To simplify future patches dealing with this code, simplify and refactor
some conditions to switch statements.

(cherry picked from commit 9d132989)

When doing an internal snapshot on a VM with sheepdog or RBD disks we
would not set a flag to mark the domain is using internal snapshots and
might end up creating a mixed snapshot. Move the setting of the variable
to avoid this problem.

(cherry picked from commit d8cf91ae)

Restarting an active libvirt-guests.service is the equivalent of
doing:

/usr/libexec/libvirt-guests.sh stop
/usr/libexec/libvirt-guests.sh start

Which in a default configuration will managedsave every running VM,
and then restore them. Certainly not something we should do every
time the libvirt-client RPM is updated.

Just drop the try-restart attempt, I don't know what purpose it
serves anyways.

https://bugzilla.redhat.com/show_bug.cgi?id=962225
(cherry picked from commit d65e0e14)

If the host side of an LXC container console disconnected
and the guest side continued to write data, until the PTY
buffer filled up, the LXC controller would busy wait. It
would repeatedly see POLLHUP from poll() and not disable
the watch.

This was due to some bogus logic detecting blocking
conditions. Upon seeing a POLLHUP we must disable all
reading & writing from the PTY, and setup the epoll to
wake us up again when the connection comes back.
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>
(cherry picked from commit 5087a5a0)

Possible fix for occasional libvirt-guests failure at boot time:

https://bugzilla.redhat.com/show_bug.cgi?id=906009
(cherry picked from commit d9203675)

Syslog is socket activated since at least systemd v35 so we can drop
this dependency. Debian's linitian otherwise complains about it.

References:

    http://www.freedesktop.org/wiki/Software/systemd/syslog/
    http://lintian.debian.org/tags/systemd-service-file-refers-to-obsolete-target.html

(cherry picked from commit 3c9e40a1)

QEMU 1.6.0 introduced new migration status: setup
Libvirt does not expect such string in QMP and refuses to migrate with error
"unexpected migration status in setup"

This patch fixes it.
Signed-off-by: NJiri Denemark <jdenemar@redhat.com>
(cherry picked from commit d35ae414)

Rather than casting the virBitmap pointer to uint8_t* and then using
the structure contents as a byte array, use the virBitmap API to determine
the bitmap size and test each bit.
Signed-off-by: NJeremy Fitzhardinge <jeremy@goop.org>
(cherry picked from commit ba1bf100)

This should resolve:

  https://bugzilla.redhat.com/show_bug.cgi?id=1012085

libvirt previously recognized NFS, GFS2, OCFS2, and AFS filesystems as
"shared", and thus eligible for exceptions to certain rules/actions
about chowning image files before handing them off to a guest. This
patch widens the definition of "shared filesystem" to include SMB and
CIFS filesystems (aka "Windows file sharing"); both of these use the
same protocol, but different drivers so there are different magic
numbers for each.

(cherry picked from commit e4e73337)

When opening a new connection to the driver, nwfilterOpen
only succeeds if the driverState has been allocated.

Move the privilege check in driver initialization before
the state allocation to disable the driver.

This changes the nwfilter-define error from:
error: cannot create config directory (null): Bad address
To:
this function is not supported by the connection driver:
virNWFilterDefineXML

https://bugzilla.redhat.com/show_bug.cgi?id=1029266
(cherry picked from commit b7829f95)

Since qemu-kvm 1.1 [1] (since 1.3. in upstream QEMU [2])
'-no-kvm-pit-reinjection' has been deprecated.
Use -global kvm-pit.lost_tick_policy=discard instead.

https://bugzilla.redhat.com/show_bug.cgi?id=978719

[1] http://git.kernel.org/cgit/virt/kvm/qemu-kvm.git/commit/?id=4e4fa39
[2] http://git.qemu.org/?p=qemu.git;a=commitdiff;h=c21fb4f

(cherry picked from commit 1569fa14)

Conflicts:
	tests/qemucapabilitiesdata/caps_1.2.2-1.caps
	tests/qemucapabilitiesdata/caps_1.2.2-1.replies
	tests/qemucapabilitiesdata/caps_1.3.1-1.caps
	tests/qemucapabilitiesdata/caps_1.3.1-1.replies
	tests/qemucapabilitiesdata/caps_1.4.2-1.caps
	tests/qemucapabilitiesdata/caps_1.4.2-1.replies
	tests/qemucapabilitiesdata/caps_1.5.3-1.caps
	tests/qemucapabilitiesdata/caps_1.5.3-1.replies
	tests/qemucapabilitiesdata/caps_1.6.0-1.caps
	tests/qemucapabilitiesdata/caps_1.6.0-1.replies
	tests/qemucapabilitiesdata/caps_1.6.50-1.caps
	tests/qemucapabilitiesdata/caps_1.6.50-1.replies
(qemucapabilitiestest is not backported)

Since 86d90b3a (yes, my patch; again) we are supporting NBD storage
migration. However, on error recovery path we got the steps reversed.
The correct order is: return NBD port to the virPortAllocator and then
either unlock the vm or remove it from the driver. Not vice versa.

==11192== Invalid write of size 4
==11192==    at 0x11488559: qemuMigrationPrepareAny (qemu_migration.c:2459)
==11192==    by 0x11488EA6: qemuMigrationPrepareDirect (qemu_migration.c:2652)
==11192==    by 0x114D1509: qemuDomainMigratePrepare3Params (qemu_driver.c:10332)
==11192==    by 0x519075D: virDomainMigratePrepare3Params (libvirt.c:7290)
==11192==    by 0x1502DA: remoteDispatchDomainMigratePrepare3Params (remote.c:4798)
==11192==    by 0x12DECA: remoteDispatchDomainMigratePrepare3ParamsHelper (remote_dispatch.h:5741)
==11192==    by 0x5212127: virNetServerProgramDispatchCall (virnetserverprogram.c:435)
==11192==    by 0x5211C86: virNetServerProgramDispatch (virnetserverprogram.c:305)
==11192==    by 0x520A8FD: virNetServerProcessMsg (virnetserver.c:165)
==11192==    by 0x520A9E1: virNetServerHandleJob (virnetserver.c:186)
==11192==    by 0x50DA78F: virThreadPoolWorker (virthreadpool.c:144)
==11192==    by 0x50DA11C: virThreadHelper (virthreadpthread.c:161)
==11192==  Address 0x1368baa0 is 576 bytes inside a block of size 688 free'd
==11192==    at 0x4A07F5C: free (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
==11192==    by 0x5079A2F: virFree (viralloc.c:580)
==11192==    by 0x11456C34: qemuDomainObjPrivateFree (qemu_domain.c:267)
==11192==    by 0x50F41B4: virDomainObjDispose (domain_conf.c:2034)
==11192==    by 0x50C2991: virObjectUnref (virobject.c:262)
==11192==    by 0x50F4CFC: virDomainObjListRemove (domain_conf.c:2361)
==11192==    by 0x1145C125: qemuDomainRemoveInactive (qemu_domain.c:2087)
==11192==    by 0x11488520: qemuMigrationPrepareAny (qemu_migration.c:2456)
==11192==    by 0x11488EA6: qemuMigrationPrepareDirect (qemu_migration.c:2652)
==11192==    by 0x114D1509: qemuDomainMigratePrepare3Params (qemu_driver.c:10332)
==11192==    by 0x519075D: virDomainMigratePrepare3Params (libvirt.c:7290)
==11192==    by 0x1502DA: remoteDispatchDomainMigratePrepare3Params (remote.c:4798)
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>
(cherry picked from commit 1f2f879e)

https://bugzilla.redhat.com/show_bug.cgi?id=1018897

If a PCI deivce is not binded to any driver (e.g. there's yet no PCI
driver in the linux kernel) but still users want to passthru the device
we fail the whole operation as we fail to resolve the 'driver' link
under the PCI device sysfs tree. Obviously, this is not a fatal error
and it shouldn't be error at all.
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>
(cherry picked from commit df4283a5)

https://bugzilla.redhat.com/show_bug.cgi?id=1027096

If there's the following snippet in the domain XML, the domain will be
lost upon the daemon restart (if the domain is started prior restart):

    <seclabel type='dynamic' relabel='yes'/>

The problem is, the 'label', 'imagelabel' and 'baselabel' are parsed
whenever the VIR_DOMAIN_XML_INACTIVE is *not* present or the label is
static. The latter is not our case, obviously. So, when libvirtd starts
up, it finds domain state xml and parse it. During parsing, many XML
flags are enabled but VIR_DOMAIN_XML_INACTIVE. Hence, our parser tries
to extract 'label', 'imagelabel' and 'baselabel' from the XML which
fails for model='none'. Err, this model - even though not specified in
XML - can be taken from qemu wide config file: /etc/libvirtd/qemu.conf.

However, in order to know we are dealing with model='none' the code in
question must be moved forward a bit. Then a new check must be
introduced. This is what the first two chunks are doing.

But this alone is not sufficient. The domain state XML won't contain the
model attribute without slight modification. The model should be
inserted into the XML even if equal to 'none' and the state XML is being
generated - what if the origin (the @security_driver variable in
qemu.conf) changes during libvirtd restarts?

At the end, a test to catch this scenario is introduced.
Signed-off-by: NMichal Privoznik <mprivozn@redhat.com>
(cherry picked from commit 9fb3f957)

For inexplicable reasons, many of the 3rd party package deps
were left against the 'libvirt-daemon' RPM when the drivers
were split out. This makes a minimal install heavier that
it should be. Push them all down into libvirt-daemon-driver-XXX
so they're only pulled in when truly needed

With this change applied, a minimal install of just the
libvirt-daemon-driver-lxc RPM is reduced by 41 MB on a
Fedora 19 host.
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>
(cherry picked from commit 23142ac9)

The following sequence

 1. Define a persistent QMEU guest
 2. Start the QEMU guest
 3. Stop libvirtd
 4. Kill the QEMU process
 5. Start libvirtd
 6. List persistent guests

At the last step, the previously running persistent guest
will be missing. This is because of a race condition in the
QEMU driver startup code. It does

 1. Load all VM state files
 2. Spawn thread to reconnect to each VM
 3. Load all VM config files

Only at the end of step 3, does the 'virDomainObjPtr' get
marked as "persistent". There is therefore a window where
the thread reconnecting to the VM will remove the persistent
VM from the list.

The easy fix is to simply switch the order of steps 2 & 3.

In addition to this though, we must only attempt to reconnect
to a VM which had a non-zero PID loaded from its state file.
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>
(cherry picked from commit f26701f5)

The 'error' cleanup block in qemuProcessReconnect() had a
'return' statement in the middle of it. This caused a leak
of virConnectPtr & virQEMUDriverConfigPtr instances. This
was identified because netcf recently started checking its
refcount in libvirtd shutdown:

netcfStateCleanup:109 : internal error: Attempt to close netcf state driver with open connections
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>
(cherry picked from commit 54a24112)

virDomainObjListLoadAllConfigs sets dom->persistent after
having released its lock on the domain object. This exposes
a possible race condition.
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>
(cherry picked from commit b260a77e)

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>
(cherry picked from commit 9cd6a57d)

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>
(cherry picked from commit 9b0af092)

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>
(cherry picked from commit c566fa1a)

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>
(cherry picked from commit 171bb129)

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>
(cherry picked from commit e22b0232)

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>
(cherry picked from commit 71b21f12)

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>
(cherry picked from commit 1e4a02bd)

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>
(cherry picked from commit 9b8f307c)

In Fedora 20, libvirt_lxc crashes immediately at startup with a
trace

 #0  0x00007f0cddb653ec in free () from /lib64/libc.so.6
 #1  0x00007f0ce0e16f4a in virFree (ptrptr=ptrptr@entry=0x7f0ce1830058) at util/viralloc.c:580
 #2  0x00007f0ce0e2764b in virResetError (err=0x7f0ce1830030) at util/virerror.c:354
 #3  0x00007f0ce0e27a5a in virResetLastError () at util/virerror.c:387
 #4  0x00007f0ce0e28858 in virEventRegisterDefaultImpl () at util/virevent.c:233
 #5  0x00007f0ce0db47c6 in main (argc=11, argv=0x7fff4596c328) at lxc/lxc_controller.c:2352

Normally virInitialize calls virErrorInitialize and
virThreadInitialize, but we don't link to libvirt.so
in libvirt_lxc, and nor did we ever call the error
or thread initializers.

I have absolutely no idea how this has ever worked, let alone
what caused it to stop working in Fedora 20.

In addition not all code paths from virLogSetFromEnv will
ensure virLogInitialize is called correctly, which is another
possible crash scenario.
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>
(cherry picked from commit 97973ebb)

The log message regex has been

[0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}:[0-9]{2}\.[0-9]{3}\+[0-9]{4}: [0-9]+: debug|info|warning|error :

The precedence of '|' is high though, so this is equivalent to matching

   [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}:[0-9]{2}\.[0-9]{3}\+[0-9]{4}: [0-9]+: debug

Or

   info

Or

   warning

Or

   error :

Which is clearly not what it should have done. This caused the code to
skip over things which are not log messages. The solution is to simply
add brackets.

A test case is also added to validate correctness.
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>
(cherry picked from commit 5787f0b9)

Many people will not want the setuid virt-login-shell binary
installed by default, so move it into a separate sub-RPM
named libvirt-login-shell. This RPM is only generated if
LXC is enabled
Signed-off-by: NDaniel P. Berrange <berrange@redhat.com>
(cherry picked from commit 8adc9269)

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>
(cherry picked from commit d665003d)

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>
(cherry picked from commit 9712c251)

Commit d76227be added functions virDomainCreateWithFiles and
virDomainCreateXMLWithFiles, but there was a little piece missing in
python bindings.  This patch fixes proper passing of file descriptors
in the overwrites of these functions.

On systems where gnutls uses libgcrypt, I'm seeing the following
build failure

libvirt.c:314: error: variable 'virTLSThreadImpl' has initializer but incomplete type
libvirt.c:319: error: 'GCRY_THREAD_OPTION_PTHREAD' undeclared here (not in a function)
...

Fix by undefining WITH_GNUTLS_GCRYPT in config-post.h

After commit 3e2f27e1, I've noticed build failures of virt-login-shell
when libapparmor-devel is installed on the build host

CCLD     virt-login-shell
../src/.libs/libvirt-setuid-rpc-client.a(libvirt_setuid_rpc_client_la-vircommand.o):
In function `virExec':
/home/jfehlig/virt/upstream/libvirt/src/util/vircommand.c:653: undefined
reference to `aa_change_profile'
collect2: error: ld returned 1 exit status

I was about to commit an easy fix under the build-breaker rule
(build-fix-1.patch), but thought to extend the notion of SECDRIVER_LIBS
to SECDRIVER_CFLAGS, and use both throughout src/Makefile.am where it
makes sense (build-fix-2.patch).

Should I just stick with the simple fix, or is something along the lines
of patch 2 preferred?

Regards,
Jim

>From a0f35945f3127ab70d051101037e821b1759b4bb Mon Sep 17 00:00:00 2001
From: Jim Fehlig <jfehlig@suse.com>
Date: Mon, 21 Oct 2013 15:30:02 -0600
Subject: [PATCH] build: fix virt-login-shell build with apparmor

With libapparmor-devel installed, virt-login-shell fails to link

CCLD     virt-login-shell
../src/.libs/libvirt-setuid-rpc-client.a(libvirt_setuid_rpc_client_la-vircommand.o): In function `virExec':
/home/jfehlig/virt/upstream/libvirt/src/util/vircommand.c:653: undefined reference to `aa_change_profile'
collect2: error: ld returned 1 exit status

Fix by linking libvirt_setuid_rpc_client with previously determined
SECDRIVER_LIBS in src/Makefile.am.  While at it, introduce SECDRIVER_CFLAGS
and use both throughout src/Makefile.am where it makes sense.
Signed-off-by: NJim Fehlig <jfehlig@suse.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>
(cherry picked from commit 3e2f27e1)

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>
(cherry picked from commit b7fcc799)

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>
(cherry picked from commit 8c3586ea)

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>
(cherry picked from commit ae53e5d1)

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>
(cherry picked from commit 57687fd6)