In the Linux kernel, the following vulnerability has been resolved:
ceph: bound num_export_targets array for mds info v2/v3
ceph_mdsmap_decode() in fs/ceph/mdsmap.c reads num_export_targets from
each per-mds info record and advances the decode cursor by
num_export_targets * sizeof(u32) without first checking that many bytes
remain. The only upper-bound check that catches a runaway cursor
(*p > info_end) is gated on info_v >= 4, because info_end is left NULL
for info_v 2 and 3. When the monitor sends an MDS map whose per-mds
info version is 2 or 3 with an oversized num_export_targets, the cursor
moves past the message front buffer and the later export-targets loop
calls the unchecked ceph_decode_32() on out-of-bounds memory.
A kernel client processes CEPH_MSG_MDS_MAP from its monitor session
(net/ceph/mon_client.c dispatches it; fs/ceph/super.c routes it to
ceph_mdsc_handle_mdsmap(), which sets end to the front buffer bound and
calls ceph_mdsmap_decode()). A malicious or compromised monitor, or an
on-path attacker on an unsigned/unencrypted messenger session, can
therefore drive an out-of-bounds read in the client kernel; on x86_64
with KASAN it is reported as a slab-out-of-bounds read in
ceph_mdsmap_decode(). The decoded values land in the internal
info->export_targets[] array, so the consequence is a kernel
out-of-bounds read, not an information leak to the attacker.
Impact: a malicious or compromised Ceph monitor sending an MDS map with
a per-mds info version of 2 or 3 and an oversized num_export_targets
field triggers an out-of-bounds read in the CephFS client kernel.
Add a ceph_decode_need() for the export-targets array before advancing
the cursor, so the bound is enforced for every info_v >= 2, not only
info_v >= 4. This mirrors the count-then-need idiom already used for
m_data_pg_pools later in the same function.
Compute the export-targets byte count with size_mul() and reuse that
checked length when advancing the cursor, so the attacker-controlled
num_export_targets multiplication fails closed on overflow rather than
relying on the later kcalloc() guard.
CVSS Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:N/A:H
CVSS Score: 9.1
AV:N - CEPH_MSG_MDS_MAP arrives over the kernel Ceph messenger TCP session from a monitor (mon_dispatch/extra_mon_dispatch) or MDS (mds_dispatch) and is decoded in ceph_mdsmap_decode(); a malicious cluster peer or on-path attacker on an unsigned msgr session reaches the bug with network data alone.
AC:L - The attacker fully controls per-mds info_v and num_export_targets; setting info_v to 2 or 3 (so info_end stays NULL) and an oversized count deterministically advances the cursor past msg->front and hits unchecked ceph_decode_32(), with no race or victim memory layout required.
PR:N - The attacker acts as the remote Ceph monitor or MDS peer (or MITM on msgr1/crc or auth_none); no local account, capability, or privilege on the victim CephFS client host is required.
UI:N - Once CephFS is mounted, the kernel client subscribes to MDS maps and ceph_mdsc_handle_mdsmap() runs automatically on each update; no further mount, open, or interactive victim action is required at exploit time.
S:U - The out-of-bounds read, any leaked heap bytes, and a possible oops stay inside the victim host kernel CephFS client and do not cross a VM, container, IOMMU, or other security authority.
C:H - An attacker-chosen num_export_targets drives an unbounded loop of unchecked ceph_decode_32() reads past the message front slab into adjacent kernel heap (KASAN slab-out-of-bounds); per kernel guidance an unbounded out-of-bounds read is Confidentiality High.
I:N - OOB words are copied only into a correctly sized kcalloc'd export_targets array; overshooting the front buffer then fails later ceph_decode_32_safe() so the map is destroyed and never installed, giving no out-of-bounds write or control-flow primitive.
A:H - Walking an attacker-chosen number of u32s past the kmalloc'd msg->front object into unmapped or guarded pages oopses the ceph-msgr worker, and a failed decode also shuts down the mount via ceph_umount_begin(); both are repeatable on each crafted map.
| Attack Vector |
Network |
Scope |
Unchanged |
| Attack Complexity |
Low |
Confidentiality Impact |
High |
| Privileges Required |
None |
Integrity Impact |
None |
| User Interaction |
None |
Availability Impact |
High |
AV:N - CEPH_MSG_MDS_MAP arrives over the kernel Ceph messenger TCP session from a monitor (mon_dispatch/extra_mon_dispatch) or MDS (mds_dispatch) and is decoded in ceph_mdsmap_decode(); a malicious cluster peer or on-path attacker on an unsigned msgr session reaches the bug with network data alone.
AC:L - The attacker fully controls per-mds info_v and num_export_targets; setting info_v to 2 or 3 (so info_end stays NULL) and an oversized count deterministically advances the cursor past msg->front and hits unchecked ceph_decode_32(), with no race or victim memory layout required.
PR:N - The attacker acts as the remote Ceph monitor or MDS peer (or MITM on msgr1/crc or auth_none); no local account, capability, or privilege on the victim CephFS client host is required.
UI:N - Once CephFS is mounted, the kernel client subscribes to MDS maps and ceph_mdsc_handle_mdsmap() runs automatically on each update; no further mount, open, or interactive victim action is required at exploit time.
S:U - The out-of-bounds read, any leaked heap bytes, and a possible oops stay inside the victim host kernel CephFS client and do not cross a VM, container, IOMMU, or other security authority.
C:H - An attacker-chosen num_export_targets drives an unbounded loop of unchecked ceph_decode_32() reads past the message front slab into adjacent kernel heap (KASAN slab-out-of-bounds); per kernel guidance an unbounded out-of-bounds read is Confidentiality High.
I:N - OOB words are copied only into a correctly sized kcalloc'd export_targets array; overshooting the front buffer then fails later ceph_decode_32_safe() so the map is destroyed and never installed, giving no out-of-bounds write or control-flow primitive.
A:H - Walking an attacker-chosen number of u32s past the kmalloc'd msg->front object into unmapped or guarded pages oopses the ceph-msgr worker, and a failed decode also shuts down the mount via ceph_umount_begin(); both are repeatable on each crafted map.
CVSS 3.1