Path: utzoo!utgpu!water!watmath!clyde!rutgers!sunybcs!bingvaxu!leah!itsgw!imagine!pawl21.pawl.rpi.edu!kyriazis
From: kyriazis@pawl21.pawl.rpi.edu (George Kyriazis)
Newsgroups: comp.lang.c
Subject: Re: Array problems.  [Was: reply to GEORGE KRYKARISS]
Message-ID: <371@imagine.PAWL.RPI.EDU>
Date: 16 Feb 88 03:44:00 GMT
References: <11829@brl-adm.ARPA>
Sender: news@imagine.PAWL.RPI.EDU
Reply-To: kyriazis@pawl21.pawl.rpi.edu (George Kyriazis)
Organization: RPI Public Access Workstation Lab - Troy, NY
Lines: 104

In article <11829@brl-adm.ARPA> kcfy@hp1.ccs.cornell.EDU (Oya Ekin) writes:
>
>GEORGE KYRIAZISSS writes ::
>

noooooo....  My name is George Kyriazis..  :-)


>>       Hello world.  I first found that peculiar behaviour trying to
>>  write a computer graphics program involving points, lines and polygons.  
>>  Anyway, here it is:  
>>
>> int	(*lines)[2];
>> main() {}
>>
>>is a dummy program that just defines this data structure.  What I actually
>>want is a pointer to arrays of 2 ints.  I believe that int *lines[2] will
>>do the reverse, ie. give me two pointers to int.
>>Anyway, when I run it thru dbx on a SUN and ask 'whatis lines' I get this:
>>
>> int (*lines)[8196];
>>
>>As far as I know, this is a *BIG* mistake?  Or am I wrong?
>>
>>
>     I do not quite understand how you are going to use a declaration like
>the one you gave. BUT You are right ... I compiled this on Vax it
>gives me    array[0..1] of (*lines) but I do not understand how you are
>going to use this even if it is declared correctly as in VAX. 
>
>MAYBE you need something like :
>
>int *lines ; 
>
>main()
>{
>       lines = calloc(2*n,sizeof(int)) ; 
>}
>
>to create arrays of 2 ints refernced as
>     for example for a point :
>     line[n]  for x   and line[n+1] for y ; 
>

You are right, this thing will work..  Even defining a structure like

struct lines {int e1,e2} *lines;

will work.  What I was thinking of doing was

lines=(int (*)[2])malloc(sizeof(int)*2*nelem);

so I can reference as lines[i][0] and lines[i][1].  (Don't worry about
the (int (*)[2]) casting...  It works..  I have no idea why, but it does..)
But here there is a problem between debuggers..  I compiled it to a vax,
an iris 3130 workstation and an Balance 21000 and here is what I got:

On the VAX:
csv.rpi.edu 10% dbx 
dbx version 3.21 of 6/5/86 16:40 (monet.Berkeley.EDU).
Type 'help' for help.
enter object file name (default is `a.out'): 
reading symbolic information ...
(dbx) whatis lines
array[0..1] of int *lines;
(dbx) 

On the iris:
iris.cs.rpi.edu 8% dbx
enter object file name (default is `a.out'): 
dbx version 3.99 of 12/19/87 21:33.
Type 'help' for help.
reading symbolic information ...
(dbx) whatis lines
array[0..4294967295] of int *lines;
(dbx) 

On the Sequent:
b21.cs.rpi.edu 8% dbx temp
dbx version 5.7 of 9/10/87 22:48 (Sequent).
Type 'help' for help.
reading symbolic information ... 
(dbx) whatis lines
array[0..4294967295] of int *lines;
(dbx) 

and finally on the SUN:
pawl10.pawl.rpi.edu 4% dbx temp
Reading symbolic information...
Read 51 symbols
(dbx) whatis lines
int (*lines)[8196];
(dbx) 

I have no problem changing my data structure, but do we have a ?major?
dbx problem here??  I don't understand..


*******************************************************
*George C. Kyriazis                                   *    Gravity is a myth
*userfe0e@mts.rpi.edu or userfe0e@rpitsmts.bitnet     *        \    /
*Electrical and Computer Systems Engineering Dept.    *         \  /
*Rensselear Polytechnic Institute, Troy, NY 12180     *          ||
*******************************************************      Earth sucks.