[llvm] [llvm-exegesis] [AArch64] Resolving "not all operands are initialized by snippet generator" (PR #142529)

[David Green via llvm-commits]

================
@@ -1317,11 +1317,23 @@ def logical_vec_hw_shift : Operand<i32> {
 }
 
 // A vector move shifter operand:
-//  {0} - imm1: #8 or #16
-def move_vec_shift : Operand<i32> {
+//  {0} - imm1: #8
+def move_vec_shift_2s : Operand<i32> {
----------------
davemgreen wrote:

I don't think you need both of these. Correct me if I am wrong but MOVIv2s_msl and MOVIv4s_msl take 2 input operands - a i8 encoded value and an encoded shift amount that can be 272 (msl 16) or 264 (msl 8). The new OPERAND_MSL_SHIFT just needs to say that one of those two values is valid.

https://github.com/llvm/llvm-project/pull/142529