Crash Logs Repository¶
Raw crash logs and kernel traces collected during MT7925 driver development and testing.
Browse All Crash Logs on GitHub
Log Categories¶
NULL Pointer Crashes¶
These crashes involve NULL pointer dereferences, typically during VIF iteration or reset work.
| Log File | Date | Issue |
|---|---|---|
| crash-2026-01-10-reset-work-null-ptr.log | Jan 10 | NULL pointer in mt76_connac_mcu_uni_add_dev during reset work after MCU timeouts |
| crash-2026-01-10-2112-reset-work-null-vif.log | Jan 10 | NULL VIF during reset work iteration |
| crash-2026-01-10-2123-reset-work-null-vif.log | Jan 10 | Same pattern, different trigger |
| crash-2026-01-10-2131-reset-work-null-vif.log | Jan 10 | Same pattern, multiple occurrences |
Root Cause: During mt7925_mac_reset_work, the driver iterates VIFs via ieee80211_iterate_active_interfaces. If a VIF is removed during iteration, the pointer becomes NULL.
Fix: Patches 0003, 0005, 0017 add NULL checks for VIF and link pointers.
MLO (Multi-Link Operation) Issues¶
Crashes and hangs related to WiFi 7 Multi-Link Operation.
| Log File | Date | Issue |
|---|---|---|
| crash-2026-01-11-1819-mlo-roam-hang.log | Jan 11 | System hang during MLO authentication on 6 GHz. Missing patches caused deadlock. |
| crash-2026-01-19-mlo-authentication-failure.log | Jan 19 | Repeated MLO auth failures with ROC timeouts on WiFi 7 network |
Key Symptoms:
nl80211: kernel reports: link ID must for MLO group key- ROC (Remain On Channel) command timeouts
- Firmware resets during MLO link establishment
Root Cause: MLO key handling and link ID management issues in the upstream driver.
Fix: Patches 0012-0017 address various MLO issues.
MCU Timeout Events¶
MCU (Microcontroller Unit) command timeouts that may or may not lead to crashes.
| Log File | Date | Issue |
|---|---|---|
| crash-2026-01-12-2210-auth-loop-mcu-timeout.log | Jan 12 | Auth loop with MCU timeout, recovered gracefully (patches working) |
| crash-2026-01-13-0030-mcu-timeout-recovery.log | Jan 13 | 5 consecutive MCU timeouts with successful firmware reload recovery |
Key Observation: With patches applied, MCU timeouts trigger firmware reload and recovery instead of kernel panic. This demonstrates the patches are working correctly.
6 GHz Band Issues¶
Issues specifically related to 6 GHz band operation.
| Log File | Date | Issue |
|---|---|---|
| behavior-2026-01-19-6ghz-comeback-roc-timeout.log | Jan 19 | Repeated pattern: 6 GHz disconnect → comeback rejection → ROC timeout → firmware reload |
Pattern Observed:
- Station on 6 GHz disconnects
- Driver attempts to connect to 5 GHz fallback
- AP sends "comeback later" (status 30)
- AP forgets auth state during comeback wait
- ROC commands timeout
- Firmware reload recovers
Status: Under investigation. May indicate firmware instability with certain AP rejection sequences.
Authentication Failures¶
Connection and authentication issues.
| Log File | Date | Issue |
|---|---|---|
| crash-2026-01-12-1959-wifi-auth-failures.log | Jan 12 | Multiple auth failures with status 37 (SAE/WPA3 rejection) |
Unrelated System Issues¶
Some logs capture crashes from other subsystems occurring at the same time.
| Log File | Date | Issue |
|---|---|---|
| crash-2026-01-14-1731-amdgpu-mes-hang.log | Jan 14 | AMDGPU MES (Micro Engine Scheduler) hang - unrelated to WiFi driver |
| crash-2026-01-10-journalctl.log | Jan 10 | General journal output from crash session |
Summary Documentation¶
| File | Description |
|---|---|
| OTHER_CRASHES.md | Detailed analysis of hard lockup and station removal deadlock bugs |
Common Error Messages¶
MCU Timeout¶
The MCU (firmware) didn't respond to a command within the timeout period.MLO Key Error¶
Multi-Link Operation key handling error - requires MLO patches.NULL Pointer Dereference¶
A pointer was NULL when the driver expected a valid object.Hung Task¶
A kernel worker thread is stuck, usually due to mutex deadlock.Contributing Crash Logs¶
If you experience a crash, please open an issue with:
- Kernel version (
uname -r) - DKMS version (
dkms status | grep mt76) - dmesg output around the crash
- What you were doing when it crashed
See Debugging Guide for instructions on collecting crash information.