Re: test failure for mappers.cc

[Michael D Godfrey]

On 11/09/2010 09:42 AM, John W. Eaton wrote:

I'm not seeing the difference, but it is probably happening because I
changed the float version of arg from

  float arg (float x) { return atan2 (0.0, x); }

to

  float arg (float x) { return atan2f (0.0f, x); }

It looks like your result is now off by about "2*eps ('single')".
Does adding that as a tolerance avoid the warning?  Even if it does,
should we make the change?  Shouldn't the atan2f function in your math
library recognize -1 as a special case and return precisely the single
precision value for pi?  Should we be tweaking test tolerances to hide
what might be considered bugs in system libraries?

jwe

I think that any "hiding" of such errors is not good.  If it is true that this
atan2f is returning a value that is approx 2*eps less than pi there is likely
a value for the argument  close to -1 which produces a larger
value, i.e. the approach to pi is not uniform.  This would be a serious error.

It would be good to verify this behavior and report the problem if necessary.

Michael