On Sun, Aug 19, 2001 at 05:53:11AM +0200, Andrea Arcangeli wrote:
> just need to do a second lookup after upgrading the lock because we
> don't have a primitive to upgrade the lock from "read" to "write"
I attempted to write such a primitive but it's impossible without a fail
path (think two readers trying to upgrade the lock at the same time, it
either deadlocks or one of the two will have to fail the atomic
upgrade and accept that something has changed under it), so it could
only improve performance saving a lookup sometime in case we are the
only readers out there, but we would still need to implement the ugly
slow path case (and avoiding the ugliness of such code was the only
reason I attempted to write such a primitive so...)
Andrea
-
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/
This archive was generated by hypermail 2b29 : Thu Aug 23 2001 - 21:00:30 EST