8bd5dbf7cc
The old `cast' didn't work right for a mismatch between a pointer GCableness and the source or target types, and it didn't work right for an GCable pointer with a non-zero offset. While those pitfalls were documented, the first of them definitely has been a source of bugs in code that I wrote. Also added `cpointer-gcable?'
84 lines
2.8 KiB
Racket
84 lines
2.8 KiB
Racket
#lang scribble/doc
|
|
@(require "utils.rkt")
|
|
|
|
@title{Miscellaneous Support}
|
|
|
|
@defproc[(list->cblock [lst list?] [type ctype?]) any]{
|
|
|
|
Allocates a memory block of an appropriate size, and initializes it
|
|
using values from @racket[lst] and the given @racket[type]. The
|
|
@racket[lst] must hold values that can all be converted to C values
|
|
according to the given @racket[type].}
|
|
|
|
|
|
@defproc[(vector->cblock [vec vector?] [type type?]) any]{
|
|
|
|
Like @racket[list->cblock], but for Racket vectors.}
|
|
|
|
|
|
@defproc[(vector->cpointer [vec vector?]) cpointer?]{
|
|
|
|
Returns a pointer to an array of @racket[_scheme] values, which is the
|
|
internal representation of @racket[vec].}
|
|
|
|
@defproc[(flvector->cpointer [flvec flvector?]) cpointer?]{
|
|
|
|
Returns a pointer to an array of @racket[_double] values, which is the
|
|
internal representation of @racket[flvec].}
|
|
|
|
@defproc[(saved-errno) exact-integer?]{
|
|
|
|
Returns the value most recently saved (in the current thread) after a
|
|
foreign call with a non-@racket[#f] @racket[#:save-errno] option (see
|
|
@racket[_fun] and @racket[_cprocedure]).}
|
|
|
|
@defproc[(lookup-errno [sym (or/c 'EINTR 'EEXIST 'EAGAIN)])
|
|
exact-integer?]{
|
|
|
|
Returns a platform-specific value corresponding to a Posix @tt{errno}
|
|
symbol. The set of supported symbols is likely to expand in the
|
|
future.}
|
|
|
|
|
|
@defproc[(cast [v any/c] [from-type ctype?] [to-type ctype?]) any/c]{
|
|
|
|
Converts @racket[v] from a value matching @racket[from-type] to a
|
|
value matching @racket[to-type], where @racket[(ctype-sizeof from-type)]
|
|
matches @racket[(ctype-sizeof to-type)].
|
|
|
|
The conversion is roughly equivalent to
|
|
|
|
@racketblock[
|
|
(let ([p (malloc from-type)])
|
|
(ptr-set! p from-type v)
|
|
(ptr-ref p to-type))
|
|
]
|
|
|
|
If @racket[v] is a cpointer, @racket[(cpointer-gcable? v)] is true,
|
|
and @racket[from-type] and @racket[to-type] are both based on
|
|
@racket[_pointer] or @racket[_gcpointer], then @racket[from-type] is
|
|
implicitly converted with @racket[_gcable] to ensure that the result
|
|
cpointer is treated as referring to memory that is managed by the
|
|
garbage collector.
|
|
|
|
If @racket[v] is a pointer with an offset component (e.g., from
|
|
@racket[ptr-add]), @racket[(cpointer-gcable? v)] is true, and the
|
|
result is a cpointer, then the result pointer has the same offset
|
|
component as @racket[v]. If @racket[(cpointer-gcable? v)] is false,
|
|
then any offset is folded into the pointer base for the result.}
|
|
|
|
|
|
@defproc[(cblock->list [cblock any/c] [type ctype?] [length exact-nonnegative-integer?])
|
|
list?]{
|
|
|
|
Converts C @racket[cblock], which is a vector of @racket[type]s, to a
|
|
Racket list. The arguments are the same as in the
|
|
@racket[list->cblock]. The @racket[length] must be specified because
|
|
there is no way to know where the block ends.}
|
|
|
|
|
|
@defproc[(cblock->vector [cblock any/c] [type ctype?] [length exact-nonnegative-integer?])
|
|
vector?]{
|
|
|
|
Like @racket[cblock->vector], but for Racket vectors.}
|