| Message ID | 1526682016-17336-1-git-send-email-changpeng.liu@intel.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Sat, May 19, 2018 at 06:20:16AM +0800, Changpeng Liu wrote: > Read only feature shouldn't be negotiable, because if the > backend device reported Read only feature supported, QEMU > host driver shouldn't change backend's RO attribute. I don't understand this patch. Does it make *all* virtio-blk devices read-only? Or is the idea that the vhost-user slave has to clear the read-only bit if it is a writable device? Stefan
> -----Original Message----- > From: Stefan Hajnoczi [mailto:stefanha@redhat.com] > Sent: Tuesday, May 22, 2018 6:11 AM > To: Liu, Changpeng <changpeng.liu@intel.com> > Cc: qemu-devel@nongnu.org; mst@redhat.com > Subject: Re: [PATCH] vhost-blk: turn on pre-defined RO feature bit > > On Sat, May 19, 2018 at 06:20:16AM +0800, Changpeng Liu wrote: > > Read only feature shouldn't be negotiable, because if the > > backend device reported Read only feature supported, QEMU > > host driver shouldn't change backend's RO attribute. > > I don't understand this patch. > > Does it make *all* virtio-blk devices read-only? If the slave target reported RO feature, the disk in OS should be RO, currently users should specify in QEMU command to enable RO. > > Or is the idea that the vhost-user slave has to clear the read-only bit > if it is a writable device? Exactly. > > Stefan
On Tue, May 22, 2018 at 02:39:28PM +0000, Liu, Changpeng wrote: > > > > -----Original Message----- > > From: Stefan Hajnoczi [mailto:stefanha@redhat.com] > > Sent: Tuesday, May 22, 2018 6:11 AM > > To: Liu, Changpeng <changpeng.liu@intel.com> > > Cc: qemu-devel@nongnu.org; mst@redhat.com > > Subject: Re: [PATCH] vhost-blk: turn on pre-defined RO feature bit > > > > On Sat, May 19, 2018 at 06:20:16AM +0800, Changpeng Liu wrote: > > > Read only feature shouldn't be negotiable, because if the > > > backend device reported Read only feature supported, QEMU > > > host driver shouldn't change backend's RO attribute. > > > > I don't understand this patch. > > > > Does it make *all* virtio-blk devices read-only? > If the slave target reported RO feature, the disk in OS should be RO, currently > users should specify in QEMU command to enable RO. > > > > Or is the idea that the vhost-user slave has to clear the read-only bit > > if it is a writable device? > Exactly. Thanks for explaining! Please implement the read-only bit in contrib/vhost-user-blk we it can be tested. Thanks, Stefan
> -----Original Message----- > From: Stefan Hajnoczi [mailto:stefanha@redhat.com] > Sent: Friday, May 25, 2018 6:20 PM > To: Liu, Changpeng <changpeng.liu@intel.com> > Cc: qemu-devel@nongnu.org; mst@redhat.com > Subject: Re: [PATCH] vhost-blk: turn on pre-defined RO feature bit > > On Tue, May 22, 2018 at 02:39:28PM +0000, Liu, Changpeng wrote: > > > > > > > -----Original Message----- > > > From: Stefan Hajnoczi [mailto:stefanha@redhat.com] > > > Sent: Tuesday, May 22, 2018 6:11 AM > > > To: Liu, Changpeng <changpeng.liu@intel.com> > > > Cc: qemu-devel@nongnu.org; mst@redhat.com > > > Subject: Re: [PATCH] vhost-blk: turn on pre-defined RO feature bit > > > > > > On Sat, May 19, 2018 at 06:20:16AM +0800, Changpeng Liu wrote: > > > > Read only feature shouldn't be negotiable, because if the > > > > backend device reported Read only feature supported, QEMU > > > > host driver shouldn't change backend's RO attribute. > > > > > > I don't understand this patch. > > > > > > Does it make *all* virtio-blk devices read-only? > > If the slave target reported RO feature, the disk in OS should be RO, currently > > users should specify in QEMU command to enable RO. > > > > > > Or is the idea that the vhost-user slave has to clear the read-only bit > > > if it is a writable device? > > Exactly. > > Thanks for explaining! > > Please implement the read-only bit in contrib/vhost-user-blk we it can > be tested. Sounds good, will add the test together. > > Thanks, > Stefan
diff --git a/hw/block/vhost-user-blk.c b/hw/block/vhost-user-blk.c index 262baca..cdc54dd 100644 --- a/hw/block/vhost-user-blk.c +++ b/hw/block/vhost-user-blk.c @@ -204,13 +204,11 @@ static uint64_t vhost_user_blk_get_features(VirtIODevice *vdev, virtio_add_feature(&features, VIRTIO_BLK_F_TOPOLOGY); virtio_add_feature(&features, VIRTIO_BLK_F_BLK_SIZE); virtio_add_feature(&features, VIRTIO_BLK_F_FLUSH); + virtio_add_feature(&features, VIRTIO_BLK_F_RO); if (s->config_wce) { virtio_add_feature(&features, VIRTIO_BLK_F_CONFIG_WCE); } - if (s->config_ro) { - virtio_add_feature(&features, VIRTIO_BLK_F_RO); - } if (s->num_queues > 1) { virtio_add_feature(&features, VIRTIO_BLK_F_MQ); } @@ -322,7 +320,6 @@ static Property vhost_user_blk_properties[] = { DEFINE_PROP_UINT16("num-queues", VHostUserBlk, num_queues, 1), DEFINE_PROP_UINT32("queue-size", VHostUserBlk, queue_size, 128), DEFINE_PROP_BIT("config-wce", VHostUserBlk, config_wce, 0, true), - DEFINE_PROP_BIT("config-ro", VHostUserBlk, config_ro, 0, false), DEFINE_PROP_END_OF_LIST(), }; diff --git a/include/hw/virtio/vhost-user-blk.h b/include/hw/virtio/vhost-user-blk.h index 5804cc9..8c7c79b 100644 --- a/include/hw/virtio/vhost-user-blk.h +++ b/include/hw/virtio/vhost-user-blk.h @@ -34,7 +34,6 @@ typedef struct VHostUserBlk { uint16_t num_queues; uint32_t queue_size; uint32_t config_wce; - uint32_t config_ro; struct vhost_dev dev; } VHostUserBlk;
Read only feature shouldn't be negotiable, because if the backend device reported Read only feature supported, QEMU host driver shouldn't change backend's RO attribute. Signed-off-by: Changpeng Liu <changpeng.liu@intel.com> --- hw/block/vhost-user-blk.c | 5 +---- include/hw/virtio/vhost-user-blk.h | 1 - 2 files changed, 1 insertion(+), 5 deletions(-)