sdcard report

Sean McNeil sean at mcneil.com
Wed Jun 25 19:23:59 CEST 2008


Andy Green wrote:
> Somebody in the thread at some point said:
>
> |> | 128MB Sandisk sdcard fails.
> |> |
> |> | Latest andy kernel.
> |>
> |> What was the testing action and what did the failure look like?  Any
> |> suspends in there?
> |
> | testing action: trying to mount it during boot
>
> This is different to Sean's situation, it's FAT format here vs ext3 for
> him, and spewing fatal hard errors to communicate with the card, rather
> than logical errors in the filesystem.
>
Actually, this looks similar to problems I had at first. Can you try 
reducing the max clock rate?

drivers/mfd/glamo/glamo-mci.c: change divisor on  mmc->f_max assignment 
to 4 or 5.

See if that makes a difference.

> | failure:
> | Buffer I/O error on device mmcblk0p1, logical block 168
> | end_request: I/O error, dev mmcblk0, sector 266
> ...
> | glamo-mci glamo-mci.0: Error after cmd: 0x8020
> | glamo-mci glamo-mci.0: Error after cmd: 0x8122
> | mmcblk0: error -84 sending read/write command  <=== failed
> | end_request: I/O error, dev mmcblk0, sector 97
> | FAT: unable to read boot sector
> ...
> | No suspend/resume involved.
>
> OK.  Well I think it's a different problem somehow.  Maybe try sticking
> your thumb on it during this to see if it is a physical connectivity
> problem.
>
> Also I see it's a spew so maybe this isn't easy, but any log messages
> with this kind of format are also interesting
>
> [    2.330000] glamo-mci glamo-mci.0: powered (vdd = 20) clk: 16666kHz
> div=2 (req: 16666kHz). Bus width=0
>
> since they tell us the clock rate, voltage we selected and whether we
> are in 1 or 4 bit mode.  If you do suspend, you should get a set of them
> again in dmesg.
>
> -Andy
>





More information about the openmoko-kernel mailing list