[llvm] AMDGPU: Use pattern to select instruction for intrinsic llvm.fptrunc.round (PR #105761)

[Matt Arsenault via llvm-commits]

================
@@ -297,7 +297,7 @@ def : GINodeEquiv<G_AMDGPU_S_BUFFER_LOAD_UBYTE, SIsbuffer_load_ubyte>;
 def : GINodeEquiv<G_AMDGPU_S_BUFFER_LOAD_SSHORT, SIsbuffer_load_short>;
 def : GINodeEquiv<G_AMDGPU_S_BUFFER_LOAD_USHORT, SIsbuffer_load_ushort>;
 
-def : GINodeEquiv<G_FPTRUNC_ROUND, SIfptrunc_round>;
+def : GINodeEquiv<G_INTRINSIC_FPTRUNC_ROUND, SIfptrunc_round>;
----------------
arsenm wrote:

Apparently it's not defined. I don't know why we have SIfptrunc_round, it should just be defined in TargetSelectionDAG.td without the target prefix 

https://github.com/llvm/llvm-project/pull/105761