Re: [PATCH net-next] netem: add cell concept to simulate special MAC behavior

[Eric Dumazet <eric.dumazet@gmail.com>]

Le samedi 10 décembre 2011 à 16:15 +0100, Hagen Paul Pfeifer a écrit :
> So I hope Eric's div/mod concerns are addressed with this version!? ;-)

Yes, finaly ;)

But please read two other comments.

> No "hot-path" div/mod operation anymore. tc_netem_rate was introduced in
> current net-next tree so there are no API f*ckups.
> 
> 
> From: Hagen Paul Pfeifer <hagen@jauu.net>
> 
> This extension can be used to simulate special link layer
> characteristics. Simulate because packet data is not modified, only the
> calculation base is changed to delay a packet based on the original
> packet size and artificial cell information.
> 
> packet_overhead can be used to simulate a link layer header compression
> scheme (e.g. set packet_overhead to -20) or with a positive
> packet_overhead value an additional MAC header can be simulated. It is
> also possible to "replace" the 14 byte Ethernet header with something
> else.
> 
> cell_size and cell_overhead can be used to simulate link layer schemes,
> based on cells, like some TDMA schemes. Another application area are MAC
> schemes using a link layer fragmentation with a (small) header each.
> Cell size is the maximum amount of data bytes within one cell. Cell
> overhead is an additional variable to change the per-cell-overhead
> (e.g.  5 byte header per fragment).
> 
> Example (5 kbit/s, 20 byte per packet overhead, cell-size 100 byte, per
> cell overhead 5 byte):
> 
>   tc qdisc add dev eth0 root netem rate 5kbit 20 100 5
> 
> Signed-off-by: Hagen Paul Pfeifer <hagen@jauu.net>
> Signed-off-by: Florian Westphal <fw@strlen.de>
> ---
>  include/linux/pkt_sched.h |    3 +++
>  net/sched/sch_netem.c     |   33 +++++++++++++++++++++++++++++----
>  2 files changed, 32 insertions(+), 4 deletions(-)
> 
> diff --git a/include/linux/pkt_sched.h b/include/linux/pkt_sched.h
> index 8786ea7..8daced3 100644
> --- a/include/linux/pkt_sched.h
> +++ b/include/linux/pkt_sched.h
> @@ -502,6 +502,9 @@ struct tc_netem_corrupt {
>  
>  struct tc_netem_rate {
>  	__u32	rate;	/* byte/s */
> +	__s32	packet_overhead;
> +	__u32	cell_size;
> +	__s32	cell_overhead;

Adding fields in existing structure is convenient, but this means a
previous tc binary wont work anymore.

(because of this line in netem_policy[] :)
[TCA_NETEM_RATE]    = { .len = sizeof(struct tc_netem_rate) },

I guess netem users are capable to fetch new iproute2 package, but its
worth to mention this in changelog.

>  };
>  
>  enum {
> diff --git a/net/sched/sch_netem.c b/net/sched/sch_netem.c
> index 3bfd733..31ab06e 100644
> --- a/net/sched/sch_netem.c
> +++ b/net/sched/sch_netem.c
> @@ -22,6 +22,7 @@
>  #include <linux/skbuff.h>
>  #include <linux/vmalloc.h>
>  #include <linux/rtnetlink.h>
> +#include <linux/reciprocal_div.h>
>  
>  #include <net/netlink.h>
>  #include <net/pkt_sched.h>
> @@ -80,6 +81,10 @@ struct netem_sched_data {
>  	u32 reorder;
>  	u32 corrupt;
>  	u32 rate;
> +	s32 packet_overhead;
> +	u32 cell_size;
> +	u32 cell_size_reciprocal;
> +	s32 cell_overhead;
>  
>  	struct crndstate {
>  		u32 last;
> @@ -299,11 +304,23 @@ static psched_tdiff_t tabledist(psched_tdiff_t mu, psched_tdiff_t sigma,
>  	return  x / NETEM_DIST_SCALE + (sigma / NETEM_DIST_SCALE) * t + mu;
>  }
>  
> -static psched_time_t packet_len_2_sched_time(unsigned int len, u32 rate)
> +static psched_time_t packet_len_2_sched_time(unsigned int len, struct netem_sched_data *q)
>  {
> -	u64 ticks = (u64)len * NSEC_PER_SEC;
> +	u64 ticks;
>  
> -	do_div(ticks, rate);
> +	len += q->packet_overhead;
> +
> +	if (q->cell_size) {
> +		u32 cells = reciprocal_divide(len, q->cell_size_reciprocal);
> +
> +		if (len > cells * q->cell_size)	/* extra cell needed for remainder */
> +			cells++;
> +		len = cells * (q->cell_size + q->cell_overhead);
> +	}
> +
> +	ticks = (u64)len * NSEC_PER_SEC;
> +
> +	do_div(ticks, q->rate);
>  	return PSCHED_NS2TICKS(ticks);
>  }
>  
> @@ -384,7 +401,7 @@ static int netem_enqueue(struct sk_buff *skb, struct Qdisc *sch)
>  		if (q->rate) {
>  			struct sk_buff_head *list = &q->qdisc->q;
>  
> -			delay += packet_len_2_sched_time(skb->len, q->rate);
> +			delay += packet_len_2_sched_time(skb->len, q);
>  
>  			if (!skb_queue_empty(list)) {
>  				/*
> @@ -568,6 +585,11 @@ static void get_rate(struct Qdisc *sch, const struct nlattr *attr)
>  	const struct tc_netem_rate *r = nla_data(attr);
>  
>  	q->rate = r->rate;
> +	q->packet_overhead = r->packet_overhead;
> +	q->cell_size = r->cell_size;
> +	q->cell_size_reciprocal = reciprocal_value(q->cell_size);

Here you get a crash if cell_size is 0. (divide by 0)


> +	q->cell_size = q->cell_size;
> +	q->cell_overhead = r->cell_overhead;
>  }
>  
>  static int get_loss_clg(struct Qdisc *sch, const struct nlattr *attr)
> @@ -909,6 +931,9 @@ static int netem_dump(struct Qdisc *sch, struct sk_buff *skb)
>  	NLA_PUT(skb, TCA_NETEM_CORRUPT, sizeof(corrupt), &corrupt);
>  
>  	rate.rate = q->rate;
> +	rate.packet_overhead = q->packet_overhead;
> +	rate.cell_size = q->cell_size;
> +	rate.cell_overhead = q->cell_overhead;
>  	NLA_PUT(skb, TCA_NETEM_RATE, sizeof(rate), &rate);
>  
>  	if (dump_loss_model(q, skb) != 0)