October 2023 - Devel - libvirt List Archives

ANNOUNCE: Virtualization linting library
by Michal Prívozník 18 Oct '23

18 Oct '23

Users of virtualization may struggle with keeping their guest configuration optimal. And while libvirt is perfect tool for its job, it merely just do whatever it is told. For instance, it does not check whether there's enough of free memory to start a guest. It just starts it and hope for the best. This led management applications to have some code that performs (more or less) similar checks. And this is my attempt to unify that code, move it out of individual projects under this umbrella project. It's named virt-lint. While it's written in Rust, it has C bindings from which bindings to other languages can be written (I've written Golang so far). I've created initial commit here: https://gitlab.com/MichalPrivoznik/virt-lint There is README.md which covers all the basic information. So far, I've written four very basic checks (see src/validators.rs): 1) whether there exists set of NUMA nodes large enough to accommodate given domain, 2) whether there exists set of NUMA nodes with enough of free memory to accommodate given domain, 3) whether the virtualization node has suitable emulator binary for given domain, 4) if domain is of Q35 machine type, whether it has at least one pcie-root-port (for hotplugging). Please keep in mind, these checks are just examples and possibly buggy. Writing proper checks is part of future work, for now I want to gather feedback on overall design, APIs, CLI, etc. I've also built RPMs and copied them over here: https://mprivozn.fedorapeople.org/rpms/virt-lint/ I appreciate any feedback. Michal

1 0

[PATCH v7 0/4] Migration deprecated parts
by Juan Quintela 18 Oct '23

18 Oct '23

Based on: Message-ID: <20231018100651.32674-1-quintela(a)redhat.com> [PULL 00/11] Migration 20231018 patches And here we are, at v7: - drop black line at the end of deprecated.rst - change qemu-iotest output due to warnings for deprecation. The only real change is the output of the qemu-iotest. That is the reason why I maintained the reviewed-by. But will be happy if anyone of the block people ack the changes. Thanks, Juan. On this v6: - Fixed Markus comments - 1st patch is reviewed - dropped the RFC ones. Later, Juan. On this v5: - Rebased on top of last migration pull requesnt: - address markus comments. Basically we recommend always blockdev-mirror + NBD. In deprecated.rst we also put the posiblity of using block-incremental and block, but we state that they are also deprecated. I know, I know, I deprecated them in the following patch. - Dropped the removal of block-migration and block-incremental I am only interested in showing why I want to remove the -b/-i options. Please review. Later, Juan. On this v4: - addressed all markus comments. - rebased on latest. - improve formatting of migration.json - print block migration status when needed. - patches 7-10 are not mean to merge, they just show why we want to deprecate block migration and remove its support. - Patch 7 just drop support for -i/-b and qmp equivalents. - Patch 8 shows all the helpers and convolutions we need to have to support that -i and -d. - patch 9 drops block-incremental migration support. - patch 9 drops block migration support. Please review. Thanks, Juan. On this v3: - Rebase on top of upstream. - Changed v8.1 to 8.2 (I left the reviewed by anyways) - missing the block deprecation code, please. Please, review. Later, Juan. On this v2: - dropped -incoming <uri> deprecation Paolo came with a better solution using keyvalues. - skipped field is already ready for next pull request, so dropped. - dropped the RFC bits, nermal PATCH. - Assessed all the review comments. - Added indentation of migration.json. - Used the documentation pointer to substitute block migration. Please review. [v1] Hi this series describe the migration parts that have to be deprecated. - It is an rfc because I doubt that I did the deprecation process right. Hello Markus O:-) - skipped field: It is older than me, I have never know what it stands for. As far as I know it has always been zero. - inc/blk migrate command options. They are only used by block migration (that I deprecate on the following patch). And they are really bad. grep must_remove_block_options. - block migration. block jobs, whatever they are called this week are way more flexible. Current code works, but we broke it here and there, and really nobody has stand up to maintain it. It is quite contained and can be left there. Is anyone really using it? - old compression method. It don't work. See last try from Lukas to make a test that works reliabely. I failed with the same task years ago. It is really slow, and if compression is good for you, multifd + zlib is going to perform/compress way more. I don't know what to do with this code, really. * Remove it for this release? It don't work, and haven't work reliabely in quite a few time. * Deprecate it and remove in another couple of releases, i.e. normal deprecation. * Ideas? - -incoming <uri> if you need to set parameters (multifd cames to mind, and preempt has the same problem), you really needs to use defer. So what should we do here? This part is not urget, because management apps have a working option that are already using "defer", and the code simplifacation if we remove it is not so big. So we can leave it until 9.0 or whatever we think fit. What do you think? Later, Juan. Juan Quintela (4): migration: migrate 'inc' command option is deprecated. migration: migrate 'blk' command option is deprecated. migration: Deprecate block migration migration: Deprecate old compression method docs/about/deprecated.rst | 35 +++++++++++++ qapi/migration.json | 93 ++++++++++++++++++++++++++-------- migration/block.c | 3 ++ migration/migration-hmp-cmds.c | 10 ++++ migration/migration.c | 10 ++++ migration/options.c | 22 +++++++- tests/qemu-iotests/183.out | 2 + 7 files changed, 152 insertions(+), 23 deletions(-) -- 2.41.0

2 7

[PATCH v6 0/5] Migration deprecated parts
by Juan Quintela 18 Oct '23

18 Oct '23

Based on: Message-ID: <20231017083003.15951-1-quintela(a)redhat.com> Migration 20231017 patches On this v6: - Fixed Markus comments - 1st patch is reviewed - dropped the RFC ones. Later, Juan. On this v5: - Rebased on top of last migration pull requesnt: - address markus comments. Basically we recommend always blockdev-mirror + NBD. In deprecated.rst we also put the posiblity of using block-incremental and block, but we state that they are also deprecated. I know, I know, I deprecated them in the following patch. - Dropped the removal of block-migration and block-incremental I am only interested in showing why I want to remove the -b/-i options. Please review. Later, Juan. On this v4: - addressed all markus comments. - rebased on latest. - improve formatting of migration.json - print block migration status when needed. - patches 7-10 are not mean to merge, they just show why we want to deprecate block migration and remove its support. - Patch 7 just drop support for -i/-b and qmp equivalents. - Patch 8 shows all the helpers and convolutions we need to have to support that -i and -d. - patch 9 drops block-incremental migration support. - patch 9 drops block migration support. Please review. Thanks, Juan. On this v3: - Rebase on top of upstream. - Changed v8.1 to 8.2 (I left the reviewed by anyways) - missing the block deprecation code, please. Please, review. Later, Juan. On this v2: - dropped -incoming <uri> deprecation Paolo came with a better solution using keyvalues. - skipped field is already ready for next pull request, so dropped. - dropped the RFC bits, nermal PATCH. - Assessed all the review comments. - Added indentation of migration.json. - Used the documentation pointer to substitute block migration. Please review. [v1] Hi this series describe the migration parts that have to be deprecated. - It is an rfc because I doubt that I did the deprecation process right. Hello Markus O:-) - skipped field: It is older than me, I have never know what it stands for. As far as I know it has always been zero. - inc/blk migrate command options. They are only used by block migration (that I deprecate on the following patch). And they are really bad. grep must_remove_block_options. - block migration. block jobs, whatever they are called this week are way more flexible. Current code works, but we broke it here and there, and really nobody has stand up to maintain it. It is quite contained and can be left there. Is anyone really using it? - old compression method. It don't work. See last try from Lukas to make a test that works reliabely. I failed with the same task years ago. It is really slow, and if compression is good for you, multifd + zlib is going to perform/compress way more. I don't know what to do with this code, really. * Remove it for this release? It don't work, and haven't work reliabely in quite a few time. * Deprecate it and remove in another couple of releases, i.e. normal deprecation. * Ideas? - -incoming <uri> if you need to set parameters (multifd cames to mind, and preempt has the same problem), you really needs to use defer. So what should we do here? This part is not urget, because management apps have a working option that are already using "defer", and the code simplifacation if we remove it is not so big. So we can leave it until 9.0 or whatever we think fit. What do you think? Later, Juan. Juan Quintela (5): migration: Print block status when needed migration: migrate 'inc' command option is deprecated. migration: migrate 'blk' command option is deprecated. migration: Deprecate block migration migration: Deprecate old compression method docs/about/deprecated.rst | 36 +++++++++++++ qapi/migration.json | 93 ++++++++++++++++++++++++++-------- migration/block.c | 3 ++ migration/migration-hmp-cmds.c | 15 ++++-- migration/migration.c | 10 ++++ migration/options.c | 22 +++++++- 6 files changed, 153 insertions(+), 26 deletions(-) -- 2.41.0

2 8

[PATCH v5 0/7] Migration deprecated parts
by Juan Quintela 18 Oct '23

18 Oct '23

Based on: Message-ID: <20231017083003.15951-1-quintela(a)redhat.com> Migration 20231017 patches On this v5: - Rebased on top of last migration pull requesnt: - address markus comments. Basically we recommend always blockdev-mirror + NBD. In deprecated.rst we also put the posiblity of using block-incremental and block, but we state that they are also deprecated. I know, I know, I deprecated them in the following patch. - Dropped the removal of block-migration and block-incremental I am only interested in showing why I want to remove the -b/-i options. Please review. Later, Juan. On this v4: - addressed all markus comments. - rebased on latest. - improve formatting of migration.json - print block migration status when needed. - patches 7-10 are not mean to merge, they just show why we want to deprecate block migration and remove its support. - Patch 7 just drop support for -i/-b and qmp equivalents. - Patch 8 shows all the helpers and convolutions we need to have to support that -i and -d. - patch 9 drops block-incremental migration support. - patch 9 drops block migration support. Please review. Thanks, Juan. On this v3: - Rebase on top of upstream. - Changed v8.1 to 8.2 (I left the reviewed by anyways) - missing the block deprecation code, please. Please, review. Later, Juan. On this v2: - dropped -incoming <uri> deprecation Paolo came with a better solution using keyvalues. - skipped field is already ready for next pull request, so dropped. - dropped the RFC bits, nermal PATCH. - Assessed all the review comments. - Added indentation of migration.json. - Used the documentation pointer to substitute block migration. Please review. [v1] Hi this series describe the migration parts that have to be deprecated. - It is an rfc because I doubt that I did the deprecation process right. Hello Markus O:-) - skipped field: It is older than me, I have never know what it stands for. As far as I know it has always been zero. - inc/blk migrate command options. They are only used by block migration (that I deprecate on the following patch). And they are really bad. grep must_remove_block_options. - block migration. block jobs, whatever they are called this week are way more flexible. Current code works, but we broke it here and there, and really nobody has stand up to maintain it. It is quite contained and can be left there. Is anyone really using it? - old compression method. It don't work. See last try from Lukas to make a test that works reliabely. I failed with the same task years ago. It is really slow, and if compression is good for you, multifd + zlib is going to perform/compress way more. I don't know what to do with this code, really. * Remove it for this release? It don't work, and haven't work reliabely in quite a few time. * Deprecate it and remove in another couple of releases, i.e. normal deprecation. * Ideas? - -incoming <uri> if you need to set parameters (multifd cames to mind, and preempt has the same problem), you really needs to use defer. So what should we do here? This part is not urget, because management apps have a working option that are already using "defer", and the code simplifacation if we remove it is not so big. So we can leave it until 9.0 or whatever we think fit. What do you think? Later, Juan. Juan Quintela (7): migration: Print block status when needed migration: migrate 'inc' command option is deprecated. migration: migrate 'blk' command option is deprecated. migration: Deprecate block migration migration: Deprecate old compression method [RFC] migration: Make -i/-b an error for hmp and qmp [RFC] migration: Remove helpers needed for -i/-b migrate options docs/about/deprecated.rst | 36 +++++++++++ qapi/migration.json | 110 ++++++++++++++++++++++++--------- migration/migration.h | 4 -- migration/options.h | 6 -- migration/block.c | 3 + migration/migration-hmp-cmds.c | 18 ++++-- migration/migration.c | 35 ++++------- migration/options.c | 63 +++++++------------ 8 files changed, 165 insertions(+), 110 deletions(-) -- 2.41.0

3 18

Re: [PATCH v2 3/3] qom: Link multiple numa nodes to device using a new object
by Alex Williamson 17 Oct '23

17 Oct '23

On Tue, 17 Oct 2023 14:00:54 +0000 Ankit Agrawal <ankita(a)nvidia.com> wrote: > >> -device vfio-pci-nohotplug,host=0009:01:00.0,bus=pcie.0,addr=04.0,rombar=0,id=dev0 \ > >> -object nvidia-acpi-generic-initiator,id=gi0,device=dev0,numa-node-start=2,numa-node-count=8 > > > > Why didn't we just implement start and count in the base object (or a > > list)? It seems like this gives the nvidia-acpi-generic-initiator two > > different ways to set gi->node, either node= of the parent or > > numa-node-start= here. Once we expose the implicit node count in the > > base object, I'm not sure the purpose of this object. I would have > > thought it for keying the build of the NVIDIA specific _DSD, but that's > > not implemented in this version. > > Agree, allowing a list of nodes to be provided to the acpi-generic-initiator > will remove the need for the nvidia-acpi-generic-initiator object. And what happened to the _DSD? Is it no longer needed? Why? > > I also don't see any programatic means for management tools to know how > > many nodes to create. For example what happens if there's a MIGv2 that > > supports 16 partitions by default and makes use of the same vfio-pci > > variant driver? Thanks, > > It is supposed to stay at 8 for all the G+H devices. Maybe this can be managed > through proper documentation in the user manual? I thought the intention here was that a management tool would automatically configure the VM with these nodes and GI object in support of the device. Planning only for Grace-Hopper isn't looking too far into the future and it's difficult to make software that can reference a user manual. This leads to a higher maintenance burden where the management tool needs to recognize not only the driver, but the device bound to the driver and update as new devices are released. The management tool will never automatically support new devices without making an assumption about the node configuration. Do we therefore need some programatic means for the kernel driver to expose the node configuration to userspace? What interfaces would libvirt like to see here? Is there an opportunity that this could begin to define flavors or profiles for variant devices like we have types for mdev devices where the node configuration would be encompassed in a device profile? Thanks, Alex

2 3

[PATCH Libvirt 0/4] Support for dirty-limit live migration
by ~hyman 17 Oct '23

17 Oct '23

In qemu>=8.1, the dirty-limit functionality for live migration was included. In the live migration scenario, it implements the force convergence using the dirty-limit approach, which results in better reliable read performance. A straightforward dirty-limit capability for live migration is added by this patchset. Users might not care about other dirty-limit arguments like "x-vcpu-dirty-limit-period" or "vcpu-dirty-limit," thus do not expose them to Libvirt and leave the default settings in place. values as default. For more details about dirty-limit. please see the following reference: https://lore.kernel.org/qemu- devel/169024923116.19090.10825599068950039132-0(a)git.sr.ht/ Please review, thanks, Yong Hyman Huang (4): Add VIR_MIGRATE_DIRTY_LIMIT flag qemu_migration: Implement VIR_MIGRATE_DIRTY_LIMIT flag virsh: Add support for VIR_MIGRATE_DIRTY_LIMIT flag NEWS: document support for dirty-limit live migration NEWS.rst | 8 ++++++++ docs/manpages/virsh.rst | 10 +++++++++- include/libvirt/libvirt-domain.h | 5 +++++ src/libvirt-domain.c | 8 ++++++++ src/qemu/qemu_migration.c | 8 ++++++++ src/qemu/qemu_migration.h | 1 + src/qemu/qemu_migration_params.c | 6 ++++++ src/qemu/qemu_migration_params.h | 1 + tools/virsh-domain.c | 10 ++++++++++ 9 files changed, 56 insertions(+), 1 deletion(-) -- 2.38.5

2 5

[PATCH] fix error in printf format string
by zhujun2 17 Oct '23

17 Oct '23

use '%u' to printf unsigned int Signed-off-by: zhujun2 <zhujun2(a)cmss.chinamobile.com> --- examples/c/misc/event-test.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/examples/c/misc/event-test.c b/examples/c/misc/event-test.c index 813bca2699..07d86500cd 100644 --- a/examples/c/misc/event-test.c +++ b/examples/c/misc/event-test.c @@ -969,7 +969,7 @@ myDomainEventMemoryFailureCallback(virConnectPtr conn G_GNUC_UNUSED, void *opaque G_GNUC_UNUSED) { printf("%s EVENT: Domain %s(%d) memory failure: recipient '%d', " - "aciont '%d', flags '%d'\n", __func__, virDomainGetName(dom), + "aciont '%d', flags '%u'\n", __func__, virDomainGetName(dom), virDomainGetID(dom), recipient, action, flags); return 0; } -- 2.17.1

1 0

[PATCH v4 00/10] Migration deprecated parts
by Juan Quintela 17 Oct '23

17 Oct '23

On this v4: - addressed all markus comments. - rebased on latest. - improve formatting of migration.json - print block migration status when needed. - patches 7-10 are not mean to merge, they just show why we want to deprecate block migration and remove its support. - Patch 7 just drop support for -i/-b and qmp equivalents. - Patch 8 shows all the helpers and convolutions we need to have to support that -i and -d. - patch 9 drops block-incremental migration support. - patch 9 drops block migration support. Please review. Thanks, Juan. On this v3: - Rebase on top of upstream. - Changed v8.1 to 8.2 (I left the reviewed by anyways) - missing the block deprecation code, please. Please, review. Later, Juan. On this v2: - dropped -incoming <uri> deprecation Paolo came with a better solution using keyvalues. - skipped field is already ready for next pull request, so dropped. - dropped the RFC bits, nermal PATCH. - Assessed all the review comments. - Added indentation of migration.json. - Used the documentation pointer to substitute block migration. Please review. [v1] Hi this series describe the migration parts that have to be deprecated. - It is an rfc because I doubt that I did the deprecation process right. Hello Markus O:-) - skipped field: It is older than me, I have never know what it stands for. As far as I know it has always been zero. - inc/blk migrate command options. They are only used by block migration (that I deprecate on the following patch). And they are really bad. grep must_remove_block_options. - block migration. block jobs, whatever they are called this week are way more flexible. Current code works, but we broke it here and there, and really nobody has stand up to maintain it. It is quite contained and can be left there. Is anyone really using it? - old compression method. It don't work. See last try from Lukas to make a test that works reliabely. I failed with the same task years ago. It is really slow, and if compression is good for you, multifd + zlib is going to perform/compress way more. I don't know what to do with this code, really. * Remove it for this release? It don't work, and haven't work reliabely in quite a few time. * Deprecate it and remove in another couple of releases, i.e. normal deprecation. * Ideas? - -incoming <uri> if you need to set parameters (multifd cames to mind, and preempt has the same problem), you really needs to use defer. So what should we do here? This part is not urget, because management apps have a working option that are already using "defer", and the code simplifacation if we remove it is not so big. So we can leave it until 9.0 or whatever we think fit. What do you think? Later, Juan. Juan Quintela (10): migration: Improve json and formatting migration: Print block status when needed migration: migrate 'inc' command option is deprecated. migration: migrate 'blk' command option is deprecated. migration: Deprecate block migration migration: Deprecate old compression method [RFC] migration: Make -i/-b an error for hmp and qmp [RFC] migration: Remove helpers needed for -i/-b migrate options [RFC] migration: Remove support for block_incremental [RFC] migration: Remove block migration support docs/about/deprecated.rst | 25 + meson.build | 2 - qapi/migration.json | 133 ++--- include/migration/misc.h | 8 - migration/block.h | 52 -- migration/migration.h | 4 - migration/options.h | 8 - migration/block.c | 1027 -------------------------------- migration/colo.c | 1 - migration/migration-hmp-cmds.c | 40 +- migration/migration.c | 47 +- migration/options.c | 89 +-- migration/ram.c | 16 - migration/savevm.c | 5 - hmp-commands.hx | 17 +- migration/meson.build | 3 - 16 files changed, 111 insertions(+), 1366 deletions(-) delete mode 100644 migration/block.h delete mode 100644 migration/block.c -- 2.41.0

2 17

[PATCH 00/31] qemu: Add accessors for 'storage' and 'format' nodenames and refactor callers ('raw' driver removal part 1)
by Peter Krempa 16 Oct '23

16 Oct '23

This is a prequel series necessary for removing the dummy 'raw' driver node if it isn't needed for performance reasons. Peter Krempa (31): qemu: domain: Identify blockjobs by storage nodename in VM status XML qemu: block: Refactor logic in qemuBlockStorageSourceGetBlockdevProps qemu: block: Rename qemuBlockStorageSourceGetBlockdevProps qemu: block: Add accessors for protocol/storage node names tests: Use 'storage' layer nodename accessors in tests qemuDomainVirStorageSourceFindByNodeName: Use proper accessor qemu: block: Use proper accessors for image formatting/creation code qemu: domain: Convert the status XML code for 'storage' nodenames to new accessors qemu: block: Convert disk 'storage' backend JSON props generator to new accessors qemu: domain: Rework assignment of 'storage' nodenames to use new accessors qemu: Refactor storage backend attach/detach setup code to use 'storage' nodename accessors qemu: Refactor storage backend 'storage' layer helepr object setup qemuDomainGetStatsBlockExportDisk: Use 'storage' node name accessors qemuDomainSetBlockThreshold: Use 'storage' node name accessor conf: Rename 'nodestorage' field of virStorageSource to 'nodenamestorage' qemu: block: Add accessors for format layer node names qemu: block: Add accessors for storage source effective nodename qemuBlockStorageSourceGetFormatProps: Use new frontend name accessor qemu: backup: Use format nodename accessors qemu: blockjob: Use 'format' nodename accessors for job naming qemu: block: Use 'format' nodename accessors in '-blockdev' setup code qemu: domain: Use 'format' layer node name accessors for nodename setup code tests: Use 'format' layer nodename accessors in test code qemu: Convert disk backend setup code to use 'format' nodename accessors qemu: driver: Convert disk stats code to use 'format' nodename accessors qemu: Use 'format' nodename accessors for block dirty bitmap operations qemu: command: Use 'format' nodename accessors for 'pflash' backend setup qemu: Convert migration setup code to use 'format' layer node name accessors qemu: migration: Use 'format' nodename accessors in dirty bitmap migration qemu: driver: Use 'format' nodename accessors for disk resize conf: Rename 'nodeformat' field of virStorageSource to 'nodenameformat' src/conf/storage_source_conf.c | 8 +- src/conf/storage_source_conf.h | 4 +- src/qemu/qemu_backup.c | 8 +- src/qemu/qemu_block.c | 238 ++++++++++++------ src/qemu/qemu_block.h | 25 +- src/qemu/qemu_blockjob.c | 24 +- src/qemu/qemu_checkpoint.c | 9 +- src/qemu/qemu_command.c | 12 +- src/qemu/qemu_domain.c | 68 ++--- src/qemu/qemu_driver.c | 25 +- src/qemu/qemu_migration.c | 17 +- src/qemu/qemu_migration_cookie.c | 5 +- src/qemu/qemu_nbdkit.c | 4 +- src/qemu/qemu_process.c | 2 +- src/qemu/qemu_snapshot.c | 6 +- tests/qemublocktest.c | 16 +- tests/qemumonitorjsontest.c | 4 +- .../blockjob-blockdev-in.xml | 12 +- tests/qemuxml2argvtest.c | 3 +- 19 files changed, 314 insertions(+), 176 deletions(-) -- 2.41.0

2 35

[libvirt PATCH] hypervisor: Move interface mgmt methods to hypervisor
by Praveen K Paladugu 16 Oct '23

16 Oct '23

Move guest interface management methods from qemu to hypervisor. These methods will be shared by networking support in ch driver. Signed-off-by: Praveen K Paladugu <prapal(a)linux.microsoft.com> --- po/POTFILES | 1 + src/hypervisor/domain_interface.c | 279 ++++++++++++++++++++++++++++++ src/hypervisor/domain_interface.h | 38 ++++ src/hypervisor/meson.build | 1 + src/libvirt_private.syms | 6 + src/qemu/qemu_command.c | 7 +- src/qemu/qemu_hotplug.c | 3 +- src/qemu/qemu_interface.c | 246 +------------------------- src/qemu/qemu_interface.h | 6 - src/qemu/qemu_process.c | 3 +- 10 files changed, 341 insertions(+), 249 deletions(-) create mode 100644 src/hypervisor/domain_interface.c create mode 100644 src/hypervisor/domain_interface.h diff --git a/po/POTFILES b/po/POTFILES index 3a51aea5cb..023c041f61 100644 --- a/po/POTFILES +++ b/po/POTFILES @@ -92,6 +92,7 @@ src/hyperv/hyperv_util.c src/hyperv/hyperv_wmi.c src/hypervisor/domain_cgroup.c src/hypervisor/domain_driver.c +src/hypervisor/domain_interface.c src/hypervisor/virhostdev.c src/interface/interface_backend_netcf.c src/interface/interface_backend_udev.c diff --git a/src/hypervisor/domain_interface.c b/src/hypervisor/domain_interface.c new file mode 100644 index 0000000000..b01b6ee767 --- /dev/null +++ b/src/hypervisor/domain_interface.c @@ -0,0 +1,279 @@ +/* + * Copyright Microsoft Corp. 2023 + * + * domain_interface.c: methods to manage guest/domain interfaces + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library. If not, see + * <http://www.gnu.org/licenses/>. + */ + +#include <config.h> + +#include "virmacaddr.h" +#include "virnetdevtap.h" +#include "virconftypes.h" +#include "domain_conf.h" +#include "domain_interface.h" +#include "domain_nwfilter.h" +#include "domain_audit.h" +#include "virebtables.h" +#include "virlog.h" +#include "virfile.h" +#include "viralloc.h" +#include "virnetdevbridge.h" +#include "network_conf.h" + +#define VIR_FROM_THIS VIR_FROM_INTERFACE + +VIR_LOG_INIT("interface.interface_connect"); + +bool +virDomainInterfaceIsVnetCompatModel(const virDomainNetDef *net) +{ + return (virDomainNetIsVirtioModel(net) || + net->model == VIR_DOMAIN_NET_MODEL_E1000E || + net->model == VIR_DOMAIN_NET_MODEL_IGB || + net->model == VIR_DOMAIN_NET_MODEL_VMXNET3); +} + +/* virDomainInterfaceEthernetConnect: + * @def: the definition of the VM + * @driver: qemu driver data + * @net: pointer to the VM's interface description + * @tapfd: array of file descriptor return value for the new device + * @tapfdsize: number of file descriptors in @tapfd + * + * Called *only* called if actualType is VIR_DOMAIN_NET_TYPE_ETHERNET + * (i.e. if the connection is made with a tap device) + */ +int +virDomainInterfaceEthernetConnect(virDomainDef *def, + virDomainNetDef *net, + ebtablesContext *ebtables, + bool macFilter, + bool privileged, + int *tapfd, + size_t tapfdSize) +{ + virMacAddr tapmac; + int ret = -1; + unsigned int tap_create_flags = VIR_NETDEV_TAP_CREATE_IFUP; + bool template_ifname = false; + const char *tunpath = "/dev/net/tun"; + const char *auditdev = tunpath; + + if (net->backend.tap) { + tunpath = net->backend.tap; + if (!privileged) { + virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", + _("cannot use custom tap device in session mode")); + goto cleanup; + } + } + VIR_WARN("PPK: interfaceEthernetConnect Called\n"); + if (virDomainInterfaceIsVnetCompatModel(net)) + tap_create_flags |= VIR_NETDEV_TAP_CREATE_VNET_HDR; + + if (net->managed_tap == VIR_TRISTATE_BOOL_NO) { + if (!net->ifname) { + virReportError(VIR_ERR_INTERNAL_ERROR, "%s", + _("target dev must be supplied when managed='no'")); + goto cleanup; + } + if (virNetDevExists(net->ifname) != 1) { + virReportError(VIR_ERR_INTERNAL_ERROR, "%s", + _("target managed='no' but specified dev doesn't exist")); + goto cleanup; + } + + tap_create_flags |= VIR_NETDEV_TAP_CREATE_ALLOW_EXISTING; + + if (virNetDevMacVLanIsMacvtap(net->ifname)) { + auditdev = net->ifname; + if (virNetDevMacVLanTapOpen(net->ifname, tapfd, tapfdSize) < 0) + goto cleanup; + if (virNetDevMacVLanTapSetup(tapfd, tapfdSize, + virDomainInterfaceIsVnetCompatModel(net)) < 0) { + goto cleanup; + } + } else { + if (virNetDevTapCreate(&net->ifname, tunpath, tapfd, tapfdSize, + tap_create_flags) < 0) + goto cleanup; + } + } else { + + if (!net->ifname) + template_ifname = true; + + if (virNetDevTapCreate(&net->ifname, tunpath, tapfd, tapfdSize, + tap_create_flags) < 0) { + goto cleanup; + } + + /* The tap device's MAC address cannot match the MAC address + * used by the guest. This results in "received packet on + * vnetX with own address as source address" error logs from + * the kernel. + */ + virMacAddrSet(&tapmac, &net->mac); + if (tapmac.addr[0] == 0xFE) + tapmac.addr[0] = 0xFA; + else + tapmac.addr[0] = 0xFE; + + if (virNetDevSetMAC(net->ifname, &tapmac) < 0) + goto cleanup; + + if (virNetDevSetOnline(net->ifname, true) < 0) + goto cleanup; + } + + if (net->script && + virNetDevRunEthernetScript(net->ifname, net->script) < 0) + goto cleanup; + + if (macFilter && + ebtablesAddForwardAllowIn(ebtables, + net->ifname, + &net->mac) < 0) + goto cleanup; + + if (net->filter && + virDomainConfNWFilterInstantiate(def->name, def->uuid, net, false) < 0) { + goto cleanup; + } + + virDomainAuditNetDevice(def, net, auditdev, true); + + ret = 0; + + cleanup: + if (ret < 0) { + size_t i; + + virDomainAuditNetDevice(def, net, auditdev, false); + for (i = 0; i < tapfdSize && tapfd[i] >= 0; i++) + VIR_FORCE_CLOSE(tapfd[i]); + if (template_ifname) + VIR_FREE(net->ifname); + } + + return ret; +} + +/** + * qemuInterfaceStartDevice: + * @net: net device to start + * + * Based upon the type of device provided, perform the appropriate + * work to completely activate the device and make it reachable from + * the rest of the network. + */ +int +virDomainInterfaceStartDevice(virDomainNetDef *net) +{ + virDomainNetType actualType = virDomainNetGetActualType(net); + + switch (actualType) { + case VIR_DOMAIN_NET_TYPE_BRIDGE: + case VIR_DOMAIN_NET_TYPE_NETWORK: + if (virDomainNetGetActualBridgeMACTableManager(net) + == VIR_NETWORK_BRIDGE_MAC_TABLE_MANAGER_LIBVIRT) { + /* libvirt is managing the FDB of the bridge this device + * is attaching to, so we have turned off learning and + * unicast_flood on the device to prevent the kernel from + * adding any FDB entries for it. This means we need to + * add an fdb entry ourselves, using the MAC address from + * the interface config. + */ + if (virNetDevBridgeFDBAdd(&net->mac, net->ifname, + VIR_NETDEVBRIDGE_FDB_FLAG_MASTER | + VIR_NETDEVBRIDGE_FDB_FLAG_TEMP) < 0) + return -1; + } + break; + + case VIR_DOMAIN_NET_TYPE_DIRECT: { + const char *physdev = virDomainNetGetActualDirectDev(net); + bool isOnline = true; + + /* set the physdev online if necessary. It may already be up, + * in which case we shouldn't re-up it just in case that causes + * some sort of "blip" in the physdev's status. + */ + if (physdev && virNetDevGetOnline(physdev, &isOnline) < 0) + return -1; + if (!isOnline && virNetDevSetOnline(physdev, true) < 0) + return -1; + + /* macvtap devices share their MAC address with the guest + * domain, and if they are set online prior to the domain CPUs + * being started, the host may send out traffic from this + * device that could confuse other entities on the network (in + * particular, if this new domain is the destination of a + * migration, and the source domain is still running, another + * host may mistakenly direct traffic for the guest to the + * destination domain rather than source domain). To prevent + * this, we create the macvtap device with IFF_UP false + * (i.e. "offline") then wait to bring it online until just as + * we are starting the domain CPUs. + */ + if (virNetDevSetOnline(net->ifname, true) < 0) + return -1; + break; + } + + case VIR_DOMAIN_NET_TYPE_ETHERNET: + if (virNetDevIPInfoAddToDev(net->ifname, &net->hostIP) < 0) + return -1; + + break; + + case VIR_DOMAIN_NET_TYPE_USER: + case VIR_DOMAIN_NET_TYPE_VHOSTUSER: + case VIR_DOMAIN_NET_TYPE_SERVER: + case VIR_DOMAIN_NET_TYPE_CLIENT: + case VIR_DOMAIN_NET_TYPE_MCAST: + case VIR_DOMAIN_NET_TYPE_UDP: + case VIR_DOMAIN_NET_TYPE_INTERNAL: + case VIR_DOMAIN_NET_TYPE_HOSTDEV: + case VIR_DOMAIN_NET_TYPE_VDPA: + case VIR_DOMAIN_NET_TYPE_NULL: + case VIR_DOMAIN_NET_TYPE_VDS: + case VIR_DOMAIN_NET_TYPE_LAST: + /* these types all require no action */ + break; + } + + return 0; +} + +/** + * virDomainInterfaceStartDevices: + * @def: domain definition + * + * Set all ifaces associated with this domain to the online state. + */ +int +virDomainInterfaceStartDevices(virDomainDef *def) +{ + size_t i; + + for (i = 0; i < def->nnets; i++) { + if (virDomainInterfaceStartDevice(def->nets[i]) < 0) + return -1; + } + return 0; +} diff --git a/src/hypervisor/domain_interface.h b/src/hypervisor/domain_interface.h new file mode 100644 index 0000000000..4e76b84990 --- /dev/null +++ b/src/hypervisor/domain_interface.h @@ -0,0 +1,38 @@ +/* + * Copyright Microsoft Corp. 2023 + * + * domain_interface.h: methods to manage guest/domain interfaces + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library. If not, see + * <http://www.gnu.org/licenses/>. + */ + +#pragma once + +#include "virebtables.h" + +int +virDomainInterfaceEthernetConnect(virDomainDef *def, + virDomainNetDef *net, + ebtablesContext *ebtables, + bool macFilter, + bool privileged, + int *tapfd, + size_t tapfdSize); + +bool +virDomainInterfaceIsVnetCompatModel(const virDomainNetDef *net); + +int virDomainInterfaceStartDevice(virDomainNetDef *net); +int virDomainInterfaceStartDevices(virDomainDef *def); diff --git a/src/hypervisor/meson.build b/src/hypervisor/meson.build index f35565b16b..819a9a82a2 100644 --- a/src/hypervisor/meson.build +++ b/src/hypervisor/meson.build @@ -1,6 +1,7 @@ hypervisor_sources = [ 'domain_cgroup.c', 'domain_driver.c', + 'domain_interface.c', 'virclosecallbacks.c', 'virhostdev.c', ] diff --git a/src/libvirt_private.syms b/src/libvirt_private.syms index 4e475d5b1a..cae0ecb857 100644 --- a/src/libvirt_private.syms +++ b/src/libvirt_private.syms @@ -1630,6 +1630,12 @@ virDomainDriverNodeDeviceReset; virDomainDriverParseBlkioDeviceStr; virDomainDriverSetupPersistentDefBlkioParams; +# hypervisor/domain_interface.h +virDomainInterfaceEthernetConnect; +virDomainInterfaceIsVnetCompatModel; +virDomainInterfaceStartDevice; +virDomainInterfaceStartDevices; + # hypervisor/virclosecallbacks.h virCloseCallbacksDomainAdd; diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index 8a7b80719f..eda5327293 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -43,6 +43,7 @@ #include "domain_nwfilter.h" #include "domain_addr.h" #include "domain_conf.h" +#include "domain_interface.h" #include "netdev_bandwidth_conf.h" #include "virnetdevopenvswitch.h" #include "device_conf.h" @@ -53,6 +54,7 @@ #include "virmdev.h" #include "virutil.h" + #include <sys/stat.h> #include <fcntl.h> @@ -8523,7 +8525,10 @@ qemuBuildInterfaceConnect(virDomainObj *vm, break; case VIR_DOMAIN_NET_TYPE_ETHERNET: - if (qemuInterfaceEthernetConnect(vm->def, priv->driver, net, + if (virDomainInterfaceEthernetConnect(vm->def, net, + priv->driver->ebtables, + priv->driver->config->macFilter, + priv->driver->privileged, tapfd, tapfdSize) < 0) return -1; vhostfd = true; diff --git a/src/qemu/qemu_hotplug.c b/src/qemu/qemu_hotplug.c index 1f7f5bdd26..8e76cb8b38 100644 --- a/src/qemu/qemu_hotplug.c +++ b/src/qemu/qemu_hotplug.c @@ -39,6 +39,7 @@ #include "qemu_virtiofs.h" #include "domain_audit.h" #include "domain_cgroup.h" +#include "domain_interface.h" #include "netdev_bandwidth_conf.h" #include "domain_nwfilter.h" #include "virlog.h" @@ -1269,7 +1270,7 @@ qemuDomainAttachNetDevice(virQEMUDriver *driver, } /* Set device online immediately */ - if (qemuInterfaceStartDevice(net) < 0) + if (virDomainInterfaceStartDevice(net) < 0) goto cleanup; qemuDomainInterfaceSetDefaultQDisc(driver, net); diff --git a/src/qemu/qemu_interface.c b/src/qemu/qemu_interface.c index 8856bb95a8..48e3422ae1 100644 --- a/src/qemu/qemu_interface.c +++ b/src/qemu/qemu_interface.c @@ -24,6 +24,7 @@ #include "network_conf.h" #include "domain_audit.h" #include "domain_nwfilter.h" +#include "domain_interface.h" #include "qemu_interface.h" #include "viralloc.h" #include "virlog.h" @@ -41,110 +42,9 @@ VIR_LOG_INIT("qemu.qemu_interface"); -/** - * qemuInterfaceStartDevice: - * @net: net device to start - * - * Based upon the type of device provided, perform the appropriate - * work to completely activate the device and make it reachable from - * the rest of the network. - */ -int -qemuInterfaceStartDevice(virDomainNetDef *net) -{ - virDomainNetType actualType = virDomainNetGetActualType(net); - - switch (actualType) { - case VIR_DOMAIN_NET_TYPE_BRIDGE: - case VIR_DOMAIN_NET_TYPE_NETWORK: - if (virDomainNetGetActualBridgeMACTableManager(net) - == VIR_NETWORK_BRIDGE_MAC_TABLE_MANAGER_LIBVIRT) { - /* libvirt is managing the FDB of the bridge this device - * is attaching to, so we have turned off learning and - * unicast_flood on the device to prevent the kernel from - * adding any FDB entries for it. This means we need to - * add an fdb entry ourselves, using the MAC address from - * the interface config. - */ - if (virNetDevBridgeFDBAdd(&net->mac, net->ifname, - VIR_NETDEVBRIDGE_FDB_FLAG_MASTER | - VIR_NETDEVBRIDGE_FDB_FLAG_TEMP) < 0) - return -1; - } - break; - - case VIR_DOMAIN_NET_TYPE_DIRECT: { - const char *physdev = virDomainNetGetActualDirectDev(net); - bool isOnline = true; - - /* set the physdev online if necessary. It may already be up, - * in which case we shouldn't re-up it just in case that causes - * some sort of "blip" in the physdev's status. - */ - if (physdev && virNetDevGetOnline(physdev, &isOnline) < 0) - return -1; - if (!isOnline && virNetDevSetOnline(physdev, true) < 0) - return -1; - - /* macvtap devices share their MAC address with the guest - * domain, and if they are set online prior to the domain CPUs - * being started, the host may send out traffic from this - * device that could confuse other entities on the network (in - * particular, if this new domain is the destination of a - * migration, and the source domain is still running, another - * host may mistakenly direct traffic for the guest to the - * destination domain rather than source domain). To prevent - * this, we create the macvtap device with IFF_UP false - * (i.e. "offline") then wait to bring it online until just as - * we are starting the domain CPUs. - */ - if (virNetDevSetOnline(net->ifname, true) < 0) - return -1; - break; - } - - case VIR_DOMAIN_NET_TYPE_ETHERNET: - if (virNetDevIPInfoAddToDev(net->ifname, &net->hostIP) < 0) - return -1; - - break; - case VIR_DOMAIN_NET_TYPE_USER: - case VIR_DOMAIN_NET_TYPE_VHOSTUSER: - case VIR_DOMAIN_NET_TYPE_SERVER: - case VIR_DOMAIN_NET_TYPE_CLIENT: - case VIR_DOMAIN_NET_TYPE_MCAST: - case VIR_DOMAIN_NET_TYPE_UDP: - case VIR_DOMAIN_NET_TYPE_INTERNAL: - case VIR_DOMAIN_NET_TYPE_HOSTDEV: - case VIR_DOMAIN_NET_TYPE_VDPA: - case VIR_DOMAIN_NET_TYPE_NULL: - case VIR_DOMAIN_NET_TYPE_VDS: - case VIR_DOMAIN_NET_TYPE_LAST: - /* these types all require no action */ - break; - } - return 0; -} -/** - * qemuInterfaceStartDevices: - * @def: domain definition - * - * Set all ifaces associated with this domain to the online state. - */ -int -qemuInterfaceStartDevices(virDomainDef *def) -{ - size_t i; - - for (i = 0; i < def->nnets; i++) { - if (qemuInterfaceStartDevice(def->nets[i]) < 0) - return -1; - } - return 0; -} /** @@ -166,7 +66,7 @@ qemuInterfaceStopDevice(virDomainNetDef *net) if (virDomainNetGetActualBridgeMACTableManager(net) == VIR_NETWORK_BRIDGE_MAC_TABLE_MANAGER_LIBVIRT) { /* remove the FDB entries that were added during - * qemuInterfaceStartDevices() + * virDomainInterfaceStartDevices() */ if (virNetDevBridgeFDBDel(&net->mac, net->ifname, VIR_NETDEVBRIDGE_FDB_FLAG_MASTER | @@ -236,14 +136,7 @@ qemuInterfaceStopDevices(virDomainDef *def) } -static bool -qemuInterfaceIsVnetCompatModel(const virDomainNetDef *net) -{ - return (virDomainNetIsVirtioModel(net) || - net->model == VIR_DOMAIN_NET_MODEL_E1000E || - net->model == VIR_DOMAIN_NET_MODEL_IGB || - net->model == VIR_DOMAIN_NET_MODEL_VMXNET3); -} + /** @@ -271,7 +164,7 @@ qemuInterfaceDirectConnect(virDomainDef *def, unsigned int macvlan_create_flags = VIR_NETDEV_MACVLAN_CREATE_WITH_TAP; qemuDomainNetworkPrivate *netpriv = QEMU_DOMAIN_NETWORK_PRIVATE(net); - if (qemuInterfaceIsVnetCompatModel(net)) + if (virDomainInterfaceIsVnetCompatModel(net)) macvlan_create_flags |= VIR_NETDEV_MACVLAN_VNET_HDR; if (virNetDevMacVLanCreateWithVPortProfile(net->ifname, @@ -409,133 +302,6 @@ qemuCreateInBridgePortWithHelper(virQEMUDriverConfig *cfg, return *tapfd < 0 ? -1 : 0; } - -/* qemuInterfaceEthernetConnect: - * @def: the definition of the VM - * @driver: qemu driver data - * @net: pointer to the VM's interface description - * @tapfd: array of file descriptor return value for the new device - * @tapfdsize: number of file descriptors in @tapfd - * - * Called *only* called if actualType is VIR_DOMAIN_NET_TYPE_ETHERNET - * (i.e. if the connection is made with a tap device) - */ -int -qemuInterfaceEthernetConnect(virDomainDef *def, - virQEMUDriver *driver, - virDomainNetDef *net, - int *tapfd, - size_t tapfdSize) -{ - virMacAddr tapmac; - int ret = -1; - unsigned int tap_create_flags = VIR_NETDEV_TAP_CREATE_IFUP; - bool template_ifname = false; - g_autoptr(virQEMUDriverConfig) cfg = virQEMUDriverGetConfig(driver); - const char *tunpath = "/dev/net/tun"; - const char *auditdev = tunpath; - - if (net->backend.tap) { - tunpath = net->backend.tap; - if (!driver->privileged) { - virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", - _("cannot use custom tap device in session mode")); - goto cleanup; - } - } - - if (qemuInterfaceIsVnetCompatModel(net)) - tap_create_flags |= VIR_NETDEV_TAP_CREATE_VNET_HDR; - - if (net->managed_tap == VIR_TRISTATE_BOOL_NO) { - if (!net->ifname) { - virReportError(VIR_ERR_INTERNAL_ERROR, "%s", - _("target dev must be supplied when managed='no'")); - goto cleanup; - } - if (virNetDevExists(net->ifname) != 1) { - virReportError(VIR_ERR_INTERNAL_ERROR, "%s", - _("target managed='no' but specified dev doesn't exist")); - goto cleanup; - } - - tap_create_flags |= VIR_NETDEV_TAP_CREATE_ALLOW_EXISTING; - - if (virNetDevMacVLanIsMacvtap(net->ifname)) { - auditdev = net->ifname; - if (virNetDevMacVLanTapOpen(net->ifname, tapfd, tapfdSize) < 0) - goto cleanup; - if (virNetDevMacVLanTapSetup(tapfd, tapfdSize, - qemuInterfaceIsVnetCompatModel(net)) < 0) { - goto cleanup; - } - } else { - if (virNetDevTapCreate(&net->ifname, tunpath, tapfd, tapfdSize, - tap_create_flags) < 0) - goto cleanup; - } - } else { - - if (!net->ifname) - template_ifname = true; - - if (virNetDevTapCreate(&net->ifname, tunpath, tapfd, tapfdSize, - tap_create_flags) < 0) { - goto cleanup; - } - - /* The tap device's MAC address cannot match the MAC address - * used by the guest. This results in "received packet on - * vnetX with own address as source address" error logs from - * the kernel. - */ - virMacAddrSet(&tapmac, &net->mac); - if (tapmac.addr[0] == 0xFE) - tapmac.addr[0] = 0xFA; - else - tapmac.addr[0] = 0xFE; - - if (virNetDevSetMAC(net->ifname, &tapmac) < 0) - goto cleanup; - - if (virNetDevSetOnline(net->ifname, true) < 0) - goto cleanup; - } - - if (net->script && - virNetDevRunEthernetScript(net->ifname, net->script) < 0) - goto cleanup; - - if (cfg->macFilter && - ebtablesAddForwardAllowIn(driver->ebtables, - net->ifname, - &net->mac) < 0) - goto cleanup; - - if (net->filter && - virDomainConfNWFilterInstantiate(def->name, def->uuid, net, false) < 0) { - goto cleanup; - } - - virDomainAuditNetDevice(def, net, auditdev, true); - - ret = 0; - - cleanup: - if (ret < 0) { - size_t i; - - virDomainAuditNetDevice(def, net, auditdev, false); - for (i = 0; i < tapfdSize && tapfd[i] >= 0; i++) - VIR_FORCE_CLOSE(tapfd[i]); - if (template_ifname) - VIR_FREE(net->ifname); - } - - return ret; -} - - /* qemuInterfaceBridgeConnect: * @def: the definition of the VM * @driver: qemu driver data @@ -578,7 +344,7 @@ qemuInterfaceBridgeConnect(virDomainDef *def, if (!net->ifname) template_ifname = true; - if (qemuInterfaceIsVnetCompatModel(net)) + if (virDomainInterfaceIsVnetCompatModel(net)) tap_create_flags |= VIR_NETDEV_TAP_CREATE_VNET_HDR; if (driver->privileged) { @@ -598,7 +364,7 @@ qemuInterfaceBridgeConnect(virDomainDef *def, * is attaching to, so we need to turn off learning and * unicast_flood on the device to prevent the kernel from * adding any FDB entries for it. We will add an fdb - * entry ourselves (during qemuInterfaceStartDevices(), + * entry ourselves (during virDomainInterfaceStartDevices(), * using the MAC address from the interface config. */ if (virNetDevBridgePortSetLearning(brname, net->ifname, false) < 0) diff --git a/src/qemu/qemu_interface.h b/src/qemu/qemu_interface.h index 6eed3e6bd7..feb299dd6c 100644 --- a/src/qemu/qemu_interface.h +++ b/src/qemu/qemu_interface.h @@ -37,12 +37,6 @@ int qemuInterfaceDirectConnect(virDomainDef *def, size_t tapfdSize, virNetDevVPortProfileOp vmop); -int qemuInterfaceEthernetConnect(virDomainDef *def, - virQEMUDriver *driver, - virDomainNetDef *net, - int *tapfd, - size_t tapfdSize); - int qemuInterfaceBridgeConnect(virDomainDef *def, virQEMUDriver *driver, virDomainNetDef *net, diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c index ae0bb7bf80..8a87e7e685 100644 --- a/src/qemu/qemu_process.c +++ b/src/qemu/qemu_process.c @@ -74,6 +74,7 @@ #include "virhostcpu.h" #include "domain_audit.h" #include "domain_cgroup.h" +#include "domain_interface.h" #include "domain_nwfilter.h" #include "domain_postparse.h" #include "domain_validate.h" @@ -3121,7 +3122,7 @@ qemuProcessStartCPUs(virQEMUDriver *driver, virDomainObj *vm, g_autoptr(virQEMUDriverConfig) cfg = virQEMUDriverGetConfig(driver); /* Bring up netdevs before starting CPUs */ - if (qemuInterfaceStartDevices(vm->def) < 0) + if (virDomainInterfaceStartDevices(vm->def) < 0) return -1; VIR_DEBUG("Using lock state '%s'", NULLSTR(priv->lockState)); -- 2.41.0

4 3