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 --- drivers/char/tpm/tpmrm-dev.c | 54 ++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 54 insertions(+) create mode 100644 drivers/char/tpm/tpmrm-dev.c (limited to 'drivers/char/tpm/tpmrm-dev.c') diff --git a/drivers/char/tpm/tpmrm-dev.c b/drivers/char/tpm/tpmrm-dev.c new file mode 100644 index 000000000..19e8f2779 --- /dev/null +++ b/drivers/char/tpm/tpmrm-dev.c @@ -0,0 +1,54 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * Copyright (C) 2017 James.Bottomley@HansenPartnership.com + */ +#include +#include "tpm-dev.h" + +struct tpmrm_priv { + struct file_priv priv; + struct tpm_space space; +}; + +static int tpmrm_open(struct inode *inode, struct file *file) +{ + struct tpm_chip *chip; + struct tpmrm_priv *priv; + int rc; + + chip = container_of(inode->i_cdev, struct tpm_chip, cdevs); + priv = kzalloc_obj(*priv); + if (priv == NULL) + return -ENOMEM; + + rc = tpm2_init_space(&priv->space, TPM2_SPACE_BUFFER_SIZE); + if (rc) { + kfree(priv); + return -ENOMEM; + } + + tpm_common_open(file, chip, &priv->priv, &priv->space); + + return nonseekable_open(inode, file); +} + +static int tpmrm_release(struct inode *inode, struct file *file) +{ + struct file_priv *fpriv = file->private_data; + struct tpmrm_priv *priv = container_of(fpriv, struct tpmrm_priv, priv); + + tpm_common_release(file, fpriv); + tpm2_del_space(fpriv->chip, &priv->space); + kfree(priv); + + return 0; +} + +const struct file_operations tpmrm_fops = { + .owner = THIS_MODULE, + .open = tpmrm_open, + .read = tpm_common_read, + .write = tpm_common_write, + .poll = tpm_common_poll, + .release = tpmrm_release, +}; -- cgit v1.3.1