diff options
Diffstat (limited to 'include/net/netns/ipv4.h')
| -rw-r--r-- | include/net/netns/ipv4.h | 242 |
1 files changed, 151 insertions, 91 deletions
diff --git a/include/net/netns/ipv4.h b/include/net/netns/ipv4.h index 8e4fcac4df72..6373e3f17da8 100644 --- a/include/net/netns/ipv4.h +++ b/include/net/netns/ipv4.h @@ -9,9 +9,9 @@ #include <linux/uidgid.h> #include <net/inet_frag.h> #include <linux/rcupdate.h> +#include <linux/seqlock.h> #include <linux/siphash.h> -struct tcpm_hash_bucket; struct ctl_table_header; struct ipv4_devconf; struct fib_rules_ops; @@ -19,8 +19,7 @@ struct hlist_head; struct fib_table; struct sock; struct local_ports { - seqlock_t lock; - int range[2]; + u32 range; /* high << 16 | low */ bool warned; }; @@ -32,15 +31,70 @@ struct ping_group_range { struct inet_hashinfo; struct inet_timewait_death_row { - atomic_t tw_count; + refcount_t tw_refcount; + /* Padding to avoid false sharing, tw_refcount can be often written */ struct inet_hashinfo *hashinfo ____cacheline_aligned_in_smp; int sysctl_max_tw_buckets; }; struct tcp_fastopen_context; +#ifdef CONFIG_IP_ROUTE_MULTIPATH +struct sysctl_fib_multipath_hash_seed { + u32 user_seed; + u32 mp_seed; +}; +#endif + +struct udp_tunnel_gro { + struct sock __rcu *sk; + struct hlist_head list; +}; + struct netns_ipv4 { + /* Cacheline organization can be found documented in + * Documentation/networking/net_cachelines/netns_ipv4_sysctl.rst. + * Please update the document when adding new fields. + */ + + /* TX readonly hotpath cache lines */ + __cacheline_group_begin(netns_ipv4_read_tx); + u8 sysctl_tcp_early_retrans; + u8 sysctl_tcp_tso_win_divisor; + u8 sysctl_tcp_tso_rtt_log; + u8 sysctl_tcp_autocorking; + int sysctl_tcp_min_snd_mss; + unsigned int sysctl_tcp_notsent_lowat; + int sysctl_tcp_limit_output_bytes; + int sysctl_tcp_min_rtt_wlen; + int sysctl_tcp_wmem[3]; + u8 sysctl_ip_fwd_use_pmtu; + __cacheline_group_end(netns_ipv4_read_tx); + + /* TXRX readonly hotpath cache lines */ + __cacheline_group_begin(netns_ipv4_read_txrx); + u8 sysctl_tcp_moderate_rcvbuf; + __cacheline_group_end(netns_ipv4_read_txrx); + + /* RX readonly hotpath cache line */ + __cacheline_group_begin(netns_ipv4_read_rx); + u8 sysctl_ip_early_demux; + u8 sysctl_tcp_early_demux; + u8 sysctl_tcp_l3mdev_accept; + /* 3 bytes hole, try to pack */ + int sysctl_tcp_reordering; + int sysctl_tcp_rmem[3]; + __cacheline_group_end(netns_ipv4_read_rx); + + struct inet_timewait_death_row tcp_death_row; + struct udp_table *udp_table; + +#if IS_ENABLED(CONFIG_NET_UDP_TUNNEL) + /* Not in a pernet subsys because need to be available at GRO stage */ + struct udp_tunnel_gro udp_tunnel_gro[2]; +#endif + #ifdef CONFIG_SYSCTL struct ctl_table_header *forw_hdr; struct ctl_table_header *frags_hdr; @@ -54,153 +108,154 @@ struct netns_ipv4 { struct mutex ra_mutex; #ifdef CONFIG_IP_MULTIPLE_TABLES struct fib_rules_ops *rules_ops; - bool fib_has_custom_rules; - unsigned int fib_rules_require_fldissect; struct fib_table __rcu *fib_main; struct fib_table __rcu *fib_default; + unsigned int fib_rules_require_fldissect; + bool fib_has_custom_rules; #endif bool fib_has_custom_local_routes; + bool fib_offload_disabled; + u8 sysctl_tcp_shrink_window; #ifdef CONFIG_IP_ROUTE_CLASSID - int fib_num_tclassid_users; + atomic_t fib_num_tclassid_users; #endif struct hlist_head *fib_table_hash; - bool fib_offload_disabled; struct sock *fibnl; + struct hlist_head *fib_info_hash; + unsigned int fib_info_hash_bits; + unsigned int fib_info_cnt; - struct sock * __percpu *icmp_sk; struct sock *mc_autojoin_sk; struct inet_peer_base *peers; - struct sock * __percpu *tcp_sk; struct fqdir *fqdir; -#ifdef CONFIG_NETFILTER - struct xt_table *iptable_filter; - struct xt_table *iptable_mangle; - struct xt_table *iptable_raw; - struct xt_table *arptable_filter; -#ifdef CONFIG_SECURITY - struct xt_table *iptable_security; -#endif - struct xt_table *nat_table; -#endif - int sysctl_icmp_echo_ignore_all; - int sysctl_icmp_echo_ignore_broadcasts; - int sysctl_icmp_ignore_bogus_error_responses; + u8 sysctl_icmp_echo_ignore_all; + u8 sysctl_icmp_echo_enable_probe; + u8 sysctl_icmp_echo_ignore_broadcasts; + u8 sysctl_icmp_ignore_bogus_error_responses; + u8 sysctl_icmp_errors_use_inbound_ifaddr; int sysctl_icmp_ratelimit; int sysctl_icmp_ratemask; - int sysctl_icmp_errors_use_inbound_ifaddr; + int sysctl_icmp_msgs_per_sec; + int sysctl_icmp_msgs_burst; + atomic_t icmp_global_credit; + u32 icmp_global_stamp; + u32 ip_rt_min_pmtu; + int ip_rt_mtu_expires; + int ip_rt_min_advmss; struct local_ports ip_local_ports; - int sysctl_tcp_ecn; - int sysctl_tcp_ecn_fallback; + u8 sysctl_tcp_ecn; + u8 sysctl_tcp_ecn_fallback; - int sysctl_ip_default_ttl; - int sysctl_ip_no_pmtu_disc; - int sysctl_ip_fwd_use_pmtu; - int sysctl_ip_fwd_update_priority; - int sysctl_ip_nonlocal_bind; - int sysctl_ip_autobind_reuse; + u8 sysctl_ip_default_ttl; + u8 sysctl_ip_no_pmtu_disc; + u8 sysctl_ip_fwd_update_priority; + u8 sysctl_ip_nonlocal_bind; + u8 sysctl_ip_autobind_reuse; /* Shall we try to damage output packets if routing dev changes? */ - int sysctl_ip_dynaddr; - int sysctl_ip_early_demux; + u8 sysctl_ip_dynaddr; #ifdef CONFIG_NET_L3_MASTER_DEV - int sysctl_raw_l3mdev_accept; + u8 sysctl_raw_l3mdev_accept; #endif - int sysctl_tcp_early_demux; - int sysctl_udp_early_demux; + u8 sysctl_udp_early_demux; - int sysctl_nexthop_compat_mode; + u8 sysctl_nexthop_compat_mode; - int sysctl_fwmark_reflect; - int sysctl_tcp_fwmark_accept; -#ifdef CONFIG_NET_L3_MASTER_DEV - int sysctl_tcp_l3mdev_accept; -#endif - int sysctl_tcp_mtu_probing; + u8 sysctl_fwmark_reflect; + u8 sysctl_tcp_fwmark_accept; + u8 sysctl_tcp_mtu_probing; int sysctl_tcp_mtu_probe_floor; int sysctl_tcp_base_mss; - int sysctl_tcp_min_snd_mss; int sysctl_tcp_probe_threshold; u32 sysctl_tcp_probe_interval; int sysctl_tcp_keepalive_time; - int sysctl_tcp_keepalive_probes; int sysctl_tcp_keepalive_intvl; + u8 sysctl_tcp_keepalive_probes; - int sysctl_tcp_syn_retries; - int sysctl_tcp_synack_retries; - int sysctl_tcp_syncookies; - int sysctl_tcp_reordering; - int sysctl_tcp_retries1; - int sysctl_tcp_retries2; - int sysctl_tcp_orphan_retries; + u8 sysctl_tcp_syn_retries; + u8 sysctl_tcp_synack_retries; + u8 sysctl_tcp_syncookies; + u8 sysctl_tcp_migrate_req; + u8 sysctl_tcp_comp_sack_nr; + u8 sysctl_tcp_backlog_ack_defer; + u8 sysctl_tcp_pingpong_thresh; + + u8 sysctl_tcp_retries1; + u8 sysctl_tcp_retries2; + u8 sysctl_tcp_orphan_retries; + u8 sysctl_tcp_tw_reuse; + unsigned int sysctl_tcp_tw_reuse_delay; int sysctl_tcp_fin_timeout; - unsigned int sysctl_tcp_notsent_lowat; - int sysctl_tcp_tw_reuse; - int sysctl_tcp_sack; - int sysctl_tcp_window_scaling; - int sysctl_tcp_timestamps; - int sysctl_tcp_early_retrans; - int sysctl_tcp_recovery; - int sysctl_tcp_thin_linear_timeouts; - int sysctl_tcp_slow_start_after_idle; - int sysctl_tcp_retrans_collapse; - int sysctl_tcp_stdurg; - int sysctl_tcp_rfc1337; - int sysctl_tcp_abort_on_overflow; - int sysctl_tcp_fack; + u8 sysctl_tcp_sack; + u8 sysctl_tcp_window_scaling; + u8 sysctl_tcp_timestamps; + int sysctl_tcp_rto_min_us; + int sysctl_tcp_rto_max_ms; + u8 sysctl_tcp_recovery; + u8 sysctl_tcp_thin_linear_timeouts; + u8 sysctl_tcp_slow_start_after_idle; + u8 sysctl_tcp_retrans_collapse; + u8 sysctl_tcp_stdurg; + u8 sysctl_tcp_rfc1337; + u8 sysctl_tcp_abort_on_overflow; + u8 sysctl_tcp_fack; /* obsolete */ int sysctl_tcp_max_reordering; - int sysctl_tcp_dsack; - int sysctl_tcp_app_win; - int sysctl_tcp_adv_win_scale; - int sysctl_tcp_frto; - int sysctl_tcp_nometrics_save; - int sysctl_tcp_no_ssthresh_metrics_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_adv_win_scale; /* obsolete */ + u8 sysctl_tcp_dsack; + u8 sysctl_tcp_app_win; + u8 sysctl_tcp_frto; + u8 sysctl_tcp_nometrics_save; + u8 sysctl_tcp_no_ssthresh_metrics_save; + u8 sysctl_tcp_workaround_signed_windows; int sysctl_tcp_challenge_ack_limit; - int sysctl_tcp_min_tso_segs; - int sysctl_tcp_min_rtt_wlen; - int sysctl_tcp_autocorking; + u8 sysctl_tcp_min_tso_segs; + u8 sysctl_tcp_reflect_tos; int sysctl_tcp_invalid_ratelimit; int sysctl_tcp_pacing_ss_ratio; int sysctl_tcp_pacing_ca_ratio; - int sysctl_tcp_wmem[3]; - int sysctl_tcp_rmem[3]; - int sysctl_tcp_comp_sack_nr; + unsigned int sysctl_tcp_child_ehash_entries; unsigned long sysctl_tcp_comp_sack_delay_ns; unsigned long sysctl_tcp_comp_sack_slack_ns; - struct inet_timewait_death_row tcp_death_row; int sysctl_max_syn_backlog; int sysctl_tcp_fastopen; const struct tcp_congestion_ops __rcu *tcp_congestion_control; struct tcp_fastopen_context __rcu *tcp_fastopen_ctx; - spinlock_t tcp_fastopen_ctx_lock; unsigned int sysctl_tcp_fastopen_blackhole_timeout; atomic_t tfo_active_disable_times; unsigned long tfo_active_disable_stamp; - int sysctl_tcp_reflect_tos; + u32 tcp_challenge_timestamp; + u32 tcp_challenge_count; + u8 sysctl_tcp_plb_enabled; + u8 sysctl_tcp_plb_idle_rehash_rounds; + u8 sysctl_tcp_plb_rehash_rounds; + u8 sysctl_tcp_plb_suspend_rto_sec; + int sysctl_tcp_plb_cong_thresh; int sysctl_udp_wmem_min; int sysctl_udp_rmem_min; + u8 sysctl_fib_notify_on_flag_change; + u8 sysctl_tcp_syn_linear_timeouts; + #ifdef CONFIG_NET_L3_MASTER_DEV - int sysctl_udp_l3mdev_accept; + u8 sysctl_udp_l3mdev_accept; #endif + u8 sysctl_igmp_llm_reports; int sysctl_igmp_max_memberships; int sysctl_igmp_max_msf; - int sysctl_igmp_llm_reports; int sysctl_igmp_qrv; struct ping_group_range ping_group_range; atomic_t dev_addr_genid; + unsigned int sysctl_udp_child_hash_entries; + #ifdef CONFIG_SYSCTL unsigned long *sysctl_local_reserved_ports; int sysctl_ip_prot_sock; @@ -215,17 +270,22 @@ struct netns_ipv4 { #endif #endif #ifdef CONFIG_IP_ROUTE_MULTIPATH - int sysctl_fib_multipath_use_neigh; - int sysctl_fib_multipath_hash_policy; + struct sysctl_fib_multipath_hash_seed sysctl_fib_multipath_hash_seed; + u32 sysctl_fib_multipath_hash_fields; + u8 sysctl_fib_multipath_use_neigh; + u8 sysctl_fib_multipath_hash_policy; #endif struct fib_notifier_ops *notifier_ops; - unsigned int fib_seq; /* protected by rtnl_mutex */ + unsigned int fib_seq; /* writes protected by rtnl_mutex */ struct fib_notifier_ops *ipmr_notifier_ops; unsigned int ipmr_seq; /* protected by rtnl_mutex */ atomic_t rt_genid; siphash_key_t ip_id_key; + struct hlist_head *inet_addr_lst; + struct delayed_work addr_chk_work; }; + #endif |