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, 319 insertions, 0 deletions

diff --git a/drivers/md/persistent-data/dm-bitset.c b/drivers/md/persistent-data/dm-bitset.c
new file mode 100644
index 000000000..00c0a3f18
--- /dev/null
+++ b/drivers/md/persistent-data/dm-bitset.c
@@ -0,0 +1,319 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2012 Red Hat, Inc.
+ *
+ * This file is released under the GPL.
+ */
+
+#include "dm-bitset.h"
+#include "dm-transaction-manager.h"
+
+#include <linux/export.h>
+#include <linux/device-mapper.h>
+
+#define DM_MSG_PREFIX "bitset"
+#define BITS_PER_ARRAY_ENTRY 64
+
+/*----------------------------------------------------------------*/
+
+static struct dm_btree_value_type bitset_bvt = {
+	.context = NULL,
+	.size = sizeof(__le64),
+	.inc = NULL,
+	.dec = NULL,
+	.equal = NULL,
+};
+
+/*----------------------------------------------------------------*/
+
+void dm_disk_bitset_init(struct dm_transaction_manager *tm,
+			 struct dm_disk_bitset *info)
+{
+	dm_array_info_init(&info->array_info, tm, &bitset_bvt);
+	info->current_index_set = false;
+}
+EXPORT_SYMBOL_GPL(dm_disk_bitset_init);
+
+int dm_bitset_empty(struct dm_disk_bitset *info, dm_block_t *root)
+{
+	return dm_array_empty(&info->array_info, root);
+}
+EXPORT_SYMBOL_GPL(dm_bitset_empty);
+
+struct packer_context {
+	bit_value_fn fn;
+	unsigned int nr_bits;
+	void *context;
+};
+
+static int pack_bits(uint32_t index, void *value, void *context)
+{
+	int r;
+	struct packer_context *p = context;
+	unsigned int bit, nr = min(64u, p->nr_bits - (index * 64));
+	uint64_t word = 0;
+	bool bv;
+
+	for (bit = 0; bit < nr; bit++) {
+		r = p->fn(index * 64 + bit, &bv, p->context);
+		if (r)
+			return r;
+
+		if (bv)
+			set_bit(bit, (unsigned long *) &word);
+		else
+			clear_bit(bit, (unsigned long *) &word);
+	}
+
+	*((__le64 *) value) = cpu_to_le64(word);
+
+	return 0;
+}
+
+int dm_bitset_new(struct dm_disk_bitset *info, dm_block_t *root,
+		  uint32_t size, bit_value_fn fn, void *context)
+{
+	struct packer_context p;
+
+	p.fn = fn;
+	p.nr_bits = size;
+	p.context = context;
+
+	return dm_array_new(&info->array_info, root, dm_div_up(size, 64), pack_bits, &p);
+}
+EXPORT_SYMBOL_GPL(dm_bitset_new);
+
+int dm_bitset_resize(struct dm_disk_bitset *info, dm_block_t root,
+		     uint32_t old_nr_entries, uint32_t new_nr_entries,
+		     bool default_value, dm_block_t *new_root)
+{
+	uint32_t old_blocks = dm_div_up(old_nr_entries, BITS_PER_ARRAY_ENTRY);
+	uint32_t new_blocks = dm_div_up(new_nr_entries, BITS_PER_ARRAY_ENTRY);
+	__le64 value = default_value ? cpu_to_le64(~0) : cpu_to_le64(0);
+
+	__dm_bless_for_disk(&value);
+	return dm_array_resize(&info->array_info, root, old_blocks, new_blocks,
+			       &value, new_root);
+}
+EXPORT_SYMBOL_GPL(dm_bitset_resize);
+
+int dm_bitset_del(struct dm_disk_bitset *info, dm_block_t root)
+{
+	return dm_array_del(&info->array_info, root);
+}
+EXPORT_SYMBOL_GPL(dm_bitset_del);
+
+int dm_bitset_flush(struct dm_disk_bitset *info, dm_block_t root,
+		    dm_block_t *new_root)
+{
+	int r;
+	__le64 value;
+
+	if (!info->current_index_set || !info->dirty)
+		return 0;
+
+	value = cpu_to_le64(info->current_bits);
+
+	__dm_bless_for_disk(&value);
+	r = dm_array_set_value(&info->array_info, root, info->current_index,
+			       &value, new_root);
+	if (r)
+		return r;
+
+	info->current_index_set = false;
+	info->dirty = false;
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(dm_bitset_flush);
+
+static int read_bits(struct dm_disk_bitset *info, dm_block_t root,
+		     uint32_t array_index)
+{
+	int r;
+	__le64 value;
+
+	r = dm_array_get_value(&info->array_info, root, array_index, &value);
+	if (r)
+		return r;
+
+	info->current_bits = le64_to_cpu(value);
+	info->current_index_set = true;
+	info->current_index = array_index;
+	info->dirty = false;
+
+	return 0;
+}
+
+static int get_array_entry(struct dm_disk_bitset *info, dm_block_t root,
+			   uint32_t index, dm_block_t *new_root)
+{
+	int r;
+	unsigned int array_index = index / BITS_PER_ARRAY_ENTRY;
+
+	if (info->current_index_set) {
+		if (info->current_index == array_index)
+			return 0;
+
+		r = dm_bitset_flush(info, root, new_root);
+		if (r)
+			return r;
+	}
+
+	return read_bits(info, root, array_index);
+}
+
+int dm_bitset_set_bit(struct dm_disk_bitset *info, dm_block_t root,
+		      uint32_t index, dm_block_t *new_root)
+{
+	int r;
+	unsigned int b = index % BITS_PER_ARRAY_ENTRY;
+
+	r = get_array_entry(info, root, index, new_root);
+	if (r)
+		return r;
+
+	set_bit(b, (unsigned long *) &info->current_bits);
+	info->dirty = true;
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(dm_bitset_set_bit);
+
+int dm_bitset_clear_bit(struct dm_disk_bitset *info, dm_block_t root,
+			uint32_t index, dm_block_t *new_root)
+{
+	int r;
+	unsigned int b = index % BITS_PER_ARRAY_ENTRY;
+
+	r = get_array_entry(info, root, index, new_root);
+	if (r)
+		return r;
+
+	clear_bit(b, (unsigned long *) &info->current_bits);
+	info->dirty = true;
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(dm_bitset_clear_bit);
+
+int dm_bitset_test_bit(struct dm_disk_bitset *info, dm_block_t root,
+		       uint32_t index, dm_block_t *new_root, bool *result)
+{
+	int r;
+	unsigned int b = index % BITS_PER_ARRAY_ENTRY;
+
+	r = get_array_entry(info, root, index, new_root);
+	if (r)
+		return r;
+
+	*result = test_bit(b, (unsigned long *) &info->current_bits);
+	return 0;
+}
+EXPORT_SYMBOL_GPL(dm_bitset_test_bit);
+
+static int cursor_next_array_entry(struct dm_bitset_cursor *c)
+{
+	int r;
+	__le64 *value;
+
+	r = dm_array_cursor_next(&c->cursor);
+	if (r)
+		return r;
+
+	dm_array_cursor_get_value(&c->cursor, (void **) &value);
+	c->array_index++;
+	c->bit_index = 0;
+	c->current_bits = le64_to_cpu(*value);
+	return 0;
+}
+
+int dm_bitset_cursor_begin(struct dm_disk_bitset *info,
+			   dm_block_t root, uint32_t nr_entries,
+			   struct dm_bitset_cursor *c)
+{
+	int r;
+	__le64 *value;
+
+	if (!nr_entries)
+		return -ENODATA;
+
+	c->info = info;
+	c->entries_remaining = nr_entries;
+
+	r = dm_array_cursor_begin(&info->array_info, root, &c->cursor);
+	if (r)
+		return r;
+
+	dm_array_cursor_get_value(&c->cursor, (void **) &value);
+	c->array_index = 0;
+	c->bit_index = 0;
+	c->current_bits = le64_to_cpu(*value);
+
+	return r;
+}
+EXPORT_SYMBOL_GPL(dm_bitset_cursor_begin);
+
+void dm_bitset_cursor_end(struct dm_bitset_cursor *c)
+{
+	return dm_array_cursor_end(&c->cursor);
+}
+EXPORT_SYMBOL_GPL(dm_bitset_cursor_end);
+
+int dm_bitset_cursor_next(struct dm_bitset_cursor *c)
+{
+	int r = 0;
+
+	if (!c->entries_remaining)
+		return -ENODATA;
+
+	c->entries_remaining--;
+	if (++c->bit_index > 63)
+		r = cursor_next_array_entry(c);
+
+	return r;
+}
+EXPORT_SYMBOL_GPL(dm_bitset_cursor_next);
+
+int dm_bitset_cursor_skip(struct dm_bitset_cursor *c, uint32_t count)
+{
+	int r;
+	__le64 *value;
+	uint32_t nr_array_skip;
+	uint32_t remaining_in_word = 64 - c->bit_index;
+
+	if (c->entries_remaining < count)
+		return -ENODATA;
+
+	if (count < remaining_in_word) {
+		c->bit_index += count;
+		c->entries_remaining -= count;
+		return 0;
+
+	} else {
+		c->entries_remaining -= remaining_in_word;
+		count -= remaining_in_word;
+	}
+
+	nr_array_skip = (count / 64) + 1;
+	r = dm_array_cursor_skip(&c->cursor, nr_array_skip);
+	if (r)
+		return r;
+
+	dm_array_cursor_get_value(&c->cursor, (void **) &value);
+	c->entries_remaining -= count;
+	c->array_index += nr_array_skip;
+	c->bit_index = count & 63;
+	c->current_bits = le64_to_cpu(*value);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(dm_bitset_cursor_skip);
+
+bool dm_bitset_cursor_get_value(struct dm_bitset_cursor *c)
+{
+	return test_bit(c->bit_index, (unsigned long *) &c->current_bits);
+}
+EXPORT_SYMBOL_GPL(dm_bitset_cursor_get_value);
+
+/*----------------------------------------------------------------*/