| Message ID | 20241218115111.2407958-4-srasheed@marvell.com (mailing list archive) |
|---|---|
| State | Changes Requested |
| Delegated to: | Netdev Maintainers |
| Headers | show |
| Series | Fix race conditions in ndo_get_stats64 | expand |
On Wed, 18 Dec 2024 03:51:10 -0800 Shinas Rasheed wrote: > ndo_get_stats64() can race with ndo_stop(), which frees input and > output queue resources. Call synchornize_net() in ndo_stop() synchronize_net() here.. > + synchronize_rcu(); > netdev_info(netdev, "Stopping the device ...\n"); .. but not in the code
diff --git a/drivers/net/ethernet/marvell/octeon_ep_vf/octep_vf_main.c b/drivers/net/ethernet/marvell/octeon_ep_vf/octep_vf_main.c index 7e6771c9cdbb..e6253f85b623 100644 --- a/drivers/net/ethernet/marvell/octeon_ep_vf/octep_vf_main.c +++ b/drivers/net/ethernet/marvell/octeon_ep_vf/octep_vf_main.c @@ -523,6 +523,7 @@ static int octep_vf_stop(struct net_device *netdev) { struct octep_vf_device *oct = netdev_priv(netdev); + synchronize_rcu(); netdev_info(netdev, "Stopping the device ...\n"); /* Stop Tx from stack */
ndo_get_stats64() can race with ndo_stop(), which frees input and output queue resources. Call synchornize_net() in ndo_stop() to avoid such races. Fixes: c3fad23cdc06 ("octeon_ep_vf: add support for ndo ops") Signed-off-by: Shinas Rasheed <srasheed@marvell.com> --- V3: - No changes V2: https://lore.kernel.org/all/20241216075842.2394606-4-srasheed@marvell.com/ - Changed sync mechanism to fix race conditions from using an atomic set_bit ops to a much simpler synchronize_net() V1: https://lore.kernel.org/all/20241203072130.2316913-4-srasheed@marvell.com/ drivers/net/ethernet/marvell/octeon_ep_vf/octep_vf_main.c | 1 + 1 file changed, 1 insertion(+)