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, 41 insertions, 0 deletions

diff --git a/net/Kconfig.debug b/net/Kconfig.debug
new file mode 100644
index 000000000..277fab8c4
--- /dev/null
+++ b/net/Kconfig.debug
@@ -0,0 +1,41 @@
+# SPDX-License-Identifier: GPL-2.0-only
+
+config NET_DEV_REFCNT_TRACKER
+	bool "Enable net device refcount tracking"
+	depends on DEBUG_KERNEL && STACKTRACE_SUPPORT && NET
+	select REF_TRACKER
+	default n
+	help
+	  Enable debugging feature to track device references.
+	  This adds memory and cpu costs.
+
+config NET_NS_REFCNT_TRACKER
+	bool "Enable networking namespace refcount tracking"
+	depends on DEBUG_KERNEL && STACKTRACE_SUPPORT && NET
+	select REF_TRACKER
+	default n
+	help
+	  Enable debugging feature to track netns references.
+	  This adds memory and cpu costs.
+
+config DEBUG_NET
+	bool "Add generic networking debug"
+	depends on DEBUG_KERNEL && NET
+	help
+	  Enable extra sanity checks in networking.
+	  This is mostly used by fuzzers, but is safe to select.
+
+config DEBUG_NET_SMALL_RTNL
+	bool "Add extra per-netns mutex inside RTNL"
+	depends on DEBUG_KERNEL && NET && LOCK_DEBUGGING_SUPPORT
+	select PROVE_LOCKING
+	default n
+	help
+	  rtnl_lock() is being replaced with rtnl_net_lock() that
+	  acquires the global RTNL and a small per-netns RTNL mutex.
+
+	  During the conversion, rtnl_net_lock() just adds an extra
+	  mutex in every RTNL scope and slows down the operations.
+
+	  Once the conversion completes, rtnl_lock() will be removed
+	  and rtnetlink will gain per-netns scalability.