Closed
Bug 1257623
Opened 8 years ago
Closed 8 years ago
CC / GC between each cycle of tpaint
Categories
(Testing :: Talos, defect)
Testing
Talos
Tracking
(firefox48 fixed)
RESOLVED
FIXED
mozilla48
Tracking | Status | |
---|---|---|
firefox48 | --- | fixed |
People
(Reporter: mconley, Unassigned)
References
(Blocks 1 open bug)
Details
Attachments
(1 file)
I think at least part of the e10s regression on tpaint is accounted for by the fact that both processes are free'd up enough to trigger CC / GC during some cycles, causing spikes in the time to open windows. We should force CC / GC between each window open.
Reporter | ||
Comment 1•8 years ago
|
||
Review commit: https://reviewboard.mozilla.org/r/40909/diff/#index_header See other reviews: https://reviewboard.mozilla.org/r/40909/
Comment 2•8 years ago
|
||
https://reviewboard.mozilla.org/r/40909/#review37475 ::: testing/talos/talos/cmdline.py:89 (Diff revision 1) > add_arg('--spsProfileInterval', dest='sps_profile_interval', type=int, > help="How frequently to take samples (ms)") > add_arg('--spsProfileEntries', dest="sps_profile_entries", type=int, > help="How many samples to take with the profiler") > add_arg('--extension', dest='extensions', action='append', > - default=['${talos}/talos-powers/talos-powers-signed.xpi', > + default=['${talos}/talos-powers/', we shouldn't remove this unless you have a trick I don't know about. ::: testing/talos/talos/config.py:184 (Diff revision 1) > 'devtools.chrome.enabled': False, > 'devtools.debugger.remote-enabled': False, > 'devtools.theme': "light", > 'devtools.timeline.enabled': False, > - 'identity.fxaccounts.migrateToDevEdition': False > + 'identity.fxaccounts.migrateToDevEdition': False, > + 'xpinstall.signatures.required': False, I assume this is debugging? ::: testing/talos/talos/startup_test/tpaint.html:40 (Diff revision 1) > if (windowIndex >= REPEAT_COUNT) { > window.setTimeout(reportTimes, 0); > return; > } > + > + TalosPowersContent.forceCCAndGC(); iirc, cc | gc requires a bit of time to come back, do we need to wait for a callback?
Reporter | ||
Comment 3•8 years ago
|
||
Yeah, sorry - this patch was a WIP (work-in-progress), which is why there's still some debugging cruft in there, and I hadn't requested review yet. A for-reals patch is coming up soon.
Reporter | ||
Comment 4•8 years ago
|
||
https://reviewboard.mozilla.org/r/40909/#review37475 > iirc, cc | gc requires a bit of time to come back, do we need to wait for a callback? I'm using a sync message here to communicate with the parent, which means that it'll return only once the parent has finished the CC / GC.
Reporter | ||
Updated•8 years ago
|
Attachment #8731890 -
Attachment description: MozReview Request: Bug 1257623 - [WIP] CC / GC between each cycle of tpaint. r?jmaher → MozReview Request: Bug 1257623 - CC / GC between each cycle of tpaint. r?jmaher
Attachment #8731890 -
Flags: review?(jmaher)
Reporter | ||
Comment 5•8 years ago
|
||
Comment on attachment 8731890 [details] MozReview Request: Bug 1257623 - CC / GC between each cycle of tpaint. r?jmaher Review request updated; see interdiff: https://reviewboard.mozilla.org/r/40909/diff/1-2/
Comment 6•8 years ago
|
||
Comment on attachment 8731890 [details] MozReview Request: Bug 1257623 - CC / GC between each cycle of tpaint. r?jmaher https://reviewboard.mozilla.org/r/40909/#review37653 ::: testing/talos/talos/talos-powers/components/TalosPowersService.js:62 (Diff revision 2) > this.receiveProfileCommand(message); > break; > } > + case "TalosPowersContent:ForceCCAndGC": { > + Cu.forceCC(); > + Cu.forceGC(); does this really do a synchronous execution, or does this fire a command and take time to properly execute?
Attachment #8731890 -
Flags: review?(jmaher)
Reporter | ||
Comment 7•8 years ago
|
||
https://reviewboard.mozilla.org/r/40909/#review37653 > does this really do a synchronous execution, or does this fire a command and take time to properly execute? I believe both forceCC and forceGC are synchronous.
Comment 8•8 years ago
|
||
Comment on attachment 8731890 [details] MozReview Request: Bug 1257623 - CC / GC between each cycle of tpaint. r?jmaher https://reviewboard.mozilla.org/r/40909/#review37661 given the assertion that forceCC and forceGC are synchronous, lets ship it
Attachment #8731890 -
Flags: review+
Comment 10•8 years ago
|
||
bugherder |
https://hg.mozilla.org/mozilla-central/rev/77caf1897d7a
Status: NEW → RESOLVED
Closed: 8 years ago
status-firefox48:
--- → fixed
Resolution: --- → FIXED
Target Milestone: --- → mozilla48
You need to log in
before you can comment on or make changes to this bug.
Description
•