From 19757cebf0c5016a1f36f7fe9810a9f0b33c0832 Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 14 Oct 2021 06:41:26 -0700 Subject: tcp: switch orphan_count to bare per-cpu counters Use of percpu_counter structure to track count of orphaned sockets is causing problems on modern hosts with 256 cpus or more. Stefan Bach reported a serious spinlock contention in real workloads, that I was able to reproduce with a netfilter rule dropping incoming FIN packets. 53.56% server [kernel.kallsyms] [k] queued_spin_lock_slowpath | ---queued_spin_lock_slowpath | --53.51%--_raw_spin_lock_irqsave | --53.51%--__percpu_counter_sum tcp_check_oom | |--39.03%--__tcp_close | tcp_close | inet_release | inet6_release | sock_close | __fput | ____fput | task_work_run | exit_to_usermode_loop | do_syscall_64 | entry_SYSCALL_64_after_hwframe | __GI___libc_close | --14.48%--tcp_out_of_resources tcp_write_timeout tcp_retransmit_timer tcp_write_timer_handler tcp_write_timer call_timer_fn expire_timers __run_timers run_timer_softirq __softirqentry_text_start As explained in commit cf86a086a180 ("net/dst: use a smaller percpu_counter batch for dst entries accounting"), default batch size is too big for the default value of tcp_max_orphans (262144). But even if we reduce batch sizes, there would still be cases where the estimated count of orphans is beyond the limit, and where tcp_too_many_orphans() has to call the expensive percpu_counter_sum_positive(). One solution is to use plain per-cpu counters, and have a timer to periodically refresh this cache. Updating this cache every 100ms seems about right, tcp pressure state is not radically changing over shorter periods. percpu_counter was nice 15 years ago while hosts had less than 16 cpus, not anymore by current standards. v2: Fix the build issue for CONFIG_CRYPTO_DEV_CHELSIO_TLS=m, reported by kernel test robot Remove unused socket argument from tcp_too_many_orphans() Fixes: dd24c00191d5 ("net: Use a percpu_counter for orphan_count") Signed-off-by: Eric Dumazet Reported-by: Stefan Bach Cc: Neal Cardwell Acked-by: Neal Cardwell Signed-off-by: David S. Miller --- include/net/tcp.h | 17 +++-------------- 1 file changed, 3 insertions(+), 14 deletions(-) (limited to 'include/net/tcp.h') diff --git a/include/net/tcp.h b/include/net/tcp.h index 4c2898ac6569..af77e6453b1b 100644 --- a/include/net/tcp.h +++ b/include/net/tcp.h @@ -48,7 +48,9 @@ extern struct inet_hashinfo tcp_hashinfo; -extern struct percpu_counter tcp_orphan_count; +DECLARE_PER_CPU(unsigned int, tcp_orphan_count); +int tcp_orphan_count_sum(void); + void tcp_time_wait(struct sock *sk, int state, int timeo); #define MAX_TCP_HEADER L1_CACHE_ALIGN(128 + MAX_HEADER) @@ -290,19 +292,6 @@ static inline bool tcp_out_of_memory(struct sock *sk) void sk_forced_mem_schedule(struct sock *sk, int size); -static inline bool tcp_too_many_orphans(struct sock *sk, int shift) -{ - struct percpu_counter *ocp = sk->sk_prot->orphan_count; - int orphans = percpu_counter_read_positive(ocp); - - if (orphans << shift > sysctl_tcp_max_orphans) { - orphans = percpu_counter_sum_positive(ocp); - if (orphans << shift > sysctl_tcp_max_orphans) - return true; - } - return false; -} - bool tcp_check_oom(struct sock *sk, int shift); -- cgit