PCIe DevCap register is actually 32 bits, not 16 bits. Since FLR is
bit 28, we clearly are failing to detect FLR support.
Known to fix device reset with some SR-IOV devices.
* src/pci.c: fix pciDetectFunctionLevelReset()
polkit was disabled by default for a reason - because we selectively
enable it on newer fedoras rather than disable it on older fedoras
Same fix needed for netcf
It's not needed at build time
Removed in Fedora by:
* Fri Jun 5 2009 Mark McLoughlin <markmc@redhat.com> - 0.6.4-2.fc12
- Remove the qemu BuildRequires
Disabled on < f12 for now until netcf is in Fedora updates
BuildRequires netcf if enabled, pass --without-netcf if its disabled
* libvirt.spec.in: enabled netcf by default
In order to hotplug a network/bridge backed NIC, we need to first create
the tap file descriptor, add the tap interface to the bridge and then
pass the file descriptor to the qemu process using the 'getfd' monitor
command.
Once the tapfd has been accepted, we create the network backend using
host_net_add, supplying the name assigned to the tapfd. If this fails,
we need to close the tapfd in qemu using the 'closefd' monitor command.
If the version of qemu does not support the getfd/closefd monitor
commands we detect "unknown command" in the getfd reply and fail the
attach operation.
* src/qemu_driver.c: add support for tapfd based hotplug in
qemudDomainAttachNetDevice()
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()
Switch from using write() to using sendmsg() on QEMU's monitor socket
so that we can add support for SCM_RIGHTS.
* src/qemu_driver.c: add sendmsg() based qemudMonitorSendUnix() and use
it when the monitor fd is a unix socket
Add a little helper function to write the monitor command followed by
carriage return in a single write.
This doesn't make any real difference, but allows us to more easily
switch to using sendmsg() when using the monitor over a unix socket.
* src/qemu_conf.c: split qemudMonitorSend() out
In subsequent patches we're going to have a file descriptor to close
too, so centralize the error handling cleanups to make things easier.
* src/qemu_conf.c: in qemudDomainAttachNetDevice() consolidate the
error handling cleanups together
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
* src/qemu_conf.h: export qemudNetworkIfaceConnect()
* src/qemu_conf.c: move vnet_hdr logic into qemudNetworkIfaceConnect()
since we need it for hotplug too
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()
e.g. <machine canonical='pc'>pc-0.11</machine>
* src/capabilities.c: output the canonical machine names in the
capabilities output, if available
* docs/schemas/capabilities.rng: add the new attribute
Not all possible emulators are actually in the capabilities, so if we
don't find the supplied emulator we should probe it directly for machine
types.
* src/qemu_driver.c: add qemudCanonicalizeMachineDirect() to directly
probe an emulator for the canonical machine type
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.
See also:
https://fedoraproject.org/wiki/Features/KVM_Stable_Guest_ABI
* src/qemu_conf.c: add qemudCanonicalizeMachine() to canonicalize
the machine type according to the machine aliases in capabilities
* src/qemu_driver.c: parse aliases in qemudParseMachineTypesStr()
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
* src/xen_internal.c: ditto
* tests/testutilsqemu.c: ditto
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()
There's no need for the hasbase/hasaltbase confusion, just store the
first binary path found in a variable.
* src/qemu_conf.c: kill hasbase/hasaltbase logic in qemudCapsInitGuest()
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.
* src/qemu_driver.c: add qemudDomainDetachNetDevice()
When we pci_add a NIC, we need to retain the PCI address assigned by
qemu for using during detach.
* src/qemu_driver.c: use qemudParsePciAddReply() to pull the PCI
address from the pci_add reply
* src/domain_conf.c: handle storing and parsing the PCI address in the
domain state XML file
The current code for parsing pci_add replies ignores the the domain and
bus numbers. Re-write the code to rectify that.
Also, since pci_add is used for NIC hotplug as well ask disk hotplug,
re-factor the code into a separate function.
* src/qemu_driver.c: add qemudParsePciAddReply() function which can
handle parsing domain and bus numbers
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
Implement basic NIC hotplug support using the 'host_net_add' and
'pci_add' qemu monitor commands.
For now, we don't support 'bridge' or 'network' types.
Also, if pci_add fails, we currently fail to remove the backend
which we added.
Finally, NIC hot-unplug support is missing.
* src/qemu_driver.c: add qemudDomainAttachNetDevice()
* src/qemu_conf.[ch]: export qemuBuildNicStr(), qemuBuildHostNetStr()
and qemuAssignNames()
* src/libvirt_private.syms: export virDomainNetTypeToString()
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()
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
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"/>
We need these so that we can remove the devices via the monitor.
* src/domain_conf.h: add nic_name and hostnet_name to virDomainNetDef
* src/domain_conf.c: free nic_name and hostnet_name
* src/qemu_conf.c: add qemuAssignNetNames(), use it if qemu has
support for the param and pass the names on the command line
* tests/qemuxml2argv*: add a test for this
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
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
Re-factor this code so that it can be used for NIC hotplug
too. The awkward arguments are needed to allow use to do
"pci_add auto nic macaddr=..."
* src/qemu_conf.c: factor the nic string formatting code into
its own function
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
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.
* include/libvirt/libvirt.h: add VIR_DOMAIN_XML_FLAGS_MASK
* src/libvirt.c: reject private flags in virDomainGetXMLDesc()
* src/domain_conf.h: add VIR_DOMAIN_XML_INTERNAL_STATUS
* src/domain_conf.c: pass the flag from virDomainObjParseXML() and
virDomainSaveStatus
* src/domain_conf.c: replace open coded chr type parsing with
virDomainChrTypeFromString(), retaining the existing semantics
where unknown types are silently mapped to the "null" type and
"pty" is used if none is specified
We keep support for the pty based monitor so that we can re-connect
to VMs started by older versions of libvirtd.
* src/domain_conf.c: handle formatting and parsing unix monitors
* src/qemu_driver.c: add qemudOpenMonitorUnix(), remove the monitor
pty path searching from qemudFindCharDevicePTYs(), switch
qemudStartVMDaemon() and qemuDomainXMLToNative() to using a unix
monitor
* tests/qemuxml2argvtest.c: switch to using a unix monitor
* tests/qemuxml2argvdata/qemuxml2argv-*.args: update test data
There are no functional changes in this patch apart from adding the
monitor type to the state XML.
The patch mostly consists of switching to use virDomainChrDef every
where to describe the monitor.
* src/domain_conf.h: replace monitorpath with monitor_chr
* src/domain_conf.c: handle parsing the monitor type and initializing
monitor chr
* src/qemu_conf.[ch]: make qemudBuildCommandLine take a
virDomainChrDefPtr and use that to build the -monitor parameter
* src/qemu_driver.c: split pty specific and common code from
qemudOpenMonitor, have qemudStartVMDaemon() initialize monitor_chr
* tests/qemuxml2argvtest.c: update for qemudBuildCommandLine() change
* src/qemu_driver.c: vm->monitorpath is already initialized in the case
of re-connect, so move the initialization for the normal startup case
out of the common code
Fixes startup of guest's with sourceless cdrom devices.
Patch from Cole Robinson originally posted here:
https://bugzilla.redhat.com/499569
but never sent upstream.
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
As pointed out by Tim Waugh here:
https://bugzilla.redhat.com/507555
We shouldn't bother trying to set the context of a file if it already
matches what we want.
(Fixed to use STREQ() and not use tabs, as pointed out by danpb)
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
The KVM version string can be one of the following:
- qemu-kvm-x.y.z in stable releases
- kvm-XX for kvm versions up to kvm-85
- qemu-kvm-devel-XX for kvm version kvm-86 and later
There are only a few of places where we need to detect
differences between KVM versions based on 0.9.1:
1) VNET_HDR introduced in kvm-74
2) -incoming tcp introduced in kvm-79
3) -incoming exec introduced in kvm-80
4) -incoming stdio in all earlier kvm versions
With qemu-kvm-0.10.x, we can now assume that (1) is available
if it's a KVM release, (2) and (3) is always available and
(4) is never available.
So, from now on we should only need to check the qemu version
number and the "is_kvm" flag for detecting feature availability.
We only need the KVM version number for older releases.
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
This patch is purely re-factoring without any functional changes
to make way for the next patch.
The main thing achieved by the refactoring is that we now have
easier access to the parenthesised string that KVM folks seem
to delight in changing.
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
This change:
Tue Mar 3 08:55:13 GMT 2009 Daniel P. Berrange <berrange@redhat.com>
Don't try to detach & reset PCI devices while running test
suite for XML-> ARGV conversion.
* src/qemu_driver.c: Add qemuPrepareHostDevices() helper to
detach and reset PCI devices.
* src/qemu_conf.c: Don't detach & reset PCI devices while
building the command line argv
accidentally did this:
- if (hostdev->managed) {
+ if (!hostdev->managed) {
Which results in managed='yes' not causing the device to be
detached when the guest is starting.
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
It's "service libvirtd start", not "service start libvirtd".
Reported-by: Robert P. J. Day <rpjday@crashcourse.ca>
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
Seems like a simple typo - it has been "format=" since the flag
was introduced, but we added it as "fmt=".
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
Fixes:
CHECK: avoid_if_before_free
src/lxc_container.c: if (oldroot) VIR_FREE(oldroot)
src/lxc_container.c: if (newroot) VIR_FREE(newroot)
Makefile.maint: found useless "if" before "free" above
and:
src/lxc_container.c:317:
Makefile.maint: found trailing blank(s)
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
Add a test to check node device XML parsing by first parsing
the XML, then re-formatting as XML and finally comparing the
resulting XML to the original XML.
Add support for parsing node device XML descriptions.
This will be used by PCI passthrough related functions to
obtain the PCI device address for a given node device.
bridge.c: In function 'brSetInetAddr':
bridge.c:665: error: dereferencing pointer '({anonymous})' does break strict-aliasing rules
bridge.c:665: note: initialized from here
bridge.c:666: error: dereferencing pointer '({anonymous})' does break strict-aliasing rules
bridge.c:666: note: initialized from here
Fix qemud.c warnings with gcc-4.4
qemud.c: In function ‘remoteListenTCP’:
qemud.c:675: error: dereferencing pointer ‘sa.72’ does break strict-aliasing rules
qemud.c:675: note: initialized from here
qemud.c:678: error: dereferencing pointer ‘sa.73’ does break strict-aliasing rules
qemud.c:678: note: initialized from here
IFF_VNET_HDR is a tun/tap flag that allows you to send and receive
large (i.e. GSO) packets and packets with partial checksums. Setting
the flag means that every packet is proceeded by the same header which
virtio uses to communicate GSO/csum metadata.
By enabling this flag on the tap fds we create, we greatly increase
the achievable throughput with virtio_net.
However, we need to be careful to only set the flag when a) QEMU has
support for this ABI and b) the value of the flag is queryable using
the TUNGETIFF ioctl.
It's nearly five months since kvm-74 - the first KVM release with this
feature - was released. Up until now, we've not added libvirt support
because there is no clean way to detect support for this in QEMU at
runtime. A brief attempt to add a "info capabilities" monitor command
to QEMU floundered. Perfect is the enemy of good enough. Probing the
KVM version will suffice for now.
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
from the post-install script:
Installing : libvirt
ln: creating symbolic link `/etc/libvirt/qemu/networks/autostart/default.xml': File exists
See https://bugzilla.redhat.com/462011
* src/qemu_conf.c: fix serial/telnet to use "server" option
rather than "listen"
* tests/qemuxml2argvdata/qemuxml2argv-serial-tcp-telnet.args:
fix the test case too
* src/internal.h: move xstrol() variants from here ...
* src/util.[ch]: ... to here and rename to virStrToLong()
* src/libvirt_sym.version: export __virStrToLong_i() for
virsh and qemud.
* src/nodeinfo.c, src/stats_linux.c, src/virsh.c,
src/xend_internal.c, qemud/qemud.c: replace xstrtol()
calls with virStrToLong()
* src/nodeinfo.h: don't include internal.h, which was only
needed for xstrtol(), but instead include libvirt.h which
is suffificient for the declarations in the header.
a proposed system for letting iptables know how to
reload our rules. The proposed system wasn't accepted
so, although there might be some other theoretical
use for this, let's just remove it.
iptables configuration using the lokkit --custom-rules command.
Basically, we write out our rules to /var/lib/libvirt/iptables
and run lokkit --custom-rules so that if e.g. iptables is
restarted or the user edits their firewall configuration, then
libvirt's rules get reloaded.
Mark McLoughlin