dmd.toobj: Move the toObjFile call in genTypeInfo to DMD gluelayer

[ibuclaw]

Almost all code in typinf.d is common front-end semantic logic for all compilers, with exception to the toObjFile() call into DMD's gluelayer.

This moves that call to a new helper function genTypeInfoForType in the dmd-specific toobj.d module.

With this, GDC no longer needs to patch any DMD front-end sources in order for it to work with the GCC glue.

[dlang-bot]

Thanks for your pull request, @ibuclaw!

Bugzilla references

Your PR doesn't reference any Bugzilla issue.

If your PR contains non-trivial changes, please reference a Bugzilla issue or create a manual changelog.

Testing this PR locally

If you don't have a local development environment setup, you can use Digger to test this PR:

dub run digger -- build "master + dmd#13690"

[ibuclaw]

There's some funny things going on with type merging. But on closer inspection, sc is only not null from one place when calling genTypeInfo, so we can refactor with that in mind instead.

[ibuclaw]

Passes, but can be further simplified, as for instance getTypeInfoType is only ever called from semantic, so doesn't need sc checking, nor extern(C++).

[kinke]

FWIW, I don't assign any TypeInfos here to modules anymore for LDC. Instead, I emit struct and ad-hoc TypeInfos into each referencing object file, solely from the glue layer. Besides potentially less inter-object-file dependencies and increased optimization potential, this also takes care of all CTFE / speculative instantiation problems.