Andrea Arcangeli (andrea@e-mind.com)
Wed, 7 Apr 1999 01:31:56 +0200 (CEST)
On Tue, 6 Apr 1999, Philip Blundell wrote:
>FWIW, I think this time is now adjustable with sysctl. Whatever behaviour we
Well, I'll upgrade to the latest code ASAP.
>choose for this it will always be possible to invent a situation where it
>goes wrong. Just disabling the loop may fix things for you (at least, it may
>get your machine performance back) but it will hurt printing speed for at
>least some people.
Ok, fine.
>I agree that having thought about the numbers a bit, 500us is possibly
>slightly on the high side as the default value. Try winding it down to about
>50 or 100 and see if that improves matters. You could also try adding a check
Ok.
>for resched_needed somewhere in parport_ieee1284_write_compat if there isn't
>one already.
>
>Are you using polling or interrupts there?
irq driven.
0x378: FIFO is 16 bytes
0x378: readIntrThreshold is 1
0x378: writeIntrThreshold is 1
0x378: PWord is 8 bits
0x378: Interrupts are ISA-Pulses
cpp_daisy: aa5500ff(18)
assign_addrs: aa5500ff(18)
parport0: PC-style at 0x378 (0x778), irq 7 [SPP,ECP,ECPEPP,ECPPS2]
lp0: using parport0 (interrupt-driven).
Thanks Philip and Tim!
Andrea Arcangeli
-- 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 Tue 06 Apr 1999 - 21:21:29 EDT