[llvm] [AMDGPU] GFX12: select @llvm.prefetch intrinsic (PR #74576)

Matt Arsenault via llvm-commits llvm-commits at lists.llvm.org
Wed Dec 6 01:51:24 PST 2023


================
@@ -3101,6 +3101,24 @@ void AMDGPURegisterBankInfo::applyMappingImpl(
       applyDefaultMapping(OpdMapper);
       constrainOpWithReadfirstlane(B, MI, 8); // M0
       return;
+    case Intrinsic::prefetch: {
----------------
arsenm wrote:

This should have a G_ instruction.

Why erase here and not in the legalizer?

https://github.com/llvm/llvm-project/pull/74576


More information about the llvm-commits mailing list