Re: UPS 3.37-beta5 available - feedback with fortran on linux

From: Ian Edwards (ian_at_concerto.demon.co.uk)
Date: Thu Apr 04 2002 - 18:11:35 BST

  • Next message: Tom Hughes: "Re: request for feature"
    Hello Satish,
    
    On 04-Apr-02 Satish Balay wrote:
    > The patch works!
    
    Good.
    
    > You might also want the following change in ups/ao_symscan.c
    > 
    >         if (strcmp(suf, "f90") == 0)
    > to
    >         if (strcmp(suf, "f90") == 0 strcmp(suf, "F90") == 0)
    
    I wondered about that.
    
    > Also - gcc 3.04 gives the following warnings at linktime:
    >>>>
    > Linking ups with gcc -pipe -g -O2 -DHAVE_CONFIG_H -I. -I/usr/X11R6/include
    > -Idevelhdrs -L/usr/X11R6/lib -lSM -lICE -lX11 ...
    > xc_builtins.o: In function `builtin_qsort':
    > /home/balay/ups-3.37-beta5/ups/xc_builtins.c:165: the use of `tmpnam' is
    > dangerous, better use `mkstemp'
    > xc_builtins.o: In function `builtin_ioctl':
    > /home/balay/ups-3.37-beta5/ups/xc_builtins.c:257: the use of `mktemp' is
    > dangerous, better use `mkstemp'
    >>>>
    
    We just have to live with these warnings.  Ups does not itself call either
    of those routines but links them in as part of the built-in 'C' interpreter.
    There seems to be no flag to GCC to tell it not to complain.
    
    Regards,
    Ian.
    
    
    
     -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
     Ian Edwards <ian_at_concerto.demon.co.uk>       04-Apr-02  18:11:35
    
     Early Music Record Labels    - http://www.recordlabels.org/
     'ups' debugger for C/C++/f77 - http://www.concerto.demon.co.uk/UPS/
     -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
    


    This archive was generated by hypermail 2.1.4 : Thu May 23 2002 - 15:54:12 BST