stable/9 panic Bad tailq NEXT(0xffffffff80e52660->tqh_last) != NULL

[John Baldwin]

On Monday, July 16, 2012 5:39:26 am Andriy Gapon wrote:
> on 13/07/2012 19:31 Sean Bruno said the following:
> > pcib7: <ACPI PCI-PCI bridge> at device 28.0 on pci0
> > pcib7:   domain            0
> > pcib7:   secondary bus     7
> > pcib7:   subordinate bus   7
> > pcib7:   no prefetched decode
> > device_attach: pcib7 attach returned 6
> 
> It seems that the following panic is caused by the above attach failure.
> Namely, it seems that there are rman members in pcib_softc via pcib_window fields.
> Those members are passed to rman_init during attach attempt, but they are not
> passed to rman_fini if the attach fails (no attach failure cleanup).
> After newbus code deallocates pcib_softc the global rman list (rman_head) becomes
> corrupted.

Oof.  None of the pcib(4) drivers really support detach or cleanup yet.

-- 
John Baldwin