RESOLVED FIXED 40312
[chromium] Track state for whether a ResourceRequest is fetched via a proxy
https://bugs.webkit.org/show_bug.cgi?id=40312
Summary [chromium] Track state for whether a ResourceRequest is fetched via a proxy
Mike Belshe
Reported 2010-06-08 10:52:08 PDT
Track whether a page is fetched via a proxy.
Attachments
Simple patch - chromium specific. (4.50 KB, patch)
2010-06-08 17:59 PDT, Mike Belshe
levin: review+
Update the reviewer field . (5.26 KB, patch)
2010-06-09 11:58 PDT, Mike Belshe
mbelshe: commit-queue-
One more try, without tagalong fixes. (4.50 KB, patch)
2010-06-09 12:00 PDT, Mike Belshe
no flags
Mike Belshe
Comment 1 2010-06-08 17:59:07 PDT
Created attachment 58204 [details] Simple patch - chromium specific.
David Levin
Comment 2 2010-06-08 18:08:20 PDT
Not sure if it matters but these fields will not be copied across threads (see PassOwnPtr<CrossThreadResourceRequestData> copyData() const;). This is called when workers do resource requests. Other: The reviewer typically isn't filled in until someone has r+'ed it.
Mike Belshe
Comment 3 2010-06-08 18:32:00 PDT
Yes, CrossThreadResourceRequestData can't deal with platform specific flags. So this is no change. re the Reviewer: I know. I take a crap-shoot at the reviewer. If Dimitry reviews it, then I don't need to regenerate a patch. A better fix would be to get rid of ChangeLogs, but I think that is an issue for the Office of Redundancy to tackle...
David Levin
Comment 4 2010-06-08 19:50:03 PDT
(In reply to comment #3) > Yes, CrossThreadResourceRequestData can't deal with platform specific flags. So this is no change. Well, it doesn't at present. I suppose it should be modified to handle platform specific flags if they matter (when it is called). Specifically, this method is called when workers do resource requests. Do these fields matter for that case?
Mike Belshe
Comment 5 2010-06-09 10:24:01 PDT
(In reply to comment #4) > Specifically, this method is called when workers do resource requests. Do these fields matter for that case? We're trying to track at a page-load-level whether a page is "proxied or not". This already has room for error, because some resources on a page could be proxied, and others may not. We use the base-page as the trigger for how to label the page load. Since we're only looking at page loads, not resources loaded by XHR, workers, etc, it doesn't matter for workers.
David Levin
Comment 6 2010-06-09 10:35:28 PDT
(In reply to comment #5) > (In reply to comment #4) > > Specifically, this method is called when workers do resource requests. Do these fields matter for that case? > > We're trying to track at a page-load-level whether a page is "proxied or not". This already has room for error, because some resources on a page could be proxied, and others may not. We use the base-page as the trigger for how to label the page load. Since we're only looking at page loads, not resources loaded by XHR, workers, etc, it doesn't matter for workers. Thanks!
Mike Belshe
Comment 7 2010-06-09 11:58:56 PDT
Created attachment 58269 [details] Update the reviewer field .
Mike Belshe
Comment 8 2010-06-09 12:00:21 PDT
Created attachment 58271 [details] One more try, without tagalong fixes.
WebKit Commit Bot
Comment 9 2010-06-10 06:56:25 PDT
Comment on attachment 58271 [details] One more try, without tagalong fixes. Clearing flags on attachment: 58271 Committed r60955: <http://trac.webkit.org/changeset/60955>
Noam Rosenthal
Comment 10 2010-07-29 09:34:01 PDT
*** Bug 43011 has been marked as a duplicate of this bug. ***
Adam Barth
Comment 11 2010-08-10 22:49:55 PDT
This patch claims to have been landed.
Note You need to log in before you can comment on or make changes to this bug.