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, 115 insertions, 0 deletions

diff --git a/drivers/misc/cb710/debug.c b/drivers/misc/cb710/debug.c
new file mode 100644
index 000000000..20d672edf
--- /dev/null
+++ b/drivers/misc/cb710/debug.c
@@ -0,0 +1,115 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ *  cb710/debug.c
+ *
+ *  Copyright by Michał Mirosław, 2008-2009
+ */
+#include <linux/cb710.h>
+#include <linux/kernel.h>
+#include <linux/module.h>
+
+#define CB710_REG_COUNT		0x80
+
+static const u16 allow[CB710_REG_COUNT/16] = {
+	0xFFF0, 0xFFFF, 0xFFFF, 0xFFFF,
+	0xFFF0, 0xFFFF, 0xFFFF, 0xFFFF,
+};
+static const char *const prefix[ARRAY_SIZE(allow)] = {
+	"MMC", "MMC", "MMC", "MMC",
+	"MS?", "MS?", "SM?", "SM?"
+};
+
+static inline int allow_reg_read(unsigned block, unsigned offset, unsigned bits)
+{
+	unsigned mask = (1 << bits/8) - 1;
+	offset *= bits/8;
+	return ((allow[block] >> offset) & mask) == mask;
+}
+
+#define CB710_READ_REGS_TEMPLATE(t)					\
+static void cb710_read_regs_##t(void __iomem *iobase,			\
+	u##t *reg, unsigned select)					\
+{									\
+	unsigned i, j;							\
+									\
+	for (i = 0; i < ARRAY_SIZE(allow); ++i, reg += 16/(t/8)) {	\
+		if (!(select & (1 << i)))					\
+			continue;					\
+									\
+		for (j = 0; j < 0x10/(t/8); ++j) {			\
+			if (!allow_reg_read(i, j, t))			\
+				continue;				\
+			reg[j] = ioread##t(iobase			\
+				+ (i << 4) + (j * (t/8)));		\
+		}							\
+	}								\
+}
+
+static const char cb710_regf_8[] = "%02X";
+static const char cb710_regf_16[] = "%04X";
+static const char cb710_regf_32[] = "%08X";
+static const char cb710_xes[] = "xxxxxxxx";
+
+#define CB710_DUMP_REGS_TEMPLATE(t)					\
+static void cb710_dump_regs_##t(struct device *dev,			\
+	const u##t *reg, unsigned select)				\
+{									\
+	const char *const xp = &cb710_xes[8 - t/4];			\
+	const char *const format = cb710_regf_##t;			\
+									\
+	char msg[100], *p;						\
+	unsigned i, j;							\
+									\
+	for (i = 0; i < ARRAY_SIZE(allow); ++i, reg += 16/(t/8)) {	\
+		if (!(select & (1 << i)))				\
+			continue;					\
+		p = msg;						\
+		for (j = 0; j < 0x10/(t/8); ++j) {			\
+			*p++ = ' ';					\
+			if (j == 8/(t/8))				\
+				*p++ = ' ';				\
+			if (allow_reg_read(i, j, t))			\
+				p += sprintf(p, format, reg[j]);	\
+			else						\
+				p += sprintf(p, "%s", xp);		\
+		}							\
+		dev_dbg(dev, "%s 0x%02X %s\n", prefix[i], i << 4, msg);	\
+	}								\
+}
+
+#define CB710_READ_AND_DUMP_REGS_TEMPLATE(t)				\
+static void cb710_read_and_dump_regs_##t(struct cb710_chip *chip,	\
+	unsigned select)						\
+{									\
+	u##t regs[CB710_REG_COUNT/sizeof(u##t)];			\
+									\
+	memset(&regs, 0, sizeof(regs));					\
+	cb710_read_regs_##t(chip->iobase, regs, select);		\
+	cb710_dump_regs_##t(cb710_chip_dev(chip), regs, select);	\
+}
+
+#define CB710_REG_ACCESS_TEMPLATES(t)		\
+  CB710_READ_REGS_TEMPLATE(t)			\
+  CB710_DUMP_REGS_TEMPLATE(t)			\
+  CB710_READ_AND_DUMP_REGS_TEMPLATE(t)
+
+CB710_REG_ACCESS_TEMPLATES(8)
+CB710_REG_ACCESS_TEMPLATES(16)
+CB710_REG_ACCESS_TEMPLATES(32)
+
+void cb710_dump_regs(struct cb710_chip *chip, unsigned select)
+{
+	if (!(select & CB710_DUMP_REGS_MASK))
+		select = CB710_DUMP_REGS_ALL;
+	if (!(select & CB710_DUMP_ACCESS_MASK))
+		select |= CB710_DUMP_ACCESS_8;
+
+	if (select & CB710_DUMP_ACCESS_32)
+		cb710_read_and_dump_regs_32(chip, select);
+	if (select & CB710_DUMP_ACCESS_16)
+		cb710_read_and_dump_regs_16(chip, select);
+	if (select & CB710_DUMP_ACCESS_8)
+		cb710_read_and_dump_regs_8(chip, select);
+}
+EXPORT_SYMBOL_GPL(cb710_dump_regs);
+