- Devel - libvirt List Archives

[libvirt] [PATCH v3 0/8] libxl: domain statistics support
by Joao Martins 15 Dec '15

15 Dec '15

Hey, [ Apologies for resending this series before receiving comments, but I found two issues (namely on patch 3 and 4) and thus I am reposting it as v3 having those fixed. ] This series bring support for various statistics about domains regarding CPU, Memory, Network Interfaces, Block and Jobs. Not all of the statistics are implemented: qdisk support is missing in this series and some of the memory statistics aren't available. With this series we further implement 7 more functions of libvirt APIs. It is organized as follows: * Patch 1, 2: implements cpu/memory statistics. * Patch 3: implements network statistics * Patch 4: adds helper method virDiskNameParse as an extension to virDiskNameToIndex (New) * Patch 5: VBD block statistics. QDisk will follow up in a separate series regarding QEMU monitor integration. * Patch 6: implement fetching all domain statistics * Patch 7, 8: implements Job information statistics. Overall it looks big but 70% of the patch is due to #5 and #6 but doesn't add necessarily more complexity to the driver I believe. Patch #7 and #8 are of special importance because GetJobInfo and GetJobStats are now used in Openstack Kilo to monitor live-migration progress. These two patches together with the p2p migration I sent earlier let us sucessfully live-migrate with Openstack Kilo. Furthermore with this series we get to support nova diagnostics. Tested this series on 4.4.3 and 4.5 setups plus Openstack Kilo. Individual patches contain the changelog and comments addressed since v1. Thanks! Joao Martins (8): libxl: implement virDomainGetCPUStats libxl: implement virDomainMemorystats libxl: implement virDomainInterfaceStats util: add virDiskNameParse to handle disk and partition idx libxl: implement virDomainBlockStats libxl: implement virConnectGetAllDomainStats libxl: implement virDomainGetJobInfo libxl: implement virDomainGetJobStats configure.ac | 2 +- src/libvirt_private.syms | 1 + src/libxl/libxl_domain.c | 54 +++ src/libxl/libxl_domain.h | 6 + src/libxl/libxl_driver.c | 960 +++++++++++++++++++++++++++++++++++++++++++++++ src/util/virutil.c | 41 +- src/util/virutil.h | 1 + tests/utiltest.c | 56 +++ 8 files changed, 1116 insertions(+), 5 deletions(-) -- 2.1.4

3 29

[libvirt] [PATCH v2] pci: Use virPCIDeviceAddress in virPCIDevice
by Andrea Bolognani 15 Dec '15

15 Dec '15

Instead of replicating the information (domain, bus, slot, function) inside the virPCIDevice structure, use the already-existing virPCIDeviceAddress structure. Outside of the module, the only visible change is that the return value of virPCIDeviceGetAddress() must no longer be freed by the caller. --- Changes in v2: * use virPCIDeviceAddress instead of virPCIDeviceAddressPtr to avoid extra allocations src/util/virhostdev.c | 4 --- src/util/virpci.c | 95 +++++++++++++++++++++++---------------------------- 2 files changed, 43 insertions(+), 56 deletions(-) diff --git a/src/util/virhostdev.c b/src/util/virhostdev.c index de029a0..173ac58 100644 --- a/src/util/virhostdev.c +++ b/src/util/virhostdev.c @@ -585,7 +585,6 @@ virHostdevPreparePCIDevices(virHostdevManagerPtr hostdev_mgr, goto cleanup; } - VIR_FREE(devAddr); if (!(devAddr = virPCIDeviceGetAddress(dev))) goto cleanup; @@ -728,7 +727,6 @@ virHostdevPreparePCIDevices(virHostdevManagerPtr hostdev_mgr, virObjectUnlock(hostdev_mgr->activePCIHostdevs); virObjectUnlock(hostdev_mgr->inactivePCIHostdevs); virObjectUnref(pcidevs); - VIR_FREE(devAddr); return ret; } @@ -1581,7 +1579,6 @@ virHostdevPCINodeDeviceDetach(virHostdevManagerPtr hostdev_mgr, out: virObjectUnlock(hostdev_mgr->inactivePCIHostdevs); virObjectUnlock(hostdev_mgr->activePCIHostdevs); - VIR_FREE(devAddr); return ret; } @@ -1613,7 +1610,6 @@ virHostdevPCINodeDeviceReAttach(virHostdevManagerPtr hostdev_mgr, out: virObjectUnlock(hostdev_mgr->inactivePCIHostdevs); virObjectUnlock(hostdev_mgr->activePCIHostdevs); - VIR_FREE(devAddr); return ret; } diff --git a/src/util/virpci.c b/src/util/virpci.c index bececb5..bb874ac 100644 --- a/src/util/virpci.c +++ b/src/util/virpci.c @@ -56,10 +56,7 @@ VIR_ENUM_IMPL(virPCIELinkSpeed, VIR_PCIE_LINK_SPEED_LAST, "", "2.5", "5", "8") struct _virPCIDevice { - unsigned int domain; - unsigned int bus; - unsigned int slot; - unsigned int function; + virPCIDeviceAddress address; char name[PCI_ADDR_LEN]; /* domain:bus:slot.function */ char id[PCI_ID_LEN]; /* product vendor */ @@ -655,10 +652,10 @@ virPCIDeviceSharesBusWithActive(virPCIDevicePtr dev, virPCIDevicePtr check, void virPCIDeviceList *inactiveDevs = data; /* Different domain, different bus, or simply identical device */ - if (dev->domain != check->domain || - dev->bus != check->bus || - (dev->slot == check->slot && - dev->function == check->function)) + if (dev->address.domain != check->address.domain || + dev->address.bus != check->address.bus || + (dev->address.slot == check->address.slot && + dev->address.function == check->address.function)) return 0; /* same bus, but inactive, i.e. about to be assigned to guest */ @@ -689,7 +686,7 @@ virPCIDeviceIsParent(virPCIDevicePtr dev, virPCIDevicePtr check, void *data) int ret = 0; int fd; - if (dev->domain != check->domain) + if (dev->address.domain != check->address.domain) return 0; if ((fd = virPCIDeviceConfigOpen(check, false)) < 0) @@ -713,7 +710,7 @@ virPCIDeviceIsParent(virPCIDevicePtr dev, virPCIDevicePtr check, void *data) /* if the secondary bus exactly equals the device's bus, then we found * the direct parent. No further work is necessary */ - if (dev->bus == secondary) { + if (dev->address.bus == secondary) { ret = 1; goto cleanup; } @@ -722,10 +719,12 @@ virPCIDeviceIsParent(virPCIDevicePtr dev, virPCIDevicePtr check, void *data) * In this case, what we need to do is look for the "best" match; i.e. * the most restrictive match that still satisfies all of the conditions. */ - if (dev->bus > secondary && dev->bus <= subordinate) { + if (dev->address.bus > secondary && dev->address.bus <= subordinate) { if (*best == NULL) { - *best = virPCIDeviceNew(check->domain, check->bus, check->slot, - check->function); + *best = virPCIDeviceNew(check->address.domain, + check->address.bus, + check->address.slot, + check->address.function); if (*best == NULL) { ret = -1; goto cleanup; @@ -745,8 +744,10 @@ virPCIDeviceIsParent(virPCIDevicePtr dev, virPCIDevicePtr check, void *data) if (secondary > best_secondary) { virPCIDeviceFree(*best); - *best = virPCIDeviceNew(check->domain, check->bus, check->slot, - check->function); + *best = virPCIDeviceNew(check->address.domain, + check->address.bus, + check->address.slot, + check->address.function); if (*best == NULL) { ret = -1; goto cleanup; @@ -970,7 +971,7 @@ virPCIDeviceReset(virPCIDevicePtr dev, ret = virPCIDeviceTryPowerManagementReset(dev, fd); /* Bus reset is not an option with the root bus */ - if (ret < 0 && dev->bus != 0) + if (ret < 0 && dev->address.bus != 0) ret = virPCIDeviceTrySecondaryBusReset(dev, fd, inactiveDevs); if (ret < 0) { @@ -1483,8 +1484,8 @@ virPCIDeviceWaitForCleanup(virPCIDevicePtr dev, const char *matcher) virStrToLong_ui(tmp + 1, &tmp, 16, &function) < 0 || *tmp != '\n') continue; - if (domain != dev->domain || bus != dev->bus || slot != dev->slot || - function != dev->function) + if (domain != dev->address.domain || bus != dev->address.bus || + slot != dev->address.slot || function != dev->address.function) continue; in_matching_device = true; match_depth = strspn(line, " "); @@ -1560,17 +1561,16 @@ virPCIDeviceNew(unsigned int domain, if (VIR_ALLOC(dev) < 0) return NULL; - dev->domain = domain; - dev->bus = bus; - dev->slot = slot; - dev->function = function; + dev->address.domain = domain; + dev->address.bus = bus; + dev->address.slot = slot; + dev->address.function = function; if (snprintf(dev->name, sizeof(dev->name), "%.4x:%.2x:%.2x.%.1x", - dev->domain, dev->bus, dev->slot, - dev->function) >= sizeof(dev->name)) { + domain, bus, slot, function) >= sizeof(dev->name)) { virReportError(VIR_ERR_INTERNAL_ERROR, _("dev->name buffer overflow: %.4x:%.2x:%.2x.%.1x"), - dev->domain, dev->bus, dev->slot, dev->function); + domain, bus, slot, function); goto error; } if (virAsprintf(&dev->path, PCI_SYSFS "devices/%s/config", @@ -1661,25 +1661,14 @@ virPCIDeviceFree(virPCIDevicePtr dev) * @dev: device to get address from * * Take a PCI device on input and return its PCI address. The - * caller must free the returned value when no longer needed. + * returned object is owned by the device and must not be freed. * * Returns NULL on failure, the device address on success. */ virPCIDeviceAddressPtr virPCIDeviceGetAddress(virPCIDevicePtr dev) { - - virPCIDeviceAddressPtr pciAddrPtr; - - if (!dev || (VIR_ALLOC(pciAddrPtr) < 0)) - return NULL; - - pciAddrPtr->domain = dev->domain; - pciAddrPtr->bus = dev->bus; - pciAddrPtr->slot = dev->slot; - pciAddrPtr->function = dev->function; - - return pciAddrPtr; + return &(dev->address); } const char * @@ -1890,12 +1879,14 @@ virPCIDeviceListFindIndex(virPCIDeviceListPtr list, virPCIDevicePtr dev) { size_t i; - for (i = 0; i < list->count; i++) - if (list->devs[i]->domain == dev->domain && - list->devs[i]->bus == dev->bus && - list->devs[i]->slot == dev->slot && - list->devs[i]->function == dev->function) + for (i = 0; i < list->count; i++) { + virPCIDevicePtr other = list->devs[i]; + if (other->address.domain == dev->address.domain && + other->address.bus == dev->address.bus && + other->address.slot == dev->address.slot && + other->address.function == dev->address.function) return i; + } return -1; } @@ -1910,10 +1901,11 @@ virPCIDeviceListFindByIDs(virPCIDeviceListPtr list, size_t i; for (i = 0; i < list->count; i++) { - if (list->devs[i]->domain == domain && - list->devs[i]->bus == bus && - list->devs[i]->slot == slot && - list->devs[i]->function == function) + virPCIDevicePtr other = list->devs[i]; + if (other->address.domain == domain && + other->address.bus == bus && + other->address.slot == slot && + other->address.function == function) return list->devs[i]; } return NULL; @@ -1944,7 +1936,8 @@ int virPCIDeviceFileIterate(virPCIDevicePtr dev, int direrr; if (virAsprintf(&pcidir, "/sys/bus/pci/devices/%04x:%02x:%02x.%x", - dev->domain, dev->bus, dev->slot, dev->function) < 0) + dev->address.domain, dev->address.bus, + dev->address.slot, dev->address.function) < 0) goto cleanup; if (!(dir = opendir(pcidir))) { @@ -2074,13 +2067,11 @@ virPCIDeviceListPtr virPCIDeviceGetIOMMUGroupList(virPCIDevicePtr dev) { virPCIDeviceListPtr groupList = virPCIDeviceListNew(); - virPCIDeviceAddress devAddr = { dev->domain, dev->bus, - dev->slot, dev->function }; if (!groupList) goto error; - if (virPCIDeviceAddressIOMMUGroupIterate(&devAddr, + if (virPCIDeviceAddressIOMMUGroupIterate(&(dev->address), virPCIDeviceGetIOMMUGroupAddOne, groupList) < 0) goto error; @@ -2294,7 +2285,7 @@ virPCIDeviceIsBehindSwitchLackingACS(virPCIDevicePtr dev) * into play since devices on the root bus can't P2P without going * through the root IOMMU. */ - if (dev->bus == 0) { + if (dev->address.bus == 0) { return 0; } else { virReportError(VIR_ERR_INTERNAL_ERROR, -- 2.5.0

2 3

[libvirt] [PATCH 0/2] notify about reverting to snapshot
by Dmitry Andreev 15 Dec '15

15 Dec '15

Reverting to snapshot may change domain configuration but there will be no events about that. Lack of the event become a problem for virt-manager https://bugzilla.redhat.com/show_bug.cgi?id=1081148 This patch-set introduces new event and emits it in qemuDomainRevertToSnapshot. Dmitry Andreev (2): Introduce new VIR_DOMAIN_EVENT_DEFINED_FROM_SNAPSHOT event qemu: emit 'defined' event after reverted to snapshot examples/object-events/event-test.c | 2 ++ include/libvirt/libvirt-domain.h | 1 + src/qemu/qemu_driver.c | 5 +++++ tools/virsh-domain.c | 3 ++- 4 files changed, 10 insertions(+), 1 deletion(-) -- 1.8.3.1

2 5

[libvirt] libvirt qcow2 internal snapshot without vm state
by Vasiliy Tolstov 15 Dec '15

15 Dec '15

Hi! I'm try to grep sources, but can't understand, how libvirt create internal snapshot in the same qcow2 file without memory state? As i understand libvirt via qmp/hmp runs savevm command via qemu. But i don't understand how libvirt skips memory state in this case? -- Vasiliy Tolstov, e-mail: v.tolstov(a)selfip.ru

2 2

Re: [libvirt] [Qemu-devel] [Patch V2 1/2] x86, mce: Basic support to add LMCE support to QEMU
by Eduardo Habkost 15 Dec '15

15 Dec '15

On Mon, Dec 14, 2015 at 07:17:27PM -0500, Raj, Ashok wrote: > On Mon, Dec 14, 2015 at 11:37:16PM +0100, Borislav Petkov wrote: > > On Mon, Dec 14, 2015 at 02:11:46PM -0500, Raj, Ashok wrote: > > > This is mostly harmless.. since the MCG_CAP space is shared and has no > > > conflict between vendors. Also just the CAP being set has no effect. > > > > Of course it does - we check SER_P in machine_check_poll() and when > > I emulate an AMD guest and inject errors into it, error handling is > > obviously wrong, see: > > > > https://lkml.kernel.org/r/20151123150355.GE5134@pd.tnic > > > > I can see how this hurts.. since the poller isn't doing cpu model specific > stuff..? > > in the LMCE case, even if you advertise MCG_LMCE_P in MCG_CAP, the guest kernel > wont call intel_init_lmce() only from mce_intel.c.. so the same problem > won't happen. > > but the issue Eduardo mentioned seems like the following. > > New QEMU_LMCE + New KVM_LMCE + New_GUEST_LMCE - No problem > > but if you were to migrage the Guest_LMCE to a non-LMCE supported KVM host > we could run into an issue.. > > is this the compatibility issue that you were looking to fix Eduardo? If I understood you correctly, yes. Also, note that currently kvm_arch_init_vcpu() simply warns about missing capabilities, instead of preventing the VM from running/migrating (as it should). We need to change that, and figure out a good way to report "feature FOO can't be enabled in this host" errors to management software[1]. The main problem is that we don't even have a QMP console available anymore if machine initialization is aborted. CCing libvir-list so they get in the loop. [1] This is similar to what we need for CPUID checks, but the new MCE feature means we need something more generic (that just reports QOM property names, probably?) -- Eduardo

1 0

[libvirt] [PATCHv2 0/3] security: misc cleanups
by Ján Tomko 15 Dec '15

15 Dec '15

v1: https://www.redhat.com/archives/libvir-list/2015-November/msg00781.html new in v2: * split by security driver * more functions renamed Ján Tomko (3): security_dac: remove extra Security from function names security_selinux: remove extra Security from function names security_stack: : remove extra security from function names src/security/security_dac.c | 242 +++++++++++++------------- src/security/security_selinux.c | 375 ++++++++++++++++++++-------------------- src/security/security_stack.c | 68 ++++---- 3 files changed, 338 insertions(+), 347 deletions(-) -- 2.4.6

2 5

[libvirt] volume download/upload but for snapshots
by Vasiliy Tolstov 15 Dec '15

15 Dec '15

Hi! I'm happy with libvirt volume download/upload ability, but also i need to download snapshot, what is the best way to do that? Define volume from snapshot? Or write libvirt function for this? -- Vasiliy Tolstov, e-mail: v.tolstov(a)selfip.ru

2 3

[libvirt] [PATCH v2 0/2] Add support for zero-write detection
by Martin Kletzander 15 Dec '15

15 Dec '15

QEMU supports detect-zeroes option since version 2.1, but we never added support for it in libvirt. If was requested by Vasiliy Tolstov in the list, so I just added it. There are two discussions to be had, optionally. One is to decide whether we should disable detect_zeros='unmap' if discard is not set to 'unmap', but this is getting very hypervisor-specific, so I just documented the behaviour. The other one is the naming. I described why I made the decision for "zeros" instead of "zeroes" the decision in the patch, but I have no problem changing it to what others like better. v2: - format detect_zeroes on the command line instead of detect_zeros v1: - https://www.redhat.com/archives/libvir-list/2015-December/msg00484.html Martin Kletzander (2): conf: Add support of zero-detection for disks qemu: Add support for zero-detection writes docs/formatdomain.html.in | 10 ++++++ docs/schemas/domaincommon.rng | 12 +++++++ src/conf/domain_conf.c | 23 +++++++++++++- src/conf/domain_conf.h | 11 +++++++ src/libvirt_private.syms | 2 ++ src/qemu/qemu_capabilities.c | 2 ++ src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_command.c | 11 +++++++ tests/qemucapabilitiesdata/caps_2.1.1-1.caps | 1 + tests/qemucapabilitiesdata/caps_2.4.0-1.caps | 1 + tests/qemucapabilitiesdata/caps_2.5.0-1.caps | 1 + .../qemuxml2argv-disk-drive-detect-zeros.args | 27 ++++++++++++++++ .../qemuxml2argv-disk-drive-detect-zeros.xml | 37 ++++++++++++++++++++++ tests/qemuxml2argvtest.c | 4 +++ tests/qemuxml2xmltest.c | 1 + 15 files changed, 143 insertions(+), 1 deletion(-) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-drive-detect-zeros.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-drive-detect-zeros.xml -- 2.6.4

3 7

[libvirt] Exclusive VM lock
by Dmitry Andreev 15 Dec '15

15 Dec '15

Hi, I'm looking for a mechanism to do a sequence of API calls as atomic operation. Is there any way for libvirt's API client to acquire an exclusive VM lock to prevent other client from changing VM state through libvirt?

3 5

[libvirt] [PATCH] qemu: Warn when using vhost-user without shared memory
by Martin Kletzander 15 Dec '15

15 Dec '15

When user configures vhost-user interface and forgets to also configure any shared memory, the search for the root cause of non-operational interface might take unpleasantly long time. Let's enhance user experience by emitting a warning in the logs. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1266982 Signed-off-by: Martin Kletzander <mkletzan(a)redhat.com> --- src/qemu/qemu_process.c | 44 ++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 44 insertions(+) diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index 420196264685..fb471342e790 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -4542,6 +4542,7 @@ qemuProcessLaunch(virConnectPtr conn, unsigned int hostdev_flags = 0; size_t nnicindexes = 0; int *nicindexes = NULL; + bool check_shmem = false; VIR_DEBUG("vm=%p name=%s id=%d asyncJob=%d " "incoming.launchURI=%s incoming.deferredURI=%s " @@ -4749,6 +4750,49 @@ qemuProcessLaunch(virConnectPtr conn, goto cleanup; } + VIR_DEBUG("Checking for any possible (non-fatal) issues"); + + /* + * For vhost-user to work, the domain has to have some type of + * shared memory configured. We're not the proper once to judge + * whether shared hugepages or shm are enough and will be in the + * future, so we'll just warn in case there is none of that + * configured. Moreover failing would give the false illusion + * that libvirt is really checking that everything works before + * running the domain and not only we are unable to do that, but + * it's also not our aim to do so. + */ + for (i = 0; i < vm->def->nnets; i++) { + if (virDomainNetGetActualType(vm->def->nets[i]) == + VIR_DOMAIN_NET_TYPE_VHOSTUSER) { + check_shmem = true; + break; + } + } + + if (check_shmem) { + bool shmem = vm->def->nshmems; + + /* + * This check is by no means complete. We merely check + * whetere there are *some* hugepages enabled and *some* NUMA + * nodes with shared memory access. + */ + if (!shmem && vm->def->mem.nhugepages) { + for (i = 0; i < virDomainNumaGetNodeCount(vm->def->numa); i++) { + if (virDomainNumaGetNodeMemoryAccessMode(vm->def->numa, i) == + VIR_NUMA_MEM_ACCESS_SHARED) + shmem = true; + break; + } + } + + if (!shmem) { + VIR_WARN("Detected vhost-user interface without any shared memory. " + "The interface might not be operational"); + } + } + VIR_DEBUG("Building emulator command line"); if (!(cmd = qemuBuildCommandLine(conn, driver, vm->def, priv->monConfig, priv->monJSON, priv->qemuCaps, -- 2.6.3

3 7