Re: another issue with ubifs on v3.6-rc2: attaching an formatted device

[Marc Kleine-Budde <mkl@pengutronix.de>]

[-- Attachment #1: Type: text/plain, Size: 2742 bytes --]

On 08/21/2012 12:15 PM, Artem Bityutskiy wrote:
> On Tue, 2012-08-21 at 11:35 +0200, Uwe Kleine-König wrote:
>> On Mon, Aug 20, 2012 at 10:11:07PM +0300, Artem Bityutskiy wrote:
>>> On Mon, 2012-08-20 at 18:07 +0200, Uwe Kleine-König wrote:
>>>> [  163.714143] uncorrectable error : 
>>>> [  163.727597] UBI warning: ubi_io_read: error -74 (ECC error) while
>>>> reading 64 bytes from PEB 0:0, read only 64 bytes, retry
>>>> [  163.740737] uncorrectable error : 
>>>> [  163.744032] UBI warning: ubi_io_read: error -74 (ECC error) while
>>>> reading 64 bytes from PEB 0:0, read only 64 bytes, retry
>>>> [  163.757119] uncorrectable error : 
>>>> [  163.760487] UBI warning: ubi_io_read: error -74 (ECC error) while
>>>> reading 64 bytes from PEB 0:0, read only 64 bytes, retry
>>>> [  163.774427] uncorrectable error : 
>>>> [  163.777808] UBI error: ubi_io_read: error -74 (ECC error) while
>>>> reading 64 bytes from PEB 0:0, read 64 bytes 
>>>
>>> This looks like your flash does not really support sub-pages while
>>> declares that it does. Check it with MTD subpage test.
>> the subpages test fails with:
>>
>> [   82.435792] mtd_subpagetest: written up to eraseblock 7424
>> [   82.905775] mtd_subpagetest: written up to eraseblock 7680
>> [   83.375201] mtd_subpagetest: written up to eraseblock 7936
>> [   83.840447] mtd_subpagetest: written 8192 eraseblocks
>> [   83.845522] mtd_subpagetest: verifying all eraseblocks
>> [   83.852740] uncorrectable error : 
>> [   83.856086] mtd_subpagetest: error: read failed at 0x0
>> [   83.861469] mtd_subpagetest: error -74 occurred
>> [   83.866093] =================================================
>> Error: could not insert module ./drivers/mtd/tests/mtd_subpagetest.ko: Bad message
>>
>> I don't have a clue yet how to debug that further. Will start with
>> checking older kernel releases. Any hints welcome.
> 
> AFAIR, the driver did not support subpages and just needed to report
> about this. And there is a bug in MTD which makes it impossible for
> drivers to tell that sub-pages are not supported. So AFAIU, we need the
> patch I pointed you, and we also need the driver to use "no subpage
> writes" option.
> 
> I think the maintainer is the best person to comment on this. I assume
> this is the gpmi driver, so that would be Huang Shijie
> <shijie8@gmail.com>.

It's an mx35, so mxc_nand.c and it's Sascha Hauer, who's on vacation
this week.

Marc
-- 
Pengutronix e.K.                  | Marc Kleine-Budde           |
Industrial Linux Solutions        | Phone: +49-231-2826-924     |
Vertretung West/Dortmund          | Fax:   +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686  | http://www.pengutronix.de   |


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 262 bytes --]