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, 77 insertions, 0 deletions

diff --git a/Documentation/admin-guide/media/ci.rst b/Documentation/admin-guide/media/ci.rst
new file mode 100644
index 000000000..ded4d8fbb
--- /dev/null
+++ b/Documentation/admin-guide/media/ci.rst
@@ -0,0 +1,77 @@
+.. SPDX-License-Identifier: GPL-2.0
+
+Digital TV Conditional Access Interface
+=======================================
+
+
+.. note::
+
+   This documentation is outdated.
+
+This document describes the usage of the high level CI API as
+in accordance to the Linux DVB API. This is a not a documentation for the,
+existing low level CI API.
+
+.. note::
+
+   For the Twinhan/Twinhan clones, the dst_ca module handles the CI
+   hardware handling. This module is loaded automatically if a CI
+   (Common Interface, that holds the CAM (Conditional Access Module)
+   is detected.
+
+ca_zap
+~~~~~~
+
+A userspace application, like ``ca_zap`` is required to handle encrypted
+MPEG-TS streams.
+
+The ``ca_zap`` userland application is in charge of sending the
+descrambling related information to the Conditional Access Module (CAM).
+
+This application requires the following to function properly as of now.
+
+a) Tune to a valid channel, with szap.
+
+  eg: $ szap -c channels.conf -r "TMC" -x
+
+b) a channels.conf containing a valid PMT PID
+
+  eg: TMC:11996:h:0:27500:278:512:650:321
+
+  here 278 is a valid PMT PID. the rest of the values are the
+  same ones that szap uses.
+
+c) after running a szap, you have to run ca_zap, for the
+   descrambler to function,
+
+  eg: $ ca_zap channels.conf "TMC"
+
+d) Hopefully enjoy your favourite subscribed channel as you do with
+   a FTA card.
+
+.. note::
+
+  Currently ca_zap, and dst_test, both are meant for demonstration
+  purposes only, they can become full fledged applications if necessary.
+
+
+Cards that fall in this category
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+At present the cards that fall in this category are the Twinhan and its
+clones, these cards are available as VVMER, Tomato, Hercules, Orange and
+so on.
+
+CI modules that are supported
+~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
+
+The CI module support is largely dependent upon the firmware on the cards
+Some cards do support almost all of the available CI modules. There is
+nothing much that can be done in order to make additional CI modules
+working with these cards.
+
+Modules that have been tested by this driver at present are
+
+(1) Irdeto 1 and 2 from SCM
+(2) Viaccess from SCM
+(3) Dragoncam