Closed GoogleCodeExporter closed 9 years ago
Maybe there is another exception happens after the first one, so two crash
reports are sent? Can you debug and tell me if this is true?
Original comment by zexspect...@gmail.com
on 26 Jun 2010 at 12:28
I wrote a test application (exe+dll). See the attachment app.zip. I tested this
application, and there was only one error report.
I suppose there was a second exception in your program, maybe in a class
destructor. This problem may be already fixed in v.1.2.5. Please try it.
Original comment by zexspect...@gmail.com
on 26 Jun 2010 at 2:12
Attachments:
Thank you for prompt response and test dll project. The problem has been sorted
out although the real reason is still not exactly understood. Anyway, it works
as expected now. Here is the recap of this issue and solution.
1 Project
My project is a C++ project that integrates TCL script interpreter.
2 Problem
CrashReport is invoked twice for a crash some times. My initial report found it
happened in DLL but not main application. It turns out it is not caused by DLL.
It really happens in those TCL commands created in my own project. As long as
the crash occurs inside the TCL command, it sends report twice no matter the
TCL command is defined in DLL or main.
3 Solution
After remove CR_INST_SIGSEGV_HANDLER flag from info.dwFlags, the application
sends report once per crash.
Original comment by Ouyang...@gmail.com
on 28 Jun 2010 at 6:04
Anyway, I think I can add some crash counter member variable to CCrashHandler
class. If this counter >0, than the second crash report won't be sent.
An exception of this rule is manually generated crash reports
(crGenerateErrorReport function). I think there shouldn't be crash counter for
such reports.
Original comment by zexspect...@gmail.com
on 29 Jun 2010 at 3:56
Fixed in v.1.2.6
Original comment by zexspect...@gmail.com
on 9 Jul 2010 at 5:05
Thanks. The counter in v1.2.6 does work as expected. I don't need to tweak
anything now with the updated version.
Original comment by Ouyang...@gmail.com
on 14 Jul 2010 at 2:24
I saw duplicate crash emails when Issue 43 was triggered.
Original comment by kevin.m....@gmail.com
on 26 Aug 2010 at 8:29
Original issue reported on code.google.com by
Ouyang...@gmail.com
on 25 Jun 2010 at 6:49