Extended the json monitor test program with support for query-cpus-fast
and added a sample file set for x86 data obtained using the it.
Signed-off-by: NViktor Mihajlovski <mihajlov@linux.vnet.ibm.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

Print out the halted property in the test output.
Signed-off-by: NViktor Mihajlovski <mihajlov@linux.vnet.ibm.com>
Reviewed-by: NJohn Ferlan <jferlan@redhat.com>

It turns out that our implementation of the hashing function is
endian-dependent and thus if used on various architectures the testsuite
may have different results. Work this around by mocking virHashCodeGen
to something which does not use bit operations instead of just setting a
deterministic seed.

Prior to qemu 2.5 the node names would not be generated, thus would be
missing from 'query-blockstats' and 'query-named-block-nodes'. Test that
the code correctly detects nothing.

Additionally make sure that a VM without disks does not cause problems.

The test case change is necessary as our test file checker does not play
well with empty files.

Test storage was created on a rhel/centos 7 node using targetcli.
Reviewed-by: NEric Blake <eblake@redhat.com>

Test file created by:

qemu-img create -f luks /var/lib/libvirt/images/luks 10M \
        -o key-secret=asdf --object secret,id=asdf,format=raw,data=asdf

Used in libvirt as:
 <disk type='file' device='disk'>
   <driver name='qemu' type='raw'/>
   <source file='/var/lib/libvirt/images/luks'/>
   <target dev='vda' bus='virtio'/>
   <encryption format='luks'>
     <secret type='passphrase' uuid='9b2c831a-fdb9-4c09-873c-1959580589e1'/>
   </encryption>
 </disk>
Reviewed-by: NEric Blake <eblake@redhat.com>

qemu 2.9 returns an extra layer in the backing data if a block job is
active. Add a test case to see whether our code properly detects and
ignores such layer.

The test data was prepared by creating a backing chain of qcow2 images
(with qemu-img and with libvirt's snapshot feature).

One of the layers was then merged back by doing a block-commit:

virsh blockcommit VM hda --top /var/lib/libvirt/images/b

and then a block-copy job was started and kept in synchronized phase:

virsh blockcopy VM hda /tmp/tgt.img --transient job
Reviewed-by: NEric Blake <eblake@redhat.com>

Add the blockstats data and fix the expected output.
Reviewed-by: NEric Blake <eblake@redhat.com>

Add the blockstats data and fix the expected output.

Test data was created as:

qemu-img create -f qcow2 img0 10M
qemu-img create -f qcow2 -o "backing_fmt=qcow2,backing_file=img0" img1
qemu-img create -f qcow2 -o "backing_fmt=qcow2,backing_file=img1" img2
...
Reviewed-by: NEric Blake <eblake@redhat.com>

With the new approach we are actually able to correctly detect node
names for the two instances of the same backing file.

Test images were created as:
qemu-img create -f qcow2 /var/lib/libvirt/images/base.qcow2 10M
qemu-img create -f qcow2 \
   -o "backing_fmt=qcow2,backing_file=/var/lib/libvirt/images/base.qcow2 \
   /var/lib/libvirt/images/a.qcow2
qemu-img create -f qcow2 \
   -o "backing_fmt=qcow2,backing_file=/var/lib/libvirt/images/base.qcow2 \
   /var/lib/libvirt/images/b.qcow2

and then used for two separate disks.
Reviewed-by: NEric Blake <eblake@redhat.com>

Store the 'drv' field both for the storage node and for the format node
and format them in the test case.
Reviewed-by: NEric Blake <eblake@redhat.com>

We can now iterate the hash table and print all detected backing chains.
This simplifies calling of the test cases.
Reviewed-by: NEric Blake <eblake@redhat.com>

Remove the complex and unreliable code which inferred the node name
hierarchy only from data returned by 'query-named-block-nodes'. It turns
out that query-blockstats contain the full hierarchy of nodes as
perceived by qemu so the inference code is not necessary.

In query blockstats, the 'parent' object corresponds to the storage
behind a storage volume and 'backing' corresponds to the lower level of
backing chain. Since all have node names this data can be really easily
used to detect node names.

In addition to the code refactoring the one remaining test case needed
to be fixed along.
Reviewed-by: NEric Blake <eblake@redhat.com>

The node name and backing file name can be inferred from the hierarchy.
This will also help when converting to detect node names using
query-blockstats data.
Reviewed-by: NEric Blake <eblake@redhat.com>

Rename 'json' and related variables to 'nodeNameJson'. Also rename the
test files along. This is a preparation for modifying how we detect node
names which will also require data from 'query-blockstats'.
Reviewed-by: NEric Blake <eblake@redhat.com>

Test cases named '1' and '2' differed only in the length of the backing
chain, so remove test case '2' and rename test '1' to 'basic'.
Reviewed-by: NEric Blake <eblake@redhat.com>

vcpu properties gathered from query-hotpluggable cpus need to be passed
back to qemu. As qemu did not use the node-id property until now and
libvirt forgot to pass it back properly (it was parsed but not passed
around) we did not honor this.

This patch adds node-id to the structures where it was missing and
passes it around as necessary.

The test data was generated with a VM with following config:
    <numa>
      <cell id='0' cpus='0,2,4,6' memory='512000' unit='KiB'/>
      <cell id='1' cpus='1,3,5,7' memory='512000' unit='KiB'/>
    </numa>

Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1452053

oVirt uses relative names with directories in them. Test such
configuration. Also tests a snapshot done with _REUSE_EXTERNAL and a
relative backing file pre-specified in the qcow2 metadata.

Since we have to match the images by filename a common backing image
will break the detection process. Add a test case to see that the code
correctly did not continue the detection process.

The code is rather magic so a test case will help making sure that
everything works well. The first case is a simple backing chain.

Commit 3f71c797 added 'qemu_id' field to track the id of the cpu
as reported by query-cpus. The patch did not include changes necessary
to propagate the id through the functions matching the data to the
libvirt cpu structures and thus all vcpus had id 0.

The field is named 'enable_id' in other structures and a patch recently
added 'qemu_id' which has different semantics. To avoid confusion in the
tests rename the field.

Return whether a vcpu entry is hotpluggable or online so that upper
layers don't have to infer the information from other data.

Advantage is that this code can be tested by unit tests.

The reported data is unusual so add it to the test suite.

Test the algorithm that extracts the order in which the vcpu entries
were plugged in on a sample of data created by plugging in vcpus
arbitrarily.

Power 8 platform's basic hotpluggable unit is a core rather than a
thread for x86_64 family. This introduces most of the complexity of the
matching code and thus needs to be tested.

The test data contain data captured from in-order cpu hotplug and
unplug operations.

During review it was reported that adding at least 11 vcpus creates a
collision of prefixes in the monitor matching algorithm. Add a test case
to verify that the problem won't happen.

As the combination algorithm is rather complex and ugly it's necessary
to make sure it works properly. Add test suite infrastructure for
testing it along with a basic test based on x86_64 platform.

CPUID instruction normally takes its parameter from EAX, but sometimes
ECX is used as an additional parameter. This patch prepares the x86 CPU
driver code for the new 'ecx_in' CPUID parameter.
Signed-off-by: NJiri Denemark <jdenemar@redhat.com>

Signed-off-by: NJiri Denemark <jdenemar@redhat.com>

CPUID instruction normally takes its parameter from EAX, but sometimes
ECX is used as an additional parameter. Let's rename 'function' to
'eax_in' in preparation for adding 'ecx_in'.
Signed-off-by: NJiri Denemark <jdenemar@redhat.com>

The qemu monitor supports retrieval of actual CPUID bits presented to
the guest using QMP monitor. Add APIs to extract these information and
tests for them.
Signed-off-by: NPeter Krempa <pkrempa@redhat.com>