Skip to main content
CVE-2026-64268VulnerabilityPublished 07/25/2026

In the Linux kernel, the following vulnerability has been resolved: RDMA/siw: bound Read Response placement to the RREAD length In… (CVE-2026-64268)

Source GCVE Database

0
Severity
High

No CVSS score published yet

Exploitation

Not in CISA KEV, no exploitation reported

EPSS
0.7% chance of exploitation in 30 days – higher than 53% of CVEsExploit Prediction Scoring System: the likelihood attackers will use this flaw soon.

What to do

AI-suggested

A fix is available that bounds each Read Response segment before placement and terminates connections that would cause buffer overruns. Users should apply the official Linux kernel patches that address this vulnerability.

Description

In the Linux kernel, the following vulnerability has been resolved: RDMA/siw: bound Read Response placement to the RREAD length In drivers/infiniband/sw/siw/siw_qp_rx.c, siw_proc_rresp() places each inbound Read Response DDP segment at sge->laddr + wqe->processed and then accumulates wqe->processed, but it never checks the running total against the sink buffer length on continuation segments. siw_check_sge() resolves and validates the sink memory only on the first fragment (the if (!*mem) branch), and siw_rresp_check_ntoh() compares the cumulative length against wqe->bytes only on the final segment (the !frx->more_ddp_segs guard). A connected siw peer that answers an outstanding RREAD with Read Response segments that keep the DDP Last flag clear, carrying more total payload than the RREAD requested, drives wqe->processed past the validated sink buffer; the next siw_rx_data() call writes out of bounds at sge->laddr + wqe->processed. siw runs iWARP over ordinary routable TCP, so the peer is the remote end of an established RDMA connection and needs no local privilege. Bound every segment before placement, exactly as siw_proc_send() and siw_proc_write() already do for their tagged and untagged paths, and terminate the connection with a base-or-bounds DDP error when the Read Response would overrun the sink buffer. This is the second receive-path length fix for this file. A separate change rejects an MPA FPDU length that underflows the per-fragment remainder in the header decode; that guard does not cover this case, because here each individual segment length is self-consistent and only the accumulated placement offset overruns the buffer.

Impact

A remote attacker with an established RDMA connection can cause out-of-bounds memory writes in kernel space by sending malicious Read Response segments that exceed the requested length. This could lead to memory corruption, potential kernel crashes, or other undefined behavior. No local privileges are required, but the attacker must be a connected RDMA peer.

Summary

The Linux kernel's siw RDMA driver (infiniband/sw/siw/siw_qp_rx.c) improperly handles the placement of inbound Read Response DDP segments by accumulating processed lengths without verifying the total against the sink buffer length on continuation segments. While initial fragment validation occurs, subsequent segments can cause wqe->processed to exceed the validated buffer size, resulting in out-of-bounds writes at sge->laddr + wqe->processed. This vulnerability can be triggered by a remote connected siw peer over an established RDMA connection without requiring local privileges. The fix involves bounding every segment before placement and terminating the connection with a DDP error if an overrun would occur.

References

Community reviews

Loading reviews…