在 dmesg 中找到与 GPU 相关的日志后，我尝试附加到正在运行的进程（python3）。ex) 西 13、31、45
但由于以下消息，我无法获得任何线索。有谁知道这个消息是什么意思？
无法写入扩展状态状态：地址错误。
已经在调试程序。杀了它？
$ cuda-gdb python3 1243
NVIDIA (R) CUDA Debugger
10.2 release
Portions Copyright (C) 2007-2019 NVIDIA Corporation
GNU gdb (GDB) 7.12
Copyright (C) 2016 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
(omit)
Reading symbols from /usr/lib64/libcuda.so.1...(no debugging symbols found)...done.
Reading symbols from /usr/local/lib64/python3.6/site-packages/PIL/_imaging.cpython-36m-x86_64-linux-gnu.so...(no debugging symbols found)...done.
Reading symbols from /usr/local/lib64/python3.6/site-packages/PIL/../Pillow.libs/libjpeg-ba7bf5af.so.9.4.0...(no debugging symbols found)...done.
Reading symbols from /usr/local/lib64/python3.6/site-packages/PIL/../Pillow.libs/libopenjp2-b3d7668a.so.2.3.1...(no debugging symbols found)...done.
Reading symbols from /usr/local/lib64/python3.6/site-packages/PIL/../Pillow.libs/libtiff-41910f6d.so.5.5.0...(no debugging symbols found)...done.
Reading symbols from /usr/local/lib64/python3.6/site-packages/PIL/../Pillow.libs/./liblzma-99449165.so.5.2.5...(no debugging symbols found)...done.
0x00007ffcd4ed4980 in clock_gettime ()
Couldn't write extended state status: Bad address.
A program is being debugged already. Kill it? (y or n) y
/home1/irteam/apps/pytorch-app/src/1243: No such file or directory.
(cuda-gdb) bt
No stack.
(cuda-gdb) exit