[En-Nut-Discussion] Race condition and stack overflow in tcpsm.c

[Harald Kipp]

Hi Philipp,

On 14.11.2014 10:33, Philipp Burch wrote:
> +#include <sys/mutex.h>

I agree that the problem exists, but I do not agree with the way you
tried to fix this.

MUTEX is application candy, which means, that it is possibly used by
some existing code, but not in the kernel. Your change would add quite
some extra bytes for the rest by creating a reference to nut/os/mutex.c.

http://lists.egnite.de/pipermail/en-nut-discussion/2009-July/025199.html

While Nut/OS is cooperative, mutual exclusion can be done with much less
effort:

int NutTcpInitStateMachine(void)
{
 static int_fast8_t initialized;
 if (!initialized) {
  initialized = 1;
  ...
  Code for initialization
  ...
 }
 return whatever;
}

Regards,

Harald