| Commit message (Collapse) | Author | Age | Files | Lines |
... | |
| |
|
|
|
|
| |
Also contains simple test suite to test indications.
|
|
|
|
|
|
|
|
| |
This is a fully functional basic implementation of the provider.
Maximum number of enumerated instances is limited, see the code.
TODO list: https://fedorahosted.org/openlmi/ticket/142
|
|
|
|
|
|
|
|
| |
Found by Coverity Scan. The change is really not needed, the iter variable
is initialized NULL and subsequent g_variant_iter_next() call should return
immediately, preventing the cycle to run.
But, just to be on a safe side...
|
| |
|
|
|
|
|
|
| |
Tracing messages logging entries and exits of functions/methods polute
the trace logs greatly. Let's make them less severe making them
easy to suppress, while keeping other interesting messages enabled.
|
|
|
|
|
|
|
|
|
|
|
|
| |
Tracing decorators are everywhere and they make debugging very
inconvenient. This allows to disable them easily using environment
variable.
To disable them, just add
export LMI_DISABLE_TRACING_DECORATORS=1
to your
/usr/libexec/pegasus/$provider-cimprovagt
script.
|
|
|
|
| |
use track instead
|
| |
|
|
|
|
| |
Added a short introduction to OpenLMI Software provider documentation.
|
|
|
|
| |
specified regular expression for Version and Release properties
|
|
|
|
|
|
|
| |
Compilation fails with older schema (preceding 2.38.0) because of
missing Values in CIM_Processor::Family's qualifiers. Let's check
whether the values get generated by konkret and fallback to custom
defines.
|
|
|
|
|
| |
Add a note about the need to change the SystemClassName in config when
sfcb is used.
|
|
|
|
|
| |
Regular expressions could parse nevra and envra strings wrongly. Also
increased the rebustness of checks for SoftwareIdentity::InstanceID.
|
|\ |
|
| |
| |
| |
| | |
The thread did not stop when AttributeError occured.
|
| | |
|
| |
| |
| |
| |
| | |
There is no reason to define own ORGID. Use the common one. Fixes some
compilation warnings.
|
|/ |
|
|
|
|
|
|
|
| |
And get rid of sblim-cmpi-base dependency.
It would be nice if our tests have an option to select Linux_ComputerSystem
or PG_ComputerSystem in the future...
|
|
|
|
|
| |
This commit sort of reverts commit
657e5bfee810e4a51b33ffd401aa307862491423.
|
|
|
|
| |
Add EnumerateInstances() and GetInstance().
|
|
|
|
|
|
|
|
|
|
| |
Setting an error on top of an existing error is considered a programmer's
bug, possibly leaking memory. So do proper cleanup before reusing that
variable.
Also, errors set while retrieving keys and values are not checked anywhere
so let's just don't set them. Returned NULL values in case of error should
be enough to skip the particular value.
|
| |
|
| |
|
|
|
|
| |
LMI_Processor_Family_Other was undefined
|
| |
|
|
|
|
| |
Bumped version in setups to 0.3.0.
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
It may happen that get_providers are called twice in single python
process. Let's prevent an error to be raised due to double instantiation
of IndicationManager which is a singleton.
|
|
|
|
| |
Let's ignore trailing slashes when comparing repository urls.
|
|
|
|
|
| |
Added AssocClass property to Associator(Name)?s calls which limits the
number of association classes being enumerated by cimom to 1.
|
|
|
|
|
|
|
|
| |
Be more benevolent when checking association class instances for a references to
ComputerSystem.
Fixed also few cim class inheritance checks to correspond with mof
files.
|
| |
|
| |
|
|
|
|
|
|
| |
Use dict parameter in first_instance method
Signed-off-by: Roman Rakus <rrakus@redhat.com>
|
|
|
|
| |
Signed-off-by: Roman Rakus <rrakus@redhat.com>
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
| |
Note that the theme is BSD-licensed.
|
| |
|
|
|
|
|
|
|
|
|
| |
In overall documentation, generated by tools/gendoc utility, we do not want
to include description of classes from individual OpenLMI providers in their
docs.
Instead, we want to have one reference with _all_ OpenLMI classes, which will
the individual providers reference.
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
User deletetion is done by DeleteUser method. Added similar method
to delete a group.
Also do the check if the group isn't a primary group of a user.
Changed documentation accordingly.
Signed-off-by: Roman Rakus <rrakus@redhat.com>
|