Connor Aird

Profiling results for `test_EXX_isol_C2H4_4proc_PBE0ERI_fullSZP_0.4_SCF` on myriad using the intel compiler and the following modules. ``` Currently Loaded Modulefiles: 1) beta-modules 5) libxc/6.2.2/intel-2022 9) git/2.41.0-lfs-3.3.0 2) gcc-libs/10.2.0 6) cmake/3.21.1 10) emacs/28.1...

@davidbowler The commits above solved the issue so far as ensuring that exx_phi_on_grid was called with an `xys` input that was always initialised to zero, which seems to works for...

After comparing `m_kern_exx_cri` and `m_kern_exx_eri`, it does not look simple to reduce their duplication through refactoring. Many similar things are happening in both but the loop structures are different. For...

After comparing `m_kern_exx_eri` and `m_kern_exx_eri_gto`, the only differences seem to be that arrays are allocated in `m_kern_exx_eri` but not `m_kern_exx_eri_gto` and that `m_kern_exx_eri_gto` calls `compute_eri_hoh` at it's [deepest point](https://github.com/OrderN/CONQUEST-release/blob/f-exx-opt/src/exx_kernel_default.f90#L1878) whereas...

`m_kern_exx_eri` and `m_kern_exx_dummy` are also very similar with the main difference being that arrays are allocated at the beginning of `m_kern_exx_dummy`. Therefore, this can also likely be refactored to deduplicate...

I believe the code inside the `do nsf2 = 1, jb%nsup` loop is the same for `m_kern_exx_cri` and `m_kern_exx_eri`. Therefore, this should be able to be pulled out into a...

Should this be `not-for-arc`?