https://gcc.gnu.org/bugzilla/show_bug.cgi?id=62283
--- Comment #29 from rguenther at suse dot de <rguenther at suse dot de> --- On Wed, 6 May 2015, ro at gcc dot gnu.org wrote: > https://gcc.gnu.org/bugzilla/show_bug.cgi?id=62283 > > --- Comment #28 from Rainer Orth <ro at gcc dot gnu.org> --- > Created attachment 35476 > --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=35476&action=edit > bb-slp-32.c.141t.slp2 dump > > A reghunt just confirmed that the patch also caused > > XPASS: gcc.dg/vect/bb-slp-32.c -flto -ffat-lto-objects scan-tree-dump slp2 > "vectorization is not profitable" > > Dump attached. Well - for vect_no_align && { ! vect_hw_misalign } it now tries to build the vector from scalar loads and that (rightfully) fails due to the cost model: Vector inside of basic block cost: 4 Vector prologue cost: 3 Vector epilogue cost: 0 Scalar cost of basic block: 4 so I bet we can now simply remove the XFAIL for all archs. Will do that.
