Worrisome USB disk messages in 4.6.0-1, but not 4.5.0-2

Brian Flaherty

unread,

Jul 29, 2016, 2:50:03 PM7/29/16

to

Hello,

On my desktop, I've been getting messages like this in dmesg connected with an external USB hard disk. This is a block of them from this morning:

Jul 29 10:21:48 pelman kernel: sd 5:0:0:0: [sdb] tag#0 FAILED Result: hostbyte=DID_ERROR driverbyte=DRIVER_SENSE
Jul 29 10:21:48 pelman kernel: sd 5:0:0:0: [sdb] tag#0 Sense Key : Hardware Error [current] [descriptor]
Jul 29 10:21:48 pelman kernel: sd 5:0:0:0: [sdb] tag#0 Add. Sense: No additional sense information
Jul 29 10:21:48 pelman kernel: sd 5:0:0:0: [sdb] tag#0 CDB: ATA command pass through(16) 85 06 2c 00 00 00 00 00 00 00 00 00 00 00 e5 00
Jul 29 10:21:48 pelman kernel: sd 5:0:0:0: [sdb] tag#0 FAILED Result: hostbyte=DID_ERROR driverbyte=DRIVER_SENSE
Jul 29 10:21:48 pelman kernel: sd 5:0:0:0: [sdb] tag#0 Sense Key : Hardware Error [current] [descriptor]
Jul 29 10:21:48 pelman kernel: sd 5:0:0:0: [sdb] tag#0 Add. Sense: No additional sense information
Jul 29 10:21:48 pelman kernel: sd 5:0:0:0: [sdb] tag#0 CDB: ATA command pass through(16) 85 06 2c 00 da 00 00 00 00 00 4f 00 c2 00 b0 00
Jul 29 10:51:55 pelman kernel: sd 5:0:0:0: [sdb] tag#0 FAILED Result: hostbyte=DID_ERROR driverbyte=DRIVER_SENSE
Jul 29 10:51:55 pelman kernel: sd 5:0:0:0: [sdb] tag#0 Sense Key : Hardware Error [current] [descriptor]
Jul 29 10:51:55 pelman kernel: sd 5:0:0:0: [sdb] tag#0 Add. Sense: No additional sense information
Jul 29 10:51:55 pelman kernel: sd 5:0:0:0: [sdb] tag#0 CDB: ATA command pass through(16) 85 06 2c 00 00 00 00 00 00 00 00 00 00 00 e5 00
Jul 29 10:51:55 pelman kernel: sd 5:0:0:0: [sdb] tag#0 FAILED Result: hostbyte=DID_ERROR driverbyte=DRIVER_SENSE
Jul 29 10:51:55 pelman kernel: sd 5:0:0:0: [sdb] tag#0 Sense Key : Hardware Error [current] [descriptor]
Jul 29 10:51:55 pelman kernel: sd 5:0:0:0: [sdb] tag#0 Add. Sense: No additional sense information
Jul 29 10:51:55 pelman kernel: sd 5:0:0:0: [sdb] tag#0 CDB: ATA command pass through(16) 85 06 2c 00 da 00 00 00 00 00 4f 00 c2 00 b0 00

I've been searching on the web (duckduckgo and google), but haven't seen this exact situation anywhere. On a bug report I read (https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1176355), the person mentioned that their disk worked on a previous release, so I attached my disk to a laptop that hasn't been updated in a while. It is running Debian unstable with linux 4.5.0-2. None of those messages have shown up. I'm running fsck -f on it right now to check it (on the laptop).

Thus, I assume there is some change in the kernel code, but that my disk is fine. Does that sound reasonable, or has some improvement been introduced in the linux code that is picking up problems that 4.5 doesn't see?

Thanks for your time and happy to include further information if it is helpful.

Brian Flaherty

unread,

Jul 29, 2016, 4:00:04 PM7/29/16

to

On Friday, July 29, 2016 11:39 AM, Brian Flaherty <bx...@yahoo.com> wrote:

>

> Hello,

>

> On my desktop, I've been getting messages like this in dmesg connected with an external USB hard disk. This is a block of them from this morning:

[ rest deleted ]

If it is worth more information, I rebooted the desktop with kernel 4.5.0-2-amd64 and I've not seen any of the messages, so it seem related to kernel 4.6.0-1.

Michael Schaffner

unread,

Jul 30, 2016, 7:00:04 AM7/30/16

to

Hi,

same here (running debian jessie with kernels from debian-backports):
With 4.5.0-0.bpo.2-amd64 there are no messages, with 4.6.0-0.bpo.1-amd64
i get them:
Jul 18 09:25:30 mvdr2 kernel: [ 461.495406] sd 10:0:0:0: [sdb] tag#0
FAILED Result: hostbyte=DID_ERROR driverbyte=DRIVER_SENSE
Jul 18 09:25:30 mvdr2 kernel: [ 461.495412] sd 10:0:0:0: [sdb] tag#0

Sense Key : Hardware Error [current] [descriptor]

Jul 18 09:25:30 mvdr2 kernel: [ 461.495415] sd 10:0:0:0: [sdb] tag#0

Add. Sense: No additional sense information

Jul 18 09:25:30 mvdr2 kernel: [ 461.495419] sd 10:0:0:0: [sdb] tag#0
CDB: ATA command pass through(12)/Blank a1 06 20 00 00 00 00 00 00 e5 00 00

>I've been searching on the web (duckduckgo and google), but haven't
>seen this exact situation anywhere.

I came across https://bugzilla.redhat.com/show_bug.cgi?id=1351305, which
seems to be related.

Bye..
Michael