[PATCH] time: Cleanup direct xtime usage in xen

[PATCH] time: Cleanup direct xtime usage in xen

[John Stultz]

Cleanup xen's direct use of internal timekeeping values.

Signed-off-by: John Stultz <johnstul@us.ibm.com>
---
 arch/x86/xen/time.c |    6 +++---
 1 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c
index 0d3f07c..6365df9 100644
--- a/arch/x86/xen/time.c
+++ b/arch/x86/xen/time.c
@@ -475,6 +475,7 @@ void xen_timer_resume(void)
 __init void xen_time_init(void)
 {
 	int cpu = smp_processor_id();
+	struct timespec tp;
 
 	clocksource_register(&xen_clocksource);
 
@@ -486,9 +487,8 @@ __init void xen_time_init(void)
 	}
 
 	/* Set initial system time with full resolution */
-	xen_read_wallclock(&xtime);
-	set_normalized_timespec(&wall_to_monotonic,
-				-xtime.tv_sec, -xtime.tv_nsec);
+	xen_read_wallclock(&tp);
+	do_settimeofday(&tp);
 
 	setup_force_cpu_cap(X86_FEATURE_TSC);
 
-- 
1.6.0.4

[PATCH] time: Cleanup warp_clock

[John Stultz]

Not critical, 2.6.35 material.

warp_clock() currently accesses timekeeping internal state directly, which
is unnecessary. Convert it to use the proper timekeeping interfaces.

Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: John Stultz <johnstul@us.ibm.com>
---
 kernel/time.c |   11 +++++------
 1 files changed, 5 insertions(+), 6 deletions(-)

diff --git a/kernel/time.c b/kernel/time.c
index 8047980..abe1c1b 100644
--- a/kernel/time.c
+++ b/kernel/time.c
@@ -133,12 +133,11 @@ SYSCALL_DEFINE2(gettimeofday, struct timeval __user *, tv,
  */
 static inline 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;
-	update_xtime_cache(0);
-	write_sequnlock_irq(&xtime_lock);
-	clock_was_set();
+	struct timespec delta, adjust;
+	delta.tv_sec = sys_tz.tz_minuteswest * 60;
+	delta.tv_nsec = 0;
+	adjust = timespec_add_safe(current_kernel_time(), delta);	
+	do_settimeofday(&adjust);
 }
 
 /*
-- 
1.6.0.4

[PATCH] time: Cleanup direct xtime usage in xen

[John Stultz]

Not critical, 2.6.35 material.

Cleanup xen's direct use of internal timekeeping values.

Cc: Jeremy Fitzhardinge <jeremy@xensource.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: John Stultz <johnstul@us.ibm.com>
---
 arch/x86/xen/time.c |    6 +++---
 1 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c
index 0d3f07c..6365df9 100644
--- a/arch/x86/xen/time.c
+++ b/arch/x86/xen/time.c
@@ -475,6 +475,7 @@ void xen_timer_resume(void)
 __init void xen_time_init(void)
 {
 	int cpu = smp_processor_id();
+	struct timespec tp;
 
 	clocksource_register(&xen_clocksource);
 
@@ -486,9 +487,8 @@ __init void xen_time_init(void)
 	}
 
 	/* Set initial system time with full resolution */
-	xen_read_wallclock(&xtime);
-	set_normalized_timespec(&wall_to_monotonic,
-				-xtime.tv_sec, -xtime.tv_nsec);
+	xen_read_wallclock(&tp);
+	do_settimeofday(&tp);
 
 	setup_force_cpu_cap(X86_FEATURE_TSC);
 
-- 
1.6.0.4

[PATCH] xtensa: Fix unnecessary setting of xtime

[John Stultz]

Not critical, 2.6.35 material.

xtensa supports read_persisitent_clock(), so there is no reason for
the arch specific code to be setting xtime. Setting xtime will be done
by the generic timekeeping code.

Please let me know if this is some sort of a workaround for an issue
instead of just an oversight.

Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Chris Zankel <chris@zankel.net>
Signed-off-by: John Stultz <johnstul@us.ibm.com>
---
 arch/xtensa/kernel/time.c |    5 -----
 1 files changed, 0 insertions(+), 5 deletions(-)

diff --git a/arch/xtensa/kernel/time.c b/arch/xtensa/kernel/time.c
index 19f7df3..19df764 100644
--- a/arch/xtensa/kernel/time.c
+++ b/arch/xtensa/kernel/time.c
@@ -60,11 +60,6 @@ static struct irqaction timer_irqaction = {
 
 void __init time_init(void)
 {
-	/* FIXME: xtime&wall_to_monotonic are set in timekeeping_init. */
-	read_persistent_clock(&xtime);
-	set_normalized_timespec(&wall_to_monotonic,
-		-xtime.tv_sec, -xtime.tv_nsec);
-
 #ifdef CONFIG_XTENSA_CALIBRATE_CCOUNT
 	printk("Calibrating CPU frequency ");
 	platform_calibrate_ccount();
-- 
1.6.0.4

Re: [PATCH] time: Cleanup direct xtime usage in xen

[Jeremy Fitzhardinge]

On 03/03/2010 07:11 PM, John Stultz wrote:
> Not critical, 2.6.35 material.
>
> Cleanup xen's direct use of internal timekeeping values.
>
> Cc: Jeremy Fitzhardinge<jeremy@xensource.com>
> Cc: Thomas Gleixner<tglx@linutronix.de>
> Cc: Andrew Morton<akpm@linux-foundation.org>
> Signed-off-by: John Stultz<johnstul@us.ibm.com>
>    
Acked-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>

> ---
>   arch/x86/xen/time.c |    6 +++---
>   1 files changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c
> index 0d3f07c..6365df9 100644
> --- a/arch/x86/xen/time.c
> +++ b/arch/x86/xen/time.c
> @@ -475,6 +475,7 @@ void xen_timer_resume(void)
>   __init void xen_time_init(void)
>   {
>   	int cpu = smp_processor_id();
> +	struct timespec tp;
>
>   	clocksource_register(&xen_clocksource);
>
> @@ -486,9 +487,8 @@ __init void xen_time_init(void)
>   	}
>
>   	/* Set initial system time with full resolution */
> -	xen_read_wallclock(&xtime);
> -	set_normalized_timespec(&wall_to_monotonic,
> -				-xtime.tv_sec, -xtime.tv_nsec);
> +	xen_read_wallclock(&tp);
> +	do_settimeofday(&tp);
>
>   	setup_force_cpu_cap(X86_FEATURE_TSC);
>
>