Spectre漏洞扩大至GPU:NVIDIA、Intel补丁下载说明

简介: 无论实际安全影响有多大,我想有些用户、系统厂商不得不更新软件了。

接上篇《【转】Intel安全说明:关于对应侧信道分析的事实


无论实际安全影响有多大,我想有些用户、系统厂商不得不更新软件了。本文整理了来自网络的一些信息,供读者朋友参考。


NVIDIA在官方网站上发布安全公告确认,NVIDIA产品不会受到Meltdown漏洞影响(因为它是Intel架构专属的),但是在另一个漏洞Spectre面前,GeForce、Quadro、NVS、Tesla、GRID全部沦陷!

其中,GeForce、Quadro、NVS系列搭配Windows、Linux、FreeBSD、Solaris系统均受影响,Tesla系列搭配Windows、Linux系统受影响,GRID搭配Windows、Windows Server Hyper-V、Linux、Citrix XenServer、VMware vSphere、Red Hat KVM受影响。

目前,NVIDIA正在为旗下产品推送新的驱动程序,其中就包含漏洞修复补丁,其中GeForce、Quadro、NVS系列的Windows新驱动版本号为390.65/386.07,Linux新驱动版本为390.12/384.111,均已经提供下载,建议更新。


下面是来自 http://nvidia.custhelp.com/app/answers/detail/a_id/4611 的声明:


Security Bulletin: NVIDIA Driver Security Updates for CPU Speculative Side Channel Vulnerabilities

Answer ID 4611

Updated 01/10/2018 11:18 AM

       

NVIDIA driver response to CPU speculative side channel vulnerabilities - CVE-2017-5753, CVE-2017-5715, CVE-2017-5754

   


Bulletin Summary

NVIDIA is providing an initial security update to mitigate aspects of Google Project Zero’s January 3, 2018 publication of novel information disclosure attacks that combine CPU speculative execution with known side channels.

NVIDIA's core business is GPU computing.

We believe our GPU hardware is immune to the reported security issue. As for our driver software, we are providing updates to help mitigate the CPU security issue.

The vulnerability has three known variants:

  • Variant 1 (CVE-2017-5753): CPU mitigations are provided with the security update included in this bulletin. NVIDIA expects to work together with its ecosystem partners on future updates to further strengthen mitigations for affected CPUs.

  • Variant 2 (CVE-2017-5715): NVIDIA’s initial analysis indicates that NVIDIA software running on affected CPUs may require further updates. NVIDIA expects to work together with its ecosystem partners on this variant.

  • Variant 3 (CVE-2017-5754): At this time, NVIDIA has no reason to believe that NVIDIA software is vulnerable to this variant when running on affected CPUs.

For updates and additional information, actively monitor the NVIDIA Product Security page.

https://www.nvidia.com/product-security


Vulnerability Details

The following sections summarize the vulnerabilities. Descriptions are as published on MITRE and risk assessments follow CVSS.

CVE-2017-5753

Computer systems with microprocessors utilizing speculative execution and branch prediction may allow unauthorized disclosure of information to an attacker with local user access via a side-channel analysis.

CVSS Base Score: 8.2
CVSS Vector: CVSS:3.0/AV:A/AC:L/PR:N/UI:N/S:C/C:H/I:L/A:N

CVE-2017-5715

Computer systems with microprocessors utilizing speculative execution and indirect branch prediction may allow unauthorized disclosure of information to an attacker with local user access via a side-channel analysis.

CVSS Base Score: 8.2
CVSS Vector: CVSS:3.0/AV:A/AC:L/PR:N/UI:N/S:C/C:H/I:L/A:N

CVE-2017-5754

Computer systems with microprocessors utilizing speculative execution may allow unauthorized disclosure of information to an attacker with local user access via a side-channel analysis of the data cache.

CVSS Base Score: 7.9
CVSS Vector: CVSS:3.0/AV:L/AC:L/PR:N/UI:N/S:C/C:H/I:L/A:N

NVIDIA’s risk assessment is based on an average of risk across a diverse set of installed systems and may not represent the true risk of your local installation. NVIDIA recommends consulting a security or IT professional to evaluate the risk of your specific configuration.

Updated Products

Product OS
GeForce, Quadro, NVS Driver Software Windows, Linux, FreeBSD, Solaris
Tesla Driver Software Windows, Linux
GRID Driver Software Windows, Windows Server with Hyper-V, Linux, Citrix XenServer, VMware vSphere, and Red Hat KVM

Security Updates for CPU Speculative Side Channel Vulnerabilities

When available, download the updates from the NVIDIA Driver Downloads page.

http://www.nvidia.com/Download/index.aspx


If you are an Enterprise Service customer using NVIDIA GRID, visit the NVIDIA Licensing Center to obtain NVIDIA GRID updates.

http://nvidia.flexnetoperations.com/


If you are an Enterprise Service customer using DGX-1 or DGX Station, visit the NVIDIA Enterprise Portal for guidance.

http://nvidia.flexnetoperations.com/


Windows

Product Product Series OS Driver Branch Updated Version
GeForce All Windows R390 390.65
Quadro, NVS All Windows R390 390.65
R384 386.07

Tesla

All Windows R390 Expected week of January 22, 2018
R384 386.07
GRID All Windows, Windows Server with Hyper-V R367 Expected before the end of January 2018

Linux

Product Product Series OS Driver Branch Updated Version
GeForce All Linux, FreeBSD, Solaris R390 390.12
R384 384.111
Quadro, NVS All Linux, FreeBSD, Solaris R390 390.12
R384 384.111
Tesla All Linux R390 Expected week of January 22, 2018
R384 384.111
GRID All Linux, Citrix XenServer, VMware vSphere, and Red Hat KVM R367 Expected before the end of January 2018

Notes

  • The Linux driver contains updates to maintain compatibility with recent Linux updates for this security issue.

  • If you are using earlier driver branches of the affected products, upgrade to a supported driver branch that contains the udpate as listed in the tables for Windows and Linux.

Acknowledgements

None.

Get the Most Up to Date Product Security Information

To learn more about the vulnerability management process followed by the NVIDIA Product Security Incident Response Team (PSIRT), see the current list of NVIDIA security bulletins, or subscribe to security bulletin notifications, go to NVIDIA Product Security.

Revision History

Revision Date Description
5.0 January 10, 2018 Added clarification that GPU hardware is not affected and that NVIDIA is updating its drivers to help mitigate the CPU security issue
4.0 January 9, 2018 Added information about updated NVIDIA GRID products
3.0 January 9, 2018 Added the driver version for Tesla Compute on Windows and Linux
2.0 January 8, 2018 Added the driver version for GeForce and Quadro, NVS on Windows
1.0 January 4, 2018 Initial release

Frequently Asked Questions (FAQs)

How do I know what driver version I have installed?

  1. Launch Windows Device Manager.

  2. Select Display Adapters.

  3. Select the NVIDIA GPU node and right-click.

  4. Go to the Driver tab.

The driver version can be deciphered as shown in the following examples: 10.18.13.6472 is 364.72 and 10.18.13.472 is 304.72


现在,Intel也发布了面向Linux系统的微代码更新补丁,可以在各个Linux发行版上修复所有受影响CPU的漏洞。

支持的操作系统涵盖了近些年的几乎所有流行发行版本:Red Hat Enterprise Linux 7.4/7.3/7.2/7.1/7.0、Red Hat Enterprise Linux 6.8/6.7/6.6/6.5/6.4/6.3/6.2、SUSE Linux Enterprise Server 12/SP3/SP2/SP1、SUSE Linux Enterprise Server 11/SP4/SP3/SP2/SP1、SUSE Linux Enterprise Server (SLES) 12.2、CentOS 7.4/7.3/7.2/7.1/7.0、Debian 8.x/7.x、Fedora 24/23、Ubuntu 16.04 LTS/14.04 LTS。


Linux* Processor Microcode Data File

https://downloadcenter.intel.com/downloads/eula/27431/Linux-Processor-Microcode-Data-File?httpDown=https%3A%2F%2Fdownloadmirror.intel.com%2F27431%2Feng%2Fmicrocode-20180108.tgz

相关实践学习
部署Stable Diffusion玩转AI绘画(GPU云服务器)
本实验通过在ECS上从零开始部署Stable Diffusion来进行AI绘画创作,开启AIGC盲盒。
目录
相关文章
|
机器学习/深度学习 虚拟化 数据中心
NVIDIA T4和A10:不同应用场景下的GPU加速器选择
在数据中心和云计算领域,GPU加速器扮演着至关重要的角色。NVIDIA T4和A10是两款适用于不同应用场景的GPU加速器。本文将比较它们的性能和适用场景,帮助读者更好地选择适合自己需求的GPU实例。
4868 0
|
弹性计算 虚拟化 异构计算
阿里云GPU服务器详细介绍(Nvidia M40/P100/P4/V100)
阿里云GPU服务器详细介绍(Nvidia M40/P100/P4/V100),阿里云GPU服务器租用价格表包括包年包月价格、一个小时收费以及学生GPU服务器租用费用,阿里云GPU计算卡包括NVIDIA V100计算卡、T4计算卡、A10计算卡和A100计算卡
1755 0
|
弹性计算 虚拟化 异构计算
阿里云GPU服务器NVIDIA T4 GPU卡租用价格表
阿里云GPU服务器NVIDIA T4 GPU卡租用价格表,阿里云GPU服务器租用价格表包括包年包月价格、一个小时收费以及学生GPU服务器租用费用,阿里云GPU计算卡包括NVIDIA V100计算卡、T4计算卡、A10计算卡和A100计算卡,GPU云服务器gn6i可享受3折优惠,阿里云百科分享阿里云GPU服务器租用价格表、GPU一个小时多少钱以及学生GPU服务器收费价格表
531 0
|
5月前
|
机器学习/深度学习 人工智能 弹性计算
阿里云GPU服务器全解析_GPU服务器租用费用_NVIDIA A10、V100、T4、P4、P100 GPU卡
阿里云GPU云服务器提供NVIDIA A10、V100、T4、P4、P100等多种GPU卡,结合高性能CPU,单实例计算性能高达5PFLOPS。支持2400万PPS及160Gbps内网带宽。实例规格多样,如A10卡GN7i(3213.99元/月)、V100-16G卡GN6v(3830.00元/月)等。适用于深度学习、科学计算、图形处理等场景。GPU软件如AIACC-Training、AIACC-Inference助力性能优化。购买方式灵活,客户案例包括深势科技、流利说、小牛翻译。
656 0
|
5月前
|
XML 机器学习/深度学习 监控
性能监控之Telegraf+InfluxDB+Grafana NVIDIA GPU实时监控
【6月更文挑战12天】性能监控之Telegraf+InfluxDB+Grafana NVIDIA GPU实时监控
137 0
|
6月前
|
弹性计算 并行计算 UED
GPU实例使用--自动安装NVIDIA GPU驱动和CUDA组件
GPU 云服务器正常工作需提前安装正确的基础设施软件,对于搭载了 NVIDIA 系列 GPU卡的实例而言,如果把 NVIDIA GPU 用作通用计算,则需安装 NVIDIA GPU 驱动、 CUDA、cuDNN等软件。
101381 3
|
6月前
|
弹性计算 并行计算 UED
带你读《弹性计算技术指导及场景应用》——4. 自动安装NVIDIA GPU驱动和CUDA组件
带你读《弹性计算技术指导及场景应用》——4. 自动安装NVIDIA GPU驱动和CUDA组件
147 0
|
6月前
|
并行计算 PyTorch 算法框架/工具
详细讲述Yolov5从下载、配置及如何使用GPU运行
详细讲述Yolov5从下载、配置及如何使用GPU运行
2951 0
|
11月前
|
Ubuntu Shell Linux
专为Intel Arc GPU设计:oneAPI的详细图文安装教程
本次我们基于Intel Arc系列的GPU,在windows的WSL上安装oneAPI, 之前发表过纯CPU,通过VM虚拟机的方式,而现在我们在WSL上实现GPU版本的安装
376 0
|
机器学习/深度学习 弹性计算 数据可视化
阿里云GPU服务器详细介绍_ NVIDIA A100_A10_V100_T4 GPU卡
阿里云GPU 云服务器(GPU Cloud Computing,GPU)是提供 GPU 算力的弹性计算服务,具有超强的计算能力,服务于深度学习、科学计算、图形可视化、视频处理多种应用场景。阿里云作为亚太第一的云服务提供商,随时为您提供触手可得的算力,有效缓解计算压力,提升您的业务效率,助您提高企业竞争力
2738 0

热门文章

最新文章