- 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 19 Jun '12

19 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