[llvm] [AMDGPU][True16][MC] Support v_swap_b16. (PR #100442)

[Georg Lehmann via llvm-commits]

================
@@ -729,7 +729,20 @@ def V_ACCVGPR_MOV_B32 : VOP1_Pseudo<"v_accvgpr_mov_b32", VOPProfileAccMov, [], 1
   let isAsCheapAsAMove = 1;
 }
 
+def VOP_SWAP_I16 : VOPProfile_True16<VOP_I16_I16> {
+  let Outs32 = (outs VOPDstOperand_t16Lo128:$vdst,
+                     VOPSrcEncodedDstOperand_t16Lo128:$vdst1);
+  let Ins32 = (ins VOPSrcEncodedDstOperand_t16Lo128:$src0,
+                   VOPDstOperand_t16Lo128:$src1);
+  let Asm32 = " $vdst, $src0";
+}
+
 let SubtargetPredicate = isGFX11Plus in {
+  def V_SWAP_B16 : VOP1_Pseudo<"v_swap_b16", VOP_SWAP_I16, [], /* VOP1Only= */ 1> {
----------------
DadSchoorse wrote:

Are you sure this opcode is VOP1 only? When I implemented support for `v_swap_b16` in our compiler, I tested it with VOP3 encoding and it worked fine. VOP3 v_swap_b16 is nessecary to address v128-v255.

https://github.com/llvm/llvm-project/pull/100442