Jean Guyader [Thu, 29 Jan 2009 22:47:42 +0000 (22:47 +0000)]
Change the way we do the keyboard switching.
- Creation of a proxy node under the private tree of a domain.
- /local/domain/<domid>/dom0_input/command
The command are:
- take to ask for the keyboard
- release to give the keyboard back to the natif domain.
- Move dom0_input to dom0_input/pos
- Add some a xenstore fonction to chmod the node from qemu.
The dom0_input/command node needs to be writable by the domain.
Jean Guyader [Tue, 9 Dec 2008 01:38:15 +0000 (01:38 +0000)]
- Direct mapping of the emulated framebuffer on the intel if both card option
are compatible.
- Force the intel graphic card to linear mode when we do a fullscreen
switching.
Ross Philipson [Wed, 26 Nov 2008 12:34:41 +0000 (12:34 +0000)]
This patch removes the HP vendor specific WLAN device and subsystem
ID check.
This was a workaround in the passthrough code to allow the WLAN
card to work w/in an HVM that did not have the correct SMBIOS
information. The patch to hvmloader to pass in the correct SMBIOS
strings makes this special HP check/mod uneeded.
Committer: Ross Philipson <ross.philipson@citrix.com>
On branch remove_wlan_hpcheck
Changes to be committed:
(use "git reset HEAD <file>..." to unstage)
Jean Guyader [Fri, 21 Nov 2008 16:00:02 +0000 (16:00 +0000)]
Fix the switch.
You can use the command node to switch the keyboard only or both
(mouse/keyboard).
If you switch the mouse and the keyboard it will also change the focus to
the new vm.
To switch only the keyboard:
write into /local/domain/0/dom0_driver "keyboard slot-id"
To switch mouse and keyboard:
write into /local/domain/0/dom0_driver "switch slot-id"
Each vm owns a slot-id. It's the value of the dom0-input config option.
Jean Guyader [Thu, 9 Oct 2008 16:51:19 +0000 (17:51 +0100)]
- Pass through some value of the real host in case of intel graphic
card. The graphic card need the know how much merroy has been stolen
from the physical RAM.
Ian Jackson [Thu, 23 Oct 2008 09:21:41 +0000 (10:21 +0100)]
Battery management - compilation fixes
* xenstore_refresh_battery_status declaration and definition
was missing void (interpreted in C as K&R syntax)
* a couple of functions returning int returned NULL on failure
Ian Jackson [Thu, 16 Oct 2008 17:31:04 +0000 (18:31 +0100)]
cirrus vga save/restore and lfb_addr/lfb_end
Cirrus VGA save and restore functions cast lfb_addr into an uint64_t
pointer while lfb_addr is only an unsigned long.
Same thing happened to lfb_end, causing pci_dev to be partially
overwritten by mistake.
Ian Jackson [Thu, 16 Oct 2008 09:29:00 +0000 (10:29 +0100)]
enable usb uhci save/restore
enables the usb uhci save and restore functions that
for some reasons were commented out.
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
[ These were commented out in 1e4146791b46f0a798589c58a09f18b2841a9212
aka svn://svn.savannah.nongnu.org/qemu/trunk@3609 c046a42c-6fe2-441c-8c8c-71466251a162
apparently by mistake. This commit reverts 1e414679
-Ian Jackson ]
Ian Jackson [Thu, 16 Oct 2008 09:29:00 +0000 (10:29 +0100)]
enable usb uhci save/restore
enables the usb uhci save and restore functions that
for some reasons were commented out.
Signed-off-by: Stefano Stabellini <stefano.stabellini@eu.citrix.com>
[ These were commented out in 1e4146791b46f0a798589c58a09f18b2841a9212
aka svn://svn.savannah.nongnu.org/qemu/trunk@3609 c046a42c-6fe2-441c-8c8c-71466251a162
apparently by mistake. This commit reverts 1e414679
-Ian Jackson ]
Ian Jackson [Tue, 14 Oct 2008 10:46:53 +0000 (11:46 +0100)]
Fix blktap device backend patch check
Regarding http://bugzilla.xensource.com/bugzilla/show_bug.cgi?id=1367,
it appears that the expected backend path check is too strict for'
blktap devices. Therefore if the devtype is `vbd' we allow the
backend to be `tap'.
Thanks to report and inspiration from Yosuke Iwamatsu.
Signed-off-by: Ian Jackson <ian.jackson@eu.citrix.com>
Jean Guyader [Thu, 9 Oct 2008 16:51:19 +0000 (17:51 +0100)]
- Pass through some value of the real host in case of intel graphic
card. The graphic card need the know how much merroy has been stolen
from the physical RAM.
Ian Jackson [Thu, 9 Oct 2008 13:50:51 +0000 (14:50 +0100)]
Fix open_disk for blktap disks to use real bdrv_new
If blktap drives are registered properly, qemu code is much less likely
to get confused by them. Use bdrv_new(), assign a device name and create
an entry in drives_table for them.
Ian Jackson [Thu, 2 Oct 2008 13:22:41 +0000 (14:22 +0100)]
Avoid trusting client-controlled areas of xenstore.
Various parts of xenstore, specifically /local/GUEST/devices, are
writeable by the guest. Information from these areas must be used
with care, to avoid the guest tricking qemu-dm into improperly using
its privileged access to dom0 resources.
To this end:
* Variables and functions referring to client-controlled data have
had `danger' added to their names.
* There is a new sanitisation/checking arrangement for getting
backend paths and information about the configuration of device
backends, given the frontend information. This is so that when
qemu is providing a data path which is parallel to the PV
frontend/backend arrangements, it only uses the configuration from
a genuine backend which is really configured to serve qemu's own
guest.
* For information which should be read from or written to the `vm'
tree, we obtain the guest's uuid from the hypervisor (the `label'
as the hypercall interface calls it) rather than reading it from
the guest-controlled areas of xenstore.
* The `phantom vbd' feature is disabled. It relies on
guest-controlled xenstore areas indicating device paths on the
guest. We do not believe this feature is currently very relevant.
* We _do_ allow the guest of a stubdom qemu to mess up the
correspondence between pv backends and emulated devices, in the
sense that we don't mind if the guest directs qemu to use a `wrong'
frontend.