- Devel - libvirt List Archives

[libvirt] Looking for project ideas and mentors for Google Summer of Code 2014
by Stefan Hajnoczi 07 Mar '14

07 Mar '14

KVM & libvirt: you are welcome to join the QEMU umbrella organization like last year. What is GSoC? Google Summer of Code 2014 (GSoC) provides funding for students to work on open source projects for 12-weeks over the summer. Open source organizations apply to participate and those accepted receive funding for one or more students. How we are participating QEMU has participated, together with KVM and libvirt, in past years and I would like to continue that tradition this year. We now need to collect a list of project ideas on our wiki. We also need mentors to volunteer. http://qemu-project.org/Google_Summer_of_Code_2014 Project ideas Please post project ideas on the wiki page below. Project ideas should be suitable as a 12-week project that a student fluent in C/Python/etc can complete. No prior knowledge of QEMU/KVM/libvirt internals can be assumed. http://qemu-project.org/Google_Summer_of_Code_2014 Mentors Please add your name to project ideas you are willing to mentor. In order to mentor you must be an established contributor (regularly contribute patches). You must be willing to spend about 5 hours per week from May 19 to August 18. I have CCed the 8 most active committers since QEMU 1.5.0 but everyone is invited. Official timeline: https://www.google-melange.com/gsoc/events/google/gsoc2014 Stefan

8 16

[libvirt] [PATCH 1/9] conf: eliminate hardcoded indent from all xml
by Laine Stump 07 Mar '14

07 Mar '14

Many of the domain xml format functions (including all of the device format functions) had hard-coded spaces, which made for incorrect indentation when those functions were called in a different context (for example, commit 2122cf39 added <interface> XML into the document provided to a network hook script, and in this case it should have been indented by 2 spaces, but was instead indented by 6 spaces). In that patch I mentioned doing a followup patch to make the device xml formatters more consistent. After doing that patch, it felt incomplete to not give the same treatment to the entire directory. The one downside to this series is that it may create merge conflicts during backports, but fortunately the conflicts should all be fairly easy to resolve. Laine Stump (9): conf: eliminate hardcoded indent from domain xml conf: eliminate hardcoded indent from domain snapshot xml conf: eliminate hardcoded indent from network xml conf: eliminate outmoded/odd indent method from interface xml conf: eliminate hardcoded indentation in nwfilter xml conf: eliminate hardcoded indentation in capabilities xml conf: eliminate hardcoded indentation in node device xml conf: eliminate hardcoded indent in volume/pool xml conf: eliminate hardcoded indentation in all remaining xml src/conf/capabilities.c | 183 ++++++----- src/conf/cpu_conf.c | 11 +- src/conf/domain_conf.c | 599 +++++++++++++++++++---------------- src/conf/interface_conf.c | 137 ++++---- src/conf/netdev_bandwidth_conf.c | 6 +- src/conf/netdev_vlan_conf.c | 6 +- src/conf/netdev_vport_profile_conf.c | 6 +- src/conf/network_conf.c | 8 +- src/conf/node_device_conf.c | 207 ++++++------ src/conf/nwfilter_conf.c | 94 ++---- src/conf/nwfilter_params.c | 6 +- src/conf/secret_conf.c | 20 +- src/conf/snapshot_conf.c | 48 +-- src/conf/storage_conf.c | 174 +++++----- src/conf/storage_encryption_conf.c | 6 +- 15 files changed, 811 insertions(+), 700 deletions(-) -- 1.8.5.3

2 22

[libvirt] Python-libvirt in a virtual environment.
by Sijo Jose 07 Mar '14

07 Mar '14

Hi, Could you guide me to install libvirt in a virtual environment..? OR How python-libvirt could be used within a virtual environment after installing it, in the system. Regards Sijo

3 4

[libvirt] [PATCH] spec: Let translations be properly updated
by Jiri Denemark 06 Mar '14

06 Mar '14

Libvirt tarball contains po/stamp-po file which prevents any po/*.gmo file to be regenerated even if a corresponding po/*.po file is newer. By removing the stamp-po file, all *.gmo files are properly updated if required. This allows downstreams to provide patches that update translations. Signed-off-by: Jiri Denemark <jdenemar(a)redhat.com> --- libvirt.spec.in | 1 + 1 file changed, 1 insertion(+) diff --git a/libvirt.spec.in b/libvirt.spec.in index 014fe5d..23374e7 100644 --- a/libvirt.spec.in +++ b/libvirt.spec.in @@ -1421,6 +1421,7 @@ driver autoreconf -if %endif +rm -f po/stamp-po %configure %{?_without_xen} \ %{?_without_qemu} \ %{?_without_openvz} \ -- 1.9.0

2 4

[libvirt] [PATCH v2 1/5] Introduce virDomainFSFreeze() and virDomainFSThaw() public API
by Tomoki Sekiyama 06 Mar '14

06 Mar '14

These will freeze and thaw filesystems within guest. The APIs take flags arguments which are currently not used, for future extensions. Signed-off-by: Tomoki Sekiyama <tomoki.sekiyama(a)hds.com> --- include/libvirt/libvirt.h.in | 6 ++++ src/driver.h | 10 ++++++ src/libvirt.c | 70 ++++++++++++++++++++++++++++++++++++++++++ src/libvirt_public.syms | 5 +++ 4 files changed, 91 insertions(+) diff --git a/include/libvirt/libvirt.h.in b/include/libvirt/libvirt.h.in index 295d551..22f373b 100644 --- a/include/libvirt/libvirt.h.in +++ b/include/libvirt/libvirt.h.in @@ -5245,6 +5245,12 @@ int virDomainFSTrim(virDomainPtr dom, unsigned long long minimum, unsigned int flags); +int virDomainFSFreeze(virDomainPtr dom, + unsigned int flags); + +int virDomainFSThaw(virDomainPtr dom, + unsigned int flags); + /** * virSchedParameterType: * diff --git a/src/driver.h b/src/driver.h index fbfaac4..7e6aec0 100644 --- a/src/driver.h +++ b/src/driver.h @@ -1130,6 +1130,14 @@ typedef int unsigned int flags, int cancelled); +typedef int +(*virDrvDomainFSFreeze)(virDomainPtr dom, + unsigned int flags); + +typedef int +(*virDrvDomainFSThaw)(virDomainPtr dom, + unsigned int flags); + typedef struct _virDriver virDriver; typedef virDriver *virDriverPtr; @@ -1341,6 +1349,8 @@ struct _virDriver { virDrvDomainMigrateFinish3Params domainMigrateFinish3Params; virDrvDomainMigrateConfirm3Params domainMigrateConfirm3Params; virDrvConnectGetCPUModelNames connectGetCPUModelNames; + virDrvDomainFSFreeze domainFSFreeze; + virDrvDomainFSThaw domainFSThaw; }; diff --git a/src/libvirt.c b/src/libvirt.c index a385935..84c9783 100644 --- a/src/libvirt.c +++ b/src/libvirt.c @@ -20552,3 +20552,73 @@ error: virDispatchError(dom->conn); return -1; } + +/** + * virDomainFSFreeze: + * @dom: a domain object + * @flags: extra flags, not used yet, so callers should always pass 0 + * + * Freeze filesystems within the guest (hence guest agent may be + * required depending on hypervisor used). + * + * Returns 0 on success, -1 otherwise. + */ +int +virDomainFSFreeze(virDomainPtr dom, + unsigned int flags) +{ + VIR_DOMAIN_DEBUG(dom, "flags=%x", flags); + + virResetLastError(); + + virCheckDomainReturn(dom, -1); + virCheckReadOnlyGoto(dom->conn->flags, error); + + if (dom->conn->driver->domainFSFreeze) { + int ret = dom->conn->driver->domainFSFreeze(dom, flags); + if (ret < 0) + goto error; + return ret; + } + + virReportUnsupportedError(); + +error: + virDispatchError(dom->conn); + return -1; +} + +/** + * virDomainFSThaw: + * @dom: a domain object + * @flags: extra flags, not used yet, so callers should always pass 0 + * + * Thaw the frozen filesystems within the guest (hence guest agent + * may be required depending on hypervisor used). + * + * Returns 0 on success, -1 otherwise. + */ +int +virDomainFSThaw(virDomainPtr dom, + unsigned int flags) +{ + VIR_DOMAIN_DEBUG(dom, "flags=%x", flags); + + virResetLastError(); + + virCheckDomainReturn(dom, -1); + virCheckReadOnlyGoto(dom->conn->flags, error); + + if (dom->conn->driver->domainFSThaw) { + int ret = dom->conn->driver->domainFSThaw(dom, flags); + if (ret < 0) + goto error; + return ret; + } + + virReportUnsupportedError(); + +error: + virDispatchError(dom->conn); + return -1; +} diff --git a/src/libvirt_public.syms b/src/libvirt_public.syms index 6ed6ce6..9e49a9d 100644 --- a/src/libvirt_public.syms +++ b/src/libvirt_public.syms @@ -645,5 +645,10 @@ LIBVIRT_1.2.1 { virConnectNetworkEventDeregisterAny; } LIBVIRT_1.1.3; +LIBVIRT_1.2.3 { + global: + virDomainFSFreeze; + virDomainFSThaw; +} LIBVIRT_1.2.1; # .... define new API here using predicted next version number ....

1 0

[libvirt] [RFC][PATCH 0/2] Fix zero cpu shares handling
by Ján Tomko 06 Mar '14

06 Mar '14

Hello, This series tries to fix the behavior of <cputune><shares>0</shares></cputune> (means default shares of 1024) and virsh schedinfo --set cpu_shares=0 (means minimum of 2) If 0 was explicitly specified, treat it as a valid value (which will get converted to 2 by the kernel) Re-read the shares value to reflect the cpu shares in live XML. In v1, I've tried to change the schedinfo behavior instead, but I have not found a way to reset the value to the kernel default (which is currently 1024 - see ROOT_TASK_GROUP_LOAD in kernel/sched/sched.h in Linux source) other than writing it to the cgroup fs. Partially fixes https://bugzilla.redhat.com/show_bug.cgi?id=998431 Ján Tomko (2): Treat zero cpu shares as a valid value Show the real cpu shares value in live xml src/conf/domain_conf.c | 12 +++++++----- src/conf/domain_conf.h | 1 + src/lxc/lxc_cgroup.c | 13 ++++++++++--- src/lxc/lxc_driver.c | 8 +++++++- src/lxc/lxc_native.c | 8 +++++--- src/qemu/qemu_cgroup.c | 12 +++++++++--- src/qemu/qemu_driver.c | 13 +++++++++++-- src/vmx/vmx.c | 1 + 8 files changed, 51 insertions(+), 17 deletions(-) -- 1.8.3.2

2 7

[libvirt] [PATCH v2] qemu: Reject unsupported tuning in session mode
by Martin Kletzander 06 Mar '14

06 Mar '14

When domain is started with setting that cannot be done, i.e. those that require cgroups, there is no error reported and it succeeds without any message whatsoever. When setting with API, virsh, an error is reported, but only due to the fact that no cgroups are mounted (priv->cgroup == NULL). Given the above it seems reasonable to reject such unsupported settings. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1023366 Signed-off-by: Martin Kletzander <mkletzan(a)redhat.com> --- Notes: v2: - Allow CPU pinning since setting affinity should still work. - Remove bogus 'cfg = ...' in SetBlockIoTune. src/qemu/qemu_command.c | 26 ++++++++++++++++++ src/qemu/qemu_driver.c | 73 ++++++++++++++++++++++++++++++++++++++++++++++++- 2 files changed, 98 insertions(+), 1 deletion(-) diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index 18d0a64..611d21d 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -7762,6 +7762,32 @@ qemuBuildCommandLine(virConnectPtr conn, emulator = def->emulator; + if (!cfg->privileged) { + /* If we have no cgroups than we can have no tunings that + * require them */ + + if (def->mem.hard_limit || def->mem.soft_limit || + def->mem.min_guarantee || def->mem.swap_hard_limit) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", + _("Memory tuning is not available in session mode")); + goto error; + } + + if (def->blkio.weight || def->blkio.ndevices) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", + _("Block I/O tuning is not available in session mode")); + goto error; + } + + if (def->cputune.shares || def->cputune.period || + def->cputune.quota || def->cputune.emulator_period || + def->cputune.emulator_quota) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", + _("CPU tuning is not available in session mode")); + goto error; + } + } + for (i = 0; i < def->ngraphics; ++i) { switch (def->graphics[i]->type) { case VIR_DOMAIN_GRAPHICS_TYPE_SDL: diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index 4fbcb27..1e50572 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -7434,6 +7434,8 @@ static char *qemuDomainGetSchedulerType(virDomainPtr dom, char *ret = NULL; virDomainObjPtr vm = NULL; qemuDomainObjPrivatePtr priv; + virQEMUDriverPtr driver = dom->conn->privateData; + virQEMUDriverConfigPtr cfg = NULL; if (!(vm = qemuDomObjFromDomain(dom))) goto cleanup; @@ -7443,6 +7445,13 @@ static char *qemuDomainGetSchedulerType(virDomainPtr dom, if (virDomainGetSchedulerTypeEnsureACL(dom->conn, vm->def) < 0) goto cleanup; + cfg = virQEMUDriverGetConfig(driver); + if (!cfg->privileged) { + virReportError(VIR_ERR_OPERATION_UNSUPPORTED, "%s", + _("CPU tuning is not available in session mode")); + goto cleanup; + } + /* Domain not running, thus no cgroups - return defaults */ if (!virDomainObjIsActive(vm)) { if (nparams) @@ -7469,6 +7478,7 @@ static char *qemuDomainGetSchedulerType(virDomainPtr dom, cleanup: if (vm) virObjectUnlock(vm); + virObjectUnref(cfg); return ret; } @@ -7683,6 +7693,12 @@ qemuDomainSetBlkioParameters(virDomainPtr dom, if (virDomainSetBlkioParametersEnsureACL(dom->conn, vm->def, flags) < 0) goto cleanup; + if (!cfg->privileged) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", + _("Block I/O tuning is not available in session mode")); + goto cleanup; + } + if (!(caps = virQEMUDriverGetCapabilities(driver, false))) goto cleanup; @@ -7863,6 +7879,7 @@ qemuDomainGetBlkioParameters(virDomainPtr dom, int ret = -1; virCapsPtr caps = NULL; qemuDomainObjPrivatePtr priv; + virQEMUDriverConfigPtr cfg = NULL; virCheckFlags(VIR_DOMAIN_AFFECT_LIVE | VIR_DOMAIN_AFFECT_CONFIG | @@ -7881,6 +7898,13 @@ qemuDomainGetBlkioParameters(virDomainPtr dom, if (virDomainGetBlkioParametersEnsureACL(dom->conn, vm->def) < 0) goto cleanup; + cfg = virQEMUDriverGetConfig(driver); + if (!cfg->privileged) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", + _("Block I/O tuning is not available in session mode")); + goto cleanup; + } + if (!(caps = virQEMUDriverGetCapabilities(driver, false))) goto cleanup; @@ -8269,6 +8293,7 @@ cleanup: if (vm) virObjectUnlock(vm); virObjectUnref(caps); + virObjectUnref(cfg); return ret; } @@ -8316,6 +8341,12 @@ qemuDomainSetMemoryParameters(virDomainPtr dom, if (virDomainSetMemoryParametersEnsureACL(dom->conn, vm->def, flags) < 0) goto cleanup; + if (!cfg->privileged) { + virReportError(VIR_ERR_OPERATION_UNSUPPORTED, "%s", + _("Memory tuning is not available in session mode")); + goto cleanup; + } + if (!(caps = virQEMUDriverGetCapabilities(driver, false))) goto cleanup; @@ -8423,6 +8454,7 @@ qemuDomainGetMemoryParameters(virDomainPtr dom, int ret = -1; virCapsPtr caps = NULL; qemuDomainObjPrivatePtr priv; + virQEMUDriverConfigPtr cfg = NULL; virCheckFlags(VIR_DOMAIN_AFFECT_LIVE | VIR_DOMAIN_AFFECT_CONFIG | @@ -8439,6 +8471,13 @@ qemuDomainGetMemoryParameters(virDomainPtr dom, if (virDomainGetMemoryParametersEnsureACL(dom->conn, vm->def) < 0) goto cleanup; + cfg = virQEMUDriverGetConfig(driver); + if (!cfg->privileged) { + virReportError(VIR_ERR_OPERATION_UNSUPPORTED, "%s", + _("Memory tuning is not available in session mode")); + goto cleanup; + } + if (!(caps = virQEMUDriverGetCapabilities(driver, false))) goto cleanup; @@ -8550,6 +8589,7 @@ cleanup: if (vm) virObjectUnlock(vm); virObjectUnref(caps); + virObjectUnref(cfg); return ret; } @@ -8975,6 +9015,13 @@ qemuDomainSetSchedulerParametersFlags(virDomainPtr dom, if (virDomainSetSchedulerParametersFlagsEnsureACL(dom->conn, vm->def, flags) < 0) goto cleanup; + cfg = virQEMUDriverGetConfig(driver); + if (!cfg->privileged) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", + _("CPU tuning is not available in session mode")); + goto cleanup; + } + if (!(caps = virQEMUDriverGetCapabilities(driver, false))) goto cleanup; @@ -9210,6 +9257,7 @@ qemuDomainGetSchedulerParametersFlags(virDomainPtr dom, virDomainDefPtr persistentDef; virCapsPtr caps = NULL; qemuDomainObjPrivatePtr priv; + virQEMUDriverConfigPtr cfg = NULL; virCheckFlags(VIR_DOMAIN_AFFECT_LIVE | VIR_DOMAIN_AFFECT_CONFIG | @@ -9226,6 +9274,13 @@ qemuDomainGetSchedulerParametersFlags(virDomainPtr dom, if (virDomainGetSchedulerParametersFlagsEnsureACL(dom->conn, vm->def) < 0) goto cleanup; + cfg = virQEMUDriverGetConfig(driver); + if (!cfg->privileged) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", + _("CPU tuning is not available in session mode")); + goto cleanup; + } + if (*nparams > 1) cpu_bw_status = virCgroupSupportsCpuBW(priv->cgroup); @@ -9320,6 +9375,7 @@ cleanup: if (vm) virObjectUnlock(vm); virObjectUnref(caps); + virObjectUnref(cfg); return ret; } @@ -15521,11 +15577,17 @@ qemuDomainSetBlockIoTune(virDomainPtr dom, if (virDomainSetBlockIoTuneEnsureACL(dom->conn, vm->def, flags) < 0) goto cleanup; + cfg = virQEMUDriverGetConfig(driver); + if (!cfg->privileged) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", + _("Block I/O tuning is not available in session mode")); + goto cleanup; + } + if (qemuDomainObjBeginJob(driver, vm, QEMU_JOB_MODIFY) < 0) goto cleanup; priv = vm->privateData; - cfg = virQEMUDriverGetConfig(driver); if (!(caps = virQEMUDriverGetCapabilities(driver, false))) goto endjob; @@ -15668,6 +15730,7 @@ qemuDomainGetBlockIoTune(virDomainPtr dom, int ret = -1; size_t i; virCapsPtr caps = NULL; + virQEMUDriverConfigPtr cfg = NULL; virCheckFlags(VIR_DOMAIN_AFFECT_LIVE | VIR_DOMAIN_AFFECT_CONFIG | @@ -15682,6 +15745,13 @@ qemuDomainGetBlockIoTune(virDomainPtr dom, if (virDomainGetBlockIoTuneEnsureACL(dom->conn, vm->def) < 0) goto cleanup; + cfg = virQEMUDriverGetConfig(driver); + if (!cfg->privileged) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", + _("Block I/O tuning is not available in session mode")); + goto cleanup; + } + if (!(caps = virQEMUDriverGetCapabilities(driver, false))) goto cleanup; @@ -15784,6 +15854,7 @@ cleanup: if (vm) virObjectUnlock(vm); virObjectUnref(caps); + virObjectUnref(cfg); return ret; } -- 1.9.0

2 4

[libvirt] virDomainShutdown() API failed with error '-1'
by Panday Ritesh Sharma (rpanday) 06 Mar '14

06 Mar '14

Hi Team, We are calling virDomainShutdown() API for shutting the VM. Sometime we see, this API is failing because of 'unknown' reason. It took almost 15 secs to bringdown the VM and at last failed. And the domain goes into shut off state. You can find the time stamp in below log. Could you please let us know in which case this API can fail, how to fix/debug this issue ? We are running 1.0.2 libvirtd version. =========== 20.53.48.532818176: Shuting down domain... 20.54.03.539286528: Failure of libvirt library call: Code: internal error Context: None Message: An error occurred, but the cause is unknown Level: ERROR 20.54.03.539291392: virDomainShutdown() call returned error -1 for CP—3 ============ =========== [host:/var/log/libvirt/qemu]$ virsh list --all Id Name State ---------------------------------------------------- - CP--3 shut off =========== ============ [host:/var/log/libvirt/qemu]$ [host:/var/log/libvirt/qemu]$ libvirtd --version libvirtd (libvirt) 1.0.2 [host:/var/log/libvirt/qemu]$ ============ Regards Ritesh Sharma

1 0

[libvirt] [PATCH 0/4] Add caching of QEMU probed capabilities
by Daniel P. Berrange 06 Mar '14

06 Mar '14

Probing capabilities takes 200-300ms per binary and we have as many as 26 binaries. This noticably slows down libvirtd startup. It does not look like performance of probing QEMU can be improved, so this series introduces caching of the capabilities information. So the first time libvirtd starts it'll be slow, but thereafter it is fast. The cache is invalidated any time the QEMU binary timestamp changes or the libvirtd binary or driver module timestamp changes. Daniel P. Berrange (4): Add helper APIs for generating cryptographic hashes Convert lock driver plugins to use new crypto APIs Cache result of QEMU capabilities extraction Refresh qemu capabilities if libvirtd binary changes .gitignore | 1 + daemon/libvirtd.c | 2 + include/libvirt/virterror.h | 1 + po/POTFILES.in | 1 + src/Makefile.am | 1 + src/driver.c | 2 + src/libvirt_private.syms | 6 + src/locking/lock_driver_lockd.c | 32 +-- src/locking/lock_driver_sanlock.c | 42 +--- src/qemu/qemu_capabilities.c | 418 +++++++++++++++++++++++++++++++++++++- src/qemu/qemu_capabilities.h | 2 + src/qemu/qemu_driver.c | 1 + src/util/vircrypto.c | 80 ++++++++ src/util/vircrypto.h | 40 ++++ src/util/virerror.c | 1 + src/util/virutil.c | 23 +++ src/util/virutil.h | 4 + tests/Makefile.am | 5 + tests/vircryptotest.c | 90 ++++++++ 19 files changed, 682 insertions(+), 70 deletions(-) create mode 100644 src/util/vircrypto.c create mode 100644 src/util/vircrypto.h create mode 100644 tests/vircryptotest.c -- 1.8.5.3

3 12

[libvirt] [RFC] support memory reserved feature and optimize mlock guest memory propose
by Zhanghailiang 06 Mar '14

06 Mar '14

Hi all： Currently, we use cgroup(memory) to support memory QoS on KVM platform, and use "mlock" on qemu to support "memory reserved". The "mlock" seems to be not appropriate. Now qemu "mlock" memory in the main thread, which would lock iothread (qemu_mutex_lock_iothread), if the memory size is large, that will consume lots of time. It means whenever we want to set a new 'mlock', the VM would be blocked for a while. Here is my optimization: 1. Add a global variable (lock_ram_size) to save the value of "memory reserved"; 2. Add a qmp commond "set_ram_minguarantee" to change lock_ram_size; 3. Create a new thread to mlock(lock_ram_zie), while is waked up by the "set_ram_minguarantee" qmp command. Flow chart: main funciton qmp command "set_ram_minguarantee" | | | | create "mlock" thread change value of lock_ram_size | | | | |------>thread wait<-------------wake up "mlock" thread | | | | | | |-------mlock(lock_ram_zie) We have tested this demo a few days, it seems to be worked well. But we are not sure is there any other problems , if the main thread and mlock thread access one memory zone at one time without a mutex lock. Is it workable?Or Is there any other idea to support "memory reserved" ? Thanks zhanghailiang

2 3

[libvirt] [PATCH v3 0/2] fix query-command-line-options
by Amos Kong 06 Mar '14

06 Mar '14

This patchset fixed some issues of query-command-line-options: * some new options that haven't argument can't be queried. (eg: -enable-fips) * some legacy options that have argument can't be queried. (eg: -vnc display) More discussion: http://marc.info/?l=qemu-devel&m=139081830416684&w=2 V2: remove duplicate option tables, update schema V3: fix typo in commitlog and export qemu_options talbe Amos Kong (2): qmp: rename query_option_descs() to get_param_infolist() query-command-line-options: query all the options in qemu-options.hx qapi-schema.json | 8 ++++++-- qemu-options.h | 18 ++++++++++++++++++ util/qemu-config.c | 41 ++++++++++++++++++++++++++++++++--------- vl.c | 17 ----------------- 4 files changed, 56 insertions(+), 28 deletions(-) -- 1.8.5.3

2 8

[libvirt] [PATCH 00/10] virFork cleanups
by Eric Blake 05 Mar '14

05 Mar '14

Some of these patches were written while working on CVE-2013-6456; we decided to reorder things and fix that problem first. While rebasing these patches to the latest tree, I found other things worth fixing. Eric Blake (10): nwfilter: don't ignore child process failures virFork: give specific status on failure prior to exec util: make it easier to reflect child exit status util: preserve exit status from mount namespace callback util: make it easier to grab only regular process exit util: make it easier to grab only regular command exit virFork: simplify semantics virt-login-shell: use single instead of double fork virt-login-shell: saner exit value virsh: report exit status of failed lxc-enter-namespace daemon/libvirtd.c | 4 +- daemon/remote.c | 7 +- docs/internals/command.html.in | 17 ++- src/access/viraccessdriverpolkit.c | 9 +- src/bhyve/bhyve_process.c | 19 +--- src/fdstream.c | 3 +- src/internal.h | 7 ++ src/libvirt.c | 4 +- src/libvirt_private.syms | 2 + src/lxc/lxc_container.c | 6 +- src/lxc/lxc_process.c | 11 +- src/nwfilter/nwfilter_ebiptables_driver.c | 89 ++++++--------- src/openvz/openvz_driver.c | 18 +--- src/qemu/qemu_capabilities.c | 1 + src/qemu/qemu_command.c | 3 +- src/storage/storage_backend_iscsi.c | 7 +- src/util/vircommand.c | 173 +++++++++++++++--------------- src/util/vircommand.h | 4 +- src/util/virebtables.c | 5 +- src/util/virfile.c | 35 ++---- src/util/viriptables.c | 7 +- src/util/virnetdevveth.c | 4 +- src/util/virprocess.c | 121 +++++++++++++++------ src/util/virprocess.h | 8 +- src/xen/xen_driver.c | 9 +- tests/commandtest.c | 126 +++++++++++++++++++++- tests/reconnect.c | 3 +- tests/statstest.c | 3 +- tests/testutils.c | 4 +- tools/virsh-domain.c | 30 +++--- tools/virsh.pod | 5 +- tools/virt-login-shell.c | 141 ++++++++++-------------- tools/virt-login-shell.pod | 25 ++++- 33 files changed, 525 insertions(+), 385 deletions(-) -- 1.8.5.3

5 21

[libvirt] [PATCH] Fix missing char dev lock path case in configure
by Roman Bogorodskiy 05 Mar '14

05 Mar '14

configure check for character devices lock path calls AC_DEFINE_UNQUOTED for VIR_CHRDEV_LOCK_FILE_PATH even if $with_chrdev_lock_files = "no". So the locking code in conf/virchrdev.c: #ifdef VIR_CHRDEV_LOCK_FILE_PATH is compiled in even if it shouldn't, because VIR_CHRDEV_LOCK_FILE_PATH is defined as "no", so it tries to create lock files with strange lock path like 'no/LCK..'. Fix that by calling AC_DEFINE_UNQUOTED only if $with_chrdev_lock_files is not 'no'. --- configure.ac | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/configure.ac b/configure.ac index a9339ce..62b74c5 100644 --- a/configure.ac +++ b/configure.ac @@ -1401,8 +1401,10 @@ if test "$with_chrdev_lock_files" != "no"; then AC_MSG_ERROR([You must specify path for the lock files on this platform]) fi - AC_DEFINE_UNQUOTED([VIR_CHRDEV_LOCK_FILE_PATH], "$with_chrdev_lock_files", - [path to directory containing UUCP device lock files]) + if test "$with_chrdev_lock_files" != "no"; then + AC_DEFINE_UNQUOTED([VIR_CHRDEV_LOCK_FILE_PATH], "$with_chrdev_lock_files", + [path to directory containing UUCP device lock files]) + fi fi AM_CONDITIONAL([VIR_CHRDEV_LOCK_FILE_PATH], [test "$with_chrdev_lock_files" != "no"]) -- 1.8.4.2

2 2

[libvirt] [PATCH 00/13] libxl: add basic support for migration
by Jim Fehlig 05 Mar '14

05 Mar '14

Based on an earlier patch from Chunyan Liu https://www.redhat.com/archives/libvir-list/2013-September/msg00667.html This patch series adds basic migration support to the libxl driver. Follow-up patches can improve pre-migration checks and add support for additional migration flags. Patches 1-12 are almost exclusively code motion, moving functions from the main driver module into the libxl_domain and libxl_conf modules. Patch13 is rather large and provides the migration implementation. If it is preferred to split patch13 further, I'm open to suggestions on ways to do that. I've done some basic migration testing of this series using NFS shared storage. Error handling only minimally tested though. I attempted a migration to host with a different bridge than the source, verifying the migration failed and the vm was resumed on the source. Jim Fehlig (13): libxl: move libxlDomainEventQueue to libxl_domain libxl: move libxlDomainManagedSavePath to libxl_domain libxl: move libxlSaveImageOpen to libxl_domain libxl: move libxlVmCleanup{,Job} to libxl_domain libxl: move libxlDomEventsRegister to libxl_domain libxl: move libxlDomainAutoCoreDump to libxl_domain libxl: move libxlDoNodeGetInfo to libxl_conf libxl: move libxlDomainSetVcpuAffinities to libxl_domain libxl: move libxlFreeMem to libxl_domain libxl: move libxlVmStart to libxl_domain libxl: include a pointer to the driver in libxlDomainObjPrivate libxl: move domain event handler to libxl_domain libxl: add migration support po/POTFILES.in | 1 + src/Makefile.am | 3 +- src/libxl/libxl_conf.c | 36 ++ src/libxl/libxl_conf.h | 10 + src/libxl/libxl_domain.c | 696 +++++++++++++++++++++++++++++++ src/libxl/libxl_domain.h | 51 ++- src/libxl/libxl_driver.c | 972 +++++++++++--------------------------------- src/libxl/libxl_migration.c | 579 ++++++++++++++++++++++++++ src/libxl/libxl_migration.h | 78 ++++ 9 files changed, 1681 insertions(+), 745 deletions(-) create mode 100644 src/libxl/libxl_migration.c create mode 100644 src/libxl/libxl_migration.h -- 1.8.1.4

1 14

[libvirt] [PATCH] qemu: monitor: Provide more information in generic block job error
by Peter Krempa 05 Mar '14

05 Mar '14

The qemuMonitorJSONBlockJob handles a few errors internally. If qemu returns a different error we would report a rather unhelpful message: $ virsh blockpull gluster-job vda --base /dev/null error: internal error: Unexpected error As the actual message from qemu contains a bit more info, let's use it to report something a little more useful: $ virsh blockpull gluster-job vda --base /dev/null error: internal error: Unexpected error: (GenericError) 'Base '/dev/null' not found' --- src/qemu/qemu_monitor_json.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/src/qemu/qemu_monitor_json.c b/src/qemu/qemu_monitor_json.c index 9a5b812..ee3ae15 100644 --- a/src/qemu/qemu_monitor_json.c +++ b/src/qemu/qemu_monitor_json.c @@ -3678,8 +3678,12 @@ qemuMonitorJSONBlockJob(qemuMonitorPtr mon, virReportError(VIR_ERR_OPERATION_INVALID, _("Command '%s' is not found"), cmd_name); } else { - virReportError(VIR_ERR_INTERNAL_ERROR, "%s", - _("Unexpected error")); + virJSONValuePtr error = virJSONValueObjectGet(reply, "error"); + + virReportError(VIR_ERR_INTERNAL_ERROR, + _("Unexpected error: (%s) '%s'"), + NULLSTR(virJSONValueObjectGetString(error, "class")), + NULLSTR(virJSONValueObjectGetString(error, "desc"))); } } -- 1.9.0

2 2

[libvirt] [PATCHv8 0/2] Implement RBD storage pool support
by Adam Walters 05 Mar '14

05 Mar '14

This is an updated version of my RBD storage pool support patch. Relevant changes since v7 are: - Fixed bug where I compared disk->srcpool->mode to VIR_DOMAIN_DISK_TYPE_NETWORK In response to a couple of items Ján mentioned, I decided not to pursue adding anything to qemuxml2argvtest. The reason I decided not to pursue this is there is not really any need. In comparison with other volume types, RBD already has the same test coverage. Namely, a base set of test-cases without using storage pools, plus a generic storage pool test. In addition, I also attempted to implement a proper check for expected vs actual secret usage, but at that particular point in the code, Ján was correct that the pool definition has not yet been parsed. Thus, I had to keep my bypass. Adam Walters (2): qemu: conf Implement domain RBD storage pool support domain: conf: Fix secret type checking for network volumes src/conf/domain_conf.c | 9 ++++++-- src/qemu/qemu_conf.c | 62 +++++++++++++++++++++++++++++++++++++++++++++++++- 2 files changed, 68 insertions(+), 3 deletions(-) -- 1.8.5.3

2 4

[libvirt] [PATCH v13 00/49] write separate module for hostdev passthrough
by Chunyan Liu 05 Mar '14

05 Mar '14

These patches implements a separate module for hostdev passthrough so that it could be shared by different drivers and can maintain a global state of a host device. Patches 1~6 are to switch existing qemu and lxc driver to use common library lists, so that to maintain a global state of every host device. Patches 7~45 are to extract general code from qemu_hostdev.c piece by piece, make them reusable common APIs. Patches 46: unit test for the virhostdev common library Patches 47: add a hostdev backend type for xen Patches 48: add pci passthrough to libxl driver based on the common library Patches 49: change lxc to use common library APIs --- changes to v12: * split "add hostdev passthrough common library" patch into small patches for easier review. * fix v12 comments * rebase to libxl changes Chunyan Liu (49): add 'driver' info to used_by qemu: reuse hostdev interfaces to avoid duplicate qemu: remove functions now used internally only from qemu_hostdev.h add virhostdev files to maintain global state of host devices qemu: use general virhostdev lists instead of its own lxc: use general virhostdev lists instead of its own qemu_hostdev: move cfg->relaxedACS as a flag qemu_hostdev: move ColdBoot as a flag qemu_hostdev: move netconfig file location to virhostdev stateDir extract general code from qemuPrepareHostdevPCIDevices rename qemu*NetConfigRestore/Replace to virHostdevNetConfigRestore/Replace rename qemuGet*PciHostDeviceList to virHostdevGet*PciHostDeviceList pass driver name as a parameter to virHostdevPrepareHostdevPCIDevices extract general code from qemuDomainReAttachHostdevDevices pass driver name as a parameter to virHostdevReAttachPCIDevices rename qemuReAttachPciDevice to virHostdevReAttachPciDevice move virHostdevPrepare(ReAttach)PCIDevices to virhostdev.c extract general code from qemuUpdateActivePciHostdevs extract general code from qemuUpdateActiveUsbHostdevs extract general code from qemuUpdateActiveScsiHostdevs pass driver_name as parameter of virHostdevUpdate*Hostdevs functions move virHostdevUpdate* functions to virhostdev.c qemuPrepareUSBDevices: code adjustment for extracting general code extract general code from qemuPrepareHostUSBDevices rename qemu*USBDevices to virHostdev*USBDevices pass driver name to virHostdevPrepareUSBDevices move virHostdevPrepareHostUSBDevices to virhostdev.c extract general code from qemuPrepareHostSCSIDevices pass driver name as parameter to virHostdevPrepareSCSIDevices move virHostdevPrepareHostSCSIDevices to virhostdev.c extract general code from qemuDomainReAttachHostUsbDevices pass driver name as paramter to virHostdevReAttachUsbHostdevs move virHostdevDomainReAttachHostUsbDevices to virhostdev.c extract general code from qemuDomainReAttachHostScsiDevices pass driver name as parameter to virHostdevReAttachScciHostdevs move virHostdevReAttachHostScsiDevices to virhostdev.c extract general code of NodeDeviceDetach extract general code of NodeDeviceReAttach extract general code of NodeDeviceReset move virHostdevNodeDevice* to virhostdev.c improve parameter name to let it more meaningful rename some function names to keep consistency improve virHostdevUpdate* parameters to make it more widely used add 3 wrapper functions for prepare/reattach/update domain hostdevs add parameter checks to common interfaces add unit test for new virhostdev common library change lxc_hostdev.c to use virhostdev common library APIs add hostdev pci backend type for xen add pci passthrough to libxl driver .gitignore | 1 + docs/schemas/domaincommon.rng | 1 + po/POTFILES.in | 1 + src/Makefile.am | 1 + src/conf/domain_conf.c | 3 +- src/conf/domain_conf.h | 1 + src/libvirt_private.syms | 19 + src/libxl/libxl_conf.c | 63 ++ src/libxl/libxl_conf.h | 4 + src/libxl/libxl_domain.c | 9 + src/libxl/libxl_driver.c | 447 +++++++++++- src/lxc/lxc_conf.h | 4 - src/lxc/lxc_driver.c | 17 +- src/lxc/lxc_hostdev.c | 315 +-------- src/qemu/qemu_command.c | 3 +- src/qemu/qemu_conf.h | 10 +- src/qemu/qemu_driver.c | 88 +-- src/qemu/qemu_hostdev.c | 1220 ++----------------------------- src/qemu/qemu_hostdev.h | 27 +- src/qemu/qemu_hotplug.c | 77 +-- src/qemu/qemu_process.c | 8 +- src/util/virhostdev.c | 1621 +++++++++++++++++++++++++++++++++++++++++ src/util/virhostdev.h | 140 ++++ src/util/virpci.c | 31 +- src/util/virpci.h | 9 +- src/util/virscsi.c | 32 +- src/util/virscsi.h | 7 +- src/util/virusb.c | 31 +- src/util/virusb.h | 8 +- tests/Makefile.am | 5 + tests/virhostdevtest.c | 507 +++++++++++++ tests/virscsitest.c | 6 +- 32 files changed, 3081 insertions(+), 1635 deletions(-) create mode 100644 src/util/virhostdev.c create mode 100644 src/util/virhostdev.h create mode 100644 tests/virhostdevtest.c

5 69

[libvirt] [PATCH v4 0/3] support dumping guest memory in compressed format
by qiaonuohan 05 Mar '14

05 Mar '14

dumping guest's memroy is introduced without compression supported, and this is a freature regression of 'virsh dump --memory-only'. This patchset is used to add support in libvirt side to make qemu dump guest's memory in kdump-compressed format and please refer the following address to see implementation of the qemu side, the lastest version of qemu side is v9(ready for being queued). http://lists.nongnu.org/archive/html/qemu-devel/2014-02/msg03016.html ChangLog: Changes from v3 to v4: 1. dropping patch "add dump_memory_format in qemu.conf" 2. fix to follow some conventions Changes from v2 to v3: 1. address Jiri Denemark's comment about adding a new public API instead of changing an old one. Changes from v1 to v2: 1. address Daniel P. Berrange's comment about using a new parameter to replace flags like VIR_DUMP_COMPRESS_ZLIB. qiaonuohan (3): add new virDomainCoreDumpWithFormat API add qemu support to virDomainCoreDumpWithFormat API allow "virsh dump --memory-only" specify dump format include/libvirt/libvirt.h.in | 31 ++++++++++++++ src/driver.h | 7 +++ src/libvirt.c | 100 +++++++++++++++++++++++++++++++++++++++++++ src/libvirt_public.syms | 5 +++ src/qemu/qemu_driver.c | 45 +++++++++++++++---- src/qemu/qemu_monitor.c | 7 +-- src/qemu/qemu_monitor.h | 3 +- src/qemu/qemu_monitor_json.c | 4 +- src/qemu/qemu_monitor_json.h | 3 +- src/remote/remote_driver.c | 1 + src/remote/remote_protocol.x | 15 ++++++- src/remote_protocol-structs | 7 +++ src/test/test_driver.c | 19 ++++++-- tests/qemumonitorjsontest.c | 2 +- tools/virsh-domain.c | 45 ++++++++++++++++++- 15 files changed, 273 insertions(+), 21 deletions(-) -- 1.8.5.3

3 12

[libvirt] [PATCHv2 0/7] various fixes related to gluster volumes
by Peter Krempa 05 Mar '14

05 Mar '14

Peter Krempa (7): virsh: volume: Fix lookup of volumes to provide better error messages storage: Use cleanup label instead of out storage: Error out when attempting to vol-upload into a remote pool storage: Avoid mangling paths of non-local filesystems when looking up storage: Don't lie about path used to look up in error message doc: storage: Explicitly state that it's possible to have non-unique key gluster: Fix "key" attribute for gluster volumes docs/formatstorage.html.in | 6 +- src/storage/storage_backend_gluster.c | 10 +- src/storage/storage_driver.c | 215 +++++++++++++++++++++------------- tools/virsh-volume.c | 12 +- 4 files changed, 154 insertions(+), 89 deletions(-) -- 1.9.0

2 15

[libvirt] libvirt xml can't specify USB physical port when passthrough USB device to VM
by Wangrui (K) 05 Mar '14

05 Mar '14

Hello, I passthrough a host usb device to a guest with a bus number and device number in the xml to identify the usb device. When I unplug and then plug back into the same USB physical port, the guest identifies the usb device failed. Because on the host ,the usb device number increases. So , I want to know that whether Libvirt will provide an interface that identifies a host usb device with bus number and physical port . As far as I know ,Qemu upstream had added a hostport property which allows to specify the host usb with bus number and physical port. the patch is : https://lists.gnu.org/archive/html/qemu-devel/2011-05/msg02341.html Best Regards

1 0

[libvirt] [Question] Will we support to set boot order online?
by Wangyufei (James) 05 Mar '14

05 Mar '14

Hello, I find that QEMU supports to set boot(QMP command "boot_set") order online, namely the virtual machine can be reboot in appointed order, but libvirt had no corresponding interface. So, I want to know if libvirt intend to support to set boot order online. If the answer is yes, I'll try to apply the patch. If no, why? Best Regards, -WangYufei

3 4

[libvirt] [PATCH] qemu: add PCI-multibus support for ppc
by Olivia Yin 05 Mar '14

05 Mar '14

Signed-off-by: Olivia Yin <hong-hua.yin(a)freescale.com> --- src/qemu/qemu_capabilities.c | 10 ++++++++++ 1 files changed, 10 insertions(+), 0 deletions(-) diff --git a/src/qemu/qemu_capabilities.c b/src/qemu/qemu_capabilities.c index 7bc1ebc..7d7791d 100644 --- a/src/qemu/qemu_capabilities.c +++ b/src/qemu/qemu_capabilities.c @@ -2209,6 +2209,11 @@ virQEMUCapsInitHelp(virQEMUCapsPtr qemuCaps, uid_t runUid, gid_t runGid) virQEMUCapsClear(qemuCaps, QEMU_CAPS_NO_ACPI); } + /* ppc support PCI-multibus */ + if (qemuCaps->arch == VIR_ARCH_PPC) { + virQEMUCapsSet(qemuCaps, QEMU_CAPS_PCI_MULTIBUS); + } + /* virQEMUCapsExtractDeviceStr will only set additional caps if qemu * understands the 0.13.0+ notion of "-device driver,". */ if (virQEMUCapsGet(qemuCaps, QEMU_CAPS_DEVICE) && @@ -2450,6 +2455,11 @@ virQEMUCapsInitQMP(virQEMUCapsPtr qemuCaps, virQEMUCapsSet(qemuCaps, QEMU_CAPS_NO_ACPI); } + /* ppc support PCI-multibus */ + if (qemuCaps->arch == VIR_ARCH_PPC) { + virQEMUCapsSet(qemuCaps, QEMU_CAPS_PCI_MULTIBUS); + } + if (virQEMUCapsProbeQMPCommands(qemuCaps, mon) < 0) goto cleanup; if (virQEMUCapsProbeQMPEvents(qemuCaps, mon) < 0) -- 1.6.4

5 7

[libvirt] [PATCH 0/4] Expose FSFreeze/FSThaw within the guest as commands
by Tomoki Sekiyama 04 Mar '14

04 Mar '14

Currently FSFreeze and FSThaw are supported by qemu guest agent and they are used internally in snapshot-create command with --quiesce option. However, when users want to utilize the native snapshot feature of storage devices (such as LVM over iSCSI, various enterprise storage systems, etc.), they need to issue fsfreeze command separately from libvirt-driven snapshots. (OpenStack cinder provides these storages' snapshot feature, but it cannot quiesce the guest filesystems automatically for now.) Although virDomainQemuGuestAgent() API could be used for this purpose, it depends too much on specific hypervisor implementation. This patchset adds virDomainFSFreeze()/virDomainFSThaw() APIs and virsh domfsfreeze/domfsthaw commands to enable the users to freeze and thaw domain's filesystems cleanly. The APIs has mountPoint and flags option currently unsupported for future extension, as virDomainFSTrim() API. Duplicated FSFreeze results in error caused by qemu guest agent. --- Tomoki Sekiyama (4): Introduce virDomainFSFreeze() public API remote: Implement virDomainFSFreeze and virDomainFSThaw qemu: Implement virDomainFSFreeze virsh: Expose new virDomainFSFreeze and virDomainFSThaw API include/libvirt/libvirt.h.in | 8 ++ src/access/viraccessperm.c | 2 - src/access/viraccessperm.h | 6 ++ src/driver.h | 12 ++++ src/libvirt.c | 92 +++++++++++++++++++++++++++ src/libvirt_public.syms | 6 ++ src/qemu/qemu_driver.c | 142 ++++++++++++++++++++++++++++++++++++++++++ src/remote/remote_driver.c | 2 + src/remote/remote_protocol.x | 26 +++++++- src/remote_protocol-structs | 12 ++++ src/rpc/gendispatch.pl | 2 + tools/virsh-domain.c | 108 ++++++++++++++++++++++++++++++++ tools/virsh.pod | 17 +++++ 13 files changed, 433 insertions(+), 2 deletions(-)

3 10

[libvirt] [PATCH 0/2] Properly error out on unsupported settings
by Martin Kletzander 04 Mar '14

04 Mar '14

<BLURB/> Martin Kletzander (2): tests: Create privileged driver config in qemuxml2argvtest qemu: Reject unsupported tuning in session mode src/qemu/qemu_command.c | 32 +++++++++++++++ src/qemu/qemu_driver.c | 103 ++++++++++++++++++++++++++++++++++++++++++++++- tests/qemuxml2argvtest.c | 2 +- 3 files changed, 135 insertions(+), 2 deletions(-) -- 1.9.0

2 7

[libvirt] [PATCH v2 0/2] Introduce max_anonymous_clients
by Michal Privoznik 04 Mar '14

04 Mar '14

https://bugzilla.redhat.com/show_bug.cgi?id=981729 So far we can limit how many clients are connected, how many are waiting in accept() line but we could not control the count of accepted but not authenticated yet. Michal Privoznik (2): virNetServer: Introduce unauth clients counter daemon: Introduce max_anonymous_clients daemon/libvirtd-config.c | 1 + daemon/libvirtd-config.h | 1 + daemon/libvirtd.aug | 1 + daemon/libvirtd.c | 1 + daemon/libvirtd.conf | 4 +++ daemon/remote.c | 21 ++++++----- daemon/test_libvirtd.aug.in | 1 + src/locking/lock_daemon.c | 3 +- src/lxc/lxc_controller.c | 2 +- src/rpc/virnetserver.c | 88 ++++++++++++++++++++++++++++++++++++++++++--- src/rpc/virnetserver.h | 3 ++ 11 files changed, 110 insertions(+), 16 deletions(-) -- 1.9.0

2 8