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, 179 insertions, 0 deletions

diff --git a/drivers/net/dsa/Kconfig b/drivers/net/dsa/Kconfig
new file mode 100644
index 000000000..676fb7dff
--- /dev/null
+++ b/drivers/net/dsa/Kconfig
@@ -0,0 +1,179 @@
+# SPDX-License-Identifier: GPL-2.0-only
+menu "Distributed Switch Architecture drivers"
+	depends on NET_DSA
+
+source "drivers/net/dsa/b53/Kconfig"
+
+config NET_DSA_BCM_SF2
+	tristate "Broadcom Starfighter 2 Ethernet switch support"
+	depends on HAS_IOMEM
+	select NET_DSA_TAG_BRCM
+	select FIXED_PHY
+	select BCM7XXX_PHY
+	select MDIO_BCM_UNIMAC
+	select B53
+	help
+	  This enables support for the Broadcom Starfighter 2 Ethernet
+	  switch chips.
+
+config NET_DSA_LOOP
+	tristate "DSA mock-up Ethernet switch chip support"
+	select NET_DSA_TAG_NONE
+	select FIXED_PHY
+	help
+	  This enables support for a fake mock-up switch chip which
+	  exercises the DSA APIs.
+
+source "drivers/net/dsa/hirschmann/Kconfig"
+
+source "drivers/net/dsa/lantiq/Kconfig"
+
+config NET_DSA_MT7530
+	tristate "MediaTek MT7530 and MT7531 Ethernet switch support"
+	select NET_DSA_TAG_MTK
+	select REGMAP_IRQ
+	imply NET_DSA_MT7530_MDIO
+	imply NET_DSA_MT7530_MMIO
+	help
+	  This enables support for the MediaTek MT7530 and MT7531 Ethernet
+	  switch chips. Multi-chip module MT7530 in MT7621AT, MT7621DAT,
+	  MT7621ST and MT7623AI SoCs, and built-in switch in MT7988 SoC are
+	  supported as well.
+
+config NET_DSA_MT7530_MDIO
+	tristate "MediaTek MT7530 MDIO interface driver"
+	depends on NET_DSA_MT7530
+	select MEDIATEK_GE_PHY
+	select PCS_MTK_LYNXI
+	help
+	  This enables support for the MediaTek MT7530 and MT7531 switch
+	  chips which are connected via MDIO, as well as multi-chip
+	  module MT7530 which can be found in the MT7621AT, MT7621DAT,
+	  MT7621ST and MT7623AI SoCs.
+
+config NET_DSA_MT7530_MMIO
+	tristate "MediaTek MT7530 MMIO interface driver"
+	depends on NET_DSA_MT7530
+	depends on HAS_IOMEM
+	imply MEDIATEK_GE_SOC_PHY
+	help
+	  This enables support for the built-in Ethernet switch found
+	  in the MediaTek MT7988 SoC.
+	  The switch is a similar design as MT7531, but the switch registers
+	  are directly mapped into the SoCs register space rather than being
+	  accessible via MDIO.
+
+config NET_DSA_MT7628
+	tristate "MediaTek MT7628 Embedded Ethernet switch support"
+	depends on HAS_IOMEM && (SOC_MT7620 || COMPILE_TEST)
+	select NET_DSA_TAG_MT7628
+	select MEDIATEK_FE_SOC_PHY
+	select REGMAP_MMIO
+	help
+	  This enables support for the built-in Ethernet switch found
+	  in the MT7628 SoC.
+
+config NET_DSA_MV88E6060
+	tristate "Marvell 88E6060 ethernet switch chip support"
+	select NET_DSA_TAG_TRAILER
+	help
+	  This enables support for the Marvell 88E6060 ethernet switch
+	  chip.
+
+source "drivers/net/dsa/microchip/Kconfig"
+
+source "drivers/net/dsa/mv88e6xxx/Kconfig"
+
+source "drivers/net/dsa/mxl862xx/Kconfig"
+
+source "drivers/net/dsa/netc/Kconfig"
+
+source "drivers/net/dsa/ocelot/Kconfig"
+
+source "drivers/net/dsa/qca/Kconfig"
+
+source "drivers/net/dsa/sja1105/Kconfig"
+
+source "drivers/net/dsa/xrs700x/Kconfig"
+
+source "drivers/net/dsa/realtek/Kconfig"
+
+config NET_DSA_RZN1_A5PSW
+	tristate "Renesas RZ/N1 A5PSW Ethernet switch support"
+	depends on OF && (ARCH_RZN1 || COMPILE_TEST)
+	select NET_DSA_TAG_RZN1_A5PSW
+	select PCS_RZN1_MIIC
+	help
+	  This driver supports the A5PSW switch, which is embedded in Renesas
+	  RZ/N1 SoC.
+
+config NET_DSA_KS8995
+	tristate "Micrel KS8995 family 5-ports 10/100 Ethernet switches"
+	depends on SPI
+	select NET_DSA_TAG_NONE
+	help
+	  This driver supports the Micrel KS8995 family of 10/100 Mbit ethernet
+	  switches, managed over SPI.
+
+config NET_DSA_SMSC_LAN9303
+	tristate
+	select NET_DSA_TAG_LAN9303
+	select REGMAP
+	imply SMSC_PHY
+	help
+	  This enables support for the Microchip LAN9303/LAN9354 3 port ethernet
+	  switch chips.
+
+config NET_DSA_SMSC_LAN9303_I2C
+	tristate "SMSC/Microchip LAN9303 3-ports 10/100 ethernet switch in I2C managed mode"
+	depends on I2C
+	depends on VLAN_8021Q || VLAN_8021Q=n
+	select NET_DSA_SMSC_LAN9303
+	select REGMAP_I2C
+	help
+	  Enable access functions if the SMSC/Microchip LAN9303 is configured
+	  for I2C managed mode.
+
+config NET_DSA_SMSC_LAN9303_MDIO
+	tristate "Microchip LAN9303/LAN9354 3-ports 10/100 ethernet switch in MDIO managed mode"
+	select NET_DSA_SMSC_LAN9303
+	depends on VLAN_8021Q || VLAN_8021Q=n
+	help
+	  Enable access functions if the Microchip LAN9303/LAN9354 is configured
+	  for MDIO managed mode.
+
+config NET_DSA_VITESSE_VSC73XX
+	tristate
+	select NET_DSA_TAG_VSC73XX_8021Q
+	select FIXED_PHY
+	select VITESSE_PHY
+	select GPIOLIB
+	help
+	  This enables support for the Vitesse VSC7385, VSC7388,
+	  VSC7395 and VSC7398 SparX integrated ethernet switches.
+
+config NET_DSA_VITESSE_VSC73XX_SPI
+	tristate "Vitesse VSC7385/7388/7395/7398 SPI mode support"
+	depends on SPI
+	select NET_DSA_VITESSE_VSC73XX
+	help
+	  This enables support for the Vitesse VSC7385, VSC7388, VSC7395
+	  and VSC7398 SparX integrated ethernet switches in SPI managed mode.
+
+config NET_DSA_VITESSE_VSC73XX_PLATFORM
+	tristate "Vitesse VSC7385/7388/7395/7398 Platform mode support"
+	depends on HAS_IOMEM
+	select NET_DSA_VITESSE_VSC73XX
+	help
+	  This enables support for the Vitesse VSC7385, VSC7388, VSC7395
+	  and VSC7398 SparX integrated ethernet switches, connected over
+	  a CPU-attached address bus and work in memory-mapped I/O mode.
+
+config NET_DSA_YT921X
+	tristate "Motorcomm YT9215 ethernet switch chip support"
+	select NET_DSA_TAG_YT921X
+	select NET_IEEE8021Q_HELPERS if DCB
+	help
+	  This enables support for the Motorcomm YT9215 ethernet switch
+	  chip.
+endmenu