- Devel - libvirt List Archives

[libvirt] error: failed to connect to the hypervisor
by Liu, Zhentao 18 Oct '09

18 Oct '09

Hello, I have installed libvirt-bin 0.6.1 on the ubuntu 9.04. When I run "virsh -c qemu:///system list", I get this error: ---------------------------------------------------------------------------- unable to connect to '/var/run/libvirt/libvirt-sock': Permission denied failed to connect to the hypervisor ----------------------------------------------------------------------------- Then I search in the Internet, it may be a bug of version 0.6.1. So I update libvirt-bin to version 0.6.3, but I still get a error: --------------------------------------------------------------------------------- root@forest:/home/zli# virsh -c qemu:///system list error: failed to connect to the hypervisor ---------------------------------------------------------------------------------- I have no idea about the reason. Any helps? Regards Zhentao

2 1

[libvirt] [PATCH] util: Make sure random data is initialized when in virRandom
by Cole Robinson 16 Oct '09

16 Oct '09

When writing some tests, I mistakenly attempted to auto-generate a UUID, which caused a segfault: virRandom was being used without calling virRandomInitialize. Make sure this case can't happen. Signed-off-by: Cole Robinson <crobinso(a)redhat.com> --- src/util/util.c | 8 ++++++++ 1 files changed, 8 insertions(+), 0 deletions(-) diff --git a/src/util/util.c b/src/util/util.c index e5135fc..1554097 100644 --- a/src/util/util.c +++ b/src/util/util.c @@ -1910,6 +1910,7 @@ int virKillProcess(pid_t pid, int sig) } +static int random_initialized = 0; static char randomState[128]; static struct random_data randomData; static virMutex randomLock; @@ -1925,6 +1926,7 @@ int virRandomInitialize(unsigned int seed) &randomData) < 0) return -1; + random_initialized = 1; return 0; } @@ -1932,6 +1934,12 @@ int virRandom(int max) { int32_t ret; + if (!random_initialized) { + /* This can error, but what's worse? Unnoticed bogus random data or + * a segfault? */ + virRandomInitialize(time(NULL) ^ getpid()); + } + virMutexLock(&randomLock); random_r(&randomData, &ret); virMutexUnlock(&randomLock); -- 1.6.5.rc2

2 2

[libvirt] node device libudev backend second look
by Dave Allan 16 Oct '09

16 Oct '09

Attached is a patch against the current head containing an implementation of node device enumeration using libudev. It is complete except for the monitor, but I'm submitting it now as I have a few questions about the implementation that I'd like advice on. They are marked XXX in comments in the patch. The other thing that's not clear to me is how the code generates the tree structure for nodedev-list --tree. I'm setting the parent pointer to what I think is correct, but the tree output is broken. I can dig through it until I understand it, but if anyone is familiar with the implementation and would be willing to take a few minutes to walk me through it, it would save me a bunch of time. I think it's also important that people get the code installed on a variety of systems as soon as possible to shake out the inevitable bugs that will arise from differing device models and code versions, and I'll have the final version with the monitor shortly. Dave >From efedc8fee8127b32fb5b761ce5035902d168a207 Mon Sep 17 00:00:00 2001 From: David Allan <dallan(a)redhat.com> Date: Fri, 25 Sep 2009 10:59:28 -0400 Subject: [PATCH 1/1] A second look at the libudev node device backend. This patch is a very much cleaned up version of the libudev based node device backend with all the device types implemented, although I need some guidance on the USB interface code as I am not especially familiar with the terminology and HAL and udev display the details quite differently. This version of the code uses libudev to retrieve all the device information, so there's no manual mucking about in sysfs except for a little bit of existing code for fibre channel adapters. There is a lot of detail work in this code, so we should try to get people running it on a wide variety of hardware so we can shake out the differences in implementation between the HAL and libudev backends. --- configure.in | 47 ++- daemon/libvirtd.c | 3 +- src/Makefile.am | 16 +- src/conf/node_device_conf.c | 15 +- src/conf/node_device_conf.h | 5 + src/node_device/node_device_driver.c | 12 +- src/node_device/node_device_driver.h | 22 + src/node_device/node_device_hal.h | 19 - ...evice_hal_linux.c => node_device_linux_sysfs.c} | 0 src/node_device/node_device_udev.c | 800 ++++++++++++++++++++ src/node_device/node_device_udev.h | 27 + src/util/util.c | 28 + src/util/util.h | 3 + 13 files changed, 967 insertions(+), 30 deletions(-) rename src/node_device/{node_device_hal_linux.c => node_device_linux_sysfs.c} (100%) create mode 100644 src/node_device/node_device_udev.c create mode 100644 src/node_device/node_device_udev.h diff --git a/configure.in b/configure.in index 2f9db72..25ed55f 100644 --- a/configure.in +++ b/configure.in @@ -1640,7 +1640,7 @@ test "$enable_shared" = no && lt_cv_objdir=. LV_LIBTOOL_OBJDIR=${lt_cv_objdir-.} AC_SUBST([LV_LIBTOOL_OBJDIR]) -dnl HAL or DeviceKit library for host device enumeration +dnl HAL, DeviceKit, or libudev library for host device enumeration HAL_REQUIRED=0.0 HAL_CFLAGS= HAL_LIBS= @@ -1734,8 +1734,46 @@ AM_CONDITIONAL([HAVE_DEVKIT], [test "x$with_devkit" = "xyes"]) AC_SUBST([DEVKIT_CFLAGS]) AC_SUBST([DEVKIT_LIBS]) +UDEV_REQUIRED=143 +UDEV_CFLAGS= +UDEV_LIBS= +AC_ARG_WITH([udev], + [ --with-udev use libudev for host device enumeration], + [], + [with_udev=check]) + +if test "$with_libvirtd" = "no" ; then + with_udev=no +fi +if test "x$with_udev" = "xyes" -o "x$with_udev" = "xcheck"; then + PKG_CHECK_MODULES(UDEV, libudev >= $UDEV_REQUIRED, + [with_udev=yes], [ + if test "x$with_udev" = "xcheck" ; then + with_udev=no + else + AC_MSG_ERROR( + [You must install udev-devel >= $UDEV_REQUIRED to compile libvirt]) + fi + ]) + if test "x$with_udev" = "xyes" ; then + AC_DEFINE_UNQUOTED([HAVE_UDEV], 1, + [use UDEV for host device enumeration]) + + old_CFLAGS=$CFLAGS + old_LDFLAGS=$LDFLAGS + CFLAGS="$CFLAGS $UDEV_CFLAGS" + LDFLAGS="$LDFLAGS $UDEV_LIBS" + AC_CHECK_FUNCS([udev_new],,[with_udev=no]) + CFLAGS="$old_CFLAGS" + LDFLAGS="$old_LDFLAGS" + fi +fi +AM_CONDITIONAL([HAVE_UDEV], [test "x$with_udev" = "xyes"]) +AC_SUBST([UDEV_CFLAGS]) +AC_SUBST([UDEV_LIBS]) + with_nodedev=no; -if test "$with_devkit" = "yes" -o "$with_hal" = "yes"; +if test "$with_devkit" = "yes" -o "$with_hal" = "yes" -o "$with_udev" = "yes"; then with_nodedev=yes AC_DEFINE_UNQUOTED([WITH_NODE_DEVICES], 1, [with node device driver]) @@ -1900,6 +1938,11 @@ AC_MSG_NOTICE([ devkit: $DEVKIT_CFLAGS $DEVKIT_LIBS]) else AC_MSG_NOTICE([ devkit: no]) fi +if test "$with_udev" = "yes" ; then +AC_MSG_NOTICE([ udev: $UDEV_CFLAGS $UDEV_LIBS]) +else +AC_MSG_NOTICE([ udev: no]) +fi if test "$with_netcf" = "yes" ; then AC_MSG_NOTICE([ netcf: $NETCF_CFLAGS $NETCF_LIBS]) else diff --git a/daemon/libvirtd.c b/daemon/libvirtd.c index 78dfb2d..0c03aae 100644 --- a/daemon/libvirtd.c +++ b/daemon/libvirtd.c @@ -833,8 +833,7 @@ static struct qemud_server *qemudInitialize(int sigread) { #ifdef WITH_STORAGE_DIR storageRegister(); #endif -#if defined(WITH_NODE_DEVICES) && \ - (defined(HAVE_HAL) || defined(HAVE_DEVKIT)) +#if defined(WITH_NODE_DEVICES) nodedevRegister(); #endif secretRegister(); diff --git a/src/Makefile.am b/src/Makefile.am index d0ef7d1..f6f63a2 100644 --- a/src/Makefile.am +++ b/src/Makefile.am @@ -254,16 +254,20 @@ SECURITY_DRIVER_APPARMOR_SOURCES = \ NODE_DEVICE_DRIVER_SOURCES = \ - node_device/node_device_driver.c node_device/node_device_driver.h + node_device/node_device_driver.c \ + node_device/node_device_driver.h \ + node_device/node_device_linux_sysfs.c NODE_DEVICE_DRIVER_HAL_SOURCES = \ node_device/node_device_hal.c \ - node_device/node_device_hal.h \ - node_device/node_device_hal_linux.c + node_device/node_device_hal.h NODE_DEVICE_DRIVER_DEVKIT_SOURCES = \ node_device/node_device_devkit.c +NODE_DEVICE_DRIVER_UDEV_SOURCES = \ + node_device/node_device_udev.c + ######################### # @@ -642,6 +646,11 @@ libvirt_driver_nodedev_la_SOURCES += $(NODE_DEVICE_DRIVER_DEVKIT_SOURCES) libvirt_driver_nodedev_la_CFLAGS += $(DEVKIT_CFLAGS) libvirt_driver_nodedev_la_LDFLAGS += $(DEVKIT_LIBS) endif +if HAVE_UDEV +libvirt_driver_nodedev_la_SOURCES += $(NODE_DEVICE_DRIVER_UDEV_SOURCES) +libvirt_driver_nodedev_la_CFLAGS += $(UDEV_CFLAGS) +libvirt_driver_nodedev_la_LDFLAGS += $(UDEV_LIBS) +endif if WITH_DRIVER_MODULES libvirt_driver_nodedev_la_LDFLAGS += -module -avoid-version @@ -689,6 +698,7 @@ EXTRA_DIST += \ $(NODE_DEVICE_DRIVER_SOURCES) \ $(NODE_DEVICE_DRIVER_HAL_SOURCES) \ $(NODE_DEVICE_DRIVER_DEVKIT_SOURCES) \ + $(NODE_DEVICE_DRIVER_UDEV_SOURCES) \ $(SECURITY_DRIVER_SELINUX_SOURCES) \ $(SECURITY_DRIVER_APPARMOR_SOURCES) \ $(SECRET_DRIVER_SOURCES) \ diff --git a/src/conf/node_device_conf.c b/src/conf/node_device_conf.c index f09f814..2322819 100644 --- a/src/conf/node_device_conf.c +++ b/src/conf/node_device_conf.c @@ -325,6 +325,9 @@ char *virNodeDeviceDefFormat(virConnectPtr conn, data->usb_if.subclass); virBufferVSprintf(&buf, " <protocol>%d</protocol>\n", data->usb_if.protocol); + if (data->usb_if.interface_name) + virBufferVSprintf(&buf, " <interface_name>%s</interface_name>\n", + data->usb_if.interface_name); if (data->usb_if.description) virBufferVSprintf(&buf, " <description>%s</description>\n", data->usb_if.description); @@ -394,10 +397,19 @@ char *virNodeDeviceDefFormat(virConnectPtr conn, "</media_available>\n", avl ? 1 : 0); virBufferVSprintf(&buf, " <media_size>%llu</media_size>\n", data->storage.removable_media_size); - virBufferAddLit(&buf, " </capability>\n"); + virBufferVSprintf(&buf, " <logical_block_size>%llu" + "</logical_block_size>\n", + data->storage.logical_block_size); + virBufferVSprintf(&buf, " <num_blocks>%llu</num_blocks>\n", + data->storage.num_blocks); } else { virBufferVSprintf(&buf, " <size>%llu</size>\n", data->storage.size); + virBufferVSprintf(&buf, " <logical_block_size>%llu" + "</logical_block_size>\n", + data->storage.logical_block_size); + virBufferVSprintf(&buf, " <num_blocks>%llu</num_blocks>\n", + data->storage.num_blocks); } if (data->storage.flags & VIR_NODE_DEV_CAP_STORAGE_HOTPLUGGABLE) virBufferAddLit(&buf, @@ -1239,6 +1251,7 @@ void virNodeDevCapsDefFree(virNodeDevCapsDefPtr caps) VIR_FREE(data->usb_dev.vendor_name); break; case VIR_NODE_DEV_CAP_USB_INTERFACE: + VIR_FREE(data->usb_if.interface_name); VIR_FREE(data->usb_if.description); break; case VIR_NODE_DEV_CAP_NET: diff --git a/src/conf/node_device_conf.h b/src/conf/node_device_conf.h index 29a4d43..95910c5 100644 --- a/src/conf/node_device_conf.h +++ b/src/conf/node_device_conf.h @@ -101,6 +101,7 @@ struct _virNodeDevCapsDef { unsigned function; unsigned product; unsigned vendor; + unsigned class; char *product_name; char *vendor_name; } pci_dev; @@ -117,10 +118,12 @@ struct _virNodeDevCapsDef { unsigned _class; /* "class" is reserved in C */ unsigned subclass; unsigned protocol; + char *interface_name; char *description; } usb_if; struct { char *address; + unsigned address_len; char *ifname; enum virNodeDevNetCapType subtype; /* LAST -> no subtype */ } net; @@ -139,6 +142,8 @@ struct _virNodeDevCapsDef { } scsi; struct { unsigned long long size; + unsigned long long num_blocks; + unsigned long long logical_block_size; unsigned long long removable_media_size; char *block; char *bus; diff --git a/src/node_device/node_device_driver.c b/src/node_device/node_device_driver.c index 14b3098..f3bd45d 100644 --- a/src/node_device/node_device_driver.c +++ b/src/node_device/node_device_driver.c @@ -70,9 +70,12 @@ static int update_caps(virNodeDeviceObjPtr dev) } -#ifdef __linux__ -static int update_driver_name(virConnectPtr conn, - virNodeDeviceObjPtr dev) +#if defined (__linux__) && defined (HAVE_HAL) +/* XXX Why does this function exist? Are there devices that change + * their drivers while running? Under libudev, most devices seem to + * provide their driver name as a property "DRIVER" */ +static int update_driver_name_hal_linux(virConnectPtr conn, + virNodeDeviceObjPtr dev) { char *driver_link = NULL; char devpath[PATH_MAX]; @@ -754,5 +757,8 @@ int nodedevRegister(void) { #ifdef HAVE_DEVKIT return devkitNodeRegister(); #endif +#ifdef HAVE_UDEV + return udevNodeRegister(); +#endif #endif } diff --git a/src/node_device/node_device_driver.h b/src/node_device/node_device_driver.h index db01624..5be0781 100644 --- a/src/node_device/node_device_driver.h +++ b/src/node_device/node_device_driver.h @@ -45,6 +45,9 @@ int halNodeRegister(void); #ifdef HAVE_DEVKIT int devkitNodeRegister(void); #endif +#ifdef HAVE_UDEV +int udevNodeRegister(void); +#endif void nodeDeviceLock(virDeviceMonitorStatePtr driver); void nodeDeviceUnlock(virDeviceMonitorStatePtr driver); @@ -53,4 +56,23 @@ void registerCommonNodeFuncs(virDeviceMonitorPtr mon); int nodedevRegister(void); +#ifdef __linux__ + +#define check_fc_host(d) check_fc_host_linux(d) +int check_fc_host_linux(union _virNodeDevCapData *d); + +#define check_vport_capable(d) check_vport_capable_linux(d) +int check_vport_capable_linux(union _virNodeDevCapData *d); + +#define read_wwn(host, file, wwn) read_wwn_linux(host, file, wwn) +int read_wwn_linux(int host, const char *file, char **wwn); + +#else /* __linux__ */ + +#define check_fc_host(d) +#define check_vport_capable(d) +#define read_wwn(host, file, wwn) + +#endif /* __linux__ */ + #endif /* __VIR_NODE_DEVICE_H__ */ diff --git a/src/node_device/node_device_hal.h b/src/node_device/node_device_hal.h index c859fe3..8ac8a35 100644 --- a/src/node_device/node_device_hal.h +++ b/src/node_device/node_device_hal.h @@ -22,23 +22,4 @@ #ifndef __VIR_NODE_DEVICE_HAL_H__ #define __VIR_NODE_DEVICE_HAL_H__ -#ifdef __linux__ - -#define check_fc_host(d) check_fc_host_linux(d) -int check_fc_host_linux(union _virNodeDevCapData *d); - -#define check_vport_capable(d) check_vport_capable_linux(d) -int check_vport_capable_linux(union _virNodeDevCapData *d); - -#define read_wwn(host, file, wwn) read_wwn_linux(host, file, wwn) -int read_wwn_linux(int host, const char *file, char **wwn); - -#else /* __linux__ */ - -#define check_fc_host(d) -#define check_vport_capable(d) -#define read_wwn(host, file, wwn) - -#endif /* __linux__ */ - #endif /* __VIR_NODE_DEVICE_HAL_H__ */ diff --git a/src/node_device/node_device_hal_linux.c b/src/node_device/node_device_linux_sysfs.c similarity index 100% rename from src/node_device/node_device_hal_linux.c rename to src/node_device/node_device_linux_sysfs.c diff --git a/src/node_device/node_device_udev.c b/src/node_device/node_device_udev.c new file mode 100644 index 0000000..32fde52 --- /dev/null +++ b/src/node_device/node_device_udev.c @@ -0,0 +1,800 @@ +/* + * node_device_udev.c: node device enumeration - libudev implementation + * + * Copyright (C) 2009 Red Hat + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * + * Author: Dave Allan <dallan(a)redhat.com> + */ + +#include <config.h> +#include <libudev.h> + +#include "node_device_udev.h" +#include "virterror_internal.h" +#include "node_device_conf.h" +#include "node_device_driver.h" +#include "driver.h" +#include "datatypes.h" +#include "logging.h" +#include "memory.h" +#include "util.h" +#include "buf.h" + +#define VIR_FROM_THIS VIR_FROM_NODEDEV + +static virDeviceMonitorStatePtr driverState = NULL; + +/* This function allocates memory from the heap for the property + * value. That memory must be later freed by some other code. */ +static int udevGetDeviceProperty(struct udev_device *udev_device, + const char *property_key, + char **property_value) +{ + const char *udev_value = NULL; + int ret = -1; + + udev_value = udev_device_get_property_value(udev_device, property_key); + if (udev_value == NULL) { + VIR_ERROR(_("udev reports property '%s' does not exist"), + property_key); + goto out; + } + + /* If this allocation is changed, the comment at the beginning + * of the function must also be changed. */ + *property_value = strdup(udev_value); + if (*property_value == NULL) { + VIR_ERROR(_("Failed to allocate memory for " + "property '%s' on device '%s'"), + property_key, udev_device_get_sysname(udev_device)); + goto out; + } + + ret = 0; + +out: + if (ret != 0) { + VIR_ERROR(_("Failed to get udev property '%s' for device '%s'"), + property_key, udev_device_get_sysname(udev_device)); + } + + return ret; +} + + +static int udevGetStringProperty(struct udev_device *udev_device, + const char *property_key, + char **value) +{ + return udevGetDeviceProperty(udev_device, property_key, value); +} + + +static int udevGetIntProperty(struct udev_device *udev_device, + const char *property_key, + int *value, + int base) +{ + char *udev_value = NULL; + int ret = 0; + + ret = udevGetDeviceProperty(udev_device, property_key, &udev_value); + + if (ret == 0) { + ret = virStrToLong_i(udev_value, NULL, base, value); + } + + VIR_FREE(udev_value); + return ret; +} + + +static int udevGetUintProperty(struct udev_device *udev_device, + const char *property_key, + unsigned int *value, + int base) +{ + char *udev_value = NULL; + int ret = 0; + + ret = udevGetDeviceProperty(udev_device, property_key, &udev_value); + + if (ret == 0) { + ret = virStrToLong_ui(udev_value, NULL, base, value); + } + + VIR_FREE(udev_value); + return ret; +} + + +static int udevGetStringSysfsAttr(struct udev_device *udev_device, + const char *attr_name, + char **value) +{ + const char *udev_value = NULL; + int ret = -1; + + udev_value = udev_device_get_sysattr_value(udev_device, attr_name); + + if (udev_value != NULL) { + *value = strdup(udev_value); + if (*value != NULL) { + ret = 0; + } + } + + return ret; +} + + +static int udevGetIntSysfsAttr(struct udev_device *udev_device, + const char *attr_name, + int *value, + int base) +{ + const char *udev_value = NULL; + int ret = -1; + + udev_value = udev_device_get_sysattr_value(udev_device, attr_name); + + if (udev_value != NULL) { + ret = virStrToLong_i(udev_value, NULL, base, value); + } + + return ret; +} + + +static int udevGetUintSysfsAttr(struct udev_device *udev_device, + const char *attr_name, + unsigned int *value, + int base) +{ + const char *udev_value = NULL; + int ret = -1; + + udev_value = udev_device_get_sysattr_value(udev_device, attr_name); + + if (udev_value != NULL) { + ret = virStrToLong_ui(udev_value, NULL, base, value); + } + + return ret; +} + + +static int udevGetUint64SysfsAttr(struct udev_device *udev_device, + const char *attr_name, + uint64_t *value) +{ + const char *udev_value = NULL; + int ret = -1; + + udev_value = udev_device_get_sysattr_value(udev_device, attr_name); + + if (udev_value != NULL) { + ret = virStrToLong_ull(udev_value, NULL, 0, value); + } + + return ret; +} + + +static void udevLogFunction(struct udev *udev ATTRIBUTE_UNUSED, + int priority ATTRIBUTE_UNUSED, + const char *file, + int line, + const char *fn, + const char *fmt, + va_list args) +{ + VIR_ERROR_INT(file, fn, line, fmt, args); +} + + +static int udevProcessPCI(struct udev_device *device, + virNodeDeviceDefPtr def) +{ + const char *devpath = NULL; + int ret = -1; + + devpath = udev_device_get_devpath(device); + + char *p = strrchr(devpath, '/'); + if (p) { + virStrToLong_ui(p+1, &p, 16, &def->caps->data.pci_dev.domain); + virStrToLong_ui(p+1, &p, 16, &def->caps->data.pci_dev.bus); + virStrToLong_ui(p+1, &p, 16, &def->caps->data.pci_dev.slot); + virStrToLong_ui(p+1, &p, 16, &def->caps->data.pci_dev.function); + } + + udevGetUintSysfsAttr(device, "vendor", &def->caps->data.pci_dev.vendor, 0); + udevGetUintSysfsAttr(device, "product", &def->caps->data.pci_dev.product, 0); + + /* FIXME: to do the vendor name and product name, we have to + * parse /usr/share/hwdata/pci.ids. Look in hal/hald/ids.c + */ + + ret = 0; + + return ret; +} + + +static int udevProcessUSBDevice(struct udev_device *device, + virNodeDeviceDefPtr def) +{ + int ret = 0; + udevGetUintProperty(device, "BUSNUM", &def->caps->data.usb_dev.bus, 0); + udevGetUintProperty(device, "DEVNUM", &def->caps->data.usb_dev.device, 0); + udevGetUintProperty(device, "ID_VENDOR_ID", &def->caps->data.usb_dev.vendor, 16); + udevGetStringProperty(device, "ID_VENDOR", &def->caps->data.usb_dev.vendor_name); + udevGetUintProperty(device, "ID_MODEL_ID", &def->caps->data.usb_dev.product, 0); + udevGetStringProperty(device, "ID_MODEL", &def->caps->data.usb_dev.product_name); + + return ret; +} + + +/* XXX I don't fully understand the properties here. Advice needed. */ +static int udevProcessUSBInterface(struct udev_device *device, + virNodeDeviceDefPtr def) +{ + int ret = 0; + char *tmp = NULL, *p = NULL; + + udevGetStringProperty(device, + "INTERFACE", + &def->caps->data.usb_if.interface_name); + + tmp = strdup(def->caps->data.usb_if.interface_name); + + p = strrchr(tmp, '/'); + if (p) { + *p = '\0'; + virStrToLong_ui(tmp, NULL, 0, &def->caps->data.usb_if.number); + virStrToLong_ui(p+1, &p, 0, &def->caps->data.usb_if._class); + virStrToLong_ui(p+1, &p, 0, &def->caps->data.usb_if.subclass); + } + + VIR_FREE(tmp); + return ret; +} + + +static int udevProcessNetworkInterface(struct udev_device *device, + virNodeDeviceDefPtr def) +{ + int ret = 0; + + udevGetStringSysfsAttr(device, "address", &def->caps->data.net.address); + udevGetUintSysfsAttr(device, "addr_len", &def->caps->data.net.address_len, 0); + + return ret; +} + + +static int udevProcessSCSIHost(struct udev_device *device ATTRIBUTE_UNUSED, + virNodeDeviceDefPtr def) +{ + int ret = 0; + char *filename = NULL; + + filename = basename(def->name); + + if (STRPREFIX(filename, "host")) { + /* XXX There's really no better way to get the host #? */ + ret = virStrToLong_ui(filename + strlen("host"), + NULL, + 0, + &def->caps->data.scsi_host.host); + check_fc_host(&def->caps->data); + check_vport_capable(&def->caps->data); + } + + return ret; +} + + +static int udevProcessSCSIDevice(struct udev_device *device ATTRIBUTE_UNUSED, + virNodeDeviceDefPtr def) +{ + int ret = 0; + char *filename = NULL, *p = NULL; + + filename = basename(def->name); + + virStrToLong_ui(filename, &p, 10, &def->caps->data.scsi.host); + virStrToLong_ui(p+1, &p, 10, &def->caps->data.scsi.bus); + virStrToLong_ui(p+1, &p, 10, &def->caps->data.scsi.target); + virStrToLong_ui(p+1, &p, 10, &def->caps->data.scsi.lun); + + return ret; +} + + +static int udevProcessDisk(struct udev_device *device, + virNodeDeviceDefPtr def) +{ + int ret = 0; + + def->caps->data.storage.drive_type = strdup("disk"); + if (def->caps->data.storage.drive_type == NULL) { + ret = -1; + goto out; + } + + udevGetUint64SysfsAttr(device, "size", &def->caps->data.storage.num_blocks); + udevGetUint64SysfsAttr(device, "queue/logical_block_size", + &def->caps->data.storage.logical_block_size); + + def->caps->data.storage.size = def->caps->data.storage.num_blocks * + def->caps->data.storage.logical_block_size; + +out: + return ret; +} + + +static int udevProcessCDROM(struct udev_device *device, + virNodeDeviceDefPtr def) +{ + int tmp_int = 0, ret = 0; + + if ((udevGetIntSysfsAttr(device, "removable", &tmp_int, 0) == 0) && + (tmp_int == 1)) { + + def->caps->data.storage.flags |= VIR_NODE_DEV_CAP_STORAGE_REMOVABLE; + + if ((udevGetIntProperty(device, "ID_CDROM_MEDIA", &tmp_int, 0) == 0) && + (tmp_int == 1)) { + + def->caps->data.storage.flags |= + VIR_NODE_DEV_CAP_STORAGE_REMOVABLE_MEDIA_AVAILABLE; + + udevGetUint64SysfsAttr(device, "size", &def->caps->data.storage.num_blocks); + udevGetUint64SysfsAttr(device, "queue/logical_block_size", + &def->caps->data.storage.logical_block_size); + + /* XXX This calculation is wrong for the qemu virtual + * cdrom which reports the size in 512 byte blocks, but + * the logical block size as 2048. I don't have a + * physical cdrom on a devel system to see how they + * behave. */ + def->caps->data.storage.removable_media_size = + def->caps->data.storage.num_blocks * + def->caps->data.storage.logical_block_size; + } + } + + return ret; +} + + +/* This function exists to deal with the case in which a driver does + * not provide a device type in the usual place, but udev told us it's + * a storage device, and we can make a good guess at what kind of + * storage device it is from other information that is provided. */ +static int udevKludgeStorageType(virNodeDeviceDefPtr def) +{ + int ret = -1; + + if (STRPREFIX(def->caps->data.storage.block, "/dev/vd")) { + /* virtio disk */ + def->caps->data.storage.drive_type = strdup("disk"); + if (def->caps->data.storage.drive_type != NULL) { + ret = 0; + } + } + + return ret; +} + + +static void udevStripSpaces(char *s) +{ + if (s == NULL) { + return; + } + + while (virFileStripSuffix(s, " ")) { + /* do nothing */ + ; + } + + return; +} + + +static int udevProcessStorage(struct udev_device *device, + virNodeDeviceDefPtr def) +{ + int ret = -1; + + def->caps->data.storage.block = strdup(udev_device_get_devnode(device)); + udevGetStringProperty(device, "DEVNAME", &def->caps->data.storage.block); + udevGetStringProperty(device, "ID_BUS", &def->caps->data.storage.bus); + udevGetStringProperty(device, "ID_SERIAL", &def->caps->data.storage.serial); + udevGetStringSysfsAttr(device, "device/vendor", &def->caps->data.storage.vendor); + udevStripSpaces(def->caps->data.storage.vendor); + udevGetStringSysfsAttr(device, "device/model", &def->caps->data.storage.model); + udevStripSpaces(def->caps->data.storage.model); + /* XXX Is there an equivalent of the HAL hotpluggable property? + * Is the hotpluggable property still relevant? */ + + if (udevGetStringProperty(device, + "ID_TYPE", + &def->caps->data.storage.drive_type) != 0) { + /* If udev doesn't have it, perhaps we can guess it. */ + if (udevKludgeStorageType(def) != 0) { + goto out; + } + } + + /* NB: drive_type has changed from HAL; now it's "cd" instead of "cdrom" */ + if (STREQ(def->caps->data.storage.drive_type, "cd")) { + ret = udevProcessCDROM(device, def); + } else if (STREQ(def->caps->data.storage.drive_type, "disk")) { + ret = udevProcessDisk(device, def); + } else { + goto out; + } + +out: + return ret; +} + + +static int udevSetDeviceType(struct udev_device *device, + virNodeDeviceDefPtr def) +{ + const char *devtype = NULL; + int ret = 0; + + if (VIR_ALLOC(def->caps) != 0) { + ret = -1; + goto out; + } + + devtype = udev_device_get_devtype(device); + + if (udevGetUintProperty(device, + "PCI_CLASS", + &def->caps->data.pci_dev.class, + 0) == 0) { + def->caps->type = VIR_NODE_DEV_CAP_PCI_DEV; + goto out; + } + + if (devtype != NULL && STREQ(devtype, "usb_device")) { + def->caps->type = VIR_NODE_DEV_CAP_USB_DEV; + goto out; + } + + if (devtype != NULL && STREQ(devtype, "usb_interface")) { + def->caps->type = VIR_NODE_DEV_CAP_USB_INTERFACE; + goto out; + } + + /* XXX Do some network interfaces set the device type property? */ + if (udevGetStringProperty(device, + "INTERFACE", + &def->caps->data.net.ifname) == 0) { + + def->caps->type = VIR_NODE_DEV_CAP_NET; + goto out; + } + + if (devtype != NULL && STREQ(devtype, "scsi_host")) { + def->caps->type = VIR_NODE_DEV_CAP_SCSI_HOST; + goto out; + } + + if (devtype != NULL && STREQ(devtype, "scsi_device")) { + def->caps->type = VIR_NODE_DEV_CAP_SCSI; + goto out; + } + + if (devtype != NULL && STREQ(devtype, "disk")) { + def->caps->type = VIR_NODE_DEV_CAP_STORAGE; + goto out; + } + + ret = -1; + +out: + return ret; +} + + +static int udevGetDeviceDetails(struct udev_device *device, + virNodeDeviceDefPtr def) +{ + int ret = 0; + + switch (def->caps->type) { + case VIR_NODE_DEV_CAP_SYSTEM: + /* What do we want to do with system--there's no libudev equivalent. */ + break; + case VIR_NODE_DEV_CAP_PCI_DEV: + ret = udevProcessPCI(device, def); + break; + case VIR_NODE_DEV_CAP_USB_DEV: + ret = udevProcessUSBDevice(device, def); + break; + case VIR_NODE_DEV_CAP_USB_INTERFACE: + ret = udevProcessUSBInterface(device, def); + break; + case VIR_NODE_DEV_CAP_NET: + ret = udevProcessNetworkInterface(device, def); + break; + case VIR_NODE_DEV_CAP_SCSI_HOST: + ret = udevProcessSCSIHost(device, def); + break; + case VIR_NODE_DEV_CAP_SCSI: + ret = udevProcessSCSIDevice(device, def); + break; + case VIR_NODE_DEV_CAP_STORAGE: + ret = udevProcessStorage(device, def); + break; + default: + ret = -1; + break; + } + + return ret; +} + + +static int udevProcessOneDevice(struct udev_device *device, + virNodeDeviceDefPtr def) +{ + struct udev_device *parent = NULL; + virNodeDeviceObjPtr dev = NULL; + const char *parent_name = NULL; + int ret = -1; + + def->name = strdup(udev_device_get_syspath(device)); + udevGetStringProperty(device, "DRIVER", &def->driver); + + /* How to handle parent devices? Shall we change the meaning of + * the parent pointer for the udev backend? What does client code + * use the parent relationship for? We really want to preserve + * the behavior of the HAL backend, but that may be difficult in + * practice. */ + parent = udev_device_get_parent(device); + if (parent != NULL) { + parent_name = udev_device_get_syspath(parent); + if (parent_name != NULL) { + def->parent = strdup(parent_name); + } + } + + if (udevSetDeviceType(device, def) == 0) { + ret = udevGetDeviceDetails(device, def); + } + + if (ret == 0) { + dev = virNodeDeviceAssignDef(NULL, &driverState->devs, def); + dev->devicePath = strdup(udev_device_get_devpath(device)); + virNodeDeviceObjUnlock(dev); + } else { + virNodeDeviceDefFree(def); + } + + return ret; +} + + +static int udevProcessDeviceListEntry(struct udev *udev, + struct udev_list_entry *list_entry) +{ + struct udev_device *device; + virNodeDeviceDefPtr def = NULL; + const char *name = NULL; + int ret = -1; + + name = udev_list_entry_get_name(list_entry); + + if (VIR_ALLOC(def) != 0) { + goto out; + } + + device = udev_device_new_from_syspath(udev, name); + if (device != NULL) { + udevProcessOneDevice(device, def); + udev_device_unref(device); + ret = 0; + } + +out: + return ret; +} + + +static int udevEnumerateDevices(struct udev *udev) +{ + struct udev_enumerate *udev_enumerate = NULL; + struct udev_list_entry *list_entry = NULL; + const char *name = NULL; + int ret = 0; + + udev_enumerate = udev_enumerate_new(udev); + + ret = udev_enumerate_scan_devices(udev_enumerate); + if (0 != ret) { + VIR_ERROR("udev scan devices returned %d\n", ret); + goto out; + } + + udev_list_entry_foreach(list_entry, + udev_enumerate_get_list_entry(udev_enumerate)) { + + udevProcessDeviceListEntry(udev, list_entry); + name = udev_list_entry_get_name(list_entry); + } + +out: + udev_enumerate_unref(udev_enumerate); + return ret; +} + + +static int udevDeviceMonitorShutdown(void) +{ + int ret = 0; + + struct udev_monitor *udev_monitor = NULL; + struct udev *udev = NULL; + + if (driverState) { + + nodeDeviceLock(driverState); + udev_monitor = DRV_STATE_UDEV_MONITOR(driverState); + + if (udev_monitor != NULL) { + udev = udev_monitor_get_udev(udev_monitor); + udev_monitor_unref(udev_monitor); + } + + if (udev != NULL) { + udev_unref(udev); + } + + virNodeDeviceObjListFree(&driverState->devs); + nodeDeviceUnlock(driverState); + virMutexDestroy(&driverState->lock); + VIR_FREE(driverState); + + } else { + ret = -1; + } + + return ret; +} + + +static int udevDeviceMonitorStartup(int privileged ATTRIBUTE_UNUSED) +{ + struct udev *udev = NULL; + struct udev_monitor *udev_monitor = NULL; + int ret = 0; + + if (VIR_ALLOC(driverState) < 0) { + ret = -1; + goto out; + } + + if (virMutexInit(&driverState->lock) < 0) { + VIR_FREE(driverState); + ret = -1; + goto out; + } + + nodeDeviceLock(driverState); + + /* + * http://www.kernel.org/pub/linux/utils/kernel/hotplug/libudev/libudev-udev.h… + * + * indicates no return value other than success, so we don't check + * its return value. + */ + udev = udev_new(); + udev_set_log_fn(udev, udevLogFunction); + + udev_monitor = udev_monitor_new_from_netlink(udev, "udev"); + if (udev_monitor == NULL) { + ret = -1; + goto out; + } + + /* udev can be retrieved from udev_monitor */ + driverState->privateData = udev_monitor; + nodeDeviceUnlock(driverState); + + /* Populate with known devices */ + + if (udevEnumerateDevices(udev) != 0) { + ret = -1; + goto out; + } + +out: + if (ret == -1) { + udevDeviceMonitorShutdown(); + } + return ret; +} + + +static int udevDeviceMonitorReload(void) +{ + return 0; +} + + +static int udevDeviceMonitorActive(void) +{ + /* Always ready to deal with a shutdown */ + return 0; +} + + +static virDrvOpenStatus udevNodeDrvOpen(virConnectPtr conn, + virConnectAuthPtr auth ATTRIBUTE_UNUSED, + int flags ATTRIBUTE_UNUSED) +{ + if (driverState == NULL) { + return VIR_DRV_OPEN_DECLINED; + } + + conn->devMonPrivateData = driverState; + + return VIR_DRV_OPEN_SUCCESS; +} + +static int udevNodeDrvClose(virConnectPtr conn) +{ + conn->devMonPrivateData = NULL; + return 0; +} + +static virDeviceMonitor udevDeviceMonitor = { + .name = "udevDeviceMonitor", + .open = udevNodeDrvOpen, + .close = udevNodeDrvClose, +}; + +static virStateDriver udevStateDriver = { + .initialize = udevDeviceMonitorStartup, + .cleanup = udevDeviceMonitorShutdown, + .reload = udevDeviceMonitorReload, + .active = udevDeviceMonitorActive, +}; + +int udevNodeRegister(void) +{ + VIR_ERROR0("Registering udev node device backend\n"); + + registerCommonNodeFuncs(&udevDeviceMonitor); + if (virRegisterDeviceMonitor(&udevDeviceMonitor) < 0) { + return -1; + } + + return virRegisterStateDriver(&udevStateDriver); +} diff --git a/src/node_device/node_device_udev.h b/src/node_device/node_device_udev.h new file mode 100644 index 0000000..bac628d --- /dev/null +++ b/src/node_device/node_device_udev.h @@ -0,0 +1,27 @@ +/* + * node_device_udev.h: node device enumeration - libudev implementation + * + * Copyright (C) 2009 Red Hat + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * + * Author: Dave Allan <dallan(a)redhat.com> + */ + +#include <libudev.h> +#include <stdint.h> + +#define SYSFS_DATA_SIZE 4096 +#define DRV_STATE_UDEV_MONITOR(ds) ((struct udev_monitor *)((ds)->privateData)) diff --git a/src/util/util.c b/src/util/util.c index 98f8a14..f63bd39 100644 --- a/src/util/util.c +++ b/src/util/util.c @@ -2114,3 +2114,31 @@ void virFileWaitForDevices(virConnectPtr conn) void virFileWaitForDevices(virConnectPtr conn ATTRIBUTE_UNUSED) {} #endif #endif + +int virBuildPathInternal(char **path, ...) +{ + char *path_component = NULL; + virBuffer buf = VIR_BUFFER_INITIALIZER; + va_list ap; + int ret = 0; + + va_start(ap, *path); + + path_component = va_arg(ap, char *); + virBufferAdd(&buf, path_component, -1); + + while ((path_component = va_arg(ap, char *)) != NULL) + { + virBufferAddChar(&buf, '/'); + virBufferAdd(&buf, path_component, -1); + } + + va_end(ap); + + *path = virBufferContentAndReset(&buf); + if (*path == NULL) { + ret = -1; + } + + return ret; +} diff --git a/src/util/util.h b/src/util/util.h index 8679636..da410e1 100644 --- a/src/util/util.h +++ b/src/util/util.h @@ -248,4 +248,7 @@ char *virFileFindMountPoint(const char *type); void virFileWaitForDevices(virConnectPtr conn); +#define virBuildPath(path, ...) virBuildPathInternal(path, __VA_ARGS__, NULL) +int virBuildPathInternal(char **path, ...) __attribute__ ((sentinel)); + #endif /* __VIR_UTIL_H__ */ -- 1.6.4.4

3 3

[libvirt] [RFC PATCH 0/3] Add MAC filtering to qemu
by Gerhard Stenzel 16 Oct '09

16 Oct '09

The following series of patches is a prototype implementation of a ebtables based MAC address filter. I hope to have addressed all the previous comments and I am still interested to get more feedback/comments. At the moment, some defaults are set when libvirtd starts and when a domain is started or destroyed. The plan is to extend this filter capability to the API level and virsh command level. The plan is also to add more filtering features like VLAN filtering and QoS filtering. Thanks, Gerhard --- Gerhard Stenzel (3): add MAC address based port filtering to qemu add build support for ebtables wrapper add ebtables wrapper configure.in | 3 src/Makefile.am | 1 src/libvirt_private.syms | 27 + src/qemu/qemu.conf | 3 src/qemu/qemu_conf.c | 85 ++++ src/qemu/qemu_conf.h | 8 src/qemu/qemu_driver.c | 44 ++ src/util/ebtables.c | 982 ++++++++++++++++++++++++++++++++++++++++++++++ src/util/ebtables.h | 134 ++++++ 9 files changed, 1287 insertions(+), 0 deletions(-) create mode 100644 src/util/ebtables.c create mode 100644 src/util/ebtables.h --

2 6

[libvirt] [PATCH 0/5] Misc changes to domain driver stuff
by Daniel P. Berrange 16 Oct '09

16 Oct '09

The patches that follow make some changes to the domain driver infrastructure, and the QEMU driver. The important goal is to improve the concurrency of the QEMU driver, by allowing multiple readers on the driver (change mutex to a rwlock) and removing the O(n) locks on the virDomainObjPtr instances These have not been heavily tested yet, beyond checking the TCK passes with QEMU driver. I just send them now for early review, while I continue to work on the rest of the QEMU monitor related patches which this series supports. The CIL object locking test also needs to be extended to understand the idea of RWlocks instead of just mutexes so it can properly validate this code Daniel

3 14

[libvirt] [PATCH] LXC complement PATH environment variable of a controller
by Ryota Ozaki 16 Oct '09

16 Oct '09

Hi, lxc controller usually executes 'ip' command to control veths, however now PATH environment variable of the controller is not set (logging code resets environment variables all) and the execution will definitely fail because the controller cannot find the executable. This patch complements PATH with driver's one. The code is just copied from qemu_conf.c Thanks, ozaki-r >From c98136628c1ea0964191be976b23ac99d0f7bb7b Mon Sep 17 00:00:00 2001 From: Ryota Ozaki <ozaki.ryota(a)gmail.com> Date: Wed, 14 Oct 2009 05:09:04 +0900 Subject: [PATCH] LXC complement PATH environment variable of a controller * src/lxc/lxc_driver.c: without PATH, the controller will definitely fail to call ip command --- src/lxc/lxc_driver.c | 13 +++++++++++++ 1 files changed, 13 insertions(+), 0 deletions(-) diff --git a/src/lxc/lxc_driver.c b/src/lxc/lxc_driver.c index 0b614e3..783dfcc 100644 --- a/src/lxc/lxc_driver.c +++ b/src/lxc/lxc_driver.c @@ -1005,6 +1005,19 @@ static int lxcControllerStart(virConnectPtr conn, lenv[lenvc++] = envval; ¥ } while (0) +#define ADD_ENV_COPY(envname) ¥ + do { ¥ + char *val = getenv(envname); ¥ + if (val != NULL) { ¥ + ADD_ENV_PAIR(envname, val); ¥ + } ¥ + } while (0) + + /* + * The controller may call ip command, so we have to remain PATH. + */ + ADD_ENV_COPY("PATH"); + log_level = virLogGetDefaultPriority(); if (virAsprintf(&tmp, "LIBVIRT_DEBUG=%d", log_level) < 0) goto no_memory; -- 1.6.2.5

4 3

[libvirt] [RFC] Support for CPUID masking
by Jiri Denemark 16 Oct '09

16 Oct '09

Hi, We need to provide support for CPU ID masking. Xen and VMware ESX are examples of current hypervisors which support such masking. My proposal is to define new 'cpuid' feature advertised in guest capabilities: <guest> ... <features> <cpuid/> </feature> </guest> When a driver supports cpuid feature, one can use it to mask/check for specific bits returned by CPU ID as follows: <domain> ... <features> <cpuid> <mask level='hex' register='eax|ebx|ecx|edx'>MASK</mask> </cpuid> </features> ... </domain> Where - level is a hexadecimal number used as an input to CPU ID (i.e. eax register), - register is one of eax, ebx, ecx or edx, - and MASK is a string with the following format: mmmm:mmmm:mmmm:mmmm:mmmm:mmmm:mmmm:mmmm with m being 1-bit mask for the corresponding bit in the register. There are three possibilities of specifying what values can be used for 'm': - let it be driver-specific, - define all possible values, - define a common set of values and allow drivers to specify their own additional values. I think the third is the way to go as it lowers the confusion of different values used by different drivers for the same purpose while maintaining the flexibility to support driver-specific masks. The following could be a good set of predefined common values: - 1 force the bit to be 1 - 0 force the bit to be 0 - x don't care, i.e., use driver's default value - T require the bit to be 1 - F require the bit to be 0 Some examples of what it could look like follow: <capabilities> ... <guest> <os_type>xen</os_type> ... <features> <pae/> <cpuid/> </features> </guest> ... </capabilities> <domain type='xen' id='42'> ... <features> <pae/> <acpi/> <apic/> <cpuid> <mask level='1' register='ebx'> xxxx:xxxx:0000:1010:xxxx:xxxx:xxxx:xxxx </mask> <mask level='1' register='ecx'> xxxx:xxxx:xxxx:xxxx:xxxx:xxxx:xx1x:xxxx </mask> <mask level='1' register='edx'> xxx1:xxxx:xxxx:xxxx:xxxx:xxxx:xxxx:xxxx </mask> <mask level='80000001' register='ecx'> xxxx:xxxx:xxxx:xxxx:xxxx:xxxx:xxxx:xx1x </mask> <mask level='80000008' register='ecx'> xxxx:xxxx:xxxx:xxxx:xxxx:xxxx:xx00:1001 </mask> </cpuid> </features> ... </domain> What are your opinions about this? Thank for all comments. Jirka

8 12

[libvirt] [PATCH] Avoid crash in virBufferEscapeString
by Laine Stump 15 Oct '09

15 Oct '09

If virBufferEscapeString is called on a buffer that has 0 bytes of space, a size of -1 will be passed to snprintf, resulting in a segmentation fault. This patch checks for 0 space, and grows the buffer if needed prior to determining size. I discovered this when I accidentally made virBufferEscapeString the first function to add something to a newly minted buffer. --- src/util/buf.c | 6 ++++++ 1 files changed, 6 insertions(+), 0 deletions(-) diff --git a/src/util/buf.c b/src/util/buf.c index c802aa2..9681635 100644 --- a/src/util/buf.c +++ b/src/util/buf.c @@ -318,6 +318,12 @@ virBufferEscapeString(const virBufferPtr buf, const char *format, const char *st } *out = 0; + if ((buf->use >= buf->size) && + virBufferGrow(buf, 100) < 0) { + VIR_FREE(escaped); + return; + } + size = buf->size - buf->use - 1; while (((count = snprintf(&buf->content[buf->use], size, format, (char *)escaped)) < 0) || (count >= size - 1)) { -- 1.6.2.5

3 3

[libvirt] [PATCH] Don't copy old machines from a domain which has none
by Mark McLoughlin 15 Oct '09

15 Oct '09

If the the qemu and kvm binaries are the same, we don't include machine types in the kvm domain info. However, the code which refreshes the machine types info from the previous capabilities structure first looks at the kvm domain's info, finds it matches and then copies the empty machine types list over for the top-level qemu domain. That doesn't make sense, we shouldn't copy an empty machin types list. * src/qemu/qemu_conf.c: qemudGetOldMachinesFromInfo(): don't copy an empty machine types list. --- src/qemu/qemu_conf.c | 3 +++ 1 files changed, 3 insertions(+), 0 deletions(-) diff --git a/src/qemu/qemu_conf.c b/src/qemu/qemu_conf.c index ac63570..b881f1e 100644 --- a/src/qemu/qemu_conf.c +++ b/src/qemu/qemu_conf.c @@ -505,6 +505,9 @@ qemudGetOldMachinesFromInfo(virCapsGuestDomainInfoPtr info, virCapsGuestMachinePtr *list; int i; + if (!info->nmachines) + return 0; + if (!info->emulator || !STREQ(emulator, info->emulator)) return 0; -- 1.6.2.5

2 1

[libvirt] [PATCH] Add ocfs2 to list of fs pool types
by Jim Fehlig 15 Oct '09

15 Oct '09

Ensures mount.ocfs2 is called when <format type="ocfs2"/> is specified in storage pool XML. Regards, Jim

3 3

[libvirt] Re: Libvir Xen and domain events
by Christopher Pyper 14 Oct '09

14 Oct '09

I have been looking into registering some domain events though the python Libvir API. However, I am getting errors and I believe it may be due to lack of event support in Xen. This is the error I am getting: libvir: Xen error : this function is not supported by the hypervisor: xenUnifiedDomainEventRegister Traceback (most recent call last): File "./testcallback.py", line 19, in ? conn.domainEventRegister(callback,None) File "/usr/lib64/python2.4/site-packages/libvirt.py", line 1578, in domainEventRegister if ret == -1: raise libvirtError ('virConnectDomainEventRegister () failed', conn=self) libvirt.libvirtError: this function is not supported by the hypervisor: xenUnifiedDomainEventRegister If this lack of support is the case, what would you recommend is the best route for catching domain events? Currently, I am thinking of just polling regularly and monitoring for changes in domain state. Cheers. Chris Pyper. P.S. Libvir is awesome.

2 2

[libvirt] Release of libvirt-0.7.2
by Daniel Veillard 14 Oct '09

14 Oct '09

This is an intermediate release to flush out one month worth of development and patches, as we have another release scheduled for the 30th October. It is available on ftp://libvirt.org/libvirt as usual and tagged in GIT The main new features are AppArmor security driver and the possibility to define new domain in the ESX driver. The code had a lot of cleanup, including a big tree refactoring and a fairly long list of bug fixes: Features: - sVirt AppArmor security driver (Jamie Strandboge) - Add public API definition for data stream handling (Daniel P. Berrange) - ESX add esxDomainDefineXML() (Matthias Bolte) - LXC: suspend/resume support (Ryota Ozaki) - Big code tree cleanup (Daniel P. Berrange) Documentation: - Documentation and examples for SVirt Apparmor driver (Jamie Strandboge) - Fix documentation and comment typos (Paolo Bonzini) - Fix up a few typos in the tree. (Chris Lalancette) - Fix a typo in virNetHasValidPciAddr() too (Mark McLoughlin) - Fix a typo in virDiskHasValidPciAddr() (Jiri Denemark) - Fix a number of small typos (Dan Kenigsberg) - add doc for graphic and video elements (Florian Vichot) - Fix up 'neccessary -> necessary' in a comment. (Chris Lalancette) - Fix up comments for domainXML{To,From}Native. (Chris Lalancette) - Simple fix of a comment in qemuStringToArgvEnv. (Chris Lalancette) - Add a README file to src/ explaining the directory structure (Daniel P. Berrange) - don't emit trailing blanks into generated and VC'd NEWS file (Jim Meyering) Portability: - Misc win32 build fixes (Daniel P. Berrange) - Don't require full daemon install for libvirt python bindings (Daniel P. Berrange) - Tweak specfile to fix RHEL6 rules & ESX/PHYP enablement (Daniel P. Berrange) - Bug Fixes: - network: Fix printing XML 'delay' attribute (Cole Robinson) - Fix virFileReadLimFD/virFileReadAll to handle EINTR (Daniel P. Berrange) - storage: Fix generating iscsi 'auth' xml (Cole Robinson) - Fix QEMU restore from file in raw format (Daniel P. Berrange) - Take domain type into account when looking up default machine (Mark McLoughlin) - Fix schema to allow missing machine type (Mark McLoughlin) - Fix stream abort upon I/O failure during migration (Daniel P. Berrange) - Create /var/log/libvirt/{lxc,uml} dirs (Mark McLoughlin) - nodedev: Add locking in nodeNumOfDevices (Cole Robinson) - test: Throw a proper error in GetBridgeName (Cole Robinson) - 526769 change logrotate config default to weekly (Daniel Veillard) - Fix emission of domain events messages (Daniel P. Berrange) - unbreak `make rpcgen' (Paolo Bonzini) - unbreak migration (Paolo Bonzini) - Fix USB device re-labelling (Mark McLoughlin) - Avoid a libvirtd crash on broken input 523418 (Daniel Veillard) - Re-label image file backing stores (Mark McLoughlin) - Fix memory leaks in libvirtd's message processing (Matthias Bolte) - Fix QEMU test suite with new VNC env variable (Daniel P. Berrange) - VBox vboxDomainDestroy forgot to wait for completion (Pritesh Kothari) - Vbox call OpenHardDisk with "" instead of NULL (Pritesh Kothari) - Avoid double free in errors in virsh (Jim Fehlig) - Fix crash in device hotplug cleanup code (Daniel P. Berrange) - Maintain value of ctxt->node in virInterfaceDefParseDhcp (Laine Stump) - Fix some XPath relative node resets (Daniel Veillard) - Fix unitialized variable in qemudDomainDetachHostPciDevice() (Charles Duffy) - ESX: Check if a datastore is accessible first (Matthias Bolte) - Fix handling of Xen(ner) detection (Daniel P. Berrange) - Fix xen driver refcounting. (Matthias Bolte) - prevent attempt to call cat -c during virDomainSave to raw (Charles Duffy) - Don't do virSetConnError when virDrvSupportsFeature is successful. (Chris Lalancette) - Fix a double-free in qemudRunLoop() (Chris Lalancette) - Fix leak in PCI hostdev hot-unplug (Mark McLoughlin) - Fix net/disk hot-unplug segfault (Mark McLoughlin) Improvements: - schema: Update storage pool schema. (Cole Robinson) - test: Activate interfaces specified through driver config file. (Cole Robinson) - Rewrite example domain events programm for python (Daniel P. Berrange) - Support a new peer-to-peer migration mode & public API (Daniel P. Berrange) - LXC add augeas support for config file (Amy Griffis) - LXC add driver config file lxc.conf (Amy Griffis) - LXC do not truncate container log files on restart (Amy Griffis) - LXC initialize logging configuration (Amy Griffis) - Add debug for envp[] in virExecWithHook() (Amy Griffis) - Add accessors for logging filters and outputs (Amy Griffis) - Add virFileAbsPath() utility (Amy Griffis) - LXC implement memory control APIs (Ryota Ozaki) - Add a domain argument to SVirt *RestoreImageLabel (Jamie Strandboge) - test: Support loading node device info from file/XML (Cole Robinson) - test: Implement node device driver. (Cole Robinson) - configure: Add explict --with-python option. (Cole Robinson) - Tunnelled migration. (Chris Lalancette) - Various monitor improvements for migration. (Chris Lalancette) - 523639 Allows a <description> tag for domains (Daniel Veillard) - Add src/util/storage_file.c to the POTFILES.in. (Chris Lalancette) - Add a qemu feature flag for unix socket migration. (Chris Lalancette) - Let remoteClientStream only do RX if requested. (Chris Lalancette) - Introduce virStorageFileMetadata structure (Mark McLoughlin) - Allow control over QEMU audio backend (Daniel P. Berrange) - Handle data streams in remote client (Daniel P. Berrange) - Handle outgoing data streams in libvirtd (Daniel P. Berrange) - Handle incoming data streams in libvirtd (Daniel P. Berrange) - Lots of cleanups and improvement on QEmu monitor code (Daniel P. Berrange) - ESX add esxVI_Occurence enum to for occurences (Matthias Bolte) - ESX add x86_64 detection based on the CPUID (Matthias Bolte) - ESX add tests for the VMX to/from domain XML mapping (Matthias Bolte) - ESX Add esxDomainXMLToNative() (Matthias Bolte) - ESX Set challenge for auth callback to hostname (Matthias Bolte) - ESX Add esxNodeGetFreeMemory() (Matthias Bolte) - network: add 'bootp' and 'tftp' config (Paolo Bonzini) - OpenVZ Fix a restriction about domain names (Yuji NISHIDA) - Make pki_check.sh into an installed & supported tool (Daniel P. Berrange) - ESX add support for vmxnet3 virtual device (Shahar Klein) Cleanups: - remote: Don't print a warning every time a remote call fails (Cole Robinson) - storage: Report errors in FindPoolSources (Cole Robinson) - LXC fix return code handling in lxcVmStart (Ryota Ozaki) - Add a target for libvirt.devhelp (Daniel Veillard) - Remove some auto-generated files (Daniel P. Berrange) - Re-arrange doTunnelMigrate to simplify cleanup code (Daniel P. Berrange) - Separate out code for sending tunnelled data (Daniel P. Berrange) - Pull connection handling code out of doTunnelMigrate (Daniel P. Berrange) - Refactor native QEMU migration code (Daniel P. Berrange) - Don't force dconn to be NULL in virDomainMigrate (Daniel P. Berrange) - Remove unneccessary uri_in parameter from virMigratePrepareTunnel (Daniel P. Berrange) - Move the VIR_DRV_FEATURE* constants (Daniel P. Berrange) - Fix configure.ac message vertical alignment (Daniel P. Berrange) - cgroup: Fix -Werror breakage (Cole Robinson) - Fix handling return value of qemuMonitorSetBalloon (Ryota Ozaki) - Fix up "make check" (Chris Lalancette) - Fix rebuilding of devhelp files (Daniel P. Berrange) - Fix ordering of <exports> in API description file (Daniel P. Berrange) - node conf: Make parsing routines consistent with other drivers (Cole Robinson) - nodedev: Break out virNodeDeviceHasCap to node_conf (Cole Robinson) - python: Add a newline after custom classes (Cole Robinson) - python: Fix generated virInterface method names (Cole Robinson) - python: Use a pure python implementation of 'vir*GetConnect' (Cole Robinson) - python: Don't generate bindings for vir*Ref (Cole Robinson) - python: Don't generate conflicting conn.createXML functions. (Cole Robinson) - python: Remove use of xmllib in generator.py (Cole Robinson) - python: Remove FastParser from generator. (Cole Robinson) - Fix typo in Makefile.am breaking NEWS file generation (Daniel P. Berrange) - Fix build in separate build directory (Jiri Denemark) - Incorrect error message in virDomainNetDefParseXML (Florian Vichot) - Fix a few 'make rpm' breakages (Daniel Veillard) - Pass remote_message_header to the dispatch functions. (Chris Lalancette) - Fix up some warnings from stream DEBUG statements. (Chris Lalancette) - Fix apibuild.py warnings (Matthias Bolte) - Change signature of remoteSendStreamData() to fix compile warning (Matthias Bolte) - Add virStorageFileGetMetadata() helper (Mark McLoughlin) - Move virStorageGetMetadataFromFD() to libvirt_util (Mark McLoughlin) - Split virStorageGetMetadataFromFD() from virStorageBackendProbeTarget() (Mark McLoughlin) - Move file format enum to libvirt_util (Mark McLoughlin) - Remove hand-crafted UUID parsers (Daniel P. Berrange) - Helper functions for processing data streams in libvirtd (Daniel P. Berrange) - Standardize debugging messages in QEMU monitor code (Daniel P. Berrange) - Remove low level monitor APIs from header file (Daniel P. Berrange) - Rename qemudMonitorSendCont to qemuMonitorStartCPUs (Daniel P. Berrange) - Pull QEMU monitor interaction out to separate file (Daniel P. Berrange) - util.h needs libvirt.h for virConnectPtr (Mark McLoughlin) - Fix API doc extractor to stop munging comment formatting (Daniel P. Berrange) - Fix secret_driver compile warning, bug. (Charles Duffy) - ESX remove phantom mode (Matthias Bolte) - ESX replace esxUtil_EqualSuffix() with virFileHasSuffix() (Matthias Bolte) - ESX Whitespace cleanup (Matthias Bolte) - Fix up "make syntax-check" after the tree restructuring. (Chris Lalancette) - Introduce virStrncpy. (Chris Lalancette) - Ignore auto-generated header file (Daniel P. Berrange) - Remove an unnecessary variable from remoteIOReadMessage(). (Chris Lalancette) - Remove auto-generated header file from repo (Daniel P. Berrange) - Move example XML files into examples/xml (Daniel P. Berrange) - Remove all generated docs from source control (Daniel P. Berrange) - Fix missing data file in qemuhelpdata (Daniel P. Berrange) - Misc syntax-check fixes (Daniel P. Berrange) - Move remote protocol definition into src/remote/ (Daniel P. Berrange) - Move all shared utility files to src/util/ (Daniel P. Berrange) - Move all XML configuration handling to src/conf/ (Daniel P. Berrange) - Re-arrange python generator to make it clear what's auto-generated (Daniel P. Berrange) - Remove obsolete files (Daniel P. Berrange) - Move docs/examples into examples/ (Daniel P. Berrange) - Remove unused images from docs/ directory (Daniel P. Berrange) - Rename daemon main code (Daniel P. Berrange) - Move config files to align with driver sources (Daniel P. Berrange) - Move virsh into tools/ directory (Daniel P. Berrange) - Move security drivers to src/security/ (Daniel P. Berrange) - Move secret driver into src/secret/ (Daniel P. Berrange) - Move netcf interface driver into src/interface/ (Daniel P. Berrange) - Move network driver into src/network (Daniel P. Berrange) - Move remote driver to src/remote/ (Daniel P. Berrange) - Move test driver into src/test/ (Daniel P. Berrange) - Move node device drivers to src/node_device/ (Daniel P. Berrange) - Move storage drivers into src/storage/ (Daniel P. Berrange) - Move OpenVZ driver to src/openvz/ (Daniel P. Berrange) - Move UML driver to src/uml/ (Daniel P. Berrange) - Move QEMU driver to src/qemu/ (Daniel P. Berrange) - Move LXC driver into src/lxc/ (Daniel P. Berrange) - Move xen driver code into src/xen/ directory (Daniel P. Berrange) - Rename qemud/ directory to daemon/ (Daniel P. Berrange) - Refactor libvirt.spec to allow client-only builds (Daniel P. Berrange) Please give it a try, and thanks everybody for reports, patches or documentation ! Daniel -- Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ daniel(a)veillard.com | Rpmfind RPM search engine http://rpmfind.net/ http://veillard.com/ | virtualization library http://libvirt.org/

1 0

[libvirt]: Not able to login to container when cgroup is enabled
by Rishikesh 14 Oct '09

14 Oct '09

Hi All, On F12 i was playing with libvirt support for container. I am seeing two issue over here: 1> Not able to create a container when start cgconfig services. 2> When i stop cgconfig, then able to create a container, but seeing few error messages to libvirtd daemon. 3> After creation of container through libvirt, i am not able to touch any file or directory. Attached: vm1.xml , through which i created container. I referred following link: http://libvirt.org/drvlxc.html http://openvz.org/pipermail/devel/2008-September/014314.html Detailed Bug Report: When i create a container without starting "cgconfig" service, i am able to create but getting some error message in libvirtd daemon: [root@mx3950 ~]# libvirtd 14:47:24.556: warning : qemudStartup:565 : Unable to create cgroup for driver: No such device or address 14:47:24.647: warning : lxcStartup:1462 : Unable to create cgroup for driver: No such device or address libvir: Linux Container error : failed to connect to client socket: No such file or directory libvir: Linux Container error : failed to connect to client socket: No such file or directory libvir: Linux Container error : failed to connect to client socket: No such file or directory But when i start cgconfig service, then i am not able to open console for my container. Below is the detailed steps which i executed. [root@mx3950 ~]# /etc/init.d/cgconfig stop Stopping cgconfig service: [ OK ] [root@mx3950 ~]# virsh -c lxc:/// define vm1.xml error: Failed to define domain from vm1.xml error: operation failed: domain 'vm1' is already defined with uuid ef558ab3-54ea-5b95-205d-8c0096ae9a4f [root@mx3950 ~]# virsh -c lxc:/// start vm1 Domain vm1 started [root@mx3950 ~]# virsh -c lxc:/// console vm1 Connected to domain vm1 Escape character is ^] sh-4.0# ps -ef UID PID PPID C STIME TTY TIME CMD root 1 0 0 15:23 pts/4 00:00:00 /bin/sh root 4 1 0 15:23 pts/4 00:00:00 ps -ef sh-4.0# touch rishi touch: cannot touch `rishi': Permission denied sh-4.0# exit exit [root@mx3950 ~]# /etc/init.d/cgconfig start Starting cgconfig service: [ OK ] [root@mx3950 ~]# virsh -c lxc:/// start vm1 Domain vm1 started [root@mx3950 ~]# virsh -c lxc:/// console vm1 Connected to domain vm1 Escape character is ^] 15:24:12.869: error : vshRunConsole:77 : unable to open tty /dev/pts/3: No such file or directory [root@mx3950 ~]# Package List: [root@mx3950 ~]# rpm -qa | grep -e libcgroup -e libvirt libcgroup-0.34-2.fc12.x86_64 libvirt-0.7.1-1.fc12.x86_64 libvirt-client-0.7.1-1.fc12.x86_64 libvirt-python-0.7.1-1.fc12.x86_64 [root@mx3950 ~]# uname -a Linux mx3950.in.ibm.com 2.6.31-14.fc12.x86_64 #1 SMP Tue Sep 15 03:48:57 EDT 2009 x86_64 x86_64 x86_64 GNU/Linux

5 17

[libvirt] Execute script on migration
by Radek Hladík 13 Oct '09

13 Oct '09

Short question: What is the best way to run some script on VM migration? Detailed description: We use own system for volume management. Its based on md and we found it very flexible. Basic idea is that we have md mirror everywhere we can. This suits great for libvirt: We have a disk on iSCSI server, its "partioned" using device mapper, the resulting "partition" is member of md raid1, that raid is exported via iSCSI to the libvirt node, it is on the node included in another md raid1 and that "final" md raid is used as storage by the VM. This provides high level of abstraction that allows us to migrate the storage to another disk, another iSCSI server, etc... The only glitch is on live migration of the VM. It is possible to assemble and run the array on two nodes. Also all superblock modifications are done on first write to the array. However it would be more secure to stop the array after the source vm stops and before destination vm starts. What is the best way to run some script in this particular moment? I know we could create a new storage driver but I am not sure that it is the best option. -- Open WebMail Project (http://openwebmail.org)

3 2

[libvirt] [PATCH] test: Activate interfaces specified through driver config file.
by Cole Robinson 13 Oct '09

13 Oct '09

This matches the behavior of domains, storage, etc. Also, fix up some whitespace issues in the area. Signed-off-by: Cole Robinson <crobinso(a)redhat.com> --- src/test/test_driver.c | 5 ++++- 1 files changed, 4 insertions(+), 1 deletions(-) diff --git a/src/test/test_driver.c b/src/test/test_driver.c index 0b866d2..0541a73 100644 --- a/src/test/test_driver.c +++ b/src/test/test_driver.c @@ -779,13 +779,16 @@ static int testOpenFromFile(virConnectPtr conn, if (!def) goto error; } else { - if ((def = virInterfaceDefParseNode(conn, xml, ifaces[i])) == NULL) + if ((def = virInterfaceDefParseNode(conn, xml, ifaces[i])) == NULL) goto error; } + if (!(iface = virInterfaceAssignDef(conn, &privconn->ifaces, def))) { virInterfaceDefFree(def); goto error; } + + iface->active = 1; virInterfaceObjUnlock(iface); } VIR_FREE(ifaces); -- 1.6.5.rc2

3 3

[libvirt] [PATCH] storage: Report errors in FindPoolSources
by Cole Robinson 13 Oct '09

13 Oct '09

Signed-off-by: Cole Robinson <crobinso(a)redhat.com> --- src/storage/storage_driver.c | 15 ++++++++++++--- 1 files changed, 12 insertions(+), 3 deletions(-) diff --git a/src/storage/storage_driver.c b/src/storage/storage_driver.c index 9ab53e1..4f8949b 100644 --- a/src/storage/storage_driver.c +++ b/src/storage/storage_driver.c @@ -440,15 +440,24 @@ storageFindPoolSources(virConnectPtr conn, char *ret = NULL; backend_type = virStoragePoolTypeFromString(type); - if (backend_type < 0) + if (backend_type < 0) { + virStorageReportError(conn, VIR_ERR_INTERNAL_ERROR, + _("unknown storage pool type %s"), type); goto cleanup; + } backend = virStorageBackendForType(backend_type); if (backend == NULL) goto cleanup; - if (backend->findPoolSources) - ret = backend->findPoolSources(conn, srcSpec, flags); + if (!backend->findPoolSources) { + virStorageReportError(conn, VIR_ERR_NO_SUPPORT, + _("pool type '%s' does not support source " + "discovery"), type); + goto cleanup; + } + + ret = backend->findPoolSources(conn, srcSpec, flags); cleanup: return ret; -- 1.6.5.rc2

3 3

[libvirt] [PATCH] remote: Don't print a warning every time a remote call fails
by Cole Robinson 13 Oct '09

13 Oct '09

$ sudo virsh pool-start idontexist 10:58:18.716: warning : processCallDispatchReply:7612 : Method call error error: failed to get pool 'idontexist' error: Storage pool not found: no pool with matching name 'idontexist' That warning doesn't server much purpose being printed via a virsh call. Change the warning to a debug message. Signed-off-by: Cole Robinson <crobinso(a)redhat.com> --- src/remote/remote_driver.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/src/remote/remote_driver.c b/src/remote/remote_driver.c index 284593a..162d682 100644 --- a/src/remote/remote_driver.c +++ b/src/remote/remote_driver.c @@ -7606,7 +7606,7 @@ processCallDispatchReply(virConnectPtr conn, struct private_data *priv, return 0; case REMOTE_ERROR: - VIR_WARN0("Method call error"); + VIR_DEBUG0("Method call error"); memset (&thecall->err, 0, sizeof thecall->err); if (!xdr_remote_error (xdr, &thecall->err)) { error (in_open ? NULL : conn, -- 1.6.5.rc2

2 2

[libvirt] [PATCH] LXC fix rc handling in lxcVmStart
by Ryota Ozaki 13 Oct '09

13 Oct '09

Hi, In lxcVmStart we have to set rc = -1 before jumping to the cleanup code when a local procedure call fails and the rc is overwritten, however, some codes don't observe the rule and so invalid cleanups are likely to happen. A simple solution for the defect is to add rc = -1 before every jumping if needed, however, IMHO, we first should not reuse rc for another purpose rather than the return value of lxcVmStart. Instead the patch introduces yet another variable for storing only the return value of the local procedure calls. By doing so, we don't need to care about resetting rc anymore. Thanks, ozaki-r >From c3eb2cfa1280b220d5b92ad2ad8bbf6cf6ca0a04 Mon Sep 17 00:00:00 2001 From: Ryota Ozaki <ozaki.ryota(a)gmail.com> Date: Tue, 13 Oct 2009 00:36:41 +0900 Subject: [PATCH] LXC fix rc handling in lxcVmStart * src/lxc/lxc_driver.c: don't reuse rc for local procedure calls --- src/lxc/lxc_driver.c | 15 ++++++--------- 1 files changed, 6 insertions(+), 9 deletions(-) diff --git a/src/lxc/lxc_driver.c b/src/lxc/lxc_driver.c index ef0d368..0b614e3 100644 --- a/src/lxc/lxc_driver.c +++ b/src/lxc/lxc_driver.c @@ -1117,7 +1117,7 @@ static int lxcVmStart(virConnectPtr conn, lxc_driver_t * driver, virDomainObjPtr vm) { - int rc = -1; + int rc = -1, r; unsigned int i; int parentTty; char *parentTtyPath = NULL; @@ -1126,8 +1126,8 @@ static int lxcVmStart(virConnectPtr conn, unsigned int nveths = 0; char **veths = NULL; - if ((rc = virFileMakePath(driver->logDir)) < 0) { - virReportSystemError(conn, rc, + if ((r = virFileMakePath(driver->logDir)) < 0) { + virReportSystemError(conn, r, _("cannot create log directory '%s'"), driver->logDir); return -1; @@ -1157,10 +1157,8 @@ static int lxcVmStart(virConnectPtr conn, goto cleanup; /* Persist the live configuration now we have veth & tty info */ - if (virDomainSaveConfig(conn, driver->stateDir, vm->def) < 0) { - rc = -1; + if (virDomainSaveConfig(conn, driver->stateDir, vm->def) < 0) goto cleanup; - } if ((logfd = open(logfile, O_WRONLY | O_APPEND | O_CREAT, S_IRUSR|S_IWUSR)) < 0) { @@ -1183,11 +1181,10 @@ static int lxcVmStart(virConnectPtr conn, goto cleanup; /* And get its pid */ - if ((rc = virFileReadPid(driver->stateDir, vm->def->name, &vm->pid)) != 0) { - virReportSystemError(conn, rc, + if ((r = virFileReadPid(driver->stateDir, vm->def->name, &vm->pid)) != 0) { + virReportSystemError(conn, r, _("Failed to read pid file %s/%s.pid"), driver->stateDir, vm->def->name); - rc = -1; goto cleanup; } -- 1.6.2.5

4 3

[libvirt] unable to start guest under kvm (revisited)
by Anton Protopopov 13 Oct '09

13 Oct '09

The problem, that I post here http://www.mail-archive.com/libvir-list@redhat.com/msg16154.html and that was solved, comes back in a different way. Consider a working kvm machine with the following description: ... <os> <type arch='x86_64' machine='pc'>hvm</type> <boot dev='hd'/> </os> ... If I delete "machine='pc'" attribute from <type>, so description becomes ... <os> <type arch='x86_64'>hvm</type> <boot dev='hd'/> </os> ... then libvirt chooses a machine='pc-0.11' value (again: qemu API version instead of kvm version) and I get virsh # start DD error: Failed to start domain DD error: internal error unable to start guest: Supported machines are: pc Standard PC (default) isapc ISA-only PC --- This problem actually occurs when using virt-manager, see https://www.redhat.com/archives/virt-tools-list/2009-October/msg00037.html

4 6

[libvirt] feature request: virDomainSendkeys API
by Dan Kenigsberg 13 Oct '09

13 Oct '09

Our QA team makes use of qemu's sendkeys monitor command to test interesting things such as Windows guest installation-from-scratch. When we move to libvirt, we loose this ability. On irc, dpb suggested using vnc to send the required keys. Are you aware of a command-line tool for that, or should we build our own? If we go down this path, we would have to build a similar tool for testing VMs with SPICE. (and have nothing to do for VMs with no graphics support, but these are not used in the product anyway) Alternatively, libvirt can supply a new API for injecting key strokes into a domain. Surely it's useful only for debug and qa, but so is virDomainMemoryPeek. Is that reasonable? Dan.

3 2

[libvirt] cannot set ownership on /nfs/mounted/image.iso: Operation not permitted
by Dan Kenigsberg 13 Oct '09

13 Oct '09

I having libvirt run qemu as non-root, but I don't think libvirt should try chowning image files. Cerntainly not if they sit on a root_squashing nfs server. Also, the behavior of libvirt in this regard is a bit surprising for local image files and even for host devices. It should be enough that an image is accessible to qemu; it does not have to be owned by it. The fact that qemuDomainSetAllDeviceOwnership() with restore=1 does not really restore ownership of device/file but force-chowns it to root:root is anouther surprise. Dan

2 1

[libvirt] [PATCH] Fix the spec file
by Daniel Veillard 13 Oct '09

13 Oct '09

I think I erased a % when trying to apply the last patch from Amy, trivial fix, will commit, Daniel -- Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ daniel(a)veillard.com | Rpmfind RPM search engine http://rpmfind.net/ http://veillard.com/ | virtualization library http://libvirt.org/

2 2

[libvirt] [PATCH] Fix QEMU restore from file in raw format
by Daniel P. Berrange 13 Oct '09

13 Oct '09

The logic for running the decompression programs was broken in commit f238709304f9f6c57204cdd943e542cbae38fa5f, so that for non-raw formats the decompression program was never run, and for raw formats, it tried to exec an argv[] with initial NULL in the program name. * src/qemu/qemu_driver.c: Fix logic in runing decompression program --- src/qemu/qemu_driver.c | 3 +-- 1 files changed, 1 insertions(+), 2 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 1e23a9b..556984a 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -3781,9 +3781,8 @@ static int qemudDomainRestore(virConnectPtr conn, goto cleanup; } - if (header.compressed != QEMUD_SAVE_FORMAT_RAW) + if (header.compressed != QEMUD_SAVE_FORMAT_RAW) { intermediate_argv[0] = prog; - else { intermediatefd = fd; fd = -1; if (virExec(conn, intermediate_argv, NULL, NULL, -- 1.6.2.5

3 2

[libvirt] [PATCH 0/9] Support disk-hotremove and controller hotplugging
by Wolfgang Mauerer 13 Oct '09

13 Oct '09

Hi, this patch reworks libvirt's disk-hotadd support and introduces support for disk controller hotplugging and disk-hotremove (see http://thread.gmane.org/gmane.comp.emulators.libvirt/15860 for more details). Currently, it targets only qemu and also requires some additions to qemu that have only recently been submitted, which is why this is cross-posted to qemu-devel. Hopefully the lack of support in qemu does not prevent the community from reviewing the code, though ;-) Notice that no documentation and testcases are included yet; I'll supply those once the interface is fixed and possible issues are settled. Thanks for your comments, Wolfgang docs/schemas/domain.rng | 145 ++++++++++---- src/domain_conf.c | 208 ++++++++++++++++++++- src/domain_conf.h | 39 ++++- src/libvirt_private.syms | 4 +- src/qemu_driver.c | 479 +++++++++++++++++++++++++++++++++++++++++++--- 5 files changed, 802 insertions(+), 73 deletions(-) -- Siemens AG, Corporate Technology, CT SE 2 Corporate Competence Center Embedded Linux

5 26

[libvirt] [PATCH] Fix virFileReadLimFD/virFileReadAll to handle EINTR
by Daniel P. Berrange 13 Oct '09

13 Oct '09

The fread_file_lim() function uses fread() but never handles EINTR results, causing unexpected failures when reading QEMU help arg info. It was unneccessarily using FILE * instead of plain UNIX file handles, which prevented use of saferead() * src/util/util.c: Switch fread_file_lim over to use saferead instead of fread, remove FILE * use, and rename --- src/util/util.c | 45 ++++++++++++--------------------------------- 1 files changed, 12 insertions(+), 33 deletions(-) diff --git a/src/util/util.c b/src/util/util.c index e5135fc..98f8a14 100644 --- a/src/util/util.c +++ b/src/util/util.c @@ -898,7 +898,7 @@ virExec(virConnectPtr conn, number of bytes. If the length of the input is <= max_len, and upon error while reading that data, it works just like fread_file. */ static char * -fread_file_lim (FILE *stream, size_t max_len, size_t *length) +saferead_lim (int fd, size_t max_len, size_t *length) { char *buf = NULL; size_t alloc = 0; @@ -906,8 +906,8 @@ fread_file_lim (FILE *stream, size_t max_len, size_t *length) int save_errno; for (;;) { - size_t count; - size_t requested; + int count; + int requested; if (size + BUFSIZ + 1 > alloc) { alloc += alloc / 2; @@ -923,12 +923,12 @@ fread_file_lim (FILE *stream, size_t max_len, size_t *length) /* Ensure that (size + requested <= max_len); */ requested = MIN (size < max_len ? max_len - size : 0, alloc - size - 1); - count = fread (buf + size, 1, requested, stream); + count = saferead (fd, buf + size, requested); size += count; if (count != requested || requested == 0) { save_errno = errno; - if (ferror (stream)) + if (count < 0) break; buf[size] = '\0'; *length = size; @@ -941,12 +941,12 @@ fread_file_lim (FILE *stream, size_t max_len, size_t *length) return NULL; } -/* A wrapper around fread_file_lim that maps a failure due to +/* A wrapper around saferead_lim that maps a failure due to exceeding the maximum size limitation to EOVERFLOW. */ -static int virFileReadLimFP(FILE *fp, int maxlen, char **buf) +int virFileReadLimFD(int fd, int maxlen, char **buf) { size_t len; - char *s = fread_file_lim (fp, maxlen+1, &len); + char *s = saferead_lim (fd, maxlen+1, &len); if (s == NULL) return -1; if (len > maxlen || (int)len != len) { @@ -960,37 +960,16 @@ static int virFileReadLimFP(FILE *fp, int maxlen, char **buf) return len; } -/* Like virFileReadLimFP, but use a file descriptor rather than a FILE*. */ -int virFileReadLimFD(int fd_arg, int maxlen, char **buf) -{ - int fd = dup (fd_arg); - if (fd >= 0) { - FILE *fp = fdopen (fd, "r"); - if (fp) { - int len = virFileReadLimFP (fp, maxlen, buf); - int saved_errno = errno; - fclose (fp); - errno = saved_errno; - return len; - } else { - int saved_errno = errno; - close (fd); - errno = saved_errno; - } - } - return -1; -} - int virFileReadAll(const char *path, int maxlen, char **buf) { - FILE *fh = fopen(path, "r"); - if (fh == NULL) { + int fd = open(path, O_RDONLY); + if (fd < 0) { virReportSystemError(NULL, errno, _("Failed to open file '%s'"), path); return -1; } - int len = virFileReadLimFP (fh, maxlen, buf); - fclose(fh); + int len = virFileReadLimFD(fd, maxlen, buf); + close(fd); if (len < 0) { virReportSystemError(NULL, errno, _("Failed to read file '%s'"), path); return -1; -- 1.6.2.5

2 2

[libvirt] Tentative schedule for next release
by Daniel Veillard 09 Oct '09

09 Oct '09

I'm suggesting to try to push 0.7.2 around Fri 16 Oct, we already have a number of cleanup patches in and serious improvements for ESX driver. If we go with that plan I think patches and big changes should be pushed in the tree by Fri 9, this includes among other things the AppArmor security driver. Daniel -- Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ daniel(a)veillard.com | Rpmfind RPM search engine http://rpmfind.net/ http://veillard.com/ | virtualization library http://libvirt.org/

2 3

[libvirt] dynamic scsi disk attach seems to be broken in qemu(-kvm)-0.11, libvirt-0.7.1
by Dustin Kirkland 09 Oct '09

09 Oct '09

We're experiencing what we believe to be a regression in qemu-kvm-0.11 and libvirt-0.7.0, from previous versions of kvm-84 and libvirt-0.6.1: Dynamically attaching sdb (a second scsi disk) fails. (Note that I have tested this against libvirt-0.7.1 built from source as well--same results.) Here are some instructions to reproduce this issue: Create some small auxiliary storage to dynamically add to a running VM. (host) $ dd if=/dev/zero of=/tmp/foo bs=1M count=64 Boot a vm in under qemu-kvm-0.11.0 like this: (host) $ kvm -drive file=server.img,if=scsi,boot=on -boot c To reproduce the problem, you must use if=scsi. This problem manifests itself when subsequent scsi storage devices are added. If the device is the first scsi device on the system, then it would succeed. It's the 2nd scsi device that causes the problem. Once the system is booted, login and in the vm, load this module: (vm) $ sudo modprobe acpiphp Check that the acpiphp slots are loaded in dmesg. And note that there is only one /dev/sd[a-z] device. Now, drop to the qemu console with ctrl-alt-2, and add the storage: (qemu) pci_add auto storage file=/tmp/foo,if=scsi OK domain 0, bus 0, slot 6, function 0 Switch back to the vm linux shell with ctrl-alt-1, and look at the dmesg output. (vm) $ dmesg | tail -n 12 [ 44.033397] pci 0000:00:06.0: reg 10 io port: [0x00-0xff] [ 44.033443] pci 0000:00:06.0: reg 14 32bit mmio: [0x000000-0x0003ff] [ 44.033486] pci 0000:00:06.0: reg 18 32bit mmio: [0x000000-0x001fff] [ 44.033899] pci 0000:00:02.0: BAR 6: bogus alignment [0x0-0x0] flags 0x2 [ 44.033975] decode_hpp: Could not get hotplug parameters. Use defaults [ 44.042277] sym53c8xx 0000:00:06.0: enabling device (0000 -> 0003) [ 44.043230] ACPI: PCI Interrupt Link [LNKB] enabled at IRQ 11 [ 44.043247] sym53c8xx 0000:00:06.0: PCI INT A -> Link[LNKB] -> GSI 11 (level, high) -> IRQ 11 [ 44.045237] sym1: <895a> rev 0x0 at pci 0000:00:06.0 irq 11 [ 44.047586] sym1: No NVRAM, ID 7, Fast-40, LVD, parity checking [ 44.055399] sym1: SCSI BUS has been reset. [ 44.063329] scsi3 : sym-2.2.3 More importantly, note that no /dev/sd[b-z] device shows up. If you now drop to the qemu console with ctrl-alt-2, and do this: (qemu) pci_add auto storage file=/tmp/foo,if=virtio OK domain 0, bus 0, slot 7, function 0 Going back to the vm with ctrl-alt-1, you can now see a new /dev/vda device registered. Now, all of that said, it is actually possible to hot-add a second scsi device. However, as far as I can tell, this method is not yet supported by libvirt. It looks to me that with modern qemu, you have to do it this way: Drop to a qemu console with ctrl-alt-2. Get the address of the current scsi bus: (qemu) info pci Look for "SCSI Controller". In my case, it's on Bus 0, device 4, function 0 Now instead of pci_add, use drive_add (qemu) drive_add 0:4 file=/tmp/foo,if=scsi OK bus 0, unit 1 This is not ideal, however, as I tried re-scan the scsi bus with rescan-scsi-bus.sh (from scsitools) without luck. It did not pick up the new sdb device. However, I did reboot the vm, and voila, I now have /dev/sdb. I would like some advice on how to proceed with this bug, and where the solution lies...in qemu or in libvirt. Ultimately, I would like the behavior we had in our previous release with kvm-84 and libvirt-0.6.1, where we could dynamically add scsi devices without a problem, using: pci_add 1 storage file=/tmp/foo,if=scsi Can anyone else reproduce this? Is this considered a regression by anyone else? Where should I look to solve this, in libvirt, or in qemu? -- :-Dustin

3 4

[libvirt] Help needed: simple python api to obtain events
by Dan Kenigsberg 09 Oct '09

09 Oct '09

Would someone help me have a shrink-wrapped solution for obtaining libvirt events in python? I'm looking for a python module with no API calls (or maybe just a few). I'd like to register a callback functions with virConnect.domainEventRegister(), and want them to be called in the context of a specialized (automatically-created?) thread. I don't mind if the module has two API calls instead: register() and startPollThread(). I learned that it takes me far too long to do it on my own, and that just modifying events-python.py (attached) reaches miserable outcomes ("libvir: Domain error : invalid domain pointer in virDomainFree", and deadlocked python: #0 0x00000037fe40d174 in __lll_lock_wait () from /lib64/libpthread.so.0 #1 0x00000037fe408aca in _L_lock_1034 () from /lib64/libpthread.so.0 #2 0x00000037fe40898c in pthread_mutex_lock () from /lib64/libpthread.so.0 #3 0x0000003cce04db84 in remoteDomainEventFired () #4 0x00002b4cc96e1612 in libvirt_virEventInvokeHandleCallback () ) Help most welcome, Dan.

4 10

[libvirt] [PATCH 0/9] Support new peer-2-peer migration mode & public API
by Daniel P. Berrange 09 Oct '09

09 Oct '09

This series is an update of http://www.redhat.com/archives/libvir-list/2009-September/msg00540.html There isn't as much functional change here as you might presume from the number of patches. Since Chris' tunnelled migration code is added, I thought it better to do alot of small refactoring steps rather than munge it all in one patch. One particular notable change since last time is that the new virDomainMigrateToURI method does *not* mandate use of the new "VIR_MIGRATE_PEER2PEER" flag anymore. I will now explain why... Traditionally migration has been a 3 step process 1. client -> dest (prepare) 2. client -> source (perform) 3. client -> dest (finish) This is why virDomainMigrate requires the extra 'dconn' parameter from the client app With the tunnelled migration / peer-to-peer migration, the extra 'dconn' parameter is no longer required from the client, because the source libvirt driver directly opens a connection to the destination libvirtd daemon on the remote host. ie libvirtd is talking to another libvirtd peer-2-peer. In considering the implementation of the PEER_TO_PEER flag for the Xen and VMWare drivers though I realized that even this is not technically neccessary, and in fact detrimental to use of libvirt for migrating with these drivers. Both Xen and VMWare have their own management daemon which is perfectly capable of handling the entire process without any need for libvirtd to be involved on the destination host. libvirt need only initiate the migration op on the source host. If you look at the impl of Prepare/Finish ops in the Xen driver this should be obvious, since they are pretty much no-ops Thus we in fact have 3 possible migration processes * Traditional libvirt client managed client -> dest (prepare) client -> source (perform) client -> dest (finish) * New peer-to-peer migration (on which tunnelled mig is built) client -> source (perform) source -> dest (open) source -> dest (prepareTunnel) source -> dest (streamSend) source -> dest (finish) source -> dest (close) * New direct migration (hypervisor managed) client -> source (perform) source -> dest (whatever the HV's native migration does) In terms of libvirt public APIs this works out as follows * Traditional libvirt client managed A hypervisor specific URI style, passed to virDomainMigrate() with no flags set * New peer-to-peer migration (on which tunnelled mig is built) The standard libvirt URI style, passed to virDomainMigrate() or virDomainMigrateToURI() with VIR_DOMAIN_MIGRATE_PEER2PEER flag set. The virDomainMigrateToURI() method is better since it avoids the unneccessary burden on the client of connecting to the remote libvirtd. * New direct migration (hypervisor managed) A hypervisor specific URI style, passed to virDomainMigrateToURI() with no flags set The QEMU driver can't support the latter, since it has no native management daemon - it requires libvirtd's help. Xen and VMWare, and probably VirtualBox can support this just fine. This is good for flexibility of usage in terms of authentication too, since each of these 3 modes has different characteristics * Traditional libvirt client managed client -> source libvirtd auth client -> dest libvirtd auth possible hypervisor -> hypervisor auth * New peer-to-peer migration, without tunnelling client -> source libvirtd auth source -> dest libvirtd auth possible hypervisor -> hypervisor auth * New peer-to-peer migration, without tunnelling client -> source libvirtd auth source -> dest libvirtd auth * New direct migration (hypervisor managed) client -> source libvirtd auth possible hypervisor -> hypervisor auth NB, 'client -> source libvirtd auth' is essentially no-op if initiating migration from the source host directly in all these cases. With with the combination of the two virDomainMigrate and virDomainMigrateToURI methods, and the VIR_MIGRATE_PEER2PEER flag, I believe we're well placed to cover all the main deployment/auth scenarios for migration in all hypervisors without imposing extra auth burden ourselves as we did in the past Further things that need to be done though: - Xen could easily support PEER2PEER and TUNNELLED flags - Add a VIR_MIGRATE_SECURE flag, to indicate that the app only wants the migration to be performed if the HV can guarentee the data channel is secure. - Document the scenarios I just outlined and give some mre guidance to app developers/administrators about the tradeoff between each scenario Regards, Daniel

3 15

[libvirt] [PATCH] Remove some auto-generated files
by Daniel P. Berrange 09 Oct '09

09 Oct '09

The original mail i tried to send was far too large, so here's it edited to remove the boring bits. Daniel > From: "Daniel P. Berrange" <berrange(a)redhat.com> > To: libvir-list(a)redhat.com > Cc: "Daniel P. Berrange" <berrange(a)redhat.com> > Subject: [PATCH] Remove some auto-generated files > Date: Thu, 8 Oct 2009 17:03:07 +0100 > > Removes some auto-generated files still under version control. > It also moves the rule for generating NEWS into the Makefile.am > that's in the same directory as the output file to avoid confusion > > * docs/libvirt-api.xml, docs/libvirt-refs.xml, NEWS: Remove > auto-generated files from source control > * Makefile.am: Add rule for generating NEWS file > * docs/Makefile.am: Remove rule for generating NEWS file > --- > Makefile.am | 9 + > NEWS | 1620 ---------- > docs/Makefile.am | 10 +- > docs/libvirt-api.xml | 2986 ------------------ > docs/libvirt-refs.xml | 8313 ------------------------------------------------- > 5 files changed, 10 insertions(+), 12928 deletions(-) > delete mode 100644 NEWS > delete mode 100644 docs/libvirt-api.xml > delete mode 100644 docs/libvirt-refs.xml > > diff --git a/Makefile.am b/Makefile.am > index add7e2a..977ad0c 100644 > --- a/Makefile.am > +++ b/Makefile.am > @@ -31,6 +31,15 @@ EXTRA_DIST = \ > pkgconfigdir = $(libdir)/pkgconfig > pkgconfig_DATA = libvirt.pc > > +NEWS: $(top_srcdir)/docs/news.xsl $(top_srcdir)/docs/news.html.in > + -@(if [ -x $(XSLTPROC) ] ; then \ > + $(XSLTPROC) --nonet $(top_srcdir)/docs/news.xsl \ > + $(top_srcdir)/docs/news.html.in \ > + | perl -0777 -pe 's/\n\n+$$/\n/' \ > + | perl -pe 's/[ \t]+$$//' \ > + > $@-t && mv $@-t $@ ; fi ); > + > + > rpm: clean > @(unset CDPATH ; $(MAKE) dist && rpmbuild -ta $(distdir).tar.gz) > > diff --git a/docs/Makefile.am b/docs/Makefile.am > index 8832b72..0c206c3 100644 > --- a/docs/Makefile.am > +++ b/docs/Makefile.am > @@ -80,7 +80,7 @@ EXTRA_DIST= \ > > CLEANFILES = $(dot_html) $(apihtml) $(devhelphtml) > > -all: web $(top_builddir)/NEWS > +all: web > > api: libvirt-api.xml libvirt-refs.xml > > @@ -133,14 +133,6 @@ libvirt-api.xml libvirt-refs.xml: apibuild.py \ > $(srcdir)/../src/libvirt.c $(srcdir)/../src/util/virterror.c > -(./apibuild.py) > > -$(top_builddir)/NEWS: $(top_srcdir)/docs/news.xsl $(top_srcdir)/docs/news.html.in > - -@(if [ -x $(XSLTPROC) ] ; then \ > - $(XSLTPROC) --nonet $(top_srcdir)/docs/news.xsl \ > - $(top_srcdir)/docs/news.html.in \ > - | perl -0777 -pe 's/\n\n+$$/\n/' \ > - | perl -pe 's/[ \t]+$$//' \ > - > $@-t && mv $@-t $@ ; fi ); > - > clean-local: > rm -f *~ *.bak *.hierarchy *.signals *-unused.txt > ----- End forwarded message ----- -- |: Red Hat, Engineering, London -o- http://people.redhat.com/berrange/ :| |: http://libvirt.org -o- http://virt-manager.org -o- http://ovirt.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: GnuPG: 7D3B9505 -o- F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 :|

2 1

[libvirt] Xen: Domain.migrate() causes invalid listDomains() output
by Thomas Treutner 09 Oct '09

09 Oct '09

Hi, I'm using Xen 3.2 (Debian Lenny) and libvirt 0.7.1. When I use Domain.migrate() from libvirt-java 0.3.0, a following Connect.listDomains() gives wrong DomainIDs (they seem to be the ones before the migration) and therefore DomainLookupByID() fails because of a nonexistent ID. Restarting libvirtd and/or xend doesn't help. xm list etc. output is still correct, but anything that uses libvirt (like virt-top) is having troubles. The migration itself is working without glitches, the domains are up&running and responsive. But at this point, libvirt is completely mistaken about the running domains and their IDs. There are some quite similar bug reports and mails here, but they are rather old. Shall I provide more debug output? tia&kr, thomas

2 4

[libvirt] libvirt/kvm config for routed networking
by William Wagner 08 Oct '09

08 Oct '09

Hello, I hope this is the right list I should post to as my question is not development related. I am trying to setup a kvm/libvirt VM on my host (Ubuntu jaunty). My host has a public static IP and my VM also has a public static IP. Unfortunately I can not use bridged networking as the hosting provider has configured their switch to only accept packets from the MAC address of the host. I want to be able to setup my VM so it has the public static IP and it appears to be directly connected to the net. I believe I can do this with routed networking. I have created a new routed network: <network> <name>routed-net</name> <bridge name="routed%d" /> <forward mode="route" dev="eth0"/> <ip address="10.255.255.2" netmask="255.255.255.255"> </ip> </network> and in my vm's config I have: <interface type='user'> <source network='routed-net'/> <mac address='54:52:00:47:a8:38'/> <model type='virtio'/> </interface> Will this mean that the VM is placed on the routed network? Then I just need to add suitable routing rules on the host and everything should work? Is there a way to get libvirt to add the rules automatically when the VM starts. I have previously used Xen where you are able to specify what the IP address of the VM is and entries are automatically added to iptables. Is there similar syntax for libvirt and if so what is it? If not how do you recommend adding the routing rules? Thanks. Will. -- ------------------------------------------------------------------------ Will Wagner will_wagner(a)carallon.com Development Manager Office Tel: +44 (0)20 7371 2032 Carallon Ltd, Studio G20, Shepherds Building, Rockley Rd, London W14 0DA ------------------------------------------------------------------------

2 1

[libvirt] [PATCH 0/2] Take 2 - live interface info + IPv6 interface config
by Laine Stump 08 Oct '09

08 Oct '09

This is similar to the (now deprecated) patches I sent on 9/28 and 10/04, but without the unpopular "source" attribute in the ip address. You can successfully build libvirt if you have the latest release of netcf installed, but you'll need to get the netcf source, apply the patches in the thread listed below, and build/install netcf to get correct/proper output. https://fedorahosted.org/pipermail/netcf-devel/2009-October/000289.html

3 10

[libvirt] [PATCH 0/7] Series short description
by Amy Griffis 08 Oct '09

08 Oct '09

The following patches fix logging for the lxc controller. The lxc controller makes use of the libvirt logging module, but because it doesn't re-initialize the logging configuration the messages are currently going nowhere. There are a couple of options in fixing this. The lxc controller could use libvirtd's logging configuration exactly, or it adopt a more independent mode of logging to the container log file. After talking with Daniel Veillard a while back, it seemed clear that both behaviors are desirable. I resolved this by adding a lxc.conf file for the lxc driver, with a boolean variable to toggle between modes. I hope this looks good to both libvirt and container folks. libvirt.spec.in | 16 ++++++ src/Makefile.am | 25 +++++++-- src/libvirt_private.syms | 3 + src/lxc/libvirtd_lxc.aug | 30 +++++++++++ src/lxc/lxc.conf | 13 +++++ src/lxc/lxc_conf.c | 24 +++++++++ src/lxc/lxc_conf.h | 1 src/lxc/lxc_controller.c | 3 + src/lxc/lxc_driver.c | 86 ++++++++++++++++++++++++++++---- src/lxc/test_libvirtd_lxc.aug | 31 ++++++++++++ src/util/logging.c | 109 ++++++++++++++++++++++++++++++++++++++--- src/util/logging.h | 14 +++++ src/util/util.c | 49 ++++++++++++++++++ src/util/util.h | 3 + 14 files changed, 379 insertions(+), 28 deletions(-) create mode 100644 src/lxc/libvirtd_lxc.aug create mode 100644 src/lxc/lxc.conf create mode 100644 src/lxc/test_libvirtd_lxc.aug Regards, Amy

6 35

[libvirt] Resubmission #2: [PATCH 0/3] sVirt AppArmor security driver
by Jamie Strandboge 08 Oct '09

08 Oct '09

Resubmitting again based on feedback from this list. Notably, the driver has been reworked to use: xml = virDomainDefFormat(conn, vm->def, ... Then the helper program uses: ctl->def = virDomainDefParseString(... xmlStr ...); This simplifies adding new files. Part of this process involved adding support for kernel, initrd, loader, serial, and console. I have commented out code for hostdev (and pci would be easy enough), but it requires a patch to to move the _usbDevice struct fully into hostusb.h. I figure once the AppArmor driver is in, I can submit a patch for that. This iteration also properly works with attach/detach. While I use virDomainDefFormat() primarily, this does not work with attach because the definition is not updated with the information yet (neither def nor newDef), so I pass the disk file on the command line. For now this is ok since SetSecurityImageLabel currently only deals with a single file, but this will need to be adjusted depending on how the storage backend patch works out. I thought about creating a new def and inserting the disk into it and sending it off to virt-aa-helper, but I couldn't find an easy way to get an accurate virCapsPtr in security_apparmor.c. I had a similar problem with virt-aa-helper, but there I just created a fake virCapsPtr (which defaults to 'hvm', which was a compromise I was willing to make at this time since sVirt only supports qemu right now). The issues with virCapsPtr aside, the code is overall cleaner and much more easily extendable, so thanks for the excellent feedback. :) The patches are against up to date trunk, compiles with no warnings, 'make check' passes for the tests I added, and the code passes syntax-check. With the exception of the three calls to strncpy I had to convert to virStr*, this is the code that is currently in Ubuntu 9.10 (0.7.0-1ubuntu8). Among other things, that means 9.10 now has your preferred licensing. [PATCH 1] patch_1_reenable-nonfile-labels.patch (Updated based on prior feedback): When James Morris originally submitted his sVirt patches (as seen in libvirt 0.6.1), he did not require on disk labelling for virSecurityDomainRestoreImageLabel. A later commit[2] changed this behavior to assume on disk labelling, which halts implementations for path-based MAC systems such as AppArmor and TOMOYO where vm->def->seclabel is required to obtain the label. This patch simply adds the 'virDomainObjPtr vm' argument back to *RestoreImageLabel. [PATCH 2] patch_2_apparmor_driver.patch (updated and merged into one patch based on prior feedback): - Updates src/security.c for AppArmor - Adds security_apparmor.c, security_apparmor.h, virt-aa-helper.c and updates po/POTFILES.in. virt-aa-helper.c is a new binary which is used exclusively by the AppArmor security driver to manipulate AppArmor. - Adds tests for virt-aa-helper and the security driver. secaatest.c is identical to seclabeltest.c except it initializes the 'apparmor' driver instead of 'selinux'. These tests are integrated into 'make check' and pass. - Updates configure.in, Makefile.am, src/Makefile.am, tests/Makefile.am, and tools/Makefile.am for AppArmor. It is based on and should operate the same as the SELinux configuration. [PATCH 3] patch_3_docs.patch (Updated based on prior feedback): Updates docs/drvqemu.html.in for AppArmor and adds profile examples to examples/apparmor. Updated based on prior feedback. Jamie On Fri, 04 Sep 2009, Jamie Strandboge wrote: > This patch series implements the AppArmor security driver for sVirt. > This implementation was developed for the Ubuntu AppArmorLibvirtProfile > specification[1], but is general enough for any AppArmor deployment > (such as Ubuntu, *SUSE and Mandriva). > > This patch has seen quite a bit of real world testing in Ubuntu 9.10 > (our development release) in our 0.7.0-1ubuntu3 package. I did make a > few small changes after going through HACKING, but mostly I got the > tests going and added documentation. > > DESIGN > ------ > When a virtual machine is started, determine if a profile is currently > defined for the machine, and use it if available. If not, generate a new > profile for the machine based on a template, which is by default a very > restrictive profile allowing access to disk files, and anything else > needed to run, such as the pid, monitor and log files. > > Virtual machines should have a unique profile specific to that machine. > To ensure uniqueness, the profile name will be derived from the UUID of > the virtual machine. These profiles should be configurable, either by > adjusting the profile template for new machines, creating/modifying the > VM profile directly or through the use of AppArmor abstractions. This > will allow for administrators to fine-tune confinement for individual > machines if desired. > > If enabled at compile time, the sVirt security model will be activated > if AppArmor is available on the host OS and a profile for the libvirtd > daemon is loaded when libvirtd is started. > > libvirtd should not be allowed to create arbitrary profiles or modify > profiles directly, so as to not allow libvirtd to potentially (ie via a > security bug in libvirtd itself) bootstrap out of AppArmor confinement. > > Because root privileges are needed to manipulate AppArmor profiles, > qemu:///session will not be supported at this time, but the > implementation must allow for a confined libvirtd with qemu:///session > guests running unconfined. This can be revisited when AppArmor supports > per-user profiles. > > Please see the specification[1] for more details. -- Jamie Strandboge | http://www.canonical.com

3 12

[libvirt] [PATCH] cgroup -Werror fix
by Cole Robinson 08 Oct '09

08 Oct '09

Building with -Werror, I'm getting this error: make[3]: *** [libvirt_lxc-cgroup.o] Error 1 make[3]: *** Waiting for unfinished jobs.... cc1: warnings being treated as errors util/cgroup.c: In function 'virCgroupGetMemoryUsage': util/cgroup.c:743: error: passing argument 4 of 'virCgroupGetValueU64' from incompatible pointer type util/cgroup.c:381: note: expected 'long long unsigned int *' but argument is of type 'uint64_t *' Patch below fixes it, though I'm not familiar with the code. - Cole diff --git a/src/util/cgroup.c b/src/util/cgroup.c index 77f19a5..f728a2f 100644 --- a/src/util/cgroup.c +++ b/src/util/cgroup.c @@ -736,7 +736,7 @@ */ int virCgroupGetMemoryUsage(virCgroupPtr group, unsigned long *kb) { - uint64_t usage_in_bytes; + long long unsigned int usage_in_bytes; int ret; ret = virCgroupGetValueU64(group, VIR_CGROUP_CONTROLLER_MEMORY,

2 1

[libvirt] [PATCH] in vbox driver, interface type bridge should really be type ethernet
by Florian Vichot 08 Oct '09

08 Oct '09

If I understand correctly, the "bridge" type for an interface means libvirt is in charge of creating a tun device and adding it to the specified bridge in the <source bridge=".."> attribute. This is not what the (badly named IHMO) "Bridged Networking" mode in vbox does: all it does is read and write its packets on the specified interface. Which is what a type "ethernet" interface does in my opinion. I joined a quick patch for that, Pritesh could you check it ? Also, I was about to patch to add support for the proper "bridge" mode in the vbox driver (because it's what I need in the end), but I see no way of storing the fact that it's not just an "ethernet" interface using the vbox API, so that the config would persist over libvirtd restarts. There is always the possibility of naming the interface a special way like "br1_myTap", and assume any type "ethernet" interface with a device containing a "br1_" prefix would actually be a type "bridge" interface with an associated "vbox_br1" bridge. Would that be acceptable, or is it too hackish ? BTW, just to make sure, I can use pretty much anything as a <target> ie. ethN, panN etc in a type "ethernet" interface, right ? If <target> is specified, it's not going to create a tun and use that ? Thanks, Florian

4 6

[libvirt] [PATCH 0/6] correctly migrate paused VMs (and a bunch of bugfixes)
by Paolo Bonzini 07 Oct '09

07 Oct '09

Currently, the QEMU driver tries to migrate paused VMs, but the effect is that the migrated virtual machine will always run on the destination. Furthermore, the state is erroneously stored as paused, so that the driver is left confused and it is not possible to re-pause the VM without doing a (useless except to make the libvirtd state consistent) "virsh resume" first. While pausing is not very used, this feature is still nice-to-have in case for example you're migrating all VMs in emergency, or (when QEMU will have asynchronous notifications) if you want to run QEMU with the werror=stop option (*). This set of patches is structured as follows: 1) patches 1 and 2 fix two bugs that were introduced by the recent reorganization. 2) patch 3 fixes a bug that happens to be in the paths touched later on; 3) patch 4 adds the possibility to migrate a VM and leave it suspended on the destination VM. I added the feature for debugging and thought I might as well submit it for inclusion. Especially if the last two patches turn out to be a no-go (at least in their current form), this patch at least provides the infrastructure and a way to test it. 4) patches 5 and 6 actually fix the bug. Unfortunately, this requires a change to the RPC protocol. I'm not sure whether it is required the libvirtd's on both sides of the migration to have the same version, or rather if this is a deal breaker. The change to the protocol is in patch 5; the bug fix is in patch 6. Please review and ack (or, for the last two, nack). Paolo (*) While the state of the migrated VM is currently fetched at the beginning of migration, the communication happens at the end of it (from Perform to Finish), so this is ready for being improved in the future once QEMU provides asynchronous notifications.

5 26

[libvirt] [PATCH] Misc win32 build fixes
by Daniel P. Berrange 07 Oct '09

07 Oct '09

* daemon/Makefile.am: Fix missing sasl rule * src/datatypes.c: Add unistd.h to avoid gnulib bug * src/util/cgroup.c: Disable mntent if not available --- daemon/Makefile.am | 3 ++ src/datatypes.c | 1 + src/util/cgroup.c | 66 ++++++++++++++++++++++++++++++++++++++++++--------- 3 files changed, 58 insertions(+), 12 deletions(-) diff --git a/daemon/Makefile.am b/daemon/Makefile.am index 26347d6..84aab04 100644 --- a/daemon/Makefile.am +++ b/daemon/Makefile.am @@ -243,6 +243,9 @@ install-data-sasl: uninstall-data-sasl: rm -f $(DESTDIR)$(sysconfdir)/sasl2/libvirt.conf rmdir $(DESTDIR)$(sysconfdir)/sasl2/ +else +install-data-sasl: +uninstall-data-sasl: endif diff --git a/src/datatypes.c b/src/datatypes.c index 6741b9e..89ad309 100644 --- a/src/datatypes.c +++ b/src/datatypes.c @@ -20,6 +20,7 @@ */ #include <config.h> +#include <unistd.h> #include "datatypes.h" #include "virterror_internal.h" diff --git a/src/util/cgroup.c b/src/util/cgroup.c index 2e646fd..66297b9 100644 --- a/src/util/cgroup.c +++ b/src/util/cgroup.c @@ -13,7 +13,9 @@ #include <stdio.h> #include <stdint.h> #include <inttypes.h> +#ifdef HAVE_MNTENT_H #include <mntent.h> +#endif #include <fcntl.h> #include <string.h> #include <errno.h> @@ -68,7 +70,7 @@ void virCgroupFree(virCgroupPtr *group) VIR_FREE(*group); } - +#ifdef HAVE_MNTENT_H /* * Process /proc/mounts figuring out what controllers are * mounted and where @@ -233,6 +235,7 @@ static int virCgroupDetect(virCgroupPtr group) return rc; } +#endif static int virCgroupPathOfController(virCgroupPtr group, @@ -317,12 +320,12 @@ static int virCgroupGetValueStr(virCgroupPtr group, static int virCgroupSetValueU64(virCgroupPtr group, int controller, const char *key, - uint64_t value) + unsigned long long int value) { char *strval = NULL; int rc; - if (virAsprintf(&strval, "%" PRIu64, value) == -1) + if (virAsprintf(&strval, "%llu", value) == -1) return -ENOMEM; rc = virCgroupSetValueStr(group, controller, key, strval); @@ -339,12 +342,12 @@ static int virCgroupSetValueU64(virCgroupPtr group, static int virCgroupSetValueI64(virCgroupPtr group, int controller, const char *key, - int64_t value) + long long int value) { char *strval = NULL; int rc; - if (virAsprintf(&strval, "%" PRIi64, value) == -1) + if (virAsprintf(&strval, "%lld", value) == -1) return -ENOMEM; rc = virCgroupSetValueStr(group, controller, key, strval); @@ -357,7 +360,7 @@ static int virCgroupSetValueI64(virCgroupPtr group, static int virCgroupGetValueI64(virCgroupPtr group, int controller, const char *key, - int64_t *value) + long long int *value) { char *strval = NULL; int rc = 0; @@ -378,7 +381,7 @@ out: static int virCgroupGetValueU64(virCgroupPtr group, int controller, const char *key, - uint64_t *value) + unsigned long long int *value) { char *strval = NULL; int rc = 0; @@ -387,7 +390,7 @@ static int virCgroupGetValueU64(virCgroupPtr group, if (rc != 0) goto out; - if (sscanf(strval, "%" SCNu64, value) != 1) + if (virStrToLong_ull(strval, NULL, 10, value) < 0) rc = -EINVAL; out: VIR_FREE(strval); @@ -396,6 +399,7 @@ out: } +#ifdef HAVE_MNTENT_H static int virCgroupCpuSetInherit(virCgroupPtr parent, virCgroupPtr group) { int i; @@ -546,7 +550,7 @@ cleanup: virCgroupFree(&rootgrp); return rc; } - +#endif /** * virCgroupRemove: @@ -617,6 +621,7 @@ int virCgroupAddTask(virCgroupPtr group, pid_t pid) * * Returns 0 on success */ +#ifdef HAVE_MNTENT_H int virCgroupForDriver(const char *name, virCgroupPtr *group, int privileged, @@ -650,6 +655,16 @@ out: return rc; } +#else +int virCgroupForDriver(const char *name ATTRIBUTE_UNUSED, + virCgroupPtr *group ATTRIBUTE_UNUSED, + int privileged ATTRIBUTE_UNUSED, + int create ATTRIBUTE_UNUSED) +{ + /* Claim no support */ + return -ENXIO; +} +#endif /** @@ -661,6 +676,7 @@ out: * * Returns 0 on success */ +#ifdef HAVE_MNTENT_H int virCgroupForDomain(virCgroupPtr driver, const char *name, virCgroupPtr *group, @@ -684,6 +700,15 @@ int virCgroupForDomain(virCgroupPtr driver, return rc; } +#else +int virCgroupForDomain(virCgroupPtr driver ATTRIBUTE_UNUSED, + const char *name ATTRIBUTE_UNUSED, + virCgroupPtr *group ATTRIBUTE_UNUSED, + int create ATTRIBUTE_UNUSED) +{ + return -ENXIO; +} +#endif /** * virCgroupSetMemory: @@ -786,6 +811,7 @@ int virCgroupAllowDeviceMajor(virCgroupPtr group, char type, int major) * * Returns: 0 on success */ +#if defined(major) && defined(minor) int virCgroupAllowDevicePath(virCgroupPtr group, const char *path) { struct stat sb; @@ -801,6 +827,14 @@ int virCgroupAllowDevicePath(virCgroupPtr group, const char *path) major(sb.st_rdev), minor(sb.st_rdev)); } +#else +int virCgroupAllowDevicePath(virCgroupPtr group ATTRIBUTE_UNUSED, + const char *path ATTRIBUTE_UNUSED) +{ + return -ENOSYS; +} +#endif + /** * virCgroupDenyDevice: @@ -861,6 +895,7 @@ int virCgroupDenyDeviceMajor(virCgroupPtr group, char type, int major) return rc; } +#if defined(major) && defined(minor) int virCgroupDenyDevicePath(virCgroupPtr group, const char *path) { struct stat sb; @@ -876,26 +911,33 @@ int virCgroupDenyDevicePath(virCgroupPtr group, const char *path) major(sb.st_rdev), minor(sb.st_rdev)); } +#else +int virCgroupDenyDevicePath(virCgroupPtr group ATTRIBUTE_UNUSED, + const char *path ATTRIBUTE_UNUSED) +{ + return -ENOSYS; +} +#endif int virCgroupSetCpuShares(virCgroupPtr group, unsigned long long shares) { return virCgroupSetValueU64(group, VIR_CGROUP_CONTROLLER_CPU, - "cpu.shares", (uint64_t)shares); + "cpu.shares", shares); } int virCgroupGetCpuShares(virCgroupPtr group, unsigned long long *shares) { return virCgroupGetValueU64(group, VIR_CGROUP_CONTROLLER_CPU, - "cpu.shares", (uint64_t *)shares); + "cpu.shares", shares); } int virCgroupGetCpuacctUsage(virCgroupPtr group, unsigned long long *usage) { return virCgroupGetValueU64(group, VIR_CGROUP_CONTROLLER_CPUACCT, - "cpuacct.usage", (uint64_t *)usage); + "cpuacct.usage", usage); } int virCgroupSetFreezerState(virCgroupPtr group, const char *state) -- 1.6.2.5

2 3

[libvirt] [PATCH] Fix handling the return value of qemuMonitorSetBalloon
by Ryota Ozaki 07 Oct '09

07 Oct '09

Hi, The current code doesn't correctly handle the positive return value of qemuMonitorSetBalloon, i.e., the success case. The return value is directly passed to the return of qemudDomainSetMemory which treats a positive value as a failure. This patch fixes the defect. Thanks, ozaki-r >From f94b4c920ea30ae4a7f55fda343b4fff3bba040f Mon Sep 17 00:00:00 2001 From: Ryota Ozaki <ozaki.ryota(a)gmail.com> Date: Mon, 5 Oct 2009 03:14:58 +0900 Subject: [PATCH] Fix handling the return value of qemuMonitorSetBalloon * src/qemu/qemu_driver.c: The positive return value of qemuMonitorSetBalloon should be handled as a success --- src/qemu/qemu_driver.c | 13 ++++++++----- 1 files changed, 8 insertions(+), 5 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 95e672b..8cf9002 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -3020,17 +3020,20 @@ static int qemudDomainSetMemory(virDomainPtr dom, unsigned long newmem) { } if (virDomainIsActive(vm)) { - ret = qemuMonitorSetBalloon(vm, newmem); - /* Turn lack of balloon support into a fatal error */ - if (ret == 0) { + int r = qemuMonitorSetBalloon(vm, newmem); + if (r < 0) + goto cleanup; + + /* Lack of balloon support is a fatal error */ + if (r == 0) { qemudReportError(dom->conn, dom, NULL, VIR_ERR_NO_SUPPORT, "%s", _("cannot set memory of an active domain")); - ret = -1; + goto cleanup; } } else { vm->def->memory = newmem; - ret = 0; } + ret = 0; cleanup: if (vm) -- 1.6.5.rc2

3 2

[libvirt] [PATCH] lxc: implement missing APIs to control the amount of memory
by Ryota Ozaki 07 Oct '09

07 Oct '09

Hi, The patch implements the missing memory control APIs for lxc, i.e., domainGetMaxMemory, domainSetMaxMemory, domainSetMemory, and improves domainGetInfo to return proper amount of used memory via cgroup. And also it includes trivial two fixes; add missing virCgroupSetMemory in libvirt_private.syms and fix invalid lxcError argument in lxcDomainGetInfo. Thanks, ozaki-r >From 530e56db9367fcb325aeaa7fb0d52a008db2f19f Mon Sep 17 00:00:00 2001 From: Ryota Ozaki <ozaki.ryota(a)gmail.com> Date: Mon, 5 Oct 2009 02:07:04 +0900 Subject: [PATCH] lxc: implement missing APIs to control the amount of memory * src/libvirt_private.syms: Export virCgroupGetMemoryUsage and add missing virCgroupSetMemory * src/lxc/lxc_driver.c: Implement missing memory functions * src/util/cgroup.c, src/util/cgroup.h: Add the function to get used memory --- src/libvirt_private.syms | 2 + src/lxc/lxc_driver.c | 122 ++++++++++++++++++++++++++++++++++++++++++++-- src/util/cgroup.c | 20 ++++++++ src/util/cgroup.h | 1 + 4 files changed, 140 insertions(+), 5 deletions(-) diff --git a/src/libvirt_private.syms b/src/libvirt_private.syms index 49bbf96..3be3724 100644 --- a/src/libvirt_private.syms +++ b/src/libvirt_private.syms @@ -53,6 +53,8 @@ virCgroupForDriver; virCgroupRemove; virCgroupFree; virCgroupAddTask; +virCgroupSetMemory; +virCgroupGetMemoryUsage; virCgroupSetCpuShares; virCgroupGetCpuShares; virCgroupDenyDevicePath; diff --git a/src/lxc/lxc_driver.c b/src/lxc/lxc_driver.c index 5fb4105..8ec4ae4 100644 --- a/src/lxc/lxc_driver.c +++ b/src/lxc/lxc_driver.c @@ -450,6 +450,7 @@ static int lxcDomainGetInfo(virDomainPtr dom, if (!virDomainIsActive(vm) || driver->cgroup == NULL) { info->cpuTime = 0; + info->memory = vm->def->memory; } else { if (virCgroupForDomain(driver->cgroup, vm->def->name, &cgroup, 0) != 0) { lxcError(dom->conn, dom, VIR_ERR_INTERNAL_ERROR, @@ -458,13 +459,18 @@ static int lxcDomainGetInfo(virDomainPtr dom, } if (virCgroupGetCpuacctUsage(cgroup, &(info->cpuTime)) < 0) { - lxcError(dom->conn, dom, VIR_ERR_OPERATION_FAILED, ("cannot read cputime for domain")); + lxcError(dom->conn, dom, VIR_ERR_OPERATION_FAILED, + "%s", _("cannot read cputime for domain")); + goto cleanup; + } + if (virCgroupGetMemoryUsage(cgroup, &(info->memory)) < 0) { + lxcError(dom->conn, dom, VIR_ERR_OPERATION_FAILED, + "%s", _("cannot read memory usage for domain")); goto cleanup; } } info->maxMem = vm->def->maxmem; - info->memory = vm->def->memory; info->nrVirtCpu = 1; ret = 0; @@ -501,6 +507,112 @@ cleanup: return ret; } +/* Returns max memory in kb, 0 if error */ +static unsigned long lxcDomainGetMaxMemory(virDomainPtr dom) { + lxc_driver_t *driver = dom->conn->privateData; + virDomainObjPtr vm; + unsigned long ret = 0; + + lxcDriverLock(driver); + vm = virDomainFindByUUID(&driver->domains, dom->uuid); + lxcDriverUnlock(driver); + + if (!vm) { + char uuidstr[VIR_UUID_STRING_BUFLEN]; + virUUIDFormat(dom->uuid, uuidstr); + lxcError(dom->conn, dom, VIR_ERR_NO_DOMAIN, + _("no domain with matching uuid '%s'"), uuidstr); + goto cleanup; + } + + ret = vm->def->maxmem; + +cleanup: + if (vm) + virDomainObjUnlock(vm); + return ret; +} + +static int lxcDomainSetMaxMemory(virDomainPtr dom, unsigned long newmax) { + lxc_driver_t *driver = dom->conn->privateData; + virDomainObjPtr vm; + int ret = -1; + + lxcDriverLock(driver); + vm = virDomainFindByUUID(&driver->domains, dom->uuid); + lxcDriverUnlock(driver); + + if (!vm) { + char uuidstr[VIR_UUID_STRING_BUFLEN]; + virUUIDFormat(dom->uuid, uuidstr); + lxcError(dom->conn, dom, VIR_ERR_NO_DOMAIN, + _("no domain with matching uuid '%s'"), uuidstr); + goto cleanup; + } + + if (newmax < vm->def->memory) { + lxcError(dom->conn, dom, VIR_ERR_INVALID_ARG, + "%s", _("cannot set max memory lower than current memory")); + goto cleanup; + } + + vm->def->maxmem = newmax; + ret = 0; + +cleanup: + if (vm) + virDomainObjUnlock(vm); + return ret; +} + +static int lxcDomainSetMemory(virDomainPtr dom, unsigned long newmem) { + lxc_driver_t *driver = dom->conn->privateData; + virDomainObjPtr vm; + virCgroupPtr cgroup = NULL; + int ret = -1; + + lxcDriverLock(driver); + vm = virDomainFindByUUID(&driver->domains, dom->uuid); + lxcDriverUnlock(driver); + if (!vm) { + char uuidstr[VIR_UUID_STRING_BUFLEN]; + virUUIDFormat(dom->uuid, uuidstr); + lxcError(dom->conn, dom, VIR_ERR_NO_DOMAIN, + _("no domain with matching uuid '%s'"), uuidstr); + goto cleanup; + } + + if (newmem > vm->def->maxmem) { + lxcError(dom->conn, dom, VIR_ERR_INVALID_ARG, + "%s", _("cannot set memory higher than max memory")); + goto cleanup; + } + + if (virDomainIsActive(vm)) { + if (virCgroupForDomain(driver->cgroup, vm->def->name, &cgroup, 0) != 0) { + lxcError(dom->conn, dom, VIR_ERR_INTERNAL_ERROR, + _("Unable to get cgroup for %s¥n"), vm->def->name); + goto cleanup; + } + + if (virCgroupSetMemory(cgroup, newmem) < 0) { + lxcError(dom->conn, dom, VIR_ERR_OPERATION_FAILED, + "%s", _("cannot set memory for domain")); + goto cleanup; + } + } else { + vm->def->memory = newmem; + } + ret = 0; + +cleanup: + if (vm) + virDomainObjUnlock(vm); + if (cgroup) + virCgroupFree(&cgroup); + return ret; +} + static char *lxcDomainDumpXML(virDomainPtr dom, int flags) { @@ -2103,9 +2215,9 @@ static virDriver lxcDriver = { NULL, /* domainReboot */ lxcDomainDestroy, /* domainDestroy */ lxcGetOSType, /* domainGetOSType */ - NULL, /* domainGetMaxMemory */ - NULL, /* domainSetMaxMemory */ - NULL, /* domainSetMemory */ + lxcDomainGetMaxMemory, /* domainGetMaxMemory */ + lxcDomainSetMaxMemory, /* domainSetMaxMemory */ + lxcDomainSetMemory, /* domainSetMemory */ lxcDomainGetInfo, /* domainGetInfo */ NULL, /* domainSave */ NULL, /* domainRestore */ diff --git a/src/util/cgroup.c b/src/util/cgroup.c index 2e646fd..e56d293 100644 --- a/src/util/cgroup.c +++ b/src/util/cgroup.c @@ -702,6 +702,26 @@ int virCgroupSetMemory(virCgroupPtr group, unsigned long kb) } /** + * virCgroupGetMemoryUsage: + * + * @group: The cgroup to change memory for + * @kb: Pointer to returned used memory in kilobytes + * + * Returns: 0 on success + */ +int virCgroupGetMemoryUsage(virCgroupPtr group, unsigned long *kb) +{ + uint64_t usage_in_bytes; + int ret; + ret = virCgroupGetValueU64(group, + VIR_CGROUP_CONTROLLER_MEMORY, + "memory.usage_in_bytes", &usage_in_bytes); + if (ret == 0) + *kb = (unsigned long) usage_in_bytes >> 10; + return ret; +} + +/** * virCgroupDenyAllDevices: * * @group: The cgroup to deny devices for diff --git a/src/util/cgroup.h b/src/util/cgroup.h index aba56c6..aa36632 100644 --- a/src/util/cgroup.h +++ b/src/util/cgroup.h @@ -41,6 +41,7 @@ int virCgroupForDomain(virCgroupPtr driver, int virCgroupAddTask(virCgroupPtr group, pid_t pid); int virCgroupSetMemory(virCgroupPtr group, unsigned long kb); +int virCgroupGetMemoryUsage(virCgroupPtr group, unsigned long *kb); int virCgroupDenyAllDevices(virCgroupPtr group); -- 1.6.5.rc2

3 2

[libvirt] [PATCH] Make dnsmasq run from libvirtd reading /etc/ethers
by Satoru SATOH 07 Oct '09

07 Oct '09

The following small patch makes dnsmasq run from libvirtd reading /etc/ethers and enables static ip address assignment with it. Libvirtd already supports static ip address assignment by //ip/dhcp/host element in network xml definition file, however the number of the assignments looks limited by ARG_MAX or similar parameters when dnsmasq is exec-ed. This patch implements a workaround for this issue and the number of static ip assignments becomes only limited by dnsmasq (I'm not sure the exact number but it should large enough for most cases, I guess). Signed-off-by: Satoru SATOH <satoru.satoh(a)gmail.com> --- src/network/bridge_driver.c | 2 ++ 1 files changed, 2 insertions(+), 0 deletions(-) diff --git a/src/network/bridge_driver.c b/src/network/bridge_driver.c index 95bc810..725d340 100644 --- a/src/network/bridge_driver.c +++ b/src/network/bridge_driver.c @@ -391,6 +391,7 @@ networkBuildDnsmasqArgv(virConnectPtr conn, 1 + /* dnsmasq */ 1 + /* --strict-order */ 1 + /* --bind-interfaces */ + 1 + /* --read-ethers */ (network->def->domain?2:0) + /* --domain name */ 2 + /* --pid-file /var/run/libvirt/network/$NAME.pid */ 2 + /* --conf-file "" */ @@ -427,6 +428,7 @@ networkBuildDnsmasqArgv(virConnectPtr conn, */ APPEND_ARG(*argv, i++, "--strict-order"); APPEND_ARG(*argv, i++, "--bind-interfaces"); + APPEND_ARG(*argv, i++, "--read-ethers"); if (network->def->domain) { APPEND_ARG(*argv, i++, "--domain"); -- 1.6.2.5

2 2

[libvirt] [PATCH] Fix up "make check"
by Chris Lalancette 07 Oct '09

07 Oct '09

While running make check, I noticed that it was actually using the virsh binary from my system, in /usr/bin/virsh, and not the one that was just compiled. This is actually caused by a bug in Makefile.am, where we didn't update the PATH to include tools. While here, I also updated all of the scripts to properly define the srcdir, abs_top_srcdir, and abs_top_builddir environment variables. This is required if you want to be able to run the tests standalone (i.e. ./test instead of from make check). I've tested this on both RHEL-5 and Fedora-10 machines, and make check works on both, as does running the individual tests by hand. Signed-off-by: Chris Lalancette <clalance(a)redhat.com> --- tests/Makefile.am | 2 +- tests/cpuset | 12 ++++++++---- tests/daemon-conf | 18 +++++++++++------- tests/define-dev-segfault | 10 ++++++---- tests/int-overflow | 20 ++++++++++---------- tests/libvirtd-fail | 10 ++++++---- tests/libvirtd-pool | 10 ++++++---- tests/read-bufsiz | 15 ++++++++------- tests/read-non-seekable | 12 ++++++++---- tests/start | 10 ++++++---- tests/test_conf.sh | 14 ++++++++------ tests/undefine | 14 +++++++++----- tests/vcpupin | 12 ++++++++---- tests/virsh-all | 12 +++++++----- tests/virsh-synopsis | 11 +++++++---- 15 files changed, 109 insertions(+), 73 deletions(-) diff --git a/tests/Makefile.am b/tests/Makefile.am index abda4ad..4caca8f 100644 --- a/tests/Makefile.am +++ b/tests/Makefile.am @@ -150,7 +150,7 @@ TESTS += nodedevxml2xmltest TESTS += interfacexml2xmltest -path_add = $$abs_top_builddir/src$(PATH_SEPARATOR)$$abs_top_builddir/daemon +path_add = $$abs_top_builddir/src$(PATH_SEPARATOR)$$abs_top_builddir/daemon$(PATH_SEPARATOR)$$abs_top_builddir/tools # NB, automake < 1.10 does not provide the real # abs_top_{src/build}dir variables, so don't rely diff --git a/tests/cpuset b/tests/cpuset index eeb0ee2..9f43269 100755 --- a/tests/cpuset +++ b/tests/cpuset @@ -16,17 +16,21 @@ # You should have received a copy of the GNU General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. +test -z "$srcdir" && srcdir=$(pwd) +test -z "$abs_top_srcdir" && abs_top_srcdir=$(pwd)/.. +test -z "$abs_top_builddir" && abs_top_builddir=$(pwd)/.. + if test "$VERBOSE" = yes; then set -x - virsh --version + $abs_top_builddir/tools/virsh --version fi -. $srcdir/test-lib.sh +. "$srcdir/test-lib.sh" fail=0 # generate input -virsh --connect test:///default dumpxml 1 > xml || fail=1 +$abs_top_builddir/tools/virsh --connect test:///default dumpxml 1 > xml || fail=1 # require the presence of the string we'll transform grep '<vcpu>' xml > /dev/null || fail=1 @@ -34,7 +38,7 @@ grep '<vcpu>' xml > /dev/null || fail=1 sed "s/vcpu>/vcpu cpuset='aaa'>/" xml > xml-invalid || fail=1 # Require failure and a diagnostic. -virsh --connect test:///default define xml-invalid > out 2>&1 && fail=1 +$abs_top_builddir/tools/virsh --connect test:///default define xml-invalid > out 2>&1 && fail=1 cat <<\EOF > exp || fail=1 error: Failed to define domain from xml-invalid error: failed Xen syscall topology cpuset syntax error diff --git a/tests/daemon-conf b/tests/daemon-conf index 071497e..722fe4e 100755 --- a/tests/daemon-conf +++ b/tests/daemon-conf @@ -1,16 +1,20 @@ #!/bin/sh # Get coverage of libvirtd's config-parsing code. +test -z "$srcdir" && srcdir=$(pwd) +test -z "$abs_top_srcdir" && abs_top_srcdir=$(pwd)/.. +test -z "$abs_top_builddir" && abs_top_builddir=$(pwd)/.. + if test "$VERBOSE" = yes; then set -x - libvirtd --version + $abs_top_builddir/daemon/libvirtd --version fi -test -z "$srcdir" && srcdir=$(pwd) -test -z "$abs_top_srcdir" && abs_top_srcdir=$(pwd)/.. -. $srcdir/test-lib.sh +. "$srcdir/test-lib.sh" + +test -z "$CONFIG_HEADER" && CONFIG_HEADER="$abs_top_builddir/config.h" -grep '^#define WITH_QEMU 1' $CONFIG_HEADER > /dev/null || +grep '^#define WITH_QEMU 1' "$CONFIG_HEADER" > /dev/null || skip_test_ "configured without QEMU support" conf="$abs_top_srcdir/daemon/libvirtd.conf" @@ -40,7 +44,7 @@ while :; do esac # Run libvirtd, expecting it to fail. - libvirtd --config=$f 2> err && fail=1 + $abs_top_builddir/daemon/libvirtd --config=$f 2> err && fail=1 case $rhs in # '"'*) msg='should be a string';; @@ -73,7 +77,7 @@ sed 's,^log_outputs.*,log_outputs="3:file:'"$(pwd)/log"'",' tmp.conf > k \ mv k tmp.conf || fail=1 printf "running libvirtd with a valid config file ($sleep_secs seconds)\n" 1>&2 -libvirtd --pid-file=pid-file --config=tmp.conf > log 2>&1 & pid=$! +$abs_top_builddir/daemon/libvirtd --pid-file=pid-file --config=tmp.conf > log 2>&1 & pid=$! sleep $sleep_secs kill $pid diff --git a/tests/define-dev-segfault b/tests/define-dev-segfault index 4ae286f..e0b953a 100755 --- a/tests/define-dev-segfault +++ b/tests/define-dev-segfault @@ -2,13 +2,15 @@ # Exercise a bug whereby defining a valid domain could kill libvirtd. # The bug can also be exercised with a simple define/dumpxml pair to virsh. +test -z "$srcdir" && srcdir=$(pwd) +test -z "$abs_top_srcdir" && abs_top_srcdir=$(pwd)/.. +test -z "$abs_top_builddir" && abs_top_builddir=$(pwd)/.. + if test "$VERBOSE" = yes; then set -x - virsh --version + $abs_top_builddir/tools/virsh --version fi -test -z "$srcdir" && srcdir=$(pwd) -test -z "$abs_top_srcdir" && abs_top_srcdir=$(pwd)/.. . "$srcdir/test-lib.sh" fail=0 @@ -62,7 +64,7 @@ cat <<\EOF > D.xml || fail=1 EOF url=test:///default -virsh --connect "$url" 'define D.xml; dumpxml D' > out 2>&1 || fail=1 +$abs_top_builddir/tools/virsh --connect "$url" 'define D.xml; dumpxml D' > out 2>&1 || fail=1 cat > exp <<EOF || fail=1 Domain D defined from D.xml diff --git a/tests/int-overflow b/tests/int-overflow index ac3119b..c9f5de9 100755 --- a/tests/int-overflow +++ b/tests/int-overflow @@ -2,16 +2,16 @@ # Ensure that an invalid domain ID isn't interpreted as a valid one. # Before, an ID of 2^32+2 would be treated just like an ID of 2. -# Boilerplate code to set up a test directory, cd into it, -# and to ensure we remove it upon completion. -this_test_() { echo "./$0" | sed 's,.*/,,'; } -t_=$(this_test_)-$$ -init_cwd_=$(pwd) -trap 'st=$?; d='"$t_"'; - cd '"$init_cwd_"' && chmod -R u+rwx "$d" && rm -rf "$d" && exit $st' 0 -trap '(exit $?); exit $?' 1 2 13 15 -mkdir "$t_" || fail=1 -cd "$t_" || fail=1 +test -z "$srcdir" && srcdir=$(pwd) +test -z "$abs_top_srcdir" && abs_top_srcdir=$(pwd)/.. +test -z "$abs_top_builddir" && abs_top_builddir=$(pwd)/.. + +if test "$VERBOSE" = yes; then + set -x + $abs_top_builddir/tools/virsh --version +fi + +. "$srcdir/test-lib.sh" echo "error: failed to get domain '4294967298'" > exp || fail=1 echo domname 4294967298 | $abs_top_builddir/tools/virsh --quiet \ diff --git a/tests/libvirtd-fail b/tests/libvirtd-fail index eeb1ee6..c6b6876 100755 --- a/tests/libvirtd-fail +++ b/tests/libvirtd-fail @@ -1,18 +1,20 @@ #!/bin/sh # Ensure that libvirt fails when given nonexistent --config=FILE +test -z "$srcdir" && srcdir=$(pwd) +test -z "$abs_top_srcdir" && abs_top_srcdir=$(pwd)/.. +test -z "$abs_top_builddir" && abs_top_builddir=$(pwd)/.. + if test "$VERBOSE" = yes; then set -x - libvirtd --version + $abs_top_builddir/daemon/libvirtd --version fi -test -z "$srcdir" && srcdir=$(pwd) -test -z "$abs_top_srcdir" && abs_top_srcdir=$(pwd)/.. . "$srcdir/test-lib.sh" fail=0 -libvirtd --config=no-such-conf --timeout=5 2> log +$abs_top_builddir/daemon/libvirtd --config=no-such-conf --timeout=5 2> log RET=$? test "$RET" != "0" && exit 0 || exit 1 diff --git a/tests/libvirtd-pool b/tests/libvirtd-pool index 59da5a6..ca1db94 100755 --- a/tests/libvirtd-pool +++ b/tests/libvirtd-pool @@ -1,20 +1,22 @@ #!/bin/sh # Get coverage of virsh pool-define-as XML formatting +test -z "$srcdir" && srcdir=$(pwd) +test -z "$abs_top_srcdir" && abs_top_srcdir=$(pwd)/.. +test -z "$abs_top_builddir" && abs_top_builddir=$(pwd)/.. + if test "$VERBOSE" = yes; then set -x - virsh --version + $abs_top_builddir/tools/virsh --version fi -test -z "$srcdir" && srcdir=$(pwd) -test -z "$abs_top_srcdir" && abs_top_srcdir=$(pwd)/.. . "$srcdir/test-lib.sh" fail=0 pwd=$(pwd) || fail=1 -virsh --connect test:///default \ +$abs_top_builddir/tools/virsh --connect test:///default \ pool-define-as --print-xml \ P dir src-host /src/path /src/dev S /target-path \ 1>out 2>&1 diff --git a/tests/read-bufsiz b/tests/read-bufsiz index 7d53735..5baa7c5 100755 --- a/tests/read-bufsiz +++ b/tests/read-bufsiz @@ -16,20 +16,21 @@ # You should have received a copy of the GNU General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. +test -z "$srcdir" && srcdir=`pwd` +test -z "$abs_top_srcdir" && abs_top_srcdir=`pwd`/.. +test -z "$abs_top_builddir" && abs_top_builddir=`pwd`/.. + if test "$VERBOSE" = yes; then set -x - virsh --version + $abs_top_builddir/tools/virsh --version fi -test -z "$srcdir" && srcdir=`pwd` -test -z "$abs_top_srcdir" && abs_top_srcdir=`pwd`/.. - -. $srcdir/test-lib.sh +. "$srcdir/test-lib.sh" fail=0 # Output a valid definition, to be used as input. -virsh -c test:///default dumpxml 1 > xml || fail=1 +$abs_top_builddir/tools/virsh -c test:///default dumpxml 1 > xml || fail=1 for i in before after; do # The largest BUFSIZ I've seen is 128K. This is slightly larger. @@ -38,7 +39,7 @@ for i in before after; do # Append or prepend enough spaces to push the size over the limit: ( test $i = before && cat sp xml || cat xml sp ) > $in || fail=1 - virsh --connect test:///default define $in > out || fail=1 + $abs_top_builddir/tools/virsh --connect test:///default define $in > out || fail=1 printf "Domain test defined from $in\n\n" > exp || fail=1 compare exp out || fail=1 done diff --git a/tests/read-non-seekable b/tests/read-non-seekable index 8a7bdcd..59c2389 100755 --- a/tests/read-non-seekable +++ b/tests/read-non-seekable @@ -16,12 +16,16 @@ # You should have received a copy of the GNU General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. +test -z "$srcdir" && srcdir=`pwd` +test -z "$abs_top_srcdir" && abs_top_srcdir=`pwd`/.. +test -z "$abs_top_builddir" && abs_top_builddir=`pwd`/.. + if test "$VERBOSE" = yes; then set -x - virsh --version + $abs_top_builddir/tools/virsh --version fi -. $srcdir/test-lib.sh +. "$srcdir/test-lib.sh" fail=0 @@ -40,11 +44,11 @@ cat <<\EOF > dom </domain> EOF -virsh -c test:///default define dom > /dev/null || fail=1 +$abs_top_builddir/tools/virsh -c test:///default define dom > /dev/null || fail=1 mkfifo_or_skip_ fifo cat dom > fifo & -virsh -c test:///default define fifo > /dev/null || fail=1 +$abs_top_builddir/tools/virsh -c test:///default define fifo > /dev/null || fail=1 (exit $fail); exit $fail diff --git a/tests/start b/tests/start index f457d59..930a6d9 100755 --- a/tests/start +++ b/tests/start @@ -16,13 +16,15 @@ # You should have received a copy of the GNU General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. +test -z "$srcdir" && srcdir=$(pwd) +test -z "$abs_top_srcdir" && abs_top_srcdir=$(pwd)/.. +test -z "$abs_top_builddir" && abs_top_builddir=$(pwd)/.. + if test "$VERBOSE" = yes; then set -x - virsh --version + $abs_top_builddir/tools/virsh --version fi -test -z "$srcdir" && srcdir=$(pwd) -test -z "$abs_top_srcdir" && abs_top_srcdir=$(pwd)/.. . "$srcdir/test-lib.sh" fail=0 @@ -30,7 +32,7 @@ fail=0 test_url=test:///default # expect this to fail -virsh -c $test_url start test > out 2> err && fail=1 +$abs_top_builddir/tools/virsh -c $test_url start test > out 2> err && fail=1 # stdout gets a newline echo > exp || fail=1 diff --git a/tests/test_conf.sh b/tests/test_conf.sh index 62c2324..682f1f5 100755 --- a/tests/test_conf.sh +++ b/tests/test_conf.sh @@ -1,17 +1,19 @@ #!/bin/sh +test -z "$srcdir" && srcdir=$(pwd) +test -z "$abs_top_srcdir" && abs_top_srcdir=$(pwd)/.. +test -z "$abs_top_builddir" && abs_top_builddir=$(pwd)/.. +test -z "$abs_srcdir" && abs_srcdir=$(pwd) +test -z "$abs_builddir" && abs_builddir=$(pwd) + if test "$VERBOSE" = yes; then set -x - virsh --version + $abs_top_builddir/tools/virsh --version fi -. $srcdir/test-lib.sh +. "$srcdir/test-lib.sh" set -e -if test "x$abs_srcdir" = x; then - abs_srcdir=`pwd` - abs_builddir=`pwd` -fi fail=0 i=1 diff --git a/tests/undefine b/tests/undefine index fafdae7..48b0ad9 100755 --- a/tests/undefine +++ b/tests/undefine @@ -16,17 +16,21 @@ # You should have received a copy of the GNU General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. +test -z "$srcdir" && srcdir=$(pwd) +test -z "$abs_top_srcdir" && abs_top_srcdir=$(pwd)/.. +test -z "$abs_top_builddir" && abs_top_builddir=$(pwd)/.. + if test "$VERBOSE" = yes; then set -x - virsh --version + $abs_top_builddir/tools/virsh --version fi -. $srcdir/test-lib.sh +. "$srcdir/test-lib.sh" fail=0 # Attempt to undefine a running domain, by domain name. -virsh -q -c test:///default undefine test > out 2>&1 +$abs_top_builddir/tools/virsh -q -c test:///default undefine test > out 2>&1 test $? = 1 || fail=1 cat <<\EOF > exp || fail=1 error: Failed to undefine domain test @@ -35,7 +39,7 @@ EOF compare exp out || fail=1 # A different diagnostic when specifying a domain ID -virsh -q -c test:///default undefine 1 > out 2>&1 +$abs_top_builddir/tools/virsh -q -c test:///default undefine 1 > out 2>&1 test $? = 1 || fail=1 cat <<\EOF > exp || fail=1 error: a running domain like 1 cannot be undefined; @@ -44,7 +48,7 @@ EOF compare exp out || fail=1 # Succeed, now: first shut down, then undefine, both via name. -virsh -q -c test:///default 'shutdown test; undefine test' > out 2>&1 +$abs_top_builddir/tools/virsh -q -c test:///default 'shutdown test; undefine test' > out 2>&1 test $? = 0 || fail=1 cat <<\EOF > exp || fail=1 Domain test is being shutdown diff --git a/tests/vcpupin b/tests/vcpupin index 79b02dc..a72ad4c 100755 --- a/tests/vcpupin +++ b/tests/vcpupin @@ -16,17 +16,21 @@ # You should have received a copy of the GNU General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. +test -z "$srcdir" && srcdir=$(pwd) +test -z "$abs_top_srcdir" && abs_top_srcdir=$(pwd)/.. +test -z "$abs_top_builddir" && abs_top_builddir=$(pwd)/.. + if test "$VERBOSE" = yes; then set -x - virsh --version + $abs_top_builddir/tools/virsh --version fi -. $srcdir/test-lib.sh +. "$srcdir/test-lib.sh" fail=0 # Invalid syntax. -virsh --connect test:///default vcpupin test a 0,1 > out 2>&1 +$abs_top_builddir/tools/virsh --connect test:///default vcpupin test a 0,1 > out 2>&1 test $? = 1 || fail=1 cat <<\EOF > exp || fail=1 error: vcpupin: Invalid or missing vCPU number. @@ -35,7 +39,7 @@ EOF compare exp out || fail=1 # An out-of-range vCPU number deserves a diagnostic, too. -virsh --connect test:///default vcpupin test 100 0,1 > out 2>&1 +$abs_top_builddir/tools/virsh --connect test:///default vcpupin test 100 0,1 > out 2>&1 test $? = 1 || fail=1 cat <<\EOF > exp || fail=1 error: vcpupin: Invalid vCPU number. diff --git a/tests/virsh-all b/tests/virsh-all index 03ea466..81b3e57 100755 --- a/tests/virsh-all +++ b/tests/virsh-all @@ -16,27 +16,29 @@ # You should have received a copy of the GNU General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. +test -z "$srcdir" && srcdir=$(pwd) +test -z "$abs_top_srcdir" && abs_top_srcdir=$(pwd)/.. +test -z "$abs_top_builddir" && abs_top_builddir=$(pwd)/.. + if test "$VERBOSE" = yes; then set -x - virsh --version + $abs_top_builddir/tools/virsh --version fi -test -z "$srcdir" && srcdir=$(pwd) -test -z "$abs_top_srcdir" && abs_top_srcdir=$(pwd)/.. . "$srcdir/test-lib.sh" fail=0 test_url=test:///default -virsh -c $test_url help > cmds || framework_failure +$abs_top_builddir/tools/virsh -c $test_url help > cmds || framework_failure cmds=$(sed -n 's/^ $[^ ][^ ]*$ .*/\1/p' cmds) || framework_failure test -n "$cmds" || framework_failure for i in $cmds; do echo testing $i... 1>&2 # For now, just run the command and ignore output and exit status. - virsh -c $test_url $i < /dev/null > /dev/null 2>&1 + $abs_top_builddir/tools/virsh -c $test_url $i < /dev/null > /dev/null 2>&1 done (exit $fail); exit $fail diff --git a/tests/virsh-synopsis b/tests/virsh-synopsis index 24038dd..d72e887 100755 --- a/tests/virsh-synopsis +++ b/tests/virsh-synopsis @@ -16,24 +16,27 @@ # You should have received a copy of the GNU General Public License # along with this program. If not, see <http://www.gnu.org/licenses/>. +test -z "$srcdir" && srcdir=$(pwd) +test -z "$abs_top_srcdir" && abs_top_srcdir=$(pwd)/.. +test -z "$abs_top_builddir" && abs_top_builddir=$(pwd)/.. + if test "$VERBOSE" = yes; then set -x - virsh --version + $abs_top_builddir/tools/virsh --version fi -test -z "$srcdir" && srcdir=$(pwd) . "$srcdir/test-lib.sh" fail=0 test_url=test:///default -virsh -c $test_url help > cmds || framework_failure +$abs_top_builddir/tools/virsh -c $test_url help > cmds || framework_failure cmds=$(sed -n 's/^ $[^ ][^ ]*$ .*/\1/p' cmds) || framework_failure test -n "$cmds" || framework_failure for i in $cmds; do - virsh -c $test_url help $i > help || fail=1 + $abs_top_builddir/tools/virsh -c $test_url help $i > help || fail=1 grep -A1 '^ SYNOPSIS$' help > synopsis \ || { echo 1>&2 missing or invalid help SYNOPSIS for $i; fail=1; } sed -n 2p synopsis > s2 || framework_failure -- 1.6.0.6

2 2

[libvirt] boot.kernel.org support in libvirt
by Itamar Reis Peixoto 07 Oct '09

07 Oct '09

Hello guy's what you think to add bko support in libvirt/virt-manager for example boot from hard disk boot from network / pxe boot from kernel.org please look --> http://boot.kernel.org/ -- ------------ Itamar Reis Peixoto e-mail/msn: itamar(a)ispbrasil.com.br sip: itamar(a)ispbrasil.com.br skype: itamarjp icq: 81053601 +55 11 4063 5033 +55 34 3221 8599

2 1

Re: [libvirt] Xen: Domain.migrate() causes invalid listDomains() output
by Thomas Treutner 07 Oct '09

07 Oct '09

On Wednesday 07 October 2009 10:25:11 Thomas Treutner wrote: > Shall I provide more debug output? Here the debug output of libvirtd: http://tt.scripty.at/tmp/node01.txt.gz http://tt.scripty.at/tmp/node02.txt.gz At start, domain wp01 is @ node01, and wp02 @ node02. Then their placement is exchanged. After migration: node01:~# xm list Name ID Mem VCPUs State Time(s) Domain-0 0 256 2 r----- 37.9 wp02 2 512 2 -b---- 0.9 node02:~# xm list Name ID Mem VCPUs State Time(s) Domain-0 0 192 2 r----- 52.4 wp01 3 512 2 -b---- 1.0 virt-top etc. don't see wp01@node02 anymore. libvirt obviously returns ID 1 for it, which can't be resolved. As there is no alternative to int[] listDomains() (f.e., return UUIDs), I'm stuck here. NB: Doing migration with xm migrate works fine, libvirt isn't confused by that. -t-

1 0

[libvirt] cann't connect kvm hypervisor
by Liu, Zhentao 07 Oct '09

07 Oct '09

Hello, I have used virsh to connect my hypervisor, what is the reason? ------------------------------------------------------------------------------- root@forest:/# virsh -c qemu:///system list error: unable to connect to '/var/run/libvirt/libvirt-sock': Connection refused error: failed to connect to the hypervisor --------------------------------------------------------------------------------- Thanks Zhentao Liu

2 1

[libvirt] [PATCH] Support for IPv6 / multiple ifaces in interface_conf.[ch]
by Laine Stump 06 Oct '09

06 Oct '09

This patch updates the xml parsing and formatting, and the associated virInterfaceDef data structure to support IPv6, along the way adding support for multiple protocols per interface, and multiple IP addresses per protocol. Note that netcf appears to not accept defining both ipv4 and ipv6 on the same interface, and still can't report live config of IPv6 (or multiple IPv4 addresses), so the usefulness of this patch is limited until those items are fixed in netcf. This patch applies on top of commit 1ceabc5a152efa8892954cb4f45b85553edfee9f, which I submitted on Sept 28. --- src/conf/interface_conf.c | 308 ++++++++++++++++++++++++++++++++++----------- src/conf/interface_conf.h | 21 +++- 2 files changed, 247 insertions(+), 82 deletions(-) diff --git a/src/conf/interface_conf.c b/src/conf/interface_conf.c index b422464..24f111c 100644 --- a/src/conf/interface_conf.c +++ b/src/conf/interface_conf.c @@ -53,9 +53,32 @@ void virInterfaceBareDefFree(virInterfaceBareDefPtr def) { VIR_FREE(def); } +static +void virInterfaceIpDefFree(virInterfaceIpDefPtr def) { + if (def == NULL) + return; + VIR_FREE(def->address); + VIR_FREE(def->source); +} + +static +void virInterfaceProtocolDefFree(virInterfaceProtocolDefPtr def) { + int ii; + + if (def == NULL) + return; + for (ii = 0; ii < def->nips; ii++) { + virInterfaceIpDefFree(def->ips[ii]); + } + VIR_FREE(def->ips); + VIR_FREE(def->family); + VIR_FREE(def->gateway); + VIR_FREE(def); +} + void virInterfaceDefFree(virInterfaceDefPtr def) { - int i; + int i, pp; if (def == NULL) return; @@ -89,11 +112,11 @@ void virInterfaceDefFree(virInterfaceDefPtr def) break; } - VIR_FREE(def->proto.family); - VIR_FREE(def->proto.address); - VIR_FREE(def->proto.gateway); - VIR_FREE(def->proto.source); - + /* free all protos */ + for (pp = 0; pp < def->nprotos; pp++) { + virInterfaceProtocolDefFree(def->protos[pp]); + } + VIR_FREE(def->protos); VIR_FREE(def); } @@ -226,22 +249,22 @@ virInterfaceDefParseBondArpValid(virConnectPtr conn, xmlXPathContextPtr ctxt) { } static int -virInterfaceDefParseDhcp(virConnectPtr conn, virInterfaceDefPtr def, +virInterfaceDefParseDhcp(virConnectPtr conn, virInterfaceProtocolDefPtr def, xmlNodePtr dhcp, xmlXPathContextPtr ctxt) { xmlNodePtr save; char *tmp; int ret = 0; - def->proto.dhcp = 1; + def->dhcp = 1; save = ctxt->node; ctxt->node = dhcp; /* Not much to do in the current version */ tmp = virXPathString(conn, "string(./@peerdns)", ctxt); if (tmp) { if (STREQ(tmp, "yes")) - def->proto.peerdns = 1; + def->peerdns = 1; else if (STREQ(tmp, "no")) - def->proto.peerdns = 0; + def->peerdns = 0; else { virInterfaceReportError(conn, VIR_ERR_XML_ERROR, _("unknown dhcp peerdns value %s"), tmp); @@ -249,88 +272,209 @@ virInterfaceDefParseDhcp(virConnectPtr conn, virInterfaceDefPtr def, } VIR_FREE(tmp); } else - def->proto.peerdns = -1; + def->peerdns = -1; ctxt->node = save; return(ret); } static int -virInterfaceDefParseIp(virConnectPtr conn, virInterfaceDefPtr def, - xmlNodePtr ip ATTRIBUTE_UNUSED, xmlXPathContextPtr ctxt) { +virInterfaceDefParseIp(virConnectPtr conn, virInterfaceIpDefPtr def, + xmlXPathContextPtr ctxt) { int ret = 0; char *tmp; long l; - tmp = virXPathString(conn, "string(./ip[1]/@address)", ctxt); - def->proto.address = tmp; + tmp = virXPathString(conn, "string(./@address)", ctxt); + def->address = tmp; if (tmp != NULL) { - ret = virXPathLong(conn, "string(./ip[1]/@prefix)", ctxt, &l); + ret = virXPathLong(conn, "string(./@prefix)", ctxt, &l); if (ret == 0) - def->proto.prefix = (int) l; + def->prefix = (int) l; else if (ret == -2) { virInterfaceReportError(conn, VIR_ERR_XML_ERROR, "%s", _("Invalid ip address prefix value")); return(-1); } - tmp = virXPathString(conn, "string(./ip[1]/@source)", ctxt); - def->proto.source = tmp; + tmp = virXPathString(conn, "string(./@source)", ctxt); + def->source = tmp; } - tmp = virXPathString(conn, "string(./route[1]/@gateway)", ctxt); - def->proto.gateway = tmp; return(0); } static int -virInterfaceDefParseProtoIPv4(virConnectPtr conn, virInterfaceDefPtr def, +virInterfaceDefParseProtoIPv4(virConnectPtr conn, virInterfaceProtocolDefPtr def, xmlXPathContextPtr ctxt) { - xmlNodePtr dhcp, ip; - int ret = 0; + xmlNodePtr dhcp; + xmlNodePtr *ipNodes = NULL; + int nIpNodes, ii, ret = 0; + char *tmp; + + tmp = virXPathString(conn, "string(./route[1]/@gateway)", ctxt); + def->gateway = tmp; dhcp = virXPathNode(conn, "./dhcp", ctxt); if (dhcp != NULL) ret = virInterfaceDefParseDhcp(conn, def, dhcp, ctxt); + if (ret != 0) + return(ret); + + nIpNodes = virXPathNodeSet(conn, "./ip", ctxt, &ipNodes); + if (ipNodes == NULL) + return 0; + + if (VIR_ALLOC_N(def->ips, nIpNodes) < 0) { + virReportOOMError(conn); + ret = -1; + goto error; + } + + def->nips = 0; + for (ii = 0; ii < nIpNodes; ii++) { + virInterfaceIpDefPtr ip; + + if (VIR_ALLOC(ip) < 0) { + virReportOOMError(conn); + break; + } + + ctxt->node = ipNodes[ii]; + ret = virInterfaceDefParseIp(conn, ip, ctxt); + if (ret != 0) { + virInterfaceIpDefFree(ip); + break; + } + def->ips[def->nips++] = ip; + } + +error: + VIR_FREE(ipNodes); + return(ret); +} + +static int +virInterfaceDefParseProtoIPv6(virConnectPtr conn, virInterfaceProtocolDefPtr def, + xmlXPathContextPtr ctxt) { + xmlNodePtr dhcp, autoconf; + xmlNodePtr *ipNodes = NULL; + int nIpNodes, ii, ret = 0; + char *tmp; + + tmp = virXPathString(conn, "string(./route[1]/@gateway)", ctxt); + def->gateway = tmp; + + autoconf = virXPathNode(conn, "./autoconf", ctxt); + if (autoconf != NULL) + def->autoconf = 1; + + dhcp = virXPathNode(conn, "./dhcp", ctxt); + if (dhcp != NULL) + ret = virInterfaceDefParseDhcp(conn, def, dhcp, ctxt); if (ret != 0) return(ret); - ip = virXPathNode(conn, "./ip", ctxt); - if (ip != NULL) - ret = virInterfaceDefParseIp(conn, def, ip, ctxt); + nIpNodes = virXPathNodeSet(conn, "./ip", ctxt, &ipNodes); + if (ipNodes == NULL) + return 0; + + if (VIR_ALLOC_N(def->ips, nIpNodes) < 0) { + virReportOOMError(conn); + ret = -1; + goto error; + } + + def->nips = 0; + for (ii = 0; ii < nIpNodes; ii++) { + + virInterfaceIpDefPtr ip; + + if (VIR_ALLOC(ip) < 0) { + virReportOOMError(conn); + break; + } + + ctxt->node = ipNodes[ii]; + ret = virInterfaceDefParseIp(conn, ip, ctxt); + if (ret != 0) { + virInterfaceIpDefFree(ip); + break; + } + def->ips[def->nips++] = ip; + } + +error: + VIR_FREE(ipNodes); return(ret); } static int virInterfaceDefParseIfAdressing(virConnectPtr conn, virInterfaceDefPtr def, xmlXPathContextPtr ctxt) { - xmlNodePtr cur, save; - int ret; + xmlNodePtr save; + xmlNodePtr *protoNodes = NULL; + int nProtoNodes, pp, ret = 0; char *tmp; - cur = virXPathNode(conn, "./protocol[1]", ctxt); - if (cur == NULL) - return(0); save = ctxt->node; - ctxt->node = cur; - tmp = virXPathString(conn, "string(./@family)", ctxt); - if (tmp == NULL) { - virInterfaceReportError(conn, VIR_ERR_XML_ERROR, - "%s", _("protocol misses the family attribute")); - ret = -1; - goto done; + + nProtoNodes = virXPathNodeSet(conn, "./protocol", ctxt, &protoNodes); + if (nProtoNodes <= 0) { + /* no protocols is an acceptable outcome */ + return 0; } - if (STREQ(tmp, "ipv4")) { - def->proto.family = tmp; - ret = virInterfaceDefParseProtoIPv4(conn, def, ctxt); - } else { - virInterfaceReportError(conn, VIR_ERR_XML_ERROR, - _("unsupported protocol family '%s'"), tmp); + + if (VIR_ALLOC_N(def->protos, nProtoNodes) < 0) { + virReportOOMError(conn); ret = -1; - VIR_FREE(tmp); + goto error; } -done: + def->nprotos = 0; + for (pp = 0; pp < nProtoNodes; pp++) { + + virInterfaceProtocolDefPtr proto; + + if (VIR_ALLOC(proto) < 0) { + virReportOOMError(conn); + break; + } + + ctxt->node = protoNodes[pp]; + tmp = virXPathString(conn, "string(./@family)", ctxt); + if (tmp == NULL) { + virInterfaceReportError(conn, VIR_ERR_XML_ERROR, + "%s", _("protocol misses the family attribute")); + virInterfaceProtocolDefFree(proto); + ret = -1; + break; + } + proto->family = tmp; + if (STREQ(tmp, "ipv4")) { + ret = virInterfaceDefParseProtoIPv4(conn, proto, ctxt); + if (ret != 0) { + virInterfaceProtocolDefFree(proto); + break; + } + } else if (STREQ(tmp, "ipv6")) { + ret = virInterfaceDefParseProtoIPv6(conn, proto, ctxt); + if (ret != 0) { + virInterfaceProtocolDefFree(proto); + break; + } + } else { + virInterfaceReportError(conn, VIR_ERR_XML_ERROR, + _("unsupported protocol family '%s'"), tmp); + ret = -1; + virInterfaceProtocolDefFree(proto); + break; + } + def->protos[def->nprotos++] = proto; + } + +error: + VIR_FREE(protoNodes); ctxt->node = save; return(ret); @@ -1005,36 +1149,48 @@ virInterfaceVlanDefFormat(virConnectPtr conn, virBufferPtr buf, static int virInterfaceProtocolDefFormat(virConnectPtr conn ATTRIBUTE_UNUSED, virBufferPtr buf, const virInterfaceDefPtr def) { - char prefixStr[16]; - if (def->proto.family == NULL) - return(0); - virBufferVSprintf(buf, " <protocol family='%s'>\n", def->proto.family); - if (def->proto.dhcp) { - if (def->proto.peerdns == 0) - virBufferAddLit(buf, " <dhcp peerdns='no'/>\n"); - else if (def->proto.peerdns == 1) - virBufferAddLit(buf, " <dhcp peerdns='yes'/>\n"); - else - virBufferAddLit(buf, " <dhcp/>\n"); - } - if (def->proto.address != NULL) { - if (def->proto.prefix != 0) - snprintf(prefixStr, sizeof(prefixStr), "%d", def->proto.prefix); - - virBufferVSprintf(buf, " <ip address='%s'%s%s%s%s%s%s/>\n", - def->proto.address, - def->proto.prefix ? " prefix='" : "", - def->proto.prefix ? prefixStr : "", - def->proto.prefix ? "'" : "", - def->proto.source ? " source='" : "", - def->proto.source ? def->proto.source : "", - def->proto.source ? "'" : ""); - } - if (def->proto.gateway != NULL) { - virBufferVSprintf(buf, " <route gateway='%s'/>\n", - def->proto.gateway); + int pp, ii; + + for (pp = 0; pp < def->nprotos; pp++) { + + virBufferVSprintf(buf, " <protocol family='%s'>\n", def->protos[pp]->family); + + if (def->protos[pp]->autoconf) { + virBufferAddLit(buf, " <autoconf/>\n"); + } + + if (def->protos[pp]->dhcp) { + if (def->protos[pp]->peerdns == 0) + virBufferAddLit(buf, " <dhcp peerdns='no'/>\n"); + else if (def->protos[pp]->peerdns == 1) + virBufferAddLit(buf, " <dhcp peerdns='yes'/>\n"); + else + virBufferAddLit(buf, " <dhcp/>\n"); + } + + for (ii = 0; ii < def->protos[pp]->nips; ii++) { + if (def->protos[pp]->ips[ii]->address != NULL) { + + virBufferVSprintf(buf, " <ip address='%s'", + def->protos[pp]->ips[ii]->address); + if (def->protos[pp]->ips[ii]->prefix != 0) { + virBufferVSprintf(buf, " prefix='%d'", + def->protos[pp]->ips[ii]->prefix); + } + if (def->protos[pp]->ips[ii]->source != NULL) { + virBufferVSprintf(buf, " source='%s'", + def->protos[pp]->ips[ii]->source); + } + virBufferAddLit(buf, "/>\n"); + } + } + if (def->protos[pp]->gateway != NULL) { + virBufferVSprintf(buf, " <route gateway='%s'/>\n", + def->protos[pp]->gateway); + } + + virBufferAddLit(buf, " </protocol>\n"); } - virBufferAddLit(buf, " </protocol>\n"); return(0); } diff --git a/src/conf/interface_conf.h b/src/conf/interface_conf.h index df871aa..49acdd6 100644 --- a/src/conf/interface_conf.h +++ b/src/conf/interface_conf.h @@ -123,16 +123,25 @@ struct _virInterfaceVlanDef { char *devname; /* device name for vlan */ }; +typedef struct _virInterfaceIpDef virInterfaceIpDef; +typedef virInterfaceIpDef *virInterfaceIpDefPtr; +struct _virInterfaceIpDef { + char *address; /* ip address */ + int prefix; /* ip prefix */ + char *source; /* source of address - "device" or "config" (default) */ +}; + + typedef struct _virInterfaceProtocolDef virInterfaceProtocolDef; typedef virInterfaceProtocolDef *virInterfaceProtocolDefPtr; struct _virInterfaceProtocolDef { - char *family; /* ipv4 only right now */ + char *family; /* ipv4 or ipv6 */ int dhcp; /* use dhcp */ int peerdns; /* dhcp peerdns ? */ - char *address; /* ip address */ - int prefix; /* ip prefix */ + int autoconf; /* only useful if family is ipv6 */ + int nips; + virInterfaceIpDefPtr *ips; /* ptr to array of ips[nips] */ char *gateway; /* route gateway */ - char *source; /* source of address - "device" or "config" (default) */ }; @@ -152,8 +161,8 @@ struct _virInterfaceDef { virInterfaceBondDef bond; } data; - /* separated as we may allow multiple of those in the future */ - virInterfaceProtocolDef proto; + int nprotos; + virInterfaceProtocolDefPtr *protos; /* ptr to array of protos[nprotos] */ }; typedef struct _virInterfaceObj virInterfaceObj; -- 1.6.2.5

3 5

[libvirt] [PATCH] Support reporting live interface IP/netmask.
by Laine Stump 06 Oct '09

06 Oct '09

From: root <root(a)vlap.laine.org> This patch adds the flag VIR_INTERFACE_XML_INACTIVE to virInterfaceGetXMLDesc's flags. When it is *not* set (the default), the live interface info will be returned in the XML. in particular, the IP address(es) and netmask(s) will be retrieved by querying the device directly, rather than just reporting what's in the config file. The backend of this is in netcf's new ncf_if_xml_state() function. Any live interface ip address info in the xml will have the property "source" set to "device", eg: <ip address='10.24.0.1' prefix='24' source='device'/> Also, if an interface is currently inactive, no ip addresses will be returned, since an inactive interface device can't be queried for IP addresses (effectively it has none). A difference in the XML from previously - it is now acceptable to have both a dhcp *and* an ip node (or neither) within the protocol node. Before you had to have one or the other, but couldn't have both. Note that you need at least netcf 0.1.2 for this to build and work, and an upcoming release (patches submitted today) for it to work exactly as described above. --- include/libvirt/libvirt.h.in | 4 +++ src/conf/interface_conf.c | 62 ++++++++++++++++++++++------------------- src/conf/interface_conf.h | 1 + src/interface/netcf_driver.c | 8 ++++- src/libvirt.c | 15 +++++++--- tools/virsh.c | 10 +++++- 6 files changed, 63 insertions(+), 37 deletions(-) diff --git a/include/libvirt/libvirt.h.in b/include/libvirt/libvirt.h.in index 4e63e48..fd0c90b 100644 --- a/include/libvirt/libvirt.h.in +++ b/include/libvirt/libvirt.h.in @@ -922,6 +922,10 @@ virInterfacePtr virInterfaceLookupByMACString (virConnectPtr conn, const char* virInterfaceGetName (virInterfacePtr iface); const char* virInterfaceGetMACString (virInterfacePtr iface); +typedef enum { + VIR_INTERFACE_XML_INACTIVE = 1 /* dump inactive interface information */ +} virInterfaceXMLFlags; + char * virInterfaceGetXMLDesc (virInterfacePtr iface, unsigned int flags); virInterfacePtr virInterfaceDefineXML (virConnectPtr conn, diff --git a/src/conf/interface_conf.c b/src/conf/interface_conf.c index e646351..b422464 100644 --- a/src/conf/interface_conf.c +++ b/src/conf/interface_conf.c @@ -92,6 +92,7 @@ void virInterfaceDefFree(virInterfaceDefPtr def) VIR_FREE(def->proto.family); VIR_FREE(def->proto.address); VIR_FREE(def->proto.gateway); + VIR_FREE(def->proto.source); VIR_FREE(def); } @@ -272,6 +273,8 @@ virInterfaceDefParseIp(virConnectPtr conn, virInterfaceDefPtr def, "%s", _("Invalid ip address prefix value")); return(-1); } + tmp = virXPathString(conn, "string(./ip[1]/@source)", ctxt); + def->proto.source = tmp; } tmp = virXPathString(conn, "string(./route[1]/@gateway)", ctxt); def->proto.gateway = tmp; @@ -282,22 +285,19 @@ virInterfaceDefParseIp(virConnectPtr conn, virInterfaceDefPtr def, static int virInterfaceDefParseProtoIPv4(virConnectPtr conn, virInterfaceDefPtr def, xmlXPathContextPtr ctxt) { - xmlNodePtr cur; - int ret; + xmlNodePtr dhcp, ip; + int ret = 0; - cur = virXPathNode(conn, "./dhcp", ctxt); - if (cur != NULL) - ret = virInterfaceDefParseDhcp(conn, def, cur, ctxt); - else { - cur = virXPathNode(conn, "./ip", ctxt); - if (cur != NULL) - ret = virInterfaceDefParseIp(conn, def, cur, ctxt); - else { - virInterfaceReportError(conn, VIR_ERR_XML_ERROR, - "%s", _("interface miss dhcp or ip adressing")); - ret = -1; - } - } + dhcp = virXPathNode(conn, "./dhcp", ctxt); + if (dhcp != NULL) + ret = virInterfaceDefParseDhcp(conn, def, dhcp, ctxt); + + if (ret != 0) + return(ret); + + ip = virXPathNode(conn, "./ip", ctxt); + if (ip != NULL) + ret = virInterfaceDefParseIp(conn, def, ip, ctxt); return(ret); } @@ -1005,6 +1005,7 @@ virInterfaceVlanDefFormat(virConnectPtr conn, virBufferPtr buf, static int virInterfaceProtocolDefFormat(virConnectPtr conn ATTRIBUTE_UNUSED, virBufferPtr buf, const virInterfaceDefPtr def) { + char prefixStr[16]; if (def->proto.family == NULL) return(0); virBufferVSprintf(buf, " <protocol family='%s'>\n", def->proto.family); @@ -1015,20 +1016,23 @@ virInterfaceProtocolDefFormat(virConnectPtr conn ATTRIBUTE_UNUSED, virBufferAddLit(buf, " <dhcp peerdns='yes'/>\n"); else virBufferAddLit(buf, " <dhcp/>\n"); - } else { - /* theorically if we don't have dhcp we should have an address */ - if (def->proto.address != NULL) { - if (def->proto.prefix != 0) - virBufferVSprintf(buf, " <ip address='%s' prefix='%d'/>\n", - def->proto.address, def->proto.prefix); - else - virBufferVSprintf(buf, " <ip address='%s'/>\n", - def->proto.address); - } - if (def->proto.gateway != NULL) { - virBufferVSprintf(buf, " <route gateway='%s'/>\n", - def->proto.gateway); - } + } + if (def->proto.address != NULL) { + if (def->proto.prefix != 0) + snprintf(prefixStr, sizeof(prefixStr), "%d", def->proto.prefix); + + virBufferVSprintf(buf, " <ip address='%s'%s%s%s%s%s%s/>\n", + def->proto.address, + def->proto.prefix ? " prefix='" : "", + def->proto.prefix ? prefixStr : "", + def->proto.prefix ? "'" : "", + def->proto.source ? " source='" : "", + def->proto.source ? def->proto.source : "", + def->proto.source ? "'" : ""); + } + if (def->proto.gateway != NULL) { + virBufferVSprintf(buf, " <route gateway='%s'/>\n", + def->proto.gateway); } virBufferAddLit(buf, " </protocol>\n"); return(0); diff --git a/src/conf/interface_conf.h b/src/conf/interface_conf.h index bb9dce4..df871aa 100644 --- a/src/conf/interface_conf.h +++ b/src/conf/interface_conf.h @@ -132,6 +132,7 @@ struct _virInterfaceProtocolDef { char *address; /* ip address */ int prefix; /* ip prefix */ char *gateway; /* route gateway */ + char *source; /* source of address - "device" or "config" (default) */ }; diff --git a/src/interface/netcf_driver.c b/src/interface/netcf_driver.c index ca14fb0..b5c3664 100644 --- a/src/interface/netcf_driver.c +++ b/src/interface/netcf_driver.c @@ -326,7 +326,7 @@ cleanup: } static char *interfaceGetXMLDesc(virInterfacePtr ifinfo, - unsigned int flags ATTRIBUTE_UNUSED) + unsigned int flags) { struct interface_driver *driver = ifinfo->conn->interfacePrivateData; struct netcf_if *iface = NULL; @@ -342,7 +342,11 @@ static char *interfaceGetXMLDesc(virInterfacePtr ifinfo, goto cleanup; } - xmlstr = ncf_if_xml_desc(iface); + if ((flags & VIR_INTERFACE_XML_INACTIVE)) { + xmlstr = ncf_if_xml_desc(iface); + } else { + xmlstr = ncf_if_xml_state(iface); + } if (!xmlstr) { const char *errmsg, *details; int errcode = ncf_error(driver->netcf, &errmsg, &details); diff --git a/src/libvirt.c b/src/libvirt.c index bcb89e1..bd712dc 100644 --- a/src/libvirt.c +++ b/src/libvirt.c @@ -5915,10 +5915,17 @@ virInterfaceGetMACString(virInterfacePtr iface) /** * virInterfaceGetXMLDesc: * @iface: an interface object - * @flags: an OR'ed set of extraction flags, not used yet + * @flags: an OR'ed set of extraction flags. Current valid bits: + * + * VIR_INTERFACE_XML_INACTIVE - return the static configuration, + * suitable for use redefining the + * interface via virInterfaceDefineXML() * - * Provide an XML description of the interface. The description may be reused - * later to redefine the interface with virInterfaceDefineXML(). + * Provide an XML description of the interface. If + * VIR_INTERFACE_XML_INACTIVE is set, the description may be reused + * later to redefine the interface with virInterfaceDefineXML(). If it + * is not set, the ip address and netmask will be the current live + * setting of the interface, not the settings from the config files. * * Returns a 0 terminated UTF-8 encoded XML instance, or NULL in case of error. * the caller must free() the returned value. @@ -5935,7 +5942,7 @@ virInterfaceGetXMLDesc(virInterfacePtr iface, unsigned int flags) virLibInterfaceError(NULL, VIR_ERR_INVALID_INTERFACE, __FUNCTION__); return (NULL); } - if (flags != 0) { + if ((flags & ~VIR_INTERFACE_XML_INACTIVE) != 0) { virLibInterfaceError(iface, VIR_ERR_INVALID_ARG, __FUNCTION__); goto error; } diff --git a/tools/virsh.c b/tools/virsh.c index 3482389..77f1d70 100644 --- a/tools/virsh.c +++ b/tools/virsh.c @@ -2763,7 +2763,7 @@ cmdInterfaceEdit (vshControl *ctl, const vshCmd *cmd) char *doc = NULL; char *doc_edited = NULL; char *doc_reread = NULL; - int flags = 0; + int flags = VIR_INTERFACE_XML_INACTIVE; if (!vshConnectionUsability(ctl, ctl->conn, TRUE)) goto cleanup; @@ -3297,6 +3297,7 @@ static const vshCmdInfo info_interface_dumpxml[] = { static const vshCmdOptDef opts_interface_dumpxml[] = { {"interface", VSH_OT_DATA, VSH_OFLAG_REQ, gettext_noop("interface name or MAC address")}, + {"inactive", VSH_OT_BOOL, 0, gettext_noop("show inactive defined XML")}, {NULL, 0, 0, NULL} }; @@ -3306,6 +3307,11 @@ cmdInterfaceDumpXML(vshControl *ctl, const vshCmd *cmd) virInterfacePtr iface; int ret = TRUE; char *dump; + int flags = 0; + int inactive = vshCommandOptBool(cmd, "inactive"); + + if (inactive) + flags |= VIR_INTERFACE_XML_INACTIVE; if (!vshConnectionUsability(ctl, ctl->conn, TRUE)) return FALSE; @@ -3313,7 +3319,7 @@ cmdInterfaceDumpXML(vshControl *ctl, const vshCmd *cmd) if (!(iface = vshCommandOptInterface(ctl, cmd, NULL))) return FALSE; - dump = virInterfaceGetXMLDesc(iface, 0); + dump = virInterfaceGetXMLDesc(iface, flags); if (dump != NULL) { printf("%s", dump); free(dump); -- 1.6.2.5

5 16

[libvirt] [RFC] Multi-IQN proposal
by Shyam_Iyer＠Dell.com 06 Oct '09

06 Oct '09

Would this proposal be acceptable ? Example XML schema for an iSCSI storage pool created -- <pool type="iscsi"> <name>dell</name> <uuid>cf354733-b01b-8870-2040-94888640e24d</uuid> <capacity>0</capacity> <allocation>0</allocation> <available>0</available> - <source> <initiator iqnname = "<initiator IQN1>"> <initiator iqnname = "<initiator IQN2>"> ........................................ ........................................ <host name="<iSCSI target hostname or Target IP address>" /> <device path="<iSCSI Target IQN name>" /> </source> - <target> <path>/dev/disk/by-path</path> - <permissions> <mode>0700</mode> <owner>0</owner> <group>0</group> </permissions> </target> </pool> Each initiator iqn name can be parsed to create the unique sessions. This should solve the following possibilities -- * possibility of multiple IQNs for a single Guest * option for Guest's own BIOS & initiator to use these IQNs (iSCSI in Guest) * option for hypervisor's initiator to use these IQNs on behalf of the guest (Multi-IQN) Compile tested only. Needs beatification. diff --git a/src/storage_conf.c b/src/storage_conf.c index cb063cc..915e897 100644 --- a/src/storage_conf.c +++ b/src/storage_conf.c @@ -116,6 +116,7 @@ enum { VIR_STORAGE_POOL_SOURCE_DIR = (1<<2), VIR_STORAGE_POOL_SOURCE_ADAPTER = (1<<3), VIR_STORAGE_POOL_SOURCE_NAME = (1<<4), + VIR_STORAGE_POOL_SOURCE_INITIATOR_IQN = (1<<5), }; @@ -537,6 +538,33 @@ virStoragePoolDefParseXML(virConnectPtr conn, goto cleanup; } } + +/* Read the conf here */ + if (options->flags & VIR_STORAGE_POOL_SOURCE_INITATOR_IQN) { + int niqn, i; + + if ((niqn = virXPathNodeSet(conn, "./source/host/initiator", ctxt, &nodeset)) < 0) { + virStorageReportError(conn, VIR_ERR_XML_ERROR, + "%s", _("cannot extract storage pool source devices")); + goto cleanup; + } + if (VIR_ALLOC_N(ret->source.niqn, niqn) < 0) { + VIR_FREE(nodeset); + virReportOOMError(conn); + goto cleanup; + } + for (i = 0 ; i < niqn ; i++ ) { + xmlChar *name = xmlGetProp(nodeset[i], BAD_CAST "iqnname"); + if (path == NULL) { + VIR_FREE(nodeset); + virStorageReportError(conn, VIR_ERR_XML_ERROR, + "%s", _("missing storage pool source device path")); + goto cleanup; + } + ret->source.initiator[i].iqnname = (char *)name; + } + } if (options->flags & VIR_STORAGE_POOL_SOURCE_DEVICE) { xmlNodePtr *nodeset = NULL; int nsource, i; diff --git a/src/storage_conf.h b/src/storage_conf.h index 421d305..c2504be 100644 --- a/src/storage_conf.h +++ b/src/storage_conf.h @@ -202,7 +202,12 @@ struct _virStoragePoolSourceDevice { } geometry; }; - +/* Initiator Attributes */ +typedef virStoragePoolSourceInitiatorAttr *virStoragePoolSourceInitiatorAttrPtr; +struct virStoragePoolSourceInitiatorAttr { + char *iqnname; + /* We could put other initiator specific things here */ +} typedef struct _virStoragePoolSource virStoragePoolSource; typedef virStoragePoolSource *virStoragePoolSourcePtr; @@ -214,6 +219,9 @@ struct _virStoragePoolSource { int ndevice; virStoragePoolSourceDevicePtr devices; + /*And either one or more initiators*/ + virStoragePoolSourceInitiatorAttrPtr initiator; + /* Or a directory */ char *dir;

4 9

[libvirt] [RFC PATCH 0/6] Series short description
by Gerhard Stenzel 06 Oct '09

06 Oct '09

The following series of patches is a prototype implementation of a ebtables based MAC address filter. This is currently not intended for inclusion, but rather to get some feedback/comments about the approach and the implementation. At the moment, some defaults are set when libvirtd starts and when a domain is started or destroyed. The plan is to extend this filter capability to the API level and virsh command level. The plan is also to add more filtering features like VLAN filtering and QoS filtering. Thanks, Gerhard --- Gerhard Stenzel (6): remove currently unused parameter to pass make check test add MAC address based port filtering to qemu add MAC address based port filtering to libvirt add the ebtables wrapper to network driver add build support for ebtables wrapper add ebtables wrapper configure.in | 3 include/libvirt/libvirt.h.in | 14 + src/Makefile.am | 1 src/conf/network_conf.h | 10 src/driver.h | 14 + src/libvirt.c | 119 +++++ src/libvirt_private.syms | 26 + src/libvirt_public.syms | 3 src/network/bridge_driver.c | 175 ++++++++ src/qemu/qemu.conf | 3 src/qemu/qemu_conf.c | 14 + src/qemu/qemu_conf.h | 2 src/qemu/qemu_driver.c | 23 + src/test/test_driver.c | 3 src/util/ebtables.c | 961 ++++++++++++++++++++++++++++++++++++++++++ src/util/ebtables.h | 134 ++++++ 16 files changed, 1505 insertions(+), 0 deletions(-) create mode 100644 src/util/ebtables.c create mode 100644 src/util/ebtables.h --

2 11