[PATCH] D103170: [CodeGen][AArch64][SVE] Use ld1r[bhsd] for vector splat from memory

[Peter Waller via Phabricator via llvm-commits]

peterwaller-arm updated this revision to Diff 355832.
peterwaller-arm edited the summary of this revision.
peterwaller-arm added a comment.

Apologies for delayed response. Should be much faster on further updates.

- Address most review comments.
  - Fix the issue of stack objects.
  - Force offset to zero in SelectAddrModeIndexedUImm.
- Add a mir test for frame lowering, which crashes on main and passes here.
- Drop addition of azextload and use more specific patterns instead. I couldn't find a sense in which they were necessary.


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D103170/new/

https://reviews.llvm.org/D103170

Files:
  llvm/lib/Target/AArch64/AArch64ISelDAGToDAG.cpp
  llvm/lib/Target/AArch64/AArch64InstrFormats.td
  llvm/lib/Target/AArch64/AArch64InstrInfo.cpp
  llvm/lib/Target/AArch64/AArch64SVEInstrInfo.td
  llvm/test/CodeGen/AArch64/sve-ld-post-inc.ll
  llvm/test/CodeGen/AArch64/sve-ld1r.ll
  llvm/test/CodeGen/AArch64/sve-ld1r.mir

-------------- next part --------------
A non-text attachment was scrubbed...
Name: D103170.355832.patch
Type: text/x-patch
Size: 51659 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20210701/03880afd/attachment.bin>