Re: [PATCH 1/1] driver core: Fix driver_sysfs_remove() order in really_probe()

[Greg Kroah-Hartman]

On Wed, Dec 29, 2021 at 12:51:59PM +0800, Lu Baolu wrote:
> The driver_sysfs_remove() should always be called after successful
> driver_sysfs_add(). Otherwise, NULL pointers will be passed to the
> sysfs_remove_link(), where it is decoded as searching sysfs root.

What null pointer is being sent to sysfs_remove_link()?  For which link?

How are you triggering this failure path and how was it tested?

> 
> Fixes: 1901fb2604fbc ("Driver core: fix "driver" symlink timing")
> Cc: stable@xxxxxxxxxxxxxxx
> Signed-off-by: Lu Baolu <baolu.lu@xxxxxxxxxxxxxxx>
> ---
>  drivers/base/dd.c | 7 ++++---
>  1 file changed, 4 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/base/dd.c b/drivers/base/dd.c
> index 68ea1f949daa..9eaaff2f556c 100644
> --- a/drivers/base/dd.c
> +++ b/drivers/base/dd.c
> @@ -577,14 +577,14 @@ static int really_probe(struct device *dev, struct device_driver *drv)
>  	if (dev->bus->dma_configure) {
>  		ret = dev->bus->dma_configure(dev);
>  		if (ret)
> -			goto probe_failed;
> +			goto pinctrl_bind_failed;

Why not call the notifier chain here?  Did you verify that this change
still works properly?

thanks,

greg k-h