MozillaWiki

Performance/Status Meetings/2008-January-23: Difference between revisions

Page Discussion

Performance/Status Meetings/2008-January-23 (view source)

Revision as of 19:54, 23 January 2008

1,030 bytes removed ,  23 January 2008
→‎Leaks: fill in bug numbers
No edit summary
(→‎Leaks: fill in bug numbers)
 
(7 intermediate revisions by 3 users not shown)
Line 15: Line 15:
* Still a few leaks left with popular extensions, but haven't triaged remaining bugs yet.
* 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.
* 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.
* We're keeping closed windows alive too long after they are closed
** {{bug|413447}}: nsXBLDocumentInfo can keep closed global window alive seem to be one source
** {{bug|394514}} and {{bug|387491}}: traversal issues related to XUL template code


=== Action Items ===
=== Action Items ===
Line 21: Line 23:
* '''AI:sicking''' Look into enabling trace-refcount, without the tracing part in nightlies.
* '''AI:sicking''' Look into enabling trace-refcount, without the tracing part in nightlies.
* '''AI:cbarrett''' Look into OS leaks on Mac.
* '''AI:cbarrett''' Look into OS leaks on Mac.
* '''AI:dbaron''' Make assertions fatal on new windows leak box.
* <s>'''AI:dbaron''' Make assertions fatal on new windows leak box.</s>
* '''AI:sicking''' Test leaked handles on windows manually using purify.
* '''AI:sicking''' Test leaked handles on windows manually using purify.
* '''AI:sayrer''' Run leak buster while looking for trace-malloc leaks. <B>[didn't get this done--sayrer]</b>
* '''AI:sayrer''' Run leak buster while looking for trace-malloc leaks. <B>[didn't get this done--sayrer]</b>
Line 27: Line 29:


== Performance ==
== Performance ==
* What Tp tests do we need to keep? Some only run on tinderbox which is being phased out. See [https://bugzilla.mozilla.org/show_bug.cgi?id=372870#c15].  
* 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
* JPEG/image speedups
Line 42: Line 44:
** linux: middle on 1.8, now fastest on 1.9 (30% improvement)
** linux: middle on 1.8, now fastest on 1.9 (30% improvement)


* massive Ts regression in 2006
* Ts improvements
** Ts analysis: {{bug|406810}}
** Ts analysis: {{bug|406810}}
** Related to thread manager landing
** {{bug|311566}} - nsCSSScanner::GatherIdent does lots of string appending
** Darin thinks it might be excessive painting.
** {{bug|412796}} - Optimize fastload system (mmap fileIO, endianness, packed structs)
** {{bug|411579}} - Optimize read file buffer sizes for faster startup times


* Profile Guided Optimization
* Profile Guided Optimization
** WPO turned on for windows ({{bug|361343}}) -- nice performance win on XP
** {{bug|361343}} - Win32 turn-on; {{bug|410994}} - PGO Spidermonkey
** Spidermonkey seems to respond well to PGO under GCC ({{bug|410994}})
** TODO: Figure out Linux (sayrer)
** sayrer has entire browser building with -fprofile-generate on Linux. no results yet.
** TODO: Fix win32 breakage (cairo; possibly fixed with cairo patch?)
** 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)
* Other
* {{bug|196843}} -- CSS fastload
** {{bug|353962}} -- JS spinlocks vs. full locks (Crashes in Fx2)


* On radar (unowned stuff that people have mentioned have mentioned we should look at at some point)
* On radar (unowned stuff that people have mentioned have mentioned we should look at at some point)
Line 63: Line 64:
** more work to speed up onSecurityChange, but further work will require SSL work and reducing the number of notifications
** 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)
** 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 [http://developer.apple.com/tools/download/ 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 ===
=== Action Items ===
Dbaron
canmove, Confirmed users, Bureaucrats and Sysops emeriti
1,334

edits

Retrieved from "https://wiki.mozilla.org/Special:MobileDiff/80670...80816"