RE: [Xen-users] 3.0.0: tg3 & sata_sil crashes

["Ian Pratt" <m+Ian.Pratt@xxxxxxxxxxxx>]

> I get the following in dmesg when booting dom0 under xen-3.0.0.
> If someone could explain why I'd very much appreciate it :-).
>
> 1.)
> Looks bad enough, but the SATA disks are usable, so 
> apparently not critical:
> 
> kobject_register failed for sata_sil (-17)
> 
> Call Trace:
> <ffffffff801ef976>{kobject_register+70}
> <ffffffff80243dab>{bus_add_driver+107}
> <ffffffff801fc063>{pci_register_driver+131}
> <ffffffff80151915>{sys_init_module+325}
> <ffffffff80112406>{system_call+134}
> <ffffffff80112380>{system_call+0}

I saw something similar to this when using a bleeding edge compiler a
while back.

Are you building your own xen/kernel? What compiler? Have you modified
the config?
 
> 2.)
> Not sure why tg3 is mentioned below - when I cat 
> /proc/interrupts, only ohci_hcd seems to be using irq 19.  In 
> fact, tg3 (tigon3) doesn't seem to have any interrupts 
> assigned, but then again the network adapter seems to be 
> working fine anyway.

It certainly wouldn't be working if it didn't have an interrupt
assigned. If its not listed in /proc/interrupts then something is deeply
confused on your system.

Getting spurious interrupts isn't that uncommon on some motherboards.

> irq 19: nobody cared!
> 
> Call Trace:
> <ffffffff80155250>{__report_bad_irq+48}
> <ffffffff80155329>{note_interrupt+89}
> <ffffffff80154b5c>{__do_IRQ+252}
> <ffffffff80115f04>{do_IRQ+52}
> <ffffffff8010db65>{evtchn_do_upcall+197}
> <ffffffff880349b0>{:bridge:br_handle_frame_finish+0}
> <ffffffff80112cf1>{do_hypervisor_callback+17}
> <ffffffff8010da2c>{force_evtchn_callback+12}
> <ffffffff8010da2c>{force_evtchn_callback+12}
> <ffffffff8800cfe1>{:tg3:tg3_interrupt_tagged+417}
> <ffffffff80154a0c>{handle_IRQ_event+76}
> <ffffffff80154b3c>{__do_IRQ+220}
> <ffffffff80115f04>{do_IRQ+52}
> <ffffffff8011a5fe>{monotonic_clock+78}
> <ffffffff8010db65>{evtchn_do_upcall+197}<ffffffff80112cf1>{do_
> hypervisor_callback+17}
> <ffffffff8011035a>{xen_idle+106}
> <ffffffff8011035a>{xen_idle+106}
> <ffffffff801103aa>{cpu_idle+58}
> <ffffffff804c871a>{start_kernel+490}
> <ffffffff804c819a>{_sinittext+410}
> 
> handlers:
> [<ffffffff802a80e0>] (usb_hcd_irq+0x0/0x70) 
> [<ffffffff802a80e0>] (usb_hcd_irq+0x0/0x70)
> 
> Disabling IRQ #19
> 
> 
> Hope that someone can bring me up to date.
> 
> Btw, out of curiosity, I'm wondering what the "+100" etc. 
> numbers above mean.
> Is it a byte offset into my particular compilation of the source code?
> If that's the case, is that actually useful to anyone?

Yes and yes. It can give you an idea of how far into the function the
call is.

Ian

_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users