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, 180 insertions, 0 deletions

diff --git a/fs/squashfs/export.c b/fs/squashfs/export.c
new file mode 100644
index 000000000..62972f0ff
--- /dev/null
+++ b/fs/squashfs/export.c
@@ -0,0 +1,180 @@
+// SPDX-License-Identifier: GPL-2.0-or-later
+/*
+ * Squashfs - a compressed read only filesystem for Linux
+ *
+ * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008
+ * Phillip Lougher <phillip@squashfs.org.uk>
+ *
+ * export.c
+ */
+
+/*
+ * This file implements code to make Squashfs filesystems exportable (NFS etc.)
+ *
+ * The export code uses an inode lookup table to map inode numbers passed in
+ * filehandles to an inode location on disk.  This table is stored compressed
+ * into metadata blocks.  A second index table is used to locate these.  This
+ * second index table for speed of access (and because it is small) is read at
+ * mount time and cached in memory.
+ *
+ * The inode lookup table is used only by the export code, inode disk
+ * locations are directly encoded in directories, enabling direct access
+ * without an intermediate lookup for all operations except the export ops.
+ */
+
+#include <linux/fs.h>
+#include <linux/vfs.h>
+#include <linux/dcache.h>
+#include <linux/exportfs.h>
+#include <linux/slab.h>
+
+#include "squashfs_fs.h"
+#include "squashfs_fs_sb.h"
+#include "squashfs_fs_i.h"
+#include "squashfs.h"
+
+/*
+ * Look-up inode number (ino) in table, returning the inode location.
+ */
+static long long squashfs_inode_lookup(struct super_block *sb, int ino_num)
+{
+	struct squashfs_sb_info *msblk = sb->s_fs_info;
+	int blk = SQUASHFS_LOOKUP_BLOCK(ino_num - 1);
+	int offset = SQUASHFS_LOOKUP_BLOCK_OFFSET(ino_num - 1);
+	u64 start;
+	__le64 ino;
+	int err;
+
+	TRACE("Entered squashfs_inode_lookup, inode_number = %d\n", ino_num);
+
+	if (ino_num == 0 || (ino_num - 1) >= msblk->inodes)
+		return -EINVAL;
+
+	start = le64_to_cpu(msblk->inode_lookup_table[blk]);
+
+	err = squashfs_read_metadata(sb, &ino, &start, &offset, sizeof(ino));
+	if (err < 0)
+		return err;
+
+	TRACE("squashfs_inode_lookup, inode = 0x%llx\n",
+		(u64) le64_to_cpu(ino));
+
+	return le64_to_cpu(ino);
+}
+
+
+static struct dentry *squashfs_export_iget(struct super_block *sb,
+	unsigned int ino_num)
+{
+	long long ino;
+	struct dentry *dentry = ERR_PTR(-ENOENT);
+
+	TRACE("Entered squashfs_export_iget\n");
+
+	ino = squashfs_inode_lookup(sb, ino_num);
+	if (ino >= 0)
+		dentry = d_obtain_alias(squashfs_iget(sb, ino, ino_num));
+
+	return dentry;
+}
+
+
+static struct dentry *squashfs_fh_to_dentry(struct super_block *sb,
+		struct fid *fid, int fh_len, int fh_type)
+{
+	if ((fh_type != FILEID_INO32_GEN && fh_type != FILEID_INO32_GEN_PARENT)
+			|| fh_len < 2)
+		return NULL;
+
+	return squashfs_export_iget(sb, fid->i32.ino);
+}
+
+
+static struct dentry *squashfs_fh_to_parent(struct super_block *sb,
+		struct fid *fid, int fh_len, int fh_type)
+{
+	if (fh_type != FILEID_INO32_GEN_PARENT || fh_len < 4)
+		return NULL;
+
+	return squashfs_export_iget(sb, fid->i32.parent_ino);
+}
+
+
+static struct dentry *squashfs_get_parent(struct dentry *child)
+{
+	struct inode *inode = d_inode(child);
+	unsigned int parent_ino = squashfs_i(inode)->parent;
+
+	return squashfs_export_iget(inode->i_sb, parent_ino);
+}
+
+
+/*
+ * Read uncompressed inode lookup table indexes off disk into memory
+ */
+__le64 *squashfs_read_inode_lookup_table(struct super_block *sb,
+		u64 lookup_table_start, u64 next_table, unsigned int inodes)
+{
+	unsigned int length = SQUASHFS_LOOKUP_BLOCK_BYTES(inodes);
+	unsigned int indexes = SQUASHFS_LOOKUP_BLOCKS(inodes);
+	int n;
+	__le64 *table;
+	u64 start, end;
+
+	TRACE("In read_inode_lookup_table, length %d\n", length);
+
+	/* Sanity check values */
+
+	/* there should always be at least one inode */
+	if (inodes == 0)
+		return ERR_PTR(-EINVAL);
+
+	/*
+	 * The computed size of the lookup table (length bytes) should exactly
+	 * match the table start and end points
+	 */
+	if (length != (next_table - lookup_table_start))
+		return ERR_PTR(-EINVAL);
+
+	table = squashfs_read_table(sb, lookup_table_start, length);
+	if (IS_ERR(table))
+		return table;
+
+	/*
+	 * table0], table[1], ... table[indexes - 1] store the locations
+	 * of the compressed inode lookup blocks.  Each entry should be
+	 * less than the next (i.e. table[0] < table[1]), and the difference
+	 * between them should be SQUASHFS_METADATA_SIZE or less.
+	 * table[indexes - 1] should  be less than lookup_table_start, and
+	 * again the difference should be SQUASHFS_METADATA_SIZE or less
+	 */
+	for (n = 0; n < (indexes - 1); n++) {
+		start = le64_to_cpu(table[n]);
+		end = le64_to_cpu(table[n + 1]);
+
+		if (start >= end
+		    || (end - start) >
+		    (SQUASHFS_METADATA_SIZE + SQUASHFS_BLOCK_OFFSET)) {
+			kfree(table);
+			return ERR_PTR(-EINVAL);
+		}
+	}
+
+	start = le64_to_cpu(table[indexes - 1]);
+	if (start >= lookup_table_start ||
+	    (lookup_table_start - start) >
+	    (SQUASHFS_METADATA_SIZE + SQUASHFS_BLOCK_OFFSET)) {
+		kfree(table);
+		return ERR_PTR(-EINVAL);
+	}
+
+	return table;
+}
+
+
+const struct export_operations squashfs_export_ops = {
+	.encode_fh = generic_encode_ino32_fh,
+	.fh_to_dentry = squashfs_fh_to_dentry,
+	.fh_to_parent = squashfs_fh_to_parent,
+	.get_parent = squashfs_get_parent
+};