May 2016 - Devel - libvirt List Archives

[libvirt] Need to re-work final "peer address" patches and re-push them
by Laine Stump 16 May '16

16 May '16

I reverted these three patches that introduced and enabled a "peer" attribute for type='ethernet' interface <ip> elements prior to the release of 1.3.4 with the intent of fixing/re-posting them after release, but forgot until today: https://www.redhat.com/archives/libvir-list/2016-April/msg01995.html I have patches for most of the bugs, but the one problem that still doesn't have resolution is the naming of the "peer" attribute. In my opinion, having the two address attributes named "address" and "peer" makes it ambiguous which address is for the guest side and which for the host side (especially since the attribute that has been named "peer" would be set to the "address" in the netlink command, and the attribute named "address" would be set to "peer" in the netlink command :-O). Since "address" is an existing attribute, and already used for the guest side IP address in lxc type='bridge' interfaces, it must remain as-is. In order to make it obvious that the new address is for the host side of the tap (or veth pair in the case of lxc), I propose calling it either "host", or "hostAddress", e.g: <ip address='192.168.123.43' host='192.168.123.1' prefix='25'/> or <ip address='192.168.123.4' hostAddress='192.168.123.1' prefix='25'/> (Vasiliy had suggested "hostPeer", but I dislike that, since it sounds like "the peer of the host", which is even more misleading). Can some of you normally-opinionated people weigh in on this? I don't like the feeling of making a unilateral decision :-) Also, I'm realizing that, although there was a patch to support setting the host-side address (hmm - "hostSide"? nah) for lxc type='bridge' interface, this is not at all useful, because anything plugged into a bridge should not have any IP on the side plugged into the bridge. The place where it would be useful for lxc would be (just as it is for qemu) with a type='ethernet' interface - the guest-side veth would have "address" and the host-side veth would have "hostAddress", and it would then properly work without needing a bridge (which I think is the entire point). Since lxc doesn't currently support type='ethernet', I think that initial support should be made for qemu only, and when type='ethernet is added to lxc, it can be made to support an IP address on both sides of the veth pair from the start. Lacking any useful responses, I'm thinking to update Vasiliy's patches to use "hostAddress" (and fix the other bugs I had found) and re-post them.

5 10

[libvirt] [PATCH 0/7] PCI Multifunction device hotplug support
by Shivaprasad G Bhat 16 May '16

16 May '16

The series implements the PCI multifunction hotplug functionality. The first two patch reverts the old commits which forbade the multifunction addresses. Patch2 is added because the two reverts necessiates additional checks added in Patch3. Actual implementation starts patch 4 onwards. The semantics is to encapsulate all the functions in <devices/> tag as was previously explained. (http://www.spinics.net/linux/fedora/libvir/msg127921.html The series aims to support virtio devices also the hostdevices in the <devices>. The auto addess assignment is implemented here works well for ppc64 and pc-i440fx machines. Couldn't get it working on q35 machines. May be follow up patches for that. Todo: 1) Rollback the Hotplug of already hoplugged functions if any of the function hotplug fails. 2) Hardening the hotplug checks to disallow multifunction cards hotplug as though they are single function cards. 3) Documentation update. 4) Patch 6 may need some work across drivers to add similar checks as added in qemu_driver.c 5) Test cases. --- Shivaprasad G Bhat (7): Revert "prevent hot unplugging multi function PCI device" Release address in function granularity than slot Validate address in virDomainPCIAddressReleaseAddr Introduce PCI Multifunction device parser Introduce virDomainPCIMultifunctionDeviceAddressAssign Allow calling virHostdevPreparePCIDevices twice for same device Enable PCI Multifunction hotplug/unplug src/conf/domain_addr.c | 244 ++++++++++++++++++++++++++++- src/conf/domain_addr.h | 4 src/conf/domain_conf.c | 227 +++++++++++++++++++++++++++ src/conf/domain_conf.h | 22 +++ src/libvirt_private.syms | 5 + src/qemu/qemu_domain_address.c | 2 src/qemu/qemu_driver.c | 331 ++++++++++++++++++++++++++++++++-------- src/qemu/qemu_hotplug.c | 74 ++------- src/util/virhostdev.c | 16 ++ 9 files changed, 781 insertions(+), 144 deletions(-) -- Signature

1 7

[libvirt] [PATCH 0/7] PCI Multifunction device hotplug support
by Shivaprasad G Bhat 16 May '16

16 May '16

The series implements the PCI multifunction hotplug functionality. The first two patch reverts the old commits which forbade the multifunction addresses. Patch2 is added because the two reverts necessiates additional checks added in Patch3. Actual implementation starts patch 4 onwards. The semantics is to encapsulate all the functions in <devices/> tag as was previously explained. (http://www.spinics.net/linux/fedora/libvir/msg127921.html) The series aims to support virtio devices and also the hostdevices in the <devices>. Auto-address assignment is implemented here works well for ppc64 and pc-i440fx machines. Couldn't get it working on q35 machines. May be follow up patches for that. TODO: 1) Rollback the Hotplug of already hoplugged functions if any of the function hotplug fails. 2) Hardening the hotplug checks to disallow multifunction cards hotplug as though they are single function cards. 3) Documentation update. 4) Patch 6 may need some work across drivers to add similar checks as added in qemu_driver.c 5) Test cases. --- Shivaprasad G Bhat (7): Revert "prevent hot unplugging multi function PCI device" Release address in function granularity than slot Validate address in virDomainPCIAddressReleaseAddr Introduce PCI Multifunction device parser Introduce virDomainPCIMultifunctionDeviceAddressAssign Allow calling virHostdevPreparePCIDevices twice for same device Enable PCI Multifunction hotplug/unplug src/conf/domain_addr.c | 244 ++++++++++++++++++++++++++++- src/conf/domain_addr.h | 4 src/conf/domain_conf.c | 227 +++++++++++++++++++++++++++ src/conf/domain_conf.h | 22 +++ src/libvirt_private.syms | 5 + src/qemu/qemu_domain_address.c | 2 src/qemu/qemu_driver.c | 331 ++++++++++++++++++++++++++++++++-------- src/qemu/qemu_hotplug.c | 74 ++------- src/util/virhostdev.c | 16 ++ 9 files changed, 781 insertions(+), 144 deletions(-) -- Signature

1 7

[libvirt] [PATCH] bhyve: implement virConnectIsAlive
by Fabian Freyer 16 May '16

16 May '16

bhyve connections are local, and a "connection will be classed as alive if it is [...] local". --- src/bhyve/bhyve_driver.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/src/bhyve/bhyve_driver.c b/src/bhyve/bhyve_driver.c index c3c572e..8b41f7a 100644 --- a/src/bhyve/bhyve_driver.c +++ b/src/bhyve/bhyve_driver.c @@ -1517,6 +1517,11 @@ bhyveConnectGetType(virConnectPtr conn) return "BHYVE"; } +static int bhyveConnectIsAlive(virConnectPtr conn ATTRIBUTE_UNUSED) +{ + return 1; +} + static virHypervisorDriver bhyveHypervisorDriver = { .name = "bhyve", .connectOpen = bhyveConnectOpen, /* 1.2.2 */ @@ -1567,6 +1572,7 @@ static virHypervisorDriver bhyveHypervisorDriver = { .connectDomainEventDeregisterAny = bhyveConnectDomainEventDeregisterAny, /* 1.2.5 */ .domainHasManagedSaveImage = bhyveDomainHasManagedSaveImage, /* 1.2.13 */ .connectGetType = bhyveConnectGetType, /* 1.3.5 */ + .connectIsAlive = bhyveConnectIsAlive, /* 1.3.5 */ }; -- 2.1.4

2 1

[libvirt] [PATCH] qemuxml2argvtest: uncoditionally mock NUMA routines
by Roman Bogorodskiy 16 May '16

16 May '16

Currently, qemuxml2argvmock.c only mocks virNumaNodeIsAvailable(), and only if libvirt was built with NUMA support. This causes some test failures where NUMA is involved. For example, memory-hotplug-dimm fails because it goes all they way down to qemuBuildMemoryBackendStr() that calls virNumaNodesetIsAvailable() for a user specified nodeset and fails, hence the test (unexpectedly) fails. To make qemuxml2argtest successfully run on non-NUMA platforms, do the following: - mock virNumaNodeIsAvailable() unconditionally - add a mock for virNumaNodesetIsAvailable() --- tests/qemuxml2argvmock.c | 23 +++++++++++++++++------ 1 file changed, 17 insertions(+), 6 deletions(-) diff --git a/tests/qemuxml2argvmock.c b/tests/qemuxml2argvmock.c index 1616eed..57e56ab 100644 --- a/tests/qemuxml2argvmock.c +++ b/tests/qemuxml2argvmock.c @@ -56,17 +56,28 @@ virNumaGetMaxNode(void) return maxnodesNum; } -#if WITH_NUMACTL && HAVE_NUMA_BITMASK_ISBITSET -/* - * In case libvirt is compiled with full NUMA support, we need to mock - * this function in order to fake what numa nodes are available. - */ bool virNumaNodeIsAvailable(int node) { return node >= 0 && node <= virNumaGetMaxNode(); } -#endif /* WITH_NUMACTL && HAVE_NUMA_BITMASK_ISBITSET */ + +bool +virNumaNodesetIsAvailable(virBitmapPtr nodeset ATTRIBUTE_UNUSED) +{ + ssize_t bit = -1; + + if (!nodeset) + return true; + + while ((bit = virBitmapNextSetBit(nodeset, bit)) >= 0) { + if (virNumaNodeIsAvailable(bit)) + continue; + + return false; + } + return true; +} char * virTPMCreateCancelPath(const char *devpath) -- 2.7.4

3 5

[libvirt] [PATCH 0/2] Address some recent coverity
by John Ferlan 16 May '16

16 May '16

The first patch is a noise cleaner, while the second one fixes a real leak. Also anyone with desire to 'debug' a build environment - seems when the STATIC_ANALYSIS is true, the building of the docs (apibuild.py[.stamp]) fails in a rather spectacular mess. I can avoid building docs, but it's not clear to me what the failure is... John Ferlan (2): qemu: More qemu_monitor_json cleanups libxl: Resolve Coverity RESOURCE_LEAK src/libxl/libxl_migration.c | 5 ++++- src/qemu/qemu_monitor_json.c | 46 ++++++++++++++++++++++++-------------------- 2 files changed, 29 insertions(+), 22 deletions(-) -- 2.5.5

2 4

[libvirt] [PATCH v2 0/4] domain: Support <address type='pci'/> allocation
by Cole Robinson 16 May '16

16 May '16

This patch series allows the user to specify bare device <address type='pci'/> to explicitly request PCI address allocation. It's only wired up for qemu, and explicitly errors for all other drivers. This has several uses, but the motivating one is providing an easy way to request PCI address allocation where it normally isn't the default address type, like for aarch64 VMs. v1 posting was here: http://www.redhat.com/archives/libvir-list/2016-March/msg00268.html This series depends on calling qemuDomainAssignAddresses from PostParse: http://www.redhat.com/archives/libvir-list/2016-May/msg01075.html Changes since v1: - Rebased to master - Dropped patches that were merged - Split out AssignAddress bits to their own series - Add formatdomain docs Cole Robinson (4): util: xml: add virXMLPropertyCount domain: Make <address type='pci'/> request address allocation qemu: Wire up address type=pci auto_allocate tests: qemu: test <address type='pci'/> with aarch64 docs/formatdomain.html.in | 5 ++- docs/schemas/domaincommon.rng | 5 ++- src/conf/domain_conf.c | 30 +++++++++++++- src/conf/domain_conf.h | 1 + src/libvirt_private.syms | 1 + src/qemu/qemu_domain_address.c | 47 ++++++++++++++++++++++ src/util/virxml.c | 17 ++++++++ src/util/virxml.h | 1 + .../generic-pci-autofill-addr.xml | 27 +++++++++++++ tests/genericxml2xmltest.c | 3 ++ ...l2argv-aarch64-virtio-pci-manual-addresses.args | 4 +- ...ml2argv-aarch64-virtio-pci-manual-addresses.xml | 5 +++ .../qemuxml2argv-pci-autofill-addr.args | 25 ++++++++++++ .../qemuxml2argv-pci-autofill-addr.xml | 35 ++++++++++++++++ tests/qemuxml2argvtest.c | 1 + ...2xmlout-aarch64-virtio-pci-manual-addresses.xml | 5 +++ .../qemuxml2xmlout-pci-autofill-addr.xml | 41 +++++++++++++++++++ tests/qemuxml2xmltest.c | 1 + 18 files changed, 250 insertions(+), 4 deletions(-) create mode 100644 tests/genericxml2xmlindata/generic-pci-autofill-addr.xml create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-pci-autofill-addr.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-pci-autofill-addr.xml create mode 100644 tests/qemuxml2xmloutdata/qemuxml2xmlout-pci-autofill-addr.xml -- 2.7.4

2 6

[libvirt] [PATCH v3 0/6] qemu: handle ',' and '=' in VM name
by Cole Robinson 16 May '16

16 May '16

This series adds qemu cli comma escaping to several places that are dependent on the VM name, to enable names with embedded commas. Patch 6 makes use of qemu -name guest=X value to allow names with '=' in them. https://bugzilla.redhat.com/show_bug.cgi?id=639926 https://bugzilla.redhat.com/show_bug.cgi?id=1276485 Note: There's likely other places that are VM name dependent that need comma escaping too, but this hits the mandatory ones. I've listed some more on the BiteSizedTasks page: http://wiki.libvirt.org/page/BiteSizedTasks#qemu:_Use_comma_escaping_for_mo… v3: Patch 1 and patch 4 are new Cole Robinson (6): qemu: command: Add qemuBufferEscapeComma qemu: command: escape commas in VM name qemu: command: escape commas in secret master path qemu: command: escape commas in VNC socket path qemu: command: escape commas in chardev socket path qemu: command: Use -name guest= if available src/qemu/qemu_capabilities.c | 2 + src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_command.c | 50 ++++++++++++++++------ tests/qemucapabilitiesdata/caps_2.1.1-1.x86_64.xml | 1 + tests/qemucapabilitiesdata/caps_2.4.0-1.x86_64.xml | 1 + tests/qemucapabilitiesdata/caps_2.5.0-1.x86_64.xml | 1 + .../qemucapabilitiesdata/caps_2.6.0-1.aarch64.xml | 1 + .../qemucapabilitiesdata/caps_2.6.0-1.ppc64le.xml | 1 + tests/qemucapabilitiesdata/caps_2.6.0-1.x86_64.xml | 1 + .../qemucapabilitiesdata/caps_2.6.0-2.aarch64.xml | 1 + tests/qemucaps2xmldata/all_1.6.0-1.caps | 1 + tests/qemucaps2xmldata/nodisksnapshot_1.6.0-1.caps | 1 + .../qemuxml2argvdata/qemuxml2argv-name-escape.args | 25 +++++++++++ .../qemuxml2argvdata/qemuxml2argv-name-escape.xml | 19 ++++++++ tests/qemuxml2argvtest.c | 3 ++ 15 files changed, 96 insertions(+), 13 deletions(-) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-name-escape.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-name-escape.xml -- 2.7.4

2 8

[libvirt] [PATCH 0/3] Fix default vram setting for libxl
by Ján Tomko 16 May '16

16 May '16

An alternative to https://www.redhat.com/archives/libvir-list/2016-May/msg00700.html 1462942877-31516-1-git-send-email-jfehlig(a)suse.com This series adds a call to DeviceDefPostParse for the implicit video which is being in the DomainDefPostParse phase, after DeviceDefPostParse has been done on all the existing devices. Ján Tomko (3): Move virDomainDefPostParseInternal after virDomainDeviceDefPostParse Call per-device post-parse callback even on implicit video Fill out default vram in DeviceDefPostParse src/conf/domain_conf.c | 105 ++++++++++++++++++++++-------------------- src/qemu/qemu_parse_command.c | 2 - 2 files changed, 56 insertions(+), 51 deletions(-) -- 2.7.3

5 9

[libvirt] [PATCH] admin: Fix passing an incorrect readonly attribute to virNetServerServiceNew
by Erik Skultety 16 May '16

16 May '16

When registering admin UNIX socket, a new service is created for it. This service is incorrectly initialized to be readonly, which is later inherited by all clients connected to the socket. In libvirt-admin's case there currently isn't any use for the attribute anyway, but since the socket has root-only access permissions, but the least we can do is to make every admin client connected to it report readonly as false. Signed-off-by: Erik Skultety <eskultet(a)redhat.com> --- daemon/libvirtd.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/daemon/libvirtd.c b/daemon/libvirtd.c index cc5190f..f24fb22 100644 --- a/daemon/libvirtd.c +++ b/daemon/libvirtd.c @@ -543,7 +543,7 @@ daemonSetupNetworking(virNetServerPtr srv, #if WITH_GNUTLS NULL, #endif - true, + false, config->admin_max_queued_clients, config->admin_max_client_requests))) goto cleanup; -- 2.5.5

2 2