[PATCH] D83395: [SVE] Code generation for fixed length vector truncates.

[Paul Walker via Phabricator via llvm-commits]

paulwalker-arm marked an inline comment as done.
paulwalker-arm added inline comments.


================
Comment at: llvm/lib/Target/AArch64/AArch64ISelLowering.cpp:1063
   setOperationAction(ISD::STORE, VT, Custom);
+  setOperationAction(ISD::TRUNCATE, VT, Custom);
 }
----------------
efriedma wrote:
> This specifically applies to the result type.  You might want to note that you're implicitly depending on the fact that we do custom legalization for NEON TRUNCATE operations for other reasons.
I wondered about that.  Do you think it would be better if I was just explicit and add the necessary setOperation calls even though they're duplicates?


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D83395/new/

https://reviews.llvm.org/D83395