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, 86 insertions, 0 deletions

diff --git a/Documentation/userspace-api/mfd_noexec.rst b/Documentation/userspace-api/mfd_noexec.rst
new file mode 100644
index 000000000..7afcc480e
--- /dev/null
+++ b/Documentation/userspace-api/mfd_noexec.rst
@@ -0,0 +1,86 @@
+.. SPDX-License-Identifier: GPL-2.0
+
+==================================
+Introduction of non-executable mfd
+==================================
+:Author:
+    Daniel Verkamp <dverkamp@chromium.org>
+    Jeff Xu <jeffxu@chromium.org>
+
+:Contributor:
+	Aleksa Sarai <cyphar@cyphar.com>
+
+Since Linux introduced the memfd feature, memfds have always had their
+execute bit set, and the memfd_create() syscall doesn't allow setting
+it differently.
+
+However, in a secure-by-default system, such as ChromeOS, (where all
+executables should come from the rootfs, which is protected by verified
+boot), this executable nature of memfd opens a door for NoExec bypass
+and enables “confused deputy attack”.  E.g, in VRP bug [1]: cros_vm
+process created a memfd to share the content with an external process,
+however the memfd is overwritten and used for executing arbitrary code
+and root escalation. [2] lists more VRP of this kind.
+
+On the other hand, executable memfd has its legit use: runc uses memfd’s
+seal and executable feature to copy the contents of the binary then
+execute them. For such a system, we need a solution to differentiate runc's
+use of executable memfds and an attacker's [3].
+
+To address those above:
+ - Let memfd_create() set X bit at creation time.
+ - Let memfd be sealed for modifying X bit when NX is set.
+ - Add a new pid namespace sysctl: vm.memfd_noexec to help applications in
+   migrating and enforcing non-executable MFD.
+
+User API
+========
+``int memfd_create(const char *name, unsigned int flags)``
+
+``MFD_NOEXEC_SEAL``
+	When MFD_NOEXEC_SEAL bit is set in the ``flags``, memfd is created
+	with NX. F_SEAL_EXEC is set and the memfd can't be modified to
+	add X later. MFD_ALLOW_SEALING is also implied.
+	This is the most common case for the application to use memfd.
+
+``MFD_EXEC``
+	When MFD_EXEC bit is set in the ``flags``, memfd is created with X.
+
+Note:
+	``MFD_NOEXEC_SEAL`` implies ``MFD_ALLOW_SEALING``. In case that
+	an app doesn't want sealing, it can add F_SEAL_SEAL after creation.
+
+
+Sysctl:
+========
+``pid namespaced sysctl vm.memfd_noexec``
+
+The new pid namespaced sysctl vm.memfd_noexec has 3 values:
+
+ - 0: MEMFD_NOEXEC_SCOPE_EXEC
+	memfd_create() without MFD_EXEC nor MFD_NOEXEC_SEAL acts like
+	MFD_EXEC was set.
+
+ - 1: MEMFD_NOEXEC_SCOPE_NOEXEC_SEAL
+	memfd_create() without MFD_EXEC nor MFD_NOEXEC_SEAL acts like
+	MFD_NOEXEC_SEAL was set.
+
+ - 2: MEMFD_NOEXEC_SCOPE_NOEXEC_ENFORCED
+	memfd_create() without MFD_NOEXEC_SEAL will be rejected.
+
+The sysctl allows finer control of memfd_create for old software that
+doesn't set the executable bit; for example, a container with
+vm.memfd_noexec=1 means the old software will create non-executable memfd
+by default while new software can create executable memfd by setting
+MFD_EXEC.
+
+The value of vm.memfd_noexec is passed to child namespace at creation
+time. In addition, the setting is hierarchical, i.e. during memfd_create,
+we will search from current ns to root ns and use the most restrictive
+setting.
+
+[1] https://crbug.com/1305267
+
+[2] https://bugs.chromium.org/p/chromium/issues/list?q=type%3Dbug-security%20memfd%20escalation&can=1
+
+[3] https://lwn.net/Articles/781013/