I've recently ported libfirm to D.

This nice C library, developed at the Karlsruhe university, allows to build compiler back-ends, using the SSA intermediate representation.

In theory it could even be used to make a new D compiler version, e.g "FDC", although this is obviously of no interest ;)

So this is mostly to make JIT scripting engines or back-ends for toy languages...

See:

- https://github.com/BBasile/libfirm-d
- https://code.dlang.org/packages/firm-d
- https://pp.ipd.kit.edu/firm/

Note that unfortunately this library is LGPL2 licensed, so much less usable than LLVM-d, which is a similar but a more mainstream library.

Basile B. wrote:

> I've recently ported libfirm to D.

great news and great work, thank you. yet i must admit that wrapping != porting. for a moment you made my heart stopped, 'cause i thought that i wasted alot of time trying to do a proper port (nope, not finished, but still...).

On Sunday, 1 July 2018 at 15:03:30 UTC, ketmar wrote:
> Basile B. wrote:
>
>> I've recently ported libfirm to D.
>
> great news and great work, thank you. yet i must admit that wrapping != porting. for a moment you made my heart stopped, 'cause i thought that i wasted alot of time trying to do a proper port (nope, not finished, but still...).

Thanks. BTW I remember you said on IRC that SSA form is much better (maybe it was once when we talked about DMD optimizations or something like that)... this is basically why i choose this one, although not sure if i'll manage to do something with it.

Basile B. wrote:

> On Sunday, 1 July 2018 at 15:03:30 UTC, ketmar wrote:
>> Basile B. wrote:
>>
>>> I've recently ported libfirm to D.
>>
>> great news and great work, thank you. yet i must admit that wrapping != porting. for a moment you made my heart stopped, 'cause i thought that i wasted alot of time trying to do a proper port (nope, not finished, but still...).
>
> Thanks. BTW I remember you said on IRC that SSA form is much better (maybe it was once when we talked about DMD optimizations or something like that)... this is basically why i choose this one, although not sure if i'll manage to do something with it.

tbh, i gave up on porting it fully, so i prolly will go with your wrapper, if anything. and yeah, SSA rox. i am temporarily on C++ side now, yet sooner or later i'll return, and will have to restart Aliced again. maybe this time with a brand new backend. ;-)

On Sunday, 1 July 2018 at 12:04:06 UTC, Basile B. wrote:
> I've recently ported libfirm to D.
>
> This nice C library, developed at the Karlsruhe university, allows to build compiler back-ends, using the SSA intermediate representation.
>
> In theory it could even be used to make a new D compiler version, e.g "FDC", although this is obviously of no interest ;)
>
> So this is mostly to make JIT scripting engines or back-ends for toy languages...
>
> See:
>
> - https://github.com/BBasile/libfirm-d
> - https://code.dlang.org/packages/firm-d
> - https://pp.ipd.kit.edu/firm/
>
> Note that unfortunately this library is LGPL2 licensed, so much less usable than LLVM-d, which is a similar but a more mainstream library.

Very nice, I remember checking this one out a while back.

I don't see the files from the ADT module though, was there a reason they weren't included?

https://github.com/libfirm/libfirm/tree/master/include/libfirm/adt

On Thursday, 5 July 2018 at 00:00:07 UTC, Chris M. wrote:
> On Sunday, 1 July 2018 at 12:04:06 UTC, Basile B. wrote:
>> I've recently ported libfirm to D.
>>
>> This nice C library, developed at the Karlsruhe university, allows to build compiler back-ends, using the SSA intermediate representation.
>
> Very nice, I remember checking this one out a while back.
>
> I don't see the files from the ADT module though, was there a reason they weren't included?
>
> https://github.com/libfirm/libfirm/tree/master/include/libfirm/adt

Yes there's a reason that is that these files mostly contain data structures, they are not needed in D and i don't remember having seen them as parameter type during translation of the headers, (although it's been done semi manually, so i could have missed one).

On Sunday, 1 July 2018 at 15:03:30 UTC, ketmar wrote:
> Basile B. wrote:
>
>> I've recently ported libfirm to D.
>
> great news and great work, thank you. yet i must admit that wrapping != porting.

Right, i have used the wrong words. Title hopefully is not ambiguous.

Basile B. wrote:

> On Sunday, 1 July 2018 at 15:03:30 UTC, ketmar wrote:
>> Basile B. wrote:
>>
>>> I've recently ported libfirm to D.
>>
>> great news and great work, thank you. yet i must admit that wrapping != porting.
>
> Right, i have used the wrong words. Title hopefully is not ambiguous.

ah, tbh, i'm just being jealous. you did my work! me is useless now!

On Thursday, 5 July 2018 at 01:11:58 UTC, Basile B. wrote:
> On Thursday, 5 July 2018 at 00:00:07 UTC, Chris M. wrote:
>> On Sunday, 1 July 2018 at 12:04:06 UTC, Basile B. wrote:
>>> [...]
>>
>> Very nice, I remember checking this one out a while back.
>>
>> I don't see the files from the ADT module though, was there a reason they weren't included?
>>
>> https://github.com/libfirm/libfirm/tree/master/include/libfirm/adt
>
> Yes there's a reason that is that these files mostly contain data structures, they are not needed in D and i don't remember having seen them as parameter type during translation of the headers, (although it's been done semi manually, so i could have missed one).

Makes sense, I didn't think you would have just missed them.

On Sunday, 1 July 2018 at 12:04:06 UTC, Basile B. wrote:
> I've recently ported libfirm to D.
>
> This nice C library, developed at the Karlsruhe university, allows to build compiler back-ends, using the SSA intermediate representation.
>
> In theory it could even be used to make a new D compiler version, e.g "FDC", although this is obviously of no interest ;)
>
> So this is mostly to make JIT scripting engines or back-ends for toy languages...
>
> See:
>
> - https://github.com/BBasile/libfirm-d
> - https://code.dlang.org/packages/firm-d
> - https://pp.ipd.kit.edu/firm/
>
> Note that unfortunately this library is LGPL2 licensed, so much less usable than LLVM-d, which is a similar but a more mainstream library.

With the recent release of dstep 1.0.0 I've managed to make a new minor release that only auto generate the sources (v0.1.0). It was obviously not a good idea to maintain all the headers up to date manually, although this permitted to learn a bit the API.

Note that building must be done at least twice the first time because of a DUB limitation.

other changes :
- dub package on code dlang is gone, it must really be a git stuff now
- it's much faster because build is optimized using git hashes.
- location changed due to GH account renaming

https://github.com/Basile-z/libfirm-d