[llvm] [LowerMemIntrinsics] Avoid udiv/urem when type size is a power of 2 (PR #81238)

[Matt Arsenault via llvm-commits]

================
@@ -59,8 +59,9 @@ define void @issue63986(i64 %0, i64 %idxprom) {
 ; CHECK-NEXT:    flat_store_byte v[8:9], v25 offset:12
 ; CHECK-NEXT:    s_cbranch_vccz .LBB0_2
 ; CHECK-NEXT:  ; %bb.3: ; %loop-memcpy-residual-header
-; CHECK-NEXT:    s_and_b32 s4, 32, 15
-; CHECK-NEXT:    s_mov_b32 s5, 0
+; CHECK-NEXT:    s_mov_b32 s4, 0
+; CHECK-NEXT:    s_mov_b32 s5, s4
+; CHECK-NEXT:    s_cmp_lg_u64 s[4:5], 0
----------------
arsenm wrote:

Ok, separately we should have constant folding for addrspacecast implemented (at least as a DAG combine, ideally we would have some IR folding too) 

https://github.com/llvm/llvm-project/pull/81238