Look up another CVE · open CVE-2026-97428 in the interactive tracker
CVE-2026-97428
drm/amdgpu: harden FRU PIA parsing with bounded helpers
Published 2026-09-24 (24 days ago) by the Linux kernel CNA. Subsystem drivers/gpu. Debian urgency: unimportant.
What to do, per Debian release
Debian's own status for CVE-2026-97428 in every release this site tracks, and what each one means in practice. Every status below is the one the Debian Security Team publishes for this CVE. source
| Release | Status | Version | What this means |
|---|---|---|---|
| Debian 12bookworm | not affected | 6.1.190-1 | Nothing to do. bookworm never shipped the vulnerable code. |
| Debian 13trixie | fixedDSA-6528-1 · 2026-09-29 · text | 6.12.111-1 | Update the kernel and reboot. DSA-6528-1 shipped the fix on 2026-09-29, in 6.12.111-1. A running kernel keeps the old code until the machine restarts. |
| Debian 14forky | fixed | 7.2.6-1 | Update the kernel and reboot. Fixed in 7.2.6-1. A running kernel keeps the old code until the machine restarts. |
| Debian unstablesid | fixed | 7.2.6-1 | Update the kernel and reboot. Fixed in 7.2.6-1. A running kernel keeps the old code until the machine restarts. |
| Debian 12bookworm · linux-6.12 | fixedDLA-4817-1 · 2026-10-04 · text | 6.12.111-1~deb12u1 | Update the kernel and reboot. DLA-4817-1 shipped the fix on 2026-10-04, in 6.12.111-1~deb12u1. A running kernel keeps the old code until the machine restarts. |
The vulnerability
What the bug is
Replace the open-coded TLV walk with fru_pia_advance() and fru_pia_copy_field() helpers that bound every read by the actual EEPROM data length, preventing out-of-bounds reads on truncated or malformed FRU data.
Where it lives
The fix touches these files. If the subsystem is one you do not use, the practical exposure is lower, though the package is still the vulnerable one.
drivers/gpu/drm/amd/amdgpu/amdgpu_fru_eeprom.c
- What this is
- RADEON and AMDGPU DRM DRIVERSMAINTAINERS section matched on drivers/gpu/drm/amd/ · amd-gfx@lists.freedesktop.org · Supported
- Kernel module
- amdgpu
Check it on the machine
grep -w CONFIG_DRM_AMDGPU /boot/config-$(uname -r)Reads the booted kernel's own configuration, on the machine and nowhere else. =y is built in, =m is a module, and an is not set line means the code is not there.
lsmod | grep -w amdgpuPrints a line if that module is loaded right now, and nothing if it is not. One that is not loaded can still be loaded later by anything that needs it.
Show whether Debian builds this, per release
Debian builds this, but not on every flavour. At least one release below builds the code on some of its kernel flavours and not on others, so the answer depends on the architecture and flavour installed.
- bookwormMnot built on amd64/cloud-amd64, arm64/cloud-arm64, armhf/armmp, armhf/armmp-lpae, armhf/rt/armmp
- trixieMnot built on amd64/cloud-amd64, arm64/cloud-arm64, armel/rpi, armhf/armmp, armhf/armmp-lpae, armhf/rt/armmp, s390x/s390x
- forky, sidMnot built on amd64/cloud-amd64, arm64/cloud-arm64, s390x/s390x
- bookworm · linux-6.12Mnot built on amd64/cloud-amd64, arm64/cloud-arm64, armhf/armmp, armhf/armmp-lpae, armhf/rt/armmp
- M
- a module where it is built at all; some flavours do not build it
- Build switch
- CONFIG_DRM_AMDGPU
From the configuration Debian publishes for every kernel flavour it builds. It covers only the architectures Debian ships a linux-config package for, and nothing about a kernel you built yourself. Debian's own status above stays authoritative about the source package.MAINTAINERS kbuild Makefiles Debian linux-config packages
Signals
- Published
- 2026-09-24 (24 days ago)source
- CVSS
- 7.7 high
CVSS:3.1/AV:L/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:Hsource recompute it - Attack vector
- local (AV:L)source
- EPSS
- 0.14%, higher than 3.0% of all CVEssource
- CISA KEV
- not listedthe KEV catalogue
- Debian urgency
- unimportantsource
- Subsystem
- drivers/gpu
Upstream fixes
The versions the CNA records as introducing and fixing this, and the commit for each. source
| Introduced | Fixed in | Commit |
|---|---|---|
| 6.2 | 6.12.111 | 813311512505 |
| 6.2 | 6.18.53 | 4396d74b0c73 |
| 6.2 | 7.2 | c990c05eb6c7 |
Debian advisories
| Advisory | Date | Shipped to | Debian |
|---|---|---|---|
| DLA-4817-1 | 2026-10-04 | bookworm 6.12.111-1~deb12u1 | tracker |
| DSA-6528-1 | 2026-09-29 | trixie 6.12.111-1 | tracker |
Why the CNA scored it this way
Quoted from the kernel CNA's own scoring file, word for word. A highlighted metric carries the most severe value CVSS v3.1 defines for it, which is what pushes the score up. source
- AV:LAttack vector, local
- The malformed IPMI FRU Product Info Area TLV lengths (pia[addr] & 0x3F) are consumed by amdgpu_fru_get_product_info() after amdgpu_eeprom_read() on adev->pm.fru_eeprom_i2c_bus, reached from amdgpu_pci_probe() → amdgpu_driver_load_kms() → amdgpu_device_init() → amdgpu_device_ip_init(). Those bytes come from the GPU I2C FRU EEPROM, not a network protocol or USB descriptor.
- AC:LAttack complexity, low
- A FRU image with valid Common Header and PIA checksums but TLV lengths that extend past the kmalloc'd pia buffer makes the pre-fix memcpy of manufacturer_name, product_name, product_number, serial, and fru_id overread on every probe. The attacker programs those EEPROM bytes; skipped Product Version/Asset Tag advances also lack a bounds check. No race.
- PR:NPrivileges required, none
- amdgpu_fru_get_product_info() runs from amdgpu_device_ip_init() during amdgpu_pci_probe() with no capable() or credential check. is_fru_eeprom_supported() only matches server SKUs (Vega20 D161/D163, Aldebaran, SMU 13.0.6/13.0.14) and skips SR-IOV VFs and APUs; a crafted FRU on a supported GPU is parsed without a host account.
- UI:NUser interaction, none
- amdgpu_pci_probe() calls amdgpu_fru_get_product_info() automatically while binding a supported AMD server GPU; no victim must open a DRM node, mount media, or write sysfs.
- S:UScope, unchanged
- The memcpy overread of the kzalloc'd pia buffer and the later 0444 sysfs emit stay in this host kernel's amdgpu driver. is_fru_eeprom_supported() returns false for amdgpu_sriov_vf(), so this is not a guest-to-host or IOMMU boundary crossing.
- C:HConfidentiality impact, high
- Pre-fix memcpy(fru_info->product_name, pia+addr+1, min(64, pia[addr]&0x3F)) and the manufacturer_name/product_number/serial/fru_id copies do not cap the length to remaining PIA bytes, so a 6-bit length near the end of pia reads up to 63 bytes of adjacent kernel heap. amdgpu_fru_product_name_show and sibling 0444 attributes then emit those fields; that is not a few-byte leak.
- I:NIntegrity impact, none
- Each memcpy destination is min'd against sizeof(fru_info->manufacturer_name/product_name/product_number/serial/fru_id), so the TLV walk does not write past those arrays or other kernel objects. This path has no out-of-bounds write or control-flow hijack primitive.
- A:HAvailability impact, high
- A short PIA (size = buf[1]*8, e.g. 8 bytes) plus a product_name field length of 63 makes memcpy() walk tens of bytes past the kzalloc'd pia object into unmapped or poisoned slab and oops inside amdgpu_fru_get_product_info() during probe.