https://gcc.gnu.org/bugzilla/show_bug.cgi?id=62283
--- Comment #15 from rguenther at suse dot de <rguenther at suse dot de> --- On Mon, 1 Sep 2014, ro at CeBiTec dot Uni-Bielefeld.DE wrote: > https://gcc.gnu.org/bugzilla/show_bug.cgi?id=62283 > > --- Comment #14 from ro at CeBiTec dot Uni-Bielefeld.DE <ro at CeBiTec dot > Uni-Bielefeld.DE> --- > > --- Comment #13 from Richard Biener <rguenth at gcc dot gnu.org> --- > > (In reply to Rainer Orth from comment #12) > >> Created attachment 33426 [details] > >> vect dump > >> > >> Sure, I meant to but forgot... > > > > Ok: > > > > /vol/gcc/src/hg/trunk/local/gcc/testsuite/gcc.dg/vect/vect-33.c:19:3: note: > > vectorization factor = 8 > > > > so we have a vector of 8 chars, v8qi but vect64 is not enabled for SPARC? > > (target-supports.exp). The { ! vect64 } / && vect64 cases were supposed > > to catch the case where vectorization uses v8qi (instead of the more > > common v16qi). > > > > So ... can you test enabling vect64 for sparc? > > While it works for the test at hand, the two other tests using vect64 > (bb-slp-11.c and bb-slp-26.c) now FAIL: > > FAIL: gcc.dg/vect/bb-slp-11.c scan-tree-dump-times slp2 "basic block > vectorized" 1 > FAIL: gcc.dg/vect/bb-slp-26.c scan-tree-dump-times slp1 "basic block > vectorized" 1 I wonder about the dump files for those... also if they fail on the other vect64 target (arm-little-endian with NEON). Richard.
