The functions in iptables.c all return -1 on failure, but all their
callers (which all happen to be in bridge_driver.c) assume that they
are returning an errno, and the logging is done accordingly. This
patch fixes all the error checking and logging to assume < 0 is an
error, and nothing else.

Later patches will add the possibility to define a network's netmask
as a prefix (0-32, or 0-128 in the case of IPv6). To make it easier to
deal with definition of both kinds (prefix or netmask), add two new
functions:

virNetworkDefNetmask: return a copy of the netmask into a
virSocketAddr. If no netmask was specified in the XML, create a
default netmask based on the network class of the virNetworkDef's IP
address.

virNetworkDefPrefix: return the netmask as numeric prefix (or the
default prefix for the network class of the virNetworkDef's IP
address, if no netmask was specified in the XML)

virSocketPrefixToNetmask: Given a 'prefix', which is the number of 1
bits in a netmask, fill in a virSocketAddr object with a netmask as an
IP address (IPv6 or IPv4).

virSocketAddrMask: Mask off the host bits in one virSocketAddr
according to the netmask in another virSocketAddr.

virSocketAddrMaskByPrefix, Mask off the host bits in a virSocketAddr
according to a prefix (number of 1 bits in netmask).

VIR_SOCKET_FAMILY: return the family of a virSocketAddr

When I run configure, I receive some syntax error.

.....
checking where to find <rpc/rpc.h>... none
checking for library containing dlopen... -ldl
/configure: line 52500: test: =: unary operator expected
/configure: line 52766: test: =: unary operator expected
checking linux/kvm.h usability... yes
checking linux/kvm.h presence... yes
.....
Signed-off-by: NWen Congyang <wency@cn.fujitsu.com>

By adding it to sitemap.html.in. Also <ul> can't be nested in <p>.

Shorten qemuDomainSnapshotWriteSnapshotMetadata function name
and make it take a snapshot pointer instead of dealing with
the current snapshot. Update other functions accordingly.

Add a qemuDomainSnapshotReparentChildren hash iterator to
reparent the children of a snapshot that is being deleted. Use
qemuDomainSnapshotWriteMetadata to write updated metadata
to disk.

This fixes a problem where outdated parent information breaks
the snapshot tree and hinders the deletion of child snapshots.

Reported by Philipp Hahn.

I began noticing a race when reserving VNC ports as described here

https://www.redhat.com/archives/libvir-list/2010-November/msg00379.html

Turns out that we were not initializing the size field of bitmap
struct when allocating the bitmap.  This subsequently caused
virBitmapSetBit() to fail since bitmap->size is 0, hence we never
actually reserved the port.

Commit ed0d9f6c added support for
automatic port allocation for SPICE but forgot to mark such ports as
unused when they are not used anymore.

Fix glitch in commit cddd2a06 (thankfully post-0.8.6, so no
released version has the glitch).

Document and try to workaround glitch in commit 46e9b0fb (in 0.8.0),
which invalidated 6 virErrorNumber values dating as far back as 0.7.1.

My audit did not find any other glitches until pre-0.1.0 days.  I'm
not sure how to add a syntax-check off the top of my head, but
hopefully the explicit numbering will make people think twice about
renumbering in the future.

* include/libvirt/virterror.h (virErrorDomain): Avoid inserting
new values in the middle, and add explicit numbering to help avoid
this in the future.
(virErrorNumber): Add explicit numbering, and document the snafu.
* src/remote/remote_driver.c (remoteIO): Compensate for the snafu.

This fixes the build from a tarball and makes autobuild.sh
work again.

This should actually have been part of this earlier commit:

  esx: Move VMX handling code out of the driver directory
  42b2f35d

Reported by Eric Blake.

All other drivers are explicitly linked to gnulib. The VMware
driver lacked this, resulting in mdir_name being an undefine
symbol.

Explicitly link the VMware driver to gnulib to fix this.

Now the VMware driver doesn't depend on the ESX driver anymore.

Add a WITH_VMX option that depends on WITH_ESX and WITH_VMWARE.
Also add a libvirt_vmx.syms file.

Move some escaping functions from esx_util.c to vmx.c.

Adapt the test suite, ESX and VMware driver to the new code layout.

Connecting to a ESX(i) server that is part of a cluster failed
when the connection also involved a vCenter.

Accept ClusterComputeResource type in addition to ComputeResource
type in the object lookup function.

Reported by Guillaume Le Louët.

* daemon/Makefile.am: Avoid spurious space before tabs.
* src/Makefile.am: Likewise.
* examples/dominfo/Makefile.am: Likewise.
* examples/domsuspend/Makefile.am: Likewise.
* tools/Makefile.am: Likewise.
* src/datatypes.h (VIR_CONNECT_MAGIC): Likewise.
* src/internal.h (TODO): Likewise.
* src/qemu/qemu_monitor.h (QEMU_MONITOR_MIGRATE): Likewise.
* src/xen/xen_hypervisor.c (XEN_V2_OP_GETAVAILHEAP): Likewise.
* src/xen/xs_internal.h: Likewise.

* src/util/command.c (virCommandRun): Don't capture output on
daemons.
* tests/commandtest.c (test18): Expose the bug.
Reported by Laine Stump.

* libvirt.spec.in: Provide vmware conditionals.

If there is a dangling symbolic link in filesystem pool, the pool
will fail to start or refresh, this patch is to fix it by ignoring
it with a warning log.

Network disks are accessed by qemu directly, and have no
associated file on the host, so checking for file ownership etc.
is unnecessary.
Signed-off-by: NJosh Durgin <joshd@hq.newdream.net>

Also added an additional menu placement for the windows page, in
order to attract further potential testers.

* configure.ac (dlopen): Cygwin dlopen is in libc; avoid spurious
failure.
(XDR_CFLAGS): Define when needed.
* src/Makefile.am (libvirt_driver_remote_la_CFLAGS): Use it.

* src/qemu/qemu_driver.c (qemudLogFD, qemudLogReadFD)
(qemudStartup, qemudGetProcessInfo): Use heap instead of stack.

When running 'make check' under a multi-cpu Dom0 xen machine,
nodeinfotest had a spurious failure it was reading from
/sys/devices/system/cpu, but xen has no notion of topology.  The test
was intended to be isolated from reading any real system files; the
regression was introduced in Mar 2010 with commit aa2f6f96.

Fix things by allowing an early exit for the testsuite.

* src/nodeinfo.c (linuxNodeInfoCPUPopulate): Add parameter.
(nodeGetInfo): Adjust caller.
* tests/nodeinfotest.c (linuxTestCompareFiles): Likewise.

I got some spurious failures when commandhelper won the race and
ran to the point of parent detection prior to the intermediate
daemonizing process getting a chance to exit.  This fixes it.

* tests/commandhelper.c (main): Checking for re-parenting to
init(1) is racy; instead check that we belong to a new session.

* tests/Makefile.am (TESTS_ENVIRONMENT, commandtest_CFLAGS)
(commandhelper_CFLAGS): Avoid $(builddir) and $(abs_builddir) in
automake 1.9.6; fixes spurious failures of commandtest.

* configure.ac (with_selinux): Check for <selinux/label.h>.
* src/security/security_selinux.c (getContext): New function.
(SELinuxRestoreSecurityFileLabel): Use it to restore compilation
when using older libselinux.

Autoconf 2.59 doesn't define ${localedir}, so libvirt was failing
to compile due to a missing LOCALEDIR until today's configmake fix.

* .gnulib: Update to latest for configmake fix.
* configure.ac (libpcap): Avoid AS_CASE.

* src/Makefile.am (VMWARE_DRIVER_SOURCES): Fix vmware_driver.h.
Signed-off-by: NWen Congyang <wency@cn.fujitsu.com>

* tools/libvirt-guests.init.in (libvirtd): Skip function library
if it is not present.
Suggested by Guido Günther.

While not technically a double free (since VIR_FREE NULLs the
pointer), this is unnecessary extra code.

This crept in when the function was converted from virRun to virCommand.

The AUTHORS file has also been updated.

formatnetwork.html has a menu item at level 3. libvirt.css
doesn't have a explicit rule for level 3 and level 3 and
level 2 items end up at the same indentation level.

Add an additional 1em indentation to level 3 menu items.

XPCOM returns an array as a pointer to an array of pointers to the
actual items. When the array isn't needed anymore the items are
released, but the actual array containing the pointers to the items
was not freed and leaked.

Free the actual array using ComUnallocMem.

This doesn't affect MSCOM as SafeArrayDestroy releases all items
and frees the array.

Add a vboxArray to hide the details from the general driver code.

Don't require dlopen, but link to ole32 and oleaut32 on Windows.

Don't expose g_pVBoxFuncs anymore. It was only used to get the
version of the API. Make VBoxCGlueInit return the version instead.
This simplifies the implementation of the MSCOM glue layer.

Get the VirtualBox version from the registry.

Add a dummy implementation of the nsIEventQueue to the MSCOM glue
as there seems to be no direct equivalent with MSCOM. It might be
implemented using the normal window message loop. This requires
additional investigation.