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, 114 insertions, 0 deletions

diff --git a/Documentation/process/index.rst b/Documentation/process/index.rst
new file mode 100644
index 000000000..aa7c959a5
--- /dev/null
+++ b/Documentation/process/index.rst
@@ -0,0 +1,114 @@
+.. raw:: latex
+
+	\renewcommand\thesection*
+	\renewcommand\thesubsection*
+
+.. _process_index:
+
+=============================================
+Working with the kernel development community
+=============================================
+
+So you want to be a Linux kernel developer?  Welcome!  While there is a lot
+to be learned about the kernel in a technical sense, it is also important
+to learn about how our community works.  Reading these documents will make
+it much easier for you to get your changes merged with a minimum of
+trouble.
+
+An introduction to how kernel development works
+-----------------------------------------------
+
+Read these documents first: an understanding of the material here will ease
+your entry into the kernel community.
+
+.. toctree::
+   :maxdepth: 1
+
+   howto
+   development-process
+   submitting-patches
+   submit-checklist
+
+Tools and technical guides for kernel developers
+------------------------------------------------
+
+This is a collection of material that kernel developers should be familiar
+with.
+
+.. toctree::
+   :maxdepth: 1
+
+   changes
+   programming-language
+   coding-style
+   maintainer-pgp-guide
+   email-clients
+   applying-patches
+   backporting
+   adding-syscalls
+   volatile-considered-harmful
+   botching-up-ioctls
+
+Policy guides and developer statements
+--------------------------------------
+
+These are the rules that we try to live by in the kernel community (and
+beyond).
+
+.. toctree::
+   :maxdepth: 1
+
+   license-rules
+   code-of-conduct
+   code-of-conduct-interpretation
+   contribution-maturity-model
+   kernel-enforcement-statement
+   kernel-driver-statement
+   stable-api-nonsense
+   stable-kernel-rules
+   management-style
+   researcher-guidelines
+   generated-content
+   coding-assistants
+   conclave
+
+Dealing with bugs
+-----------------
+
+Bugs are a fact of life; it is important that we handle them properly. The
+documents below provide general advice about debugging and describe our
+policies around the handling of a couple of special classes of bugs:
+regressions and security problems.
+
+.. toctree::
+   :maxdepth: 1
+
+   debugging/index
+   handling-regressions
+   security-bugs
+   threat-model
+   cve
+   embargoed-hardware-issues
+
+Maintainer information
+----------------------
+
+How to find the people who will accept your patches.
+
+.. toctree::
+   :maxdepth: 1
+
+   maintainer-handbooks
+   maintainers
+
+Other material
+--------------
+
+Here are some other guides to the community that are of interest to most
+developers:
+
+.. toctree::
+   :maxdepth: 1
+
+   kernel-docs
+   deprecated