John Ferlan [Tue, 5 Apr 2016 09:43:47 +0000 (05:43 -0400)]
network: Remove extraneous ATTRIBUTE_NONNULL
Commit id '3992ff14' added the prototype for networkGetActualType
with 1 parameter, but added 2 ATTRIBUTE_NONNULL's (assume from a
cut-n-paste), just remove (2).
Commit d77ffb6876 added not only reporting of the PCI header type, but
also parsing of that information. However, because there was no parsing
done for the other sub-PCI capabilities, if there was any other
capability then a valid header type name (like phys_function or
virt_functions) the parsing would fail. This prevented passing node
device XMLs that we generated into our own functions when dealing with,
e.g. with SRIOV cards.
Instead of reworking the whole parsing, just fix this one occurence and
remove a test for it for the time being. Future patches will deal with
the rest.
Signed-off-by: Martin Kletzander <mkletzan@redhat.com>
Laine Stump [Fri, 1 Apr 2016 17:18:57 +0000 (13:18 -0400)]
qemu: fix alias name for <interface type='hostdev'>
Starting with commit f8e712fe, if you start a domain that has an
<interface type='hostdev' (or that has <interface type='network'>
where the network is a pool of devices for hostdev assignment), when
you later try to add *another* interface (of any kind) with hotplug,
the function qemuAssignDeviceNetAlias() fails as soon as it sees a
"hostdevN" alias in the list of interfaces), causing the attach to
fail.
This is because (starting with f8e712fe) the device alias names are
assigned during the new function qemuProcessPrepareDomain(), which is
called *before* networkAllocateActualDevice() (which is called from
qemuProcessPrepareHost(), which is called from
qemuProcessLaunch()). Prior to that commit,
networkAllocateActualDevice() was called first.
The problem with this is that the alias for interfaces that are really
a hostdev (<interface type='hostdev'>) is of the form "hostdevN" (just
like other hostdevs), while other interfaces are "netN". But if you
don't know that the interface is going to be a hostdev at the time you
assign the alias name, you can't name it differently. (As far as I've
seen so far, the change in name by itself wouldn't have been a problem
(other than just an outwardly noticeable change in behavior) except
for the abovementioned failure to attach/detach new interfaces.
Rather than take the chance that there may be other not-yet-revealed
problems associated with changing the alias name, this patch changes
the way that aliases are assigned to restore the old behavior.
Old: In the past, assigning an alias to an interface was skipped if it
was seen that the interface was type='hostdev' - we knew that the
hostdev part of the interface was also in the list of hostdevs (that's
part of what happens in networkAllocateActualDevice()) and it would be
assigned when all the other hostdev aliases were assigned.
New: When assigning an alias to an interface, we haven't yet called
networkAllocateActualDevice() to construct the hostdev part of the
interface, so we can't just wait for the loop that creates aliases for
all the hostdevs (there's nothing on that list for this device
yet!). Instead we handle it immediately in the loop creating interface
aliases, by calling the new function networkGetActualType() to
determine if it is going to be hostdev, and if so calling
qemuAssignDeviceHostdevAlias() instead.
Some adjustments have to be made to both
qemuAssignDeviceHostdevAlias() and to qemuAssignDeviceNetAlias() to
accommodate this. In both of them, an error return from
qemuDomainDeviceAliasIndex() is no longer considered an error; instead
it's just ignored (because it almost certainly means that the alias
string for the device was "net" when we expected "hostdev" or vice
versa). in qemuAssignDeviceHostdevAlias() we have to look at all
interface aliases for hostdevN in addition to looking at all hostdev
aliases (this wasn't necessary in the past, because both the interface
entry and the hostdev entry for the device already pointed at the
device info; no longer the case since the hostdev entry hasn't yet
been setup).
Fortunately the buggy behavior hasn't yet been in any official release
of libvirt.
Laine Stump [Fri, 1 Apr 2016 14:40:23 +0000 (10:40 -0400)]
qemu: change args to qemuAssignDeviceHostdevAlias()
In certain cases, we need to assign a hostdevN-style alias in a case
when we don't have a virDomainHostdevDefPtr (instead we have a
virDomainNetDefPtr). Since qemuAssignDeviceHostdevAlias() doesn't use
anything in the virDomainHostdevDef except the alias string itself
anyway, this patch just changes the arguments to pass a pointer to the
alias pointer instead.
Laine Stump [Fri, 1 Apr 2016 13:45:51 +0000 (09:45 -0400)]
network: new function networkGetActualType
There are times when it's necessary to learn the actual type of a
network connection before any resources have been allocated
(e.g. during qemuProcessPrepareDomain()), but in the past it was
necessary to call networkAllocateActualDevice() in order to have the
actual type filled in.
This new function returns the type of network that *will be* setup
once it actually happens, but without making any changes on the host.
The paths have the domain ID in them. Without cleaning them, they would
contain the same ID even after multiple restarts. That could cause
various problems, e.g. with access.
Add function qemuDomainClearPrivatePaths() for this as a counterpart of
qemuDomainSetPrivatePaths().
Signed-off-by: Martin Kletzander <mkletzan@redhat.com>
This unbreaks launching QEMU/KVM VMs with apparmor enabled. It also adds
the directory for the qemu guest-agent socket which is not known when
parsing the domain XML.
Since we're still not persisting our bookkeeping lists across
daemon restarts, we might have lost some information
virPCIDeviceReattach() relies on, for example whether the
device needs to be unbound from the stub driver.
As a result, if the daemon has been restarted in the meantime,
the device might end up remaining bound to the stub driver even
after 'virsh nodedev-reattach' or similar has been called, with
no way of giving it back to the host short of messing with
sysfs behind libvirt's back.
Revert back to the previous behavior of always trying to bind
the device to the host driver, regardless of its status when it
was detached, until persistent bookkeeping lists have been
implemented.
According to MDN[1], 'margin-left' and similar CSS properties,
including 'margin-right', cannot be applied to the '::first-line'
pseudo-element, so this rule will never have any effect and can
be safely removed.
Pass the correct cpu count when calling virDomainGetCPUStats.
When using the --start option, the show_count should not be set to
max_id as the --start <cpu> means we dont need those many initial cpu
stats. Hence, show_count should be adjusted accordingly.
Peter Krempa [Thu, 31 Mar 2016 14:45:14 +0000 (16:45 +0200)]
qemu: perf: Tweak flags before using them
@flags have a valid modification impact only after calling
virDomainObjUpdateModificationImpact. virDomainObjGetOneDef calls it but
doesn't update them in the caller.
which I incorrectly modified before pushing as commit b5534e53. My
modification caused network devices of type hostdev to no longer
be removed. This patch changes b5534e53 to resemble Chunyan's
original, correct patch.
I made what was thought to be an improvement and pushed the patch as
commit e6336442. As it turns out, my change broke adding net devices
that are actually hostdevs to the list of nets in virDomainDef. This
patch changes e6336442 to resemble Chunyan's original, correct
patch.
Fabiano Fidêncio [Wed, 30 Mar 2016 06:52:38 +0000 (08:52 +0200)]
build: workaround broken SASL header (again)
Compilation for xdg-app failed due to a buggy SASL headers present on
the used runtime (org.gnome.Sdk 3.18).
In file included from rpc/virnetsaslcontext.h:24:0,
from rpc/virnetsaslcontext.c:25:
/usr/include/sasl/sasl.h:230:38: error: unknown type name 'size_t'
typedef void *sasl_realloc_t(void *, size_t);
^
/usr/include/sasl/sasl.h:235:5: error: unknown type name 'sasl_realloc_t'
sasl_realloc_t *,
We use _LAST items in enums to mark the last position in given
enum. Now, if and enum is passed to switch(), compiler checks
that all the values from enum occur in 'case' enumeration.
Including _LAST. But coverity spots it's a dead code. And it
really is. So to resolve this, we tend to put a comment just
above 'case ..._LAST' notifying coverity that we know this is a
dead code but we want to have it that way.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Michal Privoznik [Thu, 31 Mar 2016 14:43:08 +0000 (16:43 +0200)]
virPerfReadEvent: Prefer saferead over read
Do I really need to explain why?
Well, if read() is interrupted int the middle of reading, we will
never read the rest (even though it's highly unlikely as we are
reading just 8 bytes).
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Michal Privoznik [Wed, 30 Mar 2016 16:33:29 +0000 (18:33 +0200)]
virDomain{Get,Set}PerfEvents: support --config --live --current
Now that we have @flags we can support changing perf events just
in active or inactive configuration regardless of the other.
Previously, calling virDomainSetPerfEvents set events in both
active and inactive configuration at once. Even though we allow
users to set perf events that are to be enabled once domain is
started up. The virDomainGetPerfEvents API was flawed too. It
returned just runtime info.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Michal Privoznik [Wed, 30 Mar 2016 15:56:02 +0000 (17:56 +0200)]
virsh: Make perf accept event list separated by commas
Everywhere else we use a comma separated list. There's no good
reason to make 'perf' command an exception. Currently, it accepts
string list separated by '|'.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Michal Privoznik [Wed, 30 Mar 2016 15:40:50 +0000 (17:40 +0200)]
virDomain{Get,Set}PerfEvents: Add @flags argument
I've noticed that these APIs are missing @flags argument. Even
though we don't have a use for them, it's our policy that every
new API must have @flags.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Peter Krempa [Thu, 24 Mar 2016 14:56:16 +0000 (15:56 +0100)]
qemu: command: Split up formatting of -numa and memory devices
They recently were extracted to a separate function. They don't belong
together though. Since -numa formatting is pretty compact, move it to
the main function and rename qemuBuildNumaCommandLine to
qemuBuildMemoryDeviceCommandLine.
Peter Krempa [Thu, 24 Mar 2016 15:05:11 +0000 (16:05 +0100)]
qemu: command: Pass numad nodeset when formatting memory devices at boot
When starting up a VM libvirtd asks numad to place the VM in case of
automatic nodeset. The nodeset would not be passed to the memory device
formatter and the user would get an error.
Pavel Hrdina [Tue, 29 Mar 2016 13:17:45 +0000 (15:17 +0200)]
qemu_process: add check for hyperv features
Commit 7068b56c introduced several hyperv features. Not all hyperv
features are supported by old enough kernels and we shouldn't allow to
start a guest if kernel doesn't support any of the hyperv feature.
There is one exception, for backward compatibility we cannot error out
if one of the RELAXED, VAPIC or SPINLOCKS isn't supported, for the same
reason we ignore invtsc, to not break restoring saved domains with older
libvirt.
Pavel Hrdina [Tue, 29 Mar 2016 13:13:52 +0000 (15:13 +0200)]
qemu_process: skip only cpu features
This check is there to allow restore saved domain with older libvirt
where we included invtsc by default for host-passthrough model. Don't
skip the whole function, but only the part that checks for invtsc.
Andrea Bolognani [Tue, 29 Mar 2016 16:40:34 +0000 (18:40 +0200)]
host-validate: Fix suggestion for missing cpu cgroup
If the cpu cgroup is not found when validating an host for
LXC support, virt-host-validate will suggest to enable the
CONFIG_CGROUP_SCHED kconfig option.
The appropriate option is really CONFIG_CGROUP_CPU. The
QEMU checks already get that right, so no changes needed.
* tools/nss/libvirt_nss.[ch]: add BSD-comptabile wrappers and
register via the nss_module_register() interface
* m4/virt-nss.m4: add checks if we're building NSS for FreeBSD
* tools/Makefile.am: handle target library name differences, as
Linux needs libnss_libvirt.so.2 and FreeBSD needs
nss_libvirt.so.1. Also, different syms files have to be used
as Linux needs to export all the methods while FreeBSD
only needs to have nss_module_register()
* tests/nsstest.c, tests/nssmock.c: s/__linux__/NSS/
* tests/nssmock.c: pass int instead of mode_t to va_arg() to please
gcc 4.8
* libvirt_nss_bsd.syms: FreeBSD syms file
Chunyan Liu [Mon, 28 Mar 2016 19:55:42 +0000 (13:55 -0600)]
libxl: fix resource leaks in libxlDomainStart error paths
libxlDomainStart allocates and reserves resources that were not
being released in error paths. libxlDomainCleanup already handles
the job of releasing resources, and libxlDomainStart should call
it when encountering a failure.
Change the error handling logic to call libxlDomainCleanup on
failure. This includes acquiring the lease sooner and allowing
it to be released in libxlDomainCleanup on failure, similar to
the way other resources are reclaimed. With the lease now
released in libxlDomainCleanup, the release_dom label can be
renamed to cleanup_dom to better reflect its changed semantics.
Peter Krempa [Mon, 21 Mar 2016 16:04:35 +0000 (17:04 +0100)]
conf: decrease iterations complexity when formatting iothreads
Create a bitmap of iothreads that have scheduler info set so that the
transformation algorithm does not have to iterate the empty bitmap many
times. By reusing self-expanding bitmaps the bitmap size does not need
to be pre-calculated.
In some cases it's impractical to use the regular APIs as the bitmap
size needs to be pre-declared. These new APIs allow to use bitmaps that
self expand.
The new code adds a property to the bitmap to track the allocation of
memory so that VIR_RESIZE_N can be used.
Jiri Denemark [Tue, 29 Mar 2016 13:35:19 +0000 (15:35 +0200)]
qemu: Fix /proc/**/stat parsing
Since commit v1.3.2-119-g1e34a8f which enabled debug-threads in QEMU
qemuGetProcessInfo would fail to parse stats for any thread with a space
in its name.
Peter Krempa [Tue, 29 Mar 2016 13:47:40 +0000 (15:47 +0200)]
vsh: Introduce helper to parse --bandwidth
Historically we've used 'unsigned long' and allowed wrapping of negative
numbers for bandwidth values. Add a helper that will simplify adding
support for scaled integers and support for byte granularity while
keeping the compatibility with the older approach.
Peter Krempa [Thu, 17 Mar 2016 14:40:43 +0000 (15:40 +0100)]
vsh: Tweak error message for scaled integers
It was too similar to the non-scaled alternative.
before:
error: Numeric value 'abc' for <size> option is malformed or out of range
after:
error: Scaled numeric value 'abc' for <size> option is malformed or out of range
Peter Krempa [Fri, 11 Mar 2016 15:33:03 +0000 (16:33 +0100)]
qemu: monitor: Remove JSON impls of drive_add and drive_del
qemu won't ever add those functions directly to QMP. They will be
replaced with 'blockdev-add' and 'blockdev-del' eventually. At this time
there's no need to keep the stubs around.
Additionally the drive_del stub in JSON contained dead code in the
attempt to report errors. (VIR_ERR_OPERATION_UNSUPPORTED was never
reported). Since the text impl does have the same message it is reported
anyways.
Qiaowei Ren [Mon, 28 Mar 2016 13:30:28 +0000 (21:30 +0800)]
perf: implement a set of util functions for perf event
This patch implement a set of interfaces for perf event. Based on
these interfaces, we can implement internal driver API for perf,
and get the results of perf conuter you care about.
Michal Privoznik [Tue, 29 Mar 2016 11:19:56 +0000 (13:19 +0200)]
docs: Document NSS module
While we have a wiki page describing the feature [1] since the
feature is distributed in our .tar.gz we ought to document it. So
I went ahead, copied the wiki page and reformatted so it fits our
docs coding style.
1: http://wiki.libvirt.org/page/NSS_module
Signed-off-by: Michal Privoznik <mprivozn@redhat.com>
Nitesh Konkar [Tue, 22 Mar 2016 08:34:50 +0000 (04:34 -0400)]
storage: Initialize pool size parameters for refresh thread
If the pool creation thread happens to detect the luns in
the scsi target, the size parameters will be calculated as
part of the refreshPool called from storagePoolCreate().
This means the virStoragePoolFCRefreshThread (commit id
'512b874') waiting to run and "refresh" the pool will
essentially double the allocation and capacity values.
A separate refresh would correct the values.
To avoid this, the FCRefreshThread needs to reinitialize
the pool size values prior to calling virStorageBackendSCSIFindLUs
which eventually calls virStorageBackendSCSINewLun and
updates the size values for each volume found.
Peter Krempa [Tue, 29 Mar 2016 06:50:30 +0000 (08:50 +0200)]
storage: rbd: Fix build
After the recent commits the build didn't work for me. Fix it by
using size_t as the callback argument is using and the correct
formatter. The attempted fixup to use %llu as a formatter was wrong.
Jim Fehlig [Mon, 28 Mar 2016 18:59:36 +0000 (12:59 -0600)]
libxl: remove reference to non-existent out label
Commit e6336442 changed the 'out:' label to 'cleanup' in
libxlDomainAttachNetDevice(), but missed a comment referencing
the 'out:' label. Remove it from the comment since it is no
longer accurate anyhow.
Cole Robinson [Fri, 18 Mar 2016 19:57:11 +0000 (15:57 -0400)]
docs: website: more header spacing tweaks
- remove top padding for h1. this means page titles sit flush with the top
of the side bar (like 'The virtualization API' on the front page)
- up the top padding for the remaining sections. makes it visually easier
to tell adjacent header sections apart, especially in dense wiki pages
- use two different spacing levels for h2-h4 and h5-h6,
gives pages some more visual flexibility
- use a slightly lower bottom padding... this makes top padding stick out
more which makes it visually easier to differentiate between adjacent
header sections
John Ferlan [Mon, 28 Mar 2016 14:57:55 +0000 (10:57 -0400)]
docs: Update the hyperv feature qemu supported version
In order to follow recent comments which indicate support for specific
feature bits are supported by a specific QEMU version add the version
from whence the relaxed, vapic, and spinlocks support was added.
Maxim Nestratov [Thu, 10 Mar 2016 12:43:49 +0000 (15:43 +0300)]
conf: qemu: Add support for more HyperV Enlightenment features
This patch adds support for "vpindex", "runtime", "synic",
"stimer", and "vendor_id" features available in qemu 2.5+.
- When Hyper-V "vpindex" is on, guest can use MSR HV_X64_MSR_VP_INDEX
to get virtual processor ID.
- Hyper-V "runtime" enlightement feature allows to use MSR
HV_X64_MSR_VP_RUNTIME to get the time the virtual processor consumes
running guest code, as well as the time the hypervisor spends running
code on behalf of that guest.
- Hyper-V "synic" stands for Synthetic Interrupt Controller, which is
lapic extension controlled via MSRs.
- Hyper-V "stimer" switches on Hyper-V SynIC timers MSR's support.
Guest can setup and use fired by host events (SynIC interrupt and
appropriate timer expiration message) as guest clock events
- Hyper-V "reset" allows guest to reset VM.
- Hyper-V "vendor_id" exposes hypervisor vendor id to guest.
Signed-off-by: Nikolay Shirokovskiy <nshirokovskiy@virtuozzo.com> Signed-off-by: John Ferlan <jferlan@redhat.com>
Laine Stump [Mon, 28 Mar 2016 14:14:04 +0000 (10:14 -0400)]
util: avoid getting stuck on macvtapN name created outside libvirt
After the patches that added tracking of in-use macvtap names (commit
370608, first appearing in libvirt-1.3.2), if the function to allocate
a new macvtap device came to a device name created outside libvirt, it
would retry the same device name MACVLAN_MAX_ID (8191) times before
finally giving up in failure.
The problem was that virBitmapNextClearBit was always being called
with "0" rather than the value most recently checked (which would
increment each time through the loop), so it would always return the
same id (since we dutifully release that id after failing to create a
new device using it).
Anatole Denis [Fri, 11 Mar 2016 16:50:54 +0000 (17:50 +0100)]
tests: storagepoolxml2xmltest: Enable pool-rbd
This test was commited 4 years ago, but was never enabled in
storagepoolxml2xmltest.c. This patch reactivates it, conditionnaly on RBD
storage support being enabled
Anatole Denis [Fri, 11 Mar 2016 16:50:53 +0000 (17:50 +0100)]
tests: storagepoolxml2xmltest: Fix pool-rbd test
This test failed for two reasons:
* The uuid was missing from the input file
* The output file had the <name> in a different place from the actual output
The "if (vol->target.format != VIR_STORAGE_FILE_RAW)" check in the
createVol backend. This check is bogus because virStorageVolDefParseXML()
in conf/storage_conf.c sets target.format only if volOptions in
virStoragePoolTypeInfo has formatFromString set, and that's not the
case the zfs backend.
So the check always fails and breaks volume creation.