Avi Kivity wrote:
> I'd hoped to release the first version of virtio guest drivers with
> kvm-60, but that is delayed, and enough changes have accumulated. Most
> notable is virtio host support and Vista x64 on AMD hosts.
>
> Changes from kvm-59:
>
I missed:
- fix unbounded latency due to
Chris Wright wrote:
> * Izik Eidus ([EMAIL PROTECTED]) wrote:
>
> Just a bunch of nitpicks.
>
>
>> struct ksm *ksm;
>>
>
> static
>
>
yes
>> static int page_hash_size = 0;
>>
>
> no need to initialize to zero
>
ok
>
>> module_param(page_hash_size, int, 0);
>> MODULE_PARM_DE
Avi Kivity wrote:
> Chris Wright wrote:
>> * Izik Eidus ([EMAIL PROTECTED]) wrote:
>>
>>> this module find this identical data (pages) and merge them into one
>>> single page
>>> this new page is write protected so in any case the guest will try
>>> to write to it do_wp_page will duplicate the
On Wednesday 23 January 2008 17:46:22 Avi Kivity wrote:
> Yang, Sheng wrote:
> > On Tuesday 22 January 2008 21:54:10 Avi Kivity wrote:
> >> Yang, Sheng wrote:
> >>> +
> >>> +static int pit_get_out(struct kvm *kvm, int channel)
> >>> +{
> >>> + struct PITChannelState *c =
> >>> + &kvm->arch.
Izik Eidus wrote:
>>
>>> struct ksm *ksm;
>>>
>>
>> static
>>
>>
>
> yes
Actually the entire contents of 'struct ksm' should be module static
variables.
--
Any sufficiently difficult bug is indistinguishable from a feature.
--
Yang, Sheng wrote:
> I updated patch using ktime_get() (Is it too heavy?).
>
>
Shouldn't be. At least on modern hosts.
> And currently all the muldiv64() is used to divide 1e9, which can be
> insteaded
> by ">> 30" approximately. What's your opinion?
>
Accuracy is more important. And m
Avi Kivity wrote:
> Izik Eidus wrote:
>>>
struct ksm *ksm;
>>>
>>> static
>>>
>>>
>>
>> yes
>
>
> Actually the entire contents of 'struct ksm' should be module static
> variables.
>
i agree with you, i will fix this as well
--
woof.
-
Yang, Sheng a écrit :
> From 39551508c869f678f028655c062b3cb5fe4715bc Mon Sep 17 00:00:00 2001
> From: Sheng Yang <[EMAIL PROTECTED]>
> Date: Thu, 17 Jan 2008 15:14:33 +0800
> Subject: [PATCH] KVM: VMX: Enable Virtual Processor Identification (VPID)
>
> To allow TLB entries to be retained across V
Chris Lalancette wrote:
> All,
> Attached is a fairly simple patch to the migration code to make it handle
> errors better. In particular, if the remote side wasn't set up properly (i.e.
> the memory size didn't match), the migration would fail, but the source host
> didn't actually react p
Yang, Sheng wrote:
> To allow TLB entries to be retained across VM entry and VM exit, the VMM can
> now identify distinct address spaces through a new virtual-processor ID
> (VPID) field of the VMCS.
>
>
> +#define VPID_BITMAP_SIZE (1 << 16)*BITS_PER_BYTE
>
This is str
Chris Lalancette wrote:
> All,
> I've been doing some migration testing of KVM guests, and have been
> running
> into some problems. Let me describe the setup and what I've tried, and maybe
> somebody has some ideas about what might be going on here.
>
> When I try to do
> live migration thi
On Thu, Jan 24, 2008 at 07:56:43AM +0200, Avi Kivity wrote:
> What I need is a list of (mm, va) that map the page. kvm doesn't have
> access to that, export notifiers do. It seems reasonable that export
> notifier do that rmap walk since they are part of core mm, not kvm.
Yes. Like said in ear
Andrea Arcangeli wrote:
> On Thu, Jan 24, 2008 at 07:56:43AM +0200, Avi Kivity wrote:
>
>> What I need is a list of (mm, va) that map the page. kvm doesn't have
>> access to that, export notifiers do. It seems reasonable that export
>> notifier do that rmap walk since they are part of core m
On Tue, Jan 22, 2008 at 01:12:12PM +0100, Jan Kiszka wrote:
> Guido Guenther wrote:
> > On Tue, Jan 22, 2008 at 11:11:00AM +0100, Jan Kiszka wrote:
> >> #if defined(TARGET_I386)
> >> void qemu_system_powerdown(void)
> >> {
> >> +if (!pm_state)
> >> +exit(0);
> >> if(pm_state->pm
Guido Guenther wrote:
> If the guest can't do that, it simply can't. You can then SIGTERM the
> guest after a timeout. Note that even an acpi system might not be able
> to shut down since it hangs on a network unmount or it ignores the ACPI
> interrupt via kernel commandline or whatever. You can't
Avi Kivity wrote:
> Agree, but should try a "quit" monitor command first. Signals are racy,
> like anything that deals with pids (qemu dies, another process is
> fork()ed with the same pid, libvirt kills it).
There is no race in that specific case because qemu is started by
libvirtd. libvirtd
Gerd Hoffmann wrote:
> Avi Kivity wrote:
>
>> Agree, but should try a "quit" monitor command first. Signals are racy,
>> like anything that deals with pids (qemu dies, another process is
>> fork()ed with the same pid, libvirt kills it).
>>
>
> There is no race in that specific case becau
Avi Kivity wrote:
> Too much Kconfig for this. Since I want to do some Kconfig changes
> anyway, I'll do this another way.
I think you overlooked something here. Actually this should change too
if we don't have pio:
-...case KVM_GET_VCPU_MMAP_SIZE:
-...-...r = -EINVAL;
-...-.
The patches really fix/simplify things :)
Did you test device open->close->open? It was broken in my last test,
we cad add link indications too.
Besides that I also remembered that adding 'vq->num_free =vq->ring.num;'
after reset fixed part of the problems, I don't remember why (gonna
re-check it
Carsten Otte wrote:
> Avi Kivity wrote:
>> Too much Kconfig for this. Since I want to do some Kconfig changes
>> anyway, I'll do this another way.
> I think you overlooked something here. Actually this should change too
> if we don't have pio:
> -...case KVM_GET_VCPU_MMAP_SIZE:
> -...-..
On Wed, Jan 23, 2008 at 12:18:45PM -0800, Christoph Lameter wrote:
> On Wed, 23 Jan 2008, Andrea Arcangeli wrote:
>
> > > [..] The linux instance with the
> > > secondary mmu must call back to the exporting machine in order to
> > > reinstantiate the page. PageExported is cleared in invalidate_p
On Thu, Jan 24, 2008 at 03:34:54PM +0100, Andrea Arcangeli wrote:
> set_page_dirty can be called inside ->invalidate_page if needed. But
> I'm not against artificially setting the dirty bit of the pteval
> returned by set_page_dirty, perhaps that's more efficient as it will
^^
Andrea Arcangeli wrote:
> The remote page fault
>
As we have two names for this ('shadow' and 'remote/export') I'd like to
suggest a neutral nomenclature. How about 'secondary mmu' (and
secondary ptes (sptes), etc.)? I think this fits xpmem, kvm, rdma, and dri.
--
Any sufficiently difficu
Avi Kivity wrote:
> Andrea Arcangeli wrote:
>
>> The remote page fault
>>
>>
>
> As we have two names for this ('shadow' and 'remote/export') I'd like to
> suggest a neutral nomenclature. How about 'secondary mmu' (and
> secondary ptes (sptes), etc.)? I think this fits xpmem, kvm, rd
On Wed, Jan 23, 2008 at 12:27:47PM -0800, Christoph Lameter wrote:
> There are still dirty bit issues.
Yes, but no big issues given ->invalidate_page is fully capable of
running set_page_dirty if needed.
> > The window that you must close with that bitflag is the request coming
> > from the remot
Hi,
please find the following trivial warning fix patch for
kvm-60/user/test/x86/*
below. Feel free to merge to upstream kvm. It's diffed against kvm-60.
The patch does not change things, just shuts up the warnings.
The deal here is that we check for warnings and are alert
Bernhard Kaindl wrote:
> I did not test this patch as I did not find documentation on how to run the
> test cases and I could not find a make target to run them from make.
>
>
make -C user test_cases
user/kvmctl user/test/x86/bootstrap user/test/x86/access.flat
(we should rename user -> tes
Hi,
Trying to browse the list archives, but both gmane and sourcefrog's
interfaces are really painful to deal with.
So, any chance someone could point me at the current ia64 KVM kernel
patches? I notice they are not yet in Avi's tree.
Cheers,
Jes
On Thu, 24 Jan 2008, Avi Kivity wrote:
>
> make -C user test_cases
> user/kvmctl user/test/x86/bootstrap user/test/x86/access.flat
Ah, thanks!
> > -smp_init(ac_test_run);
> > +smp_init((void (*)(void))ac_test_run);
>
> Better to add a wrapper that conforms to the expected signature, an
* Izik Eidus ([EMAIL PROTECTED]) wrote:
> Chris Wright wrote:
>>> list_add_tail(&slot->link, &ksm->slots);
>>
>> slots_lock?
>
> good catch, i forgat to put it here
>
>>> list_add_tail(&slot->sma_link, &ksm_sma->sma_slots);
>>>
>>
>> theoretical, but if threaded, registering program co
* Avi Kivity ([EMAIL PROTECTED]) wrote:
> Actually the entire contents of 'struct ksm' should be module static
> variables.
Yeah, I agree.
-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visu
* Jes Sorensen ([EMAIL PROTECTED]) wrote:
> Trying to browse the list archives, but both gmane and sourcefrog's
> interfaces are really painful to deal with.
>
> So, any chance someone could point me at the current ia64 KVM kernel
> patches? I notice they are not yet in Avi's tree.
The last full
Uri Lublin wrote:
>
> Chris Lalancette wrote:
>> All,
>> Attached is a fairly simple patch to the migration code to make it
>> handle
>> errors better. In particular, if the remote side wasn't set up properly
>> (i.e.
>> the memory size didn't match), the migration would fail, but the sour
Chris Wright wrote:
>>> struct ksm_memory_region {
>>> __u32 npages; /* number of pages to share */
>>> __u64 addr; /* the begining of the virtual address */
>>> };
>>>
>> why isnt it compat safe?
>>
>
> 32-bit has more relaxed alignment requirement for __u64 (4 bytes)
> than
Uri Lublin wrote:
> Patch looks good.
> Why did you define MIG_STAT_DIRTY_TRACK_FAIL and not
> MIG_STAT_KVM_SET_DIRTY_TRACKING_FAILED ?
Oops, I slightly misunderstood this bit in my last e-mail. You were saying that
there is already a "MIG_STAT_KVM_SET_DIRTY_TRACKING_FAILED" error flag. In the
On Thu, 24 Jan 2008, Andrea Arcangeli wrote:
> > SetPageExported is set when a remote instance of linux establishes a
> > reference to the page (a kind of remote page fault). In the KVM scenario
> > that would occur when memory is made available.
>
> The remote page fault is exactly the thing t
This patch updates KVM's virtio implementation to the latest virtio ABI. This
includes a change in the network header and support for reset.
With this patch, the block and network driver from Rusty's latest queue are
functioning. Module unload and reload work and I no longer see an error when
re
On Thu, 24 Jan 2008, Andrea Arcangeli wrote:
> I think you should consider if you can also build a rmap per-MM like
> KVM does and index it by the virtual address like KVM does.
Yes we have that.
If we have that then we do not need the mmu_notifier.
We could call it with a page parameter and th
* Avi Kivity ([EMAIL PROTECTED]) wrote:
> Chris Wright wrote:
struct ksm_memory_region {
__u32 npages; /* number of pages to share */
__u64 addr; /* the begining of the virtual address */
};
>>> why isnt it compat safe?
>>
>> 32-bit has more relaxed alignment
On Friday 25 January 2008 00:09:07 Dor Laor wrote:
> The patches really fix/simplify things :)
Yes, that's why I like reset, it solves multiple problems.
> Did you test device open->close->open? It was broken in my last test,
Well, it's not really fair, since I moved the buffer allocation and
d
This patch adds support to QEMU for Rusty's recently introduce virtio balloon
driver. The user-facing portions of this are the introduction of a "balloon"
and "info balloon" command in the monitor.
I think using madvise unconditionally is okay but I am not sure. If madvise
is called on memory th
Anthony Liguori wrote:
> This patch adds support to QEMU for Rusty's recently introduce virtio balloon
> driver. The user-facing portions of this are the introduction of a "balloon"
> and "info balloon" command in the monitor.
>
> I think using madvise unconditionally is okay but I am not sure.
L
Hi, Jes
Yes, Anthony and I are working with kernel-ia64 and kvm community to push the
patches. Since kernel should export some interface for kvm use, we have to wait
the response from kernel-ia64. But anyway, It should be picked up in near
future. :) Compared with last push, we added smp guest
On Thursday 24 January 2008 19:15:57 Aurelien Jarno wrote:
> Yang, Sheng a écrit :
> > From 39551508c869f678f028655c062b3cb5fe4715bc Mon Sep 17 00:00:00 2001
> > From: Sheng Yang <[EMAIL PROTECTED]>
> > Date: Thu, 17 Jan 2008 15:14:33 +0800
> > Subject: [PATCH] KVM: VMX: Enable Virtual Processor Id
On Thursday 24 January 2008 19:50:20 Avi Kivity wrote:
> Yang, Sheng wrote:
> > To allow TLB entries to be retained across VM entry and VM exit, the VMM
> > can now identify distinct address spaces through a new virtual-processor
> > ID (VPID) field of the VMCS.
> >
> >
> >
> > +#define VPID_BITMAP
Hi Guys:
My System:
Linux jerinkturion 2.6.20-gentoo-r7 #20 SMP PREEMPT Thu Jan 24 14:34:23 SGT
2008 x86_64 AMD Turion(tm) 64 X2 Mobile Technology TL-60 AuthenticAMD GNU/Linux
Acer Ferarri 5000 Laptop.
I was previously runng qemu 0.9.0 with kqemu to run Windows XP-SP2.
I installed KVM-59-r1 fro
Core code for mmu notifiers.
Signed-off-by: Christoph Lameter <[EMAIL PROTECTED]>
Signed-off-by: Andrea Arcangeli <[EMAIL PROTECTED]>
---
include/linux/mm_types.h |8 ++
include/linux/mmu_notifier.h | 152 +++
include/linux/page-flags.h |9 +
The invalidation of address ranges in a mm_struct needs to be
performed when pages are removed or permissions etc change.
invalidate_range() is generally called with mmap_sem held but
no spinlocks are active.
Exceptions:
We hold i_mmap_lock in __unmap_hugepage_range and
sometimes in zap_page_ran
These notifiers here use the Linux rmaps to perform the callbacks.
In order to walk the rmaps locks must be held. Callbacks can therefore
only operate in an atomic context.
Signed-off-by: Christoph Lameter <[EMAIL PROTECTED]>
---
mm/filemap_xip.c |1 +
mm/fremap.c |1 +
mm/memory.c
Callbacks to remove individual pages if the subsystem has an
rmap capability. The pagelock is held but no spinlocks are held.
The refcount of the page is elevated so that dropping the refcount
in the subsystem will not directly free the page.
The callbacks occur after the Linux rmaps have been wal
This is a patchset implementing MMU notifier callbacks based on Andrea's
earlier work. These are needed if Linux pages are referenced from something
else than tracked by the rmaps of the kernel.
To do:
- Make locking requirements for the callbacks consistent and
document them accurately.
- Insu
Christoph Lameter wrote:
> On Thu, 24 Jan 2008, Andrea Arcangeli wrote:
>
>
>> I think you should consider if you can also build a rmap per-MM like
>> KVM does and index it by the virtual address like KVM does.
>>
>
> Yes we have that.
>
> If we have that then we do not need the mmu_notifie
Current Physicians in the USA
788,008 in total <> 17,377 emails
Physicians in many different specialties
Sort by over a dozen different fields
Now offered at the lower rate: $393
=== You get these for F-Ree with every order this week ===
<> US Pharmaceutical Company Executives
Yang, Sheng wrote:
> Here is the updated patch:
>
>
Applied, thanks.
> @@ -1714,6 +1788,8 @@ static int vmx_vcpu_reset(struct kvm_vcpu *vcpu)
> vmx_fpu_activate(&vmx->vcpu);
> update_exception_bitmap(&vmx->vcpu);
>
> + vpid_sync_all();
> +
> return 0;
>
>
This cau
On Friday 25 January 2008 14:52:47 Avi Kivity wrote:
> Yang, Sheng wrote:
> > Here is the updated patch:
>
> Applied, thanks.
>
> > @@ -1714,6 +1788,8 @@ static int vmx_vcpu_reset(struct kvm_vcpu *vcpu)
> > vmx_fpu_activate(&vmx->vcpu);
> > update_exception_bitmap(&vmx->vcpu);
> >
> > + v
Yang, Sheng wrote:
> I think it's OK for there is a judgment in __invvpid() to see if machine has
> the ability(also if it is allowed to using VPID). :)
>
Oh right, I missed that.
We can remove it now since it will only be called if vpid != 0, and that
happens only on vpid enabled machines,
56 matches
Mail list logo
