Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit 164891aa authored by Stephen Hemminger's avatar Stephen Hemminger Committed by David S. Miller
Browse files

[TCP]: Congestion control API update. 



Do some simple changes to make congestion control API faster/cleaner.
* use ktime_t rather than timeval
* merge rtt sampling into existing ack callback
  this means one indirect call versus two per ack.
* use flags bits to store options/settings

Signed-off-by: default avatarStephen Hemminger <shemminger@linux-foundation.org>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
parent 65d1b4a7

include/linux/skbuff.h

+5 −0
Original line number Diff line number Diff line
@@ -1569,6 +1569,11 @@ static inline void __net_timestamp(struct sk_buff *skb) 
	skb->tstamp = ktime_get_real();

}



static inline ktime_t net_timedelta(ktime_t t)

{

	return ktime_sub(ktime_get_real(), t);

}





extern __sum16 __skb_checksum_complete_head(struct sk_buff *skb, int len);

extern __sum16 __skb_checksum_complete(struct sk_buff *skb);

include/net/tcp.h

+5 −4
Original line number Diff line number Diff line
@@ -629,9 +629,12 @@ enum tcp_ca_event { 
#define TCP_CA_MAX	128

#define TCP_CA_BUF_MAX	(TCP_CA_NAME_MAX*TCP_CA_MAX)



#define TCP_CONG_NON_RESTRICTED 0x1

#define TCP_CONG_RTT_STAMP	0x2



struct tcp_congestion_ops {

	struct list_head	list;

	int	non_restricted;

	unsigned long flags;



	/* initialize private data (optional) */

	void (*init)(struct sock *sk);
@@ -645,8 +648,6 @@ struct tcp_congestion_ops { 
	/* do new cwnd calculation (required) */

	void (*cong_avoid)(struct sock *sk, u32 ack,

			   u32 rtt, u32 in_flight, int good_ack);

	/* round trip time sample per acked packet (optional) */

	void (*rtt_sample)(struct sock *sk, u32 usrtt);

	/* call before changing ca_state (optional) */

	void (*set_state)(struct sock *sk, u8 new_state);

	/* call when cwnd event occurs (optional) */
@@ -654,7 +655,7 @@ struct tcp_congestion_ops { 
	/* new value of cwnd after loss (optional) */

	u32  (*undo_cwnd)(struct sock *sk);

	/* hook for packet ack accounting (optional) */

	void (*pkts_acked)(struct sock *sk, u32 num_acked);

	void (*pkts_acked)(struct sock *sk, u32 num_acked, ktime_t last);

	/* get info for inet_diag (optional) */

	void (*get_info)(struct sock *sk, u32 ext, struct sk_buff *skb);

net/ipv4/tcp_bic.c

+1 −1
Original line number Diff line number Diff line
@@ -206,7 +206,7 @@ static void bictcp_state(struct sock *sk, u8 new_state) 
/* Track delayed acknowledgment ratio using sliding window

 * ratio = (15*ratio + sample) / 16

 */

static void bictcp_acked(struct sock *sk, u32 cnt)

static void bictcp_acked(struct sock *sk, u32 cnt, ktime_t last)

{

	const struct inet_connection_sock *icsk = inet_csk(sk);

net/ipv4/tcp_cong.c

+7 −7
Original line number Diff line number Diff line
@@ -126,7 +126,7 @@ int tcp_set_default_congestion_control(const char *name) 
#endif



	if (ca) {

		ca->non_restricted = 1;	/* default is always allowed */

		ca->flags |= TCP_CONG_NON_RESTRICTED;	/* default is always allowed */

		list_move(&ca->list, &tcp_cong_list);

		ret = 0;

	}
@@ -181,7 +181,7 @@ void tcp_get_allowed_congestion_control(char *buf, size_t maxlen) 
	*buf = '\0';

	rcu_read_lock();

	list_for_each_entry_rcu(ca, &tcp_cong_list, list) {

		if (!ca->non_restricted)

		if (!(ca->flags & TCP_CONG_NON_RESTRICTED))

			continue;

		offs += snprintf(buf + offs, maxlen - offs,

				 "%s%s",
@@ -212,16 +212,16 @@ int tcp_set_allowed_congestion_control(char *val) 
		}

	}



	/* pass 2 clear */

	/* pass 2 clear old values */

	list_for_each_entry_rcu(ca, &tcp_cong_list, list)

		ca->non_restricted = 0;

		ca->flags &= ~TCP_CONG_NON_RESTRICTED;



	/* pass 3 mark as allowed */

	while ((name = strsep(&val, " ")) && *name) {

		ca = tcp_ca_find(name);

		WARN_ON(!ca);

		if (ca)

			ca->non_restricted = 1;

			ca->flags |= TCP_CONG_NON_RESTRICTED;

	}

out:

	spin_unlock(&tcp_cong_list_lock);
@@ -256,7 +256,7 @@ int tcp_set_congestion_control(struct sock *sk, const char *name) 
	if (!ca)

		err = -ENOENT;



	else if (!(ca->non_restricted || capable(CAP_NET_ADMIN)))

	else if (!((ca->flags & TCP_CONG_NON_RESTRICTED) || capable(CAP_NET_ADMIN)))

		err = -EPERM;



	else if (!try_module_get(ca->owner))
@@ -371,8 +371,8 @@ u32 tcp_reno_min_cwnd(const struct sock *sk) 
EXPORT_SYMBOL_GPL(tcp_reno_min_cwnd);



struct tcp_congestion_ops tcp_reno = {

	.flags		= TCP_CONG_NON_RESTRICTED,

	.name		= "reno",

	.non_restricted = 1,

	.owner		= THIS_MODULE,

	.ssthresh	= tcp_reno_ssthresh,

	.cong_avoid	= tcp_reno_cong_avoid,

net/ipv4/tcp_cubic.c

+1 −1
Original line number Diff line number Diff line
@@ -334,7 +334,7 @@ static void bictcp_state(struct sock *sk, u8 new_state) 
/* Track delayed acknowledgment ratio using sliding window

 * ratio = (15*ratio + sample) / 16

 */

static void bictcp_acked(struct sock *sk, u32 cnt)

static void bictcp_acked(struct sock *sk, u32 cnt, ktime_t last)

{

	const struct inet_connection_sock *icsk = inet_csk(sk);