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, 428 insertions, 0 deletions

diff --git a/net/rds/cong.c b/net/rds/cong.c
new file mode 100644
index 000000000..f7634ce3f
--- /dev/null
+++ b/net/rds/cong.c
@@ -0,0 +1,428 @@
+/*
+ * Copyright (c) 2007, 2017 Oracle and/or its affiliates. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses.  You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ *     Redistribution and use in source and binary forms, with or
+ *     without modification, are permitted provided that the following
+ *     conditions are met:
+ *
+ *      - Redistributions of source code must retain the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer.
+ *
+ *      - Redistributions in binary form must reproduce the above
+ *        copyright notice, this list of conditions and the following
+ *        disclaimer in the documentation and/or other materials
+ *        provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ *
+ */
+#include <linux/slab.h>
+#include <linux/types.h>
+#include <linux/rbtree.h>
+#include <linux/bitops.h>
+#include <linux/export.h>
+
+#include "rds.h"
+
+/*
+ * This file implements the receive side of the unconventional congestion
+ * management in RDS.
+ *
+ * Messages waiting in the receive queue on the receiving socket are accounted
+ * against the sockets SO_RCVBUF option value.  Only the payload bytes in the
+ * message are accounted for.  If the number of bytes queued equals or exceeds
+ * rcvbuf then the socket is congested.  All sends attempted to this socket's
+ * address should return block or return -EWOULDBLOCK.
+ *
+ * Applications are expected to be reasonably tuned such that this situation
+ * very rarely occurs.  An application encountering this "back-pressure" is
+ * considered a bug.
+ *
+ * This is implemented by having each node maintain bitmaps which indicate
+ * which ports on bound addresses are congested.  As the bitmap changes it is
+ * sent through all the connections which terminate in the local address of the
+ * bitmap which changed.
+ *
+ * The bitmaps are allocated as connections are brought up.  This avoids
+ * allocation in the interrupt handling path which queues messages on sockets.
+ * The dense bitmaps let transports send the entire bitmap on any bitmap change
+ * reasonably efficiently.  This is much easier to implement than some
+ * finer-grained communication of per-port congestion.  The sender does a very
+ * inexpensive bit test to test if the port it's about to send to is congested
+ * or not.
+ */
+
+/*
+ * Interaction with poll is a tad tricky. We want all processes stuck in
+ * poll to wake up and check whether a congested destination became uncongested.
+ * The really sad thing is we have no idea which destinations the application
+ * wants to send to - we don't even know which rds_connections are involved.
+ * So until we implement a more flexible rds poll interface, we have to make
+ * do with this:
+ * We maintain a global counter that is incremented each time a congestion map
+ * update is received. Each rds socket tracks this value, and if rds_poll
+ * finds that the saved generation number is smaller than the global generation
+ * number, it wakes up the process.
+ */
+static atomic_t		rds_cong_generation = ATOMIC_INIT(0);
+
+/*
+ * Congestion monitoring
+ */
+static LIST_HEAD(rds_cong_monitor);
+static DEFINE_RWLOCK(rds_cong_monitor_lock);
+
+/*
+ * Yes, a global lock.  It's used so infrequently that it's worth keeping it
+ * global to simplify the locking.  It's only used in the following
+ * circumstances:
+ *
+ *  - on connection buildup to associate a conn with its maps
+ *  - on map changes to inform conns of a new map to send
+ *
+ *  It's sadly ordered under the socket callback lock and the connection lock.
+ *  Receive paths can mark ports congested from interrupt context so the
+ *  lock masks interrupts.
+ */
+static DEFINE_SPINLOCK(rds_cong_lock);
+static struct rb_root rds_cong_tree = RB_ROOT;
+
+static struct rds_cong_map *rds_cong_tree_walk(const struct in6_addr *addr,
+					       struct rds_cong_map *insert)
+{
+	struct rb_node **p = &rds_cong_tree.rb_node;
+	struct rb_node *parent = NULL;
+	struct rds_cong_map *map;
+
+	while (*p) {
+		int diff;
+
+		parent = *p;
+		map = rb_entry(parent, struct rds_cong_map, m_rb_node);
+
+		diff = rds_addr_cmp(addr, &map->m_addr);
+		if (diff < 0)
+			p = &(*p)->rb_left;
+		else if (diff > 0)
+			p = &(*p)->rb_right;
+		else
+			return map;
+	}
+
+	if (insert) {
+		rb_link_node(&insert->m_rb_node, parent, p);
+		rb_insert_color(&insert->m_rb_node, &rds_cong_tree);
+	}
+	return NULL;
+}
+
+/*
+ * There is only ever one bitmap for any address.  Connections try and allocate
+ * these bitmaps in the process getting pointers to them.  The bitmaps are only
+ * ever freed as the module is removed after all connections have been freed.
+ */
+static struct rds_cong_map *rds_cong_from_addr(const struct in6_addr *addr)
+{
+	struct rds_cong_map *map;
+	struct rds_cong_map *ret = NULL;
+	unsigned long zp;
+	unsigned long i;
+	unsigned long flags;
+
+	map = kzalloc_obj(struct rds_cong_map);
+	if (!map)
+		return NULL;
+
+	map->m_addr = *addr;
+	init_waitqueue_head(&map->m_waitq);
+	INIT_LIST_HEAD(&map->m_conn_list);
+
+	for (i = 0; i < RDS_CONG_MAP_PAGES; i++) {
+		zp = get_zeroed_page(GFP_KERNEL);
+		if (zp == 0)
+			goto out;
+		map->m_page_addrs[i] = zp;
+	}
+
+	spin_lock_irqsave(&rds_cong_lock, flags);
+	ret = rds_cong_tree_walk(addr, map);
+	spin_unlock_irqrestore(&rds_cong_lock, flags);
+
+	if (!ret) {
+		ret = map;
+		map = NULL;
+	}
+
+out:
+	if (map) {
+		for (i = 0; i < RDS_CONG_MAP_PAGES && map->m_page_addrs[i]; i++)
+			free_page(map->m_page_addrs[i]);
+		kfree(map);
+	}
+
+	rdsdebug("map %p for addr %pI6c\n", ret, addr);
+
+	return ret;
+}
+
+/*
+ * Put the conn on its local map's list.  This is called when the conn is
+ * really added to the hash.  It's nested under the rds_conn_lock, sadly.
+ */
+void rds_cong_add_conn(struct rds_connection *conn)
+{
+	unsigned long flags;
+
+	rdsdebug("conn %p now on map %p\n", conn, conn->c_lcong);
+	spin_lock_irqsave(&rds_cong_lock, flags);
+	list_add_tail(&conn->c_map_item, &conn->c_lcong->m_conn_list);
+	spin_unlock_irqrestore(&rds_cong_lock, flags);
+}
+
+void rds_cong_remove_conn(struct rds_connection *conn)
+{
+	unsigned long flags;
+
+	rdsdebug("removing conn %p from map %p\n", conn, conn->c_lcong);
+	spin_lock_irqsave(&rds_cong_lock, flags);
+	list_del_init(&conn->c_map_item);
+	spin_unlock_irqrestore(&rds_cong_lock, flags);
+}
+
+int rds_cong_get_maps(struct rds_connection *conn)
+{
+	conn->c_lcong = rds_cong_from_addr(&conn->c_laddr);
+	conn->c_fcong = rds_cong_from_addr(&conn->c_faddr);
+
+	if (!(conn->c_lcong && conn->c_fcong))
+		return -ENOMEM;
+
+	return 0;
+}
+
+void rds_cong_queue_updates(struct rds_cong_map *map)
+{
+	struct rds_connection *conn;
+	unsigned long flags;
+
+	spin_lock_irqsave(&rds_cong_lock, flags);
+
+	list_for_each_entry(conn, &map->m_conn_list, c_map_item) {
+		struct rds_conn_path *cp = &conn->c_path[0];
+
+		rcu_read_lock();
+		if (!test_and_set_bit(0, &conn->c_map_queued) &&
+		    !rds_destroy_pending(cp->cp_conn)) {
+			rds_stats_inc(s_cong_update_queued);
+			/* We cannot inline the call to rds_send_xmit() here
+			 * for two reasons (both pertaining to a TCP transport):
+			 * 1. When we get here from the receive path, we
+			 *    are already holding the sock_lock (held by
+			 *    tcp_v4_rcv()). So inlining calls to
+			 *    tcp_setsockopt and/or tcp_sendmsg will deadlock
+			 *    when it tries to get the sock_lock())
+			 * 2. Interrupts are masked so that we can mark the
+			 *    port congested from both send and recv paths.
+			 *    (See comment around declaration of rdc_cong_lock).
+			 *    An attempt to get the sock_lock() here will
+			 *    therefore trigger warnings.
+			 * Defer the xmit to rds_send_worker() instead.
+			 */
+			queue_delayed_work(cp->cp_wq, &cp->cp_send_w, 0);
+		}
+		rcu_read_unlock();
+	}
+
+	spin_unlock_irqrestore(&rds_cong_lock, flags);
+}
+
+void rds_cong_map_updated(struct rds_cong_map *map, uint64_t portmask)
+{
+	rdsdebug("waking map %p for %pI4\n",
+	  map, &map->m_addr);
+	rds_stats_inc(s_cong_update_received);
+	atomic_inc(&rds_cong_generation);
+	if (wq_has_sleeper(&map->m_waitq))
+		wake_up(&map->m_waitq);
+	if (wq_has_sleeper(&rds_poll_waitq))
+		wake_up_all(&rds_poll_waitq);
+
+	if (portmask && !list_empty(&rds_cong_monitor)) {
+		unsigned long flags;
+		struct rds_sock *rs;
+
+		read_lock_irqsave(&rds_cong_monitor_lock, flags);
+		list_for_each_entry(rs, &rds_cong_monitor, rs_cong_list) {
+			spin_lock(&rs->rs_lock);
+			rs->rs_cong_notify |= (rs->rs_cong_mask & portmask);
+			rs->rs_cong_mask &= ~portmask;
+			spin_unlock(&rs->rs_lock);
+			if (rs->rs_cong_notify)
+				rds_wake_sk_sleep(rs);
+		}
+		read_unlock_irqrestore(&rds_cong_monitor_lock, flags);
+	}
+}
+EXPORT_SYMBOL_GPL(rds_cong_map_updated);
+
+int rds_cong_updated_since(unsigned long *recent)
+{
+	unsigned long gen = atomic_read(&rds_cong_generation);
+
+	if (likely(*recent == gen))
+		return 0;
+	*recent = gen;
+	return 1;
+}
+
+/*
+ * We're called under the locking that protects the sockets receive buffer
+ * consumption.  This makes it a lot easier for the caller to only call us
+ * when it knows that an existing set bit needs to be cleared, and vice versa.
+ * We can't block and we need to deal with concurrent sockets working against
+ * the same per-address map.
+ */
+void rds_cong_set_bit(struct rds_cong_map *map, __be16 port)
+{
+	unsigned long i;
+	unsigned long off;
+
+	rdsdebug("setting congestion for %pI4:%u in map %p\n",
+	  &map->m_addr, ntohs(port), map);
+
+	i = be16_to_cpu(port) / RDS_CONG_MAP_PAGE_BITS;
+	off = be16_to_cpu(port) % RDS_CONG_MAP_PAGE_BITS;
+
+	set_bit_le(off, (void *)map->m_page_addrs[i]);
+}
+
+void rds_cong_clear_bit(struct rds_cong_map *map, __be16 port)
+{
+	unsigned long i;
+	unsigned long off;
+
+	rdsdebug("clearing congestion for %pI4:%u in map %p\n",
+	  &map->m_addr, ntohs(port), map);
+
+	i = be16_to_cpu(port) / RDS_CONG_MAP_PAGE_BITS;
+	off = be16_to_cpu(port) % RDS_CONG_MAP_PAGE_BITS;
+
+	clear_bit_le(off, (void *)map->m_page_addrs[i]);
+}
+
+static int rds_cong_test_bit(struct rds_cong_map *map, __be16 port)
+{
+	unsigned long i;
+	unsigned long off;
+
+	i = be16_to_cpu(port) / RDS_CONG_MAP_PAGE_BITS;
+	off = be16_to_cpu(port) % RDS_CONG_MAP_PAGE_BITS;
+
+	return test_bit_le(off, (void *)map->m_page_addrs[i]);
+}
+
+void rds_cong_add_socket(struct rds_sock *rs)
+{
+	unsigned long flags;
+
+	write_lock_irqsave(&rds_cong_monitor_lock, flags);
+	if (list_empty(&rs->rs_cong_list))
+		list_add(&rs->rs_cong_list, &rds_cong_monitor);
+	write_unlock_irqrestore(&rds_cong_monitor_lock, flags);
+}
+
+void rds_cong_remove_socket(struct rds_sock *rs)
+{
+	unsigned long flags;
+	struct rds_cong_map *map;
+
+	write_lock_irqsave(&rds_cong_monitor_lock, flags);
+	list_del_init(&rs->rs_cong_list);
+	write_unlock_irqrestore(&rds_cong_monitor_lock, flags);
+
+	/* update congestion map for now-closed port */
+	spin_lock_irqsave(&rds_cong_lock, flags);
+	map = rds_cong_tree_walk(&rs->rs_bound_addr, NULL);
+	spin_unlock_irqrestore(&rds_cong_lock, flags);
+
+	if (map && rds_cong_test_bit(map, rs->rs_bound_port)) {
+		rds_cong_clear_bit(map, rs->rs_bound_port);
+		rds_cong_queue_updates(map);
+	}
+}
+
+int rds_cong_wait(struct rds_cong_map *map, __be16 port, int nonblock,
+		  struct rds_sock *rs)
+{
+	if (!rds_cong_test_bit(map, port))
+		return 0;
+	if (nonblock) {
+		if (rs && rs->rs_cong_monitor) {
+			unsigned long flags;
+
+			/* It would have been nice to have an atomic set_bit on
+			 * a uint64_t. */
+			spin_lock_irqsave(&rs->rs_lock, flags);
+			rs->rs_cong_mask |= RDS_CONG_MONITOR_MASK(ntohs(port));
+			spin_unlock_irqrestore(&rs->rs_lock, flags);
+
+			/* Test again - a congestion update may have arrived in
+			 * the meantime. */
+			if (!rds_cong_test_bit(map, port))
+				return 0;
+		}
+		rds_stats_inc(s_cong_send_error);
+		return -ENOBUFS;
+	}
+
+	rds_stats_inc(s_cong_send_blocked);
+	rdsdebug("waiting on map %p for port %u\n", map, be16_to_cpu(port));
+
+	return wait_event_interruptible(map->m_waitq,
+					!rds_cong_test_bit(map, port));
+}
+
+void rds_cong_exit(void)
+{
+	struct rb_node *node;
+	struct rds_cong_map *map;
+	unsigned long i;
+
+	while ((node = rb_first(&rds_cong_tree))) {
+		map = rb_entry(node, struct rds_cong_map, m_rb_node);
+		rdsdebug("freeing map %p\n", map);
+		rb_erase(&map->m_rb_node, &rds_cong_tree);
+		for (i = 0; i < RDS_CONG_MAP_PAGES && map->m_page_addrs[i]; i++)
+			free_page(map->m_page_addrs[i]);
+		kfree(map);
+	}
+}
+
+/*
+ * Allocate a RDS message containing a congestion update.
+ */
+struct rds_message *rds_cong_update_alloc(struct rds_connection *conn)
+{
+	struct rds_cong_map *map = conn->c_lcong;
+	struct rds_message *rm;
+
+	rm = rds_message_map_pages(map->m_page_addrs, RDS_CONG_MAP_BYTES);
+	if (!IS_ERR(rm))
+		rm->m_inc.i_hdr.h_flags = RDS_FLAG_CONG_BITMAP;
+
+	return rm;
+}