FF4STATUS: Firefox 4 beta bug roundup
Christian Legnitto
[589670] - Need to help OS X users upgrade Flash for Firefox 4?
https://bugzilla.mozilla.org/show_bug.cgi?id=589670
Status: ON TRACK
Has Patch: Yes
Synopsis: The bug has morphed into a blocklisting bug for certain old Flash versions on Mac OS X 10.6. The blocklist entry has been staged and is being tested. There are some peripheral issues related to the plugincheck webpage (see bug 633285) but nothing blocking Firefox 4
--
[630947] - reloading about:home builds up 50+ DOMWindows
https://bugzilla.mozilla.org/show_bug.cgi?id=630947
Staus: ON TRACK
Has Patch: Yes
Synopsis: A performance bug where we are holding onto references too long (but not a leak as they are eventually cleaned up). It looks to have landed on mozilla-central but is not marked fixed yet. Additional/similar work was spun off into bug 614347
--
[626678] - Firefox becomes unresponsive when docking or undocking my laptop
https://bugzilla.mozilla.org/show_bug.cgi?id=626678
Status: AT RISK
Has Patch: No
Synopsis: In certain configurations (certain machines with certain graphics cards using a docking station with dual monitors) Firefox may hang. This is possibly due to graphics drivers that need to be blacklisted. We are unable to reproduce internally but there are comments in the bug that community members can reproduce easily
--
[629799] - Alpha extraction in plugin subprocesses is broken on windows
https://bugzilla.mozilla.org/show_bug.cgi?id=629799
Status: ON TRACK
Has Patch: Yes
Synopsis: This is the risky bug we have been waiting on. It enables Flash movies set as transparent to display correctly with graphics acceleration enabled. It landed on mozilla-central and bounced due to test failures. The failures have been figured out and we are waiting for the updated patch to land. We know this bug needs at least a couple days of nightly testing before we will release a beta with it
--
[632781] - Tall <canvas> element fails to scroll with the rest of the page
https://bugzilla.mozilla.org/show_bug.cgi?id=632781
Status: ON TRACK
Has Patch: Yes
Synopsis: When creating a large canvas element scrolling is messed up. The reporter has a reproducible testcase. The bug has both a fix and tests and is expected to land on mozilla-centeal shortly.
--
[611591] - Clearbox broken (incorrect image order, white instead of first image)
https://bugzilla.mozilla.org/show_bug.cgi?id=611591
Status: UNKNOWN
Has Patch: No
Synopsis: Changes in Firefox 4 seem to make a certain website behave differently by loading images in a different order. The offending changeset has been identified (bug 572520). It may be the web developers with incorrect expectations / code. It is potentially a web compatibility issue that needs to be handled with care
--
[626909] - jsd.onError and jsd.onDebug fail
https://bugzilla.mozilla.org/show_bug.cgi?id=626909
Status: ON TRACK
Has Patch: No
Synopsis: This bug looks to have been written about two issues. One looks to have never worked and has now morphed into an enhancement request. The actual regression appears to already be fixed in tracemonkey. This likely has no work remaining and the bugs just need to be cleaned up
--
[632343] - Running automated Firebug tests crashes Firefox [@ js::mjit::EnterMethodJIT]
https://bugzilla.mozilla.org/show_bug.cgi?id=632343
Status: ON TRACK
Has Patch: Yes
Synopsis: Firebug's automated tests uncovered an issue in Firefox. Dave Townsend bisected to find http://hg.mozilla.org/mozilla-central/rev/b647ff21bfed as the offending changeset. There is a patch that has a r+ once the "nits" are addressed
--
[591687] - Flash performance heavily degraded in Firefox 4 Beta 4
https://bugzilla.mozilla.org/show_bug.cgi?id=591687
Status: AT RISK
Has Patch: Yes
Synopsis: This bug is using YouTube HD videos to make sure Firefox's CPU usage is consistent with other browsers. There are some large changes to the way we draw CoreAnimation plugins. The patch has actually already landed on mozilla-central but there may be regressions (like 633826, which was fixed) and the underlying issue may not be 100% fixed
--
[626602] - Implement background copying for windowless async plugins
https://bugzilla.mozilla.org/show_bug.cgi?id=626602
Status: AT RISK
Has Patch: Yes
Synopsis: This patch changes the way we draw plugin content on Mac OS X. According to comment 245 (!) the patch is pretty much ready to land. This looks like a scary change that should have at least a couple of days of bake before being released in a beta
--
[626206] - Flash menu and games rendered with black background
https://bugzilla.mozilla.org/show_bug.cgi?id=626206
Status: ON TRACK
Has Patch: No
Synopsis: This bug is essentially a dupe of bug 626602. We are keeping it open to make sure the issues are in fact resolved when bug 626602 (above) lands.
--
[628872] - BBC iplayer fullscreen function broken in FF4 beta 9, 10 and nightly builds
https://bugzilla.mozilla.org/show_bug.cgi?id=628872
Status: ON TRACK
Has Patch: Yes
Synopsis: Jim was able to reproduce this and fix it, but his fix regressed a fix for a different player. It looks like the current patch fixes the outstanding BBC issue without regressing other Flash players!
--
[614347] - XPConnect-wrapped JSObjects must clear their gray bit when they are handed out
https://bugzilla.mozilla.org/show_bug.cgi?id=614347
Status: AT RISK
Has Patch: Yes
Synopsis: This bug is related to bug 630947 above. There is a patch that has landed and been backed out twice on tracemonkey. There are still outstanding issues and the bug is at risk. Due to the nature of the change I would suggest a couple of days bake before going to beta if this patch eventually lands
--
[633520] - """ASSERTION: wrong context on XPCJSContextStack!: 'cx == topJSContext',"" with xpcshell"
https://bugzilla.mozilla.org/show_bug.cgi?id=633520
Status: ON TRACK
Has Patch: Yes
Synopsis: This bug is fixed on tracemonkey. Kyle Huey found the changeset that caused the issue and Luke cooked up a patch that looks to have fixed the issue and is green on try
--
[613800] - A Race Condition when closing inactive tab with alerts in OnBeforeUnload
https://bugzilla.mozilla.org/show_bug.cgi?id=613800
Status: ON TRACK
Has Patch: Yes
Synopsis: This bug is fallout from modal dialogs and likely affects Panorama as well. The issue is that there is a race condition when closing inactive tabs with alerts fired by OnBeforeUnload. Mihai has a patch that Dolske has given feedback on, but it currently has some outstanding issues and has been r-. The bug is well understood, has STR, and is well on it's way to getting a r+.
Again, if you don't want to see these reports please filter on "FF4STATUS", I won't be offended in the slightest.
Thanks,
Christian
[1] The query I used for this email is https://bugzilla.mozilla.org/buglist.cgi?quicksearch=blocking2.0%3AbetaN+sw%3Ahard