RESOLVED FIXED 129227
REGRESSION (r164507): Crash beneath JSGlobalObjectInspectorController::reportAPIException at facebook.com, twitter.com, youtube.com 
https://bugs.webkit.org/show_bug.cgi?id=129227
Summary REGRESSION (r164507): Crash beneath JSGlobalObjectInspectorController::report...
mitz
Reported 2014-02-22 21:04:32 PST
<rdar://problem/16142324> Navigating to the URL causes the Web Content process to crash with a backtrace like this: 0 com.apple.JavaScriptCore 0x00000001064d75ce WTFCrash + 62 1 com.apple.JavaScriptCore 0x0000000106195fae JSC::ErrorHandlingScope::ErrorHandlingScope(JSC::VM&) + 62 2 com.apple.JavaScriptCore 0x000000010639d5c6 Inspector::JSGlobalObjectInspectorController::reportAPIException(JSC::ExecState*, JSC::JSValue) + 70 3 com.apple.JavaScriptCore 0x000000010636eb6b JSValueToObject + 331 4 com.apple.Safari.framework 0x00007fff9264aa21 controlObject(objc_object*, double, NSString*) + 207 This was caused by <http://trac.webkit.org/r164507>, the fix for bug 128776.
Attachments
Revert r164507 (60.63 KB, patch)
2014-02-22 21:07 PST, mitz 		eric.carlson: review+
DetailsFormatted DiffDiff
View All Add attachment proposed patch, testcase, etc.
mitz
Comment 1 2014-02-22 21:07:09 PST
Created attachment 224992 [details] Revert r164507
mitz
Comment 2 2014-02-22 21:39:03 PST
Committed <http://trac.webkit.org/r164554>.
Philippe Wittenbergh
Comment 3 2014-02-22 21:59:34 PST
Is bug 129221 the same issue ?
mitz
Comment 4 2014-02-22 22:10:52 PST
(In reply to comment #3) > Is bug 129221 the same issue ? Yes. Sorry I didn’t see that one!
Philippe Wittenbergh
Comment 5 2014-02-22 22:18:58 PST
*** Bug 129221 has been marked as a duplicate of this bug. ***
Joseph Pecoraro
Comment 6 2014-02-23 10:53:33 PST
I see. I kept the ErrorHandlingScope from the original code, but I guess I'm not exactly why it is needed. Thanks for rolling out.
Joseph Pecoraro
Comment 7 2014-02-23 10:54:00 PST
Correction: I'm not exactly sure* why it is needed.
Joseph Pecoraro
Comment 8 2014-02-27 12:41:17 PST
The original patch landed again in <http://trac.webkit.org/changeset/164824>. I could not reproduce the problem after the recent JSC fixes for this. I was unable to reproduce the same reportAPIException calls you guys saw, but I did produce a different one and we handled that fine. Please let me know if you are seeing crashes again after r 164824!
Note You need to log in before you can comment on or make changes to this bug.