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

Look up another CVE · open CVE-2023-54280 in the interactive tracker

CVE-2023-54280

cifs: fix potential race when tree connecting ipc

9.8 EPSS 0.5% AV:N

Published 2025-12-30 (10 months ago) by the Linux kernel CNA. Subsystem fs/cifs. Debian urgency: not yet assigned.

What to do, per Debian release

Debian's own status for CVE-2023-54280 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 12bookwormvulnerable6.1.190-1No fix published anywhere yet, neither in Debian nor upstream for the series bookworm tracks. Watch the Debian tracker page for this CVE. Advisories for bookworm have historically landed a median of 23 days after publication.
Debian 13trixiefixed6.3.7-1Update the kernel and reboot. Fixed in 6.3.7-1. A running kernel keeps the old code until the machine restarts.
Debian 14forkyfixed6.3.7-1Update the kernel and reboot. Fixed in 6.3.7-1. A running kernel keeps the old code until the machine restarts.
Debian unstablesidfixed6.3.7-1Update the kernel and reboot. Fixed in 6.3.7-1. A running kernel keeps the old code until the machine restarts.

The vulnerability

What the bug is

Protect access of TCP_Server_Info::hostname when building the ipc tree name as it might get freed in cifsd thread and thus causing an use-after-free bug in __tree_connect_dfs_target(). Also, while at it, update status of IPC tcon on success and then avoid any extra tree connects.

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.

fs/cifs/dfs.c

What this is
FILESYSTEMS (VFS and infrastructure)MAINTAINERS section matched on fs/* · linux-fsdevel@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?
?
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
2025-12-30 (10 months ago)source
CVSS
9.8 critical
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:Hsource recompute it
Attack vector
network (AV:N)source
EPSS
0.52%, higher than 42.6% of all CVEssource
CISA KEV
not listedthe KEV catalogue
Debian urgency
not yet assignedsource
Subsystem
fs/cifs

Upstream fixes

The versions the CNA records as introducing and fixing this, and the commit for each. source

IntroducedFixed inCommit
5.166.2.15536ec71ba060
5.166.3.2553476df55a1
5.166.4ee20d7c61007
5.15.8100

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:NAttack vector, network
The race is in the CIFS/SMB client's DFS reconnect path and is driven entirely by remote peer behaviour — a malicious or compromised SMB/DFS server (or an on-path attacker resetting the TCP session) controls when reconnects occur and which failover targets are returned, and the leaked heap bytes are sent back to that server in the TREE_CONNECT request. Both racing threads (cifsd demux thread and the cifsiod reconnect worker) run automatically over the network connection with no local component.
AC:LAttack complexity, low
The attacker controls both sides of the race: dropping the TCP connection forces cifsd into reconnect_dfs_server()/__reconnect_target_unlocked() which kfree()s server->hostname, while the referral target list it supplies guarantees the free happens, and the cifsiod reconnect worker concurrently reads the pointer unlocked. reconnect_dfs_server() retries in a msleep(3000) loop so the window can be hit repeatedly until it lands.
PR:NPrivileges required, none
The attacker acts as the SMB/DFS server or as a network attacker on the connection and never authenticates to or holds any account on the victim host. The vulnerable reconnect path runs in kernel threads (cifsd and the cifsiod reconnect worker), so no privileged or unprivileged local process is needed at all.
UI:NUser interaction, none
Against an already-mounted DFS share — the normal persistent state for enterprise fstab/autofs mounts — the entire trigger sequence is kernel-internal: cifsd drops and re-establishes the connection and queues smb2_reconnect_server(), which reaches __tree_connect_dfs_target() with no user process and no victim action.
S:UScope, unchanged
The use-after-free is confined to kernel heap memory of the affected host and does not cross a virtualization, IOMMU, or sandbox boundary. Impact and vulnerable component share the same security authority.
C:HConfidentiality impact, high
The freed hostname buffer is formatted with "%s", and string_nocheck() runs with precision -1 (`while (lim--)`), so it reads freed and reallocated heap memory unbounded until a NUL byte, copying up to ~500 bytes into the tree name, which is then transmitted to the remote attacker-controlled server in the SMB2 TREE_CONNECT request — a direct kernel heap disclosure primitive.
I:HIntegrity impact, high
This is a use-after-free on an attacker-influenced heap object; an attacker who grooms the reallocated kmalloc chunk controls the string used to build the IPC$ tree-connect request, steering the client's connection state, and UAF conditions on the CIFS server/session objects are generally leverageable for stronger corruption primitives.
A:HAvailability impact, high
The unbounded scan for a NUL over the freed allocation readily walks off the slab object into unmapped memory, producing a kernel oops/panic (and an immediate KASAN BUG on hardened kernels). The remote attacker can retrigger the reconnect race repeatedly until the client crashes.