Solaris 9 now passes all 93 tests. Ben, I replaced the tests/expressions/expressions.sh file with the one you sent, and now that passes the compare test. And John, setting the CFLAGS to "-g -O0" seems to do the trick for the examine.sh issue--turning off optimization causes no Abort messages now.
Ben, I believe you also sent an attachment for the FC4 issue I mentioned--I'll try that out tonight. Also, Ben, you mentioned in one of your emails about my switching from the bash to sh shell to run the tests. I did do that, because it seemed I was having a problem with only 63 tests passing the first time. However, I believe it may have been due to the path pointing to the Solaris diff instead of the Gnu diff. When running the make check tests, I made sure to "path" to the Gnu diff, ran the tests with both shells, and neither shell caused any problems with the result. Marshall -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of [EMAIL PROTECTED] Sent: Tuesday, August 02, 2005 7:01 AM To: [EMAIL PROTECTED] Cc: [email protected]; [EMAIL PROTECTED] Subject: Re: release candidate 5 available--stack trace for Solaris Importance: Low On Mon, Aug 01, 2005 at 08:23:04PM -0700, Ben Pfaff wrote: > Well we're getting somewhere. The parameters for show_summary are clearly > rediculous. In my experience sometimes gdb just gets arguments wrong. It seems "difficult" to get them right in the presence of compiler optimizations. Occasionally I have seen that happen, but I don't think it's the case in this instance. My bet is that something is trashing the global variables sometime after they've been populated. But your point about optimisation is a valid one. Marshall, can you rebuild with CFLAGS="-g -O0" and try the stack trace again? J' -- 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
