Look up another CVE · open CVE-2026-97911 in the interactive tracker
CVE-2026-97911
accel: ethosu: Ensure SRAM region size matches job
Published 2026-09-25 (24 days ago) by the Linux kernel CNA. Subsystem drivers/accel. Debian urgency: unimportant.
What to do, per Debian release
Debian's own status for CVE-2026-97911 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 | not affected | 6.12.113-1 | Nothing to do. trixie never shipped the vulnerable code. |
| Debian 14forky | fixed | 7.2.7-1 | Update the kernel and reboot. Fixed in 7.2.7-1. A running kernel keeps the old code until the machine restarts. |
| Debian unstablesid | fixed | 7.2.7-1 | Update the kernel and reboot. Fixed in 7.2.7-1. A running kernel keeps the old code until the machine restarts. |
The vulnerability
What the bug is
It is possible for userspace to set the job SRAM size to 0, but then still have SRAM accesses in the command stream. When the job SRAM size is 0, setting the region base register is skipped and a stale base address from a prior job is used. Check the region size against the job's SRAM size instead of just the size of the SRAM. The job's SRAM size was already checked against the total SRAM size.
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/accel/ethosu/ethosu_job.c
- What this is
- ARM ETHOS-U NPU DRIVERMAINTAINERS section matched on drivers/accel/ethosu/ · dri-devel@lists.freedesktop.org · Supported
Check it on the machine
grep -wE 'CONFIG_DRM_ACCEL|CONFIG_DRM_ACCEL_ARM_ETHOSU' /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.
Show whether Debian builds this, per release
Debian does not build this code in 2 releases. The build rule could not be established for 2 releases, so nothing is claimed about those.
- bookworm, trixie?
- forky, sidn
- ?
- not established - no Makefile entry for the affected file
- n
- not enabled on any flavour - this release does not build the code
Where a release above reads "not established", no build rule for the affected file was found in that release's kernel tree. That is a gap in this lookup, not a finding that the code is absent.
- Build switches
- CONFIG_DRM_ACCEL, CONFIG_DRM_ACCEL_ARM_ETHOSU
- Reads as
- all on means the file is compiled; any one off means it is not.
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
- 2026-09-25 (24 days ago)source
- CVSS
- 7.8 high
CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:Hsource recompute it - Attack vector
- local (AV:L)source
- EPSS
- 0.15%, higher than 4.1% of all CVEssource
- CISA KEV
- not listedthe KEV catalogue
- Debian urgency
- unimportantsource
- Subsystem
- drivers/accel
Upstream fixes
The versions the CNA records as introducing and fixing this, and the commit for each. source
| Introduced | Fixed in | Commit |
|---|---|---|
| 7.2 | 7.2.7 | 50c27d412fed |
| 7.2 | 7.3 | 2b39d680c9e0 |
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 input is a DRM_IOCTL_ETHOSU_SUBMIT job (sram_size=0, no region-2 BO handle) plus a command stream that uses region 2. A local process sends it through the /dev/accel/accelN ioctl into ethosu_ioctl_submit_job(). No remote protocol carries this data.
- AC:LAttack complexity, low
- The attacker sets the stale BASEP(2) value themselves: a prior job of their own binds a BO to region 2 while sram_size is 0, then they free that BO. Their next job skips the base write in ethosu_job_hw_submit(). Both steps are ordinary submissions they control, and there is no race.
- PR:LPrivileges required, low
- All ethosu ioctls, including SUBMIT, are registered with flags 0, so any unprivileged user allowed to open the accel device node (for example, the render group) can reach the flawed region-size check. Root is not required.
- UI:NUser interaction, none
- The attacker's own ioctl calls on their own BOs and command streams trigger the bug. No other user has to do anything.
- S:UScope, unchanged
- The NPU is driven by the same kernel, and no IOMMU or VM boundary is crossed (the driver sets only a 40-bit DMA mask). The impact stays within the kernel's own security authority.
- C:HConfidentiality impact, high
- The NPU reads through a stale BASEP(2): a freed BO's physical address, or SRAM holding another client's data. The command stream can copy up to sram_size bytes from there into an attacker BO, disclosing kernel pages or other users' data.
- I:HIntegrity impact, high
- NPU output operations can write through the stale region-2 base into pages that were freed and reused by the kernel, or into another client's buffer. This is a DMA use-after-free write that the attacker sizes, and it can corrupt kernel memory.
- A:HAvailability impact, high
- Uncontrolled DMA writes into reused kernel pages, or into a stale or reset base address, can corrupt kernel data structures, crash the system, or hang the NPU. The attacker can repeat this at will.