Return to Level1Techs.com

Iwlwifi bug? kernel 5.1.8

Running Fedora 30 on an Asrock X470 Taichi with the onboard wireless adapter. Since kernel 5.1.x(? not sure when exactly, but definitely with 5.1.8 i am currently running) i am seeing poor performance and driver/crash reset with microcode software error.

The adapter basically goes offline and needs a disable/re-enable to reconnect. Super annoying, have had about 6 months of hassle free wifi until recently…

Wondering if any one else running this hardware on fedora is seeing similar?

Relevant lines from dmesg:

detected hardware:

[ 6.901579] iwlwifi 0000:27:00.0: enabling device (0000 -> 0002)
[ 6.903393] iwlwifi 0000:27:00.0: loaded firmware version 29.1044073957.0 op_mode iwlmvm
[ 6.984407] iwlwifi 0000:27:00.0: Detected Intel® Dual Band Wireless AC 3168, REV=0x220

When it craps out…

[ 450.696226] iwlwifi 0000:27:00.0: Microcode SW error detected. Restarting 0x2000000.
[ 450.696394] iwlwifi 0000:27:00.0: Start IWL Error Log Dump:
[ 450.696396] iwlwifi 0000:27:00.0: Status: 0x00000100, count: 6
[ 450.696397] iwlwifi 0000:27:00.0: Loaded firmware version: 29.1044073957.0
[ 450.696399] iwlwifi 0000:27:00.0: 0x00000EDC | ADVANCED_SYSASSERT
[ 450.696400] iwlwifi 0000:27:00.0: 0x00000220 | trm_hw_status0
[ 450.696401] iwlwifi 0000:27:00.0: 0x00000000 | trm_hw_status1
[ 450.696402] iwlwifi 0000:27:00.0: 0x00043D54 | branchlink2
[ 450.696403] iwlwifi 0000:27:00.0: 0x0004AFA6 | interruptlink1
[ 450.696404] iwlwifi 0000:27:00.0: 0x00000000 | interruptlink2
[ 450.696405] iwlwifi 0000:27:00.0: 0x0ACA001C | data1
[ 450.696406] iwlwifi 0000:27:00.0: 0x2000005E | data2
[ 450.696407] iwlwifi 0000:27:00.0: 0x00001C84 | data3
[ 450.696408] iwlwifi 0000:27:00.0: 0x2DC0CDC5 | beacon time
[ 450.696409] iwlwifi 0000:27:00.0: 0x923B4245 | tsf low
[ 450.696410] iwlwifi 0000:27:00.0: 0x00000224 | tsf hi
[ 450.696411] iwlwifi 0000:27:00.0: 0x00000000 | time gp1
[ 450.696411] iwlwifi 0000:27:00.0: 0x1A633483 | time gp2
[ 450.696412] iwlwifi 0000:27:00.0: 0x00000001 | uCode revision type
[ 450.696413] iwlwifi 0000:27:00.0: 0x0000001D | uCode version major
[ 450.696414] iwlwifi 0000:27:00.0: 0x3E3B4DE5 | uCode version minor
[ 450.696415] iwlwifi 0000:27:00.0: 0x00000220 | hw version
[ 450.696416] iwlwifi 0000:27:00.0: 0x00C89200 | board version
[ 450.696417] iwlwifi 0000:27:00.0: 0x2000005E | hcmd
[ 450.696418] iwlwifi 0000:27:00.0: 0x24022002 | isr0
[ 450.696418] iwlwifi 0000:27:00.0: 0x00800000 | isr1
[ 450.696419] iwlwifi 0000:27:00.0: 0x00000002 | isr2
[ 450.696420] iwlwifi 0000:27:00.0: 0x00417CC0 | isr3
[ 450.696421] iwlwifi 0000:27:00.0: 0x00000000 | isr4
[ 450.696422] iwlwifi 0000:27:00.0: 0x0A99001C | last cmd Id
[ 450.696423] iwlwifi 0000:27:00.0: 0x00000000 | wait_event
[ 450.696424] iwlwifi 0000:27:00.0: 0x00000094 | l2p_control
[ 450.696424] iwlwifi 0000:27:00.0: 0x00018030 | l2p_duration
[ 450.696425] iwlwifi 0000:27:00.0: 0x0000000F | l2p_mhvalid
[ 450.696426] iwlwifi 0000:27:00.0: 0x00000085 | l2p_addr_match
[ 450.696427] iwlwifi 0000:27:00.0: 0x00000005 | lmpm_pmg_sel
[ 450.696428] iwlwifi 0000:27:00.0: 0x03041751 | timestamp
[ 450.696435] iwlwifi 0000:27:00.0: 0x00000010 | flow_handler
[ 450.696437] ieee80211 phy0: Hardware restart was requested
[ 685.039518] rfkill: input handler disabled
[ 819.223747] wlp39s0: deauthenticating from 40:9b:cd:a6:b9:30 by local choice (Reason: 3=DEAUTH_LEAVING)
[ 842.491706] wlp39s0: authenticate with 40:9b:cd:a6:b9:30
[ 842.495331] wlp39s0: send auth to 40:9b:cd:a6:b9:30 (try 1/3)
[ 842.603849] wlp39s0: send auth to 40:9b:cd:a6:b9:30 (try 2/3)
[ 842.604432] wlp39s0: authenticated
[ 842.604849] wlp39s0: associate with 40:9b:cd:a6:b9:30 (try 1/3)
[ 842.606310] wlp39s0: RX AssocResp from 40:9b:cd:a6:b9:30 (capab=0x11 status=0 aid=1)
[ 842.607496] wlp39s0: associated
[ 842.629480] IPv6: ADDRCONF(NETDEV_CHANGE): wlp39s0: link becomes ready

ethtool output

driver: iwlwifi
version: 5.1.8-300.fc30.x86_64
firmware-version: 29.1044073957.0
expansion-rom-version:
bus-info: 0000:27:00.0
supports-statistics: yes
supports-test: no
supports-eeprom-access: no
supports-register-dump: no
supports-priv-flags: no

@wendell do you use Intel wifi on recent kernels perchance? Guessing mostly wired… i know you’re an Asrock user and maybe even have a Taichi board?

kernel:

Linux ryzen2700x 5.1.8-300.fc30.x86_64 #1 SMP Sun Jun 9 17:09:32 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux

edit:
not sure if i should submit as a kernel bug, from here:
https://wireless.wiki.kernel.org/en/users/drivers/iwlwifi/debugging

… I’m not sure if firmware is supposed to be reported to the kernel guys as a bug any more?

Its actually a confirmed bug not kernel specific. From what I saw on both Ubuntu and fedora forums the solution though temporary was to upgraded to the latest git version of the firmware.

Its confirmed on only the lower tier cards my 8260 and the 8265 9260 etc are completely unaffectedw

After a small google search of

Microcode SW error detected. Restarting 0x2000000

I found the info its been confirmed since September of last year

1 Like

weird how it was working just fine until around the time kernel 5.1 hit. i think it was fine with earlier releases of 5.1 prior to 5.1.5

i did also find those google links however the firmware listed for download has a modification date older than what i have.

Could be the removed something that made it work…

Whats the date on your firmware

This is the latest build
iwlwifi-fw-2019-05-03 | linux-firmware-iwlwifi-fw-2019-05-03.tar.gz

If this isnt working right… shoot for a iwlwifi-fw-next build thats current circa june but its BETA so I caution you

You can find it here… It works on my 8260

iwlwifi-sent-for-review-2019-06-14 | iwlwifi-fixes-iwlwifi-sent-for-review-2019-06-14.tar.gz

Keep in mind Emmanuel is the Maintainer of Linux wifi driver along with intel. He does good work but he can make mistakes

1 Like

Thanks will check when i get home.

1 Like

Looks like the firmware may have been upgraded using the Fedora package:

  • linux-firmware.noarch-20190312-94.fc30

So i have attempted a rollback with

  • dnf downgrade linux-firmware

… which has rolled me back to : linux-firmware-20190312-94.fc30

Will see how it goes.

Definitely mid may would have been when things started flaking out.

Just thought i’d update this:

confirmed as a firmware/kernel bug (as i understand both needed tweaks), and fixed

https://bugzilla.kernel.org/show_bug.cgi?id=204151

… in case anyone is having similar issues, they should be fixed upstream real soon now.

:slight_smile:

1 Like