net: dsa: mv88e6xxx: also read STU state in mv88e6250_g1_vtu_getnext
authorRasmus Villemoes <rasmus.villemoes@prevas.dk>
Sat, 16 Jan 2021 02:39:35 +0000 (03:39 +0100)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 27 Jan 2021 10:55:21 +0000 (11:55 +0100)
commit55c869b1324f49890d2c3898b9b473e71acfa893
treeb74650e65730ebdaef4922fb110a7eef469a28f0
parent2fc06bfa701d031e27ba24c6aec24f124fe13650
net: dsa: mv88e6xxx: also read STU state in mv88e6250_g1_vtu_getnext

commit 87fe04367d842c4d97a77303242d4dd4ac351e46 upstream.

mv88e6xxx_port_vlan_join checks whether the VTU already contains an
entry for the given vid (via mv88e6xxx_vtu_getnext), and if so, merely
changes the relevant .member[] element and loads the updated entry
into the VTU.

However, at least for the mv88e6250, the on-stack struct
mv88e6xxx_vtu_entry vlan never has its .state[] array explicitly
initialized, neither in mv88e6xxx_port_vlan_join() nor inside the
getnext implementation. So the new entry has random garbage for the
STU bits, breaking VLAN filtering.

When the VTU entry is initially created, those bits are all zero, and
we should make sure to keep them that way when the entry is updated.

Fixes: 92307069a96c (net: dsa: mv88e6xxx: Avoid VTU corruption on 6097)
Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Tobias Waldekranz <tobias@waldekranz.com>
Tested-by: Tobias Waldekranz <tobias@waldekranz.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
drivers/net/dsa/mv88e6xxx/global1_vtu.c