April 2013 - Devel - libvirt List Archives

[libvirt] [PATCH 0/2] Add support for translations
by Daniel P. Berrange 10 Apr '13

10 Apr '13

libvirt-glib contains a bunch of user visible error message strings which need translation. These patches wire up the glib/intltool/gettext i18n framework I have created a transifex project to go with it https://www.transifex.com/projects/p/libvirt-glib/

3 8

Re: [libvirt] [Qemu-devel] GSoC 2013, "Introduce API to query IP addresses for given domain" Libvirt project
by Stefan Hajnoczi 10 Apr '13

10 Apr '13

On Wed, Apr 10, 2013 at 04:46:45AM +0200, Tabibel Sami wrote: > I am doing master degree of Cryptology and information security > I have good C and Python skills, and i am interested to work on > "Introduce API to query IP addresses for given domain" Libvirt > project. > > I am looking for any comment about the difficulty and the content of > the project, and also about my chances to be accepted if i apply for. Adding libvirt mailing list and CCing mprivozn who is the mentor for this project. Stefan

2 1

[libvirt] VNC Server in libvirt
by Jaime Loureiro Acuña 10 Apr '13

10 Apr '13

My name is Jaime Loureiro and I work in the Network & Applications area in University of Vigo. I am writing to you hoping you can help me with some libvirt graphics issues. I was analysing the source code of the libvirt in order to find how to access the display of each virtual machine. We are trying to generate a video streaming from the graphics content, though it is not easy to find what function is the most appropriate to do that. We tried to use the screenshot feature, but the rate is to low to generate a smooth streaming. I suposse the vnc server is reading the graphics output of each virtual machine at a low level, however I cannot find the exact point where the server is. Currently I am trying to understand the function "virDomainOpenGraphics" in libvirt.c and, at a lower level, the function "qemuProcessHandleGraphics" in qemu_process.c, but I 'm not sure if it is possible to get what we need from them. I would appreciate any help or clue you could provide regarding this issue. Thank you very much in advance. Best regards,

2 1

[libvirt] [PATCH 0/2] Add version comparison method for virsh_version output
by John Ferlan 10 Apr '13

10 Apr '13

I believe this set of patches should be applied instead of those from https://www.redhat.com/archives/libvirt-cim/2013-April/msg00006.html I suppose we could consider combining the two functions; however, there are a couple place where the return version string is compared more than once, so I kept both interfaces. The vxml change is just an informative message that showed me the name of the image that was failing - I had created 'default-kvm-dimage.img' and tests were failing to find 'default-kvm-dimage' to place into the cimtest storage pool. John Ferlan (2): vxml: Add which volume could not be found to error message Add and utilize virsh_version_cmp .../08_CreateDiskResourcePool.py | 4 ++-- .../10_create_storagevolume.py | 6 ++++-- .../11_create_dir_storagevolume_errs.py | 5 +++-- .../12_create_netfs_storagevolume_errs.py | 5 +++-- .../13_delete_storagevolume.py | 5 +++-- .../14_delete_storagevolume_errs.py | 5 +++-- .../ResourcePoolConfigurationService/15_DiskPoolAutostart.py | 4 ++-- .../cimtest/SettingsDefineCapabilities/01_forward.py | 1 - .../VirtualSystemManagementService/19_definenetwork_ers.py | 8 ++++---- suites/libvirt-cim/lib/XenKvmLib/common_util.py | 9 +++++---- suites/libvirt-cim/lib/XenKvmLib/pool.py | 6 +++--- suites/libvirt-cim/lib/XenKvmLib/rasd.py | 10 ++++++---- suites/libvirt-cim/lib/XenKvmLib/vxml.py | 4 ++-- suites/libvirt-cim/lib/XenKvmLib/xm_virt_util.py | 6 ++++++ 14 files changed, 46 insertions(+), 32 deletions(-) -- 1.8.1.4

4 5

[libvirt] [PATCH 0/2] Fix domain list race crash
by Peter Krempa 10 Apr '13

10 Apr '13

This series fixes the race between freeing and accessing of domain objects and provides a reproducer case to test it. Peter Krempa (2): conf: Fix race between looking up a domain object and freeing it DO NOT APPLY UPSTREAM: Reproducer of domain free crash src/conf/domain_conf.c | 8 ++++++++ src/qemu/qemu_driver.c | 6 ++++++ 2 files changed, 14 insertions(+) -- 1.8.1.5

3 5

[libvirt] [PATCH 00/12] A bunch of extensions to libxl driver
by Marek Marczykowski 10 Apr '13

10 Apr '13

This are some additional features to libxl driver. Some of them require change in domain config structures/syntax. Details described with each patch. The last two patches are bugfix for deadlock during daemon startup. PS Please keep me in To/Cc - I'm not subscribed to the list. Marek Marczykowski (12): libxl: allow script for any network interface, not only bridge libxl: PCI passthrough support libxl: nodeDevice* support for PCI devices libxl: populate xenstore memory entries at startup conf: add 'script' attribute to disk specification libxl: use disk 'script' attribute conf: support backend domain name in disk and network devices libxl: support backend domain setting for disk and net devices libxl: fill HVM SDL and VNC settings based on <graphics/> entries RFC: libxl: special 'stubdom-dm' emulator to use qemu in stub domain conf: virDomainObjListRemoveLocked function libxl: fix deadlock in libxlReconnectDomain docs/schemas/domaincommon.rng | 14 ++ src/conf/domain_conf.c | 54 ++++++++ src/conf/domain_conf.h | 5 + src/libvirt_private.syms | 1 + src/libxl/libxl_conf.c | 297 +++++++++++++++++++++++++++++++++--------- src/libxl/libxl_conf.h | 6 +- src/libxl/libxl_driver.c | 253 +++++++++++++++++++++++++++++++---- 7 files changed, 545 insertions(+), 85 deletions(-) -- 1.8.1.4

2 2

[libvirt] [PATCH 03/12] libxl: nodeDevice* support for PCI devices
by Marek Marczykowski 10 Apr '13

10 Apr '13

For now only for PCI devices. Mostly copy-paste from old xen driver. --- src/libxl/libxl_driver.c | 193 +++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 193 insertions(+) diff --git a/src/libxl/libxl_driver.c b/src/libxl/libxl_driver.c index 40a7a6b..011edf8 100644 --- a/src/libxl/libxl_driver.c +++ b/src/libxl/libxl_driver.c @@ -42,6 +42,7 @@ #include "libxl.h" #include "libxl_driver.h" #include "libxl_conf.h" +#include "node_device_conf.h" #include "xen_xm.h" #include "virtypedparam.h" #include "viruri.h" @@ -3666,6 +3667,195 @@ libxlDomainUpdateDeviceFlags(virDomainPtr dom, const char *xml, return libxlDomainModifyDeviceFlags(dom, xml, flags, LIBXL_DEVICE_UPDATE); } +static int +libxlNodeDeviceGetPciInfo(virNodeDevicePtr dev, + unsigned *domain, + unsigned *bus, + unsigned *slot, + unsigned *function) +{ + virNodeDeviceDefPtr def = NULL; + virNodeDevCapsDefPtr cap; + char *xml = NULL; + int ret = -1; + + xml = virNodeDeviceGetXMLDesc(dev, 0); + if (!xml) + goto out; + + def = virNodeDeviceDefParseString(xml, EXISTING_DEVICE, NULL); + if (!def) + goto out; + + cap = def->caps; + while (cap) { + if (cap->type == VIR_NODE_DEV_CAP_PCI_DEV) { + *domain = cap->data.pci_dev.domain; + *bus = cap->data.pci_dev.bus; + *slot = cap->data.pci_dev.slot; + *function = cap->data.pci_dev.function; + break; + } + + cap = cap->next; + } + + if (!cap) { + virReportError(VIR_ERR_INVALID_ARG, + _("device %s is not a PCI device"), dev->name); + goto out; + } + + ret = 0; +out: + virNodeDeviceDefFree(def); + VIR_FREE(xml); + return ret; +} + +static int +libxlNodeDeviceDettach(virNodeDevicePtr dev) +{ + virPCIDevicePtr pci; + unsigned domain, bus, slot, function; + int ret = -1; + + if (libxlNodeDeviceGetPciInfo(dev, &domain, &bus, &slot, &function) < 0) + return -1; + + pci = virPCIDeviceNew(domain, bus, slot, function); + if (!pci) + return -1; + + if (virPCIDeviceDetach(pci, NULL, NULL, "pciback") < 0) + goto out; + + ret = 0; +out: + virPCIDeviceFree(pci); + return ret; +} + +static int +libxlNodeDeviceAssignedDomainId(virNodeDevicePtr dev) +{ + int numdomains; + int numpcidevs; + int ret = -1, i, j; + int *ids = NULL; + char *bdf = NULL; + char *xref = NULL; + unsigned int domain, bus, slot, function; + libxl_device_pci *pcidevs = NULL; + virConnectPtr conn = dev->conn; + libxlDriverPrivatePtr driver = conn->privateData; + + /* Get active domains */ + numdomains = libxlNumDomains(conn); + if (numdomains < 0) { + return ret; + } + if (numdomains > 0){ + if (VIR_ALLOC_N(ids, numdomains) < 0) { + virReportOOMError(); + goto out; + } + if ((numdomains = libxlListDomains(conn, &ids[0], numdomains)) < 0) { + goto out; + } + } + + /* Get pci bdf */ + if (libxlNodeDeviceGetPciInfo(dev, &domain, &bus, &slot, &function) < 0) + goto out; + + if (virAsprintf(&bdf, "%04x:%02x:%02x.%0x", + domain, bus, slot, function) < 0) { + virReportOOMError(); + goto out; + } + + libxlDriverLock(driver); + /* Check if bdf is assigned to one of active domains */ + for (i = 0; i < numdomains; i++) { + pcidevs = libxl_device_pci_list(driver->ctx, ids[i], &numpcidevs); + if (pcidevs == NULL) { + continue; + } else { + for (j = 0; j < numpcidevs; j++) { + if (pcidevs[j].bus == bus && pcidevs[j].dev == slot && pcidevs[j].func == function) { + ret = ids[i]; + break; + } + } + VIR_FREE(pcidevs); + } + } + libxlDriverUnlock(driver); + + VIR_FREE(xref); + VIR_FREE(bdf); +out: + VIR_FREE(ids); + + return ret; +} + +static int +libxlNodeDeviceReAttach(virNodeDevicePtr dev) +{ + virPCIDevicePtr pci; + unsigned domain, bus, slot, function; + int ret = -1; + int domid; + + if (libxlNodeDeviceGetPciInfo(dev, &domain, &bus, &slot, &function) < 0) + return -1; + + pci = virPCIDeviceNew(domain, bus, slot, function); + if (!pci) + return -1; + + /* Check if device is assigned to an active guest */ + if ((domid = libxlNodeDeviceAssignedDomainId(dev)) >= 0) { + virReportError(VIR_ERR_INTERNAL_ERROR, + _("Device %s has been assigned to guest %d"), + dev->name, domid); + goto out; + } + + if (virPCIDeviceReattach(pci, NULL, NULL, "pciback") < 0) + goto out; + + ret = 0; +out: + virPCIDeviceFree(pci); + return ret; +} + +static int +libxlNodeDeviceReset(virNodeDevicePtr dev) +{ + virPCIDevicePtr pci; + unsigned domain, bus, slot, function; + int ret = -1; + + if (libxlNodeDeviceGetPciInfo(dev, &domain, &bus, &slot, &function) < 0) + return -1; + + pci = virPCIDeviceNew(domain, bus, slot, function); + if (!pci) + return -1; + + if (virPCIDeviceReset(pci, NULL, NULL) < 0) + goto out; + + ret = 0; +out: + virPCIDeviceFree(pci); + return ret; +} + static unsigned long long libxlNodeGetFreeMemory(virConnectPtr conn) { @@ -4233,6 +4423,9 @@ static virDriver libxlDriver = { .nodeGetFreeMemory = libxlNodeGetFreeMemory, /* 0.9.0 */ .domainEventRegister = libxlDomainEventRegister, /* 0.9.0 */ .domainEventDeregister = libxlDomainEventDeregister, /* 0.9.0 */ + .nodeDeviceDettach = libxlNodeDeviceDettach, /* 1.0.4 */ + .nodeDeviceReAttach = libxlNodeDeviceReAttach, /* 1.0.4 */ + .nodeDeviceReset = libxlNodeDeviceReset, /* 1.0.4 */ .domainManagedSave = libxlDomainManagedSave, /* 0.9.2 */ .domainHasManagedSaveImage = libxlDomainHasManagedSaveImage, /* 0.9.2 */ .domainManagedSaveRemove = libxlDomainManagedSaveRemove, /* 0.9.2 */ -- 1.8.1.4

1 0

[libvirt] [PATCH 04/12] libxl: populate xenstore memory entries at startup
by Marek Marczykowski 10 Apr '13

10 Apr '13

libxl uses some xenstore entries for hints in memory management (especially when starting new domain). This includes dom0 memory limit and Xen free memory margin, based on current system state. Entries are created at first usage, so force such usage at daemon startup, which most likely will be before any domain startup. --- src/libxl/libxl_driver.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/src/libxl/libxl_driver.c b/src/libxl/libxl_driver.c index 011edf8..89546a5 100644 --- a/src/libxl/libxl_driver.c +++ b/src/libxl/libxl_driver.c @@ -1115,6 +1115,7 @@ libxlStartup(bool privileged, char *log_file = NULL; virCommandPtr cmd; int status, ret = 0; + unsigned int free_mem; char ebuf[1024]; /* Disable libxl driver if non-root */ @@ -1240,6 +1241,13 @@ libxlStartup(bool privileged, goto error; } + /* This will fill xenstore info about free and dom0 memory - if missing, + * should be called before starting first domain */ + if (libxl_get_free_memory(libxl_driver->ctx, &free_mem)) { + VIR_ERROR(_("cannot get free memory info")); + goto error; + } + if (!(libxl_driver->xmlconf = virDomainXMLConfNew(&libxlDomainXMLPrivateDataCallbacks, NULL))) goto error; -- 1.8.1.4

1 0

[libvirt] [PATCH 02/12] libxl: PCI passthrough support
by Marek Marczykowski 10 Apr '13

10 Apr '13

--- src/libxl/libxl_conf.c | 72 ++++++++++++++++++++++++++++++++++++++++++++++++++ src/libxl/libxl_conf.h | 2 ++ 2 files changed, 74 insertions(+) diff --git a/src/libxl/libxl_conf.c b/src/libxl/libxl_conf.c index ffc7bbb..7668305 100644 --- a/src/libxl/libxl_conf.c +++ b/src/libxl/libxl_conf.c @@ -759,6 +759,74 @@ error: return -1; } +int +libxlMakePci(virDomainHostdevDefPtr l_hostdev, + libxl_device_pci *x_pci) +{ + if (l_hostdev->mode != VIR_DOMAIN_HOSTDEV_MODE_SUBSYS) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", _("This driver supports only subsystem host device mode (PCI devices only)")); + return -1; + } + if (l_hostdev->source.subsys.type != VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_PCI) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", _("This driver supports only PCI host devices")); + return -1; + } + + x_pci->bus = l_hostdev->source.subsys.u.pci.bus; + x_pci->dev = l_hostdev->source.subsys.u.pci.slot; + x_pci->func = l_hostdev->source.subsys.u.pci.function; + return 0; +} + +static int +libxlMakePciList(virDomainDefPtr def, + libxl_domain_config *d_config) +{ + virDomainHostdevDefPtr *l_hostdevs = def->hostdevs; + int nhostdevs = def->nhostdevs; + int npcidevs = 0; + libxl_device_pci *x_pcidevs; + int i, j; + + if (nhostdevs == 0) + return 0; + + /* count PCI devices, ignore others (USB) */ + for (i = 0; i < nhostdevs; i++) { + if (l_hostdevs[i]->mode == VIR_DOMAIN_HOSTDEV_MODE_SUBSYS && + l_hostdevs[i]->source.subsys.type == VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_PCI) + npcidevs++; + } + if (VIR_ALLOC_N(x_pcidevs, npcidevs) < 0) { + virReportOOMError(); + return -1; + } + + for (i = 0, j = 0; i < nhostdevs; i++) { + if (l_hostdevs[i]->mode != VIR_DOMAIN_HOSTDEV_MODE_SUBSYS || + l_hostdevs[i]->source.subsys.type != VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_PCI) + continue; + + libxl_device_pci_init(&x_pcidevs[j]); + + if (libxlMakePci(l_hostdevs[i], &x_pcidevs[j]) < 0) + goto error; + j++; + } + + d_config->pcidevs = x_pcidevs; + d_config->num_pcidevs = npcidevs; + + return 0; + +error: + for (i = 0; i < npcidevs; i++) { + libxl_device_pci_dispose(&x_pcidevs[i]); + } + VIR_FREE(x_pcidevs); + return -1; +} + virCapsPtr libxlMakeCapabilities(libxl_ctx *ctx) { @@ -817,6 +885,10 @@ libxlBuildDomainConfig(libxlDriverPrivatePtr driver, goto error; } + if (libxlMakePciList(def, d_config) < 0) { + goto error; + } + d_config->on_reboot = def->onReboot; d_config->on_poweroff = def->onPoweroff; d_config->on_crash = def->onCrash; diff --git a/src/libxl/libxl_conf.h b/src/libxl/libxl_conf.h index f8602b4..b3ab3bf 100644 --- a/src/libxl/libxl_conf.h +++ b/src/libxl/libxl_conf.h @@ -119,6 +119,8 @@ libxlMakeNic(virDomainNetDefPtr l_nic, libxl_device_nic *x_nic); int libxlMakeVfb(libxlDriverPrivatePtr driver, virDomainGraphicsDefPtr l_vfb, libxl_device_vfb *x_vfb); +int +libxlMakePci(virDomainHostdevDefPtr l_hostdev, libxl_device_pci *x_pci); int libxlBuildDomainConfig(libxlDriverPrivatePtr driver, -- 1.8.1.4

1 0

[libvirt] [PATCH 01/12] libxl: allow script for any network interface, not only bridge
by Marek Marczykowski 10 Apr '13

10 Apr '13

This can be useful for route or NAT networks, or any other custom network setup. Especially configuration example in documentation uses <script/> tag with type 'ethernet'. --- src/libxl/libxl_conf.c | 17 +++++------------ 1 file changed, 5 insertions(+), 12 deletions(-) diff --git a/src/libxl/libxl_conf.c b/src/libxl/libxl_conf.c index b208dd8..ffc7bbb 100644 --- a/src/libxl/libxl_conf.c +++ b/src/libxl/libxl_conf.c @@ -608,18 +608,11 @@ libxlMakeNic(virDomainNetDefPtr l_nic, libxl_device_nic *x_nic) virReportOOMError(); return -1; } - if (l_nic->script && - (x_nic->script = strdup(l_nic->script)) == NULL) { - virReportOOMError(); - return -1; - } - } else { - if (l_nic->script) { - virReportError(VIR_ERR_CONFIG_UNSUPPORTED, - _("scripts are not supported on interfaces of type %s"), - virDomainNetTypeToString(l_nic->type)); - return -1; - } + } + if (l_nic->script && + (x_nic->script = strdup(l_nic->script)) == NULL) { + virReportOOMError(); + return -1; } return 0; -- 1.8.1.4

1 0