[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: nptl 0.30



"Alexander Terekhov" <terekhov web de> schrieb am 18.03.03 21:33:10:
> 
> Ulrich Drepper <drepper redhat com> schrieb am 18.03.03 19:53:59:
> > 
> > -----BEGIN PGP SIGNED MESSAGE-----
> > Hash: SHA1
> > 
> > Alexander Terekhov wrote:
> > >>Just invert those lines to :
> > >>
> > >>lll_mutex_unlock (cbuffer->cond->__data.__lock); /* much better to call
> > >>futex_wake() with the mutex unlocked */
> > >>lll_futex_wake (futex, INT_MAX);
> > > 
> > > 
> > > I think that it's sufficient to wake just one single thread 
> > > here (in cleanup() for async.cancel region "wrapping" futex 
> > > wait(); I guess).
> > 
> > No, it's not sufficient.  There might be a thread which cannot be woken
> > because it arrived after the cond_signal o broadcast which might be woken.
> 
> I don't understand this. Would you please elaborate?

I think now I understand.

> 
> > 
> > Again, this is cancellation we are talking about.  It is not a regular
> > programming tool, it's far too costly.  It's used in exception
> > situations only.
> 
> Basically, I mean: < illustration >
> 
> http://www.terekhov.de/DESIGN-futex-CV.txt-with-async.cancelable-wait.txt

Ulrich, I've applied the patch now. Thanks. ;-)

regards,
alexander.

______________________________________________________________________________
Mit der Gruppen-SMS von WEB.DE FreeMail erreichen Sie mit einem Klick 
alle Freunde gleichzeitig! http://freemail.web.de/features/?mc=021180





[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]