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/arch/x86/lib/usercopy.c b/arch/x86/lib/usercopy.c
new file mode 100644
index 000000000..24b48af27
--- /dev/null
+++ b/arch/x86/lib/usercopy.c
@@ -0,0 +1,55 @@
+/*
+ * User address space access functions.
+ *
+ *  For licencing details see kernel-base/COPYING
+ */
+
+#include <linux/uaccess.h>
+#include <linux/export.h>
+#include <linux/instrumented.h>
+
+#include <asm/tlbflush.h>
+
+/**
+ * copy_from_user_nmi - NMI safe copy from user
+ * @to:		Pointer to the destination buffer
+ * @from:	Pointer to a user space address of the current task
+ * @n:		Number of bytes to copy
+ *
+ * Returns: The number of not copied bytes. 0 is success, i.e. all bytes copied
+ *
+ * Contrary to other copy_from_user() variants this function can be called
+ * from NMI context. Despite the name it is not restricted to be called
+ * from NMI context. It is safe to be called from any other context as
+ * well. It disables pagefaults across the copy which means a fault will
+ * abort the copy.
+ *
+ * For NMI context invocations this relies on the nested NMI work to allow
+ * atomic faults from the NMI path; the nested NMI paths are careful to
+ * preserve CR2.
+ */
+unsigned long
+copy_from_user_nmi(void *to, const void __user *from, unsigned long n)
+{
+	unsigned long ret;
+
+	if (!__access_ok(from, n))
+		return n;
+
+	if (!nmi_uaccess_okay())
+		return n;
+
+	/*
+	 * Even though this function is typically called from NMI/IRQ context
+	 * disable pagefaults so that its behaviour is consistent even when
+	 * called from other contexts.
+	 */
+	pagefault_disable();
+	instrument_copy_from_user_before(to, from, n);
+	ret = raw_copy_from_user(to, from, n);
+	instrument_copy_from_user_after(to, from, n, ret);
+	pagefault_enable();
+
+	return ret;
+}
+EXPORT_SYMBOL_GPL(copy_from_user_nmi);