https://gcc.gnu.org/bugzilla/show_bug.cgi?id=92048
Richard Biener <rguenth at gcc dot gnu.org> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |ASSIGNED Last reconfirmed| |2019-10-15 Assignee|unassigned at gcc dot gnu.org |rguenth at gcc dot gnu.org Ever confirmed|0 |1 --- Comment #1 from Richard Biener <rguenth at gcc dot gnu.org> --- Wonder how this just made a difference on armeb ... we probably now unroll the inner loop. On x86 we then vectorize the outer loop just fine. Ah, when unrolled appearantly we run into fast-math-vect-pr29925.c:16:20: missed: not vectorized: unsupported unaligned load: *_31 since we didn't unroll in the past just disabling unrolling for the testcase is the way to go.