- Devel - libvirt List Archives

[libvirt] [PATCH v2] qemu_domain: add timestamp in tainting of guests log
by Chen Hanxiao 21 Jan '17

21 Jan '17

From: Chen Hanxiao <chenhanxiao(a)gmail.com> We lacked of timestamp in tainting of guests log, which bring troubles for finding guest issues: such as whether a guest powerdown caused by qemu-monitor-command or others issues inside guests. If we had timestamp in tainting of guests log, it would be helpful when checking guest's /var/log/messages. Signed-off-by: Chen Hanxiao <chenhanxiao(a)gmail.com> --- v2: update logic flow of virDomainObjTaint check src/qemu/qemu_domain.c | 79 +++++++++++++++++++++++++++----------------------- 1 file changed, 42 insertions(+), 37 deletions(-) diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c index c676042..2bc023c 100644 --- a/src/qemu/qemu_domain.c +++ b/src/qemu/qemu_domain.c @@ -3999,46 +3999,51 @@ void qemuDomainObjTaint(virQEMUDriverPtr driver, { virErrorPtr orig_err = NULL; bool closeLog = false; + char *timestamp = NULL; + char uuidstr[VIR_UUID_STRING_BUFLEN]; - if (virDomainObjTaint(obj, taint)) { - char uuidstr[VIR_UUID_STRING_BUFLEN]; - virUUIDFormat(obj->def->uuid, uuidstr); - - VIR_WARN("Domain id=%d name='%s' uuid=%s is tainted: %s", - obj->def->id, - obj->def->name, - uuidstr, - virDomainTaintTypeToString(taint)); - - /* We don't care about errors logging taint info, so - * preserve original error, and clear any error that - * is raised */ - orig_err = virSaveLastError(); - if (logCtxt == NULL) { - logCtxt = qemuDomainLogContextNew(driver, obj, - QEMU_DOMAIN_LOG_CONTEXT_MODE_ATTACH); - if (!logCtxt) { - if (orig_err) { - virSetError(orig_err); - virFreeError(orig_err); - } - VIR_WARN("Unable to open domainlog"); - return; - } - closeLog = true; - } + if (!virDomainObjTaint(obj, taint)) + return; + + virUUIDFormat(obj->def->uuid, uuidstr); + + VIR_WARN("Domain id=%d name='%s' uuid=%s is tainted: %s", + obj->def->id, + obj->def->name, + uuidstr, + virDomainTaintTypeToString(taint)); - if (qemuDomainLogContextWrite(logCtxt, - "Domain id=%d is tainted: %s\n", - obj->def->id, - virDomainTaintTypeToString(taint)) < 0) - virResetLastError(); - if (closeLog) - qemuDomainLogContextFree(logCtxt); - if (orig_err) { - virSetError(orig_err); - virFreeError(orig_err); + if ((timestamp = virTimeStringNow()) == NULL) + goto cleanup; + + /* We don't care about errors logging taint info, so + * preserve original error, and clear any error that + * is raised */ + orig_err = virSaveLastError(); + if (logCtxt == NULL) { + logCtxt = qemuDomainLogContextNew(driver, obj, + QEMU_DOMAIN_LOG_CONTEXT_MODE_ATTACH); + if (!logCtxt) { + VIR_WARN("Unable to open domainlog"); + goto cleanup; } + closeLog = true; + } + + if (qemuDomainLogContextWrite(logCtxt, + "%s: Domain id=%d is tainted: %s\n", + timestamp, + obj->def->id, + virDomainTaintTypeToString(taint)) < 0) + virResetLastError(); + + cleanup: + VIR_FREE(timestamp); + if (closeLog) + qemuDomainLogContextFree(logCtxt); + if (orig_err) { + virSetError(orig_err); + virFreeError(orig_err); } } -- 2.7.4

2 1

[libvirt] [PATCH 0/9] i386: query-cpu-model-expansion test script
by Eduardo Habkost 20 Jan '17

20 Jan '17

This is a follow-up to the series that implements query-cpu-model-expansion. Before including the test script, the series has some fixes to allow the results of query-cpu-model-expansion to be used in the QEMU command-line. The script probably will work on s390x too, but I couldn't test it yet. This series and its dependencies can be pulled from the branch: https://github.com/ehabkost/qemu-hacks.git work/x86-query-cpu-expansion-test --- Cc: Cornelia Huck <cornelia.huck(a)de.ibm.com> Cc: Christian Borntraeger <borntraeger(a)de.ibm.com> Cc: David Hildenbrand <david(a)redhat.com> Cc: libvir-list(a)redhat.com Cc: Jiri Denemark <jdenemar(a)redhat.com> Cc: "Jason J. Herne" <jjherne(a)linux.vnet.ibm.com> Cc: Markus Armbruster <armbru(a)redhat.com> Cc: Richard Henderson <rth(a)twiddle.net> Cc: Igor Mammedov <imammedo(a)redhat.com> Cc: Eric Blake <eblake(a)redhat.com> Eduardo Habkost (9): target-i386: Move "host" properties to base class target-i386: Allow short strings to be used as vendor ID cpu: Support comma escaping when parsing -cpu qemu.py: Make logging optional qtest.py: Support QTEST_LOG environment variable qtest.py: make logging optional qtest.py: Make 'binary' parameter optional tests: Add rules to non-gtester qtest test cases tests: Test case for query-cpu-model-expansion scripts/qemu.py | 25 ++- scripts/qtest.py | 15 +- qom/cpu.c | 32 ++-- target/i386/cpu.c | 83 ++++----- tests/test-x86-cpuid-compat.c | 19 ++ tests/Makefile.include | 40 ++++- tests/query-cpu-model-test.py | 398 ++++++++++++++++++++++++++++++++++++++++++ 7 files changed, 551 insertions(+), 61 deletions(-) create mode 100755 tests/query-cpu-model-test.py -- 2.11.0.259.g40922b1

4 10

[libvirt] [PATCH python] Protect against user accidentally calling constructors directly
by Daniel P. Berrange 20 Jan '17

20 Jan '17

When using libvirt python you must never call the object constructors directly, as these are expecting to be passed a wrapped C object. For example import libvirt c = libvirt.virConnect("qemu:///system") c.listAllDomains() will mysteriously segfault. With this change the user now gets an slightly more helpful error Traceback (most recent call last): File "<stdin>", line 1, in <module> File "/home/berrange/src/virt/libvirt-python/build/libvirt.py", line 3409, in __init__ raise Exception("Expected a wrapped C Object but got %s" % type(_obj)) Exception: Expected a wrapped C Object but got <type 'str'> Signed-off-by: Daniel P. Berrange <berrange(a)redhat.com> --- generator.py | 2 ++ 1 file changed, 2 insertions(+) diff --git a/generator.py b/generator.py index e9be8b1..6bcf80b 100755 --- a/generator.py +++ b/generator.py @@ -1532,6 +1532,8 @@ def buildWrappers(module): elif classname in [ "virDomainSnapshot" ]: classes.write(" self._dom = dom\n") classes.write(" self._conn = dom.connect()\n") + classes.write(" if type(_obj).__name__ not in [\"PyCapsule\", \"PyCObject\"]:\n") + classes.write(" raise Exception(\"Expected a wrapped C Object but got %s\" % type(_obj))\n") classes.write(" self._o = _obj\n\n") destruct=None if classname in classes_destructors: -- 2.9.3

2 1

[libvirt] [PATCH] virDomainHostdevSubsysSCSIVHostDefParseXML: Don't leak @wwpn
by Michal Privoznik 20 Jan '17

20 Jan '17

==24748== 12 bytes in 2 blocks are definitely lost in loss record 25 of 84 ==24748== at 0x4C2BF80: malloc (vg_replace_malloc.c:296) ==24748== by 0x1A1E1E78: xmlStrndup (in /usr/lib64/libxml2.so.2.9.4) ==24748== by 0x18D0495F: virXMLPropString (virxml.c:506) ==24748== by 0x18D1FB3E: virDomainHostdevSubsysSCSIVHostDefParseXML (domain_conf.c:6280) ==24748== by 0x18D20350: virDomainHostdevDefParseXMLSubsys (domain_conf.c:6450) ==24748== by 0x18D34E7D: virDomainHostdevDefParseXML (domain_conf.c:13218) ==24748== by 0x18D42598: virDomainDefParseXML (domain_conf.c:17745) ==24748== by 0x18D440A9: virDomainDefParseNode (domain_conf.c:18236) ==24748== by 0x18D43EFA: virDomainDefParse (domain_conf.c:18180) ==24748== by 0x18D43FA0: virDomainDefParseFile (domain_conf.c:18206) ==24748== by 0x44EDA1: testCompareDomXML2XMLFiles (testutils.c:1140) ==24748== by 0x4365F8: testXML2XMLActive (qemuxml2xmltest.c:59) Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- src/conf/domain_conf.c | 20 +++++++++++--------- 1 file changed, 11 insertions(+), 9 deletions(-) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 52aee2b02..55d2c25b2 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -6274,13 +6274,14 @@ static int virDomainHostdevSubsysSCSIVHostDefParseXML(xmlNodePtr sourcenode, virDomainHostdevDefPtr def) { - char *protocol = NULL; + char *protocol = NULL, *wwpn = NULL; virDomainHostdevSubsysSCSIVHostPtr hostsrc = &def->source.subsys.u.scsi_host; + int ret = -1; if (!(protocol = virXMLPropString(sourcenode, "protocol"))) { virReportError(VIR_ERR_XML_ERROR, "%s", _("Missing scsi_host subsystem protocol")); - return -1; + return ret; } if ((hostsrc->protocol = @@ -6293,17 +6294,19 @@ virDomainHostdevSubsysSCSIVHostDefParseXML(xmlNodePtr sourcenode, switch ((virDomainHostdevSubsysSCSIHostProtocolType) hostsrc->protocol) { case VIR_DOMAIN_HOSTDEV_SUBSYS_SCSI_HOST_PROTOCOL_TYPE_VHOST: - if (!(hostsrc->wwpn = virXMLPropString(sourcenode, "wwpn"))) { + if (!(wwpn = virXMLPropString(sourcenode, "wwpn"))) { virReportError(VIR_ERR_XML_ERROR, "%s", _("missing vhost-scsi hostdev source wwpn")); goto cleanup; } - if (!STRPREFIX(hostsrc->wwpn, "naa.") || - !virValidateWWN(hostsrc->wwpn + 4)) { + if (!STRPREFIX(wwpn, "naa.") || + !virValidateWWN(wwpn + 4)) { virReportError(VIR_ERR_XML_ERROR, "%s", _("malformed 'wwpn' value")); goto cleanup; } + hostsrc->wwpn = wwpn; + wwpn = NULL; break; case VIR_DOMAIN_HOSTDEV_SUBSYS_SCSI_HOST_PROTOCOL_TYPE_NONE: case VIR_DOMAIN_HOSTDEV_SUBSYS_SCSI_HOST_PROTOCOL_TYPE_LAST: @@ -6314,12 +6317,11 @@ virDomainHostdevSubsysSCSIVHostDefParseXML(xmlNodePtr sourcenode, break; } - return 0; - + ret = 0; cleanup: - VIR_FREE(hostsrc->wwpn); + VIR_FREE(wwpn); VIR_FREE(protocol); - return -1; + return ret; } -- 2.11.0

2 1

[libvirt] [PATCH v3] qemu: set default vhost-user ifname
by Michal Privoznik 20 Jan '17

20 Jan '17

Based on work of Mehdi Abaakouk <sileht(a)sileht.net>. When parsing vhost-user interface XML and no ifname is found we can try to fill it in in post parse callback. The way this works is we try to make up interface name from given socket path and then ask openvswitch wether it knows the interface. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- diff to v2: - All the review points from v2 worked in src/libvirt_private.syms | 1 + src/qemu/qemu_domain.c | 21 +++++--- src/util/virnetdevopenvswitch.c | 58 ++++++++++++++++++++++ src/util/virnetdevopenvswitch.h | 4 ++ tests/Makefile.am | 7 +++ tests/qemuxml2argvmock.c | 8 +++ tests/qemuxml2xmlmock.c | 33 ++++++++++++ .../qemuxml2xmlout-net-vhostuser.xml | 2 + tests/qemuxml2xmltest.c | 2 +- 9 files changed, 129 insertions(+), 7 deletions(-) create mode 100644 tests/qemuxml2xmlmock.c diff --git a/src/libvirt_private.syms b/src/libvirt_private.syms index 2d23e462d..9e4e8f83f 100644 --- a/src/libvirt_private.syms +++ b/src/libvirt_private.syms @@ -2057,6 +2057,7 @@ virNetDevMidonetUnbindPort; # util/virnetdevopenvswitch.h virNetDevOpenvswitchAddPort; virNetDevOpenvswitchGetMigrateData; +virNetDevOpenvswitchGetVhostuserIfname; virNetDevOpenvswitchInterfaceStats; virNetDevOpenvswitchRemovePort; virNetDevOpenvswitchSetMigrateData; diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c index acfa69589..eed5745b6 100644 --- a/src/qemu/qemu_domain.c +++ b/src/qemu/qemu_domain.c @@ -41,6 +41,7 @@ #include "domain_addr.h" #include "domain_event.h" #include "virtime.h" +#include "virnetdevopenvswitch.h" #include "virstoragefile.h" #include "virstring.h" #include "virthreadjob.h" @@ -3028,12 +3029,20 @@ qemuDomainDeviceDefPostParse(virDomainDeviceDefPtr dev, def->emulator); } - if (dev->type == VIR_DOMAIN_DEVICE_NET && - dev->data.net->type != VIR_DOMAIN_NET_TYPE_HOSTDEV && - !dev->data.net->model) { - if (VIR_STRDUP(dev->data.net->model, - qemuDomainDefaultNetModel(def, qemuCaps)) < 0) - goto cleanup; + if (dev->type == VIR_DOMAIN_DEVICE_NET) { + if (dev->data.net->type != VIR_DOMAIN_NET_TYPE_HOSTDEV && + !dev->data.net->model) { + if (VIR_STRDUP(dev->data.net->model, + qemuDomainDefaultNetModel(def, qemuCaps)) < 0) + goto cleanup; + } + if (dev->data.net->type == VIR_DOMAIN_NET_TYPE_VHOSTUSER && + !dev->data.net->ifname) { + if (virNetDevOpenvswitchGetVhostuserIfname( + dev->data.net->data.vhostuser->data.nix.path, + &dev->data.net->ifname) < 0) + goto cleanup; + } } /* set default disk types and drivers */ diff --git a/src/util/virnetdevopenvswitch.c b/src/util/virnetdevopenvswitch.c index f003b3b13..d93653317 100644 --- a/src/util/virnetdevopenvswitch.c +++ b/src/util/virnetdevopenvswitch.c @@ -377,3 +377,61 @@ virNetDevOpenvswitchInterfaceStats(const char *ifname, virCommandFree(cmd); return ret; } + +/** + * virNetDevOpenvswitchVhostuserGetIfname: + * @path: the path of the unix socket + * @ifname: the retrieved name of the interface + * + * Retreives the ovs ifname from vhostuser unix socket path. + * + * Returns: 1 if interface is an openvswitch interface, + * 0 if it is not, but no other error occurred, + * -1 otherwise. + */ +int +virNetDevOpenvswitchGetVhostuserIfname(const char *path, + char **ifname) +{ + virCommandPtr cmd = NULL; + char *tmpIfname = NULL; + char **tokens = NULL; + size_t ntokens = 0; + int status; + int ret = -1; + + /* Openvswitch vhostuser path are hardcoded to + * /<runstatedir>/openvswitch/<ifname> + * for example: /var/run/openvswitch/dpdkvhostuser0 + * + * so we pick the filename and check it's a openvswitch interface + */ + if ((tokens = virStringSplitCount(path, "/", 0, &ntokens))) { + if (VIR_STRDUP(tmpIfname, tokens[ntokens - 1]) < 0) + goto cleanup; + } + + if (!tmpIfname) { + ret = 0; + goto cleanup; + } + + cmd = virCommandNewArgList(OVSVSCTL, "--timeout=5", "get", "Interface", + tmpIfname, "name", NULL); + if (virCommandRun(cmd, &status) < 0 || + status) { + /* it's not a openvswitch vhostuser interface. */ + ret = 0; + goto cleanup; + } + + *ifname = tmpIfname; + tmpIfname = NULL; + ret = 1; + + cleanup: + virStringListFreeCount(tokens, ntokens); + virCommandFree(cmd); + VIR_FREE(tmpIfname); + return ret; +} diff --git a/src/util/virnetdevopenvswitch.h b/src/util/virnetdevopenvswitch.h index 0f9e1dfa6..8f5faf14e 100644 --- a/src/util/virnetdevopenvswitch.h +++ b/src/util/virnetdevopenvswitch.h @@ -52,4 +52,8 @@ int virNetDevOpenvswitchInterfaceStats(const char *ifname, virDomainInterfaceStatsPtr stats) ATTRIBUTE_NONNULL(1) ATTRIBUTE_RETURN_CHECK; +int virNetDevOpenvswitchGetVhostuserIfname(const char *path, + char **ifname) + ATTRIBUTE_NONNULL(1) ATTRIBUTE_NONNULL(2) ATTRIBUTE_RETURN_CHECK; + #endif /* __VIR_NETDEV_OPENVSWITCH_H__ */ diff --git a/tests/Makefile.am b/tests/Makefile.am index ecd04e895..c8584fea6 100644 --- a/tests/Makefile.am +++ b/tests/Makefile.am @@ -414,6 +414,7 @@ if WITH_QEMU test_libraries += libqemumonitortestutils.la \ libqemutestdriver.la \ qemuxml2argvmock.la \ + qemuxml2xmlmock.la \ qemucaps2xmlmock.la \ qemucapsprobemock.la \ $(NULL) @@ -571,6 +572,12 @@ qemuxml2argvmock_la_CFLAGS = $(AM_CFLAGS) qemuxml2argvmock_la_LDFLAGS = $(MOCKLIBS_LDFLAGS) qemuxml2argvmock_la_LIBADD = $(MOCKLIBS_LIBS) +qemuxml2xmlmock_la_SOURCES = \ + qemuxml2xmlmock.c +qemuxml2xmlmock_la_CFLAGS = $(AM_CFLAGS) +qemuxml2xmlmock_la_LDFLAGS = $(MOCKLIBS_LDFLAGS) +qemuxml2xmlmock_la_LIBADD = $(MOCKLIBS_LIBS) + qemuxml2xmltest_SOURCES = \ qemuxml2xmltest.c testutilsqemu.c testutilsqemu.h \ testutils.c testutils.h diff --git a/tests/qemuxml2argvmock.c b/tests/qemuxml2argvmock.c index c501b59ac..177b24e0a 100644 --- a/tests/qemuxml2argvmock.c +++ b/tests/qemuxml2argvmock.c @@ -28,6 +28,7 @@ #include "virnetdev.h" #include "virnetdevip.h" #include "virnetdevtap.h" +#include "virnetdevopenvswitch.h" #include "virnuma.h" #include "virrandom.h" #include "virscsi.h" @@ -180,3 +181,10 @@ virCryptoGenerateRandom(size_t nbytes) return buf; } + +int +virNetDevOpenvswitchGetVhostuserIfname(const char *path ATTRIBUTE_UNUSED, + char **ifname) +{ + return VIR_STRDUP(*ifname, "vhost-user0"); +} diff --git a/tests/qemuxml2xmlmock.c b/tests/qemuxml2xmlmock.c new file mode 100644 index 000000000..0d3e6f2bd --- /dev/null +++ b/tests/qemuxml2xmlmock.c @@ -0,0 +1,33 @@ +/* + * Copyright (C) 2016 Red Hat, Inc. + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library. If not, see + * <http://www.gnu.org/licenses/>. + * + * Author: Michal Privoznik <mprivozn(a)redhat.com> + */ + +#include <config.h> + +#include "virnetdevopenvswitch.h" +#include "virstring.h" + +#define VIR_FROM_THIS VIR_FROM_NONE + +int +virNetDevOpenvswitchGetVhostuserIfname(const char *path ATTRIBUTE_UNUSED, + char **ifname) +{ + return VIR_STRDUP(*ifname, "vhost-user0"); +} diff --git a/tests/qemuxml2xmloutdata/qemuxml2xmlout-net-vhostuser.xml b/tests/qemuxml2xmloutdata/qemuxml2xmlout-net-vhostuser.xml index ac1d7e110..2bdcac358 100644 --- a/tests/qemuxml2xmloutdata/qemuxml2xmlout-net-vhostuser.xml +++ b/tests/qemuxml2xmloutdata/qemuxml2xmlout-net-vhostuser.xml @@ -30,12 +30,14 @@ <interface type='vhostuser'> <mac address='52:54:00:ee:96:6b'/> <source type='unix' path='/tmp/vhost0.sock' mode='server'/> + <target dev='vhost-user0'/> <model type='virtio'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x03' function='0x0'/> </interface> <interface type='vhostuser'> <mac address='52:54:00:ee:96:6c'/> <source type='unix' path='/tmp/vhost1.sock' mode='client'/> + <target dev='vhost-user0'/> <model type='virtio'/> <address type='pci' domain='0x0000' bus='0x00' slot='0x04' function='0x0'/> </interface> diff --git a/tests/qemuxml2xmltest.c b/tests/qemuxml2xmltest.c index ddd17cb1e..30931f581 100644 --- a/tests/qemuxml2xmltest.c +++ b/tests/qemuxml2xmltest.c @@ -1031,7 +1031,7 @@ mymain(void) return ret == 0 ? EXIT_SUCCESS : EXIT_FAILURE; } -VIRT_TEST_MAIN(mymain) +VIRT_TEST_MAIN_PRELOAD(mymain, abs_builddir "/.libs/qemuxml2xmlmock.so") #else -- 2.11.0

3 3

[libvirt] [PATCH] qemu: hotplug: Properly emit "DEVICE_DELETED" event when unplugging memory
by Peter Krempa 20 Jan '17

20 Jan '17

The event needs to be emitted after the last monitor call, so that it's not possible to find the device in the XML accidentally while the vm object is unlocked. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1414393 --- src/qemu/qemu_hotplug.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/src/qemu/qemu_hotplug.c b/src/qemu/qemu_hotplug.c index 5b0a3f515..a6de25407 100644 --- a/src/qemu/qemu_hotplug.c +++ b/src/qemu/qemu_hotplug.c @@ -3691,9 +3691,6 @@ qemuDomainRemoveMemoryDevice(virQEMUDriverPtr driver, VIR_DEBUG("Removing memory device %s from domain %p %s", mem->info.alias, vm, vm->def->name); - event = virDomainEventDeviceRemovedNewFromObj(vm, mem->info.alias); - qemuDomainEventQueue(driver, event); - if (virAsprintf(&backendAlias, "mem%s", mem->info.alias) < 0) return -1; @@ -3708,6 +3705,9 @@ qemuDomainRemoveMemoryDevice(virQEMUDriverPtr driver, if (rc < 0) return -1; + event = virDomainEventDeviceRemovedNewFromObj(vm, mem->info.alias); + qemuDomainEventQueue(driver, event); + if ((idx = virDomainMemoryFindByDef(vm->def, mem)) >= 0) virDomainMemoryRemove(vm->def, idx); -- 2.11.0

2 1

[libvirt] [ libvirt-python BUG ?] Memory leak in libvirt_virConnectDomainEventJobCompletedCallback
by Hailiang Zhang 20 Jan '17

20 Jan '17

Hi all, It seems that, there is a memory leak bug in libvirt_virConnectDomainEventJobCompletedCallback() function of libvirt-override.c. We didn't decrease the reference of 'pyobj_dict' in this function or any other places while pyt0bj_ret is not zero. libvirt_virConnectDomainEventJobCompletedCallback() { ... ... pyobj_ret = PyObject_CallMethod(pyobj_conn, (char*)"_dispatchDomainEventJobCompletedCallback", (char*)"OOO", pyobj_dom, pyobj_dict, pyobj_cbData); Py_DECREF(pyobj_cbData); Py_DECREF(pyobj_dom); cleanup: if (!pyobj_ret) { DEBUG("%s - ret:%p\n", __FUNCTION__, pyobj_ret); PyErr_Print(); Py_XDECREF(pyobj_dict); } else { Py_DECREF(pyobj_ret); ret = 0; } LIBVIRT_RELEASE_THREAD_STATE; return ret; } I don't think the reference of pyobj_dict will be stolen by PyObject_CallMethod when it is passed to the Python function _dispatchDomainEventJobCompletedCallback. In fact, we have a test which triggered lots of events and found that the memory of the test process will keep raising slowly and could not be reclaimed even we called gc.collect() explicitly in our codes. So, is it a bug ? Any comments ? Thanks, hailiang

2 1

[libvirt] [V4] RFC for support cache tune(CAT) in libvirt
by 乔立勇(Eli Qiao) 20 Jan '17

20 Jan '17

hi all Sorry for resend this again, I forget to add subject :( ... Thanks all for the comment for previous RFC version, I summary all input here also with Opens which I am not sure, can you please review again? #Propose Changes ## Expose cache information into capabilities Extend capabilities to expose all level cache resource information and should tell the topology of caches also tell which kinds of resources can be tuned/allocation There information comes from /sys/devices/system/cpu/cpu0/cache/index${level}/size also /sys/fs/resctrl/ (this is CAT sys interface in linux kernel) virsh capabilities <cache> <bank id='0, 'type="l3" size="56320" units="KiB" cpus="0,1,2,6,7,8"/> <--------------------- level 3 cache is per socket, so group them by socket id <control unit="KiB" min="2816"/> <bank id='1', type="l3" size="56320" units="KiB" cpus="3,4,5,9,10,11"/> <bank id='2' type="l2" size="256" units="KiB" cpus="0"/> <bank id='3' type="l2" size="256" units="KiB" cpus="1"/> <bank id='4' type="l2" size="256" units="KiB" cpus="2"/> <bank id='5' type="l2" size="256" units="KiB" cpus="3"/> <bank id='6' type="l2" size="256" units="KiB" cpus="4"/> ... <cache> Opens 1. how about add socket id to bank for bank type = l3 ? 2. do we really want to expose l2/l3 cache for now , they are per core resource and linux kernel don't support l2 yet (depend no hardware)? 3. if enable CDP in resctrl, for bank type=l3 , it will be split to l3data l3code, should expose this ability. <bank type="l3" size="56320" units="KiB" cpus="0,1,2,6,7,8"/> <--------------------- level 3 cache is per socket, so group them by socket id <control unit="KiB" min="2816" cdp="enabled"/> ## Provide a new API to get the avail cache on each bank, such as the output are: id=0 type=l3 avail=56320 total = ?? <--------- do we need this? id=1 type=l3 avail=56320 id=3 type=l2 avail=256 Opens: · Don't expose the avail cache information if the host can not do the allocation of that type cache(eg, for l2 currently) ? · We can not make all of the cache , the reservation amount is the min_cbm_len (=1) * min_unit . · do we need to expose total? ## enable CAT for a domain 1 Domain XML changes <cputune> <cache id="1" host_id="0" type="l3" size="5632" unit="KiB"/> <cache id="2" host_id="1" type="l3" size="5632" unit="KiB"/> <cpu_cache vcpus="0-3" id="1"/> <cpu_cache vcpus="4-7" id="2"/> <iothread_cache iothreads="0-1" id="1"/> <emulator_cache id="2"/> </cputune> 2. Extend cputune command ? Opens: 1. Do we accept to extend existed API ? or using new API/virsh? 2. How to calculate cache size -> CBM bit? eg: 5632/ 2816 = 2 bits 5733/ 2816 = 2 bits or 3 bits? ## Restriction for using cache tune on multiple sockets' host. The l3 cache is per socket resource, kernel need to know about what's affinity looks like, so for a VM which running on a multiple socket's host, it should have NUMA setting or vcpuset pin setting. Or cache tune will fail. [1] kernel support https://git.kernel.org/cgit/linux/kernel/git/tip/tip .git/tree/arch/x86/kernel/cpu/intel_rdt.c?h=x86/cache [2] libvirt PoC(not finished yet) https://github.com/taget/ libvirt/commits/cat_new -- Best regards - Eli 天涯无处不重逢 a leaf duckweed belongs to the sea , where not to meet in life

3 3

[libvirt] Mirror on Github failed to sync from libvirt.org
by Chen Hanxiao 20 Jan '17

20 Jan '17

Dear list, Mirror on Github: https://github.com/libvirt/libvirt did not sync from libvirt.org for almost a month. Regards, - Chen

2 1

[libvirt] [PATCH] Interface: return appropriate status for bridge device
by Lin Ma 20 Jan '17

20 Jan '17

In function udevInterfaceIsActive, The current design relies on the sys attribute 'operstate' to determine whether the interface is active. For the device node of virtual network(say virbr0), There is already a dummy tap device to maintain a fixed mac on it, So it's available and its status should be considered as active. But if no anyelse tap device connects to it, The value of operstate of virbr0 in sysfs is 'down', So udevInterfaceIsActive returns 0, It causes 'virsh iface-list --all' to report the status of virbr0 as 'inactive'. This patch fixes the issue by counting slave devices for bridge device. Signed-off-by: Lin Ma <lma(a)suse.com> --- src/interface/interface_backend_udev.c | 22 ++++++++++++++++++++++ 1 file changed, 22 insertions(+) diff --git a/src/interface/interface_backend_udev.c b/src/interface/interface_backend_udev.c index 5d0fc64..9ac4674 100644 --- a/src/interface/interface_backend_udev.c +++ b/src/interface/interface_backend_udev.c @@ -1127,6 +1127,11 @@ udevInterfaceIsActive(virInterfacePtr ifinfo) struct udev_device *dev; virInterfaceDefPtr def = NULL; int status = -1; + struct dirent **member_list = NULL; + const char *devtype; + int member_count = 0; + char *member_path; + size_t i; dev = udev_device_new_from_subsystem_sysname(udev, "net", ifinfo->name); @@ -1146,6 +1151,23 @@ udevInterfaceIsActive(virInterfacePtr ifinfo) /* Check if it's active or not */ status = STREQ(udev_device_get_sysattr_value(dev, "operstate"), "up"); + if (!status) { + devtype = udev_device_get_devtype(dev); + if (STREQ_NULLABLE(devtype, "bridge")) { + if (virAsprintf(&member_path, "%s/%s", + udev_device_get_syspath(dev), "brif") < 0) + goto cleanup; + member_count = scandir(member_path, &member_list, + udevBridgeScanDirFilter, alphasort); + if (member_count > 0) + status = 1; + for (i = 0; i < member_count; i++) + VIR_FREE(member_list[i]); + VIR_FREE(member_list); + VIR_FREE(member_path); + } + } + udev_device_unref(dev); cleanup: -- 2.9.2

1 0