- Devel - libvirt List Archives

[libvirt] [PATCH] macvtap: libvirtd forgot macvtap device name during a shutdown/restart cycle
by Stefan Berger 03 Nov '10

03 Nov '10

During a shutdown/restart cycle libvirtd forgot the macvtap device name that it had created on behalf of a VM so that a stale macvtap device remained on the host when the VM terminated. Libvirtd has to actively tear down a macvtap device and it uses its name for identifying which device to tear down. The solution is to not blank out the <target dev='...'/> completely, but only blank it out on VMs that are not active. So, if a VM is active, the device name makes it into the XML and is also being parsed. If a VM is not active, the device name is discarded. Signed-off-by: Stefan Berger <stefanb(a)us.ibm.com> --- src/conf/domain_conf.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) Index: libvirt-acl/src/conf/domain_conf.c =================================================================== --- libvirt-acl.orig/src/conf/domain_conf.c +++ libvirt-acl/src/conf/domain_conf.c @@ -2343,7 +2343,8 @@ virDomainNetDefParseXML(virCapsPtr caps, def->data.direct.linkdev = dev; dev = NULL; - VIR_FREE(ifname); + if ((flags & VIR_DOMAIN_XML_INACTIVE)) + VIR_FREE(ifname); break; @@ -5801,6 +5802,8 @@ virDomainNetDefFormat(virBufferPtr buf, virBufferAddLit(buf, "/>\n"); virVirtualPortProfileFormat(buf, &def->data.direct.virtPortProfile, " "); + if ((flags & VIR_DOMAIN_XML_INACTIVE)) + VIR_FREE(def->ifname); break; case VIR_DOMAIN_NET_TYPE_USER:

2 6

[libvirt] [PATCH 1/7] Sysinfo extension to relax-ng schemas
by Daniel Veillard 03 Nov '10

03 Nov '10

Currently includes a subset of the SMBIOS strings set Signed-off-by: Daniel Veillard <veillard(a)redhat.com> --- docs/schemas/domain.rng | 75 +++++++++++++++++++++++++++++++++++++++++++++++ 1 files changed, 75 insertions(+), 0 deletions(-)

2 2

[libvirt] Revamped Smbios/Sysinfo support
by Daniel Veillard 03 Nov '10

03 Nov '10

This patch set implements a new way to store system informations and pass them to the domain, it's modelled on the SMBIOS support in hypervisors and allow to override those informations in domains: The simplest is the use of: <smbios mode="host"/> in which case the domain will try to inherit those values from the Host own values, the classical use case is the use of OEM'ed Windows versions where one need to export SMBIOS informations from the host to the guest. But the <sysinfo> construct allows to define those informations, for example: <sysinfo type="smbios"> <bios> <entry name="vendor">QEmu/KVM</entry> <entry name="version">0.13</entry> </bios> <system> <entry name="manufacturer">Fedora</entry> <entry name="product">Virt-Manager</entry> <entry name="version">0.8.2-3.fc14</entry> <entry name="serial">32dfcb37-5af1-552b-357c-be8c3aa38310</entry> <entry name="uuid">c7a5fdbd-edaf-9455-926a-d65c16db1809</entry> </system> </sysinfo> <smbios mode="sysinfo"/> In that case libvirt hypervisor driver will try to set up those values in the domain emulation. The patch set attached implement those modes for QEmu, and based on our previous reviews with Matthias the first mode should be implementable in the ESX driver (for just a couple of settings though) and the VirtualBox driver should be abble to implement both if we understood correctly, Daniel

2 3

[libvirt] [PATCH 7/7] Add a new test for sysinfo and smbios handling
by Daniel Veillard 03 Nov '10

03 Nov '10

Signed-off-by: Daniel Veillard <veillard(a)redhat.com> --- tests/qemuxml2argvdata/qemuxml2argv-smbios.args | 1 + tests/qemuxml2argvdata/qemuxml2argv-smbios.xml | 39 +++++++++++++++++++++++ tests/qemuxml2argvtest.c | 2 + 3 files changed, 42 insertions(+), 0 deletions(-) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-smbios.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-smbios.xml

2 1

[libvirt] [PATCH 5/7] Read the host sysinfo for the QEmu driver
by Daniel Veillard 03 Nov '10

03 Nov '10

Read and store the data when initializing the driver. Signed-off-by: Daniel Veillard <veillard(a)redhat.com> --- src/qemu/qemu_conf.h | 2 ++ src/qemu/qemu_driver.c | 7 +++++++ 2 files changed, 9 insertions(+), 0 deletions(-)

2 1

[libvirt] [PATCH 3/7] Add smbios element to schema and configuration
by Daniel Veillard 03 Nov '10

03 Nov '10

the element has a mode attribute allowing only 3 values: - emulate: use the smbios emulation from the hypervisor - host: try to use the smbios values from the node - sysinfo: grab the values from the <sysinfo> fields * docs/schemas/domain.rng: extend the schemas * src/conf/domain_conf.h: add the flag to the domain config * src/conf/domain_conf.h: parse and serialize the smbios if present Signed-off-by: Daniel Veillard <veillard(a)redhat.com> --- docs/schemas/domain.rng | 16 ++++++++++++++++ src/conf/domain_conf.c | 17 +++++++++++++++++ src/conf/domain_conf.h | 11 +++++++++++ 3 files changed, 44 insertions(+), 0 deletions(-)

2 1

[libvirt] Allow nbd-backed disk images?
by Adam Litke 03 Nov '10

03 Nov '10

I am trying to use a qcow image with libvirt where the backing 'file' is a qemu-nbd server. Unfortunately the security code assumes that backingStore is always a real file so something like 'nbd:0:3333' is rejected because a file with that name cannot be accessed. A simple patch like the following works around the problem -- are there any suggestions on how to do this properly? Note that I am not worried about directly using nbd images. That would require a new disk type with XML markup, etc. I only want it to be permitted as a backingStore. diff --git a/src/util/storage_file.c b/src/util/storage_file.c index 6f48b10..7c0ea9a 100644 --- a/src/util/storage_file.c +++ b/src/util/storage_file.c @@ -292,6 +292,13 @@ qcowXGetBackingStore(char **res, return BACKING_STORE_INVALID; if (size + 1 == 0) return BACKING_STORE_INVALID; + + /* Short-circuit nbd backing files */ + if (size >= 4 && STRPREFIX((const char *)(buf + offset), "nbd:")) { + return BACKING_STORE_OK; + } + if (VIR_ALLOC_N(*res, size + 1) < 0) { virReportOOMError(); return BACKING_STORE_ERROR; -- Thanks, Adam

2 1

[libvirt] [PATCH v2] macvtap: libvirtd forgot macvtap device name during a shutdown/restart cycle
by Stefan Berger 03 Nov '10

03 Nov '10

V2: - removed change from virDomainNetDefFormat. The reported problem also does not occur. During a shutdown/restart cycle libvirtd forgot the macvtap device name that it had created on behalf of a VM so that a stale macvtap device remained on the host when the VM terminated. Libvirtd has to actively tear down a macvtap device and it uses its name for identifying which device to tear down. The solution is to not blank out the <target dev='...'/> completely, but only blank it out on VMs that are not active. So, if a VM is active, the device name makes it into the XML and is also being parsed. If a VM is not active, the device name is discarded. Signed-off-by: Stefan Berger <stefanb(a)us.ibm.com> --- src/conf/domain_conf.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) Index: libvirt-acl/src/conf/domain_conf.c =================================================================== --- libvirt-acl.orig/src/conf/domain_conf.c +++ libvirt-acl/src/conf/domain_conf.c @@ -2343,7 +2343,8 @@ virDomainNetDefParseXML(virCapsPtr caps, def->data.direct.linkdev = dev; dev = NULL; - VIR_FREE(ifname); + if ((flags & VIR_DOMAIN_XML_INACTIVE)) + VIR_FREE(ifname); break;

2 2

[libvirt] [PATCH] tests: fix daemon-conf testing failure
by Osier Yang 03 Nov '10

03 Nov '10

libvirtd.conf uses "libvirt" as the value of "unix_sock_group", however, group "libvirt" may not exist on system, in this case the case will always be failed, which will cause 'make check' , and 'make rpm' always be failed further more. As a solution, replace "libvirt" with "root" in "tmp.conf". * tests/daemon-conf --- tests/daemon-conf | 3 +++ 1 files changed, 3 insertions(+), 0 deletions(-) diff --git a/tests/daemon-conf b/tests/daemon-conf index 6c91d96..225f84c 100755 --- a/tests/daemon-conf +++ b/tests/daemon-conf @@ -25,6 +25,9 @@ grep -v '\"PARAMETER = VALUE\"' "$conf" | grep '[a-z_] *= *[^ ]' | grep -vE '^ # Start with the sample libvirtd.conf file, uncommenting all real directives. sed -n 's/^#$[^ #]$/\1/p' "$conf" > tmp.conf +sed -e 's/^$unix_sock_group =$.*/\1 "root"/g' tmp.conf > k +mv k tmp.conf + # Iterate through that list of directives, corrupting one RHS at a # time and running libvirtd with the resulting config. Each libvirtd # invocation must fail. -- 1.7.2.3

4 4

[libvirt] PCI/USB devices documentation
by Zdenek Styblik 03 Nov '10

03 Nov '10

Hello, so far unsuccessfully, I'm trying to attach PCI/USB devices to Guests (via virt-manager, soon via virsh because - whatever :]). Anyway, there is <hostdev managed='yes'> where I'm wondering what 'managed' means, because it's not documented here[1]. Is it a bug or is it documented elsewhere? btw also, 'Node devices XML format'[2] page is empty Thanks, Zdenek References: --- [1] http://libvirt.org/formatdomain.html#elementsUSB [2] http://libvirt.org/formatnode.html PS: I haven't forgotten about Slackware SSH how-to, but- *sigh* I promise! -- Zdenek Styblik Net/Linux admin OS TurnovFree.net email: stybla(a)turnovfree.net jabber: stybla(a)jabber.turnovfree.net

3 4