Messages in this thread | | | Subject | Re: [PATCH] fix race in mark_mounts_for_expiry() | From | Trond Myklebust <> | Date | Wed, 18 May 2005 07:07:09 -0400 |
| |
on den 18.05.2005 Klokka 12:53 (+0200) skreiv Miklos Szeredi: > > Yes you can. cmpxchg() is atomic. Several archs implement atomic_inc() and co > > with cmpxchg() or similar. > > > > Something like: > > > > static inline struct namespace *grab_namespace(struct namespace *n) > > { > > int old = atomic_read(&n->count); > > > > while (old > 0) { > > /* attempt to increment the counter */ > > old = cmpxchg(&n->count, old, old + 1); > > } > > > > return old > 0 ? n : NULL; > > } > > > > Ahh OK :) > > There's still the problem of cmpxchg meddling in the internals of an > atomic_t. Is that OK? Will that work on all archs?
Some archs already have an atomic_dec_if_positive() (see for instance the PPC). It won't take much work to convert that to an atomic_inc_if_positive().
For those arches that don't have that sort of thing, then writing a generic atomic_inc_if_positive() using cmpxchg() will often be possible, but there are exceptions (for instance the original 386 does not have a cmpxchg, so there you will have to use something else).
Cheers, Trond
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |