[Bug tree-optimization/58010] [4.8/4.9 Regression] ICE in vect_create_epilog_for_reduction, at tree-vect-loop.c:4378

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58010

--- Comment #7 from Richard Biener  ---
Author: rguenth
Date: Fri Aug 30 07:47:54 2013
New Revision: 202095

URL: http://gcc.gnu.org/viewcvs?rev=202095&root=gcc&view=rev
Log:
2013-08-30  Richard Biener  

PR tree-optimization/58010
* tree-vect-loop.c (vect_create_epilog_for_reduction): Remove
assert that we have a loop-closed PHI.

* gcc.dg/pr58010.c: New testcase.

Added:
trunk/gcc/testsuite/gcc.dg/pr58010.c
Modified:
trunk/gcc/ChangeLog
trunk/gcc/testsuite/ChangeLog
trunk/gcc/tree-vect-loop.c

[Bug tree-optimization/58010] [4.8/4.9 Regression] ICE in vect_create_epilog_for_reduction, at tree-vect-loop.c:4378

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58010

--- Comment #6 from Richard Biener  ---
In the loop in question we have a dead scalar cycle for which we do not
insert loop-closed PHI nodes.  Hm.

The correct fix looks easy though - simply remove the assert.  We'll vectorize
dead code, but who cares.

Testing that.

[Bug tree-optimization/58010] [4.8/4.9 Regression] ICE in vect_create_epilog_for_reduction, at tree-vect-loop.c:4378

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58010

Richard Biener  changed:

   What|Removed |Added

 Status|NEW |ASSIGNED
   Assignee|unassigned at gcc dot gnu.org  |rguenth at gcc dot 
gnu.org

--- Comment #5 from Richard Biener  ---
Mine.

[Bug tree-optimization/58010] [4.8/4.9 Regression] ICE in vect_create_epilog_for_reduction, at tree-vect-loop.c:4378

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58010

Jakub Jelinek  changed:

   What|Removed |Added

 CC||jakub at gcc dot gnu.org

--- Comment #4 from Jakub Jelinek  ---
With -O2 -funswitch-loops -fvectorize -fno-vect-const-model it started with
r178728 .
Richard, when you are back, can you please look at this?

[Bug tree-optimization/58010] [4.8/4.9 Regression] ICE in vect_create_epilog_for_reduction, at tree-vect-loop.c:4378

[wschmidt at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58010

--- Comment #3 from Bill Schmidt  ---
r189527 is probably a red herring.  That just changed the cost model to be
turned on by default at -O3.  Somebody who's actively working on the vectorizer
should probably have a look at this.

If you want to narrow it down further, try turning on -fvect-cost-model and
bisecting.  The problem will probably show up as being introduced further back
if you do this.

This also suggests a workaround for the OP of using -fno-vect-cost-model, if a
workaround is needed.

[Bug tree-optimization/58010] [4.8/4.9 Regression] ICE in vect_create_epilog_for_reduction, at tree-vect-loop.c:4378

[mpolacek at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58010

Marek Polacek  changed:

   What|Removed |Added

 CC||wschmidt at gcc dot gnu.org

--- Comment #2 from Marek Polacek  ---
Started with http://gcc.gnu.org/r189527

[Bug tree-optimization/58010] [4.8/4.9 Regression] ICE in vect_create_epilog_for_reduction, at tree-vect-loop.c:4378

[mpolacek at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58010

Marek Polacek  changed:

   What|Removed |Added

 Status|UNCONFIRMED |NEW
   Last reconfirmed||2013-08-02
 CC||mpolacek at gcc dot gnu.org
  Known to work||4.7.3
   Target Milestone|--- |4.8.2
Summary|ICE in  |[4.8/4.9 Regression] ICE in
   |vect_create_epilog_for_redu |vect_create_epilog_for_redu
   |ction, at   |ction, at
   |tree-vect-loop.c:4378   |tree-vect-loop.c:4378
 Ever confirmed|0   |1
  Known to fail||4.8.1, 4.9.0

--- Comment #1 from Marek Polacek  ---
Confirmed.