[Webkit-unassigned] [Bug 193213] Fix Resource Timing buffer edge cases for WPT

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=193213

--- Comment #32 from Yoav Weiss <yoav at yoav.ws> ---
Comment on attachment 360345
  --> https://bugs.webkit.org/attachment.cgi?id=360345
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=360345&action=review

>> Source/WebCore/page/Performance.cpp:249
>> +        if (beforeCount <= afterCount) {
> 
> If this check is true, it seems we would be loosing some information, how big an issue is it?
> Should we do a smarter check here?
> Something like, continue if we ever queue any entry. Or is there a risk with that approach?

In the spec discussions, it was agreed (with rniwa) that this is the desired behaviour. If the event fired, but its handler did not increase the buffer size or cleared the entries, running the event again (because there are entries in the secondary buffer) is likely to result in an infinite loop.
So, in this rare case when there's a handler which doesn't make any room in the buffer, we will lose entries. That seems reasonable.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20190131/34aae3a9/attachment.html>