At ply == OnePly (common case) we avoid some useless
floating point computation.
No functional change.
Signed-off-by: Marco Costalba <mcostalba@gmail.com>
// red = baseReduction + ln(moveCount) * ln(depth / 2) / reductionInhibitor;
//
logLimit = depth > OnePly ? (1.0 - baseReduction) * reductionInhibitor / ln(depth / 2) : 1000.0;
- gradient = ln(depth / 2) / reductionInhibitor;
+ gradient = depth > OnePly ? ln(depth / 2) / reductionInhibitor : 0.0;
}