Various cleanups for Dynarr code, from Unicode-internal ws dynarr.c: Add comment explaining Dynarr_largest() use. dynarr.c: In Dynarr_insert_many(), don't call Dynarr_resize() unless we actually need to resize, and note that an assert() that we are inserting at or below the current end could be wrong if code wants to access stuff between `len' and `largest'. dynarr.c: Don't just Dynarr_resize() to the right size; instead use Dynarr_reset() then Dynarr_add_many(), so that the 'len' and 'largest' and such get set properly. dynarr.c, faces.c, gutter.c, lisp.h, lread.c, lrecord.h, redisplay-output.c, redisplay.c: Rename Dynarr member 'cur' to 'len' since it's the length of the dynarr, not really a pointer to a "current insertion point". Use type_checking_assert() instead of just assert() in some places. Add additional assertions (Dynarr_verify*()) to check that we're being given positions within range. Use them in Dynarr_at, Dynarr_atp, etc. New Dynarr_atp_allow_end() for retrieving a pointer to a position that might be the element past the last one. New Dynarr_past_lastp() to retrieve a pointer to the position past the last one, using Dynarr_atp_allow_end(). Change code appropriately to use it. Rename Dynarr_end() to Dynarr_lastp() (pointer to the last element) for clarity, and change code appropriately to use it. Change code appropriately to use Dynarr_begin(). Rewrite Dynarr_add_many(). New version can accept a NULL pointer to mean "reserve space but don't put anything in it". Used by stack_like_malloc().

/* Synched up with: FSF 19.31. */

/* I don't care if this doesn't do more than including bsd4-3.h;
   Mach is not bsd4-3 and the moment you forget it chances are that
   you're in deep shit.  */

#include "bsd4-3.h"

/* The rest of this stuff is XEmacs additions. */
 
/* SYSTEM_TYPE should indicate the kind of system you are using.
 It sets the Lisp variable system-type.  */

#undef SYSTEM_TYPE
#define SYSTEM_TYPE "mach"

/* Don't send signals to subprocesses by "typing" special chars at them. */
#undef SIGNALS_VIA_CHARACTERS

/* XEmacs change */
/* unistd.h defines _POSIX_VERSION, which leads some things to believe 
   that _POSIX_PATH_MAX should be defined.  Unfortunately, it isn't. */
#ifndef NOT_C_CODE
#include <sys/param.h>
#define _POSIX_PATH_MAX MAXPATHLEN
#endif

#ifndef NOT_C_CODE
typedef int pid_t;
/* XEmacs change */
typedef unsigned short mode_t;
#endif /* NOT_C_CODE */

#if (defined(i386) || defined(ibmrt))
  /* use drem() instead of fmod() -- this is a BUG in the compiler runtime. */
# define USE_DREM
#endif