2003-08-11 17:23:43

by Justin T. Gibbs

[permalink] [raw]
Subject: Re: Linux [2.6.0-test3/mm1] aic7xxx problems.

> On a side note, the same aic7xxx drivers version 6.2.8 and 6.2.36 works
> with 2.4.21 and 2.4.22-rc1/rc2 series of kernel with above hardware.

I don't think that any of the changes between 6.2.35 and 6.2.36 will
make a difference for you, but you could try upgrading. The source
files are here:

http://people.FreeBSD.org/~gibbs/linux/SRC/

The console output you've provided makes me think that interrupts are
not working correctly in your system.

--
Justin


2003-08-11 17:42:56

by Paul Blazejowski

[permalink] [raw]
Subject: Re: Linux [2.6.0-test3/mm1] aic7xxx problems.

On Mon, 2003-08-11 at 13:22, Justin T. Gibbs wrote:

> > On a side note, the same aic7xxx drivers version 6.2.8 and 6.2.36 works
> > with 2.4.21 and 2.4.22-rc1/rc2 series of kernel with above hardware.
>
> I don't think that any of the changes between 6.2.35 and 6.2.36 will
> make a difference for you, but you could try upgrading. The source
> files are here:
>
> http://people.FreeBSD.org/~gibbs/linux/SRC/
>

Tried those already on both 2.4.21/22 and 2.6.0-test1,2,3.
While the 2.4 driver compiles and works fine the 2.5 sources does not
play nice with 2.6.0-test* kernel.I had to change the includes to blkdev
in the source and then there was few compile warnings.I can try to
provide the exact ones, unless there's a driver that applies to
2.6-test* cleanly.Once i had it compiled the kernel would just panic.

> The console output you've provided makes me think that interrupts are
> not working correctly in your system.
>

Is this due to my mainboard (nforce2),cpu,ACPI,devfs,sysfs...or all
these together?

I tried swapping the SCSI controller card in various PCI slots,used new
and original working cables,changed the BIOS IRQ's...etc

I start running of ideas here, is there other options to fully debug the
problem? What else can i try? I would appreciate any pointers.

I would like to get things working with 2.6.0 kernels...right now only
aic7xxx_old is somehow working but that's not the solution to the
problem.

> --
> Justin
>
>

Thank you for your time.

Paul



Attachments:
signature.asc (189.00 B)
This is a digitally signed message part

2003-08-11 19:23:50

by Rahul Karnik

[permalink] [raw]
Subject: Re: Linux [2.6.0-test3/mm1] aic7xxx problems.

Paul Blazejowski wrote:

> Is this due to my mainboard (nforce2),cpu,ACPI,devfs,sysfs...or all
> these together?

NForce2 + ACPI IRQ routing is a work in progress. Try turning off ACPI
for the time being and see if it helps.

Thanks,
Rahul
--
Rahul Karnik
[email protected]

2003-08-12 01:11:13

by Paul Blazejowski

[permalink] [raw]
Subject: Re: Linux [2.6.0-test3/mm1] aic7xxx problems.

> Paul Blazejowski wrote:
>
>> Is this due to my mainboard (nforce2),cpu,ACPI,devfs,sysfs...or all
>> these together?
>
> NForce2 + ACPI IRQ routing is a work in progress. Try turning off ACPI
> for the time being and see if it helps.
>
> Thanks,
> Rahul
> --
> Rahul Karnik
> [email protected]
>
>
>

Thanks Rahul, i'll try that.Are acpi=off and pci=noacpi the same options?
what about APIC, should i also use noapic option upon booting the kernel
(APIC is compiled in)?

Paul

2003-08-12 18:56:25

by Paul Blazejowski

[permalink] [raw]
Subject: Re: Linux [2.6.0-test3/mm1] aic7xxx problems.

On Mon, 2003-08-11 at 13:22, Justin T. Gibbs wrote:

>
> I don't think that any of the changes between 6.2.35 and 6.2.36 will
> make a difference for you, but you could try upgrading. The source
> files are here:
>
> http://people.FreeBSD.org/~gibbs/linux/SRC/
>
> The console output you've provided makes me think that interrupts are
> not working correctly in your system.
>
> --
> Justin
>
>

Justin,

I have a nice oops from the
http://people.freebsd.org/~gibbs/linux/SRC/aic79xx-linux-2.5-20030603-tar.gz driver on 2.6.0-test3.

I took few snaps which show the scsi controller scan process and the
oops from kernel, they are at:
http://www.blazebox.homeip.net:81/diffie/images/linux-2.6.0-test2/2.6.36/

Paul


Attachments:
signature.asc (189.00 B)
This is a digitally signed message part