Skip to main content
Vulnerability Database/CVE-2024-21924

CVE-2024-21924: AMD Platform RAS SMM Driver RCE Vulnerability

CVE-2024-21924 is a System Management Mode callout vulnerability in AMD Platform RAS SMM driver that enables ring 0 attackers to modify boot services handlers for arbitrary code execution. This post covers technical details, impact, and mitigation.

Published:

Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.

Sources: National Vulnerability Database, CISA KEV, Exploit-DB

CVE-2024-21924 Overview

CVE-2024-21924 is a System Management Mode (SMM) callout vulnerability in the AMD AmdPlatformRasSspSmm driver. A ring 0 attacker can abuse the flaw to modify boot services handlers, potentially achieving arbitrary code execution at the firmware level. The weakness is classified as [CWE-250] Execution with Unnecessary Privileges and affects the trust boundary between SMM and boot services. Successful exploitation lets an attacker execute code within SMM, the most privileged execution context on x86 platforms, which can be used to persist below the operating system and bypass endpoint controls.

Critical Impact

A local, high-privileged (ring 0) attacker can modify boot services handlers from an SMI handler, enabling arbitrary code execution with firmware-level persistence and full compromise of confidentiality, integrity, and availability.

Affected Products

  • AMD platforms shipping the AmdPlatformRasSspSmm SMM driver (see AMD Security Bulletin AMD-SB-7028 for the authoritative affected-product list)
  • OEM UEFI firmware images incorporating the vulnerable AMD RAS SMM module
  • Systems that have not applied the firmware update referenced in AMD-SB-7028

Discovery Timeline

  • 2025-02-11 - CVE-2024-21924 published to NVD
  • 2026-06-17 - Last updated in NVD database

Technical Details for CVE-2024-21924

Vulnerability Analysis

The vulnerability resides in the AmdPlatformRasSspSmm driver, which handles Reliability, Availability, and Serviceability (RAS) and Secure Service Processor (SSP) events inside System Management Mode. SMM drivers run at the highest privilege level on x86 CPUs (ring -2) and are expected to self-contain their execution. A classic SMM callout occurs when an SMI handler invokes code or services that live outside the locked SMRAM region, such as UEFI boot services or runtime services. In this driver, the SMI handler reaches out to boot services pointers that an attacker can control or that reference memory outside SMRAM. Because the handler executes with full SMM privileges, any code reached through that callout inherits SMM privileges.

Root Cause

The root cause is Execution with Unnecessary Privileges [CWE-250]: the SMM driver invokes boot services handlers from inside an SMI handler without validating that the target function pointer resides in SMRAM. An attacker who already holds ring 0 in the operating system can modify the referenced boot services table or handler before triggering the SMI, redirecting execution to attacker-controlled code that then runs in SMM context.

Attack Vector

Exploitation requires local access and high privileges, consistent with a ring 0 attacker such as a kernel driver, bootkit, or administrator-level process with direct hardware access. The attacker overwrites or redirects a boot services handler, then triggers the vulnerable SMI by issuing a software SMI through the APMC port or an equivalent platform-specific mechanism. When the AmdPlatformRasSspSmm driver calls the targeted handler, control transfers to attacker-controlled code executing in SMM. The scope change is significant because SMM compromise crosses the firmware/OS trust boundary and can disable Secure Boot, patch the kernel, or install persistent implants that survive OS reinstallation. No public proof-of-concept or in-the-wild exploitation has been reported, and the vulnerability is not listed in the CISA Known Exploited Vulnerabilities catalog.

Stop Remote Code Execution Before It Runs

Singularity™ Endpoint detects and blocks exploit attempts at machine speed, including zero-days and novel attack techniques.

Stop Exploit Attempts

Detection Methods for CVE-2024-21924

Indicators of Compromise

  • Unexpected modifications to UEFI boot services function pointers or the EFI System Table observed during pre-boot measurement
  • Kernel drivers that map physical memory ranges associated with SMRAM or the EFI boot services region
  • Unsigned or unexpected kernel-mode drivers issuing software SMIs via writes to I/O port 0xB2
  • TPM PCR measurements for firmware (PCR 0, 2, 3) that deviate from a known-good baseline after driver load

Detection Strategies

  • Monitor loaded kernel drivers for direct physical memory access APIs such as MmMapIoSpace or /dev/mem usage targeting firmware regions
  • Validate firmware integrity using vendor-provided tools and compare SMM module hashes against the versions listed in AMD-SB-7028
  • Enable and audit Windows Hypervisor-protected Code Integrity (HVCI) and Linux IMA to block the unsigned ring 0 code typically required for exploitation
  • Correlate SMI rate anomalies from platform telemetry with suspicious process or driver activity

Monitoring Recommendations

  • Collect firmware and driver load telemetry centrally and alert on unsigned or newly introduced ring 0 modules
  • Track Secure Boot state, TPM measurements, and firmware versions as part of endpoint posture reporting
  • Review OEM firmware update compliance across the fleet to confirm patched BIOS versions are deployed

How to Mitigate CVE-2024-21924

Immediate Actions Required

  • Inventory AMD-based systems and identify firmware versions that include the vulnerable AmdPlatformRasSspSmm driver
  • Apply the OEM BIOS/UEFI update that incorporates the AMD fix referenced in AMD Security Bulletin AMD-SB-7028
  • Enforce Secure Boot, HVCI, and driver signature enforcement to raise the bar for obtaining ring 0 execution
  • Restrict local administrative privileges and block loading of unsigned or vulnerable kernel drivers

Patch Information

AMD has published remediation guidance in AMD Security Bulletin AMD-SB-7028. The fix is delivered through updated AGESA firmware that OEMs integrate into platform BIOS releases. Administrators should obtain the corresponding BIOS update from their system or motherboard vendor and validate the installed firmware version matches or exceeds the fixed release listed in the bulletin.

Workarounds

  • Reduce the local attack surface by removing unnecessary administrators and preventing untrusted driver installation until firmware updates are applied
  • Enable virtualization-based security features (VBS, HVCI, SMM protections such as SMM Supervisor where supported) to constrain ring 0 code
  • Where firmware updates cannot be applied immediately, isolate affected systems from untrusted workloads and restrict physical access

Disclaimer: This content was generated using AI. While we strive for accuracy, please verify critical information with official sources.

Sources: National Vulnerability Database, CISA KEV, Exploit-DB

Experience the Most Advanced Cybersecurity Platform

See how the world’s most intelligent, autonomous cybersecurity platform can protect your organization today and into the future.