Jetson nano 执行GPU预测报错：Error Message Summary:FatalError:' Segmentation fault' is detected by the operating system.

[Irvingao]

• 版本、环境信息
  1）PaddleHub和PaddlePaddle版本：PaddleHub：1.8.0，PaddlePaddle-gpu：2.0.0
  2）系统环境：Jetson nano Ubuntu18.04，python3.6.9
• 复现信息：

from jetcam.csi_camera import CSICamera
import paddlehub as hub
import cv2
import time

def draw_bbox_image(img, data_list):
    for boxes in data_list:
        print(boxes)
        left = int(boxes["left"] * img.shape[1])
        right = int(boxes["right"] * img.shape[1])
        top = int(boxes["top"] * img.shape[0])
        bottom = int(boxes["bottom"] * img.shape[0])
        confidence = str(boxes["confidence"])
        cv2.rectangle(img, (left, top), (right, bottom), (0,255,0), 2)
        cv2.putText(img, confidence, (left, top-2), cv2.FONT_HERSHEY_SIMPLEX, 0.5, (0,255,0), 2)

if __name__ == "__main__":
    camera = CSICamera(capture_device=0, width=224, height=224)

    object_detector = hub.Module(name="ssd_mobilenet_v1_pascal")
    
    while True:
        frame = camera.read()

        tim1 = time.time()
        result = object_detector.object_detection(images=[frame], use_gpu=True)
        try:
            data_list = result[0]["data"]
            draw_bbox_image(frame, data_list)
            print("data_list:", data_list)
        except IndexError:
            pass
        cv2.imshow("result", frame)
        tim2 = time.time()
        tim = tim2 - tim1
        print("tim", tim)
        print("result", result)
        if cv2.waitKey(1) & 0xFF == ord('q'):
                break

[Irvingao]

采用单张图片

CPU预测正常：

GPU预测报错：

[haoyuying]

请您参考#1301
查看自己的cuda和cudnn及paddlepaddle-gpu三者之间的版本是否匹配，一般出现segmentation fault一般是由于cudnn和cuda版本及paddlepaddle-gpu版本不匹配导致的

[jiafuzeng]