- Devel - libvirt List Archives

[libvirt] [PATCH 00/28] IP peer address and host-side address config support
by Laine Stump 26 Jun '16

26 Jun '16

This (admittedly very long, but most of the patches are either very small or just code movement/renaming, really!) series started out to be just a replacement for a single patch posted by Vasiliy Tolstov back in April (originally pushed upstream as commit 690969af, which was reverted in commit 1d14b13f when I noticed an inconsistency in behavior between lxc and qemu). But each step forward seemed to require (or strongly suggest) a step sideways to reorganize existing code, or eliminate annoying problems I found, and I ended up with 28 patches. The patches that actually add the new functionality are 24-28 (and they are quite small), all the rest just setting up for that. In particular, the three largest patches (03, 13, 14) are purely moving code and renaming identifiers. Patches 07, 08, 12, and 18 fix minor bugs I found during all the changes. (12 is a bit strange, as it removes an undocumented behavior from the openvz driver that has been there since 2010). The one possible controversy in this series (other than the modification to openvz configuration in patch 12) is my placement of the config info for the host-side network device. I could have either put the IP info for the host-side interface as a subelement of <target> where the host-side interface is named, or I could follow the convention of other device types and make it a subelement of the <source> element of interface (since there is where you're supposed to describe how a device connects to the host side resources). <target> element is used to give the name of the tap interface itself. I went back and forth about this, and in the end decided that since the IP address is configuring how the device connects to the outside world (just as with the "network" and "bridge" attributes of <source> for those other network types), that it makes most sense to put it under <source> Please don't be put off by the large number of patches - most of them are fairly straightforward and should be quick reviews. I promise! :-) Laine Stump (27): util: move virNetDevLinkDump to virnetlink.c util: move virInterface(State|Link)/virNetDevFeature from conf to util global: consistently use IP rather than Ip in identifiers conf: clean up virDomainNetIPParseXML() tests: mock virNetDevSetIPAddress util: allow calling virSocketAddrGetIPPrefix with NULL netmask or address lxc: eliminate extraneous free of netDef->ifname_guest lxc: use correct prefix when setting veth IP address conf: new function virDomainNetDefClear qemu: don't set/clear NetDef IP addresses in qemuConnectDomainXMLToNative() qemu: eliminate memory leaks when converting NetDefs to type='ethernet' conf/openvz: eliminate incorrect/undocumented use of <source dev='blah'/> util: new files virnetdevip.[ch] for IP-related netdev functions util: move IP route & address object-related functions to virnetdevip.c conf: single object containing list of IP addresses, list of routes conf: use virNetDevIPInfo in virDomainHostdevCaps conf: use virNetDevIPInfo for guest-side <interface> config qemu: forbid setting guest-side IP address/route info of <interface> conf: clean up after adding calls to virNetDevIPInfo helpers lxc: move debug/error log when adding IP addresses to virNetDevIPAddrAdd util: new function virNetDevIPInfoAddToDev util: provide default destination IP in alternate virNetDevIPRouteAdd() util: make messages consistent in both versions of virNetDevIPRouteAdd() conf: support host-side IP/route information in <interface> util: support setting peer for virNetDevIPInfo addresses lxc: support setting host-side IP addresses/routes qemu: support setting host-side IP addresses/routes Vasiliy Tolstov (1): conf: allow setting peer address in <ip> element of <interface> docs/formatdomain.html.in | 60 +- docs/schemas/domaincommon.rng | 82 +- po/POTFILES.in | 1 + src/Makefile.am | 1 + src/conf/device_conf.c | 31 +- src/conf/device_conf.h | 44 +- src/conf/domain_conf.c | 399 +++++---- src/conf/domain_conf.h | 28 +- src/conf/interface_conf.c | 38 +- src/conf/interface_conf.h | 10 +- src/conf/network_conf.c | 96 +- src/conf/network_conf.h | 24 +- src/conf/networkcommon_conf.c | 107 +-- src/conf/networkcommon_conf.h | 55 +- src/conf/node_device_conf.h | 2 +- src/esx/esx_driver.c | 44 +- src/esx/esx_interface_driver.c | 4 +- src/esx/esx_vi.c | 4 +- src/esx/esx_vi.h | 2 +- src/libvirt_private.syms | 50 +- src/libxl/libxl_conf.c | 10 +- src/libxl/libxl_domain.c | 2 +- src/lxc/lxc_container.c | 92 +- src/lxc/lxc_controller.c | 2 +- src/lxc/lxc_native.c | 42 +- src/lxc/lxc_process.c | 8 + src/network/bridge_driver.c | 103 +-- src/network/bridge_driver_linux.c | 76 +- src/nwfilter/nwfilter_ebiptables_driver.c | 24 +- src/openvz/openvz_conf.c | 2 +- src/openvz/openvz_driver.c | 15 +- src/qemu/qemu_domain.c | 34 + src/qemu/qemu_driver.c | 76 +- src/qemu/qemu_hotplug.c | 6 +- src/qemu/qemu_interface.c | 6 +- src/uml/uml_conf.c | 4 +- src/util/virnetdev.c | 876 +------------------ src/util/virnetdev.h | 46 +- src/util/virnetdevip.c | 971 +++++++++++++++++++++ src/util/virnetdevip.h | 93 ++ src/util/virnetdevvportprofile.c | 4 +- src/util/virnetlink.c | 135 +++ src/util/virnetlink.h | 5 + src/util/virsocketaddr.c | 12 +- src/util/virsocketaddr.h | 3 +- src/vbox/vbox_common.c | 6 +- src/vbox/vbox_network.c | 8 +- src/vz/vz_sdk.c | 2 +- src/xenconfig/xen_common.c | 14 +- src/xenconfig/xen_sxpr.c | 16 +- tests/lxcxml2xmldata/lxc-ethernet-hostip.xml | 44 + tests/lxcxml2xmltest.c | 1 + .../qemuxml2argv-net-eth-hostip.args | 23 + .../qemuxml2argv-net-eth-hostip.xml | 39 + tests/qemuxml2argvmock.c | 10 +- tests/qemuxml2argvtest.c | 1 + .../qemuxml2xmlout-net-eth-hostip.xml | 44 + tests/qemuxml2xmltest.c | 1 + tests/virnetdevtest.c | 14 +- tests/xml2sexprdata/xml2sexpr-net-routed.xml | 1 - 60 files changed, 2193 insertions(+), 1760 deletions(-) create mode 100644 src/util/virnetdevip.c create mode 100644 src/util/virnetdevip.h create mode 100644 tests/lxcxml2xmldata/lxc-ethernet-hostip.xml create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-net-eth-hostip.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-net-eth-hostip.xml create mode 100644 tests/qemuxml2xmloutdata/qemuxml2xmlout-net-eth-hostip.xml -- 2.5.5

2 72

[libvirt] [PATCH 0/3] Enable admin API functionality and tar distribution
by Erik Skultety 25 Jun '16

25 Jun '16

Finally enable admin API in our code and start distributing it both via a tarball and as part of an rpm package. I also added some usage examples, however the html documentation is missing, because of our current xsl stylesheet, more specifically because of how xsltproc works. It takes a stylesheet and an xml file which it processes and produces index.html for each xml file on its input. This way, we either get TOC for libvirt APIs or just the admin. The same story goes for libvirt-qemu and libvirt-lxc which we neither generate, nor distribute. I talked to Jirka and we thought about merging the xmls somehow to a temporary file and then generate the index.html but that requires some more digging, especially into xslt capabilities. So I haven't figured this one out yet, but once I do, I'll enable it for all of our *.xml API 'recipes'. Erik Skultety (3): admin: enable both admin API functionality and tarball distribution examples: admin: Add some examples for the new admin APIs spec: distribute admin API within libvirt-client package .gitignore | 6 ++ Makefile.am | 2 +- cfg.mk | 4 +- daemon/libvirtd.c | 2 +- docs/Makefile.am | 5 +- examples/Makefile.am | 14 +++- examples/admin/client_close.c | 57 +++++++++++++ examples/admin/client_info.c | 165 +++++++++++++++++++++++++++++++++++++ examples/admin/client_limits.c | 63 ++++++++++++++ examples/admin/list_clients.c | 111 +++++++++++++++++++++++++ examples/admin/list_servers.c | 42 ++++++++++ examples/admin/threadpool_params.c | 65 +++++++++++++++ include/libvirt/Makefile.am | 6 +- libvirt.spec.in | 14 ++-- src/libvirt_admin_public.syms | 2 +- 15 files changed, 538 insertions(+), 20 deletions(-) create mode 100644 examples/admin/client_close.c create mode 100644 examples/admin/client_info.c create mode 100644 examples/admin/client_limits.c create mode 100644 examples/admin/list_clients.c create mode 100644 examples/admin/list_servers.c create mode 100644 examples/admin/threadpool_params.c -- 2.5.5

2 5

[libvirt] [PATCH 00/18] vz: support disk/net/graphics devices updates
by Nikolay Shirokovskiy 25 Jun '16

25 Jun '16

It makes sense to support other devices too as update can be on persistent config only which is always possible, but as AFAIK the main purpose of this API is to change live config the set of devices is just set of devices supported by qemu for live updates. Nikolay Shirokovskiy (18): vz: remove disk cache mode hunk vz: make prlsdkGetDisk more generic vz: reuse edit config frame in for attach/detach functions vz: add device updates vz: leverage disks parameters check on disks updates too vz: move disks checks to device post parse vz: fix memory leaks in attach/detach functions vz: give nice report if network device not found vz: dump ip addresses to domain xml vz: dump route info in domain xml vz: fix updating to no gateways vz: fix minor type safey issues with net union usage vz: trustGuestRxFilters fixes vz: move getting container video devices out from vnc code vz: support attach/detach/update/ of graphics device vz: remove exlicitly setting zeros in dumping graphics vz: support vnc password vz: always pass graphics address to sdk src/conf/networkcommon_conf.c | 2 +- src/conf/networkcommon_conf.h | 2 +- src/libvirt_private.syms | 1 + src/vz/vz_driver.c | 115 +++--- src/vz/vz_sdk.c | 897 +++++++++++++++++++++++------------------- src/vz/vz_sdk.h | 10 +- src/vz/vz_utils.c | 204 ++++++++-- src/vz/vz_utils.h | 7 +- 8 files changed, 740 insertions(+), 498 deletions(-) -- 1.8.3.1

2 36

[libvirt] [Bug] make check fails when LIBVIRT_DEBUG is defined
by Tomasz Flendrich 25 Jun '16

25 Jun '16

Hello everybody, I am resending this email as I suppose that the previous one wasn’t accepted due to a big attachment inside. If I create a double post, please excuse me, I didn’t mean to. I set up a clean libvirt repo with: git clone ./autogen.sh make and then executed tests with various values of LIBVIRT_DEBUG using the following command: LIBVIRT_DEBUG=0…5 VIR_TEST_DEBUG=1 make check With LIBVIRT_DEBUG=0 (or =5 or probably any invalid value), the logs say: "Ignoring invalid log level setting” but still, 10 tests fail due to various reasons. With LIBVIRT_DEBUG=1, there is a slight difference. The number of tests that fail is also 10, but here, commandtest fails instead of eventtest. With LIBVIRT_DEBUG=2, only 9 tests fail, because commandtest passes. With LIBVIRT_DEBUG=3, only 1 test fails: qemuargv2xmltest. The reason is the same as in the previous tests. With LIBVIRT_DEBUG=4, it’s just qemuargv2xmltest that is failing, but for a different reason than before: "qemuParseCommandLineString should have logged a warning”. Links to tests/test-syslog.log, indexed by value of LIBVIRT_DEBUG: 0: http://pastebin.com/bZrDV1K2 <http://pastebin.com/bZrDV1K2> 1: https://gist.github.com/Darge/d180ebdd497573ed633625cf136de73c <https://gist.github.com/Darge/d180ebdd497573ed633625cf136de73c> 2: http://pastebin.com/bhsq1MtC <http://pastebin.com/bhsq1MtC> 3: http://pastebin.com/BSRtR9Rq <http://pastebin.com/BSRtR9Rq> 4: http://pastebin.com/FQn79her <http://pastebin.com/FQn79her> Perhaps it’s okay that these tests fail, but still there's no harm in letting people know. Thank you rbogorodskiy for guidance on the mailing list :) Have a nice day, Tomasz

1 0

[libvirt] [PATCH 0/3] Drop some useless CPU code
by Jiri Denemark 25 Jun '16

25 Jun '16

Jiri Denemark (3): cpu: Consolidate ARM drivers cpu: Drop generic driver cpu: Drop NR_DRIVERS macro po/POTFILES.in | 1 - src/Makefile.am | 2 - src/cpu/cpu.c | 16 ++-- src/cpu/cpu_aarch64.c | 133 ----------------------------- src/cpu/cpu_aarch64.h | 31 ------- src/cpu/cpu_arm.c | 37 +++++---- src/cpu/cpu_generic.c | 226 -------------------------------------------------- src/cpu/cpu_generic.h | 32 ------- 8 files changed, 26 insertions(+), 452 deletions(-) delete mode 100644 src/cpu/cpu_aarch64.c delete mode 100644 src/cpu/cpu_aarch64.h delete mode 100644 src/cpu/cpu_generic.c delete mode 100644 src/cpu/cpu_generic.h -- 2.9.0

3 9

[libvirt] [PATCH v2 1/1] cpu_map.xml: add cmt/mbm feature to x86
by Qiaowei Ren 24 Jun '16

24 Jun '16

Some Intel processor families (e.g. the Intel Xeon processor E5 v3 family) introduced some PQos (Platform Qos) features, including CMT (Cache Monitoring technology) and MBM (Memory Bandwidth Monitoring), to monitor or control shared resource. This patch add them into x86 part of cpu_map.xml to be used for applications based on libvirt to get cpu capabilities. For example, Nova in OpenStack schedules guests based on the CPU features that the host has. Signed-off-by: Qiaowei Ren <qiaowei.ren(a)intel.com> --- src/cpu/cpu_map.xml | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/src/cpu/cpu_map.xml b/src/cpu/cpu_map.xml index bf2dfc6..6da8321 100644 --- a/src/cpu/cpu_map.xml +++ b/src/cpu/cpu_map.xml @@ -224,6 +224,9 @@ <feature name='rtm'> <cpuid eax_in='0x07' ebx='0x00000800'/> </feature> + <feature name='cmt'> + <cpuid eax_in='0x07' ebx='0x00001000'/> + </feature> <feature name='mpx'> <cpuid eax_in='0x07' ebx='0x00004000'/> </feature> @@ -266,6 +269,14 @@ <cpuid eax_in='0x0d' ecx_in='0x01' eax='0x00000008'/> </feature> +  + <feature name='mbm_total'>  + <cpuid eax_in='0x0f' ecx_in='0x01' edx='0x00000002'/> + </feature> + <feature name='mbm_local'>  + <cpuid eax_in='0x0f' ecx_in='0x01' edx='0x00000004'/> + </feature> +  <feature name='syscall'>  <cpuid eax_in='0x80000001' edx='0x00000800'/> -- 1.9.1

2 1

[libvirt] Plans for next release 2.0.0
by Daniel Veillard 24 Jun '16

24 Jun '16

So based on the new numbering scheme we should push 2.0.0 at the end of the month. Due to travel and move, I am tempted to enter the freeze on next Sunday (or Monday) to try to push for the release on Friday 1st or the week-end. Hope this works for everyone, 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 2

[libvirt] [PATCH] util: fix build in virNetDevTapGetRealDeviceName
by Roman Bogorodskiy 24 Jun '16

24 Jun '16

Commit e81de04c switched virNetDevTapGetRealDeviceName() to use virDirOpen() instead of opendir(), however it mistakenly dropped DIR *dirp declaration, so restore that to fix build. --- Pushed under the build-breaker rule. src/util/virnetdevtap.c | 1 + 1 file changed, 1 insertion(+) diff --git a/src/util/virnetdevtap.c b/src/util/virnetdevtap.c index 98e27bb..7488a4c 100644 --- a/src/util/virnetdevtap.c +++ b/src/util/virnetdevtap.c @@ -95,6 +95,7 @@ virNetDevTapGetRealDeviceName(char *ifname ATTRIBUTE_UNUSED) #ifdef TAPGIFNAME char *ret = NULL; struct dirent *dp; + DIR *dirp = NULL; char *devpath = NULL; int fd; -- 2.7.4

1 0

[libvirt] [PATCH python] Add support for storage pool refesh callback
by Daniel P. Berrange 24 Jun '16

24 Jun '16

Signed-off-by: Daniel P. Berrange <berrange(a)redhat.com> --- examples/event-test.py | 6 ++++- libvirt-override-virConnect.py | 10 ++++++++ libvirt-override.c | 54 ++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 69 insertions(+), 1 deletion(-) diff --git a/examples/event-test.py b/examples/event-test.py index 241369b..f0341b5 100755 --- a/examples/event-test.py +++ b/examples/event-test.py @@ -572,7 +572,6 @@ def storageEventToString(event): "Undefined", "Started", "Stopped", - "Refreshed", ) return storageEventStrings[event] @@ -581,6 +580,9 @@ def myStoragePoolEventLifecycleCallback(conn, pool, event, detail, opaque): storageEventToString(event), detail)) +def myStoragePoolEventRefreshCallback(conn, pool, event, detail, opaque): + print("myStoragePoolEventRefreshCallback: Storage pool %s" % pool.name()) + ########################################################################## # Set up and run the program ########################################################################## @@ -672,7 +674,9 @@ def main(): vc.domainEventRegisterAny(None, libvirt.VIR_DOMAIN_EVENT_ID_DEVICE_REMOVAL_FAILED, myDomainEventDeviceRemovalFailedCallback, None) vc.networkEventRegisterAny(None, libvirt.VIR_NETWORK_EVENT_ID_LIFECYCLE, myNetworkEventLifecycleCallback, None) + vc.storagePoolEventRegisterAny(None, libvirt.VIR_STORAGE_POOL_EVENT_ID_LIFECYCLE, myStoragePoolEventLifecycleCallback, None) + vc.storagePoolEventRegisterAny(None, libvirt.VIR_STORAGE_POOL_EVENT_ID_REFRESH, myStoragePoolEventRefreshCallback, None) vc.setKeepAlive(5, 3) diff --git a/libvirt-override-virConnect.py b/libvirt-override-virConnect.py index 1aaa586..b085b07 100644 --- a/libvirt-override-virConnect.py +++ b/libvirt-override-virConnect.py @@ -312,6 +312,16 @@ cb(self, virStoragePool(self, _obj=pool), event, detail, opaque) return 0 + def _dispatchStoragePoolEventGenericCallback(self, pool, cbData): + """Dispatches events to python user storage pool + generic event callbacks + """ + cb = cbData["cb"] + opaque = cbData["opaque"] + + cb(self, virStoragePool(self, _obj=pool), opaque) + return 0 + def storagePoolEventDeregisterAny(self, callbackID): """Removes a Storage Pool Event Callback. De-registering for a storage pool callback will disable delivery of this event type""" diff --git a/libvirt-override.c b/libvirt-override.c index 4734b7e..79e3d30 100644 --- a/libvirt-override.c +++ b/libvirt-override.c @@ -8832,6 +8832,56 @@ libvirt_virConnectStoragePoolEventLifecycleCallback(virConnectPtr conn ATTRIBUTE return ret; } +static int +libvirt_virConnectStoragePoolEventGenericCallback(virConnectPtr conn ATTRIBUTE_UNUSED, + virStoragePoolPtr pool, + void *opaque) +{ + PyObject *pyobj_cbData = (PyObject*)opaque; + PyObject *pyobj_pool; + PyObject *pyobj_ret = NULL; + PyObject *pyobj_conn; + PyObject *dictKey; + int ret = -1; + + LIBVIRT_ENSURE_THREAD_STATE; + + if (!(dictKey = libvirt_constcharPtrWrap("conn"))) + goto cleanup; + pyobj_conn = PyDict_GetItem(pyobj_cbData, dictKey); + Py_DECREF(dictKey); + + /* Create a python instance of this virNetworkPtr */ + virStoragePoolRef(pool); + if (!(pyobj_pool = libvirt_virStoragePoolPtrWrap(pool))) { + virStoragePoolFree(pool); + goto cleanup; + } + Py_INCREF(pyobj_cbData); + + /* Call the Callback Dispatcher */ + pyobj_ret = PyObject_CallMethod(pyobj_conn, + (char*)"_dispatchStoragePoolEventGenericCallback", + (char*)"OiiO", + pyobj_pool, + pyobj_cbData); + + Py_DECREF(pyobj_cbData); + Py_DECREF(pyobj_pool); + + cleanup: + if (!pyobj_ret) { + DEBUG("%s - ret:%p\n", __FUNCTION__, pyobj_ret); + PyErr_Print(); + } else { + Py_DECREF(pyobj_ret); + ret = 0; + } + + LIBVIRT_RELEASE_THREAD_STATE; + return ret; +} + static PyObject * libvirt_virConnectStoragePoolEventRegisterAny(PyObject *self ATTRIBUTE_UNUSED, PyObject *args) @@ -8863,6 +8913,10 @@ libvirt_virConnectStoragePoolEventRegisterAny(PyObject *self ATTRIBUTE_UNUSED, cb = VIR_STORAGE_POOL_EVENT_CALLBACK(libvirt_virConnectStoragePoolEventLifecycleCallback); break; + case VIR_STORAGE_POOL_EVENT_ID_REFRESH: + cb = VIR_STORAGE_POOL_EVENT_CALLBACK(libvirt_virConnectStoragePoolEventGenericCallback); + break; + case VIR_STORAGE_POOL_EVENT_ID_LAST: break; } -- 2.7.4

2 1

[libvirt] [PATCH] Promote storage pool refresh lifecycle event to top level event
by Daniel P. Berrange 24 Jun '16

24 Jun '16

The VIR_STORAGE_POOL_EVENT_REFRESHED constant does not reflect any change in the lifecycle of the storage pool. It should thus not be part of the storage pool lifecycle event set, but rather be a top level event in its own right. Thus we introduce VIR_STORAGE_POOL_EVENT_ID_REFRESH to replace it. Signed-off-by: Daniel P. Berrange <berrange(a)redhat.com> --- NB this breaks API and wire protocol, so we *must* include this before 2.0.0. Ideally we should get it into before we cut any rc release too. daemon/remote.c | 29 +++++++++++++++++ examples/object-events/event-test.c | 14 ++++++-- include/libvirt/libvirt-storage.h | 2 +- src/conf/storage_event.c | 57 +++++++++++++++++++++++++++++++++ src/conf/storage_event.h | 4 +++ src/libvirt_private.syms | 1 + src/remote/remote_driver.c | 30 +++++++++++++++++ src/remote/remote_protocol.x | 13 +++++++- src/remote_protocol-structs | 5 +++ src/storage/storage_driver.c | 12 +++---- src/test/test_driver.c | 4 +-- tests/objecteventtest.c | 29 ++++++++++++----- tools/virsh-pool.c | 64 +++++++++++++++++++++++++++++++------ 13 files changed, 232 insertions(+), 32 deletions(-) diff --git a/daemon/remote.c b/daemon/remote.c index 5ee82bb..4aa43c2 100644 --- a/daemon/remote.c +++ b/daemon/remote.c @@ -1294,8 +1294,37 @@ remoteRelayStoragePoolEventLifecycle(virConnectPtr conn, return 0; } +static int +remoteRelayStoragePoolEventRefresh(virConnectPtr conn, + virStoragePoolPtr pool, + void *opaque) +{ + daemonClientEventCallbackPtr callback = opaque; + remote_storage_pool_event_refresh_msg data; + + if (callback->callbackID < 0 || + !remoteRelayStoragePoolEventCheckACL(callback->client, conn, pool)) + return -1; + + VIR_DEBUG("Relaying storage pool refresh event callback %d", + callback->callbackID); + + /* build return data */ + memset(&data, 0, sizeof(data)); + make_nonnull_storage_pool(&data.pool, pool); + data.callbackID = callback->callbackID; + + remoteDispatchObjectEventSend(callback->client, remoteProgram, + REMOTE_PROC_STORAGE_POOL_EVENT_REFRESH, + (xdrproc_t)xdr_remote_storage_pool_event_refresh_msg, + &data); + + return 0; +} + static virConnectStoragePoolEventGenericCallback storageEventCallbacks[] = { VIR_STORAGE_POOL_EVENT_CALLBACK(remoteRelayStoragePoolEventLifecycle), + VIR_STORAGE_POOL_EVENT_CALLBACK(remoteRelayStoragePoolEventRefresh), }; verify(ARRAY_CARDINALITY(storageEventCallbacks) == VIR_STORAGE_POOL_EVENT_ID_LAST); diff --git a/examples/object-events/event-test.c b/examples/object-events/event-test.c index 215e026..f2f71cc 100644 --- a/examples/object-events/event-test.c +++ b/examples/object-events/event-test.c @@ -349,8 +349,6 @@ storagePoolEventToString(int event) return "Started"; case VIR_STORAGE_POOL_EVENT_STOPPED: return "Stopped"; - case VIR_STORAGE_POOL_EVENT_REFRESHED: - return "Refreshed"; case VIR_STORAGE_POOL_EVENT_LAST: break; } @@ -679,6 +677,17 @@ myStoragePoolEventCallback(virConnectPtr conn ATTRIBUTE_UNUSED, } +static int +myStoragePoolEventRefreshCallback(virConnectPtr conn ATTRIBUTE_UNUSED, + virStoragePoolPtr pool, + void *opaque ATTRIBUTE_UNUSED) +{ + printf("%s EVENT: Storage pool %s refresh\n", __func__, + virStoragePoolGetName(pool)); + return 0; +} + + static void eventTypedParamsPrint(virTypedParameterPtr params, int nparams) @@ -936,6 +945,7 @@ struct storagePoolEventData { struct storagePoolEventData storagePoolEvents[] = { STORAGE_POOL_EVENT(VIR_STORAGE_POOL_EVENT_ID_LIFECYCLE, myStoragePoolEventCallback), + STORAGE_POOL_EVENT(VIR_STORAGE_POOL_EVENT_ID_REFRESH, myStoragePoolEventRefreshCallback), }; /* make sure that the events are kept in sync */ diff --git a/include/libvirt/libvirt-storage.h b/include/libvirt/libvirt-storage.h index a67a97f..0974f6e 100644 --- a/include/libvirt/libvirt-storage.h +++ b/include/libvirt/libvirt-storage.h @@ -394,6 +394,7 @@ int virStoragePoolIsPersistent(virStoragePoolPtr pool); */ typedef enum { VIR_STORAGE_POOL_EVENT_ID_LIFECYCLE = 0, /* virConnectStoragePoolEventLifecycleCallback */ + VIR_STORAGE_POOL_EVENT_ID_REFRESH = 1, /* virConnectStoragePoolEventGenericCallback */ # ifdef VIR_ENUM_SENTINELS VIR_STORAGE_POOL_EVENT_ID_LAST @@ -443,7 +444,6 @@ typedef enum { VIR_STORAGE_POOL_EVENT_UNDEFINED = 1, VIR_STORAGE_POOL_EVENT_STARTED = 2, VIR_STORAGE_POOL_EVENT_STOPPED = 3, - VIR_STORAGE_POOL_EVENT_REFRESHED = 4, # ifdef VIR_ENUM_SENTINELS VIR_STORAGE_POOL_EVENT_LAST diff --git a/src/conf/storage_event.c b/src/conf/storage_event.c index c5688be..5768bab 100644 --- a/src/conf/storage_event.c +++ b/src/conf/storage_event.c @@ -48,10 +48,20 @@ struct _virStoragePoolEventLifecycle { typedef struct _virStoragePoolEventLifecycle virStoragePoolEventLifecycle; typedef virStoragePoolEventLifecycle *virStoragePoolEventLifecyclePtr; +struct _virStoragePoolEventRefresh { + virStoragePoolEvent parent; + + bool dummy; +}; +typedef struct _virStoragePoolEventRefresh virStoragePoolEventRefresh; +typedef virStoragePoolEventRefresh *virStoragePoolEventRefreshPtr; + static virClassPtr virStoragePoolEventClass; static virClassPtr virStoragePoolEventLifecycleClass; +static virClassPtr virStoragePoolEventRefreshClass; static void virStoragePoolEventDispose(void *obj); static void virStoragePoolEventLifecycleDispose(void *obj); +static void virStoragePoolEventRefreshDispose(void *obj); static int virStoragePoolEventsOnceInit(void) @@ -68,6 +78,12 @@ virStoragePoolEventsOnceInit(void) sizeof(virStoragePoolEventLifecycle), virStoragePoolEventLifecycleDispose))) return -1; + if (!(virStoragePoolEventRefreshClass = + virClassNew(virStoragePoolEventClass, + "virStoragePoolEventRefresh", + sizeof(virStoragePoolEventRefresh), + virStoragePoolEventRefreshDispose))) + return -1; return 0; } @@ -90,6 +106,14 @@ virStoragePoolEventLifecycleDispose(void *obj) static void +virStoragePoolEventRefreshDispose(void *obj) +{ + virStoragePoolEventRefreshPtr event = obj; + VIR_DEBUG("obj=%p", event); +} + + +static void virStoragePoolEventDispatchDefaultFunc(virConnectPtr conn, virObjectEventPtr event, virConnectObjectEventGenericCallback cb, @@ -115,6 +139,13 @@ virStoragePoolEventDispatchDefaultFunc(virConnectPtr conn, goto cleanup; } + case VIR_STORAGE_POOL_EVENT_ID_REFRESH: + { + ((virConnectStoragePoolEventGenericCallback)cb)(conn, pool, + cbopaque); + goto cleanup; + } + case VIR_STORAGE_POOL_EVENT_ID_LAST: break; } @@ -235,3 +266,29 @@ virStoragePoolEventLifecycleNew(const char *name, return (virObjectEventPtr)event; } + + +/** + * virStoragePoolEventRefreshNew: + * @name: name of the storage pool object the event describes + * @uuid: uuid of the storage pool object the event describes + * + * Create a new storage pool refresh event. + */ +virObjectEventPtr +virStoragePoolEventRefreshNew(const char *name, + const unsigned char *uuid) +{ + virStoragePoolEventRefreshPtr event; + + if (virStoragePoolEventsInitialize() < 0) + return NULL; + + if (!(event = virObjectEventNew(virStoragePoolEventRefreshClass, + virStoragePoolEventDispatchDefaultFunc, + VIR_STORAGE_POOL_EVENT_ID_REFRESH, + 0, name, uuid))) + return NULL; + + return (virObjectEventPtr)event; +} diff --git a/src/conf/storage_event.h b/src/conf/storage_event.h index 7c8ccb7..ea72691 100644 --- a/src/conf/storage_event.h +++ b/src/conf/storage_event.h @@ -57,4 +57,8 @@ virStoragePoolEventLifecycleNew(const char *name, int type, int detail); +virObjectEventPtr +virStoragePoolEventRefreshNew(const char *name, + const unsigned char *uuid); + #endif diff --git a/src/libvirt_private.syms b/src/libvirt_private.syms index ef30f7f..0009f89 100644 --- a/src/libvirt_private.syms +++ b/src/libvirt_private.syms @@ -881,6 +881,7 @@ virStorageVolTypeToString; # conf/storage_event.h virStoragePoolEventLifecycleNew; +virStoragePoolEventRefreshNew; virStoragePoolEventStateRegisterID; diff --git a/src/remote/remote_driver.c b/src/remote/remote_driver.c index 1f81f03..e9904f8 100644 --- a/src/remote/remote_driver.c +++ b/src/remote/remote_driver.c @@ -362,6 +362,11 @@ remoteStoragePoolBuildEventLifecycle(virNetClientProgramPtr prog ATTRIBUTE_UNUSE void *evdata, void *opaque); static void +remoteStoragePoolBuildEventRefresh(virNetClientProgramPtr prog ATTRIBUTE_UNUSED, + virNetClientPtr client ATTRIBUTE_UNUSED, + void *evdata, void *opaque); + +static void remoteConnectNotifyEventConnectionClosed(virNetClientProgramPtr prog ATTRIBUTE_UNUSED, virNetClientPtr client ATTRIBUTE_UNUSED, void *evdata, void *opaque); @@ -544,6 +549,10 @@ static virNetClientProgramEvent remoteEvents[] = { remoteStoragePoolBuildEventLifecycle, sizeof(remote_storage_pool_event_lifecycle_msg), (xdrproc_t)xdr_remote_storage_pool_event_lifecycle_msg }, + { REMOTE_PROC_STORAGE_POOL_EVENT_REFRESH, + remoteStoragePoolBuildEventRefresh, + sizeof(remote_storage_pool_event_refresh_msg), + (xdrproc_t)xdr_remote_storage_pool_event_refresh_msg }, }; static void @@ -5134,6 +5143,27 @@ remoteStoragePoolBuildEventLifecycle(virNetClientProgramPtr prog ATTRIBUTE_UNUSE } static void +remoteStoragePoolBuildEventRefresh(virNetClientProgramPtr prog ATTRIBUTE_UNUSED, + virNetClientPtr client ATTRIBUTE_UNUSED, + void *evdata, void *opaque) +{ + virConnectPtr conn = opaque; + struct private_data *priv = conn->privateData; + remote_storage_pool_event_refresh_msg *msg = evdata; + virStoragePoolPtr pool; + virObjectEventPtr event = NULL; + + pool = get_nonnull_storage_pool(conn, msg->pool); + if (!pool) + return; + + event = virStoragePoolEventRefreshNew(pool->name, pool->uuid); + virObjectUnref(pool); + + remoteEventQueue(priv, event, msg->callbackID); +} + +static void remoteDomainBuildQemuMonitorEvent(virNetClientProgramPtr prog ATTRIBUTE_UNUSED, virNetClientPtr client ATTRIBUTE_UNUSED, void *evdata, void *opaque) diff --git a/src/remote/remote_protocol.x b/src/remote/remote_protocol.x index d11bfdf..4403714 100644 --- a/src/remote/remote_protocol.x +++ b/src/remote/remote_protocol.x @@ -3126,6 +3126,11 @@ struct remote_storage_pool_event_lifecycle_msg { int detail; }; +struct remote_storage_pool_event_refresh_msg { + int callbackID; + remote_nonnull_storage_pool pool; +}; + struct remote_domain_fsfreeze_args { remote_nonnull_domain dom; remote_nonnull_string mountpoints<REMOTE_DOMAIN_FSFREEZE_MOUNTPOINTS_MAX>; /* (const char **) */ @@ -5871,5 +5876,11 @@ enum remote_procedure { * @generate: both * @acl: domain:write */ - REMOTE_PROC_DOMAIN_SET_GUEST_VCPUS = 372 + REMOTE_PROC_DOMAIN_SET_GUEST_VCPUS = 372, + + /** + * @generate: both + * @acl: none + */ + REMOTE_PROC_STORAGE_POOL_EVENT_REFRESH = 373 }; diff --git a/src/remote_protocol-structs b/src/remote_protocol-structs index 0d89b15..46e798b 100644 --- a/src/remote_protocol-structs +++ b/src/remote_protocol-structs @@ -2567,6 +2567,10 @@ struct remote_storage_pool_event_lifecycle_msg { int event; int detail; }; +struct remote_storage_pool_event_refresh_msg { + int callbackID; + remote_nonnull_storage_pool pool; +}; struct remote_domain_fsfreeze_args { remote_nonnull_domain dom; struct { @@ -3140,4 +3144,5 @@ enum remote_procedure { REMOTE_PROC_STORAGE_POOL_EVENT_LIFECYCLE = 370, REMOTE_PROC_DOMAIN_GET_GUEST_VCPUS = 371, REMOTE_PROC_DOMAIN_SET_GUEST_VCPUS = 372, + REMOTE_PROC_STORAGE_POOL_EVENT_REFRESH = 373, }; diff --git a/src/storage/storage_driver.c b/src/storage/storage_driver.c index d75c5aa..cef2124 100644 --- a/src/storage/storage_driver.c +++ b/src/storage/storage_driver.c @@ -1206,10 +1206,8 @@ storagePoolRefresh(virStoragePoolPtr obj, goto cleanup; } - event = virStoragePoolEventLifecycleNew(pool->def->name, - pool->def->uuid, - VIR_STORAGE_POOL_EVENT_REFRESHED, - 0); + event = virStoragePoolEventRefreshNew(pool->def->name, + pool->def->uuid); ret = 0; cleanup: @@ -2342,10 +2340,8 @@ virStorageVolPoolRefreshThread(void *opaque) if (backend->refreshPool(NULL, pool) < 0) VIR_DEBUG("Failed to refresh storage pool"); - event = virStoragePoolEventLifecycleNew(pool->def->name, - pool->def->uuid, - VIR_STORAGE_POOL_EVENT_REFRESHED, - 0); + event = virStoragePoolEventRefreshNew(pool->def->name, + pool->def->uuid); cleanup: if (event) diff --git a/src/test/test_driver.c b/src/test/test_driver.c index 2622335..baff085 100644 --- a/src/test/test_driver.c +++ b/src/test/test_driver.c @@ -4480,9 +4480,7 @@ testStoragePoolRefresh(virStoragePoolPtr pool, goto cleanup; } - event = virStoragePoolEventLifecycleNew(pool->name, pool->uuid, - VIR_STORAGE_POOL_EVENT_REFRESHED, - 0); + event = virStoragePoolEventRefreshNew(pool->name, pool->uuid); ret = 0; cleanup: diff --git a/tests/objecteventtest.c b/tests/objecteventtest.c index ac8cce3..5e8087b 100644 --- a/tests/objecteventtest.c +++ b/tests/objecteventtest.c @@ -67,7 +67,6 @@ typedef struct { int defineEvents; int undefineEvents; int unexpectedEvents; - int refreshEvents; } lifecycleEventCounter; static void @@ -78,7 +77,6 @@ lifecycleEventCounter_reset(lifecycleEventCounter *counter) counter->defineEvents = 0; counter->undefineEvents = 0; counter->unexpectedEvents = 0; - counter->refreshEvents = 0; } typedef struct { @@ -153,8 +151,16 @@ storagePoolLifecycleCb(virConnectPtr conn ATTRIBUTE_UNUSED, counter->defineEvents++; else if (event == VIR_STORAGE_POOL_EVENT_UNDEFINED) counter->undefineEvents++; - else if (event == VIR_STORAGE_POOL_EVENT_REFRESHED) - counter->refreshEvents++; +} + +static void +storagePoolRefreshCb(virConnectPtr conn ATTRIBUTE_UNUSED, + virStoragePoolPtr pool ATTRIBUTE_UNUSED, + void* opaque) +{ + int *counter = opaque; + + (*counter)++; } static int @@ -646,18 +652,24 @@ testStoragePoolStartStopEvent(const void *data) { const objecteventTest *test = data; lifecycleEventCounter counter; - int id; + int refreshCounter; + int id1, id2; int ret = 0; if (!test->pool) return -1; lifecycleEventCounter_reset(&counter); + refreshCounter = 0; - id = virConnectStoragePoolEventRegisterAny(test->conn, test->pool, + id1 = virConnectStoragePoolEventRegisterAny(test->conn, test->pool, VIR_STORAGE_POOL_EVENT_ID_LIFECYCLE, VIR_STORAGE_POOL_EVENT_CALLBACK(&storagePoolLifecycleCb), &counter, NULL); + id2 = virConnectStoragePoolEventRegisterAny(test->conn, test->pool, + VIR_STORAGE_POOL_EVENT_ID_REFRESH, + VIR_STORAGE_POOL_EVENT_CALLBACK(&storagePoolRefreshCb), + &refreshCounter, NULL); virStoragePoolCreate(test->pool, 0); virStoragePoolRefresh(test->pool, 0); virStoragePoolDestroy(test->pool); @@ -668,13 +680,14 @@ testStoragePoolStartStopEvent(const void *data) } if (counter.startEvents != 1 || counter.stopEvents != 1 || - counter.refreshEvents != 1 || counter.unexpectedEvents > 0) { + refreshCounter != 1 || counter.unexpectedEvents > 0) { ret = -1; goto cleanup; } cleanup: - virConnectStoragePoolEventDeregisterAny(test->conn, id); + virConnectStoragePoolEventDeregisterAny(test->conn, id1); + virConnectStoragePoolEventDeregisterAny(test->conn, id2); return ret; } diff --git a/tools/virsh-pool.c b/tools/virsh-pool.c index 18e218c..6045331 100644 --- a/tools/virsh-pool.c +++ b/tools/virsh-pool.c @@ -1899,8 +1899,7 @@ VIR_ENUM_IMPL(virshPoolEvent, N_("Defined"), N_("Undefined"), N_("Started"), - N_("Stopped"), - N_("Refreshed")) + N_("Stopped")) static const char * virshPoolEventToString(int event) @@ -1909,18 +1908,21 @@ virshPoolEventToString(int event) return str ? _(str) : _("unknown"); } +struct vshEventCallback { + const char *name; + virConnectStoragePoolEventGenericCallback cb; +}; +typedef struct vshEventCallback vshEventCallback; + struct virshPoolEventData { vshControl *ctl; bool loop; bool timestamp; int count; + vshEventCallback *cb; }; typedef struct virshPoolEventData virshPoolEventData; -VIR_ENUM_DECL(virshPoolEventId) -VIR_ENUM_IMPL(virshPoolEventId, - VIR_STORAGE_POOL_EVENT_ID_LAST, - "lifecycle") static void vshEventLifecyclePrint(virConnectPtr conn ATTRIBUTE_UNUSED, @@ -1955,6 +1957,45 @@ vshEventLifecyclePrint(virConnectPtr conn ATTRIBUTE_UNUSED, vshEventDone(data->ctl); } +static void +vshEventGenericPrint(virConnectPtr conn ATTRIBUTE_UNUSED, + virStoragePoolPtr pool, + void *opaque) +{ + virshPoolEventData *data = opaque; + + if (!data->loop && data->count) + return; + + if (data->timestamp) { + char timestamp[VIR_TIME_STRING_BUFLEN]; + + if (virTimeStringNowRaw(timestamp) < 0) + timestamp[0] = '\0'; + + vshPrint(data->ctl, _("%s: event '%s'' for storage pool %s\n"), + timestamp, + data->cb->name, + virStoragePoolGetName(pool)); + } else { + vshPrint(data->ctl, _("event '%s' for storage pool %s\n"), + data->cb->name, + virStoragePoolGetName(pool)); + } + + data->count++; + if (!data->loop) + vshEventDone(data->ctl); +} + +static vshEventCallback vshEventCallbacks[] = { + { "lifecycle", + VIR_STORAGE_POOL_EVENT_CALLBACK(vshEventLifecyclePrint), }, + { "refresh", vshEventGenericPrint, } +}; +verify(VIR_STORAGE_POOL_EVENT_ID_LAST == ARRAY_CARDINALITY(vshEventCallbacks)); + + static const vshCmdInfo info_pool_event[] = { {.name = "help", .data = N_("Storage Pool Events") @@ -2009,7 +2050,7 @@ cmdPoolEvent(vshControl *ctl, const vshCmd *cmd) size_t i; for (i = 0; i < VIR_STORAGE_POOL_EVENT_ID_LAST; i++) - vshPrint(ctl, "%s\n", virshPoolEventIdTypeToString(i)); + vshPrint(ctl, "%s\n", vshEventCallbacks[i].name); return true; } @@ -2019,7 +2060,11 @@ cmdPoolEvent(vshControl *ctl, const vshCmd *cmd) vshError(ctl, "%s", _("either --list or event type is required")); return false; } - if ((event = virshPoolEventIdTypeFromString(eventName)) < 0) { + + for (event = 0; event < VIR_STORAGE_POOL_EVENT_ID_LAST; event++) + if (STREQ(eventName, vshEventCallbacks[event].name)) + break; + if (event == VIR_STORAGE_POOL_EVENT_ID_LAST) { vshError(ctl, _("unknown event type %s"), eventName); return false; } @@ -2028,6 +2073,7 @@ cmdPoolEvent(vshControl *ctl, const vshCmd *cmd) data.loop = vshCommandOptBool(cmd, "loop"); data.timestamp = vshCommandOptBool(cmd, "timestamp"); data.count = 0; + data.cb = &vshEventCallbacks[event]; if (vshCommandOptTimeoutToMs(ctl, cmd, &timeout) < 0) return false; @@ -2037,7 +2083,7 @@ cmdPoolEvent(vshControl *ctl, const vshCmd *cmd) goto cleanup; if ((eventId = virConnectStoragePoolEventRegisterAny(priv->conn, pool, event, - VIR_STORAGE_POOL_EVENT_CALLBACK(vshEventLifecyclePrint), + data.cb->cb, &data, NULL)) < 0) goto cleanup; switch (vshEventWait(ctl)) { -- 2.7.4

2 3

[libvirt] [PATCH v2 0/3] Fix the domain capabilities wrt maxvcpus
by Shivaprasad G Bhat 24 Jun '16

24 Jun '16

This series drops the suggested cpu in the domcapabilties output which was implemented in previous version. So, this has just the fix for the domcapabililties output wrt maxvcpus. I'll follow up this series with suggested cpus and the virsh maxvcpus fix. The previous version is available here. http://www.redhat.com/archives/libvir-list/2016-June/msg00947.html --- Shivaprasad G Bhat (3): Rename kvmGetMaxVCPUs() to virHostCPUGetKVMMaxVCPUs() Check the kvm host cpu max limits in virConnectGetDomainCapabilities Document to not rely on virConnectGetMaxVcpus API src/libvirt-host.c | 5 +++- src/libvirt_private.syms | 1 + src/qemu/qemu_capabilities.c | 11 ++++++--- src/qemu/qemu_capabilities.h | 3 ++ src/qemu/qemu_driver.c | 54 ++---------------------------------------- src/util/virhostcpu.c | 36 ++++++++++++++++++++++++++++ src/util/virhostcpu.h | 2 ++ tests/domaincapstest.c | 3 ++ 8 files changed, 57 insertions(+), 58 deletions(-) -- Signature

2 6

[libvirt] [PATCH python v4 0/7] Introducing storage pool lifecycle event APIs
by Jovanka Gulicoska 24 Jun '16

24 Jun '16

Introducing implementation of storage pool event APIs. Code changes follow network event APIs. Implemented functions: virStoragePoolEventRegisterAny(), virStoragePoolEventDeregisterAny(), virStoragePoolLifeCycleEventNew(), introduced STARTED, STOPPED, DEFINE, UNDEFINE and REFRESHED. STARTED signal is emiited in storagePoolCreateXML() and storagePoolCreate() DEFINED signal is emitted in storagePoolDefineXML() UNDEFINED signal is emitted in storagePoolUndefine() STOPPED signal is emitted in storagePoolDestroy() and storagePoolRefresh() REFRESHED signal is emitted in storagePoolRefresh() There are also test as well as unittests for the new functions and signals. This is part of a GSOC project: Asynchronous lifecycle events for storage objects As part of the project there should also be implementation for storage volume events. For now there's no signal when volumes are created or deleted, they can also be implemented, but probably the easiest way is to have apps watch for REFRESH signal, and later extend storage driver code to refresh a pool after volume APIs are called. Changes since v1: small changes following comments Changes since v2: #1 - fix broken alignment in datatypes and fix typos in docstring #2 - rename STORAGE_POOL_EVENT_SOURCES to STORAGE_EVENT_SOURCES, change copyright #3 - fixed spacing issue in comments in test driver #4 - fix comment formating, change in comment, move REMOTE_PROC_STORAGE_POOL_EVENT_LIFECYCLE to buttom of list #5 - include storageEventState pointer, change version from 1.3.6 to 2.0.0, use callbackID insted of ret in storageConnectStoragePoolEventRegisterAny() #6 - use domain like lifecycle event callback Changes since v3: moved storage_conf.h from patch #2 to #6, indentation in remote_protocol-stucts (Martin's suggestion) Jovanka Gulicoska (7): Introduce storage lifecycle event APIs conf: add storage_event handling test: implement storage lifecycle event APIs remote: implement storage lifecycle event APIs storage: implement storage lifecycle event APIs event-test: support storage lifecycle event APIs virsh: Introduce virsh-pool command daemon/libvirtd.h | 2 + daemon/remote.c | 207 ++++++++++++++++++++++++++++++- examples/object-events/event-test.c | 73 +++++++++++ include/libvirt/libvirt-storage.h | 93 ++++++++++++++ src/Makefile.am | 5 + src/conf/storage_conf.h | 4 + src/conf/storage_event.c | 237 ++++++++++++++++++++++++++++++++++++ src/conf/storage_event.h | 60 +++++++++ src/datatypes.h | 13 ++ src/driver-storage.h | 13 ++ src/libvirt-storage.c | 125 +++++++++++++++++++ src/libvirt_private.syms | 5 + src/libvirt_public.syms | 6 + src/remote/remote_driver.c | 128 +++++++++++++++++++ src/remote/remote_protocol.x | 43 ++++++- src/remote_protocol-structs | 19 +++ src/storage/storage_driver.c | 110 +++++++++++++++++ src/test/test_driver.c | 71 +++++++++++ tests/objecteventtest.c | 177 +++++++++++++++++++++++++++ tools/virsh-pool.c | 184 ++++++++++++++++++++++++++++ 20 files changed, 1571 insertions(+), 4 deletions(-) create mode 100644 src/conf/storage_event.c create mode 100644 src/conf/storage_event.h -- 2.5.5

3 11

[libvirt] [PATCH] daemon: Drop dependency on libvirt-admin.so
by Michal Privoznik 24 Jun '16

24 Jun '16

Currently, the daemon requires libvirt-admin.so because the functions encoding/decoding RPC messages for admin APIs live there. But this makes it very hard to split admin API into its own separate package: if libvirt-admin.so is going to live in a separate package than the daemon, either both packages must be installed or none. Solve this by statically linking the RPC message handling functions with the daemon. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- I'm a bit torn. I'm tempted to rewrite some parts of admin API so that it follows the driver architecture we have for other areas, e.g. domain drivers. Until then, this patch is needed. daemon/Makefile.am | 2 +- src/Makefile.am | 15 +++++++++++++-- 2 files changed, 14 insertions(+), 3 deletions(-) diff --git a/daemon/Makefile.am b/daemon/Makefile.am index 927d16f..3b6aafe 100644 --- a/daemon/Makefile.am +++ b/daemon/Makefile.am @@ -152,7 +152,7 @@ libvirtd_admin_la_LDFLAGS = \ $(NO_INDIRECT_LDFLAGS) \ $(NULL) libvirtd_admin_la_LIBADD = \ - ../src/libvirt-admin.la + ../src/libvirt-admin-rpc.la man8_MANS = libvirtd.8 diff --git a/src/Makefile.am b/src/Makefile.am index 9f8b638..7c3cef6 100644 --- a/src/Makefile.am +++ b/src/Makefile.am @@ -592,7 +592,7 @@ $(srcdir)/lock_protocol-struct: \ $(srcdir)/%-struct: locking/lockd_la-%.lo $(PDWTAGS) $(srcdir)/admin_protocol-struct: \ - $(srcdir)/%-struct: admin/libvirt_admin_la-%.lo + $(srcdir)/%-struct: admin/%.lo $(PDWTAGS) else !WITH_REMOTE @@ -2165,10 +2165,20 @@ libvirt_admin.syms: libvirt_admin_public.syms $(ADMIN_SYM_FILES) \ # need to include it in the dist EXTRA_DIST += admin/admin_remote.c +noinst_LTLIBRARIES += \ + libvirt-admin-rpc.la + +libvirt_admin_rpc_la_SOURCES = \ + $(ADMIN_PROTOCOL_GENERATED) + +libvirt_admin_rpc_la_LDFLAGS = \ + $(AM_LDFLAGS) \ + $(CYGWIN_EXTRA_LDFLAGS) \ + $(MINGW_EXTRA_LDFLAGS) + lib_LTLIBRARIES += libvirt-admin.la libvirt_admin_la_SOURCES = \ libvirt-admin.c \ - $(ADMIN_PROTOCOL_GENERATED) \ $(DATATYPES_SOURCES) libvirt_admin_la_LDFLAGS = \ @@ -2180,6 +2190,7 @@ libvirt_admin_la_LDFLAGS = \ libvirt_admin_la_LIBADD = \ libvirt.la \ + libvirt-admin-rpc.la \ $(CYGWIN_EXTRA_LIBADD) libvirt_admin_la_CFLAGS = \ -- 2.8.4

5 11

[libvirt] [PATCH] sasl auth：add default mapping of credentials to machines
by longguang.yue 24 Jun '16

24 Jun '16

hi,all: in production environment, it is important to keep safe. libvirt does automatic validation, but it needs to set up mapping for each machine. this patch is to add default certificate mapping , make it easy in lange scale environment. maybe another way is using regex to do exact match, if this idea is helpful, i can continue thanks

2 3

[libvirt] [PATCH 0/3] A few more cleanups
by Ján Tomko 24 Jun '16

24 Jun '16

Ján Tomko (3): Replace some uses STREQLEN with STRPREFIX Fix error detection in virStorageBackendISCSIGetHostNumber openvz: do not open-code STRSKIP src/openvz/openvz_conf.c | 3 +-- src/qemu/qemu_command.c | 2 +- src/storage/storage_backend_iscsi.c | 42 ++++++++++++++++++------------------- src/storage/storage_backend_scsi.c | 2 +- src/xen/xen_hypervisor.c | 4 ++-- 5 files changed, 26 insertions(+), 27 deletions(-) -- 2.7.3

3 5

[libvirt] [PATCH] conf: Allow disks with identical WWN or serial
by Peter Krempa 24 Jun '16

24 Jun '16

Disallowing them broke a use case of testing multipath configurations for storage. The original intent was added as it made it impossible to use certain /dev/disk-by... links. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1349895 --- src/conf/domain_conf.c | 14 -------------- 1 file changed, 14 deletions(-) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 9443281..68e89ed 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -24906,20 +24906,6 @@ virDomainDiskDefCheckDuplicateInfo(const virDomainDiskDef *a, return -1; } - if (a->wwn && b->wwn && STREQ(a->wwn, b->wwn)) { - virReportError(VIR_ERR_CONFIG_UNSUPPORTED, - _("Disks '%s' and '%s' have identical WWN"), - a->dst, b->dst); - return -1; - } - - if (a->serial && b->serial && STREQ(a->serial, b->serial)) { - virReportError(VIR_ERR_CONFIG_UNSUPPORTED, - _("Disks '%s' and '%s' have identical serial"), - a->dst, b->dst); - return -1; - } - return 0; } -- 2.9.0

2 3

[libvirt] [PATCH 0/4] Fix the domain capabilities wrt maxvcpus
by Shivaprasad G Bhat 24 Jun '16

24 Jun '16

This series addresses the comments to my patch in http://www.redhat.com/archives/libvir-list/2016-May/msg00218.html The v2 got quite a lot of criticism to not to change the virConnectGetMaxVCPUS() instead document to use the virConnectGetDomainCapabilities(). The virConnectGetDomainCapabilities() is extended to check for the host limits and the NR_CPUs are also returned. I am planning to follow up this series with possible fixes to maxvcpus command once this merged. --- Shivaprasad G Bhat (4): Document to not rely on virConnectGetMaxVcpus API Rename and move kvmGetMaxVCPUs to utils and extend it Check for VFIO too where the Legacy passthrough is checked Check the kvm host cpu max limits in virConnectGetDomainCapabilities() docs/formatdomaincaps.html.in | 4 +-- src/conf/domain_capabilities.c | 10 +++++-- src/conf/domain_capabilities.h | 1 + src/libvirt-host.c | 6 ++++ src/libvirt_private.syms | 1 + src/qemu/qemu_capabilities.c | 13 +++++++-- src/qemu/qemu_capabilities.h | 3 +- src/qemu/qemu_driver.c | 57 +++------------------------------------- src/util/virhostcpu.c | 37 ++++++++++++++++++++++++++ src/util/virhostcpu.h | 7 +++++ tests/domaincapstest.c | 3 +- 11 files changed, 78 insertions(+), 64 deletions(-) -- Signature

2 12

[libvirt] [PATCH] docs: Fix whitespace in output
by John Ferlan 24 Jun '16

24 Jun '16

Many moons ago, commit id '8d7800a55' adjusted the format of the output to add a space on the HEADER and the DATA... the docs weren't updated to reflect that... This makes that adjustment. Signed-off-by: John Ferlan <jferlan(a)redhat.com> --- Pushed as trivial, but noted as a result of reviews done on LUKS series docs/formatsecret.html.in | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/docs/formatsecret.html.in b/docs/formatsecret.html.in index 599cb38..c39d2a7 100644 --- a/docs/formatsecret.html.in +++ b/docs/formatsecret.html.in @@ -123,9 +123,9 @@ Secret 1b40a534-8301-45d5-b1aa-11894ebb1735 created # # virsh secret-list - UUID Usage + UUID Usage ----------------------------------------------------------- - 1b40a534-8301-45d5-b1aa-11894ebb1735 cephx ceph_example + 1b40a534-8301-45d5-b1aa-11894ebb1735 cephx ceph_example # # CEPHPHRASE=`printf %s "pass phrase" | base64` # virsh secret-set-value 1b40a534-8301-45d5-b1aa-11894ebb1735 $CEPHPHRASE @@ -209,9 +209,9 @@ Secret c4dbe20b-b1a3-4ac1-b6e6-2ac97852ebb6 created # virsh secret-list - UUID Usage + UUID Usage ----------------------------------------------------------- - c4dbe20b-b1a3-4ac1-b6e6-2ac97852ebb6 iscsi libvirtiscsi + c4dbe20b-b1a3-4ac1-b6e6-2ac97852ebb6 iscsi libvirtiscsi # MYSECRET=`printf %s "mysecret" | base64` # virsh secret-set-value c4dbe20b-b1a3-4ac1-b6e6-2ac97852ebb6 $MYSECRET -- 2.5.5

1 0

[libvirt] [PATCH 0/6] storage: fix issues with pool state cleanup
by Cole Robinson 24 Jun '16

24 Jun '16

This series fixes several issues with pool shutdown cleanup regarding transient pool removal, newDef assignment, and stateFile removal. Cole Robinson (6): storage: Break out storageDriverAutostartPool storage: Add storagePoolSetInactive storage: Handle transient pool removal in driver startup storage: delete: Don't unlink stateFile storage: Unlink stateFile in storagePoolSetInactive conf: sync error reporting for object configFile unlinking src/conf/domain_conf.c | 7 +- src/conf/network_conf.c | 6 +- src/conf/nwfilter_conf.c | 6 +- src/conf/storage_conf.c | 7 +- src/conf/virsecretobj.c | 2 +- src/storage/storage_driver.c | 198 ++++++++++++++++++++++--------------------- 6 files changed, 113 insertions(+), 113 deletions(-) -- 2.7.4

2 15

[libvirt] [PATCHv2] build: increase xz compression level
by Ján Tomko 24 Jun '16

24 Jun '16

Increase the default compression level to 9 from 6. This also increases decompression memory requirements from 9 MB to 65 MB. Also turn on verbosity. --- Makefile.am | 3 +++ 1 file changed, 3 insertions(+) v2: also use -v move the setting to Makefile.am diff --git a/Makefile.am b/Makefile.am index 4b4aad3..ef88067 100644 --- a/Makefile.am +++ b/Makefile.am @@ -22,6 +22,9 @@ GENHTML = genhtml SUBDIRS = . gnulib/lib include/libvirt src daemon tools docs gnulib/tests \ tests po examples +XZ_OPT ?= -9ev +export XZ_OPT + ACLOCAL_AMFLAGS = -I m4 EXTRA_DIST = \ -- 2.7.3

2 3

[libvirt] [PATCH v2 0/5] bhyve: virConnectDomainXMLFromNative
by Fabian Freyer 24 Jun '16

24 Jun '16

Aaand I'm reposting this series again, in the hope that this time it will consistently send to the mailing list. Again, very sorry for the noise. Differences to v1: - use gnulib's reentrant getopt implementation. This is necessary for thread safety. - config-post.h: __GNUC_PREREQ is defined here, since using gnulib's getopt pulls in other gnulib headers, which rely on __GNUC_PREREQ, which doesn't exist on FreeBSD. This approach is open for discussion: I chose config-post.h as this would likely always be the first header pulled in (through config.h). Link to v1: https://www.redhat.com/archives/libvir-list/2016-June/msg00001.html Fabian Freyer (5): config-post.h: define __GNUC_PREREQ if not defined gnulib: add getopt module bhyve: implement virConnectDomainXMLFromNative bhyve: implement bhyve argument parser bhyve: implement argument parser for loader bootstrap.conf | 1 + config-post.h | 11 + m4/virt-driver-bhyve.m4 | 3 + po/POTFILES.in | 1 + src/Makefile.am | 2 + src/bhyve/bhyve_driver.c | 42 ++ src/bhyve/bhyve_parse_command.c | 875 ++++++++++++++++++++++++++++++++++++++++ src/bhyve/bhyve_parse_command.h | 30 ++ 8 files changed, 965 insertions(+) create mode 100644 src/bhyve/bhyve_parse_command.c create mode 100644 src/bhyve/bhyve_parse_command.h -- 2.7.0

2 12

[libvirt] [PATCH] virStorageBackendISCSIGetHostNumber: correctly use virDirOpen
by Ján Tomko 24 Jun '16

24 Jun '16

Incorrect conflict resolution in my commit e81de04c1 broke this. --- src/storage/storage_backend_iscsi.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) Pushed under the build-breaker rule. diff --git a/src/storage/storage_backend_iscsi.c b/src/storage/storage_backend_iscsi.c index 4a16d2b..98d1141 100644 --- a/src/storage/storage_backend_iscsi.c +++ b/src/storage/storage_backend_iscsi.c @@ -100,9 +100,7 @@ virStorageBackendISCSIGetHostNumber(const char *sysfs_path, virFileWaitForDevices(); - if (virDirOpen(&sysdir, sysfs_path) < 0) - virReportSystemError(errno, - _("Failed to opendir path '%s'"), sysfs_path); + if (virDirOpen(&sysdir, sysfs_path) < 0) { retval = -1; goto out; } -- 2.7.3

1 0

[libvirt] [PATCH 00/18] Allow domains to start with a dot
by Ján Tomko 24 Jun '16

24 Jun '16

Also introduce virDirOpen* and VIR_DIR_CLOSE helpers. https://bugzilla.redhat.com/show_bug.cgi?id=1333248 Ján Tomko (18): Do not save errno in virUSBDeviceSearch Fix error detection in virStorageBackendISCSIGetHostNumber Do not check the return value of closedir Introduce VIR_DIR_CLOSE Introduce virDirOpen Use virDirOpen Add virDirOpenIfExists Use virDirOpenIfExists Introduce virDirOpenQuiet Use virDirOpenQuiet Prohibit opendir in syntax-check Skip '.' and '..' in virDirRead Do not check for '.' and '..' after virDirRead Fix comment in virStorageBackendFileSystemRefresh Do not ignore hidden files in /sys and /proc Do not skip hidden entries when looking for a stable path Allow configs to start with a dot Do not skip files starting with a dot in leases directory cfg.mk | 7 +- src/conf/network_conf.c | 31 +++------ src/conf/nwfilter_conf.c | 15 ++--- src/conf/storage_conf.c | 30 +++------ src/conf/virdomainobjlist.c | 16 ++--- src/conf/virsecretobj.c | 14 ++-- src/libvirt_private.syms | 4 ++ src/network/bridge_driver.c | 17 ++--- src/openvz/openvz_conf.c | 5 +- src/qemu/qemu_driver.c | 14 +--- src/qemu/qemu_hostdev.c | 10 +-- src/storage/storage_backend.c | 11 ++-- src/storage/storage_backend_fs.c | 14 ++-- src/storage/storage_backend_iscsi.c | 47 +++++++------- src/storage/storage_backend_scsi.c | 30 ++------- src/util/vircgroup.c | 43 ++++--------- src/util/virfile.c | 124 ++++++++++++++++++++++++++---------- src/util/virfile.h | 9 +++ src/util/virhostcpu.c | 19 ++---- src/util/virnetdev.c | 10 +-- src/util/virnetdevtap.c | 9 +-- src/util/virnuma.c | 17 ++--- src/util/virpci.c | 36 +++-------- src/util/virprocess.c | 9 +-- src/util/virscsi.c | 22 ++----- src/util/virusb.c | 16 +---- src/util/virutil.c | 32 ++-------- src/xen/xen_inotify.c | 13 ++-- src/xen/xm_internal.c | 10 +-- tests/virschematest.c | 8 +-- tools/nss/libvirt_nss.c | 13 +--- 31 files changed, 245 insertions(+), 410 deletions(-) -- 2.7.3

2 27

[libvirt] [PATCH 1/1] cpu_map.xml: add cmt/mbm feature to x86
by Qiaowei Ren 24 Jun '16

24 Jun '16

Some Intel processor families (e.g. the Intel Xeon processor E5 v3 family) introduced some PQos (Platform Qos) features, including CMT (Cache Monitoring echnology) and MBM (Memory Bandwidth Monitoring), to monitor or control shared resource. This patch add them into x86 part of cpu_map.xml to be used for applications (like OpenStack) based on libvirt to get cpu capabilities. Signed-off-by: Qiaowei Ren <qiaowei.ren(a)intel.com> --- src/cpu/cpu_map.xml | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/src/cpu/cpu_map.xml b/src/cpu/cpu_map.xml index 08aded2..2e2cb4f 100644 --- a/src/cpu/cpu_map.xml +++ b/src/cpu/cpu_map.xml @@ -320,6 +320,9 @@ <feature name='rtm'> <cpuid function='0x00000007' ebx='0x00000800'/> </feature> + <feature name='cmt'> + <cpuid function='0x00000007' ebx='0x00001000'/> + </feature> <feature name='mpx'> <cpuid function='0x00000007' ebx='0x00004000'/> </feature> @@ -353,6 +356,14 @@ <cpuid function='0x80000007' edx='0x00000100'/> </feature> +  + <feature name='mbm_total'>  + <cpuid function='0x0000000f' edx='0x00000002'/> + </feature> + <feature name='mbm_local'>  + <cpuid function='0x0000000f' edx='0x00000004'/> + </feature> +  <model name='486'> <feature name='fpu'/> -- 1.9.1

4 6

[libvirt] [PATCH 0/2] Warn users with weird thread/memory alignment
by Martin Kletzander 24 Jun '16

24 Jun '16

People are doing weird things. Let's let them do what they want but gently warn them in case they did not want that exact behaviour and are looking for some possible problems. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1254402 Martin Kletzander (2): qemu: Add qemuProcessSetupPid() qemu: Check for thread <=> memory alignment src/qemu/qemu_driver.c | 4 +- src/qemu/qemu_process.c | 375 ++++++++++++++++++++++-------------------------- src/qemu/qemu_process.h | 6 +- 3 files changed, 181 insertions(+), 204 deletions(-) -- 2.9.0

3 9