Re: 2.6.7-mm4: regression: ieee1394: sbp2: null pointer dereference

["Matthias Urlichs" <smurf@smurf.noris.de>]

Hi,

Ben Collins:
> This oops traces back into the scsi stack, right?

Umm ... not that I know of. It basically says

                      sbp2_logout_device+0x13/0x140 [sbp2]
kernel:  [<fb2c49f1>] sbp2_update+0x21/0x80 [sbp2]

No SCSI anywhere (that I can see).

sbp2_update+0x21 points to the instruction after a call to
sbp2_logout_device(), so that matches up. GDB says the error is here:

sbp2.c:1323:	scsi_id->logout_orb->reserved1 = 0x0;

which probably means that either scsi_id or logout_orb is NULL.

>         I guess USB doesn't have too many problems since it does a
> scsi-host per device, but that's not as easy with sbp2 and 1394, since a
> single sbp2 device can have multiple LUN's, and it's just easier to
> treat that as one scsi host.
> 
So can USB; I have a card reader here which registers a different LUN
per card interface.

> I can't reproduce it, but I'll try to get into the logic of sbp2 device
> removal again to see if I can find out where and why this is occuring.

FWIW, here's the device info (from gscanbus):

SelfID Info
-----------
Physical ID: 2
Link active: Yes
Gap Count: 10
PHY Speed: S400
PHY Delay: <=144ns
IRM Capable: No
Power Class: None
Port 0: Connected to parent node
Port 1: Connected to child node
Init. reset: No

CSR ROM Info
------------
GUID: 0x00A0B80A0000144F
Node Capabilities: 0x000083C0
Vendor ID: 0x0000A0B8
Unit Spec ID: 0x0000609E
Unit SW Version: 0x00010483
Model ID: 0x00000000
Nr. Textual Leafes: 2

Vendor:  SYMBIOS LOGIC INC.
Textual Leafes: 
LSI Logic
LSI 501 rev B3

AV/C Subunits
-------------
N/A



-- 
Matthias Urlichs   |   {M:U} IT Design @ m-u-it.de   |  smurf@smurf.noris.de