TPEPerformance/Meetings/2016-08-15
From MozillaWiki
< TPEPerformance | Meetings
« previous week | index | next week »
Contents
OKR confidence
- O: Make Firefox more stable [current crash rates]
- KR: 1.2 nightly browser crashes per 100 ADI (per-month median)
- Current: 2.46
- KR: 0.3 nightly startup crashes per 100 ADI (per-month median)
- Current: 0.81
- KR: 2.0 nightly content crashes per 100 ADI (per-month median)
- Current: 14.38
- KR: 1.2 nightly browser crashes per 100 ADI (per-month median)
- O: Make crashes easier to find & fix
- KR: 90% new crashes have associated bugs opened and tracked
- Current: no metric
- KR: 90% new crashes have associated bugs opened and tracked
- O: Improve Google document performance on Firefox
- O: Understand Google document performance issues on Firefox
- KR: Reduced the problem space to a finite areas that we can improve
- Current: GC slowness, Inline script blocking content
- KR: Reduced the problem space to a finite areas that we can improve
- O: Understand Google document performance issues on Firefox
Health Checks
- Test code coverage (minus crash/no return code path)
- N/A
- Team mental health
- Good
- ADI / DAU / MAU
- N/A
Previous Week
This report is automatically generated by https://github.com/kanru/standups_weekly
thinker
- Previous commitment:
- [x] Network, focus on load resource earlier
- [ ] Improve task tracer (scriptable interface)
- Content.sink.enable_perf_mode=1 looks more responsiveness.
- NsHtml5TreeOpExecutor would suspend current parsing and resume it later if it run too long.
- Speculative loading has a great potential to improve gdoc startup time. I am looking into it, and observe its behavior.
- XHR::open would be blocked to create an actor in sync. It should be improved. Greg would study how much time it costs.
- bug 1269690 NEW:: [Perf][google docs] 45.25%(4528ms) slower than Chrome when opening 1 page table content
- Vsync seems to affect multiple 10%s of time for gdoc table case. bug 1269690
cervantes
- Previous commitment:
- [x] Study heap memory in minidump, experiment with previous crashes
- [x] e10s media crash, possibley gmp related
- Tracing GMP threading code
- bug 1279612 NEW:: Near permafailing mda tests on Win7 debug on beta in test_webvtt_disabled.html | application crashed [@ mozalloc_abort(char const * const)]
- Commented the finding to bug 1279612
- bug 1286802 NEW:: Add some heap areas to the minidump
- Bug 1286802 - testing the patch and comparing with the MiniDumpWithIndirectlyReferencedMemory type of minidump
- Bug 1286802 - testing the patch with previously fixed crashes
wcpan
- Previous commitment:
- [x] More MOZ_MUST_USE
- [ ] Fix Coverity discovered bugs
- bug 1293212 NEW:: Use MOZ_MUST_USE in uriloader/
- Bug 1293212 - review request
- bug 1046166 NEW:: [e10s] userContent.css does not work
- Bug 1046166 - review granted
ting
- Previous commitment:
- [x] Fragmentation. Experiment on previous crash to see the virtual memory layout
- [x] RelaseData() crash
- Checked 0807 nightly crashes
- bug 1292781 NEW:: [e10s] Treeherder scroll position back to the top after loading new results if HWA was disabled
- Am investigating bug 1292781
- bug 965936 NEW:: [meta] Virtual address space fragmentation
- Bug 965936 found mozjemalloc still has 1M default chunk size, bug 1202523 was for jemalloc, fixed the script
- Bug 965936 posted what i've found for a oom-small so far
- bug 1145613 NEW:: Crash in [@ ReleaseData ] & [@ ReleaseData(void*, unsigned int)]
- Filed bug 1293501 for adding debug annotations to track down bug 1145613
- bug 1202523 RESOLVED::FIXED Use the default chunk size with jemalloc 4
- Bug 965936 found mozjemalloc still has 1M default chunk size, bug 1202523 was for jemalloc, fixed the script
- bug 1293577 RESOLVED::FIXED Crash in mozilla::image::Decoder::HasProgress
- Read minidumps of bug 1293577 and bug 1293569
- bug 1293569 Invalid bug or security bug
- Read minidumps of bug 1293577 and bug 1293569
- bug 1293501 NEW:: Add debug annotation/code for the crash in ReleaseData()
- Bug 1293501 submitted a path for adding annotations for the nsHttpHeaderArray
- Filed bug 1293501 for adding debug annotations to track down bug 1145613
kanru
- Previous commitment:
- [x] Structured Clone (reduce OOM)
- [ ] Crash components
- [ ] MOZ_MUST_USE
- [ ] JS stack in crash report
- bug 1292892 RESOLVED::FIXED Stop using JSRuntime outside SpiderMonkey
- Review patch for bug 1292892
gweng
- Previous commitment:
- [ ] JavaScript GC slowness on GDoc
- [ ] Wait bailout result from QA
This Week Priorities
kanru
- Structured Clone (reduce OOM)
- Crash components
- MOZ_MUST_USE
- JS stack in crash report
- Establish graphic of crash rates and MAUs
- Predict script running time and run rendering before run script, ask support
thinker
- Sort task tracer patch to a bug.
- Work on gfx rendering bugs
- Gdoc, classify tasks into needed and un-needed tasks
cervantes
- Heap in minidump bug
- GMP shutdown crash
- Explore jemalloc, app verifier, etc
wcpan
- MOZ_MUST_USE
- Coverity bugs
ting
- Memory Fragmentation
- ReleaseData
gweng
- Check bailout test results
- External (inline) scrippt tuning
- JS timeout, executation deadline
Incoming
- Sep: Many team members plan to take PTOs for the Mid-Autumn Festival
- Nov: WebSummit/CodeSummit