[clang] [llvm] [AMDGPU] Add dot product patterns with saturating add (clamp) (PR #187945)

[via llvm-commits]

================
@@ -728,13 +728,49 @@ defm V_DOT4_F32_BF8_BF8 : VOP3PDOTF8Inst<"v_dot4_f32_bf8_bf8", int_amdgcn_dot4_f
 def : UDot2Pat<V_DOT2_U32_U16>;
 def : SDot2Pat<V_DOT2_I32_I16>;
 
+// Saturating unsigned dot2 pattern: uaddsat(a[0]*b[0] + a[1]*b[1], c)
----------------
addmisol wrote:

the UDot2SatPat and the SDot2SatPat Tablegen patterns are the only way dot2 saturating cases get matched.

https://github.com/llvm/llvm-project/pull/187945