Taku Izumi [Fri, 20 Jan 2012 06:30:47 +0000 (15:30 +0900)]
virsh: let domif-{get,set}link take target name
Other virsh domifXXX commands can accept target name
as a parameter to specify interface. From viewpoint of
consistency, virsh domif-getlink command should accept
target name as a parameter. This patch achieves this.
Eric Blake [Fri, 20 Jan 2012 21:00:58 +0000 (14:00 -0700)]
maint: enforce use of _LAST marker
When converting a linear enum to a string, we have checks in
place in the VIR_ENUM_IMPL macro to ensure that there is one
string for every value, which lets us quickly flag if a user
added a value but forgot to add a counterpart string. However,
this only works if we use the _LAST marker.
Eric Blake [Fri, 20 Jan 2012 18:43:28 +0000 (11:43 -0700)]
API: make declaration of _LAST enum values conditional
Although this is a public API break, it only affects users that
were compiling against *_LAST values, and can be trivially
worked around without impacting compilation against older
headers, by the user defining VIR_ENUM_SENTINELS before using
libvirt.h. It is not an ABI break, since enum values do not
appear as .so entry points. Meanwhile, it prevents users from
using non-stable enum values without explicitly acknowledging
the risk of doing so.
See this list discussion:
https://www.redhat.com/archives/libvir-list/2012-January/msg00804.html
* include/libvirt/libvirt.h.in: Hide all sentinels behind
LIBVIRT_ENUM_SENTINELS, and add missing sentinels.
* src/internal.h (VIR_DEPRECATED): Allow inclusion after
libvirt.h.
(LIBVIRT_ENUM_SENTINELS): Expose sentinels internally.
* daemon/libvirtd.h: Use the sentinels.
* src/remote/remote_protocol.x (includes): Don't expose sentinels.
* python/generator.py (enum): Likewise.
* tests/cputest.c (cpuTestCompResStr): Silence compiler warning.
* tools/virsh.c (vshDomainStateReasonToString)
(vshDomainControlStateToString): Likewise.
Eric Blake [Mon, 28 Nov 2011 23:13:40 +0000 (16:13 -0700)]
error: drop old-style error reporting
While we still don't want to enable gcc's new -Wformat-literal
warning, I found a rather easy case where the warning could be
reduced, by getting rid of obsolete error-reporting practices.
This is the last place where we were passing the (unused) net
and conn arguments for constructing an error.
* src/util/virterror_internal.h (virErrorMsg): Delete prototype.
(virReportError): Delete macro.
* src/util/virterror.c (virErrorMsg): Make static.
* src/libvirt_private.syms (virterror_internal.h): Drop export.
* src/util/conf.c (virConfError): Convert to macro.
(virConfErrorHelper): New function, and adjust error calls.
* src/xen/xen_hypervisor.c (virXenErrorFunc): Delete.
(xenHypervisorGetSchedulerType)
(xenHypervisorGetSchedulerParameters)
(xenHypervisorSetSchedulerParameters)
(xenHypervisorDomainBlockStats)
(xenHypervisorDomainInterfaceStats)
(xenHypervisorDomainGetOSType)
(xenHypervisorNodeGetCellsFreeMemory, xenHypervisorGetVcpus):
Update callers.
* src/lxc/lxc_driver.c (lxcSetSchedulerParametersFlags)
(lxcGetSchedulerParametersFlags, lxcDomainSetBlkioParameters)
(lxcDomainGetBlkioParameters): Use helpers.
(lxcDomainSetBlkioParameters): Allow setting live and config at
once.
Eric Blake [Thu, 12 Jan 2012 21:14:44 +0000 (14:14 -0700)]
build: silence some compiler warnings from gnulib
Gnulib claims that there are some classes of warnings that are
worth enabling during development, but where silencing those
warnings causes code bloat that is not necessary in an optimized
build. The code bloat to silence the warnings is only enabled
by -Dlint. Follow the lead of coreutils in setting up -Dlint
whenever full warnings are requested.
* m4/virt-compile-warnings.m4 (LIBVIRT_COMPILE_WARNINGS): Add
-Dlint, and move _FORTIFY_SOURCE to config.h instead of CFLAGS.
Eric Blake [Tue, 20 Dec 2011 22:06:47 +0000 (15:06 -0700)]
threads: check for failure to set thread-local value
We had a memory leak on a very arcane OOM situation (unlikely to ever
hit in practice, but who knows if libvirt.so would ever be linked
into some other program that exhausts all thread-local storage keys?).
I found it by code inspection, while analyzing a valgrind report
generated by Alex Jia.
Fix rpc generator to anchor matches for method names
The RPC generator transforms methods matching certain
patterns like 'id' or 'uuid', etc but does not anchor
its matches to the end of the word. So if a method
contains 'id' in the middle (eg virIdentity) then the
RPC generator munges that.
Added capability checking for block <iotune> setting.
There was missing capability for blkiotune and thus specifying these
settings caused libvirt to run qemu with invalid parameters and then
reporting qemu error instead of the standard libvirt one. The support
for blkiotune setting was added in upstream qemu repo under commit 0563e191516289c9d2f282a8c50f2eecef2fa773.
To deal with the fact that the root filesystem can be moved,
we need to resolve symlinks in *any* part of the filesystem
source path.
* src/libvirt_private.syms, src/util/util.c,
src/util/util.h: Add virFileResolveAllLinks to resolve
all symlinks in a path
* src/lxc/lxc_container.c: Resolve all symlinks in filesystem
paths during startup
Osier Yang [Tue, 17 Jan 2012 20:02:05 +0000 (04:02 +0800)]
qemu: Introduce inactive PCI device list
pciTrySecondaryBusReset checks if there is active device on the
same bus, however, qemu driver doesn't maintain an effective
list for the inactive devices, and it passes meaningless argument
for parameter "inactiveDevs". e.g. (qemuPrepareHostdevPCIDevices)
if (!(pcidevs = qemuGetPciHostDeviceList(hostdevs, nhostdevs)))
return -1;
..skipped...
if (pciResetDevice(dev, driver->activePciHostdevs, pcidevs) < 0)
goto reattachdevs;
NB, the "pcidevs" used above are extracted from domain def, and
thus one won't be able to attach a device of which bus has other
device even detached from host (nodedev-detach). To see more
details of the problem:
This patch is to resolve the problem by introducing an inactive
PCI device list (just like qemu_driver->activePciHostdevs), and
the whole logic is:
* Add the device to inactive list during nodedev-dettach
* Remove the device from inactive list during nodedev-reattach
* Remove the device from inactive list during attach-device
(for non-managed device)
* Add the device to inactive list after detach-device, only
if the device is not managed
With the above, we have a sufficient inactive PCI device list, and thus
we can use it for pciResetDevice. e.g.(qemuPrepareHostdevPCIDevices)
if (pciResetDevice(dev, driver->activePciHostdevs,
driver->inactivePciHostdevs) < 0)
goto reattachdevs;
Execute bit on *.stp files in examples/systemtap/ caused dependency when
building RPM packages. Disabling execute permission should help the auto
dependency resolver to see that systemtap is not needed.
Deepak C Shetty [Tue, 17 Jan 2012 12:44:18 +0000 (18:14 +0530)]
Add new attribute wrpolicy to <driver> element
This introduces new attribute wrpolicy with only supported
value as immediate. This will be an optional
attribute with no defaults. This helps specify whether
to skip the host page cache.
When wrpolicy is specified, meaning when wrpolicy=immediate
a writeback is explicitly initiated for the dirty pages in
the host page cache as part of the guest file write operation.
Jiri Denemark [Tue, 17 Jan 2012 15:21:38 +0000 (16:21 +0100)]
qemu: Don't break domain with 0:0:2.0 assigned to anything but VGA
In the past we didn't reserve 0:0:2.0 PCI address if there was no video
device assigned to a domain, which made it impossible to add a video
device later on. So we fixed it (commit v0.9.0-37-g7b2cac1) by always
reserving that address. However, that breaks existing domains without
video devices that already have another device assigned to the
problematic address.
This patch reserves address 0:0:2.0 only in case it was not explicitly
assigned to another device, which means libvirt will try to keep this
address free and will not automatically assign it new devices. But
existing domains for which older libvirt already assigned the address to
a non-video device will keep working as they used to work before 0.9.1.
Moreover, users who want to create a domain without a video device and
use its address for another device may do so by explicitly configuring
the PCI address in domain XML.
Eric Blake [Tue, 17 Jan 2012 19:09:51 +0000 (12:09 -0700)]
tests: avoid dirname in tests
qemuxml2argvtest sanitizes PATH to just /bin, but on at least
Fedora 16, dirname lives in /usr/bin instead. Regression
introduced in commit e7201afd.
* tests/qemuxml2argvdata/qemu.sh: Avoid forking a dirname call,
since dirname might not be in PATH after test sanitization.
* tests/qemuxml2argvdata/qemu-supported-cpus.sh: Likewise.
Diagnosed by Michal Privoznik.
Stefan Berger [Tue, 17 Jan 2012 17:47:41 +0000 (12:47 -0500)]
nwfilter: fix typing error in filter
Fix a typing error in the no-ip-spoofing filter.
Return DHCP request packets passing through this filter. Have
the user use another filter to actually allow DHCP requests to be
sent (action='accept').
Jiri Denemark [Thu, 18 Aug 2011 10:56:56 +0000 (12:56 +0200)]
Taint domains configured with cpu mode=host-passthrough
There are several reasons for doing this:
- the CPU specification is out of libvirt's control so we cannot
guarantee stable guest ABI
- not every feature of a CPU may actually work as expected when
advertised directly to a guest
- migration between two machines with exactly the same CPU may work but
no guarantees can be made
- this mode is not supported and its use is at one's own risk
Jiri Denemark [Mon, 19 Dec 2011 14:41:16 +0000 (15:41 +0100)]
cpu: Update guest CPU in host-* mode
VIR_DOMAIN_XML_UPDATE_CPU flag for virDomainGetXMLDesc may be used to
get updated custom mode guest CPU definition in case it depends on host
CPU. This patch implements the same behavior for host-model and
host-passthrough CPU modes.
- guest CPU should be a copy of host CPU as advertised by capabilities
XML (this is a short cut for manually copying host CPU specification
from capabilities to domain XML):
<cpu mode='host-model'/>
In case a hypervisor does not support the exact host model, libvirt
automatically falls back to a closest supported CPU model and
removes/adds features to match host. This behavior can be disabled by
<cpu mode='host-model'>
<model fallback='forbid'/>
</cpu>
- the same as previous returned by virDomainGetXMLDesc with
VIR_DOMAIN_XML_UPDATE_CPU flag:
<cpu mode='host-model' match='exact'>
<model fallback='allow'>Penryn</model> --+
<vendor>Intel</vendor> |
<topology sockets='2' cores='4' threads='1'/> + copied from
<feature policy='require' name='dca'/> | capabilities XML
<feature policy='require' name='xtpr'/> |
... --+
</cpu>
- guest CPU should be exactly the same as host CPU even in the aspects
libvirt doesn't model (such domain cannot be migrated unless both
hosts contain exactly the same CPUs):
<cpu mode='host-passthrough'/>
- the same as previous returned by virDomainGetXMLDesc with
VIR_DOMAIN_XML_UPDATE_CPU flag:
<cpu mode='host-passthrough' match='minimal'>
<model>Penryn</model> --+ copied from caps
<vendor>Intel</vendor> | XML but doesn't
<topology sockets='2' cores='4' threads='1'/> | describe all
<feature policy='require' name='dca'/> | aspects of the
<feature policy='require' name='xtpr'/> | actual guest CPU
... --+
</cpu>
Jiri Denemark [Wed, 21 Dec 2011 13:27:16 +0000 (14:27 +0100)]
cpu: Optionally forbid fallback CPU models
In case a hypervisor doesn't support the exact CPU model requested by a
domain XML, we automatically fallback to a closest CPU model the
hypervisor supports (and make sure we add/remove any additional features
if needed). This patch adds 'fallback' attribute to model element, which
can be used to disable this automatic fallback.
Jiri Denemark [Mon, 16 Jan 2012 09:24:42 +0000 (10:24 +0100)]
Clarify semantics of virDomainMigrate{,ToURI}2
Commit 5d784bd6d7b19314b0908aec6b46bfe377aeba42 was a nice attempt to
clarify the semantics by requiring domain name from dxml to either match
original name or dname. However, setting dxml domain name to dname
doesn't really work since destination host needs to know the original
domain name to be able to use it in migration cookies. This patch
requires domain name in dxml to match the original domain name. The
change should be safe and backward compatible since migration would fail
just a bit later in the process.
Osier Yang [Mon, 16 Jan 2012 14:08:07 +0000 (22:08 +0800)]
docs: Add missed RNG schema for interface
We support <interface> of type "mcast", "server", and "client",
but the RNG schema for them are missed. Attribute "address" is
optional for "server" type. And these 3 types support
<mac address='MAC'/>, too.
Eric Blake [Mon, 16 Jan 2012 17:23:32 +0000 (10:23 -0700)]
build: fix bootstrap on fresh clone
Commit 29db7a0 picked up a gnulib bug, where a change in
bootstrap meant that it would fail to run libtoolize on
projects, like libvirt, that used the older spelling
AM_PROG_LIBTOOL instead of LT_INIT for the sake of building
on RHEL 5. Now that gnulib is fixed, we should pick up that
fix.
* .gnulib: Update to latest, for bootstrap fix.
* bootstrap: Resync from gnulib.
Osier Yang [Mon, 16 Jan 2012 09:28:51 +0000 (17:28 +0800)]
docs: Expose alias tag in domain RNG schema
Though <alias> is ignored when defining a domain, it can cause
failure if one validates (e.g. virt-xml-validate) the XML dumped
from a running domain. This patch expose it in domain RNG schema
for all the devices which support it.
Michael Ellerman [Fri, 13 Jan 2012 03:39:26 +0000 (14:39 +1100)]
conf: Remove do-nothing validation functions
There are three address validation routines that do nothing:
virDomainDeviceDriveAddressIsValid()
virDomainDeviceUSBAddressIsValid()
virDomainDeviceVirtioSerialAddressIsValid()
Remove them, and replace their call sites with "1" which is what they
currently return. In some cases this means we can remove an entire
if block.
Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Michael Ellerman [Fri, 13 Jan 2012 03:39:25 +0000 (14:39 +1100)]
tests: Add qemuxml2argv tests for PPC64 pseries machine
Add four tests of the XML -> argv handling for the PPC64 pseries machine.
The first is just a basic test of a bare bones machine.
The three others test various aspects of the spapr-vio address handling.
It seems that currently we can't include network devices, doing so leads
to a segfault because the network driverState is not initialised. Working
around that leads us to the problem that the 'default' network doesn't
exist. So for now just leave network devices out.
Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Michael Ellerman [Fri, 13 Jan 2012 03:39:24 +0000 (14:39 +1100)]
tests: Teach qemuxml2argvtest about spapr-vio addresses
We can't call qemuCapsExtractVersionInfo() from test code, because it
expects to be able to call the emulator, and for testing we have fake
emulators that can't be executed. For that reason qemuxml2argvtest.c
doesn't call qemuDomainAssignPCIAddresses(), instead it open codes its
own version.
That means we can't call qemuDomainAssignAddresses() from the test code,
instead we need to manually call qemuDomainAssignSpaprVioAddresses().
Also add logic to cope with qemuDomainAssignSpaprVioAddresses() failing,
so that we can write a test that checks for a known failure in there.
Signed-off-by: Michael Ellerman <michael@ellerman.id.au>
Adam Litke [Fri, 13 Jan 2012 20:44:42 +0000 (14:44 -0600)]
events: Return the correct number of registered events
Commit d09f6ba5feb655925175dc80122ca2a1e14db2b9 introduced a regression in event
registration. virDomainEventCallbackListAddID() will only return a positive
integer if the type of event being registered is VIR_DOMAIN_EVENT_ID_LIFECYCLE.
For other event types, 0 is always returned on success. This has the
unfortunate side effect of not enabling remote event callbacks because
remoteDomainEventRegisterAny() uses the return value from the local call to
determine if an event callback needs to be registered on the remote end.
Make sure virDomainEventCallbackListAddID() returns the callback count for the
eventID being registered.
Osier Yang [Fri, 13 Jan 2012 10:03:12 +0000 (18:03 +0800)]
virsh: Two new fields for command domblklist
Disk "type" and "device" are generally interesting stuff the
user may want to known, too. To not break any scripts which
parsed the output field, a new option "--details" is introduced
to output the two introduced fields.
Osier Yang [Thu, 12 Jan 2012 09:31:14 +0000 (17:31 +0800)]
qemu: Support copy on read for disk
The new introduced optional attribute "copy_on_read</code> controls
whether to copy read backing file into the image file. The value can
be either "on" or "off". Copy-on-read avoids accessing the same backing
file sectors repeatedly and is useful when the backing file is over a
slow network. By default copy-on-read is off.
Added check for maximum number of vcpus exceeding topology limit
Earlier, when the number of vcpus was greater than the topology allowed,
libvirt didn't raise an error and continued, resulting in running qemu
with parameters making no sense. Even though qemu did not report any
error itself, the number of vcpus was set to maximum allowed by the
topology.
Eric Blake [Thu, 12 Jan 2012 00:44:49 +0000 (17:44 -0700)]
uuid: fix off-by-one
Detected by Coverity. Although unlikely, if we are ever started
with stdin closed, we could reach a situation where we open a
uuid file but then fail to close it, making that file the new
stdin for the rest of the process.
* src/util/uuid.c (getDMISystemUUID): Allow for stdin.
Re-write LXC controller end-of-file I/O handling yet again
Currently the LXC controller attempts to deal with EOF on a
tty by spawning a thread to do an edge triggered epoll_wait().
This avoids the normal event loop spinning on POLLHUP. There
is a subtle mistake though - even after seeing POLLHUP on a
master PTY, it is still perfectly possible & valid to write
data to the PTY. There is a buffer that can be filled with
data, even when no client is present.
The second mistake is that the epoll_wait() thread was not
looking for the EPOLLOUT condition, so when a new client
connects to the LXC console, it had to explicitly send a
character before any queued output would appear.
Finally, there was in fact no need to spawn a new thread to
deal with epoll_wait(). The epoll file descriptor itself
can be poll()'d on normally.
This patch attempts to deal with all these problems.
- The blocking epoll_wait() thread is replaced by a poll
on the epoll file descriptor which then does a non-blocking
epoll_wait() to handle events
- Even if POLLHUP is seen, we continue trying to write
any pending output until getting EAGAIN from write.
- Once write returns EAGAIN, we modify the epoll event
mask to also look for EPOLLOUT
* src/lxc/lxc_controller.c: Avoid stalled I/O upon
connected to an LXC console
Allow 10 chars for domain IDs & 30 chars for names in virsh list
Domain IDs are at least 16 bits for most hypervisors, theoretically
event 32-bits. 3 characters is clearly too small an alignment.
Increase alignment to 5 characters to allow 16-bit domain IDs to
display cleanly. Commonly seen with LXC where domain IDs are the
process IDs by default. Also increase the 'name' field from 20
to 30 characters to cope with longer guest names which are quite
common
Michal Privoznik [Tue, 10 Jan 2012 15:57:30 +0000 (16:57 +0100)]
stream: Check for stream EOF
If client stream does not have any data to sink and neither received
EOF, a dummy packet is sent to the daemon signalising client is ready to
sink some data. However, after we added event loop to client a race may
occur:
Thread 1 calls virNetClientStreamRecvPacket and since no data are cached
nor stream has EOF, it decides to send dummy packet to server which will
sent some data in turn. However, during this decision and actual message
exchange with server -
Thread 2 receives last stream data from server. Therefore an EOF is set
on stream and if there is a call waiting (which is not yet) it is woken
up. However, Thread 1 haven't sent anything so far, so there is no call
to be woken up. So this thread sent dummy packet to daemon, which
ignores that as no stream is associated with such packet and therefore
no reply will ever come.
Osier Yang [Wed, 11 Jan 2012 12:45:31 +0000 (20:45 +0800)]
virsh: New command print summary of all virtual interfaces
Just like command "domblklist", the command extracts "type",
"source", "target", "model", and "MAC" of all virtual interfaces
from domain XML (live or persistent).
Deepak C Shetty [Tue, 10 Jan 2012 12:53:31 +0000 (18:23 +0530)]
Do not generate security_model when fs driver is anything but 'path'
QEMU does not support security_model for anything but 'path' fs driver type.
Currently in libvirt, when security_model ( accessmode attribute) is not
specified it auto-generates it irrespective of the fs driver type, which
can result in a qemu error for drivers other than path. This patch ensures
that the qemu cmdline is correctly generated by taking into account the
fs driver type.
Signed-off-by: Deepak C Shetty <deepakcs@linux.vnet.ibm.com>
Shradha Shah [Wed, 14 Dec 2011 10:50:30 +0000 (10:50 +0000)]
Functionality to implicitly get interface pool from SR-IOV PF.
If a system has 64 or more VF's, it is quite tedious to mention each VF
in the interface pool.
The following modification will implicitly create an interface pool from
the SR-IOV PF.
Although the netcf interface driver can in theory be used by
the stateless drivers, in practice none of them want to use
it because they have different ways of dealing with interfaces.
Furthermore, if you have mingw32-netcf installed, then the
libvirt mingw32 build will fail with
../../src/interface/netcf_driver.c:644:5: error: unknown field 'close_used_without_including_unistd_h' specified in initializer
* configure.ac: disable netcf if built without libvirtd
Eric Blake [Wed, 11 Jan 2012 13:48:14 +0000 (06:48 -0700)]
build: fix build on mingw with netcf available
The autobuilder pointed out an odd failure on mingw:
../../src/interface/netcf_driver.c:644:5: error: unknown field 'close_used_without_including_unistd_h' specified in initializer
cc1: warnings being treated as errors
This is because the gnulib headers #define close to different strings,
according to which headers are included, in order to work around some
odd mingw problems with close(), and these defines happen to also
affect field members declared with a name of struct foo.close. As long
as all headers are included before both the definition and use of the
struct, the various #define doesn't matter, but the netcf file hit
an instance where things were included in a different order. Fix this
for all clients that use a struct member named 'close'.
* src/driver.h: Include <unistd.h> before using 'close'.
Eric Blake [Wed, 11 Jan 2012 13:32:52 +0000 (06:32 -0700)]
build: avoid spurious compiler warning
For some weird reason, i686-pc-mingw32-gcc version 4.6.1 at -O2 complained:
../../src/conf/nwfilter_params.c: In function 'virNWFilterVarCombIterCreate':
../../src/conf/nwfilter_params.c:346:23: error: 'minValue' may be used uninitialized in this function [-Werror=uninitialized]
../../src/conf/nwfilter_params.c:319:28: note: 'minValue' was declared here
../../src/conf/nwfilter_params.c:344:23: error: 'maxValue' may be used uninitialized in this function [-Werror=uninitialized]
../../src/conf/nwfilter_params.c:319:18: note: 'maxValue' was declared here
cc1: all warnings being treated as errors
even though all paths of the preceding switch statement either
assign the variables or return.
Stefan Berger [Wed, 11 Jan 2012 11:42:37 +0000 (06:42 -0500)]
Address side effects of accessing vars via index
Address side effect of accessing a variable via an index: Filters
accessing a variable where an element is accessed that is beyond the
size of the list (for example $TEST[10] and only 2 elements are available)
cannot instantiate that filter. Test for this and report proper error
to user.
processes the two lists 'A' and 'B' in parallel. This means that A and B
must have the same number of 'N' elements and that 'N' rules will be
instantiated (assuming all tuples from A and B are unique).
In this patch we now introduce the assignment of variables to different
iterators. Therefore a rule like
will now create every combination of elements in A with elements in B since
A has been assigned to an iterator with Id '1' and B has been assigned to an
iterator with Id '2', thus processing their value independently.
Stefan Berger [Wed, 11 Jan 2012 11:42:37 +0000 (06:42 -0500)]
Optimize the elements the iterator visits.
In this patch we introduce testing whether the iterator points to a
unique set of entries that have not been seen before at one of the previous
iterations. The point is to eliminate duplicates and with that unnecessary
filtering rules by preventing identical filtering rules from being
instantiated.
Example with two lists:
list1 = [1,2,1]
list2 = [1,3,1]
The 1st iteration would take the 1st items of each list -> 1,1
The 2nd iteration would take the 2nd items of each list -> 2,3
The 3rd iteration would take the 3rd items of each list -> 1,1 but
skip them since this same pair has already been encountered in the 1st
iteration
Implementation-wise this is solved by taking the n-th element of list1 and
comparing it against elements 1..n-1. If no equivalent is found, then there
is no possibility of this being a duplicate. In case an equivalent element
is found at position i, then the n-th element in the 2nd list is compared
against the i-th element in the 2nd list and if that is not the same, then
this is a unique pair, otherwise it is not unique and we may need to do
the same comparison on the 3rd list.
Change security driver APIs to use virDomainDefPtr instead of virDomainObjPtr
When sVirt is integrated with the LXC driver, it will be neccessary
to invoke the security driver APIs using only a virDomainDefPtr
since the lxc_container.c code has no virDomainObjPtr available.
Aside from two functions which want obj->pid, every bit of the
security driver code only touches obj->def. So we don't need to
pass a virDomainObjPtr into the security drivers, a virDomainDefPtr
is sufficient. Two functions also gain a 'pid_t pid' argument.
* src/qemu/qemu_driver.c, src/qemu/qemu_hotplug.c,
src/qemu/qemu_migration.c, src/qemu/qemu_process.c,
src/security/security_apparmor.c,
src/security/security_dac.c,
src/security/security_driver.h,
src/security/security_manager.c,
src/security/security_manager.h,
src/security/security_nop.c,
src/security/security_selinux.c,
src/security/security_stack.c: Change all security APIs to use a
virDomainDefPtr instead of virDomainObjPtr
Eric Blake [Mon, 9 Jan 2012 18:57:46 +0000 (11:57 -0700)]
snapshot: allow reuse of existing files in disk snapshot
When disk snapshots were first implemented, libvirt blindly refused
to allow an external snapshot destination that already exists, since
qemu will blindly overwrite the contents of that file during the
snapshot_blkdev monitor command, and we don't like a default of
data loss by default. But VDSM has a scenario where NFS permissions
are intentionally set so that the destination file can only be
created by the management machine, and not the machine where the
guest is running, so that libvirt will necessarily see the destination
file already existing; adding a flag will allow VDSM to force the file
reuse without libvirt complaining of possible data loss.
Eric Blake [Mon, 9 Jan 2012 20:56:42 +0000 (13:56 -0700)]
docs: standardize description of flags
We had loads of different styles in describing the @flags parameter
for various APIs, as well as several APIs that didn't list which
enums provided the bit values valid for the flags.
The end result is one of two formats:
@flags: bitwise-OR of vir...Flags
@flags: extra flags; not used yet, so callers should always pass 0
(it doesn't eliminate the failure to start, but causes libvirt to give
a better idea about the cause of the failure).
If a guest uses a kvm emulator (e.g. /usr/bin/qemu-kvm) and the guest
is started when kvm isn't available (either because virtualization is
unavailable / has been disabled in the BIOS, or the kvm modules
haven't been loaded for some reason), a semi-cryptic error message is
logged:
libvirtError: internal error Child process (LC_ALL=C
PATH=/sbin:/usr/sbin:/bin:/usr/bin /usr/bin/qemu-kvm -device ? -device
pci-assign,? -device virtio-blk-pci,? -device virtio-net-pci,?) status
unexpected: exit status 1
This patch notices at process start that a guest needs kvm, and checks
for the presence of /dev/kvm (a reasonable indicator that kvm is
available) before trying to execute the qemu binary. If kvm isn't
available, a more useful (too verbose??) error is logged.
Jim Fehlig [Tue, 3 Jan 2012 18:35:06 +0000 (11:35 -0700)]
PolicyKit: Check auth before asking client to obtain it
I previously mentioned [1] a PolicyKit issue where libvirt would
proceed with authentication even though polkit-auth failed:
testusr xen134:~> virsh list --all
Attempting to obtain authorization for org.libvirt.unix.manage.
polkit-grant-helper: given auth type (8 -> yes) is bogus
Failed to obtain authorization for org.libvirt.unix.manage.
Id Name State
----------------------------------
0 Domain-0 running
- sles11sp1-pv shut off
AFAICT, libvirt attempts to obtain a privilege it already has,
causing polkit-auth to fail with above message. Instead of calling
obtain and then checking auth, IMO the workflow should be for the
server to check auth first, and if that fails ask the client to
obtain it and check again. This workflow also allows for checking
only successful exit of polkit-auth in virConnectAuthGainPolkit().
Laine Stump [Thu, 5 Jan 2012 03:48:38 +0000 (22:48 -0500)]
qemu: add new disk device='lun' for bus='virtio' & type='block'
In the past, generic SCSI commands issued from a guest to a virtio
disk were always passed through to the underlying disk by qemu, and
the kernel would also pass them on.
As a result of CVE-2011-4127 (see:
http://seclists.org/oss-sec/2011/q4/536), qemu now honors its
scsi=on|off device option for virtio-blk-pci (which enables/disables
passthrough of generic SCSI commands), and the kernel will only allow
the commands for physical devices (not for partitions or logical
volumes). The default behavior of qemu is still to allow sending
generic SCSI commands to physical disks that are presented to a guest
as virtio-blk-pci devices, but libvirt prefers to disable those
commands in the standard virtio block devices, enabling it only when
specifically requested (hopefully indicating that the requester
understands what they're asking for). For this purpose, a new libvirt
disk device type (device='lun') has been created.
device='lun' is identical to the default device='disk', except that:
1) It is only allowed if bus='virtio', type='block', and the qemu
version is "new enough" to support it ("new enough" == qemu 0.11 or
better), otherwise the domain will fail to start and a
CONFIG_UNSUPPORTED error will be logged).
2) The option "scsi=on" will be added to the -device arg to allow
SG_IO commands (if device !='lun', "scsi=off" will be added to the
-device arg so that SG_IO commands are specifically forbidden).
Guests which continue to use disk device='disk' (the default) will no
longer be able to use SG_IO commands on the disk; those that have
their disk device changed to device='lun' will still be able to use SG_IO
commands.
*docs/formatdomain.html.in - document the new device attribute value.
*docs/schemas/domaincommon.rng - allow it in the RNG
*tests/* - update the args of several existing tests to add scsi=off, and
add one new test that will test scsi=on.
*src/conf/domain_conf.c - update domain XML parser and formatter
*src/qemu/qemu_(command|driver|hotplug).c - treat
VIR_DOMAIN_DISK_DEVICE_LUN *almost* identically to
VIR_DOMAIN_DISK_DEVICE_DISK, except as indicated above.
Note that no support for this new device value was added to any
hypervisor drivers other than qemu, because it's unclear what it might
mean (if anything) to those drivers.
Laine Stump [Tue, 29 Nov 2011 18:37:27 +0000 (13:37 -0500)]
qemu: add capabilities flags related to SG_IO
This patch adds two capabilities flags to deal with various aspects
of supporting SG_IO commands on virtio-blk-pci devices:
QEMU_CAPS_VIRTIO_BLK_SCSI
set if -device virtio-blk-pci accepts the scsi="on|off" option
When present, this is on by default, but can be set to off to disable
SG_IO functions.
QEMU_CAPS_VIRTIO_BLK_SG_IO
set if SG_IO commands are supported in the virtio-blk-pci driver
(present since qemu 0.11 according to a qemu developer, if I
understood correctly)
Laine Stump [Fri, 6 Jan 2012 17:59:47 +0000 (12:59 -0500)]
config: report error when script given for inappropriate interface type
This fixes https://bugzilla.redhat.com/show_bug.cgi?id=638633
Although scripts are not used by interfaces of type other than
"ethernet" in qemu, due to the fact that the parser stores the script
name in a union that is only valid when type is ethernet or bridge,
there is no way for anyone except the parser itself to catch the
problem of specifying an interface script for an inappropriate
interface type (by the time the parsed data gets back to the code that
called the parser, all evidence that a script was specified is
forgotten).
Since the parser itself should be agnostic to which type of interface
allows scripts (an example of why: a script specified for an interface
of type bridge is valid for xen domains, but not for qemu domains),
the solution here is to move the script out of the union(s) in the
DomainNetDef, always populate it when specified (regardless of
interface type), and let the driver decide whether or not it is
appropriate.
Currently the qemu, xen, libxml, and uml drivers recognize the script
parameter and do something with it (the uml driver only to report that
it isn't supported). Those drivers have been updated to log a
CONFIG_UNSUPPORTED error when a script is specified for an interface
type that's inappropriate for that particular hypervisor.
(NB: There was earlier discussion of solving this problem by adding a
VALIDATE flag to all libvirt APIs that accept XML, which would cause
the XML to be validated against the RNG files. One statement during
that discussion was that the RNG shouldn't contain hypervisor-specific
things, though, and a proper solution to this problem would require
that (again, because a script for an interface of type "bridge" is
accepted by xen, but not by qemu).
Eric Blake [Fri, 6 Jan 2012 23:07:34 +0000 (16:07 -0700)]
tests: work around pdwtags 1.9 failure
On rawhide, gcc is new enough to output new DWARF information that
pdwtags has not yet learned, but the resulting 'make check' output
was rather confusing:
$ make -C src check
...
GEN virkeepaliveprotocol-structs
die__process_function: DW_TAG_INVALID (0x4109) @ <0x58c> not handled!
WARNING: your pdwtags program is too old
WARNING: skipping the virkeepaliveprotocol-structs test
WARNING: install dwarves-1.3 or newer
...
$ pdwtags --version
v1.9
I've filed the pdwtags deficiency as
https://bugzilla.redhat.com/show_bug.cgi?id=772358
* src/Makefile.am (PDWTAGS): Don't leave -t file behind on version
mismatch. Soften warning message, since 1.9 is newer than 1.3.
Don't leak stderr from broken version.
Eric Blake [Fri, 6 Jan 2012 21:07:23 +0000 (14:07 -0700)]
tests: avoid test failure on rawhide gnutls
I hit a VERY weird testsuite failure on rawhide, which included
_binary_ output to stderr, followed by a hang waiting for me
to type something! (Here, using ^@ for NUL):
$ ./commandtest
TEST: commandtest
WARNING: gnome-keyring:: couldn't send data: Bad file descriptor
.WARNING: gnome-keyring:: couldn't send data: Bad file descriptor
.WARNING: gnome-keyring:: couldn't send data: Bad file descriptor
WARNING: gnome-keyring:: couldn't send data: Bad file descriptor
.8^@^@^@8^@^@^@^A^@^@^@^Bay^A^@^@^@)PRIVATE-GNOME-KEYRING-PKCS11-PROTOCOL-V-1
I finally traced it to the fact that gnome-keyring, called via
gnutls_global_init which is turn called by virNetTLSInit, opens
an internal fd that it expects to communicate to via a
pthread_atfork handler (never mind that it violates POSIX by
using non-async-signal-safe functions in that handler:
https://bugzilla.redhat.com/show_bug.cgi?id=772320).
Our problem stems from the fact that we pulled the rug out from
under the library's expectations by closing an fd that it had
just opened. While we aren't responsible for fixing the bugs
in that pthread_atfork handler, we can at least avoid the bugs
by not closing the fd in the first place.
* tests/commandtest.c (mymain): Avoid closing fds that were opened
by virInitialize.
Alex Jia [Fri, 6 Jan 2012 06:36:34 +0000 (14:36 +0800)]
qemu: Avoid memory leaks on qemuParseRBDString
Detected by valgrind. Leak introduced in commit 5745dc1.
* src/qemu/qemu_command.c: fix memory leak on failure and successful path.
* How to reproduce?
% valgrind -v --leak-check=full ./qemuargv2xmltest
* Actual result:
==2196== 80 bytes in 1 blocks are definitely lost in loss record 3 of 4
==2196== at 0x4A05FDE: malloc (vg_replace_malloc.c:236)
==2196== by 0x39CF07F6E1: strdup (in /lib64/libc-2.12.so)
==2196== by 0x419823: qemuParseRBDString (qemu_command.c:1657)
==2196== by 0x4221ED: qemuParseCommandLine (qemu_command.c:5934)
==2196== by 0x422AFB: qemuParseCommandLineString (qemu_command.c:7561)
==2196== by 0x416864: testCompareXMLToArgvHelper (qemuargv2xmltest.c:48)
==2196== by 0x417DB1: virtTestRun (testutils.c:141)
==2196== by 0x415CAF: mymain (qemuargv2xmltest.c:175)
==2196== by 0x4174A7: virtTestMain (testutils.c:696)
==2196== by 0x39CF01ECDC: (below main) (in /lib64/libc-2.12.so)
==2196==
==2196== LEAK SUMMARY:
==2196== definitely lost: 80 bytes in 1 blocks
Eric Blake [Thu, 5 Jan 2012 21:21:11 +0000 (14:21 -0700)]
build: drop check for ANSI compiler
Using automake.git (will become 1.12 someday), I got this error:
configure.ac:90: error: automatic de-ANSI-fication support has been removed
/usr/local/share/aclocal-1.11a/protos.m4:13: AM_C_PROTOTYPES is expanded from...
configure.ac:90: the top level
autom4te: /usr/bin/m4 failed with exit status: 1
In short, pre-C89 compilers are no longer a viable portability
target. Besides, our code base already requires C99, so worrying
about pre-C89 seems pointless.
* configure.ac (AM_C_PROTOTYPES): Drop, since newer automake no
longer provides it.
Hu Tao [Wed, 4 Jan 2012 09:41:43 +0000 (17:41 +0800)]
qemu: fix a bug in numatune
When setting numa nodeset for a domain which has no nodeset set
before, libvirtd crashes by dereferencing the pointer to the old
nodemask which is null in that case.
Eric Blake [Wed, 4 Jan 2012 23:01:24 +0000 (16:01 -0700)]
seclabel: fix regression in libvirtd restart
Commit b434329 has a logic bug: seclabel overrides don't set
def->type, but the default value is 0 (aka static). Restarting
libvirtd would thus reject the XML for any domain with an
override of <seclabel relabel='no'/> (which happens quite
easily if a disk image lives on NFS), with a message:
2012-01-04 22:29:40.949+0000: 6769: error : virSecurityLabelDefParseXMLHelper:2593 : XML error: security label is missing
Fix the logic to never read from an override's def->type, and
to allow a missing <label> subelement when relabel is no. There's
a lot of stupid double-negatives in the code (!norelabel) because
of the way that we want the zero-initialized defaults to behave.
* src/conf/domain_conf.c (virSecurityLabelDefParseXMLHelper): Use
type field from correct location.
command: Discard FD_SETSIZE limit for opened files
Currently, virCommand implementation uses FD_ macros from
sys/select.h. However, those cannot handle more opened files
than FD_SETSIZE. Therefore switch to generalized implementation
based on array of integers.
Jim Fehlig [Wed, 4 Jan 2012 15:47:37 +0000 (08:47 -0700)]
Support Xen domctl v8
xen-unstable c/s 23874:651aed73b39c added another member to
xen_domctl_getdomaininfo struct and bumped domctl version to 8.
Add a corresponding domctl v8 struct in xen hypervisor sub-driver
and detect domctl v8 during initialization.
Jim Fehlig [Tue, 3 Jan 2012 22:39:59 +0000 (15:39 -0700)]
Fix xenstore serial console path for HVM guests
The console path in xenstore is /local/domain/<id>/console/tty
for PV guests (PV console) and /local/domain/<id>/serial/0/tty
(serial console) for HVM guests. Similar to Xen's in-tree console
client, read the correct path for PV vs HVM.
projects / libvirt.git / log