[Beignet] [PATCH] backend: add double support to fmax

[rander]

Signed-off-by: rander <rander.wang at intel.com>
---
 backend/src/libocl/tmpl/ocl_math.tmpl.cl    | 15 +++++++++++++++
 backend/src/libocl/tmpl/ocl_math.tmpl.h     |  2 ++
 backend/src/libocl/tmpl/ocl_math_20.tmpl.cl | 11 +++++++++++
 backend/src/libocl/tmpl/ocl_math_20.tmpl.h  |  2 ++
 4 files changed, 30 insertions(+)

diff --git a/backend/src/libocl/tmpl/ocl_math.tmpl.cl b/backend/src/libocl/tmpl/ocl_math.tmpl.cl
index d526d6c..6978c92 100644
--- a/backend/src/libocl/tmpl/ocl_math.tmpl.cl
+++ b/backend/src/libocl/tmpl/ocl_math.tmpl.cl
@@ -4333,3 +4333,18 @@ OVERLOADABLE double mad(double a, double b, double c)
 	return __gen_ocl_mad(a, b, c);
 }
 
+OVERLOADABLE double fmax(double a, double b)
+{
+	ulong ua = as_ulong(a);
+	ulong ub =as_ulong(b);
+
+	if((ua & DF_ABS_MASK) > DF_MAX_NORMAL) return b;
+	if((ub & DF_ABS_MASK) > DF_MAX_NORMAL) return a;
+	if(ua == DF_POSITIVE_INF) return a;
+	if(ub == DF_POSITIVE_INF) return b;
+
+	double c = a - b;
+	return (c >= 0) ? a:b;
+}
+
+
diff --git a/backend/src/libocl/tmpl/ocl_math.tmpl.h b/backend/src/libocl/tmpl/ocl_math.tmpl.h
index 852bf47..1bf9b9a 100644
--- a/backend/src/libocl/tmpl/ocl_math.tmpl.h
+++ b/backend/src/libocl/tmpl/ocl_math.tmpl.h
@@ -236,6 +236,7 @@ OVERLOADABLE float half_tan(float x);
 OVERLOADABLE double ceil(double x);
 OVERLOADABLE double fabs(double x);
 OVERLOADABLE double floor(double x);
+OVERLOADABLE double fmax(double a, double b);
 OVERLOADABLE double fract(double x, global double *p);
 OVERLOADABLE double fract(double x, local double *p);
 OVERLOADABLE double fract(double x, private double *p);
@@ -248,3 +249,4 @@ OVERLOADABLE int ilogb(double x);
 OVERLOADABLE double mad(double a, double b, double c);
 
 
+
diff --git a/backend/src/libocl/tmpl/ocl_math_20.tmpl.cl b/backend/src/libocl/tmpl/ocl_math_20.tmpl.cl
index f459619..d1e8305 100644
--- a/backend/src/libocl/tmpl/ocl_math_20.tmpl.cl
+++ b/backend/src/libocl/tmpl/ocl_math_20.tmpl.cl
@@ -4210,6 +4210,17 @@ OVERLOADABLE double mad(double a, double b, double c)
 	return __gen_ocl_mad(a, b, c);
 }
 
+OVERLOADABLE double fmax(double a, double b)
+{
+	ulong ua = as_ulong(a);
+	ulong ub =as_ulong(b);
 
+	if((ua & DF_ABS_MASK) > DF_MAX_NORMAL) return b;
+	if((ub & DF_ABS_MASK) > DF_MAX_NORMAL) return a;
+	if(ua == DF_POSITIVE_INF) return a;
+	if(ub == DF_POSITIVE_INF) return b;
 
+	double c = a - b;
+	return (c >= 0) ? a:b;
+}
 
diff --git a/backend/src/libocl/tmpl/ocl_math_20.tmpl.h b/backend/src/libocl/tmpl/ocl_math_20.tmpl.h
index ee313af..10f0bcc 100644
--- a/backend/src/libocl/tmpl/ocl_math_20.tmpl.h
+++ b/backend/src/libocl/tmpl/ocl_math_20.tmpl.h
@@ -212,6 +212,7 @@ OVERLOADABLE float half_tan(float x);
 //------- double -----------
 OVERLOADABLE double ceil(double x);
 OVERLOADABLE double fabs(double x);
+OVERLOADABLE double fmax(double a, double b);
 OVERLOADABLE double floor(double x);
 OVERLOADABLE double fract(double x, global double *p);
 OVERLOADABLE double fract(double x, local double *p);
@@ -224,3 +225,4 @@ OVERLOADABLE double logb(double x);
 OVERLOADABLE int ilogb(double x);
 OVERLOADABLE double mad(double a, double b, double c);
 
+
-- 
2.7.4