CrashKill/2014-05-19: Difference between revisions

From MozillaWiki
Jump to navigation Jump to search
Line 29: Line 29:


==Android==
==Android==
* Crash rate for 30b4 looks better
** [https://crash-stats.mozilla.com/daily?form_selection=by_version&p=FennecAndroid&v=&v=29.0b10&v=30.0b4&v=30.0b2&hang_type=any&os=Windows&os=Mac+OS+X&os=Linux&date_range_type=report&date_start=2014-04-25&date_end=2014-05-19&submit=Generate Raw crash data]
* Looking into the increase in raw crashes for Nightly


==B2G==
==B2G==

Revision as of 17:07, 19 May 2014

General

  • Data for yesterday missing on Socorro an other places because ADI push to Socorro failed. Metrics is looking into it, possibly pulling Socorro folks in.

Previous Action Items

  • [Carryover] KaiRo/liz to evaluate options for public stability list. Mailing list could be moderated.
  • [DONE] Naoki to file a bug on the Geeksphone channel : bug 1009155, bonus bug : bug 1009151
  • [DONE] Naoki to reach out to Vance on schedule for QC symbols : new symbols should be updated later this week (and IT deleted the old ones)

Socorro

  • ADI is late. Currently investigating.
  • Symbols upload investigation found 3 blocking problems. The 2 operational problems were resolved Friday, the last problem is code an expected to ship late this week.
  • New this week: ADU by signatures handles crash_count of 0 better, new version of browser id, lots of work to support ceph prototype

Flash

Desktop

Beta

  • a few stability fixes landed in beta5 for significant crash volume bugs. Crash stats are reflecting these fixes are working.
  • bug 1012691 - crash in NS_ABORT_OOM(unsigned int) | PL_DHashTableInit(PLDHashTable*, PLDHashTableOps const*, void*, unsigned int, unsigned int) | nsCycleCollector::BeginCollection(ccType, nsICycleCollectorListener*)
    • filed as a follow up to bug 1006181 - HTTP cache v2 memory only storage area not freed on limit
  • bug 938226 - crash in nsHtml5TreeBuilder::flushCharacters()
    • Should it ride the trains or can it be safely uplifted?

Aurora

Trunk

dmajor: There was a symbols problem on nightly between the 5/8 and 5/13 builds, see bug 1003085 comment 10. It caused a bunch of signatures that don't make sense. Things should be back to normal, but keep an eye out just in case.

Android

  • Crash rate for 30b4 looks better
  • Looking into the increase in raw crashes for Nightly

B2G

Notes

Upcoming PTO

  • Next Monday, May 26th, is a US holiday, so many people are out and there will be no meeting!
  • juanb, PTO, Thur 22 - Mon 26, Texas

Action Items