🚀 go-pugleaf

It is difficult to diagnose the problem you are experiencing without
looking at the actual code. Could you please post the minimal example
that reproduces the problem?

Having said that, one posibility is that one of your threads terminates
without releasing the
ownership of a mutex object and therefore the mutex becomes abandoned.

Another thread (that is waiting) can get ownership of the abandoned mutex,
*This is what could be happening in your application* however the wait
function
that you use to request ownership will return WAIT_ABANDONED.

Whe a mutex is abandoned, the code the mutex is suppose to protect will be
in an undefined state

Rodolfo


"PaulMac" <machiave@i-stat.com> wrote in message
news:3f3a81c8@newsgroups.borland.com...
> In using MUTEX objects, I came across something unexpected.  Given a
Windows
> GUI app and a Service app, if the same named MUTEX is accessed, both
> applications can acquire it when the other has already done so.  However,
if
> both applications are service apps or both are GUI apps, then the proper
> synchronization occurs.
>
> It has been my understanding that MUTEXes, being kernel objects, are
> "global" to a system.  So, why am I seeing this behavior?
>
> Paul Mac
>
>