119812 2013-08-14 13:41:00 -0700 errorDescriptionForValue() should not assume error value is an Object 2013-08-22 14:24:34 -0700 1 1 1 Unclassified WebKit JavaScriptCore 528+ (Nightly build) Unspecified Unspecified RESOLVED FIXED P2 Normal --- 119803 1 chris_curtis chris_curtis bfulgham chris_curtis commit-queue jbriance joepeck mark.lam msaboff oliver oldest_to_newest 917268 0 chris_curtis 2013-08-14 13:41:00 -0700 +++ This bug was initially created as a clone of Bug #119803 +++ Bug 119803 <https://bugs.webkit.org/show_bug.cgi?id=119803> reported 2 crashes: 1 in a release build, the other in a debug build. The release build issues is due to errorDescriptionForValue() assuming that the passed in error value is either one of the following: null, undefined, int32, double, true, false, string, function, or object. If the value did not turn out to be one of the first 8 cases, we assumed it was an object. This turned out to not be true i.e. the steps below produced an error JSValue that was not any of the 9 cases. This bug will fix that issue. The debug crash is due to something else, and we'll let <https://bugs.webkit.org/show_bug.cgi?id=119803> track that. Steps to reproduce the crash: Visiting the website http://html5test.com using WinLauncher on Windows crashes with the following stacktrace: In release we crash as follows: > JavaScriptCore.dll!JSC::JSCell::methodTable() Line 157 C++ JavaScriptCore.dll!JSC::errorDescriptionForValue(JSC::ExecState * exec, JSC::JSValue v) Line 110 + 0x8 bytes C++ JavaScriptCore.dll!JSC::createError(JSC::ExecState * exec, JSC::JSObject * (JSC::ExecState *, const WTF::String &)* errorFactory, JSC::JSValue value, const WTF::String & message) Line 115 + 0x24 bytes C++ JavaScriptCore.dll!JSC::createNotAnObjectError(JSC::ExecState * exec, JSC::JSValue value) Line 139 + 0x28 bytes C++ JavaScriptCore.dll!JSC::JSValue::synthesizePrototype(JSC::ExecState * exec) Line 111 + 0xe bytes C++ JavaScriptCore.dll!JSC::JSValue::get(JSC::ExecState * exec, JSC::PropertyName propertyName, JSC::PropertySlot & slot) Line 637 C++ JavaScriptCore.dll!JSC::getByVal(JSC::ExecState * callFrame, JSC::JSValue baseValue, JSC::JSValue subscript, JSC::ReturnAddressPtr returnAddress) Line 1544 + 0x2b bytes C++ JavaScriptCore.dll!cti_op_get_by_val_generic(void * * args) Line 1605 C++ 0b8307d0() JavaScriptCore.dll!JSC::JITCode::execute(JSC::JSStack * stack, JSC::ExecState * callFrame, JSC::VM * vm) Line 46 + 0x20 bytes C++ JavaScriptCore.dll!JSC::Interpreter::execute(JSC::ProgramExecutable * program, JSC::ExecState * callFrame, JSC::JSObject * thisObj) Line 851 + 0x2d bytes C++ JavaScriptCore.dll!JSC::evaluate(JSC::ExecState * exec, const JSC::SourceCode & source, JSC::JSValue thisValue, JSC::JSValue * returnedException) Line 85 C++ WebKit.dll!WebCore::JSMainThreadExecState::evaluate(JSC::ExecState * exec, const JSC::SourceCode & source, JSC::JSValue thisValue, JSC::JSValue * exception) Line 74 + 0x1b bytes C++ WebKit.dll!WebCore::ScriptController::evaluateInWorld(const WebCore::ScriptSourceCode & sourceCode, WebCore::DOMWrapperWorld * world) Line 142 + 0x34 bytes C++ WebKit.dll!WebCore::ScriptController::evaluate(const WebCore::ScriptSourceCode & sourceCode) Line 158 + 0x40 bytes C++ WebKit.dll!WebCore::ScriptElement::executeScript(const WebCore::ScriptSourceCode & sourceCode) Line 316 + 0x16 bytes C++ WebKit.dll!WebCore::ScriptRunner::timerFired(WebCore::Timer<WebCore::ScriptRunner> * timer) Line 121 + 0x2a5 bytes C++ WebKit.dll!WebCore::Timer<WebCore::Settings>::fired() Line 114 + 0xb bytes C++ WebKit.dll!WebCore::ThreadTimers::sharedTimerFiredInternal() Line 132 C++ WebKit.dll!WebCore::TimerWindowWndProc(HWND__ * hWnd, unsigned int message, unsigned int wParam, long lParam) Line 111 C++ user32.dll!_InternalCallWinProc@20() + 0x23 bytes user32.dll!_UserCallWinProcCheckWow@36() + 0xbd bytes user32.dll!_DispatchMessageWorker@8() + 0xf8 bytes user32.dll!_DispatchMessageW@4() + 0x10 bytes CoreFoundation.dll!__CFRunLoopRun(__CFRunLoop * rl, __CFRunLoopMode * rlm, double seconds, unsigned char stopAfterHandle, __CFRunLoopMode * previousMode) Line 42292 C++ CoreFoundation.dll!CFRunLoopRunSpecific(__CFRunLoop * rl, const __CFString * modeName, double seconds, unsigned char returnAfterSourceHandled) Line 42413 + 0x12 bytes C++ CoreFoundation.dll!CFRunLoopRun() Line 42440 + 0x1d bytes C++ WinLauncher.dll!dllLauncherEntryPoint(HINSTANCE__ * __formal, HINSTANCE__ * __formal, HINSTANCE__ * __formal, int nCmdShow) Line 456 C++ WinLauncher.exe!004018b8() [Frames below may be incorrect and/or missing, no symbols loaded for WinLauncher.exe] msvcr100.dll!_free() + 0x1c bytes msvcr100.dll!__wsetenvp() + 0xa2 bytes msvcr100.dll!___wgetmainargs() + 0x53 bytes WinLauncher.exe!004024c9() WinLauncher.exe!00402636() kernel32.dll!@BaseThreadInitThunk@12() + 0xe bytes ntdll.dll!___RtlUserThreadStart@8() + 0x27 bytes ntdll.dll!__RtlUserThreadStart@8() + 0x1b bytes 917273 1 bfulgham 2013-08-14 13:47:41 -0700 Chris, the 'crash' under Bug #119803 is simply an assertion. We have debug asserts perform a WTFCrash, which stops in the debugger so you can see what's going on. I mentioned it in my original bug because I thought it might be a clue to the problem you resolved here. 917275 2 chris_curtis 2013-08-14 13:51:52 -0700 (In reply to comment #1) > Chris, the 'crash' under Bug #119803 is simply an assertion. We have debug asserts perform a WTFCrash, which stops in the debugger so you can see what's going on. > > I mentioned it in my original bug because I thought it might be a clue to the problem you resolved here. This is true but that assert happens in a completely different area that is unrelated to the release crash. The patch will fix the release crash, but the debug assert crash will still happen. 917355 3 208776 chris_curtis 2013-08-14 17:21:55 -0700 Created attachment 208776 patch Added a check to make sure the JSValue was an object before casting it. 917358 4 208776 oliver 2013-08-14 17:29:57 -0700 Comment on attachment 208776 patch What is the type? Is it not a cell? If it is a cell why class info does it say it has? 917384 5 208776 darin 2013-08-14 18:36:30 -0700 Comment on attachment 208776 patch View in context: https://bugs.webkit.org/attachment.cgi?id=208776&action=review We need a regression test for this. Or an explanation of why you can’t make one. > Source/JavaScriptCore/runtime/ExceptionHelpers.cpp:109 > + return jsString(exec, asObject(v)->methodTable()->className(asObject(v))); Should use “object” here instead of “asObject(v)”. > Source/JavaScriptCore/runtime/ExceptionHelpers.cpp:111 > + return jsString(exec, ""); Should be: return vm->smallStrings.emptyString(); But why specifically the empty string? 917400 6 chris_curtis 2013-08-14 20:21:10 -0700 (In reply to comment #4) > (From update of attachment 208776 [details]) > What is the type? Is it not a cell? If it is a cell why class info does it say it has? It has no type. The JSValue in this case only evaluates as a primitive. All other types evaluate as false. It could be something unique with the site because it tests features of html5 that are not yet supported. The other unique thing is that this is only present on windows. I can not duplicate error on mac. 917401 7 oliver 2013-08-14 20:24:02 -0700 (In reply to comment #6) > (In reply to comment #4) > > (From update of attachment 208776 [details] [details]) > > What is the type? Is it not a cell? If it is a cell why class info does it say it has? > > It has no type. The JSValue in this case only evaluates as a primitive. All other types evaluate as false. It could be something unique with the site because it tests features of html5 that are not yet supported. The other unique thing is that this is only present on windows. I can not duplicate error on mac. No, everything must have a type. Hence my question: does value.isCell() return true? If it does it could be that we're incorrectly claiming that it's guaranteed to be a cell and so we're blindly replacing the tag with CellTag. If it is a real cell it must have a structure, and the structure has a class info. Have you tried reproducing on 32bit mac? 917402 8 chris_curtis 2013-08-14 20:26:35 -0700 (In reply to comment #7) > (In reply to comment #6) > > (In reply to comment #4) > > > (From update of attachment 208776 [details] [details] [details]) > > > What is the type? Is it not a cell? If it is a cell why class info does it say it has? > > > > It has no type. The JSValue in this case only evaluates as a primitive. All other types evaluate as false. It could be something unique with the site because it tests features of html5 that are not yet supported. The other unique thing is that this is only present on windows. I can not duplicate error on mac. > > No, everything must have a type. Hence my question: does value.isCell() return true? If it does it could be that we're incorrectly claiming that it's guaranteed to be a cell and so we're blindly replacing the tag with CellTag. If it is a real cell it must have a structure, and the structure has a class info. > > Have you tried reproducing on 32bit mac? "Hence my question: does value.isCell() return true?" No, value.isCell() evaluates false. I reduced it down to the 32 bit build on the mac side, and still mac does not fail. 917404 9 chris_curtis 2013-08-14 20:43:41 -0700 (In reply to comment #5) > (From update of attachment 208776 [details]) > > Source/JavaScriptCore/runtime/ExceptionHelpers.cpp:111 > > + return jsString(exec, ""); > > Should be: > > return vm->smallStrings.emptyString(); > > But why specifically the empty string? At this point in the code if the JSValue had any type, it would have been picked out. This piece of code is used to stringily a JSValue for the error object message. There is also a stipulation that this function cannot be allowed to throw another error so the normal method of calling toString() can not be used. The empty string was selected because there was no type to send back. 919184 10 209223 chris_curtis 2013-08-20 13:14:06 -0700 Created attachment 209223 winLauncher Patch This patch will fix the problem on win launcher that is causing it to crash. I have created a separate bugzilla report to look into why the JSValue isEmpty() in the first place. 919185 11 chris_curtis 2013-08-20 13:15:05 -0700 (In reply to comment #10) > Created an attachment (id=209223) [details] > winLauncher Patch > > This patch will fix the problem on win launcher that is causing it to crash. I have created a separate bugzilla report to look into why the JSValue isEmpty() in the first place. The new bugzilla bug for follow up: https://bugs.webkit.org/show_bug.cgi?id=120080 919196 12 209223 ggaren 2013-08-20 13:41:48 -0700 Comment on attachment 209223 winLauncher Patch View in context: https://bugs.webkit.org/attachment.cgi?id=209223&action=review > Source/JavaScriptCore/ChangeLog:8 > + Added a check to make sure that the JSValue was an object before casting it as an object. Also, In case the JSValue was has no type "Also, In" should be "Also, in". JSValues always have a type. The type in question here is the empty value ("JSValue()"). > Source/JavaScriptCore/runtime/ExceptionHelpers.cpp:113 > + // FIXME: https://bugs.webkit.org/show_bug.cgi?id=120080 The JSValue here should never be empty here. "here" is twice in this sentence. 919219 13 209228 chris_curtis 2013-08-20 14:05:09 -0700 Created attachment 209228 patch2 919934 14 209228 ggaren 2013-08-22 12:21:30 -0700 Comment on attachment 209228 patch2 r=me 919940 15 209228 commit-queue 2013-08-22 12:43:13 -0700 Comment on attachment 209228 patch2 Clearing flags on attachment: 209228 Committed r154457: <http://trac.webkit.org/changeset/154457> 919941 16 commit-queue 2013-08-22 12:43:16 -0700 All reviewed patches have been landed. Closing bug. 919967 17 joepeck 2013-08-22 14:24:34 -0700 Why no test? Darin even mentioned we should have a test in an earlier review. 208776 2013-08-14 17:21:55 -0700 2013-08-20 13:14:06 -0700 patch fix.diff text/plain 1737 chris_curtis SW5kZXg6IFNvdXJjZS9KYXZhU2NyaXB0Q29yZS9DaGFuZ2VMb2cKPT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PQotLS0gU291 cmNlL0phdmFTY3JpcHRDb3JlL0NoYW5nZUxvZwkocmV2aXNpb24gMTU0MDgxKQorKysgU291cmNl L0phdmFTY3JpcHRDb3JlL0NoYW5nZUxvZwkod29ya2luZyBjb3B5KQpAQCAtMSwzICsxLDE3IEBA CisyMDEzLTA4LTE0ICBDaHJpcyBDdXJ0aXMgIDxjaHJpc19jdXJ0aXNAYXBwbGUuY29tPgorCisg ICAgICAgIGVycm9yRGVzY3JpcHRpb25Gb3JWYWx1ZSgpIHNob3VsZCBub3QgYXNzdW1lIGVycm9y IHZhbHVlIGlzIGFuIE9iamVjdAorICAgICAgICBodHRwczovL2J1Z3Mud2Via2l0Lm9yZy9zaG93 X2J1Zy5jZ2k/aWQ9MTE5ODEyCisgICAgICAgIAorICAgICAgICBSZXZpZXdlZCBieSBOT0JPRFkg KE9PUFMhKS4KKworICAgICAgICBUaGUgSlNWYWx1ZSBpcyB0cnlpbmcgdG8gYmUgc3RyaW5naWZp ZWQgaW4gYSBzYWZlIG1hbm5lciBmb3IgYW4gZXJyb3Igb2JqZWN0IG1lc3NhZ2UuCisgICAgICAg IFRoZSBsYXN0IGNhc2UgYXNzdW1lZCB0aGF0IGl0IHdhcyBhbiBvYmplY3QuIEZvciB0aGlzIGJ1 ZywgdGhlIEpTVmFsdWUgZXZhbHVhdGVkIGFzIAorICAgICAgICBwcmltYXRpdmUgb25seS4gCisg ICAgICAgIEFkZGVkIGEgbGFzdCByZXNvcnQgY2FzZSB0aGF0IHJldHVybnMgYW4gZW1wdHkgc3Ry aW5nLiAKKyAgICAgICAgKiBydW50aW1lL0V4Y2VwdGlvbkhlbHBlcnMuY3BwOgorICAgICAgICAo SlNDOjplcnJvckRlc2NyaXB0aW9uRm9yVmFsdWUpOgorCiAyMDEzLTA4LTE0ICBNaWNoYWVsIFNh Ym9mZiAgPG1zYWJvZmZAYXBwbGUuY29tPgogCiAgICAgICAgIEZpeGVkIGppdCBvbiBXaW42NC4K SW5kZXg6IFNvdXJjZS9KYXZhU2NyaXB0Q29yZS9ydW50aW1lL0V4Y2VwdGlvbkhlbHBlcnMuY3Bw Cj09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT0KLS0tIFNvdXJjZS9KYXZhU2NyaXB0Q29yZS9ydW50aW1lL0V4Y2VwdGlvbkhl bHBlcnMuY3BwCShyZXZpc2lvbiAxNTQwNjMpCisrKyBTb3VyY2UvSmF2YVNjcmlwdENvcmUvcnVu dGltZS9FeGNlcHRpb25IZWxwZXJzLmNwcAkod29ya2luZyBjb3B5KQpAQCAtMTA2LDggKzEwNiw5 IEBAIEpTU3RyaW5nKiBlcnJvckRlc2NyaXB0aW9uRm9yVmFsdWUoRXhlY1MKICAgICAgICAgSlNP YmplY3QqIG9iamVjdCA9IGFzT2JqZWN0KHYpOwogICAgICAgICBpZiAob2JqZWN0LT5tZXRob2RU YWJsZSgpLT5nZXRDYWxsRGF0YShvYmplY3QsIGNhbGxEYXRhKSAhPSBDYWxsVHlwZU5vbmUpCiAg ICAgICAgICAgICByZXR1cm4gdm0uc21hbGxTdHJpbmdzLmZ1bmN0aW9uU3RyaW5nKCk7CisgICAg ICAgIHJldHVybiBqc1N0cmluZyhleGVjLCBhc09iamVjdCh2KS0+bWV0aG9kVGFibGUoKS0+Y2xh c3NOYW1lKGFzT2JqZWN0KHYpKSk7CiAgICAgfQotICAgIHJldHVybiBqc1N0cmluZyhleGVjLCBh c09iamVjdCh2KS0+bWV0aG9kVGFibGUoKS0+Y2xhc3NOYW1lKGFzT2JqZWN0KHYpKSk7CisgICAg cmV0dXJuIGpzU3RyaW5nKGV4ZWMsICIiKTsKIH0KICAgICAKIEpTT2JqZWN0KiBjcmVhdGVFcnJv cihFeGVjU3RhdGUqIGV4ZWMsIEVycm9yRmFjdG9yeSBlcnJvckZhY3RvcnksIEpTVmFsdWUgdmFs dWUsIGNvbnN0IFN0cmluZyYgbWVzc2FnZSkK 209223 2013-08-20 13:14:06 -0700 2013-08-20 14:05:09 -0700 winLauncher Patch winPatch.diff text/plain 1896 chris_curtis SW5kZXg6IFNvdXJjZS9KYXZhU2NyaXB0Q29yZS9DaGFuZ2VMb2cKPT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PQotLS0gU291 cmNlL0phdmFTY3JpcHRDb3JlL0NoYW5nZUxvZwkocmV2aXNpb24gMTU0MzUyKQorKysgU291cmNl L0phdmFTY3JpcHRDb3JlL0NoYW5nZUxvZwkod29ya2luZyBjb3B5KQpAQCAtMSwzICsxLDE1IEBA CisyMDEzLTA4LTIwICBDaHJpcyBDdXJ0aXMgIDxjaHJpc19jdXJ0aXNAYXBwbGUuY29tPgorCisg ICAgICAgIGVycm9yRGVzY3JpcHRpb25Gb3JWYWx1ZSgpIHNob3VsZCBub3QgYXNzdW1lIGVycm9y IHZhbHVlIGlzIGFuIE9iamVjdAorICAgICAgICBodHRwczovL2J1Z3Mud2Via2l0Lm9yZy9zaG93 X2J1Zy5jZ2k/aWQ9MTE5ODEyCisKKyAgICAgICAgUmV2aWV3ZWQgYnkgTk9CT0RZIChPT1BTISku CisKKyAgICAgICAgQWRkZWQgYSBjaGVjayB0byBtYWtlIHN1cmUgdGhhdCB0aGUgSlNWYWx1ZSB3 YXMgYW4gb2JqZWN0IGJlZm9yZSBjYXN0aW5nIGl0IGFzIGFuIG9iamVjdC4gQWxzbywgSW4gY2Fz ZSB0aGUgSlNWYWx1ZSB3YXMgaGFzIG5vIHR5cGUKKyAgICAgICAgdGhlIGZ1bmN0aW9uIG5vdyBy ZXR1cm5zIHRoZSBlbXB0eSBzdHJpbmcuIAorICAgICAgICAqIHJ1bnRpbWUvRXhjZXB0aW9uSGVs cGVycy5jcHA6CisgICAgICAgIChKU0M6OmVycm9yRGVzY3JpcHRpb25Gb3JWYWx1ZSk6CisKIDIw MTMtMDgtMjAgIE1hcmsgSGFobmVuYmVyZyAgPG1oYWhuZW5iZXJnQGFwcGxlLmNvbT4KIAogICAg ICAgICA8aHR0cHM6Ly93ZWJraXQub3JnL2IvMTE5OTE5PiBDb25jdXJyZW50IEpJVCBjcmFzaGVz IGluIHZhcmlvdXMgZmFzdC9qcy9kZmctKiB0ZXN0cyB3aGlsZSB0aGUgbWFpbiB0aHJlYWQgaXMg c2V0dGluZyBpbm5lckhUTUwKSW5kZXg6IFNvdXJjZS9KYXZhU2NyaXB0Q29yZS9ydW50aW1lL0V4 Y2VwdGlvbkhlbHBlcnMuY3BwCj09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT0KLS0tIFNvdXJjZS9KYXZhU2NyaXB0Q29yZS9y dW50aW1lL0V4Y2VwdGlvbkhlbHBlcnMuY3BwCShyZXZpc2lvbiAxNTQzNTIpCisrKyBTb3VyY2Uv SmF2YVNjcmlwdENvcmUvcnVudGltZS9FeGNlcHRpb25IZWxwZXJzLmNwcAkod29ya2luZyBjb3B5 KQpAQCAtMTA2LDggKzEwNiwxMiBAQCBKU1N0cmluZyogZXJyb3JEZXNjcmlwdGlvbkZvclZhbHVl KEV4ZWNTCiAgICAgICAgIEpTT2JqZWN0KiBvYmplY3QgPSBhc09iamVjdCh2KTsKICAgICAgICAg aWYgKG9iamVjdC0+bWV0aG9kVGFibGUoKS0+Z2V0Q2FsbERhdGEob2JqZWN0LCBjYWxsRGF0YSkg IT0gQ2FsbFR5cGVOb25lKQogICAgICAgICAgICAgcmV0dXJuIHZtLnNtYWxsU3RyaW5ncy5mdW5j dGlvblN0cmluZygpOworICAgICAgICByZXR1cm4ganNTdHJpbmcoZXhlYywgb2JqZWN0LT5tZXRo b2RUYWJsZSgpLT5jbGFzc05hbWUob2JqZWN0KSk7CiAgICAgfQotICAgIHJldHVybiBqc1N0cmlu ZyhleGVjLCBhc09iamVjdCh2KS0+bWV0aG9kVGFibGUoKS0+Y2xhc3NOYW1lKGFzT2JqZWN0KHYp KSk7CisgICAgCisgICAgQVNTRVJUKHYuaXNFbXB0eSgpKTsKKyAgICAvLyBGSVhNRTogaHR0cHM6 Ly9idWdzLndlYmtpdC5vcmcvc2hvd19idWcuY2dpP2lkPTEyMDA4MCBUaGUgSlNWYWx1ZSBoZXJl IHNob3VsZCBuZXZlciBiZSBlbXB0eSBoZXJlLgorICAgIHJldHVybiB2bS5zbWFsbFN0cmluZ3Mu ZW1wdHlTdHJpbmcoKTsKIH0KICAgICAKIEpTT2JqZWN0KiBjcmVhdGVFcnJvcihFeGVjU3RhdGUq IGV4ZWMsIEVycm9yRmFjdG9yeSBlcnJvckZhY3RvcnksIEpTVmFsdWUgdmFsdWUsIGNvbnN0IFN0 cmluZyYgbWVzc2FnZSkK 209228 2013-08-20 14:05:09 -0700 2013-08-22 12:43:13 -0700 patch2 winPatch.diff text/plain 1914 chris_curtis SW5kZXg6IFNvdXJjZS9KYXZhU2NyaXB0Q29yZS9DaGFuZ2VMb2cKPT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PQotLS0gU291 cmNlL0phdmFTY3JpcHRDb3JlL0NoYW5nZUxvZwkocmV2aXNpb24gMTU0MzUyKQorKysgU291cmNl L0phdmFTY3JpcHRDb3JlL0NoYW5nZUxvZwkod29ya2luZyBjb3B5KQpAQCAtMSwzICsxLDE1IEBA CisyMDEzLTA4LTIwICBDaHJpcyBDdXJ0aXMgIDxjaHJpc19jdXJ0aXNAYXBwbGUuY29tPgorCisg ICAgICAgIGVycm9yRGVzY3JpcHRpb25Gb3JWYWx1ZSgpIHNob3VsZCBub3QgYXNzdW1lIGVycm9y IHZhbHVlIGlzIGFuIE9iamVjdAorICAgICAgICBodHRwczovL2J1Z3Mud2Via2l0Lm9yZy9zaG93 X2J1Zy5jZ2k/aWQ9MTE5ODEyCisKKyAgICAgICAgUmV2aWV3ZWQgYnkgTk9CT0RZIChPT1BTISku CisKKyAgICAgICAgQWRkZWQgYSBjaGVjayB0byBtYWtlIHN1cmUgdGhhdCB0aGUgSlNWYWx1ZSB3 YXMgYW4gb2JqZWN0IGJlZm9yZSBjYXN0aW5nIGl0IGFzIGFuIG9iamVjdC4gQWxzbywgaW4gY2Fz ZSB0aGUgcGFyYW1ldGVyaXplZCBKU1ZhbHVlCisgICAgICAgIGhhcyBubyB0eXBlLCB0aGUgZnVu Y3Rpb24gbm93IHJldHVybnMgdGhlIGVtcHR5IHN0cmluZy4gCisgICAgICAgICogcnVudGltZS9F eGNlcHRpb25IZWxwZXJzLmNwcDoKKyAgICAgICAgKEpTQzo6ZXJyb3JEZXNjcmlwdGlvbkZvclZh bHVlKToKKwogMjAxMy0wOC0yMCAgTWFyayBIYWhuZW5iZXJnICA8bWhhaG5lbmJlcmdAYXBwbGUu Y29tPgogCiAgICAgICAgIDxodHRwczovL3dlYmtpdC5vcmcvYi8xMTk5MTk+IENvbmN1cnJlbnQg SklUIGNyYXNoZXMgaW4gdmFyaW91cyBmYXN0L2pzL2RmZy0qIHRlc3RzIHdoaWxlIHRoZSBtYWlu IHRocmVhZCBpcyBzZXR0aW5nIGlubmVySFRNTApJbmRleDogU291cmNlL0phdmFTY3JpcHRDb3Jl L3J1bnRpbWUvRXhjZXB0aW9uSGVscGVycy5jcHAKPT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PQotLS0gU291cmNlL0phdmFT Y3JpcHRDb3JlL3J1bnRpbWUvRXhjZXB0aW9uSGVscGVycy5jcHAJKHJldmlzaW9uIDE1NDM1MikK KysrIFNvdXJjZS9KYXZhU2NyaXB0Q29yZS9ydW50aW1lL0V4Y2VwdGlvbkhlbHBlcnMuY3BwCSh3 b3JraW5nIGNvcHkpCkBAIC0xMDYsOCArMTA2LDEyIEBAIEpTU3RyaW5nKiBlcnJvckRlc2NyaXB0 aW9uRm9yVmFsdWUoRXhlY1MKICAgICAgICAgSlNPYmplY3QqIG9iamVjdCA9IGFzT2JqZWN0KHYp OwogICAgICAgICBpZiAob2JqZWN0LT5tZXRob2RUYWJsZSgpLT5nZXRDYWxsRGF0YShvYmplY3Qs IGNhbGxEYXRhKSAhPSBDYWxsVHlwZU5vbmUpCiAgICAgICAgICAgICByZXR1cm4gdm0uc21hbGxT dHJpbmdzLmZ1bmN0aW9uU3RyaW5nKCk7CisgICAgICAgIHJldHVybiBqc1N0cmluZyhleGVjLCBv YmplY3QtPm1ldGhvZFRhYmxlKCktPmNsYXNzTmFtZShvYmplY3QpKTsKICAgICB9Ci0gICAgcmV0 dXJuIGpzU3RyaW5nKGV4ZWMsIGFzT2JqZWN0KHYpLT5tZXRob2RUYWJsZSgpLT5jbGFzc05hbWUo YXNPYmplY3QodikpKTsKKyAgICAKKyAgICBBU1NFUlQodi5pc0VtcHR5KCkpOworICAgIC8vIEZJ WE1FOiBodHRwczovL2J1Z3Mud2Via2l0Lm9yZy9zaG93X2J1Zy5jZ2k/aWQ9MTIwMDgwIFRoZSBK U1ZhbHVlIHNob3VsZCBuZXZlciBiZSBlbXB0eSBpbiB0aGlzIGZ1bmN0aW9uLgorICAgIHJldHVy biB2bS5zbWFsbFN0cmluZ3MuZW1wdHlTdHJpbmcoKTsKIH0KICAgICAKIEpTT2JqZWN0KiBjcmVh dGVFcnJvcihFeGVjU3RhdGUqIGV4ZWMsIEVycm9yRmFjdG9yeSBlcnJvckZhY3RvcnksIEpTVmFs dWUgdmFsdWUsIGNvbnN0IFN0cmluZyYgbWVzc2FnZSkK