Linux 4.20 released

Kernel 4.20 work very well on manjaro i3 :slight_smile:
Boot time feel more smooth/quick then before.
(nmb need time :wink: )

1 Like

The kernel gets better every day :slight_smile:

1 Like

Hey everyone.. I'm new to Manjaro and to this forum as well. I'm hoping to get some help wtih the following:

I just built the 4.20 kernel and headers today, however, I can't seem to get the NVIDIA kernel modules to build and load. I'm building the kernel from scratch primarily because I want the BFQ scheduler and BBR TCP congestion protocol built in statically instead of as modules. I followed the instructions given here: How to compile the mainline kernel the Manjaro way

Everything worked as expected. The issue I'm having is that DKMS doesn't seem to make/update the NVIDIA kernel modules. When I try to boot the newly built kernel, I'm getting an error that says "Failed to load kernel modules". LightDM fails to start and as a result, no X. Would someone be kind enough to lend a hand with this? Thanks!

I've done this a couple of times, so no expert, but the basic procedure is to download the files from https://gitlab.manjaro.org/packages/extra/linux420-extramodules/nvidia and move them to a build folder. (Assuming you want the 415.25 driver that is the latest on Manjaro, if you want an older one then download those build files instead.)

Adjust the _linuxprefix in the PKGBUILD (line 6) to be whatever your kernel is called. Then build the package with makepkg -s.

Move the built package into /var/cache/pacman/pkg/ and install with pacman -U /var/cache/pacman/pkg/packagename where packagename is the full name of the built file. (Alternatively use 'makepkg -si' but then your package will be in a different location to the rest.)

Hope that works for you. :wink:

p.s. It is better/usual forum etiquette to start a new thread for your query/problem. You know for next time or if my help is insufficient...

3 Likes

Thank you so much grinner! That worked. Is the entire kernel building process documented somewhere? Thanks again.

the manjaro kernel is missing the new Pressure-stall Information

# CONFIG_PSI is not set

^ not sure if thats the config.

@philm is there any reason for not enabling this?


BTW for me, 4.20 is the slowest kernel for a long time. Graphics performance is much(!) better with 4.19, general FS (read) performance seems also to be worse with f2fs and xfs and network performance feels a lot more sluggish. But since this is a .0 release I'm happy that nothing is broken, like it was with 4.19, 4.18 and 4.17 for me :joy:

Just works. Thank you Manjaro!

Mixed bag here. As with @Strit, my notebook doesn't boot with 4.20. I have Dell 2018 Inspiron 13 7000 2-in-1 with an AMD Ryzen 5 2500U and Vega 8 Graphics. No issues with 4.19. Maybe I need to get updated amdgpu modules to match 4.20? Probably. On my desktop, though, I haven't noticed any issues. It's a Ryzen 5 2500G (also Vega 8, but disabled) with a Radeon RX 580 as the primary GPU. I don't recall having to do anything special on it, just updated the kernel and rebooted. :man_shrugging:

EDIT: I updated my desktop a few days ago and my notebook today. Since I didn't have any issues on my desktop, I didn't feel like the risk on my notebook was too great. It's not my primary, so if it's down a day or so, it doesn't matter.

EDIT2: I chrooted and uninstalled 4.20 and was able to boot again. No surprise. My problem may be an outdated linux-firmware. On my notebook, it's 20181026.xxx, while on my desktop it's 20181214.xxx. I have refreshed the databases, but I'm not getting available updates. Both computers are on the stable branch (verified with pacman-mirrors -G).

@philm mentioned in the update thread, that there might be problems with some AMD hardware, because of the new firmware/amd-ucode update.

4.20.0 has just let me down, with UT of a lousy 1d4h. For me it's therefore been no happier than 4.19.5+, with the usual symptoms:

Individual open pgms progressively began partially malfunctioning then completely not responding; couldn't kill them in Firejail; couldn't logout; couldn't change to TTY... but at least REISUB worked. The full-freeze timestamp was 11:13:03, but the wonkiness began several minutes prior.

Summary
[kdemeoz@GA-Z97-HD3-Tower ~]$ sudo journalctl -b -1 | grep kernel
Dec 28 11:01:16 GA-Z97-HD3-Tower kernel: rcu: INFO: rcu_preempt detected stalls on CPUs/tasks:
Dec 28 11:01:16 GA-Z97-HD3-Tower kernel: rcu:         Tasks blocked on level-0 rcu_node (CPUs 0-7): P10582
Dec 28 11:01:16 GA-Z97-HD3-Tower kernel: rcu:         (detected by 7, t=18002 jiffies, g=16536037, q=908637)
Dec 28 11:04:16 GA-Z97-HD3-Tower kernel: rcu: INFO: rcu_preempt detected stalls on CPUs/tasks:
Dec 28 11:04:16 GA-Z97-HD3-Tower kernel: rcu:         Tasks blocked on level-0 rcu_node (CPUs 0-7): P10582
Dec 28 11:04:16 GA-Z97-HD3-Tower kernel: rcu:         (detected by 4, t=72007 jiffies, g=16536037, q=1800529)
Dec 28 11:07:16 GA-Z97-HD3-Tower kernel: rcu: INFO: rcu_preempt detected stalls on CPUs/tasks:
Dec 28 11:07:16 GA-Z97-HD3-Tower kernel: rcu:         Tasks blocked on level-0 rcu_node (CPUs 0-7): P10582
Dec 28 11:07:16 GA-Z97-HD3-Tower kernel: rcu:         (detected by 4, t=126012 jiffies, g=16536037, q=2423729)
Dec 28 11:07:58 GA-Z97-HD3-Tower kernel: BUG: Bad page state in process sh  pfn:1df89a
Dec 28 11:07:58 GA-Z97-HD3-Tower kernel: page:fffff31ec77e2680 count:0 mapcount:0 mapping:0400000000000000 index:0x0
Dec 28 11:07:58 GA-Z97-HD3-Tower kernel: flags: 0xaffff0000000000()
Dec 28 11:07:58 GA-Z97-HD3-Tower kernel: raw: 0affff0000000000 dead000000000100 dead000000000200 0400000000000000
Dec 28 11:07:58 GA-Z97-HD3-Tower kernel: raw: 0000000000000000 0000000000000000 00000000ffffffff 0000000000000000
Dec 28 11:07:58 GA-Z97-HD3-Tower kernel: page dumped because: non-NULL mapping
Dec 28 11:07:58 GA-Z97-HD3-Tower kernel: BUG: Bad page state in process ksysguard  pfn:1df8b0
Dec 28 11:07:58 GA-Z97-HD3-Tower kernel: page:fffff31ec77e2c00 count:0 mapcount:0 mapping:0000000000000000 index:0x0
Dec 28 11:07:58 GA-Z97-HD3-Tower kernel: flags: 0x2ffff0000000000()
Dec 28 11:07:58 GA-Z97-HD3-Tower kernel: raw: 02ffff0000000000 dead000000000100 dead000000000200 0000000000000000
Dec 28 11:07:58 GA-Z97-HD3-Tower kernel: raw: 0000000000000000 0000000000000000 00000000ffffffff 0200000000000000
Dec 28 11:07:58 GA-Z97-HD3-Tower kernel: page dumped because: page still charged to cgroup
Dec 28 11:07:58 GA-Z97-HD3-Tower kernel: page->mem_cgroup:0200000000000000
Dec 28 11:10:16 GA-Z97-HD3-Tower kernel: rcu: INFO: rcu_preempt detected stalls on CPUs/tasks:
Dec 28 11:10:16 GA-Z97-HD3-Tower kernel: rcu:         Tasks blocked on level-0 rcu_node (CPUs 0-7): P10582
Dec 28 11:10:16 GA-Z97-HD3-Tower kernel: rcu:         (detected by 6, t=180017 jiffies, g=16536037, q=3016715)
Dec 28 11:13:16 GA-Z97-HD3-Tower kernel: rcu: INFO: rcu_preempt detected stalls on CPUs/tasks:
Dec 28 11:13:16 GA-Z97-HD3-Tower kernel: rcu:         Tasks blocked on level-0 rcu_node (CPUs 0-7): P10582
Dec 28 11:13:16 GA-Z97-HD3-Tower kernel: rcu:         (detected by 5, t=234024 jiffies, g=16536037, q=3236869)
[kdemeoz@GA-Z97-HD3-Tower ~]$ 

Sob.

PS - I have Intel, not AMD.

fwiw ........ :roll_eyes: Both Intel i5
I just installed 420 on an old lap top:
$ inxi -MC Machine: Type: Laptop System: Dell product: Latitude E6410 v: 0001 serial: <root required> Mobo: Dell model: 0667CC v: A01 serial: <root required> BIOS: Dell v: A05 date: 08/10/2010 CPU: Topology: Dual Core model: Intel Core i5 M 520 bits: 64 type: MT MCP L2 cache: 3072 KiB Speed: 1472 MHz min/max: 1199/2400 MHz Core speeds (MHz): 1: 1479 2: 1495 3: 1489 4: 1468
seems to work fine.
I've been running on my desktop nearly a week, no problem:

$ inxi -MC
Machine:   Type: Desktop System: Dell product: Inspiron 3847 v: N/A serial: <root required> 
           Mobo: Dell model: 088DT1 v: A00 serial: <root required> BIOS: Dell v: A02 date: 12/03/2013 
CPU:       Topology: Quad Core model: Intel Core i5-4440 bits: 64 type: MCP L2 cache: 6144 KiB 
           Speed: 798 MHz min/max: 800/3300 MHz Core speeds (MHz): 1: 798 2: 798 3: 798 4: 798 
Slight newer. 

:smile:

Can you be more specific? I don't notice any difference in performance between 4.19 and 4.20. SSD or HDD? Which scheduler? Wireless or wired? Graphics card?

No problems here

System:    Host: hoflo Kernel: 4.20.0-1-MANJARO x86_64 bits: 64 Desktop: KDE Plasma 5.14.4 Distro: Manjaro Linux 
Machine:   Type: Desktop System: LENOVO product: 7307A1G v: ThinkCentre M58e serial: <filter> 
           Mobo: LENOVO model: N/A serial: <filter> BIOS: LENOVO v: 5HKT43AUS date: 09/07/2009 
CPU:       Topology: Dual Core model: Intel Core2 Duo E8400 bits: 64 type: MCP L2 cache: 6144 KiB 
           Speed: 2182 MHz min/max: 2003/3003 MHz Core speeds (MHz): 1: 2451 2: 2490 
Graphics:  Device-1: Intel 4 Series Integrated Graphics driver: i915 v: kernel 
           Display: x11 server: X.Org 1.20.3 driver: intel resolution: 1920x1080~60Hz 
           OpenGL: renderer: Mesa DRI Intel G41 v: 2.1 Mesa 18.2.6 
Audio:     Device-1: Intel NM10/ICH7 Family High Definition Audio driver: snd_hda_intel 
           Sound Server: ALSA v: k4.20.0-1-MANJARO 
Network:   Device-1: Marvell 88E8057 PCI-E Gigabit Ethernet driver: sky2 
           IF: ens33 state: up speed: 1000 Mbps duplex: full mac: <filter> 
Drives:    Local Storage: total: 232.89 GiB used: 15.56 GiB (6.7%) 
           ID-1: /dev/sda vendor: Samsung model: SSD 850 EVO 250GB size: 232.89 GiB 
Partition: ID-1: / size: 220.81 GiB used: 15.56 GiB (7.0%) fs: ext4 dev: /dev/sda1 
           ID-2: swap-1 size: 8.43 GiB used: 0 KiB (0.0%) fs: swap dev: /dev/sda2 
Sensors:   System Temperatures: cpu: 40.0 C mobo: N/A 
           Fan Speeds (RPM): N/A 
Info:      Processes: 159 Uptime: 26m Memory: 3.82 GiB used: 1.22 GiB (32.0%) Shell: bash inxi: 3.0.28 

regards
caho

I dont want to make this a support thread :wink:

more specific

SSD for f2fs (SSD 860 EVO)
HDD for XFS (ST500LT012-1DG142)

Whatever is default on Manjaros 4.19 and 4.20 kernels

Wired only

AMD Radeon R9 380


➜ inxi -Fxzc0
System:    Host: balder Kernel: 4.20.0-1-MANJARO x86_64 bits: 64 compiler: gcc v: 8.2.1 Desktop: Gnome 3.30.2 
           Distro: Manjaro Linux 
Machine:   Type: Desktop System: Gigabyte product: AB350M-Gaming 3 v: N/A serial: <filter> 
           Mobo: Gigabyte model: AB350M-Gaming 3-CF v: x.x serial: <filter> UEFI: American Megatrends v: F24c date: 12/06/2018 
CPU:       Topology: 8-Core model: AMD Ryzen 7 1700 bits: 64 type: MT MCP arch: Zen rev: 1 L2 cache: 4096 KiB 
           flags: lm nx pae sse sse2 sse3 sse4_1 sse4_2 sse4a ssse3 svm bogomips: 95860 
           Speed: 1337 MHz min/max: 1550/3000 MHz Core speeds (MHz): 1: 1357 2: 1274 3: 2369 4: 2327 5: 2246 6: 2252 7: 2074 
           8: 1986 9: 3376 10: 2853 11: 1313 12: 1276 13: 2131 14: 2142 15: 1347 16: 1346 
Graphics:  Device-1: Advanced Micro Devices [AMD/ATI] Tonga PRO [Radeon R9 285/380] vendor: PC Partner Limited driver: amdgpu 
           v: kernel bus ID: 06:00.0 
           Display: wayland server: X.org 1.20.3 driver: amdgpu resolution: <xdpyinfo missing> 
           OpenGL: renderer: AMD Radeon R9 380 Series (TONGA DRM 3.27.0 4.20.0-1-MANJARO LLVM 7.0.0) v: 4.5 Mesa 18.3.1 
           direct render: Yes 
Audio:     Device-1: Advanced Micro Devices [AMD/ATI] Tonga HDMI Audio [Radeon R9 285/380] vendor: PC Partner Limited 
           driver: snd_hda_intel v: kernel bus ID: 06:00.1 
           Device-2: Advanced Micro Devices [AMD] Family 17h HD Audio vendor: Gigabyte driver: snd_hda_intel v: kernel 
           bus ID: 08:00.3 
           Device-3: Logitech HD Webcam C525 type: USB driver: snd-usb-audio,uvcvideo bus ID: 1-5:3 
           Sound Server: ALSA v: k4.20.0-1-MANJARO 
Network:   Device-1: Realtek RTL8111/8168/8411 PCI Express Gigabit Ethernet vendor: Gigabyte driver: r8169 v: kernel 
           port: f000 bus ID: 03:00.0 
           IF: enp3s0 state: up speed: 1000 Mbps duplex: full mac: <filter> 
           Device-2: Intel Wireless 7260 driver: iwlwifi v: kernel port: f000 bus ID: 04:00.0 
           IF: wlp4s0 state: down mac: <filter> 
Drives:    Local Storage: total: 1.14 TiB used: 272.46 GiB (23.3%) 
           ID-1: /dev/sda vendor: Samsung model: SSD 860 EVO 500GB @ size: 465.76 GiB 
           ID-2: /dev/sdb vendor: Seagate model: ST500LT012-1DG142 size: 465.76 GiB 
           ID-3: /dev/sdc vendor: Crucial model: CT256M550SSD1 @ size: 238.47 GiB 
Partition: ID-1: / size: 50.00 GiB used: 29.40 GiB (58.8%) fs: f2fs dev: /dev/sda2 
           ID-2: /boot size: 499.0 MiB used: 190.3 MiB (38.1%) fs: vfat dev: /dev/sda1 
           ID-3: /home size: 415.00 GiB used: 90.83 GiB (21.9%) fs: f2fs dev: /dev/sda3 
Sensors:   System Temperatures: cpu: 35.8 C mobo: 26.0 C gpu: amdgpu temp: 43 C 
           Fan Speeds (RPM): cpu: 0 fan-1: 0 fan-3: 0 gpu: amdgpu fan: 962 
           Voltages: 12v: N/A 5v: N/A 3.3v: 1.67 vbat: 1.59 
Info:      Processes: 355 Uptime: 12m Memory: 15.68 GiB used: 3.17 GiB (20.2%) Init: systemd Compilers: gcc: 8.2.1 Shell: zsh  v: 5.6.2 inxi: 3.0.28 

At the very least people with similar setups and hardware can verify your findings.

1 Like

After inastall Kernel 4.20 my lan internet speed go from 100 mib/s to 10-15mib/s,
On wifi was good.
I think that is some driver problem....
Model: "Realtek RTL8101/2/6E PCI Express Fast/Gigabit Ethernet controller"
On kernel 4.19 everithing ok.
Greetings

When playing a Youtube video using kernel 4.20 the CPU score went up to 90-95% with a stuttering video. Now I'm back on 4.19 and although the score is still around 50% with the same video it now runs smoothly.
Cause of this was systemd with 70+%
image

Installed the 4.20 kernel yesterday, and no problems so far, everything's running good as it should be :slight_smile:

Installed the 4.20 kernel as soon as it was possible and everything is working fine, I don't think that there is actual diffrence in performance, but systemd-analyze decreased slightly

After testing for a few weeks, I can confirm this also on ext4 and ntfs. On my hdd's, disk I/O is 10-15% worse on 4.20. While transferring files between a ext4 and ntfs partition on separate drives, transfers on 4.14 and 4.19 average 110-115 MB/s (as reported by midnight commander), on 4.20 it is 90-100 MB/s.

Only after turning off all spectre/meltdown mitigations is fs performance improved.

1 Like

Forum kindly sponsored by