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/io_uring/wait.h b/io_uring/wait.h
new file mode 100644
index 000000000..b7b9c46b1
--- /dev/null
+++ b/io_uring/wait.h
@@ -0,0 +1,55 @@
+// SPDX-License-Identifier: GPL-2.0
+#ifndef IOU_WAIT_H
+#define IOU_WAIT_H
+
+#include <linux/io_uring_types.h>
+
+/*
+ * ->cq_wait_nr is armed with the number of lazy task_work adds the waiter
+ * still needs, and counted down by the add side, with the add reaching zero
+ * issuing the (single) wake up for this wait cycle. Zero and below means no
+ * wake up is to be issued: IO_CQ_WAKE_INIT when no task is waiting (also
+ * what a forced wake up resets it to when claiming one), zero once the
+ * countdown has fired.
+ */
+#define IO_CQ_WAKE_INIT		(-1)
+
+struct ext_arg {
+	size_t argsz;
+	struct timespec64 ts;
+	const sigset_t __user *sig;
+	ktime_t min_time;
+	bool ts_set;
+	bool iowait;
+};
+
+int io_cqring_wait(struct io_ring_ctx *ctx, int min_events, u32 flags,
+		   struct ext_arg *ext_arg);
+int io_run_task_work_sig(struct io_ring_ctx *ctx);
+void io_cqring_do_overflow_flush(struct io_ring_ctx *ctx);
+void io_cqring_overflow_flush_locked(struct io_ring_ctx *ctx);
+
+static inline unsigned int __io_cqring_events(struct io_ring_ctx *ctx)
+{
+	struct io_rings *rings = io_get_rings(ctx);
+	return ctx->cached_cq_tail - READ_ONCE(rings->cq.head);
+}
+
+static inline unsigned int __io_cqring_events_user(struct io_ring_ctx *ctx)
+{
+	struct io_rings *rings = io_get_rings(ctx);
+
+	return READ_ONCE(rings->cq.tail) - READ_ONCE(rings->cq.head);
+}
+
+/*
+ * Reads the tail/head of the CQ ring while providing an acquire ordering,
+ * see comment at top of io_uring.c.
+ */
+static inline unsigned io_cqring_events(struct io_ring_ctx *ctx)
+{
+	smp_rmb();
+	return __io_cqring_events(ctx);
+}
+
+#endif