Linux kernel 5.8 is officially released

anon43355446 · August 5, 2020, 1:07pm

CUPS sucks, at least with my HP printer.

anon31687413 · August 5, 2020, 1:38pm

You need this patch in the kernel:
https://github.com/torvic9/linux58-vd/blob/arch/0006-vbox-58-support.patch

And these (the 3 patch files) for the host modules:
https://github.com/torvic9/linux58-vd/tree/arch/virtualbox

Known issue, reported here:
https://www.virtualbox.org/ticket/19644

manuel · August 5, 2020, 2:24pm

Thanks for the info!

So it will be fixed quite soon it seems.

anon31687413 · August 5, 2020, 2:37pm

Yes, VBox 6.1.14 should have the fixes.

(Obviously the patches above are not mine, but taken directly from the VBox ticket link, I just integrated them into my kernel)

anon31687413 · August 10, 2020, 3:19pm

Any Zen2 user on kernel 5.8 here?

I do not get any energy readings from sensors.
They were there on my custom 5.7 with backports from 5.8, but now on 5.8, I don’t have them anymore. Most probably a mistake on my side, but I can’t find it…

Here is the relevant commit:
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/drivers/hwmon?h=v5.8&id=8abee9566b7e8eecf566c4daf6be062a27369890

The config option is set: CONFIG_SENSORS_AMD_ENERGY=y

Could it be related to the BIOS update I did yesterday, introducing new AGESA?

Othoric · August 10, 2020, 3:22pm

I’ve been on AGESA V2 PI 1.0.0.2 since it was provided by ASUS back on June 29th. Let me switch to 5.8 and see what I get.

Resiliencia · August 10, 2020, 3:23pm

I’m with Linux-Hardened, but it will surely help, no problems around here:

$ uname -a
Linux judd-b365mds3h 5.7.11.a-1-hardened #1 SMP PREEMPT Wed, 29 Jul 2020 19:25:17 +0000 x86_64 GNU/Linux```

$ sensors
acpitz-acpi-0
Adapter: ACPI interface
temp1: +16.8°C (crit = +20.8°C)
temp2: +27.8°C (crit = +119.0°C)
temp3: +29.8°C (crit = +119.0°C)

coretemp-isa-0000
Adapter: ISA adapter
Package id 0: +30.0°C (high = +84.0°C, crit = +100.0°C)
Core 0: +30.0°C (high = +84.0°C, crit = +100.0°C)
Core 1: +28.0°C (high = +84.0°C, crit = +100.0°C)
Core 2: +29.0°C (high = +84.0°C, crit = +100.0°C)
Core 3: +28.0°C (high = +84.0°C, crit = +100.0°C)
Core 4: +30.0°C (high = +84.0°C, crit = +100.0°C)
Core 5: +28.0°C (high = +84.0°C, crit = +100.0°C)

Edit:I read wrong, I’m not 5.8 yet …

anon31687413 · August 10, 2020, 3:27pm

That’s Intel
En mi caso, es AMD Zen2. No hay problemas con los sensores de temperatura. Hace falta la informaciòn sobre la consomaciòn de energía, exprimida en Joule.

Othoric · August 10, 2020, 3:28pm

Readout after switching to mainline:

$ sensors
iwlwifi_1-virtual-0
Adapter: Virtual device
temp1: N/A

nvme-pci-0100
Adapter: PCI adapter
Composite: +36.9°C (low = -0.1°C, high = +89.8°C)
(crit = +94.8°C)

ucsi_source_psy_3_00081-i2c-3-08
Adapter: NVIDIA GPU I2C adapter
in0: 0.00 V (min = +0.00 V, max = +0.00 V)
curr1: 0.00 A (max = +0.00 A)

zenpower-pci-00c3
Adapter: PCI adapter
SVI2_Core: 1.39 V
SVI2_SoC: 1.07 V
Tdie: +62.8°C (high = +95.0°C)
Tctl: +62.8°C
Tccd1: +49.2°C
Tccd2: +37.0°C
SVI2_P_Core: 25.72 W
SVI2_P_SoC: 16.13 W
SVI2_C_Core: 19.11 A
SVI2_C_SoC: 15.01 A

anon31687413 · August 10, 2020, 3:30pm

Thanks.
Can you check whether you have amd_energy module loaded?
You also seem to be missing the Ecore and Esocket readings.

Othoric · August 10, 2020, 3:33pm

Negative. No amd_energy module

Othoric · August 10, 2020, 3:38pm

Upon further inspection I was also using zenpower from the AUR. Here is the readout from sensors without it on 5.8

$ sensors
iwlwifi_1-virtual-0
Adapter: Virtual device
temp1: N/A

nvme-pci-0100
Adapter: PCI adapter
Composite: +36.9°C (low = -0.1°C, high = +89.8°C)
(crit = +94.8°C)

ucsi_source_psy_3_00081-i2c-3-08
Adapter: NVIDIA GPU I2C adapter
in0: 0.00 V (min = +0.00 V, max = +0.00 V)
curr1: 0.00 A (max = +0.00 A)

k10temp-pci-00c3
Adapter: PCI adapter
Vcore: 1.49 V
Vsoc: 1.07 V
Tctl: +55.5°C
Tdie: +55.5°C
Tccd1: +32.0°C
Tccd2: +29.5°C
Icore: 13.00 A
Isoc: 12.75 A

anon31687413 · August 10, 2020, 3:44pm

Great, thanks! Quite similar to mine regarding k10temp.
Still wonder where the energy readings have gone.

Othoric · August 10, 2020, 3:48pm

For what it’s worth I switched back to 5.7.12-24-tkg-bmq and the sensors readout is the same as on mainline, so maybe something with AGESA? I also get the following during boot after updating to the latest BIOS as well:

[ 0.570803] do_IRQ: 1.55 No irq handler for vector
[ 0.571421] #2
[ 0.571421] do_IRQ: 2.55 No irq handler for vector
[ 0.572745] #3
[ 0.572745] do_IRQ: 3.55 No irq handler for vector
[ 0.574088] #4
[ 0.574088] do_IRQ: 4.55 No irq handler for vector
[ 0.575412] #5
[ 0.575412] do_IRQ: 5.55 No irq handler for vector
[ 0.576745] #6
[ 0.576745] do_IRQ: 6.55 No irq handler for vector
[ 0.578087] #7
[ 0.578087] do_IRQ: 7.55 No irq handler for vector
[ 0.579415] #8
[ 0.579415] do_IRQ: 8.55 No irq handler for vector
[ 0.580746] #9
[ 0.580746] do_IRQ: 9.55 No irq handler for vector
[ 0.582088] #10
[ 0.582088] do_IRQ: 10.55 No irq handler for vector

I’ve never actually figured out what that means or why it only shows up when using the latest AGESA

anon31687413 · August 10, 2020, 4:36pm

Yeah I’ve got these as well since the AGESA update…

EDIT: unfortunately I don’t have my 5.7 anymore (built for Manjaro), but I’m also starting to think that the readings are unavailable due to the BIOS update.
Well, not completely, they seem to be accessible via perf/sysctl, not sure if it’s the same thing:

#           time socket cpus             counts   unit events
     1.000225160 S0        1              22,92 Joules power/energy-pkg/                                           
     2.000390799 S0        1              20,18 Joules power/energy-pkg/                                           
     3.000498920 S0        1              21,84 Joules power/energy-pkg/                                           
     4.000610485 S0        1              20,12 Joules power/energy-pkg/                                           
     5.000719580 S0        1              20,82 Joules power/energy-pkg/                                           
     6.000840096 S0        1              26,10 Joules power/energy-pkg/                                           
     7.000952927 S0        1              48,57 Joules power/energy-pkg/

Othoric · August 10, 2020, 6:11pm

I actually just did a BIOS Flashback to an early version that I know was working correctly. At the very least it solves the “no irq handler vector” message I was always getting on boot-up. I was also getting a hardware error: link error in dmesg which seemed to be related to instability with SOC and VDDG voltages when using high memory speed/high fabric clock (which in my case is 3600/1800 respectfully). I had tried everything including raising the corresponding voltages higher than I would like, so I’m hoping this resolves it.

anon31687413 · August 10, 2020, 6:33pm

I had lots of memory related issues, especially boot problems, but also weird compilation errors, until circa April (don’t remember) when a BIOS update fixed it. That was before the new AGESA update however.

I also see that my board gives quite a lot of voltage to VDDG/VDDP when using the “auto” setting. Only running 3333/1667 though.
I’ll have to review these…

ricklinux · August 10, 2020, 9:04pm

I did the same thing for two reasons the no irq handler vector message and it also would not hibernate or suspend. This was on an MSI MPG X570 Gaming Plus

Edit: It had a new AGESA and another HDMI audio fix for another card which is odd because they are all in the same family of Graphics cards as i have the RX590

Othoric · August 10, 2020, 9:10pm

Yeah I definitely think there were some occasional issues with newer X570 BIOS revisions, I was having random disconnects with external USB drives as well as the other issues I mentioned. I’m going to stick with the BIOS I’m using now (BIOS 1302 on ASUS Crosshair VIII Hero Wi-Fi) and hopefully it gets resolved with a future update.