Up until now, the default has been rtl8139, but no check was in
place to make sure that device was actually available.

Now we try rtl8139, e1000 and virtio-net in turn, checking for
availability before using any of them: this means we have a much
better chance for the guest to be able to boot.

This capability can be used to detect whether or not the QEMU
binary supports the virtio-net-* network device.

This capability can be used to detect whether or not the QEMU
binary supports the e1000 network device.

This capability can be used to detect whether or not the QEMU
binary supports the rtl8139 network device.

Commit f1f68ca3 did not report an error if virFileMakePath()
returned -1.  Well, who would've guessed function with name starting
with 'vir' sets an errno instead of reporting an error the libvirt way.
Anyway, let's fix it, so the output changes from:

  $ virsh start arm
  error: Failed to start domain arm
  error: An error occurred, but the cause is unknown

to:

  $ virsh start arm
  error: Failed to start domain arm
  error: Cannot create directory '/var/lib/libvirt/qemu/domain-arm': Not
  a directory

Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1146886Signed-off-by: NMartin Kletzander <mkletzan@redhat.com>

This reverts commit e5470dd0.

This has been ACK'd by the original author in the original mail thread:
https://www.redhat.com/archives/libvir-list/2015-September/msg00310.html

The reason to revert this is due to the patch breaking the generation of
internal subsites.  The original issue still needs to be dealt with,
though.
Signed-off-by: NMartin Kletzander <mkletzan@redhat.com>

If the current live definition does not have memory hotplug enabled, but
the persistent one does libvirt would reject migration if the
destination does not support memory hotplug even if the user didn't want
to persist the VM at the destination and thus the XML containing the
memory hotplug definition would not be used. To fix this corner case the
code will check for memory hotplug in the newDef only if
VIR_MIGRATE_PERSIST_DEST was used.

Double semicolons have special meaning in makefiles, but they would have
to be combined with other rules witch such separators in order to be
used as intended.  Since there are no other rules like that, let's
clean it up.
Signed-off-by: NMartin Kletzander <mkletzan@redhat.com>

Commit 35847860 Added the virFileUnlink function, but failed to add
a version for mingw build, causing the following error:

  Cannot export virFileUnlink: symbol not defined
Signed-off-by: NMartin Kletzander <mkletzan@redhat.com>

https://bugzilla.redhat.com/show_bug.cgi?id=1260846

Introduced by 8fedbbdb, if we parse an unordered NUMA cell, will
get a segfault. This is because of a check for overlapping @cpus
sets we have there. However, since the array to hold guest NUMA
cells is allocated upfront and therefore it contains all zeros,
an out of order cell will break our assumption that cell IDs have
increasing character. At this point we try to access yet NULL
bitmap and therefore segfault.
Signed-off-by: NLuyao Huang <lhuang@redhat.com>

Running valgrind on a very simplistic program consisting only of
opening and closing admin connection (virAdmConnect{Open,Close}) shows a
leak in remoteAdminPrivNew, because the last reference to privateData is
not decremented, thus the object won't be disposed. This patch unrefs
the privateData object once we closed the active connection to daemon,
making further use of this connection  useless.

==24577==    at 0x4A089C7: calloc (in /usr/lib64/valgrind/vgpreload_***linux.so)
==24577==    by 0x4E8835F: virAllocVar (viralloc.c:560)
==24577==    by 0x4EDFA5C: virObjectNew (virobject.c:193)
==24577==    by 0x4EDFBD4: virObjectLockableNew (virobject.c:219)
==24577==    by 0x4C14DAF: remoteAdminPrivNew (libvirt-admin.c:152)
==24577==    by 0x4C1537E: virAdmConnectOpen (libvirt-admin.c:308)
==24577==    by 0x400BAD: main (listservers.c:39)

==24577== LEAK SUMMARY:
==24577==    definitely lost: 80 bytes in 1 blocks
==24577==    indirectly lost: 840 bytes in 6 blocks
==24577==      possibly lost: 0 bytes in 0 blocks
==24577==    still reachable: 12,179 bytes in 199 blocks
==24577==         suppressed: 0 bytes in 0 blocks

https://bugzilla.redhat.com/show_bug.cgi?id=1172544

https://bugzilla.redhat.com/show_bug.cgi?id=1254420Signed-off-by: NJiri Denemark <jdenemar@redhat.com>

Let's move some variables from an inside loop to global function
declaration header block. It's going to be easier for next
patches. At the same time, order the cleanup calls at the
function's end so it's easier to track which variables are freed
and which not.
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>

In e755186c we tried to introduce an example demonstrating
new virDomainRename API. Unfortunately, in the .gitignore we had
a different binary listed. It's 'rename' binary which we want git
to ignore, not 'test'.
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>

Commit e755186c added the rename example, but forgot to build some
essential files in there as well as add it to the spec file.
Signed-off-by: NMartin Kletzander <mkletzan@redhat.com>

Remove the need for a couple of sa_asserts.

Coverity claims it could be possible to call virDBusTypeStackFree with
*stack == NULL and although the two API's that call it don't appear to
allow that - I suppose it's better to be safe than sorry

In virFileNBDDeviceFindUnused if virFileNBDDeviceIsBusy returns 0,
then both branches jumped to cleanup, so just use ignore_value
since the function returns NULL or some memory and the caller
handles the error.

Commit id '692e9fac' used virProcessSetNamespaces instead of inlining
the similar functionality; however, Coverity notes that the function
prototype expects a size_t value and not an enum and complains. So,
just typecast the enum as a size_t to avoid the noise.

Commit id '2e7cea24' added a check for an error from Finish instead
of 'unexpected error'; however, if for some reason there wasn't an
error, then virGetLastError could return NULL resulting in the
NULL pointer deref to err->domain.

Signed-off-by: NMartin Kletzander <mkletzan@redhat.com>

And in the middle it prints out its name to demonstrate changes in later
patch(es).
Signed-off-by: NMartin Kletzander <mkletzan@redhat.com>

Creating ACL rules is not exactly easy and existing examples are pretty
simple. This patch adds a somewhat complex example which defines several
roles. Admins can do everything, operators can do basic operations
on any domain and several groups of users who act as operators but only
on a limited set of domains.
Signed-off-by: NJiri Denemark <jdenemar@redhat.com>

https://bugzilla.redhat.com/show_bug.cgi?id=1258361

When attaching a disk, controller, or rng using an address type ccw
or s390, we need to ensure the support is provided by both the machine.os
and the emulator capabilities (corollary to unconditional setting when
address was not provided for the correct machine.os and emulator.

For an inactive guest, an addition followed by a start would cause the
startup to fail after qemu_command builds the command line and attempts
to start the guest. For an active guest, libvirtd would crash.

Rather than have different usages of STR function in order to determine
whether the domain is s390-ccw or s390-ccw-virtio, make a single API
which will check the machine.os prefix. Then use the function.

There's no reason why debug initialization could not be made completely
hidden, just like readline initialization is. The point of the global
initializer vshInit is to make initialization of smaller features transparent
to the user/caller.

Currently, we set interactive mode as default possibly reverting the
setting after we parse the command line arguments. There's nothing
particulary wrong with that, but a call to vshReadlineInit is performed
always in the global initializer just because the default mode is interactive.
Rather than moving vshReadlineInit call somewhere else (because another client
might want to implement interactive mode only), we could make the decision
if we're about to run in interactive mode once the command line is parsed.

Commit a0b6a36f separated vshInitDebug from the original vshInit
(before virsh got split and vshInit became virshInit - commit 834c5720)
in order to be able to debug command line parsing.
After the parsing is finished, debugging is reinitialized to work properly.
There might as well be other features that require re-initialization as
the command line could specify parameters that override our defaults which
had been set prior to calling vshArgvParse.

As part of the effort to stay consistent, change the vshInit signature
from returning int to returning bool. Moreover, remove the
unnecessary error label as there is no cleanup that would make use of
it.

Remove redundant error reporting in libxlDomainMigrationPerform().
virNetSocketNewConnectTCP() is perfectly capable of reporting
sensible errors.

Commit d091518b tried to escape all strings in produced XML, but missed
this one.
Signed-off-by: NPavel Hrdina <phrdina@redhat.com>

Before libvirt sets the MAC address of the physdev (the physical
ethernet device) linked to a macvtap passthrough device, it always
saves the previous MAC address to restore when the guest is finished
(following a "leave nothing behind" policy). For a long time it
accomplished the save/restore with a combination of
ioctl(SIOCGIFHWADDR) and ioctl(SIOCSIFHWADDR), but in commit cbfe38c
(first in libvirt 1.2.15) this was changed to use netlink RTM_GETLINK
and RTM_SETLINK commands sent to the Physical Function (PF) of any
device that was detected to be a Virtual Function (VF).

We later found out that this caused problems with any devices using
the Cisco enic driver (e.g. vmfex cards) because the enic driver
hasn't implemented the function that is called to gather the
information in the IFLA_VFINFO_LIST attribute of RTM_GETLINK
(ndo_get_vf_config() for those keeping score), so we would never get
back a useful response.

In an ideal world, all drivers would implement all functions, but it
turns out that in this case we can work around this omission without
any bad side effects - since all macvtap passthrough <interface>
definitions pointing to a physdev that uses the enic driver *must*
have a <virtualport type='802.1Qbh'>, and since no other type of
ethernet devices use 802.1Qbh, libvirt can change its behavior in this
case to use the old-style.  ioctl(SIOC[GS]IFHWADDR).  That's what this
patch does.

Resolves:  https://bugzilla.redhat.com/show_bug.cgi?id=1257004

These functions were made static as a part of commit cbfe38c since
they were no longer called from outside virnetdev.c. We once again
need to call them from another file, so this patch makes them once
again public.

Well, in 8ad126e6 we tried to fix a memory corruption problem.
However, the fix was not as good as it could be. I mean, the
commit has one line more than it should. I've noticed this output
just recently:

  # ./run valgrind --leak-check=full --show-reachable=yes ./tools/virsh domblklist gentoo
  ==17019== Memcheck, a memory error detector
  ==17019== Copyright (C) 2002-2013, and GNU GPL'd, by Julian Seward et al.
  ==17019== Using Valgrind-3.10.1 and LibVEX; rerun with -h for copyright info
  ==17019== Command: /home/zippy/work/libvirt/libvirt.git/tools/.libs/virsh domblklist gentoo
  ==17019==
  Target     Source
  ------------------------------------------------
  fda        /var/lib/libvirt/images/fd.img
  vda        /var/lib/libvirt/images/gentoo.qcow2
  hdc        /home/zippy/tmp/install-amd64-minimal-20150402.iso

  ==17019== Thread 2:
  ==17019== Invalid read of size 4
  ==17019==    at 0x4EFF5B4: virObjectUnref (virobject.c:258)
  ==17019==    by 0x5038CFF: remoteClientCloseFunc (remote_driver.c:552)
  ==17019==    by 0x5069D57: virNetClientCloseLocked (virnetclient.c:685)
  ==17019==    by 0x506C848: virNetClientIncomingEvent (virnetclient.c:1852)
  ==17019==    by 0x5082136: virNetSocketEventHandle (virnetsocket.c:1913)
  ==17019==    by 0x4ECD64E: virEventPollDispatchHandles (vireventpoll.c:509)
  ==17019==    by 0x4ECDE02: virEventPollRunOnce (vireventpoll.c:658)
  ==17019==    by 0x4ECBF00: virEventRunDefaultImpl (virevent.c:308)
  ==17019==    by 0x130386: vshEventLoop (vsh.c:1864)
  ==17019==    by 0x4F1EB07: virThreadHelper (virthread.c:206)
  ==17019==    by 0xA8462D3: start_thread (in /lib64/libpthread-2.20.so)
  ==17019==    by 0xAB441FC: clone (in /lib64/libc-2.20.so)
  ==17019==  Address 0x139023f4 is 4 bytes inside a block of size 240 free'd
  ==17019==    at 0x4C2B1F0: free (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
  ==17019==    by 0x4EA8949: virFree (viralloc.c:582)
  ==17019==    by 0x4EFF6D0: virObjectUnref (virobject.c:273)
  ==17019==    by 0x4FE74D6: virConnectClose (libvirt.c:1390)
  ==17019==    by 0x13342A: virshDeinit (virsh.c:406)
  ==17019==    by 0x134A37: main (virsh.c:950)

The problem is, when registering remoteClientCloseFunc(), it's
conn->closeCallback which is ref'd. But in the function itself
it's conn->closeCallback->conn what is unref'd. This is causing
imbalance in reference counting. Moreover, there's no need for
the remote driver to increase/decrease conn refcount since it's
not used anywhere. It's just merely passed to client registered
callback. And for that purpose it's correctly ref'd in
virConnectRegisterCloseCallback() and then unref'd in
virConnectUnregisterCloseCallback().
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>

Caller is responsible for freeing the result of virStringJoin()
when no longer needed:

==10701== 1 bytes in 1 blocks are definitely lost in loss record 1 of 806
==10701==    at 0x4C29F80: malloc (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
==10701==    by 0xAADB679: strdup (in /lib64/libc-2.20.so)
==10701==    by 0x4F18655: virStrdup (virstring.c:726)
==10701==    by 0x4F175AF: virStringJoin (virstring.c:165)
==10701==    by 0x131D4D: vshReadlineInit (vsh.c:2572)
==10701==    by 0x1322DF: vshInit (vsh.c:2736)
==10701==    by 0x1347C1: main (virsh.c:907)
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>

Commit id '155ca616' added the 'refreshVol' API. In an NFS root-squash
environment it was possible that if the just created volume from XML wasn't
properly created with the right uid/gid and/or mode, then the followup
refreshVol will fail to open the volume in order to get the allocation/
capacity values. This would leave the volume still on the server and
cause a libvirtd crash because 'voldef' would be in the pool list, but
the cleanup code would free it.

Commit id '7c2d65dd' changed the default value of mode to be -1 if not
supplied in the XML, which should cause creation of the volume using the
default mode of VIR_STORAGE_DEFAULT_VOL_PERM_MODE; however, the check
made was whether mode was '0' or not to use default or provided value.

This patch fixes the issue to check if the 'mode' was provided in the XML
and use that value.