[Bug ada/114065] GNAT build with -D_TIME_BITS=64 -D_FILE_OFFSET_BITS=64 fails on 32-bit

[nicolas at debian dot org via Gcc-bugs]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=114065

--- Comment #72 from Nicolas Boulenguez <nicolas at debian dot org> ---
What are the arguments against an Ada representation clause generated from the
actual C layout?

POSIX allows additionnal fields like the Y2038 padding, so GNAT must be
informed of the C configuration in a way or another. Build-time automation
seems safer than human reasoning when there are so many possibilities.

For the record, here is the declaration in
glibc-2.44/time/bits/types/struct_timespec.h.
struct timespec
{
#ifdef __USE_TIME64_REDIRECTS
  __time64_t tv_sec;            /* Seconds.  */
#else
  __time_t tv_sec;              /* Seconds.  */
#endif
#if __WORDSIZE == 64 \
  || (defined __SYSCALL_WORDSIZE && __SYSCALL_WORDSIZE == 64) \
  || (__TIMESIZE == 32 && !defined __USE_TIME64_REDIRECTS)
  __syscall_slong_t tv_nsec;    /* Nanoseconds.  */
#else
# if __BYTE_ORDER == __BIG_ENDIAN
  int: 32;           /* Padding.  */
  long int tv_nsec;  /* Nanoseconds.  */
# else
  long int tv_nsec;  /* Nanoseconds.  */
  int: 32;           /* Padding.  */
# endif
#endif   
};