NEW 33288
platform/mac/fast/canvas/canvas-draw-xbm-image.html crashed on Leopard Commit Bot 
https://bugs.webkit.org/show_bug.cgi?id=33288
Summary platform/mac/fast/canvas/canvas-draw-xbm-image.html crashed on Leopard Commit...
Eric Seidel (no email)
Reported 2010-01-06 19:53:24 PST
platform/mac/fast/canvas/canvas-draw-xbm-image.html crashed on Leopard Commit Bot This is the first time I've seen this crash. I will attach a crash log.
Attachments
crash log (63.60 KB, text/plain)
2010-01-06 19:54 PST, Eric Seidel (no email) 		no flags
Details
View All Add attachment proposed patch, testcase, etc.
Eric Seidel (no email)
Comment 1 2010-01-06 19:54:36 PST
Created attachment 46016 [details] crash log Looks like it's trying to grab at a bad pointer in JSC. Exception Type: EXC_BAD_ACCESS (SIGBUS) Exception Codes: KERN_PROTECTION_FAILURE at 0x000000000000002c Crashed Thread: 0 Thread 0 Crashed: 0 com.apple.JavaScriptCore 0x0031ee7f JSC::JSGlobalObject::~JSGlobalObject() + 79 1 com.apple.JavaScriptCore 0x00262daa JSC::BytecodeGenerator::BytecodeGenerator(JSC::FunctionBodyNode*, JSC::Debugger const*, JSC::ScopeChain const&, WTF::HashMap<WTF::RefPtr<JSC::UStringImpl>, JSC::SymbolTableEntry, JSC::IdentifierRepHash, WTF::HashTraits<WTF::RefPtr<JSC::UStringImpl> >, JSC::SymbolTableIndexHashTraits>*, JSC::CodeBlock*) + 42 2 com.apple.JavaScriptCore 0x002932b2 JSC::FunctionExecutable::reparseExceptionInfo(JSC::JSGlobalData*, JSC::ScopeChainNode*, JSC::CodeBlock*) + 450 (BytecodeGenerator.h:375) 3 com.apple.JavaScriptCore 0x0026f49a JSC::CodeBlock::reparseForExceptionInfoIfNecessary(JSC::ExecState*) + 90 (CodeBlock.cpp:1460) 4 com.apple.JavaScriptCore 0x002f6f33 cti_op_call_NotJSFunction + 99 (MacroAssemblerCodeRef.h:149) 5 ??? 0x1c587e14 0 + 475561492
Eric Seidel (no email)
Comment 2 2010-01-06 19:55:15 PST
Oliver has been very helpful in diagnosing this type of crash in the past. I'll CC him in case he has any comments. (You should of course feel free to remove yourself.)
Eric Seidel (no email)
Comment 3 2010-01-06 19:57:08 PST
The commit-queue hit this during it's "make sure the current revision builds and passes all the tests before trying to apply the patch" phase.
Oliver Hunt
Comment 4 2010-01-08 14:23:20 PST
Eric what are the repro steps for this, does it happen on its own?
Eric Seidel (no email)
Comment 5 2010-01-08 15:11:05 PST
The bot was just running run-webkit-tests on a clean checkout. I've not tried reproducing the crash. I've only seen it the once, I filed this bug for tracking. If we don't see the crash again then we can just close this as unreproducible.
Eric Seidel (no email)
Comment 6 2010-01-08 16:05:07 PST
I ran the test with --iterations=10000 and it didn't crash. :) I suspect this was an innocent victim of some other test or stray electron. We can close this in a couple weeks if we don't see anything like it again.
Note You need to log in before you can comment on or make changes to this bug.