Most recent kernel where this bug did not occur: don't know Distribution: gentoo Hardware Environment: AMD64 Athlon X2 4200+ Software Environment: Problem Description: When system is booted with smp, after some random time (hours), usb mouse polling change - mouse is jumping. In system log I found following: May 8 08:03:30 [kernel] irq 10: nobody cared (try booting with the "irqpoll" option) May 8 08:03:30 [kernel] Call Trace: <IRQ> <ffffffff80150334>{__report_bad_irq+48} May 8 08:03:30 [kernel] <ffffffff80150559>{note_interrupt+472} <ffffffff8014fe7c>{__do_IRQ+183} May 8 08:03:30 [kernel] <ffffffff8010d49f>{do_IRQ+57} <ffffffff80108cd9>{default_idle+0} May 8 08:03:30 [kernel] <ffffffff8010af30>{ret_from_intr+0} <EOI> <ffffffff80108d04>{default_idle+43} May 8 08:03:30 [kernel] <ffffffff80108f3a>{cpu_idle+151} <ffffffff805a27d0>{start_kernel+444} May 8 08:03:30 [kernel] <ffffffff805a22a7>{_sinittext+679} May 8 08:03:30 [kernel] handlers: May 8 08:03:30 [kernel] [<ffffffff802ff630>] (acpi_irq+0x0/0x18) In /proc/interrupts I observed that ACPI interrupt (IRQ 10) started to increment with very high speed - thousands interrupts per second. Steps to reproduce: Switch on system with SMP on, and normal work.
Created attachment 8053 [details] Dmesg
Created attachment 8054 [details] lspci
Created attachment 8055 [details] kernel config file
Created attachment 8056 [details] cpuinfo
Created attachment 8057 [details] contents of /proc/interrupts
I found that there are 2 independent things. The one is with ACPI interrupt. Right now it happen 1 time. The other one is with USB interrupts, I will make separate bug report for it.
As it works for hours before the failure, I expect that the SCI on IRQ10 is working properly and that the problem at hand is some sort of run-time error. If you kill acpid and cat /proc/acpi/event and then press the power button a bunch of times, do you see an event for each press? Does the failure occur also if you boot with "maxcpus=1"?
When I killed acpid, and press power button, then after each time the button was released there was generated an event: button/power PWRB 00000080 000000xx where xx is autoincremented event No also I got several: processor CPU0 00000080 00000000 I will try later for maxcpus=1
Now I cannot repeat the Ooops again. Probably the condition was somehow associated with problem on Via EHCI. I propose to make the error invalid now. If I get it again, and have more details, I will reopen it.
I try duplicate this bug with mine (http://bugme.osdl.org/show_bug.cgi?id=6419) , Please try lastest kernel 2.6.18-rc7 with http://lkml.org/lkml/diff/2006/9/7/235/1 patch. and use nvidia opensource code drive
please re-open if this is reproducible with 2.6.21-rc5 or later.