Hello,
I was trying to enable the use of nmi watchdog within a linux guest running in
kvm. I have done it
by allowing direct access to perfmon msrs using the MSR_BITMAP field in vmcs
region.
Most of the times the NMI Watchdog Test in the guest fails, but with a finite
number of NMI's
* Balaji Rao [EMAIL PROTECTED] wrote:
I was trying to enable the use of nmi watchdog within a linux guest
running in kvm. I have done it by allowing direct access to perfmon
msrs using the MSR_BITMAP field in vmcs region.
your patch looks pretty clean and simple to me - and it would make
Anthony Liguori wrote:
This ABI breakage is worrying. While it is expected to take some
time for the ABI to congeal, we need some way to prevent mismatched
guests and hosts from running. Perhaps something like the kvm abi
version, storedin the pci revision field? Keep incrementing it
Jerone Young wrote:
This set of patches is to enable PowerPC embedded KVM capablities to be taken
advantage of by qemu.
The code currently boots a ungziped uImage of the Linux kernel complied for
PowerPC 440 AMCC Bamboo evalution board. We have not fully tested userspace,
but fixes will
Jeremy Fitzhardinge wrote:
Avi Kivity wrote:
I find it non-descriptive, and it reminds me of another hypervisor.
I suggest 'tlp' for two-level paging.
That has its own ambiguity; without other context it reads like
two-level pagetable. Anyway, using the same term for the same
Yang, Sheng wrote:
From af5d9ee2a189c2f0998c5c46d1fcd25c5cb72f8c Mon Sep 17 00:00:00 2001
From: Sheng Yang [EMAIL PROTECTED]
Date: Mon, 28 Jan 2008 05:10:22 +0800
Subject: [PATCH] KVM: In-kernel PIT model
The patch moved PIT from userspace to kernel, and increase the timer accuracy
Yang, Sheng wrote:
+void kvm_create_pit(kvm_context_t kvm)
+{
+ int r;
+
+ if (!kvm-no_pit_creation) {
+ r = ioctl(kvm-fd, KVM_CHECK_EXTENSION, KVM_CAP_PIT);
+ if (r 0) {
+ r = ioctl(kvm-vm_fd, KVM_CREATE_PIT);
+ if
Yang, Sheng wrote:
Hi,
Here is the updated in-kernel PIT model. The main improvements are:
Looking good; apart from some very minor comments all that is missing is
save/restore support.
--
error compiling committee.c: too many arguments to function
On Mon, Jan 28, 2008 at 01:48:40PM +0530, Balaji Rao wrote:
Hello,
I was trying to enable the use of nmi watchdog within a linux guest running
in kvm. I have done it
by allowing direct access to perfmon msrs using the MSR_BITMAP field in vmcs
region.
Is there a proper virtualization of
Jerone Young wrote:
The issue here is that qemu does not support u-boot firmware. What we
are doing is actually loading a u-boot image and booting Linux directly
with a pre made device tree (which is something normally u-boot firmware
would provide Linux). So this code wouldn't fit into plain
On Monday 28 January 2008 08:30:32 pm Joerg Roedel wrote:
On Mon, Jan 28, 2008 at 08:12:26PM +0530, Balaji Rao wrote:
On Monday 28 January 2008 05:59:25 pm Joerg Roedel wrote:
This patch kills the ability of KVM to migrate between AMD and Intel
because the Intel performance counters are
Avi Kivity wrote:
Rusty Russell wrote:
+
+static void vp_reset(struct virtio_device *vdev)
+{
+struct virtio_pci_device *vp_dev = to_vp_device(vdev);
+/* 0 status means a reset. */
+return iowrite8(0, vp_dev-ioaddr + VIRTIO_PCI_STATUS);
}
pci has something called FLR for
Andrea Arcangeli wrote:
On Thu, Jan 24, 2008 at 09:56:06PM -0800, Christoph Lameter wrote:
Andrea's mmu_notifier #4 - RFC V1
- Merge subsystem rmap based with Linux rmap based approach
- Move Linux rmap based notifiers out of macro
- Try to account for what locks are held while the
Hi,
The CR3 cache feature of VMX CPU's does not seem to increase
context switch performance significantly as it did in the original
implementation (http://lkml.org/lkml/2007/1/5/205).
The following is similar to the original, but it also caches roots for
4-level pagetables on x86-64, and
Anthony Liguori wrote:
I'm worried about the ramp up to 2.6.25 causing confusion among users
as before that things will break left and right, if we don't provide
a tighter check.
Well, using the PCI revision ID isn't a bad idea. It wouldn't have
helped this last problem because that
On Mon, Jan 28, 2008 at 06:10:39PM +0200, Izik Eidus wrote:
i dont understand how is that better than notification on tlb flush?
I certainly agree. The quoted call wasn't actually the one that could
be moved in a single place in the .h file though. But the 4/4 patch
could be reduced to a few
Marcelo Tosatti wrote:
So the questions are, why the benefit is not similar (vm-exit's are now
way cheaper?) and is it worth to merge this given the results?
Yes, vmexits are now about 5 times faster compared to older kvm.
+
+/*
+ * Avoid the VM exit upon cr3 load by using the cached
This is true (you got me there). I'll test it out and submit it to
qemu-devel.
On Mon, 2008-01-28 at 10:17 -0600, Anthony Liguori wrote:
From:
Anthony Liguori
[EMAIL PROTECTED]
To:
[EMAIL PROTECTED]
Anthony Liguori wrote:
PCI-e has a common reset concept (warm and cold). I've been looking
around and I can't seem to find any common reset mechanism for PCI.
Is FLR something that is per-device or a standard PCI mechanism? If
it's the former, than we've basically implemented FLR using
Dor Laor wrote:
On Mon, 2008-01-28 at 09:32 -0600, Anthony Liguori wrote:
Hi Dor,
How are you measuring performance? The numbers I've gotten with netperf
before and after your patch are:
tx - 647.27mbit
rx - 89.22
tx - 27.82
rx - 79.93
I've been testing with iperf
Hi Dor,
How are you measuring performance? The numbers I've gotten with netperf
before and after your patch are:
tx - 647.27mbit
rx - 89.22
tx - 27.82
rx - 79.93
So this patch is pretty much killing performance for netperf.
Dor Laor wrote:
There was a problem with the location of the
On Sat, 26 Jan 2008, Robin Holt wrote:
No you cannot do that because there are still callbacks that come later.
The invalidate_all may lead to invalidate_range() doing nothing for this
mm. The ops notifier and the freeing of the structure has to wait until
release().
Could you be a
On Fri, Jan 25, 2008 at 05:08:41PM +0100, Guido Guenther wrote:
On Thu, Jan 24, 2008 at 01:52:16PM +0200, Uri Lublin wrote:
Sometimes after lodavm guest network does work better with '-no-kvm-irqchip'
I am trying to use bisection to find the problematic patch.
Another migration problem that
On Mon, Jan 28, 2008 at 06:37:08PM +0100, Ingo Molnar wrote:
try running your shell on one CPU:
taskset -p 1 $$
that should give you the 'raw' single-CPU context switch results.
With pinning ~= 3.30, and on an UP guest still ~= 6. I'll check the
vmexits later, thanks.
* Marcelo Tosatti [EMAIL PROTECTED] wrote:
when i did the testing then i was able to get zero VM exits in the
lat_ctx hotpath and get the same performance as on native. The above
numbers you got suggest that this is not the case. (your lat_ctx
numbers on native are probably around 1
As Avi pointed out, as we continue to massage the virtio PCI ABI, we can make
things a little more friendly to users by utilizing the PCI revision field to
indicate which version of the ABI we're using. This is a hard ABI version
and incrementing it will cause the guest driver to break.
This is
On Mon, 2008-01-28 at 09:32 -0600, Anthony Liguori wrote:
Hi Dor,
How are you measuring performance? The numbers I've gotten with netperf
before and after your patch are:
tx - 647.27mbit
rx - 89.22
tx - 27.82
rx - 79.93
I've been testing with iperf (patched with Ingo's fix).
I
On Monday 28 January 2008 05:59:25 pm Joerg Roedel wrote:
On Mon, Jan 28, 2008 at 01:48:40PM +0530, Balaji Rao wrote:
Hello,
I was trying to enable the use of nmi watchdog within a linux guest
running in kvm. I have done it by allowing direct access to perfmon msrs
using the MSR_BITMAP
The issue here is that qemu does not support u-boot firmware. What we
are doing is actually loading a u-boot image and booting Linux directly
with a pre made device tree (which is something normally u-boot firmware
would provide Linux). So this code wouldn't fit into plain qemu just
yet.
On Sun,
On Mon, 2008-01-28 at 10:17 -0600, Anthony Liguori wrote:
Jerone Young wrote:
The issue here is that qemu does not support u-boot firmware. What we
are doing is actually loading a u-boot image and booting Linux directly
with a pre made device tree (which is something normally u-boot
On Sat, 26 Jan 2008, Robin Holt wrote:
But what if the caller is already holding the mmap_sem? Why force the
acquire into this function? Since we are dealing with a semaphore/mutex,
Then you need to call __register_mmu_notifier.
Avi Kivity wrote:
Anthony Liguori wrote:
I'm worried about the ramp up to 2.6.25 causing confusion among
users as before that things will break left and right, if we don't
provide a tighter check.
Well, using the PCI revision ID isn't a bad idea. It wouldn't have
helped this last
Bugs item #1881418, was opened at 2008-01-28 21:55
Message generated for change (Tracker Item Submitted) made by Item Submitter
You can respond by visiting:
https://sourceforge.net/tracker/?func=detailatid=893831aid=1881418group_id=180599
Please note that this message will contain a full copy of
Core code for mmu notifiers.
Signed-off-by: Christoph Lameter [EMAIL PROTECTED]
Signed-off-by: Andrea Arcangeli [EMAIL PROTECTED]
---
include/linux/list.h | 14 ++
include/linux/mm_types.h |6 +
include/linux/mmu_notifier.h | 210 +++
when a task exits we can remove all external pts at once. At that point the
extern mmu may also unregister itself from the mmu notifier chain to avoid
future calls.
Note the complications because of RCU. Other processors may not see that the
notifier was unlinked until a quiescent period has
These notifiers here use the Linux rmaps to perform the callbacks.
In order to walk the rmaps locks must be held. Callbacks can therefore
only operate in an atomic context.
Signed-off-by: Christoph Lameter [EMAIL PROTECTED]
---
mm/rmap.c | 12 +++-
1 file changed, 11 insertions(+), 1
Callbacks to remove individual pages if the subsystem has an
rmap capability. The pagelock is held but no spinlocks are held.
The refcount of the page is elevated so that dropping the refcount
in the subsystem will not directly free the page.
The callbacks occur after the Linux rmaps have been
On Mon, 28 Jan 2008, Andrea Arcangeli wrote:
With regard to the synchronize_rcu troubles they also be left to the
notifier-user to solve. Certainly having the synchronize_rcu like in
Ahh. Ok.
-
This SF.net email is
* Marcelo Tosatti [EMAIL PROTECTED] wrote:
lat_ctx numbers (output is nr-procs overhead-in-us):
cr3-cache:
size=0k ovr=1.30
2 6.63
size=0k ovr=1.31
4 7.43
size=0k ovr=1.32
8 11.02
when i did the testing then i was able to get zero VM exits in the
lat_ctx hotpath and get the same
Avi Kivity wrote:
Anthony Liguori wrote:
This ABI breakage is worrying. While it is expected to take some
time for the ABI to congeal, we need some way to prevent mismatched
guests and hosts from running. Perhaps something like the kvm abi
version, storedin the pci revision field? Keep
On Mon, Jan 28, 2008 at 08:12:26PM +0530, Balaji Rao wrote:
On Monday 28 January 2008 05:59:25 pm Joerg Roedel wrote:
This patch kills the ability of KVM to migrate between AMD and Intel
because the Intel performance counters are not available on AMD and vice
verca.
Yes. The way we
mmu core: Need to use hlist_del
Wrong type of list del in mmu_notifier_release()
Signed-off-by: Christoph Lameter [EMAIL PROTECTED]
---
mm/mmu_notifier.c |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
Index: linux-2.6/mm/mmu_notifier.c
Problem for external rmaps: There is no pagelock held on the page.
Signed-off-by: Robin Holt [EMAIL PROTECTED]
---
mm/filemap_xip.c |5 +
1 file changed, 5 insertions(+)
Index: linux-2.6/mm/filemap_xip.c
===
---
The invalidation of address ranges in a mm_struct needs to be
performed when pages are removed or permissions etc change.
Most of the VM address space changes can use the range invalidate
callback.
invalidate_range() is generally called with mmap_sem held but
no spinlocks are active. If
Avi Kivity wrote:
Anthony Liguori wrote:
PCI-e has a common reset concept (warm and cold). I've been looking
around and I can't seem to find any common reset mechanism for PCI.
Is FLR something that is per-device or a standard PCI mechanism? If
it's the former, than we've basically
# HG changeset patch
# User Hollis Blanchard [EMAIL PROTECTED]
# Date 1201563731 21600
# Node ID 6a2f4869cf5da00fa0ccacc9188ea493459ce4cb
# Parent a6dd6b5a597903a069cb9711f73ea1b5c4f0b764
This allows kvm_host.h to be #included even when struct preempt_notifier is
undefined. This is needed to
On Mon, Jan 28, 2008 at 06:17:34PM +0100, Ingo Molnar wrote:
* Marcelo Tosatti [EMAIL PROTECTED] wrote:
lat_ctx numbers (output is nr-procs overhead-in-us):
cr3-cache:
size=0k ovr=1.30
2 6.63
size=0k ovr=1.31
4 7.43
size=0k ovr=1.32
8 11.02
when i did the testing then i
On Mon, 28 Jan 2008, Andrea Arcangeli wrote:
So I'd like to know what can we do to help to merge the 4 patches from
Christoph in mainline, I'd appreciate comments on them so we can help
to address any outstanding issue!
There are still some pending issues (RCU troubles). I will post V2 today.
+void mmu_notifier_release(struct mm_struct *mm)
...
+ hlist_for_each_entry_safe_rcu(mn, n, t,
+ mm-mmu_notifier.head, hlist) {
+ if (mn-ops-release)
+ mn-ops-release(mn, mm);
+
Hi;
This patch (rediffed againg kvm-60) from Tavis Ormandy [EMAIL PROTECTED]
fixes an infinite
loop in the emulated SB16 device (See http://taviso.decsystem.org/virtsec.pdf
for more details.)
I'm not sure why qemu upstream not merged these but Xen already did.
[1]
Hi;
This patch (rediffed againg kvm-60) from Tavis Ormandy [EMAIL PROTECTED]
fixes CVE-2007-1320 [1]
Again, I'm not sure why qemu upstream not merged these but Xen already did [2].
[1] http://cve.mitre.org/cgi-bin/cvename.cgi?name=CVE-2007-1320
[2]
# HG changeset patch
# User Jerone Young [EMAIL PROTECTED]
# Date 1201568508 21600
# Node ID a568d031723942e1baf77077031d2b77795cbd8a
# Parent 5ce532cf9a1f711d1fecb42814d301abd37aa378
Cleanup extern declerations for now removed vcpu_env in Qemu
This patch removes extern decleration for vcpu_env
Bugs item #1881532, was opened at 2008-01-28 18:16
Message generated for change (Tracker Item Submitted) made by Item Submitter
You can respond by visiting:
https://sourceforge.net/tracker/?func=detailatid=893831aid=1881532group_id=180599
Please note that this message will contain a full copy of
On Mon, 28 Jan 2008, Robin Holt wrote:
USE_AFTER_FREE!!! I made this same comment as well as other relavent
comments last week.
Must have slipped somehow. Patch needs to be applied after the rcu fix.
Please repeat the other relevant comments if they are still relevant I
thought I had
Dor Laor wrote:
On Mon, 2008-01-28 at 09:32 -0600, Anthony Liguori wrote:
Hi Dor,
How are you measuring performance? The numbers I've gotten with netperf
before and after your patch are:
tx - 647.27mbit
rx - 89.22
tx - 27.82
rx - 79.93
I've been testing with iperf
# HG changeset patch
# User Jerone Young [EMAIL PROTECTED]
# Date 1201585078 21600
# Node ID 34d726145a8ee9d58a4573473e861db2ad695c37
# Parent a568d031723942e1baf77077031d2b77795cbd8a
Remove unoptimal code from qemu dcr handles for powerpc
A patch I submitted yesterday to use the call
56 matches
Mail list logo
