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/ceph/snapshot.c | 63 +++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 63 insertions(+) create mode 100644 net/ceph/snapshot.c (limited to 'net/ceph/snapshot.c') diff --git a/net/ceph/snapshot.c b/net/ceph/snapshot.c new file mode 100644 index 000000000..e24315937 --- /dev/null +++ b/net/ceph/snapshot.c @@ -0,0 +1,63 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * snapshot.c Ceph snapshot context utility routines (part of libceph) + * + * Copyright (C) 2013 Inktank Storage, Inc. + */ + +#include +#include +#include + +/* + * Ceph snapshot contexts are reference counted objects, and the + * returned structure holds a single reference. Acquire additional + * references with ceph_get_snap_context(), and release them with + * ceph_put_snap_context(). When the reference count reaches zero + * the entire structure is freed. + */ + +/* + * Create a new ceph snapshot context large enough to hold the + * indicated number of snapshot ids (which can be 0). Caller has + * to fill in snapc->seq and snapc->snaps[0..snap_count-1]. + * + * Returns a null pointer if an error occurs. + */ +struct ceph_snap_context *ceph_create_snap_context(u32 snap_count, + gfp_t gfp_flags) +{ + struct ceph_snap_context *snapc; + size_t size; + + size = sizeof (struct ceph_snap_context); + size += snap_count * sizeof (snapc->snaps[0]); + snapc = kzalloc(size, gfp_flags); + if (!snapc) + return NULL; + + refcount_set(&snapc->nref, 1); + snapc->num_snaps = snap_count; + + return snapc; +} +EXPORT_SYMBOL(ceph_create_snap_context); + +struct ceph_snap_context *ceph_get_snap_context(struct ceph_snap_context *sc) +{ + if (sc) + refcount_inc(&sc->nref); + return sc; +} +EXPORT_SYMBOL(ceph_get_snap_context); + +void ceph_put_snap_context(struct ceph_snap_context *sc) +{ + if (!sc) + return; + if (refcount_dec_and_test(&sc->nref)) { + /*printk(" deleting snap_context %p\n", sc);*/ + kfree(sc); + } +} +EXPORT_SYMBOL(ceph_put_snap_context); -- cgit v1.3.1