CVE-2026-99709: inet: frags: limit the truesize of a reassembly queue
In the Linux kernel, the following vulnerability has been resolved:
inet: frags: limit the truesize of a reassembly queue
The truesize of fragments in a reassembly queue is only bounded by the per netns ipfraghighthresh sysctl, which has no upper limit.
inetfragreasmfinish() sums the truesize of all fragments in an int. If this sum reaches 2^32, it wraps, and the final skwmemalloc adjustment of the socket owning the reassembled skb can cancel its own reference, leading to a use-after-free of this socket.
This can be triggered from an unprivileged user/network namespace, raising ipfraghighthresh and using padded TUN packets, which keep a large truesize after being trimmed by iprcvcore().
Track the sum of the truesize of fragments in each queue, and refuse a fragment that would make it bigger than INTMAX / 2. Callers kill the queue on this error. This bound is far above the needs of legitimate traffic.
Affected Software
Remediation
Recommended actions to resolve this vulnerability, in priority order.
- Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 5.5 - Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 5.11 - Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 5.16 - Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 6.2 - Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 6.7 - Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 6.9 - Upgrade
Upgrade
linux/Kernelto a version that resolves this vulnerability.Fixed in 7.3.5 - Compensating control
Track the total truesize of fragments in each reassembly queue and refuse any fragment that would make the total exceed INT_MAX / 2.
Frequently Asked Questions
Who can trigger this issue?
An unprivileged user in a user or network namespace can trigger it by raising the per-network-namespace ipfrag_high_thresh setting and sending padded TUN packets.
What conditions are needed for exploitation?
The attacker needs to create a fragment reassembly queue whose cumulative truesize reaches 2^32, causing an integer wrap during reassembly. Padded TUN packets are used because their truesize can remain large after trimming by ip_rcv_core().
What happens when the vulnerable condition is reached?
The wrapped truesize sum can cause the final sk_wmem_alloc adjustment to cancel the reference to the socket that owns the reassembled skb. This can result in a use-after-free of that socket.
How does the fix mitigate the problem?
The fix tracks cumulative fragment truesize for each reassembly queue and rejects a fragment that would raise it above INT_MAX / 2. When this occurs, callers destroy the affected reassembly queue.