On 12/06/2013 08:22 AM, Marcelo Tosatti wrote:
> On Thu, Dec 05, 2013 at 11:30:27PM +0800, Xiao Guangrong wrote:
>> In some cases, the lockless walker will do endless-walking on desc and
>> without rewalk, consider this case:
>>
>> there are two descs: desc1 and desc2 who is pointed by
On 12/06/2013 08:22 AM, Marcelo Tosatti wrote:
On Thu, Dec 05, 2013 at 11:30:27PM +0800, Xiao Guangrong wrote:
In some cases, the lockless walker will do endless-walking on desc and
without rewalk, consider this case:
there are two descs: desc1 and desc2 who is pointed by desc1-next:
On Thu, Dec 05, 2013 at 11:30:27PM +0800, Xiao Guangrong wrote:
> In some cases, the lockless walker will do endless-walking on desc and
> without rewalk, consider this case:
>
> there are two descs: desc1 and desc2 who is pointed by desc1->next:
> desc1->next = desc2.
>
> CPU 0
On Thu, Dec 05, 2013 at 11:30:27PM +0800, Xiao Guangrong wrote:
> > Is it not the case that simply moving to the slow path once a maximum of
> > rewalks has been reached enough? (looks a like a good solution).
>
> In some cases, the lockless walker will do endless-walking on desc and
> without
On Dec 5, 2013, at 9:50 PM, Marcelo Tosatti wrote:
> GOn Tue, Dec 03, 2013 at 03:10:48PM +0800, Xiao Guangrong wrote:
>> On 11/28/2013 04:53 PM, Xiao Guangrong wrote:
>>> On 11/27/2013 03:31 AM, Marcelo Tosatti wrote:
On Tue, Nov 26, 2013 at 11:21:37AM +0800, Xiao Guangrong wrote:
> On
GOn Tue, Dec 03, 2013 at 03:10:48PM +0800, Xiao Guangrong wrote:
> On 11/28/2013 04:53 PM, Xiao Guangrong wrote:
> > On 11/27/2013 03:31 AM, Marcelo Tosatti wrote:
> >> On Tue, Nov 26, 2013 at 11:21:37AM +0800, Xiao Guangrong wrote:
> >>> On 11/26/2013 02:12 AM, Marcelo Tosatti wrote:
> On
GOn Tue, Dec 03, 2013 at 03:10:48PM +0800, Xiao Guangrong wrote:
On 11/28/2013 04:53 PM, Xiao Guangrong wrote:
On 11/27/2013 03:31 AM, Marcelo Tosatti wrote:
On Tue, Nov 26, 2013 at 11:21:37AM +0800, Xiao Guangrong wrote:
On 11/26/2013 02:12 AM, Marcelo Tosatti wrote:
On Mon, Nov 25, 2013
On Dec 5, 2013, at 9:50 PM, Marcelo Tosatti mtosa...@redhat.com wrote:
GOn Tue, Dec 03, 2013 at 03:10:48PM +0800, Xiao Guangrong wrote:
On 11/28/2013 04:53 PM, Xiao Guangrong wrote:
On 11/27/2013 03:31 AM, Marcelo Tosatti wrote:
On Tue, Nov 26, 2013 at 11:21:37AM +0800, Xiao Guangrong wrote:
On Thu, Dec 05, 2013 at 11:30:27PM +0800, Xiao Guangrong wrote:
Is it not the case that simply moving to the slow path once a maximum of
rewalks has been reached enough? (looks a like a good solution).
In some cases, the lockless walker will do endless-walking on desc and
without rewalk,
On Thu, Dec 05, 2013 at 11:30:27PM +0800, Xiao Guangrong wrote:
In some cases, the lockless walker will do endless-walking on desc and
without rewalk, consider this case:
there are two descs: desc1 and desc2 who is pointed by desc1-next:
desc1-next = desc2.
CPU 0
On 11/28/2013 04:53 PM, Xiao Guangrong wrote:
> On 11/27/2013 03:31 AM, Marcelo Tosatti wrote:
>> On Tue, Nov 26, 2013 at 11:21:37AM +0800, Xiao Guangrong wrote:
>>> On 11/26/2013 02:12 AM, Marcelo Tosatti wrote:
On Mon, Nov 25, 2013 at 02:29:03PM +0800, Xiao Guangrong wrote:
>>> Also,
On 11/28/2013 04:53 PM, Xiao Guangrong wrote:
On 11/27/2013 03:31 AM, Marcelo Tosatti wrote:
On Tue, Nov 26, 2013 at 11:21:37AM +0800, Xiao Guangrong wrote:
On 11/26/2013 02:12 AM, Marcelo Tosatti wrote:
On Mon, Nov 25, 2013 at 02:29:03PM +0800, Xiao Guangrong wrote:
Also, there is no
On 11/27/2013 03:31 AM, Marcelo Tosatti wrote:
> On Tue, Nov 26, 2013 at 11:21:37AM +0800, Xiao Guangrong wrote:
>> On 11/26/2013 02:12 AM, Marcelo Tosatti wrote:
>>> On Mon, Nov 25, 2013 at 02:29:03PM +0800, Xiao Guangrong wrote:
>> Also, there is no guarantee of termination (as long as sptes
On 11/27/2013 03:58 AM, Marcelo Tosatti wrote:
> On Tue, Nov 26, 2013 at 11:10:19AM +0800, Xiao Guangrong wrote:
>> On 11/25/2013 10:23 PM, Marcelo Tosatti wrote:
>>> On Mon, Nov 25, 2013 at 02:48:37PM +0200, Avi Kivity wrote:
On Mon, Nov 25, 2013 at 8:11 AM, Xiao Guangrong
wrote:
>
On 11/27/2013 03:58 AM, Marcelo Tosatti wrote:
On Tue, Nov 26, 2013 at 11:10:19AM +0800, Xiao Guangrong wrote:
On 11/25/2013 10:23 PM, Marcelo Tosatti wrote:
On Mon, Nov 25, 2013 at 02:48:37PM +0200, Avi Kivity wrote:
On Mon, Nov 25, 2013 at 8:11 AM, Xiao Guangrong
On 11/27/2013 03:31 AM, Marcelo Tosatti wrote:
On Tue, Nov 26, 2013 at 11:21:37AM +0800, Xiao Guangrong wrote:
On 11/26/2013 02:12 AM, Marcelo Tosatti wrote:
On Mon, Nov 25, 2013 at 02:29:03PM +0800, Xiao Guangrong wrote:
Also, there is no guarantee of termination (as long as sptes are
On Tue, Nov 26, 2013 at 11:21:37AM +0800, Xiao Guangrong wrote:
> On 11/26/2013 02:12 AM, Marcelo Tosatti wrote:
> > On Mon, Nov 25, 2013 at 02:29:03PM +0800, Xiao Guangrong wrote:
> Also, there is no guarantee of termination (as long as sptes are
> deleted with the correct timing). BTW,
On Tue, Nov 26, 2013 at 11:10:19AM +0800, Xiao Guangrong wrote:
> On 11/25/2013 10:23 PM, Marcelo Tosatti wrote:
> > On Mon, Nov 25, 2013 at 02:48:37PM +0200, Avi Kivity wrote:
> >> On Mon, Nov 25, 2013 at 8:11 AM, Xiao Guangrong
> >> wrote:
> >>>
> >>> On Nov 23, 2013, at 3:14 AM, Marcelo
On Tue, Nov 26, 2013 at 11:10:19AM +0800, Xiao Guangrong wrote:
> On 11/25/2013 10:23 PM, Marcelo Tosatti wrote:
> > On Mon, Nov 25, 2013 at 02:48:37PM +0200, Avi Kivity wrote:
> >> On Mon, Nov 25, 2013 at 8:11 AM, Xiao Guangrong
> >> wrote:
> >>>
> >>> On Nov 23, 2013, at 3:14 AM, Marcelo
On Tue, Nov 26, 2013 at 11:21:37AM +0800, Xiao Guangrong wrote:
> On 11/26/2013 02:12 AM, Marcelo Tosatti wrote:
> > On Mon, Nov 25, 2013 at 02:29:03PM +0800, Xiao Guangrong wrote:
> Also, there is no guarantee of termination (as long as sptes are
> deleted with the correct timing). BTW,
On Tue, Nov 26, 2013 at 11:21:37AM +0800, Xiao Guangrong wrote:
On 11/26/2013 02:12 AM, Marcelo Tosatti wrote:
On Mon, Nov 25, 2013 at 02:29:03PM +0800, Xiao Guangrong wrote:
Also, there is no guarantee of termination (as long as sptes are
deleted with the correct timing). BTW, can't see
On Tue, Nov 26, 2013 at 11:10:19AM +0800, Xiao Guangrong wrote:
On 11/25/2013 10:23 PM, Marcelo Tosatti wrote:
On Mon, Nov 25, 2013 at 02:48:37PM +0200, Avi Kivity wrote:
On Mon, Nov 25, 2013 at 8:11 AM, Xiao Guangrong
xiaoguangr...@linux.vnet.ibm.com wrote:
On Nov 23, 2013, at 3:14 AM,
On Tue, Nov 26, 2013 at 11:10:19AM +0800, Xiao Guangrong wrote:
On 11/25/2013 10:23 PM, Marcelo Tosatti wrote:
On Mon, Nov 25, 2013 at 02:48:37PM +0200, Avi Kivity wrote:
On Mon, Nov 25, 2013 at 8:11 AM, Xiao Guangrong
xiaoguangr...@linux.vnet.ibm.com wrote:
On Nov 23, 2013, at 3:14 AM,
On Tue, Nov 26, 2013 at 11:21:37AM +0800, Xiao Guangrong wrote:
On 11/26/2013 02:12 AM, Marcelo Tosatti wrote:
On Mon, Nov 25, 2013 at 02:29:03PM +0800, Xiao Guangrong wrote:
Also, there is no guarantee of termination (as long as sptes are
deleted with the correct timing). BTW, can't see
On 11/26/2013 02:12 AM, Marcelo Tosatti wrote:
> On Mon, Nov 25, 2013 at 02:29:03PM +0800, Xiao Guangrong wrote:
Also, there is no guarantee of termination (as long as sptes are
deleted with the correct timing). BTW, can't see any guarantee of
termination for rculist nulls either (a
On 11/25/2013 10:23 PM, Marcelo Tosatti wrote:
> On Mon, Nov 25, 2013 at 02:48:37PM +0200, Avi Kivity wrote:
>> On Mon, Nov 25, 2013 at 8:11 AM, Xiao Guangrong
>> wrote:
>>>
>>> On Nov 23, 2013, at 3:14 AM, Marcelo Tosatti wrote:
>>
>>
>>
>> I'm not really following, but note that parent_pte
On 11/25/2013 10:08 PM, Marcelo Tosatti wrote:
> On Mon, Nov 25, 2013 at 02:11:31PM +0800, Xiao Guangrong wrote:
>>
>> On Nov 23, 2013, at 3:14 AM, Marcelo Tosatti wrote:
>>
>>> On Wed, Oct 23, 2013 at 09:29:25PM +0800, Xiao Guangrong wrote:
It likes nulls list and we use the pte-list as the
On Mon, Nov 25, 2013 at 02:29:03PM +0800, Xiao Guangrong wrote:
> >> Also, there is no guarantee of termination (as long as sptes are
> >> deleted with the correct timing). BTW, can't see any guarantee of
> >> termination for rculist nulls either (a writer can race with a lockless
> >> reader
GOn Mon, Nov 25, 2013 at 04:29:28PM +0200, Gleb Natapov wrote:
> On Mon, Nov 25, 2013 at 12:23:51PM -0200, Marcelo Tosatti wrote:
> > On Mon, Nov 25, 2013 at 02:48:37PM +0200, Avi Kivity wrote:
> > > On Mon, Nov 25, 2013 at 8:11 AM, Xiao Guangrong
> > > wrote:
> > > >
> > > > On Nov 23, 2013, at
On Mon, Nov 25, 2013 at 12:23:51PM -0200, Marcelo Tosatti wrote:
> On Mon, Nov 25, 2013 at 02:48:37PM +0200, Avi Kivity wrote:
> > On Mon, Nov 25, 2013 at 8:11 AM, Xiao Guangrong
> > wrote:
> > >
> > > On Nov 23, 2013, at 3:14 AM, Marcelo Tosatti wrote:
> >
> >
> >
> > I'm not really
On Mon, Nov 25, 2013 at 02:48:37PM +0200, Avi Kivity wrote:
> On Mon, Nov 25, 2013 at 8:11 AM, Xiao Guangrong
> wrote:
> >
> > On Nov 23, 2013, at 3:14 AM, Marcelo Tosatti wrote:
>
>
>
> I'm not really following, but note that parent_pte predates EPT (and
> the use of rcu in kvm), so all the
On Mon, Nov 25, 2013 at 02:11:31PM +0800, Xiao Guangrong wrote:
>
> On Nov 23, 2013, at 3:14 AM, Marcelo Tosatti wrote:
>
> > On Wed, Oct 23, 2013 at 09:29:25PM +0800, Xiao Guangrong wrote:
> >> It likes nulls list and we use the pte-list as the nulls which can help us
> >> to
> >> detect
On Mon, Nov 25, 2013 at 8:11 AM, Xiao Guangrong
wrote:
>
> On Nov 23, 2013, at 3:14 AM, Marcelo Tosatti wrote:
I'm not really following, but note that parent_pte predates EPT (and
the use of rcu in kvm), so all the complexity that is the result of
trying to pack as many list entries into a
On Mon, Nov 25, 2013 at 12:05:00PM +0100, Peter Zijlstra wrote:
> On Mon, Nov 25, 2013 at 06:59:24PM +0800, Xiao Guangrong wrote:
> > I guess Marcelo was talking about rculist_nulls.h
> > (Documentation/RCU/rculist_nulls.txt).
>
> Oh, let me have a look, I don't think I've really looked at that
On Mon, Nov 25, 2013 at 06:59:24PM +0800, Xiao Guangrong wrote:
>
> Hi Peter,
>
> On 11/25/2013 05:31 PM, Peter Zijlstra wrote:
> > On Fri, Nov 22, 2013 at 05:14:29PM -0200, Marcelo Tosatti wrote:
> >> Also, there is no guarantee of termination (as long as sptes are
> >> deleted with the correct
Hi Peter,
On 11/25/2013 05:31 PM, Peter Zijlstra wrote:
> On Fri, Nov 22, 2013 at 05:14:29PM -0200, Marcelo Tosatti wrote:
>> Also, there is no guarantee of termination (as long as sptes are
>> deleted with the correct timing). BTW, can't see any guarantee of
>> termination for rculist nulls
On 11/25/2013 06:19 PM, Gleb Natapov wrote:
> On Mon, Nov 25, 2013 at 02:11:31PM +0800, Xiao Guangrong wrote:
>>>
>>> For example, nothing prevents lockless walker to move into some
>>> parent_ptes chain, right?
>>
>> No.
>>
>> The nulls can help us to detect this case, for parent_ptes, the nulls
On Mon, Nov 25, 2013 at 02:11:31PM +0800, Xiao Guangrong wrote:
> >
> > For example, nothing prevents lockless walker to move into some
> > parent_ptes chain, right?
>
> No.
>
> The nulls can help us to detect this case, for parent_ptes, the nulls points
> to "shadow page" but for rmaps, the
On Fri, Nov 22, 2013 at 05:14:29PM -0200, Marcelo Tosatti wrote:
> Also, there is no guarantee of termination (as long as sptes are
> deleted with the correct timing). BTW, can't see any guarantee of
> termination for rculist nulls either (a writer can race with a lockless
> reader indefinately,
On Fri, Nov 22, 2013 at 05:14:29PM -0200, Marcelo Tosatti wrote:
Also, there is no guarantee of termination (as long as sptes are
deleted with the correct timing). BTW, can't see any guarantee of
termination for rculist nulls either (a writer can race with a lockless
reader indefinately,
On Mon, Nov 25, 2013 at 02:11:31PM +0800, Xiao Guangrong wrote:
For example, nothing prevents lockless walker to move into some
parent_ptes chain, right?
No.
The nulls can help us to detect this case, for parent_ptes, the nulls points
to shadow page but for rmaps, the nulls points to
On 11/25/2013 06:19 PM, Gleb Natapov wrote:
On Mon, Nov 25, 2013 at 02:11:31PM +0800, Xiao Guangrong wrote:
For example, nothing prevents lockless walker to move into some
parent_ptes chain, right?
No.
The nulls can help us to detect this case, for parent_ptes, the nulls points
to shadow
Hi Peter,
On 11/25/2013 05:31 PM, Peter Zijlstra wrote:
On Fri, Nov 22, 2013 at 05:14:29PM -0200, Marcelo Tosatti wrote:
Also, there is no guarantee of termination (as long as sptes are
deleted with the correct timing). BTW, can't see any guarantee of
termination for rculist nulls either (a
On Mon, Nov 25, 2013 at 06:59:24PM +0800, Xiao Guangrong wrote:
Hi Peter,
On 11/25/2013 05:31 PM, Peter Zijlstra wrote:
On Fri, Nov 22, 2013 at 05:14:29PM -0200, Marcelo Tosatti wrote:
Also, there is no guarantee of termination (as long as sptes are
deleted with the correct timing).
On Mon, Nov 25, 2013 at 12:05:00PM +0100, Peter Zijlstra wrote:
On Mon, Nov 25, 2013 at 06:59:24PM +0800, Xiao Guangrong wrote:
I guess Marcelo was talking about rculist_nulls.h
(Documentation/RCU/rculist_nulls.txt).
Oh, let me have a look, I don't think I've really looked at that yet.
On Mon, Nov 25, 2013 at 8:11 AM, Xiao Guangrong
xiaoguangr...@linux.vnet.ibm.com wrote:
On Nov 23, 2013, at 3:14 AM, Marcelo Tosatti mtosa...@redhat.com wrote:
snip complicated stuff about parent_pte
I'm not really following, but note that parent_pte predates EPT (and
the use of rcu in kvm),
On Mon, Nov 25, 2013 at 02:48:37PM +0200, Avi Kivity wrote:
On Mon, Nov 25, 2013 at 8:11 AM, Xiao Guangrong
xiaoguangr...@linux.vnet.ibm.com wrote:
On Nov 23, 2013, at 3:14 AM, Marcelo Tosatti mtosa...@redhat.com wrote:
snip complicated stuff about parent_pte
I'm not really following,
On Mon, Nov 25, 2013 at 02:11:31PM +0800, Xiao Guangrong wrote:
On Nov 23, 2013, at 3:14 AM, Marcelo Tosatti mtosa...@redhat.com wrote:
On Wed, Oct 23, 2013 at 09:29:25PM +0800, Xiao Guangrong wrote:
It likes nulls list and we use the pte-list as the nulls which can help us
to
detect
On Mon, Nov 25, 2013 at 12:23:51PM -0200, Marcelo Tosatti wrote:
On Mon, Nov 25, 2013 at 02:48:37PM +0200, Avi Kivity wrote:
On Mon, Nov 25, 2013 at 8:11 AM, Xiao Guangrong
xiaoguangr...@linux.vnet.ibm.com wrote:
On Nov 23, 2013, at 3:14 AM, Marcelo Tosatti mtosa...@redhat.com wrote:
GOn Mon, Nov 25, 2013 at 04:29:28PM +0200, Gleb Natapov wrote:
On Mon, Nov 25, 2013 at 12:23:51PM -0200, Marcelo Tosatti wrote:
On Mon, Nov 25, 2013 at 02:48:37PM +0200, Avi Kivity wrote:
On Mon, Nov 25, 2013 at 8:11 AM, Xiao Guangrong
xiaoguangr...@linux.vnet.ibm.com wrote:
On
On Mon, Nov 25, 2013 at 02:29:03PM +0800, Xiao Guangrong wrote:
Also, there is no guarantee of termination (as long as sptes are
deleted with the correct timing). BTW, can't see any guarantee of
termination for rculist nulls either (a writer can race with a lockless
reader indefinately,
On 11/25/2013 10:08 PM, Marcelo Tosatti wrote:
On Mon, Nov 25, 2013 at 02:11:31PM +0800, Xiao Guangrong wrote:
On Nov 23, 2013, at 3:14 AM, Marcelo Tosatti mtosa...@redhat.com wrote:
On Wed, Oct 23, 2013 at 09:29:25PM +0800, Xiao Guangrong wrote:
It likes nulls list and we use the pte-list
On 11/25/2013 10:23 PM, Marcelo Tosatti wrote:
On Mon, Nov 25, 2013 at 02:48:37PM +0200, Avi Kivity wrote:
On Mon, Nov 25, 2013 at 8:11 AM, Xiao Guangrong
xiaoguangr...@linux.vnet.ibm.com wrote:
On Nov 23, 2013, at 3:14 AM, Marcelo Tosatti mtosa...@redhat.com wrote:
snip complicated stuff
On 11/26/2013 02:12 AM, Marcelo Tosatti wrote:
On Mon, Nov 25, 2013 at 02:29:03PM +0800, Xiao Guangrong wrote:
Also, there is no guarantee of termination (as long as sptes are
deleted with the correct timing). BTW, can't see any guarantee of
termination for rculist nulls either (a writer can
On 11/25/2013 02:11 PM, Xiao Guangrong wrote:
>
> On Nov 23, 2013, at 3:14 AM, Marcelo Tosatti wrote:
>
>> On Wed, Oct 23, 2013 at 09:29:25PM +0800, Xiao Guangrong wrote:
>>> It likes nulls list and we use the pte-list as the nulls which can help us
>>> to
>>> detect whether the "desc" is
On Nov 23, 2013, at 3:14 AM, Marcelo Tosatti wrote:
> On Wed, Oct 23, 2013 at 09:29:25PM +0800, Xiao Guangrong wrote:
>> It likes nulls list and we use the pte-list as the nulls which can help us to
>> detect whether the "desc" is moved to anther rmap then we can re-walk the
>> rmap
>> if that
On Nov 23, 2013, at 3:14 AM, Marcelo Tosatti mtosa...@redhat.com wrote:
On Wed, Oct 23, 2013 at 09:29:25PM +0800, Xiao Guangrong wrote:
It likes nulls list and we use the pte-list as the nulls which can help us to
detect whether the desc is moved to anther rmap then we can re-walk the
rmap
On 11/25/2013 02:11 PM, Xiao Guangrong wrote:
On Nov 23, 2013, at 3:14 AM, Marcelo Tosatti mtosa...@redhat.com wrote:
On Wed, Oct 23, 2013 at 09:29:25PM +0800, Xiao Guangrong wrote:
It likes nulls list and we use the pte-list as the nulls which can help us
to
detect whether the desc is
On Wed, Oct 23, 2013 at 09:29:25PM +0800, Xiao Guangrong wrote:
> It likes nulls list and we use the pte-list as the nulls which can help us to
> detect whether the "desc" is moved to anther rmap then we can re-walk the rmap
> if that happened
>
> kvm->slots_lock is held when we do lockless
On Wed, Oct 23, 2013 at 09:29:25PM +0800, Xiao Guangrong wrote:
It likes nulls list and we use the pte-list as the nulls which can help us to
detect whether the desc is moved to anther rmap then we can re-walk the rmap
if that happened
kvm-slots_lock is held when we do lockless walking that
It likes nulls list and we use the pte-list as the nulls which can help us to
detect whether the "desc" is moved to anther rmap then we can re-walk the rmap
if that happened
kvm->slots_lock is held when we do lockless walking that prevents rmap
is reused (free rmap need to hold that lock) so that
It likes nulls list and we use the pte-list as the nulls which can help us to
detect whether the desc is moved to anther rmap then we can re-walk the rmap
if that happened
kvm-slots_lock is held when we do lockless walking that prevents rmap
is reused (free rmap need to hold that lock) so that we
62 matches
Mail list logo