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, 170 insertions, 0 deletions

diff --git a/net/sctp/debug.c b/net/sctp/debug.c
new file mode 100644
index 000000000..ccd773e4c
--- /dev/null
+++ b/net/sctp/debug.c
@@ -0,0 +1,170 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/* SCTP kernel implementation
+ * (C) Copyright IBM Corp. 2001, 2004
+ * Copyright (c) 1999-2000 Cisco, Inc.
+ * Copyright (c) 1999-2001 Motorola, Inc.
+ * Copyright (c) 2001 Intel Corp.
+ *
+ * This file is part of the SCTP kernel implementation
+ *
+ * This file converts numerical ID value to alphabetical names for SCTP
+ * terms such as chunk type, parameter time, event type, etc.
+ *
+ * Please send any bug reports or fixes you make to the
+ * email address(es):
+ *    lksctp developers <linux-sctp@vger.kernel.org>
+ *
+ * Written or modified by:
+ *    La Monte H.P. Yarroll <piggy@acm.org>
+ *    Karl Knutson          <karl@athena.chicago.il.us>
+ *    Xingang Guo           <xingang.guo@intel.com>
+ *    Jon Grimm             <jgrimm@us.ibm.com>
+ *    Daisy Chang	    <daisyc@us.ibm.com>
+ *    Sridhar Samudrala	    <sri@us.ibm.com>
+ */
+
+#include <net/sctp/sctp.h>
+
+/* These are printable forms of Chunk ID's from section 3.1.  */
+static const char *const sctp_cid_tbl[SCTP_NUM_BASE_CHUNK_TYPES] = {
+	"DATA",
+	"INIT",
+	"INIT_ACK",
+	"SACK",
+	"HEARTBEAT",
+	"HEARTBEAT_ACK",
+	"ABORT",
+	"SHUTDOWN",
+	"SHUTDOWN_ACK",
+	"ERROR",
+	"COOKIE_ECHO",
+	"COOKIE_ACK",
+	"ECN_ECNE",
+	"ECN_CWR",
+	"SHUTDOWN_COMPLETE",
+};
+
+/* Lookup "chunk type" debug name. */
+const char *sctp_cname(const union sctp_subtype cid)
+{
+	if (cid.chunk <= SCTP_CID_BASE_MAX)
+		return sctp_cid_tbl[cid.chunk];
+
+	switch (cid.chunk) {
+	case SCTP_CID_ASCONF:
+		return "ASCONF";
+
+	case SCTP_CID_ASCONF_ACK:
+		return "ASCONF_ACK";
+
+	case SCTP_CID_FWD_TSN:
+		return "FWD_TSN";
+
+	case SCTP_CID_AUTH:
+		return "AUTH";
+
+	case SCTP_CID_RECONF:
+		return "RECONF";
+
+	case SCTP_CID_I_DATA:
+		return "I_DATA";
+
+	case SCTP_CID_I_FWD_TSN:
+		return "I_FWD_TSN";
+
+	default:
+		break;
+	}
+
+	return "unknown chunk";
+}
+
+/* These are printable forms of the states.  */
+const char *const sctp_state_tbl[SCTP_STATE_NUM_STATES] = {
+	"STATE_CLOSED",
+	"STATE_COOKIE_WAIT",
+	"STATE_COOKIE_ECHOED",
+	"STATE_ESTABLISHED",
+	"STATE_SHUTDOWN_PENDING",
+	"STATE_SHUTDOWN_SENT",
+	"STATE_SHUTDOWN_RECEIVED",
+	"STATE_SHUTDOWN_ACK_SENT",
+};
+
+/* Events that could change the state of an association.  */
+const char *const sctp_evttype_tbl[] = {
+	"EVENT_T_unknown",
+	"EVENT_T_CHUNK",
+	"EVENT_T_TIMEOUT",
+	"EVENT_T_OTHER",
+	"EVENT_T_PRIMITIVE"
+};
+
+/* Return value of a state function */
+const char *const sctp_status_tbl[] = {
+	"DISPOSITION_DISCARD",
+	"DISPOSITION_CONSUME",
+	"DISPOSITION_NOMEM",
+	"DISPOSITION_DELETE_TCB",
+	"DISPOSITION_ABORT",
+	"DISPOSITION_VIOLATION",
+	"DISPOSITION_NOT_IMPL",
+	"DISPOSITION_ERROR",
+	"DISPOSITION_BUG"
+};
+
+/* Printable forms of primitives */
+static const char *const sctp_primitive_tbl[SCTP_NUM_PRIMITIVE_TYPES] = {
+	"PRIMITIVE_ASSOCIATE",
+	"PRIMITIVE_SHUTDOWN",
+	"PRIMITIVE_ABORT",
+	"PRIMITIVE_SEND",
+	"PRIMITIVE_REQUESTHEARTBEAT",
+	"PRIMITIVE_ASCONF",
+};
+
+/* Lookup primitive debug name. */
+const char *sctp_pname(const union sctp_subtype id)
+{
+	if (id.primitive <= SCTP_EVENT_PRIMITIVE_MAX)
+		return sctp_primitive_tbl[id.primitive];
+	return "unknown_primitive";
+}
+
+static const char *const sctp_other_tbl[] = {
+	"NO_PENDING_TSN",
+	"ICMP_PROTO_UNREACH",
+};
+
+/* Lookup "other" debug name. */
+const char *sctp_oname(const union sctp_subtype id)
+{
+	if (id.other <= SCTP_EVENT_OTHER_MAX)
+		return sctp_other_tbl[id.other];
+	return "unknown 'other' event";
+}
+
+static const char *const sctp_timer_tbl[] = {
+	"TIMEOUT_NONE",
+	"TIMEOUT_T1_COOKIE",
+	"TIMEOUT_T1_INIT",
+	"TIMEOUT_T2_SHUTDOWN",
+	"TIMEOUT_T3_RTX",
+	"TIMEOUT_T4_RTO",
+	"TIMEOUT_T5_SHUTDOWN_GUARD",
+	"TIMEOUT_HEARTBEAT",
+	"TIMEOUT_RECONF",
+	"TIMEOUT_PROBE",
+	"TIMEOUT_SACK",
+	"TIMEOUT_AUTOCLOSE",
+};
+
+/* Lookup timer debug name. */
+const char *sctp_tname(const union sctp_subtype id)
+{
+	BUILD_BUG_ON(SCTP_EVENT_TIMEOUT_MAX + 1 != ARRAY_SIZE(sctp_timer_tbl));
+
+	if (id.timeout < ARRAY_SIZE(sctp_timer_tbl))
+		return sctp_timer_tbl[id.timeout];
+	return "unknown_timer";
+}