Re-read the whole code, check everything

[phil-blain]

• [x] check all comments to see if they make sense, they are useful, and do not denote things that still need to be done.
• ~[ ] make sure we do as much code reuse as we can (especially for rheology computations)~ see https://github.com/phil-blain/CICE/issues/36
• [x] remove unused variables
• [x] remove unused subroutines
• ~[ ] check arguments vs use statements~ I did that a while ago, I won't change all interfaces again for now.
• [x] rename variables that have unclear names
• [x] compile with -Wall and fix warnings
• [x] fix whitespace inconsistencies

[phil-blain]

I compiled with ice_dyn_vp.o: FFLAGS += -Wall added to Macros.conda_macos. This activates the -Wall flag only for the ice_dyn_vp.o target, i.e. when compiling ice_dyn_vp.F90.

This flag found

• unused subroutines
• unused local variables
• unused "dummy" arguments
• unused use-imported variables in subroutines. (see https://gcc.gnu.org/onlinedocs/gcc/Warning-Options.html and https://gcc.gnu.org/onlinedocs/gfortran/Error-and-Warning-Options.html)

See 083b611, 71eba9e, 2a5f4b7, 7a8951a, 4931ac2

[phil-blain]

I fixed trailing whitespace issues, using git diff $(git empty-tree) cicecore/cicedynB/dynamics/ice_dyn_vp.F90 : 0d527f6

[phil-blain]

I fixed more whitespace issues, and uniformized whitespace for subroutine arguments and intents: 0179edf 7d46f1d 9b6c2bf b32bdfe

[phil-blain]

I did some more code cleanup in 8218e82 and eadcfa6

[phil-blain]

I re-read every code comment in ice_dyn_vp and cleaned up a lot of them, also clarified others in 5ff03ab.

[phil-blain]

Still need to :

• [x] clarify if the conv and r0 variables in FGMRES (and PGMRES) should be removed. I need to compare with the Saad implementation.

• [x] reread the FGMRES algorithm to find a better name and description for ww

[phil-blain]

I did more variable naming clean up in 1416a7e

[phil-blain]

I renamed ww to orig_basis in 9267229

[phil-blain]

I removed the conv and r0 arguments; they were unused. If we change the implementation in the future and we need them, we'll just re-add them.

[phil-blain]

there are no !phb comments anymore. I changed some of them to TODO.

[phil-blain]

I went through the code and there is a lot of repetitions (copy-paste) between different computations regarding rheology. Here is a list of the computations by subroutines:

calc_zeta_dPr (ice_dyn_vp)

• computes zeta
• computes \partial_x P_R, \partial_y Pr (derivatives of the replacement pressure)
• computes `stressp_[1-4] (only P_R part)
• then computes [cs]sigp[nsew12]
• stress_Pr -> used in calc_bvec

stress_vp (ice_dyn_vp)

• computes stresses from zeta
• stress[p,m,12]_[1-4]
• here stressp is the whole term

matvec (ice_dyn_vp)

• computes (partial stresses) from zeta
• stress[p,m,12]_[1-4]
• stressp is only "linear" part (linear in \partial^n_[x,y] u)
• computes [cs]sigp[nsew12]
• computes str

formDiag_step1 (ice_dyn_vp)

• computes strain_rates (with some coefficients = 0 or 1)
• computes stresses (stressp -> only "linear" part)
• computes Drheo (same computation as matvec, but special case for 8 corners...)

formDiag_step2 (ice_dyn_vp)

• similar to the second step of matvec, bu without ccb (i.e. only ccaimp)

stress (ice_dyn_evp)

• computes stress[p,m,12]_[1-4] -> different computations (EVP)
• computes [cs]sig[p,m,12][nsew] -> SAME as matvec
• computes str[pm]_tmp -> SAME as matvec

stress_eap (ice_dyn_eap)

• computes strain_rates
• computes deformations
• computes [cs]sig[p,m,12][nsew] -> SAME as matvec
• computes str[pm]_tmp -> SAME as matvec

We will refactor the code to unify these computations, but in a second step (after the Picard PR).

See https://github.com/phil-blain/CICE/issues/36