Re: MPS codegen

[Helmut Eller]

On Thu, Jun 13 2024, Gerd Möllmann wrote:

>> The patches below use a code generator for most of the fix an mirror
>> functions.  I dropped tree-sitter for now; maybe it can be used for
>> something later.  The code generator is about 800 lines and the
>> generated code 1600.  However the generated code is longer than hand
>> written code.  WDYT?
>
> Hm, to be honest, how do I say it? Please don't feel offended, but I
> think I don't like it that much :-).

I'm also not fully convinced that code generation is an improvement.

> I think the thing I don't like is that I believe there should be only
> one description of an object's layout. Be it C (struct), or be it
> somethings else, from which C is generated.

Sure, but that's easier said than done.


BTW, the MPS documentation says that mps_reserve allocates a block.  In
MPS, a block is not necessarily a single object.  One possible
interpretation of this could be that mps_reserve can be used to allocate
a block that contains many objects.  Perhaps we could use a single block
for the entire dump?  E.g. implement a version of dump_mmap_contiguous
that uses mps_reserve.  What's your interpretation?