Andrea Arcangeli (arcangeli@mbox.queen.it)
Mon, 9 Feb 1998 20:12:25 +0100 (CET)
You said that lp in 2.0.33 works fine. lp in 2.1.85 instead don' t work
fine. I need to know more details.
Before any try apply the patch I will sent you via private email that
fixes some not very dangerous bugs in the latest parport staff.
At first you can try to apply this patch against 2.1.85 in ordert to be
sure that this is a problem related to interrupt handling.
--- linux/include/linux/lp.h 1998/02/09 12:21:23 1.1
+++ linux/include/linux/lp.h 1998/02/09 18:57:42
@@ -159,7 +159,7 @@
*/
#define LP_DELAY 50
-#define LP_POLLING(minor) (lp_table[(minor)].dev->port->irq == PARPORT_IRQ_NONE)
+#define LP_POLLING(minor) 1
#define LP_PREEMPTED(minor) (lp_table[(minor)].dev->port->waithead != NULL)
/*
This patch force the printer to work polled. If now all is right I suspect
that the bug is related to at least ECP interrupt handling.
Now revert the patch.
As second you can try to change your bios parport settings from ECP to EPP
and try to print again.
How long is the stall during the printing? Maybe exactly 1min? If so to
quick fix the problem you only need to decrease LP_TIMEOUT_INTERRUPT in
order to decrease the stall time (60 * HZ == 60 sec, so 1 * HZ will be 1
sec). But this __don' t__ fix the bug, and since you can reproduce the bug
you can help us to really fix it. If decreasing LP_TIMEOUT_INTERRUPT will
decrease also the stall time this means that a interrupt is missed
somewhere.
Andrea[s] 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 Wed 30 Dec 1998 - 10:17:27 EST