Use Kokkos::Parallel in Solver Utils

[julianlitz]

Merge Request - GuideLine Checklist

Guideline to check code before resolve WIP and approval, respectively.
As many checkboxes as possible should be ticked.

Checks by code author:

Always to be checked:

• There is at least one issue associated with the pull request.
• New code adheres with the coding guidelines
• No large data files have been added to the repository. Maximum size for files should be of the order of KB not MB. In particular avoid adding of pdf, word, or other files that cannot be change-tracked correctly by git.

If functions were changed or functionality was added:

• Tests for new functionality has been added
• A local test was succesful

If new functionality was added:

• There is appropriate documentation of your work. (use doxygen style comments)

If new third party software is used:

• Did you pay attention to its license? Please remember to add it to the wiki after successful merging.

If new mathematical methods or epidemiological terms are used:

• Are new methods referenced? Did you provide further documentation?

Checks by code reviewer(s):

• Is the code clean of development artifacts e.g., unnecessary comments, prints, ...
• The ticket goals for each associated issue are reached or problems are clearly addressed (i.e., a new issue was introduced).
• There are appropriate unit tests and they pass.
• The git history is clean and linearized for the merge request. All reviewers should squash commits and write a simple and meaningful commit message.
• Coverage report for new code is acceptable.
• No large data files have been added to the repository. Maximum size for files should be of the order of KB not MB. In particular avoid adding of pdf, word, or other files that cannot be change-tracked correctly by git.

[julianlitz]

@EmilyBourne Can you make "ExactSolution" callable on GPU? Then all Kokkos Loops should be finished.

[EmilyBourne]

@EmilyBourne Can you make "ExactSolution" callable on GPU? Then all Kokkos Loops should be finished.

Ah, I had hoped to avoid this 😅 I will open a PR to remove the inheritance from ExactSolution so it can be called from GPU

[EmilyBourne]

@EmilyBourne Can you make "ExactSolution" callable on GPU? Then all Kokkos Loops should be finished.

Ah, I had hoped to avoid this 😅 I will open a PR to remove the inheritance from ExactSolution so it can be called from GPU

I have put this in draft temporarily until this is fixed

[EmilyBourne]

@EmilyBourne Can you make "ExactSolution" callable on GPU? Then all Kokkos Loops should be finished.

Ah, I had hoped to avoid this 😅 I will open a PR to remove the inheritance from ExactSolution so it can be called from GPU

I have put this in draft temporarily until this is fixed

Actually we can avoid removing the inheritance. We just need to put either a

void calculate_error(Vector<double> error, ConstVector<T> r, ConstVector<T> theta, ConstVector<T> exact_solution) const;

method in ExactSolver or a method:

void exact_solution(ConstVector<T> r, ConstVector<T> theta, Vector<T> exact) const;

@AbdelhadiKara will take care of this

[julianlitz]

@AbdelhadiKara I tried fixing my previous version based on your suggestions.
I think your approach does a better job at avoiding the re-allocation per iteration.
Though as it is only used for testing it doesn't really matter to me.

In my version I just used a single host vector which is then transfered to the error vector (device memory in the future). Then I just used the provided subract() function of the LinearAlgebra class.

[codecov]

Codecov Report

❌ Patch coverage is 98.07692% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 95.75%. Comparing base (a87782a) to head (1847b87).

Files with missing lines	Patch %	Lines
include/GMGPolar/solver.h	98.07%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #270      +/-   ##
==========================================
+ Coverage   95.73%   95.75%   +0.02%     
==========================================
  Files          79       79              
  Lines        8492     8535      +43     
==========================================
+ Hits         8130     8173      +43     
  Misses        362      362              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[AbdelhadiKara]

is there a possibility to avoid if condition inside the kernel ?

[julianlitz]

I don't think so.
When everything is ported to the GPU we can evaluate how the kernels can be optimized.

[julianlitz]

I thought the issue isn't not necessarily the branches but the amount a time a thread idles to wait for expensive computations in another thread execution path to finish. (Which is minimal here)
The waiting time for =* 3.0/4.0 to finish is almost none.