Avoid people introducing security flaws in their apps by
forbidding the use of libvirt.so in setuid programs, with
a check in virInitialize.
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>

Most of the usage of getuid()/getgid() is in cases where we are
considering what privileges we have. As such the code should be
using the effective IDs, not real IDs.
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>

We already have stubs for getuid, geteuid, getgid but
not for getegid. Something in gnulib already does a
check for it during configure, so we already have the
HAVE_GETEGID macro defined.
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>

We don't want setuid programs automatically spawning libvirtd,
so disable any use of autostart when setuid.
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>

We don't know enough about quality of external libraries used
for non-UNIX transports, nor do we want to spawn external
commands when setuid. Restrict to the bare minimum which is
UNIX transport for local usage. Users shouldn't need to be
running setuid if connecting to remote hypervisors in any
case.
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>

The use of getenv is typically insecure, and we want people
to use our wrappers, to force them to think about setuid
needs.
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>

Unconditional use of getenv is not secure in setuid env.
While not all libvirt code runs in a setuid env (since
much of it only exists inside libvirtd) this is not always
clear to developers. So make all the code paranoid, even
if it only ever runs inside libvirtd.
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>

When running setuid, we must be careful about what env vars
we allow commands to inherit from us. Replace the
virCommandAddEnvPass function with two new ones which do
filtering

  virCommandAddEnvPassAllowSUID
  virCommandAddEnvPassBlockSUID

And make virCommandAddEnvPassCommon use the appropriate
ones
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>

The virt-login-shell binary shouldn't need to execute programs
relying on $PATH, but just in case set a fixed $PATH value
of /bin:/usr/bin
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>

The libvirt.so library has far too many library deps to allow
linking against it from setuid programs. Those libraries can
do stuff in __attribute__((constructor) functions which is
not setuid safe.

The virt-login-shell needs to link directly against individual
files that it uses, with all library deps turned off except
for libxml2 and libselinux.

Create a libvirt-setuid-rpc-client.la library which is linked
to by virt-login-shell. A config-post.h file allows this library
to disable all external deps except libselinux and libxml2.
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>

We don't want to inherit any FDs in the new namespace
except for the stdio FDs. Explicitly close them all,
just in case some do not have the close-on-exec flag
set.
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>

We must not allow file/syslog/journald log outputs when running
setuid since they can be abused to do bad things. In particular
the 'file' output can be used to overwrite files.
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>

Care must be taken accessing env variables when running
setuid. Introduce a virGetEnvAllowSUID for env vars which
are safe to use in a setuid environment, and another
virGetEnvBlockSUID for vars which are not safe. Also add
a virIsSUID helper method for any other non-env var code
to use.
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>

The virConnectDomainXMLToNative API should require 'connect:write'
not 'connect:read', since it will trigger execution of the QEMU
binaries listed in the XML.

Also make virConnectDomainXMLFromNative API require a full
read-write connection and 'connect:write' permission. Although the
current impl doesn't trigger execution of QEMU, we should not
rely on that impl detail from an API permissioning POV.
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>

s/effect/affect
Signed-off-by: NChen Hanxiao <chenhanxiao@cn.fujitsu.com>

The meaning of one line of code was accidentally inverted.
Signed-off-by: NEric Blake <eblake@redhat.com>

Commit e3ef20d7 allows user to configure migration ports range via
qemu.conf. However, it forgot to update augeas definition file and
even the test data was malicious.
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>

host-model is a nice idea but it's current implementation make it
useless on some hosts so it should be used with care.

https://bugzilla.redhat.com/show_bug.cgi?id=1019053

https://bugzilla.redhat.com/show_bug.cgi?id=1019053

When we migrate vms concurrently, there's a chance that libvirtd on
destination assigns the same port for different migrations, which will
lead to migration failure during prepare phase on destination. So we use
virPortAllocator here to solve the problem.
Signed-off-by: NWang Yufei <james.wangyufei@huawei.com>
Signed-off-by: NJiri Denemark <jdenemar@redhat.com>

In fact, the suffix should be _QUIET not _QUIT to stress the
fact, that no OOM error is reported on error.
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>

netcfStateInitialize() initializes the driverState variable,
and when netcfStateCleanup is called, it will call virReportError()
if driverState is NULL.
This is not consistent with what other state objects are doing,
they return -1 without reporting an error in such cases.

See also
https://www.redhat.com/archives/libvir-list/2013-October/msg00809.html:

On Thu, Oct 17, 2013 at 01:40:19PM +0100, Daniel P. Berrange wrote:
> We don't want virStateCleanup to skip execution if virStateInitialize
> has failed though - every callback in virStateCleanup should be written
> to be safe if its corresponding init function hasn't run.

Introduced by 7b87a3
When I quit the process which only register VIR_DOMAIN_EVENT_ID_REBOOT,
I got error like:
"libvirt: XML-RPC error : internal error: domain event 0 not registered".
Then I add the following code, it fixed.
Signed-off-by: NZhou Yimin <zhouyimin@huawei.com>
Signed-off-by: NEric Blake <eblake@redhat.com>

If we fail to get domain, we had to judge whether
it's null or not when doing 'cleanup'.
Signed-off-by: NChen Hanxiao <chenhanxiao@cn.fujitsu.com>

The header definition didn't match the function declaration, so adjusted
header to reflect the definition.

Found during a Coverity build where STATIC_ANALYSIS is enabled resulting
in the internal.h adding __nonnull__ handling to arguments.

Commit '6d264c91' added support for the qemuMonitorJSONDrivePivot() and
commit 'fbc3adc9' added a corresponding test which ended up triggering
the build failure which I didn't notice until today!

'--print-xml' option is very useful for doing some test.
But we had to specify a real domain for it.
This patch could enable us to specify a fake domain
when using --print-xml option.
Signed-off-by: NChen Hanxiao <chenhanxiao@cn.fujitsu.com>
Signed-off-by: NEric Blake <eblake@redhat.com>

Code cleanup: remove explicit NULL comparisons like ptr == NULL and
ptr != NULL from the ESX code, replacing them with the simpler ptr
and !ptr.

Part three of three.

Code cleanup: remove explicit NULL comparisons like ptr == NULL and
ptr != NULL from the ESX code, replacing them with the simpler ptr
and !ptr.

Part two of three.

Code cleanup: remove explicit NULL comparisons like ptr == NULL and
ptr != NULL from the ESX code, replacing them with the simpler ptr
and !ptr.

Part one of three.

QEMU has support for SASL auth for SPICE guests, but libvirt
has no way to enable it. Following the example from VNC where
it is globally enabled via qemu.conf
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>

The last argument of memmove is the amount of bytes to be moved. The
amount is in Bytes. We are moving some void pointers around. However,
since sizeof(void *) is not Byte on any architecture, we've got the
arithmetic wrong.

This patch improves the error checking in the LOCAL_PEERCRED version
of virNetSocketGetUNIXIdentity, used by FreeBSD and Mac OSX.

1. The error return paths now correctly unlock the socket. This is
implemented in exactly the same way as the SO_PEERCRED version,
using "goto cleanup"

2. cr.cr_ngroups is initialised to -1, and cr.cr_ngroups is checked
for negative and overlarge values.

This means that if the getsockopt() call returns success but doesn't
actually update the xucred structure, this is now caught. This
happened previously when getsockopt was called with SOL_SOCKET
instead of SOL_LOCAL, prior to commit 5a468b38, and resulted in
random uids being accepted.
Signed-off-by: NEric Blake <eblake@redhat.com>

s/it's/its
Signed-off-by: NChen Hanxiao <chenhanxiao@cn.fujitsu.com>

prevent aclocal from preferring .m4 files under m4/ over the version
provided by gnulib, by using only one directory.

I have noticed this after './configure --help' gave me two different
versions of "--enable-threads".  This was caused by aclocal that
preferred the version of lock.m4 provided by autopoint instead of
using the newer version distributed with gnulib.

Having two different directories made sense back when we checked
gnulib files into libvirt.git, but that was ages ago.
Signed-off-by: NGiuseppe Scrivano <gscrivan@redhat.com>
Signed-off-by: NEric Blake <eblake@redhat.com>

The RNG grammar did not allow arbitrary interleaving, which makes
it harder than necessary to create a new pool from handwritten XML.

* docs/schemas/storagepool.rng: Allow interleaving.
* tests/storagepoolxml2xmlin/pool-sheepdog.xml: Test interleave.
* tests/storagepoolxml2xmlin/pool-iscsi-auth.xml: Likewise.
Signed-off-by: NEric Blake <eblake@redhat.com>

We forgot to document several pool types.

* docs/formatstorage.html.in: Add docs for scsi, mpath, rbd, and
sheepdog.
Signed-off-by: NEric Blake <eblake@redhat.com>

<...>
/* Size of message length field. Not counted in VIR_NET_MESSAGE_MAX
 * and VIR_NET_MESSAGE_INITIAL.
 */
const VIR_NET_MESSAGE_LEN_MAX = 4;
</...>

However, msg->bufferLength includes the length word. The wrong checking
was introduced by commit e914dcfd.

* src/rpc/virnetmessage.c:
  - Correct the checking in virNetMessageEncodePayloadRaw
  - Use a new variable to track the new payload length in
    virNetMessageEncodePayloadRaw

The recent patch series proposing the addition of PPC little endian
arch support to Linux defines new arch names 'ppcle' and 'ppc64le':

https://lists.ozlabs.org/pipermail/linuxppc-dev/2013-August/109908.html

This just makes libvirt know about these arch names, so it doesn't
immediately trip up if it seems these new names from uname.
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>

A typo in the setup of NBD backed filesystems meant the
/dev/nbdN device would not be added to the cgroups device
ACL.
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>