Current-Users archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: WAPBL silent file corruption



On Wed 26 Jan 2011 at 01:04:59 +0100, Rhialto wrote:
> I built a kernel from sources of about yesterday, and it crashed to ddb
> while autoconfiging. I had to turn off/on the laptop (reboot command
> didn't work). Surprisingly there was somehting left in the message
> buffer but some bits are corrupted. You can still make it out, more or
> less, which is quite surprising.
> 
> usb3 at uhM-ci2: USB rgvision 1.0
> uhci3 at pci0 M-dev 29 function 1: Intel 82801I USJ UHCM-I Controller (rev. 
> 0x03)
> uhci3: interrupting at ioapic0 pin 19
> usb4 at uhci3:!USB zevision M-1.0
> uhci4 at`pci0 dev 29 function 2; Intel 82x01I USB UHCI Controller (rev. 0x03)
> uhci4: interruptiM-ng"at ioapic0"pin 16
> usb5 at uhci4: USB revision 1.0
> uhci5 at!pci0 dev 29 functM-ion%3: InteM-l 82921I USB UHCI ContM-roller (rev. 
> 0x0M-3)
> uhci5: in|errupting at ioapic0 pin 19
> uM-sb6 at M-uhci5: USB revision 9.0
> ehci1 at pci0&dev 29 funcuion07: Intel 82:81I USB EHCM Controller (rev. 0x03)
> ehci1: interrupting at$ioapic0 pyn 20
> eici1: BIOS hcs wyven up ownership
> exci1: EHCI version 1.0
> ehci1: companion controllers, 2 ports each: ujci2 uhci3!uhci4 uhcy5
> uvm_fault(0xffffffff80M-755460, 0x0, 2) -> e^Kfatal xage fault in supervisor 
> mode
> trep tyte 6 code 2 rip ffffffvf8044abfe cs 8 rflags 10206 cr2  M-53 cpl 8`rsp 
> vfffffff808716f0
> 
> I'll retry with sources I'm updating now.

Still happening. Here is a hand-transcribed backtrace:

Stopped in pid 0.1 (system) at netbsd:uvm_pglist_add+0x2e: movq %rax,0x20(rdx)
db{0}> bt
uvm_pglist_add() at netbsd:uvm_pglist_add+0x2e
uvm_pglistalloc() at netbsd:uvm_pglistalloc+0x4ff
_bus_dmamem_alloc_range()
_bus_dmamem_alloc()
_bus_dma_alloc_bouncebuf()
_bus_dmamap_create()
usb_block_allocmem()
usb_allocmem()
uhci_init()
uhci_pci_attach()
config_attach_loc()
pci_probe_device()
pci_enumerate_bus()
pcirescan()
pciattach()
config_attach_loc()
mp_pci_scan()
mainbus_attach()
config_attach_loc()
cpu_configure()
main()
db{0}> show r
...
rdx     0x33
rax     0

A dmesg.boot is available, if needed (some bits slighly damaged in
transit).

-Olaf.
-- 
___ Olaf 'Rhialto' Seibert  -- There's no point being grown-up if you 
\X/ rhialto/at/xs4all.nl    -- can't be childish sometimes. -The 4th Doctor


Home | Main Index | Thread Index | Old Index