Modern processors dynamically control their operating frequency to optimize resource utilization, maximize energy savings, and conform to system-defined constraints. If, during the execution of a software workload, the running average of any electrical or thermal parameter exceeds its corresponding predefined threshold value, the power management architecture will reactively adjust CPU frequency to ensure safe operating conditions. In this paper, we demonstrate how such power management-based frequency throttling activity forms a source of timing side-channel information leakage, which can be exploited by an attacker to infer secret data even from a constant-cycle victim workload. The proposed frequency throttling side-channel attack can be launched by both kernel-space and user-space attackers, thus compromising security guarantees provided by isolation boundaries. We validate our attack methodology across different systems and threat models by performing experiments on a constant-cycle implementation of AES algorithm based on AES-NI instructions. The results of our experimental evaluations demonstrate that the attacker can successfully recover all bytes of an AES key by measuring encryption execution times. Finally, we discuss different options to mitigate the threat posed by frequency throttling side-channel attacks, as well as their advantages and disadvantages.

翻译：现代处理器通过动态控制其运行频率来优化资源利用率、最大化节能效果，并满足系统定义的约束条件。当软件工作负载执行期间，任何电气或热参数的运行平均值超过预设阈值时，电源管理架构将主动调整CPU频率以确保安全运行状态。本文展示了基于电源管理的频率节流活动如何构成时序侧信道信息泄露源，攻击者可利用该漏洞从恒定周期的受害者工作负载中推断出秘密数据。所提出的频率节流侧信道攻击可由内核态和用户态攻击者发起，从而破坏隔离边界提供的安全保障。我们通过基于AES-NI指令的恒定周期AES算法实现，在跨系统和威胁模型下验证了攻击方法。实验评估结果表明，攻击者可通过测量加密执行时间成功恢复AES密钥的所有字节。最后，我们讨论了缓解频率节流侧信道攻击威胁的不同方案及其优缺点。