Intel encodings for nearest/floor/ceil/trunc.

These floating point rounding operations all use the roundss/roundsd
instructions that are available in SSE 4.1.

cranelift/filetests/wasm/f32-arith.cton

@@ -35,10 +35,29 @@ ebb0(v0: f32):
     return v1
 }
 
-; function %f32_ceil(f32) -> f32
-; function %f32_floor(f32) -> f32
-; function %f32_trunc(f32) -> f32
-; function %f32_nearest (f32) -> f32
+function %f32_ceil(f32) -> f32 {
+ebb0(v0: f32):
+    v1 = ceil v0
+    return v1
+}
+
+function %f32_floor(f32) -> f32 {
+ebb0(v0: f32):
+    v1 = floor v0
+    return v1
+}
+
+function %f32_trunc(f32) -> f32 {
+ebb0(v0: f32):
+    v1 = trunc v0
+    return v1
+}
+
+function %f32_nearest (f32) -> f32 {
+ebb0(v0: f32):
+    v1 = nearest v0
+    return v1
+}
 
 ; Binary Operations
 

cranelift/filetests/wasm/f64-arith.cton

@@ -32,10 +32,29 @@ ebb0(v0: f64):
     return v1
 }
 
-; function %f64_ceil(f64) -> f64
-; function %f64_floor(f64) -> f64
-; function %f64_trunc(f64) -> f64
-; function %f64_nearest (f64) -> f64
+function %f64_ceil(f64) -> f64 {
+ebb0(v0: f64):
+    v1 = ceil v0
+    return v1
+}
+
+function %f64_floor(f64) -> f64 {
+ebb0(v0: f64):
+    v1 = floor v0
+    return v1
+}
+
+function %f64_trunc(f64) -> f64 {
+ebb0(v0: f64):
+    v1 = trunc v0
+    return v1
+}
+
+function %f64_nearest (f64) -> f64 {
+ebb0(v0: f64):
+    v1 = nearest v0
+    return v1
+}
 
 ; Binary Operations