On Fri, Aug 09, 2002 at 08:31:34PM +0100, David Laight wrote:

 > However those were for very large transfers.
 > What happens for small transfers where all the 'red tape'
 > becomes more significant than the innermost loop.
 > Even 128 or 512 bytes might be significantly slower.

Well... if you take a look at the current copyin/copyout, I'm not
sure you can convince me that the new one would be significantly
*slower*.  The current one is ... really brain-dead.

 > I expect that certain workloads contain a significant number
 > of short transfers.  It is probably possible to test misalgned
 > transfers using mismatched ibs and obs values.

In the near future, we plan on writing some code that will give us a
histogram of copyin/copyout activity.

-- 
        -- Jason R. Thorpe <thorpej@wasabisystems.com>