Firefox/Channels/Meetings/2016-05-17

From MozillaWiki
Jump to: navigation, search

Channel Meeting Details

Video/Teleconference Details - NEW

  • 650-903-0800 or 650-215-1282 x92 Conf# 99951 (US/INTL)
  • 1-800-707-2533 (pin 369) Conf# 99951 (US)
  • Vidyo Room: Release Coordination
  • Vidyo Guest URL
REMEMBER
These notes are read by people who weren't able to attend the meeting. Please make sure to include links and context so they can be understood.



Attendees

  • kmoir, lizzard, sylvestre, calixte, marco, anthony vaughn, blassey, marcia, phiippe, ryanvm, kbrosnan, jst

Previous Actions

Schedule Update

  • 47 beta 6 Desktop is live
  • beta 6 mobile uploaded in GP
  • 47 beta 7 g-t-b Thursday


Add-ons

Stability

Aurora

Overall rate: 1.79 (orange) - browser: 1.18, content: 0.62 There was a compiler change (VS2013 -> VS2015) recently which caused some signatures to be modified. This means some signatures that appear to have spiked actually haven't. I guess we're going to have the same problem when we start building beta and then release with VS2015. Aurora is probably going to switch back to VS2013, but nightly is going to stay with VS2015.

- Spike with signature shutdownhang | WaitForSingleObjectEx | WaitForSingleObject | PR_WaitCondVar | nsThread::ProcessNextEvent | NS_ProcessNextEvent | mozilla::layers::CompositorBridgeParent::ShutDown from 0 to 65 crashes in three days - filed https://bugzilla.mozilla.org/show_bug.cgi?id=1273514 - Spike with signature nsAutoCompleteController::HandleEnter (a few comments mention malware, one mentions "SafeFinder" in particular). The crashes on 48.0a2 all come from a single person, which does have a few suspicious DLLs. I've analyzed other reports, many have suspicious DLLs or addons. - Spike with signature tosqep64.dll@0x8f8c, fix now landed on inbound ( https://bugzilla.mozilla.org/show_bug.cgi?id=1268905 ) - Spike with signature shutdownhang | WaitForSingleObjectEx | WaitForSingleObject | PR_WaitCondVar | nsThread::ProcessNextEvent | NS_ProcessNextEvent | mozilla::net::nsHttpConnectionMgr::Shutdown, the related bug ( https://bugzilla.mozilla.org/show_bug.cgi?id=1158189 ) is marked as fixed in 48 as of 2016-05-04, but it looks like it isn't actually fixed - Spike with signature nss3.dll@0x1e66d0 | GetFileInfo, all crashes are from the same user

Beta

Overall rate: 1.24 (orange) - browser: 1.06, content: 0.18 - Spike with js::GCMarker::markAndTraceChildren<T>, https://bugzilla.mozilla.org/show_bug.cgi?id=1267686, and js::GCMarker::lazilyMarkChildren, https://bugzilla.mozilla.org/show_bug.cgi?id=1259214

https://crash-stats.mozilla.com/search/?product=Firefox&process_type=browser&proto_signature=~D3D11DXVA2Manager&proto_signature=~CDecodeContext&version=47.0b5&_facets=signature&_facets=user_comments&_facets=adapter_vendor_id&_facets=platform_pretty_version&_columns=date&_columns=signature&_columns=version#facet-signature

New crashes in beta 5 (maybe this is interesting...) https://crash-analysis.mozilla.com/rkaiser/datil/searchcompare/?common=product%3DFirefox%26process_type%3Dbrowser%26process_type%3Dcontent%26dom_ipc_enabled%3D__null__&p1=version%3D47.0b4&p2=version%3D47.0b5

    • Maybe JS issues

Release

Overall rate: 1.04 (orange) startup crashes 0.14, shutdownhangs: 0.2 - Spike with OOM | large | NS_ABORT_OOM | mozilla::ipc::MessageChannel::OnMessageReceivedFromLink. The linked bug is https://bugzilla.mozilla.org/show_bug.cgi?id=1156484, which is duplicated to https://bugzilla.mozilla.org/show_bug.cgi?id=1235633, which is fixed in 47 and wontfix for 46. This appears to be a e10s-only crash? OK, but it's ~300 crashes in a week out of 229,397 crash reports. Do we care about this kind of spike? (Maybe. I commented in bug 1235633) - liz - Spike with RtlActivateActivationContextUnsafeFast | UserCallWinProcCheckWow, plugin related, similar to https://bugzilla.mozilla.org/show_bug.cgi?id=723133 Here, we likely want to file a new bug, let's ask jimm - Spike with signature F1398665248_____________________________, https://bugzilla.mozilla.org/show_bug.cgi?id=774281, it doesn't seem actionable as it is a generic flash crash - Spike with nsTArray_Impl<T>::Clear | nsPluginFrame::SetEmptyWidgetConfiguration, https://bugzilla.mozilla.org/show_bug.cgi?id=1182247, fix uplifted to aurora, no way to tell if it's fixed until it goes to beta - Spike with _dwrite_draw_glyphs_to_gdi_surface_gdi, https://bugzilla.mozilla.org/show_bug.cgi?id=726206, fix in nightly, requested uplift to aurora - Spike with mozilla::layers::BasicCompositor::DrawQuad, https://bugzilla.mozilla.org/show_bug.cgi?id=1187466 has a WIP patch - Spikes with GC-related OOM crashes, waiting on https://bugzilla.mozilla.org/show_bug.cgi?id=1257387 to get more informations about the crashes (allocation size) - Spike with Media-related shutdown hangs (shutdownhang | WaitForSingleObjectEx | WaitForSingleObject | PR_WaitCondVar | nsThread::ProcessNextEvent | NS_ProcessNextEvent | mozilla::MediaShutdownManager::Shutdown)

Nightly


Mobile

Performance (Thu)

Aurora / Dev Edition

Beta

Release

QE

Nightly

Aurora / Dev Edition

Beta

Mobile

Performance

RelEng

Marketing/Communications

User Advocacy

No updates 46 Release Survey: https://docs.google.com/document/d/1n3C6i7kr6KFcf8n9shA7QoAfH1tLXbeDyxcXBDgj6rw/edit

Roundtable

Telemetry Experiments

  • Beta47
    • Plugin content blocking (Bug 1248813)
      • Start date: 47.0b2 ; End date: 47.0b8 (gtb)
    • E10s turned on for 20% beta population (Bug ???)
      • Start date: 47.0b1 ; End date: 47.0b6

System add-on updates

  • Hello 1.3.2 to ship-to- beta (yesterday?)

Special Topics

Aurora/Beta Feature Review

Post-Mortem (Tues 2wks from GA Release)

Sign Off (Thurs prior to GA Release)

Action items

Actions