Re: [PATCH v5] memcg: expose socket memory pressure in a cgroup
From: Kuniyuki Iwashima <kuniyu@google.com>
Date: 2025-10-15 18:21:33
Also in:
cgroups, linux-mm
Possibly related (same subject, not in this thread)
- 2025-10-15 · Re: [PATCH v5] memcg: expose socket memory pressure in a cgroup · Roman Gushchin <roman.gushchin@linux.dev>
- 2025-10-14 · Re: [PATCH v5] memcg: expose socket memory pressure in a cgroup · Daniel Sedlak <hidden>
- 2025-10-14 · Re: [PATCH v5] memcg: expose socket memory pressure in a cgroup · Roman Gushchin <roman.gushchin@linux.dev>
- 2025-10-09 · Re: [PATCH v5] memcg: expose socket memory pressure in a cgroup · Shakeel Butt <shakeel.butt@linux.dev>
- 2025-10-09 · Re: [PATCH v5] memcg: expose socket memory pressure in a cgroup · Roman Gushchin <roman.gushchin@linux.dev>
On Tue, Oct 14, 2025 at 1:33 PM Shakeel Butt [off-list ref] wrote:
On Mon, Oct 13, 2025 at 04:30:53PM +0200, Daniel Sedlak wrote: [...]quoted
quoted
quoted
quoted
quoted
How about we track the actions taken by the callers of mem_cgroup_sk_under_memory_pressure()? Basically if network stack reduces the buffer size or whatever the other actions it may take when mem_cgroup_sk_under_memory_pressure() returns, tracking those actions is what I think is needed here, at least for the debugging use-case.I am not against it, but I feel that conveying those tracked actions (or how to represent them) to the user will be much harder. Are there already existing APIs to push this information to the user?I discussed with Wei Wang and she suggested we should start tracking the calls to tcp_adjust_rcv_ssthresh() first. So, something like the following. I would like feedback frm networking folks as well:
I think we could simply put memcg_memory_event() in mem_cgroup_sk_under_memory_pressure() when it returns true. Other than tcp_adjust_rcv_ssthresh(), if tcp_under_memory_pressure() returns true, it indicates something bad will happen, failure to expand rcvbuf and sndbuf, need to prune out-of-order queue more aggressively, FIN deferred to a retransmitted packet. Also, we could cover mptcp and sctp too.
quoted hunk ↗ jump to hunk
From 54bd2bf6681c1c694295646532f2a62a205ee41a Mon Sep 17 00:00:00 2001 From: Shakeel Butt <shakeel.butt@linux.dev> Date: Tue, 14 Oct 2025 13:27:36 -0700 Subject: [PATCH] memcg: track network throttling due to memcg memory pressure Signed-off-by: Shakeel Butt <shakeel.butt@linux.dev> --- include/linux/memcontrol.h | 1 + mm/memcontrol.c | 2 ++ net/ipv4/tcp_input.c | 5 ++++- net/ipv4/tcp_output.c | 8 ++++++-- 4 files changed, 13 insertions(+), 3 deletions(-)diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h index 873e510d6f8d..5fe254813123 100644 --- a/include/linux/memcontrol.h +++ b/include/linux/memcontrol.h@@ -52,6 +52,7 @@ enum memcg_memory_event { MEMCG_SWAP_HIGH, MEMCG_SWAP_MAX, MEMCG_SWAP_FAIL, + MEMCG_SOCK_THROTTLED, MEMCG_NR_MEMORY_EVENTS, };diff --git a/mm/memcontrol.c b/mm/memcontrol.c index 4deda33625f4..9207bba34e2e 100644 --- a/mm/memcontrol.c +++ b/mm/memcontrol.c@@ -4463,6 +4463,8 @@ static void __memory_events_show(struct seq_file *m, atomic_long_t *events) atomic_long_read(&events[MEMCG_OOM_KILL])); seq_printf(m, "oom_group_kill %lu\n", atomic_long_read(&events[MEMCG_OOM_GROUP_KILL])); + seq_printf(m, "sock_throttled %lu\n", + atomic_long_read(&events[MEMCG_SOCK_THROTTLED])); } static int memory_events_show(struct seq_file *m, void *v)diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 31ea5af49f2d..2206968fb505 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c@@ -713,6 +713,7 @@ static void tcp_grow_window(struct sock *sk, const struct sk_buff *skb, * Adjust rcv_ssthresh according to reserved mem */ tcp_adjust_rcv_ssthresh(sk); + memcg_memory_event(sk->sk_memcg, MEMCG_SOCK_THROTTLED); } }@@ -5764,8 +5765,10 @@ static int tcp_prune_queue(struct sock *sk, const struct sk_buff *in_skb) if (!tcp_can_ingest(sk, in_skb)) tcp_clamp_window(sk); - else if (tcp_under_memory_pressure(sk)) + else if (tcp_under_memory_pressure(sk)) { tcp_adjust_rcv_ssthresh(sk); + memcg_memory_event(sk->sk_memcg, MEMCG_SOCK_THROTTLED); + } if (tcp_can_ingest(sk, in_skb)) return 0;diff --git a/net/ipv4/tcp_output.c b/net/ipv4/tcp_output.c index bb3576ac0ad7..8fe8d973d7ac 100644 --- a/net/ipv4/tcp_output.c +++ b/net/ipv4/tcp_output.c@@ -3275,8 +3275,10 @@ u32 __tcp_select_window(struct sock *sk) if (free_space < (full_space >> 1)) { icsk->icsk_ack.quick = 0; - if (tcp_under_memory_pressure(sk)) + if (tcp_under_memory_pressure(sk)) { tcp_adjust_rcv_ssthresh(sk); + memcg_memory_event(sk->sk_memcg, MEMCG_SOCK_THROTTLED); + } /* free_space might become our new window, make sure we don't * increase it due to wscale.@@ -3334,8 +3336,10 @@ u32 __tcp_select_window(struct sock *sk) if (free_space < (full_space >> 1)) { icsk->icsk_ack.quick = 0; - if (tcp_under_memory_pressure(sk)) + if (tcp_under_memory_pressure(sk)) { tcp_adjust_rcv_ssthresh(sk); + memcg_memory_event(sk->sk_memcg, MEMCG_SOCK_THROTTLED); + } /* if free space is too low, return a zero window */ if (free_space < (allowed_space >> 4) || free_space < mss || --2.47.3