Commit Graph

533 Commits

Author SHA1 Message Date
Michal Privoznik
68ade25372 qemu: Enable firmware autoselection
https://bugzilla.redhat.com/show_bug.cgi?id=1564270

Now that everything is prepared for qemu driver we can enable
parser feature to allow users define such domains.

At the same time, introduce bunch of tests to test the feature.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
2019-03-12 16:09:55 +01:00
Andrea Bolognani
2c0b8d6a08 tests: Drop unnecessary variables
In qemuxml2xmltest, both activeVcpus and blockjobs members
of the testInfo struct have been entirely unused ever since
commit d1a7fc8bb3.

Signed-off-by: Andrea Bolognani <abologna@redhat.com>
2019-03-12 14:09:49 +01:00
Andrea Bolognani
bb5c6e7a96 tests: Use WHEN_BOTH in qemuxml2xml whenever possible
There are a few cases where we are using either WHEN_ACTIVE
or WHEN_INACTIVE even though WHEN_BOTH would work perfectly
fine: for those, start using the simpler DO_TEST() macro.

Signed-off-by: Andrea Bolognani <abologna@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2019-03-05 10:41:08 +01:00
Cole Robinson
448a094717 qemu: Support scsi controller model=virtio-{non-}transitional
Add <controller type='scsi' model handling for virtio transitional
devices. Ex:

  <controller type='scsi' model='virtio-transitional'/>

* "virtio-transitional" maps to qemu "virtio-scsi-pci-transitional"
* "virtio-non-transitional" maps to qemu "virtio-scsi-non-transitional"

The naming here doesn't match the pre-existing model=virtio-scsi.
The prescence of '-scsi' there seems kind of redundant as we have
type='scsi' already, so I decided to follow the pattern of other
patches and use virtio-transitional etc.

Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Cole Robinson <crobinso@redhat.com>
2019-03-04 11:08:41 -05:00
Cole Robinson
2593a1bd1a conf: Add <input model='virtio-{non-}transitional'/>
<input> devices lack the model= attribute which is used by
most other device types. To eventually support
virtio-input-host-pci-{non-}traditional in qemu, let's add
a standard model= attribute. This just adds the domain_conf
wiring

Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Cole Robinson <crobinso@redhat.com>
2019-03-04 11:08:41 -05:00
Cole Robinson
6e64899284 qemu: Support vsock model=virtio-{non-}transitional
Add <vsock> model handling for virtio transitional devices. Ex:

  <vsock model='virtio-transitional'>
    ...
  </vsock>

* "virtio-transitional" maps to qemu "vhost-vsock-pci-transitional"
* "virtio-non-transitional" maps to qemu "vhost-vsock-pci-non-transitional"

Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Cole Robinson <crobinso@redhat.com>
2019-03-04 11:08:41 -05:00
Cole Robinson
e063707556 qemu: Support rng model=virtio-{non-}transitional
Add new <rng> model values for virtio transitional devices. Ex:

  <rng model='virtio-transitional'>
    ...
  </rng>

* "virtio-transitional" maps to qemu "virtio-rng-pci-transitional"
* "virtio-non-transitional" maps to qemu "virtio-rng-pci-non-transitional"

Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Cole Robinson <crobinso@redhat.com>
2019-03-04 11:08:41 -05:00
Cole Robinson
25d05051b3 conf: Add <disk model='virtio-{non-}transitional'/>
<disk> devices lack the model= attribute which is used by
most other device types. bus= mostly acts as one, but it
serves other purposes too like determing what target=
prefix to use, and for matching against controller type=
values.

Extending bus= to handle additional virtio transitional
devices will complicate apps lives, and it isn't a clean
mapping anyways. So let's bite the bullet and add a new
<disk model=X/> attribute, and wire up common handling
for virtio and virtio-{non-}transitional

Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Cole Robinson <crobinso@redhat.com>
2019-03-04 11:08:40 -05:00
Peter Krempa
ea13c12940 tests: qemuxml2argv: Add CDROM disks for all untested buses
Add full and empty cdroms on 'usb' and 'sd' bus to have test
coverage. Note that this does not guarantee that qemu will accept them.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
2019-02-08 08:48:11 +01:00
Peter Krempa
4bf49bc566 qemu: caps: Always assume presence of 'scsi-hd' and 'scsi-cd' device
The split of scsi-disk into the two separate devices was introduced by
qemu commit b443ae67 released in qemu v0.15.

All changes to test files are not really related to disk testing thanks
to previous refactors.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
2019-02-08 08:34:17 +01:00
Peter Krempa
61900c4cda tests: qemu: Merge 'disk-scsi-vscsi' test into 'disk-scsi'
As we support multiple scsi controllers there's no need to have a
special test for this controller.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
2019-02-08 08:32:40 +01:00
Peter Krempa
ee39011a09 tests: qemu: Merge 'disk-scsi-mptsas1068' test into 'disk-scsi'
As we support multiple scsi controllers there's no need to have a
special test for this controller.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
2019-02-08 08:32:40 +01:00
Peter Krempa
a11e333779 tests: qemu: Merge 'disk-scsi-megasas' test into 'disk-scsi'
As we support multiple scsi controllers there's no need to have a
special test for this controller.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
2019-02-08 08:32:40 +01:00
Peter Krempa
566ada91d5 tests: qemu: Rename 'disk-scsi-device' to 'disk-scsi'
Drop the 'device' suffix which is quite pointless.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
2019-02-08 08:32:40 +01:00
Peter Krempa
3eff4bc2c5 tests: qemuxml: Merge 'ioeventfd' variant of 'virtio-scsi' test into the common file
We don't need separate files for this test. Also modernize it in the
process.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
2019-02-08 08:31:15 +01:00
Peter Krempa
7d1fc9f8d6 tests: qemuxml: Merge 'max_sectors' variant of 'virtio-scsi' test into the common file
We don't need separate files for this test. Also modernize it in the
process.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
2019-02-08 08:31:15 +01:00
Peter Krempa
74a7d6fb2b tests: qemuxml: Merge 'cmd_per_lun' variant of 'virtio-scsi' test into the common file
We don't need separate files for this test. Also modernize it in the
process.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
2019-02-08 08:31:15 +01:00
Peter Krempa
ab5c7dbee3 tests: qemuxml: Merge 'num-queues' variant of 'virtio-scsi' test into the common file
We don't need separate files for this test. Also modernize it in the
process.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
2019-02-08 08:31:15 +01:00
Peter Krempa
6d8be31f2a tests: qemu: Remove 'disk-scsi-virtio-scsi' test
Now that we have a specific test for testing the 'virtio-scsi'
controller and other tests which test a combination of scsi and non-scsi
devices this test no longer makes sense.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
2019-02-08 08:31:15 +01:00
Peter Krempa
7a22fa7dce tests: qemuxml: Add a common test file for the 'virtio-scsi' controller
Add a file to aggregate testing for 'virtio-scsi' based on the modern
framework.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
2019-02-08 08:31:02 +01:00
Andrea Bolognani
c5f92bf558 tests: Add test for PCI usage on RISC-V
This shows users can now use PCI for RISC-V guests, as long
as they opt into it by manually assigning addresses.

Signed-off-by: Andrea Bolognani <abologna@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2019-02-01 11:57:52 +01:00
Cole Robinson
0ba9786d8a qemu: Move <rng> validation out of qemu_command.c
Move the rng->model == VIRTIO check to parse time. This also
allows us to remove similar checks throughout the qemu driver

Reviewed-by: Andrea Bolognani <abologna@redhat.com>
Signed-off-by: Cole Robinson <crobinso@redhat.com>
2019-01-24 18:59:38 -05:00
Luyao Zhong
db521e7d03 conf: Introduce 'readonly' element into xml for NVDIMM memory
The 'readonly' option allows users to mark vNVDIMM read-only:

<devices>
  ...
  <memory model='nvdimm' access='shared'>
      <source>
          <path>/dev/dax0.0</path>
      </source>
      <target>
          <size unit='MiB'>4094</size>
          <node>0</node>
          <label>
              <size unit='MiB'>2</size>
          </label>
          <readonly/>
      </target>
  </memory>
  ...
</devices>

Signed-off-by: Luyao Zhong <luyao.zhong@intel.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
2019-01-02 09:00:34 -05:00
Luyao Zhong
73fc8c491e conf: Introduce 'pmem' element into xml for NVDIMM memory
The 'pmem' option allows users to specify whether the backend
storage of memory-backend-file is a real persistent memory:

<devices>
  ...
  <memory model='nvdimm' access='shared'>
      <source>
          <path>/dev/dax0.0</path>
          <pmem/>
      </source>
      <target>
          <size unit='MiB'>4094</size>
          <node>0</node>
          <label>
              <size unit='MiB'>2</size>
          </label>
      </target>
  </memory>
  ...
</devices>

Signed-off-by: Luyao Zhong <luyao.zhong@intel.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
2019-01-02 09:00:29 -05:00
Luyao Zhong
80d1ed9773 conf: Introduce 'alignsize' element into xml for NVDIMM memory
NVDIMM emulation will mmap the backend file, it uses host pagesize
as the alignment of mapping address before, but some backends may
require alignments different from the pagesize. So the 'alignsize'
option is introduced to allow specification of the proper alignment:

<devices>
  ...
  <memory model='nvdimm' access='shared'>
      <source>
          <path>/dev/dax0.0</path>
          <alignsize unit='MiB'>2</alignsize>
      </source>
      <target>
          <size unit='MiB'>4094</size>
          <node>0</node>
          <label>
              <size unit='MiB'>2</size>
          </label>
      </target>
  </memory>
  ...
</devices>

Signed-off-by: Luyao Zhong <luyao.zhong@intel.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
2019-01-02 09:00:24 -05:00
Erik Skultety
dd45c2710f conf: domain: gfx: Iterate over graphics devices when doing validation
The QEMU validation code for graphics has been in place for a while, but
because it is only executed from virDomainDeviceInfoIterateInternal, it
was never run, since the iterator expects the device to have boot info
which graphics don't have. The unfortunate side effect of this whole mess
was that a few capabilities were missing from the test suite (as commit
d8266ebe1 demonstrated with graphics-spice-invalid-egl-headless test),
which in turn meant that a few graphics tests which expected a failure
happily accepted any failure the test runtime returned which made them
succeed. The impact of this was that we then allowed to start a domain
with multiple OpenGL-enabled graphics devices.

This patch enables iteration over graphics devices. Unsurprisingly,
a few tests started to fail as a result, so fix those too.

Signed-off-by: Erik Skultety <eskultet@redhat.com>
Reviewed-by: John Ferlan <jferlan@redhat.com>
2018-12-12 15:12:35 +01:00
Peter Krempa
0ba4da5871 tests: qemuxml2xml: Add few debug statements for status XML testing
Add markers for allowing test debugging if one of the steps fails
without setting a proper error.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2018-12-05 17:53:26 +01:00
Erik Skultety
5f931fe391 conf: gfx: egl-headless: Introduce a new <gl> subelement
Unlike with SPICE and SDL which use the <gl> subelement to enable OpenGL
acceleration, specifying egl-headless graphics in the XML has
essentially the same meaning, thus in case of egl-headless we don't have
a need for the 'enable' element attribute and we'll only be interested
in the 'rendernode' one further down the road.

Signed-off-by: Erik Skultety <eskultet@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2018-12-03 14:56:08 +01:00
Andrea Bolognani
0029eace52 qemu: Introduce QEMU_CAPS_MACHINE_PSERIES_CAP_NESTED_HV
Signed-off-by: Andrea Bolognani <abologna@redhat.com>
2018-11-27 17:12:18 +01:00
Yi Min Zhao
9d6be3ff79 qemu: Generate and use zPCI device in QEMU command line
Add new functions to generate zPCI command string and append it to
QEMU command line. And the related tests are added.

Signed-off-by: Yi Min Zhao <zyimin@linux.ibm.com>
Reviewed-by: Boris Fiuczynski <fiuczy@linux.ibm.com>
Reviewed-by: Stefan Zimmermann <stzi@linux.ibm.com>
Reviewed-by: Bjoern Walk <bwalk@linux.ibm.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
2018-11-15 12:32:18 +01:00
Yi Min Zhao
b4833b2c2f conf: Introduce parser, formatter for uid and fid
This patch introduces new XML parser/formatter functions. Uid is
16-bit and non-zero. Fid is 32-bit. They are the two attributes of zpci
which is introduced as PCI address element. Zpci element is parsed and
formatted along with PCI address. And add the related test cases.

Signed-off-by: Yi Min Zhao <zyimin@linux.ibm.com>
Reviewed-by: Boris Fiuczynski <fiuczy@linux.ibm.com>
Reviewed-by: Stefan Zimmermann <stzi@linux.ibm.com>
Reviewed-by: Bjoern Walk <bwalk@linux.ibm.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
2018-11-15 12:32:18 +01:00
Andrea Bolognani
5095394e1e qemu: Drop QEMU_CAPS_DEVICE_SCSI_GENERIC
It was already available in 1.5.0, so we can assume it's
present and avoid checking for it at runtime.

This commit is best viewed with 'git show -w'.

Signed-off-by: Andrea Bolognani <abologna@redhat.com>
2018-09-21 16:50:46 +02:00
Erik Skultety
5165ff0971 src: More cleanup of some system headers already contained in internal.h
All of the ones being removed are pulled in by internal.h. The only
exception is sanlock which expects the application to include <stdint.h>
before sanlock's headers, because sanlock prototypes use fixed width
int, but they don't include stdint.h themselves, so we have to leave
that one in place.

Signed-off-by: Erik Skultety <eskultet@redhat.com>
Acked-by: Michal Privoznik <mprivozn@redhat.com>
2018-09-20 10:16:39 +02:00
Erik Skultety
9403b63102 internal: Move <stdio.h> include to internal.h
It doesn't really make sense for us to have stdlib.h and string.h but
not stdio.h in the internal.h header.

Signed-off-by: Erik Skultety <eskultet@redhat.com>
Acked-by: Michal Privoznik <mprivozn@redhat.com>
2018-09-20 10:16:38 +02:00
Laine Stump
7ea7342996 conf: correct false boot order error during domain parse
virDomainDefCollectBootOrder() is called for every item on the list
for each type of device. One of the checks it makes is to gather the
order attributes from the <boot> element of all devices, and assure
that no two devices have been given the same order.

Since (internally to libvirt, *not* in the domain XML) an <interface
type='hostdev'> is on both the list of hostdev devices and the list of
network devices, it will be counted twice, and the code that checks
for multiple devices with the same boot order will give a false
positive.

To remedy this, we make sure to return early for hostdev devices that
have a parent.type != NONE.

This was introduced in commit 5b75a4, which was first in libvirt-4.4.0.

Resolves: https://bugzilla.redhat.com/1601318

Signed-off-by: Laine Stump <laine@laine.org>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2018-09-12 11:09:53 -04:00
Andrea Bolognani
7dd9d375ea tests: Add missing virtio-input capabilities
A few qemuxml2xml tests for virtio-input devices are
missing the capabilities used for the corresponding
qemuxml2argv tests: this wasn't a problem until now
because capabilities were only checked at command line
generation time, but we're going to change that later.

Signed-off-by: Andrea Bolognani <abologna@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2018-09-06 16:29:42 +02:00
Andrea Bolognani
0770587d28 tests: Fix use of virtio-serial for aarch64/virt
virtio-serial is an alias for virtio-serial-pci, which
should not have been used for a PCIe-less aarch64/virt
guest but it ended up being used anyway because the
virtio-mmio capability was missing and the algorithm
is buggy.

Fix the test case so that we can fix the algorithm next.

Signed-off-by: Andrea Bolognani <abologna@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2018-09-05 15:54:41 +02:00
Lubomir Rintel
097ff04948 tests: Add RISC-V guest
Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
2018-08-24 13:23:36 +02:00
Peter Krempa
c3543a3217 tests: qemu: Add test data for backing chains and indexes
Add test data for nested backing chains with/without indexes (used in
status XMLs) which will excercise blockdev and the related work.

Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
2018-08-21 15:46:06 +02:00
Michal Privoznik
9eae8398ed qemu: Prefer nvdimmPath over hugepages for memory-backend-file
If a domain has hugepages configured and we're currently building
memory-backend-file for a nvdimm device that domain has we will
put hugepages path onto the command line. It should have been
nvdimm path configured in the XML.

Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
2018-08-13 16:33:45 +02:00
Pavel Hrdina
0a476f1521 conf: Introduce virDomainDefPostParseMemtune
Previously we were ignoring "nodeset" attribute for hugepage pages
if there was no guest NUMA topology configured in the domain XML.
Commit <fa6bdf6afa878b8d7c5ed71664ee72be8967cdc5> partially fixed
that issue but it introduced a somehow valid regression.

In case that there is no guest NUMA topology configured and the
"nodeset" attribute is set to "0" it was accepted and was working
properly even though it was not completely valid XML.

This patch introduces a workaround that it will ignore the nodeset="0"
only in case that there is no guest NUMA topology in order not to
hit the validation error.

After this commit the following XML configuration is valid:

  <memoryBacking>
    <hugepages>
      <page size='2048' unit='KiB' nodeset='0'/>
    </hugepages>
  </memoryBacking>

but this configuration remains invalid:

  <memoryBacking>
    <hugepages>
      <page size='2048' unit='KiB' nodeset='0'/>
      <page size='1048576' unit='KiB'/>
    </hugepages>
  </memoryBacking>

The issue with the second configuration is that it was originally
working, however changing the order of the <page> elements resolved
into using different page size for the guest.  The code is written
in a way that it expect only one page configured and always uses only
the first page in case that there is no guest NUMA topology configured.
See qemuBuildMemPathStr() function for details.

Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1591235

Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
2018-08-10 17:06:48 +02:00
Pavel Hrdina
5c93dfb46d conf: Move hugepage XML validation check out of qemu_command
We can safely validate the hugepage nodeset attribute at a define time.
This validation is not done for already existing domains when the daemon
is restarted.

All the changes to the tests are necessary because we move the error
from domain start into XML parse.

Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
2018-08-10 17:06:48 +02:00
Pavel Hrdina
154b08a3a7 tests: introduce hugepages-nodeset
This use-case was broken by commit
<fa6bdf6afa878b8d7c5ed71664ee72be8967cdc5>.

We allowed this configuration and it was working as expected therefore
we can consider it as regression.  We should have never allowed such
configuration so now the best solution is in case of non-numa guest
silently ignore the 'nodeset' attribute if it's set to '0'.

That will be fixed by following patches.

Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
2018-08-10 17:06:48 +02:00
Pavel Hrdina
6ff69e936e tests: introduce hugepages-default-1G-nodeset-2M
This test case is currently working but it uncovers existing issue
in our code that the generated QEMU commandline uses the default 1G
hugepage instead of the 2M hugepage specified for exact node.

The issue in our code is that for non-numa guests we take into account
only the first hugepage.  This will be fixed as invalid configuration
since it doesn't make any sense to set default and specific hugepage
for non-numa guest.

Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
2018-08-10 17:06:48 +02:00
Pavel Hrdina
7e0ac26867 tests: rename hugepages-pages7 into pages-dimm-discard
Remove unnecessary XML elements as well.

Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
2018-08-10 17:06:48 +02:00
Pavel Hrdina
3d553f4d4b tests: rename hugepages-pages6 into hugepages-default-system-size
Remove unnecessary XML elements as well.

Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
2018-08-10 17:06:48 +02:00
Pavel Hrdina
6110770324 tests: rename hugepages-pages5 into hugepages-default-2M
Remove unnecessary XML elements as well.

Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
2018-08-10 17:06:48 +02:00
Pavel Hrdina
5534edcb9e tests: rename hugepages-pages4 into hugepages-numa-nodeset-nonexist
Remove unnecessary XML elements as well.

Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
2018-08-10 17:06:48 +02:00
Pavel Hrdina
96937ccf26 tests: rename hugepages-pages3 into hugepages-numa-nodeset-part
Remove unnecessary XML elements as well.

Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
2018-08-10 17:06:48 +02:00
Pavel Hrdina
9a6674c0ff tests: extract pages-discard-hugepages out of hugepages-pages3
Similar thing happens as for pages-discard, it is not passed to QEMU.

Signed-off-by: Pavel Hrdina <phrdina@redhat.com>
2018-08-10 17:06:48 +02:00