[X86] AMD Zen 3: same-reg SSE XMM ANDNPS is a 1-cycle(!) dep-breaking zero-idiom
authorRoman Lebedev <lebedev.ri@gmail.com>
Fri, 14 May 2021 10:37:22 +0000 (13:37 +0300)
committerRoman Lebedev <lebedev.ri@gmail.com>
Fri, 14 May 2021 11:06:23 +0000 (14:06 +0300)
commitf38dcbecb643e30931b56bbcf37254477eac3977
tree11f38d0de44352521ad2f3711e2f6c11a8badb6b
parentc79c7bb980054fa7c1ebe5aae0e90755fe9a1314
[X86] AMD Zen 3: same-reg SSE XMM ANDNPS is a 1-cycle(!) dep-breaking zero-idiom

Same as SSE XMM XORPS/XORPD, it is not zero-cycle, even though it breaks the deps.
As confirmed by the exegesis measurements, and ref docs.
llvm/lib/Target/X86/X86ScheduleZnver3.td
llvm/test/tools/llvm-mca/X86/Znver3/zero-idioms-sse-xmm.s