Using automated replacement with sed and editing I have now replaced all
occurrences of close() with VIR_(FORCE_)CLOSE() except for one, of
course. Some replacements were straight forward, others I needed to pay
attention. I hope I payed attention in all the right places... Please
have a look. This should have at least solved one more double-close
error.

Instead of storing the IP address string in virNetwork related
structs, store the parsed virSocketAddr. This will make it
easier to add IPv6 support in the future, by letting driver
code directly check what address family is present

* src/conf/network_conf.c, src/conf/network_conf.h,
  src/network/bridge_driver.c: Convert to use virSocketAddr
  in virNetwork, instead of char *.
* src/util/bridge.c, src/util/bridge.h,
  src/util/dnsmasq.c, src/util/dnsmasq.h,
  src/util/iptables.c, src/util/iptables.h: Convert to
  take a virSocketAddr instead of char * for any IP
  address parameters
* src/util/network.h: Add macros to determine if an address
  is set, and what address family is set.

The  inet_pton and inet_ntop functions are obsolete, replaced
by getaddrinfo+getnameinfo with the AI_NUMERICHOST flag set.
These can be accessed via the virSocket APIs.

The bridge.c code had methods for fetching the IP address of
a bridge which used inet_ntop. Aside from the use of inet_ntop
these methods are broken, because a NIC can have multiple
addresses and this only returns one address. Since the methods
are never used, just remove them.

* src/conf/network_conf.c, src/nwfilter/nwfilter_learnipaddr.c:
  Replace inet_pton and inet_ntop with virSocket APIs
* src/util/bridge.c, src/util/bridge.h: Remove unused methods
  which called inet_ntop.

When passing a NULL tapfd argument to brAddTap, we need to close the fd
of the tap device. If we don't, libvirt will keep the fd open
indefinitely and renders the the guest unable to configure its side of
the tap device.
Signed-off-by: NSoren Hansen <soren@linux2go.dk>

A Linux software bridge will assume the MAC address of the enslaved
interface with the numerically lowest MAC addr. When the bridge
changes MAC address there is a period of network blackout, so a
change should be avoided. The kernel gives TAP devices a completely
random MAC address. Occassionally the random TAP device MAC is lower
than that of the physical interface (eth0, eth1etc) that is enslaved,
causing the bridge to change its MAC.

This change sets an explicit MAC address for all TAP devices created
using the configured MAC from the XML, but with the high byte set
to 0xFE. This should ensure TAP device MACs are higher than any
physical interface MAC.

* src/qemu/qemu_conf.c, src/uml/uml_conf.c: Pass in a MAC addr
  for the TAP device with high byte set to 0xFE
* src/util/bridge.c, src/util/bridge.h: Set a MAC when creating
  the TAP device to override random MAC

* global: patch created by running:
for f in $(git ls-files '*.[ch]') ; do
    cppi $f > $f.t && mv $f.t $f
done

It was used for error reporting only.

Add the virStrncpy function, which takes a dst string, source string,
the number of bytes to copy and the number of bytes available in the
dest string.  If the source string is too large to fit into the
destination string, including the \0 byte, then no data is copied and
the function returns NULL.  Otherwise, this function copies n bytes
from source into dst, including the \0, and returns a pointer to the
dst string.  This function is intended to replace all unsafe uses
of strncpy in the code base, since strncpy does *not* guarantee that
the buffer terminates with a \0.
Signed-off-by: NChris Lalancette <clalance@redhat.com>

* src/bridge.c, src/bridge.h, src/buf.c, src/buf.h, src/cgroup.c,
  src/cgroup.h, src/conf.c, src/conf.h, src/event.c, src/event.h,
  src/hash.c, src/hash.h, src/hostusb.c, src/hostusb.h,
  src/iptables.c, src/iptables.h, src/logging.c, src/logging.h,
  src/memory.c, src/memory.h, src/pci.c, src/pci.h, src/qparams.c,
  src/qparams.h, src/stats_linux.c, src/stats_linux.h,
  src/threads-pthread.c, src/threads-pthread.h, src/threads-win32.c,
  src/threads-win32.h, src/threads.c, src/threads.h, src/util.c,
  src/util.h, src/uuid.c, src/uuid.h, src/virterror.c,
  src/virterror_internal.h, src/xml.c, src/xml.h: Move all files
  into src/util/
* daemon/Makefile.am: Add -Isrc/util/ to build flags
* src/Makefile.am: Add -Isrc/util/ to build flags and update for
  moved files
* src/libvirt_private.syms: Export cgroup APIs since they're now
  in util rather than linking directly to drivers
* src/xen/xs_internal.c: Disable bogus virEventRemoveHandle call
  when built under PROXY
* proxy/Makefile.am: Update for changed file locations. Remove
  bogus build of event.c
* tools/Makefile.am, tests/Makefile.am: Add -Isrc/util/ to build flags

* src/bridge.c: avoid a couple of unused var/func warnings

* src/bridge.c: fix struct ifreq ifr init in brAddTap()

* src/bridge.c: no need to format the device string in brAddTap()
  this can be delegated to the kernel and removes an arbitrary limit.

daniel

* src/bridge.c (brProbeVnetHdr)
[IFF_VNET_HDR && TUNGETFEATURES && TUNGETIFF]: Use a "(void)" case
to mark the parameter as unused.  Reported by Maximilian Wilhelm in
http://thread.gmane.org/gmane.comp.emulators.libvirt/11918/focus=11917

 bridge.c: In function 'brSetInetAddr':
 bridge.c:665: error: dereferencing pointer '({anonymous})' does break strict-aliasing rules
 bridge.c:665: note: initialized from here
 bridge.c:666: error: dereferencing pointer '({anonymous})' does break strict-aliasing rules
 bridge.c:666: note: initialized from here

Fix qemud.c warnings with gcc-4.4

 qemud.c: In function ‘remoteListenTCP’:
 qemud.c:675: error: dereferencing pointer ‘sa.72’ does break strict-aliasing rules
 qemud.c:675: note: initialized from here
 qemud.c:678: error: dereferencing pointer ‘sa.73’ does break strict-aliasing rules
 qemud.c:678: note: initialized from here

* src/bridge.c (brAddTap): Remove redundant errno=ENOMEM assignment
after failed strdup.
* src/cgroup.c (virCgroupFree): Remove redundant *group=NULL assignment
after VIR_FREE(*group).
* src/lxc_driver.c (lxcShutdown): Likewise.
* src/xen_unified.c (xenUnifiedClose): Likewise.
* src/xm_internal.c (xenXMDomainConfigFormat): Rewrite disk-
and net-config-setting code to be clearer.

IFF_VNET_HDR is a tun/tap flag that allows you to send and receive
large (i.e. GSO) packets and packets with partial checksums. Setting
the flag means that every packet is proceeded by the same header which
virtio uses to communicate GSO/csum metadata.

By enabling this flag on the tap fds we create, we greatly increase
the achievable throughput with virtio_net.

However, we need to be careful to only set the flag when a) QEMU has
support for this ABI and b) the value of the flag is queryable using
the TUNGETIFF ioctl.

It's nearly five months since kvm-74 - the first KVM release with this
feature - was released. Up until now, we've not added libvirt support
because there is no clean way to detect support for this in QEMU at
runtime. A brief attempt to add a "info capabilities" monitor command
to QEMU floundered. Perfect is the enemy of good enough. Probing the
KVM version will suffice for now.
Signed-off-by: NMark McLoughlin <markmc@redhat.com>

* configure.in: Provide a new --with-bridge option.
* src/Makefile.am (PRIVSYMFILES) [WITH_BRIDGE]: Append libvirt_bridge.syms.
* src/bridge.c: Test WITH_BRIDGE rather than particular drivers.
* src/libvirt_private.syms: Move bridge-related symbols into...
* src/libvirt_bridge.syms: ...this new file.
Author: John Levon

* src/bridge.c: patch setting MTU on tap interface to be the same
  as the value for the bridge (Eduardo Habkost)
daniel

Daniel

Done with these commands:
git grep -l Local.variab|xargs \
  perl -0x3b -pi -e 's,\n+/\*\n \* vim:(.|\n)*,\n,'

git grep -l Local.variab|xargs \
  perl -0x3b -pi -e 's,\n+/\*\n \* Local variables:\n(.|\n)*,\n,'

* src/bridge.c: fix a typo in a comment
Daniel

daniel

* Makefile.cfg (local-checks-to-skip): Remove sc_trailing_blank.
* .x-sc_trailing_blank: New file, to exempt the few binary files.

Given code like: T *var = calloc (n, sizeof (T));
Convert to this: T *var = calloc (n, sizeof (*var));
This first-cut change adjusts all malloc, calloc, and
realloc statements.

The only binary differences are in remote_internal.c
(due to the bug fix) and in xmlrpc.c (due to factorization).

* python/libvir.c: As above.
* qemud/event.c: Likewise.
* qemud/mdns.c: Likewise.
* qemud/qemud.c: Likewise.
* qemud/remote.c: Likewise.
* src/bridge.c: Likewise.
* src/buf.c: Likewise.
* src/conf.c: Likewise.
* src/hash.c: Likewise.
* src/iptables.c: Likewise.
* src/openvz_conf.c: Likewise.
* src/qemu_conf.c: Likewise.
* src/qemu_driver.c: Likewise.
* src/test.c: Likewise.
* src/xen_internal.c: Likewise.
* src/xen_unified.c: Likewise.
* src/xm_internal.c: Likewise.
* src/xml.c: Likewise.
* tests/qemuxml2argvtest.c: Likewise.
* src/xmlrpc.c (xmlRpcValuePtr): Likewise, and minor factorization.
* src/remote_internal.c (remoteAuthMakeCredentials): Use the right
type when allocating space for an array of cred _pointers_.

	* src/bridge.c, src/bridge.h, src/qemu_conf.c, src/qemu_conf.h,
	  src/qemu_driver.c, src/qemu_driver.h, src/test.c: Disable more
	  of qemu if configured --without-qemu.

	* src/bridge.c, src/qemu_driver.c: Incorrect call to
	  brSetForwardDelay changed to brSetEnableSTP.

        * src/bridge.c, src/qemu_driver.c, configure.in: Try to detect
        the paths of dnsmasq and brctl at compile time.  If found
        then compile them in, otherwise search $PATH at runtime.

        * libvirt.spec.in: BR dnsmasq and bridge-utils so we have their
        paths at compile time.

        * acinclude.m4, configure.in: ./configure option
          --disable-stack-protector.
        * src/gnutls_1_0_compat.h: Compatibility with GnuTLS 1.0.
        * src/bridge.c: If no bridge ioctls, give an error at runtime.

  stylesheet
* docs/*: regenerated
* src/bridge.c src/bridge.h src/buf.c src/iptables.c src/libvirt.c
  src/qemu_driver.c src/qemu_driver.h src/uuid.c src/uuid.h: cleanup,
  addd comments, made functions static and fixe a few bugs
Daniel