Look up another CVE · open CVE-2026-74296 in the interactive tracker
CVE-2026-74296
RDMA/mlx5: Release the HW‑provided UAR index rather than the SW one
Published 2026-08-15 (2 months ago) by the Linux kernel CNA. Subsystem drivers/infiniband. Debian urgency: not yet assigned.
What to do, per Debian release
Debian's own status for CVE-2026-74296 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 | fixed | 6.1.180-1 | Update the kernel and reboot. Fixed in 6.1.180-1. A running kernel keeps the old code until the machine restarts. |
| Debian 13trixie | fixed | 6.12.100-1 | Update the kernel and reboot. Fixed in 6.12.100-1. A running kernel keeps the old code until the machine restarts. |
| Debian 14forky | fixed | 7.1.5-1 | Update the kernel and reboot. Fixed in 7.1.5-1. A running kernel keeps the old code until the machine restarts. |
| Debian unstablesid | fixed | 7.1.5-1 | Update the kernel and reboot. Fixed in 7.1.5-1. A running kernel keeps the old code until the machine restarts. |
The vulnerability
What the bug is
Free the UAR index returned by the hardware.
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/infiniband/hw/mlx5/main.c
- What this is
- MELLANOX MLX5 IB driverMAINTAINERS section matched on drivers/infiniband/hw/mlx5/ · linux-rdma@vger.kernel.org · Supported
- Kernel module
- mlx5_ib
Check it on the machine
grep -wE 'CONFIG_INFINIBAND|CONFIG_MLX5_INFINIBAND' /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.
lsmod | grep -w mlx5_ibPrints a line if that module is loaded right now, and nothing if it is not. One that is not loaded can still be loaded later by anything that needs it.
Show whether Debian builds this, per release
Debian builds this, but not on every flavour. At least one release below builds the code on some of its kernel flavours and not on others, so the answer depends on the architecture and flavour installed.
- bookwormm
- trixieMnot built on armel/rpi
- forky, sidm
- m
- a loadable module on every flavour - check lsmod, can be blacklisted
- M
- a module where it is built at all; some flavours do not build it
- Build switches
- CONFIG_INFINIBAND, CONFIG_MLX5_INFINIBAND
- 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
- 2026-08-15 (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
- 0.18%, higher than 6.6% of all CVEssource
- CISA KEV
- not listedthe KEV catalogue
- Debian urgency
- not yet assignedsource
- Subsystem
- drivers/infiniband
Upstream fixes
The versions the CNA records as introducing and fixing this, and the commit for each. source
| Introduced | Fixed in | Commit |
|---|---|---|
| 4.16 | 6.1.178 | 80f1f49f53a4 |
| 4.16 | 6.6.145 | aabfc845838e |
| 4.16 | 6.12.97 | 6f83de384ca5 |
| 4.16 | 6.18.40 | d3ff718c0c71 |
| 4.16 | 7.1.5 | ef369446f629 |
| 4.16 | 7.2 | 449ae7927152 |
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 mmap(2) on an opened mlx5 uverbs file descriptor (/dev/infiniband/uverbs*); the bug is in uar_mmap() error cleanup and is not reachable via remote RDMA/network packet handling.
- AC:LAttack complexity, low
- An attacker who can open uverbs and issue MLX5_IB_MMAP_ALLOC_WC can reliably reach the faulty cleanup by forcing rdma_user_mmap_io()/io_remap_pfn_range() to fail (e.g., memory or VMA map pressure), and controls all syscall parameters for the race-free trigger.
- PR:LPrivileges required, low
- Any unprivileged local user with access to the mlx5 RDMA character device (default 0666 uverbs nodes on shared HPC/cloud compute nodes with ConnectX NICs) can alloc a context and mmap; no real-root or CAP_SYS_ADMIN capability is required.
- UI:NUser interaction, none
- Exploitation is fully attacker-driven through uverbs open, alloc_ucontext, and mmap syscalls; no victim interaction such as mounting filesystems or opening files is needed.
- S:UScope, unchanged
- Impact is confined to kernel/driver and RDMA device resource integrity on the host; it does not cross a VM-hypervisor or IOMMU security boundary into a different authority.
- C:HConfidentiality impact, high
- Deallocating the wrong hardware UAR index frees another context's live UAR while the attacker-allocated UAR leaks, creating a cross-context UAR use-after-free that can expose other users' MMIO doorbell/register mappings and device state.
- I:HIntegrity impact, high
- UAR pages provide userspace MMIO write access to mlx5 doorbells and control registers; corrupting or reassigning UAR ownership enables arbitrary device register/DMA manipulation and is exploitable for privilege escalation beyond simple data tampering.
- A:HAvailability impact, high
- Freeing an in-use hardware UAR from another context can cause kernel oops/panic, NIC hangs, or repeated crashes when victims continue doorbell/MMIO access through stale mappings.