On Tue, Jan 17, 2023 at 12:57:19AM +0100, Piergiorgio Beruto wrote:
> Fix problem found by syzbot dereferencing a device pointer.
>
> Signed-off-by: Piergiorgio Beruto <piergiorgio.beruto@xxxxxxxxx>
> Reported-by: syzbot+8cf35743af243e5f417e@xxxxxxxxxxxxxxxxxxxxxxxxx
> Fixes: 8580e16c28f3 ("net/ethtool: add netlink interface for the PLCA RS")
> ---
> net/ethtool/plca.c | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/net/ethtool/plca.c b/net/ethtool/plca.c
> index be7404dc9ef2..bc3d31f99998 100644
> --- a/net/ethtool/plca.c
> +++ b/net/ethtool/plca.c
> @@ -155,6 +155,8 @@ int ethnl_set_plca_cfg(struct sk_buff *skb, struct genl_info *info)
> return ret;
>
> dev = req_info.dev;
> + if(!dev)
> + return -ENODEV;
Shouldn't be necessary. The fact that you pass "true" to the
"require_dev" argument of ethnl_parse_header_dev_get() takes care
specifically of that.
Looking at that syzbot report, it looks like you solved it with commit
28dbf774bc87 ("plca.c: fix obvious mistake in checking retval"). Or was
that not the only issue?
>
> rtnl_lock();
>
> --
> 2.37.4
>
