sf_atan, sf_atan2 in mathfp won't compile when double is 32bits

[Ran Cabell]

sf_atan and sf_atan2 will not compile for targets where double is
32bits. In both cases it appears to be a simple typo; possible patches
below:


--- sf_atan2.c  Tue Nov 10 21:57:35 1998
+++ ../../../../newlib-1.8.2/newlib/libm/mathfp/sf_atan2.c      Thu Feb 
3 23:31:14 2000
@@ -26,7 +26,7 @@
 }
 
 #ifdef _DOUBLE_IS_32BITS
-double atan2f (double v, double u)
+double atan2 (double v, double u)
 {
   return (double) atangentf (0.0, (float) v, (float) u, 1);
 }



--- sf_atan.c   Tue Nov 10 21:57:35 1998
+++ ../../../../newlib-1.8.2/newlib/libm/mathfp/sf_atan.c       Thu Feb 
3 23:33:52 2000
@@ -39,7 +39,7 @@
 #ifdef _DOUBLE_IS_32BITS
 double atan (double x)
 {
-  return (double) atangentf ((float) x, 0);
+  return (double) atangentf ((float) x, 0, 0, 0);
 }
 
 #endif /* defined(_DOUBLE_IS_32BITS) */


==================

Ran Cabell
rcabell@norfolk.infi.net