From: Gleb Natapov g...@redhat.com
Signed-off-by: Gleb Natapov g...@redhat.com
Signed-off-by: Avi Kivity a...@redhat.com
diff --git a/qemu/qemu-kvm.c b/qemu/qemu-kvm.c
index 01c265b..8b9e7c5 100644
--- a/qemu/qemu-kvm.c
+++ b/qemu/qemu-kvm.c
@@ -1357,3 +1357,9 @@ int qemu_kvm_has_sync_mmu(void)
From: Gleb Natapov g...@redhat.com
Reset stack pointer to zero on S3 resume before jumping to resume vector
to prevent tpr patching on boot CPU during resume.
Signed-off-by: Gleb Natapov g...@redhat.com
Signed-off-by: Avi Kivity a...@redhat.com
diff --git a/bios/rombios.c b/bios/rombios.c
index
From: Gleb Natapov g...@redhat.com
Signed-off-by: Gleb Natapov g...@redhat.com
Signed-off-by: Avi Kivity a...@redhat.com
diff --git a/bios/rombios32.c b/bios/rombios32.c
index fd6ff71..a0c9572 100755
--- a/bios/rombios32.c
+++ b/bios/rombios32.c
@@ -38,7 +38,7 @@ typedef unsigned long long
This is with kvm-81
I'm getting a kernel panic when I modprobe kvm-amd
It used to work until I had to use the CMOS jumper to boot.
The motherboard is an Asus M3N78 PRO
The CPU is an AMD Phenom 9950 Black Edition
Here are the steps I've already tried:
- Checking that virtualization is enabled
Marcelo Tosatti wrote:
Certain clocks (such as TSC) in older 2.6 guests overaccount for lost
ticks, causing severe time drift. Interrupt reinjection magnifies the
problem.
Provide an option to disable it.
Signed-off-by: Marcelo Tosatti mtosa...@redhat.com
Index: kvm/arch/x86/kvm/i8254.c
Marcelo Tosatti wrote:
Signed-off-by: Marcelo Tosatti mtosa...@redhat.com
Index: kvm-userspace.pit/libkvm/libkvm-x86.c
===
--- kvm-userspace.pit.orig/libkvm/libkvm-x86.c
+++ kvm-userspace.pit/libkvm/libkvm-x86.c
@@ -75,6 +75,20 @@
Marcelo Tosatti wrote:
{ no-kvm-irqchip, 0, QEMU_OPTION_no_kvm_irqchip },
{ no-kvm-pit, 0, QEMU_OPTION_no_kvm_pit },
+{ kvm-pit-no-reinject, 0, QEMU_OPTION_kvm_pit_no_reinject },
no-kvm-pit-reinjection?
no-kvm-pit-compensation?
Hard to find a user friendly name for this.
--
On Tuesday 30 December 2008 18:19:29 Avi Kivity wrote:
Sheng Yang wrote:
For MSI disable feature later.
Notice I changed ABI here, but due to no userspace patch, I think it's
OK.
It's not okay, since eventually we will have userspace and it will have
to work with older kernels as well.
Sheng Yang wrote:
Which is more convenient...
Signed-off-by: Sheng Yang sh...@linux.intel.com
---
virt/kvm/kvm_main.c | 10 ++
1 files changed, 2 insertions(+), 8 deletions(-)
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index ffd261d..cd84b3e 100644
---
Sheng Yang wrote:
On Tuesday 30 December 2008 18:19:29 Avi Kivity wrote:
Sheng Yang wrote:
For MSI disable feature later.
Notice I changed ABI here, but due to no userspace patch, I think it's
OK.
It's not okay, since eventually we will have userspace and it will have
to work
On Tuesday 30 December 2008 18:31:16 Avi Kivity wrote:
Sheng Yang wrote:
On Tuesday 30 December 2008 18:19:29 Avi Kivity wrote:
Sheng Yang wrote:
For MSI disable feature later.
Notice I changed ABI here, but due to no userspace patch, I think it's
OK.
It's not okay, since
The patch series provides remaining bits for S3 support in KVM. They are
not (yet?) relevant for qemu upstream.
---
Gleb Natapov (3):
KVM does not support SMM. Disable it.
Don't return to guest after CPU issued S3 command.
Reset stack pointer to zero on S3 resume.
Reset stack pointer to zero on S3 resume before jumping to resume vector
to prevent tpr patching on boot CPU during resume.
Signed-off-by: Gleb Natapov g...@redhat.com
---
bios/rombios.c |1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/bios/rombios.c b/bios/rombios.c
Signed-off-by: Gleb Natapov g...@redhat.com
---
qemu/hw/acpi.c |2 +-
qemu/qemu-kvm.c |2 +-
qemu/qemu-kvm.h |1 +
3 files changed, 3 insertions(+), 2 deletions(-)
diff --git a/qemu/hw/acpi.c b/qemu/hw/acpi.c
index 0ff8851..219d8ac 100644
--- a/qemu/hw/acpi.c
+++ b/qemu/hw/acpi.c
Signed-off-by: Gleb Natapov g...@redhat.com
---
bios/rombios32.c |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/bios/rombios32.c b/bios/rombios32.c
index fd6ff71..a0c9572 100755
--- a/bios/rombios32.c
+++ b/bios/rombios32.c
@@ -38,7 +38,7 @@ typedef unsigned long long
Sheng Yang wrote:
Avi's purpose, to use single kvm_set_irq() to deal with all interrupt, including
MSI. So here is it.
struct gsi_msg is a mapping from a special gsi(with KVM_GSI_MSG_MASK) to
MSI/MSI-X message address/data.
Now we support up to 256 gsi_msg mapping, and gsi_msg is allocated by
Sheng Yang wrote:
MASK_MSIX, and UNMASK, every two action are in pairs but we have to use
twice bits to store them. So I'd like to use MSI_ACTION approach...
Well, it you have flags without ENABLE_MSI, doesn't it imply DISABLE_MSI?
The structure contains the state we want to reach, not
On Tuesday 30 December 2008 18:41:54 Avi Kivity wrote:
Sheng Yang wrote:
MASK_MSIX, and UNMASK, every two action are in pairs but we have to use
twice bits to store them. So I'd like to use MSI_ACTION approach...
Well, it you have flags without ENABLE_MSI, doesn't it imply
DISABLE_MSI?
Sheng Yang wrote:
Using kvm_set_irq to handle all interrupt injection.
/* This should be called with the kvm-lock mutex held */
-void kvm_set_irq(struct kvm *kvm, int irq_source_id, int irq, int level)
+void kvm_set_irq(struct kvm *kvm, int irq_source_id, u32 gsi, int level)
{
-
Glauber Costa writes ([PATCH] hook cpu running at a higher level.):
This patch removes the kvm_enabled() check from cpu-exec.c.
This file is highly tcg-specific, and we'll probably want it
out when tcg is not compiled in (coming soon, in a theathe near you)
That would be interesting,
Ian Jackson wrote:
The way we have approached these problems in the Xen tree is to supply
an alternative implementation of (say) main_loop and arrange for the
standard one not to be compiled. Is it the intent to make kvm a
run-time selectable option ? It seems to me that that given that we
Sheng Yang wrote:
For MSI disable feature later.
Notice I changed ABI here, but due to no userspace patch, I think it's OK.
It's not okay, since eventually we will have userspace and it will have
to work with older kernels as well.
No released kernel has KVM_DEV_IRQ_ASSIGN_ENABLE_MSI,
On Tuesday 30 December 2008 18:48:42 Avi Kivity wrote:
Sheng Yang wrote:
Using kvm_set_irq to handle all interrupt injection.
/* This should be called with the kvm-lock mutex held */
-void kvm_set_irq(struct kvm *kvm, int irq_source_id, int irq, int level)
+void kvm_set_irq(struct
Sheng Yang wrote:
+ mutex_lock(kvm-gsi_msg_lock);
The lock is already taken here?
Um? For gsi_msg_lock?
Sorry, my mistake. Will have to get used to all those locks.
Is there a way to avoid the lock? We're starting to complicate things...
This looks very messy.
Sheng Yang wrote:
This will call cancel_work_sync(), which may wait upon kvm-lock, which
I think we hold here - deadlock.
I think that the current code has even bigger problems (races), since a
scheduled work can arrive after the interrupt has been freed and
reallocated.
For the race
Gleb Natapov wrote:
Signed-off-by: Gleb Natapov g...@redhat.com
---
qemu/hw/acpi.c |2 +-
qemu/qemu-kvm.c |2 +-
qemu/qemu-kvm.h |1 +
3 files changed, 3 insertions(+), 2 deletions(-)
diff --git a/qemu/hw/acpi.c b/qemu/hw/acpi.c
index 0ff8851..219d8ac 100644
--- a/qemu/hw/acpi.c
On Tuesday 30 December 2008 19:16:27 Avi Kivity wrote:
Sheng Yang wrote:
This will call cancel_work_sync(), which may wait upon kvm-lock, which
I think we hold here - deadlock.
I think that the current code has even bigger problems (races), since a
scheduled work can arrive after the
On Tuesday 30 December 2008 18:28:21 Avi Kivity wrote:
Sheng Yang wrote:
Which is more convenient...
Signed-off-by: Sheng Yang sh...@linux.intel.com
---
virt/kvm/kvm_main.c | 10 ++
1 files changed, 2 insertions(+), 8 deletions(-)
diff --git a/virt/kvm/kvm_main.c
Gleb Natapov wrote:
The patch series provides remaining bits for S3 support in KVM. They are
not (yet?) relevant for qemu upstream.
---
Gleb Natapov (3):
KVM does not support SMM. Disable it.
Don't return to guest after CPU issued S3 command.
Reset stack pointer to zero on S3
On Tuesday 30 December 2008 19:07:52 Avi Kivity wrote:
Sheng Yang wrote:
+ mutex_lock(kvm-gsi_msg_lock);
The lock is already taken here?
Um? For gsi_msg_lock?
Sorry, my mistake. Will have to get used to all those locks.
Is there a way to avoid the lock? We're starting to
Sheng Yang wrote:
So still, how about deal with this at the beginning of kvm_destory_vm? Discard
currently reference count. In the function, disable all known interrupt(and
result some nested disabled but we would free them later), then
cancel_work_sync(). At this time, I think the state is
On Tue, Dec 30, 2008 at 8:24 AM, Ian Jackson The Xen qemu process
runs only in one thread which is fine because it
doesn't need to be involved with actual processor execution. In
theory parallel execution (in different threads and thus on different
physical cpus) of IO emulations requested by
--
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/majordomo-info.html
kvm_arch_sync_events is introduced to quiet down all other events may happen
contemporary with VM destroy process, like IRQ handler and work struct for
assigned device.
For kvm_arch_sync_events is called at the very beginning of kvm_destroy_vm(), so
the state of KVM here is legal and can provide
Signed-off-by: Sheng Yang sh...@linux.intel.com
---
include/linux/kvm_host.h |1 -
1 files changed, 0 insertions(+), 1 deletions(-)
diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
index d63e9a4..76cc371 100644
--- a/include/linux/kvm_host.h
+++ b/include/linux/kvm_host.h
@@
Sheng Yang wrote:
This looks good, but I'll let Marcelo (or any other interested party)
review this as well, since this is tricky stuff.
--
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
On Tue, Dec 30, 2008 at 01:19:22PM +0200, Avi Kivity wrote:
+int kvm_shutdown(void *opaque, void *data);
That's a bad API. What's the opaque? what's the data?
How about this one:
Don't return to guest CPU after qemu_system_reset_request().
Signed-off-by: Gleb Natapov g...@redhat.com
Hi everybody,
I am trying to implement a new virtio device. I have implemented the
new device in Qemu and I also have the correspondent virtio driver in the
guest. But when I get data in the device and I try to put that into the
virtqueue using virtqueue_pop, the qemu instance crash
Actually, I am not trying to put data into the queue using pop. I get an
element from the queue using virtqueue_pop, fill the element and then push
the element into the queue using virtqueue_push. When I try to do the pop is
when it crash.
Pablo
-Original Message-
From:
Gleb Natapov wrote:
On Tue, Dec 30, 2008 at 01:19:22PM +0200, Avi Kivity wrote:
+int kvm_shutdown(void *opaque, void *data);
That's a bad API. What's the opaque? what's the data?
How about this one:
Don't return to guest CPU after qemu_system_reset_request().
Applied,
Avi Kivity writes (Re: [PATCH] hook cpu running at a higher level.):
kvm is run-time selectable, both in upstream and in kvm-userspace. If
kvm is not detected (or the caller lacks sufficient privileges), we fall
back to tcg (of course we'd also like the option of not compiling tcg
where
Ian Jackson wrote:
The Xen qemu process runs only in one thread which is fine because it
doesn't need to be involved with actual processor execution. In
theory parallel execution (in different threads and thus on different
physical cpus) of IO emulations requested by different guest vcpus
might
Avi Kivity writes (Re: [PATCH] hook cpu running at a higher level.):
I thing it makes sense for you to have multiple vcpus -- that is
multiple CPUState objects. You can still have just one thread (which
would be the iothread in kvm's terminology).
Xen does have multiple vcpus (of course)
Hello fellow 2008-ers, soon to be 2009-ers.
It's time to say goodbye to the year that goes, which I do
by sending the last small series of the year.
This is another hook proposal for kvm, still in the ways to
make tcg optional.
I'm hooking vcpu creation in order to avoid tcg initialization
Allow kvm to override vcpu creation. We need to grab
a minimal amount of code that is shared, but the big
part is highly kvm/tcg specific.
Signed-off-by: Glauber Costa glom...@redhat.com
---
kvm-all.c| 18 ++
kvm.h|2 +-
target-i386/helper.c |
We don't need to pass the model string by parameter,
since it ough to be part of the CPUState anyway at
the call time. Also, don't declare it as static,
so kvm can call it in the future.
Signed-off-by: Glauber Costa glom...@redhat.com
---
target-i386/cpu.h|2 ++
target-i386/helper.c |
On Tue, Dec 30, 2008 at 02:53:36PM +1100, Nick Piggin wrote:
RSP 88011f4c7be8
---[ end trace 31811279a2e983e8 ]---
note: qemu-system-x86[4440] exited with preempt_count 2
(gdb) l *(__purge_vmap_area_lazy + 0x12c)
0x80289ca2 is in __purge_vmap_area_lazy
Marcelo Tosatti wrote:
On Tue, Dec 30, 2008 at 02:53:36PM +1100, Nick Piggin wrote:
RSP 88011f4c7be8
---[ end trace 31811279a2e983e8 ]---
note: qemu-system-x86[4440] exited with preempt_count 2
(gdb) l *(__purge_vmap_area_lazy + 0x12c)
0x80289ca2 is in __purge_vmap_area_lazy
Shan, Haitao wrote:
Well, it may make sense to provide the ROMs as virtual PCI BARs, and
have the bios do the work. This way, if some driver relies on
remapping the BAR (graphic cards?), it can still work.
I do not quite understand this. Can you elaborate?
Which part?
I propose not
On Tue, Dec 30, 2008 at 12:08:01PM +0200, Avi Kivity wrote:
Marcelo Tosatti wrote:
Certain clocks (such as TSC) in older 2.6 guests overaccount for lost
ticks, causing severe time drift. Interrupt reinjection magnifies the
problem.
Provide an option to disable it.
Signed-off-by: Marcelo
Avi Kivity wrote:
From: Gleb Natapov g...@redhat.com
Signed-off-by: Gleb Natapov g...@redhat.com
Signed-off-by: Avi Kivity a...@redhat.com
diff --git a/bios/rombios32.c b/bios/rombios32.c
index fd6ff71..a0c9572 100755
--- a/bios/rombios32.c
+++ b/bios/rombios32.c
@@ -38,7 +38,7 @@ typedef
Marcelo Tosatti wrote:
On Tue, Dec 30, 2008 at 12:08:01PM +0200, Avi Kivity wrote:
Marcelo Tosatti wrote:
Certain clocks (such as TSC) in older 2.6 guests overaccount for lost
ticks, causing severe time drift. Interrupt reinjection magnifies the
problem.
Provide an option to disable
(copying relevant people)
Tony Battersby wrote:
I have a SuperMicro C2SBX motherboard with BIOS revision 1.0b. With
vt-d enabled in the BIOS, Linux gets into an endless loop printing
DMAR:Unknown DMAR structure type when booting. Here is the DMAR ACPI
table:
DMAR @ 0x7fe86dec
: 44 4d
On Tue, Dec 30, 2008 at 10:14:09AM +0800, Sheng Yang wrote:
There is one remaining issue: kvm_assigned_dev_interrupt_work_handler
can re-enable the interrupt for KVM_ASSIGNED_DEV_GUEST_MSI case. Perhaps
you need a new flag to indicate shutdown (so the host IRQ won't be
reenabled).
Is it
I have a SuperMicro C2SBX motherboard with BIOS revision 1.0b. With
vt-d enabled in the BIOS, Linux gets into an endless loop printing
DMAR:Unknown DMAR structure type when booting. Here is the DMAR ACPI
table:
DMAR @ 0x7fe86dec
: 44 4d 41 52 98 00 00 00 01 6f 49 6e 74 65 6c 20
Some more information on this one:
I just pulled and installed the latest userspace, Now I get
a kernel panic: not syncing: IO-APIC + timer doesn't work!
Boot with apic=debug and send a report. Then try booting
with the 'noapic' option. (this is during the ubuntu CD boot
with the qemu line as:
Glauber Costa writes (Re: [Qemu-devel] Re: [PATCH] hook cpu running at a
higher level.):
On Tue, Dec 30, 2008 at 8:24 AM, Ian Jackson
The Xen qemu process runs only in one thread which is fine because it
doesn't need to be involved with actual processor execution. In
theory parallel
Addressing comments.
--
--
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/majordomo-info.html
count_load_time assignment is bogus: its supposed to contain what it
means, not the expiration time.
Signed-off-by: Marcelo Tosatti mtosa...@redhat.com
Index: kvm/arch/x86/kvm/i8254.c
===
--- kvm.orig/arch/x86/kvm/i8254.c
+++
Certain clocks (such as TSC) in older 2.6 guests overaccount for lost
ticks, causing severe time drift. Interrupt reinjection magnifies the
problem.
Provide an option to disable it.
Signed-off-by: Marcelo Tosatti mtosa...@redhat.com
Index: kvm/arch/x86/kvm/i8254.c
Userspace support for KVM_PIT_REINJECT_CONTROL.
--
--
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/majordomo-info.html
Signed-off-by: Marcelo Tosatti mtosa...@redhat.com
Index: kvm-userspace.pit/libkvm/libkvm-x86.c
===
--- kvm-userspace.pit.orig/libkvm/libkvm-x86.c
+++ kvm-userspace.pit/libkvm/libkvm-x86.c
@@ -75,6 +75,23 @@ int
Signed-off-by: Marcelo Tosatti mtosa...@redhat.com
Index: kvm-userspace.pit/qemu/qemu-kvm.c
===
--- kvm-userspace.pit.orig/qemu/qemu-kvm.c
+++ kvm-userspace.pit/qemu/qemu-kvm.c
@@ -11,6 +11,7 @@
int kvm_allowed = 1;
int kvm_irqchip
I upgraded from kvm-80 (which had some bugs with vista networking for
me) to 82 and at the same time upgraded to 2.6.28. My .config is the
same as it was before the kernel upgrade.
When I run qemu (see below) the window pops up for a split second and
the it segfaults.
Thanks in advance - Justin
Hello,
I'm running into a strange problem of 'make' not producing any *.ko
files recently (kvm-82). This seems rather odd, as 'make' still
completes successfully (echo $? returns 0), and there are *.o files
that are output as a result.
I've done a paste of my ./configure and make steps here:
On Mon, Dec 29, 2008 at 03:58:47PM +0100, Jan Kiszka wrote:
Well, in the current gdb design, current_gdbarch is consulted when
disassembling the code while target_gdbarch defines the register set
that is exchanged with the remote stub.
This is a transitional state. Really, there isn't
Setting CONFIG_KVM=m fixed the problem. KVM-82 works great.
-jk
On Tue, Dec 30, 2008 at 1:31 PM, Justin Keogh justin.ke...@gmail.com wrote:
I upgraded from kvm-80 (which had some bugs with vista networking for
me) to 82 and at the same time upgraded to 2.6.28. My .config is the
same as it
On Tue, 30 Dec 2008, Michael Park wrote:
Hello,
I'm running into a strange problem of 'make' not producing any *.ko
files recently (kvm-82). This seems rather odd, as 'make' still
completes successfully (echo $? returns 0), and there are *.o files
that are output as a result.
I've done
Hi there,
Just curious -- have you not been using the kvm modules included with
the linux kernel? (I'm assuming you mean kvm.ko and friends?)
I started out with Fedora's packaged kvm module, but at some point I
went the upstream route (thinking back now, I don't remember the
reason why I
Hi, Avi,
Option ROM already has its own BAR at 0x30h. I think the devices assignment
code now already handles this register.
Can I summary your proposals like the following: In guest BIOS, scan all the
pci devices (virtual) for existance of Option ROMs. Copy them to available BIOS
space in
On Wednesday 31 December 2008 02:32:50 Avi Kivity wrote:
Marcelo Tosatti wrote:
On Tue, Dec 30, 2008 at 02:53:36PM +1100, Nick Piggin wrote:
RSP 88011f4c7be8
---[ end trace 31811279a2e983e8 ]---
note: qemu-system-x86[4440] exited with preempt_count 2
(gdb) l
On Tue, Dec 30, 2008 at 10:34:13AM -0600, Anthony Liguori wrote:
Avi Kivity wrote:
From: Gleb Natapov g...@redhat.com
Signed-off-by: Gleb Natapov g...@redhat.com
Signed-off-by: Avi Kivity a...@redhat.com
diff --git a/bios/rombios32.c b/bios/rombios32.c
index fd6ff71..a0c9572 100755
---
72 matches
Mail list logo