- Devel - libvirt List Archives

[libvirt] [PATCH] fdstream: Report error from the I/O thread
by John Ferlan 06 Jun '18

06 Jun '18

https://bugzilla.redhat.com/show_bug.cgi?id=1529059 Commit id 0fe4aa14 added the thread specific error message reporting (or save) to virFDStreamEvent; however, as processing goes via virStream{Send|SendHole|Recv} via calls from daemonStreamHandle{WriteData|Hole|Read} the last error gets reset in the main libvirt API's thus, whatever error may have been set as last error will be cleared prior to the error paths using it resulting in the generic error on the client side. For each of the paths that check threadQuit or threadErr, check if threadErr was set and set it agian if there isn't a last error (e.g. some other failure) set so that the message can be provided back to the client. Signed-off-by: John Ferlan <jferlan(a)redhat.com> --- src/util/virfdstream.c | 26 ++++++++++++++++++++------ 1 file changed, 20 insertions(+), 6 deletions(-) diff --git a/src/util/virfdstream.c b/src/util/virfdstream.c index e4973a2bd0..8189559964 100644 --- a/src/util/virfdstream.c +++ b/src/util/virfdstream.c @@ -795,8 +795,13 @@ static int virFDStreamWrite(virStreamPtr st, const char *bytes, size_t nbytes) char *buf; if (fdst->threadQuit || fdst->threadErr) { - virReportSystemError(EBADF, "%s", - _("cannot write to stream")); + + /* virStreamSend will virResetLastError possibly set + * by virFDStreamEvent */ + if (fdst->threadErr && !virGetLastError()) + virSetError(fdst->threadErr); + else + virReportSystemError(EBADF, "%s", _("cannot write to stream")); goto cleanup; } @@ -875,8 +880,13 @@ static int virFDStreamRead(virStreamPtr st, char *bytes, size_t nbytes) while (!(msg = fdst->msg)) { if (fdst->threadQuit || fdst->threadErr) { if (nbytes) { - virReportSystemError(EBADF, "%s", - _("stream is not open")); + /* virStreamRecv will virResetLastError possibly set + * by virFDStreamEvent */ + if (fdst->threadErr && !virGetLastError()) + virSetError(fdst->threadErr); + else + virReportSystemError(EBADF, "%s", + _("stream is not open")); } else { ret = 0; } @@ -976,8 +986,12 @@ virFDStreamSendHole(virStreamPtr st, * might mess up file position for the thread. */ if (fdst->threadQuit || fdst->threadErr) { - virReportSystemError(EBADF, "%s", - _("stream is not open")); + /* virStreamSendHole will virResetLastError possibly set + * by virFDStreamEvent */ + if (fdst->threadErr && !virGetLastError()) + virSetError(fdst->threadErr); + else + virReportSystemError(EBADF, "%s", _("stream is not open")); goto cleanup; } -- 2.14.4

2 1

[libvirt] [PATCH 0/2] qemu: Fix memleak in qemuBuildNumaArgStr
by Michal Privoznik 06 Jun '18

06 Jun '18

*** BLURB HERE *** Michal Privoznik (2): qemu: Fix memleak in qemuBuildNumaArgStr qemuBuildNumaArgStr: Simplify @nodeBackends src/qemu/qemu_command.c | 11 ++++------- 1 file changed, 4 insertions(+), 7 deletions(-) -- 2.16.4

2 3

[libvirt] [PATCH] virQEMUCapsFreeHostCPUModel: Don't always free host cpuData
by Michal Privoznik 06 Jun '18

06 Jun '18

This function exists because of 5276ec712a44b36. But it is missing initial check just like virQEMUCapsInitHostCPUModel() has. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- src/qemu/qemu_capabilities.c | 4 ++++ src/qemu/qemu_capspriv.h | 1 + tests/qemuxml2argvtest.c | 6 ++++-- 3 files changed, 9 insertions(+), 2 deletions(-) diff --git a/src/qemu/qemu_capabilities.c b/src/qemu/qemu_capabilities.c index 26969ed3dc..949c32ddd2 100644 --- a/src/qemu/qemu_capabilities.c +++ b/src/qemu/qemu_capabilities.c @@ -2846,10 +2846,14 @@ virQEMUCapsNewHostCPUModel(void) void virQEMUCapsFreeHostCPUModel(virQEMUCapsPtr qemuCaps, + virArch hostArch, virDomainVirtType type) { virQEMUCapsHostCPUDataPtr cpuData = virQEMUCapsGetHostCPUData(qemuCaps, type); + if (!virQEMUCapsGuestIsNative(hostArch, qemuCaps->arch)) + return; + virQEMUCapsHostCPUDataClearModels(cpuData); } diff --git a/src/qemu/qemu_capspriv.h b/src/qemu/qemu_capspriv.h index fea039ef3a..989d183c6c 100644 --- a/src/qemu/qemu_capspriv.h +++ b/src/qemu/qemu_capspriv.h @@ -58,6 +58,7 @@ virQEMUCapsSetArch(virQEMUCapsPtr qemuCaps, void virQEMUCapsFreeHostCPUModel(virQEMUCapsPtr qemuCaps, + virArch hostArch, virDomainVirtType type); void diff --git a/tests/qemuxml2argvtest.c b/tests/qemuxml2argvtest.c index 14a994523f..4984028d38 100644 --- a/tests/qemuxml2argvtest.c +++ b/tests/qemuxml2argvtest.c @@ -388,8 +388,10 @@ testUpdateQEMUCaps(const struct testInfo *info, if (testAddCPUModels(info->qemuCaps, info->skipLegacyCPUs) < 0) goto cleanup; - virQEMUCapsFreeHostCPUModel(info->qemuCaps, VIR_DOMAIN_VIRT_KVM); - virQEMUCapsFreeHostCPUModel(info->qemuCaps, VIR_DOMAIN_VIRT_QEMU); + virQEMUCapsFreeHostCPUModel(info->qemuCaps, caps->host.arch, + VIR_DOMAIN_VIRT_KVM); + virQEMUCapsFreeHostCPUModel(info->qemuCaps, caps->host.arch, + VIR_DOMAIN_VIRT_QEMU); virQEMUCapsInitHostCPUModel(info->qemuCaps, caps->host.arch, VIR_DOMAIN_VIRT_KVM); -- 2.16.4

2 1

[libvirt] [ v3 0/4] Introduce network-backed loader & NVRAM.
by Prerna Saxena 06 Jun '18

06 Jun '18

Libvirt domain XML allows only local filepaths to specify a loader element or its matching NVRAM. Given that VMs may themselves move across hypervisor hosts, it should be possible to allocate loaders/NVRAM disks on network storage for uninterrupted access. This series extends the loader & NVRAM disk elements to be described as virStorageSource* elements, as discussed in : https://www.redhat.com/archives/libvir-list/2018-March/msg01721.html Sample XML with new annotation: <loader readonly='yes' type='pflash' backing='file'> <source file='/usr/share/OVMF/OVMF_CODE.fd'/> </loader> <nvram backing='network'> <source protocol='iscsi' name='iqn.2013-07.com.example:iscsi-nopool/0'> <host name='example.com' port='6000'/> </source> </nvram> References: ---------- v0/ Proposal: https://www.redhat.com/archives/libvir-list/2018-March/msg01721.html.v1 v1: https://www.redhat.com/archives/libvir-list/2018-April/msg02024.html v2: https://www.redhat.com/archives/libvir-list/2018-May/msg00948.html Changelog: --------- Changes since v2: - Consolidated patches with related data structures to avoid build breakage. - Passes make check & make syntax-check. Prerna Saxena (4): Schema: Introduce XML schema for network-backed loader and nvram elements. Loader: Add a more elaborate definition. Test: Add a test snippet to evaluate command line generation for loader/nvram specified via virStorageSource Documentation: Add a blurb for the newly added XML snippets for loader and nvram. docs/formatdomain.html.in | 36 +++- docs/schemas/domaincommon.rng | 108 +++++++++-- src/bhyve/bhyve_command.c | 6 +- src/conf/domain_conf.c | 250 +++++++++++++++++++++++-- src/conf/domain_conf.h | 11 +- src/qemu/qemu_cgroup.c | 13 +- src/qemu/qemu_command.c | 21 ++- src/qemu/qemu_domain.c | 31 ++- src/qemu/qemu_driver.c | 7 +- src/qemu/qemu_parse_command.c | 30 ++- src/qemu/qemu_process.c | 54 ++++-- src/security/security_dac.c | 6 +- src/security/security_selinux.c | 6 +- src/security/virt-aa-helper.c | 14 +- src/vbox/vbox_common.c | 11 +- src/xenapi/xenapi_driver.c | 4 +- src/xenconfig/xen_sxpr.c | 19 +- src/xenconfig/xen_xm.c | 9 +- tests/qemuxml2argvdata/bios-nvram-network.args | 31 +++ tests/qemuxml2argvdata/bios-nvram-network.xml | 42 +++++ tests/qemuxml2argvtest.c | 1 + 21 files changed, 606 insertions(+), 104 deletions(-) create mode 100644 tests/qemuxml2argvdata/bios-nvram-network.args create mode 100644 tests/qemuxml2argvdata/bios-nvram-network.xml -- 1.8.1.2

3 8

[libvirt] [PATCH] qemu: Honour <on_reboot/>
by Michal Privoznik 06 Jun '18

06 Jun '18

https://bugzilla.redhat.com/show_bug.cgi?id=1476866 For some reason, we completely ignore <on_reboot/> setting for domains. The implementation is simply not there. It never was. However, things are slightly more complicated. QEMU sends us two RESET events on domain reboot. Fortunately, the event contains this 'guest' field telling us who initiated the reboot. And since we don't want to destroy the domain if the reset is initiated by a user, we have to ignore those events. Whatever, just look at the code. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- src/qemu/qemu_domain.h | 1 + src/qemu/qemu_monitor.c | 4 ++-- src/qemu/qemu_monitor.h | 3 ++- src/qemu/qemu_monitor_json.c | 8 +++++++- src/qemu/qemu_process.c | 34 ++++++++++++++++++++++++++++++---- 5 files changed, 42 insertions(+), 8 deletions(-) diff --git a/src/qemu/qemu_domain.h b/src/qemu/qemu_domain.h index 4c9050aff..d865e67c7 100644 --- a/src/qemu/qemu_domain.h +++ b/src/qemu/qemu_domain.h @@ -233,6 +233,7 @@ struct _qemuDomainObjPrivate { bool agentError; bool gotShutdown; + bool gotReset; bool beingDestroyed; char *pidfile; diff --git a/src/qemu/qemu_monitor.c b/src/qemu/qemu_monitor.c index 19082d8bf..8f81a2b28 100644 --- a/src/qemu/qemu_monitor.c +++ b/src/qemu/qemu_monitor.c @@ -1344,12 +1344,12 @@ qemuMonitorEmitShutdown(qemuMonitorPtr mon, virTristateBool guest) int -qemuMonitorEmitReset(qemuMonitorPtr mon) +qemuMonitorEmitReset(qemuMonitorPtr mon, virTristateBool guest) { int ret = -1; VIR_DEBUG("mon=%p", mon); - QEMU_MONITOR_CALLBACK(mon, ret, domainReset, mon->vm); + QEMU_MONITOR_CALLBACK(mon, ret, domainReset, mon->vm, guest); return ret; } diff --git a/src/qemu/qemu_monitor.h b/src/qemu/qemu_monitor.h index 31f7e97ba..8c33f6783 100644 --- a/src/qemu/qemu_monitor.h +++ b/src/qemu/qemu_monitor.h @@ -134,6 +134,7 @@ typedef int (*qemuMonitorDomainShutdownCallback)(qemuMonitorPtr mon, void *opaque); typedef int (*qemuMonitorDomainResetCallback)(qemuMonitorPtr mon, virDomainObjPtr vm, + virTristateBool guest, void *opaque); typedef int (*qemuMonitorDomainPowerdownCallback)(qemuMonitorPtr mon, virDomainObjPtr vm, @@ -346,7 +347,7 @@ int qemuMonitorEmitEvent(qemuMonitorPtr mon, const char *event, long long seconds, unsigned int micros, const char *details); int qemuMonitorEmitShutdown(qemuMonitorPtr mon, virTristateBool guest); -int qemuMonitorEmitReset(qemuMonitorPtr mon); +int qemuMonitorEmitReset(qemuMonitorPtr mon, virTristateBool guest); int qemuMonitorEmitPowerdown(qemuMonitorPtr mon); int qemuMonitorEmitStop(qemuMonitorPtr mon); int qemuMonitorEmitResume(qemuMonitorPtr mon); diff --git a/src/qemu/qemu_monitor_json.c b/src/qemu/qemu_monitor_json.c index b8a68154a..8a1501ced 100644 --- a/src/qemu/qemu_monitor_json.c +++ b/src/qemu/qemu_monitor_json.c @@ -536,7 +536,13 @@ static void qemuMonitorJSONHandleShutdown(qemuMonitorPtr mon, virJSONValuePtr da static void qemuMonitorJSONHandleReset(qemuMonitorPtr mon, virJSONValuePtr data ATTRIBUTE_UNUSED) { - qemuMonitorEmitReset(mon); + bool guest = false; + virTristateBool guest_initiated = VIR_TRISTATE_BOOL_ABSENT; + + if (data && virJSONValueObjectGetBoolean(data, "guest", &guest) == 0) + guest_initiated = guest ? VIR_TRISTATE_BOOL_YES : VIR_TRISTATE_BOOL_NO; + + qemuMonitorEmitReset(mon, guest_initiated); } static void qemuMonitorJSONHandlePowerdown(qemuMonitorPtr mon, virJSONValuePtr data ATTRIBUTE_UNUSED) diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index 0aecce3b1..889efc7f0 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -478,27 +478,51 @@ qemuProcessFindVolumeQcowPassphrase(qemuMonitorPtr mon ATTRIBUTE_UNUSED, static int qemuProcessHandleReset(qemuMonitorPtr mon ATTRIBUTE_UNUSED, virDomainObjPtr vm, + virTristateBool guest_initiated, void *opaque) { virQEMUDriverPtr driver = opaque; - virObjectEventPtr event; + virObjectEventPtr event = NULL; qemuDomainObjPrivatePtr priv; virQEMUDriverConfigPtr cfg = virQEMUDriverGetConfig(driver); + bool callOnReboot = false; virObjectLock(vm); + priv = vm->privateData; + + /* This is a bit tricky. When a guest does 'reboot' we receive RESET event + * twice, both times it's guest initiated. However, if users call 'virsh + * reset' we still receive two events but the first one is guest_initiated + * = no, the second one is guest_initiated = yes. Therefore, to avoid + * executing onReboot action in the latter case we need this complicated + * construction. */ + if (guest_initiated == VIR_TRISTATE_BOOL_NO) { + VIR_DEBUG("Ignoring not guest initiated RESET event from domain %s", + vm->def->name); + priv->gotReset = true; + } else if (priv->gotReset && guest_initiated == VIR_TRISTATE_BOOL_YES) { + VIR_DEBUG("Ignoring second RESET event from domain %s", + vm->def->name); + priv->gotReset = false; + } else { + callOnReboot = true; + } + event = virDomainEventRebootNewFromObj(vm); - priv = vm->privateData; if (priv->agent) qemuAgentNotifyEvent(priv->agent, QEMU_AGENT_EVENT_RESET); if (virDomainSaveStatus(driver->xmlopt, cfg->stateDir, vm, driver->caps) < 0) VIR_WARN("Failed to save status on vm %s", vm->def->name); + if (callOnReboot && + guest_initiated == VIR_TRISTATE_BOOL_YES && + vm->def->onReboot == VIR_DOMAIN_LIFECYCLE_DESTROY) + qemuProcessShutdownOrReboot(driver, vm); + virObjectUnlock(vm); - qemuDomainEventQueue(driver, event); - virObjectUnref(cfg); return 0; } @@ -555,6 +579,7 @@ qemuProcessFakeReboot(void *opaque) goto endjob; } priv->gotShutdown = false; + priv->gotReset = false; event = virDomainEventLifecycleNewFromObj(vm, VIR_DOMAIN_EVENT_RESUMED, VIR_DOMAIN_EVENT_RESUMED_UNPAUSED); @@ -5320,6 +5345,7 @@ qemuProcessPrepareDomain(virConnectPtr conn, priv->monError = false; priv->monStart = 0; priv->gotShutdown = false; + priv->gotReset = false; VIR_DEBUG("Updating guest CPU definition"); if (qemuProcessUpdateGuestCPU(vm->def, priv->qemuCaps, caps, flags) < 0) -- 2.13.0

4 6

[libvirt] [RFC PATCH 0/6] qemu: Support pagesize tuning for pSeries guests
by Andrea Bolognani 06 Jun '18

06 Jun '18

The QEMU part, which is RFC as well, can be found at http://lists.nongnu.org/archive/html/qemu-devel/2018-04/msg02818.html Applies cleanly on top of c49013f26c6b40b741f4d5fc61269898f7fd25b8. Andrea Bolognani (6): conf: Reintroduce virDomainDef::hpt_resizing conf: Tweak HPT parsing and formatting qemu: Introduce QEMU_CAPS_MACHINE_PSERIES_CAP_HPT_MPS tests: Pretend we have pseries.cap-hpt-mps in 2.12 conf: Parse and format HPT maxpagesize qemu: Format pseries.cap-hpt-mps on the command line docs/schemas/domaincommon.rng | 21 +- src/conf/domain_conf.c | 72 ++++++- src/conf/domain_conf.h | 2 + src/qemu/qemu_capabilities.c | 8 + src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_command.c | 57 ++++-- src/qemu/qemu_domain.c | 2 +- .../caps_2.12.0.aarch64.replies | 24 ++- .../caps_2.12.0.aarch64.xml | 2 +- .../caps_2.12.0.ppc64.replies | 180 +++++++++++++++++- .../caps_2.12.0.ppc64.xml | 3 +- .../caps_2.12.0.s390x.replies | 26 ++- .../caps_2.12.0.s390x.xml | 2 +- .../caps_2.12.0.x86_64.replies | 30 +-- .../caps_2.12.0.x86_64.xml | 2 +- tests/qemuxml2argvdata/pseries-features.args | 3 +- tests/qemuxml2argvdata/pseries-features.xml | 18 +- tests/qemuxml2argvtest.c | 1 + tests/qemuxml2xmloutdata/pseries-features.xml | 31 ++- tests/qemuxml2xmltest.c | 1 + 20 files changed, 401 insertions(+), 85 deletions(-) mode change 120000 => 100644 tests/qemuxml2xmloutdata/pseries-features.xml -- 2.17.0

3 22

[libvirt] [PATCH] qemu: fix msg could be a wild pointer in qemuMonitorIOProcess()
by Shannon Zhao 06 Jun '18

06 Jun '18

From: Weilun Zhu <zhuweilun(a)huawei.com> As qemuMonitorJSONIOProcess() will unlock the qemu monitor, there is some extreme situation, eg qemu send message to monitor twice in a short time, where the local viriable 'msg' of qemuMonitorIOProcess() could be a wild point: 1. qemuMonitorSend() assign mon->msg to parameter 'msg', which is alse a local variable of its caller qemuMonitorJSONCommandWithFd(), cause eventloop to send message to monitor, then wait condition. 2. qemu send message to monitor for the first time immediately. 3. qemuMonitorIOProcess() is called, then wake up the qemuMonitorSend() thread, but the qemuMonitorSend() thread stuck for a while, which means the qemu monitor is still unlocked. 4. qemu send message to monitor for the second time, such as RTC_CHANGE event 5. qemuMonitorIOProcess() is called, the local viriable 'msg' is assigned to mon->msg. 6. qemuMonitorIOProcess() call qemuMonitorJSONIOProcess() to deal with the message 7. qemuMonitorJSONIOProcess() unlock the qemu monitor, qemuMonitorSend() thread get the lock and free the mon->msg, assign mon->msg to NULL. so the local viriable 'msg' of qemuMonitorIOProcess() is a wild pointer now. AFAIK, it is not harmful to call again virCondBroadcast() while msg is a wild pointer, but just in case, we fix it in this patch. --- src/qemu/qemu_monitor.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/qemu/qemu_monitor.c b/src/qemu/qemu_monitor.c index 43f1d2f..464f200 100644 --- a/src/qemu/qemu_monitor.c +++ b/src/qemu/qemu_monitor.c @@ -454,7 +454,7 @@ qemuMonitorIOProcess(qemuMonitorPtr mon) #if DEBUG_IO VIR_DEBUG("Process done %d used %d", (int)mon->bufferOffset, len); #endif - if (msg && msg->finished) + if (msg && msg == mon->msg && msg->finished) virCondBroadcast(&mon->notify); return len; } -- 1.8.3.1

3 3

[libvirt] [PATCH] docs: document mandatory signoffs in governance.html
by Ján Tomko 06 Jun '18

06 Jun '18

Amend the paragraphs about no CLAs and implicit license agreements to mention mandatory Signed-off-by tags. Signed-off-by: Ján Tomko <jtomko(a)redhat.com> --- This is duplicating information, but I did not find a reasonable way to link the numbered list in hacking.html without making it look ugly. Very probably due to my lack of trying. docs/governance.html.in | 12 ++++++++---- 1 file changed, 8 insertions(+), 4 deletions(-) diff --git a/docs/governance.html.in b/docs/governance.html.in index df5999c0f2..e9608f673c 100644 --- a/docs/governance.html.in +++ b/docs/governance.html.in @@ -141,13 +141,17 @@ than having the interest and ability to provide a contribution. The libvirt project does not require any "Contributor License Agreement" - to be signed prior to engagement with the community. + to be signed prior to engagement with the community. However for + contributing patches, providing a 'Signed-off-by' line with the + author's legal name and e-mail address to demonstrate agreement + and compliance with the <a href="https://developercertificate.org/"> + Developer Certificate of Origin</a> is required. - In making a contribution to the project, the community member is - implicitly stating that they accept the terms of the license under - which the work they are contributing to is distributed. They are + In making a non-patch contribution to the project, the community + member is implicitly stating that they accept the terms of the license + under which the work they are contributing to is distributed. They are also implicitly stating that they have the legal right to make the contribution, if doing so on behalf of a broader organization / company. Most of the project's code is distributed under the GNU -- 2.13.6

2 1

[libvirt] [PATCH] qemu: Fix double free in qemuDomainSecretAESClear
by John Ferlan 06 Jun '18

06 Jun '18

Commit id 02b031a4 added a secondary path from which the incoming @secinfo would not be free'd until the private data was freed in qemuDomainStorageSourcePrivateDispose. However, by doing this the original intention to free @*secinfo afterwards is lost and thus the pass by value of the secinfo->s.aes (or secinfo->s.plain for its method) results in not keeping the NULL setting in the various secret.{username|iv|ciphertext} fields upon return to qemuDomainSecretInfoClear and eventually will result in a double free at domain destroy: raise () abort () __libc_message () malloc_printerr () _int_free () virFree qemuDomainSecretAESClear qemuDomainSecretInfoClear qemuDomainSecretInfoFree qemuDomainStorageSourcePrivateDispose virObjectUnref virStorageSourceClear virStorageSourceFree virDomainDiskDefFree virDomainDefFree virDomainObjRemoveTransientDef qemuProcessStop qemuDomainDestroyFlags virDomainDestroy Signed-off-by: John Ferlan <jferlan(a)redhat.com> --- src/qemu/qemu_domain.c | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-) Domains w/ secrets weren't very happy when I went to destroy them today during testing... Fortunately issue is not in 4.4.0... I modified both Plain and AES just because it's probably best to avoid something like this in the future. diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c index f135117a95..1fb1ef1deb 100644 --- a/src/qemu/qemu_domain.c +++ b/src/qemu/qemu_domain.c @@ -945,23 +945,23 @@ qemuDomainMasterKeyCreate(virDomainObjPtr vm) static void -qemuDomainSecretPlainClear(qemuDomainSecretPlain secret) +qemuDomainSecretPlainClear(qemuDomainSecretPlainPtr secret) { - VIR_FREE(secret.username); - VIR_DISPOSE_N(secret.secret, secret.secretlen); + VIR_FREE(secret->username); + VIR_DISPOSE_N(secret->secret, secret->secretlen); } static void -qemuDomainSecretAESClear(qemuDomainSecretAES secret, +qemuDomainSecretAESClear(qemuDomainSecretAESPtr secret, bool keepAlias) { if (!keepAlias) - VIR_FREE(secret.alias); + VIR_FREE(secret->alias); - VIR_FREE(secret.username); - VIR_FREE(secret.iv); - VIR_FREE(secret.ciphertext); + VIR_FREE(secret->username); + VIR_FREE(secret->iv); + VIR_FREE(secret->ciphertext); } @@ -974,11 +974,11 @@ qemuDomainSecretInfoClear(qemuDomainSecretInfoPtr secinfo, switch ((qemuDomainSecretInfoType) secinfo->type) { case VIR_DOMAIN_SECRET_INFO_TYPE_PLAIN: - qemuDomainSecretPlainClear(secinfo->s.plain); + qemuDomainSecretPlainClear(&secinfo->s.plain); break; case VIR_DOMAIN_SECRET_INFO_TYPE_AES: - qemuDomainSecretAESClear(secinfo->s.aes, keepAlias); + qemuDomainSecretAESClear(&secinfo->s.aes, keepAlias); break; case VIR_DOMAIN_SECRET_INFO_TYPE_LAST: -- 2.14.4

2 1

[libvirt] [PATCH v2 0/2] nwfilter: fix IP address learning
by Daniel P. Berrangé 05 Jun '18

05 Jun '18

Changed in v2: - Use if() instead of switch() Daniel P. Berrangé (2): nwfilter: fix IP address learning nwfilter: directly use poll to wait for packets instead of pcap_next src/nwfilter/nwfilter_learnipaddr.c | 55 ++++++++++++++++++++++------- src/nwfilter/nwfilter_learnipaddr.h | 2 +- 2 files changed, 43 insertions(+), 14 deletions(-) -- 2.17.0

2 3

[libvirt] [perl PATCH 0/6] Update RPM spec to match latest Fedora
by Daniel P. Berrangé 05 Jun '18

05 Jun '18

Syncing changes back from Fedora to follow best practice Daniel P. Berrangé (6): Remove obsolete RPM group tag Replace search.cpan.org with metacpan.org Remove commented out cruft leftover from Module::Build conversion Remove obsolete %defattr statement Use %license for LICENSE file instead of %doc Modernize list of build deps to match actual required packages perl-Sys-Virt.spec.PL | 45 ++++++++++++++++++++++++++++--------------- 1 file changed, 30 insertions(+), 15 deletions(-) -- 2.17.0

2 12

[libvirt] [PATCH v2 0/3] adding virGetLastErrorCode/Domain to paritally replace virGetLastError
by ramyelkest 05 Jun '18

05 Jun '18

Changes from v1[1]: * removed virHasLastError() and s/virHasLastError/virGetLastErrorCode/g * replaced in missed files: virmodule.c and virnetlibsshsession.c * better split of patches [1] https://www.redhat.com/archives/libvir-list/2018-May/msg00259.html ramyelkest (3): util: cleanup: using virGetLastErrorMessage instead of err->message util: added virGetLastErrorCode/Domain all: replacing virGetLastError with virGetLastErrorCode where we can include/libvirt/virterror.h | 2 ++ src/libvirt_public.syms | 6 ++++++ src/locking/lock_driver_lockd.c | 3 +-- src/lxc/lxc_controller.c | 4 +--- src/qemu/qemu_agent.c | 3 +-- src/qemu/qemu_conf.c | 3 +-- src/qemu/qemu_domain.c | 2 +- src/qemu/qemu_driver.c | 12 ++++++------ src/qemu/qemu_hotplug.c | 2 +- src/qemu/qemu_migration.c | 4 ++-- src/qemu/qemu_monitor.c | 5 ++--- src/qemu/qemu_monitor_json.c | 2 +- src/qemu/qemu_process.c | 4 ++-- src/remote/remote_driver.c | 3 +-- src/rpc/virnetclient.c | 2 +- src/rpc/virnetlibsshsession.c | 4 +--- src/util/virerror.c | 42 +++++++++++++++++++++++++++++++++++++++++ src/util/virfilecache.c | 3 +-- src/util/virmodule.c | 3 +-- src/util/virxml.c | 4 ++-- tests/commandtest.c | 2 +- tests/testutils.c | 6 ++---- tests/virhostcputest.c | 2 +- tests/virstoragetest.c | 8 ++++---- tools/virsh-domain-monitor.c | 7 +++---- tools/virsh-domain.c | 4 +--- tools/virsh-util.c | 3 +-- tools/vsh.c | 2 +- 28 files changed, 90 insertions(+), 57 deletions(-) -- 2.7.4

3 8

[libvirt] [PATCH] tests: use a fixed NBD TLS certificate path
by Daniel P. Berrangé 05 Jun '18

05 Jun '18

The default NBD TLS certificate path varies based on prefix given to configure, causing tests to fail depending on build options. Signed-off-by: Daniel P. Berrangé <berrange(a)redhat.com> --- Pushed as a build fix for CI tests/qemuxml2argvdata/disk-drive-network-tlsx509.args | 4 ++-- tests/qemuxml2argvtest.c | 4 ++++ 2 files changed, 6 insertions(+), 2 deletions(-) diff --git a/tests/qemuxml2argvdata/disk-drive-network-tlsx509.args b/tests/qemuxml2argvdata/disk-drive-network-tlsx509.args index 970b8a32a6..e25f45742c 100644 --- a/tests/qemuxml2argvdata/disk-drive-network-tlsx509.args +++ b/tests/qemuxml2argvdata/disk-drive-network-tlsx509.args @@ -44,8 +44,8 @@ file.server.host=192.168.0.3,file.server.port=9999,format=raw,if=none,\ id=drive-virtio-disk2,serial=eb90327c-8302-4725-9e1b-4e85ed4dc252,cache=none \ -device virtio-blk-pci,bus=pci.0,addr=0x6,drive=drive-virtio-disk2,\ id=virtio-disk2 \ --object tls-creds-x509,id=objvirtio-disk3_tls0,dir=/etc/pki/qemu,\ -endpoint=client,verify-peer=yes \ +-object tls-creds-x509,id=objvirtio-disk3_tls0,dir=/etc/pki/libvirt-nbd/dummy,,\ +path,endpoint=client,verify-peer=yes \ -drive file.driver=nbd,file.server.type=inet,file.server.host=example.com,\ file.server.port=1234,file.tls-creds=objvirtio-disk3_tls0,format=raw,if=none,\ id=drive-virtio-disk3,cache=none \ diff --git a/tests/qemuxml2argvtest.c b/tests/qemuxml2argvtest.c index 36bff26d33..8e5504f845 100644 --- a/tests/qemuxml2argvtest.c +++ b/tests/qemuxml2argvtest.c @@ -655,6 +655,9 @@ mymain(void) VIR_FREE(driver.config->vxhsTLSx509certdir); if (VIR_STRDUP_QUIET(driver.config->vxhsTLSx509certdir, "/etc/pki/libvirt-vxhs/dummy,path") < 0) return EXIT_FAILURE; + VIR_FREE(driver.config->nbdTLSx509certdir); + if (VIR_STRDUP_QUIET(driver.config->nbdTLSx509certdir, "/etc/pki/libvirt-nbd/dummy,path") < 0) + return EXIT_FAILURE; VIR_FREE(driver.config->hugetlbfs); if (VIR_ALLOC_N(driver.config->hugetlbfs, 2) < 0) @@ -2860,6 +2863,7 @@ mymain(void) if (getenv("LIBVIRT_SKIP_CLEANUP") == NULL) virFileDeleteTree(fakerootdir); + VIR_FREE(driver.config->nbdTLSx509certdir); qemuTestDriverFree(&driver); VIR_FREE(fakerootdir); VIR_FREE(capslatest_x86_64); -- 2.17.0

1 0

[libvirt] libvirt default machine-type guarantees? (was Re: [Qemu-devel] [PATCH RFC] hw/pc: set q35 as the default x86 machine)
by Eduardo Habkost 05 Jun '18

05 Jun '18

(CCing libvir-list) On Tue, Jun 05, 2018 at 09:43:00AM +0100, Daniel P. Berrangé wrote: > On Tue, Jun 05, 2018 at 09:27:46AM +0200, Gerd Hoffmann wrote: > > Hi, > > > > > > Add to that shortcuts like -cdrom > > > > stop working, > > > > > > Maybe is fixable. > > > > Already fixed for ages. > > > > > I see marking Q35 as the default machine a first step. > > > > Maybe the better option is to go the arm route: Just don't define a > > default, so users have to specify pc or q35. That will make them notice > > there is a world beside 'pc', and we also avoid breaking things > > silently. > > If QEMU removes the default, then libvirt will have to hardcode > 'pc' as the default to maintain back compatibility, so I don't > think that ends up as a net win Is there an actual promise to never change the default machine-type documented in the libvirt API, or is this just fear of breaking existing code? -- Eduardo

4 10

[libvirt] [PATCH v3 0/4] qemu: use FD passing for chardev UNIX sockets
by Daniel P. Berrangé 05 Jun '18

05 Jun '18

This series makes use of the chardev fd passing arriving in QEMU 2.12 to get rid of the startup race wrt opening the QEMU monitor. It is actually enabled in all chardev UNIX sockets for sake of having the same codepath everywhere, but is only important for the monitor socket. Changed in v3: - Refactor UNIX socket opening code to allow it to be mocked in the unit tests to avoid creating real UNIX sockets Daniel P. Berrangé (4): qemu: probe for -chardev 'fd' parameter for FD passing qemu: support passing pre-opened UNIX socket listen FD qemu: don't retry connect() if doing FD passing qemu: remove pointless connect retry logic in agent src/qemu/qemu_agent.c | 84 ++----------------- src/qemu/qemu_capabilities.c | 4 +- src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_command.c | 64 +++++++++++++- src/qemu/qemu_command.h | 4 + src/qemu/qemu_monitor.c | 54 +++++++----- src/qemu/qemu_monitor.h | 1 + src/qemu/qemu_process.c | 27 ++++-- .../caps_2.12.0.aarch64.xml | 1 + .../caps_2.12.0.ppc64.xml | 1 + .../caps_2.12.0.s390x.xml | 1 + .../caps_2.12.0.x86_64.xml | 1 + tests/qemumonitortestutils.c | 1 + .../disk-drive-write-cache.x86_64-latest.args | 3 +- ...irtio-scsi-reservations.x86_64-latest.args | 3 +- tests/qemuxml2argvmock.c | 16 ++++ 16 files changed, 153 insertions(+), 113 deletions(-) -- 2.17.0

3 7

[libvirt] [PATCH 0/3] travis: Stop performing Ubuntu builds
by Andrea Bolognani 05 Jun '18

05 Jun '18

They're all failing at the moment anyway. We'll get the Ubuntu coverage back as soon as we add Ubuntu builders to the CentOS CI environment. Andrea Bolognani (3): travis: Stop performing Ubuntu builds travis: Remove matrix travis: Minor style tweaks .travis.yml | 118 ++++++++-------------------------------------------- 1 file changed, 17 insertions(+), 101 deletions(-) -- 2.17.1

2 4

[libvirt] [PATCH] qemu: Fix detaching from persistent def in qemuDomainDetachDeviceAliasLiveAndConfig
by Peter Krempa 05 Jun '18

05 Jun '18

The code that detaches the device from persistent definition copies the persistent definition first so that it can easily be rolled back. The actual detaching is then made in the copy which is assigned back on success (if the live operation succeeded as well). This is not the case in qemuDomainDetachDeviceAliasLiveAndConfig where the definition was copied and put back, but the detaching happened from the other object which was overwritten. Signed-off-by: Peter Krempa <pkrempa(a)redhat.com> --- src/qemu/qemu_driver.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/src/qemu/qemu_driver.c b/src/qemu/qemu_driver.c index c200c5a63d..b7b318896f 100644 --- a/src/qemu/qemu_driver.c +++ b/src/qemu/qemu_driver.c @@ -8822,14 +8822,13 @@ qemuDomainDetachDeviceAliasLiveAndConfig(virQEMUDriverPtr driver, if (persistentDef) { virDomainDeviceDef dev; - vmdef = virDomainObjCopyPersistentDef(vm, caps, driver->xmlopt); - if (!vmdef) + if (!(vmdef = virDomainObjCopyPersistentDef(vm, caps, driver->xmlopt))) goto cleanup; - if (virDomainDefFindDevice(persistentDef, alias, &dev, true) < 0) + if (virDomainDefFindDevice(vmdef, alias, &dev, true) < 0) goto cleanup; - if (qemuDomainDetachDeviceConfig(persistentDef, &dev, caps, + if (qemuDomainDetachDeviceConfig(vmdef, &dev, caps, parse_flags, driver->xmlopt) < 0) goto cleanup; } -- 2.16.2

2 1

[libvirt] [PATCH] qemu: add entry for balloon stat stat-disk-caches
by Tomáš Golembiovský 05 Jun '18

05 Jun '18

Signed-off-by: Tomáš Golembiovský <tgolembi(a)redhat.com> --- include/libvirt/libvirt-domain.h | 9 ++++++++- src/qemu/qemu_monitor_json.c | 2 ++ 2 files changed, 10 insertions(+), 1 deletion(-) diff --git a/include/libvirt/libvirt-domain.h b/include/libvirt/libvirt-domain.h index da773b76cb..b96c018a90 100644 --- a/include/libvirt/libvirt-domain.h +++ b/include/libvirt/libvirt-domain.h @@ -628,11 +628,18 @@ typedef enum { /* Timestamp of the last update of statistics, in seconds. */ VIR_DOMAIN_MEMORY_STAT_LAST_UPDATE = 9, + /* + * The amount of memory, in bytes, that can be quickly reclaimed without + * additional I/O. Typically these pages are used for caching files from + * disk. + */ + VIR_DOMAIN_MEMORY_STAT_DISK_CACHES = 10, + /* * The number of statistics supported by this version of the interface. * To add new statistics, add them to the enum and increase this value. */ - VIR_DOMAIN_MEMORY_STAT_NR = 10, + VIR_DOMAIN_MEMORY_STAT_NR = 11, # ifdef VIR_ENUM_SENTINELS VIR_DOMAIN_MEMORY_STAT_LAST = VIR_DOMAIN_MEMORY_STAT_NR diff --git a/src/qemu/qemu_monitor_json.c b/src/qemu/qemu_monitor_json.c index 42d7b9c5e9..b0a65d8af9 100644 --- a/src/qemu/qemu_monitor_json.c +++ b/src/qemu/qemu_monitor_json.c @@ -2069,6 +2069,8 @@ int qemuMonitorJSONGetMemoryStats(qemuMonitorPtr mon, VIR_DOMAIN_MEMORY_STAT_AVAILABLE, 1024); GET_BALLOON_STATS(statsdata, "stat-available-memory", VIR_DOMAIN_MEMORY_STAT_USABLE, 1024); + GET_BALLOON_STATS(statsdata, "stat-disk-caches", + VIR_DOMAIN_MEMORY_STAT_DISK_CACHES, 1024); GET_BALLOON_STATS(data, "last-update", VIR_DOMAIN_MEMORY_STAT_LAST_UPDATE, 1); ret = got; -- 2.17.0

3 4

[libvirt] [PATCH v8 00/18] Add support for TPM emulator (for 4.5)
by Stefan Berger 05 Jun '18

05 Jun '18

This series of patches adds support for the TPM emulator backend that is available in QEMU and based on swtpm + libtpms. It allows to attach a TPM 1.2 or 2 to a QEMU VM. sVirt labels are used for labeling the swtpm process, its Unix socket, and log file with the same label that the QEMU process gets. Besides that swtpm is added to the emulator cgroup to restrict its CPU usage. The device XML can be changed from a TPM 1.2 to a TPM 2 and back to a TPM 1.2. The device state is not removed during those changes but only when the domain is undefined. The swtpm needs persistent storage to store its state. For that I am using the uuid of the VM as part of the path since the name of the VM can be changed. Logfiles, PID files, and socket names are based on the name of the VM, though. Stefan v7->v8: - Delaying this series for 4.5; adjusted references to 4.4 - Fixed a test case since version='1.2' is now formatted as well - Appended patches for AppArmor and auditing - Appended patches that improve / fix existing code - patch for validating the TPM configuration rather than overwriting it; a particular case is the CRB interface does not work with a TPM 1.2 - swtpm_setup can be run for a TPM 2 in unprivileged mode as well v6->v7: - followed Jan Tomko's suggestion with resulting changing to patch 10/12. - re-added missing parts related to swtpm_setup and TPM that got lost in v4 v5->v6: - Addressed John Ferlan's comments - rebased on latest tip - Added patch 12. v4->v5: - Addressed John Ferlan's, Boris Fiuczysnki's and Marc Hartmayer's comments - rebased on latest tip v3->v4: - Addressed John Ferlan's comments - Fixed bugs I found while testing - rebased on latest tip Stefan Berger (18): conf: Add support for external swtpm TPM emulator to domain XML qemu: Extend QEMU capabilities with 'tpm-emulator' util: Implement virFileChownFiles() security: Add DAC and SELinux security for tpm-emulator qemu: Extend qemu_conf with tpm-emulator support qemu: Extend QEMU with external TPM support qemu: Add support for external swtpm TPM emulator tests: Add test cases for external swtpm TPM emulator security: Label the external swtpm with SELinux labels conf: Add support for choosing emulation of a TPM 2 qemu: Add swtpm to emulator cgroup news: Update news with new TPM emulator feature security: Add swtpm paths to the domain's AppArmor profile qemu: Run swtpm_setup in unprivileged mode for a TPM 2 qemu: Validate chosen TPM model rather than overwriting it conf: Audit TPM emulator device at domain startup conf: Use resrc=tpm in case of TPM passthrough following docs conf: Use virDomainChrSourceDefClear() rather than VIR_FREE() docs/auditlog.html.in | 2 +- docs/formatdomain.html.in | 43 + docs/news.xml | 13 + docs/schemas/domaincommon.rng | 17 + examples/apparmor/libvirt-qemu | 3 + libvirt.spec.in | 2 + src/conf/domain_audit.c | 20 +- src/conf/domain_conf.c | 49 +- src/conf/domain_conf.h | 15 + src/libvirt_private.syms | 3 + src/qemu/Makefile.inc.am | 10 + src/qemu/libvirtd_qemu.aug | 5 + src/qemu/qemu.conf | 8 + src/qemu/qemu_capabilities.c | 5 + src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_cgroup.c | 36 + src/qemu/qemu_cgroup.h | 2 + src/qemu/qemu_command.c | 34 +- src/qemu/qemu_conf.c | 43 + src/qemu/qemu_conf.h | 6 + src/qemu/qemu_domain.c | 31 +- src/qemu/qemu_extdevice.c | 180 ++++ src/qemu/qemu_extdevice.h | 59 ++ src/qemu/qemu_process.c | 16 + src/qemu/qemu_security.c | 69 ++ src/qemu/qemu_security.h | 11 + src/qemu/qemu_tpm.c | 922 +++++++++++++++++++++ src/qemu/qemu_tpm.h | 56 ++ src/qemu/test_libvirtd_qemu.aug.in | 2 + src/security/security_dac.c | 7 + src/security/security_driver.h | 7 + src/security/security_manager.c | 36 + src/security/security_manager.h | 6 + src/security/security_selinux.c | 172 ++++ src/security/security_stack.c | 40 + src/security/virt-aa-helper.c | 24 + src/util/virfile.c | 55 ++ src/util/virfile.h | 3 + tests/qemucapabilitiesdata/caps_2.11.0.s390x.xml | 1 + tests/qemucapabilitiesdata/caps_2.12.0.aarch64.xml | 1 + tests/qemucapabilitiesdata/caps_2.12.0.ppc64.xml | 1 + tests/qemucapabilitiesdata/caps_2.12.0.s390x.xml | 1 + tests/qemucapabilitiesdata/caps_2.12.0.x86_64.xml | 1 + .../tpm-emulator-tpm2.x86_64-latest.args | 33 + tests/qemuxml2argvdata/tpm-emulator-tpm2.xml | 30 + .../tpm-emulator.x86_64-latest.args | 33 + tests/qemuxml2argvdata/tpm-emulator.xml | 30 + tests/qemuxml2argvtest.c | 16 +- tests/qemuxml2xmloutdata/tpm-emulator-tpm2.xml | 34 + tests/qemuxml2xmloutdata/tpm-emulator.xml | 34 + tests/qemuxml2xmltest.c | 1 + 51 files changed, 2212 insertions(+), 17 deletions(-) create mode 100644 src/qemu/qemu_extdevice.c create mode 100644 src/qemu/qemu_extdevice.h create mode 100644 src/qemu/qemu_tpm.c create mode 100644 src/qemu/qemu_tpm.h create mode 100644 tests/qemuxml2argvdata/tpm-emulator-tpm2.x86_64-latest.args create mode 100644 tests/qemuxml2argvdata/tpm-emulator-tpm2.xml create mode 100644 tests/qemuxml2argvdata/tpm-emulator.x86_64-latest.args create mode 100644 tests/qemuxml2argvdata/tpm-emulator.xml create mode 100644 tests/qemuxml2xmloutdata/tpm-emulator-tpm2.xml create mode 100644 tests/qemuxml2xmloutdata/tpm-emulator.xml -- 2.14.3

3 44

[libvirt] [PATCH 0/8] vsock hotplug
by Ján Tomko 05 Jun '18

05 Jun '18

Followup for: https://bugzilla.redhat.com/show_bug.cgi?id=1291851 Ján Tomko (8): qemu: split out qemuBuildVsockDevStr qemuBuildVsockDevStr: allow passing a fdprefix export virDomainVsockDefFree qemu: export vsock-related functions qemu: implement vsock hotplug conf: introduce virDomainVsockDefEquals qemu: implement vsock hotunplug qemu: implement vsock coldplug/coldunplug src/conf/domain_conf.c | 18 +++++++ src/conf/domain_conf.h | 3 ++ src/libvirt_private.syms | 2 + src/qemu/qemu_alias.c | 2 +- src/qemu/qemu_alias.h | 2 + src/qemu/qemu_command.c | 39 ++++++++++---- src/qemu/qemu_command.h | 8 +++ src/qemu/qemu_driver.c | 36 +++++++++++-- src/qemu/qemu_hotplug.c | 138 ++++++++++++++++++++++++++++++++++++++++++++++- src/qemu/qemu_hotplug.h | 7 +++ src/qemu/qemu_process.c | 2 +- src/qemu/qemu_process.h | 2 + 12 files changed, 241 insertions(+), 18 deletions(-) -- 2.16.1

2 20

[libvirt] [PATCH 0/3] Require GnuTLS
by Michal Privoznik 05 Jun '18

05 Jun '18

As mentioned in discussion to my PRNG patches [1] we use GnuTLS functions widely. Therefore, make GnuTLS required at build time. This enables us to drop most of #ifdef WITH_GNULS we have. Not all of them though because we still want libvirt-setuid-rpc-client.la to build without GnuTLS. I should also mention that surprisingly this breaks travis. This time, it's Ubuntu that lacks new enough GnuTLS and not OSX. But after Peter's patches travis is broken anyway (on GnuTLS). Michal Privoznik (3): configure: Require GnuTLS build: Build gnutls related sources unconditionally src: Drop most of #ifdef WITH_GNUTLS configure.ac | 2 -- m4/virt-gnutls.m4 | 4 --- src/Makefile.am | 7 +---- src/locking/lock_daemon.c | 4 --- src/logging/log_daemon.c | 4 --- src/lxc/lxc_controller.c | 2 -- src/qemu/qemu_migration_cookie.c | 12 +++----- src/remote/remote_daemon.c | 23 --------------- src/remote/remote_daemon_dispatch.c | 2 -- src/rpc/Makefile.inc.am | 14 ++------- src/rpc/virnetdaemon.h | 4 +-- src/rpc/virnetserver.c | 6 ---- src/rpc/virnetserver.h | 6 +--- src/rpc/virnetserverclient.c | 57 +++---------------------------------- src/rpc/virnetserverclient.h | 8 ------ src/rpc/virnetserverservice.c | 24 ---------------- src/rpc/virnetserverservice.h | 10 ------- src/util/vircrypto.c | 43 ++-------------------------- tests/Makefile.am | 12 ++------ tests/qemuxml2argvtest.c | 15 ---------- tests/vircryptotest.c | 24 +++++----------- tests/virfilecachetest.c | 18 +++--------- tests/virnetdaemontest.c | 8 ------ tests/virnetserverclienttest.c | 2 -- tests/virrandommock.c | 8 ++---- 25 files changed, 32 insertions(+), 287 deletions(-) -- 2.16.4

4 10

[libvirt] [jenkins-ci PATCH 0/5] guests: Make Perl and Python paths dynamic
by Andrea Bolognani 05 Jun '18

05 Jun '18

Plus other cleanups. See patches 3/5 and 4/5 for the good stuff. Andrea Bolognani (5): guests: Add .j2 extension to existing templates guests: Remove distinction between files and templates guests: Make PERL5LIB dynamic guests: Make PYTHONPATH dynamic guests: Move remaining paths out of the inventory guests/group_vars/all/main.yml | 13 -------- guests/host_vars/libvirt-centos-6/main.yml | 3 -- guests/host_vars/libvirt-centos-7/main.yml | 3 -- guests/host_vars/libvirt-debian-8/main.yml | 3 -- guests/host_vars/libvirt-debian-9/main.yml | 3 -- guests/host_vars/libvirt-debian-sid/main.yml | 3 -- guests/host_vars/libvirt-fedora-27/main.yml | 3 -- guests/host_vars/libvirt-fedora-28/main.yml | 3 -- .../host_vars/libvirt-fedora-rawhide/main.yml | 3 -- guests/host_vars/libvirt-freebsd-10/main.yml | 3 -- guests/host_vars/libvirt-freebsd-11/main.yml | 3 -- .../libvirt-freebsd-current/main.yml | 3 -- guests/host_vars/libvirt-ubuntu-14/main.yml | 3 -- guests/host_vars/libvirt-ubuntu-16/main.yml | 3 -- guests/host_vars/libvirt-ubuntu-18/main.yml | 3 -- guests/tasks/base.yml | 12 +++---- guests/tasks/users.yml | 6 ++-- .../{bash_profile => bash_profile.j2} | 0 guests/templates/bashrc | 20 ------------ guests/templates/bashrc.j2 | 32 +++++++++++++++++++ .../ccache.conf => templates/ccache.conf.j2} | 0 .../fedora-rawhide-kernel-nodebug.repo.j2} | 0 .../jessie-backports.preferences.j2} | 0 .../jessie-backports.sources.j2} | 0 24 files changed, 41 insertions(+), 84 deletions(-) rename guests/templates/{bash_profile => bash_profile.j2} (100%) delete mode 100644 guests/templates/bashrc create mode 100644 guests/templates/bashrc.j2 rename guests/{files/ccache.conf => templates/ccache.conf.j2} (100%) rename guests/{files/fedora-rawhide-kernel-nodebug.repo => templates/fedora-rawhide-kernel-nodebug.repo.j2} (100%) rename guests/{files/jessie-backports.preferences => templates/jessie-backports.preferences.j2} (100%) rename guests/{files/jessie-backports.sources => templates/jessie-backports.sources.j2} (100%) -- 2.17.0

2 13

[libvirt] [RFC PATCH] travis: require libgnutls28-dev
by Ján Tomko 05 Jun '18

05 Jun '18

Since commit <eee8c85> we require GnuTLS >= 3.2.0. Install the aptly-named libgnutls28-dev package instead of plain libgnutls-dev. Signed-off-by: Ján Tomko <jtomko(a)redhat.com> --- RFC, since I haven't even started to figure out how to test it. Exposed by commit 3bf1959 which led to virCryptoHaveCipher(VIR_CRYPTO_CIPHER_AES256CBC) being called from the test suite, in qemuDomainSupportsEncryptedSecret. The proper fix for non-GnuTLS user is probably disabling the build of QEMU driver without GnuTLS, since capabilities caching won't work properly anyway without the hash functions. .travis.yml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/.travis.yml b/.travis.yml index f4dd0427a9..3f7c67398a 100644 --- a/.travis.yml +++ b/.travis.yml @@ -60,7 +60,7 @@ addons: - libdbus-1-dev - libdevmapper-dev - libfuse-dev - - libgnutls-dev + - libgnutls28-dev - libnetcf-dev - libnl-3-dev - libnl-route-3-dev -- 2.13.6

4 3

[libvirt] [jenkins-ci PATCH] guests: Fix os_version for development branches
by Andrea Bolognani 05 Jun '18

05 Jun '18

We currently know about three development branches: Fedora Rawhide, Debian sid and FreeBSD -CURRENT. For all three, we use the name of the development branch in the guest name; however, the value of os_version is not set consistently, making it awkward to create package mappings specific to these development branches. As an added bonus, this makes the code setting os_name and os_version more straightforward by simplifying conditionals and removing regexp-based string replacement. Signed-off-by: Andrea Bolognani <abologna(a)redhat.com> --- guests/tasks/base.yml | 20 ++++++++++++++------ 1 file changed, 14 insertions(+), 6 deletions(-) diff --git a/guests/tasks/base.yml b/guests/tasks/base.yml index 70d5abc..9f27470 100644 --- a/guests/tasks/base.yml +++ b/guests/tasks/base.yml @@ -2,25 +2,33 @@ - name: Set additional facts (OS) set_fact: os_name: '{{ ansible_distribution }}' + +- name: Set additional facts (OS) + set_fact: os_version: '{{ ansible_distribution_major_version }}' when: - - ansible_distribution != 'FreeBSD' - - ansible_distribution_release != 'Rawhide' + - ansible_distribution_major_version is defined - name: Set additional facts (OS) set_fact: - os_name: '{{ ansible_distribution }}' os_version: Rawhide when: - - ansible_distribution != 'FreeBSD' + - os_name == 'Fedora' - ansible_distribution_release == 'Rawhide' - name: Set additional facts (OS) set_fact: - os_name: '{{ ansible_distribution }}' - os_version: '{{ ansible_distribution_release|regex_replace("^(\d+)\.\d+-.*$", "\1") }}' + os_version: Sid + when: + - os_name == 'Debian' + - '{{ ansible_distribution_major_version|match("^.*/sid$") }}' + +- name: Set additional facts (OS) + set_fact: + os_version: Current when: - ansible_distribution == 'FreeBSD' + - '{{ ansible_distribution_release|match("^.*-CURRENT$") }}' - name: Set additional facts (package format) set_fact: -- 2.17.0

2 3

[libvirt] [PATCHv2] virFileAccessibleAs: Remove redundant forkRet
by Radostin Stoyanov 05 Jun '18

05 Jun '18

The variable forkRet is not used after commit 25f8781 Signed-off-by: Radostin Stoyanov <rstoyanov1(a)gmail.com> --- src/util/virfile.c | 10 ---------- 1 file changed, 10 deletions(-) diff --git a/src/util/virfile.c b/src/util/virfile.c index 523241f64f..a2c4cebb56 100644 --- a/src/util/virfile.c +++ b/src/util/virfile.c @@ -2117,7 +2117,6 @@ virFileAccessibleAs(const char *path, int mode, { pid_t pid = 0; int status, ret = 0; - int forkRet = 0; gid_t *groups; int ngroups; @@ -2152,15 +2151,6 @@ virFileAccessibleAs(const char *path, int mode, return 0; } - /* child. - * Return positive value here. Parent - * will change it to negative one. */ - - if (forkRet < 0) { - ret = errno; - goto childerror; - } - if (virSetUIDGID(uid, gid, groups, ngroups) < 0) { ret = errno; goto childerror; -- 2.17.1

2 1