[PATCH] Some clean up of the time code.

[PATCH] Some clean up of the time code.

[george anzinger]

[-- Attachment #1: Type: text/plain, Size: 1132 bytes --]

This patch does the following:

Pushs down the change from timeval to timespec in the settime routines.

Fixes two places where time was set without updating the monotonic 
clock offset.  (Changes sys_stime() to call do_settimeofday() and 
changes clock_warp to do the update directly.)  These were bugs!

Changes the uptime code to use the posix_clock_monotonic notion of 
uptime instead of the jiffies.  This time will track NTP changes and 
so should be better than your standard wristwatch (if your using ntp).

Changes posix_clock_monotonic to start at 0 on boot (was set to start 
at initial jiffies).

Adds a test in do_settimeofday() to error out (EINVAL) attempts to use 
unnormalized times.  This is passed back up to both settimeofday and 
posix_setclock().

Warning: Requires changes in .../arch/???/kernel/time.c to change 
do_settimeofday() to return an error if time is not normalized and to 
use a timespec instead of timeval for its input.
-- 
George Anzinger   george@mvista.com
High-res-timers:  http://sourceforge.net/projects/high-res-timers/
Preemption patch: http://www.kernel.org/pub/linux/kernel/people/rml

[-- Attachment #2: timecleanup-2.5.69.1.0.patch --]
[-- Type: text/plain, Size: 8706 bytes --]

diff -urP -I '\$Id:.*Exp \$' -X /usr/src/patch.exclude linux-2.5.69-kb/arch/i386/kernel/time.c linux/arch/i386/kernel/time.c
--- linux-2.5.69-kb/arch/i386/kernel/time.c	2003-05-05 15:33:31.000000000 -0700
+++ linux/arch/i386/kernel/time.c	2003-05-23 14:30:05.000000000 -0700
@@ -114,8 +114,11 @@
 	tv->tv_usec = usec;
 }
 
-void do_settimeofday(struct timeval *tv)
+int do_settimeofday(struct timespec *tv)
 {
+	if ((unsigned long)tv->tv_nsec > NSEC_PER_SEC)
+		return -EINVAL;
+
 	write_seqlock_irq(&xtime_lock);
 	/*
 	 * This is revolting. We need to set "xtime" correctly. However, the
@@ -123,17 +126,16 @@
 	 * wall time.  Discover what correction gettimeofday() would have
 	 * made, and then undo it!
 	 */
-	tv->tv_usec -= timer->get_offset();
-	tv->tv_usec -= (jiffies - wall_jiffies) * (USEC_PER_SEC / HZ);
+	tv->tv_nsec -= timer->get_offset() * NSEC_PER_USEC;
+	tv->tv_nsec -= (jiffies - wall_jiffies) * TICK_NSEC(TICK_USEC);
 
-	while (tv->tv_usec < 0) {
-		tv->tv_usec += USEC_PER_SEC;
+	while (tv->tv_nsec < 0) {
+		tv->tv_nsec += NSEC_PER_SEC;
 		tv->tv_sec--;
 	}
-	tv->tv_usec *= NSEC_PER_USEC;
 
 	wall_to_monotonic.tv_sec += xtime.tv_sec - tv->tv_sec;
-	wall_to_monotonic.tv_nsec += xtime.tv_nsec - tv->tv_usec;
+	wall_to_monotonic.tv_nsec += xtime.tv_nsec - tv->tv_nsec;
 
 	if (wall_to_monotonic.tv_nsec > NSEC_PER_SEC) {
 		wall_to_monotonic.tv_nsec -= NSEC_PER_SEC;
@@ -145,13 +147,14 @@
 	}
 
 	xtime.tv_sec = tv->tv_sec;
-	xtime.tv_nsec = tv->tv_usec;
+	xtime.tv_nsec = tv->tv_nsec;
 	time_adjust = 0;		/* stop active adjtime() */
 	time_status |= STA_UNSYNC;
 	time_maxerror = NTP_PHASE_LIMIT;
 	time_esterror = NTP_PHASE_LIMIT;
 	write_sequnlock_irq(&xtime_lock);
 	clock_was_set();
+	return 0;
 }
 
 static int set_rtc_mmss(unsigned long nowtime)
@@ -299,9 +302,9 @@
 {
 	
 	xtime.tv_sec = get_cmos_time();
-	wall_to_monotonic.tv_sec = -xtime.tv_sec + INITIAL_JIFFIES / HZ;
+	wall_to_monotonic.tv_sec = -xtime.tv_sec;
 	xtime.tv_nsec = (INITIAL_JIFFIES % HZ) * (NSEC_PER_SEC / HZ);
-	wall_to_monotonic.tv_nsec = 0;
+	wall_to_monotonic.tv_nsec = -xtime.tv_nsec;
 
 
 	timer = select_timer();
diff -urP -I '\$Id:.*Exp \$' -X /usr/src/patch.exclude linux-2.5.69-kb/fs/proc/proc_misc.c linux/fs/proc/proc_misc.c
--- linux-2.5.69-kb/fs/proc/proc_misc.c	2003-05-05 15:34:02.000000000 -0700
+++ linux/fs/proc/proc_misc.c	2003-05-23 14:52:47.000000000 -0700
@@ -137,36 +137,19 @@
 static int uptime_read_proc(char *page, char **start, off_t off,
 				 int count, int *eof, void *data)
 {
-	u64 uptime;
-	unsigned long uptime_remainder;
+	struct timespec uptime;
+	struct timespec idle;
 	int len;
+	u64 idle_jiffies = init_task.utime + init_task.stime;
 
-	uptime = get_jiffies_64() - INITIAL_JIFFIES;
-	uptime_remainder = (unsigned long) do_div(uptime, HZ);
+	do_posix_clock_monotonic_gettime(&uptime);
+	jiffies_to_timespec(idle_jiffies, &idle);
+	len = sprintf(page,"%lu.%02lu %lu.%02lu\n",
+			(unsigned long) uptime.tv_sec,
+			(uptime.tv_nsec / (NSEC_PER_SEC / 100)),
+			(unsigned long) idle.tv_sec,
+			(idle.tv_nsec / (NSEC_PER_SEC / 100)));
 
-#if HZ!=100
-	{
-		u64 idle = init_task.utime + init_task.stime;
-		unsigned long idle_remainder;
-
-		idle_remainder = (unsigned long) do_div(idle, HZ);
-		len = sprintf(page,"%lu.%02lu %lu.%02lu\n",
-			(unsigned long) uptime,
-			(uptime_remainder * 100) / HZ,
-			(unsigned long) idle,
-			(idle_remainder * 100) / HZ);
-	}
-#else
-	{
-		unsigned long idle = init_task.utime + init_task.stime;
-
-		len = sprintf(page,"%lu.%02lu %lu.%02lu\n",
-			(unsigned long) uptime,
-			uptime_remainder,
-			idle / HZ,
-			idle % HZ);
-	}
-#endif
 	return proc_calc_metrics(page, start, off, count, eof, len);
 }
 
diff -urP -I '\$Id:.*Exp \$' -X /usr/src/patch.exclude linux-2.5.69-kb/include/linux/time.h linux/include/linux/time.h
--- linux-2.5.69-kb/include/linux/time.h	2003-05-05 15:34:09.000000000 -0700
+++ linux/include/linux/time.h	2003-05-23 14:51:36.000000000 -0700
@@ -200,9 +200,10 @@
 
 #ifdef __KERNEL__
 extern void do_gettimeofday(struct timeval *tv);
-extern void do_settimeofday(struct timeval *tv);
-extern int do_sys_settimeofday(struct timeval *tv, struct timezone *tz);
+extern int do_settimeofday(struct timespec *tv);
+extern int do_sys_settimeofday(struct timespec *tv, struct timezone *tz);
 extern void clock_was_set(void); // call when ever the clock is set
+extern int do_posix_clock_monotonic_gettime(struct timespec *tp);
 extern long do_nanosleep(struct timespec *t);
 extern long do_utimes(char * filename, struct timeval * times);
 #endif
diff -urP -I '\$Id:.*Exp \$' -X /usr/src/patch.exclude linux-2.5.69-kb/kernel/posix-timers.c linux/kernel/posix-timers.c
--- linux-2.5.69-kb/kernel/posix-timers.c	2003-05-23 15:39:50.000000000 -0700
+++ linux/kernel/posix-timers.c	2003-05-23 13:31:37.000000000 -0700
@@ -1038,8 +1038,7 @@
 	if (posix_clocks[which_clock].clock_set)
 		return posix_clocks[which_clock].clock_set(&new_tp);
 
-	new_tp.tv_nsec /= NSEC_PER_USEC;
-	return do_sys_settimeofday((struct timeval *) &new_tp, NULL);
+	return do_sys_settimeofday(&new_tp, NULL);
 }
 
 asmlinkage long
diff -urP -I '\$Id:.*Exp \$' -X /usr/src/patch.exclude linux-2.5.69-kb/kernel/time.c linux/kernel/time.c
--- linux-2.5.69-kb/kernel/time.c	2003-05-05 15:32:33.000000000 -0700
+++ linux/kernel/time.c	2003-05-23 14:31:34.000000000 -0700
@@ -63,28 +63,22 @@
 
 /*
  * sys_stime() can be implemented in user-level using
- * sys_settimeofday().  Is this for backwards compatibility?  If so,
+ * sys_settimeofday().	Is this for backwards compatibility?  If so,
  * why not move it into the appropriate arch directory (for those
  * architectures that need it).
  */
  
 asmlinkage long sys_stime(int * tptr)
 {
-	int value;
+	struct timespec tv;
 
 	if (!capable(CAP_SYS_TIME))
 		return -EPERM;
-	if (get_user(value, tptr))
+	if (get_user(tv.tv_sec, tptr))
 		return -EFAULT;
-	write_seqlock_irq(&xtime_lock);
-	xtime.tv_sec = value;
-	xtime.tv_nsec = 0;
-	last_time_offset = 0;
-	time_adjust = 0;	/* stop active adjtime() */
-	time_status |= STA_UNSYNC;
-	time_maxerror = NTP_PHASE_LIMIT;
-	time_esterror = NTP_PHASE_LIMIT;
-	write_sequnlock_irq(&xtime_lock);
+
+	tv.tv_nsec = 0;
+	do_settimeofday(&tv);
 	return 0;
 }
 
@@ -124,9 +118,11 @@
 inline static void warp_clock(void)
 {
 	write_seqlock_irq(&xtime_lock);
+	wall_to_monotonic.tv_sec -= sys_tz.tz_minuteswest * 60;
 	xtime.tv_sec += sys_tz.tz_minuteswest * 60;
 	last_time_offset = 0;
 	write_sequnlock_irq(&xtime_lock);
+	clock_was_set();
 }
 
 /*
@@ -140,7 +136,7 @@
  * various programs will get confused when the clock gets warped.
  */
 
-int do_sys_settimeofday(struct timeval *tv, struct timezone *tz)
+int do_sys_settimeofday(struct timespec *tv, struct timezone *tz)
 {
 	static int firsttime = 1;
 
@@ -161,14 +157,14 @@
 		/* SMP safe, again the code in arch/foo/time.c should
 		 * globally block out interrupts when it runs.
 		 */
-		do_settimeofday(tv);
+		return do_settimeofday(tv);
 	}
 	return 0;
 }
 
 asmlinkage long sys_settimeofday(struct timeval __user *tv, struct timezone __user *tz)
 {
-	struct timeval	new_tv;
+	struct timespec	new_tv;
 	struct timezone new_tz;
 
 	if (tv) {
@@ -178,6 +174,7 @@
 	if (tz) {
 		if (copy_from_user(&new_tz, tz, sizeof(*tz)))
 			return -EFAULT;
+		new_tv.tv_nsec *= NSEC_PER_USEC;
 	}
 
 	return do_sys_settimeofday(tv ? &new_tv : NULL, tz ? &new_tz : NULL);
diff -urP -I '\$Id:.*Exp \$' -X /usr/src/patch.exclude linux-2.5.69-kb/kernel/timer.c linux/kernel/timer.c
--- linux-2.5.69-kb/kernel/timer.c	2003-05-05 15:34:10.000000000 -0700
+++ linux/kernel/timer.c	2003-05-23 14:46:47.000000000 -0700
@@ -1105,7 +1105,6 @@
 asmlinkage long sys_sysinfo(struct sysinfo __user *info)
 {
 	struct sysinfo val;
-	u64 uptime;
 	unsigned long mem_total, sav_total;
 	unsigned int mem_unit, bitcount;
 	unsigned long seq;
@@ -1113,11 +1112,25 @@
 	memset((char *)&val, 0, sizeof(struct sysinfo));
 
 	do {
+		struct timespec tp;
 		seq = read_seqbegin(&xtime_lock);
 
-		uptime = jiffies_64 - INITIAL_JIFFIES;
-		do_div(uptime, HZ);
-		val.uptime = (unsigned long) uptime;
+		/*
+		 * This is annoying.  The below is the same thing
+		 * posix_get_clock_monotonic() does, but it wants to
+		 * take the lock which we want to cover the loads stuff
+		 * too.
+		 */
+
+		do_gettimeofday((struct timeval *)&tp);
+		tp.tv_nsec *= NSEC_PER_USEC;
+		tp.tv_sec += wall_to_monotonic.tv_sec;
+		tp.tv_nsec += wall_to_monotonic.tv_nsec;
+		if (tp.tv_nsec - NSEC_PER_SEC >= 0) {
+			tp.tv_nsec = tp.tv_nsec - NSEC_PER_SEC;
+			tp.tv_sec++;
+		}
+		val.uptime = tp.tv_sec + (tp.tv_nsec ? 1 : 0);
 
 		val.loads[0] = avenrun[0] << (SI_LOAD_SHIFT - FSHIFT);
 		val.loads[1] = avenrun[1] << (SI_LOAD_SHIFT - FSHIFT);

Re: [PATCH] Some clean up of the time code.

[Andrew Morton]

george anzinger <george@mvista.com> wrote:
>
> Andrew Morton wrote:
> > george anzinger <george@mvista.com> wrote:
> > 
> >>-void do_settimeofday(struct timeval *tv)
> >> +int do_settimeofday(struct timespec *tv)
> >>  {
> >> +	if ((unsigned long)tv->tv_nsec > NSEC_PER_SEC)
> >> +		return -EINVAL;
> >> +
> > 
> > 
> > Should that be ">="?
> > 
> > Is there any reasonable way to avoid breaking existing
> > do_settimeofday() implementations? That's just more grief all round.
> 
> Hm. Giving this more thought, the main reason for the change was to 
> move to the timespec from the timeval, i.e. nanoseconds instead of 
> microseconds.  The error check was put in because the function was 
> already being changed.  The reason to move to the timespec is to 
> complete the change made to xtime and to more correctly align with the 
> POSIX clock_settime, both of which use timespec.
> 

Well if it's really the Right Thing To Do then OK.  Was just checking.

About 30 do_settimeofday() implementations need to be repaired.

Re: [PATCH] Some clean up of the time code.

[george anzinger]

Andrew Morton wrote:
> george anzinger <george@mvista.com> wrote:
> 
>>-void do_settimeofday(struct timeval *tv)
>> +int do_settimeofday(struct timespec *tv)
>>  {
>> +	if ((unsigned long)tv->tv_nsec > NSEC_PER_SEC)
>> +		return -EINVAL;
>> +
> 
> 
> Should that be ">="?
> 
> Is there any reasonable way to avoid breaking existing
> do_settimeofday() implementations? That's just more grief all round.

Hm. Giving this more thought, the main reason for the change was to 
move to the timespec from the timeval, i.e. nanoseconds instead of 
microseconds.  The error check was put in because the function was 
already being changed.  The reason to move to the timespec is to 
complete the change made to xtime and to more correctly align with the 
POSIX clock_settime, both of which use timespec.

I suspect Linus would prefer this direction...

Comments?
> 
> 

-- 
George Anzinger   george@mvista.com
High-res-timers:  http://sourceforge.net/projects/high-res-timers/
Preemption patch: http://www.kernel.org/pub/linux/kernel/people/rml

Re: [PATCH] Some clean up of the time code.

[george anzinger]

Andrew Morton wrote:
> george anzinger <george@mvista.com> wrote:
> 
>>-void do_settimeofday(struct timeval *tv)
>> +int do_settimeofday(struct timespec *tv)
>>  {
>> +	if ((unsigned long)tv->tv_nsec > NSEC_PER_SEC)
>> +		return -EINVAL;
>> +
> 
> 
> Should that be ">="?

Yep, thanks for catching that.
> 
> Is there any reasonable way to avoid breaking existing
> do_settimeofday() implementations? That's just more grief all round.
> 
> 
Of course there is a way.  The question is which way leads to the most 
grief :).  The test could be made in the calling routines, but then it 
would need to be made in both posix-timer.c and time.c.  I suppose it 
would be better to do it that way as both are in common code and the 
"arch" warnning would go away.   Tomorrow...

-- 
George Anzinger   george@mvista.com
High-res-timers:  http://sourceforge.net/projects/high-res-timers/
Preemption patch: http://www.kernel.org/pub/linux/kernel/people/rml

Re: [PATCH] Some clean up of the time code.

[Andrew Morton]

george anzinger <george@mvista.com> wrote:
>
> -void do_settimeofday(struct timeval *tv)
>  +int do_settimeofday(struct timespec *tv)
>   {
>  +	if ((unsigned long)tv->tv_nsec > NSEC_PER_SEC)
>  +		return -EINVAL;
>  +

Should that be ">="?

Is there any reasonable way to avoid breaking existing
do_settimeofday() implementations? That's just more grief all round.

[PATCH] Some clean up of the time code.

[george anzinger]

[-- Attachment #1: Type: text/plain, Size: 1497 bytes --]

NOW, on top of 2.5.70-bk9 :)

This is an update of a patch I sent some time ago.  You could not
apply it then because of some other stuff...  Let me know if I need to
use a different base system...

I will send Eric's patch soon...

This patch does the following:

Pushs down the change from timeval to timespec in the settime routines.

Fixes two places where time was set without updating the monotonic
clock offset.  (Changes sys_stime() to call do_settimeofday() and
changes clock_warp to do the update directly.)  These were bugs!

Changes the uptime code to use the posix_clock_monotonic notion of
uptime instead of the jiffies.  This time will track NTP changes and
so should be better than your standard wristwatch (if your using ntp).

Changes posix_clock_monotonic to start at 0 on boot (was set to start
at initial jiffies).

Fixes a bug (never experienced) in timer_create() in posix-timers.c
where we "could" have released timer_id 0 if "id resources" were low.

Adds a test in do_settimeofday() to error out (EINVAL) attempts to use
unnormalized times.  This is passed back up to both settimeofday and
posix_setclock().

Warning: Requires changes in .../arch/???/kernel/time.c to change
do_settimeofday() to return an error if time is not normalized and to
use a timespec instead of timeval for its input.
-- 
George Anzinger   george@mvista.com
High-res-timers:  http://sourceforge.net/projects/high-res-timers/
Preemption patch: http://www.kernel.org/pub/linux/kernel/people/rml



[-- Attachment #2: timecleanup-2.5.70-bk9-1.0.patch --]
[-- Type: text/plain, Size: 8723 bytes --]

diff -urP -I '\$Id:.*Exp \$' -X /usr/src/patch.exclude linux-2.5.70-bk9-kb/arch/i386/kernel/time.c linux/arch/i386/kernel/time.c
--- linux-2.5.70-bk9-kb/arch/i386/kernel/time.c	2003-05-05 15:33:31.000000000 -0700
+++ linux/arch/i386/kernel/time.c	2003-06-09 17:16:13.000000000 -0700
@@ -114,8 +114,11 @@
 	tv->tv_usec = usec;
 }
 
-void do_settimeofday(struct timeval *tv)
+int do_settimeofday(struct timespec *tv)
 {
+	if ((unsigned long)tv->tv_nsec > NSEC_PER_SEC)
+		return -EINVAL;
+
 	write_seqlock_irq(&xtime_lock);
 	/*
 	 * This is revolting. We need to set "xtime" correctly. However, the
@@ -123,17 +126,16 @@
 	 * wall time.  Discover what correction gettimeofday() would have
 	 * made, and then undo it!
 	 */
-	tv->tv_usec -= timer->get_offset();
-	tv->tv_usec -= (jiffies - wall_jiffies) * (USEC_PER_SEC / HZ);
+	tv->tv_nsec -= timer->get_offset() * NSEC_PER_USEC;
+	tv->tv_nsec -= (jiffies - wall_jiffies) * TICK_NSEC(TICK_USEC);
 
-	while (tv->tv_usec < 0) {
-		tv->tv_usec += USEC_PER_SEC;
+	while (tv->tv_nsec < 0) {
+		tv->tv_nsec += NSEC_PER_SEC;
 		tv->tv_sec--;
 	}
-	tv->tv_usec *= NSEC_PER_USEC;
 
 	wall_to_monotonic.tv_sec += xtime.tv_sec - tv->tv_sec;
-	wall_to_monotonic.tv_nsec += xtime.tv_nsec - tv->tv_usec;
+	wall_to_monotonic.tv_nsec += xtime.tv_nsec - tv->tv_nsec;
 
 	if (wall_to_monotonic.tv_nsec > NSEC_PER_SEC) {
 		wall_to_monotonic.tv_nsec -= NSEC_PER_SEC;
@@ -145,13 +147,14 @@
 	}
 
 	xtime.tv_sec = tv->tv_sec;
-	xtime.tv_nsec = tv->tv_usec;
+	xtime.tv_nsec = tv->tv_nsec;
 	time_adjust = 0;		/* stop active adjtime() */
 	time_status |= STA_UNSYNC;
 	time_maxerror = NTP_PHASE_LIMIT;
 	time_esterror = NTP_PHASE_LIMIT;
 	write_sequnlock_irq(&xtime_lock);
 	clock_was_set();
+	return 0;
 }
 
 static int set_rtc_mmss(unsigned long nowtime)
@@ -299,9 +302,9 @@
 {
 	
 	xtime.tv_sec = get_cmos_time();
-	wall_to_monotonic.tv_sec = -xtime.tv_sec + INITIAL_JIFFIES / HZ;
+	wall_to_monotonic.tv_sec = -xtime.tv_sec;
 	xtime.tv_nsec = (INITIAL_JIFFIES % HZ) * (NSEC_PER_SEC / HZ);
-	wall_to_monotonic.tv_nsec = 0;
+	wall_to_monotonic.tv_nsec = -xtime.tv_nsec;
 
 
 	timer = select_timer();
diff -urP -I '\$Id:.*Exp \$' -X /usr/src/patch.exclude linux-2.5.70-bk9-kb/fs/proc/proc_misc.c linux/fs/proc/proc_misc.c
--- linux-2.5.70-bk9-kb/fs/proc/proc_misc.c	2003-05-30 01:18:10.000000000 -0700
+++ linux/fs/proc/proc_misc.c	2003-06-09 17:16:13.000000000 -0700
@@ -137,36 +137,19 @@
 static int uptime_read_proc(char *page, char **start, off_t off,
 				 int count, int *eof, void *data)
 {
-	u64 uptime;
-	unsigned long uptime_remainder;
+	struct timespec uptime;
+	struct timespec idle;
 	int len;
+	u64 idle_jiffies = init_task.utime + init_task.stime;
 
-	uptime = get_jiffies_64() - INITIAL_JIFFIES;
-	uptime_remainder = (unsigned long) do_div(uptime, HZ);
+	do_posix_clock_monotonic_gettime(&uptime);
+	jiffies_to_timespec(idle_jiffies, &idle);
+	len = sprintf(page,"%lu.%02lu %lu.%02lu\n",
+			(unsigned long) uptime.tv_sec,
+			(uptime.tv_nsec / (NSEC_PER_SEC / 100)),
+			(unsigned long) idle.tv_sec,
+			(idle.tv_nsec / (NSEC_PER_SEC / 100)));
 
-#if HZ!=100
-	{
-		u64 idle = init_task.utime + init_task.stime;
-		unsigned long idle_remainder;
-
-		idle_remainder = (unsigned long) do_div(idle, HZ);
-		len = sprintf(page,"%lu.%02lu %lu.%02lu\n",
-			(unsigned long) uptime,
-			(uptime_remainder * 100) / HZ,
-			(unsigned long) idle,
-			(idle_remainder * 100) / HZ);
-	}
-#else
-	{
-		unsigned long idle = init_task.utime + init_task.stime;
-
-		len = sprintf(page,"%lu.%02lu %lu.%02lu\n",
-			(unsigned long) uptime,
-			uptime_remainder,
-			idle / HZ,
-			idle % HZ);
-	}
-#endif
 	return proc_calc_metrics(page, start, off, count, eof, len);
 }
 
diff -urP -I '\$Id:.*Exp \$' -X /usr/src/patch.exclude linux-2.5.70-bk9-kb/include/linux/time.h linux/include/linux/time.h
--- linux-2.5.70-bk9-kb/include/linux/time.h	2003-05-30 01:18:18.000000000 -0700
+++ linux/include/linux/time.h	2003-06-09 17:16:13.000000000 -0700
@@ -200,9 +200,10 @@
 
 #ifdef __KERNEL__
 extern void do_gettimeofday(struct timeval *tv);
-extern void do_settimeofday(struct timeval *tv);
-extern int do_sys_settimeofday(struct timeval *tv, struct timezone *tz);
+extern int do_settimeofday(struct timespec *tv);
+extern int do_sys_settimeofday(struct timespec *tv, struct timezone *tz);
 extern void clock_was_set(void); // call when ever the clock is set
+extern int do_posix_clock_monotonic_gettime(struct timespec *tp);
 extern long do_nanosleep(struct timespec *t);
 extern long do_utimes(char __user * filename, struct timeval * times);
 #endif
diff -urP -I '\$Id:.*Exp \$' -X /usr/src/patch.exclude linux-2.5.70-bk9-kb/kernel/posix-timers.c linux/kernel/posix-timers.c
--- linux-2.5.70-bk9-kb/kernel/posix-timers.c	2003-06-09 17:13:23.000000000 -0700
+++ linux/kernel/posix-timers.c	2003-06-09 17:37:10.000000000 -0700
@@ -409,7 +409,7 @@
 	do {
 		if (unlikely(!idr_pre_get(&posix_timers_id))) {
 			error = -EAGAIN;
-			new_timer_id = (timer_t)-1;
+			new_timer->it_id = (timer_t)-1;
 			goto out;
 		}
 		spin_lock_irq(&idr_lock);
@@ -1026,8 +1026,7 @@
 	if (posix_clocks[which_clock].clock_set)
 		return posix_clocks[which_clock].clock_set(&new_tp);
 
-	new_tp.tv_nsec /= NSEC_PER_USEC;
-	return do_sys_settimeofday((struct timeval *) &new_tp, NULL);
+	return do_sys_settimeofday(&new_tp, NULL);
 }
 
 asmlinkage long
diff -urP -I '\$Id:.*Exp \$' -X /usr/src/patch.exclude linux-2.5.70-bk9-kb/kernel/time.c linux/kernel/time.c
--- linux-2.5.70-bk9-kb/kernel/time.c	2003-06-09 17:13:23.000000000 -0700
+++ linux/kernel/time.c	2003-06-09 17:28:12.000000000 -0700
@@ -68,22 +68,15 @@
  
 asmlinkage long sys_stime(int * tptr)
 {
-	int value;
+	struct timespec tv;
 
 	if (!capable(CAP_SYS_TIME))
 		return -EPERM;
-	if (get_user(value, tptr))
+	if (get_user(tv.tv_sec, tptr))
 		return -EFAULT;
-	write_seqlock_irq(&xtime_lock);
 
-	time_interpolator_reset();
-	xtime.tv_sec = value;
-	xtime.tv_nsec = 0;
-	time_adjust = 0;	/* stop active adjtime() */
-	time_status |= STA_UNSYNC;
-	time_maxerror = NTP_PHASE_LIMIT;
-	time_esterror = NTP_PHASE_LIMIT;
-	write_sequnlock_irq(&xtime_lock);
+	tv.tv_nsec = 0;
+	do_settimeofday(&tv);
 	return 0;
 }
 
@@ -123,9 +116,11 @@
 inline static void warp_clock(void)
 {
 	write_seqlock_irq(&xtime_lock);
+	wall_to_monotonic.tv_sec -= sys_tz.tz_minuteswest * 60;
 	xtime.tv_sec += sys_tz.tz_minuteswest * 60;
 	time_interpolator_update(sys_tz.tz_minuteswest * 60 * NSEC_PER_SEC);
 	write_sequnlock_irq(&xtime_lock);
+	clock_was_set();
 }
 
 /*
@@ -139,7 +134,7 @@
  * various programs will get confused when the clock gets warped.
  */
 
-int do_sys_settimeofday(struct timeval *tv, struct timezone *tz)
+int do_sys_settimeofday(struct timespec *tv, struct timezone *tz)
 {
 	static int firsttime = 1;
 
@@ -160,14 +155,14 @@
 		/* SMP safe, again the code in arch/foo/time.c should
 		 * globally block out interrupts when it runs.
 		 */
-		do_settimeofday(tv);
+		return do_settimeofday(tv);
 	}
 	return 0;
 }
 
 asmlinkage long sys_settimeofday(struct timeval __user *tv, struct timezone __user *tz)
 {
-	struct timeval	new_tv;
+	struct timespec	new_tv;
 	struct timezone new_tz;
 
 	if (tv) {
@@ -177,6 +172,7 @@
 	if (tz) {
 		if (copy_from_user(&new_tz, tz, sizeof(*tz)))
 			return -EFAULT;
+		new_tv.tv_nsec *= NSEC_PER_USEC;
 	}
 
 	return do_sys_settimeofday(tv ? &new_tv : NULL, tz ? &new_tz : NULL);
diff -urP -I '\$Id:.*Exp \$' -X /usr/src/patch.exclude linux-2.5.70-bk9-kb/kernel/timer.c linux/kernel/timer.c
--- linux-2.5.70-bk9-kb/kernel/timer.c	2003-06-09 17:13:23.000000000 -0700
+++ linux/kernel/timer.c	2003-06-09 17:16:13.000000000 -0700
@@ -1108,7 +1108,6 @@
 asmlinkage long sys_sysinfo(struct sysinfo __user *info)
 {
 	struct sysinfo val;
-	u64 uptime;
 	unsigned long mem_total, sav_total;
 	unsigned int mem_unit, bitcount;
 	unsigned long seq;
@@ -1116,11 +1115,25 @@
 	memset((char *)&val, 0, sizeof(struct sysinfo));
 
 	do {
+		struct timespec tp;
 		seq = read_seqbegin(&xtime_lock);
 
-		uptime = jiffies_64 - INITIAL_JIFFIES;
-		do_div(uptime, HZ);
-		val.uptime = (unsigned long) uptime;
+		/*
+		 * This is annoying.  The below is the same thing
+		 * posix_get_clock_monotonic() does, but it wants to
+		 * take the lock which we want to cover the loads stuff
+		 * too.
+		 */
+
+		do_gettimeofday((struct timeval *)&tp);
+		tp.tv_nsec *= NSEC_PER_USEC;
+		tp.tv_sec += wall_to_monotonic.tv_sec;
+		tp.tv_nsec += wall_to_monotonic.tv_nsec;
+		if (tp.tv_nsec - NSEC_PER_SEC >= 0) {
+			tp.tv_nsec = tp.tv_nsec - NSEC_PER_SEC;
+			tp.tv_sec++;
+		}
+		val.uptime = tp.tv_sec + (tp.tv_nsec ? 1 : 0);
 
 		val.loads[0] = avenrun[0] << (SI_LOAD_SHIFT - FSHIFT);
 		val.loads[1] = avenrun[1] << (SI_LOAD_SHIFT - FSHIFT);

Re: [PATCH] Some clean up of the time code.

[john stultz]

On Mon, 2003-06-09 at 15:57, george anzinger wrote:
> Pushs down the change from timeval to timespec in the settime routines.

No complaints here. 

> Fixes two places where time was set without updating the monotonic
> clock offset.  (Changes sys_stime() to call do_settimeofday() and
> changes clock_warp to do the update directly.)  These were bugs!

Yea, that sys_stime bug really needed fixing. Thanks for kicking this
out.

> Changes the uptime code to use the posix_clock_monotonic notion of
> uptime instead of the jiffies.  This time will track NTP changes and
> so should be better than your standard wristwatch (if your using ntp).

Hmmm. Might want to see if the btime changes from last week could
benefit from a similar change. 

Didn't test it personally, but looks good. 

thanks
-john

[PATCH] Some clean up of the time code.

[george anzinger]

[-- Attachment #1: Type: text/plain, Size: 1469 bytes --]

This is an update of a patch I sent some time ago.  You could not 
apply it then because of some other stuff...  Let me know if I need to 
use a different base system...

I will send Eric's patch soon...

This patch does the following:

Pushs down the change from timeval to timespec in the settime routines.

Fixes two places where time was set without updating the monotonic
clock offset.  (Changes sys_stime() to call do_settimeofday() and
changes clock_warp to do the update directly.)  These were bugs!

Changes the uptime code to use the posix_clock_monotonic notion of
uptime instead of the jiffies.  This time will track NTP changes and
so should be better than your standard wristwatch (if your using ntp).

Changes posix_clock_monotonic to start at 0 on boot (was set to start
at initial jiffies).

Fixes a bug (never experienced) in timer_create() in posix-timers.c 
where we "could" have released timer_id 0 if "id resources" were low.

Adds a test in do_settimeofday() to error out (EINVAL) attempts to use
unnormalized times.  This is passed back up to both settimeofday and
posix_setclock().

Warning: Requires changes in .../arch/???/kernel/time.c to change
do_settimeofday() to return an error if time is not normalized and to
use a timespec instead of timeval for its input.
-- 
George Anzinger   george@mvista.com
High-res-timers:  http://sourceforge.net/projects/high-res-timers/
Preemption patch: http://www.kernel.org/pub/linux/kernel/people/rml


[-- Attachment #2: timecleanup-2.5.70-1.0.patch --]
[-- Type: text/plain, Size: 9138 bytes --]

diff -urP -I '\$Id:.*Exp \$' -X /usr/src/patch.exclude linux-2.5.70-kb/arch/i386/kernel/time.c linux/arch/i386/kernel/time.c
--- linux-2.5.70-kb/arch/i386/kernel/time.c	2003-05-05 15:33:31.000000000 -0700
+++ linux/arch/i386/kernel/time.c	2003-05-30 01:29:48.000000000 -0700
@@ -114,8 +114,11 @@
 	tv->tv_usec = usec;
 }
 
-void do_settimeofday(struct timeval *tv)
+int do_settimeofday(struct timespec *tv)
 {
+	if ((unsigned long)tv->tv_nsec > NSEC_PER_SEC)
+		return -EINVAL;
+
 	write_seqlock_irq(&xtime_lock);
 	/*
 	 * This is revolting. We need to set "xtime" correctly. However, the
@@ -123,17 +126,16 @@
 	 * wall time.  Discover what correction gettimeofday() would have
 	 * made, and then undo it!
 	 */
-	tv->tv_usec -= timer->get_offset();
-	tv->tv_usec -= (jiffies - wall_jiffies) * (USEC_PER_SEC / HZ);
+	tv->tv_nsec -= timer->get_offset() * NSEC_PER_USEC;
+	tv->tv_nsec -= (jiffies - wall_jiffies) * TICK_NSEC(TICK_USEC);
 
-	while (tv->tv_usec < 0) {
-		tv->tv_usec += USEC_PER_SEC;
+	while (tv->tv_nsec < 0) {
+		tv->tv_nsec += NSEC_PER_SEC;
 		tv->tv_sec--;
 	}
-	tv->tv_usec *= NSEC_PER_USEC;
 
 	wall_to_monotonic.tv_sec += xtime.tv_sec - tv->tv_sec;
-	wall_to_monotonic.tv_nsec += xtime.tv_nsec - tv->tv_usec;
+	wall_to_monotonic.tv_nsec += xtime.tv_nsec - tv->tv_nsec;
 
 	if (wall_to_monotonic.tv_nsec > NSEC_PER_SEC) {
 		wall_to_monotonic.tv_nsec -= NSEC_PER_SEC;
@@ -145,13 +147,14 @@
 	}
 
 	xtime.tv_sec = tv->tv_sec;
-	xtime.tv_nsec = tv->tv_usec;
+	xtime.tv_nsec = tv->tv_nsec;
 	time_adjust = 0;		/* stop active adjtime() */
 	time_status |= STA_UNSYNC;
 	time_maxerror = NTP_PHASE_LIMIT;
 	time_esterror = NTP_PHASE_LIMIT;
 	write_sequnlock_irq(&xtime_lock);
 	clock_was_set();
+	return 0;
 }
 
 static int set_rtc_mmss(unsigned long nowtime)
@@ -299,9 +302,9 @@
 {
 	
 	xtime.tv_sec = get_cmos_time();
-	wall_to_monotonic.tv_sec = -xtime.tv_sec + INITIAL_JIFFIES / HZ;
+	wall_to_monotonic.tv_sec = -xtime.tv_sec;
 	xtime.tv_nsec = (INITIAL_JIFFIES % HZ) * (NSEC_PER_SEC / HZ);
-	wall_to_monotonic.tv_nsec = 0;
+	wall_to_monotonic.tv_nsec = -xtime.tv_nsec;
 
 
 	timer = select_timer();
diff -urP -I '\$Id:.*Exp \$' -X /usr/src/patch.exclude linux-2.5.70-kb/fs/proc/proc_misc.c linux/fs/proc/proc_misc.c
--- linux-2.5.70-kb/fs/proc/proc_misc.c	2003-05-30 01:18:10.000000000 -0700
+++ linux/fs/proc/proc_misc.c	2003-05-30 01:29:48.000000000 -0700
@@ -137,36 +137,19 @@
 static int uptime_read_proc(char *page, char **start, off_t off,
 				 int count, int *eof, void *data)
 {
-	u64 uptime;
-	unsigned long uptime_remainder;
+	struct timespec uptime;
+	struct timespec idle;
 	int len;
+	u64 idle_jiffies = init_task.utime + init_task.stime;
 
-	uptime = get_jiffies_64() - INITIAL_JIFFIES;
-	uptime_remainder = (unsigned long) do_div(uptime, HZ);
+	do_posix_clock_monotonic_gettime(&uptime);
+	jiffies_to_timespec(idle_jiffies, &idle);
+	len = sprintf(page,"%lu.%02lu %lu.%02lu\n",
+			(unsigned long) uptime.tv_sec,
+			(uptime.tv_nsec / (NSEC_PER_SEC / 100)),
+			(unsigned long) idle.tv_sec,
+			(idle.tv_nsec / (NSEC_PER_SEC / 100)));
 
-#if HZ!=100
-	{
-		u64 idle = init_task.utime + init_task.stime;
-		unsigned long idle_remainder;
-
-		idle_remainder = (unsigned long) do_div(idle, HZ);
-		len = sprintf(page,"%lu.%02lu %lu.%02lu\n",
-			(unsigned long) uptime,
-			(uptime_remainder * 100) / HZ,
-			(unsigned long) idle,
-			(idle_remainder * 100) / HZ);
-	}
-#else
-	{
-		unsigned long idle = init_task.utime + init_task.stime;
-
-		len = sprintf(page,"%lu.%02lu %lu.%02lu\n",
-			(unsigned long) uptime,
-			uptime_remainder,
-			idle / HZ,
-			idle % HZ);
-	}
-#endif
 	return proc_calc_metrics(page, start, off, count, eof, len);
 }
 
diff -urP -I '\$Id:.*Exp \$' -X /usr/src/patch.exclude linux-2.5.70-kb/include/linux/time.h linux/include/linux/time.h
--- linux-2.5.70-kb/include/linux/time.h	2003-05-30 01:18:18.000000000 -0700
+++ linux/include/linux/time.h	2003-05-30 01:29:48.000000000 -0700
@@ -200,9 +200,10 @@
 
 #ifdef __KERNEL__
 extern void do_gettimeofday(struct timeval *tv);
-extern void do_settimeofday(struct timeval *tv);
-extern int do_sys_settimeofday(struct timeval *tv, struct timezone *tz);
+extern int do_settimeofday(struct timespec *tv);
+extern int do_sys_settimeofday(struct timespec *tv, struct timezone *tz);
 extern void clock_was_set(void); // call when ever the clock is set
+extern int do_posix_clock_monotonic_gettime(struct timespec *tp);
 extern long do_nanosleep(struct timespec *t);
 extern long do_utimes(char __user * filename, struct timeval * times);
 #endif
diff -urP -I '\$Id:.*Exp \$' -X /usr/src/patch.exclude linux-2.5.70-kb/kernel/posix-timers.c linux/kernel/posix-timers.c
--- linux-2.5.70-kb/kernel/posix-timers.c	2003-05-30 01:18:19.000000000 -0700
+++ linux/kernel/posix-timers.c	2003-06-09 13:26:30.000000000 -0700
@@ -418,7 +418,7 @@
 	do {
 		if (unlikely(!idr_pre_get(&posix_timers_id))) {
 			error = -EAGAIN;
-			new_timer_id = (timer_t)-1;
+			new_timer->it_id = (timer_t)-1;
 			goto out;
 		}
 		spin_lock_irq(&idr_lock);
@@ -1035,8 +1035,7 @@
 	if (posix_clocks[which_clock].clock_set)
 		return posix_clocks[which_clock].clock_set(&new_tp);
 
-	new_tp.tv_nsec /= NSEC_PER_USEC;
-	return do_sys_settimeofday((struct timeval *) &new_tp, NULL);
+	return do_sys_settimeofday(&new_tp, NULL);
 }
 
 asmlinkage long
@@ -1202,6 +1201,7 @@
 		if (abs || !rq_time) {
 			adjust_abs_time(&posix_clocks[which_clock], &t, abs,
 					&rq_time);
+			//	rq_time += (t.tv_sec || t.tv_nsec);
 		}
 
 		left = rq_time - get_jiffies_64();
diff -urP -I '\$Id:.*Exp \$' -X /usr/src/patch.exclude linux-2.5.70-kb/kernel/time.c linux/kernel/time.c
--- linux-2.5.70-kb/kernel/time.c	2003-05-05 15:32:33.000000000 -0700
+++ linux/kernel/time.c	2003-05-30 01:29:48.000000000 -0700
@@ -63,28 +63,22 @@
 
 /*
  * sys_stime() can be implemented in user-level using
- * sys_settimeofday().  Is this for backwards compatibility?  If so,
+ * sys_settimeofday().	Is this for backwards compatibility?  If so,
  * why not move it into the appropriate arch directory (for those
  * architectures that need it).
  */
  
 asmlinkage long sys_stime(int * tptr)
 {
-	int value;
+	struct timespec tv;
 
 	if (!capable(CAP_SYS_TIME))
 		return -EPERM;
-	if (get_user(value, tptr))
+	if (get_user(tv.tv_sec, tptr))
 		return -EFAULT;
-	write_seqlock_irq(&xtime_lock);
-	xtime.tv_sec = value;
-	xtime.tv_nsec = 0;
-	last_time_offset = 0;
-	time_adjust = 0;	/* stop active adjtime() */
-	time_status |= STA_UNSYNC;
-	time_maxerror = NTP_PHASE_LIMIT;
-	time_esterror = NTP_PHASE_LIMIT;
-	write_sequnlock_irq(&xtime_lock);
+
+	tv.tv_nsec = 0;
+	do_settimeofday(&tv);
 	return 0;
 }
 
@@ -124,9 +118,11 @@
 inline static void warp_clock(void)
 {
 	write_seqlock_irq(&xtime_lock);
+	wall_to_monotonic.tv_sec -= sys_tz.tz_minuteswest * 60;
 	xtime.tv_sec += sys_tz.tz_minuteswest * 60;
 	last_time_offset = 0;
 	write_sequnlock_irq(&xtime_lock);
+	clock_was_set();
 }
 
 /*
@@ -140,7 +136,7 @@
  * various programs will get confused when the clock gets warped.
  */
 
-int do_sys_settimeofday(struct timeval *tv, struct timezone *tz)
+int do_sys_settimeofday(struct timespec *tv, struct timezone *tz)
 {
 	static int firsttime = 1;
 
@@ -161,14 +157,14 @@
 		/* SMP safe, again the code in arch/foo/time.c should
 		 * globally block out interrupts when it runs.
 		 */
-		do_settimeofday(tv);
+		return do_settimeofday(tv);
 	}
 	return 0;
 }
 
 asmlinkage long sys_settimeofday(struct timeval __user *tv, struct timezone __user *tz)
 {
-	struct timeval	new_tv;
+	struct timespec	new_tv;
 	struct timezone new_tz;
 
 	if (tv) {
@@ -178,6 +174,7 @@
 	if (tz) {
 		if (copy_from_user(&new_tz, tz, sizeof(*tz)))
 			return -EFAULT;
+		new_tv.tv_nsec *= NSEC_PER_USEC;
 	}
 
 	return do_sys_settimeofday(tv ? &new_tv : NULL, tz ? &new_tz : NULL);
diff -urP -I '\$Id:.*Exp \$' -X /usr/src/patch.exclude linux-2.5.70-kb/kernel/timer.c linux/kernel/timer.c
--- linux-2.5.70-kb/kernel/timer.c	2003-05-30 01:18:19.000000000 -0700
+++ linux/kernel/timer.c	2003-05-30 01:29:48.000000000 -0700
@@ -1105,7 +1105,6 @@
 asmlinkage long sys_sysinfo(struct sysinfo __user *info)
 {
 	struct sysinfo val;
-	u64 uptime;
 	unsigned long mem_total, sav_total;
 	unsigned int mem_unit, bitcount;
 	unsigned long seq;
@@ -1113,11 +1112,25 @@
 	memset((char *)&val, 0, sizeof(struct sysinfo));
 
 	do {
+		struct timespec tp;
 		seq = read_seqbegin(&xtime_lock);
 
-		uptime = jiffies_64 - INITIAL_JIFFIES;
-		do_div(uptime, HZ);
-		val.uptime = (unsigned long) uptime;
+		/*
+		 * This is annoying.  The below is the same thing
+		 * posix_get_clock_monotonic() does, but it wants to
+		 * take the lock which we want to cover the loads stuff
+		 * too.
+		 */
+
+		do_gettimeofday((struct timeval *)&tp);
+		tp.tv_nsec *= NSEC_PER_USEC;
+		tp.tv_sec += wall_to_monotonic.tv_sec;
+		tp.tv_nsec += wall_to_monotonic.tv_nsec;
+		if (tp.tv_nsec - NSEC_PER_SEC >= 0) {
+			tp.tv_nsec = tp.tv_nsec - NSEC_PER_SEC;
+			tp.tv_sec++;
+		}
+		val.uptime = tp.tv_sec + (tp.tv_nsec ? 1 : 0);
 
 		val.loads[0] = avenrun[0] << (SI_LOAD_SHIFT - FSHIFT);
 		val.loads[1] = avenrun[1] << (SI_LOAD_SHIFT - FSHIFT);