Hello list,
I am running winXP English on the kvm, and sometimes after savevm when
starting with loadvm - it starts frozen (the process takes 100% CPU and
the target OS is not responsive).
The command line to run qemu (while kvm-intel is loaded) that I use is:
Dong, Eddie wrote:
- rename 'struct kvm_ioctl_pic' to 'struct kvm_pic_state'
- rename 'struct kvm_ioctl_ioapic' to 'struct kvm_ioapic_state'
- rearranged struct kvm_vcpu to reduce changes relative ro master
However, neither lapic4 nor lapic5 will boot Windows XP SP2
(ACPI HAL).
I'll
Dong, Eddie wrote:
Avi:
Any comments to this patch? Seems not in yet.
The sequence is apic-timer-TMCCT.patch, apic-timer-last5.patch and
apic-timer-pending4.patch.
thx,eddie
I'd like to get guest smp in first...
--
error compiling committee.c: too many arguments to
Luca wrote:
On 8/13/07, Avi Kivity [EMAIL PROTECTED] wrote:
Luca Tettamanti wrote:
The patch stops the unhandled wrmsr, but reboot is still not working
(guest is stuck using 100% of the CPU). The last working userspace is
KVM-28, and I tested it with recent kernel modules. Any idea
I think this is a really nice and important patch set. Just a couple
things:
On Sun, 2007-08-19 at 00:02 +0200, Luca Tettamanti wrote:
In this case the dyn-tick minimum res will be 1msec. I believe it
should
work ok since this is the case without any dyn-tick.
Actually minimum resolution
Laurent Vivier wrote:
Avi Kivity wrote:
[...]
The normal user/system accounting has the same issue, no? Whereever we
happen to land (kernel or user) gets the whole tick.
So I think it is okay to have the same limitation for guest time.
So this is how it looks like.
PATCH 1 and
Laurent Vivier wrote:
Other way round. In the patch that adds account_guest_time(), have a
CONFIG_GUEST_ACCOUNTING (defaulting to n, with no description, help, or
dependencies. Then, CONFIG_KVM can select GUEST_ACCOUNTING.
I was wondering in which Kconfig I can put it...
Looks
Laurent Vivier wrote:
So I think it is okay to have the same limitation for guest time.
OK, so we can go back to my first patch.
Who can decide to introduce this into the kernel ?
The sched.c parts are best merged by Ingo, and I can carry the kvm
parts. Alternatively, I can carry
Zhao, Yunfeng wrote:
Hi,Avi
Do you see any timer issue on Lapic5?
On lapic5 I found that linux guest has a big time lag while running kernel
build.
With network time the time cost is about 450s, but guest timer reports 899s.
I didn't meet the issue on lapic3.
No, I didn't. lapic3 and
Nitin A Kamble wrote:
On Fri, 2007-08-17 at 05:18 -0700, Avi Kivity wrote:
Applied, thanks.
Please keep opcodes in sorted order, and don't post patches in dos
format.
Hi Avi,
Attached is the patch, doing code cleanup. It keeps the emulation of
op-codes in sorted order. Also
Nitin A Kamble wrote:
On Fri, 2007-08-17 at 05:24 -0700, Avi Kivity wrote:
A file scope function would indeed be uglier. But the macro has more
serious problems; it's impossible to understand how variables are
affected since nobody expects function local variables whose address
is
not
Nitin A Kamble wrote:
Also, please keep opcodes in sorted order.
It is now.
Resending the updated push reg emulation patch.
Please Apply or comment.
Applied, thanks.
--
error compiling committee.c: too many arguments to function
Matthew Kent wrote:
If DYNAMIC_TICKS is defined qemu does not attepmt to generate SIGALRM at a
constant rate. Rather, the system timer is set to generate SIGALRM only
when it is needed. DYNAMIC_TICKS reduces the number of SIGALRMs sent to
idle dynamic-ticked guests.
Original patch from Dan
Dor Laor wrote:
The idea to shift/share mmu cache memory between guests is great.
You do need to take care of proper inter vm locking (kvm_lock lock,
don't mix the kvm-lock).
No, we need to stay away from kvm_lock on anything resembling a hot path.
--
error compiling committee.c: too
Izik Eidus wrote:
simply change kvm.h to work with the new lists and with blocks of 1MB
please note that the KVM_MMU_PAGES_DIVIDER is the magic, it control on
how much % memory will be allocated to the mmu pages,
KVM_MMU_PAGES_DIVIDER = 100 mean 1%
KVM_MMU_PAGES_DIVIDER = 50 mean 2%
Izik Eidus wrote:
simply make the slot memory allocation function to count how much it
allocate, and create blocks of mmu pages by this,
in addition it initiate the lists.
+++ new_kvm_main.c2007-08-19 06:21:45.0 +0300
@@ -297,6 +297,8 @@
Please use 'diff -p' (or simply
Izik Eidus wrote:
this patch all the functions that use the mmu pages
it teach them how to use the new lists.
--- mmu.c 2007-08-15 11:37:08.0 +0300
+++ new_mmu.c 2007-08-19 06:27:56.0 +0300
@@ -586,14 +586,25 @@
static struct kvm_mmu_page *kvm_mmu_lookup_page(struct
Anthony Liguori wrote:
I think this is a really nice and important patch set. Just a couple
things:
On Sun, 2007-08-19 at 00:02 +0200, Luca Tettamanti wrote:
In this case the dyn-tick minimum res will be 1msec. I believe it should
work ok since this is the case without any dyn-tick.
Dor Laor wrote:
The idea to shift/share mmu cache memory between guests is great.
You do need to take care of proper inter vm locking (kvm_lock lock,
don't mix the kvm-lock).
No, we need to stay away from kvm_lock on anything resembling a hot
path.
If we share the mmu cache globally we need
Amit Shah wrote:
On Saturday 18 August 2007 02:42:56 Koen de Jonge wrote:
Hi all,
here is a list of Supported Guests:
http://kvm.qumranet.com/kvmwiki/Guest_Support_Status
The tables are a bit unclear to me, I will explain using the 'Debian
Etch' example.
There's no guideline on
Weiyang Chen wrote:
Hi,
I build kvm-35 from tarball successfully on my Macbook core2due with
a 2.6.22.3 http://2.6.22.3 kernel, Ubuntu 7.04. But when I try to
start kvm with a Windows XP image, it gets crushed with a unhandled
vm exit: 0x9 error. Here are the last 2 lines from dmesg:
Koen de Jonge wrote:
It was very, very fast, even better than using ubuntu 7.04 but that
wasn't a fully 64 bit host os, and this debian is.
It's probably due to ubuntu 7.04 using a very old version of kvm.
Are you developers interested in warnings like these?
Tony Smolar wrote:
So you must use the standard HAL?
First can you please provide the cpu type, we had some problems on AMD.
Second, what's the host's OS? Is it 32/64 bit? kernel version.
Third, does -no-kvm works?
Some sanities just for making sure:
- Try running qemu with taskset -p 1
-
Rusty Russell wrote:
2) We either need huge descriptors or some chaining mechanism to
handle scatter-gather.
Or, my preference, have a small sglist in the descriptor; if the buffer
doesn't fit in the sglist follow a pointer and size (stored in the same
place as the
Heiko Carstens wrote:
The only thing remotely relevant in the list config is that 'Filter out
duplicate messages to list members (if possible)' is set as a default for
new members. Maybe this means that if a cc is also part of the list, that
cc is stripped (which seems a wierd
[Resend, since I didn't see this email on the list after 3h]
Hello list,
I am running winXP English on the kvm, and sometimes after savevm when
starting with loadvm - it starts frozen (the process takes 100% CPU and
the target OS is not responsive).
The command line to run qemu (while kvm-intel
Shahar Livne wrote:
[Resend, since I didn't see this email on the list after 3h]
[The list won't echo your own messages]
Hello list,
I am running winXP English on the kvm, and sometimes after savevm when
starting with loadvm - it starts frozen (the process takes 100% CPU and
the
The attached patch should help.
Shahar Livne wrote:
Hello list,
I am running winXP English on the kvm, and sometimes after savevm when
starting with loadvm - it starts frozen (the process takes 100% CPU and
the target OS is not responsive).
The command line to run qemu (while kvm-intel is
Markus Hitter wrote:
Am 19.08.2007 um 12:35 schrieb Avi Kivity:
- this deviates from standard Unix (and Windows) practice.
Arguably the OS X way is superior, but qemu is perhaps not the best
vehicle to introduce it to other OSes.
NeXTstep introduced this into the Unix world
Avi Kivity wrote:
This allows save/restore (and live migration on kvm) of smp guests.
Index: hw/apic.c
===
RCS file: /sources/qemu/qemu/hw/apic.c,v
retrieving revision 1.13
diff -u -r1.13 apic.c
--- hw/apic.c 3 Apr 2007
Jindrich Makovicka wrote:
On 8/17/07, Avi Kivity [EMAIL PROTECTED] wrote:
That page doesn't work for me -- it probably expects some HTTP POST
data. Can you set down instructions to reproduce? (i.e. starting from
java.sun.com or similar)
Sure.
1) go to http://java.com
2) click
Hi Uri.
Thanks for the patch.
First tests show that I still have the freeze, but I will check it further.
Regards,
Shahar
Uri Lublin wrote:
The attached patch should help.
-
This SF.net email is sponsored by:
On Sun, 19 Aug 2007 14:23:49 +0300
Avi Kivity [EMAIL PROTECTED] wrote:
Jindrich Makovicka wrote:
On 8/17/07, Avi Kivity [EMAIL PROTECTED] wrote:
That page doesn't work for me -- it probably expects some HTTP POST
data. Can you set down instructions to reproduce? (i.e. starting
from
Hi .
I'm trying to run KVM on Feisty, 64 bit Intel with the -net nic -net
tap switches so I can use it in bridged network mode.
It refuses and I have to run it sudo which for security concerns I don't
want to do.
The error message is : warning: could not configure /dev/net/tun: no
virtual
When taking a cpu down, we need to hardware_disable() it. Unfortunately,
the CPU_DYING notifier is called with interrupts disabled, which means
we can't use smp_call_function_single(). Fortunately, the CPU_DYING notifier
is always called on the dying cpu, so we don't need to use the function at
On 8/19/07, Eugene Coetzee [EMAIL PROTECTED] wrote:
Hi .
I'm trying to run KVM on Feisty, 64 bit Intel with the -net nic -net
tap switches so I can use it in bridged network mode.
It refuses and I have to run it sudo which for security concerns I don't
want to do.
The error message is :
Avi Kivity wrote:
In this case the dyn-tick minimum res will be 1msec. I believe it should
work ok since this is the case without any dyn-tick.
Actually minimum resolution depends on host HZ setting, but - yes -
essentially you have the same behaviour of the unix timer, plus the
Jamie Lokier wrote:
Avi Kivity wrote:
In this case the dyn-tick minimum res will be 1msec. I believe it should
work ok since this is the case without any dyn-tick.
Actually minimum resolution depends on host HZ setting, but - yes -
essentially you have the same
Yes, good thinking, but this should only be done if it actually impacts
something. Reducing overhead from 0.1% to 0.05% is not worthwhile if it
introduces extra complexity.
If the overhead is that small, why are we touching this code in the first
place?
Paul
Luca wrote:
On 8/19/07, Eugene Coetzee [EMAIL PROTECTED] wrote:
Hi .
I'm trying to run KVM on Feisty, 64 bit Intel with the -net nic -net
tap switches so I can use it in bridged network mode.
It refuses and I have to run it sudo which for security concerns I don't
want to do.
You need
Yes, good thinking, but this should only be done if it actually
impacts
something. Reducing overhead from 0.1% to 0.05% is not worthwhile
if
it
introduces extra complexity.
If the overhead is that small, why are we touching this code in the
first
place?
Accuracy is much more important
On 8/19/07, Eugene Coetzee [EMAIL PROTECTED] wrote:
Luca wrote:
On 8/19/07, Eugene Coetzee [EMAIL PROTECTED] wrote:
Hi .
I'm trying to run KVM on Feisty, 64 bit Intel with the -net nic -net
tap switches so I can use it in bridged network mode.
It refuses and I have to run it sudo which
On 8/19/07, Luca Tettamanti [EMAIL PROTECTED] wrote:
+static uint64_t qemu_next_deadline(void) {
+uint64_t nearest_delta_us = ULLONG_MAX;
+uint64_t vmdelta_us;
Hum, I introduced a bug here... those vars should be signed.
On the overhead introduced: how do you measure it?
Luca
Paul Brook wrote:
Yes, good thinking, but this should only be done if it actually
impacts
something. Reducing overhead from 0.1% to 0.05% is not worthwhile
if it
introduces extra complexity.
If the overhead is that small, why are we touching this code in the
first
place?
Insightful.
A
Jamie Lokier wrote:
Paul Brook wrote:
Yes, good thinking, but this should only be done if it actually impacts
something. Reducing overhead from 0.1% to 0.05% is not worthwhile if it
introduces extra complexity.
If the overhead is that small, why are we touching this code in the
Luca wrote:
On 8/19/07, Luca Tettamanti [EMAIL PROTECTED] wrote:
+static uint64_t qemu_next_deadline(void) {
+uint64_t nearest_delta_us = ULLONG_MAX;
+uint64_t vmdelta_us;
Hum, I introduced a bug here... those vars should be signed.
On the overhead introduced: how do you
Il Sun, Aug 19, 2007 at 10:36:06AM +0300, Avi Kivity ha scritto:
Luca wrote:
On 8/13/07, Avi Kivity [EMAIL PROTECTED] wrote:
Luca Tettamanti wrote:
The patch stops the unhandled wrmsr, but reboot is still not working
(guest is stuck using 100% of the CPU). The last working
Jeff Garzik wrote:
Avi Kivity wrote:
Heiko Carstens wrote:
The only thing remotely relevant in the list config is that
'Filter out duplicate messages to list members (if possible)' is
set as a default for new members. Maybe this means that if a cc
is also part of the list, that cc is
48 matches
Mail list logo