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, 95 insertions, 0 deletions

diff --git a/include/net/compat.h b/include/net/compat.h
new file mode 100644
index 000000000..84c163f40
--- /dev/null
+++ b/include/net/compat.h
@@ -0,0 +1,95 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef NET_COMPAT_H
+#define NET_COMPAT_H
+
+
+struct sock;
+
+#include <linux/compat.h>
+
+struct compat_msghdr {
+	compat_uptr_t	msg_name;	/* void * */
+	compat_int_t	msg_namelen;
+	compat_uptr_t	msg_iov;	/* struct compat_iovec * */
+	compat_size_t	msg_iovlen;
+	compat_uptr_t	msg_control;	/* void * */
+	compat_size_t	msg_controllen;
+	compat_uint_t	msg_flags;
+};
+
+struct compat_mmsghdr {
+	struct compat_msghdr msg_hdr;
+	compat_uint_t        msg_len;
+};
+
+struct compat_cmsghdr {
+	compat_size_t	cmsg_len;
+	compat_int_t	cmsg_level;
+	compat_int_t	cmsg_type;
+};
+
+struct compat_rtentry {
+	u32		rt_pad1;
+	struct sockaddr rt_dst;         /* target address               */
+	struct sockaddr rt_gateway;     /* gateway addr (RTF_GATEWAY)   */
+	struct sockaddr rt_genmask;     /* target network mask (IP)     */
+	unsigned short	rt_flags;
+	short		rt_pad2;
+	u32		rt_pad3;
+	unsigned char	rt_tos;
+	unsigned char	rt_class;
+	short		rt_pad4;
+	short		rt_metric;      /* +1 for binary compatibility! */
+	compat_uptr_t	rt_dev;         /* forcing the device at add    */
+	u32		rt_mtu;         /* per route MTU/Window         */
+	u32		rt_window;      /* Window clamping              */
+	unsigned short  rt_irtt;        /* Initial RTT                  */
+};
+
+int __get_compat_msghdr(struct msghdr *kmsg, struct compat_msghdr *msg,
+			struct sockaddr __user **save_addr);
+int get_compat_msghdr(struct msghdr *, struct compat_msghdr __user *,
+		      struct sockaddr __user **, struct iovec **);
+int put_cmsg_compat(struct msghdr*, int, int, int, void *);
+
+int cmsghdr_from_user_compat_to_kern(struct msghdr *, struct sock *,
+				     unsigned char *, int);
+
+struct compat_group_req {
+	__u32				 gr_interface;
+	struct __kernel_sockaddr_storage gr_group
+		__aligned(4);
+} __packed;
+
+struct compat_group_source_req {
+	__u32				 gsr_interface;
+	struct __kernel_sockaddr_storage gsr_group
+		__aligned(4);
+	struct __kernel_sockaddr_storage gsr_source
+		__aligned(4);
+} __packed;
+
+struct compat_group_filter {
+	union {
+		struct {
+			__u32				 gf_interface_aux;
+			struct __kernel_sockaddr_storage gf_group_aux
+				__aligned(4);
+			__u32				 gf_fmode_aux;
+			__u32				 gf_numsrc_aux;
+			struct __kernel_sockaddr_storage gf_slist[1]
+				__aligned(4);
+		} __packed;
+		struct {
+			__u32				 gf_interface;
+			struct __kernel_sockaddr_storage gf_group
+				__aligned(4);
+			__u32				 gf_fmode;
+			__u32				 gf_numsrc;
+			struct __kernel_sockaddr_storage gf_slist_flex[]
+				__aligned(4);
+		} __packed;
+	};
+} __packed;
+
+#endif /* NET_COMPAT_H */