#if defined(GGML_USE_HIPBLAS)
#include "hipblas/hipblas.h"
#include "hip/hip_runtime.h"
#define CUBLAS_COMPUTE_32F HIPBLAS_R_32F
#define CUBLAS_COMPUTE_32F_FAST_16F HIPBLAS_R_32F
#define CUBLAS_GEMM_DEFAULT HIPBLAS_GEMM_DEFAULT
#define CUBLAS_OP_N HIPBLAS_OP_N
#define CUBLAS_OP_T HIPBLAS_OP_T
#define CUBLAS_STATUS_SUCCESS HIPBLAS_STATUS_SUCCESS
#define CUBLAS_TF32_TENSOR_OP_MATH 0
#define CUDA_R_16F  HIPBLAS_R_16F
#define CUDA_R_32F  HIPBLAS_R_32F
#define cublasCreate hipblasCreate
#define cublasGemmEx hipblasGemmEx
#define cublasHandle_t hipblasHandle_t
#define cublasSetMathMode(h, m) HIPBLAS_STATUS_SUCCESS
#define cublasSetStream hipblasSetStream
#define cublasSgemm hipblasSgemm
#define cublasStatus_t hipblasStatus_t
#define cudaDeviceSynchronize hipDeviceSynchronize
#define cudaError_t hipError_t
#define cudaEventCreateWithFlags hipEventCreateWithFlags
#define cudaEventDisableTiming hipEventDisableTiming
#define cudaEventRecord hipEventRecord
#define cudaEvent_t hipEvent_t
#define cudaFree hipFree
#define cudaFreeHost hipFreeHost
#define cudaGetErrorString hipGetErrorString
#define cudaGetLastError hipGetLastError
#define cudaMalloc hipMalloc
#define cudaMallocHost hipMallocHost
#define cudaMemcpy2DAsync hipMemcpy2DAsync
#define cudaMemcpyAsync hipMemcpyAsync
#define cudaMemcpyDeviceToHost hipMemcpyDeviceToHost
#define cudaMemcpyHostToDevice hipMemcpyHostToDevice
#define cudaStreamCreateWithFlags hipStreamCreateWithFlags
#define cudaStreamNonBlocking hipStreamNonBlocking
#define cudaStreamSynchronize hipStreamSynchronize
#define cudaStreamWaitEvent hipStreamWaitEvent
#define cudaStream_t hipStream_t
#define cudaSuccess hipSuccess
#define GGML_USE_CUBLAS
#else
#include <cublas_v2.h>
#include <cuda_runtime.h>
#endif
#include "ggml.h"

#ifdef  __cplusplus
extern "C" {
#endif

void   ggml_init_cublas(void);

bool   ggml_cuda_can_mul_mat(const struct ggml_tensor * src0, const struct ggml_tensor * src1, struct ggml_tensor * dst);
size_t ggml_cuda_mul_mat_get_wsize(const struct ggml_tensor * src0, const struct ggml_tensor * src1, struct ggml_tensor * dst);
void   ggml_cuda_mul_mat(const struct ggml_tensor * src0, const struct ggml_tensor * src1, struct ggml_tensor * dst, void * wdata, size_t wsize);

// TODO: export these with GGML_API
void * ggml_cuda_host_malloc(size_t size);
void   ggml_cuda_host_free(void * ptr);

#ifdef  __cplusplus
}
#endif