Re: [BUG] mfd: rtsx_usb data corruption with SDCX microsd

[Daniel Lenski <dlenski@gmail.com>]

On 21 October 2015 at 12:02, Ulf Hansson <ulf.hansson <at> linaro.org> 
wrote:

> On 20 October 2015 at 18:29, Marcus Overhagen <marcus.overhagen <at> 
gmail.com> wrote:
> > I tested again with a 4.2 kernel but the bug is still present and
> > happens more often. So far nobody has responded.
> > I don't know what to do, and whether it's related to usb, mmc or mfd.
> > Please advise.
> 
> Sorry for the delay. I was hoping to get some input from Roger as me
> personally don't know much about this HW.
> I can't tell whether this is a regression or not. The rtsx_usb driver
> were added in Linux 3.16. Perhaps you can verify if this is a
> regression or not!?

I have been having a very similar problem, though in my case it *does* 
happen with a 32 GB microSDHC card.

[    9.854751] mmc0: new ultra high speed SDR50 SDHC card at address aaaa
[    9.862897] mmcblk0: mmc0:aaaa SU32G 29.7 GiB 
[  581.272899] mmcblk0: error -110 sending stop command, original cmd 
response 0x900, card status 0x900

At first I thought the flash memory on the card had become corrupted, but 
then realized that it still worked completely fine (a) when booting Wind8 
on the same system, (b) in an Android phone, and (c) in a Rockbox MP3 
player.

> > A 32 GB SDHC does not show data corruption with this card reader.

On my system, the error occurs _almost always_ when trying to access my 32 
GB microSDHC card, but never with 2 GB or 4 GB SD cards.

Running badblocks -n /dev/mmcblk0, it tells me that every single block on 
the card is bad.

Per your suggestion, I booted a 3.16 kernel and encountered the exact same 
errors.

> As I don't know the HW, I can just provide some guesses of how to narrow 
down the problem.
>
> If you re-build the kernel and make changes around which MMC_CAPS the 
host supports, you can try to narrow down the problem if it's related to 
speed modes.
>
> For example start by using *only* the following MMC_CAPS (updates to be 
made in rtsx_usb_init_host() - drivers/mmc/host/rtsx_usb.c):
> 
> MMC_CAP_4_BIT_DATA | MMC_CAP_SD_HIGHSPEED | MMC_CAP_NEEDS_POLL
> 
> ..and if no error, try add cap by cap to see what happens.
> 
> Sorry, not being able to help you more - but at least this is a start!

Am I right that this has been moved to rtsx_usb_sdmmc in the most recent 
kernels?
https://github.com/torvalds/linux/blob/HEAD/drivers/mmc/host/rtsx_usb_sdmmc
.c#L1335

Thanks,
Dan