[PATCH] D11518: [DAGCombiner] Convert constant AND masks to shuffle clear masks down to the byte level

[Simon Pilgrim]

RKSimon created this revision.
RKSimon added reviewers: chandlerc, qcolombet, andreadb.
RKSimon added a subscriber: llvm-commits.
RKSimon set the repository for this revision to rL LLVM.

The XformToShuffleWithZero method currently checks AND masks at the per-lane level for all-one and all-zero constants and attempts to converts them to legal shuffle clear masks.

This patch generalises XformToShuffleWithZero, splitting and checking the sub-lanes of the constants down to the byte level to see if any legal shuffle clear masks are possible. This allows a lot of masks (often from legalization or truncation) to be folded into existing shuffle patterns and removes a lot of constant mask loading.

The patch involves a number of additional minor tweaks to improve codegen, I can commit these separately or generate patches for extra review if any of you wish:

- XformToShuffleWithZero is only attempted if constant folding has failed.
- A lot more X86 byte vector blends are now generated. I've added a stage to the VPBLENDVB lowering in lowerVectorShuffleAsBlend to attempt to lower back to a AND mask using lowerVectorShuffleAsBitMask if possible. VPAND is a lot faster than VPBLENDVB.
- X86 v8i16 shuffle lowering now attempts to use lowerVectorShuffleAsBitMask before resorting to VPSHUFB (matches v16i8 shuffle lowering). VPAND is a lot faster than VPSHUFB.

There are still a few examples of poor shuffle lowering that are exposed that we can cleanup in future patches (e.g. x86 legalized v8i8 zero extension uses PMOVZX+AND+AND instead of AND+PMOVZX)

Repository:
  rL LLVM

http://reviews.llvm.org/D11518

Files:
  lib/CodeGen/SelectionDAG/DAGCombiner.cpp
  lib/Target/X86/X86ISelLowering.cpp
  test/CodeGen/X86/avx2-conversions.ll
  test/CodeGen/X86/masked_memop.ll
  test/CodeGen/X86/sse2-vector-shifts.ll
  test/CodeGen/X86/sse3.ll
  test/CodeGen/X86/vec_cast2.ll
  test/CodeGen/X86/vec_int_to_fp.ll
  test/CodeGen/X86/vector-zext.ll
  test/CodeGen/X86/vselect-avx.ll
  test/CodeGen/X86/widen_load-2.ll

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D11518.30667.patch
Type: text/x-patch
Size: 63679 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20150726/c9eedeac/attachment.bin>