Re: make check failure for octave 2.9.6 under Mac OS X 10.4

[David Bateman]

Marius Schamschula wrote:

> Hi all,
>
> I did a test build of octave 2.9.6 under Mac OS X 10.4.6 with gcc
> 4.0.1 and g95. I built the UFsparse libraries to the satisfaction of
> the octave configure script (I wish these libraries had an install
> script in their Makefile). The build went all right. To be sure I ran
> make check and got:
>
>   test_args.m ............................................ PASS   10/10 
>   test_arith.m ........................................... PASS  197/197
>   test_contin.m .......................................... PASS   17/17 
>   test_diffeq.m .......................................... PASS   11/11 
>   test_error.m ........................................... PASS   10/10 
>   test_eval-catch.m ...................................... PASS    9/9  
>   test_eval.m ............................................ PASS   10/10 
>   test_for.m ............................................. PASS    9/9  
>   test_global.m .......................................... PASS   10/10 
>   test_if.m .............................................. PASS    6/6  
>   test_index-wfi-f.m ..................................... PASS   40/40 
>   test_index-wfi-t.m ..................................... PASS   39/39 
>   test_infnan.m .......................................... PASS    9/9  
>   test_io.m .............................................. PASS  
> 65/66   FAIL 1
>   test_linalg.m .......................................... PASS   66/66 
>   test_logical-wfi-f.m ................................... PASS   44/44 
>   test_logical-wfi-t.m ................................... PASS   44/44 
>   test_matrix.m .......................................... PASS   91/91 
>   test_nonlin.m .......................................... PASS    5/5  
>   test_number.m .......................................... PASS   52/52 
>   test_poly.m ............................................ PASS   51/51 
>   test_prefer.m .......................................... PASS   29/29 
>   test_quad.m ............................................ ABNORMAL
> RETURN FROM DQAGP
> PASS    7/7  
>   test_recursion.m ....................................... PASS    2/2  
>   test_return.m .......................................... PASS    2/2  
>   test_set.m ............................................. PASS   19/19 
>   test_signal.m .......................................... PASS    8/8  
>   test_sparse.m ..........................................panic:
> Segmentation fault -- stopping myself...
> make[2]: *** [check] Segmentation fault
> make[1]: *** [check] Error 2
> make: *** [check] Error 2
>

Did using the octave-forge CVS fix this seg-fault? If it didn't then I
need to track down what might be an endian-ness problem. In that case I
might ask you to run

cd test
more off
test('test_sparse.m','verbose',stdout)

and tell me exactly which test code immediately proceeds the seg-fault...

D.