Limit threads for mris_expand

[chrisadamsonmcri]

Limits the threads for mris_expand and changes the OpenMP scheduler to dynamic.

Fixes #490.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 75.03%. Comparing base (afde296) to head (5e0b15a).
⚠️ Report is 131 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #491      +/-   ##
==========================================
+ Coverage   70.87%   75.03%   +4.16%     
==========================================
  Files          25       25              
  Lines        2067     2067              
  Branches      268      268              
==========================================
+ Hits         1465     1551      +86     
+ Misses        544      448      -96     
- Partials       58       68      +10     

Flag	Coverage Δ
ds054	44.95% <ø> (?)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 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.

[effigies]

FWIW, I would delete that _num_threads_update method entirely and update

smriprep/src/smriprep/workflows/surfaces.py

Lines 236 to 241 in afde296

	midthickness = pe.MapNode(
	MakeMidthickness(thickness=True, distance=0.5, out_name='midthickness'),
	iterfield='in_file',
	name='midthickness',
	n_procs=min(omp_nthreads, 12),
	)

to use:

     midthickness = pe.MapNode(
         MakeMidthickness(
             thickness=True,
             distance=0.5,
             out_name='midthickness',
+            environ={'OMP_SCHEDULE': 'dynamic'},
         ),
         iterfield='in_file',
         name='midthickness',
-        n_procs=min(omp_nthreads, 12),
+        n_procs=min(omp_nthreads, 8),
    )

The purpose of the _num_threads_update() was to allow the job to claim,e.g, 8 threads while actually allocating 12. If we aren't doing a trick like that, it's cleaner to put the min/max logic where the node is created. And any Nipype CommandLine interface can take an environ argument, so let's do that there, too.

Incidentally, I thought you were saying that the returns were already diminishing significantly by 4 threads. Are you sure you want to max out at 8?

[effigies]

Sorry, I thought you'd said to close the PR, but now I can't see the comment...