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 --- tools/testing/selftests/clock-helpers.h | 76 +++++++++++++++++++++++++++++++++ 1 file changed, 76 insertions(+) create mode 100644 tools/testing/selftests/clock-helpers.h (limited to 'tools/testing/selftests/clock-helpers.h') diff --git a/tools/testing/selftests/clock-helpers.h b/tools/testing/selftests/clock-helpers.h new file mode 100644 index 000000000..01451f538 --- /dev/null +++ b/tools/testing/selftests/clock-helpers.h @@ -0,0 +1,76 @@ +/* SPDX-License-Identifier: GPL-2.0 */ + +#ifndef __CLOCK_HELPERS_H +#define __CLOCK_HELPERS_H + +#include +#include + +#define MSEC_PER_SEC 1000LL +#define USEC_PER_MSEC 1000LL +#define NSEC_PER_USEC 1000LL +#define NSEC_PER_MSEC 1000000LL +#define USEC_PER_SEC 1000000LL +#define NSEC_PER_SEC 1000000000LL +#define PSEC_PER_SEC 1000000000000LL +#define FSEC_PER_SEC 1000000000000000LL + +#ifndef CLOCK_AUX +#define CLOCK_AUX 16 +#endif + +#ifndef MAX_AUX_CLOCKS +#define MAX_AUX_CLOCKS 8 +#endif + +#ifndef CLOCK_AUX_LAST +#define CLOCK_AUX_LAST (CLOCK_AUX + MAX_AUX_CLOCKS - 1) +#endif + +__attribute__((unused)) +static inline const char *clock_name(clockid_t clockid) +{ + switch (clockid) { + case CLOCK_REALTIME: + return "CLOCK_REALTIME"; + case CLOCK_MONOTONIC: + return "CLOCK_MONOTONIC"; + case CLOCK_PROCESS_CPUTIME_ID: + return "CLOCK_PROCESS_CPUTIME_ID"; + case CLOCK_THREAD_CPUTIME_ID: + return "CLOCK_THREAD_CPUTIME_ID"; + case CLOCK_MONOTONIC_RAW: + return "CLOCK_MONOTONIC_RAW"; + case CLOCK_REALTIME_COARSE: + return "CLOCK_REALTIME_COARSE"; + case CLOCK_MONOTONIC_COARSE: + return "CLOCK_MONOTONIC_COARSE"; + case CLOCK_BOOTTIME: + return "CLOCK_BOOTTIME"; + case CLOCK_REALTIME_ALARM: + return "CLOCK_REALTIME_ALARM"; + case CLOCK_BOOTTIME_ALARM: + return "CLOCK_BOOTTIME_ALARM"; + case CLOCK_TAI: + return "CLOCK_TAI"; + case CLOCK_AUX + 0: + return "CLOCK_AUX0"; + case CLOCK_AUX + 1: + return "CLOCK_AUX1"; + case CLOCK_AUX + 2: + return "CLOCK_AUX2"; + case CLOCK_AUX + 3: + return "CLOCK_AUX3"; + case CLOCK_AUX + 4: + return "CLOCK_AUX4"; + case CLOCK_AUX + 5: + return "CLOCK_AUX5"; + case CLOCK_AUX + 6: + return "CLOCK_AUX6"; + case CLOCK_AUX + 7: + return "CLOCK_AUX7"; + }; + return "UNKNOWN_CLOCKID"; +} + +#endif /* __CLOCK_HELPERS_H */ -- cgit v1.3.1