Now that committing transactions using pid == -1 means that we're
not fork()-ing to run the transaction in a specific namespace, we
can utilize the transaction processing semantics in order to
start, run a or multiple commands, and then commit the
transaction without being concerned with other interactions or
transactions interrupting the processing. This will eventually
allow us to have a single place where all the paths can be
locked, followed by relabeling and unlocking again.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: John Ferlan <jferlan@redhat.com>
It will be desirable to run transactions more often than we
currently do. Even if the domain we're relabeling the paths for
does not run in a namespace. If that's the case, there is no need
to fork() as we are already running in the right namespace. To
differentiate whether transaction code should fork() or not the
@pid argument now accepts -1 (which means do not fork).
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: John Ferlan <jferlan@redhat.com>
qemu_security: Require full wrappers for APIs that might touch a file
In the future, the transactions are not going to be optional and
they will be run regardless of domain using namespace to collect
list of paths to be relabeled.
To make sure there won't be an API that goes behind transaction
code back update the comment that serves as decision manual
whether an API must be fully implemented or plain #define is
sufficient.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: John Ferlan <jferlan@redhat.com>
Even though the current use of the functions does not require full
implementation with transactions (none of the callers passes a path
somewhere under /dev), it doesn't hurt either. Moreover, in
future patches the paradigm is going to shift so that any API
that touches a file is required to use transactions.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: John Ferlan <jferlan@redhat.com>
Even though the current use of the function does not require full
implementation with transactions (none of the callers pass a path
somewhere under /dev), it doesn't hurt either. Moreover, in
future patches the paradigm is going to shift so that any API
that touches a file is required to use transactions.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: John Ferlan <jferlan@redhat.com>
Jim Fehlig [Tue, 27 Mar 2018 17:50:14 +0000 (11:50 -0600)]
libxl: drop support for Xen < 4.6
Currently the libxl driver claims support for Xen >= 4.4, but
Xen 4.4 and 4.5 are no longer supported upstream. Let's increase
the minimum supported Xen version to 4.6 and change the defined
LIBXL_API_VERSION to 0x040500, which is the API version defined
when Xen 4.6 was released.
Since Xen 4.6 contains a pkgconfig file, drop the now unused code
that falls back to using LIBVIRT_CHECK_LIB in the absence of
pkgconfig file. In addition, bumping the LIBXL_API_VERSION
required adjusting the calls to libxl_set_vcpuaffinity to account
for the extra parameter in the 0x040500 version of the API.
Signed-off-by: Jim Fehlig <jfehlig@suse.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
Lin Ma [Thu, 13 Sep 2018 07:54:24 +0000 (15:54 +0800)]
util: Return a virArpTablePtr when the nlmsghdr for loop is over
commit b00c9c39 removed the label end_of_netlink_messages and 'return
table' statement, It causes the function virArpTableGet doesn't return
a proper virArpTable pointer.
How to reproduce:
# virsh domiflist sles12sp3
Interface Type Source Model MAC
-------------------------------------------------------
vnet0 network default virtio 52:54:00:cd:02:e6
# virsh domifaddr sles12sp3 --source arp
error: Failed to query for interfaces addresses
error: An error occurred, but the cause is unknown
It seems that the "if (nh->nlmsg_type == NLMSG_DONE)" statement won't be
meted. So this patch adds 'return table' when the iterations of nlmsghdr
for loop is over.
Signed-off-by: Lin Ma <lma@suse.com> Reviewed-by: Chen Hanxiao <chenhanxiao@gmail.com> Reviewed-by: John Ferlan <jferlan@redhat.com>
It is not a problem at all if the `tss` user/group does not exist, the code
fallbacks to the `root` user/group. However we report a warning for no reason
on every start-up. Fix this by checking if the user/group actually exists.
Signed-off-by: Martin Kletzander <mkletzan@redhat.com>
Add functions for checking if user or group exists
Instead of duplicating the code from virGet{User,Group}IDByName(), which are
static anyway, extend those functions to accept NULL pointers for the result and
a boolean for controlling the error reporting.
Signed-off-by: Martin Kletzander <mkletzan@redhat.com>
Ján Tomko [Sat, 8 Sep 2018 22:13:23 +0000 (00:13 +0200)]
tests: do not mangle real qemu caps in xml2argvtest
None of the things testUpdateQEMUCaps adjusts are applicable
for tests that use the DO_TEST_CAPS macros, i.e.
real QEMU capabilities parsed from the XML files:
The architecture must be chosen before we even open the caps
file, CPU models are already present and the expensive HostModel
computation was already done in virQEMUCapsLoadCache.
Introduce FLAG_REAL_CAPS and skip the whole testUpdateQEMUCaps
function for DO_TEST_CAPS.
Ján Tomko [Sat, 8 Sep 2018 21:21:46 +0000 (23:21 +0200)]
tests: report errors in QEMU XML->startup XML tests
Now that the function is only run if requested by
the FLAG_STEAL_VM flag, we know that missing data
is an error, not a request to skip the test.
The existence of the output file is now checked by
virTestCompareToFile, which allows usage of
the VIR_TEST_REGENERATE_OUTPUT=1 env variable
to generate new test cases.
Ján Tomko [Sat, 8 Sep 2018 21:10:23 +0000 (23:10 +0200)]
tests: only run startup XML tests if requested
Use the recently introduced flag as a witness.
This reduces the apparent number of test cases
to the real number of test cases.
Note that this does not suffer from the same problem
as commit 70255fa was fixing, because the condition
for running virTestRun does not depend on results
of previous tests.
John Ferlan [Tue, 11 Sep 2018 12:46:46 +0000 (08:46 -0400)]
conf: Move hypervisor specific nhugepage checks
Commit 82327038 moved a couple of checks out of the XML parser
into the domain validation; however, those checks seem to be more
useful as hypervisor specific checks rather than the more general
domain conf checks (nothing in the docs indicate a specific error).
Fortunately only QEMU was processing the memoryBacking, thus
add the changes to qemuDomainDefValidateMemory and change the
code a bit to make usage of the similar deref to def->mem and
the mem->nhugepages filter.
Signed-off-by: John Ferlan <jferlan@redhat.com> Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Laine Stump [Fri, 7 Sep 2018 01:09:45 +0000 (21:09 -0400)]
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.
Made redundant by commit 1e9a083 which switched to using
qemuProcessCreatePretendCmd, where capabilities are filtered
in qemuProcessInit after being fetched from the cache.
When processing the inputvol for encryption, we need to handle
the case where the inputvol is encrypted. This then allows for
the encrypted inputvol to be used either for an output encrypted
volume or an output volume of some XML provided type.
Add tests to show the various conversion options when either input
or output is encrypted. This includes when both are encrypted.
Signed-off-by: John Ferlan <jferlan@redhat.com> ACKed-by: Michal Privoznik <mprivozn@redhat.com>
John Ferlan [Tue, 21 Aug 2018 13:53:12 +0000 (09:53 -0400)]
storage: Allow for inputvol to have any format for encryption
Commit 39cef12a9 altered/fixed the inputvol processing to create
a multistep process when using an inputvol to create an encrypted
output volume; however, it unnecessarily assumed/restricted the
inputvol to be of 'raw' format only.
Modify the processing code to allow the inputvol format to be checked
and used in order to create the encrypted volume.
Signed-off-by: John Ferlan <jferlan@redhat.com> ACKed-by: Michal Privoznik <mprivozn@redhat.com>
qemu_domain: Drop namespace check from top level functions
In some cases we are checking if the mount namespace is enabled
at two places: one is at the beginning of exported function (e.g.
qemuDomainNamespaceSetupDisk()) and the other is at the beginning
of qemuDomainNamespaceMknodPaths() which is called from the
former function anyway. Then we have some other functions which
rely on the later check solely.
In order to compensate for possibly needless function call,
qemuDomainNamespaceMknodPaths() returns early if @npaths is zero.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: John Ferlan <jferlan@redhat.com>
Shi Lei [Fri, 7 Sep 2018 07:17:25 +0000 (15:17 +0800)]
util: netlink: Add some wrapper macros to get rid of redundancy
This patch adds wrapper macros around nla_nest_[start|end] and nla_put,
thus getting rid of some redundancy and making virNetlinkNewLink more
readable.
Signed-off-by: Shi Lei <shi_lei@massclouds.com> Signed-off-by: Erik Skultety <eskultet@redhat.com>
Jim Fehlig [Fri, 24 Aug 2018 21:38:14 +0000 (15:38 -0600)]
libxl: join with thread receiving migration data
It is possible the incoming VM is not fully started when the finish
phase of migration is executed. In libxlDomainMigrationDstFinish,
wait for the thread receiving the VM to complete before executing
finish phase tasks.
Signed-off-by: Jim Fehlig <jfehlig@suse.com> ACKed-by: Michal Privoznik <mprivozn@redhat.com>
Jim Fehlig [Wed, 29 Aug 2018 17:11:00 +0000 (11:11 -0600)]
libxl: fix job handling across migration phases on dst
The libxlDomainMigrationDst* functions are a bit flawed in their
handling of modify jobs. A job begins when the destination host
begins receiving the incoming VM and ends after the VM is started.
The finish phase contains another BeginJob/EndJob sequence.
This patch changes the logic to begin a job for the incoming VM
in the prepare phase and end the job in the finish phase.
Signed-off-by: Jim Fehlig <jfehlig@suse.com> ACKed-by: Michal Privoznik <mprivozn@redhat.com>
Jim Fehlig [Tue, 28 Aug 2018 23:30:18 +0000 (17:30 -0600)]
libxl: fix job handling across migration phases on src
The libxlDomainMigrationSrc* functions are a bit flawed in their
handling of modify jobs. A job begins at the start of the begin
phase but ends before the phase completes. No job is running for
the remaining phases of migration on the source host.
Change the logic to keep the job running after a successful begin
phase, and end the job in the confirm phase. The job must also end
in the perform phase in the case of error since confirm phase would
not be executed.
Signed-off-by: Jim Fehlig <jfehlig@suse.com> ACKed-by: Michal Privoznik <mprivozn@redhat.com>
Jim Fehlig [Tue, 28 Aug 2018 23:13:54 +0000 (17:13 -0600)]
libxl: fix logic in P2P migration
libxlDoMigrateSrcP2P() performs all phases of the migration
protocol for peer-to-peer migration. Unfortunately the logic
was a bit flawed since it is possible to skip the confirm
phase after a successfull begin and prepare phase. Fix the
logic to always call the confirm phase after a successful begin
and perform. Skip the confirm phase if begin or perform fail.
Signed-off-by: Jim Fehlig <jfehlig@suse.com> ACKed-by: Michal Privoznik <mprivozn@redhat.com>
Jim Fehlig [Fri, 24 Aug 2018 20:55:03 +0000 (14:55 -0600)]
libxl: migration: defer removing VM until finish phase
If for any reason the restore of a VM fails on the destination host
in a migration operation, the VM is removed (if not persistent) from
the virDomainObjList, meaning it is no longer available for additional
cleanup or processing in the finish phase. Defer removing the VM from
the virDomainObjList until the finish phase, which already contains
logic to remove the VM.
Signed-off-by: Jim Fehlig <jfehlig@suse.com> ACKed-by: Michal Privoznik <mprivozn@redhat.com>
John Ferlan [Fri, 7 Sep 2018 12:20:15 +0000 (08:20 -0400)]
test: Remove possible infinite loop in virnetsockettest
Commit 39015a6f3 modified the test to be more reliable/realistic,
but without checking the return status of virEventRunDefaultImpl
it's possible that the test could run infinitely.
Found by Coverity
Signed-off-by: John Ferlan <jferlan@redhat.com> ACKed-by: Michal Privoznik <mprivozn@redhat.com>
John Ferlan [Fri, 7 Sep 2018 20:01:27 +0000 (16:01 -0400)]
qemu: Remove duplicated qemuAgentCheckError
Commit 5b3492fadb moved qemuAgentCheckError calls into
qemuAgentCommand for various reasons; however, subsequent
commit 0977b8aa0 adding a new command made call again
So let's just remove the duplicitous call from
qemuAgentGetInterfaces.
Signed-off-by: John Ferlan <jferlan@redhat.com> ACKed-by: Michal Privoznik <mprivozn@redhat.com>
John Ferlan [Fri, 7 Sep 2018 20:01:26 +0000 (16:01 -0400)]
qemu: Remove duplicated qemuAgentCheckError
Commit 5b3492fadb moved qemuAgentCheckError calls into
qemuAgentCommand for various reasons; however, subsequent
commit b1aa91e14 restored the call. So let's just remove
the duplicitous call from qemuAgentSetVCPUsCommand.
Signed-off-by: John Ferlan <jferlan@redhat.com> ACKed-by: Michal Privoznik <mprivozn@redhat.com>
All backingStore XML definitions have a XML tag with the timestamp. This
timestamp is not defined insinde RNG volume storage schema and it is
causing some problems to validate and check volume XMLs.
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.
Wang Huaqiang [Mon, 27 Aug 2018 11:23:04 +0000 (19:23 +0800)]
conf: Renamed 'controlBuf' to 'childrenBuf'
To add CMT/MBM feature and let code be consistent in later patches,
renaming variable name from 'controlBuf' to 'childrenBuf', locates
in functions 'virCapabilitiesFormatCaches' and
'virCapabilitiesFormatMemoryBandwidth'.
Signed-off-by: Wang Huaqiang <huaqiang.wang@intel.com> Reviewed-by: John Ferlan <jferlan@redhat.com>
Shi Lei [Fri, 7 Sep 2018 08:14:46 +0000 (16:14 +0800)]
util: netlink: Replace virNetDevPutExtraHeader with nlmsg_append
nlmsg_append from the libnl library provides exactly the same
functionality, so we should rely on that instead. This also allows us to
drop the aforementioned function completely.
Signed-off-by: Shi Lei <shi_lei@massclouds.com> Reviewed-by: Erik Skultety <eskultet@redhat.com>
Shi Lei [Fri, 7 Sep 2018 08:27:39 +0000 (16:27 +0800)]
util: netdevip: Fix a memleak in virNetDevIPRouteAdd
@resp is allocated by virNetlinkCommand and the caller is responsible
for freeing the buffer. Since we already converted this module to use
VIR_AUTO{FREE,PTR} macros, let's resolve the problem by using them.
Signed-off-by: Shi Lei <shi_lei@massclouds.com> Reviewed-by: Erik Skultety <eskultet@redhat.com>
Instantiation of an nwfilter binding is only allowed when
the net->filter is defined for the network; however, the
teardown of the binding does not make this check. This
leaves open the possibility that the teardown could be
called during guest shutdown/teardown in session mode
resulting in the following error being logged:
So before going through the teardown processing, let's
be sure the network had a filter and then attempt to
get a connection. For session mode it's not even possible
create an nwfilter binding.
Signed-off-by: John Ferlan <jferlan@redhat.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
John Ferlan [Mon, 27 Aug 2018 14:41:35 +0000 (10:41 -0400)]
nwfilter: Disallow binding creation in session mode
Similar to nwfilterDefineXML, let's be sure the a filter binding
creation is not attempted in session mode and generate the proper
error message.
Failure to open nwfilter in session mode (nwfilterConnectOpen)
fails already, but that doesn't stop the free thinker from using
a different connection in order to attempt to attempt to create
the binding. Although even doing that would result in a failure:
$ virsh nwfilter-binding-create QEMUGuest1-binding.xml
error: Failed to create network filter from QEMUGuest1-binding.xml
error: internal error: Could not get access to ACL tech driver 'ebiptables'
$
Signed-off-by: John Ferlan <jferlan@redhat.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
qemu: Unify generation of command line for virtio devices
A virtio device such as
<controller type='scsi' model='virtio-scsi'/>
will be translated to one of four different QEMU devices
based on the address type. This behavior is the same for
all virtio devices, but unfortunately we have separate
ad-hoc code dealing with each and every one of them: not
only this is pointless duplication, but it turns out
that most of that code is not robust against new address
types being introduced and some of it is outright buggy.
Introduce a new function, qemuBuildVirtioDevStr(), which
deals with the issue in a generic fashion, and rewrite
all existing code to use it.
This fixes a bunch of issues such as virtio-serial-pci
being used with virtio-mmio addresses and virtio-gpu
not being usable at all with virtio-mmio addresses.
It also introduces a couple of minor regressions,
namely no longer erroring out when attempting to
use virtio-balloon and virtio-input devices with
virtio-s390 addresses; that said, virtio-s390 has
been superseded by virtio-ccw such a long time ago
that recent QEMU releases have dropped support for
the former entirely, so re-implementing such
device-specific validation is not worth it.
Signed-off-by: Andrea Bolognani <abologna@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
qemu: Check for virtio-input capabilities at validate time
The appropriate time to ensure the required capabilities are
present is validate rather than command line generation: add
a new qemuDomainDeviceDefValidateInput() function and move
all existing checks there.
Signed-off-by: Andrea Bolognani <abologna@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
So far we've only formatted it for virtio-blk-pci and
virtio-blk-ccw, but other virtio-blk devices also support
the corresponding option; moreover, we've always formatted
it for all virtio-scsi devices.
Signed-off-by: Andrea Bolognani <abologna@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
qemu: Remove duplicated option formatting for virtio devices
There are several functions where we pointlessly duplicate
parts of the format string and pass the same arguments:
refactor them so that the common parts are formatted separately
from the variable parts.
Signed-off-by: Andrea Bolognani <abologna@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
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>
If a domain is configured to use <source type='file'/> under
<memoryBacking/> we have to honour that setting and produce
-mem-path on the command line. We are not doing so if domain has
no guest NUMA nodes nor hugepages.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: John Ferlan <jferlan@redhat.com>
Julio Faracco [Wed, 5 Sep 2018 04:20:56 +0000 (01:20 -0300)]
qemu: unlink the error report from VIR_STRDUP.
The function to retrieve the file system info using QEMU-GA is using
some conditionals to retrieve the info. This is wrong because the error
of some conditionals will be raised if VIR_STRDUP return errors and not
if some problem occurred with JSON.
Signed-off-by: Julio Faracco <jcfaracco@gmail.com> Reviewed-by: John Ferlan <jferlan@redhat.com>
This commit implements the function qemuAgentGetHostname() that uses
the QEMU guest agent command 'guest-get-host-name' to retrieve the
guest hostname of virtual machine running the QEMU-GA.
Signed-off-by: Julio Faracco <jcfaracco@gmail.com> Reviewed-by: John Ferlan <jferlan@redhat.com>
tests: Add simple headless guests using latest caps
The new tests use DO_TEST_CAPS_ARCH_LATEST() with an input
XML describing a very simple headless guest and cover most
architectures and machine types we care about.
Signed-off-by: Andrea Bolognani <abologna@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
Andrea Bolognani [Thu, 30 Aug 2018 16:03:58 +0000 (18:03 +0200)]
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>
There's a single user for it which takes an existing
virPCIDeviceAddress, passes its various bits to the
function which in turn constructs a virPCIDevice and
then copies the string representation for the caller
to use: we can use virPCIDeviceAddressAsString()
instead and avoid creating the virPCIDevice in the
first place. Since the function ends up having no
users after the change, we can just drop it.
Signed-off-by: Andrea Bolognani <abologna@redhat.com> Reviewed-by: Martin Kletzander <mkletzan@redhat.com>
Michal Privoznik [Wed, 22 Aug 2018 07:56:07 +0000 (09:56 +0200)]
virLockManagerLockDaemonAddResource: Switch to cleanup label rather than error
This will help in future expansions of the code when it is be
harder to track if @newName and/or @newLockspace were already
allocated or not and thus whether it is safe to 'return' or we
need to 'goto error'. By using the 'cleanup' label those two
cases merge into a single one.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: John Ferlan <jferlan@redhat.com>
projects / libvirt.git / log