sched: _cpu_down(): Don't play with current->cpus_allowed

_cpu_down() changes the current task's affinity and then recovers it at
the end. The problems are well known: we can't restore old_allowed if it
was bound to the now-dead-cpu, and we can race with the userspace which
can change cpu-affinity during unplug.

_cpu_down() should not play with current->cpus_allowed at all. Instead,
take_cpu_down() can migrate the caller of _cpu_down() after __cpu_disable()
removes the dying cpu from cpu_online_mask.

Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Acked-by: Rafael J. Wysocki <rjw@sisk.pl>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <20100315091023.GA9148@redhat.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

3 files changed, 8 insertions, 13 deletions

diff --git a/include/linux/sched.h b/include/linux/sched.h
index 43c945152732..8bea40725c76 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1843,6 +1843,7 @@ extern void sched_clock_idle_sleep_event(void);
 extern void sched_clock_idle_wakeup_event(u64 delta_ns);
 
 #ifdef CONFIG_HOTPLUG_CPU
+extern void move_task_off_dead_cpu(int dead_cpu, struct task_struct *p);
 extern void idle_task_exit(void);
 #else
 static inline void idle_task_exit(void) {}
diff --git a/kernel/cpu.c b/kernel/cpu.c
index f8cced2692b3..8d340faac380 100644
--- a/kernel/cpu.c
+++ b/kernel/cpu.c
@@ -163,6 +163,7 @@ static inline void check_for_tasks(int cpu)
 }
 
 struct take_cpu_down_param {
+        struct task_struct *caller;
         unsigned long mod;
         void *hcpu;
 };
@@ -171,6 +172,7 @@ struct take_cpu_down_param {
 static int __ref take_cpu_down(void *_param)
 {
         struct take_cpu_down_param *param = _param;
+        unsigned int cpu = (unsigned long)param->hcpu;
         int err;
 
         /* Ensure this CPU doesn't handle any more interrupts. */
@@ -181,6 +183,8 @@ static int __ref take_cpu_down(void *_param)
         raw_notifier_call_chain(&cpu_chain, CPU_DYING | param->mod,
                                 param->hcpu);
 
+        if (task_cpu(param->caller) == cpu)
+                move_task_off_dead_cpu(cpu, param->caller);
         /* Force idle task to run as soon as we yield: it should
            immediately notice cpu is offline and die quickly. */
         sched_idle_next();
@@ -191,10 +195,10 @@ static int __ref take_cpu_down(void *_param)
 static int __ref _cpu_down(unsigned int cpu, int tasks_frozen)
 {
         int err, nr_calls = 0;
-        cpumask_var_t old_allowed;
         void *hcpu = (void *)(long)cpu;
         unsigned long mod = tasks_frozen ? CPU_TASKS_FROZEN : 0;
         struct take_cpu_down_param tcd_param = {
+                .caller = current,
                 .mod = mod,
                 .hcpu = hcpu,
         };
@@ -205,9 +209,6 @@ static int __ref _cpu_down(unsigned int cpu, int tasks_frozen)
         if (!cpu_online(cpu))
                 return -EINVAL;
 
-        if (!alloc_cpumask_var(&old_allowed, GFP_KERNEL))
-                return -ENOMEM;
-
         cpu_hotplug_begin();
         set_cpu_active(cpu, false);
         err = __raw_notifier_call_chain(&cpu_chain, CPU_DOWN_PREPARE | mod,
@@ -224,10 +225,6 @@ static int __ref _cpu_down(unsigned int cpu, int tasks_frozen)
                 goto out_release;
         }
 
-        /* Ensure that we are not runnable on dying cpu */
-        cpumask_copy(old_allowed, &current->cpus_allowed);
-        set_cpus_allowed_ptr(current, cpu_active_mask);
-
         err = __stop_machine(take_cpu_down, &tcd_param, cpumask_of(cpu));
         if (err) {
                 set_cpu_active(cpu, true);
@@ -236,7 +233,7 @@ static int __ref _cpu_down(unsigned int cpu, int tasks_frozen)
                                             hcpu) == NOTIFY_BAD)
                         BUG();
 
-                goto out_allowed;
+                goto out_release;
         }
         BUG_ON(cpu_online(cpu));
 
@@ -254,8 +251,6 @@ static int __ref _cpu_down(unsigned int cpu, int tasks_frozen)
 
         check_for_tasks(cpu);
 
-out_allowed:
-        set_cpus_allowed_ptr(current, old_allowed);
 out_release:
         cpu_hotplug_done();
         if (!err) {
@@ -263,7 +258,6 @@ out_release:
                                             hcpu) == NOTIFY_BAD)
                         BUG();
         }
-        free_cpumask_var(old_allowed);
         return err;
 }
 
diff --git a/kernel/sched.c b/kernel/sched.c
index 165b532dd8c2..11119deffa48 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -5442,7 +5442,7 @@ static int migration_thread(void *data)
 /*
  * Figure out where task on dead CPU should go, use force if necessary.
  */
-static void move_task_off_dead_cpu(int dead_cpu, struct task_struct *p)
+void move_task_off_dead_cpu(int dead_cpu, struct task_struct *p)
 {
         struct rq *rq = cpu_rq(dead_cpu);
         int needs_cpu, uninitialized_var(dest_cpu);