Public bug reported:
System: Lenovo ThinkPad X1 Yoga Gen 8 (21HQ), BIOS N3XET67W 1.42, Ubuntu
26.04.1 LTS, kernel 7.0.0-31-generic. WiFi: Intel Wi-Fi 6E AX211
(Raptor Lake PCH CNVi), PCI 0000:00:14.3, iwlwifi driver, firmware
so-a0-gf-a0-89.ucode (confirmed the newest ever published for this chip
variant — nothing above 89 exists upstream, checked against the current
linux-firmware.git WHENCE).
Two related, reproducible problems, both involving repeated iwlwifi
firmware crashes (Microcode SW error / ADVANCED_SYSASSERT):
1. RUNTIME CRASH LOOP. The adapter's firmware periodically crashes and
self-restarts during ordinary background WiFi scanning (driver-side
WARN_ON at drivers/net/wireless/intel/iwlwifi/mvm/scan.c:1667,
iwl_mvm_scan_umac_fill_ch_p_v7). Worst observed instance: 56 crashes
over ~4.5 hours on an almost exactly fixed 330-second cadence
(matching a periodic scan timer, not workload), machine idle and
unattended overnight. Once, WiFi ended up stuck fully disabled with
no error shown, requiring a manual toggle to recover.
2. S2IDLE SUSPEND HANGS. Separately, `systemctl suspend` intermittently
wedges completely: kernel log ends at "PM: suspend entry (s2idle)"
with no resume, screen black, power LED stuck on, forced power-off
required. Across 40+ manual test cycles, roughly 10-40% of attempts
fail depending on conditions. Most (not all) hangs occur on the
attempt immediately following one of the crashes above.
MITIGATION TESTED: blocking the adapter from D3cold
(d3cold_allowed=0) plus unloading iwlwifi before every suspend via a
logind delay-inhibitor script cut the hang rate substantially (~91%
success in one test run) but did not eliminate it — some hangs still
occur with no crash precursor at all, and the crash loop itself
continues regardless of this mitigation.
CROSS-OS COMPARISON (same hardware, dual-boot Windows 11): 10/10 sleep
cycles clean, full-shutdown battery drain ~0.2%/hour (normal) — versus
this laptop separately being found with a fully drained battery and
unresponsive power button after a normal Linux `poweroff`, on at least
two occasions (battery ~96% to empty over ~9 hours). Windows handles
the identical hardware cleanly across both suspend and shutdown,
pointing at the Linux driver/firmware stack rather than a hardware
defect.
RELATED: bug #1977815 (closed Invalid in 2022 after the reporter's
ucode changed 67->71) — this system runs the much newer ucode 89 and
still reproduces the identical crash signature, suggesting it was
never fully fixed or has regressed.
Full technical write-up (complete timelines, exact log excerpts,
methodology) attached as x1yoga8-iwlwifi-notes.txt.
** Affects: linux-firmware (Ubuntu)
Importance: Undecided
Status: New
** Attachment added: "x1yoga8-iwlwifi-notes.txt"
https://bugs.launchpad.net/bugs/2167666/+attachment/6000969/+files/x1yoga8-iwlwifi-notes.txt
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2167666
Title:
Intel AX211 (0000:00:14.3) iwlwifi firmware crash (ADVANCED_SYSASSERT)
after resume causes s2idle suspend to hang; does not reproduce on
Windows 11 on same hardware
To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux-firmware/+bug/2167666/+subscriptions
--
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
