search for: get_features_rev

...its depending on which > revision they're operating at. Let's give the transport a way to > re-query the device about its features when the revision has been > changed. > > Signed-off-by: Cornelia Huck <cornelia.huck at de.ibm.com> So now we have both get_features and get_features_rev, and it's never clear which revision does host_features refer to. IMHO that's just too messy. Let's add get_legacy_features and host_legacy_features instead? > --- > hw/s390x/virtio-ccw.c | 12 ++++++++++-- > hw/virtio/virtio-bus.c | 14 ++++++++++++-- &g...

...its depending on which > revision they're operating at. Let's give the transport a way to > re-query the device about its features when the revision has been > changed. > > Signed-off-by: Cornelia Huck <cornelia.huck at de.ibm.com> So now we have both get_features and get_features_rev, and it's never clear which revision does host_features refer to. IMHO that's just too messy. Let's add get_legacy_features and host_legacy_features instead? > --- > hw/s390x/virtio-ccw.c | 12 ++++++++++-- > hw/virtio/virtio-bus.c | 14 ++++++++++++-- &g...

...uint64_t requested_features, + unsigned int revision) { VirtIODevice *vdev = virtio_bus_get_device(bus); VirtioDeviceClass *k; assert(vdev != NULL); k = VIRTIO_DEVICE_GET_CLASS(vdev); + if (revision > 0 && k->get_features_rev) { + return k->get_features_rev(vdev, requested_features, revision); + } assert(k->get_features != NULL); return k->get_features(vdev, requested_features); } +uint64_t virtio_bus_get_vdev_features(VirtioBusState *bus, + uint64_t r...

...uint64_t requested_features, + unsigned int revision) { VirtIODevice *vdev = virtio_bus_get_device(bus); VirtioDeviceClass *k; assert(vdev != NULL); k = VIRTIO_DEVICE_GET_CLASS(vdev); + if (revision > 0 && k->get_features_rev) { + return k->get_features_rev(vdev, requested_features, revision); + } assert(k->get_features != NULL); return k->get_features(vdev, requested_features); } +uint64_t virtio_bus_get_vdev_features(VirtioBusState *bus, + uint64_t r...

...uint64_t requested_features, + unsigned int revision) { VirtIODevice *vdev = virtio_bus_get_device(bus); VirtioDeviceClass *k; assert(vdev != NULL); k = VIRTIO_DEVICE_GET_CLASS(vdev); + if (revision > 0 && k->get_features_rev) { + return k->get_features_rev(vdev, requested_features, revision); + } assert(k->get_features != NULL); return k->get_features(vdev, requested_features); } +uint64_t virtio_bus_get_vdev_features(VirtioBusState *bus, + uint64_t r...

...uint64_t requested_features, + unsigned int revision) { VirtIODevice *vdev = virtio_bus_get_device(bus); VirtioDeviceClass *k; assert(vdev != NULL); k = VIRTIO_DEVICE_GET_CLASS(vdev); + if (revision > 0 && k->get_features_rev) { + return k->get_features_rev(vdev, requested_features, revision); + } assert(k->get_features != NULL); return k->get_features(vdev, requested_features); } +uint64_t virtio_bus_get_vdev_features(VirtioBusState *bus, + uint64_t r...

...t. Let's give the transport a way to > > > re-query the device about its features when the revision has been > > > changed. > > > > > > Signed-off-by: Cornelia Huck <cornelia.huck at de.ibm.com> > > > > So now we have both get_features and get_features_rev, and > > it's never clear which revision does host_features refer to. > > IMHO that's just too messy. > > Let's add get_legacy_features and host_legacy_features instead? > > I wanted to avoid touching anything that does not support version 1. > And this inter...

...t. Let's give the transport a way to > > > re-query the device about its features when the revision has been > > > changed. > > > > > > Signed-off-by: Cornelia Huck <cornelia.huck at de.ibm.com> > > > > So now we have both get_features and get_features_rev, and > > it's never clear which revision does host_features refer to. > > IMHO that's just too messy. > > Let's add get_legacy_features and host_legacy_features instead? > > I wanted to avoid touching anything that does not support version 1. > And this inter...

...> re-query the device about its features when the revision has been > > > > > changed. > > > > > > > > > > Signed-off-by: Cornelia Huck <cornelia.huck at de.ibm.com> > > > > > > > > So now we have both get_features and get_features_rev, and > > > > it's never clear which revision does host_features refer to. > > > > IMHO that's just too messy. > > > > Let's add get_legacy_features and host_legacy_features instead? > > > > > > I wanted to avoid touching anything th...

...> re-query the device about its features when the revision has been > > > > > changed. > > > > > > > > > > Signed-off-by: Cornelia Huck <cornelia.huck at de.ibm.com> > > > > > > > > So now we have both get_features and get_features_rev, and > > > > it's never clear which revision does host_features refer to. > > > > IMHO that's just too messy. > > > > Let's add get_legacy_features and host_legacy_features instead? > > > > > > I wanted to avoid touching anything th...

...diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c index 9cfae66..fdc236a 100644 --- a/hw/block/virtio-blk.c +++ b/hw/block/virtio-blk.c @@ -587,6 +587,24 @@ static uint64_t virtio_blk_get_features(VirtIODevice *vdev, uint64_t features) return features; } +static uint64_t virtio_blk_get_features_rev(VirtIODevice *vdev, + uint64_t features, + unsigned int revision) +{ + if (revision == 0) { + /* legacy */ + virtio_clear_feature(&features, VIRTIO_F_VERSION_1); + return virtio_blk_get...

...vision they're operating at. Let's give the transport a way to > > re-query the device about its features when the revision has been > > changed. > > > > Signed-off-by: Cornelia Huck <cornelia.huck at de.ibm.com> > > So now we have both get_features and get_features_rev, and > it's never clear which revision does host_features refer to. > IMHO that's just too messy. > Let's add get_legacy_features and host_legacy_features instead? I wanted to avoid touching anything that does not support version 1. And this interface might still work for late...

...t a way to > > > > re-query the device about its features when the revision has been > > > > changed. > > > > > > > > Signed-off-by: Cornelia Huck <cornelia.huck at de.ibm.com> > > > > > > So now we have both get_features and get_features_rev, and > > > it's never clear which revision does host_features refer to. > > > IMHO that's just too messy. > > > Let's add get_legacy_features and host_legacy_features instead? > > > > I wanted to avoid touching anything that does not support versi...

...diff --git a/hw/block/virtio-blk.c b/hw/block/virtio-blk.c index 9cfae66..fdc236a 100644 --- a/hw/block/virtio-blk.c +++ b/hw/block/virtio-blk.c @@ -587,6 +587,24 @@ static uint64_t virtio_blk_get_features(VirtIODevice *vdev, uint64_t features) return features; } +static uint64_t virtio_blk_get_features_rev(VirtIODevice *vdev, + uint64_t features, + unsigned int revision) +{ + if (revision == 0) { + /* legacy */ + virtio_clear_feature(&features, VIRTIO_F_VERSION_1); + return virtio_blk_get...

...vision they're operating at. Let's give the transport a way to > > re-query the device about its features when the revision has been > > changed. > > > > Signed-off-by: Cornelia Huck <cornelia.huck at de.ibm.com> > > So now we have both get_features and get_features_rev, and > it's never clear which revision does host_features refer to. > IMHO that's just too messy. > Let's add get_legacy_features and host_legacy_features instead? I wanted to avoid touching anything that does not support version 1. And this interface might still work for late...

...t a way to > > > > re-query the device about its features when the revision has been > > > > changed. > > > > > > > > Signed-off-by: Cornelia Huck <cornelia.huck at de.ibm.com> > > > > > > So now we have both get_features and get_features_rev, and > > > it's never clear which revision does host_features refer to. > > > IMHO that's just too messy. > > > Let's add get_legacy_features and host_legacy_features instead? > > > > I wanted to avoid touching anything that does not support versi...

...block/virtio-blk.c > index 9cfae66..fdc236a 100644 > --- a/hw/block/virtio-blk.c > +++ b/hw/block/virtio-blk.c > @@ -587,6 +587,24 @@ static uint64_t virtio_blk_get_features(VirtIODevice *vdev, uint64_t features) > return features; > } > > +static uint64_t virtio_blk_get_features_rev(VirtIODevice *vdev, > + uint64_t features, > + unsigned int revision) > +{ > + if (revision == 0) { > + /* legacy */ > + virtio_clear_feature(&features, VIRTIO_F_VERSION_1);...

...block/virtio-blk.c > index 9cfae66..fdc236a 100644 > --- a/hw/block/virtio-blk.c > +++ b/hw/block/virtio-blk.c > @@ -587,6 +587,24 @@ static uint64_t virtio_blk_get_features(VirtIODevice *vdev, uint64_t features) > return features; > } > > +static uint64_t virtio_blk_get_features_rev(VirtIODevice *vdev, > + uint64_t features, > + unsigned int revision) > +{ > + if (revision == 0) { > + /* legacy */ > + virtio_clear_feature(&features, VIRTIO_F_VERSION_1);...

And yet another iteration of virtio-1 support in qemu, tested with the latest virtio kernel patches. Find it at git://github.com/cohuck/qemu virtio-1 Changes from v5: - fixed stupid bug in "virtio: support more feature bits": we need to define a proper prop backend for 64 bit wide handling... - don't negotiate revision 1 unless VERSION_1 is offered - use 64 bit wide features

And yet another iteration of virtio-1 support in qemu, tested with the latest virtio kernel patches. Find it at git://github.com/cohuck/qemu virtio-1 Changes from v5: - fixed stupid bug in "virtio: support more feature bits": we need to define a proper prop backend for 64 bit wide handling... - don't negotiate revision 1 unless VERSION_1 is offered - use 64 bit wide features