Don't use state-matching in a drop rule.

This also fixes a portability problem with the %a format modifier.
%a is not portable and made esxDomainDumpXML fail at runtime in
MinGW builds.

strtok_r will be used in the ESX driver to replace scanf-based code.

MinGW lacks strtok_r, so we need gnulib to provide it, but until now
strtok_r was licensed LGPL3.

This stops libvirt trying to connect to a non-existing libvirtd on the
ESX server in order to find a nwfilter driver.

* src/remote/remote_driver.c (remoteIO, remoteIOEventLoop): Report
failures on pipe used for wakeup.
Reported by Chris Lalancette.

Based on a warning from coverity.  The safe* functions
guarantee complete transactions on success, but don't guarantee
freedom from failure.

* src/util/util.h (saferead, safewrite, safezero): Add
ATTRIBUTE_RETURN_CHECK.
* src/remote/remote_driver.c (remoteIO, remoteIOEventLoop): Ignore
some failures.
(remoteIOReadBuffer): Adjust error messages on read failure.
* daemon/event.c (virEventHandleWakeup): Ignore read failure.

Regression introduced in commit 62170b99.

* src/util/memory.h: Placate cppi, and fit 80 columns.

Disk devices in QEMU have two parts, the guest device and the host
backend driver. Historically these two parts have had the same
"unique" name. With the switch to using -device though, they now
have separate names. Thus when changing CDROM media, for guests
using -device syntax, we need to prepend the QEMU_DRIVE_HOST_PREFIX
constant

* src/qemu/qemu_conf.c, src/qemu/qemu_conf.h: Add helper function
  qemuDeviceDriveHostAlias() for building a host backend alias
* src/qemu/qemu_driver.c: Use qemuDeviceDriveHostAlias() to determine
  the host backend alias for performing eject/change commands in the
  monitor

The device_add command was added in JSON mode in a way I didn't
expect. Instead of passing the normal device string to the JSON
command:

    { "execute": "device_add", "arguments": { "device": "ne2k_pci,id=nic.1,netdev=net.1" } }

We need to split up the device string into a full JSON object

    { "execute": "device_add", "arguments": { "driver": "ne2k_pci", "id": "nic.1", "netdev": "net.1" } }

* src/qemu/qemu_conf.h, src/qemu/qemu_conf.c: Rename the
  qemuCommandLineParseKeywords method to qemuParseKeywords
  and export it to monitor
* src/qemu/qemu_monitor_json.c: Split up device string into
  a JSON object for device_add command

The parameter for the qemuMonitorDeviceDel() is a device alias,
not a device config string. Rename the parameter reflect this
and avoid confusion to readers.

* src/qemu/qemu_monitor.c, src/qemu/qemu_monitor.h,
  src/qemu/qemu_monitor_json.c, src/qemu/qemu_monitor_json.h,
  src/qemu/qemu_monitor_text.c, src/qemu/qemu_monitor_text.h:
  Rename devicestr to devalias in qemuMonitorDeviceDel()

The QEMU developers have stated that they will not be porting
the commands 'pci_add', 'pci_del', 'usb_add', 'usb_del' to the
JSON mode monitor, since they're obsoleted by 'device_add'
and 'device_del'. libvirt has (untested) code that would have
supported those commands in theory, but since we already use
device_add/del where available, there's no need to keep the
legacy stuff anymore.

The text mode monitor keeps support for all commands for sake
of historical compatability.

* src/qemu/qemu_monitor_json.c: Remove 'pci_add', 'pci_del',
  'usb_add', 'usb_del' commands

To ensure that patches in the RPM don't break any functionality
it is neccessary to run the test suites during build. It currently
has 3 tests disabled

 - daemon-conf: this is totally broken, since it relies on
   being able to resolve the 'libvirt' group & being able to
   resolve hostnames at daemon startup. This isn't possible
   in a mock build root
 - seclabeltest: fails to initialize selinux in the mock
   build root. Possibly fixable
 - nodeinfotest: broken on s390 + ppc - this is a real bug

* libvirt.spec.in: Add a %check section, with 3 tests
   temporarily disabled

The QEMU driver is mistakenly calling directly into the text
mode monitor for the domain memory stats query.

* src/qemu/qemu_driver.c: Replace qemuMonitorTextGetMemoryStats with
  qemuMonitorGetMemoryStats
* src/qemu/qemu_monitor.c, src/qemu/qemu_monitor.h: Add the new
  wrapper for qemuMonitorGetMemoryStats
* src/qemu/qemu_monitor_json.c, src/qemu/qemu_monitor_json.h: Add
  qemuMonitorJSONGetMemoryStats implementation

Instead of reporting VIR_ERR_INTERNAL_ERROR use the more specific
VIR_ERR_CONFIG_UNSUPPORTED

* src/qemu/qemu_conf.c: Report VIR_ERR_CONFIG_UNSUPPORTED for
  unsupported video adapters

To avoid race-conditions, the tear down of a filter has to happen before
the tap interface disappears and another tap interface with the same
name can re-appear. This patch tries to fix this. In one place, where
communication with the qemu monitor may fail, I am only tearing the
filters down after knowing that the function did not fail.

I am also moving the tear down functions into an include file for other
drivers to reuse.

I am adding braces around the for statement that are now needed to due the
new sa_assert immediately following the for statement.

Triggered by gnulib when compiling with MinGW.

* This patch implements a memory allocator to obtain memory for
  structures whose last member is a variable length array.  C99 refers
  to these variable length objects as structs containing flexible
  array members.
* Fixed macro parentheses per Eric Blake

* src/util/conf.c (virConfParseValue): Add an sa_assert.

* src/xen/xend_internal.c (xend_parse_sexp_desc_char): Add three
uses of sa_assert, each preceding a strchr(value,... to assure
clang that "value" is non-NULL.

* src/qemu/qemu_driver.c (qemudDomainAttachSCSIDisk):
Initialize "cont" to NULL, so clang knows it's set.
Add an sa_assert so it knows it's non-NULL when dereferenced.

* src/util/util.c (virGetHostnameLocalhost): Add an sa_assert
to tell clang it's ok to dereference "info" after a non-failing
getaddrinfo call.

* src/nwfilter/nwfilter_ebiptables_driver.c (ebiptablesApplyNewRules):
Don't dereference a NULL or uninitialized pointer when given
an empty list of rules.  Add an sa_assert(inst) in each loop to
tell clang that the uses of "inst[i]" are valid.

* configure.ac (STATIC_ANALYSIS): Define when run via clang's
scan-build or coverity-prevent's cov-build.
Use the CLANG_CC and COVERITY_BUILD_COMMAND envvars as witnesses.

Among some here, there is a strong aversion to the use of "assert", yet
some others think it is essential (when applied judiciously) even --
perhaps "especially" -- at the heart of libraries and core hypervisor-
related code.
Here is a compromise that lets us make assertions about the code (e.g.,
to tell static analyzers about invariants) without even a hint of risk
of an abort.
* src/internal.h [STATIC_ANALYSIS]: Include <assert.h>.
(sa_assert): Define.  A no-op most of the time, but equivalent
to classical assert when STATIC_ANALYSIS is nonzero.

* src/esx/esx_vi_methods.h: Placate cppi.

* tests/virt-aa-helper-test (testme): Use printf instead.

* tests/schematestutils.sh: Use printf rather than echo -e.

* src/storage/storage_backend_fs.c (virStorageBackendFileSystemMount):
Use virAsprintf only when needed.  In this case, strdup works fine.

* src/storage/storage_backend_fs.c (virStorageBackendFileSystemMount):
Clang was not smart enough, and mistakenly reported that "options"
could be used uninitialized.  Initialize it.

Somehow the backend of this function was never implemented in
libvirt's netcf driver, and nobody noticed until now. (The required
netcf function was already in place, so nothing needs to change
there.)
* src/interface/netcf_driver.c: add in the backend function, and point
                                to it from the table of driver functions.

* src/openvz/openvz_driver.c (openvzGetProcessInfo): Reorganize
so that unexpected /proc/vz/vestat content cannot make us use
uninitialized variables.  Without this change, an input line with
a matching "readvps", but fewer than 4 numbers would result in our
using at least "systime" uninitialized.

* tools/virsh.c (vshCommandRun): Test only the initial value of
ctl->timing, so that static analyzers don't have to consider that
it might be changed by cmd->def->handler.

I am getting rid of determining the path to necessary CLI tools at
compile time. Instead, now the firewall driver has an initialization
function that uses virFindFileInPath() to determine the path to
necessary CLI tools and a shutdown function to free allocated memory.
The rest of the patch mostly deals with availability of the CLI tools
and to not call certain code blocks if a tool is not available and that
strings now have to be built slightly differently.

Generate almost all SOAP method mapping code.

Update the driver code to use the complete paramater list of some methods
that had parameters skipped before.

Improve the ESX_VI__METHOD marco to do automatic output deserialization
based on output occurrence. Also incorporate automatic _this binding and
output pointer check.

* src/esx/esx_vmx.c (esxVMX_GatherSCSIControllers): Do not dereference
a NULL disk->driverName.  We already detect this condition in another
case.  Check for it here, too.

When building libvirt on RHEL-5, I saw this error:

cc1: warnings being treated as errors
openvz/openvz_conf.c: In function 'openvzGetVPSUUID':
openvz/openvz_conf.c:835: warning: 'saveptr' may be used uninitialized in this function
make[3]: *** [libvirt_driver_openvz_la-openvz_conf.lo] Error 1

gcc in RHEL-5 gets upset about this usage of strtok_r (even though
it is perfectly valid).  Just set *saveptr to NULL at the
start to quiet it down.
Signed-off-by: NChris Lalancette <clalance@redhat.com>

Signed-off-by: NChris Lalancette <clalance@redhat.com>

Signed-off-by: NChris Lalancette <clalance@redhat.com>