RESOLVED FIXED141097
GC marking threads should clear malloc caches 
https://bugs.webkit.org/show_bug.cgi?id=141097
Summary GC marking threads should clear malloc caches
Geoffrey Garen
Reported 2015-01-30 13:35:57 PST
GC marking threads should clear malloc caches
Attachments
Patch (7.69 KB, patch)
2015-01-30 14:11 PST, Geoffrey Garen 		no flags
DetailsFormatted DiffDiff
Patch (1.90 KB, patch)
2015-01-30 16:27 PST, Geoffrey Garen 		sam: review+
DetailsFormatted DiffDiff
Show Obsolete (1) View All Add attachment proposed patch, testcase, etc.
Geoffrey Garen
Comment 1 2015-01-30 14:11:14 PST
Created attachment 245740 [details] Patch
Andreas Kling
Comment 2 2015-01-30 14:18:53 PST
Comment on attachment 245740 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=245740&action=review r=me > Source/JavaScriptCore/ChangeLog:8 > + This is an attempt to amelirate a potential memory use regression ameliorate
Geoffrey Garen
Comment 3 2015-01-30 14:22:52 PST
Committed r179411: <http://trac.webkit.org/changeset/179411>
Geoffrey Garen
Comment 4 2015-01-30 14:27:51 PST
Committed r179412: <http://trac.webkit.org/changeset/179412>
Mark Hahnenberg
Comment 5 2015-01-30 14:50:03 PST
Comment on attachment 245740 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=245740&action=review > Source/JavaScriptCore/heap/GCThread.cpp:79 > + WTF::releaseFastMallocFreeMemoryForThisThread(); I wonder if it makes more sense to do this at the end of collection rather than between phases? Seems a shame to clear out the per-thread malloc cache between marking and copying since copying might want to reuse some stuff from the marking phase.
Mark Hahnenberg
Comment 6 2015-01-30 14:54:11 PST
Also, maybe I dunno what scavenging is...
Geoffrey Garen
Comment 7 2015-01-30 15:47:39 PST
> I wonder if it makes more sense to do this at the end of collection rather > than between phases? Yeah, end of collection is better. How to achieve? Maybe move the code to the end of the 'case Copy' block?
Mark Hahnenberg
Comment 8 2015-01-30 16:14:46 PST
Yeah that seems ok. There shouldn't be any parallel work left to do at the end of the CopyPhase, so that sounds like a good time to clear the malloc caches on those threads.
Mark Hahnenberg
Comment 9 2015-01-30 16:17:21 PST
Another idea might be to only clear the caches during collections caused by the GCTimer.
Geoffrey Garen
Comment 10 2015-01-30 16:27:53 PST
Reopening to attach new patch.
Geoffrey Garen
Comment 11 2015-01-30 16:27:54 PST
Created attachment 245755 [details] Patch
Geoffrey Garen
Comment 12 2015-01-30 16:29:17 PST
Committed r179422: <http://trac.webkit.org/changeset/179422>
Chris Dumez
Comment 13 2015-01-30 16:34:33 PST
Looks like some JSC tests are failing on the bots after this change? https://build.webkit.org/builders/Apple%20Yosemite%20Debug%20WK2%20%28Tests%29/builds/1796/steps/jscore-test/logs/stdio
Geoffrey Garen
Comment 15 2015-01-30 17:16:34 PST
Turned out to be a sporadic failure in that date test.
Note You need to log in before you can comment on or make changes to this bug.