El Mon, Mar 19, 2007 at 09:03:21AM -0400 Robert P. J. Day ha dit:
> however, it's not clear that you can replace those (semaphore-based)
> "MUTEX"es with a real mutex, since mutexes aren't just semaphores with
> an initial count of 1, they have extra restrictions as we've already
> discussed here.
semantically correct code should use MUTEXes like mutexes and not rely
on features specifics to semaphores (release from interrupt, release by
non-owner).
> if one wanted to go through the kernel tree and replace semaphores
> with mutexes, one would have to be *very* careful to do that only
> where the semantics made sense, no?
i agree, and probably s/he should convert MUTEXes that act like binary
semapharores to SEMAPHOREs in order to avoid confusion in the future.
m.
--
Comunicar bichos a <bug-coreutils@xxxxxxx>
(LANG=es_ES uname --help)
.''`.
using free software / Debian GNU/Linux | http://debian.org : :' :
`. `'`
gpg --keyserver keys.indymedia.org --recv-keys B9A88F6F `-
--
To unsubscribe from this list: send an email with
"unsubscribe kernelnewbies" to ecartis@xxxxxxxxxxxx
Please read the FAQ at http://kernelnewbies.org/FAQ
