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, 55 insertions, 0 deletions

diff --git a/Documentation/input/notifier.rst b/Documentation/input/notifier.rst
new file mode 100644
index 000000000..824379399
--- /dev/null
+++ b/Documentation/input/notifier.rst
@@ -0,0 +1,55 @@
+=================
+Keyboard notifier
+=================
+
+One can use register_keyboard_notifier to get called back on keyboard
+events (see kbd_keycode() function for details).  The passed structure is
+keyboard_notifier_param (see <linux/keyboard.h>):
+
+- 'vc' always provide the VC for which the keyboard event applies;
+- 'down' is 1 for a key press event, 0 for a key release;
+- 'shift' is the current modifier state, mask bit indexes are KG_*;
+- 'ledstate' is the current LED state;
+- 'value' depends on the type of event.
+
+- KBD_KEYCODE events are always sent before other events, value is the keycode.
+- KBD_UNBOUND_KEYCODE events are sent if the keycode is not bound to a keysym.
+  value is the keycode.
+- KBD_UNICODE events are sent if the keycode -> keysym translation produced a
+  unicode character. value is the unicode value.
+- KBD_KEYSYM events are sent if the keycode -> keysym translation produced a
+  non-unicode character. value is the keysym.
+- KBD_POST_KEYSYM events are sent after the treatment of non-unicode keysyms.
+  That permits one to inspect the resulting LEDs for instance.
+
+For each kind of event but the last, the callback may return NOTIFY_STOP in
+order to "eat" the event: the notify loop is stopped and the keyboard event is
+dropped.
+
+In a rough C snippet, we have::
+
+    kbd_keycode(keycode) {
+	...
+	params.value = keycode;
+	if (notifier_call_chain(KBD_KEYCODE,&params) == NOTIFY_STOP)
+	    || !bound) {
+		notifier_call_chain(KBD_UNBOUND_KEYCODE,&params);
+		return;
+	}
+
+	if (unicode) {
+		param.value = unicode;
+		if (notifier_call_chain(KBD_UNICODE,&params) == NOTIFY_STOP)
+			return;
+		emit unicode;
+		return;
+	}
+
+	params.value = keysym;
+	if (notifier_call_chain(KBD_KEYSYM,&params) == NOTIFY_STOP)
+		return;
+	apply keysym;
+	notifier_call_chain(KBD_POST_KEYSYM,&params);
+    }
+
+.. note:: This notifier is usually called from interrupt context.