I can reproduce this bug in Nightly 57 x64 20170807113452 de_DE @ Windows 7 x64, too.
OS: Linux → All
Hardware: x86_64 → All
> Nightly 57 x64 20170807113452 @ Debian Testing (with Stylo enabled & disabled) Same build, a day later: WFM (wtf) > I can reproduce this bug in Nightly 57 x64 20170807113452 de_DE @ Windows 7 x64, too. Still broken (in a new profile). There is a ghost under the hood.
> > I can reproduce this bug in Nightly 57 x64 20170807113452 de_DE @ Windows 7 x64, too. > Still broken (in a new profile). On https://bugzilla.mozilla.org it works as it should. Maybe a history or HSTS/HSTS Preloading thing? I will dig further.
Nightly 57 x64 20170808114032 @ Debian Testing This only happens if you typed in a domain without a protocol, pressed enter and Ctrl+U. STR (see attached video): 1. Type in fastaim.de, press enter (or confirm the first Awesomebar result "fastaim.de -- Visit") 2. press Ctrl+U: view-source: is missing and links are unclickable 3. Close these tabs. 4. Type in "fastaim.de", but select the second Awesomebar result, e.g. "https://fastaim.de/", press enter and Ctrl+U: everything is fine
> This only happens if you typed in a domain without a protocol, pressed enter and Ctrl+U. Also if you reload a website which was opened on this way and then press Ctrl+U you will get this bug. (attachment 8894666 [details] from comment 0 was about that. Sorry for the unprecise STR there).
I had some fun with https://download-origin.cdn.mozilla.net/pub/firefox/nightly/2017/07/ I renamed the filenames after testing: firefox-45.0a1.de.linux-x86_64_unaffected.tar.bz2 firefox-53.0a2.de.linux-x86_64_unaffected.tar.bz2 firefox-56.0a1.de.linux-x86_64_2017-06-30-10-02-34_unaffected.tar.bz2 firefox-56.0a1.de.linux-x86_64_2017-07-18-16-42-11_unaffected.tar.bz2 firefox-56.0a1.de.linux-x86_64_2017-07-24-10-03-04_unaffected.tar.bz2 firefox-56.0a1.de.linux-x86_64_2017-07-26-10-02-41_unaffected.tar.bz2 firefox-56.0a1.de.linux-x86_64_2017-07-27-16-25-31_unaffected.tar.bz2 firefox-56.0a1.de.linux-x86_64_2017-07-28-10-03-58_unaffected.tar.bz2 firefox-56.0a1.de.linux-x86_64_2017-07-29-10-02-5_unaffected.tar.bz2 firefox-56.0a1.de.linux-x86_64_2017-07-30-10-03-07_affected.tar.bz2 <-------- de (first affected build) firefox-56.0a1.en-US.linux-x86_64_2017-07-30-10-03-07_affected.tar.bz2 <------- en-US https://www.mozilla.org/de/firefox/beta/all/ firefox-55.0b13_unaffected.tar.bz2 (German Linux x64) Beta 56 is not out yet. So I went here: https://download-origin.cdn.mozilla.net/pub/firefox/candidates/56.0b1-candidates/build5/linux-x86_64/de/ * affected *
Thanks for chasing it down this far! Have you used mozregression before? That should give us a very precise window, down to just a few changesets most likely. : https://mozilla.github.io/mozregression/quickstart.html
(In reply to J. Ryan Stinnett [:jryans] (use ni?) from comment #7) > Thanks for chasing it down this far! > > Have you used mozregression before? That should give us a very precise window, down to just a few changesets most likely. Amazing tool! mozregression --good 2017-07-28 --bad 2017-07-31 [...] 9:07.06 INFO: Last good revision: dd3e2939065e8c6805b1994583d5a04f39c8ab7c 9:07.06 INFO: First bad revision: 00167e9fe0c0fc573801eb8a905eb3822290c2da 9:07.06 INFO: Pushlog: https://hg.mozilla.org/integration/mozilla-inbound/pushloghtml?fromchange=dd3e2939065e8c6805b1994583d5a04f39c8ab7c&tochange=00167e9fe0c0fc573801eb8a905eb3822290c2da
Thanks for the regression window! :mayhemer, from the regression log above, it sounds like your patch might be involved here. Can you please take a look?
Assignee: nobody → honzab.moz
This looks like a proper fix. nsDocShell::LoadPage is dedicated to showing the view-source: of the page. It replaces the URI and drops the originalURI from the cloned session history entry (a session history entry clone from nsDocShell::GetCurrentDescriptor called from ViewSourceContent.loadSource). Interesting is that nsDocShell::LoadPage has never been dropping loadReplace from the entry as well - probably a long standing bug.
Attachment #8895486 - Flags: review?(bzbarsky)
Comment on attachment 8895486 [details] [diff] [review] v1 r=me
Attachment #8895486 - Flags: review?(bzbarsky) → review+
Status: NEW → ASSIGNED
[Tracking Requested - why for this release]: Regression in View Source functionality, let's be sure to uplift.
Comment on attachment 8895486 [details] [diff] [review] v1 Approval Request Comment [Feature/Bug causing the regression]: 1380012 [User impact if declined]: as this bug title says [Is this code covered by automated tests?]: apparently not :) [Has the fix been verified in Nightly?]: only locally so far [Needs manual test from QE? If yes, steps to reproduce]: yes, see comment 4 for str [List of other uplifts needed for the feature/fix]: none [Is the change risky?]: no [Why is the change risky/not risky?]: the bug is isolated to view-source and the fix is also in a code that is dedicated to view-source only, while the cause is clear and well understood [String changes made/needed]: none
Attachment #8895486 - Flags: approval-mozilla-beta?
Comment on attachment 8895486 [details] [diff] [review] v1 We definitely want view source to work! Please uplift for 56 beta 2.
Comment on attachment 8895486 [details] [diff] [review] v1 Correction, let's uplift this fix after we land it on m-c. It may still make it to beta 2.
Attachment #8895486 - Flags: approval-mozilla-beta+ → approval-mozilla-beta?
Pushed by firstname.lastname@example.org: https://hg.mozilla.org/integration/mozilla-inbound/rev/164240dd108d Drop result principal URI from a cloned session history entry when opening view-source: page. r=bz
Comment on attachment 8895486 [details] [diff] [review] v1 Fix a regression. Beta56+. Should be in 56.0b3.
I've reproduced this issue on an affected Nightly build using STR from comment 4. This is verified fixed on latest Nightly 57.0a1 (2017-08-22) and 56 Beta 6 (20170821193225) under Windows 10 x64, Mac OS X 10.11.6 and Ubuntu 16.04 x64 LTS.
You need to log in before you can comment on or make changes to this bug.