change mixed-precision kernels to use iso-ieee kinds

[Matthew Walker (mattatmet)]

PR Summary

Sci/Tech Reviewer: Benjamin Went (@MetBenjaminWent)
Code Reviewer: Ed Hone (@EdHone)

This PR is a replica of the Trac ticket #4463:

Some of the mixed precision kernels are using LFRic defined kinds of r_single and r_double. These are just labels in constants_mod.F90 and resolve to real32 and real64 respectively. However, the other mixed precision code is using the iso-ieee kinds. Moreover, we want to be explicit and unambiguous for compiler performance, especially for performance on GPUs where we may also want further reduced kinds. This ticket will re-write the kernels in the native kinds of real32 and real64.

Code Quality Checklist

• I have performed a self-review of my own code
• My code follows the project's style guidelines
• Comments have been included that aid understanding and enhance the readability of the code
• My changes generate no new warnings
• All automated checks in the CI pipeline have completed successfully

Testing

• I have tested this change locally, using the LFRic Core rose-stem suite
• If required (e.g. API changes) I have also run the LFRic Apps test suite using this branch
• If any tests fail (rose-stem or CI) the reason is understood and acceptable (e.g. kgo changes)
• I have added tests to cover new functionality as appropriate (e.g. system tests, unit tests, etc.)
• Any new tests have been assigned an appropriate amount of compute resource and have been allocated to an appropriate testing group (i.e. the developer tests are for jobs which use a small amount of compute resource and complete in a matter of minutes)

trac.log

Test Suite Results - lfric_core - mixed-precision2-core/run1

Suite Information

Item	Value
Suite Name	mixed-precision2-core/run1
Suite User	matthew.walker
Workflow Start	2026-06-12T12:47:13
Groups Run	developer

Dependency	Reference	Main Like
lfric_core	mattatmet/lfric_core@iso-ieee-kinds	False
SimSys_Scripts	MetOffice/SimSys_Scripts@cab3315	True

Task Information

✅ succeeded tasks - 422

Security Considerations

• I have reviewed my changes for potential security issues
• Sensitive data is properly handled (if applicable)
• Authentication and authorisation are properly implemented (if applicable)

Performance Impact

• Performance of the code has been considered and, if applicable, suitable performance measurements have been conducted

AI Assistance and Attribution

• Some of the content of this change has been produced with the assistance of Generative AI tool name (e.g., Met Office Github Copilot Enterprise, Github Copilot Personal, ChatGPT GPT-4, etc) and I have followed the Simulation Systems AI policy (including attribution labels)

Documentation

• Where appropriate I have updated documentation related to this change and confirmed that it builds correctly

PSyclone Approval

• If you have edited any PSyclone-related code (e.g. PSyKAl-lite, Kernel interface, optimisation scripts, LFRic data structure code) then please contact the TCD Team

Sci/Tech Review

• I understand this area of code and the changes being added
• The proposed changes correspond to the pull request description
• Documentation is sufficient (do documentation papers need updating)
• Sufficient testing has been completed

(Please alert the code reviewer via a tag when you have approved the SR)

Code Review

• All dependencies have been resolved
• Related Issues have been properly linked and addressed
• CLA compliance has been confirmed
• Code quality standards have been met
• Tests are adequate and have passed
• Documentation is complete and accurate
• Security considerations have been addressed
• Performance impact is acceptable

[Thomas Bendall (tommbendall)]

This seems like a sensible change, thanks for making it.

My only comment is that just replacing r_single with real32 and similar for r_double means that there are now quite a few code sections which aren't aligned. Sorry to be tedious, but could you go through your changed kernels and realign any changed sections? I have posted a couple of examples so you see what I mean.

[Thomas Bendall (tommbendall)]

Suggested change

	subroutine dg_inc_matrix_vector_code_real64(cell, &
	subroutine dg_inc_matrix_vector_code_real64( cell, &

[Thomas Bendall (tommbendall)]

Suggested change

	subroutine dg_inc_matrix_vector_code_real32(cell, &
	subroutine dg_inc_matrix_vector_code_real32( cell, &

[Thomas Bendall (tommbendall)]

Suggested change

	subroutine restrict_scalar_unweighted_code_real64( &
	subroutine restrict_scalar_unweighted_code_real64( &

[Thomas Bendall (tommbendall)]

Suggested change

	real(kind=real64), intent(inout) :: coarse_field(undf_coarse)
	real(kind=real64), intent(in) :: fine_field(undf_fine)
	real(kind=real64), intent(inout) :: coarse_field(undf_coarse)
	real(kind=real64), intent(in) :: fine_field(undf_fine)

[Thomas Bendall (tommbendall)]

Suggested change

	real(kind=real64) :: denom, coarse_value(nlayers-1+ndf)
	real(kind=real64) :: denom, coarse_value(nlayers-1+ndf)

[Matthew Walker (mattatmet)]

My only comment is that just replacing r_single with real32 and similar for r_double means that there are now quite a few code sections which aren't aligned. Sorry to be tedious, but could you go through your changed kernels and realign any changed sections? I have posted a couple of examples so you see what I mean.

Thanks for this Tom, I hadn't noticed that these changes had knocked everything out of alignment. Have gone through it carefully and I think I've fixed everything. Let me know if I missed anything though, cheers.

[Thomas Bendall (tommbendall)]

Great, thanks for neatening that up!

[Benjamin Went (MetBenjaminWent)]

Changes look good! I've checked this out locally and confirmed that all of these are removed from the kernels as advised in the PR summary.

I have however also noticed a few small formatting changes which could be updated?

Also whats the plan for lfric apps? I appreciate it's out of scope for this ticket, just some context present, like a linked issue apps side could be good here?

Besides these small changes, it's otherwise approved! Thanks

[Matthew Walker (mattatmet)]

I have however also noticed a few small formatting changes which could be updated?

Also whats the plan for lfric apps? I appreciate it's out of scope for this ticket, just some context present, like a linked issue apps side could be good here?

Benjamin Went (@MetBenjaminWent) Thanks Ben! I've made those changes now. With regards to lfric apps, there is no plan to extend this to apps yet I don't think.

[Benjamin Went (MetBenjaminWent)]

Thanks Matthew! Changes look good, SR approved.

[Ed Hone (EdHone)]

All changes look fine - the only request I would have would be for all the new use statements for iso_fortran_env be moved to the top in their own block, so they aren't confused with lfric modules: i.e., the first instance would become:

module dg_inc_matrix_vector_kernel_mod

  use, intrinsic :: iso_fortran_env, only: real32, real64

  use argument_mod,            only : arg_type,                  &
                                      GH_FIELD, GH_OPERATOR,     &
                                      GH_REAL, GH_READ,          &
                                      GH_READWRITE, ANY_SPACE_1, &
                                      ANY_DISCONTINUOUS_SPACE_1, &
                                      CELL_COLUMN
  use constants_mod,           only : i_def
  use kernel_mod,              only : kernel_type

[Matthew Walker (mattatmet)]

All changes look fine - the only request I would have would be for all the new use statements for iso_fortran_env be moved to the top in their own block, so they aren't confused with lfric modules: i.e., the first instance would become:

Thanks for the suggestion Ed Hone (@EdHone), I've made these changes and pushed them.

[Ed Hone (EdHone)]

There are still some style/lint errors to fix here:

https://cylchub/services/cylc-review/taskjobs/edward.hone?&suite=lfric_core-iso-ieee-kinds%2Frun1&cycles=1&task_status=submit-failed&task_status=failed

[Matthew Walker (mattatmet)]

There are still some style/lint errors to fix here:

https://cylchub/services/cylc-review/taskjobs/edward.hone?&suite=lfric_core-iso-ieee-kinds%2Frun1&cycles=1&task_status=submit-failed&task_status=failed

Oops, sorry about that. Fixed.

[Ed Hone (EdHone)]

There were a few final whitespace issues to fix but I sorted them out

[Matthew Walker (mattatmet)]

Thanks very much Ed!