On Fri, Nov 27, 2020 at 16:50:01 +, Daniel Berrange wrote:
> On Fri, Nov 27, 2020 at 04:15:44PM +0100, Peter Krempa wrote:
> > On Thu, Nov 26, 2020 at 19:51:53 +, Daniel Berrange wrote:
> > > On Thu, Nov 26, 2020 at 02:23:17PM +0100, Peter Krempa wrote:
[...]
> > I can try to send a cut-d
On Fri, Nov 27, 2020 at 04:15:44PM +0100, Peter Krempa wrote:
> On Thu, Nov 26, 2020 at 19:51:53 +, Daniel Berrange wrote:
> > On Thu, Nov 26, 2020 at 02:23:17PM +0100, Peter Krempa wrote:
> > > Gitlab allows projects to define templates for filing issues [1].
> > > Libvirt didn't have them yet
On Fri, Nov 27, 2020 at 04:37:53PM +0100, Michal PrÃvoznÃk wrote:
> On 11/27/20 12:52 PM, Roman Bolshakov wrote:
> > On Fri, Nov 27, 2020 at 10:53:49AM +0100, Michal PrÃvoznÃk wrote:
> > > On 11/23/20 11:10 PM, Roman Bolshakov wrote:
> > > > The test takes 40+ seconds on MBP 2012, MBA 2015. Cir
I have just tagged v6.10.0-rc2 in the repository and pushed signed
tarballs and source RPMs to https://libvirt.org/sources/
Please give the release candidate some testing and in case you find a
serious issue which should have a fix in the upcoming release, feel
free to reply to this thread to make
On Fri, Nov 27, 2020 at 15:53:16 +, Daniel Berrange wrote:
> On Fri, Nov 27, 2020 at 01:18:09PM +0100, Peter Krempa wrote:
> > On Fri, Nov 27, 2020 at 14:45:12 +0300, Roman Bolshakov wrote:
[...]
> IIUC, this all relies on us importing a dump of the latest QEMU
> capabilities into the libvirt
On Fri, Nov 27, 2020 at 16:44:05 +0100, Markus Armbruster wrote:
> Peter Krempa writes:
>
> > On Fri, Nov 27, 2020 at 14:45:12 +0300, Roman Bolshakov wrote:
> >> On Fri, Nov 27, 2020 at 12:21:54PM +0100, Peter Krempa wrote:
[...]
> > As you can see this has an issue when we have to add support
On Fri, Nov 27, 2020 at 01:18:09PM +0100, Peter Krempa wrote:
> On Fri, Nov 27, 2020 at 14:45:12 +0300, Roman Bolshakov wrote:
> > On Fri, Nov 27, 2020 at 12:21:54PM +0100, Peter Krempa wrote:
> > > On Fri, Nov 27, 2020 at 10:50:59 +, Daniel Berrange wrote:
> > > > Copying libvir-list for the d
Hi,
QEMU Advent Calendar 2020 is around the corner and we are looking for
volunteers to contribute disk images that showcase something cool, bring
back retro computing memories, or simply entertain with a puzzle or game.
QEMU Advent Calendar publishes a QEMU disk image each day from
December 1-2
Peter Krempa writes:
> On Fri, Nov 27, 2020 at 14:45:12 +0300, Roman Bolshakov wrote:
>> On Fri, Nov 27, 2020 at 12:21:54PM +0100, Peter Krempa wrote:
>> > On Fri, Nov 27, 2020 at 10:50:59 +, Daniel Berrange wrote:
>> > > Copying libvir-list for the deprecation warning.
>> > >
>> > >
>> > >
On 11/27/20 12:52 PM, Roman Bolshakov wrote:
On Fri, Nov 27, 2020 at 10:53:49AM +0100, Michal Prívozník wrote:
On 11/23/20 11:10 PM, Roman Bolshakov wrote:
The test takes 40+ seconds on MBP 2012, MBA 2015. Cirrus completes the
test within default timeout, just above 29 seconds but the error mar
On Thu, Nov 26, 2020 at 19:51:53 +, Daniel Berrange wrote:
> On Thu, Nov 26, 2020 at 02:23:17PM +0100, Peter Krempa wrote:
> > Gitlab allows projects to define templates for filing issues [1].
> > Libvirt didn't have them yet. We can use them to encourage users to
> > file better bug reports an
The majority of the code is prepared since virtio-pmem. The alias
for virtio-mem is new ("virtiomem"). And again, prealloc doesn't
make much sense - the whole point is to dynamically allocate
memory for guest and possibly return it to the host (e.g. if the
host is running low on memory) and this wo
Signed-off-by: Michal Privoznik
---
src/qemu/qemu_validate.c | 15 +++
1 file changed, 15 insertions(+)
diff --git a/src/qemu/qemu_validate.c b/src/qemu/qemu_validate.c
index fc9bc7a5b4..e4d12db3bd 100644
--- a/src/qemu/qemu_validate.c
+++ b/src/qemu/qemu_validate.c
@@ -4621,6 +4621,
The virtio-pmem is a virtio variant of NVDIMM and just like
NVDIMM virtio-pmem also allows accessing host pages bypassing
guest page cache. The difference is that if a regular file is
used to back guest's NVDIMM (model='nvdimm') the persistence of
guest writes might not be guaranteed while with vir
Signed-off-by: Michal Privoznik
---
src/security/security_apparmor.c | 26
src/security/security_dac.c | 42 +---
src/security/security_selinux.c | 42 ++--
src/security/virt-aa-helper.c| 22 ++---
The only test we do when checking for UUID validity is that
whether all bytes are the same (invalid UUID) or not (valid
UUID). The algorithm we use is needlessly complicated.
Also, the checked UUID is not modified and hence the argument can
be of 'const' type.
Signed-off-by: Michal Privoznik
---
Provide an user friendly way of modifying 'requested-size' of a
virtio-mem device. New --node and --alias arguments are
introduced but they are needed only if it is not clear which
virtio-mem does user mean.
Signed-off-by: Michal Privoznik
---
tools/virsh-domain.c | 126 +
The virDomainMemorySourceDefFormat() uses good old style of
formatting child buffer (virBufferAdjustIndent()). When switched
to virXMLFormatElement() we can save a couple of lines.
Signed-off-by: Michal Privoznik
---
src/conf/domain_conf.c | 20 +++-
1 file changed, 7 insertions(
Our code expects that a nvdimm has a path defined always. And the
parser does check for that. Well, not fully - only when parsing
(which is an optional element). So if the element is
not in the XML then the check is not performed and the assumption
is broken. Verify in the memory def validator tha
What code tries to achieve is that if no flags were provided to
either 'setmem' or 'setmaxmem' commands then the old (no flags)
API is called to be able to communicate with older daemons.
Well, the code can be simplified a bit.
Signed-off-by: Michal Privoznik
---
tools/virsh-domain.c | 14 --
If the QEMU driver restarts it loses the track of the actual size
of virtio-mem (because it's runtime type of information and thus
not stored in XML) and therefore, we have to refresh it when
reconnecting to the domain monitor.
Signed-off-by: Michal Privoznik
---
src/qemu/qemu_domain.c | 3
As advertised in previous commit, this event is delivered to us
when virtio-mem module changes the allocation inside the guest.
It comes with one attribute - size - which holds the new size of
the virtio-mem (well, allocated size), in bytes.
Mind you, this is not necessarily the same number as 'req
QEMU gained this new virtio-mem model. It's similar to pc-dimm
in a sense that guest uses it as memory, but in a sense very
different from it as it can dynamically change allocation,
without need for hotplug. More specifically, the device has two
attributes more (it has more of course, but these tw
Now we have everything prepared for generating the command line.
The device alias prefix was chosen to be 'virtiopmem'.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1735375
Signed-off-by: Michal Privoznik
---
src/qemu/qemu_alias.c | 44 ++
src/qemu/qemu_c
As advertised in the previous commit, we want' to be able to
change 'requested-size' attribute of virtio-mem on the fly. This
commit does exactly that. Changing anything else is checked for
and forbidden.
Signed-off-by: Michal Privoznik
---
src/conf/domain_conf.c | 23 +
src/conf/doma
So far our memory modules could go only into DIMM slots. But with
virtio model this assumption is no longer true - virtio-pmem goes
onto PCI bus. But for formatting PCI address onto command line we
already have a function - qemuBuildDeviceAddressStr(). Therefore,
mode DIMM address generation into i
Some users might wanto have virtio-pmem backed by a block device
in which case we have to create the device in domain namespace.
Signed-off-by: Michal Privoznik
---
src/qemu/qemu_namespace.c | 19 +--
1 file changed, 17 insertions(+), 2 deletions(-)
diff --git a/src/qemu/qemu_na
Some users might wanto have virtio-pmem backed by a block device
in which case we have to allow the device in CGroups.
Signed-off-by: Michal Privoznik
---
src/qemu/qemu_cgroup.c | 43 --
1 file changed, 33 insertions(+), 10 deletions(-)
diff --git a/src/q
In this commit two new capabilities are introduced that reflect
device support by give qemu:
QEMU_CAPS_DEVICE_VIRTIO_MEM_PCI, /* -device virtio-mem-pci */
QEMU_CAPS_DEVICE_VIRTIO_PMEM_PCI, /* -device virtio-pmem-pci */
Please note, virtio-pmem device was introduced sooner (QEMU 4.1)
than virtio-m
The structure has two sets of members: some for the target and
some for the source. The latter is model dependant (e.g. path to
a nvdimm device applies only to NVDIMM model). Move the members
into an union so that it is obvious which members apply to which
model. This way it's easier to maintain co
There is this function qemuDomainDefValidateMemoryHotplug() which
is called explicitly from hotplug path and the qemu's domain def
validator. This is not really necessary because we can move the
part that validates feature against qemuCaps into device
validator which is called implicitly (from qemu
The virDomainMemoryTargetDefFormat() uses good old style of
formatting child buffer (virBufferAdjustIndent()). When switched
to virXMLFormatElement() we can save a couple of lines
Signed-off-by: Michal Privoznik
---
src/conf/domain_conf.c | 21 +
1 file changed, 9 insertions(
The virDomainMemoryModel structure has a @type member which is
really type of virDomainMemoryModel but we store it as int
because the virDomainMemoryModelTypeFromString() call stores its
retval right into it. Then, to have compiler do compile time
check for us, every switch() typecasts the @type. T
On PPC platform it is required that a NVDIMM has an UUID. If none
is provided then libvirt generates one during parsing (see
v6.2.0-rc1~96 and friends). However, the example provided in our
documentation is not valid XML.
Signed-off-by: Michal Privoznik
---
docs/formatdomain.rst | 6 +++---
1 fi
The UUID is guest visible and thus shouldn't change if we want to
not break guest ABI.
Fixes: 08ed673901bb5b4f419b37bcce9b11d31ce370e6
Signed-off-by: Michal Privoznik
---
src/conf/domain_conf.c | 6 ++
1 file changed, 6 insertions(+)
diff --git a/src/conf/domain_conf.c b/src/conf/domain_con
Signed-off-by: Michal Privoznik
---
src/qemu/qemu_domain_address.c | 10 --
1 file changed, 4 insertions(+), 6 deletions(-)
diff --git a/src/qemu/qemu_domain_address.c b/src/qemu/qemu_domain_address.c
index 2788dc7fb3..d872f75b38 100644
--- a/src/qemu/qemu_domain_address.c
+++ b/src/qemu
Available also here:
https://gitlab.com/MichalPrivoznik/libvirt/-/commits/virtio_mem/
There are new virtio variants of pc-dimm and nvdimm devices. This is the
first attempt to impalement support for them in libvirt.
Thanks to David Hildenbrand for his valuable input!
Michal Prívozník (26):
vi
This macro checks whether given number is an integer power of
two.
Signed-off-by: Michal Privoznik
---
src/internal.h | 10 +
src/qemu/qemu_validate.c | 46 +---
src/util/virrandom.c | 2 +-
3 files changed, 35 insertions(+), 23 deletion
On Fri, Nov 27, 2020 at 03:05:49PM +0100, Andrea Bolognani wrote:
> On Fri, 2020-11-27 at 11:28 +0100, Erik Skultety wrote:
> > On Fri, Nov 27, 2020 at 10:44:56AM +0100, Andrea Bolognani wrote:
> > > On Fri, 2020-11-27 at 09:46 +0100, Erik Skultety wrote:
> > > > +++ b/ci/Makefile
> > > > -ci-check
On Fri, 2020-11-27 at 11:28 +0100, Erik Skultety wrote:
> On Fri, Nov 27, 2020 at 10:44:56AM +0100, Andrea Bolognani wrote:
> > On Fri, 2020-11-27 at 09:46 +0100, Erik Skultety wrote:
> > > +++ b/ci/Makefile
> > > -ci-check@%:
> > > - $(MAKE) -C $(CI_ROOTDIR) ci-build@$* CI_MAKE_ARGS="check"
> > >
On Fri, 2020-11-27 at 10:08 +, Daniel P. Berrangé wrote:
> +++ b/NEWS.rst
> @@ -44,6 +44,13 @@ v6.10.0 (unreleased)
>
> * **Bug fixes**
>
> + * remote: fixed performance regression in SSH tunnelling
> +
> +The virt-ssh-helper binary introduced in 6.9.0 had very
virt-ssh-helper was in
On Fri, Nov 27, 2020 at 14:45:12 +0300, Roman Bolshakov wrote:
> On Fri, Nov 27, 2020 at 12:21:54PM +0100, Peter Krempa wrote:
> > On Fri, Nov 27, 2020 at 10:50:59 +, Daniel Berrange wrote:
> > > Copying libvir-list for the deprecation warning.
> > >
> > >
> > > On Mon, Nov 16, 2020 at 04:10:
On Fri, Nov 27, 2020 at 10:53:49AM +0100, Michal PrÃvoznÃk wrote:
> On 11/23/20 11:10 PM, Roman Bolshakov wrote:
> > The test takes 40+ seconds on MBP 2012, MBA 2015. Cirrus completes the
> > test within default timeout, just above 29 seconds but the error margin
> > is narrow, under a second.
>
On Fri, Nov 27, 2020 at 12:21:54PM +0100, Peter Krempa wrote:
> On Fri, Nov 27, 2020 at 10:50:59 +, Daniel Berrange wrote:
> > Copying libvir-list for the deprecation warning.
> >
> >
> > On Mon, Nov 16, 2020 at 04:10:11PM +0300, Roman Bolshakov wrote:
> > > 'query-accel' QMP command should b
On Fri, Nov 27, 2020 at 10:50:59 +, Daniel Berrange wrote:
> Copying libvir-list for the deprecation warning.
>
>
> On Mon, Nov 16, 2020 at 04:10:11PM +0300, Roman Bolshakov wrote:
> > 'query-accel' QMP command should be used instead.
> >
> > Signed-off-by: Roman Bolshakov
> > ---
> > qapi
On 11/23/2020 8:14 PM, Martin Kletzander wrote:
On Mon, Nov 23, 2020 at 05:02:31PM +0800, Zhong, Luyao wrote:
On 11/20/2020 6:08 PM, Martin Kletzander wrote:
On Fri, Nov 20, 2020 at 01:17:29PM +0800, Zhong, Luyao wrote:
On 11/18/2020 8:42 PM, Martin Kletzander wrote:
So let's finish this
On 11/23/2020 8:14 PM, Martin Kletzander wrote:
On Mon, Nov 23, 2020 at 05:02:31PM +0800, Zhong, Luyao wrote:
On 11/20/2020 6:08 PM, Martin Kletzander wrote:
On Fri, Nov 20, 2020 at 01:17:29PM +0800, Zhong, Luyao wrote:
On 11/18/2020 8:42 PM, Martin Kletzander wrote:
So let's finish this
---
include/libvirt/libvirt-domain.h | 1 +
src/conf/numa_conf.c | 9 +
src/qemu/qemu_command.c | 6 ++-
src/qemu/qemu_process.c | 27 +
src/util/virnuma.c| 3 ++
.../n
support 'restrictive' mode in memory element and memnode
element in numatune:
...
...
---
docs/schemas/domaincommon.rng | 2 ++
1 file changed, 2 insertions(+)
diff --git a/docs/schemas/domaincommon.rng b/docs/schemas/domaincommon.rng
index 795b654feb.
Before this patch set, numatune only has three memory modes:
static, interleave and prefered. These memory policies are
ultimately set by mbind() system call.
Memory policy could be 'hard coded' into the kernel, but none of
above policies fit our requirment under this case. mbind() support
default
When user would like use cgroups to restrict the allowed memory
nodes, and require not setting any specific memory policy, then
'restrictive' mode is useful.
---
docs/formatdomain.rst | 7 +--
1 file changed, 5 insertions(+), 2 deletions(-)
diff --git a/docs/formatdomain.rst b/docs/formatdoma
Copying libvir-list for the deprecation warning.
On Mon, Nov 16, 2020 at 04:10:11PM +0300, Roman Bolshakov wrote:
> 'query-accel' QMP command should be used instead.
>
> Signed-off-by: Roman Bolshakov
> ---
> qapi/machine.json | 6 +-
> 1 file changed, 5 insertions(+), 1 deletion(-)
docs/
On Wed, 25 Nov 2020 12:06:48 +0100
Thomas Huth wrote:
> On s390x, devices are attached to the channel IO subsytem by default,
s/attached to the channel IO subsystem/accessed via the channel subsystem/
> so we need to look up scsi controllers via their CCW address there
> instead of using PCI.
>
On Fri, Nov 27, 2020 at 10:44:56AM +0100, Andrea Bolognani wrote:
> On Fri, 2020-11-27 at 09:46 +0100, Erik Skultety wrote:
> > +++ b/ci/Makefile
> > -ci-check@%:
> > - $(MAKE) -C $(CI_ROOTDIR) ci-build@$* CI_MAKE_ARGS="check"
> > +ci-test@%:
> > + $(MAKE) -C $(CI_ROOTDIR) ci-build@$* CI_NINJA_
On Wed, 25 Nov 2020 12:06:47 +0100
Thomas Huth wrote:
> On s390x, devices are attached to the channel IO subsytem by default,
s/attached to the channel IO subsystem/accessed via the channel subsystem/
> so we need to look up the devices via their CCW address there instead
> of using PCI.
>
> T
On Wed, 25 Nov 2020 12:06:46 +0100
Thomas Huth wrote:
> Newer versions of the QEMU guest agent will provide the CCW address
> of devices on s390x. Store this information in the qemuAgentDiskInfo
> so that we can use this later.
>
> We also map the CSSID 0 from the guest to the value 0xfe on the
Signed-off-by: Daniel P. Berrangé
---
NEWS.rst | 7 +++
1 file changed, 7 insertions(+)
diff --git a/NEWS.rst b/NEWS.rst
index aa8a217eb6..8245e34efa 100644
--- a/NEWS.rst
+++ b/NEWS.rst
@@ -44,6 +44,13 @@ v6.10.0 (unreleased)
* **Bug fixes**
+ * remote: fixed performance regression in
On 11/23/20 11:10 PM, Roman Bolshakov wrote:
Some tests in qemuxml2argvtest need opendir() from virpcimock, others
need opendir() from virfilewrapper.
But as of now, only opendir() from virpcimock has an effect.
real_opendir in virpcimock has a pointer to opendir$INODE64 in
libsystem_kernel.dyli
On 11/23/20 11:10 PM, Roman Bolshakov wrote:
There might be mocks that need to reference qemu test driver and link
with it. It's not possible now because qemu test driver is defined after
mocks.
While at it, add 'link_with' parameter to mock definition that allows to
specify a set of libraries t
On 11/23/20 11:10 PM, Roman Bolshakov wrote:
qemucapsprobemock can't find real versions of qemuMonitorSend() and
qemuMonitorJSONIOProcessLine() on macOS. That breaks qemucapsprobe.
The failure can be explained by documented behaviour of dlsym(3) on
macOS:
If dlsym() is called with the specia
On 11/23/20 11:10 PM, Roman Bolshakov wrote:
The test takes 40+ seconds on MBP 2012, MBA 2015. Cirrus completes the
test within default timeout, just above 29 seconds but the error margin
is narrow, under a second.
It'd be good to provide reasonable default timeout to avoid test suite
failure if
On 11/23/20 11:10 PM, Roman Bolshakov wrote:
There's no need to have different CI process between macOS and FreeBSD
as test suite has been fixed on macOS.
Signed-off-by: Roman Bolshakov
---
ci/cirrus/build.yml | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
Reviewed-by: Michal Pr
On Fri, 2020-11-27 at 09:46 +0100, Erik Skultety wrote:
> +++ b/ci/Makefile
> -ci-check@%:
> - $(MAKE) -C $(CI_ROOTDIR) ci-build@$* CI_MAKE_ARGS="check"
> +ci-test@%:
> + $(MAKE) -C $(CI_ROOTDIR) ci-build@$* CI_NINJA_ARGS+=test
I don't know why this last bit turned from =test in v3 to +=te
Add meson required bits to the ci logic in the repo to be able to run
a meson build in a container.
This patch also drops several environment variables we don't need with
meson anymore.
Signed-off-by: Erik Skultety
---
v3: https://www.redhat.com/archives/libvir-list/2020-November/msg01293.html
Si
64 matches
Mail list logo