Before upstream qemu commit

  commit 56c4bfb3f07f3107894c00281276aea4f5e8834d
  Author: Laszlo Ersek <lersek@redhat.com>
  Date:   Tue Aug 6 12:37:11 2013 +0200

      dump: rebase from host-private RAMBlock offsets to guest-physical
      addresses

the dump-guest-memory QMP command used to write incorrect vmcores for
x86_64 guests with more than 3.5GB RAM. Add a testcase that creates a 4GB
guest, dumps its core, copies it into the guest, and analyzes it there
with the "crash" utility.

Related RHBZ: https://bugzilla.redhat.com/show_bug.cgi?id=981582
Related RHBZ: https://bugzilla.redhat.com/show_bug.cgi?id=990118Suggested-by: NAdemar de Souza Reis Jr. <areis@redhat.com>
Signed-off-by: NLaszlo Ersek <lersek@redhat.com>

Signed-off-by: NYanbing Du <ydu@redhat.com>

Signed-off-by: NYanbing Du <ydu@redhat.com>

Enhance pci_test_cmd for block_hotplug testing 

Add rescan and list disk steps; Do rescan to ensure diskpart read latest disk
configuration and it will avoid diskpart report
'The disk you specified is not valid.' error, list disk step will record disk
info before check disk drive;
New steps will helpful for debugging when test failed at pci device test step;
Signed-off-by: NXu Tian <xutian@redhat.com>

Small fix for unattended_install.floppy_ks test

Reading ks from floppy driectly isn't works for some distr but boot option
'ks=hd:/dev/fd0' can reading ks from mounted floppy; so skip replace
this options;
Signed-off-by: NXu Tian <xutian@redhat.com>

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

* Simple/minimal implementation with passwd property, more can be added
  later.
* Thanks to Kyla - Wei Zhang <weizhan@redhat.com> for requesting and
  helping test.
* Added unittest for graphics device to verify changes are commited to
  vm_xml instance.
Signed-off-by: NChris Evich <cevich@redhat.com>

Readonly mode is an important part for virsh command,so add
a boolean parameter 'readonly' when execute virsh commnd.
Signed-off-by: NYanbing Du <ydu@redhat.com>

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

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

sr-iov:  Update PciAssignable class to support more feature.

For pci-assign we need not get all the same group device.
Signed-off-by: NFeng Yang <fyang@redhat.com>

virttest.test_setup: Fix issue that 'pci-assign' is treated as not support in some host which really support it.

In some host pci-assign is aliased as kvm-pci-assign in qemu-kvm -device ?
So pci-assign is treated as not support. Add code to fix this.
Signed-off-by: NFeng Yang <fyang@redhat.com>

It used to set which device we want to use in device assignment test
Signed-off-by: NFeng Yang <fyang@redhat.com>

Also add new function get_same_group_devs() to get devices in same
group.

Get device's driver from /sys/bus/pci/devices/[device id]/driver
Signed-off-by: NFeng Yang <fyang@redhat.com>

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

In this patch have following update:
1. Delete self.type_list and introduce self.devices.
2. Update add_device() to support mac setting for VF.
3. Update get_devs() to use new get_vf_devs() and get_pf_devs().
4. Update kvm parameter by echo to /sys folder.
5. Support set mac for VF.
6. Update the way used to update kvm module parameter.
Signed-off-by: NFeng Yang <fyang@redhat.com>

get requested vf or pf from self.pf_vf_info list.
Signed-off-by: NFeng Yang <fyang@redhat.com>

Get the PF PCI ID from self.pf_vf_info.
Return the first available PF if name is not set.
Signed-off-by: NFeng Yang <fyang@redhat.com>

This function will return corresponding pf eth name and vf num according to vf id.
We need use pf eth name and vf num when setting mac for vf.
Signed-off-by: NFeng Yang <fyang@redhat.com>

get_pf_vf_info() will get following pf information:
    pf_id: The id of the pf device.
    occupied: Whether the pf device assigned or not
    vf_ids: vf device id list of this pf.
    ethname: network device interface name in host for this pf.
Signed-off-by: NFeng Yang <fyang@redhat.com>

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

We have update vm.pci_assignable.request_devs() function call.

Now we need create device dict and use it in request_devs().

We need set device["type"], device['mac'] (for vf), device["name"] (optional)
Signed-off-by: NFeng Yang <fyang@redhat.com>

We have update vm.pci_assignable.request_devs() function call.

Now we need create device dict and use it in request_devs().

We need set device["type"], device['mac'] (for vf), device["name"] (optional)
Signed-off-by: NFeng Yang <fyang@redhat.com>

Now we need set mac for vf before using it. Or guest could not get IP.
Signed-off-by: NFeng Yang <fyang@redhat.com>

This patch will fix following issue:

"qemu: -device pci-assign,host=06:00.1,id=id_06:00.1:
 Parameter 'id' expects an identifier\n"

Also correct parameter in add_pcidevice function call.
Signed-off-by: NFeng Yang <fyang@redhat.com>

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

Run CrystalDiskMark in guest during s3 and s4.
Signed-off-by: NYiqiao Pu <ypu@redhat.com>

Add cmd /c to let the cmd session wait util the autoit scripts finished.
Otherwise it will return immediately after the autoit scripts start.
Also correct the script name.
Signed-off-by: NYiqiao Pu <ypu@redhat.com>

Signed-off-by: NXu Tian <xutian@redhat.com>

virttest.utils_misc: Convert host_setup_flag into int

Fix some extra problems with docstrings in the code tree.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

This file is also borrowed from autotest, and already
reflects the updates to the coding style of the projects
under the autotest umbrella.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

This way we can keep those scripts in sync among
virt-test and autotest.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Make virt-test to have a docstring style mostly compliant with
the sphinx's [1] format, reStructured text. This means that
we have to trade all '@' tags to be ':', mostly.
Signed-off-by: NLucas Meneghel Rodrigues <lmr@redhat.com>

Options of run_pep8 tool is changed, then the description
should be updated at the same time.
Signed-off-by: NDongsheng Yang <yangds.fnst@cn.fujitsu.com>

It's possible to install libvirt bits w/o installing kvm or xen.  This
will be reflected in 'virsh capabilities' output, but the VM class just
reports a key error.  Added two specific error logging messages to make
the problem clearer, but continue to raise the original exception.
Signed-off-by: NChris Evich <cevich@redhat.com>

virttest.libvirt_storage_unittest: remove ModuleLoadCheckVirsh class.