Firefox/Aurora/2011-06-09: Difference between revisions
Jump to navigation
Jump to search
| Line 22: | Line 22: | ||
** 6.0a2 2011/06/08: https://crash-analysis.mozilla.com/chofmann/20110607/top-plugin-crashes-6.0a2.html | ** 6.0a2 2011/06/08: https://crash-analysis.mozilla.com/chofmann/20110607/top-plugin-crashes-6.0a2.html | ||
** Top crash hang | NtUserMessageCall | RealDefWindowProcW [https://bugzilla.mozilla.org/show_bug.cgi?id=618683 Bug 618683] is considerably higher on 6.0a2 (321) vs 7.0a1 (59) | ** Top crash hang | NtUserMessageCall | RealDefWindowProcW [https://bugzilla.mozilla.org/show_bug.cgi?id=618683 Bug 618683] is considerably higher on 6.0a2 (321) vs 7.0a1 (59) | ||
* Addendum | |||
** Upon further investigation, there is a plugin hang regression for Flash 10.3 that we need to back out of Aurora. It could be that this is the root of the increases we see. We need to get that checked in ASAP so we can verify [https://bugzilla.mozilla.org/show_bug.cgi?id=658741 Bug 658741]. | |||
== Add-on compatibility bump == | == Add-on compatibility bump == | ||
Revision as of 21:48, 9 June 2011
Agenda
- Metrics for 6.0a2
- Elevated crash level?
- Status on add-on compatibility bump
Metrics for 6.0a2
- Aurora ADU's continuing to creep up - 46,751 yesterday.
Elevated crash level for 6.0a2?
- Looking at the graphs on https://crash-stats.mozilla.com/products/Firefox, 6.0a2 is way higher than any other build, even nightly
- Is this a tools issue? We wouldn't expect Aurora to be an outlier as it is pretty close to beta and central
- Analysis - browser crashes (smooney)
- 7.0a1 2011/06/07: 1.69 browser crashes per 100 ADU - https://crash-analysis.mozilla.com/chofmann/20110608/top-7.0a1.html
- 6.0a2 2011/06/07: 1.87 browser crashes per 100 ADU - https://crash-analysis.mozilla.com/chofmann/20110608/top-6.0a2.html
- For browser crashes rates are not that different.
- Looking at number for 6.0a2, nothing has significantly high volume. Some crashes that indicate 6.0a2 only but low volume <10 per day.
- One signature - js_XDRScriptAndSubscripts(JSXDRState*, JSScript**) that is NOT on trunk or beta and should have been fixed. Needs some investigation. Bug 648022. Only 19 crashes so not high enough to affect the crash rate.
- Analysis - plugin crashes (smooney)
- The plugin crashes show a different story
- 7.0a1 2011/06/08: https://crash-analysis.mozilla.com/chofmann/20110607/top-plugin-crashes-7.0a1.html
- 6.0a2 2011/06/08: https://crash-analysis.mozilla.com/chofmann/20110607/top-plugin-crashes-6.0a2.html
- Top crash hang | NtUserMessageCall | RealDefWindowProcW Bug 618683 is considerably higher on 6.0a2 (321) vs 7.0a1 (59)
- Addendum
- Upon further investigation, there is a plugin hang regression for Flash 10.3 that we need to back out of Aurora. It could be that this is the root of the increases we see. We need to get that checked in ASAP so we can verify Bug 658741.
Add-on compatibility bump
- Happened!
Action Items
- Christian will write up the meeting notes about how to try to grow the overall Aurora audience
- Christian and grace are meeting again so he is waiting to write them up in case the meeting alters the plan