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, 57 insertions, 0 deletions

diff --git a/arch/powerpc/boot/redboot.h b/arch/powerpc/boot/redboot.h
new file mode 100644
index 000000000..8f319b1ad
--- /dev/null
+++ b/arch/powerpc/boot/redboot.h
@@ -0,0 +1,57 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef _PPC_REDBOOT_H
+#define _PPC_REDBOOT_H
+
+//=========================================================================
+// include/asm-ppc/redboot.h
+//   Copyright (c) 2002, 2003 Gary Thomas (<gary@mlbassoc.com>
+//   Copyright (c) 1997 Dan Malek (dmalek@jlc.net)
+
+//
+// Board specific details, as provided by RedBoot
+//
+
+/* A Board Information structure that is given to a program when
+ * RedBoot starts it up.  Note: not all fields make sense for all
+ * architectures and it's up to the platform specific code to fill
+ * in the details.
+ */
+typedef struct bd_info {
+    unsigned int   bi_tag;        /* Should be 0x42444944 "BDID" */
+    unsigned int   bi_size;       /* Size of this structure */
+    unsigned int   bi_revision;   /* revision of this structure */
+    unsigned int   bi_bdate;      /* bootstrap date, i.e. 0x19971106 */
+    unsigned int   bi_memstart;   /* Memory start address */
+    unsigned int   bi_memsize;    /* Memory (end) size in bytes */
+    unsigned int   bi_intfreq;    /* Internal Freq, in Hz */
+    unsigned int   bi_busfreq;    /* Bus Freq, in Hz */
+    unsigned int   bi_cpmfreq;    /* CPM Freq, in Hz */
+    unsigned int   bi_brgfreq;    /* BRG Freq, in Hz */
+    unsigned int   bi_vco;        /* VCO Out from PLL */
+    unsigned int   bi_pci_freq;   /* PCI Freq, in Hz */
+    unsigned int   bi_baudrate;   /* Default console baud rate */
+    unsigned int   bi_immr;       /* IMMR when called from boot rom */
+    unsigned char  bi_enetaddr[6];
+    unsigned int   bi_flashbase;  /* Physical address of FLASH memory */
+    unsigned int   bi_flashsize;  /* Length of FLASH memory */
+    int            bi_flashwidth; /* Width (8,16,32,64) */
+    unsigned char *bi_cmdline;    /* Pointer to command line */
+    unsigned char  bi_esa[3][6];  /* Ethernet station addresses */
+    unsigned int   bi_ramdisk_begin, bi_ramdisk_end;
+    struct {                      /* Information about [main] video screen */
+        short x_res;              /*   Horizontal resolution in pixels */
+        short y_res;              /*   Vertical resolution in pixels */
+        short bpp;                /*   Bits/pixel */
+        short mode;               /*   Type of pixels (packed, indexed) */
+        unsigned long fb;         /*   Pointer to frame buffer (pixel) memory */
+    } bi_video;
+    void         (*bi_cputc)(char);   /* Write a character to the RedBoot console */
+    char         (*bi_cgetc)(void);   /* Read a character from the RedBoot console */
+    int          (*bi_ctstc)(void);   /* Test for input on the RedBoot console */
+} bd_t;
+
+#define BI_REV 0x0102    /* Version 1.02 */
+
+#define bi_pci_busfreq bi_pci_freq
+#define bi_immr_base   bi_immr
+#endif