[Bug tree-optimization/43657] [4.3/4.4/4.5/4.6 Regression] -ftree-loop-linear causes FAIL: gcc.dg/vect/vect-cond-5.c execution test
aoliva at gcc dot gnu.org
gcc-bugzilla@gcc.gnu.org
Sun Jan 16 13:23:00 GMT 2011
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43657
--- Comment #6 from Alexandre Oliva <aoliva at gcc dot gnu.org> 2011-01-16 12:29:18 UTC ---
Created attachment 22982
--> http://gcc.gnu.org/bugzilla/attachment.cgi?id=22982
Proposed patch
This avoids the transformation, but... Is the test too strict? My reasoning
is that, since we're in closed form (right?), the exit block will only have
PHIs for stuff that is changed wthin the loop and is used outside the loop.
Does that sound sound? :-)
More information about the Gcc-bugs
mailing list