Develop Reference

Marshalling exception with code from the ShareTarget second window

I'm having a hard time with some code I have that apparently struggles when called from the second window created by the ShareTarget contract (when you share something to the app, it opens in a small standalone window).
This is my code so far:
// Blur and resize the image to get the average HSL color
// Assume that stream is an IRandomAccessStream pointing to valid image data
HslColor hslMean;
using (RandomAccessStreamImageSource imageProvider = new RandomAccessStreamImageSource(stream))
using (BlurEffect blurEffect = new BlurEffect(imageProvider) { KernelSize = 256 })
{
Color mean = await DispatcherHelper.GetFromUIThreadAsync(async () =>
{
WriteableBitmap
blurred = new WriteableBitmap((int)decoder.PixelWidth, (int)decoder.PixelHeight),
result = await blurEffect.GetBitmapAsync(blurred, OutputOption.Stretch),
resized = result.Resize(1, 1, WriteableBitmapExtensions.Interpolation.Bilinear);
return resized.GetPixel(0, 0);
});
hslMean = mean.ToHsl();
}
Note: that DispatcherHelper.GetFromUIThreadAsync method just checks the thread access to the UI thread, and if needed it schedules the code to a CoreDispatcher object that was obtained with CoreApplication.MainView.CoreWindow.Dispatcher.
Problem: this code works 100% fine if my app is already open, as at that point that CoreDispatcher object has already been created by previous calls to that DispatcherHelper class, so the method just uses the stored dispatcher to schedule the work and it works fine. But, if the app is closed when the ShareTarget window is opened (so that DispatcherHelper has to create the dispatcher for the first time) the CoreApplication.MainView.CoreWindow line throws an exception. A very weird one:
COMException:
A COM call to an ASTA was blocked because the call chain originated in or passed through another ASTA. This call pattern is deadlock-prone and disallowed by apartment call control.
A COM call (IID: {638BB2DB-451D-4661-B099-414F34FFB9F1}, method index: 6) to an ASTA (thread 10276) was blocked because the call chain originated in or passed through another ASTA (thread 4112). This call pattern is deadlock-prone and disallowed by apartment call control.
So, I needed a way to make that method reliable even when being called from different windows. I've tried different options:
#1: Just invoking that code without dispatching to a different thread, as in theory I should be on the UI thread at this point ---> FAIL (The application called an interface that was marshalled for a different thread. (Exception from HRESULT: 0x8001010E (RPC_E_WRONG_THREAD)))
#2: Manually calling CoreApplication.MainView.CoreWindow.Dispatcher to dispatch that code block ---> FAIL (I get that weird COMException mentioned above)
#3: Manually using CoreApplication.MainView.Dispatcher to dispatch the code block (as it was the .CoreWindow part that spawned the exception) ---> FAIL (COMException: item not found)
#4: Using CoreApplication.GetCurrentView().CoreWindow.Dispatcher, CoreApplication.GetCurrentView().Dispatcher, Window.Current.CoreWindow.Dispatcher and Window.Current.Content.Dispatcher to schedule that code ---> FAIL (wrong thread again, I get the usual marshalling exception)
All these marshalling exception are thrown at the line result = await blurEffect.GetBitmapAsync(blurred, OutputOption.Stretch), so I suspect it might be something related to the Lumia Imaging SDK.
I mean, I'm quite sure that I am in fact on the UI thread, or otherwise I wouldn't have managed to create an instance of the WriteableBitmap class, right?
Why is it that I can create WriteableBitmap objects (and they need to be created on the UI thread as far as I know), but that GetBitmapAsync method from the Lumia SDK always throws that marshalling exception? I'm using it everywhere in my app without any problems, why is it that it just won't work from the ShareTarget window? Is there anything I need to do?
Thanks for your help!

Looks like this is a bug in the Lumia Imaging SDK (that was originally written for WP8.1, which didn't have multiple windows/dispatchers), so unless the call to the library is made from the dispatcher associated with the main app window (which of course can only be retrieved if the app is open in the background when the ShareTarget window pops up), it will just fail.
The only solution at this point is to replace that call to the Lumia SDK with some other code that doesn't rely on that particular library (in this case for example, it is possible to just get the ARGB array from the WriteableBitmap object and calculate the mean color manually).

uglify crashing after running child_process.execFile

EDIT 2
I "solved" the problem, but I don't want to post it as an answer b/c it doesn't explain what actually happened. In the code for the .NET resourceReader.exe I use
Console.OutputEncoding = System.Text.Encoding.UTF8;
to output the internationalized resources to stdout in unicode. If I reset the encoding at the end of my program with
Console.OutputEncoding = System.Text.Encoding.Default;
then I don't get any errors in Node. If I don't reset it, I get the error described in the original question. It seems that .NET is somehow messing up some of the output encoding settings on the cmd.exe and causing the subsequent node run to fail!
EDIT
I narrowed down the error to being caused by resourceReader.exe. It's a.NET program which reads some resource streams out of the .NET assembly and prints them to the stdout using Console.WriteLine. I added Console.OutputEncoding = System.Text.Encoding.UTF8 to resourceReader.exe because some of the resources are in non ASCII letters and that's whats causing the crash in grunt!
If I take that line out, the task doesn't crash, but the resources show up in non printable ASCII characters! Also, the crash only happens if I actually print non-ASCII to sdtout. If I don't print them, it doesn't error.
ORIGINAL
I added a step to my Gruntfile which uses child_process.execFile to run an read some data from an external program and uses it in the build. Now whenever I run my build, it runs fine the first time, but crashes the second time!
Here's the output from the crash (this is during the uglify task):
File build/Scripts/NDB.contacts.min.js created: 16.85 kBevents.js:85
throw er; // Unhandled 'error' event
^
Error: This socket is closed.
at WriteStream.Socket._writeGeneric (net.js:656:19)
at WriteStream.Socket._write (net.js:709:8)
at doWrite (_stream_writable.js:301:12)
at writeOrBuffer (_stream_writable.js:288:5)
at WriteStream.Writable.write (_stream_writable.js:217:11)
at WriteStream.Socket.write (net.js:634:40)
at Log._write (C:\...\node_modules\grunt\node_modules\grunt-legacy-log\index.js:161:26)
at Log.wrapper [as _write] (C:\...\node_modules\grunt\node_modules\grunt-legacy-log\node_modules\lodash\index.js:3095:19)
at Log._writeln (C:\...\node_modules\grunt\node_modules\grunt-legacy-log\index.js:166:8)
at Log.wrapper [as _writeln] (C:\...\node_modules\grunt\node_modules\grunt-legacy-log\node_modules\lodash\index.js:3095:19)
at Log.writeln (C:\...\node_modules\grunt\node_modules\grunt-legacy-log\index.js:177:8)
at Log.wrapper (C:\...\node_modules\grunt\node_modules\grunt-legacy-log\node_modules\lodash\index.js:3095:19)
at writeln (C:\...\node_modules\grunt\lib\grunt\fail.js:30:13)
at Object.fail.fatal (C:\...\node_modules\grunt\lib\grunt\fail.js:46:3)
at process.uncaughtHandler (C:\...\node_modules\grunt\lib\grunt.js:121:10)
at process.emit (events.js:129:20)
at process._fatalException (node.js:236:26)
at Task.runTaskFn (C:\...\node_modules\grunt\lib\util\task.js:250:7)
at Task.<anonymous> (C:\...\node_modules\grunt\lib\util\task.js:293:12)
at C:\...\node_modules\grunt\lib\util\task.js:220:11
at process._tickCallback (node.js:355:11)
Here's the code for the task which uses child_process.
function readAllCultures() {
var readDeferred = q.defer();
childProc.execFile("../tools/resourceReader.exe", function (err, stdout, stderr) {
if (err) throw new Error(err);
var cultures = JSON.parse(stdout);
readDeferred.resolve(cultures);
});
return readDeferred.promise;
}
Here's some things I discovered debugging that might be helpful
If I redirect the output of grunt (using either > filename or | process) it runs fine
When I redirect output, I never see the message from uglify that it created the main output only that it created the source map.
If I close and reopen my command prompt (cmd.exe) it works fine
I added a listener to the exit and close events of the child process using rdr.on("close", function() { console.log("close"); }); and the same for exit. Both events fire as expected in the first run.
Using Process Explorer, I can see node.exe open under the command prompt and close again when my command finishes running. There are no processes visibly "left open" under the command prompt.

The fact that the stack trace is ultimately showing a socket.write error is interesting. Nothing in the code you've provided suggests you are trying to write to a socket.
If we follow the stack down, we can see that it's actually grunt that is trying to write to a socket, as it's attempting to log an uncaught exception.
So, first your grunt task throws a currently unknown error, and then grunt itself errors because it can't tell you about it.
I would first try and log the error that occurs in your child process. Currently if there is an error present, you simply throw it without finding out what it is. It's likely that this is what grunt is trying and failing to log.
Replace
if (err) throw new Error(err);
With
if (err) console.error(err);
This will hopefully avoid the socket.write problem and give you specific information about what error is happening on the child process. What do you see?
Second, I would try using child_process.exec instead of child_process.execFile. This will spawn a shell and run resourceReader.exe inside of it (instead of running it directly). This might help avoid any problems you are encountering with the command still running/failing in the background, which could be the cause of the socket.write error.

Why Navigation throws null exception when returning to a page 2nd time

I have a serial port controller (for sensing any object that passes through its transport module) connected to USB. It responses with either 0 or 1 after input command t when it senses any dropping or no dropping of object through the transport. The page CshDeposit5 contains the code. Start.xaml is the page the activity starts and ends up in CshDeposit5.xaml as follows:
Start.xaml->Lang->Menu->CshDeposit5.xaml->TmpMsg.xaml or TmpErrorMsg.xaml->Start.xaml
TmpErrorMsg and TmpMsg are the pages to show the user what happened. Everything runs fine for the first time. But when it tries to come to this CshDeposit5.xaml 2nd time it throws exception on the following codes:
Dispatcher.BeginInvoke(new System.Threading.ThreadStart(delegate
{
NavigationService.Navigate(new Uri("TmpMsg.xaml", UriKind.RelativeOrAbsolute));
}));

Strange 'Index was out of range' (Multithreading issue)

I'm writing a Windows service in C# (Visual Studio 2013).
The service activates a thread for each server it connects with.
The 'server' object is defined as 'ThreadStatic' (so each thread should access it as a different object):
[ThreadStatic]
public static Server CurServer
On debug- I sometimes get the error:
ArgumentOutOfRangeException
Index was out of range. Must be non-negative and less than the size of the collection.
It occures in the following line (inside the method of the 'server' thread):
EventConn.FindString = G.CurServer.FilesList[G.CurServer.nFilesIndex].SearchString;
But the strange thing is that the debugger shows that the values are ok:
G.CurServer.FilesList.count = 1
G.CurServer.nFilesIndex = 0
So there should'nt be any error!!!
When I press F11 (debugger step), it keeps on debugging as if everything is fine and the assignment also works...
Why????? :0
Is it a bug in the Visual Studio that prompts the error before assigning the current thread's values? Or that I'm not using threads saftley (more likely)?

I can't explain the bad display of VS, but I found the cause of the error:
I used object assignment in the thread's method like this:
G.CurServer = server;
Instead of cloning the object like this (How to Clone Objects):
G.CurServer = (Server).Clone();
This first assignment caused values of G.CurServer to change in the different threads even though it is [ThreadStatic].
I hope it helps someone...
Best wishes, Eyal. http://reginiano.com

Application that uses WebBrowser control crashes after installing IE9

I've installed IE 9 last week and since, my c# .net application crashes about 20% of times. The debugger is unable to show something useful besides stopping at Program.cs Application.Run(new MyMainForm()); , btw the main form was already shown, so it's not that it's something in construction on main form.
I have Windows7.
The exception thrown is:
"An unhandled exception of type 'System.ArgumentException' occurred in System.Windows.Forms.dll
Additional information: Value does not fall within the expected range.
Screen shot of callstack -> http://img861.imageshack.us/f/ie9v.png/
When running outside of debugger, this info is shown:
Problem signature:
Problem Event Name: APPCRASH
Application Name: myexe.exe
Application Version: 6.7.6.0
Application Timestamp: 4d7fdffd
Fault Module Name: mshtml.dll
Fault Module Version: 9.0.8112.16421
Fault Module Timestamp: 4d76266c
Exception Code: c0000005
Exception Offset: 0012c848
OS Version: 6.1.7600.2.0.0.256.48
Locale ID: 1037
And sometimes instead of mshtml.dll it says StackHash_f09d
Problem Event Name: APPCRASH
Application Name: myexe.exe
Application Version: 6.7.6.0
Application Timestamp: 4d7fdffd
Fault Module Name: StackHash_f09d
Fault Module Version: 0.0.0.0
Fault Module Timestamp: 00000000
Exception Code: c0000005
Exception Offset: 00000000
OS Version: 6.1.7600.2.0.0.256.48
Locale ID: 1037
Thanks in advance
Edited:
That's what I see in windbg, with symbols:
0:000> kb
ChildEBP RetAddr Args to Child
0020eda4 64d54f83 0566c988 00001012 00000000 mshtml!CDoc::ReduceMemoryPressureTask+0x1a
0020edb4 64d54f2c c6b991e4 0020ee78 00000113 mshtml!GWYieldToMsgOnCurrentThread+0x17b
0020edfc 770086ef 00192392 00000012 0000201b mshtml!GlobalWndProc+0x1f2
0020ee28 77008876 64d54afe 00192392 00000113 USER32!InternalCallWinProc+0x23
0020eea0 770089b5 00000000 64d54afe 00192392 USER32!UserCallWinProcCheckWow+0x14b
0020ef00 77008e9c 64d54afe 00000000 0020ef2c USER32!DispatchMessageWorker+0x35e
0020ef10 03b54726 0020ef9c fa69a961 00000000 USER32!DispatchMessageW+0xf
*** WARNING: Unable to verify checksum for C:\Windows\assembly\NativeImages_v2.0.50727_32\System.Windows.Forms\f92c882fd4e7005c005e208daa04c28d\System.Windows.Forms.ni.dll
WARNING: Frame IP not in any known module. Following frames may be wrong.
0020ef2c 5af78aee 01b743e4 00000001 01ac95cc 0x3b54726
0020efe0 5af78757 00000000 ffffffff 00000000 System_Windows_Forms_ni+0x208aee
0020f038 5af785a1 01b6c610 1f3a000e 00000000 System_Windows_Forms_ni+0x208757
0020f068 5af35911 01bb7d84 0020f10c 003b73d8 System_Windows_Forms_ni+0x2085a1
0020f0e0 6f221b5c 015b1141 00000001 0020f170 System_Windows_Forms_ni+0x1c5911
0020f0f0 6f232209 0020f1c0 00000000 0020f190 mscorwks!CallDescrWorker+0x33
0020f170 6f246511 0020f1c0 00000000 0020f190 mscorwks!CallDescrWorkerWithHandler+0xa3
0020f2b4 6f246544 0032c040 0020f380 0020f34c mscorwks!MethodDesc::CallDescr+0x19c
0020f2d0 6f246562 0032c040 0020f380 0020f34c mscorwks!MethodDesc::CallTargetWorker+0x1f
0020f2e8 6f2b0c45 0020f34c d847bc11 00000000 mscorwks!MethodDescCallSite::CallWithValueTypes+0x1a
0020f44c 6f2b0b65 003239c0 00000001 0020f488 mscorwks!ClassLoader::RunMain+0x223
0020f6b4 6f2b10b5 00000000 d847b3d9 00000001 mscorwks!Assembly::ExecuteMainMethod+0xa6
0020fb84 6f2b129f 013a0000 00000000 d847b389 mscorwks!SystemDomain::ExecuteMainMethod+0x456
=====UPDATE(I'm not sure, if I'm supposed to post it as "Answer your question")============
Thanks to everyone trying to help, I appreciate it.
Being desperate, I started to remove pieces of code to understand which part of my code affects it(we use webrowser control in many forms). After removing call to LoginForm which also uses webrowser control, the problem disappeared.
The login form hosts webrowser control, it navigates to certain url, for example /login.php and if user is already logged-in, method UserLoggedIn inside form is invoked from html using ObjectForScripting. When UserLoggedIn was called, we were calling Close() to close form if the LoginForm was shown. Even though LoginForm was doing all this, we weren't always showing it. We were, showing it only if after X seconds UserLoggedIn() was not called(i.e user needs to login).
For some reason, and thanks to MS for making us able to debug into .net sources, when we were calling Close, and the form was not Visible it was actually disposing the form and all it's children because IsHandleCreated was false. Now, the Close was called from UserLoggedIn(), which is an event fired by the browser control(the callstack shows ieframe.dll, mshtml.dll etc) , so the webbrowser object was being destoyed while being called from.
The hacky way to resolve this, was to call Close, only if form was Visible. BTW, I don't know, why IsHandleCreated is false, if we don't show the form. I tried to reproduce it, by writing a sample that creates a form, which is not shown, but it's IsHandleCreated is true.
---- from Forms.cs -----
public void Close()
{
if (GetState(STATE_CREATINGHANDLE))
throw new InvalidOperationException(SR.GetString(SR.ClosingWhileCreatingHandle, "Close"));
if (IsHandleCreated) {
closeReason = CloseReason.UserClosing;
SendMessage(NativeMethods.WM_CLOSE, 0, 0);
}
else{
// MSDN: When a form is closed, all resources created within the object are closed and the form is disposed.
// For MDI child: MdiChildren collection gets updated (VSWhidbey# 368642 & 93550)
Dispose(); // THIS WAS CALLED WHEN FORM WAS NOT VISIBLE
}
}

Educated speculation to get you started- timers use thread pool threads, which are MTA. If I had to guess I'd say that it could be related to that, since activex controls can only be safely instantiated from STA threads. Perhaps try creating a thread manually rather than using the timer thread?

browser.Navigate("xxx") is called from timer
What kind of timer are you using? If the timer fire callback happens on another thread, and you are calling methods on the browser object, then you are opening yourself up random crash likelyhood. Try setting Control.CheckForIllegalCrossThreadCalls = true; in your Main() method beofre running the Application.Run(...) method and see if when it crashes, you get a more "on point" error message or the same.

This crash happens when jscript.dll is incorrectly registered. The ReduceMemoryPressureTask method expects it to be loaded and causes a null dereference otherwise. A fix is to run regsvr32 jscript.dll on the affected machine. You can check for the error by looking at the loaded module list to see if jscript.dll is missing or not. This can be done either programatically or through Process Explorer.

The exception code c0000005 usually means that is a memory problem. Maybe it has to do with 32 and 64 bits areas of your application getting confused.
I would try to make your app run fully in 64, then fully 32, see what happens.
Try running as admin, see if that changes anything.
Also you should put try catches on the points where it makes sense, and rety whenever the error is found.
Well those are suggestions, so try it out and see if you can get us more info.
Good Luck!

There might be a reason, since the WinForms runs under STA , but the web browser requires MTA. Put a try-catch above the browser control and also check if there is r/w access voilation. As stated by oakcool

I've run in to problems with this, mostly due to privilege violations.
The program was running at administrator level and the browser was running at user level, or visa-versa.
Something to look into

i get the same error/crash since i got the ie9 installed. it only happens whenever there is a youtube movie on the page. than it hangs/crashes with the exact same window you get Jack Juiceson.
but behind the "Fault Module Name" i dont get mshtml.dll, i get a d3d9.dll or something like that. veryyyyy odd.