openvswitch: meter: fix race when getting now_ms.

We have observed meters working unexpected if traffic is 3+Gbit/s
with multiple connections.

now_ms is not pretected by meter->lock, we may get a negative
long_delta_ms when another cpu updated meter->used, then:
    delta_ms = (u32)long_delta_ms;
which will be a large value.

    band->bucket += delta_ms * band->rate;
then we get a wrong band->bucket.

OpenVswitch userspace datapath has fixed the same issue[1] some
time ago, and we port the implementation to kernel datapath.

[1] https://patchwork.ozlabs.org/project/openvswitch/patch/20191025114436.9746-1-i.maximets@ovn.org/

Fixes: 96fbc13d7e77 ("openvswitch: Add meter infrastructure")
Signed-off-by: Tao Liu <thomas.liu@ucloud.cn>
Suggested-by: Ilya Maximets <i.maximets@ovn.org>
Reviewed-by: Ilya Maximets <i.maximets@ovn.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

diff --git a/net/openvswitch/meter.c b/net/openvswitch/meter.c
index 96b524ceabca4f53b07411ae4d5d5c49e7d24656..896b8f5bc885350980a9547fc921583515dff7ec 100644 (file)
--- a/net/openvswitch/meter.c
+++ b/net/openvswitch/meter.c
@@ -611,6 +611,14 @@ bool ovs_meter_execute(struct datapath *dp, struct sk_buff *skb,
        spin_lock(&meter->lock);
 
        long_delta_ms = (now_ms - meter->used); /* ms */
+       if (long_delta_ms < 0) {
+               /* This condition means that we have several threads fighting
+                * for a meter lock, and the one who received the packets a
+                * bit later wins. Assuming that all racing threads received
+                * packets at the same time to avoid overflow.
+                */
+               long_delta_ms = 0;
+       }
 
        /* Make sure delta_ms will not be too large, so that bucket will not
         * wrap around below.