c4162f6708
Use -M virt as default machine for arm/aarch64
...
Centralize the default machine logic in caps and use it consistently
2014-09-23 16:05:48 -04:00
0182ba3ead
guest: Use virtio for arm -M virt
...
In fact it's the only thing it supports
2014-09-23 15:12:02 -04:00
81fdc1e5e5
guest: Clarify a function name
2014-09-23 15:09:35 -04:00
92db9bd4dd
tests: Add some more arm tests
...
Many are broken for now, but they will help validate we are setting
the correct defaults with subsequent patches.
2014-09-23 15:08:47 -04:00
c216302743
tests: capabilities: Add aarch64 test data
2014-09-23 14:56:44 -04:00
86417d42ca
osxml: Ensure kernel/initrd/dtb are absolute paths
2014-09-23 14:32:01 -04:00
4c4bdbb8b8
cli: Allow embedding comma in kernel_args
...
And fix can_comma for aliases
2014-09-23 14:14:21 -04:00
30db9ece22
osdict: Add qemu-ga for RHEL6 and RHEL7 guests (bz 1139109)
2014-09-22 17:33:55 -04:00
e1c803dc7c
osdict: Only list latest point release in virt-manager
...
So only list RHEL6.5, and not 6.4-6.0 as well. That earlier versions
can be seen by selecting 'show all options'
2014-09-22 17:20:07 -04:00
3e5b61da86
domain: relax the UNDEFINE_NVRAM condition precisely to libvirtd's one
...
(1) A separate nvram (ie. variable store) file for the domain exists if
and only if the following condition holds:
(self.get_xmlobj().os.loader_ro is True and
self.get_xmlobj().os.loader_type == "pflash")
(Refer to libvirtd's qemuPrepareNVRAM() function, as of commit 742b08e30.)
(2) The
self.get_xmlobj().os.nvram
condition is sufficient, but not necessary, for the separate varstore file
to exist. That is, if the condition holds, then the separate varstore file
exists for sure, but if the condition doesn't hold, the file may exist
nonetheless. (Because libvirtd can auto-generate the varstore file's
pathname.)
This means that requiring condition (2) for setting UNDEFINE_NVRAM on
domain deletion will miss a subset of the cases, ie. when the necessary
and sufficient condition (1) holds, but the sufficient-only condition (2)
doesn't.
Make sure that the code uses the necessary and sufficient condition (1).
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
2014-09-22 12:32:31 -04:00
b3fa8bb650
tests: Ensure osdict.py aliases is never extended
2014-09-21 15:26:03 -04:00
662922dc95
createinterface: Use inactive XML for embeding child XML
...
The inactive XML will drop things like complex protocol output and
link state info, which netcf can choke on for sub interfaces.
2014-09-21 15:04:16 -04:00
bd0bcba0b5
createinterface: Fix 'copy child protocol' behavior (bug 1122743)
...
We should only copy the inactive protocol XML, which lists things
like dhcp, vs the running XML, which is the same format as static
addressing.
2014-09-21 15:02:27 -04:00
6a9aa6827b
console: Break out tunnel handling to its own file
2014-09-21 13:38:53 -04:00
e1b646594c
console: Hide display widget from consumers
...
There were several layering violations, make it explicit by making
display private, and providing APIs for everything the consumers
need.
2014-09-21 13:25:43 -04:00
a3bd1b26a9
console: Clarify keyboard grab default handling
2014-09-21 13:06:25 -04:00
4568cdb636
module_trace: Only show main thread APIs by default
2014-09-21 12:05:02 -04:00
f4563906ca
snapshots: Use consistent button layout with host dialog
...
Except 'refresh' is in place of 'stop'
2014-09-21 11:44:01 -04:00
df50fdc323
testdriver: Add a host-model and host-passthrough usage
2014-09-21 10:17:51 -04:00
b0d5198b7a
virt-manager: Switch 'Copy host' to use host-model (bz 1133155)
...
Similar to the previous change for virt-install
2014-09-20 20:44:14 -04:00
eda5eea549
virt-install: Switch --cpu host to use host-model, not copy (bz 1133155)
...
host-model offers us other protections now, like ensuring we don't try
to use a host cpu feature that qemu doesn't support.
http://bugzilla.redhat.com/show_bug.cgi?id=1133155
2014-09-20 20:44:13 -04:00
64106230bd
virt-install: Add hidden option --test-media-detection
...
Can help with debugging and test cases
2014-09-20 20:32:19 -04:00
3933ff101b
guest: Limit number of default usb redirdevs to 2 (bug 1135488)
...
If we use 4 devices here, we fill up all the emulated USB2 slots,
and directly assigned devices are forced to fall back to USB1
https://bugzilla.redhat.com/show_bug.cgi?id=1135488
2014-09-20 20:16:31 -04:00
99700821c6
virt-install: Reorder --help output a bit
2014-09-20 19:30:16 -04:00
d7e1a40352
virt-install: Deprecate --nodisks in favor of --disk none
2014-09-20 19:16:13 -04:00
57ae284ad4
virt-install: Deprecate --nonetworks in favor of --network none
...
Follows the pattern used by the other cli options
2014-09-20 19:02:38 -04:00
c26512926d
virt-install: Default to --graphics even if DISPLAY is unset
...
This was a weird bit of logic that tries to be smart, but there isn't
a clear corrolation between DISPLAY being unset and whether the user
wants graphics or not.
Make the default consistent, some people may need to adjust to
--nographics
2014-09-20 18:52:20 -04:00
e3de3d8b69
virt-convert: Fix pointing to relative config file
2014-09-20 18:29:20 -04:00
1f3f3fa4fc
virt-install: If virt-viewer will fail, default to --wait -1
...
This can happen if user requests explicit --graphics, but they are on
a headless display. We had a few reports of this recently.
2014-09-20 18:28:43 -04:00
85161e4c57
host: Clean up invocation of *apply from confirm dialog
...
Basically, only apply the page we are actually on
2014-09-20 17:32:10 -04:00
0717a21f65
host: Fix renaming network/storage pool
2014-09-20 17:32:09 -04:00
38738d9fa6
connection: Don't close VM dialog on rename
...
Rework things a bit to simplify everything we pass around.
The specific bug fix is making sure we update the object list in place,
otherwise the event loop detects it as the VM being deleted and closes
the details window.
2014-09-20 17:32:09 -04:00
ea97325a72
connection: Tweak some debugging output
2014-09-20 16:33:26 -04:00
f2ff378a9d
details: Avoid some misleading logging on rename
2014-09-20 16:18:10 -04:00
1bfdf0575f
test_inject: Drop old bits, add f21 and rhel7, reorg a bit
2014-09-20 15:58:52 -04:00
6a8194ebc7
docs: Document virtio-scsi better, and add explicit test cases for it
2014-09-20 15:03:19 -04:00
56c091600a
details: Move hostdev prettyifier to virtinst
2014-09-20 14:56:57 -04:00
5130d959e5
cli: Advertise the --host-device option as --hostdev
...
It annoys me that all the other CLI options map to the libvirt XML name,
except this one. Of course, keep the old option around for back compat,
just give precendence to the new option.
2014-09-20 14:29:34 -04:00
5bbd46a070
cli: Fix suddenly appearing pylint warning :/
2014-09-20 14:22:29 -04:00
c88d085d03
virt-xml: Make --remove-device --host-device <str> work
...
Need to add some infrastructure for arbitrary lookup functions
2014-09-20 14:22:29 -04:00
35c86a775a
nodedev: Simplify parameter comparison
2014-09-20 14:22:03 -04:00
cf80b4e013
cli: Show failing option via virt-xml --remove lookup
2014-09-20 13:53:27 -04:00
92f298e393
addhardware: Use virtmanager APIs for VM listing
2014-09-20 13:04:47 -04:00
a3c6637989
nodedev: Share comparison logic for cli string lookup
2014-09-20 12:58:51 -04:00
dd67c48233
connection: Handle nodedev polling too
...
And cleanup up a bunch of nodedev functions as a result
2014-09-20 12:04:49 -04:00
8cd4958a19
cli: Break out function to disable libvirt error printing
...
We need it early in the test suite
2014-09-20 12:04:37 -04:00
439e5ad0fe
details: Reorg a bunch of the global helpers
...
Some were redundant, some should be in virtinst
2014-09-20 10:23:19 -04:00
4ad50d968a
details: Put tabs on left, easier to navigate with glade
2014-09-20 10:21:27 -04:00
fa76e92f2c
virt-manager: Fix issues with ellipsizing on recent gtk
2014-09-19 22:08:22 -04:00
775ea8dc72
virt-manager: Silence gtk deprecation warnings
...
They are on by default on F21, and they are very noisy
2014-09-19 21:47:13 -04:00
Cole Robinson