;;; st_size is a long, not an off-t, because off-t is a 64-bit
;;; quantity on Alpha. And FIXME: "No one would want a file length
;;; longer than 32 bits anyway, right?":-|
+;;;
+;;; The comment about alien and 64-bit quantities has not been kept in
+;;; sync with the comment now in wrap.h (formerly wrap.c), but it's
+;;; not clear whether either comment is correct. -- RMK 2007-11-14.
(define-alien-type nil
(struct wrapped_stat
- (st-dev #!-(or mips largefile) unsigned-int
- #!+mips unsigned-long
- #!+(and largefile (not mips)) dev-t)
+ (st-dev ffi-dev-t)
(st-ino ino-t)
(st-mode mode-t)
(st-nlink nlink-t)
(st-uid uid-t)
(st-gid gid-t)
- (st-rdev #!-(or mips largefile) unsigned-int
- #!+mips unsigned-long
- #!+(and largefile (not mips)) dev-t)
- (st-size #!-(or darwin mips largefile) unsigned-int
- #!+(or darwin mips largefile) off-t)
- #!+(and darwin)
- (st-blksize unsigned-int)
- #!-(and darwin)
- (st-blksize unsigned-long)
+ (st-rdev ffi-dev-t)
+ (st-size ffi-off-t)
+ (st-blksize ffi-blksize-t)
(st-blocks unsigned-long)
(st-atime time-t)
(st-mtime time-t)
#include "runtime.h"
#include "util.h"
+#include "wrap.h"
/* Although it might seem as though this should be in some standard
Unix header, according to Perry E. Metzger, in a message on
* stat(2) stuff
*/
-/* As of 0.6.12, the FFI can't handle 64-bit values. For now, we use
- * these munged-to-32-bits values for might-be-64-bit slots of
- * stat_wrapper as a workaround, so that at least we can still work
- * when values are small.
- *
- * FIXME: But of course we should fix the FFI so that we can use the
- * actual 64-bit values instead. In fact, we probably have by now
- * (2003-10-03) on all working platforms except MIPS and HPPA; if some
- * motivated spark would simply fix those, this hack could go away.
- * -- CSR, 2003-10-03
- *
- * Some motivated spark fixed MIPS. -- ths, 2005-10-06 */
-
-#if defined(LISP_FEATURE_LARGEFILE) && !defined(LISP_FEATURE_MIPS)
-typedef dev_t ffi_dev_t;
-typedef off_t ffi_off_t;
-#elif defined(LISP_FEATURE_MIPS)
-typedef unsigned long ffi_dev_t; /* Linux/MIPS struct stat doesn't use dev_t */
-typedef off_t ffi_off_t;
-#elif defined(LISP_FEATURE_DARWIN)
-typedef dev_t ffi_dev_t;
-typedef off_t ffi_off_t;
-#else
-typedef u32 ffi_dev_t; /* since Linux dev_t can be 64 bits */
-typedef u32 ffi_off_t; /* since OpenBSD 2.8 st_size is 64 bits */
-#endif
-
-#ifdef LISP_FEATURE_OS_PROVIDES_BLKSIZE_T
-typedef blksize_t ffi_blksize_t;
-#else
-typedef unsigned long ffi_blksize_t;
-#endif
-
-/* a representation of stat(2) results which doesn't depend on CPU or OS */
-struct stat_wrapper {
- /* KLUDGE: The verbose wrapped_st_ prefixes are to protect us from
- * the C preprocessor as wielded by the fiends of OpenBSD, who do
- * things like
- * #define st_atime st_atimespec.tv_sec
- * I remember when I was young and innocent, I read about how the
- * C preprocessor isn't to be used to globally munge random
- * lowercase symbols like this, because things like this could
- * happen, and I nodded sagely. But now I know better.:-| This is
- * another entry for Dan Barlow's ongoing episodic rant about C
- * header files, I guess.. -- WHN 2001-05-10 */
- ffi_dev_t wrapped_st_dev; /* device */
- ino_t wrapped_st_ino; /* inode */
- mode_t wrapped_st_mode; /* protection */
-#ifndef LISP_FEATURE_WIN32
- nlink_t wrapped_st_nlink; /* number of hard links */
- uid_t wrapped_st_uid; /* user ID of owner */
- gid_t wrapped_st_gid; /* group ID of owner */
-#else
- short wrapped_st_nlink; /* Win32 doesn't have nlink_t */
- short wrapped_st_uid; /* Win32 doesn't have st_uid */
- short wrapped_st_gid; /* Win32 doesn't have st_gid */
-#endif
- ffi_dev_t wrapped_st_rdev; /* device type (if inode device) */
- ffi_off_t wrapped_st_size; /* total size, in bytes */
- ffi_blksize_t wrapped_st_blksize; /* blocksize for filesystem I/O */
- unsigned long wrapped_st_blocks; /* number of blocks allocated */
- time_t wrapped_st_atime; /* time_t of last access */
- time_t wrapped_st_mtime; /* time_t of last modification */
- time_t wrapped_st_ctime; /* time_t of last change */
-};
-
static void
copy_to_stat_wrapper(struct stat_wrapper *to, struct stat *from)
{
--- /dev/null
+/*
+ * Data structures used in wrap.c in this directory, moved here from
+ * wrap.c in November 2007 so that
+ * src/tools-for-build/grovel-headers.c can grovel the sizes and
+ * offsets of things.
+ */
+
+
+/* As of 0.6.12, the FFI can't handle 64-bit values. For now, we use
+ * these munged-to-32-bits values for might-be-64-bit slots of
+ * stat_wrapper as a workaround, so that at least we can still work
+ * when values are small.
+ *
+ * FIXME: But of course we should fix the FFI so that we can use the
+ * actual 64-bit values instead. In fact, we probably have by now
+ * (2003-10-03) on all working platforms except MIPS and HPPA; if some
+ * motivated spark would simply fix those, this hack could go away.
+ * -- CSR, 2003-10-03
+ *
+ * Some motivated spark fixed MIPS. -- ths, 2005-10-06 */
+/* It would seem as though the FFI would have to be able to handle
+ * 64-bit values in order for the LARGEFILE && !MIPS case below to
+ * work, so can the comment above still be right? If FFI can only
+ * handle 64-bit aliens on some platforms, maybe there should be a
+ * distinct Lisp feature for 64-bit aliens support? -- RMK,
+ * 2007-11-14 */
+
+#include "sbcl.h"
+#include "runtime.h"
+
+#if defined(LISP_FEATURE_LARGEFILE) && !defined(LISP_FEATURE_MIPS)
+typedef dev_t ffi_dev_t;
+typedef off_t ffi_off_t;
+#elif defined(LISP_FEATURE_MIPS)
+typedef unsigned long ffi_dev_t; /* Linux/MIPS struct stat doesn't use dev_t */
+typedef off_t ffi_off_t;
+#elif defined(LISP_FEATURE_DARWIN)
+typedef dev_t ffi_dev_t;
+typedef off_t ffi_off_t;
+#else
+typedef u32 ffi_dev_t; /* since Linux dev_t can be 64 bits */
+typedef u32 ffi_off_t; /* since OpenBSD 2.8 st_size is 64 bits */
+#endif
+
+#ifdef LISP_FEATURE_OS_PROVIDES_BLKSIZE_T
+typedef blksize_t ffi_blksize_t;
+#else
+typedef unsigned long ffi_blksize_t;
+#endif
+
+/* a representation of stat(2) results which doesn't depend on CPU or OS */
+struct stat_wrapper {
+ /* KLUDGE: The verbose wrapped_st_ prefixes are to protect us from
+ * the C preprocessor as wielded by the fiends of OpenBSD, who do
+ * things like
+ * #define st_atime st_atimespec.tv_sec
+ * I remember when I was young and innocent, I read about how the
+ * C preprocessor isn't to be used to globally munge random
+ * lowercase symbols like this, because things like this could
+ * happen, and I nodded sagely. But now I know better.:-| This is
+ * another entry for Dan Barlow's ongoing episodic rant about C
+ * header files, I guess.. -- WHN 2001-05-10 */
+ ffi_dev_t wrapped_st_dev; /* device */
+ ino_t wrapped_st_ino; /* inode */
+ mode_t wrapped_st_mode; /* protection */
+#ifndef LISP_FEATURE_WIN32
+ nlink_t wrapped_st_nlink; /* number of hard links */
+ uid_t wrapped_st_uid; /* user ID of owner */
+ gid_t wrapped_st_gid; /* group ID of owner */
+#else
+ short wrapped_st_nlink; /* Win32 doesn't have nlink_t */
+ short wrapped_st_uid; /* Win32 doesn't have st_uid */
+ short wrapped_st_gid; /* Win32 doesn't have st_gid */
+#endif
+ ffi_dev_t wrapped_st_rdev; /* device type (if inode device) */
+ ffi_off_t wrapped_st_size; /* total size, in bytes */
+ ffi_blksize_t wrapped_st_blksize; /* blocksize for filesystem I/O */
+ unsigned long wrapped_st_blocks; /* number of blocks allocated */
+ time_t wrapped_st_atime; /* time_t of last access */
+ time_t wrapped_st_mtime; /* time_t of last modification */
+ time_t wrapped_st_ctime; /* time_t of last change */
+};
#include "genesis/config.h"
+#include "wrap.h"
+
#define DEFTYPE(lispname,cname) { cname foo; \
printf("(define-alien-type " lispname " (%s %d))\n", (((foo=-1)<0) ? "sb!alien:signed" : "unsigned"), (8 * (sizeof foo))); }
DEFTYPE("suseconds-t", suseconds_t);
#endif
DEFTYPE("uid-t", uid_t);
+/* Types in src/runtime/wrap.h */
+ DEFTYPE("ffi-dev-t", ffi_dev_t);
+ DEFTYPE("ffi-off-t", ffi_off_t);
+ DEFTYPE("ffi-blksize-t", ffi_blksize_t);
printf("\n");
printf(";;; fcntl.h (or unistd.h on OpenBSD and NetBSD)\n");
;;; checkins which aren't released. (And occasionally for internal
;;; versions, especially for internal versions off the main CVS
;;; branch, it gets hairier, e.g. "0.pre7.14.flaky4.13".)
-"1.0.11.25"
+"1.0.11.26"