"irq 16: nobody cared!" -errors after motherboard-switch (ABIT IS7-E2 motherboard)

From: K G
Date: Wed Dec 01 2004 - 12:41:56 EST


Hello!

This is a part from /var/log/dmesg (link to full file
included later):

irq 16: nobody cared!
[<c0108366>] __report_bad_irq+0x2a/0x8b
[<c0108450>] note_interrupt+0x6f/0x9f
[<c010876e>] do_IRQ+0x161/0x192
[<c0106a40>] common_interrupt+0x18/0x20
[<c042ffd3>] cfb_imageblit+0x3b7/0x60c
[<c0106a40>] common_interrupt+0x18/0x20
[<c0426a37>] accel_putcs+0x29a/0x320
[<c0108752>] do_IRQ+0x145/0x192
[<c042007b>] cdrom_mrw_probe_pc+0xa5/0xa8
[<c04280a2>] fbcon_redraw+0x1e8/0x208
[<c042880a>] fbcon_scroll+0x748/0xc40
[<c037cbb1>] scrup+0x13b/0x14f
[<c037e5d7>] lf+0x73/0x75
[<c0380e6c>] vt_console_print+0x127/0x2ed
[<c0123a82>] __call_console_drivers+0x55/0x57
[<c0123b90>] call_console_drivers+0x85/0x11f
[<c0123f67>] release_console_sem+0x63/0xf3
[<c0123e6b>] printk+0x169/0x1bd
[<c075f834>] usb_pwc_init+0x43/0x3a7
[<c07428e5>] do_initcalls+0x28/0xb4
[<c0133ea4>] init_workqueues+0x17/0x2e
[<c0100520>] init+0xf3/0x263
[<c010042d>] init+0x0/0x263
[<c0104271>] kernel_thread_helper+0x5/0xb

handlers:
[<c0435483>] (usb_hcd_irq+0x0/0x67)
[<c0435483>] (usb_hcd_irq+0x0/0x67)
Disabling IRQ #16

I've recently switched from an "ASUS P4T533-C"
motherboard to an "ABIT IS7-E2", and got the errors
mentioned above after boot. Only the motherboard and
the ram was changed (rambus -> INFINEON 400Mhz DDR).

I've tried everything that I could do, bot non of them
helped:

I've set the Sata to "enhanced mode" in BIOS Settings,
which solved the problem for a guy with a similar ABIT
motherboard, it didn't worked for me; Actually I've
tried all Sata settings...

I've tried it with 2.6.7, 2.6.9, even 2.6.10-rc2
kernels, same error everywhere.

Slax LiveCD got the same error, my installed system
got the same error(SourceMage linux). Interestingly
UHU-Linux hasn't got any errors(maybe because the 2.4
kernel it uses).

So I've figured that the IRQ is shared between the two
onboard 1.1 USB hubs and the vga card. My vga is an
NVIDIA GeForce4 Ti4200 btw.

If I disable the onboard usb in the BIOS, I get no
errors. I also don't get any errors when I pass the
"noirqdebug" parameter to the kernel. (Maybe the
system is a bit slower I think.)

If I get the errors I can't start X with the nvidia
driver, and my usb mouse stops working. However if I
use vesa for video I can start X, and it runs just
fine. USB mouse works too, if I plug it elsewhere(not
in the motherboard's onboard usb hubs)

I have support of ACPI support in kernel, tried to
boot with "noacpi" parameter, but nothing changed.
Disabling ACPI support in kernel give the same
results.

Oh, and when I tried to install windows XP for my sis
on the same computer it hangs after it copied the
files to hdd, and rebooted the system. It stays there
with the blue background of the installer and a
cursor. Mouse moves, but nothing happens. I've just
wrote this here to give all info I can. :)

So I only want to know that why can it work flawlessly
with noirqdebug on, and why it can't without.

I've put the board in the computer according to its'
manual, and it is ok I think. Though if I messed up
something, then I REALLY don't want to waste your
time, but if someone would be so kind to explain me in
cc than I would be really thankful :) I don't want to
waste your time or the list's bandwith.

Here are the complete logs/files:

http://izone.sytes.net/logs/dmesg
http://izone.sytes.net/logs/.config
http://izone.sytes.net/logs/xorg.log

Or the compressed version with all of them:
http://izone.sytes.net/logs/logs.tar.bz2

Kálmán Gergely

PS: This is a first post, so please be gentle :)

__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/