qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Qemu-devel] [PATCH 00/17] v4: virtio-serial fixes, new abi for port dis


From: Amit Shah
Subject: [Qemu-devel] [PATCH 00/17] v4: virtio-serial fixes, new abi for port discovery
Date: Wed, 31 Mar 2010 13:03:51 +0530

Hello,

These patches rework the way ports are announced to the guests. A
control message is used to let the guest know a new port is
added. Initial port discovery and port hot-plug work via this way now.

This was done to have the host and guest port numbering in sync to
avoid surprises after several hotplug/unplug operations and
migrations.

The ability to assign a particular port number to ports is also added
so that management software can control the placement of ports.

Differences from v3:
- Removed the QMP events (will ride on top of new infrastructure by
  Luiz)

- Add flow control: ports can signal to the virtio-serial code to stop
  sending data till it's ready to accept more.

- A fix for not abusing the virtqueue_push() api: we should signal to
  the guest the number of bytes written for guest to consume. We were
  using the number to signify the bytes consumed by hosts.

- Juan's comment about re-using control queue buffers

Overall:
- Users can set the port id they wish to instantiate ports at by using
  the ,nr= parameter to 'virtserialport' and 'virtconsole' devices

- Migration fixes: refuse migration when:
  - number of active ports is different between the src and destination
  - max_nr_ports a device can support on the src is more

- If a qemu chardev connection to a port is closed on the dest while
  it was open on the src, inform the guest about this. (Also do the
  same for port closed on src but open on dest.)

- Use control messages for relaying new port information instead of
  config space (changes abi)

- Propagate error message from guest in instantiating a port or a
  device to the user.

- Handle scatter/gather for control output and data output from the
  guest

- Fix abuse of virtio api in the virtqueue_push() function

- Add an API for the ports for flow control: ports can signal when
  they're ready to accept data / stop sending data.

Amit Shah (17):
  virtio-serial: save/load: Ensure target has enough ports
  virtio-serial: save/load: Ensure nr_ports on src and dest are same.
  virtio-serial: save/load: Ensure we have hot-plugged ports
    instantiated
  virtio-serial: save/load: Send target host connection status if
    different
  virtio-serial: Use control messages to notify guest of new ports
  virtio-serial: whitespace: match surrounding code
  virtio-serial: Remove redundant check for 0-sized write request
  virtio-serial: Update copyright year to 2010
  virtio-serial: Propagate errors in initialising ports / devices in
    guest
  virtio-serial: Send out guest data to ports only if port is opened
  iov: Introduce a new file for helpers around iovs, add iov_from_buf()
  iov: Add iov_to_buf and iov_size helpers
  virtio-serial: Handle scatter-gather buffers for control messages
  virtio-serial: Handle scatter/gather input from the guest
  virtio-serial: Apps should consume all data that guest sends out /
    Fix virtio api abuse
  virtio-serial: Discard data that guest sends us when ports aren't
    connected
  virtio-serial: Implement flow control for individual ports

 Makefile.objs          |    1 +
 hw/iov.c               |   70 ++++++++++
 hw/iov.h               |   19 +++
 hw/virtio-balloon.c    |   35 +-----
 hw/virtio-console.c    |   11 +-
 hw/virtio-net.c        |   20 +---
 hw/virtio-serial-bus.c |  326 +++++++++++++++++++++++++++++++++++------------
 hw/virtio-serial.h     |   34 ++++--
 8 files changed, 368 insertions(+), 148 deletions(-)
 create mode 100644 hw/iov.c
 create mode 100644 hw/iov.h





reply via email to

[Prev in Thread] Current Thread [Next in Thread]