- Devel - libvirt List Archives

[libvirt] [PATCH v3 00/19] Add vhost-user-gpu support
by marcandre.lureau＠redhat.com 30 Aug '19

30 Aug '19

From: Marc-André Lureau <marcandre.lureau(a)redhat.com> Hi, This series adds support for running virtio GPUs in seperate processes, thanks to vhost-user backend. The QEMU support landed for 4.1. There are several benefits of running the GPU/virgl in an external process, since Mesa is rather heavy on the qemu main loop, and may block for a while, or crash. The external GPU process is started with one end of a socket pair, the other end is given to a QEMU chardev attached to a device. The external process is also added to the cgroup to limit resources usage. vhost-user is a generic mechanism that allows implementing virtio device dataplane handling in a separate userspace process. vhost-user-gpu here notably moves the virgl 3d handling out of the main qemu process. The external process will be /usr/libexec/vhost-user-gpu, which comes from qemu.git contrib/vhost-user-gpu code, first released in qemu-4.1. Part of this series deals with discovering the location on disk of the vhost-user-gpu binary, and what capabilities it provides. This uses a similar mechanism to firmware.json, described in qemu docs/interop/vhost-user.json https://github.com/qemu/qemu/blob/master/docs/interop/vhost-user.json qemu 4.1 ships a 50-qemu-gpu.json to match. I believe virtio-fs will use a similar mechanism when it lands in upstream qemu, as virtiofsd is a separate process that communicates with qemu over vhost-user. For a bit more background on vhost-user-gpu process handling and the json interop motivation, here's some of the qemu discussion: https://lists.nongnu.org/archive/html/qemu-devel/2018-08/msg02610.html https://lists.nongnu.org/archive/html/qemu-devel/2018-09/msg00807.html For this series, the XML to enable this is: <video model='virtio'> <driver name='vhostuser'/> <acceleration accel3d='yes' rendernode='/path/to/rendernode'/> </video> rendernode is optional qemu_vhost_user.c handles vhost-user.json qemu_vhost_user_gpu.c handles the process management for vhost-user-gpu v3: - rename qemu_configs -> qemu_interop_config - replace <model .. vhostuser='yes'/> with <driver name='vhostuser'/> - move vhost_user_binary to virDomainVideoDriverDef - some VIR_AUTO usage improvements - introduce qemuDomainVideoPrivate to store vhost-user fd - improved selection of -device model - use DO_TEST_CAPS_LATEST - allocate a rendernode with virHostGetDRMRenderNode() - but no clear idea how to have a test - add a patch to ignore EPERM in virDirOpenIfExists() - better domain checks/validate - added Ján r-b - no s-o-b from Cole, per request and commits taken from his git branch - rebase, indentation/style fixes, update doc, version.. v2: - rebase to master - if test file build by dropping LDADDS usage - syntax-check issues: * use #pragma once * if () bracket issues * jump label indent issues * error message %s usage * size_t for loops Marc-André Lureau (19): qemu: generalize qemuFetchConfigs util: ignore EACCESS in virDirOpenIfExists conf: format/parse/rng/docs for video <driver name='qemu|vhostuser'/> domain: add rendernode attribute on <accel> qemu-cgroup: allow accel rendernode access qemu: add vhost-user-gpu capabilities checks qemu: check that qemu is vhost-user-vga capable qemu: validate virtio-gpu with vhost-user qemu: restrict 'virgl=' option to non-vhostuser video type qemu: add vhost-user helpers qemu: add qemuSecurityStartVhostUserGPU helper conf: add privateData to virDomainVideoDef qemu: add qemuDomainVideoPrivate qemu: add vhost-user-gpu helper unit qemu: prepare domain for vhost-user GPU qemu: start/stop the vhost-user-gpu external device qemu: build vhost-user GPU devices tests: mock execv/execve tests: add vhost-user-gpu xml2argv tests docs/formatdomain.html.in | 18 +- docs/schemas/domaincommon.rng | 13 + src/conf/domain_conf.c | 87 +++- src/conf/domain_conf.h | 21 +- src/qemu/Makefile.inc.am | 6 + src/qemu/qemu_capabilities.c | 4 + src/qemu/qemu_capabilities.h | 2 + src/qemu/qemu_cgroup.c | 24 + src/qemu/qemu_command.c | 65 ++- src/qemu/qemu_domain.c | 50 ++- src/qemu/qemu_domain.h | 14 +- src/qemu/qemu_extdevice.c | 73 ++- src/qemu/qemu_extdevice.h | 5 + src/qemu/qemu_firmware.c | 144 +----- src/qemu/qemu_interop_config.c | 183 ++++++++ src/qemu/qemu_interop_config.h | 25 ++ src/qemu/qemu_process.c | 61 ++- src/qemu/qemu_security.c | 40 ++ src/qemu/qemu_security.h | 6 + src/qemu/qemu_vhost_user.c | 422 ++++++++++++++++++ src/qemu/qemu_vhost_user.h | 48 ++ src/qemu/qemu_vhost_user_gpu.c | 285 ++++++++++++ src/qemu/qemu_vhost_user_gpu.h | 50 +++ src/util/virfile.c | 13 +- tests/Makefile.am | 9 + .../caps_4.1.0.x86_64.xml | 2 + .../etc/qemu/vhost-user/40-gpu.json | 1 + .../etc/qemu/vhost-user/50-gpu.json | 0 .../qemu/vhost-user/test-vhost-user-gpu | 11 + .../usr/share/qemu/vhost-user/30-gpu.json | 1 + .../usr/share/qemu/vhost-user/50-gpu.json | 8 + .../usr/share/qemu/vhost-user/60-gpu.json | 1 + tests/qemuvhostusertest.c | 132 ++++++ ...host-user-gpu-secondary.x86_64-latest.args | 43 ++ .../vhost-user-gpu-secondary.xml | 46 ++ .../vhost-user-vga.x86_64-latest.args | 40 ++ tests/qemuxml2argvdata/vhost-user-vga.xml | 42 ++ tests/qemuxml2argvdata/virtio-options.args | 5 +- tests/qemuxml2argvdata/virtio-options.xml | 4 +- tests/qemuxml2argvtest.c | 15 +- tests/virfilewrapper.c | 22 + 41 files changed, 1832 insertions(+), 209 deletions(-) create mode 100644 src/qemu/qemu_interop_config.c create mode 100644 src/qemu/qemu_interop_config.h create mode 100644 src/qemu/qemu_vhost_user.c create mode 100644 src/qemu/qemu_vhost_user.h create mode 100644 src/qemu/qemu_vhost_user_gpu.c create mode 100644 src/qemu/qemu_vhost_user_gpu.h create mode 120000 tests/qemuvhostuserdata/etc/qemu/vhost-user/40-gpu.json create mode 100644 tests/qemuvhostuserdata/etc/qemu/vhost-user/50-gpu.json create mode 100755 tests/qemuvhostuserdata/usr/libexec/qemu/vhost-user/test-vhost-user-gpu create mode 120000 tests/qemuvhostuserdata/usr/share/qemu/vhost-user/30-gpu.json create mode 100644 tests/qemuvhostuserdata/usr/share/qemu/vhost-user/50-gpu.json create mode 120000 tests/qemuvhostuserdata/usr/share/qemu/vhost-user/60-gpu.json create mode 100644 tests/qemuvhostusertest.c create mode 100644 tests/qemuxml2argvdata/vhost-user-gpu-secondary.x86_64-latest.args create mode 100644 tests/qemuxml2argvdata/vhost-user-gpu-secondary.xml create mode 100644 tests/qemuxml2argvdata/vhost-user-vga.x86_64-latest.args create mode 100644 tests/qemuxml2argvdata/vhost-user-vga.xml -- 2.23.0

1 19

[libvirt] [PATCH v2 00/16] Add vhost-user-gpu support
by Cole Robinson 30 Aug '19

30 Aug '19

v1: https://www.redhat.com/archives/libvir-list/2019-June/msg00102.html This is v2 of Marc-André's series with minor changes. I'm not taking over this series, I just fixed these as part of the patch rebase so I can review it :) Changes since v1: - rebase to master - if test file build by dropping LDADDS usage - syntax-check issues: * use #pragma once * if () bracket issues * jump label indent issues * error message %s usage * size_t for loops I didn't know much about vhost-user-gpu before this series, here's what I've learned. vhost-user is a generic mechanism that allows implementing virtio device dataplane handling in a separate userspace process. vhost-user-gpu here notably moves the virgl 3d handling out of the main qemu process. The external process will be /usr/libexec/vhost-user-gpu, which comes from qemu.git contrib/vhost-user-gpu code, first released in qemu-4.1. Part of this series deals with discovering the location on disk of the vhost-user-gpu binary, and what capabilities it provides. This uses a similar mechanism to firmware.json, described in qemu docs/interop/vhost-user.json https://github.com/qemu/qemu/blob/master/docs/interop/vhost-user.json qemu 4.1 ships a 50-qemu-gpu.json to match. I believe virtio-fs will use a similar mechanism when it lands in upstream qemu, as virtiofsd is a separate process that communicates with qemu over vhost-user. For a bit more background on vhost-user-gpu process handling and the json interop motivation, here's some of the qemu discussion: https://lists.nongnu.org/archive/html/qemu-devel/2018-08/msg02610.html https://lists.nongnu.org/archive/html/qemu-devel/2018-09/msg00807.html For this series, the XML to enable this is: <video model='virtio' vhostuser='yes'> <acceleration accel3d='yes' rendernode='/path/to/rendernode'/> </video> rendernode is optional qemu_vhost_user.c handles vhost-user.json qemu_vhost_user_gpu.c handles the process management for vhost-user-gpu Marc-André Lureau (16): qemu: extract out qemuFetchConfigs from firmware domain: add "vhostuser" attribute to virtio video model domain: add rendernode attribute on <accel> qemu-cgroup: allow accel rendernode access qemu: add vhost-user-gpu capabilities checks qemu: check that qemu is vhost-user-vga capable qemu: validate virtio-gpu with vhost-user qemu: restrict 'virgl=' option to non-vhostuser video type qemu: add vhost-user helpers qemu: add qemuSecurityStartVhostUserGPU helper qemu: add vhost-user-gpu helper unit qemu: prepare domain for vhost-user GPU qemu: start/stop the vhost-user-gpu external device qemu: build vhost-user GPU devices tests: mock execv/execve tests: add vhost-user-gpu xml2argv tests docs/formatdomain.html.in | 11 + docs/schemas/domaincommon.rng | 16 +- src/conf/device_conf.c | 1 + src/conf/device_conf.h | 2 + src/conf/domain_conf.c | 32 +- src/conf/domain_conf.h | 2 + src/qemu/Makefile.inc.am | 6 + src/qemu/qemu_capabilities.c | 4 + src/qemu/qemu_capabilities.h | 2 + src/qemu/qemu_cgroup.c | 24 ++ src/qemu/qemu_command.c | 54 ++- src/qemu/qemu_configs.c | 183 ++++++++ src/qemu/qemu_configs.h | 28 ++ src/qemu/qemu_domain.c | 11 +- src/qemu/qemu_extdevice.c | 75 +++- src/qemu/qemu_extdevice.h | 5 + src/qemu/qemu_firmware.c | 144 +------ src/qemu/qemu_process.c | 18 +- src/qemu/qemu_security.c | 47 +++ src/qemu/qemu_security.h | 6 + src/qemu/qemu_vhost_user.c | 394 ++++++++++++++++++ src/qemu/qemu_vhost_user.h | 48 +++ src/qemu/qemu_vhost_user_gpu.c | 305 ++++++++++++++ src/qemu/qemu_vhost_user_gpu.h | 50 +++ tests/Makefile.am | 9 + .../caps_4.1.0.x86_64.xml | 2 + .../etc/qemu/vhost-user/40-gpu.json | 1 + .../etc/qemu/vhost-user/50-gpu.json | 0 .../qemu/vhost-user/test-vhost-user-gpu | 11 + .../usr/share/qemu/vhost-user/30-gpu.json | 1 + .../usr/share/qemu/vhost-user/50-gpu.json | 8 + .../usr/share/qemu/vhost-user/60-gpu.json | 1 + tests/qemuvhostusertest.c | 132 ++++++ .../vhost-user-gpu-secondary.args | 38 ++ .../vhost-user-gpu-secondary.xml | 44 ++ tests/qemuxml2argvdata/vhost-user-vga.args | 35 ++ tests/qemuxml2argvdata/vhost-user-vga.xml | 41 ++ tests/qemuxml2argvtest.c | 21 + tests/virfilewrapper.c | 22 + 39 files changed, 1676 insertions(+), 158 deletions(-) create mode 100644 src/qemu/qemu_configs.c create mode 100644 src/qemu/qemu_configs.h create mode 100644 src/qemu/qemu_vhost_user.c create mode 100644 src/qemu/qemu_vhost_user.h create mode 100644 src/qemu/qemu_vhost_user_gpu.c create mode 100644 src/qemu/qemu_vhost_user_gpu.h create mode 120000 tests/qemuvhostuserdata/etc/qemu/vhost-user/40-gpu.json create mode 100644 tests/qemuvhostuserdata/etc/qemu/vhost-user/50-gpu.json create mode 100755 tests/qemuvhostuserdata/usr/libexec/qemu/vhost-user/test-vhost-user-gpu create mode 120000 tests/qemuvhostuserdata/usr/share/qemu/vhost-user/30-gpu.json create mode 100644 tests/qemuvhostuserdata/usr/share/qemu/vhost-user/50-gpu.json create mode 120000 tests/qemuvhostuserdata/usr/share/qemu/vhost-user/60-gpu.json create mode 100644 tests/qemuvhostusertest.c create mode 100644 tests/qemuxml2argvdata/vhost-user-gpu-secondary.args create mode 100644 tests/qemuxml2argvdata/vhost-user-gpu-secondary.xml create mode 100644 tests/qemuxml2argvdata/vhost-user-vga.args create mode 100644 tests/qemuxml2argvdata/vhost-user-vga.xml -- 2.21.0

3 44

[libvirt] [PATCH 0/3] qemu: Fix formatting of (some) unsigned long long values to JSON (blockdev-add saga)
by Peter Krempa 30 Aug '19

30 Aug '19

Currently, libvirt would regress to computer-middle-ages by not being able to use more than 4GiB images for snapshots without corrupting data. Fix it by using the proper formatting modifier. Peter Krempa (3): qemu: block: Use correct type when creating image size JSON entries tests: qemublock: Use bigger numbers as dummy capacity/physical qemu: monitor: Fix formatting of 'offset' in qemuMonitorJSONSaveMemory src/qemu/qemu_block.c | 12 ++++++------ src/qemu/qemu_monitor_json.c | 2 +- tests/qemublocktest.c | 4 ++-- .../qemublocktestdata/imagecreate/luks-encopts.json | 4 ++-- tests/qemublocktestdata/imagecreate/luks-noopts.json | 4 ++-- .../imagecreate/network-gluster-qcow2.json | 4 ++-- .../imagecreate/network-rbd-qcow2.json | 4 ++-- .../imagecreate/network-sheepdog-qcow2.json | 4 ++-- .../imagecreate/network-ssh-qcow2.json | 4 ++-- .../imagecreate/qcow2-backing-luks.json | 4 ++-- .../imagecreate/qcow2-backing-raw-nbd.json | 4 ++-- .../imagecreate/qcow2-backing-raw.json | 4 ++-- .../imagecreate/qcow2-luks-encopts-backing.json | 4 ++-- .../imagecreate/qcow2-luks-encopts.json | 4 ++-- .../imagecreate/qcow2-luks-noopts.json | 4 ++-- tests/qemublocktestdata/imagecreate/qcow2.json | 4 ++-- tests/qemublocktestdata/imagecreate/raw.json | 2 +- 17 files changed, 36 insertions(+), 36 deletions(-) -- 2.21.0

3 8

[libvirt] [PATCH for 5.7.0 v2 0/3] qemu_blockjob: Remove secdriver metadata more frequently
by Michal Privoznik 30 Aug '19

30 Aug '19

v2 of: https://www.redhat.com/archives/libvir-list/2019-August/msg01418.html Patches 1/3 and 2/3 were ACKed but I did not pushed them yet - I'd like to push these all at once. diff to v1: - Instead of restoring seclabels on the backing chain, remove the metadata as we can't know if somebody else is not using some parts of the chain. Michal Prívozník (3): qemu_blockjob: Move active commit failed state handling into a function qemu_blockjob: Print image path on failed security metadata move too qemu_blockjob: Remove secdriver metadata more frequently src/qemu/qemu_blockjob.c | 78 ++++++++++++++++++++++++++++++++++++---- 1 file changed, 72 insertions(+), 6 deletions(-) -- 2.21.0

2 4

[libvirt] [PATCH for 5.7.0 0/3] qemu_blockjob: Restore seclabels more frequently
by Michal Privoznik 30 Aug '19

30 Aug '19

At some points in blockjob code we were not restoring seclabels on backing chain qemu deceased to use. This leads to perms/XATTRs leak problem. Michal Prívozník (3): qemu_blockjob: Move active commit failed state handling into a function qemu_blockjob: Print image path on failed security metadata move too qemu_blockjob: Restore seclabels more frequently on job events src/qemu/qemu_blockjob.c | 46 ++++++++++++++++++++++++++++++++++------ 1 file changed, 40 insertions(+), 6 deletions(-) -- 2.21.0

2 7

[libvirt] [PATCH] vircgroupv2: fix abort in VIR_AUTOFREE
by Pavel Hrdina 30 Aug '19

30 Aug '19

Introduced by commit <c854e0bd33c7a5afb04a36465bf04f861b2efef5> that tried to fix an issue where we would fail to parse values from files. We cannot change the original pointer that is going to be used by VIR_AUTOFREE. Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1747440 Signed-off-by: Pavel Hrdina <phrdina(a)redhat.com> --- Sigh, shame on me! src/util/vircgroupv2.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/src/util/vircgroupv2.c b/src/util/vircgroupv2.c index c62ee0d933..2aca4e5d62 100644 --- a/src/util/vircgroupv2.c +++ b/src/util/vircgroupv2.c @@ -849,6 +849,7 @@ virCgroupV2GetBlkioDeviceWeight(virCgroupPtr group, VIR_AUTOFREE(char *) path = NULL; VIR_AUTOFREE(char *) str = NULL; VIR_AUTOFREE(char *) value = NULL; + char *tmp; if (virCgroupV2PathOfController(group, VIR_CGROUP_CONTROLLER_BLKIO, "io.weight", &path) < 0) { @@ -869,7 +870,7 @@ virCgroupV2GetBlkioDeviceWeight(virCgroupPtr group, if (!str) { *weight = 0; - } else if (virStrToLong_ui(str, &str, 10, weight) < 0) { + } else if (virStrToLong_ui(str, &tmp, 10, weight) < 0) { virReportError(VIR_ERR_INTERNAL_ERROR, _("Unable to parse '%s' as an integer"), str); @@ -1576,6 +1577,7 @@ virCgroupV2GetCpuCfsQuota(virCgroupPtr group, long long *cfs_quota) { VIR_AUTOFREE(char *) str = NULL; + char *tmp; if (virCgroupGetValueStr(group, VIR_CGROUP_CONTROLLER_CPU, "cpu.max", &str) < 0) { @@ -1587,7 +1589,7 @@ virCgroupV2GetCpuCfsQuota(virCgroupPtr group, return 0; } - if (virStrToLong_ll(str, &str, 10, cfs_quota) < 0) { + if (virStrToLong_ll(str, &tmp, 10, cfs_quota) < 0) { virReportError(VIR_ERR_INTERNAL_ERROR, _("Failed to parse value '%s' from cpu.max."), str); return -1; -- 2.21.0

2 1

[libvirt] [PATCH v1 0/3] unplug timeout QEMU configuration
by Daniel Henrique Barboza 30 Aug '19

30 Aug '19

At this moment, the unplug timeout Libvirt uses is 5 seconds. Which is good enough for most cases, but can be troublesome for cases in which it is not enough. For example, PowerPC guests with lots of vcpus can experience the 'vcpu unplug request timed out' message depending on the guest workload. What happens is that the user doesn't know if there was a problem in the unplug operation in QEMU, or the process took a bit longer than 5 seconds and Libvirt timed out, with QEMU carrying on the unplug operation regardless. This series implements a new 'unplug_timeout' attribute to allow the user to set this parameter, allowing users to set bigger timeouts, if so they choose. The existing 5 seconds timeout is the default value if the attribute isn't set. 5 seconds is also the minimal timeout allowed. Daniel Henrique Barboza (3): adding unplug_timeout QEMU conf qemu_hotplug: use qemu_driver->unplugTimeout in device removal qemu: Remove qemu_hotplugpriv.h src/qemu/Makefile.inc.am | 1 - src/qemu/libvirtd_qemu.aug | 3 +++ src/qemu/qemu.conf | 4 ++++ src/qemu/qemu_conf.c | 26 ++++++++++++++++++++++++ src/qemu/qemu_conf.h | 5 +++++ src/qemu/qemu_driver.c | 2 ++ src/qemu/qemu_hotplug.c | 15 +++++--------- src/qemu/qemu_hotplugpriv.h | 32 ------------------------------ src/qemu/test_libvirtd_qemu.aug.in | 1 + tests/qemuhotplugtest.c | 3 +-- 10 files changed, 47 insertions(+), 45 deletions(-) delete mode 100644 src/qemu/qemu_hotplugpriv.h -- 2.21.0

4 10

[libvirt] [PATCH trivial] qemu_conf.c: removing unused virQEMUDriverConfigPtr variable
by Daniel Henrique Barboza 30 Aug '19

30 Aug '19

'virQEMUDriverConfigPtr cfg' is declared, initiated, but never used in virQEMUDriverCreateCapabilities(). Signed-off-by: Daniel Henrique Barboza <danielhb413(a)gmail.com> --- Accidental discovery while reading code. src/qemu/qemu_conf.c | 3 --- 1 file changed, 3 deletions(-) diff --git a/src/qemu/qemu_conf.c b/src/qemu/qemu_conf.c index 89c183e46a..d771bb6916 100644 --- a/src/qemu/qemu_conf.c +++ b/src/qemu/qemu_conf.c @@ -1266,7 +1266,6 @@ virCapsPtr virQEMUDriverCreateCapabilities(virQEMUDriverPtr driver) virSecurityManagerPtr *sec_managers = NULL; /* Security driver data */ const char *doi, *model, *lbl, *type; - virQEMUDriverConfigPtr cfg = virQEMUDriverGetConfig(driver); const int virtTypes[] = {VIR_DOMAIN_VIRT_KVM, VIR_DOMAIN_VIRT_QEMU,}; @@ -1313,13 +1312,11 @@ virCapsPtr virQEMUDriverCreateCapabilities(virQEMUDriverPtr driver) } VIR_FREE(sec_managers); - virObjectUnref(cfg); return caps; error: VIR_FREE(sec_managers); virObjectUnref(caps); - virObjectUnref(cfg); return NULL; } -- 2.21.0

3 2

[libvirt] [PATCH 0/6] security_selinux: Don't store XATTRs if FS fakes SELinux
by Michal Privoznik 30 Aug '19

30 Aug '19

For full explanation see 6/6, but here's a digest: GlusterFS via FUSE supports XATTRs but doesn't allow any SELinux label change (which is fortunate for us because migrations work at least). However, we need to treat this situation as "don't remember any seclabels" because if the source sets XATTRs and the migration destination tries to set different label this fails. Michal Prívozník (6): virSecuritySELinuxGetProcessLabel: Fix comment virSecuritySELinuxSetFileconImpl: Drop @optional argument security_selinux: DropvirSecuritySELinuxSetFileconOptional() security_selinux: Drop @optional from _virSecuritySELinuxContextItem security_selinux: Drop virSecuritySELinuxSetFileconHelper security_selinux: Play nicely with network FS that only emulates SELinux src/security/security_selinux.c | 141 ++++++++++++++++---------------- 1 file changed, 70 insertions(+), 71 deletions(-) -- 2.21.0

2 16

[libvirt] [PATCH] qemu: use 'bochs' video type by default for UEFI domains
by Jonathon Jongsma 30 Aug '19

30 Aug '19

The 'bochs' video device doesn't have any legacy vga emulation so the attack surface is much lower. It works with OVMF, so UEFI guests should not see any functional difference to VGA. https://bugzilla.redhat.com/show_bug.cgi?id=1707119 Signed-off-by: Jonathon Jongsma <jjongsma(a)redhat.com> --- NOTE: You may run into an error when trying to use the bochs video device. For example: error: internal error: process exited while connecting to monitor: 2019-08-28T21:32:20.134546Z qemu-system-x86_64: -device bochs-display,id=video0,vgamem=16384k,bus=pcie.0,addr=0x1: failed to find romfile "vgabios-bochs-display.bin" This should be solved in e.g. Fedora 31 with newer releases of seabios/qemu. As a temporary workaround, you can symlink the appropriate vgabios file under /usr/share/qemu/. src/qemu/qemu_domain.c | 19 +++++---- src/qemu/qemu_domain.h | 1 + .../video-default-nouefi.x86_64-latest.args | 36 +++++++++++++++++ .../qemuxml2argvdata/video-default-nouefi.xml | 20 ++++++++++ .../video-default-uefi.x86_64-latest.args | 40 +++++++++++++++++++ tests/qemuxml2argvdata/video-default-uefi.xml | 22 ++++++++++ tests/qemuxml2argvtest.c | 2 + 7 files changed, 133 insertions(+), 7 deletions(-) create mode 100644 tests/qemuxml2argvdata/video-default-nouefi.x86_64-latest.args create mode 100644 tests/qemuxml2argvdata/video-default-nouefi.xml create mode 100644 tests/qemuxml2argvdata/video-default-uefi.x86_64-latest.args create mode 100644 tests/qemuxml2argvdata/video-default-uefi.xml diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c index 4998474dc9..7ecb89ac84 100644 --- a/src/qemu/qemu_domain.c +++ b/src/qemu/qemu_domain.c @@ -4584,6 +4584,14 @@ qemuDomainValidateCpuCount(const virDomainDef *def, } +static bool +qemuDomainDefIsUEFI(const virDomainDef *def) +{ + return ((def->os.firmware == VIR_DOMAIN_OS_DEF_FIRMWARE_EFI || + (def->os.loader && def->os.loader->type == + VIR_DOMAIN_LOADER_TYPE_PFLASH))); +} + static int qemuDomainDefValidate(const virDomainDef *def, virCapsPtr caps ATTRIBUTE_UNUSED, @@ -4606,10 +4614,7 @@ qemuDomainDefValidate(const virDomainDef *def, } /* On x86, UEFI requires ACPI */ - if ((def->os.firmware == VIR_DOMAIN_OS_DEF_FIRMWARE_EFI || - (def->os.loader && - def->os.loader->type == VIR_DOMAIN_LOADER_TYPE_PFLASH)) && - ARCH_IS_X86(def->os.arch) && + if (qemuDomainDefIsUEFI(def) && ARCH_IS_X86(def->os.arch) && def->features[VIR_DOMAIN_FEATURE_ACPI] != VIR_TRISTATE_SWITCH_ON) { virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", _("UEFI requires ACPI on this architecture")); @@ -4619,9 +4624,7 @@ qemuDomainDefValidate(const virDomainDef *def, /* On aarch64, ACPI requires UEFI */ if (def->features[VIR_DOMAIN_FEATURE_ACPI] == VIR_TRISTATE_SWITCH_ON && def->os.arch == VIR_ARCH_AARCH64 && - (def->os.firmware != VIR_DOMAIN_OS_DEF_FIRMWARE_EFI && - (!def->os.loader || - def->os.loader->type != VIR_DOMAIN_LOADER_TYPE_PFLASH))) { + !qemuDomainDefIsUEFI(def)) { virReportError(VIR_ERR_CONFIG_UNSUPPORTED, "%s", _("ACPI requires UEFI on this architecture")); goto cleanup; @@ -7452,6 +7455,8 @@ qemuDomainDeviceVideoDefPostParse(virDomainVideoDefPtr video, qemuDomainIsRISCVVirt(def) || ARCH_IS_S390(def->os.arch)) video->type = VIR_DOMAIN_VIDEO_TYPE_VIRTIO; + else if (qemuDomainDefIsUEFI(def)) + video->type = VIR_DOMAIN_VIDEO_TYPE_BOCHS; else video->type = VIR_DOMAIN_VIDEO_TYPE_CIRRUS; } diff --git a/src/qemu/qemu_domain.h b/src/qemu/qemu_domain.h index 37a00323a7..c6deab1c52 100644 --- a/src/qemu/qemu_domain.h +++ b/src/qemu/qemu_domain.h @@ -923,6 +923,7 @@ bool qemuDomainHasBuiltinIDE(const virDomainDef *def); bool qemuDomainNeedsFDC(const virDomainDef *def); bool qemuDomainSupportsPCI(virDomainDefPtr def, virQEMUCapsPtr qemuCaps); +bool qemuDomainIsUEFI(const virDomainDef *def); void qemuDomainUpdateCurrentMemorySize(virDomainObjPtr vm); diff --git a/tests/qemuxml2argvdata/video-default-nouefi.x86_64-latest.args b/tests/qemuxml2argvdata/video-default-nouefi.x86_64-latest.args new file mode 100644 index 0000000000..f0c9e36594 --- /dev/null +++ b/tests/qemuxml2argvdata/video-default-nouefi.x86_64-latest.args @@ -0,0 +1,36 @@ +LC_ALL=C \ +PATH=/bin \ +HOME=/tmp/lib/domain--1-guest \ +USER=test \ +LOGNAME=test \ +XDG_DATA_HOME=/tmp/lib/domain--1-guest/.local/share \ +XDG_CACHE_HOME=/tmp/lib/domain--1-guest/.cache \ +XDG_CONFIG_HOME=/tmp/lib/domain--1-guest/.config \ +QEMU_AUDIO_DRV=none \ +/usr/bin/qemu-system-x86_64 \ +-name guest=guest,debug-threads=on \ +-S \ +-object secret,id=masterKey0,format=raw,\ +file=/tmp/lib/domain--1-guest/master-key.aes \ +-machine q35,accel=tcg,usb=off,dump-guest-core=off \ +-cpu Haswell \ +-m 1024 \ +-overcommit mem-lock=off \ +-smp 1,sockets=1,cores=1,threads=1 \ +-uuid 26b73eb7-f8c4-4541-ae6f-06607a1b21c3 \ +-display none \ +-no-user-config \ +-nodefaults \ +-chardev socket,id=charmonitor,fd=1729,server,nowait \ +-mon chardev=charmonitor,id=monitor,mode=control \ +-rtc base=utc \ +-no-shutdown \ +-boot strict=on \ +-device pcie-root-port,port=0x10,chassis=1,id=pci.1,bus=pcie.0,\ +multifunction=on,addr=0x2 \ +-device pcie-root-port,port=0x11,chassis=2,id=pci.2,bus=pcie.0,addr=0x2.0x1 \ +-device qemu-xhci,id=usb,bus=pci.1,addr=0x0 \ +-device cirrus-vga,id=video0,bus=pcie.0,addr=0x1 \ +-sandbox on,obsolete=deny,elevateprivileges=deny,spawn=deny,\ +resourcecontrol=deny \ +-msg timestamp=on diff --git a/tests/qemuxml2argvdata/video-default-nouefi.xml b/tests/qemuxml2argvdata/video-default-nouefi.xml new file mode 100644 index 0000000000..7db2bedf6c --- /dev/null +++ b/tests/qemuxml2argvdata/video-default-nouefi.xml @@ -0,0 +1,20 @@ +<domain type='qemu'> + <name>guest</name> + <uuid>26b73eb7-f8c4-4541-ae6f-06607a1b21c3</uuid> + <memory unit='KiB'>1048576</memory> + <vcpu placement='static'>1</vcpu> + <os> + <type arch='x86_64' machine='q35'>hvm</type> + </os> + <features> + <acpi/> + </features> + <cpu mode='custom'> + <model>Haswell</model> + </cpu> + <devices> + <emulator>/usr/bin/qemu-system-x86_64</emulator> + <memballoon model='none'/> + <video/> + </devices> +</domain> diff --git a/tests/qemuxml2argvdata/video-default-uefi.x86_64-latest.args b/tests/qemuxml2argvdata/video-default-uefi.x86_64-latest.args new file mode 100644 index 0000000000..75c599f321 --- /dev/null +++ b/tests/qemuxml2argvdata/video-default-uefi.x86_64-latest.args @@ -0,0 +1,40 @@ +LC_ALL=C \ +PATH=/bin \ +HOME=/tmp/lib/domain--1-guest \ +USER=test \ +LOGNAME=test \ +XDG_DATA_HOME=/tmp/lib/domain--1-guest/.local/share \ +XDG_CACHE_HOME=/tmp/lib/domain--1-guest/.cache \ +XDG_CONFIG_HOME=/tmp/lib/domain--1-guest/.config \ +QEMU_AUDIO_DRV=none \ +/usr/bin/qemu-system-x86_64 \ +-name guest=guest,debug-threads=on \ +-S \ +-object secret,id=masterKey0,format=raw,\ +file=/tmp/lib/domain--1-guest/master-key.aes \ +-machine q35,accel=tcg,usb=off,dump-guest-core=off \ +-cpu Haswell \ +-drive file=/usr/share/OVMF/OVMF_CODE.fd,if=pflash,format=raw,unit=0,\ +readonly=on \ +-drive file=/var/lib/libvirt/qemu/nvram/guest_VARS.fd,if=pflash,format=raw,\ +unit=1 \ +-m 1024 \ +-overcommit mem-lock=off \ +-smp 1,sockets=1,cores=1,threads=1 \ +-uuid 26b73eb7-f8c4-4541-ae6f-06607a1b21c3 \ +-display none \ +-no-user-config \ +-nodefaults \ +-chardev socket,id=charmonitor,fd=1729,server,nowait \ +-mon chardev=charmonitor,id=monitor,mode=control \ +-rtc base=utc \ +-no-shutdown \ +-boot strict=on \ +-device pcie-root-port,port=0x10,chassis=1,id=pci.1,bus=pcie.0,\ +multifunction=on,addr=0x2 \ +-device pcie-root-port,port=0x11,chassis=2,id=pci.2,bus=pcie.0,addr=0x2.0x1 \ +-device qemu-xhci,id=usb,bus=pci.1,addr=0x0 \ +-device bochs-display,id=video0,vgamem=16384k,bus=pcie.0,addr=0x1 \ +-sandbox on,obsolete=deny,elevateprivileges=deny,spawn=deny,\ +resourcecontrol=deny \ +-msg timestamp=on diff --git a/tests/qemuxml2argvdata/video-default-uefi.xml b/tests/qemuxml2argvdata/video-default-uefi.xml new file mode 100644 index 0000000000..59e880c78c --- /dev/null +++ b/tests/qemuxml2argvdata/video-default-uefi.xml @@ -0,0 +1,22 @@ +<domain type='qemu'> + <name>guest</name> + <uuid>26b73eb7-f8c4-4541-ae6f-06607a1b21c3</uuid> + <memory unit='KiB'>1048576</memory> + <vcpu placement='static'>1</vcpu> + <os> + <type arch='x86_64' machine='q35'>hvm</type> + <loader readonly='yes' type='pflash'>/usr/share/OVMF/OVMF_CODE.fd</loader> + <nvram>/var/lib/libvirt/qemu/nvram/guest_VARS.fd</nvram> + </os> + <features> + <acpi/> + </features> + <cpu mode='custom'> + <model>Haswell</model> + </cpu> + <devices> + <emulator>/usr/bin/qemu-system-x86_64</emulator> + <memballoon model='none'/> + <video/> + </devices> +</domain> diff --git a/tests/qemuxml2argvtest.c b/tests/qemuxml2argvtest.c index 9395cc19a2..671e79e631 100644 --- a/tests/qemuxml2argvtest.c +++ b/tests/qemuxml2argvtest.c @@ -2045,6 +2045,8 @@ mymain(void) DO_TEST("video-none-device", QEMU_CAPS_VNC); DO_TEST_PARSE_ERROR("video-invalid-multiple-devices", NONE); + DO_TEST_CAPS_LATEST("video-default-uefi"); + DO_TEST_CAPS_LATEST("video-default-nouefi"); DO_TEST("virtio-rng-default", QEMU_CAPS_DEVICE_VIRTIO_RNG, -- 2.21.0

6 6