Re: recursive locks (in general)

[Christian Schoenebeck]

On Freitag, 21. August 2020 15:08:09 CEST Paolo Bonzini wrote:
> On 21/08/20 13:12, Christian Schoenebeck wrote:
> > There is a golden rule with recursive locks: You always have to preserve a
> > clear hierarchy. Say you have the following recursive mutexes:
> > 
> > RecursiveMutex mutex0;
> > RecursiveMutex mutex1;
> > RecursiveMutex mutex2;
> > ...
> > RecursiveMutex mutexN;
> > 
> > where the suffix shall identify the hierarchy, i.e. h(mutex0) = 0,
> > h(mutex1) = 1, ... h(mutexN) = N. Then the golden rule is that in any call
> > stack the nested locks must always preserve the same transitive hierarchy,
> 
> > e.g.:
> That's also what you do with regular locks.

You can't do that with non-recursive mutexes if you have cyclic dependencies.

> But the difference is that with regular locks you can always do
> 
> void bar(std::unique_lock<std::mutex> &mutex3_guard)
> {
>       ...
>       mutex3_guard.unlock();
>       synchronized(mutex2) {
>       }
>       mutex3_guard.lock();
>       ...
> }

Right, if you are able to clearly judge that this unlock is really safe for 
all layers involved.

Okay never mind, I see that we'll keep split on this recursive lock issue 
anyway. Sorry for the noise Paolo! :)

Best regards,
Christian Schoenebeck