IDE problems with 2.3.49?

IDE problems with 2.3.49?

[Michael Thompson]

I've been trying to get version 2.3.49 booted on my B&W G3 (USB/IDE).

It can probe the IDE drives, and they spin up, but it isn't getting
interrupts from IDE0:

	hda: lost interrupt
	hda: lost interrupt
		.
		.
	hdb: lost interrupt
	hdb: lost interrupt

I have been able to boot with a ram disk, and the keyboard works,
so interrupts, in general, are working.

At this point, the problem seems to be isolated to IDE.
Has anyone had success with this?

Btw, I got the kernel from fsmlabs' web-site. It's rtlinux v3 pre 4.

Thanks,

Michael Thompson
mickey@berkeley.innomedia.com

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

IDE problems with 2.3.49?- same with 2.2.15pre3

[Kevin B. Hendricks]

Hi,

I just finished installing linuxppc 2000 on my brand new G4 450MHz box and
every time I try to boot into my linux side with Bootx 1.2fc1 I get the
exact same error messages:

hdc: lost interrupt
hdd: lost interrupt

and on top of that

usbcore: USB device not responding, giving up (error=-101).

The funny thing is that the installation from the ramdisk went fine and the
keyboard and things worked with no problem.

I have tried the hda=noautotune .. hdb= ... hdc= to no avail.

Is there something different about recent G4 450s that I need to worry about?

Thanks,

Kevin


At 08:32 -0800 3/9/00, Michael Thompson wrote:
>I've been trying to get version 2.3.49 booted on my B&W G3 (USB/IDE).
>
>It can probe the IDE drives, and they spin up, but it isn't getting
>interrupts from IDE0:
>
>	hda: lost interrupt
>	hda: lost interrupt
>		.
>		.
>	hdb: lost interrupt
>	hdb: lost interrupt
>
>I have been able to boot with a ram disk, and the keyboard works,
>so interrupts, in general, are working.
>
>At this point, the problem seems to be isolated to IDE.
>Has anyone had success with this?
>
>Btw, I got the kernel from fsmlabs' web-site. It's rtlinux v3 pre 4.
>
>Thanks,
>
>Michael Thompson
>mickey@berkeley.innomedia.com
>


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: IDE problems with 2.3.49?- same with 2.2.15pre3

[David A. Gatwood]

On Thu, 9 Mar 2000, Kevin B. Hendricks wrote:

> I just finished installing linuxppc 2000 on my brand new G4 450MHz box and
> every time I try to boot into my linux side with Bootx 1.2fc1 I get the
> exact same error messages:
>
> hdc: lost interrupt
> hdd: lost interrupt
>
> and on top of that
>
> usbcore: USB device not responding, giving up (error=-101).
>
> The funny thing is that the installation from the ramdisk went fine and the
> keyboard and things worked with no problem.
>
> I have tried the hda=noautotune .. hdb= ... hdc= to no avail.
>
> Is there something different about recent G4 450s that I need to worry about?

Does it work?  If so, then you're fine.  I've been running Benjamin
Herrenschmidt's iBook/* kernel on a G4/450 with the MkLinux distro, and
I've noticed these same messages at startup myself.  The box has been up
and running under heavy disk and keyboard load for 16 days now w/o a
crash.


David


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: IDE problems with 2.3.49?- same with 2.2.15pre3

[Michael Thompson]

>
> On Thu, 9 Mar 2000, Kevin B. Hendricks wrote:
>
> > I just finished installing linuxppc 2000 on my brand new G4 450MHz box and
> > every time I try to boot into my linux side with Bootx 1.2fc1 I get the
> > exact same error messages:
> >
> > hdc: lost interrupt
> > hdd: lost interrupt
> >
> > and on top of that
> >
> > usbcore: USB device not responding, giving up (error=-101).
> >
> > The funny thing is that the installation from the ramdisk went fine and the
> > keyboard and things worked with no problem.
> >
> > I have tried the hda=noautotune .. hdb= ... hdc= to no avail.
> >
> > Is there something different about recent G4 450s that I need to worry about?
>
> Does it work?  If so, then you're fine.  I've been running Benjamin
> Herrenschmidt's iBook/* kernel on a G4/450 with the MkLinux distro, and
> I've noticed these same messages at startup myself.  The box has been up
> and running under heavy disk and keyboard load for 16 days now w/o a
> crash.

For me, it doesn't work. I did just discover, however, that I can mount
the cdrom (IDE2, irq=13).

-Michael

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: IDE problems with 2.3.49?- same with 2.2.15pre3

[Benjamin Herrenschmidt]

On Thu, Mar 9, 2000, Kevin B. Hendricks <khendricks@ivey.uwo.ca> wrote:

>I just finished installing linuxppc 2000 on my brand new G4 450MHz box and
>every time I try to boot into my linux side with Bootx 1.2fc1 I get the
>exact same error messages:
>
>hdc: lost interrupt
>hdd: lost interrupt
>
>and on top of that
>
>usbcore: USB device not responding, giving up (error=-101).
>
>The funny thing is that the installation from the ramdisk went fine and the
>keyboard and things worked with no problem.
>
>I have tried the hda=noautotune .. hdb= ... hdc= to no avail.
>
>Is there something different about recent G4 450s that I need to worry about?

Use yaboot, not BootX and tell me if you still have the problem.


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: IDE problems with 2.3.49?- same with 2.2.15pre3

[Kevin B. Hendricks]

Hi Ben,

>>Is there something different about recent G4 450s that I need to worry about?
>
>Use yaboot, not BootX and tell me if you still have the problem.

I grabbed the yellowdog yaboot_0.4 (I can not get your yaboot_0.5 to
uncompress properly) and started up with with the following yaboot.conf

image=zip:vmlinux.2.2.15pre3
label=linuxppc
append="root=/dev/hda12 video=aty128fb:vmode:18,cmode:32"

And used the command-option-o-f settings to get the of prompt and then typed:

boot zip:yaboot_0.4

and at the yaboot prompt typed linuxppc

Unfortunately I could not see the screen since it got split into 3 separate
vertical panes (like the aty128fb.c problem that anthony fixed by removing
the clock setting calls if you remember).

I will try again with just the OF frame buffer and let you know what if any
error messages about lost interrupts come up.

Thanks,

Kevin


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: IDE problems with 2.3.49?- same with 2.2.15pre3

[Kevin B. Hendricks]

Hi Ben,

>>>Is there something different about recent G4 450s that I need to worry
>>>about?
>>
>>Use yaboot, not BootX and tell me if you still have the problem.


Nope, all of the booting problems went away with yaboot!

Unfortunately, booting with of is not my choice:

1. Where can I get the yaboot-0.5 version as a MacOS app that can be used
from zip

2.  I do not want to again repartition.  Is booting from a zip using the
o-f keys my only choice now?  Instead of using a zip, could I burn a cd-rom
with yaboot on it that I put in when I want to of boot and then restart and
hold down the c key?  Would it be better to put your of-booting script on
the cdrom?

Also, is there any chance BootX will be fixed to work with these machines
or am I SOL (I spend as much time using MacOS as LinuxPPC and liked the
control of simple dual booting without messing with of).

Thanks!

Kevin


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: IDE problems with 2.3.49?- same with 2.2.15pre3

[Benjamin Herrenschmidt]

On Thu, Mar 9, 2000, Kevin B. Hendricks <khendricks@ivey.uwo.ca> wrote:

>1. Where can I get the yaboot-0.5 version as a MacOS app that can be used
>from zip
>
>2.  I do not want to again repartition.  Is booting from a zip using the
>o-f keys my only choice now?  Instead of using a zip, could I burn a cd-rom
>with yaboot on it that I put in when I want to of boot and then restart and
>hold down the c key?  Would it be better to put your of-booting script on
>the cdrom?

If you used BootX, then I beleive you already have an HFS or HFS+
partition. Just put yaboot there along with a yaboot.conf file, and set
the OF "boot-device" path to point to it. If you want to dual boot MacOS
and Linux, copy also the dual boot OF script I posted some time ago (it's
also in the ybin package from Ethan Benson) and set OF to boot this
script (and don't forget to edit the OF path inside the script). Then,
use the space bar at boot to choose which OS you want.

>Also, is there any chance BootX will be fixed to work with these machines
>or am I SOL (I spend as much time using MacOS as LinuxPPC and liked the
>control of simple dual booting without messing with of).

There are more chances that we finally make a tool to configure OF and
yaboot easily than for BootX to be fixed. The BootX problem with the
iMacDV is not simple since MacOS actualy shuts down some hardware devices
completely (apparently, the clock of the ATI controller seems to be
switched off) before I take control. With other machines, we would have
to find a way to stop all devices that may do bus mastering and reset
them to a known state.


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: IDE problems with 2.3.49?- same with 2.2.15pre3

[Benjamin Herrenschmidt]

On Thu, Mar 9, 2000, Kevin B. Hendricks <khendricks@ivey.uwo.ca> wrote:

>I grabbed the yellowdog yaboot_0.4 (I can not get your yaboot_0.5 to
>uncompress properly) and started up with with the following yaboot.conf

there's a problem with Apache configuration on ppclinux.apple.com (which
is, BTW, the same problem i used to have when downloading Xpmac from your
web server). Apparently, the server tries to uncompress the gz file on
the fly. It appears however that most web browser will only download the
compressed size, and so the downloaded file is incomplete.
I used to work around this (when downloading Xpmac from your site) by
using Netscape under MacOS (apparently, it did download the entire
uncompressed file), and removing manally the ".gz" once downloaded.

If you know the solution (the server setting that need to be changed),
tell me and I'll pass the info along to Kevyn (maintainer of
ppclinux.apple.com).


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: IDE problems with 2.3.49?

[Wulf Hofbauer]

I've been experiencing IDE interrupt problems with anything more
recent than 2.1.24. On my 7220, the IDE controller doesn't seem to
be properly initialized and raises an interrupt when none is expected.
The kernel then gets stuck in an infinite interrupt loop. This occured
however only when booting from OF. When using a MacOS-based loader,
MacOS would have initialized the IDE controller before the kernel was
running.

The workaround for me has been to reset the IDE controller. The kernel
bootflags are something like ide0=reset, ide1=reset. After that, the
unexpected interrupt is discarded, though it still generates a warning.

Maybe the current problems are related to this.

- Wulf

--
 ________________________________________________________
! Dipl. Phys. Wulf Hofbauer  (wh@echo.chem.tu-berlin.de) !
! Max-Volmer-Institut     Technische Universitaet Berlin !
! Strasse des 17. Juni 135     10623 Berlin      Germany !
!________________________________________________________!

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/