2013-01-16 17:16:37 +01:00
|
|
|
/*
|
|
|
|
* Context tracking: Probe on high level context boundaries such as kernel
|
|
|
|
* and userspace. This includes syscalls and exceptions entry/exit.
|
|
|
|
*
|
|
|
|
* This is used by RCU to remove its dependency on the timer tick while a CPU
|
|
|
|
* runs in userspace.
|
|
|
|
*
|
|
|
|
* Started by Frederic Weisbecker:
|
|
|
|
*
|
|
|
|
* Copyright (C) 2012 Red Hat, Inc., Frederic Weisbecker <fweisbec@redhat.com>
|
|
|
|
*
|
|
|
|
* Many thanks to Gilad Ben-Yossef, Paul McKenney, Ingo Molnar, Andrew Morton,
|
|
|
|
* Steven Rostedt, Peter Zijlstra for suggestions and improvements.
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
2012-11-27 19:33:25 +01:00
|
|
|
#include <linux/context_tracking.h>
|
|
|
|
#include <linux/rcupdate.h>
|
|
|
|
#include <linux/sched.h>
|
|
|
|
#include <linux/hardirq.h>
|
2012-12-16 20:00:34 +01:00
|
|
|
#include <linux/export.h>
|
2014-06-14 08:47:12 +02:00
|
|
|
#include <linux/kprobes.h>
|
2012-11-27 19:33:25 +01:00
|
|
|
|
2013-07-11 20:27:43 +02:00
|
|
|
#define CREATE_TRACE_POINTS
|
|
|
|
#include <trace/events/context_tracking.h>
|
|
|
|
|
2013-07-11 19:12:32 +02:00
|
|
|
struct static_key context_tracking_enabled = STATIC_KEY_INIT_FALSE;
|
2013-07-10 02:44:35 +02:00
|
|
|
EXPORT_SYMBOL_GPL(context_tracking_enabled);
|
2013-07-11 19:12:32 +02:00
|
|
|
|
|
|
|
DEFINE_PER_CPU(struct context_tracking, context_tracking);
|
2013-07-10 02:44:35 +02:00
|
|
|
EXPORT_SYMBOL_GPL(context_tracking);
|
2012-11-27 19:33:25 +01:00
|
|
|
|
2013-07-10 00:55:25 +02:00
|
|
|
void context_tracking_cpu_set(int cpu)
|
|
|
|
{
|
2013-07-11 19:12:32 +02:00
|
|
|
if (!per_cpu(context_tracking.active, cpu)) {
|
|
|
|
per_cpu(context_tracking.active, cpu) = true;
|
|
|
|
static_key_slow_inc(&context_tracking_enabled);
|
|
|
|
}
|
2013-07-10 00:55:25 +02:00
|
|
|
}
|
|
|
|
|
2013-01-16 17:16:37 +01:00
|
|
|
/**
|
2013-07-10 02:44:35 +02:00
|
|
|
* context_tracking_user_enter - Inform the context tracking that the CPU is going to
|
|
|
|
* enter userspace mode.
|
2013-01-16 17:16:37 +01:00
|
|
|
*
|
|
|
|
* This function must be called right before we switch from the kernel
|
|
|
|
* to userspace, when it's guaranteed the remaining kernel instructions
|
|
|
|
* to execute won't use any RCU read side critical section because this
|
|
|
|
* function sets RCU in extended quiescent state.
|
|
|
|
*/
|
2013-07-10 02:44:35 +02:00
|
|
|
void context_tracking_user_enter(void)
|
2012-11-27 19:33:25 +01:00
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
|
2013-09-10 00:54:17 +02:00
|
|
|
/*
|
|
|
|
* Repeat the user_enter() check here because some archs may be calling
|
|
|
|
* this from asm and if no CPU needs context tracking, they shouldn't
|
2013-11-06 14:45:57 +01:00
|
|
|
* go further. Repeat the check here until they support the inline static
|
|
|
|
* key check.
|
2013-09-10 00:54:17 +02:00
|
|
|
*/
|
2013-11-06 14:45:57 +01:00
|
|
|
if (!context_tracking_is_enabled())
|
2013-09-10 00:54:17 +02:00
|
|
|
return;
|
|
|
|
|
2012-11-27 19:33:25 +01:00
|
|
|
/*
|
|
|
|
* Some contexts may involve an exception occuring in an irq,
|
|
|
|
* leading to that nesting:
|
|
|
|
* rcu_irq_enter() rcu_user_exit() rcu_user_exit() rcu_irq_exit()
|
|
|
|
* This would mess up the dyntick_nesting count though. And rcu_irq_*()
|
|
|
|
* helpers are enough to protect RCU uses inside the exception. So
|
|
|
|
* just return immediately if we detect we are in an IRQ.
|
|
|
|
*/
|
|
|
|
if (in_interrupt())
|
|
|
|
return;
|
|
|
|
|
2013-01-16 17:16:37 +01:00
|
|
|
/* Kernel threads aren't supposed to go to userspace */
|
2012-11-27 19:33:25 +01:00
|
|
|
WARN_ON_ONCE(!current->mm);
|
|
|
|
|
|
|
|
local_irq_save(flags);
|
2013-07-11 23:59:33 +02:00
|
|
|
if ( __this_cpu_read(context_tracking.state) != IN_USER) {
|
|
|
|
if (__this_cpu_read(context_tracking.active)) {
|
2013-07-11 20:27:43 +02:00
|
|
|
trace_user_enter(0);
|
2013-07-11 23:59:33 +02:00
|
|
|
/*
|
|
|
|
* At this stage, only low level arch entry code remains and
|
|
|
|
* then we'll run in userspace. We can assume there won't be
|
|
|
|
* any RCU read-side critical section until the next call to
|
|
|
|
* user_exit() or rcu_irq_enter(). Let's remove RCU's dependency
|
|
|
|
* on the tick.
|
|
|
|
*/
|
|
|
|
vtime_user_enter(current);
|
|
|
|
rcu_user_enter();
|
|
|
|
}
|
2013-01-16 17:16:37 +01:00
|
|
|
/*
|
2013-07-11 23:59:33 +02:00
|
|
|
* Even if context tracking is disabled on this CPU, because it's outside
|
|
|
|
* the full dynticks mask for example, we still have to keep track of the
|
|
|
|
* context transitions and states to prevent inconsistency on those of
|
|
|
|
* other CPUs.
|
|
|
|
* If a task triggers an exception in userspace, sleep on the exception
|
|
|
|
* handler and then migrate to another CPU, that new CPU must know where
|
|
|
|
* the exception returns by the time we call exception_exit().
|
|
|
|
* This information can only be provided by the previous CPU when it called
|
|
|
|
* exception_enter().
|
|
|
|
* OTOH we can spare the calls to vtime and RCU when context_tracking.active
|
|
|
|
* is false because we know that CPU is not tickless.
|
2013-01-16 17:16:37 +01:00
|
|
|
*/
|
2012-07-25 07:56:04 +02:00
|
|
|
__this_cpu_write(context_tracking.state, IN_USER);
|
2012-11-27 19:33:25 +01:00
|
|
|
}
|
|
|
|
local_irq_restore(flags);
|
|
|
|
}
|
2014-06-14 08:47:12 +02:00
|
|
|
NOKPROBE_SYMBOL(context_tracking_user_enter);
|
2012-11-27 19:33:25 +01:00
|
|
|
|
2013-01-16 17:16:37 +01:00
|
|
|
/**
|
2013-07-10 02:44:35 +02:00
|
|
|
* context_tracking_user_exit - Inform the context tracking that the CPU is
|
|
|
|
* exiting userspace mode and entering the kernel.
|
2013-01-16 17:16:37 +01:00
|
|
|
*
|
|
|
|
* This function must be called after we entered the kernel from userspace
|
|
|
|
* before any use of RCU read side critical section. This potentially include
|
|
|
|
* any high level kernel code like syscalls, exceptions, signal handling, etc...
|
|
|
|
*
|
|
|
|
* This call supports re-entrancy. This way it can be called from any exception
|
|
|
|
* handler without needing to know if we came from userspace or not.
|
|
|
|
*/
|
2013-07-10 02:44:35 +02:00
|
|
|
void context_tracking_user_exit(void)
|
2012-11-27 19:33:25 +01:00
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
|
2013-11-06 14:45:57 +01:00
|
|
|
if (!context_tracking_is_enabled())
|
2013-09-10 00:54:17 +02:00
|
|
|
return;
|
|
|
|
|
2012-11-27 19:33:25 +01:00
|
|
|
if (in_interrupt())
|
|
|
|
return;
|
|
|
|
|
|
|
|
local_irq_save(flags);
|
|
|
|
if (__this_cpu_read(context_tracking.state) == IN_USER) {
|
2013-07-11 23:59:33 +02:00
|
|
|
if (__this_cpu_read(context_tracking.active)) {
|
|
|
|
/*
|
|
|
|
* We are going to run code that may use RCU. Inform
|
|
|
|
* RCU core about that (ie: we may need the tick again).
|
|
|
|
*/
|
|
|
|
rcu_user_exit();
|
|
|
|
vtime_user_exit(current);
|
2013-07-11 20:27:43 +02:00
|
|
|
trace_user_exit(0);
|
2013-07-11 23:59:33 +02:00
|
|
|
}
|
2012-07-25 07:56:04 +02:00
|
|
|
__this_cpu_write(context_tracking.state, IN_KERNEL);
|
2012-11-27 19:33:25 +01:00
|
|
|
}
|
|
|
|
local_irq_restore(flags);
|
|
|
|
}
|
2014-06-14 08:47:12 +02:00
|
|
|
NOKPROBE_SYMBOL(context_tracking_user_exit);
|
2012-11-27 19:33:25 +01:00
|
|
|
|
2013-01-16 17:16:37 +01:00
|
|
|
/**
|
2013-07-11 19:42:13 +02:00
|
|
|
* __context_tracking_task_switch - context switch the syscall callbacks
|
2013-01-16 17:16:37 +01:00
|
|
|
* @prev: the task that is being switched out
|
|
|
|
* @next: the task that is being switched in
|
|
|
|
*
|
|
|
|
* The context tracking uses the syscall slow path to implement its user-kernel
|
|
|
|
* boundaries probes on syscalls. This way it doesn't impact the syscall fast
|
|
|
|
* path on CPUs that don't do context tracking.
|
|
|
|
*
|
|
|
|
* But we need to clear the flag on the previous task because it may later
|
|
|
|
* migrate to some CPU that doesn't do the context tracking. As such the TIF
|
|
|
|
* flag may not be desired there.
|
|
|
|
*/
|
2013-07-11 19:42:13 +02:00
|
|
|
void __context_tracking_task_switch(struct task_struct *prev,
|
|
|
|
struct task_struct *next)
|
2012-11-27 19:33:25 +01:00
|
|
|
{
|
2013-07-11 23:59:33 +02:00
|
|
|
clear_tsk_thread_flag(prev, TIF_NOHZ);
|
|
|
|
set_tsk_thread_flag(next, TIF_NOHZ);
|
2012-11-27 19:33:25 +01:00
|
|
|
}
|
2013-07-11 19:12:32 +02:00
|
|
|
|
|
|
|
#ifdef CONFIG_CONTEXT_TRACKING_FORCE
|
|
|
|
void __init context_tracking_init(void)
|
|
|
|
{
|
|
|
|
int cpu;
|
|
|
|
|
|
|
|
for_each_possible_cpu(cpu)
|
|
|
|
context_tracking_cpu_set(cpu);
|
|
|
|
}
|
|
|
|
#endif
|