Generate Multiple Kernels with Single Pragma

[shibizhao]

Hi,
I am a beginner of the CCF framework.
I wrote a simple program vadd below:

int kernel(){
	static int _VecAdd2d_buffer[128*128*1] = {};
	static int _inputa_buffer[128*128*1] = {};
	static int _inputb_buffer[128*128*1] = {};
	for(int init_idx = 0; init_idx < (128*128*1); ++init_idx){
		_VecAdd2d_buffer[init_idx] = init_idx + 1;
	}
	for(int init_idx = 0; init_idx < (128*128*1); ++init_idx){
		_inputa_buffer[init_idx] = init_idx + 2;
	}
	for(int init_idx = 0; init_idx < (128*128*1); ++init_idx){
		_inputb_buffer[init_idx] = init_idx + 3;
	}
	stage_0:{
		const int const5 = 0;
		const int const6 = 128;
		for(int incre8 = const5; incre8 < const6; ++incre8){
			const int const9 = 128;
			auto mul10 = incre8 * const9;
			const int const7 = 128;
			auto mul11 = const7 * incre8;
			const int const3 = 0;
			const int const4 = 128;
			#pragma CGRA
			for(int incre12 = const3; incre12 < const4; ++incre12){
				auto add13 = mul10 + incre12;
				auto add14 = mul11 + incre12;
				auto add15 = _inputa_buffer + add13;
				auto load16 = *(add15);
				auto add17 = _inputb_buffer + add13;
				auto load18 = *(add17);
				auto add19 = load16 + load18;
				auto add20 = _VecAdd2d_buffer + add14;
				*(add20) = add19;
			}
		} 
	}
	int _results = 0;
	for(int init_idx = 0; init_idx < (128*128*1); ++init_idx){
		_results += (int)_VecAdd2d_buffer[init_idx];
	}
	for(int init_idx = 0; init_idx < (128*128*1); ++init_idx){
		_results += (int)_inputa_buffer[init_idx];
	}
	for(int init_idx = 0; init_idx < (128*128*1); ++init_idx){
		_results += (int)_inputb_buffer[init_idx];
	}
	return _results;
}

int main(){
	kernel();
	return 0;
}

The CGRA pragma is added at loop-incre12. However, there are two loops in my CGRAExec directory.
Look forward to your reply!
Thanks!