[Bug middle-end/43057] [LTO] fold check: original tree changed by fold

[Bug middle-end/43057] [LTO] fold check: original tree changed by fold

[d.g.gorbachev at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43057

Dmitry Gorbachev <d.g.gorbachev at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
      Known to fail|                            |4.5.2, 4.6.0

--- Comment #5 from Dmitry Gorbachev <d.g.gorbachev at gmail dot com> 2010-11-20 14:19:44 UTC ---
Still fails with GCC 4.5.2 and 4.6.0.

[Bug middle-end/43057] [LTO] fold check: original tree changed by fold

[ubizjak at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43057

Uros Bizjak <ubizjak at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2010.11.20 16:13:06
     Ever Confirmed|0                           |1

--- Comment #6 from Uros Bizjak <ubizjak at gmail dot com> 2010-11-20 16:13:06 UTC ---
Following untested patch fixes the failure:

Index: fold-const.c
===================================================================
--- fold-const.c    (revision 166974)
+++ fold-const.c    (working copy)
@@ -13415,7 +13415,7 @@ fold_ternary_loc (location_t loc, enum t
          a COND, which will recurse.  In that case, the COND_EXPR
          is probably the best choice, so leave it alone.  */
       && type == TREE_TYPE (arg0))
-    return pedantic_non_lvalue_loc (loc, arg0);
+    return pedantic_non_lvalue_loc (loc, unshare_expr (arg0));

       /* Convert A ? 0 : 1 to !A.  This prefers the use of NOT_EXPR
      over COND_EXPR in cases such as floating point comparisons.  */

[Bug middle-end/43057] [LTO] fold check: original tree changed by fold

[uros at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43057

--- Comment #7 from uros at gcc dot gnu.org 2010-11-22 16:14:33 UTC ---
Author: uros
Date: Mon Nov 22 16:14:27 2010
New Revision: 167037

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=167037
Log:
    PR middle-end/43057
    * fold-const.c (pedantic_non_lvalue_loc): Unshare x before
    setting location.


Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/fold-const.c

[Bug middle-end/43057] [LTO] fold check: original tree changed by fold

[uros at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43057

--- Comment #8 from uros at gcc dot gnu.org 2010-11-22 18:03:05 UTC ---
Author: uros
Date: Mon Nov 22 18:02:58 2010
New Revision: 167042

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=167042
Log:
    PR middle-end/43057
    * fold-const.c (pedantic_non_lvalue_loc): Unshare x before
    setting location.


Modified:
    branches/gcc-4_5-branch/gcc/ChangeLog
    branches/gcc-4_5-branch/gcc/fold-const.c

[Bug middle-end/43057] [LTO] fold check: original tree changed by fold

[ubizjak at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43057

Uros Bizjak <ubizjak at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED
   Target Milestone|---                         |4.5.2

--- Comment #9 from Uros Bizjak <ubizjak at gmail dot com> 2010-11-22 18:19:55 UTC ---
Fixed.

FWIW, following this patch, at least mainline can be bootstrapped with
--enable-checking=all; I didn't bootstrap 4.5 since it takes lots of time.

[Bug middle-end/43057] [LTO] fold check: original tree changed by fold

[d.g.gorbachev at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43057

--- Comment #10 from Dmitry Gorbachev <d.g.gorbachev at gmail dot com> 2010-11-22 19:58:56 UTC ---
*** Bug 43316 has been marked as a duplicate of this bug. ***

[Bug lto/43057] New: [LTO] fold check: original tree changed by fold

[d dot g dot gorbachev at gmail dot com]

GCC 4.5.0 20100211. Configured with --enable-languages=c --enable-checking=all
--disable-bootstrap.


-- 
           Summary: [LTO] fold check: original tree changed by fold
           Product: gcc
           Version: 4.5.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: lto
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: d dot g dot gorbachev at gmail dot com
 GCC build triplet: i686-pc-linux-gnu
  GCC host triplet: i686-pc-linux-gnu
GCC target triplet: i686-pc-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43057

[Bug middle-end/43057] [LTO] fold check: original tree changed by fold

[zsojka at seznam dot cz]

------- Comment #3 from zsojka at seznam dot cz  2010-02-27 14:02 -------
Testcase fails when compiled as C++ code too (without -flto):

$ /mnt/sdb1/build-157106-checking-fold/gcc/cc1plus -O bug.c
 int main()
Analyzing compilation unit
Performing interprocedural optimizations
 <visibility> <*free_lang_data> <early_local_cleanups> <whole-program> <inline>
<static-var> <pure-const>Assembling functions:
 int main()
bug.c: In function &#8216;int main()&#8217;:
bug.c:3:5: internal compiler error: fold check: original tree changed by fold
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43057

[Bug middle-end/43057] [LTO] fold check: original tree changed by fold

[d dot g dot gorbachev at gmail dot com]

------- Comment #4 from d dot g dot gorbachev at gmail dot com  2010-03-10 00:17 -------
Created an attachment (id=20071)
 --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=20071&action=view)
Another testcase


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43057