py/objfloat: Fix abs(-0.0) so it returns 0.0.

Nan and inf (signed and unsigned) are also handled correctly by using
signbit (they were also handled correctly with "val<0", but that didn't
handle -0.0 correctly).  A test case is added for this behaviour.

tests/float/builtin_float_abs.py

@@ -0,0 +1,13 @@
+# test builtin abs function with float args
+
+for val in (
+    '1.0',
+    '-1.0',
+    '0.0',
+    '-0.0',
+    'nan',
+    '-nan',
+    'inf',
+    '-inf',
+    ):
+    print(val, abs(float(val)))