home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.unix.ultrix
- Path: sparky!uunet!mcsun!sun4nl!alchemy!ruunfs!hooft
- From: hooft@fys.ruu.nl (Rob Hooft)
- Subject: Re: DEC Fortran rtl error
- Message-ID: <1992Aug18.082102.4903@fys.ruu.nl>
- Organization: Physics Department, University of Utrecht, The Netherlands
- References: <1992Aug17.181432.7946@usho93.hou281.chevron.com> <6096@lhdsy1.lahabra.chevron.com>
- Date: Tue, 18 Aug 1992 08:21:02 GMT
- Lines: 24
-
- In <6096@lhdsy1.lahabra.chevron.com> wnw@lhdx17.NoSubdomain.NoDomain (Bill Wally) writes:
-
- >In article <1992Aug17.181432.7946@usho93.hou281.chevron.com>, hfwpa@usho43.hou281.chevron.com (F.W.Patrick) writes:
- >|>
- >|> We have been attempting a to link Fortran code compiled with
- >|> DEC Fortran for RISC ULTRIX 3.1 with some Fortran libraries
- >|> obtained from another shop. After a successful link the
- >|> following msg comes out ot the Fortran run time library:
- >|>
- >|> forrtl: severe: internal consistency check failure
- >|>
-
- >Sometimes turning off compiler optimization fixes these kinds of problems. We
- >have had similar ones with Oracle Fortran access, and now NEVER use optimization
- >in those subroutines that access the database (this is not a performance problem,
- >since the routines are mostly SQL anyway).
-
- Another fix: upgrade to V3.2 or even better V3.2.2. They are much better than
- the 3.1 release, of which I have seen such messages repeatedly. Optimization is
- still not without flaws, though.
- --
- Rob Hooft, Bijvoet Center for Biomolecular Research,
- Chemistry department University of Utrecht, the Netherlands
- hooft@hutruu54.bitnet hooft@chem.ruu.nl hooft@fys.ruu.nl hooft@cc.ruu.nl
-