TPEPerformance/Meetings/2016-10-24

« previous week | index | next week »

OKR confidence

1. O: Make Firefox more stable [current crash rates] [Crashes per 1000 usage hours]
   1. KR: 1.2 nightly browser crashes per 100 ADI (per-month median)
      • Current: 2.46 -> 3.21 -> 2.38 -> n/a -> 1.65 -> 2.83 -> 2.58
   2. KR: 0.3 nightly startup crashes per 100 ADI (per-month median)
      • Current: 0.81 -> 0.82 -> 0.72 -> n/a -> 0.6 -> 0.91 -> 0.83
   3. KR: 2.0 nightly content crashes per 100 ADI (per-month median)
      • Current: 14.38 -> 8.70 -> 8.25 -> n/a -> 7.19 -> 10.91 -> 8.90
2. O: Improve Google document performance on Firefox
   1. O: Understand Google document performance issues on Firefox
      1. KR: Implement proposed fixes and measure results
         • Current:

Health Checks

1. Test code coverage (minus crash/no return code path)
   • N/A
2. Team mental health
   • Good
3. ADI / DAU / MAU
   • N/A

Previous Week

cervantes

• Previous
  • [x] Run remaining tests with App verifier
  • [x] Look at coverity
• Trying out DrMemory on Windows
• bug 1278443 RESOLVED::FIXED ChildImpl::OpenProtocolOnMainThread crashes when attempting to send a message during shutdown
  • Bug 1278443 - debugging mochitest media crashes with the patch
  • Landed bug 1278443
  • Triggering a try build with bug 1278443
• bug 1310152 NEW:: [App Verifier] Critical section not initialized in sqlite3
  • Bug 1310152 - run xpcshell test with valgrind to debug, but didn't find the root cause because the tests are run sequentially when valgrind is used.
  • Bug 1310152 - rerun the test with debug version of sqlite3 and found the crash. Investigating
  • Debugging bug 1310152
• bug 1289994 NEW:: Use Application Verifier
  • Bug 1289994 - running marionette and web-platform-tests with App Verifier
  • Bug 1289994 - running reftest-ipc with app verifier

kanru

• Previous
  • [x] Sort out test coverage framework
• Looked 10-17 nightly crashes
• Study coverage tools
• bug 1309799 RESOLVED::FIXED Generate ipc msgtype to name string mapping function and use in IPC
  • Bug 1309799 landed
• bug 1277337 NEW:: Use hg.mozilla.org to map crashes to bug components by way of source files when possible
  • Study how to setup json-mozbuildinfo for bug 1277337

wcpan

• Previous
  • [x] MOZ_MUST_USE in network/protocol/
  • [ ] Coverity
• bug 1046166 REOPENED:: [e10s] userContent.css does not work
  • Bug 1046166 - r-, maybe shouldn't use about: protocol
• bug 1310127 NEW:: Use MOZ_MUST_USE in netwerk/protocol/http
  • Bug 1310127 - likely busted on try
  • Bug 1310127 - some problems in IPC, ni necko team
• bug 1268766 NEW:: Use MOZ_MUST_USE everywhere
  • Opening sub-bugs for bug 1268766

ting

• Previous
  • Use memory tool to find problems
• Triaged 1016 nightly crashes
• bug 1312309 NEW:: Fails to build jsoncpp with clang-cl
  • Filed bug 1312309 and bug 1312313
• bug 1310926 NEW:: Crash in TraversalTracer::onChild
  • Am checking bug 1310926 and bug 1310925
  • bug 1310926 read disassembly and commented for the reason why i think it is about gc
• bug 1312313 NEW:: Fails to build breakpad with clang-cl because of c++11 narrowing checks
  • Filed bug 1312309 and bug 1312313
• bug 965936 NEW:: [meta] Virtual address space fragmentation
  • bug 965936 jit:ExecutableAllocator::systemAlloc is the majority of 1m unaligned allocations, it allocates 64k, am wondering what if we change it to 1m aligned
  • bug 965936 ran tracefrag for 2 days, got only the logs from the parent process and don't have any signficant new stack that allocate 64k-unaligned memory, will do another round with 1M-unaligned allocation
  • bug 965936 ran tracefrag overnight, but the browser doesn't response to any input events
• bug 1310925 RESOLVED::FIXED Crash in TimerThread::PostTimerEvent
  • Am checking bug 1310926 and bug 1310925
  • bug 1310925 figured out it is the target has been freed
• bug 752004 NEW:: Allow building Firefox with clang-cl on Windows
  • Am checking bug 752004, will try to build with clang-cl
  • bug 752004 failed building m-c with the latest clang, commented on the bug for more information, and am trying to figure out what goes wrong
• bug 1301022 RESOLVED::FIXED Add an API to annotate pending IPC when push IPC message to pending queue OOM
  • bug 1301022 received r+, try looks good, is goind to autoland

gregweng

• Previous
  • [x] Trace JS compilation.
  • [x] Persuade JS peers to reply their needinfo.
• Check the patch and code about compilation priority in SpiderMonkey
• Initialize a document about dynamic config set
• Just read and studied how re-lazifying benefits code size, which is related to the first task in Ion compilation
• Need to study what will happen if a new script comes when the current one is being complied: in theory, they can be handled even more parallely

thinker

• Previous
  • [x] SVG blur test cases & fix harness
  • [ ] Study performance related documents
• bug 1306107 NEW:: Problem with z-index and/or backface-visibility when applying transform: rotateY(180deg);
  • bug 1306107, have a patch to start bing reviewed.
  • Working on bug 1306107
• bug 1248828 NEW:: Extend the reftest harness to allow testing compositor-side invalidation
  • Make bug 1248828 to be reviewed.
• bug 1271983 NEW:: Animated SVG image is blurred in FF 46
  • Make bug 1271983 to be reviewed
• bug 1301361 NEW:: LayerTreeInvalidation has a corner case for extended 3D rendering context.
  • Waiting feedbacks for bug 1301361
• bug 779598 NEW:: Consider doing animations of transforms with preserve-3d off the main thread
  • Fix regression of the patch on bug 779598.

This Week Priorities

gweng

• Study JS engine logging
• Discuss devtools next step

kanru

• Study test coverage on try server
• Implement "display components" on crash-stats

cervantes

• PTO

thinker

• PTO

wcpan

• MOZ_MUST_USE on netwerk/protocol
• Coverity

ting

• JS JIT buffer allocation, 64k or 1M align
• Investigate clang-cl

Incoming

• Oct: Oct-21--Nov-7 some team memebres are going on a two weeks around-the-island bike trip
• Nov: WebSummit/CodeSummit