Re: [PATCH net-next v2 3/3] genetlink: fit NLMSG_DONE into same read() as families

On Thu, Mar 21, 2024 at 02:56:41PM +0200, Gal Pressman wrote:

We've encountered a new issue recently which I believe is related to
this discussion.

Following Eric's patch:
9cc4cc329d30 ("ipv6: use xa_array iterator to implement inet6_dump_addr()")

Setting the interface mtu to < 1280 results in 'ip addr show eth2'
returning an error, because the ipv6 dump fails. This is a degradation
from the user's perspective.

# ip addr show eth2
4: eth2: <BROADCAST,MULTICAST> mtu 1500 qdisc noop state DOWN group
default qlen 1000
    link/ether 24:42:53:21:52:44 brd ff:ff:ff:ff:ff:ff
    altname enp6s0f0np0
# ip link set dev eth2 mtu 1000
# ip addr show eth2
RTNETLINK answers: No such device
Dump terminated

I don't think it's the same issue. Original issue was about user space
not knowing how to handle NLMSG_DONE being sent together with dump
responses. The issue you reported seems to be related to an
unintentional change in the return code when IPv6 is disabled on an
interface. Can you please test the following patch?

diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 247bd4d8ee45..92db9b474f2b 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c

@@ -5416,10 +5416,11 @@ static int inet6_dump_addr(struct sk_buff *skb, struct netlink_callback *cb,
 
                err = 0;
                if (fillargs.ifindex) {
-                       err = -ENODEV;
                        dev = dev_get_by_index_rcu(tgt_net, fillargs.ifindex);
-                       if (!dev)
+                       if (!dev) {
+                               err = -ENODEV;
                                goto done;
+                       }
                        idev = __in6_dev_get(dev);
                        if (idev)
                                err = in6_dump_addrs(idev, skb, cb,