On Fri, Oct 11, 2019 at 03:58:04PM +0200, Mikhail Sennikovsky wrote:
> > Currently offloads disabled by guest via the
> > VIRTIO_NET_CTRL_GUEST_OFFLOADS_SET
> > command are not preserved on VM migration.
> > Instead all offloads reported by guest features (via
> > VIRTIO
igned-off-by: Mikhail Sennikovsky
---
hw/net/virtio-net.c| 27 ---
include/hw/virtio/virtio-net.h | 2 ++
2 files changed, 26 insertions(+), 3 deletions(-)
diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c
index b9e1cd7..6adb0fe 100644
--- a/hw/net/virtio
ok in a device class instead. Devices should use
this unless they specifically want to verify the migration stream as
it's processed, e.g. for bounds checking.
Suggested-by: "Dr. David Alan Gilbert"
Cc: Mikhail Sennikovsky
Signed-off-by: Michael S. Tsirkin
---
hw/virtio/virtio.c
I still wonder though if this approach is really cleaner than my
original one of having an extra argument in set_features callback,
saying whether the device settings (offloads in case of virtio-net)
need to be reset.
Mikhail
Am Fr., 11. Okt. 2019 um 12:30 Uhr schrieb Mikhail Sennikovsky
n);
in the new post load hook (virtio_net_post_load_virtio) exactly like you say?
Mikhail
>
>
> On Fri, Oct 11, 2019 at 11:58:38AM +0200, Mikhail Sennikovsky wrote:
> > Note that the virtio_net_set_features gets also called from the
> > virtio_pci_common_write when guest does
Oct 11, 2019 at 11:46:22AM +0200, Mikhail Sennikovsky wrote:
> > Hi Michael,
> >
> > Unfortunately your approach will not work, because the
> > VirtIONet::curr_guest_offloads would still be reset in
> > virtio_net_set_features:
> > --
> > if (n
69eee20) at
> migration/savevm.c:2395
> #10 qemu_loadvm_state (f=0x569dc010) at migration/savevm.c:2467
> #11 process_incoming_migration_co (opaque=0x0) at migration/migration.c:449
>
> Fix this by pushing out offload initialization to the new post load hook.
>
> Repo
igned-off-by: Mikhail Sennikovsky
---
hw/net/virtio-net.c | 9 -
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/hw/net/virtio-net.c b/hw/net/virtio-net.c
index b9e1cd7..49be172 100644
--- a/hw/net/virtio-net.c
+++ b/hw/net/virtio-net.c
@@ -767,7 +767,14 @@
in Stefan, Jason and Michael who know the virtio details
> >
> > Dave
> >
> > * Mikhail Sennikovsky (mikhail.sennikovs...@cloud.ionos.com) wrote:
> >> Currently offloads disabled by guest via the
> >> VIRTIO_NET_CTRL_GUEST_OFFLOADS_SET
> >> command a
Hi all,
This is the third version of the patch, implementing the approach
based on checking the VM runstate proposed by Jason Wang.
Regards,
Mikhail
Mikhail Sennikovsky (1):
virtio-net: prevent offloads reset on migration
hw/net/virtio-net.c | 9 -
1 file changed, 8 insertions(+), 1
t;
> Dave
>
> * Mikhail Sennikovsky (mikhail.sennikovs...@cloud.ionos.com) wrote:
> > Currently offloads disabled by guest via the
> > VIRTIO_NET_CTRL_GUEST_OFFLOADS_SET
> > command are not preserved on VM migration.
> > Instead all offloads reported by guest features (v
d setting it to false
for the migration case.
Signed-off-by: Mikhail Sennikovsky
---
hw/display/virtio-gpu-base.c | 3 ++-
hw/net/virtio-net.c | 5 +++--
hw/virtio/virtio.c | 11 ++-
include/hw/virtio/virtio.h | 2 +-
4 files changed, 12 insertions(+), 9 deleti
Hi all,
This is the second version of the patch with the coding style problem fixed.
Regards,
Mikhail
Mikhail Sennikovsky (1):
virtio-net: prevent offloads reset on migration
hw/display/virtio-gpu-base.c | 3 ++-
hw/net/virtio-net.c | 5 +++--
hw/virtio/virtio.c | 11
to illustrate the problem and potentially initiate a further
discussion on the best way of fixing it.
Thanks & Regards,
Mikhail
Mikhail Sennikovsky (1):
virtio-net: prevent offloads reset on migration
hw/display/virtio-gpu-base.c | 3 ++-
hw/net/virtio-net.c | 5 +++--
d setting it to false
for the migration case.
Signed-off-by: Mikhail Sennikovsky
---
hw/display/virtio-gpu-base.c | 3 ++-
hw/net/virtio-net.c | 5 +++--
hw/virtio/virtio.c | 10 +-
include/hw/virtio/virtio.h | 2 +-
4 files changed, 11 insertions(+), 9 deleti
15 matches
Mail list logo