[LLVMdev] GCC frontend

Vikram Adve vadve at cs.uiuc.edu
Sat Jul 10 13:43:01 PDT 2004


Let me second what Chris said.  It's far easier to write an 
LLVM-to-NewIL translator (or LLVM-to-OCaml back-end), than a 
GCC-to-NewIL front-end.  And that is if you have a compelling reason to 
invent a new IL and write new analyses and transformations for it, 
rather than using an existing IL and compiler system.


On Jul 10, 2004, at 1:07 PM, Chris Lattner wrote:

> On Sat, 10 Jul 2004, Ramu Ramamurthy wrote:
>>     LLVM is an excellent piece of work,
> Thanks!
>> LLVM is one of the few project I know that has used the mysterious GCC
>> front-end. I think it will be a great contribution to the community if
>> you could writeup and document how to interface to the GCC front-end
>> data-structures (like LLVM has done). Specifically:
>> a) detail the gcc front-end tree data-structures
>> b) how to go about interfacing to them
>> c) what are changes that LLVM did to the gcc front end
> While that would be a useful document, I don't think that anyone will 
> be
> jumping to write it unfortunately.  We all have plenty of other
> interesting things to do, and writing docs that makes it easier for 
> people
> to not use LLVM is not a high-priority. :)
>> My interest is to interface to gcc data-structures to produce an
>> open-source language independent IL in Ocaml -- like LLVM but at a
>> higher level suitable for program analysis in Ocaml (like the CIL 
>> effort
>> at berkeley -- note CIL folks developed their C frontend from scratch 
>> --
>> a waste of effort when they could have leveraged the gcc frontend)
> Ok.
>> Unfortunately there is very little documentation for the gcc front end
>> interface. I have to spend a lot of time code-browsing the gcc code
>> trying to understand the gcc code and data-structures. Since you have
>> done this already, I will really appreciate if you could document your
>> efforts.
> You could look a the llvmgcc source, or you could build on what we 
> already
> have with LLVM.  I'm not sure exactly what you hope to gain with a
> OCaml-like CIL, but I suspect that you can get 95% or more of the 
> benefit
> by writing an OCaml backend for LLVM.  Please read up on LLVM and get a
> clear understanding of what we can and can not do, you might be 
> suprised.
> -Chris
> -- 
> http://llvm.cs.uiuc.edu/
> http://nondot.org/sabre/
> _______________________________________________
> LLVM Developers mailing list
> LLVMdev at cs.uiuc.edu         http://llvm.cs.uiuc.edu
> http://mail.cs.uiuc.edu/mailman/listinfo/llvmdev

More information about the llvm-dev mailing list