From: Sheng Yang sh...@linux.intel.com
In capability probing ioctl.
Signed-off-by: Sheng Yang sh...@linux.intel.com
Signed-off-by: Avi Kivity a...@redhat.com
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 4d2be16..08ec7d3 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
From: Andrea Arcangeli aarca...@redhat.com
When kvm emulates an invlpg instruction, it can drop a shadow pte, but
leaves the guest tlbs intact. This can cause memory corruption when
swapping out.
Without this the other cpu can still write to a freed host physical page.
tlb smp flush must happen
From: Sheng Yang sh...@linux.intel.com
After discussion with Marcelo, we decided to rework device assignment framework
together. The old problems are kernel logic is unnecessary complex. So Marcelo
suggest to split it into a more elegant way:
1. Split host IRQ assign and guest IRQ assign. And
Notice here is a simple solution, can be replaced later.
(update: fix kvm_check_extension() only return 0 or 1, which is inconsisent
with kvm_get_gsi_count())
Signed-off-by: Sheng Yang sh...@linux.intel.com
---
libkvm/kvm-common.h |1 +
libkvm/libkvm.c | 17 -
In capability probing ioctl.
Signed-off-by: Sheng Yang sh...@linux.intel.com
---
virt/kvm/kvm_main.c |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index d480a3b..2648096 100644
--- a/virt/kvm/kvm_main.c
+++
Sheng Yang wrote:
In capability probing ioctl.
Applied, thanks. Will queue for 2.6.30 as well.
--
error compiling committee.c: too many arguments to function
--
To unsubscribe from this list: send the line unsubscribe kvm in
the body of a message to majord...@vger.kernel.org
More
Sheng Yang wrote:
On Thursday 12 March 2009 21:36:44 Sheng Yang wrote:
From: Marcelo Tosatti mtosa...@redhat.com
Signed-off-by: Sheng Yang sh...@linux.intel.com
Oops.. Should be Marcelo's signed-off here...
Should be both. If you are forwarding someone else's patch, even with
Han, Weidong wrote:
kvm_deassign_pci_device is defined when KVM_CAP_DEVICE_DEASSIGNMENT is defined.
Following patch should fix the breakage.
Signoff please.
--
error compiling committee.c: too many arguments to function
--
To unsubscribe from this list: send the line unsubscribe kvm in
Andrea Arcangeli wrote:
It's up to you, I guess it'll make life easier with the compat code ;).
I applied it, looking now at reducing the cost.
--
error compiling committee.c: too many arguments to function
--
To unsubscribe from this list: send the line unsubscribe kvm in
the body of a
KVM flushes tlbs on remote cpus for two purposes: to protect guest pages
that it needs to collect information about, and to prevent stale tlb entries
from pointing to pages that no longer belong to the guest.
We can defer the latter flushes to the point when we actually free the pages,
which is
Jan Kiszka wrote:
Hi,
maybe someone has an immediate idea or can reproduce, I'm currently
lacking time to debug this: With latest kvm-userspace I'm seeing
sporadic qemu crashes when hard-resetting a guest that currently does
some I/O (qcow2 disk in snapshot mode). Below is the backtrace, s is
For KVM remote TLB flushes we need to check the KVM_REQ_TLB_FLUSH
request flag when the irqs are already disabled. Otherwise there is a
small window of time for a race condition where we may enter a guest
without doing a requested TLB flush.
Signed-off-by: Joerg Roedel joerg.roe...@amd.com
---
Joerg Roedel wrote:
For KVM remote TLB flushes we need to check the KVM_REQ_TLB_FLUSH
request flag when the irqs are already disabled. Otherwise there is a
small window of time for a race condition where we may enter a guest
without doing a requested TLB flush.
@@ -3108,8 +3108,6 @@ static int
Avi Kivity wrote:
Jan Kiszka wrote:
Hi,
maybe someone has an immediate idea or can reproduce, I'm currently
lacking time to debug this: With latest kvm-userspace I'm seeing
sporadic qemu crashes when hard-resetting a guest that currently does
some I/O (qcow2 disk in snapshot mode). Below is
I originally asked this on the Fedora-general list, but people note that it
seems to happen in ubuntu a well.
When running stuff in KVM, I find that my mouse regularly (but at no fixed
interval) loses focus, but the keyboard remains bound to the application, and
the window header still advises
On Mon, Mar 16, 2009 at 03:12:52PM +0200, Avi Kivity wrote:
Joerg Roedel wrote:
For KVM remote TLB flushes we need to check the KVM_REQ_TLB_FLUSH
request flag when the irqs are already disabled. Otherwise there is a
small window of time for a race condition where we may enter a guest
without
From: Mike mbu...@redhat.com
---
client/tests/kvm_runtest_2/kvm_tests.cfg.sample| 63 +
.../tests/kvm_runtest_2/steps/RHEL-3.9-i386.steps | 241 +++
.../kvm_runtest_2/steps/RHEL-3.9-x86_64.steps | 241 +++
Bugs item #2687641, was opened at 2009-03-15 14:36
Message generated for change (Comment added) made by avik
You can respond by visiting:
https://sourceforge.net/tracker/?func=detailatid=893831aid=2687641group_id=180599
Please note that this message will contain a full copy of the comment
Bugs item #2688789, was opened at 2009-03-16 17:11
Message generated for change (Tracker Item Submitted) made by Item Submitter
You can respond by visiting:
https://sourceforge.net/tracker/?func=detailatid=893831aid=2688789group_id=180599
Please note that this message will contain a full copy of
Hi,
what must I do assigning a graphic card pci-device to (i.e.
winXP-client) guest? In my case, the vga-adapter uses irq 16, so that I
think, it could be possible to mask out this irq, so that it is free for
using in the guest. Im deactivating _lots_ of linux-options so linux ist
not using the
On Mon, 16 Mar 2009 17:34:25 +0100, Oliver Rath wrote:
Hi,
what must I do assigning a graphic card pci-device to (i.e.
winXP-client) guest? In my case, the vga-adapter uses irq 16, so that I
think, it could be possible to mask out this irq, so that it is free for
using in the guest. Im
PCI passthrough for video cards isn't expected to work, even if the host
isn't using them at all. See prior on-list discussion for details; my
understanding is that the guest won't run the card's video BIOS at
boot-time.
--
To unsubscribe from this list: send the line unsubscribe kvm in
the
Signed-off-by: Mike Burns mbu...@redhat.com
---
client/tests/kvm_runtest_2/kvm_tests.cfg.sample| 63 +
.../tests/kvm_runtest_2/steps/RHEL-3.9-i386.steps | 241 +++
.../kvm_runtest_2/steps/RHEL-3.9-x86_64.steps | 241 +++
On Mon, Mar 16, 2009 at 11:10:47AM +0200, Avi Kivity wrote:
Sheng Yang wrote:
Patch 1 and 2 are new ones, all the others had been sent before.
This (huge) patchset, contained:
Patch 1..2 are new interface after reworked device assignment kernel part.
Patch 3..6 are generic capability
On Sun, Mar 15, 2009 at 10:11:29PM +0200, Izik Eidus wrote:
Marcelo Tosatti wrote:
On Thu, Mar 12, 2009 at 06:18:43PM +0100, Andrea Arcangeli wrote:
From: Andrea Arcangeli aarca...@redhat.com
While looking at invlpg out of sync code with Izik I think I noticed a
missing smp tlb flush
On Mon, Mar 16, 2009 at 03:12:52PM +0200, Avi Kivity wrote:
Joerg Roedel wrote:
For KVM remote TLB flushes we need to check the KVM_REQ_TLB_FLUSH
request flag when the irqs are already disabled. Otherwise there is a
small window of time for a race condition where we may enter a guest
without
Marcelo Tosatti wrote:
BTW, I've wondered if the local_irq_enable in svm_vcpu_run is safe:
clgi();
local_irq_enable();
There is no way that an interrupt can be handled there without an exit,
right?
clgi trumps sti, so all interrupts will be deferred until the guest is
[ Get, save, synchronize, put... hell. Reminds me: what's the schedule
for finally dropping libkvm? ]
Fix this fallout from recent qemu-trunk merge, it broke guest debugging.
Signed-off-by: Jan Kiszka jan.kis...@siemens.com
---
qemu/qemu-kvm.h |4 ++--
1 files changed, 2 insertions(+), 2
On Mon, Mar 16, 2009 at 03:30:00PM -0300, Marcelo Tosatti wrote:
On Mon, Mar 16, 2009 at 03:12:52PM +0200, Avi Kivity wrote:
Joerg Roedel wrote:
For KVM remote TLB flushes we need to check the KVM_REQ_TLB_FLUSH
request flag when the irqs are already disabled. Otherwise there is a
small
Align qemu-kvm-x86 with upstream kvm support /wrt software breakpoint
support in ROM code.
Signed-off-by: Jan Kiszka jan.kis...@siemens.com
---
qemu/qemu-kvm-x86.c | 29 -
1 files changed, 24 insertions(+), 5 deletions(-)
diff --git a/qemu/qemu-kvm-x86.c
Hello,
I'm currently testing Windows XP in a Linux host. I works well except the
USB transfer rate. Printers and USB Pen Drives can be connected and get
detected (I added host:auto...). My System only supports USB 1.1, the
transfer rate in plain linux is okay. In the Guest Windows, the
Jan Kiszka wrote:
Align qemu-kvm-x86 with upstream kvm support /wrt software breakpoint
support in ROM code.
Hmm, this might not be needed as kvm-userspace is not protecting its
ROM. That's why I didn't pushed this so far. However, aligning isn't
bad. But dropping this duplication would be
On Wed, Mar 11, 2009 at 12:07:55PM +0200, Gleb Natapov wrote:
free_mmu_pages() should only undo what alloc_mmu_pages() does.
Signed-off-by: Gleb Natapov g...@redhat.com
diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c
index 2a36f7f..b625ed4 100644
--- a/arch/x86/kvm/mmu.c
+++
Bugs item #2681442, was opened at 2009-03-11 06:08
Message generated for change (Comment added) made by bliitz
You can respond by visiting:
https://sourceforge.net/tracker/?func=detailatid=893831aid=2681442group_id=180599
Please note that this message will contain a full copy of the comment
On Mon, Mar 16, 2009 at 05:15:33PM -0300, Marcelo Tosatti wrote:
On Wed, Mar 11, 2009 at 12:07:55PM +0200, Gleb Natapov wrote:
free_mmu_pages() should only undo what alloc_mmu_pages() does.
Signed-off-by: Gleb Natapov g...@redhat.com
diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c
On Mon, Mar 16, 2009 at 10:34:01PM +0200, Gleb Natapov wrote:
Doesnt the vm shutdown path rely on the while loop you removed to free
all shadow pages before freeing the mmu kmem caches, if mmu notifiers
is disabled?
Shouldn't mmu_free_roots() on all vcpus clear all mmu pages?
No. It
On Mon, Mar 16, 2009 at 06:01:52PM -0300, Marcelo Tosatti wrote:
On Mon, Mar 16, 2009 at 10:34:01PM +0200, Gleb Natapov wrote:
Doesnt the vm shutdown path rely on the while loop you removed to free
all shadow pages before freeing the mmu kmem caches, if mmu notifiers
is disabled?
Avi Kivity avi at redhat.com writes:
While qemu uses active-high interrupts throughout, kvm is meant to be
useful with other userspace (or for emulating other chipsets) and thus
needs correct polarity emulation.
If the guest programs polarity to active-low, this indicates either a
Avi Kivity avi at redhat.com writes:
Best would be to check for a Wind River kernel as this is likely to
break on mainline Linux.
OK, I see, that's more complicated. We are currently thinking about to upgrade
our systems to an up-to-date kernel version. Then this patch is no longer
The new wiki now lives in it own domain -
http://www.linux-kvm.org/page/Main_Page
Now that we have spam-safe wiki (hopefully) it also have nicer look too.
Note that it is not related to www.linux-kvm.com.
The old content is still at http://il.qumranet.com/kvmwiki, comments
about missing
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Hey all,
I started testing live migration and seem to have his a snag... It
appears to only occasionally work. Here is my setup:
2 Identical Machines:
Intel(R) Xeon(R) CPU 3050 @ 2.13GHz
8G Ram
Debian Lenny 64bit
Vanilla 2.6.28.7 kernel (loading
Avi Kivity wrote:
Han, Weidong wrote:
kvm_deassign_pci_device is defined when KVM_CAP_DEVICE_DEASSIGNMENT
is defined. Following patch should fix the breakage.
Signoff please.
Sorry for missing it. Signed-off-by: Weidong Han weidong@intel.com
Regards,
Weidong--
To unsubscribe from
* Mike Burns mbu...@redhat.com [2009-03-16 12:52]:
Very nice! Thanks. I haven't validated RHEL3 bits yet, but tested out
RHEL4.7 and 5.3 32 and 64 install steps; working here.
Signed-off-by: Mike Burns mbu...@redhat.com
Acked-by: Ryan Harper ry...@us.ibm.com
---
Hi Matthew,
Can you please take a look at this new version? I'd like to make sure
that all concerns are addressed and I didn't miss something :-)
Thanks,
Yu
On Wed, Mar 11, 2009 at 03:25:41PM +0800, Yu Zhao wrote:
Greetings,
Following patches are intended to support SR-IOV capability in the
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Update on this...
The 32bit guests migrate fine if they are under low load. It looks like
the 64bit guests consistently become unresponsive during the migration.
Stu
Stuart Sheldon wrote:
Hey all,
I started testing live migration and seem to
Extboot submits requests with whatever buffer alignment the guest gave
to the BIOS. This breaks with O_DIRECT disks, as they require 512 byte
alignment.
Most guest bootloaders sector align their requests out of paranoia, but
the OpenBSD bootloader does not.
This patch always copies. Since
On Tuesday 17 March 2009 02:12:11 Marcelo Tosatti wrote:
On Mon, Mar 16, 2009 at 11:10:47AM +0200, Avi Kivity wrote:
Sheng Yang wrote:
Patch 1 and 2 are new ones, all the others had been sent before.
This (huge) patchset, contained:
Patch 1..2 are new interface after reworked device
From: Marcelo Tosatti mtosa...@redhat.com
Signed-off-by: Marcelo Tosatti mtosa...@redhat.com
Signed-off-by: Sheng Yang sh...@linux.intel.com
---
libkvm/libkvm.c | 38 +-
libkvm/libkvm.h | 21 +
2 files changed, 54 insertions(+), 5
From: Marcelo Tosatti mtosa...@redhat.com
Also add irq_requested_type for assigned device to track current status.
Signed-off-by: Marcelo Tosatti mtosa...@redhat.com
Signed-off-by: Sheng Yang sh...@linux.intel.com
---
qemu/hw/device-assignment.c | 12
qemu/hw/device-assignment.h
Which is used later for capability detection.
Signed-off-by: Sheng Yang sh...@linux.intel.com
---
qemu/Makefile.target |1 +
qemu/configure | 20
qemu/hw/pci.h|8
3 files changed, 29 insertions(+), 0 deletions(-)
diff --git
Update from v5:
Add a new field irq_requested_type in assigned device to clarify the
ambigious semantic of deassign irq(at least it's more clear to me, even though
we didn't check this in the kernel space).
--
libkvm/kvm-common.h |1 +
libkvm/libkvm.c | 178
Notice here is a simple solution, can be replaced later.
Signed-off-by: Sheng Yang sh...@linux.intel.com
---
libkvm/kvm-common.h |1 +
libkvm/libkvm.c | 17 -
libkvm/libkvm.h |8
3 files changed, 25 insertions(+), 1 deletions(-)
diff --git
Signed-off-by: Sheng Yang sh...@linux.intel.com
---
qemu/hw/device-assignment.c | 20
1 files changed, 12 insertions(+), 8 deletions(-)
diff --git a/qemu/hw/device-assignment.c b/qemu/hw/device-assignment.c
index 4ec655c..063e313 100644
--- a/qemu/hw/device-assignment.c
Try to figure out device capability in update_dev_cap(). Now we are only care
about MSI capability.
The function pci_find_cap_offset original function wrote by Allen for Xen.
Notice the function need root privilege to work. This depends on libpci to work.
Signed-off-by: Allen Kay
SRIOV's virtual function didn't show correct Vendor ID/Device ID in config, so
we have to fill them manually according to device/vendor file in sysfs.
Signed-off-by: Sheng Yang sh...@linux.intel.com
---
qemu/hw/device-assignment.c | 31 ++-
1 files changed, 30
This framework can be easily extended to support device capability, like
MSI/MSI-x.
Signed-off-by: Sheng Yang sh...@linux.intel.com
---
qemu/hw/pci.c | 77 +++-
qemu/hw/pci.h | 29 +
2 files changed, 104 insertions(+), 2
Signed-off-by: Sheng Yang sh...@linux.intel.com
---
libkvm/libkvm.c | 98 ---
libkvm/libkvm.h | 22
2 files changed, 101 insertions(+), 19 deletions(-)
diff --git a/libkvm/libkvm.c b/libkvm/libkvm.c
index f13cc03..36036b8
MMIO enable byte would be checked when enabling virtual function, but in fact,
the whole virtual function's command register is hard-wired to zero... So when
guest read from command register it would only get 0, specially for MMIO enable
bit. And this result in, if guest driver want to do a RMW to
(Alex: correct libpci usage)
Signed-off-by: Sheng Yang sh...@linux.intel.com
Signed-off-by: Alexander Duyck alexander.h.du...@intel.com
---
qemu/hw/device-assignment.c | 144 --
qemu/hw/device-assignment.h |9 +++
2 files changed, 146 insertions(+), 7
Signed-off-by: Sheng Yang sh...@linux.intel.com
---
qemu/hw/device-assignment.c |7 ++-
1 files changed, 6 insertions(+), 1 deletions(-)
diff --git a/qemu/hw/device-assignment.c b/qemu/hw/device-assignment.c
index 9a3104b..d404286 100644
--- a/qemu/hw/device-assignment.c
+++
The most important part here, is we emulate a page of MMIO region using a
page of memory. That's because MSI-X table was put in the region and we have to
intercept it.
(Thanks Ben to find one bug in assigned_dev_update_msix_mmio)
Signed-off-by: Sheng Yang sh...@linux.intel.com
Signed-off-by:
Signed-off-by: Sheng Yang sh...@linux.intel.com
---
libkvm/libkvm.c | 25 +
libkvm/libkvm.h |7 +++
2 files changed, 32 insertions(+), 0 deletions(-)
diff --git a/libkvm/libkvm.c b/libkvm/libkvm.c
index 542b9da..4cf779c 100644
--- a/libkvm/libkvm.c
+++
Shouldn't update assigned irq if assigned device's interrupt pin is 0, which
means uninitialized or don't support INTx.
Signed-off-by: Sheng Yang sh...@linux.intel.com
---
qemu/hw/device-assignment.c |4
1 files changed, 4 insertions(+), 0 deletions(-)
diff --git
Signed-off-by: Sheng Yang sh...@linux.intel.com
---
qemu/hw/pci.h |1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/qemu/hw/pci.h b/qemu/hw/pci.h
index 4f494e7..67fd1b9 100644
--- a/qemu/hw/pci.h
+++ b/qemu/hw/pci.h
@@ -146,6 +146,7 @@ typedef struct PCIIORegion {
#define
On Mon, 2009-03-16 at 12:35 +0800, Liu Yu wrote:
Hollis,
Finally I found the reason for unstable kvm in SMP.
Congratulations. :)
do you think the modification is OK?
They look fine to me, as far as I understand the e500 MMU. ;) Feel free
to send to Avi.
--
Hollis Blanchard
IBM Linux
On Mon, 2009-03-16 at 18:17 +0800, Liu Yu-B13201 wrote:
44x code was backported to qemu.git and e500 code was directly submmited to
qemu.git.
There is no support in kvm-userspace/qemu so far, and it's broken now.
I took a quick look, and found that the qemu.git uses kvm.h,
while
Hi Liu,
I could trace the illegal instruction during qemu startup down to this function
in cpu-all.h cpu_get_real_ticks(void) which is using the time base registers -
the crash is during vm_start - cpu_enable_ticks which calls cpu_get_real_ticks.
I believe e500 is using SPR's for this purpose
67 matches
Mail list logo