| Message ID | 591b58e7623edc3eb66dd1fcfa8c8f133d090974.1649794741.git.petrm@nvidia.com (mailing list archive) |
|---|---|
| State | Accepted |
| Commit | 23cfe941b52e2fa645bdfd770087128a74c7dbee |
| Delegated to: | Netdev Maintainers |
| Headers | show |
| Series | [net] rtnetlink: Fix handling of disabled L3 stats in RTM_GETSTATS replies | expand |
On Tue, Apr 12, 2022 at 10:25:06PM +0200, Petr Machata wrote: > When L3 stats are disabled, rtnl_offload_xstats_get_size_stats() returns > size of 0, which is supposed to be an indication that the corresponding > attribute should not be emitted. However, instead, the current code > reserves a 0-byte attribute. > > The reason this does not show up as a citation on a kasan kernel is that > netdev_offload_xstats_get(), which is supposed to fill in the data, never > ends up getting called, because rtnl_offload_xstats_get_stats() notices > that the stats are not actually used and skips the call. > > Thus a zero-length IFLA_OFFLOAD_XSTATS_L3_STATS attribute ends up in a > response, confusing the userspace. > > Fix by skipping the L3-stats related block in rtnl_offload_xstats_fill(). > > Fixes: 0e7788fd7622 ("net: rtnetlink: Add UAPI for obtaining L3 offload xstats") > Signed-off-by: Petr Machata <petrm@nvidia.com> Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Hello: This patch was applied to netdev/net.git (master) by Paolo Abeni <pabeni@redhat.com>: On Tue, 12 Apr 2022 22:25:06 +0200 you wrote: > When L3 stats are disabled, rtnl_offload_xstats_get_size_stats() returns > size of 0, which is supposed to be an indication that the corresponding > attribute should not be emitted. However, instead, the current code > reserves a 0-byte attribute. > > The reason this does not show up as a citation on a kasan kernel is that > netdev_offload_xstats_get(), which is supposed to fill in the data, never > ends up getting called, because rtnl_offload_xstats_get_stats() notices > that the stats are not actually used and skips the call. > > [...] Here is the summary with links: - [net] rtnetlink: Fix handling of disabled L3 stats in RTM_GETSTATS replies https://git.kernel.org/netdev/net/c/23cfe941b52e You are awesome, thank you!
diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c index 159c9c61e6af..d1381ea6d52e 100644 --- a/net/core/rtnetlink.c +++ b/net/core/rtnetlink.c @@ -5242,6 +5242,8 @@ static int rtnl_offload_xstats_fill(struct sk_buff *skb, struct net_device *dev, *prividx = attr_id_l3_stats; size_l3 = rtnl_offload_xstats_get_size_stats(dev, t_l3); + if (!size_l3) + goto skip_l3_stats; attr = nla_reserve_64bit(skb, attr_id_l3_stats, size_l3, IFLA_OFFLOAD_XSTATS_UNSPEC); if (!attr) @@ -5253,6 +5255,7 @@ static int rtnl_offload_xstats_fill(struct sk_buff *skb, struct net_device *dev, return err; have_data = true; +skip_l3_stats: *prividx = 0; }
When L3 stats are disabled, rtnl_offload_xstats_get_size_stats() returns size of 0, which is supposed to be an indication that the corresponding attribute should not be emitted. However, instead, the current code reserves a 0-byte attribute. The reason this does not show up as a citation on a kasan kernel is that netdev_offload_xstats_get(), which is supposed to fill in the data, never ends up getting called, because rtnl_offload_xstats_get_stats() notices that the stats are not actually used and skips the call. Thus a zero-length IFLA_OFFLOAD_XSTATS_L3_STATS attribute ends up in a response, confusing the userspace. Fix by skipping the L3-stats related block in rtnl_offload_xstats_fill(). Fixes: 0e7788fd7622 ("net: rtnetlink: Add UAPI for obtaining L3 offload xstats") Signed-off-by: Petr Machata <petrm@nvidia.com> --- net/core/rtnetlink.c | 3 +++ 1 file changed, 3 insertions(+)