>>> On 23.04.15 at 18:29, wrote:
> On 23/04/15 17:11, Jan Beulich wrote:
> On 22.04.15 at 18:00, wrote:
>>> -return handle;
>>> +v->maptrack_limit = new_mt_limit;
>>> +
>>> +return __get_maptrack_handle(lgt, v);
>>
>> With the lock dropped, nothing guarantees this to succeed, whi
On 04/17/2015 03:37 PM, Jan Beulich wrote:
On 17.04.15 at 09:23, wrote:
On 04/17/2015 02:58 PM, Jan Beulich wrote:
On 17.04.15 at 08:51, wrote:
On 04/17/2015 02:23 PM, Jan Beulich wrote:
On 17.04.15 at 05:10, wrote:
On 04/16/2015 11:42 PM, Jan Beulich wrote:
On 15.04.15 at 09:03, wrot
On Thu, 2015-04-23 at 12:05 +0100, Ian Jackson wrote:
> Pang, LongtaoX writes ("RE: [Xen-devel] [OSSTEST Nested PATCH v8 6/7] Compose
> the main recipe of nested test job"):
> > Ian Campbell [mailto:ian.campb...@citrix.com]:
> > > The arguments passed to ts-* in sg-run-job should _always_ be just
On 23/04/2015 22:09, Jan Beulich wrote:
> >>> On 23.04.16 at 15:31, wrote:
> > The intel_pstate.c file under xen/arch/x86/acpi/cpufreq/ contains all
> > the logic for selecting the current P-state. It follows its
> > implementation in the kernel. Instead of using the traditional cpufreq
> > govern
On 2015/4/23 20:59, Jan Beulich wrote:
On 23.04.15 at 14:32, wrote:
But if you already have one just please ignore this and tell me
Here's what I currently have:
Could you resend me this as an attached file? Then I can apply that
properly without any miss?
Thanks
Tiejun
introduce XENM
On 04/21/2015 03:01 AM, Jan Beulich wrote:
+ ((++dev_cnt > 0x3f) && hypercall_preempt_check()) )
+break;
+}
+
+if ( (!ret || (ret == -ENODEV)) &&
+ __copy_field_to_guest(u_sysctl, op, u.pcitopoinfo.first_dev) )
+ret = -EFAU
On 4/23/15, 04:25, "Tim Deegan" wrote:
>At 17:44 +0100 on 16 Apr (1429206250), Tim Deegan wrote:
>> update_intremap_entry_from_msi() doesn't write to its data pointer on
>> some error paths, so we copying that variable into the msg would count
>> as undefined behaviour.
>>
>> Signed-off-by: Tim
On 4/23/15, 13:06, "Konrad Rzeszutek Wilk" wrote:
>>I have tested this patch w/ staging branch booting Dom0, and this patch
>>got rid of the following error from xl dmesg:
>>(XEN) APIC error on CPU0: 00(40)
>>(XEN) APIC error on CPU2: 00(40)
>>However, when I tried starting a guest w/ PCI devic
On 4/23/15, 12:59, "Sander Eikelenboom" wrote:
>
>> On 4/23/15, 08:47, "Jan Beulich" wrote:
>
>> On 23.04.15 at 15:31, wrote:
>>>
On 4/17/15, 10:27, "Jan Beulich" wrote:
>1aeb1156fa ("x86 don't change affinity with interrupt unmasked")
>introducing RTE reads prio
Hi Dario,
2015-04-23 8:48 GMT-04:00 Dario Faggioli :
> Hey, guys,
>
> I know, I know, I'm s much late to the party! Sorry, I got trapped
> into a thing that I really needed to finish... :-/
>
> I've got no intention to resurrect this old thread, just wanted to
> pointed out a few things.
What
On Thu, Apr 23, 2015 at 05:51:05PM +, Suthikulpanit, Suravee wrote:
>
>
> On 4/23/15, 08:47, "Jan Beulich" wrote:
>
> On 23.04.15 at 15:31, wrote:
> >
> >>
> >> On 4/17/15, 10:27, "Jan Beulich" wrote:
> >>
> >>>1aeb1156fa ("x86 don't change affinity with interrupt unmasked")
> >>>i
> On 4/23/15, 08:47, "Jan Beulich" wrote:
> On 23.04.15 at 15:31, wrote:
>>
>>>
>>> On 4/17/15, 10:27, "Jan Beulich" wrote:
>>>
1aeb1156fa ("x86 don't change affinity with interrupt unmasked")
introducing RTE reads prior to the respective interrupt having got
enabled for the
On 4/23/15, 08:47, "Jan Beulich" wrote:
On 23.04.15 at 15:31, wrote:
>
>>
>> On 4/17/15, 10:27, "Jan Beulich" wrote:
>>
>>>1aeb1156fa ("x86 don't change affinity with interrupt unmasked")
>>>introducing RTE reads prior to the respective interrupt having got
>>>enabled for the first tim
On 23/04/15 16:46, Paul Durrant wrote:
> Currently, unless a (non-default) ioreq server is explicitly disabled before
> being destroyed, its gmfns will not be placed back into the p2m but still
> released back into the ioreq_gmfn mask. This is somewhat counter-intuitive
> and easily remedied by thi
On Tue, 2015-03-31 at 15:37 +0100, George Dunlap wrote:
> On 03/26/2015 09:48 AM, Justin T. Weaver wrote:
> > --- a/xen/common/sched_credit2.c
> > +++ b/xen/common/sched_credit2.c
> > @@ -194,6 +195,12 @@ int opt_overload_balance_tolerance=-3;
> > integer_param("credit2_balance_over", opt_overlo
On 23/04/15 16:46, Paul Durrant wrote:
> hvm_free_ioreq_gmfn has the sense of the ioreq_gmfn mask inverted; it
> needs to set a bit to release the gmfn, not clear it.
>
> Signed-off-by: Paul Durrant
> Cc: Keir Fraser
> Cc: Jan Beulich
> Cc: Andrew Cooper
s/mask/free_mask/ ?
For the change its
Currently, unless a (non-default) ioreq server is explicitly disabled before
being destroyed, its gmfns will not be placed back into the p2m but still
released back into the ioreq_gmfn mask. This is somewhat counter-intuitive
and easily remedied by this small patch.
Signed-off-by: Paul Durrant
Cc
The following two patches fix a bug in the ioreq server gmfn release
code and also make the semantics of server destruction more intuitive.
___
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel
hvm_free_ioreq_gmfn has the sense of the ioreq_gmfn mask inverted; it
needs to set a bit to release the gmfn, not clear it.
Signed-off-by: Paul Durrant
Cc: Keir Fraser
Cc: Jan Beulich
Cc: Andrew Cooper
---
xen/arch/x86/hvm/hvm.c |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff -
On Wed, 2015-03-25 at 23:48 -1000, Justin T. Weaver wrote:
> Functions runq_tickle and choose_cpu both have code sections that get turned
> into loops in patch 4 v3, soft affinity. Do the indenting here to make the
> patch 4 diff section easier to read.
>
Yeah, I know what you mean, an thanks for
On Wed, 2015-03-25 at 23:48 -1000, Justin T. Weaver wrote:
> Move affinity balancing related functions and defines from sched_credit.c to
> sched-if.h so other schedulers can use them. Change name prefixes from csched
> to sched since they are no longer specific to the credit scheduler.
>
It's pro
Due to the Hackathon next week, our normal April Document Day is
taking a one week bump to May 6. If anyone objects to this date,
please let me know. Otherwise, we'll allow those attending the
Hackathon to participate in Document Day upon their return.
All the information you need to participate
On 23/04/15 17:11, Jan Beulich wrote:
On 22.04.15 at 18:00, wrote:
>>
>> +v->maptrack_head = lgt->maptrack_pages * MAPTRACK_PER_PAGE;
>> +if (v->maptrack_tail == MAPTRACK_TAIL)
>> +{
>> +v->maptrack_tail = (lgt->maptrack_pages * MAPTRACK_PER_PAGE)
>> ++ MAPTR
>>> On 23.04.15 at 17:54, wrote:
> Right now, `xen-hptool mem-offline ` does not take the ,
> to which the belongs. I think this tool should take the
> as an extra parameter just as `xen-mfndump lookup-pte `
> does. When the tool construct the foreigndom for mmu, it should make
> the foreigndom
>>> On 22.04.15 at 18:00, wrote:
> static inline int
> get_maptrack_handle(
> struct grant_table *lgt)
> {
> +struct vcpu *v = current;
> int i;
> grant_handle_thandle;
> struct grant_mapping *new_mt;
> unsigned int new_mt
Hi Jan,
2015-04-23 4:14 GMT-04:00 Jan Beulich :
>
> >>> On 23.04.15 at 06:03, wrote:
> > The first question is:
> > Why should we use "foreigndom >> 16" instead of "foreigndom" to get the
> > pt_dom?
>
> Because there are possibly three domains involved here: The current
> one (issuing the hyperc
>>> On 22.04.15 at 18:00, wrote:
> This is safe because: a) the grant table version only changes once
> from 0 to 1 or 2;
gnttab_set_version() also allows it to transition between 1 and 2
afaics.
> @@ -919,18 +914,15 @@ __gnttab_unmap_common(
> }
>
> op->map = &maptrack_entry(lgt, op
>>> On 22.04.15 at 18:00, wrote:
> Subsequent changes make both these locks uncontented. Is this patch
> really necessary? -- dvrabel
Yeah, particularly if this ...
> @@ -540,6 +550,16 @@ static void mapcount(
>
> *wrc = *rdc = 0;
>
> +/*
> + * While taking the local maptrack
>>> On 23.04.15 at 16:43, wrote:
> At 14:54 +0100 on 23 Apr (1429800874), Jan Beulich wrote:
>> >>> On 23.04.15 at 14:03, wrote:
>> > At 11:11 +0100 on 21 Apr (1429614687), David Vrabel wrote:
>> >> void _spin_unlock(spinlock_t *lock)
>> >> {
>> >> +smp_mb();
>> >> preempt_enable();
>>
At 14:45 +0100 on 23 Apr (1429800338), Andrew Cooper wrote:
> On 23/04/15 14:43, David Vrabel wrote:
> > On 23/04/15 13:03, Tim Deegan wrote:
> >> Hi,
> >>
> >> At 11:11 +0100 on 21 Apr (1429614687), David Vrabel wrote:
> >>> void _spin_lock(spinlock_t *lock)
> >>> {
> >>> +spinlock_tickets_t
>>> On 23.04.15 at 16:13, wrote:
> On Thu, 23 Apr 2015, Jan Beulich wrote:
>> while I'm not really convinced of the change, commit 14655e9a18
>> ("glib-compat: fix problems with not-quite glib 2.22") at least helps
>> with building on SLE11 SP3. My concern is that the fallback seems
>> worse than
At 15:24 +0100 on 23 Apr (1429802696), Tim Deegan wrote:
> At 14:43 +0100 on 23 Apr (1429800229), David Vrabel wrote:
> > On 23/04/15 13:03, Tim Deegan wrote:
> > > Hi,
> > >
> > > At 11:11 +0100 on 21 Apr (1429614687), David Vrabel wrote:
> > >> void _spin_lock(spinlock_t *lock)
> > >> {
> > >>
On Wed, Apr 22, 2015 at 10:49:18PM -0400, Boris Ostrovsky wrote:
> When resuming, the guest needs to check whether the port has changed. HVM
> guests use this parameter to get the port number.
>
> (We can't always use xenstore where this value is also written: for example
> on Linux the console is
At 14:54 +0100 on 23 Apr (1429800874), Jan Beulich wrote:
> >>> On 23.04.15 at 14:03, wrote:
> > At 11:11 +0100 on 21 Apr (1429614687), David Vrabel wrote:
> >> void _spin_unlock(spinlock_t *lock)
> >> {
> >> +smp_mb();
> >> preempt_enable();
> >> LOCK_PROFILE_REL;
> >> -_raw_s
On Thu, 23 Apr 2015, Jan Beulich wrote:
> Stefano,
>
> while I'm not really convinced of the change, commit 14655e9a18
> ("glib-compat: fix problems with not-quite glib 2.22") at least helps
> with building on SLE11 SP3. My concern is that the fallback seems
> worse than using the backported inter
At 14:59 +0100 on 23 Apr (1429801184), Jan Beulich wrote:
> >>> On 23.04.15 at 14:32, wrote:
> > On 2015/4/16 23:40, Tim Deegan wrote:
> >> At 17:21 +0800 on 10 Apr (1428686518), Tiejun Chen wrote:
> >>> @@ -1851,7 +1857,14 @@ static int rmrr_identity_mapping(struct domain *d,
> >>> bool_t map,
>
At 14:43 +0100 on 23 Apr (1429800229), David Vrabel wrote:
> On 23/04/15 13:03, Tim Deegan wrote:
> > Hi,
> >
> > At 11:11 +0100 on 21 Apr (1429614687), David Vrabel wrote:
> >> void _spin_lock(spinlock_t *lock)
> >> {
> >> +spinlock_tickets_t tickets = { .tail = 1, };
> >> LOCK_PROFILE
>>> On 23.04.16 at 15:31, wrote:
> The intel_pstate.c file under xen/arch/x86/acpi/cpufreq/ contains all
> the logic for selecting the current P-state. It follows its
> implementation in the kernel. Instead of using the traditional cpufreq
> governors, intel_pstate implements its internal govern
On Thu, Apr 23, 2015 at 10:24:04AM +0100, Andrew Cooper wrote:
> On 23/04/15 03:49, Boris Ostrovsky wrote:
> > When resuming, the guest needs to check whether the port has changed. HVM
> > guests use this parameter to get the port number.
> >
> > (We can't always use xenstore where this value is al
>>> On 23.04.15 at 14:32, wrote:
> On 2015/4/16 23:40, Tim Deegan wrote:
>> At 17:21 +0800 on 10 Apr (1428686518), Tiejun Chen wrote:
>>> @@ -1851,7 +1857,14 @@ static int rmrr_identity_mapping(struct domain *d,
>>> bool_t map,
>>> if ( !is_hardware_domain(d) )
>>> {
>>>
The intel_pstate driver receives percentage values to set the
performance limits. This patch adds interfaces to support the
input of percentage values to control the intel_pstate driver.
Also, the "get-cpufreq-para" is modified to show percentage
based feedback info.
Signed-off-by: Wei Wang
---
The unregister notifier function is needed to support cpu hotplug.
Signed-off-by: Wei Wang
---
xen/common/cpu.c | 7 +++
xen/include/xen/cpu.h | 1 +
2 files changed, 8 insertions(+)
diff --git a/xen/common/cpu.c b/xen/common/cpu.c
index 47e8b5b..508cee5 100644
--- a/xen/common/cpu.c
+
By default, the intel_pstate driver is loaded.a If
"intel_pstate=disable" is added to the Xen booting param list,
the old pstate driver will be loaded.
Signed-off-by: Wei Wang
---
xen/arch/x86/acpi/cpufreq/cpufreq.c | 9 ++---
xen/arch/x86/acpi/cpufreq/intel_pstate.c | 16 +++
Add support in the pmstat.c so that the xenpm tool can request to
access the intel_pstate driver.
Signed-off-by: Wei Wang
---
tools/libxc/xc_pm.c | 8 +++
xen/drivers/acpi/pmstat.c | 57 +++--
xen/include/public/sysctl.h | 22 ++-
On 23/04/15 14:43, David Vrabel wrote:
> On 23/04/15 13:03, Tim Deegan wrote:
>> Hi,
>>
>> At 11:11 +0100 on 21 Apr (1429614687), David Vrabel wrote:
>>> void _spin_lock(spinlock_t *lock)
>>> {
>>> +spinlock_tickets_t tickets = { .tail = 1, };
>>> LOCK_PROFILE_VAR;
>>>
>>> check_l
The intel_pstate driver is ported following its kernel code logic
(commit: 93f0822d).
In the kernel, a user can adjust the limits via sysfs
(limits.min_sysfs_pct/max_sysfs_pct). In Xen, the
policy->min_perf_pct/max_perf_pct acts as the transit station.
A user interacts with it via xenpm.
Signed-o
In order to better support future Intel processors, intel_pstate
changes to use percentage values to tune P-states. The intel_pstate
driver uses its own internal governor, and it is recorded in the
"policy->policy" field. The setpolicy driver interface is used to
configure the intel_pstate internal
The added calculation related functions will be used in the intel_pstate.c.
Signed-off-by: Wei Wang
---
xen/include/asm-x86/div64.h | 68 +
xen/include/xen/kernel.h| 30
2 files changed, 98 insertions(+)
diff --git a/xen/inclu
Register/unregister the CPU hotplug notifier when the driver is registered, and
move the driver register/unregister function to the cpufreq.c.
Signed-off-by: Wei Wang
---
xen/drivers/cpufreq/cpufreq.c | 27 +++
xen/include/acpi/cpufreq/cpufreq.h | 28 ++-
Re-organize the cpu matching code in the mwait-idle.c, so that it can be
re-used in the intel_pstate.c.
Signed-off-by: Wei Wang
---
xen/arch/x86/cpu/common.c | 39 +++
xen/arch/x86/cpu/mwait-idle.c | 30 +-
xen/include/asm-
>>> On 23.04.15 at 14:03, wrote:
> At 11:11 +0100 on 21 Apr (1429614687), David Vrabel wrote:
>> void _spin_unlock(spinlock_t *lock)
>> {
>> +smp_mb();
>> preempt_enable();
>> LOCK_PROFILE_REL;
>> -_raw_spin_unlock(&lock->raw);
>> +lock->tickets.head++;
>
> This needs to b
On 04/23/2015 02:31 PM, Konrad Rzeszutek Wilk wrote:
> On Wed, Apr 15, 2015 at 01:26:45PM +0100, Ian Campbell wrote:
>> On Fri, 2015-04-03 at 20:43 -0400, Konrad Rzeszutek Wilk wrote:
>>> Instead of just using -c 0x we can
>>> also use: -c -, -c ,, or a
>>> combination of them, or 'all' for all cpu
On 23/04/15 13:42, Jan Beulich wrote:
>
>> @@ -3130,7 +3167,9 @@ grant_table_destroy(
>>
>> if ( t == NULL )
>> return;
>> -
>> +
>> +spin_lock(&t->lock);
>> +
>> for ( i = 0; i < nr_grant_frames(t); i++ )
>> free_xenheap_page(t->shared_raw[i]);
>> xfree(t
>>> On 23.04.15 at 15:31, wrote:
>
> On 4/17/15, 10:27, "Jan Beulich" wrote:
>
>>1aeb1156fa ("x86 don't change affinity with interrupt unmasked")
>>introducing RTE reads prior to the respective interrupt having got
>>enabled for the first time uncovered a bug in 2ca9fbd739 ("AMD IOMMU:
>>alloc
On 23/04/15 13:03, Tim Deegan wrote:
> Hi,
>
> At 11:11 +0100 on 21 Apr (1429614687), David Vrabel wrote:
>> void _spin_lock(spinlock_t *lock)
>> {
>> +spinlock_tickets_t tickets = { .tail = 1, };
>> LOCK_PROFILE_VAR;
>>
>> check_lock(&lock->debug);
>> -while ( unlikely(!_raw
Stefano,
while I'm not really convinced of the change, commit 14655e9a18
("glib-compat: fix problems with not-quite glib 2.22") at least helps
with building on SLE11 SP3. My concern is that the fallback seems
worse than using the backported interface in glib, but I also can't
see a reasonable way
Hi,
This patch series ports the intel_pstate driver from the Linux kernel to
Xen. The intel_pstate driver is used to tune P states for SandyBridge+
processors. It can be disabled by adding "intel_pstate=disable" to the
booting parameter list.
The intel_pstate.c file under xen/arch/x86/acpi/cpuf
On 4/17/15, 10:27, "Jan Beulich" wrote:
>1aeb1156fa ("x86 don't change affinity with interrupt unmasked")
>introducing RTE reads prior to the respective interrupt having got
>enabled for the first time uncovered a bug in 2ca9fbd739 ("AMD IOMMU:
>allocate IRTE entries instead of using a static m
On Wed, Apr 15, 2015 at 01:26:45PM +0100, Ian Campbell wrote:
> On Fri, 2015-04-03 at 20:43 -0400, Konrad Rzeszutek Wilk wrote:
> > Instead of just using -c 0x we can
> > also use: -c -, -c ,, or a
> > combination of them, or 'all' for all cpus.
> >
> > This new format can include just singular CP
Instead of just using -c 0x we can
also use: -c -, -c ,, or a
combination of them, or 'all' for all cpus.
This new format can include just singular CPUs: -c ,
or ranges without an start or end (and xentrace will figure out
the values), such as: -c - (which will include cpu0, cpu1,
and cpu2) or -c
Apologies for breaking the threading. The mail server ate the thread and I
couldn't manage to get it back...
>[snip]
>> OK, here it is as a proper patch. I still haven't tested it -- indeed
>> I'm not sure how to test multiple ioreq clients. Any suggestions?
>
>Paul?
>
>Thanks, Jan
Testing mul
From: George Dunlap
xentrace is the only caller at the moment. Split the cpu and event
mask setting out into seperate functions, but leave the current limit
of 32 bits for masks passed in from the command-line.
Based on a patch from Konrad Rzeszutek Wilk
Acked-by: Ian Campbell
Signed-off-by:
We export the xc_cpumap_alloc but not the bit operations.
One could include 'xc_bitops.h' but that is naughty - so instead
we just export the proper functions to do it on the xc_cpumap_t
typedef.
Signed-off-by: Konrad Rzeszutek Wilk
v2: Use our own macro to make sure ARM is not affected nega
Hey George, Ian,
Ian: I copied your comment and stuffed it in, please Ack:
[PATCH v4.1 1/3] libxl/cpumap: Add xc_cpumap_[setcpu, clearcpu,
already Acked:
[PATCH v4.1 2/3] libxc/xentrace: Use xc_cpumap_t for
George: please review (it is unchanged since v4 posting):
[PATCH v4.1 3/3] xentrace:
On Wed, Apr 22, 2015 at 10:08:52AM +0200, Olaf Hering wrote:
> On Tue, Apr 21, Konrad Rzeszutek Wilk wrote:
>
> > On Fri, Apr 17, 2015 at 08:30:58AM +, Olaf Hering wrote:
> > > +=item B
> > > +
> > > +Specifies the PVSCSI devices to be provided to the guest. PVSCSI passes
> > > +dom0 SCSI devi
On Thu, 2015-04-23 at 12:30 +0100, Ian Campbell wrote:
> This is the same problem I mentioned above arising from $l1 doing double
> duty as both host and guest. There are places in osstest which are
> pretty fluid about this (if the object has the approrpaite props you can
> use it as either), but
At 20:32 +0800 on 23 Apr (1429821151), Chen, Tiejun wrote:
> >> +if ( flag == XEN_DOMCTL_PCIDEV_RDM_TRY )
> >> +{
> >> +printk(XENLOG_G_WARNING "Some devices may work failed
> >> .\n");
> >
> > This is a bit cryptic. How about:
> > "RMRR map f
>>> On 23.04.15 at 14:32, wrote:
> But if you already have one just please ignore this and tell me
Here's what I currently have:
introduce XENMEM_reserved_device_memory_map
This is a prerequisite for punching holes into HVM and PVH guests' P2M
to allow passing through devices that are associate
Hey, guys,
I know, I know, I'm s much late to the party! Sorry, I got trapped
into a thing that I really needed to finish... :-/
I've got no intention to resurrect this old thread, just wanted to
pointed out a few things.
On Wed, 2015-04-08 at 16:52 -0400, Meng Xu wrote:
> 2015-04-08 5:13 GM
On 21/04/15 11:22, Jan Beulich wrote:
On 21.04.15 at 12:11, wrote:
>> We have analysed the affect of this series on interrupt latency (by
>> measuring the duration of irq disable/enable regions) and there is no
>> signficant impact.
>>
>> http://xenbits.xen.org/people/dvrabel/bar2_comp.png
>
>>> On 22.04.15 at 18:00, wrote:
> @@ -548,11 +542,14 @@ static void mapcount(
>
> for ( handle = 0; handle < lgt->maptrack_limit; handle++ )
> {
> +struct active_grant_entry *act;
const please (wherever possible, to document the intent to not
modify).
> -if ( active_
From: David Vrabel
Add the specification for a new migration stream format. The document
includes all the details but to summarize:
The existing (legacy) format is dependant on the word size of the
toolstack. This prevents domains from migrating from hosts running
32-bit toolstacks to hosts ru
Restore the x86 HVM specific parts of a domain. This is the HVM_CONTEXT and
HVM_PARAMS records.
There is no need for any page localisation.
This also includes writing the trailing qemu save record to a file because
this is what libxc currently does. This is intended to be moved into libxl
prope
Save a domain, calling domain type specific function at the appropriate
points. This implements the xc_domain_save2() API function which is
equivalent to the existing xc_domain_save().
This writes the image and domain headers, and writes all the PAGE_DATA records
using a "live" process.
Signed-o
These changes cause migration v2 to behave similarly enough to legacy
migration to function for HVM guests under an unmodified xl/libxl.
The migration v2 work for libxl will fix the layering issues with the
toolstack and qemu records, at which point this patch will be unneeded.
It is however incl
Save the x86 HVM specific parts of the domain. This is considerably simpler
than an x86 PV domain. Only the HVM_CONTEXT and HVM_PARAMS records are
needed.
There is no need for any page normalisation.
Signed-off-by: Andrew Cooper
Acked-by: Ian Campbell
CC: Ian Jackson
CC: Wei Liu
---
tools/
On 2015/4/16 23:05, Tim Deegan wrote:
Hi,
At 17:21 +0800 on 10 Apr (1428686516), Tiejun Chen wrote:
@@ -862,6 +862,36 @@ int set_mmio_p2m_entry(struct domain *d, unsigned long
gfn, mfn_t mfn,
return set_typed_p2m_entry(d, gfn, mfn, p2m_mmio_direct, access);
}
+int set_identity_p2m_ent
On 2015/4/16 23:40, Tim Deegan wrote:
Hi,
At 17:21 +0800 on 10 Apr (1428686518), Tiejun Chen wrote:
+/*
+ * In some cases, e.g. add a device to hwdomain, and remove a device from
+ * user domain, 'try' is fine enough since this is always safe to hwdomain.
+ */
+#define XEN_DOMCTL_PCIDEV_RDM_DEF
>>> @@ -121,6 +121,8 @@ void iommu_dt_domain_destroy(struct domain *d);
struct page_info;
+typedef int iommu_grdm_t(xen_pfn_t start, xen_ulong_t nr, u32 id, void *ctxt);
This needs a comment describing what the return values are.
Will do.
I'm not sure if yourself determine providing
On 2015/4/15 21:10, Ian Jackson wrote:
Tiejun Chen writes ("[RFC][PATCH 04/13] tools/libxl: detect and avoid conflicts with
RDM"):
While building a VM, HVM domain builder provides struct hvm_info_table{}
to help hvmloader. Currently it includes two fields to construct guest
e820 table by hvmloa
hwdom_shutdown() operates with hardware domains, use the proper wording.
Eliminate pointless braces from switch cases.
Use hardware_domain->domain_id instead of hardware_domid to print the actual
domain ID as in some cases it can differ (e.g. Dom0 dies before the actual HW
domain got created, kexe
Restore a domain from the new format. This reads and validates the domain and
image header and loads the guest memory from the PAGE_DATA records, populating
the p2m as it does so.
This provides the xc_domain_restore2() function as an alternative to the
existing xc_domain_restore().
Signed-off-by
On Thu, 2015-04-23 at 10:43 +, Hu, Robert wrote:
> > If/when the other ts-*-install need to support sizing the disk based on
> > a runvar they would need similar changes. If you want to make those now
> > for consistency then that would be great, but I don't think it needs to
> > be mandatory.
On 04/09/2015 02:12 AM, Wei Liu wrote:
> On Wed, Apr 01, 2015 at 02:41:43PM +0800, Yang Hongyang wrote:
>> From: Wen Congyang
>>
>> For PVHVM, the hypercall return code is 0, and it can be resumed
>> in a new domain context.
>>
>> For HVM, do nothing.
>>
>> Signed-off-by: Wen Congyang
>> ---
>>
On 04/22/2015 10:54 PM, Ian Campbell wrote:
> On Wed, 2015-04-01 at 14:41 +0800, Yang Hongyang wrote:
>> From: Wen Congyang
>>
>> For PVHVM, the hypercall return code is 0, and it can be resumed
>> in a new domain context.
>>
>> For HVM, do nothing.
>>
>> Signed-off-by: Wen Congyang
>> ---
>> to
Hi,
At 11:11 +0100 on 21 Apr (1429614687), David Vrabel wrote:
> void _spin_lock(spinlock_t *lock)
> {
> +spinlock_tickets_t tickets = { .tail = 1, };
> LOCK_PROFILE_VAR;
>
> check_lock(&lock->debug);
> -while ( unlikely(!_raw_spin_trylock(&lock->raw)) )
> +tickets.head_t
>>> On 21.04.15 at 12:11, wrote:
> @@ -213,27 +211,32 @@ int _spin_trylock(spinlock_t *lock)
>
> void _spin_barrier(spinlock_t *lock)
> {
> +spinlock_tickets_t sample;
> #ifdef LOCK_PROFILE
> s_time_t block = NOW();
> -u64 loop = 0;
> +#endif
>
> check_barrier(&lock->
On 04/22/2015 11:10 PM, Ian Campbell wrote:
> On Wed, 2015-04-01 at 14:41 +0800, Yang Hongyang wrote:
>> From: Wen Congyang
>>
>> This is the same set used by libxc.
>
> What is this for?
>
> libxl already exposes a fairly complete libxl_bitmap type and helpers
> for use in its own interfaces an
Restore the x86 PV specific parts. The X86_PV_INFO, the P2M_FRAMES,
SHARED_INFO, and VCPU context records.
The localise_page callback is called from the common PAGE_DATA code to convert
PFNs in page tables to MFNs.
Page tables are pinned and the guest's P2M is updated when the stream is
complete
Provide the C structures matching the binary (wire) format of the new
stream format. All header/record fields are naturally aligned and
explicit padding fields are used to ensure the correct layout (i.e.,
there is no need for any non-standard structure packing pragma or
attribute).
Provide some h
Add the context structure used to keep state during the save/restore
process.
Define the set of architecture or domain type specific operations with a
set of callbacks (save_ops, and restore_ops).
Add common functions for writing records.
Signed-off-by: Andrew Cooper
Acked-by: Ian Campbell
CC:
Save the x86 PV specific parts of a domain. This is the X86_PV_INFO record,
the P2M_FRAMES, the X86_PV_SHARED_INFO, the three different VCPU context
records, and the MSR records.
The normalise_page callback used by the common code when writing the PAGE_DATA
records, converts MFNs in page tables t
This implementation of writev_exact() will cope with an iovcnt greater than
IOV_MAX because glibc will actually let this work anyway, and it is very
useful not to have to work about this in the caller of writev_exact(). The
caller is still required to ensure that the sum of iov_len's doesn't overf
Save/restore records common to all x86 domain types (HVM, PV).
This is only the TSC_INFO record.
Signed-off-by: Andrew Cooper
Acked-by: Ian Campbell
CC: Ian Jackson
CC: Wei Liu
---
tools/libxc/Makefile |1 +
tools/libxc/xc_sr_common_x86.c | 54
Add functions common to save and restore of x86 PV guests. This includes
functions for dealing with the P2M and M2P and the VCPU context.
Signed-off-by: Andrew Cooper
Acked-by: Ian Campbell
CC: Ian Jackson
CC: Wei Liu
---
tools/libxc/Makefile |1 +
tools/libxc/xc_sr_common_x
Progress information is logged via a different logger to regular libxc log
messages, and currently can only express a range. However, not everything
which needs reporting as progress comes with a range. Extend the interface to
allow reporting of a single statement.
The programming interface now
Presented here is v10 of the Migration v2 series (libxc subset), which is able
to function when transparently inserted under an unmodified xl/libxl.
There are no major changes from v9. A docs subseries has been completed in
parallel and removed from this series, and liberal quantities of spell
ch
For testing purposes, the environmental variable "XG_MIGRATION_V2" allows the
two save/restore codepaths to coexist, and have a runtime switch.
Signed-off-by: Andrew Cooper
Acked-by: Ian Campbell
CC: Ian Jackson
CC: Wei Liu
---
v8: Adjust layout to compile with MiniOS
---
tools/libxc/Makefil
On Thu, 2015-04-23 at 07:27 +, Pang, LongtaoX wrote:
>
>
> > -Original Message-
> > From: Ian Campbell [mailto:ian.campb...@citrix.com]
> > Sent: Tuesday, April 21, 2015 6:40 PM
> > To: Pang, LongtaoX
> > Cc: xen-devel@lists.xen.org; ian.jack...@eu.citrix.com;
> > wei.l...@citrix.com
On Thu, 2015-04-23 at 17:38 +0800, Robert Hu wrote:
> > > As mentioned in [ Patch], 'nested_l1' is ident for L1 guest VM,
> > > 'nestedl1' is hostname for L1 guest VM.
> > > ' store_runvar('nested_l1',$l1->{Guest});' is used to store L1's ident
> > > and L1's hostname to database, that will be
1 - 100 of 148 matches
Mail list logo
