Le 30/11/2015 10:23, Michal Privoznik a écrit :
> On 27.11.2015 16:48, Remi Collet wrote:
>> Hi,
>>
>> arginfo are very important in PHP extensions, useful for users
>> (documentation), and various tools rely on them (through reflection).
>>
>> Attached patch add some.
>>
>> Tell me if ok, and I
On Thu, 2015-12-03 at 14:06 -0500, Laine Stump wrote:
> On 12/02/2015 12:39 PM, Andrea Bolognani wrote:
> > This is a straightforward wrapper around
> > virPCIDeviceAddressIOMMUGroupIterate() that will make some code less
> > awkward to write later on.
> > ---
> > src/libvirt_private.syms | 1 +
On Thu, Dec 03, 2015 at 11:13:06PM -0700, Jim Fehlig wrote:
> On 11/26/2015 09:59 AM, Ian Campbell wrote:
> > libxlConnectDomainXMLFromNative calls both xenParseXM and xenParseXL
> > with cfg->verInfo->xen_version_major, however AFAICT they both (either
> > inherently, or through there use of
Hi,
Thank you, it sounds good. Can you please let me know how Openstack
Ceilometer will access that wait counter (vcpu.NNN.wait)? To my
understanding Openstack uses libvirt python library to access domain data.
Regards,
- Avi Weit
"Daniel P. Berrange" wrote on
QEMU commit 75d373e disables svm by default since PC machine 2.2.
So with PC machine version > 2.1, When a user wants to expose the
svm flag to a guest, says:
..
..
The svm flag wont be passed to qemu command line, This patch fixes it.
Signed-off-by: Lin Ma
---
The initial version was released on Wednesday, not Thursday. I
would not care, but rpmlint is kind of sad:
libvirt-snmp.src: E: specfile-error warning: bogus date in %changelog: Thu Feb
2 2011 Michal Privoznik 0.0.1-1
13 packages and 0 specfiles checked; 1 errors, 0
On 12/04/2015 09:55 AM, Ján Tomko wrote:
> On Fri, Dec 04, 2015 at 09:44:20AM -0500, John Ferlan wrote:
>>
>>
>> On 12/04/2015 08:46 AM, Ján Tomko wrote:
>>> On Tue, Nov 24, 2015 at 03:57:14PM -0500, John Ferlan wrote:
Similar to the openflags VIR_STORAGE_VOL_OPEN_NOERROR processing, if
On Fri, 2015-12-04 at 16:31 +0100, Michal Privoznik wrote:
> ACK, although I'd rather wait until after release. If you (or somebody
> else) feels otherwise, do push it any time you want.
No need to rush, I'll push after release.
Thanks for the review!
--
Andrea Bolognani
Software Engineer -
The virt driver, dom name and uuid associated with a log
file are important pieces of metadata to keep around for
sake of future enhancements to virtlogd. Currently we
discard them after opening the log file, but we should
preserve them, even across restarts.
Signed-off-by: Daniel P. Berrange
W dniu 2015-11-20 o 09:13, Peter Krempa pisze:
On Thu, Nov 19, 2015 at 21:13:56 +0100, Piotr Rybicki wrote:
W dniu 2015-11-19 o 17:31, Michal Privoznik pisze:
...
==2650== 7,692,288 bytes in 2 blocks are still reachable in loss record 1,444
of 1,452
==2650==at 0x4C2BFC8: calloc
On Fri, Dec 04, 2015 at 09:44:20AM -0500, John Ferlan wrote:
>
>
> On 12/04/2015 08:46 AM, Ján Tomko wrote:
> > On Tue, Nov 24, 2015 at 03:57:14PM -0500, John Ferlan wrote:
> >> Similar to the openflags VIR_STORAGE_VOL_OPEN_NOERROR processing, if some
> >> read processing operation fails, check
On 04.12.2015 15:34, Andrea Bolognani wrote:
> When mocking filesystem access in the test suite, we have always
> assumed it to be some subdirectory of /sys depending on the needs
> of the specific test case.
>
> This limits our flexibility, and will be a problem once we need
> to start mocking
On 12/04/2015 07:30 AM, Michal Privoznik wrote:
So yet again one of integer arguments that we use as a boolean.
Since the argument count of the function is unbearably long
enough, lets turn those booleans into flags.
Signed-off-by: Michal Privoznik
---
On Fri, Dec 04, 2015 at 15:18:16 +0100, Michal Privoznik wrote:
> On 01.12.2015 15:11, Peter Krempa wrote:
> > Peter Krempa (2):
> > schema: Allow > UINT_MAX KiB of memory for NUMA nodes
> > qemu: domain: Prevent overflows in memory alignment code
> >
> > docs/schemas/domaincommon.rng
The virtlogd RPC messages all have a flags parameter. For
sake of future error reporting we should be verifying
these are all 0 for now.
Signed-off-by: Daniel P. Berrange
---
src/logging/log_daemon_dispatch.c | 5 -
src/logging/log_handler.c | 11 ++-
Having looked at the code again there is a wire-ABI incompatible
change I think we should make to virtlogd to make it more
flexible. Since we've not released 1.3.0 yet, this is the only
chance we'll have to change the wire protocol. There are also
a couple of fixes in this series we should have
The current virtlogd RPC protocol provides the ability to
handle log files associated with QEMU stdout/err. The log
protocol messages take the virt driver, domain name and
use that to form a log file path. This is quite restrictive
as it prevents us re-using the same RPC protocol messages
for
The qemuDomainLogContextNew method leaks the "logfile" path
on the non-virtlogd code path.
Signed-off-by: Daniel P. Berrange
---
src/qemu/qemu_domain.c | 6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/src/qemu/qemu_domain.c b/src/qemu/qemu_domain.c
On 12/04/2015 07:30 AM, Michal Privoznik wrote:
There are few outdated things. Firstly, we don't need to undergo
the torture of fopen, fscanf and fclose when we have nice wrapper
over that: virFileReadAll. Secondly, we can use dynamically
allocated buffer for the interface index.
Nothing
On 01.12.2015 15:11, Peter Krempa wrote:
> Peter Krempa (2):
> schema: Allow > UINT_MAX KiB of memory for NUMA nodes
> qemu: domain: Prevent overflows in memory alignment code
>
> docs/schemas/domaincommon.rng | 4 +--
> src/qemu/qemu_domain.c
So, for the multiqueue on macvtaps we are going to need to open
the device multiple times. Currently, this is not supported.
Rework the function, so that upper layers can be reworked too.
Signed-off-by: Michal Privoznik
---
src/util/virnetdevmacvlan.c | 58
On Fri, 2015-12-04 at 10:01 +, Daniel P. Berrange wrote:
> On Thu, Dec 03, 2015 at 11:13:06PM -0700, Jim Fehlig wrote:
> > On 11/26/2015 09:59 AM, Ian Campbell wrote:
> > > libxlConnectDomainXMLFromNative calls both xenParseXM and xenParseXL
> > > with cfg->verInfo->xen_version_major, however
The old name is no longer accurate, since now we're using its value as
the root of the fake filesystem.
No functional changes.
---
tests/vircgroupmock.c | 10 +-
tests/vircgrouptest.c | 16
tests/virhostdevtest.c | 16
tests/virpcimock.c | 6
https://bugzilla.redhat.com/show_bug.cgi?id=1240439
Ta-da! Now that we know how to open a macvtap device multiple
times, we can finally enable the multiqueue feature. Everything
else is already prepared (e.g. command line generation) from the
previous iteration where the feature was implemented
So, for the multiqueue on macvtaps we are going to need to open
the device multiple times. Currently, this is not supported.
Rework the function, so that upper layers can be reworked too.
Signed-off-by: Michal Privoznik
---
src/lxc/lxc_process.c | 1 +
In the non-systemd case, without socket activation, we need to proper
ordering.
---
daemon/libvirtd.init.in | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/daemon/libvirtd.init.in b/daemon/libvirtd.init.in
index ed42195..22006c4 100644
--- a/daemon/libvirtd.init.in
+++
On Fri, Dec 04, 2015 at 02:55:05PM +0100, Guido Günther wrote:
> In the non-systemd case, without socket activation, we need to proper
> ordering.
> ---
> daemon/libvirtd.init.in | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/daemon/libvirtd.init.in
On Fri, Dec 04, 2015 at 02:19:33PM +, Ian Campbell wrote:
> On Fri, 2015-12-04 at 10:01 +, Daniel P. Berrange wrote:
> > On Thu, Dec 03, 2015 at 11:13:06PM -0700, Jim Fehlig wrote:
> > > On 11/26/2015 09:59 AM, Ian Campbell wrote:
> > > > libxlConnectDomainXMLFromNative calls both
We might need to mock files living outside SYSFS_PREFIX later on,
so it's better to treat the temporary directory we are passed via
the environment as the root of the fake filesystem and create
SYSFS_PREFIX inside it.
The environment variable name will be changed to reflect the new use
we're
We might need to mock files living outside PCI_SYSFS_PREFIX later on,
so it's better to treat the temporary directory we are passed via
the environment as the root of the fake filesystem and create
PCI_SYSFS_PREFIX inside it.
The environment variable name will be changed to reflect the new use
When mocking filesystem access in the test suite, we have always
assumed it to be some subdirectory of /sys depending on the needs
of the specific test case.
This limits our flexibility, and will be a problem once we need
to start mocking eg. /dev as well, or simply two different parts
of the
Instead of fakesysfsdir, which is very generic, use fakesysfspcidir and
fakesysfscgroupdir. This makes it explicit what part of the fake sysfs
filesystem they're referring to, and also leaves open the possibility of
handling files in two unrelated parts of the fake sysfs filesystem.
No functional
This updates the test program to make it consistent with recent changes
to the mock libraries, and also opens up the possibility of mocking more
than just /sys in the future.
---
tests/scsihosttest.c | 17 -
1 file changed, 12 insertions(+), 5 deletions(-)
diff --git
The test program is not preloading any of the mock libraries that read
that environment variable, so setting it is pointless.
---
tests/scsihosttest.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/tests/scsihosttest.c b/tests/scsihosttest.c
index eecb1c3..eb5f522 100644
---
Some patches are easier to review when showed with --ignore-space-change.
Michal Privoznik (6):
virNetDevMacVLanCreateWithVPortProfile: Turn vnet_hdr into flag
virNetDevMacVLanTapOpen: Slightly rework
virNetDevMacVLanTapOpen: Rework to support multiple FDs
virNetDevMacVLanTapSetup: Rework
There are few outdated things. Firstly, we don't need to undergo
the torture of fopen, fscanf and fclose when we have nice wrapper
over that: virFileReadAll. Secondly, we can use dynamically
allocated buffer for the interface index.
Signed-off-by: Michal Privoznik
---
So yet again one of integer arguments that we use as a boolean.
Since the argument count of the function is unbearably long
enough, lets turn those booleans into flags.
Signed-off-by: Michal Privoznik
---
src/lxc/lxc_process.c | 2 +-
src/qemu/qemu_command.c | 5
So, for the multiqueue on macvtaps we are going to need to open
the device multiple times. Currently, this is not supported.
Rework the function, so that upper layers can be reworked too.
Signed-off-by: Michal Privoznik
---
src/util/virnetdevmacvlan.c | 66
On Fri, Dec 04, 2015 at 02:12:42PM +0200, Avi Weit wrote:
> Hi,
>
> Thank you, it sounds good. Can you please let me know how Openstack
> Ceilometer will access that wait counter (vcpu.NNN.wait)? To my
> understanding Openstack uses libvirt python library to access domain data.
The python API
On Tue, Nov 24, 2015 at 03:57:11PM -0500, John Ferlan wrote:
> v1 took a simplistic approach:
>
> http://www.redhat.com/archives/libvir-list/2015-October/msg00919.html
>
> but not specific enough according to the review. So adding some extra
> complexity of checking for specific errors and flag
On Tue, Nov 24, 2015 at 03:57:14PM -0500, John Ferlan wrote:
> Similar to the openflags VIR_STORAGE_VOL_OPEN_NOERROR processing, if some
> read processing operation fails, check the readflags for the corresponding
> error flag being set. If so, rather then causing an error - use VIR_WARN
> to flag
On 12/04/2015 08:46 AM, Ján Tomko wrote:
> On Tue, Nov 24, 2015 at 03:57:14PM -0500, John Ferlan wrote:
>> Similar to the openflags VIR_STORAGE_VOL_OPEN_NOERROR processing, if some
>> read processing operation fails, check the readflags for the corresponding
>> error flag being set. If so,
init_env() will return right away if fakesysfsdir is already
initialized, so this check is redundant.
---
tests/virpcimock.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/tests/virpcimock.c b/tests/virpcimock.c
index 0b49290..6d00a4f 100644
--- a/tests/virpcimock.c
+++
On 12/04/2015 07:30 AM, Michal Privoznik wrote:
So, for the multiqueue on macvtaps we are going to need to open
s/So, for the multiqueue/For multiqueue support/
the device multiple times. Currently, this is not supported.
Rework the function, so that upper layers can be reworked too.
On the very first log message we send to any output, we include
the libvirt version number and package string. In some bug reports
we have been given libvirtd.log files that came from a different
host than the corresponding /var/log/libvirt/qemu log files. So
extend the initial log message to
On 12/04/2015 07:31 AM, Michal Privoznik wrote:
So, for the multiqueue on macvtaps we are going to need to open
the device multiple times. Currently, this is not supported.
Rework the function, so that upper layers can be reworked too.
Signed-off-by: Michal Privoznik
---
The log file descriptor associated with the virRotatingFile
struct should be marked close-on-exec, as even when virtlogd
re-exec's itself it expect to open the log file fresh. It
does not need to preserve the logfile handles, only the network
client FDs.
Signed-off-by: Daniel P. Berrange
On 12/04/2015 07:31 AM, Michal Privoznik wrote:
So, for the multiqueue on macvtaps we are going to need to open
the device multiple times. Currently, this is not supported.
Rework the function, so that upper layers can be reworked too.
Either merge 3 & 4 together, or change the log message
Often when debugging bug reports one is given a copy of the file
from /var/log/libvirt/qemu/$NAME.log along with other supporting
files. In a number of cases I've been given sets of files which
were from different machines. Including the hostname in the QEMU
log file will help identify when the
Commit d2e5538b1 changes virDomainDef to create ifnames
that are autogenerated by libxl, and also clearing them up
on domain cleanup. One place that was missing was also
on migration, when domain xml is sent to dst libvirtd and
would contain ifnames from the source libvirtd. This
would lead to
On 12/04/2015 12:26 PM, Daniel P. Berrange wrote:
Often when debugging bug reports one is given a copy of the file
from /var/log/libvirt/qemu/$NAME.log along with other supporting
files. In a number of cases I've been given sets of files which
were from different machines. Including the hostname
On 12/04/2015 12:24 PM, Daniel P. Berrange wrote:
The log file descriptor associated with the virRotatingFile
struct should be marked close-on-exec, as even when virtlogd
re-exec's itself it expect to open the log file fresh. It
does not need to preserve the logfile handles, only the network
On Fri, Dec 04, 2015 at 02:00:13PM +, Daniel P. Berrange wrote:
> On Fri, Dec 04, 2015 at 02:55:05PM +0100, Guido Günther wrote:
> > In the non-systemd case, without socket activation, we need to proper
> > ordering.
> > ---
> > daemon/libvirtd.init.in | 2 +-
> > 1 file changed, 1
On 12/03/2015 12:18 PM, Joao Martins wrote:
>
> On 12/02/2015 11:27 PM, Jim Fehlig wrote:
>> On 11/10/2015 08:32 AM, Joao Martins wrote:
>>> Introduce support for VIR_MIGRATE_PEER2PEER in libxl driver
>>> for supporting migration in Openstack. Most of the changes
>>> occur at the source and no
Commit d2e5538b1 changes virDomainDef to include ifnames
that autogenerated by libxl, and that are also cleared
on domain cleanup. One place that's missing is on
migration, when domain xml is sent to dst libvirtd and
would contain ifnames from the source libvirtd. This
would lead to erronous
On 12/04/2015 10:53 AM, Daniel P. Berrange wrote:
> Having looked at the code again there is a wire-ABI incompatible
> change I think we should make to virtlogd to make it more
> flexible. Since we've not released 1.3.0 yet, this is the only
> chance we'll have to change the wire protocol. There
https://bugzilla.redhat.com/show_bug.cgi?id=1276198
Prior to commit id '98322052' failure to saferead the block device would
cause an error to be logged and the device to be skipped while attempting
to discover/create a stable target path for a new LUN (NPIV).
This was because
I found this useful while processing a volume that wouldn't end up
showing up in the resulting list of block volumes. In this case, the
partition type wasn't found in the disk_types table.
Signed-off-by: John Ferlan
---
src/storage/storage_backend.c | 4
1 file changed,
https://bugzilla.redhat.com/show_bug.cgi?id=1265694
Add the capability to recognize and parse named/documented flags for
the storage pool. Chose to make rng flags just text rather than having
flags be some list of "expected" values (such as in nwfilter.rng) mostly
to make future expansion easier.
Add a flags argument to the storage pool source to handle allowing some
feature to be passed to the lower/backend for the purpose of executing
specific code paths or making decisions on algorithms.
Signed-off-by: John Ferlan
---
src/conf/storage_conf.h | 11 +++
1
In order to be able to process disk storage pool's using a multipath
device to handle the partitions, libvirt_parthelper will need to know
about a feature available from multipath known as 'user_friendly_names'.
This feature causes multipath partitions to not include the "p"
partition separator
https://bugzilla.redhat.com/show_bug.cgi?id=1265694
Posting as an RFC mainly to get buyin/feedback over the concept of using
a field in the as a way to allow or direct the
pool/storage backend code to use specific algorithms.
The issue is described in the bz, but the shorter version is the
Similar to the openflags VIR_STORAGE_VOL_OPEN_NOERROR processing, if some
read processing operation fails, check the readflags for the corresponding
error flag being set. If so, rather then causing an error - use VIR_WARN
to flag the error, but return -2 which some callers can use to perform
Similar to the openflags which allow VIR_STORAGE_VOL_OPEN_NOERROR to be
passed to avoid open errors, add a 'readflags' variable so that in the
future read failures could also be ignored.
Signed-off-by: John Ferlan
---
src/storage/storage_backend.c | 17
Just so it's clearer what to expect upon input and what types of return
values could be generated. These were loosely copied from existing
virStorageBackendUpdateVolTargetInfoFD.
Signed-off-by: John Ferlan
---
src/storage/storage_backend.c | 30
While processing the volume for lseek, virFileReadHeaderFD, and
virStorageFileGetMetadataFromBuf - failure would cause an error,
but ret would not be set. That would result in an error message being
sent, but successful status being returned.
Signed-off-by: John Ferlan
---
v2 here:
http://www.redhat.com/archives/libvir-list/2015-November/thread.html
Based on review and well just the amount of change I figured I'd
generate a v3. The patches were mostly ACK'd, but patch 3 required
removal of some flags which resulted in patch 1 needing updates...
Then I found that
On 11/12/2015 02:23 AM, Nikolay Shirokovskiy wrote:
> virDomainSetMemory is documented to change only runtime configuration of
> running domain. This can't be implemented for vz driver as it always change
> persistent state too. Let's extend semantics of this function for changes to
> be
>
On 12/04/2015 12:38 PM, Daniel P. Berrange wrote:
On the very first log message we send to any output, we include
the libvirt version number and package string. In some bug reports
we have been given libvirtd.log files that came from a different
host than the corresponding /var/log/libvirt/qemu
On 12/04/2015 07:31 AM, Michal Privoznik wrote:
https://bugzilla.redhat.com/show_bug.cgi?id=1240439
Ta-da! Now that we know how to open a macvtap device multiple
times, we can finally enable the multiqueue feature. Everything
else is already prepared (e.g. command line generation) from the
On 12/04/2015 05:28 PM, Joao Martins wrote:
> Commit d2e5538b1 changes virDomainDef to create ifnames
> that are autogenerated by libxl, and also clearing them up
> on domain cleanup. One place that was missing was also
> on migration, when domain xml is sent to dst libvirtd and
> would contain
Signed-off-by: Pavel Hrdina
---
src/esx/esx_driver.c | 4 ++--
src/vmware/vmware_conf.c | 3 ++-
src/vmware/vmware_driver.c | 2 +-
src/vmx/vmx.c | 5 +
src/vmx/vmx.h | 1 +
tests/vmx2xmltest.c| 2 +-
6 files changed, 12
This patch series is an attempt to make the code for input devices cleaner and
readable and also follow some good practices whit rule that one function should
do only one thing. We are really bad at this and our code is sometimes really
hard to follow.
Pavel Hrdina (7):
vmware/vmx: use
On 12/04/2015 10:53 AM, Daniel P. Berrange wrote:
> The virt driver, dom name and uuid associated with a log
> file are important pieces of metadata to keep around for
> sake of future enhancements to virtlogd. Currently we
> discard them after opening the log file, but we should
> preserve
On 12/04/2015 12:07 PM, Boris Fiuczynski wrote:
> On 12/03/2015 05:56 PM, John Ferlan wrote:
>>
>>
>> On 11/30/2015 06:06 AM, Boris Fiuczynski wrote:
>>> When a SCSI disk is hotplugged to a domain that does not have the
>>> required
>>> scsi controller already defined the following internal
Signed-off-by: Pavel Hrdina
---
src/qemu/qemu_domain.c | 28 ++
.../qemuxml2argvdata/qemuxml2argv-pseries-disk.xml | 2 +-
2 files changed, 13 insertions(+), 17 deletions(-)
diff --git a/src/qemu/qemu_domain.c
Signed-off-by: Pavel Hrdina
---
src/lxc/lxc_driver.c | 5 -
src/lxc/lxc_native.c | 8 +++-
src/lxc/lxc_native.h | 4 +++-
tests/Makefile.am
The current code was a little bit odd. At first we've removed all
possible implicit input devices from domain definition to add them later
back if there was any graphics device defined while parsing XML
description. That's not all, while formating domain definition to XML
description we at first
Signed-off-by: Pavel Hrdina
---
src/libxl/libxl_driver.c | 8 ++--
src/xen/xen_driver.c | 5 +++--
src/xen/xm_internal.c| 2 +-
src/xenconfig/xen_sxpr.c
Only 'tsc' timer allows set mode and track is valid only for 'rtc' and
'platform' timers.
Signed-off-by: Pavel Hrdina
---
src/xenconfig/xen_common.c | 2 ++
src/xenconfig/xen_sxpr.c | 2 ++
2 files changed, 4 insertions(+)
diff --git a/src/xenconfig/xen_common.c
Signed-off-by: Pavel Hrdina
---
src/libxl/libxl_domain.c | 4
1 file changed, 4 insertions(+)
diff --git a/src/libxl/libxl_domain.c b/src/libxl/libxl_domain.c
index ef92974..9551f2e 100644
--- a/src/libxl/libxl_domain.c
+++ b/src/libxl/libxl_domain.c
@@ -995,6 +995,10
81 matches
Mail list logo