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>

Changes from V1:
 * Disable RHEL.5.i386 guest for this test since there
   is no iasl package for this platform.
 * Update the cmd for RHEL/Fedora guests separately.
Signed-off-by: NQingtang Zhou <qzhou@redhat.com>

Signed-off-by: NQingtang Zhou <qzhou@redhat.com>

Signed-off-by: NQingtang Zhou <qzhou@redhat.com>

These 2 files (guest_s3.cfg and guest_s4.cfg) are out of
date and should not be in here.
Signed-off-by: NQingtang Zhou <qzhou@redhat.com>

ppc64 platform does not support isa serial, so disable this.
Signed-off-by: NMike Qiu <qiudayu@linux.vnet.ibm.com>

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

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

new test to verify image file no backingfile after stream done;

changes from v1:
    update test steps in doc string;
Signed-off-by: NXu Tian <xutian@redhat.com>

raise TestFail doesn't make sense in get_backingfile function, it should
be verify in test script according case;
Signed-off-by: NXu Tian <xutian@redhat.com>

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

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

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

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

Accessing any vm_xml.devices member requiring an undefined
device xml module resulted in an extremely unhelpful error:

AttributeError: "type object 'exceptions.NotImplementedError' has no
attribute 'new_from_element'"

This commit adds a stub-class generating metaclass for undefined
devices.  The generated stub-class provides the minimal required
interface for the higher-level vm_xml.VMXMLDevices class.  Each
stub class will raise one-warning upon first use.  This is to notify
and make clear to all uses of these devices, that the interface
will likely change in the future.

Added unittests to verify functionality of stub-class metaclass.  Also
added plenty of docstrings and docstring updates.
Signed-off-by: NChris Evich <cevich@redhat.com>

* added del_seclabel accessor method so we have a complete set.

* get_seclabel() accessor method should raise exception
  for non-existant value instead of returning None value.

* set_seclabel() accessor method should not set unspecified
  imput attribute values.

* set_seclabel() should not perform non-data manipulation actions
  (self.define/undefine) since can introduce hidden side-effects.

* Made libvirt_xml_unittests fail tests for any subject-call
  which uses undefined dummy_virsh commands.

* Extended seclabel unittests for multi-mode access and
  exception checking.
Signed-off-by: NChris Evich <cevich@redhat.com>

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

This patch add the restart vm function.

Changes from v2:
update some parameter name.
Signed-off-by: NShuping Cui <scui@redhat.com>
Signed-off-by: NYiqiao Pu <ypu@redhat.com>
Acked-by: NFeng Yang <fyang@redhat.com>

Check some advance commands not include qemu-kvm package

changes from v1:
    1. instead private code with function in framework
    2. fixed typo
    3. calll monitor.verify_supported_cmds driectly to check commands
    unavaliable
changes from v2:
    1. add notes in docstring
    2. instead filter function
Signed-off-by: NXu Tian <xutian@redhat.com>

autotest.scrashme test is duplicate with autotest.flail, remove
autotest.scrashme test but no autotest.flail test because flail test's
"syscall_table" newer than scrashme test used;

Old syscalls that won't be tested anymore:

newstat
newfstat
newlstat
ni_syscall (uselib)
sysctl
umount
ni_syscall (create_module)
ni_syscall (get_kernel_syms)
ni_syscall (query_module)
ni_syscall (getpmsg)
ni_syscall (putpmsg)
ni_syscall (afs)
ni_syscall (tux)
ni_syscall (security)
ni_syscall (set_thread_area)
ni_syscall (get_thread_area)
ni_syscall (epoll_ctl_old)
ni_syscall (epoll_wait_old)
clock_retres
ni_syscall (vserver)
fstatat
timerfd
rt_tgsigqueueinfo
perf_counter_open

New syscalls that will be tested:

stat
fstat
lstat
uselib
_sysctl
umount2
create_module
get_kernel_syms
query_module
getpmsg
putpmsg
afs_syscall
tuxcall
security
set_thread_area
get_thread_area
epoll_ctl_old
epoll_wait_old
clock_getres
vserver
newfstatat
timerfd_create
Signed-off-by: NXu Tian <xutian@redhat.com>

Mainly test steps:
1. Check if the NIC number in guest matches that specified in config file.
2. Get the ip-mac pair in guest.
3. Gather the ping info by tcpdump.
4. Check if the ip-mac info matches in the tcpdump output.
5. Check if the ping count matches the given number.
Signed-off-by: NBear <byang@redhat.com>
Signed-off-by: NKen Cao <kcao@redhat.com>
Signed-off-by: NYolkfull Chow <yzhou@redhat.com>
Signed-off-by: NFeng Yang <fyang@redhat.com>
KVM-test: Multi nic test case improve
Signed-off-by: NYiqiao Pu <ypu@redhat.com>
KVM Test: multi_nics: Fix argument order for scp_between_remotes
Signed-off-by: NQingtang Zhou <qzhou@redhat.com>
KVM Test: Refactor multi_nics.py test script
Rename multi_nics.py to multi_vms_nics.py
Support multi vms and multi nics.
Signed-off-by: NFeng Yang <fyang@redhat.com>
qemu: Fix syntax error in multi_nics.py
Signed-off-by: NFeng Yang <fyang@redhat.com>

The nic driver directory is different for rhel4.9 guest
so make it configurable and update it to correct one.
Signed-off-by: NGolita Yue <gyue@redhat.com>
Acked-by: NFeng Yang <fyang@redhat.com>

Signed-off-by: NFeng Yang <fyang@redhat.com>
KVM-Test: Boot/shutdown rebased image
Signed-off-by: NSuqin Huang <shuang@redhat.com>
KVM Test: check the image after image creation.
KVM test: qemu_img: Pre-allocated metadata qcow2 format support
KVM Test: Raw device support for qemu_img test
KVM test: qemu_img add the qcow2 over LVM checking
Signed-off-by: NJason Wang <jasowang@redhat.com>
KVM Test: Add cluster_size_2048 test case.
Bug fix and add error.context in qemu_img.py script
Fix bug that guest start twice.
Signed-off-by: NFeng Yang <fyang@redhat.com>
KVM-Test: qemu_img.py: update case with new autotest api
virt: Update 'guest_name' to 'os_type' in all scripts
Signed-off-by: NQingtang Zhou <qzhou@redhat.com>
KVM-test: Update qemu_img commit test command to support Windows
Signed-off-by: NYiqiao Pu <ypu@redhat.com>

Add two parameters to let test case timedift.with_load
can be configured. When need run long time testing, can
reset the value "test_duration" & "interval_gettime".
Signed-off-by: NGolita Yue <gyue@redhat.com>

changes from v1:
 - add boot options for remove.

changes from v2:
 - also fix guest kernel cmdline in other timedrift testcases.

changes from v3:
 - add a judgement for boot options, don't run this step if there
   is nothing to be add or remove
Signed-off-by: NQingtang Zhou <qzhou@redhat.com>
Signed-off-by: NAmos Kong <akong@redhat.com>
Signed-off-by: NYiqiao Pu <ypu@redhat.com>
Acked-by: NFeng Yang <fyang@redhat.com>
Acked-by: Suqin Huang<shuang@redhat.com>

Signed-off-by: NQingtang Zhou <qzhou@redhat.com>
Acked-by: Feng Yang<fyang@redhat.com>
Acked-by: NYiqiao Pu <ypu@redhat.com>

Add timedrift with cpu online/offline test case.
 Check the guest time after offline the cpu
 Check the guest time after online the cpu

change from v1:
1. fix typo
2. add more error.context for other steps
Signed-off-by: NGolita Yue <gyue@redhat.com>

Add new case timedrift_monotonicity.py
and check the guest time monotonicity after migration.
There are two case:
1. Take the guest time during migration and check time.
2. Let guest running long time after migration and
   check the time.

change from v1:
1. delete guest script, add function to get time in test case
2. fix typo

change from v2:
1. add cmd_timeout
2. add parameter for function get_time()
Signed-off-by: NGolita Yue <gyue@redhat.com>
Acked-by: NQingtang Zhou <qzhou@redhat.com>
Acked-by: NYiqiao Pu <ypu@redhat.com>

Add stress test for multi vms with kernel compile running
inside vm.
1. boot up vms
every VM has 4G vmem, the total vmem of VMs' are
$overcommit times as host's mem.(default overcommit=1.5)
2) Launch kernel compile inside every guest.

Changes from v1:
 - modify the parameter "i" to "vmnum"
 - remove "raise error" when kernel compile failed, just return error
to avoid the long message cause virtlab crash
 - put kernel path to virtlab cfg files
 - change the test failed check to ping test, and judge if the guest
can be reach after kernel compile.
Signed-off-by: NGolita Yue <gyue@redhat.com>
Signed-off-by: NYiqiao Pu <ypu@redhat.com>
Acked-by: NAmos Kong <akong@redhat.com>

Bin folder has been removed and most file in it have been moved to
client folder.
Signed-off-by: NFeng Yang <fyang@redhat.com>

This class will setup and clean KSM config in different hosts.
You can use setup_ksm to configure it. There are several params
that can be used for ksm configure:

ksm_pages_to_scan: pages ksmd will scan in one loop
ksm_sleep_ms: million secs ksmd will sleep between two scan
ksm_run: start ksmd or not
ksm_module: if ksm is not compiled into kerenl need this parameter
            tell scripts which module should it load
disable_ksmtuned: if we need idsable ksmtuned during our test

This class also can be used to disable KSM in some special case with
set setup_ksm to yes and ksm_run to 0 or no.
Signed-off-by: NYiqiao Pu <ypu@redhat.com>

Add a case for block resize. Use monitor command to resize the block
device. And check it inside guest. This case support Windows and
Linux, qcow2 and raw, virtio_blk, virtio_scsi and ide with different
parameters.

* indentation fixed
Signed-off-by: NYiqiao Pu <ypu@redhat.com>
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

qemu: Make usb_hotplug test work on Win guests

Signed-off-by: NFeng Yang <fyang@redhat.com>