Add legalization patterns for fabs and fneg.

These sign bit manipulations need to use a -0.0 floating point constant
which we didn't have a way of materializing previously.

Add a ieee32.bits(0x...) syntax to the Python AST nodes that creates am
f32 immediate value with the exact requested bitwise representation.

cranelift/filetests/wasm/f32-arith.cton

@@ -17,8 +17,18 @@ ebb0:
 
 ; Unary operations
 
-; function %f32_abs(f32) -> f32
-; function %f32_neg(f32) -> f32
+function %f32_abs(f32) -> f32 {
+ebb0(v0: f32):
+    v1 = fabs v0
+    return v1
+}
+
+function %f32_neg(f32) -> f32 {
+ebb0(v0: f32):
+    v1 = fneg v0
+    return v1
+}
+
 ; function %f32_sqrt(f32) -> f32
 ; function %f32_ceil(f32) -> f32
 ; function %f32_floor(f32) -> f32

cranelift/filetests/wasm/f64-arith.cton

@@ -14,8 +14,18 @@ ebb0:
 
 ; Unary operations
 
-; function %f64_abs(f64) -> f64
-; function %f64_neg(f64) -> f64
+function %f64_abs(f64) -> f64 {
+ebb0(v0: f64):
+    v1 = fabs v0
+    return v1
+}
+
+function %f64_neg(f64) -> f64 {
+ebb0(v0: f64):
+    v1 = fneg v0
+    return v1
+}
+
 ; function %f64_sqrt(f64) -> f64
 ; function %f64_ceil(f64) -> f64
 ; function %f64_floor(f64) -> f64