June 2022 - Devel - libvirt List Archives

[PATCH v4 00/13] Introduce network backed NVRAM
by Peter Krempa 14 Jun '22

14 Jun '22

This version fixes multiple problems which I'd point out when reviewing: - virStorageSource is properly used in all places - logic for initializing NVRAM imgages from template is fixed to avoid touching network backed nvrams - documentation now states the correct version - some cleanups - dropped NEWS entry stashed in a patch with other stuff - fixed/simplified schema Rohit, please give it a try. I didn't yet have time to test this beyond unit tests. This series can also be fetched from my repo: git fetch https://gitlab.com/pipo.sk/libvirt.git network-nvram2 Peter Krempa (9): qemuDomainPrepareStorageSourceBlockdev: Add a variant for custom nodename qemuBuildPflashBlockdevCommandLine: Take virDomainObj instead of private data qemu: Use 'def->os.loader->nvram' directly instead of 'priv->pflash1' qemu: Properly setup the NVRAM virStorageSource qemuProcessReconnect: Don't re-instantiate pflash storage source qemuDomainInitializePflashStorageSource: Properly and fully initialize nvram source qemuFirmwareFillDomain: Don't fill in firmware for network backed nvram conf: Extract formatting of NVRAM out of virDomainLoaderDefFormat virDomainHugepagesFormat: Use virXMLFormatElementEmpty Rohit Kumar (4): conf: Convert def->os.loader->nvram a virStorageSource qemu: validate: Reject virStorageSource features we don't want to support with nvram conf: Add support to parse/format <source> for NVRAM Add unit tests for new specification of nvram. docs/formatdomain.rst | 37 +++++ src/conf/domain_conf.c | 136 ++++++++++++++---- src/conf/domain_conf.h | 3 +- src/conf/schemas/domaincommon.rng | 9 +- src/qemu/qemu_cgroup.c | 2 +- src/qemu/qemu_command.c | 22 +-- src/qemu/qemu_domain.c | 58 +++++--- src/qemu/qemu_domain.h | 9 +- src/qemu/qemu_driver.c | 5 +- src/qemu/qemu_firmware.c | 30 +++- src/qemu/qemu_namespace.c | 6 +- src/qemu/qemu_process.c | 11 +- src/qemu/qemu_validate.c | 115 +++++++++++---- src/security/security_dac.c | 19 +-- src/security/security_selinux.c | 21 +-- src/security/virt-aa-helper.c | 5 +- src/vbox/vbox_common.c | 3 +- .../bios-nvram-file.x86_64-latest.args | 37 +++++ tests/qemuxml2argvdata/bios-nvram-file.xml | 23 +++ .../bios-nvram-network-iscsi.x86_64-4.1.0.err | 1 + ...ios-nvram-network-iscsi.x86_64-latest.args | 38 +++++ .../bios-nvram-network-iscsi.xml | 31 ++++ .../bios-nvram-network-nbd.x86_64-latest.args | 37 +++++ .../bios-nvram-network-nbd.xml | 28 ++++ tests/qemuxml2argvtest.c | 4 + .../bios-nvram-file.x86_64-latest.xml | 39 +++++ ...bios-nvram-network-iscsi.x86_64-latest.xml | 44 ++++++ .../bios-nvram-network-nbd.x86_64-latest.xml | 41 ++++++ tests/qemuxml2xmltest.c | 3 + 29 files changed, 696 insertions(+), 121 deletions(-) create mode 100644 tests/qemuxml2argvdata/bios-nvram-file.x86_64-latest.args create mode 100644 tests/qemuxml2argvdata/bios-nvram-file.xml create mode 100644 tests/qemuxml2argvdata/bios-nvram-network-iscsi.x86_64-4.1.0.err create mode 100644 tests/qemuxml2argvdata/bios-nvram-network-iscsi.x86_64-latest.args create mode 100644 tests/qemuxml2argvdata/bios-nvram-network-iscsi.xml create mode 100644 tests/qemuxml2argvdata/bios-nvram-network-nbd.x86_64-latest.args create mode 100644 tests/qemuxml2argvdata/bios-nvram-network-nbd.xml create mode 100644 tests/qemuxml2xmloutdata/bios-nvram-file.x86_64-latest.xml create mode 100644 tests/qemuxml2xmloutdata/bios-nvram-network-iscsi.x86_64-latest.xml create mode 100644 tests/qemuxml2xmloutdata/bios-nvram-network-nbd.x86_64-latest.xml -- 2.35.3

3 27

[PATCH v3 0/5] Introduce network backed NVRAM
by Rohit Kumar 14 Jun '22

14 Jun '22

Libvirt domain XML currently allows only local filepaths that can be used to specify a NVRAM disk. It should be possible to support NVRAM disks on network storage as it would give flexibility to start the VM on any host without having to worry about where to get the latest nvram image. This series extends the NVRAM element to support hosting over network-backed disks. It achieves this by embedding virStorageSource pointer for nvram into _virDomainLoaderDef. It introduces a 'type' attribute for NVRAM element to specify 'file' vs 'network' backed NVRAM. XML with new annotation: <nvram type='network'> <source protocol='iscsi' name='iqn.2013-07.com.example:iscsi-nopool/0'> <host name='example.com' port='6000'/> <auth username='myname'> <secret type='iscsi' usage='mycluster_myname'/> </auth> </host> </source> </nvram> or <nvram type='network'> <source protocol='nbd' name='bar'> <host name='example.org' port='6000'/> </source> </nvram> or <nvram type='file'> <source file='/var/lib/libvirt/nvram/guest_VARS.fd'/> </nvram> Changes v1->v2: - Split the patch into smaller patches - Added unit test - Updated the doc - Addressed Peter's comment on v1 (https://listman.redhat.com/archives/libvir-list/2022-March/229684.html) Changes v2->v3: - Added authentication with 'iscsi' protocol unit test - Updated the validation logic - Addressed Peter's other comments on v2 patch(https://listman.redhat.com/archives/libvir-list/2022-April/229971.htm… Rohit Kumar (5): Make NVRAM a virStorageSource type. Add support to parse/format/validate virStorageSource type NVRAM Update schema, docs, and validation logic to support network backed NVRAM Add unit tests for network backed NVRAM Add unit test to support new 'file' type NVRAM NEWS.rst | 5 + docs/formatdomain.rst | 34 +++++- src/conf/domain_conf.c | 115 +++++++++++++++--- src/conf/domain_conf.h | 3 +- src/conf/schemas/domaincommon.rng | 21 +++- src/qemu/qemu_cgroup.c | 3 +- src/qemu/qemu_command.c | 2 +- src/qemu/qemu_domain.c | 23 +++- src/qemu/qemu_driver.c | 5 +- src/qemu/qemu_firmware.c | 23 +++- src/qemu/qemu_namespace.c | 5 +- src/qemu/qemu_process.c | 5 +- src/qemu/qemu_validate.c | 71 +++++++++++ src/security/security_dac.c | 6 +- src/security/security_selinux.c | 6 +- src/security/virt-aa-helper.c | 5 +- src/vbox/vbox_common.c | 3 +- .../bios-nvram-file.x86_64-latest.args | 37 ++++++ tests/qemuxml2argvdata/bios-nvram-file.xml | 23 ++++ .../bios-nvram-network-iscsi.x86_64-4.1.0.err | 1 + ...ios-nvram-network-iscsi.x86_64-latest.args | 38 ++++++ .../bios-nvram-network-iscsi.xml | 31 +++++ .../bios-nvram-network-nbd.x86_64-latest.args | 37 ++++++ .../bios-nvram-network-nbd.xml | 28 +++++ tests/qemuxml2argvtest.c | 4 + .../bios-nvram-file.x86_64-latest.xml | 39 ++++++ ...bios-nvram-network-iscsi.x86_64-latest.xml | 44 +++++++ .../bios-nvram-network-nbd.x86_64-latest.xml | 41 +++++++ tests/qemuxml2xmltest.c | 3 + 29 files changed, 618 insertions(+), 43 deletions(-) create mode 100644 tests/qemuxml2argvdata/bios-nvram-file.x86_64-latest.args create mode 100644 tests/qemuxml2argvdata/bios-nvram-file.xml create mode 100644 tests/qemuxml2argvdata/bios-nvram-network-iscsi.x86_64-4.1.0.err create mode 100644 tests/qemuxml2argvdata/bios-nvram-network-iscsi.x86_64-latest.args create mode 100644 tests/qemuxml2argvdata/bios-nvram-network-iscsi.xml create mode 100644 tests/qemuxml2argvdata/bios-nvram-network-nbd.x86_64-latest.args create mode 100644 tests/qemuxml2argvdata/bios-nvram-network-nbd.xml create mode 100644 tests/qemuxml2xmloutdata/bios-nvram-file.x86_64-latest.xml create mode 100644 tests/qemuxml2xmloutdata/bios-nvram-network-iscsi.x86_64-latest.xml create mode 100644 tests/qemuxml2xmloutdata/bios-nvram-network-nbd.x86_64-latest.xml -- 2.25.1

3 15

[libvirt PATCH] Fix incorrect uses of g_clear_pointer() introduced in 8.1.0
by Mark Mielke 14 Jun '22

14 Jun '22

This is a partial revert of 87a43a907f0ad4897a28ad7c216bc70f37270b93. The change to use g_clear_pointer() in more places was accidentally applied to cases involving vir_g_source_unref(). In some cases, the ordering of g_source_destroy() and vir_g_source_unref() was reversed, which resulted in the source being marked as destroyed, after it is already unreferenced. This use-after-free case might work in many cases, but with versions of glibc older than 2.64.0 it may defer unref to run within the main thread to avoid a race condition, which creates a large distance between the g_source_unref() and g_source_destroy(). In some cases, the call to vir_g_source_unref() was replaced with a second call to g_source_destroy(), leading to a memory leak or worse. In our experience, the symptoms were that use of libvirt-python became slower over time, with OpenStack nova-compute initially taking around one second to periodically query the host PCI devices, and within an hour it was taking over a minute to complete the same operation, until it is was eventually running this query back-to-back, resulting in the nova-compute process consuming 100% of one CPU thread, losing its RabbitMQ connection frequently, and showing up as down to the control plane. --- src/qemu/qemu_agent.c | 3 ++- src/qemu/qemu_monitor.c | 3 ++- src/util/vireventglib.c | 12 ++++++++---- 3 files changed, 12 insertions(+), 6 deletions(-) diff --git a/src/qemu/qemu_agent.c b/src/qemu/qemu_agent.c index f57a8d5f25..e6e92c7dc4 100644 --- a/src/qemu/qemu_agent.c +++ b/src/qemu/qemu_agent.c @@ -452,8 +452,9 @@ static void qemuAgentUnregister(qemuAgent *agent) { if (agent->watch) { + g_source_destroy(agent->watch); vir_g_source_unref(agent->watch, agent->context); - g_clear_pointer(&agent->watch, g_source_destroy); + agent->watch = NULL; } } diff --git a/src/qemu/qemu_monitor.c b/src/qemu/qemu_monitor.c index 37bcbde31e..32c993a941 100644 --- a/src/qemu/qemu_monitor.c +++ b/src/qemu/qemu_monitor.c @@ -794,8 +794,9 @@ void qemuMonitorUnregister(qemuMonitor *mon) { if (mon->watch) { + g_source_destroy(mon->watch); vir_g_source_unref(mon->watch, mon->context); - g_clear_pointer(&mon->watch, g_source_destroy); + mon->watch = NULL; } } diff --git a/src/util/vireventglib.c b/src/util/vireventglib.c index fc04d8f712..983787932f 100644 --- a/src/util/vireventglib.c +++ b/src/util/vireventglib.c @@ -228,7 +228,8 @@ virEventGLibHandleUpdate(int watch, VIR_DEBUG("Removed old handle source=%p", data->source); g_source_destroy(data->source); - g_clear_pointer(&data->source, g_source_destroy); + vir_g_source_unref(data->source, NULL); + data->source = NULL; data->events = 0; } @@ -275,8 +276,9 @@ virEventGLibHandleRemove(int watch) data, watch, data->fd); if (data->source != NULL) { + g_source_destroy(data->source); vir_g_source_unref(data->source, NULL); - g_clear_pointer(&data->source, g_source_destroy); + data->source = NULL; data->events = 0; } @@ -417,8 +419,9 @@ virEventGLibTimeoutUpdate(int timer, if (data->source == NULL) goto cleanup; + g_source_destroy(data->source); vir_g_source_unref(data->source, NULL); - g_clear_pointer(&data->source, g_source_destroy); + data->source = NULL; } cleanup: @@ -465,8 +468,9 @@ virEventGLibTimeoutRemove(int timer) data, timer); if (data->source != NULL) { + g_source_destroy(data->source); vir_g_source_unref(data->source, NULL); - g_clear_pointer(&data->source, g_source_destroy); + data->source = NULL; } /* since the actual timeout deletion is done asynchronously, a timeoutUpdate call may -- 2.36.1

2 3

[PATCH 0/2] Two <interface type='direct'/> related fixes
by Michal Privoznik 14 Jun '22

14 Jun '22

Except not really. Only the first one is strictly related to that type of <interface/>. The other one fixes a regression introduced in 8.4.0 and it just so happens that I'm able to reproduce it 100% on my (SRIOV-less) machine. Michal Prívozník (2): virNetDevSaveNetConfig: Pass mode to virFileWriteStr() qemuBuildInterfaceConnect: Initialize @tapfd array src/qemu/qemu_command.c | 2 ++ src/util/virnetdev.c | 2 +- 2 files changed, 3 insertions(+), 1 deletion(-) -- 2.35.1

4 7

[PATCH] spec: Xen arches have changed on Fedora 36+
by Cole Robinson 13 Jun '22

13 Jun '22

Latest fedora 36+ xen builds have dropped i686 and armv7hl builds. Signed-off-by: Cole Robinson <crobinso(a)redhat.com> --- libvirt.spec.in | 3 +++ 1 file changed, 3 insertions(+) diff --git a/libvirt.spec.in b/libvirt.spec.in index 59d2f96709..c4ea02fc8e 100644 --- a/libvirt.spec.in +++ b/libvirt.spec.in @@ -21,6 +21,9 @@ %define arches_systemtap_64bit %{arches_64bit} %define arches_dmidecode %{arches_x86} %define arches_xen %{arches_x86} aarch64 +%if 0%{?fedora} >= 36 +%define arches_xen x86_64 aarch64 +%endif %define arches_vbox %{arches_x86} %define arches_ceph %{arches_64bit} %define arches_zfs %{arches_x86} %{power64} %{arm} -- 2.36.1

2 4

[PATCH] kbase: launch_security_sev: Break up overly long line
by Peter Krempa 13 Jun '22

13 Jun '22

Standard text is aligned to 80 colums in all .rst files. Signed-off-by: Peter Krempa <pkrempa(a)redhat.com> --- Pushed as trivial. docs/kbase/launch_security_sev.rst | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/docs/kbase/launch_security_sev.rst b/docs/kbase/launch_security_sev.rst index 51b3e14dbf..2734832487 100644 --- a/docs/kbase/launch_security_sev.rst +++ b/docs/kbase/launch_security_sev.rst @@ -295,9 +295,9 @@ In order to make virtio devices work, we need to use ``<driver iommu='on'/>`` inside the given device XML element in order to enable DMA API in the virtio driver. -Starting with QEMU 6.0.0 QEMU will set this for us by default. For earlier versions though, you will need to explicitly enable this in the device XML as follows: - -:: +Starting with QEMU 6.0.0 QEMU will set this for us by default. For earlier +versions though, you will need to explicitly enable this in the device XML as +follows:: # virsh edit <domain> <domain> -- 2.36.1

1 0

[libvirt PATCH 0/2] qemu: virtiofs: add --thread-pool-size option
by Ján Tomko 13 Jun '22

13 Jun '22

Ján Tomko (2): conf: virtiofs: add thread_pool element qemu: virtiofs: format --thread-pool-size docs/formatdomain.rst | 6 ++++++ src/conf/domain_conf.c | 14 ++++++++++++++ src/conf/domain_conf.h | 1 + src/conf/schemas/domaincommon.rng | 9 +++++++++ src/qemu/qemu_virtiofs.c | 4 ++++ tests/qemuxml2argvdata/vhost-user-fs-fd-memory.xml | 1 + 6 files changed, 35 insertions(+) -- 2.35.1

2 4

[PATCH] docs: kbase/launch_security_sev: QEMU 6.0+ sets iommu=on for us
by Cole Robinson 13 Jun '22

13 Jun '22

Signed-off-by: Cole Robinson <crobinso(a)redhat.com> --- docs/kbase/launch_security_sev.rst | 2 ++ 1 file changed, 2 insertions(+) diff --git a/docs/kbase/launch_security_sev.rst b/docs/kbase/launch_security_sev.rst index 3ebb01ad80..9f6330a1ca 100644 --- a/docs/kbase/launch_security_sev.rst +++ b/docs/kbase/launch_security_sev.rst @@ -295,6 +295,8 @@ In order to make virtio devices work, we need to use ``<driver iommu='on'/>`` inside the given device XML element in order to enable DMA API in the virtio driver. +QEMU 6.0 and later will `set this by default <https://gitlab.com/qemu-project/qemu/-/commit/9f88a7a3df>`__. For earlier QEMU versions, you will need to explicitly enable this in the device XML: + :: # virsh edit <domain> -- 2.36.1

2 2

[libvirt PATCH] util: Fix error reporting in virProcessSetMaxMemLock
by Jiri Denemark 11 Jun '22

11 Jun '22

Commit v7.1.0-136-g6a6d6bb520 refactored virProcessSetMaxMemLock by moving its part into a new virProcessSetLimit, but lost "return -1" on error; Signed-off-by: Jiri Denemark <jdenemar(a)redhat.com> --- src/util/virprocess.c | 1 + 1 file changed, 1 insertion(+) diff --git a/src/util/virprocess.c b/src/util/virprocess.c index 5ed0b5d0db..013afd91b4 100644 --- a/src/util/virprocess.c +++ b/src/util/virprocess.c @@ -926,6 +926,7 @@ virProcessSetMaxMemLock(pid_t pid, unsigned long long bytes) _("cannot limit locked memory " "of process %lld to %llu"), (long long int)pid, bytes); + return -1; } VIR_DEBUG("Locked memory for process %lld limited to %llu bytes", -- 2.35.1

2 2

[libvirt PATCH 0/3] firmware: Small fixes and cleanups
by Andrea Bolognani 10 Jun '22

10 Jun '22

Andrea Bolognani (3): vmx: Declare support for firmware autoselection conf: Fix virDomainDefOSValidate() qemu: Simplify handling of virTristateBool values src/conf/domain_validate.c | 6 +++--- src/qemu/qemu_firmware.c | 38 ++++++++++++++++---------------------- src/vmx/vmx.c | 1 + 3 files changed, 20 insertions(+), 25 deletions(-) -- 2.35.3

2 4