cp from fdlibm and pass the cft after refined
Signed-off-by: rander <[email protected]>
---
backend/src/libocl/tmpl/ocl_math_common.tmpl.cl | 34 +++++++++++++++++++++++++
backend/src/libocl/tmpl/ocl_math_common.tmpl.h | 1 +
2 files changed, 35 insertions(+)
diff --git a/backend/src/libocl/tmpl/ocl_math_common.tmpl.cl
b/backend/src/libocl/tmpl/ocl_math_common.tmpl.cl
index a705e2f..c370864 100644
--- a/backend/src/libocl/tmpl/ocl_math_common.tmpl.cl
+++ b/backend/src/libocl/tmpl/ocl_math_common.tmpl.cl
@@ -1999,6 +1999,40 @@ OVERLOADABLE double tan(double x)
}
}
+OVERLOADABLE double tanh(double x)
+{
+ double one=1.0, two=2.0, tiny = 1.0e-300;
+ double t,z;
+ int jx,ix;
+
+ /* High word of |x|. */
+ jx = __HI(x);
+ ix = jx&0x7fffffff;
+
+ /* x is INF or NaN */
+ if(ix>=0x7ff00000) {
+ if (jx>=0) return one/x+one; /* tanh(+-inf)=+-1 */
+ else return one/x-one; /* tanh(NaN) = NaN */
+ }
+
+ /* |x| < 22 */
+ if (ix < 0x40360000) { /* |x|<22 */
+ if (ix<0x3c800000) /* |x|<2**-55 */
+ return x*(one+x); /* tanh(small) = small */
+ if (ix>=0x3ff00000) { /* |x|>=1 */
+ t = expm1(two*fabs(x));
+ z = one - two/(t+two);
+ } else {
+ t = expm1(-two*fabs(x));
+ z= -t/(t+two);
+ }
+ /* |x| > 22, return +-1 */
+ } else {
+ z = one - tiny; /* raised inexact flag */
+ }
+ return (jx>=0)? z: -z;
+}
+
OVERLOADABLE double trunc(double x)
{
double ret = floor(fabs(x));
diff --git a/backend/src/libocl/tmpl/ocl_math_common.tmpl.h
b/backend/src/libocl/tmpl/ocl_math_common.tmpl.h
index 9ac3b31..471f866 100644
--- a/backend/src/libocl/tmpl/ocl_math_common.tmpl.h
+++ b/backend/src/libocl/tmpl/ocl_math_common.tmpl.h
@@ -58,6 +58,7 @@ OVERLOADABLE double sin(double x);
OVERLOADABLE double sinh(double x);
OVERLOADABLE double sqrt(double x);
OVERLOADABLE double tan(double x);
+OVERLOADABLE double tanh(double x);
OVERLOADABLE double trunc(double x);
--
2.7.4
_______________________________________________
Beignet mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/beignet
