Re: Bug in pci_restore_msi_state

From: Thomas Meyer
Date: Sat Mar 17 2007 - 14:30:43 EST


Pekka Enberg schrieb:
> On 3/17/07, Thomas Meyer <thomas@xxxxxxxx> wrote:
>> Hello everybody.
>>
>> I get this bug after suspending to disk twice:
>>
>> http://m3y3r.de/bilder/Bug-pci_restore_msi_state.png
>>
>> This happens with current git head
>> cd05a1f818073a623455a58e756c5b419fc98db9.
>
> If you know a kernel that works, please consider doing git bisect:
>
> http://www.kernel.org/pub/software/scm/git/docs/howto/isolate-bugs-with-bisect.txt
>
>
I don't think so, because suspend to disk began to work again with the
latest commit.

But it seems to have to do with dev->msi_enabled, that is enabled, but
the msi_attrib for this device is empty (pos is zero) in the function
__pci_restore_msi_state. irq=19 and msi_attrib=0 before the bug happens
(trying to accessing msi_attrib.pos).
But 19 is not an msi interrupt, isn't it?

CPU0 CPU1
0: 168998 0 IO-APIC-edge timer
9: 9056 0 IO-APIC-fasteoi acpi
16: 43 0 IO-APIC-fasteoi uhci_hcd:usb5
17: 8278 19996 IO-APIC-fasteoi wifi0
18: 11388 0 IO-APIC-fasteoi libata, uhci_hcd:usb4
19: 3 0 IO-APIC-fasteoi uhci_hcd:usb3, ohci1394
20: 115285 0 IO-APIC-fasteoi ehci_hcd:usb1, uhci_hcd:usb2
21: 225 0 IO-APIC-fasteoi HDA Intel
218: 14365 0 PCI-MSI-edge libata
219: 1 0 PCI-MSI-edge eth0
NMI: 0 0
LOC: 49119 88723
ERR: 0
MIS: 0

Shouldn't the flag msi_enabled be set to zero in function
msi_remove_pci_irq_vectors?


-
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/