https://bugs.llvm.org/show_bug.cgi?id=46249
Simon Pilgrim <llvm-...@redking.me.uk> changed:
What |Removed |Added
----------------------------------------------------------------------------
Fixed By Commit(s)| |e7e204a373683a09978be2b1df8
| |783ba26fe43a4
Resolution|--- |FIXED
Status|NEW |RESOLVED
--- Comment #3 from Simon Pilgrim <llvm-...@redking.me.uk> ---
Fixed in rGe7e204a373683a09978be2b1df8783ba26fe43a4
Looking at the reported test code in https://t.co/o5DY0C0Z0e?amp=1
we still screw up some predicate selects that we've widened from v16i32 to
v8i64 types:
movb $-52, %al
kmovw %eax, %k3
vpmaxsd %zmm2, %zmm0, %zmm0
vmovdqa64 %zmm0, %zmm1 {%k2}
which gcc avoids, allowing to fold the select into the vpmaxsd.
--
You are receiving this mail because:
You are on the CC list for the bug._______________________________________________
llvm-bugs mailing list
llvm-bugs@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-bugs
