From: Matthijs Kooijman Date: Wed, 20 Jan 2010 17:26:29 +0000 (+0100) Subject: Add some initial content about application, choice and types. X-Git-Url: https://git.stderr.nl/gitweb?a=commitdiff_plain;h=19da243e9fb1c4cc31f6700984cdf12d72e50665;p=matthijs%2Fmaster-project%2Fdsd-paper.git Add some initial content about application, choice and types. All of this is copied directly from my thesis, examples and other stuff has been left out and it still needs cutting down. Layout is still broken in some places. --- diff --git "a/c\316\273ash.tex" "b/c\316\273ash.tex" index 9592c5e..80eeb81 100644 --- "a/c\316\273ash.tex" +++ "b/c\316\273ash.tex" @@ -452,7 +452,259 @@ foo\par bar % Won't compile without at least two paragraphs. \section{Hardware description in Haskell} -foo\par bar + To translate Haskell to hardware, every Haskell construct needs a + translation to \VHDL. There are often multiple valid translations + possible. When faced with choices, the most obvious choice has been + chosen wherever possible. In a lot of cases, when a programmer looks + at a functional hardware description it is completely clear what + hardware is described. We want our translator to generate exactly that + hardware whenever possible, to make working with Cλash as intuitive as + possible. + + \subsection{Function application} + The basic syntactic elements of a functional program are functions + and function application. These have a single obvious \VHDL\ + translation: each top level function becomes a hardware component, + where each argument is an input port and the result value is the + (single) output port. This output port can have a complex type (such + as a tuple), so having just a single output port does not pose a + limitation. + + Each function application in turn becomes component instantiation. + Here, the result of each argument expression is assigned to a + signal, which is mapped to the corresponding input port. The output + port of the function is also mapped to a signal, which is used as + the result of the application. + + Since every top level function generates its own component, the + hierarchy of of function calls is reflected in the final \VHDL\ + output as well, creating a hierarchical \VHDL\ description of the + hardware. This separation in different components makes the + resulting \VHDL\ output easier to read and debug. + + \subsection{Choice} + Although describing components and connections allows us to describe + a lot of hardware designs already, there is an obvious thing + missing: choice. We need some way to be able to choose between + values based on another value. In Haskell, choice is achieved by + \hs{case} expressions, \hs{if} expressions, pattern matching and + guards. + + However, to be able to describe our hardware in a more convenient + way, we also want to translate Haskell's choice mechanisms. The + easiest of these are of course case expressions (and \hs{if} + expressions, which can be very directly translated to \hs{case} + expressions). A \hs{case} expression can in turn simply be + translated to a conditional assignment, where the conditions use + equality comparisons against the constructors in the \hs{case} + expressions. + + A slightly more complex (but very powerful) form of choice is + pattern matching. A function can be defined in multiple clauses, + where each clause specifies a pattern. When the arguments match the + pattern, the corresponding clause will be used. + + \subsection{Types} + Translation of two most basic functional concepts has been + discussed: function application and choice. Before looking further + into less obvious concepts like higher-order expressions and + polymorphism, the possible types that can be used in hardware + descriptions will be discussed. + + Some way is needed to translate every values used to its hardware + equivalents. In particular, this means a hardware equivalent for + every \emph{type} used in a hardware description is needed + + Since most functional languages have a lot of standard types that + are hard to translate (integers without a fixed size, lists without + a static length, etc.), a number of \quote{built-in} types will be + defined first. These types are built-in in the sense that our + compiler will have a fixed VHDL type for these. User defined types, + on the other hand, will have their hardware type derived directly + from their Haskell declaration automatically, according to the rules + sketched here. + + \subsection{Built-in types} + The language currently supports the following built-in types. Of these, + only the \hs{Bool} type is supported by Haskell out of the box (the + others are defined by the Cλash package, so they are user-defined types + from Haskell's point of view). + + \begin{description} + \item[\hs{Bit}] + This is the most basic type available. It is mapped directly onto + the \texttt{std\_logic} \VHDL\ type. Mapping this to the + \texttt{bit} type might make more sense (since the Haskell version + only has two values), but using \texttt{std\_logic} is more standard + (and allowed for some experimentation with don't care values) + + \item[\hs{Bool}] + This is the only built-in Haskell type supported and is translated + exactly like the Bit type (where a value of \hs{True} corresponds to a + value of \hs{High}). Supporting the Bool type is particularly + useful to support \hs{if ... then ... else ...} expressions, which + always have a \hs{Bool} value for the condition. + + A \hs{Bool} is translated to a \texttt{std\_logic}, just like \hs{Bit}. + \item[\hs{SizedWord}, \hs{SizedInt}] + These are types to represent integers. A \hs{SizedWord} is unsigned, + while a \hs{SizedInt} is signed. These types are parametrized by a + length type, so you can define an unsigned word of 32 bits wide as + ollows: + + \begin{verbatim} + type Word32 = SizedWord D32 + \end{verbatim} + + Here, a type synonym \hs{Word32} is defined that is equal to the + \hs{SizedWord} type constructor applied to the type \hs{D32}. \hs{D32} + is the \emph{type level representation} of the decimal number 32, + making the \hs{Word32} type a 32-bit unsigned word. + + These types are translated to the \small{VHDL} \texttt{unsigned} and + \texttt{signed} respectively. + \item[\hs{Vector}] + This is a vector type, that can contain elements of any other type and + has a fixed length. It has two type parameters: its + length and the type of the elements contained in it. By putting the + length parameter in the type, the length of a vector can be determined + at compile time, instead of only at run-time for conventional lists. + + The \hs{Vector} type constructor takes two type arguments: the length + of the vector and the type of the elements contained in it. The state + type of an 8 element register bank would then for example be: + + \begin{verbatim} + type RegisterState = Vector D8 Word32 + \end{verbatim} + + Here, a type synonym \hs{RegisterState} is defined that is equal to + the \hs{Vector} type constructor applied to the types \hs{D8} (The type + level representation of the decimal number 8) and \hs{Word32} (The 32 + bit word type as defined above). In other words, the + \hs{RegisterState} type is a vector of 8 32-bit words. + + A fixed size vector is translated to a \VHDL\ array type. + \item[\hs{RangedWord}] + This is another type to describe integers, but unlike the previous + two it has no specific bit-width, but an upper bound. This means that + its range is not limited to powers of two, but can be any number. + A \hs{RangedWord} only has an upper bound, its lower bound is + implicitly zero. There is a lot of added implementation complexity + when adding a lower bound and having just an upper bound was enough + for the primary purpose of this type: type-safely indexing vectors. + + To define an index for the 8 element vector above, we would do: + + \begin{verbatim} + type RegisterIndex = RangedWord D7 + \end{verbatim} + + Here, a type synonym \hs{RegisterIndex} is defined that is equal to + the \hs{RangedWord} type constructor applied to the type \hs{D7}. In + other words, this defines an unsigned word with values from + 0 to 7 (inclusive). This word can be be used to index the + 8 element vector \hs{RegisterState} above. + + This type is translated to the \texttt{unsigned} \VHDL type. + \end{description} + \subsection{User-defined types} + There are three ways to define new types in Haskell: algebraic + data-types with the \hs{data} keyword, type synonyms with the \hs{type} + keyword and type renamings with the \hs{newtype} keyword. \GHC\ + offers a few more advanced ways to introduce types (type families, + existential typing, \small{GADT}s, etc.) which are not standard + Haskell. These will be left outside the scope of this research. + + Only an algebraic datatype declaration actually introduces a + completely new type, for which we provide the \VHDL\ translation + below. Type synonyms and renamings only define new names for + existing types (where synonyms are completely interchangeable and + renamings need explicit conversion). Therefore, these do not need + any particular \VHDL\ translation, a synonym or renamed type will + just use the same representation as the original type. The + distinction between a renaming and a synonym does no longer matter + in hardware and can be disregarded in the generated \VHDL. + + For algebraic types, we can make the following distinction: + + \begin{description} + + \item[Product types] + A product type is an algebraic datatype with a single constructor with + two or more fields, denoted in practice like (a,b), (a,b,c), etc. This + is essentially a way to pack a few values together in a record-like + structure. In fact, the built-in tuple types are just algebraic product + types (and are thus supported in exactly the same way). + + The ``product'' in its name refers to the collection of values belonging + to this type. The collection for a product type is the Cartesian + product of the collections for the types of its fields. + + These types are translated to \VHDL\ record types, with one field for + every field in the constructor. This translation applies to all single + constructor algebraic data-types, including those with just one + field (which are technically not a product, but generate a VHDL + record for implementation simplicity). + \item[Enumerated types] + An enumerated type is an algebraic datatype with multiple constructors, but + none of them have fields. This is essentially a way to get an + enumeration-like type containing alternatives. + + Note that Haskell's \hs{Bool} type is also defined as an + enumeration type, but we have a fixed translation for that. + + These types are translated to \VHDL\ enumerations, with one value for + each constructor. This allows references to these constructors to be + translated to the corresponding enumeration value. + \item[Sum types] + A sum type is an algebraic datatype with multiple constructors, where + the constructors have one or more fields. Technically, a type with + more than one field per constructor is a sum of products type, but + for our purposes this distinction does not really make a + difference, so this distinction is note made. + + The ``sum'' in its name refers again to the collection of values + belonging to this type. The collection for a sum type is the + union of the the collections for each of the constructors. + + Sum types are currently not supported by the prototype, since there is + no obvious \VHDL\ alternative. They can easily be emulated, however, as + we will see from an example: + + \begin{verbatim} + data Sum = A Bit Word | B Word + \end{verbatim} + + An obvious way to translate this would be to create an enumeration to + distinguish the constructors and then create a big record that + contains all the fields of all the constructors. This is the same + translation that would result from the following enumeration and + product type (using a tuple for clarity): + + \begin{verbatim} + data SumC = A | B + type Sum = (SumC, Bit, Word, Word) + \end{verbatim} + + Here, the \hs{SumC} type effectively signals which of the latter three + fields of the \hs{Sum} type are valid (the first two if \hs{A}, the + last one if \hs{B}), all the other ones have no useful value. + + An obvious problem with this naive approach is the space usage: the + example above generates a fairly big \VHDL\ type. Since we can be + sure that the two \hs{Word}s in the \hs{Sum} type will never be valid + at the same time, this is a waste of space. + + Obviously, duplication detection could be used to reuse a + particular field for another constructor, but this would only + partially solve the problem. If two fields would be, for + example, an array of 8 bits and an 8 bit unsigned word, these are + different types and could not be shared. However, in the final + hardware, both of these types would simply be 8 bit connections, + so we have a 100\% size increase by not sharing these. + \end{description} + \section{Cλash prototype}