On Fri, Oct 11, 2019 at 12:34:59PM +0200, Mikhail Sennikovsky wrote:
> 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.
>
>
On Fri, Oct 11, 2019 at 12:30:07PM +0200, Mikhail Sennikovsky wrote:
> Am Fr., 11. Okt. 2019 um 12:08 Uhr schrieb Michael S. Tsirkin
> :
> >...
> > And pre save hook can do n->saved_guest_offloads = n->curr_guest_offloads.
> Would you want to have the saved_guest_offloads as part of the saved stat
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
:
>
> Am
Am Fr., 11. Okt. 2019 um 12:08 Uhr schrieb Michael S. Tsirkin :
>...
> And pre save hook can do n->saved_guest_offloads = n->curr_guest_offloads.
Would you want to have the saved_guest_offloads as part of the saved state?
The curr_guest_offloads info is already there, so why would you want
to dupli
Ugh. Okay.
So let's add a new field saved_guest_offloads and have
vmstate use that instead of curr_guest_offloads.
Now the new post load hook can do
n->curr_guest_offloads = n->saved_guest_offloads
and apply that.
And pre save hook can do n->saved_guest_offloads = n->curr_guest_offloads.
Note that the virtio_net_set_features gets also called from the
virtio_pci_common_write when guest does virtio device configuration.
In that case the curr_guest_offloads are still expected to be reset.
Mikhail
Am Fr., 11. Okt. 2019 um 11:51 Uhr schrieb Michael S. Tsirkin :
>
> On Fri, Oct 11, 201
On Fri, 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->has_vnet_hdr) {
> n->curr_guest_offloads =
> v
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->has_vnet_hdr) {
n->curr_guest_offloads =
virtio_net_guest_offloads_by_features(features);
--
( https://github.com/qemu/qemu/blob/
On 2019/10/11 上午2:04, Michael S. Tsirkin 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_PCI_GUEST_FEATURES)
get enabled.
What happens is: first the Virt
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_PCI_GUEST_FEATURES)
get enabled.
What happens is: first the VirtIONet::curr_guest_offloads gets restored
and offload
10 matches
Mail list logo