COMPMID-1758: NEON: Implement Range

Change-Id: I56dff9462b85760fbed6db43224cadb90d283810
Reviewed-on: https://review.mlplatform.org/472
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>

9 files changed, 514 insertions, 138 deletions

diff --git a/arm_compute/core/CL/kernels/CLRangeKernel.h b/arm_compute/core/CL/kernels/CLRangeKernel.h
index 2da21175ce..2349b8ecad 100644
--- a/arm_compute/core/CL/kernels/CLRangeKernel.h
+++ b/arm_compute/core/CL/kernels/CLRangeKernel.h
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018 ARM Limited.
+ * Copyright (c) 2018-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -31,7 +31,7 @@ namespace arm_compute
 {
 class ICLTensor;
 
-/** Kernel class for Range()
+/** Kernel class for Range
  *
  * range generates a 1-D tensor containing a sequence of numbers that begins at 'start' and extends by increments
  * of 'step' up to but not including 'end'.
@@ -51,7 +51,7 @@ public:
     CLRangeKernel &operator=(CLRangeKernel &&) = default;
     /** Default destructor */
     ~CLRangeKernel() = default;
-    /** Initialise the kernel's output tensor, start, end and step of the sequence.
+    /** Initialize the kernel's output tensor, start, end and step of the sequence.
      *
      * @param[out] output Output tensor. Data types supported: U8/S8/QASYMM8/U16/S16/U32/S32/F16/F32.
      * @param[in]  start  The starting value of the sequence.
diff --git a/arm_compute/core/NEON/NEKernels.h b/arm_compute/core/NEON/NEKernels.h
index a32c507266..a99b7d08e1 100644
--- a/arm_compute/core/NEON/NEKernels.h
+++ b/arm_compute/core/NEON/NEKernels.h
@@ -105,6 +105,7 @@
 #include "arm_compute/core/NEON/kernels/NEPriorBoxLayerKernel.h"
 #include "arm_compute/core/NEON/kernels/NEQuantizationLayerKernel.h"
 #include "arm_compute/core/NEON/kernels/NEROIPoolingLayerKernel.h"
+#include "arm_compute/core/NEON/kernels/NERangeKernel.h"
 #include "arm_compute/core/NEON/kernels/NEReductionOperationKernel.h"
 #include "arm_compute/core/NEON/kernels/NERemapKernel.h"
 #include "arm_compute/core/NEON/kernels/NEReorgLayerKernel.h"
diff --git a/arm_compute/core/NEON/kernels/NERangeKernel.h b/arm_compute/core/NEON/kernels/NERangeKernel.h
new file mode 100644
index 0000000000..eeacf3f8f5
--- /dev/null
+++ b/arm_compute/core/NEON/kernels/NERangeKernel.h
@@ -0,0 +1,90 @@
+/*
+ * Copyright (c) 2018-2019 ARM Limited.
+ *
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#ifndef __ARM_COMPUTE_NERANGEKERNEL_H__
+#define __ARM_COMPUTE_NERANGEKERNEL_H__
+
+#include "arm_compute/core/NEON/INEKernel.h"
+#include "arm_compute/core/Types.h"
+
+namespace arm_compute
+{
+class ITensor;
+
+/** Kernel class for Range
+ *
+ * range generates a 1-D tensor containing a sequence of numbers that begins at 'start' and extends by increments
+ * of 'step' up to but not including 'end'.
+ */
+class NERangeKernel : public INEKernel
+{
+public:
+    const char *name() const override
+    {
+        return "NERangeKernel";
+    }
+    /** Default constructor */
+    NERangeKernel();
+    /** Prevent instances of this class from being copied (As this class contains pointers) */
+    NERangeKernel(const NERangeKernel &) = delete;
+    /** Prevent instances of this class from being copied (As this class contains pointers) */
+    NERangeKernel &operator=(const NERangeKernel &) = delete;
+    /** Allow instances of this class to be moved */
+    NERangeKernel(NERangeKernel &&) = default;
+    /** Allow instances of this class to be moved */
+    NERangeKernel &operator=(NERangeKernel &&) = default;
+    /** Default destructor */
+    ~NERangeKernel() = default;
+    /** Initialize the kernel's output tensor, start, end and step of the sequence.
+     *
+     * @param[out] output Output tensor. Data types supported: U8/S8/U16/S16/U32/S32/F16/F32.
+     * @param[in]  start  The starting value of the sequence.
+     * @param[in]  end    The ending (not including) value of the sequence.
+     * @param[in]  step   The gap between each pair of values in the sequence.
+     */
+    void configure(ITensor *output, float start, float end, float step);
+    /** Static function to check if given info will lead to a valid configuration of @ref NERangeKernel
+     *
+     * @param[in] output Output tensor info. Data types supported: U8/S8/U16/S16/U32/S32/F16/F32.
+     * @param[in] start  The starting value of the sequence.
+     * @param[in] end    The ending (not including) value of the sequence.
+     * @param[in] step   The gap between each pair of values in the sequence.
+     *
+     * @return a status
+     */
+    static Status validate(const ITensorInfo *output, float start, float end, float step);
+
+    // Inherited methods overridden:
+    void run(const Window &window, const ThreadInfo &info) override;
+
+private:
+    using RangeFunction = void(ITensor *output, float start, float step, const Window &window);
+
+    RangeFunction *_func;   /**< Range function to be called */
+    float          _start;  /**< Start of sequence */
+    float          _end;    /**< End of sequence */
+    float          _step;   /**< Increment/step value */
+    ITensor       *_output; /**< Destination tensor */
+};
+} // namespace arm_compute
+#endif /* __ARM_COMPUTE_NERANGEKERNEL_H__ */
diff --git a/arm_compute/core/NEON/wrapper/intrinsics/getlane.h b/arm_compute/core/NEON/wrapper/intrinsics/getlane.h
index 107ce44e0c..68267ba92a 100644
--- a/arm_compute/core/NEON/wrapper/intrinsics/getlane.h
+++ b/arm_compute/core/NEON/wrapper/intrinsics/getlane.h
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018 ARM Limited.
+ * Copyright (c) 2018-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -30,62 +30,62 @@ namespace arm_compute
 {
 namespace wrapper
 {
-#define VGETLANE_IMPL_8(stype, vtype, postfix)                \
-    inline stype vgetlane(const vtype vector, const int lane) \
-    {                                                         \
-        switch(lane)                                          \
-        {                                                     \
-            case 0:                                           \
-                return vget_lane_##postfix(vector, 0);        \
-            case 1:                                           \
-                return vget_lane_##postfix(vector, 1);        \
-            case 2:                                           \
-                return vget_lane_##postfix(vector, 2);        \
-            case 3:                                           \
-                return vget_lane_##postfix(vector, 3);        \
-            case 4:                                           \
-                return vget_lane_##postfix(vector, 4);        \
-            case 5:                                           \
-                return vget_lane_##postfix(vector, 5);        \
-            case 6:                                           \
-                return vget_lane_##postfix(vector, 6);        \
-            case 7:                                           \
-                return vget_lane_##postfix(vector, 7);        \
-            default:                                          \
-                ARM_COMPUTE_ERROR("Invalid lane");            \
-        }                                                     \
+#define VGETLANE_IMPL_8(stype, vtype, postfix)                         \
+    inline stype vgetlane(const vtype vector, const unsigned int lane) \
+    {                                                                  \
+        switch(lane)                                                   \
+        {                                                              \
+            case 0:                                                    \
+                return vget_lane_##postfix(vector, 0);                 \
+            case 1:                                                    \
+                return vget_lane_##postfix(vector, 1);                 \
+            case 2:                                                    \
+                return vget_lane_##postfix(vector, 2);                 \
+            case 3:                                                    \
+                return vget_lane_##postfix(vector, 3);                 \
+            case 4:                                                    \
+                return vget_lane_##postfix(vector, 4);                 \
+            case 5:                                                    \
+                return vget_lane_##postfix(vector, 5);                 \
+            case 6:                                                    \
+                return vget_lane_##postfix(vector, 6);                 \
+            case 7:                                                    \
+                return vget_lane_##postfix(vector, 7);                 \
+            default:                                                   \
+                ARM_COMPUTE_ERROR("Invalid lane");                     \
+        }                                                              \
     }
 
-#define VGETLANE_IMPL_4(stype, vtype, postfix)                \
-    inline stype vgetlane(const vtype vector, const int lane) \
-    {                                                         \
-        switch(lane)                                          \
-        {                                                     \
-            case 0:                                           \
-                return vget_lane_##postfix(vector, 0);        \
-            case 1:                                           \
-                return vget_lane_##postfix(vector, 1);        \
-            case 2:                                           \
-                return vget_lane_##postfix(vector, 2);        \
-            case 3:                                           \
-                return vget_lane_##postfix(vector, 3);        \
-            default:                                          \
-                ARM_COMPUTE_ERROR("Invalid lane");            \
-        }                                                     \
+#define VGETLANE_IMPL_4(stype, vtype, postfix)                         \
+    inline stype vgetlane(const vtype vector, const unsigned int lane) \
+    {                                                                  \
+        switch(lane)                                                   \
+        {                                                              \
+            case 0:                                                    \
+                return vget_lane_##postfix(vector, 0);                 \
+            case 1:                                                    \
+                return vget_lane_##postfix(vector, 1);                 \
+            case 2:                                                    \
+                return vget_lane_##postfix(vector, 2);                 \
+            case 3:                                                    \
+                return vget_lane_##postfix(vector, 3);                 \
+            default:                                                   \
+                ARM_COMPUTE_ERROR("Invalid lane");                     \
+        }                                                              \
     }
 
-#define VGETLANE_IMPL_2(stype, vtype, postfix)                \
-    inline stype vgetlane(const vtype vector, const int lane) \
-    {                                                         \
-        switch(lane)                                          \
-        {                                                     \
-            case 0:                                           \
-                return vget_lane_##postfix(vector, 0);        \
-            case 1:                                           \
-                return vget_lane_##postfix(vector, 1);        \
-            default:                                          \
-                ARM_COMPUTE_ERROR("Invalid lane");            \
-        }                                                     \
+#define VGETLANE_IMPL_2(stype, vtype, postfix)                         \
+    inline stype vgetlane(const vtype vector, const unsigned int lane) \
+    {                                                                  \
+        switch(lane)                                                   \
+        {                                                              \
+            case 0:                                                    \
+                return vget_lane_##postfix(vector, 0);                 \
+            case 1:                                                    \
+                return vget_lane_##postfix(vector, 1);                 \
+            default:                                                   \
+                ARM_COMPUTE_ERROR("Invalid lane");                     \
+        }                                                              \
     }
 
 VGETLANE_IMPL_8(uint8_t, uint8x8_t, u8)
@@ -99,90 +99,90 @@ VGETLANE_IMPL_2(float, float32x2_t, f32)
 VGETLANE_IMPL_4(float16_t, float16x4_t, f16)
 #endif // __ARM_FEATURE_FP16_VECTOR_ARITHMETIC
 
-#define VGETQLANE_IMPL_16(stype, vtype, postfix)               \
-    inline stype vgetqlane(const vtype vector, const int lane) \
-    {                                                          \
-        switch(lane)                                           \
-        {                                                      \
-            case 0:                                            \
-                return vgetq_lane_##postfix(vector, 0);        \
-            case 1:                                            \
-                return vgetq_lane_##postfix(vector, 1);        \
-            case 2:                                            \
-                return vgetq_lane_##postfix(vector, 2);        \
-            case 3:                                            \
-                return vgetq_lane_##postfix(vector, 3);        \
-            case 4:                                            \
-                return vgetq_lane_##postfix(vector, 4);        \
-            case 5:                                            \
-                return vgetq_lane_##postfix(vector, 5);        \
-            case 6:                                            \
-                return vgetq_lane_##postfix(vector, 6);        \
-            case 7:                                            \
-                return vgetq_lane_##postfix(vector, 7);        \
-            case 8:                                            \
-                return vgetq_lane_##postfix(vector, 8);        \
-            case 9:                                            \
-                return vgetq_lane_##postfix(vector, 9);        \
-            case 10:                                           \
-                return vgetq_lane_##postfix(vector, 10);       \
-            case 11:                                           \
-                return vgetq_lane_##postfix(vector, 11);       \
-            case 12:                                           \
-                return vgetq_lane_##postfix(vector, 12);       \
-            case 13:                                           \
-                return vgetq_lane_##postfix(vector, 13);       \
-            case 14:                                           \
-                return vgetq_lane_##postfix(vector, 14);       \
-            case 15:                                           \
-                return vgetq_lane_##postfix(vector, 15);       \
-            default:                                           \
-                ARM_COMPUTE_ERROR("Invalid lane");             \
-        }                                                      \
+#define VGETQLANE_IMPL_16(stype, vtype, postfix)                       \
+    inline stype vgetlane(const vtype vector, const unsigned int lane) \
+    {                                                                  \
+        switch(lane)                                                   \
+        {                                                              \
+            case 0:                                                    \
+                return vgetq_lane_##postfix(vector, 0);                \
+            case 1:                                                    \
+                return vgetq_lane_##postfix(vector, 1);                \
+            case 2:                                                    \
+                return vgetq_lane_##postfix(vector, 2);                \
+            case 3:                                                    \
+                return vgetq_lane_##postfix(vector, 3);                \
+            case 4:                                                    \
+                return vgetq_lane_##postfix(vector, 4);                \
+            case 5:                                                    \
+                return vgetq_lane_##postfix(vector, 5);                \
+            case 6:                                                    \
+                return vgetq_lane_##postfix(vector, 6);                \
+            case 7:                                                    \
+                return vgetq_lane_##postfix(vector, 7);                \
+            case 8:                                                    \
+                return vgetq_lane_##postfix(vector, 8);                \
+            case 9:                                                    \
+                return vgetq_lane_##postfix(vector, 9);                \
+            case 10:                                                   \
+                return vgetq_lane_##postfix(vector, 10);               \
+            case 11:                                                   \
+                return vgetq_lane_##postfix(vector, 11);               \
+            case 12:                                                   \
+                return vgetq_lane_##postfix(vector, 12);               \
+            case 13:                                                   \
+                return vgetq_lane_##postfix(vector, 13);               \
+            case 14:                                                   \
+                return vgetq_lane_##postfix(vector, 14);               \
+            case 15:                                                   \
+                return vgetq_lane_##postfix(vector, 15);               \
+            default:                                                   \
+                ARM_COMPUTE_ERROR("Invalid lane");                     \
+        }                                                              \
     }
 
-#define VGETQLANE_IMPL_8(stype, vtype, postfix)                \
-    inline stype vgetqlane(const vtype vector, const int lane) \
-    {                                                          \
-        switch(lane)                                           \
-        {                                                      \
-            case 0:                                            \
-                return vgetq_lane_##postfix(vector, 0);        \
-            case 1:                                            \
-                return vgetq_lane_##postfix(vector, 1);        \
-            case 2:                                            \
-                return vgetq_lane_##postfix(vector, 2);        \
-            case 3:                                            \
-                return vgetq_lane_##postfix(vector, 3);        \
-            case 4:                                            \
-                return vgetq_lane_##postfix(vector, 4);        \
-            case 5:                                            \
-                return vgetq_lane_##postfix(vector, 5);        \
-            case 6:                                            \
-                return vgetq_lane_##postfix(vector, 6);        \
-            case 7:                                            \
-                return vgetq_lane_##postfix(vector, 7);        \
-            default:                                           \
-                ARM_COMPUTE_ERROR("Invalid lane");             \
-        }                                                      \
+#define VGETQLANE_IMPL_8(stype, vtype, postfix)                        \
+    inline stype vgetlane(const vtype vector, const unsigned int lane) \
+    {                                                                  \
+        switch(lane)                                                   \
+        {                                                              \
+            case 0:                                                    \
+                return vgetq_lane_##postfix(vector, 0);                \
+            case 1:                                                    \
+                return vgetq_lane_##postfix(vector, 1);                \
+            case 2:                                                    \
+                return vgetq_lane_##postfix(vector, 2);                \
+            case 3:                                                    \
+                return vgetq_lane_##postfix(vector, 3);                \
+            case 4:                                                    \
+                return vgetq_lane_##postfix(vector, 4);                \
+            case 5:                                                    \
+                return vgetq_lane_##postfix(vector, 5);                \
+            case 6:                                                    \
+                return vgetq_lane_##postfix(vector, 6);                \
+            case 7:                                                    \
+                return vgetq_lane_##postfix(vector, 7);                \
+            default:                                                   \
+                ARM_COMPUTE_ERROR("Invalid lane");                     \
+        }                                                              \
     }
 
-#define VGETQLANE_IMPL_4(stype, vtype, postfix)                \
-    inline stype vgetqlane(const vtype vector, const int lane) \
-    {                                                          \
-        switch(lane)                                           \
-        {                                                      \
-            case 0:                                            \
-                return vgetq_lane_##postfix(vector, 0);        \
-            case 1:                                            \
-                return vgetq_lane_##postfix(vector, 1);        \
-            case 2:                                            \
-                return vgetq_lane_##postfix(vector, 2);        \
-            case 3:                                            \
-                return vgetq_lane_##postfix(vector, 3);        \
-            default:                                           \
-                ARM_COMPUTE_ERROR("Invalid lane");             \
-        }                                                      \
+#define VGETQLANE_IMPL_4(stype, vtype, postfix)                        \
+    inline stype vgetlane(const vtype vector, const unsigned int lane) \
+    {                                                                  \
+        switch(lane)                                                   \
+        {                                                              \
+            case 0:                                                    \
+                return vgetq_lane_##postfix(vector, 0);                \
+            case 1:                                                    \
+                return vgetq_lane_##postfix(vector, 1);                \
+            case 2:                                                    \
+                return vgetq_lane_##postfix(vector, 2);                \
+            case 3:                                                    \
+                return vgetq_lane_##postfix(vector, 3);                \
+            default:                                                   \
+                ARM_COMPUTE_ERROR("Invalid lane");                     \
+        }                                                              \
     }
 
 VGETQLANE_IMPL_16(uint8_t, uint8x16_t, u8)
@@ -199,6 +199,10 @@ VGETQLANE_IMPL_8(float16_t, float16x8_t, f16)
 #undef VGETLANE_IMPL_8
 #undef VGETLANE_IMPL_4
 #undef VGETLANE_IMPL_2
+
+#undef VGETQLANE_IMPL_16
+#undef VGETQLANE_IMPL_8
+#undef VGETQLANE_IMPL_4
 } // namespace wrapper
 } // namespace arm_compute
 #endif /* __ARM_COMPUTE_WRAPPER_GET_LANE_H__ */
diff --git a/arm_compute/core/NEON/wrapper/intrinsics/intrinsics.h b/arm_compute/core/NEON/wrapper/intrinsics/intrinsics.h
index 97af983e62..896e5106ab 100644
--- a/arm_compute/core/NEON/wrapper/intrinsics/intrinsics.h
+++ b/arm_compute/core/NEON/wrapper/intrinsics/intrinsics.h
@@ -52,6 +52,7 @@
 #include "arm_compute/core/NEON/wrapper/intrinsics/pmin.h"
 #include "arm_compute/core/NEON/wrapper/intrinsics/pow.h"
 #include "arm_compute/core/NEON/wrapper/intrinsics/rev64.h"
+#include "arm_compute/core/NEON/wrapper/intrinsics/setlane.h"
 #include "arm_compute/core/NEON/wrapper/intrinsics/store.h"
 #include "arm_compute/core/NEON/wrapper/intrinsics/sub.h"
 
diff --git a/arm_compute/core/NEON/wrapper/intrinsics/setlane.h b/arm_compute/core/NEON/wrapper/intrinsics/setlane.h
new file mode 100644
index 0000000000..4eba1490c3
--- /dev/null
+++ b/arm_compute/core/NEON/wrapper/intrinsics/setlane.h
@@ -0,0 +1,208 @@
+/*
+ * Copyright (c) 2018-2019 ARM Limited.
+ *
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#ifndef __ARM_COMPUTE_WRAPPER_SET_LANE_H__
+#define __ARM_COMPUTE_WRAPPER_SET_LANE_H__
+
+#include <arm_neon.h>
+
+namespace arm_compute
+{
+namespace wrapper
+{
+#define VSETLANE_IMPL_8(stype, atype, vtype, postfix)                                     \
+    inline stype vsetlane(const atype value, const vtype vector, const unsigned int lane) \
+    {                                                                                     \
+        switch(lane)                                                                      \
+        {                                                                                 \
+            case 0:                                                                       \
+                return vset_lane_##postfix(value, vector, 0);                             \
+            case 1:                                                                       \
+                return vset_lane_##postfix(value, vector, 1);                             \
+            case 2:                                                                       \
+                return vset_lane_##postfix(value, vector, 2);                             \
+            case 3:                                                                       \
+                return vset_lane_##postfix(value, vector, 3);                             \
+            case 4:                                                                       \
+                return vset_lane_##postfix(value, vector, 4);                             \
+            case 5:                                                                       \
+                return vset_lane_##postfix(value, vector, 5);                             \
+            case 6:                                                                       \
+                return vset_lane_##postfix(value, vector, 6);                             \
+            case 7:                                                                       \
+                return vset_lane_##postfix(value, vector, 7);                             \
+            default:                                                                      \
+                ARM_COMPUTE_ERROR("Invalid lane");                                        \
+        }                                                                                 \
+    }
+
+#define VSETLANE_IMPL_4(stype, atype, vtype, postfix)                                     \
+    inline stype vsetlane(const atype value, const vtype vector, const unsigned int lane) \
+    {                                                                                     \
+        switch(lane)                                                                      \
+        {                                                                                 \
+            case 0:                                                                       \
+                return vset_lane_##postfix(value, vector, 0);                             \
+            case 1:                                                                       \
+                return vset_lane_##postfix(value, vector, 1);                             \
+            case 2:                                                                       \
+                return vset_lane_##postfix(value, vector, 2);                             \
+            case 3:                                                                       \
+                return vset_lane_##postfix(value, vector, 3);                             \
+            default:                                                                      \
+                ARM_COMPUTE_ERROR("Invalid lane");                                        \
+        }                                                                                 \
+    }
+
+#define VSETLANE_IMPL_2(stype, atype, vtype, postfix)                                     \
+    inline stype vsetlane(const atype value, const vtype vector, const unsigned int lane) \
+    {                                                                                     \
+        switch(lane)                                                                      \
+        {                                                                                 \
+            case 0:                                                                       \
+                return vset_lane_##postfix(value, vector, 0);                             \
+            case 1:                                                                       \
+                return vset_lane_##postfix(value, vector, 1);                             \
+            default:                                                                      \
+                ARM_COMPUTE_ERROR("Invalid lane");                                        \
+        }                                                                                 \
+    }
+
+VSETLANE_IMPL_8(uint8x8_t, uint8_t, uint8x8_t, u8)
+VSETLANE_IMPL_8(int8x8_t, int8_t, int8x8_t, s8)
+VSETLANE_IMPL_4(uint16x4_t, uint16_t, uint16x4_t, u16)
+VSETLANE_IMPL_4(int16x4_t, int16_t, int16x4_t, s16)
+VSETLANE_IMPL_2(uint32x2_t, uint32_t, uint32x2_t, u32)
+VSETLANE_IMPL_2(int32x2_t, int32_t, int32x2_t, s32)
+VSETLANE_IMPL_2(float32x2_t, float, float32x2_t, f32)
+#ifdef __ARM_FEATURE_FP16_VECTOR_ARITHMETIC
+VSETLANE_IMPL_4(float16x4_t, float16_t, float16x4_t, f16)
+#endif // __ARM_FEATURE_FP16_VECTOR_ARITHMETIC
+
+#define VSETQLANE_IMPL_16(stype, atype, vtype, postfix)                                   \
+    inline stype vsetlane(const atype value, const vtype vector, const unsigned int lane) \
+    {                                                                                     \
+        switch(lane)                                                                      \
+        {                                                                                 \
+            case 0:                                                                       \
+                return vsetq_lane_##postfix(value, vector, 0);                            \
+            case 1:                                                                       \
+                return vsetq_lane_##postfix(value, vector, 1);                            \
+            case 2:                                                                       \
+                return vsetq_lane_##postfix(value, vector, 2);                            \
+            case 3:                                                                       \
+                return vsetq_lane_##postfix(value, vector, 3);                            \
+            case 4:                                                                       \
+                return vsetq_lane_##postfix(value, vector, 4);                            \
+            case 5:                                                                       \
+                return vsetq_lane_##postfix(value, vector, 5);                            \
+            case 6:                                                                       \
+                return vsetq_lane_##postfix(value, vector, 6);                            \
+            case 7:                                                                       \
+                return vsetq_lane_##postfix(value, vector, 7);                            \
+            case 8:                                                                       \
+                return vsetq_lane_##postfix(value, vector, 8);                            \
+            case 9:                                                                       \
+                return vsetq_lane_##postfix(value, vector, 9);                            \
+            case 10:                                                                      \
+                return vsetq_lane_##postfix(value, vector, 10);                           \
+            case 11:                                                                      \
+                return vsetq_lane_##postfix(value, vector, 11);                           \
+            case 12:                                                                      \
+                return vsetq_lane_##postfix(value, vector, 12);                           \
+            case 13:                                                                      \
+                return vsetq_lane_##postfix(value, vector, 13);                           \
+            case 14:                                                                      \
+                return vsetq_lane_##postfix(value, vector, 14);                           \
+            case 15:                                                                      \
+                return vsetq_lane_##postfix(value, vector, 15);                           \
+            default:                                                                      \
+                ARM_COMPUTE_ERROR("Invalid lane");                                        \
+        }                                                                                 \
+    }
+
+#define VSETQLANE_IMPL_8(stype, atype, vtype, postfix)                                    \
+    inline stype vsetlane(const atype value, const vtype vector, const unsigned int lane) \
+    {                                                                                     \
+        switch(lane)                                                                      \
+        {                                                                                 \
+            case 0:                                                                       \
+                return vsetq_lane_##postfix(value, vector, 0);                            \
+            case 1:                                                                       \
+                return vsetq_lane_##postfix(value, vector, 1);                            \
+            case 2:                                                                       \
+                return vsetq_lane_##postfix(value, vector, 2);                            \
+            case 3:                                                                       \
+                return vsetq_lane_##postfix(value, vector, 3);                            \
+            case 4:                                                                       \
+                return vsetq_lane_##postfix(value, vector, 4);                            \
+            case 5:                                                                       \
+                return vsetq_lane_##postfix(value, vector, 5);                            \
+            case 6:                                                                       \
+                return vsetq_lane_##postfix(value, vector, 6);                            \
+            case 7:                                                                       \
+                return vsetq_lane_##postfix(value, vector, 7);                            \
+            default:                                                                      \
+                ARM_COMPUTE_ERROR("Invalid lane");                                        \
+        }                                                                                 \
+    }
+
+#define VSETQLANE_IMPL_4(stype, atype, vtype, postfix)                                    \
+    inline stype vsetlane(const atype value, const vtype vector, const unsigned int lane) \
+    {                                                                                     \
+        switch(lane)                                                                      \
+        {                                                                                 \
+            case 0:                                                                       \
+                return vsetq_lane_##postfix(value, vector, 0);                            \
+            case 1:                                                                       \
+                return vsetq_lane_##postfix(value, vector, 1);                            \
+            case 2:                                                                       \
+                return vsetq_lane_##postfix(value, vector, 2);                            \
+            case 3:                                                                       \
+                return vsetq_lane_##postfix(value, vector, 3);                            \
+            default:                                                                      \
+                ARM_COMPUTE_ERROR("Invalid lane");                                        \
+        }                                                                                 \
+    }
+
+VSETQLANE_IMPL_16(uint8x16_t, uint8_t, uint8x16_t, u8)
+VSETQLANE_IMPL_16(int8x16_t, int8_t, int8x16_t, s8)
+VSETQLANE_IMPL_8(uint16x8_t, uint16_t, uint16x8_t, u16)
+VSETQLANE_IMPL_8(int16x8_t, int16_t, int16x8_t, s16)
+VSETQLANE_IMPL_4(uint32x4_t, uint32_t, uint32x4_t, u32)
+VSETQLANE_IMPL_4(int32x4_t, int32_t, int32x4_t, s32)
+VSETQLANE_IMPL_4(float32x4_t, float, float32x4_t, f32)
+#ifdef __ARM_FEATURE_FP16_VECTOR_ARITHMETIC
+VSETQLANE_IMPL_8(float16x8_t, float16_t, float16x8_t, f16)
+#endif // __ARM_FEATURE_FP16_VECTOR_ARITHMETIC
+
+#undef VSETLANE_IMPL_8
+#undef VSETLANE_IMPL_4
+#undef VSETLANE_IMPL_2
+
+#undef VSETQLANE_IMPL_16
+#undef VSETQLANE_IMPL_8
+#undef VSETQLANE_IMPL_4
+} // namespace wrapper
+} // namespace arm_compute
+#endif /* __ARM_COMPUTE_WRAPPER_AET_LANE_H__ */
diff --git a/arm_compute/runtime/CL/functions/CLRange.h b/arm_compute/runtime/CL/functions/CLRange.h
index 2614534f14..904e6ad20f 100644
--- a/arm_compute/runtime/CL/functions/CLRange.h
+++ b/arm_compute/runtime/CL/functions/CLRange.h
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2018 ARM Limited.
+ * Copyright (c) 2018-2019 ARM Limited.
  *
  * SPDX-License-Identifier: MIT
  *
@@ -39,7 +39,7 @@ class ICLTensor;
 class CLRange : public ICLSimpleFunction
 {
 public:
-    /** Initialise the kernel's start, end, step and output tensor.
+    /** Initialize the kernel's start, end, step and output tensor.
      *
      * @param[out] output Output tensor. Data types supported: U8/S8/QASYMM8/U16/S16/U32/S32/F16/F32.
      * @param[in]  start  The starting value of the sequence.
diff --git a/arm_compute/runtime/NEON/NEFunctions.h b/arm_compute/runtime/NEON/NEFunctions.h
index da61853785..ceebc7b844 100644
--- a/arm_compute/runtime/NEON/NEFunctions.h
+++ b/arm_compute/runtime/NEON/NEFunctions.h
@@ -107,6 +107,7 @@
 #include "arm_compute/runtime/NEON/functions/NEQuantizationLayer.h"
 #include "arm_compute/runtime/NEON/functions/NERNNLayer.h"
 #include "arm_compute/runtime/NEON/functions/NEROIPoolingLayer.h"
+#include "arm_compute/runtime/NEON/functions/NERange.h"
 #include "arm_compute/runtime/NEON/functions/NEReduceMean.h"
 #include "arm_compute/runtime/NEON/functions/NEReductionOperation.h"
 #include "arm_compute/runtime/NEON/functions/NERemap.h"
diff --git a/arm_compute/runtime/NEON/functions/NERange.h b/arm_compute/runtime/NEON/functions/NERange.h
new file mode 100644
index 0000000000..d888c5106d
--- /dev/null
+++ b/arm_compute/runtime/NEON/functions/NERange.h
@@ -0,0 +1,71 @@
+/*
+ * Copyright (c) 2018-2019 ARM Limited.
+ *
+ * SPDX-License-Identifier: MIT
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this software and associated documentation files (the "Software"), to
+ * deal in the Software without restriction, including without limitation the
+ * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
+ * sell copies of the Software, and to permit persons to whom the Software is
+ * furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in all
+ * copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+ * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ */
+#ifndef __ARM_COMPUTE_NERANGE_H__
+#define __ARM_COMPUTE_NERANGE_H__
+
+#include "arm_compute/core/NEON/kernels/NERangeKernel.h"
+#include "arm_compute/core/Types.h"
+#include "arm_compute/runtime/IFunction.h"
+
+namespace arm_compute
+{
+class ITensor;
+
+/** Basic function to run @ref NERangeKernel
+ *
+ * @note The tensor data type for the output must be U8/S8/U16/S16/U32/S32/F16/F32.
+ * @note The function performs generates a sequence with the given start, end and step.
+ */
+class NERange : public IFunction
+{
+public:
+    /** Default constructor */
+    NERange();
+    /** Initialize the kernel's start, end, step and output tensor.
+     *
+     * @param[out] output Output tensor. Data types supported: U8/S8/U16/S16/U32/S32/F16/F32.
+     * @param[in]  start  The starting value of the sequence.
+     * @param[in]  end    The ending (not including) value of the sequence.
+     * @param[in]  step   The gap between each pair of values in the sequence. Default is 1.
+     */
+    void configure(ITensor *output, float start, float end, float step = 1.f);
+    /** Static function to check if given info will lead to a valid configuration of @ref NERange
+     *
+     * @param[in] output Output tensor info. Data types supported: U8/S8/U16/S16/U32/S32/F16/F32.
+     * @param[in] start  The starting value of the sequence.
+     * @param[in] end    The ending (not including) value of the sequence.
+     * @param[in] step   The gap between each pair of values in the sequence. Default is 1.
+     *
+     * @return a status
+     */
+    static Status validate(const ITensorInfo *output, float start, float end, float step = 1.f);
+
+    // Inherited methods overridden:
+    void run() override;
+
+private:
+    NERangeKernel _kernel;
+};
+} // namespace arm_compute
+#endif /* __ARM_COMPUTE_NERANGE_H__ */