Skip to main content

Syself security assessment

CVE-2026-98068 in Syself Kubernetes

In the Linux kernel, the following vulnerability has been resolved: net/rds: don't let rds_conn_shutdown() consume a concurrent drop rds_conn_shutdown() finishes by moving the path from RDS_CONN_DISCONNECTING to RDS_CONN_DOWN, and also accepts RDS_CONN_ERROR as the starting state of that final transition, so that a FIN processed in softirq context during the teardown does not derail the shutdown into a noisy error path. But consuming that RDS_CONN_ERROR also consumes the shutdown pass that came with it: rds_conn_path_drop() sets RDS_CONN_ERROR and then queues cp_down_w, and a pass that starts on a path already in RDS_CONN_DOWN is a no-op. For the FIN case that is harmless - the socket the FIN arrived on is the very socket the teardown just released. It is not harmless for a dropper that attached something to the path first. rds_tcp_accept_one() is such a dropper. Its path claim in rds_tcp_accept_one_path() transitions RDS_CONN_DOWN -> RDS_CONN_CONNECTING, and a concurrent drop - a FIN on a previous socket in softirq context, an administrative reset - can put the path into RDS_CONN_ERROR between that claim and the state check that follows, which accepts RDS_CONN_ERROR. The accept then installs the freshly accepted socket with rds_tcp_set_callbacks() while the queued teardown - which sampled tc->t_sock before this socket existed - is still running. rds_connect_path_complete() fails its transition to RDS_CONN_UP and drops the path again, queueing the pass that should reap the socket it just installed. If the in-flight shutdown's final transition consumes that drop's RDS_CONN_ERROR, the queued pass finds the path in RDS_CONN_DOWN and does nothing. The installed socket is never torn down: it sits established with its callbacks armed and its rds_tcp_connection on rds_tcp_tc_list, the peer sees a connection that nothing ever reads, and the path is wedged in RDS_CONN_DOWN until some later event drops it again. Reproduced with widened race windows as an ever-growing receive queue on a socket owned by a path stuck in RDS_CONN_DOWN, with the peer's send path wedged behind it. Make the final transition only DISCONNECTING -> DOWN. If it fails because the path is in RDS_CONN_ERROR, a drop raced the teardown: cancel the reconnect timer and clear RDS_RECONNECT_PENDING - the one piece of the skipped tail that must not be left behind - and return, letting the pass the drop queued finish the job: it tears down whatever attached to the path in the meantime, completes the transition to RDS_CONN_DOWN, and re-arms the reconnect from its own tail. The timer quiesce in that branch matters because the racing drop does not always queue that pass: rds_conn_path_drop() returns without queueing when a destroy is pending - exactly the situation during a netns teardown or module unload, when a FIN on the dying socket is processed while rds_conn_path_destroy() flushes cp_down_w. If the flushed pass is the one that takes this return, no later pass exists, and rds_conn_path_destroy() would find cp_conn_w still armed (WARN_ON) and then free a path whose reconnect timer can still fire. With the cancel in the branch, every exit of a shutdown pass leaves the timer quiesced no matter which pass completes the transition. The FIN case keeps making progress, one pass later and still without noisy logging. Any other state keeps today's rds_conn_path_error() handling; no current cp_state writer can leave a DISCONNECTING path in anything but RDS_CONN_ERROR (every other writer is a cmpxchg from a non-DISCONNECTING state), so that branch is defensive. On kernels without the preceding patches the same hazard exists with the sample-based quiesce; the fix applies there equally.

Syself Autopilot is fully managed Kubernetes on your own Hetzner servers. Every cluster runs a versioned cluster stack release, and this page tracks how CVE-2026-98068 affects the kernel component those releases ship, based on Syself's assessments. For the general description and scoring of the vulnerability, see the references.

Not affected in 1.36 v6 Upstream severity: Medium Last assessed

Syself Kubernetes is not affected by CVE-2026-98068 in any of the 6 releases that ship kernel.

Reason: The vulnerable code is not included.

Our assessment in 1.36 v6, the newest release with this finding

kernel 6.18.49
Not affected Medium
The vulnerable code is not included

The affected code is in the Reliable Datagram Sockets protocol. That protocol is disabled and not built in the shipped kernel, so the vulnerable connection-shutdown code is absent on the deployed nodes.

Syself Security

Status by release

ReleasesComponentStatusReasonFixed inAssessment
kernel 6.18.42, 6.18.45, 6.18.49Not affectedThe vulnerable code is not included—Details

Frequently asked questions

My scanner reports CVE-2026-98068 in a Syself cluster. What should I do?

Scanners such as Trivy and Grype match package versions, so they report CVE-2026-98068 whenever the vulnerable version is present, even when it cannot be exploited. Syself publishes its assessments as a VEX (Vulnerability Exploitability eXchange) document for every release. Pass it to your scanner to drop the findings Syself has assessed as not affected or fixed:

trivy image --vex hetzner-apalla-1-36-v6-advisory.openvex.json <image>
grype <image> --vex hetzner-apalla-1-36-v6-advisory.openvex.json

Each assessment applies to the package version it names. A finding stays in the report when the document has no assessment for that package, or when the vulnerability is still affected or under investigation.

Download the OpenVEX document for 1.36 v6

What do the statuses mean?

Not affected
The component ships in the release, but Syself has assessed that the vulnerability cannot be exploited there. The reason column says why.
Fixed
The release ships a version of the component that carries the fix.
Affected
The vulnerability can affect the release. Upgrade to a release listed as fixed once one is available.
Under Investigation
Syself is still assessing whether the vulnerability affects the release.

Why can the severity differ from other sources?

The severity shown is the upstream rating of the vulnerability itself. The status is Syself's assessment of whether it can be exploited in the components Syself ships, so a critical vulnerability can still be not affected.

References

Think this assessment is wrong, or found a vulnerability in Syself? Contact us.