Laurent Vivier wrote:
Compilation fails on my system (Debian etch on x86_64)
gcc -I /home/vivierl/kvm-25/kernel/include -MMD -MF test/.bootstrap.d -g
-fomit-frame-pointer -Wall -m64 -fno-stack-protector -std=gnu99
-ffreestanding -c -nostdlib -o test/bootstrap.o test/bootstrap.S
Hi all!
KVM is great!
I am successfully running Windows Xp on KVM-24 in Fedora 7 test 4. I
have even connected a CDROM drive, which was a bit of hassle.
One problem I have is keyboard layout. I need a Norwegian keyboard
layout to work, and I have installed this in the Linux host and in the
Li, Xin B wrote:
So you will check in the patch soon?
Yes. Still need to test AMD and some odd guests.
Really a good news since this will be the base for enabling SMP.
SMP, faster paravirt drivers, TPR shadow on vmx...
--
error compiling committee.c: too many arguments to function
Uwe Kubosch wrote:
On Thu, 2007-05-24 at 12:55 +0300, Avi Kivity wrote:
One problem I have is keyboard layout. I need a Norwegian keyboard
layout to work, and I have installed this in the Linux host and in the
Windows XP guest, but no Norwegian characters show, and characters
like
Uwe Kubosch wrote:
If I skip the virt-manager, and start from the command line, could you
give me an example for using sdl? Something like
qemu-kvm nice.options.here my.image.file
qemu-kvm is the right program to use, right?
qemu-kvm -m 1024 your-image-file
For virtual machine with
On Thu, 2007-05-24 at 13:56 +0300, Avi Kivity wrote:
Uwe Kubosch wrote:
If I skip the virt-manager, and start from the command line, could you
give me an example for using sdl? Something like
qemu-kvm nice.options.here my.image.file
qemu-kvm is the right program to use, right?
On Thu, May 24, 2007 at 4:29 AM, in message [EMAIL PROTECTED],
Avi Kivity [EMAIL PROTECTED] wrote:
Gregory Haskins wrote:
Halting in userspace requires a relatively cumbersome mechanism to signal
the
halted VCPU. Implementing halt in kernel should be relatively straight
forward and it
On Thu, May 24, 2007 at 5:38 AM, in message
[EMAIL PROTECTED], Li,
Xin B [EMAIL PROTECTED] wrote:
So you will check in the patch soon?
Really a good news since this will be the base for enabling SMP.
- Xin
Our plan is to run regressions tests with it at by Monday to
check it in
if there are
CPU_DYING is called in atomic context, so don't try to take any locks.
Signed-off-by: Avi Kivity [EMAIL PROTECTED]
---
kernel/cpuset.c |3 +++
1 files changed, 3 insertions(+), 0 deletions(-)
diff --git a/kernel/cpuset.c b/kernel/cpuset.c
index f57854b..d4ab1c6 100644
--- a/kernel/cpuset.c
CPU_DYING is notified in atomic context, so no taking mutexes here.
Signed-off-by: Avi Kivity [EMAIL PROTECTED]
---
arch/i386/kernel/cpu/mcheck/therm_throt.c |6 --
1 files changed, 4 insertions(+), 2 deletions(-)
diff --git a/arch/i386/kernel/cpu/mcheck/therm_throt.c
The following patchset makes kvm more robust wrt cpu hotunplug, and
makes suspend-to-ram actually work. Suspend-to-disk benefits from
the cpu hotunplug improvements as well.
The major issue is that KVM wants to disable the virtualization
extensions at a point in time when no user processes are
KVM wants a notification when a cpu is about to die, so it can disable
hardware extensions, but at a time when user processes cannot be scheduled
on the cpu, so it doesn't try to use virtualization extensions after they
have been disabled.
This adds a CPU_DYING notification. The notification is
Only at the CPU_DYING stage can we be sure that no user process will
be scheduled onto the cpu and oops when trying to use virtualization
extensions.
Signed-off-by: Avi Kivity [EMAIL PROTECTED]
---
drivers/kvm/kvm_main.c |4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git
This defines on_one_cpu() which is similar to smp_call_function_single()
except that it works if cpu happens to be the current cpu. Can also be
seen as a complement to on_each_cpu() (which also doesn't treat the
current cpu specially).
Signed-off-by: Avi Kivity [EMAIL PROTECTED]
---
Gregory Haskins wrote:
We have a specialized signal mask for the kernel, and it is already in
effect here. See KVM_SET_SIGNAL_MASK.
KVM_SET_SIGNAL_MASK is precisely what I was trying to counteract ;).
Basically I was thinking that userspace could have set some arbitrary
So as long as any appropriate signal will kick the HLT out of this wait-queue,
we are ok. It sounds like I can just get rid of the sigprocmask all together
and this will work? If so, I will pull that out for the v8 drop I am about
to make (rebased on git-head)
On Thu, May 24, 2007 at 8:17
Gregory Haskins wrote:
So as long as any appropriate signal will kick the HLT out of this
wait-queue, we are ok. It sounds like I can just get rid of the sigprocmask
all together and this will work?
Yes.
--
error compiling committee.c: too many arguments to function
Avi Kivity wrote:
The following patchset makes kvm more robust wrt cpu hotunplug, and
makes suspend-to-ram actually work. Suspend-to-disk benefits from
the cpu hotunplug improvements as well.
Here's the patchset diffstat in case anyone's interested:
Changelog: v7 plus the following:
1) Rebased on git-head
2) Removed level-senstive patch. I re-read the spec more carefully and we had
it right the first time.
3) Got rid of sigprocmask from vcpu_halt per comments
Passes the usual tests, and offers an 11% speedup. There is another
Signed-off-by: Gregory Haskins [EMAIL PROTECTED]
---
drivers/kvm/kvm.h | 60 +++
drivers/kvm/kvm_main.c | 94 ++--
2 files changed, 142 insertions(+), 12 deletions(-)
diff --git a/drivers/kvm/kvm.h
Signed-off-by: Gregory Haskins [EMAIL PROTECTED]
---
drivers/kvm/vmx.c |6 +++---
1 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/kvm/vmx.c b/drivers/kvm/vmx.c
index bfd5f8e..3411813 100644
--- a/drivers/kvm/vmx.c
+++ b/drivers/kvm/vmx.c
@@ -1977,13 +1977,13 @@ static
The current code is geared towards using a user-mode (A)PIC. This patch adds
an irqdevice abstraction, and implements a userint model to handle the
duties of the original code. Later, we can develop other irqdevice models
to handle objects like LAPIC, IOAPIC, i8259, etc, as appropriate
The VCPU executes synchronously w.r.t. userspace today, and therefore
interrupt injection is pretty straight forward. However, we will soon need
to be able to inject interrupts asynchronous to the execution of the VCPU
due to the introduction of SMP, paravirtualized drivers, and asynchronous
Signed-off-by: Gregory Haskins [EMAIL PROTECTED]
---
drivers/kvm/Makefile |2
drivers/kvm/kernint.c | 149 +
drivers/kvm/kvm.h | 35 +
drivers/kvm/kvm_main.c | 198 ++-
drivers/kvm/lapic.c| 1418
drivers/kvm/svm.c
Signed-off-by: Gregory Haskins [EMAIL PROTECTED]
---
drivers/kvm/kvm.h |6 ++
drivers/kvm/kvm_main.c | 23 ---
drivers/kvm/vmx.c |2 ++
3 files changed, 28 insertions(+), 3 deletions(-)
diff --git a/drivers/kvm/kvm.h b/drivers/kvm/kvm.h
index
Halting in userspace requires a relatively cumbersome mechanism to signal the
halted VCPU. Implementing halt in kernel should be relatively straight
forward and it eliminates the need for the signaling
Signed-off-by: Gregory Haskins [EMAIL PROTECTED]
---
drivers/kvm/kvm.h |3 ++
v7 + rebase on git-head
-Greg
-
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
Non-performance critical code is made more awkward by having to always define
both #ifdef KVM and if (kvm_allowed). Define kvm_allowed = 0 by
default. Anthony Ligouri is credited with the idea.
Signed-off-by: Gregory Haskins [EMAIL PROTECTED]
---
qemu/qemu-kvm.c |9 -
1 files
Newer userspace may run on an older kernel. Therefore we need a way to
check the capabilities of the kernel so that we can downgrade userspace
dynamically if necessary
Signed-off-by: Gregory Haskins [EMAIL PROTECTED]
---
qemu/qemu-kvm.c | 27 +++
user/kvmctl.c |
Signed-off-by: Gregory Haskins [EMAIL PROTECTED]
---
qemu/hw/apic.c | 35 ++-
qemu/hw/pc.c| 34 --
qemu/qemu-kvm.c | 49 +++--
qemu/qemu-kvm.h |2 ++
qemu/vl.h | 11
There is still a minor issue with the in-kernel APIC (even in v8)
w.r.t. informing QEMU about irq-windows. As a result, QEMU thinks there are
more windows than there actually are, and pushes another round of interrupts
down into the kernel. Since PICs can generally only queue 1-2 interrupts,
On Thu, May 24, 2007 at 03:10:12PM +0300, Avi Kivity wrote:
This defines on_one_cpu() which is similar to smp_call_function_single()
except that it works if cpu happens to be the current cpu. Can also be
seen as a complement to on_each_cpu() (which also doesn't treat the
current cpu
Heiko Carstens wrote:
On Thu, May 24, 2007 at 03:10:12PM +0300, Avi Kivity wrote:
This defines on_one_cpu() which is similar to smp_call_function_single()
except that it works if cpu happens to be the current cpu. Can also be
seen as a complement to on_each_cpu() (which also doesn't treat
I don't see any documented restrictions about preemption being
disabled when this function is called, but...
+int on_one_cpu(int cpu, void (*func) (void *info), void *info,
+ int retry, int wait)
+{
+int ret;
+int this_cpu;
+
+this_cpu = get_cpu();
what if a
Author: Markus Rechberger [EMAIL PROTECTED]
kvm-25 won't compile against the latest git tree due missing header
files.
/devel/kvm-25/kernel# make
make -C /lib/modules/2.6.22-rc2/build M=`pwd` $@
make[1]: Entering directory `/usr/src/linux-2.6'
LD /devel/kvm-25/kernel/built-in.o
CC [M]
With some linux livecds qemu crash before i see the (grub?) start menu.
images which crash: kubuntu-7.04-desktop-i386.iso or
openSUSE-10.2-GM-LiveDVD.iso
image which dont crash: dsl-3.3.iso
Laurent Vivier wrote:
Hi,
I made some benchmarks with dbench on linux/KVM.
IDE means I use an image disk with -hda
SCSI means I use an image disk with -sda
(see the patch I proposed on qemu list:
http://www.mail-archive.com/[EMAIL PROTECTED]/msg08679.html )
threads KVM-21 KVM-21 KVM-24
Anthony Liguori wrote:
Laurent Vivier wrote:
Hi,
I made some benchmarks with dbench on linux/KVM.
IDE means I use an image disk with -hda
SCSI means I use an image disk with -sda
(see the patch I proposed on qemu list:
http://www.mail-archive.com/[EMAIL PROTECTED]/msg08679.html )
Gregory Haskins wrote:
Passes the usual tests, and offers an 11% speedup. There is another
enhancement not included here (coming shortly) which exposes another 12%
speedup over this as well. Stay tuned
This fails on amd, locks up completely. Bisecting now. lapic.c needs a
#include
Gregory Haskins wrote:
1) I think I really want *any* signal to kick the HLT, so I temporarily
unmask everything while halted.
Acutally that'll cause trouble if people have device drivers in user
space, which mask SIGIO for the CPU threads like our s390 prototype
does. Consider leaving the
Avi Kivity wrote:
Gregory Haskins wrote:
Passes the usual tests, and offers an 11% speedup. There is another
enhancement not included here (coming shortly) which exposes another 12%
speedup over this as well. Stay tuned
This fails on amd, locks up completely. Bisecting now. lapic.c
On Thu, May 24, 2007 at 10:57 AM, in message [EMAIL PROTECTED],
Carsten Otte [EMAIL PROTECTED] wrote:
Gregory Haskins wrote:
1) I think I really want *any* signal to kick the HLT, so I temporarily
unmask everything while halted.
Acutally that'll cause trouble if people have device drivers
Gregory Haskins wrote:
On Thu, May 24, 2007 at 10:58 AM, in message [EMAIL PROTECTED],
Avi Kivity [EMAIL PROTECTED] wrote:
Avi Kivity wrote:
Gregory Haskins wrote:
Passes the usual tests, and offers an 11% speedup. There is another
enhancement not included here
On Thu, May 24, 2007 at 11:06 AM, in message [EMAIL PROTECTED],
Avi Kivity [EMAIL PROTECTED] wrote:
Gregory Haskins wrote:
I bet it has something to do with the local_irq_disable vs clgi in the
kvm_run() code.
clgi() overrides local irq, so I don't see how that happens.
Im just
Laurent Vivier wrote:
Anthony Liguori wrote:
Hi Anthony,
I didn't modify aio_init limit...
It would be pretty useful to give it a try. In qemu_aio_init() where
aio_init() get's called. If you just bump aio_threads and aio_num from
1 to, say, 32. You should see a big difference in
Axel Kittenberger wrote:
Hello List!
I just downloaded kvm-25 and get following error while trying compile.
Avi just fixed this in git.
Regards,
Anthony Liguori
(Because the kvm 0.8 that came with my distro, always crashed after 1
second into trying to boot an qemu-installed Windows
On Thu, 2007-05-24 at 20:10 +0800, Avi Kivity wrote:
The following patchset makes kvm more robust wrt cpu hotunplug, and
makes suspend-to-ram actually work. Suspend-to-disk benefits from
the cpu hotunplug improvements as well.
The major issue is that KVM wants to disable the virtualization
47 matches
Mail list logo