An industry-wide issue was found in the way many modern microprocessor designs have implemented speculative execution of instructions (a commonly used performance optimization). There are three primary variants of the issue which differ in the way the speculative execution can be exploited.

Variant CVE-2017-5753 triggers the speculative execution by performing a bounds-check bypass. It relies on the presence of a precisely-defined instruction sequence in the privileged code as well as the fact that memory accesses may cause allocation into the microprocessor's data cache even for speculatively executed instructions that never actually commit (retire). As a result, an unprivileged attacker could use this flaw to cross the syscall boundary and read privileged memory by conducting targeted cache side-channel attacks.

In certain circumstances, an issue in Arm Cortex-A57, Cortex-A72 (revisions before r1p0), Cortex-A73 and Cortex-A75 may allow an adversary to gain a weak form of control over the victim's branch history.

DISPUTED The AES instructions on the ARMv8 platform do not have an algorithm that is "intrinsically resistant" to side-channel attacks. NOTE: the vendor reportedly offers the position "while power side channel attacks ... are possible, they are not directly caused by or related to the Arm architecture."

A new cache speculation vulnerability known as Branch History Injection (BHI) or Spectre-BHB was found. Spectre-BHB is similar to Spectre v2, except that malicious code uses the shared branch history (stored in the CPU Branch History Buffer, or BHB) to influence mispredicted branches within the victim's own hardware context. Once that occurs, speculation caused by mispredicted branches can be used to cause cache allocation, which can then be used to infer information that should not be accessible.

Spectre BHB is a variant of Spectre-v2 in which malicious code uses the shared branch history (stored in the CPU BHB) to influence mispredicted branches in the victim's hardware context. Speculation caused by these mispredicted branches can then potentially be used to cause cache allocation, which can then be used to infer information that should be protected.