Artsize still not quite right in CURRENT

[Russ Allbery]

Joe St Sauver <JOE at oregon.uoregon.edu> writes:

> Of course, I should add also that one can also manually #UNDEF the
> conditional code that's being triggered at date.c:66 and date.c:160,
> which also lets date.c compile cleanly.

> [As I continued to look at this, I note that there may be other areas
> where including -D_XOPEN_SOURCE=500 breaks code which is otherwise
> clean, e.g.:

> gcc -g -O2 -I../include  -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 
> -D_XOPEN_SOURCE=500 -c dbz.c
> dbz.c: In function `getcore':
> dbz.c:1299: `caddr_t' undeclared (first use in this function)
> dbz.c:1299: (Each undeclared identifier is reported only once
> dbz.c:1299: for each function it appears in.)
> dbz.c:1299: parse error before `0'

Yup, this is exactly why I dislike feature-test macros.  *sigh*

The hope is that _GNU_SOURCE won't do that.  Hm.  Okay, I agree that it's
probably the best workaround that we have; now I wonder if there's a good
way to put logic in configure.in so that it only adds it if it's
necessary.  I'm a little hesitant to just enable it on all Linux systems.

Maybe try to see if we can get a definition for pwrite without it and then
try again with it, similar to the existing NEEDS_DECLARATION stuff....

-- 
Russ Allbery (rra at stanford.edu)             <http://www.eyrie.org/~eagle/>