Re: Unsynchronized access to spi bus by mmc_rescan?

[Mark Brown <broonie-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>]

[-- Attachment #1: Type: text/plain, Size: 1050 bytes --]

On Thu, Jul 21, 2016 at 11:43:08PM +0100, Mark Brown wrote:

> This underlying issue is that we are trying to use one mutex for two
> purposes, the existing mutex is mainly being used to serialize access to
> the physical bus but the externally visible bus lock is there to ensure
> that only one caller is able to queue new transfers which is not the
> same thing at all.  It is completely irrelevant when we are pushing
> messages to the bus if we are doing so with the bus lock held.

Just sent a test patch for this, completely untested given the hour but
hopefully you can give it a spin if this is easy to reproduce for you.

Rich, I also just noticed that you didn't CC any people on any of your
postings prior to today.  As I know we've discussed before you should
*always* directly CC at least maintainers, you can not assume that
anyone is going to see anything that is only on the lists.  This was the
same reason why your top posted content free pings didn't get anywhere,
most likely none of the maintainers had seen any of your emails.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 473 bytes --]