[RFC] Forcing 64-bits __OFF_T_TYPE and __INO_T_TYPE for new 32-bit architectures?

[Zack Weinberg]

I would certainly support having off_t and ino_t always be 64-bit on
new architectures, even if they only support 32-bit pointers.  I am
not qualified to assess whether this patch is the right way to go
about it.

Shouldn't blkcnt_t, fsblkcnt_t, and fsfilcnt_t also get this
treatment?  (And anything else affected by -D_FILE_OFFSET_BITS=64 that
I missed.) And, while we're at it, time_t?

zw