"Kendrick M. Smith" wrote:
> I have the patch from your original post and will give it a try with
> 2.5.  In that post, you also mentioned having some sort of testsuite
> which would place the semaphore under heavy contention, while also
> testing basic semantics of the semaphore.  If you send this along, I
> will give it a try as well...

The testsuite patch is attached. Apply it after the rwsem trylock patch.
What it does is start ten kernel threads that randomly call down_read,
down_write, down_read_trylock, and down_write_trylock. Reboot with the
test kernel and let it run for awhile. It should give your hard drive a
good workout as the debug output is written to the log. After a few
thousand locking operations, reboot with an ordinary kernel and study
the log.

If you have questions about the debug output, let me know.

-- 
Brian Watson                | "Now I don't know, but I been told it's
Linux Kernel Developer      |  hard to run with the weight of gold,
Open SSI Clustering Project |  Other hand I heard it said, it's
Compaq Computer Corp        |  just as hard with the weight of lead."
Los Angeles, CA             |     -Robert Hunter, 1970

mailto:Brian.J.Watson@compaq.com
http://opensource.compaq.com/