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, 381 insertions, 0 deletions

diff --git a/net/ieee802154/header_ops.c b/net/ieee802154/header_ops.c
new file mode 100644
index 000000000..a9f0c8df5
--- /dev/null
+++ b/net/ieee802154/header_ops.c
@@ -0,0 +1,381 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Copyright (C) 2014 Fraunhofer ITWM
+ *
+ * Written by:
+ * Phoebe Buckheister <phoebe.buckheister@itwm.fraunhofer.de>
+ */
+
+#include <linux/ieee802154.h>
+
+#include <net/mac802154.h>
+#include <net/ieee802154_netdev.h>
+
+static int
+ieee802154_hdr_push_addr(u8 *buf, const struct ieee802154_addr *addr,
+			 bool omit_pan)
+{
+	int pos = 0;
+
+	if (addr->mode == IEEE802154_ADDR_NONE)
+		return 0;
+
+	if (!omit_pan) {
+		memcpy(buf + pos, &addr->pan_id, 2);
+		pos += 2;
+	}
+
+	switch (addr->mode) {
+	case IEEE802154_ADDR_SHORT:
+		memcpy(buf + pos, &addr->short_addr, 2);
+		pos += 2;
+		break;
+
+	case IEEE802154_ADDR_LONG:
+		memcpy(buf + pos, &addr->extended_addr, IEEE802154_ADDR_LEN);
+		pos += IEEE802154_ADDR_LEN;
+		break;
+
+	default:
+		return -EINVAL;
+	}
+
+	return pos;
+}
+
+static int
+ieee802154_hdr_push_sechdr(u8 *buf, const struct ieee802154_sechdr *hdr)
+{
+	int pos = 5;
+
+	memcpy(buf, hdr, 1);
+	memcpy(buf + 1, &hdr->frame_counter, 4);
+
+	switch (hdr->key_id_mode) {
+	case IEEE802154_SCF_KEY_IMPLICIT:
+		return pos;
+
+	case IEEE802154_SCF_KEY_INDEX:
+		break;
+
+	case IEEE802154_SCF_KEY_SHORT_INDEX:
+		memcpy(buf + pos, &hdr->short_src, 4);
+		pos += 4;
+		break;
+
+	case IEEE802154_SCF_KEY_HW_INDEX:
+		memcpy(buf + pos, &hdr->extended_src, IEEE802154_ADDR_LEN);
+		pos += IEEE802154_ADDR_LEN;
+		break;
+	}
+
+	buf[pos++] = hdr->key_id;
+
+	return pos;
+}
+
+int
+ieee802154_hdr_push(struct sk_buff *skb, struct ieee802154_hdr *hdr)
+{
+	u8 buf[IEEE802154_MAX_HEADER_LEN];
+	int pos = 2;
+	int rc;
+	struct ieee802154_hdr_fc *fc = &hdr->fc;
+
+	buf[pos++] = hdr->seq;
+
+	fc->dest_addr_mode = hdr->dest.mode;
+
+	rc = ieee802154_hdr_push_addr(buf + pos, &hdr->dest, false);
+	if (rc < 0)
+		return -EINVAL;
+	pos += rc;
+
+	fc->source_addr_mode = hdr->source.mode;
+
+	if (hdr->source.pan_id == hdr->dest.pan_id &&
+	    hdr->dest.mode != IEEE802154_ADDR_NONE)
+		fc->intra_pan = true;
+
+	rc = ieee802154_hdr_push_addr(buf + pos, &hdr->source, fc->intra_pan);
+	if (rc < 0)
+		return -EINVAL;
+	pos += rc;
+
+	if (fc->security_enabled) {
+		fc->version = 1;
+
+		rc = ieee802154_hdr_push_sechdr(buf + pos, &hdr->sec);
+		if (rc < 0)
+			return -EINVAL;
+
+		pos += rc;
+	}
+
+	memcpy(buf, fc, 2);
+
+	memcpy(skb_push(skb, pos), buf, pos);
+
+	return pos;
+}
+EXPORT_SYMBOL_GPL(ieee802154_hdr_push);
+
+int ieee802154_mac_cmd_push(struct sk_buff *skb, void *f,
+			    const void *pl, unsigned int pl_len)
+{
+	struct ieee802154_mac_cmd_frame *frame = f;
+	struct ieee802154_mac_cmd_pl *mac_pl = &frame->mac_pl;
+	struct ieee802154_hdr *mhr = &frame->mhr;
+	int ret;
+
+	skb_reserve(skb, sizeof(*mhr));
+	ret = ieee802154_hdr_push(skb, mhr);
+	if (ret < 0)
+		return ret;
+
+	skb_reset_mac_header(skb);
+	skb->mac_len = ret;
+
+	skb_put_data(skb, mac_pl, sizeof(*mac_pl));
+	skb_put_data(skb, pl, pl_len);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ieee802154_mac_cmd_push);
+
+int ieee802154_beacon_push(struct sk_buff *skb,
+			   struct ieee802154_beacon_frame *beacon)
+{
+	struct ieee802154_beacon_hdr *mac_pl = &beacon->mac_pl;
+	struct ieee802154_hdr *mhr = &beacon->mhr;
+	int ret;
+
+	skb_reserve(skb, sizeof(*mhr));
+	ret = ieee802154_hdr_push(skb, mhr);
+	if (ret < 0)
+		return ret;
+
+	skb_reset_mac_header(skb);
+	skb->mac_len = ret;
+
+	skb_put_data(skb, mac_pl, sizeof(*mac_pl));
+
+	if (mac_pl->pend_short_addr_count || mac_pl->pend_ext_addr_count)
+		return -EOPNOTSUPP;
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ieee802154_beacon_push);
+
+static int
+ieee802154_hdr_get_addr(const u8 *buf, int mode, bool omit_pan,
+			struct ieee802154_addr *addr)
+{
+	int pos = 0;
+
+	if (mode == IEEE802154_ADDR_NONE) {
+		memset(addr, 0, sizeof(*addr));
+		addr->mode = IEEE802154_ADDR_NONE;
+		return 0;
+	}
+
+	addr->mode = mode;
+
+	if (!omit_pan) {
+		memcpy(&addr->pan_id, buf + pos, 2);
+		pos += 2;
+	}
+
+	if (mode == IEEE802154_ADDR_SHORT) {
+		memcpy(&addr->short_addr, buf + pos, 2);
+		return pos + 2;
+	} else {
+		memcpy(&addr->extended_addr, buf + pos, IEEE802154_ADDR_LEN);
+		return pos + IEEE802154_ADDR_LEN;
+	}
+}
+
+static int ieee802154_hdr_addr_len(int mode, bool omit_pan)
+{
+	int pan_len = omit_pan ? 0 : 2;
+
+	switch (mode) {
+	case IEEE802154_ADDR_NONE: return 0;
+	case IEEE802154_ADDR_SHORT: return 2 + pan_len;
+	case IEEE802154_ADDR_LONG: return IEEE802154_ADDR_LEN + pan_len;
+	default: return -EINVAL;
+	}
+}
+
+static int
+ieee802154_hdr_get_sechdr(const u8 *buf, struct ieee802154_sechdr *hdr)
+{
+	int pos = 5;
+
+	memcpy(hdr, buf, 1);
+	memcpy(&hdr->frame_counter, buf + 1, 4);
+
+	switch (hdr->key_id_mode) {
+	case IEEE802154_SCF_KEY_IMPLICIT:
+		return pos;
+
+	case IEEE802154_SCF_KEY_INDEX:
+		break;
+
+	case IEEE802154_SCF_KEY_SHORT_INDEX:
+		memcpy(&hdr->short_src, buf + pos, 4);
+		pos += 4;
+		break;
+
+	case IEEE802154_SCF_KEY_HW_INDEX:
+		memcpy(&hdr->extended_src, buf + pos, IEEE802154_ADDR_LEN);
+		pos += IEEE802154_ADDR_LEN;
+		break;
+	}
+
+	hdr->key_id = buf[pos++];
+
+	return pos;
+}
+
+static int ieee802154_sechdr_lengths[4] = {
+	[IEEE802154_SCF_KEY_IMPLICIT] = 5,
+	[IEEE802154_SCF_KEY_INDEX] = 6,
+	[IEEE802154_SCF_KEY_SHORT_INDEX] = 10,
+	[IEEE802154_SCF_KEY_HW_INDEX] = 14,
+};
+
+static int ieee802154_hdr_sechdr_len(u8 sc)
+{
+	return ieee802154_sechdr_lengths[IEEE802154_SCF_KEY_ID_MODE(sc)];
+}
+
+static int ieee802154_hdr_minlen(const struct ieee802154_hdr *hdr)
+{
+	int dlen, slen;
+
+	dlen = ieee802154_hdr_addr_len(hdr->fc.dest_addr_mode, false);
+	slen = ieee802154_hdr_addr_len(hdr->fc.source_addr_mode,
+				       hdr->fc.intra_pan);
+
+	if (slen < 0 || dlen < 0)
+		return -EINVAL;
+
+	return 3 + dlen + slen + hdr->fc.security_enabled;
+}
+
+static int
+ieee802154_hdr_get_addrs(const u8 *buf, struct ieee802154_hdr *hdr)
+{
+	int pos = 0;
+
+	pos += ieee802154_hdr_get_addr(buf + pos, hdr->fc.dest_addr_mode,
+				       false, &hdr->dest);
+	pos += ieee802154_hdr_get_addr(buf + pos, hdr->fc.source_addr_mode,
+				       hdr->fc.intra_pan, &hdr->source);
+
+	if (hdr->fc.intra_pan)
+		hdr->source.pan_id = hdr->dest.pan_id;
+
+	return pos;
+}
+
+int
+ieee802154_hdr_pull(struct sk_buff *skb, struct ieee802154_hdr *hdr)
+{
+	int pos = 3, rc;
+
+	if (!pskb_may_pull(skb, 3))
+		return -EINVAL;
+
+	memcpy(hdr, skb->data, 3);
+
+	rc = ieee802154_hdr_minlen(hdr);
+	if (rc < 0 || !pskb_may_pull(skb, rc))
+		return -EINVAL;
+
+	pos += ieee802154_hdr_get_addrs(skb->data + pos, hdr);
+
+	if (hdr->fc.security_enabled) {
+		int want = pos + ieee802154_hdr_sechdr_len(skb->data[pos]);
+
+		if (!pskb_may_pull(skb, want))
+			return -EINVAL;
+
+		pos += ieee802154_hdr_get_sechdr(skb->data + pos, &hdr->sec);
+	}
+
+	skb_pull(skb, pos);
+	return pos;
+}
+EXPORT_SYMBOL_GPL(ieee802154_hdr_pull);
+
+int ieee802154_mac_cmd_pl_pull(struct sk_buff *skb,
+			       struct ieee802154_mac_cmd_pl *mac_pl)
+{
+	if (!pskb_may_pull(skb, sizeof(*mac_pl)))
+		return -EINVAL;
+
+	memcpy(mac_pl, skb->data, sizeof(*mac_pl));
+	skb_pull(skb, sizeof(*mac_pl));
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(ieee802154_mac_cmd_pl_pull);
+
+int
+ieee802154_hdr_peek_addrs(const struct sk_buff *skb, struct ieee802154_hdr *hdr)
+{
+	const u8 *buf = skb_mac_header(skb);
+	int pos = 3, rc;
+
+	if (buf + 3 > skb_tail_pointer(skb))
+		return -EINVAL;
+
+	memcpy(hdr, buf, 3);
+
+	rc = ieee802154_hdr_minlen(hdr);
+	if (rc < 0 || buf + rc > skb_tail_pointer(skb))
+		return -EINVAL;
+
+	pos += ieee802154_hdr_get_addrs(buf + pos, hdr);
+	return pos;
+}
+EXPORT_SYMBOL_GPL(ieee802154_hdr_peek_addrs);
+
+int
+ieee802154_hdr_peek(const struct sk_buff *skb, struct ieee802154_hdr *hdr)
+{
+	const u8 *buf = skb_mac_header(skb);
+	int pos;
+
+	pos = ieee802154_hdr_peek_addrs(skb, hdr);
+	if (pos < 0)
+		return -EINVAL;
+
+	if (hdr->fc.security_enabled) {
+		u8 key_id_mode = IEEE802154_SCF_KEY_ID_MODE(*(buf + pos));
+		int want = pos + ieee802154_sechdr_lengths[key_id_mode];
+
+		if (buf + want > skb_tail_pointer(skb))
+			return -EINVAL;
+
+		pos += ieee802154_hdr_get_sechdr(buf + pos, &hdr->sec);
+	}
+
+	return pos;
+}
+EXPORT_SYMBOL_GPL(ieee802154_hdr_peek);
+
+int ieee802154_max_payload(const struct ieee802154_hdr *hdr)
+{
+	int hlen = ieee802154_hdr_minlen(hdr);
+
+	if (hdr->fc.security_enabled) {
+		hlen += ieee802154_sechdr_lengths[hdr->sec.key_id_mode] - 1;
+		hlen += ieee802154_sechdr_authtag_len(&hdr->sec);
+	}
+
+	return IEEE802154_MTU - hlen - IEEE802154_MFR_SIZE;
+}
+EXPORT_SYMBOL_GPL(ieee802154_max_payload);