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, 74 insertions, 0 deletions

diff --git a/arch/x86/boot/setup.ld b/arch/x86/boot/setup.ld
new file mode 100644
index 000000000..e1d594a60
--- /dev/null
+++ b/arch/x86/boot/setup.ld
@@ -0,0 +1,74 @@
+/*
+ * setup.ld
+ *
+ * Linker script for the i386 setup code
+ */
+OUTPUT_FORMAT("elf32-i386")
+OUTPUT_ARCH(i386)
+ENTRY(_start)
+
+SECTIONS
+{
+	. = 0;
+	.bstext	: {
+		*(.bstext)
+		. = 495;
+	} =0xffffffff
+
+	.header		: { *(.header) }
+	.entrytext	: { *(.entrytext) }
+	.inittext	: { *(.inittext) }
+	.initdata	: { *(.initdata) }
+	__end_init = .;
+
+	.text		: { *(.text .text.*) }
+	.text32		: { *(.text32) }
+
+	.pecompat	: { *(.pecompat) }
+	PROVIDE(pecompat_fsize = setup_size - pecompat_fstart);
+
+	. = ALIGN(16);
+	.rodata		: { *(.rodata*) }
+
+	.videocards	: {
+		video_cards = .;
+		*(.videocards)
+		video_cards_end = .;
+	}
+
+	. = ALIGN(16);
+	.data		: { *(.data*) }
+
+	.signature	: {
+		setup_sig = .;
+		LONG(0x5a5aaa55)
+
+		setup_size = ALIGN(ABSOLUTE(.), 4096);
+		setup_sects = ABSOLUTE(setup_size / 512);
+		ASSERT(setup_sects >= 5, "The setup must be at least 5 sectors in size");
+		ASSERT(setup_sects <= 64, "The setup must be at most 64 sectors in size");
+	}
+
+	. = ALIGN(16);
+	.bss		:
+	{
+		__bss_start = .;
+		*(.bss)
+		__bss_end = .;
+	}
+	. = ALIGN(16);
+	_end = .;
+
+	/DISCARD/	: {
+		*(.note*)
+	}
+
+	/*
+	 * The ASSERT() sink to . is intentional, for binutils 2.14 compatibility:
+	 */
+	. = ASSERT(_end <= 0x8000, "Setup too big!");
+	. = ASSERT(hdr == 0x1f1, "The setup header has the wrong offset!");
+	/* Necessary for the very-old-loader check to work... */
+	. = ASSERT(__end_init <= 5*512, "init sections too big!");
+
+}