Subj : Re: Optimization to Jeffrey Richter's COptex (Windows)?
To   : comp.programming.threads
From : Alexander Terekhov
Date : Wed Jan 26 2005 07:06 pm


Joseph Seigh wrote:

[... ordinal reads ...]

> No.  If a thread releases the lock, it sets the owning thread id to zero.
> it can only see zero or the thread id  of another thread.  

Absent application synchronization with respect to termination 
and creation of threads, it may (in theory; formally speaking
based on the lack of constraint imposed on the threading impl) 
see id of another already terminated thread that happens to 
have the same value. 

t0: thread 666 unlocks the mutex and terminates;

t1: thread 000 (it can still see 666) creates another thread 
    and it gets reclaimed id 666. It "inherits" the same view 
    of memory as thread 000 (sees 666);

t3: thread 001 locks the mutex.

t4: new thread 666 attempts to lock the mutex and thinks that 
    it already owns it...

not good, oder?

regards,
alexander.

.