pkg upgrade desaster

W

winkoe

FreeBSD hplap 13.0-RELEASE FreeBSD 13.0-RELEASE #0 releng/13.0-n244733-ea31abc261f: Fri Apr 9 04:24:09 UTC 2021 root@releng1.nyi.freebsd.org:/usr/obj/usr/src/amd64.amd64/sys/GENERIC amd64

HP-laptop with Display: x11 server: X.Org 1.20.9 driver: amdgpu,ati unloaded: fbdev,modesetting,vesa resolution: 1920x1080~60Hz

After sporting a new laptop with a fresh 13.0 RELEASE and everything working near to perfection I ventured a pkg upgrade.

My previous experience with pkg upgrade on the 11-versions always ended in a minor/major disaster. Same thing now: ttyv1 is being flooded with a never ending repetition of

Code: 
amdgpu: [powerplay] pp_dpm_get_mclk_od was not implemented.
amdgpu: [powerplay] pp_dpm_get_sclk_od was not implemented.

Furthermore the zzz which used to work perfectly for sleepmode, now leads to a complete freeze requiring a hard reset.

What else is rank I have not yet discovered.

Before I start to do desperate actions - does anyone have a suggestion how to continue?

winkoe
 
winkoe said:
HP-laptop with Display: x11 server: X.Org 1.20.9 driver: amdgpu,ati unloaded: fbdev,modesetting,vesa resolution: 1920x1080~60Hz
Click to expand...

Post your /var/log/Xorg.0.log: cat /var/log/Xorg.0.log | nc termbin.com 9999 and post the resulting URL here.

winkoe said:
My previous experience with pkg upgrade on the 11-versions always ended in a minor/major disaster.
Click to expand...
This often happens when someone doesn't understand how it works and tries to force pkg(8) to do things it wasn't meant to be doing. Or when someone tries to mix ports and packages without understanding the interaction between the different versions.
 
Upon pciconf -lv | egrep -i "radeon|amd" I get a 25 fold repetition of
Code: 
   vendor     = 'Advanced Micro Devices, Inc. [AMD]'
and two times
Code: 
    vendor     = 'Advanced Micro Devices, Inc. [AMD/ATI]'
.
winkoe
 
From Xorg.0.log

Code: 
[    47.670] (II) modeset(0): glamor X acceleration enabled on AMD Radeon(TM) Vega 8 Graphics (RAVEN, DRM 3.35.0, 13.0-RELEASE, LLVM 10.0.1)
Notice 'Vega 8 Graphics Raven'.

Install besides graphics/drm-fbsd13-kmod (drm-kmod) x11-drivers/xf86-video-amdgpu. This log in the commit history of the latter mentions support for Raven Ridge APUs.

You need to create a xorg config file: /usr/local/etc/X11/xorg.conf.d/amdgpu.conf. See amdgpu(4) for details.
 
The never ending message
Code: 
amdgpu: [powerplay] pp_dpm_get_mclk_od was not implemented.
amdgpu: [powerplay] pp_dpm_get_sclk_od was not implemented.
is a consequence of a java application with intermittant calls sysctl.
Modifying the java-thing stopped the nuissance.

winkoe
 
winkoe said:
is a consequence of a java application with intermittant calls sysctl.
Modifying the java-thing stopped the nuissance.
Click to expand...
Hi winkoe,

I'm having the same error. Can you explain how did you resolve this? Really appreciate if you could specify a bit more in detail how you fixed this.

Thanks a lot!
 
Hi jcamos,

I have a few homebrew Java applications on my Desktop.

In that particular case I wanted to have the coretemp displayed via sysctl -a | grep temperature. And simultaneously sysctl ... was being used by a different application.

That coincidence was the culprit.

Swapping one of the commands to read hwstat | grep tz0 solved the issue.

Is that the answer you were asking for?

Good luck
winkoe
 
Thanks a lot! It does give me a hint for sure! My dilemma here is that I haven't made any change that would be obvious for this sudden behaviour... Either an update caused this or something was changed that I might have forget I did... need to further investigate. Still it's very annoying :/

I post something as soon as I figure out/or not

Cheers!
 
I see where the problem comes from! I was testing out Webmin and as soon as I open the browser it starts firing
Code: 
kernel: amdgpu: pp_dpm_get_sclk_od was not implemented.
Now why this is happening I have no clue... :what:

Edit:
winkoe said:
Is that the answer you were asking for?
Click to expand...
Actually it did help a lot!!!

sysctl -a | grep temperature provokes this error. This is confirmed because as soon as I manually issue the command I get this output. Now somewhere in Webmin this command is called which in turn floods my ttyv0
 
This is still a problem on 14.3 and amdgpu.

Calling sysctl -a causes the pair of console messages.

Calling sysctl dev.amdtemp.0.core0.sensor0 does not.

My guess is that probing some sysctl causes the message.

Edit: This is on an AMD Ryzen 3 3300U, likely a different processor than the forgoing and in a Lenovo M75n, not an HP Laptop

AMD Ryzen 3 PRO 3300U w/ Radeon Vega Mobile Gfx (2096.17-MHz K8-class CPU)
Code: 
$ kldstat
Id Refs Address                Size Name
 1  149 0xffffffff80200000  1f41500 kernel
 2    1 0xffffffff82142000   11d718 if_re.ko
 3    1 0xffffffff82260000     7808 cryptodev.ko
 4    1 0xffffffff82269000   5e9328 zfs.ko
 5    1 0xffffffff82853000     7618 amdtemp.ko
 6    2 0xffffffff8285b000     36e0 amdsmn.ko
 7    1 0xffffffff8285f000     5b90 ng_ether.ko
 8    8 0xffffffff82865000    16cc0 netgraph.ko
 9    1 0xffffffff8287c000     7028 ng_socket.ko
10    1 0xffffffff836e9000    3de48 if_wg.ko
11    1 0xffffffff84200000   6648d8 amdgpu.ko
12    2 0xffffffff84865000    85090 drm.ko
13    1 0xffffffff841f8000     22b8 iic.ko
14    2 0xffffffff841fb000     4120 linuxkpi_video.ko
15    3 0xffffffff848eb000     7350 dmabuf.ko
16    4 0xffffffff848f3000     3378 lindebugfs.ko
17    1 0xffffffff848f7000     c338 ttm.ko
18    1 0xffffffff84904000     2220 amdgpu_picasso_gpu_info_bin.ko
19    1 0xffffffff84907000     64e0 amdgpu_picasso_sdma_bin.ko
20    1 0xffffffff8490e000    2c2e0 amdgpu_picasso_asd_bin.ko
21    1 0xffffffff8493b000     c4e0 amdgpu_picasso_ta_bin.ko
22    1 0xffffffff84948000     7b70 amdgpu_raven_dmcu_bin.ko
23    1 0xffffffff84950000     7560 amdgpu_picasso_pfp_bin.ko
24    1 0xffffffff84958000     6560 amdgpu_picasso_me_bin.ko
25    1 0xffffffff8495f000     4560 amdgpu_picasso_ce_bin.ko
26    1 0xffffffff84964000     b9c8 amdgpu_picasso_rlc_bin.ko
27    1 0xffffffff84970000    43800 amdgpu_picasso_mec_bin.ko
28    1 0xffffffff849b4000    43800 amdgpu_picasso_mec2_bin.ko
29    1 0xffffffff849f8000    5b4c0 amdgpu_picasso_vcn_bin.ko
30    1 0xffffffff84a54000     3390 acpi_wmi.ko
31    1 0xffffffff84a58000    b1270 if_iwlwifi.ko
32    1 0xffffffff84b0a000    17314 if_iwm.ko
33    1 0xffffffff84b22000     6020 ig4.ko
34    1 0xffffffff84b29000     3220 intpm.ko
35    1 0xffffffff84b2d000     2178 smbus.ko
36    1 0xffffffff84c00000   28fe30 iwm9260fw.ko
37    1 0xffffffff84b30000     2240 cpuctl.ko
38    1 0xffffffff84b33000     7798 if_bridge.ko
39    1 0xffffffff84b3b000     60e0 bridgestp.ko
40    1 0xffffffff84b42000     85a0 if_ure.ko
41    1 0xffffffff84b4b000     3190 uether.ko
42    1 0xffffffff84b4f000     5640 ng_ubt.ko
43    3 0xffffffff84b55000     a250 ng_hci.ko
44    2 0xffffffff84b60000     2670 ng_bluetooth.ko
45    1 0xffffffff84b63000     2238 ng_eiface.ko
 
I know it doesn't help you much, but sometimes it's easier and less work overall just to switch to better hardware. :)

www.laptopmag.com

Lenovo ThinkPad X201

Lenovo's business ultraportable continues its dominance with long battery life and excellent performance in a small package.
www.laptopmag.com www.laptopmag.com


Or for something more recent... T480/T490 are good value now. There are lots on the market, don't pay too much; get 16GB RAM, and you can upgrade T490 to 40GB RAM:). And you can put 1TB NVME SSD.

www.pcmag.com

Lenovo ThinkPad T490 Review

The ThinkPad T490 is a top-shelf business laptop with impressive quality and security, although its battery life doesn't excel by today's standards.
www.pcmag.com www.pcmag.com

This one looks nice...

Out of the two, I would grab the T490 nowadays. True it doesn't have the classic tp keyboard like the X201, but the keyboard it does have is still very good. I bought T490 for myself, anyway.
 
Wondering.. have you tried compiling and installing drm-kmod from ports/source? This sounds a bit like typical problems that occur when your drm drivers are not perfectly in sync with your kernel? Particularly the one with the freeze...
 
You must log in or register to reply here.
Back
Top