Document fmal.

lib/msun/man/fma.3

@@ -29,7 +29,8 @@
 .Os
 .Sh NAME
 .Nm fma ,
-.Nm fmaf
+.Nm fmaf ,
+.Nm fmal
 .Nd fused multiply-add
 .Sh LIBRARY
 .Lb libm
@@ -39,11 +40,14 @@
 .Fn fma "double x" "double y" "double z"
 .Ft float
 .Fn fmaf "float x" "float y" "float z"
+.Ft long double
+.Fn fmal "long double x" "long double y" "long double z"
 .Sh DESCRIPTION
 The
-.Fn fma
+.Fn fma ,
+.Fn fmaf ,
 and
-.Fn fmaf
+.Fn fmal
 functions return
 .No "(x * y) + z" ,
 computed with only one rounding error.
@@ -62,23 +66,21 @@ accuracy of calculations such as dot products.
 It may also be used to improve performance on machines that implement
 it natively.
 The macros
-.Dv FP_FAST_FMA
-and
+.Dv FP_FAST_FMA ,
 .Dv FP_FAST_FMAF
+and
+.Dv FP_FAST_FMAL
 may be defined in
 .In math.h
 to indicate that
-.Fn fma
+.Fn fma ,
+.Fn fmaf ,
 and
-.Fn fmaf
+.Fn fmal
 (respectively) have comparable or faster speed than a multiply
 operation followed by an add operation.
 .Sh IMPLEMENTATION NOTES
-In general,
-.Fn fma
-and
-.Fn fmaf
-will behave as one would expect if
+In general, these routines will behave as one would expect if
 .No "x * y + z"
 were computed with unbounded precision and range,
 then rounded to the precision of the return type.
@@ -93,13 +95,22 @@ would have otherwise generated an invalid exception.
 .Xr math 3
 .Sh STANDARDS
 The
-.Fn fma
+.Fn fma ,
+.Fn fmaf ,
 and
-.Fn fmaf
+.Fn fmal
 functions conform to
 .St -isoC-99 .
 A fused multiply-add operation with virtually identical
 characteristics appears in IEEE draft standard 754R.
 .Sh HISTORY
-These routines first appeared in
-.Fx 5.4 .
+The
+.Fn fma
+and
+.Fn fmaf
+routines first appeared in
+.Fx 5.4 ,
+and
+.Fn fmal
+appeared in
+.Fx 5.5 .