[Webkit-unassigned] [Bug 103867] [Resource Timing] Record redirectStart & redirectEnd time and expose in resource timing entries

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Tue Jan 15 21:32:44 PST 2013


https://bugs.webkit.org/show_bug.cgi?id=103867





--- Comment #10 from pdeng6 <pan.deng at intel.com>  2013-01-15 21:34:29 PST ---
(From update of attachment 182559)
View in context: https://bugs.webkit.org/attachment.cgi?id=182559&action=review

>> Source/WebCore/loader/SubresourceLoader.cpp:149
>> +        m_resource->addRedirect();
> 
> We also need to check if this is a cross-origin redirect. If it is, then we must reset everything, including the start value stored in m_initiatorMap. See processing model step 19a.

thanks for your remind, it is really confused, let's sync understanding first of all.

Assume scenario, doc D req R1 -> R2 -> R3 -> R4. ( "->" means redirect, R4 is the final resource)
case 1: D, R1, R2, R3 and R4 are all from same origin.
case 2: although Ri and its adjacent are not the same origin, Ri allow its prior 'timing', or allowed timing by its posterior.
case 3: others

In case 1 and 2, redirectStart should be the initiate time of R1, redirectEnd should be last byte R3's redirect response. startTime should be initiate time of R1, fetchStart should be R4's request time.
In case 3, redirectStart should be 0, redirectEnd should be 0. startTime should be , startTime should be R4's request time, fetchStart should be R4's request time.
while for other fields, domainLookupStart, domainLookupEnd, connectStart, connectEnd, secureConnectionStart, requestStart and responseStart, should be reported if R4 and D same origin or R4 allow D timing. otherwise 0.

if above right, I would like to store redirect response chain(in CachedResource? I'm not sure how to store in m_initiatorMap of CachedResourceLoader) then feed it to  addResourceTiming(...). in addResourceTiming, pass in the correct values for startTime and redirectEnd in the constructor.

>> LayoutTests/http/tests/w3c/webperf/submission/Intel/resource-timing/test_resource_timing_same_origin_redirect.html:15
>> +            setup({explicit_done: true});
> 
> I don't think you need this if everything is tested during the load event. Same for the other tests.

seems it doesn't work well if I remove this and  the latter done().

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.


More information about the webkit-unassigned mailing list