net: ensure correct skb->tstamp in various fragmenters
Subsystems (5)
PATCH 1/1 net: ensure correct skb->tstamp in various fragmenters
Thomas found that some forwarded packets would be stuck in FQ packet scheduler because their skb->tstamp contained timestamps far in the future.
We thought we addressed this point in commit 8203e2d844d3 ("net: clear skb->tstamp in forwarding paths") but there is still an issue when/if a packet needs to be fragmented.
In order to meet EDT requirements, we have to make sure all fragments get the original skb->tstamp.
Note that this original skb->tstamp should be zero in forwarding path, but might have a non zero value in output path if user decided so.
--- net/bridge/netfilter/nf_conntrack_bridge.c | 3 +++ net/ipv4/ip_output.c | 3 +++ net/ipv6/ip6_output.c | 3 +++ net/ipv6/netfilter.c | 3 +++ 4 files changed, 12 insertions(+)
net/bridge/netfilter/nf_conntrack_bridge.c +3 -0
@@ -33,6 +33,7 @@ static int nf_br_ip_fragment(struct net *net, struct sock *sk, {int frag_max_size =BR_INPUT_SKB_CB (skb)->frag_max_size;unsigned int hlen, ll_rs, mtu; +ktime_t tstamp = skb->tstamp;struct ip_frag_state state;struct iphdr *iph;int err;
@@ -80,6 +81,7 @@ static int nf_br_ip_fragment(struct net *net, struct sock *sk,if (iter.frag)ip_fraglist_prepare (skb, &iter); + skb->tstamp = tstamp; err =output (net, sk, data, skb);if (err || !iter.frag)break ;
@@ -104,6 +106,7 @@ static int nf_br_ip_fragment(struct net *net, struct sock *sk, goto blackhole; } + skb2->tstamp = tstamp; err =output (net, sk, data, skb2);if (err) goto blackhole;
net/ipv4/ip_output.c +3 -0
@@ -771,6 +771,7 @@ int ip_do_fragment(struct net *net, struct sock *sk, struct sk_buff *skb,struct rtable *rt =skb_rtable (skb);unsigned int mtu, hlen, ll_rs;struct ip_fraglist_iter iter; +ktime_t tstamp = skb->tstamp;struct ip_frag_state state;int err =0 ;
@@ -846,6 +847,7 @@ int ip_do_fragment(struct net *net, struct sock *sk, struct sk_buff *skb,ip_fraglist_prepare (skb, &iter); } + skb->tstamp = tstamp; err =output (net, sk, skb);if (!err)
@@ -900,6 +902,7 @@ int ip_do_fragment(struct net *net, struct sock *sk, struct sk_buff *skb,/* * Put this fragment into the sending queue. */ + skb2->tstamp = tstamp; err =output (net, sk, skb2);if (err) goto fail;
net/ipv6/ip6_output.c +3 -0
@@ -768,6 +768,7 @@ int ip6_fragment(struct net *net, struct sock *sk, struct sk_buff *skb, inet6_sk(skb->sk) : NULL;struct ip6_frag_state state;unsigned int mtu, hlen, nexthdr_offset; +ktime_t tstamp = skb->tstamp;int hroom, err =0 ;__be32 frag_id;u8 *prevhdr, nexthdr =0 ;
@@ -855,6 +856,7 @@ int ip6_fragment(struct net *net, struct sock *sk, struct sk_buff *skb,if (iter.frag)ip6_fraglist_prepare (skb, &iter); + skb->tstamp = tstamp; err =output (net, sk, skb);if (!err) IP6_INC_STATS(net, ip6_dst_idev(&rt->dst),
@@ -913,6 +915,7 @@ int ip6_fragment(struct net *net, struct sock *sk, struct sk_buff *skb,/* * Put this fragment into the sending queue. */ + frag->tstamp = tstamp; err =output (net, sk, frag);if (err) goto fail;
net/ipv6/netfilter.c +3 -0
@@ -119,6 +119,7 @@ int br_ip6_fragment(struct net *net, struct sock *sk, struct sk_buff *skb, struct sk_buff *)) {int frag_max_size =BR_INPUT_SKB_CB (skb)->frag_max_size; +ktime_t tstamp = skb->tstamp;struct ip6_frag_state state;u8 *prevhdr, nexthdr =0 ;unsigned int mtu, hlen;
@@ -183,6 +184,7 @@ int br_ip6_fragment(struct net *net, struct sock *sk, struct sk_buff *skb,if (iter.frag)ip6_fraglist_prepare (skb, &iter); + skb->tstamp = tstamp; err =output (net, sk, data, skb);if (err || !iter.frag)break ;
@@ -215,6 +217,7 @@ int br_ip6_fragment(struct net *net, struct sock *sk, struct sk_buff *skb, goto blackhole; } + skb2->tstamp = tstamp; err =output (net, sk, data, skb2);if (err) goto blackhole;
Signature · 1 line
2.23.0.700.g56cf767bdb-goog
David Miller
maintainerFrom: Eric Dumazet <[email protected]>
Date: Wed, 16 Oct 2019 18:00:56 -0700
Quoted 11 lines
Thomas found that some forwarded packets would be stuck in FQ packet scheduler because their
skb->tstampcontained timestamps far in the future.We thought we addressed this point in commit
8203e2d844d3("net: clearskb->tstampin forwarding paths") but there is still an issue when/if a packet needs to be fragmented.In order to meet EDT requirements, we have to make sure all fragments get the original
skb->tstamp.
Quoted 7 lines
Note that this original
skb->tstampshould be zero in forwarding path, but might have a non zero value in output path if user decided so.Fixes fb420d5d91c1 ("tcp/fq: move back to CLOCK_MONOTONIC")Signed-off-by Eric Dumazet <[email protected]>Reported-by Thomas Bartschies <[email protected]>
Applied and queued up for -stable, thanks Eric.