Release 4.11 fs/timerfd.c

/*
 *  fs/timerfd.c
 *
 *  Copyright (C) 2007  Davide Libenzi <davidel@xmailserver.org>
 *
 *
 *  Thanks to Thomas Gleixner for code reviews and useful comments.
 *
 */

#include <linux/alarmtimer.h>
#include <linux/file.h>
#include <linux/poll.h>
#include <linux/init.h>
#include <linux/fs.h>
#include <linux/sched.h>
#include <linux/kernel.h>
#include <linux/slab.h>
#include <linux/list.h>
#include <linux/spinlock.h>
#include <linux/time.h>
#include <linux/hrtimer.h>
#include <linux/anon_inodes.h>
#include <linux/timerfd.h>
#include <linux/syscalls.h>
#include <linux/compat.h>
#include <linux/rcupdate.h>


struct timerfd_ctx {
	union {
		
struct hrtimer tmr;
		
struct alarm alarm;
	} 
t;
	
ktime_t tintv;
	
ktime_t moffs;
	
wait_queue_head_t wqh;
	
u64 ticks;
	
int clockid;
	
short unsigned expired;
	
short unsigned settime_flags;	/* to show in fdinfo */
	
struct rcu_head rcu;
	
struct list_head clist;
	
spinlock_t cancel_lock;
	
bool might_cancel;
};

static LIST_HEAD(cancel_list);
static DEFINE_SPINLOCK(cancel_lock);



static inline bool isalarm(struct timerfd_ctx *ctx)
{
	return ctx->clockid == CLOCK_REALTIME_ALARM ||
		ctx->clockid == CLOCK_BOOTTIME_ALARM;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Todd Android Poynor
19
76.00%
1
50.00%
Davide Libenzi
6
24.00%
1
50.00%
Total
25
100.00%
2
100.00%

/*
 * This gets called when the timer event triggers. We set the "expired"
 * flag, but we do not re-arm the timer (in case it's necessary,
 * tintv != 0) until the timer is accessed.
 */


static void timerfd_triggered(struct timerfd_ctx *ctx)
{
	unsigned long flags;

	spin_lock_irqsave(&ctx->wqh.lock, flags);
	ctx->expired = 1;
	ctx->ticks++;
	wake_up_locked(&ctx->wqh);
	spin_unlock_irqrestore(&ctx->wqh.lock, flags);
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Davide Libenzi
51
87.93%
3
75.00%
Todd Android Poynor
7
12.07%
1
25.00%
Total
58
100.00%
4
100.00%



static enum hrtimer_restart timerfd_tmrproc(struct hrtimer *htmr)
{
	struct timerfd_ctx *ctx = container_of(htmr, struct timerfd_ctx,
					       t.tmr);
	timerfd_triggered(ctx);
	return HRTIMER_NORESTART;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Todd Android Poynor
33
89.19%
1
50.00%
Davide Libenzi
4
10.81%
1
50.00%
Total
37
100.00%
2
100.00%



static enum alarmtimer_restart timerfd_alarmproc(struct alarm *alarm,
	ktime_t now)
{
	struct timerfd_ctx *ctx = container_of(alarm, struct timerfd_ctx,
					       t.alarm);
	timerfd_triggered(ctx);
	return ALARMTIMER_NORESTART;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Todd Android Poynor
40
100.00%
1
100.00%
Total
40
100.00%
1
100.00%

/*
 * Called when the clock was set to cancel the timers in the cancel
 * list. This will wake up processes waiting on these timers. The
 * wake-up requires ctx->ticks to be non zero, therefore we increment
 * it before calling wake_up_locked().
 */


void timerfd_clock_was_set(void)
{
	ktime_t moffs = ktime_mono_to_real(0);
	struct timerfd_ctx *ctx;
	unsigned long flags;

	rcu_read_lock();
	list_for_each_entry_rcu(ctx, &cancel_list, clist) {
		if (!ctx->might_cancel)
			continue;
		spin_lock_irqsave(&ctx->wqh.lock, flags);
		if (ctx->moffs != moffs) {
			ctx->moffs = KTIME_MAX;
			ctx->ticks++;
			wake_up_locked(&ctx->wqh);
		}
		spin_unlock_irqrestore(&ctx->wqh.lock, flags);
	}
	rcu_read_unlock();
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Thomas Gleixner
81
80.20%
2
40.00%
Davide Libenzi
14
13.86%
1
20.00%
Max Asbock
5
4.95%
1
20.00%
Arjan van de Ven
1
0.99%
1
20.00%
Total
101
100.00%
5
100.00%



static void __timerfd_remove_cancel(struct timerfd_ctx *ctx)
{
	if (ctx->might_cancel) {
		ctx->might_cancel = false;
		spin_lock(&cancel_lock);
		list_del_rcu(&ctx->clist);
		spin_unlock(&cancel_lock);
	}
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Thomas Gleixner
44
97.78%
2
66.67%
Davide Libenzi
1
2.22%
1
33.33%
Total
45
100.00%
3
100.00%



static void timerfd_remove_cancel(struct timerfd_ctx *ctx)
{
	spin_lock(&ctx->cancel_lock);
	__timerfd_remove_cancel(ctx);
	spin_unlock(&ctx->cancel_lock);
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Thomas Gleixner
32
100.00%
1
100.00%
Total
32
100.00%
1
100.00%



static bool timerfd_canceled(struct timerfd_ctx *ctx)
{
	if (!ctx->might_cancel || ctx->moffs != KTIME_MAX)
		return false;
	ctx->moffs = ktime_mono_to_real(0);
	return true;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Thomas Gleixner
39
100.00%
3
100.00%
Total
39
100.00%
3
100.00%



static void timerfd_setup_cancel(struct timerfd_ctx *ctx, int flags)
{
	spin_lock(&ctx->cancel_lock);
	if ((ctx->clockid == CLOCK_REALTIME ||
	     ctx->clockid == CLOCK_REALTIME_ALARM) &&
	    (flags & TFD_TIMER_ABSTIME) && (flags & TFD_TIMER_CANCEL_ON_SET)) {
		if (!ctx->might_cancel) {
			ctx->might_cancel = true;
			spin_lock(&cancel_lock);
			list_add_rcu(&ctx->clist, &cancel_list);
			spin_unlock(&cancel_lock);
		}
	} else {
		__timerfd_remove_cancel(ctx);
	}
	spin_unlock(&ctx->cancel_lock);
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Thomas Gleixner
98
92.45%
3
75.00%
Todd Android Poynor
8
7.55%
1
25.00%
Total
106
100.00%
4
100.00%



static ktime_t timerfd_get_remaining(struct timerfd_ctx *ctx)
{
	ktime_t remaining;

	if (isalarm(ctx))
		remaining = alarm_expires_remaining(&ctx->t.alarm);
	else
		remaining = hrtimer_expires_remaining_adjusted(&ctx->t.tmr);

	return remaining < 0 ? 0: remaining;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Thomas Gleixner
33
60.00%
3
75.00%
Todd Android Poynor
22
40.00%
1
25.00%
Total
55
100.00%
4
100.00%



static int timerfd_setup(struct timerfd_ctx *ctx, int flags,
			 const struct itimerspec *ktmr)
{
	enum hrtimer_mode htmode;
	ktime_t texp;
	int clockid = ctx->clockid;

	htmode = (flags & TFD_TIMER_ABSTIME) ?
		HRTIMER_MODE_ABS: HRTIMER_MODE_REL;

	texp = timespec_to_ktime(ktmr->it_value);
	ctx->expired = 0;
	ctx->ticks = 0;
	ctx->tintv = timespec_to_ktime(ktmr->it_interval);

	if (isalarm(ctx)) {
		alarm_init(&ctx->t.alarm,
			   ctx->clockid == CLOCK_REALTIME_ALARM ?
			   ALARM_REALTIME : ALARM_BOOTTIME,
			   timerfd_alarmproc);
	} else {
		hrtimer_init(&ctx->t.tmr, clockid, htmode);
		hrtimer_set_expires(&ctx->t.tmr, texp);
		ctx->t.tmr.function = timerfd_tmrproc;
	}

	if (texp != 0) {
		if (isalarm(ctx)) {
			if (flags & TFD_TIMER_ABSTIME)
				alarm_start(&ctx->t.alarm, texp);
			else
				alarm_start_relative(&ctx->t.alarm, texp);
		} else {
			hrtimer_start(&ctx->t.tmr, texp, htmode);
		}

		if (timerfd_canceled(ctx))
			return -ECANCELED;
	}

	ctx->settime_flags = flags & TFD_SETTIME_FLAGS;
	return 0;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Davide Libenzi
113
48.09%
2
33.33%
Todd Android Poynor
85
36.17%
1
16.67%
Thomas Gleixner
24
10.21%
1
16.67%
Cyrill V. Gorcunov
8
3.40%
1
16.67%
Arjan van de Ven
5
2.13%
1
16.67%
Total
235
100.00%
6
100.00%



static int timerfd_release(struct inode *inode, struct file *file)
{
	struct timerfd_ctx *ctx = file->private_data;

	timerfd_remove_cancel(ctx);

	if (isalarm(ctx))
		alarm_cancel(&ctx->t.alarm);
	else
		hrtimer_cancel(&ctx->t.tmr);
	kfree_rcu(ctx, rcu);
	return 0;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Davide Libenzi
40
58.82%
1
33.33%
Todd Android Poynor
20
29.41%
1
33.33%
Thomas Gleixner
8
11.76%
1
33.33%
Total
68
100.00%
3
100.00%



static unsigned int timerfd_poll(struct file *file, poll_table *wait)
{
	struct timerfd_ctx *ctx = file->private_data;
	unsigned int events = 0;
	unsigned long flags;

	poll_wait(file, &ctx->wqh, wait);

	spin_lock_irqsave(&ctx->wqh.lock, flags);
	if (ctx->ticks)
		events |= POLLIN;
	spin_unlock_irqrestore(&ctx->wqh.lock, flags);

	return events;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Davide Libenzi
84
100.00%
3
100.00%
Total
84
100.00%
3
100.00%



static ssize_t timerfd_read(struct file *file, char __user *buf, size_t count,
			    loff_t *ppos)
{
	struct timerfd_ctx *ctx = file->private_data;
	ssize_t res;
	u64 ticks = 0;

	if (count < sizeof(ticks))
		return -EINVAL;
	spin_lock_irq(&ctx->wqh.lock);
	if (file->f_flags & O_NONBLOCK)
		res = -EAGAIN;
	else
		res = wait_event_interruptible_locked_irq(ctx->wqh, ctx->ticks);

	/*
         * If clock has changed, we do not care about the
         * ticks and we do not rearm the timer. Userspace must
         * reevaluate anyway.
         */
	if (timerfd_canceled(ctx)) {
		ctx->ticks = 0;
		ctx->expired = 0;
		res = -ECANCELED;
	}

	if (ctx->ticks) {
		ticks = ctx->ticks;

		if (ctx->expired && ctx->tintv) {
			/*
                         * If tintv != 0, this is a periodic timer that
                         * needs to be re-armed. We avoid doing it in the timer
                         * callback to avoid DoS attacks specifying a very
                         * short timer period.
                         */
			if (isalarm(ctx)) {
				ticks += alarm_forward_now(
					&ctx->t.alarm, ctx->tintv) - 1;
				alarm_restart(&ctx->t.alarm);
			} else {
				ticks += hrtimer_forward_now(&ctx->t.tmr,
							     ctx->tintv) - 1;
				hrtimer_restart(&ctx->t.tmr);
			}
		}
		ctx->expired = 0;
		ctx->ticks = 0;
	}
	spin_unlock_irq(&ctx->wqh.lock);
	if (ticks)
		res = put_user(ticks, (u64 __user *) buf) ? -EFAULT: sizeof(ticks);
	return res;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Davide Libenzi
172
65.40%
4
44.44%
Todd Android Poynor
44
16.73%
1
11.11%
Thomas Gleixner
41
15.59%
3
33.33%
Michal Nazarewicz
6
2.28%
1
11.11%
Total
263
100.00%
9
100.00%

#ifdef CONFIG_PROC_FS


static void timerfd_show(struct seq_file *m, struct file *file)
{
	struct timerfd_ctx *ctx = file->private_data;
	struct itimerspec t;

	spin_lock_irq(&ctx->wqh.lock);
	t.it_value = ktime_to_timespec(timerfd_get_remaining(ctx));
	t.it_interval = ktime_to_timespec(ctx->tintv);
	spin_unlock_irq(&ctx->wqh.lock);

	seq_printf(m,
		   "clockid: %d\n"
		   "ticks: %llu\n"
		   "settime flags: 0%o\n"
		   "it_value: (%llu, %llu)\n"
		   "it_interval: (%llu, %llu)\n",
		   ctx->clockid,
		   (unsigned long long)ctx->ticks,
		   ctx->settime_flags,
		   (unsigned long long)t.it_value.tv_sec,
		   (unsigned long long)t.it_value.tv_nsec,
		   (unsigned long long)t.it_interval.tv_sec,
		   (unsigned long long)t.it_interval.tv_nsec);
}
Contributors
Person
Tokens
Prop
Commits
CommitProp
Cyrill V. Gorcunov
142
98.61%
1
50.00%
Joe Perches
2
1.39%
1
50.00%
Total
144
100.00%
2
100.00%

#else

#define timerfd_show NULL
#endif

#ifdef CONFIG_CHECKPOINT_RESTORE


static long timerfd_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
{
	struct timerfd_ctx *ctx = file->private_data;
	int ret = 0;

	switch (cmd) {
	case TFD_IOC_SET_TICKS: {
		u64 ticks;

		if (copy_from_user(&ticks, (u64 __user *)arg, sizeof(ticks)))
			return -EFAULT;
		if (!ticks)
			return -EINVAL;

		spin_lock_irq(&ctx->wqh.lock);
		if (!timerfd_canceled(ctx)) {
			ctx->ticks = ticks;
			wake_up_locked(&ctx->wqh);
		} else
			ret = -ECANCELED;
		spin_unlock_irq(&ctx->wqh.lock);
		break;
	}
	default:
		ret = -ENOTTY;
		break;
	}

	return ret;
}
Contributors
Person
Tokens
Prop
Commits
CommitProp
Cyrill V. Gorcunov
141
100.00%
1
100.00%
Total
141
100.00%
1
100.00%

#else

#define timerfd_ioctl NULL
#endif


static const struct file_operations timerfd_fops = {
	.release	= timerfd_release,
	.poll		= timerfd_poll,
	.read		= timerfd_read,
	.llseek		= noop_llseek,
	.show_fdinfo	= timerfd_show,
	.unlocked_ioctl	= timerfd_ioctl,
};



static int timerfd_fget(int fd, struct fd *p)
{
	struct fd f = fdget(fd);
	if (!f.file)
		return -EBADF;
	if (f.file->f_op != &timerfd_fops) {
		fdput(f);
		return -EINVAL;
	}
	*p = f;
	return 0;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Davide Libenzi
41
64.06%
1
33.33%
Al Viro
23
35.94%
2
66.67%
Total
64
100.00%
3
100.00%


SYSCALL_DEFINE2(timerfd_create, int, clockid, int, flags)
{
	int ufd;
	struct timerfd_ctx *ctx;

	/* Check the TFD_* constants for consistency.  */
	BUILD_BUG_ON(TFD_CLOEXEC != O_CLOEXEC);
	BUILD_BUG_ON(TFD_NONBLOCK != O_NONBLOCK);

	if ((flags & ~TFD_CREATE_FLAGS) ||
	    (clockid != CLOCK_MONOTONIC &&
	     clockid != CLOCK_REALTIME &&
	     clockid != CLOCK_REALTIME_ALARM &&
	     clockid != CLOCK_BOOTTIME &&
	     clockid != CLOCK_BOOTTIME_ALARM))
		return -EINVAL;

	if ((clockid == CLOCK_REALTIME_ALARM ||
	     clockid == CLOCK_BOOTTIME_ALARM) &&
	    !capable(CAP_WAKE_ALARM))
		return -EPERM;

	ctx = kzalloc(sizeof(*ctx), GFP_KERNEL);
	if (!ctx)
		return -ENOMEM;

	init_waitqueue_head(&ctx->wqh);
	spin_lock_init(&ctx->cancel_lock);
	ctx->clockid = clockid;

	if (isalarm(ctx))
		alarm_init(&ctx->t.alarm,
			   ctx->clockid == CLOCK_REALTIME_ALARM ?
			   ALARM_REALTIME : ALARM_BOOTTIME,
			   timerfd_alarmproc);
	else
		hrtimer_init(&ctx->t.tmr, clockid, HRTIMER_MODE_ABS);

	ctx->moffs = ktime_mono_to_real(0);

	ufd = anon_inode_getfd("[timerfd]", &timerfd_fops, ctx,
			       O_RDWR | (flags & TFD_SHARED_FCNTL_FLAGS));
	if (ufd < 0)
		kfree(ctx);

	return ufd;
}



static int do_timerfd_settime(int ufd, int flags, 
		const struct itimerspec *new,
		struct itimerspec *old)
{
	struct fd f;
	struct timerfd_ctx *ctx;
	int ret;

	if ((flags & ~TFD_SETTIME_FLAGS) ||
	    !timespec_valid(&new->it_value) ||
	    !timespec_valid(&new->it_interval))
		return -EINVAL;

	ret = timerfd_fget(ufd, &f);
	if (ret)
		return ret;
	ctx = f.file->private_data;

	if (isalarm(ctx) && !capable(CAP_WAKE_ALARM)) {
		fdput(f);
		return -EPERM;
	}

	timerfd_setup_cancel(ctx, flags);

	/*
         * We need to stop the existing timer before reprogramming
         * it to the new values.
         */
	for (;;) {
		spin_lock_irq(&ctx->wqh.lock);

		if (isalarm(ctx)) {
			if (alarm_try_to_cancel(&ctx->t.alarm) >= 0)
				break;
		} else {
			if (hrtimer_try_to_cancel(&ctx->t.tmr) >= 0)
				break;
		}
		spin_unlock_irq(&ctx->wqh.lock);
		cpu_relax();
	}

	/*
         * If the timer is expired and it's periodic, we need to advance it
         * because the caller may want to know the previous expiration time.
         * We do not update "ticks" and "expired" since the timer will be
         * re-programmed again in the following timerfd_setup() call.
         */
	if (ctx->expired && ctx->tintv) {
		if (isalarm(ctx))
			alarm_forward_now(&ctx->t.alarm, ctx->tintv);
		else
			hrtimer_forward_now(&ctx->t.tmr, ctx->tintv);
	}

	old->it_value = ktime_to_timespec(timerfd_get_remaining(ctx));
	old->it_interval = ktime_to_timespec(ctx->tintv);

	/*
         * Re-program the timer to the new value ...
         */
	ret = timerfd_setup(ctx, flags, new);

	spin_unlock_irq(&ctx->wqh.lock);
	fdput(f);
	return ret;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Davide Libenzi
163
54.88%
4
33.33%
Todd Android Poynor
55
18.52%
1
8.33%
Al Viro
42
14.14%
3
25.00%
Eric Caruso
19
6.40%
1
8.33%
Thomas Gleixner
13
4.38%
2
16.67%
Stephen D. Smalley
5
1.68%
1
8.33%
Total
297
100.00%
12
100.00%



static int do_timerfd_gettime(int ufd, struct itimerspec *t)
{
	struct fd f;
	struct timerfd_ctx *ctx;
	int ret = timerfd_fget(ufd, &f);
	if (ret)
		return ret;
	ctx = f.file->private_data;

	spin_lock_irq(&ctx->wqh.lock);
	if (ctx->expired && ctx->tintv) {
		ctx->expired = 0;

		if (isalarm(ctx)) {
			ctx->ticks +=
				alarm_forward_now(
					&ctx->t.alarm, ctx->tintv) - 1;
			alarm_restart(&ctx->t.alarm);
		} else {
			ctx->ticks +=
				hrtimer_forward_now(&ctx->t.tmr, ctx->tintv)
				- 1;
			hrtimer_restart(&ctx->t.tmr);
		}
	}
	t->it_value = ktime_to_timespec(timerfd_get_remaining(ctx));
	t->it_interval = ktime_to_timespec(ctx->tintv);
	spin_unlock_irq(&ctx->wqh.lock);
	fdput(f);
	return 0;
}

Contributors
Person
Tokens
Prop
Commits
CommitProp
Davide Libenzi
110
57.89%
2
33.33%
Todd Android Poynor
46
24.21%
1
16.67%
Al Viro
34
17.89%
3
50.00%
Total
190
100.00%
6
100.00%


SYSCALL_DEFINE4(timerfd_settime, int, ufd, int, flags,
		const struct itimerspec __user *, utmr,
		struct itimerspec __user *, otmr)
{
	struct itimerspec new, old;
	int ret;

	if (copy_from_user(&new, utmr, sizeof(new)))
		return -EFAULT;
	ret = do_timerfd_settime(ufd, flags, &new, &old);
	if (ret)
		return ret;
	if (otmr && copy_to_user(otmr, &old, sizeof(old)))
		return -EFAULT;

	return ret;
}


SYSCALL_DEFINE2(timerfd_gettime, int, ufd, struct itimerspec __user *, otmr)
{
	struct itimerspec kotmr;
	int ret = do_timerfd_gettime(ufd, &kotmr);
	if (ret)
		return ret;
	return copy_to_user(otmr, &kotmr, sizeof(kotmr)) ? -EFAULT: 0;
}

#ifdef CONFIG_COMPAT

COMPAT_SYSCALL_DEFINE4(timerfd_settime, int, ufd, int, flags,
		const struct compat_itimerspec __user *, utmr,
		struct compat_itimerspec __user *, otmr)
{
	struct itimerspec new, old;
	int ret;

	if (get_compat_itimerspec(&new, utmr))
		return -EFAULT;
	ret = do_timerfd_settime(ufd, flags, &new, &old);
	if (ret)
		return ret;
	if (otmr && put_compat_itimerspec(otmr, &old))
		return -EFAULT;
	return ret;
}


COMPAT_SYSCALL_DEFINE2(timerfd_gettime, int, ufd,
		struct compat_itimerspec __user *, otmr)
{
	struct itimerspec kotmr;
	int ret = do_timerfd_gettime(ufd, &kotmr);
	if (ret)
		return ret;
	return put_compat_itimerspec(otmr, &kotmr) ? -EFAULT: 0;
}
#endif
Overall Contributors
Person
Tokens
Prop
Commits
CommitProp
Davide Libenzi
1000
36.72%
5
14.29%
Thomas Gleixner
463
17.00%
6
17.14%
Todd Android Poynor
431
15.83%
1
2.86%
Al Viro
374
13.73%
4
11.43%
Cyrill V. Gorcunov
332
12.19%
2
5.71%
Eric Caruso
35
1.29%
1
2.86%
Ulrich Drepper
21
0.77%
4
11.43%
Heiko Carstens
17
0.62%
2
5.71%
Stephen D. Smalley
11
0.40%
1
2.86%
Michal Nazarewicz
6
0.22%
1
2.86%
Arjan van de Ven
6
0.22%
1
2.86%
Max Asbock
6
0.22%
1
2.86%
Arnd Bergmann
5
0.18%
1
2.86%
Greg Hackmann
4
0.15%
1
2.86%
Roland Dreier
4
0.15%
1
2.86%
Tejun Heo
3
0.11%
1
2.86%
Adrian Bunk
3
0.11%
1
2.86%
Joe Perches
2
0.07%
1
2.86%
Total
2723
100.00%
35
100.00%
Directory: fs

Information contained on this website is for historical information purposes only and does not indicate or represent copyright ownership.