data built 2026-10-19T13:16:53Z

Look up another CVE · open CVE-2026-53133 in the interactive tracker

CVE-2026-53133

RDMA/umem: Fix truncation for block sizes >= 4G

7.8 EPSS 0.1% AV:L

Published 2026-06-25 (4 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-2026-53133 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

ReleaseStatus VersionWhat this means
Debian 12bookwormfixedDLA-4665-1 · 2026-07-03 · text6.1.176-1Update the kernel and reboot. DLA-4665-1 shipped the fix on 2026-07-03, in 6.1.176-1. A running kernel keeps the old code until the machine restarts.
Debian 13trixiefixed6.12.94-1Update the kernel and reboot. Fixed in 6.12.94-1. A running kernel keeps the old code until the machine restarts.
Debian 14forkyfixed7.0.13-1Update the kernel and reboot. Fixed in 7.0.13-1. A running kernel keeps the old code until the machine restarts.
Debian unstablesidfixed7.0.13-1Update the kernel and reboot. Fixed in 7.0.13-1. A running kernel keeps the old code until the machine restarts.

The vulnerability

What the bug is

When the iommu is used the linearization of the mapping can give a single block that is very large split across multiple SG entries. When __rdma_block_iter_next() reassembles the split SG entries it is overflowing the 32 bit stack values and computed the wrong DMA addresses for blocks after the truncation. Use the right types to hold DMA addresses.

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/core/iter.c

What this is
INFINIBAND SUBSYSTEMMAINTAINERS section matched on drivers/infiniband/ · linux-rdma@vger.kernel.org · Supported
Kernel module
ib_core

Check it on the machine

grep -w CONFIG_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 ib_core

Prints 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 as a loadable module. Every release below compiles it as a module on every flavour, so whether the code is in the running kernel depends on whether that module is loaded.

  • bookworm, trixie, forky, sidm
m
a loadable module on every flavour - check lsmod, can be blacklisted
Build switch
CONFIG_INFINIBAND

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
2026-06-25 (4 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
0.13%, higher than 2.3% 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

IntroducedFixed inCommit
5.25.10.2592ff4b7817e5b
5.25.15.210dee2a49adeeb
5.26.1.176cc644d5608e3
5.26.6.1438fe0231adebe
5.26.12.94baf8685bcf56
5.26.18.36afd35fec9297
5.27.0.13ac1aad8e1281
5.27.115fe76e23615

Debian advisories

AdvisoryDateShipped toDebian
DLA-4671-12026-07-05bullseye 6.1.176-1~deb11u1tracker
DLA-4665-12026-07-03bookworm 6.1.176-1tracker
DLA-4664-12026-07-03bullseye 5.10.259-1tracker

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
Vulnerable code is reached through local RDMA uverbs operations such as opening `/dev/infiniband/uverbsX` and issuing memory-registration commands. Remote RDMA traffic may later exercise a bad MR, but the vulnerable iterator is invoked by local userspace registration, not by network packet parsing.
AC:LAttack complexity, low
There is no race or probabilistic condition; on an affected RDMA/IOMMU deployment the attacker controls the MR address, length, and access flags needed to request large aligned mappings. Hardware/IOMMU support is an environmental prerequisite for the affected subsystem, not an exploitation step beyond attacker control.
PR:LPrivileges required, low
The uverbs device node is created world-accessible and MR registration does not require CAP_SYS_ADMIN or CAP_NET_ADMIN, though the process must be an unprivileged local user able to use the RDMA device and pin memory under its memlock limits. No real init-namespace root privilege is required.
UI:NUser interaction, none
After local access to the RDMA device, the attacker can perform the open/context/PD/MR registration sequence directly. No separate victim action is required.
S:UScope, unchanged
The impact remains within the kernel/RDMA device memory-management security authority. This is not a guest-to-host escape, IOMMU boundary bypass by a guest, or another cross-scope boundary change.
C:HConfidentiality impact, high
The truncation can program wrong DMA addresses into RDMA memory translation tables for blocks after the overflow. With RDMA read-capable access, the resulting mistranslation can expose unintended memory reachable through the device mapping, so the higher defensible confidentiality impact is High.
I:HIntegrity impact, high
The same wrong DMA translation can be used with local or remote write-capable MR permissions to direct RDMA writes to unintended mapped memory. Because this is device DMA memory corruption rather than a bounded metadata error, the higher defensible integrity impact is High.
A:HAvailability impact, high
Incorrect DMA translations can cause IOMMU/device faults, corrupted RDMA state, or repeated failures of registered memory operations. Given the kernel/device-level DMA fault surface and the requirement to choose the higher severity when uncertain, availability impact is High.