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, 55 insertions, 0 deletions

diff --git a/tools/testing/memblock/Makefile b/tools/testing/memblock/Makefile
new file mode 100644
index 000000000..d80982ccd
--- /dev/null
+++ b/tools/testing/memblock/Makefile
@@ -0,0 +1,55 @@
+# SPDX-License-Identifier: GPL-2.0
+
+# Memblock simulator requires AddressSanitizer (libasan) and liburcu development
+# packages installed
+CFLAGS += -I. -I../../include -Wall -O2 -fsanitize=address \
+	  -fsanitize=undefined -D CONFIG_PHYS_ADDR_T_64BIT
+LDFLAGS += -fsanitize=address -fsanitize=undefined
+TARGETS = main
+TEST_OFILES = tests/alloc_nid_api.o tests/alloc_helpers_api.o tests/alloc_api.o \
+		  tests/basic_api.o tests/common.o tests/alloc_exact_nid_api.o
+DEP_OFILES = memblock.o lib/slab.o mmzone.o slab.o cmdline.o
+OFILES = main.o $(DEP_OFILES) $(TEST_OFILES)
+EXTR_SRC = ../../../mm/memblock.c
+
+ifeq ($(BUILD), 32)
+	CFLAGS += -m32
+	LDFLAGS += -m32
+endif
+
+# Process user parameters
+include scripts/Makefile.include
+
+main: $(OFILES)
+
+$(OFILES): include
+
+include: ../../../include/linux/memblock.h ../../include/linux/*.h \
+	../../include/asm/*.h
+
+	@mkdir -p linux
+	test -L linux/memblock.h || ln -s ../../../../include/linux/memblock.h linux/memblock.h
+	test -L asm/asm.h || ln -s ../../../arch/x86/include/asm/asm.h asm/asm.h
+	test -L asm/cmpxchg.h || ln -s ../../../arch/x86/include/asm/cmpxchg.h asm/cmpxchg.h
+
+memblock.c: $(EXTR_SRC)
+	test -L memblock.c || ln -s $(EXTR_SRC) memblock.c
+
+clean:
+	$(RM) $(TARGETS) $(OFILES) linux/memblock.h memblock.c asm/asm.h asm/cmpxchg.h
+
+help:
+	@echo  'Memblock simulator'
+	@echo  ''
+	@echo  'Available targets:'
+	@echo  '  main		  - Build the memblock simulator'
+	@echo  '  clean		  - Remove generated files and symlinks in the directory'
+	@echo  ''
+	@echo  'Configuration:'
+	@echo  '  make MEMBLOCK_DEBUG=1     - enable memblock_dbg() messages'
+	@echo  '  make NUMA=1               - simulate enabled NUMA'
+	@echo  '  make 32BIT_PHYS_ADDR_T=1  - Use 32 bit physical addresses'
+
+vpath %.c ../../lib
+
+.PHONY: clean include help