virDomainNetDefParseXML: Fix error message for unknown value of //mac/@type
In v6.6.0-rc1~124 we've introduced a new mechanism for MAC
addresses for ESX: ignore all checks (type='static') that libvirt
or ESX would do (and possibly fail) for specified MAC address.
Accepted values for the @type attribute are "generated" and
"static". But the error message mentions a different attribute.
Peter Krempa [Mon, 26 Oct 2020 12:06:17 +0000 (13:06 +0100)]
qemu: Update to new design of 'block-export-add'
qemu decided to modify the arguments of 'block-export-add' to include an
array of bitmaps rather than a single bitmap.
Since we've added the code prior to qemu setting the interface in stone
and thus it will be changed incompatibly and we already have tests for
the new interface we need to update the code and qemu capabilities data
at the same time.
Use a array of bitmaps as the 'bitmaps' argument instead of 'bitmap' and
bump qemu capabilities for the upcoming 5.2.0 release to v5.1.0-2827-g2c6605389c
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Eric Blake <eblake@redhat.com>
Peter Krempa [Mon, 26 Oct 2020 12:44:19 +0000 (13:44 +0100)]
docs: page.xsl: Improve generation of paragraph anchor links
Use the 'parent' axis to check whether the parent is a div with
class='section' rather than looking for 'toc-backref' anchor to see
whether to generate one of the headerlink alternatives. Both hare
docutils-specific thus apply to docs generated from RST documents.
This adds the links for pages generated from RST documents which don't
have a table of contents (and thus lack the 'toc-backref' anchors) and
thus fixes pages such as hacking.html and news.html to have reasonable
links which can be shared.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Laine Stump [Wed, 28 Oct 2020 01:43:47 +0000 (21:43 -0400)]
node_device: fix leak of DIR*
Commit 53aec799fa31 introduced the function udevGetVDPACharDev(),
which scans a directory using virDirOpenIfExists() and
virDirRead(). It unfortunately forgets to close the DIR* when it is
finished with it. This patch fixes that omission.
Fixes: 53aec799fa31711ffaeacc7ec17ec6d3c2e3cadf Signed-off-by: Laine Stump <laine@redhat.com> Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
GLibC has a really complicated way of dealing with the 'stat' function
historically, which means our mocks in turn have to look at four
different possible functions to replace, stat, stat64, __xstat,
__xstat64.
In Fedora 33 and earlier:
- libvirt.so links to __xstat64
- libc.so library exports stat, stat64, __xstat, __xstat64
- sys/stat.h header exposes stat and __xstat
In Fedora 34 rawhide:
- libvirt.so links to stat64
- libc.so library exports stat, stat64, __xstat, __xstat64
- sys/stat.h header exposes stat
Historically we only looked at the exported symbols from libc.so to
decide which to mock.
In F34 though we must not consider __xstat / __xstat64 though because
they only existance for binary compatibility. Newly built binaries
won't reference them.
Thus we must introduce a header file check into our logic for deciding
which symbol to mock. We must ignore the __xstat / __xstat64 symbols
if they don't appear in the sys/stat.h header, even if they appear
in libc.so
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
rpm: tell meson whether to use libssh or libssh2 explicitly
The %meson macro sets "--auto-features=enabled", thus any feature in the
RPM which has a "with_XXX" condition, needs to explicitly pass a
"-DXXX=state" arg to %meson to override the auto features setting.
The with_libssh and with_libssh2 conditions were not exposed to meson,
so if either was set disabled, then meson would fail the build if the
-devel packages were not found.
Reviewed-by: Pavel Hrdina <phrdina@redhat.com> Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The %meson macro sets "--auto-features=enabled", thus any feature in the
RPM which has a "with_XXX" condition, needs to explicitly pass a
"-DXXX=state" arg to %meson to override the auto features setting.
The with_bash_completion condition is always set to 1, so rather than
adding an arg to %meson, just remove the condition.
Reviewed-by: Pavel Hrdina <phrdina@redhat.com> Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Peter Krempa [Mon, 26 Oct 2020 10:26:50 +0000 (11:26 +0100)]
qemu: capabilities: Disable detection of QEMU_CAPS_BLOCK_EXPORT_ADD
We use the capability to switch to using 'block-export-add' in the
upcoming qemu release instead of the at the same time deprecated
'nbd-server-add'.
Unfortunately qemu wants to change the interface of 'block-export-add'
before the release. Since we've tried to stay up to date and added the
code before it was written in stone, we need to disable the use of the
new interface for the upcoming libvirt release so that we don't have a
version of libvirt which would not work with the upcoming qemu version.
Remove the detection of 'block-export-add' until we are more sure how
the qemu interface will look.
The cpu mask was free()'d immediately on any error and at the end of the
function, where it was expected that it would either error out and return or
goto another allocation if the code was to fail. However since commit 9514e24984ee the error path did not return in one new case which caused
double-free in such situation. In order to make the code more straightforward
just free the mask after it's been used even before checking the return code of
the call.
qemu: honour fatal errors dealing with qemu slirp helper
Currently all errors from qemuInterfacePrepareSlirp() are completely
ignored by the callers. The intention is that missing qemu-slirp binary
should cause the caller to fallback to the built-in slirp impl.
Many of the possible errors though should indeed be considered fatal.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Neal Gompa [Mon, 26 Oct 2020 21:53:55 +0000 (17:53 -0400)]
Revert "spec: Simplify setting features off by default"
As it turns out, the rather complicated structure that is
currently used for enabling or disabling features in the libvirt
build does not cleanly map well to RPM's bcond feature.
Consequently, we need these back in order to support trivially
activating these features through extra macros as build inputs.
Andrea Bolognani [Mon, 26 Oct 2020 23:25:53 +0000 (00:25 +0100)]
news: Mention virt-ssh-helper detection fix
Signed-off-by: Andrea Bolognani <abologna@redhat.com> Reviewed-by: Neal Gompa <ngompa13@gmail.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Andrea Bolognani [Mon, 26 Oct 2020 23:15:33 +0000 (00:15 +0100)]
rpc: Fix virt-ssh-helper detection
When trying to figure out whether virt-ssh-helper is available
on the remote host, we mistakenly look for the helper by the
name it had while the feature was being worked on instead of
the one that was ultimately picked, and thus end up using the
netcat fallback every single time.
Fixes: f8ec7c842df9e40c6607eae9b0223766cb226336 Signed-off-by: Andrea Bolognani <abologna@redhat.com> Reviewed-by: Neal Gompa <ngompa13@gmail.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Andrea Bolognani [Mon, 26 Oct 2020 13:37:23 +0000 (14:37 +0100)]
NEWS: Fix vertical spacing between sections
Looking at the entire repository reveals we're not too consistent
about this, but at least in this specific document we mostly have
two blank lines between sections, so let's stick with that.
Signed-off-by: Andrea Bolognani <abologna@redhat.com>
Michal Privoznik [Mon, 26 Oct 2020 10:28:13 +0000 (11:28 +0100)]
qemu_migration: Don't mangle NBD part of migration cookie
In recent commit v6.8.0-135-g518be41aaa the formatting of NBD
into migration cookie was moved into a separate function and with
it it was switched from direct printing into the output buffer to
virXMLFormatElement(). But there was a typo. The
virXMLFormatElement() accepts two buffers on input, one for
element attributes and another for child elements. Well, the line
that was supposed to add NBD port into the attributes buffer
printed the attribute directly into the output buffer which
produced this mangled XML:
Michal Privoznik [Mon, 26 Oct 2020 08:49:35 +0000 (09:49 +0100)]
qemu: Don't try to start NBD server twice
In one of recent patches the way that we start NBD server for
incoming migration was reworked (v6.8.0-rc1~298). A new boolean
was introduced that tracks whether the NBD server was started so
that we don't start it twice nor record in the port in the port
allocator twice. Well, this idea is good, but in the
implementation the boolean is never set, so we are reserving the
port twice and would be starting the NBD server twice too if it
wasn't for port reservation fail.
Fixes: e74d627bb3bc2684cbe3edc1e2f7cc745b4e1ff3 Reported-by: Vjaceslavs Klimovs <vklimovs@gmail.com> Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Peter Krempa <pkrempa@redhat.com>
zhenwei pi [Fri, 23 Oct 2020 13:53:49 +0000 (21:53 +0800)]
news: introduce memory failure event
Signed-off-by: zhenwei pi <pizhenwei@bytedance.com> Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Guests should be allowed to create hard links on mounted pathes, since
many applications rely on this functionality and would error on guest
with current "rw" AppArmor permission with 9pfs.
Signed-off-by: Christian Schoenebeck <qemu_oss@crudebyte.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Michal Privoznik [Fri, 23 Oct 2020 17:39:03 +0000 (19:39 +0200)]
qemu: Don't pass mode when opening domain log file for reading
In qemuDomainLogContextNew() the domain log file is opened.
Twice, the first time for writing, and the second time for
reading (if required by caller). When opening the log file for
reading a mode is provided. This doesn't do much harm, but is
unnecessary. Drop the mode.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Jonathon Jongsma [Wed, 14 Oct 2020 17:08:30 +0000 (12:08 -0500)]
Include vdpa devices in node device list
The current udev node device driver ignores all events related to vdpa
devices. Since libvirt now supports vDPA network devices, include these
devices in the device list.
Example output:
virsh # nodedev-list
[...ommitted long list of nodedevs...]
vdpa_vdpa0
NOTE: normally the 'parent' would be a PCI device instead of 'computer',
but this example output is from the vdpa_sim kernel module, so it
doesn't have a normal parent device.
Jonathon Jongsma [Wed, 21 Oct 2020 21:50:06 +0000 (16:50 -0500)]
qemu: fix potential resource leak
Coverity reported a potential resource leak. While it's probably not
a real-world scenario, the code could technically jump to cleanup
between the time that vdpafd is opened and when it is used. Ensure that
it gets cleaned up in that case.
Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com> Reviewed-by: Laine Stump <laine@redhat.com>
zhenwei pi [Wed, 14 Oct 2020 10:37:51 +0000 (18:37 +0800)]
qemu: implement memory failure event
Since QEMU 5.2 (commit-77b285f7f6), QEMU supports 'memory failure'
event, posts event to monitor if hitting a hardware memory error.
Fully support this feature for QEMU.
Test with commit 'libvirt: support memory failure event', build a
little complex environment(nested KVM):
1, install newly built libvirt in L1, and start a L2 vm. run command
in L1:
~# virsh event l2 --event memory-failure
2, run command in L0 to inject MCE to L1:
~# virsh qemu-monitor-command l1 --hmp mce 0 9 0xbd000000000000c0 0xd 0x62000000 0x8c
Test result in l1(recipient hypervisor case):
event 'memory-failure' for domain l2:
recipient: hypervisor
action: ignore
flags:
action required: 0
recursive: 0
Test result in l1(recipient guest case):
event 'memory-failure' for domain l2:
recipient: guest
action: inject
flags:
action required: 0
recursive: 0
Signed-off-by: zhenwei pi <pizhenwei@bytedance.com> Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
zhenwei pi [Wed, 14 Oct 2020 10:37:50 +0000 (18:37 +0800)]
libvirt: support memory failure event
Introduce memory failure event. Libvirt should monitor domain's
event, then posts it to uplayer. According to the hardware memory
corrupted message, a cloud scheduler could migrate domain to another
health physical server.
Several changes in this patch:
public API:
include/*
src/conf/*
src/remote/*
src/remote_protocol-structs
With this patch, each driver could implement its own method to run
this new event.
Signed-off-by: zhenwei pi <pizhenwei@bytedance.com> Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
Michal Privoznik [Wed, 21 Oct 2020 12:36:16 +0000 (14:36 +0200)]
hyperv: Use hypervRequestStateChange() in hypervDomainSuspend()
A few commits ago, hypervRequestStateChange() helper was
introduced which has exactly the same code as a part of
hypervDomainSuspend(). Deduplicate by calling the helper.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Matt Coleman <matt@datto.com>
Peter Krempa [Tue, 20 Oct 2020 16:43:20 +0000 (18:43 +0200)]
Replace all instances of 'virHashCreate' with 'virHashNew'
It doesn't make much sense to configure the bucket count in the hash
table for each case specifically. Replace all calls of virHashCreate
with virHashNew which has a pre-set size and remove virHashCreate
completely.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Peter Krempa [Wed, 21 Oct 2020 07:41:50 +0000 (09:41 +0200)]
tests: hash: Prepare for replacement of virHashCreate
Most callers pass a random number. We have virHashNew which doesn't give
the callers the option to configure the table. Since we are going to
switch to virHashNew replace it in tests and remove multiple instances
of the 'testHashGrow' case as it doesn't make sense with the new
semantics.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Peter Krempa [Tue, 20 Oct 2020 15:01:51 +0000 (17:01 +0200)]
util: cgroup: Use GHashTable instead of virHashTable
Rewrite using GHashTable which already has interfaces for using a number
as hash key. Glib's implementation doesn't copy the key by default, so
we need to allocate it, but overal the interface is more suited for this
case.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Peter Krempa [Thu, 22 Oct 2020 07:56:46 +0000 (09:56 +0200)]
virCgroupKillRecursive: Return -1 on failure condition
virCgroupKillRecursive sneakily initializes 'ret' to 0 rather than the
usual -1. 401030499bf moved an error condition but didn't actually
modify 'ret' return the proper error code.
Fixes: 401030499bf Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Pavel Hrdina <phrdina@redhat.com>
Michal Privoznik [Thu, 22 Oct 2020 10:55:40 +0000 (12:55 +0200)]
parthelper: Don't leak @canonical_path
The @canonical_path variable holds canonicalized path passed as
argv[1]. The canonicalized path is obtained either via
virFileResolveLink() or plain g_strdup(). Nevertheless, in both
cases it must be freed.
Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
Michal Privoznik [Thu, 22 Oct 2020 10:49:44 +0000 (12:49 +0200)]
parthelper: Initialize error object
Some functions called from parthelper can report an error. But
that means that the error object must be initialized otherwise
virResetError() (which happens as a part of virReportError())
will free random pointers.
Reported-by: Katerina Koukiou <kkoukiou@redhat.com> Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Ján Tomko <jtomko@redhat.com>
Use of the -enable-fips option is being deprecated in QEMU >= 5.2.0. If
FIPS compliance is required, QEMU must be built with libcrypt which will
unconditionally enforce it.
Thus there is no need for libvirt to pass -enable-fips to modern QEMU.
Unfortunately there was never any way to probe for -enable-fips in the
first instance, it was enabled by libvirt based on version number
originally, and then later unconditionally enabled when libvirt dropped
support for older QEMU. Similarly we now use a version number check to
decide when to stop passing -enable-fips.
Note that the qemu-5.2 capabilities are currently from the pre-release
version and will be updated once qemu-5.2 is released.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com> Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Peter Krempa [Wed, 21 Oct 2020 08:08:11 +0000 (10:08 +0200)]
qemuxml2argvtest: Sanitize testing of '-enable-fips'
Rename 'FLAG_FIPS' to 'FLAG_FIPS_HOST' to signify that we are simulating
a host supporting fips mode and use the flag to assert 'enabeFips'
argument of 'qemuProcessCreatePretendCmdBuild' rather than passing it
via QEMU_CAPS_ENABLE_FIPS.
This prepares the testsuite for testing of -enable-fips deprecation in
qemu-5.2.
Signed-off-by: Peter Krempa <pkrempa@redhat.com> Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Laine Stump [Sun, 11 Oct 2020 02:56:23 +0000 (22:56 -0400)]
util: avoid manual VIR_FREE of a g_autofree pointer in virPCIGetName()
thisPhysPortID is only used inside a conditional, so reduce its scope
to just the body of that conditional, which will eliminate the need
for the undesirable manual VIR_FREE().
Signed-off-by: Laine Stump <laine@redhat.com> Reviewed-by: Erik Skultety <eskultet@redhat.com> Reviewed-by: Peter Krempa <pkrempa@redhat.com>
Laine Stump [Sun, 11 Oct 2020 01:47:16 +0000 (21:47 -0400)]
util: simplify virPCIProbeStubDriver()
This function had a loop that was only executed twice; it was
artificially constructed with a label, a goto, and a boolean to tell
that it had already been executed once. Aside from that, the body of
the loop contained only two lines that needed to be repeated (the
second time through, everything beyond those two lines would be
skipped).
One side effect of this strange loop was that a g_autofree string was
manually freed and re-initialized; I've been told that manually
freeing a g_auto_free object is highly discouraged.
This patch refactors the function to simply repeat the 2 lines that
might possibly be executed twice, thus eliminating the ugly use of
goto to construct a loop, and also takes advantage of the fact that
virPCIDriverDir() was previously returning *exactly* the same string
both times it was called to eliminate the manual VIR_FREE of drvpath.
Signed-off-by: Laine Stump <laine@redhat.com> Reviewed-by: Erik Skultety <eskultet@redhat.com> Reviewed-by: Peter Krempa <pkrempa@redhat.com>
Laine Stump [Sat, 10 Oct 2020 21:46:20 +0000 (17:46 -0400)]
util: simplify virPCIFile() and its callers
There is no need for a temporary variable in this function, and ever
since we switched to glib for memory allocation, there is no possibility
it can return NULL, so callers don't need to check for it.
Signed-off-by: Laine Stump <laine@redhat.com> Reviewed-by: Erik Skultety <eskultet@redhat.com> Reviewed-by: Peter Krempa <pkrempa@redhat.com>
Laine Stump [Thu, 1 Oct 2020 18:33:17 +0000 (14:33 -0400)]
util: simplify virHostdevPCISysfsPath()
Apparently at some point in the past, when there were multiple types
to represent PCI addresses, the function
virPCIDeviceAddressGetSysfsFile() used one of those types, while
virDomainHostDevDef used another. It's been quite awhile since we
reduced the number of different representations of PCI address, but
this function was still creating a temporary virPCIDeviceAddress, then
copying the individual elements into this temporary object from the
same type of object in the virDomainHostDevDef.
This patch just eliminates that pointless copy.
Signed-off-by: Laine Stump <laine@redhat.com> Reviewed-by: Erik Skultety <eskultet@redhat.com> Reviewed-by: Peter Krempa <pkrempa@redhat.com>
Laine Stump [Wed, 21 Oct 2020 02:31:27 +0000 (22:31 -0400)]
util: fix very old bug/typo in virNetDevParseVfInfo()
When this function was recently changed to add in parsing of
IFLA_VF_STATS, I noticed that the checks for existence of IFLA_VF_MAC
and IFLA_VF_VLAN were looking in the *wrong array*. The array that
contains the results of parsing each IFLA_VFINFO in
tb[IFLA_VFINFO_LIST] is tb_vf, but we were checking for these in tb
(which is the array containing the results of the toplevel parsing of
the netlink message, *not* the results of parsing one of the nested
IFLA_VFINFO's.
This incorrect code has been here since the function was originally
written in 2012. It has only worked all these years due to coincidence
- the items at those indexes in tb are IFLA_ADDRESS and IFLA_BROADCAST
(of the *PF*, not of any of its VFs), and those happen to always be
present in the toplevel netlink message; since we are only looking in
the incorrect place to check for mere existence of the attribute (but
are doing the actual retrieval of the attribute from the correct
place), this bug has no real consequences other than confusing anyone
trying to understand the code.
Signed-off-by: Laine Stump <laine@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
A binutils change has caused breakage when linking the tests
/usr/bin/ld: tests/libtest_qemu_driver.so: undefined reference to `__open_missing_mode'
This is probably a regression in binutils, so disable LTO until we get
more clarity on the root cause and whether binutils or libvirt will need
changing.
Matt Coleman [Wed, 21 Oct 2020 08:46:09 +0000 (04:46 -0400)]
hyperv: implement domainReboot and domainReset
Co-authored-by: Sri Ramanujam <sramanujam@datto.com> Signed-off-by: Matt Coleman <matt@datto.com> Signed-off-by: Michal Privoznik <mprivozn@redhat.com> Reviewed-by: Michal Privoznik <mprivozn@redhat.com>
zhenwei pi [Thu, 15 Oct 2020 11:21:14 +0000 (19:21 +0800)]
util: support device stats collection for <interface type='hostdev'>
libvirt can retrieve traffic stats for emulated interfaces that are
backed by tap or macvtap devices, but this information wasn't
available for hostdev interfaces (those that are implemented by
assigning an SR-IOV VF device to a guest using vfio):
#virsh domifstat instance --interface=52:54:00:2d:b2:35
error: Failed to get interface stats instance 52:54:00:2d:b2:35
error: internal error: Interface name not provided
For some SR-IOV VF devices this information is available via the
netlink VFINFO_LIST request/response, and that is what this patch uses
to implement stats retrieval for VF. Not that this is dependent on
support in the PF driver - for example, the Mellanox ConnectX-4 Lx
(mlx5) driver reports usable stats, while Intel 82599 (ixgbe) and
82576 (igb) just report all stats as 0. (this is the same result as
"ip -s link show").
Signed-off-by: zhenwei pi <pizhenwei@bytedance.com> Reviewed-by: Laine Stump <laine@redhat.com>
Jonathon Jongsma [Wed, 14 Oct 2020 17:08:25 +0000 (12:08 -0500)]
conf: Add support for vDPA network devices
This patch adds new schema and adds support for parsing and formatting
domain configurations that include vdpa devices.
vDPA network devices allow high-performance networking in a virtual
machine by providing a wire-speed data path. These devices require a
vendor-specific host driver but the data path follows the virtio
specification.
When a device on the host is bound to an appropriate vendor-specific
driver, it will create a chardev on the host at e.g. /dev/vhost-vdpa-0.
That chardev path can then be used to define a new interface with
type='vdpa'.
Signed-off-by: Jonathon Jongsma <jjongsma@redhat.com> Reviewed-by: Laine Stump <laine@redhat.com>
projects / thirdparty / libvirt.git / log
