Peter Krempa [Thu, 14 Jun 2012 08:29:36 +0000 (10:29 +0200)]
qemu: Fix off-by-one error while unescaping monitor strings
While unescaping the commands the commands passed through to the monitor
function qemuMonitorUnescapeArg() initialized lenght of the input string
to strlen()+1 which is fine for alloc but not for iteration of the
string.
This patch fixes the off-by-one error and drops the pointless check for
a single trailing slash that is automaticaly handled by the default
branch of switch.
(cherry picked from commit 0f4660c8787cc41fe67f869984c0ae11d680037e)
Peter Krempa [Wed, 13 Jun 2012 14:55:51 +0000 (16:55 +0200)]
virsh: Don't generate invalid XML in attach-disk command
The attach-disk command used with parameter --cache created an invalid
XML snippet as the beginning of the <driver> element was not printed
when used solely with --cache and no other attribute to driver.
(cherry picked from commit 5b4740265cf1e9cd00165e54e1f5e81f9008bc63)
Eric Blake [Mon, 11 Jun 2012 20:49:03 +0000 (14:49 -0600)]
python: fix snapshot listing bugs
Python exceptions are different than libvirt errors, and we had
some corner case bugs on OOM situations.
* python/libvirt-override.c (libvirt_virDomainSnapshotListNames)
(libvirt_virDomainSnapshotListChildrenNames): Use correct error
returns, avoid segv on OOM, and avoid memory leaks on error.
(cherry picked from commit a0de5d78ef593958bcb83e563fe54137fe43c4ff)
Eric Blake [Mon, 11 Jun 2012 20:32:47 +0000 (14:32 -0600)]
python: use simpler methods
* python/libvirt-override.c (libvirt_virDomainGetVcpus)
(libvirt_virDomainGetVcpuPinInfo): Use Py_XDECREF instead of
open-coding it.
(cherry picked from commit 8566618f65d983b31efb8efec13c25c6410b1718)
Michal Privoznik [Mon, 11 Jun 2012 13:57:19 +0000 (15:57 +0200)]
qemu: Don't overwrite security labels
Currently, if qemuProcessStart fail at some point, e.g. because
domain being started wants a PCI/USB device already assigned to
a different domain, we jump to cleanup label where qemuProcessStop
is performed. This unconditionally calls virSecurityManagerRestoreAllLabel
which is wrong because the other domain is still using those devices.
However, once we successfully label all devices/paths in
qemuProcessStart() from that point on, we have to perform a rollback
on failure - that is - we have to virSecurityManagerRestoreAllLabel.
(cherry picked from commit 86032b2276ace5a7977aad2bbae73b4c33e31914)
Michal Privoznik [Mon, 11 Jun 2012 13:20:44 +0000 (15:20 +0200)]
qemuProcessStop: Switch to flags
Currently, we are passing only one boolean (migrated) so there is
no real profit in this. But it creates starting position for
next patch.
(cherry picked from commit 69dd77149cacc6b12740bd70b729b6cb7506f4e0)
Laine Stump [Mon, 11 Jun 2012 21:16:17 +0000 (17:16 -0400)]
docs: minor fixes to domain interface documentation
A few examples for <interface> had a type='direct' interface with no
sub-elements. This is not allowed - a type='direct' interface must
have at least a source element. (Most likely the example was copied
from the type='user' or type='ethernet' examples - they *do* allow an
instance with no sub-elements).
There was also one place that mistakenly used %lt; ... %gt; instead of
< ... > (for some reason, I make that typo all the time).
(cherry picked from commit 549741ee44c9d3dcaf0231d2f09838f46ac1d74d)
Eric Blake [Thu, 7 Jun 2012 19:23:49 +0000 (13:23 -0600)]
docs: link to FLOSS Weekly podcast, virt blogs
Eric Blake and Guido Günther were guests during this week's
FLOSS Weekly podcast, giving insights into libvirt as a Free
Software project. Also, there are several useful blogs on
virt-related topics.
Eric Blake [Fri, 8 Jun 2012 03:46:47 +0000 (21:46 -0600)]
snapshot: avoid virsh crash with older servers
Commits 51082301, 16d7b39, and 521cc447 introduced support for
'virsh snapshot-list --from' when talking to a server older than
0.9.5, but broke support for plain 'virsh snapshot-list' for the
same old server in the process. Because the code is not properly
gated, we end up with a SIGSEGV during a strcmp with a NULL argument.
* tools/virsh.c (cmdSnapshotList): Don't waste time on fallbacks
when --from is not present.
(cherry picked from commit d9f606659829e666803d95d05aee9ca1ecced955)
Li Zhang [Tue, 29 May 2012 08:35:17 +0000 (16:35 +0800)]
Assign correct address type to spapr-vlan and spapr-vty.
For pseries guest, spapr-vlan and spapr-vty is based
on spapr-vio address. According to model of network
device, the address type should be assigned automatically.
For serial device, serial pty device is recognized as
spapr-vty device, which is also on spapr-vio.
So this patch is to correct the address type of
spapr-vlan and spapr-vty, and build correct
command line of spapr-vty.
Signed-off-by: Li Zhang <zhlcindy@linux.vnet.ibm.com> Reviewed-by: Michael Ellerman<michaele@au1.ibm.com>
(cherry picked from commit 04a319ba4e5f82e56af47e18dc055e5735ddd74d)
Eric Blake [Thu, 7 Jun 2012 16:48:05 +0000 (10:48 -0600)]
maint: make it easier to copy FORTIFY_SOURCE snippet
While libvirt intentionally avoids -Wundef (after all, C99
guarantees sane semantics of treating undefined macros as 0),
the glibc insanity of #warning on _FORTIFY_SOURCE coupled with
what some people feel is the black magic of autoconf means
that other projects are likely to copy our snippet verbatim.
We can be nicer to other projects by making it easier to
integrate into projects that use -Wundef.
Suggested by Christophe Fergeau.
* m4/virt-compile-warnings.m4 (LIBVIRT_COMPILE_WARNINGS): Be nice
to other projects using -Wundef.
(cherry picked from commit 1c2edf0fbd4d20b3955cd5a3d8986413e94fba51)
Eric Blake [Sat, 26 May 2012 02:17:01 +0000 (20:17 -0600)]
command: avoid potential deadlock on handshake
There is a theoretical problem of an extreme bug where we can get
into deadlock due to command handshaking. Thanks to a pair of pipes,
we have a situation where the parent thinks the child reported an
error and is waiting for a message from the child to explain the
error; but at the same time the child thinks it reported success
and is waiting for the parent to acknowledge the success; so both
processes are now blocked.
Thankfully, I don't think this deadlock is possible without at
least one other bug in the code, but I did see exactly that sort
of situation prior to commit da831af - I saw a backtrace where a
double close bug in the parent caused the parent to read from the
wrong fd and assume the child failed, even though the child really
sent success.
This potential deadlock is not quite like commit 858c247 (a deadlock
due to multiple readers on one pipe preventing a write from completing),
although the solution is similar - always close unused pipe fds before
blocking, rather than after.
Eric Blake [Thu, 31 May 2012 21:50:07 +0000 (15:50 -0600)]
command: avoid deadlock on EPIPE situation
It is possible to deadlock libvirt by having a domain with XML
longer than PIPE_BUF, and by writing a hook script that closes
stdin early. This is because libvirt was keeping a copy of the
child's stdin read fd open, which means the write fd in the
parent will never see EPIPE (remember, libvirt should always be
run with SIGPIPE ignored, so we should never get a SIGPIPE signal).
Since there is no error, libvirt blocks waiting for a write to
complete, even though the only reader is also libvirt. The
solution is to ensure that only the child can act as a reader
before the parent does any writes; and then dealing with the
fallout of dealing with EPIPE.
Thankfully, this is not a security hole - since the only way to
trigger the deadlock is to install a custom hook script, anyone
that already has privileges to install a hook script already has
privileges to do any number of other equally disruptive things
to libvirt; it would only be a security hole if an unprivileged
user could install a hook script to DoS a privileged user.
* src/util/command.c (virCommandRun): Close parent's copy of child
read fd earlier.
(virCommandProcessIO): Don't let EPIPE be fatal; the child may
be done parsing input.
* tests/commandhelper.c (main): Set up a SIGPIPE situation.
* tests/commandtest.c (test20): Trigger it.
* tests/commanddata/test20.log: New file.
(cherry picked from commit 858c2476d9dccf630ad5b7f4d8e0e61e9295c2b6)
Eric Blake [Mon, 4 Jun 2012 17:33:20 +0000 (11:33 -0600)]
build: allow building with newer glibc-headers and -O0
glibc 2.15 (on Fedora 17) coupled with explicit disabling of
optimization during development dies a painful death:
In file included from /usr/include/limits.h:27:0,
from /usr/lib/gcc/x86_64-redhat-linux/4.7.0/include/limits.h:169,
from /usr/lib/gcc/x86_64-redhat-linux/4.7.0/include/syslimits.h:7,
from /usr/lib/gcc/x86_64-redhat-linux/4.7.0/include/limits.h:34,
from util/bitmap.c:26:
/usr/include/features.h:314:4: error: #warning _FORTIFY_SOURCE requires compiling with optimization (-O) [-Werror=cpp]
cc1: all warnings being treated as errors
Work around this by only conditionally defining _FORTIFY_SOURCE,
in the case where glibc can actually use it. The trick is using
AH_VERBATIM instead of AC_DEFINE.
* m4/virt-compile-warnings.m4 (LIBVIRT_COMPILE_WARNINGS): Squelch
_FORTIFY_SOURCE when needed to avoid glibc #warnings.
(cherry picked from commit 2af63b1c349114df98c163a8401fd9cf2facdabe)
Radu Caragea [Mon, 4 Jun 2012 07:32:37 +0000 (09:32 +0200)]
Fix sync issue in virNetClientStreamEventRemoveCallback
The stream lock is unlocked twice instead of being locked and then
unlocked. Probably a typo.
(cherry picked from commit 107f51b69c44b9902ed76ab0e33e89dde5d21cb1)
The problem is that an interface with type='hostdev' will have an
alias of the form "hostdev%d", while the function that looks through
existing netdevs to determine the name to use for a new addition will
fail if there's an existing entry that does not match the form
"net%d".
This is another of the handful of places that need an exception due to
the hybrid nature of <interface type='hostdev'> (which is not exactly
an <interface> or a <hostdev>, but is both at the same time).
(cherry picked from commit 6734ce7bc8ec4da7900080aa17fef20c68d401ef)
tools: make virt-pki-validate work with acls and xattrs
This patch makes virt-pki-validate work with certificates that have
acl or xattr set. Otherwise it failing due to wrong permissions.
(cherry picked from commit d4fb6694a4b7d5c613136e3bfc1bbf8d6b98ec74)
Wen Congyang [Wed, 30 May 2012 09:20:44 +0000 (17:20 +0800)]
qemu: avoid closing fd more than once
If we migrate to fd, spec->fwdType is not MIGRATION_FWD_DIRECT,
we will close spec->dest.fd.local in qemuMigrationRun(). So we
should set spec->dest.fd.local to -1 in qemuMigrationRun().
Eric Blake [Wed, 30 May 2012 15:20:37 +0000 (09:20 -0600)]
fdstream: avoid double close bug
Wen Congyang reported that we have a double-close bug if we fail
virFDStreamOpenInternal, since childfd duplicated one of the fds[]
array contents. In truth, since we always transfer both members
of fds to other variables, we should close the fds through those
other names, and just use fds[] for pipe().
* src/fdstream.c (virFDStreamOpenFileInternal): Swap scope of
childfd and fds[], to avoid a double close.
(cherry picked from commit f3cfc7c8848006b613e09bb1acd081716854527e)
Eric Blake [Tue, 29 May 2012 23:47:58 +0000 (17:47 -0600)]
command: avoid double close bugs
KAMEZAWA Hiroyuki reported a nasty double-free bug when virCommand
is used to convert a string into input to a child command. The
problem is that the poll() loop of virCommandProcessIO would close()
the write end of the pipe in order to let the child see EOF, then
the caller virCommandRun() would also close the same fd number, with
the second close possibly nuking an fd opened by some other thread
in the meantime. This in turn can have all sorts of bad effects.
The bug has been present since the introduction of virCommand in
commit f16ad06f.
This is based on his first attempt at a patch, at
https://bugzilla.redhat.com/show_bug.cgi?id=823716
* src/util/command.c (_virCommand): Drop inpipe member.
(virCommandProcessIO): Add argument, to avoid closing caller's fd
without informing caller.
(virCommandRun, virCommandNewArgs): Adjust clients.
(cherry picked from commit da831afcf2f6c0d3ed1ea3128a6208f548a05d8f)
Wen Congyang [Wed, 30 May 2012 09:20:46 +0000 (17:20 +0800)]
avoid fd leak
virCommandRunAsync() will set errfd if it succeed. We should
close it if virFDStreamOpenInternal() fails.
(cherry picked from commit 655cffa0f25be40c68624036af80bde102177612)
Eric Blake [Tue, 29 May 2012 14:10:44 +0000 (08:10 -0600)]
build: silence warning from autoconf
Autoconf 2.60 and later insist on using ${datarootdir}, rather than
the derived ${datadir} (although the latter defaults to the former,
it is possible to set configure arguments so that they differ):
config.status: creating libvirt.pc
config.status: WARNING: 'libvirt.pc.in' seems to ignore the --datarootdir setting
This patch follows the autoconf manual's suggestions for how to
support 2.59 (RHEL 5) and newer simultaneously.
* libvirt.pc.in (datarootdir): Define, so ${datadir} will not ignore
datarootdir when using newer autoconf.
(cherry picked from commit aabf77aa41e9589faa7ce44add87b97f5a6dde2e)
Michal Privoznik [Wed, 30 May 2012 12:17:26 +0000 (14:17 +0200)]
virCommand: Extend debug message for handshake
Currently, we are logging only one side of pipes we
create in virCommandRequireHandshake(); This is enough
in cases where pipe2() returns two consecutive FDs. However,
it is not guaranteed and it may return any FDs.
Therefore, it's wise to log the other ends as well.
(cherry picked from commit 7454849ec5489c7b52d3b41a8022f2bc6fda9eee)
When getting number of CPUs the host has assigned, there was always
number "1" returned. Even though all lxc domains with no pinning
launched by libvirt run on all pCPUs (by default, no matter what's the
number), we should at least return the same number as the user
specified when creating the domain.
(cherry picked from commit 87dfdb0b92f234a08408332095454260f1c86917)
Dave Allan [Tue, 29 May 2012 22:49:13 +0000 (16:49 -0600)]
examples: add consolecallback example python script
A while back I wrote the attached code to demonstrate how to use
events and serial console to create a serial console that stays up
even when the VM is down. It might need some work, as I am not
terribly strong with Python.
* examples/python/consolecallback.py: New file.
* examples/python/Makefile.am (EXTRA_DIST): Ship it.
(cherry picked from commit ffc9f7ab4f203ce6825dbdc22b75f0894ae833b2)
Stefan Berger [Tue, 29 May 2012 10:25:59 +0000 (06:25 -0400)]
leak_fix.diff
==3240== 23 bytes in 1 blocks are definitely lost in loss record 242 of 744
==3240== at 0x4C2A4CD: malloc (vg_replace_malloc.c:236)
==3240== by 0x8077537: __vasprintf_chk (vasprintf_chk.c:82)
==3240== by 0x509C677: virVasprintf (stdio2.h:199)
==3240== by 0x509C733: virAsprintf (util.c:1912)
==3240== by 0x1906583A: qemudStartup (qemu_driver.c:679)
==3240== by 0x511991D: virStateInitialize (libvirt.c:809)
==3240== by 0x40CD84: daemonRunStateInit (libvirtd.c:751)
==3240== by 0x5098745: virThreadHelper (threads-pthread.c:161)
==3240== by 0x7953D8F: start_thread (pthread_create.c:309)
==3240== by 0x805FF5C: clone (clone.S:115)
"accel3d" was specified twice, the second one is obviously "accel2d"
(also from the text down below.
(cherry picked from commit 94618eb66ac79cae49bd564f397ae3a4f1e49f98)
When you try to connect to a socket in the abstract namespace,
the error will be ECONNREFUSED for a non-listening daemon. With
the non-abstract namespace though, you instead get ENOENT. Add
a check for this extra errno when auto-spawning the daemon
Jim Meyering [Sat, 26 May 2012 09:21:47 +0000 (11:21 +0200)]
maint: avoid new automake warning about AM_PROG_CC_STDC
* configure.ac (AM_PROG_CC_STDC): Stop using this macro.
It provokes warnings from newer automake and is superseded by
autoconf's AC_PROG_CC, which we're already using.
(cherry picked from commit 0d07aa512c775d9b4e8460f203898364ba5413dd)
Add a note about setting the LIBVIRT_DRIVER_DIR env variable,
explain --system and fix example to use --disable-werror
(cherry picked from commit 4878a33125a0793b8af3a1f9837bb1c396a80420)
Eric Blake [Wed, 24 Nov 2010 20:41:50 +0000 (13:41 -0700)]
tests: run valgrind on real executables, not libtool wrappers
* tests/Makefile.am (valgrind): Use libtool's ability to bypass
libtool wrappers when running valgrind.
(cherry picked from commit 428fc2bf3138136de2f340e06fae287a37d3f5f2)
Add parsing for VIR_ENUM_IMPL & VIR_ENUM_DECL in apibuild.py
The apibuild.py parser needs to be able to parse & ignore
any VIR_ENUM_IMPL/VIR_ENUM_DECL macros in the source. Add
some special case code to deal with this rather than trying
to figure out a generic syntax for parsing macros.
The libvirt-lock-sanlock RPM requires libvirtd, so its RPM dep
should be on libvirt-daemon, not libvirt
(cherry picked from commit aad1625da693f765af60ab0a184b23069cb5cb4e)
Adding syslog.target is obsolete, avahi.target does not
exist and dbus.target is also obsolete
Reported-by: Lennart Poettering <lpoetter@redhat.com> Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
(cherry picked from commit 4c7973e184a4ce52f78e63f1a92f9597709279ab)
Jiri Denemark [Mon, 21 May 2012 14:02:05 +0000 (16:02 +0200)]
Revert "rpc: Discard non-blocking calls only when necessary"
This reverts commit b1e374a7ac56927cfe62435179bf0bba1e08b372, which was
rather bad since I failed to consider all sides of the issue. The main
things I didn't consider properly are:
- a thread which sends a non-blocking call waits for the thread with
the buck to process the call
- the code doesn't expect non-blocking calls to remain in the queue
unless they were already partially sent
Thus, the reverted patch actually breaks more than what it fixes and
clients (which may even be libvirtd during p2p migrations) will likely
end up in a deadlock.
(cherry picked from commit 63643f67abcdeaa33a0f85ea8e54da75ea9908e4)
Peter Krempa [Mon, 21 May 2012 14:31:53 +0000 (16:31 +0200)]
qemu_hotplug: Don't free the PCI device structure after hot-unplug
The pciDevice structure corresponding to the device being hot-unplugged
was freed after it was "stolen" from activeList. The pointer was still
used for eg-inactive list. This patch removes the free of the structure
and frees it only if reset fails on the device.
(cherry picked from commit db19417fc012416639c2230e5f19717b84245ce5)
Fix potential events deadlock when unref'ing virConnectPtr
When the last reference to a virConnectPtr is released by
libvirtd, it was possible for a deadlock to occur in the
virDomainEventState functions. The virDomainEventStatePtr
holds a reference on virConnectPtr for each registered
callback. When removing a callback, the virUnrefConnect
function is run. If this causes the last reference on the
virConnectPtr to be released, then virReleaseConnect can
be run, which in turns calls qemudClose. This function has
a call to virDomainEventStateDeregisterConn which is intended
to remove all callbacks associated with the virConnectPtr
instance. This will try to grab a lock on virDomainEventState
but this lock is already held. Deadlock ensues
Thread 1 (Thread 0x7fcbb526a840 (LWP 23185)):
Since each callback associated with a virConnectPtr holds a
reference on virConnectPtr, it is impossible for the qemudClose
method to be invoked while any callbacks are still registered.
Thus the call to virDomainEventStateDeregisterConn must in fact
be a no-op. Thus it is possible to just remove all trace of
virDomainEventStateDeregisterConn and avoid the deadlock.
* src/conf/domain_event.c, src/conf/domain_event.h,
src/libvirt_private.syms: Delete virDomainEventStateDeregisterConn
* src/libxl/libxl_driver.c, src/lxc/lxc_driver.c,
src/qemu/qemu_driver.c, src/uml/uml_driver.c: Remove
calls to virDomainEventStateDeregisterConn
(cherry picked from commit 2cb0899eec72376629a0583647dcad39b00c5715)
Jim Fehlig [Mon, 21 May 2012 15:23:41 +0000 (09:23 -0600)]
Fix build when configuring with polkit0
Commit 2223ea98 removed the only use of 'server' param in
remoteDispatchAuthPolkit(). Mark the parameter with ATTRIBUTE_UNUSED
to fix the build when configuring with polkit0.
(cherry picked from commit 651d7124529b997031d525ea30bbee5d112c2e88)
Eric Blake [Fri, 18 May 2012 15:42:25 +0000 (09:42 -0600)]
build: fix virnetlink on glibc 2.11
We were being lazy - virnetlink.c was getting uint32_t as a
side-effect from glibc 2.14's <unistd.h>, but older glibc 2.11
does not provide uint32_t from <unistd.h>. In fact, POSIX states
that <unistd.h> need only provide intptr_t, not all of <stdint.h>,
so the bug really is ours. Reported by Jonathan Alescio.
Michal Privoznik [Thu, 17 May 2012 11:40:52 +0000 (13:40 +0200)]
qemu: Don't delete USB device on failed qemuPrepareHostdevUSBDevices
If qemuPrepareHostdevUSBDevices fail it will roll back devices added
to the driver list of used devices. However, if it may fail because
the device is being used already. But then again - with roll back.
Therefore don't try to remove a usb device manually if the function
fail. Although, we want to remove the device if any operation
performed afterwards fail.
(cherry picked from commit 9c484e3dc5464dfbb538744360b401a0bc59c1c6)
Michal Privoznik [Wed, 16 May 2012 14:42:02 +0000 (16:42 +0200)]
qemu: Rollback on used USB devices
One of our latest USB device handling patches 05abd1507d66aabb6cad12eeafeb4c4d1911c585 introduced a regression.
That is, we first create a temporary list of all USB devices that
are to be used by domain just starting up. Then we iterate over and
check if a device from the list is in the global list of currently
assigned devices (activeUsbHostdevs). If not, we add it there and
continue with next iteration then. But if a device from temporary
list is either taken already or adding to the activeUsbHostdevs fails,
we remove all devices in temp list from the activeUsbHostdevs list.
Therefore, if a device is already taken we remove it from
activeUsbHostdevs even if we should not. Thus, next time we allow
the device to be assigned to another domain.
(cherry picked from commit 2f5fdc886ec7ed8b871ebd0576271f8ee5be1f71)
There is no 'udev.target' unit in systemd (only 'udev.service')
yet libvirtd's unit file had a dep on one. There's no compelling
reason for a dep on udev, so remove it altogether.
Set a sensible default master start port for ehci companion controllers
The uhci1, uhci2, uhci3 companion controllers for ehci1 must
have a master start port set. Since this value is predictable
we should set it automatically if the app does not supply it
(cherry picked from commit 03b804a200e214e3faf264b16f41a640c1344962)
(crobinso: Drop an element that wasn't backported to -maint branch.)
Fix logic for assigning PCI addresses to USB2 companion controllers
Currently each USB2 companion controller gets put on a separate
PCI slot. Not only is this wasteful of PCI slots, but it is not
in compliance with the spec for USB2 controllers. The master
echi1 and all companion controllers should be in the same slot,
with echi1 in function 7, and uhci1-3 in functions 0-2 respectively.
* src/qemu/qemu_command.c: Special case handling of USB2 controllers
to apply correct pci slot assignment
* tests/qemuxml2argvdata/qemuxml2argv-usb-ich9-ehci-addr.args,
tests/qemuxml2argvdata/qemuxml2argv-usb-ich9-ehci-addr.xml: Expand
test to cover automatic slot assignment
(cherry picked from commit 1ebd52cb871f87b7868503b28448e96d59e41d63)
Fix virDomainDeviceInfoIsSet() to check all struct fields
The virDomainDeviceInfoIsSet API was only checking if an
address or alias was set in the struct. Thus if only a
rom bar setting / filename, boot index, or USB master
value was set, they could be accidentally dropped when
formatting XML
(cherry picked from commit 2c195fdbf3293792e39bc4f06755ae751025b0ea)
Allow stack traces to be included with log messages
Sometimes it is useful to see the callpath for log messages.
This change enhances the log filter syntax so that stack traces
can be show by setting '1:+NAME' instead of '1:NAME'.
* docs/logging.html.in: Document new syntax
* configure.ac: Check for execinfo.h
* src/util/logging.c, src/util/logging.h: Add support for
stack traces
* tests/testutils.c: Adapt to API change
This patch is to fix the problem by using the CPU index in
caps->host.numaCell[i]->cpus[i] to set the bitmask instead of
assuming the CPU index of the NUMA nodes are always sequential.
(cherry picked from commit d1bdeca875d92c6fe9c9eb854fa1f3d891fef84a)
Matthias Bolte [Sun, 6 May 2012 17:33:59 +0000 (19:33 +0200)]
esx: Fix memory leaks in error paths related to transferred ownership
Appending an item to a list transfers ownership of that item to the
list owner. But an error can occur in between item allocation and
appending it to the list. In this case the item has to be freed
explicitly. This was not done in some special cases resulting in
possible memory leaks.
Peter Krempa [Mon, 7 May 2012 11:58:22 +0000 (13:58 +0200)]
qemu: Don't skip detection of virtual cpu's on non KVM targets
This patch lifts the limit of calling thread detection code only on KVM
guests. With upstream qemu the thread mappings are reported also on
non-KVM machines.
QEMU adopted the thread_id information from the kvm branch.
To remain compatible with older upstream versions of qemu the check is
attempted but the failure to detect threads (or even run the monitor
command - on older versions without SMP support) is treated non-fatal
and the code reports one vCPU with pid of the hypervisor (in same
fashion this was done on non-KVM guests).
(cherry picked from commit c833526924d56e14f2d0133804cf209f5f47a228)
Peter Krempa [Mon, 7 May 2012 11:56:17 +0000 (13:56 +0200)]
qemu: Re-detect virtual cpu threads after cpu hot (un)plug.
After a cpu hotplug the qemu driver did not refresh information about
virtual processors used by qemu and their corresponding threads. This
patch forces a re-detection as is done on start of QEMU.
This ensures that correct information is reported by the
virDomainGetVcpus API and "virsh vcpuinfo".
A failure to obtain the thread<->vcpu mapping is treated non-fatal and
the mapping is not updated in a case of failure as not all versions of
QEMU report this in the info cpus command.
(cherry picked from commit 3163682b585da7e894c9ea069741755fc316bdb3)
Peter Krempa [Mon, 7 May 2012 11:53:20 +0000 (13:53 +0200)]
qemu: Refactor qemuDomainSetVcpusFlags
This patch changes a switch statement into ifs when handling live vs.
configuration modifications getting rid of redundant code in case when
both live and persistent configuration gets changed.
(cherry picked from commit e99ad93d028cf7a63fb9ffd7a620dc00b0d54fce)
Guannan Ren [Fri, 11 May 2012 06:29:15 +0000 (14:29 +0800)]
usb: fix crash when failing to attach a second usb device
when failing to attach another usb device to a domain for some reason
which has one use device attached before, the libvirtd crashed.
The crash is caused by null-pointer dereference error in invoking
usbDeviceListSteal passed in NULL value usb variable.
commit 05abd1507d66aabb6cad12eeafeb4c4d1911c585 introduces the bug.
(cherry picked from commit ab5fb8f34c93661bb19b62e4ed3592fb53cd6b36)
Eric Blake [Mon, 30 Apr 2012 20:51:07 +0000 (14:51 -0600)]
docs: mention migration issue of which credentials are used
Based on a report by Seth Vidal. Just because _you_ can use virsh
to connect to both source and destinations does not mean that libvirtd
on the source (aka _root_) can likewise connect to the destination;
this matters when setting up a peer-to-peer migration instead of a
native one.
* docs/migration.html.in: Mention that in peer-to-peer, the owner
of the source libvirtd (usually root) must be able to connect to
the destination.
(cherry picked from commit 38bd605b71d43e951ebfcb247f82c0435bad397b)
error: Failed to define domain from x.xml
error: An error occurred, but the cause is unknown
With this small patch, the error is more helpful
error: Failed to define domain from x.xml
error: XML error: unknown target type 'kvm' specified for character device
(cherry picked from commit 10a87145f73068e572d14d82a13df46c95960273)
Osier Yang [Wed, 9 May 2012 04:22:58 +0000 (12:22 +0800)]
numad: Check numactl-devel if compiled with numad support
Since now we pre-set memory policy using libnuma to fully
drive numad, it needs to check numactl-devel if "with_numad"
is "yes".
configure with groups "--with-numad=yes --with-numactl=yes",
"--with-numad=no --with-numactl=yes", "--with-numad=yes
--with-numactl=yes" works fine after the change.
(cherry picked from commit b0f3244554b7b67183615a353897a2e19e0c6d68)
Eric Blake [Mon, 7 May 2012 22:56:08 +0000 (16:56 -0600)]
snapshot: allow block devices past cgroup
It turns out that when cgroups are enabled, the use of a block device
for a snapshot target was failing with EPERM due to libvirt failing
to add the block device to the cgroup whitelist. See also
https://bugzilla.redhat.com/show_bug.cgi?id=810200
* src/qemu/qemu_driver.c
(qemuDomainSnapshotCreateSingleDiskActive)
(qemuDomainSnapshotUndoSingleDiskActive): Account for cgroup.
(qemuDomainSnapshotCreateDiskActive): Update caller.
(cherry picked from commit 8be304ecb9c8294c9ca3ef7324dea8228e496f9c)
Alon Levy [Tue, 8 May 2012 17:46:45 +0000 (20:46 +0300)]
tests: add some self-documentation to tests
Alon tried './qemuxml2argvtest --help' to figure out a test failure,
but it didn't help. The information is in HACKING, but it doesn't
hurt to make the tests also provide their own help.
Jim Meyering [Mon, 7 May 2012 19:22:09 +0000 (21:22 +0200)]
virsh: avoid heap corruption leading to virsh abort
* tools/virsh.c (vshParseSnapshotDiskspec): Fix off-by-3 memmove
that would corrupt heap when parsing escaped --diskspec comma.
Bug introduced via commit v0.9.4-260-g35d52b5.
(cherry picked from commit c6694ab85c207e51c6f39cd958c4323b636d8d8d)
Laine Stump [Fri, 4 May 2012 17:48:20 +0000 (13:48 -0400)]
util: set src_pid for virNetlinkCommand when appropriate
Until now, the nl_pid of the source address of every message sent by
virNetlinkCommand has been set to the value of getpid(). Most of the
time this doesn't matter, and in the one case where it does
(communication with lldpad), it previously was the proper thing to do,
because the netlink event service (which listens on a netlink socket
for unsolicited messages from lldpad) coincidentally always happened
to bind with a local nl_pid == getpid().
that particular nl_pid is now effectively a reserved value, so the
netlink event service will always bind to something else
(coincidentally "getpid() + (1 << 22)", but it really could be
anything). The result is that communication between lldpad and
libvirtd is broken (lldpad gets a "disconnected" error when it tries
to send a directed message).
The solution to this problem caused by a solution, is to query the
netlink event service's nlhandle for its "local_port", and send that
as the source nl_pid (but only when sending to lldpad, of course - in
other cases we maintain the old behavior of sending getpid()).
There are two cases where a message is being directed at lldpad - one
in virNetDevLinkDump, and one in virNetDevVPortProfileOpSetLink.
The case of virNetDevVPortProfileOpSetLink is simplest to explain -
only if !nltarget_kernel, i.e. the message isn't targetted for the
kernel, is the dst_pid set (by calling
virNetDevVPortProfileGetLldpadPid()), so only in that case do we call
virNetlinkEventServiceLocalPid() to set src_pid.
For virNetDevLinkDump, it's a bit more complicated. The call to
virNetDevVPortProfileGetLldpadPid() was effectively up one level (in
virNetDevVPortProfileOpCommon), although obscured by an unnecessary
passing of a function pointer. This patch removes the function
pointer, and calls virNetDevVPortProfileGetLldpadPid() directly in
virNetDevVPortProfileOpCommon - if it's doing this, it knows that it
should also call virNetlinkEventServiceLocalPid() to set src_pid too;
then it just passes src_pid and dst_pid down to
virNetDevLinkDump. Since (src_pid == 0 && dst_pid == 0) implies that
the kernel is the destination, there is no longer any need to send
nltarget_kernel as an arg to virNetDevLinkDump, so it's been removed.
The disparity between src_pid being int and dst_pid being uint32_t may
be a bit disconcerting to some, but I didn't want to complicate
virNetlinkEventServiceLocalPid() by having status returned separately
from the value.
(cherry picked from commit cc0737713a8e1ad37fc9d5393b4ea514f7708138)
Note that libnl's port generation algorithm guarantees that the
nl_socket_get_local_port() will always be > 0 (since it is "getpid() +
(n << 22>" where n is always < 1024), so it is okay to cast the
uint32_t to int (thus allowing us to use -1 as an error sentinel).
(cherry picked from commit c99e93758d44e7c6d62dbd3d064b6df61d1fefb2)
Laine Stump [Fri, 4 May 2012 16:58:36 +0000 (12:58 -0400)]
util: allow specifying both src and dst pid in virNetlinkCommand
Until now, virNetlinkCommand has assumed that the nl_pid in the source
address of outgoing netlink messages should always be the return value
of getpid(). In most cases it actually doesn't matter, but in the case
of communication with lldpad, lldpad saves this info and later uses it
to send netlink messages back to libvirt. A recent patch to fix Bug
816465 changed the order of the universe such that the netlink event
service socket is no longer bound with nl_pid == getpid(), so lldpad
could no longer send unsolicited messages to libvirtd. Adding src_pid
as an argument to virNetlinkCommand() is the first step in notifying
lldpad of the proper address of the netlink event service socket.
(cherry picked from commit cca7bb1fb583459824f7a42be0406d0833a80593)
A short explanation is included in the comments of the patch itself.
Note that this patch by itself breaks communication between lldpad and
libvirtd, so the other 3 patches in the series must be applied at the
same time as this patch.
(cherry picked from commit 642973135c54b93242c4548ef27d591b52b0994c)
Guannan Ren [Sun, 6 May 2012 14:45:05 +0000 (22:45 +0800)]
qemu: call usb search function for hostdev initialization and hotplug
src/qemu/qemu_hostdev.c:
refactor qemuPrepareHostdevUSBDevices function, make it focus on
adding usb device to activeUsbHostdevs after check. After that,
the usb hotplug function qemuDomainAttachHostDevice also could use
it.
expand qemuPrepareHostUSBDevices to perform the usb search,
rollback on failure.
src/qemu/qemu_hotplug.c:
If there are multiple usb devices available with same vendorID and productID,
but with different value of "bus, device", we give an error to let user
use <address> to specify the desired one.
(cherry picked from commit 05abd1507d66aabb6cad12eeafeb4c4d1911c585)
Jiri Denemark [Fri, 4 May 2012 19:23:17 +0000 (21:23 +0200)]
qemu: Emit compatible XML when migrating a domain
When we added the default USB controller into domain XML, we efficiently
broke migration to older versions of libvirt that didn't support USB
controllers at all (0.9.4 and earlier) even for domains that don't use
anything that the older libvirt can't provide. We still want to present
the default USB controller in any XML seen by a user/app but we can
safely remove it from the domain XML used during migration. If we are
migrating to a new enough libvirt, it will add the controller XML back,
while older libvirt won't be confused with it although it will still
tell qemu to create the controller.
Similar approach can be used in the future whenever we find out we
always enabled some kind of device without properly advertising it in
domain XML.
(cherry picked from commit 409b5f549530e7b3a33f4505f2cad2e26896107c)
Jiri Denemark [Fri, 4 May 2012 19:00:13 +0000 (21:00 +0200)]
qemu: Don't use virDomainDefFormat* directly
Always use appropriate qemuDomain{,Def}Format wrapper since it may do
some additional magic based on the flags.
(cherry picked from commit cd603008b1ddde1c549e6efd40fa2ea32fbc8053)
Eric Blake [Fri, 4 May 2012 16:24:17 +0000 (10:24 -0600)]
qemu: reject blockiotune if qemu too old
Commit 4c82f09e added a capability check for qemu per-device io
throttling, but only applied it to domain startup. As mentioned
in the previous commit (98cec05), the user can still get an 'internal
error' message during a hotplug attempt, when the monitor command
doesn't exist. It is confusing to allow tuning on inactive domains
only to then be rejected when starting the domain.
* src/qemu/qemu_driver.c (qemuDomainSetBlockIoTune): Reject
offline tuning if online can't match it.
(cherry picked from commit 13f9a19326a7727e07ab689ea5392255fbd5cd4f)
projects / thirdparty / libvirt.git / log
