Jan,
> Here are some patches that are necessary to get KVM running with the -rt4
> patchset.
Thanks a lot.
Unfortunaately, there is still a last one at kernel/smp.c:288
/* Can deadlock when called with interrupts disabled */
WARN_ON_ONCE(irqs_disabled() && !oops_in_progress);
Do
On Thu, 2009-04-09 at 23:50 +0800, Avi Kivity wrote:
> Huang Ying wrote:
> > +int kvm_set_msr_common(struct kvm_vcpu *vcpu, u32 msr, u64 data)
> > +{
> > + switch (msr) {
> > + case MSR_EFER:
> > + set_efer(vcpu, data);
> > break;
> > case MSR_IA32_DEBUGCTLMSR:
> >
Hello,
I'm currently writing my bachelor's thesis on the KVM and ran a series
of benchmarks, partly "home grown" on it. This was a few months ago. I
used the kvm package from the Debian sid repository (kvm-72) with kernel
2.6.26, which at the time were both almost up-to-date :-) (upstream KVM
Signed-off-by: Beth Kon
diff --git a/bios/acpi-dsdt.dsl b/bios/acpi-dsdt.dsl
index 06ab25d..84697db 100755
--- a/bios/acpi-dsdt.dsl
+++ b/bios/acpi-dsdt.dsl
@@ -307,6 +307,24 @@ DefinitionBlock (
,, , AddressRangeMemory, TypeStatic)
})
}
+Device(
These patches resolve the irq0->inti2 override issue, and get the hpet working
on kvm with and without -no-kvm-irqchip (i.e., when hpet takes over, it
disables userspace or in-kernel pit as appropriate).
The irq0->inti2 override will always be used unless the kernel cannot do irq
routing (i.e.,
On Thu, Apr 09, 2009 at 09:37:10AM -0700, Stephen Hemminger wrote:
> > +static int tx_ringlen = 256;
> > +module_param(tx_ringlen, int, 0444);
> > +
> > +#undef PDEBUG /* undef it, just in case */
> > +#ifdef VBUS_ENET_DEBUG
> > +# define PDEBUG(fmt, args...) printk(KERN_DEBUG "vbus_en
Is this a problem in the latest (kvm-84) or just in an older version that
Fedora 10 provides?
- Original Message -
From: "Daniel Scott"
To: "Alex Williamson"
Cc: "Brian Jackson" , kvm@vger.kernel.org
Sent: Thursday, April 9, 2009 2:54:11 PM GMT -05:00 US/Canada Eastern
Subject: Re
Brian, Alex,
2009/4/9 Alex Williamson :
> On Thu, 2009-04-09 at 10:39 -0500, Brian Jackson wrote:
>> some subjects to look for:
>> * Can't boot guest with more than 3585MB when using large pages
>> There was a patch from Marcelo that was supposed to fix it.
>> * kvm-84 and guests with more than 35
Hi,
Thanks for the response. I have read through the archives (I also did
so before posting :) ) but I haven't been able to find any information
about this. Do you know where I can find it? I tried searching for 4GB
and 3525MB and various combinations. Do you know if a specific limit
is mentioned?
On Thu, Apr 09, 2009 at 07:02:02PM +0200, Riccardo Veraldi wrote:
> Hello,
> I have always created my VM under kvm manually starting them from a .sh
> script:
>
> for example machine called abbone
>
>
> # abbone.sh
> screen qemu-kvm -nographic -hda ./abbone.img -m 512 -net
> nic,model=e1000,m
Hello,
I have always created my VM under kvm manually starting them from a .sh
script:
for example machine called abbone
# abbone.sh
screen qemu-kvm -nographic -hda ./abbone.img -m 512 -net
nic,model=e1000,macaddr=00:16:3e:05:00:07 -net tap -name abbone -vnc :0
I use screen so that I am
The x86 kvm kernel module limits guest cpu count to 16, but theuserspace pc
definition says 255 still, so kvm_create_vcpu will fail for that reason with
-smp > 16 specified. This patch causes qemu-kvm to exit in that case. Without
this patch other errors get reported down the road and finally
Avi,
Gregory Haskins wrote:
>
> Todo:
> *) Develop some kind of hypercall registration mechanism for KVM so that
>we can use that as an integration point instead of directly hooking
>kvm hypercalls
>
What would you like to see here? I now remember why I removed the
original patch I ha
On Thu, 09 Apr 2009 12:31:29 -0400
Gregory Haskins wrote:
> Signed-off-by: Gregory Haskins
> ---
>
> drivers/net/Kconfig | 13 +
> drivers/net/Makefile|1
> drivers/net/vbus-enet.c | 680
> +++
> 3 files changed, 694 insertions(+), 0
We can map these over VBUS shared memory (or really any shared-memory
architecture if it supports shm-signals) to allow asynchronous
communication between two end-points. Memory is synchronized using
pure barriers (i.e. lockless), so IOQs are friendly in many contexts,
even if the memory is remote
It will be common to map an IOQ over the VBUS shared-memory interfaces,
so lets generalize their setup so we can reuse the pattern.
Signed-off-by: Gregory Haskins
---
include/linux/vbus_device.h |7 +++
include/linux/vbus_driver.h |7 +++
kernel/vbus/Kconfig |2 +
kernel/vbu
Signed-off-by: Gregory Haskins
---
drivers/net/vbus-enet.c | 249 +--
include/linux/venet.h | 39 +++
2 files changed, 275 insertions(+), 13 deletions(-)
diff --git a/drivers/net/vbus-enet.c b/drivers/net/vbus-enet.c
index 3779f77..2a190e0 10
The ioapic code currently privately manages the mapping between irq
and vector. This results in some layering violations as the support
for certain MSI operations need this info. As a result, the MSI
code itself was moved to the ioapic module. This is not really
optimal.
We now have another nee
This adds a driver to interface between the host VBUS support, and the
guest-vbus bus model.
Signed-off-by: Gregory Haskins
---
arch/x86/Kconfig|9 +
drivers/Makefile|1
drivers/vbus/proxy/Makefile |2
drivers/vbus/proxy/kvm.c| 726
We add a new virtio transport for accessing backends located on vbus. This
complements the existing transports for virtio-pci, virtio-s390, and
virtio-lguest that already exist.
Signed-off-by: Gregory Haskins
---
drivers/virtio/Kconfig | 15 +
drivers/virtio/Makefile |1
driv
This patch adds support for guest access to a VBUS assigned to the same
context as the VM. It utilizes a IOQ+IRQ to move events from host->guest,
and provides a hypercall interface to move events guest->host.
Signed-off-by: Gregory Haskins
---
arch/x86/include/asm/kvm_para.h |1
arch/x86/
This allows userspace applications to access vbus devices
Signed-off-by: Gregory Haskins
---
include/linux/vbus.h |4
include/linux/vbus_client.h|2
include/linux/vbus_userspace.h | 48
kernel/vbus/Kconfig| 10 +
kernel/vbus/Makefile |2
This patch provides the ability to dynamically declare and map an
interrupt-request handle to an x86 8-bit vector.
Problem Statement: Emulated devices (such as PCI, ISA, etc) have
interrupt routing done via standard PC mechanisms (MP-table, ACPI,
etc). However, we also want to support a new class
We need a way to detect if a VM is reset later in the series, so lets
add a capability for userspace to signal a VM reset down to the kernel.
Signed-off-by: Gregory Haskins
---
arch/x86/kvm/x86.c |1 +
include/linux/kvm.h |2 ++
include/linux/kvm_host.h |6 ++
virt/kv
Signed-off-by: Gregory Haskins
---
drivers/vbus/devices/venet-tap.c | 235 +-
1 files changed, 228 insertions(+), 7 deletions(-)
diff --git a/drivers/vbus/devices/venet-tap.c b/drivers/vbus/devices/venet-tap.c
index 148e2c8..a1f2dc6 100644
--- a/drivers/vbus
Signed-off-by: Gregory Haskins
---
drivers/net/Kconfig | 13 +
drivers/net/Makefile|1
drivers/net/vbus-enet.c | 680 +++
3 files changed, 694 insertions(+), 0 deletions(-)
create mode 100644 drivers/net/vbus-enet.c
diff --git a/drive
This module is similar in concept to a "tuntap". A tuntap module provides
a netif() interface on one side, and a char-dev interface on the other.
Packets that ingress on one interface, egress on the other (and vice versa).
This module offers a similar concept, except that it substitues the
char-d
Signed-off-by: Gregory Haskins
---
include/linux/venet.h | 47 +++
1 files changed, 47 insertions(+), 0 deletions(-)
create mode 100644 include/linux/venet.h
diff --git a/include/linux/venet.h b/include/linux/venet.h
new file mode 100644
index
We need to get hotswap events in environments which cannot use existing
facilities (e.g. inotify). So we add a notifier-chain to allow client
callbacks whenever an interface is {un}registered.
Signed-off-by: Gregory Haskins
---
include/linux/vbus.h | 15 +
kernel/vbus/core.c |
This will generally be used for hypervisors to publish any host-side
virtual devices up to a guest. The guest will have the opportunity
to consume any devices present on the vbus-proxy as if they were
platform devices, similar to existing buses like PCI.
Signed-off-by: Gregory Haskins
---
incl
This is release v2. Changes since v1:
*) Incorporated review feedback from Stephen Hemminger on vbus-enet driver
*) Added support for connecting to vbus devices from userspace
*) Added support for a virtio-vbus transport to allow virtio drivers to
work with vbus (needs testing and backend mode
We expect to have various types of connection-clients (e.g. userspace,
kvm, etc), each of which is likely to have common access patterns and
marshalling duties. Therefore we create a "client" API to simplify
client development by helping with mundane tasks such as handle-2-pointer
translation, etc
This interface provides a bidirectional shared-memory based signaling
mechanism. It can be used by any entities which desire efficient
communication via shared memory. The implementation details of the
signaling are abstracted so that they may transcend a wide variety
of locale boundaries (e.g. u
Thanks Brian,
The kvm host with the problem is running:
* Fedora release 10 (Cambridge)
* kernel: 2.6.27 (the actual fedora package is 2.6.27.21-170.2.56.fc10.x86_64)
* 4x Quad-Core AMD Opteron(tm) Processor 8347 HE
* KVM version: 74 (the actual fedora package is (kvm-74-10.fc10.x86_64)
Read
Marcelo Tosatti wrote:
Applied, thanks.
--
Do not meddle in the internals of kernels, for they are subtle and quick to
panic.
--
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.org/m
nathan binkert wrote:
Excellent. One of the things I'm trying hard to do is keep kvm from being a
'qemu accelerator' and generally useful for other projects. That is, I'm
trying to keep the userspace interface neutral, and not to model exactly the
hardware qemu provides but allow for other conf
Zhiyong Wu wrote:
Signed-off-by: Zhiyong Wu
Applied both, thanks.
--
Do not meddle in the internals of kernels, for they are subtle and quick to
panic.
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to majord...@vger.kernel.org
More majordomo
Hi, I am a few questions on Virtio, Hope somebody clarifies them
1. What is the address type that is put inside the vring descriptor in the virt
queue while placing a request? Is it the guest virtual address or guest
physical address?
2. If it is Guest physical address, how qemu converts it to i
Huang Ying wrote:
Add MCE support to KVM. The related MSRs are emulated. A new vcpu
ioctl command KVM_X86_SETUP_MCE is used to setup MCE emulation such as
the mcg_cap. MCE is injected via vcpu ioctl command
KVM_X86_SET_MCE. Extended machine-check state (MCG_EXT_P) and CMCI are
not implemented.
S
On Thu, 2009-04-09 at 10:39 -0500, Brian Jackson wrote:
> I was thinking of a bug when using large pages, which may not be your
> problem.
> There have been other reports recently, with workarounds, that are similar,
> just not the exact same number you have.
>
> some subjects to look for:
> *
I was thinking of a bug when using large pages, which may not be your problem.
There have been other reports recently, with workarounds, that are similar,
just not the exact same number you have.
some subjects to look for:
* Can't boot guest with more than 3585MB when using large pages
There was
Hello,
now I was able to start the guest vmu with disk virtio, and some of the
tests with disk involvement even improved a bit.
But the test in which a logo is added to the video stream does not
improve. I don't know why the performance is so bad?
Subtest: Reading video locally, adding a logo to
Sheng Yang wrote:
The PCI spec said...
System software reads this field to determine the MSI-X Table Size *N*,
which is encoded as *N-1*. For example, a returned value of “011”
indicates a table size of 4.
Applied, thanks.
--
Do not meddle in the internals of kernels, for they are
Sheng Yang wrote:
Some driver(e.g. bnx2) do the following to enable MSI-X:
1. Mask all vectors.
2. Write the msg data and address.
3. Enable MSI-X
4. Unmask all the vectors.
For this, check per-vector mask bit before enable MSI-X would cause device
fail to enable MSI-X. So now we only determine
Marcelo Tosatti wrote:
Commit 46ee278652f4cbd51013471b64c7897ba9bcd1b1 causes Solaris 10
to hang on boot.
Assuming that PIT counter reads should return 0 for an expired timer
is wrong: when it is active, the counter never stops (see comment on
__kpit_elapsed).
Also arm a one shot timer for mod
Hi,
Thanks for the response. I have read through the archives (I also did
so before posting :) ) but I haven't been able to find any information
about this. Do you know where I can find it? I tried searching for 4GB
and 3525MB and various combinations. Do you know if a specific limit
is mentioned?
Signed-off-by: Mike Burns
---
0 files changed, 0 insertions(+), 0 deletions(-)
mode change 100644 => 100755 client/bin/autotestd_monitor
diff --git a/client/bin/autotestd_monitor b/client/bin/autotestd_monitor
old mode 100644
new mode 100755
--
1.5.5.6
--
To unsubscribe from this list: send
On Wed, 8 Apr 2009, Jan Blunck wrote:
> This moves the get_cpu() call down to be called after we wake up the
> waiters. Therefore the waitqueue locks can savely be rt mutex.
Applied. Thanks,
tglx
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message
On Wed, 8 Apr 2009, Jan Blunck wrote:
> This patch removes the stupid "Read locks within the self-held write lock
> succeed" behaviour. This is breaking in mm_take_all_locks() since it is quite
> common to ensure that a lock is taken with
> BUG_ON(down_read_trylock(&mm->mmap_sem)).
Good catch. T
On Wed, 8 Apr 2009, Jan Blunck wrote:
> This patch converts some KVM spin-locks to be of type raw_spinlock_t.
Applied. Thanks,
tglx
--
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
Marcelo Tosatti wrote:
On Tue, Mar 31, 2009 at 03:00:03AM +0300, Izik Eidus wrote:
this flag notify that the host physical page we are pointing to from
the spte is write protected, and therefore we cant change its access
to be write unless we run get_user_pages(write = 1).
(this is needed fo
Marcelo Tosatti wrote:
On Tue, Mar 31, 2009 at 03:00:02AM +0300, Izik Eidus wrote:
When using mmu notifiers, we are allowed to remove the page count
reference tooken by get_user_pages to a specific page that is mapped
inside the shadow page tables.
This is needed so we can balance the pageco
On Tue, Mar 31, 2009 at 03:00:03AM +0300, Izik Eidus wrote:
> this flag notify that the host physical page we are pointing to from
> the spte is write protected, and therefore we cant change its access
> to be write unless we run get_user_pages(write = 1).
>
> (this is needed for change_pte suppor
On Tue, Mar 31, 2009 at 03:00:02AM +0300, Izik Eidus wrote:
> When using mmu notifiers, we are allowed to remove the page count
> reference tooken by get_user_pages to a specific page that is mapped
> inside the shadow page tables.
>
> This is needed so we can balance the pagecount against mapcoun
Acked-by: Marcelo Tosatti
On Wed, Apr 08, 2009 at 01:58:56AM +0200, Jan Blunck wrote:
> This moves the get_cpu() call down to be called after we wake up the
> waiters. Therefore the waitqueue locks can savely be rt mutex.
>
> Signed-off-by: Jan Blunck
> Signed-off-by: Sven-Thorsten Dietrich
>
Acked-by: Marcelo Tosatti
On Wed, Apr 08, 2009 at 01:58:55AM +0200, Jan Blunck wrote:
> This patch converts some KVM spin-locks to be of type raw_spinlock_t.
>
> Signed-off-by: Jan Blunck
> ---
> arch/x86/kvm/i8254.h |2 +-
> arch/x86/kvm/irq.h |2 +-
> 2 files changed, 2 insertions
H. Peter Anvin wrote:
Glauber Costa wrote:
mov ss is a non-issue, since it is executed natively.
In real mode?
Yes.
--
I have a truly marvellous patch that fixes the bug which this
signature is too narrow to contain.
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the
Glauber Costa wrote:
How does this logic work when the instruction emulated is an STI or MOV
SS instruction? In particular, when does GUEST_INTERRUPTIBILITY_INFO
sets set to reflect the *blocking* operation?
mov ss is a non-issue, since it is executed natively.
Except in big real m
H. Peter Anvin wrote:
How does this logic work when the instruction emulated is an STI or
MOV SS instruction? In particular, when does
GUEST_INTERRUPTIBILITY_INFO sets set to reflect the *blocking* operation?
The processor sets it when emulating an sti or mov ss instruction. We
ought to s
59 matches
Mail list logo