Bug 126219 – js/dom/dfg-custom-getter-throw.html and js/dom/dfg-custom-getter-throw-inlined.html hit assertions after r161051

• Status: RESOLVED
• Resolution: FIXED
• Component: JavaScriptCore

[Ryosuke Niwa]

js/dom/dfg-custom-getter-throw.html
and
js/dom/dfg-custom-getter-throw-inlined.html
have started hitting assertions after http://trac.webkit.org/changeset/161051

ASSERTION FAILED: exec == topCallFrame || exec == exec->lexicalGlobalObject()->globalExec() || exec == exec->vmEntryGlobalObject()->globalExec()
/Volumes/Data/slave/mavericks-debug/build/Source/JavaScriptCore/runtime/VM.cpp(634) : JSC::JSValue JSC::VM::throwException(JSC::ExecState *, JSC::JSValue)
1   0x10da505b0 WTFCrash
2   0x10d9f14af JSC::VM::throwException(JSC::ExecState*, JSC::JSValue)
3   0x10f6cb39f WebCore::setDOMException(JSC::ExecState*, int)
4   0x10fbe6119 WebCore::JSXMLHttpRequest::responseText(JSC::ExecState*) const
5   0x10fbe0d3f WebCore::jsXMLHttpRequestResponseText(JSC::ExecState*, long long, long long, JSC::PropertyName)
6   0x3ed880e68ffa

http://build.webkit.org/results/Apple%20Mavericks%20Debug%20WK2%20(Tests)/r161053%20(1201)/results.html

[Ryosuke Niwa]

Looks like this is a fairly recent regression in JSC.

I certainly did not encounter it at r158715, and reverting the WebCore code change in r161051 confirms that the assertion failure exists without the patch.

[Ryosuke Niwa]

Created attachment 219979 [details]
The version of js/dom/dfg-custom-getter-throw.html that reproduces the assertion failure

[Radar WebKit Bug Importer]

<rdar://problem/15723849>

[Ryosuke Niwa]

Committed r161059: <http://trac.webkit.org/changeset/161059>

[Ryosuke Niwa]

Sorry, didn't mean to close this bug.

[Mark Lam]

Testing with a recent build (r165197), I don't see this issue manifest anymore.  There also have been numerous changes and bug fixes in the area of exception handling and stack management since the time this bug was filed.  Those changes appear to have resolved the issue. 

There doesn't seem to be anything left to do for this bug.  Will close.

[Alexey Proskuryakov]

Mark, did you unskip the skipped tests?

[Mark Lam]

(In reply to comment #7)
> Mark, did you unskip the skipped tests?

They were already unskipped.  See https://bugs.webkit.org/show_bug.cgi?id=126219#c4.