#include "op_plugin/AclOpsInterface.h"
#include "op_plugin/OpApiInterface.h"
#include "op_plugin/utils/op_api_common.h"
namespace op_api {
using npu_preparation = at_npu::native::OpPreparation;
at::Tensor& gt_out(const at::Tensor& self, const at::Scalar& other, at::Tensor& result)
{
DO_COMPATIBILITY(aclnnGtScalar, acl_op::gt_out(self, other, result));
auto output_size = self.sizes();
npu_preparation::check_tensor({self}, result, output_size);
EXEC_NPU_CMD(aclnnGtScalar, self, other, result);
return result;
}
at::Tensor gt(const at::Tensor& self, const at::Scalar& other)
{
DO_COMPATIBILITY(aclnnGtScalar, acl_op::gt(self, other));
auto output_size = op_infer::input_same_output_size(self);
at::Tensor result = npu_preparation::apply_tensor_without_format(output_size, self.options().dtype(at::kBool));
EXEC_NPU_CMD(aclnnGtScalar, self, other, result);
return result;
}
at::Tensor& gt_(at::Tensor& self, const at::Scalar& other)
{
DO_COMPATIBILITY(aclnnInplaceGtScalar, acl_op::gt_(self, other));
EXEC_NPU_CMD(aclnnInplaceGtScalar, self, other);
return self;
}
at::Tensor& gt_out(const at::Tensor& self, const at::Tensor& other, at::Tensor& result)
{
DO_COMPATIBILITY(aclnnGtTensor, acl_op::gt_out(self, other, result));
auto output_size = op_infer::broadcast_ops_npu_output_size(self, other);
npu_preparation::check_tensor({self, other}, result, output_size);
if (npu_preparation::IsCPUScalar(self)) {
const at::Scalar self_scalar = self.item();
EXEC_NPU_CMD(aclnnLtScalar, other, self_scalar, result);
} else if (npu_preparation::IsCPUScalar(other)) {
const at::Scalar other_scalar = other.item();
EXEC_NPU_CMD(aclnnGtScalar, self, other_scalar, result);
} else {
EXEC_NPU_CMD(aclnnGtTensor, self, other, result);
}
return result;
}
at::Tensor gt(const at::Tensor& self, const at::Tensor& other)
{
DO_COMPATIBILITY(aclnnGtTensor, acl_op::gt(self, other));
auto output_size = op_infer::broadcast_ops_npu_output_size(self, other);
at::Tensor result = npu_preparation::apply_tensor_without_format(output_size, self.options().dtype(at::kBool));
if (npu_preparation::IsCPUScalar(self)) {
const at::Scalar self_scalar = self.item();
EXEC_NPU_CMD(aclnnLtScalar, other, self_scalar, result);
} else if (npu_preparation::IsCPUScalar(other)) {
const at::Scalar other_scalar = other.item();
EXEC_NPU_CMD(aclnnGtScalar, self, other_scalar, result);
} else {
EXEC_NPU_CMD(aclnnGtTensor, self, other, result);
}
return result;
}
at::Tensor& gt_(at::Tensor &self, const at::Tensor &other)
{
DO_COMPATIBILITY(aclnnInplaceGtTensor, acl_op::gt_(self, other));
if (other.dim() == 0 && !torch_npu::utils::is_npu(other)) {
return op_api::gt_(self, other.item());
} else {
TORCH_CHECK(self.device() == other.device(),
"Expected all tensors to be on the same device, but found at least two devices", OPS_ERROR(ErrCode::INTERNAL));
npu_preparation::CheckMemory({self, other}, {self});
EXEC_NPU_CMD(aclnnInplaceGtTensor, self, other);
return self;
}
}
}