January 2017 - Devel - libvirt List Archives

[libvirt] [PATCH v2 0/5] storage: Virtuozzo storage backend for storage pool
by Olga Krishtal 27 Jan '17

27 Jan '17

This series of patches support pool and volume management within Virtuozzo Storage. Virtuozzo Storage is a highly-available distributed software defined storage with built-in replication and disaster recovery. More information about vzstorage can be found here: https://openvz.org/Virtuozzo_Storage >From client's point of view it looks like network attached storage (NFS or GlusterFS). It supports the same volume formats as directory, nfs. Default format ifor volume is raw. Because of such similarities all a lot of functions from storage backend fs can be used with no change: all functions that deals with volumes and pool refresh function. >From the other hand, Virtuozzo storage demands special packages like vstorage-client to be installed. So all functions that work with it are in separate file. v2: - resplitted patches - all functions to work with Virtuozzo storage are in separate file - added possibility to mout cluster for pool for ures/with specific perms. - added documetation Olga Krishtal (5): storage: adds with_storage_vstorage buils option storage: vstorage empty backend storage: vstorage pool operations storage: added vstorage pool backend volume functions storage: vstorage pool documentation and simple test configure.ac | 5 + docs/formatstorage.html.in | 7 +- docs/schemas/storagepool.rng | 21 + docs/storage.html.in | 28 +- include/libvirt/libvirt-storage.h | 1 + m4/virt-storage-vstorage.m4 | 73 +++ po/POTFILES.in | 1 + src/Makefile.am | 9 + src/conf/storage_conf.c | 17 +- src/conf/storage_conf.h | 1 + src/storage/storage_backend.c | 6 + src/storage/storage_backend_vstorage.c | 888 ++++++++++++++++++++++++++ src/storage/storage_backend_vstorage.h | 27 + src/storage/storage_driver.c | 2 + tests/storagepoolxml2xmlin/pool-vstorage.xml | 10 + tests/storagepoolxml2xmlout/pool-vstorage.xml | 18 + tests/storagepoolxml2xmltest.c | 3 + tools/virsh-pool.c | 3 + tools/virsh.c | 3 + 19 files changed, 1118 insertions(+), 5 deletions(-) create mode 100644 m4/virt-storage-vstorage.m4 create mode 100644 src/storage/storage_backend_vstorage.c create mode 100644 src/storage/storage_backend_vstorage.h create mode 100644 tests/storagepoolxml2xmlin/pool-vstorage.xml create mode 100644 tests/storagepoolxml2xmlout/pool-vstorage.xml -- 1.8.3.1

3 22

[libvirt] [PATCH python 0/7] Remove unused code from generator
by Daniel P. Berrange 27 Jan '17

27 Jan '17

All this functionality is inherited from libxml, but is not required to generate libvirt python bindings. Daniel P. Berrange (7): Removed unused 'reference_keepers' code from generator Removed unused 'function_post' code from generator Removed unused 'foreign_encoding_args' code from generator Removed unused 'py_return_types' code from generator Removed unused 'classes_ancestor' code from generator Removed unused 'converter_type' code from generator Removed unused 'functions_list_exception_test' code from generator generator.py | 170 +++++++++++------------------------------------------------ 1 file changed, 30 insertions(+), 140 deletions(-) -- 2.9.3

2 9

[libvirt] [PATCH] domain_conf: vnc: preserve autoport value if no port was specified
by Pavel Hrdina 27 Jan '17

27 Jan '17

The issue is that if this graphics definition is provided: <graphics type='vnc' port='0'/> it's parsed as: <graphics type='vnc' autoport='no'> <listen type='address'/> </graphics> but if the resulting XML is parsed again the output is: <graphics type='vnc' port='-1' autoport='yes'> <listen type='address'/> </graphics> and this should not happen. The XML have to always remain the same after it was already parsed by libvirt. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1383039 Signed-off-by: Pavel Hrdina <phrdina(a)redhat.com> --- src/conf/domain_conf.c | 2 ++ .../generic-graphics-vnc-autoport-no.xml | 30 ++++++++++++++++++++++ tests/genericxml2xmltest.c | 1 + 3 files changed, 33 insertions(+) create mode 100644 tests/genericxml2xmlindata/generic-graphics-vnc-autoport-no.xml diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 26bb0fdd0a..ffd020f641 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -11716,6 +11716,8 @@ virDomainGraphicsDefParseXMLVNC(virDomainGraphicsDefPtr def, if (flags & VIR_DOMAIN_DEF_PARSE_INACTIVE) def->data.vnc.port = 0; def->data.vnc.autoport = true; + } else { + def->data.vnc.autoport = false; } } diff --git a/tests/genericxml2xmlindata/generic-graphics-vnc-autoport-no.xml b/tests/genericxml2xmlindata/generic-graphics-vnc-autoport-no.xml new file mode 100644 index 0000000000..748e7f489d --- /dev/null +++ b/tests/genericxml2xmlindata/generic-graphics-vnc-autoport-no.xml @@ -0,0 +1,30 @@ +<domain type='qemu'> + <name>QEMUGuest1</name> + <uuid>c7a5fdbd-edaf-9455-926a-d65c16db1809</uuid> + <memory unit='KiB'>219100</memory> + <currentMemory unit='KiB'>219100</currentMemory> + <vcpu placement='static'>1</vcpu> + <os> + <type arch='i686' machine='pc'>hvm</type> + <boot dev='hd'/> + </os> + <clock offset='utc'/> + <on_poweroff>destroy</on_poweroff> + <on_reboot>restart</on_reboot> + <on_crash>destroy</on_crash> + <devices> + <emulator>/usr/bin/qemu</emulator> + <controller type='usb' index='0'/> + <controller type='ide' index='0'/> + <controller type='pci' index='0' model='pci-root'/> + <input type='mouse' bus='ps2'/> + <input type='keyboard' bus='ps2'/> + <graphics type='vnc' autoport='no'> + <listen type='address'/> + </graphics> + <video> + <model type='cirrus' vram='16384' heads='1' primary='yes'/> + </video> + <memballoon model='none'/> + </devices> +</domain> diff --git a/tests/genericxml2xmltest.c b/tests/genericxml2xmltest.c index 2ea2396480..488190270f 100644 --- a/tests/genericxml2xmltest.c +++ b/tests/genericxml2xmltest.c @@ -91,6 +91,7 @@ mymain(void) DO_TEST_DIFFERENT("graphics-vnc-socket-attr-listen-socket"); DO_TEST_FULL("graphics-vnc-socket-attr-listen-socket-mismatch", 0, false, TEST_COMPARE_DOM_XML2XML_RESULT_FAIL_PARSE); + DO_TEST("graphics-vnc-autoport-no"); DO_TEST_FULL("name-slash-fail", 0, false, TEST_COMPARE_DOM_XML2XML_RESULT_FAIL_PARSE); -- 2.11.0

2 2

[libvirt] [PATCH] doc: improve VNC/SPICE password documentation
by Pavel Hrdina 27 Jan '17

27 Jan '17

If the passwd attribute is set to empty string it disables VNC/SPICE access to the guest. Signed-off-by: Pavel Hrdina <phrdina(a)redhat.com> --- docs/formatdomain.html.in | 19 +++++++++++-------- 1 file changed, 11 insertions(+), 8 deletions(-) diff --git a/docs/formatdomain.html.in b/docs/formatdomain.html.in index 3f7f87524a..1cb94d0253 100644 --- a/docs/formatdomain.html.in +++ b/docs/formatdomain.html.in @@ -5499,9 +5499,10 @@ qemu-kvm -net nic,model=? /dev/null should be auto-allocated). The <code>autoport</code> attribute is the new preferred syntax for indicating auto-allocation of the TCP port to use. The <code>passwd</code> attribute provides a VNC - password in clear text. The <code>keymap</code> attribute specifies - the keymap to use. It is possible to set a limit on the validity of - the password by giving an timestamp + password in clear text, if the <code>passwd</code> is set to empty + string the VNC access is disabled. The <code>keymap</code> attribute + specifies the keymap to use. It is possible to set a limit on the + validity of the password by giving an timestamp <code>passwdValidTo='2010-04-09T15:51:00'</code> assumed to be in UTC. The <code>connected</code> attribute allows control of connected client during password changes. VNC accepts @@ -5540,11 +5541,13 @@ qemu-kvm -net nic,model=? /dev/null an alternative secure port number. The <code>autoport</code> attribute is the new preferred syntax for indicating auto-allocation of needed port numbers. The <code>passwd</code> - attribute provides a SPICE password in clear text. The - <code>keymap</code> attribute specifies the keymap to use. It is - possible to set a limit on the validity of the password by giving - an timestamp <code>passwdValidTo='2010-04-09T15:51:00'</code> - assumed to be in UTC. + attribute provides a SPICE password in clear text, if the + <code>passwd</code> is set to empty string the SPICE access is + disabled. The <code>keymap</code> attribute specifies the keymap + to use. It is possible to set a limit on the validity of the + password by giving an timestamp + <code>passwdValidTo='2010-04-09T15:51:00'</code> assumed to be + in UTC. </p> <p> The <code>connected</code> attribute allows control of connected -- 2.11.0

2 2

[libvirt] Refactoring of storage pool
by Olga Krishtal 27 Jan '17

27 Jan '17

Hi everyone! Half a year ago we started discussion about filesystem pools: https://www.redhat.com/archives/libvir-list/2016-April/msg01941.html https://www.redhat.com/archives/libvir-list/2016-May/msg00208.html https://www.redhat.com/archives/libvir-list/2016-September/msg00463.html At the end we have decided not to use copy/paste and try to merge code: https://www.redhat.com/archives/libvir-list/2016-December/msg00051.html However, after first try it had became obvious (the discussion is in the link above) that pools can be describe as a separate object, that later will be used in storage pool, filesystem pool, vgpu pool, etc. This latter is an extension of previous discussion; I just want to separate refactoring of storage pools and the implementation of filesystem pools a bit. Before sending any patch with such huge refactoring I want to agree about pools Here is an example of pool structures and prototypes of some management functions that were discussed before (thanks a lot to John Ferlan for good advice). I tried to tie everything together and here it is: /* * virpool.h: pool utility to manage structures commonly used for pools. * */ /* The dscription of pool element * For storage pool it used to be virStorageVolDef * instead of key we use uuid, because key is specific * for storage pools, and for other pool types (filesystem pools) * or gpu pool we do nor have key. * type field is also specific - that is why we put it into * void* ElementTypeDef. * ElementTypeDef - stores internal information about pool element(volume, item, etc) * (the same as privateData for virConnectPtr) * For example, for storage pool voluems it consists from: * int type; * virStorageVolSource source; * virStorageSource target; */ #include "virhash.h" typedef enum { VIR_POOL_ELEMENT_VOLUME, VIR_POOL_ELEMENT_FILESYSTEM, VIR_POOL_ELEMENT_LAST, } virPoolElemenType; VIR_ENUM_DECL(virPoolElementType) typedef void (*virPoolElementTypeDefFree) (void *PoolElementDef); typedef struct _virPoolElementDef virPoolElementDef; typedef virPoolElementDef *virPoolElementDefPtr; struct _virPoolElementDef { char *name; char *uuid; bool building; unsigned int in_use; int type; void* ElementTypeDef; virPoolElementTypeDefFree PoolElementDefFree; }; typedef struct _virPoolElementDefList virPoolElementDefList; typedef virPoolElementDefList *virPoolElementDefListPtr; struct _virPoolElementDefList { size_t count; virPoolElementDefPtr *elements; }; /* General information about pool source*/ typedef struct _virPoolSourceHost virPoolSourceHost; typedef virPoolSourceHost *virPoolSourceHostPtr; struct _virPoolSourceHost { char *name; int port; }; typedef struct _virPoolAuthDef virPoolAuthDef; typedef virPoolAuthDef *virPoolAuthDefPtr; struct _virPoolAuthDef { char *username; char *secrettype; /* <secret type='%s' for disk source */ int authType; /* virStorageAuthType */ virSecretLookupTypeDef seclookupdef; }; typedef enum { VIR_POOL_SOURCE_DIR, VIR_POOL_SOURCE_DEVICE, VIR_POOL_SOURCE_NAME, VIR_POOL_SOURCE_LAST, } virPoolSourcePathType; VIR_ENUM_DECL(virPoolSourcePathType) typedef struct _virPoolSourcePath virPoolSourcePath; typedef virPoolSourcePath *virPoolSourcePathPtr; struct _virPoolSourcePath { virPoolSourcePathType sourcetype; char *path; }; typedef void(*virPoolPrivDataFree) (void *privePoolData); typedef struct _virPoolSource virPoolSource; typedef virPoolSource *virPoolSourcePtr; struct _virPoolSource{ /* One or more host definitions */ size_t nhost; virPoolSourceHostPtr hosts; /* Authentication information */ virPoolAuthDefPtr auth; /* One or more devices */ size_t nsrcpath; virPoolSourcePathPtr srcpaths; /* Name of the source */ char *name; /* Vendor of the source */ char *vendor; /* Product name of the source */ char *product; /*Some private data */ void *privatePoolData; virPoolPrivDataFree privDataFree; }; typedef struct _virPoolPathPerms virPoolPathPerms; typedef virPoolPathPerms *virPoolPathPermsPtr; struct _virPoolPathPerms { mode_t mode; uid_t uid; gid_t gid; char *label; } typedef struct _virPoolTarget virPoolTarget; typedef virPoolTarget *virPoolTargetPtr; struct _virPoolTarget{ char *path; /* Optional path to target */ virPoolPathPerms perms; /* Default permissions for path */ }; typedef struct _virPoolDef virPoolDef; typedef virPoolDef *virPoolDefPtr; struct _virPoolDef { char *name; unsigned char uuid[VIR_UUID_BUFLEN]; virPoolSource source; virPoolTarget target; }; typedef struct _virPoolInfo virPoolInfo; // used to be virPoolObj typedef virPoolInfo *virPoolInfoPtr; struct _virPoolInfo { char *configFile; char *autostartLink; bool active; int autostart; unsigned int asyncjobs; virPoolDefPtr def; virPoolDefPtr newDef; virPoolElementDefList elementsdef; }; /* Used to be PoolObjList * Separate struct - in case we will need additional info * upon the hash table. */ typedef struct _virPoolStore virPoolStore; typedef virPoolStore *virPoolStorePtr; struct _virPoolStore { size_t count; virHashAtomicPtr *pools; //Used to be objects obj }; /* * Pool hash and pool elements managment operations */ void virPoolElementDefFree (virPoolElementDefPtr def); void virPoolClearElements(virPoolInfoPtr poolinfo); void virPoolAuthDefFree(virPoolAuthDefPtr authdef); virPoolSourcePtr virPoolSourceNew (virPoolPrivDataFree privDataFree); void virPoolSourceClear(virPoolSourcePtr source); void virPoolDefFree(virPoolDefPtr def); void virPoolInfoFree(virPoolInfoPtr pool); virPoolStorePtr virPoolStoreNew(); * * Functions to manage pool as an object. * */ // It is reference implementation. #include "virpool.h" VIR_ENUM_IMPL(virPoolElementType, VIR_POOL_ELEMNT_LAST, "volume", "filesystem"); void virPoolElementDefFree(virPoolElementDefPtr def) { if (!def) return; VIR_FREE(def->name); VIR_FREE(def->uuid); if (def->PoolElementDefFree) def->PoolElementDefFree(def->ElementTypeDef); } void virPoolClearElements(virPoolInfoPtr poolinfo) { size_t i; for (i = 0; i < poolinfo->elementsdef.count; i++) virPoolElementDefFree(poolinfo->elementsdef.elements[i]); VIR_FREE(poolinfo->elementdef.elements); poolindo->elementsdef.count = 0; } /* General functions to work with pool */ VIR_ENUM_IMPL(virPoolSourcePathType, VIR_POOL_SOURCE_LAST, "dir", "device", "name"); void virPoolAuthDefFree(virPoolAuthDefPtr authdef) { if (!authdef) return; VIR_FREE(authdef->username); VIR_FREE(authdef->secrettype); VIR_FREE(authdef); } virPoolSourcePtr virPoolSourceNew(virPoolPrivDataFree privDataFree) { virPoolSourcePtr source; if (VIR_ALLOC(source) < 0) return -1; source->privDataFree = privDataFree; return source; } void virPoolSourceClear(virPoolSourcePtr source) { size_t i; if (!source) return; /* Free hosts */ for (i = 0; i < source->nhosts; i++) VIR_FREE(source->hosts[i].name); VIR_FREE(source->hosts); /* Free auth information * empty at the moment*/ virPoolAuthDefFree(source->auth); /* Free dev path*/ for (i = 0; i < source->nsrcpath; i++) VIR_FREE(source->srcpath[i].name); VIR_FREE(source->srcpath); VIR_FREE(source->name); VIR_FREE(source->vendor); VIR_FREE(source->product); if (source->privDataFree) source->privDataFree(source->privatePoolData); } void virPoolDefFree(virPoolDefPtr def) { if (!def) return; VIR_FREE(def->name); /*Free other fields*/ virPoolSourceClear(def->source); VIR_FREE(def->target.path); VIR_FREE(def->target.perms.lable); VIR_FREE(def); } void virPoolInfoFree(virPoolInfoPtr pool) { if (!pool) return; VIR_FREE(pool->configFile); VIR_FREE(pool->autostartLink); virPoolDefFree(pool->def); virPoolDefFree(pool->newDef); virPoolClearElements(pool); } virPoolStorePtr virPoolStoreNew() { virPoolStorePtr poolstore; if (VIR_ALLOC(poolstore) < 0) return NULL; poolstore->count = 0; poolstore->pools = virHashAtomicNew(0, virPoolInfoFree); return poolstore; } I will be glad to here any ideas and advises. -- Best regards, Olga

3 3

[libvirt] [PATCH 0/7] perf: add CACHE_L1D perf event support
by Nitesh Konkar 27 Jan '17

27 Jan '17

This patch series adds support and documentation for a generalized hardware cache event called CACHE_L1D perf event. This perf event is split into cache_l1dra, cache_l1drm, cache_l1dwa, cache_l1dwm, cache_l1dpa and cache_l1dpm perf events depending on the config value set. Nitesh Konkar (7): perf: add cache_l1dra perf event support perf: add cache_l1drm perf event support perf: add cache_l1dwa perf event support perf: add cache_l1dwm perf event support perf: add cache_l1dpa perf event support perf: add cache_l1dpm perf event support virsh: Fix manpage typo docs/formatdomain.html.in | 42 ++++++++++++++++++ docs/news.xml | 6 ++- docs/schemas/domaincommon.rng | 6 +++ include/libvirt/libvirt-domain.h | 66 +++++++++++++++++++++++++++++ src/libvirt-domain.c | 18 ++++++++ src/qemu/qemu_driver.c | 6 +++ src/remote/remote_protocol.x | 2 +- src/util/virperf.c | 34 ++++++++++++++- src/util/virperf.h | 6 +++ tests/genericxml2xmlindata/generic-perf.xml | 6 +++ tools/virsh.pod | 26 ++++++++++-- 11 files changed, 210 insertions(+), 8 deletions(-) -- 1.9.3

2 10

[libvirt] [PATCH] virnettlssessiontest.c: fix grammar
by Nitesh Konkar 27 Jan '17

27 Jan '17

Signed-off-by: Nitesh Konkar <nitkon12(a)linux.vnet.ibm.com> --- tests/virnettlssessiontest.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tests/virnettlssessiontest.c b/tests/virnettlssessiontest.c index 0d2e106..24de212 100644 --- a/tests/virnettlssessiontest.c +++ b/tests/virnettlssessiontest.c @@ -74,9 +74,9 @@ static ssize_t testRead(char *buf, size_t len, void *opaque) * This is replicating the checks that are done for an * active TLS session after handshake completes. To * simulate that we create our TLS contexts, skipping - * sanity checks. When then get a socketpair, and + * sanity checks. We then get a socketpair, and * initiate a TLS session across them. Finally do - * do actual cert validation tests + * actual cert validation tests */ static int testTLSSessionInit(const void *opaque) { -- 1.9.3

2 1

[libvirt] [PATCH] perf: Prevent enabling of already enabled perf event
by Nitesh Konkar 27 Jan '17

27 Jan '17

Currently, on every --enable perf_event command, a new event->fd is created and counting of perf event counter starts from zero and previous event->fd is lost. This patch prevents this behaviour. Signed-off-by: Nitesh Konkar <nitkon12(a)linux.vnet.ibm.com> --- src/util/virperf.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/src/util/virperf.c b/src/util/virperf.c index 788b817..57c2081 100644 --- a/src/util/virperf.c +++ b/src/util/virperf.c @@ -230,6 +230,9 @@ virPerfEventEnable(virPerfPtr perf, if (!event || !event_attr) return -1; + if (event->enabled) + return 0; + if (event_attr->attrType == 0 && (type == VIR_PERF_EVENT_CMT || type == VIR_PERF_EVENT_MBMT || type == VIR_PERF_EVENT_MBML)) { -- 1.9.3

2 1

[libvirt] [PATCH] docs: Reword virsh manpage for --uuid --name --details options
by Nitesh Konkar 27 Jan '17

27 Jan '17

This commit is similar to commit 0977ada8.The virsh manpage lists options --uuid and --name as mutually exclusive if option --details is specified when actually the option --details is mutually exclusive and can't go with options --uuid and/or --name. This patch rewords the virsh manpage to state the correct meaning. Signed-off-by: Nitesh Konkar <nitkon12(a)linux.vnet.ibm.com> --- tools/virsh.pod | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/virsh.pod b/tools/virsh.pod index 290f508..fb62923 100644 --- a/tools/virsh.pod +++ b/tools/virsh.pod @@ -3634,8 +3634,8 @@ I<--autostart> lists the autostarting pools, I<--no-autostart> lists the pools with autostarting disabled. If I<--uuid> is specified only pool's UUIDs are printed. If I<--name> is specified only pool's names are printed. If both I<--name> and I<--uuid> are specified, pool's UUID and names are printed side by side -without any header. Options I<--uuid> and I<--name> are mutually exclusive -if option I<--details> is specified. +without any header. Option I<--details> is mutually exclusive with options +I<--uuid> and I<--name>. You may also want to list pools with specified types using I<type>, the pool types must be separated by comma, e.g. --type dir,disk. The valid pool -- 1.9.3

2 1

[libvirt] [PATCH] virsh: Modify description for LIBVIRT_DEBUG=4 in manpage
by Nitesh Konkar 27 Jan '17

27 Jan '17

As LIBVIRT_DEBUG=4 logs only error messages and there are no levels above it, adjusting the description in the man page accordingly. Signed-off-by: Nitesh Konkar <nitkon12(a)linux.vnet.ibm.com> --- tools/virsh.pod | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/virsh.pod b/tools/virsh.pod index c1bdbdd..9e45f09 100644 --- a/tools/virsh.pod +++ b/tools/virsh.pod @@ -4565,7 +4565,7 @@ Messages at level WARNING or above =item * LIBVIRT_DEBUG=4 -Messages at level ERROR or above +Messages at level ERROR =back -- 1.9.3

2 1