[PATCH 0/2] Convert message queue timeouts to hrtimer

[PATCH 0/2] Convert message queue timeouts to hrtimer

[Carsten Emde]

The message queue system calls that may timeout, mq_timedsend() and
mq_timedreceive(), still use the old jiffie-based timer interface.
Convert them to hrtimer.

The first patch provides the hrtimer infrastructure, the second patch
does the actual conversion to hrtimer.

	Carsten.

[PATCH 1/2] provide-schedule_hrtimeout-for-wallclock.patch

[Carsten Emde]

[-- Attachment #1: provide-schedule_hrtimeout-for-wallclock.patch --]
[-- Type: text/plain, Size: 4836 bytes --]

The current version of schedule_hrtimeout() always uses the
monotonic clock. Some system calls such as mq_timedsend()
and mq_timedreceive(), however, require the use of the wall
clock due to the definition of the system call.

This patch provides the infrastructure to use schedule_hrtimeout() 
with a wallclock timer.

CC: Andrew Morton <akpm@linux-foundation.org>
CC: Arjan van de Veen <arjan@infradead.org>
CC: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Pradyumna Sampath <pradysam@gmail.com>
Signed-off-by: Carsten Emde <C.Emde@osadl.org>

---
 include/linux/hrtimer.h |    2 +
 kernel/hrtimer.c        |   66 +++++++++++++++++++++++++++++-------------------
 2 files changed, 42 insertions(+), 26 deletions(-)

Index: head/include/linux/hrtimer.h
===================================================================
--- head.orig/include/linux/hrtimer.h
+++ head/include/linux/hrtimer.h
@@ -422,6 +422,8 @@ extern void hrtimer_init_sleeper(struct 
 
 extern int schedule_hrtimeout_range(ktime_t *expires, unsigned long delta,
 						const enum hrtimer_mode mode);
+extern int schedule_hrtimeout_range_clock(ktime_t *expires,
+		unsigned long delta, const enum hrtimer_mode mode, int clock);
 extern int schedule_hrtimeout(ktime_t *expires, const enum hrtimer_mode mode);
 
 /* Soft interrupt function to run the hrtimer queues: */
Index: head/kernel/hrtimer.c
===================================================================
--- head.orig/kernel/hrtimer.c
+++ head/kernel/hrtimer.c
@@ -1749,35 +1749,14 @@ void __init hrtimers_init(void)
 }
 
 /**
- * schedule_hrtimeout_range - sleep until timeout
+ * schedule_hrtimeout_range_clock - sleep until timeout
  * @expires:	timeout value (ktime_t)
  * @delta:	slack in expires timeout (ktime_t)
  * @mode:	timer mode, HRTIMER_MODE_ABS or HRTIMER_MODE_REL
- *
- * Make the current task sleep until the given expiry time has
- * elapsed. The routine will return immediately unless
- * the current task state has been set (see set_current_state()).
- *
- * The @delta argument gives the kernel the freedom to schedule the
- * actual wakeup to a time that is both power and performance friendly.
- * The kernel give the normal best effort behavior for "@expires+@delta",
- * but may decide to fire the timer earlier, but no earlier than @expires.
- *
- * You can set the task state as follows -
- *
- * %TASK_UNINTERRUPTIBLE - at least @timeout time is guaranteed to
- * pass before the routine returns.
- *
- * %TASK_INTERRUPTIBLE - the routine may return early if a signal is
- * delivered to the current task.
- *
- * The current task state is guaranteed to be TASK_RUNNING when this
- * routine returns.
- *
- * Returns 0 when the timer has expired otherwise -EINTR
+ * @clock:	timer clock, CLOCK_MONOTONIC or CLOCK_REALTIME
  */
-int __sched schedule_hrtimeout_range(ktime_t *expires, unsigned long delta,
-			       const enum hrtimer_mode mode)
+int __sched schedule_hrtimeout_range_clock(ktime_t *expires,
+	unsigned long delta, const enum hrtimer_mode mode, int clock)
 {
 	struct hrtimer_sleeper t;
 
@@ -1799,7 +1778,7 @@ int __sched schedule_hrtimeout_range(kti
 		return -EINTR;
 	}
 
-	hrtimer_init_on_stack(&t.timer, CLOCK_MONOTONIC, mode);
+	hrtimer_init_on_stack(&t.timer, clock, mode);
 	hrtimer_set_expires_range_ns(&t.timer, *expires, delta);
 
 	hrtimer_init_sleeper(&t, current);
@@ -1818,6 +1797,41 @@ int __sched schedule_hrtimeout_range(kti
 
 	return !t.task ? 0 : -EINTR;
 }
+
+/**
+ * schedule_hrtimeout_range - sleep until timeout
+ * @expires:	timeout value (ktime_t)
+ * @delta:	slack in expires timeout (ktime_t)
+ * @mode:	timer mode, HRTIMER_MODE_ABS or HRTIMER_MODE_REL
+ *
+ * Make the current task sleep until the given expiry time has
+ * elapsed. The routine will return immediately unless
+ * the current task state has been set (see set_current_state()).
+ *
+ * The @delta argument gives the kernel the freedom to schedule the
+ * actual wakeup to a time that is both power and performance friendly.
+ * The kernel give the normal best effort behavior for "@expires+@delta",
+ * but may decide to fire the timer earlier, but no earlier than @expires.
+ *
+ * You can set the task state as follows -
+ *
+ * %TASK_UNINTERRUPTIBLE - at least @timeout time is guaranteed to
+ * pass before the routine returns.
+ *
+ * %TASK_INTERRUPTIBLE - the routine may return early if a signal is
+ * delivered to the current task.
+ *
+ * The current task state is guaranteed to be TASK_RUNNING when this
+ * routine returns.
+ *
+ * Returns 0 when the timer has expired otherwise -EINTR
+ */
+int __sched schedule_hrtimeout_range(ktime_t *expires, unsigned long delta,
+       const enum hrtimer_mode mode)
+{
+	return schedule_hrtimeout_range_clock(expires, delta, mode,
+	    CLOCK_MONOTONIC);
+}
 EXPORT_SYMBOL_GPL(schedule_hrtimeout_range);
 
 /**

[PATCH 2/2] let-message-queue-timeout-use-hrtimer.patch

[Carsten Emde]

[-- Attachment #1: let-message-queue-timeout-use-hrtimer.patch --]
[-- Type: text/plain, Size: 4760 bytes --]

The message queue functions mq_timedsend() and mq_timedreceive()
have not yet been converted to use the hrtimer interface.

This patch replaces the call to schedule_timeout() by a call to
schedule_hrtimeout() and transforms the expiration time from
timespec to ktime as required.

CC: Andrew Morton <akpm@linux-foundation.org>
CC: Arjan van de Veen <arjan@infradead.org>
CC: Thomas Gleixner <tglx@linutronix.de>
Tested-by: Pradyumna Sampath <pradysam@gmail.com>
Signed-off-by: Carsten Emde <C.Emde@osadl.org>

---
 ipc/mqueue.c |   74 +++++++++++++++++++----------------------------------------
 1 file changed, 25 insertions(+), 49 deletions(-)

Index: head/ipc/mqueue.c
===================================================================
--- head.orig/ipc/mqueue.c
+++ head/ipc/mqueue.c
@@ -428,7 +428,7 @@ static void wq_add(struct mqueue_inode_i
  * sr: SEND or RECV
  */
 static int wq_sleep(struct mqueue_inode_info *info, int sr,
-			long timeout, struct ext_wait_queue *ewp)
+		    ktime_t *timeout, struct ext_wait_queue *ewp)
 {
 	int retval;
 	signed long time;
@@ -439,7 +439,8 @@ static int wq_sleep(struct mqueue_inode_
 		set_current_state(TASK_INTERRUPTIBLE);
 
 		spin_unlock(&info->lock);
-		time = schedule_timeout(timeout);
+		time = schedule_hrtimeout_range_clock(timeout,
+		    HRTIMER_MODE_ABS, 0, CLOCK_REALTIME);
 
 		while (ewp->state == STATE_PENDING)
 			cpu_relax();
@@ -551,31 +552,16 @@ static void __do_notify(struct mqueue_in
 	wake_up(&info->wait_q);
 }
 
-static long prepare_timeout(struct timespec *p)
+static int prepare_timeout(const struct timespec __user *u_abs_timeout,
+			   ktime_t *expires, struct timespec *ts)
 {
-	struct timespec nowts;
-	long timeout;
-
-	if (p) {
-		if (unlikely(p->tv_nsec < 0 || p->tv_sec < 0
-			|| p->tv_nsec >= NSEC_PER_SEC))
-			return -EINVAL;
-		nowts = CURRENT_TIME;
-		/* first subtract as jiffies can't be too big */
-		p->tv_sec -= nowts.tv_sec;
-		if (p->tv_nsec < nowts.tv_nsec) {
-			p->tv_nsec += NSEC_PER_SEC;
-			p->tv_sec--;
-		}
-		p->tv_nsec -= nowts.tv_nsec;
-		if (p->tv_sec < 0)
-			return 0;
-
-		timeout = timespec_to_jiffies(p) + 1;
-	} else
-		return MAX_SCHEDULE_TIMEOUT;
+	if (copy_from_user(ts, u_abs_timeout, sizeof(struct timespec)))
+		return -EFAULT;
+	if (!timespec_valid(ts))
+		return -EINVAL;
 
-	return timeout;
+	*expires = timespec_to_ktime(*ts);
+	return 0;
 }
 
 static void remove_notification(struct mqueue_inode_info *info)
@@ -861,22 +847,21 @@ SYSCALL_DEFINE5(mq_timedsend, mqd_t, mqd
 	struct ext_wait_queue *receiver;
 	struct msg_msg *msg_ptr;
 	struct mqueue_inode_info *info;
-	struct timespec ts, *p = NULL;
-	long timeout;
+ 	ktime_t expires, *timeout = NULL;
+	struct timespec ts;
 	int ret;
 
 	if (u_abs_timeout) {
-		if (copy_from_user(&ts, u_abs_timeout, 
-					sizeof(struct timespec)))
-			return -EFAULT;
-		p = &ts;
+		int res = prepare_timeout(u_abs_timeout, &expires, &ts);
+		if (res)
+			return res;
+		timeout = &expires;
 	}
 
 	if (unlikely(msg_prio >= (unsigned long) MQ_PRIO_MAX))
 		return -EINVAL;
 
-	audit_mq_sendrecv(mqdes, msg_len, msg_prio, p);
-	timeout = prepare_timeout(p);
+	audit_mq_sendrecv(mqdes, msg_len, msg_prio, timeout ? &ts : NULL);
 
 	filp = fget(mqdes);
 	if (unlikely(!filp)) {
@@ -918,9 +903,6 @@ SYSCALL_DEFINE5(mq_timedsend, mqd_t, mqd
 		if (filp->f_flags & O_NONBLOCK) {
 			spin_unlock(&info->lock);
 			ret = -EAGAIN;
-		} else if (unlikely(timeout < 0)) {
-			spin_unlock(&info->lock);
-			ret = timeout;
 		} else {
 			wait.task = current;
 			wait.msg = (void *) msg_ptr;
@@ -953,24 +935,23 @@ SYSCALL_DEFINE5(mq_timedreceive, mqd_t, 
 		size_t, msg_len, unsigned int __user *, u_msg_prio,
 		const struct timespec __user *, u_abs_timeout)
 {
-	long timeout;
 	ssize_t ret;
 	struct msg_msg *msg_ptr;
 	struct file *filp;
 	struct inode *inode;
 	struct mqueue_inode_info *info;
 	struct ext_wait_queue wait;
-	struct timespec ts, *p = NULL;
+   	ktime_t expires, *timeout = NULL;
+	struct timespec ts;
 
 	if (u_abs_timeout) {
-		if (copy_from_user(&ts, u_abs_timeout, 
-					sizeof(struct timespec)))
-			return -EFAULT;
-		p = &ts;
+		int res = prepare_timeout(u_abs_timeout, &expires, &ts);
+		if (res)
+			return res;
+		timeout = &expires;
 	}
 
-	audit_mq_sendrecv(mqdes, msg_len, 0, p);
-	timeout = prepare_timeout(p);
+	audit_mq_sendrecv(mqdes, msg_len, 0, timeout ? &ts : NULL);
 
 	filp = fget(mqdes);
 	if (unlikely(!filp)) {
@@ -1002,11 +983,6 @@ SYSCALL_DEFINE5(mq_timedreceive, mqd_t, 
 		if (filp->f_flags & O_NONBLOCK) {
 			spin_unlock(&info->lock);
 			ret = -EAGAIN;
-			msg_ptr = NULL;
-		} else if (unlikely(timeout < 0)) {
-			spin_unlock(&info->lock);
-			ret = timeout;
-			msg_ptr = NULL;
 		} else {
 			wait.task = current;
 			wait.state = STATE_NONE;