video card stops working *ERROR* clock recovery reached max voltage

fobbyal · 2017-03-19 03:08:56

recently upgraded to 4.9.14-1-lts and my monitors started to shutoff randomly. And after they shutoff I am unable to switch them on with keyboard or mouse input. I thought it was was something related to DPMS and turned screen saver + dpms off. It was still happening after that. so investiagted with journalctl and found the following stack trace. I am a fairly new linux user and have not really done any system programming. Not really sure how to build and debug/bisect the kernel to find out what is wrong so I am asking for a little help to point me to the right direction. For the time being, I have reverted back to 4.9.13-1-lts which does not have this problem AFAIK. Can someone tell me to what I should do next? Thx in advance.

------------[ cut here ]------------
Mar 18 10:05:00 showdownarch kernel: WARNING: CPU: 3 PID: 13100 at ./include/drm/drm_crtc.h:1403 drm_helper_choose_encoder_dpms+0x8a/0x90 [drm_kms_helper]
Mar 18 10:05:00 showdownarch kernel: Modules linked in: xt_multiport iptable_filter overlay snd_hda_codec_realtek snd_hda_codec_generic snd_hda_codec_hdmi iTCO_wdt gpio_ich iTCO_ven
Mar 18 10:05:00 showdownarch kernel:  mii fb_sys_fops i2c_algo_bit button snd_timer snd mei_me soundcore mei intel_agp shpchp intel_gtt acpi_cpufreq tpm_tis tpm_tis_core tpm sch_fq_
Mar 18 10:05:00 showdownarch kernel: CPU: 3 PID: 13100 Comm: kworker/3:256 Tainted: G        W       4.9.14-1-lts #1
Mar 18 10:05:00 showdownarch kernel: Hardware name: Gigabyte Technology Co., Ltd. H55-USB3/H55-USB3, BIOS F7 08/20/2010
Mar 18 10:05:00 showdownarch kernel: Workqueue: events radeon_dp_work_func [radeon]
Mar 18 10:05:00 showdownarch kernel:  ffffc90010d63d20 ffffffff812f890d 0000000000000000 0000000000000000
Mar 18 10:05:00 showdownarch kernel:  ffffc90010d63d60 ffffffff8107cb0b 0000057b9e861d09 ffff88030a843000
Mar 18 10:05:00 showdownarch kernel:  ffff88030ffe7a00 ffff88030f5f9000 0000000000000003 ffff88030aa92660
Mar 18 10:05:00 showdownarch kernel: Call Trace:
Mar 18 10:05:00 showdownarch kernel:  [<ffffffff812f890d>] dump_stack+0x63/0x86
Mar 18 10:05:00 showdownarch kernel:  [<ffffffff8107cb0b>] __warn+0xcb/0xf0
Mar 18 10:05:00 showdownarch kernel:  [<ffffffff8107cc3d>] warn_slowpath_null+0x1d/0x20
Mar 18 10:05:00 showdownarch kernel:  [<ffffffffa044408a>] drm_helper_choose_encoder_dpms+0x8a/0x90 [drm_kms_helper]
Mar 18 10:05:00 showdownarch kernel:  [<ffffffffa04444eb>] drm_helper_connector_dpms+0x4b/0x100 [drm_kms_helper]
Mar 18 10:05:00 showdownarch kernel:  [<ffffffffa04465cb>] ? drm_dp_dpcd_read_link_status+0x1b/0x20 [drm_kms_helper]
Mar 18 10:05:00 showdownarch kernel:  [<ffffffffa05ab477>] radeon_connector_hotplug+0xf7/0x100 [radeon]
Mar 18 10:05:00 showdownarch kernel:  [<ffffffffa05b91ff>] radeon_dp_work_func+0x3f/0x60 [radeon]
Mar 18 10:05:00 showdownarch kernel:  [<ffffffff81096399>] process_one_work+0x1e9/0x440
Mar 18 10:05:00 showdownarch kernel:  [<ffffffff8109663b>] worker_thread+0x4b/0x4f0
Mar 18 10:05:00 showdownarch kernel:  [<ffffffff810965f0>] ? process_one_work+0x440/0x440
Mar 18 10:05:00 showdownarch kernel:  [<ffffffff8109c0c9>] kthread+0xd9/0xf0
Mar 18 10:05:00 showdownarch kernel:  [<ffffffff8102c74e>] ? __switch_to+0x2ce/0x5b0
Mar 18 10:05:00 showdownarch kernel:  [<ffffffff8109bff0>] ? kthread_park+0x60/0x60
Mar 18 10:05:00 showdownarch kernel:  [<ffffffff815f8a55>] ret_from_fork+0x25/0x30
Mar 18 10:05:00 showdownarch kernel: ---[ end trace 6a290eae4410f871 ]---
Mar 18 10:05:00 showdownarch kernel: [drm:radeon_dp_link_train [radeon]] *ERROR* clock recovery reached max voltage
Mar 18 10:05:00 showdownarch kernel: [drm:radeon_dp_link_train [radeon]] *ERROR* clock recovery failed

R00KIE · 2017-03-19 14:54:10

This is something you want to report upstream. I'd say the way to go here is bisect the kernel and find the commit that breaks things.

Since this is most probably a problem with the radeon driver the place to report it is here https://bugs.freedesktop.org under DRI -> DRM/Radeon.

fobbyal · 2017-03-19 16:22:11

Thx for advice. I will file a bug at freedesktop.org. On the debug side. Is there a guide for getting the git repo of linux kernel + building + testing ? I don't even know how to trigger this on demand. It just randomly happens on when the computer is on. Any suggestion?

loqs · 2017-03-19 17:09:31

https://cdn.kernel.org/pub/linux/kernel … Log-4.9.14 contains no radeon/drm changes so you might want to run for a while longer before deciding the issue was introduced between that release and 4.9.14.

https://wiki.archlinux.org/index.php/Bisecting_bugs

Follow the links from there through AUR, makepkg and PKGBUILD if you have never done so before. You will need to adjust the linux-git PKGBUILD to point to the linux-stable repo and it is probably a good idea to change the config files in the PKGBUILD to those from 4.9.14-1-lts.

fobbyal · 2017-03-22 03:41:13

okay reporting back...
So after running 4.9.13-1-lts for a while longer (30 some hours) .. the issue did pop back again. From the journalctl --list-boots i found a period that i didn't restart for a few days and that was on linux-lts-4.4.52-1. Rolled back to that, was okay for 24 hours and it started happening again (no stack trace though) but this time it only happens when i try to wake the Monitors the dpms shut them off. So I think it's probably in the driver.

530567 -rw-r--r-- 1 root root  141084 Feb  3 06:42 xf86-video-ati-1:7.8.99.r24.g80fc7863-1-x86_64.pkg.tar.xz
525988 -rw-r--r-- 1 root root  142784 Mar  1 15:37 xf86-video-ati-1:7.8.99.r38.g9035b6ab-1-x86_64.pkg.tar.xz
529187 -rw-r--r-- 1 root root  142760 Mar 16 16:41 xf86-video-ati-1:7.9.0-1-x86_64.pkg.tar.xz

The xf86-video-ati-1:7.9.0-1-x86_64.pkg.tar.xz propably did the damage. Fortunately, the AMDGPU driver has experimental support for the Video Card that i am using Radeon HD 7970. I've switch to that just now. Let's see what happens.

Going to file a bug at https://bugs.freedesktop.org

Thank you all for your help.

Arch Linux

#1 2017-03-19 03:08:56