Merge asm/ and analysis/ into arch/ but keep plugin formats.

[wargio]

DO NOT SQUASH

Your checklist for this pull request

• I've read the guidelines for contributing to this repository
• I made sure to follow the project's coding style
• I've documented or updated the documentation of every function and struct this PR changes. If not so I've explained why.
• I've added tests that prove my fix is effective or that my feature works (if possible)
• I've updated the rizin book with the relevant information (if needed)

Detailed description

This is the first step against rz_arch. this merges still the librz/asm/ and librz/analysis/ into the same folder structure but keeps the same formats.

One change that can be noticed is that now we do have a list of all deactivated plugins.

Another big change is also related how we generate the plugins headers, which before needed to be included by the developer each time you add an arch, but now is automatically generated by meson.

After merging this pr, it will be still possible to load external asm and analysis plugins as standalone libraries.

Also I have merged TMS320C64x into TMS320 because it was a duplicated architecture which was available in both tms320 and tms320c64x with code that was duplicated between the two plugins.

Old PR #4120

[ret2libc]

Why removing the common logic in librz/meson.build? Is it possible to just declare the plugins in each librz/*/meson.build and then construct the header files in librz/meson.build so we avoid duplicated meson code?

[wargio]

t possible to just declare the plugin

because now it's built as a library, so if we split stuff is easier to handle. also bin plugins has RzBinPlugin and RzBinXtrPlugin so that common logic cannot be easily kept.

[wargio]

It also broke TMS320C64x analysis

yeah. that piece of code was made of #include "file.c"

[wargio]

It also broke TMS320C64x analysis:

I merged TMS320C64x into TMS320 because it was a dup.

[ret2libc]

I don't see why we can't use the approach that was used before to avoid the duplicated meson code. We were already handling RzBinXtr and RzBin plugins. Also, we can probably also just have a single rz_plugins.h.in for all the various modules since it is the same.

[ret2libc]

what do you mean that it was never built? if it was here it should have been built. I don't get it.

[wargio]

the code was there, but it was never enabled as plugin. that is why when i added it back we had "test failing" but just because now was showing up.

[XVilka]

LGTM as is. Windows builds and rz-bindgen could be fixed separately after this jumbo PR is merged.

[wargio]

I don't see why we can't use the approach that was used before to avoid the duplicated meson code. We were already handling RzBinXtr and RzBin plugins. Also, we can probably also just have a single rz_plugins.h.in for all the various modules since it is the same.

You were not handling plugins, but just creating a list of plugins which is easier to do.

[wargio]

What's the difference?

The first one creates ad-hoc file per lib. within the header i have a #define and the export definition for each plugin.
The second one creates only one header file for all the libs which has only some #define in it.

The huge difference (and why this is needed) is that if i follow the old way to generate those, i would need to add the headers of all the libs into the config.h just to declare the export definitions. in the old code, since it was just a macro, the compiler ignores it unless is used.

So to avoid to have such horrendous config.h it is better and cleaner to have one specific header file generated by each library (subfolder) and included only by that library.

[ret2libc]

export_plugins += ['extern RzLangPlugin ' + plugin.get('base_cname') + p + ';']

Whoops, this of course needs to be fixed. it should be enough to add a plugin_type field to the various *_plugins dictionary in librz/*/meson.build files.

[ret2libc]

Better now! Thanks 👍