Paul Brook wrote:
I immediately reproduced the problem locally. It turns out that
kvm reflects packets coming from one guest NIC on another guest
NIC, and since both are connected to the same bridge we're getting
endless packet storm. To a level when kvm process becomes 100%
busy and does not r
Anthony Liguori wrote:
> Avi Kivity wrote:
> >On 11/10/2009 04:36 PM, Anthony Liguori wrote:
> >>
> >>>A stats vq might solve this more cleanly?
> >>
> >>actual and target are both really just stats. Had we implemented
> >>those with a vq, I'd be inclined to agree with you but since they're
> >>
On Wed, Nov 11, 2009 at 01:34:12AM +, Paul Brook wrote:
> On Monday 26 October 2009, Michael S. Tsirkin wrote:
> > wmb must be at least a compiler barrier, even without SMP.
>
> Why?
Because virtio code might run on a separate thread from guest.
If compiler reorders writes, guest might see in
On 11/11/09 05:06, Paul Brook wrote:
On Friday 06 November 2009, Gerd Hoffmann wrote:
Hi,
http://repo.or.cz/w/qemu/kraxel.git/shortlog/refs/heads/scsi.v6
What is in there?
(3) New interface for HBA<=> SCSIDevice interaction:
* building on the new SCSIRequest struct.
* the
On Thu, Sep 24, 2009 at 10:14:58AM -0300, Luiz Capitulino wrote:
> On Thu, 24 Sep 2009 16:07:28 +0300
> Avi Kivity wrote:
>
> > On 09/24/2009 04:05 PM, Luiz Capitulino wrote:
> > >>>I think the easiest way to have this would be to add a
> > >>> monitor command to disable capabilities. Like a
On Thu, Sep 24, 2009 at 03:28:40PM +0200, Markus Armbruster wrote:
> Luiz Capitulino writes:
>
> > On Thu, 24 Sep 2009 00:37:53 +0200
> > Markus Armbruster wrote:
> >
> >> Luiz Capitulino writes:
> >>
> >> [...]
> >> > 2.2 Server Greeting
> >> > ---
> >> >
> >> > Right when con
On Wed, Nov 11, 2009 at 09:24:09AM +, Jamie Lokier wrote:
> Anthony Liguori wrote:
> > Avi Kivity wrote:
> > >On 11/10/2009 04:36 PM, Anthony Liguori wrote:
> > >>
> > >>>A stats vq might solve this more cleanly?
> > >>
> > >>actual and target are both really just stats. Had we implemented
>
Paul Brook wrote:
> On Thursday 05 November 2009, Juan Quintela wrote:
>> Daniel Jacobowitz wrote:
>> > On Thu, Nov 05, 2009 at 05:17:46PM +0100, Juan Quintela wrote:
>> >> How are you compiling?
>> >> It works for me compiling in-tree with make -j3 (only 2 cores)
>> >
>> > I can reliably reprodu
On Tuesday 10 November 2009, Aurelien Jarno wrote:
> I have tested it by removing all the block around tb_add_jump in
> cpu_exec.c. I have a speed loss of about 2.5x in the boot time of an
> x86_64 image.
I just tried it with qemu-system-x86_64, and with that I can observe a
noticable performance
On 11/07/09 00:09, Gerd Hoffmann wrote:
Hi,
http://repo.or.cz/w/qemu/kraxel.git/shortlog/refs/heads/scsi.v6
Pushed an update, same branch.
What needs to be done?
(1) Better patch descriptions.
(2) Submit patches to the list for review.
Still to do. Also:
- fold incremental fixes into
linux-user build on fedora 11 breaks because fallocate
is broken on that system if -D_GNU_SOURCE -D_FILE_OFFSET_BITS=64
are specified, which is what QEMU uses.
We do have a configure check to catch this and disable fallocate,
however, it turns out that default QEMU_CFLAGS/LDFLAGS were assigned in
Gerd Hoffmann wrote:
> On 11/07/09 00:09, Gerd Hoffmann wrote:
>> Hi,
>>
>> http://repo.or.cz/w/qemu/kraxel.git/shortlog/refs/heads/scsi.v6
>
> Pushed an update, same branch.
>
>> What needs to be done?
>>
>> (1) Better patch descriptions.
>> (2) Submit patches to the list for review.
>
> Stil
pci code had a TODO to move apb specific
pci bridge initialization to apb_pci.
Implement this and remove the TODO.
Signed-off-by: Michael S. Tsirkin
---
hw/apb_pci.c | 22 ++
hw/pci.c | 26 ++
hw/pci.h |1 +
3 files changed, 29 insertio
On Wed, Nov 11, 2009 at 01:49:06PM +0100, Juan Quintela wrote:
> "Michael S. Tsirkin" wrote:
>
> Hi
>
> > @@ -1585,7 +1605,7 @@ int main(void)
> > return 0;
> > }
> > EOF
> > -if $cc $ARCH_CFLAGS -o $TMPE $TMPC 2> /dev/null ; then
> > +if compile_prog "$ARCH_CFLAGS" "" ; then
>
> s/$ARCH
On Wednesday 11 November 2009, Michael S. Tsirkin wrote:
> On Wed, Nov 11, 2009 at 01:34:12AM +, Paul Brook wrote:
> > On Monday 26 October 2009, Michael S. Tsirkin wrote:
> > > wmb must be at least a compiler barrier, even without SMP.
> >
> > Why?
>
> Because virtio code might run on a separ
Anthony Liguori wrote:
> Gerd Hoffmann wrote:
>> On 11/09/09 17:06, Paul Brook wrote:
>>> On Tuesday 08 September 2009, Anthony Liguori wrote:
Gerd Hoffmann wrote:
> $subject says pretty much everything.
>
> extboot.[cS] are a straight copy from the kvm tree. The windup in
> v
On 11/11/09 12:52, Hannes Reinecke wrote:
I still don't quite agree with the new scsi callback, which has
just the SCSIRequest as an argument. So when supporting more command
queueing we have no idea to which (internal) request structure
the SCSIRequest belongs, and still have to do a lookup.
Whi
pci.h declares some functions/macros which aren't
defined in pci.h. Clean up, moving things
to appropriate headers, and update all users.
Signed-off-by: Michael S. Tsirkin
---
hw/apb_pci.c |1 +
hw/apb_pci.h |9 +
hw/lsi53c895a.c |2 ++
hw/mips_malta.c |2 ++
linux-user build on fedora 11 breaks because fallocate
is broken on that system if -D_GNU_SOURCE -D_FILE_OFFSET_BITS=64
are specified, which is what QEMU uses.
We do have a configure check to catch this and disable fallocate,
however, it turns out that default QEMU_CFLAGS/LDFLAGS were assigned in
On Wed, Nov 11, 2009 at 01:01:03PM +, Paul Brook wrote:
> On Wednesday 11 November 2009, Michael S. Tsirkin wrote:
> > On Wed, Nov 11, 2009 at 01:34:12AM +, Paul Brook wrote:
> > > On Monday 26 October 2009, Michael S. Tsirkin wrote:
> > > > wmb must be at least a compiler barrier, even wit
Hannes Reinecke wrote:
But why? Why do we have to emulate the entire HBA for the BIOS?
The HBA is emulated, too, and just uses the bdrv interface
internally anyway.
So IMHO it makes far more sense to skip the HBA emulation in
the BIOS completely and just use the bdrv interface directly
here.
On Wed, 2009-11-11 at 10:12 +, Daniel P. Berrange wrote:
> This all suggests that we should only update the stats from the guest
> when something on the host actually asks for them by issuing the QEMU
> monitor command. We don't want any kind of continuous polling of stats
> at any frequency, i
Gerd Hoffmann wrote:
> On 11/11/09 12:52, Hannes Reinecke wrote:
>> I still don't quite agree with the new scsi callback, which has
>> just the SCSIRequest as an argument. So when supporting more command
>> queueing we have no idea to which (internal) request structure
>> the SCSIRequest belongs, a
qemu-system-ppc is also broken.
A bisect gives me:
c169998802505c244b8bcad562633f29de7d74a4 is first bad commit
commit c169998802505c244b8bcad562633f29de7d74a4
Author: Glauber Costa
Date: Thu Nov 5 16:05:15 2009 -0200
v3: don't call reset functions on cpu initialization
There is
On Wed, Nov 11, 2009 at 01:01:03PM +, Paul Brook wrote:
> A "compiler memory barrier" provides absolutely no guarantees in a
> multithreaded environment.
Claims including words "absolutely no" have absolutely no
chance to be correct.
--
MST
> > > > > wmb must be at least a compiler barrier, even without SMP.
> > > >
> > > > Why?
> > >
> > > Because virtio code might run on a separate thread from guest.
> > > If compiler reorders writes, guest might see inconsistent data.
> >
> > If you've got threads running in parallel (which may be
On Wednesday 11 November 2009, Anthony Liguori wrote:
> Hannes Reinecke wrote:
> > But why? Why do we have to emulate the entire HBA for the BIOS?
> > The HBA is emulated, too, and just uses the bdrv interface
> > internally anyway.
> > So IMHO it makes far more sense to skip the HBA emulation in
>
On Wed, Nov 11, 2009 at 01:45:35PM +, Paul Brook wrote:
> If you don't need real barriers, then why does the kvm code have them?
We need real barriers but AFAIK kvm does not have them :(
IOW: virtio is currently broken with kvm, and my patch did
not fix this. The comment that I added says as m
> The current qemu code *does* cache the response. scsi-disk caps the
> buffer at 128k (which is big enough for any request I've seen in my
> testing). scsi-generic has no cap.
That cap is important.
For scsi-generic you probably don't have a choice because of the way the
kernel interface works
On Wednesday 11 November 2009, Michael S. Tsirkin wrote:
> On Wed, Nov 11, 2009 at 01:45:35PM +, Paul Brook wrote:
> > If you don't need real barriers, then why does the kvm code have them?
>
> We need real barriers but AFAIK kvm does not have them :(
> IOW: virtio is currently broken with kvm
Hi all,
It has been a while coming, but we have finally started work on
Kemari's port to KVM. For those not familiar with it, Kemari provides
the basic building block to create a virtualization-based fault
tolerant machine: a virtual machine synchronization mechanism.
Traditional high availabili
This patch reduces the size of memory being cleared on every virtio-blk IO.
Improve number of IOPS when using avirtio-blk device.
On every virtio-blk IO command passed to QEMU, virtio_blk_alloc_request()
allocates and clears (with qemu_mallocz()) a VirtIOBlockReq object.
The sizeof(VirtIOBlockReq
On Wed, Nov 11, 2009 at 01:15:45AM +0100, Alexander Graf wrote:
>
> On 11.11.2009, at 01:09, Anthony Liguori wrote:
>
> > Scott Tsai wrote:
> >> On Wed, Nov 11, 2009 at 1:06 AM, Luiz Capitulino >> > wrote:
> >>
> I'd certainly like to make this code useful for something other
> than
Hi All,
I'm using the following command to have two nics in multicast on the
same vlan. I see a storm of ARP requests. Does any one have any
suggestions?
qemu.bin.kvm84 -hda /live_disks/clone-disk.img -snapshot -serial
telnet:SERVER:5,nowait,server -monitor
tcp:SERVER:51000,server,nowait,nodel
diff -uNr qemu-kvm-0.10.6/x_keymap.c qemu-kvm-0.10.6.patched/x_keymap.c
--- qemu-kvm-0.10.6/x_keymap.c 2009-08-02 18:00:28.0 +0900
+++ qemu-kvm-0.10.6.patched/x_keymap.c 2009-08-22 05:04:07.0 +0900
@@ -94,7 +94,7 @@
*/
static const uint8_t evdev_keycode_to_pc_keycode[61] = {
On Wed, Nov 11, 2009 at 02:16:00PM +, Paul Brook wrote:
> On Wednesday 11 November 2009, Michael S. Tsirkin wrote:
> > On Wed, Nov 11, 2009 at 01:45:35PM +, Paul Brook wrote:
> > > If you don't need real barriers, then why does the kvm code have them?
> >
> > We need real barriers but AFAI
On 11/11/09 14:30, Hannes Reinecke wrote:
Gerd Hoffmann wrote:
How about sticking a 'void *hba_private' element into SCSIRequest instead?
Would work for me, too.
Pushed (scsi.v7 now).
cheers,
Gerd
On 11/11/2009 03:26 PM, Adam Litke wrote:
On Wed, 2009-11-11 at 10:12 +, Daniel P. Berrange wrote:
This all suggests that we should only update the stats from the guest
when something on the host actually asks for them by issuing the QEMU
monitor command. We don't want any kind of contin
On Wed, 2009-11-11 at 13:13 +1030, Rusty Russell wrote:
> > It's not laziness, it's consistency. How is actual different than free
> > memory or any other stat?
>
> Because it's a COLLECTION of stats. For example, swap in should be < swap
> out. Now, the current Linux implementation of all_vm_
On 11/11/09 15:13, Paul Brook wrote:
The current qemu code *does* cache the response. scsi-disk caps the
buffer at 128k (which is big enough for any request I've seen in my
testing). scsi-generic has no cap.
That cap is important.
For scsi-generic you probably don't have a choice because of t
On Wed, Nov 11, 2009 at 8:51 AM, Chad wrote:
> (this is mostly to get some ideas going rather than trying to get anything
> upstream... yet!)
>
> This version of tcg_out_mov for i386's tcg-target.c filters out the
>
> mov %ebx, %edx
> mov %ebx, [some index]
> mov %edx, %ebx
The question is: what
On Wednesday 11 November 2009, Michael S. Tsirkin wrote:
> On Wed, Nov 11, 2009 at 02:16:00PM +, Paul Brook wrote:
> > On Wednesday 11 November 2009, Michael S. Tsirkin wrote:
> > > On Wed, Nov 11, 2009 at 01:45:35PM +, Paul Brook wrote:
> > > > If you don't need real barriers, then why doe
>> That cap is important.
>> For scsi-generic you probably don't have a choice because of the way the
>> kernel interface works.
>
>Exactly. And why is the cap important for scsi-disk if scsi-generic
>does fine without?
With scsi-generic you're at the mercy of what the kernel API gives you, and i
Signed-off-by: Anthony Liguori
---
Makefile |1 +
check-qfloat.c | 81
configure |1 +
3 files changed, 83 insertions(+), 0 deletions(-)
create mode 100644 check-qfloat.c
diff --git a/Makefile b/Makefile
index 8d94fda
Signed-off-by: Anthony Liguori
---
Makefile |2 +-
qbool.c | 76 +
qbool.h | 29 +++
qobject.h |1 +
4 files changed, 107 insertions(+), 1 deletions(-)
create mode 100644 qbool.c
create mode 100644
Signed-off-by: Anthony Liguori
---
Makefile |1 +
check-qjson.c | 608 +
configure |2 +-
3 files changed, 610 insertions(+), 1 deletions(-)
create mode 100644 check-qjson.c
diff --git a/Makefile b/Makefile
index 3818c51.
On Wed, Nov 11, 2009 at 10:08 PM, Michael S. Tsirkin wrote:
> On Wed, Nov 11, 2009 at 01:45:35PM +, Paul Brook wrote:
>> If you don't need real barriers, then why does the kvm code have them?
>
> We need real barriers but AFAIK kvm does not have them :(
> IOW: virtio is currently broken with k
This is the third and final stage of the JSON parser. It parses lexical tokens
performing grammar validation and creating the final QObject representation. It
uses a recursive decent parser.
Signed-off-by: Anthony Liguori
---
Makefile |2 +-
json-parser.c | 560 ++
This lets us use QString for building larger strings
Signed-off-by: Anthony Liguori
---
qstring.c | 37 -
qstring.h |4
2 files changed, 40 insertions(+), 1 deletions(-)
diff --git a/qstring.c b/qstring.c
index 6d411da..441a9e6 100644
--- a/qstring
The second stage of our JSON parser is a simple state machine that identifies
individual JSON values by counting the levels of nesting of tokens. It does
not perform grammar validation. We use this to emit a full JSON value to the
parser.
Signed-off-by: Anthony Liguori
---
Makefile|
QDECREF does not properly escape the macro arguments which can lead to
unexpected syntax errors.
Signed-off-by: Anthony Liguori
---
qobject.h |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/qobject.h b/qobject.h
index 4cc9287..76f669f 100644
--- a/qobject.h
+++ b/qobjec
This makes lists no longer invariant. It's a very useful bit of functionality
though.
To deal with the fact that lists are no longer invariant, introduce a deep
copy mechanism for lists.
Signed-off-by: Anthony Liguori
---
qlist.c | 56
Signed-off-by: Anthony Liguori
---
Makefile |3 +-
qfloat.c | 76 +
qfloat.h | 29 +++
qobject.h |1 +
4 files changed, 108 insertions(+), 1 deletions(-)
create mode 100644 qfloat.c
create mode 100644
This provides a QObject interface for creating QObjects from a JSON expression.
Signed-off-by: Anthony Liguori
---
Makefile |2 +-
qjson.c | 60
qjson.h | 23 +++
3 files changed, 84 insertions(+), 1 delet
Our JSON parser is a three stage parser. The first stage tokenizes the stream
into a set of lexical tokens. Since the lexical grammar is regular, we can
use a finite state machine to model it. The state machine will emit tokens
as they are identified.
Signed-off-by: Anthony Liguori
---
Makefi
On Wed, Nov 11, 2009 at 1:54 AM, Juan Quintela wrote:
> Blue Swirl wrote:
>> On Tue, Nov 10, 2009 at 10:50 PM, Aurelien Jarno
>> wrote:
>>> Please note that at least qemu-system-arm, qemu-system-mips and
>>> qemu-system-mipsel are broken by this commit:
>>
>> Given that none of the devices touc
On Wed, Nov 11, 2009 at 7:47 PM, Blue Swirl wrote:
> On Wed, Nov 11, 2009 at 1:54 AM, Juan Quintela wrote:
>> Blue Swirl wrote:
>>> On Tue, Nov 10, 2009 at 10:50 PM, Aurelien Jarno
>>> wrote:
Please note that at least qemu-system-arm, qemu-system-mips and
qemu-system-mipsel are broke
We will have a linux boot option rom soon, so let's take all functionality
that might be useful for both to a header file that both roms can include.
That way we only have to write fw_cfg access code once.
Signed-off-by: Alexander Graf
---
pc-bios/optionrom/multiboot.S | 79 +-
The fw_cfg interface can only handle up to 16 bits of data for its streams.
While that isn't too much of a problem when handling integers, we would
like to stream full kernel images over that interface!
So let's extend it to 32 bit length variables.
Signed-off-by: Alexander Graf
---
hw/fw_cfg.c
We should install linuxboot.bin too, so let's add it to the to-be-installed
blobs.
Signed-off-by: Alexander Graf
---
Makefile |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/Makefile b/Makefile
index 30f1c9d..a6647c2 100644
--- a/Makefile
+++ b/Makefile
@@ -256,7 +256,7
We have several rom helpers currently, but none of them can get us
code that spans several roms into a pointer.
This patch introduces a function that copies over rom contents.
Signed-off-by: Alexander Graf
---
hw/loader.c | 38 ++
hw/loader.h |1 +
2 fi
Right now we load the guest kernel to RAM, fire off the BIOS, hope it
doesn't clobber memory and run an option rom that jumps into the kernel.
That breaks with SeaBIOS, as that clears memory. So let's read all
kernel, module etc. data using the fw_cfg interface when in the int19
handler.
This pat
We already have a working multiboot implementation that uses fw_cfg to get
its kernel module etc. data in int19 runtime now.
So what's missing is a working linux boot option rom. While at it I figured it
would be a good idea to take the opcode generator out of pc.c and instead use
a proper option
SeaBIOS clears RAM between we write our -kernel image to RAM and the int19
handler gets triggered.
So in order to work around that, I sat down and implemented Avi's suggestion
of "downloading" all blobs in runtime from the fw_cfg interface
Thanks to glommer who talked me into doing it ;-).
Alexa
On Thu, Nov 12, 2009 at 01:18:11AM +0800, Scott Tsai wrote:
> On Wed, Nov 11, 2009 at 10:08 PM, Michael S. Tsirkin wrote:
> > On Wed, Nov 11, 2009 at 01:45:35PM +, Paul Brook wrote:
> >> If you don't need real barriers, then why does the kvm code have them?
> >
> > We need real barriers but AF
On Thu, Nov 12, 2009 at 01:18:11AM +0800, Scott Tsai wrote:
> On Wed, Nov 11, 2009 at 10:08 PM, Michael S. Tsirkin wrote:
> > On Wed, Nov 11, 2009 at 01:45:35PM +, Paul Brook wrote:
> >> If you don't need real barriers, then why does the kvm code have them?
> >
> > We need real barriers but AF
On Wed, 11 Nov 2009 11:29:02 -0600
Anthony Liguori wrote:
> This provides a QObject interface for creating QObjects from a JSON
> expression.
>
> Signed-off-by: Anthony Liguori
> ---
> Makefile |2 +-
> qjson.c | 60
> qjson.
On Mon, Nov 9, 2009 at 3:35 PM, Anthony Liguori wrote:
> Filip Navara wrote:
>
>> The support for multiple keyboard devices is essential for emulating
>> embedded boards where multiple input devices are present (eg. keypad and
>> rotary encoder) which are implemented using separate QEMU devices.
>
On Tuesday 10 November 2009, Aurelien Jarno wrote:
> On Tue, Nov 10, 2009 at 11:19:40PM +0200, Blue Swirl wrote:
> > On Tue, Nov 10, 2009 at 10:50 PM, Aurelien Jarno
wrote:
> > > Please note that at least qemu-system-arm, qemu-system-mips and
> > > qemu-system-mipsel are broken by this commit:
>
On Thu, Nov 12, 2009 at 2:09 AM, Michael S. Tsirkin wrote:
>> I do have a newbie question, when exactly would vrtio have to handle
>> concurrent access from multiple threads?
>> My current reading of the code suggests:
>> 1. when CONFIG_IOTHREAD is true
>> 2. when CONFIG_KVM is true and the guest
On Tue, Nov 10, 2009 at 11:19:40PM +0200, Blue Swirl wrote:
> On Tue, Nov 10, 2009 at 10:50 PM, Aurelien Jarno wrote:
> > Please note that at least qemu-system-arm, qemu-system-mips and
> > qemu-system-mipsel are broken by this commit:
>
> Given that none of the devices touched by the commit shou
From: Luiz Capitulino
This adds iterator support to QDict, it will be used by the
(to be introduced) QError module.
Signed-off-by: Luiz Capitulino
Signed-off-by: Anthony Liguori
---
qdict.c | 19 +++
qdict.h |3 +++
2 files changed, 22 insertions(+), 0 deletions(-)
diff
By reusing the qjson test suite. After checking that we can demarshal, marshal
again and compared to the expected decoded value. This doesn't work so well
for floats because they cannot be accurately represented in decimal but we
try our best.
Signed-off-by: Anthony Liguori
---
check-qjson.c |
This introduces qobject_to_json which will convert a QObject to a JSON string
representation.
Signed-off-by: Anthony Liguori
---
qjson.c | 178 +++
qjson.h |3 +
2 files changed, 181 insertions(+), 0 deletions(-)
diff --git a/qjso
Luiz Capitulino wrote:
I think we should abort() on error, assuming the only way to fail
is a bad syntax.
We'll be using qobject_from_jsonf() to parse incoming QMP traffic. We
definitely don't want to abort when we receive invalid input on the QMP
port. Best to just close the session
On Wed, 11 Nov 2009 13:39:29 -0600
Anthony Liguori wrote:
> Luiz Capitulino wrote:
> > I think we should abort() on error, assuming the only way to fail
> > is a bad syntax.
> >
>
> We'll be using qobject_from_jsonf() to parse incoming QMP traffic. We
> definitely don't want to abort when
On Thu, Nov 12, 2009 at 02:37:26AM +0800, Scott Tsai wrote:
> On Thu, Nov 12, 2009 at 2:09 AM, Michael S. Tsirkin wrote:
> >> I do have a newbie question, when exactly would vrtio have to handle
> >> concurrent access from multiple threads?
> >> My current reading of the code suggests:
> >> 1. whe
On Wed, Nov 11, 2009 at 8:57 PM, Glauber Costa wrote:
> On Tue, Nov 10, 2009 at 11:19:40PM +0200, Blue Swirl wrote:
>> On Tue, Nov 10, 2009 at 10:50 PM, Aurelien Jarno
>> wrote:
>> > Please note that at least qemu-system-arm, qemu-system-mips and
>> > qemu-system-mipsel are broken by this commit
On Wed, Nov 11, 2009 at 8:28 PM, Paul Brook wrote:
> On Tuesday 10 November 2009, Aurelien Jarno wrote:
>> On Tue, Nov 10, 2009 at 11:19:40PM +0200, Blue Swirl wrote:
>> > On Tue, Nov 10, 2009 at 10:50 PM, Aurelien Jarno
> wrote:
>> > > Please note that at least qemu-system-arm, qemu-system-mips
Luiz Capitulino wrote:
Hi,
I can't remember seeing updated versions of a RFC series, but this should
prevent Anthony's scripts from merging these patches.
This new QError version has two major changes: the static error table has
been dropped and I'm using symbolic names instead of error code
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
Hi folks,
I'm writing a virtio-rng host-side driver for qemu-kvm, and I've got
something up and running that works, and will pass data gathered from a
char device on the host through to the virtio-rng driver on a guest copy
of linux.
Ultimately it'll
Alexander Graf wrote:
The fw_cfg interface can only handle up to 16 bits of data for its streams.
While that isn't too much of a problem when handling integers, we would
like to stream full kernel images over that interface!
So let's extend it to 32 bit length variables.
Signed-off-by: Alexande
Alexander Graf wrote:
We have several rom helpers currently, but none of them can get us
code that spans several roms into a pointer.
This patch introduces a function that copies over rom contents.
Signed-off-by: Alexander Graf
---
hw/loader.c | 38 ++
hw
> > This is latent breakage introduced by 45a50b1.
> > See commits 97fe84f5 (makes breakage obvious) and f2d7497 (fixed ARM).
> > MIPS still needs fixing.
>
> I can't find 97fe84f5 or f2d7497, what commits are those?
http://git.savannah.gnu.org/cgit/qemu.git/commit/?id=97fe84f5
http://git.savanna
Anthony Liguori wrote:
> Alexander Graf wrote:
>> The fw_cfg interface can only handle up to 16 bits of data for its
>> streams.
>> While that isn't too much of a problem when handling integers, we would
>> like to stream full kernel images over that interface!
>>
>> So let's extend it to 32 bit le
Alexander Graf wrote:
Anthony Liguori wrote:
Alexander Graf wrote:
The fw_cfg interface can only handle up to 16 bits of data for its
streams.
While that isn't too much of a problem when handling integers, we would
like to stream full kernel images over that interface!
So let's extend
> I'm writing a virtio-rng host-side driver for qemu-kvm, and I've got
> something up and running that works, and will pass data gathered from a
> char device on the host through to the virtio-rng driver on a guest copy
> of linux.
Why do you need a special device? Isn't a regular serial data stre
This partly reverts 2be3783328: First, the conversion neglected to
update the opaque translation in fdc_pre_save/fdc_post_load which causes
memory corruptions on vmsave/restore. And second, we can't apply a
common translation here as DeviceState->fdctrl_t is different for sysbus
and ISA.
Signed-of
Jan Kiszka wrote:
> This partly reverts 2be3783328: First, the conversion neglected to
> update the opaque translation in fdc_pre_save/fdc_post_load which causes
> memory corruptions on vmsave/restore. And second, we can't apply a
> common translation here as DeviceState->fdctrl_t is different for
Hi,
I'm working on this patch and I have some trouble after a data IN
transaction.
I submit the urb but I don't receive any async completation.
88011e512140 2614190796 S Ii:2:008:7 -115:128 16 <
husb: data submit. ep 0x87 len 16 aurb 0xcf3850
So the state machine is blocked
Any hints?
On Wed, 11 Nov 2009 15:20:30 -0600
Anthony Liguori wrote:
> Luiz Capitulino wrote:
> > Hi,
> >
> > I can't remember seeing updated versions of a RFC series, but this should
> > prevent Anthony's scripts from merging these patches.
> >
> > This new QError version has two major changes: the stat
Hi
With this three patches (on top of the one already in staging)
I am able to get migration working with today qemu/master.
- fdc: vmstate+reset qdev change made to use an vmstate that expected
an fdctrl_t variable and received a fdctrl_isabus_t variable. You can guess
what happened.
- qemu
When code was transformed to use qdev_reset/vmstate registration, vmstate
was passed a variable of the wrong type
Signed-off-by: Juan Quintela
---
hw/fdc.c | 33 +
1 files changed, 29 insertions(+), 4 deletions(-)
diff --git a/hw/fdc.c b/hw/fdc.c
index d2bfa71.
Signed-off-by: Juan Quintela
---
vl.c |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/vl.c b/vl.c
index bf91ee1..fff8e8d 100644
--- a/vl.c
+++ b/vl.c
@@ -4044,7 +4044,6 @@ static void main_loop(void)
qemu_system_ready = 1;
qemu_cond_broadcast(&qemu_system_cond
Signed-off-by: Juan Quintela
---
hw/pci.c |4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/hw/pci.c b/hw/pci.c
index 2ab1117..a326930 100644
--- a/hw/pci.c
+++ b/hw/pci.c
@@ -273,9 +273,9 @@ static int get_pci_config_device(QEMUFile *f, void *pv,
size_t size)
/* just
On 11.11.2009, at 22:57, Anthony Liguori wrote:
Alexander Graf wrote:
We have several rom helpers currently, but none of them can get us
code that spans several roms into a pointer.
This patch introduces a function that copies over rom contents.
Signed-off-by: Alexander Graf
---
hw/loader.c
On 11.11.2009, at 23:22, Anthony Liguori wrote:
Alexander Graf wrote:
Anthony Liguori wrote:
Alexander Graf wrote:
The fw_cfg interface can only handle up to 16 bits of data for its
streams.
While that isn't too much of a problem when handling integers, we
would
like to stream full kern
Alexander Graf wrote:
Juan, I'd really love to learn some new voodoo :-).
This whole new qdev whatever based save format was supposed to make
things like this easy, right? I would've known what to do with the old
code ...
I think Juan's mentioned something about writing a doc explaining how t
When I tried to compile qemu git with --enable-io-thread on Mac OS X 10.6.2, it
botches on qemu-thread:
CCqemu-thread.o
/Users/cwbetts/makestuff/qemu/qemu-thread.c: In function ‘qemu_mutex_timedlock’:
/Users/cwbetts/makestuff/qemu/qemu-thread.c:66: warning: implicit declaration
of function
On Thu, 12 Nov 2009 01:38:34 am Adam Litke wrote:
> > But it raises the question: what stats are generally useful cross-OS?
> > Should
> > we be supplying numbers like "unused" (free) "instantly discardable" (ie.
> > clean), "discardable to disk" (ie. file-backed), "discardable to swap"
> > (ie.
1 - 100 of 129 matches
Mail list logo
