Two Complete System Freezes Within One Hour (Leap 15.2)

**dickhack** · 18-Dec-2020, 17:37

Well, this install has officially become my worst ever with openSUSE.

Back on December 6, I tried to upgrade from 15.1 to 15.2. I botched that and had to do a clean install.

Then I had the problem reported over in Multimedia about SMPlayer, VLC and other media players losing audio/video sync when I did skipping. I decided to live with that.

Then I lost sound completely for no known reason as a result of a sudden system freeze, which forced me to do another install when I couldn't recover the sound as I described over in Multimedia.

In the first reinstall, I kept getting desktop freezes, several per day. The Compositor was set to OpenGL 2.0. So I changed it to OpenGL 3.1 and didn't get any freezes for several days until I had to do the reinstall to get my sound back. The other day I had a freeze so I decided to set the Compositor to XRender to see if that made any difference.

Last night I was messing around with a problem I had where "Empty Trash" couldn't delete some files on one of my other partitions. So I went in as root and started removing them. Then I had the same problem on the root partition. In the process I accidentally deleted everything under ~/local/share. Well, I have a nightly home backup, so I copied everything back from the backup and the system didn't seem to have any problems. So I don't know if that is relevant.

Today I've had two freezes - both while inside Firefox 78.5.0esr - within an hour. Plus when I rebooted, the desktop wallpaper was missing (that may or may not be relevant, I guess, I saw that earlier, there may be a setting for that I missed.) I am suspecting that there is an issue with the video drivers, as under 15.1 I had desktop freezes occurring every couple of days. Eventually in the last month that slowed somewhat so I thought things had improved due to some system update. I suspect that setting the system to XRender did not help, although I didn't have any freezes last night for many hours.

My hardware:

Code:

rhack@localhost:~> inxi -b
System:    Host: localhost.localdomain Kernel: 5.3.18-lp152.57-preempt x86_64 bits: 64 Desktop: KDE Plasma 5.18.6 
           Distro: openSUSE Leap 15.2 
Machine:   Type: Desktop Mobo: ASUSTeK model: ROG STRIX X470-F GAMING v: Rev X.0x serial: <superuser/root required> 
           UEFI [Legacy]: American Megatrends v: 4207 date: 12/07/2018 
CPU:       6-Core: AMD Ryzen 5 2600X type: MT MCP speed: 2068 MHz min/max: 2200/3600 MHz 
Graphics:  Device-1: Advanced Micro Devices [AMD/ATI] Baffin [Radeon RX 550 640SP / RX 560/560X] driver: amdgpu v: kernel 
           Display: x11 server: X.Org 1.20.3 driver: amdgpu,ati unloaded: fbdev,modesetting,vesa resolution: 1920x1080~60Hz 
           OpenGL: renderer: Radeon RX 550 Series (POLARIS11 DRM 3.33.0 5.3.18-lp152.57-preempt LLVM 9.0.1) v: 4.5 Mesa 19.3.4 
Network:   Device-1: Intel I211 Gigabit Network driver: igb 
Drives:    Local Storage: total: 26.41 TiB used: 18.53 TiB (70.2%) 
Info:      Processes: 341 Uptime: N/A Memory: 15.56 GiB used: 2.21 GiB (14.2%) Shell: bash inxi: 3.1.00

Code:

rhack@localhost:~> inxi -G
Graphics:  Device-1: Advanced Micro Devices [AMD/ATI] Baffin [Radeon RX 550 640SP / RX 560/560X] driver: amdgpu v: kernel 
           Display: x11 server: X.Org 1.20.3 driver: amdgpu,ati unloaded: fbdev,modesetting,vesa resolution: 1920x1080~60Hz 
           OpenGL: renderer: Radeon RX 550 Series (POLARIS11 DRM 3.33.0 5.3.18-lp152.57-preempt LLVM 9.0.1) v: 4.5 Mesa 19.3.4

Problem I have with that list is that I have set the Compositor to XRender - not OpenGL... Any idea why inxi is not showing XRender?

What I need from the community is some idea of how to debug what is actually going on with these freezes. I need a step-by-step about the following:
1) What logs are relevant?
2) How can I make the relevant logs persistent across boots, so I can find out what happened before and during the freeze?
3) What Compositor is recommended for an AMD Radeon Graphics Card?
4) Are there any issues with the amdgpu drivers related to either 15.1 or the Compositor setting?
5) Which server should I be using: X11, Wayland, Full Wayland, what?

This freeze thing is extremely irritating and I need to dig in and fix it - or I'll be looking at another reinstall with no guarantee that will solve anything.

Greatly appreciate any help anyone can give me - preferably with a suggestion for how to approach this methodically.

**dickhack** · 18-Dec-2020, 20:48

Just had one more total system freeze.

This time it happened exactly like the time I lost my sound completely before I did this last clean install:

I had Dolphin open, jEdit, Konsole, Kplayer and Youtube-DLG. Youtube-DLG was downloading some Youtube videos. I was laying on the bed listening to epic music using KPlayer. The song cuts off instantly. I get up, go over to the computer, nada: no cursor, no Alt-Tab, nothing moving on the screen (Youtube-DLG no longer shows anything downloading). Nothing else was being done on the system at the time other than whatever goes on normally in the background. This is exactly how it happened the time before I lost my sound. This time I still have sound.

Before doing these downloads I had reset the Compositor from XRender back to OpenGL 3.1. So I guess now I'll try resetting it back to OpenGL 2.0. Question: Why are there two versions of OpenGL installed in a default install?

**dcurtisfra** · 19-Dec-2020, 11:58

@dickhack:

Please check the following for anything unusual:

Code:

 > cat /proc/cmdline

Code:

 # zypper verify
 # rpm --verify --all
 # rpmconfigcheck
 # journalctl --this-boot

How large is the system partition? – The one with /tmp and /var – if they're not in their own partitions …

**dickhack** · 19-Dec-2020, 22:00

Originally Posted by dcurtisfra

@dickhack:

Please check the following for anything unusual:

Code:

 > cat /proc/cmdline

Code:

 # zypper verify
 # rpm --verify --all
 # rpmconfigcheck
 # journalctl --this-boot

How large is the system partition? – The one with /tmp and /var – if they're not in their own partitions …

Thanks for your reply.

Command line is as follows:

Code:

BOOT_IMAGE=/boot/vmlinuz-5.3.18-lp152.57-preempt 
root=UUID=f0f32d8c-bb81-4755-9491-5b2e01dcf6b8 
isofrom=/dev/disk/by-uuid/e0fc3b4e-9a4a-4473-baf2-70b5675f9cd7:/openSUSE-Leap-15.2-DVD-x86_64.iso 
isofrom_device=/dev/disk/by-uuid/e0fc3b4e-9a4a-4473-baf2-70b5675f9cd7 
isofrom_system=/openSUSE-Leap-15.2-DVD-x86_64.iso 
loader=syslinux splash=silent 
resume=/dev/disk/by-id/nvme-ADATA_SX8200PNP_2J1720070214-part3 mitigations=auto 
quiet

rpmverify showed the following:

Code:

rhack@localhost:/Data2/Work> sudo rpm --verify --all
.M.......  g /run/tmux
S.5....T.  c /etc/systemd/journald.conf
.M.......  g /var/log/boot.log
....L....  d /usr/share/man/man1/ftp.1.gz
.M.......  c /var/log/NetworkManager
S.5....T.  c /etc/chrony.conf
.M.......  g /var/lib/chrony/drift
SM5....T.  c /etc/fonts/conf.d/30-metric-aliases.conf
.M....G..  g /var/log/lastlog
.M.......  g /var/lib/flatpak
missing     /var/cache/libx11/compose/l4_030_313cb605_00280cc0
.M.......  g /etc/xml/catalog-d.xml
S.5....T.  c /etc/unixODBC/odbcinst.ini
.M.......  g /run/avahi-daemon
.M.......  g /usr/share/fonts/100dpi/encodings.dir
.M.......  g /usr/share/fonts/100dpi/fonts.scale
.M.......  g /usr/share/fonts/75dpi/encodings.dir
.M.......  g /usr/share/fonts/75dpi/fonts.scale
....L....    /etc/ImageMagick-7
S.5....T.  c /etc/fonts/conf.d/10-rendering-options.conf
S.5....T.  c /etc/fonts/conf.d/58-family-prefer-local.conf
........P    /usr/lib64/libexec/ksysguard/ksgrd_network_helper
.......T.  c /etc/login.defs
.M.......  g /var/lib/pulseaudio
.M.......  g /run/mcelog
....L....  c /etc/pam.d/common-account
....L....  c /etc/pam.d/common-auth
....L....  c /etc/pam.d/common-password
....L....  c /etc/pam.d/common-session
.....UG..  g /run/sddm
.M.......  g /etc/iscsi/initiatorname.iscsi
.M.......  g /run/netconfig
.M.......  g /run/netconfig/resolv.conf
.M.......  g /run/netconfig/yp.conf
S.5....T.  c /etc/zypp/repos.d/teamviewer.repo
.M.......  g /var/log/alternatives.log
.M.......  g /var/lib/ca-certificates/ca-bundle.pem
.M.......  g /var/lib/ca-certificates/java-cacerts
.M...U...    /var/cache/cups
.......T.  c /etc/YaST2/control.xml
..5....T.    /var/lib/nordvpn/data/insights.dat
S.5....T.    /var/lib/nordvpn/data/servers.dat
.M.......  g /usr/share/fonts/misc/encodings.dir
.M.......  g /usr/share/fonts/misc/fonts.scale
.M.......  g /etc/mpv/scripts
.M.......  c /var/log/audit/audit.log
.......T.    /usr/lib64/gconv/gconv-modules.cache
....L....    /usr/bin/vi
S.5....T.  c /etc/sane.d/dll.conf
.M.......  g /run/cryptsetup
.M.......  g /boot/grub2/grub.cfg
S.5....T.  c /etc/default/grub
missing     /usr/bin/pip
.M.......  g /usr/share/fonts/Type1/encodings.dir
.M.......  g /usr/share/fonts/cyrillic/encodings.dir
.M.......  g /usr/share/fonts/cyrillic/fonts.scale
.M.......  g /usr/share/fonts/truetype/encodings.dir
.M.......  g /etc/udev/hwdb.bin
.M.......  g /usr/lib/udev/compat-symlink-generation
S.5....T.  c /etc/postfix/main.cf
S.5....T.  c /etc/postfix/master.cf

plus a ****-ton of missing python2.7 pip packages for some reason:

Code:

missing     /usr/bin/pip
missing     /usr/bin/pip2
missing     /usr/bin/pip2.7
missing     /usr/lib/python2.7/site-packages/pip
missing     /usr/lib/python2.7/site-packages/pip-10.0.1-py2.7.egg-info
and many, many more

rpmconfigcheck found nothing.

I assume you don't want to see the entire output of journalctl --this-boot as it's enormous. In case you do, here is the suse.paste link: https://paste.opensuse.org/b5ee2a98

I trimmed all of the journalctrl output from today. I also trimmed a lot of "spam" that Mega's MegaSync does which puts "overlays" on every directory on the system (I have thousands). I just disabled that as I'm not syncing to Mega anyway. If you see anything in there of interest, let me know. I couldn't see anything. I'm not even sure it covers a period when I had one of the freezes. When the next freeze occurs, I'll be able to check it as I've figured out how to set the journal persistent.

The system partition is 34.22GB with 21.4GB free or 37% full. /home is 15GB, only 17% full - I use external drives to hold most of my data. /home is only for settings and such.

Hope this helps.

**mrmazda** · 20-Dec-2020, 17:36

If XMP is enabled in the BIOS, try disabling it.

**dickhack** · 20-Dec-2020, 23:51

Originally Posted by mrmazda

If XMP is enabled in the BIOS, try disabling it.

I'll check, but I don't remember enabling it when I built this system. Apparently it's called "DOCP" for ASUS AMD mobos. I did make adjustments to the motherboard, mostly to try to lower the operating temperature (AMD CPUs tend to run hot), but I think the settings I've had for the last two years have been fine. No problems at all with the hardware so far.

I'll let you know what I find.

**dickhack** · 21-Dec-2020, 02:19

Ah, well, yes, it is set. The setting shows only two options: 5V or D.O.C.P. DDR4-3000 15-17-17-3. The Memory Frequency is set to DDR4-3000 MHz. The memory I have in it is CORSAIR Vengeance LPX 16GB (2 x 8GB) 288-Pin DDR4 SDRAM DDR4 3000 (PC4 24000) Memory Kit Model CMK16GX4M2B3000C15 from Newegg. Should I change it to 5V?

The changes I made back when I built the box were to the VDDCR CPU Voltage at 1.275V and the VDDCR SOC Voltage to 0.900V. Everything else was set to auto. I spent quite some time working on those settings in order to get what I thought was excessive heat at idling - the board shows 47 degrees Centigrade when I'm in the BIOS, but a sensors run shows 35 degrees when operating. I've never understood why the thing is hotter in the BIOS than while running. I do have seven fans in the box: 2 front input, one side input, two top output, one rear output, plus the PSU, CPU and GPU fans. So the box should be fine heat-wise.

I know a lot of people had problems working with this motherboard early on, back in 2018, getting stable overclocks. But I never overclocked. I simply adjusted the voltage to try to get the heat down. Eventually I gave up because it appeared that was the normal condition of these AMD CPUs - the boost capability causes them to jack up the voltage temporarily, then it falls back. Every time I work with wallpaper images, I can hear the fans ramp up, then settle back down. I have the fan profile set very aggressively in the motherboard settings.

So, what happens if I reset the D.O.C.P. to 5V? I can try it, but I have little knowledge about any possible effects on anything. I assume the memory speed will drop back to 2133 or something. I probably won't really notice that. But I don't want to do anything that might make the system even more unstable. The system works fine ninety-nine percent of the time - it just freezes every couple days. I've had no freeze at all today - yet.

The 15.1 version didn't freeze - it just crashed the desktop every couple days. Executing programs continued to run fine; I could let it finish any downloads in progress. If these current freezes are hardware related, why didn't they exist before December 6 when I upgraded? Is 15.2 doing something significantly different from 15.1 that could trigger memory issues? Either the X11 or KDE? By the way, I'm not running Wayland sessions, if that matters.

**dickhack** · 21-Dec-2020, 03:58

Well, here's another circus...

So a few minutes after I posted the last post where I said I hadn't had a freeze today - yet. Guess what? Freeze... I thought of possibly updating the BIOS since I haven't done that since I built the box in December, 2018, so I was going to the ASUS Web site in Firefox when total freeze again.

So I said, the hell with it. I went into the BIOS and reset the DOCP to Auto, which reset several other things. I then went in to the VDDCR CPU and SOC voltages and reset them to auto. So now I have a box where the temperatures *start* at 52 degrees Centigrade instead of the 40-45 I had previously. The fans ramp up more now because the box is hotter.

So I have the thought that maybe I should update the BIOS because maybe they've made improvements which would control the voltage and the heat better. So I go to their site, look up the right BIOS update and download it. I unzip it, format a USB drive with FAT32 as they instruct, copied the file, went into the BIOS, ran the EZ tool, selected the update - and the EZ tool says "not a proper BIOS". Some Google research shows this happens to a *lot* of people. So after several tries, renaming the file to remove dashes, and the like, none of which worked, I decided the hell with it, I'll update over the Internet which the EZ tool supports. Fine, I select that, it says there's a BIOS update, I tell it to go, it downloads it - then says it's not a proper BIOS! Their own **** file is not a proper BIOS! Thanks so much, ASUS, you morons!

OK, so the voltage and memory are all back to stock. Let's see if I have any more freezes. If I do, then it's not the hardware, it's either XII, the amdgpu, or KDE or something else. If I get another freeze, I'll reset the hardware back to where I had it, since that was fine for two years.

**dcurtisfra** · 21-Dec-2020, 09:57

Originally Posted by dickhack

Hope this helps.

Any particular reason for using the Preempt Kernel as opposed to the Default Kernel?

N.B.: Due to a recent patch, quite a few folks now have the Preempt Kernel “accidentality” in use rather than, the Default Kernel …

**dcurtisfra** · 21-Dec-2020, 10:10

Originally Posted by dickhack

rpmverify showed the following:

The fact that, ‘/usr/bin/pip’ is missing means that, something has been “accidentally” deleted …

Code:

 > rpm --query --whatprovides /usr/bin/pip
python3-pip-10.0.1-lp152.4.6.1.noarch
 > l /usr/bin/pip
lrwxrwxrwx 1 root root 21 30. Nov 23:26 /usr/bin/pip -> /etc/alternatives/pip*
 > which -a pip
/usr/bin/pip
 > l /etc/alternatives/pip
lrwxrwxrwx 1 root root 15 18. Aug 15:09 /etc/alternatives/pip -> /usr/bin/pip3.6*
 > l /usr/bin/pip3.6
-rwxr-xr-x 1 root root 371 30. Nov 23:26 /usr/bin/pip3.6*
 > rpm --query --whatprovides /usr/bin/pip3.6
python3-pip-10.0.1-lp152.4.6.1.noarch
 >

You need to forcibly re-install the “python3-pip” package:

Code:

 # zypper install --force python3-pip

But, you seem to have both Python 2 and, Python 3, installed – for whatever reason …

You may well have to, also, forcibly reinstall the “python2-pip” package …
You may well have to check the rest of your Python 3 plus Python 2 installation …
One is forced to make the assumption that, you know how to switch between Python 3 and Python 2 by means of the “alternatives” mechanism …