Performance/Status Meetings/2008-January-23
Gecko 1.9 Performance and Leaks Meeting Details
- Wednesdays - Firefox 3 - 10:00am Pacific, 1:00pm Eastern
- Script Conference Room, Building S, Mountain View
- 650-903-0800 or 650-215-1282 x91 Conf# 217 (US/INTL)
- 1-800-707-2533 (pin 369) Conf# 217 (US)
- irc.mozilla.org #perf for backchannel
Leaks
- Most (!!) extension leaks fixed by peters and igors work in bug 412491 and bug 398219.
- Still a few leaks left with popular extensions, but haven't triaged remaining bugs yet.
- Peterv has a patch to fix bug in cycle collector that causes many random crashes bug 407034, still awaiting review.
- We're keeping closed windows alive too long after they are closed. bug 413447 seem to be one source. RDF bug another.
Action Items
- AI:sicking Look into enabling trace-refcount, without the tracing part in nightlies.
- AI:cbarrett Look into OS leaks on Mac.
- AI:dbaron Make assertions fatal on new windows leak box.
- AI:sicking Test leaked handles on windows manually using purify.
- AI:sayrer Run leak buster while looking for trace-malloc leaks. [didn't get this done--sayrer]
- AI:sicking to go through and prioritize leak bugs
Performance
- What Tp tests do we need to keep? Some only run on tinderbox which is being phased out. See bug 372870, comment#15 .
- JPEG/image speedups
- bug 411379: JPEG color conversion MMX/SSE2
- Shows perf wins on win32, OSX has bug 412396
- bug 411718: avoid buffer copy on JPEG decode
- Perf Bugs for FF3:
- Txul platform differences according to mac mini machines
mac: slowest on 1.8, slower still on 1.9 (30% regression)Looks even or better than branch- windows: fastest on 1.8, now in second place (30% regression) -- improving, still slower
- linux: middle on 1.8, now fastest on 1.9 (30% improvement)
- massive Ts regression in 2006
- Ts analysis: bug 406810
- Related to thread manager landing
- Darin thinks it might be excessive painting.
- Profile Guided Optimization
- WPO turned on for windows (bug 361343) -- nice performance win on XP
- Spidermonkey seems to respond well to PGO under GCC (bug 410994)
- sayrer has entire browser building with -fprofile-generate on Linux. no results yet.
- dbaron: shipped fx 1.0 with PGO on linux as well, was a 10% perf win... but compiler upgrade + build system breakage stopped us. We need to restart this on linux (and see about OSX).
- -GL breaks some things on win32
- bug 353962 -- JS spinlocks vs. full locks (Crashes in Fx2)
- bug 196843 -- CSS fastload
- On radar (unowned stuff that people have mentioned have mentioned we should look at at some point)
- Theme optimization
- roc's scrollbar issues
- startup performance (fastload perf, loading libXUL and other files, etc.) (unowned)
- more work to speed up onSecurityChange, but further work will require SSL work and reducing the number of notifications
- XPConnect accessor speed (bug 412531) (request for priority+owner by dietrich)
- Start/Stop Shark from JS (bug 411365 Patch)
- Gets really clean profiles.
- Install newest CHUD
- Start Shark App.
- Good setting for brief events is config->advanced 20us sample rate
- Put shark in remote mode
- Example from bug 412396 test:
<html>
<head>
<title>Test JPEG Rendering Time - (Hacked from scragz' test)</title>
<script type="text/javascript">
connectShark();
startShark(); // Start sampling
startDate = new Date();
startTime = startDate.getTime() / 1000;
window.onload = function () {
var ignoreMe = document.body.offsetWidth;
var endTime = new Date();
};
window.scrollBy(0,1);
function calceRenderingTime() {
stopShark(); // Stop sampling
disconnectShark();
endDate = new Date();
endTime = endDate.getTime() / 1000;
alert(endTime - startTime);
}
</script>
</head>
<body onload="calceRenderingTime();">
...
Action Items
- make faster
Fragmentation
- bug 407459 -- Hook up jemalloc to our build
- ported to windows, mac and linux. integrated to build system on mac and linux. windows work continuing. Hope to have numbers by end of the week.
- bug 403830 -- Content arena
- bug 408720 -- Fallout
- New arenas
- Other bugs (should have perf and footprint keywords)
- bug 403354 -- Get rid of nsCStringArray ParseString usage (maybe fixed?)
- bug 407442 -- nsEventListenerManager allocates a lot [patch]
- bug 408113 -- JS reallocate context->stackPool arenas way too often [patch]
- bug 408301 -- XPConnect wrappers w/o XPConnect proto don't share JS scopes with their protos. [patch]
- bug 410250 -- nsID/nsJSID stupid mallocs [patch]
- bug 410357 -- use nsAutoTArray for nsCSSDeclaration::mOrder [patch]
- bug 410392 -- stack allocate gfx objects [patch, needs work]
- bug 408921 -- Arenas size+header are too big for size classification
Action Items
- fix things