Specify behavior of Math.pow

[ghost]

Library doc needs to specify the behavior of Math.pow, both for doubles and integers. In the absence of documentation Dart attempts to implement behavior defined in ECMA-262 15.8.2.13.

[lrhn]

The ES5 standard is pretty arbitrary. It doesn't match any of the three power functions in IEEE-754-2008.
I think we should just use the IEEE pow function:

For the pow function (integral exponents get special treatment):
 pow(x, ±0) is 1 for any x (even a zero, quiet NaN, or infinity)
 pow(±0, y) is ±∞ and signals the divideByZero exception for y an odd integer <0
 pow(±0, −∞) is +∞ with no exception
 pow(±0, +∞) is +0 with no exception
 pow(±0, y) is +∞ and signals the divideByZero exception for finite y<0 and not an odd integer
 pow(±0, y) is ±0 for finite y>0 an odd integer
 pow(±0, y) is +0 for finite y>0 and not an odd integer
 pow(−1, ±∞) is 1 with no exception
 pow(+1, y) is 1 for any y (even a quiet NaN)
 pow(x, y) signals the invalid operation exception for finite x<0 and finite non-integer y.

where the exception of the last line is handled by returning NaN.

This matches ECMAScript Math.pow except for the two next-to-last cases (-1 ** [-]Infinity, 1 ** anything), and it's likely to have an efficient implementation in native libraries.

---

cc @floitschG.
Set owner to @lrhn.

[floitschG]

Sounds good to me.
If I read this correctly we will guarantee an integer result for x:int, y:int and y>= 0. For all other input we return a double.

[lrhn]

Pow is now documented. It matches the cases above, and specifies some more special cases.
VM matches documentation, dart2js may not (it wasn't running lib/math tests at all).

[lrhn]

Added Fixed label.