On Thu, 2003-11-27 at 09:19, Olaf Hering wrote:
> What should be used for sig_atomic_t in the future?
> klibc doesnt define it anymore. Maybe all archs have to define a
> sig_atomic_t?
> 
> In file included from udevdb.c:39:
> tdb/tdb.h:131: parse error before '*' token
> tdb/tdb.h:131: warning: function declaration isn't a prototype
> make: *** [udevdb.o] Error 1
> Bad exit status from /var/tmp/rpm-tmp.80116 (%build)
> olaf@mandarine:/usr/src/packages/SOURCES> grep -wr sig_atomic_t  ../BUILD/udev-007/
> ../BUILD/udev-007/klibc/klibc/arch/ia64/include/klibc/archsignal.h:typedef int sig_atomic_t;
> ../BUILD/udev-007/tdb/tdb.c:static sig_atomic_t *palarm_fired;
> ../BUILD/udev-007/tdb/tdb.c:void tdb_set_lock_alarm(sig_atomic_t *palarm)
> ../BUILD/udev-007/tdb/tdb.h:void tdb_set_lock_alarm(sig_atomic_t *palarm);
> ../BUILD/udev-007/klibc-0.87/klibc/arch/ia64/include/klibc/archsignal.h:typedef int sig_atomic_t;

I just submitted a patch recently that added it for ia64 (as seen in
your grep), which Peter Anvin took.  I just assumed that all other
arches already defined it somewhere.

Maybe this is something that should be in klibc's main header files,
instead of in arch/?

mh

-- 
Martin Hicks || mort@bork.org || PGP/GnuPG: 0x4C7F2BEE