Mailing List Archive: [PATCH 3/3] xen/privcmd: add IOCTL_PRIVCMD_SYNC

[PATCH 3/3] xen/privcmd: add IOCTL_PRIVCMD_SYNC_WALLCLOCK to sync Xen's wallclock

Oct 12, 2012, 5:57 AM

Post #1 of 9 (565 views)

From: David Vrabel <david.vrabel@citrix.com>

Add a new ioctl to synchronize Xen's wallclock with the current system
time.

This may be used by the tools to ensure that newly created domains see
the correct wallclock time if NTP is not used in dom0 or if domains
are started before NTP has synchronized.

Signed-off-by: David Vrabel <david.vrabel@citrix.com>
---
arch/x86/xen/time.c | 25 ++++++++++++++++++-------
drivers/xen/privcmd.c | 12 ++++++++++++
include/xen/privcmd.h | 8 ++++++++
include/xen/xen-ops.h | 2 ++
4 files changed, 40 insertions(+), 7 deletions(-)

diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c
index 11770d0..d481ac9 100644
--- a/arch/x86/xen/time.c
+++ b/arch/x86/xen/time.c
@@ -8,6 +8,7 @@
* Jeremy Fitzhardinge <jeremy@xensource.com>, XenSource Inc, 2007
*/
#include <linux/kernel.h>
+#include <linux/export.h>
#include <linux/interrupt.h>
#include <linux/clocksource.h>
#include <linux/clockchips.h>
@@ -192,6 +193,19 @@ static void xen_read_wallclock(struct timespec *ts)
put_cpu_var(xen_vcpu);
}

+int xen_write_wallclock(const struct timespec *ts)
+{
+ struct xen_platform_op op;
+
+ op.cmd = XENPF_settime;
+ op.u.settime.secs = ts->tv_sec;
+ op.u.settime.nsecs = ts->tv_nsec;
+ op.u.settime.system_time = xen_clocksource_read();
+
+ return HYPERVISOR_dom0_op(&op);
+}
+EXPORT_SYMBOL_GPL(xen_write_wallclock);
+
static unsigned long xen_get_wallclock(void)
{
struct timespec ts;
@@ -202,7 +216,7 @@ static unsigned long xen_get_wallclock(void)

static int xen_set_wallclock(unsigned long now)
{
- struct xen_platform_op op;
+ struct timespec ts;
int rc;

/* do nothing for domU */
@@ -218,12 +232,9 @@ static int xen_set_wallclock(unsigned long now)
* update_persistent_wallclock() is called ~500 ms after 'now'
* so add an extra 500 ms.
*/
- op.cmd = XENPF_settime;
- op.u.settime.secs = now;
- op.u.settime.nsecs = NSEC_PER_SEC / 2;
- op.u.settime.system_time = xen_clocksource_read();
-
- rc = HYPERVISOR_dom0_op(&op);
+ ts.tv_sec = now;
+ ts.tv_nsec = NSEC_PER_SEC / 2;
+ rc = xen_write_wallclock(&ts);
WARN(rc != 0, "XENPF_settime failed: now=%ld\n", now);

return rc;
diff --git a/drivers/xen/privcmd.c b/drivers/xen/privcmd.c
index ccee0f1..ed2caf3 100644
--- a/drivers/xen/privcmd.c
+++ b/drivers/xen/privcmd.c
@@ -338,6 +338,14 @@ out:
return ret;
}

+static long privcmd_ioctl_sync_wallclock(void)
+{
+ struct timespec ts;
+
+ getnstimeofday(&ts);
+ return xen_write_wallclock(&ts);
+}
+
static long privcmd_ioctl(struct file *file,
unsigned int cmd, unsigned long data)
{
@@ -357,6 +365,10 @@ static long privcmd_ioctl(struct file *file,
ret = privcmd_ioctl_mmap_batch(udata);
break;

+ case IOCTL_PRIVCMD_SYNC_WALLCLOCK:
+ ret = privcmd_ioctl_sync_wallclock();
+ break;
+
default:
ret = -EINVAL;
break;
diff --git a/include/xen/privcmd.h b/include/xen/privcmd.h
index 17857fb..d17d087 100644
--- a/include/xen/privcmd.h
+++ b/include/xen/privcmd.h
@@ -66,6 +66,12 @@ struct privcmd_mmapbatch {
* @cmd: IOCTL_PRIVCMD_HYPERCALL
* @arg: &privcmd_hypercall_t
* Return: Value returned from execution of the specified hypercall.
+ *
+ * @cmd: IOCTL_PRIVCMD_SYNC_WALLCLOCK
+ * @arg: Unused.
+ * Synchronizes the Xen wallclock with the current system time.
+ * Return: 0 on success, or -1 on error with errno set to EPERM or
+ * EACCES.
*/
#define IOCTL_PRIVCMD_HYPERCALL \
_IOC(_IOC_NONE, 'P', 0, sizeof(struct privcmd_hypercall))
@@ -73,5 +79,7 @@ struct privcmd_mmapbatch {
_IOC(_IOC_NONE, 'P', 2, sizeof(struct privcmd_mmap))
#define IOCTL_PRIVCMD_MMAPBATCH \
_IOC(_IOC_NONE, 'P', 3, sizeof(struct privcmd_mmapbatch))
+#define IOCTL_PRIVCMD_SYNC_WALLCLOCK \
+ _IOC(_IOC_NONE, 'P', 5, 0)

#endif /* __LINUX_PUBLIC_PRIVCMD_H__ */
diff --git a/include/xen/xen-ops.h b/include/xen/xen-ops.h
index 6a198e4..eefed22 100644
--- a/include/xen/xen-ops.h
+++ b/include/xen/xen-ops.h
@@ -29,4 +29,6 @@ int xen_remap_domain_mfn_range(struct vm_area_struct *vma,
unsigned long mfn, int nr,
pgprot_t prot, unsigned domid);

+int xen_write_wallclock(const struct timespec *ts);
+
#endif /* INCLUDE_XEN_OPS_H */
--
1.7.2.5

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

Re: [PATCH 3/3] xen/privcmd: add IOCTL_PRIVCMD_SYNC_WALLCLOCK to sync Xen's wallclock [ In reply to ]

konrad at kernel

Oct 12, 2012, 6:41 AM

Post #2 of 9 (557 views)

Permalink

On Fri, Oct 12, 2012 at 01:57:14PM +0100, David Vrabel wrote:
> From: David Vrabel <david.vrabel@citrix.com>
>
> Add a new ioctl to synchronize Xen's wallclock with the current system
> time.
>
> This may be used by the tools to ensure that newly created domains see
> the correct wallclock time if NTP is not used in dom0 or if domains
> are started before NTP has synchronized.

So... how does this work with NTPD? As in does ntpd _not_ update the
hwclock enough?

>
> Signed-off-by: David Vrabel <david.vrabel@citrix.com>
> ---
> arch/x86/xen/time.c | 25 ++++++++++++++++++-------
> drivers/xen/privcmd.c | 12 ++++++++++++
> include/xen/privcmd.h | 8 ++++++++
> include/xen/xen-ops.h | 2 ++
> 4 files changed, 40 insertions(+), 7 deletions(-)
>
> diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c
> index 11770d0..d481ac9 100644
> --- a/arch/x86/xen/time.c
> +++ b/arch/x86/xen/time.c
> @@ -8,6 +8,7 @@
> * Jeremy Fitzhardinge <jeremy@xensource.com>, XenSource Inc, 2007
> */
> #include <linux/kernel.h>
> +#include <linux/export.h>
> #include <linux/interrupt.h>
> #include <linux/clocksource.h>
> #include <linux/clockchips.h>
> @@ -192,6 +193,19 @@ static void xen_read_wallclock(struct timespec *ts)
> put_cpu_var(xen_vcpu);
> }
>
> +int xen_write_wallclock(const struct timespec *ts)
> +{
> + struct xen_platform_op op;
> +
> + op.cmd = XENPF_settime;
> + op.u.settime.secs = ts->tv_sec;
> + op.u.settime.nsecs = ts->tv_nsec;
> + op.u.settime.system_time = xen_clocksource_read();
> +
> + return HYPERVISOR_dom0_op(&op);
> +}
> +EXPORT_SYMBOL_GPL(xen_write_wallclock);
> +
> static unsigned long xen_get_wallclock(void)
> {
> struct timespec ts;
> @@ -202,7 +216,7 @@ static unsigned long xen_get_wallclock(void)
>
> static int xen_set_wallclock(unsigned long now)
> {
> - struct xen_platform_op op;
> + struct timespec ts;
> int rc;
>
> /* do nothing for domU */
> @@ -218,12 +232,9 @@ static int xen_set_wallclock(unsigned long now)
> * update_persistent_wallclock() is called ~500 ms after 'now'
> * so add an extra 500 ms.
> */
> - op.cmd = XENPF_settime;
> - op.u.settime.secs = now;
> - op.u.settime.nsecs = NSEC_PER_SEC / 2;
> - op.u.settime.system_time = xen_clocksource_read();
> -
> - rc = HYPERVISOR_dom0_op(&op);
> + ts.tv_sec = now;
> + ts.tv_nsec = NSEC_PER_SEC / 2;
> + rc = xen_write_wallclock(&ts);
> WARN(rc != 0, "XENPF_settime failed: now=%ld\n", now);
>
> return rc;
> diff --git a/drivers/xen/privcmd.c b/drivers/xen/privcmd.c
> index ccee0f1..ed2caf3 100644
> --- a/drivers/xen/privcmd.c
> +++ b/drivers/xen/privcmd.c
> @@ -338,6 +338,14 @@ out:
> return ret;
> }
>
> +static long privcmd_ioctl_sync_wallclock(void)
> +{
> + struct timespec ts;
> +
> + getnstimeofday(&ts);
> + return xen_write_wallclock(&ts);
> +}
> +
> static long privcmd_ioctl(struct file *file,
> unsigned int cmd, unsigned long data)
> {
> @@ -357,6 +365,10 @@ static long privcmd_ioctl(struct file *file,
> ret = privcmd_ioctl_mmap_batch(udata);
> break;
>
> + case IOCTL_PRIVCMD_SYNC_WALLCLOCK:
> + ret = privcmd_ioctl_sync_wallclock();
> + break;
> +
> default:
> ret = -EINVAL;
> break;
> diff --git a/include/xen/privcmd.h b/include/xen/privcmd.h
> index 17857fb..d17d087 100644
> --- a/include/xen/privcmd.h
> +++ b/include/xen/privcmd.h
> @@ -66,6 +66,12 @@ struct privcmd_mmapbatch {
> * @cmd: IOCTL_PRIVCMD_HYPERCALL
> * @arg: &privcmd_hypercall_t
> * Return: Value returned from execution of the specified hypercall.
> + *
> + * @cmd: IOCTL_PRIVCMD_SYNC_WALLCLOCK
> + * @arg: Unused.
> + * Synchronizes the Xen wallclock with the current system time.
> + * Return: 0 on success, or -1 on error with errno set to EPERM or
> + * EACCES.
> */
> #define IOCTL_PRIVCMD_HYPERCALL \
> _IOC(_IOC_NONE, 'P', 0, sizeof(struct privcmd_hypercall))
> @@ -73,5 +79,7 @@ struct privcmd_mmapbatch {
> _IOC(_IOC_NONE, 'P', 2, sizeof(struct privcmd_mmap))
> #define IOCTL_PRIVCMD_MMAPBATCH \
> _IOC(_IOC_NONE, 'P', 3, sizeof(struct privcmd_mmapbatch))
> +#define IOCTL_PRIVCMD_SYNC_WALLCLOCK \
> + _IOC(_IOC_NONE, 'P', 5, 0)
>
> #endif /* __LINUX_PUBLIC_PRIVCMD_H__ */
> diff --git a/include/xen/xen-ops.h b/include/xen/xen-ops.h
> index 6a198e4..eefed22 100644
> --- a/include/xen/xen-ops.h
> +++ b/include/xen/xen-ops.h
> @@ -29,4 +29,6 @@ int xen_remap_domain_mfn_range(struct vm_area_struct *vma,
> unsigned long mfn, int nr,
> pgprot_t prot, unsigned domid);
>
> +int xen_write_wallclock(const struct timespec *ts);
> +
> #endif /* INCLUDE_XEN_OPS_H */
> --
> 1.7.2.5
>
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xen.org
> http://lists.xen.org/xen-devel
>

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

Re: [PATCH 3/3] xen/privcmd: add IOCTL_PRIVCMD_SYNC_WALLCLOCK to sync Xen's wallclock [ In reply to ]

david.vrabel at citrix

Oct 12, 2012, 7:02 AM

Post #3 of 9 (565 views)

Permalink

On 12/10/12 14:41, Konrad Rzeszutek Wilk wrote:
> On Fri, Oct 12, 2012 at 01:57:14PM +0100, David Vrabel wrote:
>> From: David Vrabel <david.vrabel@citrix.com>
>>
>> Add a new ioctl to synchronize Xen's wallclock with the current system
>> time.
>>
>> This may be used by the tools to ensure that newly created domains see
>> the correct wallclock time if NTP is not used in dom0 or if domains
>> are started before NTP has synchronized.
>
> So... how does this work with NTPD? As in does ntpd _not_ update the
> hwclock enough?

Once NTPD is synchronized then the kernel updates the wallclock (and the
RTC with patch #1) every 11 mins. I assume this is often enough given
how NTP adjusts the system time.

You only really need the tools to sync wallclock if system time was
stepped at start of day. e.g., init scripts could do something like:

ntpdate pool.ntp.org
hwclock --systohc
xen-wallclock --systowc

David

>> Signed-off-by: David Vrabel <david.vrabel@citrix.com>
>> ---
>> arch/x86/xen/time.c | 25 ++++++++++++++++++-------
>> drivers/xen/privcmd.c | 12 ++++++++++++
>> include/xen/privcmd.h | 8 ++++++++
>> include/xen/xen-ops.h | 2 ++
>> 4 files changed, 40 insertions(+), 7 deletions(-)
>>
>> diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c
>> index 11770d0..d481ac9 100644
>> --- a/arch/x86/xen/time.c
>> +++ b/arch/x86/xen/time.c
>> @@ -8,6 +8,7 @@
>> * Jeremy Fitzhardinge <jeremy@xensource.com>, XenSource Inc, 2007
>> */
>> #include <linux/kernel.h>
>> +#include <linux/export.h>
>> #include <linux/interrupt.h>
>> #include <linux/clocksource.h>
>> #include <linux/clockchips.h>
>> @@ -192,6 +193,19 @@ static void xen_read_wallclock(struct timespec *ts)
>> put_cpu_var(xen_vcpu);
>> }
>>
>> +int xen_write_wallclock(const struct timespec *ts)
>> +{
>> + struct xen_platform_op op;
>> +
>> + op.cmd = XENPF_settime;
>> + op.u.settime.secs = ts->tv_sec;
>> + op.u.settime.nsecs = ts->tv_nsec;
>> + op.u.settime.system_time = xen_clocksource_read();
>> +
>> + return HYPERVISOR_dom0_op(&op);
>> +}
>> +EXPORT_SYMBOL_GPL(xen_write_wallclock);
>> +
>> static unsigned long xen_get_wallclock(void)
>> {
>> struct timespec ts;
>> @@ -202,7 +216,7 @@ static unsigned long xen_get_wallclock(void)
>>
>> static int xen_set_wallclock(unsigned long now)
>> {
>> - struct xen_platform_op op;
>> + struct timespec ts;
>> int rc;
>>
>> /* do nothing for domU */
>> @@ -218,12 +232,9 @@ static int xen_set_wallclock(unsigned long now)
>> * update_persistent_wallclock() is called ~500 ms after 'now'
>> * so add an extra 500 ms.
>> */
>> - op.cmd = XENPF_settime;
>> - op.u.settime.secs = now;
>> - op.u.settime.nsecs = NSEC_PER_SEC / 2;
>> - op.u.settime.system_time = xen_clocksource_read();
>> -
>> - rc = HYPERVISOR_dom0_op(&op);
>> + ts.tv_sec = now;
>> + ts.tv_nsec = NSEC_PER_SEC / 2;
>> + rc = xen_write_wallclock(&ts);
>> WARN(rc != 0, "XENPF_settime failed: now=%ld\n", now);
>>
>> return rc;
>> diff --git a/drivers/xen/privcmd.c b/drivers/xen/privcmd.c
>> index ccee0f1..ed2caf3 100644
>> --- a/drivers/xen/privcmd.c
>> +++ b/drivers/xen/privcmd.c
>> @@ -338,6 +338,14 @@ out:
>> return ret;
>> }
>>
>> +static long privcmd_ioctl_sync_wallclock(void)
>> +{
>> + struct timespec ts;
>> +
>> + getnstimeofday(&ts);
>> + return xen_write_wallclock(&ts);
>> +}
>> +
>> static long privcmd_ioctl(struct file *file,
>> unsigned int cmd, unsigned long data)
>> {
>> @@ -357,6 +365,10 @@ static long privcmd_ioctl(struct file *file,
>> ret = privcmd_ioctl_mmap_batch(udata);
>> break;
>>
>> + case IOCTL_PRIVCMD_SYNC_WALLCLOCK:
>> + ret = privcmd_ioctl_sync_wallclock();
>> + break;
>> +
>> default:
>> ret = -EINVAL;
>> break;
>> diff --git a/include/xen/privcmd.h b/include/xen/privcmd.h
>> index 17857fb..d17d087 100644
>> --- a/include/xen/privcmd.h
>> +++ b/include/xen/privcmd.h
>> @@ -66,6 +66,12 @@ struct privcmd_mmapbatch {
>> * @cmd: IOCTL_PRIVCMD_HYPERCALL
>> * @arg: &privcmd_hypercall_t
>> * Return: Value returned from execution of the specified hypercall.
>> + *
>> + * @cmd: IOCTL_PRIVCMD_SYNC_WALLCLOCK
>> + * @arg: Unused.
>> + * Synchronizes the Xen wallclock with the current system time.
>> + * Return: 0 on success, or -1 on error with errno set to EPERM or
>> + * EACCES.
>> */
>> #define IOCTL_PRIVCMD_HYPERCALL \
>> _IOC(_IOC_NONE, 'P', 0, sizeof(struct privcmd_hypercall))
>> @@ -73,5 +79,7 @@ struct privcmd_mmapbatch {
>> _IOC(_IOC_NONE, 'P', 2, sizeof(struct privcmd_mmap))
>> #define IOCTL_PRIVCMD_MMAPBATCH \
>> _IOC(_IOC_NONE, 'P', 3, sizeof(struct privcmd_mmapbatch))
>> +#define IOCTL_PRIVCMD_SYNC_WALLCLOCK \
>> + _IOC(_IOC_NONE, 'P', 5, 0)
>>
>> #endif /* __LINUX_PUBLIC_PRIVCMD_H__ */
>> diff --git a/include/xen/xen-ops.h b/include/xen/xen-ops.h
>> index 6a198e4..eefed22 100644
>> --- a/include/xen/xen-ops.h
>> +++ b/include/xen/xen-ops.h
>> @@ -29,4 +29,6 @@ int xen_remap_domain_mfn_range(struct vm_area_struct *vma,
>> unsigned long mfn, int nr,
>> pgprot_t prot, unsigned domid);
>>
>> +int xen_write_wallclock(const struct timespec *ts);
>> +
>> #endif /* INCLUDE_XEN_OPS_H */
>> --
>> 1.7.2.5
>>
>>
>> _______________________________________________
>> Xen-devel mailing list
>> Xen-devel@lists.xen.org
>> http://lists.xen.org/xen-devel
>>

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

Re: [PATCH 3/3] xen/privcmd: add IOCTL_PRIVCMD_SYNC_WALLCLOCK to sync Xen's wallclock [ In reply to ]

konrad at kernel

Oct 12, 2012, 7:29 AM

Post #4 of 9 (559 views)

Permalink

On Fri, Oct 12, 2012 at 10:02 AM, David Vrabel <david.vrabel@citrix.com> wrote:
> On 12/10/12 14:41, Konrad Rzeszutek Wilk wrote:
>> On Fri, Oct 12, 2012 at 01:57:14PM +0100, David Vrabel wrote:
>>> From: David Vrabel <david.vrabel@citrix.com>
>>>
>>> Add a new ioctl to synchronize Xen's wallclock with the current system
>>> time.
>>>
>>> This may be used by the tools to ensure that newly created domains see
>>> the correct wallclock time if NTP is not used in dom0 or if domains
>>> are started before NTP has synchronized.
>>
>> So... how does this work with NTPD? As in does ntpd _not_ update the
>> hwclock enough?
>
> Once NTPD is synchronized then the kernel updates the wallclock (and the
> RTC with patch #1) every 11 mins. I assume this is often enough given
> how NTP adjusts the system time.
>
> You only really need the tools to sync wallclock if system time was
> stepped at start of day. e.g., init scripts could do something like:
>
> ntpdate pool.ntp.org
> hwclock --systohc
> xen-wallclock --systowc

I think I am missing something. The hwclock should end up in the
xen_set_wallclock call. And from there on, the ntpd would update the
wallclock if it got skewed enough? Or is the system time not calling
the wall-clock enough? If that is the case, would just adding this in
the crontab be enough:

*/11 * * * * hwclock --systohc

?

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

Re: [PATCH 3/3] xen/privcmd: add IOCTL_PRIVCMD_SYNC_WALLCLOCK to sync Xen's wallclock [ In reply to ]

david.vrabel at citrix

Oct 12, 2012, 7:59 AM

Post #5 of 9 (558 views)

Permalink

On 12/10/12 15:29, Konrad Rzeszutek Wilk wrote:
> On Fri, Oct 12, 2012 at 10:02 AM, David Vrabel <david.vrabel@citrix.com> wrote:
>> On 12/10/12 14:41, Konrad Rzeszutek Wilk wrote:
>>> On Fri, Oct 12, 2012 at 01:57:14PM +0100, David Vrabel wrote:
>>>> From: David Vrabel <david.vrabel@citrix.com>
>>>>
>>>> Add a new ioctl to synchronize Xen's wallclock with the current system
>>>> time.
>>>>
>>>> This may be used by the tools to ensure that newly created domains see
>>>> the correct wallclock time if NTP is not used in dom0 or if domains
>>>> are started before NTP has synchronized.
>>>
>>> So... how does this work with NTPD? As in does ntpd _not_ update the
>>> hwclock enough?
>>
>> Once NTPD is synchronized then the kernel updates the wallclock (and the
>> RTC with patch #1) every 11 mins. I assume this is often enough given
>> how NTP adjusts the system time.
>>
>> You only really need the tools to sync wallclock if system time was
>> stepped at start of day. e.g., init scripts could do something like:
>>
>> ntpdate pool.ntp.org
>> hwclock --systohc
>> xen-wallclock --systowc
>
> I think I am missing something. The hwclock should end up in the
> xen_set_wallclock call. And from there on, the ntpd would update the
> wallclock if it got skewed enough? Or is the system time not calling
> the wall-clock enough? If that is the case, would just adding this in
> the crontab be enough:

hwclock talks to /dev/rtc which writes to the CMOS directly and does not
call update_persistent_clock() (or xen_set_wallclock()).

David

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

Re: [PATCH 3/3] xen/privcmd: add IOCTL_PRIVCMD_SYNC_WALLCLOCK to sync Xen's wallclock [ In reply to ]

konrad.wilk at oracle

Oct 12, 2012, 8:02 AM

Post #6 of 9 (559 views)

Permalink

On Fri, Oct 12, 2012 at 03:59:25PM +0100, David Vrabel wrote:
> On 12/10/12 15:29, Konrad Rzeszutek Wilk wrote:
> > On Fri, Oct 12, 2012 at 10:02 AM, David Vrabel <david.vrabel@citrix.com> wrote:
> >> On 12/10/12 14:41, Konrad Rzeszutek Wilk wrote:
> >>> On Fri, Oct 12, 2012 at 01:57:14PM +0100, David Vrabel wrote:
> >>>> From: David Vrabel <david.vrabel@citrix.com>
> >>>>
> >>>> Add a new ioctl to synchronize Xen's wallclock with the current system
> >>>> time.
> >>>>
> >>>> This may be used by the tools to ensure that newly created domains see
> >>>> the correct wallclock time if NTP is not used in dom0 or if domains
> >>>> are started before NTP has synchronized.
> >>>
> >>> So... how does this work with NTPD? As in does ntpd _not_ update the
> >>> hwclock enough?
> >>
> >> Once NTPD is synchronized then the kernel updates the wallclock (and the
> >> RTC with patch #1) every 11 mins. I assume this is often enough given
> >> how NTP adjusts the system time.
> >>
> >> You only really need the tools to sync wallclock if system time was
> >> stepped at start of day. e.g., init scripts could do something like:
> >>
> >> ntpdate pool.ntp.org
> >> hwclock --systohc
> >> xen-wallclock --systowc
> >
> > I think I am missing something. The hwclock should end up in the
> > xen_set_wallclock call. And from there on, the ntpd would update the
> > wallclock if it got skewed enough? Or is the system time not calling
> > the wall-clock enough? If that is the case, would just adding this in
> > the crontab be enough:
>
> hwclock talks to /dev/rtc which writes to the CMOS directly and does not
> call update_persistent_clock() (or xen_set_wallclock()).

/me scratches his head.

I recall that the update_persistent_clock() was being called.. with
hwclock -w (which is the same as --systohc).

That seems like a bug in the generic code - I would think that
hwclock would update the wallclock, not just the RTC.

Oh, it is whoever calls 'adjtimex' syscall ends up calling in
update_persistent_clock(). So .. ntpdate or ntpd don't call that?

Somebody does.. I hope?

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

Re: [PATCH 3/3] xen/privcmd: add IOCTL_PRIVCMD_SYNC_WALLCLOCK to sync Xen's wallclock [ In reply to ]

david.vrabel at citrix

Oct 12, 2012, 9:14 AM

Post #7 of 9 (556 views)

Permalink

On 12/10/12 16:02, Konrad Rzeszutek Wilk wrote:
> On Fri, Oct 12, 2012 at 03:59:25PM +0100, David Vrabel wrote:
>> On 12/10/12 15:29, Konrad Rzeszutek Wilk wrote:
>>> On Fri, Oct 12, 2012 at 10:02 AM, David Vrabel <david.vrabel@citrix.com> wrote:
>>>> On 12/10/12 14:41, Konrad Rzeszutek Wilk wrote:
>>>>> On Fri, Oct 12, 2012 at 01:57:14PM +0100, David Vrabel wrote:
>>>>>> From: David Vrabel <david.vrabel@citrix.com>
>>>>>>
>>>>>> Add a new ioctl to synchronize Xen's wallclock with the current system
>>>>>> time.
>>>>>>
>>>>>> This may be used by the tools to ensure that newly created domains see
>>>>>> the correct wallclock time if NTP is not used in dom0 or if domains
>>>>>> are started before NTP has synchronized.
>>>>>
>>>>> So... how does this work with NTPD? As in does ntpd _not_ update the
>>>>> hwclock enough?
>>>>
>>>> Once NTPD is synchronized then the kernel updates the wallclock (and the
>>>> RTC with patch #1) every 11 mins. I assume this is often enough given
>>>> how NTP adjusts the system time.
>>>>
>>>> You only really need the tools to sync wallclock if system time was
>>>> stepped at start of day. e.g., init scripts could do something like:
>>>>
>>>> ntpdate pool.ntp.org
>>>> hwclock --systohc
>>>> xen-wallclock --systowc
>>>
>>> I think I am missing something. The hwclock should end up in the
>>> xen_set_wallclock call. And from there on, the ntpd would update the
>>> wallclock if it got skewed enough? Or is the system time not calling
>>> the wall-clock enough? If that is the case, would just adding this in
>>> the crontab be enough:
>>
>> hwclock talks to /dev/rtc which writes to the CMOS directly and does not
>> call update_persistent_clock() (or xen_set_wallclock()).
>
> /me scratches his head.
>
> I recall that the update_persistent_clock() was being called.. with
> hwclock -w (which is the same as --systohc).

No,

> That seems like a bug in the generic code - I would think that
> hwclock would update the wallclock, not just the RTC.

If we wanted hwclock to also update the xen wallclock we would need a
xen-specific rtc driver that updated both rtc and wallclock.

> Oh, it is whoever calls 'adjtimex' syscall ends up calling in
> update_persistent_clock(). So .. ntpdate or ntpd don't call that?

ntpd does call adjtimex so if you are running ntpd and it is
synchronized to its clock source you do get the periodic sync of the
wallclock.

David

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

Re: [PATCH 3/3] xen/privcmd: add IOCTL_PRIVCMD_SYNC_WALLCLOCK to sync Xen's wallclock [ In reply to ]

konrad.wilk at oracle

Oct 12, 2012, 9:16 AM

Post #8 of 9 (559 views)

Permalink

On Fri, Oct 12, 2012 at 05:14:16PM +0100, David Vrabel wrote:
> On 12/10/12 16:02, Konrad Rzeszutek Wilk wrote:
> > On Fri, Oct 12, 2012 at 03:59:25PM +0100, David Vrabel wrote:
> >> On 12/10/12 15:29, Konrad Rzeszutek Wilk wrote:
> >>> On Fri, Oct 12, 2012 at 10:02 AM, David Vrabel <david.vrabel@citrix.com> wrote:
> >>>> On 12/10/12 14:41, Konrad Rzeszutek Wilk wrote:
> >>>>> On Fri, Oct 12, 2012 at 01:57:14PM +0100, David Vrabel wrote:
> >>>>>> From: David Vrabel <david.vrabel@citrix.com>
> >>>>>>
> >>>>>> Add a new ioctl to synchronize Xen's wallclock with the current system
> >>>>>> time.
> >>>>>>
> >>>>>> This may be used by the tools to ensure that newly created domains see
> >>>>>> the correct wallclock time if NTP is not used in dom0 or if domains
> >>>>>> are started before NTP has synchronized.
> >>>>>
> >>>>> So... how does this work with NTPD? As in does ntpd _not_ update the
> >>>>> hwclock enough?
> >>>>
> >>>> Once NTPD is synchronized then the kernel updates the wallclock (and the
> >>>> RTC with patch #1) every 11 mins. I assume this is often enough given
> >>>> how NTP adjusts the system time.
> >>>>
> >>>> You only really need the tools to sync wallclock if system time was
> >>>> stepped at start of day. e.g., init scripts could do something like:
> >>>>
> >>>> ntpdate pool.ntp.org
> >>>> hwclock --systohc
> >>>> xen-wallclock --systowc
> >>>
> >>> I think I am missing something. The hwclock should end up in the
> >>> xen_set_wallclock call. And from there on, the ntpd would update the
> >>> wallclock if it got skewed enough? Or is the system time not calling
> >>> the wall-clock enough? If that is the case, would just adding this in
> >>> the crontab be enough:
> >>
> >> hwclock talks to /dev/rtc which writes to the CMOS directly and does not
> >> call update_persistent_clock() (or xen_set_wallclock()).
> >
> > /me scratches his head.
> >
> > I recall that the update_persistent_clock() was being called.. with
> > hwclock -w (which is the same as --systohc).
>
> No,
>
> > That seems like a bug in the generic code - I would think that
> > hwclock would update the wallclock, not just the RTC.
>
> If we wanted hwclock to also update the xen wallclock we would need a
> xen-specific rtc driver that updated both rtc and wallclock.
>
> > Oh, it is whoever calls 'adjtimex' syscall ends up calling in
> > update_persistent_clock(). So .. ntpdate or ntpd don't call that?
>
> ntpd does call adjtimex so if you are running ntpd and it is
> synchronized to its clock source you do get the periodic sync of the
> wallclock.

So, having both ntpd (which calls call adjtimex) - which would run in the
background; and hwclock run at startup - (which updates the RTC, which ends
- I hope - being trapped by the hypervisor), we end up with the correct
time, right?

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

Re: [PATCH 3/3] xen/privcmd: add IOCTL_PRIVCMD_SYNC_WALLCLOCK to sync Xen's wallclock [ In reply to ]

david.vrabel at citrix

Oct 12, 2012, 9:30 AM

Post #9 of 9 (554 views)

Permalink

On 12/10/12 17:16, Konrad Rzeszutek Wilk wrote:
>
> So, having both ntpd (which calls call adjtimex) - which would run in the
> background; and hwclock run at startup - (which updates the RTC, which ends
> - I hope - being trapped by the hypervisor), we end up with the correct
> time, right?

Yes, unless you start a guest before ntpd has synchronized.

David

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel