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, 108 insertions, 0 deletions

diff --git a/Documentation/driver-api/soundwire/locking.rst b/Documentation/driver-api/soundwire/locking.rst
new file mode 100644
index 000000000..3a7ffb3d8
--- /dev/null
+++ b/Documentation/driver-api/soundwire/locking.rst
@@ -0,0 +1,108 @@
+=================
+SoundWire Locking
+=================
+
+This document explains locking mechanism of the SoundWire Bus. Bus uses
+following locks in order to avoid race conditions in Bus operations on
+shared resources.
+
+  - Bus lock
+
+  - Message lock
+
+Bus lock
+========
+
+SoundWire Bus lock is a mutex and is part of Bus data structure
+(sdw_bus) which is used for every Bus instance. This lock is used to
+serialize each of the following operations(s) within SoundWire Bus instance.
+
+  - Addition and removal of Slave(s), changing Slave status.
+
+  - Prepare, Enable, Disable and De-prepare stream operations.
+
+  - Access of Stream data structure.
+
+Message lock
+============
+
+SoundWire message transfer lock. This mutex is part of
+Bus data structure (sdw_bus). This lock is used to serialize the message
+transfers (read/write) within a SoundWire Bus instance.
+
+Below examples show how locks are acquired.
+
+Example 1
+---------
+
+Message transfer.
+
+  1. For every message transfer
+
+     a. Acquire Message lock.
+
+     b. Transfer message (Read/Write) to Slave1 or broadcast message on
+        Bus in case of bank switch.
+
+     c. Release Message lock
+
+     ::
+
+	+----------+                    +---------+
+	|          |                    |         |
+	|   Bus    |                    | Master  |
+	|          |                    | Driver  |
+	|          |                    |         |
+	+----+-----+                    +----+----+
+	     |                               |
+	     |     bus->ops->xfer_msg()      |
+	     <-------------------------------+   a. Acquire Message lock
+	     |                               |   b. Transfer message
+	     |                               |
+	     +------------------------------->   c. Release Message lock
+	     |    return success/error       |   d. Return success/error
+	     |                               |
+	     +                               +
+
+Example 2
+---------
+
+Prepare operation.
+
+  1. Acquire lock for Bus instance associated with Master 1.
+
+  2. For every message transfer in Prepare operation
+
+     a. Acquire Message lock.
+
+     b. Transfer message (Read/Write) to Slave1 or broadcast message on
+        Bus in case of bank switch.
+
+     c. Release Message lock.
+
+  3. Release lock for Bus instance associated with Master 1 ::
+
+	+----------+                    +---------+
+	|          |                    |         |
+	|   Bus    |                    | Master  |
+	|          |                    | Driver  |
+	|          |                    |         |
+	+----+-----+                    +----+----+
+	     |                               |
+	     |    sdw_prepare_stream()       |
+	     <-------------------------------+   1. Acquire bus lock
+	     |                               |   2. Perform stream prepare
+	     |                               |
+	     |                               |
+	     |     bus->ops->xfer_msg()      |
+	     <-------------------------------+   a. Acquire Message lock
+	     |                               |   b. Transfer message
+	     |                               |
+	     +------------------------------->   c. Release Message lock
+	     |    return success/error       |   d. Return success/error
+	     |                               |
+	     |                               |
+	     |    return success/error       |   3. Release bus lock
+	     +------------------------------->   4. Return success/error
+	     |                               |
+	     +                               +