Re: [PATCH] of: overlay: fix memory leak related to duplicated property

[Frank Rowand <frowand.list-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>]

Hi Rob,

On 10/12/17 20:07, Lixin Wang wrote:
> From: alawang <alan.1.wang-FTnzSCDXTkYjo0HpFSRKWA@public.gmane.org>
> 
> Hello,
> 
> Sorry It was my fault in last email that wrote the wrong subject and sign off name.
> Correct them this time.
> Thanks
> 
> Function of_changeset_add_property or of_changeset_update_property may
> fails. In this case the property just allocated is never deallocated.
> 
> Signed-off-by: Lixin Wang <alan.1.wang-FTnzSCDXTkYjo0HpFSRKWA@public.gmane.org>
> ---
>  drivers/of/overlay.c | 15 +++++++++++----
>  1 file changed, 11 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/of/overlay.c b/drivers/of/overlay.c
> index 8ecfee3..af3b9a1 100644
> --- a/drivers/of/overlay.c
> +++ b/drivers/of/overlay.c
> @@ -162,6 +162,7 @@ static int of_overlay_apply_single_property(struct of_overlay *ov,
>  		bool is_symbols_node)
>  {
>  	struct property *propn = NULL, *tprop;
> +	int ret = 0;
>  
>  	/* NOTE: Multiple changes of single properties not supported */
>  	tprop = of_find_property(target, prop->name, NULL);
> @@ -186,10 +187,16 @@ static int of_overlay_apply_single_property(struct of_overlay *ov,
>  
>  	/* not found? add */
>  	if (tprop == NULL)
> -		return of_changeset_add_property(&ov->cset, target, propn);
> -
> -	/* found? update */
> -	return of_changeset_update_property(&ov->cset, target, propn);
> +		ret = of_changeset_add_property(&ov->cset, target, propn);
> +	else /* found? update */
> +		ret = of_changeset_update_property(&ov->cset, target, propn);
> +
> +	if (ret) {
> +		kfree(propn->name);
> +		kfree(propn->value);
> +		kfree(propn);
> +	}
> +	return ret;
>  }
>  
>  static int of_overlay_apply_single_device_node(struct of_overlay *ov,
> 

Just a heads up.

This will conflict with my patch series "[PATCH 00/12] of: overlay: clean up
device tree overlay code" [1].  The issue that Lixin has identified will
still remain after applying my patch series, and can be fixed in the same
manner as his patch, just different context, including variable names.

[1] https://lkml.org/lkml/2017/10/2/679

-Frank
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html