Improve edgecase handling for empty trees

[lukasdreyer]

Describe your changes here:
When trying to run our testsuite on 8 procs, several tests failed, so this PR fixes some of the problems related to empty trees.

All these boxes must be checked by the reviewers before merging the pull request:

As a reviewer please read through all the code lines and make sure that the code is fully understood, bug free, well-documented and well-structured.

General

• The reviewer executed the new code features at least once and checked the results manually

• The code follows the t8code coding guidelines

• New source/header files are properly added to the Makefiles

• The code is well documented

• All function declarations, structs/classes and their members have a proper doxygen documentation

• All new algorithms and data structures are sufficiently optimal in terms of memory and runtime (If this should be merged, but there is still potential for optimization, create a new issue)

Tests

• The code is covered in an existing or new test case using Google Test

Github action

• The code compiles without warning in debugging and release mode, with and without MPI (this should be executed automatically in a github action)

• All tests pass (in various configurations, this should be executed automatically in a github action)

  If the Pull request introduces code that is not covered by the github action (for example coupling with a new library):

  • Should this use case be added to the github action?
  • If not, does the specific use case compile and all tests pass (check manually)

Scripts and Wiki

• If a new directory with source-files is added, it must be covered by the script/find_all_source_files.scp to check the indentation of these files.
• If this PR introduces a new feature, it must be covered in an example/tutorial and a Wiki article.

Licence

• The author added a BSD statement to doc/ (or already has one)

[jmark]

@lukasdreyer I ran the test with 3 ranks (also with 8 ranks) in debug mode and it failed. For one and two ranks the following assertion does not fail:

mpirun -n 3 fix-empty_tree_handling/test/t8_gtest_iterate_replace
[...]
[t8 0] Building global first descendants for forest 0x13289a0
[t8 1] Building global first descendants for forest 0x25e2320
[t8 2] Building global first descendants for forest 0xa37220
[libsc 0] Abort: Assertion 'last_desc_id < first_desc_id'
[libsc 0] Abort: /home/mark_jo/codes/t8code/main/src/t8_forest/t8_forest_partition.cxx:245
[libsc 0] Abort: Obtained 21 stack frames
[...]

See following for full output.
out.txt

[lukasdreyer]

@lukasdreyer I ran the test with 3 ranks (also with 8 ranks) in debug mode and it failed. For one and two ranks the following assertion does not fail:

mpirun -n 3 fix-empty_tree_handling/test/t8_gtest_iterate_replace
[...]
[t8 0] Building global first descendants for forest 0x13289a0
[t8 1] Building global first descendants for forest 0x25e2320
[t8 2] Building global first descendants for forest 0xa37220
[libsc 0] Abort: Assertion 'last_desc_id < first_desc_id'
[libsc 0] Abort: /home/mark_jo/codes/t8code/main/src/t8_forest/t8_forest_partition.cxx:245
[libsc 0] Abort: Obtained 21 stack frames
[...]

See following for full output. out.txt

The assertion is definitely wrong and maybe we can talk on the phone how to write a better assertion there. I think during debugging I already once discovered that this assertion is wrong and simply deleted it, but forgot to include it in the PR.

[lukasdreyer]

@lukasdreyer I ran the test with 3 ranks (also with 8 ranks) in debug mode and it failed. For one and two ranks the following assertion does not fail:

mpirun -n 3 fix-empty_tree_handling/test/t8_gtest_iterate_replace
[...]
[t8 0] Building global first descendants for forest 0x13289a0
[t8 1] Building global first descendants for forest 0x25e2320
[t8 2] Building global first descendants for forest 0xa37220
[libsc 0] Abort: Assertion 'last_desc_id < first_desc_id'
[libsc 0] Abort: /home/mark_jo/codes/t8code/main/src/t8_forest/t8_forest_partition.cxx:245
[libsc 0] Abort: Obtained 21 stack frames
[...]

See following for full output. out.txt

The assertion is definitely wrong and maybe we can talk on the phone how to write a better assertion there. I think during debugging I already once discovered that this assertion is wrong and simply deleted it, but forgot to include it in the PR.

I adapted the assertion, so that t8_gtest_iterate_replace now should run on multiple procs.
Altough i still think that there can be cases, where this assertion is too strict, but I added a corresponding comment in the commit.