- Devel - libvirt List Archives

[libvirt] [PATCH 0/3] Three simple bugfixes
by Laine Stump 22 May '15

22 May '15

Patch 1 was posted several days ago, but may have gotten lost in the stream, so I'm reposting it. (The patches are unrelated other than that they are nominally related to networking, they were all in my queue at the same time, and they're all three very short. I'm just posting them together to make it easier to keep track of them.) Laine Stump (3): netdev: fail when setting up an SRIOV VF if PF is offline interface: allow multiple IPv4 addresses + dhcp on a single interface util: better error message after failure to initialize firewall backend docs/schemas/interface.rng | 28 ++++++++++------------ src/util/virfirewall.c | 10 +++++++- src/util/virnetdev.c | 22 +++++++++++++++++ .../ethernet-dhcp-and-multi-static.xml | 9 +++++++ tests/interfacexml2xmltest.c | 1 + 5 files changed, 54 insertions(+), 16 deletions(-) create mode 100644 tests/interfaceschemadata/ethernet-dhcp-and-multi-static.xml -- 2.1.0

2 5

[libvirt] [PATCH] conf: fix issue on virCPUDefCopy
by Thibaut Collet 22 May '15

22 May '15

This patch is only for maintenance branches v1.2.9-maint, v1.2.10-maint, v1.2.11-maint and v1.2.12-maint Thibaut Collet (1): conf: fix issue on virCPUDefCopy src/conf/cpu_conf.c | 1 + 1 file changed, 1 insertion(+) -- 1.7.10.4

2 2

[libvirt] [RFC PATCH 0/2] Clear capabilities cache options
by John Ferlan 22 May '15

22 May '15

Posting this as an RFC since I'm trying to get upstream feedback on whether either approach is better or whether perhaps both approaches should be used. The problem occurred primarily for downstream installations where it seems the order of the build libvirtd caused issues regarding whether certain capabilities bits were present. Although the history of the case doesn't "prove" my theory exactly other issues around the same time that were part of a private bz or email stream (I forget now) seem to indicate that the order of libvirtd builds and installations would be very important as they relate to when the cache is rebuilt. In the instances seen there was lots of testing going on related to whether patches were found in a 'maintenance release' and the 'mainline release'. Since the process is build the mainline first and then backport the patch(es) to the maintenance release, it's possible the date of the maintenance libvirtd would be later than that of the mainline release. If testing installs the maintenance release first and the cache gets updated based on the 'ctime' of libvirtd being later than what was currently installed and then the mainline release is installed with a libvirtd 'ctime' that would possibly be prior to that of the maintentance release, then the cache wouldn't be updated. The fix for testing has been just to delete the cache. In looking at the code and history, I see a couple possible solutions. The first seems to be a "big hammer" - on every libvirtd installation we delete the entire cache - this would force the reread. I believe I got that right using %post of the install to clear out anything in the capabilities directory (whether just created or already existing from some prior installation). Another solution was to adjust the libvirtd 'ctime' check to be if different than what is found in the cache. When looking at the original commit, this seems to be what the intention was in the commit message, but the actual code was slightly different. Upstream review comments for changes didn't mention the difference either, see: http://www.redhat.com/archives/libvir-list/2014-March/msg00629.html and https://www.redhat.com/archives/libvir-list/2014-March/msg00301.html John Ferlan (2): spec: Remove capabilities cache during %post install qemu: Force capabilities cache read if libvirtd date is different libvirt.spec.in | 11 +++++++++++ src/qemu/qemu_capabilities.c | 6 ++++-- 2 files changed, 15 insertions(+), 2 deletions(-) -- 2.1.0

4 15

[libvirt] RFC: exposing qemu's block-set-write-threshold
by Eric Blake 22 May '15

22 May '15

I'm trying to wire up libvirt to do event reporting based on qemu 2.3's BLOCK_WRITE_THRESHOLD event. Doing this will allow management applications to do event-based notification on when to enlarge LVM (or other) storage underlying a qcow2 volume, rather than their current requirement to frequently poll block statistics. But I'm stuck on the best way to expose the new parameter: One idea is to treat it as part of the domain XML, and have virDomainSetBlockIoTune add one more typed parameter for a disk's current write threshold. Doing this could allow setting a threshold even for an offline domain (although the threshold is only meaningful for a running domain), but might get weird because qemu's event is one-shot (you have to re-arm a new threshold every time an existing threshold fires - so every time it fires, the domain XML is rewritten, even though it is not guest-visible ABI that was changing). At least with this approach, it is also easy for a client to poll the current setting of the threshold, via virDomainGetBlockIoTune. But the threshold isn't quite a tuning parameter (it isn't throttling how fast the guest can write to the block device, only how full the host side can get in order to allow transparent resizing of the host storage prior to running out of space). Another idea is to add a completely new API, maybe named virDomainBlockSetWriteThreshold(virDomainPtr dom, const char *disk, long long int threshold, unsigned int flags) (with threshold in bytes). Here, virDomainBlockStatsFlags() could be a way to query the current threshold. And if desired, we could add a flag value to treat threshold as a percentage instead of a byte value (but is 1% too large of granularity, and how would you scale the percentage to anything finer while still keeping the parameter as long long int rather than double?) Of course, I'd want virConnectGetAllDomainStats() to list the current threshold setting (0 if no threshold or if the event has already fired, non-zero if the threshold is still set waiting to fire), so that clients can query thresholds for multiple domains and multiple disks per domain in one API call. But I don't know if we have any good way to set multiple thresholds in one call (at least virDomainSetBlockIoTune must be called once per disk; it might be possible for my proposed virDomainBlockStatsFlags() to set a threshold for multiple disks if the disk name is passed as NULL - but then we're back to the question of what happens if the guest has multiple disks of different sizes; it's better to set per-disk thresholds than to assume all disks must be at the same byte or percentage threshold). I'm also worried about what happens across libvirtd restarts - if the qemu event fires while libvirtd is unconnected, should libvirt be tracking that a threshold was registered in the XML, and upon reconnection check if qemu still has the threshold? If qemu no longer has a threshold, then libvirt can assume it missed the event, and generate one as part of reconnecting to the domain. Thoughts? If I don't hear anything, I'll base my first implementation on adding a new API, rather than reusing virDomainSetBlockIoTune. -- Eric Blake eblake redhat com +1-919-301-3266 Libvirt virtualization library http://libvirt.org

4 8

[libvirt] [PATCH] conf: fix forget restore ctxt->node when parse memory device success
by Luyao Huang 22 May '15

22 May '15

https://bugzilla.redhat.com/show_bug.cgi?id=1223631 When set a memory device in the xml, <sysinfo> in xml will be lost. Because we forgot restore ctxt->node to the oldnode after parse memory device, this will make the parse function after virDomainMemoryDefParseXML cannot find a node they need when parse a full xml(virDomainDefParseXML). Signed-off-by: Luyao Huang <lhuang(a)redhat.com> --- src/conf/domain_conf.c | 1 + 1 file changed, 1 insertion(+) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index bfdc94e..7ddc1ea 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -11828,6 +11828,7 @@ virDomainMemoryDefParseXML(xmlNodePtr memdevNode, if (virDomainDeviceInfoParseXML(memdevNode, NULL, &def->info, flags) < 0) goto error; + ctxt->node = save; return def; error: -- 1.8.3.1

4 4

[libvirt] [PATCH] virsh: fix a typos issue cause error not right
by Luyao Huang 22 May '15

22 May '15

Signed-off-by: Luyao Huang <lhuang(a)redhat.com> --- tools/virsh-host.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/virsh-host.c b/tools/virsh-host.c index a72fd05..e939dac 100644 --- a/tools/virsh-host.c +++ b/tools/virsh-host.c @@ -500,7 +500,7 @@ cmdAllocpages(vshControl *ctl, const vshCmd *cmd) if (vshCommandOptScaledInt(cmd, "pagesize", &tmp, 1024, UINT_MAX) < 0) { vshError(ctl, _("Numeric value for <%s> option is malformed or out of range"), - "cellno"); + "pagesize"); return false; } pageSizes[0] = VIR_DIV_UP(tmp, 1024); -- 1.8.3.1

3 2

[libvirt] [PATCH v2 0/5] storage: fs: tweak dir perms handling on build
by Cole Robinson 21 May '15

21 May '15

Consider the following issue - Using virt-manager with qemu:///session - User adds a storage pool pointing at /tmp. No explicit permissions are requested in the XML - virt-manager calls PoolDefine, then PoolBuild - libvirt tries to unconditionally chmod 755 /tmp. This fails because my user doesn't own root. Pool build fails, virt-manager reports failure Yes there's a couple ways we could avoid this specific case in virt-manager, but I think it makes more sense to have pool.build on a directory be a no-op in this case. The following patches address this. I already pushed some bits of v1 of the series. Now the patches are: - Patch 1 (new) fixes a bug in the previous patches I pushed - Patch 2 (new) changes storage XML to not output -1 for owner/group - Patch 3 changes storage XML to not hardcode a <mode> value if the user didn't specify one in. - Patch 4 (new) drops a redundant flag to virDirCreate - Patch 5 makes pool.build skip dir chmod unless the user explicitly requested <mode> via the XML. However if a mode is required for mkdir, continue to use the previous default. v2: Add patches 1, 3, 4, adjust other patches to match Style fix in patch 5 Cole Robinson (5): virfile: virDirCreate: Fix ALLOW_EXIST conditional storage: conf: Don't output owner/group -1 storage: conf: Don't set any default <mode> in the XML virfile: virDirCreate: Drop redundand FORCE_PERMS flag storage: fs: Only force directory permissions if required docs/schemas/storagecommon.rng | 36 ++++++++------ src/conf/storage_conf.c | 57 +++++++++++----------- src/storage/storage_backend.c | 20 ++++++-- src/storage/storage_backend.h | 3 ++ src/storage/storage_backend_fs.c | 21 +++++--- src/storage/storage_backend_logical.c | 4 +- src/util/virfile.c | 8 ++- src/util/virfile.h | 3 +- tests/storagepoolxml2xmlout/pool-dir-naming.xml | 2 - tests/storagepoolxml2xmlout/pool-dir.xml | 2 - tests/storagepoolxml2xmlout/pool-netfs-gluster.xml | 3 -- tests/storagevolxml2xmlin/vol-file.xml | 4 +- .../vol-gluster-dir-neg-uid.xml | 2 - tests/storagevolxml2xmlout/vol-gluster-dir.xml | 3 -- tests/storagevolxml2xmlout/vol-sheepdog.xml | 3 -- 15 files changed, 91 insertions(+), 80 deletions(-) -- 2.4.0

4 15

[libvirt] [PATCH] build fix: fix libvirt build if WITH_HAL is defined
by Maxim Nestratov 21 May '15

21 May '15

--- src/node_device/node_device_hal.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/node_device/node_device_hal.c b/src/node_device/node_device_hal.c index 2d3bc17..6d18a87 100644 --- a/src/node_device/node_device_hal.c +++ b/src/node_device/node_device_hal.c @@ -372,7 +372,7 @@ gather_capability(LibHalContext *ctx, const char *udi, virNodeDevCapsDefPtr caps; if (VIR_ALLOC(caps) < 0) return ENOMEM; - caps->type = entry->type; + caps->data.type = entry->type; if (entry->gather_fn) { int rv = (*entry->gather_fn)(ctx, udi, &caps->data); if (rv != 0) { -- 2.1.0

2 1

[libvirt] [PATCH v2] sysinfo: Fix reports on ARM
by Michal Privoznik 21 May '15

21 May '15

Due to a kernel commit (b4b8f770e), cpuinfo format has changed on ARMs. Firstly, 'Processor: ...' may not be reported, it's replaced by 'model name: ...'. Secondly, the "Processor" string may occur in CPU name, e.g. 'ARMv7 Processor rev 5 (v7l)'. Therefore, we must firstly look for 'model name' and then for 'Processor' if not found. Moreover, lines in the cpuinfo file are shuffled, so we better not manipulate the pointer to start of internal buffer as we may lost some info. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- diff to v1: - Jan's comments worked in - added new test case (yes, there really is a space at EOL in cpuinfo on my ARM) src/util/virsysinfo.c | 9 +++---- tests/sysinfodata/arm-rpi2cpuinfo.data | 43 ++++++++++++++++++++++++++++++++ tests/sysinfodata/arm-rpi2sysinfo.expect | 18 +++++++++++++ tests/sysinfotest.c | 22 ++++++++++++---- 4 files changed, 82 insertions(+), 10 deletions(-) create mode 100644 tests/sysinfodata/arm-rpi2cpuinfo.data create mode 100644 tests/sysinfodata/arm-rpi2sysinfo.expect diff --git a/src/util/virsysinfo.c b/src/util/virsysinfo.c index 8bb17f0..40390ab 100644 --- a/src/util/virsysinfo.c +++ b/src/util/virsysinfo.c @@ -289,16 +289,15 @@ virSysinfoParseProcessor(const char *base, virSysinfoDefPtr ret) virSysinfoProcessorDefPtr processor; char *processor_type = NULL; - if (!(tmp_base = strstr(base, "Processor"))) + if (!(tmp_base = strstr(base, "model name")) && + !(tmp_base = strstr(base, "Processor"))) return 0; - base = tmp_base; - eol = strchr(base, '\n'); - cur = strchr(base, ':') + 1; + eol = strchr(tmp_base, '\n'); + cur = strchr(tmp_base, ':') + 1; virSkipSpaces(&cur); if (eol && VIR_STRNDUP(processor_type, cur, eol - cur) < 0) goto error; - base = cur; while ((tmp_base = strstr(base, "processor")) != NULL) { base = tmp_base; diff --git a/tests/sysinfodata/arm-rpi2cpuinfo.data b/tests/sysinfodata/arm-rpi2cpuinfo.data new file mode 100644 index 0000000..41fde00 --- /dev/null +++ b/tests/sysinfodata/arm-rpi2cpuinfo.data @@ -0,0 +1,43 @@ +processor : 0 +model name : ARMv7 Processor rev 5 (v7l) +BogoMIPS : 38.40 +Features : half thumb fastmult vfp edsp neon vfpv3 tls vfpv4 idiva idivt vfpd32 lpae evtstrm +CPU implementer : 0x41 +CPU architecture: 7 +CPU variant : 0x0 +CPU part : 0xc07 +CPU revision : 5 + +processor : 1 +model name : ARMv7 Processor rev 5 (v7l) +BogoMIPS : 38.40 +Features : half thumb fastmult vfp edsp neon vfpv3 tls vfpv4 idiva idivt vfpd32 lpae evtstrm +CPU implementer : 0x41 +CPU architecture: 7 +CPU variant : 0x0 +CPU part : 0xc07 +CPU revision : 5 + +processor : 2 +model name : ARMv7 Processor rev 5 (v7l) +BogoMIPS : 38.40 +Features : half thumb fastmult vfp edsp neon vfpv3 tls vfpv4 idiva idivt vfpd32 lpae evtstrm +CPU implementer : 0x41 +CPU architecture: 7 +CPU variant : 0x0 +CPU part : 0xc07 +CPU revision : 5 + +processor : 3 +model name : ARMv7 Processor rev 5 (v7l) +BogoMIPS : 38.40 +Features : half thumb fastmult vfp edsp neon vfpv3 tls vfpv4 idiva idivt vfpd32 lpae evtstrm +CPU implementer : 0x41 +CPU architecture: 7 +CPU variant : 0x0 +CPU part : 0xc07 +CPU revision : 5 + +Hardware : BCM2709 +Revision : a01041 +Serial : 00000000c9e9323d diff --git a/tests/sysinfodata/arm-rpi2sysinfo.expect b/tests/sysinfodata/arm-rpi2sysinfo.expect new file mode 100644 index 0000000..acb3ad9 --- /dev/null +++ b/tests/sysinfodata/arm-rpi2sysinfo.expect @@ -0,0 +1,18 @@ +<sysinfo type='smbios'> + <processor> + <entry name='socket_destination'>0</entry> + <entry name='type'>ARMv7 Processor rev 5 (v7l)</entry> + </processor> + <processor> + <entry name='socket_destination'>1</entry> + <entry name='type'>ARMv7 Processor rev 5 (v7l)</entry> + </processor> + <processor> + <entry name='socket_destination'>2</entry> + <entry name='type'>ARMv7 Processor rev 5 (v7l)</entry> + </processor> + <processor> + <entry name='socket_destination'>3</entry> + <entry name='type'>ARMv7 Processor rev 5 (v7l)</entry> + </processor> +</sysinfo> diff --git a/tests/sysinfotest.c b/tests/sysinfotest.c index d8266a7..74e5f71 100644 --- a/tests/sysinfotest.c +++ b/tests/sysinfotest.c @@ -166,11 +166,23 @@ VIRT_TEST_MAIN(test_x86) static int test_arm(void) { - return sysinfotest_run("arm sysinfo", - NULL, - NULL, - "/sysinfodata/armcpuinfo.data", - "/sysinfodata/armsysinfo.expect"); + int ret = EXIT_SUCCESS; + + if (sysinfotest_run("arm sysinfo", + NULL, + NULL, + "/sysinfodata/armcpuinfo.data", + "/sysinfodata/armsysinfo.expect") != EXIT_SUCCESS) + ret = EXIT_FAILURE; + + if (sysinfotest_run("Raspberry Pi 2 sysinfo", + NULL, + NULL, + "/sysinfodata/arm-rpi2cpuinfo.data", + "/sysinfodata/arm-rpi2sysinfo.expect") != EXIT_SUCCESS) + ret = EXIT_FAILURE; + + return ret; } VIRT_TEST_MAIN(test_arm) -- 2.3.6

3 2

[libvirt] [PATCH] util: Avoid shadow of 'ulong' in virMemoryMaxValue
by Peter Krempa 21 May '15

21 May '15

Old compilers whine: src/util/virutil.c: In function 'virMemoryMaxValue': src/util/virutil.c:2612: error: declaration of 'ulong' shadows a global declaration [-Wshadow] /usr/include/sys/types.h:151: error: shadowed declaration is here [-Wshadow] s/ulong/capped/ to work around the problem ---1 Pushed under the trivial && build-breaker rules. src/util/virutil.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/src/util/virutil.c b/src/util/virutil.c index 638d6e2..e479cce 100644 --- a/src/util/virutil.c +++ b/src/util/virutil.c @@ -2603,17 +2603,17 @@ virMemoryLimitIsSet(unsigned long long value) /** * virMemoryMaxValue * - * @ulong: whether the value must fit into unsigned long + * @capped: whether the value must fit into unsigned long * (long long is assumed otherwise) * * Returns the maximum possible memory value in bytes. */ unsigned long long -virMemoryMaxValue(bool ulong) +virMemoryMaxValue(bool capped) { /* On 32-bit machines, our bound is 0xffffffff * KiB. On 64-bit * machines, our bound is off_t (2^63). */ - if (ulong && sizeof(unsigned long) < sizeof(long long)) + if (capped && sizeof(unsigned long) < sizeof(long long)) return 1024ull * ULONG_MAX; else return LLONG_MAX; -- 2.3.5

1 0

[libvirt] [PATCH 0/3] Support changing user passwords in the guest
by Ján Tomko 21 May '15

21 May '15

The guest-set-user-password command was added to qemu-agent in the 2.3.0 release by Daniel Berrange (some of you might have heard of him): commit 215a2771a7b6b29037ee8deba484815d816b6fdd qga: add guest-set-user-password command http://git.qemu.org/?p=qemu.git;a=commitdiff;h=215a277 This mini-series introduces a libvirt API, set_password ACL, a virsh command and wires up the API in the QEMU driver. Any bikesh^Wsuggestions regarding the API naming and the order of the virsh parameters are welcome. Ján Tomko (3): Introduce virDomainSetUserPassword API virsh: add set-user-password command qemu: wire up virDomainSetUserPassword include/libvirt/libvirt-domain.h | 9 +++++ src/access/viraccessperm.c | 2 +- src/access/viraccessperm.h | 6 ++++ src/driver-hypervisor.h | 7 ++++ src/libvirt-domain.c | 47 +++++++++++++++++++++++++ src/libvirt_public.syms | 5 +++ src/qemu/qemu_agent.c | 39 +++++++++++++++++++++ src/qemu/qemu_agent.h | 4 +++ src/qemu/qemu_driver.c | 55 +++++++++++++++++++++++++++++ src/remote/remote_driver.c | 1 + src/remote/remote_protocol.x | 15 +++++++- src/remote_protocol-structs | 7 ++++ tools/virsh-domain.c | 76 ++++++++++++++++++++++++++++++++++++++++ tools/virsh.pod | 11 ++++++ 14 files changed, 282 insertions(+), 2 deletions(-) -- 2.0.5

3 9

[libvirt] [PATCH 0/3] parallels: fix prlsdkLoadDomain and prlsdkAddDomain
by Maxim Nestratov 21 May '15

21 May '15

Maxim Nestratov (3): parallels: fix prlsdkAddDomain parallels: move up updating autostart parameter in prlsdkLoadDomain parallels: fix possible crash in case of errors in prlsdkLoadDomain src/parallels/parallels_sdk.c | 39 ++++++++++++++++++++++++++------------- 1 file changed, 26 insertions(+), 13 deletions(-) -- 2.1.0

5 15

[libvirt] [PATCH 0/2] Add and use for virDomainDiskIndexBy*
by Jiri Denemark 21 May '15

21 May '15

Sometimes the only thing we need is the pointer to virDomainDiskDef and having to call virDomainDiskIndexBy* APIs, storing the disk index, and looking it up in the disks array is ugly. After this patch, we can just call virDomainDiskBy* and get the pointer in one step. Jiri Denemark (2): Add wrappers for virDomainDiskIndexBy* Use virDomainDiskByName where appropriate src/conf/domain_conf.c | 20 +++++++++++++++ src/conf/domain_conf.h | 8 ++++++ src/libvirt_private.syms | 2 ++ src/libxl/libxl_driver.c | 4 +-- src/lxc/lxc_driver.c | 10 +++----- src/qemu/qemu_agent.c | 9 ++++--- src/qemu/qemu_blockjob.c | 5 ++-- src/qemu/qemu_driver.c | 62 ++++++++++++++++++----------------------------- src/qemu/qemu_migration.c | 6 ++--- src/qemu/qemu_process.c | 23 ++++-------------- 10 files changed, 72 insertions(+), 77 deletions(-) -- 2.4.1

4 9

[libvirt] [PATCH] threadpool: Switch to detached threads
by Jiri Denemark 21 May '15

21 May '15

Using joinable threads does not help anything, but it can lead to memory leaks. When a worker thread exits, it decreases nWorkers or nPrioWorkers and once both nWorkers and nPrioWorkers are zero (i.e., the last worker is gone), quit_cond is signaled. When freeing the pool we first tell all threads to die and then we are waiting for both nWorkers and nPrioWorkers to become zero. At this point we already know all threads are gone. So the only reason for calling virThreadJoin of all workers is to free the memory allocated for joinable threads. If we avoid allocating this memory, we don't need to take care of freeing it. Moreover, any memory associated with a worker thread which died before we asked it to die (e.g., because virCondWait failed in the thread) would be lost anyway since virThreadPoolFree calls virThreadJoin only for threads which were running at the time virThreadPoolFree was called. Signed-off-by: Jiri Denemark <jdenemar(a)redhat.com> --- src/util/virthreadpool.c | 17 +++-------------- 1 file changed, 3 insertions(+), 14 deletions(-) diff --git a/src/util/virthreadpool.c b/src/util/virthreadpool.c index ed22486..f640448 100644 --- a/src/util/virthreadpool.c +++ b/src/util/virthreadpool.c @@ -201,7 +201,7 @@ virThreadPoolNewFull(size_t minWorkers, data->cond = &pool->cond; if (virThreadCreateFull(&pool->workers[i], - true, + false, virThreadPoolWorker, pool->jobFuncName, true, @@ -225,7 +225,7 @@ virThreadPoolNewFull(size_t minWorkers, data->priority = true; if (virThreadCreateFull(&pool->prioWorkers[i], - true, + false, virThreadPoolWorker, pool->jobFuncName, true, @@ -249,16 +249,11 @@ void virThreadPoolFree(virThreadPoolPtr pool) { virThreadPoolJobPtr job; bool priority = false; - size_t i; - size_t nWorkers; - size_t nPrioWorkers; if (!pool) return; virMutexLock(&pool->mutex); - nWorkers = pool->nWorkers; - nPrioWorkers = pool->nPrioWorkers; pool->quit = true; if (pool->nWorkers > 0) virCondBroadcast(&pool->cond); @@ -275,12 +270,6 @@ void virThreadPoolFree(virThreadPoolPtr pool) VIR_FREE(job); } - for (i = 0; i < nWorkers; i++) - virThreadJoin(&pool->workers[i]); - - for (i = 0; i < nPrioWorkers; i++) - virThreadJoin(&pool->prioWorkers[i]); - VIR_FREE(pool->workers); virMutexUnlock(&pool->mutex); virMutexDestroy(&pool->mutex); @@ -338,7 +327,7 @@ int virThreadPoolSendJob(virThreadPoolPtr pool, data->cond = &pool->cond; if (virThreadCreateFull(&pool->workers[pool->nWorkers - 1], - true, + false, virThreadPoolWorker, pool->jobFuncName, true, -- 2.4.1

2 4

[libvirt] [PATCH] qemu: Log error if domain uses security driver which is not loaded
by Erik Skultety 21 May '15

21 May '15

When starting a domain, if we find out domain requests security drivers we do not have loaded, we fail. However we don't check for this during reconnect, so any operation relying on security driver functionality would fail. If someone e.g. starts a domain with selinux driver loaded, then they turn off the security driver in config, restart the daemon and call dump/save/.., QEMU returns an error. As we shouldn't kill the domain, we should at least log an error to let the user know that domain reconnect wasn't completely clean. https://bugzilla.redhat.com/show_bug.cgi?id=1183893 --- src/qemu/qemu_process.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index 56719eb..8da79e5 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -3747,6 +3747,12 @@ qemuProcessReconnect(void *opaque) if ((qemuDomainAssignAddresses(obj->def, priv->qemuCaps, obj)) < 0) goto error; + /* if domain requests security driver we haven't loaded, report error, but + * do not kill the domain + */ + ignore_value(virSecurityManagerCheckAllLabel(driver->securityManager, + obj->def)); + if (virSecurityManagerReserveLabel(driver->securityManager, obj->def, obj->pid) < 0) goto error; -- 1.9.3

3 2

[libvirt] [PATCH 0/3] virsh: Fix and enhance handling of numeric options
by Andrea Bolognani 21 May '15

21 May '15

The first commit improves error messages by making them much more consistent and a little bit more helpful to the user. The other two commits contain smaller fixes and enhancement. Andrea Bolognani (3): virsh: Improve error message on integer value parsing failure. virsh: Fix dommemstat --period option type. virsh: Improve handling of send-process-signal --pid. tests/vcpupin | 4 +- tools/virsh-domain-monitor.c | 12 +++-- tools/virsh-domain.c | 114 ++++++++++++++++++++++++++++++------------- tools/virsh-host.c | 44 ++++++++++++----- tools/virsh-interface.c | 4 +- tools/virsh-network.c | 4 +- tools/virsh-volume.c | 16 ++++-- 7 files changed, 141 insertions(+), 57 deletions(-) -- 2.1.0

2 6

[libvirt] [PATCH 0/2] Fix virsh nodeset on shutoff domains
by Michal Privoznik 20 May '15

20 May '15

This goes on the top of Martin's (ACKed) patch: https://www.redhat.com/archives/libvir-list/2015-May/msg00585.html Michal Privoznik (2): virDomainNumatuneGetMode: Report if numatune was defined qemuDomainGetNumaParameters: Don't report spurious info src/conf/numa_conf.c | 38 ++++++++++++++++++++++-------- src/conf/numa_conf.h | 5 ++-- src/lxc/lxc_cgroup.c | 5 ++-- src/lxc/lxc_controller.c | 30 ++++++++++++------------ src/parallels/parallels_sdk.c | 5 ++-- src/qemu/qemu_cgroup.c | 15 +++++++----- src/qemu/qemu_command.c | 4 ++-- src/qemu/qemu_driver.c | 54 ++++++++++++++++++++++++++++--------------- src/qemu/qemu_process.c | 28 +++++++++++----------- 9 files changed, 113 insertions(+), 71 deletions(-) -- 2.3.6

2 6

[libvirt] [libvirt-python 0/3] reproducible builds
by Guido Günther 20 May '15

20 May '15

Debian tries to make builds reporducible[0] therefore we need predictable output (i.e. no timestamps, predictable iteration over hashes, etc.). This patch (mostly by Chris Lamb) to libvirt-python makes the necessary changes. [0] https://wiki.debian.org/ReproducibleBuilds Guido Günther (3): Sort dictionary keys Simplify sorting Sort tuples on both items generator.py | 19 ++++++++----------- 1 file changed, 8 insertions(+), 11 deletions(-) -- 2.1.4

2 7

[libvirt] [PATCH v3 0/4] Introduce pci-serial
by Michal Privoznik 20 May '15

20 May '15

diff to v2: - Allocate address on domain define diff to v1: - Fix hot(un-)plug Michal Privoznik (4): Introduce pci-serial qemu: Implement pci-serial qemuDomainAttachChrDevice: Fix chardev hotplug qemuDomainDetachChrDevice: Fix chardev hot-unplug docs/formatdomain.html.in | 16 ++++++---- docs/schemas/domaincommon.rng | 1 + src/conf/domain_conf.c | 3 +- src/conf/domain_conf.h | 1 + src/qemu/qemu_capabilities.c | 2 ++ src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_command.c | 30 +++++++++++++++++- src/qemu/qemu_hotplug.c | 36 ++++++++++++++++----- tests/qemucapabilitiesdata/caps_1.3.1-1.caps | 1 + tests/qemucapabilitiesdata/caps_1.4.2-1.caps | 1 + tests/qemucapabilitiesdata/caps_1.5.3-1.caps | 1 + tests/qemucapabilitiesdata/caps_1.6.0-1.caps | 1 + tests/qemucapabilitiesdata/caps_1.6.50-1.caps | 1 + tests/qemucapabilitiesdata/caps_2.1.1-1.caps | 1 + .../qemuxml2argv-pci-serial-dev-chardev.args | 7 ++++ .../qemuxml2argv-pci-serial-dev-chardev.xml | 37 ++++++++++++++++++++++ tests/qemuxml2argvtest.c | 3 ++ tests/qemuxml2xmltest.c | 1 + 18 files changed, 128 insertions(+), 16 deletions(-) create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-pci-serial-dev-chardev.args create mode 100644 tests/qemuxml2argvdata/qemuxml2argv-pci-serial-dev-chardev.xml -- 2.3.6

2 8

[libvirt] [PATCH] conf: Catch memory size overflow earlier
by Peter Krempa 20 May '15

20 May '15

virDomainParseMemory parses the size and then rounds up while converting it to kibibytes. Since the number is limit-checked before the rounding it's possible to use a number that would be correctly parsed the first time, but not the second time. For numbers not limited to 32 bit systems the magic is 9223372036854775807 bytes. That number then can't be parsed back in kibibytes. To solve the issue add a second overflow check for the few values that would cause the problem. Since virDomainParseMemory is used in config parsing, this avoids vanishing VMs. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1221504 --- src/conf/domain_conf.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index bfdc94e..ba4f430 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -7237,6 +7237,13 @@ virDomainParseMemory(const char *xpath, /* Yes, we really do use kibibytes for our internal sizing. */ *mem = VIR_DIV_UP(bytes, 1024); + + if (*mem >= VIR_DIV_UP(max, 1024)) { + virReportError(VIR_ERR_OVERFLOW, "%s", _("size value too large")); + ret = -1; + goto cleanup; + } + ret = 0; cleanup: return ret; -- 2.3.5

2 2

[libvirt] [PATCH] Taint domains using cdrom-passthrough
by John Ferlan 20 May '15

20 May '15

https://bugzilla.redhat.com/show_bug.cgi?id=976387 For a domain configured using the host cdrom, we should taint the domain due to problems encountered when the host and guest try to control the tray. Signed-off-by: John Ferlan <jferlan(a)redhat.com> --- src/conf/domain_conf.c | 3 ++- src/conf/domain_conf.h | 1 + src/qemu/qemu_domain.c | 6 ++++++ 3 files changed, 9 insertions(+), 1 deletion(-) diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c index add857c..a67e200 100644 --- a/src/conf/domain_conf.c +++ b/src/conf/domain_conf.c @@ -101,7 +101,8 @@ VIR_ENUM_IMPL(virDomainTaint, VIR_DOMAIN_TAINT_LAST, "disk-probing", "external-launch", "host-cpu", - "hook-script"); + "hook-script", + "cdrom-passthrough"); VIR_ENUM_IMPL(virDomainVirt, VIR_DOMAIN_VIRT_LAST, "qemu", diff --git a/src/conf/domain_conf.h b/src/conf/domain_conf.h index 2cd105a7..0867e8b 100644 --- a/src/conf/domain_conf.h +++ b/src/conf/domain_conf.h @@ -2280,6 +2280,7 @@ typedef enum { VIR_DOMAIN_TAINT_EXTERNAL_LAUNCH, /* Externally launched guest domain */ VIR_DOMAIN_TAINT_HOST_CPU, /* Host CPU passthrough in use */ VIR_DOMAIN_TAINT_HOOK, /* Domain (possibly) changed via hook script */ + VIR_DOMAIN_TAINT_CDROM_PASSTHROUGH,/* CDROM passthrough */ VIR_DOMAIN_TAINT_LAST } virDomainTaintFlags; diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c index fa8229f..b66ee89 100644 --- a/src/qemu/qemu_domain.c +++ b/src/qemu/qemu_domain.c @@ -2031,6 +2031,12 @@ void qemuDomainObjCheckDiskTaint(virQEMUDriverPtr driver, qemuDomainObjTaint(driver, obj, VIR_DOMAIN_TAINT_HIGH_PRIVILEGES, logFD); + if (disk->device == VIR_DOMAIN_DISK_DEVICE_CDROM && + virStorageSourceGetActualType(disk->src) == VIR_STORAGE_TYPE_BLOCK && + disk->src->path) + qemuDomainObjTaint(driver, obj, VIR_DOMAIN_TAINT_CDROM_PASSTHROUGH, + logFD); + virObjectUnref(cfg); } -- 2.1.0

2 4

[libvirt] [PATCH 0/3] Introduce yet another type to SMBIOS
by Michal Privoznik 20 May '15

20 May '15

The specification is here: http://www.dmtf.org/sites/default/files/standards/documents/DSP0134_3.0.0.p… The first two patches rework the code a bit, and the last one adds something useful. Michal Privoznik (3): virSysinfoDef: Exempt BIOS variables virSysinfoDef: Exempt SYSTEM variables virSysinfo: Introduce SMBIOS type 2 support docs/formatdomain.html.in | 37 +- docs/schemas/domaincommon.rng | 24 ++ src/conf/domain_conf.c | 252 ++++++++--- src/libvirt_private.syms | 3 + src/qemu/qemu_command.c | 112 +++-- src/util/virsysinfo.c | 528 +++++++++++++++++++----- src/util/virsysinfo.h | 51 ++- tests/qemuxml2argvdata/qemuxml2argv-smbios.args | 2 + tests/qemuxml2argvdata/qemuxml2argv-smbios.xml | 9 + 9 files changed, 804 insertions(+), 214 deletions(-) -- 2.3.6

2 7

[libvirt] [python PATCHv2] examples: Add example to make guest agent lifecycle event useful
by Peter Krempa 19 May '15

19 May '15

This example allows to use the guest agent event and metadata to track vCPU count set via the guest agent (agent-based onlining/offlining) and keep it persistent accross domain restarts. The daemon listens for the agent lifecycle event, and if it's received it looks into doman's metadata to see whether a desired count was set and issues the guest agent command. --- MANIFEST.in | 2 + examples/README | 2 + examples/guest-vcpus/guest-vcpu-daemon.py | 160 ++++++++++++++++++++++++++++++ examples/guest-vcpus/guest-vcpu.py | 74 ++++++++++++++ 4 files changed, 238 insertions(+) create mode 100755 examples/guest-vcpus/guest-vcpu-daemon.py create mode 100755 examples/guest-vcpus/guest-vcpu.py diff --git a/MANIFEST.in b/MANIFEST.in index ad08207..230baea 100644 --- a/MANIFEST.in +++ b/MANIFEST.in @@ -11,6 +11,8 @@ include examples/domsave.py include examples/domstart.py include examples/esxlist.py include examples/event-test.py +include examples/guest-vcpus/guest-vcpu-daemon.py +include examples/guest-vcpus/guest-vcpu.py include examples/topology.py include generator.py include libvirt-lxc-override-api.xml diff --git a/examples/README b/examples/README index 1d4b425..0cb4513 100644 --- a/examples/README +++ b/examples/README @@ -12,6 +12,8 @@ esxlist.py - list active domains of an VMware ESX host and print some info. also demonstrates how to use the libvirt.openAuth() method dhcpleases.py - list dhcp leases for a given virtual network domipaddrs.py - list IP addresses for guest domains +guest-vcpus - two helpers to make the guest agent event useful with agent based + vCPU state modification The XML files in this directory are examples of the XML format that libvirt expects, and will have to be adapted for your setup. They are only needed diff --git a/examples/guest-vcpus/guest-vcpu-daemon.py b/examples/guest-vcpus/guest-vcpu-daemon.py new file mode 100755 index 0000000..c7c08a8 --- /dev/null +++ b/examples/guest-vcpus/guest-vcpu-daemon.py @@ -0,0 +1,160 @@ +#!/usr/bin/env python + +import libvirt +import threading +from xml.dom import minidom +import time +import sys +import getopt +import os + +uri = "qemu:///system" +customXMLuri = "guest-cpu.python.libvirt.org" +connectRetryTimeout = 5 + +def usage(): + print("usage: "+os.path.basename(sys.argv[0])+" [-h] [uri]") + print(" uri will default to qemu:///system") + print(" --help, -h Print(this help message") + print("") + print("This service waits for the guest agent lifecycle event and reissues " + + "guest agent calls to modify the cpu count according to the metadata " + + "set by guest-vcpu.py example") + +class workerData: + def __init__(self): + self.doms = list() + self.conn = None + self.cond = threading.Condition() + + def notify(self): + self.cond.acquire() + self.cond.notify() + self.cond.release() + + def waitNotify(self): + self.cond.acquire() + self.cond.wait() + self.cond.release() + + def addDomainNotify(self, dom): + self.doms.append(dom) + self.notify() + + def closeConnectNotify(self): + conn = self.conn + self.conn = None + conn.close() + self.notify() + + def setConnect(self, conn): + self.conn = conn + + def hasConn(self): + return self.conn is not None + + def hasDom(self): + return len(self.doms) > 0 + + def getDom(self): + return self.doms.pop() + + def setDoms(self, doms): + self.doms = doms + + +def virEventLoopNativeRun(): + while True: + libvirt.virEventRunDefaultImpl() + +def handleAgentLifecycleEvent(conn, dom, state, reason, opaque): + if state == libvirt.VIR_CONNECT_DOMAIN_EVENT_AGENT_LIFECYCLE_STATE_CONNECTED: + opaque.addDomainNotify(dom) + +def handleConnectClose(conn, reason, opaque): + print('Disconnected from ' + uri) + opaque.closeConnectNotify() + +def handleLibvirtLibraryError(opaque, error): + pass + +def processAgentConnect(dom): + try: + cpus = dom.metadata(libvirt.VIR_DOMAIN_METADATA_ELEMENT, customXMLuri, + libvirt.VIR_DOMAIN_AFFECT_LIVE) + doc = minidom.parseString(cpus) + ncpus = int(doc.getElementsByTagName('ncpus')[0].getAttribute('count')) + except: + return + + try: + dom.setVcpusFlags(ncpus, libvirt.VIR_DOMAIN_AFFECT_LIVE | libvirt.VIR_DOMAIN_VCPU_GUEST) + print("set vcpu count for domain " + dom.name() + " to " + str(ncpus)) + except: + print("failed to set vcpu count for domain " + dom.name()) + +def work(): + data = workerData() + + print("Using uri: " + uri) + + while True: + if not data.hasConn(): + try: + conn = libvirt.open(uri) + except: + print('Failed to connect to ' + uri + ', retry in ' + str(connectRetryTimeout)) + ' seconds' + time.sleep(connectRetryTimeout) + continue + + print('Connected to ' + uri) + + data.setConnect(conn) + conn.registerCloseCallback(handleConnectClose, data) + conn.setKeepAlive(5, 3) + conn.domainEventRegisterAny(None, + libvirt.VIR_DOMAIN_EVENT_ID_AGENT_LIFECYCLE, + handleAgentLifecycleEvent, + data) + + data.setDoms(conn.listAllDomains(libvirt.VIR_CONNECT_LIST_DOMAINS_ACTIVE)) + + while data.hasConn() and data.hasDom(): + processAgentConnect(data.getDom()) + + data.waitNotify() + +def main(): + libvirt.virEventRegisterDefaultImpl() + libvirt.registerErrorHandler(handleLibvirtLibraryError, None) + + worker = threading.Thread(target=work) + worker.setDaemon(True) + worker.start() + + eventLoop = threading.Thread(target=virEventLoopNativeRun) + eventLoop.setDaemon(True) + eventLoop.start() + + while True: + time.sleep(1) + +if __name__ == "__main__": + try: + opts, args = getopt.getopt(sys.argv[1:], "h", ["help"]) + except getopt.GetoptError as err: + print(str(err)) + usage() + sys.exit(2) + for o, a in opts: + if o in ("-h", "--help"): + usage() + sys.exit() + + if len(args) > 1: + usage() + sys.exit(1) + elif len(args) == 1: + uri = args[0] + + main() diff --git a/examples/guest-vcpus/guest-vcpu.py b/examples/guest-vcpus/guest-vcpu.py new file mode 100755 index 0000000..8faba87 --- /dev/null +++ b/examples/guest-vcpus/guest-vcpu.py @@ -0,0 +1,74 @@ +#!/usr/bin/env python + +import libvirt +import sys +import getopt +import os + +customXMLuri = "guest-cpu.python.libvirt.org" + +def usage(): + print("usage: "+os.path.basename(sys.argv[0])+" [-hcl] domain count [uri]") + print(" uri will default to qemu:///system") + print(" --help, -h Print(this help message") + print(" --config, -c Modify persistent domain configuration") + print(" --live, -l Modify live domain configuration") + print("") + print("Sets the vCPU count via the guest agent and sets the metadata element " + + "used by guest-vcpu-daemon.py example") + +uri = "qemu:///system" +flags = 0 +live = False; +config = False; + +try: + opts, args = getopt.getopt(sys.argv[1:], "hcl", ["help", "config", "live"]) +except getopt.GetoptError as err: + # print help information and exit: + print(str(err)) # will print something like "option -a not recognized" + usage() + sys.exit(2) +for o, a in opts: + if o in ("-h", "--help"): + usage() + sys.exit() + if o in ("-c", "--config"): + config = True + flags |= libvirt.VIR_DOMAIN_AFFECT_CONFIG + if o in ("-l", "--live"): + live = True + flags |= libvirt.VIR_DOMAIN_AFFECT_LIVE + +if len(args) < 2: + usage() + sys.exit(1) +elif len(args) >= 3: + uri = args[2] + +domain = args[0] +count = int(args[1]) + +conn = libvirt.open(uri) +dom = conn.lookupByName(domain) + +if flags == 0 or config: + confvcpus = dom.vcpusFlags(libvirt.VIR_DOMAIN_AFFECT_CONFIG) + + if confvcpus < count: + print("Persistent domain configuration has only " + str(confvcpus) + " vcpus configured") + sys.exit(1) + +if flags == 0 or live: + livevcpus = dom.vcpusFlags(libvirt.VIR_DOMAIN_AFFECT_LIVE) + + if livevcpus < count: + print("Live domain configuration has only " + str(livevcpus) + " vcpus configured") + sys.exit(1) + +if flags == 0 or live: + dom.setVcpusFlags(count, libvirt.VIR_DOMAIN_AFFECT_LIVE | libvirt.VIR_DOMAIN_VCPU_GUEST) + +meta = "<ncpus count='" + str(count) + "'/>" + +dom.setMetadata(libvirt.VIR_DOMAIN_METADATA_ELEMENT, meta, "guestvcpudaemon", customXMLuri, flags) -- 2.3.5

2 2

[libvirt] [PATCHv2] docs: add dimm address document in docs
by Luyao Huang 19 May '15

19 May '15

As we have a new device address type 'dimm', add some document. As we do not want user touch this address, so no need add an XML example in the doc. Signed-off-by: Luyao Huang <lhuang(a)redhat.com> --- v2: remove the xml example and values range, and point out no need specified this address by users. docs/formatdomain.html.in | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/docs/formatdomain.html.in b/docs/formatdomain.html.in index e0b6ba7..4adf827 100644 --- a/docs/formatdomain.html.in +++ b/docs/formatdomain.html.in @@ -2876,6 +2876,14 @@ attributes: <code>iobase</code> and <code>irq</code>. <span class="since">Since 1.2.1</span> </dd> + <dt><code>type='dimm'</code></dt> + <dd>DIMM addresses, for memory device, have the following additional + attributes: <code>slot</code> and <code>base</code>. No need specified + address when use memory device usually, these 2 attributes' values are + determined automatically when start the vm or hot-plug the memory + device. + <span class="since">Since 1.2.14</span> + </dd> </dl> <h4><a name="elementsControllers">Controllers</a></h4> -- 1.8.3.1

1 0

[libvirt] [PATCH] qemu: Fix numatune nodeset reporting
by Martin Kletzander 19 May '15

19 May '15

Since af2a1f0587d88656f2c14265a63fbc11ecbd924e, qemuDomainGetNumaParameters() returns invalid value for a running guest. The problem is that it is getting the information from cgroups, but the parent cgroup is being left alone since the mentioned commit. Since the running guest's XML is in sync with cgroups, there is no need to look into cgroups (unless someone changes the configuration behind libvirt's back). Returning the info from the definition fixes a bug and is also a cleanup. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1221047 Signed-off-by: Martin Kletzander <mkletzan(a)redhat.com> --- src/qemu/qemu_driver.c | 35 ++++++++++------------------------- 1 file changed, 10 insertions(+), 25 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 8c00cdc..0cc0a29 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -10523,7 +10523,7 @@ qemuDomainGetNumaParameters(virDomainPtr dom, char *nodeset = NULL; int ret = -1; virCapsPtr caps = NULL; - qemuDomainObjPrivatePtr priv; + virDomainDefPtr def = NULL; virCheckFlags(VIR_DOMAIN_AFFECT_LIVE | VIR_DOMAIN_AFFECT_CONFIG | @@ -10537,8 +10537,6 @@ qemuDomainGetNumaParameters(virDomainPtr dom, if (!(vm = qemuDomObjFromDomain(dom))) return -1; - priv = vm->privateData; - if (virDomainGetNumaParametersEnsureACL(dom->conn, vm->def) < 0) goto cleanup; @@ -10555,6 +10553,11 @@ qemuDomainGetNumaParameters(virDomainPtr dom, goto cleanup; } + if (flags & VIR_DOMAIN_AFFECT_CONFIG) + def = persistentDef; + else + def = vm->def; + for (i = 0; i < QEMU_NB_NUMA_PARAM && i < *nparams; i++) { virMemoryParameterPtr param = &params[i]; @@ -10564,35 +10567,17 @@ qemuDomainGetNumaParameters(virDomainPtr dom, VIR_TYPED_PARAM_INT, 0) < 0) goto cleanup; - if (flags & VIR_DOMAIN_AFFECT_CONFIG) - param->value.i = virDomainNumatuneGetMode(persistentDef->numa, -1); - else - param->value.i = virDomainNumatuneGetMode(vm->def->numa, -1); + param->value.i = virDomainNumatuneGetMode(def->numa, -1); break; case 1: /* fill numa nodeset here */ - if (flags & VIR_DOMAIN_AFFECT_CONFIG) { - nodeset = virDomainNumatuneFormatNodeset(persistentDef->numa, - NULL, -1); - if (!nodeset) - goto cleanup; - } else { - if (!virCgroupHasController(priv->cgroup, - VIR_CGROUP_CONTROLLER_CPUSET) || - virCgroupGetCpusetMems(priv->cgroup, &nodeset) < 0) { - nodeset = virDomainNumatuneFormatNodeset(vm->def->numa, - NULL, -1); - if (!nodeset) - goto cleanup; - } - } - - if (virTypedParameterAssign(param, VIR_DOMAIN_NUMA_NODESET, + nodeset = virDomainNumatuneFormatNodeset(def->numa, NULL, -1); + if (!nodeset || + virTypedParameterAssign(param, VIR_DOMAIN_NUMA_NODESET, VIR_TYPED_PARAM_STRING, nodeset) < 0) goto cleanup; nodeset = NULL; - break; /* coverity[dead_error_begin] */ -- 2.4.0

2 1