sbcl.sbcl/doc/ffi.xml
Kevin Rosenberg 3c5c2c2656 0.8.6.1:
* clean.sh: Allow recursive includes in src/runtime/Config
	* doc/*: Convert from SGML format and processing to XML
	Currently, catalogs are in place for Debian and RedHat which
	allow automated use of local files instead of the tools
	downloading DTDs and XSLs from their canonical URLs.
	I'm glad to help setup catalogs for other operating systems.
2003-11-26 02:11:09 +00:00

1256 lines
49 KiB
XML

<?xml version="1.0" encoding="iso-8859-1"?>
<!DOCTYPE book PUBLIC "-//OASIS//DTD DocBook XML V4.2//EN"
"http://www.oasis-open.org/docbook/xml/4.2/docbookx.dtd" [
<!ENTITY % myents SYSTEM "entities.inc">
%myents;
]>
<chapter id="ffi"><title>The Foreign Function Interface</title>
<para>This chapter describes &SBCL;'s interface to C programs and
libraries (and, since C interfaces are a sort of <foreignphrase>lingua
franca</foreignphrase> of the Unix world, to other programs and libraries in
general.)</para>
<note><para>In the modern Lisp world, the usual term for this
functionality is Foreign Function Interface, or <acronym>FFI</acronym>, where
despite the mention of <quote>function</quote> in this term, <acronym>FFI</acronym> also
refers to direct manipulation of C data structures as well as
functions. The traditional &CMUCL; terminology is Alien Interface, and
while that older terminology is no longer used much in the system
documentation, it still reflected in names in the
implementation, notably in the name of the <literal>SB-ALIEN</literal>
package.</para></note>
<sect1 id="ffi-intro"><title>Introduction to the Foreign Function Interface</title>
<!-- AKA "Introduction to Aliens" in the CMU CL manual -->
<para>
Because of Lisp's emphasis on dynamic memory allocation and garbage
collection, Lisp implementations use non-C-like memory representations
for objects. This representation mismatch creates friction when a Lisp
program must share objects with programs which expect C data. There
are three common approaches to establishing communication:
<itemizedlist>
<listitem><para>The burden can be placed on the foreign program
(and programmer) by requiring the knowledge and use of the
representations used internally by the Lisp implementation.
This can require a considerable amount of <quote>glue</quote> code on the
C side, and that code tends to be sensitively dependent on the
internal implementation details of the Lisp system.</para></listitem>
<listitem><para>The Lisp system can automatically convert objects
back and forth between the Lisp and foreign representations.
This is convenient, but translation becomes prohibitively slow
when large or complex data structures must be shared. This approach
is supported by the &SBCL; <acronym>FFI</acronym>, and used automatically
by the when passing integers and strings.</para></listitem>
<listitem><para>The Lisp program can directly manipulate foreign
objects through the use of extensions to the Lisp language.
</para></listitem>
</itemizedlist>
</para>
<para>&SBCL;, like &CMUCL; before it, relies primarily on the
automatic conversion and direct manipulation approaches. The SB-ALIEN
package provices a facility wherein foreign values of simple scalar
types are automatically converted and complex types are directly
manipulated in their foreign representation. Additionally the
lower-level System Area Pointers (or SAPs) can be used where
necessary to provide untyped access to foreign memory.</para>
<para>Any foreign objects that can't automatically be converted into
Lisp values are represented by objects of type <type>alien-value</type>.
Since Lisp is a dynamically typed language, even foreign objects must
have a run-time type; this type information is provided by
encapsulating the raw pointer to the foreign data within an
<type>alien-value</type> object.</para>
<para>The type language and operations on foreign types are
intentionally similar to those of the C language.</para>
</sect1>
<sect1 id="foreign-types"><title>Foreign Types</title>
<!-- AKA "Alien Types" in the CMU CL manual -->
<para>Alien types have a description language based on nested list
structure. For example the C type
<programlisting>struct foo {
int a;
struct foo *b[100];
};</programlisting>
has the corresponding &SBCL; FFI type
<programlisting>(struct foo
(a int)
(b (array (* (struct foo)) 100)))</programlisting>
</para>
<sect2><title>Defining Foreign Types</title>
<para>
Types may be either named or anonymous. With structure and union
types, the name is part of the type specifier, allowing recursively
defined types such as:
<programlisting>(struct foo (a (* (struct foo))))</programlisting>
An anonymous structure or union type is specified by using the name
<literal>nil</literal>. The <function>with-alien</function> macro defines a local
scope which <quote>captures</quote> any named type definitions. Other types
are not inherently named, but can be given named abbreviations using
the <function>define-alien-type</function> macro.
</para>
</sect2>
<sect2><title>Foreign Types and Lisp Types</title>
<para>
The foreign types form a subsystem of the &SBCL; type system. An
<type>alien</type> type specifier provides a way to use any foreign type as a
Lisp type specifier. For example,
<programlisting>(typep foo '(alien (* int)))</programlisting>
can be used to determine whether <varname>foo</varname> is a pointer to a foreign
<type>int</type>. <type>alien</type> type specifiers can be used in the same ways
as ordinary Lisp type specifiers (like <type>string</type>.) Alien type
declarations are subject to the same
precise type checking <!-- FIXME: should be linked to id="precisetypechecking" -->
as any other declaration.
</para>
<para>
Note that the type identifiers used in the
foreign type system overlap with native Lisp type
specifiers in some cases. For example, the type specifier
<type>(alien single-float)</type> is identical to <type>single-float</type>, since
foreign floats are automatically converted to Lisp floats. When
<function>type-of</function> is called on an alien value that is not automatically
converted to a Lisp value, then it will return an <type>alien</type> type
specifier.
</para>
</sect2>
<sect2><title>Foreign Type Specifiers</title>
<note><para>
All foreign type names are exported from the <literal>sb-alien</literal>
package. Some foreign type names are also symbols in
the <literal>common-lisp</literal> package, in which case they are
reexported from the <literal>sb-alien</literal> package, so that
e.g. it is legal to refer to <type>sb-alien:single-float</type>.
</para></note>
<para>
These are the basic foreign type specifiers:
<!-- FIXME: There must be some better way of formatting definitions
in DocBook than this. I haven't found it yet, but suggestions
or patches would be welcome. -->
<itemizedlist>
<listitem>
<para>
The foreign type specifier <type>(* foo)</type> describes a
pointer to an object of type <type>foo</type>. A pointed-to type
<type>foo</type> of <type>t</type> indicates a pointer to anything,
similar to <type>void *</type> in ANSI C. A null alien pointer can
be detected with the <function>sb-alien:null-alien</function>
function.
</para>
</listitem>
<listitem>
<para>
The foreign type specifier <type>(array foo &amp;optional dimensions)</type>
describes array of the specified <literal>dimensions</literal>, holding
elements of type <type>foo</type>. Note that (unlike in C) <type>(* foo)</type>
<type>(array foo)}</type> are considered to be different types when
type checking is done. If equivalence of pointer and array types
is desired, it may be explicitly coerced using
<function>sb-alien:cast</function>.
</para>
<para>
Arrays are accessed using <function>sb-alien:deref</function>, passing
the indices as additional arguments. Elements are stored in
column-major order (as in C), so the first dimension determines
only the size of the memory block, and not the layout of the
higher dimensions. An array whose first dimension is variable
may be specified by using <literal>nil</literal> as the first dimension.
Fixed-size arrays can be allocated as array elements, structure
slots or <function>sb-alien:with-alien</function> variables. Dynamic
arrays can only be allocated using <function>sb-alien:make-alien</function>.
</para>
</listitem>
<listitem>
<para>
The foreign type specifier
<type>(sb-alien:struct name &amp;rest fields)</type>
describes a structure type with the specified <varname>name</varname> and
<varname>fields</varname>. Fields are allocated at the same offsets
used by the implementation's C compiler. If <varname>name</varname>
is <literal>nil</literal> then the structure is anonymous.
</para>
<para>
If a named foreign <type>struct</type> specifier is passed to
<function>define-alien-type</function> or <function>with-alien</function>,
then this defines, respectively, a new global or local foreign
structure type. If no <varname>fields</varname> are specified, then
the fields are taken from the current (local or global) alien
structure type definition of <varname>name</varname>.
</para>
</listitem>
<listitem>
<para>
The foreign type specifier
<type>(sb-alien:union name &amp;rest fields)</type>
is similar to <type>sb-alien:struct</type>, but describes a union type.
All fields are allocated at the same offset, and the size of the
union is the size of the largest field. The programmer must
determine which field is active from context.
</para>
</listitem>
<listitem>
<para>
The foreign type specifier <type>(sb-alien:enum name &amp;rest specs)</type>
describes an enumeration type that maps between integer values
and keywords. If <varname>name</varname> is <literal>nil</literal>, then the
type is anonymous. Each element of the <varname>specs</varname>
list is either a Lisp keyword, or a list <literal>(keyword value)</literal>.
<varname>value</varname> is an integer. If <varname>value</varname> is not
supplied, then it defaults to one greater than the value for
the preceding spec (or to zero if it is the first spec.)
</para>
</listitem>
<listitem>
<para>
The foreign type specifier <type>(sb-alien:signed &amp;optional bits)</type>
specifies a signed integer with the specified number of
<varname>bits</varname> precision. The upper limit on integer
precision is determined by the machine's word
size. If <varname>bits</varname> is not specified, the maximum
size will be used.
</para>
</listitem>
<listitem>
<para>
The foreign type specifier <type>(integer &amp;optional bits)</type> is
equivalent to the corresponding type specifier using
<type>sb-alien:signed</type> instead of <type>integer</type>.
</para>
</listitem>
<listitem>
<para>
The foreign type specifier
<type>(sb-alien:unsigned &amp;optional bits)</type>
is like corresponding type specifier using <type>sb-alien:signed</type>
except that the variable is treated as an unsigned integer.
</para>
</listitem>
<listitem>
<para>
The foreign type specifier <type>(boolean &amp;optional bits)</type> is
similar to an enumeration type, but maps from Lisp <literal>nil</literal>
and <literal>t</literal> to C <literal>0</literal> and <literal>1</literal>
respectively. <varname>bits</varname> determines the amount of
storage allocated to hold the truth value.
</para>
</listitem>
<listitem>
<para>
The foreign type specifier <type>single-float</type> describes a
floating-point number in IEEE single-precision format.
</para>
</listitem>
<listitem>
<para>
The foreign type specifier <type>double-float</type> describes a
floating-point number in IEEE double-precision format.
</para>
</listitem>
<listitem>
<para>
The foreign type specifier
<type>(function result-type &amp;rest arg-types)</type>
describes a foreign function that takes arguments of the specified
<varname>arg-types</varname> and returns a result of type <type>result-type</type>.
Note that the only context where a foreign <type>function</type> type
is directly specified is in the argument to
<function>sb-alien:alien-funcall</function>.
In all other contexts, foreign functions are represented by
foreign function pointer types: <type>(* (function ...))</type>.
</para>
</listitem>
<listitem>
<para>
The foreign type specifier <type>sb-alien:system-area-pointer</type>
describes a pointer which is represented in Lisp as a
<type>system-area-pointer</type> object. &SBCL; exports this type from
<literal>sb-alien</literal> because &CMUCL; did, but tentatively (as of
the first draft of this section of the manual, &SBCL; 0.7.6) it is
deprecated, since it doesn't seem to be required by user code.
</para>
</listitem>
<listitem>
<para>
The foreign type specifier <type>sb-alien:void</type> is
used in function types to declare that no useful value
is returned. Using <function>alien-funcall</function>
to call a <type>void</type> foreign function will return
zero values.
</para>
</listitem>
<listitem>
<para>
The foreign type specifier <type>sb-alien:c-string</type>
is similar to <type>(* char)</type>, but is interpreted as a
null-terminated string, and is automatically converted into a
Lisp string when accessed; or if the pointer is C <literal>NULL</literal>
or <literal>0</literal>, then accessing it gives Lisp <literal>nil</literal>.
Lisp strings are stored with a trailing NUL termination, so no
copying (either by the user or the implementation) is necessary
when passing them to foreign code.
</para>
<para>
Assigning a Lisp string to a <type>c-string</type> structure field or
variable stores the contents of the string to the memory already
pointed to by that variable. When a foreign object of type
<type>(* char)</type> is assigned to a <type>c-string</type>, then the
<type>c-string</type> pointer is assigned to. This allows
<type>c-string</type> pointers to be initialized. For example:
<programlisting>(cl:in-package "CL-USER") ; which USEs package "SB-ALIEN"
(define-alien-type nil (struct foo (str c-string)))
(defun make-foo (str) (let ((my-foo (make-alien (struct foo))))
(setf (slot my-foo 'str) (make-alien char (length str))
(slot my-foo 'str) str) my-foo))</programlisting>
Storing Lisp <literal>NIL</literal> in a <type>c-string</type> writes C
<literal>NULL</literal> to the variable.
</para>
</listitem>
<listitem>
<para>
<literal>sb-alien</literal> also exports translations of these C type
specifiers as foreign type specifiers:
<type>sb-alien:char</type>,
<type>sb-alien:short</type>,
<type>sb-alien:int</type>,
<type>sb-alien:long</type>,
<type>sb-alien:unsigned-char</type>,
<type>sb-alien:unsigned-short</type>,
<type>sb-alien:unsigned-int</type>,
<type>sb-alien:unsigned-long</type>,
<type>sb-alien:float</type>, and
<type>sb-alien:double</type>.
</para>
</listitem>
</itemizedlist>
</para>
</sect2>
</sect1>
<sect1 id="foreign-values"><title>Operations On Foreign Values</title>
<!-- AKA "Alien Operations" in the CMU CL manual -->
<para>This section describes how to read foreign values as Lisp
values, how to coerce foreign values to different kinds of foreign values, and
how to dynamically allocate and free foreign variables.</para>
<sect2><title>Accessing Foreign Values</title>
<synopsis>(sb-alien:deref pointer-or-array &amp;rest indices)</synopsis>
<para>The <function>sb-alien:deref</function> function returns the value pointed to by
a foreign pointer, or the value of a foreign array element. When
dereferencing a pointer, an optional single index can be specified to
give the equivalent of C pointer arithmetic; this index is scaled by
the size of the type pointed to. When dereferencing an array, the
number of indices must be the same as the number of dimensions in the
array type. <function>deref</function> can be set with <function>setf</function> to
assign a new value.</para>
<synopsis>(sb-alien:slot struct-or-union &amp;rest slot-names)</synopsis>
<para>The <function>sb-alien:slot</function> function extracts the value of
the slot named <varname>slot-name</varname> from a foreign <type>struct</type> or
<type>union</type>. If <varname>struct-or-union</varname> is a pointer to a
structure or union, then it is automatically dereferenced.
<function>sb-alien:slot</function> can be set with <function>setf</function> to assign
a new value. Note that <varname>slot-name</varname> is evaluated, and need
not be a compile-time constant (but only constant slot accesses are
efficiently compiled.)</para>
<sect3><title>Untyped memory</title>
<para>As noted at the beginning of the chapter, the System Area
Pointer facilities allow untyped access to foreign memory. SAPs can
be converted to and from the usual typed foreign values using
<function>sap-alien</function> and <function>alien-sap</function>
(described elsewhere), and also to and from integers - raw machine
addresses. They should thus be used with caution; corrupting the Lisp
heap or other memory with SAPs is trivial.</para>
<synopsis>(sb-sys:int-sap machine-address)</synopsis>
<para>Creates a SAP pointing at the virtual address
<varname>machine-address</varname>. </para>
<synopsis>(sb-sys:sap-ref-32 sap offset)</synopsis>
<para>Access the value of the memory location at
<varname>offset</varname> bytes from <varname>sap</varname>. This form
may also be used with <function>setf</function> to alter the memory at
that location.</para>
<synopsis>(sb-sys:sap= sap1 sap2)</synopsis>
<para>Compare <varname>sap1</varname> and <varname>sap2</varname> for
equality.</para>
<para>Similarly named functions exist for accessing other sizes of
word, other comparisons, and other conversions. The reader is invited
to use <function>apropos</function> and <function>describe</function>
for more details</para>
<programlisting>
(apropos "sap" :sb-sys)
</programlisting>
</sect3></sect2>
<sect2><title>Coercing Foreign Values</title>
<synopsis>(sb-alien:addr alien-expr)</synopsis>
<para>
The <function>sb-alien:addr</function> macro
returns a pointer to the location specified by
<varname>alien-expr</varname>, which must be either a foreign variable, a use of
<function>sb-alien:deref</function>, a use of <function>sb-alien:slot</function>, or a use of
<function>sb-alien:extern-alien</function>.
</para>
<synopsis>(sb-alien:cast foreign-value new-type)</synopsis>
<para>The <function>sb-alien:cast</function>
converts <varname>foreign-value</varname> to a new foreign value with the specified
<varname>new-type</varname>. Both types, old and new, must be foreign pointer,
array or function types. Note that the resulting Lisp
foreign variable object
is not <function>eq</function> to the
argument, but it does refer to the same foreign data bits.</para>
<synopsis>(sb-alien:sap-alien sap type)</synopsis>
<para>The <function>sb-alien:sap-alien</function> function converts <varname>sap</varname>
(a system area pointer) to a foreign value with the specified
<varname>type</varname>. <varname>type</varname> is not evaluated.
</para>
<para>The <varname>type</varname> must be some foreign pointer, array, or
record type.</para>
<synopsis>(sb-alien:alien-sap foreign-value type)</synopsis>
<para>The <function>sb-alien:alien-sap</function> function
returns the SAP which points to <varname>alien-value</varname>'s data.
</para>
<para>The <varname>foreign-value</varname> must be of some foreign pointer,
array, or record type.</para>
</sect2>
<sect2><title>Foreign Dynamic Allocation</title>
<para>Lisp code can call the C standard library functions
<function>malloc</function> and <function>free</function> to dynamically allocate and
deallocate foreign variables. The Lisp code shares the same allocator
with foreign C code, so it's OK for foreign code to call
<function>free</function> on the result of Lisp
<function>sb-alien:make-alien</function>, or for Lisp code to call
<function>sb-alien:free-alien</function> on foreign objects allocated by C
code.</para>
<synopsis>(sb-alien:make-alien type size)</synopsis>
<para>The <function>sb-alien:make-alien</function> macro
returns a dynamically allocated foreign value of the specified
<varname>type</varname> (which is not evaluated.) The allocated memory is not
initialized, and may contain arbitrary junk. If supplied,
<varname>size</varname> is an expression to evaluate to compute the size of the
allocated object. There are two major cases:
<itemizedlist>
<listitem>
<para>When <varname>type</varname> is a foreign array type, an array of
that type is allocated and a pointer to it is returned. Note that you
must use <function>deref</function> to change the result to an array before you
can use <function>deref</function> to read or write elements:
<programlisting>
(cl:in-package "CL-USER") ; which USEs package "SB-ALIEN"
(defvar *foo* (make-alien (array char 10)))
(type-of *foo*) => (alien (* (array (signed 8) 10)))
(setf (deref (deref foo) 0) 10) => 10</programlisting>
If supplied, <varname>size</varname> is used as the first dimension for the
array.</para>
</listitem>
<listitem>
<para>When <varname>type</varname> is any other foreign type, then an
object for that type is allocated, and a pointer to it is
returned. So <function>(make-alien int)</function> returns a <type>(* int)</type>.
If <varname>size</varname> is specified, then a block of that many
objects is allocated, with the result pointing to the first one.</para>
</listitem>
</itemizedlist>
</para>
<synopsis>(sb-alien:free-alien foreign-value)</synopsis>
<para>The <function>sb-alien:free-alien</function> function
frees the storage for <varname>foreign-value</varname>,
which must have been allocated with Lisp <function>make-alien</function>
or C <function>malloc</function>.</para>
<para>See also the <function>sb-alien:with-alien</function> macro, which
allocates foreign values on the stack.</para>
</sect2>
</sect1>
<sect1 id="foreign-variables"><title>Foreign Variables</title>
<!-- AKA "Alien Variables" in the CMU CL manual -->
<para>
Both local (stack allocated) and external (C global) foreign variables are
supported.
</para>
<sect2><title>Local Foreign Variables</title>
<synopsis>(sb-alien:with-alien var-definitions &amp;body body)</synopsis>
<para>The <function>with-alien</function>
macro establishes local
foreign variables
with the specified
alien types and names.
This form is analogous to defining a local variable in C: additional
storage is allocated, and the initial value is copied.
This form is less
analogous to LET-allocated Lisp variables, since the variables
can't be captured in closures: they live only for the dynamic extent
of the body, and referring to them outside is a gruesome error.
</para>
<para>The <varname>var-definitions</varname> argument is a list of
variable definitions, each of the form
<programlisting>(name type &amp;optional initial-value)</programlisting>
The names of the variables are established as symbol-macros; the bindings have
lexical scope, and may be assigned with <function>setq</function>
or <function>setf</function>.
</para>
<para>The <function>with-alien</function> macro also establishes
a new scope for named structures
and unions. Any <varname>type</varname> specified for a variable may contain
named structure or union types with the slots specified. Within the
lexical scope of the binding specifiers and body, a locally defined
foreign structure type <type>foo</type> can be referenced by its name using
<type>(struct foo)</type>.
</para>
</sect2>
<sect2><title>External Foreign Variables</title>
<para>
External foreign names are strings, and Lisp names are symbols. When
an external foreign value is represented using a Lisp variable, there
must be a way to convert from one name syntax into the other. The
macros <function>extern-alien</function>, <function>define-alien-variable</function> and
<function>define-alien-routine</function> use this conversion heuristic:
<itemizedlist>
<listitem><para>Alien names are converted to Lisp names by uppercasing and
replacing underscores with hyphens.</para></listitem>
<listitem><para>Conversely, Lisp names are converted to alien names by
lowercasing and replacing hyphens with underscores.</para></listitem>
<listitem><para>Both the Lisp symbol and alien string names may be
separately specified by using a list of the form
<programlisting>(alien-string lisp-symbol)</programlisting></para></listitem>
</itemizedlist>
</para>
<synopsis>(sb-alien:define-alien-variable name type)</synopsis>
<para>
The <function>define-alien-variable</function> macro
defines <varname>name</varname> as an external foreign variable of the
specified foreign <type>type</type>. <varname>name</varname> and <type>type</type> are not
evaluated. The Lisp name of the variable (see above) becomes a
global alien variable. Global alien variables
are effectively ``global symbol macros''; a reference to the
variable fetches the contents of the external variable. Similarly,
setting the variable stores new contents---the new contents must be
of the declared <type>type</type>. Someday, they may well be implemented
using the &ANSI; <function>define-symbol-macro</function> mechanism, but
as of &SBCL; 0.7.5, they are still implemented using an older
more-or-less parallel mechanism inherited from &CMUCL;.
</para>
<para>
For example, to access a C-level counter <varname>foo</varname>, one could
write
<programlisting>
(define-alien-variable "foo" int)
;; Now it is possible to get the value of the C variable foo simply by
;; referencing that Lisp variable:
(print foo)
(setf foo 14)
(incf foo)</programlisting>
</para>
<synopsis>(sb-alien:get-errno)</synopsis>
<para>
Since in modern C libraries, the <varname>errno</varname> "variable" is typically
no longer a variable, but some bizarre artificial construct
which behaves superficially like a variable within a given thread,
it can no longer reliably be accessed through the ordinary
<varname>define-alien-variable</varname> mechanism. Instead, &SBCL; provides
the operator <function>sb-alien:get-errno</function> to allow Lisp code to read it.
</para>
<synopsis>(sb-alien:extern-alien name type)</synopsis>
<para>
The <function>extern-alien</function> macro
returns an alien with the specified <type>type</type> which
points to an externally defined value. <varname>name</varname> is not evaluated,
and may be either a string or a symbol. <type>type</type> is
an unevaluated alien type specifier.
</para>
</sect2>
</sect1>
<sect1 id="foreign-data-structure"><title>Foreign Data Structure Examples</title>
<!-- AKA "Alien Data Structure Example" in the CMU CL manual -->
<para>
Now that we have alien types, operations and variables, we can manipulate
foreign data structures. This C declaration
<programlisting>
struct foo {
int a;
struct foo *b[100];
};</programlisting>
can be translated into the following alien type:
<programlisting>(define-alien-type nil
(struct foo
(a int)
(b (array (* (struct foo)) 100))))</programlisting>
</para>
<para>
Once the <type>foo</type> alien type has been defined as above,
the C expression
<programlisting>
struct foo f;
f.b[7].a</programlisting>
can be translated in this way:
<programlisting>
(with-alien ((f (struct foo)))
(slot (deref (slot f 'b) 7) 'a)
;;
;; Do something with f...
)</programlisting>
</para>
<para>
Or consider this example of an external C variable and some accesses:
<programlisting>
struct c_struct {
short x, y;
char a, b;
int z;
c_struct *n;
};
extern struct c_struct *my_struct;
my_struct->x++;
my_struct->a = 5;
my_struct = my_struct->n;</programlisting>
which can be manipulated in Lisp like this:
<programlisting>
(define-alien-type nil
(struct c-struct
(x short)
(y short)
(a char)
(b char)
(z int)
(n (* c-struct))))
(define-alien-variable "my_struct" (* c-struct))
(incf (slot my-struct 'x))
(setf (slot my-struct 'a) 5)
(setq my-struct (slot my-struct 'n))</programlisting>
</para>
</sect1>
<sect1 id="load-object"><title>Loading Unix Object Files</title>
<para>
Foreign object files can be loaded into the running Lisp process by
calling the functions <function>load-foreign</function> or
<function>load-1-foreign</function>.
</para>
<para> The <function>sb-alien:load-1-foreign</function> function is the more
primitive of the two operations. It loads a single object file. into
the currently running Lisp. The external symbols defining routines and
variables are made available for future external references (e.g. by
<function>extern-alien</function>). Forward references to foreign symbols
aren't supported: <function>load-1-foreign</function> must be run before any
of the defined symbols are referenced.
</para>
<para><function>sb-alien:load-foreign</function> is built in terms of
<function>load-1-foreign</function> and some other machinery
like <function>sb-ext:run-program</function>.
It accepts a list of files and libraries,
and runs the linker on the files and
libraries, creating an absolute Unix object file which is then
processed by <function>load-1-foreign</function>.</para>
<note><para>As of &SBCL; 0.7.5, all foreign code (code loaded
with <function>load-1-function</function> or <function>load-function</function>) is
lost when a Lisp core is saved with
<function>sb-ext:save-lisp-and-die</function>, and no attempt is made to
restore it when the core is loaded. Historically this has been an
annoyance both for &SBCL; users and for &CMUCL; users.
It's hard to solve this problem completely cleanly, but some
generally-reliable partial solution might be useful. Once someone in
either camp gets sufficiently annoyed to create it, &SBCL; is
likely to adopt some mechanism for automatically restoring foreign
code when a saved core is loaded.</para></note>
</sect1>
<sect1 id="foreign-function-calls"><title>Foreign Function Calls</title>
<para>
The foreign function call interface allows a Lisp program to call
many functions written in languages that use the C calling convention.
</para>
<para>
Lisp sets up various signal handling routines and other environment
information when it first starts up, and expects these to be in place
at all times. The C functions called by Lisp should not change the
environment, especially the signal handlers: the signal handlers
installed by Lisp typically have interesting flags set (e.g to request
machine context information, or for signal delivery on an alternate
stack) which the Lisp runtime relies on for correct operation.
Precise details of how this works may change without notice between
versions; the source, or the brain of a friendly &SBCL; developer,
is the only documentation. Users of a Lisp built with the :sb-thread
feature should also read the Threading section
<!-- FIXME I'm sure docbook has some syntax for internal links -->
of this manual</para>
<sect2><title>The <function>alien-funcall</function> Primitive</title>
<synopsis>(sb-alien:alien-funcall alien-function &amp;rest arguments)</synopsis>
<para>
The <function>alien-funcall</function> function is the foreign function call
primitive: <varname>alien-function</varname> is called with the supplied
<varname>arguments</varname> and its C return value is returned as a Lisp value.
The <varname>alien-function</varname> is an arbitrary
run-time expression; to refer to a constant function, use
<function>extern-alien</function> or a value defined by
<function>define-alien-routine</function>.
</para>
<para>
The type of <function>alien-function</function>
must be <type>(alien (function ...))</type>
or <type>(alien (* (function ...)))</type>.
The function type is used to
determine how to call the function (as though it was declared with
a prototype.) The type need not be known at compile time, but only
known-type calls are efficiently compiled. Limitations:
<itemizedlist>
<listitem><para>Structure type return values are not implemented.</para></listitem>
<listitem><para>Passing of structures by value is not implemented.</para></listitem>
</itemizedlist>
</para>
<para>
Here is an example which allocates a <type>(struct foo)</type>, calls a foreign
function to initialize it, then returns a Lisp vector of all the
<type>(* (struct foo))</type> objects filled in by the foreign call:
<programlisting>
;; Allocate a foo on the stack.
(with-alien ((f (struct foo)))
;; Call some C function to fill in foo fields.
(alien-funcall (extern-alien "mangle_foo" (function void (* foo)))
(addr f))
;; Find how many foos to use by getting the A field.
(let* ((num (slot f 'a))
(result (make-array num)))
;; Get a pointer to the array so that we don't have to keep extracting it:
(with-alien ((a (* (array (* (struct foo)) 100)) (addr (slot f 'b))))
;; Loop over the first N elements and stash them in the result vector.
(dotimes (i num)
(setf (svref result i) (deref (deref a) i)))
;; Voila.
result)))</programlisting>
</para>
</sect2>
<sect2><title>The <function>define-alien-routine</function> Macro</title>
<synopsis>(sb-alien:define-alien-routine} name result-type &amp;rest arg-specifiers)</synopsis>
<para>
The <function>define-alien-routine</function> macro is a convenience
for automatically generating Lisp
interfaces to simple foreign functions. The primary feature is the
parameter style specification, which translates the C
pass-by-reference idiom into additional return values.
</para>
<para>
<varname>name</varname> is usually a string external symbol, but may also be a
symbol Lisp name or a list of the foreign name and the Lisp name.
If only one name is specified, the other is automatically derived
as for <function>extern-alien</function>.
<varname>result-type</varname> is the alien type of the return value.
</para>
<para>
Each element of the <varname>arg-specifiers</varname> list
specifies an argument to the foreign function, and is
of the form
<programlisting>(aname atype &amp;optional style)</programlisting>
<varname>aname</varname> is the symbol name of the argument to the constructed
function (for documentation). <varname>atype</varname> is the alien type of
corresponding foreign argument. The semantics of the actual call
are the same as for <function>alien-funcall</function>. <varname>style</varname>
specifies how this argument should be handled at call and return time,
and should be one of the following
<itemizedlist>
<listitem><para><varname>:in</varname>specifies that the argument is
passed by value. This is the default. <varname>:in</varname> arguments
have no corresponding return value from the Lisp function.
</para></listitem>
<listitem><para><varname>:copy</varname> is similar to <varname>:in</varname>,
but the argument is copied
to a pre-allocated object and a pointer to this object is passed
to the foreign routine.</para></listitem>
<listitem><para><varname>:out</varname> specifies a pass-by-reference
output value. The type of the argument must be a pointer to
a fixed-sized object (such as an integer or pointer).
<varname>:out</varname> and <varname>:in-out</varname> style cannot
be used with pointers to arrays, records or functions. An
object of the correct size is allocated on the stack, and
its address is passed to the foreign function. When the
function returns, the contents
of this location are returned as one of the values of the Lisp
function (and the location is automatically deallocated).
</para></listitem>
<listitem><para><varname>:in-out</varname> is a combination of
<varname>:copy</varname> and <varname>:out</varname>.
The argument is copied to a pre-allocated object and a pointer to
this object is passed to the foreign routine. On return, the
contents of this location is returned as an additional value.
</para></listitem>
</itemizedlist>
</para>
<note>
<para>
Any efficiency-critical foreign interface function should be inline
expanded, which can be done by preceding the
<function>define-alien-routine</function> call with:
<programlisting>(declaim (inline lisp-name))</programlisting>
In addition to avoiding the Lisp call overhead, this allows
pointers, word-integers and floats to be passed using non-descriptor
representations, avoiding consing.)
</para>
</note>
</sect2>
<sect2><title><function>define-alien-routine</function> Example</title>
<para>
Consider the C function <function>cfoo</function>
with the following calling convention:
<programlisting>
void
cfoo (str, a, i)
char *str;
char *a; /* update */
int *i; /* out */
{
/* body of cfoo(...) */
}</programlisting>
This can be described by the following call to
<function>define-alien-routine</function>:
<programlisting>
(define-alien-routine "cfoo" void
(str c-string)
(a char :in-out)
(i int :out))</programlisting>
The Lisp function <function>cfoo</function> will have
two arguments (<varname>str</varname> and <varname>a</varname>)
and two return values (<varname>a</varname> and <varname>i</varname>).
</para>
</sect2>
<sect2><title>Calling Lisp From C</title>
<para>
Calling Lisp functions from C is sometimes possible, but is extremely
hackish and poorly supported as of &SBCL; 0.7.5.
See <function>funcall0</function> ... <function>funcall3</function> in
the runtime system. The
arguments must be valid &SBCL; object descriptors (so that
e.g. fixnums must be
left-shifted by 2.) As of &SBCL; 0.7.5, the format
of object descriptors is documented only by the source code and, in parts,
by the old &CMUCL; "INTERNALS" documentation.</para>
<para> Note that the garbage collector moves objects, and won't be
able to fix up any references in C variables. There are three
mechanisms for coping with this:
<orderedlist>
<listitem><para>The <function>sb-ext:purify</function> moves all live Lisp
data into static or read-only areas such that it will never be moved
(or freed) again in the life of the Lisp session</para></listitem>
<listitem><para><function>sb-sys:with-pinned-objects</function> is a
macro which arranges for some set of objects to be pinned in memory
for the dynamic extent of its body forms. On ports which use the
generational garbage collector (as of &SBCL; 0.8.3, only the x86) this
has a page granularity - i.e. the entire 4k page or pages containing
the objects will be locked down. On other ports it is implemented by
turning off GC for the duration (so could be said to have a
whole-world granularity). </para></listitem>
<listitem><para>Disable GC, using the <function>without-gcing</function>
macro or <function>gc-off</function> call.</para></listitem>
</orderedlist>
</para>
<!-- FIXME: This is a "changebar" section from the CMU CL manual.
I (WHN 2002-07-14) am not very familiar with this content, so
I'm not immediately prepared to try to update it for SBCL, and
I'm not feeling masochistic enough to work to encourage this
kind of low-level hack anyway. However, I acknowledge that callbacks
are sometimes really really necessary, so I include the original
text in case someone is hard-core enough to benefit from it. If
anyone brings the information up to date for SBCL, it belong
either in the main manual or on a CLiki SBCL Internals page.
LaTeX \subsection{Accessing Lisp Arrays}
LaTeX
LaTeX Due to the way \cmucl{} manages memory, the amount of memory that can
LaTeX be dynamically allocated by \code{malloc} or \funref{make-alien} is
LaTeX limited\footnote{\cmucl{} mmaps a large piece of memory for it's own
LaTeX use and this memory is typically about 8 MB above the start of the C
LaTeX heap. Thus, only about 8 MB of memory can be dynamically
LaTeX allocated.}.
Empirically determined to be considerably >8Mb on this x86 linux
machine, but I don't know what the actual values are - dan 2003.09.01
Note that this technique is used in SB-GROVEL in the SBCL contrib
LaTeX
LaTeX To overcome this limitation, it is possible to access the content of
LaTeX Lisp arrays which are limited only by the amount of physical memory
LaTeX and swap space available. However, this technique is only useful if
LaTeX the foreign function takes pointers to memory instead of allocating
LaTeX memory for itself. In latter case, you will have to modify the
LaTeX foreign functions.
LaTeX
LaTeX This technique takes advantage of the fact that \cmucl{} has
LaTeX specialized array types (\pxlref{specialized-array-types}) that match
LaTeX a typical C array. For example, a \code{(simple-array double-float
LaTeX (100))} is stored in memory in essentially the same way as the C
LaTeX array \code{double x[100]} would be. The following function allows us
LaTeX to get the physical address of such a Lisp array:
LaTeX \begin{example}
LaTeX (defun array-data-address (array)
LaTeX "Return the physical address of where the actual data of an array is
LaTeX stored.
LaTeX
LaTeX ARRAY must be a specialized array type in CMU Lisp. This means ARRAY
LaTeX must be an array of one of the following types:
LaTeX
LaTeX double-float
LaTeX single-float
LaTeX (unsigned-byte 32)
LaTeX (unsigned-byte 16)
LaTeX (unsigned-byte 8)
LaTeX (signed-byte 32)
LaTeX (signed-byte 16)
LaTeX (signed-byte 8)
LaTeX "
LaTeX (declare (type (or #+signed-array (array (signed-byte 8))
LaTeX #+signed-array (array (signed-byte 16))
LaTeX #+signed-array (array (signed-byte 32))
LaTeX (array (unsigned-byte 8))
LaTeX (array (unsigned-byte 16))
LaTeX (array (unsigned-byte 32))
LaTeX (array single-float)
LaTeX (array double-float))
LaTeX array)
LaTeX (optimize (speed 3) (safety 0))
LaTeX (ext:optimize-interface (safety 3)))
LaTeX ;; with-array-data will get us to the actual data. However, because
LaTeX ;; the array could have been displaced, we need to know where the
LaTeX ;; data starts.
LaTeX (lisp::with-array-data ((data array)
LaTeX (start)
LaTeX (end))
LaTeX (declare (ignore end))
LaTeX ;; DATA is a specialized simple-array. Memory is laid out like this:
LaTeX ;;
LaTeX ;; byte offset Value
LaTeX ;; 0 type code (should be 70 for double-float vector)
LaTeX ;; 4 4 * number of elements in vector
LaTeX ;; 8 1st element of vector
LaTeX ;; ... ...
LaTeX ;;
LaTeX (let ((addr (+ 8 (logandc1 7 (kernel:get-lisp-obj-address data))))
LaTeX (type-size (let ((type (array-element-type data)))
LaTeX (cond ((or (equal type '(signed-byte 8))
LaTeX (equal type '(unsigned-byte 8)))
LaTeX 1)
LaTeX ((or (equal type '(signed-byte 16))
LaTeX (equal type '(unsigned-byte 16)))
LaTeX 2)
LaTeX ((or (equal type '(signed-byte 32))
LaTeX (equal type '(unsigned-byte 32)))
LaTeX 4)
LaTeX ((equal type 'single-float)
LaTeX 4)
LaTeX ((equal type 'double-float)
LaTeX 8)
LaTeX (t
LaTeX (error "Unknown specialized array element type"))))))
LaTeX (declare (type (unsigned-byte 32) addr)
LaTeX (optimize (speed 3) (safety 0) (ext:inhibit-warnings 3)))
LaTeX (system:int-sap (the (unsigned-byte 32)
LaTeX (+ addr (* type-size start)))))))
LaTeX \end{example}
LaTeX
LaTeX Assume we have the C function below that we wish to use:
LaTeX \begin{example}
LaTeX double dotprod(double* x, double* y, int n)
LaTeX \{
LaTeX int k;
LaTeX double sum = 0;
LaTeX
LaTeX for (k = 0; k < n; ++k) \{
LaTeX sum += x[k] * y[k];
LaTeX \}
LaTeX \}
LaTeX \end{example}
LaTeX The following example generates two large arrays in Lisp, and calls the C
LaTeX function to do the desired computation. This would not have been
LaTeX possible using \code{malloc} or \code{make-alien} since we need about
LaTeX 16 MB of memory to hold the two arrays.
LaTeX \begin{example}
LaTeX (define-alien-routine "dotprod" double
LaTeX (x (* double-float) :in)
LaTeX (y (* double-float) :in)
LaTeX (n int :in))
LaTeX
LaTeX (let ((x (make-array 1000000 :element-type 'double-float))
LaTeX (y (make-array 1000000 :element-type 'double-float)))
LaTeX ;; Initialize X and Y somehow
LaTeX (let ((x-addr (system:int-sap (array-data-address x)))
LaTeX (y-addr (system:int-sap (array-data-address y))))
LaTeX (dotprod x-addr y-addr 1000000)))
LaTeX \end{example}
LaTeX In this example, it may be useful to wrap the inner \code{let}
LaTeX expression in an \code{unwind-protect} that first turns off garbage
LaTeX collection and then turns garbage collection on afterwards. This will
LaTeX prevent garbage collection from moving \code{x} and \code{y} after we
LaTeX have obtained the (now erroneous) addresses but before the call to
LaTeX \code{dotprod} is made.
LaTeX
-->
</sect2>
</sect1>
<sect1 id="ffi-example"><title>Step-By-Step Example of the Foreign Function Interface</title>
<para>
This section presents a complete example of an interface to a somewhat
complicated C function.
</para>
<para>
Suppose you have the following C function which you want to be able to
call from Lisp in the file <filename>test.c</filename>
<programlisting>
struct c_struct
{
int x;
char *s;
};
struct c_struct *c_function (i, s, r, a)
int i;
char *s;
struct c_struct *r;
int a[10];
{
int j;
struct c_struct *r2;
printf("i = %d\n", i);
printf("s = %s\n", s);
printf("r->x = %d\n", r->x);
printf("r->s = %s\n", r->s);
for (j = 0; j &lt; 10; j++) printf("a[%d] = %d.\n", j, a[j]);
r2 = (struct c_struct *) malloc (sizeof(struct c_struct));
r2->x = i + 5;
r2->s = "a C string";
return(r2);
};</programlisting>
</para>
<para>
It is possible to call this C function from Lisp using the file
<filename>test.lisp</filename> containing
<programlisting>
(cl:defpackage "TEST-C-CALL" (:use "CL" "SB-ALIEN" "SB-C-CALL"))
(cl:in-package "TEST-C-CALL")
;;; Define the record C-STRUCT in Lisp.
(define-alien-type nil
(struct c-struct
(x int)
(s c-string)))
;;; Define the Lisp function interface to the C routine. It returns a
;;; pointer to a record of type C-STRUCT. It accepts four parameters:
;;; I, an int; S, a pointer to a string; R, a pointer to a C-STRUCT
;;; record; and A, a pointer to the array of 10 ints.
;;;
;;; The INLINE declaration eliminates some efficiency notes about heap
;;; allocation of alien values.
(declaim (inline c-function))
(define-alien-routine c-function
(* (struct c-struct))
(i int)
(s c-string)
(r (* (struct c-struct)))
(a (array int 10)))
;;; a function which sets up the parameters to the C function and
;;; actually calls it
(defun call-cfun ()
(with-alien ((ar (array int 10))
(c-struct (struct c-struct)))
(dotimes (i 10) ; Fill array.
(setf (deref ar i) i))
(setf (slot c-struct 'x) 20)
(setf (slot c-struct 's) "a Lisp string")
(with-alien ((res (* (struct c-struct))
(c-function 5 "another Lisp string" (addr c-struct) ar)))
(format t "~&amp;back from C function~%")
(multiple-value-prog1
(values (slot res 'x)
(slot res 's))
;; Deallocate result. (after we are done referring to it:
;; "Pillage, *then* burn.")
(free-alien res)))))</programlisting>
</para>
<para>
To execute the above example, it is necessary to compile the C routine,
e.g.:
<userinput>cc -c test.c</userinput>
(In order to enable incremental loading with some linkers, you may need
to say
<userinput>cc -G 0 -c test.c</userinput>)
</para>
<para>
Once the C code has been compiled, you can start up Lisp and load it in:
<userinput>sbcl</userinput>.
Lisp should start up with its normal prompt.</para>
<para>
Within Lisp,
compile the Lisp file. (This step can be done separately. You don't
have to recompile every time.)
<userinput>(compile-file "test.lisp")</userinput>
</para>
<para>
Within Lisp, load the foreign object file to define the necessary
symbols:
<userinput>(load-foreign "test.o")</userinput>.
This must be done before loading any code that refers
to these symbols.
</para>
<para>
Now you can load the compiled Lisp ("fasl") file into Lisp:
<userinput>(load "test.fasl")</userinput>
And once the Lisp file is loaded, you can call the
Lisp routine that sets up the parameters and calls the C
function:
<userinput>(test-c-call::call-cfun)</userinput>
</para>
<para>
The C routine should print the following information to standard output:
<!-- FIXME: What should be here is a verbatim environment for computer
output, but since I don't know one in DocBook, I made do with
PROGRAMLISTING for now... -->
<programlisting>i = 5
s = another Lisp string
r->x = 20
r->s = a Lisp string
a[0] = 0.
a[1] = 1.
a[2] = 2.
a[3] = 3.
a[4] = 4.
a[5] = 5.
a[6] = 6.
a[7] = 7.
a[8] = 8.
a[9] = 9.</programlisting>
After return from the C function,
the Lisp wrapper function should print the following output:
<programlisting>back from C function</programlisting>
And upon return from the Lisp wrapper function,
before the next prompt is printed, the
Lisp read-eval-print loop should print the following return values:
<!-- FIXME: As above, it's not a program listing, but computer output... -->
<programlisting>
10
"a C string"
</programlisting>
</para>
</sect1>
</chapter>