UNCONFIRMED 19387
WebKit causes spins under normal usage patterns 
https://bugs.webkit.org/show_bug.cgi?id=19387
Summary WebKit causes spins under normal usage patterns
Michael J. Cohen
Reported 2008-06-03 21:22:51 PDT
opened jdate and a few other tabs, nothing that would put my mac pro to shame, and I get spins like the attached ones.
Attachments
5 second spin observed. (26.91 KB, text/plain)
2008-06-03 21:27 PDT, Michael J. Cohen 		no flags
Details
additional spin, 7 seconds, same loaded website content. (170.49 KB, text/plain)
2008-06-03 21:32 PDT, Michael J. Cohen 		no flags
Details
Second spin, 1.59 seconds. (48.71 KB, text/plain)
2008-06-03 21:34 PDT, Michael J. Cohen 		no flags
Details
0.71 seconds, same site. (49.39 KB, text/plain)
2008-06-03 21:38 PDT, Michael J. Cohen 		no flags
Details
Same site, different behaviors to trigger. (24.57 KB, text/plain)
2008-06-03 21:41 PDT, Michael J. Cohen 		no flags
Details
Show Obsolete (1) View All Add attachment proposed patch, testcase, etc.
Michael J. Cohen
Comment 1 2008-06-03 21:27:53 PDT
Created attachment 21491 [details] 5 second spin observed.
Michael J. Cohen
Comment 2 2008-06-03 21:32:28 PDT
Created attachment 21492 [details] additional spin, 7 seconds, same loaded website content.
Michael J. Cohen
Comment 3 2008-06-03 21:33:16 PDT
Comment on attachment 21492 [details] additional spin, 7 seconds, same loaded website content. not webkit. whoops
Michael J. Cohen
Comment 4 2008-06-03 21:34:32 PDT
Created attachment 21493 [details] Second spin, 1.59 seconds.
Michael J. Cohen
Comment 5 2008-06-03 21:38:45 PDT
Created attachment 21495 [details] 0.71 seconds, same site.
Michael J. Cohen
Comment 6 2008-06-03 21:41:52 PDT
Created attachment 21496 [details] Same site, different behaviors to trigger.
Mark Rowe (bdash)
Comment 7 2008-06-03 22:56:11 PDT
Of the five logs attached, only the first and third look to have anything particularly WebKit-related executing at the time of the sample. Both of these samples show the time being spent below CFURLConnectionSendSynchronousRequest, which suggests to me that the site in question is doing a synchronous XHR.
Note You need to log in before you can comment on or make changes to this bug.