cranelift: Align Scalar and SIMD shift semantics (#4520)

* cranelift: Reorganize test suite

Group some SIMD operations by instruction.

* cranelift: Deduplicate some shift tests

Also, new tests with the mod behaviour

* aarch64: Lower shifts with mod behaviour

* x64: Lower shifts with mod behaviour

* wasmtime: Don't mask SIMD shifts

cranelift/filetests/filetests/runtests/simd-vselect.clif

@@ -72,3 +72,15 @@ block0(v0: b64x2, v1: i64x2, v2: i64x2):
     return v3
 }
 ; run: %vselect_p_i64x2([true false], [1 2], [100000000000 200000000000]) == [1 200000000000]
+
+
+function %vselect_i32x4(i32x4, i32x4) -> i32x4 {
+block0(v1: i32x4, v2: i32x4):
+    ; `make_trampoline` still does not know how to convert boolean vector types
+    ; so we load the value directly here.
+    v0 = vconst.b32x4 [true true false false]
+    v3 = vselect v0, v1, v2
+    return v3
+}
+; Remember that vselect accepts: 1) the selector vector, 2) the "if true" vector, and 3) the "if false" vector.
+; run: %vselect_i32x4([1 2 -1 -1], [-1 -1 3 4]) == [1 2 3 4]