search for: annotatedtypes

A couple people have asked about this, so I wanted to describe my proposals for improving LLVM's garbage collection intrinsics. Currently the llvm.gcroot intrinsic associates a set of GC metadata with a value, specifically a value produced by an alloca instruction. In the proposed scheme, GC metadata would instead be associated with a *type*. The most general approach would be to introduce a

...'d be happy to run with it if it checks out. In fact, I would have eventually ran into exactly the same problem with discriminated unions that you describe; Idris, like any other modern statically-typed functional language, has algebraic datatypes after all. Getting optimization passes to treat AnnotatedTypes the same as their contained type would be necessary for this to pay off completely; can anyone comment on what, if any, difficulties might be involved there? > On Fri, Dec 28, 2012 at 1:09 PM, Benjamin Saunders <ralith at gmail.com> wrote: >> >> I'm working on an LLVM back...

I'm working on an LLVM backend for Idris, a garbage-collected pure functional programming language, and have experienced some frustration that LLVM's GC support, specifically with regard to mapping roots, operates only on allocas. This entails a lot of otherwise unnecessary stack allocation (especially in a pure language, where in-place mutation is rare) and imposes limitations on what

First of all, thanks for looking into this! As you've no doubt discovered, I'm one of the people who has talked a lot about this issue in the past, and have been frustrated with the lack of progress in this area. I completely agree with your point about wanting to be able to attach GC metadata to a type (rather than attaching it to a value, as is done now). In the past, there have been