Thanks a lot for all the great work Zac and no problems. Your DKMS module still works like a charm even on 7.0.0 kernel line :-)
Status update for the 7.0 series (resolute `linux`, noble `linux- hwe-7.0`) checked on 7.0.0-31.31~24.04.1, Framework Laptop 16, MT7925. 1. The deadlock half of this bug is fixed in Ubuntu 7.0. "wifi: mt76: mt7925: fix potential deadlock in mt7925_roc_abort_sync") has been in since 7.0.0-28.28 — it is listed in the changelog under that upload and came in with the v7.0.x stable updates. 2. The NULL dereference this bug was opened for is not fixed in 7.0.0-31. It is fixed upstream by bd8b2ec83818 "wifi: mt76: mt7925: fix crash in reset link replay" by (Sean Wang, in v7.2) 3. Two more upstream crash fixes in the same area are missing from 7.0.0-31: 341639421954 "wifi: mt76: fix list corruption in mt76_wcid_cleanup" (v7.1) Zac's patch 02/12 from v6 of his series, merged upstream. sta_poll_list corruption after a chip reset. Not MLO-specific. 0fff5b5e2786 "wifi: mt76: mt7925: drop WCID reinit after publish" (v7.1) Same defect as Zac's "fix double wcid initialization race condition". 4. Still not upstream as of 7.3-rc3, but potentially resolved: None of the 7 patches in Zac's v7 series (2026-01-29) merged verbatim. As of 7.3-rc3, mainline still has no NULL checks in mt792x_tx() for mlink/conf/link_sta, no mutex around ieee80211_iterate_active_interfaces() in mt7925_mac_reset_work() / mt7925_set_runtime_pm(), and mt7925_mac_link_sta_assoc() / _add() still dereference the results of the link lookups unchecked. Upstream instead changed link lifetime and publication ordering in v7.1 (Sean Wang's series: 6ace866cf6d2, db134691924f, beec58f36983, 75e2d6bfd9ac, 9e4d518a4707, ...). I don't quite know whether that makes the NULL cases Zac reported unreachable or merely less likely — that would need review by someone who knows the mt76 link lifecycle, or testing on MLO hardware. The three commits above (bd8b2ec83818, 341639421954, 0fff5b5e2786) will not reach the 7.0 kernel through stable updates. Upstream linux-7.0.y is EOL (kernel.org no longer lists it) and 7.0.0-31 already includes its final release, v7.0.14. Only bd8b2ec83818 carries a Fixes: tag. All three are upstream now, which I believe satisfies the request in #12, so could they be considered for SRU to resolute / noble linux-hwe-7.0? -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/2137291 Title: mt7925e: Kernel panic NULL pointer dereference in mt76_connac_mcu_uni_add_dev during WiFi reset - Framework Desktop To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/2137291/+subscriptions -- ubuntu-bugs mailing list [email protected] https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
