Re: sqrt function, NaN and complex number

[Dmitri A. Sergatskov]

Dmitri A. Sergatskov wrote:
> John W. Eaton wrote:
>
> > In any case, I don't think it is really a bug in Octave.
> >
> > What happens if you try a simple C++ program to test sqrt(NaN+NaN*i)?
>
> That does seem to work OK:

I guess I spoke too soon. It appears that it breaks if I use
any optimization flag to g++:

address@hidden divzero]$ g++ -Wall -march=athlon-mp test4.cc -o test4
test4.cc: In function `int main()':
test4.cc:6: warning: division by zero in `1.0e+0 / 0.'
test4.cc:6: warning: division by zero in `-1.0e+0 / 0.'
address@hidden divzero]$ ./test4
z1: (inf,-inf)
z2: (nan,nan)
z3: (nan,nan)
address@hidden divzero]$ g++ -Wall -O -march=athlon-mp test4.cc -o test4
test4.cc: In function `int main()':
test4.cc:6: warning: division by zero in `1.0e+0 / 0.'
test4.cc:6: warning: division by zero in `-1.0e+0 / 0.'
address@hidden divzero]$ ./test4
(it hangs here)

(I am off to bugzilla.)

Sincerely,

Dmitri.
--



-------------------------------------------------------------
Octave is freely available under the terms of the GNU GPL.

Octave's home on the web:  http://www.octave.org
How to fund new projects:  http://www.octave.org/funding.html
Subscription information:  http://www.octave.org/archive.html
-------------------------------------------------------------