- Devel - libvirt List Archives

[libvirt] [PATCH] qemu: Minimalize global driver accesses
by Michal Privoznik 27 Oct '16

27 Oct '16

Whilst working on another issue, I've noticed that in some functions we have a local @driver variable among with access to global @qemu_driver variable. This makes no sense. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- src/qemu/qemu_driver.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index bacabd2..3495cbf 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -3831,7 +3831,7 @@ qemuDomainScreenshot(virDomainPtr dom, } unlink_tmp = true; - virSecurityManagerSetSavedStateLabel(qemu_driver->securityManager, vm->def, tmp); + virSecurityManagerSetSavedStateLabel(driver->securityManager, vm->def, tmp); qemuDomainObjEnterMonitor(driver, vm); if (qemuMonitorScreendump(priv->mon, tmp) < 0) { @@ -11449,7 +11449,7 @@ qemuDomainMemoryPeek(virDomainPtr dom, goto endjob; } - virSecurityManagerSetSavedStateLabel(qemu_driver->securityManager, vm->def, tmp); + virSecurityManagerSetSavedStateLabel(driver->securityManager, vm->def, tmp); priv = vm->privateData; qemuDomainObjEnterMonitor(driver, vm); -- 2.8.4

2 1

[libvirt] [PATCH] virsh: Fix on_shutdown to on_poweroff
by Martin Kletzander 27 Oct '16

27 Oct '16

The 'on_shutdown' was the name in Xen, there was never such option in libvirt's domain XML. Reported-by: Ruben Kerkhof <ruben(a)rubenkerkhof.com> Signed-off-by: Martin Kletzander <mkletzan(a)redhat.com> --- Pushed as trivial. tools/virsh.pod | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/virsh.pod b/tools/virsh.pod index 8236a36fc2af..f278fecafde3 100644 --- a/tools/virsh.pod +++ b/tools/virsh.pod @@ -2398,7 +2398,7 @@ succeed, and may take a variable length of time depending on what services must be shutdown in the domain. The exact behavior of a domain when it shuts down is set by the -I<on_shutdown> parameter in the domain's XML definition. +I<on_poweroff> parameter in the domain's XML definition. If I<domain> is transient, then the metadata of any snapshots will be lost once the guest stops running, but the snapshot contents still -- 2.10.1

1 0

[libvirt] [PATCH 1/1] qemu: host NUMA hugepage policy without guest NUMA
by Sam Bobroff 27 Oct '16

27 Oct '16

At the moment, guests that are backed by hugepages in the host are only able to use policy to control the placement of those hugepages on a per-(guest-)CPU basis. Policy applied globally is ignored. Such guests would use <memoryBacking><hugepages/></memoryBacking> and a <numatune> block with <memory mode=... nodeset=.../> but no <memnode .../> elements. This patch corrects this by, in this specific case, changing the QEMU command line from "-mem-prealloc -mem-path=..." (which cannot specify NUMA policy) to "-object memory-backend-file ..." (which can). Note: This is not visible to the guest and does not appear to create a migration incompatibility. Signed-off-by: Sam Bobroff <sam.bobroff(a)au1.ibm.com> --- There was some discussion leading up to this patch, here: https://www.redhat.com/archives/libvir-list/2016-October/msg00033.html During that discussion it seemed that fixing this issue would cause migration incompatibility but after some careful testing, it appears that it only does when a memory-backend object is attached to a guest NUMA node and that is not the case here. If only one is created, and used globally (not attached via mem=<id>), the migration data does not seem to be changed and so it seems reasonable to try something like this patch. This patch does work for my test cases but I don't claim a deep understanding of the libvirt code so this is at least partly a RFC. Comments welcome :-) Cheers, Sam. src/qemu/qemu_command.c | 55 ++++++++++++++++++++++++++++++++++++++++--------- src/qemu/qemu_command.h | 2 +- 2 files changed, 46 insertions(+), 11 deletions(-) diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index 0804133..c28c8f2 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -3143,7 +3143,7 @@ qemuBuildMemoryBackendStr(unsigned long long size, int guestNode, virBitmapPtr userNodeset, virBitmapPtr autoNodeset, - virDomainDefPtr def, + const virDomainDefPtr def, virQEMUCapsPtr qemuCaps, virQEMUDriverConfigPtr cfg, const char **backendType, @@ -7129,12 +7129,18 @@ qemuBuildSmpCommandLine(virCommandPtr cmd, static int qemuBuildMemPathStr(virQEMUDriverConfigPtr cfg, - const virDomainDef *def, + const virDomainDefPtr def, virQEMUCapsPtr qemuCaps, - virCommandPtr cmd) + virCommandPtr cmd, + virBitmapPtr auto_nodeset) { const long system_page_size = virGetSystemPageSizeKB(); char *mem_path = NULL; + virBitmapPtr nodemask = NULL; + const char *backendType = NULL; + char *backendStr = NULL; + virJSONValuePtr props; + int rv = -1; /* * No-op if hugepages were not requested. @@ -7159,18 +7165,47 @@ qemuBuildMemPathStr(virQEMUDriverConfigPtr cfg, if (qemuGetHupageMemPath(cfg, def->mem.hugepages[0].size, &mem_path) < 0) return -1; - virCommandAddArgList(cmd, "-mem-prealloc", "-mem-path", mem_path, NULL); + if (virDomainNumatuneMaybeGetNodeset(def->numa, auto_nodeset, + &nodemask, -1) < 0) + return -1; + if (nodemask && virQEMUCapsGet(qemuCaps, QEMU_CAPS_OBJECT_MEMORY_FILE)) { + props = virJSONValueNewObject(); + if (qemuBuildMemoryBackendStr(virDomainDefGetMemoryInitial(def), + 0, -1, NULL, auto_nodeset, + def, qemuCaps, cfg, &backendType, + &props, false) < 0) + goto cleanup; + if (!(backendStr = virQEMUBuildObjectCommandlineFromJSON(backendType, + "mem", + props))) + goto cleanup; + virCommandAddArgList(cmd, "-object", backendStr, NULL); + rv = 0; +cleanup: + VIR_FREE(backendStr); + VIR_FREE(props); + } + else { + if (nodemask || 1) + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", + _("Memory file backend objects are " + "unsupported by QEMU binary. Global NUMA " + "hugepage policy will be ignored.")); + virCommandAddArgList(cmd, "-mem-prealloc", "-mem-path", mem_path, NULL); + rv = 0; + } VIR_FREE(mem_path); - return 0; + return rv; } static int qemuBuildMemCommandLine(virCommandPtr cmd, virQEMUDriverConfigPtr cfg, - const virDomainDef *def, - virQEMUCapsPtr qemuCaps) + const virDomainDefPtr def, + virQEMUCapsPtr qemuCaps, + virBitmapPtr auto_nodeset) { if (qemuDomainDefValidateMemoryHotplug(def, qemuCaps, NULL) < 0) return -1; @@ -7194,7 +7229,7 @@ qemuBuildMemCommandLine(virCommandPtr cmd, * there is no numa node specified. */ if (!virDomainNumaGetNodeCount(def->numa) && - qemuBuildMemPathStr(cfg, def, qemuCaps, cmd) < 0) + qemuBuildMemPathStr(cfg, def, qemuCaps, cmd, auto_nodeset) < 0) return -1; if (def->mem.locked && !virQEMUCapsGet(qemuCaps, QEMU_CAPS_REALTIME_MLOCK)) { @@ -7331,7 +7366,7 @@ qemuBuildNumaArgStr(virQEMUDriverConfigPtr cfg, } if (!needBackend && - qemuBuildMemPathStr(cfg, def, qemuCaps, cmd) < 0) + qemuBuildMemPathStr(cfg, def, qemuCaps, cmd, auto_nodeset) < 0) goto cleanup; for (i = 0; i < ncells; i++) { @@ -9381,7 +9416,7 @@ qemuBuildCommandLine(virQEMUDriverPtr driver, if (!migrateURI && !snapshot && qemuDomainAlignMemorySizes(def) < 0) goto error; - if (qemuBuildMemCommandLine(cmd, cfg, def, qemuCaps) < 0) + if (qemuBuildMemCommandLine(cmd, cfg, def, qemuCaps, nodeset) < 0) goto error; if (qemuBuildSmpCommandLine(cmd, def) < 0) diff --git a/src/qemu/qemu_command.h b/src/qemu/qemu_command.h index 2f2a6ff..1e97d76 100644 --- a/src/qemu/qemu_command.h +++ b/src/qemu/qemu_command.h @@ -126,7 +126,7 @@ int qemuBuildMemoryBackendStr(unsigned long long size, int guestNode, virBitmapPtr userNodeset, virBitmapPtr autoNodeset, - virDomainDefPtr def, + const virDomainDefPtr def, virQEMUCapsPtr qemuCaps, virQEMUDriverConfigPtr cfg, const char **backendType, -- 2.10.0.297.gf6727b0

3 12

[libvirt] on shutdown
by Ruben Kerkhof 27 Oct '16

27 Oct '16

Hi all, virsh(1) has this to say about virsh shutdown: "The exact behavior of a domain when it shuts down is set by the on_shutdown parameter in the domain's XML definition." on_shutdown isn't documented in https://libvirt.org/formatdomain.html Is the virsh(1) manpage wrong? Kind regards, Ruben Kerkhof

2 4

[libvirt] [PATCH] Resuming running domain raise error message now
by Sławek Kapłoński 27 Oct '16

27 Oct '16

This patch introduce raising XML error when user tries to resume domain (Qemu or LXC) which is already running. Closes bug: https://bugzilla.redhat.com/show_bug.cgi?id=1009008 Sławek Kapłoński (1): Resuming running domain raise error message src/lxc/lxc_driver.c | 8 +++++++- src/qemu/qemu_driver.c | 4 ++++ 2 files changed, 11 insertions(+), 1 deletion(-) -- 2.10.1

2 5

[libvirt] [PATCH 0/5] wireshark: Fix $(prefix) handling
by Andrea Bolognani 27 Oct '16

27 Oct '16

Well, almost :) There are still some cases that are not handled correctly, but at least this will unbreak 'make rpm' while I work on the rest. Tested by running 'make rpm' successfully on Fedora 23, Fedora 24 and Fedora rawhide. Andrea Bolognani (5): wireshark: Introduce $ws_modversion wireshark: Hoist $ws_prefix declaration wireshark: Strip prefix correctly wireshark: Inject $(prefix) at the right time wireshark: Rename plugindir to ws_plugindir m4/virt-wireshark.m4 | 29 ++++++++++++++++++----------- tools/Makefile.am | 2 +- 2 files changed, 19 insertions(+), 12 deletions(-) -- 2.7.4

3 8

[libvirt] [PATCH] virt-wireshark.m4: Defer $(prefix) substitution
by Michal Privoznik 27 Oct '16

27 Oct '16

The problem with evaluating $(prefix) in configure phase is that autoconf does a lot of magic with this variable. Firstly, if the --prefix argument is not set for the configure script, the variable will have value "NONE". Wait, wat? Secondly, even autoconf scripts then have special cases for when variable is "NONE". It's even worse - whenever they need to construct new path based on prefix, they save its current value, substitute it with proper value (like /usr/local/), do what they need and then restore the original value. Le sigh. This mad behaviour is, however, suppressed if there's --prefix on the configure command line. Now, the problem is that we tried to construct path for wireshark plugin in the m4 file. Based on the knowledge described above, we shouldn't do it and just construct the full path in Makefile later where prefix has proper value. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- m4/virt-wireshark.m4 | 2 +- tools/Makefile.am | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/m4/virt-wireshark.m4 b/m4/virt-wireshark.m4 index f383e2b..e1e4a59 100644 --- a/m4/virt-wireshark.m4 +++ b/m4/virt-wireshark.m4 @@ -38,7 +38,7 @@ AC_DEFUN([LIBVIRT_CHECK_WIRESHARK],[ if test "x$ws_prefix" = "x" ; then ws_prefix="/usr"; fi - plugindir="${prefix}${plugindir#$ws_prefix}" + plugindir="${plugindir#$ws_prefix}" fi elif test "x$with_ws_plugindir" = "xno" || test "x$with_ws_plugindir" = "xyes"; then AC_MSG_ERROR([ws-plugindir must be used only with valid path]) diff --git a/tools/Makefile.am b/tools/Makefile.am index e7e42c3..08e1680 100644 --- a/tools/Makefile.am +++ b/tools/Makefile.am @@ -398,7 +398,7 @@ EXTRA_DIST += \ if WITH_WIRESHARK_DISSECTOR -ws_plugindir = $(plugindir) +ws_plugindir = $(prefix)$(plugindir) ws_plugin_LTLIBRARIES = wireshark/src/libvirt.la wireshark_src_libvirt_la_CPPFLAGS = \ -I wireshark/src $(WIRESHARK_DISSECTOR_CFLAGS) -- 2.8.4

4 6

[libvirt] [PATCH v11 0/5] Add native TLS encrypted chardev TCP support
by John Ferlan 26 Oct '16

26 Oct '16

v10: http://www.redhat.com/archives/libvir-list/2016-October/msg00866.html Changes since v10... This is essentially the adding the secret to/for the TLS chardev TCP. Lots of messy stuff. Patches 1-3 are new... In particular, I kept looking at the RNG removal code and kept thinking, something just doesn't look right when compared to the RemoveVirtioDisk code. If I'm incorrect with order of removal it's an easy enough change. Kept trying to think what depends on what and for the RNG it would seem a need a to unplug the backend before removing the front end object would be more logical. Patches 4-5 are the old patch 3-4. They (hopefully) encompass the previous review plus other things since discovered. John Ferlan (5): qemu: Move TLS object remove from DetachChr to RemoveChr qemu: Swap order of RNG hot unplug removal qemu: Need to remove TLS object in RemoveRNGDevice qemu: Add a secret object to/for a char source dev qemu: Add secret object hotplug for TCP chardev TLS src/qemu/qemu_command.c | 30 +++- src/qemu/qemu_command.h | 1 + src/qemu/qemu_domain.c | 173 +++++++++++++++++- src/qemu/qemu_domain.h | 17 +- src/qemu/qemu_driver.c | 6 +- src/qemu/qemu_hotplug.c | 195 +++++++++++++++++---- src/qemu/qemu_hotplug.h | 9 +- src/qemu/qemu_process.c | 9 +- tests/qemuhotplugtest.c | 2 +- ...xml2argv-serial-tcp-tlsx509-secret-chardev.args | 38 ++++ ...uxml2argv-serial-tcp-tlsx509-secret-chardev.xml | 50 ++++++ tests/qemuxml2argvtest.c | 17 ++ 12 files changed, 495 insertions(+), 52 deletions(-) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-serial-tcp-tlsx509-secret-chardev.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-serial-tcp-tlsx509-secret-chardev.xml -- 2.7.4

2 16

[libvirt] [PATCH v5 00/15] Use more PCIe less legacy PCI
by Laine Stump 26 Oct '16

26 Oct '16

These are just the rebased versions of the patches from v4 that I haven't yet pushed, to make review easier. Patch 1 and 2 (and several others) are already ACKed, but included for completeness. Laine Stump (15): qemu: new functions qemuDomainMachineHasPCI[e]Root() qemu: new functions to calculate/set device pciConnectFlags qemu: set/use info->pciConnectFlags when validating/assigning PCI addresses qemu: set/use proper pciConnectFlags during hotplug qemu: set pciConnectFlags to 0 instead of PCI|HOTPLUGGABLE if device isn't PCI qemu: assign virtio devices to PCIe slot when appropriate qemu: assign e1000e network devices to PCIe slots when appropriate qemu: assign nec-xhci (USB3) controller to a PCIe address when appropriate qemu: only force an available legacy-PCI slot on domains with pci-root qemu: auto-add pcie-root-port/dmi-to-pci-bridge controllers as needed [RFC] qemu: if pci-bridge is in PCIe config w/o dmi-to-pci-bridge, add it qemu: don't force-add a dmi-to-pci-bridge just on principle qemu: add a USB3 controller to Q35 domains by default qemu: try to put ich9 sound device at 00:1B.0 qemu: initially reserve one open pcie-root-port for hotplug src/conf/device_conf.h | 5 + src/conf/domain_addr.c | 112 ++- src/conf/domain_addr.h | 3 +- src/qemu/qemu_domain.c | 52 +- src/qemu/qemu_domain.h | 2 + src/qemu/qemu_domain_address.c | 805 +++++++++++++++++---- src/qemu/qemu_domain_address.h | 4 + src/qemu/qemu_hotplug.c | 20 +- tests/qemuxml2argvdata/qemuxml2argv-autoindex.args | 10 +- .../qemuxml2argv-bios-nvram-secure.args | 1 + .../qemuxml2argv-machine-smm-opt.args | 1 + tests/qemuxml2argvdata/qemuxml2argv-pcie-root.args | 3 +- .../qemuxml2argv-q35-default-devices-only.args | 23 + .../qemuxml2argv-q35-default-devices-only.xml | 18 + .../qemuxml2argv-q35-pcie-autoadd.args | 57 ++ .../qemuxml2argv-q35-pcie-autoadd.xml | 51 ++ tests/qemuxml2argvdata/qemuxml2argv-q35-pcie.args | 58 ++ tests/qemuxml2argvdata/qemuxml2argv-q35-pcie.xml | 67 ++ .../qemuxml2argv-q35-pm-disable-fallback.args | 1 + .../qemuxml2argv-q35-pm-disable.args | 1 + .../qemuxml2argv-q35-virt-manager-basic.args | 56 ++ .../qemuxml2argv-q35-virt-manager-basic.xml | 76 ++ .../qemuxml2argv-q35-virtio-pci.args | 58 ++ .../qemuxml2argv-q35-virtio-pci.xml | 1 + tests/qemuxml2argvtest.c | 164 ++++- .../qemuxml2xmlout-autoindex.xml | 10 +- .../qemuxml2xmlout-pcie-root.xml | 4 - .../qemuxml2xmlout-q35-default-devices-only.xml | 45 ++ .../qemuxml2xmlout-q35-pcie-autoadd.xml | 148 ++++ .../qemuxml2xmloutdata/qemuxml2xmlout-q35-pcie.xml | 152 ++++ .../qemuxml2xmlout-q35-virt-manager-basic.xml | 121 ++++ .../qemuxml2xmlout-q35-virtio-pci.xml | 152 ++++ tests/qemuxml2xmltest.c | 136 +++- 33 files changed, 2189 insertions(+), 228 deletions(-) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-q35-default-devices-only.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-q35-default-devices-only.xml create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-q35-pcie-autoadd.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-q35-pcie-autoadd.xml create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-q35-pcie.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-q35-pcie.xml create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-q35-virt-manager-basic.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-q35-virt-manager-basic.xml create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-q35-virtio-pci.args create mode 120000 tests/qemuxml2argvdata/qemuxml2argv-q35-virtio-pci.xml create mode 100644 tests/qemuxml2xmloutdata/qemuxml2xmlout-q35-default-devices-only.xml create mode 100644 tests/qemuxml2xmloutdata/qemuxml2xmlout-q35-pcie-autoadd.xml create mode 100644 tests/qemuxml2xmloutdata/qemuxml2xmlout-q35-pcie.xml create mode 100644 tests/qemuxml2xmloutdata/qemuxml2xmlout-q35-virt-manager-basic.xml create mode 100644 tests/qemuxml2xmloutdata/qemuxml2xmlout-q35-virtio-pci.xml -- 2.7.4

2 18

[libvirt] [PATCH] qemu: Fix depedency order in qemuRemoveDiskDevice
by John Ferlan 26 Oct '16

26 Oct '16

Need to remove the drive first, then the secobj and/or encobj if they exist. This is because the drive has a dependency on secobj (or the secret for the networked storage server) and/or the encobj (or the secret for the LUKS encrypted volume). Deleting either object first leaves an drive without it's respective objects. Signed-off-by: John Ferlan <jferlan(a)redhat.com> --- src/qemu/qemu_hotplug.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/src/qemu/qemu_hotplug.c b/src/qemu/qemu_hotplug.c index 7a21dc6..30565d5 100644 --- a/src/qemu/qemu_hotplug.c +++ b/src/qemu/qemu_hotplug.c @@ -3176,6 +3176,9 @@ qemuDomainRemoveDiskDevice(virQEMUDriverPtr driver, qemuDomainObjEnterMonitor(driver, vm); + qemuMonitorDriveDel(priv->mon, drivestr); + VIR_FREE(drivestr); + /* If it fails, then so be it - it was a best shot */ if (objAlias) ignore_value(qemuMonitorDelObject(priv->mon, objAlias)); @@ -3186,8 +3189,6 @@ qemuDomainRemoveDiskDevice(virQEMUDriverPtr driver, ignore_value(qemuMonitorDelObject(priv->mon, encAlias)); VIR_FREE(encAlias); - qemuMonitorDriveDel(priv->mon, drivestr); - VIR_FREE(drivestr); if (qemuDomainObjExitMonitor(driver, vm) < 0) return -1; -- 2.7.4

2 1