- Devel - libvirt List Archives

[libvirt] [PATCH] sanlock: Re-add lockspace unconditionally
by Michal Privoznik 14 Dec '12

14 Dec '12

Currently, if sanlock is already registering a lockspace other libvirtd instances (from other hosts) obtain -EINPROGRESS. On sufficiently new sanlock, sanlock_inq_lockspace() is called, which suspend execution until lockspace state is changed. With current libvirt implementation, we fail to retry adding the lockspace again but continue in error path. Therefore we produce meaningless error message: virLockManagerSanlockSetupLockspace:363 : Unable to add lockspace /var/lib/libvirt/sanlock/__LIBVIRT__DISKS__: Success qemudLoadDriverConfig:558 : Failed to load lock manager sanlock We should try to re-add the lockspace after its state change to be sure it was added successfully. In fact, with sufficiently new sanlock we can just avoid dummy usleep() which is used if there's no inquire API. --- src/locking/lock_driver_sanlock.c | 20 ++++++++------------ 1 files changed, 8 insertions(+), 12 deletions(-) diff --git a/src/locking/lock_driver_sanlock.c b/src/locking/lock_driver_sanlock.c index 430e11e..a072343 100644 --- a/src/locking/lock_driver_sanlock.c +++ b/src/locking/lock_driver_sanlock.c @@ -197,9 +197,7 @@ static int virLockManagerSanlockSetupLockspace(void) struct sanlk_lockspace ls; char *path = NULL; char *dir = NULL; -#ifndef HAVE_SANLOCK_INQ_LOCKSPACE int retries = LOCKSPACE_RETRIES; -#endif if (virAsprintf(&path, "%s/%s", driver->autoDiskLeasePath, @@ -332,25 +330,23 @@ static int virLockManagerSanlockSetupLockspace(void) * either call a sanlock API that blocks us until lockspace changes state, * or we can fallback to polling. */ -#ifndef HAVE_SANLOCK_INQ_LOCKSPACE retry: -#endif if ((rv = sanlock_add_lockspace(&ls, 0)) < 0) { - if (-rv == EINPROGRESS) { + if (-rv == EINPROGRESS && --retries) { #ifdef HAVE_SANLOCK_INQ_LOCKSPACE /* we have this function which blocks until lockspace change the * state. It returns 0 if lockspace has been added, -ENOENT if it - * hasn't. XXX should we goto retry? */ + * hasn't. */ VIR_DEBUG("Inquiring lockspace"); - rv = sanlock_inq_lockspace(&ls, SANLK_INQ_WAIT); + if (sanlock_inq_lockspace(&ls, SANLK_INQ_WAIT) < 0) + VIR_DEBUG("Unable to inquire lockspace"); #else /* fall back to polling */ - if (retries--) { - usleep(LOCKSPACE_SLEEP * 1000); - VIR_DEBUG("Retrying to add lockspace (left %d)", retries); - goto retry; - } + VIR_DEBUG("Sleeping for %dms", LOCKSPACE_SLEEP); + usleep(LOCKSPACE_SLEEP * 1000); #endif + VIR_DEBUG("Retrying to add lockspace (left %d)", retries); + goto retry; } if (-rv != EEXIST) { if (rv <= -200) -- 1.7.8.6

2 2

[libvirt] [PATCH] install: fix virtlockd installation
by Eric Blake 14 Dec '12

14 Dec '12

The virtlockd daemon scripts were lousy, when compared to their counterparts in daemon/Makefile.am. In particular, when init scripts were selected, this resulted in 'make distcheck' failing due to failure to clean up src/virtlockd.init. * src/Makefile.am (install-systemd): Fix dependencies. Use MKDIR_P. (uninstall-systemd): Remove empty directory. Use fewer processes. (install-init, install-sysconfig): Use MKDIR_P. (uninstall-init): Remove correct file, and also empty directory. (uninstall-sysconfig): Remove empty directory. (CLEANFILES): Clean up built sources. --- src/Makefile.am | 18 +++++++++++------- 1 file changed, 11 insertions(+), 7 deletions(-) diff --git a/src/Makefile.am b/src/Makefile.am index 25b6e3f..dd7e219 100644 --- a/src/Makefile.am +++ b/src/Makefile.am @@ -1628,24 +1628,26 @@ EXTRA_DIST += locking/virtlockd.sysconf \ locking/test_libvirt_lockd.aug.in install-sysconfig: - mkdir -p $(DESTDIR)$(sysconfdir)/sysconfig + $(MKDIR_P) $(DESTDIR)$(sysconfdir)/sysconfig $(INSTALL_DATA) $(srcdir)/locking/virtlockd.sysconf \ $(DESTDIR)$(sysconfdir)/sysconfig/virtlockd uninstall-sysconfig: rm -f $(DESTDIR)$(sysconfdir)/sysconfig/virtlockd + rmdir $(DESTDIR)$(sysconfdir)/sysconfig || : EXTRA_DIST += locking/virtlockd.init.in if WITH_LIBVIRTD if LIBVIRT_INIT_SCRIPT_RED_HAT install-init:: virtlockd.init install-sysconfig - mkdir -p $(DESTDIR)$(sysconfdir)/rc.d/init.d + $(MKDIR_P) $(DESTDIR)$(sysconfdir)/rc.d/init.d $(INSTALL_SCRIPT) virtlockd.init \ $(DESTDIR)$(sysconfdir)/rc.d/init.d/virtlockd uninstall-init:: uninstall-sysconfig - rm -f $(DESTDIR)$(sysconfdir)/rc.d/init.d/libvirtd + rm -f $(DESTDIR)$(sysconfdir)/rc.d/init.d/virtlockd + rmdir $(DESTDIR)$(sysconfdir)/rc.d/init.d || : BUILT_SOURCES += virtlockd.init else @@ -1677,16 +1679,17 @@ SYSTEMD_UNIT_DIR = /lib/systemd/system BUILT_SOURCES += virtlockd.service virtlockd.socket -install-systemd: virtlockd.init install-sysconfig - mkdir -p $(DESTDIR)$(SYSTEMD_UNIT_DIR) +install-systemd: virtlockd.service virtlockd.socket install-sysconfig + $(MKDIR_P) $(DESTDIR)$(SYSTEMD_UNIT_DIR) $(INSTALL_SCRIPT) virtlockd.service \ $(DESTDIR)$(SYSTEMD_UNIT_DIR)/ $(INSTALL_SCRIPT) virtlockd.socket \ $(DESTDIR)$(SYSTEMD_UNIT_DIR)/ uninstall-systemd: uninstall-sysconfig - rm -f $(DESTDIR)$(SYSTEMD_UNIT_DIR)/virtlockd.service - rm -f $(DESTDIR)$(SYSTEMD_UNIT_DIR)/virtlockd.socket + rm -f $(DESTDIR)$(SYSTEMD_UNIT_DIR)/virtlockd.service \ + $(DESTDIR)$(SYSTEMD_UNIT_DIR)/virtlockd.socket + rmdir $(DESTDIR)$(SYSTEMD_UNIT_DIR) || : else install-systemd: uninstall-systemd: @@ -2050,5 +2053,6 @@ endif rmdir "$(DESTDIR)$(localstatedir)/lib/libvirt" ||: CLEANFILES += *.gcov .libs/*.gcda .libs/*.gcno *.gcno *.gcda *.i *.s +CLEANFILES += $(BUILT_SOURCES) DISTCLEANFILES += $(GENERATED_SYM_FILES) MAINTAINERCLEANFILES += $(REMOTE_DRIVER_GENERATED) $(VIR_NET_RPC_GENERATED) -- 1.8.0.2

3 6

[libvirt] [PATCH] docs: Fix location of libvirt.conf and auth.conf
by Michał Łomnicki 14 Dec '12

14 Dec '12

For a unprivileged user libvirt.conf and auth.conf are looked up in $XDG_CONFIG_HOME but the docs incorrectly state that it's $XDG_CONFIG_DIR. --- docs/auth.html.in | 2 +- docs/uri.html.in | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/docs/auth.html.in b/docs/auth.html.in index d792eb1..929afd0 100644 --- a/docs/auth.html.in +++ b/docs/auth.html.in @@ -25,7 +25,7 @@ for the authentication file using the following sequence: variable.</li> <li>The file path specified by the "authfile=/some/file" URI query parameter</li> - <li>The file $XDG_CONFIG_DIR/libvirt/auth.conf</li> + <li>The file $XDG_CONFIG_HOME/libvirt/auth.conf</li> <li>The file /etc/libvirt/auth.conf</li> </ol> diff --git a/docs/uri.html.in b/docs/uri.html.in index 5812ca9..97a1cdb 100644 --- a/docs/uri.html.in +++ b/docs/uri.html.in @@ -30,7 +30,7 @@ virConnectPtr conn = virConnectOpenReadOnly (<b>"test:///default"</b>); <p> To simplify life for administrators, it is possible to setup URI aliases in a libvirt client configuration file. The configuration file is <code>/etc/libvirt/libvirt.conf</code> -for the root user, or <code>$XDG_CONFIG_DIR/libvirt/libvirt.conf</code> for any unprivileged user. +for the root user, or <code>$XDG_CONFIG_HOME/libvirt/libvirt.conf</code> for any unprivileged user. In this file, the following syntax can be used to setup aliases </p> -- 1.7.9.5

2 1

[libvirt] [PATCH 0/2] Another round of build fixes for lockd
by Jiri Denemark 14 Dec '12

14 Dec '12

make rpm finally succeeds with these two patches. Jiri Denemark (2): build: Install both qemu-lockd.conf and qemu-sanlock.conf spec: Include lockd files in libvirt-daemon package libvirt.spec.in | 4 ++++ src/Makefile.am | 5 +++-- 2 files changed, 7 insertions(+), 2 deletions(-) -- 1.8.0.2

2 5

[libvirt] [test-API][PATCH v2] Add 2 vcpupin cases cover config and live flags
by Wayne Sun 14 Dec '12

14 Dec '12

v1: add 2 vcpupin cases * use pinVcpuFlags to pin domain vcpu to host cpu * 2 cases cover config and live flags * cpulist with '^', '-' and ',' is supported to give multiple host cpus * vcpus and vcpuPinInfo are used as part of the checking * a sample conf is added v2: move format cpulist functions to utils * the format cpulist functions could be reused for cases need parse param with '-', '^' and ','. Signed-off-by: Wayne Sun <gsun(a)redhat.com> --- cases/vcpupin.conf | 67 +++++++++++++++++++++++ repos/setVcpus/vcpupin_config.py | 109 ++++++++++++++++++++++++++++++++++++++ repos/setVcpus/vcpupin_live.py | 101 +++++++++++++++++++++++++++++++++++ utils/utils.py | 71 ++++++++++++++++++++++++ 4 files changed, 348 insertions(+), 0 deletions(-) create mode 100644 cases/vcpupin.conf create mode 100644 repos/setVcpus/vcpupin_config.py create mode 100644 repos/setVcpus/vcpupin_live.py diff --git a/cases/vcpupin.conf b/cases/vcpupin.conf new file mode 100644 index 0000000..880247f --- /dev/null +++ b/cases/vcpupin.conf @@ -0,0 +1,67 @@ +domain:install_linux_cdrom + guestname + $defaultname + guestos + $defaultos + guestarch + $defaultarch + vcpu + 4 + memory + $defaultmem + hddriver + $defaulthd + nicdriver + $defaultnic + imageformat + qcow2 + +setVcpus:vcpupin_live + guestname + $defaultname + vcpu + 0 + cpulist + 2,4-6,^4 + +setVcpus:vcpupin_live + guestname + $defaultname + vcpu + 1 + cpulist + 3 + +domain:destroy + guestname + $defaultname + +setVcpus:vcpupin_config + guestname + $defaultname + vcpu + 2 + cpulist + 0-8,^1 + +setVcpus:vcpupin_config + guestname + $defaultname + vcpu + 3 + cpulist + ^2,0-8 + +domain:start + guestname + $defaultname + +domain:destroy + guestname + $defaultname + +domain:undefine + guestname + $defaultname + +options cleanup=enable diff --git a/repos/setVcpus/vcpupin_config.py b/repos/setVcpus/vcpupin_config.py new file mode 100644 index 0000000..80df659 --- /dev/null +++ b/repos/setVcpus/vcpupin_config.py @@ -0,0 +1,109 @@ +#!/usr/bin/env python +# Test domain vcpu pin with flag VIR_DOMAIN_AFFECT_CONFIG, check +# domain config xml with vcpupin configuration. + +import re +from xml.dom import minidom + +import libvirt +from libvirt import libvirtError + +from src import sharedmod +from utils import utils + +required_params = ('guestname', 'vcpu', 'cpulist',) +optional_params = {} + +def vcpupin_check(domobj, vcpu, cpumap): + """check domain config xml with vcpupin element + """ + guestxml = domobj.XMLDesc(2) + logger.debug("domain %s xml :\n%s" %(domobj.name(), guestxml)) + + doc = minidom.parseString(guestxml) + vcpupin = doc.getElementsByTagName('vcpupin') + if not vcpupin: + logger.error("no vcpupin element in domain xml") + return 1 + + for i in range(len(vcpupin)): + if vcpupin[i].hasAttribute('vcpu') and \ + vcpupin[i].hasAttribute('cpuset'): + vcpu_attr = vcpupin[i].getAttributeNode('vcpu') + cpu_attr = vcpupin[i].getAttributeNode('cpuset') + if int(vcpu_attr.nodeValue) == vcpu: + cpulist = cpu_attr.nodeValue + if cpulist == '': + cpumap_tmp = () + for i in range(maxcpu): + cpumap_tmp += (False,) + else: + cpumap_tmp = utils.param_to_tuple(cpulist, maxcpu) + + if cpumap_tmp == cpumap: + logger.info("cpuset is as expected in domain xml") + return 0 + else: + logger.error("cpuset is not as expected in domain xml") + return 1 + + if i == len(vcpupin) - 1: + logger.error("the vcpupin element with given vcpu is not found") + return 1 + +def vcpupin_config(params): + """pin domain vcpu to host cpu with config flag + """ + global logger + logger = params['logger'] + params.pop('logger') + guestname = params['guestname'] + vcpu = int(params['vcpu']) + cpulist = params['cpulist'] + + logger.info("the name of virtual machine is %s" % guestname) + logger.info("the given vcpu is %s" % vcpu) + logger.info("the given cpulist is %s" % cpulist) + + global maxcpu + maxcpu = utils.get_host_cpus() + logger.info("%s physical cpu on host" % maxcpu) + + conn = sharedmod.libvirtobj['conn'] + + try: + domobj = conn.lookupByName(guestname) + cpumap = utils.param_to_tuple(cpulist, maxcpu) + + if not cpumap: + logger.error("cpulist: Invalid format") + return 1 + + logger.debug("cpumap for vcpu pin is:") + logger.debug(cpumap) + + logger.info("pin domain vcpu %s to host cpulist %s with flag: %s" % + (vcpu, cpulist, libvirt.VIR_DOMAIN_AFFECT_CONFIG)) + domobj.pinVcpuFlags(vcpu, cpumap, libvirt.VIR_DOMAIN_AFFECT_CONFIG) + + logger.info("check vcpu pin info") + ret = domobj.vcpuPinInfo(libvirt.VIR_DOMAIN_AFFECT_CONFIG) + logger.debug("vcpu pin info is:") + logger.debug(ret) + if ret[vcpu] == cpumap: + logger.info("vcpu pin info is expected") + else: + logger.error("vcpu pin info is not expected") + return 1 + except libvirtError, e: + logger.error("libvirt call failed: " + str(e)) + return 1 + + logger.info("check domain vcpupin configuration in xml") + ret = vcpupin_check(domobj, vcpu, cpumap) + if ret: + logger.error("domain vcpu pin check failed") + return 1 + else: + logger.info("domain vcpu pin check succeed") + return 0 diff --git a/repos/setVcpus/vcpupin_live.py b/repos/setVcpus/vcpupin_live.py new file mode 100644 index 0000000..c3dfe8e --- /dev/null +++ b/repos/setVcpus/vcpupin_live.py @@ -0,0 +1,101 @@ +#!/usr/bin/env python +# Test domain vcpu pin with flag VIR_DOMAIN_AFFECT_LIVE, check +# vcpu subprocess status under domain task list on host. + +import re + +import libvirt +from libvirt import libvirtError + +from src import sharedmod +from utils import utils + +required_params = ('guestname', 'vcpu', 'cpulist',) +optional_params = {} + +def vcpupin_check(guestname, vcpu, cpumap): + """check vcpu subprocess status of the running virtual machine + grep Cpus_allowed_list /proc/PID/task/*/status + """ + tmp_str = '' + cmd = "cat /var/run/libvirt/qemu/%s.pid" % guestname + status, pid = utils.exec_cmd(cmd, shell=True) + if status: + logger.error("failed to get the pid of domain %s" % guestname) + return 1 + + cmd = "grep Cpus_allowed_list /proc/%s/task/*/status" % pid[0] + status, output = utils.exec_cmd(cmd, shell=True) + logger.debug("command '%s' output is:" % cmd) + for i in range(len(output)): + tmp_str += ''.join(output[i]) + '\n' + logger.debug(tmp_str) + + task_list = output[1:] + vcpu_task = task_list[int(vcpu)] + cpulist = vcpu_task.split('\t')[1] + ret = utils.param_to_tuple(cpulist, maxcpu) + + if ret == cpumap: + logger.info("vcpu process cpus allowed list is expected") + return 0 + else: + logger.error("vcpu process cpus allowed list is not expected") + return 1 + +def vcpupin_live(params): + """pin domain vcpu to host cpu with live flag + """ + global logger + logger = params['logger'] + params.pop('logger') + guestname = params['guestname'] + vcpu = int(params['vcpu']) + cpulist = params['cpulist'] + + logger.info("the name of virtual machine is %s" % guestname) + logger.info("the given vcpu is %s" % vcpu) + logger.info("the given cpulist is %s" % cpulist) + + global maxcpu + maxcpu = utils.get_host_cpus() + logger.info("%s physical cpu on host" % maxcpu) + + conn = sharedmod.libvirtobj['conn'] + + try: + domobj = conn.lookupByName(guestname) + cpumap = utils.param_to_tuple(cpulist, maxcpu) + if not cpumap: + logger.error("cpulist: Invalid format") + return 1 + + logger.debug("cpumap for vcpu pin is:") + logger.debug(cpumap) + + logger.info("pin domain vcpu %s to host cpu %s with flag: %s" % + (vcpu, cpulist, libvirt.VIR_DOMAIN_AFFECT_LIVE)) + domobj.pinVcpuFlags(vcpu, cpumap, libvirt.VIR_DOMAIN_AFFECT_LIVE) + + logger.info("check vcpus info") + ret = domobj.vcpus() + logger.debug("vcpus info is:") + logger.debug(ret) + if ret[1][vcpu] == cpumap: + logger.info("vcpus info is expected") + else: + logger.error("vcpus info is not expected") + return 1 + + except libvirtError, e: + logger.error("libvirt call failed: " + str(e)) + return 1 + + logger.info("check vcpu pin status on host") + ret = vcpupin_check(guestname, vcpu, cpumap) + if ret: + logger.error("domain vcpu pin failed") + return 1 + else: + logger.info("domain vcpu pin succeed") + return 0 diff --git a/utils/utils.py b/utils/utils.py index e242847..36171cf 100644 --- a/utils/utils.py +++ b/utils/utils.py @@ -645,6 +645,77 @@ def run_mount_app(hostname, username, password, print "mount fail" return 1 +def format_parammap(paramlist, map_test, length): + """paramlist contains numbers which can be divided by '-', '^' and + ',', map_test is a tuple for getting it's content (True or False) + and form the new tuple base on numbers in paramlist, length is + the length of the return tuple + """ + parammap = () + + try: + if re.match('\^', paramlist): + unuse = int(re.split('\^', paramlist)[1]) + for i in range(length): + if i == unuse: + parammap += (False,) + else: + parammap += (map_test[i],) + + elif '-' in paramlist: + param = re.split('-', paramlist) + if not len(param) == 2: + return False + if not int(param[1]) < length: + print "paramlist: out of max range" + return False + if int(param[1]) < int(param[0]): + return False + + for i in range(length): + if i in range(int(param[0]), int(param[1])+1): + parammap += (True,) + else: + parammap += (map_test[i],) + + else: + for i in range(length): + if i == int(paramlist): + parammap += (True,) + else: + parammap += (map_test[i],) + + return parammap + except ValueError, e: + print "ValueError: " + str(e) + return False + +def param_to_tuple(paramlist, length): + """paramlist contains numbers which can be divided by '-', '^' and + ',', length is the length of the return tuple, return tuple only + have True or False value + """ + map_test = () + for i in range(length): + map_test += (False,) + + if ',' in paramlist: + param = re.split(',', paramlist) + for i in range(len(param)): + parammap = format_parammap(param[i], map_test, length) + if parammap: + map_test = parammap + else: + return False + return parammap + + else: + parammap = format_parammap(paramlist, map_test, length) + if parammap: + return parammap + else: + return False + def run_wget_app(hostname, username, password, file_url, logger): """Simple test for wget app on specified host""" cmd_line = "wget -P /tmp %s -o /tmp/wget.log" % (file_url) -- 1.7.1

2 1

[libvirt] [PATCH] qemu: don't fail update netdev on bridge detach failure
by Laine Stump 14 Dec '12

14 Dec '12

When a network device's bridge connection is changed by virDomainUpdateDevice, libvirt first removes the netdev's tap from its old bridge, then adds it to the new bridge. Sometimes, due to a network being destroyed while a guest device is still attached, the tap may already be "removed" from the old bridge (or the old bridge may not even exist any more); the existing code was needlessly failing the update when this happened, making it impossible to recover from the situation without completely detaching (i.e. removing) the netdev from the guest and re-attaching. Instead of failing the entire operation when removal of the tap from the old bridge fails, this patch changes qemuDomainChangeNetBridge to just log a warning and continue, allowing a reasonable recover from the situation. (you'll appreciate this change if you ever accidentally destroy a network while your guests are still using it). --- src/qemu/qemu_hotplug.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/src/qemu/qemu_hotplug.c b/src/qemu/qemu_hotplug.c index 3dfdb65..e9a5e35 100644 --- a/src/qemu/qemu_hotplug.c +++ b/src/qemu/qemu_hotplug.c @@ -1306,8 +1306,14 @@ qemuDomainChangeNetBridge(virConnectPtr conn, if (oldbridge) { ret = virNetDevBridgeRemovePort(oldbridge, olddev->ifname); virDomainAuditNet(vm, olddev, NULL, "detach", ret == 0); - if (ret < 0) - goto cleanup; + if (ret < 0) { + /* warn but continue - possibly the old network + * had been destroyed and reconstructed, leaving the + * tap device orphaned. + */ + VIR_WARN("Unable to detach device %s from bridge %s", + olddev->ifname, oldbridge); + } } ret = virNetDevBridgeAddPort(newbridge, olddev->ifname); -- 1.7.11.7

2 2

[libvirt] [PATCH] build: use fewer cat processes
by Eric Blake 14 Dec '12

14 Dec '12

* src/Makefile.am (libvirt.syms): Let cat loop for us. --- I'll wait for a review before pushing, rather than claiming it is trivial. src/Makefile.am | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/src/Makefile.am b/src/Makefile.am index 33b4ab0..cf6096a 100644 --- a/src/Makefile.am +++ b/src/Makefile.am @@ -1457,9 +1457,7 @@ libvirt.syms: libvirt_public.syms $(USED_SYM_FILES) \ printf '\n\n# Private symbols\n\n' >>$@-tmp && \ printf 'LIBVIRT_PRIVATE_$(VERSION) {\n\n' >>$@-tmp && \ printf 'global:\n\n' >>$@-tmp && \ - for file in $(USED_SYM_FILES); do \ - cat $$file >>$@-tmp; \ - done && \ + cat $(USED_SYM_FILES) >>$@-tmp && \ printf '\n\nlocal:\n*;\n\n};' >>$@-tmp && \ chmod a-w $@-tmp && \ mv $@-tmp libvirt.syms -- 1.8.0.2

2 2

[libvirt] [PATCH] build: Distribute more files
by Jiri Denemark 14 Dec '12

14 Dec '12

--- src/Makefile.am | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) diff --git a/src/Makefile.am b/src/Makefile.am index 2791cb6..33b4ab0 100644 --- a/src/Makefile.am +++ b/src/Makefile.am @@ -367,6 +367,7 @@ check-symfile: endif check-symsorting: $(AM_V_GEN)$(PERL) $(srcdir)/check-symsorting.pl $(USED_SYM_FILES) +EXTRA_DIST += check-symfile.pl check-symsorting.pl PROTOCOL_STRUCTS = \ $(srcdir)/remote_protocol-structs \ @@ -389,7 +390,7 @@ else !WITH_REMOTE # re-generated when configured --without-remote. check-protocol: endif -EXTRA_DIST += $(PROTOCOL_STRUCTS) check-symfile.pl +EXTRA_DIST += $(PROTOCOL_STRUCTS) check-local: check-protocol check-symfile check-symsorting .PHONY: check-protocol $(PROTOCOL_STRUCTS:structs=struct) @@ -1575,6 +1576,9 @@ lockd_la_SOURCES = \ lockd_la_CFLAGS = -I$(top_srcdir)/src/conf $(AM_CFLAGS) lockd_la_LDFLAGS = -module -avoid-version lockd_la_LIBADD = ../gnulib/lib/libgnu.la libvirt-net-rpc.la libvirt-net-rpc-client.la +augeas_DATA += locking/libvirt_lockd.aug +augeastest_DATA += test_libvirt_lockd.aug +CLEANFILES += test_libvirt_lockd.aug if WITH_DTRACE_PROBES lockd_la_LIBADD += libvirt_probes.lo endif @@ -1620,7 +1624,10 @@ EXTRA_DIST += $(LOCK_DAEMON_SOURCES) \ $(LOCK_DRIVER_LOCKD_SOURCES) endif -EXTRA_DIST += locking/virtlockd.sysconf +EXTRA_DIST += locking/virtlockd.sysconf \ + locking/lockd.conf \ + locking/libvirt_lockd.aug \ + locking/test_libvirt_lockd.aug.in install-sysconfig: mkdir -p $(DESTDIR)$(sysconfdir)/sysconfig -- 1.8.0.2

2 3

[libvirt] [PATCH] selinux: fix NULL dereference in GetSecurityMountOptions
by Ján Tomko 14 Dec '12

14 Dec '12

In the case of an OOM error in virDomainDefGetSecurityLabelDef, secdef is set to NULL, then dereferenced while printing the debug message. --- src/security/security_selinux.c | 3 ++- 1 files changed, 2 insertions(+), 1 deletions(-) diff --git a/src/security/security_selinux.c b/src/security/security_selinux.c index 5409e32..9134bc8 100644 --- a/src/security/security_selinux.c +++ b/src/security/security_selinux.c @@ -1993,7 +1993,8 @@ virSecuritySELinuxGetSecurityMountOptions(virSecurityManagerPtr mgr, return NULL; } - VIR_DEBUG("imageLabel=%s opts=%s", secdef->imagelabel, opts); + VIR_DEBUG("imageLabel=%s opts=%s", + secdef ? secdef->imagelabel : "(null)", opts); return opts; } -- 1.7.8.6

3 2

[libvirt] [PATCH] build: Fix VPATH build
by Jiri Denemark 14 Dec '12

14 Dec '12

$(srcdir) is already part of $$file since commit f1f9a7ac7e. --- Pushed as a build-breaker. src/Makefile.am | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/Makefile.am b/src/Makefile.am index bb80992..7c2d8ed 100644 --- a/src/Makefile.am +++ b/src/Makefile.am @@ -1451,7 +1451,7 @@ libvirt.syms: libvirt_public.syms $(USED_SYM_FILES) \ printf 'LIBVIRT_PRIVATE_$(VERSION) {\n\n' >>$@-tmp && \ printf 'global:\n\n' >>$@-tmp && \ for file in $(USED_SYM_FILES); do \ - cat $(srcdir)/$$file >>$@-tmp; \ + cat $$file >>$@-tmp; \ done && \ printf '\n\nlocal:\n*;\n\n};' >>$@-tmp && \ chmod a-w $@-tmp && \ -- 1.8.0.2

2 1

[libvirt] [PATCH] network: prevent dnsmasq from listening on localhost
by Laine Stump 14 Dec '12

14 Dec '12

This patch resolves the problem reported in: https://bugzilla.redhat.com/show_bug.cgi?id=886663 The source of the problem was the fix for CVE 2011-3411: https://bugzilla.redhat.com/show_bug.cgi?id=833033 which was originally committed upstream in commit 753ff83a50263d6975f88d6605d4b5ddfcc97560. That commit improperly removed the "--except-interface lo" from dnsmasq commandlines when --bind-dynamic was used (based on comments in the latter bug). It turns out that the problem reported in the CVE could be eliminated without removing "--except-interface lo", and removing it actually caused each instance of dnsmasq to listen on localhost on port 53, which created a new problem: If another instance of dnsmasq using "bind-interfaces" (instead of "bind-dynamic") had already been started (or if another instance started later used "bind-dynamic"), this wouldn't have any immediately visible ill effects, but if you tried to start another dnsmasq instance using "bind-interfaces" *after* starting any libvirt networks, the new dnsmasq would fail to start, because there was already another process listening on port 53. (Subsequent to the CVE fix, another patch changed the network driver to put dnsmasq options in a conf file rather than directly on the dnsmasq commandline, but preserved the same options.) This patch changes the network driver to *always* add "except-interface=lo" to dnsmasq conf files, regardless of whether we use bind-dynamic or bind-interfaces. This way no libvirt dnsmasq instances are listening on localhost (and the CVE is still fixed). The actual code change is miniscule, but must be propogated through all of the test files as well. --- src/network/bridge_driver.c | 7 ++++--- tests/networkxml2confdata/dhcp6-nat-network.conf | 1 + tests/networkxml2confdata/dhcp6-network.conf | 1 + tests/networkxml2confdata/dhcp6host-routed-network.conf | 1 + tests/networkxml2confdata/isolated-network.conf | 2 +- tests/networkxml2confdata/nat-network-dns-hosts.conf | 1 + tests/networkxml2confdata/nat-network-dns-srv-record-minimal.conf | 2 +- tests/networkxml2confdata/nat-network-dns-srv-record.conf | 1 + tests/networkxml2confdata/nat-network-dns-txt-record.conf | 1 + tests/networkxml2confdata/nat-network.conf | 1 + tests/networkxml2confdata/netboot-network.conf | 2 +- tests/networkxml2confdata/netboot-proxy-network.conf | 2 +- tests/networkxml2confdata/routed-network.conf | 1 + 13 files changed, 16 insertions(+), 7 deletions(-) diff --git a/src/network/bridge_driver.c b/src/network/bridge_driver.c index 4e1958d..a32755d 100644 --- a/src/network/bridge_driver.c +++ b/src/network/bridge_driver.c @@ -689,6 +689,9 @@ networkDnsmasqConfContents(virNetworkObjPtr network, if (pidfile) virBufferAsprintf(&configbuf, "pid-file=%s\n", pidfile); + /* dnsmasq will *always* listen on localhost unless told otherwise */ + virBufferAddLit(&configbuf, "except-interface=lo\n"); + if (dnsmasqCapsGet(caps, DNSMASQ_CAPS_BIND_DYNAMIC)) { /* using --bind-dynamic with only --interface (no * --listen-address) prevents dnsmasq from responding to dns @@ -702,9 +705,7 @@ networkDnsmasqConfContents(virNetworkObjPtr network, "interface=%s\n", network->def->bridge); } else { - virBufferAddLit(&configbuf, - "bind-interfaces\n" - "except-interface=lo\n"); + virBufferAddLit(&configbuf, "bind-interfaces\n"); /* * --interface does not actually work with dnsmasq < 2.47, * due to DAD for ipv6 addresses on the interface. diff --git a/tests/networkxml2confdata/dhcp6-nat-network.conf b/tests/networkxml2confdata/dhcp6-nat-network.conf index d488900..050f3db 100644 --- a/tests/networkxml2confdata/dhcp6-nat-network.conf +++ b/tests/networkxml2confdata/dhcp6-nat-network.conf @@ -7,6 +7,7 @@ strict-order domain-needed local=// +except-interface=lo bind-dynamic interface=virbr0 dhcp-range=192.168.122.2,192.168.122.254 diff --git a/tests/networkxml2confdata/dhcp6-network.conf b/tests/networkxml2confdata/dhcp6-network.conf index 5c1030c..5fde07f 100644 --- a/tests/networkxml2confdata/dhcp6-network.conf +++ b/tests/networkxml2confdata/dhcp6-network.conf @@ -9,6 +9,7 @@ domain-needed domain=mynet expand-hosts local=/mynet/ +except-interface=lo bind-dynamic interface=virbr0 dhcp-range=2001:db8:ac10:fd01::1:10,2001:db8:ac10:fd01::1:ff diff --git a/tests/networkxml2confdata/dhcp6host-routed-network.conf b/tests/networkxml2confdata/dhcp6host-routed-network.conf index cb4d0cc..f8f05c2 100644 --- a/tests/networkxml2confdata/dhcp6host-routed-network.conf +++ b/tests/networkxml2confdata/dhcp6host-routed-network.conf @@ -7,6 +7,7 @@ strict-order domain-needed local=// +except-interface=lo bind-dynamic interface=virbr1 dhcp-range=192.168.122.1,static diff --git a/tests/networkxml2confdata/isolated-network.conf b/tests/networkxml2confdata/isolated-network.conf index 55a44d3..f8997bd 100644 --- a/tests/networkxml2confdata/isolated-network.conf +++ b/tests/networkxml2confdata/isolated-network.conf @@ -7,8 +7,8 @@ strict-order domain-needed local=// -bind-interfaces except-interface=lo +bind-interfaces listen-address=192.168.152.1 dhcp-option=3 no-resolv diff --git a/tests/networkxml2confdata/nat-network-dns-hosts.conf b/tests/networkxml2confdata/nat-network-dns-hosts.conf index ae8f8c5..2577882 100644 --- a/tests/networkxml2confdata/nat-network-dns-hosts.conf +++ b/tests/networkxml2confdata/nat-network-dns-hosts.conf @@ -9,6 +9,7 @@ domain-needed domain=example.com expand-hosts local=/example.com/ +except-interface=lo bind-dynamic interface=virbr0 addn-hosts=/var/lib/libvirt/dnsmasq/default.addnhosts diff --git a/tests/networkxml2confdata/nat-network-dns-srv-record-minimal.conf b/tests/networkxml2confdata/nat-network-dns-srv-record-minimal.conf index faa36e6..1e9b59c 100644 --- a/tests/networkxml2confdata/nat-network-dns-srv-record-minimal.conf +++ b/tests/networkxml2confdata/nat-network-dns-srv-record-minimal.conf @@ -7,8 +7,8 @@ strict-order domain-needed local=// -bind-interfaces except-interface=lo +bind-interfaces listen-address=192.168.122.1 listen-address=192.168.123.1 listen-address=fc00:db8:ac10:fe01::1 diff --git a/tests/networkxml2confdata/nat-network-dns-srv-record.conf b/tests/networkxml2confdata/nat-network-dns-srv-record.conf index 6079912..53d044a 100644 --- a/tests/networkxml2confdata/nat-network-dns-srv-record.conf +++ b/tests/networkxml2confdata/nat-network-dns-srv-record.conf @@ -7,6 +7,7 @@ strict-order domain-needed local=// +except-interface=lo bind-dynamic interface=virbr0 srv-host=name.tcp.test-domain-name,.,1024,10,10 diff --git a/tests/networkxml2confdata/nat-network-dns-txt-record.conf b/tests/networkxml2confdata/nat-network-dns-txt-record.conf index c448bdc..921cae1 100644 --- a/tests/networkxml2confdata/nat-network-dns-txt-record.conf +++ b/tests/networkxml2confdata/nat-network-dns-txt-record.conf @@ -7,6 +7,7 @@ strict-order domain-needed local=// +except-interface=lo bind-dynamic interface=virbr0 txt-record=example,example value diff --git a/tests/networkxml2confdata/nat-network.conf b/tests/networkxml2confdata/nat-network.conf index 8f28fef..beb714b 100644 --- a/tests/networkxml2confdata/nat-network.conf +++ b/tests/networkxml2confdata/nat-network.conf @@ -7,6 +7,7 @@ strict-order domain-needed local=// +except-interface=lo bind-dynamic interface=virbr0 dhcp-range=192.168.122.2,192.168.122.254 diff --git a/tests/networkxml2confdata/netboot-network.conf b/tests/networkxml2confdata/netboot-network.conf index 83dd2b3..b6f3c23 100644 --- a/tests/networkxml2confdata/netboot-network.conf +++ b/tests/networkxml2confdata/netboot-network.conf @@ -9,8 +9,8 @@ domain-needed domain=example.com expand-hosts local=/example.com/ -bind-interfaces except-interface=lo +bind-interfaces listen-address=192.168.122.1 dhcp-range=192.168.122.2,192.168.122.254 dhcp-no-override diff --git a/tests/networkxml2confdata/netboot-proxy-network.conf b/tests/networkxml2confdata/netboot-proxy-network.conf index b266d81..1e969fa 100644 --- a/tests/networkxml2confdata/netboot-proxy-network.conf +++ b/tests/networkxml2confdata/netboot-proxy-network.conf @@ -9,8 +9,8 @@ domain-needed domain=example.com expand-hosts local=/example.com/ -bind-interfaces except-interface=lo +bind-interfaces listen-address=192.168.122.1 dhcp-range=192.168.122.2,192.168.122.254 dhcp-no-override diff --git a/tests/networkxml2confdata/routed-network.conf b/tests/networkxml2confdata/routed-network.conf index dc53a4e..62ffd7a 100644 --- a/tests/networkxml2confdata/routed-network.conf +++ b/tests/networkxml2confdata/routed-network.conf @@ -7,6 +7,7 @@ strict-order domain-needed local=// +except-interface=lo bind-dynamic interface=virbr1 addn-hosts=/var/lib/libvirt/dnsmasq/local.addnhosts -- 1.7.11.7

3 5

[libvirt] [PATCH] locking: Fix VPATH build and distribute generated files
by Jiri Denemark 14 Dec '12

14 Dec '12

--- src/Makefile.am | 18 ++++++++++++------ 1 file changed, 12 insertions(+), 6 deletions(-) diff --git a/src/Makefile.am b/src/Makefile.am index 7c2d8ed..5a9fb84 100644 --- a/src/Makefile.am +++ b/src/Makefile.am @@ -149,19 +149,22 @@ LOCK_DRIVER_SANLOCK_HELPER_SOURCES = \ locking/sanlock_helper.c LOCK_PROTOCOL_GENERATED = \ - locking/lock_protocol.h \ - locking/lock_protocol.c \ + $(srcdir)/locking/lock_protocol.h \ + $(srcdir)/locking/lock_protocol.c \ $(NULL) -EXTRA_DIST += locking/lock_protocol.x +LOCK_PROTOCOL = $(srcdir)/locking/lock_protocol.x +EXTRA_DIST += $(LOCK_PROTOCOL) \ + $(LOCK_PROTOCOL_GENERATED) BUILT_SOURCES += $(LOCK_PROTOCOL_GENERATED) MAINTAINERCLEANFILES += $(LOCK_PROTOCOL_GENERATED) LOCK_DAEMON_GENERATED = \ - locking/lock_daemon_dispatch_stubs.h + $(srcdir)/locking/lock_daemon_dispatch_stubs.h $(NULL) BUILT_SOURCES += $(LOCK_DAEMON_GENERATED) +EXTRA_DIST += $(LOCK_DAEMON_GENERATED) MAINTAINERCLEANFILES += $(LOCK_DAEMON_GENERATED) LOCK_DRIVER_LOCKD_SOURCES = \ @@ -177,8 +180,11 @@ LOCK_DAEMON_SOURCES = \ locking/lock_daemon_dispatch.h \ $(NULL) -$(srcdir)/locking/lock_daemon_dispatch_stubs.h: locking/lock_protocol.x $(srcdir)/rpc/gendispatch.pl Makefile.am - $(AM_V_GEN)perl -w $(srcdir)/rpc/gendispatch.pl -b virLockSpaceProtocol VIR_LOCK_SPACE_PROTOCOL $< > $@ +$(srcdir)/locking/lock_daemon_dispatch_stubs.h: $(LOCK_PROTOCOL) \ + $(srcdir)/rpc/gendispatch.pl Makefile.am + $(AM_V_GEN)perl -w $(srcdir)/rpc/gendispatch.pl \ + -b virLockSpaceProtocol VIR_LOCK_SPACE_PROTOCOL \ + $(LOCK_PROTOCOL) > $@ NETDEV_CONF_SOURCES = \ -- 1.8.0.2

3 3

[libvirt] [PATCH] build: Distribute more files
by Jiri Denemark 14 Dec '12

14 Dec '12

--- src/Makefile.am | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-) diff --git a/src/Makefile.am b/src/Makefile.am index 2791cb6..33b4ab0 100644 --- a/src/Makefile.am +++ b/src/Makefile.am @@ -367,6 +367,7 @@ check-symfile: endif check-symsorting: $(AM_V_GEN)$(PERL) $(srcdir)/check-symsorting.pl $(USED_SYM_FILES) +EXTRA_DIST += check-symfile.pl check-symsorting.pl PROTOCOL_STRUCTS = \ $(srcdir)/remote_protocol-structs \ @@ -389,7 +390,7 @@ else !WITH_REMOTE # re-generated when configured --without-remote. check-protocol: endif -EXTRA_DIST += $(PROTOCOL_STRUCTS) check-symfile.pl +EXTRA_DIST += $(PROTOCOL_STRUCTS) check-local: check-protocol check-symfile check-symsorting .PHONY: check-protocol $(PROTOCOL_STRUCTS:structs=struct) @@ -1575,6 +1576,9 @@ lockd_la_SOURCES = \ lockd_la_CFLAGS = -I$(top_srcdir)/src/conf $(AM_CFLAGS) lockd_la_LDFLAGS = -module -avoid-version lockd_la_LIBADD = ../gnulib/lib/libgnu.la libvirt-net-rpc.la libvirt-net-rpc-client.la +augeas_DATA += locking/libvirt_lockd.aug +augeastest_DATA += test_libvirt_lockd.aug +CLEANFILES += test_libvirt_lockd.aug if WITH_DTRACE_PROBES lockd_la_LIBADD += libvirt_probes.lo endif @@ -1620,7 +1624,10 @@ EXTRA_DIST += $(LOCK_DAEMON_SOURCES) \ $(LOCK_DRIVER_LOCKD_SOURCES) endif -EXTRA_DIST += locking/virtlockd.sysconf +EXTRA_DIST += locking/virtlockd.sysconf \ + locking/lockd.conf \ + locking/libvirt_lockd.aug \ + locking/test_libvirt_lockd.aug.in install-sysconfig: mkdir -p $(DESTDIR)$(sysconfdir)/sysconfig -- 1.8.0.2

1 1

[libvirt] [PATCH v2] Fix error reporting when fetching SCSI/LVM keys
by Daniel P. Berrange 14 Dec '12

14 Dec '12

From: "Daniel P. Berrange" <berrange(a)redhat.com> The current virStorageFileGet{LVM,SCSI}Key methods return the key as the return value. Unfortunately it is desirable for "NULL" to be a valid return value, as well as an error indicator. Thus the returned key must instead be provided as an out-parameter. When we invoke lvs or scsi_id to extract ID for block devices, we don't want virCommandWait logging errors messages. Thus we must explicitly check 'status != 0', rather than letting virCommandWait do it. Signed-off-by: Daniel P. Berrange <berrange(a)redhat.com> --- src/util/storage_file.c | 74 ++++++++++++++++++++++++++++++++----------------- src/util/storage_file.h | 6 ++-- 2 files changed, 53 insertions(+), 27 deletions(-) diff --git a/src/util/storage_file.c b/src/util/storage_file.c index 17a47cf..fe7e8b8 100644 --- a/src/util/storage_file.c +++ b/src/util/storage_file.c @@ -1192,62 +1192,75 @@ int virStorageFileIsClusterFS(const char *path) } #ifdef LVS -char *virStorageFileGetLVMKey(const char *path) +int virStorageFileGetLVMKey(const char *path, + char **key) { /* * # lvs --noheadings --unbuffered --nosuffix --options "uuid" LVNAME * 06UgP5-2rhb-w3Bo-3mdR-WeoL-pytO-SAa2ky */ - char *key = NULL; + int status; virCommandPtr cmd = virCommandNewArgList( LVS, "--noheadings", "--unbuffered", "--nosuffix", "--options", "uuid", path, NULL ); + int ret = -1; + + *key = NULL; /* Run the program and capture its output */ - virCommandSetOutputBuffer(cmd, &key); - if (virCommandRun(cmd, NULL) < 0) + virCommandSetOutputBuffer(cmd, key); + if (virCommandRun(cmd, &status) < 0) goto cleanup; - if (key) { + /* Explicitly check status == 0, rather than passing NULL + * to virCommandRun because we don't want to raise an actual + * error in this scenario, just return a NULL key. + */ + + if (status == 0 && *key) { char *nl; - char *tmp = key; + char *tmp = *key; /* Find first non-space character */ while (*tmp && c_isspace(*tmp)) { tmp++; } /* Kill leading spaces */ - if (tmp != key) - memmove(key, tmp, strlen(tmp)+1); + if (tmp != *key) + memmove(*key, tmp, strlen(tmp)+1); /* Kill trailing newline */ - if ((nl = strchr(key, '\n'))) + if ((nl = strchr(*key, '\n'))) *nl = '\0'; } - if (key && STREQ(key, "")) - VIR_FREE(key); + ret = 0; cleanup: + if (*key && STREQ(*key, "")) + VIR_FREE(*key); + virCommandFree(cmd); - return key; + return ret; } #else -char *virStorageFileGetLVMKey(const char *path) +int virStorageFileGetLVMKey(const char *path, + char **key) { virReportSystemError(ENOSYS, _("Unable to get LVM key for %s"), path); - return NULL; + return -1; } #endif #ifdef HAVE_UDEV -char *virStorageFileGetSCSIKey(const char *path) +int virStorageFileGetSCSIKey(const char *path, + char **key) { - char *key = NULL; + int status; virCommandPtr cmd = virCommandNewArgList( "/lib/udev/scsi_id", "--replace-whitespace", @@ -1255,30 +1268,41 @@ char *virStorageFileGetSCSIKey(const char *path) "--device", path, NULL ); + int ret = -1; + + *key = NULL; /* Run the program and capture its output */ - virCommandSetOutputBuffer(cmd, &key); - if (virCommandRun(cmd, NULL) < 0) + virCommandSetOutputBuffer(cmd, key); + if (virCommandRun(cmd, &status) < 0) goto cleanup; - if (key && STRNEQ(key, "")) { - char *nl = strchr(key, '\n'); + /* Explicitly check status == 0, rather than passing NULL + * to virCommandRun because we don't want to raise an actual + * error in this scenario, just return a NULL key. + */ + if (status == 0 && *key) { + char *nl = strchr(*key, '\n'); if (nl) *nl = '\0'; - } else { - VIR_FREE(key); } + ret = 0; + cleanup: + if (*key && STREQ(*key, "")) + VIR_FREE(*key); + virCommandFree(cmd); - return key; + return ret; } #else -char *virStorageFileGetSCSIKey(const char *path) +int virStorageFileGetSCSIKey(const char *path, + char **key) { virReportSystemError(ENOSYS, _("Unable to get SCSI key for %s"), path); - return NULL; + return -1; } #endif diff --git a/src/util/storage_file.h b/src/util/storage_file.h index 9e4516e..6fbd275 100644 --- a/src/util/storage_file.h +++ b/src/util/storage_file.h @@ -101,7 +101,9 @@ int virStorageFileIsClusterFS(const char *path); int virStorageFileIsSharedFSType(const char *path, int fstypes); -char *virStorageFileGetLVMKey(const char *path); -char *virStorageFileGetSCSIKey(const char *path); +int virStorageFileGetLVMKey(const char *path, + char **key); +int virStorageFileGetSCSIKey(const char *path, + char **key); #endif /* __VIR_STORAGE_FILE_H__ */ -- 1.7.11.7

2 1

[libvirt] [PATCH] conf: Remove duplicate declaration of virNetworkDNSDefPtr
by Jiri Denemark 14 Dec '12

14 Dec '12

--- Pushed as trivial build-breaker. src/conf/network_conf.h | 2 -- 1 file changed, 2 deletions(-) diff --git a/src/conf/network_conf.h b/src/conf/network_conf.h index 49fc0ca..8adc2a4 100644 --- a/src/conf/network_conf.h +++ b/src/conf/network_conf.h @@ -106,24 +106,22 @@ struct _virNetworkDNSHostDef { typedef struct _virNetworkDNSDef virNetworkDNSDef; typedef virNetworkDNSDef *virNetworkDNSDefPtr; struct _virNetworkDNSDef { size_t ntxts; virNetworkDNSTxtDefPtr txts; size_t nhosts; virNetworkDNSHostDefPtr hosts; size_t nsrvs; virNetworkDNSSrvDefPtr srvs; }; -typedef struct _virNetworkDNSDef *virNetworkDNSDefPtr; - typedef struct _virNetworkIpDef virNetworkIpDef; typedef virNetworkIpDef *virNetworkIpDefPtr; struct _virNetworkIpDef { char *family; /* ipv4 or ipv6 - default is ipv4 */ virSocketAddr address; /* Bridge IP address */ /* One or the other of the following two will be used for a given * IP address, but never both. The parser guarantees this. * Use virNetworkIpDefPrefix/virNetworkIpDefNetmask rather * than accessing the data directly - these utility functions * will convert one into the other as necessary. -- 1.8.0

2 1

[libvirt] [PATCH 0/2] virsh: catch disconnections for non-default connections too
by Ján Tomko 14 Dec '12

14 Dec '12

Right now virsh says 'Failed to reconnect to the hypervisor' even for the first connection to the default URI and it fails to catch disconnections for connections to the non-default URI. Ján Tomko (2): virsh: don't lie about reconnection in vshReconnect virsh: use vshReconnect for non-default connections too tools/virsh.c | 11 +++++------ 1 files changed, 5 insertions(+), 6 deletions(-) -- 1.7.8.6

2 4

[libvirt] [PATCH] Support network boot for HVM guests in libxl
by Jim Fehlig 13 Dec '12

13 Dec '12

The libxl driver ignored boot devices in the domain config, preventing PXE booting HVM domains. This patch accounts for user-specified boot devices when building the libxl domain configuration. --- I'm inclined to call this a bug fix since it is not possible to network boot HVM guests without this patch. Would it be ok to push this for 1.0.1? src/libxl/libxl_conf.c | 30 ++++++++++++++++++++++++++++++ 1 file changed, 30 insertions(+) diff --git a/src/libxl/libxl_conf.c b/src/libxl/libxl_conf.c index 5b6d6fb..4c0e961 100644 --- a/src/libxl/libxl_conf.c +++ b/src/libxl/libxl_conf.c @@ -438,6 +438,8 @@ libxlMakeDomBuildInfo(virDomainDefPtr def, libxl_domain_config *d_config) b_info->max_memkb = def->mem.max_balloon; b_info->target_memkb = def->mem.cur_balloon; if (hvm) { + char bootorder[VIR_DOMAIN_BOOT_LAST+1]; + libxl_defbool_set(&b_info->u.hvm.pae, def->features & (1 << VIR_DOMAIN_FEATURE_PAE)); libxl_defbool_set(&b_info->u.hvm.apic, @@ -450,6 +452,34 @@ libxlMakeDomBuildInfo(virDomainDefPtr def, libxl_domain_config *d_config) libxl_defbool_set(&b_info->u.hvm.hpet, 1); } } + for (i = 0 ; i < def->os.nBootDevs ; i++) { + switch (def->os.bootDevs[i]) { + case VIR_DOMAIN_BOOT_FLOPPY: + bootorder[i] = 'a'; + break; + default: + case VIR_DOMAIN_BOOT_DISK: + bootorder[i] = 'c'; + break; + case VIR_DOMAIN_BOOT_CDROM: + bootorder[i] = 'd'; + break; + case VIR_DOMAIN_BOOT_NET: + bootorder[i] = 'n'; + break; + } + } + if (def->os.nBootDevs == 0) { + bootorder[0] = 'c'; + bootorder[1] = '\0'; + } + else { + bootorder[def->os.nBootDevs] = '\0'; + } + if ((b_info->u.hvm.boot = strdup(bootorder)) == NULL) { + virReportOOMError(); + goto error; + } /* * The following comment and calculation were taken directly from -- 1.8.0.1

2 2

[libvirt] [PATCH 00/14 v4] Implement virtlockd daemon
by Daniel P. Berrange 13 Dec '12

13 Dec '12

This is an update of https://www.redhat.com/archives/libvir-list/2012-September/msg00816.html This series was previously fully acked, but before pushing I noticed some problems and stopped. Since then I've done quite a few more changes to rebase to latest GIT & follow new best practice. I've also added support for a config file for the lockd plugin and the ability to use custom lockspaces for LVM and SCSI IDs. I've also added support to acquire leases based on a SHA256 sum of the file, instead of directly on the file As such I don't feel comfortable pushing, without further reviews. While we're past the freeze date, I'd like to request an exception, for all except the 14th patch. Without the 14th patch, this series has minimal impact on existing code that is used in libvirt so little regression possibility. Further I discovered yesterday that we already have people using this code as custom add-on patches, so I'd like to help them by including it in 1.0.1 .gitignore | 5 bootstrap.conf | 1 cfg.mk | 9 libvirt.spec.in | 16 po/POTFILES.in | 4 run.in | 2 src/Makefile.am | 214 ++++- src/internal.h | 22 src/libvirt_private.syms | 2 src/locking/libvirt_lockd.aug | 33 src/locking/lock_daemon.c | 1441 ++++++++++++++++++++++++++++++++++ src/locking/lock_daemon.h | 56 + src/locking/lock_daemon_config.c | 193 ++++ src/locking/lock_daemon_config.h | 50 + src/locking/lock_daemon_dispatch.c | 434 ++++++++++ src/locking/lock_daemon_dispatch.h | 31 src/locking/lock_driver_lockd.c | 834 +++++++++++++++++++ src/locking/lock_manager.c | 20 src/locking/lock_manager.h | 3 src/locking/lock_protocol.x | 95 ++ src/locking/lockd.conf | 68 + src/locking/test_libvirt_lockd.aug.in | 7 src/locking/virtlockd.init.in | 93 ++ src/locking/virtlockd.service.in | 13 src/locking/virtlockd.socket.in | 8 src/locking/virtlockd.sysconf | 3 src/qemu/qemu.conf | 17 src/qemu/qemu_conf.c | 12 src/qemu/qemu_conf.h | 3 src/qemu/qemu_driver.c | 24 src/qemu/test_libvirtd_qemu.aug.in | 2 src/util/storage_file.c | 30 src/util/storage_file.h | 4 33 files changed, 3697 insertions(+), 52 deletions(-)

4 41

[libvirt] Fix handling of QED file format
by Daniel P. Berrange 13 Dec '12

13 Dec '12

Libvirt mis-detected the QED file format as raw

2 8

[libvirt] [PATCH] Change virCgroupGetAppRoot stub on non-Linux to avoid unused param warning
by Daniel P. Berrange 13 Dec '12

13 Dec '12

From: "Daniel P. Berrange" <berrange(a)redhat.com> Fully stub out the virCgroupGetAppRoot method as done with other methods in the file, rather than just the body. This lets us annotate the unused parameter to avoid a warning Pushed as a Win32 build breaker fix --- src/util/cgroup.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/src/util/cgroup.c b/src/util/cgroup.c index f867fb7..8f3b8b7 100644 --- a/src/util/cgroup.c +++ b/src/util/cgroup.c @@ -974,14 +974,17 @@ int virCgroupForDriver(const char *name ATTRIBUTE_UNUSED, * * Returns 0 on success */ +#if defined HAVE_MNTENT_H && defined HAVE_GETMNTENT_R int virCgroupGetAppRoot(virCgroupPtr *group) { -#if defined HAVE_MNTENT_H && defined HAVE_GETMNTENT_R return virCgroupNew("/", group); +} #else +int virCgroupGetAppRoot(virCgroupPtr *group ATTRIBUTE_UNUSED) +{ return -ENXIO; -#endif } +#endif /** * virCgroupForDomain: -- 1.7.11.7

1 0

[libvirt] [PATCH 00/12 v6] Unprivleged SG_IO support
by Osier Yang 13 Dec '12

13 Dec '12

As a result of RFC [1], this implements the unprivleged SG_IO support. v4 - v5: * Per discussion in [2], this set uses a hash table to store the shared disk's info. See 1/12 for more details. Osier Yang (12): qemu: Add a hash table for the shared disks docs: Add docs and rng schema for new XML cdbfilter conf: Parse and format the new XML tag cdbfilter util: Prepare helpers for unpriv_sgio setting qemu: Manage disk's unpriv_sgio in domain lifecyle qemu: Do not restore the sysfs unpriv_sgio if the disk is being shared qemu: Error out when domain starting if the cdbfilter setting conflicts qemu: Set unpriv_sgio when attaching disk qemu: Restore unpriv_sgio when detaching disk qemu: Error out if the shared disk conf conflicts with others when attaching qemu: Do not restore unpriv_sgio when detaching disk if it's still shared conf: Save disk's original unpriv_sgio state into status XML docs/formatdomain.html.in | 13 ++- docs/schemas/domaincommon.rng | 52 +++++-- src/conf/domain_conf.c | 93 ++++++++++--- src/conf/domain_conf.h | 11 ++ src/libvirt_private.syms | 4 + src/qemu/qemu_conf.c | 60 ++++++++ src/qemu/qemu_conf.h | 22 +++ src/qemu/qemu_driver.c | 72 ++++++++++ src/qemu/qemu_process.c | 104 ++++++++++++++- src/qemu/qemu_process.h | 3 + src/util/util.c | 145 ++++++++++++++++++++ src/util/util.h | 8 + ...ml2argv-disk-scsi-lun-passthrough-cdbfilter.xml | 32 +++++ tests/qemuxml2xmltest.c | 1 + 14 files changed, 585 insertions(+), 35 deletions(-) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-disk-scsi-lun-passthrough-cdbfilter.xml [1] https://www.redhat.com/archives/libvir-list/2012-November/msg00988.html [2] https://www.redhat.com/archives/libvir-list/2012-December/msg00325.html Regards, Osier

3 26

[libvirt] [PATCH v1 rebase 0/6] disable cgroup cpuset
by Hu Tao 13 Dec '12

13 Dec '12

This is merely a rebased version, so still tagged with v1. The reason to disable cgroup cpuset by default is: currently, sub-directories of cgroup cpuset won't be auto-updated if cpu hot-plug happens. For example, if we now have 0-3,5-7 pcpus online, the value of /sys/fs/cgroup/cpuset/libvirt/qemu/mydom/cpuset.cpus is 0-3,5-7, then pcpu 4 is hot-plugged, but the value of cpuset.cpus is still 0-3,5-7. This will cause a problem that we can't pin to a hot-plugged pcpu. Users can still enable cpuset by editing qemu.conf. Hu Tao (6): cgroup: Add a flag VIR_CGROUP_DISABLE_CPUSET create cgroup controllers for driver according to config create cgroup controllers for domain according to config create cgroup controllers for vcpu according to config create cgroup controllers for emulator according to config disable cgroup cpuset by default src/lxc/lxc_cgroup.c | 4 +- src/lxc/lxc_driver.c | 22 +++++------ src/lxc/lxc_process.c | 4 +- src/qemu/qemu.conf | 2 +- src/qemu/qemu_cgroup.c | 15 ++++---- src/qemu/qemu_conf.c | 1 - src/qemu/qemu_driver.c | 93 ++++++++++++++++++++++++++++++----------------- src/qemu/qemu_hotplug.c | 9 +++-- src/qemu/qemu_migration.c | 2 +- src/util/cgroup.c | 49 ++++++++++++++++++++----- src/util/cgroup.h | 12 ++++-- 11 files changed, 137 insertions(+), 76 deletions(-) -- 1.8.0.1.240.ge8a1f5a

4 10

[libvirt] [PATCH v1 0/6] disable cgroup cpuset
by Hu Tao 13 Dec '12

13 Dec '12

The reason to disable cgroup cpuset by default is: currently, sub-directories of cgroup cpuset won't be auto-updated if cpu hot-plug happens. For example, if we now have 0-3,5-7 pcpus online, the value of /sys/fs/cgroup/cpuset/libvirt/qemu/mydom/cpuset.cpus is 0-3,5-7, then pcpu 4 is hot-plugged, but the value of cpuset.cpus is still 0-3,5-7. This will cause a problem that we can't pin to a hot-plugged pcpu. Users can still enable cpuset by editing qemu.conf. Hu Tao (6): cgroup: Add a flag VIR_CGROUP_DISABLE_CPUSET create cgroup controllers for driver according to config create cgroup controllers for domain according to config create cgroup controllers for vcpu according to config create cgroup controllers for emulator according to config disable cgroup cpuset by default src/lxc/lxc_cgroup.c | 4 +-- src/lxc/lxc_driver.c | 22 ++++++------ src/lxc/lxc_process.c | 4 +-- src/qemu/qemu.conf | 2 +- src/qemu/qemu_cgroup.c | 15 ++++---- src/qemu/qemu_conf.c | 1 - src/qemu/qemu_driver.c | 92 +++++++++++++++++++++++++++++------------------ src/qemu/qemu_hotplug.c | 9 +++-- src/qemu/qemu_migration.c | 2 +- src/util/cgroup.c | 49 +++++++++++++++++++------ src/util/cgroup.h | 12 ++++--- 11 files changed, 136 insertions(+), 76 deletions(-) -- 1.7.11.7

2 11

[libvirt] [PATCH] network: match xml warning message
by Eric Blake 13 Dec '12

13 Dec '12

I noticed that /var/lib/libvirt/dnsmasq/*.conf used the wrong word; it was intended to match the wording in src/util/xml.c. * src/network/bridge_driver.c (networkDnsmasqConfContents): Fix typo. --- Pushing under the trivial rule. src/network/bridge_driver.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/network/bridge_driver.c b/src/network/bridge_driver.c index 3dd98ad..4e1958d 100644 --- a/src/network/bridge_driver.c +++ b/src/network/bridge_driver.c @@ -669,7 +669,7 @@ networkDnsmasqConfContents(virNetworkObjPtr network, "##OVERWRITTEN AND LOST. Changes to this " "configuration should be made using:\n" "## virsh net-edit %s\n" - "## of other applications using the libvirt API.\n" + "## or other application using the libvirt API.\n" "##\n## dnsmasq conf file created by libvirt\n" "strict-order\n" "domain-needed\n", -- 1.8.0.2

1 1

[libvirt] RFC: Death to the BQDL (Big QEMU Driver Lock)
by Daniel P. Berrange 13 Dec '12

13 Dec '12

Many years ago when the QEMU driver was first written for libvirt the daemon was single threaded, so we didn't have to worry about locking at all. Then we introduced threads and so we had to have locking. Since then locking has been done at two levels. The big QEMU driver lock must be acquired first, then the per-VM lock could be acquired. If no field in the driver struct was used, we could quickly release the QEMU driver lock and only hold the per-VM lock. Over time though, our code has got more & more complicated to the point where we have to hold the big QEMU driver lock in the vast majority of methods. We mitigate this by releasing the locks when sleeping on monitor API calls, but this is still a huge bottleneck. This is particularly apparent when you look at concurrency when starting/stopping guests. This scalability limitation is at a point where it is unacceptable for us to continue as we do today. Band-aids no longer suffice. To fix this for the long term, we need to dramatically change the locking approach we use in the QEMU driver. The only way we can achieve this is by dramatically changing the way we update/access QEMU driver state. The core problem is that the qemu driver struct holds a vast array of (often unrelated) data, with differing access patterns. It is clear that a single lock is not a suitable level of granularity for this. Looking at what is in the struct we can classify data into a number of buckets 1. Read-only data that never changes for lifetime of libvirtd. eg - char *configDir - char *libDir - bool privileged; - const char *uri; - virThreadPoolPtr workerPool 2. Read-only data that never changes for as long as the config file is not reloaded. (Currently equivalent to previous bucket since we don't support reloading qemu.conf - we need to support that in the future) eg - uid_t user; - uid_t group; - int dynamicOwnership; - unsigned int vncTLS :1; - unsigned int vncSASL :1; - char *vncListen; 3. Read-write data that changes at arbitrary times eg - virDomainObjList domains - size_t nactive - pciDeviceList *activePciHostdevs; - usbDeviceList *activeUsbHostdevs; - virHashTablePtr closeCallbacks; My proposal for dealing with things is as follows 1. Read-only data that never changes for lifetime of libvirtd. Turn the current driver mutex into a driver RW-lock. All API calls will always acquire a read-lock at start, hold it for the lifetime of their execution and release it at the end. API calls will never directly acquire write locks. The QEMU driver startup / shutdown global initializers will acquire write-locks. This ensures the daemon can't shutdown while any APIs are being executed. 2. Read-only data that never changes for as long as the config file is not reloaded. Move all of this data out into a new virQEMUDriverConfigPtr struct, which is an instance of virObject. The virQEMUDriverPtr will hold the primary reference to the config. The contents of this object struct will be considered immutable once initialized. When an API needs to access config file, it will obtain a reference on the config object. Obtaining the reference will involve acquiring & releasing the driver lock. If the QEMU driver needs to reload the config, it will populate a completely new virQEMUDriverConfigPtr instance, and unref the existing one. Thus access to data in virQEMUDriverConfigPtr can be completely lockless once an instance has been acquired, despite the possiblity of the config being updated at an arbitrary time. cf RCU (read-copy-update) 3. Read-write data that changes at arbitrary times. All data that can be changed must be stored in a dedicated virObject based instance. Each object must provide its own internal locking mechanisms targetted to the type of data being stored. Some objects may need some re-architecting to allow them to operate effectively without the protection of the long lived QEMU driver lock. For example during domain startup, we rely on the QEMU driver lock to protect against races between the time we check for an existing VM with (name,uuid), and the time we actually finish starting the new VM & store it in the domain list. To deal with this the virDomainObjList will need to have some concept of a 'reserved name,uuid' so safety is ensured, despite not holding a lock for the whole start operation. So I lied slightly when I said this was the death of the big QEMU driver lock. The big QEMU driver lock still exists, but API calls only ever need to have read-locks. Write-locks are only held for libvirtd startup/shutdown, and for the tiny time window it takes to grab a reference to a virQEMUDriverConfigPtr. Access to config params is completely lockless, even allowing for their live update. All the remaining exclusive locks will be pushed down into individual objects which need them, hopefully ensuring high concurrency of operation. Implementing this all is a non-trivial job, so I envisage the following order of attack 1. Create the virQEMUDriverConfigPtr object & move config file parameters into that. 2. Encapsulte all read-writable state into objects with dedicated locking 3. Turn QEMU driver mutex into a read-write lock 4. Convert all APIs to only hold read-locks on QEMU driver. Regards, Daniel -- |: http://berrange.com -o- http://www.flickr.com/photos/dberrange/ :| |: http://libvirt.org -o- http://virt-manager.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :|

2 2