- Devel - libvirt List Archives

[libvirt] [PATCH] storage backend rbd: Do not prefix rbd: on volume names.
by Wido den Hollander 25 Jun '12

25 Jun '12

We used to prefix 'rbd:' to volume names, this is not necessary. Qemu takes RBD devices in this way, like: qemu -drive rbd:pool/image When attaching a network disk like RBD to a guest we however do not use this prefix. Currently you can't map a RBD volume name directly to a domain without removing the prefix. Signed-off-by: Wido den Hollander <wido(a)widodh.nl> --- src/storage/storage_backend_rbd.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/storage/storage_backend_rbd.c b/src/storage/storage_backend_rbd.c index 3e7464c..7615dcc 100644 --- a/src/storage/storage_backend_rbd.c +++ b/src/storage/storage_backend_rbd.c @@ -237,7 +237,7 @@ static int volStorageBackendRBDRefreshVolInfo(virStorageVolDefPtr vol, vol->type = VIR_STORAGE_VOL_NETWORK; VIR_FREE(vol->target.path); - if (virAsprintf(&vol->target.path, "rbd:%s/%s", + if (virAsprintf(&vol->target.path, "%s/%s", pool->def->source.name, vol->name) == -1) { virReportOOMError(); -- 1.7.9.5

3 2

[libvirt] [PATCH 1/2] storage: Switch to new def on pool-destroy
by Michal Privoznik 25 Jun '12

25 Jun '12

Currently, we share the idea of old & new def with domains. Users can *-edit an object (domain, pool) which spawns a new internal representation for them. This is referenced via {domainObj,poolObj}->newDef [compared to ->def]. However, for pool we were never overwriting def with newDef. This must be done on pool-destroy (like we do analogically in domain detroy). --- src/storage/storage_driver.c | 5 ++++- 1 files changed, 4 insertions(+), 1 deletions(-) diff --git a/src/storage/storage_driver.c b/src/storage/storage_driver.c index 88f3809..3b95c70 100644 --- a/src/storage/storage_driver.c +++ b/src/storage/storage_driver.c @@ -794,6 +794,10 @@ storagePoolDestroy(virStoragePoolPtr obj) { if (pool->configFile == NULL) { virStoragePoolObjRemove(&driver->pools, pool); pool = NULL; + } else if (pool->newDef) { + virStoragePoolDefFree(pool->def); + pool->def = pool->newDef; + pool->newDef = NULL; } ret = 0; @@ -804,7 +808,6 @@ cleanup: return ret; } - static int storagePoolDelete(virStoragePoolPtr obj, unsigned int flags) { -- 1.7.8.5

3 7

[libvirt] [PATCH] qemu: Improve error if setmem fails for lacking of balloon support
by Osier Yang 25 Jun '12

25 Jun '12

"cannot set memory of an active domain" is misleading, it sounds like setting memory of active domain is not supported. --- src/qemu/qemu_driver.c | 3 ++- 1 files changed, 2 insertions(+), 1 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 5713650..bb33fd5 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -2037,7 +2037,8 @@ static int qemudDomainSetMemoryFlags(virDomainPtr dom, unsigned long newmem, /* Lack of balloon support is a fatal error */ if (r == 0) { qemuReportError(VIR_ERR_OPERATION_INVALID, "%s", - _("cannot set memory of an active domain")); + _("Lacks of balloon support, perhaps" + " the balloon device is disabled")); goto endjob; } } -- 1.7.7.3

2 2

[libvirt] This patch creates a mount point for libvirt-lxc containers to mount on the destination system if it does not exist.
by Daniel J Walsh 25 Jun '12

25 Jun '12

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Currently libvirt-lxc checks to see if the destination exists and is a directory. If it is not a directory then the mount fails. Since libvirt-lxc can bind mount files on an inode, this patch is needed to allow us to bind mount files on files. Currently we want to bind mount on top of /etc/machine-id, and /etc/adjtime If the destination of the mount point does not exists, it checks if the src is a directory and then attempts to create a directory, otherwise it creates an empty file for the destination. The code will then bind mount over the destination. Current code blows up if the destination was not a directory. We want to be able to bind mount files on files. Sorry if you are seeing this patch for the second time, since I sent it to the wrong list. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.12 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAk/kel4ACgkQrlYvE4MpobP7uwCfXUPMJP4bNZiBGTwnJ70dezVf KPwAnjt3MYSlHxkcwZCe5H1X4C0P4ky/ =1te/ -----END PGP SIGNATURE-----

2 2

[libvirt] [PATCH libvirt-gconfig] Add support for domain filesystems with type=bind|ram
by Daniel P. Berrange 25 Jun '12

25 Jun '12

From: "Daniel P. Berrange" <berrange(a)redhat.com> This adds support for filesystems with type=bind or type=ram, eg <filesystem type='bind'> <source dir="/some/guest/file"/> <target dir=/other/guest/file"/> </filesystem> And <filesystem type='ram'> <source usage='102400'/> <target dir='/run'/> </filesystem> Signed-off-by: Daniel P. Berrange <berrange(a)redhat.com> --- libvirt-gconfig/libvirt-gconfig-domain-filesys.c | 19 +++++++++++++++++++ libvirt-gconfig/libvirt-gconfig-domain-filesys.h | 4 ++++ 2 files changed, 23 insertions(+) diff --git a/libvirt-gconfig/libvirt-gconfig-domain-filesys.c b/libvirt-gconfig/libvirt-gconfig-domain-filesys.c index c9c7aa8..54f124a 100644 --- a/libvirt-gconfig/libvirt-gconfig-domain-filesys.c +++ b/libvirt-gconfig/libvirt-gconfig-domain-filesys.c @@ -146,6 +146,7 @@ void gvir_config_domain_filesys_set_source(GVirConfigDomainFilesys *filesys, switch (filesys->priv->type) { case GVIR_CONFIG_DOMAIN_FILESYS_MOUNT: + case GVIR_CONFIG_DOMAIN_FILESYS_BIND: attribute_name = "dir"; break; case GVIR_CONFIG_DOMAIN_FILESYS_FILE: @@ -157,6 +158,9 @@ void gvir_config_domain_filesys_set_source(GVirConfigDomainFilesys *filesys, case GVIR_CONFIG_DOMAIN_FILESYS_TEMPLATE: attribute_name = "name"; break; + case GVIR_CONFIG_DOMAIN_FILESYS_RAM: + g_return_if_reached(); + default: g_return_if_reached(); } @@ -166,6 +170,21 @@ void gvir_config_domain_filesys_set_source(GVirConfigDomainFilesys *filesys, attribute_name, source); } +void gvir_config_domain_filesys_set_ram_usage(GVirConfigDomainFilesys *filesys, + guint64 bytes) +{ + g_return_if_fail(GVIR_CONFIG_IS_DOMAIN_FILESYS(filesys)); + GVirConfigObject *src; + + src = gvir_config_object_replace_child(GVIR_CONFIG_OBJECT(filesys), + "source"); + + gvir_config_object_set_attribute_with_type(src, + "usage", G_TYPE_UINT64, bytes, + "units", G_TYPE_STRING, "bytes", + NULL); + +} void gvir_config_domain_filesys_set_target(GVirConfigDomainFilesys *filesys, const char *path) diff --git a/libvirt-gconfig/libvirt-gconfig-domain-filesys.h b/libvirt-gconfig/libvirt-gconfig-domain-filesys.h index 2ceb2a8..e4b7c20 100644 --- a/libvirt-gconfig/libvirt-gconfig-domain-filesys.h +++ b/libvirt-gconfig/libvirt-gconfig-domain-filesys.h @@ -61,6 +61,8 @@ typedef enum { GVIR_CONFIG_DOMAIN_FILESYS_BLOCK, GVIR_CONFIG_DOMAIN_FILESYS_FILE, GVIR_CONFIG_DOMAIN_FILESYS_TEMPLATE, + GVIR_CONFIG_DOMAIN_FILESYS_RAM, + GVIR_CONFIG_DOMAIN_FILESYS_BIND, } GVirConfigDomainFilesysType; typedef enum { @@ -89,6 +91,8 @@ void gvir_config_domain_filesys_set_driver_type(GVirConfigDomainFilesys *filesys GVirConfigDomainFilesysDriverType type); void gvir_config_domain_filesys_set_source(GVirConfigDomainFilesys *filesys, const char *source); +void gvir_config_domain_filesys_set_ram_usage(GVirConfigDomainFilesys *filesys, + guint64 bytes); void gvir_config_domain_filesys_set_target(GVirConfigDomainFilesys *filesys, const char *target); void gvir_config_domain_filesys_set_readonly(GVirConfigDomainFilesys *filesys, -- 1.7.10.2

2 1

[libvirt] [PATCH] conf: Fill listenAddr from driver if none set
by Michal Privoznik 25 Jun '12

25 Jun '12

Currently, if domain's graphic has no listenAddr set we provide the one from qemu.conf. However, we must make this transient to not overwrite future changes to the config file. Moreover, we have to store this into domain state XML so we remain consistent on eventual config file edits. --- Since users will never see new 'driverListenAddress' attribute I don't think update of RNG schema is necessary. src/conf/domain_conf.c | 14 ++++++++++++++ src/conf/domain_conf.h | 6 ++++++ src/qemu/qemu_command.c | 18 ++++++++++++++++-- 3 files changed, 36 insertions(+), 2 deletions(-) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 4651765..29fa215 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -909,6 +909,7 @@ void virDomainGraphicsDefFree(virDomainGraphicsDefPtr def) for (ii = 0; ii < def->nListens; ii++) virDomainGraphicsListenDefClear(&def->listens[ii]); VIR_FREE(def->listens); + VIR_FREE(def->driverListenAddress); VIR_FREE(def); } @@ -5976,6 +5977,7 @@ virDomainGraphicsDefParseXML(xmlNodePtr node, int nListens; xmlNodePtr *listenNodes = NULL; char *listenAddr = NULL; + char *driverListenAddress = NULL; xmlNodePtr save = ctxt->node; if (VIR_ALLOC(def) < 0) { @@ -5999,6 +6001,12 @@ virDomainGraphicsDefParseXML(xmlNodePtr node, goto error; } + if (flags & VIR_DOMAIN_XML_INTERNAL_STATUS) { + driverListenAddress = virXMLPropString(node, "driverListenAddress"); + if (driverListenAddress) + def->driverListenAddress = driverListenAddress; + } + if (def->type == VIR_DOMAIN_GRAPHICS_TYPE_VNC || def->type == VIR_DOMAIN_GRAPHICS_TYPE_RDP || def->type == VIR_DOMAIN_GRAPHICS_TYPE_SPICE) { @@ -12323,9 +12331,15 @@ virDomainGraphicsDefFormat(virBufferPtr buf, break; } } + if (!listenAddr && !(flags & VIR_DOMAIN_XML_INACTIVE)) + listenAddr = def->driverListenAddress; virBufferAsprintf(buf, " <graphics type='%s'", type); + if (flags & VIR_DOMAIN_XML_INTERNAL_STATUS && def->driverListenAddress) + virBufferAsprintf(buf, " driverListenAddress='%s'", + def->driverListenAddress); + switch (def->type) { case VIR_DOMAIN_GRAPHICS_TYPE_VNC: if (def->data.vnc.socket) { diff --git a/src/conf/domain_conf.h b/src/conf/domain_conf.h index 3712785..6ad48fd 100644 --- a/src/conf/domain_conf.h +++ b/src/conf/domain_conf.h @@ -1273,6 +1273,12 @@ struct _virDomainGraphicsDef { * simplify parsing code.*/ size_t nListens; virDomainGraphicsListenDefPtr listens; + + /* In qemu driver if none of 'listens' contain usable listen address, + * the one from qemu.conf file is taken. However, we cannot store it + * into inactive XML as it would discard all future changes made to + * the config file. */ + char * driverListenAddress; }; enum virDomainRedirdevBus { diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index 6549f57..9878a66 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -5496,8 +5496,15 @@ qemuBuildCommandLine(virConnectPtr conn, break; } - if (!listenAddr) + if (!listenAddr) { listenAddr = driver->vncListen; + if (listenAddr && + !(def->graphics[0]->driverListenAddress = + strdup(listenAddr))) { + virReportOOMError(); + goto error; + } + } escapeAddr = strchr(listenAddr, ':') != NULL; if (escapeAddr) @@ -5647,8 +5654,15 @@ qemuBuildCommandLine(virConnectPtr conn, break; } - if (!listenAddr) + if (!listenAddr) { listenAddr = driver->spiceListen; + if (listenAddr && + !(def->graphics[0]->driverListenAddress = + strdup(listenAddr))) { + virReportOOMError(); + goto error; + } + } if (listenAddr) virBufferAsprintf(&opt, ",addr=%s", listenAddr); -- 1.7.8.5

2 5

[libvirt] [libvirt-glib] mainloop integration fixes
by Christophe Fergeau 25 Jun '12

25 Jun '12

Hey, While testing Boxes, I've noticed that if I started a VM and then let it in the background for a while, it usually crashed after a while. When this happens, the backtrace is: Program received signal SIGSEGV, Segmentation fault. 0x00007ffff6671898 in gvir_event_handle_dispatch (source=0x7fffdc001f60, condition=G_IO_IN, opaque=0x7fffdc0021a0) at libvirt-glib-event.c:135 135 (data->cb)(data->watch, data->fd, events, data->opaque); (gdb) bt #0 0x00007ffff6671898 in gvir_event_handle_dispatch (source=0x7fffdc001f60, condition=G_IO_IN, opaque=0x7fffdc0021a0) at libvirt-glib-event.c:135 #1 0x00007ffff231dcb4 in g_io_unix_dispatch (source=0x40e34a0, callback=0x7ffff66717d8 <gvir_event_handle_dispatch>, user_data=0x7fffdc0021a0) at giounix.c:166 #2 0x00007ffff22cef15 in g_main_dispatch (context=0x7007f0) at gmain.c:2539 #3 0x00007ffff22cfbda in g_main_context_dispatch (context=0x7007f0) at gmain.c:3075 #4 0x00007ffff22cfdbd in g_main_context_iterate (context=0x7007f0, block=1, dispatch=1, self=0x15d3180) at gmain.c:3146 #5 0x00007ffff22cfe81 in g_main_context_iteration (context=0x7007f0, may_block=1) at gmain.c:3207 #6 0x00007ffff2ad7e7c in g_application_run (application=0x70b890, argc=0, argv=0x0) at gapplication.c:1607 #7 0x000000000041df98 in boxes_app_run (self=0x1527010) at app.c:1254 #8 0x0000000000450134 in _vala_main (args=0x7fffffffd708, args_length1=1) at main.c:729 #9 0x000000000045019e in main (argc=1, argv=0x7fffffffd708) at main.c:740 and the corresponding gvir_event_handle structure is corrupted: (gdb) p *data $3 = {watch = 0, fd = 0, events = 1886417008, enabled = 1886417008, channel = 0x7070707070707070, source = 1886417008, cb = 0x7070707070707070, opaque = 0x7070707070707070, ff = 0x7ffff60cfba0 <virNetSocketEventFree>} (gdb) p/x *data $4 = {watch = 0x0, fd = 0x0, events = 0x70707070, enabled = 0x70707070, channel = 0x7070707070707070, source = 0x70707070, cb = 0x7070707070707070, opaque = 0x7070707070707070, ff = 0x7ffff60cfba0} (I'm running with MALLOC_PERTURB_ set so the 0x70707070 pattern indicates uninitialized memory). Right before this happens, the libvirt-glib event debugging log has: (gnome-boxes:29577): Libvirt.GLib-DEBUG: Remove handle 0x7fffdc0021a0 1 27 (gnome-boxes:29577): Libvirt.GLib-DEBUG: Update handle 0x7fffdc0021a0 1 27 1 (gnome-boxes:29577): Libvirt.GLib-DEBUG: Update for missing handle watch 1 (gnome-boxes:29577): Libvirt.GLib-DEBUG: Dispatch handler 0x7fffdc0021a0 0 0 1 0x7070707070707070 that is, libvirt removes a handle watch, and then updates this handle (in other words, it calls virEventPollUpdateHandle after having called virEventPollRemoveHandle on the same handle). I don't know if it's legit for libvirt to do that, but we can make libvirt-glib robust against this situation, which is what patch 5/5 does. The first 3 patches are small cleanups, and 4/5 fixes a potential race which I found while looking at this code. Christophe

2 13

[libvirt] [libvirt-glib 1/8] Setters/adders for GVirConfigCapabilitiesCpu props
by Zeeshan Ali (Khattak) 25 Jun '12

25 Jun '12

From: "Zeeshan Ali (Khattak)" <zeeshanak(a)gnome.org> - gvir_config_capabilities_cpu_add_feature - gvir_config_capabilities_cpu_set_topology --- libvirt-gconfig/libvirt-gconfig-capabilities-cpu.c | 32 ++++++++++++++++++++ libvirt-gconfig/libvirt-gconfig-capabilities-cpu.h | 6 ++++ libvirt-gconfig/libvirt-gconfig.sym | 2 ++ 3 files changed, 40 insertions(+) diff --git a/libvirt-gconfig/libvirt-gconfig-capabilities-cpu.c b/libvirt-gconfig/libvirt-gconfig-capabilities-cpu.c index df77364..e517a20 100644 --- a/libvirt-gconfig/libvirt-gconfig-capabilities-cpu.c +++ b/libvirt-gconfig/libvirt-gconfig-capabilities-cpu.c @@ -56,6 +56,21 @@ gvir_config_capabilities_cpu_get_arch(GVirConfigCapabilitiesCpu *cpu) return gvir_config_object_get_node_content(GVIR_CONFIG_OBJECT(cpu), "arch"); } +/** + * gvir_config_capabilities_cpu_add_feature: + * + * Adds a new feature to CPU. + */ +void gvir_config_capabilities_cpu_add_feature(GVirConfigCapabilitiesCpu *cpu, + GVirConfigCapabilitiesCpuFeature *feature) +{ + g_return_if_fail(GVIR_CONFIG_IS_CAPABILITIES_CPU(cpu)); + g_return_if_fail(GVIR_CONFIG_IS_CAPABILITIES_CPU_FEATURE(feature)); + + gvir_config_object_attach_add(GVIR_CONFIG_OBJECT(cpu), + GVIR_CONFIG_OBJECT(feature)); +} + struct GetFeatureData { GVirConfigXmlDoc *doc; const gchar *schema; @@ -133,3 +148,20 @@ gvir_config_capabilities_cpu_get_topology(GVirConfigCapabilitiesCpu *cpu) return GVIR_CONFIG_CAPABILITIES_CPU_TOPOLOGY(object); } + +/** + * gvir_config_capabilities_cpu_set_topology: + * + * Sets the topology of the cpu. + */ +void +gvir_config_capabilities_cpu_set_topology(GVirConfigCapabilitiesCpu *cpu, + GVirConfigCapabilitiesCpuTopology *topology) +{ + g_return_if_fail(GVIR_CONFIG_IS_CAPABILITIES_CPU(cpu)); + g_return_if_fail(GVIR_CONFIG_IS_CAPABILITIES_CPU_TOPOLOGY(topology)); + + gvir_config_object_attach_replace(GVIR_CONFIG_OBJECT(cpu), + "topology", + GVIR_CONFIG_OBJECT(topology)); +} diff --git a/libvirt-gconfig/libvirt-gconfig-capabilities-cpu.h b/libvirt-gconfig/libvirt-gconfig-capabilities-cpu.h index 4d896ef..67fe607 100644 --- a/libvirt-gconfig/libvirt-gconfig-capabilities-cpu.h +++ b/libvirt-gconfig/libvirt-gconfig-capabilities-cpu.h @@ -29,6 +29,7 @@ #define __LIBVIRT_GCONFIG_CAPABILITIES_CPU_H__ #include "libvirt-gconfig-capabilities-cpu-topology.h" +#include "libvirt-gconfig-capabilities-cpu-feature.h" G_BEGIN_DECLS @@ -63,10 +64,15 @@ GType gvir_config_capabilities_cpu_get_type(void); const gchar * gvir_config_capabilities_cpu_get_arch(GVirConfigCapabilitiesCpu *cpu); +void gvir_config_capabilities_cpu_add_feature(GVirConfigCapabilitiesCpu *cpu, + GVirConfigCapabilitiesCpuFeature *feature); GList * gvir_config_capabilities_cpu_get_features(GVirConfigCapabilitiesCpu *cpu); GVirConfigCapabilitiesCpuTopology * gvir_config_capabilities_cpu_get_topology(GVirConfigCapabilitiesCpu *cpu); +void +gvir_config_capabilities_cpu_set_topology(GVirConfigCapabilitiesCpu *cpu, + GVirConfigCapabilitiesCpuTopology *topology); G_END_DECLS diff --git a/libvirt-gconfig/libvirt-gconfig.sym b/libvirt-gconfig/libvirt-gconfig.sym index 5473148..a9e5da5 100644 --- a/libvirt-gconfig/libvirt-gconfig.sym +++ b/libvirt-gconfig/libvirt-gconfig.sym @@ -389,9 +389,11 @@ LIBVIRT_GCONFIG_0.0.9 { gvir_config_capabilities_get_guests; gvir_config_capabilities_cpu_get_type; + gvir_config_capabilities_cpu_add_feature; gvir_config_capabilities_cpu_get_arch; gvir_config_capabilities_cpu_get_features; gvir_config_capabilities_cpu_get_topology; + gvir_config_capabilities_cpu_set_topology; gvir_config_capabilities_cpu_feature_get_type; gvir_config_capabilities_cpu_feature_get_name; -- 1.7.10.4

3 13

[libvirt] [PATCH v2] storage: Introduce --inactive for pool-dumpxml
by Michal Privoznik 25 Jun '12

25 Jun '12

Storage is one of the last domains in libvirt where we don't fully utilize inactive and live XML. Okay, it might be because we don't have support for that. So implement such support. However, we need to fallback when talking to old daemon which doesn't support this new flag called VIR_STORAGE_XML_INACTIVE. --- include/libvirt/libvirt.h.in | 4 ++++ src/libvirt.c | 2 +- src/storage/storage_driver.c | 10 ++++++++-- tools/virsh.c | 24 ++++++++++++++++++++++-- tools/virsh.pod | 4 +++- 5 files changed, 38 insertions(+), 6 deletions(-) diff --git a/include/libvirt/libvirt.h.in b/include/libvirt/libvirt.h.in index 024c4ec..6e8d5dd 100644 --- a/include/libvirt/libvirt.h.in +++ b/include/libvirt/libvirt.h.in @@ -2445,6 +2445,10 @@ struct _virStorageVolInfo { typedef virStorageVolInfo *virStorageVolInfoPtr; +typedef enum { + VIR_STORAGE_XML_INACTIVE = (1 << 0), /* dump inactive pool/volume information */ +} virStorageXMLFlags; + /* * Get connection from pool. */ diff --git a/src/libvirt.c b/src/libvirt.c index 0aa50cb..ef5ac47 100644 --- a/src/libvirt.c +++ b/src/libvirt.c @@ -12169,7 +12169,7 @@ error: /** * virStoragePoolGetXMLDesc: * @pool: pointer to storage pool - * @flags: bitwise-OR of virDomainXMLFlags + * @flags: bitwise-OR of virStorageXMLFlags * * Fetch an XML document describing all aspects of the * storage pool. This is suitable for later feeding back diff --git a/src/storage/storage_driver.c b/src/storage/storage_driver.c index 3b95c70..fbc630d 100644 --- a/src/storage/storage_driver.c +++ b/src/storage/storage_driver.c @@ -960,9 +960,10 @@ storagePoolGetXMLDesc(virStoragePoolPtr obj, { virStorageDriverStatePtr driver = obj->conn->storagePrivateData; virStoragePoolObjPtr pool; + virStoragePoolDefPtr def; char *ret = NULL; - virCheckFlags(0, NULL); + virCheckFlags(VIR_STORAGE_XML_INACTIVE, NULL); storageDriverLock(driver); pool = virStoragePoolObjFindByUUID(&driver->pools, obj->uuid); @@ -974,7 +975,12 @@ storagePoolGetXMLDesc(virStoragePoolPtr obj, goto cleanup; } - ret = virStoragePoolDefFormat(pool->def); + if ((flags & VIR_STORAGE_XML_INACTIVE) && pool->newDef) + def = pool->newDef; + else + def = pool->def; + + ret = virStoragePoolDefFormat(def); cleanup: if (pool) diff --git a/tools/virsh.c b/tools/virsh.c index 567a5f6..2b4cb2c 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -10807,6 +10807,7 @@ static const vshCmdInfo info_pool_dumpxml[] = { static const vshCmdOptDef opts_pool_dumpxml[] = { {"pool", VSH_OT_DATA, VSH_OFLAG_REQ, N_("pool name or uuid")}, + {"inactive", VSH_OT_BOOL, 0, N_("show inactive defined XML")}, {NULL, 0, 0, NULL} }; @@ -10815,15 +10816,20 @@ cmdPoolDumpXML(vshControl *ctl, const vshCmd *cmd) { virStoragePoolPtr pool; bool ret = true; + bool inactive = vshCommandOptBool(cmd, "inactive"); + unsigned int flags = 0; char *dump; + if (inactive) + flags |= VIR_STORAGE_XML_INACTIVE; + if (!vshConnectionUsability(ctl, ctl->conn)) return false; if (!(pool = vshCommandOptPool(ctl, cmd, "pool", NULL))) return false; - dump = virStoragePoolGetXMLDesc(pool, 0); + dump = virStoragePoolGetXMLDesc(pool, flags); if (dump != NULL) { vshPrint(ctl, "%s", dump); VIR_FREE(dump); @@ -16127,7 +16133,8 @@ cmdPoolEdit(vshControl *ctl, const vshCmd *cmd) bool ret = false; virStoragePoolPtr pool = NULL; virStoragePoolPtr pool_edited = NULL; - unsigned int flags = 0; + unsigned int flags = VIR_STORAGE_XML_INACTIVE; + char *tmp_desc = NULL; if (!vshConnectionUsability(ctl, ctl->conn)) goto cleanup; @@ -16136,6 +16143,19 @@ cmdPoolEdit(vshControl *ctl, const vshCmd *cmd) if (pool == NULL) goto cleanup; + /* Some old daemons don't support _INACTIVE flag */ + if (!(tmp_desc = virStoragePoolGetXMLDesc(pool, flags))) { + if (last_error->code == VIR_ERR_INVALID_ARG) { + flags &= ~VIR_STORAGE_XML_INACTIVE; + virFreeError(last_error); + last_error = NULL; + } else { + goto cleanup; + } + } else { + VIR_FREE(tmp_desc); + } + #define EDIT_GET_XML virStoragePoolGetXMLDesc(pool, flags) #define EDIT_NOT_CHANGED \ vshPrint(ctl, _("Pool %s XML configuration not changed.\n"), \ diff --git a/tools/virsh.pod b/tools/virsh.pod index f83a29d..191c9f2 100644 --- a/tools/virsh.pod +++ b/tools/virsh.pod @@ -2112,9 +2112,11 @@ Destroy the resources used by a given I<pool> object. This operation is non-recoverable. The I<pool> object will still exist after this command, ready for the creation of new storage volumes. -=item B<pool-dumpxml> I<pool-or-uuid> +=item B<pool-dumpxml> [I<--inactive>] I<pool-or-uuid> Returns the XML information about the I<pool> object. +I<--inactive> tells virsh to dump pool configuration that will be used +on next start of the pool as opposed to the current pool configuration. =item B<pool-edit> I<pool-or-uuid> -- 1.7.8.5

2 1

[libvirt] [PATCH] sanlock: Properly indent
by Michal Privoznik 25 Jun '12

25 Jun '12

One of latest patches (0fce94fe) didn't properly indented #define making syntax-check fail. --- Pushed under build-breaker rule. src/locking/lock_driver_sanlock.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/src/locking/lock_driver_sanlock.c b/src/locking/lock_driver_sanlock.c index 9668ca9..26bdde8 100644 --- a/src/locking/lock_driver_sanlock.c +++ b/src/locking/lock_driver_sanlock.c @@ -60,7 +60,7 @@ * too old to provide that define, and probably the functionality too */ #ifndef SANLK_RES_SHARED -#define SANLK_RES_SHARED 0x4 +# define SANLK_RES_SHARED 0x4 #endif typedef struct _virLockManagerSanlockDriver virLockManagerSanlockDriver; -- 1.7.8.5

1 0

[libvirt] [PATCH] Fix compilation on older sanlock
by Daniel Veillard 25 Jun '12

25 Jun '12

This is clearly not ideal, but we ought to check the version of sanlock and act accordingly. We may end up with autoconf.ac and libvirt.spec.in change for the real fix. Pushing under the build breaker rule Temporary fix since compilation broke with older version of sanlock following acbd4965c44c4dbc676dfe89aff970052e376073 diff --git a/src/locking/lock_driver_sanlock.c b/src/locking/lock_driver_sanlock.c index 16941c9..9668ca9 100644 --- a/src/locking/lock_driver_sanlock.c +++ b/src/locking/lock_driver_sanlock.c @@ -55,6 +55,14 @@ #define VIR_LOCK_MANAGER_SANLOCK_AUTO_DISK_LOCKSPACE "__LIBVIRT__DISKS__" +/* + * temporary fix for the case where the sanlock devel package is + * too old to provide that define, and probably the functionality too + */ +#ifndef SANLK_RES_SHARED +#define SANLK_RES_SHARED 0x4 +#endif + typedef struct _virLockManagerSanlockDriver virLockManagerSanlockDriver; typedef virLockManagerSanlockDriver *virLockManagerSanlockDriverPtr; -- Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ daniel(a)veillard.com | Rpmfind RPM search engine http://rpmfind.net/ http://veillard.com/ | virtualization library http://libvirt.org/

1 0

[libvirt] This patch mounts /run and /var/run before running a sandbox container
by Daniel J Walsh 25 Jun '12

25 Jun '12

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 We do not want to share /run between containers and the host. This patch mounts a tmpfs on /run and then bind mounts /run on /var/run. Reason this is pach 0003 is that this is a part of a previous patch set that was never applied. Sorry to those who are receiving this patch for the second time, as I sent it first to the wrong list. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.12 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAk/keZcACgkQrlYvE4MpobNnagCgyHc9igEYxPT/Bblf1XVPkwMX EM0Anj9xzzf9k88He1gkoXKBEAija52U =4Uny -----END PGP SIGNATURE-----

2 1

[libvirt] regression ?
by Harsh Bora 25 Jun '12

25 Jun '12

Hi, I was testing an under-development patch on upstream libvirt using domxml-to-native and domxml-from-native and realised that the domxml-from-native fails for a successfully generated qemu cmd line using domxml-to-native. To isolate the issue, I tried generating the qemu cmd line for tests/qemuxml2argvdata/qemuxml2argv-disk-virtio.xml like this: $virsh domxml-to-native qemu-argv tests/qemuxml2argvdata/qemuxml2argv-disk-virtio.xml > qemu-args Now, when I try to generate the XML back for the same CLI, it fails: $virsh domxml-from-native qemu-argv qem-args error: internal error missing index/unit/bus parameter in drive 'file=/dev/HostVG/QEMUGuest1,if=none,id=drive-ide0-0-0,format=raw' Does anyone have any idea, what could be causing this ? regards, Harsh

2 1

[libvirt] [PATCH] virsh: Use virXPath wrappers for vncdisplay cmd
by Doug Goldstein 25 Jun '12

25 Jun '12

Update the vncdisplay command to use the virXPath wrappers as well as check if the domain is up rather than using the port set to -1 to mean the domain is not up. Signed-off-by: Doug Goldstein <cardoe(a)cardoe.com> --- tools/virsh.c | 30 +++++++++++++++--------------- 1 files changed, 15 insertions(+), 15 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index 0354822..a6649f4 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -13844,12 +13844,12 @@ static bool cmdVNCDisplay(vshControl *ctl, const vshCmd *cmd) { xmlDocPtr xml = NULL; - xmlXPathObjectPtr obj = NULL; xmlXPathContextPtr ctxt = NULL; virDomainPtr dom; bool ret = false; int port = 0; char *doc; + char *listen_addr; if (!vshConnectionUsability(ctl, ctl->conn)) return false; @@ -13857,6 +13857,12 @@ cmdVNCDisplay(vshControl *ctl, const vshCmd *cmd) if (!(dom = vshCommandOptDomain(ctl, cmd, NULL))) return false; + /* Check if the domain is active and don't rely on -1 for this */ + if (!virDomainIsActive(dom)) { + vshError(ctl, _("Domain is not running")); + goto cleanup; + } + doc = virDomainGetXMLDesc(dom, 0); if (!doc) goto cleanup; @@ -13866,29 +13872,23 @@ cmdVNCDisplay(vshControl *ctl, const vshCmd *cmd) if (!xml) goto cleanup; - obj = xmlXPathEval(BAD_CAST "string(/domain/devices/graphics[@type='vnc']/@port)", ctxt); - if (obj == NULL || obj->type != XPATH_STRING || - obj->stringval == NULL || obj->stringval[0] == 0) { + /* Get the VNC port */ + if (virXPathInt("string(/domain/devices/graphics[@type='vnc']/@port)", + ctxt, &port)) { + vshError(ctl, _("Failed to get VNC port. Is this domain using VNC?")); goto cleanup; } - if (virStrToLong_i((const char *)obj->stringval, NULL, 10, &port) || port < 0) - goto cleanup; - xmlXPathFreeObject(obj); - obj = xmlXPathEval(BAD_CAST "string(/domain/devices/graphics[@type='vnc']/@listen)", ctxt); - if (obj == NULL || obj->type != XPATH_STRING || - obj->stringval == NULL || obj->stringval[0] == 0 || - STREQ((const char*)obj->stringval, "0.0.0.0")) { + listen_addr = virXPathString("string(/domain/devices/graphics" + "[@type='vnc']/@listen)", ctxt); + if (listen_addr == NULL || STREQ((const char *)listen_addr, "0.0.0.0")) { vshPrint(ctl, ":%d\n", port-5900); } else { - vshPrint(ctl, "%s:%d\n", (const char *)obj->stringval, port-5900); + vshPrint(ctl, "%s:%d\n", (const char *)listen_addr, port-5900); } - xmlXPathFreeObject(obj); - obj = NULL; ret = true; cleanup: - xmlXPathFreeObject(obj); xmlXPathFreeContext(ctxt); xmlFreeDoc(xml); virDomainFree(dom); -- 1.7.3.4

2 1

[libvirt] [PATCH] Add support for guest bind mounts with LXC
by Daniel P. Berrange 25 Jun '12

25 Jun '12

From: "Daniel P. Berrange" <berrange(a)redhat.com> Currently you can configure LXC to bind a host directory to a guest directory, but not to bind a guest directory to a guest directory. While the guest container init could do this itself, allowing it in the libvirt XML means a stricter SELinux policy can be written --- src/conf/domain_conf.c | 7 +++++-- src/conf/domain_conf.h | 11 ++++++----- src/lxc/lxc_container.c | 7 +++++++ 3 files changed, 18 insertions(+), 7 deletions(-) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index a653fe6..8ce122e 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -263,7 +263,8 @@ VIR_ENUM_IMPL(virDomainFS, VIR_DOMAIN_FS_TYPE_LAST, "block", "file", "template", - "ram") + "ram", + "bind") VIR_ENUM_IMPL(virDomainFSDriverType, VIR_DOMAIN_FS_DRIVER_TYPE_LAST, "default", @@ -4260,7 +4261,8 @@ virDomainFSDefParseXML(xmlNodePtr node, if (!source && xmlStrEqual(cur->name, BAD_CAST "source")) { - if (def->type == VIR_DOMAIN_FS_TYPE_MOUNT) + if (def->type == VIR_DOMAIN_FS_TYPE_MOUNT || + def->type == VIR_DOMAIN_FS_TYPE_BIND) source = virXMLPropString(cur, "dir"); else if (def->type == VIR_DOMAIN_FS_TYPE_FILE) source = virXMLPropString(cur, "file"); @@ -11335,6 +11337,7 @@ virDomainFSDefFormat(virBufferPtr buf, switch (def->type) { case VIR_DOMAIN_FS_TYPE_MOUNT: + case VIR_DOMAIN_FS_TYPE_BIND: virBufferEscapeString(buf, " <source dir='%s'/>\n", def->src); break; diff --git a/src/conf/domain_conf.h b/src/conf/domain_conf.h index b8d9c87..bdb1f75 100644 --- a/src/conf/domain_conf.h +++ b/src/conf/domain_conf.h @@ -656,11 +656,12 @@ struct _virDomainControllerDef { /* Two types of disk backends */ enum virDomainFSType { - VIR_DOMAIN_FS_TYPE_MOUNT, /* Better named 'bind' */ - VIR_DOMAIN_FS_TYPE_BLOCK, - VIR_DOMAIN_FS_TYPE_FILE, - VIR_DOMAIN_FS_TYPE_TEMPLATE, - VIR_DOMAIN_FS_TYPE_RAM, + VIR_DOMAIN_FS_TYPE_MOUNT, /* Mounts (binds) a host dir on a guest dir */ + VIR_DOMAIN_FS_TYPE_BLOCK, /* Mounts a host block dev on a guest dir */ + VIR_DOMAIN_FS_TYPE_FILE, /* Loopback mounts a host file on a guest dir */ + VIR_DOMAIN_FS_TYPE_TEMPLATE, /* Expands a OS template to a guest dir */ + VIR_DOMAIN_FS_TYPE_RAM, /* Mount a RAM filesystem on a guest dir */ + VIR_DOMAIN_FS_TYPE_BIND, /* Binds a guest dir to another guest dir */ VIR_DOMAIN_FS_TYPE_LAST }; diff --git a/src/lxc/lxc_container.c b/src/lxc/lxc_container.c index b69255e..bf67ba1 100644 --- a/src/lxc/lxc_container.c +++ b/src/lxc/lxc_container.c @@ -1025,7 +1025,14 @@ static int lxcContainerMountFS(virDomainFSDefPtr fs, if (lxcContainerMountFSTmpfs(fs) < 0) return -1; break; + case VIR_DOMAIN_FS_TYPE_BIND: + if (lxcContainerMountFSBind(fs, "") < 0) + return -1; + break; case VIR_DOMAIN_FS_TYPE_FILE: + /* We do actually support this, but the lxc controller + * should have associated the file with a loopback + * device and changed this to TYPE_BLOCK for us */ lxcError(VIR_ERR_INTERNAL_ERROR, _("Unexpected filesystem type %s"), virDomainFSTypeToString(fs->type)); -- 1.7.10.2

2 1

[libvirt] [PATCH] qemu: Support attaching controller device persistently
by Osier Yang 25 Jun '12

25 Jun '12

src/conf/domain_conf.*: (Introduce virDomainControllerFind) src/libvirt_private.syms: (Add virDomainControllerFind to private symbol) src/qemu/qemu_driver.c: (Use of virDomainControllerFind) src/qemu/qemu_driver.c: (Support attaching controller device persistently) --- src/conf/domain_conf.c | 17 +++++++++++++++++ src/conf/domain_conf.h | 2 +- src/libvirt_private.syms | 1 + src/qemu/qemu_driver.c | 17 +++++++++++++++++ src/qemu/qemu_hotplug.c | 14 +++++--------- 5 files changed, 41 insertions(+), 10 deletions(-) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 4651765..f2e8271 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -7520,6 +7520,23 @@ void virDomainControllerInsertPreAlloced(virDomainDefPtr def, def->ncontrollers++; } +bool +virDomainControllerFind(virDomainDefPtr def, + virDomainControllerDefPtr controller) +{ + int i; + bool found = false; + + for (i = 0 ; i < def->ncontrollers ; i++) { + if ((def->controllers[i]->type == controller->type) && + (def->controllers[i]->idx == controller->idx)) { + found = true; + break; + } + } + + return found; +} int virDomainLeaseIndex(virDomainDefPtr def, virDomainLeaseDefPtr lease) diff --git a/src/conf/domain_conf.h b/src/conf/domain_conf.h index 3712785..f7e7dca 100644 --- a/src/conf/domain_conf.h +++ b/src/conf/domain_conf.h @@ -2036,7 +2036,7 @@ int virDomainControllerInsert(virDomainDefPtr def, virDomainControllerDefPtr controller); void virDomainControllerInsertPreAlloced(virDomainDefPtr def, virDomainControllerDefPtr controller); - +bool virDomainControllerFind(virDomainDefPtr def, virDomainControllerDefPtr controller); int virDomainLeaseIndex(virDomainDefPtr def, virDomainLeaseDefPtr lease); diff --git a/src/libvirt_private.syms b/src/libvirt_private.syms index 2fe5068..a75086b 100644 --- a/src/libvirt_private.syms +++ b/src/libvirt_private.syms @@ -263,6 +263,7 @@ virDomainClockOffsetTypeFromString; virDomainClockOffsetTypeToString; virDomainConfigFile; virDomainControllerDefFree; +virDomainControllerFind; virDomainControllerInsert; virDomainControllerInsertPreAlloced; virDomainControllerModelSCSITypeFromString; diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 2177c30..5713650 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -5529,6 +5529,7 @@ qemuDomainAttachDeviceConfig(virDomainDefPtr vmdef, virDomainNetDefPtr net; virDomainHostdevDefPtr hostdev; virDomainLeaseDefPtr lease; + virDomainControllerDefPtr controller; switch (dev->type) { case VIR_DOMAIN_DEVICE_DISK: @@ -5600,6 +5601,22 @@ qemuDomainAttachDeviceConfig(virDomainDefPtr vmdef, dev->data.lease = NULL; break; + case VIR_DOMAIN_DEVICE_CONTROLLER: + controller = dev->data.controller; + if (virDomainControllerFind(vmdef, controller)) { + qemuReportError(VIR_ERR_INVALID_ARG, + _("Target already exists")); + return -1; + } + + if (virDomainControllerInsert(vmdef, controller) < 0) + return -1; + dev->data.controller = NULL; + + if (qemuDomainAssignAddresses(vmdef) < 0) + return -1; + break; + default: qemuReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", _("persistent attach of device is not supported")); diff --git a/src/qemu/qemu_hotplug.c b/src/qemu/qemu_hotplug.c index c2fa75b..a9d9ff9 100644 --- a/src/qemu/qemu_hotplug.c +++ b/src/qemu/qemu_hotplug.c @@ -308,21 +308,17 @@ int qemuDomainAttachPciControllerDevice(struct qemud_driver *driver, virDomainObjPtr vm, virDomainControllerDefPtr controller) { - int i; int ret = -1; const char* type = virDomainControllerTypeToString(controller->type); char *devstr = NULL; qemuDomainObjPrivatePtr priv = vm->privateData; bool releaseaddr = false; - for (i = 0 ; i < vm->def->ncontrollers ; i++) { - if ((vm->def->controllers[i]->type == controller->type) && - (vm->def->controllers[i]->idx == controller->idx)) { - qemuReportError(VIR_ERR_OPERATION_FAILED, - _("target %s:%d already exists"), - type, controller->idx); - return -1; - } + if (virDomainControllerFind(vm->def, controller)) { + qemuReportError(VIR_ERR_OPERATION_FAILED, + _("target %s:%d already exists"), + type, controller->idx); + return -1; } if (qemuCapsGet(priv->qemuCaps, QEMU_CAPS_DEVICE)) { -- 1.7.7.3

1 0

[libvirt] [PATCH] Add support for RAM filesystems for LXC
by Daniel P. Berrange 25 Jun '12

25 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='10' units='MiB'/> <target dir='/mnt'/> </filesystem> The usage units default to KiB 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 | 18 +++++++++++++ src/conf/domain_conf.c | 58 +++++++++++++++++++++++++++-------------- src/conf/domain_conf.h | 5 ++++ src/lxc/lxc_container.c | 49 ++++++++++++++++++++++++++++++++++ 5 files changed, 118 insertions(+), 21 deletions(-) diff --git a/docs/formatdomain.html.in b/docs/formatdomain.html.in index 1f30772..99ca26e 100644 --- a/docs/formatdomain.html.in +++ b/docs/formatdomain.html.in @@ -1630,6 +1630,12 @@ 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 kibibytes. + <span class="since"> (since 0.9.13)</span></dd> </dl> The filesystem block has an optional attribute <code>accessmode</code> @@ -1669,7 +1675,8 @@ 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 46e539d..4a04a65 100644 --- a/docs/schemas/domaincommon.rng +++ b/docs/schemas/domaincommon.rng @@ -1289,6 +1289,24 @@ </element> </interleave> </group> + <group> + <attribute name="type"> + <value>ram</value> + </attribute> + <interleave> + <element name="source"> + <attribute name="usage"> + <ref name="unsignedLong"/> + </attribute> + <optional> + <attribute name='unit'> + <ref name='unit'/> + </attribute> + </optional> + <empty/> + </element> + </interleave> + </group> </choice> <interleave> <element name="target"> diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 5ea264f..a653fe6 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -262,7 +262,8 @@ VIR_ENUM_IMPL(virDomainFS, VIR_DOMAIN_FS_TYPE_LAST, "mount", "block", "file", - "template") + "template", + "ram") VIR_ENUM_IMPL(virDomainFSDriverType, VIR_DOMAIN_FS_DRIVER_TYPE_LAST, "default", @@ -4210,6 +4211,7 @@ virDomainFSDefParseXML(xmlNodePtr node, char *target = NULL; char *accessmode = NULL; char *wrpolicy = NULL; + char *usage = NULL; ctxt->node = node; @@ -4297,7 +4299,8 @@ 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; @@ -4309,6 +4312,16 @@ virDomainFSDefParseXML(xmlNodePtr node, goto error; } + if (def->type == VIR_DOMAIN_FS_TYPE_RAM) { + def->usage = VIR_DOMAIN_FS_RAM_DEFAULT_USAGE; + if (virDomainParseScaledValue("./source[0]", ctxt, + &def->usage, + 1024, + ULONG_LONG_MAX, + false) < 0) + goto error; + } + def->src = source; source = NULL; def->dst = target; @@ -4325,6 +4338,7 @@ cleanup: VIR_FREE(source); VIR_FREE(accessmode); VIR_FREE(wrpolicy); + VIR_FREE(usage); return def; @@ -11319,27 +11333,31 @@ 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' units='KiB'/>\n", + def->usage); + break; } virBufferEscapeString(buf, " <target dir='%s'/>\n", diff --git a/src/conf/domain_conf.h b/src/conf/domain_conf.h index 86c1e63..b8d9c87 100644 --- a/src/conf/domain_conf.h +++ b/src/conf/domain_conf.h @@ -660,6 +660,7 @@ enum virDomainFSType { VIR_DOMAIN_FS_TYPE_BLOCK, VIR_DOMAIN_FS_TYPE_FILE, VIR_DOMAIN_FS_TYPE_TEMPLATE, + VIR_DOMAIN_FS_TYPE_RAM, VIR_DOMAIN_FS_TYPE_LAST }; @@ -690,11 +691,15 @@ enum virDomainFSWrpolicy { VIR_DOMAIN_FS_WRPOLICY_LAST }; +/* Allow 2 MB ram usage */ +#define VIR_DOMAIN_FS_RAM_DEFAULT_USAGE (1024 * 2) + struct _virDomainFSDef { int type; int fsdriver; int accessmode; int wrpolicy; /* enum virDomainFSWrpolicy */ + unsigned long long usage; char *src; char *dst; unsigned int readonly : 1; diff --git a/src/lxc/lxc_container.c b/src/lxc/lxc_container.c index 24b1017..b69255e 100644 --- a/src/lxc/lxc_container.c +++ b/src/lxc/lxc_container.c @@ -335,6 +335,8 @@ 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", @@ -966,6 +968,47 @@ 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", MS_NOSUID|MS_NODEV, 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) { @@ -978,6 +1021,10 @@ 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"), @@ -1441,6 +1488,8 @@ 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; -- 1.7.10.2

2 1

[libvirt] [PATCH] [libvirt-test-API] modify test dir and log dir path in autotest.
by Lin Qing 25 Jun '12

25 Jun '12

In autotest,test dir match test class name,which should be 'libvirt_test_api' for libvirt-test-api test suit.And the default tar extract is 'src'.So we should modify the autotest_testdir as 'tests/libvirt_test_api',and autotetst log dir as 'tests/libvirt_test_api/src/log'.Otherwise,libvirt-test-api can not work in autotest. Signed-off-by: Qing Lin<qinglbj(a)linux.vnet.ibm.com> Signed-off-by: Li Zhang<zhlcindy(a)linux.vnet.ibm.com> --- libvirt-test-api | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/libvirt-test-api b/libvirt-test-api index cec7679..017343c 100755 --- a/libvirt-test-api +++ b/libvirt-test-api @@ -141,8 +141,8 @@ class Main(object): testid = logname[-3:] log_xml_parser.add_test_xml(testrunid, testid) if os.environ.has_key('AUTODIR'): - autotest_testdir = os.path.join(os.environ['AUTODIR'],'tests/libvirt_test_API') - logfile = os.path.join('%s/libvirt-test-API/log/%s' % (autotest_testdir, testrunid), logname) + autotest_testdir = os.path.join(os.environ['AUTODIR'],'tests/libvirt_test_api') + logfile = os.path.join('%s/src/log/%s' % (autotest_testdir, testrunid), logname) else: logfile = os.path.join('log/%s' % testrunid, logname) procs.append(generator.FuncGen(cases_func_ref_dict, -- 1.7.10.2

2 2

[libvirt] libvirtd deadlock on shutdown
by Jim Fehlig 22 Jun '12

22 Jun '12

I'm looking into a libvirtd deadlock on daemon shutdown. The deadlock occurs when shutting down virNetServer. If the handling of a job is in flight in virNetServerHandleJob(), the virNetServer lock is acquired when freeing job->prog (src/rpc/virnetserver.c:167). But the lock is already held in virNetServerFree(), which is blocked in virThreadPoolFree() waiting for all the workers to finish. No progress can be made. The attached hack fixes the problem, but I'm not convinced this is an appropriate fix. Is it necessary to hold the virNetServer lock when calling virNetServerProgramFree(job->prog)? I notice the lock is not held in the error path of virNetServerHandleJob(). Thanks, Jim

5 11

[libvirt] qemu -numa option and non-contiguous CPU ranges
by Eduardo Habkost 22 Jun '12

22 Jun '12

Hi, I just noticed libvirt tries to use the -numa option in a way that qemu never understood: if a node is configured to have a non-contiguous set of CPUs, it tries to generate a command-line option that looks like: "-numa node,nodeid=...,cpus=0,2,4,mem=..." ^^^^^ But this format was never supported by qemu. This format is even a bit weird, as "," is an option separator, and it is being used as a separator _inside_ an option. My question is: should we support this option format in qemu, or should we change libvirt to use another format (that has yet to be implemented, because currently there's no way to specify a non-contiguous set of CPUs for a NUMA node). Any suggestions? -- Eduardo

3 3

[libvirt] [PATCHv2 0/5] Virtio support for S390
by Viktor Mihajlovski 22 Jun '12

22 Jun '12

Resending as a thread, same content. Please ignore previous submission. This series adds support for the s390 flavor of virtio devices. Since the s390 virtio devices are not implemented as PCI devices it is necessary to refactor some of the device address assignment code. Viktor Mihajlovski (5): S390: Extended qemuDomainAssignAddresses to be callable from everywhere. S390: Change tests to use (modified) qemuDomainAssignAddresses S390: Add support for virtio-s390 devices. S390: Domain Schema for s390-virtio machines. S390: Adding testcases for s390 docs/schemas/domaincommon.rng | 20 +++ src/conf/domain_conf.c | 11 +- src/conf/domain_conf.h | 1 + src/qemu/qemu_capabilities.c | 7 + src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_command.c | 140 ++++++++++++++++++-- src/qemu/qemu_command.h | 6 +- src/qemu/qemu_driver.c | 14 +- src/qemu/qemu_process.c | 42 +------ .../qemuxml2argv-console-virtio-s390.args | 9 ++ .../qemuxml2argv-console-virtio-s390.xml | 24 ++++ .../qemuxml2argv-disk-virtio-s390.args | 5 + .../qemuxml2argv-disk-virtio-s390.xml | 22 +++ .../qemuxml2argv-minimal-s390.args | 5 + .../qemuxml2argvdata/qemuxml2argv-minimal-s390.xml | 21 +++ .../qemuxml2argv-net-virtio-s390.args | 5 + .../qemuxml2argv-net-virtio-s390.xml | 22 +++ tests/qemuxml2argvtest.c | 20 ++-- tests/qemuxmlnstest.c | 13 +-- tests/testutilsqemu.c | 31 +++++ 20 files changed, 333 insertions(+), 86 deletions(-) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-console-virtio-s390.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-console-virtio-s390.xml create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-virtio-s390.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-virtio-s390.xml create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-minimal-s390.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-minimal-s390.xml create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-net-virtio-s390.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-net-virtio-s390.xml

1 5

[libvirt] [PATCH 0/5] Basic Enablement of s390
by Viktor Mihajlovski 22 Jun '12

22 Jun '12

Hi, although there's already some code in libvirt for qemu on s390 it has experienced a certain level of degradation over time. Or, to put it short: it is not compatible with upstream qemu/kvm. This is the first series of patches aimed to enable libvirt to manage qemu/kvm domains hosted on the s390 platform. The patches in this series address s390 specifics of node and hypervisor. Patches 2,3 and 5 are written by Thang Pham <thang.pham(a)us.ibm.com> the others are my doing. Since they all should go together, I am sending in Thang's as well, hoping to facilitate the review a bit this way. Thang Pham (3): S390: CPU support for s390(x) S390: Fixed Parser for /proc/cpuinfo needs to be adapted for your architecture S390: Added sysinfo for host on s390x. Viktor Mihajlovski (2): S390: Override QEMU_CAPS_NO_ACPI for s390x S390: Fixed core identification for s390 src/Makefile.am | 1 + src/cpu/cpu.c | 2 + src/cpu/cpu_s390x.c | 80 ++++++++++++++++++++++ src/cpu/cpu_s390x.h | 31 +++++++++ src/nodeinfo.c | 11 +++ src/qemu/qemu_capabilities.c | 5 ++ src/util/sysinfo.c | 154 ++++++++++++++++++++++++++++++++++++++++++ 7 files changed, 284 insertions(+), 0 deletions(-) create mode 100644 src/cpu/cpu_s390x.c create mode 100644 src/cpu/cpu_s390x.h -- Mit freundlichen Grüßen/Kind Regards Viktor Mihajlovski IBM Deutschland Research & Development GmbH Vorsitzender des Aufsichtsrats: Martin Jetter Geschäftsführung: Dirk Wittkopp Sitz der Gesellschaft: Böblingen Registergericht: Amtsgericht Stuttgart, HRB 243294

2 2

[libvirt] [PATCH][libvirt-sandbox] configure: Require higher version of glib
by Michal Privoznik 22 Jun '12

22 Jun '12

Since commit cfd4460b we must require glib-2.32.0 at least because we are using g_value_set_schar which was introduced in that release. --- configure.ac | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/configure.ac b/configure.ac index f8956c4..8d91440 100644 --- a/configure.ac +++ b/configure.ac @@ -11,7 +11,7 @@ AC_CANONICAL_HOST AM_SILENT_RULES([yes]) GIO_UNIX_REQUIRED=2.28.0 -GOBJECT_REQUIRED=2.28.0 +GOBJECT_REQUIRED=2.32.0 LIBVIRT_GLIB_REQUIRED=0.0.4 LIBVIRT_GOBJECT_REQUIRED=0.0.7 GOBJECT_INTROSPECTION_REQUIRED=0.10.8 -- 1.7.8.5

2 4

[libvirt] [PATCH 1/5] S390: Override QEMU_CAPS_NO_ACPI for s390x
by Viktor Mihajlovski 22 Jun '12

22 Jun '12

Starting a KVM guest on s390 fails immediately. This is because "qemu --help" reports -no-acpi even for the s390(x) architecture but -no-acpi isn't supported there. Workaround is to remove QEMU_CAPS_NO_ACPI from the capability set after the version/capability extraction. Signed-off-by: Viktor Mihajlovski <mihajlov(a)linux.vnet.ibm.com> --- src/qemu/qemu_capabilities.c | 5 +++++ 1 files changed, 5 insertions(+), 0 deletions(-) diff --git a/src/qemu/qemu_capabilities.c b/src/qemu/qemu_capabilities.c index 4308833..0c01cb0 100644 --- a/src/qemu/qemu_capabilities.c +++ b/src/qemu/qemu_capabilities.c @@ -1511,6 +1511,11 @@ int qemuCapsExtractVersionInfo(const char *qemu, const char *arch, qemuCapsSet(flags, QEMU_CAPS_PCI_MULTIBUS); } + /* S390 and probably other archs do not support no-acpi - + maybe the qemu option parsing should be re-thought. */ + if (STREQLEN(arch, "s390x", 5)) + qemuCapsClear(flags, QEMU_CAPS_NO_ACPI); + /* qemuCapsExtractDeviceStr will only set additional flags if qemu * understands the 0.13.0+ notion of "-device driver,". */ if (qemuCapsGet(flags, QEMU_CAPS_DEVICE) && -- 1.7.0.4 -- Mit freundlichen Grüßen/Kind Regards Viktor Mihajlovski IBM Deutschland Research & Development GmbH Vorsitzender des Aufsichtsrats: Martin Jetter Geschäftsführung: Dirk Wittkopp Sitz der Gesellschaft: Böblingen Registergericht: Amtsgericht Stuttgart, HRB 243294

3 3

[libvirt] [PATCH libvirt-glib] Add async version of gvir_domain_start
by Marc-André Lureau 22 Jun '12

22 Jun '12

--- libvirt-gobject/libvirt-gobject-domain.c | 75 ++++++++++++++++++++++++++++++ libvirt-gobject/libvirt-gobject-domain.h | 9 ++++ libvirt-gobject/libvirt-gobject.sym | 3 ++ 3 files changed, 87 insertions(+) diff --git a/libvirt-gobject/libvirt-gobject-domain.c b/libvirt-gobject/libvirt-gobject-domain.c index 59af63f..088cd33 100644 --- a/libvirt-gobject/libvirt-gobject-domain.c +++ b/libvirt-gobject/libvirt-gobject-domain.c @@ -345,6 +345,81 @@ gboolean gvir_domain_start(GVirDomain *dom, return TRUE; } +typedef struct { + guint flags; +} DomainStartData; + +static void domain_start_data_free(DomainStartData *data) +{ + g_slice_free(DomainStartData, data); +} + +static void +gvir_domain_start_helper(GSimpleAsyncResult *res, + GObject *object, + GCancellable *cancellable G_GNUC_UNUSED) +{ + GVirDomain *dom = GVIR_DOMAIN(object); + DomainStartData *data; + GError *err = NULL; + + data = g_simple_async_result_get_op_res_gpointer(res); + + if (!gvir_domain_start(dom, data->flags, &err)) + g_simple_async_result_take_error(res, err); +} + +/** + * gvir_domain_start_async: + * @dom: the domain + * @flags: the flags + * @cancellable: (allow-none)(transfer none): cancellation object + * @callback: (scope async): completion callback + * @user_data: (closure): opaque data for callback + * + * Asynchronous variant of #gvir_domain_start. + */ +void gvir_domain_start_async(GVirDomain *dom, + guint flags, + GCancellable *cancellable, + GAsyncReadyCallback callback, + gpointer user_data) +{ + GSimpleAsyncResult *res; + DomainStartData *data; + + g_return_if_fail(GVIR_IS_DOMAIN(dom)); + g_return_if_fail((cancellable == NULL) || G_IS_CANCELLABLE(cancellable)); + + data = g_slice_new0(DomainStartData); + data->flags = flags; + + res = g_simple_async_result_new(G_OBJECT(dom), + callback, + user_data, + gvir_domain_start_async); + g_simple_async_result_set_op_res_gpointer (res, data, (GDestroyNotify)domain_start_data_free); + g_simple_async_result_run_in_thread(res, + gvir_domain_start_helper, + G_PRIORITY_DEFAULT, + cancellable); + g_object_unref(res); +} + +gboolean gvir_domain_start_finish(GVirDomain *dom, + GAsyncResult *result, + GError **err) +{ + g_return_val_if_fail(GVIR_IS_DOMAIN(dom), FALSE); + g_return_val_if_fail(g_simple_async_result_is_valid(result, G_OBJECT(dom), gvir_domain_start_async), FALSE); + g_return_val_if_fail(err == NULL || *err == NULL, FALSE); + + if (g_simple_async_result_propagate_error(G_SIMPLE_ASYNC_RESULT(result), err)) + return FALSE; + + return TRUE; +} + /** * gvir_domain_resume: * @dom: the domain diff --git a/libvirt-gobject/libvirt-gobject-domain.h b/libvirt-gobject/libvirt-gobject-domain.h index 677fbe6..87b94f4 100644 --- a/libvirt-gobject/libvirt-gobject-domain.h +++ b/libvirt-gobject/libvirt-gobject-domain.h @@ -124,6 +124,15 @@ gint gvir_domain_get_id(GVirDomain *dom, gboolean gvir_domain_start(GVirDomain *dom, guint flags, GError **err); +void gvir_domain_start_async(GVirDomain *dom, + guint flags, + GCancellable *cancellable, + GAsyncReadyCallback callback, + gpointer user_data); +gboolean gvir_domain_start_finish(GVirDomain *dom, + GAsyncResult *result, + GError **err); + gboolean gvir_domain_resume(GVirDomain *dom, GError **err); gboolean gvir_domain_stop(GVirDomain *dom, diff --git a/libvirt-gobject/libvirt-gobject.sym b/libvirt-gobject/libvirt-gobject.sym index 7121794..94e441a 100644 --- a/libvirt-gobject/libvirt-gobject.sym +++ b/libvirt-gobject/libvirt-gobject.sym @@ -165,6 +165,9 @@ LIBVIRT_GOBJECT_0.0.9 { gvir_connection_get_capabilities; gvir_connection_get_capabilities_async; gvir_connection_get_capabilities_finish; + + gvir_domain_start_async; + gvir_domain_start_finish; } LIBVIRT_GOBJECT_0.0.8; # .... define new API here using predicted next version number .... -- 1.7.10.2

2 1