On Tuesday 03 November 2009, Benjamin Herrenschmidt wrote:
(Though glibc can be nasty, afaik it might load up optimized variants of
some routines with hard wired cache line sizes based on the CPU type)
You can also get application with hand-coded cache optimizations
that are even harder, if not
On Wed, 2009-11-04 at 09:43 +0100, Arnd Bergmann wrote:
On Tuesday 03 November 2009, Benjamin Herrenschmidt wrote:
(Though glibc can be nasty, afaik it might load up optimized
variants of
some routines with hard wired cache line sizes based on the CPU
type)
You can also get application
On 11/03/2009 05:45 PM, Harald Dunkel wrote:
Hi folks,
I am trying to use a bonding network interface as a bridge
for a virtual machine (kvm). Host and guest are both running
2.6.31.5. Problem: The guest does not receive the DHCPOFFER
reply sent by my dhcp server. There is no such problem if
On 11/03/2009 05:12 PM, Marcelo Tosatti wrote:
On Wed, Oct 28, 2009 at 01:23:07PM -0400, David Windsor wrote:
Hi,
I've encountered a situation in which I would like to allow userspace
to set the MSRs which KVM should not emulate and instead implement
these as no-ops.
I have not seen any
On 11/02/2009 01:21 AM, Michael McStarfighter wrote:
Hi,
I've got a Kubuntu host and two dedicated graphic cards. Now I want to
create a Windows Seven guest for KVM and give the guest one of the
graphic cards (it isn't in business on my host). The informations I
got is only that PCI passthrough
Ross Boylan wrote:
On Sat, 2009-10-31 at 15:21 +0300, Michael Tokarev wrote:
[]
Switch to alsa to get your audio working.
I don't see an alsa option for kvm/qemu. I'm already running alsa, but
under KDE which tends to grab the device.
Which device you think KDE grabs?
OSS device (/dev/dsp)
On 10/29/2009 08:37 PM, Danny ter Haar wrote:
Update:
I compiled/installed 2.6.32-rc5-git3 on this machine
I manually start kvm:
vhost1:~# kvm -m 512 -cdrom /vz/template/iso/debian-503-amd64-netinst.iso
The bootscreen comes up, i hit enter to install and i get these messages
(copied from
Usually userspace will freeze the guest so we can inspect it, but some
internal state is not available. Add extra data to internal error
reporting so we can expose it to the debugger. Extra data is specific
to the suberror.
Signed-off-by: Avi Kivity a...@redhat.com
---
arch/x86/kvm/mmu.c |
These happen when we trap an exception when another exception is being
delivered; we only expect these with MCEs and page faults. If something
unexpected happens, things probably went south and we're better off reporting
an internal error and freezing.
Signed-off-by: Avi Kivity a...@redhat.com
Currently if we hit an error processing simultaneous exceptions we mumble
something in dmesg and continue. AFAIK this situation always deteriorates,
so attempt to report it to userspace to give us a chance to debug it.
Avi Kivity (2):
KVM: Allow internal errors reported to userspace to carry
So we have a chance to debug it.
Signed-off-by: Avi Kivity a...@redhat.com
---
qemu-kvm.c |3 ++-
1 files changed, 2 insertions(+), 1 deletions(-)
diff --git a/qemu-kvm.c b/qemu-kvm.c
index 4b22537..5568706 100644
--- a/qemu-kvm.c
+++ b/qemu-kvm.c
@@ -878,7 +878,8 @@ static int
Signed-off-by: Avi Kivity a...@redhat.com
---
kvm/include/linux/kvm.h |4
1 files changed, 4 insertions(+), 0 deletions(-)
diff --git a/kvm/include/linux/kvm.h b/kvm/include/linux/kvm.h
index f8f8900..c518b36 100644
--- a/kvm/include/linux/kvm.h
+++ b/kvm/include/linux/kvm.h
@@ -187,6
If a kvm internal error occurs, freeze the vm so we can inspect it. Also
provide any extra data the kernel gives us.
Avi Kivity (4):
Move internal error handling into a function
Freeze guest on internal error
Report extra data on internal errors
Import internal error data header updates
Signed-off-by: Avi Kivity a...@redhat.com
---
qemu-kvm.c | 19 +--
1 files changed, 13 insertions(+), 6 deletions(-)
diff --git a/qemu-kvm.c b/qemu-kvm.c
index 809fd65..4b22537 100644
--- a/qemu-kvm.c
+++ b/qemu-kvm.c
@@ -869,6 +869,18 @@ int
Avi Kivity wrote:
Can you tcpdump on bond0, br0, vnet0, and the guest's interface to see
where the packet is lost?
Sure. Using the tcpdump command line:
tcpdump -i br0 -w /var/tmp/tcpdump.br0 ether host 00:16:36:2f:f1:d2
(similar for other interfaces) I can see the DHCPOFFER coming
from
On 11/04/2009 01:02 PM, Harald Dunkel wrote:
Avi Kivity wrote:
Can you tcpdump on bond0, br0, vnet0, and the guest's interface to see
where the packet is lost?
Sure. Using the tcpdump command line:
tcpdump -i br0 -w /var/tmp/tcpdump.br0 ether host 00:16:36:2f:f1:d2
(similar for
My thoughts are only to give the Windows guest more graphic power.
Isn't it possible to get this without a dedicated monitor? My basic
idea is a Kubuntu host. And on this I will try to use as many windows
apps via Wine as possible (also games). And all windows apps I still
need and and want but
On 11/02/2009 06:20 PM, Jan Kiszka wrote:
Add a new IOCTL pair to retrieve or set the VCPU state in one chunk.
More precisely, the IOCTL is able to process a list of substates to be
read or written. This list is easily extensible without breaking the
existing ABI, thus we will no longer have to
On 11/02/2009 06:20 PM, Jan Kiszka wrote:
This new substate exports all yet user-invisible states related to
exceptions, interrupts, and NMIs. Together with appropriate user space
changes, this fixes sporadic problems of vmsave/restore, live migration
and system reset.
+
On 11/04/2009 01:14 PM, Michael McStarfighter wrote:
My thoughts are only to give the Windows guest more graphic power.
Isn't it possible to get this without a dedicated monitor?
I'm no expert on graphics, but unless the two cards somehow multiplex
the monitor, anything Windows outputs
By the way: How good is the KVM own graphic card? Is it possible to
use / share the host graphics power?
2009/11/4 Avi Kivity a...@redhat.com:
On 11/04/2009 01:14 PM, Michael McStarfighter wrote:
My thoughts are only to give the Windows guest more graphic power.
Isn't it possible to get this
On 04.11.2009, at 09:47, Benjamin Herrenschmidt wrote:
On Wed, 2009-11-04 at 09:43 +0100, Arnd Bergmann wrote:
On Tuesday 03 November 2009, Benjamin Herrenschmidt wrote:
(Though glibc can be nasty, afaik it might load up optimized
variants of
some routines with hard wired cache line sizes
Avi Kivity wrote:
On 11/02/2009 06:20 PM, Jan Kiszka wrote:
Add a new IOCTL pair to retrieve or set the VCPU state in one chunk.
More precisely, the IOCTL is able to process a list of substates to be
read or written. This list is easily extensible without breaking the
existing ABI, thus we
On Tue, Nov 03, 2009 at 10:11:12PM +0100, Eric Dumazet wrote:
Michael S. Tsirkin a écrit :
Paul, you acked this previously. Should I add you acked-by line so
people calm down? If you would rather I replace
rcu_dereference/rcu_assign_pointer with rmb/wmb, I can do this.
Or maybe patch
On Wed, Nov 04, 2009 at 12:08:47PM +0100, Andi Kleen wrote:
Michael S. Tsirkin m...@redhat.com writes:
Haven't really read the whole thing, just noticed something at a glance.
+/* Expects to be always run from workqueue - which acts as
+ * read-size critical section for our kind of RCU.
On 11/04/2009 01:34 PM, Jan Kiszka wrote:
we're missing has_error_code here.
Isn't it statically defined by the arch? exception_has_error_code()
regenerates it (just like user land does).
In real mode exceptions lose their error code, so we have to look at
cr0, and now we depend
On 11/04/2009 01:35 PM, Jan Kiszka wrote:
Avi Kivity wrote:
On 11/02/2009 06:20 PM, Jan Kiszka wrote:
Add a new IOCTL pair to retrieve or set the VCPU state in one chunk.
More precisely, the IOCTL is able to process a list of substates to be
read or written. This list is easily
Fine?
I cannot say -- are there paths that could drop the device beforehand?
(as in do you hold a reference to it?)
-Andi
--
a...@linux.intel.com -- Speaking for myself only.
--
To unsubscribe from this list: send the line unsubscribe kvm in
the body of a message to majord...@vger.kernel.org
On Wed, Nov 04, 2009 at 01:59:57PM +0100, Andi Kleen wrote:
Fine?
I cannot say -- are there paths that could drop the device beforehand?
Do you mean drop the mm reference?
(as in do you hold a reference to it?)
By design I think I always have a reference to mm before I use it.
This works
On Wed, Nov 04, 2009 at 03:08:28PM +0200, Michael S. Tsirkin wrote:
On Wed, Nov 04, 2009 at 01:59:57PM +0100, Andi Kleen wrote:
Fine?
I cannot say -- are there paths that could drop the device beforehand?
Do you mean drop the mm reference?
No the reference to the device, which owns
On Wed, Nov 04, 2009 at 02:15:33PM +0100, Andi Kleen wrote:
On Wed, Nov 04, 2009 at 03:08:28PM +0200, Michael S. Tsirkin wrote:
On Wed, Nov 04, 2009 at 01:59:57PM +0100, Andi Kleen wrote:
Fine?
I cannot say -- are there paths that could drop the device beforehand?
Do you mean
Passthrough graphic card to guest can satisfy your requirement. Currently it's
not supported in kvm, but we have a plan to support it, but you still need wait
for a while.
I think sharing the host graphics power is not easy to implement.
Regards,
Weidong
Michael McStarfighter wrote:
By the
From: Michael S. Tsirkin m...@redhat.com
Date: Tue, 3 Nov 2009 19:24:00 +0200
Assuming it's okay with davem, I think it makes sense to merge this
patch through Rusty's tree because vhost is the first user of the new
interface. Posted here for completeness.
I'm fine with that, please add my:
On Wed, Nov 04, 2009 at 03:17:36PM +0200, Michael S. Tsirkin wrote:
On Wed, Nov 04, 2009 at 02:15:33PM +0100, Andi Kleen wrote:
On Wed, Nov 04, 2009 at 03:08:28PM +0200, Michael S. Tsirkin wrote:
On Wed, Nov 04, 2009 at 01:59:57PM +0100, Andi Kleen wrote:
Fine?
I cannot say --
On Wed, Nov 04, 2009 at 02:37:28PM +0100, Andi Kleen wrote:
On Wed, Nov 04, 2009 at 03:17:36PM +0200, Michael S. Tsirkin wrote:
On Wed, Nov 04, 2009 at 02:15:33PM +0100, Andi Kleen wrote:
On Wed, Nov 04, 2009 at 03:08:28PM +0200, Michael S. Tsirkin wrote:
On Wed, Nov 04, 2009 at
In case of a single video card in the system, it would be wonderful to
be able to suddenly give the video card as passthrough to the guest
removing it from the host. Since the video card internal state is
unknown to the guest, the guest should then re-initialize it with some
help of a guest
Add a command to setuprss.bat that disables the password prompt on resuming
from hibernation.
Note that for this change to take effect the local winutils.iso should be
updated as well.
Signed-off-by: Michael Goldish mgold...@redhat.com
---
client/tests/kvm/deps/setuprss.bat |1 +
1 files
Don't select the start word, which is usually rendered with antialiased
characters, but is sometimes aliased for some reason. I'm not sure if this is
just weird Windows behavior or some sort of qemu/KVM bug.
Signed-off-by: Michael Goldish mgold...@redhat.com
---
1. Run a process in the background and make sure it keeps running after
migration (similarly to guest_s4).
2. When logging in after migration allow the procedure to take up to 30s
instead of only 10s. We can afford this because if the guest reboots during
those 30s the background process will not
Use the file browser to run setuprss.bat as administrator.
Signed-off-by: Michael Goldish mgold...@redhat.com
---
client/tests/kvm/steps/WinVista-32-rss.steps | 59 +-
client/tests/kvm/steps/WinVista-64-rss.steps | 59 +-
2 files changed, 58
Also, properly handle closed file descriptors in read_until_output_matches().
Signed-off-by: Michael Goldish mgold...@redhat.com
---
client/tests/kvm/kvm_subprocess.py | 11 ---
1 files changed, 8 insertions(+), 3 deletions(-)
diff --git a/client/tests/kvm/kvm_subprocess.py
This speeds up SSH logins because by default other time consuming
authentication methods are attempted first.
Signed-off-by: Michael Goldish mgold...@redhat.com
---
client/tests/kvm/kvm_utils.py |9 ++---
1 files changed, 6 insertions(+), 3 deletions(-)
diff --git
The current implementation of get_user_desc() sign extends
the return value because of integer promotion rules. For
the most part, this doesn't matter, because the top bit of
base2 is usually 0. If, however, that bit is 1, then the
entire value will be 0x... which is probably not what
the
On Tue, Nov 03, 2009 at 11:38:18AM +0200, Avi Kivity wrote:
On 11/03/2009 01:25 PM, Vincent Hanquez wrote:
not sure if i'm missing the point here, but couldn't it be hypothetically
extended to stuff 3d (or video more 2d accel ?) commands too ? I can't
imagine the cirrus or stdvga driver be
Avi Kivity wrote:
On 11/04/2009 01:34 PM, Jan Kiszka wrote:
we're missing has_error_code here.
Isn't it statically defined by the arch? exception_has_error_code()
regenerates it (just like user land does).
In real mode exceptions lose their error code, so we have to look at
Ok, I think I've addressed all comments so far here.
Rusty, I'd like this to go into linux-next, through your tree, and
hopefully 2.6.33. What do you think?
---
This implements vhost: a kernel-level backend for virtio,
The main motivation for this work is to reduce virtualization
overhead for
Tun device looks similar to a packet socket
in that both pass complete frames from/to userspace.
This patch fills in enough fields in the socket underlying tun driver
to support sendmsg/recvmsg operations, and message flags
MSG_TRUNC and MSG_DONTWAIT, and exports access to this socket
to modules.
vhost net module wants to do copy to/from user from a kernel thread,
which needs use_mm. Export it to modules.
Acked-by: Andrea Arcangeli aarca...@redhat.com
Signed-off-by: Michael S. Tsirkin m...@redhat.com
---
mm/mmu_context.c |3 +++
1 files changed, 3 insertions(+), 0 deletions(-)
diff
What it is: vhost net is a character device that can be used to reduce
the number of system calls involved in virtio networking.
Existing virtio net code is used in the guest without modification.
There's similarity with vringfd, with some differences and reduced scope
- uses eventfd for
Michael S. Tsirkin wrote:
Ok, I think I've addressed all comments so far here.
Rusty, I'd like this to go into linux-next, through your tree, and
hopefully 2.6.33. What do you think?
I think the benchmark data is a prerequisite for merge consideration, IMO.
Do you have anything for us to
Vincent Hanquez wrote:
On Tue, Nov 03, 2009 at 11:38:18AM +0200, Avi Kivity wrote:
On 11/03/2009 01:25 PM, Vincent Hanquez wrote:
not sure if i'm missing the point here, but couldn't it be hypothetically
extended to stuff 3d (or video more 2d accel ?) commands too ? I can't
imagine
On Wed, Nov 04, 2009 at 11:02:15AM -0500, Gregory Haskins wrote:
Michael S. Tsirkin wrote:
Ok, I think I've addressed all comments so far here.
Rusty, I'd like this to go into linux-next, through your tree, and
hopefully 2.6.33. What do you think?
I think the benchmark data is a
On Wed, Nov 04, 2009 at 03:41:47PM +0200, Michael S. Tsirkin wrote:
On Wed, Nov 04, 2009 at 02:37:28PM +0100, Andi Kleen wrote:
On Wed, Nov 04, 2009 at 03:17:36PM +0200, Michael S. Tsirkin wrote:
On Wed, Nov 04, 2009 at 02:15:33PM +0100, Andi Kleen wrote:
On Wed, Nov 04, 2009 at
Host is ext3
Linux monkamu 2.6.31-14-generic #48-Ubuntu SMP Fri Oct 16 14:04:26 UTC
2009 i686 GNU/Linux
QEMU PC emulator version 0.11.0 (qemu-kvm-0.11.0), Copyright (c)
2003-2008 Fabrice Bellard
Avi Kivity wrote:
On 11/04/2009 05:45 AM, Andrew Olney wrote:
OK, if I convert the raw
Right now, by default autotest does:
* Drop the caches of the client machine
* Queue the package manager for all packages installed
On every test execution, which takes a long time for
tests to complete, which is good in many scenarios, but
could be configurable with minimal fuss for other
On Wed, Nov 04, 2009 at 01:57:29PM +0200, Michael S. Tsirkin wrote:
On Tue, Nov 03, 2009 at 03:57:44PM -0800, Paul E. McKenney wrote:
On Tue, Nov 03, 2009 at 01:14:06PM -0500, Gregory Haskins wrote:
Gregory Haskins wrote:
Eric Dumazet wrote:
Michael S. Tsirkin a écrit :
+static
Paul E. McKenney a écrit :
(Sorry, but, as always, I could not resist!)
Yes :)
Thanks Paul for this masterpiece of diplomatic Acked-by ;)
--
To unsubscribe from this list: send the line unsubscribe kvm in
the body of a message to majord...@vger.kernel.org
More majordomo info at
This patch reduces the size of memory being cleared on every virtio-blk IO.
On every virtio-blk IO passed to QEMU, virtio_blk_alloc_request()
allocates and clears (with qemu_mallocz()) a VirtIOBlockReq object.
The sizeof(VirtIOBlockReq) equals 41040 bytes on my x86-64 machine.
By moving the
On Tuesday 03 November 2009, Arnd Bergmann wrote:
index 3f5fd52..404abe0 100644
--- a/include/linux/if_tun.h
+++ b/include/linux/if_tun.h
@@ -86,4 +86,18 @@ struct tun_filter {
__u8 addr[0][ETH_ALEN];
};
+#ifdef __KERNEL__
+#if defined(CONFIG_TUN) ||
On Wed, Nov 04, 2009 at 09:25:42AM -0800, Paul E. McKenney wrote:
(Sorry, but, as always, I could not resist!)
Thanx, Paul
Thanks Paul!
Jonathan: are you reading this?
Another one for your quotes of the week collection :)
--
MST
--
To
Michael S. Tsirkin wrote:
On Wed, Nov 04, 2009 at 09:25:42AM -0800, Paul E. McKenney wrote:
(Sorry, but, as always, I could not resist!)
Thanx, Paul
Thanks Paul!
Jonathan: are you reading this?
Another one for your quotes of the week
Michael S. Tsirkin wrote:
On Wed, Nov 04, 2009 at 11:02:15AM -0500, Gregory Haskins wrote:
Michael S. Tsirkin wrote:
Ok, I think I've addressed all comments so far here.
Rusty, I'd like this to go into linux-next, through your tree, and
hopefully 2.6.33. What do you think?
I think the
On Wed, Nov 04, 2009 at 02:15:42PM -0500, Gregory Haskins wrote:
Michael S. Tsirkin wrote:
On Wed, Nov 04, 2009 at 11:02:15AM -0500, Gregory Haskins wrote:
Michael S. Tsirkin wrote:
Ok, I think I've addressed all comments so far here.
Rusty, I'd like this to go into linux-next, through
+ case OP_31_XOP_EIOIO:
+ break;
Have you always executed an eieio or sync when you get here, or
do you just not allow direct access to I/O devices? Other context
synchronising insns are not enough, they do not broadcast on the
bus.
There is no device
Asdo wrote:
In case of a single video card in the system, it would be wonderful to
be able to suddenly give the video card as passthrough to the guest
removing it from the host. Since the video card internal state is
unknown to the guest, the guest should then re-initialize it with some
help
65 matches
Mail list logo