ipv4: Invalidate neighbour for broadcast address upon address addition
authorIdo Schimmel <idosch@nvidia.com>
Sat, 19 Feb 2022 15:45:19 +0000 (17:45 +0200)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 13 Apr 2022 18:59:05 +0000 (20:59 +0200)
commit049072749a5e1fe277ee64746bbe9fca5c55f3f9
treeee483653f64bd0652a1129ae3320d2c315fa456e
parente45d1d19a0bb29f6454e2fa87559325d62538f5f
ipv4: Invalidate neighbour for broadcast address upon address addition

[ Upstream commit 0c51e12e218f20b7d976158fdc18019627326f7a ]

In case user space sends a packet destined to a broadcast address when a
matching broadcast route is not configured, the kernel will create a
unicast neighbour entry that will never be resolved [1].

When the broadcast route is configured, the unicast neighbour entry will
not be invalidated and continue to linger, resulting in packets being
dropped.

Solve this by invalidating unresolved neighbour entries for broadcast
addresses after routes for these addresses are internally configured by
the kernel. This allows the kernel to create a broadcast neighbour entry
following the next route lookup.

Another possible solution that is more generic but also more complex is
to have the ARP code register a listener to the FIB notification chain
and invalidate matching neighbour entries upon the addition of broadcast
routes.

It is also possible to wave off the issue as a user space problem, but
it seems a bit excessive to expect user space to be that intimately
familiar with the inner workings of the FIB/neighbour kernel code.

[1] https://lore.kernel.org/netdev/55a04a8f-56f3-f73c-2aea-2195923f09d1@huawei.com/

Reported-by: Wang Hai <wanghai38@huawei.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Tested-by: Wang Hai <wanghai38@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Sasha Levin <sashal@kernel.org>
include/net/arp.h
net/ipv4/arp.c
net/ipv4/fib_frontend.c