CVE-2026-98347
N/A
Summary
In the Linux kernel, the following vulnerability has been resolved:
IB/IPoIB: Avoid restoring OPER_UP after multicast flush
ipoib_ib_dev_flush_light() temporarily clears IPOIB_FLAG_OPER_UP to prevent multicast joins while ipoib_mcast_dev_flush() is running, and restores the flag afterwards if it was previously set.
This restore races with ipoib_ib_dev_down(). If the interface is brought down while the flush is in progress, ipoib_ib_dev_down() clears IPOIB_FLAG_OPER_UP, but the flush path may set it again after the device has already gone down.
Since commit 894021a75291 ("IB/ipoib: Make the carrier_on_task race aware"), ipoib_mcast_carrier_on_task() relies on IPOIB_FLAG_OPER_UP being cleared to terminate its rtnl_trylock() retry loop. If the flag is left set after shutdown, the workqueue retries forever, causing teardown to deadlock when ipoib_ndo_uninit() waits in destroy_workqueue() while holding RTNL.
Instead of overloading IPOIB_FLAG_OPER_UP to block multicast joins during a light flush, introduce a dedicated IPOIB_FLAG_MCAST_FLUSH flag. Use it together with IPOIB_FLAG_OPER_UP to determine whether multicast joins are allowed, avoiding the race with device shutdown.
Affected Software
| Vendor | Product | Version Range | Status |
|---|---|---|---|
| Linux | Linux | 344bacca8cd811809fc33a249f2738ab757d327f < f973769fb7cb33ba0d4a64b7dcfee0cabaead938 | affected |
| Linux | Linux | 344bacca8cd811809fc33a249f2738ab757d327f < b72f38929f4b5aa83161fd3b22a7327207e51ec5 | affected |
| Linux | Linux | 344bacca8cd811809fc33a249f2738ab757d327f < 1b11e4b55b41d9e69a8e8d07622614202e2eaca9 | affected |
| Linux | Linux | 344bacca8cd811809fc33a249f2738ab757d327f < 326f7d34bd7e64de535566b65c0533b640df5a81 | affected |
| Linux | Linux | 344bacca8cd811809fc33a249f2738ab757d327f < 5ad925d45f28464a5e043c9620880bb6a030da89 | affected |
| Linux | Linux | 344bacca8cd811809fc33a249f2738ab757d327f < 188b334a6db36a8e0bfaaf59bc4020639a0f3aad | affected |
| Linux | Linux | 344bacca8cd811809fc33a249f2738ab757d327f < 1ff3add37c329704ec46181b4ae4f00f9f16ce6f | affected |
| Linux | Linux | 344bacca8cd811809fc33a249f2738ab757d327f < 9a141d3dc869d18b2eab35e999f4790a9b84e40f | affected |
| Linux | Linux | a289c65ca49d2680a3d797e633e57a45573d1df2 | affected |
| Linux | Linux | b81459c78935e4579a577b2366c5f8878d3c7fee | affected |
| Linux | Linux | 8a9d8dc9ce83a90bfbbe69181acc4cf1beff46f5 | affected |
| Linux | Linux | dfe809702a11bd7fb91c882fb55620e39ce22109 | affected |
| Linux | Linux | 87160fb51bc343793cc8f5f031a87f1a35aecb82 | affected |
| Linux | Linux | 749fd55dd210f9676f0d445a6efaaa3bec65174a | affected |
| Linux | Linux | 3.2.84 < 3.3 | affected |
| Linux | Linux | 3.10.105 < 3.11 | affected |
| Linux | Linux | 3.12.65 < 3.13 | affected |
| Linux | Linux | 3.16.39 < 3.17 | affected |
| Linux | Linux | 4.4.24 < 4.5 | affected |
| Linux | Linux | 4.7.7 < 4.8 | affected |
| Linux | Linux | 4.8 | affected |
| Linux | Linux | 0 < 4.8 | unaffected |
| Linux | Linux | 5.10.271 <= 5.10.* | unaffected |
| Linux | Linux | 5.15.222 <= 5.15.* | unaffected |
| Linux | Linux | 6.1.189 <= 6.1.* | unaffected |
| Linux | Linux | 6.6.158 <= 6.6.* | unaffected |
| Linux | Linux | 6.12.112 <= 6.12.* | unaffected |
| Linux | Linux | 6.18.54 <= 6.18.* | unaffected |
| Linux | Linux | 7.2.8 <= 7.2.* | unaffected |
| Linux | Linux | 7.3-rc4 <= * | unaffected |
Weaknesses
References
- https://git.kernel.org/stable/c/f973769fb7cb33ba0d4a64b7dcfee0cabaead938
- https://git.kernel.org/stable/c/b72f38929f4b5aa83161fd3b22a7327207e51ec5
- https://git.kernel.org/stable/c/1b11e4b55b41d9e69a8e8d07622614202e2eaca9
- https://git.kernel.org/stable/c/326f7d34bd7e64de535566b65c0533b640df5a81
- https://git.kernel.org/stable/c/5ad925d45f28464a5e043c9620880bb6a030da89
- https://git.kernel.org/stable/c/188b334a6db36a8e0bfaaf59bc4020639a0f3aad
- https://git.kernel.org/stable/c/1ff3add37c329704ec46181b4ae4f00f9f16ce6f
- https://git.kernel.org/stable/c/9a141d3dc869d18b2eab35e999f4790a9b84e40f
Feedback
Was this page helpful?
Glad to hear it! Please tell us how we can improve.
Sorry to hear that. Please tell us how we can improve.