- Devel - libvirt List Archives

[libvirt] [PATCH] qemu: Report warning if QoS is set for vhostuser interface
by Michal Privoznik 01 Nov '18

01 Nov '18

https://bugzilla.redhat.com/show_bug.cgi?id=1524230 Because of historical reasons, we are not denying starting a domain which has QoS set for unsupported type of device. We do report just a warning instead. And even though we perhaps used to do so for vhostuser it got lost somewhere. Bring it back. Signed-off-by: Michal Privoznik <mprivozn(a)redhat.com> --- src/qemu/qemu_command.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c index 6e3ff67660..489e8bc689 100644 --- a/src/qemu/qemu_command.c +++ b/src/qemu/qemu_command.c @@ -8246,6 +8246,12 @@ qemuBuildVhostuserCommandLine(virQEMUDriverPtr driver, goto cleanup; } + if (virDomainNetGetActualBandwidth(net)) { + VIR_WARN("setting bandwidth on interfaces of " + "type '%s' is not implemented yet", + virDomainNetTypeToString(VIR_DOMAIN_NET_TYPE_VHOSTUSER)); + } + switch ((virDomainChrType)net->data.vhostuser->type) { case VIR_DOMAIN_CHR_TYPE_UNIX: if (!(chardev = qemuBuildChrChardevStr(logManager, secManager, -- 2.18.1

2 3

[libvirt] [PATCH libvirt-python] event-test.py: Report ERROR events
by Philipp Hahn 01 Nov '18

01 Nov '18

VIR_DOMAIN_EVENT_ID_IO_ERROR and VIR_DOMAIN_EVENT_ID_IO_ERROR_REASON callbacks receive the same 'action' parameter, so also translate that numeric action to a descriptive text for the first callback. Signed-off-by: Philipp Hahn <hahn(a)univention.de> --- examples/event-test.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/examples/event-test.py b/examples/event-test.py index dabf4b0..709277b 100755 --- a/examples/event-test.py +++ b/examples/event-test.py @@ -526,8 +526,8 @@ def myDomainEventWatchdogCallback(conn, dom, action, opaque): def myDomainEventIOErrorCallback(conn, dom, srcpath, devalias, action, opaque): - print("myDomainEventIOErrorCallback: Domain %s(%s) %s %s %d" % ( - dom.name(), dom.ID(), srcpath, devalias, action)) + print("myDomainEventIOErrorCallback: Domain %s(%s) %s %s %s" % ( + dom.name(), dom.ID(), srcpath, devalias, ERROR_EVENTS[action])) def myDomainEventIOErrorReasonCallback(conn, dom, srcpath, devalias, action, reason, opaque): -- 2.11.0

2 1

[libvirt] [PATCH libvirt-python] event-test.py: Fix ERROR event
by Philipp Hahn 01 Nov '18

01 Nov '18

ERROR_EVENTS translates the numeric 'action' argument to a description, not the 'reason' argument which already contains a descriptive string like 'enospc'. > Traceback (most recent call last): > File "/usr/lib/python2.7/dist-packages/libvirt.py", line 4661, in _dispatchDomainEventIOErrorReasonCallback > reason, opaque) > File "libvirt-python/examples/event-test.py", line 536, in myDomainEventIOErrorReasonCallback > dom.name(), dom.ID(), srcpath, devalias, action, ERROR_EVENTS[reason])) > File "libvirt-python/examples/event-test.py", line 474, in __getitem__ > data = self.args[item] > TypeError: tuple indices must be integers, not str Fixes: f5928c6711654f1496707ca77f626b3192843d57 Signed-off-by: Philipp Hahn <hahn(a)univention.de> --- examples/event-test.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/examples/event-test.py b/examples/event-test.py index eeb2777..dabf4b0 100755 --- a/examples/event-test.py +++ b/examples/event-test.py @@ -531,8 +531,8 @@ def myDomainEventIOErrorCallback(conn, dom, srcpath, devalias, action, opaque): def myDomainEventIOErrorReasonCallback(conn, dom, srcpath, devalias, action, reason, opaque): - print("myDomainEventIOErrorReasonCallback: Domain %s(%s) %s %s %d %s" % ( - dom.name(), dom.ID(), srcpath, devalias, action, ERROR_EVENTS[reason])) + print("myDomainEventIOErrorReasonCallback: Domain %s(%s) %s %s %s %s" % ( + dom.name(), dom.ID(), srcpath, devalias, ERROR_EVENTS[action], reason)) def myDomainEventGraphicsCallback(conn, dom, phase, localAddr, remoteAddr, authScheme, subject, opaque): -- 2.11.0

2 1

[libvirt] [PATCH v2] vhost-user: define conventions for vhost-user backends
by Marc-André Lureau 01 Nov '18

01 Nov '18

As discussed during "[PATCH v4 00/29] vhost-user for input & GPU" review, let's define a common set of backend conventions to help with management layer implementation, and interoperability. v2: - drop --pidfile - add some notes about daemonizing & stdin/out/err Cc: libvir-list(a)redhat.com Cc: Gerd Hoffmann <kraxel(a)redhat.com> Cc: Daniel P. Berrangé <berrange(a)redhat.com> Cc: Changpeng Liu <changpeng.liu(a)intel.com> Cc: Dr. David Alan Gilbert <dgilbert(a)redhat.com> Cc: Felipe Franciosi <felipe(a)nutanix.com> Cc: Gonglei <arei.gonglei(a)huawei.com> Cc: Maxime Coquelin <maxime.coquelin(a)redhat.com> Cc: Michael S. Tsirkin <mst(a)redhat.com> Cc: Victor Kaplansky <victork(a)redhat.com> Signed-off-by: Marc-André Lureau <marcandre.lureau(a)redhat.com> --- docs/interop/vhost-user.txt | 109 +++++++++++++++++++++++++++++++++++- 1 file changed, 107 insertions(+), 2 deletions(-) diff --git a/docs/interop/vhost-user.txt b/docs/interop/vhost-user.txt index ba5e37d714..339b335e9c 100644 --- a/docs/interop/vhost-user.txt +++ b/docs/interop/vhost-user.txt @@ -17,8 +17,13 @@ The protocol defines 2 sides of the communication, master and slave. Master is the application that shares its virtqueues, in our case QEMU. Slave is the consumer of the virtqueues. -In the current implementation QEMU is the Master, and the Slave is intended to -be a software Ethernet switch running in user space, such as Snabbswitch. +In the current implementation QEMU is the Master, and the Slave is the +external process consuming the virtio queues, for example a software +Ethernet switch running in user space, such as Snabbswitch, or a block +device backend processing read & write to a virtual disk. In order to +facilitate interoperability between various backend implementations, +it is recommended to follow the "Backend program conventions" +described in this document. Master and slave can be either a client (i.e. connecting) or server (listening) in the socket communication. @@ -859,3 +864,103 @@ resilient for selective requests. For the message types that already solicit a reply from the client, the presence of VHOST_USER_PROTOCOL_F_REPLY_ACK or need_reply bit being set brings no behavioural change. (See the 'Communication' section for details.) + +Backend program conventions +--------------------------- + +vhost-user backends provide various services and they may need to be +configured manually depending on the use case. However, it is a good +idea to follow the conventions listed here when possible. Users, QEMU +or libvirt, can then rely on some common behaviour to avoid +heterogenous configuration and management of the backend program and +facilitate interoperability. + +In order to be discoverable, default vhost-user backends should be +located under "/usr/libexec", and be named "vhost-user-$device" where +"$device" is the device name in lower-case following the name listed +in the Linux virtio_ids.h header (ex: the VIRTIO_ID_RPROC_SERIAL +backend would be named "vhost-user-rproc-serial"). + +Mechanisms to list, and to select among alternatives implementations +or modify the default backend are not described at this point (a +distribution may use update-alternatives, for example, to list and to +pick a different default backend). + +The backend program must not daemonize itself, but it may be +daemonized by the management layer. It may also have a restricted +access to the system. + +File descriptors 0, 1 and 2 will exist, and have regular +stdin/stdout/stderr usage (they may be redirected to /dev/null by the +management layer, or to a log handler). + +The backend program must end (as quickly and cleanly as possible) when +the SIGTERM signal is received. Eventually, it may be SIGKILL by the +management layer after a few seconds. + +The following command line options have an expected behaviour. They +are mandatory, unless explicitly said differently: + +* --socket-path=PATH + +This option specify the location of the vhost-user Unix domain socket. +It is incompatible with --fd. + +* --fd=FDNUM + +When this argument is given, the backend program is started with the +vhost-user socket as file descriptor FDNUM. It is incompatible with +--socket-path. + +* --print-capabilities + +Output to stdout a line-seperated list of backend capabilities, and +then exit successfully. Other options and arguments should be ignored, +and the backend program should not perform its normal function. + +At the time of writing, there are no common capabilities. Some +device-specific capabilities are listed in the respective sections. By +convention, device-specific capabilities are prefixed by their device +name. + +vhost-user-input program conventions +------------------------------------ + +Capabilities: + +input-evdev-path + + The --evdev-path command line option is supported. + +input-no-grab + + The --no-grab command line option is supported. + +* --evdev-path=PATH (optional) + +Specify the linux input device. + +* --no-grab (optional) + +Do no request exclusive access to the input device. + +vhost-user-gpu program conventions +---------------------------------- + +Capabilities: + +gpu-render-node + + The --render-node command line option is supported. + +gpu-virgl + + The --virgl command line option is supported. + +* --render-node=PATH (optional) + +Specify the GPU DRM render node. + +* --virgl (optional) + +Enable virgl rendering support. -- 2.19.0

5 16

[libvirt] [PATCH] libxl: add support for soft reset
by Jim Fehlig 31 Oct '18

31 Oct '18

The pvops Linux kernel implements machine_ops.crash_shutdown as static void xen_hvm_crash_shutdown(struct pt_regs *regs) { native_machine_crash_shutdown(regs); xen_reboot(SHUTDOWN_soft_reset); } but currently the libxl driver does not handle the soft reset shutdown event. As a result, the guest domain never proceeds past xen_reboot(), making it impossible for HVM domains to save a crash dump using kexec. This patch adds support for handling the soft reset event by calling libxl_domain_soft_reset() and re-enabling domain death events, which is similar to the xl tool handling of soft reset shutdown event. Signed-off-by: Jim Fehlig <jfehlig(a)suse.com> --- src/libxl/libxl_domain.c | 30 ++++++++++++++++++++++++++++++ 1 file changed, 30 insertions(+) diff --git a/src/libxl/libxl_domain.c b/src/libxl/libxl_domain.c index 0032b9dd11..295ec04a85 100644 --- a/src/libxl/libxl_domain.c +++ b/src/libxl/libxl_domain.c @@ -447,8 +447,10 @@ libxlDomainShutdownThread(void *opaque) virObjectEventPtr dom_event = NULL; libxl_shutdown_reason xl_reason = ev->u.domain_shutdown.shutdown_reason; libxlDriverConfigPtr cfg; + libxl_domain_config d_config; cfg = libxlDriverConfigGet(driver); + libxl_domain_config_init(&d_config); vm = virDomainObjListFindByID(driver->domains, ev->domid); if (!vm) { @@ -532,6 +534,33 @@ libxlDomainShutdownThread(void *opaque) * after calling libxl_domain_suspend() are handled by it's callers. */ goto endjob; + } else if (xl_reason == LIBXL_SHUTDOWN_REASON_SOFT_RESET) { + libxlDomainObjPrivatePtr priv = vm->privateData; + int res; + + if (libxl_retrieve_domain_configuration(cfg->ctx, vm->def->id, + &d_config) != 0) { + VIR_ERROR(_("Failed to retrieve config for VM '%s'. " + "Unable to perform soft reset. Destroying VM"), + vm->def->name); + goto destroy; + } + + if (priv->deathW) { + libxl_evdisable_domain_death(cfg->ctx, priv->deathW); + priv->deathW = NULL; + } + + res = libxl_domain_soft_reset(cfg->ctx, &d_config, vm->def->id, + NULL, NULL); + if (res != 0) { + VIR_ERROR(_("Failed to soft reset VM '%s'. Destroying VM"), + vm->def->name); + goto destroy; + } + libxl_evenable_domain_death(cfg->ctx, vm->def->id, 0, &priv->deathW); + libxl_domain_unpause(cfg->ctx, vm->def->id); + goto endjob; } else { VIR_INFO("Unhandled shutdown_reason %d", xl_reason); goto endjob; @@ -565,6 +594,7 @@ libxlDomainShutdownThread(void *opaque) virObjectEventStateQueue(driver->domainEventState, dom_event); libxl_event_free(cfg->ctx, ev); VIR_FREE(shutdown_info); + libxl_domain_config_dispose(&d_config); virObjectUnref(cfg); } -- 2.18.0

2 3

[libvirt] [PATCH v2 0/3] qemu: guest dedicated crypto adapters
by Boris Fiuczynski 31 Oct '18

31 Oct '18

v2: + added singleton check for vfio-ap mediated device This patch series introduces initial libvirt support for guest dedicated crypto adapters on S390. It allows to specify a vfio-ap mediated device in a domain. Extensive documentation about AP is available in patch 6 of the QEMU patch series. KVM/kernel: guest dedicated crypto adapters https://lkml.org/lkml/2018/9/26/25 QEMU: s390x: vfio-ap: guest dedicated crypto adapters https://lists.gnu.org/archive/html/qemu-devel/2018-10/msg01993.html The qemu patch series has been included into master. https://git.qemu.org/?p=qemu.git;a=commit;h=69ac8c4cb93f2685839ff7b857cef30… Boris Fiuczynski (3): qemu: add vfio-ap capability qemu: vfio-ap device support news: Update news for vfio-ap support docs/formatdomain.html.in | 3 ++- docs/news.xml | 9 +++++++++ docs/schemas/domaincommon.rng | 1 + src/conf/domain_conf.c | 28 ++++++++++++++++++++++++++++ src/qemu/qemu_capabilities.c | 2 ++ src/qemu/qemu_capabilities.h | 1 + src/qemu/qemu_command.c | 8 ++++++++ src/qemu/qemu_domain_address.c | 4 ++++ src/util/virmdev.c | 3 ++- src/util/virmdev.h | 1 + 10 files changed, 58 insertions(+), 2 deletions(-) -- 2.17.0

5 19

[libvirt] [PATCH 0/3] RfC: add support for deprecated devices.
by Gerd Hoffmann 30 Oct '18

30 Oct '18

Works simliar to machine types, by adding a deprecation_reason field (using the same name for easy grepping). Also deprecate two devices: adlib and cirrus. This is tagged as RfC because the deprecation notice is not added to introspecction yet. Gerd Hoffmann (3): qdev: add deprecation_reason to DeviceClass adlib: mark as insecure and deprecated. cirrus: mark as deprecated hw/audio/adlib.c | 1 + hw/core/qdev.c | 9 ++++++++- hw/display/cirrus_vga.c | 2 ++ hw/display/cirrus_vga_isa.c | 2 ++ include/hw/qdev-core.h | 1 + qdev-monitor.c | 7 +++++++ qemu-deprecated.texi | 8 ++++++++ 7 files changed, 29 insertions(+), 1 deletion(-) -- 2.9.3

9 34

[libvirt] RFC: "Determine if the domain has been updated."?
by Philipp Hahn 30 Oct '18

30 Oct '18

Hello, <https://libvirt.org/html/libvirt-libvirt-domain.html#virDomainIsUpdated> states: > int virDomainIsUpdated (virDomainPtr dom) > > Determine if the domain has been updated. > > dom: pointer to the domain object > Returns: 1 if updated, 0 if not, -1 on error Not very enlightening :-( So what's the meaning of "updated" here: "inactive XML got changed but the currently running VM was started from a different XML"? Only very few drivers seem to mention it at all and they seem to mean different things. Any nobody seems to update the variable, only 4 readers: > $ git grep -ne '->updated' > libxl/libxl_driver.c:4844: ret = vm->updated; > lxc/lxc_driver.c:343: ret = obj->updated; > qemu/qemu_driver.c:1628: ret = obj->updated; > uml/uml_driver.c:1456: ret = obj->updated; Thank you for the clarification. Philipp -- Philipp Hahn Open Source Software Engineer Univention GmbH be open. Mary-Somerville-Str. 1 D-28359 Bremen Tel.: +49 421 22232-0 Fax : +49 421 22232-99 hahn(a)univention.de http://www.univention.de/ Geschäftsführer: Peter H. Ganten HRB 20755 Amtsgericht Bremen Steuer-Nr.: 71-597-02876

1 0

Re: [libvirt] [PATCH v3 0/3] Deprecate QMP & HMP `cpu-add`; document vCPU hotplug
by Kashyap Chamarthy 30 Oct '18

30 Oct '18

[Cc: libvir-list] On Tue, Oct 30, 2018 at 01:35:23PM +0100, Kashyap Chamarthy wrote: > The first patch deprecates the QMP `cpu-add`, the second its HMP > equivalent, and the third documents vCPU hotplug procedure using QMP > `device_add` et al. > > v2: https://lists.gnu.org/archive/html/qemu-devel/2018-09/msg03191.html > > Kashyap Chamarthy (3): > Deprecate QMP `cpu-add` > Deprecate HMP `cpu-add` > docs: Document vCPU hotplug procedure > > docs/cpu-hotplug.rst | 142 +++++++++++++++++++++++++++++++++++++++++++ > hmp-commands.hx | 6 +- > hmp.c | 2 + > qapi/misc.json | 8 ++- > qemu-deprecated.texi | 5 ++ > 5 files changed, 160 insertions(+), 3 deletions(-) > create mode 100644 docs/cpu-hotplug.rst > > -- > 2.17.1 > -- /kashyap

2 1

[libvirt] [PATCH 0/2] Fix compatibility with MTTCG
by Daniel P. Berrangé 30 Oct '18

30 Oct '18

QEMU is using MTTCG by default on an increasingly large set of host/guest combinations. This allows us to use the normal vCPU pinning support we already have for KVM. We just need to stop throwing away the PID info, and stop artificially blocking pinning APIs. Daniel P. Berrangé (2): qemu: fix recording of vCPU pids for MTTCG Revert "qemu: Forbid pinning vCPUs for TCG domain" src/qemu/qemu_domain.c | 81 ++++++++++++++++++++++++++---------------- src/qemu/qemu_driver.c | 7 ---- 2 files changed, 51 insertions(+), 37 deletions(-) -- 2.17.2

3 8