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, 79 insertions, 0 deletions

diff --git a/Documentation/dev-tools/ubsan.rst b/Documentation/dev-tools/ubsan.rst
new file mode 100644
index 000000000..e3591f8e9
--- /dev/null
+++ b/Documentation/dev-tools/ubsan.rst
@@ -0,0 +1,79 @@
+.. SPDX-License-Identifier: GPL-2.0
+
+Undefined Behavior Sanitizer - UBSAN
+====================================
+
+UBSAN is a runtime undefined behaviour checker.
+
+UBSAN uses compile-time instrumentation to catch undefined behavior (UB).
+Compiler inserts code that perform certain kinds of checks before operations
+that may cause UB. If check fails (i.e. UB detected) __ubsan_handle_*
+function called to print error message.
+
+GCC has that feature since 4.9.x [1_] (see ``-fsanitize=undefined`` option and
+its suboptions). GCC 5.x has more checkers implemented [2_].
+
+Report example
+--------------
+
+::
+
+	 ================================================================================
+	 UBSAN: Undefined behaviour in ../include/linux/bitops.h:110:33
+	 shift exponent 32 is to large for 32-bit type 'unsigned int'
+	 CPU: 0 PID: 0 Comm: swapper Not tainted 4.4.0-rc1+ #26
+	  0000000000000000 ffffffff82403cc8 ffffffff815e6cd6 0000000000000001
+	  ffffffff82403cf8 ffffffff82403ce0 ffffffff8163a5ed 0000000000000020
+	  ffffffff82403d78 ffffffff8163ac2b ffffffff815f0001 0000000000000002
+	 Call Trace:
+	  [<ffffffff815e6cd6>] dump_stack+0x45/0x5f
+	  [<ffffffff8163a5ed>] ubsan_epilogue+0xd/0x40
+	  [<ffffffff8163ac2b>] __ubsan_handle_shift_out_of_bounds+0xeb/0x130
+	  [<ffffffff815f0001>] ? radix_tree_gang_lookup_slot+0x51/0x150
+	  [<ffffffff8173c586>] _mix_pool_bytes+0x1e6/0x480
+	  [<ffffffff83105653>] ? dmi_walk_early+0x48/0x5c
+	  [<ffffffff8173c881>] add_device_randomness+0x61/0x130
+	  [<ffffffff83105b35>] ? dmi_save_one_device+0xaa/0xaa
+	  [<ffffffff83105653>] dmi_walk_early+0x48/0x5c
+	  [<ffffffff831066ae>] dmi_scan_machine+0x278/0x4b4
+	  [<ffffffff8111d58a>] ? vprintk_default+0x1a/0x20
+	  [<ffffffff830ad120>] ? early_idt_handler_array+0x120/0x120
+	  [<ffffffff830b2240>] setup_arch+0x405/0xc2c
+	  [<ffffffff830ad120>] ? early_idt_handler_array+0x120/0x120
+	  [<ffffffff830ae053>] start_kernel+0x83/0x49a
+	  [<ffffffff830ad120>] ? early_idt_handler_array+0x120/0x120
+	  [<ffffffff830ad386>] x86_64_start_reservations+0x2a/0x2c
+	  [<ffffffff830ad4f3>] x86_64_start_kernel+0x16b/0x17a
+	 ================================================================================
+
+Usage
+-----
+
+To enable UBSAN, configure the kernel with::
+
+  CONFIG_UBSAN=y
+
+To exclude files from being instrumented use::
+
+  UBSAN_SANITIZE_main.o := n
+
+and to exclude all targets in one directory use::
+
+  UBSAN_SANITIZE := n
+
+When disabled for all targets, specific files can be enabled using::
+
+  UBSAN_SANITIZE_main.o := y
+
+Detection of unaligned accesses controlled through the separate option -
+CONFIG_UBSAN_ALIGNMENT. It's off by default on architectures that support
+unaligned accesses (CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS=y). One could
+still enable it in config, just note that it will produce a lot of UBSAN
+reports.
+
+References
+----------
+
+.. _1: https://gcc.gnu.org/onlinedocs/gcc-4.9.0/gcc/Debugging-Options.html
+.. _2: https://gcc.gnu.org/onlinedocs/gcc/Debugging-Options.html
+.. _3: https://clang.llvm.org/docs/UndefinedBehaviorSanitizer.html