Cygwin Binary on Wiki has been updated

[Angelo Graziosi]

Jerry DeLisle wrote:
> Regression testing shows one unexpected failure of atan2_1.f90

I wrote:
> May this be a 'PR323 effect'?

Following [*], another example...

$ gfortran-4.5 -c set_math_double_precision.c
$ gfortran-4.5 test_atan2_fpu.f90 -o test_atan2_fpu 
set_math_double_precision.o

$ ./test_atan2_fpu.exe
  -4.50335449E-08
   4.03570652E-08
   2.59082231E-08
   2.39143301E-08
   4.87428657E-08
  -2.40507294E-08
  -1.97886663E-09
   1.57936935E-08
  -2.46570160E-08
   2.18556977E-08

Only the first 'print' (for single precision) 'fails'! I do not know 
something for *set_math_single_precision.c* :(

Attached all you need.

Ciao,
Angelo.
---
[*] http://gcc.gnu.org/bugzilla/show_bug.cgi?id=323#c60
-------------- next part --------------
A non-text attachment was scrubbed...
Name: test_atan2-PR323.tar.bz2
Type: application/octet-stream
Size: 2308 bytes
Desc: not available
URL: <http://gcc.gnu.org/pipermail/fortran/attachments/20100324/4d939725/attachment.obj>