[llvm] [AMDGPU][DAG] Remove AssertZext before some intrinsics (PR #146052)

[Jay Foad via llvm-commits]

================
@@ -3985,6 +3985,24 @@ SDValue AMDGPUTargetLowering::performAssertSZExtCombine(SDNode *N,
     }
   }
 
+  // AssertZext in front of these intrinsics is not necessary, the lowering of
+  // the intrinsics into a register read will insert one if it is needed.
----------------
jayfoad wrote:

> Would it be better if I added a AMDGPU specific combine to simplify and+assert_zext instead?

Why would it have to be AMDGPU-specific? Why isn't some generic KnownBits-based optimization handling this already?

https://github.com/llvm/llvm-project/pull/146052