1 \chapter[chap:context]{Context}
2 An obvious question that arises when starting any research is \quote{Hasn't
3 this been done before?} Using a functional language for describing hardware
4 is not a new idea at all. In fact, there has been research into functional
5 hardware description even before the conventional hardware description
6 languages were created. However, functional languages were not nearly as
7 advanced as they are now, and functional hardware description never really
10 Recently, there have been some renewed efforts, especially using the Haskell
11 language. Examples are Lava, ForSyde, ..., which are all a form of an
12 embedded domain specific language. Each of these have a slightly different
13 approach, but all of these do some trickery inside the Haskell language
14 itself, meaning you write a program that generates a hardware circuit,
15 instead of describing the circuit directly (either by running the haskell
16 code after compilation, or using Template Haskell to inspect parts of the
17 code you have written). This allows the full power of Haskell for generating
18 a circuit, but only it also creates severe limitations in the use of the
19 language (you can't use case statements in Lava, since they would be
20 executed only once during circuit generation) and extra notational overhead.
25 \section{Conventional hardware description languages}
26 Considering that we already have some hardware description language like
27 \small{VHDL} and Verilog, why would we need anything else? By introducing
28 the functional style to hardware description, we hope to obtain a hardware
29 description language that is:
31 \item More consise. Functional programs are known for their conciseness,
32 mostly caused by the ability to abstract just about any behaviour into a
33 helper function. This is largely enabled by features like an advanced
34 type system with polymorphism and higher order functions.
35 \item Type-safer. Functional programs typically have a highly expressive
36 type system, which makes it harder to write incorrect code. This is
37 probably not only directly caused by the type system, so perhaps this
38 advantage does not apply in hardware descriptions.
39 \item Easy to process. Functional languages have nice properties like
40 purity \refdef{purity} and single binding behaviour, which make it easy
41 to apply program transformations and optimizations and could potentially
42 simplify program verification.
45 \section{Existing functional hardware description languages}
46 As noted above, we're not the first to walk this path. However, current
47 embedded functional hardware description languages (in particular those
48 using Haskell) are limited by:
50 \item Not all of Haskell's constructs can be captured by embedded domain
51 specific languages. For example, an if or case expression is typically
52 executed only once and only its result is reflected in the embedded
53 description, not the if or case expression itself. Also, sharing and
54 loops are non-trivial do properly and safely translate (though there is
55 some work to fix this, but that has not been possible in a completely
56 reliable way yet. TODO: ref
57 http://www.ittc.ku.edu/~andygill/paper.php?label=DSLExtract09).
58 \item Some things are verbose to express. Especially ForSyDe suffers
59 from a lot of notational overhead due to the Template Haskell approach
60 used. Since conditional statements are not supported, a lot of Haskell's
61 syntax sugar (if expressions, pattern matching, guards) cannot be used
62 either, leading to more verbose notation as well.
63 \item Polymorphism and higher order values are not supported within the
64 embedded language. The use of Haskell as a host language allows the use
65 of polymorphism and higher order functions at circuit generation time
66 (even for free, without any additional cost on the \small{EDSL}
67 programmers), but the described circuits do not have any polymorphism
68 or higher order functions, which can be limiting (TODO: How true or
69 appropriate is this point?).