UBUNTU-CVE-2026-99296: Medium severity ubuntu/linux vulnerability
In the Linux kernel, the following vulnerability has been resolved: libceph: bound moncommand ack decode to front len handlecommandack() bounds the decode with msg->frontalloclen, the size of the reply buffer, instead of msg->front.iovlen, the number of bytes the monitor sent. monallocmsg() routes every CEPHMSGMONCOMMANDACK to getgenericreply(), which returns the preallocated ->reply of whichever generic request carries the same tid, so the buffer keeps its fixed size however short the reply is. An ack with hdr.frontlen below sizeof(struct cephmonrequestheader) + sizeof(u32), including 0, still passes cephdecodeneed(), and cephdecode32() returns four bytes that were never received: uninitialized slab memory, stale bytes of an earlier message, or bytes the peer planted by sending a reply that lands in the same buffer and is then rejected. The read stays within the allocation, so KASAN does not flag it. req->result reaches userspace through cephmoncdostatfs(), cephmoncgetversion() and cephmoncblocklistadd(); rbd hands it back from write() on /sys/bus/rbd/addsinglemajor, and rbdgetclient() passes it to ERRPTR() unchecked, which faults: rbd: failed to get latest osdmap: -186318427 BUG: unable to handle page fault for address: fffffbfffe9ca034 Oops: Oops: 0000 [#1] SMP KASAN NOPTI RIP: 0010:dorbdadd (drivers/block/rbd.c:7105) Call Trace: ... kernfsfopwriteiter (fs/kernfs/file.c:345) vfswrite (fs/readwrite.c:595 fs/readwrite.c:687) ksyswrite (fs/readwrite.c:739) dosyscall64 (arch/x86/entry/syscall64.c:61 arch/x86/entry/syscall64.c:84) entrySYSCALL64afterhwframe (arch/x86/entry/entry64.S:121) Kernel panic - not syncing: Fatal exception Use msg->front.iovlen, matching handlesubscribeack(), handlestatfsreply(), handlegetversionreply() and handleauthreply(). A well-formed ack has front.iovlen >= 22 and is unaffected; a short one takes the existing bad: path. This applies commit d3c32939fa0e ("libceph: bound getversion reply decode to front len") to the last decode-side use of frontalloclen in this file.
Affected Software
Remediation
Recommended actions to resolve this vulnerability, in priority order.
- Compensating control
In handle_command_ack(), bound the decode using msg->front.iov_len instead of msg->front_alloc_len.
Event History
Frequently Asked Questions
Which Ubuntu kernel package families are identified as affected?
The affected software list includes Ubuntu linux, linux-fips, linux-aws, linux-azure, linux-gcp, linux-oracle, linux-aws-fips, and linux-azure-fips. No fixed package versions are provided in the available data.
What must an attacker be able to do to trigger the issue?
The attacker must be able to cause libceph to process a malformed CEPH_MSG_MON_COMMAND_ACK reply whose front length is too short for the fields decoded by handle_command_ack(). Such a reply can cause the kernel to use bytes that were not received when setting the command result.
What observable impact can result from a malformed reply?
The decoded result can reach userspace through Ceph monitor operations including statfs, version retrieval, and blocklist addition. In the rbd path, an unchecked result passed to ERR_PTR() can fault, with the description indicating an rbd failure path.
Will KASAN reliably detect exploitation of this flaw?
No. The invalid read remains within the allocated reply buffer, so KASAN does not flag it even though the bytes being decoded were not received in the reply.