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, 197 insertions, 0 deletions

diff --git a/io_uring/xattr.c b/io_uring/xattr.c
new file mode 100644
index 000000000..5303df3f2
--- /dev/null
+++ b/io_uring/xattr.c
@@ -0,0 +1,197 @@
+// SPDX-License-Identifier: GPL-2.0
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/fs.h>
+#include <linux/file.h>
+#include <linux/mm.h>
+#include <linux/slab.h>
+#include <linux/namei.h>
+#include <linux/io_uring.h>
+#include <linux/xattr.h>
+
+#include <uapi/linux/io_uring.h>
+
+#include "../fs/internal.h"
+
+#include "io_uring.h"
+#include "xattr.h"
+
+struct io_xattr {
+	struct file			*file;
+	struct kernel_xattr_ctx		ctx;
+	struct delayed_filename		filename;
+};
+
+void io_xattr_cleanup(struct io_kiocb *req)
+{
+	struct io_xattr *ix = io_kiocb_to_cmd(req, struct io_xattr);
+
+	dismiss_delayed_filename(&ix->filename);
+	kfree(ix->ctx.kname);
+	kvfree(ix->ctx.kvalue);
+}
+
+static void io_xattr_finish(struct io_kiocb *req, int ret)
+{
+	req->flags &= ~REQ_F_NEED_CLEANUP;
+
+	io_xattr_cleanup(req);
+	io_req_set_res(req, ret, 0);
+}
+
+static int __io_getxattr_prep(struct io_kiocb *req,
+			      const struct io_uring_sqe *sqe)
+{
+	struct io_xattr *ix = io_kiocb_to_cmd(req, struct io_xattr);
+	const char __user *name;
+	int ret;
+
+	INIT_DELAYED_FILENAME(&ix->filename);
+	ix->ctx.kvalue = NULL;
+	name = u64_to_user_ptr(READ_ONCE(sqe->addr));
+	ix->ctx.value = u64_to_user_ptr(READ_ONCE(sqe->addr2));
+	ix->ctx.size = READ_ONCE(sqe->len);
+	ix->ctx.flags = READ_ONCE(sqe->xattr_flags);
+
+	if (ix->ctx.flags)
+		return -EINVAL;
+
+	ix->ctx.kname = kmalloc_obj(*ix->ctx.kname);
+	if (!ix->ctx.kname)
+		return -ENOMEM;
+
+	ret = import_xattr_name(ix->ctx.kname, name);
+	if (ret) {
+		kfree(ix->ctx.kname);
+		return ret;
+	}
+
+	req->flags |= REQ_F_NEED_CLEANUP;
+	req->flags |= REQ_F_FORCE_ASYNC;
+	return 0;
+}
+
+int io_fgetxattr_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)
+{
+	return __io_getxattr_prep(req, sqe);
+}
+
+int io_getxattr_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)
+{
+	struct io_xattr *ix = io_kiocb_to_cmd(req, struct io_xattr);
+	const char __user *path;
+	int ret;
+
+	if (unlikely(req->flags & REQ_F_FIXED_FILE))
+		return -EBADF;
+
+	ret = __io_getxattr_prep(req, sqe);
+	if (ret)
+		return ret;
+
+	path = u64_to_user_ptr(READ_ONCE(sqe->addr3));
+
+	return delayed_getname(&ix->filename, path);
+}
+
+int io_fgetxattr(struct io_kiocb *req, unsigned int issue_flags)
+{
+	struct io_xattr *ix = io_kiocb_to_cmd(req, struct io_xattr);
+	int ret;
+
+	WARN_ON_ONCE(issue_flags & IO_URING_F_NONBLOCK);
+
+	ret = file_getxattr(req->file, &ix->ctx);
+	io_xattr_finish(req, ret);
+	return IOU_COMPLETE;
+}
+
+int io_getxattr(struct io_kiocb *req, unsigned int issue_flags)
+{
+	struct io_xattr *ix = io_kiocb_to_cmd(req, struct io_xattr);
+	CLASS(filename_complete_delayed, name)(&ix->filename);
+	int ret;
+
+	WARN_ON_ONCE(issue_flags & IO_URING_F_NONBLOCK);
+
+	ret = filename_getxattr(AT_FDCWD, name, LOOKUP_FOLLOW, &ix->ctx);
+	io_xattr_finish(req, ret);
+	return IOU_COMPLETE;
+}
+
+static int __io_setxattr_prep(struct io_kiocb *req,
+			const struct io_uring_sqe *sqe)
+{
+	struct io_xattr *ix = io_kiocb_to_cmd(req, struct io_xattr);
+	const char __user *name;
+	int ret;
+
+	INIT_DELAYED_FILENAME(&ix->filename);
+	name = u64_to_user_ptr(READ_ONCE(sqe->addr));
+	ix->ctx.cvalue = u64_to_user_ptr(READ_ONCE(sqe->addr2));
+	ix->ctx.kvalue = NULL;
+	ix->ctx.size = READ_ONCE(sqe->len);
+	ix->ctx.flags = READ_ONCE(sqe->xattr_flags);
+
+	ix->ctx.kname = kmalloc_obj(*ix->ctx.kname);
+	if (!ix->ctx.kname)
+		return -ENOMEM;
+
+	ret = setxattr_copy(name, &ix->ctx);
+	if (ret) {
+		kfree(ix->ctx.kname);
+		return ret;
+	}
+
+	req->flags |= REQ_F_NEED_CLEANUP;
+	req->flags |= REQ_F_FORCE_ASYNC;
+	return 0;
+}
+
+int io_setxattr_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)
+{
+	struct io_xattr *ix = io_kiocb_to_cmd(req, struct io_xattr);
+	const char __user *path;
+	int ret;
+
+	if (unlikely(req->flags & REQ_F_FIXED_FILE))
+		return -EBADF;
+
+	ret = __io_setxattr_prep(req, sqe);
+	if (ret)
+		return ret;
+
+	path = u64_to_user_ptr(READ_ONCE(sqe->addr3));
+
+	return delayed_getname(&ix->filename, path);
+}
+
+int io_fsetxattr_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe)
+{
+	return __io_setxattr_prep(req, sqe);
+}
+
+int io_fsetxattr(struct io_kiocb *req, unsigned int issue_flags)
+{
+	struct io_xattr *ix = io_kiocb_to_cmd(req, struct io_xattr);
+	int ret;
+
+	WARN_ON_ONCE(issue_flags & IO_URING_F_NONBLOCK);
+
+	ret = file_setxattr(req->file, &ix->ctx);
+	io_xattr_finish(req, ret);
+	return IOU_COMPLETE;
+}
+
+int io_setxattr(struct io_kiocb *req, unsigned int issue_flags)
+{
+	struct io_xattr *ix = io_kiocb_to_cmd(req, struct io_xattr);
+	CLASS(filename_complete_delayed, name)(&ix->filename);
+	int ret;
+
+	WARN_ON_ONCE(issue_flags & IO_URING_F_NONBLOCK);
+
+	ret = filename_setxattr(AT_FDCWD, name, LOOKUP_FOLLOW, &ix->ctx);
+	io_xattr_finish(req, ret);
+	return IOU_COMPLETE;
+}