[llvm] [AArch64][SME2] Add stridedorcontiguous pseudos for multi-vector-stores (PR #211551)

[Benjamin Maxwell via llvm-commits]

================
@@ -205,6 +364,25 @@ define void @st1_x4_i32(<vscale x 16 x i8> %unused, <vscale x 4 x i32> %zn0, <vs
 ; CHECK-NEXT:    addvl sp, sp, #1
 ; CHECK-NEXT:    ldr x29, [sp], #16 // 8-byte Folded Reload
 ; CHECK-NEXT:    ret
+;
----------------
MacDue wrote:

I think we need to look at some slightly less artificial cases for the hints too. For this case, it'd be better (as it's a very simple function with no loops etc), but that might be different if this is within a deeply nested loop etc. 

https://github.com/llvm/llvm-project/pull/211551