[llvm-branch-commits] [llvm] [AMDGPU] Make <2 x bfloat> fneg legal (PR #142870)

[Matt Arsenault via llvm-branch-commits]

================
@@ -1835,6 +1835,11 @@ def : GCNPat <
   (S_XOR_B32 SReg_32:$src, (S_MOV_B32 (i32 0x80008000)))
 >;
 
+def : GCNPat <
----------------
arsenm wrote:

Can you do the foreach over the types with the v2f16 case 

https://github.com/llvm/llvm-project/pull/142870