Daniel E. Ungar (dan@pgt.com)
Fri, 16 Oct 1998 10:05:11 -0400
Attn: David Campbell, Roger Schreiter
Thank you for your prompt response to my query.
I must, right off, confess that the FPGA-implemented port on our x-ray
spectrometer is not at all *compliant* with any revision of the IEEE 1284
spec'. It merely implements the barest essentials of single-peripheral
ECP mode, and if the host does not support that, it cannot in fall back to
any
other 1284 mode. It does not know anything about daisy chaining. [Neither
do I; the '1284 spec we have kicking around here is dated February 2, 1993.
:-( ] Our hardware performs nominal ECP-mode negotiation, setup and
termination and forward phase and reverse phase ECP data transfers.
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. To quote my boss, Rick, "We can't live with the
processor
promiscuously fondling every byte that comes along, much less leaning on a
lamppost whistling periodically."
I may end up doing a custom ECP-only driver for Linux. We had to debug the
third-party driver that we got with a SunOS add-in parallel port card,
so we do have some experience in this field. I'll make what I do available,
but I'm unlikely to find the time to make it properly '1284 compliant as far
as fallback to non-ECP modes.
Can anyone point me to a Linux driver that illustrates the use of DMA?
Dan Ungar
dan@pgt.com
-- 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 - 10:18:35 EST