Bug 5486 - irq nobody cared - usb not working >= 2.6.13
Summary: irq nobody cared - usb not working >= 2.6.13
Status: CLOSED CODE_FIX
Alias: None
Product: ACPI
Classification: Unclassified
Component: Config-Interrupts (show other bugs)
Hardware: i386 Linux
: P2 normal
Assignee: Diego Calleja
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2005-10-24 01:45 UTC by Gernot Pansy
Modified: 2006-08-10 11:55 UTC (History)
3 users (show)

See Also:
Kernel Version: 2.6.13
Subsystem:
Regression: ---
Bisected commit-id:


Attachments
cat /proc/interrupts (596 bytes, text/plain)
2005-10-24 01:47 UTC, Gernot Pansy
Details
lcpci -v (6.85 KB, text/plain)
2005-10-24 01:49 UTC, Gernot Pansy
Details
dmesg (14.37 KB, text/plain)
2005-10-24 01:50 UTC, Gernot Pansy
Details
debug patch (769 bytes, patch)
2005-10-24 19:00 UTC, Shaohua
Details | Diff
dmesg with patch 6383 (14.20 KB, text/plain)
2005-10-25 03:36 UTC, Gernot Pansy
Details
dmesg acpi=off (18.37 KB, text/plain)
2005-10-26 08:18 UTC, Gernot Pansy
Details
dmesg pci=noacpi (18.88 KB, text/plain)
2005-10-26 08:19 UTC, Gernot Pansy
Details

Description Gernot Pansy 2005-10-24 01:45:15 UTC
Most recent kernel where this bug did not occur:    
Distribution:    
Hardware Environment: MSI MegaBook S270 (amd64)   
Software Environment: Kernel >= 2.6.13    
Problem Description: since kernel 2.16.13 (also 2.6.14-rc5) i'm unable to use  
usb on my notebook, due the irq nobidy card failure. booting with irqpoll is  
working, but than my notebook is very unstable. all other boot options  
(noapic,pci=routeirq,noacpi) are not working. also there is an other problem  
with die ATI IXP chipset, that the clock ticks double speed. but booting with  
option no_timer_check helps out.  
    
Steps to reproduce:
Comment 1 Gernot Pansy 2005-10-24 01:47:17 UTC
Created attachment 6368 [details]
cat /proc/interrupts
Comment 2 Gernot Pansy 2005-10-24 01:49:00 UTC
Created attachment 6369 [details]
lcpci -v
Comment 3 Gernot Pansy 2005-10-24 01:50:48 UTC
Created attachment 6370 [details]
dmesg
Comment 4 Shaohua 2005-10-24 19:00:05 UTC
Created attachment 6383 [details]
debug patch

>ACPI: PCI Interrupt 0000:02:04.0[A] -> GSI 19 (level, low) -> IRQ 177
>ACPI: PCI Interrupt 0000:02:04.1[B] -> GSI 20 (level, low) -> IRQ 185
I suspect enabling card bus bridge is too early. Could you please try the debug
patch and post the dmesg with appling the patch?
Comment 5 Gernot Pansy 2005-10-25 03:36:15 UTC
Created attachment 6384 [details]
dmesg with patch 6383

i have forgotton to say. if i plug-in my usb mouse or enable my built-in
bluetooth dongle after boot, the irq nobody cared problem occurs not during
boot, but at the time plugging the device in.
Comment 6 Gernot Pansy 2005-10-25 15:12:42 UTC
if i disable ehci_hcd in kernel, i can boot without problems... 
Comment 7 Shaohua 2005-10-25 18:27:35 UTC
pci=noacpi doesn't work for you. BTW, does acpi=off work?
Comment 8 Gernot Pansy 2005-10-26 08:18:21 UTC
Created attachment 6392 [details]
dmesg acpi=off

dmesg with acpi=off -> not working
Comment 9 Gernot Pansy 2005-10-26 08:19:09 UTC
Created attachment 6393 [details]
dmesg pci=noacpi

dmesg pci=noacpi -> not working
Comment 10 Joachim Rudolph 2005-11-19 07:36:22 UTC
exactly same bug here on ACER Travelmate 529TXV
Comment 11 Diego Calleja 2006-07-30 11:07:57 UTC
Does it work with recent kernel versions?
Comment 12 Gernot Pansy 2006-08-10 00:57:21 UTC
sorry, for my late response, it's fixed since 2.6.15 or 2.6.14

Note You need to log in before you can comment on or make changes to this bug.