home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!elroy.jpl.nasa.gov!news.claremont.edu!ucivax!orion.oac.uci.edu!news.service.uci.edu!hardy
- From: hardy@golem.ps.uci.edu (Meinhard E. Mayer (Hardy))
- Subject: Re: that HPUX cc compiler
- Nntp-Posting-Host: golem.ps.uci.edu
- Message-ID: <HARDY.92Aug31112334@golem.ps.uci.edu>
- In-reply-to: als@bohra.cpg.oz.au's message of 31 Aug 92 10:49:50 GMT
- Newsgroups: comp.sys.hp
- Organization: Department of Physics, UC Irvine, CA 92717-4575, USA
- Lines: 24
- References: <1992Aug31.104950.1108@bohra.cpg.oz.au>
- Date: 31 Aug 92 18:23:37 GMT
-
- In article <1992Aug31.104950.1108@bohra.cpg.oz.au> als@bohra.cpg.oz.au (Anthony Shipman) writes:
-
-
- Here are some programs for which the HP cc generates incorrect code.
-
- xfig.2.1.4, contact bvsmith@lbl.gov
-
- When compiled with -O incorrect code is generated for the
- int_spline_bound() function. This results in floating point exceptions
- while working with splines. The program works without -O.
-
- .....more of the same
-
- I found that most programs compile perfectly with the +O1 option,
- quite a few with +O3 (they take quite a while -- you may have given up
- too early), unless I run out of memory. I have put +O1 as the default
- in my hp.cf to use with xmkmf (also, some of the newer X-stuff will
- not link unless I include both the X11R4 ann X11R5 include files.
- --
- Hardy
- -----
- Meinhard E. Mayer, Department of Physics, UC Irvine
- e-mail: hardy@golem.ps.uci.edu (preferred) or MMAYER@UCI.BITNET
-
-