- Devel - libvirt List Archives

[libvirt] [RFC PATCH] threshold: new API virDomainBlockSetWriteThreshold
by Eric Blake 26 May '15

26 May '15

qemu 2.3 added a new QMP command block-set-write-threshold, which allows callers to get an interrupt when a file hits a write threshold, rather than the current approach of repeatedly polling for file allocation. This patch prepares the API for callers to register to receive the event, as well as a way to query the threshold. The event is one-shot in qemu - a guest must re-register a new threshold after each time it triggers. However, the virConnectDomainEventRegisterAny() call does not allow parameterization, so callers must use a pair of APIs - one to register the callback (one-time call), and another to repeatedly set the desired threshold (repeated each time a threshold changes). Note that the threshold is registered as a double, but reported as an unsigned long long. This is intentional, as it allows the use of a flag for registering a threshold via a percentage, but once registered, the threshold is normalized according to the current size of the disk. To make the patch series more digestible, this patch intentionally omits remote support, by using a couple of placeholders at a point where the compiler forces the addition of a case within a switch statement. Signed-off-by: Eric Blake <eblake(a)redhat.com> --- Posting now to get feedback on the proposed API, before the actual implementation is complete. daemon/remote.c | 2 + include/libvirt/libvirt-domain.h | 48 +++++++++++++++++++ src/conf/domain_event.c | 4 +- src/driver-hypervisor.h | 7 +++ src/libvirt-domain.c | 101 +++++++++++++++++++++++++++++++++++++++ src/libvirt_public.syms | 1 + tools/virsh-domain.c | 23 +++++++++ tools/virsh.pod | 1 + 8 files changed, 186 insertions(+), 1 deletion(-) diff --git a/daemon/remote.c b/daemon/remote.c index e259a76..51d7de5 100644 --- a/daemon/remote.c +++ b/daemon/remote.c @@ -1102,6 +1102,8 @@ static virConnectDomainEventGenericCallback domainEventCallbacks[] = { VIR_DOMAIN_EVENT_CALLBACK(remoteRelayDomainEventTunable), VIR_DOMAIN_EVENT_CALLBACK(remoteRelayDomainEventAgentLifecycle), VIR_DOMAIN_EVENT_CALLBACK(remoteRelayDomainEventDeviceAdded), + /* TODO: Implement RPC support for this */ + VIR_DOMAIN_EVENT_CALLBACK(NULL), }; verify(ARRAY_CARDINALITY(domainEventCallbacks) == VIR_DOMAIN_EVENT_ID_LAST); diff --git a/include/libvirt/libvirt-domain.h b/include/libvirt/libvirt-domain.h index d851225..0c4fd62 100644 --- a/include/libvirt/libvirt-domain.h +++ b/include/libvirt/libvirt-domain.h @@ -515,6 +515,15 @@ typedef virDomainBlockStatsStruct *virDomainBlockStatsPtr; # define VIR_DOMAIN_BLOCK_STATS_ERRS "errs" /** + * VIR_DOMAIN_BLOCK_STATS_WRITE_THRESHOLD: + * + * Macro represents the typed parameter, as an llong, that reports + * the threshold in bytes at which the block device will trigger a + * VIR_DOMAIN_EVENT_ID_WRITE_THRESHOLD, or 0 if no threshold is registered. + */ +# define VIR_DOMAIN_BLOCK_STATS_WRITE_THRESHOLD "write-threshold" + +/** * virDomainInterfaceStats: * * Network interface stats for virDomainInterfaceStats. @@ -1297,6 +1306,16 @@ int virDomainBlockStatsFlags (virDomainPtr dom, virTypedParameterPtr params, int *nparams, unsigned int flags); + +typedef enum { + /* threshold is a percentage rather than byte limit */ + VIR_DOMAIN_BLOCK_SET_WRITE_THRESHOLD_PERCENTAGE = (1 << 0), +} virDomainBlockSetWriteThresholdFlags; +int virDomainBlockSetWriteThreshold(virDomainPtr dom, + const char *disk, + double threshold, + unsigned int flags); + int virDomainInterfaceStats (virDomainPtr dom, const char *path, virDomainInterfaceStatsPtr stats, @@ -3246,6 +3265,34 @@ typedef void (*virConnectDomainEventDeviceAddedCallback)(virConnectPtr conn, void *opaque); /** + * virConnectDomainEventWriteThresholdCallback: + * @conn: connection object + * @dom: domain on which the event occurred + * @devAlias: device alias + * @threshold: threshold that was exceeded, in bytes + * @length: length beyond @threshold that was involved in the triggering write + * @opaque: application specified data + * + * The callback signature to use when registering for an event of type + * VIR_DOMAIN_EVENT_ID_WRITE_THRESHOLD with virConnectDomainEventRegisterAny() + * + * This callback occurs when a block device detects a write event that + * exceeds a non-zero threshold set by + * virDomainBlockSetWriteThreshold(). When this event occurs, the + * threshold is reset to 0, and a new limit must be installed to see + * the event again on the same device. The intent of this event is to + * allow time for the underlying storage to be resized dynamically + * prior to the point where the guest would be paused due to running + * out of space, without having to poll for allocation values. + */ +typedef void (*virConnectDomainEventWriteThresholdCallback)(virConnectPtr conn, + virDomainPtr dom, + const char *devAlias, + unsigned long long threshold, + unsigned long long length, + void *opaque); + +/** * VIR_DOMAIN_TUNABLE_CPU_VCPUPIN: * * Macro represents formatted pinning for one vcpu specified by id which is @@ -3528,6 +3575,7 @@ typedef enum { VIR_DOMAIN_EVENT_ID_TUNABLE = 17, /* virConnectDomainEventTunableCallback */ VIR_DOMAIN_EVENT_ID_AGENT_LIFECYCLE = 18,/* virConnectDomainEventAgentLifecycleCallback */ VIR_DOMAIN_EVENT_ID_DEVICE_ADDED = 19, /* virConnectDomainEventDeviceAddedCallback */ + VIR_DOMAIN_EVENT_ID_WRITE_THRESHOLD = 20, /* virConnectDomainEventWriteThreshold */ # ifdef VIR_ENUM_SENTINELS VIR_DOMAIN_EVENT_ID_LAST diff --git a/src/conf/domain_event.c b/src/conf/domain_event.c index 20d66e1..c43799f 100644 --- a/src/conf/domain_event.c +++ b/src/conf/domain_event.c @@ -1,7 +1,7 @@ /* * domain_event.c: domain event queue processing helpers * - * Copyright (C) 2010-2014 Red Hat, Inc. + * Copyright (C) 2010-2015 Red Hat, Inc. * Copyright (C) 2008 VirtualIron * Copyright (C) 2013 SUSE LINUX Products GmbH, Nuernberg, Germany. * @@ -1614,6 +1614,8 @@ virDomainEventDispatchDefaultFunc(virConnectPtr conn, goto cleanup; } + case VIR_DOMAIN_EVENT_ID_WRITE_THRESHOLD: + /* TODO: Implement RPC support for this */ case VIR_DOMAIN_EVENT_ID_LAST: break; } diff --git a/src/driver-hypervisor.h b/src/driver-hypervisor.h index 3275343..81b80d3 100644 --- a/src/driver-hypervisor.h +++ b/src/driver-hypervisor.h @@ -484,6 +484,12 @@ typedef int unsigned int flags); typedef int +(*virDrvDomainBlockSetWriteThreshold)(virDomainPtr domain, + const char *disk, + double threshold, + unsigned int flags); + +typedef int (*virDrvDomainInterfaceStats)(virDomainPtr domain, const char *path, virDomainInterfaceStatsPtr stats); @@ -1324,6 +1330,7 @@ struct _virHypervisorDriver { virDrvDomainBlockResize domainBlockResize; virDrvDomainBlockStats domainBlockStats; virDrvDomainBlockStatsFlags domainBlockStatsFlags; + virDrvDomainBlockSetWriteThreshold domainBlockSetWriteThreshold; virDrvDomainInterfaceStats domainInterfaceStats; virDrvDomainSetInterfaceParameters domainSetInterfaceParameters; virDrvDomainGetInterfaceParameters domainGetInterfaceParameters; diff --git a/src/libvirt-domain.c b/src/libvirt-domain.c index 2edba1a..a211de0 100644 --- a/src/libvirt-domain.c +++ b/src/libvirt-domain.c @@ -5743,6 +5743,105 @@ virDomainBlockStatsFlags(virDomainPtr dom, /** + * virDomainBlockSetWriteThreshold: + * @dom: pointer to domain object + * @disk: path to the block device, or device shorthand + * @threshold: limit at which a write threshold event can trigger + * @flags: bitwise-OR of virDomainBlockSetWriteThresholdFlags + * + * This function is used to set a one-shot write threshold. It should + * be used in tandem with virConnectDomainEventRegisterAny() + * installing a handler for VIR_DOMAIN_EVENT_ID_WRITE_THRESHOLD. If + * the hypervisor detects that a write request (whether guest data, or + * host metadata) would exceed the host byte offset specified in + * @threshold, then an event is raised, and the threshold is reset to + * 0 at that time. The event registration is only needed once, but + * this function must be called each time a new threshold is desired; + * the event will only fire if a non-zero threshold is + * exceeded. + * + * By default, @threshold is specified in bytes, and must not exceed + * the size of the block device. However, when @flags includes + * VIR_DOMAIN_BLOCK_SET_WRITE_THRESHOLD_PERCENTAGE, @threshold is + * instead a value between 0.0 and 100.0, and the driver will compute + * a byte value based on the current size of the disk. A driver may + * round the requested threshold to a granularity that can actually + * be supported. + * + * Setting a threshold allows event-based resizing of host resources + * that back a guest disk without having to poll the current disk + * allocation, while still having enough time to complete the resize + * before the guest would end up halting due to insufficient space. + * Calling this function to set the threshold back to zero will stop + * further firing of the event. virDomainBlockStatsFlags() and + * virConnectGetAllDomainStats() can be used to track the current + * threshold value, always in the form normalized to bytes. + * + * The @disk parameter is either the device target shorthand (the + * <target dev='...'/> sub-element, such as "vda"), or (since 0.9.8) + * an unambiguous source name of the block device (the <source + * file='...'/> sub-element, such as "/path/to/image"). Valid names + * can be found by calling virDomainGetXMLDesc() and inspecting + * elements within //domain/devices/disk. Some drivers might also + * accept strings such as "vda[1]" for setting the threshold of a + * backing image, useful when doing a block commit into the backing + * image. + * + * Domains may have more than one block device. To set thresholds for + * each you should make multiple calls to this function. If write + * thresholds are not supported, an application will have to instead + * poll virDomainGetBlockInfo() or similar to track allocation. + * + * Returns -1 in case of error, 0 in case of success. + */ +int +virDomainBlockSetWriteThreshold(virDomainPtr dom, + const char *disk, + double threshold, + unsigned int flags) +{ + virConnectPtr conn; + + VIR_DOMAIN_DEBUG(dom, "disk=%s, threshold=%g, flags=%x", + disk, threshold, flags); + + virResetLastError(); + + virCheckDomainReturn(dom, -1); + virCheckNonNullArgGoto(disk, error); + virCheckNonNegativeArgGoto(threshold, error); + if (flags & VIR_DOMAIN_BLOCK_SET_WRITE_THRESHOLD_PERCENTAGE) { + if (threshold > 100.0) + virReportError(VIR_ERR_INVALID_ARG, + _("threshold in %s must be a percentage"), + __FUNCTION__); + goto error; + } else { + if (threshold > LLONG_MAX) + virReportError(VIR_ERR_INVALID_ARG, + _("threshold in %s must fit in unsigned long long"), + __FUNCTION__); + goto error; + } + conn = dom->conn; + + if (conn->driver->domainBlockSetWriteThreshold) { + int ret; + ret = conn->driver->domainBlockSetWriteThreshold(dom, disk, threshold, + flags); + if (ret < 0) + goto error; + return ret; + } + virReportUnsupportedError(); + + error: + virDispatchError(dom->conn); + return -1; +} + + +/** * virDomainInterfaceStats: * @dom: pointer to the domain object * @path: path to the interface @@ -11201,6 +11300,8 @@ virConnectGetDomainCapabilities(virConnectPtr conn, * unsigned long long. * "block.<num>.errors" - Xen only: the 'oo_req' value as * unsigned long long. + * "block.<num>.write-threshold" - byte at which a write threshold event + * will fire, as unsigned long long. * "block.<num>.allocation" - offset of the highest written sector * as unsigned long long. * "block.<num>.capacity" - logical size in bytes of the block device backing diff --git a/src/libvirt_public.syms b/src/libvirt_public.syms index 716dd2f..5131fcf 100644 --- a/src/libvirt_public.syms +++ b/src/libvirt_public.syms @@ -712,6 +712,7 @@ LIBVIRT_1.2.15 { LIBVIRT_1.2.16 { global: + virDomainBlockSetWriteThreshold; virDomainSetUserPassword; } LIBVIRT_1.2.15; diff --git a/tools/virsh-domain.c b/tools/virsh-domain.c index 91a1ca2..98db6a3 100644 --- a/tools/virsh-domain.c +++ b/tools/virsh-domain.c @@ -12150,6 +12150,27 @@ vshEventDeviceAddedPrint(virConnectPtr conn ATTRIBUTE_UNUSED, } static void +vshEventWriteThresholdPrint(virConnectPtr conn ATTRIBUTE_UNUSED, + virDomainPtr dom, + const char *alias, + unsigned long long threshold, + unsigned long long length, + void *opaque) +{ + vshDomEventData *data = opaque; + + if (!data->loop && *data->count) + return; + vshPrint(data->ctl, + _("event 'write-threshold' for domain %s disk %s: threshold %lld " + "exceeded by %lld bytes\n"), + virDomainGetName(dom), alias, threshold, length); + (*data->count)++; + if (!data->loop) + vshEventDone(data->ctl); +} + +static void vshEventTunablePrint(virConnectPtr conn ATTRIBUTE_UNUSED, virDomainPtr dom, virTypedParameterPtr params, @@ -12256,6 +12277,8 @@ static vshEventCallback vshEventCallbacks[] = { VIR_DOMAIN_EVENT_CALLBACK(vshEventAgentLifecyclePrint), }, { "device-added", VIR_DOMAIN_EVENT_CALLBACK(vshEventDeviceAddedPrint), }, + { "write-threshold", + VIR_DOMAIN_EVENT_CALLBACK(vshEventWriteThresholdPrint), }, }; verify(VIR_DOMAIN_EVENT_ID_LAST == ARRAY_CARDINALITY(vshEventCallbacks)); diff --git a/tools/virsh.pod b/tools/virsh.pod index d588e5a..14dd2c3 100644 --- a/tools/virsh.pod +++ b/tools/virsh.pod @@ -919,6 +919,7 @@ local file or block device, "block.<num>.fl.reqs" - total flush requests, "block.<num>.fl.times" - total time (ns) spent on cache flushing, "block.<num>.errors" - Xen only: the 'oo_req' value, +"block.<num>.write-threshold" - write threshold event trigger, in bytes, "block.<num>.allocation" - offset of highest written sector in bytes, "block.<num>.capacity" - logical size of source file in bytes, "block.<num>.physical" - physical size of source file in bytes -- 2.1.0

3 4

[libvirt] [PATCH] docs: update github project name
by Daniel P. Berrange 26 May '15

26 May '15

The github project was renamed from libvirtproject to libvirt --- docs/downloads.html.in | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) Pushed as a trivial docs fix diff --git a/docs/downloads.html.in b/docs/downloads.html.in index 73599d6..13a6db1 100644 --- a/docs/downloads.html.in +++ b/docs/downloads.html.in @@ -81,7 +81,7 @@ </p> <pre> - <a href="https://github.com/libvirtproject/libvirt">https://github.com/libvirtproject/libvirt</a> + <a href="https://github.com/libvirt/libvirt">https://github.com/libvirt/libvirt</a> <a href="http://repo.or.cz/w/libvirt.git">http://repo.or.cz/w/libvirt.git</a> <a href="https://gitlab.com/libvirt/libvirt">https://gitlab.com/libvirt/libvirt</a></pre> -- 2.1.0

1 0

[libvirt] [PATCH] storage: Fix problem with disk backend pool allocation calculation
by John Ferlan 26 May '15

26 May '15

https://bugzilla.redhat.com/show_bug.cgi?id=1224018 The disk pool recalculates the pool allocation, capacity, and available values each time through processing a newly created disk partition. However, there were two issues with doing so. The process generally is to read the partition table via virStorageBackendDiskReadPartitions and process the volume. This is called during create and refresh backend API's with the only difference being create passes a specific volume to be processed while refresh processes all volumes (passing NULL for the volume value). The first issue via virStorageBackendDiskCreateVol was that the pool's available value was cleared, so when virStorageBackendDiskMakeVol processes the (new) volume it will ignore any other partitions on the disk, thus after returning the pool allocation would only include the newly added volume. The second is while processing a partition, the adjustment of the available value depends on the type of partition. For "primary" and "logical" partitions, the available value was adjusted. However, for "extended" partitions, the available value wasn't adjusted. Since the calling function(s) storageVolCreateXML[From] will only adjust the pool available value when they determine that the backend code didn't this resulted in the available value being incorrectly adjusted by that code. If a pool refresh was executed, the "correct" value showed up. To fix the first issue, only clear the available value when we're being processed from the refresh path (vol will be NULL). To fix the second issue, increment available by 1 so that the calling function won't adjust after we're done. That could leave the appearance of a single byte being used by a pool that only has an extended partition, but that's better than having it show up as the size of the partition until someone refreshes. The refresh path will also nick by the same value so it'll be consistent. Signed-off-by: John Ferlan <jferlan(a)redhat.com> --- src/storage/storage_backend_disk.c | 18 +++++++++++++++++- 1 file changed, 17 insertions(+), 1 deletion(-) diff --git a/src/storage/storage_backend_disk.c b/src/storage/storage_backend_disk.c index 6394dac..bc14aab 100644 --- a/src/storage/storage_backend_disk.c +++ b/src/storage/storage_backend_disk.c @@ -164,8 +164,18 @@ virStorageBackendDiskMakeDataVol(virStoragePoolObjPtr pool, return -1; } + /* For "data", adjust the pool's allocation by the size of the volume + * For "metadata" (aka "extended" volume), we haven't yet used the space; + * however, the calling createVol path will adjust the pool allocation + * by the volume allocation if it determines a storage backend doesn't + * adjust the pool's allocation value. So bump the allocation to avoid + * having the calling code misrepresent the value. The refresh path + * would resolve/change the value. + */ if (STRNEQ(groups[2], "metadata")) pool->def->allocation += vol->target.allocation; + else + pool->def->allocation++; if (vol->source.extents[0].end > pool->def->capacity) pool->def->capacity = vol->source.extents[0].end; @@ -309,7 +319,13 @@ virStorageBackendDiskReadPartitions(virStoragePoolObjPtr pool, pool->def->source.devices[0].path, NULL); - pool->def->allocation = pool->def->capacity = pool->def->available = 0; + /* The reload/restart/refresh path will recalculate everything; + * otherwise, keep allocation as is as capacity and available + * will be adjusted + */ + if (!vol) + pool->def->allocation = 0; + pool->def->capacity = pool->def->available = 0; ret = virCommandRunNul(cmd, 6, -- 2.1.0

2 2

[libvirt] Strange test build failure
by Dimitri John Ledkov 26 May '15

26 May '15

# VIR_TEST_VERBOSE=1 VIR_TEST_DEBUG=1 ./qemuxml2argvtest 2>&1 | grep NUMA 61) QEMU XML-2-ARGV hugepages-pages ... libvirt: error : unsupported configuration: NUMA node 1 is unavailable 64) QEMU XML-2-ARGV hugepages-shared ... libvirt: error : unsupported configuration: NUMA node 1 is unavailable 331) QEMU XML-2-ARGV numatune-memnode ... libvirt: error : unsupported configuration: NUMA node 1 is unavailable 333) QEMU XML-2-ARGV numatune-memnode-no-memory ... libvirt: error : unsupported configuration: NUMA node 3 is unavailable 336) QEMU XML-2-ARGV numatune-auto-prefer ... libvirt: error : unsupported configuration: NUMA node 1 is unavailable 449) QEMU XML-2-ARGV memory-hotplug-dimm ... libvirt: error : unsupported configuration: NUMA node 1 is unavailable 450) QEMU XML-2-ARGV memory-hotplug-dimm-addr ... libvirt: error : unsupported configuration: NUMA node 1 is unavailable So the test fails, but I don't believe I'm compiling libvirt with numad support... So I don't understand what is being asserted here. -- Regards, Dimitri. Pura Vida! https://clearlinux.org Open Source Technology Center Intel Corporation (UK) Ltd. - Co. Reg. #1134945 - Pipers Way, Swindon SN3 1RJ.

4 8

[libvirt] [PATCH 00/10] Get ready for migration events
by Jiri Denemark 26 May '15

26 May '15

QEMU will soon get support for migration events which will finally allows us to get rid of polling query-migrate every 50ms. However, we first need to be able to wait for all events related to migration (migration status changes, block job events, async abort requests) at once. This series prepares the infrastructure which will allows us to do it (patches 1 and 2) and then makes use of it for block job events (patch 4) and abort requests (patch 10). Patch 8 fixes a bug in migration recovery after libvirtd restarts. The rest of the patches are mostly cleanups or bugfixes or other small changes required by the 5 main patches. Jiri Denemark (10): util: Introduce thread queues thread: Create thread local condition for every thread qemu: Introduce qemuBlockJobUpdate qemu: Properly report failed migration qemu: Use thread queues for synchronous block jobs qemu: Don't mess up with disk->mirrorState Pass domain object to private data formatter/parser qemu: Make qemuMigrationCancelDriveMirror usable without async job qemu: Cancel disk mirrors after libvirtd restart qemu: Use thread queues for asyncAbort po/POTFILES.in | 2 +- src/Makefile.am | 2 + src/conf/domain_conf.c | 4 +- src/conf/domain_conf.h | 6 +- src/libvirt_private.syms | 17 ++ src/libxl/libxl_domain.c | 10 +- src/lxc/lxc_domain.c | 12 +- src/qemu/qemu_blockjob.c | 177 +++++---------------- src/qemu/qemu_blockjob.h | 15 +- src/qemu/qemu_domain.c | 97 +++++++++--- src/qemu/qemu_domain.h | 8 +- src/qemu/qemu_driver.c | 24 ++- src/qemu/qemu_migration.c | 392 +++++++++++++++++++++++++++++----------------- src/qemu/qemu_migration.h | 3 + src/qemu/qemu_process.c | 17 +- src/util/virthread.c | 21 ++- src/util/virthreadqueue.c | 343 ++++++++++++++++++++++++++++++++++++++++ src/util/virthreadqueue.h | 54 +++++++ 18 files changed, 841 insertions(+), 363 deletions(-) create mode 100644 src/util/virthreadqueue.c create mode 100644 src/util/virthreadqueue.h -- 2.4.1

4 22

[libvirt] [PATCH] qemu: Fix compilation error when enum variable size differs from 'int'
by Peter Krempa 26 May '15

26 May '15

Since commit bcd9a564b631aa virDomainNumatuneGetMode returns the value via a pointer rather than in the return value. The change triggered problems with platforms where the compiler decides to use a data type of size different than integer at the point where we typecast it. Work around the issue by using an intermediate variable of the correct type that gets casted back by the default typecasting rules. --- src/qemu/qemu_driver.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index aa0acde..2b9f125 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -10566,14 +10566,16 @@ qemuDomainGetNumaParameters(virDomainPtr dom, virMemoryParameterPtr param = &params[i]; switch (i) { - case 0: /* fill numa mode here */ + case 0: { /* fill numa mode here */ + virDomainNumatuneMemMode tmp; + virDomainNumatuneGetMode(def->numa, -1, &tmp); + if (virTypedParameterAssign(param, VIR_DOMAIN_NUMA_MODE, - VIR_TYPED_PARAM_INT, 0) < 0) + VIR_TYPED_PARAM_INT, tmp) < 0) goto cleanup; - virDomainNumatuneGetMode(def->numa, -1, - (virDomainNumatuneMemMode *) &param->value.i); break; + } case 1: /* fill numa nodeset here */ nodeset = virDomainNumatuneFormatNodeset(def->numa, NULL, -1); -- 2.4.1

3 5

[libvirt] [PATCH 0/5] Selective block device migration implementation
by Pavel Boldin 26 May '15

26 May '15

The patchset represented in the mail thread implements the selective block device migration for the QEMU driver. This closes the referenced bug [1]. First the supplementary API implemented allowing for multiple key-values pair in the virTypedParameter arrays. This is used to pass the list of block devices to migrate in the following patches. Unit tests for this are added as well. This is the subject of the first three patches. Fourth patch is adding the necessary parameter `migrate_disks' and passes it through the QEMU driver call stack. The introduced `qemuMigrateBlockDevice' function is then checks that the disk is to be migrated because it is in the list. If there is no list specified then the legacy check is used: the device is not shared, not readonly and has a source. Fifth and the last patch is adding the necessary code to the `virsh' utility making it able for user to specify a comma-separated list of the block device names that are to be migrated. The implemented Python bindings patch is to be presented. [1] https://bugzilla.redhat.com/show_bug.cgi?id=1203032 Pavel Boldin (5): util: multi-value virTypedParameter util: virTypedParamsPick* multikey API util: add virTypedParamsPackStrings qemu: migration: selective block device migration virsh: selective block device migration include/libvirt/libvirt-domain.h | 9 ++ include/libvirt/libvirt-host.h | 24 ++++ src/libvirt_public.syms | 3 + src/qemu/qemu_driver.c | 63 ++++++--- src/qemu/qemu_migration.c | 197 ++++++++++++++++---------- src/qemu/qemu_migration.h | 23 ++-- src/util/virtypedparam.c | 290 +++++++++++++++++++++++++++++++-------- tests/Makefile.am | 6 + tests/virtypedparamtest.c | 289 ++++++++++++++++++++++++++++++++++++++ tools/virsh-domain.c | 44 ++++++ 10 files changed, 788 insertions(+), 160 deletions(-) create mode 100644 tests/virtypedparamtest.c -- 1.9.1

3 27

[libvirt] [PATCH v3 0/4] storage: fs: tweak dir perms handling on build
by Cole Robinson 26 May '15

26 May '15

Consider the following issue - Using virt-manager with qemu:///session - User adds a storage pool pointing at /tmp. No explicit permissions are requested in the XML - virt-manager calls PoolDefine, then PoolBuild - libvirt tries to unconditionally chmod 755 /tmp. This fails because my user doesn't own root. Pool build fails, virt-manager reports failure Yes there's a couple ways we could avoid this specific case in virt-manager, but I think it makes more sense to have pool.build on a directory be a no-op in this case. The following patches address this. I already pushed some bits of v2 of the series. Now the patches are: - Patch 1 (new) update existing docs about storage <permissions> - Patch 2 changes storage XML to not hardcode a <mode> value if the user didn't specify one in. - Patch 3 (new) don't output empty <permissions> block - Patch 4 makes pool.build skip dir chmod unless the user explicitly requested <mode> via the XML. However if a mode is required for mkdir, continue to use the previous default. Cole Robinson (4): docs: formatstorage: Update <permissions> docs storage: conf: Don't set any default <mode> in the XML conf: storage: Don't emit empty <permissions> block storage: fs: Only force directory permissions if required docs/formatstorage.html.in | 40 ++++++++----- docs/schemas/storagecommon.rng | 8 ++- src/conf/storage_conf.c | 70 ++++++++++++---------- src/storage/storage_backend.c | 20 +++++-- src/storage/storage_backend.h | 3 + src/storage/storage_backend_fs.c | 32 ++++++---- src/storage/storage_backend_logical.c | 4 +- src/util/virfile.c | 4 +- tests/storagepoolxml2xmlout/pool-netfs-gluster.xml | 3 - tests/storagevolxml2xmlout/vol-gluster-dir.xml | 3 - tests/storagevolxml2xmlout/vol-sheepdog.xml | 3 - 11 files changed, 113 insertions(+), 77 deletions(-) -- 2.4.1

2 13

[libvirt] [PATCH 0/8] parallels: add vz driver
by Maxim Nestratov 25 May '15

25 May '15

This new driver 'vz' will be used as a substitution for parallels driver. New domains will be created with 'vz' domain type. Old 'parallels' domains remain supported. Connection to the driver can be made either way: vz:///system or parallels:///system. Maxim Nestratov (8): parallels: introduce vz driver constant and string parallels: use newly introduced VIR_DOMAIN_VIRT_VZ parallels: add new guest capabilities assigned to vz driver parallels: accept vz as a driver uri and name parallels: add a new vz connection driver and hypervisor structures parallels: increment the number of connection drivers parallels: recommend to connect to vz:///system when connection fails parallels: set VIR_DOMAIN_VIRT_VZ virtType for new domains src/conf/domain_conf.c | 19 +++++++++----- src/conf/domain_conf.h | 1 + src/libvirt.c | 2 +- src/parallels/parallels_driver.c | 52 ++++++++++++++++++++++++++++++++++----- src/parallels/parallels_network.c | 3 ++- src/parallels/parallels_sdk.c | 2 +- src/parallels/parallels_storage.c | 3 ++- 7 files changed, 66 insertions(+), 16 deletions(-) -- 2.1.0

2 9

[libvirt] [PATCH 0/2 v2] parallels: fix prlsdkLoadDomain
by Maxim Nestratov 25 May '15

25 May '15

Maxim Nestratov (2): parallels: move up updating parameter in prlsdkLoadDomain parallels: fix possible crash in case of errors in prlsdkLoadDomain src/parallels/parallels_sdk.c | 35 +++++++++++++++++++++++------------ 1 file changed, 23 insertions(+), 12 deletions(-) -- 2.1.0

2 4

[libvirt] [PATCH] rpc: add testing of RPC JSON (de)serialization
by Daniel P. Berrange 25 May '15

25 May '15

The virNetServer class has the ability to serialize its state to a JSON file, and then re-load that data after an in-place execve() call to re-connect to active file handles. This data format is critical ABI that must have compatibility across releases, so it should be tested... --- src/libvirt_remote.syms | 1 + src/rpc/virnetserver.c | 4 +- src/rpc/virnetserver.h | 3 + src/rpc/virnetserverclient.c | 13 +- src/rpc/virnetserverservice.c | 6 +- tests/Makefile.am | 7 + tests/virnetserverdata/README | 14 + .../virnetserverdata/input-data-anon-clients.json | 63 +++++ tests/virnetserverdata/input-data-initial.json | 62 +++++ .../virnetserverdata/output-data-anon-clients.json | 63 +++++ tests/virnetserverdata/output-data-initial.json | 63 +++++ tests/virnetservertest.c | 290 +++++++++++++++++++++ 12 files changed, 579 insertions(+), 10 deletions(-) create mode 100644 tests/virnetserverdata/README create mode 100644 tests/virnetserverdata/input-data-anon-clients.json create mode 100644 tests/virnetserverdata/input-data-initial.json create mode 100644 tests/virnetserverdata/output-data-anon-clients.json create mode 100644 tests/virnetserverdata/output-data-initial.json create mode 100644 tests/virnetservertest.c diff --git a/src/libvirt_remote.syms b/src/libvirt_remote.syms index 950e56e..bdd68f6 100644 --- a/src/libvirt_remote.syms +++ b/src/libvirt_remote.syms @@ -77,6 +77,7 @@ xdr_virNetMessageError; # rpc/virnetserver.h +virNetServerAddClient; virNetServerAddProgram; virNetServerAddService; virNetServerAddShutdownInhibition; diff --git a/src/rpc/virnetserver.c b/src/rpc/virnetserver.c index 42427dc..091a1dc 100644 --- a/src/rpc/virnetserver.c +++ b/src/rpc/virnetserver.c @@ -259,8 +259,8 @@ static int virNetServerDispatchNewMessage(virNetServerClientPtr client, } -static int virNetServerAddClient(virNetServerPtr srv, - virNetServerClientPtr client) +int virNetServerAddClient(virNetServerPtr srv, + virNetServerClientPtr client) { virObjectLock(srv); diff --git a/src/rpc/virnetserver.h b/src/rpc/virnetserver.h index 8c5ae07..4b452be 100644 --- a/src/rpc/virnetserver.h +++ b/src/rpc/virnetserver.h @@ -79,6 +79,9 @@ int virNetServerAddService(virNetServerPtr srv, virNetServerServicePtr svc, const char *mdnsEntryName); +int virNetServerAddClient(virNetServerPtr srv, + virNetServerClientPtr client); + int virNetServerAddProgram(virNetServerPtr srv, virNetServerProgramPtr prog); diff --git a/src/rpc/virnetserverclient.c b/src/rpc/virnetserverclient.c index 34c1994..0e3a71f 100644 --- a/src/rpc/virnetserverclient.c +++ b/src/rpc/virnetserverclient.c @@ -536,13 +536,14 @@ virJSONValuePtr virNetServerClientPreExecRestart(virNetServerClientPtr client) goto error; } - if (client->privateData && client->privateDataPreExecRestart && - !(child = client->privateDataPreExecRestart(client, client->privateData))) - goto error; + if (client->privateData && client->privateDataPreExecRestart) { + if (!(child = client->privateDataPreExecRestart(client, client->privateData))) + goto error; - if (virJSONValueObjectAppend(object, "privateData", child) < 0) { - virJSONValueFree(child); - goto error; + if (virJSONValueObjectAppend(object, "privateData", child) < 0) { + virJSONValueFree(child); + goto error; + } } virObjectUnlock(client); diff --git a/src/rpc/virnetserverservice.c b/src/rpc/virnetserverservice.c index d3cf31a..4df26cb 100644 --- a/src/rpc/virnetserverservice.c +++ b/src/rpc/virnetserverservice.c @@ -303,12 +303,15 @@ virNetServerServicePtr virNetServerServiceNewFD(int fd, /* IO callback is initially disabled, until we're ready * to deal with incoming clients */ + virObjectRef(svc); if (virNetSocketAddIOCallback(svc->socks[i], 0, virNetServerServiceAccept, svc, - virObjectFreeCallback) < 0) + virObjectFreeCallback) < 0) { + virObjectUnref(svc); goto error; + } } @@ -388,7 +391,6 @@ virNetServerServicePtr virNetServerServiceNewPostExecRestart(virJSONValuePtr obj svc, virObjectFreeCallback) < 0) { virObjectUnref(svc); - virObjectUnref(sock); goto error; } } diff --git a/tests/Makefile.am b/tests/Makefile.am index 8e2dbec..c9e2c8a 100644 --- a/tests/Makefile.am +++ b/tests/Makefile.am @@ -188,6 +188,7 @@ if WITH_REMOTE test_programs += \ virnetmessagetest \ virnetsockettest \ + virnetservertest \ virnetserverclienttest \ $(NULL) if WITH_GNUTLS @@ -921,6 +922,12 @@ virnetsockettest_SOURCES = \ virnetsockettest.c testutils.h testutils.c virnetsockettest_LDADD = $(LDADDS) +virnetservertest_SOURCES = \ + virnetservertest.c \ + testutils.h testutils.c +virnetservertest_CFLAGS = $(XDR_CFLAGS) $(AM_CFLAGS) +virnetservertest_LDADD = $(LDADDS) + virnetserverclienttest_SOURCES = \ virnetserverclienttest.c \ testutils.h testutils.c diff --git a/tests/virnetserverdata/README b/tests/virnetserverdata/README new file mode 100644 index 0000000..d6d79d2 --- /dev/null +++ b/tests/virnetserverdata/README @@ -0,0 +1,14 @@ + virnetservertest data files + =========================== + +The various input-data-*.json files are a record of all the historical +formats that libvirt has been able to produce data for. Everytime a +new field is added to the JSON output, a *new* input data file should +be created. We must not add new fields to existing input-data files, +nor must we ever re-structure them if code changes, as we must check +new code handles the legacy formats. + +The various output-data-*.json files are the record of what the *new* +JSON output should look like for the correspondingly named input-data +file. It is permissible to change the existing output-data-*.json +files if the format we save in is updated. diff --git a/tests/virnetserverdata/input-data-anon-clients.json b/tests/virnetserverdata/input-data-anon-clients.json new file mode 100644 index 0000000..ed91b57 --- /dev/null +++ b/tests/virnetserverdata/input-data-anon-clients.json @@ -0,0 +1,63 @@ +{ + "min_workers": 10, + "max_workers": 50, + "priority_workers": 5, + "max_clients": 100, + "max_anonymous_clients": 10, + "keepaliveInterval": 120, + "keepaliveCount": 5, + "keepaliveRequired": true, + "mdnsGroupName": "libvirtTest", + "services": [ + { + "auth": 0, + "readonly": true, + "nrequests_client_max": 2, + "socks": [ + { + "fd": 100, + "errfd": -1, + "pid": 0, + "isClient": false + } + ] + }, + { + "auth": 2, + "readonly": false, + "nrequests_client_max": 5, + "socks": [ + { + "fd": 101, + "errfd": -1, + "pid": 0, + "isClient": false + } + ] + } + ], + "clients": [ + { + "auth": 1, + "readonly": true, + "nrequests_max": 15, + "sock": { + "fd": 102, + "errfd": -1, + "pid": -1, + "isClient": true + } + }, + { + "auth": 2, + "readonly": true, + "nrequests_max": 66, + "sock": { + "fd": 103, + "errfd": -1, + "pid": -1, + "isClient": true + } + } + ] +} diff --git a/tests/virnetserverdata/input-data-initial.json b/tests/virnetserverdata/input-data-initial.json new file mode 100644 index 0000000..7956225 --- /dev/null +++ b/tests/virnetserverdata/input-data-initial.json @@ -0,0 +1,62 @@ +{ + "min_workers": 10, + "max_workers": 50, + "priority_workers": 5, + "max_clients": 100, + "keepaliveInterval": 120, + "keepaliveCount": 5, + "keepaliveRequired": true, + "mdnsGroupName": "libvirtTest", + "services": [ + { + "auth": 0, + "readonly": true, + "nrequests_client_max": 2, + "socks": [ + { + "fd": 100, + "errfd": -1, + "pid": 0, + "isClient": false + } + ] + }, + { + "auth": 2, + "readonly": false, + "nrequests_client_max": 5, + "socks": [ + { + "fd": 101, + "errfd": -1, + "pid": 0, + "isClient": false + } + ] + } + ], + "clients": [ + { + "auth": 1, + "readonly": true, + "nrequests_max": 15, + "sock": { + "fd": 102, + "errfd": -1, + "pid": -1, + "isClient": true + } + }, + { + "auth": 2, + "readonly": true, + "nrequests_max": 66, + "sock": { + "fd": 103, + "errfd": -1, + "pid": -1, + "isClient": true + } + } + ] +} diff --git a/tests/virnetserverdata/output-data-anon-clients.json b/tests/virnetserverdata/output-data-anon-clients.json new file mode 100644 index 0000000..ed91b57 --- /dev/null +++ b/tests/virnetserverdata/output-data-anon-clients.json @@ -0,0 +1,63 @@ +{ + "min_workers": 10, + "max_workers": 50, + "priority_workers": 5, + "max_clients": 100, + "max_anonymous_clients": 10, + "keepaliveInterval": 120, + "keepaliveCount": 5, + "keepaliveRequired": true, + "mdnsGroupName": "libvirtTest", + "services": [ + { + "auth": 0, + "readonly": true, + "nrequests_client_max": 2, + "socks": [ + { + "fd": 100, + "errfd": -1, + "pid": 0, + "isClient": false + } + ] + }, + { + "auth": 2, + "readonly": false, + "nrequests_client_max": 5, + "socks": [ + { + "fd": 101, + "errfd": -1, + "pid": 0, + "isClient": false + } + ] + } + ], + "clients": [ + { + "auth": 1, + "readonly": true, + "nrequests_max": 15, + "sock": { + "fd": 102, + "errfd": -1, + "pid": -1, + "isClient": true + } + }, + { + "auth": 2, + "readonly": true, + "nrequests_max": 66, + "sock": { + "fd": 103, + "errfd": -1, + "pid": -1, + "isClient": true + } + } + ] +} diff --git a/tests/virnetserverdata/output-data-initial.json b/tests/virnetserverdata/output-data-initial.json new file mode 100644 index 0000000..da02aac --- /dev/null +++ b/tests/virnetserverdata/output-data-initial.json @@ -0,0 +1,63 @@ +{ + "min_workers": 10, + "max_workers": 50, + "priority_workers": 5, + "max_clients": 100, + "max_anonymous_clients": 100, + "keepaliveInterval": 120, + "keepaliveCount": 5, + "keepaliveRequired": true, + "mdnsGroupName": "libvirtTest", + "services": [ + { + "auth": 0, + "readonly": true, + "nrequests_client_max": 2, + "socks": [ + { + "fd": 100, + "errfd": -1, + "pid": 0, + "isClient": false + } + ] + }, + { + "auth": 2, + "readonly": false, + "nrequests_client_max": 5, + "socks": [ + { + "fd": 101, + "errfd": -1, + "pid": 0, + "isClient": false + } + ] + } + ], + "clients": [ + { + "auth": 1, + "readonly": true, + "nrequests_max": 15, + "sock": { + "fd": 102, + "errfd": -1, + "pid": -1, + "isClient": true + } + }, + { + "auth": 2, + "readonly": true, + "nrequests_max": 66, + "sock": { + "fd": 103, + "errfd": -1, + "pid": -1, + "isClient": true + } + } + ] +} diff --git a/tests/virnetservertest.c b/tests/virnetservertest.c new file mode 100644 index 0000000..08431d7 --- /dev/null +++ b/tests/virnetservertest.c @@ -0,0 +1,290 @@ +/* + * Copyright (C) 2015 Red Hat, Inc. + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library. If not, see + * <http://www.gnu.org/licenses/>. + * + * Author: Daniel P. Berrange <berrange(a)redhat.com> + */ + +#include <config.h> + +#include "testutils.h" +#include "virerror.h" +#include "rpc/virnetserver.h" + +#define VIR_FROM_THIS VIR_FROM_RPC + +#ifdef HAVE_SOCKETPAIR +static virNetServerPtr +testCreateServer(const char *host, int family) +{ + virNetServerPtr srv = NULL; + virNetServerServicePtr svc1 = NULL, svc2 = NULL; + virNetServerClientPtr cln1 = NULL, cln2 = NULL; + virNetSocketPtr sk1 = NULL, sk2 = NULL; + int fdclient[2]; + + if (socketpair(PF_UNIX, SOCK_STREAM, 0, fdclient) < 0) { + virReportSystemError(errno, "%s", + "Cannot create socket pair"); + goto cleanup; + } + + if (!(srv = virNetServerNew(10, 50, 5, 100, 10, + 120, 5, true, + "libvirtTest", + NULL, + NULL, + NULL, + NULL))) + goto error; + + if (!(svc1 = virNetServerServiceNewTCP(host, + NULL, + family, + VIR_NET_SERVER_SERVICE_AUTH_NONE, +# ifdef WITH_GNUTLS + NULL, +# endif + true, + 5, + 2))) + goto error; + + if (!(svc2 = virNetServerServiceNewTCP(host, + NULL, + VIR_NET_SERVER_SERVICE_AUTH_POLKIT, + family, +# ifdef WITH_GNUTLS + NULL, +# endif + false, + 25, + 5))) + goto error; + + if (virNetServerAddService(srv, svc1, "libvirt-ro") < 0) + goto error; + if (virNetServerAddService(srv, svc2, "libvirt-ro") < 0) + goto error; + + if (virNetSocketNewConnectSockFD(fdclient[0], &sk1) < 0) + goto error; + if (virNetSocketNewConnectSockFD(fdclient[1], &sk2) < 0) + goto error; + + if (!(cln1 = virNetServerClientNew(sk1, + VIR_NET_SERVER_SERVICE_AUTH_SASL, + true, + 15, +# ifdef WITH_GNUTLS + NULL, +# endif + NULL, NULL, NULL, NULL))) + goto error; + + if (!(cln2 = virNetServerClientNew(sk2, + VIR_NET_SERVER_SERVICE_AUTH_POLKIT, + true, + 66, +# ifdef WITH_GNUTLS + NULL, +# endif + NULL, NULL, NULL, NULL))) + goto error; + + if (virNetServerAddClient(srv, cln1) < 0) + goto error; + + if (virNetServerAddClient(srv, cln2) < 0) + goto error; + + cleanup: + virObjectUnref(cln1); + virObjectUnref(cln2); + virObjectUnref(svc1); + virObjectUnref(svc2); + return srv; + + error: + virObjectUnref(srv); + srv = NULL; + goto cleanup; +} + +static char *testGenerateJSON(void) +{ + virNetServerPtr srv = NULL; + virJSONValuePtr json = NULL; + char *jsonstr = NULL; + bool has_ipv4, has_ipv6; + + /* Our pre-saved JSON file is created so that each service + * only has one socket. If we let libvirt bind to IPv4 and + * IPv6 we might end up with two sockets, so force one or + * the other based on what's available on thehost + */ + if (virNetSocketCheckProtocols(&has_ipv4, + &has_ipv6) < 0) + return NULL; + + if (!has_ipv4 && !has_ipv6) + return NULL; + + if (!(srv = testCreateServer( + has_ipv4 ? "127.0.0.1" : "::1", + has_ipv4 ? AF_INET : AF_INET6))) + goto cleanup; + + if (!(json = virNetServerPreExecRestart(srv))) + goto cleanup; + + if (!(jsonstr = virJSONValueToString(json, true))) + goto cleanup; + fprintf(stderr, "%s\n", jsonstr); + cleanup: + virNetServerClose(srv); + virObjectUnref(srv); + virJSONValueFree(json); + return jsonstr; +} + + +struct testExecRestartData { + const char *jsonfile; +}; + +static int testExecRestart(const void *opaque) +{ + int ret = -1; + virNetServerPtr srv = NULL; + const struct testExecRestartData *data = opaque; + char *infile = NULL, *outfile = NULL; + char *injsonstr = NULL, *outjsonstr = NULL; + virJSONValuePtr injson = NULL, outjson = NULL; + int fdclient[2] = { -1, -1 }, fdserver[2] = { -1, -1 }; + + if (socketpair(PF_UNIX, SOCK_STREAM, 0, fdclient) < 0) { + virReportSystemError(errno, "%s", + "Cannot create socket pair"); + goto cleanup; + } + + if (socketpair(PF_UNIX, SOCK_STREAM, 0, fdserver) < 0) { + virReportSystemError(errno, "%s", + "Cannot create socket pair"); + goto cleanup; + } + + /* We're blindly assuming the test case isn't using + * fds 100->103 for something else, which is probably + * fairly reasonable in general + */ + dup2(fdserver[0], 100); + dup2(fdserver[1], 101); + dup2(fdclient[0], 102); + dup2(fdclient[1], 103); + + if (virAsprintf(&infile, "%s/virnetserverdata/input-data-%s.json", + abs_srcdir, data->jsonfile) < 0) + goto cleanup; + + if (virAsprintf(&outfile, "%s/virnetserverdata/output-data-%s.json", + abs_srcdir, data->jsonfile) < 0) + goto cleanup; + + if (virFileReadAll(infile, 8192, &injsonstr) < 0) + goto cleanup; + + if (!(injson = virJSONValueFromString(injsonstr))) + goto cleanup; + + if (!(srv = virNetServerNewPostExecRestart(injson, + NULL, NULL, NULL, + NULL, NULL))) + goto cleanup; + + if (!(outjson = virNetServerPreExecRestart(srv))) + goto cleanup; + + if (!(outjsonstr = virJSONValueToString(outjson, true))) + goto cleanup; + + if (virtTestCompareToFile(outjsonstr, outfile) < 0) + goto fail; + + ret = 0; + + cleanup: + if (ret < 0) { + virErrorPtr err = virGetLastError(); + fprintf(stderr, "%s\n", err->message); + } + fail: + VIR_FREE(infile); + VIR_FREE(outfile); + VIR_FREE(injsonstr); + VIR_FREE(outjsonstr); + virJSONValueFree(injson); + virJSONValueFree(outjson); + virObjectUnref(srv); + VIR_FORCE_CLOSE(fdserver[0]); + VIR_FORCE_CLOSE(fdserver[1]); + VIR_FORCE_CLOSE(fdclient[0]); + VIR_FORCE_CLOSE(fdclient[1]); + return ret; +} + + +static int +mymain(void) +{ + int ret = 0; + + /* Hack to make it easier to generate new JSON files when + * the RPC classes change. Just set this env var, save + * the generated JSON, and replace the file descriptor + * numbers with 100, 101, 102, 103. + */ + if (getenv("VIR_GENERATE_JSON")) { + char *json = testGenerateJSON(); + fprintf(stdout, "%s\n", json); + VIR_FREE(json); + return EXIT_SUCCESS; + } + +# define EXEC_RESTART_TEST(file) \ + do { \ + struct testExecRestartData data = { \ + file \ + }; \ + if (virtTestRun("ExecRestart " file, \ + testExecRestart, &data) < 0) \ + ret = -1; \ + } while (0) + + EXEC_RESTART_TEST("initial"); + EXEC_RESTART_TEST("anon-clients"); + + return ret == 0 ? EXIT_SUCCESS : EXIT_FAILURE; +} +#else +static int +mymain(void) +{ + return EXIT_AM_SKIP; +} +#endif +VIRT_TEST_MAIN(mymain); -- 2.1.0

2 1

[libvirt] [PATCH 0/3] Patches for a couple of Coverity errors
by John Ferlan 25 May '15

25 May '15

The first one is one I forgot the last time I sent Coverity patches a few weeks ago. The next two are from changes made recently. John Ferlan (3): storage: Resolve Coverity FORWARD_NULL conf: Resolve Coverity FORWARD_NULL network: Resolve Coverity FORWARD_NULL src/conf/domain_conf.c | 2 ++ src/network/bridge_driver.c | 2 +- src/storage/storage_backend_disk.c | 2 +- 3 files changed, 4 insertions(+), 2 deletions(-) -- 2.1.0

3 14

[libvirt] [PATCH] virsh: reject negative values for scaled integer
by Pavel Hrdina 25 May '15

25 May '15

Some virsh commands have a size parameter, which is handled as scaled integer. We don't have any *feature* that would allow to use '-1' as maximum size, so it's safe to reject any negative values for those commands. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1159171 Signed-off-by: Pavel Hrdina <phrdina(a)redhat.com> --- tools/virsh.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/virsh.c b/tools/virsh.c index 4425774..1005ba8 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -1811,7 +1811,7 @@ vshCommandOptScaledInt(const vshCmd *cmd, const char *name, ret = vshCommandOptString(cmd, name, &str); if (ret <= 0) return ret; - if (virStrToLong_ull(str, &end, 10, value) < 0 || + if (virStrToLong_ullp(str, &end, 10, value) < 0 || virScaleInteger(value, end, scale, max) < 0) return -1; return 1; -- 2.3.6

2 2

[libvirt] [PATCH] libxl: load on FreeBSD
by Roman Bogorodskiy 24 May '15

24 May '15

The libxl tries to check if it's running in dom0 by parsing /proc/xen/capabilities and if that fails it doesn't load. There's no procfs interface in Xen on FreeBSD, so this check always fails. Instead of using procfs, check if /dev/xen/xenstored, that's enough to check if we're running in dom0 in FreeBSD case. -- The 'HYPERVISOR_CAPABILITIES' name could be misleading now, however, I'd prefer to use a common variable to avoid duplicating of the file checking code. Maybe it should be renamed to something like HYPERVISOR_CONTROL? --- src/libxl/libxl_driver.c | 41 +++++++++++++++++++++++++---------------- 1 file changed, 25 insertions(+), 16 deletions(-) diff --git a/src/libxl/libxl_driver.c b/src/libxl/libxl_driver.c index 12be816..c848210 100644 --- a/src/libxl/libxl_driver.c +++ b/src/libxl/libxl_driver.c @@ -73,7 +73,11 @@ VIR_LOG_INIT("libxl.libxl_driver"); #define LIBXL_CONFIG_FORMAT_XM "xen-xm" #define LIBXL_CONFIG_FORMAT_SEXPR "xen-sxpr" -#define HYPERVISOR_CAPABILITIES "/proc/xen/capabilities" +#ifdef __FreeBSD__ +# define HYPERVISOR_CAPABILITIES "/dev/xen/xenstored" +#else +# define HYPERVISOR_CAPABILITIES "/proc/xen/capabilities" +#endif /* Number of Xen scheduler parameters */ #define XEN_SCHED_CREDIT_NPARAM 2 @@ -427,8 +431,6 @@ static bool libxlDriverShouldLoad(bool privileged) { bool ret = false; - int status; - char *output = NULL; /* Don't load if non-root */ if (!privileged) { @@ -441,21 +443,28 @@ libxlDriverShouldLoad(bool privileged) " does not exist"); return ret; } - /* - * Don't load if not running on a Xen control domain (dom0). It is not - * sufficient to check for the file to exist as any guest can mount - * xenfs to /proc/xen. - */ - status = virFileReadAll(HYPERVISOR_CAPABILITIES, 10, &output); - if (status >= 0) - status = strncmp(output, "control_d", 9); - VIR_FREE(output); - if (status) { - VIR_INFO("No Xen capabilities detected, probably not running " - "in a Xen Dom0. Disabling libxenlight driver"); +#if !defined(__FreeBSD__) + { + int status; + char *output = NULL; - return ret; + /* + * Don't load if not running on a Xen control domain (dom0). It is not + * sufficient to check for the file to exist as any guest can mount + * xenfs to /proc/xen. + */ + status = virFileReadAll(HYPERVISOR_CAPABILITIES, 10, &output); + if (status >= 0) + status = strncmp(output, "control_d", 9); + VIR_FREE(output); + if (status) { + VIR_INFO("No Xen capabilities detected, probably not running " + "in a Xen Dom0. Disabling libxenlight driver"); + + return ret; + } } +#endif /* Don't load if legacy xen toolstack (xend) is in use */ if (virFileExists("/usr/sbin/xend")) { -- 2.3.7

2 2

[libvirt] [PATCH] docs: Document new RO repo mirrors
by Michal Privoznik 24 May '15

24 May '15

In the upstream discussion on creating a github mirror [1], it turned out that there are some read-only mirrors of our repository. Lets advertise them on our downloads page. But do it wisely and discourage people in sending a pull requests on GitHub. 1: https://www.redhat.com/archives/libvir-list/2015-May/msg00775.html Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- docs/downloads.html.in | 14 ++++++++++++++ 1 file changed, 14 insertions(+) diff --git a/docs/downloads.html.in b/docs/downloads.html.in index 435c2f1..a0ef7e6 100644 --- a/docs/downloads.html.in +++ b/docs/downloads.html.in @@ -71,6 +71,20 @@ <pre> <a href="http://libvirt.org/git/?p=libvirt.git;a=summary">http://libvirt.org/git/?p=libvirt.git;a=summary</a></pre> + <p> + In addition to this repository, there are following read-only + repositories which mirror the master one. Note that we currently may not + be using the full set of features on these mirrors (e.g. pull requests on + GitHub, please don't use them). We want the discussion and patch review + to happen at one place - on the list. Also note that some repositories + listed below allows HTTP checkouts too. + </p> + + <pre> + <a href="https://github.com/libvirtproject/libvirt">https://github.com/libvirtproject/libvirt</a> + <a href="http://repo.or.cz/w/libvirt.git">http://repo.or.cz/w/libvirt.git</a> + <a href="https://gitlab.com/libvirt/libvirt">https://gitlab.com/libvirt/libvirt</a></pre> + <br /> <h1>libvirt Application Development Guide</h1> -- 2.3.6

2 1

[libvirt] [PATCH] zfs: fix storagepoolxml2xml test
by Roman Bogorodskiy 24 May '15

24 May '15

Commit c4d27bd dropped output of owner/group -1. Update zfs tests accordingly. --- tests/storagepoolxml2xmlout/pool-zfs-sourcedev.xml | 2 -- tests/storagepoolxml2xmlout/pool-zfs.xml | 2 -- 2 files changed, 4 deletions(-) diff --git a/tests/storagepoolxml2xmlout/pool-zfs-sourcedev.xml b/tests/storagepoolxml2xmlout/pool-zfs-sourcedev.xml index bbd2e9f..89dcdbf 100644 --- a/tests/storagepoolxml2xmlout/pool-zfs-sourcedev.xml +++ b/tests/storagepoolxml2xmlout/pool-zfs-sourcedev.xml @@ -12,8 +12,6 @@ <path>/dev/zvol/testpool</path> <permissions> <mode>0755</mode> - <owner>-1</owner> - <group>-1</group> </permissions> </target> </pool> diff --git a/tests/storagepoolxml2xmlout/pool-zfs.xml b/tests/storagepoolxml2xmlout/pool-zfs.xml index ff02329..c9e5df9 100644 --- a/tests/storagepoolxml2xmlout/pool-zfs.xml +++ b/tests/storagepoolxml2xmlout/pool-zfs.xml @@ -11,8 +11,6 @@ <path>/dev/zvol/testpool</path> <permissions> <mode>0755</mode> - <owner>-1</owner> - <group>-1</group> </permissions> </target> </pool> -- 2.3.7

2 2

[libvirt] [PATCH] bhyve: fix build with gcc48
by Roman Bogorodskiy 24 May '15

24 May '15

Build with gcc 4.8 fails with: bhyve/bhyve_monitor.c: In function 'bhyveMonitorIO': bhyve/bhyve_monitor.c:51:18: error: missing initializer for field 'tv_sec' of 'const struct timespec' [-Werror=missing-field-initializers] const struct timespec zerowait = {}; Explicitly initialize zerowait to fix the build. --- src/bhyve/bhyve_monitor.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/bhyve/bhyve_monitor.c b/src/bhyve/bhyve_monitor.c index 7f19c6e..1316720 100644 --- a/src/bhyve/bhyve_monitor.c +++ b/src/bhyve/bhyve_monitor.c @@ -48,7 +48,7 @@ struct _bhyveMonitor { static void bhyveMonitorIO(int watch, int kq, int events ATTRIBUTE_UNUSED, void *opaque) { - const struct timespec zerowait = {}; + const struct timespec zerowait = { 0, 0 }; bhyveMonitorPtr mon = opaque; struct kevent kev; int rc, status; -- 2.3.7

3 3

[libvirt] [PATCH] Add missing XDR_FLAGS
by Pavel Fedin 23 May '15

23 May '15

Fixes build problems on x86_64-cygwin host for aarch64 target: CC lxc/libvirt_driver_lxc_impl_la-lxc_monitor_protocol.lo In file included from lxc/lxc_monitor_protocol.c:7:0: lxc/lxc_monitor_protocol.h:9:21: fatal error: rpc/rpc.h: No such file or directory CC rpc/libvirt_setuid_rpc_client_la-virnetmessage.lo In file included from rpc/virnetmessage.h:24:0, from rpc/virnetmessage.c:26: rpc/virnetprotocol.h:9:21: fatal error: rpc/rpc.h: No such file or directory CC lxc/libvirt_lxc-lxc_monitor_protocol.o In file included from lxc/lxc_monitor_protocol.c:7:0: lxc/lxc_monitor_protocol.h:9:21: fatal error: rpc/rpc.h: No such file or directory Signed-off-by: Pavel Fedin <p.fedin(a)samsung.com> --- src/Makefile.am | 3 +++ 1 file changed, 3 insertions(+) diff --git a/src/Makefile.am b/src/Makefile.am index 579421d..0d1f58b 100644 --- a/src/Makefile.am +++ b/src/Makefile.am @@ -1290,6 +1290,7 @@ endif ! WITH_DRIVER_MODULES libvirt_driver_lxc_impl_la_CFLAGS = \ $(LIBNL_CFLAGS) \ $(FUSE_CFLAGS) \ + $(XDR_CFLAGS) \ -I$(srcdir)/access \ -I$(srcdir)/conf \ $(AM_CFLAGS) @@ -2216,6 +2217,7 @@ libvirt_setuid_rpc_client_la_CFLAGS = \ -I$(srcdir)/rpc \ $(AM_CFLAGS) \ $(SECDRIVER_CFLAGS) \ + $(XDR_CFLAGS) \ $(NULL) endif WITH_LXC @@ -2665,6 +2667,7 @@ libvirt_lxc_CFLAGS = \ $(LIBNL_CFLAGS) \ $(FUSE_CFLAGS) \ $(DBUS_CFLAGS) \ + $(XDR_CFLAGS) \ $(NULL) if WITH_BLKID libvirt_lxc_CFLAGS += $(BLKID_CFLAGS) -- 1.9.5.msysgit.0

2 1

[libvirt] [PATCH 00/13] Admin API
by Martin Kletzander 23 May '15

23 May '15

This is a working part of the administration API with some usability helpers. We're still missing documentation and proper client (for which virsh needs to be split), but apart from the last three patches, this series is ready to go in if we decide that the documentation and client work can be done later on. Martin Kletzander (13): util: add virJSONValueCopy Move daemon-related parts of virNetServer to virNetDaemon Teach gendispatch how to handle admin dispatching files Add admin protocol Build client headers for admin protocol Add admin error domain Add libvirt-admin library Add XML files with admin API specification Add configuration options for permissions on daemon's admin socket Add support for admin API in libvirt daemon rpc: Add virNetServerGetNClients admin: Add virAdmHello function Example virt-admin .gitignore | 5 + Makefile.am | 4 +- cfg.mk | 9 +- configure.ac | 11 +- daemon/Makefile.am | 34 +- daemon/admin_server.c | 139 +++++++ daemon/admin_server.h | 36 ++ daemon/libvirtd-config.c | 5 +- daemon/libvirtd-config.h | 1 + daemon/libvirtd.aug | 1 + daemon/libvirtd.c | 205 +++++++--- daemon/libvirtd.conf | 8 + daemon/libvirtd.h | 14 +- daemon/test_libvirtd.aug.in | 1 + docs/Makefile.am | 23 +- docs/apibuild.py | 10 +- docs/internals.html.in | 4 +- docs/internals/rpc.html.in | 7 + include/libvirt/Makefile.am | 4 +- include/libvirt/libvirt-admin.h | 63 ++++ include/libvirt/virterror.h | 3 +- libvirt-admin.pc.in | 13 + libvirt.spec.in | 10 + po/POTFILES.in | 4 + src/Makefile.am | 127 ++++++- src/admin/admin_protocol.x | 78 ++++ src/admin_protocol-structs | 18 + src/datatypes.c | 30 ++ src/datatypes.h | 37 ++ src/internal.h | 1 + src/libvirt-admin.c | 412 ++++++++++++++++++++ src/libvirt_admin.syms | 19 + src/libvirt_private.syms | 1 + src/libvirt_remote.syms | 29 +- src/locking/lock_daemon.c | 63 ++-- src/locking/lock_daemon_config.c | 2 +- src/locking/lock_daemon_dispatch.c | 4 +- src/lxc/lxc_controller.c | 65 ++-- src/rpc/gendispatch.pl | 130 ++++--- src/rpc/virnetdaemon.c | 746 +++++++++++++++++++++++++++++++++++++ src/rpc/virnetdaemon.h | 82 ++++ src/rpc/virnetserver.c | 536 ++++---------------------- src/rpc/virnetserver.h | 46 +-- src/rpc/virnetserverprogram.h | 3 + src/util/virerror.c | 1 + src/util/virjson.c | 65 +++- src/util/virjson.h | 4 +- tests/confdata/libvirtd.conf | 6 + tests/confdata/libvirtd.out | 5 + tests/jsontest.c | 111 ++++++ tools/virt-admin/Makefile.am | 70 ++++ tools/virt-admin/virt-admin.c | 72 ++++ tools/virt-admin/virt-admin.pod | 43 +++ 53 files changed, 2735 insertions(+), 685 deletions(-) create mode 100644 daemon/admin_server.c create mode 100644 daemon/admin_server.h create mode 100644 include/libvirt/libvirt-admin.h create mode 100644 libvirt-admin.pc.in create mode 100644 src/admin/admin_protocol.x create mode 100644 src/admin_protocol-structs create mode 100644 src/libvirt-admin.c create mode 100644 src/libvirt_admin.syms create mode 100644 src/rpc/virnetdaemon.c create mode 100644 src/rpc/virnetdaemon.h create mode 100644 tools/virt-admin/Makefile.am create mode 100644 tools/virt-admin/virt-admin.c create mode 100644 tools/virt-admin/virt-admin.pod -- 2.4.0

3 39

[libvirt] [PATCHv2] netdev: fail when setting up an SRIOV VF if PF is offline
by Laine Stump 23 May '15

23 May '15

If an SRIOV PF is offline, the kernel won't complain if you set the mac address and vlan tag for a VF via this PF, and it will even let you assign the VF to a guest using PCI device assignment or macvtap passthrough. But in this case (the PF isn't online), the device won't be usable in the guest. Silently setting the PF online would solve the connectivity problem, but as pointed out by Dan Berrange, when an interface is set online with no associated config, the kernel will by default turn on IPv6 autoconf, which could create unexpected security problems for the host. For this reason, this patch instead logs an error and fails the operation. This resolves: https://bugzilla.redhat.com/show_bug.cgi?id=893738 Originally filed against RHEL6, but present in every version of libvirt until today. --- src/util/virnetdev.c | 22 ++++++++++++++++++++++ 1 file changed, 22 insertions(+) diff --git a/src/util/virnetdev.c b/src/util/virnetdev.c index e14b401..d0580a0 100644 --- a/src/util/virnetdev.c +++ b/src/util/virnetdev.c @@ -2258,6 +2258,28 @@ virNetDevReplaceVfConfig(const char *pflinkdev, int vf, char macstr[VIR_MAC_STRING_BUFLEN]; char *fileData = NULL; int ifindex = -1; + bool pfIsOnline; + + /* Assure that PF is online prior to twiddling with the VF. It + * *should* be, but if the PF isn't online the changes made to the + * VF via the PF won't take effect, yet there will be no error + * reported. In the case that it isn't online, fail and report the + * error, since setting an unconfigured interface online + * automatically turns on IPv6 autoconfig, which may not be what + * the admin expects, so we want them to explicitly enable the PF + * in the host system network config. + */ + if (virNetDevGetOnline(pflinkdev, &pfIsOnline) < 0) + goto cleanup; + if (!pfIsOnline) { + virReportError(VIR_ERR_INTERNAL_ERROR, + "Unable to configure VF %d of PF '%s' " + "because the PF is not online. Please " + "change host network config to put the " + "PF online.", + vf, pflinkdev); + goto cleanup; + } if (virNetDevGetVfConfig(pflinkdev, vf, &oldmac, &oldvlanid) < 0) goto cleanup; -- 2.1.0

2 1

[libvirt] Plans for next release
by Daniel Veillard 23 May '15

23 May '15

if we want to get it by next month, we should probably freeze on Tuesday for an 1.2.16 on June 1st, we 'only' have 137 commits since 1.2.15 but sticking to the monthly release is important. I hope this works for everybody, thanks, Daniel -- Daniel Veillard | Open Source and Standards, Red Hat veillard(a)redhat.com | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ http://veillard.com/ | virtualization library http://libvirt.org/

1 0

[libvirt] official libvirt mirror on github?
by Cole Robinson 22 May '15

22 May '15

Hey all, Anyone considered setting up libvirt*.git mirrors on github? Given the popularity of github these days, IMO it's unfortunate we don't have an official mirror on there. As far as the actual mirroring though, we'd probably need to set up hooks on libvirt.org to push new commits up to github, there doesn't appear to be any better way than that. Thoughts? - Cole

9 18

[libvirt] [PATCHv2] util: make it more robust to calculate timeout value
by Wang Yufei 22 May '15

22 May '15

From: Zhang Bo <oscar.zhangbo(a)huawei.com> When we change system clock to years ago, a certain CPU may use up 100% cputime. The reason is that in function virEventPollCalculateTimeout(), we assign the unsigned long long result to an INT variable, *timeout = then - now; // timeout is INT, and then/now are long long if (*timeout < 0) *timeout = 0; there's a chance that variable @then minus variable @now may be a very large number that overflows INT value expression, then *timeout will be negative and be assigned to 0. Next the 'poll' in function virEventPollRunOnce() will get into an 'endless' while loop there. thus, the cpu that virEventPollRunOnce() thread runs on will go up to 100%. Although as we discussed before in https://www.redhat.com/archives/libvir-list/2015-May/msg00400.html it should be prohibited to set-time while other applications are running, but it does seems to have no harm to make the codes more robust. Signed-off-by: Wang Yufei <james.wangyufei(a)huawei.com> Signed-off-by: Zhang Bo <oscar.zhangbo(a)huawei.com> --- src/util/vireventpoll.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/src/util/vireventpoll.c b/src/util/vireventpoll.c index ffda206..9c9e7af 100644 --- a/src/util/vireventpoll.c +++ b/src/util/vireventpoll.c @@ -357,9 +357,12 @@ static int virEventPollCalculateTimeout(int *timeout) return -1; EVENT_DEBUG("Schedule timeout then=%llu now=%llu", then, now); - *timeout = then - now; - if (*timeout < 0) + if (then <= now) { *timeout = 0; + } else { + *timeout = (int) (then - now); + *timeout = (*timeout > 0) ? (*timeout) : (*timeout)*(-1); + } } else { *timeout = -1; } -- 1.7.12.4

2 1

[libvirt] [PATCH] Fix to list online cpus using virsh capabilities
by Kothapally Madhu Pavan 22 May '15

22 May '15

Virsh capabilities will list offline cpus as online when libvirt is compiled with numactl option disabled. This fix will list correct set of online cpus. Signed-off-by: Kothapally Madhu Pavan <kmp(a)linux.vnet.ibm.com> --- src/nodeinfo.c | 36 ++++++++++++++++++++++++------------ 1 file changed, 24 insertions(+), 12 deletions(-) diff --git a/src/nodeinfo.c b/src/nodeinfo.c index 22df95c..602c76c 100644 --- a/src/nodeinfo.c +++ b/src/nodeinfo.c @@ -1639,28 +1639,38 @@ nodeCapsInitNUMAFake(virCapsPtr caps ATTRIBUTE_UNUSED) { virNodeInfo nodeinfo; virCapsHostNUMACellCPUPtr cpus; - int ncpus; + int ncpus, onlinecpus; int s, c, t; - int id; + int id, cid; + char *sysfs_cpudir = NULL; if (nodeGetInfo(&nodeinfo) < 0) return -1; ncpus = VIR_NODEINFO_MAXCPUS(nodeinfo); + onlinecpus = nodeinfo.cpus; + + if (VIR_ALLOC_N(cpus, onlinecpus) < 0) + return -1; - if (VIR_ALLOC_N(cpus, ncpus) < 0) + if (virAsprintf(&sysfs_cpudir, SYSFS_CPU_PATH) < 0) return -1; id = 0; + cid = 0; for (s = 0; s < nodeinfo.sockets; s++) { for (c = 0; c < nodeinfo.cores; c++) { for (t = 0; t < nodeinfo.threads; t++) { - cpus[id].id = id; - cpus[id].socket_id = s; - cpus[id].core_id = c; - if (!(cpus[id].siblings = virBitmapNew(ncpus))) - goto error; - ignore_value(virBitmapSetBit(cpus[id].siblings, id)); + if (virNodeGetCpuValue(sysfs_cpudir, id, "online", 0)) { + cpus[cid].id = id; + cpus[cid].socket_id = s; + cpus[cid].core_id = c; + if (!(cpus[cid].siblings = virBitmapNew(ncpus))) + goto error; + ignore_value(virBitmapSetBit(cpus[cid].siblings, id)); + cid++; + } + id++; } } @@ -1668,17 +1678,19 @@ nodeCapsInitNUMAFake(virCapsPtr caps ATTRIBUTE_UNUSED) if (virCapabilitiesAddHostNUMACell(caps, 0, nodeinfo.memory, - ncpus, cpus, + onlinecpus, cpus, 0, NULL, 0, NULL) < 0) goto error; + VIR_FREE(sysfs_cpudir); return 0; error: - for (; id >= 0; id--) - virBitmapFree(cpus[id].siblings); + for (; cid >= 0; cid--) + virBitmapFree(cpus[cid].siblings); VIR_FREE(cpus); + VIR_FREE(sysfs_cpudir); return -1; }

2 1

[libvirt] [PATCH v5] inspector: recognize ppc64 and ppc64le archs (RHBZ#1211996)
by Maros Zatko 22 May '15

22 May '15

Patch extracts MSB/LSB info from /bin/file output and passes it as separate parameret from elf_arch. Then it is sent to impl_file_architecture and checked/ Modified magic_for_file to handle regex changes. Now with tests for file_architecture, and error reporting for unknown endianness. Fixes: RHBZ#1211996 Maros Zatko (1): inspector: recognize ppc64 and ppc64le archs (RHBZ#1211996) generator/actions.ml | 8 ++++++++ src/filearch.c | 28 +++++++++++++++++++--------- 2 files changed, 27 insertions(+), 9 deletions(-) -- 1.9.3

2 3