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, 134 insertions, 0 deletions

diff --git a/drivers/mfd/ac100.c b/drivers/mfd/ac100.c
new file mode 100644
index 000000000..8f47c392c
--- /dev/null
+++ b/drivers/mfd/ac100.c
@@ -0,0 +1,134 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * MFD core driver for X-Powers' AC100 Audio Codec IC
+ *
+ * The AC100 is a highly integrated audio codec and RTC subsystem designed
+ * for mobile applications. It has 3 I2S/PCM interfaces, a 2 channel DAC,
+ * a 2 channel ADC with 5 inputs and a builtin mixer. The RTC subsystem has
+ * 3 clock outputs.
+ *
+ * The audio codec and RTC parts are completely separate, sharing only the
+ * host interface for access to its registers.
+ *
+ * Copyright (2016) Chen-Yu Tsai
+ *
+ * Author: Chen-Yu Tsai <wens@csie.org>
+ */
+
+#include <linux/interrupt.h>
+#include <linux/kernel.h>
+#include <linux/mfd/core.h>
+#include <linux/mfd/ac100.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/regmap.h>
+#include <linux/sunxi-rsb.h>
+
+static const struct regmap_range ac100_writeable_ranges[] = {
+	regmap_reg_range(AC100_CHIP_AUDIO_RST, AC100_I2S_SR_CTRL),
+	regmap_reg_range(AC100_I2S1_CLK_CTRL, AC100_I2S1_MXR_GAIN),
+	regmap_reg_range(AC100_I2S2_CLK_CTRL, AC100_I2S2_MXR_GAIN),
+	regmap_reg_range(AC100_I2S3_CLK_CTRL, AC100_I2S3_SIG_PATH_CTRL),
+	regmap_reg_range(AC100_ADC_DIG_CTRL, AC100_ADC_VOL_CTRL),
+	regmap_reg_range(AC100_HMIC_CTRL1, AC100_HMIC_STATUS),
+	regmap_reg_range(AC100_DAC_DIG_CTRL, AC100_DAC_MXR_GAIN),
+	regmap_reg_range(AC100_ADC_APC_CTRL, AC100_LINEOUT_CTRL),
+	regmap_reg_range(AC100_ADC_DAP_L_CTRL, AC100_ADC_DAP_OPT),
+	regmap_reg_range(AC100_DAC_DAP_CTRL, AC100_DAC_DAP_OPT),
+	regmap_reg_range(AC100_ADC_DAP_ENA, AC100_DAC_DAP_ENA),
+	regmap_reg_range(AC100_SRC1_CTRL1, AC100_SRC1_CTRL2),
+	regmap_reg_range(AC100_SRC2_CTRL1, AC100_SRC2_CTRL2),
+	regmap_reg_range(AC100_CLK32K_ANALOG_CTRL, AC100_CLKOUT_CTRL3),
+	regmap_reg_range(AC100_RTC_RST, AC100_RTC_UPD),
+	regmap_reg_range(AC100_ALM_INT_ENA, AC100_ALM_INT_STA),
+	regmap_reg_range(AC100_ALM_SEC, AC100_RTC_GP(15)),
+};
+
+static const struct regmap_range ac100_volatile_ranges[] = {
+	regmap_reg_range(AC100_CHIP_AUDIO_RST, AC100_PLL_CTRL2),
+	regmap_reg_range(AC100_HMIC_STATUS, AC100_HMIC_STATUS),
+	regmap_reg_range(AC100_ADC_DAP_L_STA, AC100_ADC_DAP_L_STA),
+	regmap_reg_range(AC100_SRC1_CTRL1, AC100_SRC1_CTRL1),
+	regmap_reg_range(AC100_SRC1_CTRL3, AC100_SRC2_CTRL1),
+	regmap_reg_range(AC100_SRC2_CTRL3, AC100_SRC2_CTRL4),
+	regmap_reg_range(AC100_RTC_RST, AC100_RTC_RST),
+	regmap_reg_range(AC100_RTC_SEC, AC100_ALM_INT_STA),
+	regmap_reg_range(AC100_ALM_SEC, AC100_ALM_UPD),
+};
+
+static const struct regmap_access_table ac100_writeable_table = {
+	.yes_ranges	= ac100_writeable_ranges,
+	.n_yes_ranges	= ARRAY_SIZE(ac100_writeable_ranges),
+};
+
+static const struct regmap_access_table ac100_volatile_table = {
+	.yes_ranges	= ac100_volatile_ranges,
+	.n_yes_ranges	= ARRAY_SIZE(ac100_volatile_ranges),
+};
+
+static const struct regmap_config ac100_regmap_config = {
+	.reg_bits	= 8,
+	.val_bits	= 16,
+	.wr_table	= &ac100_writeable_table,
+	.volatile_table	= &ac100_volatile_table,
+	.max_register	= AC100_RTC_GP(15),
+	.cache_type	= REGCACHE_MAPLE,
+};
+
+static struct mfd_cell ac100_cells[] = {
+	{
+		.name		= "ac100-codec",
+		.of_compatible	= "x-powers,ac100-codec",
+	}, {
+		.name		= "ac100-rtc",
+		.of_compatible	= "x-powers,ac100-rtc",
+	},
+};
+
+static int ac100_rsb_probe(struct sunxi_rsb_device *rdev)
+{
+	struct ac100_dev *ac100;
+	int ret;
+
+	ac100 = devm_kzalloc(&rdev->dev, sizeof(*ac100), GFP_KERNEL);
+	if (!ac100)
+		return -ENOMEM;
+
+	ac100->dev = &rdev->dev;
+	sunxi_rsb_device_set_drvdata(rdev, ac100);
+
+	ac100->regmap = devm_regmap_init_sunxi_rsb(rdev, &ac100_regmap_config);
+	if (IS_ERR(ac100->regmap)) {
+		ret = PTR_ERR(ac100->regmap);
+		dev_err(ac100->dev, "regmap init failed: %d\n", ret);
+		return ret;
+	}
+
+	ret = devm_mfd_add_devices(ac100->dev, PLATFORM_DEVID_NONE, ac100_cells,
+				   ARRAY_SIZE(ac100_cells), NULL, 0, NULL);
+	if (ret) {
+		dev_err(ac100->dev, "failed to add MFD devices: %d\n", ret);
+		return ret;
+	}
+
+	return 0;
+}
+
+static const struct of_device_id ac100_of_match[] = {
+	{ .compatible = "x-powers,ac100" },
+	{ },
+};
+MODULE_DEVICE_TABLE(of, ac100_of_match);
+
+static struct sunxi_rsb_driver ac100_rsb_driver = {
+	.driver = {
+		.name	= "ac100",
+		.of_match_table	= of_match_ptr(ac100_of_match),
+	},
+	.probe	= ac100_rsb_probe,
+};
+module_sunxi_rsb_driver(ac100_rsb_driver);
+
+MODULE_DESCRIPTION("Audio codec MFD core driver for AC100");
+MODULE_AUTHOR("Chen-Yu Tsai <wens@csie.org>");
+MODULE_LICENSE("GPL v2");