In netperf-2.6.0, we can assign two arguments (local send size &
remote send size) for TCP_MAERTS by '-m $local_send_size,remote_send_size'

Currently we only assigned useless local send size, this patch corrects
the parameters for TCP_MAERTS.
Signed-off-by: NAmos Kong <akong@redhat.com>

I already updated the netperf package to 2.6.0,
this patch updated the config file and file path.

wait_before_data.patch is no longer requested.
Signed-off-by: NAmos Kong <akong@redhat.com>

Signed-off-by: NLi Yang <liyang.fnst@cn.fujitsu.com>

Since shutdown can raise an error now, modify the
shutdown method to take this into account.
Signed-off-by: NLi Yang <liyang.fnst@cn.fujitsu.com>

Signed-off-by: NLi Yang <liyang.fnst@cn.fujitsu.com>

Signed-off-by: NLi Yang <liyang.fnst@cn.fujitsu.com>

If the JeOS image changed in the repos, we require the
latest one, therefore, this image has to be downloaded.
Since it will take some minutes, it will make people
nervously wonder what is going on.

It seems that my patch that prints a message for the
user to wait ends up being ignored, so, let's print
messages coming from the setup, which includes the
JeOS download progress bar. With the progress bar,
people will feel more reassured.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Signed-off-by: NLi Yang <liyang.fnst@cn.fujitsu.com>

Signed-off-by: NLi Yang <liyang.fnst@cn.fujitsu.com>

Signed-off-by: NLi Yang <liyang.fnst@cn.fujitsu.com>

Signed-off-by: NLi Yang <liyang.fnst@cn.fujitsu.com>

Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Recent Linux guests (including our JeOS) changed quite
drastically ifconfig output, making the current regexp
match against the command's output unfeasible. However,
it is possible to look on /sys/net/class to get the same
info, so let's do that if the regexp match didn't work.

Also, this method is called in a bunch of network tests,
that don't really treat its output, it is in our best
interest to make it raise a TestError exception when
it did not work (what happens today is that tests will
happily take None as a result of this function and
proceed, which will later result in a less comprehensible
test failure.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Verified all JeOS related tests that require root.
Obviously, there are some that did not get the
proper label, but we can get to that.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Pass nic nettype to hotplug method, also, correctly
handle hot plugging of nics using user mode. This fixes
running as non root, with JeOS.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

In order to display one more bit of useful information,
print the test number on the execution list in the
form (current test/total tests).
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

If colors are supported, when listing tests, use colors
in the test numbers (blue) and, if the test requires
root, print that in yellow, making it easier to process
the information being displayed.

In order to do that, do a little refactor in the Bcolors
class, so we expose colors as class attributes, and also,
move the print test list function to standalone_test.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

When listing the tests in the test list, it is useful
to know whether a given item does require root to run,
so print that information in the test list.

Unfortunately, only the kvm subtest allows to run a
subset of the tests as non root, so reflect that in
the runner heuristics. We plan on improving that.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Some tests do require admin privileges, so, when running
as a regular user and reaching one of those tests, it is
good to just skip the test instead of trying to execute
it.

Therefore, introduce a utility function to verify whether
we're running as root (actually, uid=0), and if not,
throw an exception. Also add a postprocess param, that
will throw an error if the param 'requires_root' is set.

This will also make possible to tell in the test runner
list whether a test requires root or not, making one
more useful information clear to the users.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

When no previous installer was found, no modules were
found to load. However, depending on the subtest type,
we can infer modules to load (see the kvm example, we
can load/unload kvm and kvm-[cpu_vendor] modules).

So do the possible to infer the modules, and load/unload
them safely.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Signed-off-by: NChris Evich <cevich@redhat.com>

The floppy test, in order to work among Windows and Linux,
was using a diff command to determine whether a file copied
to the floppy had its integrity preserved. The diff command
was not present on JeOS, but fortunately it is not necessary,
since we can simply evaluate the md5sum of the files.

Also, fix the relative positioning of the floppy files, using
the data_dir API.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

So the cdrom test can actually work under JeOS.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Use data_dir.get_data_dir() to get the actual path of
the cdrom file, instead of test.tmpdir.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

It is not necessary, considering that the qemu monitor
code already records and debug prints monitor inputs
and outputs.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

So it's easier to separate the configs for each subtest.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

*Fix parameters' typo for libvirt test modules.
 (Make paramemters unique in configuration)
*Use complete_uri to combinate a complete uri for virsh list test.
Signed-off-by: NYu Mingfei <yumingfei@cn.fujitsu.com>

* VirshPersistent.close_session() was not removing the session_id
  after successfully closing or encountering a closed/non-responsive
  session.
* Fixed some 'no space after comma' pylint complaints
* Added required '**dargs' parameter to snapshot_* functions
  so they will work with Virsh and VirshPersistant class
* Fixed snapshot_* functions return variable name to keep pylint happy.
* Updated docstrings for snapshot_* functions
Signed-off-by: NChris Evich <cevich@redhat.com>

The ElementTree installed in the system will take precedence
over the virttests version. So explicitly tell the virttest
version is the one to be used.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Signed-off-by: NChris Evich <cevich@redhat.com>

Signed-off-by: NChris Evich <cevich@redhat.com>

Signed-off-by: NChris Evich <cevich@redhat.com>

Signed-off-by: NChris Evich <cevich@redhat.com>

Signed-off-by: NChris Evich <cevich@redhat.com>

The same information can be obtained by directly reading
the arp cache file /proc/net/arp. The regexp used to
do the match should work even with no translation of
flags and addresses.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

If one wants to use bridge, those really have to be present.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Adding variants for invoking virsh_snapshot test.
Signed-off-by: NMiroslav Rezanina <mrezanin@redhat.com>

Testing snapshot handling procedure:

- create snapshots
- getting snapshot's information
- switching snapshots
- removing snapshots
Signed-off-by: NMiroslav Rezanina <mrezanin@redhat.com>

This patch adds following functions used for handling guest snapshots:

snapshot_create
snapshot_current
snapshot_list
snapshot_info
snapshot_revert
snapshot_delete

These functions require guest name and (if applicable) snapshot name.
Signed-off-by: NMiroslav Rezanina <mrezanin@redhat.com>

Autotest wait for guest shutdown at the end of unattended install. However,
when image is imported, guest is just setup to use existing image and booted.
There's no invocation of shutdown.

This patch add call for vm.shutdown in case of import method.
Signed-off-by: NMiroslav Rezanina <mrezanin@redhat.com>