- Devel - libvirt List Archives

[libvirt] [PATCH] util: avoid getting stuck on macvtapN name created outside libvirt
by Laine Stump 28 Mar '16

28 Mar '16

After the patches that added tracking of in-use macvtap names (commit 370608, first appearing in libvirt-1.3.2), if the function to allocate a new macvtap device came to a device name created outside libvirt, it would retry the same device name MACVLAN_MAX_I (8191) times before finally giving up in failure. The problem was that virBitmapNextClearBit was always being called with "0" rather than the value most recently checked (which would increment each time through the loop). Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1321546 --- I'm not sure how this bug snuck in, because this is one of the scenarios I tested. The only thing I can think of is that I screwed up somehow during a rebase after testing and before posting/pushing (or possibly I had tested these cases in an earlier iteration, but forgot to do so with the final version) :-/ Sigh. src/util/virnetdevmacvlan.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/util/virnetdevmacvlan.c b/src/util/virnetdevmacvlan.c index 2409113..d755b93 100644 --- a/src/util/virnetdevmacvlan.c +++ b/src/util/virnetdevmacvlan.c @@ -131,7 +131,7 @@ virNetDevMacVLanReserveID(int id, unsigned int flags, } if ((id < 0 || nextFree) && - (id = virBitmapNextClearBit(bitmap, 0)) < 0) { + (id = virBitmapNextClearBit(bitmap, id)) < 0) { virReportError(VIR_ERR_INTERNAL_ERROR, _("no unused %s names available"), (flags & VIR_NETDEV_MACVLAN_CREATE_WITH_TAP) ? -- 2.5.5

2 1

[libvirt] [PATCH 0/2] vz: fix handling inputs in config
by Nikolay Shirokovskiy 28 Mar '16

28 Mar '16

Input devices are not stored in SDK but we report them on dumpxml if vnc is on. Thus we need to handle input devices on define xml as well. The latter part have some problems. Let's fix them. Nikolay Shirokovskiy (2): vz: add default input devices on post parse vz: fix config input device check src/vz/vz_driver.c | 26 +++++++++++++++++++++++++- src/vz/vz_sdk.c | 27 ++++++++++++++++++--------- 2 files changed, 43 insertions(+), 10 deletions(-) -- 1.8.3.1

1 2

[libvirt] [PATCH] qemu: ignore EACCES error on hot unplug of PCI devices with VFIO
by Kothapally Madhu Pavan 28 Mar '16

28 Mar '16

When the hostdev managed=true, the VFIO devices are reattached to the host driver before coming to qemuTeardownHostdevCgroup(). So, the /dev/vfio/<iommu> path would not exist. We ignore the EACCES error here. Signed-off-by: Kothapally Madhu Pavan <kmp(a)linux.vnet.ibm.com> --- src/qemu/qemu_cgroup.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/qemu/qemu_cgroup.c b/src/qemu/qemu_cgroup.c index c76d585..2850346 100644 --- a/src/qemu/qemu_cgroup.c +++ b/src/qemu/qemu_cgroup.c @@ -427,7 +427,7 @@ qemuTeardownHostdevCgroup(virDomainObjPtr vm, VIR_DEBUG("Cgroup deny %s for PCI device assignment", path); rv = virCgroupDenyDevicePath(priv->cgroup, path, - VIR_CGROUP_DEVICE_RWM, false); + VIR_CGROUP_DEVICE_RWM, dev->managed); virDomainAuditCgroupPath(vm, priv->cgroup, "deny", path, "rwm", rv == 0); if (rv < 0)

2 1

[libvirt] [PATCH] storage: revert bogus raw volumes support check
by Roman Bogorodskiy 28 Mar '16

28 Mar '16

Commits bb5f2dc and 6682d62 added a check like: if (vol->target.format != VIR_STORAGE_FILE_RAW) to the CreateVol method for zfs and logical storage backends respectively. However, this check is bogus because virStorageVolDefParseXML() in conf/storage_conf.c sets target.format only if volOptions in virStoragePoolTypeInfo has formatFromString set, and that's not the case for zfs and logical backends. So the check always fails and breaks volume creation. Thus, revert those commits. --- src/storage/storage_backend_logical.c | 6 ------ src/storage/storage_backend_zfs.c | 6 ------ 2 files changed, 12 deletions(-) diff --git a/src/storage/storage_backend_logical.c b/src/storage/storage_backend_logical.c index 39e8b80..90a194e 100644 --- a/src/storage/storage_backend_logical.c +++ b/src/storage/storage_backend_logical.c @@ -912,12 +912,6 @@ virStorageBackendLogicalCreateVol(virConnectPtr conn, struct stat sb; bool created = false; - if (vol->target.format != VIR_STORAGE_FILE_RAW) { - virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", - _("only RAW volumes are supported by this storage pool")); - return -1; - } - if (vol->target.encryption != NULL) { virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", _("storage pool does not support encrypted " diff --git a/src/storage/storage_backend_zfs.c b/src/storage/storage_backend_zfs.c index 5238ecc..c66a808 100644 --- a/src/storage/storage_backend_zfs.c +++ b/src/storage/storage_backend_zfs.c @@ -301,12 +301,6 @@ virStorageBackendZFSCreateVol(virConnectPtr conn ATTRIBUTE_UNUSED, int ret = -1; int volmode_needed = -1; - if (vol->target.format != VIR_STORAGE_FILE_RAW) { - virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", - _("only RAW volumes are supported by this storage pool")); - return -1; - } - if (vol->target.encryption != NULL) { virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", _("storage pool does not support encrypted " -- 2.7.2

2 2

[libvirt] "Only raw volumes are supported"
by Richard Laager 27 Mar '16

27 Mar '16

I think the two commits below need to be reverted. This is now preventing me from creating ZFS volumes (and presumably would block creating LVM volumes too). I don't know how this passed my testing before and isn't working now, but that seems to be the case. When I run `virsh vol-create-as --pool=zfs --name=sparse1 --capacity=1G --format=raw`, target.format is 0 (VIR_STORAGE_FILE_NONE). Also note that the equivalent code in RBD came in a commit with more code. So I'm thinking I was on the wrong track with these two patches. commit bb5f2dc91f43f48ff726aa589bdac546738d2980 Author: Richard Laager <rlaager(a)wiktel.com> Date: Tue Mar 15 01:17:35 2016 -0500 zfs: Only raw volumes are supported commit 6682d6219d4595b51f06867e0c30e7efd745f682 Author: Richard Laager <rlaager(a)wiktel.com> Date: Tue Mar 15 01:17:34 2016 -0500 logical: Only raw volumes are supported -- Richard

3 2

[libvirt] I am Xin and I am a newcomer
by fu du 26 Mar '16

26 Mar '16

Hello : My name is Dangxin Hou, my email address is conquergmial(a)gmail.com and my nickname is Xin on the #virt IRC channel at OFTC. I got my bachelor’s degree from Xidian University at July 2015, and majored in software engineering. I am a postgraduate student studying in Peking University, major in computer science. I am interested in the project “Making virsh more bash like”.The project will add an intelligent auto-completer into virsh to auto complete the right objects that a command works with. First arrived in your land, please give most concorn!

2 1

[libvirt] [PATCH 0/7] qemu: block job/disk related cleanups
by Peter Krempa 25 Mar '16

25 Mar '16

I've accumulated quite a few patches in my branch to add support for allowing to use 'detect zeroes' for the block copy job. These are pathches that make sense to be posted separately. Peter Krempa (7): qemu: hotplug: Assume support for -device in qemuDomainAttachSCSIDisk qemu: monitor: Drop qemuMonitorAttachDrive and leaves in call tree qemu: monitor: Remove JSON impls of drive_add and drive_del qemu: Kill qemuDiskPathToAlias qemu: Split image access revoking from qemuDomainPrepareDiskChainElement qemu: Refactor qemuDomainPrepareDiskChainElement qemu: domain: Move and export qemuDomainDiskChainElement(Prepare|Revoke) src/qemu/qemu_alias.c | 16 ++++ src/qemu/qemu_alias.h | 2 + src/qemu/qemu_domain.c | 81 +++++++++++++++++++ src/qemu/qemu_domain.h | 11 +++ src/qemu/qemu_driver.c | 187 ++++++++++++------------------------------- src/qemu/qemu_hotplug.c | 63 +++------------ src/qemu/qemu_monitor.c | 35 +++----- src/qemu/qemu_monitor.h | 6 -- src/qemu/qemu_monitor_json.c | 29 ------- src/qemu/qemu_monitor_json.h | 6 -- src/qemu/qemu_monitor_text.c | 102 +---------------------- src/qemu/qemu_monitor_text.h | 5 -- 12 files changed, 187 insertions(+), 356 deletions(-) -- 2.7.3

2 10

[libvirt] [PATCH] test: Fix typo in testutils.h header guard
by Christophe Fergeau 25 Mar '16

25 Mar '16

This changes __VIT_TEST_UTILS_H__ to __VIR_TEST_UTILS_H__ --- tests/testutils.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/tests/testutils.h b/tests/testutils.h index 752fa52..058be55 100644 --- a/tests/testutils.h +++ b/tests/testutils.h @@ -20,8 +20,8 @@ * Karel Zak <kzak(a)redhat.com> */ -#ifndef __VIT_TEST_UTILS_H__ -# define __VIT_TEST_UTILS_H__ +#ifndef __VIR_TEST_UTILS_H__ +# define __VIR_TEST_UTILS_H__ # include <stdio.h> # include "viralloc.h" @@ -145,4 +145,4 @@ int testCompareDomXML2XMLFiles(virCapsPtr caps, const void *opaque, unsigned int parseFlags); -#endif /* __VIT_TEST_UTILS_H__ */ +#endif /* __VIR_TEST_UTILS_H__ */ -- 2.5.5

2 1

[libvirt] [PATCH] storage/rbd: Use correct printf-modifier for uint64
by Christophe Fergeau 25 Mar '16

25 Mar '16

%zu is for size_t variables, not uint64 ones. This causes a warning when building on a 32 bit linux. --- src/storage/storage_backend_rbd.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/storage/storage_backend_rbd.c b/src/storage/storage_backend_rbd.c index 7d04b39..d895cc6 100644 --- a/src/storage/storage_backend_rbd.c +++ b/src/storage/storage_backend_rbd.c @@ -334,7 +334,7 @@ virStorageBackendRBDSetAllocation(virStorageVolDefPtr vol, goto cleanup; } - VIR_DEBUG("Found %zu bytes allocated for RBD image %s", + VIR_DEBUG("Found %llu bytes allocated for RBD image %s", allocation, vol->name); vol->target.allocation = allocation; -- 2.5.0

2 2

[libvirt] [PATCH] nss: FreeBSD support
by Roman Bogorodskiy 25 Mar '16

25 Mar '16

* tools/nss/libvirt_nss.[ch]: add BSD-comptabile wrappers and register via the nss_module_register() interface * m4/virt-nss.m4: add checks if we're building NSS for FreeBSD * tools/Makefile.am: handle target library name differences, as Linux needs libnss_libvirt.so.2 and FreeBSD needs nss_libvirt.so.1. Also, different syms files have to be used as Linux needs to export all the methods while FreeBSD only needs to have nss_module_register() * libvirt_nss_bsd.syms: FreeBSD syms file --- m4/virt-nss.m4 | 12 +++- tools/Makefile.am | 16 ++++- tools/nss/libvirt_nss.c | 140 +++++++++++++++++++++++++++++++++++++++-- tools/nss/libvirt_nss.h | 9 +++ tools/nss/libvirt_nss_bsd.syms | 9 +++ 5 files changed, 179 insertions(+), 7 deletions(-) create mode 100644 tools/nss/libvirt_nss_bsd.syms diff --git a/m4/virt-nss.m4 b/m4/virt-nss.m4 index 3fa4ad3..a8ed8b9 100644 --- a/m4/virt-nss.m4 +++ b/m4/virt-nss.m4 @@ -23,6 +23,7 @@ AC_DEFUN([LIBVIRT_CHECK_NSS],[ [enable Name Servie Switch plugin for resolving guest IP addresses])], [], [with_nss_plugin=check]) + bsd_nss=no fail=0 if test "x$with_nss_plugin" != "xno" ; then AC_CHECK_HEADERS([nss.h], [ @@ -39,11 +40,20 @@ AC_DEFUN([LIBVIRT_CHECK_NSS],[ if test "x$with_nss_plugin" = "xyes" ; then AC_DEFINE_UNQUOTED([NSS], 1, [whether nss plugin is enabled]) + + AC_CHECK_TYPES([ns_mtab, nss_module_unregister_fn], + [AC_DEFINE([HAVE_BSD_NSS], + [1], + [whether using BSD style NSS]) + bsd_nss=yes + ], + [], + [#include <nsswitch.h>]) fi fi AM_CONDITIONAL(WITH_NSS, [test "x$with_nss_plugin" = "xyes"]) - + AM_CONDITIONAL(WITH_BSD_NSS, [test "x$bsd_nss" = "xyes"]) ]) AC_DEFUN([LIBVIRT_RESULT_NSS],[ diff --git a/tools/Makefile.am b/tools/Makefile.am index 4320040..6005b8b 100644 --- a/tools/Makefile.am +++ b/tools/Makefile.am @@ -417,8 +417,22 @@ CLEANFILES += wireshark/src/plugin.c endif WITH_WIRESHARK_DISSECTOR +if WITH_BSD_NSS +LIBVIRT_NSS_SYMBOL_FILE = \ + $(srcdir)/nss/libvirt_nss_bsd.syms +NSS_SO_VER = 1 + +install-exec-hook: + cd $(DESTDIR)$(libdir) && \ + $(LN_S) libnss_libvirt.so.$(NSS_SO_VER) nss_libvirt.so.$(NSS_SO_VER) + +uninstall-local: + rm $(DESTDIR)$(libdir)/libnss_libvirt.so.$(NSS_SO_VER) +else ! WITH_BSD_NSS LIBVIRT_NSS_SYMBOL_FILE = \ $(srcdir)/nss/libvirt_nss.syms +NSS_SO_VER = 2 +endif ! WITH_BSD_NSS LIBVIRT_NSS_SOURCES = \ nss/libvirt_nss.c \ @@ -449,7 +463,7 @@ nss_libnss_libvirt_la_LDFLAGS = \ -export-dynamic \ -avoid-version \ -shared \ - -shrext .so.2 + -shrext .so.$(NSS_SO_VER) nss_libnss_libvirt_la_LIBADD = \ nss/libnss_libvirt_impl.la diff --git a/tools/nss/libvirt_nss.c b/tools/nss/libvirt_nss.c index 218c62a..0ff1348 100644 --- a/tools/nss/libvirt_nss.c +++ b/tools/nss/libvirt_nss.c @@ -1,5 +1,4 @@ /* - * libvirt_nss: Name Service Switch plugin * * The aim is to enable users and applications to translate * domain names into IP addresses. However, this is currently @@ -29,11 +28,16 @@ #include "libvirt_nss.h" +#include <netinet/in.h> #include <resolv.h> #include <sys/types.h> #include <dirent.h> #include <arpa/inet.h> +#if defined(HAVE_BSD_NSS) +# include <nsswitch.h> +#endif + #include "virlease.h" #include "viralloc.h" #include "virfile.h" @@ -65,7 +69,7 @@ do { \ #define LEASEDIR LOCALSTATEDIR "/lib/libvirt/dnsmasq/" -#define ALIGN(x) (((x) + __SIZEOF_POINTER__ - 1) & ~(__SIZEOF_POINTER__ - 1)) +#define LIBVIRT_ALIGN(x) (((x) + __SIZEOF_POINTER__ - 1) & ~(__SIZEOF_POINTER__ - 1)) #define FAMILY_ADDRESS_SIZE(family) ((family) == AF_INET6 ? 16 : 4) typedef struct { @@ -256,7 +260,7 @@ static inline void * move_and_align(void *buf, size_t len, size_t *idx) { char *buffer = buf; - size_t move = ALIGN(len); + size_t move = LIBVIRT_ALIGN(len); if (!idx) return buffer + move; @@ -321,7 +325,7 @@ _nss_libvirt_gethostbyname3_r(const char *name, int af, struct hostent *result, * b) alias * c) addresses * d) NULL stem */ - need = ALIGN(nameLen + 1) + naddr * ALIGN(alen) + (naddr + 2) * sizeof(char*); + need = LIBVIRT_ALIGN(nameLen + 1) + naddr * LIBVIRT_ALIGN(alen) + (naddr + 2) * sizeof(char*); if (buflen < need) { *errnop = ENOMEM; @@ -383,6 +387,7 @@ _nss_libvirt_gethostbyname3_r(const char *name, int af, struct hostent *result, return ret; } +#ifdef HAVE_STRUCT_GAIH_ADDRTUPLE enum nss_status _nss_libvirt_gethostbyname4_r(const char *name, struct gaih_addrtuple **pat, char *buffer, size_t buflen, int *errnop, @@ -426,7 +431,7 @@ _nss_libvirt_gethostbyname4_r(const char *name, struct gaih_addrtuple **pat, /* We need space for: * a) name * b) addresses */ - need = ALIGN(nameLen + 1) + naddr * ALIGN(sizeof(struct gaih_addrtuple)); + need = LIBVIRT_ALIGN(nameLen + 1) + naddr * LIBVIRT_ALIGN(sizeof(struct gaih_addrtuple)); if (buflen < need) { *errnop = ENOMEM; @@ -474,3 +479,128 @@ _nss_libvirt_gethostbyname4_r(const char *name, struct gaih_addrtuple **pat, cleanup: return ret; } +#endif /* HAVE_STRUCT_GAIH_ADDRTUPLE */ + +#if defined(HAVE_BSD_NSS) +NSS_METHOD_PROTOTYPE(_nss_compat_getaddrinfo); +NSS_METHOD_PROTOTYPE(_nss_compat_gethostbyname2_r); + +ns_mtab methods[] = { + { NSDB_HOSTS, "getaddrinfo", _nss_compat_getaddrinfo, NULL }, + { NSDB_HOSTS, "gethostbyname", _nss_compat_gethostbyname2_r, NULL }, + { NSDB_HOSTS, "gethostbyname2_r", _nss_compat_gethostbyname2_r, NULL }, +}; + +static void +aiforaf(const char *name, int af, struct addrinfo *pai, struct addrinfo **aip) +{ + int ret; + struct hostent resolved; + char buf[1024] = { 0 }; + int err, herr; + struct addrinfo hints, *res0, *res; + char **addrList; + + if ((ret = _nss_libvirt_gethostbyname2_r(name, af, &resolved, + buf, sizeof(buf), + &err, &herr)) != NS_SUCCESS) + return; + + addrList = resolved.h_addr_list; + while (*addrList) { + virSocketAddr sa; + char *ipAddr = NULL; + void *address = *addrList; + + memset(&sa, 0, sizeof(sa)); + if (resolved.h_addrtype == AF_INET) { + virSocketAddrSetIPv4AddrNetOrder(&sa, *((uint32_t *) address)); + } else { + virSocketAddrSetIPv6AddrNetOrder(&sa, address); + } + + ipAddr = virSocketAddrFormat(&sa); + + hints = *pai; + hints.ai_flags = AI_NUMERICHOST; + hints.ai_family = af; + + if (getaddrinfo(ipAddr, NULL, &hints, &res0)) { + addrList++; + continue; + } + + for (res = res0; res; res = res->ai_next) + res->ai_flags = pai->ai_flags; + + (*aip)->ai_next = res0; + while ((*aip)->ai_next) + *aip = (*aip)->ai_next; + + addrList++; + } +} + +int +_nss_compat_getaddrinfo(void *retval, void *mdata ATTRIBUTE_UNUSED, va_list ap) +{ + struct addrinfo sentinel, *cur, *ai; + const char *name; + + name = va_arg(ap, char *); + ai = va_arg(ap, struct addrinfo *); + + memset(&sentinel, 0, sizeof(sentinel)); + cur = &sentinel; + + if ((ai->ai_family == AF_UNSPEC) || (ai->ai_family == AF_INET6)) + aiforaf(name, AF_INET6, ai, &cur); + if ((ai->ai_family == AF_UNSPEC) || (ai->ai_family == AF_INET)) + aiforaf(name, AF_INET, ai, &cur); + + if (sentinel.ai_next == NULL) { + h_errno = HOST_NOT_FOUND; + return NS_NOTFOUND; + } + *((struct addrinfo **)retval) = sentinel.ai_next; + + return NS_SUCCESS; +} + +int +_nss_compat_gethostbyname2_r(void *retval, void *mdata ATTRIBUTE_UNUSED, va_list ap) +{ + int ret; + + const char *name; + int af; + struct hostent *result; + char *buffer; + size_t buflen; + int *errnop; + int *herrnop; + + name = va_arg(ap, const char *); + af = va_arg(ap, int); + result = va_arg(ap, struct hostent *); + buffer = va_arg(ap, char *); + buflen = va_arg(ap, size_t); + errnop = va_arg(ap, int *); + herrnop = va_arg(ap, int *); + + ret = _nss_libvirt_gethostbyname2_r( + name, af, result, buffer, buflen, errnop, herrnop); + *(struct hostent **)retval = (ret == NS_SUCCESS) ? result : NULL; + + return ret; +} + +ns_mtab* +nss_module_register(const char *name ATTRIBUTE_UNUSED, unsigned int *size, + nss_module_unregister_fn *unregister) +{ + *size = sizeof(methods) / sizeof(methods[0]); + *unregister = NULL; + return methods; +} +#endif /* HAVE_BSD_NSS */ diff --git a/tools/nss/libvirt_nss.h b/tools/nss/libvirt_nss.h index 589c1e6..e025e63 100644 --- a/tools/nss/libvirt_nss.h +++ b/tools/nss/libvirt_nss.h @@ -45,8 +45,17 @@ enum nss_status _nss_libvirt_gethostbyname3_r(const char *name, int af, struct hostent *result, char *buffer, size_t buflen, int *errnop, int *herrnop, int32_t *ttlp, char **canonp); +# ifdef HAVE_STRUCT_GAIH_ADDRTUPLE enum nss_status _nss_libvirt_gethostbyname4_r(const char *name, struct gaih_addrtuple **pat, char *buffer, size_t buflen, int *errnop, int *herrnop, int32_t *ttlp); +# endif /* HAVE_STRUCT_GAIH_ADDRTUPLE */ + +# if defined(HAVE_BSD_NSS) +ns_mtab* +nss_module_register(const char *name, unsigned int *size, + nss_module_unregister_fn *unregister); +# endif /* HAVE_BSD_NSS */ + #endif /* __LIBVIRT_NSS_H__ */ diff --git a/tools/nss/libvirt_nss_bsd.syms b/tools/nss/libvirt_nss_bsd.syms new file mode 100644 index 0000000..7da3926 --- /dev/null +++ b/tools/nss/libvirt_nss_bsd.syms @@ -0,0 +1,9 @@ +# +# Officially exported symbols. +# + +{ +global: + nss_module_register; +local: *; +}; -- 2.4.6

2 2

[libvirt] [PATCH] admin_server: Avoid accessing unallocated memory
by Michal Privoznik 25 Mar '16

25 Mar '16

In 68b726b93c11cc90 we tried to fix a mem leak. However, it wasn't done quite well. Problem is, virNetDaemonGetServers() may fail in which case virObjectListFreeCount() would be called with -1 objects to free. But the number of elements is taken in unsigned rather than signed integer. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- Pushed under trivial rule. daemon/admin_server.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/daemon/admin_server.c b/daemon/admin_server.c index 85cec46..1d30ea5 100644 --- a/daemon/admin_server.c +++ b/daemon/admin_server.c @@ -54,7 +54,8 @@ adminConnectListServers(virNetDaemonPtr dmn, srvs = NULL; } cleanup: - virObjectListFreeCount(srvs, ret); + if (ret > 0) + virObjectListFreeCount(srvs, ret); return ret; } -- 2.7.3

1 0

[libvirt] [PATCH] virTestSetEnvPath: Avoid clearing out PATH
by Michal Privoznik 25 Mar '16

25 Mar '16

If the abs_builddir path already is in PATH and it's in the first position, due to a bug in our code PATH would be cleared out. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- Pushed under trivial rule. tests/testutils.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/tests/testutils.c b/tests/testutils.c index d6cd193..fc4c339 100644 --- a/tests/testutils.c +++ b/tests/testutils.c @@ -822,7 +822,8 @@ virTestSetEnvPath(void) goto cleanup; } - if (setenv("PATH", new_path, 1) < 0) + if (new_path && + setenv("PATH", new_path, 1) < 0) goto cleanup; ret = 0; -- 2.7.3

1 0

[libvirt] [PATCH] admin_server: fix memory leak
by Pavel Hrdina 25 Mar '16

25 Mar '16

If caller of adminConnectListServers() pass NULL instead of servers we need to free the list we've received from virNetDaemonGetServers(). Signed-off-by: Pavel Hrdina <phrdina(a)redhat.com> --- daemon/admin_server.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/daemon/admin_server.c b/daemon/admin_server.c index 1d16bc9..2f26488 100644 --- a/daemon/admin_server.c +++ b/daemon/admin_server.c @@ -52,6 +52,8 @@ adminConnectListServers(virNetDaemonPtr dmn, if (servers) { *servers = srvs; srvs = NULL; + } else { + virObjectListFreeCount(srvs, ret); } cleanup: return ret; -- 2.7.4

3 3

[libvirt] [PATCH v2] qemu: Replace some VIR_ERROR with vir*Error
by Jovanka Gulicoska 25 Mar '16

25 Mar '16

qemuStateInitialize uses a mix of VIR_ERROR and standard vir*Error calls. Prefer the standard vir*Error --- src/qemu/qemu_driver.c | 37 ++++++++++++++++++------------------- 1 file changed, 18 insertions(+), 19 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index e5ed6b5..161a20f 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -638,7 +638,6 @@ qemuStateInitialize(bool privileged, { char *driverConf = NULL; virConnectPtr conn = NULL; - char ebuf[1024]; virQEMUDriverConfigPtr cfg; uid_t run_uid = -1; gid_t run_gid = -1; @@ -649,7 +648,8 @@ qemuStateInitialize(bool privileged, return -1; if (virMutexInit(&qemu_driver->lock) < 0) { - VIR_ERROR(_("cannot initialize mutex")); + virReportError(VIR_ERR_INTERNAL_ERROR, "%s", + _("cannot initialize mutex")); VIR_FREE(qemu_driver); return -1; } @@ -682,44 +682,43 @@ qemuStateInitialize(bool privileged, VIR_FREE(driverConf); if (virFileMakePath(cfg->stateDir) < 0) { - VIR_ERROR(_("Failed to create state dir '%s': %s"), - cfg->stateDir, virStrerror(errno, ebuf, sizeof(ebuf))); + virReportSystemError(errno, _("Failed to create state dir %s"), + cfg->stateDir); goto error; } if (virFileMakePath(cfg->libDir) < 0) { - VIR_ERROR(_("Failed to create lib dir '%s': %s"), - cfg->libDir, virStrerror(errno, ebuf, sizeof(ebuf))); + virReportSystemError(errno, _("Failed to create lib dir %s"), + cfg->libDir); goto error; } if (virFileMakePath(cfg->cacheDir) < 0) { - VIR_ERROR(_("Failed to create cache dir '%s': %s"), - cfg->cacheDir, virStrerror(errno, ebuf, sizeof(ebuf))); + virReportSystemError(errno, _("Failed to create cache dir %s"), + cfg->cacheDir); goto error; } if (virFileMakePath(cfg->saveDir) < 0) { - VIR_ERROR(_("Failed to create save dir '%s': %s"), - cfg->saveDir, virStrerror(errno, ebuf, sizeof(ebuf))); + virReportSystemError(errno, _("Failed to create save dir %s"), + cfg->saveDir); goto error; } if (virFileMakePath(cfg->snapshotDir) < 0) { - VIR_ERROR(_("Failed to create save dir '%s': %s"), - cfg->snapshotDir, virStrerror(errno, ebuf, sizeof(ebuf))); + virReportSystemError(errno, _("Failed to create save dir %s"), + cfg->snapshotDir); goto error; } if (virFileMakePath(cfg->autoDumpPath) < 0) { - VIR_ERROR(_("Failed to create dump dir '%s': %s"), - cfg->autoDumpPath, virStrerror(errno, ebuf, sizeof(ebuf))); + virReportSystemError(errno, _("Failed to create dump dir %s"), + cfg->autoDumpPath); goto error; } if (virFileMakePath(cfg->channelTargetDir) < 0) { - VIR_ERROR(_("Failed to create channel target dir '%s': %s"), - cfg->channelTargetDir, - virStrerror(errno, ebuf, sizeof(ebuf))); + virReportSystemError(errno, _("Failed to create channel target dir %s"), + cfg->channelTargetDir); goto error; } if (virFileMakePath(cfg->nvramDir) < 0) { - VIR_ERROR(_("Failed to create nvram dir '%s': %s"), - cfg->nvramDir, virStrerror(errno, ebuf, sizeof(ebuf))); + virReportSystemError(errno, _("Failed to create nvram dir %s"), + cfg->nvramDir); goto error; } -- 2.5.0

2 1

[libvirt] [PATCH v2] qemu: Don't duplicate virGetLastErrorMessage
by Jovanka Gulicoska 25 Mar '16

25 Mar '16

These uses of virGetLastError message are just duplicating virGetLastErrorMessage. --- src/qemu/qemu_capabilities.c | 14 ++++---------- src/qemu/qemu_driver.c | 9 ++------- 2 files changed, 6 insertions(+), 17 deletions(-) diff --git a/src/qemu/qemu_capabilities.c b/src/qemu/qemu_capabilities.c index b223837..2823843 100644 --- a/src/qemu/qemu_capabilities.c +++ b/src/qemu/qemu_capabilities.c @@ -3089,10 +3089,8 @@ virQEMUCapsInitCached(virQEMUCapsPtr qemuCaps, const char *cacheDir) if (virQEMUCapsLoadCache(qemuCaps, capsfile, &qemuctime, &selfctime, &selfvers) < 0) { - virErrorPtr err = virGetLastError(); VIR_WARN("Failed to load cached caps from '%s' for '%s': %s", - capsfile, qemuCaps->binary, err ? NULLSTR(err->message) : - _("unknown error")); + capsfile, qemuCaps->binary, virGetLastErrorMessage()); virResetLastError(); ret = 0; virQEMUCapsReset(qemuCaps); @@ -3320,9 +3318,8 @@ virQEMUCapsInitQMPMonitor(virQEMUCapsPtr qemuCaps, /* @mon is supposed to be locked by callee */ if (qemuMonitorSetCapabilities(mon) < 0) { - virErrorPtr err = virGetLastError(); VIR_DEBUG("Failed to set monitor capabilities %s", - err ? err->message : "<unknown problem>"); + virGetLastErrorMessage()); ret = 0; goto cleanup; } @@ -3330,9 +3327,8 @@ virQEMUCapsInitQMPMonitor(virQEMUCapsPtr qemuCaps, if (qemuMonitorGetVersion(mon, &major, &minor, &micro, &package) < 0) { - virErrorPtr err = virGetLastError(); VIR_DEBUG("Failed to query monitor version %s", - err ? err->message : "<unknown problem>"); + virGetLastErrorMessage()); ret = 0; goto cleanup; } @@ -3561,15 +3557,13 @@ virQEMUCapsLogProbeFailure(const char *binary) { .key = "LIBVIRT_QEMU_BINARY", .s = binary, .iv = 0 }, { .key = NULL }, }; - virErrorPtr err = virGetLastError(); virLogMessage(&virLogSelf, VIR_LOG_WARN, __FILE__, __LINE__, __func__, meta, _("Failed to probe capabilities for %s: %s"), - binary, err && err->message ? err->message : - _("unknown failure")); + binary, virGetLastErrorMessage()); } diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 29c3bc8..6e0e19e 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -277,7 +277,6 @@ qemuAutostartDomain(virDomainObjPtr vm, void *opaque) { struct qemuAutostartData *data = opaque; - virErrorPtr err; int flags = 0; virQEMUDriverConfigPtr cfg = virQEMUDriverGetConfig(data->driver); int ret = -1; @@ -291,19 +290,15 @@ qemuAutostartDomain(virDomainObjPtr vm, if (vm->autostart && !virDomainObjIsActive(vm)) { if (qemuProcessBeginJob(data->driver, vm) < 0) { - err = virGetLastError(); VIR_ERROR(_("Failed to start job on VM '%s': %s"), - vm->def->name, - err ? err->message : _("unknown error")); + vm->def->name, virGetLastErrorMessage()); goto cleanup; } if (qemuDomainObjStart(data->conn, data->driver, vm, flags, QEMU_ASYNC_JOB_START) < 0) { - err = virGetLastError(); VIR_ERROR(_("Failed to autostart VM '%s': %s"), - vm->def->name, - err ? err->message : _("unknown error")); + vm->def->name, virGetLastErrorMessage()); } qemuProcessEndJob(data->driver, vm); -- 2.5.0

2 1

[libvirt] Usability Enhancement: Import/Export VMs GUI
by bancfc＠openmailbox.org 25 Mar '16

25 Mar '16

The single most important usability feature missed by our less technical users who migrate from VirtualBox is a one click import/export of VMs and their config settings. I was optimistic about the Gnome Boxes effort on Govf lib but unfortunately it was never realized and I would hesitate to recommend it because libvirt/virt-manager has the security advantages of sVirt. An ideal solution would work across all KVM frontends.

4 8

[libvirt] [PATCH 0/3] virlog: Refactor logging filters parser
by Erik Skultety 24 Mar '16

24 Mar '16

This series continues where series 0b231195 left and refactors logging filter parser. Series 0b231195 also introduced a small bug where if a specific testcase passes, but should fail, the original parser return value is returned instead of -1. That way, the virlogtest as a whole still appears as passed, even though it failed. Thus, this series also addresses this issue. Erik Skultety (3): tests: virlogtest: Fix testLogParseOutputs return value tests: Add new testcases to test parsing of log filters in virlogtest virlog: Refactor virLogParseFilters src/util/virlog.c | 95 +++++++++++++++++++++++++++++++++++++----------------- tests/virlogtest.c | 50 +++++++++++++++++++++++++--- 2 files changed, 111 insertions(+), 34 deletions(-) -- 2.4.3

2 6

[libvirt] [PATCH 0/2] Couple of small fixes
by Michal Privoznik 24 Mar '16

24 Mar '16

*** SOME BLURB HERE *** Michal Privoznik (2): nsstest: Drop useless @data check testutils: Adapt to highly unlikely case tests/nsstest.c | 3 --- tests/testutils.c | 13 ++++++++++--- 2 files changed, 10 insertions(+), 6 deletions(-) -- 2.7.3

2 3

[libvirt] current master failed to compile
by Vasiliy Tolstov 24 Mar '16

24 Mar '16

libvirt.c:323:15: error: 'gcry_thread_cbs' is deprecated [-Werror=deprecated-declarations] static struct gcry_thread_cbs virTLSThreadImpl = { ^ In file included from libvirt.c:58:0: /usr/x86_64-pc-linux-gnu/include/gcrypt.h:211:8: note: declared here struct gcry_thread_cbs ^ libvirt.c:331:5: error: excess elements in struct initializer [-Werror] virTLSMutexInit, ^ libvirt.c:331:5: note: (near initialization for 'virTLSThreadImpl') libvirt.c:332:5: error: excess elements in struct initializer [-Werror] virTLSMutexDestroy, ^ libvirt.c:332:5: note: (near initialization for 'virTLSThreadImpl') libvirt.c:333:5: error: excess elements in struct initializer [-Werror] virTLSMutexLock, ^ libvirt.c:333:5: note: (near initialization for 'virTLSThreadImpl') libvirt.c:334:5: error: excess elements in struct initializer [-Werror] virTLSMutexUnlock, ^ libvirt.c:334:5: note: (near initialization for 'virTLSThreadImpl') my configure line is: ./configure --build=x86_64-pc-linux-gnu --host=x86_64-pc-linux-gnu --prefix=/usr/x86_64-pc-linux-gnu --bindir=/usr/x86_64-pc-linux-gnu/bin --sbindir=/usr/x86_64-pc-linux-gnu/bin --libdir=/usr/x86_64-pc-linux-gnu/lib --datadir=/usr/share --datarootdir=/usr/share --docdir=/usr/share/doc/libvirt-scm --infodir=/usr/share/info --mandir=/usr/share/man --sysconfdir=/etc --localstatedir=/var/lib --disable-dependency-tracking --disable-silent-rules --enable-fast-install --enable-nls --localstatedir=/var --with-blkid --with-chrdev-lock-files=/run/lock --with-curl --with-dbus --with-firewalld --with-gnutls --with-html-dir=/usr/share/doc/libvirt-scm --with-html-subdir=./ --with-init-script=systemd --with-interface --with-libvirtd --with-macvtap --with-pciaccess --with-qemu-group=kvm --with-readline --with-remote --with-secrets --with-ssh2 --with-storage-dir --with-storage-fs --with-sysctl --with-test --with-udev --with-virtualport --without-apparmor --without-apparmor-mount --without-apparmor-profiles --without-audit --without-driver-modules --without-dtrace --without-esx --without-hal --without-hyperv --without-included-regex --without-libxl --without-netcf --without-openvz --without-openwsman --without-phyp --without-rhel5-api --without-sanlock --without-secdriver-apparmor --without-secdriver-selinux --without-selinux --without-selinux-mount --without-storage-gluster --without-storage-rbd --without-storage-zfs --without-vz --without-xen --without-xen-inotify --without-xenapi --enable-debug --without-avahi --with-attr --without-storage-iscsi --with-qemu --with-yajl --with-capng --without-storage-lvm --without-storage-mpath --with-lxc --with-fuse --without-network --without-numactl --without-numad --without-libpcap --without-pm-utils --without-polkit --without-storage-disk --without-sasl --without-storage-scsi --without-storage-sheepdog --without-uml --without-vbox --without-vmware --without-wireshark-dissector -- Vasiliy Tolstov, e-mail: v.tolstov(a)selfip.ru

4 6

[libvirt] [PATCH 0/3] Couple of coverity fixes
by Michal Privoznik 24 Mar '16

24 Mar '16

These all have been spotted by coverity. Michal Privoznik (3): event-test: Use goto cleanup instead of if else jungle event-test: Check for virConnectRegisterCloseCallback return value gendispatch: Avoid comparing signed and unsigned vars examples/object-events/event-test.c | 138 +++++++++++++++++++----------------- src/rpc/gendispatch.pl | 2 +- 2 files changed, 75 insertions(+), 65 deletions(-) -- 2.4.10

2 5

[libvirt] [PATCH v6 0/4] ARM: add query-gic-capabilities SMP command
by Peter Xu 24 Mar '16

24 Mar '16

v6 changes: - patch 1 (squashed into patch 2) - explain more about the following in commit message: why we need this command, and what does the entries mean [Markus] - explain what GIC is [Eric] - squash this patch into patch 2 [Eric] - patch 2 (new patch 1) - fix "does not implement" english error [Eric] - remove useless headers in target-arm/monitor.c [Markus] - remove this command from whitelist [Markus, Eric] - return dict rather than array, with single key points to the original array results [Markus, Eric] - patch 5 (new patch 4) - tiny change in implementation to suite the new interface. v5 changes: - patch 2: moved to target-arm/monitor.c (from target-arm/machine.c) [Peter] - patch 3: splitted into three patches: [all from Peter's comments] - patch 3 (new): leverage kvm_arm_create_scratch_host_vcpu(), tiny enhancement of old one to suite our need - patch 4: introduce kvm_support_device() in kvm-all.c - patch 5: do the implementation. v4 changes: - all: rename query-gic-capability to query-gic-capabilities [Andrea] - patch 3: rename helper function to kvm_support_device, make it inline and lighter. [Drew] v3 changes: - patch 2: remove func declaration, add qmp header [Drew] - patch 3: being able to detect KVM GIC capabilities even without kvm enabled [Andrea]: this is a little bit hacky, need some more review on this. v2 changes: - result layout change: use array and dict for the capability bits rather than a single array of strings [Andrea/Markus] - spelling out what GIC is in doc [Eric] This patch is to add ARM-specific command "query-gic-capability". The new command can report which kind of GIC device the host/QEMU support. The returned result is in the form of array. Sample command and output: {"execute": "query-gic-capability"} {"return": [{"emulated": false, "version": 3, "kernel": false}, {"emulated": true, "version": 2, "kernel": true}]} Testing: Smoke tests on both x86 (emulated) and another moonshot ARM server. Peter Xu (4): arm: qmp: add query-gic-capabilities interface arm: enhance kvm_arm_create_scratch_host_vcpu kvm: add kvm_support_device() helper function arm: implement query-gic-capabilities include/sysemu/kvm.h | 9 +++++ kvm-all.c | 15 +++++++++ monitor.c | 8 +++++ qapi-schema.json | 50 ++++++++++++++++++++++++++++ qmp-commands.hx | 28 ++++++++++++++++ target-arm/Makefile.objs | 2 +- target-arm/kvm.c | 10 +++++- target-arm/kvm_arm.h | 6 ++-- target-arm/monitor.c | 86 ++++++++++++++++++++++++++++++++++++++++++++++++ 9 files changed, 210 insertions(+), 4 deletions(-) create mode 100644 target-arm/monitor.c -- 2.4.3

4 17

[libvirt] [PATCH] remote: Fix error message
by Cole Robinson 24 Mar '16

24 Mar '16

We were printing the wrong MAX value --- Pushed as trivial daemon/remote.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/daemon/remote.c b/daemon/remote.c index 2bf9e83..3bf94fa 100644 --- a/daemon/remote.c +++ b/daemon/remote.c @@ -5724,11 +5724,11 @@ remoteDispatchConnectGetAllDomainStats(virNetServerPtr server ATTRIBUTE_UNUSED, if (nrecords > REMOTE_CONNECT_GET_ALL_DOMAIN_STATS_MAX) { virReportError(VIR_ERR_INTERNAL_ERROR, _("Number of domain stats records is %d, " "which exceeds max limit: %d"), - nrecords, REMOTE_DOMAIN_LIST_MAX); + nrecords, REMOTE_CONNECT_GET_ALL_DOMAIN_STATS_MAX); goto cleanup; } if (nrecords) { if (VIR_ALLOC_N(ret->retStats.retStats_val, nrecords) < 0) -- 2.5.0

1 0

[libvirt] Block replication driver
by Simon Kollberg 24 Mar '16

24 Mar '16

Hi! I'm working on supporting a new FT/HA solution for qemu called COLO (http://wiki.qemu.org/Features/COLO) The part that is currently being focused on for libvirt integration is Block Replication (http://wiki.qemu.org/Features/BlockReplication) which enables guest state synchronization for disks. Right now there are three issues that I'd like to get your input on: 1. As you can see on the block replication wiki-page we need to reference the secondary disk id. Example from the wiki: -drive if=none,driver=raw,file.filename=1.raw,id=colo1 \ -drive if=xxx,driver=replication,mode=secondary,\ ... file.backing.backing=colo1 My initial thought was to manually set the alias of the disk and add a new reference element to the backingStore: <disk type='file' device='disk'> ... <alias name='colo1'/> </disk> <disk type='file' device='disk'> ... <backingStore type='file'> ... <reference name='colo1'/> </backingStore> </disk> Though, I quickly realized that setting the alias is only done by the hypervisor and is therefore not an option with the current code. Would it be bad letting the user set the alias, and if so, do you have any ideas of how to solve the referencing? 2. The format of the disk and the driver type currently shares the same attribute in libvirt (the type attribute on driver XML element). However, with the new replication disk driver you need to be able to set both the disk format and also the driver name. Example from the wiki: -drive if=xxx,driver=replication,mode=secondary,\ file.file.filename=active_disk.qcow2,\ file.driver=qcow2,\ ... I saw that there was a function in libvirt called virStorageFileProbeFormat that could let us get the format of the disk without stating it in the XML. But as I'm sure you know, it's strongly advised not to be used since you can trick the function by modifying the disk file. 3. When using the replication driver the secondary disk is supposed to be added but not attached. Example from the wiki: -drive if=none,driver=raw,file.filename=1.raw,id=colo1 \ -drive if=xxx,driver=replication,mode=secondary,\ ... Clearly, trying to setup a disk without a target is not allowed at the moment. Is there any better way of doing it? Any comments and insights in general are also greatly appreciated! /Simon

4 5

[libvirt] [PATCH lxc_ethernet] lxc:support network type with ethernet
by zhongguocheng1＠163.com 24 Mar '16

24 Mar '16

From: z00209963 <z00209963(a)szxbz472.huaweiobz.com> --- src/lxc/lxc_process.c | 35 +++++++++++++++++++++++++++++++++++ src/lxc/lxc_process.h | 3 +++ 2 files changed, 38 insertions(+) diff --git a/src/lxc/lxc_process.c b/src/lxc/lxc_process.c index 57e3880..1ffd9ac 100644 --- a/src/lxc/lxc_process.c +++ b/src/lxc/lxc_process.c @@ -359,6 +359,38 @@ char *virLXCProcessSetupInterfaceDirect(virConnectPtr conn, return ret; } +char *virLXCProcessSetupInterfaceEthernet(virDomainDefPtr vm, + virDomainNetDefPtr net) +{ + char *ret = NULL; + char *parentVeth; + char *containerVeth = NULL; + + VIR_DEBUG("calling vethCreate()"); + parentVeth = net->ifname; + if (virNetDevVethCreate(&parentVeth, &containerVeth) < 0) + goto cleanup; + VIR_DEBUG("parentVeth: %s, containerVeth: %s", parentVeth, containerVeth); + + if (net->ifname == NULL) + net->ifname = parentVeth; + + if (virNetDevSetMAC(containerVeth, &net->mac) < 0) + goto cleanup; + + if (virNetDevSetOnline(parentVeth, true) < 0) + goto cleanup; + + if (net->filter && + virDomainConfNWFilterInstantiate(vm->uuid, net) < 0) + goto cleanup; + + ret = containerVeth; + + cleanup: + return ret; +} + static const char *nsInfoLocal[VIR_LXC_DOMAIN_NAMESPACE_LAST] = { [VIR_LXC_DOMAIN_NAMESPACE_SHARENET] = "net", [VIR_LXC_DOMAIN_NAMESPACE_SHAREIPC] = "ipc", @@ -559,6 +591,9 @@ static int virLXCProcessSetupInterfaces(virConnectPtr conn, break; case VIR_DOMAIN_NET_TYPE_ETHERNET: + if (!(veth = virLXCProcessSetupInterfaceEthernet(def, + net))) + goto cleanup; break; case VIR_DOMAIN_NET_TYPE_USER: diff --git a/src/lxc/lxc_process.h b/src/lxc/lxc_process.h index b6c8083..cc8e707 100644 --- a/src/lxc/lxc_process.h +++ b/src/lxc/lxc_process.h @@ -53,5 +53,8 @@ char *virLXCProcessSetupInterfaceBridged(virDomainDefPtr vm, char *virLXCProcessSetupInterfaceDirect(virConnectPtr conn, virDomainDefPtr def, virDomainNetDefPtr net); +char *virLXCProcessSetupInterfaceEthernet(virDomainDefPtr vm, + virDomainNetDefPtr net); + #endif /* __LXC_PROCESS_H__ */ -- 2.5.1.windows.1

4 4

[libvirt] [RFC PATCH] hostdev: add support for "managed='detach'"
by Laine Stump 24 Mar '16

24 Mar '16

Suggested by Alex Williamson. If you plan to assign a GPU to a virtual machine, but that GPU happens to be the host system console, you likely want it to start out using the host driver (so that boot messages/etc will be displayed), then later have the host driver replaced with vfio-pci for assignment to the virtual machine. However, in at least some cases (e.g. Intel i915) once the device has been detached from the host driver and attached to vfio-pci, attempts to reattach to the host driver only lead to "grief" (ask Alex for details). This means that simply using "managed='yes'" in libvirt won't work. And if you set "managed='no'" in libvirt then either you have to manually run virsh nodedev-detach prior to the first start of the guest, or you have to have a management application intelligent enough to know that it should detach from the host driver, but never reattach to it. This patch makes it simple/automatic to deal with such a case - it adds a third "managed" mode for assigned PCI devices, called "detach". It will detach ("unbind" in driver parlance) the device from the host driver prior to assigning it to the guest, but when the guest is finished with the device, will leave it bound to vfio-pci. This allows re-using the device for another guest, without requiring initial out-of-band intervention to unbind the host driver. --- I'm sending this with the "RFC" tag because I'm concerned it might be considered "feature creep" by some (although I think it makes at least as much sense as "managed='yes'") and also because, even once (if) it is ACKed, I wouldn't want to push it until abologna is finished hacking around with the driver bind/unbind code - he has enough grief to deal with without me causing a bunch of merge conflicts :-) docs/formatdomain.html.in | 29 ++++++++++++----- docs/schemas/basictypes.rng | 8 +++++ docs/schemas/domaincommon.rng | 4 +-- docs/schemas/network.rng | 2 +- src/conf/domain_conf.c | 21 +++++++----- src/conf/domain_conf.h | 2 +- src/conf/network_conf.c | 18 +++++------ src/conf/network_conf.h | 5 +-- src/libvirt_private.syms | 2 ++ src/network/bridge_driver.c | 2 +- src/qemu/qemu_parse_command.c | 4 +-- src/util/virhostdev.c | 10 +++--- src/util/virpci.c | 14 +++++--- src/util/virpci.h | 14 ++++++-- src/xenconfig/xen_common.c | 3 +- src/xenconfig/xen_sxpr.c | 9 +++--- tests/networkxml2xmlin/hostdev-managed-detach.xml | 10 ++++++ tests/networkxml2xmlout/hostdev-managed-detach.xml | 10 ++++++ tests/networkxml2xmltest.c | 1 + .../qemuxml2argv-hostdev-managed-detach.args | 22 +++++++++++++ .../qemuxml2argv-hostdev-managed-detach.xml | 30 ++++++++++++++++++ tests/qemuxml2argvtest.c | 3 ++ .../qemuxml2xmlout-hostdev-managed-detach.xml | 37 ++++++++++++++++++++++ tests/qemuxml2xmltest.c | 1 + tests/virhostdevtest.c | 8 ++--- 25 files changed, 215 insertions(+), 54 deletions(-) create mode 100644 tests/networkxml2xmlin/hostdev-managed-detach.xml create mode 100644 tests/networkxml2xmlout/hostdev-managed-detach.xml create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-hostdev-managed-detach.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-hostdev-managed-detach.xml create mode 100644 tests/qemuxml2xmloutdata/qemuxml2xmlout-hostdev-managed-detach.xml diff --git a/docs/formatdomain.html.in b/docs/formatdomain.html.in index 41f2488..a732c1c 100644 --- a/docs/formatdomain.html.in +++ b/docs/formatdomain.html.in @@ -3323,15 +3323,28 @@ hot-unplugged. </dd> <dt>pci</dt> - <dd>For PCI devices, when <code>managed</code> is "yes" it is - detached from the host before being passed on to the guest - and reattached to the host after the guest exits. If + <dd>For PCI devices, when <code>managed</code> is "yes" it + is detached from the host driver (and attached to the + appropriate driver for the planned type of device + assignment, e.g. vfio-pci for VFIO, pci-stub for legacy + kvm, or pciback for Xen device passthrough) before being + passed to the guest, and reattached to the host driver + after the guest exits. If <code>managed</code> is + "detach", then the device is detached from the host driver + before passing to the guest, but is left attached to + vfio-pci or pci-stub when the guest exits. If <code>managed</code> is omitted or "no", the user is - responsible to call <code>virNodeDeviceDetachFlags</code> - (or <code>virsh nodedev-detach</code> before starting the guest - or hot-plugging the device and <code>virNodeDeviceReAttach</code> - (or <code>virsh nodedev-reattach</code>) after hot-unplug or - stopping the guest. + responsible for assuring that the device is attached to + the proper driver for the desired device assignment mode + before starting the guest or hot-plugging the device. This + can be done in the libvirt API with + <code>virNodeDeviceDetachFlags</code> (or by running + <code>virsh nodedev-detach</code>). Likewise, once the + guest is finished using the device, it will be up to the + user to re-attach it to the host driver (if that is + desired), or example by using + libvirt's <code>virNodeDeviceReAttach</code> API or + running <code>virsh nodedev-reattach</code>. </dd> <dt>scsi</dt> <dd>For SCSI devices, user is responsible to make sure the device diff --git a/docs/schemas/basictypes.rng b/docs/schemas/basictypes.rng index a83063a..50ba9c7 100644 --- a/docs/schemas/basictypes.rng +++ b/docs/schemas/basictypes.rng @@ -88,6 +88,14 @@ </optional> </define> + <define name="virPCIManagedMode"> + <choice> + <value>no</value> + <value>yes</value> + <value>detach</value> + </choice> + </define> +    diff --git a/docs/schemas/domaincommon.rng b/docs/schemas/domaincommon.rng index 8c6287d..3701f4f 100644 --- a/docs/schemas/domaincommon.rng +++ b/docs/schemas/domaincommon.rng @@ -2241,7 +2241,7 @@ </attribute> <optional> <attribute name="managed"> - <ref name="virYesNo"/> + <ref name="virPCIManagedMode"/> </attribute> </optional> <interleave> @@ -3756,7 +3756,7 @@ </optional> <optional> <attribute name="managed"> - <ref name="virYesNo"/> + <ref name="virPCIManagedMode"/> </attribute> </optional> <choice> diff --git a/docs/schemas/network.rng b/docs/schemas/network.rng index 4edb6eb..f5e049b 100644 --- a/docs/schemas/network.rng +++ b/docs/schemas/network.rng @@ -112,7 +112,7 @@ <optional> <attribute name="managed"> - <ref name="virYesNo"/> + <ref name="virPCIManagedMode"/> </attribute> </optional> <interleave> diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index 7d68096..2332b69 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -39,6 +39,7 @@ #include "viruuid.h" #include "virbuffer.h" #include "virlog.h" +#include "virpci.h" #include "nwfilter_conf.h" #include "storage_conf.h" #include "virstoragefile.h" @@ -5599,9 +5600,11 @@ virDomainHostdevDefParseXMLSubsys(xmlNodePtr node, * element that might be (pure hostdev, or higher level device * (e.g. <interface>) with type='hostdev') */ - if ((managed = virXMLPropString(node, "managed")) != NULL) { - if (STREQ(managed, "yes")) - def->managed = true; + if ((managed = virXMLPropString(node, "managed")) && + (def->managed = virPCIManagedModeTypeFromString(managed)) < 0) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, + _("unknown managed mode '%s'"), managed); + goto error; } sgio = virXMLPropString(node, "sgio"); @@ -19808,8 +19811,9 @@ virDomainActualNetDefFormat(virBufferPtr buf, virBufferAsprintf(buf, "<actual type='%s'", typeStr); if (type == VIR_DOMAIN_NET_TYPE_HOSTDEV) { virDomainHostdevDefPtr hostdef = virDomainNetGetActualHostdev(def); - if (hostdef && hostdef->managed) - virBufferAddLit(buf, " managed='yes'"); + if (hostdef && (hostdef->managed != VIR_PCI_MANAGED_NO)) + virBufferAsprintf(buf, " managed='%s'", + virPCIManagedModeTypeToString(hostdef->managed)); } if (def->trustGuestRxFilters) virBufferAsprintf(buf, " trustGuestRxFilters='%s'", @@ -19979,8 +19983,9 @@ virDomainNetDefFormat(virBufferPtr buf, } virBufferAsprintf(buf, "<interface type='%s'", typeStr); - if (hostdef && hostdef->managed) - virBufferAddLit(buf, " managed='yes'"); + if (hostdef && (hostdef->managed != VIR_PCI_MANAGED_NO)) + virBufferAsprintf(buf, " managed='%s'", + virPCIManagedModeTypeToString(hostdef->managed)); if (def->trustGuestRxFilters) virBufferAsprintf(buf, " trustGuestRxFilters='%s'", virTristateBoolTypeToString(def->trustGuestRxFilters)); @@ -21430,7 +21435,7 @@ virDomainHostdevDefFormat(virBufferPtr buf, mode, type); if (def->mode == VIR_DOMAIN_HOSTDEV_MODE_SUBSYS) { virBufferAsprintf(buf, " managed='%s'", - def->managed ? "yes" : "no"); + virPCIManagedModeTypeToString(def->managed)); if (def->source.subsys.type == VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_SCSI && scsisrc->sgio) diff --git a/src/conf/domain_conf.h b/src/conf/domain_conf.h index 85c4f55..42c7d36 100644 --- a/src/conf/domain_conf.h +++ b/src/conf/domain_conf.h @@ -543,7 +543,7 @@ struct _virDomainHostdevDef { virDomainDeviceDef parent; /* higher level Def containing this */ int mode; /* enum virDomainHostdevMode */ int startupPolicy; /* enum virDomainStartupPolicy */ - bool managed; + int managed; /* enum virPCIManagedMode */ bool missing; bool readonly; bool shareable; diff --git a/src/conf/network_conf.c b/src/conf/network_conf.c index 4fb2e2a..a40f3e0 100644 --- a/src/conf/network_conf.c +++ b/src/conf/network_conf.c @@ -1,7 +1,7 @@ /* * network_conf.c: network XML handling * - * Copyright (C) 2006-2015 Red Hat, Inc. + * Copyright (C) 2006-2016 Red Hat, Inc. * Copyright (C) 2006-2008 Daniel P. Berrange * * This library is free software; you can redistribute it and/or @@ -43,6 +43,7 @@ #include "virbuffer.h" #include "c-ctype.h" #include "virfile.h" +#include "virpci.h" #include "virstring.h" #define VIR_FROM_THIS VIR_FROM_NETWORK @@ -1825,10 +1826,11 @@ virNetworkForwardDefParseXML(const char *networkName, VIR_FREE(type); } - forwardManaged = virXPathString("string(./@managed)", ctxt); - if (forwardManaged != NULL && - STRCASEEQ(forwardManaged, "yes")) { - def->managed = true; + if ((forwardManaged = virXPathString("string(./@managed)", ctxt)) && + (def->managed = virPCIManagedModeTypeFromString(forwardManaged)) < 0) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, + _("unknown managed mode '%s'"), forwardManaged); + goto cleanup; } forwardDriverName = virXPathString("string(./driver/@name)", ctxt); @@ -2723,10 +2725,8 @@ virNetworkDefFormatBuf(virBufferPtr buf, virBufferEscapeString(buf, " dev='%s'", dev); virBufferAsprintf(buf, " mode='%s'", mode); if (def->forward.type == VIR_NETWORK_FORWARD_HOSTDEV) { - if (def->forward.managed) - virBufferAddLit(buf, " managed='yes'"); - else - virBufferAddLit(buf, " managed='no'"); + virBufferAsprintf(buf, " managed='%s'", + virPCIManagedModeTypeToString(def->forward.managed)); } shortforward = !(def->forward.nifs || def->forward.npfs || VIR_SOCKET_ADDR_VALID(&def->forward.addr.start) diff --git a/src/conf/network_conf.h b/src/conf/network_conf.h index b72257b..3b86781 100644 --- a/src/conf/network_conf.h +++ b/src/conf/network_conf.h @@ -1,7 +1,7 @@ /* * network_conf.h: network XML handling * - * Copyright (C) 2006-2015 Red Hat, Inc. + * Copyright (C) 2006-2016 Red Hat, Inc. * Copyright (C) 2006-2008 Daniel P. Berrange * * This library is free software; you can redistribute it and/or @@ -187,7 +187,8 @@ typedef struct _virNetworkForwardDef virNetworkForwardDef; typedef virNetworkForwardDef *virNetworkForwardDefPtr; struct _virNetworkForwardDef { int type; /* One of virNetworkForwardType constants */ - bool managed; /* managed attribute for hostdev mode */ + int managed; /* managed attribute for hostdev mode */ + /* enum virPCIManagedMode */ int driverName; /* enum virNetworkForwardDriverNameType */ /* If there are multiple forward devices (i.e. a pool of diff --git a/src/libvirt_private.syms b/src/libvirt_private.syms index 55c3047..3dd00ec 100644 --- a/src/libvirt_private.syms +++ b/src/libvirt_private.syms @@ -2010,6 +2010,8 @@ virPCIGetVirtualFunctionIndex; virPCIGetVirtualFunctionInfo; virPCIGetVirtualFunctions; virPCIIsVirtualFunction; +virPCIManagedModeTypeFromString; +virPCIManagedModeTypeToString; virPCIStubDriverTypeFromString; virPCIStubDriverTypeToString; diff --git a/src/network/bridge_driver.c b/src/network/bridge_driver.c index 01c2ed6..bd1f57c 100644 --- a/src/network/bridge_driver.c +++ b/src/network/bridge_driver.c @@ -4085,7 +4085,7 @@ networkAllocateActualDevice(virDomainDefPtr dom, iface->data.network.actual->data.hostdev.def.parent.data.net = iface; iface->data.network.actual->data.hostdev.def.info = &iface->info; iface->data.network.actual->data.hostdev.def.mode = VIR_DOMAIN_HOSTDEV_MODE_SUBSYS; - iface->data.network.actual->data.hostdev.def.managed = netdef->forward.managed ? 1 : 0; + iface->data.network.actual->data.hostdev.def.managed = netdef->forward.managed; iface->data.network.actual->data.hostdev.def.source.subsys.type = dev->type; iface->data.network.actual->data.hostdev.def.source.subsys.u.pci.addr = dev->device.pci; diff --git a/src/qemu/qemu_parse_command.c b/src/qemu/qemu_parse_command.c index 60e3d69..060bf66 100644 --- a/src/qemu/qemu_parse_command.c +++ b/src/qemu/qemu_parse_command.c @@ -1192,7 +1192,7 @@ qemuParseCommandLinePCI(const char *val) } def->mode = VIR_DOMAIN_HOSTDEV_MODE_SUBSYS; - def->managed = true; + def->managed = VIR_PCI_MANAGED_YES; def->source.subsys.type = VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_PCI; def->source.subsys.u.pci.addr.bus = bus; def->source.subsys.u.pci.addr.slot = slot; @@ -1255,7 +1255,7 @@ qemuParseCommandLineUSB(const char *val) } def->mode = VIR_DOMAIN_HOSTDEV_MODE_SUBSYS; - def->managed = false; + def->managed = VIR_PCI_MANAGED_NO; def->source.subsys.type = VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_USB; if (*end == '.') { usbsrc->bus = first; diff --git a/src/util/virhostdev.c b/src/util/virhostdev.c index 098207e..4baf884 100644 --- a/src/util/virhostdev.c +++ b/src/util/virhostdev.c @@ -1,6 +1,6 @@ /* virhostdev.c: hostdev management * - * Copyright (C) 2006-2007, 2009-2015 Red Hat, Inc. + * Copyright (C) 2006-2007, 2009-2016 Red Hat, Inc. * Copyright (C) 2006 Daniel P. Berrange * Copyright (C) 2014 SUSE LINUX Products GmbH, Nuernberg, Germany. * @@ -592,7 +592,7 @@ virHostdevPreparePCIDevices(virHostdevManagerPtr hostdev_mgr, for (i = 0; i < virPCIDeviceListCount(pcidevs); i++) { virPCIDevicePtr dev = virPCIDeviceListGet(pcidevs, i); - if (virPCIDeviceGetManaged(dev)) { + if (virPCIDeviceGetManaged(dev) != VIR_PCI_MANAGED_NO) { VIR_DEBUG("Detaching managed PCI device %s", virPCIDeviceGetName(dev)); if (virPCIDeviceDetach(dev, @@ -726,7 +726,7 @@ virHostdevPreparePCIDevices(virHostdevManagerPtr hostdev_mgr, for (i = 0; i < virPCIDeviceListCount(pcidevs); i++) { virPCIDevicePtr dev = virPCIDeviceListGet(pcidevs, i); - if (virPCIDeviceGetManaged(dev)) { + if (virPCIDeviceGetManaged(dev) == VIR_PCI_MANAGED_YES) { VIR_DEBUG("Reattaching managed PCI device %s", virPCIDeviceGetName(dev)); ignore_value(virPCIDeviceReattach(dev, @@ -756,7 +756,7 @@ virHostdevReattachPCIDevice(virPCIDevicePtr dev, virHostdevManagerPtr mgr) /* If the device is not managed and was attached to guest * successfully, it must have been inactive. */ - if (!virPCIDeviceGetManaged(dev)) { + if (virPCIDeviceGetManaged(dev) != VIR_PCI_MANAGED_YES) { VIR_DEBUG("Adding unmanaged PCI device %s to inactive list", virPCIDeviceGetName(dev)); if (virPCIDeviceListAdd(mgr->inactivePCIHostdevs, dev) < 0) @@ -1312,7 +1312,7 @@ virHostdevPrepareSCSIHostDevices(virDomainHostdevDefPtr hostdev, virSCSIDevicePtr scsi; int ret = -1; - if (hostdev->managed) { + if (hostdev->managed != VIR_PCI_MANAGED_NO) { virReportError(VIR_ERR_XML_ERROR, "%s", _("SCSI host device doesn't support managed mode")); goto cleanup; diff --git a/src/util/virpci.c b/src/util/virpci.c index 1854318..7164824 100644 --- a/src/util/virpci.c +++ b/src/util/virpci.c @@ -1,7 +1,7 @@ /* * virpci.c: helper APIs for managing host PCI devices * - * Copyright (C) 2009-2015 Red Hat, Inc. + * Copyright (C) 2009-2016 Red Hat, Inc. * * This library is free software; you can redistribute it and/or * modify it under the terms of the GNU Lesser General Public @@ -62,6 +62,11 @@ VIR_ENUM_IMPL(virPCIStubDriver, VIR_PCI_STUB_DRIVER_LAST, "vfio-pci", /* VFIO */ ); +VIR_ENUM_IMPL(virPCIManagedMode, VIR_PCI_MANAGED_LAST, + "no", + "yes", + "detach"); + struct _virPCIDevice { virPCIDeviceAddress address; @@ -77,7 +82,7 @@ struct _virPCIDevice { unsigned int pci_pm_cap_pos; bool has_flr; bool has_pm_reset; - bool managed; + virPCIManagedMode managed; virPCIStubDriver stubDriver; @@ -1702,12 +1707,13 @@ virPCIDeviceGetName(virPCIDevicePtr dev) return dev->name; } -void virPCIDeviceSetManaged(virPCIDevicePtr dev, bool managed) +void virPCIDeviceSetManaged(virPCIDevicePtr dev, + virPCIManagedMode managed) { dev->managed = managed; } -bool +virPCIManagedMode virPCIDeviceGetManaged(virPCIDevicePtr dev) { return dev->managed; diff --git a/src/util/virpci.h b/src/util/virpci.h index 55329c8..1969e29 100644 --- a/src/util/virpci.h +++ b/src/util/virpci.h @@ -53,6 +53,16 @@ typedef enum { VIR_ENUM_DECL(virPCIStubDriver); typedef enum { + VIR_PCI_MANAGED_NO, + VIR_PCI_MANAGED_YES, + VIR_PCI_MANAGED_DETACH, + + VIR_PCI_MANAGED_LAST +} virPCIManagedMode; + +VIR_ENUM_DECL(virPCIManagedMode); + +typedef enum { VIR_PCIE_LINK_SPEED_NA = 0, VIR_PCIE_LINK_SPEED_25, VIR_PCIE_LINK_SPEED_5, @@ -98,8 +108,8 @@ int virPCIDeviceReset(virPCIDevicePtr dev, virPCIDeviceListPtr inactiveDevs); void virPCIDeviceSetManaged(virPCIDevice *dev, - bool managed); -bool virPCIDeviceGetManaged(virPCIDevice *dev); + virPCIManagedMode managed); +virPCIManagedMode virPCIDeviceGetManaged(virPCIDevice *dev); void virPCIDeviceSetStubDriver(virPCIDevicePtr dev, virPCIStubDriver driver); virPCIStubDriver virPCIDeviceGetStubDriver(virPCIDevicePtr dev); diff --git a/src/xenconfig/xen_common.c b/src/xenconfig/xen_common.c index 828c8e9..db4745c 100644 --- a/src/xenconfig/xen_common.c +++ b/src/xenconfig/xen_common.c @@ -32,6 +32,7 @@ #include "virerror.h" #include "virconf.h" #include "viralloc.h" +#include "virpci.h" #include "viruuid.h" #include "count-one-bits.h" #include "xenxs_private.h" @@ -461,7 +462,7 @@ xenParsePCI(virConfPtr conf, virDomainDefPtr def) if (!(hostdev = virDomainHostdevDefAlloc())) return -1; - hostdev->managed = false; + hostdev->managed = VIR_PCI_MANAGED_NO; hostdev->source.subsys.type = VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_PCI; hostdev->source.subsys.u.pci.addr.domain = domainID; hostdev->source.subsys.u.pci.addr.bus = busID; diff --git a/src/xenconfig/xen_sxpr.c b/src/xenconfig/xen_sxpr.c index fdfec2b..51b32a4 100644 --- a/src/xenconfig/xen_sxpr.c +++ b/src/xenconfig/xen_sxpr.c @@ -1,7 +1,7 @@ /* * xen_sxpr.c: Xen SEXPR parsing functions * - * Copyright (C) 2010-2014 Red Hat, Inc. + * Copyright (C) 2010-2014, 2016 Red Hat, Inc. * Copyright (C) 2011 Univention GmbH * Copyright (C) 2005 Anthony Liguori <aliguori(a)us.ibm.com> * @@ -32,6 +32,7 @@ #include "virerror.h" #include "virconf.h" #include "viralloc.h" +#include "virpci.h" #include "verify.h" #include "viruuid.h" #include "virlog.h" @@ -1114,7 +1115,7 @@ xenParseSxprPCI(virDomainDefPtr def, goto error; dev->mode = VIR_DOMAIN_HOSTDEV_MODE_SUBSYS; - dev->managed = false; + dev->managed = VIR_PCI_MANAGED_NO; dev->source.subsys.type = VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_PCI; dev->source.subsys.u.pci.addr.domain = domainID; dev->source.subsys.u.pci.addr.bus = busID; @@ -2002,7 +2003,7 @@ xenFormatSxprOnePCI(virDomainHostdevDefPtr def, virBufferPtr buf, int detach) { - if (def->managed) { + if (def->managed != VIR_PCI_MANAGED_NO) { virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", _("managed PCI devices not supported with XenD")); return -1; @@ -2062,7 +2063,7 @@ xenFormatSxprAllPCI(virDomainDefPtr def, for (i = 0; i < def->nhostdevs; i++) { if (def->hostdevs[i]->mode == VIR_DOMAIN_HOSTDEV_MODE_SUBSYS && def->hostdevs[i]->source.subsys.type == VIR_DOMAIN_HOSTDEV_SUBSYS_TYPE_PCI) { - if (def->hostdevs[i]->managed) { + if (def->hostdevs[i]->managed != VIR_PCI_MANAGED_NO) { virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", _("managed PCI devices not supported with XenD")); return -1; diff --git a/tests/networkxml2xmlin/hostdev-managed-detach.xml b/tests/networkxml2xmlin/hostdev-managed-detach.xml new file mode 100644 index 0000000..3a5fac4 --- /dev/null +++ b/tests/networkxml2xmlin/hostdev-managed-detach.xml @@ -0,0 +1,10 @@ +<network> + <name>hostdev</name> + <uuid>81ff0d90-c91e-6742-64da-4a736edb9a9b</uuid> + <forward mode='hostdev' managed='detach'> + <address type='pci' domain='0x0000' bus='0x03' slot='0x00' function='0x1'/> + <address type='pci' domain='0x0000' bus='0x03' slot='0x00' function='0x2'/> + <address type='pci' domain='0x0000' bus='0x03' slot='0x00' function='0x3'/> + <address type='pci' domain='0x0000' bus='0x03' slot='0x00' function='0x4'/> + </forward> +</network> diff --git a/tests/networkxml2xmlout/hostdev-managed-detach.xml b/tests/networkxml2xmlout/hostdev-managed-detach.xml new file mode 100644 index 0000000..3a5fac4 --- /dev/null +++ b/tests/networkxml2xmlout/hostdev-managed-detach.xml @@ -0,0 +1,10 @@ +<network> + <name>hostdev</name> + <uuid>81ff0d90-c91e-6742-64da-4a736edb9a9b</uuid> + <forward mode='hostdev' managed='detach'> + <address type='pci' domain='0x0000' bus='0x03' slot='0x00' function='0x1'/> + <address type='pci' domain='0x0000' bus='0x03' slot='0x00' function='0x2'/> + <address type='pci' domain='0x0000' bus='0x03' slot='0x00' function='0x3'/> + <address type='pci' domain='0x0000' bus='0x03' slot='0x00' function='0x4'/> + </forward> +</network> diff --git a/tests/networkxml2xmltest.c b/tests/networkxml2xmltest.c index 8d60aa8..ec9a47f 100644 --- a/tests/networkxml2xmltest.c +++ b/tests/networkxml2xmltest.c @@ -108,6 +108,7 @@ mymain(void) DO_TEST("openvswitch-net"); DO_TEST_FULL("passthrough-pf", VIR_NETWORK_XML_INACTIVE); DO_TEST("hostdev"); + DO_TEST("hostdev-managed-detach"); DO_TEST_FULL("hostdev-pf", VIR_NETWORK_XML_INACTIVE); DO_TEST("passthrough-address-crash"); DO_TEST("nat-network-explicit-flood"); diff --git a/tests/qemuxml2argvdata/qemuxml2argv-hostdev-managed-detach.args b/tests/qemuxml2argvdata/qemuxml2argv-hostdev-managed-detach.args new file mode 100644 index 0000000..9e37ed5 --- /dev/null +++ b/tests/qemuxml2argvdata/qemuxml2argv-hostdev-managed-detach.args @@ -0,0 +1,22 @@ +LC_ALL=C \ +PATH=/bin \ +HOME=/home/test \ +USER=test \ +LOGNAME=test \ +QEMU_AUDIO_DRV=none \ +/usr/bin/qemu \ +-name QEMUGuestManagedUnbind \ +-S \ +-M pc \ +-m 214 \ +-smp 1 \ +-uuid c7a5fdbd-edaf-9466-926a-d65c16db1809 \ +-nographic \ +-nodefconfig \ +-nodefaults \ +-monitor unix:/tmp/lib/domain--1-QEMUGuestManagedUnbi/monitor.sock,server,nowait \ +-no-acpi \ +-boot c \ +-usb \ +-device vfio-pci,host=06:12.5,id=hostdev0,bus=pci.0,addr=0x3 \ +-device virtio-balloon-pci,id=balloon0,bus=pci.0,addr=0x4 diff --git a/tests/qemuxml2argvdata/qemuxml2argv-hostdev-managed-detach.xml b/tests/qemuxml2argvdata/qemuxml2argv-hostdev-managed-detach.xml new file mode 100644 index 0000000..2d12d98 --- /dev/null +++ b/tests/qemuxml2argvdata/qemuxml2argv-hostdev-managed-detach.xml @@ -0,0 +1,30 @@ +<domain type='qemu'> + <name>QEMUGuestManagedUnbind</name> + <uuid>c7a5fdbd-edaf-9466-926a-d65c16db1809</uuid> + <memory unit='KiB'>219100</memory> + <currentMemory unit='KiB'>219100</currentMemory> + <vcpu placement='static'>1</vcpu> + <os> + <type arch='x86_64' machine='pc'>hvm</type> + <boot dev='hd'/> + </os> + <clock offset='utc'/> + <on_poweroff>destroy</on_poweroff> + <on_reboot>restart</on_reboot> + <on_crash>destroy</on_crash> + <devices> + <emulator>/usr/bin/qemu</emulator> + <controller type='usb' index='0'/> + <controller type='ide' index='0'/> + <controller type='pci' index='0' model='pci-root'/> + <input type='mouse' bus='ps2'/> + <input type='keyboard' bus='ps2'/> + <hostdev mode='subsystem' type='pci' managed='detach'> + <driver name='vfio'/> + <source> + <address domain='0x0000' bus='0x06' slot='0x12' function='0x5'/> + </source> + </hostdev> + <memballoon model='virtio'/> + </devices> +</domain> diff --git a/tests/qemuxml2argvtest.c b/tests/qemuxml2argvtest.c index e15da37..433e3a7 100644 --- a/tests/qemuxml2argvtest.c +++ b/tests/qemuxml2argvtest.c @@ -1278,6 +1278,9 @@ mymain(void) DO_TEST_FAILURE("hostdev-vfio-multidomain", QEMU_CAPS_PCIDEVICE, QEMU_CAPS_NODEFCONFIG, QEMU_CAPS_DEVICE_VFIO_PCI); + DO_TEST("hostdev-managed-detach", + QEMU_CAPS_PCIDEVICE, QEMU_CAPS_NODEFCONFIG, + QEMU_CAPS_DEVICE_VFIO_PCI); DO_TEST("pci-rom", QEMU_CAPS_PCIDEVICE, QEMU_CAPS_NODEFCONFIG, QEMU_CAPS_PCI_ROMBAR); diff --git a/tests/qemuxml2xmloutdata/qemuxml2xmlout-hostdev-managed-detach.xml b/tests/qemuxml2xmloutdata/qemuxml2xmlout-hostdev-managed-detach.xml new file mode 100644 index 0000000..e639e91 --- /dev/null +++ b/tests/qemuxml2xmloutdata/qemuxml2xmlout-hostdev-managed-detach.xml @@ -0,0 +1,37 @@ +<domain type='qemu'> + <name>QEMUGuestManagedUnbind</name> + <uuid>c7a5fdbd-edaf-9466-926a-d65c16db1809</uuid> + <memory unit='KiB'>219100</memory> + <currentMemory unit='KiB'>219100</currentMemory> + <vcpu placement='static'>1</vcpu> + <os> + <type arch='x86_64' machine='pc'>hvm</type> + <boot dev='hd'/> + </os> + <clock offset='utc'/> + <on_poweroff>destroy</on_poweroff> + <on_reboot>restart</on_reboot> + <on_crash>destroy</on_crash> + <devices> + <emulator>/usr/bin/qemu</emulator> + <controller type='usb' index='0'> + <address type='pci' domain='0x0000' bus='0x00' slot='0x01' function='0x2'/> + </controller> + <controller type='ide' index='0'> + <address type='pci' domain='0x0000' bus='0x00' slot='0x01' function='0x1'/> + </controller> + <controller type='pci' index='0' model='pci-root'/> + <input type='mouse' bus='ps2'/> + <input type='keyboard' bus='ps2'/> + <hostdev mode='subsystem' type='pci' managed='detach'> + <driver name='vfio'/> + <source> + <address domain='0x0000' bus='0x06' slot='0x12' function='0x5'/> + </source> + <address type='pci' domain='0x0000' bus='0x00' slot='0x03' function='0x0'/> + </hostdev> + <memballoon model='virtio'> + <address type='pci' domain='0x0000' bus='0x00' slot='0x04' function='0x0'/> + </memballoon> + </devices> +</domain> diff --git a/tests/qemuxml2xmltest.c b/tests/qemuxml2xmltest.c index 0735677..980c468 100644 --- a/tests/qemuxml2xmltest.c +++ b/tests/qemuxml2xmltest.c @@ -484,6 +484,7 @@ mymain(void) DO_TEST("hostdev-usb-address"); DO_TEST("hostdev-pci-address"); DO_TEST("hostdev-vfio"); + DO_TEST("hostdev-managed-detach"); DO_TEST("pci-rom"); DO_TEST("pci-serial-dev-chardev"); diff --git a/tests/virhostdevtest.c b/tests/virhostdevtest.c index 5eb2e7e..af175ca 100644 --- a/tests/virhostdevtest.c +++ b/tests/virhostdevtest.c @@ -166,7 +166,7 @@ testVirHostdevPreparePCIHostdevs_unmanaged(void) size_t active_count, inactive_count, i; for (i = 0; i < nhostdevs; i++) - hostdevs[i]->managed = false; + hostdevs[i]->managed = VIR_PCI_MANAGED_NO; active_count = virPCIDeviceListCount(mgr->activePCIHostdevs); inactive_count = virPCIDeviceListCount(mgr->inactivePCIHostdevs); @@ -225,7 +225,7 @@ testVirHostdevReAttachPCIHostdevs_unmanaged(void) size_t active_count, inactive_count, i; for (i = 0; i < nhostdevs; i++) { - if (hostdevs[i]->managed != false) { + if (hostdevs[i]->managed != VIR_PCI_MANAGED_NO) { VIR_DEBUG("invalid test"); return -1; } @@ -259,7 +259,7 @@ testVirHostdevPreparePCIHostdevs_managed(void) size_t active_count, inactive_count, i; for (i = 0; i < nhostdevs; i++) - hostdevs[i]->managed = true; + hostdevs[i]->managed = VIR_PCI_MANAGED_YES; active_count = virPCIDeviceListCount(mgr->activePCIHostdevs); inactive_count = virPCIDeviceListCount(mgr->inactivePCIHostdevs); @@ -310,7 +310,7 @@ testVirHostdevReAttachPCIHostdevs_managed(void) size_t active_count, inactive_count, i; for (i = 0; i < nhostdevs; i++) { - if (hostdevs[i]->managed != true) { + if (hostdevs[i]->managed != VIR_PCI_MANAGED_YES) { VIR_DEBUG("invalid test"); return -1; } -- 2.5.0

7 22

[libvirt] [ 0/3] more useful network ethernet type devices
by Vasiliy Tolstov 24 Mar '16

24 Mar '16

By default libvirt does not manages ethernet devices, bypass all control to qemu. This series enable creating tap devices via libvirt, assign mac, ip address and routes. Also when use specif linkstate down in xml, phisicaly down tap device. This is very useful in case of using bird, quagga or somethink like this to forward traffic to vm. Main question about peer address assign. How to deal with absent peer address? Now i'm simplify check that peer address is valid ip address, but may be this is wrong. Vasiliy Tolstov (3): autocreate tap device for ethernet network type allow to specify peer to ip element in domain network assign peer address if it provided in network docs/schemas/interface.rng | 6 ++ include/libvirt/libvirt-domain.h | 1 + src/conf/domain_conf.c | 14 +++- src/conf/domain_conf.h | 1 + src/conf/network_conf.c | 26 ++++++- src/conf/network_conf.h | 1 + src/lxc/lxc_container.c | 2 +- src/network/bridge_driver.c | 2 +- src/qemu/qemu_command.c | 38 +++++----- src/qemu/qemu_hotplug.c | 36 +++++++-- src/qemu/qemu_interface.c | 158 +++++++++++++++++++++++++++++++++++++++ src/qemu/qemu_interface.h | 5 ++ src/qemu/qemu_process.c | 6 ++ src/util/virnetdev.c | 43 +++++++---- src/util/virnetdev.h | 1 + 15 files changed, 300 insertions(+), 40 deletions(-) -- 2.7.3

4 13

[libvirt] [libvirt-php] PHP 7 compatibility
by Remi Collet 23 Mar '16

23 Mar '16

FYI, see https://gitlab.com/Conan_Kudo/libvirt-php7/commits/php7 It build but need tests. Remi.

2 1

[libvirt] [RFC 0/3] Master secret key support
by John Ferlan 23 Mar '16

23 Mar '16

Posted as RFC primarily to make sure there's buyin for something that will become the basis for more patches to add support for using the master secret to encrypt sensitive data using objects. This work is related to the QEMU 2.6 commit 'ac1d8878' (not a libvirt commit, rather a qemu.git commit id). Hopefully I've properly read the qemu checkin notes as they relate to key generation. One thing of note that differs from other descriptions for a master secret. Most bz's where it's described indicate using /var/lib/libvirt/qemu/ $GUEST-master.key; whereas, these patches chose a slightly different tact using the generated libDir (/var/lib/libvirt/qemu/domain-#-$GUEST/ master.key). This file will be generated in qemu process launch regardless of whether the emulator supports it or not (because we have the domain private object to access the masterKey, but not in qemuBuildCommandLine). The only odd part for me was the realizing that libDir is created after successfully completing qemuBuildCommandLine - so checking if the file exists before adding it to the command line wasn't possible, but it seems that's no different to other libDir usages. Internally, the I chose to store the secret to be used as a base64 value since 1. it's going to be saved in the domain objects XML output (for libvirtd restart) and 2. it's easy enough to decode if we do need to later on. Not sure it's appropriate to store the non encoded secret in the domain object XML file. One extra area I need help on is the capabilitiesdata setup... That is how to generate the tests/qemucapabilitiesdata/caps_2.6.0-1.replies so that the 'secret' object exists and so that if I add "<flag name='secret'>" to the .caps file I won't get a test failure since the existing one doesn't list the secrets object. I've done a bit of testing locally... Starting with running domains (both persistent and transient) then restarting libvirtd and with starting the same domains with the code running. I didn't update my qemu locally to see the -object on the command line, but the test added does show the -object added (although I it took a double take and some thinking whether /tmp/lib/domain--1-$NAME/ should be used - note the /tmp and the domain--# prefix). John Ferlan (3): qemu: Create domain master key qemu: Add capability bit for qemu secret object qemu: Introduce qemuBuildMasterKeyCommandLine src/qemu/qemu_capabilities.c | 2 + src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_command.c | 67 ++++++++ src/qemu/qemu_domain.c | 175 +++++++++++++++++++++ src/qemu/qemu_domain.h | 11 ++ src/qemu/qemu_process.c | 13 ++ .../qemuxml2argvdata/qemuxml2argv-master-key.args | 23 +++ tests/qemuxml2argvdata/qemuxml2argv-master-key.xml | 30 ++++ tests/qemuxml2argvtest.c | 2 + 9 files changed, 324 insertions(+) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-master-key.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-master-key.xml -- 2.5.0

3 10

[libvirt] [PATCH v2 0/4] hostdev: Cleanups and fixes
by Andrea Bolognani 23 Mar '16

23 Mar '16

Respin of patches 20, 22, 23 and 24 from https://www.redhat.com/archives/libvir-list/2016-March/msg00209.html Patches 01-19 have already been pushed; patch 21 has been converted into the tiny series https://www.redhat.com/archives/libvir-list/2016-March/msg00792.html which this one depends on. Not many changes, basically just shuffling patches around to avoid a problem found during review and other small nits. Cheers. Andrea Bolognani (4): hostdev: Streamline device ownership tracking hostdev: Save netdev configuration of actual device hostdev: Use actual device when reattaching tests: hostdev: Add more tests src/util/virhostdev.c | 181 +++++++++++++++++++++++++++---------------------- tests/virhostdevtest.c | 84 +++++++++++++++++++++-- 2 files changed, 178 insertions(+), 87 deletions(-) -- 2.5.0

2 9

[libvirt] nwfilter and ipset
by Vasiliy Tolstov 23 Mar '16

23 Mar '16

Hi! I found, that nwfilter supports ipset, does it possible to fill ipset via libvirt with some values? -- Vasiliy Tolstov, e-mail: v.tolstov(a)selfip.ru

1 0

[libvirt] [PATCH 0/2] hostdev: Small usability improvements
by Andrea Bolognani 23 Mar '16

23 Mar '16

Patch 2/2 is basically https://www.redhat.com/archives/libvir-list/2016-March/msg00230.html and patch 1/2 is an implementation of the idea outlined in https://www.redhat.com/archives/libvir-list/2016-March/msg00729.html to address the flaw discovered by John. Spun off from the main series because it can be merged on its own, and I still need to rework the remaining patches a bit before posting v2 of that. Cheers. Andrea Bolognani (2): hostdev: Detect untracked inactive devices hostdev: Stop early if unmanaged devices have not been detached src/util/virhostdev.c | 53 ++++++++++++++++++++++++++++++++++++++++++++++++--- 1 file changed, 50 insertions(+), 3 deletions(-) -- 2.5.0

2 7

[libvirt] [RFC PATCH 0/4] Implementation of a dependency relation between domains
by Valentin BOUSSON 23 Mar '16

23 Mar '16

This patch series adds a feature that enables libvirt domains to have other domains as dependencies. One possible target is the simulation of a complex computing system composed of multiple CPUs, each of which represented by a different libvirt domain. CPUs in this example are virtualized separately, by a dedicated QEMU instance, because all of them are booting independently. Moreover, if those CPU must share some physical property (for example power supply or ACPI), it is meaningful to consider them as a group. Another possible target is Asymmetric Multi Processor (AMP) systems. Where a master processor uses one or more slave processors to offload computation. Master and slave have often different architectures. This patch series makes possible to define the slave domains as dependencies of the master domain. And the whole system to be booted with one single libvirt command. A use-case for this libvirt extension are the recent patch series in QEMU that enable modeling an AMP-like system with multiple QEMU instances. [RFC v2 0/6] SDM Interface [RFC v2 0/6] QEMU shared-memory-backend [RFC v2 1/1] backend: multi-client-socket Features : ---------- - When a domain A depends on a domain B (A=>B), signals are forward-propagated from A to B. So, when a user asks the domain A to be: - instantiated - destroyed - suspended - resumed domain B will automatically receive the same signal. - Single-level dependency: By construction domain dependencies are restricted to one single level, so to avoid wrong recursive (infinite) domain dependencies. This is also because there is no clear benefit in supporting multiple level of dependencies between domains. More specifically, given to domains, A and B, if A depends on B (A=>B), then B can not have any dependency (B => C is forbidden). How to test this patch series : ------------------------------- The source code can be downloaded by: git clone https://git.virtualopensystems.com/dev/libvirt.git \ -b domain-dependency Note: The patch series relies on the commit 95ca4fe, but with the patch proposed by Michal Privoznik, "[PATCH]qemuBuildVideoCommandLine: Don't access def->videos without check" that resolves a critical bug. To add a dependency between two domains (for example between domaninA and domainB), it suffices to add in the XML description of the first one: domainA.xml: <devices> ... <domaindependency> /absolute/path/to/the/domainB.xml </domaindependency> ... </devices> Note: The two domains mus be virtualized thanks to QEMU, and be hosted on the same filesystem. Finally, instanciate domainA with the command: virsh create domainA.xml And observe that both domains are running in the same time with: virsh list You can trigger the implemented signals and check that both domains are working as one: virsh suspend domainA virsh resume domainA virsh destroy domainA Note: as mentioned, all signals are forward-propagated, so if you manipulate the domainB directly, domainA will not be affected. This work has been sponsored by Huawei Technologies Duesseldorf GmbH. Valentin BOUSSON (4): Add XML description of the new dependency device Add the inner structure managing dependency devices Add primitives for manipulation of dependency structures Implement signal propagation to dependancy domains docs/schemas/domaincommon.rng | 5 ++ src/conf/domain_conf.c | 83 +++++++++++++++++++++++++- src/conf/domain_conf.h | 18 ++++++ src/qemu/qemu_driver.c | 135 ++++++++++++++++++++++++++++++++++++++++-- src/qemu/qemu_hotplug.c | 1 + 5 files changed, 237 insertions(+), 5 deletions(-) -- 1.9.1

2 5

[libvirt] [PATCH] qemuxml2argv: Mock virSCSIDeviceGetSgName
by Michal Privoznik 23 Mar '16

23 Mar '16

When constructing SCSI hostdev command line for qemu, the /sys/bus/scsi/devices/... dir is scanned. Unfortunately, even in the tests. This is needed to determine the name of SCSI device to passthrough to qemu, because in the domain XML we were given its address instead. Anyway, we should not be touching live system data in our test suite as it produced unpredictable results. The test is regressing from 1e9a083742efe on. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- tests/qemuxml2argvmock.c | 14 ++++++++++++++ 1 file changed, 14 insertions(+) diff --git a/tests/qemuxml2argvmock.c b/tests/qemuxml2argvmock.c index 8426108..b7dfebb 100644 --- a/tests/qemuxml2argvmock.c +++ b/tests/qemuxml2argvmock.c @@ -26,6 +26,7 @@ #include "virutil.h" #include "virstring.h" #include "virtpm.h" +#include "virscsi.h" #include <time.h> #include <unistd.h> @@ -84,3 +85,16 @@ virMemoryMaxValue(bool capped ATTRIBUTE_UNUSED) { return LLONG_MAX; } + +char * +virSCSIDeviceGetSgName(const char *sysfs_prefix ATTRIBUTE_UNUSED, + const char *adapter ATTRIBUTE_UNUSED, + unsigned int bus ATTRIBUTE_UNUSED, + unsigned int target ATTRIBUTE_UNUSED, + unsigned long long unit ATTRIBUTE_UNUSED) +{ + char *ret; + + ignore_value(VIR_STRDUP(ret, "sg0")); + return ret; +} -- 2.7.3

2 1

[libvirt] [libvirt-sandbox PATCH] Ignore generated file
by Guido Günther 23 Mar '16

23 Mar '16

--- .gitignore | 1 + 1 file changed, 1 insertion(+) diff --git a/.gitignore b/.gitignore index 390d65c..83831bb 100644 --- a/.gitignore +++ b/.gitignore @@ -31,6 +31,7 @@ po/stamp-it po/libvirt-sandbox.pot po/*.gmo stamp-h1 +build-aux/test-driver INSTALL libvirt-sandbox-1.0.pc libvirt-sandbox.spec -- 2.8.0.rc3

1 0

[libvirt] [PATCH 1/3 trivial] fix typo in virt-aa-helper helptext
by Serge Hallyn 23 Mar '16

23 Mar '16

it's --dryrun not --dry-run Signed-off-by: Serge Hallyn <serge.hallyn(a)ubuntu.com> --- src/security/virt-aa-helper.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/security/virt-aa-helper.c b/src/security/virt-aa-helper.c index a2d7226..b466626 100644 --- a/src/security/virt-aa-helper.c +++ b/src/security/virt-aa-helper.c @@ -107,7 +107,7 @@ vah_usage(void) " Options:\n" " -a | --add load profile\n" " -c | --create create profile from template\n" - " -d | --dry-run dry run\n" + " -d | --dryrun dry run\n" " -D | --delete unload and delete profile\n" " -f | --add-file <file> add file to profile\n" " -F | --append-file <file> append file to profile\n" -- 2.7.0

3 5

[libvirt] [libvirt-sandbox PATCH] Allow to switch to a specific user id
by Guido Günther 23 Mar '16

23 Mar '16

Useful if you want to run programs as the same user than outside the sandbox. --- bin/virt-sandbox.c | 26 ++++++++++++++++++++++++++ 1 file changed, 26 insertions(+) diff --git a/bin/virt-sandbox.c b/bin/virt-sandbox.c index 4c400d5..9495e85 100644 --- a/bin/virt-sandbox.c +++ b/bin/virt-sandbox.c @@ -24,6 +24,8 @@ #include <libvirt-sandbox/libvirt-sandbox.h> #include <glib/gi18n.h> +#include <sys/types.h> +#include <pwd.h> static gboolean do_close(GVirSandboxConsole *con G_GNUC_UNUSED, gboolean error G_GNUC_UNUSED, @@ -92,6 +94,7 @@ int main(int argc, char **argv) { gchar *kernver = NULL; gchar *kernpath = NULL; gchar *kmodpath = NULL; + gchar *switchto = NULL; gboolean verbose = FALSE; gboolean debug = FALSE; gboolean shell = FALSE; @@ -126,6 +129,8 @@ int main(int argc, char **argv) { N_("security properties"), "PATH", }, { "privileged", 'p', 0, G_OPTION_ARG_NONE, &privileged, N_("run the command privileged"), NULL }, + { "switchto", 'S', 0, G_OPTION_ARG_STRING, &switchto, + N_("swith to the given user"), "USER" }, { "shell", 'l', 0, G_OPTION_ARG_NONE, &shell, N_("start a shell"), NULL, }, { "kernver", 0, 0, G_OPTION_ARG_STRING, &kernver, @@ -139,6 +144,7 @@ int main(int argc, char **argv) { { NULL, 0, 0, G_OPTION_ARG_NONE, NULL, NULL, NULL } }; const char *help_msg = N_("Run 'virt-sandbox --help' to see a full list of available command line options"); + struct passwd *pw; setlocale(LC_ALL, ""); bindtextdomain(PACKAGE, LOCALEDIR); @@ -198,10 +204,25 @@ int main(int argc, char **argv) { if (kmodpath) gvir_sandbox_config_set_kmodpath(cfg, kmodpath); + if (privileged && switchto) { + g_printerr(_("'switchto' and 'privileged' are incompatible options\n")); + goto cleanup; + } + if (privileged) { gvir_sandbox_config_set_userid(cfg, 0); gvir_sandbox_config_set_groupid(cfg, 0); gvir_sandbox_config_set_username(cfg, "root"); + } else if (switchto) { + pw = getpwnam(switchto); + if (!pw) { + g_printerr(_("Failed to resolve user %s\n"), switchto); + goto cleanup; + } + gvir_sandbox_config_set_userid(cfg, pw->pw_uid); + gvir_sandbox_config_set_groupid(cfg, pw->pw_gid); + gvir_sandbox_config_set_username(cfg, pw->pw_name); + gvir_sandbox_config_set_homedir(cfg, pw->pw_dir); } if (envs && @@ -541,6 +562,11 @@ to this path to locate the modules. Retain root privileges inside the sandbox, rather than dropping privileges to match the current user identity. +=item B<-S USER>, B<--switchto=USER> + +Swith to the given user inside the sandbox and setup $HOME +accordingly. + =item B<-l>, B<--shell> Launch an interactive shell on a secondary console device -- 2.8.0.rc3

2 2

[libvirt] [PATCH] nss: use more strict check for automatic enabling
by Roman Bogorodskiy 23 Mar '16

23 Mar '16

Currently nss plugin is automatically enabled if nss.h header exists and --without-nss-plugin is not specified. However, presence of the nss.h header does not guarantee a compatible interface, so check for struct gaih_addrtuple to make sure it's enabled only on platforms where it could be built. --- m4/virt-nss.m4 | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/m4/virt-nss.m4 b/m4/virt-nss.m4 index 3fa4ad3..bc054db 100644 --- a/m4/virt-nss.m4 +++ b/m4/virt-nss.m4 @@ -25,12 +25,14 @@ AC_DEFUN([LIBVIRT_CHECK_NSS],[ fail=0 if test "x$with_nss_plugin" != "xno" ; then - AC_CHECK_HEADERS([nss.h], [ + AC_CHECK_TYPE([struct gaih_addrtuple], [ with_nss_plugin=yes ],[ if test "x$with_nss_plugin" = "xyes" ; then fail = 1 fi + ],[ + #include <nss.h> ]) if test $fail = 1 ; then -- 2.4.6

2 3

[libvirt] Does the proposed solution belong in KVM input devices?
by bancfc＠openmailbox.org 23 Mar '16

23 Mar '16

qemu-discuss doesn't seem the best place to reach the devs. I saw most of the work on KVM input devices was done by Gerd Hoffman and Dan Berrange. Dan do you think the QEMU virtual keyboard is the right place for a solution similar to the one described below? https://paul.reviews/behavioral-profiling-the-password-you-cant-change/ Would this be considered feature creep and best implemented in its own program?

3 3

[libvirt] [libvirt-php 0/3] small improvements
by Vasiliy Tolstov 22 Mar '16

22 Mar '16

Add some more libvirt functions support to binding. Also add some checks to prevent malloc SIZE_MAX memory in case of errors Vasiliy Tolstov (3): add some checks to prevent overflow add block_commit support and needed const add libvirt_domain_block_job_info src/libvirt-php.c | 152 ++++++++++++++++++++++++++++++++++++++++++++++++------ src/libvirt-php.h | 2 + 2 files changed, 138 insertions(+), 16 deletions(-) -- 2.7.0

4 12

[libvirt] [PATCH 00/12] qemu_process cleanups and test improvement
by Pavel Hrdina 22 Mar '16

22 Mar '16

The first patch is unrelated but I've found it while working on this series. Pavel Hrdina (12): qemu_process: check for correct return value while starting domain nvram: generate it's path in qemuDomainDefPostParse tests: cleanup qemuxml2argvtest qemu_process: introduce qemuProcessPrepareDomain qemu_process: introduce qemuProcessPrepareHost qemu: update callers of qemuProcessStartValidate to use virDomainObjPtr qemu_process: move checks to qemuProcessStartValidate qemu_process: move qemuDomainSetPrivatePaths to qemuProcessInit qemu_process: introduce qemuProcessCreateCmd qemu_driver: cleanup qemuConnectDomainXMLToNative qemuxml2argvtest: use driver.config and priv for qemuDomainSetPrivatePaths qemuxml2argvtest: cleanup test src/qemu/qemu_domain.c | 10 + src/qemu/qemu_driver.c | 61 +- src/qemu/qemu_migration.c | 9 +- src/qemu/qemu_process.c | 706 +++++++++++---------- src/qemu/qemu_process.h | 26 +- .../qemuxml2argv-channel-virtio-unix.xml | 2 +- tests/qemuxml2argvtest.c | 126 ++-- tests/testutilsqemu.c | 7 + 8 files changed, 494 insertions(+), 453 deletions(-) -- 2.7.2

3 27

[libvirt] [PATCH 0/2] qemu: support SPICE+unix socket
by Cole Robinson 22 Mar '16

22 Mar '16

qemu has supported SPICE+unix socket listening since 2.3.0. Patch #1 wires it up from domain_conf.c down to qemu, patch #2 adds qemu.conf spice_auto_unix_socket option Full disclosure: this duplicates the logic of VNC socket usage, which has some pre-existing warts: - When the VM is running with graphics socket=FOO, the runtime XML still lists <listen type='address' address='127.0.0.1'/>. - If using vnc/spice auto socket, it looks like ports can still be allocated in the code, though they are never used. - In fact I suspect there's lots of issues with stray XML port/listen being parsed and ending up in the device config, but selectively being ignored later. We should probably explicitly ignore/error about those combos when parsing the XML so we never have to ignore specific values later on. - The auto socket path is set in qemu_command.c but it should probably be set in qemu_process.c with the port stuff. Though there are references to vncListen in qemu_command.c which should probably be moved as well. The issue with moving them to qemu_process.c is now there isn't any pre-existing way to test those bits like via qemuxml2argvtest for example. Though finding a way to test the graphics prep stuff would be valuable anyways - teuf suggests: converting <graphics socket=X/> to <graphics><listen type='socket' path='FOO'/></graphics> like was done with listen= address - tuef: virsh domdisplay doesn't handle sockets To avoid yet another rabbit hole I didn't touch any of the above issues, which basically means spice+unix is no less operationally worse than vnc+unix :) Cole Robinson (2): qemu: Support SPICE listen over unix socket qemu: Add qemu.conf option spice_auto_unix_socket docs/schemas/domaincommon.rng | 5 ++++ src/conf/domain_conf.c | 26 +++++++++++++-------- src/conf/domain_conf.h | 1 + src/qemu/libvirtd_qemu.aug | 1 + src/qemu/qemu.conf | 11 +++++++++ src/qemu/qemu_command.c | 15 +++++++++--- src/qemu/qemu_conf.c | 1 + src/qemu/qemu_conf.h | 1 + src/qemu/qemu_process.c | 3 +++ src/qemu/test_libvirtd_qemu.aug.in | 1 + src/security/virt-aa-helper.c | 5 ++++ .../qemuxml2argv-graphics-spice-unix-auto.args | 21 +++++++++++++++++ .../qemuxml2argv-graphics-spice-unix-auto.xml | 27 ++++++++++++++++++++++ .../qemuxml2argv-graphics-spice-unix.args | 21 +++++++++++++++++ .../qemuxml2argv-graphics-spice-unix.xml | 27 ++++++++++++++++++++++ tests/qemuxml2argvtest.c | 10 ++++++++ 16 files changed, 163 insertions(+), 13 deletions(-) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-graphics-spice-unix-auto.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-graphics-spice-unix-auto.xml create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-graphics-spice-unix.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-graphics-spice-unix.xml -- 2.5.0

3 9

[libvirt] [PATCH v2] qemu-hotplug: fix eject media
by Pavel Hrdina 22 Mar '16

22 Mar '16

QEMU changed the error message to: "Tray of device 'drive-sata0-0-1' is not open" and they may change the error massage in the future. This updates the code to not depend on the text from the error message but only on error itself. Signed-off-by: Pavel Hrdina <phrdina(a)redhat.com> --- src/qemu/qemu_hotplug.c | 11 +++++------ src/qemu/qemu_monitor_json.c | 12 +----------- src/qemu/qemu_monitor_text.c | 5 +---- 3 files changed, 7 insertions(+), 21 deletions(-) diff --git a/src/qemu/qemu_hotplug.c b/src/qemu/qemu_hotplug.c index b580283..a78bc60 100644 --- a/src/qemu/qemu_hotplug.c +++ b/src/qemu/qemu_hotplug.c @@ -202,15 +202,14 @@ qemuDomainChangeEjectableMedia(virQEMUDriverPtr driver, if (qemuDomainObjExitMonitor(driver, vm) < 0) goto cleanup; - if (rc == -2) { + if (rc < 0) { /* we've already tried, error out */ if (ejectRetry) goto error; - VIR_DEBUG("tray is locked, wait for the guest to unlock " - "the tray and try to eject it again"); + ejectRetry = true; - } else if (rc < 0) { - goto error; + VIR_DEBUG("tray may be locked, wait for the guest to unlock " + "the tray and try to eject it again"); } if (virTimeMillisNow(&now) < 0) @@ -220,7 +219,7 @@ qemuDomainChangeEjectableMedia(virQEMUDriverPtr driver, if (virDomainObjWaitUntil(vm, now + CHANGE_MEDIA_TIMEOUT) != 0) goto error; } - } while (ejectRetry && rc != 0); + } while (rc < 0); if (!virStorageSourceIsEmpty(newsrc)) { if (qemuGetDriveSourceString(newsrc, conn, &sourcestr) < 0) diff --git a/src/qemu/qemu_monitor_json.c b/src/qemu/qemu_monitor_json.c index fe9288a..57c34f0 100644 --- a/src/qemu/qemu_monitor_json.c +++ b/src/qemu/qemu_monitor_json.c @@ -2272,8 +2272,7 @@ int qemuMonitorJSONSetCPU(qemuMonitorPtr mon, * Run QMP command to eject a media from ejectable device. * * Returns: - * -2 on error, when the tray is locked - * -1 on all other errors + * -1 on error * 0 on success */ int qemuMonitorJSONEjectMedia(qemuMonitorPtr mon, @@ -2294,15 +2293,6 @@ int qemuMonitorJSONEjectMedia(qemuMonitorPtr mon, if (ret == 0) ret = qemuMonitorJSONCheckError(cmd, reply); - if (ret < 0) { - virJSONValuePtr error = virJSONValueObjectGet(reply, "error"); - if (error) { - const char *errorStr = virJSONValueObjectGetString(error, "desc"); - if (errorStr && c_strcasestr(errorStr, "is locked")) - ret = -2; - } - } - virJSONValueFree(cmd); virJSONValueFree(reply); return ret; diff --git a/src/qemu/qemu_monitor_text.c b/src/qemu/qemu_monitor_text.c index bb87397..3129427 100644 --- a/src/qemu/qemu_monitor_text.c +++ b/src/qemu/qemu_monitor_text.c @@ -1170,8 +1170,7 @@ int qemuMonitorTextSetCPU(qemuMonitorPtr mon, int cpu, bool online) * Run HMP command to eject a media from ejectable device. * * Returns: - * -2 on error, when the tray is locked - * -1 on all other errors + * -1 on error * 0 on success */ int qemuMonitorTextEjectMedia(qemuMonitorPtr mon, @@ -1192,8 +1191,6 @@ int qemuMonitorTextEjectMedia(qemuMonitorPtr mon, * device not found, device is locked ... * No message is printed on success it seems */ if (c_strcasestr(reply, "device ")) { - if (c_strcasestr(reply, "is locked")) - ret = -2; virReportError(VIR_ERR_OPERATION_FAILED, _("could not eject media on %s: %s"), dev_name, reply); goto cleanup; -- 2.7.4

2 1

[libvirt] [PATCH] qemu-hotplug: fix eject media
by Pavel Hrdina 22 Mar '16

22 Mar '16

QEMU changed the error message to: "Tray of device 'drive-sata0-0-1' is not open" and they may change the error massage in the future. This updates the code to not depend on the text from the error message but only on error itself. One more improvement is that we will store the original error in case if it's not about tray is not open. Signed-off-by: Pavel Hrdina <phrdina(a)redhat.com> --- src/qemu/qemu_hotplug.c | 17 +++++++++-------- src/qemu/qemu_monitor_json.c | 12 +----------- src/qemu/qemu_monitor_text.c | 5 +---- 3 files changed, 11 insertions(+), 23 deletions(-) diff --git a/src/qemu/qemu_hotplug.c b/src/qemu/qemu_hotplug.c index b580283..8122367 100644 --- a/src/qemu/qemu_hotplug.c +++ b/src/qemu/qemu_hotplug.c @@ -173,8 +173,8 @@ qemuDomainChangeEjectableMedia(virQEMUDriverPtr driver, qemuDomainObjPrivatePtr priv = vm->privateData; const char *format = NULL; char *sourcestr = NULL; - bool ejectRetry = false; unsigned long long now; + virErrorPtr ejectError = NULL; if (!disk->info.alias) { virReportError(VIR_ERR_INTERNAL_ERROR, @@ -202,15 +202,16 @@ qemuDomainChangeEjectableMedia(virQEMUDriverPtr driver, if (qemuDomainObjExitMonitor(driver, vm) < 0) goto cleanup; - if (rc == -2) { + if (rc < 0) { /* we've already tried, error out */ - if (ejectRetry) + if (ejectError) { + virSetError(ejectError); + virFreeError(ejectError); goto error; - VIR_DEBUG("tray is locked, wait for the guest to unlock " + } + ejectError = virSaveLastError(); + VIR_DEBUG("tray may be locked, wait for the guest to unlock " "the tray and try to eject it again"); - ejectRetry = true; - } else if (rc < 0) { - goto error; } if (virTimeMillisNow(&now) < 0) @@ -220,7 +221,7 @@ qemuDomainChangeEjectableMedia(virQEMUDriverPtr driver, if (virDomainObjWaitUntil(vm, now + CHANGE_MEDIA_TIMEOUT) != 0) goto error; } - } while (ejectRetry && rc != 0); + } while (rc < 0); if (!virStorageSourceIsEmpty(newsrc)) { if (qemuGetDriveSourceString(newsrc, conn, &sourcestr) < 0) diff --git a/src/qemu/qemu_monitor_json.c b/src/qemu/qemu_monitor_json.c index 8352e53..66a2922 100644 --- a/src/qemu/qemu_monitor_json.c +++ b/src/qemu/qemu_monitor_json.c @@ -2272,8 +2272,7 @@ int qemuMonitorJSONSetCPU(qemuMonitorPtr mon, * Run QMP command to eject a media from ejectable device. * * Returns: - * -2 on error, when the tray is locked - * -1 on all other errors + * -1 on error * 0 on success */ int qemuMonitorJSONEjectMedia(qemuMonitorPtr mon, @@ -2294,15 +2293,6 @@ int qemuMonitorJSONEjectMedia(qemuMonitorPtr mon, if (ret == 0) ret = qemuMonitorJSONCheckError(cmd, reply); - if (ret < 0) { - virJSONValuePtr error = virJSONValueObjectGet(reply, "error"); - if (error) { - const char *errorStr = virJSONValueObjectGetString(error, "desc"); - if (errorStr && c_strcasestr(errorStr, "is locked")) - ret = -2; - } - } - virJSONValueFree(cmd); virJSONValueFree(reply); return ret; diff --git a/src/qemu/qemu_monitor_text.c b/src/qemu/qemu_monitor_text.c index bb87397..3129427 100644 --- a/src/qemu/qemu_monitor_text.c +++ b/src/qemu/qemu_monitor_text.c @@ -1170,8 +1170,7 @@ int qemuMonitorTextSetCPU(qemuMonitorPtr mon, int cpu, bool online) * Run HMP command to eject a media from ejectable device. * * Returns: - * -2 on error, when the tray is locked - * -1 on all other errors + * -1 on error * 0 on success */ int qemuMonitorTextEjectMedia(qemuMonitorPtr mon, @@ -1192,8 +1191,6 @@ int qemuMonitorTextEjectMedia(qemuMonitorPtr mon, * device not found, device is locked ... * No message is printed on success it seems */ if (c_strcasestr(reply, "device ")) { - if (c_strcasestr(reply, "is locked")) - ret = -2; virReportError(VIR_ERR_OPERATION_FAILED, _("could not eject media on %s: %s"), dev_name, reply); goto cleanup; -- 2.7.3

2 2

[libvirt] [PATCH v4] Use-correct-pci-addresses-during-interface-detach
by Nitesh Konkar 22 Mar '16

22 Mar '16

The virsh attach virsh detach interface command fails when both live and config are set and when the interface gets attached to different pci slots on live and config xml respectively. When we attach an interface with both --live and --config, the first time they get the same PCI slots, but the second time onwards it differs and hence the virsh detach-interface --live --config command fails. This patch makes sure that when both --live --config are set , qemuDomainDetachDeviceFlags is called twice, once with config xml and once with live xml. --- Change log: v4: * Code unchanged, updated with commit message,change log and steps to reproduce the issue.. v3: * Created function vshDomainDetachInterface and called it once/twice from cmdDetachInterface depending on number of flags set (live/config/both). Passed the correct xml(live/persistent) to it. v2: * Changes only in cmdDetachInterface to pass the right domain xml depending on the flag set (live/config/both). v1: * Changes only in qemuDomainDetachDeviceFlags to set the right value in dev and dev_copy. Steps to see the issue: virsh attach-interface --domain DomainName --type network --source default --mac 52:54:00:4b:76:5f --live --config virsh detach-interface --domain DomainName --type network --mac 52:54:00:4b:76:5f --live --config virsh attach-interface --domain DomainName --type network --source default --mac 52:54:00:4b:76:5f --live --config virsh detach-interface --domain DomainName --type network --mac 52:54:00:4b:76:5f --live --config tools/virsh-domain.c | 104 +++++++++++++++++++++++++++++++-------------------- 1 file changed, 64 insertions(+), 40 deletions(-) diff --git a/tools/virsh-domain.c b/tools/virsh-domain.c index 62acecb..a6abaf5 100644 --- a/tools/virsh-domain.c +++ b/tools/virsh-domain.c @@ -10801,40 +10801,21 @@ static const vshCmdOptDef opts_detach_interface[] = { {.name = NULL} }; -static bool -cmdDetachInterface(vshControl *ctl, const vshCmd *cmd) +static bool +vshDomainDetachInterface(char *doc, unsigned int flags, virDomainPtr dom, vshControl *ctl, const vshCmd *cmd) { - virDomainPtr dom = NULL; xmlDocPtr xml = NULL; xmlXPathObjectPtr obj = NULL; xmlXPathContextPtr ctxt = NULL; xmlNodePtr cur = NULL, matchNode = NULL; - char *detach_xml = NULL; const char *mac = NULL, *type = NULL; - char *doc = NULL; + char *detach_xml = NULL; + bool current = vshCommandOptBool(cmd, "current"); char buf[64]; int diff_mac; size_t i; int ret; bool functionReturn = false; - unsigned int flags = VIR_DOMAIN_AFFECT_CURRENT; - bool current = vshCommandOptBool(cmd, "current"); - bool config = vshCommandOptBool(cmd, "config"); - bool live = vshCommandOptBool(cmd, "live"); - bool persistent = vshCommandOptBool(cmd, "persistent"); - - VSH_EXCLUSIVE_OPTIONS_VAR(persistent, current); - - VSH_EXCLUSIVE_OPTIONS_VAR(current, live); - VSH_EXCLUSIVE_OPTIONS_VAR(current, config); - - if (config || persistent) - flags |= VIR_DOMAIN_AFFECT_CONFIG; - if (live) - flags |= VIR_DOMAIN_AFFECT_LIVE; - - if (!(dom = virshCommandOptDomain(ctl, cmd, NULL))) - return false; if (vshCommandOptStringReq(ctl, cmd, "type", &type) < 0) goto cleanup; @@ -10842,15 +10823,6 @@ cmdDetachInterface(vshControl *ctl, const vshCmd *cmd) if (vshCommandOptStringReq(ctl, cmd, "mac", &mac) < 0) goto cleanup; - if (persistent && - virDomainIsActive(dom) == 1) - flags |= VIR_DOMAIN_AFFECT_LIVE; - - if (flags & VIR_DOMAIN_AFFECT_CONFIG) - doc = virDomainGetXMLDesc(dom, VIR_DOMAIN_XML_INACTIVE); - else - doc = virDomainGetXMLDesc(dom, 0); - if (!doc) goto cleanup; @@ -10918,23 +10890,75 @@ cmdDetachInterface(vshControl *ctl, const vshCmd *cmd) else ret = virDomainDetachDevice(dom, detach_xml); - if (ret != 0) { - vshError(ctl, "%s", _("Failed to detach interface")); - } else { - vshPrint(ctl, "%s", _("Interface detached successfully\n")); + if (ret == 0) functionReturn = true; - } cleanup: - VIR_FREE(doc); VIR_FREE(detach_xml); - virDomainFree(dom); + xmlFreeDoc(xml); xmlXPathFreeObject(obj); xmlXPathFreeContext(ctxt); - xmlFreeDoc(xml); return functionReturn; } +static bool +cmdDetachInterface(vshControl *ctl, const vshCmd *cmd) +{ + virDomainPtr dom = NULL; + unsigned int flags = VIR_DOMAIN_AFFECT_CURRENT; + char *doc_live = NULL, *doc_config = NULL; + bool current = vshCommandOptBool(cmd, "current"); + bool config = vshCommandOptBool(cmd, "config"); + bool live = vshCommandOptBool(cmd, "live"); + bool persistent = vshCommandOptBool(cmd, "persistent"); + bool functionReturn = false; + + VSH_EXCLUSIVE_OPTIONS_VAR(persistent, current); + + VSH_EXCLUSIVE_OPTIONS_VAR(current, live); + VSH_EXCLUSIVE_OPTIONS_VAR(current, config); + + if (config || persistent) + flags |= VIR_DOMAIN_AFFECT_CONFIG; + + if (!(dom = virshCommandOptDomain(ctl, cmd, NULL))) + return false; + + if (persistent && + virDomainIsActive(dom) == 1) + flags |= VIR_DOMAIN_AFFECT_LIVE; + + if (flags & VIR_DOMAIN_AFFECT_CONFIG) { + doc_config = virDomainGetXMLDesc(dom, VIR_DOMAIN_XML_INACTIVE); + if (!(functionReturn = vshDomainDetachInterface(doc_config, flags, dom, ctl, cmd))) + goto cleanup; + } + + if (live) + flags |= VIR_DOMAIN_AFFECT_LIVE; + + if (flags & VIR_DOMAIN_AFFECT_LIVE) { + doc_live = virDomainGetXMLDesc(dom, 0); + + if (flags & VIR_DOMAIN_AFFECT_CONFIG) + flags &= (~VIR_DOMAIN_AFFECT_CONFIG); + + functionReturn = vshDomainDetachInterface(doc_live, flags, dom, ctl, cmd); + } + + cleanup: + if (functionReturn == false) { + vshError(ctl, "%s", _("Failed to detach interface")); + } else { + vshPrint(ctl, "%s", _("Interface detached successfully\n")); + functionReturn = true; + } + VIR_FREE(doc_live); + VIR_FREE(doc_config); + virDomainFree(dom); + return functionReturn; +} + typedef enum { VIRSH_FIND_DISK_NORMAL, VIRSH_FIND_DISK_CHANGEABLE, -- 1.8.3.1

1 1

Re: [libvirt] [PATCH v5 0/5] ARM: add query-gic-capabilities SMP command
by Peter Xu 22 Mar '16

22 Mar '16

On Mon, Mar 21, 2016 at 04:56:07PM +0100, Andrea Bolognani wrote: > On Fri, 2016-03-18 at 11:27 +0800, Peter Xu wrote: > > v5 changes: > > - patch 2: moved to target-arm/monitor.c (from target-arm/machine.c) > > [Peter] > > - patch 3: splitted into three patches: [all from Peter's comments] > > - patch 3 (new): leverage kvm_arm_create_scratch_host_vcpu(), tiny > > enhancement of old one to suite our need > > - patch 4: introduce kvm_support_device() in kvm-all.c > > - patch 5: do the implementation. > > Tested on two separate aarch64 hosts, seems to work fine. Thanks Andrea! Sorry I forgot to CC libvir-list. To avoid duplication, not re-sending but CCing in this reply. If anyone interested, please review in the following link: https://lists.gnu.org/archive/html/qemu-devel/2016-03/msg04465.html (which points to exactly current thread.) Sorry for the inconvenience! -- peterx

1 0

[libvirt] virtual machine won't autostart when using LVM with cache
by David Hlacik 22 Mar '16

22 Mar '16

Hello guys, OS Fedora 23 with latest updates Recently I have switched to use LVM cache feature on logical volume /dev/hdd/windata1 to improve it performace using 32GB partition from SSD disk. However, when my computer will start, my virtual machine won't autostart Mar 21 10:48:57 brutus-coreos libvirtd[956]: Cannot access storage file '/dev/hdd/windata1' (as uid:107, gid:107): No such file or directory Mar 21 10:48:57 brutus-coreos libvirtd[956]: Failed to autostart VM 'winos1': Cannot access storage file '/dev/hdd/windata1' (as uid:107, gid:107): No such file or directory Mar 21 10:48:57 brutus-coreos libvirtd[956]: Cannot access storage file '/dev/hdd/windata1' (as uid:107, gid:107): No such file or directory It seems that when using LVM cache, one have to wait till it will inicialize? When I will manually start virtual machine afterwards, everything works OK : [root@brutus-coreos ~]# virsh start winos1 Domain winos1 started I have tried to remove LVM cache from /dev/hdd/windata1 and afterward autostart works! So it must be LVM cache related. Can you please help me to solve this ? Thanks in advance David Hlacik +420-777-307-745 | david(a)hlacik.cz [image: View David Hlacik's profile on LinkedIn] <http://cz.linkedin.com/in/hlacik>

2 2

[libvirt] [PATCH v2] nss: Make aligning look nicer
by Martin Kletzander 22 Mar '16

22 Mar '16

Every aligning requires at least one cast and it's hard to read. Let's make a function that makes sure the pointer is moved according to the alignment and use that to move throughout the data buffer. Signed-off-by: Martin Kletzander <mkletzan(a)redhat.com> --- v2: - Initialize @idx in _nss_libvirt_gethostbyname4_r as well tools/nss/libvirt_nss.c | 53 +++++++++++++++++++++++++++++++------------------ 1 file changed, 34 insertions(+), 19 deletions(-) diff --git a/tools/nss/libvirt_nss.c b/tools/nss/libvirt_nss.c index ba3bb31b3569..218c62a64c88 100644 --- a/tools/nss/libvirt_nss.c +++ b/tools/nss/libvirt_nss.c @@ -252,17 +252,31 @@ _nss_libvirt_gethostbyname2_r(const char *name, int af, struct hostent *result, errnop, herrnop, NULL, NULL); } +static inline void * +move_and_align(void *buf, size_t len, size_t *idx) +{ + char *buffer = buf; + size_t move = ALIGN(len); + + if (!idx) + return buffer + move; + + *idx += move; + + return buffer + *idx; +} + enum nss_status _nss_libvirt_gethostbyname3_r(const char *name, int af, struct hostent *result, char *buffer, size_t buflen, int *errnop, int *herrnop, int32_t *ttlp, char **canonp) { enum nss_status ret = NSS_STATUS_UNAVAIL; - char *r_name, **r_aliases, *r_addr, **r_addr_list; + char *r_name, **r_aliases, *r_addr, *r_addr_next, **r_addr_list; leaseAddress *addr = NULL; size_t naddr, i; bool found = false; - size_t nameLen, need, idx; + size_t nameLen, need, idx = 0; int alen; int r; @@ -319,23 +333,27 @@ _nss_libvirt_gethostbyname3_r(const char *name, int af, struct hostent *result, /* First, append name */ r_name = buffer; memcpy(r_name, name, nameLen + 1); - idx = ALIGN(nameLen + 1); + + r_aliases = move_and_align(buffer, nameLen + 1, &idx); /* Second, create aliases array */ - r_aliases = (char **) buffer + idx; r_aliases[0] = NULL; - idx += sizeof(char*); /* Third, append address */ - r_addr = buffer + idx; - for (i = 0; i < naddr; i++) - memcpy(r_addr + i * ALIGN(alen), addr[i].addr, alen); - idx += naddr * ALIGN(alen); + r_addr = move_and_align(buffer, sizeof(char *), &idx); + r_addr_next = r_addr; + for (i = 0; i < naddr; i++) { + memcpy(r_addr_next, addr[i].addr, alen); + r_addr_next = move_and_align(buffer, alen, &idx); + } + r_addr_list = move_and_align(buffer, 0, &idx); + r_addr_next = r_addr; /* Third, append address pointer array */ - r_addr_list = (char **) buffer + idx; - for (i = 0; i < naddr; i++) - r_addr_list[i] = r_addr + i * ALIGN(alen); + for (i = 0; i < naddr; i++) { + r_addr_list[i] = r_addr_next; + r_addr_next = move_and_align(r_addr_next, alen, NULL); + } r_addr_list[i] = NULL; idx += (naddr + 1) * sizeof(char*); @@ -375,7 +393,7 @@ _nss_libvirt_gethostbyname4_r(const char *name, struct gaih_addrtuple **pat, size_t naddr, i; bool found = false; int r; - size_t nameLen, need, idx; + size_t nameLen, need, idx = 0; struct gaih_addrtuple *r_tuple, *r_tuple_first = NULL; char *r_name; @@ -420,25 +438,22 @@ _nss_libvirt_gethostbyname4_r(const char *name, struct gaih_addrtuple **pat, /* First, append name */ r_name = buffer; memcpy(r_name, name, nameLen + 1); - idx = ALIGN(nameLen + 1); - /* Second, append addresses */ - r_tuple_first = (struct gaih_addrtuple*) (buffer + idx); + r_tuple_first = move_and_align(buffer, nameLen + 1, &idx); for (i = 0; i < naddr; i++) { int family = addr[i].af; - r_tuple = (struct gaih_addrtuple*) (buffer + idx); + r_tuple = move_and_align(buffer, 0, &idx); if (i == naddr - 1) r_tuple->next = NULL; else - r_tuple->next = (struct gaih_addrtuple*) ((char*) r_tuple + ALIGN(sizeof(struct gaih_addrtuple))); + r_tuple->next = move_and_align(buffer, sizeof(struct gaih_addrtuple), &idx); r_tuple->name = r_name; r_tuple->family = family; r_tuple->scopeid = 0; memcpy(r_tuple->addr, addr[i].addr, FAMILY_ADDRESS_SIZE(family)); - idx += ALIGN(sizeof(struct gaih_addrtuple)); } /* At this point, idx == need */ -- 2.7.3

3 3

[libvirt] [libvirt-sandbox PATCH] sandbox-builder: Don't try g_file_enumerator_next_file if enumerator is NULL
by Guido Günther 22 Mar '16

22 Mar '16

Passing the NULL enumerator to g_file_enumerator_next_file makes us overwrite error information otherwise: $ sudo bin/virt-sandbox-service create --unitfile foo.service foo (process:18529): GLib-GIO-CRITICAL **: g_file_enumerator_next_file: assertion 'G_IS_FILE_ENUMERATOR (enumerator)' failed (process:18529): GLib-GIO-CRITICAL **: g_file_enumerator_close: assertion 'G_IS_FILE_ENUMERATOR (enumerator)' failed bin/virt-sandbox-service:318: Warning: GError set over the top of a previous GError or uninitialized memory. This indicates a bug in someone's code. You must ensure an error is NULL before it's set. The overwriting error message was: Error removing file: No such file or directory context.undefine() bin/virt-sandbox-service:318: Warning: g_object_unref: assertion 'G_IS_OBJECT (object)' failed context.undefine() Cleanup failed: g-io-error-quark: No such file or directory (1) bin/virt-sandbox-service: g-io-error-quark: No such file or directory (1) also make sure we don't unref NULL in this case. --- libvirt-sandbox/libvirt-sandbox-builder.c | 37 +++++++++++++++++-------------- 1 file changed, 20 insertions(+), 17 deletions(-) diff --git a/libvirt-sandbox/libvirt-sandbox-builder.c b/libvirt-sandbox/libvirt-sandbox-builder.c index 1745c88..0c43b62 100644 --- a/libvirt-sandbox/libvirt-sandbox-builder.c +++ b/libvirt-sandbox/libvirt-sandbox-builder.c @@ -727,32 +727,35 @@ gboolean gvir_sandbox_builder_clean_post_stop(GVirSandboxBuilder *builder, errno != ENOENT) ret = FALSE; - if (!(enumerator = g_file_enumerate_children(libsFile, "*", G_FILE_QUERY_INFO_NONE, - NULL, error)) && - (*error)->code != G_IO_ERROR_NOT_FOUND) { - ret = FALSE; - goto cleanup; - } + if ((enumerator = g_file_enumerate_children(libsFile, "*", G_FILE_QUERY_INFO_NONE, + NULL, error))) { + while ((info = g_file_enumerator_next_file(enumerator, NULL, error))) { + child = g_file_enumerator_get_child(enumerator, info); + if (!g_file_delete(child, NULL, error)) + ret = FALSE; + g_object_unref(child); + child = NULL; + g_object_unref(info); + info = NULL; + } - while ((info = g_file_enumerator_next_file(enumerator, NULL, error))) { - child = g_file_enumerator_get_child(enumerator, info); - if (!g_file_delete(child, NULL, error)) + if (!g_file_enumerator_close(enumerator, NULL, error)) + ret = FALSE; + } else { + if ((*error)->code != G_IO_ERROR_NOT_FOUND) { ret = FALSE; - g_object_unref(child); - child = NULL; - g_object_unref(info); - info = NULL; + goto cleanup; + } + g_clear_error(error); } - if (!g_file_enumerator_close(enumerator, NULL, error)) - ret = FALSE; - if (!g_file_delete(libsFile, NULL, error) && (*error)->code != G_IO_ERROR_NOT_FOUND) ret = FALSE; cleanup: - g_object_unref(enumerator); + if (enumerator) + g_object_unref(enumerator); g_object_unref(libsFile); g_free(libsdir); g_free(dskfile); -- 2.7.0

2 2

[libvirt] [PATCH] tests: Produce predictable results in nsstest
by Michal Privoznik 22 Mar '16

22 Mar '16

Problem is that in the test any status file matching tests/nssdata/*.status is loaded as it contains IP addresses that are parsed. However, there's no order specified in which the files are loaded. Therefore on different systems the order may be different. This is then producing an unexpected results. Instead of defining an order in which the files are loaded, make the code that checks for missing IP addresses (or redundant ones) cope with unordered list of addresses. The reasoning behind is that the code doing the parsing is used in real NSS module where we don't care for ordering. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- tests/nsstest.c | 25 ++++++++++++------------- 1 file changed, 12 insertions(+), 13 deletions(-) diff --git a/tests/nsstest.c b/tests/nsstest.c index 68f1c60..9d7bc40 100644 --- a/tests/nsstest.c +++ b/tests/nsstest.c @@ -49,7 +49,7 @@ testGetHostByName(const void *opaque) char buf[BUF_SIZE] = { 0 }; char **addrList; int rv, tmp_errno = 0, tmp_herrno = 0; - size_t i = 0; + size_t i = 0, j = 0; if (!data) goto cleanup; @@ -141,30 +141,29 @@ testGetHostByName(const void *opaque) goto cleanup; } - if (!data->ipAddr[i]) { + for (j = 0; data->ipAddr[j]; j++) { + if (STREQ(data->ipAddr[j], ipAddr)) + break; + } + + if (!data->ipAddr[j]) { virReportError(VIR_ERR_INTERNAL_ERROR, "Unexpected address %s", ipAddr); VIR_FREE(ipAddr); goto cleanup; } - - if (STRNEQ(data->ipAddr[i], ipAddr)) { - virReportError(VIR_ERR_INTERNAL_ERROR, - "Address mismatch. Expected %s got %s", - data->ipAddr[i], ipAddr); - VIR_FREE(ipAddr); - goto cleanup; - } VIR_FREE(ipAddr); addrList++; i++; } - if (data->ipAddr[i]) { + for (j = 0; data->ipAddr[j]; j++) + ; + + if (i != j) { virReportError(VIR_ERR_INTERNAL_ERROR, - "Address mismatch. Expected %s got nothing", - data->ipAddr[i]); + "Expected %zu addresses, got %zu", j, i); goto cleanup; } -- 2.7.3

3 2

[libvirt] nss build failure on arm32
by Cole Robinson 22 Mar '16

22 Mar '16

Just a note, I'm getting build failures with the nss module on Fedora 23 arm 32bit: nss/libvirt_nss.c: In function '_nss_libvirt_gethostbyname3_r': nss/libvirt_nss.c:325:17: error: cast increases required alignment of target type [-Werror=cast-align] r_aliases = (char **) buffer + idx; ^ nss/libvirt_nss.c:336:19: error: cast increases required alignment of target type [-Werror=cast-align] r_addr_list = (char **) buffer + idx; ^ nss/libvirt_nss.c: In function '_nss_libvirt_gethostbyname4_r': nss/libvirt_nss.c:427:21: error: cast increases required alignment of target type [-Werror=cast-align] r_tuple_first = (struct gaih_addrtuple*) (buffer + idx); ^ nss/libvirt_nss.c:431:19: error: cast increases required alignment of target type [-Werror=cast-align] r_tuple = (struct gaih_addrtuple*) (buffer + idx); ^ nss/libvirt_nss.c:435:29: error: cast increases required alignment of target type [-Werror=cast-align] r_tuple->next = (struct gaih_addrtuple*) ((char*) r_tuple + ALIGN(sizeof(struct gaih_addrtuple))); - Cole

2 1