batman-adv: fix aggregation timing bug
PATCH 1/2 batman-adv: fix aggregation timing bug
From: Marek Lindner <[email protected]>
batman-adv aggregates routing packets to reduce the number of packets in the air. Every outgoing packet is compared with other packets in the buffer to determine whether it can be aggregated or not. Packets sent at a lower interval can be held back longer to maximize the aggregation. Due to insufficient checking batman-adv held back all packets for a certain time depending on its own lowest interval rate which slowed down all other nodes.
--- aggregation.c | 12 ++++++++++++ send.c | 10 +--------- 2 files changed, 13 insertions(+), 9 deletions(-)
batman-adv-kernelland/aggregation.c +12 -0
@@ -52,6 +52,8 @@ static bool can_aggregate_with(struct batman_packet *new_batman_packet,*/ if (time_before (send_time, forw_packet->send_time) && +time_after_eq (send_time +msecs_to_jiffies (MAX_AGGREGATION_MS ), + forw_packet->send_time) && (aggregated_bytes <=MAX_AGGREGATION_BYTES )) {/**
@@ -196,6 +198,16 @@ void add_bat_packet_to_list(unsigned char *packet_buff, int packet_len,if (forw_packet_aggr ==NULL ) {/* the following section can run without the lock */ spin_unlock_irqrestore (&forw_bat_list_lock, flags); + +/** +* if we could not aggregate this packet with one of the others +* we hold it back for a while, so that it might be aggregated +* later on +*/ +if ((!own_packet) && + (atomic_read (&bat_priv->aggregation_enabled))) + send_time +=msecs_to_jiffies (MAX_AGGREGATION_MS ); +new_aggregated_packet (packet_buff, packet_len, send_time, direct_link, if_incoming, own_packet);
batman-adv-kernelland/send.c +1 -9
@@ -49,15 +49,7 @@ static unsigned long own_send_time(void)/* when do we schedule a forwarded packet to be sent */ static unsigned long forward_send_time (struct bat_priv *bat_priv) { -unsigned long send_time = jiffies;/* Starting now plus... */ - -if (atomic_read (&bat_priv->aggregation_enabled)) - send_time += (((MAX_AGGREGATION_MS - (JITTER /2 ) + - (random32 () %JITTER )) *HZ ) /1000 ); -else - send_time += (((random32 () % (JITTER /2 )) *HZ ) /1000 ); - -return send_time; +return jiffies + (((random32 () % (JITTER /2 )) *HZ ) /1000 ); }/* send out an already prepared packet to the given address via the
Signature · 1 line
1.7.0
PATCH 2/2 batman-adv: Fix aggregation direct-link bug
So far, neighbour's secondary interface OGMs can involuntarily piggyback on primary interface OGMs that arrived on the same secondary interface before. Secondary interface OGMs should NEVER leave their direct neighbour broadcast domain! This patch ensures that secondary interface OGMs can only be aggregated to other secondary interface OGMs.
--- aggregation.c | 10 ++++++++-- 1 files changed, 8 insertions(+), 2 deletions(-)
batman-adv-kernelland/aggregation.c +8 -2
@@ -81,9 +81,15 @@ static bool can_aggregate_with(struct batman_packet *new_batman_packet,* interface only - we still can aggregate */ if ((directlink) && (new_batman_packet->ttl ==1 ) && - (forw_packet->if_incoming == if_incoming)) + (forw_packet->if_incoming == if_incoming) && + +/* packets from direct neighbors or +* own secondary interface packets +* (= secondary interface packets in general) */ + (batman_packet->flags &DIRECTLINK || + (forw_packet->own && + forw_packet->if_incoming->if_num !=0 )))return true; - }return false;
Signature · 1 line
1.7.0
Re: [B.A.T.M.A.N.] [PATCH 2/2] batman-adv: Fix aggregation direct-link bug
And there are two more things I'm wondering about:
- Why should primary interface OGMs from a different originator which already travelled over 49 hops not be aggregated to other primary interface OGMs?
- Why should secondary interface OGMs from neighbours not be aggregated on our own, neighboring secondary interface OGMs?
As far as I can tell, those two combinations are not being aggregated at the moment. Is there a reason why they shouldn't?
Cheers, Linus
Re: [B.A.T.M.A.N.] [PATCH 2/2] batman-adv: Fix aggregation direct-link bug
Hi,
thanks for your patch and for testing mine. I applied both of them.
Quoted 8 lines
And there are two more things I'm wondering about:
- Why should primary interface OGMs from a different originator which already travelled over 49 hops not be aggregated to other primary interface OGMs?
- Why should secondary interface OGMs from neighbours not be aggregated on our own, neighboring secondary interface OGMs?
As far as I can tell, those two combinations are not being aggregated at the moment. Is there a reason why they shouldn't?
Well, well - we happily apply more of your patches. ;-)
Cheers,
Marek