On Tue, 27 Jul 2021, Uwe Kleine-König wrote:

> The driver core only calls a remove callback when the device was
> successfully bound (aka probed) before. So dev->driver is never NULL.
> 

Are you sure dev->driver is non-NULL for the lifetime of the device?
A quick glance at device_reprobe() makes me wonder about that.

> Apart from that, the compiler might already assume dev->driver being
> non-NULL after to_nubus_driver(dev->driver) was called.
> 

I don't understand how a compiler can make that assumption. But then, I 
don't know why compilers do a lot of the things they do...

> Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
> ---
>  drivers/nubus/bus.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/nubus/bus.c b/drivers/nubus/bus.c
> index d9d04f27f89b..17fad660032c 100644
> --- a/drivers/nubus/bus.c
> +++ b/drivers/nubus/bus.c
> @@ -33,7 +33,7 @@ static void nubus_device_remove(struct device *dev)
>  {
>  	struct nubus_driver *ndrv = to_nubus_driver(dev->driver);
>  
> -	if (dev->driver && ndrv->remove)
> +	if (ndrv->remove)
>  		ndrv->remove(to_nubus_board(dev));
>  }
>  
>