- Devel - libvirt List Archives

[libvirt] [PATCHv3 00/16] Network configuration for lxc containers
by Cédric Bosdonnat 02 Dec '14

02 Dec '14

Hi all, Here is a rebased version of v2. Nothing changed except the 'since' version number in the added doc that has been updated. -- Cedric Cédric Bosdonnat (16): Forgot to cleanup ifname_guest* in domain network def parsing Domain conf: allow more than one IP address for net devices LXC: set IP addresses to veth devices in the container lxc conf2xml: convert IP addresses Allow network capabilities hostdev to configure IP addresses lxc conf2xml: convert ip addresses for hostdev NICs Domain network devices can now have a <gateway> element lxc conf2xml: convert lxc.network.ipv[46].gateway LXC: use the new net devices gateway definition LXC: honour network devices link state Wrong place for virDomainNetIpsFormat virNetDevSetIPv4Address: libnl implementation Renamed virNetDevSetIPv4Address to virNetDevSetIPAddress virNetDevAddRoute: implementation using netlink virNetDevClearIPv4Address: netlink implementation Renamed virNetDevClearIPv4Address to virNetDevClearIPAddress docs/formatdomain.html.in | 39 +++ docs/schemas/domaincommon.rng | 55 +++- src/conf/domain_conf.c | 214 +++++++++++++-- src/conf/domain_conf.h | 22 +- src/libvirt_private.syms | 7 +- src/lxc/lxc_container.c | 74 ++++- src/lxc/lxc_native.c | 173 ++++++++---- src/network/bridge_driver.c | 4 +- src/openvz/openvz_conf.c | 2 +- src/openvz/openvz_driver.c | 6 +- src/qemu/qemu_driver.c | 25 +- src/qemu/qemu_hotplug.c | 6 +- src/uml/uml_conf.c | 2 +- src/util/virnetdev.c | 305 ++++++++++++++++++--- src/util/virnetdev.h | 12 +- src/util/virnetlink.c | 38 +++ src/util/virnetlink.h | 2 + src/vbox/vbox_common.c | 3 +- src/xenconfig/xen_common.c | 15 +- src/xenconfig/xen_sxpr.c | 12 +- .../lxcconf2xmldata/lxcconf2xml-physnetwork.config | 4 + tests/lxcconf2xmldata/lxcconf2xml-physnetwork.xml | 3 + tests/lxcconf2xmldata/lxcconf2xml-simple.config | 4 + tests/lxcconf2xmldata/lxcconf2xml-simple.xml | 3 + tests/lxcxml2xmldata/lxc-hostdev.xml | 3 + tests/lxcxml2xmldata/lxc-idmap.xml | 3 + 26 files changed, 880 insertions(+), 156 deletions(-) -- 1.8.4.5

6 39

[libvirt] [PATCH v2 0/4] Precreate storage on migration
by Michal Privoznik 02 Dec '14

02 Dec '14

Version two of: https://www.redhat.com/archives/libvir-list/2014-November/msg01048.html diff to v1: -Switch from qemuDomainGetBlockInfoImpl() to qemuMonitorBlockStatsUpdateCapacity() -s/alloc/capacity/ -small fixes raised by Peter Michal Privoznik (4): storage: Introduce storagePoolLookupByPath qemuMonitorJSONBlockStatsUpdateCapacity: Don't skip disks qemu_migration: Send disk sizes to the other side qemu_migration: Precreate missing storage src/qemu/qemu_migration.c | 313 ++++++++++++++++++++++++++++++++++++++++--- src/qemu/qemu_monitor_json.c | 12 +- src/storage/storage_driver.c | 36 +++++ src/storage/storage_driver.h | 4 + 4 files changed, 342 insertions(+), 23 deletions(-) -- 2.0.4

2 8

[libvirt] [PATCH] nwfilter: fix crash when adding non-existing nwfilter
by Pavel Hrdina 02 Dec '14

02 Dec '14

Adding non-existing nwfilter to a network interface device without any nwfilter specified will crash libvirt daemon with segfault. The reason is that the nwfilter is not found an libvirt will try to restore old nwfilter configuration but there is no nwfilter specified. Signed-off-by: Pavel Hrdina <phrdina(a)redhat.com> --- src/conf/nwfilter_conf.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/conf/nwfilter_conf.c b/src/conf/nwfilter_conf.c index 074d745..317792e 100644 --- a/src/conf/nwfilter_conf.c +++ b/src/conf/nwfilter_conf.c @@ -2778,7 +2778,7 @@ virNWFilterObjFindByName(virNWFilterObjListPtr nwfilters, const char *name) for (i = 0; i < nwfilters->count; i++) { virNWFilterObjLock(nwfilters->objs[i]); - if (STREQ(nwfilters->objs[i]->def->name, name)) + if (STREQ_NULLABLE(nwfilters->objs[i]->def->name, name)) return nwfilters->objs[i]; virNWFilterObjUnlock(nwfilters->objs[i]); } -- 2.0.4

2 2

[libvirt] [PATCH] qemu: add tun devices to cgroup ACL
by rongqing.li＠windriver.com 02 Dec '14

02 Dec '14

From: Roy Li <rongqing.li(a)windriver.com> When systemd is used as init, CGroup will be enabled by default, and lead to virsh unable to start VM(which is using the tun/tap), since virsh can not access /dev/net/tun device Signed-off-by: Roy Li <rongqing.li(a)windriver.com> --- src/qemu/qemu.conf | 3 ++- src/qemu/qemu_cgroup.c | 2 +- src/qemu/test_libvirtd_qemu.aug.in | 1 + 3 files changed, 4 insertions(+), 2 deletions(-) diff --git a/src/qemu/qemu.conf b/src/qemu/qemu.conf index c6db568..225cd0f 100644 --- a/src/qemu/qemu.conf +++ b/src/qemu/qemu.conf @@ -272,7 +272,8 @@ # "/dev/null", "/dev/full", "/dev/zero", # "/dev/random", "/dev/urandom", # "/dev/ptmx", "/dev/kvm", "/dev/kqemu", -# "/dev/rtc","/dev/hpet", "/dev/vfio/vfio" +# "/dev/rtc","/dev/hpet", "/dev/vfio/vfio", +# "/dev/net/tun" #] # # RDMA migration requires the following extra files to be added to the list: diff --git a/src/qemu/qemu_cgroup.c b/src/qemu/qemu_cgroup.c index 0e94cae..7c15c07 100644 --- a/src/qemu/qemu_cgroup.c +++ b/src/qemu/qemu_cgroup.c @@ -45,7 +45,7 @@ static const char *const defaultDeviceACL[] = { "/dev/random", "/dev/urandom", "/dev/ptmx", "/dev/kvm", "/dev/kqemu", "/dev/rtc", "/dev/hpet", "/dev/vfio/vfio", - NULL, + "/dev/net/tun", NULL, }; #define DEVICE_PTY_MAJOR 136 #define DEVICE_SND_MAJOR 116 diff --git a/src/qemu/test_libvirtd_qemu.aug.in b/src/qemu/test_libvirtd_qemu.aug.in index 30fd27e..2ab7a8d 100644 --- a/src/qemu/test_libvirtd_qemu.aug.in +++ b/src/qemu/test_libvirtd_qemu.aug.in @@ -48,6 +48,7 @@ module Test_libvirtd_qemu = { "9" = "/dev/rtc" } { "10" = "/dev/hpet" } { "11" = "/dev/vfio/vfio" } + { "12" = "/dev/net/tun" } } { "save_image_format" = "raw" } { "dump_image_format" = "raw" } -- 2.1.0

2 1

[libvirt] [PATCH] conf: fix a comment typo in virDomainVideoDefaultRAM
by Wang Rui 02 Dec '14

02 Dec '14

Signed-off-by: Wang Rui <moon.wangrui(a)huawei.com> --- src/conf/domain_conf.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index fe9b986..2d81c37 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -10221,7 +10221,7 @@ virDomainVideoDefaultRAM(const virDomainDef *def, return 4 * 1024; case VIR_DOMAIN_VIDEO_TYPE_QXL: - /* QEMU use 64M as the minimal video video memory for qxl device */ + /* QEMU use 64M as the minimal video memory for qxl device */ return 64 * 1024; default: -- 1.7.12.4

2 1

[libvirt] [PATCH] libvirt.c: Move virDomainGetFSInfo to libvirt-domain.c
by Michal Privoznik 02 Dec '14

02 Dec '14

Since our big split of libvirt.c there are only a few functions living there. The majority was moved to corresponding subfile, e.g. domain functions were moved to libvirt-domain.c. However, the patches for virDomainGetFSInfo() and virDomainFSInfoFree() introduction were posted prior the big split and merged after. This resulted in two domain functions landing in wrong file. Move them to the correct one. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- src/libvirt-domain.c | 67 ++++++++++++++++++++++++++++++++++++++++++++++++++++ src/libvirt.c | 67 ---------------------------------------------------- 2 files changed, 67 insertions(+), 67 deletions(-) diff --git a/src/libvirt-domain.c b/src/libvirt-domain.c index 2b0defc..01bd9e4 100644 --- a/src/libvirt-domain.c +++ b/src/libvirt-domain.c @@ -11111,3 +11111,70 @@ virDomainStatsRecordListFree(virDomainStatsRecordPtr *stats) VIR_FREE(stats); } + + +/** + * virDomainGetFSInfo: + * @dom: a domain object + * @info: a pointer to a variable to store an array of mount points information + * @flags: extra flags; not used yet, so callers should always pass 0 + * + * Get a list of mapping information for each mounted file systems within the + * specified guest and the disks. + * + * Returns the number of returned mount points, or -1 in case of error. + * On success, the array of the information is stored into @info. The caller is + * responsible for calling virDomainFSInfoFree() on each array element, then + * calling free() on @info. On error, @info is set to NULL. + */ +int +virDomainGetFSInfo(virDomainPtr dom, + virDomainFSInfoPtr **info, + unsigned int flags) +{ + VIR_DOMAIN_DEBUG(dom, "info=%p, flags=%x", info, flags); + + virResetLastError(); + + virCheckDomainReturn(dom, -1); + virCheckReadOnlyGoto(dom->conn->flags, error); + virCheckNonNullArgGoto(info, error); + *info = NULL; + + if (dom->conn->driver->domainGetFSInfo) { + int ret = dom->conn->driver->domainGetFSInfo(dom, info, flags); + if (ret < 0) + goto error; + return ret; + } + + virReportUnsupportedError(); + + error: + virDispatchError(dom->conn); + return -1; +} + + +/** + * virDomainFSInfoFree: + * @info: pointer to a FSInfo object + * + * Frees all the memory occupied by @info. + */ +void +virDomainFSInfoFree(virDomainFSInfoPtr info) +{ + size_t i; + + if (!info) + return; + + VIR_FREE(info->mountpoint); + VIR_FREE(info->name); + VIR_FREE(info->fstype); + + for (i = 0; i < info->ndevAlias; i++) + VIR_FREE(info->devAlias[i]); + VIR_FREE(info->devAlias); +} diff --git a/src/libvirt.c b/src/libvirt.c index 86b0daa..d51c014 100644 --- a/src/libvirt.c +++ b/src/libvirt.c @@ -1400,73 +1400,6 @@ virConnectOpenAuth(const char *name, } -/** - * virDomainGetFSInfo: - * @dom: a domain object - * @info: a pointer to a variable to store an array of mount points information - * @flags: extra flags; not used yet, so callers should always pass 0 - * - * Get a list of mapping information for each mounted file systems within the - * specified guest and the disks. - * - * Returns the number of returned mount points, or -1 in case of error. - * On success, the array of the information is stored into @info. The caller is - * responsible for calling virDomainFSInfoFree() on each array element, then - * calling free() on @info. On error, @info is set to NULL. - */ -int -virDomainGetFSInfo(virDomainPtr dom, - virDomainFSInfoPtr **info, - unsigned int flags) -{ - VIR_DOMAIN_DEBUG(dom, "info=%p, flags=%x", info, flags); - - virResetLastError(); - - virCheckDomainReturn(dom, -1); - virCheckReadOnlyGoto(dom->conn->flags, error); - virCheckNonNullArgGoto(info, error); - *info = NULL; - - if (dom->conn->driver->domainGetFSInfo) { - int ret = dom->conn->driver->domainGetFSInfo(dom, info, flags); - if (ret < 0) - goto error; - return ret; - } - - virReportUnsupportedError(); - - error: - virDispatchError(dom->conn); - return -1; -} - - -/** - * virDomainFSInfoFree: - * @info: pointer to a FSInfo object - * - * Frees all the memory occupied by @info. - */ -void -virDomainFSInfoFree(virDomainFSInfoPtr info) -{ - size_t i; - - if (!info) - return; - - VIR_FREE(info->mountpoint); - VIR_FREE(info->name); - VIR_FREE(info->fstype); - - for (i = 0; i < info->ndevAlias; i++) - VIR_FREE(info->devAlias[i]); - VIR_FREE(info->devAlias); -} - - /** * virConnectClose: -- 2.0.4

2 1

[libvirt] [libvirt-python PATCH] override: iterate virDomainFSInfo.devAliases using ndevAliases
by Tomoki Sekiyama 02 Dec '14

02 Dec '14

Currently devAliases in virDomainFSInfo struct are iterated as a NULL-terminated list, but that is not guaranteed. It should use ndevAliases which stores the number of the items in devAliases. Signed-off-by: Tomoki Sekiyama <tomoki.sekiyama(a)hds.com> --- libvirt-override.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/libvirt-override.c b/libvirt-override.c index 7ef66b7..e51c44d 100644 --- a/libvirt-override.c +++ b/libvirt-override.c @@ -8334,8 +8334,8 @@ libvirt_virDomainGetFSInfo(PyObject *self ATTRIBUTE_UNUSED, PyObject *args) { PyObject *pyobj_domain; unsigned int flags; virDomainFSInfoPtr *fsinfo = NULL; - char **dev; int c_retval, i; + size_t j; PyObject *py_retval = NULL; if (!PyArg_ParseTuple(args, (char *)"Oi:virDomainFSInfo", @@ -8373,8 +8373,9 @@ libvirt_virDomainGetFSInfo(PyObject *self ATTRIBUTE_UNUSED, PyObject *args) { PyTuple_SetItem(info, 2, libvirt_constcharPtrWrap(fs->fstype)); PyTuple_SetItem(info, 3, alias); - for (dev = fs->devAlias; dev && *dev; dev++) - if (PyList_Append(alias, libvirt_constcharPtrWrap(*dev)) < 0) + for (j = 0; j < fs->ndevAlias; j++) + if (PyList_Append(alias, + libvirt_constcharPtrWrap(fs->devAlias[j])) < 0) goto cleanup; }

2 1

[libvirt] [PATCH] Fix race condition in qemuGetProcessInfo
by Eduardo Costa 01 Dec '14

01 Dec '14

There is a race condition between the fopen and fscanf calls in qemuGetProcessInfo. If fopen succeeds, there is a small possibility that the file no longer exists before reading from it. Now, if either fopen or fscanf calls fail, the function will behave just as only fopen had failed. --- src/qemu/qemu_driver.c | 43 ++++++++++++++++--------------------------- 1 file changed, 16 insertions(+), 27 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index ad5f70a..9ab1a81 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -1310,9 +1310,9 @@ qemuGetProcessInfo(unsigned long long *cpuTime, int *lastCpu, long *vm_rss, { char *proc; FILE *pidinfo; - unsigned long long usertime, systime; - long rss; - int cpu; + unsigned long long usertime = 0, systime = 0; + long rss = 0; + int cpu = 0; int ret; /* In general, we cannot assume pid_t fits in int; but /proc parsing @@ -1324,33 +1324,24 @@ qemuGetProcessInfo(unsigned long long *cpuTime, int *lastCpu, long *vm_rss, if (ret < 0) return -1; - if (!(pidinfo = fopen(proc, "r"))) { - /* VM probably shut down, so fake 0 */ - if (cpuTime) - *cpuTime = 0; - if (lastCpu) - *lastCpu = 0; - if (vm_rss) - *vm_rss = 0; - VIR_FREE(proc); - return 0; - } + pidinfo = fopen(proc, "r"); VIR_FREE(proc); /* See 'man proc' for information about what all these fields are. We're * only interested in a very few of them */ - if (fscanf(pidinfo, - /* pid -> stime */ - "%*d %*s %*c %*d %*d %*d %*d %*d %*u %*u %*u %*u %*u %llu %llu" - /* cutime -> endcode */ - "%*d %*d %*d %*d %*d %*d %*u %*u %ld %*u %*u %*u" - /* startstack -> processor */ - "%*u %*u %*u %*u %*u %*u %*u %*u %*u %*u %*d %d", - &usertime, &systime, &rss, &cpu) != 4) { + if (pidinfo) { + if (fscanf(pidinfo, + /* pid -> stime */ + "%*d %*s %*c %*d %*d %*d %*d %*d %*u %*u %*u %*u %*u %llu %llu" + /* cutime -> endcode */ + "%*d %*d %*d %*d %*d %*d %*u %*u %ld %*u %*u %*u" + /* startstack -> processor */ + "%*u %*u %*u %*u %*u %*u %*u %*u %*u %*u %*d %d", + &usertime, &systime, &rss, &cpu) != 4) { + VIR_WARN("cannot parse process status data"); + usertime = systime = 0; rss = 0; cpu = 0; + } VIR_FORCE_FCLOSE(pidinfo); - VIR_WARN("cannot parse process status data"); - errno = -EINVAL; - return -1; } /* We got jiffies @@ -1374,8 +1365,6 @@ qemuGetProcessInfo(unsigned long long *cpuTime, int *lastCpu, long *vm_rss, VIR_DEBUG("Got status for %d/%d user=%llu sys=%llu cpu=%d rss=%ld", (int) pid, tid, usertime, systime, cpu, rss); - VIR_FORCE_FCLOSE(pidinfo); - return 0; } -- 1.7.9.5

2 1

[libvirt] [PATCH] qemu: Use AAVMF as UEFI default on ARMs
by Michal Privoznik 01 Dec '14

01 Dec '14

One of the things I've spotted while playing with OVMF is, that on different architectures the OVMF firmware has different names. Like arm and aarch64: it's AAVMF_{CODE,VARS}.fd. We should reflect that. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- src/qemu/qemu.conf | 5 ++++- src/qemu/qemu_conf.c | 9 +++++++-- 2 files changed, 11 insertions(+), 3 deletions(-) diff --git a/src/qemu/qemu.conf b/src/qemu/qemu.conf index c6db568..36cacf9 100644 --- a/src/qemu/qemu.conf +++ b/src/qemu/qemu.conf @@ -507,5 +507,8 @@ # a list of strings when a single item is in form of: # ${PATH_TO_UEFI_FW}:${PATH_TO_UEFI_VARS}. # Later, when libvirt creates per domain variable store, this -# list is searched for the master image. +# list is searched for the master image. The UEFI firmware can be +# called differently on different architectures. For instance, +# it's OVMF on x86_64 and i686, but it's AAVMF on aarch64. The +# libvirt default follows this scheme. #nvram = [ "/usr/share/OVMF/OVMF_CODE.fd:/usr/share/OVMF/OVMF_VARS.fd" ] diff --git a/src/qemu/qemu_conf.c b/src/qemu/qemu_conf.c index 4764bef..6c038c3 100644 --- a/src/qemu/qemu_conf.c +++ b/src/qemu/qemu_conf.c @@ -107,8 +107,13 @@ void qemuDomainCmdlineDefFree(qemuDomainCmdlineDefPtr def) VIR_FREE(def); } -#define VIR_QEMU_LOADER_FILE_PATH "/usr/share/OVMF/OVMF_CODE.fd" -#define VIR_QEMU_NVRAM_FILE_PATH "/usr/share/OVMF/OVMF_VARS.fd" +#if defined(__arm__) || defined(__aarch64__) +# define VIR_QEMU_LOADER_FILE_PATH "/usr/share/AAVMF/AAVMF_CODE.fd" +# define VIR_QEMU_NVRAM_FILE_PATH "/usr/share/AAVMF/AAVMF_VARS.fd" +#else +# define VIR_QEMU_LOADER_FILE_PATH "/usr/share/OVMF/OVMF_CODE.fd" +# define VIR_QEMU_NVRAM_FILE_PATH "/usr/share/OVMF/OVMF_VARS.fd" +#endif virQEMUDriverConfigPtr virQEMUDriverConfigNew(bool privileged) { -- 2.0.4

3 2

[libvirt] [PATCH 0/4] Resolve some Coverity issues
by John Ferlan 01 Dec '14

01 Dec '14

The issues found in the first 3 patches aren't "usually" seen, but I've had them sitting around awhile and figured they should be posted anyway since the last issue was seen from a recent change... John Ferlan (4): hotplug: Resolve Coverity FORWARD_NULL virhook: Resolve Coverity NULL_RETURNS domain_conf: Resolve Coverity CHECKED_RETURN libxl: Resolve Coverity CHECKED_RETURN src/conf/domain_conf.c | 2 +- src/libxl/libxl_migration.c | 3 +-- src/qemu/qemu_hotplug.c | 12 ++++++------ src/util/virhook.c | 3 ++- 4 files changed, 10 insertions(+), 10 deletions(-) -- 1.9.3

2 5