May 2016 - Devel - libvirt List Archives

[libvirt] [PATCH 0/4] rpc: Fix fd leak and potential crash
by Cole Robinson 05 May '16

05 May '16

Patch 1 was already sent, I'd like to get it in for the release, it should be quite safe. Remaining patches centralize some cleanup code and fix a potential crasher, but since it should be rare I'd like to wait till after the release. Previous thread: http://www.redhat.com/archives/libvir-list/2016-April/msg01618.html Ben Gray (1): rpc: Don't leak fd via CreateXMLWithFiles Cole Robinson (3): rpc: Add virNetMessageClearPayload rpc: Clear more in virNetMessageClearPayload rpc: use virNetMessageClearPayload in client src/libvirt_remote.syms | 1 + src/rpc/virnetclient.c | 5 +---- src/rpc/virnetmessage.c | 28 +++++++++++++++++++--------- src/rpc/virnetmessage.h | 2 ++ 4 files changed, 23 insertions(+), 13 deletions(-) -- 2.7.4

2 10

[libvirt] [PATCH v2 0/8] Add IV Secret Object support
by John Ferlan 05 May '16

05 May '16

v1 here (specifically patches 11 & 12) http://www.redhat.com/archives/libvir-list/2016-April/msg01077.html What's new/changed: Patches 1-3 address comments made by Jan in his review of v1. Patch 4 is essentially following through on the Patch 3 comment Patch 5 was part of patch 11, but I split it out for easier review Patch 6 is the remainder of the former patch 11. Other than using the ...SECRET_INFO_TYPE_xxx nomenclature, no other real changes. Patch 7 was part of patch 12, but I split it out to reduce the amount to review. Beyond the aforementioned ...SECRET_INFO_TYPE_xxx changes, I removed qemuDomainSecretInfoGetAlias. Initially created to help with any hotplug code, but that's not necessary, so remove it for now. It was brought up on Jan's comments as well. Create qemuDomainSecretSetup in order to address Jan's other comment about a common API This code could be merged with patch 8, but separating it *and* leaving qemuDomainSecretHaveEncrypt essentially brain-dead worked made for a shorter pile to look at Patch 8 is the remainder of patch 12. I investigated extracting the iSCSI and RBD code out, but it really just didn't make practice sense. I altered qemuBuildSecretIVCommandLine to just access the IV alias directly since the callers were adjusted to only try to build the IV object if necessary. The qemuBuild{Disk|Hostdev}iSCSICommandLine were renamed to qemuBuild{Disk|Hostdev}SecinfoCommandLine and each encapsulated the call to qemuBuildSecretIVCommandLine rather than inlining it in each of the callers The qemuDomainSecretHaveEncrypt will not be braindead. None of the tests changed, so it doesn't seem I messed anything up with all the requested adjustments. John Ferlan (8): qemu: Adjust names of qemuDomainSecretInfoType enums qemu: Split out the master key create and write qemu: Move qemuDomainSecretPrepare to qemuProcessPrepareDomain qemu: Move qemuDomainSecretDestroy to qemuProcessLaunch qemu: Separate network URI command building code qemu: Introduce qemuDomainSecretIV qemu: Introduce new Secret IV API's qemu: Utilize qemu secret objects for SCSI/RBD auth/secret configure.ac | 1 + src/qemu/qemu_alias.c | 23 ++ src/qemu/qemu_alias.h | 2 + src/qemu/qemu_command.c | 359 +++++++++++++++++++-- src/qemu/qemu_domain.c | 242 ++++++++++++-- src/qemu/qemu_domain.h | 34 +- src/qemu/qemu_hotplug.c | 6 +- src/qemu/qemu_process.c | 23 +- ...uxml2argv-disk-drive-network-iscsi-auth-IV.args | 39 +++ ...muxml2argv-disk-drive-network-iscsi-auth-IV.xml | 43 +++ ...emuxml2argv-disk-drive-network-rbd-auth-IV.args | 31 ++ ...qemuxml2argv-disk-drive-network-rbd-auth-IV.xml | 42 +++ ...emuxml2argv-hostdev-scsi-lsi-iscsi-auth-IV.args | 41 +++ ...qemuxml2argv-hostdev-scsi-lsi-iscsi-auth-IV.xml | 48 +++ ...xml2argv-hostdev-scsi-virtio-iscsi-auth-IV.args | 43 +++ ...uxml2argv-hostdev-scsi-virtio-iscsi-auth-IV.xml | 48 +++ tests/qemuxml2argvmock.c | 31 +- tests/qemuxml2argvtest.c | 19 ++ 18 files changed, 1009 insertions(+), 66 deletions(-) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-drive-network-iscsi-auth-IV.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-drive-network-iscsi-auth-IV.xml create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-drive-network-rbd-auth-IV.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-drive-network-rbd-auth-IV.xml create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-hostdev-scsi-lsi-iscsi-auth-IV.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-hostdev-scsi-lsi-iscsi-auth-IV.xml create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-hostdev-scsi-virtio-iscsi-auth-IV.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-hostdev-scsi-virtio-iscsi-auth-IV.xml -- 2.5.5

3 20

[libvirt] Fwd: Dump network traffic from each VM.
by Julio Faracco 05 May '16

05 May '16

Forwarding to devel-list. Julio Cesar Faracco ---------- Forwarded message ---------- From: Julio Faracco <jcfaracco(a)gmail.com> Date: 2016-05-04 12:37 GMT-03:00 Subject: Dump network traffic from each VM. To: libvirt-users(a)redhat.com Hi guys, Does anybody know how to dump the network trafic from each virtual machine using libvirt? Here, we usually use VPN softwares and other network tools inside Virtual Machines (Windows and Linux). Sometimes, this applications does not work properly, so we need to dump the network traffic of each Virtual Machine to compare and debug. How can I enable it using libvirt? I know that if you use a QEMU instance, you can enable the dump using "-netdev dump,...". Thanks! :-) Julio Cesar Faracco

1 0

[libvirt] [PATCH 00/27] Remove many conditionals from the RPM
by Daniel P. Berrange 05 May '16

05 May '16

Cole previously suggested that we should simplify the RPM spec by removing conditionals that we never need to toggle for RHEL / Fedora builds, rather than having conditionals for every single feature that exists. This series does just that.... Daniel P. Berrange (27): libvirt.spec.in: explicitly disable xenapi & vz drivers libvirt.spec.in: drop Fedora < 20 and RHEL < 6 libvirt.spec.in: remove client only build option libvirt.spec.in: remove conditional for driver modules libvirt.spec.in: remove libvirtd conditional libvirt.spec.in: remove avahi conditional libvirt.spec.in: remove hal conditional libvirt.spec.in: remove apparmor conditional libvirt.spec.in: remove capng conditional libvirt.spec.in: remove udev conditional libvirt.spec.in: remove netcf conditional libvirt.spec.in: remove audit conditional libvirt.spec.in: remove dtrace conditional libvirt.spec.in: remove sasl conditionals libvirt.spec.in: remove polkit conditional libvirt.spec.in: remove yajl conditional libvirt.spec.in: remove selinux conditional libvirt.spec.in: remove interface driver conditional libvirt.spec.in: remove network driver conditiaonl libvirt.spec.in: remove most storage conditionals libvirt.spec.in: remove nodedev conditional libvirt.spec.in: remove nwfilter driver conditional libvirt.spec.in: remove libpcap conditional libvirt.spec.in: remove macvtap conditional libvirt.spec.in: remove libnl conditional libvirt.spec.in: remove nss plugin conditional libvirt.spec.in: use explicit --with-xxx for all features libvirt.spec.in | 1205 +++++++++++++++---------------------------------------- 1 file changed, 333 insertions(+), 872 deletions(-) -- 2.5.5

2 30

[libvirt] [PATCH 0/2] Two trivial fixes
by Michal Privoznik 05 May '16

05 May '16

*** BLURB HERE *** Michal Privoznik (2): virNetServerClientNewPostExecRestart: Drop useless typecasts virNetServerClientNewPostExecRestart: Avoid align problems src/rpc/virnetserverclient.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) -- 2.8.1

2 6

[libvirt] [PATCH v2] Get the maxvcpus from both the qemuCaps and /dev/kvm
by Shivaprasad G Bhat 05 May '16

05 May '16

For some archs and machines, the maxvcpus defined in the kernel can be different from the qemu-caps maxvcpus. Just reporting the kernel defined maxvcpus is not be sufficient. virsh domacapabilities and virsh maxvcpus --type kvm return different maxvcpus values and is confusing as to know what actually works. The minimum of the two values is what actually works. For example on PPC64, the KVM_MAX_VCPUS is defined to be 1024 in kernel where as qemu has MAX_CPUMASK_BITS defined at 255 in include/sysemu/sysemu.h. The guest can go upto 256 vcpus here. Signed-off-by: Shivaprasad G Bhat <sbhat(a)linux.vnet.ibm.com> --- src/qemu/qemu_driver.c | 27 ++++++++++++++++++++++++--- 1 file changed, 24 insertions(+), 3 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 3d0c7c8..2716af8 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -1255,10 +1255,31 @@ static int qemuConnectIsAlive(virConnectPtr conn ATTRIBUTE_UNUSED) static int -kvmGetMaxVCPUs(void) +kvmGetMaxVCPUs(virConnectPtr conn) { int fd; int ret; + int qemuCapsMaxVcpus = 0; + virArch arch = virArchFromHost(); + virQEMUCapsPtr qemuCaps = NULL; + virQEMUDriverPtr driver = conn->privateData; + const char *machine; + + if (!(qemuCaps = virQEMUCapsCacheLookupByArch(driver->qemuCapsCache, + arch))) { + virReportError(VIR_ERR_INVALID_ARG, + _("unable to find any emulator to serve '%s' " + "architecture"), virArchToString(arch)); + return -1; + } + + if (!(machine = virQEMUCapsGetDefaultMachine(qemuCaps))) { + virObjectUnref(qemuCaps); + return -1; + } + + qemuCapsMaxVcpus = virQEMUCapsGetMachineMaxCpus(qemuCaps, machine); + virObjectUnref(qemuCaps); if ((fd = open(KVM_DEVICE, O_RDONLY)) < 0) { virReportSystemError(errno, _("Unable to open %s"), KVM_DEVICE); @@ -1282,7 +1303,7 @@ kvmGetMaxVCPUs(void) cleanup: VIR_FORCE_CLOSE(fd); - return ret; + return ret > qemuCapsMaxVcpus ? qemuCapsMaxVcpus : ret; } @@ -1323,7 +1344,7 @@ qemuConnectGetMaxVcpus(virConnectPtr conn ATTRIBUTE_UNUSED, const char *type) return 16; if (STRCASEEQ(type, "kvm")) - return kvmGetMaxVCPUs(); + return kvmGetMaxVCPUs(conn); if (STRCASEEQ(type, "kqemu")) return 1;

5 11

[libvirt] [PATCH] Add qemucapsprobe in .gitignore
by Jiri Denemark 05 May '16

05 May '16

Signed-off-by: Jiri Denemark <jdenemar(a)redhat.com> --- Pushed as trivial. I forgot to squash this into the patch which adds qemucapsprobe :( .gitignore | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/.gitignore b/.gitignore index ba537d8..fc21bad 100644 --- a/.gitignore +++ b/.gitignore @@ -159,8 +159,9 @@ /tests/*.log /tests/*.pid /tests/*.trs -/tests/commandhelper /tests/*test +/tests/commandhelper +/tests/qemucapsprobe !/tests/*schematest !/tests/virt-aa-helper-test /tests/objectlocking -- 2.8.2

1 0

[libvirt] [PATCH 0/6] tests: Add qemucapsprobe helper
by Jiri Denemark 05 May '16

05 May '16

Adding new *.replies files for qemucapabilitiestest or updating the files when libvirt adds an additional QMP command into the probing process is quite painful. The goal of the new qemucapsprobe command is to make this process as easy as tests/qemucapsprobe /path/to/qemu/binary >caps.replies Jiri Denemark (6): tests: Create a shared library with qemu driver tests: Decouple preload code from main() qemu: Make virQEMUCapsNewForBinary usable from tests qemu: Make qemuMonitorJSONIOProcessLine available for tests virjson: Make pretty format more compact tests: Add qemucapsprobe helper src/qemu/qemu_capabilities.c | 38 ++- src/qemu/qemu_capspriv.h | 8 + src/qemu/qemu_monitor_json.c | 2 +- src/qemu/qemu_monitor_json.h | 4 + src/util/virjson.c | 4 +- tests/Makefile.am | 23 ++ tests/qemucapabilitiestest.c | 5 + tests/qemucapsprobe.c | 82 +++++++ tests/qemucapsprobemock.c | 93 ++++++++ tests/testutils.h | 9 +- .../virnetdaemondata/output-data-admin-nomdns.json | 254 ++++++++++----------- .../output-data-admin-server-names.json | 254 ++++++++++----------- .../virnetdaemondata/output-data-anon-clients.json | 128 +++++------ tests/virnetdaemondata/output-data-client-ids.json | 128 +++++------ .../output-data-client-timestamp.json | 132 +++++------ .../output-data-initial-nomdns.json | 128 +++++------ tests/virnetdaemondata/output-data-initial.json | 130 +++++------ .../output-data-no-keepalive-required.json | 254 ++++++++++----------- 18 files changed, 960 insertions(+), 716 deletions(-) create mode 100644 tests/qemucapsprobe.c create mode 100644 tests/qemucapsprobemock.c -- 2.8.2

2 13

[libvirt] [PATCH 1/2] configure: split out UML driver checks
by Roman Bogorodskiy 05 May '16

05 May '16

Move all the UML driver related routines into m4/virt-driver-uml.m4. --- configure.ac | 28 ++-------------------------- m4/virt-driver-uml.m4 | 47 +++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 49 insertions(+), 26 deletions(-) create mode 100644 m4/virt-driver-uml.m4 diff --git a/configure.ac b/configure.ac index 88e2e20..d19c1a9 100644 --- a/configure.ac +++ b/configure.ac @@ -537,10 +537,6 @@ AC_ARG_WITH([qemu], [AS_HELP_STRING([--with-qemu], [add QEMU/KVM support @<:@default=yes@:>@])]) m4_divert_text([DEFAULTS], [with_qemu=yes]) -AC_ARG_WITH([uml], - [AS_HELP_STRING([--with-uml], - [add UML support @<:@default=check@:>@])]) -m4_divert_text([DEFAULTS], [with_uml=check]) AC_ARG_WITH([openvz], [AS_HELP_STRING([--with-openvz], [add OpenVZ support @<:@default=check@:>@])]) @@ -1561,27 +1557,7 @@ dnl dnl Checks for the UML driver dnl -if test "$with_libvirtd" = "no" ; then - with_uml=no -fi -if test "$with_uml" = "yes" || test "$with_uml" = "check"; then - AC_CHECK_HEADER([sys/inotify.h], [ - with_uml=yes - ], [ - if test "$with_uml" = "check"; then - with_uml=no - AC_MSG_NOTICE([<sys/inotify.h> is required for the UML driver, disabling it]) - else - AC_MSG_ERROR([The <sys/inotify.h> is required for the UML driver. Upgrade your libc6.]) - fi - ]) -fi -if test "$with_uml" = "yes" ; then - AC_DEFINE_UNQUOTED([WITH_UML], 1, [whether UML driver is enabled]) -fi -AM_CONDITIONAL([WITH_UML], [test "$with_uml" = "yes"]) - - +LIBVIRT_DRIVER_CHECK_UML dnl dnl check for PHYP @@ -2747,7 +2723,7 @@ AC_MSG_NOTICE([Drivers]) AC_MSG_NOTICE([]) AC_MSG_NOTICE([ Xen: $with_xen]) AC_MSG_NOTICE([ QEMU: $with_qemu]) -AC_MSG_NOTICE([ UML: $with_uml]) +LIBVIRT_DRIVER_RESULT_UML AC_MSG_NOTICE([ OpenVZ: $with_openvz]) AC_MSG_NOTICE([ VMware: $with_vmware]) AC_MSG_NOTICE([ VBox: $with_vbox]) diff --git a/m4/virt-driver-uml.m4 b/m4/virt-driver-uml.m4 new file mode 100644 index 0000000..4132315 --- /dev/null +++ b/m4/virt-driver-uml.m4 @@ -0,0 +1,47 @@ +dnl The UML driver +dnl +dnl Copyright (C) 2005-2015 Red Hat, Inc. +dnl +dnl This library is free software; you can redistribute it and/or +dnl modify it under the terms of the GNU Lesser General Public +dnl License as published by the Free Software Foundation; either +dnl version 2.1 of the License, or (at your option) any later version. +dnl +dnl This library is distributed in the hope that it will be useful, +dnl but WITHOUT ANY WARRANTY; without even the implied warranty of +dnl MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +dnl Lesser General Public License for more details. +dnl +dnl You should have received a copy of the GNU Lesser General Public +dnl License along with this library. If not, see +dnl <http://www.gnu.org/licenses/>. +dnl + +AC_DEFUN([LIBVIRT_DRIVER_CHECK_UML],[ + AC_ARG_WITH([uml], + [AS_HELP_STRING([--with-uml], + [add UML support @<:@default=check@:>@])]) + m4_divert_text([DEFAULTS], [with_uml=check]) + + if test "$with_uml" = "yes" || test "$with_uml" = "check"; then + AC_CHECK_HEADER([sys/inotify.h], [ + with_uml=yes + ], [ + if test "$with_uml" = "check"; then + with_uml=no + AC_MSG_NOTICE([<sys/inotify.h> is required for the UML driver, disabling it]) + else + AC_MSG_ERROR([The <sys/inotify.h> is required for the UML driver. Upgrade your libc6.]) + fi + ]) + fi + + if test "$with_uml" = "yes" ; then + AC_DEFINE_UNQUOTED([WITH_UML], 1, [whether UML driver is enabled]) + fi + AM_CONDITIONAL([WITH_UML], [test "$with_uml" = "yes"]) +]) + +AC_DEFUN([LIBVIRT_DRIVER_RESULT_UML],[ + AC_MSG_NOTICE([ UML: $with_uml]) +]) -- 2.7.4

1 2

[libvirt] [PATCH v3 0/2] bhyve: implement domainShutdown
by Roman Bogorodskiy 05 May '16

05 May '16

Changes from v2: - Split domainDestroy changes and domainShutdown addition into two patches - In bhyveDestroy, close bhyve monitor after running 'bhyvectl --destroy' (reasoning in the commit message) Roman Bogorodskiy (2): bhyve: drop virProcessKillPainfully() from destroy bhyve: implement domainShutdown src/bhyve/bhyve_driver.c | 27 +++++++++++++++++++++++++++ src/bhyve/bhyve_process.c | 43 +++++++++++++++++++++++++++++-------------- src/bhyve/bhyve_process.h | 2 ++ 3 files changed, 58 insertions(+), 14 deletions(-) -- 2.7.4

2 5