- Devel - libvirt List Archives

[libvirt] ANNOUNCE: libvirt 3.2.1 maintenance release
by Cole Robinson 11 May '17

11 May '17

libvirt 3.2.1 is now available. This is a maintenance release of libvirt 3.2 with additional bugfixes that have accumulated upstream since the initial release. This release can be downloaded at: http://libvirt.org/sources/stable_updates/libvirt-3.2.1.tar.xz Changes in this version: * spec: Update version check for maint Source URL * mdev: Cleanup code after commits @daf5081b and @2739a983 * Don't use ceph-devel on Fedora * mdev: Fix mingw build by adding a check for non-NULL pointer * client: Report proper close reason * qemu: Fix persistent migration of transient domains * Fix padding of encrypted data * network: better log message when network is inactive during reconnect * qemu: don't kill qemu process on restart if networkNotify fails * conf: format only relevant attributes for graphics based on listen type * qemu: Move freeing of PCI address list to qemuProcessStop * qemu: process: Clean up priv->migTLSAlias * qemu: process: Don't leak priv->usbaddrs after VM restart * qemu: process: Clean automatic NUMA/cpu pinning information on shutdown * qemu: Remove extra messages for vhost-scsi hotplug * qemu: Remove extra messages from virtio-scsi hotplug * qemu: Check return code from qemuHostdevPrepareSCSIDevices * qemu: numa: Don't return automatic nodeset for inactive domain * qemu: Ignore missing query-migrate-parameters * daemon: Fix domain name leak in error path * qemu: fix argument of virDomainNetGetActualDirectMode * rpc: fix keep alive timer segfault * util: allow ignoring SIOCSIFHWADDR when errno is EPERM * util: check ifa_addr pointer before accessing its elements * Increase default task limit for libvirtd * Fix error reporting when poll returns POLLHUP/POLLERR * spec: Avoid RPM verification errors on nwfilter XMLs * xenconfig: avoid double free on OOM testing * xenFormatXLDomainDisks: avoid double free on OOM testing * virConfSaveValue: protect against a NULL pointer reference * conf: Add check for non scsi_host parent during vport delete * util: Fix resource leak * test: Remove unused variate @maxcpu in testDomainGetVcpus * esx: Fix memory leak * esx: Fix incorrect memory compare size in esxStoragePoolLookupByUUID * qemu: snapshot: Skip empty drives with internal snapshots * qemu: do not crash on USB address with no port and invalid bus * man: Align vol-resize arguments with the output of help * qemu: conf: Don't leak snapshot image format conf variable * qemu: Fix mdev checking for VFIO support * util: systemd: Don't strlen a possibly NULL string * interface: Fix resource leak in netcfConnectListAllInterfaces error path * virsh: don't leak @cpumap in virshVcpuPinQuery * tests: fix some resource leaks * rpc: fix resource leak * src: fix multiple resource leaks in loops * conf/domain_capabilities: fix resource leak * qemu: Fix two use-after-free situations * disk: Force usage of parted when checking disk format for "bsd" * disk: Resolve issues with disk partition build/start checks * conf: create new RemovalFailed event using correct class * qemu: fix memory leak and check mdevPath * qemu: Properly reset TLS in qemuProcessRecoverMigrationIn * Properly ignore files in build-aux directory * conf: Fix possible memleak in capabilities * Split out -Wframe-larger-than warning from WARN_CLFAGS * virISCSIGetSession: Don't leak memory * virStorageSourceClear: Don't leave dangling pointers behind * qemu: Break endless loop if qemuMigrationResetTLS fails * storage: gluster: Implement 'checkPool' method so that state is restored * docs: Document limitation of maximum vcpu count used with <topology> * qemu: Fix resource leak in qemuDomainAddChardevTLSObjects error path * qemu: Initialize 'data' argument * storage: util: Pass pool type to virStorageBackendFindGlusterPoolSources * util: ignore -Wcast-align in virNetlinkDumpCommand * qemu: hotplug: Clear vcpu ordering for coldplug of vcpus * qemu: hotplug: Fix formatting strings in qemuDomainFilterHotplugVcpuEntities * qemu: hotplug: Iterate over vcpu 0 in individual vcpu hotplug code * qemu: Add device id for mediated devices on qemu command line * storage: Fix capacity value for LUKS encrypted volumes * virNetDevIPCheckIPv6ForwardingCallback fixes * storage: driver: Remove unavailable transient pools after restart * storage: driver: Split out code fixing pool state after deactivation * storage: backend: Use correct stringifier for pool type * mdev: Fix daemon crash on domain shutdown after reconnect * util: mdev: Use a local variable instead of a direct pointer access * qemu: Fix regression when hyperv/vendor_id feature is used * vz: fix typo that breaks build For info about past maintenance releases, see: http://wiki.libvirt.org/page/Maintenance_Releases Thanks, Cole

1 0

[libvirt] [PATCH] tests: stub out virfilewrapper.c on Win32
by Daniel P. Berrange 11 May '17

11 May '17

The Win32 platform can not do link time overrides in the same way that we can on POSIX / ELF based platforms, so we cannot build the virfilewrapper.c code reliably. Just stub it out on Win32 so it is a no-op. Tests that use this file are already written to skip on Win32. Signed-off-by: Daniel P. Berrange <berrange(a)redhat.com> --- Pushed as a mingw build fix tests/virfilewrapper.c | 29 ++++++++++++++++------------- 1 file changed, 16 insertions(+), 13 deletions(-) diff --git a/tests/virfilewrapper.c b/tests/virfilewrapper.c index 5ea70b3..bf2fa69 100644 --- a/tests/virfilewrapper.c +++ b/tests/virfilewrapper.c @@ -18,15 +18,17 @@ #include <config.h> -#include <stdio.h> -#include <stdlib.h> -#include <fcntl.h> +#ifndef WIN32 -#include "viralloc.h" -#include "virfile.h" -#include "virfilewrapper.h" -#include "virmock.h" -#include "virstring.h" +# include <stdio.h> +# include <stdlib.h> +# include <fcntl.h> + +# include "viralloc.h" +# include "virfile.h" +# include "virfilewrapper.h" +# include "virmock.h" +# include "virstring.h" /* Mapping for prefix overrides */ @@ -139,7 +141,7 @@ virFileWrapperOverridePrefix(const char *path) } -#define PATH_OVERRIDE(newpath, path) \ +# define PATH_OVERRIDE(newpath, path) \ do { \ init_syms(); \ \ @@ -177,7 +179,7 @@ int access(const char *path, int mode) return ret; } -#ifdef HAVE___LXSTAT +# ifdef HAVE___LXSTAT int __lxstat(int ver, const char *path, struct stat *sb) { int ret = -1; @@ -191,7 +193,7 @@ int __lxstat(int ver, const char *path, struct stat *sb) return ret; } -#endif /* HAVE___LXSTAT */ +# endif /* HAVE___LXSTAT */ int lstat(const char *path, struct stat *sb) { @@ -207,7 +209,7 @@ int lstat(const char *path, struct stat *sb) return ret; } -#ifdef HAVE___XSTAT +# ifdef HAVE___XSTAT int __xstat(int ver, const char *path, struct stat *sb) { int ret = -1; @@ -221,7 +223,7 @@ int __xstat(int ver, const char *path, struct stat *sb) return ret; } -#endif /* HAVE___XSTAT */ +# endif /* HAVE___XSTAT */ int stat(const char *path, struct stat *sb) { @@ -278,3 +280,4 @@ DIR *opendir(const char *path) return ret; } +#endif -- 2.9.3

2 4

[libvirt] [PATCH] rpc: improve error message for bounds check
by Daniel P. Berrange 11 May '17

11 May '17

If we exceed a fixed limit in RPC code we get a horrible message like this, if the parameter type is a 'string', because we forgot to initialize the error message type field: $ virsh snapshot-list ostack1 error: too many remote undefineds: 1329 > 1024 It would also be useful to know which RPC call and field was exceeded. So this patch makes us report: $ virsh snapshot-list ostack1 error: too many remote undefineds: 1329 > 1024, in parameter 'names' for 'virDomainSnapshotListNames' Signed-off-by: Daniel P. Berrange <berrange(a)redhat.com> --- src/rpc/gendispatch.pl | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/src/rpc/gendispatch.pl b/src/rpc/gendispatch.pl index 173189c..0c5e4ba 100755 --- a/src/rpc/gendispatch.pl +++ b/src/rpc/gendispatch.pl @@ -1501,6 +1501,7 @@ elsif ($mode eq "client") { $single_ret_list_name = $1; $single_ret_list_max_var = "max$1"; $single_ret_list_max_define = $2; + $single_ret_list_error_msg_type = "string"; } elsif ($ret_member =~ m/^(admin|remote)_nonnull_string (\S+)<\S+>;/) { # error out on unannotated arrays die "$1_nonnull_string array without insert@<offset> annotation: $ret_member"; @@ -1773,7 +1774,8 @@ elsif ($mode eq "client") { print "\n"; print " if ($single_ret_list_max_var > $single_ret_list_max_define) {\n"; print " virReportError(VIR_ERR_RPC,\n"; - print " _(\"too many remote ${single_ret_list_error_msg_type}s: %d > %d\"),\n"; + print " _(\"too many remote ${single_ret_list_error_msg_type}s: %d > %d,\"\n"; + print " \"in parameter '$single_ret_list_name' for 'vir$call->{ProcName}'\"),\n"; print " $single_ret_list_max_var, $single_ret_list_max_define);\n"; print " goto done;\n"; print " }\n"; @@ -1839,7 +1841,8 @@ elsif ($mode eq "client") { $modern_ret_as_list) { print " if (ret.$single_ret_list_name.${single_ret_list_name}_len > $single_ret_list_max_var) {\n"; print " virReportError(VIR_ERR_RPC,\n"; - print " _(\"too many remote ${single_ret_list_error_msg_type}s: %d > %d\"),\n"; + print " _(\"too many remote ${single_ret_list_error_msg_type}s: %d > %d,\"\n"; + print " \"in parameter '$single_ret_list_name' for 'vir$call->{ProcName}'\"),\n"; print " ret.$single_ret_list_name.${single_ret_list_name}_len, $single_ret_list_max_var);\n"; print " goto cleanup;\n"; print " }\n"; -- 2.9.3

3 2

[libvirt] Release of libvirt-php-0.5.3
by Michal Privoznik 11 May '17

11 May '17

I just did the release of libvirt-php-0.5.3. You can download it here: http://libvirt.org/sources/php/libvirt-php-0.5.3.tar.gz There are 77 patches in this release (nearly 2x as much as in the previous release). There were several bug fixes, style conversions, etc. New features include: - New APIs - examples enhancement - build system improvements and many others. Thanks anybody who contributed! Michal

1 0

[libvirt] [PATCH] remote: increase max storage pools, nwfilters & snapshots to 16384
by Daniel P. Berrange 11 May '17

11 May '17

Most other top level objects have already had their limits increased to 16384. Increase the storage pool, nwfilter & snapshot object limits to match. For snapshots at least, we have seen hosts which exceeded the current limit Signed-off-by: Daniel P. Berrange <berrange(a)redhat.com> --- src/remote/remote_protocol.x | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/src/remote/remote_protocol.x b/src/remote/remote_protocol.x index 87b2bd3..4e3fa60 100644 --- a/src/remote/remote_protocol.x +++ b/src/remote/remote_protocol.x @@ -80,7 +80,7 @@ const REMOTE_NETWORK_LIST_MAX = 16384; const REMOTE_INTERFACE_LIST_MAX = 16384; /* Upper limit on lists of storage pools. */ -const REMOTE_STORAGE_POOL_LIST_MAX = 4096; +const REMOTE_STORAGE_POOL_LIST_MAX = 16384; /* Upper limit on lists of storage vols. */ const REMOTE_STORAGE_VOL_LIST_MAX = 16384; @@ -92,7 +92,7 @@ const REMOTE_NODE_DEVICE_LIST_MAX = 65536; const REMOTE_NODE_DEVICE_CAPS_LIST_MAX = 65536; /* Upper limit on lists of network filters. */ -const REMOTE_NWFILTER_LIST_MAX = 1024; +const REMOTE_NWFILTER_LIST_MAX = 16384; /* Upper limit on list of scheduler parameters. */ const REMOTE_DOMAIN_SCHEDULER_PARAMETERS_MAX = 16; @@ -137,7 +137,7 @@ const REMOTE_AUTH_TYPE_LIST_MAX = 20; const REMOTE_DOMAIN_MEMORY_STATS_MAX = 1024; /* Upper limit on lists of domain snapshots. */ -const REMOTE_DOMAIN_SNAPSHOT_LIST_MAX = 1024; +const REMOTE_DOMAIN_SNAPSHOT_LIST_MAX = 16384; /* Maximum length of a block peek buffer message. * Note applications need to be aware of this limit and issue multiple -- 2.9.3

2 1

[libvirt] [RFE] Raise enumeration limit for virDomain-{ListAllSnapshots, SnapshotListNames}?
by Kashyap Chamarthy 11 May '17

11 May '17

I've got a server with multiple guests, and wanted to quickly check how many of them had snapshots (and in turn, how many snapshots). So I quickly wrote the below script to enumerate the snapshots using the virDomainSnapshotListNames() API, and it turned out, one of the guests had 1329 snaphots (created by a script for testing): libvirt.libvirtError: too many remote undefineds: 1329 > 1024 DanPB on IRC said we could likely raise that limit: "we can likely increase that limit, since the rpc message size would allow many more before the ultimate limit is hit" He also points out the error message could be improved and is a sign of a bug: "'undefined' is a perl NULL that must come from a bug in the rpc generator" * * * Script I used to test: [ Version: libvirt-daemon-kvm-3.2.0-1.fc25.x86_64 ] ----------------------------------------------------------------------- #!/usr/bin/python import libvirt import sys conn = libvirt.openReadOnly('qemu:///system') if conn == None: print 'Failed to open connection to the hypervisor' sys.exit(1) domains = conn.listAllDomains(0) for name in domains: try: dom = conn.lookupByName(name.name()) except libvirt.libvirtError: print "Domain %s not found" % name sys.exit(1) print("Name: %s, ID: %d, OSType: %s" % (name.name(), dom.ID(), dom.OSType())) # print(dom.listAllSnapshots()) print(dom.snapshotListNames()) ----------------------------------------------------------------------- Executing the above script throws (since the 'ostack1' guest apparently had 1329 snapshots): [...] Name: cirros, ID: -1, OSType: hvm [] Name: compute, ID: -1, OSType: hvm [] Name: ostack1, ID: -1, OSType: hvm Traceback (most recent call last): File "enumerate-snapshots.py", line 23, in <module> print(dom.snapshotListNames()) File "/usr/lib64/python2.7/site-packages/libvirt.py", line 2527, in snapshotListNames if ret is None: raise libvirtError ('virDomainSnapshotListNames() failed', dom=self) libvirt.libvirtError: too many remote undefineds: 1329 > 1024 [...] -- /kashyap

1 1

[libvirt] CLI management tool
by Michal Privoznik 11 May '17

11 May '17

Dear list, you might have seen a discussion about virsh, and adding some new features to it [1]. While the feature was rejected, it got me thinking. What options do we offer for sysadmins that: a) want to stay in command line b) want higher level mgmt of their domains c) yet want to manage a single host Basically, virsh is just too low level for some operations (and using it in non-interactive mode from a script can mean hundreds of connections). Then we have virt-manager, which suits b) and c), but it's not a CLI tool. Therefore I was thinking whether we should start a new project on the top of libvirt that would fit all three points. Personally, I've never been a sysadmin, so perhaps I am not the best one to write the tool. But I'm open for suggestions. What do you think? Michal 1: https://www.redhat.com/archives/libvir-list/2017-May/msg00051.html

4 3

[libvirt] [PATCH] virsh: Add --tls description for the virsh man page
by John Ferlan 11 May '17

11 May '17

https://bugzilla.redhat.com/show_bug.cgi?id=1448806 Commit id '6a8d898d' neglected to update the man page. Signed-off-by: John Ferlan <jferlan(a)redhat.com> --- tools/virsh.pod | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/tools/virsh.pod b/tools/virsh.pod index cd1f25f..727acf6 100644 --- a/tools/virsh.pod +++ b/tools/virsh.pod @@ -1672,7 +1672,7 @@ I<domain> I<desturi> [I<migrateuri>] [I<graphicsuri>] [I<listen-address>] [I<dna [I<--compressed>] [I<--comp-methods> B<method-list>] [I<--comp-mt-level>] [I<--comp-mt-threads>] [I<--comp-mt-dthreads>] [I<--comp-xbzrle-cache>] [I<--auto-converge>] [I<auto-converge-initial>] -[I<auto-converge-increment>] [I<--persistent-xml> B<file>] +[I<auto-converge-increment>] [I<--persistent-xml> B<file>] [I<--tls>] Migrate domain to another host. Add I<--live> for live migration; <--p2p> for peer-2-peer migration; I<--direct> for direct migration; or I<--tunnelled> @@ -1753,6 +1753,11 @@ compression. I<--comp-mt-threads> and I<--comp-mt-dthreads> set the number of compress threads on source and the number of decompress threads on target respectively. I<--comp-xbzrle-cache> sets size of page cache in bytes. +Providing I<--tls> causes the migration to use the host configured TLS setup +(see migrate_tls_x509_cert_dir in /etc/libvirt/qemu.conf) in order to perform +the migration of the domain. Usage requires proper TLS setup for both source +and target. + Running migration can be canceled by interrupting virsh (usually using C<Ctrl-C>) or by B<domjobabort> command sent from another virsh instance. -- 2.9.3

2 1

[libvirt] [libvirt-php PATCH 0/2] implement a few network API bindings
by Dawid Zamirski 11 May '17

11 May '17

Hello, The followinig two patches add a few useful API bindings for dealing with libvirt networks: * virConnectListAllNetworks - allows for better control when filtering and returns array of handles to virNetworkPtr which can then be used to get other info for each network such as name, uuid, autostart etc. * virNetworkGetUUIDString, virNetworkGetUUID, virNetworkGetName to be able to easily get said info when working with virNetworkPtr handles. Dawid Zamirski (2): implement binding for virConnectListAllNetworks. implement additional network functions. src/libvirt-php.c | 149 ++++++++++++++++++++++++++++++++++++++++++++++++++++++ src/libvirt-php.h | 4 ++ 2 files changed, 153 insertions(+) -- 2.12.2

3 5

[libvirt] src/remote_protocol-structs file
by Vasiliy Tolstov 11 May '17

11 May '17

Hi. While i'm writing go binding for libvirt rpc api i'm parse src/remote/remote_protocol.x and generate needed code for go, but as i see libvirt have src/remote_protocol-structs file. That contains all needed data in mo simple syntax. But i don't see such files for qemu_protocol.x and lxc_protocol.x do you plan to add files for this xdr files, or src/remote_protocol-structs will be deprecated? Thanks! -- Vasiliy Tolstov, e-mail: v.tolstov(a)selfip.ru

2 2