Bugzilla Bug 250406

Here are the latest kernel messages:

usb-storage: device found at 4
usb-storage: waiting for device to settle before scanning
usb-storage: device scan complete
scsi scan: INQUIRY result too short (5), using 36
scsi 4:0:0:0: Direct-Access     PENTAX   DIGITAL_CAMERA   1.00 PQ: 0 ANSI: 0
sd 4:0:0:0: [sdc] 3985409 512-byte hardware sectors (2041 MB)
sd 4:0:0:0: [sdc] Write Protect is off
sd 4:0:0:0: [sdc] Mode Sense: 0a 0a 00 01
sd 4:0:0:0: [sdc] Assuming drive cache: write through
sd 4:0:0:0: [sdc] 3985409 512-byte hardware sectors (2041 MB)
sd 4:0:0:0: [sdc] Write Protect is off
sd 4:0:0:0: [sdc] Mode Sense: 0a 0a 00 01
sd 4:0:0:0: [sdc] Assuming drive cache: write through
 sdc: sdc1
sd 4:0:0:0: [sdc] Attached SCSI removable disk
sd 4:0:0:0: Attached scsi generic sg2 type 0
usb 2-1: reset full speed USB device using uhci_hcd and address 4
usb 2-1: reset full speed USB device using uhci_hcd and address 4
usb 2-1: reset full speed USB device using uhci_hcd and address 4
usb 2-1: reset full speed USB device using uhci_hcd and address 4

However, no /dev/sdc exists for this what I think should be a mass storage
device. I'm using sys-fs/udev-124-r1 and sys-kernel/gentoo-sources-2.6.27-r5.

Created an attachment (id=174734) [details]
lsusb output for the camera

Created an attachment (id=174737) [details]
kernel configuration

In addition to the kernel messages above, the following kernel messages had
also appeared before I shut down the camera in the end:

usb 2-1: reset full speed USB device using uhci_hcd and address 4
usb 2-1: reset full speed USB device using uhci_hcd and address 4
sd 4:0:0:0: [sdc] Result: hostbyte=DID_ABORT driverbyte=DRIVER_OK,SUGGEST_OK
end_request: I/O error, dev sdc, sector 3985408
Buffer I/O error on device sdc, logical block 3985408
usb 2-1: reset full speed USB device using uhci_hcd and address 4
usb 2-1: reset full speed USB device using uhci_hcd and address 4
usb 2-1: reset full speed USB device using uhci_hcd and address 4
usb 2-1: reset full speed USB device using uhci_hcd and address 4
usb 2-1: reset full speed USB device using uhci_hcd and address 4
usb 2-1: reset full speed USB device using uhci_hcd and address 4
sd 4:0:0:0: [sdc] Result: hostbyte=DID_ABORT driverbyte=DRIVER_OK,SUGGEST_OK
end_request: I/O error, dev sdc, sector 3985408
Buffer I/O error on device sdc, logical block 3985408
usb 2-1: reset full speed USB device using uhci_hcd and address 4
usb 2-1: reset full speed USB device using uhci_hcd and address 4
usb 2-1: reset full speed USB device using uhci_hcd and address 4
usb 2-1: reset full speed USB device using uhci_hcd and address 4
usb 2-1: reset full speed USB device using uhci_hcd and address 4
usb 2-1: reset full speed USB device using uhci_hcd and address 4
sd 4:0:0:0: [sdc] Result: hostbyte=DID_ABORT driverbyte=DRIVER_OK,SUGGEST_OK
end_request: I/O error, dev sdc, sector 3985408
Buffer I/O error on device sdc, logical block 3985408
usb 2-1: reset full speed USB device using uhci_hcd and address 4
usb 2-1: reset full speed USB device using uhci_hcd and address 4
usb 2-1: reset full speed USB device using uhci_hcd and address 4
usb 2-1: reset full speed USB device using uhci_hcd and address 4
usb 2-1: reset full speed USB device using uhci_hcd and address 4
sd 4:0:0:0: [sdc] Result: hostbyte=DID_ABORT driverbyte=DRIVER_OK,SUGGEST_OK
end_request: I/O error, dev sdc, sector 3985408
Buffer I/O error on device sdc, logical block 3985408
usb 2-1: reset full speed USB device using uhci_hcd and address 4
usb 2-1: reset full speed USB device using uhci_hcd and address 4
usb 2-1: reset full speed USB device using uhci_hcd and address 4
usb 2-1: reset full speed USB device using uhci_hcd and address 4
usb 2-1: USB disconnect, address 4
scsi 4:0:0:0: [sdc] Result: hostbyte=DID_ERROR driverbyte=DRIVER_OK,SUGGEST_OK
end_request: I/O error, dev sdc, sector 3985408
Buffer I/O error on device sdc, logical block 3985408
scsi 4:0:0:0: [sdc] Result: hostbyte=DID_NO_CONNECT
driverbyte=DRIVER_OK,SUGGEST_OK
end_request: I/O error, dev sdc, sector 735
Buffer I/O error on device sdc1, logical block 488
Buffer I/O error on device sdc1, logical block 489
Buffer I/O error on device sdc1, logical block 490
Buffer I/O error on device sdc1, logical block 491
Buffer I/O error on device sdc1, logical block 492
Buffer I/O error on device sdc1, logical block 493
Buffer I/O error on device sdc1, logical block 494
Buffer I/O error on device sdc1, logical block 495
Buffer I/O error on device sdc1, logical block 496
scsi 4:0:0:0: rejecting I/O to dead device

PS: The device shows no problems on Windows XP, although I don't know whether
that specific windows installation had any Pentax drivers installed (if such
drivers exist at all).

I don't believe that this has something to do with our gentoo-sources or udev.
Probably, your usb physical connection(cable, connectors etc.) is problematic.
Try to CONFIG_USB_DEBUG=y and CONFIG_USB_STORAGE_DEBUG=y in order to see some
more verbose messages about what really happens. Maybe you need to enable
CONFIG_USB_<something> or CONFIG_USB_STORAGE_<something>. I don't want to
assign any bug to the kernel team before we 're sure that this isn't a
misconfiguration or something of your side(cables etc.). Reply here when you're
sure that it's a bug in udev or gentoo-sources. 

Created an attachment (id=174848) [details]
kernel (debug) messages

Ok, I recompiled the kernel with USB and USB mass storage verbose debug
messages. Here are all kernel messages that I think might be related to this
case. This is what I changed in the kernel configuration just in case. I'm
quite certain the other USB options presented by the kernels xconfig are
irrelevant.

CONFIG_USB_DEBUG=y
CONFIG_USB_SUSPEND=y
CONFIG_USB_STORAGE_DEBUG=y
CONFIG_USB_LD=y

According to drivers/usb/storage/unusual_devs.h, some Pentax Cameras have also
been found not strictly follow the USB mass storage specifications. Perhaps
that's the case?

I still don't believe that this is gentoo specific. Probably you would
experience the same problem with vanilla-sources. But I'm not a kernel expert,
so I will let the proper team to reply. Assigned to kernel team.

Created an attachment (id=174857) [details]
possible fix

Thanks for the detailed debugging logs - helps us get straight to the problem.
Your device is crashing when Linux tries to access the very final sector. This
is a common problem where the device doesn't follow the SCSI specs -- it is
supposed to report the address of the final sector, but it instead reports the
number of sectors on the device (which is 1 higher, because sectors start at
0).

This patch should fix it, please give it a go.

Please also post the /proc/bus/usb/devices entry for your device (you have
already posted equivalent info, but this is the form that kernel developers
prefer to see included in the patch description)

The patch works like a charm. Thanks! =)

Here's the entry from /proc/bus/usb/devices:

T:  Bus=02 Lev=01 Prnt=01 Port=00 Cnt=01 Dev#=  2 Spd=12  MxCh= 0
D:  Ver= 2.00 Cls=00(>ifc ) Sub=00 Prot=00 MxPS=64 #Cfgs=  1
P:  Vendor=0a17 ProdID=006e Rev= 1.00
S:  Manufacturer=PENTAX Corporation
S:  Product=K10D
C:* #Ifs= 1 Cfg#= 1 Atr=c0 MxPwr=  2mA
I:* If#= 0 Alt= 0 #EPs= 3 Cls=08(stor.) Sub=06 Prot=50 Driver=usb-storage
E:  Ad=81(I) Atr=02(Bulk) MxPS=  64 Ivl=0ms
E:  Ad=02(O) Atr=02(Bulk) MxPS=  64 Ivl=0ms
E:  Ad=83(I) Atr=03(Int.) MxPS=  64 Ivl=100ms

Patch was acked by upstream, so I've included it for the next 2.6.26/2.6.27
releases

Fixed in gentoo-sources-2.6.27-r6, closing this bug now as I don't think we'll
do more 2.6.26 releases. (but if we do, this patch will be in there)