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, 131 insertions, 0 deletions

diff --git a/sound/soc/sof/stream-ipc.c b/sound/soc/sof/stream-ipc.c
new file mode 100644
index 000000000..78758cf58
--- /dev/null
+++ b/sound/soc/sof/stream-ipc.c
@@ -0,0 +1,131 @@
+// SPDX-License-Identifier: (GPL-2.0-only OR BSD-3-Clause)
+//
+// This file is provided under a dual BSD/GPLv2 license.  When using or
+// redistributing this file, you may do so under either license.
+//
+// Copyright(c) 2019 Intel Corporation
+//
+// Authors: Guennadi Liakhovetski <guennadi.liakhovetski@linux.intel.com>
+
+/* Generic SOF IPC code */
+
+#include <linux/device.h>
+#include <linux/export.h>
+#include <linux/module.h>
+#include <linux/types.h>
+
+#include <sound/pcm.h>
+#include <sound/sof/stream.h>
+
+#include "ops.h"
+#include "sof-priv.h"
+#include "sof-audio.h"
+
+struct sof_stream {
+	size_t posn_offset;
+};
+
+/* Mailbox-based Generic IPC implementation */
+int sof_ipc_msg_data(struct snd_sof_dev *sdev,
+		     struct snd_sof_pcm_stream *sps,
+		     void *p, size_t sz)
+{
+	if (!sps || !sdev->stream_box.size) {
+		snd_sof_dsp_mailbox_read(sdev, sdev->dsp_box.offset, p, sz);
+	} else {
+		size_t posn_offset;
+
+		if (sps->substream) {
+			struct sof_stream *stream = sps->substream->runtime->private_data;
+
+			/* The stream might already be closed */
+			if (!stream)
+				return -ESTRPIPE;
+
+			posn_offset = stream->posn_offset;
+		} else if (sps->cstream) {
+
+			struct sof_compr_stream *sstream = sps->cstream->runtime->private_data;
+
+			if (!sstream)
+				return -ESTRPIPE;
+
+			posn_offset = sstream->posn_offset;
+
+		} else {
+			dev_err(sdev->dev, "%s: No stream opened\n", __func__);
+			return -EINVAL;
+		}
+
+		snd_sof_dsp_mailbox_read(sdev, posn_offset, p, sz);
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL(sof_ipc_msg_data);
+
+int sof_set_stream_data_offset(struct snd_sof_dev *sdev,
+			       struct snd_sof_pcm_stream *sps,
+			       size_t posn_offset)
+{
+	/* check if offset is overflow or it is not aligned */
+	if (posn_offset > sdev->stream_box.size ||
+	    posn_offset % sizeof(struct sof_ipc_stream_posn) != 0)
+		return -EINVAL;
+
+	posn_offset += sdev->stream_box.offset;
+
+	if (sps->substream) {
+		struct sof_stream *stream = sps->substream->runtime->private_data;
+
+		stream->posn_offset = posn_offset;
+		dev_dbg(sdev->dev, "pcm: stream dir %d, posn mailbox offset is %zu",
+			sps->substream->stream, posn_offset);
+	} else if (sps->cstream) {
+		struct sof_compr_stream *sstream = sps->cstream->runtime->private_data;
+
+		sstream->posn_offset = posn_offset;
+		dev_dbg(sdev->dev, "compr: stream dir %d, posn mailbox offset is %zu",
+			sps->cstream->direction, posn_offset);
+	} else {
+		dev_err(sdev->dev, "No stream opened");
+		return -EINVAL;
+	}
+
+	return 0;
+}
+EXPORT_SYMBOL(sof_set_stream_data_offset);
+
+int sof_stream_pcm_open(struct snd_sof_dev *sdev,
+			struct snd_pcm_substream *substream)
+{
+	struct sof_stream *stream = kmalloc_obj(*stream);
+
+	if (!stream)
+		return -ENOMEM;
+
+	/* binding pcm substream to hda stream */
+	substream->runtime->private_data = stream;
+
+	/* align to DMA minimum transfer size */
+	snd_pcm_hw_constraint_step(substream->runtime, 0, SNDRV_PCM_HW_PARAM_PERIOD_BYTES, 4);
+
+	/* avoid circular buffer wrap in middle of period */
+	snd_pcm_hw_constraint_integer(substream->runtime,
+				      SNDRV_PCM_HW_PARAM_PERIODS);
+
+	return 0;
+}
+EXPORT_SYMBOL(sof_stream_pcm_open);
+
+int sof_stream_pcm_close(struct snd_sof_dev *sdev,
+			 struct snd_pcm_substream *substream)
+{
+	struct sof_stream *stream = substream->runtime->private_data;
+
+	substream->runtime->private_data = NULL;
+	kfree(stream);
+
+	return 0;
+}
+EXPORT_SYMBOL(sof_stream_pcm_close);