[Nauty] problem compiling with os x 10.4

[Josh Guffin]

Switching to gcc 3.3 solved the compilation issue... however,
the output still doesn't match the given output in the user's
guide... specifically,

on OS X.4, gcc 3.3;

----------------
/Users/guffin/Desktop/nauty22 % dreadnaut
Dreadnaut version 2.0beta (BSDUNIX).
 > c -a -m
 > n=12 g
0 : 1; 2; 0;
3 : 4; 5; 6; 3;
7 : 8; 9; 10; 11; 7.
 > x @
3 orbits; grpsize=480; 6 gens; 52 nodes (2 bad leaves); maxlev=7
tctotal=142; canupdates=3; cpu time = 0.00 seconds
 > g
0 : 1; 2; 3; 4; 0;
5 : 6; 7; 8; 5;
9 : 10; 11; 9.
 > x
3 orbits; grpsize=480; 6 gens; 34 nodes (2 bad leaves); maxlev=7
tctotal=80; canupdates=2; cpu time = 0.00 seconds
 > ##
h and h' are identical.
0-9 1-10 2-11 3-5 4-6 5-7 6-8 7-0 8-1 9-2 10-3 11-4
 >
----------------

while on a linux machine with gcc 2.96,

-----------------
home/theorist/guffin/nauty22 % ./dreadnaut
Dreadnaut version 2.2 (32 bits).
 > c -a -m
 > n=12 g
0 : 1; 2; 0; 4; 5;6;3;8;9;10;11;7.
 > x @
3 orbits; grpsize=480; 6 gens; 31 nodes (3 bad leaves); maxlev=7
tctotal=88; canupdates=1; cpu time = 0.00 seconds
 > g
0 : 1;2;3;4;0;
5 : 6;7;8;5;
9 : 10;11;9.
 > x
3 orbits; grpsize=480; 6 gens; 50 nodes (2 bad leaves); maxlev=7
tctotal=124; canupdates=4; cpu time = 0.00 seconds
 > ##
h and h' are identical.
0-9 1-10 2-11 3-5 4-6 5-7 6-8 7-0 8-1 9-2 10-3 11-4
 > q
--------------------

which matches what was given in the user guide.  Are the numbers
of nodes and bad leaves expected to be different on machines with
disparate processors?

Thanks for the quick response,

Josh



On 28 Jun, 2005, at 10:08 PM, Gordon Royle wrote:

> I can confirm this issue... I just tried it out on my Tiger  
> machine, and had the same problem..
>
> The issue is gcc 4.0 which appears to cause a number of problems.
>
> If you use gcc 3.3 then the problem will go away..  (use sudo  
> gcc_select 3.3 to make this the default)
>
> Cheers
>
> Gordon
>
>
> On 29/06/2005, at 9:37 AM, Josh Guffin wrote:
>
>> Hello, I've just tried to compile Nauty using gcc that comes with
>> OS X 10.4.  The program seems to compile alright, but 'make check'
>> fails, and running dreadtest yields "illegal instruction"
>