Optimize the unary cuda kernels for the contiguous case.

kernels/src/affine.cu

@@ -6,22 +6,28 @@ extern "C" __global__ void FN_NAME(  \
     const size_t numel,  \
     const size_t num_dims, \
     const size_t *info, \
-    const TYPENAME *x, \
+    const TYPENAME *inp, \
-    TYPENAME *y, \
+    TYPENAME *out, \
     const TYPENAME mul, \
     const TYPENAME add \
 ) {  \
     const size_t *dims = info; \
     const size_t *strides = info + num_dims; \
-    unsigned int i = blockIdx.x * blockDim.x + threadIdx.x;  \
+    if (is_contiguous(num_dims, dims, strides)) { \
-    if (i >= numel) {  \
+        for (unsigned int i = blockIdx.x * blockDim.x + threadIdx.x; i < numel; i += blockDim.x * gridDim.x) { \
-        return;  \
+            TYPENAME x = inp ? inp[i] : out[i]; \
+            out[i] = x * mul + add; \
         } \
+    } \
+    else { \
+        for (unsigned int i = blockIdx.x * blockDim.x + threadIdx.x; i < numel; i += blockDim.x * gridDim.x) { \
             unsigned strided_i = get_strided_index(i, num_dims, dims, strides); \
-    y[strided_i] = x[i] * mul + add; \
+            TYPENAME x = inp ? inp[strided_i] : out[i]; \
+            out[i] = x * mul + add; \
+        } \
+    } \
 } \
 
 AFFINE_OP(float, affine_f32)
 AFFINE_OP(double, affine_f64)
 AFFINE_OP(uint32_t, affine_u32)
-

kernels/src/cuda_utils.cuh

@@ -1,6 +1,22 @@
 #include "cuda_fp16.h"
 #include "compatibility.cuh"
 
+__device__ bool is_contiguous(
+    const size_t num_dims,
+    const size_t *dims,
+    const size_t *strides
+) {
+    size_t acc = 1;
+    for (unsigned int d = 0; d < num_dims; d++) {
+        unsigned int dim_idx = num_dims - 1 - d;
+        if (acc != strides[dim_idx]) {
+            return false;
+        }
+        acc *= dims[dim_idx];
+    }
+    return true;
+}
+
 __device__ unsigned int get_strided_index(
     unsigned int idx,
     const size_t num_dims,

kernels/src/unary.cu

@@ -10,12 +10,20 @@ extern "C" __global__ void FN_NAME( \
 ) { \
     const size_t *dims = info; \
     const size_t *strides = info + num_dims; \
+    if (is_contiguous(num_dims, dims, strides)) { \
+        for (unsigned int i = blockIdx.x * blockDim.x + threadIdx.x; i < numel; i += blockDim.x * gridDim.x) { \
+            TYPENAME x = inp ? inp[i] : out[i]; \
+            out[i] = FUNC; \
+        } \
+    } \
+    else { \
         for (unsigned int i = blockIdx.x * blockDim.x + threadIdx.x; i < numel; i += blockDim.x * gridDim.x) { \
             unsigned strided_i = get_strided_index(i, num_dims, dims, strides); \
             TYPENAME x = inp ? inp[strided_i] : out[i]; \
             out[i] = FUNC; \
         } \
     } \
+} \
 
 template<typename T>
 __device__ T gelu_fwd(T x) {