sched/cache: Keep nr_pref_llc_running in the runnable domain, to fix LLC mis-scheduling buggrafted

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 <zhanxusheng@xiaomi.com>
Suggested-by: Chen Yu <yu.c.chen@intel.com>
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Kayra Cizmeci <kayracizmeci@gmail.com>
Cc: <stable@kernel.org> # v7.2.x
Link: https://patch.msgid.link/06af61afedac32e6477f57feb4d658f6c411c3af.1790035273.git.tim.c.chen@linux.intel.com

1 files changed, 85 insertions, 0 deletions

diff --git a/include/drm/drm_accel.h b/include/drm/drm_accel.h
new file mode 100644
index 000000000..20a665ec6
--- /dev/null
+++ b/include/drm/drm_accel.h
@@ -0,0 +1,85 @@
+/* SPDX-License-Identifier: GPL-2.0
+ *
+ * Copyright 2022 HabanaLabs, Ltd.
+ * All Rights Reserved.
+ *
+ */
+
+#ifndef DRM_ACCEL_H_
+#define DRM_ACCEL_H_
+
+#include <drm/drm_file.h>
+
+#define ACCEL_MAJOR		261
+#define ACCEL_MAX_MINORS	256
+
+/**
+ * DRM_ACCEL_FOPS - Default drm accelerators file operations
+ *
+ * This macro provides a shorthand for setting the accelerator file ops in the
+ * &file_operations structure.  If all you need are the default ops, use
+ * DEFINE_DRM_ACCEL_FOPS instead.
+ */
+#define DRM_ACCEL_FOPS \
+	.open		= accel_open,\
+	.release	= drm_release,\
+	.unlocked_ioctl	= drm_ioctl,\
+	.compat_ioctl	= drm_compat_ioctl,\
+	.poll		= drm_poll,\
+	.read		= drm_read,\
+	.llseek		= noop_llseek, \
+	.mmap		= drm_gem_mmap, \
+	.fop_flags	= FOP_UNSIGNED_OFFSET
+
+/**
+ * DEFINE_DRM_ACCEL_FOPS() - macro to generate file operations for accelerators drivers
+ * @name: name for the generated structure
+ *
+ * This macro autogenerates a suitable &struct file_operations for accelerators based
+ * drivers, which can be assigned to &drm_driver.fops. Note that this structure
+ * cannot be shared between drivers, because it contains a reference to the
+ * current module using THIS_MODULE.
+ *
+ * Note that the declaration is already marked as static - if you need a
+ * non-static version of this you're probably doing it wrong and will break the
+ * THIS_MODULE reference by accident.
+ */
+#define DEFINE_DRM_ACCEL_FOPS(name) \
+	static const struct file_operations name = {\
+		.owner		= THIS_MODULE,\
+		DRM_ACCEL_FOPS,\
+	}
+
+#if IS_ENABLED(CONFIG_DRM_ACCEL)
+
+extern struct xarray accel_minors_xa;
+
+void accel_core_exit(void);
+int accel_core_init(void);
+void accel_set_device_instance_params(struct device *kdev, int index);
+int accel_open(struct inode *inode, struct file *filp);
+void accel_debugfs_register(struct drm_device *dev);
+
+#else
+
+static inline void accel_core_exit(void)
+{
+}
+
+static inline int __init accel_core_init(void)
+{
+	/* Return 0 to allow drm_core_init to complete successfully */
+	return 0;
+}
+
+static inline void accel_set_device_instance_params(struct device *kdev, int index)
+{
+}
+
+static inline void accel_debugfs_register(struct drm_device *dev)
+{
+}
+
+#endif /* IS_ENABLED(CONFIG_DRM_ACCEL) */
+
+#endif /* DRM_ACCEL_H_ */