* [PATCH] x86/pvclock-xen: zero last_value on resume
@ 2010-10-26 16:59 ` Jeremy Fitzhardinge
0 siblings, 0 replies; 22+ messages in thread
From: Jeremy Fitzhardinge @ 2010-10-26 16:59 UTC (permalink / raw)
To: H. Peter Anvin
Cc: Linux Virtualization, Glauber Costa, Avi Kivity,
the arch/x86 maintainers, Linux Virtualization,
Linux Kernel Mailing List, Xen-devel, kvm-devel,
Cédric Schieli, Eelco Dolstra, Olivier Hanesse
[-- Attachment #1: Type: text/plain, Size: 1899 bytes --]
If the guest domain has been suspend/resumed or migrated, then the
system clock backing the pvclock clocksource may revert to a smaller
value (ie, can be non-monotonic across the migration/save-restore).
Make sure we zero last_value in that case so that the domain
continues to see clock updates.
[ I don't know if kvm needs an analogous fix or not. ]
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Cc: Stable Kernel <stable@kernel.org>
Reported-by: Eelco Dolstra <e.dolstra@tudelft.nl>
Reported-by: Olivier Hanesse <olivier.hanesse@gmail.com>
Bisected-by: Cédric Schieli <cschieli@gmail.com>
Tested-by: Cédric Schieli <cschieli@gmail.com>
diff --git a/arch/x86/include/asm/pvclock.h b/arch/x86/include/asm/pvclock.h
index cd02f32..6226870 100644
--- a/arch/x86/include/asm/pvclock.h
+++ b/arch/x86/include/asm/pvclock.h
@@ -11,5 +11,6 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src);
void pvclock_read_wallclock(struct pvclock_wall_clock *wall,
struct pvclock_vcpu_time_info *vcpu,
struct timespec *ts);
+void pvclock_resume(void);
#endif /* _ASM_X86_PVCLOCK_H */
diff --git a/arch/x86/kernel/pvclock.c b/arch/x86/kernel/pvclock.c
index 239427c..a4f07c1 100644
--- a/arch/x86/kernel/pvclock.c
+++ b/arch/x86/kernel/pvclock.c
@@ -120,6 +120,11 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src)
static atomic64_t last_value = ATOMIC64_INIT(0);
+void pvclock_resume(void)
+{
+ atomic64_set(&last_value, 0);
+}
+
cycle_t pvclock_clocksource_read(struct pvclock_vcpu_time_info *src)
{
struct pvclock_shadow_time shadow;
diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c
index b2bb5aa..5da5e53 100644
--- a/arch/x86/xen/time.c
+++ b/arch/x86/xen/time.c
@@ -426,6 +426,8 @@ void xen_timer_resume(void)
{
int cpu;
+ pvclock_resume();
+
if (xen_clockevent != &xen_vcpuop_clockevent)
return;
[-- Attachment #2: x86-pvclock-resume.patch --]
[-- Type: text/plain, Size: 2064 bytes --]
From 29acbb4e1d93e719250648db1ce8c7a24144fd86 Mon Sep 17 00:00:00 2001
From: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Date: Mon, 25 Oct 2010 16:53:46 -0700
Subject: [PATCH] x86/pvclock: zero last_value on resume
If the guest domain has been suspend/resumed or migrated, then the
system clock backing the pvclock clocksource may revert to a smaller
value (ie, can be non-monotonic across the migration/save-restore).
Make sure we zero last_value in that case so that the domain
continues to see clock updates.
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Cc: Stable Kernel <stable@kernel.org>
Reported-by: Eelco Dolstra <e.dolstra@tudelft.nl>
Reported-by: Olivier Hanesse <olivier.hanesse@gmail.com>
Bisected-by: Cédric Schieli <cschieli@gmail.com>
Tested-by: Cédric Schieli <cschieli@gmail.com>
diff --git a/arch/x86/include/asm/pvclock.h b/arch/x86/include/asm/pvclock.h
index cd02f32..6226870 100644
--- a/arch/x86/include/asm/pvclock.h
+++ b/arch/x86/include/asm/pvclock.h
@@ -11,5 +11,6 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src);
void pvclock_read_wallclock(struct pvclock_wall_clock *wall,
struct pvclock_vcpu_time_info *vcpu,
struct timespec *ts);
+void pvclock_resume(void);
#endif /* _ASM_X86_PVCLOCK_H */
diff --git a/arch/x86/kernel/pvclock.c b/arch/x86/kernel/pvclock.c
index 239427c..a4f07c1 100644
--- a/arch/x86/kernel/pvclock.c
+++ b/arch/x86/kernel/pvclock.c
@@ -120,6 +120,11 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src)
static atomic64_t last_value = ATOMIC64_INIT(0);
+void pvclock_resume(void)
+{
+ atomic64_set(&last_value, 0);
+}
+
cycle_t pvclock_clocksource_read(struct pvclock_vcpu_time_info *src)
{
struct pvclock_shadow_time shadow;
diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c
index b2bb5aa..5da5e53 100644
--- a/arch/x86/xen/time.c
+++ b/arch/x86/xen/time.c
@@ -426,6 +426,8 @@ void xen_timer_resume(void)
{
int cpu;
+ pvclock_resume();
+
if (xen_clockevent != &xen_vcpuop_clockevent)
return;
^ permalink raw reply related [flat|nested] 22+ messages in thread
* [PATCH] x86/pvclock-xen: zero last_value on resume
@ 2010-10-26 16:59 ` Jeremy Fitzhardinge
0 siblings, 0 replies; 22+ messages in thread
From: Jeremy Fitzhardinge @ 2010-10-26 16:59 UTC (permalink / raw)
To: H. Peter Anvin
Cc: Linux Virtualization, Glauber Costa, Avi Kivity,
the arch/x86 maintainers
[-- Attachment #1: Type: text/plain, Size: 1899 bytes --]
If the guest domain has been suspend/resumed or migrated, then the
system clock backing the pvclock clocksource may revert to a smaller
value (ie, can be non-monotonic across the migration/save-restore).
Make sure we zero last_value in that case so that the domain
continues to see clock updates.
[ I don't know if kvm needs an analogous fix or not. ]
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Cc: Stable Kernel <stable@kernel.org>
Reported-by: Eelco Dolstra <e.dolstra@tudelft.nl>
Reported-by: Olivier Hanesse <olivier.hanesse@gmail.com>
Bisected-by: Cédric Schieli <cschieli@gmail.com>
Tested-by: Cédric Schieli <cschieli@gmail.com>
diff --git a/arch/x86/include/asm/pvclock.h b/arch/x86/include/asm/pvclock.h
index cd02f32..6226870 100644
--- a/arch/x86/include/asm/pvclock.h
+++ b/arch/x86/include/asm/pvclock.h
@@ -11,5 +11,6 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src);
void pvclock_read_wallclock(struct pvclock_wall_clock *wall,
struct pvclock_vcpu_time_info *vcpu,
struct timespec *ts);
+void pvclock_resume(void);
#endif /* _ASM_X86_PVCLOCK_H */
diff --git a/arch/x86/kernel/pvclock.c b/arch/x86/kernel/pvclock.c
index 239427c..a4f07c1 100644
--- a/arch/x86/kernel/pvclock.c
+++ b/arch/x86/kernel/pvclock.c
@@ -120,6 +120,11 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src)
static atomic64_t last_value = ATOMIC64_INIT(0);
+void pvclock_resume(void)
+{
+ atomic64_set(&last_value, 0);
+}
+
cycle_t pvclock_clocksource_read(struct pvclock_vcpu_time_info *src)
{
struct pvclock_shadow_time shadow;
diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c
index b2bb5aa..5da5e53 100644
--- a/arch/x86/xen/time.c
+++ b/arch/x86/xen/time.c
@@ -426,6 +426,8 @@ void xen_timer_resume(void)
{
int cpu;
+ pvclock_resume();
+
if (xen_clockevent != &xen_vcpuop_clockevent)
return;
[-- Attachment #2: x86-pvclock-resume.patch --]
[-- Type: text/plain, Size: 2064 bytes --]
From 29acbb4e1d93e719250648db1ce8c7a24144fd86 Mon Sep 17 00:00:00 2001
From: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Date: Mon, 25 Oct 2010 16:53:46 -0700
Subject: [PATCH] x86/pvclock: zero last_value on resume
If the guest domain has been suspend/resumed or migrated, then the
system clock backing the pvclock clocksource may revert to a smaller
value (ie, can be non-monotonic across the migration/save-restore).
Make sure we zero last_value in that case so that the domain
continues to see clock updates.
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Cc: Stable Kernel <stable@kernel.org>
Reported-by: Eelco Dolstra <e.dolstra@tudelft.nl>
Reported-by: Olivier Hanesse <olivier.hanesse@gmail.com>
Bisected-by: Cédric Schieli <cschieli@gmail.com>
Tested-by: Cédric Schieli <cschieli@gmail.com>
diff --git a/arch/x86/include/asm/pvclock.h b/arch/x86/include/asm/pvclock.h
index cd02f32..6226870 100644
--- a/arch/x86/include/asm/pvclock.h
+++ b/arch/x86/include/asm/pvclock.h
@@ -11,5 +11,6 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src);
void pvclock_read_wallclock(struct pvclock_wall_clock *wall,
struct pvclock_vcpu_time_info *vcpu,
struct timespec *ts);
+void pvclock_resume(void);
#endif /* _ASM_X86_PVCLOCK_H */
diff --git a/arch/x86/kernel/pvclock.c b/arch/x86/kernel/pvclock.c
index 239427c..a4f07c1 100644
--- a/arch/x86/kernel/pvclock.c
+++ b/arch/x86/kernel/pvclock.c
@@ -120,6 +120,11 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src)
static atomic64_t last_value = ATOMIC64_INIT(0);
+void pvclock_resume(void)
+{
+ atomic64_set(&last_value, 0);
+}
+
cycle_t pvclock_clocksource_read(struct pvclock_vcpu_time_info *src)
{
struct pvclock_shadow_time shadow;
diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c
index b2bb5aa..5da5e53 100644
--- a/arch/x86/xen/time.c
+++ b/arch/x86/xen/time.c
@@ -426,6 +426,8 @@ void xen_timer_resume(void)
{
int cpu;
+ pvclock_resume();
+
if (xen_clockevent != &xen_vcpuop_clockevent)
return;
^ permalink raw reply related [flat|nested] 22+ messages in thread
* Re: [PATCH] x86/pvclock-xen: zero last_value on resume
2010-10-26 16:59 ` Jeremy Fitzhardinge
@ 2010-10-26 17:48 ` Glauber Costa
-1 siblings, 0 replies; 22+ messages in thread
From: Glauber Costa @ 2010-10-26 17:48 UTC (permalink / raw)
To: Jeremy Fitzhardinge
Cc: H. Peter Anvin, Linux Virtualization, Avi Kivity,
the arch/x86 maintainers, Linux Kernel Mailing List, Xen-devel,
kvm-devel, Cédric Schieli, Eelco Dolstra, Olivier Hanesse
On Tue, 2010-10-26 at 09:59 -0700, Jeremy Fitzhardinge wrote:
>
> If the guest domain has been suspend/resumed or migrated, then the
> system clock backing the pvclock clocksource may revert to a smaller
> value (ie, can be non-monotonic across the migration/save-restore).
> Make sure we zero last_value in that case so that the domain
> continues to see clock updates.
>
> [ I don't know if kvm needs an analogous fix or not. ]
After migration, save/restore, etc, we issue an ioctl where we tell
the host the last clock value. That (in theory) guarantees monotonicity.
I am not opposed to this patch in any way, however.
^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: [PATCH] x86/pvclock-xen: zero last_value on resume
@ 2010-10-26 17:48 ` Glauber Costa
0 siblings, 0 replies; 22+ messages in thread
From: Glauber Costa @ 2010-10-26 17:48 UTC (permalink / raw)
To: Jeremy Fitzhardinge
Cc: Xen-devel, kvm-devel, Eelco Dolstra, the arch/x86 maintainers,
Linux Kernel Mailing List, Linux Virtualization, Olivier Hanesse,
Avi Kivity, H. Peter Anvin
On Tue, 2010-10-26 at 09:59 -0700, Jeremy Fitzhardinge wrote:
>
> If the guest domain has been suspend/resumed or migrated, then the
> system clock backing the pvclock clocksource may revert to a smaller
> value (ie, can be non-monotonic across the migration/save-restore).
> Make sure we zero last_value in that case so that the domain
> continues to see clock updates.
>
> [ I don't know if kvm needs an analogous fix or not. ]
After migration, save/restore, etc, we issue an ioctl where we tell
the host the last clock value. That (in theory) guarantees monotonicity.
I am not opposed to this patch in any way, however.
^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: [PATCH] x86/pvclock-xen: zero last_value on resume
2010-10-26 16:59 ` Jeremy Fitzhardinge
(?)
(?)
@ 2010-10-26 17:48 ` Glauber Costa
-1 siblings, 0 replies; 22+ messages in thread
From: Glauber Costa @ 2010-10-26 17:48 UTC (permalink / raw)
To: Jeremy Fitzhardinge
Cc: Xen-devel, kvm-devel, Eelco Dolstra, the arch/x86 maintainers,
Linux Kernel Mailing List, Linux Virtualization, Olivier Hanesse,
Avi Kivity, H. Peter Anvin
On Tue, 2010-10-26 at 09:59 -0700, Jeremy Fitzhardinge wrote:
>
> If the guest domain has been suspend/resumed or migrated, then the
> system clock backing the pvclock clocksource may revert to a smaller
> value (ie, can be non-monotonic across the migration/save-restore).
> Make sure we zero last_value in that case so that the domain
> continues to see clock updates.
>
> [ I don't know if kvm needs an analogous fix or not. ]
After migration, save/restore, etc, we issue an ioctl where we tell
the host the last clock value. That (in theory) guarantees monotonicity.
I am not opposed to this patch in any way, however.
^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: [Xen-devel] Re: [PATCH] x86/pvclock-xen: zero last_value on resume
2010-10-26 17:48 ` Glauber Costa
(?)
(?)
@ 2010-10-27 18:53 ` Jeremy Fitzhardinge
2010-10-27 20:59 ` H. Peter Anvin
2010-10-27 20:59 ` H. Peter Anvin
-1 siblings, 2 replies; 22+ messages in thread
From: Jeremy Fitzhardinge @ 2010-10-27 18:53 UTC (permalink / raw)
To: Glauber Costa
Cc: Xen-devel, kvm-devel, Eelco Dolstra, the arch/x86 maintainers,
Linux Kernel Mailing List, Linux Virtualization, Olivier Hanesse,
Avi Kivity, H. Peter Anvin
On 10/26/2010 10:48 AM, Glauber Costa wrote:
> On Tue, 2010-10-26 at 09:59 -0700, Jeremy Fitzhardinge wrote:
>> If the guest domain has been suspend/resumed or migrated, then the
>> system clock backing the pvclock clocksource may revert to a smaller
>> value (ie, can be non-monotonic across the migration/save-restore).
>> Make sure we zero last_value in that case so that the domain
>> continues to see clock updates.
>>
>> [ I don't know if kvm needs an analogous fix or not. ]
> After migration, save/restore, etc, we issue an ioctl where we tell
> the host the last clock value. That (in theory) guarantees monotonicity.
>
> I am not opposed to this patch in any way, however.
Thanks.
HPA, do you want to take this, or shall I send it on?
Thanks,
J
^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: [Xen-devel] Re: [PATCH] x86/pvclock-xen: zero last_value on resume
2010-10-26 17:48 ` Glauber Costa
(?)
@ 2010-10-27 18:53 ` Jeremy Fitzhardinge
-1 siblings, 0 replies; 22+ messages in thread
From: Jeremy Fitzhardinge @ 2010-10-27 18:53 UTC (permalink / raw)
To: Glauber Costa
Cc: Xen-devel, kvm-devel, Eelco Dolstra, the arch/x86 maintainers,
Linux Kernel Mailing List, Linux Virtualization, Olivier Hanesse,
Avi Kivity, H. Peter Anvin
On 10/26/2010 10:48 AM, Glauber Costa wrote:
> On Tue, 2010-10-26 at 09:59 -0700, Jeremy Fitzhardinge wrote:
>> If the guest domain has been suspend/resumed or migrated, then the
>> system clock backing the pvclock clocksource may revert to a smaller
>> value (ie, can be non-monotonic across the migration/save-restore).
>> Make sure we zero last_value in that case so that the domain
>> continues to see clock updates.
>>
>> [ I don't know if kvm needs an analogous fix or not. ]
> After migration, save/restore, etc, we issue an ioctl where we tell
> the host the last clock value. That (in theory) guarantees monotonicity.
>
> I am not opposed to this patch in any way, however.
Thanks.
HPA, do you want to take this, or shall I send it on?
Thanks,
J
^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: [Xen-devel] Re: [PATCH] x86/pvclock-xen: zero last_value on resume
2010-10-27 18:53 ` Jeremy Fitzhardinge
2010-10-27 20:59 ` H. Peter Anvin
@ 2010-10-27 20:59 ` H. Peter Anvin
2010-11-03 12:46 ` Ian Campbell
2010-11-03 12:46 ` Ian Campbell
1 sibling, 2 replies; 22+ messages in thread
From: H. Peter Anvin @ 2010-10-27 20:59 UTC (permalink / raw)
To: Jeremy Fitzhardinge, Glauber Costa
Cc: Xen-devel, kvm-devel, Eelco Dolstra, the arch/x86 maintainers,
Linux Kernel Mailing List, Linux Virtualization, Olivier Hanesse,
Avi Kivity
I'll check it this evening when I'm at a working network again :(
"Jeremy Fitzhardinge" <jeremy@goop.org> wrote:
> On 10/26/2010 10:48 AM, Glauber Costa wrote:
>> On Tue, 2010-10-26 at 09:59 -0700, Jeremy Fitzhardinge wrote:
>>> If the guest domain has been suspend/resumed or migrated, then the
>>> system clock backing the pvclock clocksource may revert to a smaller
>>> value (ie, can be non-monotonic across the migration/save-restore).
>>> Make sure we zero last_value in that case so that the domain
>>> continues to see clock updates.
>>>
>>> [ I don't know if kvm needs an analogous fix or not. ]
>> After migration, save/restore, etc, we issue an ioctl where we tell
>> the host the last clock value. That (in theory) guarantees
>monotonicity.
>>
>> I am not opposed to this patch in any way, however.
>
>Thanks.
>
>HPA, do you want to take this, or shall I send it on?
>
>Thanks,
> J
--
Sent from my mobile phone. Please pardon any lack of formatting.
^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: [Xen-devel] Re: [PATCH] x86/pvclock-xen: zero last_value on resume
2010-10-27 18:53 ` Jeremy Fitzhardinge
@ 2010-10-27 20:59 ` H. Peter Anvin
2010-10-27 20:59 ` H. Peter Anvin
1 sibling, 0 replies; 22+ messages in thread
From: H. Peter Anvin @ 2010-10-27 20:59 UTC (permalink / raw)
To: Jeremy Fitzhardinge, Glauber Costa
Cc: Xen-devel, kvm-devel, Eelco Dolstra, the arch/x86 maintainers,
Linux Kernel Mailing List, Linux Virtualization, Olivier Hanesse,
Avi Kivity
I'll check it this evening when I'm at a working network again :(
"Jeremy Fitzhardinge" <jeremy@goop.org> wrote:
> On 10/26/2010 10:48 AM, Glauber Costa wrote:
>> On Tue, 2010-10-26 at 09:59 -0700, Jeremy Fitzhardinge wrote:
>>> If the guest domain has been suspend/resumed or migrated, then the
>>> system clock backing the pvclock clocksource may revert to a smaller
>>> value (ie, can be non-monotonic across the migration/save-restore).
>>> Make sure we zero last_value in that case so that the domain
>>> continues to see clock updates.
>>>
>>> [ I don't know if kvm needs an analogous fix or not. ]
>> After migration, save/restore, etc, we issue an ioctl where we tell
>> the host the last clock value. That (in theory) guarantees
>monotonicity.
>>
>> I am not opposed to this patch in any way, however.
>
>Thanks.
>
>HPA, do you want to take this, or shall I send it on?
>
>Thanks,
> J
--
Sent from my mobile phone. Please pardon any lack of formatting.
^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: [Xen-devel] Re: [PATCH] x86/pvclock-xen: zero last_value on resume
2010-10-27 20:59 ` H. Peter Anvin
2010-11-03 12:46 ` Ian Campbell
@ 2010-11-03 12:46 ` Ian Campbell
1 sibling, 0 replies; 22+ messages in thread
From: Ian Campbell @ 2010-11-03 12:46 UTC (permalink / raw)
To: H. Peter Anvin
Cc: Jeremy Fitzhardinge, Glauber Costa, Xen-devel, kvm-devel,
Eelco Dolstra, the arch/x86 maintainers,
Linux Kernel Mailing List, Linux Virtualization, Olivier Hanesse,
Avi Kivity
On Wed, 2010-10-27 at 13:59 -0700, H. Peter Anvin wrote:
> I'll check it this evening when I'm at a working network again :(
Did this get applied? It seems to affect 2.6.32.x too
(http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=602273) so can we tag
it for stable as well?
Thanks,
Ian.
>
> "Jeremy Fitzhardinge" <jeremy@goop.org> wrote:
>
> > On 10/26/2010 10:48 AM, Glauber Costa wrote:
> >> On Tue, 2010-10-26 at 09:59 -0700, Jeremy Fitzhardinge wrote:
> >>> If the guest domain has been suspend/resumed or migrated, then the
> >>> system clock backing the pvclock clocksource may revert to a smaller
> >>> value (ie, can be non-monotonic across the migration/save-restore).
> >>> Make sure we zero last_value in that case so that the domain
> >>> continues to see clock updates.
> >>>
> >>> [ I don't know if kvm needs an analogous fix or not. ]
> >> After migration, save/restore, etc, we issue an ioctl where we tell
> >> the host the last clock value. That (in theory) guarantees
> >monotonicity.
> >>
> >> I am not opposed to this patch in any way, however.
> >
> >Thanks.
> >
> >HPA, do you want to take this, or shall I send it on?
> >
> >Thanks,
> > J
>
--
Ian Campbell
BOFH excuse #191:
Just type 'mv * /dev/null'.
^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: Re: [PATCH] x86/pvclock-xen: zero last_value on resume
@ 2010-11-03 12:46 ` Ian Campbell
0 siblings, 0 replies; 22+ messages in thread
From: Ian Campbell @ 2010-11-03 12:46 UTC (permalink / raw)
To: H. Peter Anvin
Cc: Jeremy Fitzhardinge, Xen-devel, kvm-devel, Dolstra,
the arch/x86 maintainers, Linux Kernel Mailing List,
Virtualization, Hanesse, Eelco, Olivier, Linux, Glauber Costa,
Avi Kivity
On Wed, 2010-10-27 at 13:59 -0700, H. Peter Anvin wrote:
> I'll check it this evening when I'm at a working network again :(
Did this get applied? It seems to affect 2.6.32.x too
(http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=602273) so can we tag
it for stable as well?
Thanks,
Ian.
>
> "Jeremy Fitzhardinge" <jeremy@goop.org> wrote:
>
> > On 10/26/2010 10:48 AM, Glauber Costa wrote:
> >> On Tue, 2010-10-26 at 09:59 -0700, Jeremy Fitzhardinge wrote:
> >>> If the guest domain has been suspend/resumed or migrated, then the
> >>> system clock backing the pvclock clocksource may revert to a smaller
> >>> value (ie, can be non-monotonic across the migration/save-restore).
> >>> Make sure we zero last_value in that case so that the domain
> >>> continues to see clock updates.
> >>>
> >>> [ I don't know if kvm needs an analogous fix or not. ]
> >> After migration, save/restore, etc, we issue an ioctl where we tell
> >> the host the last clock value. That (in theory) guarantees
> >monotonicity.
> >>
> >> I am not opposed to this patch in any way, however.
> >
> >Thanks.
> >
> >HPA, do you want to take this, or shall I send it on?
> >
> >Thanks,
> > J
>
--
Ian Campbell
BOFH excuse #191:
Just type 'mv * /dev/null'.
^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: [Xen-devel] Re: [PATCH] x86/pvclock-xen: zero last_value on resume
2010-10-27 20:59 ` H. Peter Anvin
@ 2010-11-03 12:46 ` Ian Campbell
2010-11-03 12:46 ` Ian Campbell
1 sibling, 0 replies; 22+ messages in thread
From: Ian Campbell @ 2010-11-03 12:46 UTC (permalink / raw)
To: H. Peter Anvin
Cc: Xen-devel, kvm-devel, Dolstra, the arch/x86 maintainers,
Linux Kernel Mailing List, Virtualization, Hanesse, Eelco,
Olivier, Linux, Glauber Costa, Avi Kivity
On Wed, 2010-10-27 at 13:59 -0700, H. Peter Anvin wrote:
> I'll check it this evening when I'm at a working network again :(
Did this get applied? It seems to affect 2.6.32.x too
(http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=602273) so can we tag
it for stable as well?
Thanks,
Ian.
>
> "Jeremy Fitzhardinge" <jeremy@goop.org> wrote:
>
> > On 10/26/2010 10:48 AM, Glauber Costa wrote:
> >> On Tue, 2010-10-26 at 09:59 -0700, Jeremy Fitzhardinge wrote:
> >>> If the guest domain has been suspend/resumed or migrated, then the
> >>> system clock backing the pvclock clocksource may revert to a smaller
> >>> value (ie, can be non-monotonic across the migration/save-restore).
> >>> Make sure we zero last_value in that case so that the domain
> >>> continues to see clock updates.
> >>>
> >>> [ I don't know if kvm needs an analogous fix or not. ]
> >> After migration, save/restore, etc, we issue an ioctl where we tell
> >> the host the last clock value. That (in theory) guarantees
> >monotonicity.
> >>
> >> I am not opposed to this patch in any way, however.
> >
> >Thanks.
> >
> >HPA, do you want to take this, or shall I send it on?
> >
> >Thanks,
> > J
>
--
Ian Campbell
BOFH excuse #191:
Just type 'mv * /dev/null'.
^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: Re: [PATCH] x86/pvclock-xen: zero last_value on resume
@ 2010-11-03 12:46 ` Ian Campbell
0 siblings, 0 replies; 22+ messages in thread
From: Ian Campbell @ 2010-11-03 12:46 UTC (permalink / raw)
To: H. Peter Anvin
Cc: Jeremy Fitzhardinge, Xen-devel, kvm-devel, Dolstra,
the arch/x86 maintainers, Linux Kernel Mailing List,
Virtualization, Hanesse, Eelco, OlivierLinux, Glauber Costa,
Avi Kivity
On Wed, 2010-10-27 at 13:59 -0700, H. Peter Anvin wrote:
> I'll check it this evening when I'm at a working network again :(
Did this get applied? It seems to affect 2.6.32.x too
(http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=602273) so can we tag
it for stable as well?
Thanks,
Ian.
>
> "Jeremy Fitzhardinge" <jeremy@goop.org> wrote:
>
> > On 10/26/2010 10:48 AM, Glauber Costa wrote:
> >> On Tue, 2010-10-26 at 09:59 -0700, Jeremy Fitzhardinge wrote:
> >>> If the guest domain has been suspend/resumed or migrated, then the
> >>> system clock backing the pvclock clocksource may revert to a smaller
> >>> value (ie, can be non-monotonic across the migration/save-restore).
> >>> Make sure we zero last_value in that case so that the domain
> >>> continues to see clock updates.
> >>>
> >>> [ I don't know if kvm needs an analogous fix or not. ]
> >> After migration, save/restore, etc, we issue an ioctl where we tell
> >> the host the last clock value. That (in theory) guarantees
> >monotonicity.
> >>
> >> I am not opposed to this patch in any way, however.
> >
> >Thanks.
> >
> >HPA, do you want to take this, or shall I send it on?
> >
> >Thanks,
> > J
>
--
Ian Campbell
BOFH excuse #191:
Just type 'mv * /dev/null'.
^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: [PATCH] x86/pvclock-xen: zero last_value on resume
2010-10-26 16:59 ` Jeremy Fitzhardinge
` (3 preceding siblings ...)
(?)
@ 2010-11-22 20:16 ` Jeremy Fitzhardinge
2010-11-26 10:40 ` Ingo Molnar
2010-11-26 10:40 ` Ingo Molnar
-1 siblings, 2 replies; 22+ messages in thread
From: Jeremy Fitzhardinge @ 2010-11-22 20:16 UTC (permalink / raw)
To: H. Peter Anvin
Cc: Xen-devel, kvm-devel, Glauber Costa, the arch/x86 maintainers,
Linux Kernel Mailing List, Linux Virtualization, Olivier Hanesse,
Avi Kivity, Eelco Dolstra
On 10/26/2010 09:59 AM, Jeremy Fitzhardinge wrote:
>
> If the guest domain has been suspend/resumed or migrated, then the
> system clock backing the pvclock clocksource may revert to a smaller
> value (ie, can be non-monotonic across the migration/save-restore).
> Make sure we zero last_value in that case so that the domain
> continues to see clock updates.
Ping? Looks like this fell through the gaps.
J
> [ I don't know if kvm needs an analogous fix or not. ]
>
> Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
> Cc: Stable Kernel <stable@kernel.org>
> Reported-by: Eelco Dolstra <e.dolstra@tudelft.nl>
> Reported-by: Olivier Hanesse <olivier.hanesse@gmail.com>
> Bisected-by: Cédric Schieli <cschieli@gmail.com>
> Tested-by: Cédric Schieli <cschieli@gmail.com>
>
> diff --git a/arch/x86/include/asm/pvclock.h b/arch/x86/include/asm/pvclock.h
> index cd02f32..6226870 100644
> --- a/arch/x86/include/asm/pvclock.h
> +++ b/arch/x86/include/asm/pvclock.h
> @@ -11,5 +11,6 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src);
> void pvclock_read_wallclock(struct pvclock_wall_clock *wall,
> struct pvclock_vcpu_time_info *vcpu,
> struct timespec *ts);
> +void pvclock_resume(void);
>
> #endif /* _ASM_X86_PVCLOCK_H */
> diff --git a/arch/x86/kernel/pvclock.c b/arch/x86/kernel/pvclock.c
> index 239427c..a4f07c1 100644
> --- a/arch/x86/kernel/pvclock.c
> +++ b/arch/x86/kernel/pvclock.c
> @@ -120,6 +120,11 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src)
>
> static atomic64_t last_value = ATOMIC64_INIT(0);
>
> +void pvclock_resume(void)
> +{
> + atomic64_set(&last_value, 0);
> +}
> +
> cycle_t pvclock_clocksource_read(struct pvclock_vcpu_time_info *src)
> {
> struct pvclock_shadow_time shadow;
> diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c
> index b2bb5aa..5da5e53 100644
> --- a/arch/x86/xen/time.c
> +++ b/arch/x86/xen/time.c
> @@ -426,6 +426,8 @@ void xen_timer_resume(void)
> {
> int cpu;
>
> + pvclock_resume();
> +
> if (xen_clockevent != &xen_vcpuop_clockevent)
> return;
>
>
>
>
>
> _______________________________________________
> Virtualization mailing list
> Virtualization@lists.linux-foundation.org
> https://lists.linux-foundation.org/mailman/listinfo/virtualization
^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: [PATCH] x86/pvclock-xen: zero last_value on resume
2010-10-26 16:59 ` Jeremy Fitzhardinge
` (2 preceding siblings ...)
(?)
@ 2010-11-22 20:16 ` Jeremy Fitzhardinge
-1 siblings, 0 replies; 22+ messages in thread
From: Jeremy Fitzhardinge @ 2010-11-22 20:16 UTC (permalink / raw)
To: H. Peter Anvin
Cc: Xen-devel, kvm-devel, Glauber Costa, the arch/x86 maintainers,
Linux Kernel Mailing List, Linux Virtualization, Olivier Hanesse,
Avi Kivity, Eelco Dolstra
On 10/26/2010 09:59 AM, Jeremy Fitzhardinge wrote:
>
> If the guest domain has been suspend/resumed or migrated, then the
> system clock backing the pvclock clocksource may revert to a smaller
> value (ie, can be non-monotonic across the migration/save-restore).
> Make sure we zero last_value in that case so that the domain
> continues to see clock updates.
Ping? Looks like this fell through the gaps.
J
> [ I don't know if kvm needs an analogous fix or not. ]
>
> Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
> Cc: Stable Kernel <stable@kernel.org>
> Reported-by: Eelco Dolstra <e.dolstra@tudelft.nl>
> Reported-by: Olivier Hanesse <olivier.hanesse@gmail.com>
> Bisected-by: Cédric Schieli <cschieli@gmail.com>
> Tested-by: Cédric Schieli <cschieli@gmail.com>
>
> diff --git a/arch/x86/include/asm/pvclock.h b/arch/x86/include/asm/pvclock.h
> index cd02f32..6226870 100644
> --- a/arch/x86/include/asm/pvclock.h
> +++ b/arch/x86/include/asm/pvclock.h
> @@ -11,5 +11,6 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src);
> void pvclock_read_wallclock(struct pvclock_wall_clock *wall,
> struct pvclock_vcpu_time_info *vcpu,
> struct timespec *ts);
> +void pvclock_resume(void);
>
> #endif /* _ASM_X86_PVCLOCK_H */
> diff --git a/arch/x86/kernel/pvclock.c b/arch/x86/kernel/pvclock.c
> index 239427c..a4f07c1 100644
> --- a/arch/x86/kernel/pvclock.c
> +++ b/arch/x86/kernel/pvclock.c
> @@ -120,6 +120,11 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src)
>
> static atomic64_t last_value = ATOMIC64_INIT(0);
>
> +void pvclock_resume(void)
> +{
> + atomic64_set(&last_value, 0);
> +}
> +
> cycle_t pvclock_clocksource_read(struct pvclock_vcpu_time_info *src)
> {
> struct pvclock_shadow_time shadow;
> diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c
> index b2bb5aa..5da5e53 100644
> --- a/arch/x86/xen/time.c
> +++ b/arch/x86/xen/time.c
> @@ -426,6 +426,8 @@ void xen_timer_resume(void)
> {
> int cpu;
>
> + pvclock_resume();
> +
> if (xen_clockevent != &xen_vcpuop_clockevent)
> return;
>
>
>
>
>
> _______________________________________________
> Virtualization mailing list
> Virtualization@lists.linux-foundation.org
> https://lists.linux-foundation.org/mailman/listinfo/virtualization
_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/virtualization
^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: [PATCH] x86/pvclock-xen: zero last_value on resume
2010-11-22 20:16 ` Jeremy Fitzhardinge
@ 2010-11-26 10:40 ` Ingo Molnar
2010-11-26 10:40 ` Ingo Molnar
1 sibling, 0 replies; 22+ messages in thread
From: Ingo Molnar @ 2010-11-26 10:40 UTC (permalink / raw)
To: Jeremy Fitzhardinge
Cc: H. Peter Anvin, Xen-devel, kvm-devel, Glauber Costa,
the arch/x86 maintainers, Linux Kernel Mailing List,
Linux Virtualization, Olivier Hanesse, Avi Kivity, Eelco Dolstra,
Thomas Gleixner
* Jeremy Fitzhardinge <jeremy@goop.org> wrote:
> On 10/26/2010 09:59 AM, Jeremy Fitzhardinge wrote:
> >
> > If the guest domain has been suspend/resumed or migrated, then the
> > system clock backing the pvclock clocksource may revert to a smaller
> > value (ie, can be non-monotonic across the migration/save-restore).
> > Make sure we zero last_value in that case so that the domain
> > continues to see clock updates.
>
> Ping? Looks like this fell through the gaps.
Does not apply cleanly here - mind resending the latest version?
Thanks,
Ingo
^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: [PATCH] x86/pvclock-xen: zero last_value on resume
@ 2010-11-26 10:40 ` Ingo Molnar
0 siblings, 0 replies; 22+ messages in thread
From: Ingo Molnar @ 2010-11-26 10:40 UTC (permalink / raw)
To: Jeremy Fitzhardinge
Cc: Xen-devel, kvm-devel, Glauber Costa, the arch/x86 maintainers,
Linux Kernel Mailing List, Linux Virtualization, Olivier Hanesse,
Avi Kivity, H. Peter Anvin, Thomas Gleixner, Eelco Dolstra
* Jeremy Fitzhardinge <jeremy@goop.org> wrote:
> On 10/26/2010 09:59 AM, Jeremy Fitzhardinge wrote:
> >
> > If the guest domain has been suspend/resumed or migrated, then the
> > system clock backing the pvclock clocksource may revert to a smaller
> > value (ie, can be non-monotonic across the migration/save-restore).
> > Make sure we zero last_value in that case so that the domain
> > continues to see clock updates.
>
> Ping? Looks like this fell through the gaps.
Does not apply cleanly here - mind resending the latest version?
Thanks,
Ingo
^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: [PATCH] x86/pvclock-xen: zero last_value on resume
2010-11-22 20:16 ` Jeremy Fitzhardinge
@ 2010-11-26 10:40 ` Ingo Molnar
2010-11-26 10:40 ` Ingo Molnar
1 sibling, 0 replies; 22+ messages in thread
From: Ingo Molnar @ 2010-11-26 10:40 UTC (permalink / raw)
To: Jeremy Fitzhardinge
Cc: Xen-devel, kvm-devel, Glauber Costa, the arch/x86 maintainers,
Linux Kernel Mailing List, Linux Virtualization, Olivier Hanesse,
Avi Kivity, H. Peter Anvin, Thomas Gleixner, Eelco Dolstra
* Jeremy Fitzhardinge <jeremy@goop.org> wrote:
> On 10/26/2010 09:59 AM, Jeremy Fitzhardinge wrote:
> >
> > If the guest domain has been suspend/resumed or migrated, then the
> > system clock backing the pvclock clocksource may revert to a smaller
> > value (ie, can be non-monotonic across the migration/save-restore).
> > Make sure we zero last_value in that case so that the domain
> > continues to see clock updates.
>
> Ping? Looks like this fell through the gaps.
Does not apply cleanly here - mind resending the latest version?
Thanks,
Ingo
^ permalink raw reply [flat|nested] 22+ messages in thread
* Re: [PATCH] x86/pvclock-xen: zero last_value on resume
2010-11-26 10:40 ` Ingo Molnar
@ 2010-11-26 22:29 ` Jeremy Fitzhardinge
-1 siblings, 0 replies; 22+ messages in thread
From: Jeremy Fitzhardinge @ 2010-11-26 22:29 UTC (permalink / raw)
To: Ingo Molnar
Cc: H. Peter Anvin, Xen-devel, kvm-devel, Glauber Costa,
the arch/x86 maintainers, Linux Kernel Mailing List,
Linux Virtualization, Olivier Hanesse, Avi Kivity, Eelco Dolstra,
Thomas Gleixner
On 11/26/2010 02:40 AM, Ingo Molnar wrote:
> * Jeremy Fitzhardinge <jeremy@goop.org> wrote:
>
>> On 10/26/2010 09:59 AM, Jeremy Fitzhardinge wrote:
>>> If the guest domain has been suspend/resumed or migrated, then the
>>> system clock backing the pvclock clocksource may revert to a smaller
>>> value (ie, can be non-monotonic across the migration/save-restore).
>>> Make sure we zero last_value in that case so that the domain
>>> continues to see clock updates.
>> Ping? Looks like this fell through the gaps.
> Does not apply cleanly here - mind resending the latest version?
It rebased cleanly to 2.6.37-rc3. You can pull it from:
git://git.kernel.org/pub/scm/linux/kernel/git/jeremy/xen.git pvclock-resume
Jeremy Fitzhardinge (1):
x86/pvclock: zero last_value on resume
arch/x86/include/asm/pvclock.h | 1 +
arch/x86/kernel/pvclock.c | 5 +++++
arch/x86/xen/time.c | 2 ++
3 files changed, 8 insertions(+), 0 deletions(-)
Thanks,
J
Subject: [PATCH] x86/pvclock: zero last_value on resume
If the guest domain has been suspend/resumed or migrated, then the
system clock backing the pvclock clocksource may revert to a smaller
value (ie, can be non-monotonic across the migration/save-restore).
Make sure we zero last_value in that case so that the domain
continues to see clock updates.
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
diff --git a/arch/x86/include/asm/pvclock.h b/arch/x86/include/asm/pvclock.h
index 7f7e577..31d84ac 100644
--- a/arch/x86/include/asm/pvclock.h
+++ b/arch/x86/include/asm/pvclock.h
@@ -11,6 +11,7 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src);
void pvclock_read_wallclock(struct pvclock_wall_clock *wall,
struct pvclock_vcpu_time_info *vcpu,
struct timespec *ts);
+void pvclock_resume(void);
/*
* Scale a 64-bit delta by scaling and multiplying by a 32-bit fraction,
diff --git a/arch/x86/kernel/pvclock.c b/arch/x86/kernel/pvclock.c
index 008b91e..42eb330 100644
--- a/arch/x86/kernel/pvclock.c
+++ b/arch/x86/kernel/pvclock.c
@@ -83,6 +83,11 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src)
static atomic64_t last_value = ATOMIC64_INIT(0);
+void pvclock_resume(void)
+{
+ atomic64_set(&last_value, 0);
+}
+
cycle_t pvclock_clocksource_read(struct pvclock_vcpu_time_info *src)
{
struct pvclock_shadow_time shadow;
diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c
index b2bb5aa..5da5e53 100644
--- a/arch/x86/xen/time.c
+++ b/arch/x86/xen/time.c
@@ -426,6 +426,8 @@ void xen_timer_resume(void)
{
int cpu;
+ pvclock_resume();
+
if (xen_clockevent != &xen_vcpuop_clockevent)
return;
^ permalink raw reply related [flat|nested] 22+ messages in thread
* Re: [PATCH] x86/pvclock-xen: zero last_value on resume
@ 2010-11-26 22:29 ` Jeremy Fitzhardinge
0 siblings, 0 replies; 22+ messages in thread
From: Jeremy Fitzhardinge @ 2010-11-26 22:29 UTC (permalink / raw)
To: Ingo Molnar
Cc: Xen-devel, kvm-devel, Glauber Costa, the arch/x86 maintainers,
Linux Kernel Mailing List, Linux Virtualization, Olivier Hanesse,
Avi Kivity, H. Peter Anvin, Thomas Gleixner, Eelco Dolstra
On 11/26/2010 02:40 AM, Ingo Molnar wrote:
> * Jeremy Fitzhardinge <jeremy@goop.org> wrote:
>
>> On 10/26/2010 09:59 AM, Jeremy Fitzhardinge wrote:
>>> If the guest domain has been suspend/resumed or migrated, then the
>>> system clock backing the pvclock clocksource may revert to a smaller
>>> value (ie, can be non-monotonic across the migration/save-restore).
>>> Make sure we zero last_value in that case so that the domain
>>> continues to see clock updates.
>> Ping? Looks like this fell through the gaps.
> Does not apply cleanly here - mind resending the latest version?
It rebased cleanly to 2.6.37-rc3. You can pull it from:
git://git.kernel.org/pub/scm/linux/kernel/git/jeremy/xen.git pvclock-resume
Jeremy Fitzhardinge (1):
x86/pvclock: zero last_value on resume
arch/x86/include/asm/pvclock.h | 1 +
arch/x86/kernel/pvclock.c | 5 +++++
arch/x86/xen/time.c | 2 ++
3 files changed, 8 insertions(+), 0 deletions(-)
Thanks,
J
Subject: [PATCH] x86/pvclock: zero last_value on resume
If the guest domain has been suspend/resumed or migrated, then the
system clock backing the pvclock clocksource may revert to a smaller
value (ie, can be non-monotonic across the migration/save-restore).
Make sure we zero last_value in that case so that the domain
continues to see clock updates.
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
diff --git a/arch/x86/include/asm/pvclock.h b/arch/x86/include/asm/pvclock.h
index 7f7e577..31d84ac 100644
--- a/arch/x86/include/asm/pvclock.h
+++ b/arch/x86/include/asm/pvclock.h
@@ -11,6 +11,7 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src);
void pvclock_read_wallclock(struct pvclock_wall_clock *wall,
struct pvclock_vcpu_time_info *vcpu,
struct timespec *ts);
+void pvclock_resume(void);
/*
* Scale a 64-bit delta by scaling and multiplying by a 32-bit fraction,
diff --git a/arch/x86/kernel/pvclock.c b/arch/x86/kernel/pvclock.c
index 008b91e..42eb330 100644
--- a/arch/x86/kernel/pvclock.c
+++ b/arch/x86/kernel/pvclock.c
@@ -83,6 +83,11 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src)
static atomic64_t last_value = ATOMIC64_INIT(0);
+void pvclock_resume(void)
+{
+ atomic64_set(&last_value, 0);
+}
+
cycle_t pvclock_clocksource_read(struct pvclock_vcpu_time_info *src)
{
struct pvclock_shadow_time shadow;
diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c
index b2bb5aa..5da5e53 100644
--- a/arch/x86/xen/time.c
+++ b/arch/x86/xen/time.c
@@ -426,6 +426,8 @@ void xen_timer_resume(void)
{
int cpu;
+ pvclock_resume();
+
if (xen_clockevent != &xen_vcpuop_clockevent)
return;
^ permalink raw reply related [flat|nested] 22+ messages in thread
* Re: [PATCH] x86/pvclock-xen: zero last_value on resume
2010-11-26 10:40 ` Ingo Molnar
(?)
@ 2010-11-26 22:29 ` Jeremy Fitzhardinge
-1 siblings, 0 replies; 22+ messages in thread
From: Jeremy Fitzhardinge @ 2010-11-26 22:29 UTC (permalink / raw)
To: Ingo Molnar
Cc: Xen-devel, kvm-devel, Glauber Costa, the arch/x86 maintainers,
Linux Kernel Mailing List, Linux Virtualization, Olivier Hanesse,
Avi Kivity, H. Peter Anvin, Thomas Gleixner, Eelco Dolstra
On 11/26/2010 02:40 AM, Ingo Molnar wrote:
> * Jeremy Fitzhardinge <jeremy@goop.org> wrote:
>
>> On 10/26/2010 09:59 AM, Jeremy Fitzhardinge wrote:
>>> If the guest domain has been suspend/resumed or migrated, then the
>>> system clock backing the pvclock clocksource may revert to a smaller
>>> value (ie, can be non-monotonic across the migration/save-restore).
>>> Make sure we zero last_value in that case so that the domain
>>> continues to see clock updates.
>> Ping? Looks like this fell through the gaps.
> Does not apply cleanly here - mind resending the latest version?
It rebased cleanly to 2.6.37-rc3. You can pull it from:
git://git.kernel.org/pub/scm/linux/kernel/git/jeremy/xen.git pvclock-resume
Jeremy Fitzhardinge (1):
x86/pvclock: zero last_value on resume
arch/x86/include/asm/pvclock.h | 1 +
arch/x86/kernel/pvclock.c | 5 +++++
arch/x86/xen/time.c | 2 ++
3 files changed, 8 insertions(+), 0 deletions(-)
Thanks,
J
Subject: [PATCH] x86/pvclock: zero last_value on resume
If the guest domain has been suspend/resumed or migrated, then the
system clock backing the pvclock clocksource may revert to a smaller
value (ie, can be non-monotonic across the migration/save-restore).
Make sure we zero last_value in that case so that the domain
continues to see clock updates.
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
diff --git a/arch/x86/include/asm/pvclock.h b/arch/x86/include/asm/pvclock.h
index 7f7e577..31d84ac 100644
--- a/arch/x86/include/asm/pvclock.h
+++ b/arch/x86/include/asm/pvclock.h
@@ -11,6 +11,7 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src);
void pvclock_read_wallclock(struct pvclock_wall_clock *wall,
struct pvclock_vcpu_time_info *vcpu,
struct timespec *ts);
+void pvclock_resume(void);
/*
* Scale a 64-bit delta by scaling and multiplying by a 32-bit fraction,
diff --git a/arch/x86/kernel/pvclock.c b/arch/x86/kernel/pvclock.c
index 008b91e..42eb330 100644
--- a/arch/x86/kernel/pvclock.c
+++ b/arch/x86/kernel/pvclock.c
@@ -83,6 +83,11 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src)
static atomic64_t last_value = ATOMIC64_INIT(0);
+void pvclock_resume(void)
+{
+ atomic64_set(&last_value, 0);
+}
+
cycle_t pvclock_clocksource_read(struct pvclock_vcpu_time_info *src)
{
struct pvclock_shadow_time shadow;
diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c
index b2bb5aa..5da5e53 100644
--- a/arch/x86/xen/time.c
+++ b/arch/x86/xen/time.c
@@ -426,6 +426,8 @@ void xen_timer_resume(void)
{
int cpu;
+ pvclock_resume();
+
if (xen_clockevent != &xen_vcpuop_clockevent)
return;
^ permalink raw reply related [flat|nested] 22+ messages in thread
* [PATCH] x86/pvclock-xen: zero last_value on resume
@ 2010-10-26 16:59 Jeremy Fitzhardinge
0 siblings, 0 replies; 22+ messages in thread
From: Jeremy Fitzhardinge @ 2010-10-26 16:59 UTC (permalink / raw)
To: H. Peter Anvin
Cc: Xen-devel, kvm-devel, Glauber Costa, the arch/x86 maintainers,
Linux Kernel Mailing List, Linux Virtualization, Olivier Hanesse,
Avi Kivity, Eelco Dolstra
[-- Attachment #1: Type: text/plain, Size: 1958 bytes --]
If the guest domain has been suspend/resumed or migrated, then the
system clock backing the pvclock clocksource may revert to a smaller
value (ie, can be non-monotonic across the migration/save-restore).
Make sure we zero last_value in that case so that the domain
continues to see clock updates.
[ I don't know if kvm needs an analogous fix or not. ]
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Cc: Stable Kernel <stable@kernel.org>
Reported-by: Eelco Dolstra <e.dolstra@tudelft.nl>
Reported-by: Olivier Hanesse <olivier.hanesse@gmail.com>
Bisected-by: Cédric Schieli <cschieli@gmail.com>
Tested-by: Cédric Schieli <cschieli@gmail.com>
diff --git a/arch/x86/include/asm/pvclock.h b/arch/x86/include/asm/pvclock.h
index cd02f32..6226870 100644
--- a/arch/x86/include/asm/pvclock.h
+++ b/arch/x86/include/asm/pvclock.h
@@ -11,5 +11,6 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src);
void pvclock_read_wallclock(struct pvclock_wall_clock *wall,
struct pvclock_vcpu_time_info *vcpu,
struct timespec *ts);
+void pvclock_resume(void);
#endif /* _ASM_X86_PVCLOCK_H */
diff --git a/arch/x86/kernel/pvclock.c b/arch/x86/kernel/pvclock.c
index 239427c..a4f07c1 100644
--- a/arch/x86/kernel/pvclock.c
+++ b/arch/x86/kernel/pvclock.c
@@ -120,6 +120,11 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src)
static atomic64_t last_value = ATOMIC64_INIT(0);
+void pvclock_resume(void)
+{
+ atomic64_set(&last_value, 0);
+}
+
cycle_t pvclock_clocksource_read(struct pvclock_vcpu_time_info *src)
{
struct pvclock_shadow_time shadow;
diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c
index b2bb5aa..5da5e53 100644
--- a/arch/x86/xen/time.c
+++ b/arch/x86/xen/time.c
@@ -426,6 +426,8 @@ void xen_timer_resume(void)
{
int cpu;
+ pvclock_resume();
+
if (xen_clockevent != &xen_vcpuop_clockevent)
return;
[-- Attachment #2: x86-pvclock-resume.patch --]
[-- Type: text/plain, Size: 2064 bytes --]
From 29acbb4e1d93e719250648db1ce8c7a24144fd86 Mon Sep 17 00:00:00 2001
From: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Date: Mon, 25 Oct 2010 16:53:46 -0700
Subject: [PATCH] x86/pvclock: zero last_value on resume
If the guest domain has been suspend/resumed or migrated, then the
system clock backing the pvclock clocksource may revert to a smaller
value (ie, can be non-monotonic across the migration/save-restore).
Make sure we zero last_value in that case so that the domain
continues to see clock updates.
Signed-off-by: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
Cc: Stable Kernel <stable@kernel.org>
Reported-by: Eelco Dolstra <e.dolstra@tudelft.nl>
Reported-by: Olivier Hanesse <olivier.hanesse@gmail.com>
Bisected-by: Cédric Schieli <cschieli@gmail.com>
Tested-by: Cédric Schieli <cschieli@gmail.com>
diff --git a/arch/x86/include/asm/pvclock.h b/arch/x86/include/asm/pvclock.h
index cd02f32..6226870 100644
--- a/arch/x86/include/asm/pvclock.h
+++ b/arch/x86/include/asm/pvclock.h
@@ -11,5 +11,6 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src);
void pvclock_read_wallclock(struct pvclock_wall_clock *wall,
struct pvclock_vcpu_time_info *vcpu,
struct timespec *ts);
+void pvclock_resume(void);
#endif /* _ASM_X86_PVCLOCK_H */
diff --git a/arch/x86/kernel/pvclock.c b/arch/x86/kernel/pvclock.c
index 239427c..a4f07c1 100644
--- a/arch/x86/kernel/pvclock.c
+++ b/arch/x86/kernel/pvclock.c
@@ -120,6 +120,11 @@ unsigned long pvclock_tsc_khz(struct pvclock_vcpu_time_info *src)
static atomic64_t last_value = ATOMIC64_INIT(0);
+void pvclock_resume(void)
+{
+ atomic64_set(&last_value, 0);
+}
+
cycle_t pvclock_clocksource_read(struct pvclock_vcpu_time_info *src)
{
struct pvclock_shadow_time shadow;
diff --git a/arch/x86/xen/time.c b/arch/x86/xen/time.c
index b2bb5aa..5da5e53 100644
--- a/arch/x86/xen/time.c
+++ b/arch/x86/xen/time.c
@@ -426,6 +426,8 @@ void xen_timer_resume(void)
{
int cpu;
+ pvclock_resume();
+
if (xen_clockevent != &xen_vcpuop_clockevent)
return;
[-- Attachment #3: Type: text/plain, Size: 184 bytes --]
_______________________________________________
Virtualization mailing list
Virtualization@lists.linux-foundation.org
https://lists.linux-foundation.org/mailman/listinfo/virtualization
^ permalink raw reply related [flat|nested] 22+ messages in thread
end of thread, other threads:[~2010-11-26 22:29 UTC | newest]
Thread overview: 22+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2010-10-26 16:59 [PATCH] x86/pvclock-xen: zero last_value on resume Jeremy Fitzhardinge
2010-10-26 16:59 ` Jeremy Fitzhardinge
2010-10-26 17:48 ` Glauber Costa
2010-10-26 17:48 ` Glauber Costa
2010-10-27 18:53 ` [Xen-devel] " Jeremy Fitzhardinge
2010-10-27 18:53 ` Jeremy Fitzhardinge
2010-10-27 20:59 ` H. Peter Anvin
2010-10-27 20:59 ` H. Peter Anvin
2010-11-03 12:46 ` Ian Campbell
2010-11-03 12:46 ` Ian Campbell
2010-11-03 12:46 ` Ian Campbell
2010-11-03 12:46 ` Ian Campbell
2010-10-26 17:48 ` Glauber Costa
2010-11-22 20:16 ` Jeremy Fitzhardinge
2010-11-22 20:16 ` Jeremy Fitzhardinge
2010-11-26 10:40 ` Ingo Molnar
2010-11-26 10:40 ` Ingo Molnar
2010-11-26 10:40 ` Ingo Molnar
2010-11-26 22:29 ` Jeremy Fitzhardinge
2010-11-26 22:29 ` Jeremy Fitzhardinge
2010-11-26 22:29 ` Jeremy Fitzhardinge
-- strict thread matches above, loose matches on Subject: below --
2010-10-26 16:59 Jeremy Fitzhardinge
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.