http://www.geodynamics.org/hg/magma/3D/stgMADDs/StGermain/ en-us magma/3D/stgMADDs/StGermain Changelog magma/3D/stgMADDs/StGermain Changelog Converted the ObjectList Component's tests from old system to PCU. http://www.geodynamics.org/hg/magma/3D/stgMADDs/StGermain/rev/3d8e62e76e2f
Was helped by the fact that someone (Alan?) had already used the Regresstor functions in this test, which is quite similar to PCU.]]> Patrick Sunter <patdevelop@gmail.com> Thu, 02 Apr 2009 17:42:07 +1030 Converted the MPIRoutines component tests into new pcu system. http://www.geodynamics.org/hg/magma/3D/stgMADDs/StGermain/rev/8d5ecd4843f2
Note: PCU doesn't seem to be completely ready to handle some issues of parallel testing. Will emailLuke and Jules about how to resolve these.]]> Patrick Sunter <patdevelop@gmail.com> Wed, 01 Apr 2009 12:25:25 +1030 Completed transfer of functionality from old testIndexSet into new PCU IndexSetSuite. http://www.geodynamics.org/hg/magma/3D/stgMADDs/StGermain/rev/89cb2e142706
Removing obsolete testIndexSet stuff.]]> Patrick Sunter <patdevelop@gmail.com> Tue, 31 Mar 2009 20:23:31 +1030 Testing revealed an inconsistency in IndexSet_Merge_AND() function when first IndexSet provided was longer than the second, due to nature of bit operations used. http://www.geodynamics.org/hg/magma/3D/stgMADDs/StGermain/rev/a6b066b55f86
Have decided to resolve inconsistency by changing code so that in the case that the first list is longer than the second, all entries after the end of the second list are set to zero. I believe this is most logical for an AND operation. Have updated header file description of the function to explain this. Don't believe it'll cause an issue with other parts of the code, since I don't expect this function has been used with different length lists before.]]> Patrick Sunter <patdevelop@gmail.com> Tue, 31 Mar 2009 20:14:36 +1030 Adding work so far on IndexSetSuite, test suite for IndexSet using new pcu test framework. http://www.geodynamics.org/hg/magma/3D/stgMADDs/StGermain/rev/d259c2125a10
Note: parts that are still commented out are yet to be fully converted to new format.]]> Patrick Sunter <patdevelop@gmail.com> Sat, 28 Mar 2009 15:05:47 +1030 Fixed a small bug in the function IndexSet_IsMember, that appeared through testing. http://www.geodynamics.org/hg/magma/3D/stgMADDs/StGermain/rev/d7674fcc35f8
Basically, the function is defined to return a Bool. However, if the input index was found in the IndexSet,
it was returning a non-zero number based on the IS_MEMBER macro, but this wasn't always 1 (True). Fixed
this by normalising the result of IS_MEMBER.]]> Patrick Sunter <patdevelop@gmail.com> Sat, 28 Mar 2009 14:10:07 +1030 Updated 1 bug and 1 formatting issue in the PCU scripts responsible for creating testStGermain.h: http://www.geodynamics.org/hg/magma/3D/stgMADDs/StGermain/rev/bbc5ed1d5921 - Made sure headers for each test suite have a newline at the end, so the file compiles properly
- added 3 spaces before the "addSuite" lines, so they line up nicely.
(Pat S).]]> pds@localhost Sat, 28 Mar 2009 13:19:02 +1030 Luke fixed a bug in the StGermain test related to the ProjectSuite http://www.geodynamics.org/hg/magma/3D/stgMADDs/StGermain/rev/4cce6e8bf83d dave@lavaeolous Thu, 26 Mar 2009 18:12:38 +1100 Found bugs causing binaries to be installed to the wrong http://www.geodynamics.org/hg/magma/3D/stgMADDs/StGermain/rev/ed58874fd551 locations.]]> LukeHodkinson <luke@vpac.org> Mon, 16 Mar 2009 17:58:23 +1100 Forgot to add these two files to the previous commit. http://www.geodynamics.org/hg/magma/3D/stgMADDs/StGermain/rev/73c0393f4f1e Luke Hodkinson <luke@vpac.org> Wed, 11 Mar 2009 15:19:54 +1100