[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: 19990907: udunits-1.11.7 on NEC SX-4

This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.

  • Subject: Re: 19990907: udunits-1.11.7 on NEC SX-4
  • Date: Tue, 07 Sep 1999 09:26:56 -0600
Len,

> cc: address@hidden,
> cc: address@hidden
> From: "Len Makin, CSIRO Maths & Info Sci, Melbourne." <address@hidden>
> Subject: udunits-1.11.7 on NEC SX-4
> Organization: CSIRO
> Keywords: 199909060545.XAA11766 UDUNITS NEC SX-4

In the above message, you wrote:

>       Have just installed the above (passed all tests) with environment:
>       OS is SUPER-UX unix 9.1 Rev1
>       C compiler is C/SX V2 Version 1.0 (Rev.4.01)
>       Fortran compiler is Fortran90/SX Version 1.0 (Rev 160)
>       perl is version 5.004_04
> 
>       CPPFLAGS= -DDOUBLE_PRECISION=double
>       CFLAGS=  -h2 -Xc -hacct -h stack=nan -pvctl,fullmsg
>       FFLAGS= F90FLAGS= -ew 
>       FLMOD=0

Congratulations!  What was the setting for the CC environment variable?

> The following problems were dealt with:
> 
> 1. src/lib/utlib.c had function calls as lvalues (non-ANSI)
> line 958 now
>            (void)strncpy(name, cp, n); name[n] = '\0';
> was
>            strncpy(name, cp, n)[n]     = 0;
> line 963 now
>            (void)strncpy(buf, cp, n); buf[n]   = '\0';
> was
>            strncpy(buf, cp, n)[n]      = 0;

I believe that the original statements were strictly conforming and that
it's likely that the C/SX compiler is non-conforming in this respect
if it can't handle them (see section 3.2.2.1 in ANSI X3.159-1989).
I've found nothing in the Standard that contradicts this.  If you know
of something in the Standard that does, would you please send me a
reference.

> 2. Somehow the configure and src/port/misc/config procedure stuffed up
> the header creation in src/port/misc . They all resulted in headers like
> for example:
> #define UD_SYSTEM_SEARCH_H      "/dev/null"
> which resulted in compile problems due to definitions not being found.
> I changed them all back to *.h.in formulation:
> #define UD_SYSTEM_SEARCH_H      "///usr/include/search.h"
> 
> and all compiled and tested correctly. Not sure how to correct this more 
> portably.

Interesting.  I'd like to investigate this.  Would you please send me
the files "configure.log" (the standard output and standard error of the
configure script) and "config.log".

> Regards,
>       Len Makin
> |address@hidden     | CSIRO Mathematics & Information Sciences |
> |          Manager, User Services, CSIRO/Bureau of Meteorology          |
> |          High Performance Computing and Communications Centre         |
> |          24th floor, 150 Lonsdale St, Melbourne, VIC 3000, Australia  |
> | Telephone: +61 3 9669 8109 |      Fax: +61 3 9669 8112                |

--------
Steve Emmerson   <http://www.unidata.ucar.edu>