All of lore.kernel.org
 help / color / mirror / Atom feed
* [parisc-linux] Congratulations, remarks and questions
@ 1999-11-03 12:30 Ulrich Strelow
  1999-11-04 20:55 ` Alex deVries
  1999-11-06 18:24 ` Philipp Rumpf
  0 siblings, 2 replies; 3+ messages in thread
From: Ulrich Strelow @ 1999-11-03 12:30 UTC (permalink / raw)
  To: parisc-linux

Hi all,

first of all congratulations for the success of
getting to the shell prompt. I hope that you have
celebrated this event accordingly. I finally made it
too, I build a kernel for my 715/33 which boots until
the sash-prompt. To document the necessary steps for
others, here is my recipe:

1. get a recent kernel source (I took 19991031)
2. apply the patch from Thomas Bogendoerfer to
arch/parisc/kernel/setup.c (see his mail iodc console
patch). The patch to pdc_cons.c seems to be no longer
necessary.
3. get ramdisk.bin-19991019-workingsash and copy it to
/usr/src/linux as ramdisk.bin.
4. run make config and change the following options:
CONFIG_IODC_CONSOLE=y
CONFIG_NETDEVICES=n 
CONFIG_VT=y
CONFIG_VT_CONSOLE=y
CONFIG_SERIAL=n
5. run make dep, make vmlinux, make Image
6. boot the kernel with the "isl" option and change
the command line from ttyS0 to tty (or change the
console parameter of DEFAULT_COMMANDLINE in
arch/parisc/boot/boot_code/ipl_c.c).  

Now some remarks: I have noticed the development of
the I/O-drivers (GSC/Gecko, DINO, LASI, PCI) in the
past to make them stable, so that they can be included
even if the machine does not have these features. This
is also true for my 715/33 and the 19991031 kernel
tree. However, saying no to CONFIG_GECKO_DINO,
CONFIG_GECKO_LASI and CONFIG_PCI leads to missing
symbols inb and outb during the link stage which are
defined in arch/parisc/kernel/pci.c. And saying no to
CONFIG_GECKO is impossible because gecko.a is defined
in DRIVERS in the central Makefile without a
surrounding ifdef. I don't think that this is a
problem right now, but I would like to mention these
points anyway.

My first try was to build the kernel without
specifying CONFIG_NETDEVICES=n. This leads to the
following problem: the kernel stops booting with these
messages:  

Starting lasi lan at 0xf0826000
lasi_alloc_irq(0xc0127968): No LASI found in system
yet
IRQ not found for lasi lan at 0xf0826000
Warning : device (10, 0xb, 0x0, 0x72, 0x0) NOT claimed
by Lasi i82596(???) LAN 700 or similiar (old)

Maybe this is of interest to someone, too.

Now the questions: the kernel boots to the
sash-prompt, but no keyboard input is possible
(neither the HIL-keyboard, nor the keyboard from the
serial terminal I was using). Is this because of the
missing support for the ASP chip of the 715/33 ? Are
all devices (HIL, SCSI and LAN) connected via that
chip ? And finally: has HP found and published the
docs of that chip so that someone is able to write the
supporting routines ?

Greetings from Berlin
Ulrich

__________________________________________________
Do You Yahoo!?
Bid and sell for free at http://auctions.yahoo.com

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [parisc-linux] Congratulations, remarks and questions
  1999-11-03 12:30 [parisc-linux] Congratulations, remarks and questions Ulrich Strelow
@ 1999-11-04 20:55 ` Alex deVries
  1999-11-06 18:24 ` Philipp Rumpf
  1 sibling, 0 replies; 3+ messages in thread
From: Alex deVries @ 1999-11-04 20:55 UTC (permalink / raw)
  To: Ulrich Strelow; +Cc: parisc-linux


In theory, most of these changes are, I think, already in the kernel tree
in CVS.

On Wed, 3 Nov 1999, Ulrich Strelow wrote:
> My first try was to build the kernel without
> specifying CONFIG_NETDEVICES=n. This leads to the
> following problem: the kernel stops booting with these
> messages:  
> 
> Starting lasi lan at 0xf0826000
> lasi_alloc_irq(0xc0127968): No LASI found in system
> yet
> IRQ not found for lasi lan at 0xf0826000
> Warning : device (10, 0xb, 0x0, 0x72, 0x0) NOT claimed
> by Lasi i82596(???) LAN 700 or similiar (old)

The problem here is that the i82596 expects to be connected to a Lasi, not
the ASP thing that's in a 715.  We have no docs on ASP, so we can't really
do anything about this until later.

> Now the questions: the kernel boots to the
> sash-prompt, but no keyboard input is possible
> (neither the HIL-keyboard, nor the keyboard from the
> serial terminal I was using). Is this because of the
> missing support for the ASP chip of the 715/33 ? 

This is because:
- right now no terminal we have supports a keyboard
- we don't have a HIL keyboard driver because of ASP

> Are
> all devices (HIL, SCSI and LAN) connected via that
> chip ? And finally: has HP found and published the
> docs of that chip so that someone is able to write the
> supporting routines ?

ASP support is something we might be able to get from openbsd or mklinux
source, I haven't checked.  It's on our list of documentation we want from
HP though.

- Alex

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [parisc-linux] Congratulations, remarks and questions
  1999-11-03 12:30 [parisc-linux] Congratulations, remarks and questions Ulrich Strelow
  1999-11-04 20:55 ` Alex deVries
@ 1999-11-06 18:24 ` Philipp Rumpf
  1 sibling, 0 replies; 3+ messages in thread
From: Philipp Rumpf @ 1999-11-06 18:24 UTC (permalink / raw)
  To: Ulrich Strelow; +Cc: parisc-linux

> 4. run make config and change the following options:
> CONFIG_IODC_CONSOLE=y
> CONFIG_NETDEVICES=n 
> CONFIG_VT=y
> CONFIG_VT_CONSOLE=y
> CONFIG_SERIAL=n

you're seeing the prompt with the IODC console ?

> Now the questions: the kernel boots to the
> sash-prompt, but no keyboard input is possible
> (neither the HIL-keyboard, nor the keyboard from the
> serial terminal I was using). Is this because of the
> missing support for the ASP chip of the 715/33 ? Are

Unfortunately, the IODC console is not designed to be full-duplex for
multitasking OSes.  Thus, you need to get the serial / HIL driver going
on your hardware to be able to input things or add input to the IODC
console driver which will cause bad 5 second delays (according to the
documentation) when there is no input.

	Philipp Rumpf

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~1999-11-06 19:49 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
1999-11-03 12:30 [parisc-linux] Congratulations, remarks and questions Ulrich Strelow
1999-11-04 20:55 ` Alex deVries
1999-11-06 18:24 ` Philipp Rumpf

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.