Should be followed with qemuDomainObjExitMonitor only if
qemuDomainObjEnterMonitorAsync returns 0.
---
src/qemu/qemu_migration.c | 2 +-
src/qemu/qemu_process.c | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/qemu/qemu_migration.c b/src/qemu/qemu_migration.c
index c23f
On Wed, Jul 12, 2017 at 05:09:37PM +0530, Shivaprasad G Bhat wrote:
>
>
> On 07/12/2017 04:25 PM, Andrea Bolognani wrote:
> > [libvir-list added to the loop]
> >
> > On Tue, 2017-07-04 at 10:47 +0200, Greg Kurz wrote:
> > > On Tue, 4 Jul 2017 17:29:01 +1000 David Gibson
> > > wrote:
> > > > On
On Mon, Jul 17, 2017 at 4:50 PM, John Ferlan wrote:
>
>
> On 07/12/2017 06:09 PM, ashish mittal wrote:
>> On Fri, Jun 30, 2017 at 1:29 PM, John Ferlan wrote:
>>>
>>>
>>> On 06/29/2017 10:02 PM, Ashish Mittal wrote:
From: Ashish Mittal
Add a new TLS X.509 certificate type - "vxhs".
@map_device, allocated by virAsprintf in virStorageBackendCreateVols,
was not freed and leaked.
Signed-off-by: Zhipeng Lu
---
src/storage/storage_backend_mpath.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/src/storage/storage_backend_mpath.c
b/src/storage/storage_backend_mpath.c
index 6
@remote_file, allocated by virAsprintf, was not freed and leaked.
Signed-off-by: Zhipeng Lu
---
src/phyp/phyp_driver.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/src/phyp/phyp_driver.c b/src/phyp/phyp_driver.c
index 9121581..662e2f7 100644
--- a/src/phyp/phyp_driver.c
+++ b/src/phyp/phy
On 06/28/2017 02:49 PM, Cole Robinson wrote:
> This allows drivers to set their own default. But if a driver neglects
> to fill one in, we still error like we previously would at parse time.
>
> Signed-off-by: Cole Robinson
> ---
> src/conf/domain_conf.c | 29 +
> 1
On 06/28/2017 02:49 PM, Cole Robinson wrote:
> And not generic domain_conf code. We will need qemu private functions
> in a bit.
>
> Signed-off-by: Cole Robinson
> ---
> src/conf/domain_conf.c | 3 ---
> src/qemu/qemu_domain.c | 7 +++
> 2 files changed, 7 insertions(+), 3 deletions(-)
>
On 06/28/2017 02:49 PM, Cole Robinson wrote:
> arm/aarch64 -M virt on KVM doesn't and will never work with standard
> VGA card emulation. The recommended method is to use type=virtio, so
> let's make it the default for video devices without an explicit type
> set by the user
>
> https://bugzilla
On 06/28/2017 02:49 PM, Cole Robinson wrote:
> Will be needed for future patches to pull the default video type
> setting out of XML parsing routines.
>
> Signed-off-by: Cole Robinson
> ---
> src/conf/domain_conf.c | 1 +
> src/conf/domain_conf.h | 1 +
> src/qemu/qem
On 06/28/2017 02:49 PM, Cole Robinson wrote:
> To handle setting a default heads value. Convert callers that were
> doing it by hand
>
> Signed-off-by: Cole Robinson
> ---
> src/conf/domain_conf.c| 21 -
> src/conf/domain_conf.h| 1 +
> src/libvirt_private.
On 06/28/2017 02:49 PM, Cole Robinson wrote:
> For the ram/vram monitor wrappers, just add a default: clause...
> seems like it should be rarely extended so this saves every committer
> from needing to update
>
> For the validation switch, fill in the missing values
>
> Signed-off-by: Cole Robi
On 06/28/2017 02:49 PM, Cole Robinson wrote:
> And into DeviceDefValidate which is the expected place
>
> Signed-off-by: Cole Robinson
> ---
> src/qemu/qemu_domain.c | 55
> +++---
> 1 file changed, 30 insertions(+), 25 deletions(-)
>
Reviewed-by:
On 06/28/2017 02:49 PM, Cole Robinson wrote:
> The ram/vram = 0 bits aren't needed, and PostParse will fill in the
> needed QXL default
>
> Signed-off-by: Cole Robinson
> ---
> src/qemu/qemu_command.c | 4
> src/qemu/qemu_parse_command.c | 11 +--
> 2 files changed, 1 inser
On 07/18/2017 03:18 PM, Laine Stump wrote:
On 07/18/2017 01:42 PM, Jim Fehlig wrote:
/domain/devices/disk/driver/@name is not a required or mandatory
attribute according to formatdomain, and indeed it was agreed on
IRC that the attribute is "optional for input, recommended (but
not required) for
On 07/18/2017 01:42 PM, Jim Fehlig wrote:
> /domain/devices/disk/driver/@name is not a required or mandatory
> attribute according to formatdomain, and indeed it was agreed on
> IRC that the attribute is "optional for input, recommended (but
> not required) for output". Currently the schema require
On 07/11/2017 05:05 AM, Michal Privoznik wrote:
> The purpose of this function is to tell if the current position
> in given FD is in data section or a hole and how much bytes there
> is remaining until the end of the section. This is achieved by
> couple of lseeks(). The most important part is t
Now that nwfilters object list is self locking, it's no longer
necessary to hold the driver level lock for certain API's.
However, for the DefineXML, Undefine, and Reload processing keeping
that lock ensures for serialization required in order to process
the filter Instantiation properly.
Signed-
Now that we have a bit more control, let's convert our object into a
lockable object and let that magic handle the create, lock/unlock,
and reference counting.
Because we have the need for instantiation in a recursive manner,
introduce the virObjectTryLock to handle the virNWFilterObjTryLock
proce
The current algorithm required usage of recursive locks because
there was no other mechanism available to ensure that the object
wasn't deleted whilst the instantiation was taking place.
Unfortunately common objects (virObjectLockable) do not allow for
recursive locks; however, since this is a ver
Commit id '46a811db07' added code to check if the filter by Name
already existed with a different UUID, to go along with the existing
found filter by UUID and compare the Names match thus making it
impossible to reach this find by Name condition without both the
Name and UUID already matching, so r
Perhaps a bit out of order from the normal convert driver object to
virObjectLockable, then convert the driver object list. However, as
it turns out nwfilter objects have been using a recursive mutex for
which the common virObject code does not want to use.
So, if we convert the nwfilter object li
v1: https://www.redhat.com/archives/libvir-list/2017-June/msg00079.html
Changes since v1 (if I can recall all of them!):
Patches 1, 4, 8-13 were pushed
Patches 2, 3, 5-7 are dropped
This this is a rework of patches 14-17
Patch 1 (former patch14):
* Requested adjustments made to ACK'd patch, bu
"Simple" conversion to the virObjectLockable object isn't quite possible
yet because the nwfilter lock requires usage of recursive locking due to
algorithms handing filter ""'s and ""'s. The list search
logic would also benefit from using hash tables for lookups. So this patch
is step 1 in the proc
ping?
Tks -
John
On 06/29/2017 10:32 AM, John Ferlan wrote:
> https://bugzilla.redhat.com/show_bug.cgi?id=1458630
>
> Introduce virQEMUDriverConfigSetCertDir which will handle reading the
> qemu.conf config file specific setting for default, vnc, spice, chardev,
> and migrate. If a setting is
2017-07-18 18:54 GMT+02:00 Andrea Bolognani :
> On Tue, 2017-06-27 at 15:13 -0400, Sri Ramanujam wrote:
>> This commit adds support for invoking methods on remote objects
>> via hypervInvokeMethod.
>> ---
>> src/hyperv/hyperv_wmi.c | 590
>>
>> src
On 07/18/2017 11:10 AM, Laine Stump wrote:
On 07/17/2017 10:42 AM, Christian Ehrhardt wrote:
In recent libvirt libxl does fill non-existant driver elements with type='raw',
but lacks the name='qemu' attribute that it had in the past.
Without that the xml fails to validate, therefore old XML file
/domain/devices/disk/driver/@name is not a required or mandatory
attribute according to formatdomain, and indeed it was agreed on
IRC that the attribute is "optional for input, recommended (but
not required) for output". Currently the schema requires the
attribute, causing virt-xml-validate to fail
Hi Roman,
Sorry for the delay (I didn't setup a correct filter for the mailing list and
missed your follow-up mail)
Updated patch looks awesome. thank you for your effort!
> Note 1: while adding tests I noticed that port allocator will actually
> skip already bound ports, so I'm wondering
On 07/17/2017 10:42 AM, Christian Ehrhardt wrote:
> In recent libvirt libxl does fill non-existant driver elements with
> type='raw',
> but lacks the name='qemu' attribute that it had in the past.
> Without that the xml fails to validate, therefore old XML files can't be
> edited/defined anymore.
On Tue, 2017-06-27 at 15:13 -0400, Sri Ramanujam wrote:
> This commit adds support for invoking methods on remote objects
> via hypervInvokeMethod.
> ---
> src/hyperv/hyperv_wmi.c | 590
>
> src/hyperv/hyperv_wmi.h | 8 +-
> src/hyperv/openwsman.h
On Tue, Jul 18, 2017 at 12:51 PM, Peter Krempa wrote:
> The refactor to split up storage driver into modules broke the apparmor
> helper program, since that did not initialize the storage driver
> properly and thus detection of the backing chain could not work.
>
> Register the storage driver bac
On Mon, Jun 26, 2017 at 06:01:20PM +0200, Ján Tomko wrote:
> Introduced by commit 0832c58, with the intention to link with
> the stack protector library.
>
> Another instance introduced by commit 4cbc15d which separated
> commandhelper_LDADD from LDADDS.
>
> Not needed because per commit 71b54636
v1: https://www.redhat.com/archives/libvir-list/2017-July/msg00029.html
Changes in v2:
Rewrote patch 1 to resolve (new) bz1472277. This can also be cherry
picked into 3.2-maint (and any others that get created).
Added patch 2 - it's obvious why
Adjusted patch 3 to make the change in storage rat
https://bugzilla.redhat.com/show_bug.cgi?id=1458708
When originally designed in commit id '42a021c1', providing the
'parent' attribute to the was checked to make sure that the "parent" of
the wwnn/wwpn matched that of the provided parent "just in case"
someone created the node device first, then
https://bugzilla.redhat.com/show_bug.cgi?id=1472277
Commit id '106930aaa' altered the order of checking for an existing
vHBA (e.g something created via nodedev-create functionality outside
of the storage pool logic) which inadvertantly broke the code to
decide whether to alter/force the fchost->ma
It's no longer needed since the checkParent code moved back to
storage_backend_scsi.c
Signed-off-by: John Ferlan
---
src/conf/node_device_conf.c| 8 +++-
src/conf/node_device_conf.h| 3 +--
src/storage/storage_backend_scsi.c | 2 +-
3 files changed, 5 insertions(+), 8 deletio
On Wed, 28 Jun 2017 15:21:29 +0100
"Daniel P. Berrange" wrote:
> On Wed, Jun 28, 2017 at 03:56:36PM +0200, Wim Ten Have wrote:
> > From: Wim ten Have
> >
> > Add libvirtd NUMA cell domain administration functionality to
> > describe underlying cell id sibling distances in full fashion
> > when
On 07/18/2017 03:03 AM, Erik Skultety wrote:
> On Mon, Jul 03, 2017 at 02:52:05PM -0400, John Ferlan wrote:
>> Only alter the managed setting if it wasn't provided. If someone provided
>> 'no', then honor that rather than overwriting.
>
> IMO this deserves to be tracked by a BZ, since the curren
On Tue, Jul 18, 2017 at 10:34:09AM +0200, Michal Privoznik wrote:
In Wireshark commit of 7cd6906056922e4b8 (contained in v2.4.0)
the tvb_new_subset() function was renamed to
tvb_new_subset_length_caplen(). However, we can take the extra
step and rename to tvb_new_subset_remaining() directly (see
On Mon, Jul 10, 2017 at 14:46:48 +0200, Pavel Hrdina wrote:
> This will store private data that will be used by following patches
> when switching to virFileCache.
>
> Signed-off-by: Pavel Hrdina
> ---
> src/qemu/qemu_capabilities.c | 51
> ++--
> src/qem
On Mon, Jul 10, 2017 at 14:46:47 +0200, Pavel Hrdina wrote:
> Preparation for switching to virFileCache where there are two callbacks,
> one to get a new data and second one to load a cached data.
>
> This also removes virQEMUCapsReset which is no longer required.
>
> Signed-off-by: Pavel Hrdina
On Mon, Jul 10, 2017 at 14:46:46 +0200, Pavel Hrdina wrote:
> It's not required and following patches will change the code.
>
> Signed-off-by: Pavel Hrdina
> ---
> src/qemu/qemu_capabilities.c | 29 +++--
> 1 file changed, 11 insertions(+), 18 deletions(-)
>
> diff --git
On Mon, Jul 10, 2017 at 11:12:37AM +0200, Boris Fiuczynski wrote:
Adjust qemu 2.9 s390 capabilites xml and enable qemu capabilities test.
Signed-off-by: Boris Fiuczynski
---
V2: removed s390x 2.9.50 caps and added 2.9.0 caps (Jan)
.../qemucapabilitiesdata/caps_2.9.0.s390x.replies | 14445
Commit id '9c5d98fd8' missed changing this call to use driver->devs
rather than @driver->devs.
Signed-off-by: John Ferlan
---
Pushed under build breaker rule (debian-8 and centos-6)
src/node_device/node_device_hal.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/node_
[Adding MST who wrote qemu patches]
On 07/18/2017 12:21 PM, Peter Krempa wrote:
> On Tue, Jul 18, 2017 at 09:01:31 +0200, Michal Privoznik wrote:
>> On 07/18/2017 08:23 AM, Peter Krempa wrote:
>>> On Mon, Jul 17, 2017 at 15:39:56 +0200, Michal Privoznik wrote:
https://bugzilla.redhat.com/show
On Mon, Jul 03, 2017 at 02:52:06PM -0400, John Ferlan wrote:
> https://bugzilla.redhat.com/show_bug.cgi?id=1458708
>
> When originally designed in commit id '42a021c1', providing the 'parent'
> attribute to the
> was checked to make sure that the "parent" of the wwnn/wwpn matched that
> of the pro
Roman Bogorodskiy wrote:
> From: Alexander Nusov
>
> This patch adds support for automatic VNC port assignment for bhyve guests.
> ---
> Changes from v1:
>
> * Call virPortAllocatorRelease() in virBhyveProcessStop() to release
>VNC port; that's done unconditionally of using autoport
> *
The refactor to split up storage driver into modules broke the apparmor
helper program, since that did not initialize the storage driver
properly and thus detection of the backing chain could not work.
Register the storage driver backends explicitly. Unfortunately it's now
necessary to link with t
On Tue, Jul 18, 2017 at 09:01:31 +0200, Michal Privoznik wrote:
> On 07/18/2017 08:23 AM, Peter Krempa wrote:
> > On Mon, Jul 17, 2017 at 15:39:56 +0200, Michal Privoznik wrote:
> >> https://bugzilla.redhat.com/show_bug.cgi?id=1462653
> >
> > This bugzilla is not public.
>
> Okay, I'll drop it fr
In Wireshark commit of 7cd6906056922e4b8 (contained in v2.4.0)
the tvb_new_subset() function was renamed to
tvb_new_subset_length_caplen(). However, we can take the extra
step and rename to tvb_new_subset_remaining() directly (see
Wireshark commit 0ecfc7280cf3d7). The reasoning is that there is
no
On Wed, 2017-07-12 at 17:09 +0530, Shivaprasad G Bhat wrote:
> Setting numa_node for emulated devices is the benefit for now. On x86, I
> figured there is
> no way to set the numa_node for the root controller and the emulated
> devices sitting there
> all have numa_node set to -1. Only the devices
On Mon, Jul 17, 2017 at 12:42:12PM -0400, Cole Robinson wrote:
> On 07/17/2017 12:35 PM, Daniel P. Berrange wrote:
> > On Mon, Jul 17, 2017 at 12:31:50PM -0400, Cole Robinson wrote:
> >> For a logged in user this a path like /dev/dri/renderD128 will have
> >> default ownership root:video which won'
On Mon, Jul 03, 2017 at 02:52:05PM -0400, John Ferlan wrote:
> Only alter the managed setting if it wasn't provided. If someone provided
> 'no', then honor that rather than overwriting.
IMO this deserves to be tracked by a BZ, since the current behaviour is in
contrast to what the documentation sa
On Tue, Jul 18, 2017 at 08:57:36 +0200, Christian Ehrhardt wrote:
> On Mon, Jul 17, 2017 at 8:40 PM, Christian Ehrhardt <
> christian.ehrha...@canonical.com> wrote:
>
> >
> >
> > On Mon, Jul 17, 2017 at 8:17 PM, Christian Ehrhardt <
> > christian.ehrha...@canonical.com> wrote:
> >>
> >> So it is t
On 07/18/2017 08:23 AM, Peter Krempa wrote:
> On Mon, Jul 17, 2017 at 15:39:56 +0200, Michal Privoznik wrote:
>> https://bugzilla.redhat.com/show_bug.cgi?id=1462653
>
> This bugzilla is not public.
Okay, I'll drop it from the commit message.
>
>>
>> Just like I've added support for setting rx_q
55 matches
Mail list logo