[Bug c/53153] New: ice in tree_low_cst, at tree.c:6569

[Bug c/53153] New: ice in tree_low_cst, at tree.c:6569

[dcb314 at hotmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53153

             Bug #: 53153
           Summary: ice in tree_low_cst, at tree.c:6569
    Classification: Unclassified
           Product: gcc
           Version: 4.8.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: dcb314@hotmail.com


Created attachment 27257
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=27257
C source code

I just tried to compile the package chktex-1.6.4-10
on gcc-4.8 trunk dated 20120426 on an AMD x86_64 box.

The compiler said

FindErrs.c: In function 'FindErr':
FindErrs.c:851:13: internal compiler error: in tree_low_cst, at tree.c:6569
             switch (Char)
             ^
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.

Preprocessed source code attached. Flag -O2 required.

[Bug c/53153] ice in tree_low_cst, at tree.c:6569

[marc.glisse at normalesup dot org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53153

Marc Glisse <marc.glisse at normalesup dot org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |marc.glisse at normalesup
                   |                            |dot org

--- Comment #1 from Marc Glisse <marc.glisse at normalesup dot org> 2012-04-28 09:11:45 UTC ---
Reduced:

void f (char *BufPtr) {
  int Char = *BufPtr;
  switch (Char) {
    case 'a':
    case 181:
    case ~(0xff & (~180)):
      PrintError();
  }
}

$ gcc a.c -c -O2
a.c: In function 'f':
a.c:3:3: internal compiler error: in tree_low_cst, at tree.c:6569
   switch (Char) {
   ^

The regression is recent.

[Bug c/53153] [4.8 Regression] ice in tree_low_cst, at tree.c:6569

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53153

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2012-04-28
   Target Milestone|---                         |4.8.0
            Summary|ice in tree_low_cst, at     |[4.8 Regression] ice in
                   |tree.c:6569                 |tree_low_cst, at
                   |                            |tree.c:6569
     Ever Confirmed|0                           |1
              Alias|                            |steven

--- Comment #2 from H.J. Lu <hjl.tools at gmail dot com> 2012-04-28 15:13:23 UTC ---
It is caused by revision 186579:

http://gcc.gnu.org/ml/gcc-cvs/2012-04/msg00530.html

[Bug c/53153] [4.8 Regression] ice in tree_low_cst, at tree.c:6569

[steven at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53153

Steven Bosscher <steven at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED
                 CC|steven at gcc dot gnu.org   |
         AssignedTo|unassigned at gcc dot       |steven at gcc dot gnu.org
                   |gnu.org                     |

[Bug c/53153] [4.8 Regression] ice in tree_low_cst, at tree.c:6569

[steven at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53153

--- Comment #3 from Steven Bosscher <steven at gcc dot gnu.org> 2012-04-30 17:11:15 UTC ---
This is the early forwprop eliminating the promotion cast from char to int. I
don't think this is a valid transformation.

[Bug c/53153] [4.8 Regression] ice in tree_low_cst, at tree.c:6569

[steven at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53153

--- Comment #4 from Steven Bosscher <steven at gcc dot gnu.org> 2012-04-30 21:43:33 UTC ---
Created attachment 27272
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=27272
Eliminate unreachable case labels

This is basically what the code in stmt.c used to do, and the new code in
gimplify.c also does: GIMPLE passes that change the type of the index
expression must also update the case label vector.

[Bug middle-end/53153] [4.8 Regression] ice in tree_low_cst, at tree.c:6569

[steven at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53153

--- Comment #5 from Steven Bosscher <steven at gcc dot gnu.org> 2012-05-02 12:57:17 UTC ---
Author: steven
Date: Wed May  2 12:57:10 2012
New Revision: 187048

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=187048
Log:
gcc/
    PR middle-end/53153
    * gimplify.c (preprocess_case_label_vec_for_gimple): New function,
    split out from ...
    (gimplify_switch_expr): ... here.
    * gimple.h (preprocess_case_label_vec_for_gimple): Add prototype.
    * tree-ssa-forwprop.c (simplify_gimple_switch_label_vec): New function
    to clean up case labels with values outside the index type range.
    (simplify_gimple_switch): Call it if something changed.
    Remove strange and unnecessary assert.

testsuite/
    PR middle-end/53153
    * gcc.dg/pr53153.c: New test.

Added:
    trunk/gcc/testsuite/gcc.dg/pr53153.c
Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/gimple.h
    trunk/gcc/gimplify.c
    trunk/gcc/testsuite/ChangeLog
    trunk/gcc/tree-ssa-forwprop.c

[Bug middle-end/53153] [4.8 Regression] ice in tree_low_cst, at tree.c:6569

[steven at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53153

Steven Bosscher <steven at gcc dot gnu.org> changed:

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

--- Comment #6 from Steven Bosscher <steven at gcc dot gnu.org> 2012-05-02 12:59:18 UTC ---
.