From 67806cc642cb666fb59fec60115fee37b80ecb3a Mon Sep 17 00:00:00 2001
From: Anthony Xu [EMAIL PROTECTED]
Date: Wed, 18 Jun 2008 14:32:46 -0400
Subject: [PATCH] Irq assignment
1. use bimodal _PRT
2. pci device can use irq 15, reduce interrupt sharing
3. test by running linux guest in kvm-ia64,
Le mercredi 18 juin 2008 à 00:22 -0500, Carlo Marcelo Arenas Belon a
écrit :
complement 2680f3df932a3119880a692609943587ecc8f409 to export the
definition for kvm_init_coalesced_mmio as it is used by the per
arch libkvm as shown by the following warning :
libkvm-x86.c: In function
with KVM-70
my host:
dual quad core Intel E5320
4G ram
Debian Etch
2.6.26-rc6 x86_64
my guest:
debian etch
2.6.26-rc6 x86_64
512 Mb ram
I rebuild my guest kernels without CONFIG_KVM_CLOCK, just with
CONFIG_HAVE_KVM=y
It now seems far more stable with -smp 4
started with:
Amit,
With the current implementation we have an issue if the driver on the
host was never loaded.
To be able to run kvm with passthrough we have to load and then unload
the driver on the host at least once. After that it works ok.
Note that after doing the load and unload the irq as reported
(catching up on unread email)
Chris Wright wrote:
Is this with pinning? We at least know we're losing small bits on
migration. From my measurements it's ~3000 (outliers are 10-20k).
Also, what happens if you roll back to kvm-userspace 7f5c4d15ece5?
I'm using this:
diff -up
Jerone Young wrote:
These patches remove all traces of --disable-cpu-emulation configuration flag.
I kind of like --disable-cpu-emulation. It means we have the choice of
kvm-only (low footprint, shuts up the security people), qemu-only
(running cross-arch or on older cpu), and mixed
Avi Kivity wrote:
Jerone Young wrote:
These patches remove all traces of --disable-cpu-emulation
configuration flag.
I kind of like --disable-cpu-emulation. It means we have the choice
of kvm-only (low footprint, shuts up the security people), qemu-only
(running cross-arch or on older
Mark McLoughlin wrote:
Set the VIRTIO_F_NOTIFY_ON_EMPTY feature bit so the
guest can rely on us notifying them when the queue
is empty.
Also, only notify when the available queue is empty
*and* when we've finished with all the buffers we
had detached. Right now, when the queue is empty,
we
Xu, Anthony wrote:
From 7b2c2612b21b895cd14e632fea845c03b6e1dedc Mon Sep 17 00:00:00 2001
From: Anthony Xu [EMAIL PROTECTED]
Date: Thu, 29 May 2008 13:14:56 -0400
Subject: [PATCH] Make qemu compile for kvm-ia64
Since merging with Qemu upsteram, it can't be compiled
for kvm-ia64
@@ -620,7
Akio Takebe wrote:
Hi,
This patch improves jp106 keyboard.
You can use '-k ja' option of qemu.
Please send this to qemu-devel, so non-kvm users can also make use of
this patch. Thanks. kvm will get it on the next merge.
--
I have a truly marvellous patch that fixes the bug which this
Anthony Liguori wrote:
We know exactly what the problem is. KVM activates VT
unconditionally. There's no hardware mechanism to arbitrate access to
VT. KVM is the only thing in the Linux kernel that uses VT so we
don't have a software mechanism to arbitrate access to VT.
If the VMware
Hugh Roper wrote:
Dear List,
I'm using a dual-monitor setup. When switching to fullscreen mode
(via ctrl-alt-f) is it possible to expand to the active monitor? At
the moment it centres between the monitors, which have different
resolutions.
I use Twin View Xinerama for the dual-screen
Rusty Russell wrote:
Yeah, every time I get close to benchmarking I find another bug :( But I've
spent some time optimising the normal lguest net device, so we'll have a fair
comparison.
In theory vringfd will get us zero copy from guest sendfile out to external
machines. For anything else
Mark McLoughlin wrote:
On Fri, 2008-06-13 at 17:09 +0300, Avi Kivity wrote:
Mark McLoughlin wrote:
- .ioctl = vring_ioctl,
+ .unlocked_ioctl = vring_ioctl,
+ .compat_ioctl = vring_compat_ioctl,
I think you can set compat_ioctl = vring_ioctl
Sukanto Ghosh wrote:
I saw that KVM has guest SMP support upto 4 VCPUs. Are these VCPUs
modelled as threads of 'the process representing the VM' ?
Yes.
If it is, then, when a signal is sent to the that process by the QEMU
process, shouldn't all of the VCPUs (for that VM) exit to the
Brian Jackson wrote:
Don't know if this is a kvm bug or not, but figured I'd send it anyway. Just
in case.
http://www.theiggy.com/tmp/hrtimers_bug.png
It doesn't seem to be a bug at all. Perhaps it should be converted to a
printk().
--
I have a truly marvellous patch that fixes the bug
Farkas Levente wrote:
Avi Kivity wrote:
This release was cooking for far too long, and as a result there are
plenty of changes. Windows XP SMP performance is much improved for
AMD hosts or Intel hosts without FlexPriority (performance with
FlexPriority should already be very good). There
Akio Takebe wrote:
Hi,
I got a Oops message when I shutdown a guest on x86_32 machine.
The patch fixes this issue.
Applied, thanks.
--
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
Carlo Marcelo Arenas Belon wrote:
complement 2680f3df932a3119880a692609943587ecc8f409 to export the
definition for kvm_init_coalesced_mmio as it is used by the per
arch libkvm as shown by the following warning :
Applied, thanks.
--
I have a truly marvellous patch that fixes the bug which
Inform C2 state support via ACPI's CST per-processor package.
Signed-off-by: Marcelo Tosatti [EMAIL PROTECTED]
Index: kvm-userspace.tip/bios/acpi-dsdt.dsl
===
--- kvm-userspace.tip.orig/bios/acpi-dsdt.dsl
+++
Write an invalid latency value the first time the guest attempts to idle
via P_LVL2 port.
This way the TSC is considered unreliable, and we get away with the
costs relative to APIC timer broadcasts on enter/exit necessary for C1+.
Signed-off-by: Marcelo Tosatti [EMAIL PROTECTED]
Index:
This makes it possible for PIO handlers (such as the ACPI PLVL2 one) force
the current vcpu to main_loop().
Signed-off-by: Marcelo Tosatti [EMAIL PROTECTED]
Index: kvm-userspace.tip/libkvm/libkvm.c
===
---
Enable the in-kernel C2 emulation after the _CST notification has been
triggered.
Emulate in userspace for -no-kvm-irqchip case.
Signed-off-by: Marcelo Tosatti [EMAIL PROTECTED]
Index: kvm-userspace.tip/qemu/hw/acpi.c
===
---
On Wed, Jun 18, 2008 at 11:50:52AM -0500, Charles Duffy wrote:
CONFIG_KALLSYMS is indeed defined, but kallsyms_lookup_name doesn't
appear to be exported in this kernel.
Ok I'm aware that kallsyms_lookup_name isn't exported in recent
kernels, and when I implemented the patch I verified that the
On Wed, 2008-06-18 at 06:31 -0700, Avi Kivity wrote:
Mark McLoughlin wrote:
Set the VIRTIO_F_NOTIFY_ON_EMPTY feature bit so the
guest can rely on us notifying them when the queue
is empty.
Also, only notify when the available queue is empty
*and* when we've finished with all the
i`d like to give a comment about those claims in this thread (see below),
vmware kernel part would be closed source or binary module.
i think, they arenŽt. they just are not available via public/opensource code
repository.
just go and download workstation 6.5 beta from:
Ben-Ami Yassour wrote:
On Tue, 2008-06-17 at 16:29 -0500, Anthony Liguori wrote:
I think the current VT-d code needs some reworking.
We should build the table as the shadow page table gets built. We
should suppress iotlb flushes unless the table is actually being updated.
I'm not
Avi Kivity wrote:
Avi Kivity wrote:
Jerone Young wrote:
These patches remove all traces of --disable-cpu-emulation
configuration flag.
I kind of like --disable-cpu-emulation. It means we have the choice
of kvm-only (low footprint, shuts up the security people), qemu-only
(running
Avi Kivity wrote:
Anthony Liguori wrote:
We know exactly what the problem is. KVM activates VT
unconditionally. There's no hardware mechanism to arbitrate access
to VT. KVM is the only thing in the Linux kernel that uses VT so we
don't have a software mechanism to arbitrate access to VT.
On Wed, Jun 18, 2008 at 04:02:39PM -0500, Anthony Liguori wrote:
Have we yet determined why the TSC is so unstable in the first place?
In theory, it should be relatively stable on single-node Intel and
Barcelona chips.
If the host enters C2/C3, or changes CPU frequency, it becomes
Muli Ben-Yehuda wrote:
On Wed, Jun 18, 2008 at 03:48:33PM -0500, Anthony Liguori wrote:
Right. But this is not ideal. Instead of pinning up-front, it
would make more sense IMHO to build the VT-d table as the shadow
page table gets faulted in. In certain circumstances, this will
result in
Marcelo Tosatti wrote:
On Wed, Jun 18, 2008 at 04:02:39PM -0500, Anthony Liguori wrote:
Have we yet determined why the TSC is so unstable in the first place?
In theory, it should be relatively stable on single-node Intel and
Barcelona chips.
If the host enters C2/C3, or
On Wed, Jun 18, 2008 at 04:42:40PM -0500, Anthony Liguori wrote:
Marcelo Tosatti wrote:
On Wed, Jun 18, 2008 at 04:02:39PM -0500, Anthony Liguori wrote:
Have we yet determined why the TSC is so unstable in the first
place? In theory, it should be relatively stable on single-node
Intel
On Wed, 2008-06-18 at 19:41 -0300, Marcelo Tosatti wrote:
On Wed, Jun 18, 2008 at 04:42:40PM -0500, Anthony Liguori wrote:
Marcelo Tosatti wrote:
On Wed, Jun 18, 2008 at 04:02:39PM -0500, Anthony Liguori wrote:
Have we yet determined why the TSC is so unstable in the first
place?
On 6/18/2008 3:57:16 PM, john stultz wrote:
On Wed, 2008-06-18 at 19:41 -0300, Marcelo Tosatti wrote:
On Wed, Jun 18, 2008 at 04:42:40PM -0500, Anthony Liguori wrote:
Marcelo Tosatti wrote:
On Wed, Jun 18, 2008 at 04:02:39PM -0500, Anthony Liguori wrote:
Have we yet
Hi Martin,
This lockup is fixed in the latest dot releases of Workstation, Player,
and Server.
The vmmon module now checks if VT is already enabled (actually it just
tests if CR4.VMXE is set since checking if the cpu is in VMX mode is a
pain). The vmmon module enables VMX mode when world
Hi Avi,
This builds on your preprocessor work for cleaning up the DSDT and makes
it trivially easy to change the CPU and PCI hotplug slot count. I'm
thinking about splitting a few SSDTs out of the monolithic DSDT, and
generating some of the code at compile time should make that easier.
This
Avi:
We did not get a chance to do this at the Forum. I'd be interested in
whatever options you have for reducing the scan time further (e.g., try
to get scan time down to 1-2 seconds).
thanks,
david
Avi Kivity wrote:
David S. Ahern wrote:
I gave a shot at implementing your suggestion, but
On Wed, Jun 18, 2008 at 09:29:22PM -0600, Alex Williamson wrote:
Hi Avi,
This builds on your preprocessor work for cleaning up the DSDT and makes
it trivially easy to change the CPU and PCI hotplug slot count. I'm
thinking about splitting a few SSDTs out of the monolithic DSDT, and
39 matches
Mail list logo
