- Devel - libvirt List Archives

[libvirt] [PATCHv2 variant 1] snapshot: implement new APIs for esx and vbox
by Eric Blake 20 Jun '12

20 Jun '12

The two new APIs are rather trivial; based on bits and pieces of other existing APIs. But rather than blindly return 0 or 1 for HasMetadata, I chose to first validate that the snapshot in question in fact exists. In the process, I noticed other APIs that were blindly succeeding instead of checking for existence. * src/esx/esx_driver.c (esxDomainSnapshotIsCurrent) (esxDomainSnapshotHasMetadata): New functions. (esxDomainIsUpdated, esxDomainIsPersistent): Add existence checks. * src/vbox/vbox_tmpl.c (vboxDomainSnapshotIsCurrent) (vboxDomainSnapshotHasMetadata): New functions. (vboxDomainIsPersistent, vboxDomainIsUpdated): Add existence checks. --- v2: add existence checks to other functions I like this version better. However, while it compiles, I'm completely unable to runtime test it, so I'd appreciate a good review. src/esx/esx_driver.c | 116 ++++++++++++++++++++++++++++++++++++++-- src/vbox/vbox_tmpl.c | 145 ++++++++++++++++++++++++++++++++++++++++++++++++-- 2 files changed, 253 insertions(+), 8 deletions(-) diff --git a/src/esx/esx_driver.c b/src/esx/esx_driver.c index b3f1948..db2144c 100644 --- a/src/esx/esx_driver.c +++ b/src/esx/esx_driver.c @@ -4243,10 +4243,28 @@ esxDomainIsActive(virDomainPtr domain) static int -esxDomainIsPersistent(virDomainPtr domain ATTRIBUTE_UNUSED) +esxDomainIsPersistent(virDomainPtr domain) { - /* ESX has no concept of transient domains, so all of them are persistent */ - return 1; + /* ESX has no concept of transient domains, so all of them are + * persistent. However, we do want to check for existence. */ + int result = -1; + esxPrivate *priv = domain->conn->privateData; + esxVI_ObjectContent *virtualMachine = NULL; + + if (esxVI_EnsureSession(priv->primary) < 0) + return -1; + + if (esxVI_LookupVirtualMachineByUuid(priv->primary, domain->uuid, + NULL, &virtualMachine, + esxVI_Occurrence_RequiredItem) < 0) + goto cleanup; + + result = 1; + +cleanup: + esxVI_ObjectContent_Free(&virtualMachine); + + return result; } @@ -4254,7 +4272,26 @@ esxDomainIsPersistent(virDomainPtr domain ATTRIBUTE_UNUSED) static int esxDomainIsUpdated(virDomainPtr domain ATTRIBUTE_UNUSED) { - return 0; + /* ESX domains never have a persistent state that differs from + * current state. However, we do want to check for existence. */ + int result = -1; + esxPrivate *priv = domain->conn->privateData; + esxVI_ObjectContent *virtualMachine = NULL; + + if (esxVI_EnsureSession(priv->primary) < 0) + return -1; + + if (esxVI_LookupVirtualMachineByUuid(priv->primary, domain->uuid, + NULL, &virtualMachine, + esxVI_Occurrence_RequiredItem) < 0) + goto cleanup; + + result = 0; + +cleanup: + esxVI_ObjectContent_Free(&virtualMachine); + + return result; } @@ -4707,6 +4744,75 @@ esxDomainSnapshotCurrent(virDomainPtr domain, unsigned int flags) } +static int +esxDomainSnapshotIsCurrent(virDomainSnapshotPtr snapshot, unsigned int flags) +{ + esxPrivate *priv = snapshot->domain->conn->privateData; + esxVI_VirtualMachineSnapshotTree *currentSnapshotTree = NULL; + esxVI_VirtualMachineSnapshotTree *rootSnapshotList = NULL; + esxVI_VirtualMachineSnapshotTree *snapshotTree = NULL; + int ret = -1; + + virCheckFlags(0, -1); + + if (esxVI_EnsureSession(priv->primary) < 0) { + return -1; + } + + /* Check that snapshot exists. */ + if (esxVI_LookupRootSnapshotTreeList(priv->primary, snapshot->domain->uuid, + &rootSnapshotList) < 0 || + esxVI_GetSnapshotTreeByName(rootSnapshotList, snapshot->name, + &snapshotTree, NULL, + esxVI_Occurrence_RequiredItem) < 0) { + goto cleanup; + } + + if (esxVI_LookupCurrentSnapshotTree(priv->primary, snapshot->domain->uuid, + &currentSnapshotTree, + esxVI_Occurrence_RequiredItem) < 0) { + goto cleanup; + } + + ret = STREQ(snapshot->name, currentSnapshotTree->name); + +cleanup: + esxVI_VirtualMachineSnapshotTree_Free(&currentSnapshotTree); + esxVI_VirtualMachineSnapshotTree_Free(&rootSnapshotList); + return ret; +} + + +static int +esxDomainSnapshotHasMetadata(virDomainSnapshotPtr snapshot, unsigned int flags) +{ + esxPrivate *priv = snapshot->domain->conn->privateData; + esxVI_VirtualMachineSnapshotTree *rootSnapshotList = NULL; + esxVI_VirtualMachineSnapshotTree *snapshotTree = NULL; + int ret = -1; + + virCheckFlags(0, -1); + + if (esxVI_EnsureSession(priv->primary) < 0) { + return -1; + } + + /* Check that snapshot exists. If so, there is no metadata. */ + if (esxVI_LookupRootSnapshotTreeList(priv->primary, snapshot->domain->uuid, + &rootSnapshotList) < 0 || + esxVI_GetSnapshotTreeByName(rootSnapshotList, snapshot->name, + &snapshotTree, NULL, + esxVI_Occurrence_RequiredItem) < 0) { + goto cleanup; + } + + ret = 0; + +cleanup: + esxVI_VirtualMachineSnapshotTree_Free(&rootSnapshotList); + return ret; +} + static int esxDomainRevertToSnapshot(virDomainSnapshotPtr snapshot, unsigned int flags) @@ -5021,6 +5127,8 @@ static virDriver esxDriver = { .domainSnapshotGetParent = esxDomainSnapshotGetParent, /* 0.9.7 */ .domainSnapshotCurrent = esxDomainSnapshotCurrent, /* 0.8.0 */ .domainRevertToSnapshot = esxDomainRevertToSnapshot, /* 0.8.0 */ + .domainSnapshotIsCurrent = esxDomainSnapshotIsCurrent, /* 0.9.13 */ + .domainSnapshotHasMetadata = esxDomainSnapshotHasMetadata, /* 0.9.13 */ .domainSnapshotDelete = esxDomainSnapshotDelete, /* 0.8.0 */ .isAlive = esxIsAlive, /* 0.9.8 */ }; diff --git a/src/vbox/vbox_tmpl.c b/src/vbox/vbox_tmpl.c index 4b0ee2e..8b7415a 100644 --- a/src/vbox/vbox_tmpl.c +++ b/src/vbox/vbox_tmpl.c @@ -1490,14 +1490,54 @@ static int vboxDomainIsActive(virDomainPtr dom) { } -static int vboxDomainIsPersistent(virDomainPtr dom ATTRIBUTE_UNUSED) { - /* All domains are persistent. */ - return 1; +static int vboxDomainIsPersistent(virDomainPtr dom ATTRIBUTE_UNUSED) +{ + /* All domains are persistent. However, we do want to check for + * existence. */ + VBOX_OBJECT_CHECK(dom->conn, int, -1); + vboxIID iid = VBOX_IID_INITIALIZER; + IMachine *machine = NULL; + nsresult rc; + + vboxIIDFromUUID(&iid, dom->uuid); + rc = VBOX_OBJECT_GET_MACHINE(iid.value, &machine); + if (NS_FAILED(rc)) { + vboxError(VIR_ERR_NO_DOMAIN, "%s", + _("no domain with matching UUID")); + goto cleanup; + } + + ret = 1; + +cleanup: + VBOX_RELEASE(machine); + vboxIIDUnalloc(&iid); + return ret; } static int vboxDomainIsUpdated(virDomainPtr dom ATTRIBUTE_UNUSED) { - return 0; + /* VBox domains never have a persistent state that differs from + * current state. However, we do want to check for existence. */ + VBOX_OBJECT_CHECK(dom->conn, int, -1); + vboxIID iid = VBOX_IID_INITIALIZER; + IMachine *machine = NULL; + nsresult rc; + + vboxIIDFromUUID(&iid, dom->uuid); + rc = VBOX_OBJECT_GET_MACHINE(iid.value, &machine); + if (NS_FAILED(rc)) { + vboxError(VIR_ERR_NO_DOMAIN, "%s", + _("no domain with matching UUID")); + goto cleanup; + } + + ret = 0; + +cleanup: + VBOX_RELEASE(machine); + vboxIIDUnalloc(&iid); + return ret; } static int vboxDomainSuspend(virDomainPtr dom) { @@ -6426,6 +6466,101 @@ cleanup: return ret; } +static int +vboxDomainSnapshotIsCurrent(virDomainSnapshotPtr snapshot, + unsigned int flags) +{ + virDomainPtr dom = snapshot->domain; + VBOX_OBJECT_CHECK(dom->conn, int, -1); + vboxIID iid = VBOX_IID_INITIALIZER; + IMachine *machine = NULL; + ISnapshot *snap = NULL; + ISnapshot *current = NULL; + PRUnichar *nameUtf16 = NULL; + char *name = NULL; + nsresult rc; + + virCheckFlags(0, -1); + + vboxIIDFromUUID(&iid, dom->uuid); + rc = VBOX_OBJECT_GET_MACHINE(iid.value, &machine); + if (NS_FAILED(rc)) { + vboxError(VIR_ERR_NO_DOMAIN, "%s", + _("no domain with matching UUID")); + goto cleanup; + } + + if (!(snap = vboxDomainSnapshotGet(data, dom, machine, snapshot->name))) + goto cleanup; + + rc = machine->vtbl->GetCurrentSnapshot(machine, &current); + if (NS_FAILED(rc)) { + vboxError(VIR_ERR_INTERNAL_ERROR, "%s", + _("could not get current snapshot")); + goto cleanup; + } + if (!current) { + ret = 0; + goto cleanup; + } + + rc = current->vtbl->GetName(current, &nameUtf16); + if (NS_FAILED(rc) || !nameUtf16) { + vboxError(VIR_ERR_INTERNAL_ERROR, "%s", + _("could not get current snapshot name")); + goto cleanup; + } + + VBOX_UTF16_TO_UTF8(nameUtf16, &name); + if (!name) { + virReportOOMError(); + goto cleanup; + } + + ret = STREQ(snapshot->name, name); + +cleanup: + VBOX_UTF8_FREE(name); + VBOX_UTF16_FREE(nameUtf16); + VBOX_RELEASE(current); + VBOX_RELEASE(machine); + vboxIIDUnalloc(&iid); + return ret; +} + +static int +vboxDomainSnapshotHasMetadata(virDomainSnapshotPtr snapshot, + unsigned int flags) +{ + virDomainPtr dom = snapshot->domain; + VBOX_OBJECT_CHECK(dom->conn, int, -1); + vboxIID iid = VBOX_IID_INITIALIZER; + IMachine *machine = NULL; + ISnapshot *snap = NULL; + nsresult rc; + + virCheckFlags(0, -1); + + vboxIIDFromUUID(&iid, dom->uuid); + rc = VBOX_OBJECT_GET_MACHINE(iid.value, &machine); + if (NS_FAILED(rc)) { + vboxError(VIR_ERR_NO_DOMAIN, "%s", + _("no domain with matching UUID")); + goto cleanup; + } + + /* Check that snapshot exists. If so, there is no metadata. */ + if (!(snap = vboxDomainSnapshotGet(data, dom, machine, snapshot->name))) + goto cleanup; + + ret = 0; + +cleanup: + VBOX_RELEASE(machine); + vboxIIDUnalloc(&iid); + return ret; +} + #if VBOX_API_VERSION < 3001 static int vboxDomainSnapshotRestore(virDomainPtr dom, @@ -9172,6 +9307,8 @@ virDriver NAME(Driver) = { .domainHasCurrentSnapshot = vboxDomainHasCurrentSnapshot, /* 0.8.0 */ .domainSnapshotGetParent = vboxDomainSnapshotGetParent, /* 0.9.7 */ .domainSnapshotCurrent = vboxDomainSnapshotCurrent, /* 0.8.0 */ + .domainSnapshotIsCurrent = vboxDomainSnapshotIsCurrent, /* 0.9.13 */ + .domainSnapshotHasMetadata = vboxDomainSnapshotHasMetadata, /* 0.9.13 */ .domainRevertToSnapshot = vboxDomainRevertToSnapshot, /* 0.8.0 */ .domainSnapshotDelete = vboxDomainSnapshotDelete, /* 0.8.0 */ .isAlive = vboxIsAlive, /* 0.9.8 */ -- 1.7.10.2

3 5

[libvirt] ANNOUNCE: libvirt 0.9.11.4 maintenance release
by Cole Robinson 20 Jun '12

20 Jun '12

libvirt 0.9.11.4 maintenance release is now available. This is libvirt 0.9.11 with additional bugfixes that have accumulated upstream since the initial release. This release can be downloaded at: http://libvirt.org/sources/stable_updates/libvirt-0.9.11.4.tar.gz List of changes since libvirt-0.9.11.3: * Stable release 0.9.11.4 * build: fix build of fresh checkout * build: fix 'make distcheck' issues * build: fix 'make dist' on virgin checkout * Improve error message diagnosing incorrect XML CPU mode * qemu: Enable disconnecting SPICE clients without changing password * qemu: Fix off-by-one error while unescaping monitor strings * virsh: Don't generate invalid XML in attach-disk command * Fix typo in RPM specfile * python: fix snapshot listing bugs * python: use simpler methods * qemu: Don't overwrite security labels * qemuProcessStop: Switch to flags * docs: minor fixes to domain interface documentation * docs: link to FLOSS Weekly podcast, virt blogs * Two RPM conditional fixes for RHEL-7 * snapshot: avoid virsh crash with older servers * Update AUTHORS * LXC: fix memory leak in lxcContainerMountFSBlockAuto * LXC: fix incorrect parameter of mount in lxcContainerMountFSBind * Only check for cluster fs if we're using a filesystem * Fix missing ) in 2 strings * Assign correct address type to spapr-vlan and spapr-vty. * maint: make it easier to copy FORTIFY_SOURCE snippet * command: avoid potential deadlock on handshake * spec: Build against systemd for udev * virsh: Back out if the argument for vol-create-as is malformed (cherry picked from commit ee58b581c4b275f06904253285b7ad562dc09745) * virsh: Accept UUID as an argument for net-info and net-start (cherry picked from commit 68fcfdb8bd33ef323c6c4c5b9a92b1a44829eb6a) * virsh: Accept UUID as an argument for storage commands * Fix for parallel port passthrough for QEMU * maint: command.c whitespace cleanups * command: avoid deadlock on EPIPE situation * build: allow building with newer glibc-headers and -O0 * command: Fix debug message during handshake * Fix sync issue in virNetClientStreamEventRemoveCallback * qemu: fix netdev alias name assignment wrt type='hostdev' * tools: make virt-pki-validate work with acls and xattrs * qemu: avoid closing fd more than once * command: check for fork error before closing fd * fdstream: avoid double close bug * command: avoid double close bugs * avoid fd leak * avoid closing uninitialized fd * build: silence warning from autoconf * virCommand: Extend debug message for handshake * lxc: return correct number of CPUs * examples: add consolecallback example python script * leak_fix.diff * docs: typo in acceleration element * Re-order config options in qemu driver augeas lens * Fix mistakes in augeas lens * Standardize whitespace used in example config files * Fix check for socket existance / daemon spawn * Remove last usage of PATH_MAX and ban its future use * maint: avoid new automake warning about AM_PROG_CC_STDC * Improve docs about compiling libvirt from GIT * tests: run valgrind on real executables, not libtool wrappers * qemu augeas: Add spice_tls/spice_tls_x509_cert_dir * tests: back to short test names * Add parsing for VIR_ENUM_IMPL & VIR_ENUM_DECL in apibuild.py * Add stub impl of virNetlinkEventServiceLocalPid for Win32 * Fix dep from libvirt-lock-sanlock RPM * Remove more bogus systemd service dependencies * Revert "rpc: Discard non-blocking calls only when necessary" * qemu_hotplug: Don't free the PCI device structure after hot-unplug * build: fix unused variable after last patch * Fix potential events deadlock when unref'ing virConnectPtr * Fix build when configuring with polkit0 * build: fix virnetlink on glibc 2.11 * qemu: Don't delete USB device on failed qemuPrepareHostdevUSBDevices * qemu: Rollback on used USB devices * Reject any non-option command line arguments * Remove bogus udev.target dep from libvirtd unit * Set a sensible default master start port for ehci companion controllers * Fix logic for assigning PCI addresses to USB2 companion controllers * Fix virDomainDeviceInfoIsSet() to check all struct fields * Allow stack traces to be included with log messages * Add bundled(gnulib) to RPM specfile * libvirt-guests: Remove LISTFILE if it's empty when stopping service * qemu: Use the CPU index in capabilities to map NUMA node to cpu list. * Assign spapr-vio bus address to ibmvscsi controller * esx: Fix memory leaks in error paths related to transferred ownership * qemu: Don't skip detection of virtual cpu's on non KVM targets * qemu: Re-detect virtual cpu threads after cpu hot (un)plug. * qemu: Refactor qemuDomainSetVcpusFlags * usb: fix crash when failing to attach a second usb device * docs: mention migration issue of which credentials are used * build: Fix the typo in configure.ac * qemu: fix build when !HAVE_NUMACTL * Report error when parsing character device target type * numad: Update comments in libvirt.spec.in * numad: Check numactl-devel if compiled with numad support * snapshot: allow block devices past cgroup * tests: add some self-documentation to tests * build: avoid link failure on Windows * virsh: avoid heap corruption leading to virsh abort * util: set src_pid for virNetlinkCommand when appropriate * util: function to get local nl_pid used by netlink event socket * util: allow specifying both src and dst pid in virNetlinkCommand * util: fix libvirtd startup failure due to netlink error * qemu: call usb search function for hostdev initialization and hotplug * usb: create functions to search usb device accurately * rpm: Handle different source URLs for maint releases (cherry picked from commit f4345ac21fead319a22a5761e86a46752df23402) * qemu: Emit compatible XML when migrating a domain * qemu: Don't use virDomainDefFormat* directly * qemu: reject blockiotune if qemu too old * qemu: don't modify domain on failed blockiotune * util: remove error log from stubs of virNetlinkEventServiceStart|Stop * node_device: fix possible non-terminated string * uuid: fix possible non-terminated string * tests: fix resource leak * qemu: fix resource leak * vmx: fix resource leak * Coverity: Fix resource leak in virnetlink.c (cherry picked from commit fd2b41574e05510ddeffbf9acbf06584acb3c2b2) * Coverity: Fix resource leak in nodeinfo.c * Coverity: Fix resource leak in test driver * Coverity: Fix resource leak in xen driver * Coverity: Fix resource leaks in phyp driver * Coverity: Fix the forward_null error in Python binding codes * build: fix build on cygwin * Correct indent errors in the function qemuDomainNetsRestart * build: update pid_t type static check * build: fix output of pid values * virsh: make -h always give help * build: make ATTRIBUTE_NONNULL() a NOP unless STATIC_ANALYSIS is on * Make lxcContainerSetStdio the last thing to be called in container startup * Ensure logging is initialized early in libvirt_lxc * Ensure LXC security driver is set unconditonally * Ensure libvirt_lxc process loads the live XML config * maint: avoid false positives on unmarked diagnostics * qemu: allow snapshotting of sheepdog and rbd disks * qemu: change rbd auth_supported separation character to ; * util: Avoid libvirtd crash in virNetDevTapCreate * storage: Break out the loop if duplicate pool is found * qemu: Make sure qemu can access its directory in hugetlbfs * qemu_agent: Report error class at least * More coverity findings addressed * lxc: Fix coverity findings * build: fix stamp file name * Revert "building: remove libvirt_dbus.syms from EXTRA_DIST" For info about past maintenance releases, see: http://wiki.libvirt.org/page/Maintenance_Releases Thanks, Cole

2 1

[libvirt] [PATCH v3 0/5] file descriptor passing using pass-fd
by Corey Bryant 20 Jun '12

20 Jun '12

libvirt's sVirt security driver provides SELinux MAC isolation for Qemu guest processes and their corresponding image files. In other words, sVirt uses SELinux to prevent a QEMU process from opening files that do not belong to it. sVirt provides this support by labeling guests and resources with security labels that are stored in file system extended attributes. Some file systems, such as NFS, do not support the extended attribute security namespace, and therefore cannot support sVirt isolation. A solution to this problem is to provide fd passing support, where libvirt opens files and passes file descriptors to QEMU. This, along with SELinux policy to prevent QEMU from opening files, can provide image file isolation for NFS files stored on the same NFS mount. This patch series adds the pass-fd QMP monitor command, which allows an fd to be passed via SCM_RIGHTS, and returns the received file descriptor. Support is also added to the block layer to allow QEMU to dup the fd when the filename is of the /dev/fd/X format. This is useful if MAC policy prevents QEMU from opening specific types of files. One nice thing about this approach is that no new SELinux policy is required to prevent open of NFS files (files with type nfs_t). The virt_use_nfs boolean type simply needs to be set to false, and open will be prevented (and dup will be allowed). For example: # setsebool virt_use_nfs 0 # getsebool virt_use_nfs virt_use_nfs --> off Corey Bryant (5): qapi: Convert getfd and closefd qapi: Add pass-fd QMP command osdep: Enable qemu_open to dup pre-opened fd block: Convert open calls to qemu_open block: Prevent /dev/fd/X filename from being detected as floppy block/raw-posix.c | 22 ++++++++++--------- block/raw-win32.c | 4 ++-- block/vdi.c | 5 +++-- block/vmdk.c | 21 ++++++++---------- block/vpc.c | 2 +- block/vvfat.c | 21 +++++++++--------- hmp-commands.hx | 6 ++---- hmp.c | 18 ++++++++++++++++ hmp.h | 2 ++ monitor.c | 61 +++++++++++++++++++++++++++++++++++++++-------------- osdep.c | 13 ++++++++++++ qapi-schema.json | 54 +++++++++++++++++++++++++++++++++++++++++++++++ qmp-commands.hx | 48 +++++++++++++++++++++++++++++++++++++---- 13 files changed, 216 insertions(+), 61 deletions(-) -- 1.7.10.2

5 35

[libvirt] [PATCH] Add support for RAM filesystems for LXC
by Daniel P. Berrange 20 Jun '12

20 Jun '12

From: "Daniel P. Berrange" <berrange(a)redhat.com> Introduce a new syntax for filesystems to allow use of a RAM filesystem <filesystem type='ram'> <source usage='1024'/> <target dir='/mnt'/> </filesystem> The usasge is in KB to limit consumption of host memory. * docs/formatdomain.html.in: Document new syntax * docs/schemas/domaincommon.rng: Add new attributes * src/conf/domain_conf.c: Parsing/formatting of RAM filesystems * src/lxc/lxc_container.c: Mounting of RAM filesystems Signed-off-by: Daniel P. Berrange <berrange(a)redhat.com> --- docs/formatdomain.html.in | 9 +++++++- docs/schemas/domaincommon.rng | 13 +++++++++++ src/conf/domain_conf.c | 43 ++++++++++++++++++++---------------- src/lxc/lxc_container.c | 49 +++++++++++++++++++++++++++++++++++++++++ 4 files changed, 94 insertions(+), 20 deletions(-) diff --git a/docs/formatdomain.html.in b/docs/formatdomain.html.in index e1fe0c4..9d1e02b 100644 --- a/docs/formatdomain.html.in +++ b/docs/formatdomain.html.in @@ -1617,6 +1617,12 @@ format will be autodetected. Only used by LXC driver <span class="since">(since 0.9.5)</span>. </dd> + <dt><code>type='ram'</code></dt> + <dd> + An in-memory filesystem, using memory from the host OS. + The source element has a single attribute <code>usage</code> + which gives the memory usage limit in kilobytes. + <span class="since"> (since 0.9.13)</span></dd> </dl> The filesystem block has an optional attribute <code>accessmode</code> @@ -1656,7 +1662,8 @@ The resource on the host that is being accessed in the guest. The <code>name</code> attribute must be used with <code>type='template'</code>, and the <code>dir</code> attribute must - be used with <code>type='mount'</code> + be used with <code>type='mount'</code>. The <code>usage</code> attribute + is used with <code>type='ram'</code> to set the memory limit in KB. </dd> <dt><code>target</code></dt> diff --git a/docs/schemas/domaincommon.rng b/docs/schemas/domaincommon.rng index 8419ccc..884680a 100644 --- a/docs/schemas/domaincommon.rng +++ b/docs/schemas/domaincommon.rng @@ -1271,6 +1271,19 @@ </element> </interleave> </group> + <group> + <attribute name="type"> + <value>ram</value> + </attribute> + <interleave> + <element name="source"> + <attribute name="usage"> + <ref name="unsignedLong"/> + </attribute> + <empty/> + </element> + </interleave> + </group> </choice> <interleave> <element name="target"> diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index c82971a..5d692d6 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -4177,7 +4177,8 @@ virDomainFSDefParseXML(xmlNodePtr node, def->wrpolicy = VIR_DOMAIN_FS_WRPOLICY_DEFAULT; } - if (source == NULL) { + if (source == NULL && + def->type != VIR_DOMAIN_FS_TYPE_RAM) { virDomainReportError(VIR_ERR_NO_SOURCE, target ? "%s" : NULL, target); goto error; @@ -11086,27 +11087,31 @@ virDomainFSDefFormat(virBufferPtr buf, virBufferAddLit(buf, "/>\n"); } - if (def->src) { - switch (def->type) { - case VIR_DOMAIN_FS_TYPE_MOUNT: - virBufferEscapeString(buf, " <source dir='%s'/>\n", - def->src); - break; + switch (def->type) { + case VIR_DOMAIN_FS_TYPE_MOUNT: + virBufferEscapeString(buf, " <source dir='%s'/>\n", + def->src); + break; - case VIR_DOMAIN_FS_TYPE_BLOCK: - virBufferEscapeString(buf, " <source dev='%s'/>\n", - def->src); - break; + case VIR_DOMAIN_FS_TYPE_BLOCK: + virBufferEscapeString(buf, " <source dev='%s'/>\n", + def->src); + break; - case VIR_DOMAIN_FS_TYPE_FILE: - virBufferEscapeString(buf, " <source file='%s'/>\n", - def->src); - break; + case VIR_DOMAIN_FS_TYPE_FILE: + virBufferEscapeString(buf, " <source file='%s'/>\n", + def->src); + break; - case VIR_DOMAIN_FS_TYPE_TEMPLATE: - virBufferEscapeString(buf, " <source name='%s'/>\n", - def->src); - } + case VIR_DOMAIN_FS_TYPE_TEMPLATE: + virBufferEscapeString(buf, " <source name='%s'/>\n", + def->src); + break; + + case VIR_DOMAIN_FS_TYPE_RAM: + virBufferAsprintf(buf, " <source usage='%lld'/>\n", + def->usage); + break; } virBufferEscapeString(buf, " <target dir='%s'/>\n", diff --git a/src/lxc/lxc_container.c b/src/lxc/lxc_container.c index 0636eab..f35b8f9 100644 --- a/src/lxc/lxc_container.c +++ b/src/lxc/lxc_container.c @@ -338,6 +338,8 @@ static int lxcContainerPivotRoot(virDomainFSDefPtr root) ret = -1; + VIR_DEBUG("Pivot via %s", root->src); + /* root->parent must be private, so make / private. */ if (mount("", "/", NULL, MS_PRIVATE|MS_REC, NULL) < 0) { virReportSystemError(errno, "%s", @@ -995,6 +997,47 @@ cleanup: } +static int lxcContainerMountFSTmpfs(virDomainFSDefPtr fs) +{ + int ret = -1; + char *data = NULL; + + if (virAsprintf(&data, "size=%lldk", fs->usage) < 0) { + virReportOOMError(); + goto cleanup; + } + + if (virFileMakePath(fs->dst) < 0) { + virReportSystemError(errno, + _("Failed to create %s"), + fs->dst); + goto cleanup; + } + + if (mount("tmpfs", fs->dst, "tmpfs", 0, data) < 0) { + virReportSystemError(errno, + _("Failed to mount directory %s as tmpfs"), + fs->dst); + goto cleanup; + } + + if (fs->readonly) { + VIR_DEBUG("Binding %s readonly", fs->dst); + if (mount(fs->dst, fs->dst, NULL, MS_BIND|MS_REMOUNT|MS_RDONLY, NULL) < 0) { + virReportSystemError(errno, + _("Failed to make directory %s readonly"), + fs->dst); + } + } + + ret = 0; + +cleanup: + VIR_FREE(data); + return ret; +} + + static int lxcContainerMountFS(virDomainFSDefPtr fs, const char *srcprefix) { @@ -1007,6 +1050,10 @@ static int lxcContainerMountFS(virDomainFSDefPtr fs, if (lxcContainerMountFSBlock(fs, srcprefix) < 0) return -1; break; + case VIR_DOMAIN_FS_TYPE_RAM: + if (lxcContainerMountFSTmpfs(fs) < 0) + return -1; + break; case VIR_DOMAIN_FS_TYPE_FILE: lxcError(VIR_ERR_INTERNAL_ERROR, _("Unexpected filesystem type %s"), @@ -1196,6 +1243,8 @@ static int lxcContainerResolveSymlinks(virDomainDefPtr vmDef) for (i = 0 ; i < vmDef->nfss ; i++) { virDomainFSDefPtr fs = vmDef->fss[i]; + if (!fs->src) + continue; if (virFileResolveAllLinks(fs->src, &newroot) < 0) return -1; -- 1.7.10.1

4 7

[libvirt] [PATCHv3 00/12] Add api for atomic listing of domains
by Peter Krempa 20 Jun '12

20 Jun '12

This is the third respin of the series. I've incorporated suggestions from Eric's review and cleaned up some nits that I found while polishing the series. This respin also contains two new cleanup patches: driver: Clean up driver header to space indentation maint: include ignore-value in internal.h Peter Krempa (12): lib: Add public api to enable atomic listing of guest virsh: add support for virConnectListAllDomains and clean up cmdList python: add API exports for virConnectListAllDomains() remote: implement remote protocol for virConnectListAllDomains() conf: Store managed save image existence in virDomainObj conf: Add helper for listing domains on drivers supporting virDomainObj drivers: Implement virListAllDomains for drivers using virDomainObj vbox: Add support for virConnectListAllDomains() hyperv: Add implementation for virConnectListAllDomains() esx: Add implementation for virConnectListAllDomains() driver: Clean up driver header to space indentation maint: include ignore-value in internal.h daemon/remote.c | 54 +++ include/libvirt/libvirt.h.in | 36 ++- python/generator.py | 1 + python/libvirt-override-api.xml | 12 +- python/libvirt-override-virConnect.py | 12 + python/libvirt-override.c | 50 +++- src/Makefile.am | 8 +- src/conf/domain_audit.c | 1 - src/conf/domain_conf.c | 1 - src/conf/domain_conf.h | 2 + src/conf/virdomainlist.c | 181 ++++++++ src/conf/virdomainlist.h | 66 +++ src/driver.h | 729 +++++++++++++++++---------------- src/esx/esx_driver.c | 194 +++++++++ src/hyperv/hyperv_driver.c | 136 ++++++ src/internal.h | 1 + src/libvirt.c | 124 ++++++- src/libvirt_private.syms | 4 + src/libvirt_public.syms | 1 + src/libxl/libxl_driver.c | 61 +++- src/lxc/lxc_driver.c | 19 + src/network/bridge_driver.c | 1 - src/node_device/node_device_hal.c | 1 - src/openvz/openvz_conf.c | 1 - src/openvz/openvz_driver.c | 19 + src/qemu/qemu_domain.c | 1 - src/qemu/qemu_driver.c | 67 +++- src/qemu/qemu_monitor_json.c | 1 - src/remote/remote_driver.c | 64 +++ src/remote/remote_protocol.x | 14 +- src/remote_protocol-structs | 12 + src/test/test_driver.c | 19 + src/uml/uml_driver.c | 18 + src/util/command.c | 1 - src/util/event_poll.c | 1 - src/util/logging.c | 1 - src/util/memory.c | 1 - src/util/threadpool.c | 1 - src/util/virfile.h | 1 - src/util/virnetdevbandwidth.c | 1 - src/vbox/vbox_tmpl.c | 170 ++++++++ src/vmware/vmware_driver.c | 20 + src/xenapi/xenapi_driver.c | 1 - tests/shunloadtest.c | 1 - tools/virsh.c | 555 +++++++++++++++++-------- tools/virsh.pod | 91 +++-- 46 files changed, 2155 insertions(+), 601 deletions(-) create mode 100644 src/conf/virdomainlist.c create mode 100644 src/conf/virdomainlist.h -- 1.7.3.4

2 32

[libvirt] [sandbox][PATCH] Add module directory prefix selection feature
by Radu Caragea 20 Jun '12

20 Jun '12

Add module directory prefix selection feature This is useful when running as a non-privileged user if we want to boot a custom compiled kernel: we might not have rights to install in /lib/modules/<kernel release> so when compiling the kernel we can use "make modules_install INSTALL_MOD_PATH=/path" which installs in /path/lib/modules/<kernel release>. By setting with gvir_sandbox_config_set_moddirprefix(cfg, "/path") we can now achieve just that.

2 5

[libvirt] [qpid PATCH] doc: fixed search form action
by Martin Kletzander 20 Jun '12

20 Jun '12

Search was giving 404 because of missing {$href_base} in "action" property. Also one whitespace character got removed from end of the line because my before-save-hook. --- Pushed under the 'trivial' rule. doc/architecture.html | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/doc/architecture.html b/doc/architecture.html index a564fa8..5b0f2bd 100644 --- a/doc/architecture.html +++ b/doc/architecture.html @@ -1,12 +1,12 @@ <?xml version="1.0" encoding="ISO-8859-1"?> <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Transitional//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd"> -<html xmlns="http://www.w3.org/1999/xhtml"><head><meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-1" /><link rel="stylesheet" type="text/css" href="http://libvirt.org/qpid/main.css" /><link rel="SHORTCUT ICON" href="/32favicon.png" /><title>Architecture</title><meta name="description" content="libvirt, virtualization, virtualization API, qpid" /></head><body><div id="header"><div id="headerLogo"></div><div id="headerSearch"><form action="search.php" enctype="application/x-www-form-urlencoded" method="get"><div><input id="query" name="query" type="text" size="12" value="" /><input id="submit" name="submit" type="submit" value="Search" /></div></form></div></div><div id="body"><div id="menu"><ul class="l0"><li><div><a href="index.html" class="inactive">Home</a></div></li><li><a href="http://libvirt.org/" class="inactive">libvirt</a></li><li><div><a href="releases.html" class="inactive">Releases</a></div></li><li><div><a href="architecture.html" class="inac! tive">Architecture</a></div></li><li><a href="https://www.redhat.com/mailman/listinfo/libvir-list" class="inactive">Mailing list</a></li></ul></div><div id="content"><h1 class="style1">Architecture</h1><p> +<html xmlns="http://www.w3.org/1999/xhtml"><head><meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-1" /><link rel="stylesheet" type="text/css" href="http://libvirt.org/qpid/main.css" /><link rel="SHORTCUT ICON" href="/32favicon.png" /><title>Architecture</title><meta name="description" content="libvirt, virtualization, virtualization API, qpid" /></head><body><div id="header"><div id="headerLogo"></div><div id="headerSearch"><form action="{$href_base}search.php" enctype="application/x-www-form-urlencoded" method="get"><div><input id="query" name="query" type="text" size="12" value="" /><input id="submit" name="submit" type="submit" value="Search" /></div></form></div></div><div id="body"><div id="menu"><ul class="l0"><li><div><a href="index.html" class="inactive">Home</a></div></li><li><a href="http://libvirt.org/" class="inactive">libvirt</a></li><li><div><a href="releases.html" class="inactive">Releases</a></div></li><li><div><a href="architecture.html"! class="inactive">Architecture</a></div></li><li><a href="https://www.redhat.com/mailman/listinfo/libvir-list" class="inactive">Mailing list</a></li></ul></div><div id="content"><h1 class="style1">Architecture</h1><p> libvirt-qpid is an agent that runs on a given host. It will connect to both the local libvirtd and to a qpid broker. It then queries libvirtd every 5 seconds and maintains objects of various classes that reflect the activity and configuration of libvirtd. </p><p> - The classes include the node (the host), domains, storage pools, + The classes include the node (the host), domains, storage pools, and storage volumes. A complete synchronization takes place each iteration so that any state changes that happen outside of the channel provided by libvirt-qpid will be reflected in the objects -- 1.7.8.6

1 0

[libvirt] NEED HELP on libvirt guest resource control
by Zhimou Peng 20 Jun '12

20 Jun '12

Hi all, I'm from libvirt Section 3 I'm tracking "guest resource control" test from kvm, and i found kvm support more than libvirt. they support these: https://riddler.bne.redhat.com/IEEE_Test_Plans/en-US/RHEL/6.3/html-single/K… guest blkio bandwidth and blkio throttle limitation, guest memory limitation, migrate guest from one memory to another, guest CPU bandwidth limitation, guest freezing, pin guest to certain host CPU, guest net limitation. I want to know what the current libvirt support status and if libvirt will support all these feature in the future? BR zhpeng

1 0

[libvirt] [PATCH] conf: Format numatune XML correctly while placement is none
by Osier Yang 20 Jun '12

20 Jun '12

setNumaParameters tunes the numa setting using cgroup, it's another entry except libnuma/numad for numa tuning. And it doesn't set the placement, and further more, the formating codes doesn't take this into consideration. How to reproduce: conn = libvirt.open(None) dom = conn.lookupByName('linux') param = {'numa_nodeset': '0', 'numa_mode': 1} dom.setNumaParameters(param, 2) % virsh start linux error: Failed to start domain rhel6.3rc error: (domain_definition):8: error parsing attribute name <memory mode='preferred' </numatune> -------------------------------^ --- By the way, I see problems of setNumaParameters too. conn = libvirt.open(None) dom = conn.lookupByName('linux') param = {'numa_mode': 1} dom.setNumaParameters(param, 2) The numa 'mode' will be just ignored, and no 'numatune' XML is formated, as neither 'nodeset' nor 'placement' is specified. I'd think it's right to ignore it when formating, it's meaningless to only specify the 'mode'. However, we might have to fix setNumaParameters to prevent setting the numa mode without nodeset, and error out, as it's really a bad user experience to see the API call succeeded, but the expected XML doesn't show up in the end. --- src/conf/domain_conf.c | 17 ++++++++++------- 1 files changed, 10 insertions(+), 7 deletions(-) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 81c6308..c44d89d 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -12795,23 +12795,26 @@ virDomainDefFormatInternal(virDomainDefPtr def, const char *placement; mode = virDomainNumatuneMemModeTypeToString(def->numatune.memory.mode); - virBufferAsprintf(buf, " <memory mode='%s' ", mode); + virBufferAsprintf(buf, " <memory mode='%s'", mode); - if (def->numatune.memory.placement_mode == - VIR_DOMAIN_NUMATUNE_MEM_PLACEMENT_MODE_STATIC) { + if (def->numatune.memory.nodemask) { nodemask = virDomainCpuSetFormat(def->numatune.memory.nodemask, - VIR_DOMAIN_CPUMASK_LEN); + VIR_DOMAIN_CPUMASK_LEN); if (nodemask == NULL) { virDomainReportError(VIR_ERR_INTERNAL_ERROR, "%s", _("failed to format nodeset for " "NUMA memory tuning")); goto cleanup; } - virBufferAsprintf(buf, "nodeset='%s'/>\n", nodemask); + virBufferAsprintf(buf, " nodeset='%s'/>\n", nodemask); VIR_FREE(nodemask); - } else if (def->numatune.memory.placement_mode) { + } else if (def->numatune.memory.placement_mode == + VIR_DOMAIN_NUMATUNE_MEM_PLACEMENT_MODE_AUTO) { placement = virDomainNumatuneMemPlacementModeTypeToString(def->numatune.memory.placement_mode); - virBufferAsprintf(buf, "placement='%s'/>\n", placement); + virBufferAsprintf(buf, " placement='%s'/>\n", placement); + } else { + /* Should not hit here. */ + virBufferAddLit(buf, "/>\n"); } virBufferAddLit(buf, " </numatune>\n"); } -- 1.7.7.3

2 1

[libvirt] [PATCH] openvz: Fix wordsize on 64 bit architectures
by Guido Günther 20 Jun '12

20 Jun '12

The word size there is 64 bit not 8. --- Came across this while browsing the source. O.k. to apply? -- Guido src/openvz/openvz_conf.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/openvz/openvz_conf.c b/src/openvz/openvz_conf.c index ad4ed74..007f9fe 100644 --- a/src/openvz/openvz_conf.c +++ b/src/openvz/openvz_conf.c @@ -195,7 +195,7 @@ virCapsPtr openvzCapsInit(void) if ((guest = virCapabilitiesAddGuest(caps, "exe", utsname.machine, - sizeof(int) == 4 ? 32 : 8, + sizeof(int) == 4 ? 32 : 64, NULL, NULL, 0, -- 1.7.10.4

3 7

[libvirt] [PATCH 1/2] Fix default USB controller for ppc64
by Dipankar Sarma 20 Jun '12

20 Jun '12

From: Dipankar Sarma <dipankar(a)in.ibm.com> Fix the default usb controller for pseries systems if none specified. Signed-off-by: Dipankar Sarma <dipankar(a)in.ibm.com> --- src/qemu/qemu_command.c | 13 +++++++++---- 1 files changed, 9 insertions(+), 4 deletions(-) diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index a34c707..bd4f96a 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -2605,7 +2605,8 @@ qemuControllerModelUSBToCaps(int model) static int -qemuBuildUSBControllerDevStr(virDomainControllerDefPtr def, +qemuBuildUSBControllerDevStr(virDomainDefPtr domainDef, + virDomainControllerDefPtr def, virBitmapPtr qemuCaps, virBuffer *buf) { @@ -2614,8 +2615,12 @@ qemuBuildUSBControllerDevStr(virDomainControllerDefPtr def, model = def->model; - if (model == -1) - model = VIR_DOMAIN_CONTROLLER_MODEL_USB_PIIX3_UHCI; + if (model == -1) { + if (STREQ(domainDef->os.arch, "ppc64")) + model = VIR_DOMAIN_CONTROLLER_MODEL_USB_PCI_OHCI; + else + model = VIR_DOMAIN_CONTROLLER_MODEL_USB_PIIX3_UHCI; + } smodel = qemuControllerModelUSBTypeToString(model); caps = qemuControllerModelUSBToCaps(model); @@ -2701,7 +2706,7 @@ qemuBuildControllerDevStr(virDomainDefPtr domainDef, break; case VIR_DOMAIN_CONTROLLER_TYPE_USB: - if (qemuBuildUSBControllerDevStr(def, qemuCaps, &buf) == -1) + if (qemuBuildUSBControllerDevStr(domainDef, def, qemuCaps, &buf) == -1) goto error; if (nusbcontroller)

2 4

[libvirt] Fwd: libvirt secret support password or encryption keys?
by Zhimou Peng 20 Jun '12

20 Jun '12

FWD Hi, all PLS help, i'm reviewing man doc of libvirt secrets # man virsh ... SECRET COMMMANDS The following commands manipulate "secrets" (e.g. passwords, passphrases -----> secret-set-value only support base64 "passphrases" and encryption keys). Libvirt can store secrets independently from their use, and other objects (e.g. volumes or domains) can refer to the secrets for encryption or possibly other uses. Secrets are identified using an UUID. See <http://libvirt.org/formatsecret.html> for documentation of the XML format used to represent properties of secrets. ... So, I want to know that if libvirt support other 2 kinds of "secrets". And if so, How to use password and encrytion keys ? If libvirt not support that, i will file a bug of this man doc. BR zhpeng ----- Forwarded Message ----- From: "Zhimou Peng" <zhpeng(a)redhat.com> To: "Daniel Berrange" <berrange(a)redhat.com> Sent: Thursday, June 14, 2012 1:38:09 PM Subject: Fwd: libvirt secret support password or encryption keys? Hi, Daniel I saw your blog:http://berrange.com/posts/2009/12/02/using-qcow2-disk-encryption-with-… and there is only "type=passphrases", so how can i use password or encrytion keys? zhpeng BR ----- Forwarded Message ----- From: "Zhimou Peng" <zhpeng(a)redhat.com> To: "Daniel Berrange" <berrange(a)redhat.com> Cc: "s3-bug-review" <s3-bug-review(a)redhat.com> Sent: Tuesday, June 12, 2012 2:03:05 PM Subject: libvirt secret support password or encryption keys? Hi all, # man virsh ... SECRET COMMMANDS The following commands manipulate "secrets" (e.g. passwords, passphrases -----> secret-set-value only support base64 "passphrases" and encryption keys). Libvirt can store secrets independently from their use, and other objects (e.g. volumes or domains) can refer to the secrets for encryption or possibly other uses. Secrets are identified using an UUID. See <http://libvirt.org/formatsecret.html> for documentation of the XML format used to represent properties of secrets. ... So, I want to know that if libvirt support other 2 kinds of "secrets". And if so, How to use password and encrytion keys ? If libvirt not support that, i will file a bug of this man doc. zhpeng BR

2 2

[libvirt] [PATCHv2 00/13] list all snapshots
by Eric Blake 19 Jun '12

19 Jun '12

This is a v2 of the remaining patches in: https://www.redhat.com/archives/libvir-list/2012-June/msg00273.html https://www.redhat.com/archives/libvir-list/2012-June/msg00284.html It assumes that Peter's list all domain patches have been applied already, at least through patch 6/12: https://www.redhat.com/archives/libvir-list/2012-June/msg00316.html In other words, this series will not apply to current libvirt.git at the time of this mail. And since I developed against v3 of Peter's patches and he has some tweaks to make, I will probably have to make minor tweaks to rebase on top of his eventual v4. I will follow up with a link to a git repo with Peter's v3 and my series, and/or my rebase on top of his v4. Overall changes since v1: actually tested this time address review comments lots of bug fixes due to corner case testing rearrange the patch series so that metaroot comes before filtering pick up some good ideas from Peter's v3 More details in some of the individual patches, if I can remember everything I did. In particular, I'm quite pleased that the overall diffstat for domain_conf.c reduced in size in spite of adding features :) Eric Blake (13): snapshot: new virsh function factored from snapshot-list snapshot: use new virsh function for snapshot-list snapshot: use metaroot node to simplify management snapshot: merge domain and snapshot computation snapshot: merge count and name collectoin snapshot: add additional filters when getting lists snapshot: expose new flags in virsh list: add virDomainListAllSnapshots API list: use the new snapshot API in virsh when possible list: provide python bindings for snapshots list: provide RPC call for snapshots list: new helper function to collect snapshots list: add qemu snapshot list support daemon/remote.c | 126 ++++++ include/libvirt/libvirt.h.in | 35 +- python/Makefile.am | 2 + python/generator.py | 2 + python/libvirt-override-api.xml | 16 +- python/libvirt-override-virDomain.py | 11 + python/libvirt-override-virDomainSnapshot.py | 11 + python/libvirt-override.c | 92 +++++ src/conf/domain_conf.c | 250 ++++------- src/conf/domain_conf.h | 19 +- src/conf/virdomainlist.c | 44 +- src/conf/virdomainlist.h | 18 + src/driver.h | 12 + src/libvirt.c | 275 +++++++++++-- src/libvirt_private.syms | 3 +- src/libvirt_public.syms | 2 + src/qemu/qemu_driver.c | 137 +++++-- src/qemu/qemu_migration.c | 3 +- src/remote/remote_driver.c | 126 ++++++ src/remote/remote_protocol.x | 26 +- src/remote_protocol-structs | 26 ++ tools/virsh.c | 570 +++++++++++++++++--------- tools/virsh.pod | 14 +- 23 files changed, 1371 insertions(+), 449 deletions(-) create mode 100644 python/libvirt-override-virDomain.py create mode 100644 python/libvirt-override-virDomainSnapshot.py -- 1.7.10.2

2 37

[libvirt] [PATCH] domain_conf: fix possible memory leak
by Martin Kletzander 19 Jun '12

19 Jun '12

Until now, it was possible to crash libvirtd when defining domain with channel device with missing source element. When creating new virDomainChrDef, target.port is set to -1, but unfortunately it is an union with addresses that virDomainChrDefFree tries to free in case the deviceType is channel. Having the port set to -1 is intended, however the cleanest way to get around the problems with the crash seems to be renumbering the VIR_DOMAIN_CHR_CHANNEL_ target types to cover new NONE type (with value 0) being the default (no target type yet). --- src/conf/domain_conf.c | 1 + src/conf/domain_conf.h | 3 ++- 2 files changed, 3 insertions(+), 1 deletions(-) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 7bd61a5..41f22d2 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -307,6 +307,7 @@ VIR_ENUM_IMPL(virDomainNetInterfaceLinkState, VIR_DOMAIN_NET_INTERFACE_LINK_STAT VIR_ENUM_IMPL(virDomainChrChannelTarget, VIR_DOMAIN_CHR_CHANNEL_TARGET_TYPE_LAST, + "none", "guestfwd", "virtio") diff --git a/src/conf/domain_conf.h b/src/conf/domain_conf.h index d627ad8..531176c 100644 --- a/src/conf/domain_conf.h +++ b/src/conf/domain_conf.h @@ -855,7 +855,8 @@ enum virDomainChrDeviceType { }; enum virDomainChrChannelTargetType { - VIR_DOMAIN_CHR_CHANNEL_TARGET_TYPE_GUESTFWD = 0, + VIR_DOMAIN_CHR_CHANNEL_TARGET_TYPE_NONE = 0, + VIR_DOMAIN_CHR_CHANNEL_TARGET_TYPE_GUESTFWD, VIR_DOMAIN_CHR_CHANNEL_TARGET_TYPE_VIRTIO, VIR_DOMAIN_CHR_CHANNEL_TARGET_TYPE_LAST, -- 1.7.8.6

2 2

[libvirt] [PATCH] virsh: fix few typos on desc command
by Martin Kletzander 19 Jun '12

19 Jun '12

virsh help fix: - <--title> can also /get/ the title virsh man page: - missing <domain-id> - <new_desc> should be <new-desc> --- tools/virsh.c | 2 +- tools/virsh.pod | 7 ++++--- 2 files changed, 5 insertions(+), 4 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index 4d34d49..036e018 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -1227,7 +1227,7 @@ static const vshCmdOptDef opts_desc[] = { {"live", VSH_OT_BOOL, 0, N_("modify/get running state")}, {"config", VSH_OT_BOOL, 0, N_("modify/get persistent configuration")}, {"current", VSH_OT_BOOL, 0, N_("modify/get current state configuration")}, - {"title", VSH_OT_BOOL, 0, N_("modify the title instead of description")}, + {"title", VSH_OT_BOOL, 0, N_("modify/get the title instead of description")}, {"edit", VSH_OT_BOOL, 0, N_("open an editor to modify the description")}, {"new-desc", VSH_OT_ARGV, 0, N_("message")}, {NULL, 0, 0, NULL} diff --git a/tools/virsh.pod b/tools/virsh.pod index 910a187..252512a 100644 --- a/tools/virsh.pod +++ b/tools/virsh.pod @@ -497,8 +497,9 @@ Define a domain from an XML <file>. The domain definition is registered but not started. If domain is already running, the changes will take effect on the next boot. -=item B<desc> [[I<--live>] [I<--config>] | [I<--current>]] [I<--title>] - [I<--edit>] [I<--new-desc> New description or title message] +=item B<desc> I<domain-id> [[I<--live>] [I<--config>] | + [I<--current>]] [I<--title>] [I<--edit>] [I<--new-desc> + New description or title message] Show or modify description and title of a domain. These values are user fields that allow to store arbitrary textual data to allow easy @@ -516,7 +517,7 @@ description or title should be opened and the contents saved back afterwards. Flag I<--title> selects operation on the title field instead of description. -If neither of I<--edit> and I<--new_desc> are specified the note or description +If neither of I<--edit> and I<--new-desc> are specified the note or description is displayed instead of being modified. =item B<destroy> I<domain-id> [I<--graceful>] -- 1.7.8.6

2 2

[libvirt] [PATCH] error: Fix macro checking for NULL arguments
by Peter Krempa 19 Jun '12

19 Jun '12

Macro virCheckNullArgGoto is supposed to check for NULL argument but checks non-NULL instead. --- src/internal.h | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/src/internal.h b/src/internal.h index 1b1598b..b09609e 100644 --- a/src/internal.h +++ b/src/internal.h @@ -241,21 +241,21 @@ # define virCheckNonNullArgReturn(argname, retval) \ do { \ if (argname == NULL) { \ virReportInvalidNullArg(argname); \ return retval; \ } \ } while (0) # define virCheckNullArgGoto(argname, label) \ do { \ - if (argname == NULL) { \ + if (argname != NULL) { \ virReportInvalidNullArg(argname); \ goto label; \ } \ } while (0) # define virCheckNonNullArgGoto(argname, label) \ do { \ if (argname == NULL) { \ virReportInvalidNonNullArg(argname); \ goto label; \ } \ -- 1.7.8.6

2 4

Re: [libvirt] ask for help to explain the network about tap
by Yong Sheng Gong 19 Jun '12

19 Jun '12

Can any one help? thanks -----Yong Sheng Gong/China/IBM wrote: ----- To: Eric Blake <eblake(a)redhat.com> From: Yong Sheng Gong/China/IBM Date: 06/17/2012 11:11AM Cc: libvir-list(a)redhat.com Subject: Re: [libvirt] ask for help to explain the network about tap Hi Eric, Thanks for the reply. I have a new question. I found I cannot ping the VM with predefined tap device in domain xml from other host(means the machine other than the hyper visor is running on): <interface type='ethernet'> <mac address='f6:19:e0:6c:84:c3'/> <script path=''/> <target dev='tap10'/> <alias name='net0'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x03' function='0x0'/> </interface> But I can ping the VM with libvirtd defined tap device in domain xml from other host: <interface type='bridge'> <mac address='52:54:00:e4:2e:c1'/> <source bridge='br200'/> <target dev='vnet0'/> <model type='virtio'/> <alias name='net0'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x03' function='0x0'/> </interface> [root@robinlinux ~]# brctl show bridge name bridge id STP enabled interfaces br100 8000.0021cc6d4440 no eth0 br200 8000.0021cc6d4443 no eth0.1000 gw200 tap10 vnet0 And I can ping both of VMs from local host. What does libvirtd do to allow the bridge to forward the packet to the tap device attached by target VM? As far as I know all the packets destined to the bridge's ports from outside will be consumed by bridge itself, not xmit to particular port. Thanks Yong Sheng Gong -----Eric Blake <eblake(a)redhat.com> wrote: ----- To: Yong Sheng Gong/China/IBM@IBMCN From: Eric Blake <eblake(a)redhat.com> Date: 06/15/2012 12:36PM Cc: libvir-list(a)redhat.com Subject: Re: [libvirt] ask for help to explain the network about tap On 06/14/2012 08:49 PM, Yong Sheng Gong wrote: > > Hi, > I don't know who helps qemu-kvm to create and open the tap file and when if I define an interface of a kvm vm like: > <interface type='bridge'> > <mac address='52:54:00:e4:2e:c1'/> > <source bridge='br200'/> > <model type='virtio'/> > <address type='pci' domain='0x0000' bus='0x00' slot='0x03' function='0x0'/> > </interface> > > The qemu-kvm process's command line will be: <snip> > > Note the -netdev tap,fd=24 That's libvirtd that opened the tap file, and passed it by inheritance to the new qemu process. Unfortunately, this means that you can't recreate the same qemu process without also opening the tap device yourself the same way that libvirtd would do it. But this approach of fd passing is essential to security, since libvirtd intentionally removes capabilities that prevent qemu from opening the tap itself, so libvirtd must pass in the fd pre-opened. Did you have specific questions about which functions in libvirt source code open the tap device? Also, note that libvirtd will issue an audit log event for every tapfd that it opens and gives to qemu. The fd number is unpredictable, but the fact that an fd was opened as backed by a tap device is pretty easy to follow in the audit logs. -- Eric Blake eblake(a)redhat.com +1-919-301-3266 Libvirt virtualization library http://libvirt.org [attachment "signature.asc" removed by Yong Sheng Gong/China/IBM]

1 0

[libvirt] [PATCH] virsh: make domiftune interface help string consistent
by Martin Kletzander 19 Jun '12

19 Jun '12

Append '(MAC Address)' after the help string of domiftune virsh command as it takes the same type of argument as domif-{get,set}link which have it specified. --- Pushed under 'trivial' rule. tools/virsh.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index 1e1de39..fffa482 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -2019,7 +2019,7 @@ static const vshCmdInfo info_domiftune[] = { static const vshCmdOptDef opts_domiftune[] = { {"domain", VSH_OT_DATA, VSH_OFLAG_REQ, N_("domain name, id or uuid")}, - {"interface", VSH_OT_DATA, VSH_OFLAG_REQ, N_("interface device")}, + {"interface", VSH_OT_DATA, VSH_OFLAG_REQ, N_("interface device (MAC Address)")}, {"inbound", VSH_OT_DATA, VSH_OFLAG_NONE, N_("control domain's incoming traffics")}, {"outbound", VSH_OT_DATA, VSH_OFLAG_NONE, N_("control domain's outgoing traffics")}, {"config", VSH_OT_BOOL, VSH_OFLAG_NONE, N_("affect next boot")}, -- 1.7.8.6

1 0

[libvirt] [sandbox][PATCH] Typo and example fix
by Radu Caragea 19 Jun '12

19 Jun '12

I fixed a typo in the strace debug feature, if you specified LIBVIRT_SANDBOX_STRACE=poll it would write "strace =poll" in the kernel command line and consequently it wouldn't get picked up because of that extra space. Also, the example virt-sandbox.py was a bit outdated so I updated it. The shell.py doesn't work as it has serial1 hardcoded for qemu. I haven't checked the rest.

2 1

[libvirt] [PATCH] add handling EINTR to test example of event loop
by lvroyce＠linux.vnet.ibm.com 19 Jun '12

19 Jun '12

From: lvroyce <lvroyce(a)linux.vnet.ibm.com> some system call and signal will interrupt poll, making event loop stops and fails to react events and keepalive message from libvirt. adding handling EINTR to poll to make it more robust Signed-off-by: lvroyce <lvroyce(a)linux.vnet.ibm.com> --- examples/domain-events/events-python/event-test.py | 8 +++++++- 1 files changed, 7 insertions(+), 1 deletions(-) diff --git a/examples/domain-events/events-python/event-test.py b/examples/domain-events/events-python/event-test.py index 96dc268..b446c21 100644 --- a/examples/domain-events/events-python/event-test.py +++ b/examples/domain-events/events-python/event-test.py @@ -188,7 +188,13 @@ class virEventLoopPure: sleep = (next - now) / 1000.0 debug("Poll with a sleep of %d" % sleep) - events = self.poll.poll(sleep) + try: + events = self.poll.poll(sleep) + except select.error, e: + self.runningPoll = False + if not e.errno in (errno.EINTR, errno.EAGAIN): + raise + return # Dispatch any file handle events that occurred for (fd, revents) in events: -- 1.7.7.6

3 3

[libvirt] [PATCH 02/02] libvirt-java add domain flags and error constants up to 0.9, 2
by Stefan Majer 19 Jun '12

19 Jun '12

-- Stefan Majer

1 0

[libvirt] [PATCH 01/02] libvirt-java add domain flags and error constants up to 0.9, 2
by Stefan Majer 19 Jun '12

19 Jun '12

-- Stefan Majer

1 0

[libvirt] [PATCH v2] LXC: avoid useless duplicate memory free
by Gao feng 19 Jun '12

19 Jun '12

when lxcContainerIdentifyCGroups failed, the memory it allocated has been freed, so we should not free this memory again in lxcContainerSetupPivortRoot and lxcContainerSetupExtraMounts. Signed-off-by: Gao feng <gaofeng(a)cn.fujitsu.com> --- src/lxc/lxc_container.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/src/lxc/lxc_container.c b/src/lxc/lxc_container.c index 9045a7d..24b1017 100644 --- a/src/lxc/lxc_container.c +++ b/src/lxc/lxc_container.c @@ -1323,7 +1323,7 @@ static int lxcContainerSetupPivotRoot(virDomainDefPtr vmDef, /* Before pivoting we need to identify any * cgroups controllers that are mounted */ if (lxcContainerIdentifyCGroups(&mounts, &nmounts) < 0) - goto cleanup; + return -1; /* Gives us a private root, leaving all parent OS mounts on /.oldroot */ if (lxcContainerPivotRoot(root) < 0) @@ -1406,7 +1406,7 @@ static int lxcContainerSetupExtraMounts(virDomainDefPtr vmDef, /* Before replacing /sys we need to identify any * cgroups controllers that are mounted */ if (lxcContainerIdentifyCGroups(&mounts, &nmounts) < 0) - goto cleanup; + return -1; /* Gets rid of any existing stuff under /proc, since we need new * namespace aware versions of those. We must do /proc second -- 1.7.7.6

2 2

[libvirt] [PATCHv2] adding handling EINTR to poll to make it more robust
by lvroyce＠linux.vnet.ibm.com 19 Jun '12

19 Jun '12

From: Royce Lv <lvroyce(a)linux.vnet.ibm.com> some system call and signal will interrupt poll, making event loop stops and fails to react events and keepalive message from libvirt. adding handling EINTR to poll to make it more robust Signed-off-by: Royce Lv <lvroyce(a)linux.vnet.ibm.com> --- examples/domain-events/events-python/event-test.py | 95 +++++++++++--------- 1 files changed, 51 insertions(+), 44 deletions(-) diff --git a/examples/domain-events/events-python/event-test.py b/examples/domain-events/events-python/event-test.py index 96dc268..7832d13 100644 --- a/examples/domain-events/events-python/event-test.py +++ b/examples/domain-events/events-python/event-test.py @@ -2,9 +2,9 @@ # # # -################################################################################# +############################################################################## # Start off by implementing a general purpose event loop for anyones use -################################################################################# +############################################################################## import sys import getopt @@ -27,11 +27,14 @@ import threading use_pure_python_event_loop = True do_debug = False + + def debug(msg): global do_debug if do_debug: print msg + # # This general purpose event loop will support waiting for file handle # I/O and errors events, as well as scheduling repeatable timers with @@ -97,7 +100,6 @@ class virEventLoopPure: self.cb(self.timer, self.opaque) - def __init__(self): self.poll = select.poll() self.pipetrick = os.pipe() @@ -178,48 +180,53 @@ class virEventLoopPure: def run_once(self): sleep = -1 self.runningPoll = True - next = self.next_timeout() - debug("Next timeout due at %d" % next) - if next > 0: - now = int(time.time() * 1000) - if now >= next: - sleep = 0 - else: - sleep = (next - now) / 1000.0 - - debug("Poll with a sleep of %d" % sleep) - events = self.poll.poll(sleep) - - # Dispatch any file handle events that occurred - for (fd, revents) in events: - # See if the events was from the self-pipe - # telling us to wakup. if so, then discard - # the data just continue - if fd == self.pipetrick[0]: - self.pendingWakeup = False - data = os.read(fd, 1) - continue - - h = self.get_handle_by_fd(fd) - if h: - debug("Dispatch fd %d handle %d events %d" % (fd, h.get_id(), revents)) - h.dispatch(self.events_from_poll(revents)) + try: + next = self.next_timeout() + debug("Next timeout due at %d" % next) + if next > 0: + now = int(time.time() * 1000) + if now >= next: + sleep = 0 + else: + sleep = (next - now) / 1000.0 + + debug("Poll with a sleep of %d" % sleep) + events = self.poll.poll(sleep) + + # Dispatch any file handle events that occurred + for (fd, revents) in events: + # See if the events was from the self-pipe + # telling us to wakup. if so, then discard + # the data just continue + if fd == self.pipetrick[0]: + self.pendingWakeup = False + data = os.read(fd, 1) + continue + + h = self.get_handle_by_fd(fd) + if h: + debug("Dispatch fd %d handle %d events %d" % (fd, h.get_id(), revents)) + h.dispatch(self.events_from_poll(revents)) - now = int(time.time() * 1000) - for t in self.timers: - interval = t.get_interval() - if interval < 0: - continue - - want = t.get_last_fired() + interval - # Deduct 20ms, since schedular timeslice - # means we could be ever so slightly early - if now >= (want-20): - debug("Dispatch timer %d now %s want %s" % (t.get_id(), str(now), str(want))) - t.set_last_fired(now) - t.dispatch() - - self.runningPoll = False + now = int(time.time() * 1000) + for t in self.timers: + interval = t.get_interval() + if interval < 0: + continue + + want = t.get_last_fired() + interval + # Deduct 20ms, since schedular timeslice + # means we could be ever so slightly early + if now >= (want-20): + debug("Dispatch timer %d now %s want %s" % (t.get_id(), str(now), str(want))) + t.set_last_fired(now) + t.dispatch() + + except (os.error, select.error), e: + if e.errno != errno.EINTR: + raise + finally: + self.runningPoll = False # Actually the event loop forever -- 1.7.7.6

1 0

[libvirt] pool-create/pool-create-as should include build pool then start?
by Zhimou Peng 19 Jun '12

19 Jun '12

Hi all, According to https://bugzilla.redhat.com/show_bug.cgi?id=830056#c5, i think we should talk about this: Description If pool environment is not built, pool-create/pool-create-as will always fail. I think pool-create should build the pool. Version qemu-kvm-0.12.1.2-2.295.el6.x86_64 kernel-2.6.32-276.el6.x86_64 libvirt-0.9.10-21.el6.x86_64 Steps For example: dir pool virsh # pool-create dirpool.xml error: Failed to create pool from dirpool.xml error: cannot open path '/var/lib/libvirt/dir': No such file or directory disk pool virsh # pool-create diskpool.xml error: Failed to create pool from diskpool.xml error: internal error Child process (/usr/libexec/libvirt_parthelper /dev/sdb) status unexpected: exit status 2 fs pool virsh # pool-create fspool.xml error: Failed to create pool from fspool.xml error: internal error Child process (/bin/mount -t ext4 /dev/sda7 /var/lib/libvirt/images/fspool) status unexpected: exit status 32 Steps are the same of other pools

2 2