Can anyone shed light on what the following entries in /var/log/syslog actually mean?
Jun 16 11:43:45 mark-pc kernel: [92204.212656] sd 11:0:0:2: [sdf] tag#0 FAILED Result: hostbyte=DID_ERROR driverbyte=DRIVER_SENSE Jun 16 11:43:45 mark-pc kernel: [92204.212669] sd 11:0:0:2: [sdf] tag#0 Sense Key : Hardware Error [current] Jun 16 11:43:45 mark-pc kernel: [92204.212675] sd 11:0:0:2: [sdf] tag#0 Add. Sense: No additional sense information Jun 16 11:43:45 mark-pc kernel: [92204.212683] sd 11:0:0:2: [sdf] tag#0 CDB: Read(10) 28 00 06 87 47 30 00 00 08 00 Jun 16 11:43:45 mark-pc kernel: [92204.212688] blk_update_request: I/O error, dev sdf, sector 109528880 Jun 16 11:43:45 mark-pc kernel: [92204.212696] Buffer I/O error on dev sdf1, logical block 13690854, async page read
Background: This a pretty new Sandisk [1] 128GB MicroSD (UHS) card, taken out of my phone and plugged into my PC via a card reader (USB2). A previous card caused issues in my phone so I replaced it with this one, and it now is not behaving well (usually OK but sometimes if I try to look at photos on the card the phone will lock up temporarily).
I'm trying to back everything up onto my PC so I can copy it across onto a Samsung card to see if that's better. The above errors crop up periodically in syslog whilst copying.
Is the issue the card, the card reader, the PC, the drivers, ...?
[1] I believe it to be genuine Sandisk, it was purchased from Amazon. Advice on how to verify it's not a fake welcomed!
Having been previously bitten, it seems the only way to spot the fake cards is when they don't work properly.
One thought is do the devices support cards of that size?
-- Sent from my Psion 5MX Original Message From: Mark Rogers Sent: Thursday, 16 June 2016 10:55 To: alug-main Subject: [ALUG] MicroSD card read issues
Can anyone shed light on what the following entries in /var/log/syslog actually mean?
Jun 16 11:43:45 mark-pc kernel: [92204.212656] sd 11:0:0:2: [sdf] tag#0 FAILED Result: hostbyte=DID_ERROR driverbyte=DRIVER_SENSE Jun 16 11:43:45 mark-pc kernel: [92204.212669] sd 11:0:0:2: [sdf] tag#0 Sense Key : Hardware Error [current] Jun 16 11:43:45 mark-pc kernel: [92204.212675] sd 11:0:0:2: [sdf] tag#0 Add. Sense: No additional sense information Jun 16 11:43:45 mark-pc kernel: [92204.212683] sd 11:0:0:2: [sdf] tag#0 CDB: Read(10) 28 00 06 87 47 30 00 00 08 00 Jun 16 11:43:45 mark-pc kernel: [92204.212688] blk_update_request: I/O error, dev sdf, sector 109528880 Jun 16 11:43:45 mark-pc kernel: [92204.212696] Buffer I/O error on dev sdf1, logical block 13690854, async page read
Background: This a pretty new Sandisk [1] 128GB MicroSD (UHS) card, taken out of my phone and plugged into my PC via a card reader (USB2). A previous card caused issues in my phone so I replaced it with this one, and it now is not behaving well (usually OK but sometimes if I try to look at photos on the card the phone will lock up temporarily).
I'm trying to back everything up onto my PC so I can copy it across onto a Samsung card to see if that's better. The above errors crop up periodically in syslog whilst copying.
Is the issue the card, the card reader, the PC, the drivers, ...?
[1] I believe it to be genuine Sandisk, it was purchased from Amazon. Advice on how to verify it's not a fake welcomed!
On 16 June 2016 at 12:00, huge@huge.org.uk wrote:
One thought is do the devices support cards of that size?
The card reader should (it's a new USB3 Anker card reader) but it's a good point I hadn't considered.
The other device is a Samsung S7 Edge so it should have no issues.
I'll see how I get on with the Samsung card. But I'd still like to understand what the messages in syslog mean.
On 16 June 2016 13:16:32 GMT+01:00, Mark Rogers mark@more-solutions.co.uk wrote:
But I'd still like to understand what the messages in syslog mean.
What do you mean what does it mean? ;)
It's trying to read from the SCSI device and getting an error code in reply. More detail than that probably needs me to read the SCSI protocol descriptions and driver source code.
On 22 June 2016 at 14:24, MJ Ray mjr@phonecoop.coop wrote:
On 16 June 2016 13:16:32 GMT+01:00, Mark Rogers mark@more-solutions.co.uk wrote:
But I'd still like to understand what the messages in syslog mean.
What do you mean what does it mean? ;)
Well at the macro level, is the uSD card rubbish, or is the adapter or USB interface the problem, or are these problems down to something else?
But at the micro level, I am interested to know what part of the system is saying what and which parts of the error are meaningful, in as much as if I see a broadly similar error message in a Google search result, which bits have to match for it to be relevant, etc.
It's trying to read from the SCSI device and getting an error code in reply. More detail than that probably needs me to read the SCSI protocol descriptions and driver source code.
OK, so which bits mean that?
In particular, what does Sense Key mean? ... except that I've just established for myself that this is the SCSI error code (in this case "Hardware Error", right?). Supposition on my part: Add = Additional, where additional info might be included with the error code (sense key) but isn't here? With the next couple of lines indicating where on the device the error occurred?
But then I am thrown by the Buffer I/O error - to me a buffer error implies it's not the media that's at fault but something in the interface between it and the O/S, suggesting that the problem is in the PC hardware (or USB card reader) not the card itself - but now I'm just guessing....
Jun 16 11:43:45 mark-pc kernel: [92204.212656] sd 11:0:0:2: [sdf] tag#0 FAILED Result: hostbyte=DID_ERROR driverbyte=DRIVER_SENSE Jun 16 11:43:45 mark-pc kernel: [92204.212669] sd 11:0:0:2: [sdf] tag#0 Sense Key : Hardware Error [current] Jun 16 11:43:45 mark-pc kernel: [92204.212675] sd 11:0:0:2: [sdf] tag#0 Add. Sense: No additional sense information Jun 16 11:43:45 mark-pc kernel: [92204.212683] sd 11:0:0:2: [sdf] tag#0 CDB: Read(10) 28 00 06 87 47 30 00 00 08 00 Jun 16 11:43:45 mark-pc kernel: [92204.212688] blk_update_request: I/O error, dev sdf, sector 109528880 Jun 16 11:43:45 mark-pc kernel: [92204.212696] Buffer I/O error on dev sdf1, logical block 13690854, async page read
On Mon, Jun 27, 2016 at 12:45:13PM +0100, Mark Rogers wrote:
On 22 June 2016 at 14:24, MJ Ray mjr@phonecoop.coop wrote:
It's trying to read from the SCSI device and getting an error code in reply. More detail than that probably needs me to read the SCSI protocol descriptions and driver source code.
OK, so which bits mean that?
In particular, what does Sense Key mean? ... except that I've just established for myself that this is the SCSI error code (in this case "Hardware Error", right?). Supposition on my part: Add = Additional, where additional info might be included with the error code (sense key) but isn't here? With the next couple of lines indicating where on the device the error occurred?
When a SCSI device hits an error it returns 3 bytes of data about what that was; the sense key (SK), then potentially an additional sense code (ASC) and an additional sense code qualifier (ASCQ).
The "CDB" in the error below is the SCSI command that was sent to the device - a "Read(10)" which means a 10 byte read command (there are read6 and read16 read commands - largely about how many bytes there are for describing the sector to read).
T10/SBC3 explains the READ10 command:
28 = READ10 00 = (flags) 06 87 47 30 = Sector 109528880 00 = (reserved) 00 08 = transfer length, 8 blocks (8 * 512 = 4k) 00 = (control byte)
So your kernel sent down a read, there was an return code 04/00/00 i.e. a hardware error with no additional information, and as a result the upper layers said there was an error filling the read buffer.
But then I am thrown by the Buffer I/O error - to me a buffer error implies it's not the media that's at fault but something in the interface between it and the O/S, suggesting that the problem is in the PC hardware (or USB card reader) not the card itself - but now I'm just guessing....
Jun 16 11:43:45 mark-pc kernel: [92204.212656] sd 11:0:0:2: [sdf] tag#0 FAILED Result: hostbyte=DID_ERROR driverbyte=DRIVER_SENSE Jun 16 11:43:45 mark-pc kernel: [92204.212669] sd 11:0:0:2: [sdf] tag#0 Sense Key : Hardware Error [current] Jun 16 11:43:45 mark-pc kernel: [92204.212675] sd 11:0:0:2: [sdf] tag#0 Add. Sense: No additional sense information Jun 16 11:43:45 mark-pc kernel: [92204.212683] sd 11:0:0:2: [sdf] tag#0 CDB: Read(10) 28 00 06 87 47 30 00 00 08 00 Jun 16 11:43:45 mark-pc kernel: [92204.212688] blk_update_request: I/O error, dev sdf, sector 109528880 Jun 16 11:43:45 mark-pc kernel: [92204.212696] Buffer I/O error on dev sdf1, logical block 13690854, async page read
J.
I've used h2testw for Windows to prove cards I bought from eBay\Amazon were fake in the past. I don't buy these things from eBay or Amazon now because of these issues - I find I get better results when buying from trusted retailers such as: . My Memory . eBuyer . 7 Day Shop . etc
You can download h2testw from: http://www.heise.de/download/product/h2testw-50539
I've also seen recommendations for F3 on Linux: http://oss.digirati.com.br/f3/
The Open Source Photrec software produces good results for data recovery as does the commercial R-Studio: . http://www.cgsecurity.org/wiki/PhotoRec . http://www.r-studio.com/
Beware though that most of these data testing tools have destructive tests that overwrite all blocks, so be careful what options you choose!
On Mon, 27 Jun 2016 at 13:25 Jonathan McDowell noodles@earth.li wrote:
On Mon, Jun 27, 2016 at 12:45:13PM +0100, Mark Rogers wrote:
On 22 June 2016 at 14:24, MJ Ray mjr@phonecoop.coop wrote:
It's trying to read from the SCSI device and getting an error code in reply. More detail than that probably needs me to read the SCSI protocol descriptions and driver source code.
OK, so which bits mean that?
In particular, what does Sense Key mean? ... except that I've just established for myself that this is the SCSI error code (in this case "Hardware Error", right?). Supposition on my part: Add = Additional, where additional info might be included with the error code (sense key) but isn't here? With the next couple of lines indicating where on the device the error occurred?
When a SCSI device hits an error it returns 3 bytes of data about what that was; the sense key (SK), then potentially an additional sense code (ASC) and an additional sense code qualifier (ASCQ).
The "CDB" in the error below is the SCSI command that was sent to the device - a "Read(10)" which means a 10 byte read command (there are read6 and read16 read commands - largely about how many bytes there are for describing the sector to read).
T10/SBC3 explains the READ10 command:
28 = READ10 00 = (flags) 06 87 47 30 = Sector 109528880 00 = (reserved) 00 08 = transfer length, 8 blocks (8 * 512 = 4k) 00 = (control byte)
So your kernel sent down a read, there was an return code 04/00/00 i.e. a hardware error with no additional information, and as a result the upper layers said there was an error filling the read buffer.
But then I am thrown by the Buffer I/O error - to me a buffer error implies it's not the media that's at fault but something in the interface between it and the O/S, suggesting that the problem is in the PC hardware (or USB card reader) not the card itself - but now I'm just guessing....
Jun 16 11:43:45 mark-pc kernel: [92204.212656] sd 11:0:0:2: [sdf] tag#0 FAILED Result: hostbyte=DID_ERROR driverbyte=DRIVER_SENSE Jun 16 11:43:45 mark-pc kernel: [92204.212669] sd 11:0:0:2: [sdf] tag#0 Sense Key : Hardware Error [current] Jun 16 11:43:45 mark-pc kernel: [92204.212675] sd 11:0:0:2: [sdf] tag#0 Add. Sense: No additional sense information Jun 16 11:43:45 mark-pc kernel: [92204.212683] sd 11:0:0:2: [sdf] tag#0 CDB: Read(10) 28 00 06 87 47 30 00 00 08 00 Jun 16 11:43:45 mark-pc kernel: [92204.212688] blk_update_request: I/O error, dev sdf, sector 109528880 Jun 16 11:43:45 mark-pc kernel: [92204.212696] Buffer I/O error on dev sdf1, logical block 13690854, async page read
J.
-- /-\ | I'm not allowed to talk to you any |@/ Debian GNU/Linux Developer | more. - |
main@lists.alug.org.uk http://www.alug.org.uk/ http://lists.alug.org.uk/mailman/listinfo/main Unsubscribe? See message headers or the web site above!
On Mon, 2016-06-27 at 12:56 +0000, Karl Foley wrote:
I've used h2testw for Windows to prove cards I bought from eBay\Amazon were fake in the past. I don't buy these things from eBay or Amazon now because of these issues
I agree about eBay, but you have to distinguish between things actually sold by Amazon and people using Amazon Marketplace to sell their cr@p.
On 30 June 2016 at 11:25, Huge huge@huge.org.uk wrote:
I agree about eBay, but you have to distinguish between things actually sold by Amazon and people using Amazon Marketplace to sell their cr@p.
Agreed, although the card I'm having a problem was did come from Amazon directly (not another seller), and is the second such card to give me problems. A Sumsung card is working fine though.
No be clear, at this point I have no reason to think that the card is fake.
Mark
-
Huge -
huge＠huge.org.uk
-
Jonathan McDowell -
Karl Foley -
Mark Rogers -
MJ Ray