April 2010 - Devel - libvirt List Archives

[libvirt] [PATCH TCK] Add a minimal test case for CPU hotplug
by Daniel P. Berrange 23 Apr '10

23 Apr '10

Adds a minimal test case which verifies that it is possible to set the CPU count to max, and then back to the min. This really needs libguestfs integration to validate that the guest sees the change --- scripts/domain/130-cpu-hotplug.t | 78 ++++++++++++++++++++++++++++++++++++++ 1 files changed, 78 insertions(+), 0 deletions(-) create mode 100644 scripts/domain/130-cpu-hotplug.t diff --git a/scripts/domain/130-cpu-hotplug.t b/scripts/domain/130-cpu-hotplug.t new file mode 100644 index 0000000..2976f2d --- /dev/null +++ b/scripts/domain/130-cpu-hotplug.t @@ -0,0 +1,78 @@ +# -*- perl -*- +# +# Copyright (C) 2009-2010 Red Hat, Inc. +# Copyright (C) 2009-2010 Daniel P. Berrange +# +# This program is free software; You can redistribute it and/or modify +# it under the GNU General Public License as published by the Free +# Software Foundation; either version 2, or (at your option) any +# later version +# +# The file "LICENSE" distributed along with this file provides full +# details of the terms and conditions +# + +=pod + +=head1 NAME + +domain/130-cpu-hotplug.t - whether CPU count can be changed + +=head1 DESCRIPTION + +The test case validates the it is possible to change the CPU +on a running guest. + +XXX: Need libguestfs integration to check that it has truely +worked. + +=cut + +use strict; +use warnings; + +use Test::More tests => 7; +use Test::Exception; + +use Sys::Virt::TCK; + +my $tck = Sys::Virt::TCK->new(); +my $conn = eval { $tck->setup(); }; +BAIL_OUT "failed to setup test harness: $@" if $@; +END { $tck->cleanup if $tck; } + + +my $xml = $tck->generic_domain("tck")->as_xml; + + +diag "Creating a new transient domain"; +my $dom; +ok_domain(sub { $dom = $conn->create_domain($xml) }, "created transient domain object"); + +my $max; +lives_ok(sub { $max = $dom->get_max_vcpus() }, "queried max vcpus"); + +SKIP: { + skip "SMP guests not supported", 4 unless $max > 1; + + diag "Increasing CPU count to max"; + lives_ok(sub { $dom->set_vcpus($max); }, "set vcpus to $max"); + + my $info = $dom->get_info(); + + is($info->{nrVirtCpu}, $max, "cpu count $info->{nrVirtCpu} is $max"); + + diag "Decreasing CPU count to min"; + lives_ok(sub { $dom->set_vcpus(1); }, "set vcpus to 1"); + + $info = $dom->get_info(); + is($info->{nrVirtCpu}, 1, "cpu count $info->{nrVirtCpu} is 1"); +} + +diag "Destroying the transient domain"; +$dom->destroy; + +diag "Checking that transient domain has gone away"; +ok_error(sub { $conn->get_domain_by_name("tck") }, "NO_DOMAIN error raised from missing domain", 42); + +# end -- 1.6.5.2

2 2

[libvirt] [PATCH] configure.ac SELinux fixes
by spencer＠beyondabstraction.net 23 Apr '10

23 Apr '10

* Fix a logic error in configure.ac that prevented --with-selinux=no from being used with --with-secdriver-selinux=no. * Fix some strings to clarify the difference between --with-selinux and --with-secdriver-selinux. Signed-off-by: Spencer Shimko <sshimko(a)tresys.com> --- configure.ac | 12 ++++++------ 1 files changed, 6 insertions(+), 6 deletions(-) diff --git a/configure.ac b/configure.ac index 42c09b4..68dca54 100644 --- a/configure.ac +++ b/configure.ac @@ -869,14 +869,14 @@ if test "$with_selinux" != "no"; then AC_CHECK_HEADER([selinux/selinux.h],[],[fail=1]) AC_CHECK_LIB([selinux], [fgetfilecon],[],[fail=1]) test $fail = 1 && - AC_MSG_ERROR([You must install the SELinux development package in order to compile libvirt]) + AC_MSG_ERROR([You must install the libselinux development package in order to compile libvirt with basic SELinux support]) fi CFLAGS="$old_cflags" LIBS="$old_libs" fi if test "$with_selinux" = "yes"; then SELINUX_LIBS="-lselinux" - AC_DEFINE_UNQUOTED([HAVE_SELINUX], 1, [whether SELinux is available for security]) + AC_DEFINE_UNQUOTED([HAVE_SELINUX], 1, [whether basic SELinux functionality is available]) fi AM_CONDITIONAL([HAVE_SELINUX], [test "$with_selinux" != "no"]) AC_SUBST([SELINUX_CFLAGS]) @@ -891,9 +891,9 @@ AC_ARG_WITH([secdriver-selinux], if test "$with_selinux" != "yes" ; then if test "$with_secdriver_selinux" = "check" ; then with_secdriver_selinux=no - else - AC_MSG_ERROR([You must install the SELinux development package in order to compile libvirt]) - fi + elif test "$with_secdriver_selinux" = "yes"; then + AC_MSG_ERROR([You must install the libselinux development package and enable SELinux with the --with-selinux=yes in order to compile libvirt --with-secdriver-selinux=yes]) + fi else old_cflags="$CFLAGS" old_libs="$LIBS" @@ -910,7 +910,7 @@ else if test "$with_secdriver_selinux" = "check" ; then with_secdriver_selinux=no else - AC_MSG_ERROR([You must install the SELinux development package in order to compile libvirt]) + AC_MSG_ERROR([You must install libselinux development package >= 2.0.82 in order to compile libvirt --with-secdriver-selinux=yes]) fi else with_secdriver_selinux=yes -- 1.6.6.1

2 1

[libvirt] seg fault when running snapshot-create
by Stephen Shaw 23 Apr '10

23 Apr '10

I'm getting a seg fault when running virsh snapshot-create 1, but only when virt-manager is open and connected. Here is some of the debug info I was able to come up with - http://fpaste.org/9GO6/ (bt) http://fpaste.org/7gkH/ ('thread apply all bt) * After the crash (gdb) p mon->msg $1 = (qemuMonitorMessagePtr) 0x0 nibbler:~ # libvirtd --version libvirtd (libvirt) 0.8.0 Please let me know if there is any other information you need. Stephen

4 13

[libvirt] [PATCH] Fix handling of security driver restore failures in QEMU domain save
by Daniel P. Berrange 23 Apr '10

23 Apr '10

In cases where the security driver failed to restore a label after a guest has saved, we mistakenly jumped to the error cleanup paths. This is not good, because the operation has in fact completed and cannot be rolled back completely. Label restore is non-critical, so just log the problem instead. Also add a missing restore call in the error cleanup path * src/qemu/qemu_driver.c: Fix handling of security driver restore failures in QEMU domain save --- src/qemu/qemu_driver.c | 48 +++++++++++++++++++++++++----------------------- 1 files changed, 25 insertions(+), 23 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 3bea7e7..7b92965 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -5050,16 +5050,13 @@ static int qemudDomainSaveFlag(virDomainPtr dom, const char *path, driver->securityDriver && driver->securityDriver->domainRestoreSavedStateLabel && driver->securityDriver->domainRestoreSavedStateLabel(vm, path) == -1) - goto endjob; + VIR_WARN("failed to restore save state label on %s", path); if (cgroup != NULL) { rc = virCgroupDenyDevicePath(cgroup, path); - if (rc != 0) { - virReportSystemError(-rc, - _("Unable to deny device %s for %s"), - path, vm->def->name); - goto endjob; - } + if (rc != 0) + VIR_WARN("Unable to deny device %s for %s %d", + path, vm->def->name, rc); } ret = 0; @@ -5078,24 +5075,29 @@ static int qemudDomainSaveFlag(virDomainPtr dom, const char *path, endjob: if (vm) { - if (ret != 0 && header.was_running && priv->mon) { - qemuDomainObjEnterMonitorWithDriver(driver, vm); - rc = qemuMonitorStartCPUs(priv->mon, dom->conn); - qemuDomainObjExitMonitorWithDriver(driver, vm); - if (rc < 0) - VIR_WARN0("Unable to resume guest CPUs after save failure"); - else - vm->state = VIR_DOMAIN_RUNNING; - } + if (ret != 0) { + if (header.was_running && priv->mon) { + qemuDomainObjEnterMonitorWithDriver(driver, vm); + rc = qemuMonitorStartCPUs(priv->mon, dom->conn); + qemuDomainObjExitMonitorWithDriver(driver, vm); + if (rc < 0) + VIR_WARN0("Unable to resume guest CPUs after save failure"); + else + vm->state = VIR_DOMAIN_RUNNING; + } - if (ret != 0 && cgroup != NULL) { - rc = virCgroupDenyDevicePath(cgroup, path); - if (rc != 0) { - virReportSystemError(-rc, - _("Unable to deny device %s for %s"), - path, vm->def->name); - goto endjob; + if (cgroup != NULL) { + rc = virCgroupDenyDevicePath(cgroup, path); + if (rc != 0) + VIR_WARN("Unable to deny device %s for %s: %d", + path, vm->def->name, rc); } + + if ((!bypassSecurityDriver) && + driver->securityDriver && + driver->securityDriver->domainRestoreSavedStateLabel && + driver->securityDriver->domainRestoreSavedStateLabel(vm, path) == -1) + VIR_WARN("failed to restore save state label on %s", path); } if (qemuDomainObjEndJob(vm) == 0) -- 1.6.5.2

1 0

[libvirt] [PATCH] Fix printing of event detail in python events demo program
by Daniel P. Berrange 23 Apr '10

23 Apr '10

The events demo program is slightly misleading printing myDomainEventCallback1 EVENT: Domain f14i686(-1) Added which is not distinguishing Add vs Update events. It should have been doing myDomainEventCallback1 EVENT: Domain f14i686(-1) Defined Updated * examples/domain-events/events-python/event-test.py: Fully print event detail info string --- examples/domain-events/events-python/event-test.py | 27 ++++++++++++++----- 1 files changed, 20 insertions(+), 7 deletions(-) diff --git a/examples/domain-events/events-python/event-test.py b/examples/domain-events/events-python/event-test.py index 0c6e2f0..7c5af68 100644 --- a/examples/domain-events/events-python/event-test.py +++ b/examples/domain-events/events-python/event-test.py @@ -396,21 +396,34 @@ def virEventLoopPureStart(): # Everything that now follows is a simple demo of domain lifecycle events ########################################################################## def eventToString(event): - eventStrings = ( "Added", - "Removed", + eventStrings = ( "Defined", + "Undefined", "Started", "Suspended", "Resumed", - "Stopped", - "Saved", - "Restored" ); + "Stopped" ); return eventStrings[event]; +def detailToString(event, detail): + eventStrings = ( + ( "Added", "Updated" ), + ( "Removed" ), + ( "Booted", "Migrated", "Restored", "Snapshot" ), + ( "Paused", "Migrated", "IOError", "Watchdog" ), + ( "Unpaused", "Migrated"), + ( "Shutdown", "Destroyed", "Crashed", "Migrated", "Saved", "Failed", "Snapshot") + ) + return eventStrings[event][detail] + def myDomainEventCallback1 (conn, dom, event, detail, opaque): - print "myDomainEventCallback1 EVENT: Domain %s(%s) %s %d" % (dom.name(), dom.ID(), eventToString(event), detail) + print "myDomainEventCallback1 EVENT: Domain %s(%s) %s %s" % (dom.name(), dom.ID(), + eventToString(event), + detailToString(event, detail)) def myDomainEventCallback2 (conn, dom, event, detail, opaque): - print "myDomainEventCallback2 EVENT: Domain %s(%s) %s %d" % (dom.name(), dom.ID(), eventToString(event), detail) + print "myDomainEventCallback2 EVENT: Domain %s(%s) %s %s" % (dom.name(), dom.ID(), + eventToString(event), + detailToString(event, detail)) def myDomainEventRebootCallback(conn, dom, opaque): print "myDomainEventRebootCallback: Domain %s(%s)" % (dom.name(), dom.ID()) -- 1.6.6.1

2 1

[libvirt] [PATCH 1/2] Fix CPU hotplug command names
by Daniel P. Berrange 23 Apr '10

23 Apr '10

Probably as a result of a merge error, the CPU hotplug command names were completely wrong. * src/qemu/qemu_monitor_json.c, src/qemu/qemu_monitor_text.c: Fix the CPU hotplug command names --- src/qemu/qemu_monitor_json.c | 2 +- src/qemu/qemu_monitor_text.c | 4 ++-- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/src/qemu/qemu_monitor_json.c b/src/qemu/qemu_monitor_json.c index cde9899..96f246f 100644 --- a/src/qemu/qemu_monitor_json.c +++ b/src/qemu/qemu_monitor_json.c @@ -1181,7 +1181,7 @@ int qemuMonitorJSONSetCPU(qemuMonitorPtr mon, int cpu, int online) { int ret; - virJSONValuePtr cmd = qemuMonitorJSONMakeCommand("balloon", + virJSONValuePtr cmd = qemuMonitorJSONMakeCommand("cpu_set", "U:cpu", (unsigned long long)cpu, "s:state", online ? "online" : "offline", NULL); diff --git a/src/qemu/qemu_monitor_text.c b/src/qemu/qemu_monitor_text.c index 6ad07b1..db3806d 100644 --- a/src/qemu/qemu_monitor_text.c +++ b/src/qemu/qemu_monitor_text.c @@ -800,7 +800,7 @@ int qemuMonitorTextSetBalloon(qemuMonitorPtr mon, /* - * Returns: 0 if balloon not supported, +1 if balloon adjust worked + * Returns: 0 if CPU hotplug not supported, +1 if CPU hotplug worked * or -1 on failure */ int qemuMonitorTextSetCPU(qemuMonitorPtr mon, int cpu, int online) @@ -809,7 +809,7 @@ int qemuMonitorTextSetCPU(qemuMonitorPtr mon, int cpu, int online) char *reply = NULL; int ret = -1; - if (virAsprintf(&cmd, "set_cpu %d %s", cpu, online ? "online" : "offline") < 0) { + if (virAsprintf(&cmd, "cpu_set %d %s", cpu, online ? "online" : "offline") < 0) { virReportOOMError(); return -1; } -- 1.6.6.1

3 5

[libvirt] Plan for a 0.8.1 by end of next week
by Daniel Veillard 23 Apr '10

23 Apr '10

Checking git history since the release of 0.8.0, I'm seeing an impressive set of bug fixes and cleanups, and relatively few actual features changes. So I think it's sensible to try to make a new release 0.8.1 basically a 0.8.0 fixed version, and in a relatively short timeframe. To try to realign with our initial goal of releasing on Friday at the end of the month, I suggest we plan to make 0.8.1 by next Friday, the 30th April. Having a release with the snapshot feature fixed would IMHO suffice to justify this release, but git history right now makes a good case for such an early push. So unless we get serious objections, I suggest to continue the bug fixing trend, limit features commit next week and, have a cleaned up 0.8.1 available end of next week. Daniel -- Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ daniel(a)veillard.com | Rpmfind RPM search engine http://rpmfind.net/ http://veillard.com/ | virtualization library http://libvirt.org/

1 0

[libvirt] [PATCH] portability fixes to tools/virt-pki-validate.in
by Dustin Kirkland 23 Apr '10

23 Apr '10

portability fixes to tools/virt-pki-validate.in A few fixes will help make tools/virt-pki-validate.in useful on Debian and Ubuntu. And one fix should be useful to everyone (see #4). 1) note our gnutls-bin package (in addition to your gnutls-utils package) in the no-certtool error text 2) adjust the ORG-setting sed regular expressions to work with both Red Hat and Ubuntu style cacert.pem Issuer format 3) fix a bashism, == should be = in the case where /bin/sh is a symlink to dash 4) $(SYSCONFDIR) cannot evaluate; set a single shell SYSCONFDIR variable to the autoconf @SYSCONFDIR@ value, and use $SYSCONFDIR everywhere This was tested against gnutls 2.8.5. And against both Red Hat and Ubuntu cacert.pem files. Credit Jamie Strandboge with the regular expression changes. Bug report: * https://bugs.edge.launchpad.net/ubuntu/+source/libvirt/+bug/562266 Signed-off-by: Dustin Kirkland <kirkland(a)canonical.com> diff --git a/tools/virt-pki-validate.in b/tools/virt-pki-validate.in index ee7b79d..d335997 100755 --- a/tools/virt-pki-validate.in +++ b/tools/virt-pki-validate.in @@ -16,8 +16,8 @@ PORT=16514 CERTOOL=`which certtool 2>/dev/null` if [ ! -x $CERTOOL ] then - echo Could not locate the certtool program - echo make sure the gnutls-utils package is installed + echo "Could not locate the certtool program" + echo "make sure the gnutls-utils (or gnutls-bin) package is installed" exit 1 fi echo Found $CERTOOL @@ -25,7 +25,8 @@ echo Found $CERTOOL # # Check the directory structure # -PKI="@SYSCONFDIR@/pki" +SYSCONFDIR="@SYSCONFDIR@" +PKI="$SYSCONFDIR/pki" if [ ! -d $PKI ] then echo the $PKI directory is missing, it is usually @@ -129,8 +130,8 @@ then echo "as root do: chmod 644 $CA/cacert.pem" exit 1 fi -ORG=`$CERTOOL -i --infile $CA/cacert.pem | grep Issuer | sed 's+Issuer: CN=++'` -if [ "$ORG" == "" ] +ORG=`$CERTOOL -i --infile $CA/cacert.pem | grep Issuer | sed -e 's+\s*Issuer: .*CN=++' -e 's+,EMAIL=.*++'` +if [ "$ORG" = "" ] then echo the CA certificate $CA/cacert.pem does not define the organization echo it should probably regenerated @@ -240,19 +241,19 @@ fi if [ "$SERVER" = "1" ] then - if [ -r $(SYSCONFDIR)/sysconfig/libvirtd ] + if [ -r $SYSCONFDIR/sysconfig/libvirtd ] then - if [ "`grep '^LIBVIRTD_ARGS' $(SYSCONFDIR)/sysconfig/libvirtd | grep -- '--listen'`" = "" ] + if [ "`grep '^LIBVIRTD_ARGS' $SYSCONFDIR/sysconfig/libvirtd | grep -- '--listen'`" = "" ] then - echo Make sure $(SYSCONFDIR)/sysconfig/libvirtd is setup to listen to + echo Make sure $SYSCONFDIR/sysconfig/libvirtd is setup to listen to echo TCP/IP connections and restart the libvirtd service fi fi - if [ -r $(SYSCONFDIR)/sysconfig/iptables ] + if [ -r $SYSCONFDIR/sysconfig/iptables ] then - if [ "`grep $PORT $(SYSCONFDIR)/sysconfig/iptables`" = "" ] + if [ "`grep $PORT $SYSCONFDIR/sysconfig/iptables`" = "" ] then - echo Make sure $(SYSCONFDIR)/sysconfig/iptables is setup to allow + echo Make sure $SYSCONFDIR/sysconfig/iptables is setup to allow echo incoming TCP/IP connections on port $PORT and echo restart the iptables service fi

3 6

Re: [libvirt] Failed to insert a cd in a virtual machine running
by Jorge Alejandro Rodríguez 23 Apr '10

23 Apr '10

the problem is because of new version of qemu, the name of devices isn't ide0-0-, is drive-ide0-0-1. 2010/4/22 Jorge Alejandro Rodríguez <jrodriguez(a)dit.upm.es> > I can't attach a cd when a vm is running with qemu 0.12.3 (ubuntu 10.04), > the problem is not with qemu because when I start the vm with kvm command > (without libvirt) I can change the cd with qemu-monitor without any problem > and eject it too (libvirt say me that it isn't supported by hypervisor). > > the versions of libvirt that I have tried are 0.7.7 and 0.8.0. > > Any ideas? > > Thanks. > Jorge >

1 0

[libvirt] [Resend][PATCH 0/3] Add dnsmasq module
by Satoru SATOH 22 Apr '10

22 Apr '10

I posted this just before 0.8.0 is out and it was bad timing, I suppose. So try to resend. The following series of patches is a prototype implementation of dnsmasq module. It implements an idea originally suggested by Dan-san and can address the problem that too many --dhcp-host args hitting ARG_MAX limit I reported last year [1]. [1] https://www.redhat.com/archives/libvir-list/2009-October/msg00216.html Thanks, Satoru SATOH

3 6