On Wed, 3 Jan 2007 15:04:33 -0800 "Luck, Tony" <[EMAIL PROTECTED]> wrote:
> > > It's simply impossible to implement cmpxchg() in the general case on > > > parisc. This was hashed out in great detail recently ... > > > > That's a bug in the time_interpolator implementation. Fixable, I assume, > > by whacking a spinlock into struct time_interpolator. > > The cmpxchg is already a bit painful if you have a lot of cpus (and a bunch > of users that call gettimeofday() incessantly). A spinlock would most likely > be even worse (wouldn't it?). > It might be slightly worse, although most of the cost will be common between the two implementations. Perhaps some design change is needed in there (a sequence lock?) But cmpxchg just isn't suitable for use outside per-arch code. - To unsubscribe from this list: send the line "unsubscribe linux-arch" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
