华为NPU卡ubuntu驱动安装记录

前奏：文中的绝大多数情况都是因为服务器没有网络服务，如果读者的电脑有完全的网络服务可能参考意义不是很大。不过，处理的方法和排查的思路可以作为一个借鉴。

root@ubuntu:/home/ubuntu# bash A300-3010-npu-driver_20.2.0_ubuntu18.04-x86_64.run
Verifying archive integrity... 100% SHA256 checksums are OK. All good.
Uncompressing Atlas300_software_version_20.2.0 100%
[Driver] [2021-05-31 10:50:35] [ERROR]ERR_NO:0x0004;ERR_DES: Unrecognized parameters. Try './xxx.run --help' for more information.
root@ubuntu:/home/ubuntu# ./A300-3010-npu-driver_20.2.0_ubuntu18.04-x86_64.run
Verifying archive integrity... 100% SHA256 checksums are OK. All good.
Uncompressing Atlas300_software_version_20.2.0 100%
[Driver] [2021-05-31 10:51:36] [ERROR]ERR_NO:0x0004;ERR_DES: Unrecognized parameters. Try './xxx.run --help' for more information.

chmod+x

root@ubuntu:/home/ubuntu#
root@ubuntu:/home/ubuntu# chmod +x A300-3010-npu-driver_20.2.0_ubuntu18.04-x86_64.run

gcc

root@ubuntu:/home/ubuntu# ./A300-3010-npu-driver_20.2.0_ubuntu18.04-x86_64.run --full
Verifying archive integrity... 100% SHA256 checksums are OK. All good.
Uncompressing Atlas300_software_version_20.2.0 100%
[Driver] [2021-05-31 11:19:50] [INFO]Start time: 2021-05-31 11:19:50
[Driver] [2021-05-31 11:19:50] [INFO]LogFile: /var/log/ascend_seclog/ascend_install.log
[Driver] [2021-05-31 11:19:50] [INFO]OperationLogFile: /var/log/ascend_seclog/operation.log
[Driver] [2021-05-31 11:19:51] [INFO]base version is none.
[Driver] [2021-05-31 11:19:51] [INFO]set username and usergroup, HwHiAiUser:HwHiAiUser
[ERROR]gcc: command not found
[Driver] [2021-05-31 11:19:52] [ERROR]Drv_dkms_env_check failed, details in : /var/log/ascend_seclog/ascend_install.log
[Driver] [2021-05-31 11:19:52] [INFO]Install driver failed, please retry after uninstall and reboot!
[Driver] [2021-05-31 11:19:52] [INFO]End time: 2021-05-31 11:19:52

update

#如果有网络：
根据提示输入apt-get update
更新完成后再重新安装gcc

输入代码gcc --version查看gcc版本，就成功啦！
root@ubuntu:/home/ubuntu# apt install gcc
Reading package lists... Done
Building dependency tree
Reading state information... Done
The following additional packages will be installed:
binutils binutils-common binutils-x86-64-linux-gnu cpp cpp-7 gcc-7 gcc-7-base libasan4 libatomic1 libbinutils libc-dev-bin libc6-dev libcc1-0 libcilkrts5
libgcc-7-dev libgomp1 libisl19 libitm1 liblsan0 libmpc3 libmpx2 libquadmath0 libtsan0 libubsan0 linux-libc-dev manpages-dev
Suggested packages:
binutils-doc cpp-doc gcc-7-locales gcc-multilib make autoconf automake libtool flex bison gdb gcc-doc gcc-7-multilib gcc-7-doc libgcc1-dbg libgomp1-dbg
libitm1-dbg libatomic1-dbg libasan4-dbg liblsan0-dbg libtsan0-dbg libubsan0-dbg libcilkrts5-dbg libmpx2-dbg libquadmath0-dbg glibc-doc
The following NEW packages will be installed:
binutils binutils-common binutils-x86-64-linux-gnu cpp cpp-7 gcc gcc-7 gcc-7-base libasan4 libatomic1 libbinutils libc-dev-bin libc6-dev libcc1-0 libcilkrts5
libgcc-7-dev libgomp1 libisl19 libitm1 liblsan0 libmpc3 libmpx2 libquadmath0 libtsan0 libubsan0 linux-libc-dev manpages-dev
0 upgraded, 27 newly installed, 0 to remove and 0 not upgraded.
Need to get 30.6 MB of archives.
After this operation, 118 MB of additional disk space will be used.
Do you want to continue? [Y/n] y
Ign:1 http://archive.ubuntu.com/ubuntu bionic-updates/main amd64 binutils-common amd64 2.30-21ubuntu1~18.04.4
Ign:2 http://archive.ubuntu.com/ubuntu bionic-updates/main amd64 libbinutils amd64 2.30-21ubuntu1~18.04.4
Ign:3 http://archive.ubuntu.com/ubuntu bionic-updates/main amd64 binutils-x86-64-linux-gnu amd64 2.30-21ubuntu1~18.04.4
Ign:4 http://archive.ubuntu.com/ubuntu bionic-updates/main amd64 binutils amd64 2.30-21ubuntu1~18.04.4
Ign:5 http://archive.ubuntu.com/ubuntu bionic-updates/main amd64 gcc-7-base amd64 7.5.0-3ubuntu1~18.04
Err:6 http://archive.ubuntu.com/ubuntu bionic/main amd64 libisl19 amd64 0.19-1

deb包现在安装

#服务器没有网络连接：
下载了一堆deb包
root@ubuntu:/home/ubuntu/ubuntu_packages# dpkg -i gcc-7-base_7.5.0-3ubuntu1~18.04_amd64.deb

进入目录之后：
sudo dpkg -i ./*

root@ubuntu:/home/ubuntu# ./A300-3010-npu-driver_20.2.0_ubuntu18.04-x86_64.run --full
Verifying archive integrity... 100% SHA256 checksums are OK. All good.
Uncompressing Atlas300_software_version_20.2.0 100%
[Driver] [2021-05-31 13:39:54] [INFO]Start time: 2021-05-31 13:39:54
[Driver] [2021-05-31 13:39:54] [INFO]LogFile: /var/log/ascend_seclog/ascend_install.log
[Driver] [2021-05-31 13:39:54] [INFO]OperationLogFile: /var/log/ascend_seclog/operation.log
[Driver] [2021-05-31 13:39:54] [INFO]base version is none.
[Driver] [2021-05-31 13:39:54] [INFO]set username and usergroup, HwHiAiUser:HwHiAiUser
[ERROR]make: command not found
[Driver] [2021-05-31 13:39:55] [ERROR]Drv_dkms_env_check failed, details in : /var/log/ascend_seclog/ascend_install.log
[Driver] [2021-05-31 13:39:55] [INFO]Install driver failed, please retry after uninstall and reboot!
[Driver] [2021-05-31 13:39:55] [INFO]End time: 2021-05-31 13:39:55

make

root@ubuntu:/home/ubuntu# make
Command 'make' not found, but can be installed with:
apt install make
apt install make-guile

下载make包：
make_install.tar
解压：
tar -xvf make_install.tar
继续：
sudo dpkg -i ./*
[ERROR]make: command not found
[Driver] [2021-05-31 13:55:41] [ERROR]Drv_dkms_env_check failed, details in : /var/log/ascend_seclog/ascend_install.log

查看日志：
/var/log/ascend_seclog/ascend_install.log

root@ubuntu:/var/log/ascend_seclog# vim ascend_install.log

[Driver] [2021-05-12 16:40:06] [INFO]runPackagePath =/home/ubuntu
[Driver] [2021-05-12 16:40:06] [INFO]Start time: 2021-05-12 16:40:06
[Driver] [2021-05-12 16:40:06] [INFO]UserCommand: A300-3010-npu-driver_20.2.0_ubuntu18.04-x86_64.run --full
[Driver] [2021-05-12 16:40:06] [INFO]base version was destroyed or not exist.
[Driver] [2021-05-12 16:40:06] [INFO]sklogd is down
[Driver] [2021-05-12 16:40:06] [ERROR]ERR_NO:0x0091;ERR_DES:HwHiAiUser not exists! Please add HwHiAiUser
[Driver] [2021-05-12 16:40:06] [INFO]End time: 2021-05-12 16:40:06
[Driver] [2021-05-12 16:40:25] [INFO]runPackagePath =/home/ubuntu
[Driver] [2021-05-12 16:40:25] [INFO]Start time: 2021-05-12 16:40:25
[Driver] [2021-05-12 16:40:25] [INFO]UserCommand: A300-3010-npu-driver_20.2.0_ubuntu18.04-x86_64.run --full
[Driver] [2021-05-12 16:40:25] [INFO]base version was destroyed or not exist.
[Driver] [2021-05-12 16:40:25] [INFO]sklogd is down
[Driver] [2021-05-12 16:40:25] [ERROR]ERR_NO:0x0091;ERR_DES:HwHiAiUser not exists! Please add HwHiAiUser
[Driver] [2021-05-12 16:40:25] [INFO]End time: 2021-05-12 16:40:25
[Driver] [2021-05-12 16:42:19] [INFO]runPackagePath =/home/ubuntu
[Driver] [2021-05-12 16:42:19] [INFO]Start time: 2021-05-12 16:42:19
[Driver] [2021-05-12 16:42:19] [INFO]UserCommand: A300-3010-npu-driver_20.2.0_ubuntu18.04-x86_64.run --uninstall
[Driver] [2021-05-12 16:42:19] [INFO]base version was destroyed or not exist.
[Driver] [2021-05-12 16:42:19] [INFO]sklogd is down
[Driver] [2021-05-12 16:42:19] [INFO]FEATURE_HOT_RESET is : FEATURE_HOT_RESET=n

或者：
tail（head） -f /var/log/ascend_seclog/ascend_install.log

./A300-3010-npu-driver_20.2.0_ubuntu18.04-x86_64.run --full

指定用户

# 指定一个用户
./A300-3010-npu-driver_20.2.0_ubuntu18.04-x86_64.run --full --install-username=ubuntu --install-usergroup=ubuntu

root@ubuntu:/home/ubuntu/ubuntu_packages# make -h

Command 'make' not found, but can be installed with:

apt install make
apt install make-guile

root@ubuntu:/home/ubuntu/ubuntu_packages#

make还是有问题
继续处理。

尝试卸载

./A300-3010-npu-driver_20.2.0_ubuntu18.04-x86_64.run --install-username=ubuntu --install-usergroup=ubuntu --full
支持工程师说可以先卸载，，，
./A300-3010-npu-driver_20.2.0_ubuntu18.04-x86_64.run --uninstall

卸载时有新的错误:
root@ubuntu:/home/ubuntu# ./A300-3010-npu-driver_20.2.0_ubuntu18.04-x86_64.run --uninstall
Verifying archive integrity... 100% SHA256 checksums are OK. All good.
Uncompressing Atlas300_software_version_20.2.0 100%
[Driver] [2021-05-31 16:50:16] [INFO]Start time: 2021-05-31 16:50:16
[Driver] [2021-05-31 16:50:16] [INFO]LogFile: /var/log/ascend_seclog/ascend_install.log
[Driver] [2021-05-31 16:50:16] [INFO]OperationLogFile: /var/log/ascend_seclog/operation.log
[Driver] [2021-05-31 16:50:16] [INFO]base version is none.
[Driver] [2021-05-31 16:50:16] [ERROR]ERR_NO:0x0090;ERR_DES:uninstall driver failed;Detail message in /var/log/ascend_seclog/ascend_install.log
[Driver] [2021-05-31 16:50:16] [INFO]End time: 2021-05-31 16:50:16
#尝试重启大法：
shutdown -r now 立刻重启

另外一台服务器尝试新建虚拟机

重新作了一台类似的虚拟机之后OK
先安装：
./A300-3010-npu-driver_20.2.0_ubuntu18.04-x86_64.run --install-username=ubuntu --install-usergroup=ubuntu --full
接着安装：
./A300-3000-3010-npu-firmware_1.76.22.3.220.run --install-username=ubuntu --install-usergroup=ubuntu --full

本台服务器新建虚拟机

#安装到90%的时候，系统会异常关机；

最后尝试在虚拟机里面关掉所有的PCI之后系统重启安装则OK，

但是尝试重新挂载PCI则虚拟机无法启动

#最后，卸载虚拟机、直接在裸机上安装ubuntu机器；

使用npu-smi info查看NPU的情况

总结：

tar xvf A300.tar
tar xvf install_gcc&make.tar
sudo dpkg -i *.deb
./A300-3010-npu-driver_20.2.0_ubuntu18.04-x86_64.run --install-username=ubuntu --install-usergroup=ubuntu --full

注意：

ubuntu安装需要在安装系统的过程中安装ssh、因为系统默认不安装、如果跳过后期安装需要直接在服务器上处理，无法远程连接和操作。

finalshell也是一个优秀的远程连接工具；

通过Esxi的ip地址可以直接进入vmware的管理界面，vmware workstation也是通过Esxi访问到服务器的虚拟化服务的；

参考：gcc离线安装 ubuntu 不用编译_在Ubuntu系统上手动安装GCC环境
参考：Ubuntu下deb包的安装方法
参考：Ubuntu 16.04 amd64下deb安装gcc5.4和所有依赖
参考：Ubuntu使用apt-get安装本地deb包
参考：linux-Ubuntu如何安装tar.gz文件
参考：ubuntu重启、关机命令
参考：NPU

参考：Atlas 800 训练服务器 npu-smi 命令参考 (型号9000) 0

华为NPU卡ubuntu（无网络连接情况）驱动安装记录相关推荐

华为mate30老是显示无法连接服务器,华为mate30pro应用提示无网络连接如何解决
华为mate30pro应用提示无网络连接如何解决,很多同学都不是很了解,下面iefans小编整理了解决应用提示无网络连接的方法讲解.希望能够解决各位用户的问题! 华为mate30pro应用提示无网络连 ...
华为手机手机应用无网络连接网络连接服务器,华为网络正常app连不上网络
大家好,我是时间财富网智能客服时间君,上述问题将由我为大家进行解答. 华为网络正常app连不上网络的原因: 1.请检查手机管家内应用联网权限是否开启,如果安装了第三方手机管家,请检查应用的联网权限是否 ...
ubuntu英伟达显卡驱动安装记录2
这里是重装英伟达显卡.即之前已经成功安装了显卡.因此这里少了一些必要配置. sudo ./NVIDIA-Linux-x86_64-495.46.run nvidia显卡驱动没有卸载干净的时候,选择ye ...
荣耀8x显示无法连接到服务器,华为荣耀8X手机升级更新217版本系统后经常显示无网络连接...
这个华为荣耀8X手机升级更新217版本的系统后经常显示无网络连接,也就是升级后移动网络断流严重,不知道是手机的问题还是系统的问题,升级217两天发现,移动网络断流严重,有4G信号,但是经常显示无网络连 ...
谷歌浏览器无网络连接打不开网页解决办法
不知道用户在使用的时候有没有遇到无法连接网络的问题,今天小编就会与大家一起来分享一下,在使用谷歌浏览器时,如果出现无网络连接的问题时,要如何解决,首先用户可以检查一下自己的网络情况,如果这些都没有问题 ...
在VMware Workstation中安装Ubuntu设置网络连接
在VMware中安装Ubuntu,网络连接方式一般有4种: 一.Bridged:Connected directly to the physical network (设置Ubuntu的网络连接方式为 ...
ipv6联网几十分钟后显示无网络连接，v4网络正常的解决方法
教育网,同时拥有v4v6地址,之前有一段时间突然出现这个问题,开机v6无法连接,这时候确定你勾选了ipv6协议,再通过打开任务管理器-服务-IP转换配置服务后,重新禁用启用网卡解决. 这时候遇到了第二 ...
托利多bcom怎么查看连接网络_遇到手机APP提示无网络连接，应该怎么办？
是不是有时候打开某些app后会提示"无网络连接"或者"当前网络不可用"提示,但是此提示很快会自己消失或者手动刷新后消失.怀疑过网络问题?手机问题?~~~ 想必有 ...
如何在无网络的情况下给Dbeaver安装数据库驱动
如何在无网络的情况下给Dbeaver安装数据库驱动如何在无网络的情况下给Dbeaver安装数据库驱动-以安装Oracle驱动为例参考:https://www.cnblogs.com/levi125 ...

华为NPU卡ubuntu（无网络连接情况）驱动安装记录