[PATCH] D52347: [X86][BtVer2] Fix latency and resource cycles of AVX 256-bit zero-idioms.

[Andrea Di Biagio via Phabricator via llvm-commits]

andreadb created this revision.
andreadb added reviewers: RKSimon, spatel, craig.topper, mattd, courbet.
Herald added a subscriber: gbedwell.

This patch introduces a SchedWriteVariant to describe zero-idiom VXORP(S|D)Yrr and VANDNP(S|D)Yrr.

This is a follow-up of r342555.

On Jaguar, a VXORPSYrr is 2 macro opcodes. Only one opcode is eliminated at register-renaming stage. The other opcode has to be executed to set the upper half of the destination YMM.
Same for VANDNP(S|D)Yrr.


https://reviews.llvm.org/D52347

Files:
  lib/Target/X86/X86ScheduleBtVer2.td
  test/CodeGen/X86/avx-schedule.ll
  test/tools/llvm-mca/X86/BtVer2/zero-idioms-avx-256.s

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D52347.166424.patch
Type: text/x-patch
Size: 16561 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20180921/831a8ffb/attachment.bin>