- Devel - libvirt List Archives

[libvirt] [PATCH v2 0/2] Fix parsing and formatting of 'UnixSocketAddress' qapi type
by Peter Krempa 14 Feb '18

14 Feb '18

v2: Don't 'fix' it for gluster disks which use old syntax not conforming to the schema in qemu. Peter Krempa (2): storage: Fix formatting and parsing of qemu type 'UnixSocketAddress' virstoragetest: Add test case for NBD over unix socket with new syntax src/qemu/qemu_block.c | 13 +++++++++++-- src/util/virstoragefile.c | 8 +++++++- tests/virstoragetest.c | 9 +++++++++ 3 files changed, 27 insertions(+), 3 deletions(-) -- 2.15.0

2 3

[libvirt] [jenkins-ci PATCH] jobs: Shorten name for python-distutils jobs
by Andrea Bolognani 14 Feb '18

14 Feb '18

Instead of passing the full name of the Python binary as 'python', only pass the major version as 'pyver' and build everything else based on that. Doing so allows us to make a few things, most notably job names, slightly shorter and nicer. Signed-off-by: Andrea Bolognani <abologna(a)redhat.com> --- I wouldn't bother changing this if the previous change to Python jobs had already been deployed, but since it hasn't yet might as well have nicer names instead :) jobs/python-distutils.yaml | 22 +++++++++++----------- projects/libvirt-python.yaml | 16 ++++++++-------- projects/virt-manager.yaml | 14 +++++++------- 3 files changed, 26 insertions(+), 26 deletions(-) diff --git a/jobs/python-distutils.yaml b/jobs/python-distutils.yaml index 122c759..ff68c29 100644 --- a/jobs/python-distutils.yaml +++ b/jobs/python-distutils.yaml @@ -1,11 +1,11 @@ - job-template: id: python-distutils-build-job - name: '{name}-{branch}-{python}-build' + name: '{name}-{branch}-py{pyver}-build' project-type: matrix - description: '{title} Build ({python})' + description: '{title} Build (Python {pyver})' command_pre_build: '' - workspace: '{name}-{branch}-{python}' + workspace: '{name}-{branch}-py{pyver}' child-workspace: '.' block-downstream: true block-upstream: true @@ -43,8 +43,8 @@ {global_env} {local_env} {command_pre_build} - {python} ./setup.py build - {python} ./setup.py install --prefix=$VIRT_PREFIX + python{pyver} ./setup.py build + python{pyver} ./setup.py install --prefix=$VIRT_PREFIX publishers: - email: recipients: '{obj:spam}' @@ -54,9 +54,9 @@ - job-template: id: python-distutils-check-job - name: '{name}-{branch}-{python}-check' + name: '{name}-{branch}-py{pyver}-check' project-type: matrix - description: '{title} Check ({python})' + description: '{title} Check (Python {pyver})' workspace: '{name}-{branch}' child-workspace: '.' block-downstream: true @@ -83,7 +83,7 @@ - shell: | {global_env} {local_env} - {python} ./setup.py test + python{pyver} ./setup.py test publishers: - email: recipients: '{obj:spam}' @@ -92,9 +92,9 @@ - job-template: id: python-distutils-rpm-job - name: '{name}-{branch}-{python}-rpm' + name: '{name}-{branch}-py{pyver}-rpm' project-type: matrix - description: '{title} RPM ({python})' + description: '{title} RPM (Python {pyver})' workspace: '{name}-{branch}' child-workspace: '.' block-downstream: true @@ -122,7 +122,7 @@ {global_env} {local_env} sed -i -e 's/BuildRequires: libvirt.*devel.*//' *.spec.in - {python} ./setup.py rpm + python{pyver} ./setup.py rpm publishers: - email: recipients: '{obj:spam}' diff --git a/projects/libvirt-python.yaml b/projects/libvirt-python.yaml index e065c5d..1c29321 100644 --- a/projects/libvirt-python.yaml +++ b/projects/libvirt-python.yaml @@ -14,10 +14,10 @@ title: Libvirt Python jobs: - python-distutils-build-job: - python: python2 + pyver: 2 parent_jobs: 'libvirt-master-build' - python-distutils-build-job: - python: python3 + pyver: 3 parent_jobs: 'libvirt-master-build' machines: - libvirt-debian-8 @@ -28,11 +28,11 @@ - libvirt-freebsd-10 - libvirt-freebsd-11 - python-distutils-check-job: - python: python2 - parent_jobs: 'libvirt-python-master-{python}-build' + pyver: 2 + parent_jobs: 'libvirt-python-master-py{pyver}-build' - python-distutils-check-job: - python: python3 - parent_jobs: 'libvirt-python-master-{python}-build' + pyver: 3 + parent_jobs: 'libvirt-python-master-py{pyver}-build' machines: - libvirt-debian-8 - libvirt-debian-9 @@ -42,8 +42,8 @@ - libvirt-freebsd-10 - libvirt-freebsd-11 - python-distutils-rpm-job: - python: python2 - parent_jobs: 'libvirt-python-master-{python}-check' + pyver: 2 + parent_jobs: 'libvirt-python-master-py{pyver}-check' machines: - libvirt-centos-6 - libvirt-centos-7 diff --git a/projects/virt-manager.yaml b/projects/virt-manager.yaml index b5b0df5..8f3112e 100644 --- a/projects/virt-manager.yaml +++ b/projects/virt-manager.yaml @@ -11,18 +11,18 @@ title: Virtual Machine Manager jobs: - python-distutils-build-job: - python: python3 + pyver: 3 parent_jobs: - - 'libvirt-python-master-{python}-build' + - 'libvirt-python-master-py{pyver}-build' - 'libosinfo-master-build' command_pre_build: | - {python} ./setup.py configure --prefix=$VIRT_PREFIX + python{pyver} ./setup.py configure --prefix=$VIRT_PREFIX - python-distutils-check-job: - python: python3 - parent_jobs: 'virt-manager-master-{python}-build' + pyver: 3 + parent_jobs: 'virt-manager-master-py{pyver}-build' - python-distutils-rpm-job: - python: python3 - parent_jobs: 'virt-manager-master-{python}-check' + pyver: 3 + parent_jobs: 'virt-manager-master-py{pyver}-check' machines: - libvirt-fedora-26 - libvirt-fedora-27 -- 2.14.3

2 1

[libvirt] [PATCH 0/3] Fix parsing and formatting of 'UnixSocketAddress' qapi type
by Peter Krempa 14 Feb '18

14 Feb '18

Fix the naming mistake, add tests and remove useless comment. Peter Krempa (3): storage: Fix formatting and parsing of qemu type 'UnixSocketAddress' virstoragetest: Add test case for NBD over unix socket with new syntax qemu: block: Remove misleading part of comment in qemuBlockStorageSourceBuildJSONSocketAddress src/qemu/qemu_block.c | 7 +------ src/util/virstoragefile.c | 2 +- tests/virstoragetest.c | 13 +++++++++++-- 3 files changed, 13 insertions(+), 9 deletions(-) -- 2.15.0

2 7

[libvirt] [PATCHv2 0/1] Function declarations match definitions
by Chris Venteicher 14 Feb '18

14 Feb '18

Interested in fixing mismatches between function parameter names in declarations (header files) and definitions (.c files)? This is a cosmetic / redability fix and does not change functionality. This is patch 1 of larger set of 22 patches that fix parameter names across the libvirt source code. If there is interest I can send the entire patch set. These are cosmetic redability fixes automatically generated by clang-tidy. make check and make syntax-check pass successfully after the 22 patches are applied. This is my first submission so please point out any issues and I will be glad to fix. Chris Venteicher (1): include: function parameter names same in declaration include/libvirt/libvirt-domain.h | 18 +++++++------- include/libvirt/libvirt-event.h | 4 ++-- include/libvirt/libvirt-host.h | 4 ++-- include/libvirt/libvirt-network.h | 4 ++-- include/libvirt/libvirt-nwfilter.h | 2 +- include/libvirt/libvirt-qemu.h | 2 +- include/libvirt/libvirt-secret.h | 4 ++-- include/libvirt/libvirt-storage.h | 12 +++++----- include/libvirt/libvirt-stream.h | 22 ++++++++--------- src/libvirt-domain.c | 48 +++++++++++++++++++------------------- src/libvirt-network.c | 10 ++++---- 11 files changed, 65 insertions(+), 65 deletions(-) -- 2.14.1

4 6

[libvirt] Default USB controller model vs PCI address assignment
by Daniel P. Berrangé 13 Feb '18

13 Feb '18

GCC8 identified a critical flaw in the QEMU post-parse callback They switch() based on the virDomainControllDef 'model', carefully listing every single enum case..... but no default statement. Unfortunately it is valid for "model" to have the valid "-1" which does not map to any enum constant. So the lack of any "default" statement, means when dealing with the USB controller we mistakenly fall through to the IDE controller code. The SCSI controller simililarly falls through to the virtio serial controller case. I tried just adding the obvious default case: diff --git a/src/qemu/qemu_domain_address.c b/src/qemu/qemu_domain_address.c index e28119e4b..3eaa22e39 100644 --- a/src/qemu/qemu_domain_address.c +++ b/src/qemu/qemu_domain_address.c @@ -532,6 +532,7 @@ qemuDomainDeviceCalculatePCIConnectFlags(virDomainDeviceDefPtr dev, case VIR_DOMAIN_CONTROLLER_MODEL_USB_QUSB2: /* xen only */ case VIR_DOMAIN_CONTROLLER_MODEL_USB_NONE: case VIR_DOMAIN_CONTROLLER_MODEL_USB_LAST: + default: return 0; } @@ -553,6 +554,7 @@ qemuDomainDeviceCalculatePCIConnectFlags(virDomainDeviceDefPtr dev, return pciFlags; case VIR_DOMAIN_CONTROLLER_MODEL_SCSI_LAST: + default: return 0; } This causes many of the QEMU test cases to fail now with a message "internal error: Cannot automatically add a new PCI bus for a device with connect flags 00" I could theoretically make the "default:" case return "pciFlags" when model == -1, but this seems dubious because it feels like it is blindly assuming that any unknown controller model is PCI based, but it would be functionally equivalent to the current behaviour where USB accidentally falls through to the IDE code. Thoughts ? Regards, Daniel -- |: https://berrange.com -o- https://www.flickr.com/photos/dberrange :| |: https://libvirt.org -o- https://fstop138.berrange.com :| |: https://entangle-photo.org -o- https://www.instagram.com/dberrange :|

2 1

[libvirt] [PATCH v2 0/3] Fix compat with GCC8
by Daniel P. Berrangé 13 Feb '18

13 Feb '18

In this second version I enable -Wswitch-enum to strongly validate all switches cover all enum cases. Daniel P. Berrangé (3): Turn on -Wswitch-enum and fix all problems it identifies Fix more switch fallthrough identified by gcc8 Disable gcc8 -Wcast-function-type warnings from -Wextra m4/virt-compile-warnings.m4 | 9 ++++-- src/conf/domain_audit.c | 1 + src/conf/domain_conf.c | 47 +++++++++++++++++++++++++++---- src/conf/nwfilter_conf.c | 32 ++++++++++++++++++++- src/esx/esx_driver.c | 1 + src/esx/esx_vi.c | 11 +++++--- src/esx/esx_vi_types.c | 9 +++--- src/hyperv/hyperv_driver.c | 18 ++++++++++-- src/libxl/libxl_driver.c | 2 +- src/lxc/lxc_container.c | 7 +++-- src/lxc/lxc_controller.c | 10 ++++++- src/lxc/lxc_driver.c | 40 +++++++++++++++++++++++--- src/nwfilter/nwfilter_ebiptables_driver.c | 16 +++++++---- src/nwfilter/nwfilter_learnipaddr.c | 6 +++- src/qemu/qemu_command.c | 28 ++++++++++++------ src/qemu/qemu_domain.c | 35 ++++++++++++++++++----- src/qemu/qemu_domain_address.c | 11 ++++++++ src/qemu/qemu_driver.c | 27 ++++++++++++------ src/qemu/qemu_hotplug.c | 36 +++++++++++++++++++---- src/qemu/qemu_migration.c | 11 +++++++- src/qemu/qemu_process.c | 2 ++ src/rpc/virnetclient.c | 2 ++ src/rpc/virnetclientprogram.c | 1 + src/rpc/virnetserverprogram.c | 4 +++ src/security/security_driver.c | 1 + src/util/virconf.c | 13 ++++++++- src/util/virfirewall.c | 7 +++-- src/util/virlog.c | 10 ++++++- src/util/virnetdevvportprofile.c | 11 +++++++- src/vmx/vmx.c | 27 ++++++++++++++++-- src/xenconfig/xen_common.c | 17 +++++++++-- src/xenconfig/xen_xl.c | 8 +++++- tools/virt-host-validate-qemu.c | 3 +- 33 files changed, 389 insertions(+), 74 deletions(-) -- 2.16.1

1 4

[libvirt] [PATCH 0/6] Cleanups in QEMU driver wrt virConnectPtr
by Daniel P. Berrangé 13 Feb '18

13 Feb '18

There are many places we can stop passing around virConnectPtr now and directly open secondary drivers where required instead. Daniel P. Berrangé (6): conf: reimplement virDomainNetResolveActualType in terms of public API qemu: stop passing virConnectPtr into qemuMonitorStartCPUs conf: stop passing virConnectPtr into virDomainDiskTranslateSourcePool qemu: don't pass virConnectPtr around for secrets qemu: stop passing in virConnectPtr for looking up networks qemu: remove virConnectPtr from some more startup code paths src/conf/domain_conf.c | 90 +++++++++++++++++++++++++---- src/conf/domain_conf.h | 14 +---- src/network/bridge_driver.c | 76 +----------------------- src/qemu/qemu_conf.c | 3 +- src/qemu/qemu_conf.h | 3 +- src/qemu/qemu_domain.c | 111 +++++++++++++++++------------------ src/qemu/qemu_domain.h | 15 ++--- src/qemu/qemu_driver.c | 126 +++++++++++++++------------------------- src/qemu/qemu_hotplug.c | 66 +++++++++------------ src/qemu/qemu_hotplug.h | 15 ++--- src/qemu/qemu_migration.c | 18 +++--- src/qemu/qemu_monitor.c | 10 ++-- src/qemu/qemu_monitor.h | 11 +--- src/qemu/qemu_monitor_json.c | 3 +- src/qemu/qemu_monitor_json.h | 3 +- src/qemu/qemu_monitor_text.c | 9 +-- src/qemu/qemu_monitor_text.h | 3 +- src/qemu/qemu_process.c | 134 +++++++++++++++++-------------------------- src/qemu/qemu_process.h | 12 ++-- tests/qemuhotplugtest.c | 4 +- tests/qemumonitorjsontest.c | 2 +- 21 files changed, 298 insertions(+), 430 deletions(-) -- 2.14.3

2 12

[libvirt] Developing libvirt for virtual devices like virtual printers
by Deepti S 13 Feb '18

13 Feb '18

Hi, I have a couple of questions. - I am a new libvirt code developer. I have just cloned the repo and I am looking for a suitable debugger/IDE which will help me debug the code. - I work in developing and creating virtual printers with display, with QEMU as the hypervisor (no KVM-only Qemu). Currently we have our own application to manage the virtual printers. Our plan is to migrate to libvirt, and enable our application to use the libvirt APIs. libvirt code currently looks to be specialized in pure OS-specific guests (like PCs), and I hit snags where there are manadatory devices added which a virtual printer does not need. Are there any developers who are looking into customized virtual devices? I would like to knowledge share with them. The intention is to provide new libvirt APIs whenever there is such a snag and thus enable libvirt to be used for managing virtual printers. - One issue that we are seeing is with regard to USB Host Controller: We have a customized model of ehci USB Host Controller (named nr52_ehci_usb) in qemu (locally built). When I add the custom value (n552_ehci_usb) of the host controller as the "model" type in my dom xml, virsh define <xml> fails with invalid model type error. I remove the usb controller entry from the xml, and add "-usb" as a "qemucommandline" parameter, but then libvirt adds the device lsi (-device lsi,id=scsi0 -usb), and the virsh create command fails with: "-device lsi,id=scsi0 No PCI bus found for lsi53c895a". The board we have created in QEMU does not have PCI bus implemented. I am interested in the code related to host controllers. Is there any way I can get my usb host controller to be recognized by libvirt? Is there a way I can prevent lsi device from getting added when I make -usb a qemu command line parameter? Regards, Deepti

2 1

[libvirt] [PATCH libvirt] tests: run virshtest independent of current pwd
by Bjoern Walk 13 Feb '18

13 Feb '18

virshtest execves the virsh binary. Make sure that it finds the binary's location independent of the current working directory by specifying the absolute path as determined by the build environment. Reviewed-by: Marc Hartmayer <mhartmay(a)linux.vnet.ibm.com> Reviewed-by: Boris Fiuczynski <fiuczy(a)linux.vnet.ibm.com> Signed-off-by: Bjoern Walk <bwalk(a)linux.vnet.ibm.com> --- tests/virshtest.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tests/virshtest.c b/tests/virshtest.c index 67453bd9..94548a82 100644 --- a/tests/virshtest.c +++ b/tests/virshtest.c @@ -84,13 +84,13 @@ testCompareOutputLit(const char *expectData, return result; } -# define VIRSH_DEFAULT "../tools/virsh", \ +# define VIRSH_DEFAULT abs_topbuilddir "/tools/virsh", \ "--connect", \ "test:///default" static char *custom_uri; -# define VIRSH_CUSTOM "../tools/virsh", \ +# define VIRSH_CUSTOM abs_topbuilddir "/tools/virsh", \ "--connect", \ custom_uri -- 2.13.4

2 2

[libvirt] [RFC] virSysinfo: Introduce SMBIOS type 3 support
by Zhuangyanying 13 Feb '18

13 Feb '18

From: Zhuang Yanying <ann.zhuangyanying(a)huawei.com> Some applications inside VM need to access SMBIOS Chassis Asset Tag, which should be emulated. It has already been realized in qemu, SMBIOS: Build aggregate smbios tables and entry point https://git.qemu.org/?p=qemu.git;a=commit;h=c97294ec1b9e36887e119589d456557… but not in libvirt. we realize it here. As an example, you could use something like <chassis> <entry name='manufacturer'>LENOVO</entry> <entry name='version'>0B98401 Pro</entry> <entry name='serial'>W1KS427111E</entry> <entry name='asset'>344dfTYH#</entry> </chassis> --- src/conf/domain_conf.c | 52 +++++++++++++++++++++++++++++++++++++++++++++++++ src/qemu/qemu_command.c | 49 ++++++++++++++++++++++++++++++++++++++++++++++ src/util/virsysinfo.c | 37 +++++++++++++++++++++++++++++++++++ src/util/virsysinfo.h | 13 +++++++++++++ 4 files changed, 151 insertions(+) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 34aae82..b0d5d68 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -14506,6 +14506,47 @@ virSysinfoBaseBoardParseXML(xmlXPathContextPtr ctxt, return ret; } +static int +virSysinfoChassisParseXML(xmlNodePtr node, + xmlXPathContextPtr ctxt, + virSysinfoChassisDefPtr *chassisdef) +{ + int ret = -1; + virSysinfoChassisDefPtr def; + + if (!xmlStrEqual(node->name, BAD_CAST "chassis")) { + virReportError(VIR_ERR_XML_ERROR, "%s", + _("XML does not contain expected 'chassis' element")); + return ret; + } + + if (VIR_ALLOC(def) < 0) + goto cleanup; + + def->manufacturer = + virXPathString("string(entry[@name='manufacturer'])", ctxt); + def->version = + virXPathString("string(entry[@name='version'])", ctxt); + def->serial = + virXPathString("string(entry[@name='serial'])", ctxt); + def->asset = + virXPathString("string(entry[@name='asset'])", ctxt); + def->sku = + virXPathString("string(entry[@name='sku'])", ctxt); + + if (!def->manufacturer && !def->version && + !def->serial && !def->asset && !def->sku) { + virSysinfoChassisDefFree(def); + def = NULL; + } + + *chassisdef = def; + def = NULL; + ret = 0; + cleanup: + virSysinfoChassisDefFree(def); + return ret; +} static int virSysinfoOEMStringsParseXML(xmlXPathContextPtr ctxt, @@ -14600,6 +14641,17 @@ virSysinfoParseXML(xmlNodePtr node, if (virSysinfoBaseBoardParseXML(ctxt, &def->baseBoard, &def->nbaseBoard) < 0) goto error; + /* Extract chassis metadata */ + if ((tmpnode = virXPathNode("./chassis[1]", ctxt)) != NULL) { + oldnode = ctxt->node; + ctxt->node = tmpnode; + if (virSysinfoChassisParseXML(tmpnode, ctxt, &def->chassis) < 0) { + ctxt->node = oldnode; + goto error; + } + ctxt->node = oldnode; + } + /* Extract system related metadata */ if ((tmpnode = virXPathNode("./oemStrings[1]", ctxt)) != NULL) { oldnode = ctxt->node; diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index 24b434a..cf88e75 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -5813,6 +5813,49 @@ qemuBuildSmbiosBaseBoardStr(virSysinfoBaseBoardDefPtr def) return NULL; } +static char * +qemuBuildSmbiosChassisStr(virSysinfoChassisDefPtr def) +{ + virBuffer buf = VIR_BUFFER_INITIALIZER; + + if (!def) + return NULL; + + virBufferAddLit(&buf, "type=3"); + + /* 2:Manufacturer */ + virBufferAddLit(&buf, ",manufacturer="); + virQEMUBuildBufferEscapeComma(&buf, def->manufacturer); + /* 2:Version */ + if (def->version) { + virBufferAddLit(&buf, ",version="); + virQEMUBuildBufferEscapeComma(&buf, def->version); + } + /* 2:Serial Number */ + if (def->serial) { + virBufferAddLit(&buf, ",serial="); + virQEMUBuildBufferEscapeComma(&buf, def->serial); + } + /* 2:Asset Tag */ + if (def->asset) { + virBufferAddLit(&buf, ",asset="); + virQEMUBuildBufferEscapeComma(&buf, def->asset); + } + /* 2:Sku */ + if (def->sku) { + virBufferAddLit(&buf, ",sku="); + virQEMUBuildBufferEscapeComma(&buf, def->sku); + } + + if (virBufferCheckError(&buf) < 0) + goto error; + + return virBufferContentAndReset(&buf); + + error: + virBufferFreeAndReset(&buf); + return NULL; +} static char * qemuBuildSmbiosOEMStringsStr(virSysinfoOEMStringsDefPtr def) @@ -5905,6 +5948,12 @@ qemuBuildSmbiosCommandLine(virCommandPtr cmd, VIR_FREE(smbioscmd); } + smbioscmd = qemuBuildSmbiosChassisStr(source->chassis); + if (smbioscmd != NULL) { + virCommandAddArgList(cmd, "-smbios", smbioscmd, NULL); + VIR_FREE(smbioscmd); + } + if (source->oemStrings) { if (!(smbioscmd = qemuBuildSmbiosOEMStringsStr(source->oemStrings))) return -1; diff --git a/src/util/virsysinfo.c b/src/util/virsysinfo.c index e8d3371..af434e0 100644 --- a/src/util/virsysinfo.c +++ b/src/util/virsysinfo.c @@ -108,6 +108,18 @@ void virSysinfoBaseBoardDefClear(virSysinfoBaseBoardDefPtr def) VIR_FREE(def->location); } +void virSysinfoChassisDefFree(virSysinfoChassisDefPtr def) +{ + if (def == NULL) + return; + + VIR_FREE(def->manufacturer); + VIR_FREE(def->version); + VIR_FREE(def->serial); + VIR_FREE(def->asset); + VIR_FREE(def->sku); +} + void virSysinfoOEMStringsDefFree(virSysinfoOEMStringsDefPtr def) { size_t i; @@ -143,6 +155,8 @@ void virSysinfoDefFree(virSysinfoDefPtr def) virSysinfoBaseBoardDefClear(def->baseBoard + i); VIR_FREE(def->baseBoard); + virSysinfoChassisDefFree(def->chassis); + for (i = 0; i < def->nprocessor; i++) { VIR_FREE(def->processor[i].processor_socket_destination); VIR_FREE(def->processor[i].processor_type); @@ -1203,6 +1217,28 @@ virSysinfoBaseBoardFormat(virBufferPtr buf, } static void +virSysinfoChassisFormat(virBufferPtr buf, virSysinfoChassisDefPtr def) +{ + if (!def) + return; + + virBufferAddLit(buf, "<chassis>\n"); + virBufferAdjustIndent(buf, 2); + virBufferEscapeString(buf, "<entry name='manufacturer'>%s</entry>\n", + def->manufacturer); + virBufferEscapeString(buf, "<entry name='version'>%s</entry>\n", + def->version); + virBufferEscapeString(buf, "<entry name='serial'>%s</entry>\n", + def->serial); + virBufferEscapeString(buf, "<entry name='asset'>%s</entry>\n", + def->asset); + virBufferEscapeString(buf, "<entry name='sku'>%s</entry>\n", + def->sku); + virBufferAdjustIndent(buf, -2); + virBufferAddLit(buf, "</chassis>\n"); +} + +static void virSysinfoProcessorFormat(virBufferPtr buf, virSysinfoDefPtr def) { size_t i; @@ -1354,6 +1390,7 @@ virSysinfoFormat(virBufferPtr buf, virSysinfoDefPtr def) virSysinfoBIOSFormat(&childrenBuf, def->bios); virSysinfoSystemFormat(&childrenBuf, def->system); virSysinfoBaseBoardFormat(&childrenBuf, def->baseBoard, def->nbaseBoard); + virSysinfoChassisFormat(&childrenBuf, def->chassis); virSysinfoProcessorFormat(&childrenBuf, def); virSysinfoMemoryFormat(&childrenBuf, def); virSysinfoOEMStringsFormat(&childrenBuf, def->oemStrings); diff --git a/src/util/virsysinfo.h b/src/util/virsysinfo.h index ecb3a36..00a15db 100644 --- a/src/util/virsysinfo.h +++ b/src/util/virsysinfo.h @@ -98,6 +98,16 @@ struct _virSysinfoBaseBoardDef { /* XXX board type */ }; +typedef struct _virSysinfoChassisDef virSysinfoChassisDef; +typedef virSysinfoChassisDef *virSysinfoChassisDefPtr; +struct _virSysinfoChassisDef { + char *manufacturer; + char *version; + char *serial; + char *asset; + char *sku; +}; + typedef struct _virSysinfoOEMStringsDef virSysinfoOEMStringsDef; typedef virSysinfoOEMStringsDef *virSysinfoOEMStringsDefPtr; struct _virSysinfoOEMStringsDef { @@ -116,6 +126,8 @@ struct _virSysinfoDef { size_t nbaseBoard; virSysinfoBaseBoardDefPtr baseBoard; + virSysinfoChassisDefPtr chassis; + size_t nprocessor; virSysinfoProcessorDefPtr processor; @@ -130,6 +142,7 @@ virSysinfoDefPtr virSysinfoRead(void); void virSysinfoBIOSDefFree(virSysinfoBIOSDefPtr def); void virSysinfoSystemDefFree(virSysinfoSystemDefPtr def); void virSysinfoBaseBoardDefClear(virSysinfoBaseBoardDefPtr def); +void virSysinfoChassisDefFree(virSysinfoChassisDefPtr def); void virSysinfoOEMStringsDefFree(virSysinfoOEMStringsDefPtr def); void virSysinfoDefFree(virSysinfoDefPtr def); -- 1.8.3.1

2 1

[libvirt] [PATCH 00/11] Fixes and improvements to feature handling
by Andrea Bolognani 13 Feb '18

13 Feb '18

This series is the result of shaving many a yak while implementing support for pSeries optional features (first attempt at [1], respin coming right after I send this). Quick summary of the changes: * handle more sanely cases where the GIC capability has been enabled but no GIC version has been provided; * use the switch construct more, so that the compiler can warn us when we forget to handle some feature; * where possible (eg. the value is mandatory and there is only a limited number of options to choose from) avoid storing whether a feature is enabled and its actual value separately. Save for the first one, none of the changes should be visible to the user, but they will hopefully improve things for developers. [1] https://www.redhat.com/archives/libvir-list/2018-January/msg00779.html Andrea Bolognani (11): qemu: Move feature verification from PostParse() to Validate() qemu: Use switch in qemuDomainDefValidateFeatures() qemu: Move GIC checks to qemuDomainDefValidateFeatures() conf: Use switch in virDomainDefFeaturesCheckABIStability() conf: Validate VIR_DOMAIN_FEATURE_CAPABILITIES properly conf: Integrate all features ABI checks in the switch tests: Improve GIC tests qemu: Fix GIC behavior for the default case conf: Improve IOAPIC feature handling conf: Improve HPT feature handling tests: Clean up HPT tests src/conf/domain_conf.c | 174 +++++++++++++-------- src/conf/domain_conf.h | 15 +- src/libvirt_private.syms | 2 + src/qemu/qemu_command.c | 16 +- src/qemu/qemu_domain.c | 117 +++++++++----- .../qemuxml2argvdata/aarch64-gic-default-both.args | 1 + .../qemuxml2argvdata/aarch64-gic-default-both.xml | 1 + tests/qemuxml2argvdata/aarch64-gic-default-v2.args | 1 + tests/qemuxml2argvdata/aarch64-gic-default-v2.xml | 1 + tests/qemuxml2argvdata/aarch64-gic-default-v3.args | 1 + tests/qemuxml2argvdata/aarch64-gic-default-v3.xml | 1 + ...hpt-resizing.args => pseries-features-hpt.args} | 1 - .../pseries-features-hpt.xml} | 0 ...ne.xml => pseries-features-invalid-machine.xml} | 2 +- tests/qemuxml2argvdata/pseries-hpt-resizing.xml | 19 --- tests/qemuxml2argvtest.c | 29 ++-- .../aarch64-gic-default-both.xml | 1 + .../qemuxml2xmloutdata/aarch64-gic-default-v2.xml | 1 + .../qemuxml2xmloutdata/aarch64-gic-default-v3.xml | 1 + tests/qemuxml2xmloutdata/pseries-features-hpt.xml | 1 + tests/qemuxml2xmltest.c | 9 +- 21 files changed, 230 insertions(+), 164 deletions(-) create mode 120000 tests/qemuxml2argvdata/aarch64-gic-default-both.args create mode 120000 tests/qemuxml2argvdata/aarch64-gic-default-both.xml create mode 120000 tests/qemuxml2argvdata/aarch64-gic-default-v2.args create mode 120000 tests/qemuxml2argvdata/aarch64-gic-default-v2.xml create mode 120000 tests/qemuxml2argvdata/aarch64-gic-default-v3.args create mode 120000 tests/qemuxml2argvdata/aarch64-gic-default-v3.xml rename tests/qemuxml2argvdata/{pseries-hpt-resizing.args => pseries-features-hpt.args} (96%) rename tests/{qemuxml2xmloutdata/pseries-hpt-resizing.xml => qemuxml2argvdata/pseries-features-hpt.xml} (100%) rename tests/qemuxml2argvdata/{pseries-hpt-resizing-invalid-machine.xml => pseries-features-invalid-machine.xml} (86%) delete mode 100644 tests/qemuxml2argvdata/pseries-hpt-resizing.xml create mode 120000 tests/qemuxml2xmloutdata/aarch64-gic-default-both.xml create mode 120000 tests/qemuxml2xmloutdata/aarch64-gic-default-v2.xml create mode 120000 tests/qemuxml2xmloutdata/aarch64-gic-default-v3.xml create mode 120000 tests/qemuxml2xmloutdata/pseries-features-hpt.xml -- 2.14.3

2 30

[libvirt] [PATCH] Fix build with GCC 8 new warnings
by Daniel P. Berrangé 13 Feb '18

13 Feb '18

GCC 8 added a -Wcast-function-type warning to -Wextra by default. This complains if you try to explicitly cast one function signature to another function signature with incompatible args. This is something we do many times in libvirt especially with event callbacks. The hack to silence the warning requires casting via a void(*)(void) signature before casting to the desired type. This gross, so we just disable this new warning. GCC 8 also became more fussy about detecting switch fallthroughs. First it doesn't like it if you have a fallthrough attribute that is not before a case statement. e.g. FOO: BAR: WIZZ: ATTRIBUTE_FALLTHROUGH; Is unacceptable as there's no final case statement, so while FOO & BAR are falling through, WIZZ is not falling through. IOW, GCC wants us to write FOO: BAR: ATTRIBUTE_FALLTHROUGH; WIZZ: Second, it will report risk of fallthrough even if you have a case statement for every single enum value, but only if the switch is nested inside another switch and the outer case statement has no final break. This is is arguably valid because despite the fact that we have cast from "int" to the enum typedef, nothing guarantees that the variable we're switching on only contains values that have corresponding switch labels. e.g. int domstate = 87539319; switch ((virDomainState)domstate) { ... } will not match enum value, but also not raise any kind of compiler warning. So it is right to complain about risk of fallthrough if no default: is present. Signed-off-by: Daniel P. Berrangé <berrange(a)redhat.com> --- m4/virt-compile-warnings.m4 | 2 ++ src/qemu/qemu_domain.c | 14 +++++++------- src/qemu/qemu_domain_address.c | 11 +++++++++++ 3 files changed, 20 insertions(+), 7 deletions(-) diff --git a/m4/virt-compile-warnings.m4 b/m4/virt-compile-warnings.m4 index b9c974842..4b35a6f6b 100644 --- a/m4/virt-compile-warnings.m4 +++ b/m4/virt-compile-warnings.m4 @@ -177,6 +177,8 @@ AC_DEFUN([LIBVIRT_COMPILE_WARNINGS],[ # with gl_MANYWARN_COMPLEMENT # So we have -W enabled, and then have to explicitly turn off... wantwarn="$wantwarn -Wno-sign-compare" + # We do "bad" function casts all the time for event callbacks + wantwarn="$wantwarn -Wno-cast-function-type" # GNULIB expects this to be part of -Wc++-compat, but we turn # that one off, so we need to manually enable this again diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c index 178ec24ae..560436f8e 100644 --- a/src/qemu/qemu_domain.c +++ b/src/qemu/qemu_domain.c @@ -175,9 +175,9 @@ qemuDomainAsyncJobPhaseToString(qemuDomainAsyncJob job, case QEMU_ASYNC_JOB_NONE: case QEMU_ASYNC_JOB_LAST: ATTRIBUTE_FALLTHROUGH; + default: + return "none"; } - - return "none"; } int @@ -199,12 +199,12 @@ qemuDomainAsyncJobPhaseFromString(qemuDomainAsyncJob job, case QEMU_ASYNC_JOB_NONE: case QEMU_ASYNC_JOB_LAST: ATTRIBUTE_FALLTHROUGH; + default: + if (STREQ(phase, "none")) + return 0; + else + return -1; } - - if (STREQ(phase, "none")) - return 0; - else - return -1; } diff --git a/src/qemu/qemu_domain_address.c b/src/qemu/qemu_domain_address.c index e28119e4b..42c7c0b12 100644 --- a/src/qemu/qemu_domain_address.c +++ b/src/qemu/qemu_domain_address.c @@ -532,6 +532,7 @@ qemuDomainDeviceCalculatePCIConnectFlags(virDomainDeviceDefPtr dev, case VIR_DOMAIN_CONTROLLER_MODEL_USB_QUSB2: /* xen only */ case VIR_DOMAIN_CONTROLLER_MODEL_USB_NONE: case VIR_DOMAIN_CONTROLLER_MODEL_USB_LAST: + default: return 0; } @@ -553,6 +554,7 @@ qemuDomainDeviceCalculatePCIConnectFlags(virDomainDeviceDefPtr dev, return pciFlags; case VIR_DOMAIN_CONTROLLER_MODEL_SCSI_LAST: + default: return 0; } @@ -562,6 +564,7 @@ qemuDomainDeviceCalculatePCIConnectFlags(virDomainDeviceDefPtr dev, case VIR_DOMAIN_CONTROLLER_TYPE_FDC: case VIR_DOMAIN_CONTROLLER_TYPE_CCID: case VIR_DOMAIN_CONTROLLER_TYPE_LAST: + default: /* should be 0 */ return pciFlags; } @@ -605,6 +608,7 @@ qemuDomainDeviceCalculatePCIConnectFlags(virDomainDeviceDefPtr dev, case VIR_DOMAIN_SOUND_MODEL_PCSPK: case VIR_DOMAIN_SOUND_MODEL_USB: case VIR_DOMAIN_SOUND_MODEL_LAST: + default: return 0; } @@ -622,6 +626,7 @@ qemuDomainDeviceCalculatePCIConnectFlags(virDomainDeviceDefPtr dev, case VIR_DOMAIN_DISK_BUS_SATA: case VIR_DOMAIN_DISK_BUS_SD: case VIR_DOMAIN_DISK_BUS_LAST: + default: return 0; } @@ -734,6 +739,7 @@ qemuDomainDeviceCalculatePCIConnectFlags(virDomainDeviceDefPtr dev, case VIR_DOMAIN_MEMBALLOON_MODEL_XEN: case VIR_DOMAIN_MEMBALLOON_MODEL_NONE: case VIR_DOMAIN_MEMBALLOON_MODEL_LAST: + default: return 0; } @@ -743,6 +749,7 @@ qemuDomainDeviceCalculatePCIConnectFlags(virDomainDeviceDefPtr dev, return virtioFlags; case VIR_DOMAIN_RNG_MODEL_LAST: + default: return 0; } @@ -755,6 +762,7 @@ qemuDomainDeviceCalculatePCIConnectFlags(virDomainDeviceDefPtr dev, case VIR_DOMAIN_WATCHDOG_MODEL_IB700: case VIR_DOMAIN_WATCHDOG_MODEL_DIAG288: case VIR_DOMAIN_WATCHDOG_MODEL_LAST: + default: return 0; } @@ -775,6 +783,7 @@ qemuDomainDeviceCalculatePCIConnectFlags(virDomainDeviceDefPtr dev, case VIR_DOMAIN_VIDEO_TYPE_DEFAULT: case VIR_DOMAIN_VIDEO_TYPE_GOP: case VIR_DOMAIN_VIDEO_TYPE_LAST: + default: return 0; } @@ -791,6 +800,7 @@ qemuDomainDeviceCalculatePCIConnectFlags(virDomainDeviceDefPtr dev, case VIR_DOMAIN_INPUT_BUS_XEN: case VIR_DOMAIN_INPUT_BUS_PARALLELS: case VIR_DOMAIN_INPUT_BUS_LAST: + default: return 0; } @@ -806,6 +816,7 @@ qemuDomainDeviceCalculatePCIConnectFlags(virDomainDeviceDefPtr dev, case VIR_DOMAIN_CHR_SERIAL_TARGET_TYPE_SCLP: case VIR_DOMAIN_CHR_SERIAL_TARGET_TYPE_NONE: case VIR_DOMAIN_CHR_SERIAL_TARGET_TYPE_LAST: + default: return 0; } -- 2.16.1

3 3

[libvirt] [PATCH] qemu: Check for down limit of SLIRP prefix too
by Michal Privoznik 13 Feb '18

13 Feb '18

https://bugzilla.redhat.com/show_bug.cgi?id=1515533 We're already checking if IPv4 prefix isn't too long. But we are not checking if it isn't too short. QEMU supports prefixes longer than 4 (including). I haven't find anything similar related to IPv6 in qemu sources. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- src/qemu/qemu_domain.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c index df433c2f0..4fc4db68b 100644 --- a/src/qemu/qemu_domain.c +++ b/src/qemu/qemu_domain.c @@ -3744,6 +3744,12 @@ qemuDomainDeviceDefValidateNetwork(const virDomainNetDef *net) _("prefix too long")); return -1; } + + if (ip->prefix < 4) { + virReportError(VIR_ERR_XML_ERROR, "%s", + _("prefix too short")); + return -1; + } } if (VIR_SOCKET_ADDR_IS_FAMILY(&ip->address, AF_INET6)) { -- 2.13.6

2 1

[libvirt] [PATCH 0/1] Function declarations match definitions
by Chris Venteicher 13 Feb '18

13 Feb '18

Interested in fixing mismatches between function parameter names in declarations (header files) and definitions (.c files)? This is a cosmetic / redability fix and does not change functionality. This is patch 1 of larger set of 22 patches that fix parameter names across the libvirt source code. If there is interest I can send the entire patch set. These are cosmetic redability fixes automatically generated by clang-tidy. make check and make syntax-check pass successfully after the 22 patches are applied. This is my first submission so please point out any issues and I will be glad to fix. Chris Venteicher (1): includes: function parameter names same in headers include/libvirt/libvirt-domain.h | 26 +++++++++++++------------- include/libvirt/libvirt-event.h | 4 ++-- include/libvirt/libvirt-host.h | 4 ++-- include/libvirt/libvirt-interface.h | 4 ++-- include/libvirt/libvirt-network.h | 6 +++--- include/libvirt/libvirt-nwfilter.h | 2 +- include/libvirt/libvirt-qemu.h | 2 +- include/libvirt/libvirt-secret.h | 4 ++-- include/libvirt/libvirt-storage.h | 12 ++++++------ include/libvirt/libvirt-stream.h | 22 +++++++++++----------- 10 files changed, 43 insertions(+), 43 deletions(-) -- 2.14.1

2 3

[libvirt] [PATCH REBASE 0/2] qemu: Clean up PCI controller options
by Andrea Bolognani 12 Feb '18

12 Feb '18

Rebase of https://www.redhat.com/archives/libvir-list/2017-December/msg00154.html on top of master, now that patch 1/3 has been merged as part of John's work to move controller checks from command line generation to guest validation. Andrea Bolognani (2): tests: Add some tests for PCI controller options qemu: Clean up PCI controller options src/qemu/qemu_domain.c | 117 +++++++++++++++++++++ .../i440fx-controllers-pciopts.args | 24 +++++ .../i440fx-controllers-pciopts.xml | 36 +++++++ .../pseries-controllers-pciopts.args | 22 ++++ .../pseries-controllers-pciopts.xml | 36 +++++++ .../qemuxml2argvdata/q35-controllers-pciopts.args | 28 +++++ tests/qemuxml2argvdata/q35-controllers-pciopts.xml | 60 +++++++++++ tests/qemuxml2argvtest.c | 17 +++ .../i440fx-controllers-pciopts.xml | 42 ++++++++ tests/qemuxml2xmloutdata/pcie-expander-bus.xml | 4 +- .../pseries-controllers-pciopts.xml | 41 ++++++++ .../qemuxml2xmloutdata/q35-controllers-pciopts.xml | 63 +++++++++++ tests/qemuxml2xmltest.c | 17 +++ 13 files changed, 504 insertions(+), 3 deletions(-) create mode 100644 tests/qemuxml2argvdata/i440fx-controllers-pciopts.args create mode 100644 tests/qemuxml2argvdata/i440fx-controllers-pciopts.xml create mode 100644 tests/qemuxml2argvdata/pseries-controllers-pciopts.args create mode 100644 tests/qemuxml2argvdata/pseries-controllers-pciopts.xml create mode 100644 tests/qemuxml2argvdata/q35-controllers-pciopts.args create mode 100644 tests/qemuxml2argvdata/q35-controllers-pciopts.xml create mode 100644 tests/qemuxml2xmloutdata/i440fx-controllers-pciopts.xml create mode 100644 tests/qemuxml2xmloutdata/pseries-controllers-pciopts.xml create mode 100644 tests/qemuxml2xmloutdata/q35-controllers-pciopts.xml -- 2.14.3

2 7

[libvirt] [PATCH] qemu: Fix indentation in qemuBuildControllerDevStr()
by Andrea Bolognani 12 Feb '18

12 Feb '18

Add braces around the multi-line body as well, in compliance with our coding style. Signed-off-by: Andrea Bolognani <abologna(a)redhat.com> --- Pushed as trivial. src/qemu/qemu_command.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index f7925c93a..6c73cd7bf 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -2743,9 +2743,10 @@ qemuBuildControllerDevStr(const virDomainDef *domainDef, virBufferAsprintf(&buf, "%s,bus_nr=%d,id=%s", modelName, pciopts->busNr, def->info.alias); - if (pciopts->numaNode != -1) - virBufferAsprintf(&buf, ",numa_node=%d", - pciopts->numaNode); + if (pciopts->numaNode != -1) { + virBufferAsprintf(&buf, ",numa_node=%d", + pciopts->numaNode); + } break; case VIR_DOMAIN_CONTROLLER_MODEL_DMI_TO_PCI_BRIDGE: case VIR_DOMAIN_CONTROLLER_MODEL_PCIE_SWITCH_UPSTREAM_PORT: -- 2.14.3

1 0

[libvirt] Google Summer of Code 2018
by Michal Privoznik 12 Feb '18

12 Feb '18

Dear list, Libvirt has been accepted to this year's GSoC. Yay! For those interested in mentoring let me know (if you haven't already). For those interested in participating as students now it's the best time to view our ideas page [1] and contact corresponding mentors. If you have your own idea that's even better! Let me know and we will find you a mentor. Also, it might be worth reading page dedicated to this year's run [2]. Michal 1: https://wiki.libvirt.org/page/Google_Summer_of_Code_Ideas 2: https://wiki.libvirt.org/page/Google_Summer_of_Code_2018

1 0

[libvirt] [PATCH 00/14] Basic implementation of persistent reservations
by Michal Privoznik 12 Feb '18

12 Feb '18

QEMU added support for SCSI persistent reservations. The way QEMU implemented that makes it slightly harder for libvirt to adapt to - it's a small binary that needs to start before qemu so that it can connect to it. Basically, what libvirt does is: 1) start qemu-pr-helper (the small binary from above), 2) records its PID for tracking purposes 3) start qemu and when shutting down a domain the process is reversed: 1) shut down qemu process, 2) kill qemu-pr-helper (PID was saved earlier). Now, here also lies the last missing piece to he puzzle - what if qemu-pr-helper dies while qemu is running? After some discussion with Paolo it was agreed that qemu will emit an event for libvirt so that we can respawn the process. Anyway, that shouldn't be a show stopper for these patches (if it was the patch set might get too big). Please test, review and comment. As usual, you can find the patches on my github too: https://github.com/zippy2/libvirt/tree/pr Michal Privoznik (14): virstoragefile: Introduce virStoragePRDef qemuDomainDiskChangeSupported: Deny changing reservations qemu: Introduce pr-manager-helper capability qemu_domain: Introduce qemuDomainDiskPRObject qemu: Generate alias for pr-helper qemu: Store prAlias and prPath in status XML qemu: Generate cmd line at startup qemu: Introduce pr_helper to qemu.conf qemu: Start PR daemons on domain startup qemu: Track PR daemons PIDs in status XML qemu_domain: Introduce qemuDomainGetPRUsageCount qemu_process.c: Introduce qemuProcessSetupPRDaemon qemu_hotplug: Hotplug of reservations qemu_hotplug: Hotunplug of reservations docs/formatdomain.html.in | 25 +- docs/schemas/domaincommon.rng | 19 +- docs/schemas/storagecommon.rng | 34 ++ m4/virt-driver-qemu.m4 | 5 + src/conf/domain_conf.c | 36 ++ src/libvirt_private.syms | 4 + src/qemu/libvirtd_qemu.aug | 1 + src/qemu/qemu.conf | 4 + src/qemu/qemu_capabilities.c | 2 + src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_command.c | 59 +++ src/qemu/qemu_conf.c | 7 +- src/qemu/qemu_conf.h | 1 + src/qemu/qemu_domain.c | 435 ++++++++++++++++++++- src/qemu/qemu_domain.h | 29 ++ src/qemu/qemu_hotplug.c | 110 ++++++ src/qemu/qemu_process.c | 160 ++++++++ src/qemu/qemu_process.h | 4 + src/qemu/test_libvirtd_qemu.aug.in | 1 + src/util/virstoragefile.c | 166 ++++++++ src/util/virstoragefile.h | 17 + .../disk-virtio-scsi-reservations-not-managed.args | 28 ++ .../disk-virtio-scsi-reservations-not-managed.xml | 40 ++ .../disk-virtio-scsi-reservations.args | 29 ++ .../disk-virtio-scsi-reservations.xml | 38 ++ tests/qemuxml2argvtest.c | 8 + .../disk-virtio-scsi-reservations-not-managed.xml | 1 + .../disk-virtio-scsi-reservations.xml | 1 + tests/qemuxml2xmltest.c | 4 + 29 files changed, 1250 insertions(+), 19 deletions(-) create mode 100644 tests/qemuxml2argvdata/disk-virtio-scsi-reservations-not-managed.args create mode 100644 tests/qemuxml2argvdata/disk-virtio-scsi-reservations-not-managed.xml create mode 100644 tests/qemuxml2argvdata/disk-virtio-scsi-reservations.args create mode 100644 tests/qemuxml2argvdata/disk-virtio-scsi-reservations.xml create mode 120000 tests/qemuxml2xmloutdata/disk-virtio-scsi-reservations-not-managed.xml create mode 120000 tests/qemuxml2xmloutdata/disk-virtio-scsi-reservations.xml -- 2.13.6

6 34

[libvirt] [PATCH] Revert "qemu: Expose rx/tx_queue_size in qemu.conf too"
by Daniel P. Berrangé 12 Feb '18

12 Feb '18

This reverts commit 038eb472a0d970a17ccf4343ead0666df5c92f9d. On reflection adding defaults for arbitrary guest XML device config settings to the qemu.conf is not a sustainable path. Removing the support for rx/tx queue size so that it doesn't set a bad precedent. Signed-off-by: Daniel P. Berrangé <berrange(a)redhat.com> --- docs/formatdomain.html.in | 14 ++-------- src/qemu/libvirtd_qemu.aug | 4 --- src/qemu/qemu.conf | 6 ----- src/qemu/qemu_command.c | 55 +++++++++++--------------------------- src/qemu/qemu_command.h | 3 +-- src/qemu/qemu_conf.c | 4 --- src/qemu/qemu_conf.h | 3 --- src/qemu/qemu_hotplug.c | 2 +- src/qemu/test_libvirtd_qemu.aug.in | 2 -- 9 files changed, 19 insertions(+), 74 deletions(-) diff --git a/docs/formatdomain.html.in b/docs/formatdomain.html.in index 6707744bda..3ec1173c6f 100644 --- a/docs/formatdomain.html.in +++ b/docs/formatdomain.html.in @@ -5456,12 +5456,7 @@ qemu-kvm -net nic,model=? /dev/null some restrictions on actual value. For instance, latest QEMU (as of 2016-09-01) requires value to be a power of two from [256, 1024] range. - Since 2.3.0 (QEMU and KVM only) - Additionally, since 4.1.0 the - value can be set in the <code>qemu.conf</code> file in order - to override the hypervisor default value. Note that XML has - higher precedence because it's more specific. - + Since 2.3.0 (QEMU and KVM only) In general you should leave this option alone, unless you are very certain you know what you are doing. @@ -5477,12 +5472,7 @@ qemu-kvm -net nic,model=? /dev/null range. In addition to that, this may work only for a subset of interface types, e.g. aforementioned QEMU enables this option only for <code>vhostuser</code> type. - Since 3.7.0 (QEMU and KVM only) - Additionally, since 4.1.0 the - value can be set in the <code>qemu.conf</code> file in order - to override the hypervisor default value. Note that XML has - higher precedence because it's more specific. - + Since 3.7.0 (QEMU and KVM only) In general you should leave this option alone, unless you are very certain you know what you are doing. diff --git a/src/qemu/libvirtd_qemu.aug b/src/qemu/libvirtd_qemu.aug index 084290296a..c19bf3a43a 100644 --- a/src/qemu/libvirtd_qemu.aug +++ b/src/qemu/libvirtd_qemu.aug @@ -118,9 +118,6 @@ module Libvirtd_qemu = let vxhs_entry = bool_entry "vxhs_tls" | str_entry "vxhs_tls_x509_cert_dir" - let virtio_entry = int_entry "rx_queue_size" - | int_entry "tx_queue_size" - (* Each entry in the config is one of the following ... *) let entry = default_tls_entry | vnc_entry @@ -140,7 +137,6 @@ module Libvirtd_qemu = | gluster_debug_level_entry | memory_entry | vxhs_entry - | virtio_entry let comment = [ label "#comment" . del /#[ \t]*/ "# " . store /([^ \t\n][^\n]*)?/ . del /\n/ "\n" ] let empty = [ label "#empty" . eol ] diff --git a/src/qemu/qemu.conf b/src/qemu/qemu.conf index 62c4265ea9..43dd561cca 100644 --- a/src/qemu/qemu.conf +++ b/src/qemu/qemu.conf @@ -775,9 +775,3 @@ # This directory is used for memoryBacking source if configured as file. # NOTE: big files will be stored here #memory_backing_dir = "/var/lib/libvirt/qemu/ram" - -# The following two values set the default RX/TX ring buffer size for virtio -# interfaces. These values are taken unless overridden in domain XML. For more -# info consult docs to corresponding attributes from domain XML. -#rx_queue_size = 1024 -#tx_queue_size = 1024 diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index 2bcaa5fc22..f7925c93a8 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -3351,8 +3351,7 @@ qemuBuildNicStr(virDomainNetDefPtr net, char * -qemuBuildNicDevStr(virQEMUDriverConfigPtr cfg, - virDomainDefPtr def, +qemuBuildNicDevStr(virDomainDefPtr def, virDomainNetDefPtr net, int vlan, unsigned int bootindex, @@ -3472,41 +3471,21 @@ qemuBuildNicDevStr(virQEMUDriverConfigPtr cfg, virBufferAsprintf(&buf, ",mq=on,vectors=%zu", 2 * vhostfdSize + 2); } } - if (usingVirtio) { - unsigned int rx_queue_size = net->driver.virtio.rx_queue_size; - - if (rx_queue_size == 0) - rx_queue_size = cfg->rx_queue_size; - - if (rx_queue_size) { - if (!virQEMUCapsGet(qemuCaps, QEMU_CAPS_VIRTIO_NET_RX_QUEUE_SIZE)) { - virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", - _("virtio rx_queue_size option is " - "not supported with this QEMU binary")); - goto error; - } - - net->driver.virtio.rx_queue_size = rx_queue_size; - virBufferAsprintf(&buf, ",rx_queue_size=%u", rx_queue_size); + if (usingVirtio && net->driver.virtio.rx_queue_size) { + if (!virQEMUCapsGet(qemuCaps, QEMU_CAPS_VIRTIO_NET_RX_QUEUE_SIZE)) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", + _("virtio rx_queue_size option is not supported with this QEMU binary")); + goto error; } + virBufferAsprintf(&buf, ",rx_queue_size=%u", net->driver.virtio.rx_queue_size); } - if (usingVirtio) { - unsigned int tx_queue_size = net->driver.virtio.tx_queue_size; - - if (tx_queue_size == 0) - tx_queue_size = cfg->tx_queue_size; - - if (tx_queue_size) { - if (!virQEMUCapsGet(qemuCaps, QEMU_CAPS_VIRTIO_NET_TX_QUEUE_SIZE)) { - virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", - _("virtio tx_queue_size option is " - "not supported with this QEMU binary")); - goto error; - } - - net->driver.virtio.tx_queue_size = tx_queue_size; - virBufferAsprintf(&buf, ",tx_queue_size=%u", tx_queue_size); + if (usingVirtio && net->driver.virtio.tx_queue_size) { + if (!virQEMUCapsGet(qemuCaps, QEMU_CAPS_VIRTIO_NET_TX_QUEUE_SIZE)) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", + _("virtio tx_queue_size option is not supported with this QEMU binary")); + goto error; } + virBufferAsprintf(&buf, ",tx_queue_size=%u", net->driver.virtio.tx_queue_size); } if (usingVirtio && net->mtu) { @@ -8192,7 +8171,7 @@ qemuBuildVhostuserCommandLine(virQEMUDriverPtr driver, virCommandAddArg(cmd, netdev); VIR_FREE(netdev); - if (!(nic = qemuBuildNicDevStr(cfg, def, net, -1, bootindex, + if (!(nic = qemuBuildNicDevStr(def, net, -1, bootindex, queues, qemuCaps))) { virReportError(VIR_ERR_INTERNAL_ERROR, "%s", _("Error generating NIC -device string")); @@ -8229,7 +8208,6 @@ qemuBuildInterfaceCommandLine(virQEMUDriverPtr driver, int **nicindexes, bool chardevStdioLogd) { - virQEMUDriverConfigPtr cfg = NULL; int ret = -1; char *nic = NULL, *host = NULL; int *tapfd = NULL; @@ -8291,8 +8269,6 @@ qemuBuildInterfaceCommandLine(virQEMUDriverPtr driver, return -1; } - cfg = virQEMUDriverGetConfig(driver); - switch (actualType) { case VIR_DOMAIN_NET_TYPE_NETWORK: case VIR_DOMAIN_NET_TYPE_BRIDGE: @@ -8488,7 +8464,7 @@ qemuBuildInterfaceCommandLine(virQEMUDriverPtr driver, virCommandAddArgList(cmd, "-netdev", host, NULL); } if (qemuDomainSupportsNicdev(def, net)) { - if (!(nic = qemuBuildNicDevStr(cfg, def, net, vlan, bootindex, + if (!(nic = qemuBuildNicDevStr(def, net, vlan, bootindex, vhostfdSize, qemuCaps))) goto cleanup; virCommandAddArgList(cmd, "-device", nic, NULL); @@ -8532,7 +8508,6 @@ qemuBuildInterfaceCommandLine(virQEMUDriverPtr driver, VIR_FREE(host); VIR_FREE(tapfdName); VIR_FREE(vhostfdName); - virObjectUnref(cfg); return ret; } diff --git a/src/qemu/qemu_command.h b/src/qemu/qemu_command.h index 6449883291..31c9da673c 100644 --- a/src/qemu/qemu_command.h +++ b/src/qemu/qemu_command.h @@ -90,8 +90,7 @@ char *qemuBuildNicStr(virDomainNetDefPtr net, int vlan); /* Current, best practice */ -char *qemuBuildNicDevStr(virQEMUDriverConfigPtr cfg, - virDomainDefPtr def, +char *qemuBuildNicDevStr(virDomainDefPtr def, virDomainNetDefPtr net, int vlan, unsigned int bootindex, diff --git a/src/qemu/qemu_conf.c b/src/qemu/qemu_conf.c index 2fa96431fa..af503d31cb 100644 --- a/src/qemu/qemu_conf.c +++ b/src/qemu/qemu_conf.c @@ -912,10 +912,6 @@ int virQEMUDriverConfigLoadFile(virQEMUDriverConfigPtr cfg, if (virConfGetValueString(conf, "memory_backing_dir", &cfg->memoryBackingDir) < 0) goto cleanup; - if (virConfGetValueUInt(conf, "rx_queue_size", &cfg->rx_queue_size) < 0 || - virConfGetValueUInt(conf, "tx_queue_size", &cfg->tx_queue_size) < 0) - goto cleanup; - ret = 0; cleanup: diff --git a/src/qemu/qemu_conf.h b/src/qemu/qemu_conf.h index 3f38a76c26..a553e30e2e 100644 --- a/src/qemu/qemu_conf.h +++ b/src/qemu/qemu_conf.h @@ -206,9 +206,6 @@ struct _virQEMUDriverConfig { bool vxhsTLS; char *vxhsTLSx509certdir; - - unsigned int rx_queue_size; - unsigned int tx_queue_size; }; /* Main driver state */ diff --git a/src/qemu/qemu_hotplug.c b/src/qemu/qemu_hotplug.c index 3291ce6130..c9868de778 100644 --- a/src/qemu/qemu_hotplug.c +++ b/src/qemu/qemu_hotplug.c @@ -1118,7 +1118,7 @@ qemuDomainAttachNetDevice(virQEMUDriverPtr driver, for (i = 0; i < vhostfdSize; i++) VIR_FORCE_CLOSE(vhostfd[i]); - if (!(nicstr = qemuBuildNicDevStr(cfg, vm->def, net, vlan, 0, + if (!(nicstr = qemuBuildNicDevStr(vm->def, net, vlan, 0, queueSize, priv->qemuCaps))) goto try_remove; diff --git a/src/qemu/test_libvirtd_qemu.aug.in b/src/qemu/test_libvirtd_qemu.aug.in index 4fc4e2f4ec..688e5b9fda 100644 --- a/src/qemu/test_libvirtd_qemu.aug.in +++ b/src/qemu/test_libvirtd_qemu.aug.in @@ -100,5 +100,3 @@ module Test_libvirtd_qemu = { "1" = "mount" } } { "memory_backing_dir" = "/var/lib/libvirt/qemu/ram" } -{ "rx_queue_size" = "1024" } -{ "tx_queue_size" = "1024" } -- 2.14.3

2 1

[libvirt] [PATCH v2] vcpupin: add clear feature
by Yi Wang 12 Feb '18

12 Feb '18

We can't clear vcpupin settings of XML once we did vcpupin command, this is not convenient under some condition such as migration to a host with less CPUs. This patch introduces clear feature, which can clear vcpuin setting of XML using a 'c' option. Signed-off-by: Yi Wang <wang.yi59(a)zte.com.cn> Signed-off-by: Xi Xu <xu.xi8(a)zte.com.cn> --- include/libvirt/libvirt-domain.h | 11 +++++++++++ src/qemu/qemu_driver.c | 32 ++++++++++++++++++++++++-------- tools/virsh-domain.c | 5 ++++- tools/virsh.pod | 1 + 4 files changed, 40 insertions(+), 9 deletions(-) diff --git a/include/libvirt/libvirt-domain.h b/include/libvirt/libvirt-domain.h index 4048acf..46f4e77 100644 --- a/include/libvirt/libvirt-domain.h +++ b/include/libvirt/libvirt-domain.h @@ -1847,6 +1847,17 @@ int virDomainSetVcpusFlags (virDomainPtr domain, int virDomainGetVcpusFlags (virDomainPtr domain, unsigned int flags); +/* Flags for controlling virtual CPU pinning. */ +typedef enum { + /* See virDomainModificationImpact for these flags. */ + VIR_DOMAIN_VCPU_PIN_CURRENT = VIR_DOMAIN_AFFECT_CURRENT, + VIR_DOMAIN_VCPU_PIN_LIVE = VIR_DOMAIN_AFFECT_LIVE, + VIR_DOMAIN_VCPU_PIN_CONFIG = VIR_DOMAIN_AFFECT_CONFIG, + + /* Additionally, these flags may be bitwise-OR'd in. */ + VIR_DOMAIN_VCPU_PIN_CLEAR = (1 << 2), /* Clear vcpus pin info */ +} virDomainVcpuPinFlags; + int virDomainPinVcpu (virDomainPtr domain, unsigned int vcpu, unsigned char *cpumap, diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index bbce5bd..fe1f62f 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -5056,7 +5056,8 @@ qemuDomainPinVcpuLive(virDomainObjPtr vm, int vcpu, virQEMUDriverPtr driver, virQEMUDriverConfigPtr cfg, - virBitmapPtr cpumap) + virBitmapPtr cpumap, + bool clear) { virBitmapPtr tmpmap = NULL; virDomainVcpuDefPtr vcpuinfo; @@ -5069,6 +5070,7 @@ qemuDomainPinVcpuLive(virDomainObjPtr vm, int eventNparams = 0; int eventMaxparams = 0; int ret = -1; + virBitmapPtr targetMap = NULL; if (!qemuDomainHasVcpuPids(vm)) { virReportError(VIR_ERR_OPERATION_INVALID, @@ -5083,10 +5085,15 @@ qemuDomainPinVcpuLive(virDomainObjPtr vm, goto cleanup; } - if (!(tmpmap = virBitmapNewCopy(cpumap))) - goto cleanup; + if (clear) { + targetMap = virHostCPUGetOnlineBitmap(); + } else { + targetMap = cpumap; + if (!(tmpmap = virBitmapNewCopy(cpumap))) + goto cleanup; + } - if (!(str = virBitmapFormat(cpumap))) + if (!(str = virBitmapFormat(targetMap))) goto cleanup; if (vcpuinfo->online) { @@ -5095,11 +5102,11 @@ qemuDomainPinVcpuLive(virDomainObjPtr vm, if (virCgroupNewThread(priv->cgroup, VIR_CGROUP_THREAD_VCPU, vcpu, false, &cgroup_vcpu) < 0) goto cleanup; - if (qemuSetupCgroupCpusetCpus(cgroup_vcpu, cpumap) < 0) + if (qemuSetupCgroupCpusetCpus(cgroup_vcpu, targetMap) < 0) goto cleanup; } - if (virProcessSetAffinity(qemuDomainGetVcpuPid(vm, vcpu), cpumap) < 0) + if (virProcessSetAffinity(qemuDomainGetVcpuPid(vm, vcpu), targetMap) < 0) goto cleanup; } @@ -5128,6 +5135,8 @@ qemuDomainPinVcpuLive(virDomainObjPtr vm, virCgroupFree(&cgroup_vcpu); VIR_FREE(str); qemuDomainEventQueue(driver, event); + if (clear) + virBitmapFree(targetMap); return ret; } @@ -5148,9 +5157,11 @@ qemuDomainPinVcpuFlags(virDomainPtr dom, virBitmapPtr pcpumap = NULL; virDomainVcpuDefPtr vcpuinfo = NULL; virQEMUDriverConfigPtr cfg = NULL; + bool clear = false; virCheckFlags(VIR_DOMAIN_AFFECT_LIVE | - VIR_DOMAIN_AFFECT_CONFIG, -1); + VIR_DOMAIN_AFFECT_CONFIG | + VIR_DOMAIN_VCPU_PIN_CLEAR, -1); cfg = virQEMUDriverGetConfig(driver); @@ -5166,6 +5177,8 @@ qemuDomainPinVcpuFlags(virDomainPtr dom, if (virDomainObjGetDefs(vm, flags, &def, &persistentDef) < 0) goto endjob; + clear = !!(flags & VIR_DOMAIN_VCPU_PIN_CLEAR); + if ((def && def->virtType == VIR_DOMAIN_VIRT_QEMU) || (persistentDef && persistentDef->virtType == VIR_DOMAIN_VIRT_QEMU)) { virReportError(VIR_ERR_OPERATION_FAILED, "%s", @@ -5191,7 +5204,7 @@ qemuDomainPinVcpuFlags(virDomainPtr dom, } if (def && - qemuDomainPinVcpuLive(vm, def, vcpu, driver, cfg, pcpumap) < 0) + qemuDomainPinVcpuLive(vm, def, vcpu, driver, cfg, pcpumap, clear) < 0) goto endjob; if (persistentDef) { @@ -5199,6 +5212,9 @@ qemuDomainPinVcpuFlags(virDomainPtr dom, vcpuinfo->cpumask = pcpumap; pcpumap = NULL; + if (clear) + virBitmapFree(vcpuinfo->cpumask); + ret = virDomainSaveConfig(cfg->configDir, driver->caps, persistentDef); goto endjob; } diff --git a/tools/virsh-domain.c b/tools/virsh-domain.c index 5a0e0c1..4981ecc 100644 --- a/tools/virsh-domain.c +++ b/tools/virsh-domain.c @@ -6863,7 +6863,7 @@ virshParseCPUList(vshControl *ctl, int *cpumaplen, unsigned char *cpumap = NULL; virBitmapPtr map = NULL; - if (cpulist[0] == 'r') { + if (cpulist[0] == 'r' || cpulist[0] == 'c') { if (!(map = virBitmapNew(maxcpu))) return NULL; virBitmapSetAll(map); @@ -6941,6 +6941,9 @@ cmdVcpuPin(vshControl *ctl, const vshCmd *cmd) goto cleanup; } + if (STREQ(cpulist, "c")) + flags |= VIR_DOMAIN_VCPU_PIN_CLEAR; + /* Pin mode: pinning specified vcpu to specified physical cpus*/ if (!(cpumap = virshParseCPUList(ctl, &cpumaplen, cpulist, maxcpu))) goto cleanup; diff --git a/tools/virsh.pod b/tools/virsh.pod index 69cc423..d5a1779 100644 --- a/tools/virsh.pod +++ b/tools/virsh.pod @@ -2857,6 +2857,7 @@ I<cpulist> is a list of physical CPU numbers. Its syntax is a comma separated list and a special markup using '-' and '^' (ex. '0-4', '0-3,^2') can also be allowed. The '-' denotes the range and the '^' denotes exclusive. For pinning the I<vcpu> to all physical cpus specify 'r' as a I<cpulist>. +For clearing pinning info, specify 'c' as a I<cpulist>. If I<--live> is specified, affect a running guest. If I<--config> is specified, affect the next boot of a persistent guest. If I<--current> is specified, affect the current guest state. -- 1.8.3.1

3 4

[libvirt] [PATCH] log: fix deadlock obtaining hostname (related CVE-2018-6764)
by Daniel P. Berrangé 12 Feb '18

12 Feb '18

The fix for CVE-2018-6764 introduced a potential deadlock scenario that gets triggered by the NSS module when virGetHostname() calls getaddrinfo to resolve the hostname: #0 0x00007f6e714b57e7 in futex_wait #1 futex_wait_simple #2 __pthread_once_slow #3 0x00007f6e71d16e7d in virOnce #4 0x00007f6e71d0997c in virLogInitialize #5 0x00007f6e71d0a09a in virLogVMessage #6 0x00007f6e71d09ffd in virLogMessage #7 0x00007f6e71d0db22 in virObjectNew #8 0x00007f6e71d0dbf1 in virObjectLockableNew #9 0x00007f6e71d0d3e5 in virMacMapNew #10 0x00007f6e71cdc50a in findLease #11 0x00007f6e71cdcc56 in _nss_libvirt_gethostbyname4_r #12 0x00007f6e724631fc in gaih_inet #13 0x00007f6e72464697 in __GI_getaddrinfo #14 0x00007f6e71d19e81 in virGetHostnameImpl #15 0x00007f6e71d1a057 in virGetHostnameQuiet #16 0x00007f6e71d09936 in virLogOnceInit #17 0x00007f6e71d09952 in virLogOnce #18 0x00007f6e714b5829 in __pthread_once_slow #19 0x00007f6e71d16e7d in virOnce #20 0x00007f6e71d0997c in virLogInitialize #21 0x00007f6e71d0a09a in virLogVMessage #22 0x00007f6e71d09ffd in virLogMessage #23 0x00007f6e71d0db22 in virObjectNew #24 0x00007f6e71d0dbf1 in virObjectLockableNew #25 0x00007f6e71d0d3e5 in virMacMapNew #26 0x00007f6e71cdc50a in findLease #27 0x00007f6e71cdc839 in _nss_libvirt_gethostbyname3_r #28 0x00007f6e71cdc724 in _nss_libvirt_gethostbyname2_r #29 0x00007f6e7248f72f in __gethostbyname2_r #30 0x00007f6e7248f494 in gethostbyname2 #31 0x000056348c30c36d in hosts_keys #32 0x000056348c30b7d2 in main Fortunately the extra stuff virGetHostname does is totally irrelevant to the needs of the logging code, so we can just inline a call to the native hostname() syscall directly. Signed-off-by: Daniel P. Berrangé <berrange(a)redhat.com> --- src/util/virlog.c | 20 ++++++++++++++------ 1 file changed, 14 insertions(+), 6 deletions(-) diff --git a/src/util/virlog.c b/src/util/virlog.c index 8f1e4800dd..4f66cc5e5c 100644 --- a/src/util/virlog.c +++ b/src/util/virlog.c @@ -64,7 +64,7 @@ VIR_LOG_INIT("util.log"); static regex_t *virLogRegex; -static char *virLogHostname; +static char virLogHostname[HOST_NAME_MAX+1]; #define VIR_LOG_DATE_REGEX "[0-9]{4}-[0-9]{2}-[0-9]{2}" @@ -261,6 +261,8 @@ virLogPriorityString(virLogPriority lvl) static int virLogOnceInit(void) { + int r; + if (virMutexInit(&virLogMutex) < 0) return -1; @@ -275,8 +277,17 @@ virLogOnceInit(void) /* We get and remember the hostname early, because at later time * it might not be possible to load NSS modules via getaddrinfo() * (e.g. at container startup the host filesystem will not be - * accessible anymore. */ - virLogHostname = virGetHostnameQuiet(); + * accessible anymore. + * Must not use virGetHostname though as that causes re-entrancy + * problems if it triggers logging codepaths + */ + r = gethostname(virLogHostname, sizeof(virLogHostname)); + if (r == -1) { + ignore_value(virStrcpy(virLogHostname, + "(unknown)", sizeof(virLogHostname))); + } else { + NUL_TERMINATE(virLogHostname); + } virLogUnlock(); return 0; @@ -475,9 +486,6 @@ virLogHostnameString(char **rawmsg, { char *hoststr; - if (!virLogHostname) - return -1; - if (virAsprintfQuiet(&hoststr, "hostname: %s", virLogHostname) < 0) return -1; -- 2.14.3

2 2

[libvirt] [PATCH v2] qemu: Expose rx/tx_queue_size in qemu.conf too
by Michal Privoznik 12 Feb '18

12 Feb '18

In 2074ef6cd4a2 and c56cdf259 (and friends) we've added two attributes to virtio NICs: rx_queue_size and tx_queue_size. However, sysadmins might want to set these on per-host basis but don't necessarily have an access to domain XML (e.g. because they are generated by some other app). So let's expose them under qemu.conf (the settings from domain XML still take precedence as they are more specific ones). Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- diff to v1: - Reworded docs and config file comment - Simplified logic in qemuBuildNicDevStr a bit - Make the values require qemu with corresponding capabilities docs/formatdomain.html.in | 14 ++++++++-- src/qemu/libvirtd_qemu.aug | 4 +++ src/qemu/qemu.conf | 6 +++++ src/qemu/qemu_command.c | 55 +++++++++++++++++++++++++++----------- src/qemu/qemu_command.h | 3 ++- src/qemu/qemu_conf.c | 4 +++ src/qemu/qemu_conf.h | 3 +++ src/qemu/qemu_hotplug.c | 2 +- src/qemu/test_libvirtd_qemu.aug.in | 2 ++ 9 files changed, 74 insertions(+), 19 deletions(-) diff --git a/docs/formatdomain.html.in b/docs/formatdomain.html.in index 3ec1173c6..6707744bd 100644 --- a/docs/formatdomain.html.in +++ b/docs/formatdomain.html.in @@ -5456,7 +5456,12 @@ qemu-kvm -net nic,model=? /dev/null some restrictions on actual value. For instance, latest QEMU (as of 2016-09-01) requires value to be a power of two from [256, 1024] range. - Since 2.3.0 (QEMU and KVM only) + Since 2.3.0 (QEMU and KVM only) + Additionally, since 4.1.0 the + value can be set in the <code>qemu.conf</code> file in order + to override the hypervisor default value. Note that XML has + higher precedence because it's more specific. + In general you should leave this option alone, unless you are very certain you know what you are doing. @@ -5472,7 +5477,12 @@ qemu-kvm -net nic,model=? /dev/null range. In addition to that, this may work only for a subset of interface types, e.g. aforementioned QEMU enables this option only for <code>vhostuser</code> type. - Since 3.7.0 (QEMU and KVM only) + Since 3.7.0 (QEMU and KVM only) + Additionally, since 4.1.0 the + value can be set in the <code>qemu.conf</code> file in order + to override the hypervisor default value. Note that XML has + higher precedence because it's more specific. + In general you should leave this option alone, unless you are very certain you know what you are doing. diff --git a/src/qemu/libvirtd_qemu.aug b/src/qemu/libvirtd_qemu.aug index c19bf3a43..084290296 100644 --- a/src/qemu/libvirtd_qemu.aug +++ b/src/qemu/libvirtd_qemu.aug @@ -118,6 +118,9 @@ module Libvirtd_qemu = let vxhs_entry = bool_entry "vxhs_tls" | str_entry "vxhs_tls_x509_cert_dir" + let virtio_entry = int_entry "rx_queue_size" + | int_entry "tx_queue_size" + (* Each entry in the config is one of the following ... *) let entry = default_tls_entry | vnc_entry @@ -137,6 +140,7 @@ module Libvirtd_qemu = | gluster_debug_level_entry | memory_entry | vxhs_entry + | virtio_entry let comment = [ label "#comment" . del /#[ \t]*/ "# " . store /([^ \t\n][^\n]*)?/ . del /\n/ "\n" ] let empty = [ label "#empty" . eol ] diff --git a/src/qemu/qemu.conf b/src/qemu/qemu.conf index 43dd561cc..62c4265ea 100644 --- a/src/qemu/qemu.conf +++ b/src/qemu/qemu.conf @@ -775,3 +775,9 @@ # This directory is used for memoryBacking source if configured as file. # NOTE: big files will be stored here #memory_backing_dir = "/var/lib/libvirt/qemu/ram" + +# The following two values set the default RX/TX ring buffer size for virtio +# interfaces. These values are taken unless overridden in domain XML. For more +# info consult docs to corresponding attributes from domain XML. +#rx_queue_size = 1024 +#tx_queue_size = 1024 diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index 899f0cbbb..543270028 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -3751,7 +3751,8 @@ qemuBuildNicStr(virDomainNetDefPtr net, char * -qemuBuildNicDevStr(virDomainDefPtr def, +qemuBuildNicDevStr(virQEMUDriverConfigPtr cfg, + virDomainDefPtr def, virDomainNetDefPtr net, int vlan, unsigned int bootindex, @@ -3871,21 +3872,41 @@ qemuBuildNicDevStr(virDomainDefPtr def, virBufferAsprintf(&buf, ",mq=on,vectors=%zu", 2 * vhostfdSize + 2); } } - if (usingVirtio && net->driver.virtio.rx_queue_size) { - if (!virQEMUCapsGet(qemuCaps, QEMU_CAPS_VIRTIO_NET_RX_QUEUE_SIZE)) { - virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", - _("virtio rx_queue_size option is not supported with this QEMU binary")); - goto error; + if (usingVirtio) { + unsigned int rx_queue_size = net->driver.virtio.rx_queue_size; + + if (rx_queue_size == 0) + rx_queue_size = cfg->rx_queue_size; + + if (rx_queue_size) { + if (!virQEMUCapsGet(qemuCaps, QEMU_CAPS_VIRTIO_NET_RX_QUEUE_SIZE)) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", + _("virtio rx_queue_size option is " + "not supported with this QEMU binary")); + goto error; + } + + net->driver.virtio.rx_queue_size = rx_queue_size; + virBufferAsprintf(&buf, ",rx_queue_size=%u", rx_queue_size); } - virBufferAsprintf(&buf, ",rx_queue_size=%u", net->driver.virtio.rx_queue_size); } - if (usingVirtio && net->driver.virtio.tx_queue_size) { - if (!virQEMUCapsGet(qemuCaps, QEMU_CAPS_VIRTIO_NET_TX_QUEUE_SIZE)) { - virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", - _("virtio tx_queue_size option is not supported with this QEMU binary")); - goto error; + if (usingVirtio) { + unsigned int tx_queue_size = net->driver.virtio.tx_queue_size; + + if (tx_queue_size == 0) + tx_queue_size = cfg->tx_queue_size; + + if (tx_queue_size) { + if (!virQEMUCapsGet(qemuCaps, QEMU_CAPS_VIRTIO_NET_TX_QUEUE_SIZE)) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", + _("virtio tx_queue_size option is " + "not supported with this QEMU binary")); + goto error; + } + + net->driver.virtio.tx_queue_size = tx_queue_size; + virBufferAsprintf(&buf, ",tx_queue_size=%u", tx_queue_size); } - virBufferAsprintf(&buf, ",tx_queue_size=%u", net->driver.virtio.tx_queue_size); } if (usingVirtio && net->mtu) { @@ -8571,7 +8592,7 @@ qemuBuildVhostuserCommandLine(virQEMUDriverPtr driver, virCommandAddArg(cmd, netdev); VIR_FREE(netdev); - if (!(nic = qemuBuildNicDevStr(def, net, -1, bootindex, + if (!(nic = qemuBuildNicDevStr(cfg, def, net, -1, bootindex, queues, qemuCaps))) { virReportError(VIR_ERR_INTERNAL_ERROR, "%s", _("Error generating NIC -device string")); @@ -8608,6 +8629,7 @@ qemuBuildInterfaceCommandLine(virQEMUDriverPtr driver, int **nicindexes, bool chardevStdioLogd) { + virQEMUDriverConfigPtr cfg = NULL; int ret = -1; char *nic = NULL, *host = NULL; int *tapfd = NULL; @@ -8669,6 +8691,8 @@ qemuBuildInterfaceCommandLine(virQEMUDriverPtr driver, return -1; } + cfg = virQEMUDriverGetConfig(driver); + switch (actualType) { case VIR_DOMAIN_NET_TYPE_NETWORK: case VIR_DOMAIN_NET_TYPE_BRIDGE: @@ -8864,7 +8888,7 @@ qemuBuildInterfaceCommandLine(virQEMUDriverPtr driver, virCommandAddArgList(cmd, "-netdev", host, NULL); } if (qemuDomainSupportsNicdev(def, net)) { - if (!(nic = qemuBuildNicDevStr(def, net, vlan, bootindex, + if (!(nic = qemuBuildNicDevStr(cfg, def, net, vlan, bootindex, vhostfdSize, qemuCaps))) goto cleanup; virCommandAddArgList(cmd, "-device", nic, NULL); @@ -8908,6 +8932,7 @@ qemuBuildInterfaceCommandLine(virQEMUDriverPtr driver, VIR_FREE(host); VIR_FREE(tapfdName); VIR_FREE(vhostfdName); + virObjectUnref(cfg); return ret; } diff --git a/src/qemu/qemu_command.h b/src/qemu/qemu_command.h index 31c9da673..644988329 100644 --- a/src/qemu/qemu_command.h +++ b/src/qemu/qemu_command.h @@ -90,7 +90,8 @@ char *qemuBuildNicStr(virDomainNetDefPtr net, int vlan); /* Current, best practice */ -char *qemuBuildNicDevStr(virDomainDefPtr def, +char *qemuBuildNicDevStr(virQEMUDriverConfigPtr cfg, + virDomainDefPtr def, virDomainNetDefPtr net, int vlan, unsigned int bootindex, diff --git a/src/qemu/qemu_conf.c b/src/qemu/qemu_conf.c index af503d31c..2fa96431f 100644 --- a/src/qemu/qemu_conf.c +++ b/src/qemu/qemu_conf.c @@ -912,6 +912,10 @@ int virQEMUDriverConfigLoadFile(virQEMUDriverConfigPtr cfg, if (virConfGetValueString(conf, "memory_backing_dir", &cfg->memoryBackingDir) < 0) goto cleanup; + if (virConfGetValueUInt(conf, "rx_queue_size", &cfg->rx_queue_size) < 0 || + virConfGetValueUInt(conf, "tx_queue_size", &cfg->tx_queue_size) < 0) + goto cleanup; + ret = 0; cleanup: diff --git a/src/qemu/qemu_conf.h b/src/qemu/qemu_conf.h index a553e30e2..3f38a76c2 100644 --- a/src/qemu/qemu_conf.h +++ b/src/qemu/qemu_conf.h @@ -206,6 +206,9 @@ struct _virQEMUDriverConfig { bool vxhsTLS; char *vxhsTLSx509certdir; + + unsigned int rx_queue_size; + unsigned int tx_queue_size; }; /* Main driver state */ diff --git a/src/qemu/qemu_hotplug.c b/src/qemu/qemu_hotplug.c index 2aa927e62..0672ed130 100644 --- a/src/qemu/qemu_hotplug.c +++ b/src/qemu/qemu_hotplug.c @@ -1120,7 +1120,7 @@ qemuDomainAttachNetDevice(virQEMUDriverPtr driver, for (i = 0; i < vhostfdSize; i++) VIR_FORCE_CLOSE(vhostfd[i]); - if (!(nicstr = qemuBuildNicDevStr(vm->def, net, vlan, 0, + if (!(nicstr = qemuBuildNicDevStr(cfg, vm->def, net, vlan, 0, queueSize, priv->qemuCaps))) goto try_remove; diff --git a/src/qemu/test_libvirtd_qemu.aug.in b/src/qemu/test_libvirtd_qemu.aug.in index 688e5b9fd..4fc4e2f4e 100644 --- a/src/qemu/test_libvirtd_qemu.aug.in +++ b/src/qemu/test_libvirtd_qemu.aug.in @@ -100,3 +100,5 @@ module Test_libvirtd_qemu = { "1" = "mount" } } { "memory_backing_dir" = "/var/lib/libvirt/qemu/ram" } +{ "rx_queue_size" = "1024" } +{ "tx_queue_size" = "1024" } -- 2.13.6

4 4

[libvirt] [PATCH] tests: fix running of qemuxml2argvtest program
by Daniel P. Berrangé 12 Feb '18

12 Feb '18

The previous commit: commit a455d41e3e1c1af3a36ccdbb2e3f2356cc58993e Author: Daniel P. Berrangé <berrange(a)redhat.com> Date: Thu Jan 25 09:35:50 2018 +0000 conf: expand network device callbacks to cover resolving NIC type mistakenly dropped qemuxml2argvtest from the tests due to a typo. Signed-off-by: Daniel P. Berrangé <berrange(a)redhat.com> --- tests/Makefile.am | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tests/Makefile.am b/tests/Makefile.am index d013aed5eb..4ff3fa742a 100644 --- a/tests/Makefile.am +++ b/tests/Makefile.am @@ -291,7 +291,7 @@ test_programs += qemuxml2xmltest \ if WITH_NETWORK # Dep on the network driver callback for resolving NIC # actual type. XXX remove this dep. -test_programs += qemuxml2xmltest +test_programs += qemuxml2argvtest endif WITH_NETWORK test_helpers += qemucapsprobe test_libraries += libqemumonitortestutils.la \ -- 2.14.3

2 1

[libvirt] [PATCH] virutil: Introduce virGetHostnameSimple()
by Michal Privoznik 12 Feb '18

12 Feb '18

After 759b4d1b0fe5f we are getting hostname in virLogOnceInit(). Problem with this approach is in the NSS module because the module calls some internal APIs which occasionally want to log something. This results in virLogInitialize() to be called which in turn ends up calling virGetHostnameQuiet() and effectively the control gets to NSS plugin again which calls some internal APIs which occasionally want to log something. You can see the deadlock now. One way out of this is to call only gethostname() and not whole virGetHostnameQuiet() machinery. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- src/libvirt_private.syms | 1 + src/util/virlog.c | 2 +- src/util/virutil.c | 46 ++++++++++++++++++++++++++++++++++------------ src/util/virutil.h | 1 + 4 files changed, 37 insertions(+), 13 deletions(-) diff --git a/src/libvirt_private.syms b/src/libvirt_private.syms index 3b14d7d15..3ef55f809 100644 --- a/src/libvirt_private.syms +++ b/src/libvirt_private.syms @@ -3016,6 +3016,7 @@ virGetGroupList; virGetGroupName; virGetHostname; virGetHostnameQuiet; +virGetHostnameSimple; virGetListenFDs; virGetSelfLastChanged; virGetSystemPageSize; diff --git a/src/util/virlog.c b/src/util/virlog.c index 8f1e4800d..fc854ffeb 100644 --- a/src/util/virlog.c +++ b/src/util/virlog.c @@ -276,7 +276,7 @@ virLogOnceInit(void) * it might not be possible to load NSS modules via getaddrinfo() * (e.g. at container startup the host filesystem will not be * accessible anymore. */ - virLogHostname = virGetHostnameQuiet(); + virLogHostname = virGetHostnameSimple(); virLogUnlock(); return 0; diff --git a/src/util/virutil.c b/src/util/virutil.c index cd6fbf2f3..22adecd53 100644 --- a/src/util/virutil.c +++ b/src/util/virutil.c @@ -684,26 +684,23 @@ static char * virGetHostnameImpl(bool quiet) { int r; - char hostname[HOST_NAME_MAX+1], *result = NULL; + char *result; struct addrinfo hints, *info; - r = gethostname(hostname, sizeof(hostname)); - if (r == -1) { + if (!(result = virGetHostnameSimple())) { if (!quiet) virReportSystemError(errno, "%s", _("failed to determine host name")); return NULL; } - NUL_TERMINATE(hostname); - if (STRPREFIX(hostname, "localhost") || strchr(hostname, '.')) { + if (STRPREFIX(result, "localhost") || strchr(result, '.')) { /* in this case, gethostname returned localhost (meaning we can't * do any further canonicalization), or it returned an FQDN (and * we don't need to do any further canonicalization). Return the * string as-is; it's up to callers to check whether "localhost" * is allowed. */ - ignore_value(VIR_STRDUP_QUIET(result, hostname)); goto cleanup; } @@ -714,12 +711,11 @@ virGetHostnameImpl(bool quiet) memset(&hints, 0, sizeof(hints)); hints.ai_flags = AI_CANONNAME|AI_CANONIDN; hints.ai_family = AF_UNSPEC; - r = getaddrinfo(hostname, NULL, &hints, &info); + r = getaddrinfo(result, NULL, &hints, &info); if (r != 0) { if (!quiet) VIR_WARN("getaddrinfo failed for '%s': %s", - hostname, gai_strerror(r)); - ignore_value(VIR_STRDUP_QUIET(result, hostname)); + result, gai_strerror(r)); goto cleanup; } @@ -727,15 +723,16 @@ virGetHostnameImpl(bool quiet) sa_assert(info); if (info->ai_canonname == NULL || - STRPREFIX(info->ai_canonname, "localhost")) + STRPREFIX(info->ai_canonname, "localhost")) { /* in this case, we tried to canonicalize and we ended up back with * localhost. Ignore the canonicalized name and just return the * original hostname */ - ignore_value(VIR_STRDUP_QUIET(result, hostname)); - else + } else { /* Caller frees this string. */ + VIR_FREE(result); ignore_value(VIR_STRDUP_QUIET(result, info->ai_canonname)); + } freeaddrinfo(info); @@ -760,6 +757,31 @@ virGetHostnameQuiet(void) } +/** + * virGetHostnameSimple: + * + * Plain wrapper over gethostname(). The difference to + * virGetHostname() is that this function doesn't try to + * canonicalize the hostname. + * + * Returns: hostname string (caller must free), + * NULL on error. + */ +char * +virGetHostnameSimple(void) +{ + char hostname[HOST_NAME_MAX+1]; + char *ret; + + if (gethostname(hostname, sizeof(hostname)) == -1) + return NULL; + + NUL_TERMINATE(hostname); + ignore_value(VIR_STRDUP_QUIET(ret, hostname)); + return ret; +} + + char * virGetUserDirectory(void) { diff --git a/src/util/virutil.h b/src/util/virutil.h index be0f6b0ea..57148374b 100644 --- a/src/util/virutil.h +++ b/src/util/virutil.h @@ -136,6 +136,7 @@ static inline int pthread_sigmask(int how, char *virGetHostname(void); char *virGetHostnameQuiet(void); +char *virGetHostnameSimple(void); char *virGetUserDirectory(void); char *virGetUserDirectoryByUID(uid_t uid); -- 2.13.6

2 2

[libvirt] [PATCH v3 00/15] Misc build refactoring / isolation work
by Daniel P. Berrangé 12 Feb '18

12 Feb '18

This was triggered by the recent Fedora change to add '-z defs' to RPM builds by default which breaks libvirt. Various make rule changes can fix much of the problem, but it also requires source refactoring to get rid of places where virt drivers directly call into the storage/network drivers. Co-incidentally this work will also be useful in allowing us to separate out drivers to distinct daemons. In v3: - Fixed a few build problems identified by travis In v2: - Fixed header file name comment - Resolve conflicts - Fix unit tests - Fix bisectable build by moving libvirt_lxc build patch earlier - Update syntax check header include rule Daniel P. Berrangé (15): storage: extract storage file backend from main storage driver backend storage: move storage file backend framework into util directory rpc: don't link in second copy of RPC code to libvirtd & lockd plugin build: link libvirt_lxc against libvirt.so conf: introduce callback registration for domain net device allocation conf: expand network device callbacks to cover bandwidth updates qemu: replace networkGetNetworkAddress with public API calls conf: expand network device callbacks to cover resolving NIC type network: remove conditional declarations conf: move virStorageTranslateDiskSourcePool into domain conf storage: export virStoragePoolLookupByTargetPath as a public API build: explicitly link all modules with libvirt.so build: provide a AM_FLAGS_MOD for loadable modules build: passing the "-z defs" linker flag to prevent undefined symbols cfg: forbid includes of headers in network and storage drivers again cfg.mk | 2 +- configure.ac | 1 + daemon/Makefile.am | 3 +- include/libvirt/libvirt-storage.h | 2 + m4/virt-linker-no-undefined.m4 | 32 ++ po/POTFILES.in | 2 +- src/Makefile.am | 150 ++++---- src/bhyve/bhyve_command.c | 7 +- src/conf/domain_conf.c | 355 +++++++++++++++++++ src/conf/domain_conf.h | 71 ++++ src/driver-storage.h | 5 + src/libvirt-storage.c | 40 +++ src/libvirt_private.syms | 29 ++ src/libvirt_public.syms | 6 + src/libvirt_remote.syms | 11 +- src/libxl/libxl_domain.c | 5 +- src/libxl/libxl_driver.c | 7 +- src/lxc/lxc_driver.c | 5 +- src/lxc/lxc_process.c | 7 +- src/network/bridge_driver.c | 124 +------ src/network/bridge_driver.h | 72 ---- src/qemu/qemu_alias.c | 3 +- src/qemu/qemu_command.c | 1 - src/qemu/qemu_domain.c | 3 - src/qemu/qemu_domain_address.c | 3 +- src/qemu/qemu_driver.c | 15 +- src/qemu/qemu_hotplug.c | 18 +- src/qemu/qemu_migration.c | 3 +- src/qemu/qemu_process.c | 115 +++++- src/remote/remote_driver.c | 1 + src/remote/remote_protocol.x | 17 +- src/remote_protocol-structs | 7 + src/security/virt-aa-helper.c | 2 - src/storage/storage_backend.c | 66 ---- src/storage/storage_backend.h | 75 ---- src/storage/storage_backend_fs.c | 8 +- src/storage/storage_backend_gluster.c | 4 +- src/storage/storage_driver.c | 256 +------------- src/storage/storage_driver.h | 3 - src/storage/storage_source.c | 645 ---------------------------------- src/storage/storage_source.h | 59 ---- src/util/virstoragefile.c | 609 +++++++++++++++++++++++++++++++- src/util/virstoragefile.h | 32 ++ src/util/virstoragefilebackend.c | 108 ++++++ src/util/virstoragefilebackend.h | 104 ++++++ src/vz/vz_sdk.c | 1 - tests/Makefile.am | 11 +- tests/qemuxml2argvtest.c | 4 + tests/virstoragetest.c | 1 - tools/Makefile.am | 1 + 50 files changed, 1679 insertions(+), 1432 deletions(-) create mode 100644 m4/virt-linker-no-undefined.m4 delete mode 100644 src/storage/storage_source.c delete mode 100644 src/storage/storage_source.h create mode 100644 src/util/virstoragefilebackend.c create mode 100644 src/util/virstoragefilebackend.h -- 2.14.3

3 19