[clang] [WIP][AMDGPU] Support for type inferring image load/store builtins for AMDGPU (PR #140210)

[Matt Arsenault via cfe-commits]

================
@@ -112,11 +112,12 @@ bool SemaAMDGPU::CheckAMDGCNBuiltinFunctionCall(unsigned BuiltinID,
   case AMDGPU::BI__builtin_amdgcn_image_load_mip_3d_v4f16_i32:
   case AMDGPU::BI__builtin_amdgcn_image_load_mip_cube_v4f32_i32:
   case AMDGPU::BI__builtin_amdgcn_image_load_mip_cube_v4f16_i32: {
+    StringRef FeatureList(getASTContext().BuiltinInfo.getRequiredFeatures(BuiltinID));
     bool HasImageInsts = Builtin::evaluateRequiredTargetFeatures("image-insts", CallerFeatureMap);
-
-    if(!HasImageInsts){
+    if (!Builtin::evaluateRequiredTargetFeatures(
----------------
arsenm wrote:

I don't understand why this is calling evaluateRequiredTargetFeatures twice 

https://github.com/llvm/llvm-project/pull/140210