2950x random gui crashes and reboot on Fedora and Manjaro

gros319 · February 14, 2019, 11:44am

Hi guys need some help here. My new rig’s crashing left right and center!

So here are the parts :

cpu 2950x
MB MSI MEG CREATION
RAM GSkill Ripjaws V 3200 32Gb 4*8gb
GPU RX 590
PSU Corsair rm850w
SSD 970 evo

Here are the issues

Either in fedora gnome or manjaro KDE the computer randomly crashes / freezes or reboot and all error messages are around Xorg or Wayland crashes, today the entire manjaro gui broke leaving booting in cli directly… I haven’t been able to find a stable set up for 5 days now.

PS I just did a fresh install of manjaro kde and I’m already experiencing troubles as update won’t go through sending timeout messages.

Every thought will be appreciated.

Methylzero · February 14, 2019, 2:49pm

Please see my PSAs regarding issues I have encountered on AM4 Ryzen. I do not have any experience with TR, but maybe they are useful.

PSA: testing for and solving random freezes on Ryzen CPU

So I have ran into some random freezing issues, when setting up a Ryzen 1700X system for server-like (headless, unattended, ECC RAM, no OC) 24/7/365 operation. Not good at all. I have written a small stress testing utility that can reproduce this issue quite reliably in less than 48 hours. It consists a copy of the Intel Linpack benchmark utility, that has been modified to run on AMD CPUs (just hexedited the binary and replaced some strings) and a small bash script that consists of a loop alternating between idling, single core load and 8 core load. You can download it HERE This will stress create a very large number of transitions between idle, single core boost and all core clocks and voltages, thus stressing the reliability of power management/power saving features on the CPU and the motherboard. In my case (Asrock x370 Taichi) the test reproducibly induced a complete system lockup in less then 48h. If you are also affected by this issue, the solution that worked for me is a UEFI setting: Advanced > AMD CBS > Zen Common Options > Power Supply Idle Control This should be set to Typical Current Idle. The default/auto setting seems to be Low Current Idle, that saves an additional couple of watts when completely idle, but also seems to cause some sort of power glitching, that results in occasional freezes. From what I can tell, this setting disables Package C6 state, while leaving all other power saving features such as Core C6 state enabled (you may have to update your UEFI for this setting to appear, for the x370 Taichi I would recommend UEFI version 4.40, the later versions may or may not have regressions)

PSA: The ASM1061 SATA controller on the x370 Taichi is probably garbage Motherboards

On the X370 Taichi you have 10 SATA ports, 8 is presumably provided by the X470 chipset, and the remaining 2 by an ASMedia ASM1061 SATA controller. The 8 native ports seem to work fine, but the ASM1061 might be a complete heap of garbage. Still investigating, but the controller seems to randomly lock up or glitch out when under sustained heavy load, resulting in a dropout of all SATA devices connected to it. Does not seem to cause kernel panics, but this essentially make those 2 SATA ports unusable. I will update this thread when I can confirm that the problem was indeed caused by the ASM1061 controller, but at this point I have strong suspicion. PS: fogot to add, but once the ASM1061 locks up, it seems to remain FUBAR until system restart

PSA: ASPM causes PCI-E bus errors under heavy SATA load on AM4 Motherboards

If you have an AM4 system and you are seeing errors in your kernel logs like this after an extended period of heavy SATA load : kernel: [323826.023666] pcieport 0000:00:03.2: AER: Multiple Corrected error received: id=0000 kernel: [323826.023680] pcieport 0000:00:03.2: PCIe Bus Error: severity=Corrected, type=Data Link Layer, id=001a(Transmitter ID) kernel: [323826.023687] pcieport 0000:00:03.2: device [1022:1453] error status/mask=00001100/00006000 kernel: [323826.023690] pcieport 0000:00:03.2: [ 8] RELAY_NUM Rollover kernel: [323826.023692] pcieport 0000:00:03.2: [12] Replay Timer Timeout Then you are being affected by some sort of PCI-E bug. I have seen this on two systems, both equipped with a 1700X, an X370 Taichi, 32GB of ECC RAM and 2 WD Black drives in RAID 0 (mdraid), running up to date Lubuntu 18.04 LTS. Others have seen similar errors on Threadripper and APU systems. In my case, the errors seem to have been corrected and no adverse effects have been observed. The cause of these issues seems to be ASPM, a power saving feature that can reduce power consumption a little bit, by messing with the PCI-E links. AFAIK, this is a part of the PCI-E 3.0 standard, so it should work, but in practice it often does not, due to buggy implementations. The problem manifests when you have heavy SATA activity going on for extended periods of time. I believe it is caused by the PCI-E link between the CPU and the chipset having a flaky ASPM implementation. Disabling ASPM using the pcie_aspm=off kernel boot option appears to solve this issue completely.

gros319 · February 14, 2019, 7:48pm

Thanks for your answers but I checked all my UEFI parameters and the kernel log and it appears that my crashes are not related to yours. I really thinks they are related to a graphical problem because it’s always sending messages about xorg crashing or X11 ou sddm having disappear… I got a black screen a lauch after login 10mn ago and right now the screen is frozen again in the middle of a system update and there is no particular error code on the motherboard just 00.

Methylzero · February 14, 2019, 9:19pm

Do you have a spare GPU you could try?

wendell · February 14, 2019, 9:26pm

Can you disable pcie aspm? And what bios version?

gros319 · February 14, 2019, 11:53pm

I have the latest version from the msi support page

Version 7B92v12 Release Date 2018-11-15 File Size8.66 MB

gonna try disabling pcie aspm when I got home from work thanks for the tip

gros319 · February 14, 2019, 11:54pm

got a rx 570 same outcome!

gros319 · February 15, 2019, 4:39am

new developement I came back home 15mn age the computer won’t even turn on now the power switch on the motherboard lights red normally but nothing happens when I press it or the front I/O power button either.

gros319 · February 15, 2019, 5:23am

I took the build apart and put it on a test bench with just the bare minimum still got light on the power button but nothing when I press it. I tested the psu on a other computer it worked do you think the motherboard’s dead?

Methylzero · February 15, 2019, 7:28am

Are you sure the CPU is seated properly? From what I heard the TR4 socket is a bitch, especially the Foxconn sockects have a reputation of being a complete pain in the ass to properly close.

gros319 · February 15, 2019, 8:07am

this could cause the motherboard to not turn on? this suddently?

gros319 · February 16, 2019, 3:24am

I reseated the cpu and the computer turned back on but I keep getting freezes on my new manjaro install

gros319 · February 16, 2019, 3:30am

I cannot find pcie aspm in my bios

gros319 · February 16, 2019, 4:36am

so now the case change a little bit no more freezing but more random rebooting I tried changing the kernel but no effect. Maybe the cpu is still not seated really tight in the socket?

wendell · February 16, 2019, 5:21am

Yep it’s not in there. I got mine out now too.

Can add pcie_aspm=off to the kernel line tho.

Make sure both atx12v 8 pin power is installed well on both ends? Clipped in good? Un and replug?

gros319 · February 16, 2019, 5:34am

Yep everything’s pluged in properly. I’m starting to loose my mind a little bit here. I keep getting reboot and frozen screens… I logged in for 45mn about hours ago tried to set up the distro a bit but I couldn’t find any new package I installed even though pamac was saying everything’s right. So I tried to go back to fedora burnt a newly downloaded image on to a flash drive but couldn’t even get to the live os and only got a blinking cursor at the top of my screen. Maybe the cpu is still out of position I’m waiting for new thermale paste to arrive and redo the install completly.
Just another dumb question how do I add the line to the kernel? lol
Again Many many thanks for your help

wendell · February 16, 2019, 2:21pm

For giggles enable pbo and set CPU to like 200w or 300w? This changes a bunch of stuff in the power delivery algorithm

Pcie_aspm: It’s in your grub file permanently or at boot on the kernel selection screen just arrow down to a line, hit e to edit and add after the end which probably has splash/quiet on the kernel line

MisteryAngel · February 16, 2019, 3:04pm

Already tried to re-seat the cpu?

MisteryAngel · February 16, 2019, 3:06pm

Also welcome to Manjaro really.
Maybe you could try to use the terminal for updating instead of pamac / octophi.

gros319 · February 16, 2019, 8:20pm

I re-seated the cpu allready and updated through terminal. The thing is I don’t really know which crashes are hardware related and which are distro related.