Segmentation fault on openSUSE Tumbleweed

[kpatzwald]

Hi,

I installed linux-npu-driver with zypper on openSUSE Tumbleweed. When I run npu-umd-test I get a segmentation error.

My goal is to setup local-ai as docker-image. When I start the docker container I get this message:

Intel GPU detected
Intel GPU detected, but Intel GPU drivers are not installed. GPU acceleration will not be available.

I am not sure if I need the npu-driver.

My system:
Zenbook UX3405MA with Intel(R) Core(TM) Ultra 9 185H

System:
  Kernel: 6.19.5-1-default arch: x86_64 bits: 64 compiler: gcc v: 15.2.1
    clocksource: tsc avail: acpi_pm
    parameters: BOOT_IMAGE=/opensuse-tumbleweed/6.19.5-1-default/linux-5c84a20dba054adf62623f78c02d0dea6c9f7905
    splash=silent resume=/dev/system/swap mitigations=auto quiet
    security=selinux selinux=1 i915.force_probe=!7d55 xe.force_probe=7d55
    root=UUID=814e1d91-0970-43e4-9b7a-fafb0c95e0ec
    rootflags=subvol=@/.snapshots/81/snapshot
  Desktop: KDE Plasma v: 6.6.2 tk: Qt v: N/A info: frameworks v: 6.23.0
    wm: kwin_wayland with: docker,krunner tools: avail: xscreensaver vt: 3
    dm: SDDM Distro: openSUSE Tumbleweed 20260304
Graphics:
  Device-1: Intel Meteor Lake-P [Intel Arc Graphics] vendor: ASUSTeK
    driver: xe v: kernel alternate: i915 arch: Xe-LPG process: Intel 4 (7nm+)
    built: 2023+ ports: active: eDP-1 empty: DP-1,DP-2,HDMI-A-1
    bus-ID: 0000:00:02.0 chip-ID: 8086:7d55 class-ID: 0300
  Device-2: Shinetech USB2.0 FHD UVC WebCam driver: uvcvideo type: USB
    rev: 2.0 speed: 480 Mb/s lanes: 1 mode: 2.0 bus-ID: 3-9:2 chip-ID: 3277:0055
    class-ID: fe01 serial: <filter>
  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 0x419d built: 2022 res: mode: 2880x1800
    hz: 120 scale: 175% (1.75) to: 1646x1029 dpi: 242 gamma: 1.2
    size: 302x189mm (11.89x7.44") diag: 356mm (14") ratio: 16:10
    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: 26.0.1 glx-v: 1.4
    direct-render: yes renderer: Mesa Intel Arc Graphics (MTL)
    device-ID: 8086:7d55 memory: 29.99 GiB unified: yes display-ID: :1.0
  API: Vulkan v: 1.4.341 layers: 5 device: 0 type: integrated-gpu name: Intel
    Arc Graphics (MTL) driver: mesa intel v: 26.0.1 device-ID: 8086:7d55
    surfaces: N/A device: 1 type: cpu name: llvmpipe (LLVM 21.1.8 256 bits)
    driver: mesa llvmpipe v: 26.0.1 (LLVM 21.1.8) device-ID: 10005:0000
    surfaces: N/A
  Info: Tools: api: clinfo, eglinfo, glxinfo, vulkaninfo
    de: kscreen-console,kscreen-doctor gpu: gputop, intel_gpu_top, lsgpu
    wl: wayland-info x11: xdpyinfo, xprop, xrandr

Test version npu-linux-driver-dev-1.28.0.dev-0 on openSUSE Tumbleweed (6.19.5-1-default)
[==========] Running 220 tests from 39 test suites.
[----------] Global test environment set-up.
[1]    7658 segmentation fault (core dumped)  npu-umd-test

Hint: You are currently not seeing messages from other users and the system.
      Users in the 'systemd-journal' group can see all messages. Pass -q to
      turn off this notice.
           PID: 7658 (npu-umd-test)
           UID: 1000 (kay)
           GID: 1000 (kay)
        Signal: 11 (SEGV)
     Timestamp: Sun 2026-03-08 20:17:52 CET (4s ago)
  Command Line: npu-umd-test
    Executable: /usr/bin/npu-umd-test
 Control Group: /user.slice/user-1000.slice/user@1000.service/app.slice/app-org.kde.konsole-4821.scope/tab(4864).scope
          Unit: user@1000.service
     User Unit: app-org.kde.konsole-4821.scope
         Slice: user-1000.slice
     Owner UID: 1000 (kay)
       Boot ID: 4823c32c52fd46e7be3198ec9c1a0b6b
    Machine ID: abb8aa880f7e4228a47077463ecba1c3
      Hostname: ZenbookOpenSUSE
       Storage: /var/lib/systemd/coredump/core.npu-umd-test.1000.4823c32c52fd46e7be3198ec9c1a0b6b.7658.1772997472000000.zst (present)
  Size on Disk: 138.6K
       Message: Process 7658 (npu-umd-test) of user 1000 dumped core.
                
                Stack trace of thread 7658:
                #0  0x00007f4a03b9154b n/a (libze_validation_layer.so.1 + 0x4154b)
                #1  0x00007f4a03b91e49 n/a (libze_validation_layer.so.1 + 0x41e49)
                #2  0x00007f4a046c538e call_init (ld-linux-x86-64.so.2 + 0x638e)
                #3  0x00007f4a046c546c _dl_init (ld-linux-x86-64.so.2 + 0x646c)
                #4  0x00007f4a046c1549 _dl_catch_exception (ld-linux-x86-64.so.2 + 0x2549)
                #5  0x00007f4a046cc598 dl_open_worker (ld-linux-x86-64.so.2 + 0xd598)
                #6  0x00007f4a046c14b6 _dl_catch_exception (ld-linux-x86-64.so.2 + 0x24b6)
                #7  0x00007f4a046cca74 _dl_open (ld-linux-x86-64.so.2 + 0xda74)
                #8  0x00007f4a03e97a1c dlopen_doit (libc.so.6 + 0x97a1c)
                #9  0x00007f4a046c14b6 _dl_catch_exception (ld-linux-x86-64.so.2 + 0x24b6)
                #10 0x00007f4a046c15f3 _dl_catch_error (ld-linux-x86-64.so.2 + 0x25f3)
                #11 0x00007f4a03e97474 _dlerror_run (libc.so.6 + 0x97474)
                #12 0x00007f4a03e97ae9 dlopen@GLIBC_2.2.5 (libc.so.6 + 0x97ae9)
                #13 0x00007f4a04630fe9 _ZN6loader9context_t4initEv (libze_loader.so.1 + 0x80fe9)
                #14 0x00007f4a046117bd n/a (libze_loader.so.1 + 0x617bd)
                #15 0x00007f4a045e9679 n/a (libze_loader.so.1 + 0x39679)
                #16 0x00007f4a03ea1326 __pthread_once_slow.isra.0 (libc.so.6 + 0xa1326)
                #17 0x00007f4a03ea13b9 pthread_once@GLIBC_2.2.5 (libc.so.6 + 0xa13b9)
                #18 0x00007f4a045e3b23 zeInitDrivers (libze_loader.so.1 + 0x33b23)
                #19 0x0000561b1d71d42a n/a (/usr/bin/npu-umd-test + 0x7842a)
                #20 0x00007f4a045897de _ZN7testing8internal12UnitTestImpl11RunAllTestsEv (libgtest.so.1.17.0 + 0x517de)
                #21 0x00007f4a04589b58 _ZN7testing8UnitTest3RunEv (libgtest.so.1.17.0 + 0x51b58)
                #22 0x0000561b1d6fa5d9 n/a (/usr/bin/npu-umd-test + 0x555d9)
                #23 0x00007f4a03e2b2fb __libc_start_call_main (libc.so.6 + 0x2b2fb)
                #24 0x00007f4a03e2b3cb __libc_start_main@@GLIBC_2.34 (libc.so.6 + 0x2b3cb)
                #25 0x0000561b1d711a65 n/a (/usr/bin/npu-umd-test + 0x6ca65)
                ELF object binary architecture: AMD x86-64

I tried this little test-program

(openvino_env) > python3 -c "from openvino import Core; print(Core().available_devices)" 
['CPU', 'GPU']

Any help is appreciated
Thanks
Kay

[jwludzik]

@kpatzwald , can you share a back trace from core dump generated on npu-umd-test segmentation fault?

[kpatzwald]

Sorry for the delay. I had to reinstall my system. Here is the backtrace.

backtrace_npu.txt

[jwludzik]

@kpatzwald , what level zero loader do you use? Can you update it to the latest one and try again?

[kpatzwald]

I have installed this version. Should I install level-zero from source?

Repository : openSUSE-Tumbleweed-Oss
Name : level-zero
Version : 1.28.2-1.1
Arch : x86_64
Anbieter : openSUSE
Installierte Größe : 3,5 MiB
Installiert : Ja (automatisch)
Status : aktuell
Quellpaket : level-zero-1.28.2-1.1.src
Upstream-URL : https://github.com/oneapi-src/level-zero

[jwludzik]

Can you try to rebuild the level-zero with LTO disabled and try again? See https://en.opensuse.org/openSUSE:LTO

 %_lto_cflags %nil

I am observing similar issue in Ubuntu. It is fixed with level-zero-loader 1.27.0 package that has the LTO disabled

[kpatzwald]

I created the following response a few days before but forget to send it. Should I still rebuild the driver with LTO disabled?

Okay, I've made some progress.

It works with sudo npu-umd-test.

The python-script recognizes my NPU.

I've tried some jupyter notbooks from the openVINO project. All tested versions (i.e. Hello NPU) gave the following error:
ZE_RESULT_ERROR_UNSUPPORTED_FEATURE, code 0x78000003 - generic error code for unsupported features .

The generic local-ai (docker) still doesn't accept my GPU.
Intel GPU detected, but Intel GPU drivers are not installed. GPU acceleration will not be available.

I now will tried it with latest-aio-gpu-intel.

I noticed there is a new linux-npu-driver available (1.30) but it is not available from openSUSE right now. I will wait for the update.

[jwludzik]

I am suspecting that issue is not connected directly with linux-npu-driver based on backtrace that you shared:

                #12 0x00007f4a03e97ae9 dlopen@GLIBC_2.2.5 (libc.so.6 + 0x97ae9)
                #13 0x00007f4a04630fe9 _ZN6loader9context_t4initEv (libze_loader.so.1 + 0x80fe9)

It fails on dlopen within libze_loader.so.1. This is the part of level-zero package.

Yes, please try to rebuild level-zero with LTO disabled