Commit 871da0a7 authored by David S. Miller's avatar David S. Miller
Browse files

Merge branch 'tcp-more-perns-sysctls' 



Eric Dumazet says:

====================
tcp: move 12 sysctls to namespaces

Ideally all TCP sysctls should be per netns.
This patch series takes care of 12 sysctls.

Remains the ones that need discussion :

sysctl_tcp_mem, sysctl_tcp_rmem, sysctl_tcp_wmem, and sysctl_tcp_max_orphans
====================

Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
parents 267146d4 c26e91f8

include/net/netns/ipv4.h

+12 −0
Original line number Diff line number Diff line
@@ -142,6 +142,18 @@ struct netns_ipv4 { 
	int sysctl_tcp_app_win;

	int sysctl_tcp_adv_win_scale;

	int sysctl_tcp_frto;

	int sysctl_tcp_nometrics_save;

	int sysctl_tcp_moderate_rcvbuf;

	int sysctl_tcp_tso_win_divisor;

	int sysctl_tcp_workaround_signed_windows;

	int sysctl_tcp_limit_output_bytes;

	int sysctl_tcp_challenge_ack_limit;

	int sysctl_tcp_min_tso_segs;

	int sysctl_tcp_min_rtt_wlen;

	int sysctl_tcp_autocorking;

	int sysctl_tcp_invalid_ratelimit;

	int sysctl_tcp_pacing_ss_ratio;

	int sysctl_tcp_pacing_ca_ratio;

	struct inet_timewait_death_row tcp_death_row;

	int sysctl_max_syn_backlog;

	int sysctl_tcp_fastopen;

include/net/tcp.h

+2 −14
Original line number Diff line number Diff line
@@ -247,22 +247,9 @@ extern int sysctl_tcp_max_orphans; 
extern long sysctl_tcp_mem[3];

extern int sysctl_tcp_wmem[3];

extern int sysctl_tcp_rmem[3];

extern int sysctl_tcp_nometrics_save;

extern int sysctl_tcp_moderate_rcvbuf;

extern int sysctl_tcp_tso_win_divisor;

extern int sysctl_tcp_workaround_signed_windows;



#define TCP_RACK_LOSS_DETECTION  0x1 /* Use RACK to detect losses */



extern int sysctl_tcp_limit_output_bytes;

extern int sysctl_tcp_challenge_ack_limit;

extern int sysctl_tcp_min_tso_segs;

extern int sysctl_tcp_min_rtt_wlen;

extern int sysctl_tcp_autocorking;

extern int sysctl_tcp_invalid_ratelimit;

extern int sysctl_tcp_pacing_ss_ratio;

extern int sysctl_tcp_pacing_ca_ratio;



extern atomic_long_t tcp_memory_allocated;

extern struct percpu_counter tcp_sockets_allocated;

extern unsigned long tcp_memory_pressure;
@@ -1305,7 +1292,8 @@ static inline void tcp_slow_start_after_idle_check(struct sock *sk) 
}



/* Determine a window scaling and initial window to offer. */

void tcp_select_initial_window(int __space, __u32 mss, __u32 *rcv_wnd,

void tcp_select_initial_window(const struct sock *sk, int __space,

			       __u32 mss, __u32 *rcv_wnd,

			       __u32 *window_clamp, int wscale_ok,

			       __u8 *rcv_wscale, __u32 init_rcv_wnd);

net/ipv4/syncookies.c

+1 −1
Original line number Diff line number Diff line
@@ -385,7 +385,7 @@ struct sock *cookie_v4_check(struct sock *sk, struct sk_buff *skb) 
	/* Try to redo what tcp_v4_send_synack did. */

	req->rsk_window_clamp = tp->window_clamp ? :dst_metric(&rt->dst, RTAX_WINDOW);



	tcp_select_initial_window(tcp_full_space(sk), req->mss,

	tcp_select_initial_window(sk, tcp_full_space(sk), req->mss,

				  &req->rsk_rcv_wnd, &req->rsk_window_clamp,

				  ireq->wscale_ok, &rcv_wscale,

				  dst_metric(&rt->dst, RTAX_INITRWND));

net/ipv4/sysctl_net_ipv4.c

+92 −92
Original line number Diff line number Diff line
@@ -437,13 +437,6 @@ static struct ctl_table ipv4_table[] = { 
		.proc_handler	= proc_dointvec_minmax,

		.extra1		= &one,

	},

	{

		.procname	= "tcp_min_rtt_wlen",

		.data		= &sysctl_tcp_min_rtt_wlen,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec

	},

	{

		.procname	= "tcp_low_latency",

		.data		= &sysctl_tcp_low_latency,
@@ -451,54 +444,12 @@ static struct ctl_table ipv4_table[] = { 
		.mode		= 0644,

		.proc_handler	= proc_dointvec

	},

	{

		.procname	= "tcp_no_metrics_save",

		.data		= &sysctl_tcp_nometrics_save,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec,

	},

	{

		.procname	= "tcp_moderate_rcvbuf",

		.data		= &sysctl_tcp_moderate_rcvbuf,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec,

	},

	{

		.procname	= "tcp_tso_win_divisor",

		.data		= &sysctl_tcp_tso_win_divisor,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec,

	},

	{

		.procname	= "tcp_congestion_control",

		.mode		= 0644,

		.maxlen		= TCP_CA_NAME_MAX,

		.proc_handler	= proc_tcp_congestion_control,

	},

	{

		.procname	= "tcp_workaround_signed_windows",

		.data		= &sysctl_tcp_workaround_signed_windows,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec

	},

	{

		.procname	= "tcp_limit_output_bytes",

		.data		= &sysctl_tcp_limit_output_bytes,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec

	},

	{

		.procname	= "tcp_challenge_ack_limit",

		.data		= &sysctl_tcp_challenge_ack_limit,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec

	},

#ifdef CONFIG_NETLABEL

	{

		.procname	= "cipso_cache_enable",
@@ -541,49 +492,6 @@ static struct ctl_table ipv4_table[] = { 
		.mode		= 0644,

		.proc_handler   = proc_allowed_congestion_control,

	},

	{

		.procname	= "tcp_min_tso_segs",

		.data		= &sysctl_tcp_min_tso_segs,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec_minmax,

		.extra1		= &one,

		.extra2		= &gso_max_segs,

	},

	{

		.procname	= "tcp_pacing_ss_ratio",

		.data		= &sysctl_tcp_pacing_ss_ratio,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec_minmax,

		.extra1		= &zero,

		.extra2		= &thousand,

	},

	{

		.procname	= "tcp_pacing_ca_ratio",

		.data		= &sysctl_tcp_pacing_ca_ratio,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec_minmax,

		.extra1		= &zero,

		.extra2		= &thousand,

	},

	{

		.procname	= "tcp_autocorking",

		.data		= &sysctl_tcp_autocorking,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec_minmax,

		.extra1		= &zero,

		.extra2		= &one,

	},

	{

		.procname	= "tcp_invalid_ratelimit",

		.data		= &sysctl_tcp_invalid_ratelimit,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec_ms_jiffies,

	},

	{

		.procname	= "tcp_available_ulp",

		.maxlen		= TCP_ULP_BUF_MAX,
@@ -1145,6 +1053,98 @@ static struct ctl_table ipv4_net_table[] = { 
		.mode		= 0644,

		.proc_handler	= proc_dointvec

	},

	{

		.procname	= "tcp_no_metrics_save",

		.data		= &init_net.ipv4.sysctl_tcp_nometrics_save,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec,

	},

	{

		.procname	= "tcp_moderate_rcvbuf",

		.data		= &init_net.ipv4.sysctl_tcp_moderate_rcvbuf,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec,

	},

	{

		.procname	= "tcp_tso_win_divisor",

		.data		= &init_net.ipv4.sysctl_tcp_tso_win_divisor,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec,

	},

	{

		.procname	= "tcp_workaround_signed_windows",

		.data		= &init_net.ipv4.sysctl_tcp_workaround_signed_windows,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec

	},

	{

		.procname	= "tcp_limit_output_bytes",

		.data		= &init_net.ipv4.sysctl_tcp_limit_output_bytes,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec

	},

	{

		.procname	= "tcp_challenge_ack_limit",

		.data		= &init_net.ipv4.sysctl_tcp_challenge_ack_limit,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec

	},

	{

		.procname	= "tcp_min_tso_segs",

		.data		= &init_net.ipv4.sysctl_tcp_min_tso_segs,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec_minmax,

		.extra1		= &one,

		.extra2		= &gso_max_segs,

	},

	{

		.procname	= "tcp_min_rtt_wlen",

		.data		= &init_net.ipv4.sysctl_tcp_min_rtt_wlen,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec

	},

	{

		.procname	= "tcp_autocorking",

		.data		= &init_net.ipv4.sysctl_tcp_autocorking,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec_minmax,

		.extra1		= &zero,

		.extra2		= &one,

	},

	{

		.procname	= "tcp_invalid_ratelimit",

		.data		= &init_net.ipv4.sysctl_tcp_invalid_ratelimit,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec_ms_jiffies,

	},

	{

		.procname	= "tcp_pacing_ss_ratio",

		.data		= &init_net.ipv4.sysctl_tcp_pacing_ss_ratio,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec_minmax,

		.extra1		= &zero,

		.extra2		= &thousand,

	},

	{

		.procname	= "tcp_pacing_ca_ratio",

		.data		= &init_net.ipv4.sysctl_tcp_pacing_ca_ratio,

		.maxlen		= sizeof(int),

		.mode		= 0644,

		.proc_handler	= proc_dointvec_minmax,

		.extra1		= &zero,

		.extra2		= &thousand,

	},

	{ }

};

net/ipv4/tcp.c

+1 −5
Original line number Diff line number Diff line
@@ -285,10 +285,6 @@ 


#include <trace/events/tcp.h>



int sysctl_tcp_min_tso_segs __read_mostly = 2;



int sysctl_tcp_autocorking __read_mostly = 1;



struct percpu_counter tcp_orphan_count;

EXPORT_SYMBOL_GPL(tcp_orphan_count);
@@ -699,7 +695,7 @@ static bool tcp_should_autocork(struct sock *sk, struct sk_buff *skb, 
				int size_goal)

{

	return skb->len < size_goal &&

	       sysctl_tcp_autocorking &&

	       sock_net(sk)->ipv4.sysctl_tcp_autocorking &&

	       skb != tcp_write_queue_head(sk) &&

	       refcount_read(&sk->sk_wmem_alloc) > skb->truesize;

}

CRA Git | Maintained and supported by SUSTech CRA and CCSE