This function is not used anymore. Let's remove it.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Martin Kletzander <mkletzan@redhat.com>
It's funny how this went unnoticed for such a long time. Long
story short, if a domain is configured with
VIR_DOMAIN_NUMATUNE_MEM_STRICT libvirt doesn't really honour
that. This is because of 7e72ac7878 after which libvirt allowed
qemu to allocate memory just anywhere and only after that it used
some magic involving cpuset.memory_migrate and cpuset.mems to
move the memory to desired NUMA nodes. This was done in order to
work around some KVM bug where KVM would fail if there wasn't a
DMA zone available on the NUMA node. Well, while the work around
might stopped libvirt tickling the KVM bug it also caused a bug
on libvirt side: if there is not enough memory on configured NUMA
node(s) then any attempt to start a domain must fail. Because of
the way we play with guest memory domains can start just happily.
The solution is to move the child we've just forked into emulator
cgroup, set up cpuset.mems and exec() qemu only after that.
This basically reverts 7e72ac7878 which was a workaround
for kernel bug. This bug was apparently fixed because I've tested
this successfully with recent kernel.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Martin Kletzander <mkletzan@redhat.com>
This tries to fix the same problem as f1d6585300 but it's doing
so in a less invasive way.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Tested-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Jim Fehlig <jfehlig@suse.com>
The call to resolve the actual network type will turn any NICs with
type=network into one of the other types. Thus there should be no need
to handle type=network in later switch() statements jumping off the
actual type.
Reviewed-by: Cole Robinson <crobinso@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Ports allocated on virtual networks with type=nat|route|open all get
given an actual type of 'network'.
Only ports in networks with type=bridge use an actual type of 'bridge'.
This distinction makes little sense since the virtualization drivers
will treat both actual types in exactly the same way, as they're all
just bridge devices a VM needs to be connected to.
This doesn't affect user visible XML since the "actual" device XML
is internal only, but we need code to convert the data upgrades.
Reviewed-by: Laine Stump <laine@laine.org>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Firstly, VIR_STRDUP() accepts NULL, so there is no need to check
if the string we want to duplicate is not-NULL. Secondly,
virDomainNetSetModelString() also accepts NULL. Thirdly, we have
VIR_AUTOFREE().
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
This requires drivers to opt in to handle the raw modelstr
network model, all others will error if a passed in XML value
is not in the model enum.
Enable this feature for libxl/xen/xm and qemu drivers
Acked-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Cole Robinson <crobinso@redhat.com>
This converts the qemu driver to the net model enum, for all
the model values that we have hardcoded for various checks,
which adds e1000e, virtio-transitional, virtio-non-transitional,
usb-net, spapr-vlan, lan9118, smc91c111
Because the qemu driver has historically also allowed the raw
model string onto the qemu command line, this isn't a full
conversion. Unwinding that will require more thought. However
for all new driver code we should be adding explicit enum
values for any model name we have special handling for.
Remove the now unused virDomainNetStreqModelString
Acked-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Cole Robinson <crobinso@redhat.com>
This adds a network model enum. The virDomainNetDef property
is named 'model' like most other devices.
When the XML parser or a driver calls NetSetModelString, if
the passed string is in the enum, we will set net->model,
otherwise we copy the string into net->modelstr
Add a single example for the 'netfront' xen model, and wire
that up, just to verify it's all working
Acked-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Cole Robinson <crobinso@redhat.com>
To ease converting the net->model value to an enum, add
the wrapper functions:
virDomainNetGetModelString
virDomainNetSetModelString
virDomainNetStreqModelString
virDomainNetStrcaseeqModelString
Acked-by: Michal Privoznik <mprivozn@redhat.com>
Signed-off-by: Cole Robinson <crobinso@redhat.com>
hostdevs have a link back to the original network device. This is fairly
generic accepting any type of device, however, we don't intend to make
use of this approach in future. It can thus be specialized to network
devices.
Reviewed-by: Cole Robinson <crobinso@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The APIs for allocating/notifying/removing network ports just take
an internal domain interface struct right now. As a step towards
turning these into public facing APIs, add a virNetworkPtr argument
to all of them.
Reviewed-by: Cole Robinson <crobinso@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The port allocation APIs are currently called unconditionally for all
types of NIC, but (mostly) only do anything for NICs with type=network.
The exception is the port allocate API which does some validation even
for NICs with type!=network. Relying on this validation is flawed,
however, since the network driver may not even be installed. IOW virt
drivers must not delegate validation to the network driver for NICs
with type != network.
This change allows us to report errors when the virtual network driver
is not registered.
Reviewed-by: Cole Robinson <crobinso@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
This helps in a scenarios where vCPUs run with a priority that is so high they
might starve the emulator thread. And it also fits with the rest of the
settings.
Signed-off-by: Martin Kletzander <mkletzan@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Qemu commit 767abe7 ("chardev: forbid 'wait' option with client
sockets") effectively deprecates usage of "wait" with client sockets
starting with qemu 4.0, and earlier versions ignored the value.
Cc: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Signed-off-by: Ján Tomko <jtomko@redhat.com>
My earlier commit be46f61326 was incomplete. It removed caching of
microcode version in the CPU driver, which means the capabilities XML
will see the correct microcode version. But it is also cached in the
QEMU capabilities cache where it is used to detect whether we need to
reprobe QEMU. By missing the second place, the original commit
be46f61326 made the situation even worse since libvirt would report
correct microcode version while still using the old host CPU model
(visible in domain capabilities XML).
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Support for kqemu was dropped in libvirt by commit 8e91a400c and even
back then we never set these capabilities when doing QMP probing.
Since no QEMU we aim to support has these, drop them completely.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
Reviewed-by: Andrea Bolognani <abologna@redhat.com>
The NVIDIA V100 GPU has an onboard RAM that is mapped into the
host memory and accessible as normal RAM via an NVLink2 bridge. When
passed through in a guest, QEMU puts the NVIDIA RAM window in a
non-contiguous area, above the PCI MMIO area that starts at 32TiB.
This means that the NVIDIA RAM window starts at 64TiB and go all the
way to 128TiB.
This means that the guest might request a 64-bit window, for each PCI
Host Bridge, that goes all the way to 128TiB. However, the NVIDIA RAM
window isn't counted as regular RAM, thus this window is considered
only for the allocation of the Translation and Control Entry (TCE).
For more information about how NVLink2 support works in QEMU,
refer to the accepted implementation [1].
This memory layout differs from the existing VFIO case, requiring its
own formula. This patch changes the PPC64 code of
@qemuDomainGetMemLockLimitBytes to:
- detect if we have a NVLink2 bridge being passed through to the
guest. This is done by using the @ppc64VFIODeviceIsNV2Bridge function
added in the previous patch. The existence of the NVLink2 bridge in
the guest means that we are dealing with the NVLink2 memory layout;
- if an IBM NVLink2 bridge exists, passthroughLimit is calculated in a
different way to account for the extra memory the TCE table can alloc.
The 64TiB..128TiB window is more than enough to fit all possible
GPUs, thus the memLimit is the same regardless of passing through 1 or
multiple V100 GPUs.
Further reading explaining the background
[1] https://lists.gnu.org/archive/html/qemu-devel/2019-03/msg03700.html
[2] https://www.redhat.com/archives/libvir-list/2019-March/msg00660.html
[3] https://www.redhat.com/archives/libvir-list/2019-April/msg00527.html
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Erik Skultety <eskultet@redhat.com>
The NVLink2 support in QEMU implements the detection of NVLink2
capable devices by verifying the attributes of the VFIO mem region
QEMU allocates for the NVIDIA GPUs. To properly allocate an
adequate amount of memLock, Libvirt needs this information before
a QEMU instance is even created, thus querying QEMU is not
possible and opening a VFIO window is too much.
An alternative is presented in this patch. Making the following
assumptions:
- if we want GPU RAM to be available in the guest, an NVLink2 bridge
must be passed through;
- an unknown PCI device can be classified as a NVLink2 bridge
if its device tree node has 'ibm,gpu', 'ibm,nvlink',
'ibm,nvlink-speed' and 'memory-region'.
This patch introduces a helper called @ppc64VFIODeviceIsNV2Bridge
that checks the device tree node of a given PCI device and
check if it meets the criteria to be a NVLink2 bridge. This
new function will be used in a follow-up patch that, using the
first assumption, will set up the rlimits of the guest
accordingly.
Signed-off-by: Daniel Henrique Barboza <danielhb413@gmail.com>
This does not cause a problem in usual scenarios thanks to us allowing
CAP_DAC_OVERRIDE for the qemu process, however in some scenarios this might be
an issue because the directory is created with mkdtemp(3) which explicitly
creates that with 0700 permissions and qemu running as non-root cannot access
that.
The scenarios include:
- Builds without CAPNG
- Running libvirtd in certain container configurations [1]
- and possibly others.
[1] https://github.com/kubevirt/kubevirt/pull/2181#issuecomment-481840304
Signed-off-by: Martin Kletzander <mkletzan@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
The structure can only be used for CPUID data now. Adding a type
indicator and moving the data into a union will let us store alternative
data types.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
The following patches introduce CPU features read from MSR in addition
to those queried via CPUID instruction. Let's introduce a container
struct which will be able to describe either feature type.
Signed-off-by: Jiri Denemark <jdenemar@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Vim has trouble figuring out the filetype automatically because
the name doesn't follow existing conventions; annotations like
the ones we already have in Makefile.ci help it out.
Signed-off-by: Andrea Bolognani <abologna@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
The latter is deprecated and will be removed soon. The advised
replacement is '-overcommit mem-lock=on|off'.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Added in QEMU commit of v3.0.0-rc0~48^2~9 (then fixed by
v3.1.0-rc0~119^2~37) QEMU is replacing '-realtime mlock' with
'-overcommit mem-lock'. Add a capability to tell if we're dealing
new new enough qemu to use the replacement.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
The '-realtime mlock' cmd line argument was introduced in QEMU
commit v1.5.0-rc0~190 which matches minimal QEMU version we
require. Therefore, the capability will always be present.
Apparently, nearly none of our xml2argv test cases had the
capability hence slightly bigger change under qemuxml2argvdata/.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Standardize on putting the _LAST enum value on the second line
of VIR_ENUM_IMPL invocations. Later patches that add string labels
to VIR_ENUM_IMPL will push most of these to the second line anyways,
so this saves some noise.
Signed-off-by: Cole Robinson <crobinso@redhat.com>
Demonstrate how VIR_RETURN_PTR is used by refactoring qemu_block.c
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
This is a locally used helper struct but we can make use of automatic
freeing for it.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Commit 6864d8f740 moved this one level up
for qemuBuildMemoryBackendProps but left qemuBuildMemPathStr intact.
Signed-off-by: Ján Tomko <jtomko@redhat.com>
If a management application wants to use firmware auto selection
feature it can't currently know if the libvirtd it's talking to
support is or not. Moreover, it doesn't know which values that
are accepted for the @firmware attribute of <os/> when parsing
will allow successful start of the domain later, i.e. if the mgmt
application wants to use 'bios' whether there exists a FW
descriptor in the system that describes bios.
This commit then adds 'firmware' enum to <os/> element in
<domainCapabilities/> XML like this:
<enum name='firmware'>
<value>bios</value>
<value>efi</value>
</enum>
We can see both 'bios' and 'efi' listed which means that there
are descriptors for both found in the system (matched with the
machine type and architecture reported in the domain capabilities
earlier and not shown here).
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
The point of this API is to fetch all FW descriptors, parse them
and return list of supported interfaces and SMM feature for given
combination of machine type and guest architecture.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Acked-by: Laszlo Ersek <lersek@redhat.com>
This reverts commit a5e1602090.
Getting rid of unistd.h from our headers will require more work than
just fixing the broken mingw build. Revert it until I have a more
complete proposal.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
util/virutil.h bogously included unistd.h. Drop it and replace it by
including it directly where needed.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
virutil.(c|h) is a very gross collection of random code. Remove the enum
handlers from there so we can limit the scope where virtutil.h is used.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
'viralloc.h' does not provide any type or macro which would be necessary
in headers. Prevent leakage of the inclusion.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
Keeping them with viralloc.h forcibly pulls in the other stuff from
viralloc.h into other header files. This in turn creates a mess
as more and more headers pull in the 'viral' header file.
If we want to make 'viralloc.h' omnipresent we should pick a different
approach.
Signed-off-by: Peter Krempa <pkrempa@redhat.com>
Reviewed-by: Ján Tomko <jtomko@redhat.com>
