El Mon, May 29, 2006 at 06:28:02PM +0200, Pierre THIERRY ens deleità amb les següents paraules: > Scribit Bas Wijnen dies 29/05/2006 hora 14:28: >>> I wonder if a write notice flag could be interesting. >> Perhaps. I suspect that it will cost performance in situations where >> it isn't used though, which is unacceptable IMO (because it will >> hardly ever be used, I suppose). > > I'm pretty sure it can be built so that when not used, it has no impact. > The notifier could be some sort of proxy, and it then it would just be > absent when not used. In most cases, write notification would only be > needed once (just to know when initial integrity is broken), and the > system could be made to remove that proxy then.
As of what I know, a proxy could only play a role when a page fault occurs, so it wouldn't be able to proxy _all_ accesses (and thus not all writes). A compromise solution, which doesn't penalise the common case (no notice), would be to use pages that are always on COW. Although it would have a bad performance when used (but no zero penalty when not used), it allows to capture all write accesses to the page and let a "proxy" send the notice. Read you, Lluis -- "And it's much the same thing with knowledge, for whenever you learn something new, the whole world becomes that much richer." -- The Princess of Pure Reason, as told by Norton Juster in The Phantom Tollbooth Listening: Dream Theater (Six Degrees Of Inner Turbulence - Disc One) - 3. Misunderstood _______________________________________________ L4-hurd mailing list [email protected] http://lists.gnu.org/mailman/listinfo/l4-hurd
