Otter profiling code for #1774

[LonelyCat124]

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Project coverage is 99.85%. Comparing base (12addeb) to head (94beae1). Report is 2016 commits behind head on master.

Additional details and impacted files

@@           Coverage Diff           @@
##           master    #2535   +/-   ##
=======================================
  Coverage   99.85%   99.85%           
=======================================
  Files         356      356           
  Lines       47861    47942   +81     
=======================================
+ Hits        47793    47874   +81     
  Misses         68       68           

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

[LonelyCat124]

Waiting on #2537 and #2538 to continue this.

[LonelyCat124]

Preprocessor support added so code generation can now work correctly.

I want to force some names to remain fixed, and I want to have a unit test for the OmpSerialRegion as well.

Edit: That test is added, just force fixed names remains when that PR is done.

[LonelyCat124]

@arporter I think this is ready for a first look - there are some statement orderings that are slightly different to what I did by hand (when the code adds taskwaits in itself) but they don't change the functionality so I have no issue leaving them as is.

[LonelyCat124]

We don't need this in a release for the end of the tasking project - so there's no reason to rush completion of this PR. I think the logic is all here for this and we can decide how we want to fit this into PSyclone (perhaps as Joerg suggests in the examples, though I think we still end up with some extensions of the transformation to be able to do this correctly inside the example script)

[arporter]

We don't need this in a release for the end of the tasking project - so there's no reason to rush completion of this PR. I think the logic is all here for this and we can decide how we want to fit this into PSyclone (perhaps as Joerg suggests in the examples, though I think we still end up with some extensions of the transformation to be able to do this correctly inside the example script)

Great. I think we go for an example then. @hiker, feel free to add your comments about this on here.

[arporter]

Since this modifies core PSyIR code for a tool (Otter) that is currently experimental/in development, we've agreed that this PR is not the way to go. However, we'll keep the branch in case it becomes useful in future.