Unhandled exception - LeaveCriticalSection

[abu]

Hello guys,

I am writing a small simulation program where I make thousands of a function calls to simulate a location based problem. When I use the program to simulate , for instance, at about 60 locations the program works fine. However when I tried to simulate it for 100 locations the program crashed showing the error message below, it does not point out the exact line though. I have tried to simulate each individual points separately it works fine. The problem occurs when increasing the number of simulation locations just above 60. I have tried to carefully release all the dynamically allocated memory to avoid any memory leak. Can someone help me figure out the problem please.


Unhandled exception at 0x7519c41f in myProg.exe: Microsoft C++ exception: std::bad_alloc at memory location 0x006ca3a8..

First-chance exception at 0x7519c41f in myProg.exe: Microsoft C++ exception: std::bad_alloc at memory location 0x006ca3a8..

void __cdecl _unlock (int locknum)
{

        /*

         * leave the critical section.

         */

        LeaveCriticalSection( _locktable[locknum].lock );

}

Thank You!

[Script Coder]

You have not given us enough code, to even begin helping you.

[coder777]

it sounds like an index out of bounds.

It could also an uninitialized varible/pointer.

Or

Deleting the same object more than once.

...

[MikeyBoy]

it sounds like an index out of bounds.

It could also an uninitialized varible/pointer.

Or

Deleting the same object more than once.

The reference page for the std::bad_alloc exception doesn't seem to fit any of those cases. Where are you getting these from?

According to:

http://www.cplusplus.com/reference/new/bad_alloc/

std::badalloc is thrown by new when it fails to allocate memory correctly.

To the OP:

How big are the objects representing these locations that are being instantiated? Is it possible they're big enough that your system is simply running out of memory?

[ne555]

> the program crashed showing the error message below, it does not point out the exact line though

$ g++ -ggdb main.cpp -o program.bin
$ gdb program.bin
> run
(crash)
> backtrace

[JLBorges]

> it sounds like an index out of bounds.
> It could also an uninitialized varible/pointer.
> Deleting the same object more than once.

Yes. That or something similar is possible (very probable).

An allocation failure resulting in a std::bad_alloc being thrown can be caused by corruption of the heap.

[coder777]

The reference page for the std::bad_alloc exception doesn't seem to fit any of those cases. Where are you getting these from?

This all could cause heap corruption. What would happen if locknum is/was out of bounds?


an interesting ling for detecting heap problems:

http://stackoverflow.com/questions/1010106/how-to-debug-heap-corruption-errors

[MikeyBoy]

This all could cause heap corruption. What would happen if locknum is/was out of bounds?

Fair point - guess I wasn't really thinking about it carefully. Sorry!

[abu]

Thank you so much guys for your suggestions.