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/1601318Signed-off-by: NLaine Stump <laine@laine.org>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Broken by v3.10.0-68-gaed3d038.
Signed-off-by: NJiri Denemark <jdenemar@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Attempting to use a chardev definition like

  <serial type='unix'>
    <target type='isa-serial'/>
  </serial>

correctly results in an error being reported, since the source
path - a required piece of information - is missing; however,
the very similar

  <serial type='unix'>
    <target type='pci-serial'/>
  </serial>

was happily accepted by libvirt, only to result in libvirtd
crashing as soon as the guest was started.

The issue was caused by checking the chardev's targetType
against whitelisted values from virDomainChrChannelTargetType
without first checking the chardev's deviceType to make sure
it is actually a channel, for which the check makes sense,
rather than a different type of chardev.

The only reason this wasn't spotted earlier is that the
whitelisted values just so happen to correspond to USB and
PCI serial devices and Xen and UML consoles respectively,
all of which are fairly uncommon.

https://bugzilla.redhat.com/show_bug.cgi?id=1609720Signed-off-by: NAndrea Bolognani <abologna@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

The corresponding structs are declared there.
Signed-off-by: NAndrea Bolognani <abologna@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

The function is called on a virDomainDeviceInfo, so it
should be declared along with it.

Moving this function requires moving and making public
virDomainDeviceCCWAddressIsValid() as well, but that's
perfectly fine since the same reasoning above also
applies to it, due to virDomainDeviceCCWAddress being
(correctly) declared in device_conf.
Signed-off-by: NAndrea Bolognani <abologna@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

It's used in virDomainDeviceInfo, which makes
domain_conf the wrong place to declare it.
Signed-off-by: NAndrea Bolognani <abologna@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

https://bugzilla.redhat.com/show_bug.cgi?id=1623157

Changing MTU on a running guest is not possible and trying to do
so made us face many problems. That's why we forbid it in
5f44d7e3. However, there is still one possible path where
users can sneak in change: migration XML.
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>
Reviewed-by: NDaniel P. Berrangé <berrange@redhat.com>

https://bugzilla.redhat.com/show_bug.cgi?id=1621910

When introducing this check back in 4ad54a41 my mindset was
that if an element is missing in update XML then user is
requesting for removal of the corresponding setting. For
instance, if <bandwidth/> is not present in update XML any QoS
previously set on <interface/> is cleared out. Well this
assumption is correct but only to some extent.

Turns out, we have some users who when updating path to ISO
image construct very minimalistic disk XML and pass it to device
update API. Such XML is lacking a lot of information, and alias
is one of them. This triggers error in
virDomainDefCompatibleDevice() because we think that user is
requesting to remove the alias. Well, they are not.
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

None of the existing models is suitable for use with
RISC-V virt guests, and we don't want information about
the serial console to be missing from the XML.

The name is based on comments in qemu/hw/riscv/virt.c:

  RISC-V machine with 16550a UART and VirtIO MMIO

and in qemu/hw/char/serial.c:

  QEMU 16550A UART emulation

along with the output of dmesg in the guest:

  Serial: 8250/16550 driver, 4 ports, IRQ sharing disabled
  10000000.uart: ttyS0 at MMIO 0x10000000 (irq = 13,
    base_baud= 230400) is a 16550A
Signed-off-by: NAndrea Bolognani <abologna@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Commit deb057fd added a switch without a default case.
Add it and call virReportEnumRangeError for _LAST too.
Signed-off-by: NJán Tomko <jtomko@redhat.com>
Reviewed-by: NAndrea Bolognani <abologna@redhat.com>

Similarly to backing store indexes which will become stable eventually
we need also to be able to format and store in the status XML for later
use the index for the top level of the backing chain.

Add XML formatter, parser, schema and docs.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

Allow storing of private data in the status XML for disks.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

If a user configures the backing chain in the XML we should not ignore
it. We already do parse it but don't format it out. As a
safety-precaution don't attempt to format detected chain into the
inactive XML.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>
Reviewed-by: NJán Tomko <jtomko@redhat.com>

https://bugzilla.redhat.com/show_bug.cgi?id=1610072

Due to historical reasons we were not parsing device info on
guestfwd channel. Sure, it doesn't make much sense to parse
<address/> but it surely makes sense to parse its alias (which
might be an user alias).

This reverts commit 47a3dd46
which fixed https://bugzilla.redhat.com/show_bug.cgi?id=1172526.
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>
Reviewed-by: NAndrea Bolognani <abologna@redhat.com>

Qemu-3.0 supports Hyper-V-style PV TLB flush, Windows guests can benefit
from this feature as KVM knows which vCPUs are not currently scheduled (and
thus don't require any immediate action).
Signed-off-by: NVitaly Kuznetsov <vkuznets@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Qemu-3.0 supports so-called 'Reenlightenment' notifications and this (in
conjunction with 'hv-frequencies') can be used make Hyper-V on KVM pass
stable TSC page clocksource to L2 guests.
Signed-off-by: NVitaly Kuznetsov <vkuznets@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Qemu-2.12 gained 'hv-frequencies' cpu flag to enable Hyper-V frequency
MSRs. These MSRs are required (but not sufficient) to make Hyper-V on
KVM pass stable TSC page clocksource to L2 guests.
Signed-off-by: NVitaly Kuznetsov <vkuznets@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Add return value check to virResctrlAllocForeachCache in
virDomainCachetuneDefFormat. The virResctrlAllocForeachCache does have
return value, so need check return value to make sure function executed
without error.
Signed-off-by: NBing Niu <bing.niu@intel.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Introduce a new section memorytune to support memory bandwidth allocation.
This is consistent with existing cachetune. As the example:
below:
  <cputune>
    ......
    <memorytune vcpus='0'>
      <node id='0' bandwidth='30'/>
    </memorytune>
  </cputune>

vpus      --- vpus subjected to this memory bandwidth.
id        --- on which node memory bandwidth to be set.
bandwidth --- the memory bandwidth percent to set.
Signed-off-by: NBing Niu <bing.niu@intel.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Factor out vcpus virDomainResctrlDef update from
virDomainCachetuneDefParse and introduce virDomainResctrlAppend.
virDomainResctrlAppend will format vcpus string and append a new
virDomainResctrlDef to virDomainDefPtr. So that this logic can
be reusable.
Signed-off-by: NBing Niu <bing.niu@intel.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Factor out vcpus overlapping detecting part from
virDomainCachetuneDefParse and introduce virDomainResctrlVcpuMatch.
Instead of allocating virResctrlAllocPtr by default, allocating
virResctrlAllocPtr after confirm vcpus not overlap with existing ones.
And virDomainResctrlVcpuMatch can be reused by other resource control
technologies. virDomainResctrlVcpuMatch can clarify old vcpus overlap
error whether an overlap or a redefinition.
Signed-off-by: NBing Niu <bing.niu@intel.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Extract vcpus parsing part from virDomainCachetuneDefParse into one
function called virDomainResctrlParseVcpus. So that vcpus parsing logic
can be reused by other resource control technologies. Adjust error
message and use node->name so that the error message can fit to all
technologies.
Signed-off-by: NBing Niu <bing.niu@intel.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Resctrl not only supports cache tuning, but also memory bandwidth
tuning. Renaming cachetune to resctrl to reflect that. With resctrl,
all allocation for different resources (cache, memory bandwidth) are
aggregated and represented by a virResctrlAllocPtr inside
virDomainResctrlDef.
Signed-off-by: NBing Niu <bing.niu@intel.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Some functions in virresctrl are for CAT only, while some of other
functions are for resource allocation, not just CAT. So change
their names to reflect the reality.
Signed-off-by: NBing Niu <bing.niu@intel.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Previously we were ignoring "nodeset" attribute for hugepage pages
if there was no guest NUMA topology configured in the domain XML.
Commit <fa6bdf6a> 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=1591235Signed-off-by: NPavel Hrdina <phrdina@redhat.com>

Signed-off-by: NPavel Hrdina <phrdina@redhat.com>

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: NPavel Hrdina <phrdina@redhat.com>

Signed-off-by: NPavel Hrdina <phrdina@redhat.com>

We cannot simply used the same code as for iscsi storage pool because
the default mode is 'host' which is not possible with iscsi-direct.
Reviewed-by: NJán Tomko <jtomko@redhat.com>
Signed-off-by: NPavel Hrdina <phrdina@redhat.com>

Validate that the provided XML shmem name is not directory specific to "."  or
".." as well as ensure that there is no path separator '/' in the name.

https://bugzilla.redhat.com/show_bug.cgi?id=1192400Signed-off-by: NSimon Kobyda <skobyda@redhat.com>
Reviewed-by: NMartin Kletzander <mkletzan@redhat.com>

Introducing the pool as a noop. Integration inside the build
system. Implementation will be in the following commits.
Signed-off-by: NClementine Hayat <clem@lse.epita.fr>
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>

https://bugzilla.redhat.com/show_bug.cgi?id=1591151

Add function virDomainInputDefValidate to validate input devices.
Make sure evdev attribute of source element is not used by mouse,
keyboard, and tablet input device.
Signed-off-by: NHan Han <hhan@redhat.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Use of enum types for struct fields is generally avoided since it causes
warnings if the compiler assumes the enum is unsigned. For example

  commit 8e2982b5
  Author: Cole Robinson <crobinso@redhat.com>
  Date:   Tue Jul 24 16:27:54 2018 -0400

    conf: Clean up virDomainDefParseCaps

Introduced a line:

  if ((def->virtType = virDomainVirtTypeFromString(virttype)) < 0) {

which causes a build failure with CLang

  conf/domain_conf.c:19143:65: error: comparison of unsigned enum expression < 0 is always false [-Werror,-Wtautological-compare]

as the compiler is free to optimize away the "< 0" check due to the
assumption that the enum type is unsigned and always in range.
Signed-off-by: NDaniel P. Berrangé <berrange@redhat.com>

SKIP_OSTYPE_CHECKS only hides some error reporting at this point,
so it can be foled into SKIP_VALIDATE
Acked-by: NMichal Privoznik <mprivozn@redhat.com>
Signed-off-by: NCole Robinson <crobinso@redhat.com>

We should still make an effort to fill in data, just not raise
an error if say an ostype/virttype combo disappeared from caps.
Acked-by: NMichal Privoznik <mprivozn@redhat.com>
Signed-off-by: NCole Robinson <crobinso@redhat.com>

The comment says:

    /* If the logic here seems fairly arbitrary, that's because it is :)
     * This is duplicating how the code worked before
     * CapabilitiesDomainDataLookup was added. We can simplify this,
     * but it would take a bit of work because the test suite fails
     * in numerous minor ways. */

Nowadays the test suite changes appear quite simple, just extending
test capabilities data a bit so that we aren't trying to define
invalid arch/os/virtType/machine combos
Acked-by: NMichal Privoznik <mprivozn@redhat.com>
Signed-off-by: NCole Robinson <crobinso@redhat.com>

- Convert to 'cleanup' label naming
- Use more than one 'tmp' string and do all freeing at the end
- Make the code easier to follow
Acked-by: NMichal Privoznik <mprivozn@redhat.com>
Signed-off-by: NCole Robinson <crobinso@redhat.com>

Handles parse virtType, os.type, bootloader bits, arch, machine,
emulator
Acked-by: NMichal Privoznik <mprivozn@redhat.com>
Signed-off-by: NCole Robinson <crobinso@redhat.com>

https://bugzilla.redhat.com/show_bug.cgi?id=1607825

Introduced by commit d48813e8.
Signed-off-by: NHan Han <hhan@redhat.com>
Reviewed-by: NErik Skultety <eskultet@redhat.com>

With 'switch' we can utilize the compile time enum checks which we can't
rely on with plain 'if' conditions.
Signed-off-by: NShi Lei <shilei.massclouds@gmx.com>
Reviewed-by: NErik Skultety <eskultet@redhat.com>