Summary: | JSC GC does not coalesce blocks during minor collections | ||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | Filip Pizlo <fpizlo> | ||||||||
Component: | JavaScriptCore | Assignee: | Filip Pizlo <fpizlo> | ||||||||
Status: | NEW --- | ||||||||||
Severity: | Normal | CC: | eric, fpizlo, oliver, slewis, webkit.review.bot | ||||||||
Priority: | P2 | ||||||||||
Version: | 528+ (Nightly build) | ||||||||||
Hardware: | All | ||||||||||
OS: | All | ||||||||||
Attachments: |
|
Description
Filip Pizlo
2011-08-03 20:03:25 PDT
Created attachment 102870 [details]
the patch (work in progress)
Tests still running. It would be good if I could get a confirmation that the performance is not regressed, and that memory usage is good, since this patch does a lot of damage to the sweeping code.
Attachment 102870 [details] did not pass style-queue:
Failed to run "['Tools/Scripts/check-webkit-style', '--diff-files', u'Source/JavaScriptCore/ChangeLog', u'Source..." exit_code: 1
Source/JavaScriptCore/heap/NewSpace.h:174: Should have only a single space after a punctuation in a comment. [whitespace/comments] [5]
Source/JavaScriptCore/heap/NewSpace.h:176: Should have only a single space after a punctuation in a comment. [whitespace/comments] [5]
Total errors found: 2 in 6 files
If any of these errors are false positives, please file a bug against check-webkit-style.
Created attachment 102877 [details]
the patch (work in progress, fix style)
Still testing...
Comment on attachment 102877 [details]
the patch (work in progress, fix style)
This appears to pass tests. I'd still like to get a confirmation that the performance and memory usage is in good shape but other than that, it's good to review.
Comment on attachment 102877 [details] the patch (work in progress, fix style) View in context: https://bugs.webkit.org/attachment.cgi?id=102877&action=review > Source/JavaScriptCore/ChangeLog:50 > + objects are 64MB, this change is not currently expected to be a big 64k maybe? (In reply to comment #5) > (From update of attachment 102877 [details]) > View in context: https://bugs.webkit.org/attachment.cgi?id=102877&action=review > > > Source/JavaScriptCore/ChangeLog:50 > > + objects are 64MB, this change is not currently expected to be a big > > 64k maybe? Good point. :-) Created attachment 102987 [details]
the patch
I'm going to postpone committing this patch until some more tests happen, since this patch does potentially risky things to the way we call destructors, and the way we reclaim memory. Comment on attachment 102877 [details] the patch (work in progress, fix style) Cleared Oliver Hunt's review+ from obsolete attachment 102877 [details] so that this bug does not appear in http://webkit.org/pending-commit. Comment on attachment 102987 [details] the patch View in context: https://bugs.webkit.org/attachment.cgi?id=102987&action=review > Source/JavaScriptCore/heap/NewSpace.h:-179 > - functor(block); Just testing. > Source/JavaScriptCore/heap/NewSpace.h:277 > + sizeClass.currentBlock = block; Yeah. Testing testing 1-2-3. Attachment 102987 [details] was posted by a committer and has review+, assigning to Filip Pizlo for commit.
|