Skip to content
Snippets Groups Projects
Select Git revision
  • feature/libtcg-rebase default protected
1 result
Search by author
  • Any Author
  • Anton Anton anjo
1 author
  1. Jun 29, 2017
  3. Jun 23, 2017
  5. Jun 21, 2017
  7. Jun 20, 2017
  9. Jun 14, 2017
  11. Jun 13, 2017
  13. Jun 02, 2017
  15. May 23, 2017
    • Eric Blake's avatar
      shutdown: Add source information to SHUTDOWN and RESET · cf83f140
      Eric Blake authored 
      
Time to wire up all the call sites that request a shutdown or
reset to use the enum added in the previous patch.

It would have been less churn to keep the common case with no
arguments as meaning guest-triggered, and only modified the
host-triggered code paths, via a wrapper function, but then we'd
still have to audit that I didn't miss any host-triggered spots;
changing the signature forces us to double-check that I correctly
categorized all callers.

Since command line options can change whether a guest reset request
causes an actual reset vs. a shutdown, it's easy to also add the
information to reset requests.

Signed-off-by: default avatarEric Blake <eblake@redhat.com>
Acked-by: David Gibson <david@gibson.dropbear.id.au> [ppc parts]
Reviewed-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> [SPARC part]
Reviewed-by: Cornelia Huck <cornelia.huck@de.ibm.com> [s390x parts]
Message-Id: <20170515214114.15442-5-eblake@redhat.com>
Reviewed-by: default avatarMarkus Armbruster <armbru@redhat.com>
Signed-off-by: default avatarMarkus Armbruster <armbru@redhat.com>
      cf83f140
  17. May 19, 2017
  19. May 12, 2017
  21. May 11, 2017
    • Philippe Voinov's avatar
      ui: input-linux: Add absolute event support · d755defd
      Philippe Voinov authored 
      
This patch adds support for absolute pointer events to the input-linux
subsystem. This support was omitted from the original input-linux patch,
however most of the code required for it is already in place.

Support for absolute events is especially useful for guests with vga
passthrough. Since they have a physical monitor, none of normal channels
for sending video output (vnc, etc) are used, meaning they also can't be
used to send absolute input events. This leaves QMP as the only option
to send absolute input into vga passthrough guests, which is not its
intended use and is not efficient.

This patch allows, for example, uinput to be used to create virtual
absolute input devices. This lets you build external systems which share
physical input devices between guests. Without absolute input
capability, such external systems can't seamlessly share pointer devices
between guests.

Signed-off-by: default avatarPhilippe Voinov <philippevoinov@gmail.com>
Message-id: 20170505134231.30210-1-philippevoinov@gmail.com
Signed-off-by: default avatarGerd Hoffmann <kraxel@redhat.com>
      d755defd
    • Philippe Voinov's avatar
      ui: Support non-zero minimum values for absolute input axes · 9cfa7ab9
      Philippe Voinov authored 
      
This patch refactors ui/input.c to support absolute axis
minimum values other than 0. All dependent calls to qemu_input_queue_abs
have been updated to explicitly supply 0 as the axis minimum value.

Signed-off-by: default avatarPhilippe Voinov <philippevoinov@gmail.com>
Message-id: 20170505133952.29885-1-philippevoinov@gmail.com
Signed-off-by: default avatarGerd Hoffmann <kraxel@redhat.com>
      9cfa7ab9
  23. May 09, 2017
  25. May 04, 2017
  27. May 03, 2017
    • Marc-André Lureau's avatar
      input: don't queue delay if paused · 05c6638b
      Marc-André Lureau authored 
      qemu_input_event_send() discards key event when the guest is paused,
but not the delay.

The delay ends up in the input queue, and qemu_input_event_send_key()
will further fill the queue with upcoming events.

VNC uses qemu_input_event_send_key_delay(), not SPICE, which results
in a different input behaviour on pause: VNC will queue the events
(except the first that is discarded), SPICE will discard all events.

Don't queue delay if paused, and provide same behaviour on SPICE and
VNC clients on resume (and potentially avoid over-allocating the
buffer queue)

Fixes:
https://bugzilla.redhat.com/show_bug.cgi?id=1444326



Signed-off-by: default avatarMarc-André Lureau <marcandre.lureau@redhat.com>
Message-id: 20170425130520.31819-1-marcandre.lureau@redhat.com
Signed-off-by: default avatarGerd Hoffmann <kraxel@redhat.com>
      05c6638b
    • Gerd Hoffmann's avatar
      input: limit kbd queue depth · fa18f36a
      Gerd Hoffmann authored 
      
Apply a limit to the number of items we accept into the keyboard queue.

Impact: Without this limit vnc clients can exhaust host memory by
sending keyboard events faster than qemu feeds them to the guest.

Fixes: CVE-2017-8379
Cc: P J P <ppandit@redhat.com>
Cc: Huawei PSIRT <PSIRT@huawei.com>
Reported-by: default avatar <jiangxin1@huawei.com>
Signed-off-by: default avatarGerd Hoffmann <kraxel@redhat.com>
Message-id: 20170428084237.23960-1-kraxel@redhat.com
      fa18f36a
  29. Apr 25, 2017
  31. Apr 24, 2017
  33. Apr 03, 2017