Search found 19 matches
- Wed Apr 03, 2013 11:54 am
- Forum: madExcept
- Topic: Access violation in madStackTrace with possible fix
- Replies: 1
- Views: 3327
Access violation in madStackTrace with possible fix
Hello Madshi, Recently i've been using madExcept very extensively to generate bugreports and it has been a great help in solving bugs in our application. Still on occassion (maybe once every 20 bugreports) part of the stacktrace seems to fail. I've been able to track this down to the following code ...
- Thu Jan 24, 2013 3:02 pm
- Forum: madExcept
- Topic: NewException and RegisterBugReportCallBack
- Replies: 1
- Views: 3471
NewException and RegisterBugReportCallBack
Hello madshi, I'm sorry for having to trouble you yet again but i can not figure out how to correctly use RegisterBugReportCallBack in combination with NewException. I have tried the following: lvMEException := NewException(etNormal); lvMEException.RegisterBugReportCallBack(CallBack, True); showmess...
- Wed Jan 23, 2013 12:06 pm
- Forum: madExcept
- Topic: GetBugReport threadsafe?
- Replies: 10
- Views: 9389
Re: GetBugReport threadsafe?
I suppose that when you want all your calls to madexcept to be executed from a different thread (compared to where the exception occured) this would be a problem and you would have to done any of those things you suggested. In my case its just a matter of catching consecutive exceptions. When a thre...
- Tue Jan 22, 2013 3:54 pm
- Forum: madExcept
- Topic: GetBugReport threadsafe?
- Replies: 10
- Views: 9389
Re: GetBugReport threadsafe?
Pretty silly, i already know the answer to my own question. Should call GetBugReport(False) Of course this does work: procedure TForm1.Button1Click(Sender: TObject); var lvMEException : IMeException; lvStringList: TStringList; begin try Raise Exception.Create('test'); except On E: Exception do begin...
- Tue Jan 22, 2013 3:05 pm
- Forum: madExcept
- Topic: GetBugReport threadsafe?
- Replies: 10
- Views: 9389
Re: GetBugReport threadsafe?
I've looked a little futher into this and i think now i have been mistaken. The problem is not one of threading, but as it seems the exception object is being freed before the callstack was generated. And because of that the IMEInterface instance references an invalid pointer. I can reproduce this p...
- Fri Jan 18, 2013 2:21 pm
- Forum: madExcept
- Topic: GetBugReport threadsafe?
- Replies: 10
- Views: 9389
Re: GetBugReport threadsafe?
The code breaked on a call to GetThreadStack (line 14662). I've tried stepping through the code, but of course when i do that i'm influencing the threading and the exception does not occur, or it occurs somewhere else. I'm sorry for not being able to provide more information, but this has me very mu...
- Fri Jan 18, 2013 12:54 pm
- Forum: madExcept
- Topic: GetBugReport threadsafe?
- Replies: 10
- Views: 9389
Re: GetBugReport threadsafe?
During my stress test every exception calls NewException. The TIMEExceptions objects are valid as long as i need them and destroyed when no longer referenced. So this seems allright. Calling GetBugReport on a single IMEException interface instance has never failed on me. It is only when i'm calling ...
- Fri Jan 18, 2013 11:55 am
- Forum: madExcept
- Topic: GetBugReport threadsafe?
- Replies: 10
- Views: 9389
GetBugReport threadsafe?
Hello Madshi, I'm attempting something probably a little onorthodox. In my exceptionhandlers i'm creating IMEExceptions using NewException with the Exception object passed as parameter. For various reasons however i don't want to immediately process the exception. Instead i'm using a thread to captu...
- Fri Jul 13, 2012 1:24 pm
- Forum: madExcept
- Topic: XE 2 Access Violation with MESettings
- Replies: 13
- Views: 11670
Re: XE 2 Access Violation with MESettings
I'm running into a similar problem, though the exception seems to get handled inside madexcept. The following code does run, but if madexcept is not enabled the second showmessage will display a '0' where i would expect it to display '1' showmessage(IntToStr(madExcept.MESettings.AutoClose)); madExce...
- Wed May 23, 2007 2:29 pm
- Forum: madKernel
- Topic: Tracking when an application is "done" with a file
- Replies: 5
- Views: 14619
That sounds really cool, though i must admit that i not know much about api hooking yet. Would it be possible to system wide hook the ntClose and track the handle back to the location of the file? I have managed to get the correct filename from an IHandle interface so obtaining the IHandle interface...
- Tue May 22, 2007 1:23 pm
- Forum: madKernel
- Topic: Tracking when an application is "done" with a file
- Replies: 5
- Views: 14619
Well, to be more specific, i'm working on a document/file management system for multiple users. A user can edit a file after he/she has checked-out the file first (to prevent multiple users from working in the same file). When the file is checked out the appropiate program wil open the file. To make...
- Tue May 15, 2007 10:01 am
- Forum: madKernel
- Topic: Tracking when an application is "done" with a file
- Replies: 5
- Views: 14619
Tracking when an application is "done" with a file
Hi Madshi, I have the following problem: I need to be able to see when a user has closed a file in another application (e.g. Ms Word). (we use this to store the file in a database). I'm trying this using madKernel by checking wether a handle remains valid. However when the file is saved in Ms Word (...
- Tue May 15, 2007 7:36 am
- Forum: madShell
- Topic: RegisterShellEvent on network drive problem
- Replies: 2
- Views: 14725
The problem might occur after the first shell event, or it may not happen for another 100 shell events. Generating network activity after a shell notification seems to reduce the occurence. I agree that the problem most likely exists within the OS. A timer that starts after a notification and resets...
- Fri Apr 27, 2007 12:01 pm
- Forum: madShell
- Topic: RegisterShellEvent on network drive problem
- Replies: 2
- Views: 14725
RegisterShellEvent on network drive problem
Hi, I'm experiencing some problem when using Registershellevent on a network drive to monitor seItemCreated events (other events may be problematic as well, but i have not tested those). Sometimes the notifications stop coming through. Resetting the registershellevent seems to be the only solution w...
- Tue Jan 10, 2006 4:55 pm
- Forum: madCodeHook
- Topic: IPC and synchronisation
- Replies: 2
- Views: 3496