From 0d6526f82c3cdefcca47f73f5fc08dc6f335eac6 Mon Sep 17 00:00:00 2001 From: Tim Chen Date: Mon, 21 Sep 2026 17:37:22 -0700 Subject: sched/cache: Keep nr_pref_llc_running in the runnable domain, to fix LLC mis-scheduling bug alb_break_llc() decides whether to break LLC preference during active load balance. It does so by testing that every runnable fair task on the source rq prefers its LLC: env->src_rq->nr_pref_llc_running == env->src_rq->cfs.h_nr_runnable But the two counters cover different sets. nr_pref_llc_running is updated in account_llc_enqueue()/account_llc_dequeue(), next to cfs_rq->nr_queued, so it follows queued tasks. h_nr_runnable is updated in set_delayed()/ clear_delayed() and drops delay-dequeued tasks. So under DELAY_DEQUEUE, a preferring task that goes to sleep stays counted in nr_pref_llc_running while h_nr_runnable falls. The equality then breaks, alb_break_llc() returns false, and active balance is free to pull a task off its preferred LLC. Active balance only moves runnable tasks, and this is the only LLC check it consults: once the stopper runs, LBF_ACTIVE_LB skips the per-task test in can_migrate_task(). The runnable set is the one we want. Fix it on the counter side. A task should be counted in nr_pref_llc_running exactly while it is both queued on its preferred LLC (pref_llc_queued) and runnable (!sched_delayed). Define that membership once in task_pref_llc_runnable(), and adjust the counter only through pref_llc_running_inc()/pref_llc_running_dec() from the four sites that change either input: account_llc_enqueue(), account_llc_dequeue(), set_delayed() and clear_delayed(). Gating every update on the same predicate keeps the delay, wake and dequeue paths from double-counting or underflowing; see the comments at those sites for the ordering. nr_llc_running and sd->llc_counts are not touched and stay on queued semantics. Fixes: 714059f79ff0 ("sched/cache: Handle moving single tasks to/from their preferred LLC") Closes: https://lore.kernel.org/lkml/20260827135000.735138-1-zhanxusheng@xiaomi.com/ Reported-by: Zhan Xusheng Suggested-by: Chen Yu Signed-off-by: Tim Chen Signed-off-by: Peter Zijlstra (Intel) Signed-off-by: Ingo Molnar Reviewed-by: Kayra Cizmeci Cc: # v7.2.x Link: https://patch.msgid.link/06af61afedac32e6477f57feb4d658f6c411c3af.1790035273.git.tim.c.chen@linux.intel.com --- net/mptcp/fastopen.c | 65 ++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 65 insertions(+) create mode 100644 net/mptcp/fastopen.c (limited to 'net/mptcp/fastopen.c') diff --git a/net/mptcp/fastopen.c b/net/mptcp/fastopen.c new file mode 100644 index 000000000..f71775090 --- /dev/null +++ b/net/mptcp/fastopen.c @@ -0,0 +1,65 @@ +// SPDX-License-Identifier: GPL-2.0 +/* MPTCP Fast Open Mechanism + * + * Copyright (c) 2021-2022, Dmytro SHYTYI + */ + +#include "protocol.h" + +void mptcp_fastopen_subflow_synack_set_params(struct mptcp_subflow_context *subflow, + struct request_sock *req) +{ + struct sock *sk, *ssk; + struct sk_buff *skb; + struct tcp_sock *tp; + bool has_rxtstamp; + + /* on early fallback the subflow context is deleted by + * subflow_syn_recv_sock() + */ + if (!subflow) + return; + + ssk = subflow->tcp_sock; + sk = subflow->conn; + tp = tcp_sk(ssk); + + /* A valid TFO cookie does not guarantee SYN data. */ + skb = skb_peek(&ssk->sk_receive_queue); + if (!skb) + return; + + subflow->is_mptfo = 1; + + /* dequeue the skb from sk receive queue */ + __skb_unlink(skb, &ssk->sk_receive_queue); + skb_ext_reset(skb); + + mptcp_subflow_lend_fwdmem(subflow, skb); + + /* We copy the fastopen data, but that don't belong to the mptcp sequence + * space, need to offset it in the subflow sequence, see mptcp_subflow_get_map_offset() + */ + tp->copied_seq += skb->len; + subflow->ssn_offset += skb->len; + has_rxtstamp = TCP_SKB_CB(skb)->has_rxtstamp; + + /* Only the sequence delta is relevant */ + MPTCP_SKB_CB(skb)->map_seq = -skb->len; + MPTCP_SKB_CB(skb)->end_seq = 0; + MPTCP_SKB_CB(skb)->offset = 0; + MPTCP_SKB_CB(skb)->has_rxtstamp = has_rxtstamp; + MPTCP_SKB_CB(skb)->cant_coalesce = 1; + + mptcp_data_lock(sk); + DEBUG_NET_WARN_ON_ONCE(sock_owned_by_user_nocheck(sk)); + + mptcp_borrow_fwdmem(sk, skb); + skb_set_owner_r(skb, sk); + __skb_queue_tail(&sk->sk_receive_queue, skb); + mptcp_sk(sk)->bytes_received += skb->len; + + sk->sk_data_ready(sk); + + mptcp_data_unlock(sk); +} -- cgit v1.3.1