[RFC] Make _FILE_OFFSET_BITS=64 default.

[Russ Allbery]

Florian Weimer <fweimer@redhat.com> writes:
> On 03/05/2014 07:07 PM, Rich Felker wrote:

>> No, it's just a change in the default exposed features. There's no ABI
>> issue because programs built with _FILE_OFFSET_BITS=64 access different
>> symbols (the nasty namespace-violating ones with the 64 suffix) rather
>> than the ordinary ones.

> Doesn't it changes the definition of off_t and ino_t as well?

I'm fairly sure you're right.

#ifndef __off_t_defined
# ifndef __USE_FILE_OFFSET64
typedef __off_t off_t;
# else
typedef __off64_t off_t;
# endif
# define __off_t_defined
#endif

> While these aren't supposed to be used in public header files, such a
> change seems fairly risky to me.

Yes, if the lengths of those types change, any software that uses off_t
and ino_t in on-disk binary data structures will find that a newly-built
version of that software will no longer be able to read the old data
files.  Now obviously that's a badly-designed disk format, shouldn't have
been done that way, etc., but I know that people have done that.

This is not specifically a glibc ABI change, but it would be quite
surprising.

I think one might also run into trouble when one rebuilds software with
this change that links with a shared library that predates the change and
uses off_t or ino_t in public ABIs, or vice versa.

-- 
Russ Allbery (eagle@eyrie.org)              <http://www.eyrie.org/~eagle/>