From 3e7ca1a55b5ff20a221ecabf6c90e041fd3cc75e Mon Sep 17 00:00:00 2001
From: critsium-xy <tsfxwbbzxy@163.com>
Date: Thu, 2 Jan 2025 20:20:46 +0800
Subject: [PATCH 01/15] Initial commit

---
 source/module_base/blas_connector.cpp | 207 +++++++++++++++++++++++---
 1 file changed, 183 insertions(+), 24 deletions(-)

diff --git a/source/module_base/blas_connector.cpp b/source/module_base/blas_connector.cpp
index 85ea4584e9..fc99819d01 100644
--- a/source/module_base/blas_connector.cpp
+++ b/source/module_base/blas_connector.cpp
@@ -5,32 +5,101 @@
 #include "module_base/global_variable.h"
 #endif
 
+#ifdef __CUDA
+#include <base/macros/macros.h>
+#include <cuda_runtime.h>
+#include <thrust/complex.h>
+#include <thrust/execution_policy.h>
+#include <thrust/inner_product.h>
+#include "module_base/tool_quit.h"
+
+#include "cublas_v2.h"
+
+namespace BlasUtils{
+
+	static cublasHandle_t cublas_handle = nullptr;
+
+	void createGpuBlasHandle(){
+		if (cublas_handle == nullptr) {
+			cublasErrcheck(cublasCreate(&cublas_handle));
+		}
+	}
+
+	void destoryBLAShandle(){
+		if (cublas_handle != nullptr) {
+			cublasErrcheck(cublasDestroy(cublas_handle));
+			cublas_handle = nullptr;
+		}
+	}
+
+
+	cublasOperation_t judge_trans(bool is_complex, const char& trans, const char* name)
+	{
+		if (trans == 'N')
+		{
+			return CUBLAS_OP_N;
+		}
+		else if(trans == 'T')
+		{
+			return CUBLAS_OP_T;
+		}
+		else if(is_complex && trans == 'C')
+		{
+			return CUBLAS_OP_C;
+		}
+		return CUBLAS_OP_N;
+	}
+
+} // namespace BlasUtils
+
+#endif
+
 void BlasConnector::axpy( const int n, const float alpha, const float *X, const int incX, float *Y, const int incY, base_device::AbacusDevice_t device_type)
 {
 	if (device_type == base_device::AbacusDevice_t::CpuDevice) {
 		saxpy_(&n, &alpha, X, &incX, Y, &incY);
-}
+	}
+	else if (device_type == base_device::AbacusDevice_t::GpuDevice){
+#ifdef __CUDA
+		cublasErrcheck(cublasSaxpy(BlasUtils::cublas_handle, n, &alpha, X, incX, Y, incY));
+#endif
+	}
 }
 
 void BlasConnector::axpy( const int n, const double alpha, const double *X, const int incX, double *Y, const int incY, base_device::AbacusDevice_t device_type)
 {
 	if (device_type == base_device::AbacusDevice_t::CpuDevice) {
 		daxpy_(&n, &alpha, X, &incX, Y, &incY);
-}
+	}
+	else if (device_type == base_device::AbacusDevice_t::GpuDevice){
+#ifdef __CUDA
+		cublasErrcheck(cublasDaxpy(BlasUtils::cublas_handle, n, &alpha, X, incX, Y, incY));
+#endif
+	}
 }
 
 void BlasConnector::axpy( const int n, const std::complex<float> alpha, const std::complex<float> *X, const int incX, std::complex<float> *Y, const int incY, base_device::AbacusDevice_t device_type)
 {
 	if (device_type == base_device::AbacusDevice_t::CpuDevice) {
 		caxpy_(&n, &alpha, X, &incX, Y, &incY);
-}
+	}
+	else if (device_type == base_device::AbacusDevice_t::GpuDevice){
+#ifdef __CUDA
+		cublasErrcheck(cublasCaxpy(BlasUtils::cublas_handle, n, (float2*)&alpha, (float2*)X, incX, (float2*)Y, incY));
+#endif
+	}
 }
 
 void BlasConnector::axpy( const int n, const std::complex<double> alpha, const std::complex<double> *X, const int incX, std::complex<double> *Y, const int incY, base_device::AbacusDevice_t device_type)
 {
 	if (device_type == base_device::AbacusDevice_t::CpuDevice) {
 		zaxpy_(&n, &alpha, X, &incX, Y, &incY);
-}
+	}
+	else if (device_type == base_device::AbacusDevice_t::GpuDevice){
+#ifdef __CUDA
+		cublasErrcheck(cublasZaxpy(BlasUtils::cublas_handle, n, (double2*)&alpha, (double2*)X, incX, (double2*)Y, incY));
+#endif
+	}
 }
 
 
@@ -39,28 +108,48 @@ void BlasConnector::scal( const int n,  const float alpha, float *X, const int i
 {
 	if (device_type == base_device::AbacusDevice_t::CpuDevice) {
 		sscal_(&n, &alpha, X, &incX);
-}
+	}
+	else if (device_type == base_device::AbacusDevice_t::GpuDevice) {
+#ifdef __CUDA
+		cublasErrcheck(cublasSscal(BlasUtils::cublas_handle, n, &alpha, X, incX));
+#endif
+	}
 }
 
 void BlasConnector::scal( const int n, const double alpha, double *X, const int incX, base_device::AbacusDevice_t device_type)
 {
 	if (device_type == base_device::AbacusDevice_t::CpuDevice) {
 		dscal_(&n, &alpha, X, &incX);
-}
+	}
+	else if (device_type == base_device::AbacusDevice_t::GpuDevice) {
+#ifdef __CUDA
+		cublasErrcheck(cublasDscal(BlasUtils::cublas_handle, n, &alpha, X, incX));
+#endif
+	}
 }
 
 void BlasConnector::scal( const int n, const std::complex<float> alpha, std::complex<float> *X, const int incX, base_device::AbacusDevice_t device_type)
 {
 	if (device_type == base_device::AbacusDevice_t::CpuDevice) {
 		cscal_(&n, &alpha, X, &incX);
-}
+	}
+	else if (device_type == base_device::AbacusDevice_t::GpuDevice) {
+#ifdef __CUDA
+		cublasErrcheck(cublasCscal(BlasUtils::cublas_handle, n, (float2*)&alpha, (float2*)X, incX));
+#endif
+	}
 }
 
 void BlasConnector::scal( const int n, const std::complex<double> alpha, std::complex<double> *X, const int incX, base_device::AbacusDevice_t device_type)
 {
 	if (device_type == base_device::AbacusDevice_t::CpuDevice) {
 		zscal_(&n, &alpha, X, &incX);
-}
+	}
+	else if (device_type == base_device::AbacusDevice_t::GpuDevice) {
+#ifdef __CUDA
+		cublasErrcheck(cublasZscal(BlasUtils::cublas_handle, n, (double2*)&alpha, (double2*)X, incX));
+#endif
+	}
 }
 
 
@@ -70,6 +159,13 @@ float BlasConnector::dot( const int n, const float *X, const int incX, const flo
 	if (device_type == base_device::AbacusDevice_t::CpuDevice) {
 		return sdot_(&n, X, &incX, Y, &incY);
 	}
+	else if (device_type == base_device::AbacusDevice_t::GpuDevice){
+#ifdef __CUDA
+		float result = 0.0;
+		cublasErrcheck(cublasSdot(BlasUtils::cublas_handle, n, X, incX, Y, incY, &result));
+		return result;
+#endif
+	}
 	return sdot_(&n, X, &incX, Y, &incY);
 }
 
@@ -78,6 +174,13 @@ double BlasConnector::dot( const int n, const double *X, const int incX, const d
 	if (device_type == base_device::AbacusDevice_t::CpuDevice) {
 		return ddot_(&n, X, &incX, Y, &incY);
 	}
+	else if (device_type == base_device::AbacusDevice_t::GpuDevice){
+#ifdef __CUDA
+		double result = 0.0;
+		cublasErrcheck(cublasDdot(BlasUtils::cublas_handle, n, X, incX, Y, incY, &result));
+		return result;
+#endif
+	}
 	return ddot_(&n, X, &incX, Y, &incY);
 }
 
@@ -92,13 +195,20 @@ void BlasConnector::gemm(const char transa, const char transb, const int m, cons
 		&alpha, b, &ldb, a, &lda,
 		&beta, c, &ldc);
 	}
-	#ifdef __DSP
+#ifdef __DSP
 	else if (device_type == base_device::AbacusDevice_t::DspDevice){
 		sgemm_mth_(&transb, &transa, &n, &m, &k,
 		&alpha, b, &ldb, a, &lda,
 		&beta, c, &ldc, GlobalV::MY_RANK);
 	}
-	#endif
+#endif
+	else if (device_type == base_device::AbacusDevice_t::GpuDevice){
+#ifdef __CUDA
+		cublasOperation_t cutransA = judge_trans(false, transa, "gemm_op");
+		cublasOperation_t cutransB = judge_trans(false, transb, "gemm_op");
+		cublasErrcheck(cublasSgemm(BlasUtils::cublas_handle, cutransA, cutransB, n, m, k, &alpha, b, ldb, a, lda, &beta, c, ldc));
+#endif
+	}
 }
 
 void BlasConnector::gemm(const char transa, const char transb, const int m, const int n, const int k,
@@ -110,13 +220,20 @@ void BlasConnector::gemm(const char transa, const char transb, const int m, cons
 		&alpha, b, &ldb, a, &lda,
 		&beta, c, &ldc);
 	}
-	#ifdef __DSP
+#ifdef __DSP
 	else if (device_type == base_device::AbacusDevice_t::DspDevice){
 		dgemm_mth_(&transb, &transa, &n, &m, &k,
 		&alpha, b, &ldb, a, &lda,
 		&beta, c, &ldc, GlobalV::MY_RANK);
 	}
-	#endif
+#endif
+	else if (device_type == base_device::AbacusDevice_t::GpuDevice){
+#ifdef __CUDA
+		cublasOperation_t cutransA = judge_trans(false, transa, "gemm_op");
+		cublasOperation_t cutransB = judge_trans(false, transb, "gemm_op");
+		cublasErrcheck(cublasDgemm(BlasUtils::cublas_handle, cutransA, cutransB, n, m, k, &alpha, b, ldb, a, lda, &beta, c, ldc));
+#endif
+	}
 }
 
 void BlasConnector::gemm(const char transa, const char transb, const int m, const int n, const int k,
@@ -128,13 +245,20 @@ void BlasConnector::gemm(const char transa, const char transb, const int m, cons
         &alpha, b, &ldb, a, &lda,
         &beta, c, &ldc);
 	}
-	#ifdef __DSP
+#ifdef __DSP
 	else if (device_type == base_device::AbacusDevice_t::DspDevice) {
     	cgemm_mth_(&transb, &transa, &n, &m, &k,
         &alpha, b, &ldb, a, &lda,
         &beta, c, &ldc, GlobalV::MY_RANK);
 	}
-	#endif
+#endif
+	else if (device_type == base_device::AbacusDevice_t::GpuDevice){
+#ifdef __CUDA
+		cublasOperation_t cutransA = judge_trans(false, transa, "gemm_op");
+		cublasOperation_t cutransB = judge_trans(false, transb, "gemm_op");
+		cublasErrcheck(cublasCgemm(BlasUtils::cublas_handle, cutransA, cutransB, n, m, k, (float2*)&alpha, (float2*)b, ldb, (float2*)a, lda, (float2*)&beta, (float2*)c, ldc));
+#endif
+	}
 }
 
 void BlasConnector::gemm(const char transa, const char transb, const int m, const int n, const int k,
@@ -146,13 +270,20 @@ void BlasConnector::gemm(const char transa, const char transb, const int m, cons
 		&alpha, b, &ldb, a, &lda,
 		&beta, c, &ldc);
 	}
-	#ifdef __DSP
+#ifdef __DSP
 	else if (device_type == base_device::AbacusDevice_t::DspDevice) {
     	zgemm_mth_(&transb, &transa, &n, &m, &k,
         &alpha, b, &ldb, a, &lda,
         &beta, c, &ldc, GlobalV::MY_RANK);
 	}
-	#endif
+#endif
+	else if (device_type == base_device::AbacusDevice_t::GpuDevice){
+#ifdef __CUDA
+		cublasOperation_t cutransA = judge_trans(false, transa, "gemm_op");
+		cublasOperation_t cutransB = judge_trans(false, transb, "gemm_op");
+		cublasErrcheck(cublasZgemm(BlasUtils::cublas_handle, cutransA, cutransB, n, m, k, (double2*)&alpha, (double2*)b, ldb, (double2*)a, lda, (double2*)&beta, (double2*)c, ldc));
+#endif
+	}
 }
 
 // Col-Major part
@@ -165,13 +296,20 @@ void BlasConnector::gemm_cm(const char transa, const char transb, const int m, c
 		&alpha, a, &lda, b, &ldb,
 		&beta, c, &ldc);
 	}
-	#ifdef __DSP
+#ifdef __DSP
 	else if (device_type == base_device::AbacusDevice_t::DspDevice){
 		sgemm_mth_(&transb, &transa, &m, &n, &k,
 		&alpha, a, &lda, b, &ldb,
 		&beta, c, &ldc, GlobalV::MY_RANK);
 	}
-	#endif
+#endif
+	else if (device_type == base_device::AbacusDevice_t::GpuDevice){
+#ifdef __CUDA
+		cublasOperation_t cutransA = judge_trans(false, transa, "gemm_op");
+		cublasOperation_t cutransB = judge_trans(false, transb, "gemm_op");
+		cublasErrcheck(cublasSgemm(BlasUtils::cublas_handle, cutransA, cutransB, m, n, k, &alpha, a, lda, b, ldb, &beta, c, ldc));
+#endif
+	}
 }
 
 void BlasConnector::gemm_cm(const char transa, const char transb, const int m, const int n, const int k,
@@ -183,13 +321,20 @@ void BlasConnector::gemm_cm(const char transa, const char transb, const int m, c
 		&alpha, a, &lda, b, &ldb,
 		&beta, c, &ldc);
 	}
-	#ifdef __DSP
+#ifdef __DSP
 	else if (device_type == base_device::AbacusDevice_t::DspDevice){
 		dgemm_mth_(&transa, &transb, &m, &n, &k,
 		&alpha, a, &lda, b, &ldb,
 		&beta, c, &ldc, GlobalV::MY_RANK);
 	}
-	#endif
+#endif
+	else if (device_type == base_device::AbacusDevice_t::GpuDevice){
+#ifdef __CUDA
+		cublasOperation_t cutransA = judge_trans(false, transa, "gemm_op");
+		cublasOperation_t cutransB = judge_trans(false, transb, "gemm_op");
+		cublasErrcheck(cublasDgemm(BlasUtils::cublas_handle, cutransA, cutransB, m, n, k, &alpha, a, lda, b, ldb, &beta, c, ldc));
+#endif
+	}
 }
 
 void BlasConnector::gemm_cm(const char transa, const char transb, const int m, const int n, const int k,
@@ -201,13 +346,20 @@ void BlasConnector::gemm_cm(const char transa, const char transb, const int m, c
         &alpha, a, &lda, b, &ldb,
         &beta, c, &ldc);
 	}
-	#ifdef __DSP
+#ifdef __DSP
 	else if (device_type == base_device::AbacusDevice_t::DspDevice) {
     	cgemm_mth_(&transa, &transb, &m, &n, &k,
         &alpha, a, &lda, b, &ldb,
         &beta, c, &ldc, GlobalV::MY_RANK);
 	}
-	#endif
+#endif
+	else if (device_type == base_device::AbacusDevice_t::GpuDevice){
+#ifdef __CUDA
+		cublasOperation_t cutransA = judge_trans(false, transa, "gemm_op");
+		cublasOperation_t cutransB = judge_trans(false, transb, "gemm_op");
+		cublasErrcheck(cublasCgemm(BlasUtils::cublas_handle, cutransA, cutransB, m, n, k, (float2*)&alpha, (float2*)a, lda, (float2*)b, ldb, (float2*)&beta, (float2*)c, ldc));
+#endif
+	}
 }
 
 void BlasConnector::gemm_cm(const char transa, const char transb, const int m, const int n, const int k,
@@ -219,13 +371,20 @@ void BlasConnector::gemm_cm(const char transa, const char transb, const int m, c
 		&alpha, a, &lda, b, &ldb,
 		&beta, c, &ldc);
 	}
-	#ifdef __DSP
+#ifdef __DSP
 	else if (device_type == base_device::AbacusDevice_t::DspDevice) {
     	zgemm_mth_(&transa, &transb, &m, &n, &k,
         &alpha, a, &lda, b, &ldb,
         &beta, c, &ldc, GlobalV::MY_RANK);
 	}
-	#endif
+#endif
+	else if (device_type == base_device::AbacusDevice_t::GpuDevice){
+#ifdef __CUDA
+		cublasOperation_t cutransA = judge_trans(false, transa, "gemm_op");
+		cublasOperation_t cutransB = judge_trans(false, transb, "gemm_op");
+		cublasErrcheck(cublasZgemm(BlasUtils::cublas_handle, cutransA, cutransB, m, n, k, (double2*)&alpha, (double2*)a, lda, (double2*)b, ldb, (double2*)&beta, (double2*)c, ldc));
+#endif
+	}
 }
 
 // Symm and Hemm part. Only col-major is supported.

From 5198e2db58ceca4dc3dff98b304f9a52c731c621 Mon Sep 17 00:00:00 2001
From: critsium-xy <tsfxwbbzxy@163.com>
Date: Thu, 2 Jan 2025 22:02:14 +0800
Subject: [PATCH 02/15] Modify CMakeLists

---
 source/module_base/blas_connector.cpp  | 32 +++++++++++++-------------
 source/module_base/test/CMakeLists.txt | 30 ++++++++++++------------
 2 files changed, 31 insertions(+), 31 deletions(-)

diff --git a/source/module_base/blas_connector.cpp b/source/module_base/blas_connector.cpp
index fc99819d01..3bb91e2f01 100644
--- a/source/module_base/blas_connector.cpp
+++ b/source/module_base/blas_connector.cpp
@@ -204,8 +204,8 @@ void BlasConnector::gemm(const char transa, const char transb, const int m, cons
 #endif
 	else if (device_type == base_device::AbacusDevice_t::GpuDevice){
 #ifdef __CUDA
-		cublasOperation_t cutransA = judge_trans(false, transa, "gemm_op");
-		cublasOperation_t cutransB = judge_trans(false, transb, "gemm_op");
+		cublasOperation_t cutransA = BlasUtils::judge_trans(false, transa, "gemm_op");
+		cublasOperation_t cutransB = BlasUtils::judge_trans(false, transb, "gemm_op");
 		cublasErrcheck(cublasSgemm(BlasUtils::cublas_handle, cutransA, cutransB, n, m, k, &alpha, b, ldb, a, lda, &beta, c, ldc));
 #endif
 	}
@@ -229,8 +229,8 @@ void BlasConnector::gemm(const char transa, const char transb, const int m, cons
 #endif
 	else if (device_type == base_device::AbacusDevice_t::GpuDevice){
 #ifdef __CUDA
-		cublasOperation_t cutransA = judge_trans(false, transa, "gemm_op");
-		cublasOperation_t cutransB = judge_trans(false, transb, "gemm_op");
+		cublasOperation_t cutransA = BlasUtils::judge_trans(false, transa, "gemm_op");
+		cublasOperation_t cutransB = BlasUtils::judge_trans(false, transb, "gemm_op");
 		cublasErrcheck(cublasDgemm(BlasUtils::cublas_handle, cutransA, cutransB, n, m, k, &alpha, b, ldb, a, lda, &beta, c, ldc));
 #endif
 	}
@@ -254,8 +254,8 @@ void BlasConnector::gemm(const char transa, const char transb, const int m, cons
 #endif
 	else if (device_type == base_device::AbacusDevice_t::GpuDevice){
 #ifdef __CUDA
-		cublasOperation_t cutransA = judge_trans(false, transa, "gemm_op");
-		cublasOperation_t cutransB = judge_trans(false, transb, "gemm_op");
+		cublasOperation_t cutransA = BlasUtils::judge_trans(false, transa, "gemm_op");
+		cublasOperation_t cutransB = BlasUtils::judge_trans(false, transb, "gemm_op");
 		cublasErrcheck(cublasCgemm(BlasUtils::cublas_handle, cutransA, cutransB, n, m, k, (float2*)&alpha, (float2*)b, ldb, (float2*)a, lda, (float2*)&beta, (float2*)c, ldc));
 #endif
 	}
@@ -279,8 +279,8 @@ void BlasConnector::gemm(const char transa, const char transb, const int m, cons
 #endif
 	else if (device_type == base_device::AbacusDevice_t::GpuDevice){
 #ifdef __CUDA
-		cublasOperation_t cutransA = judge_trans(false, transa, "gemm_op");
-		cublasOperation_t cutransB = judge_trans(false, transb, "gemm_op");
+		cublasOperation_t cutransA = BlasUtils::judge_trans(false, transa, "gemm_op");
+		cublasOperation_t cutransB = BlasUtils::judge_trans(false, transb, "gemm_op");
 		cublasErrcheck(cublasZgemm(BlasUtils::cublas_handle, cutransA, cutransB, n, m, k, (double2*)&alpha, (double2*)b, ldb, (double2*)a, lda, (double2*)&beta, (double2*)c, ldc));
 #endif
 	}
@@ -305,8 +305,8 @@ void BlasConnector::gemm_cm(const char transa, const char transb, const int m, c
 #endif
 	else if (device_type == base_device::AbacusDevice_t::GpuDevice){
 #ifdef __CUDA
-		cublasOperation_t cutransA = judge_trans(false, transa, "gemm_op");
-		cublasOperation_t cutransB = judge_trans(false, transb, "gemm_op");
+		cublasOperation_t cutransA = BlasUtils::judge_trans(false, transa, "gemm_op");
+		cublasOperation_t cutransB = BlasUtils::judge_trans(false, transb, "gemm_op");
 		cublasErrcheck(cublasSgemm(BlasUtils::cublas_handle, cutransA, cutransB, m, n, k, &alpha, a, lda, b, ldb, &beta, c, ldc));
 #endif
 	}
@@ -330,8 +330,8 @@ void BlasConnector::gemm_cm(const char transa, const char transb, const int m, c
 #endif
 	else if (device_type == base_device::AbacusDevice_t::GpuDevice){
 #ifdef __CUDA
-		cublasOperation_t cutransA = judge_trans(false, transa, "gemm_op");
-		cublasOperation_t cutransB = judge_trans(false, transb, "gemm_op");
+		cublasOperation_t cutransA = BlasUtils::judge_trans(false, transa, "gemm_op");
+		cublasOperation_t cutransB = BlasUtils::judge_trans(false, transb, "gemm_op");
 		cublasErrcheck(cublasDgemm(BlasUtils::cublas_handle, cutransA, cutransB, m, n, k, &alpha, a, lda, b, ldb, &beta, c, ldc));
 #endif
 	}
@@ -355,8 +355,8 @@ void BlasConnector::gemm_cm(const char transa, const char transb, const int m, c
 #endif
 	else if (device_type == base_device::AbacusDevice_t::GpuDevice){
 #ifdef __CUDA
-		cublasOperation_t cutransA = judge_trans(false, transa, "gemm_op");
-		cublasOperation_t cutransB = judge_trans(false, transb, "gemm_op");
+		cublasOperation_t cutransA = BlasUtils::judge_trans(false, transa, "gemm_op");
+		cublasOperation_t cutransB = BlasUtils::judge_trans(false, transb, "gemm_op");
 		cublasErrcheck(cublasCgemm(BlasUtils::cublas_handle, cutransA, cutransB, m, n, k, (float2*)&alpha, (float2*)a, lda, (float2*)b, ldb, (float2*)&beta, (float2*)c, ldc));
 #endif
 	}
@@ -380,8 +380,8 @@ void BlasConnector::gemm_cm(const char transa, const char transb, const int m, c
 #endif
 	else if (device_type == base_device::AbacusDevice_t::GpuDevice){
 #ifdef __CUDA
-		cublasOperation_t cutransA = judge_trans(false, transa, "gemm_op");
-		cublasOperation_t cutransB = judge_trans(false, transb, "gemm_op");
+		cublasOperation_t cutransA = BlasUtils::judge_trans(false, transa, "gemm_op");
+		cublasOperation_t cutransB = BlasUtils::judge_trans(false, transb, "gemm_op");
 		cublasErrcheck(cublasZgemm(BlasUtils::cublas_handle, cutransA, cutransB, m, n, k, (double2*)&alpha, (double2*)a, lda, (double2*)b, ldb, (double2*)&beta, (double2*)c, ldc));
 #endif
 	}
diff --git a/source/module_base/test/CMakeLists.txt b/source/module_base/test/CMakeLists.txt
index 09b77c7404..15b52c048e 100644
--- a/source/module_base/test/CMakeLists.txt
+++ b/source/module_base/test/CMakeLists.txt
@@ -2,7 +2,7 @@ remove_definitions(-D__MPI)
 install(DIRECTORY data DESTINATION ${CMAKE_CURRENT_BINARY_DIR})
 AddTest(
   TARGET base_blas_connector
-  LIBS parameter  ${math_libs}
+  LIBS parameter  ${math_libs} base device
   SOURCES blas_connector_test.cpp ../blas_connector.cpp
 )
 AddTest(
@@ -31,7 +31,7 @@ AddTest(
 )
 ADDTest(
   TARGET base_global_function
-  LIBS parameter ${math_libs} 
+  LIBS parameter ${math_libs} base device
   SOURCES global_function_test.cpp ../blas_connector.cpp ../global_function.cpp ../tool_quit.cpp ../global_variable.cpp ../global_file.cpp ../memory.cpp ../timer.cpp
 )
 AddTest(
@@ -41,7 +41,7 @@ AddTest(
 )
 AddTest(
   TARGET base_matrix3
-  LIBS parameter  ${math_libs}
+  LIBS parameter  ${math_libs} base device
   SOURCES matrix3_test.cpp ../matrix3.cpp ../matrix.cpp ../tool_quit.cpp ../global_variable.cpp ../global_file.cpp ../global_function.cpp ../memory.cpp ../timer.cpp ../blas_connector.cpp
 )
 AddTest(
@@ -56,7 +56,7 @@ AddTest(
 )
 AddTest(
  TARGET base_matrix
-  LIBS parameter  ${math_libs}
+  LIBS parameter  ${math_libs} base device
   SOURCES matrix_test.cpp ../blas_connector.cpp ../matrix.cpp ../tool_quit.cpp ../global_variable.cpp ../global_file.cpp ../global_function.cpp ../memory.cpp ../timer.cpp
 )
 AddTest(
@@ -66,7 +66,7 @@ AddTest(
 )
 AddTest(
   TARGET base_complexmatrix
-  LIBS parameter  ${math_libs}
+  LIBS parameter  ${math_libs} base device
   SOURCES complexmatrix_test.cpp ../blas_connector.cpp ../complexmatrix.cpp ../matrix.cpp
 )
 AddTest(
@@ -93,12 +93,12 @@ AddTest(
 )
 AddTest(
   TARGET base_mathzone
-  LIBS parameter  ${math_libs}
+  LIBS parameter  ${math_libs} base device
   SOURCES mathzone_test.cpp ../matrix3.cpp ../matrix.cpp ../tool_quit.cpp ../global_variable.cpp ../global_file.cpp ../global_function.cpp ../memory.cpp ../timer.cpp ../blas_connector.cpp
 )
 AddTest(
   TARGET base_mathzone_add1
-  LIBS parameter  ${math_libs}
+  LIBS parameter  ${math_libs} base device
   SOURCES mathzone_add1_test.cpp ../blas_connector.cpp ../mathzone_add1.cpp ../math_sphbes.cpp ../matrix3.cpp ../matrix.cpp ../tool_quit.cpp ../global_variable.cpp ../global_file.cpp ../global_function.cpp ../memory.cpp ../timer.cpp 
 )
 AddTest(
@@ -108,7 +108,7 @@ AddTest(
 )
 AddTest(
   TARGET base_gram_schmidt_orth
-  LIBS parameter  ${math_libs}
+  LIBS parameter  ${math_libs} base device
   SOURCES gram_schmidt_orth_test.cpp ../blas_connector.cpp ../gram_schmidt_orth.h ../gram_schmidt_orth-inl.h ../global_function.h ../math_integral.cpp
 )
 AddTest(
@@ -118,7 +118,7 @@ AddTest(
 )
 AddTest(
   TARGET base_inverse_matrix
-  LIBS parameter  ${math_libs}
+  LIBS parameter  ${math_libs} device
   SOURCES inverse_matrix_test.cpp ../blas_connector.cpp ../inverse_matrix.cpp ../complexmatrix.cpp ../matrix.cpp ../timer.cpp 
 )
 AddTest(
@@ -140,19 +140,19 @@ AddTest(
 
 AddTest(
   TARGET base_lapack_connector
-  LIBS parameter  ${math_libs}
+  LIBS parameter  ${math_libs} base device
   SOURCES lapack_connector_test.cpp ../blas_connector.cpp ../lapack_connector.h
 )
 
 AddTest(
   TARGET base_opt_CG
-  LIBS parameter  ${math_libs}
+  LIBS parameter  ${math_libs} base device
   SOURCES opt_CG_test.cpp opt_test_tools.cpp ../blas_connector.cpp ../opt_CG.cpp ../opt_DCsrch.cpp ../global_variable.cpp ../parallel_reduce.cpp
 )
 
 AddTest(
   TARGET base_opt_TN
-  LIBS parameter  ${math_libs}
+  LIBS parameter  ${math_libs} base device
   SOURCES opt_TN_test.cpp opt_test_tools.cpp ../blas_connector.cpp ../opt_CG.cpp ../opt_DCsrch.cpp ../global_variable.cpp ../parallel_reduce.cpp
 )
 
@@ -195,13 +195,13 @@ AddTest(
 AddTest(
   TARGET spherical_bessel_transformer
   SOURCES spherical_bessel_transformer_test.cpp ../blas_connector.cpp ../spherical_bessel_transformer.cpp ../math_sphbes.cpp ../math_integral.cpp ../timer.cpp
-  LIBS parameter  ${math_libs}
+  LIBS parameter  ${math_libs} base device
 )
 
 AddTest(
   TARGET cubic_spline
   SOURCES cubic_spline_test.cpp ../blas_connector.cpp ../cubic_spline.cpp
-  LIBS parameter  ${math_libs}
+  LIBS parameter  ${math_libs} base device
 )
 
 AddTest(
@@ -215,7 +215,7 @@ AddTest(
 AddTest(
   TARGET assoc_laguerre_test
   SOURCES assoc_laguerre_test.cpp ../blas_connector.cpp ../assoc_laguerre.cpp ../tool_quit.cpp ../global_variable.cpp ../global_file.cpp ../global_function.cpp ../memory.cpp ../timer.cpp
-  LIBS parameter  ${math_libs}
+  LIBS parameter  ${math_libs} base device
 )
 
 AddTest(

From 2aa45242966d753861ad550c84cf0d9af1620dee Mon Sep 17 00:00:00 2001
From: critsium-xy <tsfxwbbzxy@163.com>
Date: Thu, 2 Jan 2025 22:19:53 +0800
Subject: [PATCH 03/15] Complete CMakeLists in module_base

---
 source/module_base/test/CMakeLists.txt | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/source/module_base/test/CMakeLists.txt b/source/module_base/test/CMakeLists.txt
index 15b52c048e..18baf6d572 100644
--- a/source/module_base/test/CMakeLists.txt
+++ b/source/module_base/test/CMakeLists.txt
@@ -81,7 +81,7 @@ AddTest(
 )
 AddTest(
   TARGET base_ylmreal
-  LIBS parameter  ${math_libs} device
+  LIBS parameter  ${math_libs} base device
   SOURCES math_ylmreal_test.cpp ../blas_connector.cpp ../math_ylmreal.cpp ../complexmatrix.cpp ../global_variable.cpp ../ylm.cpp ../realarray.cpp ../timer.cpp ../matrix.cpp ../vector3.h
           ../parallel_reduce.cpp ../parallel_global.cpp ../parallel_comm.cpp ../parallel_common.cpp
           ../memory.cpp ../libm/branred.cpp ../libm/sincos.cpp
@@ -118,7 +118,7 @@ AddTest(
 )
 AddTest(
   TARGET base_inverse_matrix
-  LIBS parameter  ${math_libs} device
+  LIBS parameter  ${math_libs} base device
   SOURCES inverse_matrix_test.cpp ../blas_connector.cpp ../inverse_matrix.cpp ../complexmatrix.cpp ../matrix.cpp ../timer.cpp 
 )
 AddTest(
@@ -134,7 +134,7 @@ AddTest(
 
 AddTest(
   TARGET base_math_chebyshev
-  LIBS parameter  ${math_libs} device container
+  LIBS parameter  ${math_libs} base device container
   SOURCES math_chebyshev_test.cpp  ../blas_connector.cpp ../math_chebyshev.cpp ../tool_quit.cpp ../global_variable.cpp ../timer.cpp ../global_file.cpp ../global_function.cpp ../memory.cpp ../parallel_reduce.cpp
 )
 
@@ -209,7 +209,7 @@ AddTest(
   SOURCES clebsch_gordan_coeff_test.cpp ../blas_connector.cpp ../clebsch_gordan_coeff.cpp ../intarray.cpp ../realarray.cpp ../complexmatrix.cpp ../matrix.cpp ../timer.cpp
   ../math_ylmreal.cpp  ../global_variable.cpp ../ylm.cpp  ../timer.cpp  ../vector3.h ../parallel_reduce.cpp ../parallel_global.cpp ../parallel_comm.cpp ../parallel_common.cpp
   ../memory.cpp ../libm/branred.cpp ../libm/sincos.cpp ../inverse_matrix.cpp ../lapack_connector.h
-  LIBS parameter  ${math_libs} device
+  LIBS parameter  ${math_libs} base device
 )
 
 AddTest(

From 3c86d0bdadfdcc937f096a1f29ebe15ee1dbf546 Mon Sep 17 00:00:00 2001
From: critsium-xy <tsfxwbbzxy@163.com>
Date: Thu, 2 Jan 2025 22:53:01 +0800
Subject: [PATCH 04/15] Add blas_connector.cpp definition

---
 source/module_base/blas_connector.cpp | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

diff --git a/source/module_base/blas_connector.cpp b/source/module_base/blas_connector.cpp
index 3bb91e2f01..5bd369ed62 100644
--- a/source/module_base/blas_connector.cpp
+++ b/source/module_base/blas_connector.cpp
@@ -1,3 +1,5 @@
+#ifndef BLAS_CONNECTOR_CPP
+#define BLAS_CONNECTOR_CPP
 #include "blas_connector.h"
 
 #ifdef __DSP
@@ -531,4 +533,6 @@ void BlasConnector::copy(const long n, const std::complex<double> *a, const int
 	if (device_type == base_device::AbacusDevice_t::CpuDevice) {
 		zcopy_(&n, a, &incx, b, &incy);
 	}
-}
\ No newline at end of file
+}
+
+#endif
\ No newline at end of file

From f310a90f39f8a37b5757255cf81396acf21b653d Mon Sep 17 00:00:00 2001
From: critsium-xy <tsfxwbbzxy@163.com>
Date: Fri, 3 Jan 2025 00:28:30 +0800
Subject: [PATCH 05/15] Fix module_base tests

---
 source/module_base/blas_connector.cpp         |  6 +--
 source/module_base/test/CMakeLists.txt        | 40 +++++++++----------
 .../test/clebsch_gordan_coeff_test.cpp        | 12 ------
 .../module_base/test/complexmatrix_test.cpp   |  6 ---
 .../module_base/test/inverse_matrix_test.cpp  |  6 ---
 source/module_base/test/math_ylmreal_test.cpp | 10 -----
 6 files changed, 19 insertions(+), 61 deletions(-)

diff --git a/source/module_base/blas_connector.cpp b/source/module_base/blas_connector.cpp
index 5bd369ed62..3bb91e2f01 100644
--- a/source/module_base/blas_connector.cpp
+++ b/source/module_base/blas_connector.cpp
@@ -1,5 +1,3 @@
-#ifndef BLAS_CONNECTOR_CPP
-#define BLAS_CONNECTOR_CPP
 #include "blas_connector.h"
 
 #ifdef __DSP
@@ -533,6 +531,4 @@ void BlasConnector::copy(const long n, const std::complex<double> *a, const int
 	if (device_type == base_device::AbacusDevice_t::CpuDevice) {
 		zcopy_(&n, a, &incx, b, &incy);
 	}
-}
-
-#endif
\ No newline at end of file
+}
\ No newline at end of file
diff --git a/source/module_base/test/CMakeLists.txt b/source/module_base/test/CMakeLists.txt
index 18baf6d572..e0f61062b4 100644
--- a/source/module_base/test/CMakeLists.txt
+++ b/source/module_base/test/CMakeLists.txt
@@ -3,7 +3,7 @@ install(DIRECTORY data DESTINATION ${CMAKE_CURRENT_BINARY_DIR})
 AddTest(
   TARGET base_blas_connector
   LIBS parameter  ${math_libs} base device
-  SOURCES blas_connector_test.cpp ../blas_connector.cpp
+  SOURCES blas_connector_test.cpp
 )
 AddTest(
   TARGET base_atom_in
@@ -32,7 +32,7 @@ AddTest(
 ADDTest(
   TARGET base_global_function
   LIBS parameter ${math_libs} base device
-  SOURCES global_function_test.cpp ../blas_connector.cpp ../global_function.cpp ../tool_quit.cpp ../global_variable.cpp ../global_file.cpp ../memory.cpp ../timer.cpp
+  SOURCES global_function_test.cpp
 )
 AddTest(
   TARGET base_vector3
@@ -42,7 +42,7 @@ AddTest(
 AddTest(
   TARGET base_matrix3
   LIBS parameter  ${math_libs} base device
-  SOURCES matrix3_test.cpp ../matrix3.cpp ../matrix.cpp ../tool_quit.cpp ../global_variable.cpp ../global_file.cpp ../global_function.cpp ../memory.cpp ../timer.cpp ../blas_connector.cpp
+  SOURCES matrix3_test.cpp
 )
 AddTest(
   TARGET base_intarray
@@ -57,7 +57,7 @@ AddTest(
 AddTest(
  TARGET base_matrix
   LIBS parameter  ${math_libs} base device
-  SOURCES matrix_test.cpp ../blas_connector.cpp ../matrix.cpp ../tool_quit.cpp ../global_variable.cpp ../global_file.cpp ../global_function.cpp ../memory.cpp ../timer.cpp
+  SOURCES matrix_test.cpp
 )
 AddTest(
   TARGET base_complexarray
@@ -67,7 +67,7 @@ AddTest(
 AddTest(
   TARGET base_complexmatrix
   LIBS parameter  ${math_libs} base device
-  SOURCES complexmatrix_test.cpp ../blas_connector.cpp ../complexmatrix.cpp ../matrix.cpp
+  SOURCES complexmatrix_test.cpp
 )
 AddTest(
   TARGET base_integral
@@ -82,9 +82,7 @@ AddTest(
 AddTest(
   TARGET base_ylmreal
   LIBS parameter  ${math_libs} base device
-  SOURCES math_ylmreal_test.cpp ../blas_connector.cpp ../math_ylmreal.cpp ../complexmatrix.cpp ../global_variable.cpp ../ylm.cpp ../realarray.cpp ../timer.cpp ../matrix.cpp ../vector3.h
-          ../parallel_reduce.cpp ../parallel_global.cpp ../parallel_comm.cpp ../parallel_common.cpp
-          ../memory.cpp ../libm/branred.cpp ../libm/sincos.cpp
+  SOURCES math_ylmreal_test.cpp ../libm/branred.cpp ../libm/sincos.cpp
 )
 AddTest(
   TARGET base_math_sphbes
@@ -94,12 +92,12 @@ AddTest(
 AddTest(
   TARGET base_mathzone
   LIBS parameter  ${math_libs} base device
-  SOURCES mathzone_test.cpp ../matrix3.cpp ../matrix.cpp ../tool_quit.cpp ../global_variable.cpp ../global_file.cpp ../global_function.cpp ../memory.cpp ../timer.cpp ../blas_connector.cpp
+  SOURCES mathzone_test.cpp
 )
 AddTest(
   TARGET base_mathzone_add1
   LIBS parameter  ${math_libs} base device
-  SOURCES mathzone_add1_test.cpp ../blas_connector.cpp ../mathzone_add1.cpp ../math_sphbes.cpp ../matrix3.cpp ../matrix.cpp ../tool_quit.cpp ../global_variable.cpp ../global_file.cpp ../global_function.cpp ../memory.cpp ../timer.cpp 
+  SOURCES mathzone_add1_test.cpp
 )
 AddTest(
   TARGET base_math_polyint
@@ -109,7 +107,7 @@ AddTest(
 AddTest(
   TARGET base_gram_schmidt_orth
   LIBS parameter  ${math_libs} base device
-  SOURCES gram_schmidt_orth_test.cpp ../blas_connector.cpp ../gram_schmidt_orth.h ../gram_schmidt_orth-inl.h ../global_function.h ../math_integral.cpp
+  SOURCES gram_schmidt_orth_test.cpp
 )
 AddTest(
   TARGET base_math_bspline
@@ -119,7 +117,7 @@ AddTest(
 AddTest(
   TARGET base_inverse_matrix
   LIBS parameter  ${math_libs} base device
-  SOURCES inverse_matrix_test.cpp ../blas_connector.cpp ../inverse_matrix.cpp ../complexmatrix.cpp ../matrix.cpp ../timer.cpp 
+  SOURCES inverse_matrix_test.cpp 
 )
 AddTest(
   TARGET base_mymath
@@ -135,25 +133,25 @@ AddTest(
 AddTest(
   TARGET base_math_chebyshev
   LIBS parameter  ${math_libs} base device container
-  SOURCES math_chebyshev_test.cpp  ../blas_connector.cpp ../math_chebyshev.cpp ../tool_quit.cpp ../global_variable.cpp ../timer.cpp ../global_file.cpp ../global_function.cpp ../memory.cpp ../parallel_reduce.cpp
+  SOURCES math_chebyshev_test.cpp
 )
 
 AddTest(
   TARGET base_lapack_connector
   LIBS parameter  ${math_libs} base device
-  SOURCES lapack_connector_test.cpp ../blas_connector.cpp ../lapack_connector.h
+  SOURCES lapack_connector_test.cpp
 )
 
 AddTest(
   TARGET base_opt_CG
   LIBS parameter  ${math_libs} base device
-  SOURCES opt_CG_test.cpp opt_test_tools.cpp ../blas_connector.cpp ../opt_CG.cpp ../opt_DCsrch.cpp ../global_variable.cpp ../parallel_reduce.cpp
+  SOURCES opt_CG_test.cpp opt_test_tools.cpp
 )
 
 AddTest(
   TARGET base_opt_TN
   LIBS parameter  ${math_libs} base device
-  SOURCES opt_TN_test.cpp opt_test_tools.cpp ../blas_connector.cpp ../opt_CG.cpp ../opt_DCsrch.cpp ../global_variable.cpp ../parallel_reduce.cpp
+  SOURCES opt_TN_test.cpp opt_test_tools.cpp
 )
 
 AddTest(
@@ -194,27 +192,25 @@ AddTest(
 
 AddTest(
   TARGET spherical_bessel_transformer
-  SOURCES spherical_bessel_transformer_test.cpp ../blas_connector.cpp ../spherical_bessel_transformer.cpp ../math_sphbes.cpp ../math_integral.cpp ../timer.cpp
+  SOURCES spherical_bessel_transformer_test.cpp
   LIBS parameter  ${math_libs} base device
 )
 
 AddTest(
   TARGET cubic_spline
-  SOURCES cubic_spline_test.cpp ../blas_connector.cpp ../cubic_spline.cpp
+  SOURCES cubic_spline_test.cpp
   LIBS parameter  ${math_libs} base device
 )
 
 AddTest(
   TARGET clebsch_gordan_coeff_test
-  SOURCES clebsch_gordan_coeff_test.cpp ../blas_connector.cpp ../clebsch_gordan_coeff.cpp ../intarray.cpp ../realarray.cpp ../complexmatrix.cpp ../matrix.cpp ../timer.cpp
-  ../math_ylmreal.cpp  ../global_variable.cpp ../ylm.cpp  ../timer.cpp  ../vector3.h ../parallel_reduce.cpp ../parallel_global.cpp ../parallel_comm.cpp ../parallel_common.cpp
-  ../memory.cpp ../libm/branred.cpp ../libm/sincos.cpp ../inverse_matrix.cpp ../lapack_connector.h
+  SOURCES clebsch_gordan_coeff_test.cpp
   LIBS parameter  ${math_libs} base device
 )
 
 AddTest(
   TARGET assoc_laguerre_test
-  SOURCES assoc_laguerre_test.cpp ../blas_connector.cpp ../assoc_laguerre.cpp ../tool_quit.cpp ../global_variable.cpp ../global_file.cpp ../global_function.cpp ../memory.cpp ../timer.cpp
+  SOURCES assoc_laguerre_test.cpp
   LIBS parameter  ${math_libs} base device
 )
 
diff --git a/source/module_base/test/clebsch_gordan_coeff_test.cpp b/source/module_base/test/clebsch_gordan_coeff_test.cpp
index 16efa091b5..888249765f 100644
--- a/source/module_base/test/clebsch_gordan_coeff_test.cpp
+++ b/source/module_base/test/clebsch_gordan_coeff_test.cpp
@@ -16,18 +16,6 @@
  *     - functions: gen_rndm_r and compute_ap
  */
 
-namespace ModuleBase
-{
-void WARNING_QUIT(const std::string& file, const std::string& description)
-{
-    return;
-}
-void WARNING(const std::string& file, const std::string& description)
-{
-    return;
-}
-} // namespace ModuleBase
-
 TEST(ClebschGordanTest, ClebschGordanExit)
 {
     int lmaxkb = -2;
diff --git a/source/module_base/test/complexmatrix_test.cpp b/source/module_base/test/complexmatrix_test.cpp
index 0adc52363a..da11fafcfd 100644
--- a/source/module_base/test/complexmatrix_test.cpp
+++ b/source/module_base/test/complexmatrix_test.cpp
@@ -38,12 +38,6 @@
  *
  */
 
-//a mock function of WARNING_QUIT, to avoid the uncorrected call by matrix.cpp at line 37.
-namespace ModuleBase
-{
-    void WARNING_QUIT(const std::string &file,const std::string &description) {exit(1);}
-}
-
 inline void EXPECT_COMPLEX_EQ(const std::complex<double>& a,const std::complex<double>& b)
 {
     EXPECT_DOUBLE_EQ(a.real(),b.real());
diff --git a/source/module_base/test/inverse_matrix_test.cpp b/source/module_base/test/inverse_matrix_test.cpp
index a871f906cd..b88e556af1 100644
--- a/source/module_base/test/inverse_matrix_test.cpp
+++ b/source/module_base/test/inverse_matrix_test.cpp
@@ -16,12 +16,6 @@
  *     - computes the inverse of a dim*dim real matrix
  */
 
-//a mock function of WARNING_QUIT, to avoid the uncorrected call by matrix.cpp at line 37.
-namespace ModuleBase
-{
-    void WARNING_QUIT(const std::string &file,const std::string &description) {exit(1);}
-}
-
 TEST(InverseMatrixComplexTest, InverseMatrixComplex)
 {
     int dim = 10;
diff --git a/source/module_base/test/math_ylmreal_test.cpp b/source/module_base/test/math_ylmreal_test.cpp
index c973d8cd28..891c948f7e 100644
--- a/source/module_base/test/math_ylmreal_test.cpp
+++ b/source/module_base/test/math_ylmreal_test.cpp
@@ -36,16 +36,6 @@
  *
  */
 
-
-
-//mock functions of WARNING_QUIT and WARNING
-namespace ModuleBase
-{
-    void WARNING_QUIT(const std::string &file,const std::string &description) {exit(1);}
-    void WARNING(const std::string &file,const std::string &description) {return ;}
-}
-
-
 class YlmRealTest : public testing::Test
 {
     protected:

From 0464af2cda4ab6b9244f9dc87c89463561f14e77 Mon Sep 17 00:00:00 2001
From: critsium-xy <tsfxwbbzxy@163.com>
Date: Fri, 3 Jan 2025 13:09:23 +0800
Subject: [PATCH 06/15] Fix tests failure

---
 source/module_base/test/CMakeLists.txt           | 16 ++++++++--------
 source/module_base/test/global_function_test.cpp |  1 -
 2 files changed, 8 insertions(+), 9 deletions(-)

diff --git a/source/module_base/test/CMakeLists.txt b/source/module_base/test/CMakeLists.txt
index e0f61062b4..ffe15daf80 100644
--- a/source/module_base/test/CMakeLists.txt
+++ b/source/module_base/test/CMakeLists.txt
@@ -31,8 +31,8 @@ AddTest(
 )
 ADDTest(
   TARGET base_global_function
-  LIBS parameter ${math_libs} base device
-  SOURCES global_function_test.cpp
+  LIBS parameter ${math_libs}
+  SOURCES global_function_test.cpp ../global_function.cpp ../tool_quit.cpp ../global_variable.cpp ../global_file.cpp ../memory.cpp ../timer.cpp
 )
 AddTest(
   TARGET base_vector3
@@ -132,8 +132,8 @@ AddTest(
 
 AddTest(
   TARGET base_math_chebyshev
-  LIBS parameter  ${math_libs} base device container
-  SOURCES math_chebyshev_test.cpp
+  LIBS parameter  ${math_libs} device container
+  SOURCES math_chebyshev_test.cpp ../math_chebyshev.cpp ../tool_quit.cpp ../global_variable.cpp ../timer.cpp ../global_file.cpp ../global_function.cpp ../memory.cpp ../parallel_reduce.cpp
 )
 
 AddTest(
@@ -144,14 +144,14 @@ AddTest(
 
 AddTest(
   TARGET base_opt_CG
-  LIBS parameter  ${math_libs} base device
-  SOURCES opt_CG_test.cpp opt_test_tools.cpp
+  LIBS parameter  ${math_libs}
+  SOURCES opt_CG_test.cpp opt_test_tools.cpp ../opt_CG.cpp ../opt_DCsrch.cpp ../global_variable.cpp ../parallel_reduce.cpp
 )
 
 AddTest(
   TARGET base_opt_TN
-  LIBS parameter  ${math_libs} base device
-  SOURCES opt_TN_test.cpp opt_test_tools.cpp
+  LIBS parameter  ${math_libs}
+  SOURCES opt_TN_test.cpp opt_test_tools.cpp ../opt_CG.cpp ../opt_DCsrch.cpp ../global_variable.cpp ../parallel_reduce.cpp
 )
 
 AddTest(
diff --git a/source/module_base/test/global_function_test.cpp b/source/module_base/test/global_function_test.cpp
index 013396d6b1..9b2c49fba6 100644
--- a/source/module_base/test/global_function_test.cpp
+++ b/source/module_base/test/global_function_test.cpp
@@ -4,7 +4,6 @@
 #include "module_parameter/parameter.h"
 #undef private
 #include "../vector3.h"
-#include "../blas_connector.h"
 #include "../tool_quit.h"
 #include <string>
 #include <cstring>

From 2a0cfd6115ec5ccd2b5bcd7fa290b41bac0dbb19 Mon Sep 17 00:00:00 2001
From: critsium-xy <tsfxwbbzxy@163.com>
Date: Fri, 3 Jan 2025 13:44:23 +0800
Subject: [PATCH 07/15] fix opt_test

---
 source/module_base/test/CMakeLists.txt | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/source/module_base/test/CMakeLists.txt b/source/module_base/test/CMakeLists.txt
index ffe15daf80..d2e30ee8d7 100644
--- a/source/module_base/test/CMakeLists.txt
+++ b/source/module_base/test/CMakeLists.txt
@@ -133,7 +133,7 @@ AddTest(
 AddTest(
   TARGET base_math_chebyshev
   LIBS parameter  ${math_libs} device container
-  SOURCES math_chebyshev_test.cpp ../math_chebyshev.cpp ../tool_quit.cpp ../global_variable.cpp ../timer.cpp ../global_file.cpp ../global_function.cpp ../memory.cpp ../parallel_reduce.cpp
+  SOURCES math_chebyshev_test.cpp ../blas_connector.cpp ../math_chebyshev.cpp ../tool_quit.cpp ../global_variable.cpp ../timer.cpp ../global_file.cpp ../global_function.cpp ../memory.cpp ../parallel_reduce.cpp
 )
 
 AddTest(
@@ -144,14 +144,14 @@ AddTest(
 
 AddTest(
   TARGET base_opt_CG
-  LIBS parameter  ${math_libs}
-  SOURCES opt_CG_test.cpp opt_test_tools.cpp ../opt_CG.cpp ../opt_DCsrch.cpp ../global_variable.cpp ../parallel_reduce.cpp
+  LIBS parameter  ${math_libs} device
+  SOURCES opt_CG_test.cpp opt_test_tools.cpp ../blas_connector.cpp ../opt_CG.cpp ../opt_DCsrch.cpp ../global_variable.cpp ../parallel_reduce.cpp
 )
 
 AddTest(
   TARGET base_opt_TN
-  LIBS parameter  ${math_libs}
-  SOURCES opt_TN_test.cpp opt_test_tools.cpp ../opt_CG.cpp ../opt_DCsrch.cpp ../global_variable.cpp ../parallel_reduce.cpp
+  LIBS parameter  ${math_libs} device
+  SOURCES opt_TN_test.cpp opt_test_tools.cpp ../blas_connector.cpp ../opt_CG.cpp ../opt_DCsrch.cpp ../global_variable.cpp ../parallel_reduce.cpp
 )
 
 AddTest(

From f558e070df50817433b834c8657ee45f6faaa27f Mon Sep 17 00:00:00 2001
From: critsium-xy <tsfxwbbzxy@163.com>
Date: Fri, 3 Jan 2025 13:51:36 +0800
Subject: [PATCH 08/15] OPTFIX2

---
 source/module_base/opt_CG.h            | 4 ++++
 source/module_base/test/CMakeLists.txt | 4 ++--
 2 files changed, 6 insertions(+), 2 deletions(-)

diff --git a/source/module_base/opt_CG.h b/source/module_base/opt_CG.h
index 32707d6ca7..8014271cf5 100644
--- a/source/module_base/opt_CG.h
+++ b/source/module_base/opt_CG.h
@@ -5,7 +5,11 @@
 
 #include <iostream>
 
+#ifdef __CUDA
+#undef __CUDA
 #include "global_function.h"
+#define __CUDA
+#endif
 #include "module_base/parallel_reduce.h"
 
 namespace ModuleBase
diff --git a/source/module_base/test/CMakeLists.txt b/source/module_base/test/CMakeLists.txt
index d2e30ee8d7..4e2bb170fc 100644
--- a/source/module_base/test/CMakeLists.txt
+++ b/source/module_base/test/CMakeLists.txt
@@ -144,13 +144,13 @@ AddTest(
 
 AddTest(
   TARGET base_opt_CG
-  LIBS parameter  ${math_libs} device
+  LIBS parameter  ${math_libs}
   SOURCES opt_CG_test.cpp opt_test_tools.cpp ../blas_connector.cpp ../opt_CG.cpp ../opt_DCsrch.cpp ../global_variable.cpp ../parallel_reduce.cpp
 )
 
 AddTest(
   TARGET base_opt_TN
-  LIBS parameter  ${math_libs} device
+  LIBS parameter  ${math_libs}
   SOURCES opt_TN_test.cpp opt_test_tools.cpp ../blas_connector.cpp ../opt_CG.cpp ../opt_DCsrch.cpp ../global_variable.cpp ../parallel_reduce.cpp
 )
 

From bab57273ad907d76934e165687cd7a0bf360470a Mon Sep 17 00:00:00 2001
From: critsium-xy <tsfxwbbzxy@163.com>
Date: Fri, 3 Jan 2025 13:57:21 +0800
Subject: [PATCH 09/15] Return all changes

---
 source/module_base/opt_CG.h            |  4 ----
 source/module_base/test/CMakeLists.txt | 12 ++++++------
 2 files changed, 6 insertions(+), 10 deletions(-)

diff --git a/source/module_base/opt_CG.h b/source/module_base/opt_CG.h
index 8014271cf5..32707d6ca7 100644
--- a/source/module_base/opt_CG.h
+++ b/source/module_base/opt_CG.h
@@ -5,11 +5,7 @@
 
 #include <iostream>
 
-#ifdef __CUDA
-#undef __CUDA
 #include "global_function.h"
-#define __CUDA
-#endif
 #include "module_base/parallel_reduce.h"
 
 namespace ModuleBase
diff --git a/source/module_base/test/CMakeLists.txt b/source/module_base/test/CMakeLists.txt
index 4e2bb170fc..ade269a1e6 100644
--- a/source/module_base/test/CMakeLists.txt
+++ b/source/module_base/test/CMakeLists.txt
@@ -132,8 +132,8 @@ AddTest(
 
 AddTest(
   TARGET base_math_chebyshev
-  LIBS parameter  ${math_libs} device container
-  SOURCES math_chebyshev_test.cpp ../blas_connector.cpp ../math_chebyshev.cpp ../tool_quit.cpp ../global_variable.cpp ../timer.cpp ../global_file.cpp ../global_function.cpp ../memory.cpp ../parallel_reduce.cpp
+  LIBS parameter  ${math_libs} base device container
+  SOURCES math_chebyshev_test.cpp ../math_chebyshev.cpp ../tool_quit.cpp ../global_variable.cpp ../timer.cpp ../global_file.cpp ../global_function.cpp ../memory.cpp ../parallel_reduce.cpp
 )
 
 AddTest(
@@ -144,14 +144,14 @@ AddTest(
 
 AddTest(
   TARGET base_opt_CG
-  LIBS parameter  ${math_libs}
-  SOURCES opt_CG_test.cpp opt_test_tools.cpp ../blas_connector.cpp ../opt_CG.cpp ../opt_DCsrch.cpp ../global_variable.cpp ../parallel_reduce.cpp
+  LIBS parameter  ${math_libs} base device
+  SOURCES opt_CG_test.cpp opt_test_tools.cpp ../opt_CG.cpp ../opt_DCsrch.cpp ../global_variable.cpp ../parallel_reduce.cpp
 )
 
 AddTest(
   TARGET base_opt_TN
-  LIBS parameter  ${math_libs}
-  SOURCES opt_TN_test.cpp opt_test_tools.cpp ../blas_connector.cpp ../opt_CG.cpp ../opt_DCsrch.cpp ../global_variable.cpp ../parallel_reduce.cpp
+  LIBS parameter  ${math_libs} base device
+  SOURCES opt_TN_test.cpp opt_test_tools.cpp ../opt_CG.cpp ../opt_DCsrch.cpp ../global_variable.cpp ../parallel_reduce.cpp
 )
 
 AddTest(

From 8a8ec04024c0fb9ab074afaee1cb7bf39bd8af39 Mon Sep 17 00:00:00 2001
From: critsium-xy <tsfxwbbzxy@163.com>
Date: Fri, 3 Jan 2025 14:00:52 +0800
Subject: [PATCH 10/15] Fix global_func_text

---
 source/module_base/test/global_function_test.cpp | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/source/module_base/test/global_function_test.cpp b/source/module_base/test/global_function_test.cpp
index 9b2c49fba6..05d4d70877 100644
--- a/source/module_base/test/global_function_test.cpp
+++ b/source/module_base/test/global_function_test.cpp
@@ -691,6 +691,9 @@ TEST_F(GlobalFunctionTest,MemAvailable)
 
 TEST_F(GlobalFunctionTest,BlockHere)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
 	std::string output2;
 	std::string block_in="111";
 	GlobalV::MY_RANK=1;
@@ -705,6 +708,9 @@ TEST_F(GlobalFunctionTest,BlockHere)
 
 TEST_F(GlobalFunctionTest,BlockHere2)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
 	std::string output2;
 	std::string block_in="111";
 	GlobalV::MY_RANK=0;
@@ -723,6 +729,9 @@ TEST_F(GlobalFunctionTest,BlockHere2)
 
 TEST_F(GlobalFunctionTest,BlockHere3)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
 	std::string output2;
 	std::string block_in="111";
 	GlobalV::MY_RANK=0;

From 1cd1ffc2bd6c96c75739f653fbb16f2572e4909a Mon Sep 17 00:00:00 2001
From: critsium-xy <tsfxwbbzxy@163.com>
Date: Fri, 3 Jan 2025 14:08:50 +0800
Subject: [PATCH 11/15] Fix MPI Bug

---
 source/module_base/test/math_chebyshev_test.cpp |  3 +++
 source/module_base/test/opt_CG_test.cpp         | 15 +++++++++++++++
 source/module_base/test/opt_TN_test.cpp         |  6 ++++++
 3 files changed, 24 insertions(+)

diff --git a/source/module_base/test/math_chebyshev_test.cpp b/source/module_base/test/math_chebyshev_test.cpp
index 125dbdaeaa..3c3aef26c4 100644
--- a/source/module_base/test/math_chebyshev_test.cpp
+++ b/source/module_base/test/math_chebyshev_test.cpp
@@ -336,6 +336,9 @@ TEST_F(MathChebyshevTest, tracepolyA)
 
 TEST_F(MathChebyshevTest, checkconverge)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     const int norder = 100;
     p_chetest = new ModuleBase::Chebyshev<double>(norder);
     auto fun_sigma_y
diff --git a/source/module_base/test/opt_CG_test.cpp b/source/module_base/test/opt_CG_test.cpp
index 4b324c7cbb..d2faf8763b 100644
--- a/source/module_base/test/opt_CG_test.cpp
+++ b/source/module_base/test/opt_CG_test.cpp
@@ -143,6 +143,9 @@ class CG_test : public testing::Test
 
 TEST_F(CG_test, Stand_Solve_LinearEq)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     CG_Solve_LinearEq();
     EXPECT_NEAR(x[0], 0.5, DOUBLETHRESHOLD);
     EXPECT_NEAR(x[1], 1.6429086563584579739e-18, DOUBLETHRESHOLD);
@@ -153,6 +156,9 @@ TEST_F(CG_test, Stand_Solve_LinearEq)
 
 TEST_F(CG_test, PR_Solve_LinearEq)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     Solve(1, 0);
     EXPECT_NEAR(x[0], 0.50000000000003430589, DOUBLETHRESHOLD);
     EXPECT_NEAR(x[1], -3.4028335704761047964e-14, DOUBLETHRESHOLD);
@@ -163,6 +169,9 @@ TEST_F(CG_test, PR_Solve_LinearEq)
 
 TEST_F(CG_test, HZ_Solve_LinearEq)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     Solve(2, 0);
     EXPECT_NEAR(x[0], 0.49999999999999944489, DOUBLETHRESHOLD);
     EXPECT_NEAR(x[1], -9.4368957093138305936e-16, DOUBLETHRESHOLD);
@@ -173,6 +182,9 @@ TEST_F(CG_test, HZ_Solve_LinearEq)
 
 TEST_F(CG_test, PR_Min_Func)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     Solve(1, 1);
     EXPECT_NEAR(x[0], 4.0006805979150792396, DOUBLETHRESHOLD);
     EXPECT_NEAR(x[1], 2.0713759992720870429, DOUBLETHRESHOLD);
@@ -183,6 +195,9 @@ TEST_F(CG_test, PR_Min_Func)
 
 TEST_F(CG_test, HZ_Min_Func)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     Solve(2, 1);
     EXPECT_NEAR(x[0], 4.0006825378033568086, DOUBLETHRESHOLD);
     EXPECT_NEAR(x[1], 2.0691732100663737803, DOUBLETHRESHOLD);
diff --git a/source/module_base/test/opt_TN_test.cpp b/source/module_base/test/opt_TN_test.cpp
index db523b53e9..b136194881 100644
--- a/source/module_base/test/opt_TN_test.cpp
+++ b/source/module_base/test/opt_TN_test.cpp
@@ -110,6 +110,9 @@ class TN_test : public testing::Test
 
 TEST_F(TN_test, TN_Solve_LinearEq)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     Solve(0);
     EXPECT_NEAR(x[0], 0.50000000000003430589, DOUBLETHRESHOLD);
     EXPECT_NEAR(x[1], -3.4028335704761047964e-14, DOUBLETHRESHOLD);
@@ -120,6 +123,9 @@ TEST_F(TN_test, TN_Solve_LinearEq)
 
 TEST_F(TN_test, TN_Min_Func)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     Solve(1);
     EXPECT_NEAR(x[0], 4.0049968540891525137, DOUBLETHRESHOLD);
     EXPECT_NEAR(x[1], 2.1208751163987624722, DOUBLETHRESHOLD);

From 0a40350d707d59a30310ecbb4cd460e5c7cbc58f Mon Sep 17 00:00:00 2001
From: critsium-xy <tsfxwbbzxy@163.com>
Date: Fri, 3 Jan 2025 14:17:31 +0800
Subject: [PATCH 12/15] return base_math_chebyshev

---
 source/module_base/test/CMakeLists.txt | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/source/module_base/test/CMakeLists.txt b/source/module_base/test/CMakeLists.txt
index ade269a1e6..0c8fd53461 100644
--- a/source/module_base/test/CMakeLists.txt
+++ b/source/module_base/test/CMakeLists.txt
@@ -133,7 +133,7 @@ AddTest(
 AddTest(
   TARGET base_math_chebyshev
   LIBS parameter  ${math_libs} base device container
-  SOURCES math_chebyshev_test.cpp ../math_chebyshev.cpp ../tool_quit.cpp ../global_variable.cpp ../timer.cpp ../global_file.cpp ../global_function.cpp ../memory.cpp ../parallel_reduce.cpp
+  SOURCES math_chebyshev_test.cpp
 )
 
 AddTest(
@@ -145,13 +145,13 @@ AddTest(
 AddTest(
   TARGET base_opt_CG
   LIBS parameter  ${math_libs} base device
-  SOURCES opt_CG_test.cpp opt_test_tools.cpp ../opt_CG.cpp ../opt_DCsrch.cpp ../global_variable.cpp ../parallel_reduce.cpp
+  SOURCES opt_CG_test.cpp opt_test_tools.cpp
 )
 
 AddTest(
   TARGET base_opt_TN
   LIBS parameter  ${math_libs} base device
-  SOURCES opt_TN_test.cpp opt_test_tools.cpp ../opt_CG.cpp ../opt_DCsrch.cpp ../global_variable.cpp ../parallel_reduce.cpp
+  SOURCES opt_TN_test.cpp opt_test_tools.cpp
 )
 
 AddTest(

From 2391a681f27807c172b5fb1b254a842bcb9e31f1 Mon Sep 17 00:00:00 2001
From: critsium-xy <tsfxwbbzxy@163.com>
Date: Fri, 3 Jan 2025 15:08:11 +0800
Subject: [PATCH 13/15] Fix MPI bug

---
 .../module_base/test/math_chebyshev_test.cpp  | 48 +++++++++++++++++++
 source/module_base/test/opt_CG_test.cpp       |  3 ++
 source/module_base/test/opt_test_tools.cpp    |  3 ++
 3 files changed, 54 insertions(+)

diff --git a/source/module_base/test/math_chebyshev_test.cpp b/source/module_base/test/math_chebyshev_test.cpp
index 3c3aef26c4..c937dc7fcf 100644
--- a/source/module_base/test/math_chebyshev_test.cpp
+++ b/source/module_base/test/math_chebyshev_test.cpp
@@ -107,6 +107,9 @@ class MathChebyshevTest : public testing::Test
 
 TEST_F(MathChebyshevTest, calcoef_real)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     auto fun_x6 = [&](double x) { return fun.x6(x); };
     auto fun_x7 = [&](double x) { return fun.x7(x); };
     p_chetest = new ModuleBase::Chebyshev<double>(10);
@@ -130,6 +133,9 @@ TEST_F(MathChebyshevTest, calcoef_real)
 
 TEST_F(MathChebyshevTest, calcoef_pair)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     auto fun_x6 = [&](double x) { return fun.x6(x); };
     auto fun_x7 = [&](double x) { return fun.x7(x); };
     p_chetest = new ModuleBase::Chebyshev<double>(10);
@@ -148,6 +154,9 @@ TEST_F(MathChebyshevTest, calcoef_pair)
 
 TEST_F(MathChebyshevTest, calcoef_complex)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     auto fun_expi = [&](std::complex<double> x) { return fun.expi(x); };
     const int norder = 100;
     const double PI = 3.14159265358979323846;
@@ -169,6 +178,9 @@ TEST_F(MathChebyshevTest, calcoef_complex)
 
 TEST_F(MathChebyshevTest, calfinalvec_real)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     const int norder = 100;
     const double E = 2.718281828459046;
     p_chetest = new ModuleBase::Chebyshev<double>(norder);
@@ -203,6 +215,9 @@ TEST_F(MathChebyshevTest, calfinalvec_real)
 
 TEST_F(MathChebyshevTest, calfinalvec_complex)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     const int norder = 100;
     const double E = 2.718281828459046;
     p_chetest = new ModuleBase::Chebyshev<double>(norder);
@@ -237,6 +252,9 @@ TEST_F(MathChebyshevTest, calfinalvec_complex)
 
 TEST_F(MathChebyshevTest, calpolyvec_complex)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     const int norder = 100;
     const double E = 2.718281828459046;
     p_chetest = new ModuleBase::Chebyshev<double>(norder);
@@ -284,6 +302,9 @@ TEST_F(MathChebyshevTest, calpolyvec_complex)
 
 TEST_F(MathChebyshevTest, tracepolyA)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     const int norder = 100;
     p_chetest = new ModuleBase::Chebyshev<double>(norder);
 
@@ -384,6 +405,9 @@ TEST_F(MathChebyshevTest, checkconverge)
 
 TEST_F(MathChebyshevTest, recurs)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     testing::internal::CaptureStdout();
     EXPECT_EXIT(ModuleBase::Chebyshev<double> noneche(0), ::testing::ExitedWithCode(1), "");
     std::string output = testing::internal::GetCapturedStdout();
@@ -399,6 +423,9 @@ TEST_F(MathChebyshevTest, recurs)
 #ifdef __ENABLE_FLOAT_FFTW
 TEST_F(MathChebyshevTest, calcoef_real_float)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     auto fun_x6f = [&](float x) { return fun.x6(x); };
     auto fun_x7f = [&](float x) { return fun.x7(x); };
     p_fchetest = new ModuleBase::Chebyshev<float>(10);
@@ -421,6 +448,9 @@ TEST_F(MathChebyshevTest, calcoef_real_float)
 
 TEST_F(MathChebyshevTest, calcoef_pair_float)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     auto fun_x6f = [&](float x) { return fun.x6(x); };
     auto fun_x7f = [&](float x) { return fun.x7(x); };
     p_fchetest = new ModuleBase::Chebyshev<float>(10);
@@ -439,6 +469,9 @@ TEST_F(MathChebyshevTest, calcoef_pair_float)
 
 TEST_F(MathChebyshevTest, calcoef_complex_float)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     auto fun_expif = [&](std::complex<float> x) { return fun.expi(x); };
     const int norder = 100;
     const float PI = 3.14159265358979323846;
@@ -460,6 +493,9 @@ TEST_F(MathChebyshevTest, calcoef_complex_float)
 
 TEST_F(MathChebyshevTest, calfinalvec_real_float)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     const int norder = 100;
     const float E = 2.718281828459046;
     p_fchetest = new ModuleBase::Chebyshev<float>(norder);
@@ -494,6 +530,9 @@ TEST_F(MathChebyshevTest, calfinalvec_real_float)
 
 TEST_F(MathChebyshevTest, calfinalvec_complex_float)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     const int norder = 100;
     const float E = 2.718281828459046;
     p_fchetest = new ModuleBase::Chebyshev<float>(norder);
@@ -528,6 +567,9 @@ TEST_F(MathChebyshevTest, calfinalvec_complex_float)
 
 TEST_F(MathChebyshevTest, calpolyvec_float)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     const int norder = 100;
     const float E = 2.718281828459046;
     p_fchetest = new ModuleBase::Chebyshev<float>(norder);
@@ -575,6 +617,9 @@ TEST_F(MathChebyshevTest, calpolyvec_float)
 
 TEST_F(MathChebyshevTest, tracepolyA_float)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     const int norder = 100;
     p_fchetest = new ModuleBase::Chebyshev<float>(norder);
 
@@ -624,6 +669,9 @@ TEST_F(MathChebyshevTest, tracepolyA_float)
 
 TEST_F(MathChebyshevTest, checkconverge_float)
 {
+#ifdef __MPI
+#undef __MPI
+#endif
     const int norder = 100;
     p_fchetest = new ModuleBase::Chebyshev<float>(norder);
 
diff --git a/source/module_base/test/opt_CG_test.cpp b/source/module_base/test/opt_CG_test.cpp
index d2faf8763b..2661d997f7 100644
--- a/source/module_base/test/opt_CG_test.cpp
+++ b/source/module_base/test/opt_CG_test.cpp
@@ -1,3 +1,6 @@
+#ifdef __MPI
+#undef __MPI
+#endif
 #include "gtest/gtest.h"
 #include "../opt_CG.h"
 #include "../opt_DCsrch.h"
diff --git a/source/module_base/test/opt_test_tools.cpp b/source/module_base/test/opt_test_tools.cpp
index 1c90b79bca..71e136b3ef 100644
--- a/source/module_base/test/opt_test_tools.cpp
+++ b/source/module_base/test/opt_test_tools.cpp
@@ -1,3 +1,6 @@
+#ifdef __MPI
+#undef __MPI
+#endif
 #include "./opt_test_tools.h"
 #include <math.h>
 

From 735329559f6e7c6824c7356e65eaa0a7eb0c7aa2 Mon Sep 17 00:00:00 2001
From: critsium-xy <tsfxwbbzxy@163.com>
Date: Fri, 3 Jan 2025 16:59:24 +0800
Subject: [PATCH 14/15] Finish

---
 .../module_base/test/math_chebyshev_test.cpp  | 51 +------------------
 source/module_base/test/opt_CG_test.cpp       | 15 ++++--
 source/module_base/test/opt_TN_test.cpp       |  6 ++-
 3 files changed, 16 insertions(+), 56 deletions(-)

diff --git a/source/module_base/test/math_chebyshev_test.cpp b/source/module_base/test/math_chebyshev_test.cpp
index c937dc7fcf..a7ea215266 100644
--- a/source/module_base/test/math_chebyshev_test.cpp
+++ b/source/module_base/test/math_chebyshev_test.cpp
@@ -107,9 +107,6 @@ class MathChebyshevTest : public testing::Test
 
 TEST_F(MathChebyshevTest, calcoef_real)
 {
-#ifdef __MPI
-#undef __MPI
-#endif
     auto fun_x6 = [&](double x) { return fun.x6(x); };
     auto fun_x7 = [&](double x) { return fun.x7(x); };
     p_chetest = new ModuleBase::Chebyshev<double>(10);
@@ -133,9 +130,6 @@ TEST_F(MathChebyshevTest, calcoef_real)
 
 TEST_F(MathChebyshevTest, calcoef_pair)
 {
-#ifdef __MPI
-#undef __MPI
-#endif
     auto fun_x6 = [&](double x) { return fun.x6(x); };
     auto fun_x7 = [&](double x) { return fun.x7(x); };
     p_chetest = new ModuleBase::Chebyshev<double>(10);
@@ -154,9 +148,6 @@ TEST_F(MathChebyshevTest, calcoef_pair)
 
 TEST_F(MathChebyshevTest, calcoef_complex)
 {
-#ifdef __MPI
-#undef __MPI
-#endif
     auto fun_expi = [&](std::complex<double> x) { return fun.expi(x); };
     const int norder = 100;
     const double PI = 3.14159265358979323846;
@@ -178,9 +169,6 @@ TEST_F(MathChebyshevTest, calcoef_complex)
 
 TEST_F(MathChebyshevTest, calfinalvec_real)
 {
-#ifdef __MPI
-#undef __MPI
-#endif
     const int norder = 100;
     const double E = 2.718281828459046;
     p_chetest = new ModuleBase::Chebyshev<double>(norder);
@@ -215,9 +203,6 @@ TEST_F(MathChebyshevTest, calfinalvec_real)
 
 TEST_F(MathChebyshevTest, calfinalvec_complex)
 {
-#ifdef __MPI
-#undef __MPI
-#endif
     const int norder = 100;
     const double E = 2.718281828459046;
     p_chetest = new ModuleBase::Chebyshev<double>(norder);
@@ -252,9 +237,6 @@ TEST_F(MathChebyshevTest, calfinalvec_complex)
 
 TEST_F(MathChebyshevTest, calpolyvec_complex)
 {
-#ifdef __MPI
-#undef __MPI
-#endif
     const int norder = 100;
     const double E = 2.718281828459046;
     p_chetest = new ModuleBase::Chebyshev<double>(norder);
@@ -302,9 +284,6 @@ TEST_F(MathChebyshevTest, calpolyvec_complex)
 
 TEST_F(MathChebyshevTest, tracepolyA)
 {
-#ifdef __MPI
-#undef __MPI
-#endif
     const int norder = 100;
     p_chetest = new ModuleBase::Chebyshev<double>(norder);
 
@@ -359,7 +338,6 @@ TEST_F(MathChebyshevTest, checkconverge)
 {
 #ifdef __MPI
 #undef __MPI
-#endif
     const int norder = 100;
     p_chetest = new ModuleBase::Chebyshev<double>(norder);
     auto fun_sigma_y
@@ -401,13 +379,12 @@ TEST_F(MathChebyshevTest, checkconverge)
 
     delete[] v;
     delete p_chetest;
+#define __MPI
+#endif
 }
 
 TEST_F(MathChebyshevTest, recurs)
 {
-#ifdef __MPI
-#undef __MPI
-#endif
     testing::internal::CaptureStdout();
     EXPECT_EXIT(ModuleBase::Chebyshev<double> noneche(0), ::testing::ExitedWithCode(1), "");
     std::string output = testing::internal::GetCapturedStdout();
@@ -423,9 +400,6 @@ TEST_F(MathChebyshevTest, recurs)
 #ifdef __ENABLE_FLOAT_FFTW
 TEST_F(MathChebyshevTest, calcoef_real_float)
 {
-#ifdef __MPI
-#undef __MPI
-#endif
     auto fun_x6f = [&](float x) { return fun.x6(x); };
     auto fun_x7f = [&](float x) { return fun.x7(x); };
     p_fchetest = new ModuleBase::Chebyshev<float>(10);
@@ -448,9 +422,6 @@ TEST_F(MathChebyshevTest, calcoef_real_float)
 
 TEST_F(MathChebyshevTest, calcoef_pair_float)
 {
-#ifdef __MPI
-#undef __MPI
-#endif
     auto fun_x6f = [&](float x) { return fun.x6(x); };
     auto fun_x7f = [&](float x) { return fun.x7(x); };
     p_fchetest = new ModuleBase::Chebyshev<float>(10);
@@ -469,9 +440,6 @@ TEST_F(MathChebyshevTest, calcoef_pair_float)
 
 TEST_F(MathChebyshevTest, calcoef_complex_float)
 {
-#ifdef __MPI
-#undef __MPI
-#endif
     auto fun_expif = [&](std::complex<float> x) { return fun.expi(x); };
     const int norder = 100;
     const float PI = 3.14159265358979323846;
@@ -493,9 +461,6 @@ TEST_F(MathChebyshevTest, calcoef_complex_float)
 
 TEST_F(MathChebyshevTest, calfinalvec_real_float)
 {
-#ifdef __MPI
-#undef __MPI
-#endif
     const int norder = 100;
     const float E = 2.718281828459046;
     p_fchetest = new ModuleBase::Chebyshev<float>(norder);
@@ -530,9 +495,6 @@ TEST_F(MathChebyshevTest, calfinalvec_real_float)
 
 TEST_F(MathChebyshevTest, calfinalvec_complex_float)
 {
-#ifdef __MPI
-#undef __MPI
-#endif
     const int norder = 100;
     const float E = 2.718281828459046;
     p_fchetest = new ModuleBase::Chebyshev<float>(norder);
@@ -567,9 +529,6 @@ TEST_F(MathChebyshevTest, calfinalvec_complex_float)
 
 TEST_F(MathChebyshevTest, calpolyvec_float)
 {
-#ifdef __MPI
-#undef __MPI
-#endif
     const int norder = 100;
     const float E = 2.718281828459046;
     p_fchetest = new ModuleBase::Chebyshev<float>(norder);
@@ -617,9 +576,6 @@ TEST_F(MathChebyshevTest, calpolyvec_float)
 
 TEST_F(MathChebyshevTest, tracepolyA_float)
 {
-#ifdef __MPI
-#undef __MPI
-#endif
     const int norder = 100;
     p_fchetest = new ModuleBase::Chebyshev<float>(norder);
 
@@ -669,9 +625,6 @@ TEST_F(MathChebyshevTest, tracepolyA_float)
 
 TEST_F(MathChebyshevTest, checkconverge_float)
 {
-#ifdef __MPI
-#undef __MPI
-#endif
     const int norder = 100;
     p_fchetest = new ModuleBase::Chebyshev<float>(norder);
 
diff --git a/source/module_base/test/opt_CG_test.cpp b/source/module_base/test/opt_CG_test.cpp
index 2661d997f7..ef16ee4a0d 100644
--- a/source/module_base/test/opt_CG_test.cpp
+++ b/source/module_base/test/opt_CG_test.cpp
@@ -148,64 +148,69 @@ TEST_F(CG_test, Stand_Solve_LinearEq)
 {
 #ifdef __MPI
 #undef __MPI
-#endif
     CG_Solve_LinearEq();
     EXPECT_NEAR(x[0], 0.5, DOUBLETHRESHOLD);
     EXPECT_NEAR(x[1], 1.6429086563584579739e-18, DOUBLETHRESHOLD);
     EXPECT_NEAR(x[2], 1.5, DOUBLETHRESHOLD);
     ASSERT_EQ(final_iter, 4);
     ASSERT_EQ(cg.get_iter(), 4);
+#define __MPI
+#endif
 }
 
 TEST_F(CG_test, PR_Solve_LinearEq)
 {
 #ifdef __MPI
 #undef __MPI
-#endif
     Solve(1, 0);
     EXPECT_NEAR(x[0], 0.50000000000003430589, DOUBLETHRESHOLD);
     EXPECT_NEAR(x[1], -3.4028335704761047964e-14, DOUBLETHRESHOLD);
     EXPECT_NEAR(x[2], 1.5000000000000166533, DOUBLETHRESHOLD);
     ASSERT_EQ(final_iter, 3);
     ASSERT_EQ(cg.get_iter(), 3);
+#define __MPI
+#endif
 }
 
 TEST_F(CG_test, HZ_Solve_LinearEq)
 {
 #ifdef __MPI
 #undef __MPI
-#endif
     Solve(2, 0);
     EXPECT_NEAR(x[0], 0.49999999999999944489, DOUBLETHRESHOLD);
     EXPECT_NEAR(x[1], -9.4368957093138305936e-16, DOUBLETHRESHOLD);
     EXPECT_NEAR(x[2], 1.5000000000000011102, DOUBLETHRESHOLD);
     ASSERT_EQ(final_iter, 3);
     ASSERT_EQ(cg.get_iter(), 3);
+#define __MPI
+#endif
 }
 
 TEST_F(CG_test, PR_Min_Func)
 {
 #ifdef __MPI
 #undef __MPI
-#endif
     Solve(1, 1);
     EXPECT_NEAR(x[0], 4.0006805979150792396, DOUBLETHRESHOLD);
     EXPECT_NEAR(x[1], 2.0713759992720870429, DOUBLETHRESHOLD);
     EXPECT_NEAR(x[2], 9.2871067233169171118, DOUBLETHRESHOLD);
     ASSERT_EQ(final_iter, 18);
     ASSERT_EQ(cg.get_iter(), 18);
+#define __MPI
+#endif
 }
 
 TEST_F(CG_test, HZ_Min_Func)
 {
 #ifdef __MPI
 #undef __MPI
-#endif
     Solve(2, 1);
     EXPECT_NEAR(x[0], 4.0006825378033568086, DOUBLETHRESHOLD);
     EXPECT_NEAR(x[1], 2.0691732100663737803, DOUBLETHRESHOLD);
     EXPECT_NEAR(x[2], 9.2780872787668311474, DOUBLETHRESHOLD);
     ASSERT_EQ(final_iter, 18);
     ASSERT_EQ(cg.get_iter(), 18);
+#define __MPI
+#endif
 }
 // g++ -std=c++11 ../opt_CG.cpp ../opt_DCsrch.cpp ./CG_test.cpp ./test_tools.cpp  -lgtest -lpthread -lgtest_main -o test.exe
\ No newline at end of file
diff --git a/source/module_base/test/opt_TN_test.cpp b/source/module_base/test/opt_TN_test.cpp
index b136194881..ee3b10ad7b 100644
--- a/source/module_base/test/opt_TN_test.cpp
+++ b/source/module_base/test/opt_TN_test.cpp
@@ -112,24 +112,26 @@ TEST_F(TN_test, TN_Solve_LinearEq)
 {
 #ifdef __MPI
 #undef __MPI
-#endif
     Solve(0);
     EXPECT_NEAR(x[0], 0.50000000000003430589, DOUBLETHRESHOLD);
     EXPECT_NEAR(x[1], -3.4028335704761047964e-14, DOUBLETHRESHOLD);
     EXPECT_NEAR(x[2], 1.5000000000000166533, DOUBLETHRESHOLD);
     ASSERT_EQ(final_iter, 1);
     ASSERT_EQ(tn.get_iter(), 1);
+#define __MPI
+#endif
 }
 
 TEST_F(TN_test, TN_Min_Func)
 {
 #ifdef __MPI
 #undef __MPI
-#endif
     Solve(1);
     EXPECT_NEAR(x[0], 4.0049968540891525137, DOUBLETHRESHOLD);
     EXPECT_NEAR(x[1], 2.1208751163987624722, DOUBLETHRESHOLD);
     EXPECT_NEAR(x[2], 9.4951527720891863993, DOUBLETHRESHOLD);
     ASSERT_EQ(final_iter, 6);
     ASSERT_EQ(tn.get_iter(), 6);
+#define __MPI
+#endif
 }
\ No newline at end of file

From d03b7b69b074bc730bc94e61e275dc4c08b70b73 Mon Sep 17 00:00:00 2001
From: "pre-commit-ci-lite[bot]"
 <117423508+pre-commit-ci-lite[bot]@users.noreply.github.com>
Date: Fri, 3 Jan 2025 09:38:55 +0000
Subject: [PATCH 15/15] [pre-commit.ci lite] apply automatic fixes

---
 source/module_base/test/opt_CG_test.cpp | 23 ++++++++++++++---------
 source/module_base/test/opt_TN_test.cpp | 18 +++++++++++-------
 2 files changed, 25 insertions(+), 16 deletions(-)

diff --git a/source/module_base/test/opt_CG_test.cpp b/source/module_base/test/opt_CG_test.cpp
index ef16ee4a0d..b8abeb5760 100644
--- a/source/module_base/test/opt_CG_test.cpp
+++ b/source/module_base/test/opt_CG_test.cpp
@@ -21,10 +21,10 @@ class CG_test : public testing::Test
     double residual = 10.;
     double tol = 1e-5;
     int final_iter = 0;
-    char *task = NULL;
-    double *Ap = NULL;
-    double *p = NULL;
-    double *x = NULL;
+    char *task = nullptr;
+    double *Ap = nullptr;
+    double *p = nullptr;
+    double *x = nullptr;
 
     void SetUp()
     {
@@ -68,7 +68,8 @@ class CG_test : public testing::Test
             tools.le.get_Ap(tools.le.A, p, Ap);
             int ifPD = 0;
             step = cg.step_length(Ap, p, ifPD);
-            for (int i = 0; i < 3; ++i) x[i] += step * p[i]; 
+            for (int i = 0; i < 3; ++i) { x[i] += step * p[i]; 
+}
             residual = cg.get_residual();
         }
     }
@@ -105,14 +106,16 @@ class CG_test : public testing::Test
         {
             tools.dfuncdx(x, gradient, func_label);
             residual = 0;
-            for (int i = 0; i<3 ;++i) residual += gradient[i] * gradient[i];
+            for (int i = 0; i<3 ;++i) { residual += gradient[i] * gradient[i];
+}
             if (residual < tol) 
             {
                 final_iter = iter;
                 break;
             }
             cg.next_direct(gradient, cg_label, p);
-            for (int i = 0; i < 3; ++i) temp_x[i] = x[i];
+            for (int i = 0; i < 3; ++i) { temp_x[i] = x[i];
+}
             task[0] = 'S'; task[1] = 'T'; task[2] = 'A'; task[3] = 'R'; task[4] = 'T';
             while (true)
             {
@@ -121,7 +124,8 @@ class CG_test : public testing::Test
                 ds.dcSrch(f, g, step, task);
                 if (task[0] == 'F' && task[1] == 'G')
                 {
-                    for (int j = 0; j < 3; ++j) temp_x[j] = x[j] + step * p[j];
+                    for (int j = 0; j < 3; ++j) { temp_x[j] = x[j] + step * p[j];
+}
                     continue;
                 }
                 else if (task[0] == 'C' && task[1] == 'O')
@@ -137,7 +141,8 @@ class CG_test : public testing::Test
                     break;
                 }
             }
-            for (int i = 0; i < 3; ++i) x[i] += step * p[i];
+            for (int i = 0; i < 3; ++i) { x[i] += step * p[i];
+}
         }
         delete[] temp_x;
         delete[] gradient;
diff --git a/source/module_base/test/opt_TN_test.cpp b/source/module_base/test/opt_TN_test.cpp
index ee3b10ad7b..1fc5b7f2d6 100644
--- a/source/module_base/test/opt_TN_test.cpp
+++ b/source/module_base/test/opt_TN_test.cpp
@@ -17,9 +17,9 @@ class TN_test : public testing::Test
     double tol = 1e-5;
     int final_iter = 0;
     int flag = 0;
-    char *task = NULL;
-    double *p = NULL;
-    double *x = NULL;
+    char *task = nullptr;
+    double *p = nullptr;
+    double *x = nullptr;
 
     void SetUp()
     {
@@ -61,7 +61,8 @@ class TN_test : public testing::Test
         {
             tools.dfuncdx(x, gradient, func_label);
             residual = 0;
-            for (int i = 0; i<3 ;++i) residual += gradient[i] * gradient[i];
+            for (int i = 0; i<3 ;++i) { residual += gradient[i] * gradient[i];
+}
             if (residual < tol) 
             {
                 final_iter = iter;
@@ -75,7 +76,8 @@ class TN_test : public testing::Test
             {
                 tn.next_direct(x, gradient, flag, p, &(tools.mf), &ModuleESolver::ESolver_OF::dfuncdx);
             }
-            for (int i = 0; i < 3; ++i) temp_x[i] = x[i];
+            for (int i = 0; i < 3; ++i) { temp_x[i] = x[i];
+}
             task[0] = 'S'; task[1] = 'T'; task[2] = 'A'; task[3] = 'R'; task[4] = 'T';
             while (true)
             {
@@ -84,7 +86,8 @@ class TN_test : public testing::Test
                 ds.dcSrch(f, g, step, task);
                 if (task[0] == 'F' && task[1] == 'G')
                 {
-                    for (int j = 0; j < 3; ++j) temp_x[j] = x[j] + step * p[j];
+                    for (int j = 0; j < 3; ++j) { temp_x[j] = x[j] + step * p[j];
+}
                     continue;
                 }
                 else if (task[0] == 'C' && task[1] == 'O')
@@ -100,7 +103,8 @@ class TN_test : public testing::Test
                     break;
                 }
             }
-            for (int i = 0; i < 3; ++i) x[i] += step * p[i];
+            for (int i = 0; i < 3; ++i) { x[i] += step * p[i];
+}
         }
         delete[] temp_x;
         delete[] gradient;