- Devel - libvirt List Archives

[libvirt] [PATCH 2/14] Add build support
by Stefan Berger 18 Mar '10

18 Mar '10

This patch adds build support for the network filtering framework. Signed-off-by: Stefan Berger <stefanb(a)us.ibm.com>

2 2

[libvirt] [PATCH] util: ensure virMutexInit is not recursive
by Eric Blake 18 Mar '10

18 Mar '10

POSIX states that creation of a mutex with default attributes is unspecified whether the mutex is recursive or non-recursive. We specifically want non-recursive (deadlock is desirable in flushing out coding bugs that used our mutex incorrectly). * src/util/threads-pthread.c (virMutexInit): Specifically request non-recursive mutex, rather than relying on unspecified default. --- src/util/threads-pthread.c | 7 +++++-- 1 files changed, 5 insertions(+), 2 deletions(-) diff --git a/src/util/threads-pthread.c b/src/util/threads-pthread.c index ad42483..2f0746b 100644 --- a/src/util/threads-pthread.c +++ b/src/util/threads-pthread.c @@ -1,7 +1,7 @@ /* * threads-pthread.c: basic thread synchronization primitives * - * Copyright (C) 2009 Red Hat, Inc. + * Copyright (C) 2009-2010 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 @@ -36,7 +36,10 @@ void virThreadOnExit(void) int virMutexInit(virMutexPtr m) { int ret; - if ((ret = pthread_mutex_init(&m->lock, NULL)) != 0) { + pthread_mutexattr_t attr; + pthread_mutexattr_init(&attr); + pthread_mutexattr_settype(&attr, PTHREAD_MUTEX_NORMAL); + if ((ret = pthread_mutex_init(&m->lock, &attr)) != 0) { errno = ret; return -1; } -- 1.6.6.1

3 2

[libvirt] [PATCH] maint: make Red Hat copyright notices consistent
by Eric Blake 18 Mar '10

18 Mar '10

Spell out 'Red Hat, Inc.': git grep -i 'Copyright.*Red Hat' | grep -v Inc Include (C) consistently: git grep -i 'Copyright [^(].*Red Hat' * src/lxc/lxc_container.c: Update copyright formatting. * src/node_device/node_device_udev.c: Likewise. * src/node_device/node_device_udev.h: Likewise. * src/xen/xend_internal.h: Likewise. * src/xen/xm_internal.c: Likewise. * src/xen/xm_internal.h: Likewise. * tests/xmconfigtest.c: Likewise. * tests/object-locking.ml: Likewise. --- src/lxc/lxc_container.c | 4 ++-- src/node_device/node_device_udev.c | 2 +- src/node_device/node_device_udev.h | 2 +- src/xen/xend_internal.h | 2 +- src/xen/xm_internal.c | 2 +- src/xen/xm_internal.h | 2 +- tests/object-locking.ml | 2 +- tests/xmconfigtest.c | 2 +- 8 files changed, 9 insertions(+), 9 deletions(-) diff --git a/src/lxc/lxc_container.c b/src/lxc/lxc_container.c index 68b4656..706c796 100644 --- a/src/lxc/lxc_container.c +++ b/src/lxc/lxc_container.c @@ -1,6 +1,6 @@ /* - * Copyright IBM Corp. 2008 - * Copyright Red Hat 2008-2009 + * Copyright (C) 2008-2010 Red Hat, Inc. + * Copyright (C) 2008 IBM Corp. * * lxc_container.c: file description * diff --git a/src/node_device/node_device_udev.c b/src/node_device/node_device_udev.c index 11e27e0..e3114fa 100644 --- a/src/node_device/node_device_udev.c +++ b/src/node_device/node_device_udev.c @@ -1,7 +1,7 @@ /* * node_device_udev.c: node device enumeration - libudev implementation * - * Copyright (C) 2009-2010 Red Hat + * Copyright (C) 2009-2010 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 diff --git a/src/node_device/node_device_udev.h b/src/node_device/node_device_udev.h index 8367494..cdaa142 100644 --- a/src/node_device/node_device_udev.h +++ b/src/node_device/node_device_udev.h @@ -1,7 +1,7 @@ /* * node_device_udev.h: node device enumeration - libudev implementation * - * Copyright (C) 2009 Red Hat + * Copyright (C) 2009-2010 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 diff --git a/src/xen/xend_internal.h b/src/xen/xend_internal.h index 483253f..028a293 100644 --- a/src/xen/xend_internal.h +++ b/src/xen/xend_internal.h @@ -1,12 +1,12 @@ /* * xend_internal.h * + * Copyright (C) 2006-2008, 2010 Red Hat, Inc. * Copyright (C) 2005,2006 * * Anthony Liguori <aliguori(a)us.ibm.com> * Daniel Veillard <veillard(a)redhat.com> * - * Copyright 2006-2008 Red Hat * * This file is subject to the terms and conditions of the GNU Lesser General * Public License. See the file COPYING in the main directory of this archive diff --git a/src/xen/xm_internal.c b/src/xen/xm_internal.c index 3d4624d..74bf0b6 100644 --- a/src/xen/xm_internal.c +++ b/src/xen/xm_internal.c @@ -1,7 +1,7 @@ /* * xm_internal.h: helper routines for dealing with inactive domains * - * Copyright (C) 2006-2007, 2009, 2010 Red Hat + * Copyright (C) 2006-2007, 2009-2010 Red Hat, Inc. * Copyright (C) 2006 Daniel P. Berrange * * This library is free software; you can redistribute it and/or diff --git a/src/xen/xm_internal.h b/src/xen/xm_internal.h index 37132ef..3ad3456 100644 --- a/src/xen/xm_internal.h +++ b/src/xen/xm_internal.h @@ -1,7 +1,7 @@ /* * xm_internal.h: helper routines for dealing with inactive domains * - * Copyright (C) 2006-2007 Red Hat + * Copyright (C) 2006-2007, 2010 Red Hat, Inc. * Copyright (C) 2006 Daniel P. Berrange * * This library is free software; you can redistribute it and/or diff --git a/tests/object-locking.ml b/tests/object-locking.ml index a4cd77d..1fddd51 100644 --- a/tests/object-locking.ml +++ b/tests/object-locking.ml @@ -1,7 +1,7 @@ (* * Analyse libvirt driver API methods for mutex locking mistakes * - * Copyright 2008-2009 Red Hat, Inc + * Copyright (C) 2008-2010 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 diff --git a/tests/xmconfigtest.c b/tests/xmconfigtest.c index acb6d18..a199c72 100644 --- a/tests/xmconfigtest.c +++ b/tests/xmconfigtest.c @@ -1,7 +1,7 @@ /* * xmconfigtest.c: Test backend for xm_internal config file handling * - * Copyright (C) 2007 Red Hat + * Copyright (C) 2007, 2010 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 -- 1.6.6.1

2 7

Re: [libvirt] virtio serial config
by Matthew Booth 18 Mar '10

18 Mar '10

On 18/03/10 14:31, Matthew Whitehead wrote: > Matthew, > Dan B referred me to you. I want to create 4 (or more) virtual serial devices between two KVM virtual hosts. The output of one serial device would be the input to the other. Virt-manager doesn't provide this, so I'm prepared to hack the config files manually. Unfortunately, I can't find documentation on the syntax. > > Can you assist? I'm normally the one complaining about lack of documentation :) That's quite an oversight. The basic syntax for a virtio-serial channel is: <channel type='pty'> <target type='virtio' name='org.linux-kvm.port.1'/> </channel> You'll want 4 <channel/> entries per guest. Give them all different names. The interesting bit is going to be plugging the character devices together on the host. I'm pretty sure you'll have to use pipes. Note that afaik the libvirt documentation on this is wrong and misleading. Don't read it, just test instead :) IIRC, the way pipes actually work is that it expects to be able to use 2 pipes based on the path you specify, called foo.in and foo.out. I seem to recall you need to create these manually too. The following advice is completely untested. It may also be responsible for global warming and the rise of reality TV. Host: # mkfifo /tmp/guesta-virtioserial{1,2}.{in,out} # ln /tmp/guesta-virtioserial1.in /tmp/guestb-virtioserial1.out # ln /tmp/guesta-virtioserial1.out /tmp/guestb-virtioserial1.in # ln /tmp/guesta-virtioserial2.in /tmp/guestb-virtioserial2.out # ln /tmp/guesta-virtioserial2.out /tmp/guestb-virtioserial2.in Guest A: <channel type='pipe'> <source path="/tmp/guesta-virtioserial1"/> <target type='virtio' name='org.linux-kvm.port.1'/> </channel> <channel type='pipe'> <source path="/tmp/guesta-virtioserial2"/> <target type='virtio' name='org.linux-kvm.port.2'/> </channel> Guest B: <channel type='pipe'> <source path="/tmp/guestb-virtioserial1"/> <target type='virtio' name='org.linux-kvm.port.1'/> </channel> <channel type='pipe'> <source path="/tmp/guestb-virtioserial2"/> <target type='virtio' name='org.linux-kvm.port.2'/> </channel> At this point, both virtio-serial devices in each guest should be full-duplex. Matt -- Matthew Booth, RHCA, RHCSS Red Hat Engineering, Virtualisation Team M: +44 (0)7977 267231 GPG ID: D33C3490 GPG FPR: 3733 612D 2D05 5458 8A8A 1600 3441 EA19 D33C 3490

1 0

[libvirt] [APPENDIX] Sample filters
by Stefan Berger 18 Mar '10

18 Mar '10

Attached are some sample filter templates. Some of these should probably become available through libvirt repository later on as 'standard filters'. One may copy the XML files into /etc/libvirt/nwfilter or use 'virsh nwfilter-define <filename>' to make them known to libvirt. Using 'virsh nwfilter-list' one can then list the filters. The simpleloop.xml filter will not be accepted since it would directly introduce a loop. More complex loops are also detected. Regards, Stefan, Gerhard

1 0

[libvirt] [PATCH 13/14] Add IPv6 support for ebtables layer
by Stefan Berger 18 Mar '10

18 Mar '10

This patch adds IPv6 support for the ebtables layer. Since the parser etc. are all parameterized, it was fairly easy to add this... Signed-off-by: Stefan Berger <stefanb(a)us.ibm.com>

1 0

[libvirt] [PATCH 12/14] Core driver implementation with ebtables support
by Stefan Berger 18 Mar '10

18 Mar '10

This patch implements the core driver and provides - management functionality for managing the filter XMLs - compiling the internal filter representation into ebtables rules - applying ebtables rules on a network (tap,macvtap) interface - tearing down ebtables rules that were applied on behalf of an interface - updating of filters while VMs are running and causing the firewalls to be rebuilt - other bits and pieces Signed-off-by: Stefan Berger <stefanb(a)us.ibm.com>

1 0

[libvirt] [PATCH 11/14] Add qemu support
by Stefan Berger 18 Mar '10

18 Mar '10

Add support for Qemu to have firewall rules applied and removed on VM startup and shutdown respectively. This patch also provides support for the updating of a filter that causes all VMs that reference the filter to have their ebtables/iptables rules updated. Signed-off-by: Stefan Berger <stefanb(a)us.ibm.com>

1 0

[libvirt] [PATCH 10/14] Add XML parser extensions to the domain XML processing
by Stefan Berger 18 Mar '10

18 Mar '10

This patch extends the domain XML processing to parse the top level referenced filter along with potentially provided parameters and also converts the internal data back into XML representation. Signed-off-by: Stefan Berger <stefanb(a)us.ibm.com> Signed-off-by: Gerhard Stenzel <gerhard.stenzel(a)de.ibm.com>

1 0

[libvirt] [PATCH 9/14] Add virsh support for new CLI commands
by Stefan Berger 18 Mar '10

18 Mar '10

This patch adds virsh support for the five new CLI commands to manage network filters. Signed-off-by: Stefan Berger <stefanb(a)us.ibm.com>

1 0

[libvirt] [PATCH 8/14] Implementation of the RPC call dispatch
by Stefan Berger 18 Mar '10

18 Mar '10

This patch extends the RPC dispatcher to support the newly added RPC calls for network filtering (ACL) support. Signed-off-by: Stefan Berger <stefanb(a)us.ibm.com>

1 0

[libvirt] [PATCH 7/14] Implementation of the RPC client
by Stefan Berger 18 Mar '10

18 Mar '10

This patch extends the RPC client for the new network filtering (ACL) functionality. Signed-off-by: Stefan Berger <stefanb(a)us.ibm.com>

1 0

[libvirt] [PATCH 6/14] Definition of the wire format
by Stefan Berger 18 Mar '10

18 Mar '10

This patch adds the definition of the wire format for RPC calls for the new network filtering (ACL) functionality added to libvirt. Signed-off-by: Stefan Berger <stefanb(a)us.ibm.com>

1 0

[libvirt] [PATCH 4/14] Add internal API
by Stefan Berger 18 Mar '10

18 Mar '10

This patch adds the internal API extensions for network filtering (ACL) support. Signed-off-by: Stefan Berger <stefanb(a)us.ibm.com>

1 0

[libvirt] [PATCH 3/14] Add public API
by Stefan Berger 18 Mar '10

18 Mar '10

This patch adds extensions to libvirt's public API necessary for controlling the new functionality from remote for example. Signed-off-by: Stefan Berger <stefanb(a)us.ibm.com>

1 0

[libvirt] [PATCH 0/14 [RFC] Network filtering (ACL) extensions for libvirt
by Stefan Berger 18 Mar '10

18 Mar '10

Hi! This is a repost of this set of patches with some of the fixes recommended by Daniel Berrange applied and ipv6 support on the ebtables layer added. The following set of patches add network filtering (ACL) extensions to libvirt and enable network traffic filtering for VMs using ebtables and, depending on the networking technology being used (tap, but not macvtap), also iptables. Usage of either is optional and controlled through filters that a VM is referencing. The ebtables-level filtering is based on the XML derived from the CIM network slide 10 (filtering) from the DMTF website (http://www.dmtf.org/standards/cim/cim_schema_v2230/CIM_Network.pdf) The XML we derived from this was discussed on the list before. On the ebtables level we currently handle filtering of IPv4 and ARP traffic. The iptables-level filtering is based on similar XML where XML nodes described the particular protocol to filter for. Its extensions enable the filtering of traffic using iptables for tcp, udp, icmp, igmp, sctp and 'all' types of traffic. This list of protocols maps to the features supported by iptables and only excludes protocols like 'esp', 'ah' and 'udplite'. Currently only bridging mode is supported and based on availability of the physdev match. The filtering framework adds new libvirt virsh commands for managing the filters. The 5 new commands are: - virsh nwfilter-list - virsh nwfilter-dumpxml <name of filter> - virsh nwfilter-define <name of file containing filter desc.> - virsh nwfilter-undefine <name of filter> - virsh nwfilter-edit <name of filter> Above commands are similar to commands for already existing pools and as such much of the code directly related to the above commands could be borrowed from other drivers. The network filters can either contain rules using the above mentioned XML or contain references to other filters in order to build more complex filters that form some sort of filter tree or can contain both. An example for a filter referencing other filters would be this one here: <filter name='demofilter4' chain='root'> <uuid>66f62d1d-34c1-1421-824f-c62d5ee5e8b6</uuid> <filterref filter='no-mac-spoofing'/> <filterref filter='no-mac-broadcast'/> <filterref filter='no-arp-spoofing'/> <filterref filter='allow-dhcp'> <parameter name='DHCPSERVER' value='10.0.0.1'/> </filterref> <filterref filter='no-other-l2-traffic'/> <filterref filter='recv-only-vm-ipaddress'/> <filterref filter='recv-only-vm-macaddress'/> <filterref filter='l3-test'/> <filterref filter='ipv6test'/> </filter> A filter containing actual rules would look like this: <filter name='no-mac-broadcast' chain='ipv4'> <uuid>ffe2ccd6-edec-7360-1852-6b5ccb553234</uuid> <rule action='drop' direction='out' priority='500'> <mac dstmacaddr='ff:ff:ff:ff:ff:ff'/> </rule> </filter> The filter XML now also holds a priority attribute in the rule. This provides control over the ordering of the applied ebtables/iptables rules beyond their appearance in the XML. The domain XML has been extended to reference a top level filter from within each <interface> XML node. A valid reference to such a top level filter looks like this: <interface type='bridge'> <source bridge='static'/> <filterref filter='demofilter4'> <parameter name='IP' value='9.59.241.151'/> </filterref> </interface> In this XML a parameter IP is passed for instantiation of the referenced filters, that may require the availability of this parameter. In the above case the IP parameter's value describes the value of the IP address of the VM and allows to enable those filters to be instantiated that require this 'IP' variable. If a filter requires a parameter that is not provided, the VM will not start or the interface will not attach to a running VM. Any names of parameters can be provided for instantiation of filters and their names and values only need to pass a regular expression test. In a subsequent patch we will be adding capability to allow users to omit the IP parameter (only) and enable libvirt to learn the IP address of the VM and have it instantiate the filter once it knows it. While virtual machines are running, it is possible to update their filters. For that all running VMs' filter 'trees' are traversed to detect whether the updated filter is referenced by the VM. If so, its ebtables/iptable rules are applied. If one of the VMs' update fails allupdates are rolled back and the filter XML update is rejected. One comment about the instantiation of the rules: Since the XML allows to create nearly any possible combination of parameters to ebtables or iptables commands, I haven't used the ebtables or iptables wrappers. Instead, I am writing ebtables/iptables command into a buffer, add command line options to each one of them as described in the rule's XML, write the buffer into a file and run it as a script. For those commands that are not allowed to fail I am using the following format to run them: cmd="ebtables <some options>" r=`${cmd}` if [ $? -ne 0 ]; then echo "Failure in command ${cmd}." exit 1 fi cmd="..." [...] If one of the command fails in such a batch, the libvirt code is going pick up the error code '1', tear down anything previously established and report an error back. The actual error message shown above is currently not reported back, but can be later on with some changes to the commands running external programs that need to read the script's stdout. One comment to patch 14: It currently #include's a .c file into a .c file only for the reason so I don't have to change too much code once I change code in the underlying patch. So this has to be changed. The patch series works without patch 13, but then only supports ebtables. The patches apply to the current tip. They pass 'make syntax-check' and have been frequently run in valgrind for memory leak checks. The order in which I apply the patches is as follows: add_recursive_locks.diff add_build_support.diff add_public_api.diff add_internal_api.diff impl_pub_api.diff def_wire_protocol_format.diff impl_rpc_client.c impl_srv_dispatch.diff add_virsh_support.diff add_xml_parsing.diff add_qemu_support.diff impl_driver.diff add_ipv6_support.diff add_iptables_support.diff Looking forward to your feedback on the patches. Thanks and regards, Stefan and Gerhard

1 0

[libvirt] [PATCH] website: Increase text size
by Cole Robinson 18 Mar '10

18 Mar '10

Personally I find the text so small it is difficult to read, especially in the documentation pages where we can have a large wall of text. Here is a before and after shot of the main page on my machine (scaled down): http://fedorapeople.org/~crobinso/tmp/libvirt-web-before-after.png Text size is now similar to linux-kvm.org, which I find much easier to read. Signed-off-by: Cole Robinson <crobinso(a)redhat.com> --- docs/generic.css | 10 +++++----- docs/libvirt.css | 1 + 2 files changed, 6 insertions(+), 5 deletions(-) diff --git a/docs/generic.css b/docs/generic.css index dbf7b56..d9cf49e 100644 --- a/docs/generic.css +++ b/docs/generic.css @@ -43,22 +43,22 @@ div.footer { } h1 { - font-size: 2em; + font-size: 1.6em; } h2 { - font-size: 1.6em; + font-size: 1.4em; } h3 { - font-size: 1.4em; + font-size: 1.2em; } h4 { - font-size: 1.2em; + font-size: 1.1em; } h5 { font-size: 1em; } h6 { - font-size: 0.8em; + font-size: 0.9em; } dl dt { diff --git a/docs/libvirt.css b/docs/libvirt.css index dfc93c6..46e9a3f 100644 --- a/docs/libvirt.css +++ b/docs/libvirt.css @@ -33,6 +33,7 @@ h2, h3, h4, h5, h6 { margin-right: 1em; padding: 0px; padding-bottom: 1em; + font-size: larger; } #menu { -- 1.6.6.1

3 3

[libvirt] [PATCH 0/5] Introduce virDomainMigrateSetDowntime API
by Jiri Denemark 18 Mar '10

18 Mar '10

This API call is supposed to change maximum tolerable downtime for live migrations. Jiri Denemark (5): Internal driver API for virDomainMigrateSetDowntime Public virDomainMigrateSetDowntime API Implement virDomainMigrateSetDowntime in remote driver Wire protocol and dispatcher for virDomainMigrateSetDowntime Implement virDomainMigrateSetDowntime in qemu driver daemon/remote.c | 29 ++++++++++ daemon/remote_dispatch_args.h | 1 + daemon/remote_dispatch_prototypes.h | 8 +++ daemon/remote_dispatch_table.h | 5 ++ include/libvirt/libvirt.h.in | 3 + src/driver.h | 5 ++ src/esx/esx_driver.c | 1 + src/libvirt.c | 46 ++++++++++++++++ src/libvirt_public.syms | 5 ++ src/lxc/lxc_driver.c | 1 + src/opennebula/one_driver.c | 1 + src/openvz/openvz_driver.c | 1 + src/phyp/phyp_driver.c | 1 + src/qemu/qemu_driver.c | 47 +++++++++++++++++ src/qemu/qemu_monitor.c | 15 +++++ src/qemu/qemu_monitor.h | 3 + src/qemu/qemu_monitor_json.c | 29 ++++++++++ src/qemu/qemu_monitor_json.h | 3 + src/qemu/qemu_monitor_text.c | 27 ++++++++++ src/qemu/qemu_monitor_text.h | 3 + src/remote/remote_driver.c | 27 ++++++++++ src/remote/remote_protocol.c | 11 ++++ src/remote/remote_protocol.h | 97 +++++++++++++++++++---------------- src/remote/remote_protocol.x | 9 +++- src/test/test_driver.c | 1 + src/uml/uml_driver.c | 1 + src/vbox/vbox_tmpl.c | 1 + src/xen/xen_driver.c | 1 + src/xenapi/xenapi_driver.c | 1 + 29 files changed, 338 insertions(+), 45 deletions(-)

4 18

[libvirt] [PATCH] maint: enforce recent N_ usage
by Eric Blake 18 Mar '10

18 Mar '10

* cfg.mk (sc_prohibit_gettext_noop): New rule. --- cfg.mk | 5 +++++ 1 files changed, 5 insertions(+), 0 deletions(-) diff --git a/cfg.mk b/cfg.mk index 3fd9f7b..5b4d6ed 100644 --- a/cfg.mk +++ b/cfg.mk @@ -105,6 +105,11 @@ sc_prohibit_gethostname: msg='use virGetHostname, not gethostname' \ $(_prohibit_regexp) +sc_prohibit_gettext_noop: + @re='gettext_noop *\(' \ + msg='use _N, not gettext_noop' \ + $(_prohibit_regexp) + sc_prohibit_VIR_ERR_NO_MEMORY: @re='\<V''IR_ERR_NO_MEMORY\>' \ msg='use virReportOOMError, not V'IR_ERR_NO_MEMORY \ -- 1.6.6.1

2 2

[libvirt] View graphical desktop of linux distribution through libvirt.
by Kumar L Srikanth-B22348 18 Mar '10

18 Mar '10

Hi Daniel, Is it possible to view the graphical desktop(startx) of any linux distribution in libvirt through Linux Container(LXC)? If it is possible, what are the necessary dependents we need to take care in the Domain XML or in the minimal Root file system of the linux distribution. Can you please let me know. Regards, Srikanth.

1 0

[libvirt] [PATCH] website: Drop static FAQ, point to http://wiki.libvirt.org/page/FAQ
by Cole Robinson 18 Mar '10

18 Mar '10

The static FAQ was from the days before even QEMU support. I added a few questions to the wiki FAQ about the software license and how to download and install (basically just pointing to downloads.html). The remaining questions on the static page aren't anything that I think is really 'frequently asked' (changing socket perms for regular user xen access, and issues building against libvirt). Signed-off-by: Cole Robinson <crobinso(a)redhat.com> --- docs/FAQ.html.in | 144 -------------------------------------------------- docs/Makefile.am | 3 +- docs/search.php | 2 +- docs/sitemap.html.in | 2 +- 4 files changed, 3 insertions(+), 148 deletions(-) delete mode 100644 docs/FAQ.html.in diff --git a/docs/FAQ.html.in b/docs/FAQ.html.in deleted file mode 100644 index 50f798d..0000000 --- a/docs/FAQ.html.in +++ /dev/null @@ -1,144 +0,0 @@ -<?xml version="1.0"?> -<html> - <body> - <h1 >FAQ</h1> - Table of Contents: - <ul> - <li> - <a href="FAQ.html#License">License(s)</a> - </li> - <li> - <a href="FAQ.html#Installati">Installation</a> - </li> - <li> - <a href="FAQ.html#Compilatio">Compilation</a> - </li> - <li> - <a href="FAQ.html#Developer">Developer corner</a> - </li> - </ul> - <h3><a name="License" id="License">License</a>(s)</h3> - <ol> - <li> - Licensing Terms for libvirt - libvirt is released under the <a href="http://www.opensource.org/licenses/lgpl-license.html">GNU Lesser - General Public License</a>, see the file COPYING.LIB in the distribution - for the precise wording. The only library that libvirt depends upon is - the Xen store access library which is also licenced under the LGPL. - </li> - <li> - Can I embed libvirt in a proprietary application ? - Yes. The LGPL allows you to embed libvirt into a proprietary - application. It would be graceful to send-back bug fixes and improvements - as patches for possible incorporation in the main development tree. It - will decrease your maintenance costs anyway if you do so. - </li> - </ol> - <h3> - <a name="Installati" id="Installati">Installation</a> - </h3> - <ol> - <li>Where can I get libvirt ? - The original distribution comes from <a href="ftp://libvirt.org/libvirt/">ftp://libvirt.org/libvirt/</a>. - </li> - <li> - I can't install the libvirt/libvirt-devel RPM packages due to - failed dependencies - The most generic solution is to re-fetch the latest src.rpm , and - rebuild it locally with - <code>rpm --rebuild libvirt-xxx.src.rpm</code>. - If everything goes well it will generate two binary rpm packages (one - providing the shared libs and virsh, and the other one, the -devel - package, providing includes, static libraries and scripts needed to build - applications with libvirt that you can install locally. - One can also rebuild the RPMs from a tarball: - - <code>rpmbuild -ta libdir-xxx.tar.gz</code> - - Or from a configured tree with: - - <code>make rpm</code> - - </li> - <li> - Failure to use the API for non-root users - Large parts of the API may only be accessible with root privileges, - however the read only access to the xenstore data doesnot have to be - forbidden to user, at least for monitoring purposes. If "virsh dominfo" - fails to run as an user, change the mode of the xenstore read-only socket - with: - - <code>chmod 666 /var/run/xenstored/socket_ro</code> - - and also make sure that the Xen Daemon is running correctly with local - HTTP server enabled, this is defined in - <code>/etc/xen/xend-config.sxp</code> which need the following line to be - enabled: - - <code>(xend-http-server yes)</code> - - If needed restart the xend daemon after making the change with the - following command run as root: - - <code>service xend restart</code> - - </li> - </ol> - <h3> - <a name="Compilatio" id="Compilatio">Compilation</a> - </h3> - <ol> - <li> - What is the process to compile libvirt ? - As most UNIX libraries libvirt follows the "standard": - - <code>gunzip -c libvirt-xxx.tar.gz | tar xvf -</code> - - - <code>cd libvirt-xxxx</code> - - - <code>./configure --help</code> - - to see the options, then the compilation/installation proper - - <code>./configure [possible options]</code> - - - <code>make</code> - - - <code>make install</code> - - At that point you may have to rerun ldconfig or a similar utility to - update your list of installed shared libs. - </li> - <li> - What other libraries are needed to compile/install libvirt ? - Libvirt requires libxenstore, which is usually provided by the xen - packages as well as the public headers to compile against libxenstore. - </li> - <li> - I use the GIT version and there is no configure script - The configure script (and other Makefiles) are generated. Use the - autogen.sh script to regenerate the configure script and Makefiles, - like: - - <code>./autogen.sh --prefix=/usr --disable-shared</code> - - </li> - </ol> - <h3><a name="Developer" id="Developer">Developer</a> corner</h3> - <ol> - <li> - Troubles compiling or linking programs using libvirt - To simplify the process of reusing the library, libvirt comes with - pkgconfig support, which can be used directly from autoconf support or - via the pkg-config command line tool, like: - - <code>pkg-config libvirt --libs</code> - - </li> - </ol> - </body> -</html> diff --git a/docs/Makefile.am b/docs/Makefile.am index edf6fc8..57f3d3d 100644 --- a/docs/Makefile.am +++ b/docs/Makefile.am @@ -147,8 +147,7 @@ rebuild: api all install-data-local: $(mkinstalldirs) $(DESTDIR)$(HTML_DIR) - -$(INSTALL) -m 0644 $(srcdir)/FAQ.html \ - $(srcdir)/Libxml2-Logo-90x34.gif $(DESTDIR)$(HTML_DIR) + -$(INSTALL) -m 0644 $(srcdir)/Libxml2-Logo-90x34.gif $(DESTDIR)$(HTML_DIR) $(mkinstalldirs) $(DESTDIR)$(HTML_DIR)/html for h in $(apihtml); do \ $(INSTALL) -m 0644 $(srcdir)/$$h $(DESTDIR)$(HTML_DIR)/html; done diff --git a/docs/search.php b/docs/search.php index a6c1def..bbd652a 100644 --- a/docs/search.php +++ b/docs/search.php @@ -258,7 +258,7 @@ </li><li> <a title="User contributed content" class="inactive" href="http://wiki.libvirt.org">Wiki</a> </li><li> - <a title="Frequently asked questions" class="inactive" href="FAQ.html">FAQ</a> + <a title="Frequently asked questions" class="inactive" href="http://wiki.libvirt.org/page/FAQ">FAQ</a> </li><li> <a title="How and where to report bugs and request features" class="inactive" href="bugs.html">Bug reports</a> </li><li> diff --git a/docs/sitemap.html.in b/docs/sitemap.html.in index 76d8faa..aaa0966 100644 --- a/docs/sitemap.html.in +++ b/docs/sitemap.html.in @@ -247,7 +247,7 @@ User contributed content </li> <li> - <a href="FAQ.html">FAQ</a> + <a href="http://wiki.libvirt.org/page/FAQ">FAQ</a> Frequently asked questions </li> <li> -- 1.6.6.1

2 1

[libvirt] [PATCH] website: Remove old repos from download section
by Cole Robinson 18 Mar '10

18 Mar '10

We haven't been using CVS for quite a while now, so I think we can safely drop the reference to the old server and git mirror. Signed-off-by: Cole Robinson <crobinso(a)redhat.com> --- docs/downloads.html.in | 33 --------------------------------- 1 files changed, 0 insertions(+), 33 deletions(-) diff --git a/docs/downloads.html.in b/docs/downloads.html.in index 493923c..2bfb459 100644 --- a/docs/downloads.html.in +++ b/docs/downloads.html.in @@ -43,18 +43,6 @@ <a href="http://libvirt.org/git/?p=libvirt.git;a=summary">http://libvirt.org/git/?p=libvirt.git;a=summary</a> </pre> - <h2>CVS repository access (Deprecated) </h2> - - The master source repository used to be under <a href="http://ximbiot.com/cvs/cvshome/docs/">CVS</a> - with anonymous access at: - - <pre> - - # cvs -d :pserver:anoncvs@libvirt.org:2401/data/cvs co libvirt - </pre> - The server is maintainened temporarilly for existing uses, but all - changes are only commited to git now and we expect to deprecate the CVS - server during summer 2009. <h2>Building from a source code checkout</h2> The libvirt build process uses GNU autotools, so after obtaining a @@ -69,26 +57,5 @@ make install </pre> - <h2>GIT repository mirror</h2> - - - Jim Mereying was maintaining a CVS to git mirror on - <a href="http://git.et.redhat.com/?p=libvirt.git">git.et.redhat.com</a>. - Existing users should migrate to the new libvirt.org git server, as the - old one is now deprecated. For the sake of old links including now- - rewritten SHA1s, we'll leave the old repository on-line for some time. - It is available as: - - <pre> - - git clone git://git.et.redhat.com/libvirt.git - </pre> - - It can also be browsed at - - <pre> - <a href="http://git.et.redhat.com/?p=libvirt.git;a=summary">http://git.et.redhat.com/?p=libvirt.git;a=summary</a> - </pre> - </body> </html> -- 1.6.6.1

2 1

[libvirt] [PATCH] Change logrotate to be per-hypervisor logs
by Daniel Veillard 18 Mar '10

18 Mar '10

Original bug is: https://bugzilla.redhat.com/show_bug.cgi?id=547514 maybe this could be done in slightly different way, possibly more generic, but I think doing a simple split is good enough for now. Change logrotate to be per-hypervisor logs Having a single logrotate configuration file for all hypervisors did not work as logrotate would get confused if an hypervisor not supported on that platform was still listed. Simplest is to split the logrotate as separate per hypervisor files and change the spec file to only install the ones compiled in. * daemon/libvirtd.lxc.logrotate.in daemon/libvirtd.qemu.logrotate.in daemon/libvirtd.uml.logrotate.in: copy and split the original daemon/libvirtd.logrotate.in file * daemon/Makefile.am: update to support the different files * libvirt.spec.in: only install the relevant logrotate configs diff --git a/daemon/Makefile.am b/daemon/Makefile.am index 958a4f6..d429c71 100644 --- a/daemon/Makefile.am +++ b/daemon/Makefile.am @@ -25,7 +25,9 @@ EXTRA_DIST = \ libvirtd.sasl \ libvirtd.sysconf \ libvirtd.aug \ - libvirtd.logrotate.in \ + libvirtd.qemu.logrotate.in \ + libvirtd.lxc.logrotate.in \ + libvirtd.uml.logrotate.in \ test_libvirtd.aug \ $(AVAHI_SOURCES) \ $(DAEMON_SOURCES) @@ -177,20 +179,37 @@ remote_dispatch_args.h: $(srcdir)/remote_generate_stubs.pl $(REMOTE_PROTOCOL) remote_dispatch_ret.h: $(srcdir)/remote_generate_stubs.pl $(REMOTE_PROTOCOL) perl -w $(srcdir)/remote_generate_stubs.pl -r $(REMOTE_PROTOCOL) > $@ -BUILT_SOURCES += libvirtd.logrotate +LOGROTATE_CONFS = libvirtd.qemu.logrotate libvirtd.lxc.logrotate \ + libvirtd.uml.logrotate -libvirtd.logrotate: libvirtd.logrotate.in +BUILT_SOURCES += $(LOGROTATE_CONFS) + +libvirtd.qemu.logrotate: libvirtd.qemu.logrotate.in + sed \ + -e s!\@localstatedir\@!@localstatedir@!g \ + < $< > $@-t + mv $@-t $@ + +libvirtd.lxc.logrotate: libvirtd.lxc.logrotate.in + sed \ + -e s!\@localstatedir\@!@localstatedir@!g \ + < $< > $@-t + mv $@-t $@ + +libvirtd.uml.logrotate: libvirtd.uml.logrotate.in sed \ -e s!\@localstatedir\@!@localstatedir@!g \ < $< > $@-t mv $@-t $@ -install-logrotate: libvirtd.logrotate +install-logrotate: $(LOGROTATE_CONFS) mkdir -p $(DESTDIR)$(localstatedir)/log/libvirt/qemu/ mkdir -p $(DESTDIR)$(localstatedir)/log/libvirt/lxc/ mkdir -p $(DESTDIR)$(localstatedir)/log/libvirt/uml/ mkdir -p $(DESTDIR)$(sysconfdir)/logrotate.d/ - $(INSTALL_DATA) $< $(DESTDIR)$(sysconfdir)/logrotate.d/libvirtd + $(INSTALL_DATA) libvirtd.qemu.logrotate $(DESTDIR)$(sysconfdir)/logrotate.d/libvirtd.qemu + $(INSTALL_DATA) libvirtd.lxc.logrotate $(DESTDIR)$(sysconfdir)/logrotate.d/libvirtd.lxc + $(INSTALL_DATA) libvirtd.uml.logrotate $(DESTDIR)$(sysconfdir)/logrotate.d/libvirtd.uml if LIBVIRT_INIT_SCRIPT_RED_HAT install-init: libvirtd.init diff --git a/daemon/libvirtd.logrotate.in b/daemon/libvirtd.logrotate.in deleted file mode 100644 index 0c51fd3..0000000 --- a/daemon/libvirtd.logrotate.in +++ /dev/null @@ -1,9 +0,0 @@ -@localstatedir@/log/libvirt/qemu/*.log @localstatedir@/log/libvirt/uml/*.log @localstatedir@/log/libvirt/lxc/*.log { - weekly - missingok - rotate 4 - compress - delaycompress - copytruncate - minsize 100k -} diff --git a/daemon/libvirtd.lxc.logrotate.in b/daemon/libvirtd.lxc.logrotate.in new file mode 100644 index 0000000..4e7d314 --- /dev/null +++ b/daemon/libvirtd.lxc.logrotate.in @@ -0,0 +1,9 @@ +@localstatedir@/log/libvirt/lxc/*.log { + weekly + missingok + rotate 4 + compress + delaycompress + copytruncate + minsize 100k +} diff --git a/daemon/libvirtd.qemu.logrotate.in b/daemon/libvirtd.qemu.logrotate.in new file mode 100644 index 0000000..15cf019 --- /dev/null +++ b/daemon/libvirtd.qemu.logrotate.in @@ -0,0 +1,9 @@ +@localstatedir@/log/libvirt/qemu/*.log { + weekly + missingok + rotate 4 + compress + delaycompress + copytruncate + minsize 100k +} diff --git a/daemon/libvirtd.uml.logrotate.in b/daemon/libvirtd.uml.logrotate.in new file mode 100644 index 0000000..135a37d --- /dev/null +++ b/daemon/libvirtd.uml.logrotate.in @@ -0,0 +1,9 @@ +@localstatedir@/log/libvirt/uml/*.log { + weekly + missingok + rotate 4 + compress + delaycompress + copytruncate + minsize 100k +} diff --git a/libvirt.spec.in b/libvirt.spec.in index b1c1c99..17102d9 100644 --- a/libvirt.spec.in +++ b/libvirt.spec.in @@ -668,16 +668,20 @@ fi %{_sysconfdir}/rc.d/init.d/libvirtd %config(noreplace) %{_sysconfdir}/sysconfig/libvirtd %config(noreplace) %{_sysconfdir}/libvirt/libvirtd.conf -%config(noreplace) %{_sysconfdir}/logrotate.d/libvirtd %dir %attr(0700, root, root) %{_localstatedir}/log/libvirt/qemu/ %dir %attr(0700, root, root) %{_localstatedir}/log/libvirt/lxc/ %dir %attr(0700, root, root) %{_localstatedir}/log/libvirt/uml/ %if %{with_qemu} %config(noreplace) %{_sysconfdir}/libvirt/qemu.conf +%config(noreplace) %{_sysconfdir}/logrotate.d/libvirtd.qemu %endif %if %{with_lxc} %config(noreplace) %{_sysconfdir}/libvirt/lxc.conf +%config(noreplace) %{_sysconfdir}/logrotate.d/libvirtd.lxc +%endif +%if %{with_uml} +%config(noreplace) %{_sysconfdir}/logrotate.d/libvirtd.uml %endif %dir %{_datadir}/libvirt/ -- 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/

4 8

[libvirt] [PATCH] Small fix for LSB compilance of init script
by Daniel Veillard 18 Mar '10

18 Mar '10

A trivial small fix: Fix LSB compliance of init script https://bugzilla.redhat.com/show_bug.cgi?id=538701 * daemon/libvirtd.init.in: daemon/libvirtd.init.in were not mentionned in the usage message and if a missing or wrong argument is given it should return 2, not 1 diff --git a/daemon/libvirtd.init.in b/daemon/libvirtd.init.in index b808ab3..4c8821b 100644 --- a/daemon/libvirtd.init.in +++ b/daemon/libvirtd.init.in @@ -106,8 +106,8 @@ case "$1" in [ -f @localstatedir@/lock/subsys/$SERVICE ] && restart || : ;; *) - echo $"Usage: $0 {start|stop|status|restart|condrestart|reload}" - exit 1 + echo $"Usage: $0 {start|stop|status|restart|condrestart|reload|force-reload|try-restart}" + exit 2 ;; esac exit $RETVAL -- 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 1

[libvirt] [PATCH 0/13] [RFC] Network filtering (ACL) extensions for libvirt
by Stefan Berger 18 Mar '10

18 Mar '10

Hi! The following set of patches add network filtering (ACL) extensions to libvirt and enable network traffic filtering for VMs using ebtables and, depending on the networking technology being used (tap, but not macvtap), also iptables. Usage of either is optional and controlled through filters that a VM is referencing. The ebtables-level filtering is based on the XML derived from the CIM network slide 10 (filtering) from the DMTF website (http://www.dmtf.org/standards/cim/cim_schema_v2230/CIM_Network.pdf) The XML we derived from this was discussed on the list before. On the ebtables level we currently handle filtering of IPv4 and ARP traffic. The iptables-level filtering is based on similar XML where XML nodes described the particular protocol to filter for. Its extensions enable the filtering of traffic using iptables for tcp, udp, icmp, igmp, sctp and 'all' types of traffic. This list of protocols maps to the features supported by iptables and only excludes protocols like 'esp' and 'ah'. Currently only bridging mode is supported and based on availability of the physdev match. The filtering framework adds new libvirt virsh commands for managing the filters. The 5 new commands are: - virsh nwfilter-list - virsh nwfilter-dumpxml <name of filter> - virsh nwfilter-define <name of file containing filter desc.> - virsh nwfilter-undefine <name of filter> - virsh nwfilter-edit <name of filter> Above commands are similar to commands for already existing pools and as such much of the code directly related to the above commands could be borrowed from other drivers. The network filters can either contain rules using the above mentioned XML or contain references to other filters in order to build more complex filters that form some sort of filter tree or can contain both. An example for a filter referencing other filters would be this one here: <filter name='demofilter4' chain='root'> <uuid>66f62d1d-34c1-1421-824f-c62d5ee5e8b6</uuid> <filterref filter='no-mac-spoofing'/> <filterref filter='no-mac-broadcast'/> <filterref filter='no-arp-spoofing'/> <filterref filter='allow-dhcp'> <parameter name='DHCPSERVER' value='10.0.0.1'/> </filterref> <filterref filter='no-other-l2-traffic'/> <filterref filter='recv-only-vm-ipaddress'/> <filterref filter='recv-only-vm-macaddress'/> <filterref filter='l3-test'/> </filter> A filter containing actual rules would look like this: <filter name='no-mac-broadcast' chain='ipv4'> <uuid>ffe2ccd6-edec-7360-1852-6b5ccb553234</uuid> <rule action='drop' direction='out' priority='500'> <mac dstmacaddr='ff:ff:ff:ff:ff:ff'/> </rule> </filter> The filter XML now also holds a priority attribute in the rule. This provides control over the ordering of the applied ebtables/iptables rules beyond their appearance in the XML. The domain XML has been extended to reference a top level filter from within each <interface> XML node. A valid reference to such a top level filter looks like this: <interface type='bridge'> <source bridge='static'/> <filterref filter='demofilter4'> <parameter name='IP' value='9.59.241.151'/> </filterref> </interface> In this XML a parameter IP is passed for instantiation of the referenced filters, that may require the availability of this parameter. In the above case the IP parameter's value describes the value of the IP address of the VM and allows to enable those filters to be instantiated that require this 'IP' variable. If a filter requires a parameter that is not provided, the VM will not start or the interface will not attach to a running VM. Any names of parameters can be provided for instantiation of filters and their names and values only need to pass a regular expression test. Currently only MAC and IP addresses and port numbers can be replaced with variables inside the filter XML. In a subsequent patch we will be adding capability to allow users to omit the IP parameter (only) and enable libvirt to learn the IP address of the VM and have it instantiate the filter once it knows it. While virtual machines are running, it is possible to update their filters. For that all running VMs' filter 'trees' are traversed to detect whether the updated filter is referenced by the VM. If so, its ebtables/iptable rules are applied. If one of the VMs' update fails allupdates are rolled back and the filter XML update is rejected. One comment about the instantiation of the rules: Since the XML allows to create nearly any possible combination of parameters to ebtables or iptables commands, I haven't used the ebtables or iptables wrappers. Instead, I am writing ebtables/iptables command into a buffer, add command line options to each one of them as described in the rule's XML, write the buffer into a file and run it as a script. For those commands that are not allowed to fail I am using the following format to run them: cmd="ebtables <some options>" r=`${cmd}` if [ $? -ne 0 ]; then echo "Failure in command ${cmd}." exit 1 fi cmd="..." [...] If one of the command fails in such a batch, the libvirt code is going pick up the error code '1', tear down anything previously established and report an error back. The actual error message shown above is currently not reported back, but can be later on with some changes to the commands running external programs that need to read the script's stdout. One comment to patch 13: It currently #include's a .c file into a .c file only for the reason so I don't have to change too much code once I change code in the underlying patch. So this has to be changed. The patch series works without patch 13, but then only supports ebtables. The patches apply to the current tip. They pass 'make syntax-check' and have been frequently run in valgrind for memory leak checks. The order in which I apply the patches is as follows: add_recursive_locks.diff add_build_support.diff add_public_api.diff add_internal_api.diff impl_pub_api.diff def_wire_protocol_format.diff impl_rpc_client.c impl_srv_dispatch.diff add_virsh_support.diff add_xml_parsing.diff add_qemu_support.diff impl_driver.diff add_iptables_support.diff Looking forward to your feedback on the patches. Thanks and regards, Stefan and Gerhard

2 6

[libvirt] Network XML for static IP address.
by Kumar L Srikanth-B22348 18 Mar '10

18 Mar '10

Hi, I want to assign a static IP address to one of the interfaces created through libvirt. Can anyone please let me know the network XML format? I explored lot of sites on this, but I only found assigning IP address through DHCP rather than Static. Please help me. Regards, Srikanth.

2 1

[libvirt] virStoragePoolGetXMLDesc - how to specify format type
by Sharadha Prabhakar (3P) 18 Mar '10

18 Mar '10

Hi, I'm trying to write virStoragePoolGetXMLDesc() for XenAPI remote storage. I'd like to produce an XML similar to this <pool type="netfs"> <name>....</name> <uuid>....</uuid> <source> <format type="nfs"/> <host name="telos"/> <dir path="/images"/> </source> </pool> I'm trying to fill in the virStoragePoolDefPtr for this. I need to know if struct _virStoragePoolSource->format Is the one to fill for format type="nfs". It's seemingly an integer. Is there any enum for format types for Nfs and ext3? I couldn't find any in storage_conf.h My next query is, when would I have to fill in device path? What is it used for And which pool types use it for remote storage? Could someone explain? Regards, Sharadha

1 0

[libvirt] Assigning Static IP through libvirt.
by Kumar L Srikanth-B22348 18 Mar '10

18 Mar '10

Hi, I want to assign a static IP address to one of the interfaces created through libvirt. Can anyone please let me know the network XML format? I explored lot of sites on this, but I only found assigning IP address through DHCP rather than Static. Can u please help me. Regards, Srikanth.

1 0

[libvirt] [PATCH 9/13] Add virsh support for new CLI commands
by Stefan Berger 18 Mar '10

18 Mar '10

This patch adds virsh support for the five new CLI commands to manage network filters. Signed-off-by: Stefan Berger <stefanb(a)us.ibm.com>

3 3

[libvirt] [PATCH] docs: <pre> cannot be nested in 
by Matthias Bolte 17 Mar '10

17 Mar '10

xsltproc complained about this. --- docs/hacking.html.in | 7 ++++--- 1 files changed, 4 insertions(+), 3 deletions(-) diff --git a/docs/hacking.html.in b/docs/hacking.html.in index f5ec635..03a1bee 100644 --- a/docs/hacking.html.in +++ b/docs/hacking.html.in @@ -128,20 +128,19 @@ For variadic macros, stick with C99 syntax: + <pre> #define vshPrint(_ctl, ...) fprintf(stdout, __VA_ARGS__) </pre> - Use parenthesis when checking if a macro is defined, and use indentation to track nesting: - + <pre> #if defined(HAVE_POSIX_FALLOCATE) && !defined(HAVE_FALLOCATE) # define fallocate(a,ignored,b,c) posix_fallocate(a,b,c) #endif </pre> - <h2><a href="types">C types</a></h2> @@ -552,9 +551,11 @@ also rebuild locally, run 'make check syntax-check', and make sure you don't raise errors. Try to look for warnings too; for example, configure with + <pre> --enable-compile-warnings=error </pre> + which adds -Werror to compile flags, so no warnings get missed -- 1.6.3.3

2 2

Re: [libvirt] virsh stucked in virDomainCreate
by Matthias Bolte 17 Mar '10

17 Mar '10

[Let's keep it on the list, so other could help too] 2010/3/15 Anthony Lannuzel <lannuzel(a)gmail.com>: > On Sat, Mar 13, 2010 at 6:51 PM, Matthias Bolte > <matthias.bolte(a)googlemail.com> wrote: >> 2010/3/13 Anthony Lannuzel <lannuzel(a)gmail.com>: >>> Hi, >>> >>> I'm having an issue when trying to start a (previously created) vbox >>> item : virsh gets stucked in virDomainCreate and I see no error >>> message (LIBVIRT_DEBUG=1). >>> Is there any mean to get more information on what is happening ? >>> >>> Moreover, now that virsh has stalled, running "connect >>> vbox:///session" on another virsh instance gets an error from "secret >>> driver". Is this normal, as if I'm only supposed to have 1 connection >>> at a time ? >>> >>> Thanks >>> Regards >>> >>> Anthony >>> >> >> I tested starting vbox guest using virsh and it works as expected and >> I can use multiple connections at the same time as expected. >> >> What libvirt and vbox versions do you use? >> >> Can you give some more details about what you did to trigger this problem? >> >> Matthias >> > > Hi, > > I'm using virsh 0.7.6 with VirtualBox 3.1.4 on a 2.6.24 kernel with > openVZ, unionfs and squashfs patches. > > Here is what I do : > > The domain is created with virDomainDefineXML. Here is the XML file content: > > <domain type="vbox"> > <name>c7c0ab9f-9a19-4a45-92a1-799367789509</name> > <uuid>c7c0ab9f-9a19-4a45-92a1-799367789509</uuid> > <vcpu>1</vcpu> > <memory>196608</memory> > <on_poweroff>destroy</on_poweroff> > <on_reboot>destroy</on_reboot> > <on_crash>destroy</on_crash> > <os> > <type>hvm</type> > </os> > <features> > <pae></pae> > <acpi></acpi> > <apic></apic> > </features> > <devices> > <disk device="disk" type="file"> > <source file="/data/c7c0ab9f-9a19-4a45-92a1-799367789509/ubuntuDesktop-8.04.vdi"/> > <target bus="ide" dev="hda"/> > </disk> > <interface type="bridge"> > <source bridge="hnsTap0_tap"/> > <mac address="00:22:94:ae:a1:86"/> > <model type="82543gc"/> > </interface> > <graphics port="12000" type="rdp"/> > </devices> > </domain> > > > > Then I try to start it with virsh: > > root@hynesim-live:/home/hynesim# virsh > 10:09:06.685: debug : virInitialize:336 : register drivers > 10:09:06.685: debug : virRegisterDriver:837 : registering Test as driver 0 > 10:09:06.685: debug : virRegisterNetworkDriver:675 : registering Test > as network driver 0 > 10:09:06.685: debug : virRegisterInterfaceDriver:706 : registering > Test as interface driver 0 > 10:09:06.685: debug : virRegisterStorageDriver:737 : registering Test > as storage driver 0 > 10:09:06.685: debug : virRegisterDeviceMonitor:768 : registering Test > as device driver 0 > 10:09:06.685: debug : virRegisterSecretDriver:799 : registering Test > as secret driver 0 > 10:09:06.685: debug : virRegisterDriver:837 : registering OPENVZ as driver 1 > 10:09:06.709: debug : vboxRegister:87 : VBoxCGlueInit found API > version: 3.1.4 (3001004) > 10:09:06.718: debug : vboxRegister:104 : VirtualBox API version: 3.1 > 10:09:06.718: debug : virRegisterDriver:837 : registering VBOX as driver 2 > 10:09:06.718: debug : virRegisterNetworkDriver:675 : registering VBOX > as network driver 1 > 10:09:06.718: debug : virRegisterStorageDriver:737 : registering VBOX > as storage driver 1 > 10:09:06.718: debug : virRegisterDriver:837 : registering remote as driver 3 > 10:09:06.718: debug : virRegisterNetworkDriver:675 : registering > remote as network driver 2 > 10:09:06.719: debug : virRegisterInterfaceDriver:706 : registering > remote as interface driver 1 > 10:09:06.719: debug : virRegisterStorageDriver:737 : registering > remote as storage driver 2 > 10:09:06.719: debug : virRegisterDeviceMonitor:768 : registering > remote as device driver 1 > 10:09:06.719: debug : virRegisterSecretDriver:799 : registering remote > as secret driver 1 > 10:09:06.719: debug : virConnectOpenAuth:1355 : name=(null), > auth=0x2b78d7561c80, flags=0 > 10:09:06.719: debug : do_open:1112 : no name, allowing driver auto-select > 10:09:06.719: debug : do_open:1120 : trying driver 0 (Test) ... > 10:09:06.719: debug : do_open:1126 : driver 0 Test returned DECLINED > 10:09:06.719: debug : do_open:1120 : trying driver 1 (OPENVZ) ... > 10:09:06.759: debug : virExecWithHook:637 : LC_ALL=C /usr/sbin/vzctl --help > 10:09:06.771: debug : do_open:1126 : driver 1 OPENVZ returned SUCCESS > 10:09:06.771: debug : do_open:1146 : network driver 0 Test returned DECLINED > 10:09:06.771: debug : do_open:1146 : network driver 1 VBOX returned DECLINED > 10:09:06.772: debug : doRemoteOpen:564 : proceeding with name = openvz:///system > 10:09:06.774: debug : remoteIO:8429 : Do proc=66 serial=0 length=28 wait=(nil) > 10:09:06.774: debug : remoteIO:8491 : We have the buck 66 > 0x2b78dbcf2010 0x2b78dbcf2010 > 10:09:06.776: debug : remoteIODecodeMessageLength:7913 : Got length, > now need 64 total (60 more) > 10:09:06.776: debug : remoteIOEventLoop:8355 : Giving up the buck 66 > 0x2b78dbcf2010 (nil) > 10:09:06.776: debug : remoteIO:8522 : All done with our call 66 (nil) > 0x2b78dbcf2010 > 10:09:06.776: debug : remoteIO:8429 : Do proc=1 serial=1 length=56 wait=(nil) > 10:09:06.776: debug : remoteIO:8491 : We have the buck 1 0x63fe50 0x63fe50 > 10:09:06.848: debug : remoteIODecodeMessageLength:7913 : Got length, > now need 56 total (52 more) > 10:09:06.848: debug : remoteIOEventLoop:8355 : Giving up the buck 1 > 0x63fe50 (nil) > 10:09:06.848: debug : remoteIO:8522 : All done with our call 1 (nil) 0x63fe50 > 10:09:06.848: debug : doRemoteOpen:891 : Adding Handler for remote events > 10:09:06.848: debug : doRemoteOpen:898 : virEventAddHandle failed: No > addHandleImpl defined. continuing without events. > 10:09:06.848: debug : do_open:1146 : network driver 2 remote returned SUCCESS > 10:09:06.848: debug : do_open:1165 : interface driver 0 Test returned DECLINED > 10:09:06.849: debug : doRemoteOpen:564 : proceeding with name = openvz:///system > 10:09:06.850: debug : remoteIO:8429 : Do proc=66 serial=0 length=28 wait=(nil) > 10:09:06.850: debug : remoteIO:8491 : We have the buck 66 0x67ff00 0x67ff00 > 10:09:06.852: debug : remoteIODecodeMessageLength:7913 : Got length, > now need 64 total (60 more) > 10:09:06.852: debug : remoteIOEventLoop:8355 : Giving up the buck 66 > 0x67ff00 (nil) > 10:09:06.852: debug : remoteIO:8522 : All done with our call 66 (nil) 0x67ff00 > 10:09:06.853: debug : remoteIO:8429 : Do proc=1 serial=1 length=56 wait=(nil) > 10:09:06.853: debug : remoteIO:8491 : We have the buck 1 0x67ff00 0x67ff00 > 10:09:06.901: debug : remoteIODecodeMessageLength:7913 : Got length, > now need 56 total (52 more) > 10:09:06.901: debug : remoteIOEventLoop:8355 : Giving up the buck 1 > 0x67ff00 (nil) > 10:09:06.901: debug : remoteIO:8522 : All done with our call 1 (nil) 0x67ff00 > 10:09:06.901: debug : doRemoteOpen:891 : Adding Handler for remote events > 10:09:06.901: debug : doRemoteOpen:898 : virEventAddHandle failed: No > addHandleImpl defined. continuing without events. > 10:09:06.901: debug : do_open:1165 : interface driver 1 remote returned SUCCESS > 10:09:06.901: debug : do_open:1185 : storage driver 0 Test returned DECLINED > 10:09:06.901: debug : do_open:1185 : storage driver 1 VBOX returned DECLINED > 10:09:06.901: debug : do_open:1185 : storage driver 2 remote returned SUCCESS > 10:09:06.901: debug : do_open:1205 : node driver 0 Test returned DECLINED > 10:09:06.901: debug : do_open:1205 : node driver 1 remote returned SUCCESS > 10:09:06.901: debug : do_open:1232 : secret driver 0 Test returned DECLINED > 10:09:06.901: debug : do_open:1232 : secret driver 1 remote returned SUCCESS > Bienvenue dans virsh, le terminal de virtualisation interactif. > > Taper : « help » pour l'aide ou « help » avec la commande > « quit » pour quitter > > virsh # connect vbox:///session > 10:09:13.993: debug : virConnectClose:1381 : conn=0x63b060 > 10:09:13.994: debug : virUnrefConnect:259 : unref connection 0x63b060 1 > 10:09:13.994: debug : remoteIO:8429 : Do proc=2 serial=2 length=28 wait=(nil) > 10:09:13.994: debug : remoteIO:8491 : We have the buck 2 0x680170 0x680170 > 10:09:13.995: debug : remoteIODecodeMessageLength:7913 : Got length, > now need 56 total (52 more) > 10:09:13.995: debug : remoteIOEventLoop:8355 : Giving up the buck 2 > 0x680170 (nil) > 10:09:13.995: debug : remoteIO:8522 : All done with our call 2 (nil) 0x680170 > 10:09:13.996: debug : remoteIO:8429 : Do proc=2 serial=2 length=28 wait=(nil) > 10:09:13.996: debug : remoteIO:8491 : We have the buck 2 0x680170 0x680170 > 10:09:13.997: debug : remoteIODecodeMessageLength:7913 : Got length, > now need 56 total (52 more) > 10:09:13.997: debug : remoteIOEventLoop:8355 : Giving up the buck 2 > 0x680170 (nil) > 10:09:13.997: debug : remoteIO:8522 : All done with our call 2 (nil) 0x680170 > 10:09:13.997: debug : virDomainObjUnref:680 : obj=0x63ebf0 refs=0 > 10:09:13.997: debug : virDomainObjFree:658 : obj=0x63ebf0 > 10:09:13.997: debug : virReleaseConnect:216 : release connection 0x63b060 > 10:09:13.997: debug : virConnectOpenAuth:1355 : name=vbox:///session, > auth=0x2b78d7561c80, flags=0 > 10:09:13.997: debug : do_open:1110 : name "vbox:///session" to URI components: > scheme vbox > opaque (null) > authority (null) > server (null) > user (null) > port 0 > path /session > > 10:09:13.997: debug : do_open:1120 : trying driver 0 (Test) ... > 10:09:13.997: debug : do_open:1126 : driver 0 Test returned DECLINED > 10:09:13.997: debug : do_open:1120 : trying driver 1 (OPENVZ) ... > 10:09:13.997: debug : do_open:1126 : driver 1 OPENVZ returned DECLINED > 10:09:13.997: debug : do_open:1120 : trying driver 2 (VBOX) ... > 10:09:14.021: debug : vboxOpen:836 : in vboxOpen > 10:09:14.021: debug : do_open:1126 : driver 2 VBOX returned SUCCESS > 10:09:14.021: debug : do_open:1146 : network driver 0 Test returned DECLINED > 10:09:14.021: debug : vboxNetworkOpen:5434 : network initialized > 10:09:14.021: debug : do_open:1146 : network driver 1 VBOX returned SUCCESS > 10:09:14.022: debug : do_open:1165 : interface driver 0 Test returned DECLINED > 10:09:14.022: debug : doRemoteOpen:564 : proceeding with name = vbox:///session > 10:09:14.024: debug : remoteIO:8429 : Do proc=66 serial=0 length=28 wait=(nil) > 10:09:14.024: debug : remoteIO:8491 : We have the buck 66 0x6eaa90 0x6eaa90 > 10:09:14.025: debug : remoteIODecodeMessageLength:7913 : Got length, > now need 64 total (60 more) > 10:09:14.025: debug : remoteIOEventLoop:8355 : Giving up the buck 66 > 0x6eaa90 (nil) > 10:09:14.025: debug : remoteIO:8522 : All done with our call 66 (nil) 0x6eaa90 > 10:09:14.026: debug : remoteIO:8429 : Do proc=1 serial=1 length=56 wait=(nil) > 10:09:14.026: debug : remoteIO:8491 : We have the buck 1 0x6eaa90 0x6eaa90 > 10:09:14.027: debug : remoteIODecodeMessageLength:7913 : Got length, > now need 56 total (52 more) > 10:09:14.027: debug : remoteIOEventLoop:8355 : Giving up the buck 1 > 0x6eaa90 (nil) > 10:09:14.027: debug : remoteIO:8522 : All done with our call 1 (nil) 0x6eaa90 > 10:09:14.027: debug : doRemoteOpen:891 : Adding Handler for remote events > 10:09:14.027: debug : doRemoteOpen:898 : virEventAddHandle failed: No > addHandleImpl defined. continuing without events. > 10:09:14.027: debug : do_open:1165 : interface driver 1 remote returned SUCCESS > 10:09:14.027: debug : do_open:1185 : storage driver 0 Test returned DECLINED > 10:09:14.027: debug : vboxStorageOpen:6186 : vbox storage initialized > 10:09:14.027: debug : do_open:1185 : storage driver 1 VBOX returned SUCCESS > 10:09:14.028: debug : do_open:1205 : node driver 0 Test returned DECLINED > 10:09:14.028: debug : doRemoteOpen:564 : proceeding with name = vbox:///session > 10:09:14.030: debug : remoteIO:8429 : Do proc=66 serial=0 length=28 wait=(nil) > 10:09:14.037: debug : remoteIO:8491 : We have the buck 66 0x72ab40 0x72ab40 > 10:09:14.038: debug : remoteIODecodeMessageLength:7913 : Got length, > now need 64 total (60 more) > 10:09:14.038: debug : remoteIOEventLoop:8355 : Giving up the buck 66 > 0x72ab40 (nil) > 10:09:14.038: debug : remoteIO:8522 : All done with our call 66 (nil) 0x72ab40 > 10:09:14.038: debug : remoteIO:8429 : Do proc=1 serial=1 length=56 wait=(nil) > 10:09:14.039: debug : remoteIO:8491 : We have the buck 1 0x72ab40 0x72ab40 > 10:09:14.040: debug : remoteIODecodeMessageLength:7913 : Got length, > now need 56 total (52 more) > 10:09:14.040: debug : remoteIOEventLoop:8355 : Giving up the buck 1 > 0x72ab40 (nil) > 10:09:14.040: debug : remoteIO:8522 : All done with our call 1 (nil) 0x72ab40 > 10:09:14.040: debug : doRemoteOpen:891 : Adding Handler for remote events > 10:09:14.040: debug : doRemoteOpen:898 : virEventAddHandle failed: No > addHandleImpl defined. continuing without events. > 10:09:14.040: debug : do_open:1205 : node driver 1 remote returned SUCCESS > 10:09:14.040: debug : do_open:1232 : secret driver 0 Test returned DECLINED > 10:09:14.040: debug : doRemoteOpen:564 : proceeding with name = vbox:///session > 10:09:14.041: debug : remoteIO:8429 : Do proc=66 serial=0 length=28 wait=(nil) > 10:09:14.041: debug : remoteIO:8491 : We have the buck 66 0x76abf0 0x76abf0 > 10:09:14.042: debug : remoteIODecodeMessageLength:7913 : Got length, > now need 64 total (60 more) > 10:09:14.042: debug : remoteIOEventLoop:8355 : Giving up the buck 66 > 0x76abf0 (nil) > 10:09:14.042: debug : remoteIO:8522 : All done with our call 66 (nil) 0x76abf0 > 10:09:14.043: debug : remoteIO:8429 : Do proc=1 serial=1 length=56 wait=(nil) > 10:09:14.043: debug : remoteIO:8491 : We have the buck 1 0x76abf0 0x76abf0 > 10:09:14.044: debug : remoteIODecodeMessageLength:7913 : Got length, > now need 56 total (52 more) > 10:09:14.044: debug : remoteIOEventLoop:8355 : Giving up the buck 1 > 0x76abf0 (nil) > 10:09:14.044: debug : remoteIO:8522 : All done with our call 1 (nil) 0x76abf0 > 10:09:14.044: debug : doRemoteOpen:891 : Adding Handler for remote events > 10:09:14.044: debug : doRemoteOpen:898 : virEventAddHandle failed: No > addHandleImpl defined. continuing without events. > 10:09:14.044: debug : do_open:1232 : secret driver 1 remote returned SUCCESS > > virsh # list --all > 10:09:19.563: debug : virConnectNumOfDomains:1749 : conn=0x63ebf0 > 10:09:19.568: debug : virConnectNumOfDefinedDomains:4609 : conn=0x63ebf0 > 10:09:19.568: debug : virConnectListDefinedDomains:4648 : > conn=0x63ebf0, names=0x692430, maxnames=1 > ID Nom État > ---------------------------------- > 10:09:19.570: debug : virDomainLookupByName:2010 : conn=0x63ebf0, > name=c7c0ab9f-9a19-4a45-92a1-799367789509 > 10:09:19.571: debug : virGetDomain:345 : New hash entry 0x6a6000 > 10:09:19.571: debug : virDomainGetInfo:2867 : domain=0x6a6000, > info=0x7fffd3a519e0 > - c7c0ab9f-9a19-4a45-92a1-799367789509 closed > 10:09:19.573: debug : virDomainFree:2098 : domain=0x6a6000 > 10:09:19.573: debug : virUnrefDomain:422 : unref domain 0x6a6000 > c7c0ab9f-9a19-4a45-92a1-799367789509 1 > 10:09:19.573: debug : virReleaseDomain:376 : release domain 0x6a6000 > c7c0ab9f-9a19-4a45-92a1-799367789509 > 10:09:19.573: debug : virReleaseDomain:392 : unref connection 0x63ebf0 2 > > virsh # start c7c0ab9f-9a19-4a45-92a1-799367789509 > 10:09:26.108: debug : virDomainLookupByName:2010 : conn=0x63ebf0, > name=c7c0ab9f-9a19-4a45-92a1-799367789509 > 10:09:26.112: debug : virGetDomain:345 : New hash entry 0x68bd70 > 10:09:26.112: debug : virDomainGetID:2658 : domain=0x68bd70 > 10:09:26.112: debug : virDomainCreate:4690 : domain=0x68bd70 > 10:09:26.112: debug : virDomainCreate:4693 : 1 0x68bd70 > 10:09:26.112: debug : virDomainCreate:4700 : 2 0x68bd70 > 10:09:26.112: debug : virDomainCreate:4707 : 3 0x68bd70 > 10:09:26.112: debug : virDomainCreate:4710 : 3.1 0x68bd70 > 10:09:26.223: debug : vboxDomainCreate:3321 : before close 0x68bd70 > > > I have added some debug lines, that lead to: > > in libvirt.c: > function int virDomainCreate(virDomainPtr domain) > ret = conn->driver->domainCreate (domain); // does not return > > in vbox/vbox_tmpl.c: > function static int vboxDomainCreate(virDomainPtr dom) > data->vboxSession->vtbl->Close(data->vboxSession); > // does not return > > Thanks > At that point the domain should already be running. I wonder why closing the session handle would block. You could try variations of the domain XML config, assuming that a specify part of the config triggers this problem. For example use your posted domain XML config, but leave out the graphics part or the interface part or the disk part to see if that makes a difference. Matthias

1 0

[libvirt] Userdata for libvirt?
by Yushu Yao 17 Mar '10

17 Mar '10

Hi Experts, Forgive me if not clear enough, will try to be describe what I need. What I need: Start a VM on a remote libvirtd server based on an existing image (Might or might not exist on server), plus some userdata. It is pretty much the same like EC2 or Eucalyptus, but we need to run on our private cluster with only libvirtd. (Eucalyptus doesn't support using physical disk in VMs, that is what we need). I went through the Remote Libvirt control section of the Documentation, but couldn't find a way to pass the userdata to the VM started. Does libvirtd already include such feature of do I need to make my own. Thanks -Yushu

2 2

[libvirt] Just pushed a few small fixes
by Cole Robinson 17 Mar '10

17 Mar '10

Hi all, I've just pushed three small fixes: $ git log -3 --pretty=short commit 2ef091efcc5cd02bbd496972b141cf253f713fdd Author: Philip Hahn <hahn(a)univention.de> python: Fix networkLookupByUUID commit 0ef58c315506d094a0a4d64f022119a04a1916c0 Author: Cole Robinson <crobinso(a)redhat.com> .gitignore: Ignore generated daemon/libvirtd.logrotate commit 89d8cdfc7e0ba7b2d9cf16a34b8e4a035c851d28 Author: Cole Robinson <crobinso(a)redhat.com> Fix make dist with XenAPI changes The last chronological change was from: https://bugzilla.redhat.com/show_bug.cgi?id=574366 Apparently networkLookupByUUID has always been busted in the python bindings, but since we don't use it in virtinst/virt-manager, it wasn't noticed. - Cole

1 0

[libvirt] disk XML for error policy
by Dave Allan 17 Mar '10

17 Mar '10

Hi Dan, What do you think of this for the disk XML specifying what to do on error: <disk type='file' device='disk'> <source file='/storage/guest_disks/libgdu'/> <target dev='vda' bus='virtio'/> <on_err rerror='stop' werror='stop'/> </disk> Dave

2 2

[libvirt] [PATCH] do not require two ./autogen.sh runs to permit "make"
by Jim Meyering 17 Mar '10

17 Mar '10

I tracked down the source of the two-autogen.sh-runs-required bug. Here's the fix: >From 0583de2de2038d4f8875268391fd4994329d39bf Mon Sep 17 00:00:00 2001 From: Jim Meyering <meyering(a)redhat.com> Date: Tue, 16 Mar 2010 21:08:31 +0100 Subject: [PATCH] do not require two ./autogen.sh runs to permit "make" * autogen.sh (bootstrap_hash): New function. Running bootstrap may update the gnulib SHA1, yet we were computing t=$(git submodule status ...) *prior* to running bootstrap, and then recording that sometimes-stale value in the stamp file upon a successful bootstrap run. That would require two (lengthy!) bootstrap runs to update the stamp file. --- autogen.sh | 21 ++++++++++++++------- 1 files changed, 14 insertions(+), 7 deletions(-) diff --git a/autogen.sh b/autogen.sh index ff94678..b93cdba 100755 --- a/autogen.sh +++ b/autogen.sh @@ -62,20 +62,27 @@ else fi fi +# Compute the hash we'll use to determine whether rerunning bootstrap +# is required. The first is just the SHA1 that selects a gnulib snapshot. +# The second ensures that whenever we change the set of gnulib modules used +# by this package, we rerun bootstrap to pull in the matching set of files. +bootstrap_hash() +{ + git submodule status | sed 's/^[ +-]//;s/ .*//' + git hash-object bootstrap.conf +} + # Ensure that whenever we pull in a gnulib update or otherwise change to a # different version (i.e., when switching branches), we also rerun ./bootstrap. curr_status=.git-module-status -t=$(git submodule status|sed 's/^[ +-]//;s/ .*//'; \ - git hash-object bootstrap.conf) +t=$(bootstrap_hash) if test "$t" = "$(cat $curr_status 2>/dev/null)"; then : # good, it's up to date, all we need is autoreconf autoreconf -if else - echo running bootstrap... - ./bootstrap && echo "$t" > $curr_status || { - echo "Failed to bootstrap gnulib, please investigate." - exit 1; - } + echo running bootstrap... + ./bootstrap && bootstrap_hash > $curr_status \ + || { echo "Failed to bootstrap gnulib, please investigate."; exit 1; } fi cd "$THEDIR" -- 1.7.0.1

3 4

[libvirt] [PATCH] Allow suspend during live migration
by Jiri Denemark 17 Mar '10

17 Mar '10

Currently no command can be sent to a qemu process while another job is active. This patch adds support for signaling long-running jobs (such as migration) so that other threads may request predefined operations to be done during such jobs. Two signals are defined so far: - QEMU_JOB_SIGNAL_CANCEL - QEMU_JOB_SIGNAL_SUSPEND The first one is used by qemuDomainAbortJob. The second one is used by qemudDomainSuspend for suspending a domain during migration, which allows for changing live migration into offline migration. However, there is a small issue in the way qemudDomainSuspend is currently implemented for migrating domains. The API calls returns immediately after signaling migration job which means it is asynchronous in this specific case. Signed-off-by: Jiri Denemark <jdenemar(a)redhat.com> --- src/qemu/qemu_driver.c | 149 ++++++++++++++++++++++++++++++++--------------- 1 files changed, 101 insertions(+), 48 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index f8ab545..5b2c26a 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -87,14 +87,26 @@ #define VIR_FROM_THIS VIR_FROM_QEMU +/* Only 1 job is allowed at any time + * A job includes *all* monitor commands, even those just querying + * information, not merely actions */ +enum qemuDomainJob { + QEMU_JOB_NONE = 0, /* Always set to 0 for easy if (jobActive) conditions */ + QEMU_JOB_UNSPECIFIED, + QEMU_JOB_MIGRATION, +}; + +enum qemuDomainJobSignals { + QEMU_JOB_SIGNAL_CANCEL = 1 << 0, /* Request job cancellation */ + QEMU_JOB_SIGNAL_SUSPEND = 1 << 1, /* Request VM suspend to finish live migration offline */ +}; + typedef struct _qemuDomainObjPrivate qemuDomainObjPrivate; typedef qemuDomainObjPrivate *qemuDomainObjPrivatePtr; struct _qemuDomainObjPrivate { virCond jobCond; /* Use in conjunction with main virDomainObjPtr lock */ - unsigned int jobActive : 1; /* Non-zero if a job is active. Only 1 job is allowed at any time - * A job includes *all* monitor commands, even those just querying - * information, not merely actions */ - unsigned int jobCancel : 1; /* Non-zero if a cancel request from client has arrived */ + enum qemuDomainJob jobActive; /* Currently running job */ + unsigned int jobSignals; /* Signals for running job */ virDomainJobInfo jobInfo; unsigned long long jobStart; @@ -338,8 +350,8 @@ static int qemuDomainObjBeginJob(virDomainObjPtr obj) return -1; } } - priv->jobActive = 1; - priv->jobCancel = 0; + priv->jobActive = QEMU_JOB_UNSPECIFIED; + priv->jobSignals = 0; priv->jobStart = (now.tv_sec * 1000ull) + (now.tv_usec / 1000); memset(&priv->jobInfo, 0, sizeof(priv->jobInfo)); @@ -385,8 +397,8 @@ static int qemuDomainObjBeginJobWithDriver(struct qemud_driver *driver, return -1; } } - priv->jobActive = 1; - priv->jobCancel = 0; + priv->jobActive = QEMU_JOB_UNSPECIFIED; + priv->jobSignals = 0; priv->jobStart = (now.tv_sec * 1000ull) + (now.tv_usec / 1000); memset(&priv->jobInfo, 0, sizeof(priv->jobInfo)); @@ -410,8 +422,8 @@ static int ATTRIBUTE_RETURN_CHECK qemuDomainObjEndJob(virDomainObjPtr obj) { qemuDomainObjPrivatePtr priv = obj->privateData; - priv->jobActive = 0; - priv->jobCancel = 0; + priv->jobActive = QEMU_JOB_NONE; + priv->jobSignals = 0; priv->jobStart = 0; memset(&priv->jobInfo, 0, sizeof(priv->jobInfo)); virCondSignal(&priv->jobCond); @@ -3560,6 +3572,7 @@ static int qemudDomainSuspend(virDomainPtr dom) { virDomainObjPtr vm; int ret = -1; virDomainEventPtr event = NULL; + qemuDomainObjPrivatePtr priv; qemuDriverLock(driver); vm = virDomainFindByUUID(&driver->domains, dom->uuid); @@ -3571,30 +3584,48 @@ static int qemudDomainSuspend(virDomainPtr dom) { _("no domain with matching uuid '%s'"), uuidstr); goto cleanup; } - if (qemuDomainObjBeginJobWithDriver(driver, vm) < 0) - goto cleanup; - if (!virDomainObjIsActive(vm)) { qemuReportError(VIR_ERR_OPERATION_INVALID, "%s", _("domain is not running")); - goto endjob; + goto cleanup; } - if (vm->state != VIR_DOMAIN_PAUSED) { - qemuDomainObjPrivatePtr priv = vm->privateData; - qemuDomainObjEnterMonitorWithDriver(driver, vm); - if (qemuMonitorStopCPUs(priv->mon) < 0) { - qemuDomainObjExitMonitorWithDriver(driver, vm); + + priv = vm->privateData; + + if (priv->jobActive == QEMU_JOB_MIGRATION) { + if (vm->state != VIR_DOMAIN_PAUSED) { + VIR_DEBUG("Requesting domain pause on %s", + vm->def->name); + priv->jobSignals |= QEMU_JOB_SIGNAL_SUSPEND; + } + ret = 0; + goto cleanup; + } else { + if (qemuDomainObjBeginJobWithDriver(driver, vm) < 0) + goto cleanup; + + if (!virDomainObjIsActive(vm)) { + qemuReportError(VIR_ERR_OPERATION_INVALID, + "%s", _("domain is not running")); goto endjob; } - qemuDomainObjExitMonitorWithDriver(driver, vm); - vm->state = VIR_DOMAIN_PAUSED; - event = virDomainEventNewFromObj(vm, - VIR_DOMAIN_EVENT_SUSPENDED, - VIR_DOMAIN_EVENT_SUSPENDED_PAUSED); + if (vm->state != VIR_DOMAIN_PAUSED) { + int rc; + + qemuDomainObjEnterMonitorWithDriver(driver, vm); + rc = qemuMonitorStopCPUs(priv->mon); + qemuDomainObjExitMonitorWithDriver(driver, vm); + if (rc < 0) + goto endjob; + vm->state = VIR_DOMAIN_PAUSED; + event = virDomainEventNewFromObj(vm, + VIR_DOMAIN_EVENT_SUSPENDED, + VIR_DOMAIN_EVENT_SUSPENDED_PAUSED); + } + if (virDomainSaveStatus(driver->caps, driver->stateDir, vm) < 0) + goto endjob; + ret = 0; } - if (virDomainSaveStatus(driver->caps, driver->stateDir, vm) < 0) - goto endjob; - ret = 0; endjob: if (qemuDomainObjEndJob(vm) == 0) @@ -3946,6 +3977,35 @@ cleanup: } +/** qemuDomainMigrateOffline: + * Pause domain for non-live migration. + */ +static int +qemuDomainMigrateOffline(struct qemud_driver *driver, + virDomainObjPtr vm) +{ + qemuDomainObjPrivatePtr priv = vm->privateData; + int ret; + + qemuDomainObjEnterMonitorWithDriver(driver, vm); + ret = qemuMonitorStopCPUs(priv->mon); + qemuDomainObjExitMonitorWithDriver(driver, vm); + + if (ret == 0) { + virDomainEventPtr event; + + vm->state = VIR_DOMAIN_PAUSED; + event = virDomainEventNewFromObj(vm, + VIR_DOMAIN_EVENT_SUSPENDED, + VIR_DOMAIN_EVENT_SUSPENDED_MIGRATED); + if (event) + qemuDomainEventQueue(driver, event); + } + + return ret; +} + + static int qemuDomainWaitForMigrationComplete(struct qemud_driver *driver, virDomainObjPtr vm) { @@ -3964,8 +4024,8 @@ qemuDomainWaitForMigrationComplete(struct qemud_driver *driver, virDomainObjPtr struct timeval now; int rc; - if (priv->jobCancel) { - priv->jobCancel = 0; + if (priv->jobSignals & QEMU_JOB_SIGNAL_CANCEL) { + priv->jobSignals ^= QEMU_JOB_SIGNAL_CANCEL; VIR_DEBUG0("Cancelling migration at client request"); qemuDomainObjEnterMonitorWithDriver(driver, vm); rc = qemuMonitorMigrateCancel(priv->mon); @@ -3973,6 +4033,11 @@ qemuDomainWaitForMigrationComplete(struct qemud_driver *driver, virDomainObjPtr if (rc < 0) { VIR_WARN0("Unable to cancel migration"); } + } else if (priv->jobSignals & QEMU_JOB_SIGNAL_SUSPEND) { + priv->jobSignals ^= QEMU_JOB_SIGNAL_SUSPEND; + VIR_DEBUG0("Pausing domain for non-live migration"); + if (qemuDomainMigrateOffline(driver, vm) < 0) + VIR_WARN0("Unable to pause domain"); } qemuDomainObjEnterMonitorWithDriver(driver, vm); @@ -8979,7 +9044,7 @@ qemudDomainMigratePerform (virDomainPtr dom, virDomainObjPtr vm; virDomainEventPtr event = NULL; int ret = -1; - int paused = 0; + int resume = 0; qemuDomainObjPrivatePtr priv; qemuDriverLock(driver); @@ -8995,6 +9060,7 @@ qemudDomainMigratePerform (virDomainPtr dom, if (qemuDomainObjBeginJobWithDriver(driver, vm) < 0) goto cleanup; + priv->jobActive = QEMU_JOB_MIGRATION; if (!virDomainObjIsActive(vm)) { qemuReportError(VIR_ERR_OPERATION_INVALID, @@ -9005,23 +9071,10 @@ qemudDomainMigratePerform (virDomainPtr dom, memset(&priv->jobInfo, 0, sizeof(priv->jobInfo)); priv->jobInfo.type = VIR_DOMAIN_JOB_UNBOUNDED; + resume = vm->state == VIR_DOMAIN_RUNNING; if (!(flags & VIR_MIGRATE_LIVE) && vm->state == VIR_DOMAIN_RUNNING) { - /* Pause domain for non-live migration */ - qemuDomainObjEnterMonitorWithDriver(driver, vm); - if (qemuMonitorStopCPUs(priv->mon) < 0) { - qemuDomainObjExitMonitorWithDriver(driver, vm); + if (qemuDomainMigrateOffline(driver, vm) < 0) goto endjob; - } - qemuDomainObjExitMonitorWithDriver(driver, vm); - paused = 1; - - vm->state = VIR_DOMAIN_PAUSED; - event = virDomainEventNewFromObj(vm, - VIR_DOMAIN_EVENT_SUSPENDED, - VIR_DOMAIN_EVENT_SUSPENDED_MIGRATED); - if (event) - qemuDomainEventQueue(driver, event); - event = NULL; } if ((flags & (VIR_MIGRATE_TUNNELLED | VIR_MIGRATE_PEER2PEER))) { @@ -9035,7 +9088,7 @@ qemudDomainMigratePerform (virDomainPtr dom, /* Clean up the source domain. */ qemudShutdownVMDaemon(driver, vm); - paused = 0; + resume = 0; event = virDomainEventNewFromObj(vm, VIR_DOMAIN_EVENT_STOPPED, @@ -9049,7 +9102,7 @@ qemudDomainMigratePerform (virDomainPtr dom, ret = 0; endjob: - if (paused) { + if (resume && vm->state == VIR_DOMAIN_PAUSED) { /* we got here through some sort of failure; start the domain again */ qemuDomainObjEnterMonitorWithDriver(driver, vm); if (qemuMonitorStartCPUs(priv->mon, dom->conn) < 0) { @@ -9425,7 +9478,7 @@ static int qemuDomainAbortJob(virDomainPtr dom) { if (virDomainObjIsActive(vm)) { if (priv->jobActive) { VIR_DEBUG("Requesting cancellation of job on vm %s", vm->def->name); - priv->jobCancel = 1; + priv->jobSignals |= QEMU_JOB_SIGNAL_CANCEL; } else { qemuReportError(VIR_ERR_OPERATION_INVALID, "%s", _("no job is active on the domain")); -- 1.7.0.2

2 2

[libvirt] [PATCH 1/13] Adding recursive locks
by Stefan Berger 17 Mar '10

17 Mar '10

This patch adds recursive locks necessary due to the processing of network filter XML that can reference other network filters, including references that cause looks. Loops in the XML are prevented but their detection requires recursive locks.

2 2

[libvirt] [RFC][PATCH v2 0/2] Dynamic backend setup for macvtap interfaces
by Ed Swierk 17 Mar '10

17 Mar '10

I posted this RFC/patch set a few weeks ago but didn't receive any response. My implementation works, but I'd like to hear from anyone more familiar with libvirt concurrency than I (i.e. nearly everyone) about how it might be improved. In v2 I rebased against libvirt-0.7.7 and made a bunch of minor changes. - Using a bridge to connect a qemu NIC to a host interface offers a fair amount of flexibility to reconfigure the host without restarting the VM. For example, if the bridge connects host interface eth0 to the qemu tap0 interface, eth0 can be hot-removed and hot-plugged without affecting the VM. Similarly, if the bridge connects host VLAN interface vlan0 to the qemu tap0 interface, the admin can easily replace vlan0 with vlan1 without the VM noticing. Using the macvtap driver instead of a kernel bridge, the host interface is much more tightly tied to the VM. Qemu communicates with the macvtap interface through a file descriptor, and the macvtap interface is bound permanently to a specific host interface when it is first created. What's more, if the underlying host interface disappears, the macvtap interface vanishes along with it, leaving the VM holding a file descriptor for a deleted file. To avoid race conditions during system startup, I would like libvirt to allow starting up the VM with a NIC even if the underlying host interface doesn't yet exist, deferring creation of the macvtap interface (analogous to starting up the VM with a tap interface bound to an orphan bridge). To support adding and removing a host interface without restarting the VM, I would like libvirt to react to the (re)appearance of the underlying host interface, creating a new macvtap interface and passing the new fd to qemu to reconnect to the NIC. (It would also be nice if libvirt allowed the user to change which underlying host interface the qemu NIC is connected to. I'm ignoring this issue for now, except to note that implementing the above features should make this easier.) The libvirt API already supports domainAttachDevice and domainDetachDevice to add or remove an interface while the VM is running. In the qemu implementation, these commands add or remove the VM NIC device as well as reconfiguring the host side. This works only if the OS and application running in the VM can handle PCI hotplug and dynamically reconfigure its network. I would like to isolate the VM from changes to the host network setup, whether you use macvtap or a bridge. The changes I think are needed to implement this include: 1. Refactor qemudDomainAttachNetDevice/qemudDomainDetachNetDevice, which currently handle both backend (host) setup and adding/removing the VM NIC device; move the backend setup code into separate functions that can called separately without affecting VM devices. 2. Implement a thread or task that watches for changes to the underlying host interface for each configured macvtap interface, and reacts by invoking the appropriate backend setup code. 3. Change qemudBuildCommandLine to defer backend setup if qemu supports the necessary features for doing it later (e.g. the host_net_add monitor command). 4. Implement appropriate error handling and reporting, and any necessary changes to the configuration schema. The following patches are a partial implementation of the above as a proof of concept. Patch 1 implements change (1) above, moving the backend setup code to new functions qemudDomainConnectNetBackend/qemudDomainDisconnectNetBackend, and calling these functions from the existing qemudDomainAttachNetDevice/qemudDomainDetachNetDevice. I think this change is useful on its own: it breaks up two monster functions into more manageable pieces, and eliminates some code duplication (e.g. the try_remove clause at the end of qemudDomainAttachNetDevice). Patch 2 is a godawful hack roughly implementing changes (2) and (3) above (did I mention that this is a proof of concept?). It spawns a thread that simply tries reconnecting the backend of each macvtap interface once a second. As long as the interface is already up, the reconnection fails. If the macvtap interface goes away because the underlying host interface disappears, the reconnection fails until the host interface reappears. I ran into two major issues while implementing (2) and (3): - Can we use the existing virEvent functions to invoke the reconnection process, triggered either by a timer or by an event from the host? It seems like this ought to work, but it appears that communication between libvirt and the qemu monitor relies on an event, and since all events run in the same thread, there's no way for an event to call the monitor. - Should the reconnection process use udev or hal to get notifications, or leverage the node device code which itself uses udev or hal? Currently there doesn't appear to be a way to get notifications of changes to node devices; if there were, we'd still need to address the threading issue. If we use node devices, what changes to the configuration schema would be needed to associate a macvtap interface with the underlying node device? I'd appreciate input on item (4) as well (e.g. does it always make sense to ignore the missing host interface on the assumption that it could show up later?). --Ed

3 3

[libvirt] [PATCH] fix two "make syntax check" failures
by Jim Meyering 16 Mar '10

16 Mar '10

Laine mentioned that there were two syntax-check failures. I've just pushed this fix: >From a31bc6750347a79dbd43b0aacecf86d204e6e160 Mon Sep 17 00:00:00 2001 From: Jim Meyering <meyering(a)redhat.com> Date: Tue, 16 Mar 2010 19:32:05 +0100 Subject: [PATCH] fix two "make syntax check" failures * src/xenapi/xenapi_driver.c (xenapiOpen): Remove useless-if-before-free. * po/POTFILES.in: Add src/xenapi/xenapi_utils.c. --- po/POTFILES.in | 1 + src/xenapi/xenapi_driver.c | 2 +- 2 files changed, 2 insertions(+), 1 deletions(-) diff --git a/po/POTFILES.in b/po/POTFILES.in index ac169be..1845572 100644 --- a/po/POTFILES.in +++ b/po/POTFILES.in @@ -78,5 +78,6 @@ src/xen/xen_hypervisor.c src/xen/xen_inotify.c src/xen/xm_internal.c src/xen/xs_internal.c +src/xenapi/xenapi_utils.c tools/console.c tools/virsh.c diff --git a/src/xenapi/xenapi_driver.c b/src/xenapi/xenapi_driver.c index 153582d..ad77068 100644 --- a/src/xenapi/xenapi_driver.c +++ b/src/xenapi/xenapi_driver.c @@ -102,7 +102,7 @@ xenapiOpen (virConnectPtr conn, virConnectAuthPtr auth, int flags ATTRIBUTE_UNUS } if (!passwd || !conn->uri->user) { xenapiSessionErrorHandler(conn, VIR_ERR_AUTH_FAILED, "Username/Password not valid"); - if (passwd) VIR_FREE(passwd); + VIR_FREE(passwd); return VIR_DRV_OPEN_ERROR; } if (VIR_ALLOC(privP) < 0) { -- 1.7.0.2.445.g0ae494

1 0

[libvirt] [PATCH] Use WARN_CFLAGS when compiling virsh.c
by Jiri Denemark 16 Mar '10

16 Mar '10

Signed-off-by: Jiri Denemark <jdenemar(a)redhat.com> --- tools/Makefile.am | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/tools/Makefile.am b/tools/Makefile.am index 941e93e..46107f6 100644 --- a/tools/Makefile.am +++ b/tools/Makefile.am @@ -48,6 +48,7 @@ virsh_CFLAGS = \ -I$(top_srcdir)/src/util \ -DGETTEXT_PACKAGE=\"$(PACKAGE)\" \ -DLOCALEBASEDIR=\""$(datadir)/locale"\" \ + $(WARN_CFLAGS) \ $(COVERAGE_CFLAGS) \ $(LIBXML_CFLAGS) \ $(READLINE_CFLAGS) -- 1.7.0.2

2 2

[libvirt] Fix slow storage file allocation with O_DSYNC
by Jiri Denemark 16 Mar '10

16 Mar '10

Hi. By opening storage file with O_DSYNC before allocating disk blocks for it we made the whole operation terribly slow when the allocation is done using posix_fallocate() on a filesystem which does not support fallocate(), i.e. anything but ext4. And by terribly slow I mean 45 minutes (RHEL5) to 10+ hours (Fedora Rawhide) for a 12GB storage. To fix this issue, we have two options. Either avoid using fallocate() emulation inside posix_fallocate() or stop using O_DSYNC. I prepared a patch for each of these options and they will follow as replies to this introduction. Some numbers for the same 12GB file: - libvirt's internal fallocate emulation using - mmap ~3.5 minutes (will be used whenever mmap() is present) - write ~5 minutes - fsync() instead of O_DSYNC - ~3.5 minutes So basically the two options are equivalent wrt to time consumption. The second one seems to be less hacky at the expense of non-lienar behavior which makes progress reporting less useful. Jirka

3 11

[libvirt] 'build' on FS pool now unconditionally formats?
by Cole Robinson 16 Mar '10

16 Mar '10

Hi guys, Looking at the new FS pool build options and talking with Dave, I see that calling PoolBuild on an FS pool now unconditionally calls mkfs. This is really bad when mixed with virt-manager: previously, we assumed the FS build command was always non destructive (at most it created a directory), so we called it every time, and didn't even allow users to opt out, since there wasn't a use case that called for it. This new formatting behavior really needs to be opt in, otherwise all virt-manager versions creating an FS pool can destroy data. Just FYI, for disk pools (and certain LVM configurations) where this operation has always been destructive, we default to build=off, and loudly warn the user if they choose otherwise. We can do that with this new option as well, but the previous behavior really needs to be reinstated IMO (and before the new release). I fully accept that this could be a bug in virt-manager's assumptions of the build command, but even consider a virsh user: previously build just created a directory, now it formats a partition, without any XML change. Thanks, Cole

4 11

[libvirt] XenAPI remote storage - target path
by Sharadha Prabhakar (3P) 16 Mar '10

16 Mar '10

Hi, I'm trying to write a Remote Storage driver for XenAPI. I see that target-path is used for both storage pools and volumes. In the case of XenAPI remote storage, the storage is not mounted on the local host where libvirt is running. Storage is maintained in a remote location only. In this case how do I specify target-path and how do I go about creating a VM with storage using libvirt APIs with virsh and virt-manager. Virt-manager expects me to give an absolute path for target-path. But In my case I don't have an absolute path. Can I have a target path like this for a particular storage volume "/storage pool uuid/storage-vol uuid"? This will help me identify which storage pool is Libvirt talking about and which volume in it. Using this information I can fetch data from The remote location and give it back to libvirt. Is this approach ok or does libvirt support my specific remote storage case in some way. Could someone clarify please? Regards, Sharadha

2 2

[libvirt] RFC: Add further domain event callbacks
by Daniel P. Berrange 16 Mar '10

16 Mar '10

In the current domain APIs, we currently have support for getting notified of domain lifecycle transition events. THis is done using two methods int virConnectDomainEventRegister(virConnectPtr conn, virConnectDomainEventCallback cb, void *opaque, virFreeCallback freecb); int virConnectDomainEventDeregister(virConnectPtr conn, virConnectDomainEventCallback cb); This allows an app to register a callback that looks like this typedef int (*virConnectDomainEventCallback)(virConnectPtr conn, virDomainPtr dom, int event, int detail, void *opaque); Where 'event' is the lifecycle transition (suspended, stopped, started, etc) and 'detail' is the cause of he transition (pause, migration, shutdown, etc) I have outstanding feature requests to add a lot more event notifications to the libvirt API. In particular - IO Errors. Parameter 'alias' the name of the block device with the error - Reboot. No parameters You can argue whether this should be part of the lifecycle events. On the one hand it is not guest visible, because the guest does a internal machine reset without the host ever seeing a change. This is different from other cases where QEMU itself is stopping/starting. - Watchdog. Parameter 'action', saying what is going to happen to the guest due to this watchdog firing (ignored, shutdown, paused, etc) - VNC client. In fact three events, connect, authenticated and disconnect. Parameters, TCP address & port number of client, and also of the server. Optionally a SASL username and TLS certificate name of the authenticated user - Guest user. Logon/logoff events. This requires co-operation from a guest agent, so it may not actually be of scope of libvirt. - Disk 'high watermark'. Emitted when a QCow volume grows beyond a certain physical allocation. THis allows an app to enlarge the underlying storage holding the qcow volume before an 'out of space' occurs. Parameter is the disk alias name. So we can see there are events with a wide variety of parameters and we need to figure out how to represent this in the API. Option 1 -------- Follow the existing lifecycle event model. For each new event, add a virConnectXXXXXEventRegister & virConnectEventDeregister method, and typedef a new callback for them. eg typedef int (*virConnectDomainBlockIOEventCallback)(virConnectPtr conn, virDomainPtr dom, const char *diskname, const char *alias, void *opaque); int virConnectDomainBlockIOEventRegister(virConnectPtr conn, virConnectDomainBlockIOEventCallback cb, void *opaque, virFreeCallback freecb); int virConnectDomainBlockIOEventDeregister(virConnectPtr conn, virConnectDomainEventCallback cb); So we'll have 2 extra APIs for every event + a new typedef. We'll also need to add new APIs in src/conf/domain_event.h to cope with dispatch Option 2 -------- GLib/GObject take a very loosely typed approach to registering/unregistering events. The have a single pair of methods that work for any event & a generic callback signature, requiring application casts. typedef int (*virConnectEventCallback)(void *opaque); int virConnectEventRegister(virConnectPtr conn, const char *eventname, virConnectEventCallback cb, void *opaque, virFreeCallback freecb); int virCOnnectEventUnregister(virConnectPtr conn, int eventID); In this model, the register method returns a unique integer ID for the callback which can be used to unregister it. Application's using this will still need a strongly typed callback for receiving the event, but when calling virConnectEventRegister(), the would do an explicit 'bad' cast to 'virConnectEventCallback' Option 3 -------- A hybrid of both approaches. Have a new 'register' method for each type of event that takes a strongly typed callback, but have a generic 'unregister' method that just uses the 'int eventID' int virConnectDomainBlockIOEventRegister(virConnectPtr conn, virConnectDomainBlockIOEventCallback cb, void *opaque, virFreeCallback freecb); int virCOnnectEventUnregister(virConnectPtr conn, int eventID); Option 4 -------- Have one pair of register/unregister events, but instead of passing diffeerent parameters to each callback, have a generic callback that takes a single parameter. This parameter would be declared as a union. So depending on the type of event being received, you'd access different parts of the union typedef union { virConnectDomainBlockIOEvent blockio; virConnectDomainWatchdogEvent watchdog; ...other events... } virConnectEvent; Either we could include a dummy member in the union with padding to 1024 bytes in size for future expansion, or we could simply declare that apps must never allocate this data type themselves, thus allowing us to enlarge it at will. typedef int (*virConnectEventCallback)(int eventType, virConnectEvent, void *opaque); int virConnectEventRegister(virConnectPtr conn, const char *eventname, virConnectEventCallback cb, void *opaque, virFreeCallback freecb); int virConnectEventUnregister(virConnectPtr conn, int eventID); There is one final question unrelated to these 4 options. For the lifecycle events we always registered against the 'virConnectPtr' since that is needed to capture 'domain created' events where there's no virDomainPtr to register a callback against yet. Do we want to always register all events aganist the virConnectPtr, and then pass a 'virDomainPtr' as a parameter to the callbacks as needed. Or should we allow registering events against the virDomainPtr directly. The latter might make it simpler to map libvirt into GLib/GObjects event system in the future. Daniel -- |: Red Hat, Engineering, London -o- http://people.redhat.com/berrange/ :| |: http://libvirt.org -o- http://virt-manager.org -o- http://deltacloud.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: GnuPG: 7D3B9505 -o- F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|

2 2

[libvirt] emulation failed (pagetable) rip
by Adam Mooz 15 Mar '10

15 Mar '10

Good morning list, I run a few VM's using libvirt on a C2Q 6600, have been steadily and happily since November, most of the VM's have been happily chugging away since then as well. I woke on Thursday to find one of the VM's had crashed, couldn't login and wasn't responding to the "virsh shutdown" commands. Now when I try to start it; it comes up as 'running' according to virsh but doesn't even request an IP from my DHCP server. I've enabled info logging and the log output looks like this: > Mar 14 09:52:55 Godfather kernel: [24816.021257] __ratelimit: 932597 callbacks suppressed > Mar 14 09:52:55 Godfather kernel: [24816.021262] emulation failed (pagetable) rip 7f2019b7aafb 66 0f 7f 07 > Mar 14 09:52:55 Godfather kernel: [24816.021297] emulation failed (pagetable) rip 7f2019b7aafb 66 0f 7f 07 > Mar 14 09:52:55 Godfather kernel: [24816.021321] emulation failed (pagetable) rip 7f2019b7aafb 66 0f 7f 07 > Mar 14 09:52:55 Godfather kernel: [24816.021344] emulation failed (pagetable) rip 7f2019b7aafb 66 0f 7f 07 Any ideas on fixing this? ----------------------------------------------------------------- Adam Mooz Adam.Mooz(a)gmail.com http://www.AdamMooz.com

2 2

[libvirt] [PATCH] security: Set permissions for kernel/initrd
by Cole Robinson 15 Mar '10

15 Mar '10

Fixes URL installs when running virt-install as root on Fedora. Signed-off-by: Cole Robinson <crobinso(a)redhat.com> --- src/qemu/qemu_security_dac.c | 21 +++++++++++++++++++++ src/security/security_selinux.c | 16 ++++++++++++++++ 2 files changed, 37 insertions(+), 0 deletions(-) diff --git a/src/qemu/qemu_security_dac.c b/src/qemu/qemu_security_dac.c index 6911f48..1883fbe 100644 --- a/src/qemu/qemu_security_dac.c +++ b/src/qemu/qemu_security_dac.c @@ -332,6 +332,15 @@ qemuSecurityDACRestoreSecurityAllLabel(virDomainObjPtr vm) vm->def->disks[i]) < 0) rc = -1; } + + if (vm->def->os.kernel && + qemuSecurityDACRestoreSecurityFileLabel(vm->def->os.kernel) < 0) + rc = -1; + + if (vm->def->os.initrd && + qemuSecurityDACRestoreSecurityFileLabel(vm->def->os.initrd) < 0) + rc = -1; + return rc; } @@ -356,6 +365,18 @@ qemuSecurityDACSetSecurityAllLabel(virDomainObjPtr vm) return -1; } + if (vm->def->os.kernel && + qemuSecurityDACSetOwnership(vm->def->os.kernel, + driver->user, + driver->group) < 0) + return -1; + + if (vm->def->os.initrd && + qemuSecurityDACSetOwnership(vm->def->os.initrd, + driver->user, + driver->group) < 0) + return -1; + return 0; } diff --git a/src/security/security_selinux.c b/src/security/security_selinux.c index b2c8581..975b315 100644 --- a/src/security/security_selinux.c +++ b/src/security/security_selinux.c @@ -616,6 +616,14 @@ SELinuxRestoreSecurityAllLabel(virDomainObjPtr vm) rc = -1; } + if (vm->def->os.kernel && + SELinuxRestoreSecurityFileLabel(vm->def->os.kernel) < 0) + rc = -1; + + if (vm->def->os.initrd && + SELinuxRestoreSecurityFileLabel(vm->def->os.initrd) < 0) + rc = -1; + return rc; } @@ -736,6 +744,14 @@ SELinuxSetSecurityAllLabel(virDomainObjPtr vm) return -1; } + if (vm->def->os.kernel && + SELinuxSetFilecon(vm->def->os.kernel, default_content_context) < 0) + return -1; + + if (vm->def->os.initrd && + SELinuxSetFilecon(vm->def->os.initrd, default_content_context) < 0) + return -1; + return 0; } -- 1.6.6.1

2 2

[libvirt] [PATCH] qemu: Fix USB by product with security enabled
by Cole Robinson 15 Mar '10

15 Mar '10

We need to call PrepareHostdevs to determine the USB device path before any security calls. PrepareHostUSBDevices was also incorrectly skipping all USB devices. Signed-off-by: Cole Robinson <crobinso(a)redhat.com> --- src/qemu/qemu_driver.c | 11 ++++++----- 1 files changed, 6 insertions(+), 5 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 26b5600..0fa9b2f 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -2360,7 +2360,7 @@ qemuPrepareHostUSBDevices(struct qemud_driver *driver ATTRIBUTE_UNUSED, if (hostdev->mode != VIR_DOMAIN_HOSTDEV_MODE_SUBSYS) continue; - if (hostdev->source.subsys.type != VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_PCI) + if (hostdev->source.subsys.type != VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_USB) continue; /* Resolve a vendor/product to bus/device */ @@ -2703,6 +2703,11 @@ static int qemudStartVMDaemon(virConnectPtr conn, return -1; } + DEBUG0("Preparing host devices"); + /* Must be run before security labelling */ + if (qemuPrepareHostDevices(driver, vm->def) < 0) + goto cleanup; + DEBUG0("Generating domain security label (if required)"); /* If you are using a SecurityDriver with dynamic labelling, then generate a security label for isolation */ @@ -2769,10 +2774,6 @@ static int qemudStartVMDaemon(virConnectPtr conn, if (qemuSetupCgroup(driver, vm) < 0) goto cleanup; - DEBUG0("Preparing host devices"); - if (qemuPrepareHostDevices(driver, vm->def) < 0) - goto cleanup; - if (VIR_ALLOC(priv->monConfig) < 0) { virReportOOMError(); goto cleanup; -- 1.6.6

3 3

[libvirt] compile libvirt in windows
by Luis Costa 15 Mar '10

15 Mar '10

Dear friends It would like to know as I I make to compile libvirt in the Windows? I already installed the MingGW, you I have some cake prescription? Regards, -- Luís Carlos Moreira da Costa Eclipse RAP, RCP, GMF, OSGI, Spring-DM and Pentaho Developer Regional Communities/Brazil http://wiki.eclipse.org/Regional_Communities/Brazil Java, write once, run everywhere™

2 1

[libvirt] libvirt modifying iptables rules
by Varrun Ramani 15 Mar '10

15 Mar '10

Hi I am right now undertaking a project which deals with verification of firewall rules. I wish to know which applications/libraries modify/query firewall rules. I came to know that libvirt modifies iptables rules. Can anyone let me know for what purposes/how libvirt modifies the rules? Thanks & Regards Varrun Ramani

4 3

[libvirt] [PATCH] qemu: pass the information when disks are read-only
by Daniel Veillard 15 Mar '10

15 Mar '10

The problem is that even when the user indicates that a drive has to be opened as read-only (for example a cd-rom) qemu didn't get the information and tried to open read-write. This could get really nasty in case SELinux is monitoring that no read-write access is attempted: https://bugzilla.redhat.com/show_bug.cgi?id=536760 Patch is relatively simple and based on teh assumption that if the -device qemu options are allowed then qemu also knows about the readonly option (this was added to the help string quite later so we should not rely on that in this case). * src/qemu/qemu_conf.c: add the ",readonly=on" for read-only disks and also parse it back in qemuParseCommandLineDisk() * tests/qemuxml2argvtest.c tests/qemuxml2argvdata/qemuxml2argv-disk-drive-readonly-disk.args tests/qemuxml2argvdata/qemuxml2argv-disk-drive-readonly-disk.xml: add a specific regression test diff --git a/src/qemu/qemu_conf.c b/src/qemu/qemu_conf.c index 40ca221..fb23c52 100644 --- a/src/qemu/qemu_conf.c +++ b/src/qemu/qemu_conf.c @@ -2398,6 +2398,9 @@ qemuBuildDriveStr(virDomainDiskDefPtr disk, if (bootable && disk->device == VIR_DOMAIN_DISK_DEVICE_DISK) virBufferAddLit(&opt, ",boot=on"); + if (disk->readonly && + qemuCmdFlags & QEMUD_CMD_FLAG_DEVICE) + virBufferAddLit(&opt, ",readonly=on"); if (disk->driverType && disk->type != VIR_DOMAIN_DISK_TYPE_DIR && qemuCmdFlags & QEMUD_CMD_FLAG_DRIVE_FORMAT) @@ -4694,6 +4697,9 @@ qemuParseCommandLineDisk(const char *val, _("cannot parse drive unit '%s'"), val); goto cleanup; } + } else if (STREQ(keywords[i], "readonly")) { + if ((values[i] == NULL) || STREQ(values[i], "on")) + def->readonly = 1; } } diff --git a/tests/qemuxml2argvdata/qemuxml2argv-disk-drive-readonly-disk.args b/tests/qemuxml2argvdata/qemuxml2argv-disk-drive-readonly-disk.args new file mode 100644 index 0000000..fbc3226 --- /dev/null +++ b/tests/qemuxml2argvdata/qemuxml2argv-disk-drive-readonly-disk.args @@ -0,0 +1 @@ +LC_ALL=C PATH=/bin HOME=/home/test USER=test LOGNAME=test /usr/bin/qemu -S -M pc -m 214 -smp 1 -nographic -nodefaults -monitor unix:/tmp/test-monitor,server,nowait -no-acpi -boot c -drive file=/dev/HostVG/QEMUGuest1,if=none,id=drive-ide0-0-0 -device ide-drive,bus=ide.0,unit=0,drive=drive-ide0-0-0,id=ide0-0-0 -drive file=/dev/sr0,if=none,media=cdrom,id=drive-ide0-1-0,readonly=on -device ide-drive,bus=ide.1,unit=0,drive=drive-ide0-1-0,id=ide0-1-0 -usb -device virtio-balloon-pci,id=balloon0,bus=pci.0,addr=0x3 diff --git a/tests/qemuxml2argvdata/qemuxml2argv-disk-drive-readonly-disk.xml b/tests/qemuxml2argvdata/qemuxml2argv-disk-drive-readonly-disk.xml new file mode 100644 index 0000000..39c3a1c --- /dev/null +++ b/tests/qemuxml2argvdata/qemuxml2argv-disk-drive-readonly-disk.xml @@ -0,0 +1,30 @@ +<domain type='qemu'> + <name>QEMUGuest1</name> + <uuid>c7a5fdbd-edaf-9455-926a-d65c16db1809</uuid> + <memory>219200</memory> + <currentMemory>219200</currentMemory> + <vcpu>1</vcpu> + <os> + <type arch='i686' machine='pc'>hvm</type> + <boot dev='hd'/> + </os> + <clock offset='utc'/> + <on_poweroff>destroy</on_poweroff> + <on_reboot>restart</on_reboot> + <on_crash>destroy</on_crash> + <devices> + <emulator>/usr/bin/qemu</emulator> + <disk type='block' device='disk'> + <source dev='/dev/HostVG/QEMUGuest1'/> + <target dev='hda' bus='ide'/> + <address type='drive' controller='0' bus='0' unit='0'/> + </disk> + <disk type='block' device='cdrom'> + <source dev='/dev/sr0'/> + <target dev='hdc' bus='ide'/> + <readonly/> + <address type='drive' controller='0' bus='1' unit='0'/> + </disk> + <controller type='ide' index='0'/> + </devices> +</domain> diff --git a/tests/qemuxml2argvtest.c b/tests/qemuxml2argvtest.c index 4ca946a..e3762c9 100644 --- a/tests/qemuxml2argvtest.c +++ b/tests/qemuxml2argvtest.c @@ -249,6 +249,8 @@ mymain(int argc, char **argv) QEMUD_CMD_FLAG_DRIVE_BOOT | QEMUD_CMD_FLAG_DRIVE_FORMAT); DO_TEST("disk-drive-fat", QEMUD_CMD_FLAG_DRIVE | QEMUD_CMD_FLAG_DRIVE_BOOT | QEMUD_CMD_FLAG_DRIVE_FORMAT); + DO_TEST("disk-drive-readonly-disk", QEMUD_CMD_FLAG_DRIVE | + QEMUD_CMD_FLAG_DEVICE); DO_TEST("disk-drive-fmt-qcow", QEMUD_CMD_FLAG_DRIVE | QEMUD_CMD_FLAG_DRIVE_BOOT | QEMUD_CMD_FLAG_DRIVE_FORMAT); DO_TEST("disk-drive-shared", QEMUD_CMD_FLAG_DRIVE | -- 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/

2 2