re PR tree-optimization/60656 (x86 vectorization produces wrong code)
authorCong Hou <congh@google.com>
Sat, 5 Apr 2014 01:27:21 +0000 (21:27 -0400)
committerCong Hou <congh@gcc.gnu.org>
Sat, 5 Apr 2014 01:27:21 +0000 (21:27 -0400)
commitebc047a210e17c5facf31ff11eddaac9db4d4a95
treed1e18f32f3b8c5ae7bca649f5c026af77d3cd5ed
parent604fb578094200cc3771cc84c267cffa2fc47a6d
re PR tree-optimization/60656 (x86 vectorization produces wrong code)

2014-04-04  Cong Hou  <congh@google.com>

    PR tree-optimization/60656
    * tree-vect-stmts.c (supportable_widening_operation):
      Fix a bug that elements in a vector with vect_used_by_reduction
      property are incorrectly reordered when the operation on it is not
      consistant with the one in reduction operation.

2014-04-04  Cong Hou  <congh@google.com>

    PR tree-optimization/60656
    * gcc.dg/vect/pr60656.c: New test.

From-SVN: r209138
gcc/ChangeLog
gcc/testsuite/ChangeLog
gcc/testsuite/gcc.dg/vect/pr60656.c [new file with mode: 0644]
gcc/tree-vect-stmts.c