[llvm] [AArch64] Make use of byte FPR stores for bytes extracted from vectors (PR #134117)

Paul Walker via llvm-commits llvm-commits at lists.llvm.org
Fri Apr 4 03:11:04 PDT 2025


================
@@ -338,6 +338,8 @@ def amdgpuBufferFatPointer : ValueType<160, 234>;
 // FIXME: Remove this and the getPointerType() override if MVT::i82 is added.
 def amdgpuBufferStridedPointer : ValueType<192, 235>;
 
+def vi8       : ValueType<8,  236>;  // 8-bit integer in FPR (AArch64)
----------------
paulwalker-arm wrote:

Working on the principle that we need a new MVT, let's go with `aarch64mfp8` to match the type we've added to clang.

https://github.com/llvm/llvm-project/pull/134117


More information about the llvm-commits mailing list