CVE-2024-53072

What it actually does

The amd_pmc driver fails to validate the S2D_PHYS_ADDR values returned by SMU commands before calling ioremap, permitting a zero address that triggers a kernel warning when enable_stb=1 is used on systems without STB support.

Why it matters to the business

System administrators on affected AMD platforms may see kernel ring-buffer warnings and module-load messages during routine diagnostics, increasing noise in logs without leading to crashes, data loss, or service outages; low EPSS and absence from KEV indicate negligible operational urgency despite high ecosystem prevalence of the Linux kernel.

How an attack unfolds

An attacker with local root privileges loads the amd_pmc module using the enable_stb=1 parameter on an AMD system where the STB is unavailable; the driver receives a zero address from the SMU, performs an invalid ioremap, and generates warnings, but the flaw provides no code execution or privilege escalation path and is not known to be leveraged by the listed threat actors.

How to fix it

Update the Linux kernel to a version containing commit a50863dd1f92d43c975ab2ecc3476617fe98a66e; for t-coffee on Ubuntu 16.04 LTS upgrade to 11.00.8cbe486-2.

How to confirm the fix worked

Confirm the patched kernel is running, reload amd_pmc with enable_stb=1 on hardware lacking STB, and verify that neither the SMU error nor the ioremap warning appears in the kernel ring buffer.

Typical attack path

1. Obtain local access on an AMD-based Linux system
2. Load amd_pmc with the enable_stb=1 module parameter
3. Observe SMU command failure returning address 0
4. Trigger the ioremap warning in the kernel log

How to detect it

Search kernel logs and dmesg output for the exact messages emitted when the invalid address is passed to ioremap; correlate with module-load events containing enable_stb=1.

• amd_pmc AMDI0009:00: SMU cmd failed. err: 0xff
• ioremap on RAM at 0x0000000000000000 - 0x0000000000ffffff
• WARNING: CPU:.* at arch/x86/mm/ioremap.c:.* __ioremap_caller

What your security team should do

• Apply the referenced kernel commit or distribution update at the next maintenance window
• Review any custom scripts that pass enable_stb=1 to amd_pmc
• Monitor kernel logs for the documented warning strings as a temporary detection measure