[Webkit-unassigned] [Bug 24615] JavascriptCore leaks when creating and releasing context groups in a tight loop

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Sat Jun 1 01:52:03 PDT 2013


https://bugs.webkit.org/show_bug.cgi?id=24615





--- Comment #11 from Scott <bugs.webkit.org at scottdavies.net>  2013-06-01 01:50:35 PST ---
(In reply to comment #10)
> Set the DYLD_FRAMEWORK_PATH environment variable to the path containing JavaScriptCore.framework before running your program (e.g., path/to/WebKit.app/Contents/Frameworks/10.8 if you're working with a nightly build and on OS X 10.8.x).

Yeah, I'd already tried that, but even with that variable set appropriately, running otools -L on my executable was still listing the old system framework as a dependency rather than the new one, so I wasn't sure it was actually going to do the right thing.  On the other hand, if I set DYLD_FRAMEWORK_PATH to the new framework's path, the leak does go away, so I suppose (1) otools is probably just lying to me about what will actually get used, and (2) the leak *has* actually been fixed.  Whew.  

A previous nightly I downloaded and compiled just a month or two, on the other hand, did leak memory on the same test at an absolutely furious pace...much faster than the preinstalled system framework leaks memory, even.  Is there any way to identify WebKit "releases" that are reasonably recent but also reasonably stable?  I've never even gotten a nightly to actually build WebKit all the way through...every time I've tried in the last month or two, including the one I downloaded just last night, the build has failed (specifically on NetscapePluginHostManager.mm, with "fatal error: 'WebKitSystemInterface.h' file not found
#import "WebKitSystemInterface.h"...pretty sure it's been the same build failure every time).

Anyhow, sorry for mostly wasting your time...thanks!

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.


More information about the webkit-unassigned mailing list