- Devel - libvirt List Archives

[libvirt] [PATCH] tests: run schema checks in sorted order
by Eric Blake 22 Dec '11

22 Dec '11

Having a test that depends on file system timestamps and/or inode allocation order is non-deterministic in its output. * tests/schematestutils.sh: Run test in deterministic order. --- Pushing under the trivial rule. Evidence of the non-determinism can be found at the location of the three failures reported here: https://www.redhat.com/archives/libvir-list/2011-December/msg00908.html tests/schematestutils.sh | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/tests/schematestutils.sh b/tests/schematestutils.sh index f2b3b50..ec6452a 100644 --- a/tests/schematestutils.sh +++ b/tests/schematestutils.sh @@ -13,7 +13,7 @@ for dir in $DIRS do XML=`find $abs_srcdir/$dir -name '*.xml'` || exit 1 - for xml in $XML + for xml in `echo $XML | sort` do n=`expr $n + 1` cmd="xmllint --relaxng $SCHEMA --noout $xml" -- 1.7.7.4

1 1

[libvirt] [PATCH] qemu: Support for overriding NOFILE limit
by Michal Privoznik 22 Dec '11

22 Dec '11

This patch adds max_files option to qemu.conf which can be used to override system default limit on number of opened files that are allowed for qemu user. --- src/qemu/libvirtd_qemu.aug | 1 + src/qemu/qemu.conf | 4 +++- src/qemu/qemu_conf.c | 4 ++++ src/qemu/qemu_conf.h | 1 + src/qemu/qemu_process.c | 16 ++++++++++++++-- src/qemu/test_libvirtd_qemu.aug | 4 ++++ 6 files changed, 27 insertions(+), 3 deletions(-) diff --git a/src/qemu/libvirtd_qemu.aug b/src/qemu/libvirtd_qemu.aug index ad34e42..078e9c4 100644 --- a/src/qemu/libvirtd_qemu.aug +++ b/src/qemu/libvirtd_qemu.aug @@ -50,6 +50,7 @@ module Libvirtd_qemu = | bool_entry "allow_disk_format_probing" | bool_entry "set_process_name" | int_entry "max_processes" + | int_entry "max_files" | str_entry "lock_manager" | int_entry "max_queued" | int_entry "keepalive_interval" diff --git a/src/qemu/qemu.conf b/src/qemu/qemu.conf index c3f264f..6125d3b 100644 --- a/src/qemu/qemu.conf +++ b/src/qemu/qemu.conf @@ -301,9 +301,11 @@ # If max_processes is set to a positive integer, libvirt will use it to set # maximum number of processes that can be run by qemu user. This can be used to -# override default value set by host OS. +# override default value set by host OS. The same apply to max_files which +# set limit to maximum number of opened files. # # max_processes = 0 +# max_files = 0 # To enable 'Sanlock' project based locking of the file # content (to prevent two VMs writing to the same diff --git a/src/qemu/qemu_conf.c b/src/qemu/qemu_conf.c index 3766119..bc0a646 100644 --- a/src/qemu/qemu_conf.c +++ b/src/qemu/qemu_conf.c @@ -443,6 +443,10 @@ int qemudLoadDriverConfig(struct qemud_driver *driver, CHECK_TYPE("max_processes", VIR_CONF_LONG); if (p) driver->maxProcesses = p->l; + p = virConfGetValue(conf, "max_files"); + CHECK_TYPE("max_files", VIR_CONF_LONG); + if (p) driver->maxFiles = p->l; + p = virConfGetValue (conf, "lock_manager"); CHECK_TYPE ("lock_manager", VIR_CONF_STRING); if (p && p->str) { diff --git a/src/qemu/qemu_conf.h b/src/qemu/qemu_conf.h index 8161269..f5a0f60 100644 --- a/src/qemu/qemu_conf.h +++ b/src/qemu/qemu_conf.h @@ -105,6 +105,7 @@ struct qemud_driver { unsigned int setProcessName : 1; int maxProcesses; + int maxFiles; int max_queued; diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index 9123f4c..105b895 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -2166,9 +2166,9 @@ qemuProcessPrepareChardevDevice(virDomainDefPtr def ATTRIBUTE_UNUSED, static int qemuProcessLimits(struct qemud_driver *driver) { - if (driver->maxProcesses > 0) { - struct rlimit rlim; + struct rlimit rlim; + if (driver->maxProcesses > 0) { rlim.rlim_cur = rlim.rlim_max = driver->maxProcesses; if (setrlimit(RLIMIT_NPROC, &rlim) < 0) { virReportSystemError(errno, @@ -2178,6 +2178,18 @@ qemuProcessLimits(struct qemud_driver *driver) } } + if (driver->maxFiles > 0) { + /* Max number of opened files is one greater than + * actual limit. See man setrlimit */ + rlim.rlim_cur = rlim.rlim_max = driver->maxFiles + 1; + if (setrlimit(RLIMIT_NOFILE, &rlim) < 0) { + virReportSystemError(errno, + _("cannot set max opened files to %d"), + driver->maxFiles); + return -1; + } + } + return 0; } diff --git a/src/qemu/test_libvirtd_qemu.aug b/src/qemu/test_libvirtd_qemu.aug index f7476ae..f841936 100644 --- a/src/qemu/test_libvirtd_qemu.aug +++ b/src/qemu/test_libvirtd_qemu.aug @@ -114,6 +114,8 @@ vnc_auto_unix_socket = 1 max_processes = 12345 +max_files = 67890 + lock_manager = \"fcntl\" keepalive_interval = 1 @@ -242,6 +244,8 @@ keepalive_count = 42 { "#empty" } { "max_processes" = "12345" } { "#empty" } +{ "max_files" = "67890" } +{ "#empty" } { "lock_manager" = "fcntl" } { "#empty" } { "keepalive_interval" = "1" } -- 1.7.3.4

2 2

[libvirt] [PATCH v2] virsh: Add two new commands to change media
by Osier Yang 22 Dec '11

22 Dec '11

eject-media to eject the media from, and insert-media to insert media into CDROM or floppy drive. --- Basically this patch is just rebaing of https://www.redhat.com/archives/libvir-list/2011-June/msg01516.html, with a bit improvement, as I explained in the mail thread, we don't have to wait for the new QEMU monitor commands, what we need is just separate the tray and media management in virsh, if one day QEMU gets the patch for those new commands pushed, we can switch to honor them easily (regardless of whether we do it inside updateDeviceFlags API or in new APIs). --- tools/virsh.c | 369 +++++++++++++++++++++++++++++++++++++++++++++++++++++++ tools/virsh.pod | 33 +++++ 2 files changed, 402 insertions(+), 0 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index 02f2e0d..228a9dc 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -13453,6 +13453,373 @@ cmdAttachDisk(vshControl *ctl, const vshCmd *cmd) } /* + * "eject-media" command + */ +static const vshCmdInfo info_eject_media[] = { + {"help", N_("eject media from CD or floppy drive")}, + {"desc", N_("Eject media from CD or floppy drive.")}, + {NULL, NULL} +}; + +static const vshCmdOptDef opts_eject_media[] = { + {"domain", VSH_OT_DATA, VSH_OFLAG_REQ, N_("domain name, id or uuid")}, + {"target", VSH_OT_DATA, VSH_OFLAG_REQ, N_("target of disk device")}, + {"current", VSH_OT_BOOL, 0, N_("can be either or both of --live and --config, " + "depends on implementation of hypervisor driver")}, + {"live", VSH_OT_BOOL, 0, N_("alter live configuration of running domain")}, + {"config", VSH_OT_BOOL, 0, N_("alter persistent configuration, effect observed on next boot")}, + {"force", VSH_OT_BOOL, 0, N_("force media ejection")}, + {NULL, 0, 0, NULL} +}; + +static bool +cmdEjectMedia(vshControl *ctl, const vshCmd *cmd) +{ + xmlDocPtr xml = NULL; + xmlXPathObjectPtr obj=NULL; + xmlXPathContextPtr ctxt = NULL; + xmlNodePtr cur = NULL; + xmlBufferPtr xml_buf = NULL; + virDomainPtr dom = NULL; + const char *target = NULL; + char *doc; + int i = 0, diff_tgt; + int ret; + bool functionReturn = false; + int flags = 0; + int config = vshCommandOptBool(cmd, "config"); + int live = vshCommandOptBool(cmd, "live"); + int current = vshCommandOptBool(cmd, "current"); + int force = vshCommandOptBool(cmd, "force"); + bool has_source = false; + + if (current) { + if (live || config) { + vshError(ctl, "%s", _("--current must be specified exclusively")); + return false; + } + flags = VIR_DOMAIN_AFFECT_CURRENT; + } else { + if (config) + flags |= VIR_DOMAIN_AFFECT_CONFIG; + if (live) + flags |= VIR_DOMAIN_AFFECT_LIVE; + } + + if (force) + flags |= VIR_DOMAIN_DEVICE_MODIFY_FORCE; + + if (!vshConnectionUsability(ctl, ctl->conn)) + goto cleanup; + + if (!(dom = vshCommandOptDomain(ctl, cmd, NULL))) + goto cleanup; + + if (vshCommandOptString(cmd, "target", &target) <= 0) + goto cleanup; + + if (flags & VIR_DOMAIN_AFFECT_CONFIG) + doc = virDomainGetXMLDesc(dom, VIR_DOMAIN_XML_INACTIVE); + else + doc = virDomainGetXMLDesc(dom, 0); + if (!doc) + goto cleanup; + + xml = xmlReadDoc((const xmlChar *) doc, "domain.xml", NULL, + XML_PARSE_NOENT | XML_PARSE_NONET | + XML_PARSE_NOWARNING); + VIR_FREE(doc); + if (!xml) { + vshError(ctl, "%s", _("Failed to get disk information")); + goto cleanup; + } + ctxt = xmlXPathNewContext(xml); + if (!ctxt) { + vshError(ctl, "%s", _("Failed to get disk information")); + goto cleanup; + } + + obj = xmlXPathEval(BAD_CAST "/domain/devices/disk", ctxt); + if ((obj == NULL) || (obj->type != XPATH_NODESET) || + (obj->nodesetval == NULL) || (obj->nodesetval->nodeNr == 0)) { + vshError(ctl, "%s", _("Failed to get disk information")); + goto cleanup; + } + + /* search target */ + for (; i < obj->nodesetval->nodeNr; i++) { + xmlNodePtr n = obj->nodesetval->nodeTab[i]; + bool is_supported = false; + + /* Check if the disk is CDROM or floppy disk */ + if (xmlStrEqual(n->name, BAD_CAST "disk")) { + char *device_value = virXMLPropString(n, "device"); + + if (STREQ(device_value, "cdrom") || + STREQ(device_value, "floppy")) + is_supported = true; + + VIR_FREE(device_value); + } + + cur = obj->nodesetval->nodeTab[i]->children; + while (cur != NULL) { + if (cur->type == XML_ELEMENT_NODE && + xmlStrEqual(cur->name, BAD_CAST "target")) { + char *tmp_tgt = virXMLPropString(cur, "dev"); + + diff_tgt = STREQ(tmp_tgt, target); + VIR_FREE(tmp_tgt); + + if (diff_tgt && is_supported) { + goto hit; + } + } + + cur = cur->next; + } + } + vshError(ctl, _("No found CDROM or floppy disk whose target is %s"), target); + goto cleanup; + + hit: + xml_buf = xmlBufferCreate(); + if (!xml_buf) { + vshError(ctl, "%s", _("Failed to allocate memory")); + goto cleanup; + } + + cur = obj->nodesetval->nodeTab[i]->children; + while (cur != NULL) { + if (cur->type == XML_ELEMENT_NODE && + xmlStrEqual(cur->name, BAD_CAST "source")) { + xmlUnlinkNode(cur); + xmlFreeNode(cur); + has_source = true; + break; + } + + cur = cur->next; + } + + if (!has_source) { + vshError(ctl, _("The disk device whose target is '%s' doesn't " + "have media"), target); + goto cleanup; + } + + if (xmlNodeDump(xml_buf, xml, obj->nodesetval->nodeTab[i], 0, 0) < 0) { + vshError(ctl, "%s", _("Failed to create XML")); + goto cleanup; + } + + ret = virDomainUpdateDeviceFlags(dom, (char *)xmlBufferContent(xml_buf), flags); + + if (ret != 0) { + vshError(ctl, "%s", _("Failed to eject media")); + } else { + vshPrint(ctl, "%s", _("Media ejected successfully\n")); + functionReturn = true; + } + + cleanup: + xmlXPathFreeObject(obj); + xmlXPathFreeContext(ctxt); + xmlFreeDoc(xml); + xmlBufferFree(xml_buf); + if (dom) + virDomainFree(dom); + return functionReturn; +} + +/* + * "insert-media" command + */ +static const vshCmdInfo info_insert_media[] = { + {"help", N_("insert media into CD or floppy drive")}, + {"desc", N_("Insert media into CD or floppy drive.")}, + {NULL, NULL} +}; + +static const vshCmdOptDef opts_insert_media[] = { + {"domain", VSH_OT_DATA, VSH_OFLAG_REQ, N_("domain name, id or uuid")}, + {"target", VSH_OT_DATA, VSH_OFLAG_REQ, N_("target of disk device")}, + {"source", VSH_OT_DATA, VSH_OFLAG_REQ, N_("source of the media")}, + {"current", VSH_OT_BOOL, 0, N_("can be either or both of --live and --config, " + "depends on implementation of hypervisor driver")}, + {"live", VSH_OT_BOOL, 0, N_("alter live configuration of running domain")}, + {"config", VSH_OT_BOOL, 0, N_("alter persistent configuration, effect observed on next boot")}, + {"force", VSH_OT_BOOL, 0, N_("force media insertion")}, + {NULL, 0, 0, NULL} +}; + +static bool +cmdInsertMedia(vshControl *ctl, const vshCmd *cmd) +{ + xmlDocPtr xml = NULL; + xmlXPathObjectPtr obj=NULL; + xmlXPathContextPtr ctxt = NULL; + xmlNodePtr cur = NULL; + xmlNodePtr new_node = NULL; + xmlBufferPtr xml_buf = NULL; + virDomainPtr dom = NULL; + const char *target = NULL; + const char *source = NULL; + char *doc; + int i = 0, diff_tgt; + int ret; + bool functionReturn = false; + int flags = 0; + const char *disk_type = NULL; + int config = vshCommandOptBool(cmd, "config"); + int live = vshCommandOptBool(cmd, "live"); + int current = vshCommandOptBool(cmd, "current"); + int force = vshCommandOptBool(cmd, "force"); + + if (current) { + if (live || config) { + vshError(ctl, "%s", _("--current must be specified exclusively")); + return false; + } + flags = VIR_DOMAIN_AFFECT_CURRENT; + } else { + if (config) + flags |= VIR_DOMAIN_AFFECT_CONFIG; + if (live) + flags |= VIR_DOMAIN_AFFECT_LIVE; + } + + if (force) + flags |= VIR_DOMAIN_DEVICE_MODIFY_FORCE; + + if (!vshConnectionUsability(ctl, ctl->conn)) + goto cleanup; + + if (!(dom = vshCommandOptDomain(ctl, cmd, NULL))) + goto cleanup; + + if (vshCommandOptString(cmd, "target", &target) <= 0) + goto cleanup; + + if (vshCommandOptString(cmd, "source", &source) <= 0) + goto cleanup; + + if (flags & VIR_DOMAIN_AFFECT_CONFIG) + doc = virDomainGetXMLDesc(dom, VIR_DOMAIN_XML_INACTIVE); + else + doc = virDomainGetXMLDesc(dom, 0); + if (!doc) + goto cleanup; + + xml = xmlReadDoc((const xmlChar *) doc, "domain.xml", NULL, + XML_PARSE_NOENT | XML_PARSE_NONET | + XML_PARSE_NOWARNING); + VIR_FREE(doc); + if (!xml) { + vshError(ctl, "%s", _("Failed to get disk information")); + goto cleanup; + } + ctxt = xmlXPathNewContext(xml); + if (!ctxt) { + vshError(ctl, "%s", _("Failed to get disk information")); + goto cleanup; + } + + obj = xmlXPathEval(BAD_CAST "/domain/devices/disk", ctxt); + if ((obj == NULL) || (obj->type != XPATH_NODESET) || + (obj->nodesetval == NULL) || (obj->nodesetval->nodeNr == 0)) { + vshError(ctl, "%s", _("Failed to get disk information")); + goto cleanup; + } + + /* search target */ + for (; i < obj->nodesetval->nodeNr; i++) { + xmlNodePtr n = obj->nodesetval->nodeTab[i]; + bool is_supported = false; + + /* Check if the disk is CDROM or floppy disk */ + if (xmlStrEqual(n->name, BAD_CAST "disk")) { + disk_type = virXMLPropString(n, "type"); + char *device_value = virXMLPropString(n, "device"); + + if (STREQ(device_value, "cdrom") || + STREQ(device_value, "floppy")) + is_supported = true; + + VIR_FREE(device_value); + } + + cur = obj->nodesetval->nodeTab[i]->children; + while (cur != NULL) { + if (cur->type == XML_ELEMENT_NODE && + xmlStrEqual(cur->name, BAD_CAST "target")) { + char *tmp_tgt = virXMLPropString(cur, "dev"); + + diff_tgt = STREQ(tmp_tgt, target); + VIR_FREE(tmp_tgt); + + if (diff_tgt && is_supported) { + goto hit; + } + } + + cur = cur->next; + } + VIR_FREE(disk_type); + } + vshError(ctl, _("No found CDROM or floppy disk whose target is %s"), target); + goto cleanup; + + hit: + xml_buf = xmlBufferCreate(); + if (!xml_buf) { + vshError(ctl, "%s", _("Failed to allocate memory")); + goto cleanup; + } + + cur = obj->nodesetval->nodeTab[i]->children; + while (cur != NULL) { + if (cur->type == XML_ELEMENT_NODE && + xmlStrEqual(cur->name, BAD_CAST "source")) { + vshError(ctl, _("The disk device whose target is '%s' has media, " + "you may need to eject it first"), target); + goto cleanup; + } + + cur = cur->next; + } + + /* Insert node <source> */ + new_node = xmlNewNode(NULL, BAD_CAST "source"); + xmlNewProp(new_node, (const xmlChar *)disk_type, (const xmlChar *)source); + VIR_FREE(disk_type); + xmlAddChild(obj->nodesetval->nodeTab[i], new_node); + + if (xmlNodeDump(xml_buf, xml, obj->nodesetval->nodeTab[i], 0, 0) < 0) { + vshError(ctl, "%s", _("Failed to create XML")); + goto cleanup; + } + + ret = virDomainUpdateDeviceFlags(dom, (char *)xmlBufferContent(xml_buf), flags); + + if (ret != 0) { + vshError(ctl, "%s", _("Failed to insert media")); + } else { + vshPrint(ctl, "%s", _("Media inserted successfully\n")); + functionReturn = true; + } + + cleanup: + xmlXPathFreeObject(obj); + xmlXPathFreeContext(ctxt); + xmlFreeDoc(xml); + xmlBufferFree(xml_buf); + if (dom) + virDomainFree(dom); + return functionReturn; +} + +/* * "detach-disk" command */ static const vshCmdInfo info_detach_disk[] = { @@ -15600,7 +15967,9 @@ static const vshCmdDef domManagementCmds[] = { {"dump", cmdDump, opts_dump, info_dump, 0}, {"dumpxml", cmdDumpXML, opts_dumpxml, info_dumpxml, 0}, {"edit", cmdEdit, opts_edit, info_edit, 0}, + {"eject-media", cmdEjectMedia, opts_eject_media, info_eject_media, 0}, {"inject-nmi", cmdInjectNMI, opts_inject_nmi, info_inject_nmi, 0}, + {"insert-media", cmdInsertMedia, opts_insert_media, info_insert_media, 0}, {"send-key", cmdSendKey, opts_send_key, info_send_key}, {"managedsave", cmdManagedSave, opts_managedsave, info_managedsave, 0}, {"managedsave-remove", cmdManagedSaveRemove, opts_managedsaveremove, diff --git a/tools/virsh.pod b/tools/virsh.pod index 6c10245..45a3fde 100644 --- a/tools/virsh.pod +++ b/tools/virsh.pod @@ -1387,6 +1387,39 @@ option can be used to force device update, e.g., to eject a CD-ROM even if it is locked/mounted in the domain. See the documentation to learn about libvirt XML format for a device. +=item B<eject-media> I<domain-id> I<target> [I<--current>] [I<--live>] +[I<--config>] [I<--force>] + +Eject media from CD or floppy drive. The I<target> is the device as seen +from the domain. + +If I<--live> is specified, alter live configuration of running guest. +If I<--config> is specified, alter persistent configuration, effect observed +on next boot. +I<--current> can be either or both of I<live> and I<config>, depends on +the hypervisor's implementation. +Both I<--live> and I<--config> flags may be given, but I<--current> is +exclusive. If no flag is specified, behavior is different depending +on hypervisor. +The I<--force> option can be used to force media ejection. + +=item B<insert-media> I<domain-id> I<target> I<source> [I<--current>] +[I<--live>] [I<--config>] [I<--force>] + +Insert media into CD or floppy drive. The I<target> is the device as seen +from the domain. The I<source> specifies the path of the media which is to +be inserted. + +If I<--live> is specified, alter live configuration of running guest. +If I<--config> is specified, alter persistent configuration, effect observed +on next boot. +I<--current> can be either or both of I<live> and I<config>, depends on +the hypervisor's implementation. +Both I<--live> and I<--config> flags may be given, but I<--current> is +exclusive. If no flag is specified, behavior is different depending +on hypervisor. +The I<--force> option can be used to force media ejection. + =back =head1 NODEDEV COMMANDS -- 1.7.7.3

1 0

[libvirt] [PATCH] qemu: Support copy on read for disk
by Osier Yang 22 Dec '11

22 Dec '11

Upstream QEMU supports it since commit fb0490f69. --- docs/formatdomain.html.in | 9 +++++++++ docs/schemas/domaincommon.rng | 11 +++++++++++ src/conf/domain_conf.c | 24 +++++++++++++++++++++++- src/conf/domain_conf.h | 10 ++++++++++ src/libvirt_private.syms | 2 ++ src/qemu/qemu_capabilities.c | 3 +++ src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_command.c | 11 +++++++++++ 8 files changed, 70 insertions(+), 1 deletions(-) diff --git a/docs/formatdomain.html.in b/docs/formatdomain.html.in index 9cf0f12..779317f 100644 --- a/docs/formatdomain.html.in +++ b/docs/formatdomain.html.in @@ -1187,6 +1187,15 @@ <b>In general you should leave this option alone, unless you are very certain you know what you are doing.</b> </li> + <li> + The optional <code>copy_on_read</code> attribute controls + wether to copy read backing file into the image file. The + value can be either "on" or "off". + Copy-on-read avoids accessing the same backing file sectors + repeatedly and is useful when the backing file is over a slow + network. By default copy-on-read is off. + <span class='since'>Since 0.9.9 (QEMU and KVM only)</span> + </li> </ul> </dd> <dt><code>boot</code></dt> diff --git a/docs/schemas/domaincommon.rng b/docs/schemas/domaincommon.rng index 553a6f0..b511ec2 100644 --- a/docs/schemas/domaincommon.rng +++ b/docs/schemas/domaincommon.rng @@ -959,6 +959,14 @@ </choice> </attribute> </define> + <define name="copy_on_read"> + <attribute name='copy_on_read'> + <choice> + <value>on</value> + <value>off</value> + </choice> + </attribute> + </define> <define name="controller"> <element name="controller"> <choice> @@ -1308,6 +1316,9 @@ <optional> <ref name="event_idx"/> </optional> + <optional> + <ref name="copy_on_read"/> + </optional> <empty/> </element> </optional> diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 2897b4a..a51eb40 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -208,6 +208,11 @@ VIR_ENUM_IMPL(virDomainVirtioEventIdx, VIR_DOMAIN_VIRTIO_EVENT_IDX_LAST, "on", "off") +VIR_ENUM_IMPL(virDomainDiskCopyOnRead, VIR_DOMAIN_DISK_COPY_ON_READ_LAST, + "default", + "on", + "off") + VIR_ENUM_IMPL(virDomainDiskSnapshot, VIR_DOMAIN_DISK_SNAPSHOT_LAST, "default", "no", @@ -2621,6 +2626,7 @@ virDomainDiskDefParseXML(virCapsPtr caps, char *iotag = NULL; char *ioeventfd = NULL; char *event_idx = NULL; + char *copy_on_read = NULL; char *devaddr = NULL; virStorageEncryptionPtr encryption = NULL; char *serial = NULL; @@ -2749,6 +2755,7 @@ virDomainDiskDefParseXML(virCapsPtr caps, iotag = virXMLPropString(cur, "io"); ioeventfd = virXMLPropString(cur, "ioeventfd"); event_idx = virXMLPropString(cur, "event_idx"); + copy_on_read = virXMLPropString(cur, "copy_on_read"); } else if (xmlStrEqual(cur->name, BAD_CAST "auth")) { authUsername = virXMLPropString(cur, "username"); if (authUsername == NULL) { @@ -3053,6 +3060,17 @@ virDomainDiskDefParseXML(virCapsPtr caps, def->event_idx = idx; } + if (copy_on_read) { + int cor; + if ((cor = virDomainDiskCopyOnReadTypeFromString(copy_on_read)) <= 0) { + virDomainReportError(VIR_ERR_CONFIG_UNSUPPORTED, + _("unknown disk copy_on_read mode '%s'"), + copy_on_read); + goto error; + } + def->copy_on_read = cor; + } + if (devaddr) { if (virDomainParseLegacyDeviceAddress(devaddr, &def->info.addr.pci) < 0) { @@ -3146,6 +3164,7 @@ cleanup: VIR_FREE(iotag); VIR_FREE(ioeventfd); VIR_FREE(event_idx); + VIR_FREE(copy_on_read); VIR_FREE(devaddr); VIR_FREE(serial); virStorageEncryptionFree(encryption); @@ -9768,6 +9787,7 @@ virDomainDiskDefFormat(virBufferPtr buf, const char *iomode = virDomainDiskIoTypeToString(def->iomode); const char *ioeventfd = virDomainIoEventFdTypeToString(def->ioeventfd); const char *event_idx = virDomainVirtioEventIdxTypeToString(def->event_idx); + const char *copy_on_read = virDomainVirtioEventIdxTypeToString(def->copy_on_read); const char *startupPolicy = virDomainStartupPolicyTypeToString(def->startupPolicy); char uuidstr[VIR_UUID_STRING_BUFLEN]; @@ -9808,7 +9828,7 @@ virDomainDiskDefFormat(virBufferPtr buf, virBufferAddLit(buf, ">\n"); if (def->driverName || def->driverType || def->cachemode || - def->ioeventfd || def->event_idx) { + def->ioeventfd || def->event_idx || def->copy_on_read) { virBufferAddLit(buf, " <driver"); if (def->driverName) virBufferAsprintf(buf, " name='%s'", def->driverName); @@ -9826,6 +9846,8 @@ virDomainDiskDefFormat(virBufferPtr buf, virBufferAsprintf(buf, " ioeventfd='%s'", ioeventfd); if (def->event_idx) virBufferAsprintf(buf, " event_idx='%s'", event_idx); + if (def->copy_on_read) + virBufferAsprintf(buf, " copy_on_read='%s'", copy_on_read); virBufferAddLit(buf, "/>\n"); } diff --git a/src/conf/domain_conf.h b/src/conf/domain_conf.h index 1f6e442..4575ab2 100644 --- a/src/conf/domain_conf.h +++ b/src/conf/domain_conf.h @@ -288,6 +288,14 @@ enum virDomainVirtioEventIdx { VIR_DOMAIN_VIRTIO_EVENT_IDX_LAST }; +enum virDomainDiskCopyOnRead { + VIR_DOMAIN_DISK_COPY_ON_READ_DEFAULT = 0, + VIR_DOMAIN_DISK_COPY_ON_READ_ON, + VIR_DOMAIN_DISK_COPY_ON_READ_OFF, + + VIR_DOMAIN_DISK_COPY_ON_READ_LAST +}; + enum virDomainDiskSnapshot { VIR_DOMAIN_DISK_SNAPSHOT_DEFAULT = 0, VIR_DOMAIN_DISK_SNAPSHOT_NO, @@ -363,6 +371,7 @@ struct _virDomainDiskDef { int iomode; int ioeventfd; int event_idx; + int copy_on_read; int snapshot; /* enum virDomainDiskSnapshot */ int startupPolicy; /* enum virDomainStartupPolicy */ unsigned int readonly : 1; @@ -1968,6 +1977,7 @@ VIR_ENUM_DECL(virDomainDiskSecretType) VIR_ENUM_DECL(virDomainDiskSnapshot) VIR_ENUM_DECL(virDomainIoEventFd) VIR_ENUM_DECL(virDomainVirtioEventIdx) +VIR_ENUM_DECL(virDomainDiskCopyOnRead) VIR_ENUM_DECL(virDomainController) VIR_ENUM_DECL(virDomainControllerModelSCSI) VIR_ENUM_DECL(virDomainControllerModelUSB) diff --git a/src/libvirt_private.syms b/src/libvirt_private.syms index 4a86bdc..63c2947 100644 --- a/src/libvirt_private.syms +++ b/src/libvirt_private.syms @@ -290,6 +290,8 @@ virDomainDeviceTypeToString; virDomainDiskBusTypeToString; virDomainDiskCacheTypeFromString; virDomainDiskCacheTypeToString; +virDomainDiskCopyOnReadTypeFromString; +virDomainDiskCopyOnReadTypeToString; virDomainDiskDefAssignAddress; virDomainDiskDefForeachPath; virDomainDiskDefFree; diff --git a/src/qemu/qemu_capabilities.c b/src/qemu/qemu_capabilities.c index 43c7578..f32c792 100644 --- a/src/qemu/qemu_capabilities.c +++ b/src/qemu/qemu_capabilities.c @@ -144,6 +144,7 @@ VIR_ENUM_IMPL(qemuCaps, QEMU_CAPS_LAST, "ich9-ahci", "no-acpi", "fsdev-readonly", + "drive-copy-on-read", ); struct qemu_feature_flags { @@ -1023,6 +1024,8 @@ qemuCapsComputeCmdFlags(const char *help, qemuCapsSet(flags, QEMU_CAPS_DRIVE_READONLY); if (strstr(help, "aio=threads|native")) qemuCapsSet(flags, QEMU_CAPS_DRIVE_AIO); + if (strstr(help, "copy-on-read=on|off")) + qemuCapsSet(flags, QEMU_CAPS_DRIVE_COPY_ON_READ); } if ((p = strstr(help, "-vga")) && !strstr(help, "-std-vga")) { const char *nl = strstr(p, "\n"); diff --git a/src/qemu/qemu_capabilities.h b/src/qemu/qemu_capabilities.h index c759baf..38106dd 100644 --- a/src/qemu/qemu_capabilities.h +++ b/src/qemu/qemu_capabilities.h @@ -117,6 +117,7 @@ enum qemuCapsFlags { QEMU_CAPS_ICH9_AHCI = 77, /* -device ich9-ahci */ QEMU_CAPS_NO_ACPI = 78, /* -no-acpi */ QEMU_CAPS_FSDEV_READONLY =79, /* -fsdev readonly supported */ + QEMU_CAPS_DRIVE_COPY_ON_READ = 80, /* -drive copy-on-read */ QEMU_CAPS_LAST, /* this must always be the last item */ }; diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index ea1b763..629f13d 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -1926,6 +1926,17 @@ qemuBuildDriveStr(virConnectPtr conn ATTRIBUTE_UNUSED, virBufferAddLit(&opt, ",cache=off"); } + if (disk->copy_on_read) { + if (qemuCapsGet(qemuCaps, QEMU_CAPS_DRIVE_COPY_ON_READ)) { + virBufferAsprintf(&opt, ",copy-on-read=%s", + virDomainDiskCopyOnReadTypeToString(disk->copy_on_read)); + } else { + qemuReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", + _("copy_on_read is not supported by this QEMU binary")); + goto error; + } + } + if (qemuCapsGet(qemuCaps, QEMU_CAPS_MONITOR_JSON)) { const char *wpolicy = NULL, *rpolicy = NULL; -- 1.7.7.3

2 2

[libvirt] [PATCH] qemuhelptest: Add new qemuCap flag
by Michal Privoznik 22 Dec '11

22 Dec '11

Latest patch a1a83c587443 introduces new qemu capability flag QEMU_CAPS_FSDEV_READONLY. However, it was missing in qemuhelptest making test for qemu-1.0 fail. --- Pushing under build-breaker rule. tests/qemuhelptest.c | 3 ++- 1 files changed, 2 insertions(+), 1 deletions(-) diff --git a/tests/qemuhelptest.c b/tests/qemuhelptest.c index f31c903..60155e7 100644 --- a/tests/qemuhelptest.c +++ b/tests/qemuhelptest.c @@ -647,7 +647,8 @@ mymain(void) QEMU_CAPS_NO_SHUTDOWN, QEMU_CAPS_PCI_ROMBAR, QEMU_CAPS_ICH9_AHCI, - QEMU_CAPS_NO_ACPI); + QEMU_CAPS_NO_ACPI, + QEMU_CAPS_FSDEV_READONLY); return ret == 0 ? EXIT_SUCCESS : EXIT_FAILURE; } -- 1.7.3.4

2 1

[libvirt] [PATCH] Add new attribute writeout to <filesystem> element
by Deepak C Shetty 22 Dec '11

22 Dec '11

This introduces new attribute writeout with only supported value as immediate. This will be an optional attribute with no defaults. This helps specify whether to skip the host page cache. When writeout is specified, meaning when writeout=immediate a writeback is explicitly initiated for the dirty pages in the host page cache as part of the guest file write operation. Usage: <filesystem type='mount' accessmode='passthrough' writeout='immediate'> <source dir='/export/to/guest'/> <target dir='mount_tag'/> </filesystem> Currently this only works with type='mount' for the QEMU/KVM driver. Signed-off-by: Deepak C Shetty <deepakcs(a)linux.vnet.ibm.com> --- docs/formatdomain.html.in | 8 +++++++- docs/schemas/domaincommon.rng | 5 +++++ src/conf/domain_conf.c | 29 +++++++++++++++++++++++++++-- src/conf/domain_conf.h | 10 ++++++++++ src/qemu/qemu_command.c | 9 +++++++++ 5 files changed, 58 insertions(+), 3 deletions(-) diff --git a/docs/formatdomain.html.in b/docs/formatdomain.html.in index c57b7b3..803db8e 100644 --- a/docs/formatdomain.html.in +++ b/docs/formatdomain.html.in @@ -1303,7 +1303,7 @@ <source name='my-vm-template'/> <target dir='/'/> </filesystem> - <filesystem type='mount' accessmode='passthrough'> + <filesystem type='mount' accessmode='passthrough' writeout='immediate'> <driver type='path'/> <source dir='/export/to/guest'/> <target dir='/import/from/host'/> @@ -1379,6 +1379,12 @@ </dd> </dl> + The filesystem block has an optional attribute <code>writeout='immediate'</code> + to help specify whether to skip the host page cache. When writeout is specified, meaning when + writeout=immediate a writeback is explicitly initiated for the dirty pages in + the host page cache as part of the guest file write operation. When this attribute is not + specified, there are no defaults, meaning explicit writeback won't be initiated. + </dd> <dt><code>source</code></dt> diff --git a/docs/schemas/domaincommon.rng b/docs/schemas/domaincommon.rng index 27ec1da..2298994 100644 --- a/docs/schemas/domaincommon.rng +++ b/docs/schemas/domaincommon.rng @@ -1106,6 +1106,11 @@ <value>squash</value> </choice> </attribute> + <attribute name="writeout"> + <choice> + <value>immediate</value> + </choice> + </attribute> </optional> </element> </define> diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 2c91f82..80b8eab 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -256,6 +256,9 @@ VIR_ENUM_IMPL(virDomainFSAccessMode, VIR_DOMAIN_FS_ACCESSMODE_LAST, "mapped", "squash") +VIR_ENUM_IMPL(virDomainFSWriteout, VIR_DOMAIN_FS_WRITEOUT_LAST, + "default", + "immediate") VIR_ENUM_IMPL(virDomainNet, VIR_DOMAIN_NET_TYPE_LAST, "user", @@ -3258,6 +3261,7 @@ virDomainFSDefParseXML(xmlNodePtr node, char *source = NULL; char *target = NULL; char *accessmode = NULL; + char *writeout = NULL; if (VIR_ALLOC(def) < 0) { virReportOOMError(); @@ -3286,6 +3290,17 @@ virDomainFSDefParseXML(xmlNodePtr node, def->accessmode = VIR_DOMAIN_FS_ACCESSMODE_PASSTHROUGH; } + writeout = virXMLPropString(node, "writeout"); + if (writeout) { + if ((def->writeout = virDomainFSWriteoutTypeFromString(writeout)) < 0) { + virDomainReportError(VIR_ERR_INTERNAL_ERROR, + _("unknown filesystem writeout '%s'"), writeout); + goto error; + } + } else { + def->writeout = VIR_DOMAIN_FS_WRITEOUT_DEFAULT; + } + cur = node->children; while (cur != NULL) { if (cur->type == XML_ELEMENT_NODE) { @@ -3348,6 +3363,7 @@ cleanup: VIR_FREE(target); VIR_FREE(source); VIR_FREE(accessmode); + VIR_FREE(writeout); return def; @@ -10006,6 +10022,7 @@ virDomainFSDefFormat(virBufferPtr buf, const char *type = virDomainFSTypeToString(def->type); const char *accessmode = virDomainFSAccessModeTypeToString(def->accessmode); const char *fsdriver = virDomainFSDriverTypeTypeToString(def->fsdriver); + const char *writeout = virDomainFSWriteoutTypeToString(def->writeout); if (!type) { virDomainReportError(VIR_ERR_INTERNAL_ERROR, @@ -10022,12 +10039,20 @@ virDomainFSDefFormat(virBufferPtr buf, if (def->fsdriver == VIR_DOMAIN_FS_DRIVER_TYPE_PATH || def->fsdriver == VIR_DOMAIN_FS_DRIVER_TYPE_DEFAULT) { virBufferAsprintf(buf, - " <filesystem type='%s' accessmode='%s'>\n", + " <filesystem type='%s' accessmode='%s'", type, accessmode); } else { virBufferAsprintf(buf, - " <filesystem type='%s'>\n", type); + " <filesystem type='%s'", type); } + + /* Dont generate anything if writeout is set to default */ + if (def->writeout) { + virBufferAsprintf(buf, " writeout='%s'", writeout); + } + + /* close the filesystem element */ + virBufferAddLit(buf,">\n"); if (def->fsdriver) { virBufferAsprintf(buf, " <driver type='%s'/>\n", fsdriver); diff --git a/src/conf/domain_conf.h b/src/conf/domain_conf.h index 3229a6f..bf4d79b 100644 --- a/src/conf/domain_conf.h +++ b/src/conf/domain_conf.h @@ -450,12 +450,21 @@ enum virDomainFSAccessMode { VIR_DOMAIN_FS_ACCESSMODE_LAST }; +/* Filesystem Writeout */ +enum virDomainFSWriteout { + VIR_DOMAIN_FS_WRITEOUT_DEFAULT = 0, + VIR_DOMAIN_FS_WRITEOUT_IMMEDIATE, + + VIR_DOMAIN_FS_WRITEOUT_LAST +}; + typedef struct _virDomainFSDef virDomainFSDef; typedef virDomainFSDef *virDomainFSDefPtr; struct _virDomainFSDef { int type; int fsdriver; int accessmode; + int writeout; char *src; char *dst; unsigned int readonly : 1; @@ -1973,6 +1982,7 @@ VIR_ENUM_DECL(virDomainControllerModelUSB) VIR_ENUM_DECL(virDomainFS) VIR_ENUM_DECL(virDomainFSDriverType) VIR_ENUM_DECL(virDomainFSAccessMode) +VIR_ENUM_DECL(virDomainFSWriteout) VIR_ENUM_DECL(virDomainNet) VIR_ENUM_DECL(virDomainNetBackend) VIR_ENUM_DECL(virDomainNetVirtioTxMode) diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index 1f70eb1..ccfd092 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -108,6 +108,10 @@ VIR_ENUM_IMPL(qemuDomainFSDriver, VIR_DOMAIN_FS_DRIVER_TYPE_LAST, "local", "handle"); +VIR_ENUM_DECL(qemuDomainFSWriteout) +VIR_ENUM_IMPL(qemuDomainFSWriteout, VIR_DOMAIN_FS_WRITEOUT_LAST, + "default", + "immediate"); static void uname_normalize (struct utsname *ut) @@ -1990,6 +1994,7 @@ char *qemuBuildFSStr(virDomainFSDefPtr fs, { virBuffer opt = VIR_BUFFER_INITIALIZER; const char *driver = qemuDomainFSDriverTypeToString(fs->fsdriver); + const char *writeout = qemuDomainFSWriteoutTypeToString(fs->writeout); if (fs->type != VIR_DOMAIN_FS_TYPE_MOUNT) { qemuReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", @@ -2014,6 +2019,10 @@ char *qemuBuildFSStr(virDomainFSDefPtr fs, virBufferAddLit(&opt, ",security_model=none"); } } + + if (fs->writeout != VIR_DOMAIN_FS_WRITEOUT_DEFAULT) { + virBufferAsprintf(&opt, ",writeout=%s", writeout); + } virBufferAsprintf(&opt, ",id=%s%s", QEMU_FSDEV_HOST_PREFIX, fs->info.alias); virBufferAsprintf(&opt, ",path=%s", fs->src);

3 5

[libvirt] question of kvm
by hero 22 Dec '11

22 Dec '11

I want to back up my VM including my own data, but the VM must be 'PASUED' or 'SHUT OFF' if I use 'virt-clone', I want to know how to back up my VM when it's running like make a ghost of windows. Thansk! SunYongGang

2 1

[libvirt] [PATCH 0/4] virsh: Make more commands cancelable via ^C
by Michal Privoznik 22 Dec '11

22 Dec '11

Currently, only migration command allows job progress reporting and cancel via SIGINT. However, there are more job oriented commands which take long time to finish and user might want to cancel them like he can do in migration. Such commands are: dump, save, managedsave. Michal Privoznik (4): virsh: Move job watch code to a separate function virsh: Use do_job_watch in cmdDump virsh: Use do_job_watch in cmdSave virsh: Use do_job_watch in cmdManagedSave tools/virsh.c | 425 +++++++++++++++++++++++++++++++++++++++++-------------- tools/virsh.pod | 24 ++-- 2 files changed, 330 insertions(+), 119 deletions(-) -- 1.7.3.4

2 10

[libvirt] [libvirt-glib 1/2] Add python example for LibvirtGConfig
by Christophe Fergeau 22 Dec '11

22 Dec '11

--- examples/config-test.py | 96 +++++++++++++++++++++++++++++++++++++++++++++++ 1 files changed, 96 insertions(+), 0 deletions(-) create mode 100644 examples/config-test.py diff --git a/examples/config-test.py b/examples/config-test.py new file mode 100644 index 0000000..92e8a89 --- /dev/null +++ b/examples/config-test.py @@ -0,0 +1,96 @@ +#!/usr/bin/python + +from gi.repository import LibvirtGConfig; + +domain = LibvirtGConfig.Domain.new() +domain.set_name("foo") +domain.set_memory(1024*1024*1024) +domain.set_vcpus(2) +domain.set_lifecycle(LibvirtGConfig.DomainLifecycleEvent.ON_POWEROFF, + LibvirtGConfig.DomainLifecycleAction.DESTROY) + +clock = LibvirtGConfig.DomainClock.new() +clock.set_offset(LibvirtGConfig.DomainClockOffset.UTC) +domain.set_clock(clock) + +os = LibvirtGConfig.DomainOs.new() +os.set_os_type(LibvirtGConfig.DomainOsType.HVM) +os.set_arch("x86_64") +devices = [ LibvirtGConfig.DomainOsBootDevice.CDROM, + LibvirtGConfig.DomainOsBootDevice.NETWORK ] +os.set_boot_devices(devices) +domain.set_os(os) + +disk = LibvirtGConfig.DomainDisk.new() +disk.set_type(LibvirtGConfig.DomainDiskType.FILE) +disk.set_guest_device_type(LibvirtGConfig.DomainDiskGuestDeviceType.DISK) +disk.set_source("/tmp/foo/bar") +disk.set_driver_name("qemu") +disk.set_driver_type("qcow2") +disk.set_target_bus(LibvirtGConfig.DomainDiskBus.IDE) +disk.set_target_dev("hda") +domain.add_device(disk) + +interface = LibvirtGConfig.DomainInterfaceNetwork.new() +interface.set_source("default") +domain.add_device(interface) + +interface = LibvirtGConfig.DomainInterfaceUser.new() +interface.set_ifname("eth0") +interface.set_link_state(LibvirtGConfig.DomainInterfaceLinkState.UP) +interface.set_mac("00:11:22:33:44:55") +interface.set_model("foo") +domain.add_device(interface) + +input = LibvirtGConfig.DomainInput.new() +input.set_device_type(LibvirtGConfig.DomainInputDeviceType.TABLET) +input.set_bus(LibvirtGConfig.DomainInputBus.USB) +domain.add_device(input) + +graphics = LibvirtGConfig.DomainGraphicsSpice.new() +graphics.set_port(1234) +domain.add_device(graphics) + +video = LibvirtGConfig.DomainVideo.new() +video.set_model(LibvirtGConfig.DomainVideoModel.QXL) +domain.add_device(video) + +console = LibvirtGConfig.DomainConsole.new() +pty = LibvirtGConfig.DomainChardevSourcePty.new() +console.set_source(pty) +domain.add_device(console) + +print domain.to_xml() + + +pool = LibvirtGConfig.StoragePool.new() +pool.set_pool_type(LibvirtGConfig.StoragePoolType.DIR) + +pool_source = LibvirtGConfig.StoragePoolSource.new() +pool_source.set_directory("/foo/bar") +pool.set_source(pool_source) + +perms = LibvirtGConfig.StoragePermissions.new() +perms.set_owner(1001) +perms.set_group(1005) +perms.set_mode(0744) +perms.set_label("virt_image_t") + +pool_target = LibvirtGConfig.StoragePoolTarget.new() +pool_target.set_path("/dev/disk/by-path") +pool_target.set_permissions(perms) +pool.set_target(pool_target) + +print pool.to_xml() + + +vol = LibvirtGConfig.StorageVol.new() +vol.set_name("my-vol") +vol.set_capacity(0xdeadbeef) + +vol_target = LibvirtGConfig.StorageVolTarget.new() +vol_target.set_format("qcow2") +vol_target.set_permissions(perms) +vol.set_target(vol_target) + +print vol.to_xml() -- 1.7.7.4

2 4

[libvirt] [PATCH] qemu: Support readonly filesystem passthrough
by Osier Yang 22 Dec '11

22 Dec '11

Upstream QEMU starts to support it from commit 2c74c2cb. --- docs/formatdomain.html.in | 5 +++-- src/qemu/qemu_capabilities.c | 7 ++++++- src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_command.c | 11 +++++++++++ 4 files changed, 21 insertions(+), 3 deletions(-) diff --git a/docs/formatdomain.html.in b/docs/formatdomain.html.in index c57b7b3..9cf0f12 100644 --- a/docs/formatdomain.html.in +++ b/docs/formatdomain.html.in @@ -1399,8 +1399,9 @@ <dt><code>readonly</code></dt> <dd> - An optional <code>readonly</code> attribute is available but currently - unused. + Enables exporting filesytem as a readonly mount for guest, by + default read-write access is given (currently only works for + QEMU/KVM driver). </dd> </dl> diff --git a/src/qemu/qemu_capabilities.c b/src/qemu/qemu_capabilities.c index de2bc13..43c7578 100644 --- a/src/qemu/qemu_capabilities.c +++ b/src/qemu/qemu_capabilities.c @@ -143,6 +143,7 @@ VIR_ENUM_IMPL(qemuCaps, QEMU_CAPS_LAST, "rombar", "ich9-ahci", "no-acpi", + "fsdev-readonly", ); struct qemu_feature_flags { @@ -981,6 +982,7 @@ qemuCapsComputeCmdFlags(const char *help, virBitmapPtr flags) { const char *p; + const char *fsdev; if (strstr(help, "-no-kqemu")) qemuCapsSet(flags, QEMU_CAPS_KQEMU); @@ -1077,8 +1079,11 @@ qemuCapsComputeCmdFlags(const char *help, qemuCapsSet(flags, QEMU_CAPS_NESTING); if (strstr(help, ",menu=on")) qemuCapsSet(flags, QEMU_CAPS_BOOT_MENU); - if (strstr(help, "-fsdev")) + if ((fsdev = strstr(help, "-fsdev"))) { qemuCapsSet(flags, QEMU_CAPS_FSDEV); + if (strstr(fsdev, "readonly")) + qemuCapsSet(flags, QEMU_CAPS_FSDEV_READONLY); + } if (strstr(help, "-smbios type")) qemuCapsSet(flags, QEMU_CAPS_SMBIOS_TYPE); diff --git a/src/qemu/qemu_capabilities.h b/src/qemu/qemu_capabilities.h index 08d8457..c759baf 100644 --- a/src/qemu/qemu_capabilities.h +++ b/src/qemu/qemu_capabilities.h @@ -116,6 +116,7 @@ enum qemuCapsFlags { QEMU_CAPS_PCI_ROMBAR = 76, /* -device rombar=0|1 */ QEMU_CAPS_ICH9_AHCI = 77, /* -device ich9-ahci */ QEMU_CAPS_NO_ACPI = 78, /* -no-acpi */ + QEMU_CAPS_FSDEV_READONLY =79, /* -fsdev readonly supported */ QEMU_CAPS_LAST, /* this must always be the last item */ }; diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index 9d3bc23..ea1b763 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -2108,6 +2108,17 @@ char *qemuBuildFSStr(virDomainFSDefPtr fs, virBufferAsprintf(&opt, ",id=%s%s", QEMU_FSDEV_HOST_PREFIX, fs->info.alias); virBufferAsprintf(&opt, ",path=%s", fs->src); + if (fs->readonly) { + if (qemuCapsGet(qemuCaps, QEMU_CAPS_FSDEV_READONLY)) { + virBufferAddLit(&opt, ",readonly"); + } else { + qemuReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", + _("readonly filesystem is not supported by this " + "QEMU binary")); + goto error; + } + } + if (virBufferError(&opt)) { virReportOOMError(); goto error; -- 1.7.7.3

2 2

[libvirt] question on kvm
by hero 22 Dec '11

22 Dec '11

on redhat5.6 and redhat 6.1, I can got the MAC address of the kvm VMs,but I don't know how to get the ip address of the VMs,and virsh has not got a command,can you help me to resolve it.

2 1

[libvirt] [PATCH] nwfilter: Do not require DHCP requests to be broadcast
by Stefan Berger 21 Dec '11

21 Dec '11

Remove the requirement that DHCP messages have to be broadcasted. --- src/nwfilter/nwfilter_ebiptables_driver.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) Index: libvirt-acl/src/nwfilter/nwfilter_ebiptables_driver.c =================================================================== --- libvirt-acl.orig/src/nwfilter/nwfilter_ebiptables_driver.c +++ libvirt-acl/src/nwfilter/nwfilter_ebiptables_driver.c @@ -3245,9 +3245,8 @@ ebtablesApplyDHCPOnlyRules(const char *i virBufferAsprintf(&buf, CMD_DEF("$EBT -t nat -A %s" - " -s %s -d Broadcast " + " -s %s" " -p ipv4 --ip-protocol udp" - " --ip-src 0.0.0.0 --ip-dst 255.255.255.255" " --ip-sport 68 --ip-dport 67" " -j ACCEPT") CMD_SEPARATOR CMD_EXEC

2 2

[libvirt] [PATCH 0/4] RFC: grant KVM guests retain arbitrary capabilities
by Taku Izumi 21 Dec '11

21 Dec '11

Hi all, This patchset adds an option for KVM guests to retain arbitrary capabilities. I want KVM guests to retain "cap_sys_rawio" capability, so I tried to run qemu as root user. However because libvirt clears all capability of KVM guest by default, even if guest is running as root user, it doesn't have any capability. I can fulfill my requirement by disabling "clear_emulator_capabilities" option, but it's not good idea considering security risk. I'm happy libvirt could clear unnecessary capabilities instead of clearing all. That is a motivator for creating this patch. By adding "domain_capabilities" element and to domain XML, its domain can retain specified capabilities like the following: ; VM can retain cap_sys_rawio capability # virsh edit VM ... </features> <domain_capabilities> <cap_sys_rawio/> </domain_capabilities> <clock offset='utc'/> ... # virsh start VM # cat /proc/<VM's PID/status ... CapInh: 0000000000000000 CapPrm: fffffffc00020000 CapEff: fffffffc00020000 CapBnd: fffffffc00020000 ... *[PATCH 1/4] conf: add XML schema for domain capabilities *[PATCH 2/4] util: add functions to keep capabilities *[PATCH 3/4] util: extend virExecWithHook() *[PATCH 4/4] qemu: make qemu processes to retain capabilities -- Best regards, Taku Izumi <izumi.taku(a)jp.fujitsu.com>

2 7

[libvirt] [PATCH 1/4] Make drive unit attribute optional in the XML schema
by Michael Ellerman 21 Dec '11

21 Dec '11

The "unit" attribute of a drive address is optional in the code, so should also be in the XML schema. Signed-off-by: Michael Ellerman <michael(a)ellerman.id.au> --- docs/schemas/domaincommon.rng | 8 +++++--- 1 files changed, 5 insertions(+), 3 deletions(-) diff --git a/docs/schemas/domaincommon.rng b/docs/schemas/domaincommon.rng index 553a6f0..f35eb0b 100644 --- a/docs/schemas/domaincommon.rng +++ b/docs/schemas/domaincommon.rng @@ -2241,9 +2241,11 @@ <ref name="driveBus"/> </attribute> </optional> - <attribute name="unit"> - <ref name="driveUnit"/> - </attribute> + <optional> + <attribute name="unit"> + <ref name="driveUnit"/> + </attribute> + </optional> </define> <define name="virtioserialaddress"> <attribute name="controller"> -- 1.7.7.3

1 4

[libvirt] [PATCH] qemu: Release the lock on domobj if fails on finding the disk path
by Osier Yang 21 Dec '11

21 Dec '11

--- src/qemu/qemu_driver.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 660d771..c156794 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -7181,7 +7181,7 @@ qemuDomainBlockResize (virDomainPtr dom, if ((i = virDomainDiskIndexByName(vm->def, path, false)) < 0) { qemuReportError(VIR_ERR_INVALID_ARG, _("invalid path: %s"), path); - goto cleanup; + goto endjob; } disk = vm->def->disks[i]; -- 1.7.7.3

2 2

[libvirt] [PATCH 1/6] Add address type for SPAPR VIO devices
by Michael Ellerman 21 Dec '11

21 Dec '11

For QEMU PPC64 we have a machine type ("pseries") which has a virtual bus called "spapr-vio". We need to be able to create devices on this bus, and as such need a way to specify the address for those devices. This patch adds a new address type "spapr-vio", which achieves this. The addressing is specified with a "reg" property in the address definition. The reg is optional, if it is not specified QEMU will auto-assign an address for the device. Signed-off-by: Michael Ellerman <michael(a)ellerman.id.au> --- src/conf/domain_conf.c | 43 ++++++++++++++++++++++++++++++++++++++++++- src/conf/domain_conf.h | 9 +++++++++ src/qemu/qemu_command.c | 5 +++++ 3 files changed, 56 insertions(+), 1 deletions(-) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index d68ab10..8dfcebc 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -139,7 +139,8 @@ VIR_ENUM_IMPL(virDomainDeviceAddress, VIR_DOMAIN_DEVICE_ADDRESS_TYPE_LAST, "drive", "virtio-serial", "ccid", - "usb") + "usb", + "spapr-vio") VIR_ENUM_IMPL(virDomainDeviceAddressPciMulti, VIR_DOMAIN_DEVICE_ADDRESS_PCI_MULTI_LAST, @@ -1909,6 +1910,11 @@ virDomainDeviceInfoFormat(virBufferPtr buf, info->addr.usb.port); break; + case VIR_DOMAIN_DEVICE_ADDRESS_TYPE_SPAPRVIO: + if (info->addr.spaprvio.has_reg) + virBufferAsprintf(buf, " reg='0x%llx'", info->addr.spaprvio.reg); + break; + default: virDomainReportError(VIR_ERR_INTERNAL_ERROR, _("unknown address type '%d'"), info->type); @@ -2168,6 +2174,34 @@ cleanup: } static int +virDomainDeviceSpaprVioAddressParseXML(xmlNodePtr node, + virDomainDeviceSpaprVioAddressPtr addr) +{ + char *reg; + int ret; + + memset(addr, 0, sizeof(*addr)); + addr->has_reg = false; + + reg = virXMLPropString(node, "reg"); + if (reg) { + if (virStrToLong_ull(reg, NULL, 16, &addr->reg) < 0) { + virDomainReportError(VIR_ERR_INTERNAL_ERROR, "%s", + _("Cannot parse <address> 'reg' attribute")); + ret = -1; + goto cleanup; + } + + addr->has_reg = true; + } + + ret = 0; +cleanup: + VIR_FREE(reg); + return ret; +} + +static int virDomainDeviceUSBMasterParseXML(xmlNodePtr node, virDomainDeviceUSBMasterPtr master) { @@ -2280,6 +2314,11 @@ virDomainDeviceInfoParseXML(xmlNodePtr node, goto cleanup; break; + case VIR_DOMAIN_DEVICE_ADDRESS_TYPE_SPAPRVIO: + if (virDomainDeviceSpaprVioAddressParseXML(address, &info->addr.spaprvio) < 0) + goto cleanup; + break; + default: /* Should not happen */ virDomainReportError(VIR_ERR_INTERNAL_ERROR, @@ -3174,6 +3213,7 @@ virDomainControllerDefParseXML(xmlNodePtr node, } if (def->info.type != VIR_DOMAIN_DEVICE_ADDRESS_TYPE_NONE && + def->info.type != VIR_DOMAIN_DEVICE_ADDRESS_TYPE_SPAPRVIO && def->info.type != VIR_DOMAIN_DEVICE_ADDRESS_TYPE_PCI) { virDomainReportError(VIR_ERR_INTERNAL_ERROR, "%s", _("Controllers must use the 'pci' address type")); @@ -3566,6 +3606,7 @@ virDomainNetDefParseXML(virCapsPtr caps, /* XXX what about ISA/USB based NIC models - once we support * them we should make sure address type is correct */ if (def->info.type != VIR_DOMAIN_DEVICE_ADDRESS_TYPE_NONE && + def->info.type != VIR_DOMAIN_DEVICE_ADDRESS_TYPE_SPAPRVIO && def->info.type != VIR_DOMAIN_DEVICE_ADDRESS_TYPE_PCI) { virDomainReportError(VIR_ERR_INTERNAL_ERROR, "%s", _("Network interfaces must use 'pci' address type")); diff --git a/src/conf/domain_conf.h b/src/conf/domain_conf.h index d6ed898..0ed6aba 100644 --- a/src/conf/domain_conf.h +++ b/src/conf/domain_conf.h @@ -69,6 +69,7 @@ enum virDomainDeviceAddressType { VIR_DOMAIN_DEVICE_ADDRESS_TYPE_VIRTIO_SERIAL, VIR_DOMAIN_DEVICE_ADDRESS_TYPE_CCID, VIR_DOMAIN_DEVICE_ADDRESS_TYPE_USB, + VIR_DOMAIN_DEVICE_ADDRESS_TYPE_SPAPRVIO, VIR_DOMAIN_DEVICE_ADDRESS_TYPE_LAST }; @@ -121,6 +122,13 @@ struct _virDomainDeviceUSBAddress { char *port; }; +typedef struct _virDomainDeviceSpaprVioAddress virDomainDeviceSpaprVioAddress; +typedef virDomainDeviceSpaprVioAddress *virDomainDeviceSpaprVioAddressPtr; +struct _virDomainDeviceSpaprVioAddress { + unsigned long long reg; + bool has_reg; +}; + enum virDomainControllerMaster { VIR_DOMAIN_CONTROLLER_MASTER_NONE, VIR_DOMAIN_CONTROLLER_MASTER_USB, @@ -145,6 +153,7 @@ struct _virDomainDeviceInfo { virDomainDeviceVirtioSerialAddress vioserial; virDomainDeviceCcidAddress ccid; virDomainDeviceUSBAddress usb; + virDomainDeviceSpaprVioAddress spaprvio; } addr; int mastertype; union { diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index d33d7c8..7128b18 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -1256,6 +1256,8 @@ qemuAssignDevicePCISlots(virDomainDefPtr def, qemuDomainPCIAddressSetPtr addrs) def->controllers[i]->idx == 0) continue; + if (def->controllers[i]->info.type == VIR_DOMAIN_DEVICE_ADDRESS_TYPE_SPAPRVIO) + continue; if (def->controllers[i]->info.type != VIR_DOMAIN_DEVICE_ADDRESS_TYPE_NONE) continue; if (qemuDomainPCIAddressSetNextAddr(addrs, &def->controllers[i]->info) < 0) @@ -1406,6 +1408,9 @@ qemuBuildDeviceAddressStr(virBufferPtr buf, virBufferAsprintf(buf, ",bus="); qemuUsbId(buf, info->addr.usb.bus); virBufferAsprintf(buf, ".0,port=%s", info->addr.usb.port); + } else if (info->type == VIR_DOMAIN_DEVICE_ADDRESS_TYPE_SPAPRVIO) { + if (info->addr.spaprvio.has_reg) + virBufferAsprintf(buf, ",reg=0x%llx", info->addr.spaprvio.reg); } return 0; -- 1.7.7.3

3 21

[libvirt] [PATCH] threads: check for failure to set thread-local value
by Eric Blake 20 Dec '11

20 Dec '11

We had a memory leak on a very arcane OOM situation (unlikely to ever hit in practice, but who knows if libvirt.so would ever be linked into some other program that exhausts all thread-local storage keys?). I found it by code inspection, while analyzing a valgrind report generated by Alex Jia. * src/util/threads.h (virThreadLocalSet): Alter signature. * src/util/threads-pthread.c (virThreadHelper): Reduce allocation lifetime. (virThreadLocalSet): Detect failure. * src/util/threads-win32.c (virThreadLocalSet): Likewise. (virCondWait): Fix caller. * src/util/virterror.c (virLastErrorObject): Likewise. --- src/util/threads-pthread.c | 14 +++++++++++--- src/util/threads-win32.c | 9 ++++++--- src/util/threads.h | 2 +- src/util/virterror.c | 5 +++-- 4 files changed, 21 insertions(+), 9 deletions(-) diff --git a/src/util/threads-pthread.c b/src/util/threads-pthread.c index 5b8fd5b..ea64887 100644 --- a/src/util/threads-pthread.c +++ b/src/util/threads-pthread.c @@ -154,8 +154,11 @@ struct virThreadArgs { static void *virThreadHelper(void *data) { struct virThreadArgs *args = data; - args->func(args->opaque); + struct virThreadArgs local = *args; + + /* Free args early, rather than tying it up during the entire thread. */ VIR_FREE(args); + local.func(local.opaque); return NULL; } @@ -249,7 +252,12 @@ void *virThreadLocalGet(virThreadLocalPtr l) return pthread_getspecific(l->key); } -void virThreadLocalSet(virThreadLocalPtr l, void *val) +int virThreadLocalSet(virThreadLocalPtr l, void *val) { - pthread_setspecific(l->key, val); + int err = pthread_setspecific(l->key, val); + if (err) { + errno = err; + return -1; + } + return 0; } diff --git a/src/util/threads-win32.c b/src/util/threads-win32.c index 63f699b..1c33826 100644 --- a/src/util/threads-win32.c +++ b/src/util/threads-win32.c @@ -155,7 +155,10 @@ int virCondWait(virCondPtr c, virMutexPtr m) if (!event) { return -1; } - virThreadLocalSet(&virCondEvent, event); + if (virThreadLocalSet(&virCondEvent, event) < 0) { + CloseHandle(event); + return -1; + } } virMutexLock(&c->lock); @@ -376,7 +379,7 @@ void *virThreadLocalGet(virThreadLocalPtr l) return TlsGetValue(l->key); } -void virThreadLocalSet(virThreadLocalPtr l, void *val) +int virThreadLocalSet(virThreadLocalPtr l, void *val) { - TlsSetValue(l->key, val); + return TlsSetValue(l->key, val) == 0 ? -1 : 0; } diff --git a/src/util/threads.h b/src/util/threads.h index e52f3a9..e5000ea 100644 --- a/src/util/threads.h +++ b/src/util/threads.h @@ -104,7 +104,7 @@ typedef void (*virThreadLocalCleanup)(void *); int virThreadLocalInit(virThreadLocalPtr l, virThreadLocalCleanup c) ATTRIBUTE_RETURN_CHECK; void *virThreadLocalGet(virThreadLocalPtr l); -void virThreadLocalSet(virThreadLocalPtr l, void*); +int virThreadLocalSet(virThreadLocalPtr l, void*) ATTRIBUTE_RETURN_CHECK; # ifdef WIN32 # include "threads-win32.h" diff --git a/src/util/virterror.c b/src/util/virterror.c index 380dc56..8205516 100644 --- a/src/util/virterror.c +++ b/src/util/virterror.c @@ -267,7 +267,8 @@ virLastErrorObject(void) if (!err) { if (VIR_ALLOC(err) < 0) return NULL; - virThreadLocalSet(&virLastErr, err); + if (virThreadLocalSet(&virLastErr, err) < 0) + VIR_FREE(err); } return err; } @@ -612,7 +613,7 @@ virDispatchError(virConnectPtr conn) virErrorFunc handler = virErrorHandler; void *userData = virUserData; - /* Should never happen, but doesn't hurt to check */ + /* Can only happen on OOM. */ if (!err) return; -- 1.7.7.4

1 0

Re: [libvirt] possible 0.9.8 regression?
by Jim Fehlig 20 Dec '11

20 Dec '11

xhu wrote: > On 12/16/2011 11:33 AM, Jim Fehlig wrote: >> Hi All, >> >> I've noticed a regression in libvirt 0.9.8 on some of my kvm test machines >> >> # virsh start opensuse12 >> error: Failed to start domain opensuse12 >> error: Cannot open network interface control socket: Permission denied > For I can't reproduce it on my machine with 0.9.8, can you provide me > the detailed steps? Nothing special, basic domain config using file-backed disk and connecting to a bridge. > Also your os, libvirt, qemu-kvm and kernel version? Yeah, it has something to do with the kernel, glibc, or other such component. qemu-kvm isn't the problem as the error occurs before it is invoked. kernel 3.1.0, glibc 2.14.1 (openSUSE12.1): With libvirt 0.9.7, starting the domain works. This version of libvirt opens control socket with 'socket(AF_INET, SOCK_STREAM, 0)'. With libvirt 0.9.8, the domain does not start. In this version, the control socket is opened with 'socket(AF_PACKET, SOCK_DGRAM, 0)', which fails with EACCES. kernel 3.0.13, glibc 2.11.3 (SLES11 SP2): Regression between libvirt 0.9.7 and 0.9.8 not observed. Initially, I assumed the bug was in glibc. But I can open packet(7) sockets in a test program running as uid=euid=0, just not within libvirtd running with same privileges. Regards, Jim > Thanks! >> Opening a control socket for setting MAC addr, etc. failed with EACCES. >> In 0.9.7, the socket was opened with domain AF_INET, type SOCK_STREAM, >> which of course works on this system. In 0.9.8, the socket is opened >> with AF_PACKET, SOCK_DGRAM. Interestingly, a small test program calling >> 'socket(AF_PACKET, SOCK_DGRAM, 0)' works on this system. >> >> libvirt is built with '--without-capng --without-apparmor >> --without-selinux' and libvirtd is running with uid=euid=0. >> >> I'm really baffled why this fails in libvirtd but works otherwise. Any >> ideas? >> >> Thanks, >> Jim >> >> >> >> >> -- >> libvir-list mailing list >> libvir-list(a)redhat.com >> https://www.redhat.com/mailman/listinfo/libvir-list >

2 4

[libvirt] [PATCH] docs: improve documents for domxml-from-native and domxml-to-native commands
by ajia＠redhat.com 20 Dec '11

20 Dec '11

From: Alex Jia <ajia(a)redhat.com> * tools/virsh.pod: improve virsh man page for domxml-from-native and domxml-to-native commands. Signed-off-by: Alex Jia <ajia(a)redhat.com> --- tools/virsh.pod | 8 ++++++-- 1 files changed, 6 insertions(+), 2 deletions(-) diff --git a/tools/virsh.pod b/tools/virsh.pod index dbe5165..ca0b993 100644 --- a/tools/virsh.pod +++ b/tools/virsh.pod @@ -670,12 +670,16 @@ seconds elapsed since the control interface entered its current state. =item B<domxml-from-native> I<format> I<config> Convert the file I<config> in the native guest configuration format -named by I<format> to a domain XML format. +named by I<format> to a domain XML format. For QEMU/KVM hypervisor, +the I<format> argument must be B<qemu-argv>. For Xen hypervisor, the +I<format> argument may be B<xen-xm> or B<xen-sxpr>. =item B<domxml-to-native> I<format> I<xml> Convert the file I<xml> in domain XML format to the native guest -configuration format named by I<format>. +configuration format named by I<format>. For QEMU/KVM hypervisor, +the I<format> argument must be B<qemu-argv>. For Xen hypervisor, the +I<format> argument may be B<xen-xm> or B<xen-sxpr>. =item B<dump> I<domain-id> I<corefilepath> [I<--bypass-cache>] { [I<--live>] | [I<--crash>] | [I<--reset>] } -- 1.7.1

2 1

[libvirt] [PATCH] virsh: plug memory leaks on cmdDomXMLFromNative and cmdDomXMLToNative
by ajia＠redhat.com 20 Dec '11

20 Dec '11

From: Alex Jia <ajia(a)redhat.com> Detected by valgrind. Leaks introduced in commit 4d5383f. * tools/virsh.c: fix memory leaks on cmdDomXMLFromNative and cmdDomXMLToNative. * how to reproduce? % virsh dumpxml ${guest} > foo.xml % valgrind -v --leak-check=full virsh domxml-from-native qemu-argv foo.xml % valgrind -v --leak-check=full virsh domxml-to-native qemu-argv foo.xml * actual valgrind results: ==9724== 8,193 bytes in 1 blocks are definitely lost in loss record 31 of 33 ==9724== at 0x4A05FDE: malloc (vg_replace_malloc.c:236) ==9724== by 0x4A06167: realloc (vg_replace_malloc.c:525) ==9724== by 0x4C7510B: virReallocN (memory.c:161) ==9724== by 0x4C84679: virFileReadLimFD (util.c:394) ==9724== by 0x4C84815: virFileReadAll (util.c:455) ==9724== by 0x41A89F: cmdDomXMLFromNative (virsh.c:5532) ==9724== by 0x414872: vshCommandRun (virsh.c:16464) ==9724== by 0x425623: main (virsh.c:17971) ==9724== ==9724== LEAK SUMMARY: ==9724== definitely lost: 8,193 bytes in 1 blocks ==9724== indirectly lost: 0 bytes in 0 blocks ==9724== possibly lost: 0 bytes in 0 blocks ==9724== still reachable: 127,128 bytes in 1,347 blocks ==7409== 8,193 bytes in 1 blocks are definitely lost in loss record 31 of 33 ==7409== at 0x4A05FDE: malloc (vg_replace_malloc.c:236) ==7409== by 0x4A06167: realloc (vg_replace_malloc.c:525) ==7409== by 0x4C7510B: virReallocN (memory.c:161) ==7409== by 0x4C84679: virFileReadLimFD (util.c:394) ==7409== by 0x4C84815: virFileReadAll (util.c:455) ==7409== by 0x41A7AF: cmdDomXMLToNative (virsh.c:5578) ==7409== by 0x414892: vshCommandRun (virsh.c:16463) ==7409== by 0x425633: main (virsh.c:17970) ==7409== ==7409== LEAK SUMMARY: ==7409== definitely lost: 8,193 bytes in 1 blocks ==7409== indirectly lost: 0 bytes in 0 blocks ==7409== possibly lost: 0 bytes in 0 blocks ==7409== still reachable: 127,128 bytes in 1,347 blocks Signed-off-by: Alex Jia <ajia(a)redhat.com> --- tools/virsh.c | 2 ++ 1 files changed, 2 insertions(+), 0 deletions(-) diff --git a/tools/virsh.c b/tools/virsh.c index 583ec6d..784e4f2 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -5540,6 +5540,7 @@ cmdDomXMLFromNative(vshControl *ctl, const vshCmd *cmd) ret = false; } + VIR_FREE(configData); return ret; } @@ -5586,6 +5587,7 @@ cmdDomXMLToNative(vshControl *ctl, const vshCmd *cmd) ret = false; } + VIR_FREE(xmlData); return ret; } -- 1.7.1

2 1

[libvirt] [PATCHv2] console: plug memory leaks
by ajia＠redhat.com 20 Dec '11

20 Dec '11

From: Alex Jia <ajia(a)redhat.com> Using 'virReallocN' to allocate memory on virConsoleEventOnStdin, virConsoleEventOnStdout and virConsoleEventOnStream, however, the cleanup function virConsoleShutdown hasn't released these memory. * tools/console.c: fix memory leaks on virConsoleShutdown. https://bugzilla.redhat.com/show_bug.cgi?id=767488 Signed-off-by: Alex Jia <ajia(a)redhat.com> --- tools/console.c | 4 ++++ 1 files changed, 4 insertions(+), 0 deletions(-) diff --git a/tools/console.c b/tools/console.c index e6118a0..63e51ce 100644 --- a/tools/console.c +++ b/tools/console.c @@ -101,6 +101,10 @@ virConsoleShutdown(virConsolePtr con) virStreamAbort(con->st); virStreamFree(con->st); } + if (con->streamToTerminal.data) + VIR_FREE(con->streamToTerminal.data); + if (con->terminalToStream.data) + VIR_FREE(con->terminalToStream.data); if (con->stdinWatch != -1) virEventRemoveHandle(con->stdinWatch); if (con->stdoutWatch != -1) -- 1.7.1

2 1

[libvirt] [PATCH] qemu: Add a capability flag for -no-acpi
by Michael Ellerman 20 Dec '11

20 Dec '11

Currently non-x86 guests must have <acpi/> defined in <features> to prevent libvirt from running qemu with -no-acpi. Although it works, it is a hack. Instead add a capability flag which indicates whether qemu understands the -no-acpi option. Use it to control whether libvirt emits -no-acpi. Current versions of qemu always display -no-acpi in their help output, so this patch has no effect. However the development version of qemu has been modified such that -no-acpi is only displayed when it is actually supported. Signed-off-by: Michael Ellerman <michael(a)ellerman.id.au> --- src/qemu/qemu_capabilities.c | 3 +++ src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_command.c | 6 ++++-- 3 files changed, 8 insertions(+), 2 deletions(-) diff --git a/src/qemu/qemu_capabilities.c b/src/qemu/qemu_capabilities.c index e69d601..de2bc13 100644 --- a/src/qemu/qemu_capabilities.c +++ b/src/qemu/qemu_capabilities.c @@ -142,6 +142,7 @@ VIR_ENUM_IMPL(qemuCaps, QEMU_CAPS_LAST, "cache-unsafe", /* 75 */ "rombar", "ich9-ahci", + "no-acpi", ); struct qemu_feature_flags { @@ -1066,6 +1067,8 @@ qemuCapsComputeCmdFlags(const char *help, qemuCapsSet(flags, QEMU_CAPS_RTC_TD_HACK); if (strstr(help, "-no-hpet")) qemuCapsSet(flags, QEMU_CAPS_NO_HPET); + if (strstr(help, "-no-acpi")) + qemuCapsSet(flags, QEMU_CAPS_NO_ACPI); if (strstr(help, "-no-kvm-pit-reinjection")) qemuCapsSet(flags, QEMU_CAPS_NO_KVM_PIT); if (strstr(help, "-tdf")) diff --git a/src/qemu/qemu_capabilities.h b/src/qemu/qemu_capabilities.h index 00e5214..08d8457 100644 --- a/src/qemu/qemu_capabilities.h +++ b/src/qemu/qemu_capabilities.h @@ -115,6 +115,7 @@ enum qemuCapsFlags { QEMU_CAPS_DRIVE_CACHE_UNSAFE = 75, /* Is cache=unsafe supported? */ QEMU_CAPS_PCI_ROMBAR = 76, /* -device rombar=0|1 */ QEMU_CAPS_ICH9_AHCI = 77, /* -device ich9-ahci */ + QEMU_CAPS_NO_ACPI = 78, /* -no-acpi */ QEMU_CAPS_LAST, /* this must always be the last item */ }; diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index 92a5955..b241147 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -4035,8 +4035,10 @@ qemuBuildCommandLine(virConnectPtr conn, if (monitor_json && qemuCapsGet(qemuCaps, QEMU_CAPS_NO_SHUTDOWN)) virCommandAddArg(cmd, "-no-shutdown"); - if (!(def->features & (1 << VIR_DOMAIN_FEATURE_ACPI))) - virCommandAddArg(cmd, "-no-acpi"); + if (qemuCapsGet(qemuCaps, QEMU_CAPS_NO_ACPI)) { + if (!(def->features & (1 << VIR_DOMAIN_FEATURE_ACPI))) + virCommandAddArg(cmd, "-no-acpi"); + } if (!def->os.bootloader) { /* -- 1.7.7.3

2 1

[libvirt] [PATCHv3 0/8] add command numatune
by Hu Tao 20 Dec '11

20 Dec '11

This series adds a new command numatune to get/set numatune parameters. Besides libnuma, cpuset cgroup parameters are also set according to numatune parameters. But for now, only cpuset.mems is supported. Hu Tao (8): Add functions to set/get cgroup cpuset parameters use cpuset to manage numa add new API virDomain{G,S}etNumaParameters Implement main entries of virDomain{G,S}etNumaParameters Add virDomain{G,S}etNumaParameters support to the remote driver Implement virDomain{G,S}etNumaParameters for the qemu driver add new command numatune to virsh Add document for new command numatune. daemon/remote.c | 64 ++++++++ include/libvirt/libvirt.h.in | 23 +++ python/generator.py | 2 + src/driver.h | 15 ++ src/libvirt.c | 113 ++++++++++++++ src/libvirt_private.syms | 2 + src/libvirt_public.syms | 6 + src/qemu/qemu_cgroup.c | 19 +++ src/qemu/qemu_driver.c | 334 ++++++++++++++++++++++++++++++++++++++++++ src/remote/remote_driver.c | 50 +++++++ src/remote/remote_protocol.x | 24 +++- src/remote_protocol-structs | 16 ++ src/util/cgroup.c | 32 ++++ src/util/cgroup.h | 3 + tools/virsh.c | 181 +++++++++++++++++++++++ tools/virsh.pod | 13 ++ 16 files changed, 896 insertions(+), 1 deletions(-) -- 1.7.3.1

3 35

[libvirt] [PATCH] console: plug memory leaks
by ajia＠redhat.com 20 Dec '11

20 Dec '11

From: Alex Jia <ajia(a)redhat.com> Using 'virReallocN' to allocate memory on virConsoleEventOnStdin, virConsoleEventOnStdout and virConsoleEventOnStream, however, the cleanup function virConsoleShutdown hasn't released these memory. * tools/console.c: fix memory leaks on virConsoleShutdown. https://bugzilla.redhat.com/show_bug.cgi?id=767488 Signed-off-by: Alex Jia <ajia(a)redhat.com> --- tools/console.c | 6 ++++++ 1 files changed, 6 insertions(+), 0 deletions(-) diff --git a/tools/console.c b/tools/console.c index e6118a0..86376e4 100644 --- a/tools/console.c +++ b/tools/console.c @@ -101,6 +101,12 @@ virConsoleShutdown(virConsolePtr con) virStreamAbort(con->st); virStreamFree(con->st); } + if (con->streamToTerminal.data) { + VIR_FREE(con->streamToTerminal.data); + } + if (con->terminalToStream.data) { + VIR_FREE(con->terminalToStream.data); + } if (con->stdinWatch != -1) virEventRemoveHandle(con->stdinWatch); if (con->stdoutWatch != -1) -- 1.7.1

3 2

[libvirt] [PATCHv4 0/6] add numatune command
by Eric Blake 20 Dec '11

20 Dec '11

Not quite polished - see my complaints that will require a v5: https://www.redhat.com/archives/libvir-list/2011-December/msg00843.html I'm hoping Hu can help cover the remaining points, as I've already spent quite a bit of time cleaning it up to here. Hu Tao (6): Add functions to set/get cgroup cpuset parameters use cpuset to manage numa add new API virDomain{G, S}etNumaParameters Add virDomain{G, S}etNumaParameters support to the remote driver Implement virDomain{G, S}etNumaParameters for the qemu driver add new command numatune to virsh daemon/remote.c | 64 ++++++++++ include/libvirt/libvirt.h.in | 39 ++++++ python/generator.py | 2 + src/conf/domain_conf.h | 8 -- src/driver.h | 15 +++ src/libvirt.c | 129 +++++++++++++++++++++ src/libvirt_private.syms | 2 + src/libvirt_public.syms | 6 + src/qemu/qemu_cgroup.c | 19 +++ src/qemu/qemu_driver.c | 262 ++++++++++++++++++++++++++++++++++++++++++ src/remote/remote_driver.c | 50 ++++++++ src/remote/remote_protocol.x | 24 ++++- src/remote_protocol-structs | 22 ++++ src/util/cgroup.c | 32 +++++ src/util/cgroup.h | 3 + tools/virsh.c | 159 +++++++++++++++++++++++++ tools/virsh.pod | 19 +++ 17 files changed, 846 insertions(+), 9 deletions(-) -- 1.7.7.4

2 10