FS#56436 - [linux] wifi issues with 4.14.2-1

Attached to Project: Arch Linux
Opened by Tronje Krabbe (tronje) - Friday, 24 November 2017, 20:56 GMT
Last edited by Jan Alexander Steffens (heftig) - Saturday, 02 December 2017, 11:31 GMT
Task Type Bug Report
Category Kernel
Status Closed
Assigned To Tobias Powalowski (tpowa)
Jan Alexander Steffens (heftig)
Architecture All
Severity High
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 4
Private No

Details

Description:
Newest [testing] kernel seems to have issues with (my) wifi hardware? I'm not exactly well-versed with this kind of thing.
Here's the gist: dmesg is full of entries about "iwlwifi" (will attach), and I don't have wifi. Wired internet connection works, and booting into 4.9.64-1 (current [testing] linux-lts package) also works fine.
In addition to the wifi problem, I'm noticing some general slow-ness of the system that I don't usually notice.


Additional info:
* Hardware: Lenovo Thinkpad T460s, with Intel i5
* Intel Dual Band Wireless-AC 8260
* I've attached all of dmesg; the iwlwifi statements keep repeating throughout, they're very hard to miss


Steps to reproduce:
* boot the newset [testing] kernel, 4.14.2-1, I guess
   dmesg (324.5 KiB)
This task depends upon

Closed by  Jan Alexander Steffens (heftig)
Saturday, 02 December 2017, 11:31 GMT
Reason for closing:  Fixed
Additional comments about closing:  4.14.3
Comment by Tronje Krabbe (tronje) - Friday, 24 November 2017, 21:13 GMT
Found a maybe relevant-looking snippet in dmesg:

Nov 24 22:09:42 cyril kernel: WARNING: CPU: 2 PID: 50 at drivers/gpu/drm/i915/intel_display.c:12176 intel_atomic_com
Nov 24 22:09:42 cyril kernel: Modules linked in: snd_hda_codec_hdmi uvcvideo btusb videobuf2_vmalloc btrtl videobuf2
Nov 24 22:09:42 cyril kernel: snd_pcm rtsx_pci_ms psmouse mac_hid ptp snd_timer nvram i2c_i801 tpm_tis pps_core mem
Nov 24 22:09:42 cyril kernel: CPU: 2 PID: 50 Comm: kworker/u8:1 Tainted: G W O 4.14.2-1-ARCH #1
Nov 24 22:09:42 cyril kernel: Hardware name: LENOVO 20FAS05Q00/20FAS05Q00, BIOS N1CET37W (1.05 ) 01/15/2016
Nov 24 22:09:42 cyril kernel: Workqueue: events_unbound intel_atomic_commit_work [i915]
Nov 24 22:09:42 cyril kernel: task: ffff89dd98bf9e80 task.stack: ffff97a0c0e44000
Nov 24 22:09:42 cyril kernel: RIP: 0010:intel_atomic_commit_tail+0xe3c/0xe50 [i915]
Nov 24 22:09:42 cyril kernel: RSP: 0018:ffff97a0c0e47de0 EFLAGS: 00010282
Nov 24 22:09:42 cyril kernel: RAX: 000000000000001c RBX: 0000000000000000 RCX: 0000000000000000
Nov 24 22:09:42 cyril kernel: RDX: 0000000000000000 RSI: ffff89dda250dc98 RDI: ffff89dda250dc98
Nov 24 22:09:42 cyril kernel: RBP: 000000000001279b R08: 0000000000000001 R09: 0000000000006199
Nov 24 22:09:42 cyril kernel: R10: ffff97a0c0e47de0 R11: 0000000000000000 R12: ffff89dd98ad0000
Nov 24 22:09:42 cyril kernel: R13: ffff89dd8e560800 R14: 0000000000000001 R15: 0000000000000000
Nov 24 22:09:42 cyril kernel: FS: 0000000000000000(0000) GS:ffff89dda2500000(0000) knlGS:0000000000000000
Nov 24 22:09:42 cyril kernel: CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Nov 24 22:09:42 cyril kernel: CR2: 00007fe0c203d9b8 CR3: 0000000044c09006 CR4: 00000000003606e0
Nov 24 22:09:42 cyril kernel: Call Trace:
Nov 24 22:09:42 cyril kernel: ? pick_next_task_fair+0x2d2/0x560
Nov 24 22:09:42 cyril kernel: ? wait_woken+0x80/0x80
Nov 24 22:09:42 cyril kernel: process_one_work+0x1db/0x410
Nov 24 22:09:42 cyril kernel: worker_thread+0x2b/0x3d0
Nov 24 22:09:42 cyril kernel: ? process_one_work+0x410/0x410
Nov 24 22:09:42 cyril kernel: kthread+0x118/0x130
Nov 24 22:09:42 cyril kernel: ? kthread_create_on_node+0x70/0x70
Nov 24 22:09:42 cyril kernel: ret_from_fork+0x25/0x30
Nov 24 22:09:42 cyril kernel: Code: 24 88 01 00 00 48 8d 74 24 30 48 83 c7 08 e8 8c 96 7a ca 4d 85 ff 0f 85 df fa ff
Nov 24 22:09:42 cyril kernel: ---[ end trace a949da795ec633eb ]---


Also I hope it's ok that I set the severity to 'High'. I guess maybe not having wifi isn't the biggest issue ever.
Comment by Tronje Krabbe (tronje) - Friday, 24 November 2017, 21:38 GMT
Looks like this might be related? https://www.spinics.net/lists/linux-wireless/msg167767.html
Reply says a fix is in the works for the next stable release of 4.14
Comment by loqs (loqs) - Friday, 24 November 2017, 23:22 GMT
I think your system is affected by at least two different issues.
Could you please try applying the referenced patch https://bugzilla.kernel.org/attachment.cgi?id=260597&action=diff&collapsed=&headers=1&format=raw to confirm if it does fix the iwlwifi issue.
See https://wiki.archlinux.org/index.php/Patching_in_ABS#Applying_patches for details to alter the PKGBUILD.
Comment by Jan Alexander Steffens (heftig) - Saturday, 25 November 2017, 07:39 GMT
Delete /usr/lib/firmware/iwlwifi-8000C-34.ucode and reboot; this should work around the problem.
Comment by Tronje Krabbe (tronje) - Saturday, 25 November 2017, 09:09 GMT
Hey guys, thanks for the replies! I'll try the patch later today. I'll skip the work-around for now because I can simply boot the LTS kernel in a pinch; unless, of course, you would like to know what happens when I delete /usr/lib/firmware/iwlwifi-8000C-34.ucode.
Comment by Laurențiu Nicola (lnicola) - Saturday, 25 November 2017, 11:09 GMT
I can confirm that on 4.14.2-1-ARCH I'm affected by https://bugzilla.kernel.org/show_bug.cgi?id=197591 and removing the firmware file works. I haven't tested the patch from Bugzilla.
Comment by Tronje Krabbe (tronje) - Saturday, 25 November 2017, 12:46 GMT
I can confirm that this[0] patch works! I'm writing this comment with the patched 4.14.2-1-ARCH Kernel via wifi!

[0] https://bugzilla.kernel.org/attachment.cgi?id=260597&action=diff&collapsed=&headers=1&format=raw
Comment by loqs (loqs) - Saturday, 25 November 2017, 21:19 GMT
Is the "pipe A vblank wait timed out" warning from drivers/gpu/drm/i915/intel_display.c:12176 intel_atomic_commit_tail+0xe3c/0xe50 [i915] still present?
Comment by Laurențiu Nicola (lnicola) - Saturday, 25 November 2017, 21:22 GMT
> Is the "pipe A vblank wait timed out" warning from drivers/gpu/drm/i915/intel_display.c:12176 intel_atomic_commit_tail+0xe3c/0xe50 [i915] still present?

I don't think I've seen that one.
Comment by Tronje Krabbe (tronje) - Saturday, 25 November 2017, 21:24 GMT
> Is the "pipe A vblank wait timed out" warning from drivers/gpu/drm/i915/intel_display.c:12176 intel_atomic_commit_tail+0xe3c/0xe50 [i915] still present?

No, no longer present.
Comment by Sandeep (sandy8925) - Sunday, 26 November 2017, 16:49 GMT
I was wondering what the plan was for fixing this? IMO this is an Arch packaging issue, since the latest version of the linux-firmware package contains newer firmware that needs new kernel code not available in any stable Linux kernel release (including 4.14.2).
Comment by loqs (loqs) - Tuesday, 28 November 2017, 17:05 GMT
Fix is queued for 4.14.3 https://git.kernel.org/pub/scm/linux/kernel/git/stable/stable-queue.git/tree/queue-4.14/iwlwifi-mvm-support-version-7-of-the-scan_req_umac-fw-command.patch?id=f6e1b0c9e11ef58600cd0553461b71e2aae043ba
Comment by Sandeep (sandy8925) - Wednesday, 29 November 2017, 04:28 GMT
Ah cool - as long as 4.14.3 is released to core repo before linux-firmware, or some kind of version dependency is established, then it won't end up breaking WiFi for people like me.
Comment by loqs (loqs) - Saturday, 02 December 2017, 09:27 GMT
Can anyone affected confirm if the issue is resolved with linux 4.14.3-1?
Comment by Laurențiu Nicola (lnicola) - Saturday, 02 December 2017, 09:36 GMT
Yes, it's fixed for me.
Comment by Sandeep (sandy8925) - Saturday, 02 December 2017, 09:36 GMT
I will test soon and report back.
Comment by Tronje Krabbe (tronje) - Saturday, 02 December 2017, 10:13 GMT
Fixed for me also.
Comment by Matthias Lisin (matthias.lisin) - Saturday, 02 December 2017, 11:05 GMT
Fixes for me too.

Loading...