bug c/39588] New: internal compiler error: in trunc_int_for_mode, at explow.c:55 olivier dot roussel at cril dot univ-artois dot fr
  2009-03-30 13:48 ` [Bug c/39588] " olivier dot roussel at cril dot univ-artois dot fr
  2009-03-30 14:24 ` [Bug rtl-optimization/39588] [4.3/4.4/4.5 Regression] " rguenth at gcc dot gnu dot org
@ 2009-03-30 14:56 ` rguenth at gcc dot gnu dot org
  2009-03-31 15:44 ` ebotcazou at gcc dot gnu dot org
                   ` (4 subsequent siblings)
  7 siblings, 0 replies; 9+ messages in thread
From: rguenth at gcc dot gnu dot org @ 2009-03-30 14:56 UTC (permalink / raw)
  To: gcc-bugs



------- Comment #3 from rguenth at gcc dot gnu dot org  2009-03-30 14:56 -------
We want to simplify

(ashift:V4SI (neg:V4SI (reg:V4SI 87))
    (const_int 1 [0x1]))

to

(neg:V4SI (ashift:V4Si (reg:V4SI 87)
    (const_int 1 [0x1]))

but the code obviously doesn't expect vector modes.  For the particular
case of shifts we pass down zero, so the following would fix^Wworkaround
the issue

Index: gcc/combine.c
===================================================================
--- gcc/combine.c       (revision 145291)
+++ gcc/combine.c       (working copy)
@@ -8974,7 +8974,8 @@ merge_outer_ops (enum rtx_code *pop0, HO
   /* ??? Slightly redundant with the above mask, but not entirely.
      Moving this above means we'd have to sign-extend the mode mask
      for the final test.  */
-  const0 = trunc_int_for_mode (const0, mode);
+  if (const0 != 0)
+    const0 = trunc_int_for_mode (const0, mode);

   *pop0 = op0;
   *pconst0 = const0;


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39588


^ permalink raw reply	[flat|nested] 9+ messages in thread

* [Bug rtl-optimization/39588] [4.3/4.4/4.5 Regression] internal compiler error: in trunc_int_for_mode, at explow.c:55
  2009-03-30 13:29 [Bug c/39588] New: internal compiler error: in trunc_int_for_mode, at explow.c:55 olivier dot roussel at cril dot univ-artois dot fr
                   ` (2 preceding siblings ...)
  2009-03-30 14:56 ` rguenth at gcc dot gnu dot org
@ 2009-03-31 15:44 ` ebotcazou at gcc dot gnu dot org
  2009-04-01 20:46 ` ebotcazou at gcc dot gnu dot org
                   ` (3 subsequent siblings)
  7 siblings, 0 replies; 9+ messages in thread
From: ebotcazou at gcc dot gnu dot org @ 2009-03-31 15:44 UTC (permalink / raw)
  To: gcc-bugs



-- 

ebotcazou at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
         AssignedTo|unassigned at gcc dot gnu   |ebotcazou at gcc dot gnu dot
                   |dot org                     |org
             Status|NEW                         |ASSIGNED
   Last reconfirmed|2009-03-30 14:24:15         |2009-03-31 15:43:52
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39588


^ permalink raw reply	[flat|nested] 9+ messages in thread

* [Bug rtl-optimization/39588] [4.3/4.4/4.5 Regression] internal compiler error: in trunc_int_for_mode, at explow.c:55
  2009-03-30 13:29 [Bug c/39588] New: internal compiler error: in trunc_int_for_mode, at explow.c:55 olivier dot roussel at cril dot univ-artois dot fr
                   ` (3 preceding siblings ...)
  2009-03-31 15:44 ` ebotcazou at gcc dot gnu dot org
@ 2009-04-01 20:46 ` ebotcazou at gcc dot gnu dot org
  2009-04-01 20:48 ` ebotcazou at gcc dot gnu dot org
                   ` (2 subsequent siblings)
  7 siblings, 0 replies; 9+ messages in thread
From: ebotcazou at gcc dot gnu dot org @ 2009-04-01 20:46 UTC (permalink / raw)
  To: gcc-bugs



------- Comment #4 from ebotcazou at gcc dot gnu dot org  2009-04-01 20:46 -------
Subject: Bug 39588

Author: ebotcazou
Date: Wed Apr  1 20:46:30 2009
New Revision: 145430

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=145430
Log:
        PR rtl-optimization/39588
        * combine.c (merge_outer_ops): Do not set the constant when this
        is not necessary.
        (simplify_shift_const_1): Do not modify it either in this case.

Added:
    trunk/gcc/testsuite/gcc.c-torture/compile/20090401-1.c
Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/combine.c
    trunk/gcc/testsuite/ChangeLog


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39588


^ permalink raw reply	[flat|nested] 9+ messages in thread

* [Bug rtl-optimization/39588] [4.3/4.4/4.5 Regression] internal compiler error: in trunc_int_for_mode, at explow.c:55
  2009-03-30 13:29 [Bug c/39588] New: internal compiler error: in trunc_int_for_mode, at explow.c:55 olivier dot roussel at cril dot univ-artois dot fr
                   ` (4 preceding siblings ...)
  2009-04-01 20:46 ` ebotcazou at gcc dot gnu dot org
@ 2009-04-01 20:48 ` ebotcazou at gcc dot gnu dot org
  2009-04-01 20:48 ` ebotcazou at gcc dot gnu dot org
  2009-04-01 20:51 ` ebotcazou at gcc dot gnu dot org
  7 siblings, 0 replies; 9+ messages in thread
From: ebotcazou at gcc dot gnu dot org @ 2009-04-01 20:48 UTC (permalink / raw)
  To: gcc-bugs



------- Comment #6 from ebotcazou at gcc dot gnu dot org  2009-04-01 20:48 -------
Subject: Bug 39588

Author: ebotcazou
Date: Wed Apr  1 20:48:33 2009
New Revision: 145432

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=145432
Log:
        PR rtl-optimization/39588
        * combine.c (merge_outer_ops): Do not set the constant when this
        is not necessary.
        (simplify_shift_const_1): Do not modify it either in this case.

Added:
    branches/gcc-4_3-branch/gcc/testsuite/gcc.c-torture/compile/20090401-1.c
Modified:
    branches/gcc-4_3-branch/gcc/ChangeLog
    branches/gcc-4_3-branch/gcc/combine.c
    branches/gcc-4_3-branch/gcc/testsuite/ChangeLog


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39588


^ permalink raw reply	[flat|nested] 9+ messages in thread

* [Bug rtl-optimization/39588] [4.3/4.4/4.5 Regression] internal compiler error: in trunc_int_for_mode, at explow.c:55
  2009-03-30 13:29 [Bug c/39588] New: internal compiler error: in trunc_int_for_mode, at explow.c:55 olivier dot roussel at cril dot univ-artois dot fr
                   ` (5 preceding siblings ...)
  2009-04-01 20:48 ` ebotcazou at gcc dot gnu dot org
@ 2009-04-01 20:48 ` ebotcazou at gcc dot gnu dot org
  2009-04-01 20:51 ` ebotcazou at gcc dot gnu dot org
  7 siblings, 0 replies; 9+ messages in thread
From: ebotcazou at gcc dot gnu dot org @ 2009-04-01 20:48 UTC (permalink / raw)
  To: gcc-bugs



------- Comment #5 from ebotcazou at gcc dot gnu dot org  2009-04-01 20:47 -------
Subject: Bug 39588

Author: ebotcazou
Date: Wed Apr  1 20:47:37 2009
New Revision: 145431

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=145431
Log:
        PR rtl-optimization/39588
        * combine.c (merge_outer_ops): Do not set the constant when this
        is not necessary.
        (simplify_shift_const_1): Do not modify it either in this case.

Added:
    branches/gcc-4_4-branch/gcc/testsuite/gcc.c-torture/compile/20090401-1.c
Modified:
    branches/gcc-4_4-branch/gcc/ChangeLog
    branches/gcc-4_4-branch/gcc/combine.c
    branches/gcc-4_4-branch/gcc/testsuite/ChangeLog


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39588


^ permalink raw reply	[flat|nested] 9+ messages in thread

* [Bug rtl-optimization/39588] [4.3/4.4/4.5 Regression] internal compiler error: in trunc_int_for_mode, at explow.c:55
  2009-03-30 13:29 [Bug c/39588] New: internal compiler error: in trunc_int_for_mode, at explow.c:55 olivier dot roussel at cril dot univ-artois dot fr
                   ` (6 preceding siblings ...)
  2009-04-01 20:48 ` ebotcazou at gcc dot gnu dot org
@ 2009-04-01 20:51 ` ebotcazou at gcc dot gnu dot org
  7 siblings, 0 replies; 9+ messages in thread
From: ebotcazou at gcc dot gnu dot org @ 2009-04-01 20:51 UTC (permalink / raw)
  To: gcc-bugs



------- Comment #7 from ebotcazou at gcc dot gnu dot org  2009-04-01 20:51 -------
Thanks for the reduced testcase.


-- 

ebotcazou at gcc dot gnu dot org changed:

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


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39588


^ permalink raw reply	[flat|nested] 9+ messages in thread

end of thread, other threads:[~2009-04-01 20:51 UTC | newest]

Thread overview: 9+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2009-03-30 13:29 [Bug c/39588] New: internal compiler error: in trunc_int_for_mode, at explow.c:55 olivier dot roussel at cril dot univ-artois dot fr
2009-03-30 13:48 ` [Bug c/39588] " olivier dot roussel at cril dot univ-artois dot fr
2009-03-30 14:24 ` [Bug rtl-optimization/39588] [4.3/4.4/4.5 Regression] " rguenth at gcc dot gnu dot org
2009-03-30 14:56 ` rguenth at gcc dot gnu dot org
2009-03-31 15:44 ` ebotcazou at gcc dot gnu dot org
2009-04-01 20:46 ` ebotcazou at gcc dot gnu dot org
2009-04-01 20:48 ` ebotcazou at gcc dot gnu dot org
2009-04-01 20:48 ` ebotcazou at gcc dot gnu dot org
2009-04-01 20:51 ` ebotcazou at gcc dot gnu dot org

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).