Re: [PARPORT] parport-scsi release 0.02

Tim Waugh (
Mon, 19 Oct 1998 19:29:15 +0100 (BST)

On Mon, 19 Oct 1998, Philip Blundell wrote:

> I'd be interested in measurements of performance for your direct I/O method
> against going through parport. If there really is a problem then I'll
> certainly have a think about ways to get the overhead down.

Another thing is this: David, you have lots of inline asm in ppa for doing
port-accesses. Ages ago, you claimed it was because the parport functions
were too slow. I kept meaning to compare the .s files, but I still
haven't gotten around to it.

Maybe egcs does a better job? It really shouldn't be necessary to code
the asm yourself.


-- To unsubscribe, send mail to: --
-- with the single word "unsubscribe" in the body of the message. --

This archive was generated by hypermail 2.0b3 on Wed 30 Dec 1998 - 10:18:38 EST