Hi,
I’m running a Lunar Lake 130V using the xe driver and I’m seeing a huge drop in performance when gaming. Games I get 120fps in other distros are getting 40 in Tumbleweed,
I’ve tried Fedora, Arch, Gentoo and they all get the higher framerate, it’s just oSUS, I can’t get any debug output at the moment from oSUS as the download site is down and I don’t have an iso handy to reformat but I will as soon as I can. Does anyone have ideas in the meantime?
Thank you!
It would be helpful to get output from the environment of interest. When you do have it installed again, please share the output from
inxi -GSaz
glxinfo | grep "OpenGL renderer"
Also zypper se -si mesaresults.
Looks like the site came back up, I will grab this shortly.
Clean installed with KDE on Wayland, same issue, about 1/3rd of the framerate that I’m seeing in other distros, tried on the same Kernel/Mesa.
test@localhost:~> inxi -GSaz
System:
Kernel: 6.19.8-1-default arch: x86_64 bits: 64 compiler: gcc v: 15.2.1
clocksource: tsc avail: acpi_pm
parameters: initrd=\opensuse-tumbleweed\6.19.8-1-default\initrd-ea27e233b380063260eafa073b7ac8ec81371cd0
root=/dev/mapper/cr_root splash=silent mitigations=auto quiet
security=selinux selinux=1
systemd.machine_id=0628b6c9c05f415cb8cb8142d66fb233
Desktop: KDE Plasma v: 6.6.2 tk: Qt v: N/A info: frameworks v: 6.24.0
wm: kwin_wayland tools: avail: xscreensaver vt: 3 dm: SDDM Distro: openSUSE
Tumbleweed 20260318
Graphics:
Device-1: Intel Lunar Lake [Intel Arc Graphics 130V / 140V] vendor: Lenovo
driver: xe v: kernel arch: Xe2 process: TSMC n3 (3nm) built: 2024+ ports:
active: eDP-1 empty: DP-1,DP-2,HDMI-A-1 bus-ID: 00:02.0 chip-ID: 8086:64a0
class-ID: 0300
Display: wayland server: X.org v: 1.21.1.21 with: Xwayland v: 24.1.9
compositor: kwin_wayland driver: X: loaded: modesetting unloaded: vesa
alternate: fbdev,intel dri: iris gpu: xe display-ID: 0
Monitor-1: eDP-1 model: Samsung ATNA53JB01-0 built: 2024 res:
mode: 2880x1800 hz: 120 scale: 145% (1.45) to: 1986x1241 dpi: 222 gamma: 1.2
size: 330x210mm (12.99x8.27") diag: 391mm (15.4") modes: 2880x1800
API: EGL v: 1.5 hw: drv: intel iris platforms: device: 0 drv: iris
device: 1 drv: swrast gbm: drv: iris surfaceless: drv: iris wayland:
drv: iris x11: drv: iris
API: OpenGL v: 4.6 compat-v: 4.5 vendor: intel mesa v: 26.0.2 glx-v: 1.4
direct-render: yes renderer: Mesa Intel Graphics (LNL) device-ID: 8086:64a0
memory: 30.14 GiB unified: yes display-ID: :1.0
API: Vulkan v: 1.4.341 layers: 1 device: 0 type: integrated-gpu
name: Intel Graphics (LNL) driver: mesa intel v: 26.0.2 device-ID: 8086:64a0
surfaces: N/A device: 1 type: cpu name: llvmpipe (LLVM 21.1.8 256 bits)
driver: mesa llvmpipe v: 26.0.2 (LLVM 21.1.8) device-ID: 10005:0000
surfaces: N/A
Info: Tools: api: eglinfo, glxinfo, vulkaninfo
de: kscreen-console,kscreen-doctor wl: wayland-info
x11: xdpyinfo, xprop, xrandr
test@localhost:~> glxinfo | grep -i "opengl renderer"
OpenGL renderer string: Mesa Intel(R) Graphics (LNL)
test@localhost:~> zypper se -si mesa
S | Name | Type | Version | Arch | Repository
---+---------------------------+---------+------------+--------+------------------------
i | Mesa | package | 26.0.2-1.1 | x86_64 | openSUSE-Tumbleweed-Oss
i | Mesa-demo-egl | package | 9.0.0-7.2 | x86_64 | openSUSE-Tumbleweed-Oss
i | Mesa-demo-x | package | 9.0.0-7.2 | x86_64 | openSUSE-Tumbleweed-Oss
i | Mesa-dri | package | 26.0.2-1.1 | x86_64 | openSUSE-Tumbleweed-Oss
i | Mesa-libEGL1 | package | 26.0.2-1.1 | x86_64 | openSUSE-Tumbleweed-Oss
i | Mesa-libGL1 | package | 26.0.2-1.1 | x86_64 | openSUSE-Tumbleweed-Oss
i | Mesa-libva | package | 26.0.2-1.1 | x86_64 | openSUSE-Tumbleweed-Oss
i | Mesa-vulkan-device-select | package | 26.0.2-1.1 | x86_64 | openSUSE-Tumbleweed-Oss
test@localhost:~> lsmod | grep xe
xe 4218880 96
drm_ttm_helper 20480 1 xe
ttm 131072 2 drm_ttm_helper,xe
i2c_algo_bit 24576 1 xe
drm_suballoc_helper 16384 1 xe
drm_buddy 32768 1 xe
video 81920 2 thinkpad_acpi,xe
gpu_sched 73728 1 xe
drm_gpuvm 57344 1 xe
drm_exec 12288 2 drm_gpuvm,xe
drm_gpusvm_helper 40960 1 xe
drm_display_helper 303104 1 xe
cec 102400 2 drm_display_helper,xe
intel_vsec 28672 3 intel_pmc_ssram_telemetry,pmt_telemetry,xe
Ok, at a quick glance that all checks out as expected. Do you get the same kind of frame rate performance using a Plasma X11 session? Try that and report back.
Firmware messaging may also be useful…
sudo dmesg | grep -iE "xe|guc|huc|firmware"
Graphics performance can vary depending on whether a game is using Vulkan or OpenGL, and Vulkan will generally perform better on newer hardware. Which games are noticeably slower on openSUSE, and are they running via the Steam platform?
I can try X11 when I get a spare second, I’m running Hades 2 on Proton Experimental, through Steam, I’ve tried the native client and Distrobox (no difference in perf there).
Nothing that interesting in dmesg:
test@localhost:~> sudo dmesg | grep -iE "xe|guc|huc|firmware"
[sudo] password for root:
[ 0.000000] [ T0] NX (Execute Disable) protection: active
[ 0.000012] [ T0] MTRR map: 6 entries (3 fixed + 3 variable; max 23), built from 10 variable MTRRs
[ 0.044497] [ T1] ... fixed-purpose counters: 4
[ 0.044497] [ T1] ... fixed-purpose bitmap: 000000000000000f
[ 0.007296] [ T0] ... fixed-purpose counters: 6
[ 0.007296] [ T0] ... fixed-purpose bitmap: 0000000000000077
[ 0.208484] [ T1] ACPI: [Firmware Bug]: BIOS _OSI(Linux) query ignored
[ 0.367611] [ T1] pci 0000:00:1f.4: BAR 4 [io 0xefa0-0xefbf]
[ 0.559415] [ T1] RAPL PMU: API unit is 2^-32 Joules, 3 fixed counters, 655360 ms ovfl timer
[ 0.730520] [ T1] Relocating firmware framebuffer to offset 0x0000000000800000[d] within [mem 0x2000000000-0x200fffffff flags 0x14220c]
[ 0.862710] [ T1] evm: security.SMACK64EXEC (disabled)
[ 1.152697] [ T1] systemd[1]: /usr/lib/systemd/system/plymouth-start.service:15: Unit uses KillMode=none. This is unsafe, as it disables systemd's process lifecycle management for the service. Please update the service to use a safer KillMode=, such as 'mixed' or 'control-group'. Support for KillMode=none is deprecated and will eventually be removed.
[ 1.587009] [ T531] wmi_bus wmi_bus-PNP0C14:09: [Firmware Bug]: WQ,\x1e data block query control method not found
[ 1.606579] [ T499] intel_vpu 0000:00:0b.0: [drm] Firmware: intel/vpu/vpu_40xx_v1.bin, version: Nov 13 2025*NPU40xx*ci_tag_ud202548_vpu_rc_20251112_1901-1-g72f907ffc78*72f907ffc780df5579a2fed65afc4944da8b0e44
[ 1.756397] [ T527] xe 0000:00:02.0: vgaarb: deactivate vga console
[ 1.756946] [ T527] xe 0000:00:02.0: [drm] Found lunarlake (device ID 64a0) integrated display version 20.00 stepping B0
[ 1.760022] [ T527] xe 0000:00:02.0: vgaarb: VGA decodes changed: olddecodes=io+mem,decodes=none:owns=mem
[ 1.777268] [ T11] xe 0000:00:02.0: [drm] Finished loading DMC firmware i915/xe2lpd_dmc.bin (v2.29)
[ 1.811208] [ T527] xe 0000:00:02.0: [drm] Tile0: GT0: Using GuC firmware from xe/lnl_guc_70.bin version 70.58.0
[ 2.175323] [ T63] intel_ish_ipc 0000:00:12.0: ISH loader: load firmware: intel/ish/ish_lnlm.bin
[ 2.466044] [ T527] xe 0000:00:02.0: [drm] Tile0: GT0: ccs1 fused off
[ 2.466047] [ T527] xe 0000:00:02.0: [drm] Tile0: GT0: ccs2 fused off
[ 2.466048] [ T527] xe 0000:00:02.0: [drm] Tile0: GT0: ccs3 fused off
[ 2.490434] [ T527] xe 0000:00:02.0: [drm] Tile0: GT1: Using GuC firmware from xe/lnl_guc_70.bin version 70.58.0
[ 2.500842] [ T527] xe 0000:00:02.0: [drm] Tile0: GT1: Using HuC firmware from xe/lnl_huc.bin version 9.4.13
[ 2.506829] [ T527] xe 0000:00:02.0: [drm] Tile0: GT1: Using GSC firmware from xe/lnl_gsc_1.bin version 104.0.5.1429
[ 2.528499] [ T527] xe 0000:00:02.0: [drm] Tile0: GT1: vcs1 fused off
[ 2.528503] [ T527] xe 0000:00:02.0: [drm] Tile0: GT1: vcs2 fused off
[ 2.528504] [ T527] xe 0000:00:02.0: [drm] Tile0: GT1: vcs3 fused off
[ 2.528504] [ T527] xe 0000:00:02.0: [drm] Tile0: GT1: vcs4 fused off
[ 2.528505] [ T527] xe 0000:00:02.0: [drm] Tile0: GT1: vcs5 fused off
[ 2.528506] [ T527] xe 0000:00:02.0: [drm] Tile0: GT1: vcs6 fused off
[ 2.528506] [ T527] xe 0000:00:02.0: [drm] Tile0: GT1: vcs7 fused off
[ 2.528507] [ T527] xe 0000:00:02.0: [drm] Tile0: GT1: vecs1 fused off
[ 2.528508] [ T527] xe 0000:00:02.0: [drm] Tile0: GT1: vecs2 fused off
[ 2.528508] [ T527] xe 0000:00:02.0: [drm] Tile0: GT1: vecs3 fused off
[ 2.600203] [ T527] xe 0000:00:02.0: [drm] Registered 3 planes with drm panic
[ 2.600208] [ T527] [drm] Initialized xe 1.1.0 for 0000:00:02.0 on minor 0
[ 2.617757] [ T527] fbcon: xedrmfb (fb0) is primary device
[ 2.698304] [ T12] xe 0000:00:02.0: [drm] Tile0: GT1: found GSC cv104.1.0
[ 3.822923] [ T527] xe 0000:00:02.0: [drm] Reducing the compressed framebuffer size. This may lead to less power savings than a non-reduced-size. Try to increase stolen memory size if available in BIOS.
[ 3.875194] [ T527] xe 0000:00:02.0: [drm] fb0: xedrmfb frame buffer device
[ 13.989390] [ T1] SELinux: Permission firmware_load in class system not defined in policy.
[ 13.989396] [ T1] SELinux: Permission kexec_image_load in class system not defined in policy.
[ 13.989397] [ T1] SELinux: Permission kexec_initramfs_load in class system not defined in policy.
[ 13.990844] [ T1] SELinux: policy capability ioctl_skip_cloexec=0
[ 14.324405] [ T1] systemd[1]: /usr/lib/systemd/system/plymouth-start.service:15: Unit uses KillMode=none. This is unsafe, as it disables systemd's process lifecycle management for the service. Please update the service to use a safer KillMode=, such as 'mixed' or 'control-group'. Support for KillMode=none is deprecated and will eventually be removed.
[ 14.403174] [ T1] systemd[1]: Set up automount Arbitrary Executable File Formats File System Automount Point.
[ 14.417663] [ T1] systemd[1]: Clear Stale Hibernate Storage Info skipped, unmet condition check ConditionPathExists=/sys/firmware/efi/efivars/HibernateLocation-8cf2644b-4b0b-428f-9387-6d876050dc67
[ 14.740554] [ T124] faux_driver regulatory: Direct firmware load for regulatory.db failed with error -2
[ 14.768037] [ T1677] intel-ipu7 0000:00:05.0: firmware cpd file: intel/ipu/ipu7_fw.bin
[ 14.822247] [ T49] iwlwifi 0000:00:14.3: loaded firmware version 101.6e695a70.0 bz-b0-fm-c0-c101.ucode op_mode iwlmld
[ 14.852023] [ T126] Bluetooth: hci0: Minimum firmware build 1 week 10 2014
[ 14.858873] [ T126] Bluetooth: hci0: Found device firmware: intel/ibt-0190-0291-iml.sfi
[ 14.858878] [ T126] Bluetooth: hci0: Firmware Version: 15-3.26
[ 14.910844] [ T1687] mei_gsc_proxy 0000:00:16.0-0f73db04-97ab-4125-b893-e904ad0d5464: bound 0000:00:02.0 (ops xe_gsc_proxy_component_ops [xe])
[ 14.951377] [ T1662] sof-audio-pci-intel-lnl 0000:00:1f.3: bound 0000:00:02.0 (ops intel_audio_component_bind_ops [xe])
[ 14.991125] [ T90] sof-audio-pci-intel-lnl 0000:00:1f.3: Firmware paths/files for ipc type 1:
[ 14.991130] [ T90] sof-audio-pci-intel-lnl 0000:00:1f.3: Firmware file: intel/sof-ipc4/lnl/sof-lnl.ri
[ 14.991130] [ T90] sof-audio-pci-intel-lnl 0000:00:1f.3: Firmware lib path: intel/sof-ipc4-lib/lnl
[ 14.991582] [ T90] sof-audio-pci-intel-lnl 0000:00:1f.3: Loaded firmware library: ADSPFW, version: 2.14.1.1
[ 15.014991] [ T126] Bluetooth: hci0: Waiting for firmware download to complete
[ 15.015201] [ T126] Bluetooth: hci0: Firmware loaded in 152658 usecs
[ 15.051113] [ T126] Bluetooth: hci0: Found device firmware: intel/ibt-0190-0291-usb.sfi
[ 15.051123] [ T126] Bluetooth: hci0: Firmware Version: 15-3.26
[ 15.273277] [ T90] sof-audio-pci-intel-lnl 0000:00:1f.3: Booted firmware version: 2.14.1.1
[ 15.793261] [ T762] sof-audio-pci-intel-lnl 0000:00:1f.3: Loaded firmware library: ADSPFW, version: 2.14.1.1
[ 16.845876] [ T126] Bluetooth: hci0: Waiting for firmware download to complete
[ 16.846208] [ T126] Bluetooth: hci0: Firmware loaded in 1753020 usecs
[ 16.906221] [ T126] Bluetooth: hci0: Firmware timestamp 2026.3 buildtype 1 build 9743
[ 16.906233] [ T126] Bluetooth: hci0: Firmware SHA1: 0xd68504ef
[ 16.910238] [ T126] Bluetooth: hci0: Fseq executed: 00.00.04.202
[ 17.332778] [ T12] cs35l56 spi-cs35l56-left: DSP1: Firmware: 1a00d6 vendor: 0x2 v3.11.20, 41 algorithms
[ 18.560090] [ T99] cs35l56 spi-cs35l56-right: DSP1: Firmware: 1a00d6 vendor: 0x2 v3.11.20, 41 algorithms
I didn’t see any difference in perf running under X11, either.
Ok, this suggests not an API or compositor issue at least.
Run sudo intel_gpu_top while playing a game, to check GPU utilization.
If it is reasonably low when playing games but FPS is not good, it may pay to check the CPU governor settings…
cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor
If it shows powersave, try changing to performance…
sudo cpupower frequency-set -g performance
It was boosting up to 1950mhz, I did notice CPU per-core utilization was quite low, just checking on other distros now to see what it looks like.
Will check the cpu power too, thank you!
Watch the CPU frequency under load: watch -n1 "grep MHz /proc/cpuinfo"
When you compare with other distributions, it would be useful to check CPU frequency under load as well as the governor. That can affect how quickly work is fed to the GPU (even if overall CPU usage appears low).
There is no such thing here.
Neither is it funny.
I checked and download.o.o is and was not down. But the info you provide is just anecdotic, so we have nothing to check.
Sorry, I’m confused here: I’m not sure why it would be?
I checked and download.o.o is and was not down. But the info you provide is just anecdotic, so we have nothing to check.
It wasn’t down for long, the main site was up, https://get.opensuse.org was returning errors.
Ah, well, I never use that. If I need the latest TW for some reason I do:
knurpht@Lenovo-P16:~/Install/ISO> wget https://download.opensuse.org/tumbleweed/iso/openSUSE-Tumbleweed-DVD-x86_64-Current.iso
I probably should’ve done the same! I will do some more testing on the laptop tomorrow, I’ve tried to cover all bases with the distros I’ve checked but I spent a lot of my day installing and comparing, it seems to not be related to kernel, mesa version, or whether it’s on x/wayland. I’ll try some other games too and check the frequencies like you said. Thanks for the help so far.
Nothing looks off with the cpu frequencies, I tried setting everything to performance and nothing changed, I did see with vulkaninfo a second GPU named llvmpipe was present. This seems wrong to me but it’s not being used in game.
It looks like the GPU isn’t hitting full usage, however intel_gpu_top doesn’t work as I guess it’s only for i915. Unfortunately I might have to leave it at this as I’ve spent a lot of time already trying to figure it out. Thanks a lot for your input though, I think it might require a bit more skill to investigate than I have right now.
No, I think that is normal. The llvmpipe device (reported by vulkaninfo) is just a software renderer provided by Mesa and will always show up alongside the real GPU.
We eliminated a few things at least. I’m at the limits of my knowledge with this too (and I’m not a gamer). Hopefully users with similar hardware can chime in with further advice.
@Shortwave8115 nvtop has xe support, I have a couple of ARC (A310 and A380) setups one works fine on the xe driver.
It’s strange, because I don’t see that llvmpipe device on other distros. Even on other distros the performance isn’t great, but for some reason OpenSUSE Tumbleweed doesn’t even hit 50% of that. I haven’t tried in other applications but I think it might be GPU related, I’ll see if there’s CPU benchmarks I can try.
Thanks, I looked at nvtop briefly and didn’t see anything odd. I’ll see what else I can do to try and debug it, but I think Lunar Lake in general is still quite new in terms of Linux support.
There are the likes of furmark, vkpeak and mprime to name a few…