Implement Shuffle for the interpreter

Implemented `Shuffle` for the Cranelift interpreter, to shuffle two SIMD
vectors together based on an immediate mask of 16 bytes.

Copyright (c) 2021, Arm Limited

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

@@ -0,0 +1,12 @@
+test interpret
+test run
+target aarch64
+set enable_simd
+target x86_64
+
+function %shuffle_i8x16(i8x16, i8x16) -> i8x16 {
+block0(v0: i8x16, v1: i8x16):
+    v2 = shuffle v0, v1, [3 0 31 26 100 6 12 11 23 13 24 4 2 15 17 5]
+    return v2
+}
+; run: %shuffle_i8x16([1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16], [17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32]) == [4 1 32 27 0 7 13 12 24 14 25 5 3 16 18 6]

cranelift/interpreter/src/step.rs

@@ -776,7 +776,30 @@ where
             arg(0)?,
             ValueConversionKind::RoundNearestEven(ctrl_ty),
         )?),
-        Opcode::Shuffle => unimplemented!("Shuffle"),
+        Opcode::Shuffle => {
+            if let InstructionData::Shuffle { mask, .. } = inst {
+                let mask = state
+                    .get_current_function()
+                    .dfg
+                    .immediates
+                    .get(mask)
+                    .unwrap()
+                    .as_slice();
+                let a = Value::into_array(&arg(0)?)?;
+                let b = Value::into_array(&arg(1)?)?;
+                let mut new = [0u8; 16];
+                for i in 0..mask.len() {
+                    if (mask[i] as usize) < a.len() {
+                        new[i] = a[mask[i] as usize];
+                    } else if (mask[i] as usize - a.len()) < b.len() {
+                        new[i] = b[mask[i] as usize - a.len()];
+                    } // else leave as 0.
+                }
+                assign(Value::vector(new, ctrl_ty)?)
+            } else {
+                unreachable!();
+            }
+        }
         Opcode::Swizzle => {
             let x = Value::into_array(&arg(0)?)?;
             let s = Value::into_array(&arg(1)?)?;