c-lambda
c-declare
c-include
Bibliography
Version: 5.0

c-lambda: C FFI via raco ctool

The compiler/cffi module relies on a C compiler to statically construct an interface to C code through directives embedded in a Racket program. The library implements a subset of Gambit-C’s foreign-function interface [Feeley98].

The ffi/unsafe library is a better interface for most tasks; see FFI: Racket Foreign Interface for more information on ffi/unsafe. See also Inside: Racket C API, which describes Racket’s C-level API for extending the run-time system.

The compiler/cffi library defines three forms: c-lambda, c-declare, and c-include. When interpreted directly or compiled to byte code, c-lambda produces a function that always raises exn:fail, and c-declare and c-include raise exn:fail. When compiled by raco ctool --extension, the forms provide access to C. Thus, compiler/cffi is normally required by a module to be compiled via raco ctool. In addition, the raco ctool compiler implicitly imports compiler/cffi into the top-level environment for non-module compilation.

The c-lambda form creates a Racket procedure whose body is implemented in C. Instead of declaring argument names, a c-lambda form declares argument types, as well as a return type. The implementation can be simply the name of a C function, as in the following definition of fmod:

  (define fmod (c-lambda (double double) double "fmod"))

Alternatively, the implementation can be C code to serve as the body of a function, where the arguments are bound to ___arg1 (three underscores), etc., and the result is installed into ___result (three underscores):

  (define machine-string->float
    (c-lambda (char-string) float
       "___result = *(float *)___arg1;"))

The c-lambda form provides only limited conversions between C and Racket data. For example, the following function does not reliably produce a string of four characters:

  (define broken-machine-float->string
    (c-lambda (float) char-string
       "char b[5]; *(float *)b = ___arg1; b[4] = 0; ___result = b;"))

because the representation of a float can contain null bytes, which terminate the string. However, the full Racket API, which is described in Inside: Racket C API, can be used in a function body:

  (define machine-float->string
    (c-lambda (float) racket-object
       "char b[4];"
       "*(float *)b = ___arg1;"
       "___result = racket_make_sized_byte_string(b, 4, 1);"))

The c-declare form declares arbitrary C code to appear after "escheme.h" or "scheme.h" is included, but before any other code in the compilation environment of the declaration. It is often used to declare C header file inclusions. For example, a proper definition of fmod needs the "math.h" header file:

  (c-declare "#include <math.h>")
  (define fmod (c-lambda (double double) double "fmod"))

The c-declare form can also be used to define helper C functions to be called through c-lambda.

The c-include form expands to a c-declare form using the content of a specified file. Use (c-include file) instead of (c-declare "#include file") when it’s easier to have Racket resolve the file path than to have the C compiler resolve it.

The "collects/mzscheme/examples" directory in the Racket distribution contains additional examples.

When compiling for Racket 3m (see Inside: Racket C API), C code inserted by c-lambda, c-declare, and c-include will be transformed in the same was as raco ctool’s --xform mode (which may or may not be enough to make the code work correctly in Racket 3m; see Inside: Racket C API for more information).

(c-lambda (argument-type ...) return-type impl-string ...+)
Creates a Racket procedure whose body is implemented in C. The procedure takes as many arguments as the supplied argument-types, and it returns one value. If return-type is void, the procedure’s result is always void. The impl-string is either the name of a C function (or macro) or the body of a C function.

If a single impl-string is provided, and if it is a string containing only alphanumeric characters and _, then the created Racket procedure passes all of its arguments to the named C function (or macro) and returns the function’s result. Each argument to the Racket procedure is converted according to the corresponding argument-type (as described below) to produce an argument to the C function. Unless return-type is void, the C function’s result is converted according to return-type for the Racket procedure’s result.

If more than impl-string is provided, or if it contains more than alphanumeric characters and _, then the concatenated impl-strings must contain C code to implement the function body. The converted arguments for the function will be in variables ___arg1, ___arg2, ... (with three underscores in each name) in the context where the impl-strings are placed for compilation. Unless return-type is void, the impl-strings code should assign a result to the variable ___result (three underscores), which will be declared but not initialized. The impl-strings code should not return explicitly; control should always reach the end of the body. If the impl-strings code defines the pre-processor macro ___AT_END (with three leading underscores), then the macro’s value should be C code to execute after the value ___result is converted to a Racket result, but before the result is returned, all in the same block; defining ___AT_END is primarily useful for deallocating a string in ___result that has been copied by conversion. The impl-strings code will start on a new line at the beginning of a block in its compilation context, and ___AT_END will be undefined after the code.

In addition to ___arg1, etc., the variable argc is bound in impl-strings to the number of arguments supplied to the function, and argv is bound to a Racket_Object* array of length argc containing the function arguments as Racket values. The argv and argc variables are mainly useful for error reporting (e.g., with racket_wrong_type).

Each argument-type must be one of the following, which are recognized symbolically:

The return-type must be void or one of the arg-type keywords.

(c-declare code-string)
Declares arbitrary C code to appear after "escheme.h" or "scheme.h" is included, but before any other code in the compilation environment of the declaration. A c-declare form can appear only at the top-level or within a module’s top-level sequence.

The code code will appear on a new line in the file for C compilation. Multiple c-include declarations are concatenated (with newlines) in order to produces a sequence of declarations.

(c-include path-spec)
Expands to a use of c-declare with the content of path-spec. The path-spec has the same form as for mzlib/include’s include.

Bibliography

[Feeley98] Marc Feeley, “Gambit-C, version 3.0.” 1998.