Failure w/ latest ACPI patch...

Failure w/ latest ACPI patch...

[Aaron Gaudio]

[-- Attachment #1: Type: text/plain, Size: 2074 bytes --]

After applying the 20020726 patch to the 2.4.18 kernel, my Sony
Vaio R505E locked up during boot. It locks up somewhere after
loading the usb controller and before initializing the HID driver
for the usb mouse which was attached.

The version of the kernel was a clean 2.4.18 with just the ACPI
patch and a patch for agpgart to support high-resolution modes
in X on the i830 chipset.

Other than this, I am running a re-configured, re-built kernel
based off of Red Hat's latest 2.4.18-5 kernel sources (which is
basically the 2.4.19 release candidate + ac fixes + tons of other
patches). While this kernel boots up fine, it fails to assign an
IRQ to several devices which should have IRQs of 9 (at least, they
do in Windows- where all the PCI devices are apparently assigned
IRQ of 9, for better or worse).

I notice that in the acpi-patched version, there is some (possibly
BIOS?) message before the actual dmesg output gets printed, and
for at least one of those devices for which an IRQ is not found, it
says it will use IRQ 255. Incidentally, there are 3 USB controllers
on the laptop, and in the RH-based kernel, only 2 get loaded properly,
while the third does not have an IRQ assigned to it (perhaps this is
related to the lock-up during HID initialization in the acpi-patched
kernel).

Unfortunately, because the boot sequence does not complete, I don't
have the output of dmesg available for the failed boot (I can always
attach one for the RH-based kernel, if anyone would like to see that).

BTW, I'm trying to find userland utilities which will allow me to put
the laptop into the varying suspend states (including hibernate-- I
have preserved the Windows partition where the hibernate file is
located). I can't seem to find any one that is ACPI-compatible. Any
pointers?

-- 

Aaron Gaudio                           prothontar @ tarnation.dyndns.org
                   http://tarnation.dyndns.org/~aaron
                            ----------------
  "From fullness, aspect. From aspect, being. From being, emptiness."

[-- Attachment #2: Type: application/pgp-signature, Size: 240 bytes --]

Re: Failure w/ latest ACPI patch...

[Dominik Brodowski]

Hi,

Intersting problem you're reporting here:

On Mon, Jul 29, 2002 at 09:03:41PM -0400, Aaron Gaudio wrote:
> I notice that in the acpi-patched version, there is some (possibly
> BIOS?) message before the actual dmesg output gets printed
Even before "uncompressing kernel...." ??? 

, and
> for at least one of those devices for which an IRQ is not found, it
> says it will use IRQ 255.
That's BAD. IRQ 255 does not exist on any known laptop, I dare to say. 

> Unfortunately, because the boot sequence does not complete, I don't
> have the output of dmesg available for the failed boot (I can always
> attach one for the RH-based kernel, if anyone would like to see that).
Could you pass pci=noacpi to the ACPI kernel, and try if it boots, please?

Dominik


-------------------------------------------------------
This sf.net email is sponsored by: Dice - The leading online job board
for high-tech professionals. Search and apply for tech jobs today!
http://seeker.dice.com/seeker.epl?rel_code=31

Re: Failure w/ latest ACPI patch...

[Aaron Gaudio]

[-- Attachment #1: Type: text/plain, Size: 1899 bytes --]

And lo, upon Tue, Jul 30, 2002 at 04:54:33PM +0200, Dominik Brodowski spaketh thusly:
> Hi,
> 
> Intersting problem you're reporting here:
> 
> On Mon, Jul 29, 2002 at 09:03:41PM -0400, Aaron Gaudio wrote:
> > I notice that in the acpi-patched version, there is some (possibly
> > BIOS?) message before the actual dmesg output gets printed
> Even before "uncompressing kernel...." ??? 

Sorry, I should have been clearer. I meant that it froze up before
I could save the dmesg output to a file to attach. I could see the
actual dmesg output on screen.

> 
> , and
> > for at least one of those devices for which an IRQ is not found, it
> > says it will use IRQ 255.
> That's BAD. IRQ 255 does not exist on any known laptop, I dare to say. 
> 
> > Unfortunately, because the boot sequence does not complete, I don't
> > have the output of dmesg available for the failed boot (I can always
> > attach one for the RH-based kernel, if anyone would like to see that).
> Could you pass pci=noacpi to the ACPI kernel, and try if it boots, please?
>

Actually, thanks to advice from Ted Serreyn, it turns out the problem
was that the ACPI was in fact working :) The USB controller that previously
was not getting loaded with an IRQ was now (IRQ 9, like everything else),
and that USB controller happened to be the one for the memory stick on
the laptop. The memory stick apparently is not supported yet in the
stock kernel. Ted turned me on to a patch to enable this device, and,
upon applying it, I was able to complete the boot successfully. Now I'm
fighting to get my DVD drive to be recognized over FireWire :)
 

-- 

Aaron Gaudio                           prothontar @ tarnation.dyndns.org
                   http://tarnation.dyndns.org/~aaron
                            ----------------
  "From fullness, aspect. From aspect, being. From being, emptiness."

[-- Attachment #2: Type: application/pgp-signature, Size: 240 bytes --]

Re: Failure w/ latest ACPI patch...

[Sérgio Monteiro Basto]

> Could you pass pci=noacpi to the ACPI kernel, and try if it boots, please?

/boot/grub/grub.conf
title Red Hat Linux (2.4.18)
root (hd0,1)
kernel /vmlinuz-2.4.18 ro root=/dev/hda5 pci=noacpi

result:
PCI: Unknown option `noacpi'

this question can be a little out of the subject, but when you say pass pci=noacpi the
ACPI kernel, what is suppose to do?
adding to parameter to kernel ? 


Sérgiomb



-------------------------------------------------------
This sf.net email is sponsored by: Dice - The leading online job board
for high-tech professionals. Search and apply for tech jobs today!
http://seeker.dice.com/seeker.epl?rel_code1