Halos - making better use of halos to improve parallel performance

[edoddridge]

As noted in #218, there are a lot of exchange calls that could be removed if the halos were being used more efficiently. This PR does just that, and partially closes #218.

There are a few bugs fixed in this PR:

• SOR solver was using eta_star values from the halo that had never been computed. Oops. It's fixed by wrapping eta_star in that subroutine. The performance hit is irrelevant, since the internal solver is orders of magnitude slower, and can only be used on a single core.
• diagnostic outputs did not exclude the halos when computing statistics. They do now. This required blessing new diagnostic output.

The real meat of this PR is flushing many exchange calls, which leads to substantial improvements in scaling and performance improvement when running in parallel. The 1.5 layer version is now faster when running on multiple cores than it is on a single core, which is good. The previous scaling was embarrassing.

Still to do:

• [x] update docs with new plots.

[codecov[bot]]

Codecov Report

Merging #220 into master will decrease coverage by 0.05%. The diff coverage is 100%.

@@            Coverage Diff             @@
##           master     #220      +/-   ##
==========================================
- Coverage   91.82%   91.76%   -0.06%     
==========================================
  Files          20       20              
  Lines        1822     1810      -12     
  Branches       96       95       -1     
==========================================
- Hits         1673     1661      -12     
  Misses        124      124              
  Partials       25       25

Impacted Files	Coverage Δ
src/model_main.f90	96.66% <ø> (-0.06%)	:arrow_down:
src/bernoulli.f90	100% <100%> (ø)	:arrow_up:
src/momentum.f90	92.1% <100%> (-0.14%)	:arrow_down:
src/aronnax.f90	94.18% <100%> (ø)	:arrow_up:
src/barotropic_mode.f90	98.59% <100%> (ø)	:arrow_up:
src/advection_schemes.f90	100% <100%> (ø)	:arrow_up:
src/io.f90	94.73% <100%> (ø)	:arrow_up:
src/thickness.f90	97.61% <100%> (-0.06%)	:arrow_down:
src/boundaries.f90	92.77% <100%> (+0.08%)	:arrow_up:
src/vorticity.f90	100% <100%> (ø)	:arrow_up:
... and 4 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 50215d9...efdff6f. Read the comment docs.