CrashKill/2013-01-28
From MozillaWiki
Contents
General
- bug 806820 is dominating all desktop branches and is even rising on mobile.
Socorro
- Issues last week causing a dip in crash data for at least 22nd - 24th, see bug 834403, should be backfilled today.
Flash
Looks like Adobe has been busy fixing a lot of 11.5 and 11.6 regressions in 11.6 betas, they promise more of those in this week's beta.
Desktop
Beta
- Maybe we should really still take the UUID bump on nsIPrefBranch (bug 828296) on 19?
- bug 817568 (StringCchCopyNW) is now #10 in 19.0b3, a number of those startup crashes seem to be from a single user, though.
Aurora
Release
18 crash rates still look pretty bad.
- bug 806820, "the Facebook crasher", has become significantly worse, #1 with >20% of all 18.0.1 crashes over the last 3 days.
- Patch in bug 831626 apparently helped on trunk, we need to get this uplifted where we can, maybe even do an 18.0.2 for it.
- bug 783369, "the Norton crash", seems to have improved, but still #3 with >3%.
- Others like bug 803963 (Avira, #18), bug 828184 (Websense, #198), maybe even malware-related bug 801394 are caused by the missing UUID bump on nsIPrefBranch (bug 828296).
- bug 801394, "the malware crash", is continuing to drop, now #24 with 0.3%
- Another variant of the malware was found and possibly should be submitted to AV vendors.
Trunk
- bug 827946 is a Linux (!) crash that now is #3 overall on 21.0a1 over the last week.
Mobile
- Fennec Graph
- Also to note there was a dip in ADU from the channel change:
- ADU
- being backfilled.
- Nightly and Aurora on the climb.
- Both show:
- bug 827171 nsPluginStreamListenerPeer::OnStartRequest(nsIRequest*, nsISupports*)
- bug 705641 mozalloc_abort(char const*) | NS_DebugBreak_P | mozilla::layers::LayerManagerOGL
- bug 778175 - Startup crash on devices incompatible with OpenGL ES 2.0 (e.g. HTC Wildfire)
- bug 771774 - crash with eglMakeCurrent failed (EGL error 3000) and abort message: "OpenGL-accelerated layers are a hard requirement on this platform [...]" on Vivante GPUs (e.g. in Rockchip rk29board, imapx200 and Vimicro chipsets)
- Aurora also shows:
- bug 830131 - java.lang.NullPointerException: at org.mozilla.gecko.InputMethods.needsSoftResetWorkaround(InputMethods.java)
- Both show:
- Beta :
- bug 800838 - Android crash in nsNPAPIPluginInstance::UnscheduleTimer @ nsNPAPIPluginInstance::TimerWithID
- bug was reopened and fixed again (1/24);need push to aurora : see comment 51
- bug 823236 - crash in mozilla::layers::ShmemYCbCrImage::IsValid on ICS and above
- bug 800838 - Android crash in nsNPAPIPluginInstance::UnscheduleTimer @ nsNPAPIPluginInstance::TimerWithID
- Release
- bug 752828 - android.database.sqlite.SQLiteDatabaseLockedException: database is locked at android.database.sqlite.SQLiteStatement.native_executeSql(Native Method) on ICS
- a lot more libdvm.so crashes have happened.
- A Search by OS per crash signature would be helpful ; filed bug 835444
B2G
- Main tracking bugs in relation to b2g crash reporting:
- bug 833574 - (b2g-frontend-crash-reporting) [Meta] Other Crash Reporting Bugs
- bug 761905 - (b2g-crash-reporting) Tracking: Crash reporting for b2g (gonk)
- Possible graph per carrier
- QA Work:
- Rough Draft of B2G Crash Reporting
- Need a cumulative tally script for the week or at least 3 days.
- For beta channel those would come from socorro.
- Top crashes:
- bug 834164 - [Camera] Camera crash after opened video app and played a video
- bug 834372 - crash in mozilla::layers::BasicShadowableThebesLayer::CreateBuffer with abort message: "creating ThebesLayer 'back buffer' failed! width=2621440, height=2064384, type=3000"
- Crash Related:
- bug 827833 - B2G crash in [@ mozalloc_abort | NS_DebugBreak_P | mozilla::layers::PImageContainerChild::FatalError]
- should be fixed
- bug 827833 - B2G crash in [@ mozalloc_abort | NS_DebugBreak_P | mozilla::layers::PImageContainerChild::FatalError]