Look up another CVE · open CVE-2026-74575 in the interactive tracker
CVE-2026-74575
thunderbolt: Prevent XDomain delayed work use-after-free on disconnect
Published 2026-08-15 (2 months ago) by the Linux kernel CNA. Subsystem drivers/thunderbolt. Debian urgency: not yet assigned.
What to do, per Debian release
Debian's own status for CVE-2026-74575 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 | fixedDLA-4777-1 · 2026-09-08 · text | 6.1.187-1 | Update the kernel and reboot. DLA-4777-1 shipped the fix on 2026-09-08, in 6.1.187-1. A running kernel keeps the old code until the machine restarts. |
| Debian 13trixie | fixedDSA-6466-1 · 2026-08-25 · text | 6.12.105-1 | Update the kernel and reboot. DSA-6466-1 shipped the fix on 2026-08-25, in 6.12.105-1. A running kernel keeps the old code until the machine restarts. |
| Debian 14forky | fixed | 7.1.8-1 | Update the kernel and reboot. Fixed in 7.1.8-1. A running kernel keeps the old code until the machine restarts. |
| Debian unstablesid | fixed | 7.1.8-1 | Update the kernel and reboot. Fixed in 7.1.8-1. A running kernel keeps the old code until the machine restarts. |
| Debian 12bookworm · linux-6.12 | fixedDLA-4788-1 · 2026-09-19 · text | 6.12.107-1~deb12u1 | Update the kernel and reboot. DLA-4788-1 shipped the fix on 2026-09-19, in 6.12.107-1~deb12u1. A running kernel keeps the old code until the machine restarts. |
The vulnerability
What the bug is
tb_xdp_handle_request() runs on system_wq and queues xd->state_work via queue_delayed_work() in three request handlers: PROPERTIES_CHANGED_REQUEST, UUID_REQUEST (via start_handshake), and LINK_STATE_CHANGE_REQUEST. Similarly, update_xdomain() queues xd->properties_changed_work when local properties change. Concurrently, tb_xdomain_remove() calls stop_handshake() which does cancel_delayed_work_sync() on both delayed works. Later, tb_xdomain_unregister() calls device_unregister() which eventually frees the xdomain. Since commit 559c1e1e0134 ("thunderbolt: Run tb_xdp_handle_request() in system workqueue") moved the request handler off tb->wq, the handler and the remove path are no longer serialized. If queue_delayed_work() executes after cancel_delayed_work_sync() but before the xdomain is freed, the delayed work fires on a freed object. Add xd->removing that tb_xdomain_remove() sets under xd->lock before calling stop_handshake(). Each external queue site holds the same lock and checks removing before calling queue_delayed_work(). This provides the mutual exclusion needed: either the queue site acquires the lock first and queues work that the subsequent cancel will see, or the remove path acquires the lock first and the queue site observes removing == true and skips the queue.
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/thunderbolt/xdomain.c
include/linux/thunderbolt.h
- What this is
- THUNDERBOLT DRIVERMAINTAINERS section matched on drivers/thunderbolt/ · linux-usb@vger.kernel.org · Maintained
Show whether Debian builds this, per release
Not established. This site could not find the build rule for the affected file in the kernel trees Debian ships, so it has no answer about whether Debian compiles this code. Read that as unknown, never as absent.
- bookworm, trixie, forky, sid, bookworm · linux-6.12?
- ?
- not established - no Makefile entry for the affected file
Where a release above reads "not established", no build rule for the affected file was found in that release's kernel tree. That is a gap in this lookup, not a finding that the code is absent.
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
Signals
- Published
- 2026-08-15 (2 months ago)source
- CVSS
- 8.8 high
CVSS:3.1/AV:A/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:Hsource recompute it - Attack vector
- adjacent network (AV:A)source
- EPSS
- 0.24%, higher than 13.8% of all CVEssource
- CISA KEV
- not listedthe KEV catalogue
- Debian urgency
- not yet assignedsource
- Subsystem
- drivers/thunderbolt
Upstream fixes
The versions the CNA records as introducing and fixing this, and the commit for each. source
| Introduced | Fixed in | Commit |
|---|---|---|
| 5.2 | 5.10.266 | d4fa0d544c04 |
| 5.2 | 5.15.217 | cfbd2dba3d86 |
| 5.2 | 6.1.184 | dc11d5118f9d |
| 5.2 | 6.6.151 | 91b40862a020 |
| 5.2 | 6.12.103 | 33c0ee18cf86 |
| 5.2 | 6.18.44 | 54a62153c765 |
| 5.2 | 7.1.8 | 2aa2cde2cc79 |
| 5.2 | 7.2 | 2c5d2d3c3f70 |
Debian advisories
| Advisory | Date | Shipped to | Debian |
|---|---|---|---|
| DLA-4788-1 | 2026-09-19 | bookworm 6.12.107-1~deb12u1 | tracker |
| DLA-4777-1 | 2026-09-08 | bookworm 6.1.187-1 | tracker |
| DSA-6466-1 | 2026-08-25 | trixie 6.12.105-1 | tracker |
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:AAttack vector, adjacent network
- XDomain discovery packets reach the kernel over the Thunderbolt/USB4 control channel from a connected peer (adjacent host, dock, or inline device) on the shared physical link; no remote IP/network service is involved, but the attacker need not touch the victim chassis if they control the far end of an established TB connection.
- AC:LAttack complexity, low
- The bug is a race between system_wq request handling and disconnect teardown; the peer attacker controls both sides by sending PROPERTIES_CHANGED, UUID, or LINK_STATE_CHANGE requests while forcing unplug/disconnect and can retry timing until delayed work is queued after cancel_delayed_work_sync().
- PR:NPrivileges required, none
- No local account, capability, or root on the victim is required; any malicious or compromised Thunderbolt XDomain peer that can exchange discovery control-plane packets can trigger the vulnerable queue_delayed_work() paths without OS authentication.
- UI:NUser interaction, none
- During an active XDomain session the peer can send the triggering requests and force link teardown without any further victim action at exploit time; no additional mount, login, or sysfs operation is needed beyond the already-established Thunderbolt link.
- S:UScope, unchanged
- Impact is kernel heap use-after-free and memory corruption within kernel context; it does not directly cross VM, container, or IOMMU boundaries, though successful exploitation can yield standard local privilege escalation.
- C:HConfidentiality impact, high
- Use-after-free on struct tb_xdomain lets attacker-influenced delayed work read freed kernel memory, enabling arbitrary kernel information disclosure and kernel pointer leaks that support further exploitation of the corruption primitive.
- I:HIntegrity impact, high
- Delayed work on the freed xdomain executes handshake state transitions, property/link updates, and hardware operations on attacker-reclaimed memory, enabling heap grooming and control-flow hijack for arbitrary kernel write/code execution.
- A:HAvailability impact, high
- Accessing freed tb_xdomain via state_work or properties_changed_work causes kernel oops/panic or hang during disconnect, and a peer can repeat disconnect/request storms to deny Thunderbolt/XDomain services.