GHSA-8XJR-H523-R9WR
Vulnerability from github – Published: 2026-09-25 12:31 – Updated: 2026-09-25 15:31In the Linux kernel, the following vulnerability has been resolved:
net/rds: acquire the fastpath locks in rds_conn_shutdown()
rds_conn_shutdown() quiesces the transmit and receive-refill paths by waiting for RDS_IN_XMIT and RDS_RECV_REFILL to be sampled clear, and then runs the transport shutdown and rds_conn_path_reset(). Sampling the bits clear is not the same as owning them: the moment after the wait_event() returns, rds_send_xmit() can re-acquire RDS_IN_XMIT (or rds_ib_recv_refill() can re-acquire RDS_RECV_REFILL) and run concurrently with the teardown.
The sender does recheck the connection state after taking the lock, but that recheck is a classic store-buffering pattern: teardown writes the state and reads the bit while the sender writes the bit and reads the state. acquire_in_xmit() is only an acquire operation, so on weakly ordered architectures both sides can miss each other's write, and the transmit path then runs while the transport zeroes its rings (e.g. rds_ib_ring_init()) and rds_send_path_reset() rewrites the transmit state under it.
Oracle UEK fixed the same class of crashes - a 14-year tail of BUG_ON()s in rds_ib_sub_signaled(), unexpected op-codes and NULL dereferences in rds_ib_send_cqe_handler() during failover testing - by making the teardown path acquire the fastpath bit locks instead of testing them ("rds: Make sure transmit path and connection tear-down does not run concurrently"). Ownership of a single word is decided by RMW atomicity, so no cross-variable ordering is needed.
Do the same here: take both locks before calling the transport shutdown, hold them across rds_conn_path_reset(), and release them explicitly with a wake-up afterwards. Both are released with clear_bit_unlock(), so that the ring re-initialization done by the transport shutdown and the transmit state rewritten by rds_send_path_reset() are ordered before either bit is seen clear by the next acquire_in_xmit() or acquire_refill().
The fastpath users of these bits - rds_send_xmit() and rds_ib_recv_refill() - are trylock style and back off while teardown owns the locks, so no new lock dependency is introduced for them. rds_tcp_reset_callbacks() is different: since the previous patch it acquires RDS_IN_XMIT as well, and it blocks doing so, so its wait now spans the teardown instead of at most one send batch. That waiter runs from rds_tcp_accept_one() on the single-threaded krdsd workqueue and holds rds_tcp_accept_lock and t_conn_path_lock while it waits, so a duelling SYN accepted while its path is being torn down parks accept processing for the duration of the teardown - for TCP bounded by the (up to 5 s) drain loop in rds_tcp_conn_path_shutdown(). An IB path's drain in rds_ib_conn_path_shutdown() has no round cap, but no blocking waiter either: rds_tcp_reset_callbacks() is the only blocking acquirer of these bits and waits only on its own TCP path, and the fastpaths are trylock-and-back-off on both transports, so a long IB drain lengthens only that path's own quiesce. The window is narrow: the accept-side state check has to pass before the teardown moves the path to RDS_CONN_DISCONNECTING.
Because krdsd is a single global workqueue, everything else queued there - accept processing for other connections and network namespaces, and the flush_workqueue(rds_wq) in rds_tcp_listen_stop() during namespace teardown - waits behind the parked accept worker for that time. It cannot deadlock, although the waits do point at each other: the teardown blocks until the bit's holder releases it, and the holder may be that krdsd accept worker. The holder finishes without needing anything the teardown owns: the sync cancels rds_tcp_reset_callbacks() issues target cp_send_w and cp_recv_w on the path's ordered cp_wq, whose only execution slot is occupied by the blocked cp_down_w itself, so they are pending at most and cancel without flushing - a reliance on cp_wq being ordered that is now noted next to those cancels (on ---truncated---
{
"affected": [],
"aliases": [
"CVE-2026-98069"
],
"database_specific": {
"cwe_ids": [],
"github_reviewed": false,
"github_reviewed_at": null,
"nvd_published_at": "2026-09-25T11:17:36Z",
"severity": "HIGH"
},
"details": "In the Linux kernel, the following vulnerability has been resolved:\n\nnet/rds: acquire the fastpath locks in rds_conn_shutdown()\n\nrds_conn_shutdown() quiesces the transmit and receive-refill paths by\nwaiting for RDS_IN_XMIT and RDS_RECV_REFILL to be sampled clear, and\nthen runs the transport shutdown and rds_conn_path_reset(). Sampling\nthe bits clear is not the same as owning them: the moment after the\nwait_event() returns, rds_send_xmit() can re-acquire RDS_IN_XMIT (or\nrds_ib_recv_refill() can re-acquire RDS_RECV_REFILL) and run\nconcurrently with the teardown.\n\nThe sender does recheck the connection state after taking the lock,\nbut that recheck is a classic store-buffering pattern: teardown writes\nthe state and reads the bit while the sender writes the bit and reads\nthe state. acquire_in_xmit() is only an acquire operation, so on\nweakly ordered architectures both sides can miss each other\u0027s write,\nand the transmit path then runs while the transport zeroes its rings\n(e.g. rds_ib_ring_init()) and rds_send_path_reset() rewrites the\ntransmit state under it.\n\nOracle UEK fixed the same class of crashes - a 14-year tail of\nBUG_ON()s in rds_ib_sub_signaled(), unexpected op-codes and NULL\ndereferences in rds_ib_send_cqe_handler() during failover testing -\nby making the teardown path *acquire* the fastpath bit locks instead\nof testing them (\"rds: Make sure transmit path and connection\ntear-down does not run concurrently\"). Ownership of a single word is\ndecided by RMW atomicity, so no cross-variable ordering is needed.\n\nDo the same here: take both locks before calling the transport\nshutdown, hold them across rds_conn_path_reset(), and release them\nexplicitly with a wake-up afterwards. Both are released with\nclear_bit_unlock(), so that the ring re-initialization done by the\ntransport shutdown and the transmit state rewritten by\nrds_send_path_reset() are ordered before either bit is seen clear by\nthe next acquire_in_xmit() or acquire_refill().\n\nThe fastpath users of these bits - rds_send_xmit() and\nrds_ib_recv_refill() - are trylock style and back off while teardown\nowns the locks, so no new lock dependency is introduced for them.\nrds_tcp_reset_callbacks() is different: since the previous patch it\nacquires RDS_IN_XMIT as well, and it blocks doing so, so its wait now\nspans the teardown instead of at most one send batch. That waiter\nruns from rds_tcp_accept_one() on the single-threaded krdsd workqueue\nand holds rds_tcp_accept_lock and t_conn_path_lock while it waits, so\na duelling SYN accepted while its path is being torn down parks\naccept processing for the duration of the teardown - for TCP bounded\nby the (up to 5 s) drain loop in rds_tcp_conn_path_shutdown(). An IB\npath\u0027s drain in rds_ib_conn_path_shutdown() has no round cap, but no\nblocking waiter either: rds_tcp_reset_callbacks() is the only blocking\nacquirer of these bits and waits only on its own TCP path, and the\nfastpaths are trylock-and-back-off on both transports, so a long IB\ndrain lengthens only that path\u0027s own quiesce. The\nwindow is narrow: the accept-side state check has to pass before the\nteardown moves the path to RDS_CONN_DISCONNECTING.\n\nBecause krdsd is a single global workqueue, everything else queued\nthere - accept processing for other connections and network\nnamespaces, and the flush_workqueue(rds_wq) in rds_tcp_listen_stop()\nduring namespace teardown - waits behind the parked accept worker for\nthat time. It cannot deadlock, although the waits do point at each\nother: the teardown blocks until the bit\u0027s holder releases it, and\nthe holder may be that krdsd accept worker. The holder finishes\nwithout needing anything the teardown owns: the sync cancels\nrds_tcp_reset_callbacks() issues target cp_send_w and cp_recv_w on\nthe path\u0027s ordered cp_wq, whose only execution slot is occupied by\nthe blocked cp_down_w itself, so they are pending at most and cancel\nwithout flushing - a reliance on cp_wq being ordered that is now\nnoted next to those cancels (on \n---truncated---",
"id": "GHSA-8xjr-h523-r9wr",
"modified": "2026-09-25T15:31:44Z",
"published": "2026-09-25T12:31:35Z",
"references": [
{
"type": "ADVISORY",
"url": "https://nvd.nist.gov/vuln/detail/CVE-2026-98069"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/1fe627e5db5c3f53a9f9f9c8a66671755d306955"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/7febb113795d5de5b690b208f4b0e64a5fad1201"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/813f3582ac7ae9f60f917937d54660e0952d5f2d"
},
{
"type": "WEB",
"url": "https://git.kernel.org/stable/c/900e96c9749a06833801f60c393aa1d405ea226c"
}
],
"schema_version": "1.4.0",
"severity": [
{
"score": "CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:H/I:H/A:H",
"type": "CVSS_V3"
}
]
}
Sightings
| Author | Source | Type | Date | Other |
|---|
Nomenclature
- Seen: The vulnerability was mentioned, discussed, or observed by the user.
- Confirmed: The vulnerability has been validated from an analyst's perspective.
- Published Proof of Concept: A public proof of concept is available for this vulnerability.
- Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
- Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
- Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
- Not confirmed: The user expressed doubt about the validity of the vulnerability.
- Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.
The approach is described in our paper Mapping CVEs to MITRE ATT&CK Techniques: A Curated Gold-Set Classifier and the Limits of LLM-Assisted Label Expansion.
Browse all ATT&CK techniques and the vulnerabilities related to each.
Related by attack behaviour
Vulnerabilities whose description is nearest to this one in the vector space of the CIRCL/vulnerability-attack-technique-biencoder model. This is a similarity search over the bi-encoder space (plain cosine), not a classification, and it has no measured accuracy.