[llvm] [DAG] Remove OneUse restriction when folding (shl (add x, c1), c2) (PR #69105)

[Wang Pengcheng via llvm-commits]

================
@@ -7,14 +7,14 @@
 define <vscale x 4 x i1> @srem_eq_fold_nxv4i8(<vscale x 4 x i8> %va) {
 ; CHECK-LABEL: srem_eq_fold_nxv4i8:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    li a0, 42
+; CHECK-NEXT:    li a0, -85
 ; CHECK-NEXT:    vsetvli a1, zero, e8, mf2, ta, ma
-; CHECK-NEXT:    vmv.v.x v9, a0
-; CHECK-NEXT:    li a1, -85
-; CHECK-NEXT:    vmacc.vx v9, a1, v8
-; CHECK-NEXT:    vsll.vi v8, v9, 7
-; CHECK-NEXT:    vsrl.vi v9, v9, 1
-; CHECK-NEXT:    vor.vv v8, v9, v8
+; CHECK-NEXT:    vmul.vx v8, v8, a0
----------------
wangpc-pp wrote:

`vmv.v.x+vmacc.vx` to `vmul.vx+vadd.vx` may be a regression.

https://github.com/llvm/llvm-project/pull/69105