Re: [PATCH] remoteproc: core: check rproc->power value before decreasing it

[Mathieu Poirier <mathieu.poirier@linaro.org>]

Hi Shengjiu,

On Fri, Mar 18, 2022 at 05:51:36PM +0800, Shengjiu Wang wrote:
> There are possibility that two 'stop' operation happen
> in parallel, then the rproc->power may be decreased to
> -1, that this reference count will be in wrong state.
> So check rproc->power to make sure it is larger than
> zero before decreasing it.
> 
> Signed-off-by: Shengjiu Wang <shengjiu.wang@nxp.com>
> ---
>  drivers/remoteproc/remoteproc_core.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/drivers/remoteproc/remoteproc_core.c b/drivers/remoteproc/remoteproc_core.c
> index c510125769b9..84e065ad8743 100644
> --- a/drivers/remoteproc/remoteproc_core.c
> +++ b/drivers/remoteproc/remoteproc_core.c
> @@ -2075,6 +2075,9 @@ int rproc_shutdown(struct rproc *rproc)
>  		return ret;
>  	}
>  
> +	if (atomic_read(&rproc->power) <= 0)
> +		goto out;
> +

Although the proposed solution will likely work, I think it is best to use
rproc->state for this.  Simply do the same test as in state_store()[1] and exit
if the conditions are not satisfied.  Please do the same thing for
rproc_detach().

Thanks,
Mathieu

[1]. https://elixir.bootlin.com/linux/v5.17/source/drivers/remoteproc/remoteproc_sysfs.c#L205

>  	/* if the remote proc is still needed, bail out */
>  	if (!atomic_dec_and_test(&rproc->power))
>  		goto out;
> -- 
> 2.17.1
>