CVE-2026-31400
Description
In the Linux kernel, the following vulnerability has been resolved:
sunrpc: fix cache_request leak in cache_release
When a reader's file descriptor is closed while in the middle of reading a cache_request (rp->offset != 0), cache_release() decrements the request's readers count but never checks whether it should free the request.
In cache_read(), when readers drops to 0 and CACHE_PENDING is clear, the cache_request is removed from the queue and freed along with its buffer and cache_head reference. cache_release() lacks this cleanup.
The only other path that frees requests with readers == 0 is cache_dequeue(), but it runs only when CACHE_PENDING transitions from set to clear. If that transition already happened while readers was still non-zero, cache_dequeue() will have skipped the request, and no subsequent call will clean it up.
Add the same cleanup logic from cache_read() to cache_release(): after decrementing readers, check if it reached 0 with CACHE_PENDING clear, and if so, dequeue and free the cache_request.
AI Insight
LLM-synthesized narrative grounded in this CVE's description and references.
Memory leak in Linux kernel's sunrpc cache when a reader's file descriptor is closed mid-read; fix adds cleanup in cache_release.
Vulnerability
CVE-2026-31400 is a memory leak vulnerability in the Linux kernel's sunrpc subsystem. When a reader closes its file descriptor while in the middle of reading a cache_request (indicated by rp->offset != 0), the cache_release() function decrements the request's reader count but fails to check whether the request should be freed. This occurs because the cleanup logic present in cache_read()—which frees the request when readers reach zero and CACHE_PENDING is clear—is missing in cache_release(). Additionally, cache_dequeue() only cleans up when CACHE_PENDING transitions from set to clear; if that transition happens while readers are still non-zero, the request is never freed [1][2].
Exploitation
An attacker with the ability to open and read sunrpc cache files (typically requiring local access or a container breakout) can cause repeated partial reads and close the file descriptor before completing a read. This action triggers the leak each time, as the kernel fails to release the memory allocated for the cache_request structure and its associated buffer and cache_head reference [3]. No special privileges beyond read access to the cache are needed.
Impact
By repeatedly opening cache files and closing them mid-read, an attacker can exhaust kernel memory, leading to a denial-of-service (DoS) condition. The leak is per-operation, so sustained exploitation can eventually cause system instability or panic due to memory starvation. The CVSS v3 score of 5.5 reflects the medium severity, with local access required and availability impact [4].
Mitigation
The vulnerability is fixed in Linux kernel stable releases that include commits 373457de1428, 301670dcd098, be5c35960e5a, and f18c1f2a88ca [1][2][3][4]. Users should update their kernels to the latest patched versions. No workaround is available; the fix must be applied via a kernel update.
AI Insight generated on May 20, 2026. Synthesized from this CVE's description and the cited reference URLs; citations are validated against the source bundle.
Affected products
1Patches
0No patches discovered yet.
Vulnerability mechanics
AI mechanics synthesis has not run for this CVE yet.
References
8- git.kernel.org/stable/c/17ad31b3a43b72aec3a3d83605891e1397d0d065nvdPatch
- git.kernel.org/stable/c/1dfedb293943e491379c9302b428e6f920a73d12nvdPatch
- git.kernel.org/stable/c/301670dcd098c1fe5c2fe90fb3c7a8f4814d2351nvdPatch
- git.kernel.org/stable/c/373457de14281c1fc7cace6fc4c8a267fc176673nvdPatch
- git.kernel.org/stable/c/41f6ba6c98a618043d2cd71030bf9a752dfab8b2nvdPatch
- git.kernel.org/stable/c/7bcd5e318876ac638c8ceade7a648e76ac8c48e1nvdPatch
- git.kernel.org/stable/c/be5c35960e5ead70862736161836e2d1bc7352dcnvdPatch
- git.kernel.org/stable/c/f18c1f2a88ca91357916997cdb0f7adaf14fc497nvdPatch
News mentions
0No linked articles in our index yet.