On 05/06, Stanislav Fomichev wrote:
On 05/06, Cosmin Ratiu wrote:
__netdev_update_features() expects the netdevice to be ops-locked, but it gets called recursively on the lower level netdevices to sync their features, and nothing locks those.
This commit fixes that, with the assumption that it shouldn't be possible for both higher-level and lover-level netdevices to require the instance lock, because that would lead to lock dependency warnings.
Without this, playing with higher level (e.g. vxlan) netdevices on top of netdevices with instance locking enabled can run into issues:
Mentioning vxlan is a bit confusing here; it shouldn't let you flip lro (I think). Which upper are you testing against?
Trying to understand if we can cover this case in the selftests. netdevsim also doesn't expose F_LRO feature... (yet?)
Ok, yeah, I can see it with the teaming on top of netdevsim and the following patch applied. Might try to send a testcase to cover this and the promisc part that I missed.
Acked-by: Stanislav Fomichev sdf@fomichev.me
diff --git a/drivers/net/netdevsim/netdev.c b/drivers/net/netdevsim/netdev.c index 2aa999345fe1..af545d42961c 100644 --- a/drivers/net/netdevsim/netdev.c +++ b/drivers/net/netdevsim/netdev.c @@ -881,11 +881,13 @@ static void nsim_setup(struct net_device *dev) NETIF_F_SG | NETIF_F_FRAGLIST | NETIF_F_HW_CSUM | + NETIF_F_LRO | NETIF_F_TSO; dev->hw_features |= NETIF_F_HW_TC | NETIF_F_SG | NETIF_F_FRAGLIST | NETIF_F_HW_CSUM | + NETIF_F_LRO | NETIF_F_TSO; dev->max_mtu = ETH_MAX_MTU; dev->xdp_features = NETDEV_XDP_ACT_HW_OFFLOAD;