Re: [PATCH RFT] mmc: tmio: avoid concurrent runs of mmc_request_done()

[Dirk Behme <dirk.behme@de.bosch.com>]

Hi Wolfram,

On 28.02.2024 11:03, Wolfram Sang wrote:
> With the to-be-fixed commit, the reset_work handler cleared 'host->mrq'
> outside of the spinlock protected critical section. That leaves a small
> race window during execution of 'tmio_mmc_reset()' where the done_work
> handler could grab a pointer to the now invalid 'host->mrq'. Both would
> use it to call mmc_request_done() causing problems (see Link).
> 
> However, 'host->mrq' cannot simply be cleared earlier inside the
> critical section. That would allow new mrqs to come in asynchronously
> while the actual reset of the controller still needs to be done. So,
> like 'tmio_mmc_set_ios()', an ERR_PTR is used to prevent new mrqs from
> coming in but still avoiding concurrency between work handlers.
> 
> Reported-by: Dirk Behme <dirk.behme@de.bosch.com>
> Closes: https://lore.kernel.org/all/20240220061356.3001761-1-dirk.behme@de.bosch.com/
> Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
> Fixes: df3ef2d3c92c ("mmc: protect the tmio_mmc driver against a theoretical race")

Tested-by: Dirk Behme <dirk.behme@de.bosch.com>
Reviewed-by: Dirk Behme <dirk.behme@de.bosch.com>

> ---
> 
> Dirk: could you get this tested on your affected setups? I am somewhat
> optimistic that this is already enough. For sure, it is a needed first
> step.

Testing looks good :) Many thanks!

At least the issues we observed before are not seen any more. As we are 
not exactly sure on the root cause, of course this is not a 100% proof. 
But as the change looks good, looks like it won't break something and 
the system behaves good with it I would say we are good to go.

I think we could add anything like

Cc: stable@vger.kernel.org # 3.0+

?

>   drivers/mmc/host/tmio_mmc_core.c | 2 ++
>   1 file changed, 2 insertions(+)
> 
> diff --git a/drivers/mmc/host/tmio_mmc_core.c b/drivers/mmc/host/tmio_mmc_core.c
> index be7f18fd4836..c253d176db69 100644
> --- a/drivers/mmc/host/tmio_mmc_core.c
> +++ b/drivers/mmc/host/tmio_mmc_core.c
> @@ -259,6 +259,8 @@ static void tmio_mmc_reset_work(struct work_struct *work)
>   	else
>   		mrq->cmd->error = -ETIMEDOUT;
>   
> +	/* No new calls yet, but disallow concurrent tmio_mmc_done_work() */
> +	host->mrq = ERR_PTR(-EBUSY);
>   	host->cmd = NULL;
>   	host->data = NULL;
Thanks again!

Dirk