* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-01-27 9:11 ` Alexander Graf
0 siblings, 0 replies; 45+ messages in thread
From: Alexander Graf @ 2014-01-27 9:11 UTC (permalink / raw)
To: Aneesh Kumar K.V
Cc: Liu Ping Fan, linuxppc-dev, kvm-ppc, kvm-devel, Ben Herrenschmidt,
Paul Mackerras
On 21.01.2014, at 10:42, Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> wrote:
> Liu Ping Fan <kernelfans@gmail.com> writes:
>
>> To make sure that on host, the pages marked with _PAGE_NUMA result in a fault
>> when guest access them, we should force the checking when guest uses hypercall
>> to setup hpte.
>>
>> Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
>
> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
>
> When we mark pte with _PAGE_NUMA we already call mmu_notifier_invalidate_range_start and
> mmu_notifier_invalidate_range_end, which will mark existing guest hpte
> entry as HPTE_V_ABSENT. Now we need to do that when we are inserting new
> guest hpte entries. This patch does that.
So what happens next? We insert a page into the HTAB without HPTE_V_VALID set, so the guest will fail to use it. If the guest does an H_READ on it it will suddenly turn to V_VALID though?
I might need a crash course in the use of HPTE_V_ABSENT.
Alex
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-01-27 9:11 ` Alexander Graf
0 siblings, 0 replies; 45+ messages in thread
From: Alexander Graf @ 2014-01-27 9:11 UTC (permalink / raw)
To: Aneesh Kumar K.V
Cc: kvm-devel, kvm-ppc, Paul Mackerras, Liu Ping Fan, linuxppc-dev
On 21.01.2014, at 10:42, Aneesh Kumar K.V =
<aneesh.kumar@linux.vnet.ibm.com> wrote:
> Liu Ping Fan <kernelfans@gmail.com> writes:
>=20
>> To make sure that on host, the pages marked with _PAGE_NUMA result in =
a fault
>> when guest access them, we should force the checking when guest uses =
hypercall
>> to setup hpte.
>>=20
>> Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
>=20
> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
>=20
> When we mark pte with _PAGE_NUMA we already call =
mmu_notifier_invalidate_range_start and
> mmu_notifier_invalidate_range_end, which will mark existing guest hpte
> entry as HPTE_V_ABSENT. Now we need to do that when we are inserting =
new
> guest hpte entries. This patch does that.=20
So what happens next? We insert a page into the HTAB without =
HPTE_V_VALID set, so the guest will fail to use it. If the guest does an =
H_READ on it it will suddenly turn to V_VALID though?
I might need a crash course in the use of HPTE_V_ABSENT.
Alex
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
2014-01-27 9:11 ` Alexander Graf
(?)
@ 2014-01-27 10:28 ` Aneesh Kumar K.V
-1 siblings, 0 replies; 45+ messages in thread
From: Aneesh Kumar K.V @ 2014-01-27 10:28 UTC (permalink / raw)
To: Alexander Graf
Cc: kvm-devel, kvm-ppc, Paul Mackerras, Liu Ping Fan, linuxppc-dev
Alexander Graf <agraf@suse.de> writes:
> On 21.01.2014, at 10:42, Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> wrote:
>
>> Liu Ping Fan <kernelfans@gmail.com> writes:
>>
>>> To make sure that on host, the pages marked with _PAGE_NUMA result in a fault
>>> when guest access them, we should force the checking when guest uses hypercall
>>> to setup hpte.
>>>
>>> Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
>>
>> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
>>
>> When we mark pte with _PAGE_NUMA we already call mmu_notifier_invalidate_range_start and
>> mmu_notifier_invalidate_range_end, which will mark existing guest hpte
>> entry as HPTE_V_ABSENT. Now we need to do that when we are inserting new
>> guest hpte entries. This patch does that.
>
> So what happens next? We insert a page into the HTAB without
> HPTE_V_VALID set, so the guest will fail to use it. If the guest does
> an H_READ on it it will suddenly turn to V_VALID though?
As per the guest the entry is valid, so yes an hread should return a
valid entry. But in real hpte we would mark it not valid.
>
> I might need a crash course in the use of HPTE_V_ABSENT.
When guest tries to access the address, the host will handle the fault.
kvmppc_hpte_hv_fault should give more info
-aneesh
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-01-27 10:28 ` Aneesh Kumar K.V
0 siblings, 0 replies; 45+ messages in thread
From: Aneesh Kumar K.V @ 2014-01-27 10:40 UTC (permalink / raw)
To: Alexander Graf
Cc: Liu Ping Fan, linuxppc-dev, kvm-ppc, kvm-devel, Ben Herrenschmidt,
Paul Mackerras
Alexander Graf <agraf@suse.de> writes:
> On 21.01.2014, at 10:42, Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> wrote:
>
>> Liu Ping Fan <kernelfans@gmail.com> writes:
>>
>>> To make sure that on host, the pages marked with _PAGE_NUMA result in a fault
>>> when guest access them, we should force the checking when guest uses hypercall
>>> to setup hpte.
>>>
>>> Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
>>
>> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
>>
>> When we mark pte with _PAGE_NUMA we already call mmu_notifier_invalidate_range_start and
>> mmu_notifier_invalidate_range_end, which will mark existing guest hpte
>> entry as HPTE_V_ABSENT. Now we need to do that when we are inserting new
>> guest hpte entries. This patch does that.
>
> So what happens next? We insert a page into the HTAB without
> HPTE_V_VALID set, so the guest will fail to use it. If the guest does
> an H_READ on it it will suddenly turn to V_VALID though?
As per the guest the entry is valid, so yes an hread should return a
valid entry. But in real hpte we would mark it not valid.
>
> I might need a crash course in the use of HPTE_V_ABSENT.
When guest tries to access the address, the host will handle the fault.
kvmppc_hpte_hv_fault should give more info
-aneesh
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-01-27 10:28 ` Aneesh Kumar K.V
0 siblings, 0 replies; 45+ messages in thread
From: Aneesh Kumar K.V @ 2014-01-27 10:28 UTC (permalink / raw)
To: Alexander Graf
Cc: Liu Ping Fan, linuxppc-dev, kvm-ppc, kvm-devel, Ben Herrenschmidt,
Paul Mackerras
Alexander Graf <agraf@suse.de> writes:
> On 21.01.2014, at 10:42, Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> wrote:
>
>> Liu Ping Fan <kernelfans@gmail.com> writes:
>>
>>> To make sure that on host, the pages marked with _PAGE_NUMA result in a fault
>>> when guest access them, we should force the checking when guest uses hypercall
>>> to setup hpte.
>>>
>>> Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
>>
>> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
>>
>> When we mark pte with _PAGE_NUMA we already call mmu_notifier_invalidate_range_start and
>> mmu_notifier_invalidate_range_end, which will mark existing guest hpte
>> entry as HPTE_V_ABSENT. Now we need to do that when we are inserting new
>> guest hpte entries. This patch does that.
>
> So what happens next? We insert a page into the HTAB without
> HPTE_V_VALID set, so the guest will fail to use it. If the guest does
> an H_READ on it it will suddenly turn to V_VALID though?
As per the guest the entry is valid, so yes an hread should return a
valid entry. But in real hpte we would mark it not valid.
>
> I might need a crash course in the use of HPTE_V_ABSENT.
When guest tries to access the address, the host will handle the fault.
kvmppc_hpte_hv_fault should give more info
-aneesh
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
2014-01-27 10:28 ` Aneesh Kumar K.V
(?)
@ 2014-01-27 10:42 ` Alexander Graf
-1 siblings, 0 replies; 45+ messages in thread
From: Alexander Graf @ 2014-01-27 10:42 UTC (permalink / raw)
To: Aneesh Kumar K.V
Cc: Liu Ping Fan, linuxppc-dev, kvm-ppc, kvm-devel, Ben Herrenschmidt,
Paul Mackerras
On 27.01.2014, at 11:28, Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> wrote:
> Alexander Graf <agraf@suse.de> writes:
>
>> On 21.01.2014, at 10:42, Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> wrote:
>>
>>> Liu Ping Fan <kernelfans@gmail.com> writes:
>>>
>>>> To make sure that on host, the pages marked with _PAGE_NUMA result in a fault
>>>> when guest access them, we should force the checking when guest uses hypercall
>>>> to setup hpte.
>>>>
>>>> Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
>>>
>>> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
>>>
>>> When we mark pte with _PAGE_NUMA we already call mmu_notifier_invalidate_range_start and
>>> mmu_notifier_invalidate_range_end, which will mark existing guest hpte
>>> entry as HPTE_V_ABSENT. Now we need to do that when we are inserting new
>>> guest hpte entries. This patch does that.
>>
>> So what happens next? We insert a page into the HTAB without
>> HPTE_V_VALID set, so the guest will fail to use it. If the guest does
>> an H_READ on it it will suddenly turn to V_VALID though?
>
> As per the guest the entry is valid, so yes an hread should return a
> valid entry. But in real hpte we would mark it not valid.
Ah, yes.
>
>>
>> I might need a crash course in the use of HPTE_V_ABSENT.
>
> When guest tries to access the address, the host will handle the fault.
>
> kvmppc_hpte_hv_fault should give more info
Thanks for the pointer. So we fault it in lazily. Is there any particular reason we can't do that on h_enter already? After all this just means an additional roundtrip because the guest is pretty likely to use the page it just entered, no?
Alex
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-01-27 10:42 ` Alexander Graf
0 siblings, 0 replies; 45+ messages in thread
From: Alexander Graf @ 2014-01-27 10:42 UTC (permalink / raw)
To: Aneesh Kumar K.V
Cc: Liu Ping Fan, linuxppc-dev, kvm-ppc, kvm-devel, Ben Herrenschmidt,
Paul Mackerras
On 27.01.2014, at 11:28, Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> wrote:
> Alexander Graf <agraf@suse.de> writes:
>
>> On 21.01.2014, at 10:42, Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> wrote:
>>
>>> Liu Ping Fan <kernelfans@gmail.com> writes:
>>>
>>>> To make sure that on host, the pages marked with _PAGE_NUMA result in a fault
>>>> when guest access them, we should force the checking when guest uses hypercall
>>>> to setup hpte.
>>>>
>>>> Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
>>>
>>> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
>>>
>>> When we mark pte with _PAGE_NUMA we already call mmu_notifier_invalidate_range_start and
>>> mmu_notifier_invalidate_range_end, which will mark existing guest hpte
>>> entry as HPTE_V_ABSENT. Now we need to do that when we are inserting new
>>> guest hpte entries. This patch does that.
>>
>> So what happens next? We insert a page into the HTAB without
>> HPTE_V_VALID set, so the guest will fail to use it. If the guest does
>> an H_READ on it it will suddenly turn to V_VALID though?
>
> As per the guest the entry is valid, so yes an hread should return a
> valid entry. But in real hpte we would mark it not valid.
Ah, yes.
>
>>
>> I might need a crash course in the use of HPTE_V_ABSENT.
>
> When guest tries to access the address, the host will handle the fault.
>
> kvmppc_hpte_hv_fault should give more info
Thanks for the pointer. So we fault it in lazily. Is there any particular reason we can't do that on h_enter already? After all this just means an additional roundtrip because the guest is pretty likely to use the page it just entered, no?
Alex
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-01-27 10:42 ` Alexander Graf
0 siblings, 0 replies; 45+ messages in thread
From: Alexander Graf @ 2014-01-27 10:42 UTC (permalink / raw)
To: Aneesh Kumar K.V
Cc: kvm-devel, kvm-ppc, Paul Mackerras, Liu Ping Fan, linuxppc-dev
On 27.01.2014, at 11:28, Aneesh Kumar K.V =
<aneesh.kumar@linux.vnet.ibm.com> wrote:
> Alexander Graf <agraf@suse.de> writes:
>=20
>> On 21.01.2014, at 10:42, Aneesh Kumar K.V =
<aneesh.kumar@linux.vnet.ibm.com> wrote:
>>=20
>>> Liu Ping Fan <kernelfans@gmail.com> writes:
>>>=20
>>>> To make sure that on host, the pages marked with _PAGE_NUMA result =
in a fault
>>>> when guest access them, we should force the checking when guest =
uses hypercall
>>>> to setup hpte.
>>>>=20
>>>> Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
>>>=20
>>> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
>>>=20
>>> When we mark pte with _PAGE_NUMA we already call =
mmu_notifier_invalidate_range_start and
>>> mmu_notifier_invalidate_range_end, which will mark existing guest =
hpte
>>> entry as HPTE_V_ABSENT. Now we need to do that when we are inserting =
new
>>> guest hpte entries. This patch does that.=20
>>=20
>> So what happens next? We insert a page into the HTAB without
>> HPTE_V_VALID set, so the guest will fail to use it. If the guest does
>> an H_READ on it it will suddenly turn to V_VALID though?
>=20
> As per the guest the entry is valid, so yes an hread should return a
> valid entry. But in real hpte we would mark it not valid.
Ah, yes.
>=20
>>=20
>> I might need a crash course in the use of HPTE_V_ABSENT.
>=20
> When guest tries to access the address, the host will handle the =
fault.
>=20
> kvmppc_hpte_hv_fault should give more info
Thanks for the pointer. So we fault it in lazily. Is there any =
particular reason we can't do that on h_enter already? After all this =
just means an additional roundtrip because the guest is pretty likely to =
use the page it just entered, no?
Alex
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
2014-01-27 10:42 ` Alexander Graf
(?)
@ 2014-01-27 15:57 ` Aneesh Kumar K.V
-1 siblings, 0 replies; 45+ messages in thread
From: Aneesh Kumar K.V @ 2014-01-27 15:57 UTC (permalink / raw)
To: Alexander Graf
Cc: kvm-devel, kvm-ppc, Paul Mackerras, Liu Ping Fan, linuxppc-dev
Alexander Graf <agraf@suse.de> writes:
> On 27.01.2014, at 11:28, Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> wrote:
>
>> Alexander Graf <agraf@suse.de> writes:
>>
>>> On 21.01.2014, at 10:42, Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> wrote:
>>>
>>>> Liu Ping Fan <kernelfans@gmail.com> writes:
>>>>
>>>>> To make sure that on host, the pages marked with _PAGE_NUMA result in a fault
>>>>> when guest access them, we should force the checking when guest uses hypercall
>>>>> to setup hpte.
>>>>>
>>>>> Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
>>>>
>>>> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
>>>>
>>>> When we mark pte with _PAGE_NUMA we already call mmu_notifier_invalidate_range_start and
>>>> mmu_notifier_invalidate_range_end, which will mark existing guest hpte
>>>> entry as HPTE_V_ABSENT. Now we need to do that when we are inserting new
>>>> guest hpte entries. This patch does that.
>>>
>>> So what happens next? We insert a page into the HTAB without
>>> HPTE_V_VALID set, so the guest will fail to use it. If the guest does
>>> an H_READ on it it will suddenly turn to V_VALID though?
>>
>> As per the guest the entry is valid, so yes an hread should return a
>> valid entry. But in real hpte we would mark it not valid.
>
> Ah, yes.
>
>>
>>>
>>> I might need a crash course in the use of HPTE_V_ABSENT.
>>
>> When guest tries to access the address, the host will handle the fault.
>>
>> kvmppc_hpte_hv_fault should give more info
>
> Thanks for the pointer. So we fault it in lazily. Is there any
> particular reason we can't do that on h_enter already? After all this
> just means an additional roundtrip because the guest is pretty likely
> to use the page it just entered, no?
We could get wrong numa fault information if we didn't do h_enter from
the right node from which we faulted.
-aneesh
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-01-27 15:57 ` Aneesh Kumar K.V
0 siblings, 0 replies; 45+ messages in thread
From: Aneesh Kumar K.V @ 2014-01-27 15:57 UTC (permalink / raw)
To: Alexander Graf
Cc: Liu Ping Fan, linuxppc-dev, kvm-ppc, kvm-devel, Ben Herrenschmidt,
Paul Mackerras
Alexander Graf <agraf@suse.de> writes:
> On 27.01.2014, at 11:28, Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> wrote:
>
>> Alexander Graf <agraf@suse.de> writes:
>>
>>> On 21.01.2014, at 10:42, Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> wrote:
>>>
>>>> Liu Ping Fan <kernelfans@gmail.com> writes:
>>>>
>>>>> To make sure that on host, the pages marked with _PAGE_NUMA result in a fault
>>>>> when guest access them, we should force the checking when guest uses hypercall
>>>>> to setup hpte.
>>>>>
>>>>> Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
>>>>
>>>> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
>>>>
>>>> When we mark pte with _PAGE_NUMA we already call mmu_notifier_invalidate_range_start and
>>>> mmu_notifier_invalidate_range_end, which will mark existing guest hpte
>>>> entry as HPTE_V_ABSENT. Now we need to do that when we are inserting new
>>>> guest hpte entries. This patch does that.
>>>
>>> So what happens next? We insert a page into the HTAB without
>>> HPTE_V_VALID set, so the guest will fail to use it. If the guest does
>>> an H_READ on it it will suddenly turn to V_VALID though?
>>
>> As per the guest the entry is valid, so yes an hread should return a
>> valid entry. But in real hpte we would mark it not valid.
>
> Ah, yes.
>
>>
>>>
>>> I might need a crash course in the use of HPTE_V_ABSENT.
>>
>> When guest tries to access the address, the host will handle the fault.
>>
>> kvmppc_hpte_hv_fault should give more info
>
> Thanks for the pointer. So we fault it in lazily. Is there any
> particular reason we can't do that on h_enter already? After all this
> just means an additional roundtrip because the guest is pretty likely
> to use the page it just entered, no?
We could get wrong numa fault information if we didn't do h_enter from
the right node from which we faulted.
-aneesh
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-01-27 15:57 ` Aneesh Kumar K.V
0 siblings, 0 replies; 45+ messages in thread
From: Aneesh Kumar K.V @ 2014-01-27 15:57 UTC (permalink / raw)
To: Alexander Graf
Cc: Liu Ping Fan, linuxppc-dev, kvm-ppc, kvm-devel, Ben Herrenschmidt,
Paul Mackerras
Alexander Graf <agraf@suse.de> writes:
> On 27.01.2014, at 11:28, Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> wrote:
>
>> Alexander Graf <agraf@suse.de> writes:
>>
>>> On 21.01.2014, at 10:42, Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> wrote:
>>>
>>>> Liu Ping Fan <kernelfans@gmail.com> writes:
>>>>
>>>>> To make sure that on host, the pages marked with _PAGE_NUMA result in a fault
>>>>> when guest access them, we should force the checking when guest uses hypercall
>>>>> to setup hpte.
>>>>>
>>>>> Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
>>>>
>>>> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
>>>>
>>>> When we mark pte with _PAGE_NUMA we already call mmu_notifier_invalidate_range_start and
>>>> mmu_notifier_invalidate_range_end, which will mark existing guest hpte
>>>> entry as HPTE_V_ABSENT. Now we need to do that when we are inserting new
>>>> guest hpte entries. This patch does that.
>>>
>>> So what happens next? We insert a page into the HTAB without
>>> HPTE_V_VALID set, so the guest will fail to use it. If the guest does
>>> an H_READ on it it will suddenly turn to V_VALID though?
>>
>> As per the guest the entry is valid, so yes an hread should return a
>> valid entry. But in real hpte we would mark it not valid.
>
> Ah, yes.
>
>>
>>>
>>> I might need a crash course in the use of HPTE_V_ABSENT.
>>
>> When guest tries to access the address, the host will handle the fault.
>>
>> kvmppc_hpte_hv_fault should give more info
>
> Thanks for the pointer. So we fault it in lazily. Is there any
> particular reason we can't do that on h_enter already? After all this
> just means an additional roundtrip because the guest is pretty likely
> to use the page it just entered, no?
We could get wrong numa fault information if we didn't do h_enter from
the right node from which we faulted.
-aneesh
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
2014-01-27 9:11 ` Alexander Graf
(?)
@ 2014-01-27 10:41 ` Paul Mackerras
-1 siblings, 0 replies; 45+ messages in thread
From: Paul Mackerras @ 2014-01-27 10:41 UTC (permalink / raw)
To: Alexander Graf
Cc: Aneesh Kumar K.V, Liu Ping Fan, linuxppc-dev, kvm-ppc, kvm-devel,
Ben Herrenschmidt
On Mon, Jan 27, 2014 at 10:11:40AM +0100, Alexander Graf wrote:
>
> On 21.01.2014, at 10:42, Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> wrote:
>
> > Liu Ping Fan <kernelfans@gmail.com> writes:
> >
> >> To make sure that on host, the pages marked with _PAGE_NUMA result in a fault
> >> when guest access them, we should force the checking when guest uses hypercall
> >> to setup hpte.
> >>
> >> Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
> >
> > Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
> >
> > When we mark pte with _PAGE_NUMA we already call mmu_notifier_invalidate_range_start and
> > mmu_notifier_invalidate_range_end, which will mark existing guest hpte
> > entry as HPTE_V_ABSENT. Now we need to do that when we are inserting new
> > guest hpte entries. This patch does that.
>
> So what happens next? We insert a page into the HTAB without HPTE_V_VALID set, so the guest will fail to use it. If the guest does an H_READ on it it will suddenly turn to V_VALID though?
>
> I might need a crash course in the use of HPTE_V_ABSENT.
HPTE_V_ABSENT means present from the point of view of the guest but
not present from the host's point of view, so yes H_READ turns
HPTE_V_ABSENT into HPTE_V_VALID.
Paul.
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-01-27 10:41 ` Paul Mackerras
0 siblings, 0 replies; 45+ messages in thread
From: Paul Mackerras @ 2014-01-27 10:41 UTC (permalink / raw)
To: Alexander Graf
Cc: Aneesh Kumar K.V, Liu Ping Fan, linuxppc-dev, kvm-ppc, kvm-devel,
Ben Herrenschmidt
On Mon, Jan 27, 2014 at 10:11:40AM +0100, Alexander Graf wrote:
>
> On 21.01.2014, at 10:42, Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> wrote:
>
> > Liu Ping Fan <kernelfans@gmail.com> writes:
> >
> >> To make sure that on host, the pages marked with _PAGE_NUMA result in a fault
> >> when guest access them, we should force the checking when guest uses hypercall
> >> to setup hpte.
> >>
> >> Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
> >
> > Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
> >
> > When we mark pte with _PAGE_NUMA we already call mmu_notifier_invalidate_range_start and
> > mmu_notifier_invalidate_range_end, which will mark existing guest hpte
> > entry as HPTE_V_ABSENT. Now we need to do that when we are inserting new
> > guest hpte entries. This patch does that.
>
> So what happens next? We insert a page into the HTAB without HPTE_V_VALID set, so the guest will fail to use it. If the guest does an H_READ on it it will suddenly turn to V_VALID though?
>
> I might need a crash course in the use of HPTE_V_ABSENT.
HPTE_V_ABSENT means present from the point of view of the guest but
not present from the host's point of view, so yes H_READ turns
HPTE_V_ABSENT into HPTE_V_VALID.
Paul.
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-01-27 10:41 ` Paul Mackerras
0 siblings, 0 replies; 45+ messages in thread
From: Paul Mackerras @ 2014-01-27 10:41 UTC (permalink / raw)
To: Alexander Graf
Cc: kvm-devel, kvm-ppc, Aneesh Kumar K.V, Liu Ping Fan, linuxppc-dev
On Mon, Jan 27, 2014 at 10:11:40AM +0100, Alexander Graf wrote:
>
> On 21.01.2014, at 10:42, Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> wrote:
>
> > Liu Ping Fan <kernelfans@gmail.com> writes:
> >
> >> To make sure that on host, the pages marked with _PAGE_NUMA result in a fault
> >> when guest access them, we should force the checking when guest uses hypercall
> >> to setup hpte.
> >>
> >> Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
> >
> > Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
> >
> > When we mark pte with _PAGE_NUMA we already call mmu_notifier_invalidate_range_start and
> > mmu_notifier_invalidate_range_end, which will mark existing guest hpte
> > entry as HPTE_V_ABSENT. Now we need to do that when we are inserting new
> > guest hpte entries. This patch does that.
>
> So what happens next? We insert a page into the HTAB without HPTE_V_VALID set, so the guest will fail to use it. If the guest does an H_READ on it it will suddenly turn to V_VALID though?
>
> I might need a crash course in the use of HPTE_V_ABSENT.
HPTE_V_ABSENT means present from the point of view of the guest but
not present from the host's point of view, so yes H_READ turns
HPTE_V_ABSENT into HPTE_V_VALID.
Paul.
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
2014-01-27 9:11 ` Alexander Graf
(?)
@ 2014-04-03 2:36 ` Liu ping fan
-1 siblings, 0 replies; 45+ messages in thread
From: Liu ping fan @ 2014-04-03 2:36 UTC (permalink / raw)
To: Alexander Graf
Cc: kvm-devel, kvm-ppc, Paul Mackerras, Aneesh Kumar K.V,
linuxppc-dev
Hi Alex, could you help to pick up this patch? since v3.14 kernel can
enable numa fault for powerpc.
Thx,
Fan
On Mon, Jan 27, 2014 at 5:11 PM, Alexander Graf <agraf@suse.de> wrote:
>
> On 21.01.2014, at 10:42, Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> wrote:
>
>> Liu Ping Fan <kernelfans@gmail.com> writes:
>>
>>> To make sure that on host, the pages marked with _PAGE_NUMA result in a fault
>>> when guest access them, we should force the checking when guest uses hypercall
>>> to setup hpte.
>>>
>>> Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
>>
>> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
>>
>> When we mark pte with _PAGE_NUMA we already call mmu_notifier_invalidate_range_start and
>> mmu_notifier_invalidate_range_end, which will mark existing guest hpte
>> entry as HPTE_V_ABSENT. Now we need to do that when we are inserting new
>> guest hpte entries. This patch does that.
>
> So what happens next? We insert a page into the HTAB without HPTE_V_VALID set, so the guest will fail to use it. If the guest does an H_READ on it it will suddenly turn to V_VALID though?
>
> I might need a crash course in the use of HPTE_V_ABSENT.
>
>
> Alex
>
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-04-03 2:36 ` Liu ping fan
0 siblings, 0 replies; 45+ messages in thread
From: Liu ping fan @ 2014-04-03 2:36 UTC (permalink / raw)
To: Alexander Graf
Cc: kvm-devel, kvm-ppc, Paul Mackerras, Aneesh Kumar K.V,
linuxppc-dev
Hi Alex, could you help to pick up this patch? since v3.14 kernel can
enable numa fault for powerpc.
Thx,
Fan
On Mon, Jan 27, 2014 at 5:11 PM, Alexander Graf <agraf@suse.de> wrote:
>
> On 21.01.2014, at 10:42, Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> wrote:
>
>> Liu Ping Fan <kernelfans@gmail.com> writes:
>>
>>> To make sure that on host, the pages marked with _PAGE_NUMA result in a fault
>>> when guest access them, we should force the checking when guest uses hypercall
>>> to setup hpte.
>>>
>>> Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
>>
>> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
>>
>> When we mark pte with _PAGE_NUMA we already call mmu_notifier_invalidate_range_start and
>> mmu_notifier_invalidate_range_end, which will mark existing guest hpte
>> entry as HPTE_V_ABSENT. Now we need to do that when we are inserting new
>> guest hpte entries. This patch does that.
>
> So what happens next? We insert a page into the HTAB without HPTE_V_VALID set, so the guest will fail to use it. If the guest does an H_READ on it it will suddenly turn to V_VALID though?
>
> I might need a crash course in the use of HPTE_V_ABSENT.
>
>
> Alex
>
_______________________________________________
Linuxppc-dev mailing list
Linuxppc-dev@lists.ozlabs.org
https://lists.ozlabs.org/listinfo/linuxppc-dev
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-04-03 2:36 ` Liu ping fan
0 siblings, 0 replies; 45+ messages in thread
From: Liu ping fan @ 2014-04-03 2:36 UTC (permalink / raw)
To: Alexander Graf
Cc: kvm-devel, kvm-ppc, Paul Mackerras, Aneesh Kumar K.V,
linuxppc-dev
Hi Alex, could you help to pick up this patch? since v3.14 kernel can
enable numa fault for powerpc.
Thx,
Fan
On Mon, Jan 27, 2014 at 5:11 PM, Alexander Graf <agraf@suse.de> wrote:
>
> On 21.01.2014, at 10:42, Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com> wrote:
>
>> Liu Ping Fan <kernelfans@gmail.com> writes:
>>
>>> To make sure that on host, the pages marked with _PAGE_NUMA result in a fault
>>> when guest access them, we should force the checking when guest uses hypercall
>>> to setup hpte.
>>>
>>> Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
>>
>> Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
>>
>> When we mark pte with _PAGE_NUMA we already call mmu_notifier_invalidate_range_start and
>> mmu_notifier_invalidate_range_end, which will mark existing guest hpte
>> entry as HPTE_V_ABSENT. Now we need to do that when we are inserting new
>> guest hpte entries. This patch does that.
>
> So what happens next? We insert a page into the HTAB without HPTE_V_VALID set, so the guest will fail to use it. If the guest does an H_READ on it it will suddenly turn to V_VALID though?
>
> I might need a crash course in the use of HPTE_V_ABSENT.
>
>
> Alex
>
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
2014-04-03 2:36 ` Liu ping fan
(?)
@ 2014-04-03 11:36 ` Alexander Graf
-1 siblings, 0 replies; 45+ messages in thread
From: Alexander Graf @ 2014-04-03 11:36 UTC (permalink / raw)
To: Liu ping fan, Alexander Graf
Cc: Aneesh Kumar K.V, linuxppc-dev, kvm-ppc, kvm-devel,
Ben Herrenschmidt, Paul Mackerras
On 03.04.14 04:36, Liu ping fan wrote:
> Hi Alex, could you help to pick up this patch? since v3.14 kernel can
> enable numa fault for powerpc.
What bad happens without this patch? We map a page even though it was
declared to get NUMA migrated? What happens next?
I'm trying to figure out whether I need to mark this with a stable tag
for 3.14.
Alex
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-04-03 11:36 ` Alexander Graf
0 siblings, 0 replies; 45+ messages in thread
From: Alexander Graf @ 2014-04-03 11:36 UTC (permalink / raw)
To: Liu ping fan, Alexander Graf
Cc: Aneesh Kumar K.V, linuxppc-dev, kvm-ppc, kvm-devel,
Ben Herrenschmidt, Paul Mackerras
On 03.04.14 04:36, Liu ping fan wrote:
> Hi Alex, could you help to pick up this patch? since v3.14 kernel can
> enable numa fault for powerpc.
What bad happens without this patch? We map a page even though it was
declared to get NUMA migrated? What happens next?
I'm trying to figure out whether I need to mark this with a stable tag
for 3.14.
Alex
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-04-03 11:36 ` Alexander Graf
0 siblings, 0 replies; 45+ messages in thread
From: Alexander Graf @ 2014-04-03 11:36 UTC (permalink / raw)
To: Liu ping fan, Alexander Graf
Cc: kvm-devel, kvm-ppc, Paul Mackerras, Aneesh Kumar K.V,
linuxppc-dev
On 03.04.14 04:36, Liu ping fan wrote:
> Hi Alex, could you help to pick up this patch? since v3.14 kernel can
> enable numa fault for powerpc.
What bad happens without this patch? We map a page even though it was
declared to get NUMA migrated? What happens next?
I'm trying to figure out whether I need to mark this with a stable tag
for 3.14.
Alex
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
2014-04-03 11:36 ` Alexander Graf
(?)
@ 2014-04-03 11:38 ` Alexander Graf
-1 siblings, 0 replies; 45+ messages in thread
From: Alexander Graf @ 2014-04-03 11:38 UTC (permalink / raw)
To: Liu ping fan, Alexander Graf
Cc: Aneesh Kumar K.V, linuxppc-dev, kvm-ppc, kvm-devel,
Ben Herrenschmidt, Paul Mackerras
On 03.04.14 13:36, Alexander Graf wrote:
>
> On 03.04.14 04:36, Liu ping fan wrote:
>> Hi Alex, could you help to pick up this patch? since v3.14 kernel can
>> enable numa fault for powerpc.
>
> What bad happens without this patch? We map a page even though it was
> declared to get NUMA migrated? What happens next?
>
> I'm trying to figure out whether I need to mark this with a stable tag
> for 3.14.
Also, what about the other uses of pre_present()?
Alex
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-04-03 11:38 ` Alexander Graf
0 siblings, 0 replies; 45+ messages in thread
From: Alexander Graf @ 2014-04-03 11:38 UTC (permalink / raw)
To: Liu ping fan, Alexander Graf
Cc: Aneesh Kumar K.V, linuxppc-dev, kvm-ppc, kvm-devel,
Ben Herrenschmidt, Paul Mackerras
On 03.04.14 13:36, Alexander Graf wrote:
>
> On 03.04.14 04:36, Liu ping fan wrote:
>> Hi Alex, could you help to pick up this patch? since v3.14 kernel can
>> enable numa fault for powerpc.
>
> What bad happens without this patch? We map a page even though it was
> declared to get NUMA migrated? What happens next?
>
> I'm trying to figure out whether I need to mark this with a stable tag
> for 3.14.
Also, what about the other uses of pre_present()?
Alex
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-04-03 11:38 ` Alexander Graf
0 siblings, 0 replies; 45+ messages in thread
From: Alexander Graf @ 2014-04-03 11:38 UTC (permalink / raw)
To: Liu ping fan, Alexander Graf
Cc: kvm-devel, kvm-ppc, Paul Mackerras, Aneesh Kumar K.V,
linuxppc-dev
On 03.04.14 13:36, Alexander Graf wrote:
>
> On 03.04.14 04:36, Liu ping fan wrote:
>> Hi Alex, could you help to pick up this patch? since v3.14 kernel can
>> enable numa fault for powerpc.
>
> What bad happens without this patch? We map a page even though it was
> declared to get NUMA migrated? What happens next?
>
> I'm trying to figure out whether I need to mark this with a stable tag
> for 3.14.
Also, what about the other uses of pre_present()?
Alex
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
2014-04-03 11:38 ` Alexander Graf
(?)
@ 2014-04-03 11:43 ` Alexander Graf
-1 siblings, 0 replies; 45+ messages in thread
From: Alexander Graf @ 2014-04-03 11:43 UTC (permalink / raw)
To: Liu ping fan, Alexander Graf
Cc: Aneesh Kumar K.V, linuxppc-dev, kvm-ppc, kvm-devel,
Ben Herrenschmidt, Paul Mackerras
On 03.04.14 13:38, Alexander Graf wrote:
>
> On 03.04.14 13:36, Alexander Graf wrote:
>>
>> On 03.04.14 04:36, Liu ping fan wrote:
>>> Hi Alex, could you help to pick up this patch? since v3.14 kernel can
>>> enable numa fault for powerpc.
>>
>> What bad happens without this patch? We map a page even though it was
>> declared to get NUMA migrated? What happens next?
>>
>> I'm trying to figure out whether I need to mark this with a stable
>> tag for 3.14.
>
> Also, what about the other uses of pre_present()?
Oh how I love to reply to myself. The patch doesn't apply to the latest
code anymore either. Also please rework the wording of your patch
subject and patch description to explain the actual problem at hand.
Alex
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-04-03 11:43 ` Alexander Graf
0 siblings, 0 replies; 45+ messages in thread
From: Alexander Graf @ 2014-04-03 11:43 UTC (permalink / raw)
To: Liu ping fan, Alexander Graf
Cc: Aneesh Kumar K.V, linuxppc-dev, kvm-ppc, kvm-devel,
Ben Herrenschmidt, Paul Mackerras
On 03.04.14 13:38, Alexander Graf wrote:
>
> On 03.04.14 13:36, Alexander Graf wrote:
>>
>> On 03.04.14 04:36, Liu ping fan wrote:
>>> Hi Alex, could you help to pick up this patch? since v3.14 kernel can
>>> enable numa fault for powerpc.
>>
>> What bad happens without this patch? We map a page even though it was
>> declared to get NUMA migrated? What happens next?
>>
>> I'm trying to figure out whether I need to mark this with a stable
>> tag for 3.14.
>
> Also, what about the other uses of pre_present()?
Oh how I love to reply to myself. The patch doesn't apply to the latest
code anymore either. Also please rework the wording of your patch
subject and patch description to explain the actual problem at hand.
Alex
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-04-03 11:43 ` Alexander Graf
0 siblings, 0 replies; 45+ messages in thread
From: Alexander Graf @ 2014-04-03 11:43 UTC (permalink / raw)
To: Liu ping fan, Alexander Graf
Cc: kvm-devel, kvm-ppc, Paul Mackerras, Aneesh Kumar K.V,
linuxppc-dev
On 03.04.14 13:38, Alexander Graf wrote:
>
> On 03.04.14 13:36, Alexander Graf wrote:
>>
>> On 03.04.14 04:36, Liu ping fan wrote:
>>> Hi Alex, could you help to pick up this patch? since v3.14 kernel can
>>> enable numa fault for powerpc.
>>
>> What bad happens without this patch? We map a page even though it was
>> declared to get NUMA migrated? What happens next?
>>
>> I'm trying to figure out whether I need to mark this with a stable
>> tag for 3.14.
>
> Also, what about the other uses of pre_present()?
Oh how I love to reply to myself. The patch doesn't apply to the latest
code anymore either. Also please rework the wording of your patch
subject and patch description to explain the actual problem at hand.
Alex
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
2014-04-03 11:36 ` Alexander Graf
(?)
@ 2014-04-07 7:42 ` Aneesh Kumar K.V
-1 siblings, 0 replies; 45+ messages in thread
From: Aneesh Kumar K.V @ 2014-04-07 7:42 UTC (permalink / raw)
To: Alexander Graf, Liu ping fan, Alexander Graf
Cc: Paul Mackerras, linuxppc-dev, kvm-devel, kvm-ppc
Alexander Graf <agraf@suse.com> writes:
> On 03.04.14 04:36, Liu ping fan wrote:
>> Hi Alex, could you help to pick up this patch? since v3.14 kernel can
>> enable numa fault for powerpc.
>
> What bad happens without this patch? We map a page even though it was
> declared to get NUMA migrated? What happens next?
Nothing much, we won't be properly accounting the numa access in the
host. What we want to achieve is to convert a guest access of the page to
a host fault so that we can do proper numa access accounting in the
host. This would enable us to migrate the page to the correct numa
node.
>
> I'm trying to figure out whether I need to mark this with a stable tag
> for 3.14.
>
>
-aneesh
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-04-07 7:42 ` Aneesh Kumar K.V
0 siblings, 0 replies; 45+ messages in thread
From: Aneesh Kumar K.V @ 2014-04-07 7:54 UTC (permalink / raw)
To: Alexander Graf, Liu ping fan, Alexander Graf
Cc: linuxppc-dev, kvm-ppc, kvm-devel, Ben Herrenschmidt,
Paul Mackerras
Alexander Graf <agraf@suse.com> writes:
> On 03.04.14 04:36, Liu ping fan wrote:
>> Hi Alex, could you help to pick up this patch? since v3.14 kernel can
>> enable numa fault for powerpc.
>
> What bad happens without this patch? We map a page even though it was
> declared to get NUMA migrated? What happens next?
Nothing much, we won't be properly accounting the numa access in the
host. What we want to achieve is to convert a guest access of the page to
a host fault so that we can do proper numa access accounting in the
host. This would enable us to migrate the page to the correct numa
node.
>
> I'm trying to figure out whether I need to mark this with a stable tag
> for 3.14.
>
>
-aneesh
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-04-07 7:42 ` Aneesh Kumar K.V
0 siblings, 0 replies; 45+ messages in thread
From: Aneesh Kumar K.V @ 2014-04-07 7:42 UTC (permalink / raw)
To: Alexander Graf, Liu ping fan, Alexander Graf
Cc: linuxppc-dev, kvm-ppc, kvm-devel, Ben Herrenschmidt,
Paul Mackerras
Alexander Graf <agraf@suse.com> writes:
> On 03.04.14 04:36, Liu ping fan wrote:
>> Hi Alex, could you help to pick up this patch? since v3.14 kernel can
>> enable numa fault for powerpc.
>
> What bad happens without this patch? We map a page even though it was
> declared to get NUMA migrated? What happens next?
Nothing much, we won't be properly accounting the numa access in the
host. What we want to achieve is to convert a guest access of the page to
a host fault so that we can do proper numa access accounting in the
host. This would enable us to migrate the page to the correct numa
node.
>
> I'm trying to figure out whether I need to mark this with a stable tag
> for 3.14.
>
>
-aneesh
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
2014-04-07 7:42 ` Aneesh Kumar K.V
(?)
@ 2014-04-07 8:36 ` Alexander Graf
-1 siblings, 0 replies; 45+ messages in thread
From: Alexander Graf @ 2014-04-07 8:36 UTC (permalink / raw)
To: Aneesh Kumar K.V, Alexander Graf, Liu ping fan
Cc: linuxppc-dev, kvm-ppc, kvm-devel, Ben Herrenschmidt,
Paul Mackerras
On 07.04.14 09:42, Aneesh Kumar K.V wrote:
> Alexander Graf <agraf@suse.com> writes:
>
>> On 03.04.14 04:36, Liu ping fan wrote:
>>> Hi Alex, could you help to pick up this patch? since v3.14 kernel can
>>> enable numa fault for powerpc.
>> What bad happens without this patch? We map a page even though it was
>> declared to get NUMA migrated? What happens next?
> Nothing much, we won't be properly accounting the numa access in the
> host. What we want to achieve is to convert a guest access of the page to
> a host fault so that we can do proper numa access accounting in the
> host. This would enable us to migrate the page to the correct numa
> node.
Ok, so no breakages, just less performance. I wouldn't consider it
stable material then :).
Alex
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-04-07 8:36 ` Alexander Graf
0 siblings, 0 replies; 45+ messages in thread
From: Alexander Graf @ 2014-04-07 8:36 UTC (permalink / raw)
To: Aneesh Kumar K.V, Alexander Graf, Liu ping fan
Cc: linuxppc-dev, kvm-ppc, kvm-devel, Ben Herrenschmidt,
Paul Mackerras
On 07.04.14 09:42, Aneesh Kumar K.V wrote:
> Alexander Graf <agraf@suse.com> writes:
>
>> On 03.04.14 04:36, Liu ping fan wrote:
>>> Hi Alex, could you help to pick up this patch? since v3.14 kernel can
>>> enable numa fault for powerpc.
>> What bad happens without this patch? We map a page even though it was
>> declared to get NUMA migrated? What happens next?
> Nothing much, we won't be properly accounting the numa access in the
> host. What we want to achieve is to convert a guest access of the page to
> a host fault so that we can do proper numa access accounting in the
> host. This would enable us to migrate the page to the correct numa
> node.
Ok, so no breakages, just less performance. I wouldn't consider it
stable material then :).
Alex
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-04-07 8:36 ` Alexander Graf
0 siblings, 0 replies; 45+ messages in thread
From: Alexander Graf @ 2014-04-07 8:36 UTC (permalink / raw)
To: Aneesh Kumar K.V, Alexander Graf, Liu ping fan
Cc: Paul Mackerras, linuxppc-dev, kvm-devel, kvm-ppc
On 07.04.14 09:42, Aneesh Kumar K.V wrote:
> Alexander Graf <agraf@suse.com> writes:
>
>> On 03.04.14 04:36, Liu ping fan wrote:
>>> Hi Alex, could you help to pick up this patch? since v3.14 kernel can
>>> enable numa fault for powerpc.
>> What bad happens without this patch? We map a page even though it was
>> declared to get NUMA migrated? What happens next?
> Nothing much, we won't be properly accounting the numa access in the
> host. What we want to achieve is to convert a guest access of the page to
> a host fault so that we can do proper numa access accounting in the
> host. This would enable us to migrate the page to the correct numa
> node.
Ok, so no breakages, just less performance. I wouldn't consider it
stable material then :).
Alex
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
2014-04-07 8:36 ` Alexander Graf
(?)
@ 2014-04-10 3:28 ` Liu ping fan
-1 siblings, 0 replies; 45+ messages in thread
From: Liu ping fan @ 2014-04-10 3:28 UTC (permalink / raw)
To: Alexander Graf
Cc: Aneesh Kumar K.V, Alexander Graf, linuxppc-dev, kvm-ppc,
kvm-devel, Ben Herrenschmidt, Paul Mackerras
On Mon, Apr 7, 2014 at 4:36 PM, Alexander Graf <agraf@suse.de> wrote:
>
> On 07.04.14 09:42, Aneesh Kumar K.V wrote:
>>
>> Alexander Graf <agraf@suse.com> writes:
>>
>>> On 03.04.14 04:36, Liu ping fan wrote:
>>>>
>>>> Hi Alex, could you help to pick up this patch? since v3.14 kernel can
>>>> enable numa fault for powerpc.
>>>
>>> What bad happens without this patch? We map a page even though it was
>>> declared to get NUMA migrated? What happens next?
>>
>> Nothing much, we won't be properly accounting the numa access in the
>> host. What we want to achieve is to convert a guest access of the page to
>> a host fault so that we can do proper numa access accounting in the
>> host. This would enable us to migrate the page to the correct numa
>> node.
>
>
> Ok, so no breakages, just less performance. I wouldn't consider it stable
> material then :).
>
Sorry to reply late, since I am half out of office during this period.
I am puzzling about you reply. Without this patch, the guest can
NOT sense the numa changes and ask host to put the pages in right
place. So the pages which is used by guest will be always misplaced.
The numa-fault method is inspired by real requirement to improve
performance, so we should also consider the performance drop of guest.
Right?
Regards,
Fan
>
> Alex
>
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-04-10 3:28 ` Liu ping fan
0 siblings, 0 replies; 45+ messages in thread
From: Liu ping fan @ 2014-04-10 3:28 UTC (permalink / raw)
To: Alexander Graf
Cc: Aneesh Kumar K.V, Alexander Graf, linuxppc-dev, kvm-ppc,
kvm-devel, Ben Herrenschmidt, Paul Mackerras
On Mon, Apr 7, 2014 at 4:36 PM, Alexander Graf <agraf@suse.de> wrote:
>
> On 07.04.14 09:42, Aneesh Kumar K.V wrote:
>>
>> Alexander Graf <agraf@suse.com> writes:
>>
>>> On 03.04.14 04:36, Liu ping fan wrote:
>>>>
>>>> Hi Alex, could you help to pick up this patch? since v3.14 kernel can
>>>> enable numa fault for powerpc.
>>>
>>> What bad happens without this patch? We map a page even though it was
>>> declared to get NUMA migrated? What happens next?
>>
>> Nothing much, we won't be properly accounting the numa access in the
>> host. What we want to achieve is to convert a guest access of the page to
>> a host fault so that we can do proper numa access accounting in the
>> host. This would enable us to migrate the page to the correct numa
>> node.
>
>
> Ok, so no breakages, just less performance. I wouldn't consider it stable
> material then :).
>
Sorry to reply late, since I am half out of office during this period.
I am puzzling about you reply. Without this patch, the guest can
NOT sense the numa changes and ask host to put the pages in right
place. So the pages which is used by guest will be always misplaced.
The numa-fault method is inspired by real requirement to improve
performance, so we should also consider the performance drop of guest.
Right?
Regards,
Fan
>
> Alex
>
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-04-10 3:28 ` Liu ping fan
0 siblings, 0 replies; 45+ messages in thread
From: Liu ping fan @ 2014-04-10 3:28 UTC (permalink / raw)
To: Alexander Graf
Cc: kvm-devel, kvm-ppc, Paul Mackerras, Aneesh Kumar K.V,
linuxppc-dev, Alexander Graf
On Mon, Apr 7, 2014 at 4:36 PM, Alexander Graf <agraf@suse.de> wrote:
>
> On 07.04.14 09:42, Aneesh Kumar K.V wrote:
>>
>> Alexander Graf <agraf@suse.com> writes:
>>
>>> On 03.04.14 04:36, Liu ping fan wrote:
>>>>
>>>> Hi Alex, could you help to pick up this patch? since v3.14 kernel can
>>>> enable numa fault for powerpc.
>>>
>>> What bad happens without this patch? We map a page even though it was
>>> declared to get NUMA migrated? What happens next?
>>
>> Nothing much, we won't be properly accounting the numa access in the
>> host. What we want to achieve is to convert a guest access of the page to
>> a host fault so that we can do proper numa access accounting in the
>> host. This would enable us to migrate the page to the correct numa
>> node.
>
>
> Ok, so no breakages, just less performance. I wouldn't consider it stable
> material then :).
>
Sorry to reply late, since I am half out of office during this period.
I am puzzling about you reply. Without this patch, the guest can
NOT sense the numa changes and ask host to put the pages in right
place. So the pages which is used by guest will be always misplaced.
The numa-fault method is inspired by real requirement to improve
performance, so we should also consider the performance drop of guest.
Right?
Regards,
Fan
>
> Alex
>
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
2014-04-10 3:28 ` Liu ping fan
(?)
@ 2014-04-10 10:02 ` Alexander Graf
-1 siblings, 0 replies; 45+ messages in thread
From: Alexander Graf @ 2014-04-10 10:02 UTC (permalink / raw)
To: Liu ping fan
Cc: Aneesh Kumar K.V, Alexander Graf, linuxppc-dev, kvm-ppc,
kvm-devel, Ben Herrenschmidt, Paul Mackerras
On 10.04.14 05:28, Liu ping fan wrote:
> On Mon, Apr 7, 2014 at 4:36 PM, Alexander Graf <agraf@suse.de> wrote:
>> On 07.04.14 09:42, Aneesh Kumar K.V wrote:
>>> Alexander Graf <agraf@suse.com> writes:
>>>
>>>> On 03.04.14 04:36, Liu ping fan wrote:
>>>>> Hi Alex, could you help to pick up this patch? since v3.14 kernel can
>>>>> enable numa fault for powerpc.
>>>> What bad happens without this patch? We map a page even though it was
>>>> declared to get NUMA migrated? What happens next?
>>> Nothing much, we won't be properly accounting the numa access in the
>>> host. What we want to achieve is to convert a guest access of the page to
>>> a host fault so that we can do proper numa access accounting in the
>>> host. This would enable us to migrate the page to the correct numa
>>> node.
>>
>> Ok, so no breakages, just less performance. I wouldn't consider it stable
>> material then :).
>>
> Sorry to reply late, since I am half out of office during this period.
> I am puzzling about you reply. Without this patch, the guest can
> NOT sense the numa changes and ask host to put the pages in right
> place. So the pages which is used by guest will be always misplaced.
> The numa-fault method is inspired by real requirement to improve
> performance, so we should also consider the performance drop of guest.
> Right?
The patch will get into Linux, I just consider a non-working new feature
not a regression that warrants us to CC stable@vger :). After all
performance shouldn't be worse than without the numa migration feature,
correct?
Alex
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-04-10 10:02 ` Alexander Graf
0 siblings, 0 replies; 45+ messages in thread
From: Alexander Graf @ 2014-04-10 10:02 UTC (permalink / raw)
To: Liu ping fan
Cc: Aneesh Kumar K.V, Alexander Graf, linuxppc-dev, kvm-ppc,
kvm-devel, Ben Herrenschmidt, Paul Mackerras
On 10.04.14 05:28, Liu ping fan wrote:
> On Mon, Apr 7, 2014 at 4:36 PM, Alexander Graf <agraf@suse.de> wrote:
>> On 07.04.14 09:42, Aneesh Kumar K.V wrote:
>>> Alexander Graf <agraf@suse.com> writes:
>>>
>>>> On 03.04.14 04:36, Liu ping fan wrote:
>>>>> Hi Alex, could you help to pick up this patch? since v3.14 kernel can
>>>>> enable numa fault for powerpc.
>>>> What bad happens without this patch? We map a page even though it was
>>>> declared to get NUMA migrated? What happens next?
>>> Nothing much, we won't be properly accounting the numa access in the
>>> host. What we want to achieve is to convert a guest access of the page to
>>> a host fault so that we can do proper numa access accounting in the
>>> host. This would enable us to migrate the page to the correct numa
>>> node.
>>
>> Ok, so no breakages, just less performance. I wouldn't consider it stable
>> material then :).
>>
> Sorry to reply late, since I am half out of office during this period.
> I am puzzling about you reply. Without this patch, the guest can
> NOT sense the numa changes and ask host to put the pages in right
> place. So the pages which is used by guest will be always misplaced.
> The numa-fault method is inspired by real requirement to improve
> performance, so we should also consider the performance drop of guest.
> Right?
The patch will get into Linux, I just consider a non-working new feature
not a regression that warrants us to CC stable@vger :). After all
performance shouldn't be worse than without the numa migration feature,
correct?
Alex
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH v2] powernv: kvm: make _PAGE_NUMA take effect
@ 2014-04-10 10:02 ` Alexander Graf
0 siblings, 0 replies; 45+ messages in thread
From: Alexander Graf @ 2014-04-10 10:02 UTC (permalink / raw)
To: Liu ping fan
Cc: kvm-devel, kvm-ppc, Paul Mackerras, Aneesh Kumar K.V,
linuxppc-dev, Alexander Graf
On 10.04.14 05:28, Liu ping fan wrote:
> On Mon, Apr 7, 2014 at 4:36 PM, Alexander Graf <agraf@suse.de> wrote:
>> On 07.04.14 09:42, Aneesh Kumar K.V wrote:
>>> Alexander Graf <agraf@suse.com> writes:
>>>
>>>> On 03.04.14 04:36, Liu ping fan wrote:
>>>>> Hi Alex, could you help to pick up this patch? since v3.14 kernel can
>>>>> enable numa fault for powerpc.
>>>> What bad happens without this patch? We map a page even though it was
>>>> declared to get NUMA migrated? What happens next?
>>> Nothing much, we won't be properly accounting the numa access in the
>>> host. What we want to achieve is to convert a guest access of the page to
>>> a host fault so that we can do proper numa access accounting in the
>>> host. This would enable us to migrate the page to the correct numa
>>> node.
>>
>> Ok, so no breakages, just less performance. I wouldn't consider it stable
>> material then :).
>>
> Sorry to reply late, since I am half out of office during this period.
> I am puzzling about you reply. Without this patch, the guest can
> NOT sense the numa changes and ask host to put the pages in right
> place. So the pages which is used by guest will be always misplaced.
> The numa-fault method is inspired by real requirement to improve
> performance, so we should also consider the performance drop of guest.
> Right?
The patch will get into Linux, I just consider a non-working new feature
not a regression that warrants us to CC stable@vger :). After all
performance shouldn't be worse than without the numa migration feature,
correct?
Alex
^ permalink raw reply [flat|nested] 45+ messages in thread