Andrea Arcangeli wrote:
On Tue, Jan 22, 2008 at 04:38:49PM +0200, Avi Kivity wrote:
Andrea Arcangeli wrote:
This is arch independent code, I'm surprised mmu_lock is visible here?
The mmu_lock is arch independent as far as I can tell. Pretty much
like the mm-page_table_lock is also
Zhao, Yunfeng wrote:
Hi, all,
We did a test on KVM60RC1, and 7 issues have been found in the test.
One issue fixed:
1. Fails to boot 64bit guest with 4BG mem --- Has already been fixed
by 29ae41a859a12aa71baacd35a9648511f335848e.
3. With using qcow images xpsp2 with 2vpus may fail to boot
https://sourceforge.net/tracker/index.php?funcÞtailaid05017group_id059
9atid3831
https://sourceforge.net/tracker/index.php?funcÞtailaid05017group_id05
99atid3831
I am working on a fix for this, but I can't reproduce the
Avi Kivity wrote:
3. With using qcow images xpsp2 with 2vpus may fail to boot
https://sourceforge.net/tracker/index.php?func=detailaid=1805017group_id=180599atid=893831
https://sourceforge.net/tracker/index.php?func=detailaid=1805017group_id=180599atid=893831
I am working on a fix
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:
- fix compilation with older kernels (Andi Kleen)
- virtio host support
Yang, Sheng wrote:
On Tuesday 22 January 2008 21:54:10 Avi Kivity wrote:
Yang, Sheng wrote:
+
+/* Compute with 96 bit intermediate result: (a*b)/c */
+static u64 muldiv64(u64 a, u32 b, u32 c)
Why do we need such high accuracy for the pit?
The direct reason is we
Christoph Lameter wrote:
Ahhh. Good to hear. But we will still end in a situation where only
the remote ptes point to the page. Maybe the remote instance will dirty
the page at that point?
When the spte is dropped, its dirty bit is transferred to the page.
sharing code, and for you
On Wed, Jan 23, 2008 at 12:27:57PM +0200, Avi Kivity wrote:
The approach with the export notifier is page based not based on the
mm_struct. We only need a single page count for a page that is exported to
a number of remote instances of linux. The page count is dropped when all
the remote
Andi Kleen wrote:
While starting a kernel
Unable to handle kernel NULL pointer dereference at 0008 RIP:
[885698fd] :kvm_amd:svm_vcpu_run+0x35/0x30d
Can you disassemble svm_vcpu_run() so we see where this is?
--
error compiling committee.c: too many arguments to
Avi Kivity wrote:
Avi Kivity wrote:
3. With using qcow images xpsp2 with 2vpus may fail to boot
https://sourceforge.net/tracker/index.php?func=detailaid=1805017group_id=180599atid=893831
https://sourceforge.net/tracker/index.php?func=detailaid=1805017group_id=180599atid=893831
Hi Christoph,
On Tue, Jan 22, 2008 at 02:53:12PM -0800, Christoph Lameter wrote:
On Tue, 22 Jan 2008, Andrea Arcangeli wrote:
First it makes me optimistic this can be merged sooner than later to
see a second brand new implementation of this ;).
Brand new? Well this is borrowing as much
On Wed, Jan 23, 2008 at 04:52:47AM -0600, Robin Holt wrote:
But 100 callouts holding spinlocks will not work for our implementation
and even if the callouts are made with spinlocks released, we would very
strongly prefer a single callout which messages the range to the other
side.
But you
-Original Message-
From: Avi Kivity [mailto:[EMAIL PROTECTED]
Sent: 2008年1月23日 19:40
To: Zhao, Yunfeng
Cc: kvm-devel
Subject: Re: [kvm-devel] Test result for KVM60RC1
Avi Kivity wrote:
Avi Kivity wrote:
3. With using qcow images xpsp2 with 2vpus may fail to boot
Christoph, Maybe you can clear one thing up. Was this proposal an
addition to or replacement of Andrea's? I assumed an addition. I am
going to try to restrict my responses to ones appropriate for that
assumption.
The remote instance is like a secondary TLB what you're doing in your
code is
Why don't you stick to mm+va and use get_user_pages and let the VM do
the swapins etc...?
Our concern is also with memory protections for the physical page.
Additionally, we need to support exporting of memory in a VM_PFNMAP
address ranges (specifically, memory mapped using sgi_fetchop,
Zhao, Yunfeng wrote:
And it's now fixed, kvm-userspace 76640d2815c628f23351e2c96c48c9e505e2d83e.
Quick fix again.:) We will retest it in next testing.
Will this fix affect performance of the guest?
Hopefully not. Windows smp boots quickly, and I also ran a uniprocessor
I/O
Hi,
Jumping in here, looks like this could develop into a direction useful
for Xen.
Background: Xen has a mechanism called grant tables for page sharing.
Guest #1 can issue a grant for another guest #2, which in turn then
can use that grant to map the page owned by guest #1 into its address
On Wednesday 23 January 2008 12:28:06 Avi Kivity wrote:
Andi Kleen wrote:
While starting a kernel
Unable to handle kernel NULL pointer dereference at 0008 RIP:
[885698fd] :kvm_amd:svm_vcpu_run+0x35/0x30d
Can you disassemble svm_vcpu_run() so we see where this ist
Andi Kleen wrote:
On Wednesday 23 January 2008 12:28:06 Avi Kivity wrote:
Andi Kleen wrote:
While starting a kernel
Unable to handle kernel NULL pointer dereference at 0008 RIP:
[885698fd] :kvm_amd:svm_vcpu_run+0x35/0x30d
Can you disassemble
Andi Kleen wrote:
On Tue, Jan 15, 2008 at 12:56:52PM +0200, Avi Kivity wrote:
Andi Kleen wrote:
FWIW it seems things are broken even without -kernel in -59 too. If I try
to boot an existing image with just -hda ... the VGA screen just stays
black while the process runs at 99% CPU.
On Wed, Jan 23, 2008 at 01:51:23PM +0100, Gerd Hoffmann wrote:
Jumping in here, looks like this could develop into a direction useful
for Xen.
Background: Xen has a mechanism called grant tables for page sharing.
Guest #1 can issue a grant for another guest #2, which in turn then
can use
Izik Eidus wrote:
Andi Kleen wrote:
On Tue, Jan 15, 2008 at 12:56:52PM +0200, Avi Kivity wrote:
Andi Kleen wrote:
FWIW it seems things are broken even without -kernel in -59 too. If
I try
to boot an existing image with just -hda ... the VGA screen just stays
black while the process
Le mercredi 23 janvier 2008 à 15:29 +0200, Izik Eidus a écrit :
Izik Eidus wrote:
Andi Kleen wrote:
On Tue, Jan 15, 2008 at 12:56:52PM +0200, Avi Kivity wrote:
Andi Kleen wrote:
FWIW it seems things are broken even without -kernel in -59 too. If
I try
to boot an existing
Le mercredi 23 janvier 2008 à 14:38 +0100, Laurent Vivier a écrit :
Le mercredi 23 janvier 2008 à 15:29 +0200, Izik Eidus a écrit :
Izik Eidus wrote:
Andi Kleen wrote:
On Tue, Jan 15, 2008 at 12:56:52PM +0200, Avi Kivity wrote:
Andi Kleen wrote:
FWIW it seems things are
Joerg Roedel wrote:
Change the Makefile in kernel/ to explicitly call GNU awk if its not the
default awk installed (like in Ubuntu). This fixes a build error seen on
Ubuntu
7.10.
Applied, thanks.
--
error compiling committee.c: too many arguments to function
Hi,
That would render the notifies useless for Xen too. Xen needs to
intercept the actual pte clear and instead of just zapping it use the
hypercall to do the unmap and release the grant.
We are tackling that by having our own page table hanging off the
structure representing our seg
Jan Engelhardt wrote:
Signed-off-by: Jan Engelhardt [EMAIL PROTECTED]
---
drivers/kvm/kvm_main.c |4 ++--
Applied, but please generate patches against kvm.git as kvm moves
quickly. In this case, the file has moved to virt/kvm/.
--
error compiling committee.c: too many arguments to
Robin Holt wrote:
On Wed, Jan 23, 2008 at 01:51:23PM +0100, Gerd Hoffmann wrote:
Jumping in here, looks like this could develop into a direction useful
for Xen.
Background: Xen has a mechanism called grant tables for page sharing.
Guest #1 can issue a grant for another guest #2, which
On Wed, Jan 23, 2008 at 03:12:36PM +0100, Gerd Hoffmann wrote:
Hi,
That would render the notifies useless for Xen too. Xen needs to
intercept the actual pte clear and instead of just zapping it use the
hypercall to do the unmap and release the grant.
We are tackling that by having
Since we want to reset the device to remove them, this is simpler
(device is reset for us on driver remove).
Signed-off-by: Rusty Russell [EMAIL PROTECTED]
---
drivers/net/virtio_net.c | 44 +---
1 file changed, 25 insertions(+), 19 deletions(-)
diff -r
(One big patch, I'll merge it in properly later).
A reset function solves three problems:
1) It allows us to renegotiate features, eg. if we want to upgrade a
guest driver without rebooting the guest.
2) It gives us a clean way of shutting down virtqueues: after a reset,
we know that the
Robin Holt wrote:
We have a seg structure which is similar to some structure you probably
have which describes the grant. One of the things hanging off that
seg structure is essentially a page table containing PFNs with their
respective flags (XPMEM specific and not the same as the pfn flags
Hi Kraxel,
On Wed, Jan 23, 2008 at 01:51:23PM +0100, Gerd Hoffmann wrote:
That would render the notifies useless for Xen too. Xen needs to
intercept the actual pte clear and instead of just zapping it use the
hypercall to do the unmap and release the grant.
I think it has yet to be
On Wed, Jan 23, 2008 at 03:35:54PM +0100, Gerd Hoffmann wrote:
Robin Holt wrote:
We have a seg structure which is similar to some structure you probably
have which describes the grant. One of the things hanging off that
seg structure is essentially a page table containing PFNs with their
Izik Eidus wrote:
the bellow patch should fix this opss
(it is targeted for 2.6.24-rc7)
sorry i forgat one break in this patch
this is the fixed patch.
Thanks, queued for 2.6.24.
--
error compiling committee.c: too many arguments to function
On Mon, 21 Jan 2008 18:05:53 +0200
Izik Eidus [EMAIL PROTECTED] wrote:
i added 2 new functions to the kernel
one:
page_wrprotect() make the page as read only by setting the ptes point to
it as read only.
second:
replace_page() - replace the pte mapping related to vm area between two
pages
On Wed, Jan 23, 2008 at 06:32:30AM -0600, Robin Holt wrote:
Christoph, Maybe you can clear one thing up. Was this proposal an
addition to or replacement of Andrea's? I assumed an addition. I am
going to try to restrict my responses to ones appropriate for that
assumption.
It wasn't
On Mon, 21 Jan 2008 18:09:29 +0200
Izik Eidus [EMAIL PROTECTED] wrote:
Acked-by: Rik van Riel [EMAIL PROTECTED]
--
All rights reversed.
-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R)
Andrea Arcangeli wrote:
Like Avi said, Xen is dealing with the linux pte only, so there's no
racy smp page fault to serialize against. Perhaps we can add another
notifier for Xen though.
But I think it's still not enough for Xen to have a method called
before the ptep_clear_flush: rmap.c
On Wed, Jan 23, 2008 at 12:05:10PM -0500, Rik van Riel wrote:
On Mon, 21 Jan 2008 18:05:53 +0200
Izik Eidus [EMAIL PROTECTED] wrote:
i added 2 new functions to the kernel
one:
page_wrprotect() make the page as read only by setting the ptes point to
it as read only.
second:
On Mon, 21 Jan 2008 18:10:50 +0200
Izik Eidus [EMAIL PROTECTED] wrote:
What prevents another CPU from freeing newpage while we run through
the start of replace_page() ?
--
All rights reversed.
-
This SF.net email is
Andrea Arcangeli wrote:
On Wed, Jan 23, 2008 at 12:05:10PM -0500, Rik van Riel wrote:
On Mon, 21 Jan 2008 18:05:53 +0200
Izik Eidus [EMAIL PROTECTED] wrote:
i added 2 new functions to the kernel
one:
page_wrprotect() make the page as read only by setting the ptes point to
it as
Rik van Riel wrote:
On Mon, 21 Jan 2008 18:10:50 +0200
Izik Eidus [EMAIL PROTECTED] wrote:
What prevents another CPU from freeing newpage while we run through
the start of replace_page() ?
before calling to replace_page one have to call to get_page() and to do
put_page() after it
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 properly, so the migration would
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.
Setup:
2 identical Intel SDV boxes, Intel(R) Core(TM)2 Duo CPU
On Wed, 23 Jan 2008, Robin Holt wrote:
That won't work for kvm. If we have a hundred virtual machines, that means
99 no-op notifications.
But 100 callouts holding spinlocks will not work for our implementation
and even if the callouts are made with spinlocks released, we would very
On Wed, 23 Jan 2008, Andrea Arcangeli wrote:
On Wed, Jan 23, 2008 at 04:52:47AM -0600, Robin Holt wrote:
But 100 callouts holding spinlocks will not work for our implementation
and even if the callouts are made with spinlocks released, we would very
strongly prefer a single callout which
Hi,
Quick question: is kvm able to handle guests using pae when the host
runs a non-pae 32bit kernel?
cheers,
Gerd
-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
On Wed, Jan 23, 2008 at 11:48:43AM -0800, Christoph Lameter wrote:
On Wed, 23 Jan 2008, Andrea Arcangeli wrote:
On Wed, Jan 23, 2008 at 04:52:47AM -0600, Robin Holt wrote:
But 100 callouts holding spinlocks will not work for our implementation
and even if the callouts are made with
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_page() so
the other linux instance will be told that the page is not available.
On Wed, 23 Jan 2008, Andrea Arcangeli wrote:
You want to be able to tell the mmu_notifier that you want the flush
repeated without locks later? Sorry but then if you're always going to
set the bitflag unconditionally, why don't you simply implement a
second notifier in addition of my current
On Wed, 23 Jan 2008, Andrea Arcangeli wrote:
I think it has yet to be demonstrated that doing the invalidate
_before_ clearing the linux pte is workable at all for
shadow-pte/RDMA. Infact even doing it _after_ still requires some form
of serialization but it's less obviously broken and
Sending You My Love http://89.110.13.74/
-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/
* 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 page
What happens if you've merged more pages than you can
On Saturday 22 December 2007 02:27:04 Dor Laor wrote:
From 20a01e0c24f47e9805a88bebc9310163f99b5e99 Mon Sep 17 00:00:00 2001
From: Dor Laor [EMAIL PROTECTED]
Date: Sun, 16 Dec 2007 23:35:35 +0200
Subject: [PATCH] Handle module unload Add the device release function.
Hi Dor,
It looks
On Thursday 03 January 2008 08:01:32 Anthony Liguori wrote:
The virtio modules do not advertise themselves as GPL but rely on exported
GPL symbols. This makes them unloadable as modules.
This patch adds the appropriate MODULE_LICENSE().
Hmm, this is already at the bottom of virtio_blk and
Expand the mmu_notifiers to allow for lockless callers. To accomplish
this, the function receiving notifications needs to implement an rmap
equivalent. The notification function is also responsible for tracking
page dirty state.
With this patch, I am getting fairly close to not needing the
Here's the latest. Hope this works for everyone (putting in a oom handler or
shrinker requires a lock, but can be done quite easily).
Untested.
===
After discussions with Anthony Liguori, it seems that the virtio
balloon can be made even simpler. Here's my attempt.
Signed-off-by: Rusty Russell
On Wed, 2008-01-23 at 16:17 +0200, Avi Kivity wrote:
Robin Holt wrote:
On Wed, Jan 23, 2008 at 01:51:23PM +0100, Gerd Hoffmann wrote:
Jumping in here, looks like this could develop into a direction useful
for Xen.
Background: Xen has a mechanism called grant tables for page
Expand the mmu_notifiers to allow for lockless callers. To accomplish
this, the function receiving notifications needs to implement an rmap
equivalent. The notification function is also responsible for tracking
page dirty state.
Version 2 brings with it __xip_unmap and do_wp_page so this is
Rik van Riel wrote:
On Mon, 21 Jan 2008 18:05:53 +0200
Izik Eidus [EMAIL PROTECTED] wrote:
i added 2 new functions to the kernel
one:
page_wrprotect() make the page as read only by setting the ptes point to
it as read only.
second:
replace_page() - replace the pte mapping related to
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 page
What happens if
Gerd Hoffmann wrote:
Another maybe workable approach for Xen is to go through pv_ops
(although pte_clear doesn't go through pv_ops right now, so this would
be an additional hook too ...).
I think that's the way. Xen is not a secondary mmu but rather a primary
mmu with some magic
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 VM entry and VM exit, the VMM can
H. Peter Anvin wrote:
Just for fun, I tried to boot OS/2 Warp 4.0 under KVM (KVM-59 with the
latest git kernel from Linus as of yesterday, slightly post 2.6.24-rc8.)
I found that it crashes very early, apparently because KVM doesn't
handle an #UD received in user mode. It appears that
Jan Kiszka wrote:
Jan Kiszka wrote:
At this chance, can anyone comment on these uint32_t types?
static int subpage_register (subpage_t *mmio, uint32_t start, uint32_t end,
int memory);
static void *subpage_init (target_phys_addr_t base, uint32_t *phys,
Gerd Hoffmann wrote:
Another maybe workable approach for Xen is to go through pv_ops
(although pte_clear doesn't go through pv_ops right now, so this would
be an additional hook too ...).
It does for 32-bit PAE. Making pte_clear uniform across all pagetable
modes would be a nice cleanup.
Avi Kivity wrote:
H. Peter Anvin wrote:
Just for fun, I tried to boot OS/2 Warp 4.0 under KVM (KVM-59 with the
latest git kernel from Linus as of yesterday, slightly post
2.6.24-rc8.)I found that it crashes very early, apparently because
KVM doesn't handle an #UD received in user mode.
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.
Setup:
2 identical Intel SDV boxes,
Gerd Hoffmann wrote:
Quick question: is kvm able to handle guests using pae when the host
runs a non-pae 32bit kernel
Yes.
--
Any sufficiently difficult bug is indistinguishable from a feature.
-
This SF.net email
* Izik Eidus ([EMAIL PROTECTED]) wrote:
Just a bunch of nitpicks.
struct ksm *ksm;
static
static int page_hash_size = 0;
no need to initialize to zero
module_param(page_hash_size, int, 0);
MODULE_PARM_DESC(page_hash_size, Hash table size for the pages checksum);
static int
71 matches
Mail list logo