Ok, an update on compiling under Solaris: Doing a ./configure CFLAGS=-ansi takes care of the union/structure that was present in casefile.c issue. Thanks for that tip, John.
A new issue pops up with cmdline.c, namely the use of <getopt.h>. Solaris doesn't have the GNU getopt.h "package"--so, I had to set up the GNU libgetop and "getopt.h" in this file for this to compile ok. I've gotten a executable pspp file now, but when I do a make check I now get the following for each of the tests: Making check in tests make[1]: Entering directory `/export/home/mdb/srcfiles/pspp-0.4.0/tests' /usr/local/bin/make check-TESTS make[2]: Entering directory `/export/home/mdb/srcfiles/pspp-0.4.0/tests' ./command/aggregate.sh: STAT_CONFIG_PATH=/export/home/mdb/srcfiles/pspp-0.4.0/config: is not an identifier FAIL: command/aggregate.sh Marshall -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: Sunday, July 24, 2005 8:19 PM To: [EMAIL PROTECTED] Cc: [email protected] Subject: Re: Solaris 9: Progress and new issue for pspp 0.4.0rc3 Importance: Low On second thoughts it might be the __STDC__ that's the problem. Try compiling without the -ansi flag. On Sat, Jul 23, 2005 at 09:48:13PM +0800, John Darrington wrote: #if __STDC__ - 0 == 0 && !defined(_NO_LONGLONG) typedef long long longlong_t; typedef unsigned long long u_longlong_t; #else /* used to reserve space and generate alignment */ typedef union { double _d; int32_t _l[2]; } longlong_t; typedef union { double _d; uint32_t _l[2]; } u_longlong_t; #endif /* __STDC__ - 0 == 0 && !defined(_NO_LONGLONG) */ So it looks like _NO_LONGLONG is getting defined somewhere, when it shouldn't be. Try #undefing it and see what happens. -- PGP Public key ID: 1024D/2DE827B3 fingerprint = 8797 A26D 0854 2EAB 0285 A290 8A67 719C 2DE8 27B3 See http://pgp.mit.edu or any PGP keyserver for public key. _______________________________________________ pspp-dev mailing list [email protected] http://lists.gnu.org/mailman/listinfo/pspp-dev
