RESOLVED FIXED 21949
No need to clobber all ResourceRequest fields in FrameLoader::reload()
https://bugs.webkit.org/show_bug.cgi?id=21949
Summary No need to clobber all ResourceRequest fields in FrameLoader::reload()
Darin Fisher (:fishd, Google)
Reported 2008-10-29 11:35:24 PDT
No need to clobber all ResourceRequest fields in FrameLoader::reload() Presently, FrameLoader::reload clobbers all fields of the initialRequest when there is an unreachableURL associated with the current DocumentLoader. This causes problems for the Chromium port since we store some extra data on the ResourceRequest object that we need to have preserved across a FrameLoader::reload call. FrameLoader::reloadAllowingStaleData is careful to only call setURL when there is an unreachableURL, and I think that FrameLoader::reload should do the same. I suspect that it was not intentional to make FrameLoader::reload behave differently in this respect. Please see the attached patch for further details.
Attachments
v1 patch (1.71 KB, patch)
2008-10-29 11:39 PDT, Darin Fisher (:fishd, Google)
andersca: review+
Darin Fisher (:fishd, Google)
Comment 1 2008-10-29 11:39:57 PDT
Created attachment 24751 [details] v1 patch
Darin Fisher (:fishd, Google)
Comment 2 2008-10-29 11:42:06 PDT
To clarify, this causes problems for us because the HTTP method is forced to be GET. If the previous request was something else, then that information is lost. I guess this doesn't cause a problem for Safari, but it still seems like an error to me.
Darin Fisher (:fishd, Google)
Comment 3 2008-10-30 13:00:59 PDT
Comment on attachment 24751 [details] v1 patch Anders, can you please take a look at this patch? Thanks!
Anders Carlsson
Comment 4 2008-11-04 09:25:15 PST
Comment on attachment 24751 [details] v1 patch Looks good, I can't think of a reason why it would break things (but I've been wrong before ;) r=me!
Darin Fisher (:fishd, Google)
Comment 5 2008-11-04 11:41:18 PST
Note You need to log in before you can comment on or make changes to this bug.