[Webkit-unassigned] [Bug 87183] Web Inspector: CodeGeneratorInspector.py: protect typed API from C++ implicit float to int cast

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Tue May 29 13:50:31 PDT 2012


https://bugs.webkit.org/show_bug.cgi?id=87183





--- Comment #18 from Peter Rybin <prybin at chromium.org>  2012-05-29 13:50:30 PST ---
(In reply to comment #17)
> Thank you very much. Looks very relevant, yet I don't see what could go wrong.
> Looking into it.

What bugs me here is that stacktraces in all crashes look quite correct, but the top frame:

1. WebKit!WebCore::TypeBuilder::Memory::MemoryBlock::Builder<0>::setName+0x3c [c:\cygwin\home\buildbot\slave\win-release\build\webkitbuild\release\obj\webcore\derivedsources\inspectortypebuilder.h @ 625]

2. WebKit!WebCore::PageRuntimeAgent::notifyContextCreated+0xe5 [c:\cygwin\home\buildbot\slave\win-release\build\source\webcore\inspector\pageruntimeagent.cpp @ 161]

...

While "MemoryBlock::Builder<0>::setName" method exists and line #625 is absolutely meaningful, PageRuntimeAgent.cpp has nothing to do with this method at all. The file never mentions MemoryBlock type. It does call "setName" method of completely different type (Runtime::ExecutionContextDescription::Builder). Moreover, MemoryBlock::Builder<0> is a template instantiation and this instantiation could only occur from the different file – InspectorMemoryAgent.cpp.

As crazy as it sounds, right now it looks like a corrupted stacktrace or a linker problem, not even compiler (weird).

-- 
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