Performance optimization for NEC SX AURORA

What is the bug

The subroutines init_zero_4d_[dp,sp,i4] do not vectorize properly on NEC SX AURORA.

How do you fix it

Loop collapsing is enforced by a compiler directive for the subroutines init_zero_4d_[dp,sp,i4], in order to ensure proper vectorization on NEC SX AURORA.

How urgent is the bugfix

• I need it as soon as possible
• I can wait for a couple of days
• None of my current codes is directly affected

Mandatory steps before review

• Gitlab CI passes (Hint: use make format for linting)
• Bugfix is covered by additional unit tests
• Mark the merge request as ready by removing Draft:

Mandatory steps before merge

• Test coverage does not decrease
• Reviewed by a maintainer
• Incorporate review suggestions
• Prior to merging, please remove any boilerplate from the MR description, retaining only the What is the bug and How do you fix it section to maintain
• Remember to edit the commit message and select the proper changelog category (feature/bugfix/other)

You are not supposed to merge this request by yourself, the maintainers of fortan-support take care of this action!

src/mo_fortran_tools.F90

@@ -1201,6 +1201,7 @@ CONTAINS
 #else
 !$omp do collapse(4)
 #endif
+!NEC$ forced_collapse
     DO i4 = 1, m4
       DO i3 = 1, m3
         DO i2 = 1, m2
@@ -1235,6 +1236,7 @@ CONTAINS
 #else
 !$omp do collapse(4)
 #endif
+!NEC$ forced_collapse
     DO i4 = 1, m4
       DO i3 = 1, m3
         DO i2 = 1, m2
@@ -1269,6 +1271,7 @@ CONTAINS
 #else
 !$omp do collapse(4)
 #endif
+!NEC$ forced_collapse
     DO i4 = 1, m4
       DO i3 = 1, m3
         DO i2 = 1, m2