search for: virtio_legacy_is_cross_endian

...are available here: http://lists.linuxfoundation.org/pipermail/virtualization/2015-April/029885.html Please comment. --- C?dric Le Goater (1): vhost_net: re-enable when cross endian Greg Kurz (6): virtio: relax feature check linux-headers: sync vhost.h virtio: introduce virtio_legacy_is_cross_endian() vhost: set vring endianness for legacy virtio tap: add VNET_LE/VNET_BE operations vhost-net: tell tap backend about the vnet endianness hw/net/vhost_net.c | 50 +++++++++++++++++++++++-------------- hw/virtio/vhost.c | 50 +++++++++++++++++++...

...are available here: http://lists.linuxfoundation.org/pipermail/virtualization/2015-April/029885.html Please comment. --- C?dric Le Goater (1): vhost_net: re-enable when cross endian Greg Kurz (6): virtio: relax feature check linux-headers: sync vhost.h virtio: introduce virtio_legacy_is_cross_endian() vhost: set vring endianness for legacy virtio tap: add VNET_LE/VNET_BE operations vhost-net: tell tap backend about the vnet endianness hw/net/vhost_net.c | 50 +++++++++++++++++++++++-------------- hw/virtio/vhost.c | 50 +++++++++++++++++++...

...struct VirtIODevice *vdev, struct vhost_virtqueue *vq, @@ -677,6 +700,16 @@ static int vhost_virtqueue_start(struct vhost_dev *dev, return -errno; } + if (!virtio_has_feature(vdev, VIRTIO_F_VERSION_1) && + virtio_legacy_is_cross_endian(vdev)) { + r = vhost_virtqueue_set_vring_endian_legacy(dev, + virtio_is_big_endian(vdev), + vhost_vq_index); + if (r) { + return -errno; + } + } + s = l = vi...

...the virtio-1 base support (more feature bits etc.) or get a big fat comment and be touched up later. I'd prefer the first solution so it does not get forgotten, but I'm not sure when Michael plans to proceed with the virtio-1 patches (I think they're mostly fine already). > + virtio_legacy_is_cross_endian(vdev)) { > + r = vhost_virtqueue_set_vring_endian_legacy(dev, > + virtio_is_big_endian(vdev), > + vhost_vq_index); > + if (r) { > + return -errno; > +...

...the virtio-1 base support (more feature bits etc.) or get a big fat comment and be touched up later. I'd prefer the first solution so it does not get forgotten, but I'm not sure when Michael plans to proceed with the virtio-1 patches (I think they're mostly fine already). > + virtio_legacy_is_cross_endian(vdev)) { > + r = vhost_virtqueue_set_vring_endian_legacy(dev, > + virtio_is_big_endian(vdev), > + vhost_vq_index); > + if (r) { > + return -errno; > +...

...VER_OK or set it very late, and the fact that we tied work-arounds for even older drivers which dont' set pci bus master to the DRIVER_OK bit. I tried, and I'm close to giving up and just checking guest ack for virtio 1, and ignoring DRIVER_OK requirement if not there. > > + virtio_legacy_is_cross_endian(vdev)) { > > + r = vhost_virtqueue_set_vring_endian_legacy(dev, > > + virtio_is_big_endian(vdev), > > + vhost_vq_index); > > + if (r) { > > +...