On Monday 17 May 2010 22:23:46 Chris Wright wrote:
> Please send in any agenda items you are interested in covering.
>
> If we have a lack of agenda items I'll cancel the week's call.
Perceived long standing bugs that nobody seems to care about. There are a few,
one of which is the > 1TB [1] bu
From: "Michael S. Tsirkin"
Date: Tue, 18 May 2010 06:12:52 +0300
> DaveM, just to clarify, this patch is on top of the series
> we are working on with David L Stevens. It's not for your net tree.
Understood.
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a messa
From: "Michael S. Tsirkin"
Date: Tue, 18 May 2010 04:19:31 +0300
> With PUBLISH_USED_IDX, guest tells us which used entries
> it has consumed. This can be used to reduce the number
> of interrupts: after we write a used entry, if the guest has not yet
> consumed the previous entry, or if the gues
Please send in any agenda items you are interested in covering.
If we have a lack of agenda items I'll cancel the week's call.
thanks,
-chris
--
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 http://vger
DaveM, just to clarify, this patch is on top of the series
we are working on with David L Stevens. It's not for your net tree.
--
MST
--
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 http://vger.kernel.
Bugs item #2976863, was opened at 2010-03-26 14:44
Message generated for change (Comment added) made by haoxudong
You can respond by visiting:
https://sourceforge.net/tracker/?func=detail&atid=893831&aid=2976863&group_id=180599
Please note that this message will contain a full copy of the comment
With PUBLISH_USED_IDX, guest tells us which used entries
it has consumed. This can be used to reduce the number
of interrupts: after we write a used entry, if the guest has not yet
consumed the previous entry, or if the guest has already consumed the
new entry, we do not need to interrupt.
This imp
On Mon, May 17, 2010 at 02:16:36PM -0700, David L Stevens wrote:
> This patch enforces single-buffer allocation when
> mergeable rx buffers is not enabled.
>
> Reported-by: Michael S. Tsirkin
> Signed-off-by: David L Stevens
Thanks! Why are you resending?
> diff --git a/drivers/vhost/net.c b/d
With PUBLISH_USED_IDX, guest tells us which used entries
it has consumed. This can be used to reduce the number
of interrupts: after we write a used entry, if the guest has not yet
consumed the previous entry, or if the guest has already consumed the
new entry, we do not need to interrupt.
This imp
On Fri, 2010-05-14 at 17:43 +0800, Feng Yang wrote:
> Emulate the powercut under IO workload(dd so far) using kill -9.
> Then check image in post command.
> This case want to make sure powercut under IO workload will not
> break qcow2 image.
The big question that came to my mind here is: Are we re
On Mon, 2010-05-17 at 15:28 +0800, Jason Wang wrote:
> Michael Goldish wrote:
> > On 05/11/2010 12:04 PM, Jason Wang wrote:
> >
> >> This checker serves as the post_command to find the panic information
> >> in the file which contains the content of guest serial console.
> >>
> >> Signed-off-by:
This patch enforces single-buffer allocation when
mergeable rx buffers is not enabled.
Reported-by: Michael S. Tsirkin
Signed-off-by: David L Stevens
diff --git a/drivers/vhost/net.c b/drivers/vhost/net.c
index 309c570..c346304 100644
--- a/drivers/vhost/net.c
+++ b/drivers/vhost/net.c
@@ -361,
On Fri, 2010-05-14 at 02:42 -0700, Nicholas A. Bellinger wrote:
> On Fri, 2010-05-14 at 09:22 +0200, Hannes Reinecke wrote:
> > Nicholas A. Bellinger wrote:
> > > Greetings Hannes and co,
> > >
>
> > Let's see if I can find some time working on the megasas emulation.
> > Maybe I find something.
>
On Mon, 2010-05-10 at 15:41 +0800, Feng Yang wrote:
> Our pxe case always fail. The problem is that the bridge takes some time
> to enter the forwarding state. Before that, packages are simply dropped by
> the bridge.
> If one sets:
> $ brctl setfd 0 # forward delay = 0
>
> It works.
Also, per
On Sat, 2010-05-08 at 18:01 +0100, Lukas Doktor wrote:
> Signed-off-by: Lukas Doktor
The new patch set looks good, commited them all, see:
http://autotest.kernel.org/changeset/4492
Thanks!
> ---
> client/tests/kvm/kvm_vm.py |2 +-
> 1 files changed, 1 insertions(+), 1 deletions(-)
>
> di
On 05/17/2010 05:36 AM, Glauber Costa wrote:
On Fri, May 14, 2010 at 04:07:43PM -1000, Zachary Amsden wrote:
I believe this fixes the root cause of the kvmclock warp. It's
quite a plausible phenomenon, and explains why it was so easy to
produce.
You mean this is the case for both SM
On 05/12/2010 04:24 PM, Marcelo Tosatti wrote:
The following changes since commit 54d7cf136f040713095cbc064f62d753bff6f9d2:
Markus Armbruster (1):
doc: Clean up monitor command function index
Pulled. Thanks.
Regards,
Anthony Liguori
are available in the git repository at:
On Fri, May 14, 2010 at 04:07:43PM -1000, Zachary Amsden wrote:
> I believe this fixes the root cause of the kvmclock warp. It's
> quite a plausible phenomenon, and explains why it was so easy to
> produce.
>
You mean this is the case for both SMP and UP, or just UP as we talked
before?
I don't
On 05/07/2010 01:26 PM, Feng Yang wrote:
> 1. In remote_scp(), if SCP connetion stalled for some reason, following
> code will be ran.
> else: # match == None
>
> logging.debug("Timeout elapsed or process terminated")
> status = sub.get_status()
> sub.c
On 05/17/2010 04:35 PM, Lucas Meneghel Rodrigues wrote:
> On Mon, 2010-05-17 at 16:29 +0300, Michael Goldish wrote:
>> To disable tcpdump, set 'run_tcpdump = no' in a config file.
>> If 'run_tcpdump' isn't set at all, it defaults to 'yes'.
>>
>> (Currently TAP mode cannot be used without tcpdump.)
On Mon, 2010-05-17 at 16:29 +0300, Michael Goldish wrote:
> To disable tcpdump, set 'run_tcpdump = no' in a config file.
> If 'run_tcpdump' isn't set at all, it defaults to 'yes'.
>
> (Currently TAP mode cannot be used without tcpdump.)
Maybe we can just tie tcpdump execution to tap mode - if run
On 2010-05-17 at 15:13:35 +0200, Michael S. Tsirkin wrote:
> On Mon, May 17, 2010 at 03:12:49PM +0200, Tobias Klauser wrote:
> > The C99 specification states in section 6.11.5:
> >
> > The placement of a storage-class specifier other than at the beginning
> > of the declaration specifiers in a de
To disable tcpdump, set 'run_tcpdump = no' in a config file.
If 'run_tcpdump' isn't set at all, it defaults to 'yes'.
(Currently TAP mode cannot be used without tcpdump.)
Signed-off-by: Michael Goldish
---
client/tests/kvm/kvm_preprocessing.py |2 +-
client/tests/kvm/tests_base.cfg.sample
In order to support multiple versions of qemu which use different command line
options or syntaxes, wrap all command line options in small helper functions,
which append text to the command line according to the output of 'qemu -help'.
Signed-off-by: Michael Goldish
---
client/tests/kvm/kvm_vm.p
Print results clearly even if test names are very long.
Also, for consistency, use the same quote character everywhere.
Signed-off-by: Michael Goldish
---
client/tests/kvm/scan_results.py | 49 ++---
1 files changed, 29 insertions(+), 20 deletions(-)
diff --git
The C99 specification states in section 6.11.5:
The placement of a storage-class specifier other than at the beginning
of the declaration specifiers in a declaration is an obsolescent
feature.
Signed-off-by: Tobias Klauser
---
drivers/vhost/net.c |2 +-
1 files changed, 1 insertions(+), 1 d
On Mon, May 17, 2010 at 03:12:49PM +0200, Tobias Klauser wrote:
> The C99 specification states in section 6.11.5:
>
> The placement of a storage-class specifier other than at the beginning
> of the declaration specifiers in a declaration is an obsolescent
> feature.
>
> Signed-off-by: Tobias Klau
On Mon, May 17, 2010 at 9:20 PM, Kevin Wolf wrote:
> Am 17.05.2010 14:19, schrieb MORITA Kazutaka:
>> At Mon, 17 May 2010 13:08:08 +0200,
>> Kevin Wolf wrote:
>>>
>>> Am 17.05.2010 12:19, schrieb MORITA Kazutaka:
int bdrv_snapshot_goto(BlockDriverState *bs,
Hi,
these two patches implement the workaround for AMD Erratum 383 into KVM.
This is necessary to prevent the host to crash if a guest triggers the
erratum.
For details on the erratum please see page 96 of
http://support.amd.com/us/Processor_TechDocs/41322.pdf
The workaround implemented in these
This patch implements a workaround for AMD erratum 383 into
KVM. Without this erratum fix it is possible for a guest to
kill the host machine. This patch implements the suggested
workaround for hypervisors which will be published by the
next revision guide update.
Cc: sta...@kernel.org
Signed-off-
This patch moves handling of the MC vmexits to an earlier
point in the vmexit. The handle_exit function is too late
because the vcpu might alreadry have changed its physical
cpu.
Cc: sta...@kernel.org
Signed-off-by: Joerg Roedel
---
arch/x86/kvm/svm.c | 15 ++-
1 files changed, 14
Am 17.05.2010 14:19, schrieb MORITA Kazutaka:
> At Mon, 17 May 2010 13:08:08 +0200,
> Kevin Wolf wrote:
>>
>> Am 17.05.2010 12:19, schrieb MORITA Kazutaka:
>>>
>>> int bdrv_snapshot_goto(BlockDriverState *bs,
>>> const char *snapshot_id)
>>> {
>>> BlockDriver *drv =
At Mon, 17 May 2010 13:08:08 +0200,
Kevin Wolf wrote:
>
> Am 17.05.2010 12:19, schrieb MORITA Kazutaka:
> >
> > int bdrv_snapshot_goto(BlockDriverState *bs,
> > const char *snapshot_id)
> > {
> > BlockDriver *drv = bs->drv;
> > +int ret, open_ret;
> > +
> >
Am 17.05.2010 12:19, schrieb MORITA Kazutaka:
> When snapshot handlers are not defined in the format driver, it is
> better to call the ones of the protocol driver. This enables us to
> implement snapshot support in the protocol driver.
>
> We need to call bdrv_close() and bdrv_open() handlers of
Juan Quintela wrote:
> Jan Kiszka wrote:
>> Juan Quintela wrote:
>>> Lack of "proper" subsections. IDE is something like:
>>>
>>> const VMStateDescription vmstate_ide_drive = {
>>> .version_id = 4,
>>>
>>> }
>>>
>>> static const VMStateDescription vmstate_bmdma = {
>>> .name = "ide b
The Buildbot has detected a new failure of default_x86_64_out_of_tree on
qemu-kvm.
Full details are available at:
http://buildbot.b1-systems.de/qemu-kvm/builders/default_x86_64_out_of_tree/builds/337
Buildbot URL: http://buildbot.b1-systems.de/qemu-kvm/
Buildslave for this Build: b1_qemu_kvm_1
The Buildbot has detected a new failure of default_i386_out_of_tree on qemu-kvm.
Full details are available at:
http://buildbot.b1-systems.de/qemu-kvm/builders/default_i386_out_of_tree/builds/335
Buildbot URL: http://buildbot.b1-systems.de/qemu-kvm/
Buildslave for this Build: b1_qemu_kvm_2
Bui
The Buildbot has detected a new failure of default_i386_debian_5_0 on qemu-kvm.
Full details are available at:
http://buildbot.b1-systems.de/qemu-kvm/builders/default_i386_debian_5_0/builds/398
Buildbot URL: http://buildbot.b1-systems.de/qemu-kvm/
Buildslave for this Build: b1_qemu_kvm_2
Build
The Buildbot has detected a new failure of default_x86_64_debian_5_0 on
qemu-kvm.
Full details are available at:
http://buildbot.b1-systems.de/qemu-kvm/builders/default_x86_64_debian_5_0/builds/396
Buildbot URL: http://buildbot.b1-systems.de/qemu-kvm/
Buildslave for this Build: b1_qemu_kvm_1
Linux 2.6.34 is out, and here comes the corresponding kvm-kmod package.
Download is available at
http://downloads.sourceforge.net/project/kvm/kvm-kmod/2.6.34/kvm-kmod-2.6.34.tar.bz2
KVM changes since kvm-kmod-2.6.34-rc3:
- VMX: blocked-by-sti must not defer NMI injections
- x86: Call vcpu_load
Hi,
Thank you very much for the reviewing!
At Fri, 14 May 2010 13:08:06 +0200,
Kevin Wolf wrote:
> > +
> > +struct sd_req {
> > + uint8_t proto_ver;
> > + uint8_t opcode;
> > + uint16_tflags;
> > + uint32_tepoch;
> > + uint32_tid;
> > + uint32_
This patch calls the close handler of the block driver before the qemu
process exits.
This is necessary because the sheepdog block driver releases the lock
of VM images in the close handler.
Signed-off-by: MORITA Kazutaka
---
block.c |9 +
block.h |1 +
vl.c|1 +
3 files
Hi all,
This patch adds a block driver for Sheepdog distributed storage
system.
Changes from v2 to v3 are:
- add drv->bdrv_close() and drv->bdrv_open() before and after
bdrv_snapshot_goto() call of the protocol.
- address the review comments on the sheepdog driver code.
I'll send the det
Sheepdog is a distributed storage system for QEMU. It provides highly
available block level storage volumes to VMs like Amazon EBS. This
patch adds a qemu block driver for Sheepdog.
Sheepdog features are:
- No node in the cluster is special (no metadata node, no control
node, etc)
- Linear scal
When snapshot handlers are not defined in the format driver, it is
better to call the ones of the protocol driver. This enables us to
implement snapshot support in the protocol driver.
We need to call bdrv_close() and bdrv_open() handlers of the format
driver before and after bdrv_snapshot_goto()
On 05/17/2010 12:08 PM, Sheng Yang wrote:
Convert KVM to use generic FPU API.
void kvm_load_guest_fpu(struct kvm_vcpu *vcpu)
{
if (vcpu->guest_fpu_loaded)
@@ -5156,7 +5134,7 @@ void kvm_load_guest_fpu(struct kvm_vcpu *vcpu)
vcpu->guest_fpu_loaded = 1;
unlazy_fpu(curr
On 17.05.2010, at 11:11, Michael Tokarev wrote:
> 17.05.2010 13:07, Juan Quintela wrote:
>> Michael Tokarev wrote:
> []
>>> http://www.mail-archive.com/kvm@vger.kernel.org/msg34051.html
>>> I wonder why it is not noticed before -- it's broken since 0.12...
>>
>> People has become rich and every
On 05/17/2010 12:08 PM, Sheng Yang wrote:
Change from v1:
Use unlazy_fpu() to handle host FPU, avoiding save/restore of host FPU states.
Looks good, will wait a bit for more reviews and apply.
--
Do not meddle in the internals of kernels, for they are subtle and quick to
panic.
--
To un
On 05/17/2010 12:19 PM, Sheng Yang wrote:
On Monday 17 May 2010 17:19:02 Avi Kivity wrote:
On 05/17/2010 12:08 PM, Sheng Yang wrote:
Also add some constants.
+++ b/arch/x86/kernel/process.c
@@ -28,6 +28,7 @@ unsigned long idle_nomwait;
EXPORT_SYMBOL(idle_nomwait);
struct kme
Also add some constants.
Signed-off-by: Sheng Yang
---
arch/x86/include/asm/i387.h |2 ++
arch/x86/include/asm/xsave.h |3 +++
arch/x86/kernel/i387.c |3 ++-
arch/x86/kernel/process.c|1 +
4 files changed, 8 insertions(+), 1 deletions(-)
diff --git a/arch/x86/include/
On Monday 17 May 2010 17:19:02 Avi Kivity wrote:
> On 05/17/2010 12:08 PM, Sheng Yang wrote:
> > Also add some constants.
> >
> >
> > +++ b/arch/x86/kernel/process.c
> > @@ -28,6 +28,7 @@ unsigned long idle_nomwait;
> >
> > EXPORT_SYMBOL(idle_nomwait);
> >
> > struct kmem_cache *task_xsta
On 05/17/2010 12:08 PM, Sheng Yang wrote:
Also add some constants.
+++ b/arch/x86/kernel/process.c
@@ -28,6 +28,7 @@ unsigned long idle_nomwait;
EXPORT_SYMBOL(idle_nomwait);
struct kmem_cache *task_xstate_cachep;
+EXPORT_SYMBOL(task_xstate_cachep);
int arch_dup_task_struct(struct t
17.05.2010 13:07, Juan Quintela wrote:
Michael Tokarev wrote:
[]
http://www.mail-archive.com/kvm@vger.kernel.org/msg34051.html
I wonder why it is not noticed before -- it's broken since 0.12...
People has become rich and everybody has a 64 bit hardware :-)
Actually I don't think there's a
Also add some constants.
Signed-off-by: Sheng Yang
---
arch/x86/include/asm/i387.h |2 ++
arch/x86/include/asm/xsave.h |3 +++
arch/x86/kernel/i387.c |3 ++-
arch/x86/kernel/process.c|1 +
4 files changed, 8 insertions(+), 1 deletions(-)
diff --git a/arch/x86/include/
We can avoid unnecessary fpu load when userspace process
didn't use FPU frequently.
Derived from Avi's idea.
Signed-off-by: Sheng Yang
---
arch/x86/include/asm/kvm_host.h |1 -
arch/x86/kvm/x86.c | 18 ++
2 files changed, 2 insertions(+), 17 deletions(-)
diff
Convert KVM to use generic FPU API.
Signed-off-by: Sheng Yang
---
arch/x86/include/asm/kvm_host.h | 17 +
arch/x86/kvm/x86.c | 52 ---
2 files changed, 17 insertions(+), 52 deletions(-)
diff --git a/arch/x86/include/asm/kvm_host.h
Change from v1:
Use unlazy_fpu() to handle host FPU, avoiding save/restore of host FPU states.
Sheng Yang (3):
x86: Export FPU API for KVM use
KVM: x86: Use unlazy_fpu() for host FPU
KVM: x86: Use FPU API
arch/x86/include/asm/i387.h |2 +
arch/x86/include/asm/kvm_host.h | 18 +---
Michael Tokarev wrote:
> 17.05.2010 11:00, Avi Kivity wrote:
>> On 05/16/2010 11:04 PM, Juan Quintela wrote:
> []
We've regressed from failing some migrations to failing all migrations.
>>> Humm, 0.12.4 -> 0.12.4 should work. My advise is just revert the patch
>>> and live with it for another
User allocated bitmaps have the advantage of reducing pinned memory.
However we have plenty more pinned memory allocated in memory slots, so
by itself, user allocated bitmaps don't justify this change.
In that sense, what do you think about the question I sent last week?
=== REPOST 1 ===
>>
>
On Monday 17 May 2010 16:18:22 Avi Kivity wrote:
> On 05/17/2010 11:09 AM, Sheng Yang wrote:
> > Convert KVM to use generic FPU API.
> >
> > Signed-off-by: Sheng Yang
> > ---
> > Like this? (Drop patch 1)
>
> Will be more readable with a patch that converts host_fx_image to
> unlazy_fpu(), and a
17.05.2010 11:00, Avi Kivity wrote:
On 05/16/2010 11:04 PM, Juan Quintela wrote:
[]
We've regressed from failing some migrations to failing all migrations.
Humm, 0.12.4 -> 0.12.4 should work. My advise is just revert the patch
and live with it for another week, what do you think?
A week is f
On 05/17/2010 11:09 AM, Sheng Yang wrote:
Convert KVM to use generic FPU API.
Signed-off-by: Sheng Yang
---
Like this? (Drop patch 1)
Will be more readable with a patch that converts host_fx_image to
unlazy_fpu(), and a second patch that converts guest_fx_image to the fpu
API.
I think
Convert KVM to use generic FPU API.
Signed-off-by: Sheng Yang
---
Like this? (Drop patch 1)
arch/x86/include/asm/kvm_host.h | 18 +-
arch/x86/kvm/x86.c | 70 ++-
2 files changed, 19 insertions(+), 69 deletions(-)
diff --git a/arch/x8
Michael Goldish wrote:
On 05/11/2010 12:04 PM, Jason Wang wrote:
This checker serves as the post_command to find the panic information
in the file which contains the content of guest serial console.
Signed-off-by: Jason Wang
---
client/tests/kvm/scripts/check_serial.py | 41 +
rtc test is a test for rtc driver which could also be useful for
testing the rtc emulation in kvm.
Signed-off-by: Jason Wang
---
0 files changed, 0 insertions(+), 0 deletions(-)
diff --git a/client/tests/kvm/autotest_control/rtc.control
b/client/tests/kvm/autotest_control/rtc.control
new file
This new autotest module implements a simple test for the driver of realtime
clock. It do the testing of interrupt, date reading, alarm and
frequency.
Please refer the Documentation/rtc.txt for details.
Signed-off-by: Jason Wang
---
client/tests/rtc/src/Makefile | 19 +++
client/tests/rtc/sr
In common cases, guest SRAO MCE will cause corresponding poisoned page
be un-mapped and SIGBUS be sent to QEMU-KVM, then QEMU-KVM will relay
the MCE to guest OS.
But it is reported that if the poisoned page is accessed in guest
after un-mapped and before MCE is relayed to guest OS, QEMU-KVM will
b
On 05/16/2010 11:04 PM, Juan Quintela wrote:
So, to make the story short: I know what is happening, and I know how to
fix it, just that fix is not trivial. I just need time.
Meanwhile, we have a broken 0.12.4. Is there a quick'n'dirty
workaround that will be forward compatible with
68 matches
Mail list logo