On 02/09/2018 06:57 PM, Greg Kurz wrote:
On Fri, 9 Feb 2018 10:06:05 -0600
Eric Blake wrote:
On 02/09/2018 09:13 AM, Greg Kurz wrote:
On Thu, 8 Feb 2018 19:00:18 +0100
wrote:
From: Antonios Motakis
To support multiple devices on the 9p share, and avoid
qid path collisions we take
On 02/09/2018 02:03 PM, Greg Kurz wrote:
On Thu, 8 Feb 2018 19:00:17 +0100
wrote:
From: Antonios Motakis
The QID path should uniquely identify a file. However, the
inode of a file is currently used as the QID path, which
on its own only uniquely identifies wiles within a device
On 02/09/2018 01:37 PM, Greg Kurz wrote:
On Thu, 8 Feb 2018 19:00:16 +0100
wrote:
From: Antonios Motakis
There is no need for signedness on these QID fields for 9p.
Signed-off-by: Antonios Motakis
---
I agree these should be unsigned, but you have some more places to adapt
to this
On 02/09/2018 10:47 PM, Emilio G. Cota wrote:
On Fri, Feb 09, 2018 at 16:22:33 +0100, Greg Kurz wrote:
On Thu, 8 Feb 2018 19:00:19 +0100
wrote:
(snip)
/* stat_to_qid needs to map inode number (64 bits) and device id (32 bits)
* to a unique QID path (64 bits). To avoid having to map and
On 02/09/2018 10:58 PM, Emilio G. Cota wrote:
On Fri, Feb 09, 2018 at 16:13:26 +0100, Greg Kurz wrote:
On Thu, 8 Feb 2018 19:00:18 +0100
wrote:
(snip)
+/* creative abuse of tb_hash_func7, which is based on xxhash */
+static uint32_t qpp_hash(QppEntry e)
+{
+return tb_hash_func7(e.ino_pr
On 01/24/2018 02:30 PM, Greg Kurz wrote:
Thanks Emilio for providing these valuable suggestions ! :)
On Sat, 20 Jan 2018 17:03:49 -0500
"Emilio G. Cota" wrote:
On Fri, Jan 19, 2018 at 19:05:06 -0500, Emilio G. Cota wrote:
On Fri, 12 Jan 2018 19:32:10 +0800
Antonios Motakis wro
On 13-Jan-18 00:14, Greg Kurz wrote:
> On Fri, 12 Jan 2018 19:32:10 +0800
> Antonios Motakis wrote:
>
>> Hello all,
>>
>
> Hi Antonios,
>
> I see you have attached a patch to this email... this really isn't the
> preferred
> way to do things sinc
ay to eliminate these issues completely.
This is the extent that we were able to analyze the issue from our side, we
would like to hear if there are some better ideas, or which approach would be
more appropriate for upstream.
Best regards,
--
Antonios Motakis
Virtualization Engineer
Huawei Technol
;
> >> @@ -380,7 +492,7 @@ static int vfio_mmap_region(VFIODevice *vdev,
> VFIORegion *region,
> >> goto error;
> >> }
> >>
> >> -memory_region_init_ram_ptr(submem, OBJECT(vdev), name, size,
> *map);
> >> + m
On Tue, Mar 4, 2014 at 7:38 PM, Michael S. Tsirkin wrote:
> On Tue, Mar 04, 2014 at 07:22:56PM +0100, Antonios Motakis wrote:
> > This will be used in a following patch to ensure that a vhost-user
> > client reconnecting to QEMU supports the features that were exposed
> >
Hello Andreas,
On Tue, Mar 4, 2014 at 7:39 PM, Andreas Färber wrote:
> Am 04.03.2014 19:23, schrieb Antonios Motakis:
> > This test creates a 'server' chardev to listen for vhost-user messages.
> > Once VHOST_USER_SET_MEM_TABLE is received it mmaps each received regi
Hello,
On Tue, Mar 4, 2014 at 7:45 PM, Michael S. Tsirkin wrote:
> On Tue, Mar 04, 2014 at 07:22:53PM +0100, Antonios Motakis wrote:
> > On stopping the vhost, a call to VHOST_GET_VRING_BASE is issued. The
> > received value is stored as last_avail_idx, so the virtque
Extend -mem-path with additional property:
- share=on|off - default off, memory is mmapped with MAP_SHARED flag
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
exec.c | 21 +++--
include/exec/cpu-all.h | 1 -
qemu-options.hx| 8
a special handling in case of I/O
failure - return -1 to indicate the upper layer that it failed.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/virtio/Makefile.objs | 2 +-
hw/virtio/vhost-backend.c | 5 +
hw/virtio/vhost-user.c| 356
tify_event
- vc_init
and this array:
- serial_hds
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
stubs/Makefile.objs | 8
stubs/bdrv-commit-all.c | 7 +++
stubs/chr-msmouse.c | 7 +++
stubs/get-next-serial.c | 3 +++
stubs/is-daemonized
This will be used in a following patch to ensure that a vhost-user
client reconnecting to QEMU supports the features that were exposed
by the first client that initiated the virtio-net session.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/net/vhost_net.c| 10
Use vhost_set_backend_type to initialise a proper vhost_ops structure.
In vhost_net_init and vhost_net_start_one call conditionally TAP related
initialisation depending on the vhost backend type.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/net/vhost_net.c
This function will attempt to read data from the chardev trying
to fill the buffer up to the given length.
Add tcp_chr_disconnect to reuse disconnection code where needed.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
include/sysemu/char.h | 14 +
qemu-char.c
This will set an array of file descriptors to the internal structures.
The next time a message is send the array will be send as ancillary
data. This feature works on the UNIX domain socket backend only.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
include/sysemu/char.h
, share and unlink properties
- Set 1 sec timeout when read/write to the unix domain socket
- Fix file descriptor leak
Changes from v2:
- Reconnect when the backend disappears
Changes from v1:
- Implementation of vhost-user netdev backend
- Code improvements
Antonios Motakis (20):
On Tue, Mar 4, 2014 at 7:29 PM, Paolo Bonzini wrote:
> Il 04/03/2014 19:22, Antonios Motakis ha scritto:
>
> In this patch series we would like to introduce our approach for putting a
>> virtio-net backend in an external userspace process. Our eventual target
>> is to
>&
pass the fd.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/net/vhost_net.c| 23 ---
hw/scsi/vhost-scsi.c | 10 +-
hw/virtio/vhost.c | 12 +++-
include/hw/virtio/vhost.h | 2 +-
include/net/vhost_net.h | 8
This function allows to get the current available ring index.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/virtio/virtio.c | 5 +
include/hw/virtio/virtio.h | 1 +
2 files changed, 6 insertions(+)
diff --git a/hw/virtio/virtio.c b/hw/virtio/virtio.c
index
This document describes the basic message format used by vhost-user
for communication over a unix domain socket. The protocol is based
on the existing ioctl interface used for the kernel version of vhost.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
docs/specs/vhost
Add a function to check if the eventfd capability is present in KVM in
the host kernel.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
include/sysemu/kvm.h | 11 +++
kvm-all.c| 4
kvm-stub.c | 1 +
3 files changed, 16 insertions
compatible with the first one.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
include/net/vhost-user.h | 17 ++
net/Makefile.objs| 2 +-
net/clients.h| 3 +
net/vhost-user.c | 150 +++
ount
point defaults to '/hugetlbfs' and can be specified via the environment
variable QTEST_HUGETLBFS_PATH.
The rom pc-bios/pxe-virtio.rom is used to instantiate a virtio pcicontroller.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
tests/Makefile | 4
Decouple vhost from the Linux kernel by introducing vhost_ops. The
intention is to provide different backends - a 'kernel' backend based on
the ioctl interface, and an 'user' backend based on a UNIX domain socket
and shared memory interface.
Signed-off-by: Antonios Motakis
Si
This extends the existing qemu_chr_fe_get_msgfd by allowing to read a set
of fds. The function for receiving the fds - unix_process_msgfd is extended
to allocate the needed array size.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
include/sysemu/char.h | 15
checks for validity:
- requires `-mempath ...,share=on`
- requires `-device virtio-net-*`
The `vhostforce` option is used to force vhost-net when we deal with
non-MSIX guests.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hmp-commands.hx| 4 +-
hw/net/vhost_net.c
This is used to detect that the remote end has disconnected. Just call
tcp_char_disconnect on receiving this event.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
qemu-char.c | 16
1 file changed, 16 insertions(+)
diff --git a/qemu-char.c b/qemu-char.c
still we keep a sane value and can continue on reconnect.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/virtio/vhost.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/hw/virtio/vhost.c b/hw/virtio/vhost.c
index 9e336ad..322e2c0 100644
--- a/hw
This decouples virtio-net from the TAP netdev backend and allows support
for other backends to be implemented.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/net/vhost_net.c | 30 +++---
hw/net/virtio-net.c | 29
The poll callback needs to be called when bringing up or down
the vhost_net instance. As it is not mandatory for an NetClient
to implement it, invoke it only when it is set.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/net/vhost_net.c | 13 ++---
1 file
On Mon, Feb 17, 2014 at 8:56 AM, Michael S. Tsirkin wrote:
> On Mon, Feb 17, 2014 at 12:42:45AM +0100, Paolo Bonzini wrote:
> > Il 15/02/2014 19:10, Michael Tokarev ha scritto:
> > > 13 февраля 2014 г. 16:03:12 GMT+04:00, Antonios Motakis <
> a.mota...@virtu
ount
point defaults to '/hugetlbfs' and can be specified via the environment
variable QTEST_HUGETLBFS_PATH.
The rom pc-bios/pxe-virtio.rom is used to instantiate a virtio pcicontroller.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
tests/Makefile | 4
tify_event
- vc_init
and this array:
- serial_hds
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
stubs/Makefile.objs | 8
stubs/bdrv-commit-all.c | 7 +++
stubs/chr-msmouse.c | 7 +++
stubs/get-next-serial.c | 3 +++
stubs/is-daemonized
slave is compatible with the first one.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
include/net/vhost-user.h | 17 ++
net/Makefile.objs| 2 +-
net/clients.h| 3 +
net/vhost-user.c | 150 +++
Decouple vhost from the Linux kernel by introducing vhost_ops. The
intention is to provide different backends - 'kernel' backend based on
the ioctl interface, and 'user' backend based on an Unix domain socket and
shared memory.
Signed-off-by: Antonios Motakis
Signed-off-
This will be used in a following patch to ensure that a vhost-user
client reconnecting to QEMU supports the features that were exposed
by the first client that initiated the virtio-net session.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/net/vhost_net.c| 10
This document describes the basic message format used by vhost-user
for communication over a unix domain socket. The protocol is based
on the existing ioctl interface used for the kernel version of vhost.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
docs/specs/vhost
pass the fd.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/net/vhost_net.c| 24 +---
hw/scsi/vhost-scsi.c | 10 +-
hw/virtio/vhost.c | 12 +++-
include/hw/virtio/vhost.h | 2 +-
include/net/vhost_net.h | 8
checks for validity:
- requires `-mempath ...,share=on`
- requires `-device virtio-net-*`
The `vhostforce` option is uset to force vhost-net when we deal with no-MSIX
guest.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hmp-commands.hx | 4 +-
hw/net/vhost_net.c
global ram_list for ram blocks with a valid fd field set. This would
be set when the -mem-path option with shared=on property is used.
Upon receiving VHOST_USER_GET_FEATURES reply, the new features value is
compared to the mandatory features in the vhost_dev.
Signed-off-by: Antonios Motakis
Signed
This decouples virtio-net from the TAP netdev backend and allows support
for other backends to be implemented.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/net/vhost_net.c | 30 +++---
hw/net/virtio-net.c | 39
This extends the existing qemu_chr_fe_get_msgfd by allowing to read a set
of fds. The function for receiving the fds - unix_process_msgfd is extended
to allocate the needed array size.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
include/sysemu/char.h | 15
The poll callback needs to be called when bringing up or down
the vhost_net instance. As it is not mandatory for an NetClient
to implement it, invoke it only when it is set.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/net/vhost_net.c | 13 ++---
1 file
Use vhost_set_backend_type to initialise a proper vhost_ops structure.
In vhost_net_init and vhost_net_start_one call conditionally TAP related
initialisation depending on the vhost backend type.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/net/vhost_net.c
This function will attempt to read data from the chardev trying
to fill the buffer up to the given length.
Add tcp_chr_disconnect to reuse disconnection code where needed.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
include/sysemu/char.h | 14 +
qemu-char.c
Extend -mem-path with additional properties:
- prealloc=on|off - default off, same as -mem-prealloc
- share=on|off - default off, memory is mmapped with MAP_SHARED flag
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
exec.c | 30
This will set an array of file descriptors to the internal structures.
The next time a message is sent the array will be sent as ancillary
data. This feature works with the Unix domain socket backend only.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
include/sysemu
This is used to detect that the remote end has disconnected. Just call
tcp_char_disconnect on receiving this event.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
qemu-char.c | 16
1 file changed, 16 insertions(+)
diff --git a/qemu-char.c b/qemu-char.c
3:
- Convert -mem-path to QemuOpts with prealloc, share and unlink properties
- Set 1 sec timeout when read/write to the unix domain socket
- Fix file descriptor leak
Changes from v2:
- Reconnect when the backend disappears
Changes from v1:
- Implementation of vhost-user netdev backend
- C
On Mon, Feb 10, 2014 at 9:42 AM, Michael S. Tsirkin wrote:
> On Fri, Jan 31, 2014 at 06:34:40PM +0100, Antonios Motakis wrote:
>> Add a new QEMU netdev backend that is intended to invoke vhost_net with the
>> vhost-user backend.
>>
>> At runtime the netdev will detect
Hello,
On Mon, Feb 10, 2014 at 9:57 AM, Michael S. Tsirkin wrote:
>
> On Fri, Jan 31, 2014 at 06:34:29PM +0100, Antonios Motakis wrote:
> > In this patch series we would like to introduce our approach for putting a
> > virtio-net backend in an external userspace process. Our
Extend -mem-path with additional properties:
- prealloc=on|off - default off, same as -mem-prealloc
- share=on|off - default off, memory is mmapped with MAP_SHARED flag
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
exec.c | 30
This function will attempt to read data from the chardev trying
to fill the buffer up to the given length.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
include/sysemu/char.h | 14 +++
qemu-char.c | 65
Decouple vhost from the Linux kernel by introducing vhost_ops. The
intention is to provide different backends - 'kernel' backend based on
the ioctl interface, and 'user' backend based on a unix domain socket and
shared memory.
Signed-off-by: Antonios Motakis
Signed-off-
This decouples virtio-net from the TAP netdev backend and allows support
for other backends to be implemented.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/net/vhost_net.c | 30 +++---
hw/net/virtio-net.c | 39
-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hmp-commands.hx | 4 +--
hw/net/vhost_net.c | 4 +++
hw/net/virtio-net.c | 3 ++
net/hub.c | 1 +
net/net.c | 2 ++
net/vhost-user.c| 91 +++--
qapi
d unlink properties
- Set 1 sec timeout when read/write to the unix domain socket
- Fix file descriptor leak
Changes from v2:
- Reconnect when the backend disappears
Changes from v1:
- Implementation of vhost-user netdev backend
- Code improvements
Antonios Motakis (13):
Convert -me
This document describes the basic message format used by vhost-user
for communication over a unix domain socket. The protocol is based
on the existing ioctl interface used for the kernel version of vhost.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
docs/specs/vhost
The poll callback needs to be called when bringing up or down
the vhost_net instance. As it is not mandatory for an NetClient
to implement it, invoke it only when it is set.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/net/vhost_net.c | 13 ++---
1 file
This will set an array of file descriptors to the internal structures.
The next time a message is send the array will be send as ancillary
data. This feature works on unix domain socket backend only.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
include/sysemu/char.h
Use vhost_set_backend_type to initialise a proper vhost_ops structure.
In vhost_net_init and vhost_net_start_one call conditionally TAP related
initialisation depending on the vhost backend type.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/net/vhost_net.c
pass the fd.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/net/vhost_net.c| 24 +---
hw/scsi/vhost-scsi.c | 10 +-
hw/virtio/vhost.c | 12 +++-
include/hw/virtio/vhost.h | 2 +-
include/net/vhost_net.h | 8
Close the chardev on receiving this event.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
qemu-char.c | 15 +++
1 file changed, 15 insertions(+)
diff --git a/qemu-char.c b/qemu-char.c
index c2e599e..1c34b2b 100644
--- a/qemu-char.c
+++ b/qemu-char.c
@@ -2643,6
: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
include/net/vhost-user.h | 17 +++
net/Makefile.objs| 2 +-
net/clients.h| 3 ++
net/vhost-user.c | 130 +++
4 files changed, 151 insertions(+), 1 deletion(-)
create
global ram_list for ram blocks with a valid fd field set. This would
be set when the -mem-path option with shared=on property is used.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/virtio/Makefile.objs | 2 +-
hw/virtio/vhost-backend.c | 5 +
hw/virtio/vhost-user.c
Hello,
On Mon, Jan 27, 2014 at 5:49 PM, Michael S. Tsirkin wrote:
>
> On Mon, Jan 27, 2014 at 05:37:02PM +0100, Antonios Motakis wrote:
> > Hello again,
> >
> >
> > On Wed, Jan 15, 2014 at 3:49 PM, Michael S. Tsirkin wrote:
> > >
> > > On Wed, Ja
Hello again,
On Wed, Jan 15, 2014 at 3:49 PM, Michael S. Tsirkin wrote:
>
> On Wed, Jan 15, 2014 at 01:50:47PM +0100, Antonios Motakis wrote:
> >
> >
> >
> > On Wed, Jan 15, 2014 at 10:07 AM, Michael S. Tsirkin
> > wrote:
> >
> > On Tue, Ja
Hello,
On Thu, Jan 23, 2014 at 9:27 AM, Lei Li wrote:
> This patch series tries to refactor the functions used for
> exchange of FD in current code, provide common methods
> for it.
>
> I just tested it through page flipping migration, and tap/
> bridge-helper a bit, but have some environment p
Hello,
On Wed, Jan 15, 2014 at 11:05 AM, Michael S. Tsirkin wrote:
>
> On Mon, Jan 13, 2014 at 03:25:11PM +0100, Antonios Motakis wrote:
> > In this patch series we would like to introduce our approach for putting a
> > virtio-net backend in an external userspace process. Our
On Wed, Jan 15, 2014 at 10:07 AM, Michael S. Tsirkin wrote:
> On Tue, Jan 14, 2014 at 07:13:43PM +0100, Antonios Motakis wrote:
>>
>>
>>
>> On Tue, Jan 14, 2014 at 12:33 PM, Michael S. Tsirkin wrote:
>>
>> On Mon, Jan 13, 2014 at 03:25:11PM +0100, A
On Wed, Jan 15, 2014 at 3:49 PM, Michael S. Tsirkin wrote:
> On Wed, Jan 15, 2014 at 01:50:47PM +0100, Antonios Motakis wrote:
> >
> >
> >
> > On Wed, Jan 15, 2014 at 10:07 AM, Michael S. Tsirkin
> wrote:
> >
> > On Tue, Jan 14, 2014 a
On Wed, Jan 15, 2014 at 10:07 AM, Michael S. Tsirkin wrote:
> On Tue, Jan 14, 2014 at 07:13:43PM +0100, Antonios Motakis wrote:
> >
> >
> >
> > On Tue, Jan 14, 2014 at 12:33 PM, Michael S. Tsirkin
> wrote:
> >
> > On Mon, Jan 13, 2014 a
On Tue, Jan 14, 2014 at 12:10 PM, Michael S. Tsirkin wrote:
> On Mon, Jan 13, 2014 at 03:25:16PM +0100, Antonios Motakis wrote:
> > Add structures for passing vhost-user messages over a unix domain socket.
> > This is the equivalent of the existing vhost-kernel ioctls.
> &g
On Tue, Jan 14, 2014 at 12:21 PM, Michael S. Tsirkin wrote:
> On Mon, Jan 13, 2014 at 03:25:17PM +0100, Antonios Motakis wrote:
> > Each ioctl request of vhost-kernel has a vhost-user message equivalent,
> > which is sent over the control socket.
> >
> > The general a
On Tue, Jan 14, 2014 at 12:16 PM, Michael S. Tsirkin wrote:
> On Mon, Jan 13, 2014 at 03:25:13PM +0100, Antonios Motakis wrote:
> > The unlink option allows the created file to be externally deleted
> > after QEMU is terminated.
> >
> > - unlink=on|off - defaul
On Tue, Jan 14, 2014 at 12:33 PM, Michael S. Tsirkin wrote:
> On Mon, Jan 13, 2014 at 03:25:11PM +0100, Antonios Motakis wrote:
> > In this patch series we would like to introduce our approach for putting
> a
> > virtio-net backend in an external userspace process. Our event
Hello,
On Tue, Jan 14, 2014 at 12:14 PM, Michael S. Tsirkin wrote:
> On Mon, Jan 13, 2014 at 03:25:11PM +0100, Antonios Motakis wrote:
> > In this patch series we would like to introduce our approach for putting
> a
> > virtio-net backend in an external userspace process. O
Add a new QEMU netdev backend that is intended to invoke vhost_net
with the vhost-user backend. Also decouple virtio-net from the tap
backend.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hmp-commands.hx | 4 +-
hw/net/vhost_net.c | 66
for inclusion in
the sendmsd control header.
VHOST_SET_MEM_TABLE ignores the supplied vhost_memory structure and scans
the global ram_list for ram blocks with a valid fd field set. This would
be set when the -mem-path option with shared=on property is used.
Signed-off-by: Antonios Motakis
Signed
SCM_RIGHTS type in the message control header.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/virtio/vhost-backend.c | 306 +-
1 file changed, 301 insertions(+), 5 deletions(-)
diff --git a/hw/virtio/vhost-backend.c b/hw/virtio/vhost
Add empty vhost_call, init and cleanup for the vhost-user backend.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/net/vhost_net.c| 57 ++-
hw/virtio/vhost-backend.c | 35
include/hw
The unlink option allows the created file to be externally deleted
after QEMU is terminated.
- unlink=on|off - default on, unlink the file after opening it
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
exec.c | 18 +-
qemu-options.hx | 7
option:
-netdev vhost-user,path=/path/to/sock[,poll_time=poll_time]
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/net/vhost_net.c| 16 +
hw/virtio/vhost-backend.c | 25 --
include/hw/virtio/vhost-backend.h | 2 ++
include
Extend -mem-path with additional properties:
- prealloc=on|off - default off, same as -mem-prealloc
- share=on|off - default off, memory is mmapped with MAP_SHARED flag
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
exec.c | 41
nly in the vhost-backend (ioctl, open, close).
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/net/vhost_net.c| 13 +---
hw/scsi/vhost-scsi.c | 13 +---
hw/virtio/Makefile.objs | 2 +-
hw/virtio/vhost-backend.c |
n the backend disappears
Changes from v1:
- Implementation of vhost-user netdev backend
- Code improvements
Antonios Motakis (8):
Convert -mem-path to QemuOpts and add prealloc and share properties
New -mem-path option - unlink.
Decouple vhost from kernel interface
Add vhost-user skeleton
On Thu, Jan 9, 2014 at 4:31 PM, Michael S. Tsirkin wrote:
> On Thu, Jan 09, 2014 at 03:59:58PM +0100, Antonios Motakis wrote:
>> Add structures for passing vhost-user messages over a unix domain socket.
>> This is the equivalent of the existing vhost-kernel ioctls.
>>
>>
On Thu, Jan 9, 2014 at 4:47 PM, Michael S. Tsirkin wrote:
> On Thu, Jan 09, 2014 at 03:59:59PM +0100, Antonios Motakis wrote:
> > Each ioctl request of vhost-kernel has a vhost-user message equivalent,
> > which is sent it over the control socket.
> >
> > The general a
On Thu, Jan 9, 2014 at 5:01 PM, Michael S. Tsirkin wrote:
> On Thu, Jan 09, 2014 at 03:59:55PM +0100, Antonios Motakis wrote:
> > Extend -mem-path with additional properties:
> >
> > - prealloc=on|off - default off, same as -mem-prealloc
> > - share=on|off - default
On Thu, Jan 9, 2014 at 5:14 PM, Michael S. Tsirkin wrote:
> On Thu, Jan 09, 2014 at 04:00:00PM +0100, Antonios Motakis wrote:
> > Add a new QEMU netdev backend that is intended to invoke vhost_net
> > with the vhost-user backend. Also decouple virtio-net from the tap
> > ba
On Thu, Jan 9, 2014 at 5:16 PM, Michael S. Tsirkin wrote:
> On Thu, Jan 09, 2014 at 04:00:01PM +0100, Antonios Motakis wrote:
> > At runtime vhost-user netdev will detect if the vhost backend is up or
> down.
> > Upon disconnection it will set link_down accordingly and n
On Thu, Jan 9, 2014 at 5:11 PM, Michael S. Tsirkin wrote:
> On Thu, Jan 09, 2014 at 03:59:54PM +0100, Antonios Motakis wrote:
> > In this patch series we would like to introduce our approach for putting
> a
> > virtio-net backend in an external userspace process. Our event
fds array for inclusion in
the sendmsd control header.
VHOST_SET_MEM_TABLE ignores the supplied vhost_memory structure and scans
the global ram_list for ram blocks wiht a valid fd field set. This would
be set when the -mem-path option with shared=on property is used.
Signed-off-by: Antonios
nly in the vhost-backend (ioctl, open, close).
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/net/vhost_net.c| 13 +---
hw/scsi/vhost-scsi.c | 13 +---
hw/virtio/Makefile.objs | 2 +-
hw/virtio/vhost-backend.c |
At runtime vhost-user netdev will detect if the vhost backend is up or down.
Upon disconnection it will set link_down accordingly and notify virtio-net.
Signed-off-by: Antonios Motakis
Signed-off-by: Nikolay Nikolaev
---
hw/net/vhost_net.c| 16 +++
hw/virtio/vhost
le descriptor leak
Changes from v2:
- Reconnect when the backend disappears
Changes from v1:
- Implementation of vhost-user netdev backend
- Code improvements
Antonios Motakis (7):
Convert -mem-path to QemuOpts and add prealloc,share and unlink
properties
Decouple vhost from kernel interf
1 - 100 of 136 matches
Mail list logo
