[Gcl-devel] Bug in mpz_get_d

[Torbjorn Granlund]

Enrique Perez-Terron <enrio at online.no> writes:

  On Fri, 2009-12-11 at 15:12 -0500, Camm Maguire wrote:
  > Greetings!  Workaround?
  > 
  > >(setq a (numerator **))
  > 
  > 10715086071862673209484250490600018105614048117055336074437503883703510511249361224931983788156958581275946729175531468251871452856923140435984577574698574803934567774824230985421074605062371141877954182153046474983581941267398767559165543946077062914571196477686542167660429831652624386837205668069375
  > 
  > >(float a)
  > 
  > 1.0715086071862672E301
  > 
  > >(float (1+ a))
  > 
  > 1.0715086071862673E301
  > 
  > >(float (1- a))
  > 
  > 1.0715086071862672E301
  
This (LISP) session does not tell me anything, sorry.

  Testing GMP and MPFR directly
  $ ./test-gmp 
  Using GMP and mpf_get_d:
  ================
  num = 10715086071862673209484250490600018105614048117055336074437503883703510511249361224931983788156958581275946729175531468251871452856923140435984577574698574803934567774824230985421074605062371141877954182153046474983581941267398767559165543946077062914571196477686542167660429831652624386837205668069375
  1.071508607186267202E+301    
  
  1 + num = 10715086071862673209484250490600018105614048117055336074437503883703510511249361224931983788156958581275946729175531468251871452856923140435984577574698574803934567774824230985421074605062371141877954182153046474983581941267398767559165543946077062914571196477686542167660429831652624386837205668069376
  1.0715086071862673209E+301   
  
  1 - num = -10715086071862673209484250490600018105614048117055336074437503883703510511249361224931983788156958581275946729175531468251871452856923140435984577574698574803934567774824230985421074605062371141877954182153046474983581941267398767559165543946077062914571196477686542167660429831652624386837205668069374
  -1.071508607186267202E+301   
  
OK, some numbers...  Anything wrong there?

  Using MPFR:

THis is the wrong list for mpfr bugs.

Please explain what you get and what you expect.  *One* example of
incorrect+correct data is what we need.

But we need more information than that.  Please read

http://gmplib.org/manual/Reporting-Bugs.html#Reporting-Bugs

and send a bug report upon which we have any chance to act!

-- 
Torbjörn