Re: USB related crash on 5.99.01

["Arnaud Lacombe" <lacombar%gmail.com@localhost>]

On Tue, Dec 9, 2008 at 4:47 PM, Jukka Salmi <j+nbsd%2008.salmi.ch@localhost> 
wrote:
> Hello,
>
> a 5.99.01 i386 system just crashed when I connected a USB device:
>
> $ dmesg
> [...]
> umass0 at uhub0 port 2 configuration 1 interface 0
> umass0: Rockbox.org Rockbox media player, rev 1.10/1.00, addr 2
> umass0: using SCSI over Bulk-Only
> scsibus0 at umass0: 2 targets, 1 lun per target
> sd0 at scsibus0 target 0 lun 0: <Rockbox, Internal Storage, 0.00> disk 
> removable
> sd0: drive offline
> umass0: at uhub0 port 2 (addr 2) disconnected
> sd0(umass0:0:0:0): generic HBA error
> sd0(umass0:0:0:0): generic HBA error
> sd0: unable to open device, error = 5
> sd0(umass0:0:0:0): generic HBA error
> sd0(umass0:0:0:0): generic HBA error
> sd0: detached
> scsibus0: detached
> umass0: detached
> uvm_fault(0xc0524f80, 0, 1) -> 0xe
> fatal page fault in supervisor mode
> trap type 6 code 0 eip c01b5a7e cs 8 eflags 10296 cr2 4 ilevel 4
> kernel: supervisor trap page fault, code=0
> Stopped in pid 0.26 (system) at netbsd:uhci_device_ctrl_start+0xe:      movl  
>   0x4(%esi),%edx
> db{0}> db{0}> db{0}> 
> uhci_device_ctrl_start(c1f90500,cc5d5e1c,6,0,6,cc8e,0,1,4,c1f90400) at 
> netbsd:uhci_device_ctrl_start+0xe
> usb_transfer_complete(c1f90400,0,c0482885,0,0,0,cb4e7c4c,c02ed68e,c2349980,cb38c000)
>  at netbsd:usb_transfer_complete+0x10f
> uhci_abort_xfer(c1f90400,0,0,c040ba91,c2349980,c1971544,cb4e7c8c,c040ee31,c2349980,0)
>  at netbsd:uhci_abort_xfer+0x20e
> usbd_abort_pipe(c2349980,0,cb4e7cac,c040efd6,4,10,cb4e7cac,c040ee78,c0524140,c1971544)
>  at netbsd:usbd_abort_pipe+0x39
> usbd_kill_pipe(c0524140,c1971544,cbd0d41c,10,c1971544,c2349180,cb4e7cfc,c0411924,c2349180,cbd0d400)
>  at netbsd:usbd_kill_pipe+0x11
> usb_free_device(c2349180,cbd0d400,10,c02f5279,0,cabd3560,c1971500,cbd0d200,2,108)
>  at netbsd:usb_free_device+0x18
> uhub_explore(c1971f00,cb38cd30,cb4e7d2c,c040b317,1770,0,c04b9274,1770,0,cb38cd30)
>  at netbsd:uhub_explore+0x124
> usb_discover(1770,0,c04b9274,1770,0,cb38cd30,0,c01002e1,cb38cd30,0) at 
> netbsd:usb_discover+0x2a
> usb_event_thread(cb38cd30,0,c01002cd,0,c01002cd,0,0,0,0,0) at 
> netbsd:usb_event_thread+0x47
> db{0}> syncing disks... 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 giving up
>
> dumping to dev 0,1 offset 3146535
> dump Copyright (c) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005,
> [...]
>
> Any chance that this has been fixed in the meantime?
>
this is a race in the usb code, same trace as kern/39651.

I didn't had much time debugging it as it happen only on a box I need
alive. When the panic fires, is the machine under load (cpu-wise,
disk-wise, or both) ?

Thanks,

 - Arnaud