[webkit-reviews] review denied: [Bug 67980] [Chromium/Mac] fast/events/constructors/progress-event-constructor.html is failing : [Attachment 112459] Patch

[bugzilla-daemon at webkit.org]

Adam Barth <abarth at webkit.org> has denied Kentaro Hara <haraken at chromium.org>'s
request for review:
Bug 67980: [Chromium/Mac]
fast/events/constructors/progress-event-constructor.html is failing
https://bugs.webkit.org/show_bug.cgi?id=67980

Attachment 112459: Patch
https://bugs.webkit.org/attachment.cgi?id=112459&action=review

------- Additional Comments from Adam Barth <abarth at webkit.org>
View in context: https://bugs.webkit.org/attachment.cgi?id=112459&action=review


> Source/WebCore/bindings/v8/OptionsObject.cpp:238
> +	   // Calculate d % 2^{64}.
> +	   // -2^{64} < fmodValue < 2^{64}.
> +	   double fmodValue = fmod(trunc(d), std::numeric_limits<unsigned long
long>::max() + 1.0);
> +	   if (fmodValue >= 0) {
> +	       // 0 <= fmodValue < 2^{64}.
> +	       // 0 <= value < 2^{64}. This cast causes no loss.
> +	       value = static_cast<unsigned long long>(fmodValue);
> +	   } else {
> +	       // -2^{64} < fmodValue < 0.
> +	       // 0 < fmodValueInUnsignedLongLong < 2^{64}. This cast causes no
loss.
> +	       unsigned long long fmodValueInUnsignedLongLong =
static_cast<unsigned long long>(-fmodValue);
> +	       // -1 < (std::numeric_limits<unsigned long long>::max() -
fmodValueInUnsignedLongLong) < 2^{64} - 1.
> +	       // 0 < value < 2^{64}.
> +	       value = std::numeric_limits<unsigned long long>::max() -
fmodValueInUnsignedLongLong + 1;
> +	   }

This looks cool, but it should be factored into a more general location so that
folks can re-use it.  If it's V8-specific logic, it should probably go in
V8Utilities.  If it's more general than V8, it should go in WTF somewhere.