Radeon 6700xt performance issues on garuda after switching from 1080 cont

garuda-inxi
System:
  Kernel: 5.18.0-1-mainline arch: x86_64 bits: 64 compiler: gcc v: 12.1.0
    parameters: BOOT_IMAGE=/@/boot/vmlinuz-linux-mainline
    root=UUID=53e270ad-dfd5-4874-a125-69d21ada6b8d rw rootflags=subvol=@
    rd.udev.log_priority=3 vt.global_cursor_default=0 loglevel=3
    sysrq_always_enabled=1 amdgpu.ppfeaturemask=0xffffffff nowatchdog
    amdgpu.aspm=0 pcie_aspm=off
  Desktop: KDE Plasma v: 5.24.5 tk: Qt v: 5.15.4 info: latte-dock
    wm: kwin_x11 vt: 1 dm: SDDM Distro: Garuda Linux base: Arch Linux
Machine:
  Type: Desktop Mobo: ASRock model: X470 Taichi serial: <superuser required>
    UEFI: American Megatrends v: P4.86 date: 01/27/2022
Battery:
  Device-1: hidpp_battery_0 model: Logitech Wireless Mouse MX Master 3
    serial: <filter> charge: 100% (should be ignored) rechargeable: yes
    status: discharging
  Device-2: ps-controller-battery-4c:b9:9b:1b:68:38 model: N/A serial: N/A
    charge: N/A status: discharging
CPU:
  Info: model: AMD Ryzen 5 2600X bits: 64 type: MT MCP arch: Zen+
    family: 0x17 (23) model-id: 8 stepping: 2 microcode: 0x800820D
  Topology: cpus: 1x cores: 6 tpc: 2 threads: 12 smt: enabled cache:
    L1: 576 KiB desc: d-6x32 KiB; i-6x64 KiB L2: 3 MiB desc: 6x512 KiB
    L3: 16 MiB desc: 2x8 MiB
  Speed (MHz): avg: 4045 high: 4050 min/max: 2200/3600 boost: enabled
    scaling: driver: acpi-cpufreq governor: performance cores: 1: 4048 2: 4047
    3: 4050 4: 4039 5: 4050 6: 4046 7: 4048 8: 4050 9: 4050 10: 4015 11: 4050
    12: 4050 bogomips: 86438
  Flags: avx avx2 ht lm nx pae sse sse2 sse3 sse4_1 sse4_2 sse4a ssse3 svm
  Vulnerabilities:
  Type: itlb_multihit status: Not affected
  Type: l1tf status: Not affected
  Type: mds status: Not affected
  Type: meltdown status: Not affected
  Type: spec_store_bypass
    mitigation: Speculative Store Bypass disabled via prctl
  Type: spectre_v1
    mitigation: usercopy/swapgs barriers and __user pointer sanitization
  Type: spectre_v2
    mitigation: Retpolines, IBPB: conditional, STIBP: disabled, RSB filling
  Type: srbds status: Not affected
  Type: tsx_async_abort status: Not affected
Graphics:
  Device-1: AMD Navi 22 [Radeon RX 6700/6700 XT/6750 XT / 6800M]
    vendor: Gigabyte driver: amdgpu v: kernel pcie: gen: 4 speed: 16 GT/s
    lanes: 16 ports: active: DP-1,HDMI-A-1 empty: DP-2,HDMI-A-2
    bus-ID: 0f:00.0 chip-ID: 1002:73df class-ID: 0300
  Display: x11 server: X.Org v: 21.1.3 with: Xwayland v: 22.1.2
    compositor: kwin_x11 driver: X: loaded: amdgpu gpu: amdgpu display-ID: :0
    screens: 1
  Screen-1: 0 s-res: 6400x2160 s-dpi: 96 s-size: 1691x571mm (66.57x22.48")
    s-diag: 1785mm (70.27")
  Monitor-1: DP-1 mapped: DisplayPort-0 pos: primary,left
    model: LG (GoldStar) HDR WFHD serial: <filter> built: 2021 res: 2560x1080
    hz: 60 dpi: 81 gamma: 1.2 size: 798x334mm (31.42x13.15")
    diag: 869mm (34.2") modes: max: 2560x1080 min: 640x480
  Monitor-2: HDMI-A-1 mapped: HDMI-A-0 pos: right model: Samsung
    serial: <filter> built: 2019 res: 3840x2160 hz: 60 dpi: 52 gamma: 1.2
    size: 1872x1053mm (73.7x41.46") diag: 1093mm (43") ratio: 16:9 modes:
    max: 3840x2160 min: 720x400
  OpenGL: renderer: AMD Radeon RX 6700 XT (navy_flounder LLVM 13.0.1 DRM
    3.46 5.18.0-1-mainline)
    v: 4.6 Mesa 22.1.0 direct render: Yes
Audio:
  Device-1: AMD Navi 21/23 HDMI/DP Audio driver: snd_hda_intel v: kernel
    pcie: gen: 4 speed: 16 GT/s lanes: 16 bus-ID: 0f:00.1 chip-ID: 1002:ab28
    class-ID: 0403
  Sound Server-1: ALSA v: k5.18.0-1-mainline running: yes
  Sound Server-2: sndio v: N/A running: no
  Sound Server-3: PulseAudio v: 15.0 running: no
  Sound Server-4: PipeWire v: 0.3.51 running: yes
Network:
  Device-1: Intel I211 Gigabit Network vendor: ASRock driver: igb v: kernel
    pcie: gen: 1 speed: 2.5 GT/s lanes: 1 port: d000 bus-ID: 09:00.0
    chip-ID: 8086:1539 class-ID: 0200
  IF: enp9s0 state: up speed: 1000 Mbps duplex: full mac: <filter>
Bluetooth:
  Device-1: Edimax Bluetooth Adapter type: USB driver: btusb v: 0.8
    bus-ID: 1-8:4 chip-ID: 7392:c611 class-ID: e001 serial: <filter>
  Report: bt-adapter ID: hci0 rfk-id: 0 state: up address: <filter>
Drives:
  Local Storage: total: 2.96 TiB used: 3.2 TiB (108.1%)
  SMART Message: Unable to run smartctl. Root privileges required.
  ID-1: /dev/nvme0n1 maj-min: 259:3 vendor: Samsung
    model: SSD 970 EVO 250GB size: 232.89 GiB block-size: physical: 512 B
    logical: 512 B speed: 31.6 Gb/s lanes: 4 type: SSD serial: <filter>
    rev: 2B2QEXE7 temp: 37.9 C scheme: GPT
  ID-2: /dev/nvme1n1 maj-min: 259:0 vendor: Samsung model: SSD 980 1TB
    size: 931.51 GiB block-size: physical: 512 B logical: 512 B
    speed: 31.6 Gb/s lanes: 4 type: SSD serial: <filter> rev: 1B4QFXO7
    temp: 36.9 C scheme: GPT
  ID-3: /dev/sda maj-min: 8:0 vendor: Seagate model: ST2000DX002-2DV164
    size: 1.82 TiB block-size: physical: 4096 B logical: 512 B speed: 6.0 Gb/s
    type: HDD rpm: 7200 serial: <filter> rev: CC41 scheme: GPT
Partition:
  ID-1: / raw-size: 931.22 GiB size: 931.22 GiB (100.00%)
    used: 777.84 GiB (83.5%) fs: btrfs dev: /dev/nvme1n1p2 maj-min: 259:2
  ID-2: /boot/efi raw-size: 300 MiB size: 299.4 MiB (99.80%)
    used: 16.3 MiB (5.5%) fs: vfat dev: /dev/nvme1n1p1 maj-min: 259:1
  ID-3: /home raw-size: 931.22 GiB size: 931.22 GiB (100.00%)
    used: 777.84 GiB (83.5%) fs: btrfs dev: /dev/nvme1n1p2 maj-min: 259:2
  ID-4: /var/log raw-size: 931.22 GiB size: 931.22 GiB (100.00%)
    used: 777.84 GiB (83.5%) fs: btrfs dev: /dev/nvme1n1p2 maj-min: 259:2
  ID-5: /var/tmp raw-size: 931.22 GiB size: 931.22 GiB (100.00%)
    used: 777.84 GiB (83.5%) fs: btrfs dev: /dev/nvme1n1p2 maj-min: 259:2
Swap:
  Kernel: swappiness: 133 (default 60) cache-pressure: 100 (default)
  ID-1: swap-1 type: zram size: 31.27 GiB used: 89 MiB (0.3%) priority: 100
    dev: /dev/zram0
Sensors:
  System Temperatures: cpu: N/A mobo: 0.0 C gpu: amdgpu temp: 54.0 C
    mem: 56.0 C
  Fan Speeds (RPM): fan-3: 0 fan-5: 0 gpu: amdgpu fan: 0
Info:
  Processes: 449 Uptime: 17m wakeups: 5 Memory: 31.27 GiB
  used: 13.28 GiB (42.5%) Init: systemd v: 251 tool: systemctl Compilers:
  gcc: 12.1.0 clang: 13.0.1 Packages: pacman: 2334 lib: 582 flatpak: 0
  Shell: fish v: 3.4.1 default: Bash v: 5.1.16 running-in: alacritty
  inxi: 3.3.16
Garuda (2.6.3-2):
  System install date:     2022-05-22
  Last full system update: 2022-05-27
  Is partially upgraded:   No
  Relevant software:       NetworkManager
  Windows dual boot:       Probably (Run as root to verify)
  Snapshots:               Snapper
  Failed units:

basically, this is a continuation of Radeon 6700xt performance issues on garuda after switching from 1080 - #10 by zany130 originally I though just disabling mangohud vsync and fps settings would fix my issues but I was wrong and my issue persists.

I am having problems getting it to work correctly on any game I have tried (tomb raider, trails of cold steel 3, GTAV, and a hat in time.

On X I get really bad tearing regardless of vsync unless I run xrandr --output HDMI-A-0 --set TearFree on (or use an xorg config to set it permanently which is what I am doing now) but then that introduces stuttering (tried lowering graphical settings to see if that helps and it does not, it stutters regardless of gpu load)

on Wayland tearing is fixed but I get stuttering

I tried using corctl to set the performance mode to high and also added amdgpu.ppfeaturemask=0xffffffff to my grub settings and no dice.

also found this issue which seems to be related? As Im using a fps limiter (although running uncapped also gives some stuttering issues) power saving too agressive on RX 6800, causes stuttering with fps limiters (#1500) Β· Issues Β· drm / amd Β· GitLab

on Windows my GPU works without issues so I don't think its defective hardware.

Any help would be greatly appreciated

I have tried different kernels , also tried setting the gpu min clocks to 100 mhz less then its max clocks in corectl to force the gpu into higher frequencies. Tried with and without gamescope to see if that would maybe help (edited)

Have you tried disabling the compositor or changing any settings there? I think vsync is on auto by default.
Here's garuda-inxi from my PC that doesn't have any issues with the 6700 XT:

System:
  Kernel: 5.17.9-zen1-1-zen arch: x86_64 bits: 64 compiler: gcc v: 12.1.0
    parameters: BOOT_IMAGE=/@/boot/vmlinuz-linux-zen
    root=UUID=1748cca7-e422-4ed9-bd5c-3d12856f8671 rw rootflags=subvol=@
    quiet splash rd.udev.log_priority=3 vt.global_cursor_default=0
    systemd.unified_cgroup_hierarchy=1
    resume=UUID=dbb7a323-f52a-40d2-8ad1-ca512a750c83 loglevel=3
    sysrq_always_enabled=1
  Desktop: KDE Plasma v: 5.24.5 tk: Qt v: 5.15.4 info: latte-dock
    wm: kwin_x11 vt: 1 dm: SDDM Distro: Garuda Linux base: Arch Linux
Machine:
  Type: Desktop Mobo: ASRock model: X570 Taichi serial: <superuser required>
    UEFI: American Megatrends v: P4.30 date: 04/14/2021
CPU:
  Info: model: AMD Ryzen 9 3900XT bits: 64 type: MT MCP arch: Zen 2
    family: 0x17 (23) model-id: 0x71 (113) stepping: 0 microcode: 0x8701021
  Topology: cpus: 1x cores: 12 tpc: 2 threads: 24 smt: enabled cache:
    L1: 768 KiB desc: d-12x32 KiB; i-12x32 KiB L2: 6 MiB desc: 12x512 KiB
    L3: 64 MiB desc: 4x16 MiB
  Speed (MHz): avg: 4148 high: 4270 min/max: 2200/4776 boost: enabled
    scaling: driver: acpi-cpufreq governor: performance cores: 1: 3940 2: 4249
    3: 3876 4: 4144 5: 3893 6: 4270 7: 4239 8: 4250 9: 4241 10: 4241 11: 4255
    12: 4028 13: 4221 14: 3849 15: 4223 16: 4191 17: 4160 18: 4234 19: 4194
    20: 4184 21: 4229 22: 4037 23: 4222 24: 4187 bogomips: 182408
  Flags: avx avx2 ht lm nx pae sse sse2 sse3 sse4_1 sse4_2 sse4a ssse3 svm
  Vulnerabilities:
  Type: itlb_multihit status: Not affected
  Type: l1tf status: Not affected
  Type: mds status: Not affected
  Type: meltdown status: Not affected
  Type: spec_store_bypass
    mitigation: Speculative Store Bypass disabled via prctl
  Type: spectre_v1
    mitigation: usercopy/swapgs barriers and __user pointer sanitization
  Type: spectre_v2
    mitigation: Retpolines, IBPB: conditional, STIBP: conditional, RSB filling
  Type: srbds status: Not affected
  Type: tsx_async_abort status: Not affected
Graphics:
  Device-1: AMD Navi 22 [Radeon RX 6700/6700 XT/6750 XT / 6800M]
    vendor: ASRock driver: amdgpu v: kernel pcie: gen: 4 speed: 16 GT/s
    lanes: 16 ports: active: DP-1, DP-2, DP-3, HDMI-A-1 empty: none
    bus-ID: 0f:00.0 chip-ID: 1002:73df class-ID: 0300
  Device-2: Logitech C922 Pro Stream Webcam type: USB
    driver: snd-usb-audio,uvcvideo bus-ID: 3-4:3 chip-ID: 046d:085c
    class-ID: 0102 serial: <filter>
  Display: x11 server: X.Org v: 21.1.3 with: Xwayland v: 22.1.2
    compositor: kwin_x11 driver: X: loaded: amdgpu unloaded: modesetting,radeon
    alternate: fbdev,vesa gpu: amdgpu display-ID: :0.0 screens: 1
  Screen-1: 0 s-res: 7680x1971 s-dpi: 96 s-size: 2032x521mm (80.00x20.51")
    s-diag: 2098mm (82.59")
  Monitor-1: DP-1 mapped: DisplayPort-0 pos: 3-4 model: Samsung SyncMaster
    serial: <filter> built: 2006 res: 1280x1024 hz: 60 dpi: 86 gamma: 1.2
    size: 376x301mm (14.8x11.85") diag: 482mm (19") ratio: 5:4 modes:
    max: 1280x1024 min: 720x400
  Monitor-2: DP-2 mapped: DisplayPort-1 pos: 2-1 model: Acer G246HL
    serial: <filter> built: 2013 res: 1920x1080 hz: 60 dpi: 92 gamma: 1.2
    size: 531x299mm (20.91x11.77") diag: 609mm (24") ratio: 16:9 modes:
    max: 1920x1080 min: 720x400
  Monitor-3: DP-3 mapped: DisplayPort-2 pos: primary,1-2
    model: LG (GoldStar) QHD serial: <filter> built: 2021 res: 2560x1440
    hz: 144 dpi: 93 gamma: 1.2 size: 697x392mm (27.44x15.43")
    diag: 800mm (31.5") ratio: 16:9 modes: max: 2560x1440 min: 640x480
  Monitor-4: HDMI-A-1 mapped: HDMI-A-0 pos: 4-3 model: LCD TV built: 2011
    res: 1920x1080 hz: 60 dpi: 3048 gamma: 1.2 size: 16x9mm (0.63x0.35")
    diag: 6mm (0.2") ratio: 15:9 modes: max: 1920x1080 min: 640x480
  OpenGL: renderer: AMD Radeon RX 6700 XT (navy_flounder LLVM 13.0.1 DRM
    3.44 5.17.9-zen1-1-zen)
    v: 4.6 Mesa 22.1.0 direct render: Yes
Audio:
  Device-1: AMD Navi 21/23 HDMI/DP Audio driver: snd_hda_intel v: kernel
    pcie: gen: 4 speed: 16 GT/s lanes: 16 bus-ID: 0f:00.1 chip-ID: 1002:ab28
    class-ID: 0403
  Device-2: AMD Starship/Matisse HD Audio vendor: ASRock
    driver: snd_hda_intel v: kernel pcie: gen: 4 speed: 16 GT/s lanes: 16
    bus-ID: 11:00.4 chip-ID: 1022:1487 class-ID: 0403
  Device-3: Logitech C922 Pro Stream Webcam type: USB
    driver: snd-usb-audio,uvcvideo bus-ID: 3-4:3 chip-ID: 046d:085c
    class-ID: 0102 serial: <filter>
  Device-4: C-Media USB PnP Sound Device type: USB
    driver: hid-generic,snd-usb-audio,usbhid bus-ID: 5-3:4 chip-ID: 0d8c:013a
    class-ID: 0300
  Sound Server-1: ALSA v: k5.17.9-zen1-1-zen running: yes
  Sound Server-2: PulseAudio v: 15.0 running: no
  Sound Server-3: PipeWire v: 0.3.51 running: yes
Network:
  Device-1: Intel Wi-Fi 6 AX200 driver: iwlwifi v: kernel pcie: gen: 2
    speed: 5 GT/s lanes: 1 bus-ID: 05:00.0 chip-ID: 8086:2723 class-ID: 0280
  IF: wlp5s0 state: up mac: <filter>
  Device-2: Intel I211 Gigabit Network vendor: ASRock driver: igb v: kernel
    pcie: gen: 1 speed: 2.5 GT/s lanes: 1 port: f000 bus-ID: 07:00.0
    chip-ID: 8086:1539 class-ID: 0200
  IF: enp7s0 state: up speed: 1000 Mbps duplex: full mac: <filter>
  IF-ID-1: tailscale0 state: unknown speed: -1 duplex: full mac: N/A
  IF-ID-2: virbr0 state: down mac: <filter>
Bluetooth:
  Device-1: Intel AX200 Bluetooth type: USB driver: btusb v: 0.8
    bus-ID: 3-2:2 chip-ID: 8087:0029 class-ID: e001
  Report: bt-adapter ID: hci0 rfk-id: 0 state: up address: <filter>
Drives:
  Local Storage: total: 2.3 TiB used: 1.59 TiB (69.2%)
  SMART Message: Unable to run smartctl. Root privileges required.
  ID-1: /dev/nvme0n1 maj-min: 259:0 vendor: Samsung
    model: SSD 970 EVO Plus 2TB size: 1.82 TiB block-size: physical: 512 B
    logical: 512 B speed: 31.6 Gb/s lanes: 4 type: SSD serial: <filter>
    rev: 2B2QEXM7 temp: 47.9 C scheme: GPT
  ID-2: /dev/sda maj-min: 8:0 vendor: Samsung model: SSD 840 EVO 500GB
    size: 465.76 GiB block-size: physical: 512 B logical: 512 B speed: 6.0 Gb/s
    type: SSD serial: <filter> rev: DB6Q scheme: GPT
  ID-3: /dev/sdb maj-min: 8:16 type: USB vendor: Kingston
    model: UHSII uSD Reader size: 29.72 GiB block-size: physical: 512 B
    logical: 512 B type: N/A serial: <filter> rev: 0013 scheme: MBR
  SMART Message: Unknown USB bridge. Flash drive/Unsupported enclosure?
Partition:
  ID-1: / raw-size: 1.75 TiB size: 1.75 TiB (100.00%) used: 1.59 TiB (91.0%)
    fs: btrfs dev: /dev/nvme0n1p2 maj-min: 259:2
  ID-2: /boot/efi raw-size: 256 MiB size: 252 MiB (98.46%)
    used: 546 KiB (0.2%) fs: vfat dev: /dev/nvme0n1p1 maj-min: 259:1
  ID-3: /home raw-size: 1.75 TiB size: 1.75 TiB (100.00%)
    used: 1.59 TiB (91.0%) fs: btrfs dev: /dev/nvme0n1p2 maj-min: 259:2
  ID-4: /var/log raw-size: 1.75 TiB size: 1.75 TiB (100.00%)
    used: 1.59 TiB (91.0%) fs: btrfs dev: /dev/nvme0n1p2 maj-min: 259:2
  ID-5: /var/tmp raw-size: 1.75 TiB size: 1.75 TiB (100.00%)
    used: 1.59 TiB (91.0%) fs: btrfs dev: /dev/nvme0n1p2 maj-min: 259:2
Swap:
  Kernel: swappiness: 133 (default 60) cache-pressure: 100 (default)
  ID-1: swap-1 type: zram size: 62.79 GiB used: 371.8 MiB (0.6%)
    priority: 100 dev: /dev/zram0
  ID-2: swap-2 type: partition size: 69.07 GiB used: 0 KiB (0.0%)
    priority: -2 dev: /dev/nvme0n1p3 maj-min: 259:3
Sensors:
  System Temperatures: cpu: 38.5 C mobo: 38.0 C gpu: amdgpu temp: 55.0 C
    mem: 54.0 C
  Fan Speeds (RPM): fan-1: 0 fan-2: 989 fan-3: 0 fan-4: 1421 fan-5: 1112
    fan-6: 2153 fan-7: 0 gpu: amdgpu fan: 0
Info:
  Processes: 638 Uptime: 3d 19h 11m wakeups: 0 Memory: 62.79 GiB
  used: 20.49 GiB (32.6%) Init: systemd v: 251 tool: systemctl Compilers:
  gcc: 12.1.0 clang: 13.0.1 Packages: pacman: 2820 lib: 621 flatpak: 0
  Shell: Bash v: 5.1.16 running-in: yakuake inxi: 3.3.16
e[1;34mGaruda (2.6.3-2):e[0m
e[1;34m  System install date:e[0m     2021-06-18
e[1;34m  Last full system update:e[0m 2022-05-28 e[1;31m↻
e[1;34m  Is partially upgraded:  e[0m No
e[1;34m  Relevant software:      e[0m NetworkManager
e[1;34m  Windows dual boot:      e[0m No/Undetected
e[1;34m  Snapshots:              e[0m Snapper
e[1;34m  Failed units:           e[0m 

2 Likes

well I have set up my system so it auto disables the compositor when a game is launched (using autocompose kwin script i think its called and I double checked that it is indeed disabling the compositor)

I have also tried changing the vsync options to only on cheap but i haven;t tried any of the other options since I thought they give worse performance (at least thats what the warning says)

only two things standout between our inxi and thats that you using a zen 2 cpu while I am using a zen+ 2600x so don't know if thats an issue I have read online that this combo should work though.

and you not using amdgpu.ppfeaturemask=0xffffffff or even the amdgpu.aspm=0 pcie_aspm=off flags. i needed those (plus corectl) to prevent the clocks going down. (i was having cases as you can see in an image in the other post where it was downclocking and fps was not stable or bellow 60)

1 Like

Possible fixed it. Things i did was update the zen kernel to 5.18 and enable resizable bar on my bios. Both didn't fix my issue so i decided to try gaming only with my 4k tv enabled and voila no stutter yet. I see @Technetium also has multi monitors? Do you also have to disable them? Or are you able to run them all at the same time?

EDIT: oh and I'm no longer using gamescope as the whole reason for that was so i could play with both monitors enabled as without it games wouldnt scale correctly or show up on the correct screen

The resizable bar being on or off definitely should not have caused any issue, but it's good you turned it on - and also good you updated BIOS!

I'm not yet using the new kernel so I can't vouch for functionality there. I have never had to disable any of my monitors for any reason. Sometimes I have a game start on the wrong display, but it's easy enough to move over with a meta-key drag.

Maybe gamescope was part of your issue here?

1 Like

Hmm not sure . Without gamescope i was having trouble setting the right resolution as it was getting my ultrawides resolution instead of my tv.

Something I'm trying now is settings tear free to on for my tv and off for the rest of my displays on game start with gamemode then i revert it back to auto when i exit the game for all monitors.

This way i don't have to disable any displays and i also use gamescope to make sure it opens on my tv and runs at 4k. (Well for this game I'm running 1440p upscaled to 4k with FSR.)

it still stutters every now and then like this but it's definitely alot better then what it was before.

Still not as smooth as when i was running with my 1080 though

Tried to get a video but OBS makes the stuttering even worse and more consistent

So tried a few more things

Gamescope with only one monitor runs slightly worse than no gamescope 9n a single monitor.

However, and this is strange, dual monitors and no gamescope throttles my GPU (clocks go down and it struggles to hit 40fps where before it hit 60 easily)

Dual screen with gamescope runs about the same as single screen with gamescope and has no throttling issues.

I don't understand why there this discrepancy.

I tried updating to the git packages for gamescope and didn't seem to make a difference. Also read online that you can get better performance in gamescope by running
sudo setcap 'cap_sys_nice=eip' /usr/bin/gamescope but that breaks steam input for me (even with the gamescope -e flag which is supposed to fix gamescope breaking steam input)

I'm really regretting switching to AMD. :weary:

Have you done the most recent updates? Does it stutter with no FSR? What about with and without gamescope and no FSR? If you live-boot any other distros, do they give you the same issues? Maybe we can see what it works on, and find a difference in the configurations somewhere.

This is somewhat unrelated, but just in case maybe you can confirm your disk is healthy with smartctl. Corruption will do weird things sometimes.

Ironically I moved to AMD because of big struggles with Nvidia, you should not have to struggle so hard...

1 Like

FSR doesn't seem to make to much of a difference on any of the configurations it does add a little bit of load on the GPU when enabled but that is expected. on the other thread, I have (link in the first post) that I marked as solved to quickly I did try ubuntu really quickly and I noticed way better performance and this was with both monitors enabled and no game scope, something that would cause my GPU to go to its knees here on a garuda. the problem is I use a few packages that are not available or are too old on ubuntu so I would prefer not to have to switch over... I did try fedora really quickly but couldn't get that to boot. (think it might of been Plymouth that was also causing issues for me on garuda so I uninstalled it)

as for SMART info

sudo smartctl -a /dev/nvme1n1
smartctl 7.3 2022-02-28 r5338 [x86_64-linux-5.18.1-zen1-1-zen] (local build)
Copyright (C) 2002-22, Bruce Allen, Christian Franke, www.smartmontools.org

=== START OF INFORMATION SECTION ===
Model Number:                       Samsung SSD 980 1TB
Serial Number:                      S64ANJ0R930800J
Firmware Version:                   1B4QFXO7
PCI Vendor/Subsystem ID:            0x144d
IEEE OUI Identifier:                0x002538
Total NVM Capacity:                 1,000,204,886,016 [1.00 TB]
Unallocated NVM Capacity:           0
Controller ID:                      5
NVMe Version:                       1.4
Number of Namespaces:               1
Namespace 1 Size/Capacity:          1,000,204,886,016 [1.00 TB]
Namespace 1 Utilization:            736,245,497,856 [736 GB]
Namespace 1 Formatted LBA Size:     512
Namespace 1 IEEE EUI-64:            002538 d91148b62e
Local Time is:                      Thu Jun  2 12:43:04 2022 EDT
Firmware Updates (0x16):            3 Slots, no Reset required
Optional Admin Commands (0x0017):   Security Format Frmw_DL Self_Test
Optional NVM Commands (0x0055):     Comp DS_Mngmt Sav/Sel_Feat Timestmp
Log Page Attributes (0x0f):         S/H_per_NS Cmd_Eff_Lg Ext_Get_Lg Telmtry_Lg
Maximum Data Transfer Size:         512 Pages
Warning  Comp. Temp. Threshold:     82 Celsius
Critical Comp. Temp. Threshold:     85 Celsius
Namespace 1 Features (0x10):        NP_Fields

Supported Power States
St Op     Max   Active     Idle   RL RT WL WT  Ent_Lat  Ex_Lat
 0 +     5.24W       -        -    0  0  0  0        0       0
 1 +     4.49W       -        -    1  1  1  1        0       0
 2 +     2.19W       -        -    2  2  2  2        0     500
 3 -   0.0500W       -        -    3  3  3  3      210    1200
 4 -   0.0050W       -        -    4  4  4  4     1000    9000

Supported LBA Sizes (NSID 0x1)
Id Fmt  Data  Metadt  Rel_Perf
 0 +     512       0         0

=== START OF SMART DATA SECTION ===
SMART overall-health self-assessment test result: PASSED

SMART/Health Information (NVMe Log 0x02)
Critical Warning:                   0x00
Temperature:                        35 Celsius
Available Spare:                    100%
Available Spare Threshold:          10%
Percentage Used:                    2%
Data Units Read:                    48,448,854 [24.8 TB]
Data Units Written:                 51,057,613 [26.1 TB]
Host Read Commands:                 346,050,960
Host Write Commands:                615,522,965
Controller Busy Time:               1,774
Power Cycles:                       381
Power On Hours:                     589
Unsafe Shutdowns:                   100
Media and Data Integrity Errors:    2
Error Information Log Entries:      2
Warning  Comp. Temperature Time:    333
Critical Comp. Temperature Time:    0
Temperature Sensor 1:               35 Celsius
Temperature Sensor 2:               42 Celsius
Thermal Temp. 2 Transition Count:   15877
Thermal Temp. 2 Total Time:         15275

Error Information (NVMe Log 0x01, 16 of 64 entries)
No Errors Logged

Did a Btrfs check from an arch iso here is the result
http://dpaste.com/9UUE7GWMM

1 Like

WOW so you must have frequent power outages?
I have 6,137 Power On Hours and 41 unsafe shutdowns, you already have 100 unsafe shutdowns in the first 589 Power On Hours. That’s not very good.

Now the part under that is directly related:

Media and Data Integrity Errors: 2

That is often a bad thing. It means that it was unable to finish writing something from the disk’s cache to disk. The direct implication is usually that there is lost data. It may not be data that you care about, like something logging that it’s about to shutdown, but it could have been important files.

You can try to avoid this in the future by doing a sync before shutting down. I have an ugly alias for this which is useful, especially for slow flash storage you want to remove safely: alias watchsync='watch -n 1 grep -e Dirty: -e Writeback: /proc/meminfo The issue with relying on sync alone is that the disk controller will do it’s own thing with the cache on a controller level (especially NVMe that relies on cache for speed), and unless you’ve got an enterprise disk that has capacitors on board for reserve power, the power loss will equal data loss for things that have not moved from volatile (cache) to non-volatile storage. You might wonder why we have the ability to lose data so easily, it’s because it’s so much faster to use volatile storage this way. As you can see in that link, it’s a bad idea to turn off your volatile write cache, so your best bet is to sync before shutdowns if you’ve just committed a significant amount of data to disk, and hope the write cache properly flushes before you reach a shutdown state.

This is a good result, but it’s preferred to use Garuda’s PrivateBin for sharing pastes :slight_smile:


The next step is probably using btrfs-scrub, and lucky for us dalto made btrfs assistant and it’s installed by default, though you might want btrfs-assistant-git. I would prefer that someone with more experience than myself confirmmed that scrubbing is a good next step before you spend time doing it. It should be safe to do, moreso considering the check you did came out clean, but please don’t just take my word for it.


Maybe try Rocky / PopOS / MX as well?

4 Likes

yeah I did it from a terminal interface and didn’t know how to upload to privatebin from there so I just used wgetpaste

as for

that’s what happens when nvidia brings down your system after a few hours of playing and you have to force shutdown. that’s the main reason i decided to switch to the 6700xt instead of trying to fix my 1080 fans i thought and drivers would give me less headache (well there are in the sense its not bringing down my whole system lol)

I wiped the drive and reinstalled garuda though so i don’t think i have any data loss and i think those errors will stay there even though I reformated my drive.

anyway yeah it may be interesting to try a few other distros to see if it works better there. currently I’m looking at nobara since that seems to be the closest to what i want.

anyway another idea i had is trying the git mesa drivers but i get this when I try to install them

sudo pacman -S mesa-tkg-git lib32-mesa-tkg-git
resolving dependencies...
looking for conflicting packages...
:: mesa-tkg-git and mesa are in conflict. Remove mesa? [y/N] y
:: mesa-tkg-git and vulkan-intel are in conflict. Remove vulkan-intel? [y/N] y
:: mesa-tkg-git and vulkan-radeon are in conflict. Remove vulkan-radeon? [y/N] y
:: mesa-tkg-git and libva-mesa-driver are in conflict. Remove libva-mesa-driver? [y/N] y
:: mesa-tkg-git and mesa-vdpau are in conflict. Remove mesa-vdpau? [y/N] y
:: lib32-mesa-tkg-git and lib32-mesa are in conflict. Remove lib32-mesa? [y/N] y
:: lib32-mesa-tkg-git and lib32-vulkan-intel are in conflict. Remove lib32-vulkan-intel? [y/N] y
:: lib32-mesa-tkg-git and lib32-vulkan-radeon are in conflict. Remove lib32-vulkan-radeon? [y/N] y
:: lib32-mesa-tkg-git and lib32-libva-mesa-driver are in conflict. Remove lib32-libva-mesa-driver? [y/N] y
:: lib32-mesa-tkg-git and lib32-mesa-vdpau are in conflict. Remove lib32-mesa-vdpau? [y/N] y
:: lib32-mesa-tkg-git and lib32-vulkan-mesa-layers are in conflict (lib32-vulkan-mesa-layer). Remove lib32-vulkan-mesa-layers? [y/N] y
:: mesa-tkg-git and vulkan-mesa-layers are in conflict (vulkan-mesa-layer). Remove vulkan-mesa-layers? [y/N] y
error: failed to prepare transaction (could not satisfy dependencies)
:: removing vulkan-mesa-layers breaks dependency 'vulkan-mesa-layers' required by garuda-video-linux-config
:: removing lib32-vulkan-mesa-layers breaks dependency 'lib32-vulkan-mesa-layers' required by garuda-video-linux-config

EDIT: from searching the forms it seems I need to sudo pacman -R garuda-video-linux-config then I can install mesa-tkg-git

If it’s only fans to fix, you can do it yourself for $20 including cost of tiny screwdriver! Very sad to see so many crashes :frowning: .

If you fully wiped the drive, then that’s good and you can ignore the errors. They will not reset, and that’s okay. It’s the nature of them intentionally.

Probably do not yet, we’re currently changing those around, as llvm-git is required, and is kinda broken atm. We’re working on that, see and watch progress here: [Outdated] mesa-git Β· Issue #1450 Β· chaotic-aur/packages Β· GitHub

EDIT: tkg build should not be broken

2 Likes

tkg git pakages didn't help. at this point I think I may not be worth the trouble anymore, as i have two options that minimize stutter considerably to the point where its not even really visible on the mangohud frame graph it just seems to stutter (other options the stutter is very visible and undeniable) (and could be that the stuttering in these two options is even normal and I'm just being to picky cause i got a new GPU and expected it to be 100% bug free which i guess is not realistic lol)

OPTION 1

Using gamemode (and no gamescope) I can disable the other monitor and that mostly fixes the stuttering. I still get some but I'm thinking maybe that's normal as it happens when the GPU usage goes above 80% the only problem with this is that disable my monitor and then renableing it latter causes some glitches with plasma like the desktop or plasmoid sometimes vanishing so its not very elegant.


OPTION 2

the other solution is to use gamescope and then use gamemode to set tearfree to ON for my TV and then OFF for my other monitor. that allows me to use both monitors and doesn't have the kde glitches i mentioned above though it seems to stutter a little bit more than option 1 (again might just be)


if anything the perceived performance loss from option 2 could also be from either

  1. from game scope not having cap_sys_nice=eip but like I said in a above comment that breaks steam input

  2. I am using GitHub - catsout/wallpaper-engine-kde-plugin: A kde wallpaper plugin integrating wallpaper engine which I'm not sure but may be using the GPU a bit ( couldn't find an alt to nvtop to list programs that are using the gpu. in option one that gets disabled since I only have one screen and it has an option to pause the wallpaper when a window is active. On option 2 this does not work possible because the game is on a different monitor than the wallpaper, as i can see the wallpaper still animating while i am gaming with option 2.

EDIT: I thought that maybe wallpaper engine would make some impact on my game performance but it was the sole cause for my stutters when using option 2. I updated to the latest git of the wallpaper plugin which has support for disableing the wallpaper when a window is in focus on another screen and wow GPU usage went down almost 20%

  • Can you reproduce this using a different secondary monitor?
  • Can you give the output of xrandr --listmonitors --verbose
  • This should help with plasma being weird after a new monitor connects if you have latte.. latte-dock --replace &
3 Likes

well I don't have any other monitors so can't rule out an issue with my Samsung q60 ( I have had plenty of issues (ranging from flickering to audio glitches and artifacts at one point I went through several cables trying to diagnose) using this tv with my 1080 so wouldn't surprise me its doing something funny here)

here is the output of the xrandr Garuda's PrivateBin I was actually looking for a way to get detailed info on my display as I found this issue by reading the discussions page of the AMDGPU archwiki Setting the amdgpu HDMI Pixel Format on Linux - Wesley Moore

in here that guy mentions he gets some perceived latency due to possible conversion from YCbCr to RGB so it caught my attention since it kinda relates back to the previous signal issues i would have with my 1080 above.

as for plasma weirdness yes I am using latte dock so that command may be helpful to add to my workaround scripts where i disable the display though since yesterday I'm not using them anymore. this is all I need to get a stutter free gaming experience

───────┬─────────────────────────────────────────────────────────────────────────────────────────────────
       β”‚ File: Scripts/setTearFree.sh
       β”‚ Size: 166 B
───────┼─────────────────────────────────────────────────────────────────────────────────────────────────
   1   β”‚ #!/bin/sh
   2   β”‚ xrandr --output HDMI-A-0 --set TearFree on
   3   β”‚ echo "HDMI set TearFree on"
   4   β”‚ xrandr --output DisplayPort-0 --set TearFree off
   5   β”‚ echo "DisplayPort set TearFree off"

which I run on gamemode start

and then

───────┬─────────────────────────────────────────────────────────────────────────────────────────────────
       β”‚ File: Scripts/revertTearFree.sh
       β”‚ Size: 172 B
───────┼─────────────────────────────────────────────────────────────────────────────────────────────────
   1   β”‚ #!/bin/sh
   2   β”‚ xrandr --output HDMI-A-0 --set TearFree auto
   3   β”‚ echo "HDMI set TearFree auto"
   4   β”‚ xrandr --output DisplayPort-0 --set TearFree auto
   5   β”‚ echo "DisplayPort set TearFree auto"

on gamemode end. that along with using gamescope and

and the latest wallpaper-engine-kde-plugin built from git ( unfortantly all the aur versions are not up to date see here ) as leaving an animated wallpaper running while gaming was actually making a huge difference.

like this i been gaming flawlessly

prob could do the samething in wayland and i wouldn't need those scripts there but a few apps I use still don't work in wayland (like mangohud keyboard shortcuts)

EDIT: tbh don't know if its totally nessicary to set tear free to off for my non gaming display but the theory there was it was vsyncing to my ultrawide refresh rate and not to my 4k tv where I game and again wayland should theroticly fix this for me

1 Like

Even a TV might be enough to confirm the issue, but I could only hope as much.
Can you sync your refresh rates, or find the lowest common denominator? If freesync is supported can you disable it display side?
EDIT: Nothing in xrandr log looks too weird or bad.

1 Like

yeah the problem is if you look at the xrandr output on my tv 60hz is 60.0 but on my ultrawide its 59.98 so there is a slight variation. that's why I tough it might be syncing to the wrong display.

that's why I only enable tear free on my tv and disable it on my monitor the hope is it would sync to my tv 60.0 and not the monitors 59.98
EDIT:
and it seems to be working

the other thing was my animated wallpaper was eating up some gpu rescorces (surprisingly a lot saw a difference of 20% in GPU usage

EDIT2: oh yeah and about freesynce this tv doesn't support it unfortently (even though the 50 inch of this same model does so kinda dumb the 43 inch doesn't ) my monitor does support freesync but I have it disabled since I don't game there ( I just love the comfort a TV gives you :sweat_smile:

If curious this is the exact model of my TV was a pretty good budget(ish) tv back in 2019 ( though horrible PC support) 2019 QLED 4K Q60R 43" - Specs & Price | Samsung US

1 Like

The reason I ask about Freesync is that there are many issues you might not expect and can definitely trip you, even silly things about it - like it only working over DisplayPort right now Variable refresh rate - ArchWiki.

That slight variation shouldn't actually be an issue there, while they're not exactly the same, they're effectively identical. I see the specs of that TV say '4K AI Upscaling', is there a way to turn that off for testing?

Samsung can make an alright display, but their support definitely sucks. The smart tv's are basically spyware at this point, especially since many manufacturers tend to make as much or more from ads than the tvs.

ahh so even freesync is not all peaches and roses it even says freesynch with a multi monitor setup does not work on X.

as for the '4k ai upsacling` I have no clue what they mean by that and see no option i wouldn’t be surprised if there β€˜4k ai upsacling’ is just the regualar awfull upscaling all displays do :face_with_raised_eyebrow:

EDIT:
oh and about it being praticly spyware i totaly agree and the β€œsmart features” of samsung (at least on this tv sucks (no apps horrible internet conection even over ethernet) so I use a nvidia shield for multi-media

1 Like

This topic was automatically closed 2 days after the last reply. New replies are no longer allowed.