Merge branch 'timers-for-linus' of git://git.kernel.org/pub/scm/linux…

…/kernel/git/tip/linux-2.6-tip

* 'timers-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
  hrtimers: Reorder clock bases
  hrtimers: Avoid touching inactive timer bases
  hrtimers: Make struct hrtimer_cpu_base layout less stupid
  timerfd: Manage cancelable timers in timerfd
  clockevents: Move C3 stop test outside lock
  alarmtimer: Drop device refcount after rtc_open()
  alarmtimer: Check return value of class_find_device()
  timerfd: Allow timers to be cancelled when clock was set
  hrtimers: Prepare for cancel on clock was set timers

fs/timerfd.c

@@ -22,16 +22,24 @@
 #include <linux/anon_inodes.h>
 #include <linux/timerfd.h>
 #include <linux/syscalls.h>
+#include <linux/rcupdate.h>
 
 struct timerfd_ctx {
 	struct hrtimer tmr;
 	ktime_t tintv;
+	ktime_t moffs;
 	wait_queue_head_t wqh;
 	u64 ticks;
 	int expired;
 	int clockid;
+	struct rcu_head rcu;
+	struct list_head clist;
+	bool might_cancel;
 };
 
+static LIST_HEAD(cancel_list);
+static DEFINE_SPINLOCK(cancel_lock);
+
 /*
  * 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,
@@ -51,6 +59,63 @@ static enum hrtimer_restart timerfd_tmrproc(struct hrtimer *htmr)
 	return HRTIMER_NORESTART;
 }
 
+/*
+ * Called when the clock was set to cancel the timers in the cancel
+ * list.
+ */
+void timerfd_clock_was_set(void)
+{
+	ktime_t moffs = ktime_get_monotonic_offset();
+	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.tv64 != moffs.tv64) {
+			ctx->moffs.tv64 = KTIME_MAX;
+			wake_up_locked(&ctx->wqh);
+		}
+		spin_unlock_irqrestore(&ctx->wqh.lock, flags);
+	}
+	rcu_read_unlock();
+}
+
+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);
+	}
+}
+
+static bool timerfd_canceled(struct timerfd_ctx *ctx)
+{
+	if (!ctx->might_cancel || ctx->moffs.tv64 != KTIME_MAX)
+		return false;
+	ctx->moffs = ktime_get_monotonic_offset();
+	return true;
+}
+
+static void timerfd_setup_cancel(struct timerfd_ctx *ctx, int flags)
+{
+	if (ctx->clockid == CLOCK_REALTIME && (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 if (ctx->might_cancel) {
+		timerfd_remove_cancel(ctx);
+	}
+}
+
 static ktime_t timerfd_get_remaining(struct timerfd_ctx *ctx)
 {
 	ktime_t remaining;
@@ -59,11 +124,12 @@ static ktime_t timerfd_get_remaining(struct timerfd_ctx *ctx)
 	return remaining.tv64 < 0 ? ktime_set(0, 0): remaining;
 }
 
-static void timerfd_setup(struct timerfd_ctx *ctx, int flags,
-			  const struct itimerspec *ktmr)
+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;
@@ -72,19 +138,24 @@ static void timerfd_setup(struct timerfd_ctx *ctx, int flags,
 	ctx->expired = 0;
 	ctx->ticks = 0;
 	ctx->tintv = timespec_to_ktime(ktmr->it_interval);
-	hrtimer_init(&ctx->tmr, ctx->clockid, htmode);
+	hrtimer_init(&ctx->tmr, clockid, htmode);
 	hrtimer_set_expires(&ctx->tmr, texp);
 	ctx->tmr.function = timerfd_tmrproc;
-	if (texp.tv64 != 0)
+	if (texp.tv64 != 0) {
 		hrtimer_start(&ctx->tmr, texp, htmode);
+		if (timerfd_canceled(ctx))
+			return -ECANCELED;
+	}
+	return 0;
 }
 
 static int timerfd_release(struct inode *inode, struct file *file)
 {
 	struct timerfd_ctx *ctx = file->private_data;
 
+	timerfd_remove_cancel(ctx);
 	hrtimer_cancel(&ctx->tmr);
-	kfree(ctx);
+	kfree_rcu(ctx, rcu);
 	return 0;
 }
 
@@ -118,8 +189,21 @@ static ssize_t timerfd_read(struct file *file, char __user *buf, size_t count,
 		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.tv64) {
 			/*
 			 * If tintv.tv64 != 0, this is a periodic timer that
@@ -183,6 +267,7 @@ SYSCALL_DEFINE2(timerfd_create, int, clockid, int, flags)
 	init_waitqueue_head(&ctx->wqh);
 	ctx->clockid = clockid;
 	hrtimer_init(&ctx->tmr, clockid, HRTIMER_MODE_ABS);
+	ctx->moffs = ktime_get_monotonic_offset();
 
 	ufd = anon_inode_getfd("[timerfd]", &timerfd_fops, ctx,
 			       O_RDWR | (flags & TFD_SHARED_FCNTL_FLAGS));
@@ -199,6 +284,7 @@ SYSCALL_DEFINE4(timerfd_settime, int, ufd, int, flags,
 	struct file *file;
 	struct timerfd_ctx *ctx;
 	struct itimerspec ktmr, kotmr;
+	int ret;
 
 	if (copy_from_user(&ktmr, utmr, sizeof(ktmr)))
 		return -EFAULT;
@@ -213,6 +299,8 @@ SYSCALL_DEFINE4(timerfd_settime, int, ufd, int, flags,
 		return PTR_ERR(file);
 	ctx = file->private_data;
 
+	timerfd_setup_cancel(ctx, flags);
+
 	/*
 	 * We need to stop the existing timer before reprogramming
 	 * it to the new values.
@@ -240,14 +328,14 @@ SYSCALL_DEFINE4(timerfd_settime, int, ufd, int, flags,
 	/*
 	 * Re-program the timer to the new value ...
 	 */
-	timerfd_setup(ctx, flags, &ktmr);
+	ret = timerfd_setup(ctx, flags, &ktmr);
 
 	spin_unlock_irq(&ctx->wqh.lock);
 	fput(file);
 	if (otmr && copy_to_user(otmr, &kotmr, sizeof(kotmr)))
 		return -EFAULT;
 
-	return 0;
+	return ret;
 }
 
 SYSCALL_DEFINE2(timerfd_gettime, int, ufd, struct itimerspec __user *, otmr)

include/linux/hrtimer.h

@@ -143,19 +143,18 @@ struct hrtimer_sleeper {
  */
 struct hrtimer_clock_base {
 	struct hrtimer_cpu_base	*cpu_base;
-	clockid_t		index;
+	int			index;
+	clockid_t		clockid;
 	struct timerqueue_head	active;
 	ktime_t			resolution;
 	ktime_t			(*get_time)(void);
 	ktime_t			softirq_time;
-#ifdef CONFIG_HIGH_RES_TIMERS
 	ktime_t			offset;
-#endif
 };
 
 enum  hrtimer_base_type {
-	HRTIMER_BASE_REALTIME,
 	HRTIMER_BASE_MONOTONIC,
+	HRTIMER_BASE_REALTIME,
 	HRTIMER_BASE_BOOTTIME,
 	HRTIMER_MAX_CLOCK_BASES,
 };
@@ -164,7 +163,7 @@ enum  hrtimer_base_type {
  * struct hrtimer_cpu_base - the per cpu clock bases
  * @lock:		lock protecting the base and associated clock bases
  *			and timers
- * @clock_base:		array of clock bases for this cpu
+ * @active_bases:	Bitfield to mark bases with active timers
  * @expires_next:	absolute time of the next event which was scheduled
  *			via clock_set_next_event()
  * @hres_active:	State of high resolution mode
@@ -173,10 +172,11 @@ enum  hrtimer_base_type {
  * @nr_retries:		Total number of hrtimer interrupt retries
  * @nr_hangs:		Total number of hrtimer interrupt hangs
  * @max_hang_time:	Maximum time spent in hrtimer_interrupt
+ * @clock_base:		array of clock bases for this cpu
  */
 struct hrtimer_cpu_base {
 	raw_spinlock_t			lock;
-	struct hrtimer_clock_base	clock_base[HRTIMER_MAX_CLOCK_BASES];
+	unsigned long			active_bases;
 #ifdef CONFIG_HIGH_RES_TIMERS
 	ktime_t				expires_next;
 	int				hres_active;
@@ -186,6 +186,7 @@ struct hrtimer_cpu_base {
 	unsigned long			nr_hangs;
 	ktime_t				max_hang_time;
 #endif
+	struct hrtimer_clock_base	clock_base[HRTIMER_MAX_CLOCK_BASES];
 };
 
 static inline void hrtimer_set_expires(struct hrtimer *timer, ktime_t time)
@@ -256,8 +257,6 @@ static inline ktime_t hrtimer_expires_remaining(const struct hrtimer *timer)
 #ifdef CONFIG_HIGH_RES_TIMERS
 struct clock_event_device;
 
-extern void clock_was_set(void);
-extern void hres_timers_resume(void);
 extern void hrtimer_interrupt(struct clock_event_device *dev);
 
 /*
@@ -291,16 +290,8 @@ extern void hrtimer_peek_ahead_timers(void);
 # define MONOTONIC_RES_NSEC	LOW_RES_NSEC
 # define KTIME_MONOTONIC_RES	KTIME_LOW_RES
 
-/*
- * clock_was_set() is a NOP for non- high-resolution systems. The
- * time-sorted order guarantees that a timer does not expire early and
- * is expired in the next softirq when the clock was advanced.
- */
-static inline void clock_was_set(void) { }
 static inline void hrtimer_peek_ahead_timers(void) { }
 
-static inline void hres_timers_resume(void) { }
-
 /*
  * In non high resolution mode the time reference is taken from
  * the base softirq time variable.
@@ -316,10 +307,18 @@ static inline int hrtimer_is_hres_active(struct hrtimer *timer)
 }
 #endif
 
+extern void clock_was_set(void);
+#ifdef CONFIG_TIMERFD
+extern void timerfd_clock_was_set(void);
+#else
+static inline void timerfd_clock_was_set(void) { }
+#endif
+extern void hrtimers_resume(void);
+
 extern ktime_t ktime_get(void);
 extern ktime_t ktime_get_real(void);
 extern ktime_t ktime_get_boottime(void);
-
+extern ktime_t ktime_get_monotonic_offset(void);
 
 DECLARE_PER_CPU(struct tick_device, tick_cpu_device);
 

include/linux/thread_info.h

@@ -29,7 +29,7 @@ struct restart_block {
 		} futex;
 		/* For nanosleep */
 		struct {
-			clockid_t index;
+			clockid_t clockid;
 			struct timespec __user *rmtp;
 #ifdef CONFIG_COMPAT
 			struct compat_timespec __user *compat_rmtp;

include/linux/timerfd.h

@@ -19,13 +19,14 @@
  * shared O_* flags.
  */
 #define TFD_TIMER_ABSTIME (1 << 0)
+#define TFD_TIMER_CANCEL_ON_SET (1 << 1)
 #define TFD_CLOEXEC O_CLOEXEC
 #define TFD_NONBLOCK O_NONBLOCK
 
 #define TFD_SHARED_FCNTL_FLAGS (TFD_CLOEXEC | TFD_NONBLOCK)
 /* Flags for timerfd_create.  */
 #define TFD_CREATE_FLAGS TFD_SHARED_FCNTL_FLAGS
 /* Flags for timerfd_settime.  */
-#define TFD_SETTIME_FLAGS TFD_TIMER_ABSTIME
+#define TFD_SETTIME_FLAGS (TFD_TIMER_ABSTIME | TFD_TIMER_CANCEL_ON_SET)
 
 #endif /* _LINUX_TIMERFD_H */