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, 50 insertions, 0 deletions

diff --git a/arch/mips/mm/physaddr.c b/arch/mips/mm/physaddr.c
new file mode 100644
index 000000000..a6b1bf820
--- /dev/null
+++ b/arch/mips/mm/physaddr.c
@@ -0,0 +1,50 @@
+// SPDX-License-Identifier: GPL-2.0
+#include <linux/bug.h>
+#include <linux/export.h>
+#include <linux/types.h>
+#include <linux/mmdebug.h>
+#include <linux/mm.h>
+
+#include <asm/addrspace.h>
+#include <asm/sections.h>
+#include <asm/io.h>
+#include <asm/page.h>
+#include <asm/dma.h>
+
+static inline bool __debug_virt_addr_valid(unsigned long x)
+{
+	/*
+	 * MAX_DMA_ADDRESS is a virtual address that may not correspond to an
+	 * actual physical address. Enough code relies on
+	 * virt_to_phys(MAX_DMA_ADDRESS) that we just need to work around it
+	 * and always return true.
+	 */
+	if (x == MAX_DMA_ADDRESS)
+		return true;
+
+	return x >= PAGE_OFFSET && (KSEGX(x) < KSEG2 ||
+	       IS_ENABLED(CONFIG_EVA) ||
+	       !IS_ENABLED(CONFIG_HIGHMEM));
+}
+
+phys_addr_t __virt_to_phys(volatile const void *x)
+{
+	WARN(!__debug_virt_addr_valid((unsigned long)x),
+	     "virt_to_phys used for non-linear address: %p (%pS)\n",
+	     x, x);
+
+	return __virt_to_phys_nodebug(x);
+}
+EXPORT_SYMBOL(__virt_to_phys);
+
+phys_addr_t __phys_addr_symbol(unsigned long x)
+{
+	/* This is bounds checking against the kernel image only.
+	 * __pa_symbol should only be used on kernel symbol addresses.
+	 */
+	VIRTUAL_BUG_ON(x < (unsigned long)_text ||
+		       x > (unsigned long)_end);
+
+	return __pa_symbol_nodebug(x);
+}
+EXPORT_SYMBOL(__phys_addr_symbol);