[llvm] [AMDGPU] Overload resource descriptor in image intrinsics. (PR #107255)

Jay Foad via llvm-commits llvm-commits at lists.llvm.org
Fri Sep 27 02:22:40 PDT 2024


================
@@ -8209,9 +8209,17 @@ SDValue SITargetLowering::lowerImage(SDValue Op,
     append_range(Ops, VAddrs);
   else
     Ops.push_back(VAddr);
-  Ops.push_back(Op.getOperand(ArgOffset + Intr->RsrcIndex));
-  if (BaseOpcode->Sampler)
-    Ops.push_back(Op.getOperand(ArgOffset + Intr->SampIndex));
+  SDValue Rsrc = Op.getOperand(ArgOffset + Intr->RsrcIndex);
+  EVT VT = Rsrc.getValueType();
----------------
jayfoad wrote:

Nit: I'd prefer RsrcVT since there are lots of different VTs in this function.

https://github.com/llvm/llvm-project/pull/107255


More information about the llvm-commits mailing list