Re: [ANNOUNCE]: LuaThreads 1.0-work, pre-release

[Diego Fernandes Nehab <diego@...>]

Hi Peter,

> [[...]]
> This can make design of larger scale systems difficult, at best, and
> the problem is only worse when dealing with other people's libraries.
> Basically short of searching the source code (if you have it), you
> really must assume that it is not safe to call a function that affects
> an object protected by a mutex you currently hold; which goes against
> the idea of the mutex in the first place.

I agree  that recursive mutexes are  nice for these examples,  but think
about the  effect on condition variables.  If you use a  recursive mutex
and you  have two locks  on the mutex  associated to the  predicate, the
implicit  unlock performed  by the  wait function  will not  release the
mutex.  If this  happens, no  other thread  will be  able to  change the
predicate, and we get a deadlock. I am not sure which bug will be harder
to find.

I will see how hard it is  to provide recursive mutexes to LuaThreads in
Linux, Solaris, AIX and IRIX. recursive  mutexes are not portable as far
as I know. If I manage to get  them all working, I might add a parameter
to newmutex(), to specify the kind of  mutex to be created. I don't like
the extra complexity,  though. Besides, I am  not sure if it  is easy to
create a non-recursive (and reasonably fast) mutex in Win32.

Any sugestions?

Regards,
Diego.