Description

In the Linux kernel, the following vulnerability has been resolved: libceph: reject buckets with mismatched CRUSH ids crush_decode() stores bucket data by array slot, and the mapper later derives the per-bucket workspace index from the decoded bucket id. A malformed map can therefore make one bucket reuse another bucket's workspace by encoding an id different from -1 - slot. For uniform buckets, the second replica selection expands the source bucket's permutation into that aliased workspace buffer. If the source bucket is larger than the aliased bucket, the write runs past the smaller permutation array and can escape the kvmalloc'd CRUSH workspace. KASAN reports a slab OOB write of 4 bytes in bucket_perm_choose(). Reject buckets whose encoded id does not match their array slot. Valid CRUSH maps already use the canonical negative id corresponding to the bucket slot, so this restores the invariant expected by work->work[-1 - in->id] without changing valid map behavior.

Executive Summary

Technical Details

Mitigations

References

Additional Metadata

{
  "score": "9.8",
  "pubDate": "2026-09-11T20:19:51.813Z",
  "pubdate": "2026-09-11T20:19:51.813Z",
  "executiveSummary": "This vulnerability involves an out-of-bounds (OOB) memory write within the Linux kernel's libceph CRUSH (Controlled Replication Under Scalable Hashing) map decoding logic.\nThe issue stems from an insufficient validation of CRUSH bucket identifiers, which allows a malformed CRUSH map to induce a workspace index aliasing scenario.\nBy manipulating the bucket ID during the decoding process, an attacker can force the system to reuse a workspace buffer intended for a different bucket, leading to heap memory corruption.\nThe vulnerability manifests as a slab out-of-bounds write of 4 bytes in the bucket_perm_choose() function when processing uniform buckets.\nImpacts include potential kernel memory corruption, system instability (kernel panic), and potential arbitrary code execution or privilege escalation if the memory corruption can be weaponized by a malicious actor capable of injecting a crafted CRUSH map into the Ceph cluster configuration.\nExploitation requires the ability to provide a malformed CRUSH map to the Ceph client or server components, which typically requires administrative privileges within the Ceph infrastructure.",
  "technicalDetails": "The vulnerability resides in the interaction between crush_decode() and the workspace allocation logic in libceph. The CRUSH algorithm relies on a workspace structure to store per-bucket permutations during replica selection. The indexing of this workspace is derived directly from the bucket ID stored within the CRUSH map: work->work[-1 - in->id].\nDuring map decoding, crush_decode() assigns bucket data into specific array slots. The vulnerability exists because the CRUSH map format does not enforce a strict one-to-one mapping between the array slot index and the encoded bucket ID. An attacker can craft a malformed CRUSH map where the encoded bucket ID does not correspond to the canonical -1 - slot identifier. This discrepancy enables a bucket to claim the workspace buffer of another bucket.\nThe exploitation flow proceeds as follows: 1) A malicious CRUSH map is loaded into the kernel. 2) During the decoding process, crush_decode() reads the fraudulent bucket ID, causing the kernel to associate the bucket with an incorrect workspace index. 3) When bucket_perm_choose() is called for a uniform bucket, it performs a permutation expansion. 4) If the malicious bucket is configured to be larger than the buffer of the aliased (smaller) bucket, the write operation proceeds past the bounds of the smaller buffer.\nThis leads to an out-of-bounds slab write in kernel space. Because the overflow writes past the allocated permutation array, it corrupts adjacent heap memory objects managed by the kernel slab allocator. This can lead to the corruption of adjacent structures, metadata, or function pointers.\nThe vulnerability is triggered without requiring complex interaction; it relies on the kernel parsing a malicious map structure, which is then used for subsequent I/O requests. The lack of an invariant check between the bucket ID and the array index effectively bypasses the memory safety expected during CRUSH map operations. The resulting KASAN report confirms a 4-byte slab OOB write during the expansion phase, demonstrating that the kernel fails to perform bounds checking against the actual allocated size of the target workspace buffer, instead relying on the trusted, yet malformed, bucket ID.\nThe impact is critical in environments where CRUSH map updates are sourced from potentially untrusted or compromised management interfaces, as successful exploitation results in kernel-level memory corruption."
}