The motivation behind this foreign function interface (FFI) specification is to make it possible to describe in Haskell source code the interface to foreign functionality in a Haskell system independent manner. It builds on experiences made with the previous foreign function interfaces provided by GHC and Hugs. However, the FFI specified in this document is not in the market of trying to completely bridge the gap between the actual type of an external function, and what is a convenient type for that function to the Haskell programmer. That is the domain of tools like HaskellDirect or GreenCard, both of which are capable of generating Haskell code that uses this FFI. In the following, we will discuss the language extensions of the FFI. The extensions can be split up into two complementary halves; one half that provides Haskell constructs for importing foreign functionality into Haskell, the other which lets you expose Haskell functions to the outside world. We start with the former, how to import external functionality into Haskell. To bind a Haskell variable name and type to an external function, we introduce a new construct: foreign import. It defines the type of a Haskell function together with the name of an external function that actually implements it. The syntax of foreign import construct is as follows: topdecl : ... .. | 'foreign' 'import' [callconv] [ext_fun] ['unsafe'] varid '::' prim_type A foreign import declaration is only allowed as a toplevel declaration. It consists of two parts, one giving the Haskell type (prim_type), Haskell name (varid) and a flag indicating whether the primitive is unsafe, the other giving details of the name of the external function (ext_fun) and its calling interface (callconv.) Giving a Haskell name and type to an external entry point is clearly an unsafe thing to do, as the external name will in most cases be untyped. The onus is on the programmer using foreign import to ensure that the Haskell type given correctly maps on to the type of the external function. specifies the mapping from Haskell types to external types. The external function has to be given a Haskell name. The name must be a Haskell varid, so the language rules regarding variable names must be followed, i.e., it must start with a lower case letter followed by a sequence of alphanumeric (`in the Unicode sense') characters or '. Notice that with Haskell 98, underscore ('_') is included in the character class small. varid : small ( small | large | udigit | ' )* The name of the external function is a string: ext_fun : string For example, foreign import stdcall "RegCloseKey" regCloseKey :: Ptr a -> IO () states that the external function named RegCloseKey should be bound to the Haskell name regCloseKey. The details of where exactly the external name can be found, such as whether or not it is dynamically linked, and which library it might come from, are implementation dependent. This information is expected to be provided using a compiler-specific method (eg. GHC uses either packages or command-line options to specify libraries and extra include files). If the Haskell name of the imported function is identical to the external name, the ext_fun can be omitted. e.g.: foreign import sin :: Double -> IO Double is identical to foreign import "sin" sin :: Double -> IO Double The number of calling conventions supported is fixed: callconv : ccall | stdcall ccall The 'default' calling convention on a platform, i.e., the one used to do (C) function calls. In the case of x86 platforms, the caller pushes function arguments from right to left on the C stack before calling. The caller is responsible for popping the arguments off of the C stack on return. stdcall A Win32 specific calling convention. The same as ccall, except that the callee cleans up the C stack before returning. The stdcall is a Microsoft Win32 specific wrinkle; it's used throughout the Win32 API, for instance. On platforms where stdcall isn't meaningful, it should be treated as being equal to ccall. Some remarks: Interoperating well with external code is the name of the game here, so the guiding principle when deciding on what calling conventions to include in callconv is that there's a demonstrated need for a particular calling convention. Should it emerge that the inclusion of other calling conventions will generally improve the quality of this Haskell FFI, they will be considered for future inclusion in callconv. Supporting stdcall (and perhaps other platform-specific calling conventions) raises the issue of whether a Haskell FFI should allow the user to write platform-specific Haskell code. The calling convention is clearly an integral part of an external function's interface, so if the one used differs from the standard one specified by the platform's ABI and that convention is used by a non-trivial amount of external functions, the view of the FFI authors is that a Haskell FFI should support it. For foreign import (and other foreign declarations), supplying the calling convention is optional. If it isn't supplied, it is treated as if ccall was specified. Users are encouraged to leave out the specification of the calling convention, if possible. The range of types that can be passed as arguments to an external function is restricted (as are the range of results coming back): prim_type : IO prim_result | prim_result | prim_arg '->' prim_type If you associate a non-IO type with an external function, you have the same 'proof obligations' as when you make use of IOExts.unsafePerformIO in your Haskell programs. The external function is strict in all its arguments. defines prim_result; defines prim_arg. The external function expects zero or more arguments. The set of legal argument types is restricted to the following set: prim_arg : ext_ty | new_ty | ForeignPtr a new_ty : a Haskell newtype of a prim_arg. ext_ty : int_ty | word_ty | float_ty | Ptr a | Char | StablePtr a | Bool int_ty : Int | Int8 | Int16 | Int32 | Int64 word_ty : Word8 | Word16 | Word32 | Word64 float_ty : Float | Double ext_ty represent the set of basic types supported by C-like languages, although the numeric types are explicitly sized. The stable pointer StablePtr type looks out of place in this list of C-like types, but it has a well-defined and simple C mapping, see for details. prim_arg represent the set of permissible argument types. In addition to ext_ty, ForeignPtr is also included. The ForeignPtr type represent values that are pointers to some external entity/object. It differs from the Ptr type in that ForeignPtrs are finalized, i.e., once the garbage collector determines that a ForeignPtr is unreachable, it will invoke a finalising procedure attached to the ForeignPtr to notify the outside world that we're through with using it. Haskell newtypes that wrap up a prim_arg type can also be passed to external functions. Haskell type synonyms for any of the above can also be used in foreign import declarations. Qualified names likewise, i.e. Word.Word32 is legal. foreign import does not support the binding to external constants/variables. A foreign import declaration that takes no arguments represent a binding to a function with no arguments. A GHC extension is the support for unboxed types: prim_arg : ... | unboxed_h_ty ext_ty : .... | unboxed_ext_ty unboxed_ext_ty : Int# | Word# | Char# | Float# | Double# | Addr# | StablePtr# a unboxed_h_ty : MutableByteArray# | ForeignObj# | ByteArray# Clearly, if you want to be portable across Haskell systems, using system-specific extensions such as this is not advisable; avoid using them if you can. (Support for using unboxed types might be withdrawn sometime in the future.) An external function is permitted to return the following range of types: prim_result : ext_ty | new_ext_ty | () new_ext_ty : a Haskell newtype of an ext_ty. where () represents void / no result. External functions cannot raise exceptions (IO exceptions or non-IO ones.) It is the responsibility of the foreign import user to layer any error handling on top of an external function. Only external types (ext_ty) can be passed back, i.e., returning ForeignPtrs is not supported/allowed. Haskell newtypes that wrap up ext_ty are also permitted. For the FFI to be of any practical use, the properties and sizes of the various types that can be communicated between the Haskell world and the outside, needs to be precisely defined. We do this by presenting a mapping to C, as it is commonly used and most other languages define a mapping to it. Table defines the mapping between Haskell and C types. Haskell type C type requirement range (9) Char HsChar unspec. integral type HS_CHAR_MIN .. HS_CHAR_MAX Int HsInt signed integral of unspec. size(4) HS_INT_MIN .. HS_INT_MAX Int8 (2) HsInt8 8 bit signed integral HS_INT8_MIN .. HS_INT8_MAX Int16 (2) HsInt16 16 bit signed integral HS_INT16_MIN .. HS_INT16_MAX Int32 (2) HsInt32 32 bit signed integral HS_INT32_MIN .. HS_INT32_MAX Int64 (2,3) HsInt64 64 bit signed integral (3) HS_INT64_MIN .. HS_INT64_MAX Word8 (2) HsWord8 8 bit unsigned integral 0 .. HS_WORD8_MAX Word16 (2) HsWord16 16 bit unsigned integral 0 .. HS_WORD16_MAX Word32 (2) HsWord32 32 bit unsigned integral 0 .. HS_WORD32_MAX Word64 (2,3) HsWord64 64 bit unsigned integral (3) 0 .. HS_WORD64_MAX Float HsFloat floating point of unspec. size (5) (10) Double HsDouble floating point of unspec. size (5) (10) Bool HsBool unspec. integral type (11) Ptr a HsPtr void* (6) ForeignPtr a HsForeignPtr void* (7) StablePtr a HsStablePtr void* (8) Some remarks: A Haskell system that implements the FFI will supply a header file HsFFI.h that includes target platform specific definitions for the above types and values. The sized numeric types Hs{Int,Word}{8,16,32,64} have a 1-1 mapping to ISO C 99's {,u}int{8,16,32,64}_t. For systems that doesn't support this revision of ISO C, a best-fit mapping onto the supported C types is provided. An implementation which does not support 64 bit integral types on the C side should implement Hs{Int,Word}64 as a struct. In this case the bounds HS_INT64_{MIN,MAX} and HS_WORD64_MAX are undefined. A valid Haskell representation of Int has to be equal to or wider than 30 bits. The HsInt synonym is guaranteed to map onto a C type that satisifies Haskell's requirement for Int. It is guaranteed that Hs{Float,Double} are one of C's floating-point types float/double/long double. It is guaranteed that HsAddr is of the same size as void*, so any other pointer type can be converted to and from HsAddr without any loss of information (K&R, Appendix A6.8). Foreign objects are handled like Ptr by the FFI, so there is again the guarantee that HsForeignPtr is the same as void*. The separate name is meant as a reminder that there is a finalizer attached to the object pointed to. Stable pointers are passed as addresses by the FFI, but this is only because a void* is used as a generic container in most APIs, not because they are real addresses. To make this special case clear, a separate C type is used here. The bounds are preprocessor macros, so they can be used in #if and for array bounds. Floating-point limits are a little bit more complicated, so preprocessor macros mirroring ISO C's float.h are provided: HS_{FLOAT,DOUBLE}_RADIX HS_{FLOAT,DOUBLE}_ROUNDS HS_{FLOAT,DOUBLE}_EPSILON HS_{FLOAT,DOUBLE}_DIG HS_{FLOAT,DOUBLE}_MANT_DIG HS_{FLOAT,DOUBLE}_MIN HS_{FLOAT,DOUBLE}_MIN_EXP HS_{FLOAT,DOUBLE}_MIN_10_EXP HS_{FLOAT,DOUBLE}_MAX HS_{FLOAT,DOUBLE}_MAX_EXP HS_{FLOAT,DOUBLE}_MAX_10_EXP It is guaranteed that Haskell's False/True map to C's 0/1, respectively, and vice versa. The mapping of any other integral value to Bool is left unspecified. To avoid name clashes, identifiers starting with Hs and macros starting with HS_ are reserved for the FFI. GHC only: The GHC specific types ByteArray and MutableByteArray both map to char*. By default, a foreign import function is safe. A safe external function may cause a Haskell garbage collection as a result of being called. This will typically happen when the imported function end up calling Haskell functions that reside in the same 'Haskell world' (i.e., shares the same storage manager heap) -- see for details of how the FFI let's you call Haskell functions from the outside. If the programmer can guarantee that the imported function won't call back into Haskell, the foreign import can be marked as 'unsafe' (see for details of how to do this.) Unsafe calls are cheaper than safe ones, so distinguishing the two classes of external calls may be worth your while if you're extra conscious about performance. A foreign imported function should clearly not need to know that it is being called from Haskell. One consequence of this is that the lifetimes of the arguments that are passed from Haskell must equal that of a normal C call. For instance, for the following decl, foreign import "mumble" mumble :: ForeignPtr a -> IO () f :: Ptr a -> IO () f ptr = do fo <- newForeignObj ptr myFinalizer mumble fo The ForeignPtr must live across the call to mumble even if it is not subsequently used/reachable. Why the insistence on this? Consider what happens if mumble calls a function which calls back into the Haskell world to execute a function, behind our back as it were. This evaluation may possibly cause a garbage collection, with the result that fo may end up being finalised. By guaranteeing that fo will be considered live across the call to mumble, the unfortunate situation where fo is finalised (and hence the reference passed to mumble is suddenly no longer valid) is avoided. A foreign import declaration imports an external function into Haskell. (The name of the external function is statically known, but the loading/linking of it may very well be delayed until run-time.) A foreign import declaration is then (approximately) just a type cast of an external function with a statically known name. An extension of foreign import is the support for dynamic type casts of external names/addresses: topdecl : ... .. | 'foreign' 'import' [callconv] 'dynamic' ['unsafe'] varid :: Addr -> (prim_args -> IO prim_result) i.e., identical to a foreign import declaration, but for the specification of dynamic instead of the name of an external function. The presence of dynamic indicates that when an application of varid is evaluated, the function pointed to by its first argument will be invoked, passing it the rest of varid's arguments. What are the uses of this? Native invocation of COM methods, Or the interfacing to any other software component technologies. Haskell libraries that want to be dressed up as C libs (and hence may have to support C callbacks), Haskell code that need to dynamically load and execute code. So far we've provided the Haskell programmer with ways of importing external functions into the Haskell world. The other half of the FFI coin is how to expose Haskell functionality to the outside world. So, dual to the foreign import declaration is foreign export: topdecl : ... .. | 'foreign' 'export' callconv [ext_name] varid :: prim_type A foreign export declaration tells the compiler to expose a locally defined Haskell function to the outside world, i.e., wrap it up behind a calling interface that's useable from C. It is only permitted at the toplevel, where you have to specify the type at which you want to export the function, along with the calling convention to use. For instance, the following export declaration: foreign export ccall "foo" bar :: Int -> Addr -> IO Double will cause a Haskell system to generate the following C callable function: HsDouble foo(HsInt arg1, HsAddr arg2); When invoked, it will call the Haskell function bar, passing it the two arguments that was passed to foo(). The range of types that can be passed as arguments and results is restricted, since varid has got a prim_type. It is not possible to directly export operator symbols. The type checker will verify that the type given for the foreign export declaration is compatible with the type given to function definition itself. The type in the foreign export may be less general than that of the function itself. For example, this is legal: f :: Num a => a -> a foreign export ccall "fInt" f :: Int -> Int foreign export ccall "fFloat" f :: Float -> Float These declarations export two C-callable procedures fInt and fFloat, both of which are implemented by the (overloaded) Haskell function f. The foreign exported IO action must catch all exceptions, as the FFI does not address how to signal Haskell exceptions to the outside world. The foreign export declaration gives the C programmer access to statically defined Haskell functions. It does not allow you to conveniently expose dynamically-created Haskell function values as C function pointers though. To permit this, the FFI supports dynamic foreign exports: topdecl : ... .. | 'foreign' 'export' [callconv] 'dynamic' varid :: prim_type -> IO Addr A foreign export dynamic declaration declares a C function pointer generator. Given a Haskell function value of some restricted type, the generator wraps it up behind an externally callable interface, returning an Addr to an externally callable (C) function pointer. When that function pointer is eventually called, the corresponding Haskell function value is applied to the function pointer's arguments and evaluated, returning the result (if any) back to the caller. The mapping between the argument to a foreign export dynamic declaration and its corresponding C function pointer type, is as follows: typedef cType[[Res]] (*Varid_FunPtr) (cType[[Ty_1]] ,.., cType[[Ty_n]]); where cType[[]] is the Haskell to C type mapping presented in . To make it all a bit more concrete, here's an example: foreign export dynamic mkCallback :: (Int -> IO Int) -> IO Addr foreign import registerCallback :: Addr -> IO () exportCallback :: (Int -> IO Int) -> IO () exportCallback f = do fx <- mkCallback f registerCallback fx The exportCallback lets you register a Haskell function value as a callback function to some external library. The C type of the callback that the external library expects in registerCallback(), is: An FFI implementation is encouraged to generate the C typedef corresponding to a foreign export dynamic declaration, but isn't required to do so. typedef HsInt (*mkCallback_FunPtr) (HsInt arg1); Creating the view of a Haskell closure as a C function pointer entails registering the Haskell closure as a 'root' with the underlying Haskell storage system, so that it won't be garbage collected. The FFI implementation takes care of this, but when the outside world is through with using a C function pointer generated by a foreign export dynamic declaration, it needs to be explicitly freed. This is done by calling: void freeHaskellFunctionPtr(void *ptr); In the event you need to free these function pointers from within Haskell, a standard 'foreign import'ed binding of the above C entry point is also provided, Foreign.freeHaskellFunctionPtr :: Addr -> IO () The foreign import declaration allows us to invoke an external function by name from within the comforts of the Haskell world, while foreign import dynamic lets us invoke an external function by address. However, there's no way of getting at the code address of some particular external label though, which is at times useful, e.g. for the construction of method tables for, say, Haskell COM components. To support this, the FFI has got foreign labels: foreign label "freeAtLast" addrOf_freeAtLast :: Addr The meaning of this declaration is that addrOf_freeAtLast will now contain the address of the label freeAtLast.