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, 65 insertions, 0 deletions

diff --git a/include/crypto/aes-ctr.h b/include/crypto/aes-ctr.h
new file mode 100644
index 000000000..8e214a857
--- /dev/null
+++ b/include/crypto/aes-ctr.h
@@ -0,0 +1,65 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * AES-CTR and AES-XCTR stream ciphers
+ *
+ * Copyright 2026 Google LLC
+ */
+#ifndef _CRYPTO_AES_CTR_H
+#define _CRYPTO_AES_CTR_H
+
+#include <crypto/aes.h>
+
+/**
+ * aes_ctr() - AES-CTR en/decryption
+ * @dst: The destination buffer.  Can be in-place or out-of-place.  For other
+ *	 overlaps the behavior is unspecified.
+ * @src: The source data
+ * @len: Number of bytes to en/decrypt
+ * @ctr: The counter.  It will be incremented by ceil(@len / AES_BLOCK_SIZE).
+ * @key: The key, already prepared using aes_preparekey() or aes_prepareenckey()
+ *
+ * This implements AES in counter mode with a 128-bit big endian counter.
+ *
+ * This exists only for use by the implementation of modes built on top of CTR
+ * (e.g., GCM and CCM) and some legacy protocols that use CTR mode directly.
+ * Callers are expected to know how to use CTR mode appropriately, including
+ * choosing (key, counter) pairs appropriately to avoid keystream reuse.
+ *
+ * This supports incremental en/decryption.  The length of each non-final chunk
+ * must be a multiple of AES_BLOCK_SIZE, and the updated @ctr must be passed in
+ * each time.
+ *
+ * Context: Any context.
+ */
+void aes_ctr(u8 *dst, const u8 *src, size_t len,
+	     u8 ctr[at_least AES_BLOCK_SIZE], aes_encrypt_arg key);
+
+/**
+ * aes_xctr() - AES-XCTR en/decryption
+ * @dst: The destination buffer.  Can be in-place or out-of-place.  For other
+ *	 overlaps the behavior is unspecified.
+ * @src: The source data
+ * @len: Number of bytes to en/decrypt
+ * @ctr: The block counter (in host endianness).  For the first call, set it to
+ *	 1.  It will be incremented by ceil(@len / AES_BLOCK_SIZE).
+ * @iv: The initialization vector
+ * @key: The key, already prepared using aes_preparekey() or aes_prepareenckey()
+ *
+ * This implements AES in XOR Counter mode, as specified in the paper
+ * "Length-preserving encryption with HCTR2"
+ * (https://eprint.iacr.org/2021/1441.pdf).
+ *
+ * This exists only for use by the implementation of modes built on top of XCTR.
+ * Callers are expected to know how to use XCTR mode appropriately, including
+ * choosing (key, IV) pairs appropriately to avoid keystream reuse.
+ *
+ * This supports incremental en/decryption.  The length of each non-final chunk
+ * must be a multiple of AES_BLOCK_SIZE, and the updated @ctr must be passed in
+ * each time.
+ *
+ * Context: Any context.
+ */
+void aes_xctr(u8 *dst, const u8 *src, size_t len, u64 *ctr,
+	      const u8 iv[at_least AES_BLOCK_SIZE], aes_encrypt_arg key);
+
+#endif /* _CRYPTO_AES_CTR_H */