CrashKill/2012-12-10
From MozillaWiki
Contents
General
- Any New/spiked Windows 8 crashes after the latest release
- For Reference: Holiday coverage in December
Socorro
- No crash correlation issue - https://bugzilla.mozilla.org/show_bug.cgi?id=819881? [marcia]
- Related to bug 817718
- Actively being worked on, but manual re-runs of the cron are failing right now. Suspect network problems
Flash
- Flash 11.6 crashes [marcia]
- Looks as if we need symbols for npswf32_11_6_602_105.dll as I see crashes in Beta 3 in the top 30
- KaiRo contacted the responsible Adobe people about that today.
- bug 810797 is the top 15 crashes for Beta.
- Looks as if we need symbols for npswf32_11_6_602_105.dll as I see crashes in Beta 3 in the top 30
- KaiRo is going through hang bugs and closing the ones for browser side reports due to the ongoing changes in hang reporting.
- Main flash contact for uploading symbols is out for December - Hitomi is the contact until he gets back.
Desktop
Beta
- bug 785348 for the #5 topcrash in 18.0b3 did get review now.
Aurora
- New bugs filed by scoobidiver
- bug 805745 (gfxASurface::Release) is still #2
- bug 789933, still #3, has a fix that landed on Aurora just yesterday.
Release
- Release uptake is fast now, 52M ADI yesterday
- We probably have a fix (in our/XBL code!) for the Yandex issue (topcrash #2, 5% of all 17.0.1 crashes), see bug 770238
Trunk
- Trunk crash rate is a bit high - looks as if we are recovering from bug 818962[marcia]
- New trunk crashes filed by scoobidiver:
- https://bugzilla.mozilla.org/show_bug.cgi?id=819903 - Audio/Video regression
- https://bugzilla.mozilla.org/show_bug.cgi?id=819745 - DOM bug that appears to be a regression
- https://bugzilla.mozilla.org/show_bug.cgi?id=819633 - Layout regression
Mobile
- bug 780831 (JB crash in dvm) is still the #1 issue on beta and rising on release, we need STR.
- bug 782223 (JB crash in tegra libs, mostly Nexus 7) is high on beta as well, also needs STR.
- bug 818415 was high on trunk, but is fixed now.
B2G
- bug 817946 - We seem to corrupt memory somewhere, which leads to pretty useless stacks in crash reports
- bug 814078 - Submitting all pending crashes - just landed on trunk, waiting for getting onto beta
- nhirata and marcia working on end to end crash reporting - they will have info by 12/14.