Bug 220053 - iwlwifi: 8265: frequent disconnection - WIFI-655515
Summary: iwlwifi: 8265: frequent disconnection - WIFI-655515
Status: NEEDINFO
Alias: None
Product: Drivers
Classification: Unclassified
Component: network-wireless-intel (show other bugs)
Hardware: Intel Linux
: P3 normal
Assignee: Default virtual assignee for network-wireless-intel
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2025-04-24 16:40 UTC by Karakurt
Modified: 2025-05-11 09:30 UTC (History)
2 users (show)

See Also:
Kernel Version: 6.14.3
Subsystem:
Regression: Yes
Bisected commit-id:


Attachments
journalctl -k output containing the error (213.76 KB, text/plain)
2025-04-24 16:40 UTC, Karakurt
Details
journalctl -k output containing the reconnections (398.48 KB, text/plain)
2025-04-24 16:42 UTC, Karakurt
Details
attachment-6026-0.html (1.09 KB, text/html)
2025-04-27 14:33 UTC, Karakurt
Details

Description Karakurt 2025-04-24 16:40:38 UTC
Created attachment 308014 [details]
journalctl -k output containing the error

I am using 6.14.3-arch1-1 kernel and got stuck in a re-connection loop yesterday, and now this (log file). Did not reproduce it on 6.12.24-1-lts, but that can be due to inconsistency of the issue.
By the looks of things it should have been already resolved on 5-th kernel version (Bug IDs 208425, 207409, 205387, 205035), so I'm creating a new one for 6-th version.

Also, anecdotally it took down my system entierly - got errors on nvme, it was remounted read-only and after that KDE gradually failed (that part is cut off).

Feel free to ask any addtional information.

lspci:
3a:00.0 Network controller: Intel Corporation Wireless 8265 / 8275 (rev 78)


Here is the short version:

[29559.344088] iwlwifi 0000:3a:00.0: Queue 10 is active on fifo 1 and stuck for 10000 ms. SW [226, 239] HW [226, 239] FH TRB=0x0c010a0ee
[29559.345179] iwlwifi 0000:3a:00.0: Microcode SW error detected.  Restarting 0x2000000.
[29559.345320] iwlwifi 0000:3a:00.0: Start IWL Error Log Dump:
[29559.345323] iwlwifi 0000:3a:00.0: Transport status: 0x0000004A, valid: 6
[29559.345326] iwlwifi 0000:3a:00.0: Loaded firmware version: 36.ca7b901d.0 8265-36.ucode
...
[29559.345808] iwlwifi 0000:3a:00.0: Collecting data: trigger 2 fired.
[29559.345814] iwlwifi 0000:3a:00.0: Device error - SW reset
[29559.345818] ieee80211 phy0: Hardware restart was requested
...
[29560.082296] wlp58s0: HW problem - can not stop rx aggregation for 00:eb:d8:d1:8d:7e tid 5
...
[29560.082623] iwlwifi 0000:3a:00.0: restart completed
[29571.632540] iwlwifi 0000:3a:00.0: Queue 10 is active on fifo 1 and stuck for 10000 ms. SW [19, 51] HW [19, 51] FH TRB=0x0c010a022
[29571.633626] iwlwifi 0000:3a:00.0: Microcode SW error detected.  Restarting 0x2000000.
[29571.633754] iwlwifi 0000:3a:00.0: Start IWL Error Log Dump:
[29571.633756] iwlwifi 0000:3a:00.0: Transport status: 0x0000004A, valid: 6
[29571.633758] iwlwifi 0000:3a:00.0: Loaded firmware version: 36.ca7b901d.0 8265-36.ucode
Comment 1 Karakurt 2025-04-24 16:42:22 UTC
Created attachment 308015 [details]
journalctl -k output containing the reconnections

Might be usefull?
Comment 2 Emmanuel Grumbach 2025-04-27 14:12:36 UTC
Did this work better in the past? 

It looks like we keep losing the AP's beacons.
Comment 3 Karakurt 2025-04-27 14:33:56 UTC
Created attachment 308034 [details]
attachment-6026-0.html

I did not have had any issues with WiFi on this laptop before.
Also, a couple of days have past and I did not encounter any problems on LTS kernel (6.12.25-1), so it almost confirms that this is a 6.14 kernel problem

27 квітня 2025 р. 17:12:36 GMT+03:00, bugzilla-daemon@kernel.org пише:
>https://bugzilla.kernel.org/show_bug.cgi?id=220053
>
>Emmanuel Grumbach (emmanuel.grumbach@intel.com) changed:
>
>           What    |Removed                     |Added
>----------------------------------------------------------------------------
>                 CC|                            |emmanuel.grumbach@intel.com
>
>--- Comment #2 from Emmanuel Grumbach (emmanuel.grumbach@intel.com) ---
>Did this work better in the past? 
>
>It looks like we keep losing the AP's beacons.
>
>-- 
>You may reply to this email to add a comment.
>
>You are receiving this mail because:
>You reported the bug.
Comment 4 Emmanuel Grumbach 2025-04-27 15:27:15 UTC
Can you test patches?
Comment 5 Emmanuel Grumbach 2025-04-27 16:04:22 UTC
FWIW: I can see two issues:

1) a TFD queue hang: meaning, we can't transmit (probably because of noise). This looks like this in your log:
кві 24 18:02:43 archlinux kernel: iwlwifi 0000:3a:00.0: Queue 10 is active on fifo 1 and stuck for 10000 ms. SW [226, 239] HW [226, 239] FH TRB=0x0c010a0ee
кві 24 18:02:43 archlinux kernel: iwlwifi 0000:3a:00.0: Microcode SW error detected.  Restarting 0x2000000.
кві 24 18:02:43 archlinux kernel: iwlwifi 0000:3a:00.0: Start IWL Error Log Dump:
кві 24 18:02:43 archlinux kernel: iwlwifi 0000:3a:00.0: Transport status: 0x0000004A, valid: 6
кві 24 18:02:43 archlinux kernel: iwlwifi 0000:3a:00.0: Loaded firmware version: 36.ca7b901d.0 8265-36.ucode

2) we can't hear the beacons of the AP:
кві 23 23:45:32 archlinux kernel: iwlwifi 0000:3a:00.0: missed beacons exceeds threshold, but receiving data. Stay connected, Expect bugs.
кві 23 23:45:32 archlinux kernel: iwlwifi 0000:3a:00.0: missed_beacons:30, missed_beacons_since_rx:1
кві 23 23:45:32 archlinux kernel: iwlwifi 0000:3a:00.0: missed beacons exceeds threshold, but receiving data. Stay connected, Expect bugs.
кві 23 23:45:32 archlinux kernel: iwlwifi 0000:3a:00.0: missed_beacons:31, missed_beacons_since_rx:1
кві 23 23:45:32 archlinux kernel: iwlwifi 0000:3a:00.0: missed beacons exceeds threshold, but receiving data. Stay connected, Expect bugs.
кві 23 23:45:32 archlinux kernel: iwlwifi 0000:3a:00.0: missed_beacons:32, missed_beacons_since_rx:1
кві 23 23:45:32 archlinux kernel: iwlwifi 0000:3a:00.0: missed beacons exceeds threshold, but receiving data. Stay connected, Expect bugs.
кві 23 23:45:32 archlinux kernel: iwlwifi 0000:3a:00.0: missed_beacons:33, missed_beacons_since_rx:2
кві 23 23:45:32 archlinux kernel: iwlwifi 0000:3a:00.0: missed beacons exceeds threshold, but receiving data. Stay connected, Expect bugs.
кві 23 23:45:32 archlinux kernel: iwlwifi 0000:3a:00.0: missed_beacons:34, missed_beacons_since_rx:3
кві 23 23:45:32 archlinux kernel: wlp58s0: Connection to AP 7c:f1:7e:32:56:50 lost
Comment 6 Karakurt 2025-04-28 09:21:34 UTC
(In reply to Emmanuel Grumbach from comment #4)
> Can you test patches?

Which ones?
And also I don't have a reproducible setup. If you know of a way to trigger this problem, or increase the chances of it occurring, I'd love to hear about it.
Comment 7 Emmanuel Grumbach 2025-04-28 09:43:05 UTC
We introduced a mechanism to stay connected to an AP even we no longer hear its beacons as long as we have traffic from that AP. Those explain the prints you see:

кві 23 23:45:32 archlinux kernel: iwlwifi 0000:3a:00.0: missed_beacons:30, missed_beacons_since_rx:1
кві 23 23:45:32 archlinux kernel: iwlwifi 0000:3a:00.0: missed beacons exceeds threshold, but receiving data. Stay connected, Expect bugs.

I'd like to disconnect in this case and see if this improves your experience (by connecting to another AP maybe)
Comment 8 Emmanuel Grumbach 2025-05-11 09:30:15 UTC
Sorry, this fell between the cracks

Can you please try the master branch of:
https://git.kernel.org/pub/scm/linux/kernel/git/iwlwifi/backport-iwlwifi.git/

And then you can try this:

diff --git a/drivers/net/wireless/intel/iwlwifi/mvm/mac-ctxt.c b/drivers/net/wireless/intel/iwlwifi/mvm/mac-ctxt.c
index 44c790b2b387..6c438d48326d 100644
--- a/drivers/net/wireless/intel/iwlwifi/mvm/mac-ctxt.c
+++ b/drivers/net/wireless/intel/iwlwifi/mvm/mac-ctxt.c
@@ -1653,7 +1653,7 @@ iwl_mvm_handle_missed_beacons_notif(struct iwl_mvm *mvm,
         * and/or in case of a CS flow on one of the other AP vifs.
         */
        if (rx_missed_bcon >= IWL_MVM_MISSED_BEACONS_THRESHOLD_LONG) {
-               if (rx_missed_bcon_since_rx >= IWL_MVM_MISSED_BEACONS_SINCE_RX_THOLD) {
+               if (rx_missed_bcon_since_rx >= IWL_MVM_MISSED_BEACONS_SINCE_RX_THOLD || true) {
                        iwl_mvm_connection_loss(mvm, vif, "missed beacons");
                } else {
                        IWL_WARN(mvm,

Note You need to log in before you can comment on or make changes to this bug.