---
id: "CVE-2026-99078"
kind: "vulnerability"
aliases: ["CVE-2026-99078"]
title: "Linux kernel: mlx5e AF_XDP retry can double-free receive buffers"
severity: "unknown"
cvss: null
epss: 0.00225
exploited: false
status: "open"
category: "infrastructure"
products: []
ecosystems: []
tags: ["af-xdp", "infrastructure", "linux-kernel", "mlx5", "nvidia-connectx"]
not_affected_if: ["no-af-xdp-zero-copy", "no-mlx5-nic"]
published_at: "2026-10-23T13:17:31+00:00"
first_seen_at: "2026-10-23T21:57:34+00:00"
updated_at: "2026-10-25T12:38:50+00:00"
sources: [{"source": "nvd", "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-99078", "first_seen_at": "2026-10-23T21:57:34+00:00"}, {"source": "epss", "url": "https://api.first.org/data/v1/epss?cve=CVE-2026-99078", "first_seen_at": "2026-10-25T12:38:50+00:00"}]
---

# Linux kernel: mlx5e AF_XDP retry can double-free receive buffers

In the Linux kernel's mlx5e driver for Mellanox ConnectX NICs, an AF_XDP zero-copy receive path that failed a buffer allocation could retry and release buffers already returned to the XSK pool. The sources describe a use-after-free-style stale release, but do not state an attacker-facing impact.

_Summary, check and fix written by a model from the sources; labelled as such._

## How to check

- `uname -r` and compare against the kernel version that contains the upstream fix
- Check whether the mlx5_core driver is loaded with `lsmod | grep mlx5` and whether AF_XDP zero-copy sockets are in use
- unknown

## Fix

- Update the Linux kernel to a release that includes the upstream commit for net/mlx5e XSK MPWQE refill retry (fixed version unknown from sources)

## Timeline

- 2026-10-23T21:57:34+00:00 nvd https://nvd.nist.gov/vuln/detail/CVE-2026-99078
- 2026-10-25T12:38:50+00:00 epss https://api.first.org/data/v1/epss?cve=CVE-2026-99078

## From the sources

In the Linux kernel, the following vulnerability has been resolved:

net/mlx5e: Prevent stale XSK buffer release on MPWQE refill retry

With AF_XDP on a striding RQ, mlx5e defers releasing XSK buffers until
an MPWQE is refilled. If XSK allocation then returns -ENOMEM,
actual_wq_head is not advanced and a later NAPI poll retries the same
WQE.

mlx5e_free_rx_mpwqe() leaves each released slot marked as releasable. On
retry it can therefore call xsk_buff_free() again through stale pointers
after the frames have returned to the XSK pool and been reallocated.

Set all skip_release_bitmap bits in the common error path of
mlx5e_xsk_alloc_rx_mpwqe(). This matches mlx5e_alloc_rx_mpwqe(). A
successful allocation already clears the bitmap after replacing every
buffer, so retries become idempotent without changing the success path.

Fault injection forced three consecutive failures for one selected MPWQE.
Both an early allocation failure and a partial 8-of-16-buffer unwind
released the original 16 XSK buffers only once. Each error left a full
bitmap, the following NAPI retry skipped the release, and a later
successful allocation cleared it. A 20-second AF_XDP zero-copy pressure
run exercised 1,575,262 buffer allocation failures without invalid
descriptors, WQE errors, or kernel warnings.

---

_Free to read at a human pace. An API with keys and an MCP server are in preparation; early access: hello@infrastack.ai_