From: Xie Yongji <xieyon...@baidu.com>
This patchset is aimed at supporting qemu to reconnect
vhost-user-blk backend after vhost-user-blk backend crash or
restart.
The patch 1,2,3 introduce started flag to VirtioDevice to
fix possible guest hung after migration for virtio 1.0
transitional devices.
The patch 4,5,6 support vhost-user-blk to reconnect backend
when connection closed.
The patch 7 introduces VHOST_USER_PROTOCOL_F_SLAVE_SHMFD
to vhost-user-blk backend which is used to tell qemu that
we support reconnecting now.
To use it, we could start qemu with:
qemu-system-x86_64 \
-chardev socket,id=char0,path=/path/vhost.socket,reconnect=1, \
-device vhost-user-blk-pci,chardev=char0 \
and start vhost-user-blk backend with:
vhost-user-blk -b /path/file -s /path/vhost.socket
Then we can restart vhost-user-blk at any time during VM running.
v8:
- Remove merged vhost-user part
- Introduce started flag to VirtioDevice to fix some bugs
- Rework reconnect logic in vhost_user_blk_device_realize()
v7:
- Introduce a 64-bit counter to struct DescStateSplit/DescStatePacked
to preserve the order of fetching available descriptors
- Add support to resubmit inflight I/O in order in libvhost-user.c
- Rename process_head to last_batch_head in struct DescStateSplit
v6:
- Document the layout in inflight buffer for packed virtqueue
- Rework the layout in inflight buffer for split virtqueue
- Remove version field in VhostUserInflight
- Add a patch to remove some redundant check in
contrib/libvhost-user.c
- Document more details in vhost-user.txt
v5:
- Drop patch that enables "nowait" option on client sockets
- Support resubmitting inflight I/O in order
- Make inflight I/O tracking more robust
- Remove align field and add queue size field in VhostUserInflight
- Document more details in vhost-user.txt
v4:
- Drop messages VHOST_USER_GET_SHM_SIZE and VHOST_USER_SET_SHM_FD
- Introduce two new messages VHOST_USER_GET_INFLIGHT_FD
and VHOST_USER_SET_INFLIGHT_FD
- Allocate inflight buffer in backend rather than in qemu
- Document a recommended format for inflight buffer
v3:
- Using exisiting wait/nowait options to control connection on
client sockets instead of introducing "disconnected" option.
- Support the case that vhost-user backend restart during initialzation
of vhost-user-blk device.
v2:
- Introduce "disconnected" option for chardev instead of reuse "wait"
option
- Support the case that QEMU starts before vhost-user backend
- Drop message VHOST_USER_SET_VRING_INFLIGHT
- Introduce two new messages VHOST_USER_GET_SHM_SIZE
and VHOST_USER_SET_SHM_FD
Xie Yongji (7):
virtio: Introduce started flag to VirtioDevice
virtio: Use started flag in virtio_vmstate_change()
vhost-user-blk: Use started flag in vhost_user_blk_set_status()
vhost-user-blk: Only start vhost-user backend with the first kick
vhost-user-blk: Add return value for vhost_user_blk_start()
vhost-user-blk: Add support to reconnect backend
contrib/vhost-user-blk: enable inflight I/O tracking
contrib/vhost-user-blk/vhost-user-blk.c | 3 +-
hw/block/vhost-user-blk.c | 175 +++++++++++++++++++-----
hw/virtio/virtio.c | 54 +++++++-
include/hw/virtio/vhost-user-blk.h | 3 +
include/hw/virtio/virtio.h | 2 +
5 files changed, 202 insertions(+), 35 deletions(-)
--
2.17.1
