1. add support for python3.6
2. decode process output to text
3. update linux netperf to latest 2.7.1
4. re-design the logic for getting server_ip/server_ctl/server_ctl_ip
Signed-off-by: NWenli Quan <wquan@redhat.com>

add different block size, which will execute multiple dd test
based on the block sizes provided.
1. use extra parameter to specify partition size.
2. add readonly=off.
3. add dd I/O flag: direct.
Signed-off-by: Nlolyu <lolyu@redhat.com>

1. add function 'is_mounted_dir_acessible' to ensure nfs
mounted dir is resumed before cont vm.

2. refactor original code of iptables rule generation.

3. refactor space check code.

4. add oflag=direct for dd command.
Signed-off-by: Nlolyu <lolyu@redhat.com>

Stress app should not be killed without stress_duration defined
Signed-off-by: NQianqian Zhu <qizhu@redhat.com>

1. virtual-host tuned profile on server and client
2. virtual-guest tuned profile on vm
3. disable watchdog
4. hyper-v enlightenment for windows guest with -cpu ,hv_time,
hv_relaxed,hv_vapic,hv_spinlocks=0x1fff
Signed-off-by: NWenli Quan <wquan@redhat.com>

Signed-off-by: NYumei Huang <yuhuang@redhat.com>

Signed-off-by: NSitong Liu <siliu@redhat.com>

Since we will drop using ntpdate to get local time, so drop ntp
variant as well.
Signed-off-by: NSitong Liu <siliu@redhat.com>

1. Let load run on different cpus.
2. Add instance setting for default load scenario and full load scenario.
Signed-off-by: NSitong Liu <siliu@redhat.com>

The test would call `pxe_boot` and run parallelly, and that would
result in error which was caused by race conditions. The patch
attends to fix this issue by making the script indenpendent.
Signed-off-by: NXu Han <xuhan@redhat.com>

Existing linux stress was supporting single vm with single stress
tool. This commit extends that testcase to now support multiple
guests running same/different stress tools inside for given duration
Signed-off-by: NSrikanth Aithal <sraithal@linux.vnet.ibm.com>

hwclock test fails on JeOS, Fedora etc with default pattern,
make an ability to define custom date pattern for other OSes
Signed-off-by: NSergey Bronnikov <sergeyb@bronevichok.ru>

yum_update.py had support to perform os update for RHEL based guests,
refactor it to support Ubuntu and adopt MultiVM. The patch also renames
yum_update.py -> os_update.py similarly cfg file is renamed.
Signed-off-by: NBalamuruhan S <bala24@linux.vnet.ibm.com>

With the py3k changes some returns are not strings, but decoded
unicodes. Currently this mainly breaks monitor returns, where human
monitor gets recognized as qmp one. Let's use "six.string_types" that
provides the right string types for given python version.
Signed-off-by: NLukáš Doktor <ldoktor@redhat.com>

Signed-off-by: NSitong Liu <siliu@redhat.com>

Add additional explicit decode from bytes to str for python3
Signed-off-by: Nybduan <yduan@redhat.com>

Due to recent change in utils_test relating to VMStress, couple of qemu tests
were failing. This patch fixes up 2 qemu testcases.
Signed-off-by: NSrikanth Aithal <sraithal@linux.vnet.ibm.com>

By default kdump test sets crashkernel reserve memory with offset of 16M. In powerpc, offest is auto calculated, setting explicit offset will result in crash kernel failing to reserve memory leading to kdump service start failure. We need to make def_kernel_param_cmd generic by omitting offset (kdump will offset the reserved memory automatically) which works across archs.
Signed-off-by: NSrikanth Aithal <sraithal@linux.vnet.ibm.com>

Signed-off-by: Nqizhu <qizhu@redhat.com>

Signed-off-by: Nqizhu <qizhu@redhat.com>

Signed-off-by: NXueqiang Wei <xuwei@redhat.com>

Variants jumbo is duplicated with jumbo.py, so replace it with
with_jumbo.
Signed-off-by: NWenli Quan <wquan@redhat.com>

This determines the pass/fail of a test based on the execution
of a command.  This is the predecessor of Avocado SIMPLE tests,
and as such, can be removed in favor of the former.
Signed-off-by: NCleber Rosa <crosa@redhat.com>

The test's "srcdir" property is deprecated and is planned to be
removed no later than May 11 2018. So replaced test.srcdir with
test.workdir.
Also changed variable name accordingly.

Please refer to:
https://github.com/avocado-framework/avocado/pull/2509Signed-off-by: NHaotong Chen <hachen@redhat.com>

Signed-off-by: NHaotong Chen <hachen@redhat.com>

Signed-off-by: NHaotong Chen <hachen@redhat.com>

Python 3 dese not support foo.fucn_name,
using foo.__name__ instead.
Signed-off-by: NHaotong Chen <hachen@redhat.com>

Signed-off-by: NHaotong Chen <hachen@redhat.com>

Signed-off-by: NHaotong Chen <hachen@redhat.com>

Correct the parameter name.
Signed-off-by: NXu Han <xuhan@redhat.com>

Signed-off-by: NHaotong Chen <hachen@redhat.com>

Signed-off-by: NXu Han <xuhan@redhat.com>

Signed-off-by: NHaotong Chen <hachen@redhat.com>

Signed-off-by: NHaotong Chen <hachen@redhat.com>

Signed-off-by: NHaotong Chen <hachen@redhat.com>

Signed-off-by: NHaotong Chen <hachen@redhat.com>

This patch removes autotest dependency and uses
APIs from virttest
Signed-off-by: NMallesh Koti <mallesh@linux.vnet.ibm.com>

1. in RHEL guest with virtio1.0, when set ctrl_mac_addr=off, mac
can be changed to new one in guest but display the old one while
'info network'
2. in RHEL guest with virtio1.0, when set ctrl_mac_addr=on,
mac can be changed to new ones both in guest and 'info
network'
Signed-off-by: NXiyue Wang <xiywang@redhat.com>

1 Add session return for windrv_check_running_verifier
2 Modify all the related scripts
Signed-off-by: NYu Wang <wyu@redhat.com>

1 Add jumbo setting for macvtap devices
2 Use second ethernet for login guest since host could not
communicate guest through first one which is attached from macvtap
3 fix no attribute 'stop_tcpdump' error
Signed-off-by: NWenli Quan <wquan@redhat.com>