From 2b729fe7f3e9478a21a336231daf35768e7cf37b Mon Sep 17 00:00:00 2001
From: Tejun Heo <tj@kernel.org>
Date: Fri, 3 Apr 2020 11:32:13 -0400
Subject: [PATCH] Revert "cpuset: Make cpuset hotplug synchronous"
This reverts commit a49e4629b5ed ("cpuset: Make cpuset hotplug synchronous") as
it may deadlock with cpu hotplug path.
Link: http://lkml.kernel.org/r/F0388D99-84D7-453B-9B6B-EEFF0E7BE4CC@lca.pw
Signed-off-by: Tejun Heo <tj@kernel.org>
Reported-by: Qian Cai <cai@lca.pw>
Cc: Prateek Sood <prsood@codeaurora.org>
---
include/linux/cpuset.h | 3 +++
kernel/cgroup/cpuset.c | 35 ++++++++++++++---------------------
kernel/power/process.c | 2 ++
3 files changed, 19 insertions(+), 21 deletions(-)
diff --git a/include/linux/cpuset.h b/include/linux/cpuset.h
index cede4cb98b78..04c20de66afc 100644
--- a/include/linux/cpuset.h
+++ b/include/linux/cpuset.h
@@ -54,6 +54,7 @@ extern int cpuset_init(void);
extern void cpuset_init_smp(void);
extern void cpuset_force_rebuild(void);
extern void cpuset_update_active_cpus(void);
+extern void cpuset_wait_for_hotplug(void);
extern void cpuset_read_lock(void);
extern void cpuset_read_unlock(void);
extern void cpuset_cpus_allowed(struct task_struct *p, struct cpumask *mask);
@@ -175,6 +176,8 @@ static inline void cpuset_update_active_cpus(void)
partition_sched_domains(1, NULL, NULL);
}
+static inline void cpuset_wait_for_hotplug(void) { }
+
static inline void cpuset_read_lock(void) { }
static inline void cpuset_read_unlock(void) { }
diff --git a/kernel/cgroup/cpuset.c b/kernel/cgroup/cpuset.c
index cafd4d2ff882..58f5073acff7 100644
--- a/kernel/cgroup/cpuset.c
+++ b/kernel/cgroup/cpuset.c
@@ -3101,7 +3101,7 @@ update_tasks:
}
/**
- * cpuset_hotplug - handle CPU/memory hotunplug for a cpuset
+ * cpuset_hotplug_workfn - handle CPU/memory hotunplug for a cpuset
*
* This function is called after either CPU or memory configuration has
* changed and updates cpuset accordingly. The top_cpuset is always
@@ -3116,7 +3116,7 @@ update_tasks:
* Note that CPU offlining during suspend is ignored. We don't modify
* cpusets across suspend/resume cycles at all.
*/
-static void cpuset_hotplug(bool use_cpu_hp_lock)
+static void cpuset_hotplug_workfn(struct work_struct *work)
{
static cpumask_t new_cpus;
static nodemask_t new_mems;
@@ -3201,32 +3201,25 @@ static void cpuset_hotplug(bool use_cpu_hp_lock)
/* rebuild sched domains if cpus_allowed has changed */
if (cpus_updated || force_rebuild) {
force_rebuild = false;
- if (use_cpu_hp_lock)
- rebuild_sched_domains();
- else {
- /* Acquiring cpu_hotplug_lock is not required.
- * When cpuset_hotplug() is called in hotplug path,
- * cpu_hotplug_lock is held by the hotplug context
- * which is waiting for cpuhp_thread_fun to indicate
- * completion of callback.
- */
- percpu_down_write(&cpuset_rwsem);
- rebuild_sched_domains_locked();
- percpu_up_write(&cpuset_rwsem);
- }
+ rebuild_sched_domains();
}
free_cpumasks(NULL, ptmp);
}
-static void cpuset_hotplug_workfn(struct work_struct *work)
-{
- cpuset_hotplug(true);
-}
-
void cpuset_update_active_cpus(void)
{
- cpuset_hotplug(false);
+ /*
+ * We're inside cpu hotplug critical region which usually nests
+ * inside cgroup synchronization. Bounce actual hotplug processing
+ * to a work item to avoid reverse locking order.
+ */
+ schedule_work(&cpuset_hotplug_work);
+}
+
+void cpuset_wait_for_hotplug(void)
+{
+ flush_work(&cpuset_hotplug_work);
}
/*
diff --git a/kernel/power/process.c b/kernel/power/process.c
index 08f7019357ee..4b6a54da7e65 100644
--- a/kernel/power/process.c
+++ b/kernel/power/process.c
@@ -204,6 +204,8 @@ void thaw_processes(void)
__usermodehelper_set_disable_depth(UMH_FREEZING);
thaw_workqueues();
+ cpuset_wait_for_hotplug();
+
read_lock(&tasklist_lock);
for_each_process_thread(g, p) {
/* No other threads should have PF_SUSPEND_TASK set */