Dave Dillow (dillowd@ornl.gov)
Fri, 16 Oct 1998 15:06:52 -0400
Daniel E. Ungar wrote:
> What we *do* need is true DMA operation for a data stream of at about
> 750 KBytes/sec. Any per-byte response to IRQ, or, worse yet, device
> polling, would be too slow.
I've got an ECP version of lp that may help you. It does DMA to the
printer, but not from it. It could probably be changed to do so. It also
supports poll() so that you can tell when the printer is ready to send
data without getting blocked in a read().
I'm currently getting 350Kbytes/sec to the printer, and I know I can go
faster, if the printer wasn't so damn slow!
Of course, for your case, s/printer/spectrometer.
The only problem is getting the O.K. to release it from my company. If
anyone's interested, let me know and I'll talk to my boss on Monday. No
promises though, and it may take a while. Gotta give the lawyers
*something* to do... :/
Cheers,
Dave Dillow
dillowd@ornl.gov
-- To unsubscribe, send mail to: linux-parport-request@torque.net --
-- with the single word "unsubscribe" in the body of the message. --
This archive was generated by hypermail 2.0b3 on Wed 30 Dec 1998 - 10:18:36 EST