On 11/9/2020 7:21 PM, Martin Kletzander wrote:
On Sat, Nov 07, 2020 at 10:41:52AM +0800, Zhong, Luyao wrote:
On 11/4/2020 9:02 PM, Martin Kletzander wrote:
On Fri, Oct 16, 2020 at 10:38:51PM +0800, Zhong, Luyao wrote:
On 10/16/2020 9:32 PM, Zang, Rui wrote:
How about if “migratable” is
When migration fails, qemuMigrationDstPrepareAny will call qemuProcessStop
to restore labels only after all labels are successfully set. If some labels
fail to set, the labels that have been set will not be restore.
Signed-off-by: Jin Yan
---
src/qemu/qemu_security.c | 11 +--
1 file
On 11/11/20 5:38 AM, Michal Privoznik wrote:
Every time we create new virCommand of OVS_VSCTL it must be
followed by virNetDevOpenvswitchAddTimeout() call which adds the
--timeout=X argument to freshly created cmd. Instead of having
this as two separate function calls it can be just one.
On 11/11/20 5:38 AM, Michal Privoznik wrote:
There are two type of vhostuser ports:
s/type/types
dpdkvhostuser - OVS creates the socket and QEMU connects to it
dpdkvhostuserclient - QEMU creates the socket and OVS connects to it
But of course ovs-vsctl syntax for fetching ifname
On 11/11/20 9:09 AM, Jiri Denemark wrote:
Some CPUs provide a way to read exact TSC frequency, while measuring it
is required on other CPUs. However, measuring is never exact and the
result may slightly differ across reboots. For this reason both Linux
kernel and QEMU recently started
Related issue: https://gitlab.com/libvirt/libvirt/-/issues/16
Added in support for the following parameters in attach-disk:
--source-name
--source-protocol
--source-host-name
--source-host-socket
--source-host-transport
Allowed for multiple hosts to be added to a single source.
Multiple hosts can
Related issue: https://gitlab.com/libvirt/libvirt/-/issues/16
Added in support for the following parameters in attach-disk:
--source-name
--source-protocol
--source-host-name
--source-host-socket
--source-host-transport
Allowed for multiple hosts to be added to a single source.
Multiple hosts can
This function has no callers left after we moved the mem module
alignment to PostParse callbacks.
Signed-off-by: Daniel Henrique Barboza
---
src/qemu/qemu_domain.c | 20
src/qemu/qemu_domain.h | 2 --
2 files changed, 22 deletions(-)
diff --git a/src/qemu/qemu_domain.c
The loop in which 'hotplugmem' is calculated in qemuDomainAlignMemorySizes()
is throwing a max size overflow error. This is a reminiscent of
the time where this loop was doing the memory module alignment.
It doesn't hurt to leave the check here, but it makes more sense to
move it to the places
Similar to the previous patch, the alignment of the mem object in
qemuDomainAttachMemory() is obsolete after recent changes.
qemuDomainAttachMemory() is used by qemuDomainAttachDeviceLive(),
which is called by qemuDomainAttachDeviceLiveAndConfig() after
a call to virDomainDeviceDefParse(), where
The alignment of the mem object in qemuDomainDetachPrepMemory()
allows for the virDomainMemoryFindByDef() call to find the
right memory object for detach.
Turns out that, after recent changes, the alignment of the mem
object is already granted at this point. This function is used by
Document that we're not reflecting the aligned value in the
live domain XML, and that we fixed a bug that occurred in
pSeries guests where a guest might start with an extra 256MiB
of memory.
Signed-off-by: Daniel Henrique Barboza
---
NEWS.rst | 14 ++
1 file changed, 14
This is already being done for pSeries guests, so let's extend
the treatment for x86. The difference here is that the existing
x86 alignment in the QEMU driver might be restricted only to
QEMU, meaning that we can't put it in virDomainMemoryDefPostParse()
without infering that all hypervisors
The function is no longer used for ppc64 alignment. This allow us
to exclude the 'def' argument and simplify its usage. Let's also
make it clearer in the comment that the function is now returning
x86 specific values.
Signed-off-by: Daniel Henrique Barboza
---
src/qemu/qemu_domain.c | 14
Hi,
This is a work that I intended to do after pSeries NVDIMM changes
I made in commit v6.7.0-302-gd3f3c2c97f. The reasoning is that
if ppc64 NVDIMM alignment can be done in PostParse time, regular
ppc64 DIMMs alignment can also be done the same way. After that
I realized that the same logic can
This simplifies the handling of ppc64 alignment in other parts
of the code, with a single function to handle all cases for the
architecture.
No functional changes were made just yet. Function was renamed to
reflect the broader scope.
Signed-off-by: Daniel Henrique Barboza
---
qemuDomainAlignMemorySizes() has an operation order problem. We are
calculating 'initialmem' without aligning the memory modules first.
Since we're aligning the dimms afterwards this can create inconsistencies
in the end result. x86 has alignment of 1-2MiB and it's not severely
impacted by it, but
Until now, we were aligning just ppc64 NVDIMM modules in post parse
time. Let's extend the treatment to all ppc64 memory models. This
is possible because the alignment restrictions for the architecture
are imposed by the platform specification, being hypervisor-agnostic.
This will remove the need
On Wed, Nov 11, 2020 at 22:47:34 +0100, Martin Kletzander wrote:
> On Wed, Nov 11, 2020 at 06:00:52PM +0100, Jiri Denemark wrote:
> >Although the code in qemuProcessStartValidateTSC works as if the
> >timer frequency was already unsigned long long (by using an appropriate
> >temporary variable),
On Wed, Nov 11, 2020 at 06:00:52PM +0100, Jiri Denemark wrote:
Although the code in qemuProcessStartValidateTSC works as if the
timer frequency was already unsigned long long (by using an appropriate
temporary variable), the virDomainTimerDef structure actually defines
frequency as unsigned
On Wed, Nov 11, 2020 at 15:18:51 -0600, Jonathon Jongsma wrote:
> Signed-off-by: Jonathon Jongsma
> ---
> src/util/virhash.c | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/src/util/virhash.c b/src/util/virhash.c
> index 5d5b6389b5..d105820bfa 100644
> ---
Signed-off-by: Jonathon Jongsma
---
src/util/virhash.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/util/virhash.c b/src/util/virhash.c
index 5d5b6389b5..d105820bfa 100644
--- a/src/util/virhash.c
+++ b/src/util/virhash.c
@@ -147,8 +147,8 @@ virHashFree(GHashTable
Although the code in qemuProcessStartValidateTSC works as if the
timer frequency was already unsigned long long (by using an appropriate
temporary variable), the virDomainTimerDef structure actually defines
frequency as unsigned long, which is not guaranteed to be 64b.
Fixes support for
On 11/11/20 4:55 PM, Erik Skultety wrote:
On Wed, Nov 11, 2020 at 01:45:19PM +0100, Boris Fiuczynski wrote:
All refactoring patches of this series in v1 were accepted except for
the patch enabling the mdev_types support on CSS devices.
v2 added one more refactoring patch in docs before the
On Wed, Nov 11, 2020 at 01:45:19PM +0100, Boris Fiuczynski wrote:
> All refactoring patches of this series in v1 were accepted except for
> the patch enabling the mdev_types support on CSS devices.
> v2 added one more refactoring patch in docs before the already
> sent enablement patch follows
On Wed, 26 Aug 2020 07:48:15 +0200
Erik Skultety wrote:
> On Tue, Aug 25, 2020 at 03:59:47PM +0100, Daniel P. Berrangé wrote:
> > On Tue, Aug 25, 2020 at 04:55:06PM +0200, Erik Skultety wrote:
> > > On Tue, Aug 25, 2020 at 01:44:40PM +0100, Daniel P. Berrangé
> > > wrote:
> > > > On Tue, Aug
On a Wednesday in 2020, Jiri Denemark wrote:
The unsignedInt XML schema type allows for values up to 2^32 - 1, i.e.,
using 4294967296 or greater TSC frequency would fail schema validation.
Signed-off-by: Jiri Denemark
---
docs/schemas/domaincommon.rng | 2 +-
Synchronize with the latest release candidate.
Signed-off-by: Peter Krempa
---
.../caps_5.2.0.x86_64.replies | 684 +-
.../caps_5.2.0.x86_64.xml | 6 +-
2 files changed, 349 insertions(+), 341 deletions(-)
diff --git
On 11/11/20 2:35 PM, Masayoshi Mizuma wrote:
From: Masayoshi Mizuma
A qemu guest which has virtiofs config fails to start if the previous
starting failed because of invalid option or something.
That's because the virtiofsd isn't killed by virPidFileForceCleanupPath()
on the former failure
On 11/11/20 9:06 AM, Peter Krempa wrote:
On Tue, Nov 10, 2020 at 16:11:41 +0100, Michal Privoznik wrote:
When setting up a new guest or when a management software wants
to allow access to an existing guest the
virDomainSetUserPassword() API can be used, but that might be not
good enough if user
On Wed, Nov 11, 2020 at 12:37 AM Han Han wrote:
>
>
>
> On Wed, Nov 11, 2020 at 10:24 AM Jason Dillaman wrote:
>>
>> On Tue, Nov 10, 2020 at 8:43 PM Han Han wrote:
>> >
>> > Signed-off-by: Han Han
>> > ---
>> > docs/formatdomain.rst | 16 ++
>> > src/conf/domain_conf.c | 47
From: Masayoshi Mizuma
A qemu guest which has virtiofs config fails to start if the previous
starting failed because of invalid option or something.
That's because the virtiofsd isn't killed by virPidFileForceCleanupPath()
on the former failure because the pidfile was already removed by
On 11/11/20 14:08, Kevin Wolf wrote:
These aliases only work the command line, but not in QMP. Command line
QAPIfication involves writing some compatibility glue for them, which
I'm doing, but I think it's desirable to unify accepted values of both
paths. So deprecate the aliases so that we can
The aliases "tty" and "parport" are only valid on the command line, QMP
commands like chardev-add don't know them. query-chardev-backends should
describe QMP and therefore not include them in the list of available
backends.
Signed-off-by: Kevin Wolf
---
chardev/char.c | 21 ++---
These aliases only work the command line, but not in QMP. Command line
QAPIfication involves writing some compatibility glue for them, which
I'm doing, but I think it's desirable to unify accepted values of both
paths. So deprecate the aliases so that we can drop the compatibility
glue later.
In
QAPI doesn't know the aliases 'tty' and 'parport' and there is no
reason to prefer them to the real names of the backends 'serial' and
'parallel'.
Signed-off-by: Kevin Wolf
---
docs/system/deprecated.rst | 6 ++
chardev/char.c | 11 ++-
2 files changed, 16
If starting an container fails, the virLXCProcessStop() is
called. But since vm->def->id is not set until libvirt_lxc is
spawned (the domain's ID is PID of that process),
virLXCProcessStop() returns early as virDomainObjIsActive()
returns false. But doing so leaves behind resources reserved for
All refactoring patches of this series in v1 were accepted except for
the patch enabling the mdev_types support on CSS devices.
v2 added one more refactoring patch in docs before the already
sent enablement patch follows again.
v3 adds another tiny renaming patch and changes some references in the
To prevent copying the mdev_types description multiple times
it is refactored into a new paragraph for easy reuse.
Signed-off-by: Boris Fiuczynski
Reviewed-by: Bjoern Walk
Reviewed-by: Erik Skultety
---
docs/drvnodedev.html.in | 2 +-
docs/formatnode.html.in | 69
Signed-off-by: Boris Fiuczynski
---
docs/formatnode.html.in | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/docs/formatnode.html.in b/docs/formatnode.html.in
index 3fad842c53..cff249fb2b 100644
--- a/docs/formatnode.html.in
+++ b/docs/formatnode.html.in
@@ -157,7 +157,7
Add detection of mdev_types capability to channel subsystem devices.
Signed-off-by: Boris Fiuczynski
Reviewed-by: Bjoern Walk
---
docs/drvnodedev.html.in | 2 +-
docs/formatnode.html.in | 23 -
docs/schemas/nodedev.rng | 4
On Wed, Nov 11, 2020 at 13:04:19 +0100, Michal Privoznik wrote:
> On 11/11/20 11:32 AM, Peter Krempa wrote:
> > On Tue, Nov 10, 2020 at 16:11:40 +0100, Michal Privoznik wrote:
> > > Marc-André posted a patch that implements agent handling. I've written
> > > the rest.
> > >
> > > Marc-André
Some CPUs provide a way to read exact TSC frequency, while measuring it
is required on other CPUs. However, measuring is never exact and the
result may slightly differ across reboots. For this reason both Linux
kernel and QEMU recently started allowing for guests TSC frequency to
fall into +/- 250
On 11/11/20 11:32 AM, Peter Krempa wrote:
On Tue, Nov 10, 2020 at 16:11:40 +0100, Michal Privoznik wrote:
Marc-André posted a patch that implements agent handling. I've written
the rest.
Marc-André Lureau (1):
qemu_agent: add qemuAgentSSH{Add,Remove,Get}AuthorizedKeys
Michal Prívozník (5):
On 11/6/20 4:32 AM, Matt Coleman wrote:
Most of this is making functions void that unnecessarily return an int.
It also includes some conversion to GLib.
Feel free to squash related commits, if you'd like. I left them separate
to make it easier to review.
Matt Coleman (28):
domain_conf:
On Wed, Nov 11, 2020 at 12:43:34 +0100, Jiri Denemark wrote:
> The unsignedInt XML schema type allows for values up to 2^32 - 1, i.e.,
> using 4294967296 or greater TSC frequency would fail schema validation.
>
> Signed-off-by: Jiri Denemark
> ---
Reviewed-by: Peter Krempa
The unsignedInt XML schema type allows for values up to 2^32 - 1, i.e.,
using 4294967296 or greater TSC frequency would fail schema validation.
Signed-off-by: Jiri Denemark
---
docs/schemas/domaincommon.rng | 2 +-
.../cpu-tsc-high-frequency.x86_64-latest.args | 38
> On Nov 9, 2020, at 9:43 AM, Michal Privoznik wrote:
>
> Nice cleanup. But as I say in 03/28 I think we might want glib adoption to be
> done in bigger chunks. Usually we rewrite VIR_ALLOC/VIR_REALLOC_N ->
> g_new0()/g_renew() in one patch (might be coupled with g_free() except
> VIR_FREE()
Signed-off-by: Matt Coleman
---
NEWS.rst | 7 +++
1 file changed, 7 insertions(+)
diff --git a/NEWS.rst b/NEWS.rst
index 0e56f5dbca..8b60a5da11 100644
--- a/NEWS.rst
+++ b/NEWS.rst
@@ -13,6 +13,13 @@ v6.10.0 (unreleased)
* **New features**
+ * hyperv: implement new APIs
+
+The
Co-authored-by: Sri Ramanujam
Signed-off-by: Matt Coleman
---
src/hyperv/hyperv_driver.c | 83 ++
1 file changed, 83 insertions(+)
diff --git a/src/hyperv/hyperv_driver.c b/src/hyperv/hyperv_driver.c
index e05c10d194..9fda0d6047 100644
---
* use hypervMsvmVSMSModifyResourceSettings()
* improve the error message: say which property it failed to set
* remove usage of VIR_FREE()
Signed-off-by: Matt Coleman
---
src/hyperv/hyperv_driver.c | 26 +-
1 file changed, 5 insertions(+), 21 deletions(-)
diff --git
Co-authored-by: Sri Ramanujam
Signed-off-by: Matt Coleman
---
src/hyperv/hyperv_driver.c | 24 +++-
1 file changed, 19 insertions(+), 5 deletions(-)
diff --git a/src/hyperv/hyperv_driver.c b/src/hyperv/hyperv_driver.c
index e779b67d00..6d7bc4b80f 100644
---
Signed-off-by: Matt Coleman
---
src/hyperv/hyperv_wmi.c | 37 +
src/hyperv/hyperv_wmi.h | 8
2 files changed, 45 insertions(+)
diff --git a/src/hyperv/hyperv_wmi.c b/src/hyperv/hyperv_wmi.c
index 72a68c0591..866b347bc2 100644
---
This matches their placement in struct _virHypervisorDriver.
Signed-off-by: Matt Coleman
---
src/hyperv/hyperv_driver.c | 144 ++---
1 file changed, 72 insertions(+), 72 deletions(-)
diff --git a/src/hyperv/hyperv_driver.c b/src/hyperv/hyperv_driver.c
index
Here's a GitLab merge request, if you'd prefer to review it there:
https://gitlab.com/iammattcoleman/libvirt/-/merge_requests/9
Matt Coleman (7):
hyperv: implement domainGetMaxMemory
hyperv: move hypervDomainSetMemory and hypervDomainSetMemoryFlags
hyperv: implement domainSetMaxMemory
Co-authored-by: Sri Ramanujam
Signed-off-by: Matt Coleman
---
src/hyperv/hyperv_driver.c | 28
1 file changed, 28 insertions(+)
diff --git a/src/hyperv/hyperv_driver.c b/src/hyperv/hyperv_driver.c
index b789ef58e8..67b8b771d9 100644
--- a/src/hyperv/hyperv_driver.c
On Tue, Nov 10, 2020 at 16:11:40 +0100, Michal Privoznik wrote:
> Marc-André posted a patch that implements agent handling. I've written
> the rest.
>
> Marc-André Lureau (1):
> qemu_agent: add qemuAgentSSH{Add,Remove,Get}AuthorizedKeys
>
> Michal PrÃvoznÃk (5):
> Introduce OpenSSH
On Wed, Nov 11, 2020 at 11:15:33AM +0100, Boris Fiuczynski wrote:
> On 11/11/20 10:39 AM, Erik Skultety wrote:
> > On Tue, Nov 10, 2020 at 07:09:06PM +0100, Boris Fiuczynski wrote:
> > > Add detection of mdev_types capability to channel subsystem devices.
> > >
> > > Signed-off-by: Boris
On 11/11/20 10:39 AM, Erik Skultety wrote:
On Tue, Nov 10, 2020 at 07:09:06PM +0100, Boris Fiuczynski wrote:
Add detection of mdev_types capability to channel subsystem devices.
Signed-off-by: Boris Fiuczynski
Reviewed-by: Bjoern Walk
---
docs/drvnodedev.html.in | 5
On Tue, Nov 10, 2020 at 21:12:33 +0100, Michal Privoznik wrote:
> On 11/7/20 10:41 AM, Hao Wang wrote:
> > Introduce virDomainDirtyRateInfo structure used for domain's memory dirty
> > rate query.
> >
> > Signed-off-by: Hao Wang
> > Reviewed-by: Chuan Zheng
> > ---
> >
On 11/11/20 10:34 AM, Erik Skultety wrote:
On Tue, Nov 10, 2020 at 07:09:05PM +0100, Boris Fiuczynski wrote:
To prevent copying the mdev_types description multiple times
it is refactored into a new paragraph for easy reuse.
Signed-off-by: Boris Fiuczynski
---
docs/formatnode.html.in | 70
On Tue, Nov 10, 2020 at 07:09:06PM +0100, Boris Fiuczynski wrote:
> Add detection of mdev_types capability to channel subsystem devices.
>
> Signed-off-by: Boris Fiuczynski
> Reviewed-by: Bjoern Walk
> ---
> docs/drvnodedev.html.in | 5 +-
> docs/formatnode.html.in
On Tue, Nov 10, 2020 at 07:09:05PM +0100, Boris Fiuczynski wrote:
> To prevent copying the mdev_types description multiple times
> it is refactored into a new paragraph for easy reuse.
>
> Signed-off-by: Boris Fiuczynski
> ---
> docs/formatnode.html.in | 70
The first patch fixes the actual problem, the second is just a
cleanup I've came up with whilst looking at the code.
Michal Prívozník (2):
virnetdevopenvswitch: Get names for dpdkvhostuserclient too
virnetdevopenvswitch: Simplify OVS_VSCTL cmd creation
src/qemu/qemu_command.c | 1 +
There are two type of vhostuser ports:
dpdkvhostuser - OVS creates the socket and QEMU connects to it
dpdkvhostuserclient - QEMU creates the socket and OVS connects to it
But of course ovs-vsctl syntax for fetching ifname is different.
So far, we've implemented the former. The lack of
Every time we create new virCommand of OVS_VSCTL it must be
followed by virNetDevOpenvswitchAddTimeout() call which adds the
--timeout=X argument to freshly created cmd. Instead of having
this as two separate function calls it can be just one.
Signed-off-by: Michal Privoznik
---
On Tue, Nov 10, 2020 at 16:11:43 +0100, Michal Privoznik wrote:
> The new virsh commands are:
>
> get-user-sshkeys
> set-user-sshkeys
>
> Signed-off-by: Michal Privoznik
> ---
> docs/manpages/virsh.rst | 37 ++
> tools/virsh-domain.c| 152
On Tue, Nov 10, 2020 at 16:11:42 +0100, Michal Privoznik wrote:
> Since both APIs accept/return an array of strings we can't have
> client/server dispatch code generated. But implementation is
> fairly trivial, although verbose.
>
> Signed-off-by: Michal Privoznik
> ---
>
On Tue, Nov 10, 2020 at 16:11:41 +0100, Michal Privoznik wrote:
> When setting up a new guest or when a management software wants
> to allow access to an existing guest the
> virDomainSetUserPassword() API can be used, but that might be not
> good enough if user want to ssh into the guest. Not
69 matches
Mail list logo