[Bug tree-optimization/47355] New: [4.6 Regression] ICE: verify_ssa failed with -O2 -fipa-cp-clone

[Bug tree-optimization/47355] New: [4.6 Regression] ICE: verify_ssa failed with -O2 -fipa-cp-clone

[roman at binarylife dot net]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=47355

           Summary: [4.6 Regression] ICE: verify_ssa failed with -O2
                    -fipa-cp-clone
           Product: gcc
           Version: 4.6.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: tree-optimization
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: roman@binarylife.net


Created attachment 23025
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=23025
"reduced" test case

$ g++ -c -std=c++0x -O2 -fipa-cp-clone test.cpp             
test.cpp: In function 'vector<green> function()':
test.cpp:200:15: error: definition in block 29 does not dominate use in block
28
test.cpp:200:15: error: no immediate_use list
for SSA_NAME: D.2314_94 in statement:
# .MEM_135 = VDEF <.MEM_109>
zfree (D.2314_94);
test.cpp:200:15: internal compiler error: verify_ssa failed
...

[Bug tree-optimization/47355] [4.6 Regression] ICE: verify_ssa failed with -O2 -fipa-cp-clone

[roman at binarylife dot net]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=47355

--- Comment #1 from Roman Kononov <roman at binarylife dot net> 2011-01-19 01:45:20 UTC ---
Created attachment 23026
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=23026
this is the real test case; ignore the other one please

[Bug tree-optimization/47355] [4.6 Regression] ICE: verify_ssa failed with -O2 -fipa-cp-clone

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=47355

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2011.01.19 03:59:10
                 CC|                            |jamborm at gcc dot gnu.org
   Target Milestone|---                         |4.6.0
     Ever Confirmed|0                           |1

--- Comment #2 from H.J. Lu <hjl.tools at gmail dot com> 2011-01-19 03:59:10 UTC ---
It is caused by revision 162911:

http://gcc.gnu.org/ml/gcc-cvs/2010-08/msg00122.html

[Bug tree-optimization/47355] [4.6 Regression] ICE: verify_ssa failed with -O2 -fipa-cp-clone

[jamborm at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=47355

--- Comment #3 from Martin Jambor <jamborm at gcc dot gnu.org> 2011-01-19 12:48:57 UTC ---
>From what I have seen, it's pass_cleanup_eh removing an SSA
(names$_storage_94) name that is still being referenced in the IL (as
a parameter to zfree).

[Bug tree-optimization/47355] [4.6 Regression] ICE: verify_ssa failed with -O2 -fipa-cp-clone

[zsojka at seznam dot cz]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=47355

--- Comment #4 from Zdenek Sojka <zsojka at seznam dot cz> 2011-01-19 13:27:34 UTC ---
Created attachment 23029
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=23029
more reduced testcase

$ gcc -O -fipa-cp -fipa-cp-clone testcase.C
testcase.C: In function 'void foo()':
testcase.C:30:1: error: definition in block 15 does not dominate use in block
14
testcase.C:30:1: error: no immediate_use list
for SSA_NAME: D.2189_30 in statement:
names$_storage_26 = D.2189_30 + D.2160_19;
testcase.C:30:1: internal compiler error: verify_ssa failed
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.

(r168850, x86_64-linux)

[Bug tree-optimization/47355] [4.6 Regression] ICE: verify_ssa failed with -O2 -fipa-cp-clone

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=47355

--- Comment #5 from Jakub Jelinek <jakub at gcc dot gnu.org> 2011-01-20 10:45:43 UTC ---
Created attachment 23042
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=23042
gcc46-pr47355.patch

Untested fix.  Seems ehcleanup assumes nop is used just in the following phi
and not somewhere else, because it doesn't attempt to rename it.  I'll see with
this patch how much it triggers during bootstraps/regtests.  We can't just
replace the other uses in this case, because it is changed in the loop.

[Bug tree-optimization/47355] [4.6 Regression] ICE: verify_ssa failed with -O2 -fipa-cp-clone

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=47355

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

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

[Bug tree-optimization/47355] [4.6 Regression] ICE: verify_ssa failed with -O2 -fipa-cp-clone

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=47355

--- Comment #6 from Jakub Jelinek <jakub at gcc dot gnu.org> 2011-01-21 15:15:42 UTC ---
Author: jakub
Date: Fri Jan 21 15:15:40 2011
New Revision: 169094

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=169094
Log:
    PR tree-optimization/47355
    * tree-eh.c (cleanup_empty_eh_merge_phis): Give up if
    NOP has non-debug uses beyond PHIs in new_bb.

    * g++.dg/opt/pr47355.C: New test.

Added:
    trunk/gcc/testsuite/g++.dg/opt/pr47355.C
Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/testsuite/ChangeLog
    trunk/gcc/tree-eh.c

[Bug tree-optimization/47355] [4.6 Regression] ICE: verify_ssa failed with -O2 -fipa-cp-clone

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=47355

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

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

--- Comment #7 from Jakub Jelinek <jakub at gcc dot gnu.org> 2011-01-21 15:39:49 UTC ---
Fixed.