blockDim.x*gridDim.x 跳过一个grid
int <<<参数1，参数2>>>（int *a,int * b,int * c);
如果是一维的，参数1表示一个grid里面有多少个block块，参数2表示一个block块里面有多少个thread线程
namespace caffe {
template <typename Dtype>
__global__ void LogwxlForward(const float scale_,const int n,const Dtype * in,Dtype * out){
CUDA_KERNEL_LOOP(index,n){
out[index]=scale_*in[index];
}
}
template <typename Dtype>
__global__ void LogwxlBackward(const float scale_,const int n,const Dtype *in_diff,Dtype* out_diff){
CUDA_KERNEL_LOOP(index,n) {
out_diff[index]=in_diff[index]*scale_;
}
}
template <typename Dtype>
void LogwxlLayer<Dtype>::Forward_gpu(
const vector<Blob<Dtype>*>& bottom, const vector<Blob<Dtype>*>& top) {
//this->Forward_cpu(bottom, top);
const Dtype* bottom_data=bottom[0]->gpu_data();
Dtype * top_data=top[0]->mutable_gpu_data();
const int count=bottom[0]->count();
LogwxlForward<Dtype><<<CAFFE_GET_BLOCKS(count),CAFFE_CUDA_NUM_THREADS>>>(scale_,count,bottom_data,top_data);
CUDA_POST_KERNEL_CHECK;
}
template <typename Dtype>
void LogwxlLayer<Dtype>::Backward_gpu(
const vector<Blob<Dtype>*>& top,
const vector<bool>& propagate_down,
const vector<Blob<Dtype>*>& bottom) {
//this->Backward_cpu(top, propagate_down, bottom);
if(propagate_down[0]){
const Dtype *top_diff=top[0]->gpu_diff();
Dtype * bottom_diff=bottom[0]->mutable_gpu_diff();
const int count=bottom[0]->count();
LogwxlBackward<Dtype><<<CAFFE_GET_BLOCKS(count),CAFFE_CUDA_NUM_THREADS>>>(scale_,count,top_diff,bottom_diff);
CUDA_POST_KERNEL_CHECK;
}
}
INSTANTIATE_LAYER_GPU_FUNCS(LogwxlLayer);
} // namespace caffe