https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70729
--- Comment #22 from rguenther at suse dot de <rguenther at suse dot de> --- On Tue, 7 Jun 2016, ysrumyan at gmail dot com wrote: > https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70729 > > --- Comment #21 from Yuri Rumyantsev <ysrumyan at gmail dot com> --- > Richard! > > Are you planning to prepare the second part of the patch (zeroing safelen and > testing it in loop invariant motion phase as you proposed)? I'm not exactly sure when zeroing safelen is required. We do use safelen for #pragma GCC ivdep as well I think. That is, if the loop does not use GOMP_SIMD_LANE is it still required to zero safelen? That is, for example I don't see any GOMP_SIMD_LANE stuff in gcc.dg/vect/vect-simd-clone-5.c but it was failing. Thus - can you take the 2nd part and propose a patch please?
