CVE-2026-72466

xprtrdma: Fix bcall rep leak and unbounded peek

Published Aug 15, 2026 CVSS 3.1 · 9.8 Critical EPSS < 1% KEV · Not listed
Description
In the Linux kernel, the following vulnerability has been resolved:

xprtrdma: Fix bcall rep leak and unbounded peek

rpcrdma_is_bcall() decodes a reply's first words to decide whether
the frame is a backchannel call. Two issues in that decode path
let a short or malformed reply leak the receive buffer and drain
the Receive queue.

First, the speculative peek

p = xdr_inline_decode(xdr, 0);
/* five p++ reads follow */

asks xdr_inline_decode() for zero bytes, which returns xdr->p
without consulting xdr->end. The five subsequent __be32 reads can
then walk up to 20 bytes past the wire payload into stale regbuf
contents and misclassify the reply as a backchannel call.

Second, after the post-peek

p = xdr_inline_decode(xdr, 3 * sizeof(*p));
if (unlikely(!p))
return true;

the short-header arm returns true without calling
rpcrdma_bc_receive_call(). The contract with the caller is that a
true return transfers ownership of rep to the backchannel path:

rpcrdma_reply_handler()
if (rpcrdma_is_bcall(r_xprt, rep))
return; /* bare return, skips out_post */
...
out_post:
rpcrdma_post_recvs(r_xprt, credits + ...);

Because rpcrdma_bc_receive_call() never ran, no one took rep, but
rpcrdma_reply_handler still bare-returns past rpcrdma_rep_put()
and rpcrdma_post_recvs(). The rep, with its persistently
DMA-mapped receive buffer, is orphaned on rb_all_reps and freed
only at transport teardown. This completion reposts nothing, so
its slot is reclaimed only when a later forward-channel reply
reaches out_post and rpcrdma_post_recvs() allocates a fresh rep to
backfill; absent that traffic the Receive queue drains and the
peer's Sends draw RNR NAKs.

Fix by consulting xdr->end after the zero-length peek so the five
__be32 reads cannot run unless 20 bytes of wire payload remain. A
byte-precise comparison against xdr->end is required because a
non-4-aligned receive rounds the stream's word count up past the
true payload. Also return false from the short-header arm so the
reply falls through the normal out_norqst cleanup chain
(rpcrdma_rep_put() plus rpcrdma_post_recvs()).

Remediation

No solution or workaround provided in the CVE record

Tracking

Sign in to view the affected projects.

Advisories

Source ID Title
Ubuntu USN Ubuntu USN USN-8781-1 Linux kernel (NVIDIA Tegra) vulnerabilities
Ubuntu USN Ubuntu USN USN-8816-1 Linux kernel vulnerabilities
Ubuntu USN Ubuntu USN USN-8816-2 Linux kernel vulnerabilities
Ubuntu USN Ubuntu USN USN-8816-3 Linux kernel (Oracle) vulnerabilities
Ubuntu USN Ubuntu USN USN-8816-4 Linux kernel (GKE) vulnerabilities
Ubuntu USN Ubuntu USN USN-8876-1 Linux kernel (Azure CVM) vulnerabilities
Ubuntu USN Ubuntu USN USN-8875-1 Linux kernel vulnerabilities
Ubuntu USN Ubuntu USN USN-8877-1 Linux kernel (GCP) vulnerabilities
Ubuntu USN Ubuntu USN USN-8878-1 Linux kernel (GCP) vulnerabilities
Ubuntu USN Ubuntu USN USN-8879-1 Linux kernel (Oracle) vulnerabilities
Ubuntu USN Ubuntu USN USN-8888-1 Linux kernel (Azure) vulnerabilities
Ubuntu USN Ubuntu USN USN-8888-2 Linux kernel (Azure) vulnerabilities
Ubuntu USN Ubuntu USN USN-8875-2 Linux kernel (NVIDIA) vulnerabilities
Ubuntu USN Ubuntu USN USN-8917-1 Linux kernel vulnerabilities
Ubuntu USN Ubuntu USN USN-8918-1 Linux kernel (Azure) vulnerabilities
Ubuntu USN Ubuntu USN USN-9003-1 Linux kernel (NVIDIA) vulnerabilities
Ubuntu USN Ubuntu USN USN-9024-1 Linux kernel vulnerabilities
Ubuntu USN Ubuntu USN USN-9025-1 Linux kernel vulnerabilities
Ubuntu USN Ubuntu USN USN-9026-1 Linux kernel (Azure) vulnerabilities
Ubuntu USN Ubuntu USN USN-9029-1 Linux kernel (Raspberry Pi) vulnerabilities
Ubuntu USN Ubuntu USN USN-9040-1 Linux kernel (Low Latency NVIDIA) vulnerabilities
Ubuntu USN Ubuntu USN USN-9043-1 Linux kernel (NVIDIA) vulnerabilities
Ubuntu USN Ubuntu USN USN-9066-1 Linux kernel (HWE) vulnerabilities
History
Type Values Removed Values Added
Metrics ssvc

{'options': {'Automatable': 'yes', 'Exploitation': 'none', 'Technical Impact': 'total'}, 'version': '2.0.3'}

Type Values Removed Values Added
Weaknesses CWE-119
CWE-400
Type Values Removed Values Added
Weaknesses CWE-119
CWE-400
Type Values Removed Values Added
Weaknesses CWE-125
Type Values Removed Values Added
Metrics cvssV3_1

{'score': 9.8, 'vector': 'CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H'}

Type Values Removed Values Added
Weaknesses CWE-125
Type Values Removed Values Added
Description In the Linux kernel, the following vulnerability has been resolved: xprtrdma: Fix bcall rep leak and unbounded peek rpcrdma_is_bcall() decodes a reply's first words to decide whether the frame is a backchannel call. Two issues in that decode path let a short or malformed reply leak the receive buffer and drain the Receive queue. First, the speculative peek p = xdr_inline_decode(xdr, 0); /* five p++ reads follow */ asks xdr_inline_decode() for zero bytes, which returns xdr->p without consulting xdr->end. The five subsequent __be32 reads can then walk up to 20 bytes past the wire payload into stale regbuf contents and misclassify the reply as a backchannel call. Second, after the post-peek p = xdr_inline_decode(xdr, 3 * sizeof(*p)); if (unlikely(!p)) return true; the short-header arm returns true without calling rpcrdma_bc_receive_call(). The contract with the caller is that a true return transfers ownership of rep to the backchannel path: rpcrdma_reply_handler() if (rpcrdma_is_bcall(r_xprt, rep)) return; /* bare return, skips out_post */ ... out_post: rpcrdma_post_recvs(r_xprt, credits + ...); Because rpcrdma_bc_receive_call() never ran, no one took rep, but rpcrdma_reply_handler still bare-returns past rpcrdma_rep_put() and rpcrdma_post_recvs(). The rep, with its persistently DMA-mapped receive buffer, is orphaned on rb_all_reps and freed only at transport teardown. This completion reposts nothing, so its slot is reclaimed only when a later forward-channel reply reaches out_post and rpcrdma_post_recvs() allocates a fresh rep to backfill; absent that traffic the Receive queue drains and the peer's Sends draw RNR NAKs. Fix by consulting xdr->end after the zero-length peek so the five __be32 reads cannot run unless 20 bytes of wire payload remain. A byte-precise comparison against xdr->end is required because a non-4-aligned receive rounds the stream's word count up past the true payload. Also return false from the short-header arm so the reply falls through the normal out_norqst cleanup chain (rpcrdma_rep_put() plus rpcrdma_post_recvs()).
Title xprtrdma: Fix bcall rep leak and unbounded peek
First Time appeared Linux
Linux linux Kernel
CPEs cpe:2.3:o:linux:linux_kernel:*:*:*:*:*:*:*:*
Vendors & Products Linux
Linux linux Kernel
References
Weaknesses

Sources

Status
PUBLISHED
Assigner
Linux
Published
Updated
Oct 16, 2026, 6:55 PM
Reserved
Aug 9, 2026, 3:40 AM
Status
Received
Published
Aug 15, 2026, 6:22 AM
Modified
Oct 16, 2026, 7:17 PM
Severity
Moderate
Public date
Aug 15, 2026, 12:00 AM