On 2025-06-20, Mark Kettenis wrote:
> > Date: Fri, 20 Jun 2025 13:25:15 +0000
> > From: Job Snijders <job@openbsd.org>
> > 
> > Dear Guehnther / Tedu,
> > 
> > As I understand it, a mutex must be used to protect the condition
> > variable itself. So I think the below might help clarify the
> > relationship between conditions and mutexes. OK?
> 
> Yes, see pthread_cond_wait(3).  But I don't think adding this here
> helps in any way.

There should be a link somewhere. Someone can probably figure it
out, but man pthread_mutex_t doesn't work. I think it would be nice to
have at least one link that actually goes to a mutex man page.

Putting the text in pthread_cond_init is as good as anywhere, 
since you're likely going to init the mutex at the same time.

> 
> > Index: ./man/pthread_cond_init.3
> > ===================================================================
> > RCS file: /cvs/src/lib/libpthread/man/pthread_cond_init.3,v
> > diff -u -p -r1.13 pthread_cond_init.3
> > --- ./man/pthread_cond_init.3	7 Jun 2025 00:16:52 -0000	1.13
> > +++ ./man/pthread_cond_init.3	20 Jun 2025 13:20:56 -0000
> > @@ -50,6 +50,8 @@ If
> >  is
> >  .Dv NULL ,
> >  the default attributes are used.
> > +.Pp
> > +A condition variable must always be used in conjunction with a mutex.
> >  .Sh RETURN VALUES
> >  If successful, the
> >  .Fn pthread_cond_init
> > @@ -76,7 +78,8 @@ variable.
> >  .Xr pthread_cond_destroy 3 ,
> >  .Xr pthread_cond_signal 3 ,
> >  .Xr pthread_cond_timedwait 3 ,
> > -.Xr pthread_cond_wait 3
> > +.Xr pthread_cond_wait 3 ,
> > +.Xr pthread_mutex_init 3
> >  .Sh STANDARDS
> >  .Fn pthread_cond_init
> >  conforms to
> > 
> > 
> 
>