- Devel - libvirt List Archives

[libvirt] [PATCH] Fix bridge configuration when OUTPUT policy is DROP on the host
by Lénaïc Huard 07 Jan '14

07 Jan '14

When the host is configured with very restrictive firewall (default policy is DROP for all chains, including OUTPUT), the bridge driver for Linux adds netfilter entries to allow DHCP and DNS requests to go from the VM to the dnsmasq of the host. The issue that this commit fixes is the fact that a DROP policy on the OUTPUT chain blocks the DHCP replies from the host’s dnsmasq to the VM. As DHCP replies are sent in UDP, they are not caught by any --ctstate ESTABLISHED rule and so, need to be explicitly allowed. Signed-off-by: Lénaïc Huard <lenaic(a)lhuard.fr.eu.org> --- src/libvirt_private.syms | 2 ++ src/network/bridge_driver_linux.c | 35 +++++++++++++--------- src/util/viriptables.c | 61 +++++++++++++++++++++++++++++++++++++++ src/util/viriptables.h | 7 +++++ 4 files changed, 92 insertions(+), 13 deletions(-) diff --git a/src/libvirt_private.syms b/src/libvirt_private.syms index 2dbb8f8..c26ce29 100644 --- a/src/libvirt_private.syms +++ b/src/libvirt_private.syms @@ -1290,6 +1290,7 @@ iptablesAddForwardRejectOut; iptablesAddOutputFixUdpChecksum; iptablesAddTcpInput; iptablesAddUdpInput; +iptablesAddUdpOutput; iptablesRemoveDontMasquerade; iptablesRemoveForwardAllowCross; iptablesRemoveForwardAllowIn; @@ -1301,6 +1302,7 @@ iptablesRemoveForwardRejectOut; iptablesRemoveOutputFixUdpChecksum; iptablesRemoveTcpInput; iptablesRemoveUdpInput; +iptablesRemoveUdpOutput; # util/virjson.h diff --git a/src/network/bridge_driver_linux.c b/src/network/bridge_driver_linux.c index 066779a..dcdd033 100644 --- a/src/network/bridge_driver_linux.c +++ b/src/network/bridge_driver_linux.c @@ -564,6 +564,13 @@ int networkAddGeneralFirewallRules(virNetworkObjPtr network) goto err2; } + if (iptablesAddUdpOutput(AF_INET, network->def->bridge, 68) < 0) { + virReportError(VIR_ERR_SYSTEM_ERROR, + _("failed to add iptables rule to allow DHCP replies to '%s'"), + network->def->bridge); + goto err3; + } + /* If we are doing local DHCP service on this network, attempt to * add a rule that will fixup the checksum of DHCP response * packets back to the guests (but report failure without @@ -582,14 +589,14 @@ int networkAddGeneralFirewallRules(virNetworkObjPtr network) virReportError(VIR_ERR_SYSTEM_ERROR, _("failed to add iptables rule to allow DNS requests from '%s'"), network->def->bridge); - goto err3; + goto err4; } if (iptablesAddUdpInput(AF_INET, network->def->bridge, 53) < 0) { virReportError(VIR_ERR_SYSTEM_ERROR, _("failed to add iptables rule to allow DNS requests from '%s'"), network->def->bridge); - goto err4; + goto err5; } /* allow TFTP requests through to dnsmasq if necessary */ @@ -598,7 +605,7 @@ int networkAddGeneralFirewallRules(virNetworkObjPtr network) virReportError(VIR_ERR_SYSTEM_ERROR, _("failed to add iptables rule to allow TFTP requests from '%s'"), network->def->bridge); - goto err5; + goto err6; } /* Catch all rules to block forwarding to/from bridges */ @@ -607,14 +614,14 @@ int networkAddGeneralFirewallRules(virNetworkObjPtr network) virReportError(VIR_ERR_SYSTEM_ERROR, _("failed to add iptables rule to block outbound traffic from '%s'"), network->def->bridge); - goto err6; + goto err7; } if (iptablesAddForwardRejectIn(AF_INET, network->def->bridge) < 0) { virReportError(VIR_ERR_SYSTEM_ERROR, _("failed to add iptables rule to block inbound traffic to '%s'"), network->def->bridge); - goto err7; + goto err8; } /* Allow traffic between guests on the same bridge */ @@ -622,31 +629,33 @@ int networkAddGeneralFirewallRules(virNetworkObjPtr network) virReportError(VIR_ERR_SYSTEM_ERROR, _("failed to add iptables rule to allow cross bridge traffic on '%s'"), network->def->bridge); - goto err8; + goto err9; } /* add IPv6 general rules, if needed */ if (networkAddGeneralIp6tablesRules(network) < 0) { - goto err9; + goto err10; } return 0; /* unwind in reverse order from the point of failure */ -err9: +err10: iptablesRemoveForwardAllowCross(AF_INET, network->def->bridge); -err8: +err9: iptablesRemoveForwardRejectIn(AF_INET, network->def->bridge); -err7: +err8: iptablesRemoveForwardRejectOut(AF_INET, network->def->bridge); -err6: +err7: if (ipv4def && ipv4def->tftproot) { iptablesRemoveUdpInput(AF_INET, network->def->bridge, 69); } -err5: +err6: iptablesRemoveUdpInput(AF_INET, network->def->bridge, 53); -err4: +err5: iptablesRemoveTcpInput(AF_INET, network->def->bridge, 53); +err4: + iptablesRemoveUdpOutput(AF_INET, network->def->bridge, 68); err3: iptablesRemoveUdpInput(AF_INET, network->def->bridge, 67); err2: diff --git a/src/util/viriptables.c b/src/util/viriptables.c index a1711bb..9b78d86 100644 --- a/src/util/viriptables.c +++ b/src/util/viriptables.c @@ -184,6 +184,28 @@ iptablesInput(int family, NULL); } +static int +iptablesOutput(int family, + const char *iface, + int port, + int action, + int tcp) +{ + char portstr[32]; + + snprintf(portstr, sizeof(portstr), "%d", port); + portstr[sizeof(portstr) - 1] = '\0'; + + return iptablesAddRemoveRule("filter", "OUTPUT", + family, + action, + "--out-interface", iface, + "--protocol", tcp ? "tcp" : "udp", + "--destination-port", portstr, + "--jump", "ACCEPT", + NULL); +} + /** * iptablesAddTcpInput: * @ctx: pointer to the IP table context @@ -262,6 +284,45 @@ iptablesRemoveUdpInput(int family, return iptablesInput(family, iface, port, REMOVE, 0); } +/** + * iptablesAddUdpOutput: + * @ctx: pointer to the IP table context + * @iface: the interface name + * @port: the UDP port to add + * + * Add an output to the IP table allowing access to the given @port from + * the given @iface interface for UDP packets + * + * Returns 0 in case of success or an error code in case of error + */ + +int +iptablesAddUdpOutput(int family, + const char *iface, + int port) +{ + return iptablesOutput(family, iface, port, ADD, 0); +} + +/** + * iptablesRemoveUdpOutput: + * @ctx: pointer to the IP table context + * @iface: the interface name + * @port: the UDP port to remove + * + * Removes an output from the IP table, hence forbidding access to the given + * @port from the given @iface interface for UDP packets + * + * Returns 0 in case of success or an error code in case of error + */ +int +iptablesRemoveUdpOutput(int family, + const char *iface, + int port) +{ + return iptablesOutput(family, iface, port, REMOVE, 0); +} + static char *iptablesFormatNetwork(virSocketAddr *netaddr, unsigned int prefix) diff --git a/src/util/viriptables.h b/src/util/viriptables.h index 6dfb992..2f9a212 100644 --- a/src/util/viriptables.h +++ b/src/util/viriptables.h @@ -40,6 +40,13 @@ int iptablesRemoveUdpInput (int family, const char *iface, int port); +int iptablesAddUdpOutput (int family, + const char *iface, + int port); +int iptablesRemoveUdpOutput (int family, + const char *iface, + int port); + int iptablesAddForwardAllowOut (virSocketAddr *netaddr, unsigned int prefix, const char *iface, -- 1.8.5.1

2 1

[libvirt] [PATCH] qemu: auto-create pci controller alias if missing from domain status
by Laine Stump 07 Jan '14

07 Jan '14

https://bugzilla.redhat.com/show_bug.cgi?id=1047429 The problem occurs when a host with a running guest is upgraded from pre-1.0.2 libvirt, and then a later attempt is made to hotplug a new PCI device (emulated or passthrough) to that guest without having power cycled / save-restored / migrated the guest. The hotplug fails with this error: internal error: Device alias was not set for PCI controller with index 0 required for device at address 0000:00:xx.x This happens because the pci controller newly auto-created for the status will not contain an alias because 1) assigning an alias is normally left up to the hypervisor driver to do at domain startup, or controller attach time. 2) under the older libvirt, the PCI controller was implicit, so it was never assigned an alias. 3) when libvirtd is restarted and re-attaches to each domain, there is no "re-attach device" operation, so the no-longer-implicit PCI controller will not get an alias assigned either. When explicit PCI controllers were first added (shortly before libvirt 1.0.2), the code was written such that each controller would get an alias named "pciN" placed in its status, but the commandlines would all ignore that and instead use "pci.N". This was changed in commit 01b8812 so that 1) the proper "pci.N" alias was stored in the status, 2) the alias name from the status was used when creating the qemu commandline, and 3) when a PCI device was being hotplugged, if the status for its PCI controller had no alias assigned, an internal error would be generated. This patch changes that error condition to instead make a temporary string to contain the required alias "pci.N" and put that in the proper place in the commandline. It does not fill in the alias in the status object just in case that might create some other problem (it didn't seem very polite for the qemuBuildDeviceAddressStr() for one device to be modifying the status object for a different device). Note that this does *not* address a similar problem that would occur when libvirt was upgraded from a version between 1.0.2 and 1.1.2 to a newer version - in these cases the status would contain "pciN" as the alias name, so qemuBuildDeviceAddressStr() would be successful, but the resulting command would fail (since the qemu process has no controller named "pciN", but instead has one named "pci.N"); the problem can anyway be worked around by restarting / save-restoring / migrating the domain, and solving this second problem would require special-casing a particular alias name pattern rather than just handling a missing name, which seems just a bit too ugly to have permanently cluttering the code. --- (Yes, another exceedingly long commit log message for a very short patch :-) I'm not 100% convinced that it's necessary to handle *this* case either (upgrading from pre-1.0.2), but thought I should send the patch to the list anyway to see what others think. My opinion is that it's creating permanent clutter in the code in order to handle a situation that will come up just once in the entire lifetime of a host (and even then only if libvirt is upgraded without somehow restarting a guest, and *then* a device is hotplugged to that guest), so it may be better if we just left the workaround ("restart/save-restore/migrate the domain if you need to hotplug a device") as a footnote somewhere instead. src/qemu/qemu_command.c | 20 +++++++++++++------- 1 file changed, 13 insertions(+), 7 deletions(-) diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index d723dc8..a562e18 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -1,7 +1,7 @@ /* * qemu_command.c: QEMU command generation * - * Copyright (C) 2006-2013 Red Hat, Inc. + * Copyright (C) 2006-2014 Red Hat, Inc. * Copyright (C) 2006 Daniel P. Berrange * * This library is free software; you can redistribute it and/or @@ -3021,6 +3021,7 @@ qemuBuildDeviceAddressStr(virBufferPtr buf, { int ret = -1; char *devStr = NULL; + char *tempAlias = NULL; if (info->type == VIR_DOMAIN_DEVICE_ADDRESS_TYPE_PCI) { const char *contAlias = NULL; @@ -3035,12 +3036,16 @@ qemuBuildDeviceAddressStr(virBufferPtr buf, cont->idx == info->addr.pci.bus) { contAlias = cont->info.alias; if (!contAlias) { - virReportError(VIR_ERR_INTERNAL_ERROR, - _("Device alias was not set for PCI " - "controller with index %u required " - "for device at address %s"), - info->addr.pci.bus, devStr); - goto cleanup; + /* this could be a case where the domain had been + * started with an earlier version of libvirt that + * had no device entry for the PCI bus + * (pre-1.0.2), and in that case no alias would be + * set in the status XML, so we create one using + * the standard "pci.N" pattern. + */ + if (virAsprintf(&tempAlias, "pci.%u", info->addr.pci.bus) < 0) + goto cleanup; + contAlias = tempAlias; } break; } @@ -3118,6 +3123,7 @@ qemuBuildDeviceAddressStr(virBufferPtr buf, ret = 0; cleanup: VIR_FREE(devStr); + VIR_FREE(tempAlias); return ret; } -- 1.8.4.2

2 3

[libvirt] [PATCH] lxc_controller: Don't leak @name
by Michal Privoznik 07 Jan '14

07 Jan '14

The @name variable is VIR_STRDUP()-ed into, but never freed. In fact, there's no need to duplicate a command line argument since all places where @name is used expect const char. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- src/lxc/lxc_controller.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/src/lxc/lxc_controller.c b/src/lxc/lxc_controller.c index c013147..1b43bd2 100644 --- a/src/lxc/lxc_controller.c +++ b/src/lxc/lxc_controller.c @@ -2249,7 +2249,7 @@ int main(int argc, char *argv[]) { pid_t pid; int rc = -1; - char *name = NULL; + const char *name = NULL; size_t nveths = 0; char **veths = NULL; int handshakeFd = -1; @@ -2300,8 +2300,7 @@ int main(int argc, char *argv[]) break; case 'n': - if (VIR_STRDUP(name, optarg) < 0) - goto cleanup; + name = optarg; break; case 'v': -- 1.8.5.1

2 3

[libvirt] [PATCH] qemuProcessStop: Don't overwrite any errors
by Michal Privoznik 07 Jan '14

07 Jan '14

Currently, the qemuProcessStop tries to open the domain log file and saves the original error afterwards. Then all the cleanup is done after which the error is restored back. This has however one flaw: if opening of the log file fails an error is reported, which results in previous error being overwritten (the useful one, e.g. "PCI device XXXX:XXXX could not be found"). Hence, user sees something like: error: failed to create logfile /var/log/libvirt/qemu/ovirt_usb.log: No such file or directory instead of: error: internal error: Did not find USB device 8644:8003 Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> Reported-by: Zhou Yimin <zhouyimin(a)huawei.com> --- src/qemu/qemu_process.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index d0fde54..9331744 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -4224,6 +4224,10 @@ void qemuProcessStop(virQEMUDriverPtr driver, return; } + /* This method is routinely used in clean up paths. Disable error + * reporting so we don't squash a legit error. */ + orig_err = virSaveLastError(); + /* * We may unlock the vm in qemuProcessKill(), and another thread * can lock the vm, and then call qemuProcessStop(). So we should @@ -4256,10 +4260,6 @@ void qemuProcessStop(virQEMUDriverPtr driver, virStrerror(errno, ebuf, sizeof(ebuf))); } - /* This method is routinely used in clean up paths. Disable error - * reporting so we don't squash a legit error. */ - orig_err = virSaveLastError(); - virDomainConfVMNWFilterTeardown(vm); if (cfg->macFilter) { -- 1.8.5.1

2 1

[libvirt] [sandbox PATCH] Fix FSF address on bin/virt-sandbox-service
by Cédric Bosdonnat 07 Jan '14

07 Jan '14

--- bin/virt-sandbox-service | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/bin/virt-sandbox-service b/bin/virt-sandbox-service index 2dcbfb8..a7e71cd 100755 --- a/bin/virt-sandbox-service +++ b/bin/virt-sandbox-service @@ -16,7 +16,8 @@ # # You should have received a copy of the GNU General Public License # along with this program; if not, write to the Free Software -# Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. +# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, +# MA 02110-1301 USA. # from gi.repository import LibvirtGObject -- 1.8.5.2

3 2

[libvirt] [PATCH] maint: fix comment typos in qemu numa code
by Eric Blake 07 Jan '14

07 Jan '14

Introduced in commit 81fae6b9. * src/qemu/qemu_driver.c (qemuDomainSetNumaParamsLive): Fix typos. Signed-off-by: Eric Blake <eblake(a)redhat.com> --- Pushing under the trivial rule. src/qemu/qemu_driver.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 7e45ffc..004ec88 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -1,7 +1,7 @@ /* * qemu_driver.c: core driver methods for managing qemu guests * - * Copyright (C) 2006-2013 Red Hat, Inc. + * Copyright (C) 2006-2014 Red Hat, Inc. * Copyright (C) 2006 Daniel P. Berrange * * This library is free software; you can redistribute it and/or @@ -8161,7 +8161,7 @@ qemuDomainSetNumaParamsLive(virDomainObjPtr vm, goto cleanup; } - /*Get Exisitng nodeset values */ + /* Get existing nodeset values */ if (virCgroupGetCpusetMems(priv->cgroup, &nodeset_str) < 0 || virBitmapParse(nodeset_str, 0, &temp_nodeset, VIR_DOMAIN_CPUMASK_LEN) < 0) @@ -8192,7 +8192,7 @@ qemuDomainSetNumaParamsLive(virDomainObjPtr vm, goto cleanup; VIR_FREE(nodeset_str); - /* Ensure the cpuset string is formated before passing to cgroup */ + /* Ensure the cpuset string is formatted before passing to cgroup */ if (!(nodeset_str = virBitmapFormat(nodeset))) { virReportError(VIR_ERR_INTERNAL_ERROR, "%s", _("Failed to format nodeset")); -- 1.8.4.2

1 0

[libvirt] [PATCH v3] BSD: implement nodeGetMemoryStats
by Roman Bogorodskiy 07 Jan '14

07 Jan '14

Add a BSD implementation of nodeGetMemoryStats based on sysctl(3). --- src/nodeinfo.c | 82 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 81 insertions(+), 1 deletion(-) diff --git a/src/nodeinfo.c b/src/nodeinfo.c index 1838547..05bc038 100644 --- a/src/nodeinfo.c +++ b/src/nodeinfo.c @@ -96,7 +96,85 @@ appleFreebsdNodeGetMemorySize(unsigned long *memory) return 0; } -#endif +#endif /* defined(__FreeBSD__) || defined(__APPLE__) */ + +#ifdef __FreeBSD__ +# define BSD_MEMORY_STATS_ALL 4 + +static int +freebsdNodeGetMemoryStats(virNodeMemoryStatsPtr params, + int *nparams) +{ + size_t i, j = 0; + unsigned long pagesize = getpagesize() >> 10; + long bufpages; + size_t bufpages_size = sizeof(bufpages); + struct field_sysctl_map { + const char *field; + const char *sysctl_name; + } sysctl_map[] = { + {VIR_NODE_MEMORY_STATS_TOTAL, "vm.stats.vm.v_page_count"}, + {VIR_NODE_MEMORY_STATS_FREE, "vm.stats.vm.v_free_count"}, + {VIR_NODE_MEMORY_STATS_CACHED, "vm.stats.vm.v_cache_count"}, + {NULL, NULL} + }; + + if ((*nparams) == 0) { + *nparams = BSD_MEMORY_STATS_ALL; + return 0; + } + + if ((*nparams) != BSD_MEMORY_STATS_ALL) { + virReportInvalidArg(nparams, + _("nparams in %s must be %d"), + __FUNCTION__, BSD_MEMORY_STATS_ALL); + return -1; + } + + for (i = 0; sysctl_map[i].field != NULL; i++) { + u_int value; + size_t value_size = sizeof(value); + virNodeMemoryStatsPtr param; + + if (sysctlbyname(sysctl_map[i].sysctl_name, &value, + &value_size, NULL, 0) < 0) { + virReportSystemError(errno, + _("sysctl failed for '%s'"), + sysctl_map[i].sysctl_name); + return -1; + } + + param = &params[j++]; + if (virStrcpyStatic(param->field, sysctl_map[i].field) == NULL) { + virReportError(VIR_ERR_INTERNAL_ERROR, + _("Field '%s' too long for destination"), + sysctl_map[i].field); + return -1; + } + param->value = (unsigned long long)value * pagesize; + } + + { + virNodeMemoryStatsPtr param = &params[j++]; + + if (sysctlbyname("vfs.bufspace", &bufpages, &bufpages_size, NULL, 0) < 0) { + virReportSystemError(errno, + _("sysctl failed for '%s'"), + "vfs.bufspace"); + return -1; + } + if (virStrcpyStatic(param->field, VIR_NODE_MEMORY_STATS_BUFFERS) == NULL) { + virReportError(VIR_ERR_INTERNAL_ERROR, + _("Field '%s' too long for destination"), + VIR_NODE_MEMORY_STATS_BUFFERS); + return -1; + } + param->value = (unsigned long long)bufpages >> 10; + } + + return 0; +} +#endif /* __FreeBSD__ */ #ifdef __linux__ # define CPUINFO_PATH "/proc/cpuinfo" @@ -1041,6 +1119,8 @@ int nodeGetMemoryStats(int cellNum ATTRIBUTE_UNUSED, return ret; } +#elif defined(__FreeBSD__) + return freebsdNodeGetMemoryStats(params, nparams); #else virReportError(VIR_ERR_NO_SUPPORT, "%s", _("node memory stats not implemented on this platform")); -- 1.8.4.3

3 3

[libvirt] [PATCH] qemuStateInitialize: Create qemu log directory
by Wangyufei (James) 07 Jan '14

07 Jan '14

>From 51782a44eecf5801e5100920e5965d8dfc46e4cc Mon Sep 17 00:00:00 2001 From: Zhou Yimin <zhouyimin(a)huawei.com> Date: Thu, 2 Jan 2014 16:32:46 +0800 Subject: [PATCH] qemuStateInitialize: Create qemu log directory When an error occurred in qemuProcessStart before qemu log directory created, it will fail to create qemu log file in qemuProcessStop during qemuProcessStart's cleanup phase. So when I start a VM whose PCI device not exist, the last error will be overwrited as "failed to create logfile XXX: No such file or directory" not the real reason "Device XXX not found:could not access XXX: No such file or directory". I fix it as following. Signed-off-by: Zhou Yimin <zhouyimin(a)huawei.com> Reviewed-by: Wang Yufei <james.wangyufei(a)huawei.com> --- src/qemu/qemu_driver.c | 5 +++++ src/qemu/qemu_process.c | 14 -------------- 2 files changed, 5 insertions(+), 14 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 45d11cd..ca3208e 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -615,6 +615,11 @@ qemuStateInitialize(bool privileged, goto error; VIR_FREE(driverConf); + if (virFileMakePath(cfg->logDir) < 0) { + VIR_ERROR(_("Failed to create log dir '%s': %s"), + cfg->logDir, virStrerror(errno, ebuf, sizeof(ebuf))); + goto error; + } if (virFileMakePath(cfg->stateDir) < 0) { VIR_ERROR(_("Failed to create state dir '%s': %s"), cfg->stateDir, virStrerror(errno, ebuf, sizeof(ebuf))); diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index d0fde54..0382c06 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -3737,13 +3737,6 @@ int qemuProcessStart(virConnectPtr conn, } } - if (virFileMakePath(cfg->logDir) < 0) { - virReportSystemError(errno, - _("cannot create log directory %s"), - cfg->logDir); - goto cleanup; - } - VIR_DEBUG("Creating domain log file"); if ((logfile = qemuDomainCreateLog(driver, vm, false)) < 0) goto cleanup; @@ -4503,13 +4496,6 @@ int qemuProcessAttach(virConnectPtr conn ATTRIBUTE_UNUSED, driver->inhibitCallback(true, driver->inhibitOpaque); active = true; - if (virFileMakePath(cfg->logDir) < 0) { - virReportSystemError(errno, - _("cannot create log directory %s"), - cfg->logDir); - goto error; - } - VIR_FREE(priv->pidfile); if (VIR_STRDUP(priv->pidfile, pidfile) < 0) goto error; -- 1.7.3.1.msysgit.0 Best Regards, -WangYufei

3 2

[libvirt] [PATCH] Fix argument order at qemuMigrationPerformJob().
by Minoru Usui 07 Jan '14

07 Jan '14

Hi, everyone. I found a small bug at qemuMigrationPerformJob(). The order of argument, @listenAddress and @cookiein, should be changed place, because these order are mismatched caller and callee. Signed-off-by: Minoru Usui <usui(a)mxm.nes.nec.co.jp> --- src/qemu/qemu_migration.c | 2 +- 1 files changed, 1 insertions(+), 1 deletions(-) diff --git a/src/qemu/qemu_migration.c b/src/qemu/qemu_migration.c index 9342062..53a4ae6 100644 --- a/src/qemu/qemu_migration.c +++ b/src/qemu/qemu_migration.c @@ -4045,8 +4045,8 @@ qemuMigrationPerformJob(virQEMUDriverPtr driver, const char *dconnuri, const char *uri, const char *graphicsuri, - const char *cookiein, const char *listenAddress, + const char *cookiein, int cookieinlen, char **cookieout, int *cookieoutlen, -- 1.7.2.3 -- Minoru Usui <usui(a)mxm.nes.nec.co.jp>

2 2

[libvirt] [PATCH 0/2] Fixes on shareable SCSI host device
by Osier Yang 07 Jan '14

07 Jan '14

It's easy to reproduce with two domains using the same SCSI generic device, both with <shareable> specified. It's expected to work. But the fact is: % virsh start f20-1 error: Failed to start domain b error: Requested operation is not valid: SCSI device 1:0:0:0 is in use by domain f20-0 Osier Yang (2): util: Add "shareable" field for virSCSIDevice struct qemu: Don't fail if the SCSI host device is shareable between domains src/libvirt_private.syms | 1 + src/qemu/qemu_cgroup.c | 3 ++- src/qemu/qemu_hostdev.c | 42 +++++++++++++++++++++++----------------- src/security/security_apparmor.c | 3 ++- src/security/security_dac.c | 6 ++++-- src/security/security_selinux.c | 6 ++++-- src/util/virscsi.c | 11 ++++++++++- src/util/virscsi.h | 4 +++- 8 files changed, 50 insertions(+), 26 deletions(-) -- 1.8.1.4

2 5