On 12/30/18 11:15 AM, ba...@gandi.net wrote:
> From: Arthur Gautier <ba...@gandi.net>
>
> This commit fixes a regression in AF_INET/RTM_GETADDR and
> AF_INET6/RTM_GETADDR. The AF_UNSPEC/RTM_GETADDR was not affected by this as
> rtnl_dump_all would just hide the errno.
>
> Before this commit, the kernel would stop dumping addresses once the first
> skb was full and end the stream with NLMSG_DONE(-EMSGSIZE). The callback
> should be kept alive as the userspace is expected to call back with a new
> empty skb.
>
> Fixes: d7e38611b81e ("net/ipv4: Put target net when address dump fails due to
> bad attributes")
> Fixes: 242afaa6968c ("net/ipv6: Put target net when address dump fails due to
> bad attributes")
>
> Cc: David Ahern <dsah...@gmail.com>
> Cc: "David S . Miller" <da...@davemloft.net>
> Cc: netdev@vger.kernel.org
> Signed-off-by: Arthur Gautier <ba...@gandi.net>
> ---
> net/netlink/af_netlink.c | 9 +++++++++
> 1 file changed, 9 insertions(+)
>
> diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c
> index 3c023d6120f65..0a48bca246bc4 100644
> --- a/net/netlink/af_netlink.c
> +++ b/net/netlink/af_netlink.c
> @@ -2245,6 +2245,15 @@ static int netlink_dump(struct sock *sk)
> cb->extack = NULL;
> }
>
> + /* Should the dump method run out of space, we have to
> + * keep the dumper running until completion.
> + * Just ignore the error, userspace should call back with a
> + * new skb until dump is complete
> + */
> + if (nlk->dump_done_errno == -EMSGSIZE) {
> + nlk->dump_done_errno = skb->len;
> + }
> +
> if (nlk->dump_done_errno > 0 ||
> skb_tailroom(skb) < nlmsg_total_size(sizeof(nlk->dump_done_errno)))
> {
> mutex_unlock(nlk->cb_mutex);
>
This should be in inet_dump_ifaddr and inet6_dump_addr. Instead of:
return err < 0 ? err : skb->len;
it should be
return skb->len ? : err;
