Mark McLoughlin [Wed, 22 Jul 2009 19:17:14 +0000 (20:17 +0100)]
Add SCM_RIGHTS support to QEMU monitor code
Add qemudMonitorCommandWithFd() which allows a file descriptor to be
sent to qemu over a unix monitor socket using SCM_RIGHTS. See the
unix(7) and cmsg(3) man pages.
* src/qemu_conf.c: add a scm_fd param to qemudMonitorCommandExtra(),
add qemudMonitorCommandWithFd(), implement SCM_RIGHTS support in
qemudMonitorSendUnix()
Mark McLoughlin [Wed, 22 Jul 2009 19:17:14 +0000 (20:17 +0100)]
Make qemuBuildHostNetStr() take tapfd as a string
With hotplug, we're going to want to pass a tapfd name rather than an
actual file descriptor, so prepare the way by passing a string tapfd to
qemuBuildHostNetStr().
* src/qemu_conf.h: qemuBuildHostNetStr() takes a string tapfd now
* src/qemu_conf.c: pass qemuBuildHostNetStr() a string rather than an
actual file descriptor
* src/qemu_driver.c: update qemudDomainAttachNetDevice() for change
Mark McLoughlin [Fri, 24 Jul 2009 11:34:20 +0000 (12:34 +0100)]
Only probe qemu for machine types when binary changes
By probing for qemu machine types, we increased the time of a
GetCapabilities call from 100us to a whopping 60ms.
This patch takes the approach of only probing for machine types
when the mtime of the emulator binary changed since the last time
the capabilities were generated.
* src/capabilities.h: cache the emulator binary mtime
* src/qemu_conf.c: add qemudGetOldMachines() to copy the machine
types from the old caps struct if the mtime for the binary hasn't
changed
* src/qemu_conf.h, src/qemu_driver.c: pass the old caps pointer to
qemudCapsInit()
Mark McLoughlin [Thu, 23 Jul 2009 17:31:34 +0000 (18:31 +0100)]
Canonicalize qemu machine types
In qemu-0.11 there is a 'pc-0.10' machine type which allows you to run
guests with a machine which is compatible with the pc machine in
qemu-0.10 - e.g. using the original PCI class for virtio-blk and
virtio-console and disabling MSI support in virtio-net. The idea here
is that we don't want to suprise guests by changing the hardware when
qemu is updated.
I've just posted some patches for qemu-0.11 which allows libvirt to
canonicalize the 'pc' machine alias to the latest machine version.
This patches makes us use that so that when a guest is configured to
use the 'pc' machine type, we resolve that to 'pc-0.11' machine and
save that in the guest XML.
Mark McLoughlin [Thu, 23 Jul 2009 17:31:34 +0000 (18:31 +0100)]
Add virCapsGuestMachine structure
A subsequent commit will add a "canonical" field to this structure,
this patch basically just prepares the way for that.
The new type is added, along with virCapabilitiesAlloc/FreeMachines()
helpers and a whole bunch of code to make the transition.
One quirk is that virCapabilitiesAddGuestDomain() and
virCapabilitiesAddGuest() take ownership of the machine list rather
than duping it. This makes sense to avoid needless copying.
* src/capabilities.h: add the virCapsGuestMachine struct and use it
in virCapsGuestDomainInfo, add prototypes for new functions and
update the AddGuest() prototypes
* src/capabilities.c: add code for allocating and freeing the new
type, change the machines parameter to AddGuest() etc.
* src/libvirt_private.syms: export the new helpers
* src/qemu_conf.c: update all the machine type code to use the new
struct
Mark McLoughlin [Thu, 23 Jul 2009 17:31:34 +0000 (18:31 +0100)]
Probe for QEMU machine types
Currently we hardcode the QEMU machine types. We should really just
parse the output of 'qemu -M ?' so the lists don't get out of sync.
xenner doesn't support '-M ?', so we still need to hardcode that.
The horrible (const char *const *) is removed in a subsequent patch.
* src/qemu_conf.c: kill the arch_info*machines tables, retain the
hardcoded xenner machine type, add qemudProbeMachineTypes() to
run and parse 'qemu -M ?' and use it in qemudCapsInitGuest()
* src/esx/esx_driver.c src/esx/esx_vi.c src/esx/esx_vi.h
src/esx/esx_vmx.c src/esx/esx_vmx.h: extend the VI API version checks
to accept version 4.0 and takes care of the virtualHW.version change
from 4 to 7.
* src/esx/esx_driver.c src/esx/esx_util.c src/esx/esx_util.h
src/esx/esx_vi.c src/esx/esx_vi.h: adds a no_verify query parameter to
stop libcurl from verifying theserver certificate for the https
transport.
Don't allow NULL paths for BlockStats and InterfaceStats
Do the check in libvirt.c, to save drivers from the burden. This changes
behavior slightly in the qemu driver: we no longer explictly error if
passed an empty string. An error will still be thrown when the device
lookup fails.
Add support for VBox 3 and event callbacks on vbox
* src/vbox/vbox_driver.c: remove some old 2.5 switches and plug the
3.0 driver
* src/vbox/vbox_V3_0.c src/vbox/vbox_CAPI_v3_0.h: the driver for
VirtualBox 3.0
* src/vbox/vbox_tmpl.c: handle the new driver and add event support
* src/Makefile.am: plug in the new module
Features supported:
- Connects to HMC/VIOS or IVM systems.
- Life cycle commands (resume and shutdown).
- dumpxml
- 'list' and 'list --all'
What is being implemented:
- better and centralized control for UUID
- definexml
- CPU management commands
* src/domain_conf.c src/domain_conf.h: first version of the driver
* configure.in src/Makefile.am include/libvirt/virterror.h
src/domain_conf.[ch] src/libvirt.c src/virterror.c: glue the driver
in the general framework
* src/esx/esx_*.[ch]: the driver, uses a remote minimal SOAP client
to talk to the VI services on ESX nodes.
* configure.in include/libvirt/virterror.h src/Makefile.am src/driver.h
src/libvirt.c src/virterror.c: glue in the new driver
* qemud/libvirtd_qemu.aug, qemud/test_libvirtd_qemu.aug,
src/qemu.conf: Add 'cgroups_controllers' and 'cgroups_device_acl'
parameters
* src/qemu_conf.h, src/qemu_conf.c: Load & parse configuration params
for cgroups
* src/qemu_driver.c: Only use cgroups controllers that are activated,
and use configured device whitelist instead of default, if set.
Use cgroups for block device whitelisting in QEMU guests
* src/qemu_driver.c: Set a restrictive block device whitelist for
all QEMU guests. Update whitelist when hotplugging disks.
* src/cgroup.h, src/cgroup.c: Add some more convenience methods
for dealing with block device whitelists.
* src/qemu_driver.c: Add driver methods qemuGetSchedulerType,
qemuGetSchedulerParameters, qemuSetSchedulerParameters
* src/lxc_driver.c: Fix to use unsigned long long consistently
for schedular parameters
* src/cgroup.h, src/cgroup.c: Fix cpu_shares to take unsigned
long long
* src/util.c, src/util.h, src/libvirt_private.syms: Add a
virStrToDouble helper
* src/virsh.c: Fix handling of --set arg to schedinfo command
to honour the designated data type of each schedular tunable
as declared by the driver
Refactor cgroups to allow a group per driver to be managed directly
Allow the driver level cgroup to be managed explicitly by the
hypervisor drivers, in order to detect whether to enable or
disable cgroup support for domains. Provides better error
reporting of failures. Also allow for creation of cgroups for
unprivileged drivers if controller is accessible by the user.
* src/cgroup.c, src/cgroup.h: Add an API to obtain a driver cgroup
* src/lxc_conf.h, src/lxc_controller.c, src/lxc_driver.c:
Obtain a driver cgroup at startup and use that instead of
re-creating everytime.
* src/util.c, src/util.h, src/libvirt_private.syms: Add a
virGetUserName() helper
* src/cgroup.c: Detect the mount location of every controller at
time a virCgroupPtr is created. Detect current process' placement
within group to avoid assuming it is in the root. Pass controller
ID into SetValueStr/GetValueStr to enable much duplicated code to
be eliminated
Laine Stump [Thu, 23 Jul 2009 16:27:47 +0000 (18:27 +0200)]
Add bare format string to printf-derivatives troubles
* src/datatypes.c src/domain_conf.c src/interface_conf.c
src/lxc_driver.c src/qemu_driver.c src/storage_backend.c src/virsh.c:
add bare %s format string to printf-derivatives called with no format
string
GCC >= 4.4 assumes the 'printf' attribute refers to the native
runtime libraries format specifiers. Thanks to gnulib, libvirt
has GNU format specifiers everywhere. This means we need to
use 'gnu_printf' with GCC >= 4.4 to get correct compiler
checking of printf format specifiers.
* HACKING: Document new rules for ATTRIBUTE_FMT_PRINTF
* autobuild.sh, mingw32-libvirt.spec.in: Disable OpenNebula
driver on mingw32 builds
* qemud/dispatch.h, qemud/qemu.h, src/buf.h src/internal.h,
src/logging.h, src/security.h, src/sexpr.h, src/util.h,
src/virterror_internal.h, src/xend_internal.c: Change
over to ATTRIBUTE_FMT_PRINTF.
* src/virsh.c: Disable 'cd' and 'pwd' commands on Win32
since they don't compile
* src/threads-win32.c: Add missing return value check
Jim Paris [Wed, 22 Jul 2009 18:51:28 +0000 (20:51 +0200)]
Always add -no-kvm and -no-kqemu, for qemu domains
If the qemu binary supports "-no-kvm" and/or "-no-kqemu", they should
always be added for plain "qemu" domains. Previously, we omitted them
whenever the host and guest architectures implied that they would be
disabled automatically, but that logic was flawed in some cases
(such as i686 and x86_64).
* src/qemu_conf.c: fix the conditions for adding "-no-kvm" and/or "-no-kqemu"
* include/libvirt/libvirt.h include/libvirt/libvirt.h.in: adds the new
flag VIR_MEMORY_PHYSICAL for virDomainMemoryPeek
* src/libvirt.c: update the front-end checking
* src/qemu_driver.c: extend the QEmu driver
Laine Stump [Wed, 22 Jul 2009 14:15:52 +0000 (16:15 +0200)]
Release conn lock before reporting interface errors
* src/datatypes.c: fix a lock problem on error handling, as the
error report takes the lock, it must be released before, fixes
the problem but just for Interface objects
Mark McLoughlin [Fri, 17 Jul 2009 21:08:34 +0000 (22:08 +0100)]
Add support for network device detach
qemu network devices are hot-unplugged in two stages - first the PCI NIC
is removed using 'pci_del <pci_addr>' and then the backend is removed
using 'host_net_remove <vlan> <name>'.
In order to perform these operations we need to have retained the
PCI address, backend name and vlan number.
Mark McLoughlin [Fri, 17 Jul 2009 21:08:34 +0000 (22:08 +0100)]
Remove the network backend if NIC hotplug fails
If we fail to pci_add a NIC, we should remove the network backend and
leave things the way we found them. To do that, we pre-allocate a
host_net_remove monitor command and issue that if the pci_add fails.
If the remove fails, we just log a warning.
We can only do this if we have a name for the network backend and
we know the vlan number its associated with.
* src/qemu_driver.c: host_net_remove the network backend if the
pci_add fails
Mark McLoughlin [Fri, 17 Jul 2009 21:08:34 +0000 (22:08 +0100)]
Make qemuCmdFlags available in qemudDomainAttachDevice()
qemudDomainChangeEjectableMedia() currently extracts the qemu command
line flags, but other device attaching code might need it, so move
the qemudExtractVersionInfo() call up a frame.
* src/qemu_driver.c: move the qemudExtractVersionInfo() call from
qemudDomainChangeEjectableMedia() to qemudDomainAttachDevice()
Mark McLoughlin [Fri, 17 Jul 2009 21:08:34 +0000 (22:08 +0100)]
Store the interface vlan number in the domain state
Currently, an interface's vlan number corresponds to its index in
the table of network interfaces. That is no longer true when we
allow devices to be removed.
To fix this, we store the vlan number in the domain's state XML
so that it survives libvirtd restarts.
* src/domain_conf.h: add vlan number to virDomainNetDef
* src/domain_conf.c: store it in XML as <state vlan='N'/>, defaulting
to -1 if this is state saved by a previous version of libvirt
* src/qemu_conf.c: assign vlan numbers before starting qemu
Mark McLoughlin [Fri, 17 Jul 2009 21:08:34 +0000 (22:08 +0100)]
Add NIC and hostnet names to domain state XML
The qemu driver needs to assign and keep track of identifiers for
network devices so that it can remove them. We need to keep this state
across libvirtd restarts, but it's not configuration that needs to
be kept across guest restarts.
* src/domain_conf.c: parse and format <state nic="foo" hostnet="bar"/>
Mark McLoughlin [Fri, 17 Jul 2009 21:08:33 +0000 (22:08 +0100)]
Add checks for some NIC hotplug related features added in qemu-0.10.0
Add QEMUD_CMD_FLAG_NET_NAME to indicate that '-net ...,name=foo' is
supported and QEMUD_CMD_FLAG_HOST_NET_ADD to indicate that the
'host_net_add' monitor command is available.
Set both these flags if the qemu version is greater than 0.10.0.
Checking via the '-help' output would not work for the monitor command
and even for the command line arg, it would be quite fragile.
* src/qemu_conf.h: add new flags as aliases of QEMUD_CMD_FLAG_0_10
* src/qemu_conf.c: set QEMUD_CMD_FLAG_0_10 for versions >= 0.10.0
* tests/qemuhelptest.c: set QEMUD_CMD_FLAG_0_10 for the appropriate
qemu versions
Mark McLoughlin [Fri, 17 Jul 2009 21:08:33 +0000 (22:08 +0100)]
Factor qemuBuildHostNetStr() out from qemuBuildCommandLine()
Re-factor this code so that it can be used for NIC hotplug
too. The awkward prefix and type_sep arguments are needed to
allow us to do "host_net_add tap vlan=..."
* src/qemu_conf.c: factor the net backend string formatting
code into its own function
Mark McLoughlin [Fri, 17 Jul 2009 21:08:33 +0000 (22:08 +0100)]
Retain disk PCI address across libvirtd restarts
When we hot-plug a disk device into a qemu guest, we need to retain its
PCI address so that it can be removed again later. Currently, we do
retain the slot number, but not across libvirtd restarts.
Add <state devaddr="xxxx:xx:xx"/> to the disk device XML config when the
VIR_DOMAIN_XML_INTERNAL_STATUS flag is used. We still don't parse the
domain and bus number, but the format allows us to do that in future.
* src/domain_conf.h: replace slotnum with pci_addr struct, add helper
for testing whether the address is valid
* src/domain_conf.c: handle formatting and parsing the address
* src/qemu_driver.c: store the parsed slot number as a full PCI address,
and use this address with the pci_del monitor command
* src/vbox/vbox_tmpl.c: we're debug printing slotnum here even though
it can never be set, just delete it
Mark McLoughlin [Fri, 17 Jul 2009 21:08:33 +0000 (22:08 +0100)]
Add internal XML parsing/formatting flag
We need to store things like device names and PCI slot numbers in the
qemu domain state file so that we don't lose that information on
libvirtd restart. Add a flag to indicate that this information should
be parsed or formatted.
Make bit 16 and above of the flags bitmask for internal use only and
consume the first bit for this new status flag.
Daniel Veillard [Tue, 21 Jul 2009 14:15:39 +0000 (16:15 +0200)]
Activate the interface drivers, and cleanups
* src/libvirt.c: activate the interface drivers
* po/POTFILES.in: add the netcf driver as a source of localization strings
* src/interface_driver.c: NETCF_ENOMEM -> VIR_ERR_NO_MEMORY mapping was
breaking syntax checking
Laine Stump [Tue, 21 Jul 2009 14:02:16 +0000 (16:02 +0200)]
Netcf based interface driver implementation
* src/interface_driver.c src/interface_driver.h: the new driver
* src/Makefile.am qemud/Makefile.am qemud/qemud.c: hook the new driver
in the build system and get ti activated by the daemon
* src/libvirt_private.syms: export needed symbols internally
Laine Stump [Tue, 21 Jul 2009 13:37:20 +0000 (15:37 +0200)]
virGetinterface matching of MAC and interface name
MAC address of a particular interface may change over time, and the
reduced virInterface object (which contains just name and mac) needs
to reflect these changes. Since we can't modify the mac address of an
existing virInterface (some other thread may currently be using it) we
just create a new virInterface, and let the old one die a dignified
death when its refct goes to 0.
* src/datatypes.c: fix the matching and lifetime of virInterface object
accordingly
Laine Stump [Tue, 21 Jul 2009 13:13:32 +0000 (15:13 +0200)]
Add an error code for conflicting mac addresses
* include/libvirt/virterror.h src/virterror.c: if a driver's
virInterfaceLookupByMACString() function finds more than one interface
with the desired MAC Address, this new error is raised.
Daniel Veillard [Tue, 21 Jul 2009 09:16:15 +0000 (11:16 +0200)]
rpm spec cleanup and split off client only package
* libvirt.spec.in: make a client rpm with shared libs, client binaries
and resources needed by those, and a small number of fixes and
cleanups in the spec file.
Jim Meyering [Thu, 16 Jul 2009 20:46:43 +0000 (22:46 +0200)]
build: do not emit a trailing blank line into VC'd file, NEWS
* docs/Makefile.am ($(top_builddir)/NEWS): Adjust rule to filter
out any trailing blank lines when generating this file.
* NEWS: Regenerate, so that it contains entries for 0.6.5.
Laine Stump [Thu, 16 Jul 2009 19:44:10 +0000 (21:44 +0200)]
Add virsh commands for network interface management
* src/virsh.c: add a number of interface related commands:
iface-list, iface-name, iface-mac, iface-dumpxml, iface-define,
iface-undefine, iface-edit, iface-start and iface-destroy
qemu: Try multiple times to open unix monitor socket
Unlike the pty monitor (which we know exists since we scrape its path from
stdout), we have no way of knowing that the unix monitor socket should exist/
be initialized. As a result, some of my KVM guests randomly fail to start on
F10 host.
Try to open the unix socket in a 3 second timeout loop. Ignore EACCES (path
does not exist if a first time run) and ECONNREFUSED (leftover socket from
a previous run hasn't been removed yet). Fixes things for me.
Have storage building functions be definitions of
virStorageBackendBuildVolFrom: we will need to do this in the future anyways
if we ever support the flags attribute.
Currently, if no format is specified for a new disk volume, we pass the
invalid value "none" as the FS type to 'parted mkpart'.
There doesn't seem to be a way to have parted not format the drive, so
just default to using 'ext2' in this case: this shouldn't cause any harm,
since we are creating a new partition in the first place.