- Devel - libvirt List Archives

[libvirt PATCH] run: set abs_builddir variables
by Ján Tomko 11 Aug '21

11 Aug '21

test-lib.sh needs these to be set. Export them so that the virsh-* tests can be run using: builddir$ ./run srcdir/tests/virsh-snapshot Signed-off-by: Ján Tomko <jtomko(a)redhat.com> --- meson.build | 1 + run.in | 3 +++ 2 files changed, 4 insertions(+) diff --git a/meson.build b/meson.build index 32ad688c9c..5af09d319a 100644 --- a/meson.build +++ b/meson.build @@ -2233,6 +2233,7 @@ configure_file(output: 'meson-config.h', configuration: conf) # generate run helper run_conf = configuration_data() run_conf.set('abs_builddir', meson.build_root()) +run_conf.set('abs_top_builddir', meson.build_root()) configure_file( input: 'run.in', output: '@BASENAME@', diff --git a/run.in b/run.in index 9aae27a79e..64f99751fe 100644 --- a/run.in +++ b/run.in @@ -80,6 +80,9 @@ env["LIBVIRT_DIR_OVERRIDE"] = "1" # read problems when using glibc. env["MALLOC_PERTURB_"] = "%d" % random.randint(1, 255) +env["abs_builddir"] = "@abs_builddir@" +env["abs_top_builddir"] = "@abs_top_builddir@" + modular_daemons = [ "virtinterfaced", "virtlxcd", -- 2.31.1

2 1

[libvirt PATCH 00/10] virsh: moments: use g_auto more (glib chronicles)
by Ján Tomko 11 Aug '21

11 Aug '21

Ján Tomko (10): vsh: table: define cleanup function virsh: add cleanup function for virshDomainCheckpoint virsh: add cleanup function for virshDomainSnapshot virsh: checkpoint: move variables inside the loop virsh: snapshot: move variables inside the loop virsh: cmdCheckpointInfo: delete unused variables virsh: checkpoint: use g_auto where possible virsh: checkpoint: remove pointless cleanup labels virsh: snapshot: use g_auto where possible virsh: snapshot: remove pointless cleanup labels tools/virsh-checkpoint.c | 188 ++++++++----------------- tools/virsh-snapshot.c | 288 +++++++++++++-------------------------- tools/virsh-util.h | 4 + tools/vsh-table.h | 1 + 4 files changed, 156 insertions(+), 325 deletions(-) -- 2.31.1

2 11

[libvirt PATCH 0/5] vsh: use g_auto more (glib chronicles)
by Ján Tomko 11 Aug '21

11 Aug '21

Except for vshCommandParse. -EPARSE Ján Tomko (5): vsh: do not cast away const vsh: cmdEcho: use separate variable for argument vsh: use g_auto where possible vsh: remove pointless cleanup labels vsh: use g_clear_pointer tools/vsh.c | 91 ++++++++++++++++++----------------------------------- 1 file changed, 31 insertions(+), 60 deletions(-) -- 2.31.1

2 11

[libvirt PATCH] meson: Always use the / operator to join paths
by Andrea Bolognani 10 Aug '21

10 Aug '21

This is the preferred way to do it, but there were a few instances in which some of the path components had embedded slashes instead. Signed-off-by: Andrea Bolognani <abologna(a)redhat.com> --- docs/internals/meson.build | 2 +- docs/kbase/meson.build | 2 +- docs/manpages/meson.build | 2 +- src/security/meson.build | 2 +- src/storage/meson.build | 4 ++-- 5 files changed, 6 insertions(+), 6 deletions(-) diff --git a/docs/internals/meson.build b/docs/internals/meson.build index 5aa67ded5c..2d0ae6195c 100644 --- a/docs/internals/meson.build +++ b/docs/internals/meson.build @@ -12,7 +12,7 @@ html_xslt_gen = [] foreach name : internals_in_files html_xslt_gen += { 'name': name, - 'source': 'docs/internals' / name + '.html.in', + 'source': 'docs' / 'internals' / name + '.html.in', } endforeach diff --git a/docs/kbase/meson.build b/docs/kbase/meson.build index 8d8fd695ea..73dc11837f 100644 --- a/docs/kbase/meson.build +++ b/docs/kbase/meson.build @@ -31,7 +31,7 @@ foreach name : docs_kbase_files html_xslt_gen += { 'name': name, 'file': docs_rst2html5_gen.process(rst_file), - 'source': 'docs/kbase' / rst_file, + 'source': 'docs' / 'kbase' / rst_file, } endforeach diff --git a/docs/manpages/meson.build b/docs/manpages/meson.build index fe91a43614..cafe29a545 100644 --- a/docs/manpages/meson.build +++ b/docs/manpages/meson.build @@ -125,7 +125,7 @@ foreach data : docs_man_files html_xslt_gen += { 'name': data['name'], 'file': html_in, - 'source': 'docs/manpages' / rst_in_file, + 'source': 'docs' / 'manpages' / rst_in_file, } endforeach diff --git a/src/security/meson.build b/src/security/meson.build index 6f5e1dec1d..6230b34aa4 100644 --- a/src/security/meson.build +++ b/src/security/meson.build @@ -47,7 +47,7 @@ if conf.has('WITH_LIBVIRTD') and conf.has('WITH_APPARMOR') conf_inc_dir, hypervisor_inc_dir, include_directories('.'), - include_directories('../storage_file'), + include_directories('..' / 'storage_file'), ], } endif diff --git a/src/storage/meson.build b/src/storage/meson.build index 915ae46f61..d1aa66b147 100644 --- a/src/storage/meson.build +++ b/src/storage/meson.build @@ -79,7 +79,7 @@ if conf.has('WITH_STORAGE') ], include_directories: [ conf_inc_dir, - include_directories('../storage_file'), + include_directories('..' / 'storage_file'), ], ) @@ -163,7 +163,7 @@ if conf.has('WITH_STORAGE_GLUSTER') glusterfs_dep, ], 'include': [ - include_directories('../storage_file'), + include_directories('..' / 'storage_file'), ], 'install_dir': storage_backend_install_dir, } -- 2.31.1

2 1

[PATCH v2 0/4] interface define: add support for validation against schema
by Kristina Hanicova 10 Aug '21

10 Aug '21

This is v2 of: https://listman.redhat.com/archives/libvir-list/2021-August/msg00187.html Changes since v1 (suggested by Peter): * split of the patch containing api and virsh adjustments into separate patches * change of documentation (which I forgot to do before) Kristina Hanicova (4): conf: propagate xmlDocPtr and flags into virInterfaceDefParseXML() api: add virInterfaceDefineFlags virsh: add support for '--validate' option in define interface interface_conf: add validation against schema in define docs/manpages/virsh.rst | 5 ++-- include/libvirt/libvirt-interface.h | 4 +++ src/conf/interface_conf.c | 38 ++++++++++++++++++------- src/conf/interface_conf.h | 6 ++-- src/conf/virinterfaceobj.c | 2 +- src/interface/interface_backend_netcf.c | 6 ++-- src/libvirt-interface.c | 2 +- src/test/test_driver.c | 4 +-- tests/interfacexml2xmltest.c | 2 +- tools/virsh-interface.c | 10 ++++++- 10 files changed, 55 insertions(+), 24 deletions(-) -- 2.31.1

2 6

[PATCH 0/4] qemu: Open chardev logfile on behalf of QEMU
by Michal Privoznik 10 Aug '21

10 Aug '21

See 4/4 for explanation. Michal Prívozník (4): virDomainDefGetSecurityLabelDef: Fix const correctness qemuDomainOpenFile: Take virDomainDef instead of virDomainObj qemuDomainOpenFile: Take @cfg instead of driver qemu: Open chardev logfile on behalf of QEMU src/conf/domain_conf.c | 2 +- src/conf/domain_conf.h | 2 +- src/qemu/qemu_command.c | 68 +++++++++++++++++++++++++++++---------- src/qemu/qemu_domain.c | 13 ++++---- src/qemu/qemu_domain.h | 4 +-- src/qemu/qemu_driver.c | 4 +-- src/qemu/qemu_saveimage.c | 5 +-- 7 files changed, 66 insertions(+), 32 deletions(-) -- 2.31.1

3 6

[PATCH v2] domain_conf: Fix memory leak when free'ing virDomainObj
by Luke Yue 10 Aug '21

10 Aug '21

Free dom->deprecations when free'ing virDomainObj Signed-off-by: Luke Yue <lukedyue(a)gmail.com> --- src/conf/domain_conf.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 09da4ab952..7e1d30bdb2 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -3730,6 +3730,16 @@ void virDomainDefFree(virDomainDef *def) g_free(def); } +static void +virDomainObjDeprecationFree(virDomainObj *dom) +{ + size_t i = 0; + for (i = 0; i < dom->ndeprecations; i++) { + g_free(dom->deprecations[i]); + } + g_free(dom->deprecations); +} + static void virDomainObjDispose(void *obj) { virDomainObj *dom = obj; @@ -3742,6 +3752,7 @@ static void virDomainObjDispose(void *obj) if (dom->privateDataFreeFunc) (dom->privateDataFreeFunc)(dom->privateData); + virDomainObjDeprecationFree(dom); virDomainSnapshotObjListFree(dom->snapshots); virDomainCheckpointObjListFree(dom->checkpoints); } -- 2.32.0

2 1

[PATCH 0/2] virDomain: fix problems when setting qos on ovs managed interface
by zhangjl02 10 Aug '21

10 Aug '21

Two problems are found and fixed below: 1. Ingress rules is not clean on previous version of virNetDevOpenvswitchInterfaceClearQos. 2. If errors occurs when removing inbound qos on multi interfaces vm, some rules may not be delete as aspected. Fix by: Instead of cleaning all qos rules each time new qos is set, tx and rx's qos are set or cleaned respectively. Replace virReportError with VIR_WARN to let the cleaning process continue when error occurs. Add ifname into ovs querying statements, which will reduce failure of removing qos on the other interfaces of the same vm. Test virNetDevOpenvswitchInterfaceSetQos and virNetDevOpenvswitchInterfaceClearQos with dryrun method. Since commands in tests are not actually run, it is difficult to emulate some complex senario, such as set and then update qos. So basic tests are added in patches. Jinsheng Zhang (2): virDomain: interface: add virNetDevOpenvswitchInterfaceClearTxQos and virNetDevOpenvswitchInterfaceClearRxQos. tests: add test on virNetDevOpenvswitchInterfaceSetQos and virNetDevOpenvswitchInterfaceClearQos src/libvirt_private.syms | 2 + src/util/virnetdevopenvswitch.c | 69 +++++++++-- src/util/virnetdevopenvswitch.h | 7 ++ tests/virnetdevopenvswitchtest.c | 196 ++++++++++++++++++++++++++++++- 4 files changed, 262 insertions(+), 12 deletions(-) -- 2.30.2.windows.1

2 3

[PATCH] domain_conf: Fix memory leak when free'ing virDomainObj
by Luke Yue 10 Aug '21

10 Aug '21

Free dom->deprecations when free'ing virDomainObj Signed-off-by: Luke Yue <lukedyue(a)gmail.com> --- src/conf/domain_conf.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 09da4ab952..2527cd1d70 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -3733,6 +3733,7 @@ void virDomainDefFree(virDomainDef *def) static void virDomainObjDispose(void *obj) { virDomainObj *dom = obj; + size_t i = 0; VIR_DEBUG("obj=%p", dom); virCondDestroy(&dom->cond); @@ -3742,6 +3743,11 @@ static void virDomainObjDispose(void *obj) if (dom->privateDataFreeFunc) (dom->privateDataFreeFunc)(dom->privateData); + for (i = 0; i < dom->ndeprecations; i++) { + g_free(dom->deprecations[i]); + } + g_free(dom->deprecations); + virDomainSnapshotObjListFree(dom->snapshots); virDomainCheckpointObjListFree(dom->checkpoints); } -- 2.32.0

2 2

[PATCH 0/3] interface define: add support for validation against schema
by Kristina Hanicova 06 Aug '21

06 Aug '21

This is just the first series of patches of many more. I will send them soon, I just wanted to know if I am on the right path as they will follow the same pattern. Partially resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1972586 Kristina Hanicova (3): conf: propagate xmlDocPtr and flags into virInterfaceDefParseXML() virsh: add support for '--validate' option in define interface interface_conf: add validation against schema in define include/libvirt/libvirt-interface.h | 4 +++ src/conf/interface_conf.c | 38 ++++++++++++++++++------- src/conf/interface_conf.h | 6 ++-- src/conf/virinterfaceobj.c | 2 +- src/interface/interface_backend_netcf.c | 6 ++-- src/test/test_driver.c | 4 +-- tests/interfacexml2xmltest.c | 2 +- tools/virsh-interface.c | 10 ++++++- 8 files changed, 51 insertions(+), 21 deletions(-) -- 2.31.1

2 4

[PATCH] src: stop checking virIdentityNew return value
by Daniel P. Berrangé 06 Aug '21

06 Aug '21

This method will always succeed. Signed-off-by: Daniel P. Berrangé <berrange(a)redhat.com> --- src/remote/remote_daemon_dispatch.c | 5 +---- src/rpc/virnetserverclient.c | 5 +---- src/util/viridentity.c | 5 +---- 3 files changed, 3 insertions(+), 12 deletions(-) diff --git a/src/remote/remote_daemon_dispatch.c b/src/remote/remote_daemon_dispatch.c index 69d0a015ba..0e38b314a2 100644 --- a/src/remote/remote_daemon_dispatch.c +++ b/src/remote/remote_daemon_dispatch.c @@ -2192,7 +2192,7 @@ remoteDispatchConnectSetIdentity(virNetServer *server G_GNUC_UNUSED, int nparams = 0; int rv = -1; virConnectPtr conn = remoteGetHypervisorConn(client); - g_autoptr(virIdentity) ident = NULL; + g_autoptr(virIdentity) ident = virIdentityNew(); if (!conn) goto cleanup; @@ -2209,9 +2209,6 @@ remoteDispatchConnectSetIdentity(virNetServer *server G_GNUC_UNUSED, if (virConnectSetIdentityEnsureACL(conn) < 0) goto cleanup; - if (!(ident = virIdentityNew())) - goto cleanup; - if (virIdentitySetParameters(ident, params, nparams) < 0) goto cleanup; diff --git a/src/rpc/virnetserverclient.c b/src/rpc/virnetserverclient.c index 236702ced6..7050430e17 100644 --- a/src/rpc/virnetserverclient.c +++ b/src/rpc/virnetserverclient.c @@ -747,10 +747,7 @@ virNetServerClientCreateIdentity(virNetServerClient *client) g_autofree char *username = NULL; g_autofree char *groupname = NULL; g_autofree char *seccontext = NULL; - g_autoptr(virIdentity) ret = NULL; - - if (!(ret = virIdentityNew())) - return NULL; + g_autoptr(virIdentity) ret = virIdentityNew(); if (client->sock && virNetSocketIsLocal(client->sock)) { gid_t gid; diff --git a/src/util/viridentity.c b/src/util/viridentity.c index c18326c8cb..e36e54ae4b 100644 --- a/src/util/viridentity.c +++ b/src/util/viridentity.c @@ -313,15 +313,12 @@ virIdentity *virIdentityGetSystem(void) g_autofree char *username = NULL; g_autofree char *groupname = NULL; unsigned long long startTime; - g_autoptr(virIdentity) ret = NULL; + g_autoptr(virIdentity) ret = virIdentityNew(); #if WITH_SELINUX char *con; #endif g_autofree char *token = NULL; - if (!(ret = virIdentityNew())) - return NULL; - if (virIdentitySetProcessID(ret, getpid()) < 0) return NULL; -- 2.31.1

2 1

[PATCH 00/13] Clean up string list freeing
by Peter Krempa 06 Aug '21

06 Aug '21

Switch remaining users of virStringListFreeCount to g_auto(GStrv) and clean up some usage of string lists. Depends on the refactor of the virtual function code I've posted earlier. Pipeline: https://gitlab.com/pipo.sk/libvirt/-/pipelines/348627403 Peter Krempa (13): testQemuAgentSSHKeys: Refactor cleanup remote: dispatch: Don't use virStringListFreeCount for NULL terminated lists qemu: firmware: Store machine types as a NULL-terminated string list qemu: domain: Store passthrough environment variables in a struct qemu: domain: Store capability overrides in NULL-terminated string list qemu: domain: Store passthrough arguments in NULL-terminated string list network: bridge: Store dnsmasq passthrough options in NULL-terminated string list qemuNamespacePrepareOneItem: Restructure code to avoid temporary variables qemuNamespacePrepareOneItem: Don't pass count of elements qemuNamespaceMknodPaths: Remove 'ndevMountsPath' qemuDomainUnshareNamespace: Use automatic memory clearing for string lists lxcContainerSetReadOnly: Refactor cleanup handling util: virstring: Remove unused virStringListFreeCount src/libvirt_private.syms | 1 - src/lxc/lxc_container.c | 27 ++++----- src/network/bridge_driver.c | 23 ++++---- src/qemu/qemu_command.c | 9 +-- src/qemu/qemu_domain.c | 91 +++++++++++++++-------------- src/qemu/qemu_domain.h | 12 ++-- src/qemu/qemu_firmware.c | 4 +- src/qemu/qemu_namespace.c | 46 ++++++--------- src/qemu/qemu_process.c | 13 ++--- src/remote/remote_daemon_dispatch.c | 8 +-- src/util/virstring.c | 23 -------- src/util/virstring.h | 3 - tests/qemuagenttest.c | 54 ++++++++--------- 13 files changed, 135 insertions(+), 179 deletions(-) -- 2.31.1

2 15

[PATCH 0/6] util: Refactor fetching of virtual function list
by Peter Krempa 06 Aug '21

06 Aug '21

Note that this applies on top of the 'VIR_APPEND_ELEMENT' refactor I've sent earlier. Pipeline: https://gitlab.com/pipo.sk/libvirt/-/pipelines/348524810 Peter Krempa (6): virNetDevGetVirtualFunctions: Remove 'max_vfs' argument virPCIGetVirtualFunctions: Simplify cleanup of returned data virPCIGetVirtualFunctionIndex: Refactor cleanup virPCIGetNetName: Make 'physPortID' argument const virPCIGetVirtualFunctions: Fetch also network interface name if needed virNetDevGetVirtualFunctions: Directly return virPCIVirtualFunctionList src/conf/node_device_conf.c | 16 +++-- src/libvirt_private.syms | 2 + src/network/bridge_driver.c | 23 ++----- src/util/virnetdev.c | 65 +++--------------- src/util/virnetdev.h | 8 +-- src/util/virpci.c | 127 +++++++++++++++++++++--------------- src/util/virpci.h | 24 +++++-- 7 files changed, 122 insertions(+), 143 deletions(-) -- 2.31.1

2 7

[PATCH 0/7] vmx: Support super wide SCSI bus
by Michal Privoznik 06 Aug '21

06 Aug '21

As of vSphere 6.7 there can be up to 64 units on a SCSI bus. Honestly, I'm not a big fan of 5/7 but we want to support three different values and one of them doesn't depend on the driver that's assigning the disk address, but at a runtime information. Thus using xmlopt is not good enough. Any clever idea is very welcome. Michal Prívozník (7): vmx: Fill virtualHW.version to ESX version mapping vmx2xmltest: Add a test case vmx: Rework disk def allocation vmx: Drop needless check in virVMXParseDisk() conf: Store SCSI bus length in virDomainDef vmx: Support super wide SCSI bus DO NOT MERGE src/bhyve/bhyve_parse_command.c | 2 +- src/conf/domain_conf.c | 33 +++--- src/conf/domain_conf.h | 10 +- src/hyperv/hyperv_driver.c | 4 +- src/libxl/libxl_driver.c | 2 +- src/libxl/xen_xl.c | 2 +- src/libxl/xen_xm.c | 2 +- src/lxc/lxc_native.c | 2 +- src/openvz/openvz_conf.c | 2 +- src/qemu/qemu_process.c | 2 +- src/vbox/vbox_common.c | 8 +- src/vmx/vmx.c | 108 +++++++++--------- src/vz/vz_sdk.c | 2 +- tests/genericxml2xmltest.c | 2 +- tests/openvzutilstest.c | 2 +- tests/qemublocktest.c | 2 +- tests/qemumonitortestutils.c | 2 +- tests/securityselinuxtest.c | 2 +- .../vmx2xml-esx-in-the-wild-11.vmx | 97 ++++++++++++++++ .../vmx2xml-esx-in-the-wild-11.xml | 43 +++++++ tests/vmx2xmltest.c | 1 + 21 files changed, 239 insertions(+), 91 deletions(-) create mode 100644 tests/vmx2xmldata/vmx2xml-esx-in-the-wild-11.vmx create mode 100644 tests/vmx2xmldata/vmx2xml-esx-in-the-wild-11.xml -- 2.31.1

2 9

[libvirt PATCH 0/7] Fix some mutex related issues
by Tim Wiederhake 05 Aug '21

05 Aug '21

The calls to pthread_mutex_{lock,unlock,destroy} in "virthread.c" ignore the return value of these functions. This series fixes all "double unlock mutex" and "destroy locked mutex" issues I came across and adds checks to the calls to pthread_mutex_*, so no new issues of this type are introduced. Patches 1, 2, 3, and 4 fix the issues, patches 5, 6, and 7 add the checks. See https://gitlab.com/twiederh/libvirt/-/pipelines/348505233 and https://gitlab.com/twiederh/libvirt/-/pipelines/348505233 (only patches 5, 6, and 7). Regards, Tim Tim Wiederhake (7): tests: qemuMonitorTestFree: Unlock virDomainObj before freeing qemusecuritytest: Unlock virDomainObj before disposal qemuxml2argvtest: Unlock virDomainObj before disposal qemuhotplugtest: Lock reused virDomainObj Meson: Define "DYNAMIC_ANALYSIS" if built with sanitizers enabled virMutex*: Warn on error virMutex*: Fail loudly meson.build | 4 ++++ src/util/virthread.c | 24 +++++++++++++++++++++--- tests/qemuhotplugtest.c | 1 + tests/qemumonitortestutils.c | 2 +- tests/qemusecuritytest.c | 3 ++- tests/qemuxml2argvtest.c | 12 +++++++----- 6 files changed, 36 insertions(+), 10 deletions(-) -- 2.31.1

3 20

[PATCH 00/23] Remove return value from VIR_APPEND_ELEMENT and cleanups
by Peter Krempa 05 Aug '21

05 Aug '21

Since this touches all kind of conditionally compiled code you can find the pipeline build results here: https://gitlab.com/pipo.sk/libvirt/-/pipelines/347734973 Peter Krempa (23): virInsertElementsN: Rename 'add' argument virInsertElementsN: Split out actual insertion code util: alloc: Introduce virAppendElement helper util: alloc: Reimplement VIR_APPEND_ELEMENT_(COPY_)INPLACE using virAppendElement util: alloc: Reimplement VIR_APPEND_ELEMENT_COPY using virAppendElement util: alloc: Reimplement VIR_APPEND_ELEMENT_QUIET using virAppendElement util: alloc: Reimplement VIR_APPEND_ELEMENT using virAppendElement util: alloc: Completely replace VIR_APPEND_ELEMENT_QUIET by VIR_APPEND_ELEMENT prlsdkAddDomainVideoInfoCt: Remove pointless cleanup section xenParseXMDisk: Use automatic memory clearing and remove 'ret' variable qemuDomainUSBAddressAddHubs: Refactor cleanup virSecuritySELinuxContextListAppend: Remove unreachable cleanup qemuDomainAttachDeviceConfig: Remove pointless assignment virNWFilterRuleDefToRuleInst: Remove pointless assignment virObjectEventCallbackListAddID: Remove pointless cleanup of 'cb' virNWFilterRuleDefToRuleInst: Restructure code to avoid cleanup virNWFilterIncludeDefToRuleInst: Refactor cleanup qemuProcessSetupHotpluggableVcpus: Use automatic memory freeing virNetServerGetClients: Remove pointless cleanup lxcNetworkParseDataIPs: Automatically free string list virStorageBackendLogicalParseVolExtents: Declare one variable per line virStorageBackendLogicalParseVolExtents: Move 'extents' inside the loop virStorageBackendLogicalParseVolExtents: Remove 'cleanup' and 'ret' src/bhyve/bhyve_parse_command.c | 18 +-- src/conf/capabilities.c | 11 +- src/conf/domain_conf.c | 95 ++++++---------- src/conf/network_conf.c | 9 +- src/conf/node_device_conf.c | 24 ++-- src/conf/nwfilter_conf.c | 12 +- src/conf/object_event.c | 9 +- src/conf/storage_conf.c | 6 +- src/conf/virdomainobjlist.c | 6 +- src/conf/virnwfilterobj.c | 8 +- src/cpu/cpu_arm.c | 6 +- src/cpu/cpu_ppc64.c | 6 +- src/cpu/cpu_x86.c | 21 ++-- src/esx/esx_driver.c | 7 +- src/esx/esx_network_driver.c | 3 +- src/hyperv/hyperv_driver.c | 22 +--- src/hyperv/hyperv_network_driver.c | 3 +- src/libvirt_private.syms | 1 + src/libxl/xen_common.c | 17 +-- src/libxl/xen_xl.c | 21 +--- src/libxl/xen_xm.c | 21 +--- src/logging/log_handler.c | 6 +- src/lxc/lxc_container.c | 5 +- src/lxc/lxc_controller.c | 3 +- src/lxc/lxc_native.c | 12 +- src/network/bridge_driver.c | 4 +- src/node_device/node_device_driver.c | 3 +- src/nwfilter/nwfilter_ebiptables_driver.c | 3 +- src/nwfilter/nwfilter_gentech_driver.c | 57 ++++------ src/openvz/openvz_conf.c | 8 +- src/qemu/qemu_block.c | 9 +- src/qemu/qemu_command.c | 14 +-- src/qemu/qemu_domain.c | 6 +- src/qemu/qemu_domain_address.c | 9 +- src/qemu/qemu_driver.c | 17 +-- src/qemu/qemu_firmware.c | 3 +- src/qemu/qemu_hotplug.c | 6 +- src/qemu/qemu_monitor_json.c | 3 +- src/qemu/qemu_namespace.c | 5 +- src/qemu/qemu_process.c | 17 +-- src/remote/remote_daemon_dispatch.c | 56 ++++------ src/rpc/virnetdaemon.c | 4 +- src/rpc/virnetmessage.c | 3 +- src/rpc/virnetserver.c | 12 +- src/security/security_dac.c | 3 +- src/security/security_manager.c | 5 +- src/security/security_selinux.c | 15 +-- src/storage/storage_backend_logical.c | 38 +++---- src/storage/storage_backend_rbd.c | 3 +- src/util/viralloc.c | 129 +++++++++++++++++----- src/util/viralloc.h | 48 ++++---- src/util/virfirewall.c | 8 +- src/util/viriscsi.c | 3 +- src/util/virjson.c | 5 +- src/util/virlog.c | 10 +- src/util/virmdev.c | 7 +- src/util/virnetdev.c | 3 +- src/util/virnvme.c | 7 +- src/util/virpci.c | 13 +-- src/util/virprocess.c | 3 +- src/util/virresctrl.c | 6 +- src/util/virscsi.c | 7 +- src/util/virscsivhost.c | 4 +- src/util/virusb.c | 4 +- src/util/virutil.c | 9 +- src/vbox/vbox_common.c | 13 +-- src/vmx/vmx.c | 3 +- src/vz/vz_sdk.c | 32 ++---- tests/nwfilterxml2firewalltest.c | 13 +-- tests/qemumonitortestutils.c | 11 +- tests/testutils.c | 10 +- tests/virfilewrapper.c | 9 +- tests/virpcimock.c | 10 +- tools/virsh-domain.c | 3 +- tools/vsh-table.c | 9 +- 75 files changed, 404 insertions(+), 630 deletions(-) -- 2.31.1

2 25

[libvirt PATCH v2 0/5] add support for Fibre Channel VMID
by Pavel Hrdina 05 Aug '21

05 Aug '21

changes in v2: - refactor of resource parsing and formatting code - use <fibrechannel vmid=''/> element - use stat() directly - report only single system error Pavel Hrdina (5): vircgroup: introduce virCgroupGetInode function conf: refactor virDomainResourceDefParse conf: refactor virDomainResourceDefFormat conf: introduce support for Fibre Channel VMID qemu: implement support for Fibre Channel VMID docs/formatdomain.rst | 21 ++++++++++ docs/schemas/domaincommon.rng | 21 ++++++++-- src/conf/domain_conf.c | 42 +++++++++++-------- src/conf/domain_conf.h | 1 + src/conf/domain_validate.c | 19 +++++++++ src/libvirt_private.syms | 1 + src/qemu/qemu_cgroup.c | 27 ++++++++++++ src/util/vircgroup.c | 31 ++++++++++++++ src/util/vircgroup.h | 2 + .../fibrechannel-vmid.xml | 21 ++++++++++ tests/genericxml2xmltest.c | 2 + 11 files changed, 167 insertions(+), 21 deletions(-) create mode 100644 tests/genericxml2xmlindata/fibrechannel-vmid.xml -- 2.31.1

1 5

[PATCH] tests: Don't leak cpu defs
by Michal Privoznik 05 Aug '21

05 Aug '21

There are cpu definitions that are allocated in qemuTestDriverInit() but are missing corresponding virCPUDefFree() call in qemuTestDriverFree(). It's safe to call the free function because the definitions contain a refcounter and thus even if they were still in use the refcounter would be just decreased. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- tests/testutilsqemu.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/tests/testutilsqemu.c b/tests/testutilsqemu.c index fab676c070..9a0666724a 100644 --- a/tests/testutilsqemu.c +++ b/tests/testutilsqemu.c @@ -332,6 +332,11 @@ void qemuTestDriverFree(virQEMUDriver *driver) virObjectUnref(driver->caps); virObjectUnref(driver->config); virObjectUnref(driver->securityManager); + + virCPUDefFree(cpuDefault); + virCPUDefFree(cpuHaswell); + virCPUDefFree(cpuPower8); + virCPUDefFree(cpuPower9); } int qemuTestCapsCacheInsert(virFileCache *cache, -- 2.31.1

2 1

[PATCH] rpm: add conditionals around post scripts
by Daniel P. Berrangé 05 Aug '21

05 Aug '21

The hypervisor drivers can be disabled in certain build scenarios, so their corresponding post scripts need to match. Signed-off-by: Daniel P. Berrangé <berrange(a)redhat.com> --- libvirt.spec.in | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/libvirt.spec.in b/libvirt.spec.in index a162bc333d..011f5dd49e 100644 --- a/libvirt.spec.in +++ b/libvirt.spec.in @@ -1458,6 +1458,7 @@ fi %libvirt_daemon_perform_restart virtstoraged +%if %{with_qemu} %post daemon-driver-qemu %if %{with_modular_daemons} %libvirt_daemon_systemd_post virtqemud @@ -1469,8 +1470,10 @@ fi %posttrans daemon-driver-qemu %libvirt_daemon_perform_restart virtqemud +%endif +%if %{with_lxc} %post daemon-driver-lxc %if %{with_modular_daemons} %libvirt_daemon_systemd_post virtlxcd @@ -1482,8 +1485,10 @@ fi %posttrans daemon-driver-lxc %libvirt_daemon_perform_restart virtlxcd +%endif +%if %{with_vbox} %post daemon-driver-vbox %if %{with_modular_daemons} %libvirt_daemon_systemd_post virtvboxd @@ -1495,8 +1500,10 @@ fi %posttrans daemon-driver-vbox %libvirt_daemon_perform_restart virtvboxd +%endif +%if %{with_libxl} %post daemon-driver-libxl %if %{with_modular_daemons} %libvirt_daemon_systemd_post virtxend @@ -1508,6 +1515,7 @@ fi %posttrans daemon-driver-libxl %libvirt_daemon_perform_restart virtxend +%endif %post daemon-config-network -- 2.31.1

2 1

[PATCH] util: directly query KVM for TSC scaling support
by Daniel P. Berrangé 05 Aug '21

05 Aug '21

We currently query the host MSRs to determine if TSC scaling is supported. This works OK when running privileged and can open the /dev/cpu/0/msr. When unprivileged we fallback to querying MSRs from /dev/kvm. This is incorrect because /dev/kvm only reports accurate info for MSRs that are valid to use from inside a guest. The TSC scaling support MSR is not, thus we always end up reporting lack of TSC scaling when unprivileged. The solution to this is easy, because KVM can directly report whether TSC scaling is available, which matches what QEMU will do at startup. https://gitlab.com/libvirt/libvirt/-/issues/188 Reported-by: Roman Mohr <rmohr(a)redhat.com> Signed-off-by: Daniel P. Berrangé <berrange(a)redhat.com> --- src/util/virhostcpu.c | 19 +++++++------------ 1 file changed, 7 insertions(+), 12 deletions(-) diff --git a/src/util/virhostcpu.c b/src/util/virhostcpu.c index 7aa92ad11d..f140610b47 100644 --- a/src/util/virhostcpu.c +++ b/src/util/virhostcpu.c @@ -1349,11 +1349,10 @@ virHostCPUGetMSR(unsigned long index, virHostCPUTscInfo * virHostCPUGetTscInfo(void) { - virHostCPUTscInfo *info; + g_autofree virHostCPUTscInfo *info = g_new0(virHostCPUTscInfo, 1); VIR_AUTOCLOSE kvmFd = -1; VIR_AUTOCLOSE vmFd = -1; VIR_AUTOCLOSE vcpuFd = -1; - uint64_t msr = 0; int rc; if ((kvmFd = open(KVM_DEVICE, O_RDONLY)) < 0) { @@ -1378,23 +1377,19 @@ virHostCPUGetTscInfo(void) _("Unable to probe TSC frequency")); return NULL; } - - info = g_new0(virHostCPUTscInfo, 1); - info->frequency = rc * 1000ULL; - if (virHostCPUGetMSR(VMX_PROCBASED_CTLS2_MSR, &msr) == 0) { - /* High 32 bits of the MSR value indicate whether specific control - * can be set to 1. */ - msr >>= 32; - - info->scaling = virTristateBoolFromBool(!!(msr & VMX_USE_TSC_SCALING)); + if ((rc = ioctl(kvmFd, KVM_CHECK_EXTENSION, KVM_CAP_TSC_CONTROL)) < 0) { + virReportSystemError(errno, "%s", + _("Unable to query TSC scaling support")); + return NULL; } + info->scaling = rc ? VIR_TRISTATE_BOOL_YES : VIR_TRISTATE_BOOL_NO; VIR_DEBUG("Detected TSC frequency %llu Hz, scaling %s", info->frequency, virTristateBoolTypeToString(info->scaling)); - return info; + return g_steal_pointer(&info); } #else -- 2.31.1

2 1

[PATCH] Revert "remote: remove probing logic from virtproxyd dispatcher"
by Daniel P. Berrangé 05 Aug '21

05 Aug '21

This reverts commit 05bd8db60b35b7706100d9cbbaeb13992965e0f2. It is true that the remote driver client now contains logic for probing the driver to connect to when using modular daemons. This logic, however, only runs when the remote driver is NOT running inside a daemon since we don't want it activated inside libvirtd. Since the same remote driver build is used in all daemons, we can't rely on it in virtproxyd either. Thus we need to keep the virtproxyd probing logic Signed-off-by: Daniel P. Berrangé <berrange(a)redhat.com> --- src/remote/remote_daemon_dispatch.c | 77 +++++++++++++++++++++++++++++ 1 file changed, 77 insertions(+) diff --git a/src/remote/remote_daemon_dispatch.c b/src/remote/remote_daemon_dispatch.c index 65aa20f7d1..36d4d00b79 100644 --- a/src/remote/remote_daemon_dispatch.c +++ b/src/remote/remote_daemon_dispatch.c @@ -1948,6 +1948,73 @@ void *remoteClientNew(virNetServerClient *client, /*----- Functions. -----*/ +#ifdef VIRTPROXYD +/* + * When running in virtproxyd regular auto-probing of drivers + * does not work as we don't have any drivers present (except + * stateless ones inside libvirt.so). All the interesting + * drivers are in separate daemons. Thus when we get a NULL + * URI we need to simulate probing that virConnectOpen would + * previously do. We use the existence of the UNIX domain + * socket as our hook for probing. + * + * This assumes no stale sockets left over from a now dead + * daemon, but that's reasonable since libvirtd unlinks + * sockets it creates on shutdown, or uses systemd activation + * + * We only try to probe for primary hypervisor drivers, + * not the secondary drivers. + */ +static int +remoteDispatchProbeURI(bool readonly, + char **probeduri) +{ + g_autofree char *driver = NULL; + const char *suffix; + *probeduri = NULL; + VIR_DEBUG("Probing for driver daemon sockets"); + + /* + * If running root, either the daemon is running and the socket + * exists, or we're using socket activation so the socket exists + * too. + * + * If running non-root, the daemon may or may not already be + * running, and socket activation probably isn't relevant. + * So if no viable socket exists, we need to check which daemons + * are actually installed. This is not a big deal as only QEMU & + * VBox run as non-root, anyway. + */ + if (geteuid() != 0) { + if (remoteProbeSessionDriverFromSocket(false, &driver) < 0) + return -1; + + if (driver == NULL && + remoteProbeSessionDriverFromBinary(&driver) < 0) + return -1; + + suffix = "session"; + } else { + if (remoteProbeSystemDriverFromSocket(readonly, &driver) < 0) + return -1; + + suffix = "system"; + } + + /* Even if we didn't probe any socket, we won't + * return error. Just let virConnectOpen's normal + * logic run which will likely return an error anyway + */ + if (!driver) + return 0; + + *probeduri = g_strdup_printf("%s:///%s", driver, suffix); + VIR_DEBUG("Probed URI %s for driver %s", *probeduri, driver); + return 0; +} +#endif /* VIRTPROXYD */ + + static int remoteDispatchConnectOpen(virNetServer *server G_GNUC_UNUSED, virNetServerClient *client, @@ -1956,6 +2023,9 @@ remoteDispatchConnectOpen(virNetServer *server G_GNUC_UNUSED, struct remote_connect_open_args *args) { const char *name; +#ifdef VIRTPROXYD + g_autofree char *probeduri = NULL; +#endif unsigned int flags; struct daemonClientPrivate *priv = virNetServerClientGetPrivateData(client); int rv = -1; @@ -1984,6 +2054,13 @@ remoteDispatchConnectOpen(virNetServer *server G_GNUC_UNUSED, priv->readonly = flags & VIR_CONNECT_RO; #ifdef VIRTPROXYD + if (!name || STREQ(name, "")) { + if (remoteDispatchProbeURI(priv->readonly, &probeduri) < 0) + goto cleanup; + + name = probeduri; + } + preserveIdentity = true; #endif /* VIRTPROXYD */ -- 2.31.1

2 1

[libvirt PATCHv2 0/6] kill whole virtiofsd process group (virtio-fs epopee)
by Ján Tomko 04 Aug '21

04 Aug '21

v2 of: https://listman.redhat.com/archives/libvir-list/2021-June/msg00570.html Ján Tomko (6): virProcessKillPainfullyDelay: use 'rc' variable util: Introduce virProcessGroupKill util: introduce virProcessGroupGet util: virPidFileForceCleanupPath: add group argument qemu: virtiofs: kill the whole process group util: fix typo src/libvirt_private.syms | 1 + src/qemu/qemu_process.c | 3 ++- src/qemu/qemu_virtiofs.c | 2 +- src/util/virpidfile.c | 15 ++++++++++-- src/util/virpidfile.h | 2 ++ src/util/virprocess.c | 53 +++++++++++++++++++++++++++++++++++----- src/util/virprocess.h | 5 +++- 7 files changed, 70 insertions(+), 11 deletions(-) -- 2.31.1

3 13

[libvirt PATCH 0/3] add support for Fibre Channel VMID
by Pavel Hrdina 04 Aug '21

04 Aug '21

Pavel Hrdina (3): vircgroup: introduce virCgroupGetInode function conf: introduce support for Fibre Channel VMID qemu: implement support for Fibre Channel VMID docs/formatdomain.rst | 13 ++++++++++ docs/schemas/domaincommon.rng | 11 +++++++++ src/conf/domain_conf.c | 5 ++++ src/conf/domain_conf.h | 2 ++ src/conf/domain_validate.c | 19 +++++++++++++++ src/libvirt_private.syms | 1 + src/qemu/qemu_cgroup.c | 32 +++++++++++++++++++++++++ src/util/vircgroup.c | 37 +++++++++++++++++++++++++++++ src/util/vircgroup.h | 2 ++ tests/genericxml2xmlindata/vmid.xml | 19 +++++++++++++++ tests/genericxml2xmltest.c | 2 ++ 11 files changed, 143 insertions(+) create mode 100644 tests/genericxml2xmlindata/vmid.xml -- 2.31.1

3 16

Re: [PATCH for-6.1? v2 5/7] job: Add job_cancel_requested()
by Kevin Wolf 04 Aug '21

04 Aug '21

[ Peter, the question for you is at the end. ] Am 04.08.2021 um 10:07 hat Max Reitz geschrieben: > On 03.08.21 16:25, Kevin Wolf wrote: > > Am 26.07.2021 um 16:46 hat Max Reitz geschrieben: > > > Most callers of job_is_cancelled() actually want to know whether the job > > > is on its way to immediate termination. For example, we refuse to pause > > > jobs that are cancelled; but this only makes sense for jobs that are > > > really actually cancelled. > > > > > > A mirror job that is cancelled during READY with force=false should > > > absolutely be allowed to pause. This "cancellation" (which is actually > > > a kind of completion) may take an indefinite amount of time, and so > > > should behave like any job during normal operation. For example, with > > > on-target-error=stop, the job should stop on write errors. (In > > > contrast, force-cancelled jobs should not get write errors, as they > > > should just terminate and not do further I/O.) > > > > > > Therefore, redefine job_is_cancelled() to only return true for jobs that > > > are force-cancelled (which as of HEAD^ means any job that interprets the > > > cancellation request as a request for immediate termination), and add > > > job_cancel_request() as the general variant, which returns true for any > > > jobs which have been requested to be cancelled, whether it be > > > immediately or after an arbitrarily long completion phase. > > > > > > Buglink: https://gitlab.com/qemu-project/qemu/-/issues/462 > > > Signed-off-by: Max Reitz <mreitz(a)redhat.com> > > > --- > > > include/qemu/job.h | 8 +++++++- > > > block/mirror.c | 10 ++++------ > > > job.c | 7 ++++++- > > > 3 files changed, 17 insertions(+), 8 deletions(-) > > > > > > diff --git a/include/qemu/job.h b/include/qemu/job.h > > > index 8aa90f7395..032edf3c5f 100644 > > > --- a/include/qemu/job.h > > > +++ b/include/qemu/job.h > > > @@ -436,9 +436,15 @@ const char *job_type_str(const Job *job); > > > /** Returns true if the job should not be visible to the management layer. */ > > > bool job_is_internal(Job *job); > > > -/** Returns whether the job is scheduled for cancellation. */ > > > +/** Returns whether the job is being cancelled. */ > > > bool job_is_cancelled(Job *job); > > > +/** > > > + * Returns whether the job is scheduled for cancellation (at an > > > + * indefinite point). > > > + */ > > > +bool job_cancel_requested(Job *job); > > I don't think non-force blockdev-cancel for mirror should actually be > > considered cancellation, so what is the question that this function > > answers? > > > > "Is this a cancelled job, or a mirror block job that is supposed to > > complete soon, but only if it doesn't switch over the users to the > > target on completion"? > > Well, technically yes, but it was more intended as “Has the user ever > invoked (block-)job-cancel on this job?”. I understand this, but is this much more useful to know than "Has the user ever called HMP 'change'?", if you know what I mean? > > Is this ever a reasonable question to ask, except maybe inside the > > mirror implementation itself? > > I asked myself the same for v3, but found two places in job.c where I > would like to keep it: > > First, there’s an assertion in job_completed_txn_abort(). All jobs > other than @job have been force-cancelled, and so job_is_cancelled() > would be true for them. As for @job itself, the function is mostly > called when the job’s return value is not 0, but a soft-cancelled > mirror does have a return value of 0 and so would not end up in that > function. > But job_cancel() invokes job_completed_txn_abort() if the job has been > deferred to the main loop, which mostly correlates with the job having > been completed (in which case the assertion is skipped), but not 100 % > (there’s a small window between setting deferred_to_main_loop and the > job changing to a completed state). > So I’d prefer to keep the assertion as-is functionally, i.e. to only > check job->cancelled. Well, you don't. It's still job_is_cancelled() after this patch. So the scenario you're concerned about is a job that has just finished successfully (job->ret = 0) and then gets a cancel request? With force=false, I'm pretty sure the code is wrong anyway because calling job_completed_txn_abort() is not the right response. It should return an error because you're trying to complete twice, possibly with conflicting completion modes. Second best is just ignoring the cancel request because we obviously already fulfilled the request of completing the job (the completion mode might be different, though). With force=true, arguably still letting the job fail is correct. However, letting it fail involves more than just letting the transaction fail. We would have to call job_update_rc() as well so that instead of reporting success for the job, ECANCELED is returned and the job transitions to JOB_STATUS_ABORTING (after which job_is_completed() returns true). So, just bugs to be fixed. After this, I think we could even assert(job->ret != 0 || job->status == JOB_STATUS_PENDING) in job_completed_txn_abort(). ret == 0 can only happen when called from job_do_finalize(), when the job is only failing because other jobs in the same transaction have failed in their .prepare callback. > Second, job_complete() refuses to let a job complete that has been > cancelled. This function is basically only invoked by the user > (through qmp_block_job_complete()/qmp_job_complete(), or > job_complete_sync(), which comes from qemu-img), so I believe that it > should correspond to the external interface we have right now; i.e., > if the user has invoked (block-)job-cancel at one point, > job_complete() should generally return an error. True. But it should also return an error if the user has invoked job-complete at some point. The distinction between complete and non-force cancel doesn't make sense there. And cancelling with force=false should fail, too, when either job-cancel or job-complete was called for the job before. > > job_complete() is the only function outside of mirror that seems to use > > it. But even there, it feels wrong to make a difference. Either we > > accept redundant completion requests, or we don't. It doesn't really > > matter how the job reconfigures the graph on completion. (Also, I feel > > this should really have been part of the state machine, but I'm not sure > > if we want to touch it now...) > > Well, yes, I don’t think it makes a difference because I don’t think > anyone will first tell the job via block-job-cancel to complete > without pivoting, and then change their mind and call > block-job-complete after all. (Not least because that’s an error > pre-series.) Right, I'm just arguing that we shouldn't allow the opposite order either. Currently I think we do, and it's buggy, as explained above. > Also, I’m not even sure whether completing after a soft cancel request > works. I don’t think any of our code accounts for such a case, so I’d > rather avoid allowing it if there’s no need to allow it anyway. Yes, definitely avoid it. We should allow only one completion request (be it with job-complete or block-job-cancel) and return an error for all future completion requests for the same job. We could in theory keep allowing redundant completion requests when the completion mode doesn't conflict, but I don't see the point of that. Unless libvirt can actually issue multiple completion requests (again, this includes both (block-)job-complete and non-force block-job-cancel for mirror) for the same block job - Peter, I hope it doesn't? Kevin

3 2

[libvirt PATCH 0/6] rpm: enable modular daemons by default in Fedora 35 / RHEL 9
by Daniel P. Berrangé 04 Aug '21

04 Aug '21

Daniel P. Berrangé (6): rpm: macroize logic for restarting daemons post-transaction rpm: restart virtnwfilter/virnetworkd if configs change rpm: restart modular daemons on upgrade rpm: macroize logic for enabling/disabling daemons post/postun-install rpm: handle enabling/disabling modular daemons post/postun-install rpm: use direct remote connection for Fedora >= 35 / RHEL >= 9 libvirt.spec.in | 225 ++++++++++++++++++++++++++++++++++++++++-------- 1 file changed, 190 insertions(+), 35 deletions(-) -- 2.31.1

2 7