> 
> On Sun, 11 Feb 1996 07:09:20 +0000 () 
>  "John S. Dyson" <toor@dyson.iquest.net> wrote:
> 
>  > It would be likely on a 20MB machine or even a 32MB machine that you would
>  > not be able to measure a striking perf difference between a fully optimal
>  > scheme and a simple, straightforward bouncing scheme.  I do suggest a
>  > SCSI device independent scheme, even though I know that many are interested
>  > primarily in the aha1542 type adapters, some have bt545s's also.
> 
> Of course, one would hope that the implementation isn't limited to SCSI 
> adapters...
> 
It is (currently) limited to block devices in general, with very little
modification of the driver needed (about 3lines of code.)  There is
also an additional shim for dmaable malloced memory.

Note that I debugged the original code on an IDE based system, and
special checks in the driver to panic on invalid memory (>16MByte).

John