/* sqrt function. PowerPC32 version.
- Copyright (C) 2007, 2011 Free Software Foundation, Inc.
+ Copyright (C) 2007-2016 Free Software Foundation, Inc.
This file is part of the GNU C Library.
The GNU C Library is free software; you can redistribute it and/or
/* double [fp1] sqrt (double x [fp1])
Power4 (ISA V2.0) and above implement sqrt in hardware (not optional).
The fsqrt instruction generates the correct value for all inputs and
- sets the appropriate floating point exceptions. Extented checking is
+ sets the appropriate floating point exceptions. Extended checking is
only needed to set errno (via __kernel_standard) if the input value
is negative.
-
+
So compare the input value against the absolute value of itself.
This will compare equal unless the value is negative (EDOM) or a NAN,
in which case we branch to the extend wrapper. If equal we can return
the result directly.
-
+
This part of the function looks like a leaf routine, so no need to
stack a frame or execute prologue/epilogue code. It is safe to
branch directly to w_sqrt as long as the input value (f1) is
preserved. Putting the sqrt result into f2 (float parameter 2)
allows passing both the input value and sqrt result into the extended
wrapper so there is no need to recompute.
-
+
This tactic avoids the overhead of stacking a frame for the normal
(non-error) case. Until gcc supports prologue shrink-wrapping
this is the best we can do. */