Re: Unsynchronized access to spi bus by mmc_rescan?

[Rich Felker <dalias@xxxxxxxx>]

On Wed, May 04, 2016 at 06:44:45PM -0400, Rich Felker wrote:
> Ping.
> 
> On Fri, Apr 22, 2016 at 05:28:42PM -0400, Rich Felker wrote:
> > I'm working on a driver for the J-core (http://j-core.org) SPI master,
> > which is currently limited to PIO and using the spi-bitbang framework
> > (probably not the right thing to use, but planned to change, and seems
> > orthogonal to the issue at hand). We're using it to access an SD card
> > via the mmc_spi mmc host driver, and experiencing crashes/corruption
> > that I tracked down to mmc_rescan (we don't yet have an interrupt for
> > media change) happening during SPI message transfers. Which locks are
> > supposed to preclude this from happening? Is it likely something wrong
> > our driver is using, or is there possibly a general bug in the MMC/SPI
> > subsystem(s)?

Ping. I'm also experiencing this issue with multiple devices on the
same spi bus; access seems completely unsynchronized. While it's hard
to rule out the possibility of it being specific to the spi master
driver, the driver just implements transfer_one and set_cs, so
synchronization would necessarily have to happen at a higher level.
Any tips on how I could further track down the cause?

Rich
--
To unsubscribe from this list: send the line "unsubscribe linux-spi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html