KernelScan.io

HIGH Introduced in 6.15

io_uring ZcrxOffset Overflow

CVE-2026-99266

CVSS 8.0 / 10.0 KernelScan AI

CVSS:3.1/AV:L/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:H

01

In the Linux kernel, the following vulnerability has been resolved: io_uring/zcrx: fix buffer offsets in areas above 4GB io_zcrx_queue_cqe() computes the offset of the buffer in its area from net_iov_idx(), which is an unsigned int, so the shift is done in 32 bits and only the result is widened to the u64 that goes into the CQE. An area can be up to 1TB in size. For every niov beyond the first 4GB the CQE points the application at a buffer 4GB, or a multiple of that, below the one which has the data. It reads the wrong data, and then returns a buffer it doesn't own via the refill queue while the real one is never given back.

02

Engine v0.10.0

Risk summary

A local unprivileged user can exploit an integer truncation in io_uring's zero-copy receive path to read arbitrary kernel buffer data and corrupt the kernel's buffer pool management. This leads to information disclosure, potential memory corruption, and denial of service. The bug is triggered when using large ZCRX buffer areas exceeding 4GB.

Affectedio_uring/zcrx.c (io_uring zero-copy receive)

Vulnerability analysis

The io_uring zero-copy receive (ZCRX) implementation computes buffer offsets by shifting a 32-bit index value before widening it to 64 bits. For buffer areas larger than 4GB (up to 1TB supported), this shift truncates the high bits, causing the kernel to report incorrect buffer offsets to userspace. The application then reads from the wrong buffer — potentially exposing data from other connections or kernel memory — and returns the wrong buffer identifier to the kernel's refill queue. This corrupts the buffer pool tracking, leaving the real buffer unreleased and the wrong buffer marked as available, which can lead to use-after-free, double-free, or further memory corruption. Any local user with access to io_uring (typically unprivileged) can trigger this by registering a large ZCRX area and receiving enough data to cross the 4GB boundary.

03

BranchIntroducedFixed inPatch commit
6.186.156.18.609bfb315e2d09
7.26.157.2.1451f4ee29bf05
mainline6.157.4-rc1ab4536bf2e56