[PATCH] D66489: [X86] Add a DAG combine to transform (i8 (bitcast (v8i1 (extract_subvector (v16i1 X), 0)))) -> (i8 (trunc (i16 (bitcast (v16i1 X))))) on KNL target

[Craig Topper via Phabricator via llvm-commits]

craig.topper created this revision.
craig.topper added reviewers: RKSimon, spatel.
Herald added a subscriber: hiraditya.
Herald added a project: LLVM.

Without AVX512DQ we don't have KMOVB so we can't really copy 8-bits of a k-register to a GPR. We have to copy 16 bits instead. We do this even if the DAG copy is from v8i1->v16i1. If we detect the (i8 (bitcast (v8i1 (extract_subvector (v16i1 X), 0)))) we should rewrite the types to match the copy we do support. By doing this, we can help known bits to propagate without losing the upper 8 bits of the input to the extract_subvector. This allows some zero extends to be removed since we have an isel pattern to use kmovw for (zero_extend (i16 (bitcast (v16i1 X))).


https://reviews.llvm.org/D66489

Files:
  llvm/lib/Target/X86/X86ISelLowering.cpp
  llvm/test/CodeGen/X86/avx512-intrinsics.ll
  llvm/test/CodeGen/X86/avx512vl-intrinsics-upgrade.ll
  llvm/test/CodeGen/X86/avx512vl-vec-masked-cmp.ll
  llvm/test/CodeGen/X86/avx512vlcd-intrinsics-fast-isel.ll
  llvm/test/CodeGen/X86/win32-eh-available-externally.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D66489.216207.patch
Type: text/x-patch
Size: 48622 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20190820/6e6cd3d0/attachment.bin>