[Nauty-list] Building nauty on MinGW, missing sys/times.h

William August Rummler w.a.rummler at gmail.com
Tue Sep 11 08:26:09 EST 2007


Inserting those directives worked to dodge the cputime problem, and 
compilation succeeded for all targets but shortg, as I suppose could be 
expected from a non-Unix system. I then ran "make -k checks" which went 
through all tests giving this kind of error for each one:

rm dreadtest__
rm: cannot unlink `dreadtest__': No such file or directory
make: *** [dreadtest__] Error 1

... where, for each test in the sequence, the __ is replaced by B, S, 
4K, 1, W1, L1, and L, respectively. Frankly, I can't say if this is a 
big issue, as my dreadnaut appears to be working fine. Just reporting an 
apparent problem.

- Will

Brendan McKay wrote:
> I don't know if MinGW has the facilities for measuring cputime. If
> you find them, I can try to incorporate them. Meanwhile, at least we
> can avoid the compile-time error.
>
> In naututil-h.in, before the line
>     #endif /*NAUTY_CPU_DEFINED*/
> (about line 224), insert this:
>
> #ifndef CPUTIME
> #define CPUTIME 0.0
> #endif
>
> Then do ./configure again and try compiling.  All cpu times will
> be reported as 0, but that is better than nothing.
>
> Please report problems.
>
> Brendan.
>
> * William August Rummler <w.a.rummler at gmail.com> [070910 16:26]:
>   
>> Hello to all,
>>
>> I recently tried to build nauty with MinGW, using just "./configure" and 
>> "make all". After some successful progress, I received this error:
>>
>> gcc -o labelg -O4 -march=i686 labelg.c gtools.o gtnauty.o \
>>        nauty.o nautil.o naugraph.o nautinv.o
>> labelg.c: In function `main':
>> labelg.c:218: error: `CPUTIME' undeclared (first use in this function)
>> labelg.c:218: error: (Each undeclared identifier is reported only once
>> labelg.c:218: error: for each function it appears in.)
>> make: *** [labelg] Error 1
>>
>> I tracked the cause (or at least part of the cause) to the fact that 
>> MinGW lacks the sys/times.h file and the times() function located therein.
>>
>> Does anyone know what to do or where to begin to resolve this issue?
>>
>> Thanks,
>> Will
>>
>>
>> _______________________________________________
>> Nauty-list mailing list
>> Nauty-list at cs.anu.edu.au
>> http://dcsmail.anu.edu.au/cgi-bin/mailman/listinfo/nauty-list
>>     
>
>   

-- 
_ William August Rummler
_ Student, VCSK-6
_ Department of Computer Science
_ Rochester Institute of Technology
_ Email: w.a.rummler at gmail.com
_ Phone: 225.933.0195





More information about the Nauty mailing list