问题描述
我正在尝试在Linux Ubuntu上运行caffe。安装后，我在gpu中运行caffe，错误是
I0910 13:28:13.606891 10629 caffe.cpp:296] Use GPU with device ID 0
modprobe: ERROR: could not insert 'nvidia_352': No such device
F0910 13:28:13.728612 10629 common.cpp:142] Check failed: error == cudaSuccess (38 vs. 0) no CUDA-capable device is detected
*** Check failure stack trace: ***
@ 0x7ffd3b9a7daa (unknown)
@ 0x7ffd3b9a7ce4 (unknown)
@ 0x7ffd3b9a76e6 (unknown)
@ 0x7ffd3b9aa687 (unknown)
@ 0x7ffd3bf91cb5 caffe::Caffe::SetDevice()
@ 0x40a5a7 time()
@ 0x4080f8 main
@ 0x7ffd3aeb9ec5 (unknown)
@ 0x408618 (unknown)
@ (nil) (unknown)
Aborted (core dumped)
我的NVIDIA驱动程序是352.41。我安装了352，并且安装了最新版本。
sudo apt-get install nvidia-352[sudo]
Reading package lists... Done
Building dependency tree
Reading state information... Done
nvidia-352 is already the newest version.
The following packages were automatically installed and are no longer required:
account-plugin-windows-live libupstart1
Use 'apt-get autoremove' to remove them.
0 upgraded, 0 newly installed, 0 to remove and 31 not upgraded.
我的Ubuntu具有NVIDIA驱动程序352，为什么我会出现类似错误
I0910 13:28:13.606891 10629 caffe.cpp:296] Use GPU with device ID 0
modprobe: ERROR: could not insert 'nvidia_352': No such device
F0910 13:28:13.728612 10629 common.cpp:142] Check failed: error == cudaSuccess (38 vs. 0) no CUDA-capable device is detected
我检查了我是否有支持CUDA的设备，例如
lspci | grep -i nvidia
05:00.0 VGA compatible controller: NVIDIA Corporation GK107GL [Quadro K2000] (rev a1)
05:00.1 Audio device: NVIDIA Corporation GK107 HDMI Audio Controller (rev a1)
我有支持CUDA的设备，为什么会出现错误？
编辑1：是的，我对./deviceQuery的测试失败。
../NVIDIA_CUDA-7.5_Samples/bin/x86_64/linux/release/deviceQuery Starting...
CUDA Device Query (Runtime API) version (CUDART static linking)
cudaGetDeviceCount returned 38
-> no CUDA-capable device is detected
Result = FAIL
我在dev /文件夹中签到，我有nvidia0。
crwxrwxrwx 1 root root 195, 0 Sep 10 16:51 nvidia0
crw-rw-rw- 1 root root 195, 255 Sep 10 16:51 nvidiactl
我的nvcc -V检查给了我
li@li-HP-Z420-Workstation:/dev$ nvcc -V
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2015 NVIDIA Corporation
Built on Tue_Aug_11_14:27:32_CDT_2015
Cuda compilation tools, release 7.5, V7.5.17
然后我的版本检查
li@li-HP-Z420-Workstation:/dev$ cat /proc/driver/nvidia/version
NVRM version: NVIDIA UNIX x86_64 Kernel Module 352.41 Fri Aug 21 23:09:52 PDT 2015
GCC version: gcc version 4.8.4 (Ubuntu 4.8.4-2ubuntu1~14.04)
有什么事吗
最佳方案
现在问题解决了。我检查发现sudo dpkg --list | grep nvidia，因为我的内核具有352.41，但是客户端具有304.12。所以我做了sudo apt-get remove --purge nvidia-*。它删除了所有软件包。然后，将352.41安装为
$ sudo add-apt-repository ppa:xorg-edgers/ppa -y
$ sudo apt-get update
$ sudo apt-get install nvidia-352
之后
$ sudo dpkg --list | grep nvidia
rc nvidia-304 304.128-0ubuntu0~gpu14.04.2 amd64 NVIDIA legacy binary driver - version 304.128
rc nvidia-304-updates 304.125-0ubuntu0.0.2 amd64 NVIDIA legacy binary driver - version 304.125
ii nvidia-352 352.41-0ubuntu0~gpu14.04.1 amd64 NVIDIA binary driver - version 352.41
rc nvidia-opencl-icd-304 304.128-0ubuntu0~gpu14.04.2 amd64 NVIDIA OpenCL ICD
rc nvidia-opencl-icd-304-updates 304.125-0ubuntu0.0.2 amd64 NVIDIA OpenCL ICD
ii nvidia-opencl-icd-352 352.41-0ubuntu0~gpu14.04.1 amd64 NVIDIA OpenCL ICD
ii nvidia-prime 0.6.2 amd64 Tools to enable NVIDIA's Prime
ii nvidia-settings 355.11-0ubuntu0~gpu14.04.1 amd64 Tool for configuring the NVIDIA graphics driver
现在版本匹配。然后./deviceQuery和所有按预期工作。谢谢
次佳方案
我可以做的另一种方法是使用.run文件进行安装。那需要先杀死X服务器。 X服务器被杀死，如下所示。
Make sure you are logged out.
Hit CTRL+ALT+F1 and login using your credentials.
kill your current X server session by typing sudo service lightdm stop or sudo stop lightdm
Enter runlevel 3 (or 5) by typing sudo init 3 (or sudo init 5) and install your .run file.
You might be required to reboot when the installation finishes. If not, run sudo service start lightdm or sudo start lightdm to start your X server again.
然后run .run file as sudo sh xxxxx.run
您可能会收到The distribution-provided pre-install script failed! Are you sure you want to continue?错误。然后中止安装并
disable the "Nouveau kernel driver" as sudo update-initramfs -u
然后重新启动系统和redo stop X server, enter runlevel 3 and do sudo sh xxxx.run again.
这次，您可以忽略该消息，并继续执行该失败的规定消息。然后，您将可以从.run文件安装Nvidia驱动程序。