Oakley Brunt

> Also, what was the reason for having the 'bare' undf name - I'm sure there was one but I can't remember! It was something that I had put in...

> > > On master, there is no check that this isn't a Builtin - we always added a field argument. This may be because this method is never called...

> > I don't think it is, no. However, this is the change that fixed that problem with the examples failing (I don't know how, it just did). > >...

> > I don't think it is, no. However, this is the change that fixed that problem with the examples failing (I don't know how, it just did). > >...

Ah, I removed the `isinstance` check from `KernCallArgList` and it caused the `src/psyclone/tests/domain/lfric/transformations/dynamo0p3_transformations_test.py::test_accenterdata_builtin` test to fail. The test file it is calling has a builtin and a cell_column kernel in...

Is there a possibility that when ACC transformations happen, the `KernCallArgList` is remade with no checks for where the arguments originate from? I'm not entirely sure what's happening

> I think the sub-class `KernCallACCArgList` is used to work out what data needs to be transferred to the GPU. Perhaps this method gets called by that class? That is...

As I am working on making the dof kernel (`src/psyclone/tests/test_files/dynamo0p3/testkern_dofs_mod.f90`) compile, I noticed again that the arguments passed to the subroutine are single dofs, real or integer. This makes me...

> Pretty much done now. The only thing left is to add a bit more context in a couple of the tests so that we can be sure that the...

If `LFRicKern` and `LFRicKernMetadata` were split, there would have to be some way of determining which kernel type to initialise for each invoke. This could maybe live in `LFRicAlg::kernel_from_metadata()` with...