This patch moves some code in the qemuDomainAttachSCSIDisk
function. The check for the existence of a PCI address assigned
to the SCSI controller was moved in order to be executed only
when needed. The PCI address of a controller is not necessary
if QEMU_CAPS_DEVICE is supported.
This fixes issues with the hotplug of SCSI disks on pseries guests.
This patch (and the two patches that precede it) resolve:
https://bugzilla.redhat.com/show_bug.cgi?id=1005682
When libvirt was changed to delay the final cleanup of device removal
until the qemu process had signaled it with a DEVICE_DELETED event for
that device, the hostdev removal function
(qemuDomainRemoveHostDevice()) was written to properly handle the
removal of a hostdev that was actually an SRIOV virtual function
(defined with <interface type='hostdev'>). However, the function used
to search for a device matching the alias name provided in the
DEVICE_DELETED message (virDomainDefFindDevice()) would search through
the list of netdevs before hostdevs, so qemuDomainRemoveHostDevice()
was never called; instead the netdev function,
qemuDomainRemoveNetDevice() (which *doesn't* properly cleanup after
removal of <interface type='hostdev'>), was called.
(As a reminder - each <interface type='hostdev'> results in a
virDomainNetDef which contains a virDomainHostdevDef having a parent
type of VIR_DOMAIN_DEVICE_NET, and parent.data.net pointing back to
the virDomainNetDef; both Defs point to the same device info object
(and the info contains the device's "alias", which is used by qemu to
identify the device). The virDomainHostdevDef is added to the domain's
hostdevs list *and* the virDomainNetDef is added to the domain's nets
list, so searching either list for a particular alias will yield a
positive result.)
This function modifies the qemuDomainRemoveNetDevice() to short
circuit itself and call qemu DomainRemoveHostDevice() instead when the
actual device is a VIR_DOMAIN_NET_TYPE_HOSTDEV (similar logic to what
is done in the higher level qemuDomainDetachNetDevice())
Note that even if virDomainDefFindDevice() changes in the future so
that it finds the hostdev entry first, the current code will continue
to work properly.
This function was called in three places, and in each the call was
qualified by a slightly different conditional. In reality, this
function should only be called for a hostdev if all of the following
are true:
1) mode='subsystem'
2) type='pci'
3) there is a parent device definition which is an <interface>
(VIR_DOMAIN_DEVICE_NET)
We can simplify the callers and make them more consistent by checking
these conditions at the top ov qemuDomainHostdevNetConfigRestore and
returning 0 if one of them isn't satisfied.
The location of the call to qemuDomainHostdevNetConfigRestore() has
also been changed in the hot-plug case - it is moved into the caller
of its previous location (i.e. from qemuDomainRemovePCIHostDevice() to
qemuDomainRemoveHostDevice()). This was done to be more consistent
about which functions pay attention to whether or not this is one of
the special <interface> hostdevs or just a normal hostdev -
qemuDomainRemoveHostDevice() already contained a call to
networkReleaseActualDevice() and virDomainNetDefFree(), so it makes
sense for it to also handle the resetting of the device's MAC address
and vlan tag (which is what's done by
qemuDomainHostdevNetConfigRestore()).
Prefer using VFIO (if available) to the legacy KVM device passthrough.
With this patch a PCI passthrough device without the driver configured
will be started with VFIO if it's available on the host. If not legacy
KVM passthrough is checked and error is reported if it's not available.
The qemuDomainChangeNet() is called when 'virsh update-device' is
invoked on a NIC. Currently, we fail to update the QoS even though
we have routines for that.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
The return value of virDomainControllerFind >=0 means that
the specific controller was found.
But some functions invoke it and treat 0 as not found.
This patch fix these incorrect invocation.
Signed-off-by: Chen Hanxiao <chenhanxiao@cn.fujitsu.com>
When using a <interface type="network"> that points to a network with
hostdev forwarding mode a hostdev alias is created for the network. This
allias is inserted into the hostdev list, but is backed with a part of
the network object that it is connected to.
When a VM is being stopped qemuProcessStop() calls
networkReleaseActualDevice() which eventually frees the memory for the
hostdev object. Afterwards when the domain definition is being freed by
virDomainDefFree() an invalid pointer is accessed by
virDomainHostdevDefFree() and may cause a crash of the daemon.
This patch removes the entry in the hostdev list before freeing the
depending memory to avoid this issue.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1000973
If there's no hard_limit set and domain uses VFIO we still must lock
the guest memory (prerequisite from qemu). Hence, we should compute
the amount to be locked from max_balloon.
If user requested multiqueue networking, beside multiple /dev/tap and
/dev/vhost-net openings, we forgot to pass mq=on onto the -device
virtio-net-pci command line. This is advised at:
http://www.linux-kvm.org/page/Multiqueue#Enable_MQ_feature
This function is to guess the correct limit for maximal memory
usage by qemu for given domain. This can never be guessed
correctly, not to mention all the pains and sleepless nights this
code has caused. Once somebody discovers algorithm to solve the
Halting Problem, we can compute the limit algorithmically. But
till then, this code should never see the light of the release
again.
Hotplugging a single SCSI device works, but adding additional ones
result in an error from QEMU:
[root@gpok197 ~]# virsh attach-device guest01 blah.xml
Device attached successfully
[root@gpok197 ~]# virsh attach-device guest01 blah2.xml
error: Failed to attach device from blah2.xml
error: internal error unable to execute QEMU command 'device_add': Duplicate ID 'hostdev0' for device
The hostdev ID that is created is always set to zero, regardless
of the contents of the XML. Changing the index in the hotplug case
to a negative one so the next available index is used.
Signed-off-by: Eric Farman <farman@linux.vnet.ibm.com>
Reviewed-by: Viktor Mihajlovski <mihajlov@linux.vnet.ibm.com>
We had been setting the device alias in the devinceinfo for pci
controllers to "pci%u", but then hardcoding "pci.%u" when creating the
device address for other devices using that pci bus. This all worked
just fine until we encountered the built-in "pcie.0" bus (the PCIe
root complex) in Q35 machines.
In order to create the correct commandline for this one case, this
patch:
1) sets the alias for PCI controllers correctly, to "pci.%u" (or
"pcie.%u" for the pcie-root controller)
2) eliminates the hardcoded "pci.%u" for pci controllers when
generatuing device address strings, and instead uses the controller's
alias.
3) plumbs a pointer to the virDomainDef all the way down to
qemuBuildDeviceAddressStr. This was necessary in order to make the
aliase of the controller *used by a device* available (previously
qemuBuildDeviceAddressStr only had the deviceinfo of the device
itself, *not* of the controller it was connecting to). This made for a
larger than desired diff, but at least in the future we won't have to
do it again, since all the information we could possibly ever need for
future enhancements is in the virDomainDef. (right?)
This should be done for *all* controllers, but for now we just do it
in the case of PCI controllers, to reduce the likelyhood of
regression.
Introduced in commit 24b08219; compilation on RHEL 6.4 complained:
qemu/qemu_hotplug.c: In function 'qemuDomainAttachChrDevice':
qemu/qemu_hotplug.c:1257: error: declaration of 'remove' shadows a global declaration [-Wshadow]
/usr/include/stdio.h:177: error: shadowed declaration is here [-Wshadow]
* src/qemu/qemu_hotplug.c (qemuDomainAttachChrDevice): Avoid the
name 'remove'.
Signed-off-by: Eric Blake <eblake@redhat.com>
There are two levels on which a device may be hotplugged: config
and live. The config level requires just an insert or remove from
internal domain definition structure, which is exactly what this
patch does. There is currently no implementation for a chardev
update action, as there's not much to be updated. But more
importantly, the only thing that can be updated is path or socket
address by which chardevs are distinguished. So the update action
is currently not supported.
If an error occurs during qemuDomainAttachNetDevice after the macvtap
was created in qemuPhysIfaceConnect, the macvtap device gets left behind.
This patch adds code to the cleanup routine to delete the macvtap.
Signed-off-by: Matthew Rosato <mjrosato@linux.vnet.ibm.com>
Reviewed-by: Viktor Mihajlovski <mihajlov@linux.vnet.ibm.com>
I recently patches the callers to virPCIDeviceReset() to not call it
if the current driver for a device was vfio-pci (since that driver
will always reset the device itself when appropriate. At the time, Dan
Berrange suggested that I could instead modify virPCIDeviceReset
to check the currently bound driver for the device, and decide
for itself whether or not to go ahead with the reset.
This patch removes the previously added checks, and replaces them with
a check down in virPCIDeviceReset(), as suggested.
The functional difference here is that previously we were deciding
based on either the hostdev configuration or the value of
stubDriverName in the virPCIDevice object, but now we are actually
comparing to the "driver" link in the device's sysfs entry
directly. In practice, both should be the same.
Convert the type of loop iterators named 'i', 'j', k',
'ii', 'jj', 'kk', to be 'size_t' instead of 'int' or
'unsigned int', also santizing 'ii', 'jj', 'kk' to use
the normal 'i', 'j', 'k' naming
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
I just learned that VFIO resets PCI devices when they are assigned to
guests / returned to the host, so it is redundant for libvirt to reset
the devices. This patch inhibits calling virPCIDeviceReset to devices
that will be/were assigned using VFIO.
Commit 752596b5 broke the build with -Werror
qemu/qemu_hotplug.c: In function 'qemuDomainChangeGraphics':
qemu/qemu_hotplug.c:1980:39: error: declaration of 'listen' shadows a
global declaration [-Werror=shadow]
Fix with s/listen/newlisten/
Currently, we have a bug when updating a graphics device. A graphics device can
have a listen address set. This address is either defined by user (in which case
it's type is VIR_DOMAIN_GRAPHICS_LISTEN_TYPE_ADDRESS) or it can be inherited
from a network (in which case it's type is
VIR_DOMAIN_GRAPHICS_LISTEN_TYPE_NETWORK). However, in both cases we have a
listen address to process (e.g. during migration, as I've tried to fix in
7f15ebc7).
Later, when a user tries to update the graphics device (e.g. set a password),
we check if listen addresses match the original as qemu doesn't know how to
change listen address yet. Hence, users are required to not change the listen
address. The implementation then just dumps listen addresses and compare them.
Previously, while dumping the listen addresses, NULL was returned for NETWORK.
After my patch, this is no longer true, and we get a listen address for olddev
even if it is a type of NETWORK. So we have a real string on one side, the NULL
from user's XML on the other side and hence we think user wants to change the
listen address and we refuse it.
Therefore, we must take the type of listen address into account as well.
If we are just ejecting media, ret == -1 even after the retry loop
determines that the tray is open, as requested. This means media
disconnect always report's error.
Fix it, and fix some other mini issues:
- Don't overwrite the 'eject' error message if the retry loop fails
- Move the retries decrement inside the loop, otherwise the final loop
might succeed, yet retries == 0 and we will raise error
- Setting ret = -1 in the disk->src check is unneeded
- Fix comment typos
cc: mprivozn@redhat.com
In order to learn libvirt multiqueue several things must be done:
1) The '/dev/net/tun' device needs to be opened multiple times with
IFF_MULTI_QUEUE flag passed to ioctl(fd, TUNSETIFF, &ifr);
2) Similarly, '/dev/vhost-net' must be opened as many times as in 1)
in order to keep 1:1 ratio recommended by qemu and kernel folks.
3) The command line construction code needs to switch from 'fd=X' to
'fds=X:Y:...:Z' and from 'vhostfd=X' to 'vhostfds=X:Y:...:Z'.
4) The monitor handling code needs to learn to pass multiple FDs.
In 84c59ffa I've tried to fix changing ejectable media process. The
process should go like this:
1) we need to call 'eject' on the monitor
2) we should wait for 'DEVICE_TRAY_MOVED' event
3) now we can issue 'change' command
However, while waiting in step 2) the domain monitor was locked. So
even if qemu reported the desired event, the proper callback was not
called immediately. The monitor handling code needs to lock the
monitor in order to read the event. So that's the first lock we must
not hold while waiting. The second one is the domain lock. When
monitor handling code reads an event, the appropriate callback is
called then. The first thing that each callback does is locking the
corresponding domain as a domain or its device is about to change
state. So we need to unlock both monitor and VM lock. Well, holding
any lock while sleep()-ing is not the best thing to do anyway.
Since 0d70656afd, it starts to access the sysfs files to build
the qemu command line (by virSCSIDeviceGetSgName, which is to find
out the scsi generic device name by adpater🚌target:unit), there
is no way to work around, qemu wants to see the scsi generic device
like "/dev/sg6" anyway.
And there might be other places which need to access sysfs files
when building qemu command line in future.
Instead of increasing the arguments of qemuBuildCommandLine, this
introduces a new callback for qemuBuildCommandLine, and thus tests
can register their own callbacks for sysfs test input files accessing.
* src/qemu/qemu_command.h: (New callback struct
qemuBuildCommandLineCallbacks;
extern buildCommandLineCallbacks)
* src/qemu/qemu_command.c: (wire up the callback struct)
* src/qemu/qemu_driver.c: (Use the new syntax of qemuBuildCommandLine)
* src/qemu/qemu_hotplug.c: Likewise
* src/qemu/qemu_process.c: Likewise
* tests/testutilsqemu.[ch]: (Helper testSCSIDeviceGetSgName;
callback struct testCallbacks;)
* tests/qemuxml2argvtest.c: (Use testCallbacks)
* src/tests/qemuxmlnstest.c: (Like above)
This adds both attachment and detachment support for scsi host
device.
Signed-off-by: Han Cheng <hanc.fnst@cn.fujitsu.com>
Signed-off-by: Osier Yang <jyang@redhat>
It's better to put the usb related codes into qemuDomainAttachHostUsbDevice
instead of qemuDomainAttachHostDevice.
And in the old qemuDomainAttachHostDevice, just stealing the "usb" from
driver->activeUsbHostdevs leaks the memory.
The source code base needs to be adapted as well. Some files
include virutil.h just for the string related functions (here,
the include is substituted to match the new file), some include
virutil.h without any need (here, the include is removed), and
some require both.
The USB-specific cgroup setup had been inserted inline in
qemuDomainAttachHostUsbDevice and qemuSetupCgroup, but now there is a
common cgroup setup function called for all hostdevs, so it makes sens
to put the usb-specific setup there and just rely on that function
being called.
The one thing I'm uncertain of here (and a reason for not pushing
until after release) is that previously hostdev->missing was checked
only when starting a domain (and cgroup setup for the device skipped
if missing was true), but with this consolidation, it is now checked
in the case of hotplug as well. I don't know if this will have any
practical effect (does it make sense to hotplug a "missing" usb
device?)
PCIO device assignment using VFIO requires read/write access by the
qemu process to /dev/vfio/vfio, and /dev/vfio/nn, where "nn" is the
VFIO group number that the assigned device belongs to (and can be
found with the function virPCIDeviceGetVFIOGroupDev)
/dev/vfio/vfio can be accessible to any guest without danger
(according to vfio developers), so it is added to the static ACL.
The group device must be dynamically added to the cgroup ACL for each
vfio hostdev in two places:
1) for any devices in the persistent config when the domain is started
(done during qemuSetupCgroup())
2) at device attach time for any hotplug devices (done in
qemuDomainAttachHostDevice)
The group device must be removed from the ACL when a device it
"hot-unplugged" (in qemuDomainDetachHostDevice())
Note that USB devices are already doing their own cgroup setup and
teardown in the hostdev-usb specific function. I chose to make the new
functions generic and call them in a common location though. We can
then move the USB-specific code (which is duplicated in two locations)
to this single location. I'll be posting a followup patch to do that.
This isn't strictly speaking a bugfix, but I realized I'd gotten a bit
too verbose when I chose the names for
VIR_DOMAIN_HOSTDEV_PCI_BACKEND_TYPE_*. This shortens them all a bit.
<source type='bridge'> uses a helper application to do the necessary
TUN/TAP setup to use an existing network bridge, thus letting
unprivileged users use TUN/TAP interfaces.
However, libvirt should be preventing QEMU from running any setuid
programs at all, which would include this helper program. From
a security POV, any setuid helper needs to be run by libvirtd itself,
not QEMU.
This is what this patch does. libvirt now invokes the setuid helper,
gets the TAP fd and then passes it to QEMU in the normal manner.
The path to the helper is specified in qemu.conf.
As a small advantage, this adds a <target dev='tap0'/> element to the
XML of an active domain using <interface type='bridge'>.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
VFIO requires all of the guest's memory and IO space to be lockable in
RAM. The domain's max_balloon is the maximum amount of memory the
domain can have (in KiB). We add a generous 1GiB to that for IO space
(still much better than KVM device assignment, where the KVM module
actually *ignores* the process limits and locks everything anyway),
and convert from KiB to bytes.
In the case of hotplug, we are changing the limit for the already
existing qemu process (prlimit() is used under the hood), and for
regular commandline additions of vfio devices, we schedule a call to
setrlimit() that will happen after the qemu process is forked.
The device option for vfio-pci is nearly identical to that for
pci-assign - only the configfd parameter isn't supported (or needed).
Checking for presence of the bootindex parameter is done separately
from constructing the commandline, similar to how it is done for
pci-assign.
This patch contains tests to check for proper commandline
construction. It also includes tests for parser-formatter-parser
roundtrips (xml2xml), because those tests use the same data files, and
would have failed had they been included before now.
qemu: xml/args tests for VFIO hostdev and <interface type='hostdev'/>
These should be squashed in with the patch that adds commandline
handling of vfio (they would fail at any earlier time).
There will soon be other items related to pci hostdevs that need to be
in the same part of the hostdevsubsys union as the pci address (which
is currently a single member called "pci". This patch replaces the
single member named pci with a struct named pci that contains a single
member named "addr".
Instead of calling virCgroupForDomain every time we need
the virCgrouPtr instance, just do it once at Vm startup
and cache a reference to the object in qemuDomainObjPrivatePtr
until shutdown of the VM. Removing the virCgroupPtr from
the QEMU driver state also means we don't have stale mount
info, if someone mounts the cgroups filesystem after libvirtd
has been started
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
f946462e14 changed behavior by settings
VIR_DOMAIN_DEVICE_ADDRESS_TYPE_PCI upfront. If we do so before invoking
qemuDomainPCIAddressEnsureAddr we merely try to set the PCI slot via
qemuDomainPCIAddressReserveSlot instead reserving a new address via
qemuDomainPCIAddressSetNextAddr which fails with
$ ~/run-tck-test domain/200-disk-hotplug.t
./scripts/domain/200-disk-hotplug.t .. # Creating a new transient domain
./scripts/domain/200-disk-hotplug.t .. 1/5 # Attaching the new disk /var/lib/jenkins/jobs/libvirt-tck-build/workspace/scratchdir/200-disk-hotplug/extra.img
# Failed test 'disk has been attached'
# at ./scripts/domain/200-disk-hotplug.t line 67.
# died: Sys::Virt::Error (libvirt error code: 1, message: internal error unable to reserve PCI address 0:0:0.0
# )
The VIR_ERR_NO_SUPPORT error code is reserved for cases where an
API is not implemented in a driver. It definitely should not be
used when an API execution fails due to unsupported operation.
We didn't yet expose the virtio device attach and detach functionality
for s390 domains as the device hotplug was very limited with the old
virtio-s390 bus. With the CCW bus there's full hotplug support for
virtio devices in QEMU, so we are adding this to libvirt too.
Since the virtio hotplug isn't limited to PCI anymore, we change the
function names from xxxPCIyyy to xxxVirtioyyy, where we handle all
three virtio bus types.
Signed-off-by: J.B. Joret <jb@linux.vnet.ibm.com>
Signed-off-by: Viktor Mihajlovski <mihajlov@linux.vnet.ibm.com>
For both AttachDevice and UpdateDevice APIs, if the disk device
is 'cdrom' or 'floppy', the operations could be ejecting, updating,
and inserting. For either ejecting or updating, the shared disk
entry of the original disk src has to be removed, because it's
not useful anymore.
And since the original disk def will be changed, new disk def passed
as argument will be free'ed in qemuDomainChangeEjectableMedia, so
we need to copy the orignal disk def before
qemuDomainChangeEjectableMedia, to use it for qemuRemoveSharedDisk.
Some functions were using virDomainDeviceInfo where virDevicePCIAddress
would suffice. Some were only using integers for slots and functions,
assuming the bus numbers are always 0.
Switch from virDomainDeviceInfoPtr to virDevicePCIAddressPtr:
qemuPCIAddressAsString
qemuDomainPCIAddressCheckSlot
qemuDomainPCIAddressReserveAddr
qemuDomainPCIAddressReleaseAddr
Switch from int slot to virDevicePCIAddressPtr:
qemuDomainPCIAddressReserveSlot
qemuDomainPCIAddressReleaseSlot
qemuDomainPCIAddressGetNextSlot
Deleted functions (they would take the same parameters
as ReserveAddr/ReleaseAddr do now.)
qemuDomainPCIAddressReserveFunction
qemuDomainPCIAddressReleaseFunction
With the majority of fields in the virQEMUDriverPtr struct
now immutable or self-locking, there is no need for practically
any methods to be using the QEMU driver lock. Only a handful
of helper APIs in qemu_conf.c now need it
From qemu's point of view these are still just tap devices, so there's
no reason they shouldn't work with vhost-net; as a matter of fact,
Raja Sivaramakrishnan <srajag00@yahoo.com> verified on libvir-list
that at least the qemu_command.c part of this patch works:
https://www.redhat.com/archives/libvir-list/2012-December/msg01314.html
(the hotplug case is extrapolation on my part).
To avoid confusion between 'virCapsPtr' and 'qemuCapsPtr'
do some renaming of various fucntions/variables. All
instances of 'qemuCapsPtr' are renamed to 'qemuCaps'. To
avoid that clashing with the 'qemuCaps' typedef though,
rename the latter to virQEMUCaps.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Currently the activePciHostdevs, inactivePciHostdevsd and
activeUsbHostdevs lists are all implicitly protected by the
QEMU driver lock. Now that the lists all inherit from the
virObjectLockable, we can make the locking explicit, removing
the dependency on the QEMU driver lock for correctness.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
To allow modifications to the lists to be synchronized, convert
virPCIDeviceList and virUSBDeviceList into virObjectLockable
classes. The locking, however, will not be self-contained. The
users of these classes will have to call virObjectLock/Unlock
in the critical regions.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Currently the virQEMUDriverPtr struct contains an wide variety
of data with varying access needs. Move all the static config
data into a dedicated virQEMUDriverConfigPtr object. The only
locking requirement is to hold the driver lock, while obtaining
an instance of virQEMUDriverConfigPtr. Once a reference is held
on the config object, it can be used completely lockless since
it is immutable.
NB, not all APIs correctly hold the driver lock while getting
a reference to the config object in this patch. This is safe
for now since the config is never updated on the fly. Later
patches will address this fully.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
https://bugzilla.redhat.com/show_bug.cgi?id=892289
It seems like with new udev within guest OS, the tray is locked,
so we need to:
- 'eject'
- wait for tray to open
- 'change'
Moreover, even when doing bare 'eject', we should check for
'tray_open' as guest may have locked the tray. However, the
waiting phase shouldn't be unbounded, so I've chosen 10 retries
maximum, each per 500ms. This should give enough time for guest
to eject a media and open the tray.
This avoids "Event negative_returns: A negative constant "-1" is passed as
an argument to a parameter that cannot be negative.". The called function
uses -1 to determine whether it needs to traverse all the hostdevs.
When LXC labels USB devices during hotplug, it is running in
host context, so it needs to pass in a vroot path to the
container root.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
When a network device's bridge connection is changed by
virDomainUpdateDevice, libvirt first removes the netdev's tap from its
old bridge, then adds it to the new bridge. Sometimes, due to a
network being destroyed while a guest device is still attached, the
tap may already be "removed" from the old bridge (or the old bridge
may not even exist any more); the existing code was needlessly failing
the update when this happened, making it impossible to recover from
the situation without completely detaching (i.e. removing) the netdev
from the guest and re-attaching.
Instead of failing the entire operation when removal of the tap from
the old bridge fails, this patch changes qemuDomainChangeNetBridge to
just log a warning and continue, allowing a reasonable recover from
the situation.
(you'll appreciate this change if you ever accidentally destroy a
network while your guests are still using it).
This fixes a problem that showed up during testing of:
https://bugzilla.redhat.com/show_bug.cgi?id=881480
Due to a logic error in the function that gets the name of the bridge
an interface connects to, any time a bridge was specified directly
(type='bridge') rather than indirectly (type='network'), An error
would be logged (although the operation would then complete
successfully):
Network type 6 is not supported
The final virReportError() in the function
qemuDomainNetGetBridgeName() was apparently avoided in the past with a
"goto cleanup" at the end of each case, but the case of bridge somehow
no longer has that final goto cleanup.
The proper solution is anyway to not rely on goto's, but put the error
log inside an else {} clause, so that it's executed only if the type
is neither bridge nor network (in reality, this function should only
ever be called for those two types, that's why this is an internal
error).
While making this change, the error message was also tuned to be more
correct (since it's not really the type of the network, but the type
of the interface, and it *is* otherwise supported, it's just that the
interface type in question doesn't *have* a bridge device associated
with it, or at least we don't know how to get it).
Since we can't (currently) rely on the ability to provide blanket
support for all possible network changes by calling the toplevel
netdev hostside disconnect/connect functions (due to qemu only
supporting a lockstep between initialization of host side and guest
side of devices), in order to support live change of an interface's
nwfilter we need to make a special purpose function to only call the
nwfilter teardown and setup functions if the filter for an interface
(or its parameters) changes. The pattern is nearly identical to that
used to change the bridge that an interface is connected to.
This patch was inspired by a request from Guido Winkelmann
<guido@sagersystems.de>, who tested an earlier version.
https://bugzilla.redhat.com/show_bug.cgi?id=876828
Commit 38c4a9cc introduced a regression in hot unplugging of disks
from qemu, where cgroup device ACLs were no longer being revoked
(thankfully not a security hole: cgroup ACLs only prevent open()
of the disk; so reverting the ACL prevents future abuse but doesn't
stop abuse from an fd that was already opened before the ACL change).
Commit 1b2ebf95 overlooked that there were two spots affected.
* src/qemu/qemu_hotplug.c (qemuDomainDetachDiskDevice):
Transfer backing chain before deletion.
* src/qemu/qemu_driver.c (qemuDomainDetachDeviceDiskLive): Fix
spacing (partly to ensure a different-looking patch).
Remove the obsolete 'qemud' naming prefix and underscore
based type name. Introduce virQEMUDriverPtr as the replacement,
in common with LXC driver naming style
https://bugzilla.redhat.com/show_bug.cgi?id=876828
Commit 38c4a9cc introduced a regression in hot unplugging of disks
from qemu, where cgroup device ACLs were no longer being revoked
(thankfully not a security hole: cgroup ACLs only prevent open()
of the disk; so reverting the ACL prevents future abuse but doesn't
stop abuse from an fd that was already opened before the ACL change).
The actual regression is due to a latent bug. The hot unplug code
was computing the set of files needing cgroup ACL revocation based
on the XML passed in by the user, rather than based on the domain's
details on which disk was being deleted. As long as the revoke
path was always recomputing the backing chain, this didn't really
matter; but now that we want to compute the chain exactly once and
remember that computation, we need to hang on to the backing chain
until after the revoke has happened.
* src/qemu/qemu_hotplug.c (qemuDomainDetachPciDiskDevice):
Transfer backing chain before deletion.
The libvirt coding standard is to use 'function(...args...)'
instead of 'function (...args...)'. A non-trivial number of
places did not follow this rule and are fixed in this patch.
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
For now, disk migration via block copy job is not implemented in
libvirt. But when we do implement it, we have to deal with the
fact that qemu does not yet provide an easy way to re-start a qemu
process with mirroring still intact. Paolo has proposed an idea
for a persistent dirty bitmap that might make this possible, but
until that design is complete, it's hard to say what changes
libvirt would need. Even something like 'virDomainSave' becomes
hairy, if you realize the implications that 'virDomainRestore'
would be stuck with recreating the same mirror layout.
But if we step back and look at the bigger picture, we realize that
the initial client of live storage migration via disk mirroring is
oVirt, which always uses transient domains, and that if a transient
domain is destroyed while a mirror exists, oVirt can easily restart
the storage migration by creating a new domain that visits just the
source storage, with no loss in data.
We can make life a lot easier by being cowards for now, forbidding
certain operations on a domain. This patch guarantees that we
never get in a state where we would have to restart a domain with
a mirroring block copy, by preventing saves, snapshots, migration,
hot unplug of a disk in use, and conversion to a persistent domain
(thankfully, it is still relatively easy to 'virsh undefine' a
running domain to temporarily make it transient, run tests on
'virsh blockcopy', then 'virsh define' to restore the persistence).
Later, if the qemu design is enhanced, we can relax our code.
The change to qemudDomainDefine looks a bit odd for undoing an
assignment, rather than probing up front to avoid the assignment,
but this is because of how virDomainAssignDef combines both a
lookup and assignment into a single function call.
* src/conf/domain_conf.h (virDomainHasDiskMirror): New prototype.
* src/conf/domain_conf.c (virDomainHasDiskMirror): New function.
* src/libvirt_private.syms (domain_conf.h): Export it.
* src/qemu/qemu_driver.c (qemuDomainSaveInternal)
(qemuDomainSnapshotCreateXML, qemuDomainRevertToSnapshot)
(qemuDomainBlockJobImpl, qemudDomainDefine): Prevent dangerous
actions while block copy is already in action.
* src/qemu/qemu_hotplug.c (qemuDomainDetachDiskDevice): Likewise.
* src/qemu/qemu_migration.c (qemuMigrationIsAllowed): Likewise.
This resolves:
https://bugzilla.redhat.com/show_bug.cgi?id=862515
which describes inconsistencies in dealing with duplicate mac
addresses on network devices in a domain.
(at any rate, it resolves *almost* everything, and prints out an
informative error message for the one problem that isn't solved, but
has a workaround.)
A synopsis of the problems:
1) you can't do a persistent attach-interface of a device with a mac
address that matches an existing device.
2) you *can* do a live attach-interface of such a device.
3) you *can* directly edit a domain and put in two devices with
matching mac addresses.
4) When running virsh detach-device (live or config), only MAC address
is checked when matching the device to remove, so the first device
with the desired mac address will be removed. This isn't always the
one that's wanted.
5) when running virsh detach-interface (live or config), the only two
items that can be specified to match against are mac address and model
type (virtio, etc) - if multiple netdevs match both of those
attributes, it again just finds the first one added and assumes that
is the only match.
Since it is completely valid to have multiple network devices with the
same MAC address (although it can cause problems in many cases, there
*are* valid use cases), what is needed is:
1) remove the restriction that prohibits doing a persistent add of a
netdev with a duplicate mac address.
2) enhance the backend of virDomainDetachDeviceFlags to check for
something that *is* guaranteed unique (but still work with just mac
address, as long as it yields only a single results.
This patch does three things:
1) removes the check for duplicate mac address during a persistent
netdev attach.
2) unifies the searching for both live and config detach of netdevices
in the subordinate functions of qemuDomainModifyDeviceFlags() to use the
new function virDomainNetFindIdx (which matches mac address and PCI
address if available, checking for duplicates if only mac address was
specified). This function returns -2 if multiple matches are found,
allowing the callers to print out an appropriate message.
Steps 1 & 2 are enough to fully fix the problem when using virsh
attach-device and detach-device (which require an XML description of
the device rather than a bunch of commandline args)
3) modifies the virsh detach-interface command to check for multiple
matches of mac address and show an error message suggesting use of the
detach-device command in cases where there are multiple matching mac
addresses.
Later we should decide how we want to input a PCI address on the virsh
commandline, and enhance detach-interface to take a --address option,
eliminating the need to use detach-device
* src/conf/domain_conf.c
* src/conf/domain_conf.h
* src/libvirt_private.syms
* added new virDomainNetFindIdx function
* removed now unused virDomainNetIndexByMac and
virDomainNetRemoveByMac
* src/qemu/qemu_driver.c
* remove check for duplicate max from qemuDomainAttachDeviceConfig
* use virDomainNetFindIdx/virDomainNetRemove instead
of virDomainNetRemoveByMac in qemuDomainDetachDeviceConfig
* use virDomainNetFindIdx instead of virDomainIndexByMac
in qemuDomainUpdateDeviceConfig
* src/qemu/qemu_hotplug.c
* use virDomainNetFindIdx instead of a homespun loop in
qemuDomainDetachNetDevice.
* tools/virsh-domain.c: modified detach-interface command as described
above
This patch resolves:
https://bugzilla.redhat.com/show_bug.cgi?id=805071
to the extent that it can be resolved with current qemu functionality.
It attempts to detect as many situations as possible when the simple
operation of disconnecting an existing tap device from one bridge and
attaching it to another will satisfy the change requested in
virDomainUpdateDeviceFlags() for a network device. Before this patch,
that situation could only be detected if the pre-change interface
*and* the post-change interface definition were both "type='bridge'".
After this patch, it can also be detected if the before or after
interfaces are any combination of type='bridge' and type='network'
(the networks can be <forward mode='nat|route|bridge'>, as long as
they use a Linux host bridge and not macvtap connections).
This extra effort is especially useful since the recent discovery that
a netdev_del+netdev_add combo (to reconnect the network device with
completely different hostside configuration) doesn't work properly
with current qemu (1.2) unless it is accompanied by the matching
device_del+device_add - see this mailing list message for details:
http://lists.nongnu.org/archive/html/qemu-devel/2012-10/msg02355.html
(A slight modification of the patch referenced there has been prepared
to apply on top of this patch, but won't be pushed until qemu can be
made to work with it.)
* qemuDomainChangeNet needs access to the virDomainDeviceDef that
holds the new netdef (so that it can clear out the virDomainDeviceDef
if it ends up using the NetDef to replace the original), so the
virDomainNetDefPtr arg is replaced with a virDomainDeviceDefPtr.
* qemuDomainChangeNet previously checked for *some* changes to the
interface config, but this check was by no means complete. It was also
a bit disorganized.
This refactoring of the code is (I believe) complete in its check of
all NetDef attributes that might be changed, and either returns a
failure (for changes that are simply impossible), or sets one of three
flags:
needLinkStateChange - if the device link state needs to go up/down
needBridgeChange - if everything else is the same, but it needs
to be connected to a difference linux host
bridge
needReconnect - if the entire host side of the device needs
to be torn down and reconstructed (currently
non-working, as mentioned above)
Note that this function will refuse to make any change that requires
the *guest* side of the device to be detached (e.g. changing the PCI
address or mac address). Those would be disruptive enough to the guest
that it's reasonable to require an explicit detach/attach sequence
from the management application.
* As mentioned above, qemuDomainChangeNet also does its best to
understand when a simple change in attached bridge for the existing
tap device will work vs. the need to completely tear down/reconstruct
the host side of the device (including tap device).
This patch *does not* implement the "reconnect" code anyway - there is
a placeholder that turns that into an error. Rather, the purpose of
this patch is to replicate existing behavior with code that is ready
to have that functionality plugged in in a later patch.
* The expanded uses for qemuDomainChangeNetBridge meant that it needed
to be enhanced as well - it no longer replaces the original brname
string in olddev with the new brname; instead, it relies on the
caller to replace the *entire* olddev with newdev (since we've gone
to great lengths to assure they are functionally identical other
than the name of the bridge, this is now not only safe, but more
correct). Additionally, qemuDomainNetChangeBridge can now set the
bridge for type='network' interfaces as well as plain type='bridge'
interfaces. (Note that I had to make this change simultaneous to the
reorganization of qemuDomainChangeNet because the two are too
closely intertwined to separate).
All USB device lookup functions emit an error when they cannot find the
requested device. With this patch, their caller can choose if a missing
device is an error or normal condition.
The code which looks up a USB device specified by hostdev is duplicated
in two places. This patch creates a dedicated function that can be
called in both places.
Many parts of virDomainDefPtr were using 'int' variables as
array length counts. Replace all these with size_t and update
various format strings & API signatures to adapt
Signed-off-by: Daniel P. Berrange <berrange@redhat.com>
Disk hotplug is a two phase action: qemuMonitorAddDrive followed by
qemuMonitorAddDevice. When the first part succeeds but the second one
fails, we need to rollback the drive addition.