April 2009 - Devel - libvirt List Archives

[libvirt] qemu+tls server certificate validation failure (The certificate is not trusted)
by Scott Beardsley 28 Apr '09

28 Apr '09

I'm having a problem with remote TLS libvirt connections from an Ubuntu Jaunty client. I've reported the bug here[1] but haven't had any hits yet so I thought I'd come to the source. Let me know if ya'll have any ideas or know of any bugs in the versions I'm using (see below). I just upgraded my client to Jaunty from Intrepid and I can no longer connect to Hardy or Intrepid libvirt servers that have TLS enabled. I get the following errors: $ virt-viewer -c qemu+tls://example.com/system virt.example.com libvir: Remote error : server certificate failed validation: The certificate is not trusted. libvir: Remote error : unable to connect to 'example.com': Invalid argument unable to connect to libvirt qemu+tls://example.com/system $ In the past (ie hardy, intrepid) I was able to use the following command. Now I get an error: $ virt-viewer -c qemu://example.com/system virt.example.com libvir: error : could not connect to qemu://example.com/system unable to connect to libvirt qemu://example.com/system $ The server's config has not changed (I've tested against libvirt-bin versions 0.4.4-3ubuntu3.1 and 0.4.0-2ubuntu8.1 on the server side). I have the CA certificate installed on both server and client (in /etc/pki/CA/cacert.pem). That cert signed both my x509 client cert and the server cert. Here is some proof that it *should* work: $ openssl s_client -CAfile /etc/pki/CA/cacert.pem -cert /etc/pki/libvirt/clientcert.pem -key /etc/pki/libvirt/private/clientkey.pem -connect example.com:16514 2>/dev/null|sed -ne '/-BEGIN CERTIFICATE-/,/-END CERTIFICATE-/p' >servercert.pem $ openssl verify -CAfile /etc/pki/CA/cacert.pem servercert.pem servercert.pem: OK $ openssl verify -CAfile /etc/pki/CA/cacert.pem /etc/pki/libvirt/clientcert.pem /etc/pki/libvirt/clientcert.pem: OK $ When I run strace against virt-viewer I can see that it is accessing and (successfully opening) the correct certs/keys: $ grep /etc/pki /tmp/out stat64("/etc/pki/CA/cacert.pem", {st_mode=S_IFREG|0644, st_size=1716, ...}) = 0 stat64("/etc/pki/libvirt/private/clientkey.pem", {st_mode=S_IFREG|0644, st_size=887, ...}) = 0 stat64("/etc/pki/libvirt/clientcert.pem", {st_mode=S_IFREG|0644, st_size=1172, ...}) = 0 open("/etc/pki/CA/cacert.pem", O_RDONLY) = 5 open("/etc/pki/libvirt/private/clientkey.pem", O_RDONLY) = 5 open("/etc/pki/libvirt/clientcert.pem", O_RDONLY) = 5 $ Thanks in advance, Scott ------------ [1] https://bugs.launchpad.net/ubuntu/+source/libvirt/+bug/366455

2 3

[libvirt] PATCH: Fix compiler flag check for -Wformat-security
by Daniel P. Berrange 27 Apr '09

27 Apr '09

Certain versions of GCC will ignore the -Wformat-security flag unless you have also set the -Wformat flag. Unfortunately when checking flags, we only check one at a time in isolation, so it was thinking -Wformat-security is not supported in Fedora. The fix is to include all previously verified flags during the check process, but not for -Werror because for added fun, the autoconf test program itself causes compile warnings with some flags we have. The patch also fixes 3 harmless format problems in virsh Daniel Index: m4/compiler-flags.m4 =================================================================== RCS file: /data/cvs/libvirt/m4/compiler-flags.m4,v retrieving revision 1.2 diff -u -p -r1.2 compiler-flags.m4 --- m4/compiler-flags.m4 14 Jan 2009 15:23:26 -0000 1.2 +++ m4/compiler-flags.m4 27 Apr 2009 10:34:43 -0000 @@ -24,7 +24,15 @@ AC_DEFUN([gl_COMPILER_FLAGS], [AC_MSG_CHECKING(whether compiler accepts $1) AC_SUBST(COMPILER_FLAGS) ac_save_CFLAGS="$CFLAGS" - CFLAGS="$CFLAGS $1" + dnl Some flags are dependant, so we set all previously checked + dnl flags when testing. Except for -Werror which we have to + dnl check on its own, because some of our compiler flags cause + dnl warnings from the autoconf test program! + if test "$1" = "-Werror" ; then + CFLAGS="$CFLAGS $1" + else + CFLAGS="$CFLAGS $COMPILER_FLAGS $1" + fi AC_TRY_LINK([], [], has_option=yes, has_option=no,) echo 'int x;' >conftest.c $CC $CFLAGS -c conftest.c 2>conftest.err Index: src/virsh.c =================================================================== RCS file: /data/cvs/libvirt/src/virsh.c,v retrieving revision 1.201 diff -u -p -r1.201 virsh.c --- src/virsh.c 15 Apr 2009 20:09:09 -0000 1.201 +++ src/virsh.c 27 Apr 2009 10:34:45 -0000 @@ -4463,7 +4463,7 @@ cmdNodeListDevicesPrint(vshControl *ctl, } /* Print this device */ - vshPrint(ctl, indentBuf); + vshPrint(ctl, "%s", indentBuf); vshPrint(ctl, "%s\n", devices[devid]); @@ -4487,7 +4487,7 @@ cmdNodeListDevicesPrint(vshControl *ctl, /* If there is a child device, then print another blank line */ if (nextlastdev != -1) { - vshPrint(ctl, indentBuf); + vshPrint(ctl, "%s", indentBuf); vshPrint(ctl, " |\n"); } @@ -4511,7 +4511,7 @@ cmdNodeListDevicesPrint(vshControl *ctl, /* If there was no child device, and we're the last in * a list of devices, then print another blank line */ if (nextlastdev == -1 && devid == lastdev) { - vshPrint(ctl, indentBuf); + vshPrint(ctl, "%s", indentBuf); vshPrint(ctl, "\n"); } } -- |: Red Hat, Engineering, London -o- http://people.redhat.com/berrange/ :| |: http://libvirt.org -o- http://virt-manager.org -o- http://ovirt.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: GnuPG: 7D3B9505 -o- F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|

3 4

[libvirt] libvirt Driver support Matrix.
by Abhishek K Jha 27 Apr '09

27 Apr '09

Hi, I needed to know if libvirt supports migration via the QEMU/KVM driver. ( virDomainMigrate) Could someone point me to the latest driver support matrix for libvirt. It seems the one at http://libvirt.org/hvsupport.html was last updated on 2008-06-05. Appreciate your help! Thanks much, Abhishek Jha. IBM India Software Labs, Pune 91-20-66592139(Direct) 91-20-26982424 Extn: 2139

5 6

[libvirt] 802.1q vlan-tagging support for libvirt
by Klaus Heinrich Kiwi 27 Apr '09

27 Apr '09

Are there any plans to support configuring 802.1q VLAN tagging through libvirt? My understanding is that libvirt would use vconfig to create tagged interfaces, while using a physical interface as trunk (e.g., eth0 is the trunk, eth0.20 the interface with the '20' vlan tag). Then it would add the tagged interface (eth0.20) to a bridge with the guest virtual interface. I don't know how libvirt would need to be extended to be able to properly 'manage' those 802.1q-enabled virtual switches. Another question would be how libvirt-CIM would support this. Is anyone interested/working on this? Thanks, -Klaus -- Klaus Heinrich Kiwi <klausk(a)linux.vnet.ibm.com> Linux Security Development, IBM Linux Technology Center

7 11

[libvirt] Roadmap, beyond 0.6.3
by Daniel Veillard 25 Apr '09

25 Apr '09

Okay now that one is out, let's see what's next: First to try to keep up with monthly releases more or less, I would suggest a new release around May 25, which would mean entering code feature freeze around May 18, so that's nearly 4 weeks for development. I would assume it's 0.6.4 unless something major lands worth indicating a new level. There is a number of things we should try to do by then: - the review and inclusion of OpenNebula driver - the review and inclusion of Power Hypervisor driver - the netcf based networking setup APIs, hopefully this could help on the frequent networking support questions - finising NPIV now that the interface for node storage create and destroy is part of the API, maybe other storage devices could be plugged too - improve the documentation, it's something we (i.e. Red Hat) are commited to do, but it need to be a full part of the project too, rewiewing and increasing documentation will be helpful. - work on the test suite and augment it, Dan Berrange posted about libvirt-TCK framework, help augmenting it is welcome http://www.mail-archive.com/libvir-list@redhat.com/msg12703.html http://libvirt.org/hg/libvirt-tck - and there is number of small fixes and buglet we would need to find and fix but hopefully 0.6.3 is good enough we won't spend too much time chasing new problems ! If we could get most of this in 4 weeks time that would be impressive, but let's work on it :-) Daniel -- Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ daniel(a)veillard.com | Rpmfind RPM search engine http://rpmfind.net/ http://veillard.com/ | virtualization library http://libvirt.org/

1 0

[libvirt] Release of libvirt-0.6.3
by Daniel Veillard 24 Apr '09

24 Apr '09

As planned it has landed on the server at: ftp://libvirt.org/libvirt/ The main points are the VirtualBox driver and a number of bug fixes: * New features: - VirtualBox driver support (Pritesh Kothari) - virt-xml-validate new command (Daniel Berrange) * Portability: - patch to build on Centos (Joseph Shraibman) - build breakage (Anton Protopopov) * Documentation: - Linux Containers documentation (Serge Hallyn) - improvement and updates of architecture pages - fix virNodeGetFreeMemory documentation to reflect reality - man page cleanups (Daniel Berrange) - man page typo (Robert P. J. Day) - VirtualBox Documentation (Pritesh Kothari) * Bug fixes: - veth off-by-one error (Dan Smith) - vcpupin to inactive Xen crash (Takahashi Tomohiro) - virsh ttyconsole return value - use format= not fmt= on QEmu commandline (Mark McLoughlin) - use UUID for internal domain lookups (Daniel Berrange) - remote domain ID related bugs (Daniel Berrange) - QEmu pidfile handling bugs (Daniel Berrange) - network config handling on old Xen (Daniel Berrange) * Improvements: - add SCSI storage rescan (David Allan) - rootless LXC containers support improvements (Serge Hallyn) - getHostname support for LXC (Dan Smith) - cleanup and logging output of some domain functions (Guido Gunther) - drop pool lock when allocating volumes (Cole Robinson) - LXC handle kernel without CLONE_NEWUSER support (Serge Hallyn) - cpu pinning on defined Xen domains (Takahashi Tomohiro) - dynamic bridge names support (Soren Hansen) - LXC use of private /dev/pts when available (Daniel Berrange) - virNodeDeviceCreateXML and virNodeDeviceDestroy entry points (Dave Allan) * Cleanups: - don't hardcode getgrnam_r buffer to 1024 bytes (Guido Gunther) - qemudBuildCommandLine API cleanup (Daniel Berrange) Thanks to everybody who helped with the release, and welcome to the peple who joined the project recently ! Daniel -- Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ daniel(a)veillard.com | Rpmfind RPM search engine http://rpmfind.net/ http://veillard.com/ | virtualization library http://libvirt.org/

1 0

[libvirt] PATCH: Public API plumbing for virNodeDeviceCreateXML/Destroy
by Daniel P. Berrange 24 Apr '09

24 Apr '09

As previously discussed, Dave Allen's reworking the NPIV patches to do creation/deleation of vports via the node device APIs. Thus, this patch adds the 2 new APIs required for this approach. Daniel diff -r 2d1278bdf31f include/libvirt/libvirt.h --- a/include/libvirt/libvirt.h Fri Apr 24 11:01:11 2009 +0100 +++ b/include/libvirt/libvirt.h Fri Apr 24 13:06:12 2009 +0100 @@ -1124,6 +1124,12 @@ int virNodeDeviceDet int virNodeDeviceReAttach (virNodeDevicePtr dev); int virNodeDeviceReset (virNodeDevicePtr dev); +virNodeDevicePtr virNodeDeviceCreateXML (virConnectPtr conn, + const char *xmlDesc, + unsigned int flags); + +int virNodeDeviceDestroy (virNodeDevicePtr dev); + /* * Domain Event Notification */ diff -r 2d1278bdf31f include/libvirt/libvirt.h.in --- a/include/libvirt/libvirt.h.in Fri Apr 24 11:01:11 2009 +0100 +++ b/include/libvirt/libvirt.h.in Fri Apr 24 13:06:12 2009 +0100 @@ -1124,6 +1124,12 @@ int virNodeDeviceDet int virNodeDeviceReAttach (virNodeDevicePtr dev); int virNodeDeviceReset (virNodeDevicePtr dev); +virNodeDevicePtr virNodeDeviceCreateXML (virConnectPtr conn, + const char *xmlDesc, + unsigned int flags); + +int virNodeDeviceDestroy (virNodeDevicePtr dev); + /* * Domain Event Notification */ diff -r 2d1278bdf31f qemud/remote.c --- a/qemud/remote.c Fri Apr 24 11:01:11 2009 +0100 +++ b/qemud/remote.c Fri Apr 24 13:06:12 2009 +0100 @@ -4323,6 +4323,54 @@ remoteDispatchNodeDeviceReset (struct qe } +static int +remoteDispatchNodeDeviceCreateXml(struct qemud_server *server ATTRIBUTE_UNUSED, + struct qemud_client *client ATTRIBUTE_UNUSED, + virConnectPtr conn, + remote_error *rerr, + remote_node_device_create_xml_args *args, + remote_node_device_create_xml_ret *ret) +{ + virNodeDevicePtr dev; + + dev = virNodeDeviceCreateXML (conn, args->xml_desc, args->flags); + if (dev == NULL) { + remoteDispatchConnError(rerr, conn); + return -1; + } + + make_nonnull_node_device (&ret->dev, dev); + virNodeDeviceFree(dev); + + return 0; +} + + +static int +remoteDispatchNodeDeviceDestroy(struct qemud_server *server ATTRIBUTE_UNUSED, + struct qemud_client *client ATTRIBUTE_UNUSED, + virConnectPtr conn, + remote_error *rerr, + remote_node_device_destroy_args *args, + void *ret ATTRIBUTE_UNUSED) +{ + virNodeDevicePtr dev; + + dev = virNodeDeviceLookupByName(conn, args->name); + if (dev == NULL) { + remoteDispatchFormatError(rerr, "%s", _("node_device not found")); + return -1; + } + + if (virNodeDeviceDestroy(dev) == -1) { + remoteDispatchConnError(rerr, conn); + return -1; + } + + return 0; +} + + /************************** * Async Events **************************/ diff -r 2d1278bdf31f qemud/remote_dispatch_args.h --- a/qemud/remote_dispatch_args.h Fri Apr 24 11:01:11 2009 +0100 +++ b/qemud/remote_dispatch_args.h Fri Apr 24 13:06:12 2009 +0100 @@ -103,3 +103,5 @@ remote_node_device_re_attach_args val_remote_node_device_re_attach_args; remote_node_device_reset_args val_remote_node_device_reset_args; remote_domain_get_security_label_args val_remote_domain_get_security_label_args; + remote_node_device_create_xml_args val_remote_node_device_create_xml_args; + remote_node_device_destroy_args val_remote_node_device_destroy_args; diff -r 2d1278bdf31f qemud/remote_dispatch_prototypes.h --- a/qemud/remote_dispatch_prototypes.h Fri Apr 24 11:01:11 2009 +0100 +++ b/qemud/remote_dispatch_prototypes.h Fri Apr 24 13:06:12 2009 +0100 @@ -527,6 +527,20 @@ static int remoteDispatchNetworkUndefine remote_error *err, remote_network_undefine_args *args, void *ret); +static int remoteDispatchNodeDeviceCreateXml( + struct qemud_server *server, + struct qemud_client *client, + virConnectPtr conn, + remote_error *err, + remote_node_device_create_xml_args *args, + remote_node_device_create_xml_ret *ret); +static int remoteDispatchNodeDeviceDestroy( + struct qemud_server *server, + struct qemud_client *client, + virConnectPtr conn, + remote_error *err, + remote_node_device_destroy_args *args, + void *ret); static int remoteDispatchNodeDeviceDettach( struct qemud_server *server, struct qemud_client *client, diff -r 2d1278bdf31f qemud/remote_dispatch_ret.h --- a/qemud/remote_dispatch_ret.h Fri Apr 24 11:01:11 2009 +0100 +++ b/qemud/remote_dispatch_ret.h Fri Apr 24 13:06:12 2009 +0100 @@ -88,3 +88,4 @@ remote_node_device_list_caps_ret val_remote_node_device_list_caps_ret; remote_domain_get_security_label_ret val_remote_domain_get_security_label_ret; remote_node_get_security_model_ret val_remote_node_get_security_model_ret; + remote_node_device_create_xml_ret val_remote_node_device_create_xml_ret; diff -r 2d1278bdf31f qemud/remote_dispatch_table.h --- a/qemud/remote_dispatch_table.h Fri Apr 24 11:01:11 2009 +0100 +++ b/qemud/remote_dispatch_table.h Fri Apr 24 13:06:12 2009 +0100 @@ -607,13 +607,23 @@ .args_filter = (xdrproc_t) xdr_remote_node_device_reset_args, .ret_filter = (xdrproc_t) xdr_void, }, -{ /* DomainGetSecurityLabel => 118 */ +{ /* DomainGetSecurityLabel => 121 */ .fn = (dispatch_fn) remoteDispatchDomainGetSecurityLabel, .args_filter = (xdrproc_t) xdr_remote_domain_get_security_label_args, .ret_filter = (xdrproc_t) xdr_remote_domain_get_security_label_ret, }, -{ /* NodeGetSecurityModel => 119 */ +{ /* NodeGetSecurityModel => 122 */ .fn = (dispatch_fn) remoteDispatchNodeGetSecurityModel, .args_filter = (xdrproc_t) xdr_void, .ret_filter = (xdrproc_t) xdr_remote_node_get_security_model_ret, }, +{ /* NodeDeviceCreateXml => 123 */ + .fn = (dispatch_fn) remoteDispatchNodeDeviceCreateXml, + .args_filter = (xdrproc_t) xdr_remote_node_device_create_xml_args, + .ret_filter = (xdrproc_t) xdr_remote_node_device_create_xml_ret, +}, +{ /* NodeDeviceDestroy => 124 */ + .fn = (dispatch_fn) remoteDispatchNodeDeviceDestroy, + .args_filter = (xdrproc_t) xdr_remote_node_device_destroy_args, + .ret_filter = (xdrproc_t) xdr_void, +}, diff -r 2d1278bdf31f qemud/remote_protocol.c --- a/qemud/remote_protocol.c Fri Apr 24 11:01:11 2009 +0100 +++ b/qemud/remote_protocol.c Fri Apr 24 13:06:12 2009 +0100 @@ -2230,6 +2230,35 @@ xdr_remote_node_device_reset_args (XDR * } bool_t +xdr_remote_node_device_create_xml_args (XDR *xdrs, remote_node_device_create_xml_args *objp) +{ + + if (!xdr_remote_nonnull_string (xdrs, &objp->xml_desc)) + return FALSE; + if (!xdr_int (xdrs, &objp->flags)) + return FALSE; + return TRUE; +} + +bool_t +xdr_remote_node_device_create_xml_ret (XDR *xdrs, remote_node_device_create_xml_ret *objp) +{ + + if (!xdr_remote_nonnull_node_device (xdrs, &objp->dev)) + return FALSE; + return TRUE; +} + +bool_t +xdr_remote_node_device_destroy_args (XDR *xdrs, remote_node_device_destroy_args *objp) +{ + + if (!xdr_remote_nonnull_string (xdrs, &objp->name)) + return FALSE; + return TRUE; +} + +bool_t xdr_remote_domain_events_register_ret (XDR *xdrs, remote_domain_events_register_ret *objp) { diff -r 2d1278bdf31f qemud/remote_protocol.h --- a/qemud/remote_protocol.h Fri Apr 24 11:01:11 2009 +0100 +++ b/qemud/remote_protocol.h Fri Apr 24 13:06:12 2009 +0100 @@ -1255,6 +1255,22 @@ struct remote_node_device_reset_args { }; typedef struct remote_node_device_reset_args remote_node_device_reset_args; +struct remote_node_device_create_xml_args { + remote_nonnull_string xml_desc; + int flags; +}; +typedef struct remote_node_device_create_xml_args remote_node_device_create_xml_args; + +struct remote_node_device_create_xml_ret { + remote_nonnull_node_device dev; +}; +typedef struct remote_node_device_create_xml_ret remote_node_device_create_xml_ret; + +struct remote_node_device_destroy_args { + remote_nonnull_string name; +}; +typedef struct remote_node_device_destroy_args remote_node_device_destroy_args; + struct remote_domain_events_register_ret { int cb_registered; }; @@ -1397,6 +1413,8 @@ enum remote_procedure { REMOTE_PROC_NODE_DEVICE_RESET = 120, REMOTE_PROC_DOMAIN_GET_SECURITY_LABEL = 121, REMOTE_PROC_NODE_GET_SECURITY_MODEL = 122, + REMOTE_PROC_NODE_DEVICE_CREATE_XML = 123, + REMOTE_PROC_NODE_DEVICE_DESTROY = 124, }; typedef enum remote_procedure remote_procedure; @@ -1629,6 +1647,9 @@ extern bool_t xdr_remote_node_device_li extern bool_t xdr_remote_node_device_dettach_args (XDR *, remote_node_device_dettach_args*); extern bool_t xdr_remote_node_device_re_attach_args (XDR *, remote_node_device_re_attach_args*); extern bool_t xdr_remote_node_device_reset_args (XDR *, remote_node_device_reset_args*); +extern bool_t xdr_remote_node_device_create_xml_args (XDR *, remote_node_device_create_xml_args*); +extern bool_t xdr_remote_node_device_create_xml_ret (XDR *, remote_node_device_create_xml_ret*); +extern bool_t xdr_remote_node_device_destroy_args (XDR *, remote_node_device_destroy_args*); extern bool_t xdr_remote_domain_events_register_ret (XDR *, remote_domain_events_register_ret*); extern bool_t xdr_remote_domain_events_deregister_ret (XDR *, remote_domain_events_deregister_ret*); extern bool_t xdr_remote_domain_event_ret (XDR *, remote_domain_event_ret*); @@ -1840,6 +1861,9 @@ extern bool_t xdr_remote_node_device_lis extern bool_t xdr_remote_node_device_dettach_args (); extern bool_t xdr_remote_node_device_re_attach_args (); extern bool_t xdr_remote_node_device_reset_args (); +extern bool_t xdr_remote_node_device_create_xml_args (); +extern bool_t xdr_remote_node_device_create_xml_ret (); +extern bool_t xdr_remote_node_device_destroy_args (); extern bool_t xdr_remote_domain_events_register_ret (); extern bool_t xdr_remote_domain_events_deregister_ret (); extern bool_t xdr_remote_domain_event_ret (); diff -r 2d1278bdf31f qemud/remote_protocol.x --- a/qemud/remote_protocol.x Fri Apr 24 11:01:11 2009 +0100 +++ b/qemud/remote_protocol.x Fri Apr 24 13:06:12 2009 +0100 @@ -1109,6 +1109,19 @@ struct remote_node_device_reset_args { remote_nonnull_string name; }; +struct remote_node_device_create_xml_args { + remote_nonnull_string xml_desc; + int flags; +}; + +struct remote_node_device_create_xml_ret { + remote_nonnull_node_device dev; +}; + +struct remote_node_device_destroy_args { + remote_nonnull_string name; +}; + /** * Events Register/Deregister: @@ -1270,7 +1283,10 @@ enum remote_procedure { REMOTE_PROC_NODE_DEVICE_RESET = 120, REMOTE_PROC_DOMAIN_GET_SECURITY_LABEL = 121, - REMOTE_PROC_NODE_GET_SECURITY_MODEL = 122 + REMOTE_PROC_NODE_GET_SECURITY_MODEL = 122, + + REMOTE_PROC_NODE_DEVICE_CREATE_XML = 123, + REMOTE_PROC_NODE_DEVICE_DESTROY = 124 }; /* Custom RPC structure. */ diff -r 2d1278bdf31f src/driver.h --- a/src/driver.h Fri Apr 24 11:01:11 2009 +0100 +++ b/src/driver.h Fri Apr 24 13:06:12 2009 +0100 @@ -684,6 +684,11 @@ typedef int (*virDevMonDeviceListCaps)(v char **const names, int maxnames); +typedef virNodeDevicePtr (*virDrvNodeDeviceCreateXML)(virConnectPtr conn, + const char *xmlDesc, + unsigned int flags); +typedef int (*virDrvNodeDeviceDestroy)(virNodeDevicePtr dev); + /** * _virDeviceMonitor: * @@ -702,6 +707,8 @@ struct _virDeviceMonitor { virDevMonDeviceGetParent deviceGetParent; virDevMonDeviceNumOfCaps deviceNumOfCaps; virDevMonDeviceListCaps deviceListCaps; + virDrvNodeDeviceCreateXML deviceCreateXML; + virDrvNodeDeviceDestroy deviceDestroy; }; /* diff -r 2d1278bdf31f src/libvirt.c --- a/src/libvirt.c Fri Apr 24 11:01:11 2009 +0100 +++ b/src/libvirt.c Fri Apr 24 13:06:12 2009 +0100 @@ -7491,6 +7491,103 @@ error: } +/** + * virNodeDeviceCreateXML: + * @conn: pointer to the hypervisor connection + * @xmlDesc: string containing an XML description of the device to be created + * @flags: callers should always pass 0 + * + * Create a new device on the VM host machine, for example, virtual + * HBAs created using vport_create. + * + * Returns a node device object if successful, NULL in case of failure + */ +virNodeDevicePtr +virNodeDeviceCreateXML(virConnectPtr conn, + const char *xmlDesc, + unsigned int flags) +{ + VIR_DEBUG("conn=%p, xmlDesc=%s, flags=%d", conn, xmlDesc, flags); + + virResetLastError(); + + if (!VIR_IS_CONNECT(conn)) { + virLibConnError(NULL, VIR_ERR_INVALID_CONN, __FUNCTION__); + return NULL; + } + + if (conn->flags & VIR_CONNECT_RO) { + virLibConnError(conn, VIR_ERR_OPERATION_DENIED, __FUNCTION__); + goto error; + } + + if (xmlDesc == NULL) { + virLibConnError(conn, VIR_ERR_INVALID_ARG, __FUNCTION__); + goto error; + } + + if (conn->deviceMonitor->deviceCreateXML) { + virNodeDevicePtr dev = conn->deviceMonitor->deviceCreateXML(conn, xmlDesc, flags); + if (dev == NULL) + goto error; + return dev; + } + + virLibConnError (conn, VIR_ERR_NO_SUPPORT, __FUNCTION__); + +error: + /* Copy to connection error object for back compatability */ + virSetConnError(conn); + return NULL; +} + + +/** + * virNodeDeviceDestroy: + * @dev: a device object + * + * Destroy the device object. The virtual device is removed from the host operating system. + * This function may require privileged access + * + * Returns 0 in case of success and -1 in case of failure. + */ +int +virNodeDeviceDestroy(virNodeDevicePtr dev) +{ + int retval = 0; + + DEBUG("dev=%p", dev); + + virResetLastError(); + + if (!VIR_IS_CONNECTED_NODE_DEVICE(dev)) { + virLibNodeDeviceError(NULL, VIR_ERR_INVALID_NODE_DEVICE, __FUNCTION__); + return (-1); + } + + if (dev->conn->flags & VIR_CONNECT_RO) { + virLibConnError(dev->conn, VIR_ERR_OPERATION_DENIED, __FUNCTION__); + goto error; + } + + if (dev->conn->deviceMonitor->deviceDestroy) { + retval = dev->conn->deviceMonitor->deviceDestroy(dev); + if (retval < 0) { + goto error; + } + + return 0; + } + + virLibConnError (dev->conn, VIR_ERR_NO_SUPPORT, __FUNCTION__); + +error: + /* Copy to connection error object for back compatability */ + virSetConnError(dev->conn); + return -1; +} + + /* * Domain Event Notification */ diff -r 2d1278bdf31f src/libvirt_public.syms --- a/src/libvirt_public.syms Fri Apr 24 11:01:11 2009 +0100 +++ b/src/libvirt_public.syms Fri Apr 24 13:06:12 2009 +0100 @@ -258,4 +258,10 @@ LIBVIRT_0.6.1 { virNodeGetSecurityModel; } LIBVIRT_0.6.0; +LIBVIRT_0.6.3 { + global: + virNodeDeviceCreateXML; + virNodeDeviceDestroy; +} LIBVIRT_0.6.1; + # .... define new API here using predicted next version number .... diff -r 2d1278bdf31f src/remote_internal.c --- a/src/remote_internal.c Fri Apr 24 11:01:11 2009 +0100 +++ b/src/remote_internal.c Fri Apr 24 13:06:12 2009 +0100 @@ -4987,6 +4987,59 @@ done: } +static virNodeDevicePtr +remoteNodeDeviceCreateXML(virConnectPtr conn, + const char *xmlDesc, + unsigned int flags) +{ + remote_node_device_create_xml_args args; + remote_node_device_create_xml_ret ret; + virNodeDevicePtr dev = NULL; + struct private_data *priv = conn->privateData; + + remoteDriverLock(priv); + + memset(&ret, 0, sizeof ret); + args.xml_desc = (char *)xmlDesc; + args.flags = flags; + + if (call(conn, priv, 0, REMOTE_PROC_NODE_DEVICE_CREATE_XML, + (xdrproc_t) xdr_remote_node_device_create_xml_args, (char *) &args, + (xdrproc_t) xdr_remote_node_device_create_xml_ret, (char *) &ret) == -1) + goto done; + + dev = get_nonnull_node_device(conn, ret.dev); + xdr_free ((xdrproc_t) xdr_remote_node_device_create_xml_ret, (char *) &ret); + +done: + remoteDriverUnlock(priv); + return dev; +} + +static int +remoteNodeDeviceDestroy(virNodeDevicePtr dev) +{ + int rv = -1; + remote_node_device_destroy_args args; + struct private_data *priv = dev->conn->privateData; + + remoteDriverLock(priv); + + args.name = dev->name; + + if (call(dev->conn, priv, 0, REMOTE_PROC_NODE_DEVICE_RESET, + (xdrproc_t) xdr_remote_node_device_destroy_args, (char *) &args, + (xdrproc_t) xdr_void, (char *) NULL) == -1) + goto done; + + rv = 0; + +done: + remoteDriverUnlock(priv); + return rv; +} + + /*----------------------------------------------------------------------*/ static int @@ -6991,6 +7044,8 @@ static virDeviceMonitor dev_monitor = { .deviceGetParent = remoteNodeDeviceGetParent, .deviceNumOfCaps = remoteNodeDeviceNumOfCaps, .deviceListCaps = remoteNodeDeviceListCaps, + .deviceCreateXML = remoteNodeDeviceCreateXML, + .deviceDestroy = remoteNodeDeviceDestroy }; -- |: Red Hat, Engineering, London -o- http://people.redhat.com/berrange/ :| |: http://libvirt.org -o- http://virt-manager.org -o- http://ovirt.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: GnuPG: 7D3B9505 -o- F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|

3 4

[libvirt] Xen network interface behavior with 0.6.2
by Kaitlin Rupert 24 Apr '09

24 Apr '09

Hi, I'm using libvirt 0.6.2 to create a Xen guest with a network type interface (see XML below). When the guest is defined, the interface is converted to an ethernet type interface. If the guest is started, the interface is then converted to a bridge type interface. The progression from network type to ethernet type to bridge type seems odd to me. With older versions of libvirt, the network type interface was converted to bridge type during the define, so I'm curious what the reasoning is here.. Is this the expected behavior? Also, the XML for the bridge interface in 0.6.2 doesn't include the source network. This information should be available, so I'm not sure why its not being set in the XML.. Thanks! ================= # virsh list --all Id Name State ---------------------------------- 0 Domain-0 running - rstest_domain shut off # cat tmp <domain type="xen"> <name>hd_domain</name> <on_poweroff>destroy</on_poweroff> <on_crash>destroy</on_crash> <uuid>f99fd6b6-1434-4bc2-88e0-1ed9c8c6f8e9</uuid> <os> <type>linux</type> <kernel>/tmp/default-xen-kernel</kernel> <initrd>/tmp/default-xen-initrd</initrd> <cmdline/> </os> <currentMemory>131072</currentMemory> <memory>131072</memory> <vcpu>1</vcpu> <devices> <disk type="file" device="disk"> <source file="/tmp/default-xen-dimage"/> <target dev="xvda"/> </disk> <interface type="network"> <mac address="00:11:22:33:44:aa"/> <source network="default"/> </interface> <input type="mouse" bus="xen"/> <graphics type="vnc" port="-1" listen="127.0.0.1" keymap="en-us"/> </devices> </domain> # virsh define tmp Domain hd_domain defined from tmp # virsh list --all Id Name State ---------------------------------- 0 Domain-0 running - hd_domain shut off - rstest_domain shut off [root@elm3b25 cimtest]# virsh dumpxml hd_domain <domain type='xen'> <name>hd_domain</name> <uuid>f99fd6b6-1434-4bc2-88e0-1ed9c8c6f8e9</uuid> <memory>131072</memory> <currentMemory>131072</currentMemory> <vcpu>1</vcpu> <os> <type arch='x86_64' machine='xenpv'>linux</type> <kernel>/tmp/default-xen-kernel</kernel> <initrd>/tmp/default-xen-initrd</initrd> </os> <clock offset='utc'/> <on_poweroff>destroy</on_poweroff> <on_reboot>restart</on_reboot> <on_crash>destroy</on_crash> <devices> <disk type='file' device='disk'> <driver name='file'/> <source file='/tmp/default-xen-dimage'/> <target dev='xvda' bus='xen'/> </disk> <interface type='ethernet'> <mac address='00:11:22:33:44:aa'/> </interface> <console type='pty'> <target port='0'/> </console> <input type='mouse' bus='xen'/> <graphics type='vnc' port='-1' autoport='yes' listen='127.0.0.1' keymap='en-us'/> </devices> </domain> # virsh start hd_domain Domain hd_domain started [root@elm3b25 cimtest]# virsh dumpxml hd_domain <domain type='xen' id='61'> <name>hd_domain</name> <uuid>f99fd6b6-1434-4bc2-88e0-1ed9c8c6f8e9</uuid> <memory>131072</memory> <currentMemory>131072</currentMemory> <vcpu>1</vcpu> <os> <type>linux</type> <kernel>/tmp/default-xen-kernel</kernel> <initrd>/tmp/default-xen-initrd</initrd> </os> <clock offset='utc'/> <on_poweroff>destroy</on_poweroff> <on_reboot>restart</on_reboot> <on_crash>destroy</on_crash> <devices> <disk type='file' device='disk'> <driver name='file'/> <source file='/tmp/default-xen-dimage'/> <target dev='xvda' bus='xen'/> </disk> <interface type='bridge'> <mac address='00:11:22:33:44:aa'/> <script path='vif-bridge'/> <target dev='vif61.0'/> </interface> <console type='pty' tty='/dev/pts/4'> <source path='/dev/pts/4'/> <target port='0'/> </console> <input type='mouse' bus='xen'/> <graphics type='vnc' port='5900' autoport='yes' listen='127.0.0.1' keymap='en-us'/> </devices> </domain> -- Kaitlin Rupert IBM Linux Technology Center kaitlin(a)linux.vnet.ibm.com

3 7

[libvirt] [PATCH 1/1] Support for RDP
by Pritesh Kothari 24 Apr '09

24 Apr '09

Hi All, I have added support for RDP. Just forwarding the patch for review. Regards, Pritesh

5 15

[libvirt] PATCH: Fix mem corruption after detaching Xen device
by Daniel P. Berrange 24 Apr '09

24 Apr '09

The XM driver's implementation of the 'detach device' command is causing memory corruption, but passing the wrong size information to memmove() by passing number of elements to move, instead of amount of memory to move. This patch fixes it Daniel diff -rup libvirt-0.6.2.orig/src/xm_internal.c libvirt-0.6.2.new/src/xm_internal.c --- libvirt-0.6.2.orig/src/xm_internal.c 2009-04-03 15:04:28.000000000 +0100 +++ libvirt-0.6.2.new/src/xm_internal.c 2009-04-23 10:53:10.000000000 +0100 @@ -2912,7 +2912,8 @@ xenXMDomainDetachDevice(virDomainPtr dom if (i < (def->ndisks - 1)) memmove(def->disks + i, def->disks + i + 1, - def->ndisks - (i + 1)); + sizeof(*def->disks) * + (def->ndisks - (i + 1))); break; } } @@ -2929,7 +2930,8 @@ xenXMDomainDetachDevice(virDomainPtr dom if (i < (def->nnets - 1)) memmove(def->nets + i, def->nets + i + 1, - def->nnets - (i + 1)); + sizeof(*def->nets) * + (def->nnets - (i + 1))); break; } } -- |: Red Hat, Engineering, London -o- http://people.redhat.com/berrange/ :| |: http://libvirt.org -o- http://virt-manager.org -o- http://ovirt.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: GnuPG: 7D3B9505 -o- F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|

2 1