Re: 2.1.56 lp module problems

Juhani Rautiainen (jrauti@sasu1.carelian.fi)
Tue, 23 Sep 1997 07:28:28 +0300 (EET DST)


On Mon, 22 Sep 1997, Tim Waugh wrote:

> On Mon, 22 Sep 1997, Bernhard Rosenkraenzer wrote:
>
> > Hello,
> > Trying to modprobe lp.o for the 2.1.56 kernel hangs my system - my
> > parallel port is the internal one of a Gigabyte TX mainboard, set to
> > ECP/EPP mode.
>
> Erk! It completely hangs? Did this happen for any other kernel
> revisions? Is it possible to change the mode of the port to something
> other than ECP/EPP?
>

I've got same kind of problem but with ppa-driver. It exhibits itself
both in .55 and .56 kernels. With either EPP or ECP alone it seems to work
but when bios is set to EPP+ECP strange things start to happen.
I've been trying to debug it but those faster than speed of light reboots
and constant filesystem checks with 2 GB partition make it slow
progress. What I've found out so far is this (my machine is at home so
please bear with me). BTW my motherboard is Asus T2P4 (haven't checked
chipset yet). Parport_pc thinks that IRQ 8 is irq it should use. I've
tracked this down to programmable_irq_support where irq gets value 1 from
hardware probe (pc_read_configb()). After that parport_pc
promptly does irq+=7 in switch block which results into irq 8.
BTW it believes this more than value that I give with insmod parameters.
I changed this so that value 1 gives PARPORT_IRQ_NONE. After that
I could set irq with insmod. Now everything seemed to be fine from
parport side but doing insmod ppa reboots machine in less than one second.
I've got K6@225 and what ever scrolls on screen goes by so fast that
it is unreadable and leaves no traces in log. Only thing I reliably saw
was "EPP timeout" from ppa. Ppa works fine with 2.0.30 when it does things
without parport. Because there were no other reports of problems I thought
that I dig bit deeper into this before reporting this.

> Tim.

Juhani

--
Juhani Rautiainen			jrauti@iki.fi