On 9/9/19 5:52 PM, Eric Blake wrote:
Commit f10562799 introduced a regression: if reverting to a snapshot
fails early (such as when we refuse to revert to an external
snapshot), we lose track of the domain's current snapshot.
See: https://bugzilla.redhat.com/1738747
Signed-off-by: Eric Blake
Commit f10562799 introduced a regression: if reverting to a snapshot
fails early (such as when we refuse to revert to an external
snapshot), we lose track of the domain's current snapshot.
See: https://bugzilla.redhat.com/1738747
Signed-off-by: Eric Blake
---
src/qemu/qemu_driver.c | 2 --
1
On 9/9/19 1:00 PM, Michal Privoznik wrote:
Signed-off-by: Michal Privoznik
---
Reviewed-by: Daniel Henrique Barboza
src/bhyve/bhyve_conf.c| 10 +---
src/libvirt-admin.c | 3 +-
src/libvirt.c | 4 +-
src/libxl/libxl_conf.c
On 9/9/19 1:00 PM, Michal Privoznik wrote:
Signed-off-by: Michal Privoznik
---
Reviewed-by: Daniel Henrique Barboza
src/lxc/lxc_native.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/lxc/lxc_native.c b/src/lxc/lxc_native.c
index 018eec6977..bb8c359147 100644
On 9/9/19 1:00 PM, Michal Privoznik wrote:
Signed-off-by: Michal Privoznik
---
Reviewed-by: Daniel Henrique Barboza
src/qemu/qemu_conf.c | 139 ---
1 file changed, 63 insertions(+), 76 deletions(-)
diff --git a/src/qemu/qemu_conf.c
On 9/9/19 1:00 PM, Michal Privoznik wrote:
Signed-off-by: Michal Privoznik
---
make syntax-check wasn't happy about this patch, at least in my
box. This adjustment make it happy again:
[danielhb@rekt libvirt]$ git diff
diff --git a/src/qemu/qemu_conf.c b/src/qemu/qemu_conf.c
index
On 9/9/19 1:00 PM, Michal Privoznik wrote:
Our naming rules prefer qemuObjectOperation() scheme rather than
qemuOperationObject() for function names. These were not honoured
in recent commits to qemu_conf.c.
Signed-off-by: Michal Privoznik
---
Reviewed-by: Daniel Henrique Barboza
On 9/9/19 1:00 PM, Michal Privoznik wrote:
There are two 'cleanup' labels - one in
virQEMUDriverConfigHugeTLBFSInit() and the other in
virQEMUDriverConfigSetDefaults() that do nothing more than
return and integer value. No memory freeing or anything important
is done there. Drop them in
AppArmorGetSecurityProcessLabel copies the VM's profile name to the
label member of virSecurityLabel struct. If the profile is not loaded,
the name is set empty before calling virStrcpy to copy it. However,
virStrcpy will fail if src is empty (0 length), causing
AppArmorGetSecurityProcessLabel to
Our naming rules prefer qemuObjectOperation() scheme rather than
qemuOperationObject() for function names. These were not honoured
in recent commits to qemu_conf.c.
Signed-off-by: Michal Privoznik
---
src/qemu/qemu_conf.c | 18 +-
1 file changed, 9 insertions(+), 9 deletions(-)
Signed-off-by: Michal Privoznik
---
src/qemu/qemu_conf.c | 139 ---
1 file changed, 63 insertions(+), 76 deletions(-)
diff --git a/src/qemu/qemu_conf.c b/src/qemu/qemu_conf.c
index c3255a6f54..f805991872 100644
--- a/src/qemu/qemu_conf.c
+++
Signed-off-by: Michal Privoznik
---
src/lxc/lxc_native.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/src/lxc/lxc_native.c b/src/lxc/lxc_native.c
index 018eec6977..bb8c359147 100644
--- a/src/lxc/lxc_native.c
+++ b/src/lxc/lxc_native.c
@@ -1190,7 +1190,7 @@
Signed-off-by: Michal Privoznik
---
src/qemu/qemu_conf.c | 130 +--
1 file changed, 40 insertions(+), 90 deletions(-)
diff --git a/src/qemu/qemu_conf.c b/src/qemu/qemu_conf.c
index f11df03cf8..c3255a6f54 100644
--- a/src/qemu/qemu_conf.c
+++
There are two 'cleanup' labels - one in
virQEMUDriverConfigHugeTLBFSInit() and the other in
virQEMUDriverConfigSetDefaults() that do nothing more than
return and integer value. No memory freeing or anything important
is done there. Drop them in favour of returning immediately.
Signed-off-by:
Signed-off-by: Michal Privoznik
---
src/bhyve/bhyve_conf.c| 10 +---
src/libvirt-admin.c | 3 +-
src/libvirt.c | 4 +-
src/libxl/libxl_conf.c| 22 +++-
src/libxl/libxl_driver.c | 8 +--
src/libxl/xen_xl.c|
Michal Prívozník (6):
qemu_conf.c: Fix naming of *AddRemove* functions
qemu_conf: Drop a pair of needless 'cleanup' labels
qemu_conf: Use more of VIR_AUTOFREE()
qemu_conf: Use more of VIR_AUTOUNREF()
lxcParseConfigString: Don't return success if post parse callback
fails
lib:
On 8/29/19 9:18 PM, Daniel Henrique Barboza wrote:
changes in v3:
- rebased after all PCI test changes went to master.
- a couple of patches were dropped due to redundancy with what
Michal already pushed upstream.
- patches 3 and 5 were already R-b by Michal in v2.
Hi,
In this new version I'm
On 8/29/19 9:18 PM, Daniel Henrique Barboza wrote:
This patch adds mock of the /dev/vfio path, needed for proper
implementation of the support for multifunction/multiple devices
per iommu groups.
To do that, the existing bind and unbind operations were adapted
to operate with the mocked
In f08e6883cb4 I've made @pcidevs in
virHostdevReAttachPCIDevices() to be automatically unrefed using
VIR_AUTOUNREF() but I forgot to remove the line that explicitly
unrefs the object at the end of the function.
Signed-off-by: Michal Privoznik
---
Pushed as trivial.
src/util/virhostdev.c | 1
On Tue, Sep 03, 2019 at 10:09:48PM +0200, Eric Farman wrote:
> When starting a domain, we use the presence of a vfio-pci or
> mdev hostdev to determine if the memlock maximum needs to be
> increased. But if we hotplug either of these devices, only the
> vfio-pci path gets that love. This means
On Mon, 2019-09-09 at 13:49 +0200, Fabiano Fidêncio wrote:
> Let's refresh the archived Dockerfiles as "locales" dependency has
> to be added to all Debian, Ubuntu, and Fedora files.
>
> Signed-off-by: Fabiano Fidêncio
> ---
> buildenv-libosinfo-debian-10.zip | Bin 559 -> 629
On Mon, 2019-09-09 at 13:47 +0200, Fabiano Fidêncio wrote:
> osinfo-db tests require "en_US.UTF-8" locate to be set. Since commit
s/locate/locale/
> 3b682c416b0d the locales are installed but we have to actually generate
> it for the containers.
>
> The reason we don't have to do that for the
On Mon, Sep 09, 2019 at 02:11:36PM +0100, Daniel P. Berrangé wrote:
> On Mon, Sep 09, 2019 at 03:47:21PM +0400, Marc-André Lureau wrote:
> > Hi
> >
> > On Fri, Sep 6, 2019 at 7:37 PM Michal Privoznik wrote:
> > >
> > > This reverts commit 39dded7bb61444bb608fadd3f82f6fe93d08fd0e.
> > >
> > >
On Mon, Sep 09, 2019 at 03:47:21PM +0400, Marc-André Lureau wrote:
> Hi
>
> On Fri, Sep 6, 2019 at 7:37 PM Michal Privoznik wrote:
> >
> > This reverts commit 39dded7bb61444bb608fadd3f82f6fe93d08fd0e.
> >
> > This commit broke virpolkittest on Ubuntu 18 which has an old
> > dbus (v1.12.2). Any
On 9/9/19 9:55 AM, Michal Privoznik wrote:
On 9/4/19 1:06 AM, Daniel Henrique Barboza wrote:
After the previous commits, qemuAddSharedDevice() and
qemuRemoveSharedDevice() are now the same code with a different
flag to call the internal functions.
This patch aggregates the common code into a
On Mon, 2019-09-09 at 13:18 +0200, Fabiano Fidêncio wrote:
> On Mon, Sep 9, 2019 at 12:00 PM Andrea Bolognani wrote:
> > Ideally we'd do this not just for containers, but for virtual
> > machines as well; in fact, I'm quite surprised the osinfo-db-check
> > job it didn't start failing there too.
On 9/4/19 1:06 AM, Daniel Henrique Barboza wrote:
Following the same idea of avoid code repetition from the
previous patch, this commit introduces a new function that
aggregates the functions of qemuAddSharedDisk() and
qemuRemoveSharedDisk() into a single place, using a flag to
switch between
On 9/4/19 1:06 AM, Daniel Henrique Barboza wrote:
qemuAddSharedHostdev() has a code similar to
qemuRemoveSharedHostdev(), with exception of one line that
defines the operation (add or remove).
This patch introduces a new function that aggregates the common
code, using a flag to switch between
On 9/4/19 1:06 AM, Daniel Henrique Barboza wrote:
After the previous commits, qemuAddSharedDevice() and
qemuRemoveSharedDevice() are now the same code with a different
flag to call the internal functions.
This patch aggregates the common code into a new function called
On Tue, Sep 03, 2019 at 08:06:04PM -0300, Daniel Henrique Barboza wrote:
> Stumbled on this while changing qemu_conf.c for other
> reasons. Moved code around here and there to avoid
> repetition.
>
> Final result is 23 less lines for the same functionality,
> which is nothing to write home about
Hi
On Mon, Sep 9, 2019 at 2:49 PM Ján Tomko wrote:
>
> On Mon, Sep 09, 2019 at 11:23:59AM +0200, Michal Privoznik wrote:
> >On 9/6/19 5:55 PM, Ján Tomko wrote:
> >>[cc-ing Marc-André]
> >>
> >>On Fri, Sep 06, 2019 at 04:25:19PM +0200, Michal Privoznik wrote:
> >>>The qemu side is not merged in
Let's refresh the archived Dockerfiles as "locales" dependency has
to be added to all Debian, Ubuntu, and Fedora files.
Signed-off-by: Fabiano Fidêncio
---
buildenv-libosinfo-debian-10.zip | Bin 559 -> 629 bytes
buildenv-libosinfo-debian-sid.zip | Bin 559 -> 629
Hi
On Fri, Sep 6, 2019 at 7:37 PM Michal Privoznik wrote:
>
> This reverts commit 39dded7bb61444bb608fadd3f82f6fe93d08fd0e.
>
> This commit broke virpolkittest on Ubuntu 18 which has an old
> dbus (v1.12.2). Any other distro with the recent one works
> (v1.12.16) which hints its a bug in dbus
osinfo-db tests require "en_US.UTF-8" locate to be set. Since commit
3b682c416b0d the locales are installed but we have to actually generate
it for the containers.
The reason we don't have to do that for the virtual machines is because
the preseed file used for installed them already forces it:
We want to keep this flag across libvirtd restarts.
Signed-off-by: Nikolay Shirokovskiy
---
src/conf/domain_conf.c | 6 ++
1 file changed, 6 insertions(+)
diff --git a/src/conf/domain_conf.c b/src/conf/domain_conf.c
index c200af050c..862ca4bd3a 100644
--- a/src/conf/domain_conf.c
+++
It is possible for libvirtd to go down before DEVICE_DELETED event is
delivered upon usb hostdev unplug and to receive the event after the
libvirtd is up. In order to handle this case we need to save
usb hostdev deleteAction is status file.
Signed-off-by: Nikolay Shirokovskiy
---
Somebody can easily unplug usb device from host while libvirtd is being
stopped. Also usb device can be plugged or unplugged/plugged back and so
forth. Let's handle such cases.
Signed-off-by: Nikolay Shirokovskiy
---
src/qemu/qemu_process.c | 55 +
1 file
I guess this is the missing piece for [1]. It did not hurt before (like
we didn't even see any errors/warns in logs) because in
qemuProcessUpdateDevices function virDomainDefFindDevice does not find
device deleted from libvirt config. But now in case of unpluggind usb
device from host we leave
Now when code handling attaching/detaching usb hostdev is appropriately
changed use it to handle host usb device udev add/del events.
Signed-off-by: Nikolay Shirokovskiy
---
src/qemu/Makefile.inc.am | 2 +
src/qemu/qemu_conf.h | 3 +
src/qemu/qemu_domain.c | 2 +
We are going to use qemuDomainAttachHostUSBDevice when
host usb device is plugged back to node.
Signed-off-by: Nikolay Shirokovskiy
---
src/qemu/qemu_hotplug.c | 29 -
1 file changed, 24 insertions(+), 5 deletions(-)
diff --git a/src/qemu/qemu_hotplug.c
Imagine host usb device is unplugged from host and as a result we send
command to qemu to delete appropriate device. Then before qemu device is
deleted host usb device is plugged back. Currenly code supposes there is
no remnant device in qemu and will try to add new device and the attempt
will
*Notes*
Deleting usb device from qemu is synchronous operation (although it
is not stated in qemu API). I did not used this knowledge in the
series.
The last patch is remnant of previus version of the series yet it is useful.
Diff to previous[1] version:
- don't use dummy device while host
If hostdev is unplugged we don't need to delete it's correspondent device
from qemu etc. Just remove it from config immediately.
Signed-off-by: Nikolay Shirokovskiy
---
src/qemu/qemu_hotplug.c | 16 +++-
src/util/virhostdev.c | 2 ++
2 files changed, 17 insertions(+), 1
First I don't want to add code to handle dummy device that is used when
host usb device is not present at the moment of starting/migrating etc.
Second supporting non mandatory policies would require to handle races
when host usb device is plugged to host and libvirtd starts adding
device but if in
Handle host usb device unplug in DEVICE_DELETED handle execution
path.
Signed-off-by: Nikolay Shirokovskiy
---
src/qemu/qemu_hotplug.c | 38 +++---
1 file changed, 31 insertions(+), 7 deletions(-)
diff --git a/src/qemu/qemu_hotplug.c b/src/qemu/qemu_hotplug.c
We are going to call qemuDomainDetachDeviceLive when usb device
is unplugged from host. But later when DEVICE_DELETED event is delivered
we need to keep device in libvirt config. For this purpuse let's save
delete intention in device config.
Signed-off-by: Nikolay Shirokovskiy
---
On Tue, Sep 03, 2019 at 08:06:05PM -0300, Daniel Henrique Barboza wrote:
> qemuAddSharedHostdev() has a code similar to
> qemuRemoveSharedHostdev(), with exception of one line that
> defines the operation (add or remove).
>
> This patch introduces a new function that aggregates the common
> code,
On Mon, Sep 9, 2019 at 12:00 PM Andrea Bolognani wrote:
>
> On Fri, 2019-09-06 at 14:12 +0200, Fabiano Fidêncio wrote:
> > osinfo-db tests require "en_US.UTF-8" locale to be set. Unfortunately,
> > our containers do not contain the needed locale file.
> >
> > After a discussion on libosinfo
On Fri, Sep 06, 2019 at 09:54:29AM -0300, Daniel Henrique Barboza wrote:
>
>
> On 9/4/19 2:23 PM, Jonathon Jongsma wrote:
> > Since commit fd9ef3b31e, virDomainFindByUUIDRef() no longer exists and
> > all virDomainObjListFindBy*() functions now increment the reference
> > count.
> >
> >
On Thu, 2019-09-05 at 13:34 +0100, Daniel P. Berrangé wrote:
> We currently generate two completely separate API references for the
> libvirt public API. One at 'docs/html/' and one at 'docs/devhelp/'.
> Both are published on the website, but we only link to content in
> the 'docs/html/' pages.
>
On Mon, Sep 09, 2019 at 11:23:59AM +0200, Michal Privoznik wrote:
On 9/6/19 5:55 PM, Ján Tomko wrote:
[cc-ing Marc-André]
On Fri, Sep 06, 2019 at 04:25:19PM +0200, Michal Privoznik wrote:
The qemu side is not merged in yet, so there is a chance that the
interface will change. Don't detect the
On Mon, Sep 09, 2019 at 12:38:25PM +0200, Pavel Hrdina wrote:
> On Thu, Sep 05, 2019 at 11:52:09AM +0100, Daniel P. Berrangé wrote:
> > The Perl bindings for libvirt use the test driver for unit tests. This
> > tries to load the cpu_map/index.xml file, and when run from an
> > uninstalled build
On Thu, Sep 05, 2019 at 11:52:09AM +0100, Daniel P. Berrangé wrote:
> The Perl bindings for libvirt use the test driver for unit tests. This
> tries to load the cpu_map/index.xml file, and when run from an
> uninstalled build will fail.
>
> The problem is that virFileActivateDirOverride is called
On Fri, 2019-09-06 at 14:39 +0200, Fabiano Fidêncio wrote:
> Let's refresh the archived Dockerfiles as "locales" dependency has
> to be added to all Debian, Ubuntu, and Fedora files.
>
> Signed-off-by: Fabiano Fidêncio
> ---
> This commit is available in the following branch:
>
On Fri, 2019-09-06 at 14:12 +0200, Fabiano Fidêncio wrote:
> osinfo-db tests require "en_US.UTF-8" locale to be set. Unfortunately,
> our containers do not contain the needed locale file.
>
> After a discussion on libosinfo mailing list[0], it's been agreed on
> having the locale as part of our
On Mon, Sep 09, 2019 at 10:19:22AM +0100, Daniel P. Berrangé wrote:
> On Fri, Sep 06, 2019 at 11:36:17PM +0200, eater wrote:
> > Socket path in 5.7 would result in libvirt//var/run/libvirt-sock instead of
> > /var/run/libvirt/libvirt-sock
> >
> > Signed-off-by: eater <=@eater.me>
>
> Thanks for
On 9/6/19 5:55 PM, Ján Tomko wrote:
[cc-ing Marc-André]
On Fri, Sep 06, 2019 at 04:25:19PM +0200, Michal Privoznik wrote:
The qemu side is not merged in yet, so there is a chance that the
interface will change. Don't detect the capability just yet then.
Signed-off-by: Michal Privoznik
---
On Fri, Sep 06, 2019 at 11:36:17PM +0200, eater wrote:
> Socket path in 5.7 would result in libvirt//var/run/libvirt-sock instead of
> /var/run/libvirt/libvirt-sock
>
> Signed-off-by: eater <=@eater.me>
Thanks for your contribution, but I'm afraid we can't accept patches with
psuedonyms / fake
Socket path in 5.7 would result in libvirt//var/run/libvirt-sock instead of
/var/run/libvirt/libvirt-sock
Signed-off-by: eater <=@eater.me>
---
src/remote/remote_daemon.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/src/remote/remote_daemon.c
On 9/9/19 10:52 AM, Ján Tomko wrote:
On Mon, Sep 09, 2019 at 09:49:38AM +0200, Michal Privoznik wrote:
An example how configure looks like with these changes applied:
https://travis-ci.org/zippy2/libvirt/jobs/582549099#L4247
Michal Prívozník (4):
configure: Prefer LIBVIRT_RESULT over
On Mon, Sep 09, 2019 at 09:49:38AM +0200, Michal Privoznik wrote:
An example how configure looks like with these changes applied:
https://travis-ci.org/zippy2/libvirt/jobs/582549099#L4247
Michal Prívozník (4):
configure: Prefer LIBVIRT_RESULT over AC_MSG_NOTICE
virt-result.m4: Align string
On Thu, Sep 05, 2019 at 01:34:37PM +0100, Daniel P. Berrangé wrote:
> We currently generate two completely separate API references for the
> libvirt public API. One at 'docs/html/' and one at 'docs/devhelp/'.
> Both are published on the website, but we only link to content in
> the 'docs/html/'
An example how configure looks like with these changes applied:
https://travis-ci.org/zippy2/libvirt/jobs/582549099#L4247
Michal Prívozník (4):
configure: Prefer LIBVIRT_RESULT over AC_MSG_NOTICE
virt-result.m4: Align string more generously
configure: Colorize output
virt-result.m4:
The LIBVIRT_RESULT function takes two or three arguments. The
first one is the name of the result (aka CHECK_NAME). It is
printed before the colon character. The rest of the arguments is
printed after the character. To produce colourized output a
couple of changes needs to be made.
Firstly, we
If we're running from a TTY we can put some colors around 'yes',
'no' and other messages.
Shamelessly copied from Ruby source code and modified a bit to
comply with syntax-check.
https://github.com/ruby/ruby/commit/e4879592873abd4cd8aeed56f4cbaa360a3d3736
Signed-off-by: Michal Privoznik
---
The times, when we had small CRTs are long gone. Now, in the era
of wide screens we can be more generous when it comes to aligning
the output of configure. The longest string before the colon is
'wireshark_dissector' which counts 19 characters. Therefore,
align the strings at 20.
At the same
One of the advantages is that LIBVIRT_RESULT aligns the resulting
message for us. The other is that in near future we will colour
some parts of the message and thus it helps if we get arguments
split in two.
Signed-off-by: Michal Privoznik
---
configure.ac | 8
67 matches
Mail list logo
