On Jan 9, 2015, at 5:29 AM, René J.V. Bertin <[email protected]> wrote:

> Unless I'm mistaken and it's the other way round, dragonegg is a gcc 
> front-end to llvm.

It's a GCC plugin that swaps out the backend.

> Is there any interest to such a thing, given that clang aims to be a drop-in 
> replacement for gcc, and that code relying on gcc-specific code 
> generation/optimisation will fail the same way in clang and dragonegg?

Clang is only a replacement for C-family languages. DragonEgg enables the use 
of Fortran and Ada with the LLVM backend.

> Judging from their site (dragonegg.llvm.org), the project isn't particularly 
> well kept up-to-date, as if even the authors fail to see any interest in it?

It's not being maintained currently, as far as I know.

http://llvm.org/bugs/show_bug.cgi?id=18324#c4

> However, if as they claim "fortran works very well", how complicated would it 
> be to provide a dragonegg-3.5-gcc-4.9 containing at least a fortran front-end 
> to llvm? (IIUC, dragonegg-3.4-gcc-4.7 is already an endeavour by Jeremy or 
> another MacPorts dev...)

To provide a port? Not complicated. To provide working software that doesn't 
have upstream backing? Different story.

vq
_______________________________________________
macports-users mailing list
[email protected]
https://lists.macosforge.org/mailman/listinfo/macports-users

Reply via email to