August 2015 - Devel - libvirt List Archives

[libvirt] [PATCH v3 0/5] domainRename API implementation
by Tomas Meszaros 14 Aug '15

14 Aug '15

This is an effort to implement domain rename API. Presented patch series consists of the following: virDomainRename API implementation for qemu, implementation of the virsh command domrename and the additional support code. The idea behind this endeavor is to provide convenient and safe way to rename a domain. Instead of the: virsh dumpxml domain > domain.xml (change domain name in domain.xml) virsh undefine domain virsh define domain.xml user can simply type: virsh domrename foo bar or call virDomainRename() API and domain "foo" will be renamed to "bar". We currently support only renaming inactive domains without snapshots. Renaming procedure takes care of domain log, config, guest agent path and should be able to recover in case of failure. I've been working on this functionality in collaboration with Michal Privoznik who is my mentor during the GSoC 2015. If you have any questions, ideas or criticism feel free to join the discussion. v2: - removed guest agent path rename code - removed rename permission - added code for emitting undefined+renamed event for the old domain v3: - removed domain rename permission - fixed virDomainRename doc comment - added @flags parameter to the virDomainRename API Tomas Meszaros (5): Introduce virDomainRename API virsh: Implement "domrename" command domain_conf: Introducde virDomainObjListRenameAddNew() & virDomainObjListRenameRemove() Introduce new VIR_DOMAIN_EVENT_DEFINED_RENAMED event qemu: Implement virDomainRename examples/object-events/event-test.c | 4 + include/libvirt/libvirt-domain.h | 6 ++ src/conf/domain_conf.c | 35 +++++++++ src/conf/domain_conf.h | 5 ++ src/driver-hypervisor.h | 6 ++ src/libvirt-domain.c | 34 +++++++++ src/libvirt_private.syms | 2 + src/libvirt_public.syms | 5 ++ src/qemu/qemu_driver.c | 145 ++++++++++++++++++++++++++++++++++++ src/remote/remote_driver.c | 1 + src/remote/remote_protocol.x | 18 ++++- src/remote_protocol-structs | 8 ++ tools/virsh-domain.c | 63 +++++++++++++++- tools/virsh.pod | 7 ++ 14 files changed, 336 insertions(+), 3 deletions(-) -- 2.1.0

3 11

[libvirt] [PATCHv2] nodedev: Fix gfeature size to be according to running kernel
by Moshe Levi 14 Aug '15

14 Aug '15

This patch add virNetDevGetGFeaturesSize to get the supported gfeature size from the kernel --- src/util/virnetdev.c | 79 ++++++++++++++++++++++++++++++++++++++++++++----- 1 files changed, 71 insertions(+), 8 deletions(-) diff --git a/src/util/virnetdev.c b/src/util/virnetdev.c index 2f3690e..582efda 100644 --- a/src/util/virnetdev.c +++ b/src/util/virnetdev.c @@ -90,7 +90,8 @@ VIR_LOG_INIT("util.netdev"); #define RESOURCE_FILE_LEN 4096 #if HAVE_DECL_ETHTOOL_GFEATURES # define TX_UDP_TNL 25 -# define GFEATURES_SIZE 2 +# define DIV_ROUND_UP(n, d) (((n) + (d) - 1) / (d)) +# define FEATURE_BITS_TO_BLOCKS(n_bits) DIV_ROUND_UP(n_bits, 32U) # define FEATURE_WORD(blocks, index, field) ((blocks)[(index) / 32U].field) # define FEATURE_FIELD_FLAG(index) (1U << (index) % 32U) # define FEATURE_BIT_IS_SET(blocks, index, field) \ @@ -3110,6 +3111,59 @@ virNetDevGFeatureAvailable(const char *ifname, struct ethtool_gfeatures *cmd) ret = FEATURE_BIT_IS_SET(cmd->features, TX_UDP_TNL, active); return ret; } + + +/** + * virNetDevGetGFeaturesSize + * This function return the number of gfeatures supported in + * the kernel + * + * @ifname: name of the interface + * + * Returns gfeature size on success, 0 not supported and -1 on failure. + */ +static int + virNetDevGetGFeaturesSize(const char *ifname) +{ + struct { + struct ethtool_sset_info hdr; + uint32_t buf[1]; + } sset_info; + struct ethtool_gstrings *strings; + uint32_t size = 0; + int ret = -1; + + sset_info.hdr.cmd = ETHTOOL_GSSET_INFO; + sset_info.hdr.reserved = 0; + sset_info.hdr.sset_mask = 1ULL << ETH_SS_FEATURES; + + if (virNetDevSendEthtoolIoctl(ifname, &sset_info) == -1) + return ret; + size = sset_info.hdr.sset_mask ? sset_info.hdr.data[0] : 0; + if (VIR_ALLOC_VAR(strings, struct ethtool_gstrings, sizeof(*strings) + size * ETH_GSTRING_LEN)) + return ret; + + strings->cmd = ETHTOOL_GSTRINGS; + strings->string_set = ETH_SS_FEATURES; + strings->len = size; + if (size != 0 && virNetDevSendEthtoolIoctl(ifname, strings) == -1) { + if (errno == EOPNOTSUPP || errno == EINVAL || errno == EPERM) { + ret = 0; + size = 0; + } + goto cleanup; + } + + ret = 0; + size = FEATURE_BITS_TO_BLOCKS(size); + + cleanup: + VIR_FREE(strings); + if (ret != -1) + ret = size; + return ret; +} + # endif @@ -3131,6 +3185,7 @@ virNetDevGetFeatures(const char *ifname, struct ethtool_value cmd = { 0 }; # if HAVE_DECL_ETHTOOL_GFEATURES struct ethtool_gfeatures *g_cmd; + uint32_t size = 0; # endif struct elem{ const int cmd; @@ -3188,14 +3243,22 @@ virNetDevGetFeatures(const char *ifname, # endif # if HAVE_DECL_ETHTOOL_GFEATURES - if (VIR_ALLOC_VAR(g_cmd, - struct ethtool_get_features_block, GFEATURES_SIZE) < 0) + size = virNetDevGetGFeaturesSize(ifname); + if (size == -1) { + VIR_DEBUG("Failure to get GFeatures size on ifname %s", ifname); return -1; - g_cmd->cmd = ETHTOOL_GFEATURES; - g_cmd->size = GFEATURES_SIZE; - if (virNetDevGFeatureAvailable(ifname, g_cmd)) - ignore_value(virBitmapSetBit(*out, VIR_NET_DEV_FEAT_TXUDPTNL)); - VIR_FREE(g_cmd); + } else if (size == 0) { + VIR_DEBUG("GFeatures unsupported on ifname %s", ifname); + } else { + if (VIR_ALLOC_VAR(g_cmd, + struct ethtool_get_features_block, size) < 0) + return -1; + g_cmd->cmd = ETHTOOL_GFEATURES; + g_cmd->size = size; + if (virNetDevGFeatureAvailable(ifname, g_cmd)) + ignore_value(virBitmapSetBit(*out, VIR_NET_DEV_FEAT_TXUDPTNL)); + VIR_FREE(g_cmd); + } # endif if (virNetDevRDMAFeature(ifname, out) < 0) -- 1.7.1

3 2

[libvirt] Entering freeze for libvirt-1.2.17
by Daniel Veillard 13 Aug '15

13 Aug '15

Following discussions on Friday, I applied the patches to deactivate the subset of Admin APIs and revert from 1.3.0 to 1.2.17. I then tagged in git and pushed signed tarballs and rpms to the usual place: ftp://libvirt.org/pub/libvirt/ I didn't run my usual tests on that one, my infra is in flux, so even more reasons for people to give it a try :-) I'm likely to make a candidate release 2 on Tuesday and if all goes well we can push 1.2.17 on Thursday, thanks, Daniel -- Daniel Veillard | Open Source and Standards, Red Hat veillard(a)redhat.com | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ http://veillard.com/ | virtualization library http://libvirt.org/

3 10

[libvirt] [PATCH 1/2] Check if qemu-bridge-helper exists and is executable
by Guido Günther 13 Aug '15

13 Aug '15

Otherwise the error is just error: Failed to create domain from test1.xml error: failed to retrieve file descriptor for interface: Transport endpoint is not connected since we don't get a sensible error after the fork. --- src/qemu/qemu_command.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index 5a92ad4..bddc955 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -295,6 +295,12 @@ static int qemuCreateInBridgePortWithHelper(virQEMUDriverConfigPtr cfg, return -1; } + if (!virFileIsExecutable(cfg->bridgeHelperName)) { + virReportSystemError(errno, _("'%s' not a suitable bridge helper"), + cfg->bridgeHelperName); + return -1; + } + cmd = virCommandNew(cfg->bridgeHelperName); if (flags & VIR_NETDEV_TAP_CREATE_VNET_HDR) virCommandAddArgFormat(cmd, "--use-vnet"); -- 2.1.4

2 3

[libvirt] [PATCH v3 0/4] qemu: Return true pinning info
by Martin Kletzander 13 Aug '15

13 Aug '15

v3: - Added test for the hint keeping patch - Fixed pre-existing crasher found thanks to the test (PATCH 1/4) - Fixed problem in the parsing function found by the test added: - Use nnumaCel_max instead of nnumaCells - Fill in nnumaCell_max in test capabilities v2: - Added more info to commit message of patch 2/3 - Changed the commit subject of patch 3/3 Martin Kletzander (4): qemu: Fix segfault when parsing private domain data conf: Pass private data to Parse function of XML options qemu: Keep numad hint after reboot qemu: Use numad information when getting pin information src/conf/domain_conf.c | 2 +- src/conf/domain_conf.h | 17 +++++++++-------- src/libxl/libxl_domain.c | 3 ++- src/lxc/lxc_domain.c | 3 ++- src/qemu/qemu_domain.c | 36 ++++++++++++++++++++++++++++++++++-- src/qemu/qemu_driver.c | 11 +++++++++++ tests/qemuxml2xmltest.c | 3 ++- tests/testutilsqemu.c | 2 ++ 8 files changed, 63 insertions(+), 14 deletions(-) -- 2.5.0

3 10

[libvirt] [patch] Add support for OVMF in virt-aa-helper [Was: [apparmor] virt-aa-helper: does not support OVMF?]
by intrigeri＠debian.org 13 Aug '15

13 Aug '15

Hi, [please Cc me any reply, I'm not subscribed to libvir-list.] it was reported [1] to Ubuntu that virt-aa-helper blocks access to the OVMF files needed to boot UEFI virtual machines in QEMU. After I've confirmed that on Debian sid, Jamie Strandboge suggested a fix. I've successfully tested in my environment (applied on top of 1.2.18) so I'm forwarding it here. [1] https://bugs.launchpad.net/ubuntu/+source/libvirt/+bug/1483071 Cheers, -- intrigeri

4 3

[libvirt] [PATCH v2 0/3] qemu: Return true pinning info
by Martin Kletzander 13 Aug '15

13 Aug '15

First two patches just prepare the ground for the third one that explains what needs to be fixed and ho it's done. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1162947 v2: - Added more info to commit message of patch 2/3 - Changed the commit subject of patch 3/3 Martin Kletzander (3): conf: Pass private data to Parse function of XML options qemu: Keep numad hint after reboot qemu: Use numad information when getting pin information src/conf/domain_conf.c | 2 +- src/conf/domain_conf.h | 17 +++++++++-------- src/libxl/libxl_domain.c | 3 ++- src/lxc/lxc_domain.c | 3 ++- src/qemu/qemu_domain.c | 33 ++++++++++++++++++++++++++++++++- src/qemu/qemu_driver.c | 11 +++++++++++ 6 files changed, 57 insertions(+), 12 deletions(-) -- 2.5.0

1 4

[libvirt] [PATCH] conf: Remove 'vmdef' from virDomainHostdevDefParseXML
by John Ferlan 12 Aug '15

12 Aug '15

Since it's not used, let's remove it to avoid any future usage. Signed-off-by: John Ferlan <jferlan(a)redhat.com> --- src/conf/domain_conf.c | 9 +++------ 1 file changed, 3 insertions(+), 6 deletions(-) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index b743bdd..f66fccd 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -12054,8 +12054,7 @@ virDomainVideoDefParseXML(xmlNodePtr node, } static virDomainHostdevDefPtr -virDomainHostdevDefParseXML(const virDomainDef *vmdef ATTRIBUTE_UNUSED, - xmlNodePtr node, +virDomainHostdevDefParseXML(xmlNodePtr node, xmlXPathContextPtr ctxt, virHashTablePtr bootHash, unsigned int flags) @@ -12122,7 +12121,6 @@ virDomainHostdevDefParseXML(const virDomainDef *vmdef ATTRIBUTE_UNUSED, "address type")); goto error; } - if (virXPathBoolean("boolean(./readonly)", ctxt)) def->readonly = true; if (virXPathBoolean("boolean(./shareable)", ctxt)) @@ -12626,7 +12624,7 @@ virDomainDeviceDefParse(const char *xmlStr, goto error; break; case VIR_DOMAIN_DEVICE_HOSTDEV: - if (!(dev->data.hostdev = virDomainHostdevDefParseXML(def, node, ctxt, + if (!(dev->data.hostdev = virDomainHostdevDefParseXML(node, ctxt, NULL, flags))) goto error; break; @@ -16152,8 +16150,7 @@ virDomainDefParseXML(xmlDocPtr xml, for (i = 0; i < n; i++) { virDomainHostdevDefPtr hostdev; - hostdev = virDomainHostdevDefParseXML(def, nodes[i], ctxt, - bootHash, flags); + hostdev = virDomainHostdevDefParseXML(nodes[i], ctxt, bootHash, flags); if (!hostdev) goto error; -- 2.1.0

2 1

[libvirt] [PATCH v2 00/12] Adjust SCSI generated device address checks
by John Ferlan 12 Aug '15

12 Aug '15

v1 here: http://www.redhat.com/archives/libvir-list/2015-June/msg01104.html Some followups into July resulted in the request to move the Hostdev and Disk default (or _NONE) address creation/assignment into domain/ device post processing rather than during XML parsing. Changes in v2 are numerous, quite a bit of patch and code motion in order to accomplish the requested task in small enough and reviewable chunks John Ferlan (12): conf: Remove extraneous check in virDomainHostdevAssignAddress conf: Add 'bus' and 'target' to SCSI address conflict checks conf: Move hostdev and disk address validations conf: Add xmlopt to virDomainDeviceDefPostParseInternal conf: Add check for host address type while checking in use conf: Try controller add when searching hostdev bus for unit conf: Change when virDomainHostdevAssignAddress is called conf: Remove unused param from virDomainHostdevDefParseXML conf: Add SCSI hostdev check for disk drive address already in use conf: Change when virDomainDiskDefAssignAddress is called conf: Create locals for virDomainDiskDefAssignAddress conf: Check for hostdev conflicts when assign default disk address docs/formatdomain.html.in | 4 +- src/conf/domain_conf.c | 396 +++++++++++++++++++++++++++------------------- src/conf/domain_conf.h | 3 +- src/qemu/qemu_command.c | 4 +- src/vmx/vmx.c | 22 +-- src/vmx/vmx.h | 3 +- 6 files changed, 253 insertions(+), 179 deletions(-) -- 2.1.0

2 22

[libvirt] [PATCH] Drive hot-unplug: reliable parsing of HMP results
by Frank Schreuder 12 Aug '15

12 Aug '15

Hot-unplugging a disk from a guest that supports hot-unplugging generates an error in the libvirt log when running QEMU with the "-msg timestamp=on" flag. 2015-08-06 10:48:59.945+0000: 11662: error : qemuMonitorTextDriveDel:2594 : operation failed: deleting drive-virtio-disk4 drive failed: 2015-08-06T10:48:59.945058Z Device 'drive-virtio-disk4' not found This error is caused because the HMP results are getting prefixed with a timestamp. Parsing the output is not reliable with STRPREFIX as the results can be prefixed with a timestamp. Using strstr ensures that parsing the output works whether the results are prefixed or not. Cc: Stefan Hajnoczi <stefanha(a)redhat.com> Cc: Daniel P. Berrange <berrange(a)redhat.com> Signed-off-by: Frank Schreuder <fschreuder(a)transip.nl> --- src/qemu/qemu_monitor_text.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/qemu/qemu_monitor_text.c b/src/qemu/qemu_monitor_text.c index 2aa0460..d5ef089 100644 --- a/src/qemu/qemu_monitor_text.c +++ b/src/qemu/qemu_monitor_text.c @@ -2586,7 +2586,7 @@ int qemuMonitorTextDriveDel(qemuMonitorPtr mon, /* (qemu) drive_del wark * Device 'wark' not found */ - } else if (STRPREFIX(reply, "Device '") && (strstr(reply, "not found"))) { + } else if (strstr(reply, "Device '") && strstr(reply, "not found")) { /* NB: device not found errors mean the drive was auto-deleted and we * ignore the error */ } else if (STRNEQ(reply, "")) { -- 2.4.5

5 5