[libvirt] [PATCH 00/12] Split up qemu_command.c
by John Ferlan
Before disappearing too far down the rabbit hole...
My recent foray into qemu_command.c to split out qemu_parse_command.c
got me thinking that the module could use quite a bit of a reorganization.
It's about 11,400 lines of all sorts of various functions. This started out
as an attempt to order functions as they are used, but started turning into
extracting certain API's/Helpers into other qemu_*.c modules.
There's still quite a bit to do, but rather than hold onto it too long
I figured I'd send what I have so far now (and keep working through while
waiting for reviews).
This is essentially all code motion with some function renames along the way.
John Ferlan (12):
qemu-command: Move qemuBuildTPMBackendStr
qemu-command: Move qemuVirCommandGetFDSet
qemu-command: Move qemuBuildTPMDevStr
qemu-command: Move qemuVirCommandGetDevSet
qemu: Move qemuPhysIfaceConnect to qemu_interface.c and rename
qemu: Move qemuNetworkIfaceConnect to qemu_interface.c and rename
qemu-command: Move qemuDomainSupports* functions
qemu-command: Move qemuDomain*Address* functions to qemu_domain.c
qemu-command: Move remaining qemuDomain* functions
qemu-command: Move and rename qemuOpenVhostNet
qemu-command: Move qemuNetworkPrepareDevices
qemu-command: Move qemuAssign*Alias* API's into their own module
po/POTFILES.in | 2 +
src/Makefile.am | 1 +
src/qemu/qemu_assign_alias.c | 468 +++++++
src/qemu/qemu_assign_alias.h | 60 +
src/qemu/qemu_command.c | 3073 +++++-------------------------------------
src/qemu/qemu_command.h | 48 -
src/qemu/qemu_domain.c | 1646 ++++++++++++++++++++++
src/qemu/qemu_domain.h | 25 +
src/qemu/qemu_driver.c | 1 +
src/qemu/qemu_hotplug.c | 19 +-
src/qemu/qemu_interface.c | 378 ++++++
src/qemu/qemu_interface.h | 22 +
src/qemu/qemu_process.c | 56 +-
tests/qemuhotplugtest.c | 1 +
tests/qemuxml2argvtest.c | 1 +
15 files changed, 2978 insertions(+), 2823 deletions(-)
create mode 100644 src/qemu/qemu_assign_alias.c
create mode 100644 src/qemu/qemu_assign_alias.h
--
2.5.0
8 years, 9 months
[libvirt] [PATCH] conf: Use a temporary int variable to store GIC version
by Andrea Bolognani
Since no value in the virGICVersion enumeration is negative, a clever
enough compiler can report an error such as
src/conf/domain_conf.c:15337:75: error: comparison of unsigned enum
expression < 0 is always false [-Werror,-Wtautological-compare]
if ((def->gic_version = virGICVersionTypeFromString(tmp)) < 0 ||
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ^ ~
virGICVersionTypeFromString() can, however, return a negative value if
the input string is not part of the enumeration, so we definitely need
that check.
Work around the problem by storing the return value in a temporary int
variable.
---
Pushed under the build-breaker rule.
src/conf/domain_conf.c | 7 ++++---
1 file changed, 4 insertions(+), 3 deletions(-)
diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c
index 295bc1b..0d2957b 100644
--- a/src/conf/domain_conf.c
+++ b/src/conf/domain_conf.c
@@ -14728,7 +14728,7 @@ virDomainDefParseXML(xmlDocPtr xml,
xmlNodePtr *nodes = NULL, node = NULL;
char *tmp = NULL;
size_t i, j;
- int n, virtType;
+ int n, virtType, gic_version;
long id = -1;
virDomainDefPtr def;
bool uuid_generated = false;
@@ -15334,12 +15334,13 @@ virDomainDefParseXML(xmlDocPtr xml,
node = ctxt->node;
ctxt->node = nodes[i];
if ((tmp = virXPathString("string(./@version)", ctxt))) {
- if ((def->gic_version = virGICVersionTypeFromString(tmp)) < 0 ||
- def->gic_version == VIR_GIC_VERSION_NONE) {
+ gic_version = virGICVersionTypeFromString(tmp);
+ if (gic_version < 0 || gic_version == VIR_GIC_VERSION_NONE) {
virReportError(VIR_ERR_XML_ERROR,
_("malformed gic version: %s"), tmp);
goto error;
}
+ def->gic_version = gic_version;
VIR_FREE(tmp);
}
def->features[val] = VIR_TRISTATE_SWITCH_ON;
--
2.5.0
8 years, 9 months
[libvirt] [PATCH 0/7] Improve GIC support
by Andrea Bolognani
This series was really just supposed to enable guests to use the
special "host" GIC version, but I ended up changing a bunch of other
stuff and adding a whole lot of new test cases.
I've also made it so the GIC availability is always reflected in the
domain XML, the same way other implicit devices and features work.
The GIC-related definitions are in their own file: depending on
whether we end up probing for host GIC support ourselves or relying
on QEMU this might turn out to be a huge overkill :)
Cheers.
Andrea Bolognani (7):
gic: Introduce virGICVersion enumeration
schema: List allowed GIC versions
conf: Use virGICVersion enumeration in virDomainDef
qemu: Default to GIC v2
qemu: Always enable GIC on ARM virt machines
tests: Reorganize and simplify GIC test cases
tests: Add more GIC test cases
docs/schemas/domaincommon.rng | 6 +++-
src/Makefile.am | 1 +
src/conf/domain_conf.c | 15 ++++----
src/conf/domain_conf.h | 3 +-
src/libvirt_private.syms | 5 +++
src/qemu/qemu_command.c | 8 +++--
src/qemu/qemu_domain.c | 29 ++++++++++++++++
src/util/virgic.c | 33 ++++++++++++++++++
src/util/virgic.h | 38 ++++++++++++++++++++
.../qemuxml2argv-aarch64-aavmf-virtio-mmio.xml | 1 +
.../qemuxml2argv-aarch64-gic-default.args | 1 +
.../qemuxml2argv-aarch64-gic-default.xml | 22 ++++++++++++
...gic.args => qemuxml2argv-aarch64-gic-host.args} | 13 +++----
.../qemuxml2argv-aarch64-gic-host.xml | 22 ++++++++++++
.../qemuxml2argv-aarch64-gic-invalid.xml | 22 ++++++++++++
.../qemuxml2argv-aarch64-gic-none.args | 1 +
.../qemuxml2argv-aarch64-gic-none.xml | 19 ++++++++++
.../qemuxml2argv-aarch64-gic-not-arm.xml | 22 ++++++++++++
.../qemuxml2argv-aarch64-gic-not-virt.xml | 22 ++++++++++++
...gicv3.args => qemuxml2argv-aarch64-gic-v2.args} | 12 +++----
...h64-gic.xml => qemuxml2argv-aarch64-gic-v2.xml} | 14 ++------
.../qemuxml2argv-aarch64-gic-v3.args | 20 +++++++++++
...4-gicv3.xml => qemuxml2argv-aarch64-gic-v3.xml} | 14 ++------
tests/qemuxml2argvtest.c | 40 ++++++++++++++++++----
.../qemuxml2xmlout-aarch64-gic-default.xml | 1 +
.../qemuxml2xmlout-aarch64-gic-none.xml | 1 +
tests/qemuxml2xmltest.c | 7 ++--
27 files changed, 331 insertions(+), 61 deletions(-)
create mode 100644 src/util/virgic.c
create mode 100644 src/util/virgic.h
create mode 120000 tests/qemuxml2argvdata/qemuxml2argv-aarch64-gic-default.args
create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-aarch64-gic-default.xml
rename tests/qemuxml2argvdata/{qemuxml2argv-aarch64-gic.args => qemuxml2argv-aarch64-gic-host.args} (57%)
create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-aarch64-gic-host.xml
create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-aarch64-gic-invalid.xml
create mode 120000 tests/qemuxml2argvdata/qemuxml2argv-aarch64-gic-none.args
create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-aarch64-gic-none.xml
create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-aarch64-gic-not-arm.xml
create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-aarch64-gic-not-virt.xml
rename tests/qemuxml2argvdata/{qemuxml2argv-aarch64-gicv3.args => qemuxml2argv-aarch64-gic-v2.args} (55%)
rename tests/qemuxml2argvdata/{qemuxml2argv-aarch64-gic.xml => qemuxml2argv-aarch64-gic-v2.xml} (61%)
create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-aarch64-gic-v3.args
rename tests/qemuxml2argvdata/{qemuxml2argv-aarch64-gicv3.xml => qemuxml2argv-aarch64-gic-v3.xml} (61%)
create mode 120000 tests/qemuxml2xmloutdata/qemuxml2xmlout-aarch64-gic-default.xml
create mode 120000 tests/qemuxml2xmloutdata/qemuxml2xmlout-aarch64-gic-none.xml
--
2.5.0
8 years, 9 months
[libvirt] [PATCH] qemu: fix hot unplug of PCI devices with VFIO
by Ludovic Beliveau
Currently, on hot unplug of PCI devices with VFIO driver for QEMU, libvirt is
trying to restore the host devices to it's previous value (basically a chown
on the previous user/group).
However for devices with VFIO driver, when the device is unbinded it is
removed from the /dev/vfio file system causing the restore label to fail.
The fix is to not restore the label for those PCI devices since they are going
to be teared down anyway.
Signed-off-by: Ludovic Beliveau <ludovic.beliveau(a)windriver.com>
---
diff --git a/src/qemu/qemu_hotplug.c b/src/qemu/qemu_hotplug.c
index f8db960..f5beabd 100644
--- a/src/qemu/qemu_hotplug.c
+++ b/src/qemu/qemu_hotplug.c
@@ -2996,6 +2996,8 @@ qemuDomainRemoveHostDevice(virQEMUDriverPtr driver,
int ret = -1;
qemuDomainObjPrivatePtr priv = vm->privateData;
char *drivestr = NULL;
+ virDomainHostdevSubsysPCIPtr pcisrc = NULL;
+ bool is_vfio = false;
VIR_DEBUG("Removing host device %s from domain %p %s",
hostdev->info->alias, vm, vm->def->name);
@@ -3039,6 +3041,8 @@ qemuDomainRemoveHostDevice(virQEMUDriverPtr driver,
switch ((virDomainHostdevSubsysType) hostdev->source.subsys.type) {
case VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_PCI:
+ pcisrc = &hostdev->source.subsys.u.pci;
+ is_vfio = pcisrc->backend == VIR_DOMAIN_HOSTDEV_PCI_BACKEND_VFIO;
qemuDomainRemovePCIHostDevice(driver, vm, hostdev);
/* QEMU might no longer need to lock as much memory, eg. we just
* detached the last VFIO device, so adjust the limit here */
@@ -3058,7 +3062,8 @@ qemuDomainRemoveHostDevice(virQEMUDriverPtr driver,
if (qemuTeardownHostdevCgroup(vm, hostdev) < 0)
VIR_WARN("Failed to remove host device cgroup ACL");
- if (virSecurityManagerRestoreHostdevLabel(driver->securityManager,
+ if (!is_vfio &&
+ virSecurityManagerRestoreHostdevLabel(driver->securityManager,
vm->def, hostdev, NULL) < 0) {
VIR_WARN("Failed to restore host device labelling");
}
8 years, 9 months
[libvirt] [PATCH] Spell VMware with a lowercase w
by Ján Tomko
Replace all occurrences of VMWare outside the news.
---
docs/formatdomain.html.in | 2 +-
docs/migration.html.in | 2 +-
docs/storage.html.in | 2 +-
libvirt.spec.in | 2 +-
src/util/virerror.c | 2 +-
src/util/virstoragefile.c | 2 +-
tools/virsh.c | 2 +-
7 files changed, 7 insertions(+), 7 deletions(-)
diff --git a/docs/formatdomain.html.in b/docs/formatdomain.html.in
index b3187bb..e96798f 100644
--- a/docs/formatdomain.html.in
+++ b/docs/formatdomain.html.in
@@ -1569,7 +1569,7 @@
<dt><code>vmport</code></dt>
<dd>Depending on the <code>state</code> attribute (values <code>on</code>,
<code>off</code>, default <code>on</code>) enable or disable
- the emulation of VMWare IO port, for vmmouse etc.
+ the emulation of VMware IO port, for vmmouse etc.
<span class="since">Since 1.2.16</span>
</dd>
<dt><code>gic</code></dt>
diff --git a/docs/migration.html.in b/docs/migration.html.in
index 6986cc0..8efdd7f 100644
--- a/docs/migration.html.in
+++ b/docs/migration.html.in
@@ -475,7 +475,7 @@
</pre>
<p>
- Supported by Xen, QEMU, VMWare and VirtualBox drivers
+ Supported by Xen, QEMU, VMware and VirtualBox drivers
</p>
<h3><a name="scenarionativepeer2peer">Native migration, client to and peer2peer between, two libvirtd servers</a></h3>
diff --git a/docs/storage.html.in b/docs/storage.html.in
index 1f01330..b0efe69 100644
--- a/docs/storage.html.in
+++ b/docs/storage.html.in
@@ -161,7 +161,7 @@
<li><code>qcow</code>: QEMU v1 disk image format</li>
<li><code>qcow2</code>: QEMU v2 disk image format</li>
<li><code>qed</code>: QEMU Enhanced Disk image format</li>
- <li><code>vmdk</code>: VMWare disk image format</li>
+ <li><code>vmdk</code>: VMware disk image format</li>
<li><code>vpc</code>: VirtualPC disk image format</li>
</ul>
<p>
diff --git a/libvirt.spec.in b/libvirt.spec.in
index dd54dfc..fc2e2cf 100644
--- a/libvirt.spec.in
+++ b/libvirt.spec.in
@@ -190,7 +190,7 @@
%endif
# RHEL doesn't ship OpenVZ, VBox, UML, PowerHypervisor,
-# VMWare, libxenserver (xenapi), libxenlight (Xen 4.1 and newer),
+# VMware, libxenserver (xenapi), libxenlight (Xen 4.1 and newer),
# or HyperV.
%if 0%{?rhel}
%define with_openvz 0
diff --git a/src/util/virerror.c b/src/util/virerror.c
index 3a3ddef..e1bcf52 100644
--- a/src/util/virerror.c
+++ b/src/util/virerror.c
@@ -105,7 +105,7 @@ VIR_ENUM_IMPL(virErrorDomain, VIR_ERR_DOMAIN_LAST,
"Audit Utils",
"Sysinfo Utils",
"I/O Stream Utils",
- "VMWare Driver",
+ "VMware Driver",
"Event Loop", /* 40 */
"Xen Light Driver",
diff --git a/src/util/virstoragefile.c b/src/util/virstoragefile.c
index 3e79ae1..49b1745 100644
--- a/src/util/virstoragefile.c
+++ b/src/util/virstoragefile.c
@@ -490,7 +490,7 @@ vmdk4GetBackingStore(char **res,
*res = NULL;
/*
* Technically this should have been VMDK, since
- * VMDK spec / VMWare impl only support VMDK backed
+ * VMDK spec / VMware impl only support VMDK backed
* by VMDK. QEMU isn't following this though and
* does probing on VMDK backing files, hence we set
* AUTO
diff --git a/tools/virsh.c b/tools/virsh.c
index b96dbda..eb84dd0 100644
--- a/tools/virsh.c
+++ b/tools/virsh.c
@@ -550,7 +550,7 @@ virshShowVersion(vshControl *ctl ATTRIBUTE_UNUSED)
vshPrint(ctl, " OpenVZ");
#endif
#ifdef WITH_VMWARE
- vshPrint(ctl, " VMWare");
+ vshPrint(ctl, " VMware");
#endif
#ifdef WITH_PHYP
vshPrint(ctl, " PHYP");
--
2.4.10
8 years, 9 months
[libvirt] PING: [RFC] vhost-user + shared memory + NUMA
by Pavel Fedin
Hello! Sorry, but i did not get any answer to the last question. Would it be OK to require <memoryBacking> and implicitly add only
shared mode ?
Kind regards,
Pavel Fedin
Senior Engineer
Samsung Electronics Research center Russia
> -----Original Message-----
> From: libvir-list-bounces(a)redhat.com [mailto:libvir-list-bounces@redhat.com] On Behalf Of
> Pavel Fedin
> Sent: Thursday, February 11, 2016 5:36 PM
> To: 'Daniel P. Berrange'
> Cc: 'Libvirt'
> Subject: Re: [libvirt] [RFC] vhost-user + shared memory + NUMA
>
> Hello!
>
> > > Ok, then would it be a good compromise if we require <memoryBacking>, and only
> implicitly
> > add "shared" if we have vhost-user
> > > devices? This way we would not change the way the guest memory is allocated.
> >
> > Adding shared implicitly *will* change the way guest memory is allocated,
> > as it will have to use tmpfs to make it shared.
>
> You perhaps didn't get my idea. I meant - we will still need to specify <memoryBacking> with
> huge pages, just no <numa>. Therefore, the memory will be allocated via file backend from
> hugetlbfs. Only mode will be changed implicitly (private -> shared).
>
> > > IMHO being able to manually specify "shared" both in <numa> and
> > > in <memoryBacking> would be ambiguous.
> >
> > That's not really any different to what we have already with NUMA.
> > The top level setting would apply as the default, and the NUMA level
> > settings override it if needed.
>
> Well, the only little drawback would be necessity to add "shared" by itself. This would
> require additional patching to clients (e. g. openstack).
>
> Kind regards,
> Pavel Fedin
> Senior Engineer
> Samsung Electronics Research center Russia
>
>
> --
> libvir-list mailing list
> libvir-list(a)redhat.com
> https://www.redhat.com/mailman/listinfo/libvir-list
8 years, 9 months
[libvirt] [PATCH 0/2] virsh: fix usage of vshPrint
by Peter Krempa
Peter Krempa (2):
vsh: Simplify bailing out on OOM conditions
vsh: Replace vshPrint macro with function
tools/virsh-domain.c | 16 +++++++++-------
tools/vsh.c | 34 +++++++++++++++++++++++++++++-----
tools/vsh.h | 5 ++---
3 files changed, 40 insertions(+), 15 deletions(-)
--
2.6.2
8 years, 9 months
[libvirt] [PATCH 0/4] Couple of storage related fixes
by Michal Privoznik
After an e-mail on libvirt-users [1] I've realized there are some things that
we can do better. The first two patches introduce a new feature. The other two
are just some fixes.
Michal Privoznik (4):
cmdVolCreateAs: Rework to follow usual func pattern
virsh: Teach vol-create-as to --print-xml
storageVolCreateXML: Swap order of two operations
storageVolCreateXMLFrom: Check if backend knows how to createVol
src/storage/storage_driver.c | 15 +++++++++++----
tools/virsh-volume.c | 33 +++++++++++++++++++++------------
tools/virsh.pod | 6 ++++--
3 files changed, 36 insertions(+), 18 deletions(-)
--
2.4.10
8 years, 9 months
[libvirt] ARM KVM GICv3 Support
by Christoffer Dall
Hi all,
I'm trying to figure out what the correct solution for libvirt support
for ARM KVM guests with GICv3 is.
The challenge is that QEMU's virt machine model defaults to GICv2 unless
you specify an additional "-machine gic-version=host" parameter to QEMU,
in which case your VM gets the same GIC version as your host.
Now, I'm told that this is inconvenient for libvirt, because it
typically does not pass any -machine options, is that correct?
Further, there are two additional complications: First, some guest
images may not include GICv3 support. For Linux, this is relatively old
kernels (prior to v3.17), but other guests may only support GICv2.
Second, some hardware platforms only support GICv2 guests, some only
support GICv3 guests, and some support both GICv2 and GICv3 guests
(those hosts with a GICv3 with the backwards compatibility support).
It is unclear to me how libvirt users will relate to these constraints.
For example, in an OpenStack deployment, will it somehow be easy to tell
which nodes support which GIC version such that administrators can
choose compatible VM images, or how does that work?
The end goal here is to determine if we need to add any functionality to
QEMU to better support libvirt, and what kind of features/glue need to
be added to libvirt for this to work.
Much thanks for any feedback!
-Christoffer
8 years, 9 months
[libvirt] [libvirt-php] Status of Development?
by Markus Petzsch
Hello,
I've noticed that the PHP libvirt binding is missing some functions
which are present in other implementations. Namely things like
attach-interface which allow live modification of a running VM. How is
the current status of the PHP binding? Will we see new features in the
future or is learning Python the better option? :-)
Markus
8 years, 9 months
