2017-04-06 8:35 GMT+03:00 Michal Privoznik :
> Still, you will lack support for some drivers (e.g. ESX, Hyperv, XEN, local
> test driver). Why do you want to hook on the RPC anyway?
My use case only qemu, so all the simple. Also i think that
integration tests covers changes from libvirt side and
On Wed, Apr 05, 2017 at 12:40:09 -0400, John Ferlan wrote:
>
>
> On 04/05/2017 12:28 PM, Andrea Bolognani wrote:
> > The contents of volname would be leaked if the function were
> > to be passed an invalid pooltype by the caller.
> >
> > Make sure the memory is released instead.
> > ---
> > Init
> +
> +VIR_ENUM_IMPL(virCache, VIR_CACHE_TYPE_LAST,
> + "unified",
> + "instruction",
> + "data")
> +
> };
>
> +typedef enum {
> + VIR_CACHE_TYPE_DATA,
> + VIR_CACHE_TYPE_INSTRUCTION,
> + VIR_CACHE_TYPE_UNIFIED,
> +
> + VIR_CACHE_TYPE_LAST
> +} virCacheType;
> +
The sequence is wrong, it shoul
On 04/05/2017 09:11 PM, Vasiliy Tolstov wrote:
2017-04-05 17:53 GMT+03:00 Andrea Bolognani :
That is true as long as you stick with the public API.
If you start messing with the internal RPC and a change
in libvirt breaks your code, you get to keep both pieces.
I think in case of internal rp
On 04/05/2017 10:34 PM, John Ferlan wrote:
>
>
> On 04/05/2017 04:16 PM, Martin Kletzander wrote:
>> On Wed, Apr 05, 2017 at 03:11:52PM -0400, John Ferlan wrote:
>>>
>>>
>>> On 04/05/2017 04:50 AM, Michal Privoznik wrote:
This function runs an iscsi command and parses its output.
Howeve
On Wednesday, 5 April 2017 at 10:36 PM, Martin Kletzander wrote:
> Add info from yet another machine, this time with resctrl data so that
> we can extend tests easily in a test-driven way.
>
> Signed-off-by: Martin Kletzander (mailto:mklet...@redhat.com)>
> ---
> tests/vircaps2xmldata/linux-re
On 04/05/2017 04:16 PM, Martin Kletzander wrote:
> On Wed, Apr 05, 2017 at 03:11:52PM -0400, John Ferlan wrote:
>>
>>
>> On 04/05/2017 04:50 AM, Michal Privoznik wrote:
>>> This function runs an iscsi command and parses its output.
>>> However, due to the nature of things, virISCSIExtractSession(
On Wed, Apr 05, 2017 at 03:11:52PM -0400, John Ferlan wrote:
On 04/05/2017 04:50 AM, Michal Privoznik wrote:
This function runs an iscsi command and parses its output.
However, due to the nature of things, virISCSIExtractSession()
callback can be called multiple times. In each run it would
all
2017-04-05 17:53 GMT+03:00 Andrea Bolognani :
> That is true as long as you stick with the public API.
>
> If you start messing with the internal RPC and a change
> in libvirt breaks your code, you get to keep both pieces.
I think in case of internal rpc i need to monitor each libvirt release
and
On 04/05/2017 04:50 AM, Michal Privoznik wrote:
> This function runs an iscsi command and parses its output.
> However, due to the nature of things, virISCSIExtractSession()
> callback can be called multiple times. In each run it would
> allocate new memory and overwrite the variable where we kee
On 04/05/2017 12:28 PM, Andrea Bolognani wrote:
> The contents of volname would be leaked if the function were
> to be passed an invalid pooltype by the caller.
>
> Make sure the memory is released instead.
> ---
> Initializing volname to NULL is strictly speaking not
> necessary, but I like it
On 04/04/2017 05:25 PM, Boris Fiuczynski wrote:
> Fixing make syntax-check broken by commit 4da534c0b9.
>
> Signed-off-by: Boris Fiuczynski
> Reviewed-by: Marc Hartmayer
> Reviewed-by: Bjoern Walk
> ---
> cfg.mk | 4 ++--
> 1 file changed, 2 insertions(+), 2 deletions(-)
>
ACKed and pushed.
The contents of volname would be leaked if the function were
to be passed an invalid pooltype by the caller.
Make sure the memory is released instead.
---
Initializing volname to NULL is strictly speaking not
necessary, but I like it better that way :)
src/storage/storage_util.c | 12 +--
On Wed, 2017-04-05 at 16:17 +0300, Vasiliy Tolstov wrote:
> I know that, but also i know that libvirt does not break compatibility
> with older clients =).
That is true as long as you stick with the public API.
If you start messing with the internal RPC and a change
in libvirt breaks your code, y
https://bugzilla.redhat.com/show_bug.cgi?id=1233129
The virStoragePoolObjSourceFindDuplicate logic used by PoolCreateXML
and PoolDefineXML avoids comparing the new definition against "other"
pool types. This can cause unexpected corruption if two different pool
source types used the same source de
Refactor virStoragePoolObjSourceFindDuplicate into smaller units
separated by the "supported" pool source type. The DIR, GLUSTER,
and NETFS pools all can use "... ... ".
Alter the logic slightly to return the matching pool or NULL rather
than setting matchpool = pool and break. Easier to read tha
In the effort to reduce the virStoragePoolObjSourceFindDuplicate logic,
create a new helper which will handle all the ISCSI type differences.
Alter things just a little bit to return NULL or pool rather than
using breaks and matchpool = pool, then break. Also rather than creating
variables withing
Refactor virStoragePoolObjSourceFindDuplicate into smaller units
separated by the "supported" pool source type. The ISCSI, FS,
LOGICAL, DISK, and ZFS pools can use "... ...
".
Alter the logic slightly to return the matching pool or NULL rather
than setting matchpool = pool and break. Easier to re
https://bugzilla.redhat.com/show_bug.cgi?id=1233129
Refactor the virStoragePoolSourceFindDuplicateDevices code a bit to
make it a bit easier to read, then allow pool def type's that use a
pool source device to define the source device to perform cross pool
source device conflicts.
John Ferlan (4)
It is no longer needed thanks to the great virfilemock.c. And this
way we don't have to add a new set of functions for each prefixed
path.
While on that, add two functions that weren't there before, string and
scaled integer reading ones. Also increase the length of the string
being read by one
More directories will need to be mocked, so let's prepare for that.
Signed-off-by: Martin Kletzander
---
tests/vircaps2xmltest.c | 20 +---
1 file changed, 13 insertions(+), 7 deletions(-)
diff --git a/tests/vircaps2xmltest.c b/tests/vircaps2xmltest.c
index 8f02314afe8e..31d0279
This mock (which is actually not mock at all, see later) can redirect
all accesses to a path into another path. There is no need to
create mocks for particular directories, you just create a directory
with all the data a redirect the test there.
In the future, this should also be able to register
Commit a0fdd2f6f9a0cc77ae285c289e2c16d314b6a907 added some data from
the system but forgot 3 files for each cache.
Signed-off-by: Martin Kletzander
---
tests/vircaps2xmldata/linux-caches/cpu/cpu0/cache/index0/id | 1 +
.../vircaps2xmldata/linux-caches/cpu/cpu0/cache/index0/shared_cpu
And use it in virFileRead*
Signed-off-by: Martin Kletzander
---
src/util/virfile.c| 18 +++---
src/util/virhostcpu.c | 4 ++--
src/util/virstring.h | 8
src/util/virsysfs.c | 2 ++
4 files changed, 19 insertions(+), 13 deletions(-)
diff --git a/src/util/virfile.c
This header file has been created so that we can expose
internal functions to the test suite without making them
public: those in qemu_capabilities.h bearing the comment
/* Only for use by test suite */
are obvious candidates for being moved over.
---
src/qemu/qemu_capabilities.h | 19
If formatting NUMA topology fails, the function returns immediatelly,
but the buffer structure allocated on the stack references lot of
heap-allocated memory and that would get lost in such case.
Signed-off-by: Martin Kletzander
---
src/conf/capabilities.c | 6 +-
1 file changed, 5 insertion
Signed-off-by: Martin Kletzander
---
tests/virschematest.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/tests/virschematest.c b/tests/virschematest.c
index 119d4ba395e2..ffed2175fd37 100644
--- a/tests/virschematest.c
+++ b/tests/virschematest.c
@@ -215,7 +215,8 @@ mymain
Signed-off-by: Martin Kletzander
---
docs/schemas/capability.rng | 34 ++
1 file changed, 34 insertions(+)
diff --git a/docs/schemas/capability.rng b/docs/schemas/capability.rng
index 88e08d299ac9..208095302a63 100644
--- a/docs/schemas/capability.rng
+++ b/docs/s
Signed-off-by: Martin Kletzander
---
I'm adding only info about L3 caches, we can add more later, but for
now that's more than enough.
src/conf/capabilities.c | 200
src/conf/capabilities.h | 29
src/libvirt_private.s
Added only in drivers that were already calling
virCapabilitiesInitNUMA(). Instead of refactoring all the callers to
behave the same way in case of error, just follow what the callers are
doing for all the functions.
Signed-off-by: Martin Kletzander
---
src/lxc/lxc_conf.c | 3 +++
src
Signed-off-by: Martin Kletzander
---
src/util/virfile.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/util/virfile.c b/src/util/virfile.c
index a91c2c349501..c0f448d3437d 100644
--- a/src/util/virfile.c
+++ b/src/util/virfile.c
@@ -2861,11 +2861,11 @@ virDirOpenQuiet
So during a discussion [1] with Erik, I got an idea. And this is
basically it. Reading and using kernel-interface-like files in an easy
fashion and being able to easily mock them were two goals that, in my
not so humble opinion, I reached.
The previous version [2] is actually nor related to this
Ján Tomko wrote:
> Introduce STRICT_FRAME_LIMIT_CFLAGS that will be used for
> the library code and RELAXED_FRAME_LIMIT_CFLAGS for daemon code
> and the test code.
>
> Raising the limit for tests allows building them with clang
> with optimizations disabled.
> ---
> v2:
> * use the strict limit
2017-04-05 16:09 GMT+03:00 Michal Privoznik :
> On 04/05/2017 12:35 AM, Vasiliy Tolstov wrote:
>> Hi. I know that libvirt devs does not like devs that used internal rpc
>> of libvirt...
>
> Not entirely true. We like all devs, but we just discourage people to
> use our RPC as it's internal. It does
We need to make sure to reset all migration parameters to prevent future
operations calling the "migrate" QMP command from accidentally using
stale parameters.
This series resets only TLS parameters and all capabilities. A followup
series will reset all other parameters (such as compression) too.
Let's use a bool variable to create a single shared path returning 0.
Signed-off-by: Jiri Denemark
---
src/qemu/qemu_process.c | 31 ---
1 file changed, 16 insertions(+), 15 deletions(-)
diff --git a/src/qemu/qemu_process.c b/src/qemu/qemu_process.c
index a37496107..
qemuProcessRecoverMigrationOut doesn't explicitly call
qemuMigrationResetTLS relying on two things:
- qemuMigrationCancel resets TLS parameters
- our migration code resets TLS before entering
QEMU_MIGRATION_PHASE_PERFORM3_DONE phase
But this is not obvious and the assumptions will b
On 04/05/2017 12:35 AM, Vasiliy Tolstov wrote:
> Hi. I know that libvirt devs does not like devs that used internal rpc
> of libvirt...
Not entirely true. We like all devs, but we just discourage people to
use our RPC as it's internal. It doesn't offer the full set of features
provided by libvirt
This new API is supposed to reset all migration parameters to make sure
future migrations won't accidentally use them. This patch makes the
first step and moves qemuMigrationResetTLS call inside
qemuMigrationReset.
Signed-off-by: Jiri Denemark
---
src/qemu/qemu_migration.c | 31 +
It's only called from qemuMigrationReset now so it doesn't need to be
exported and {tls,sec}Alias are always NULL.
Signed-off-by: Jiri Denemark
---
src/qemu/qemu_migration.c | 37 +
src/qemu/qemu_migration.h | 7 ---
2 files changed, 13 insertions(+), 31
So far only QEMU_MONITOR_MIGRATION_CAPS_POSTCOPY was reset, but only in
a single code path leaving post-copy enabled in quite a few cases.
https://bugzilla.redhat.com/show_bug.cgi?id=1425003
Signed-off-by: Jiri Denemark
---
src/qemu/qemu_migration.c | 10 +++---
1 file changed, 7 insertions
Migration parameters are either reset by the main migration code path or
from qemuProcessRecoverMigration* in case libvirtd is restarted during
migration.
Signed-off-by: Jiri Denemark
---
src/qemu/qemu_migration.c | 3 ---
1 file changed, 3 deletions(-)
diff --git a/src/qemu/qemu_migration.c b/
There is no async job running when a freshly started libvirtd is trying
to recover from an interrupted incoming migration. While at it, let's
call qemuMigrationResetTLS every time we don't kill the domain. This is
not strictly necessary since TLS is not supported when v2 migration
protocol is used,
Finished qemuMigrationRun does not mean the migration itself finished
(it might have just switched to post-copy mode). While resetting TLS
parameters is probably OK at this point even if migration is still
running, we want to consolidate the code which resets various migration
parameters. Thus qemu
Introduce STRICT_FRAME_LIMIT_CFLAGS that will be used for
the library code and RELAXED_FRAME_LIMIT_CFLAGS for daemon code
and the test code.
Raising the limit for tests allows building them with clang
with optimizations disabled.
---
v2:
* use the strict limit for tools and daemon too
* append the
On Tue, Apr 04, 2017 at 19:51:26 +0200, Jiri Denemark wrote:
> There is no async job running when a freshly started libvirtd is trying
> to recover from an interrupted incoming migration. While at it, let's
> call qemuMigrationResetTLS every time we don't kill the domain. This is
> not strictly nec
On Wed, 2017-04-05 at 12:00 +0200, Ján Tomko wrote:
[...]
> > @@ -4096,10 +4100,15 @@ virQEMUCapsInitHelp(virQEMUCapsPtr qemuCaps, uid_t
> > runUid, gid_t runGid, const c
> > qmperr) < 0)
> > goto cleanup;
> >
> > -/* -no-acpi is not supported on non-x8
On Wed, Apr 05, 2017 at 02:21:48PM +0200, Peter Krempa wrote:
QEMU_CHECK_MONITOR_* already logs the object and vm name
---
src/qemu/qemu_monitor.c | 10 +-
1 file changed, 1 insertion(+), 9 deletions(-)
ACK
Jan
signature.asc
Description: Digital signature
--
libvir-list mailing list
QEMU_CHECK_MONITOR_* already logs the object and vm name
---
src/qemu/qemu_monitor.c | 10 +-
1 file changed, 1 insertion(+), 9 deletions(-)
diff --git a/src/qemu/qemu_monitor.c b/src/qemu/qemu_monitor.c
index a4fa6eca6..2fc6192dc 100644
--- a/src/qemu/qemu_monitor.c
+++ b/src/qemu/qemu_m
On Wed, Apr 05, 2017 at 07:37:34 -0400, John Ferlan wrote:
>
>
> On 03/17/2017 12:36 PM, Jiri Denemark wrote:
> > The test takes
> >
> > x86-cpuid-Something-guest.xml CPU (the CPU libvirt would use for
> > host-model on a CPU described by x86_64-cpuid-Something.xml without
> > talking
On Wed, 2017-04-05 at 12:04 +0200, Ján Tomko wrote:
[...]
> > @@ -510,6 +510,8 @@ int virQEMUCapsParseHelpStr(const char *qemu,
> > const char *qmperr);
> > /* Only for use by test suite */
> > int virQEMUCapsParseDeviceStr(virQEMUCapsPtr qemuCaps, const char *str);
> >
2017-04-05 1:35 GMT+03:00 Vasiliy Tolstov :
> Hi. I know that libvirt devs does not like devs that used internal rpc
> of libvirt... I need to create stream to use with upload/download and
> as i see it does not created via rpc call, can you suggest me what
> files i need to check to understand how
On Wed, Apr 05, 2017 at 12:29:03PM +0100, Daniel P. Berrange wrote:
On Wed, Apr 05, 2017 at 12:38:23PM +0200, Martin Kletzander wrote:
[...]
How about we enable it by default for build with new enough GCC (so that
it reports these errors without silently dropping the non-NULL checks)?
Yes,
On 03/17/2017 12:36 PM, Jiri Denemark wrote:
> The test takes
>
> x86-cpuid-Something-guest.xml CPU (the CPU libvirt would use for
> host-model on a CPU described by x86_64-cpuid-Something.xml without
> talking to QEMU about what it supports on the host)
>
> and updates it according t
---
Pushed under the build-breaker rule.
docs/Makefile.am | 1 +
1 file changed, 1 insertion(+)
diff --git a/docs/Makefile.am b/docs/Makefile.am
index 62e069e15..105fe6894 100644
--- a/docs/Makefile.am
+++ b/docs/Makefile.am
@@ -220,6 +220,7 @@ news.html.in: \
fi
EXTRA_DIST += \
On Wed, Apr 05, 2017 at 12:38:23PM +0200, Martin Kletzander wrote:
> On Wed, Apr 05, 2017 at 10:06:22AM +0100, Daniel P. Berrange wrote:
> > On Tue, Apr 04, 2017 at 09:51:47PM +0200, Martin Kletzander wrote:
> > > On Tue, Apr 04, 2017 at 04:10:59PM +0100, Daniel P. Berrange wrote:
> > > > On Tue, M
On Wed, Apr 05, 2017 at 06:53:10AM -0400, John Ferlan wrote:
On 04/05/2017 05:06 AM, Daniel P. Berrange wrote:
On Tue, Apr 04, 2017 at 09:51:47PM +0200, Martin Kletzander wrote:
On Tue, Apr 04, 2017 at 04:10:59PM +0100, Daniel P. Berrange wrote:
On Tue, Mar 28, 2017 at 01:46:31PM +0200, Mart
On Wed, Apr 05, 2017 at 10:50:52AM +0200, Michal Privoznik wrote:
Mostly mem leaks. However, the first one is an actual crasher.
Michal Privoznik (3):
virStorageSourceClear: Don't leave dangling pointers behind
virNetworkObjDispose: Don't leak virMacMap object
virISCSIGetSession: Don't leak m
On Wed, Apr 05, 2017 at 07:03:39 -0400, John Ferlan wrote:
>
>
> On 04/04/2017 08:20 AM, Peter Krempa wrote:
> > For native gluster pools the field denotes a directory inside the
> > pool. For the actual pool name the field has to be used.
> > ---
> > src/storage/storage_util.c
On 04/04/2017 08:20 AM, Peter Krempa wrote:
> For native gluster pools the field denotes a directory inside the
> pool. For the actual pool name the field has to be used.
> ---
> src/storage/storage_util.c | 16
> ++--
> tests/virstorageutildata/glust
On 04/05/2017 05:06 AM, Daniel P. Berrange wrote:
> On Tue, Apr 04, 2017 at 09:51:47PM +0200, Martin Kletzander wrote:
>> On Tue, Apr 04, 2017 at 04:10:59PM +0100, Daniel P. Berrange wrote:
>>> On Tue, Mar 28, 2017 at 01:46:31PM +0200, Martin Kletzander wrote:
The attribute (defined as ATTRI
On Wed, Apr 05, 2017 at 10:06:22AM +0100, Daniel P. Berrange wrote:
On Tue, Apr 04, 2017 at 09:51:47PM +0200, Martin Kletzander wrote:
On Tue, Apr 04, 2017 at 04:10:59PM +0100, Daniel P. Berrange wrote:
> On Tue, Mar 28, 2017 at 01:46:31PM +0200, Martin Kletzander wrote:
> > The attribute (defin
On Thu, Mar 30, 2017 at 02:06:33PM +0200, Peter Krempa wrote:
After restart of libvirtd the 'checkPool' method is supposed to validate
that the pool is online. Since libvirt then refreshes the pool contents
anyways just return whether the pool was supposed to be online so that
the code can be rea
On Wed, Mar 29, 2017 at 04:08:53PM +0200, Andrea Bolognani wrote:
The capabilities used in test cases should match those used
during normal operation for the tests to make any sense.
---
src/qemu/qemu_capabilities.c | 2 +-
src/qemu/qemu_capabilities.h
On Wed, Mar 29, 2017 at 04:08:54PM +0200, Andrea Bolognani wrote:
So far, libvirt has assumed that only x86 supports ACPI,
but that's inaccurate since aarch64 supports it too.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1429509
---
src/qemu/qemu_capabilities.c | 2
On Tue, Apr 04, 2017 at 09:51:47PM +0200, Martin Kletzander wrote:
> On Tue, Apr 04, 2017 at 04:10:59PM +0100, Daniel P. Berrange wrote:
> > On Tue, Mar 28, 2017 at 01:46:31PM +0200, Martin Kletzander wrote:
> > > The attribute (defined as ATTRIBUTE_NONNULL) was added long time
> > > ago (2009), bu
This function runs an iscsi command and parses its output.
However, due to the nature of things, virISCSIExtractSession()
callback can be called multiple times. In each run it would
allocate new memory and overwrite the variable where we keep
pointer to it and thus leaking old allocations.
Signed-
Imagine that this function is called twice over the same disk
source. While in the first run all allocated memory is freed, not
all pointers are set to NULL (e.g. def->srcpool). So when called
again, these poitners are freed again resulting in double free.
Signed-off-by: Michal Privoznik
---
src
Even though the virMacMap object is not necessarily created at
the same time as the network object, the former makes no sense
without the latter and thus should be unref'd in the network
object dispose function.
Signed-off-by: Michal Privoznik
---
src/conf/network_conf.c | 1 +
1 file changed, 1
Mostly mem leaks. However, the first one is an actual crasher.
Michal Privoznik (3):
virStorageSourceClear: Don't leave dangling pointers behind
virNetworkObjDispose: Don't leak virMacMap object
virISCSIGetSession: Don't leak memory
src/conf/network_conf.c | 1 +
src/util/viriscsi.c
On Wed, Apr 05, 2017 at 08:12:31AM +0200, Erik Skultety wrote:
> On Tue, Apr 04, 2017 at 04:23:18PM +0100, Daniel P. Berrange wrote:
> > On Wed, Mar 29, 2017 at 02:51:10PM +0200, Erik Skultety wrote:
> > > This series enables the node device driver to report information about the
> > > existing med
On Wed, Apr 05, 2017 at 09:13:00 +0100, Daniel Berrange wrote:
> On Wed, Apr 05, 2017 at 09:56:13AM +0200, Peter Krempa wrote:
> > docs/schemas directory is meant for schemas which are installed on the
> > system. The schema for the news file does not need to be installed.
> > Store it along with t
On Wed, Apr 05, 2017 at 09:56:13AM +0200, Peter Krempa wrote:
> docs/schemas directory is meant for schemas which are installed on the
> system. The schema for the news file does not need to be installed.
> Store it along with the file it describes for simplicity.
> ---
> Patch for the alternative
On 04/04/2017 04:31 PM, Daniel P. Berrange wrote:
This patch series is a followup to Wojtek's v2 posting:
https://www.redhat.com/archives/libvir-list/2017-March/msg00838.html
It contains the fixes I pointed out in v1 / v2, along with a bunch of
updates to the event-test.py example, so we can
docs/schemas directory is meant for schemas which are installed on the
system. The schema for the news file does not need to be installed.
Store it along with the file it describes for simplicity.
---
Patch for the alternative of moving news rng to docs/ along with the file it
should describe.
do
On 04/04/2017 04:05 PM, Peter Krempa wrote:
qemu requires that the topology equals to the maximum vcpu count.
Document this along with the API to set maximum vcpu count and the XML
element.
Resolves: https://bugzilla.redhat.com/show_bug.cgi?id=1426220
---
docs/formatdomain.html.in | 5 -
sr
Peter Krempa [2017-04-05, 08:44AM +0200]:
On Wed, Apr 05, 2017 at 07:58:36 +0200, Bjoern Walk wrote:
Daniel P. Berrange [2017-04-04, 07:24PM +0200]:
> The news.rng file gets picked up by the wildcard match on docs/schemas,
> and so gets installed into /usr/share/libvirt, along with the schemas
77 matches
Mail list logo