Look up another CVE · open CVE-2026-74708 in the interactive tracker
CVE-2026-74708
xsk: validate launch-time metadata size
Published 2026-08-22 (2 months ago) by the Linux kernel CNA. Subsystem include/net. Debian urgency: unimportant.
What to do, per Debian release
Debian's own status for CVE-2026-74708 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 | 7.1.9-1 | Update the kernel and reboot. Fixed in 7.1.9-1. A running kernel keeps the old code until the machine restarts. |
| Debian unstablesid | fixed | 7.1.9-1 | Update the kernel and reboot. Fixed in 7.1.9-1. A running kernel keeps the old code until the machine restarts. |
The vulnerability
What the bug is
Launch-time metadata extends beyond the first 16 bytes of struct xsk_tx_metadata. Reject the request when the registered metadata area does not contain the complete field. Snapshot the validated flags for the generic transmit path and use that snapshot for request and completion processing, avoiding inconsistent decisions if user space changes the flags concurrently. Note that only xsk_skb_metadata is properly using the flags, __xsk_buff_get_metadata ignores them. Next commits address that.
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.
include/net/xdp_sock_drv.h
net/xdp/xsk.c
- What this is
- XDP SOCKETS (AF_XDP)MAINTAINERS section matched on include/net/xdp_sock* · netdev@vger.kernel.org · Maintained
Check it on the machine
grep -w CONFIG_XDP_SOCKETS /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 builds this code. Every release below compiles the affected file into the kernels it ships.
- bookworm, trixie, forky, sidy
- y
- built into the kernel image on every flavour - only a patch removes it
- Build switch
- CONFIG_XDP_SOCKETS
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
- 2026-08-22 (2 months 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
- 0.17%, higher than 6.3% of all CVEssource
- CISA KEV
- not listedthe KEV catalogue
- Debian urgency
- unimportantsource
- Subsystem
- include/net
Upstream fixes
The versions the CNA records as introducing and fixing this, and the commit for each. source
| Introduced | Fixed in | Commit |
|---|---|---|
| 6.15 | 6.18.45 | af511afa1d29 |
| 6.15 | 7.1.9 | bc63d47611c0 |
| 6.15 | 7.2 | 439ce2dddf3d |
| 6.14.2 | 0 | 0 |
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
- Exploitation requires a local process to create an AF_XDP socket, register UMEM/TX rings via setsockopt, bind to a netdev queue, and submit TX descriptors; the bug is hit in kernel transmit handling, not by remote packet delivery alone.
- AC:LAttack complexity, low
- An attacker fully controls UMEM layout, metadata flags, and TX ring timing, so registering tx_metadata_len=16, setting XDP_TXMD_FLAGS_LAUNCH_TIME, and racing concurrent flag writes can reliably trigger the out-of-bounds launch_time read without uncontrollable kernel preconditions.
- PR:LPrivileges required, low
- AF_XDP socket creation checks CAP_NET_RAW in the socket network namespace (xsk_create), which unprivileged local users can obtain via user/network namespaces; no init-namespace root or admin rights are required beyond that capability.
- UI:NUser interaction, none
- No victim interaction is needed after the attacker can run code with CAP_NET_RAW; they directly configure UMEM metadata, enqueue TX descriptors, and drive transmission via sendmsg/poll without requiring another user to open files or mount filesystems.
- S:UScope, unchanged
- Impact stays within the kernel networking stack and attacker-controlled AF_XDP context on the same host; it does not cross VM, container-to-host, or IOMMU security boundaries into a different authority.
- C:HConfidentiality impact, high
- With only 16 bytes of registered TX metadata, enabling launch-time offload makes the kernel read meta->request.launch_time eight bytes beyond the declared metadata area; this is an out-of-bounds read whose contents are then consumed by kernel TX processing.
- I:HIntegrity impact, high
- The out-of-bounds launch_time value is written into skb->skb_mstamp_ns on the generic path and can be passed to NIC launch-time hardware offload on zerocopy drivers, letting an attacker influence kernel packet scheduling state and device transmit timing from attacker-chosen memory.
- A:HAvailability impact, high
- Repeated transmission with attacker-controlled launch-time values can disrupt TX scheduling and stress hardware offload paths (mlx5/igc/stmmac), enabling denial of service through stalled or mis-timed transmission rather than only a single benign misread.