Summary: | crash reports seen in isLoadingMultipartContent | ||||||
---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | Darin Adler <darin> | ||||
Component: | Page Loading | Assignee: | Darin Adler <darin> | ||||
Status: | RESOLVED FIXED | ||||||
Severity: | Normal | CC: | scott | ||||
Priority: | P2 | Keywords: | InRadar | ||||
Version: | 523.x (Safari 3) | ||||||
Hardware: | Mac | ||||||
OS: | OS X 10.4 | ||||||
Attachments: |
|
Created attachment 17001 [details]
patch
Comment on attachment 17001 [details]
patch
r=me, although I'd keep the ASSERT in hope of some day encountering the bug in a debug build and understanding it.
Committed revision 27379. It would appear this isn't fixed. See Bug 16097. More precisely, it appears there's another crash that now happens in the same circumstance now that the first crash is fixed. |
Apple's crash tracer has reports of crashes that look like this: >#0 com.apple.WebCore 0x95417289 WebCore::DocumentLoader::isLoadingMultipartContent() const + 9 #1 com.apple.WebCore 0x95416eea WebCore::ImageTokenizer::finish() + 138 #2 com.apple.WebCore 0x95658b44 WebCore::FrameLoader::endIfNotLoadingMainResource() + 116 #3 com.apple.WebCore 0x95260053 WebCore::Document::close() + 35 #4 com.apple.WebCore 0x953134cc WebCore::JSHTMLDocumentPrototypeFunction::callAsFunction(KJS::ExecState*, KJS::JSObject*, KJS::List const&) + 732 I think it's just a simple missing null check. Lets do a speculative fix.