- Devel - libvirt List Archives

[libvirt] [jenkins-ci PATCH v2] guests: Clean up packages after update
by Andrea Bolognani 16 Apr '18

16 Apr '18

The package cache can grow to eat up a lot of disk space, and not removing unused packages can lead to upgrade issues down the line for fast moving distributions such as Fedora Rawhide. Signed-off-by: Andrea Bolognani <abologna(a)redhat.com> --- Changes from [v1]: * call 'yum autoremove' when possible. [v1] https://www.redhat.com/archives/libvir-list/2018-April/msg00970.html guests/tasks/base.yml | 28 ++++++++++++++++++++++++++++ 1 file changed, 28 insertions(+) diff --git a/guests/tasks/base.yml b/guests/tasks/base.yml index 53cbd65..616db0a 100644 --- a/guests/tasks/base.yml +++ b/guests/tasks/base.yml @@ -117,6 +117,34 @@ when: - package_format == 'pkg' +- name: Clean up packages after update + command: yum clean packages -y + args: + warn: no + when: + - package_format == 'rpm' + +- name: Clean up packages after update + command: yum autoremove -y + args: + warn: no + when: + - package_format == 'rpm' + - not ( os_name == 'CentOS' and + os_version == '6' ) + +- name: Clean up packages after update + apt: + autoclean: yes + autoremove: yes + when: + - package_format == 'deb' + +- name: Clean up packages after update + shell: pkg clean -y && pkg autoremove -y + when: + - package_format == 'pkg' + - name: Configure hostname hostname: name: '{{ inventory_hostname }}' -- 2.14.3

2 1

[libvirt] [jenkins-ci PATCH] guests: Introduce libvirt+mingw project
by Andrea Bolognani 16 Apr '18

16 Apr '18

Instead of cramming the dependencies that we need for a MinGW build of libvirt along with those we need for a native build, create a new project for the former. Signed-off-by: Andrea Bolognani <abologna(a)redhat.com> --- guests/host_vars/libvirt-fedora-rawhide/main.yml | 1 + guests/vars/projects/libvirt+mingw.yml | 28 ++++++++++++++++++++++++ guests/vars/projects/libvirt.yml | 24 -------------------- 3 files changed, 29 insertions(+), 24 deletions(-) create mode 100644 guests/vars/projects/libvirt+mingw.yml diff --git a/guests/host_vars/libvirt-fedora-rawhide/main.yml b/guests/host_vars/libvirt-fedora-rawhide/main.yml index 3f69f80..c905fb0 100644 --- a/guests/host_vars/libvirt-fedora-rawhide/main.yml +++ b/guests/host_vars/libvirt-fedora-rawhide/main.yml @@ -5,6 +5,7 @@ PYTHONPATH: $VIRT_PREFIX/lib64/python3.6/site-packages projects: - libosinfo - libvirt + - libvirt+mingw - libvirt-cim - libvirt-dbus - libvirt-glib diff --git a/guests/vars/projects/libvirt+mingw.yml b/guests/vars/projects/libvirt+mingw.yml new file mode 100644 index 0000000..0dacb30 --- /dev/null +++ b/guests/vars/projects/libvirt+mingw.yml @@ -0,0 +1,28 @@ +--- +packages: + - mingw32-curl + - mingw32-dbus + - mingw32-dlfcn + - mingw32-gcc + - mingw32-gettext + - mingw32-gnutls + - mingw32-libssh2 + - mingw32-libxml2 + - mingw32-openssl + - mingw32-pkg-config + - mingw32-portablexdr + - mingw32-readline + - mingw64-curl + - mingw64-dbus + - mingw64-dlfcn + - mingw64-gcc + - mingw64-gettext + - mingw64-gnutls + - mingw64-libssh2 + - mingw64-libxml2 + - mingw64-openssl + - mingw64-pkg-config + - mingw64-portablexdr + - mingw64-readline + - rpcgen + - xsltproc diff --git a/guests/vars/projects/libvirt.yml b/guests/vars/projects/libvirt.yml index 598dfc4..ad42fd5 100644 --- a/guests/vars/projects/libvirt.yml +++ b/guests/vars/projects/libvirt.yml @@ -36,30 +36,6 @@ packages: - libudev - libxml2 - lvm2 - - mingw32-curl - - mingw32-dbus - - mingw32-dlfcn - - mingw32-gcc - - mingw32-gettext - - mingw32-gnutls - - mingw32-libssh2 - - mingw32-libxml2 - - mingw32-openssl - - mingw32-pkg-config - - mingw32-portablexdr - - mingw32-readline - - mingw64-curl - - mingw64-dbus - - mingw64-dlfcn - - mingw64-gcc - - mingw64-gettext - - mingw64-gnutls - - mingw64-libssh2 - - mingw64-libxml2 - - mingw64-openssl - - mingw64-pkg-config - - mingw64-portablexdr - - mingw64-readline - netcf - numad - openwsman -- 2.14.3

2 1

[libvirt] [dbus PATCH] tests: Avoid hidding TimeoutError in main loop
by Katerina Koukiou 16 Apr '18

16 Apr '18

Currently the error was not propagated from the timeout handler properly, and we had PASSED tests even with timeouts. Signed-off-by: Katerina Koukiou <kkoukiou(a)redhat.com> --- test/libvirttest.py | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/test/libvirttest.py b/test/libvirttest.py index 7168e66..d1b71cc 100644 --- a/test/libvirttest.py +++ b/test/libvirttest.py @@ -60,11 +60,14 @@ class BaseTestClass(): def timeout(): self.loop.quit() del self.loop - raise TimeoutError() + self.timeout = True + self.timeout = False self.loop = GLib.MainLoop() GLib.timeout_add(2000, timeout) self.loop.run() + if self.timeout: + raise TimeoutError() def domain(self): path = self.connect.ListDomains(0)[0] -- 2.15.0

2 1

[libvirt] [PATCHv3 0/6] Use query-cpus-fast instead of query-cpus
by Viktor Mihajlovski 16 Apr '18

16 Apr '18

The QEMU monitor commmand query-cpus is deprecated starting with QEMU 2.12.0 because it can adversely affect the performance of a running virtual machine. This series enables libvirt to use the new query-cpus-fast interface if supported by the local QEMU instance and is required in order to support QEMU once the interface has been removed. query-cpus-fast doesn't return the halted state for a virtual CPU, meaning that the vcpu.<n>.halted value won't be reported with 'virsh domstats' anymore. This is OK, as stats values are not guaranteed to be reported under all circumstances and on all architectures. Upstream discussion consensus was that the halted state was problematic anyway, as it had different semantics per architecture. The only known exploitation happened for s390, for this architecture the halted state will be computed based on an architecture-specific cpu value returned in query-cpus-fast. v2 -> v3: ========= Patch changes: 1/6: - formatting changes as requested by John Ferlan - updated the qemucapabilitiestest XML files - dropped A/R-bys (due to change above) 2/6: - fixed comment to also account for query-cpus-fast 3/6: - updated json monitor tests to account for props returned by query-cpus and query-cpus-fast - updated json monitor tests to handle the halted property (moved here from patch 5) 4/6: - removed stale text (about callbacks) from commit message - simplified qemuMonitorJSONExtractCPUArchInfo call - enhanced sample JSON response in comment - fixed inter-function spacing 5/6: - kept s390-specific tests and moved the rest to patch 3 - reformatted the JSON responses (indented "props") - dropped R-b v1 -> v2: ========= Patch changes: 1/6: - add Acked-by: Peter Krempa 2/6: - evaluate capability outside of monitor code (changes signatures of qemuMonitorGetCPUInfo and qemuMonitorGetCpuHalted) - remove ternary expressions as requested by review - remove Reviewed-bys due to code changes 3/6 - adapt test cases to changes above - remove Reviewed-bys due to code changes 4/6 - replace callbacks for architecture data extraction with direct function calls - remove Reviewed-bys due to code changes Viktor Mihajlovski (6): qemu: add capability detection for query-cpus-fast qemu: use query-cpus-fast in JSON monitor tests: add qemumonitorjson tests for query-cpus-fast qemu: add architecture-specific CPU info handling tests: add testcase for s390 query-cpus-fast qemu: refresh vcpu halted state only via query-cpus-fast src/qemu/qemu_capabilities.c | 2 + src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_domain.c | 23 +++- src/qemu/qemu_monitor.c | 30 +++-- src/qemu/qemu_monitor.h | 7 +- src/qemu/qemu_monitor_json.c | 133 +++++++++++++++++++-- src/qemu/qemu_monitor_json.h | 3 +- tests/qemucapabilitiesdata/caps_2.12.0.aarch64.xml | 1 + tests/qemucapabilitiesdata/caps_2.12.0.ppc64.xml | 1 + tests/qemucapabilitiesdata/caps_2.12.0.s390x.xml | 1 + tests/qemucapabilitiesdata/caps_2.12.0.x86_64.xml | 1 + .../qemumonitorjson-cpuinfo-ppc64-hotplug-4.data | 8 ++ .../qemumonitorjson-cpuinfo-s390-fast-cpus.json | 25 ++++ .../qemumonitorjson-cpuinfo-s390-fast-hotplug.json | 21 ++++ .../qemumonitorjson-cpuinfo-s390-fast.data | 19 +++ ...emumonitorjson-cpuinfo-x86-basic-pluggable.data | 5 + ...qemumonitorjson-cpuinfo-x86-full-fast-cpus.json | 126 +++++++++++++++++++ ...umonitorjson-cpuinfo-x86-full-fast-hotplug.json | 115 ++++++++++++++++++ .../qemumonitorjson-cpuinfo-x86-full-fast.data | 109 +++++++++++++++++ .../qemumonitorjson-cpuinfo-x86-node-full.data | 2 + tests/qemumonitorjsontest.c | 123 +++++++++++++++---- tests/qemumonitortestutils.c | 7 ++ tests/qemumonitortestutils.h | 1 + 23 files changed, 706 insertions(+), 58 deletions(-) create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-cpuinfo-s390-fast-cpus.json create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-cpuinfo-s390-fast-hotplug.json create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-cpuinfo-s390-fast.data create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-cpuinfo-x86-full-fast-cpus.json create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-cpuinfo-x86-full-fast-hotplug.json create mode 100644 tests/qemumonitorjsondata/qemumonitorjson-cpuinfo-x86-full-fast.data -- 1.9.1

3 11

[libvirt] [PATCH] qemu: remove obsolete qemuMonitorMigrateToCommand
by Daniel P. Berrangé 15 Apr '18

15 Apr '18

The last use of qemuMonitorMigrateToCommand was removed years back in commit 2e90c9daf9eabe15d826a48c06607acb9d18fda6 Author: Daniel P. Berrange <berrange(a)redhat.com> Date: Fri Nov 6 16:50:26 2015 +0000 qemu: assume support for all migration protocols except rdma Prior to that commit, 'exec:' to used to replicate the 'unix:' protocol by spawning 'nc'. Signed-off-by: Daniel P. Berrangé <berrange(a)redhat.com> --- src/qemu/qemu_monitor.c | 31 ------------------------------- src/qemu/qemu_monitor.h | 4 ---- 2 files changed, 35 deletions(-) diff --git a/src/qemu/qemu_monitor.c b/src/qemu/qemu_monitor.c index 7b647525b3..22f05222db 100644 --- a/src/qemu/qemu_monitor.c +++ b/src/qemu/qemu_monitor.c @@ -2730,37 +2730,6 @@ qemuMonitorMigrateToHost(qemuMonitorPtr mon, } -int -qemuMonitorMigrateToCommand(qemuMonitorPtr mon, - unsigned int flags, - const char * const *argv) -{ - char *argstr; - char *dest = NULL; - int ret = -1; - VIR_DEBUG("argv=%p flags=0x%x", argv, flags); - - QEMU_CHECK_MONITOR(mon); - - argstr = virArgvToString(argv); - if (!argstr) - goto cleanup; - - if (virAsprintf(&dest, "exec:%s", argstr) < 0) - goto cleanup; - - if (mon->json) - ret = qemuMonitorJSONMigrate(mon, flags, dest); - else - ret = qemuMonitorTextMigrate(mon, flags, dest); - - cleanup: - VIR_FREE(argstr); - VIR_FREE(dest); - return ret; -} - - int qemuMonitorMigrateCancel(qemuMonitorPtr mon) { diff --git a/src/qemu/qemu_monitor.h b/src/qemu/qemu_monitor.h index d04148e568..9556a51341 100644 --- a/src/qemu/qemu_monitor.h +++ b/src/qemu/qemu_monitor.h @@ -784,10 +784,6 @@ int qemuMonitorMigrateToHost(qemuMonitorPtr mon, const char *hostname, int port); -int qemuMonitorMigrateToCommand(qemuMonitorPtr mon, - unsigned int flags, - const char * const *argv); - int qemuMonitorMigrateCancel(qemuMonitorPtr mon); int qemuMonitorGetDumpGuestMemoryCapability(qemuMonitorPtr mon, -- 2.14.3

2 1

[libvirt] [PATCH] apparmor: add rules to use qemu_bridge_helper
by Christian Ehrhardt 14 Apr '18

14 Apr '18

Due to mediation of socket and signal activity currently qemu:///session connections calling qemu_bridge_helper fail. We need the profile for libvirtd itself and the subprofile for qemu-bridge-helper to be able to talk/notify to each other via unix socket and signals. Fixes: https://bugs.launchpad.net/ubuntu/+source/libvirt/+bug/1754871 Signed-off-by: Christian Ehrhardt <christian.ehrhardt(a)canonical.com> --- examples/apparmor/usr.sbin.libvirtd | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/examples/apparmor/usr.sbin.libvirtd b/examples/apparmor/usr.sbin.libvirtd index be4fabf..3102cab 100644 --- a/examples/apparmor/usr.sbin.libvirtd +++ b/examples/apparmor/usr.sbin.libvirtd @@ -65,6 +65,10 @@ signal (read, send) peer=libvirt-*, signal (send) set=("kill", "term") peer=unconfined, + # For communication/control to qemu-bridge-helper + unix (send, receive) type=stream addr=none peer=(label=/usr/sbin/libvirtd//qemu_bridge_helper), + signal (send) set=("term") peer=/usr/sbin/libvirtd//qemu_bridge_helper, + # Very lenient profile for libvirtd since we want to first focus on confining # the guests. Guests will have a very restricted profile. / r, @@ -112,6 +116,10 @@ network inet stream, + # For communication/control from libvirtd + unix (send, receive) type=stream addr=none peer=(label=/usr/sbin/libvirtd), + signal (receive) set=("term") peer=/usr/sbin/libvirtd, + /dev/net/tun rw, /etc/qemu/** r, owner @{PROC}/*/status r, -- 2.7.4

3 3

[libvirt] [PATCH] acl: remove various left over Xen ACL whitelist entries
by Daniel P. Berrangé 13 Apr '18

13 Apr '18

The legacy xen driver is removed, so these ACLL hacks can be removed too now. Signed-off-by: Daniel P. Berrangé <berrange(a)redhat.com> --- src/check-aclrules.pl | 23 ++++------------------- 1 file changed, 4 insertions(+), 19 deletions(-) diff --git a/src/check-aclrules.pl b/src/check-aclrules.pl index 5b6c711dc8..374abaa722 100755 --- a/src/check-aclrules.pl +++ b/src/check-aclrules.pl @@ -61,26 +61,12 @@ my %whitelist = ( "interfaceClose" => 1, ); -# Temp hack - remove it once xen driver is fixed +# XXX this vzDomainMigrateConfirm3Params looks +# bogus - determine why it doesn't have a valid +# ACL check. my %implwhitelist = ( - "xenUnifiedDomainRestore" => 1, - "xenUnifiedDomainRestoreFlags" => 1, - "xenUnifiedDomainMigratePrepare" => 1, - "xenUnifiedNodeDeviceDettach" => 1, - "xenUnifiedNodeDeviceDetachFlags" => 1, - "xenUnifiedNodeDeviceReset" => 1, - "xenUnifiedDomainIsActive" => 1, - "xenUnifiedDomainIsPersistent" => 1, - "xenUnifiedDomainIsUpdated" => 1, - "xenUnifiedDomainOpenConsole" => 1, "vzDomainMigrateConfirm3Params" => 1, ); -my %filterimplwhitelist = ( - "xenUnifiedConnectListDomains" => 1, - "xenUnifiedConnectNumOfDomains" => 1, - "xenUnifiedConnectListDefinedDomains" => 1, - "xenUnifiedConnectNumOfDefinedDomains" => 1, - ); my $lastfile; @@ -234,8 +220,7 @@ while (<>) { } if (exists $filtered{$api} && - !exists $aclfilters{$impl} && - !exists $filterimplwhitelist{$impl}) { + !exists $aclfilters{$impl}) { print "$ARGV:$. Missing ACL filter in function '$impl' for '$api'\n"; $status = 1; } -- 2.14.3

2 1

[libvirt] [PATCH 0/2] Fix vmware driver for VMware Fusion 10
by Rainer Müller 13 Apr '18

13 Apr '18

These changes were required to get VMware Fusion working at all. I tested with VMware Fusion 10.1.1 on macOS 10.12 Sierra. I am not sure whether calling virCapabilitiesInitCaches() makes sense at all on macOS. This function looks highly specific to Linux as it relies on /sys/devices/system, so it should probably be wrapped with an appropriate #ifdef. I do not feel proficient enough with the libvirt internals to make this change, though. Rainer Müller (2): vmware: Fix initialization of VMware Fusion vmware: Failures in cache info init are non-fatal src/vmware/vmware_conf.c | 7 ++++++- src/vmware/vmware_driver.c | 9 ++++++++- 2 files changed, 14 insertions(+), 2 deletions(-) -- 2.16.3

2 5

[libvirt] [PATCH jenkins-ci 0/2] Enable virt-viewer mingw build
by Daniel P. Berrangé 13 Apr '18

13 Apr '18

Daniel P. Berrangé (2): Enable mingw build for libvirt-glib project Enable mingw build for virt-viewer project guests/vars/mappings.yml | 72 +++++++++++++++++++++++++++++++++++ guests/vars/projects/libvirt-glib.yml | 2 + guests/vars/projects/virt-viewer.yml | 24 ++++++++++++ projects/libvirt-glib.yaml | 4 ++ projects/virt-viewer.yaml | 4 ++ 5 files changed, 106 insertions(+) -- 2.14.3

2 16

[libvirt] [RFC] passing standard input to host bootloaders
by Fabian Freyer 13 Apr '18

13 Apr '18

Hello list, Some host boot loaders, e.g. grub-bhyve when using the bhyve driver, take commands on stdin. While there is the <bootloader_args> tag to set arbitrary command line tags, there is no <bootloader_stdin> or similar to provide standard input to the boot loader. Typical input could be something along the lines of e.g. the following grub commands: kernel (cd)/path/to/vmlinuz some-cmdline initrd (root)/path/to/host/initrd boot Using e.g. the (root) device is especially useful on the bhyve driver when creating diskless VMs. Before I start implementing this, I’d appreciate some feedback on the following two points: 1) should this be an attr on the <bootloader> tag, e.g. <bootloader_stdin=“/path/to/file”>/path/to/bootloader</bootloader> or rather, as there already exists a <bootloader_args> tag a separate <bootloader_stdin> tag? 2) should the input be passed verbatim in the domain def, e.g. using a CDATA block or from a file? Should this be user-specified, e.g. in an attr? <bootloader_stdin from=“/path/to/file”/> <bootloader_stdin> <![CDATA[ kernel (host)/path/to/kernel with cmdline initrd (host)/path/to/ramdisk boot ]]> </bootloader_stdin> Regards, Fabian

1 0