x64: Fix codegen for the select instruction with v128 (#4317)

This commit fixes a bug in the previous codegen for the `select`
instruction when the operations of the `select` were of the `v128` type.
Previously teh `XmmCmove` instruction only stored an `OperandSize` of 32
or 64 for a 64 or 32-bit move, but this was also used for these 128-bit
types which meant that when used the wrong move instruction was
generated. The fix applied here is to store the whole `Type` being moved
so the 128-bit variant can be selected as well.

cranelift/codegen/src/isa/x64/inst/emit.rs

@@ -1112,7 +1112,7 @@ pub(crate) fn emit(
         }
 
         Inst::XmmCmove {
-            size,
+            ty,
             cc,
             consequent,
             alternative,
@@ -1130,10 +1130,15 @@ pub(crate) fn emit(
             // Jump if cc is *not* set.
             one_way_jmp(sink, cc.invert(), next);
 
-            let op = if *size == OperandSize::Size64 {
-                SseOpcode::Movsd
-            } else {
-                SseOpcode::Movss
+            let op = match *ty {
+                types::F64 => SseOpcode::Movsd,
+                types::F32 => SseOpcode::Movsd,
+                types::F32X4 => SseOpcode::Movaps,
+                types::F64X2 => SseOpcode::Movapd,
+                ty => {
+                    debug_assert!(ty.is_vector() && ty.bytes() == 16);
+                    SseOpcode::Movdqa
+                }
             };
             let inst = Inst::xmm_unary_rm_r(op, consequent, Writable::from_reg(dst));
             inst.emit(&[], sink, info, state);