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, 97 insertions, 0 deletions

diff --git a/arch/powerpc/boot/opal.c b/arch/powerpc/boot/opal.c
new file mode 100644
index 000000000..b69818ce5
--- /dev/null
+++ b/arch/powerpc/boot/opal.c
@@ -0,0 +1,97 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Copyright (c) 2016 IBM Corporation.
+ */
+
+#include "ops.h"
+#include "stdio.h"
+#include "io.h"
+#include <libfdt.h>
+#include "../include/asm/opal-api.h"
+
+/* Global OPAL struct used by opal-call.S */
+struct opal {
+	u64 base;
+	u64 entry;
+} opal;
+
+static u32 opal_con_id;
+
+/* see opal-wrappers.S */
+int64_t opal_console_write(int64_t term_number, u64 *length, const u8 *buffer);
+int64_t opal_console_read(int64_t term_number, uint64_t *length, u8 *buffer);
+int64_t opal_console_write_buffer_space(uint64_t term_number, uint64_t *length);
+int64_t opal_console_flush(uint64_t term_number);
+int64_t opal_poll_events(uint64_t *outstanding_event_mask);
+
+void opal_kentry(unsigned long fdt_addr, void *vmlinux_addr);
+
+static int opal_con_open(void)
+{
+	/*
+	 * When OPAL loads the boot kernel it stashes the OPAL base and entry
+	 * address in r8 and r9 so the kernel can use the OPAL console
+	 * before unflattening the devicetree. While executing the wrapper will
+	 * probably trash r8 and r9 so this kentry hook restores them before
+	 * entering the decompressed kernel.
+	 */
+	platform_ops.kentry = opal_kentry;
+	return 0;
+}
+
+static void opal_con_putc(unsigned char c)
+{
+	int64_t rc;
+	uint64_t olen, len;
+
+	do {
+		rc = opal_console_write_buffer_space(opal_con_id, &olen);
+		len = be64_to_cpu(olen);
+		if (rc)
+			return;
+		opal_poll_events(NULL);
+	} while (len < 1);
+
+
+	olen = cpu_to_be64(1);
+	opal_console_write(opal_con_id, &olen, &c);
+}
+
+static void opal_con_close(void)
+{
+	opal_console_flush(opal_con_id);
+}
+
+static void opal_init(void)
+{
+	void *opal_node;
+
+	opal_node = finddevice("/ibm,opal");
+	if (!opal_node)
+		return;
+	if (getprop(opal_node, "opal-base-address", &opal.base, sizeof(u64)) < 0)
+		return;
+	opal.base = be64_to_cpu(opal.base);
+	if (getprop(opal_node, "opal-entry-address", &opal.entry, sizeof(u64)) < 0)
+		return;
+	opal.entry = be64_to_cpu(opal.entry);
+}
+
+int opal_console_init(void *devp, struct serial_console_data *scdp)
+{
+	opal_init();
+
+	if (devp) {
+		int n = getprop(devp, "reg", &opal_con_id, sizeof(u32));
+		if (n != sizeof(u32))
+			return -1;
+		opal_con_id = be32_to_cpu(opal_con_id);
+	} else
+		opal_con_id = 0;
+
+	scdp->open = opal_con_open;
+	scdp->putc = opal_con_putc;
+	scdp->close = opal_con_close;
+
+	return 0;
+}