WSL2 Ubuntu uses llvmpipe instead of Nvidia GPU (3090)

Gaming and Visualization Technologies General Topics and Other SDKs Drivers - Linux, Windows, MacOS
, ,
1

Hi,

I’m fairly new to linux and trying to set up GPU passthrough in my Windows Subsystem for Linux 2 (WSL2) environment to utilize my 3090 for a GUI. Specifically, I’ve installed the CUDA Toolkit to enable GPU acceleration for both CUDA-based applications and Vulkan/OpenGL rendering inside WSL2. However, no matter what I do, glxinfo shows that llvmpipe is being utilised instead.

Here’s my software versions:
WSL version: 2.3.26.0
Kernel version: 5.15.167.4-1
WSLg version: 1.0.65
MSRDC version: 1.2.5620
Direct3D version: 1.611.1-81528511
DXCore version: 10.0.26100.1-240331-1435.ge-release
Windows version: 10.0.19045.5247

Here’s glxinfo -B output:

name of display: :0
display: :0  screen: 0
direct rendering: Yes
Extended renderer info (GLX_MESA_query_renderer):
    Vendor: Mesa (0xffffffff)
    Device: llvmpipe (LLVM 17.0.6, 256 bits) (0xffffffff)
    Version: 24.3.3
    Accelerated: no
    Video memory: 15945MB
    Unified memory: yes
    Preferred profile: core (0x1)
    Max core profile version: 4.5
    Max compat profile version: 4.5
    Max GLES1 profile version: 1.1
    Max GLES[23] profile version: 3.2
Memory info (GL_ATI_meminfo):
    VBO free memory - total: 0 MB, largest block: 0 MB
    VBO free aux. memory - total: 13998 MB, largest block: 13998 MB
    Texture free memory - total: 0 MB, largest block: 0 MB
    Texture free aux. memory - total: 13998 MB, largest block: 13998 MB
    Renderbuffer free memory - total: 0 MB, largest block: 0 MB
    Renderbuffer free aux. memory - total: 13998 MB, largest block: 13998 MB
Memory info (GL_NVX_gpu_memory_info):
    Dedicated video memory: 0 MB
    Total available memory: 15945 MB
    Currently available dedicated video memory: 0 MB
OpenGL vendor string: Mesa
OpenGL renderer string: llvmpipe (LLVM 17.0.6, 256 bits)
OpenGL core profile version string: 4.5 (Core Profile) Mesa 24.3.3 - kisak-mesa PPA
OpenGL core profile shading language version string: 4.50
OpenGL core profile context flags: (none)
OpenGL core profile profile mask: core profile

OpenGL version string: 4.5 (Compatibility Profile) Mesa 24.3.3 - kisak-mesa PPA
OpenGL shading language version string: 4.50
OpenGL context flags: (none)
OpenGL profile mask: compatibility profile

OpenGL ES profile version string: OpenGL ES 3.2 Mesa 24.3.3 - kisak-mesa PPA
OpenGL ES profile shading language version string: OpenGL ES GLSL ES 3.20

and this is despite nvidia-smi working:

+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 565.77.01              Driver Version: 566.36         CUDA Version: 12.7     |
|-----------------------------------------+------------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id          Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |           Memory-Usage | GPU-Util  Compute M. |
|                                         |                        |               MIG M. |
|=========================================+========================+======================|
|   0  NVIDIA GeForce RTX 3090        On  |   00000000:0B:00.0  On |                  N/A |
|  0%   40C    P8             39W /  390W |    2193MiB /  24576MiB |      1%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+

+-----------------------------------------------------------------------------------------+
| Processes:                                                                              |
|  GPU   GI   CI        PID   Type   Process name                              GPU Memory |
|        ID   ID                                                               Usage      |
|=========================================================================================|
|    0   N/A  N/A       382      G   /Xwayland                                   N/A      |
+-----------------------------------------------------------------------------------------+

vulkaninfo --summary gives the following:

WARNING: dzn is not a conformant Vulkan implementation, testing use only.
==========
VULKANINFO
==========

Vulkan Instance Version: 1.3.275


Instance Extensions: count = 24
-------------------------------
VK_EXT_acquire_drm_display             : extension revision 1
VK_EXT_acquire_xlib_display            : extension revision 1
VK_EXT_debug_report                    : extension revision 10
VK_EXT_debug_utils                     : extension revision 2
VK_EXT_direct_mode_display             : extension revision 1
VK_EXT_display_surface_counter         : extension revision 1
VK_EXT_headless_surface                : extension revision 1
VK_EXT_surface_maintenance1            : extension revision 1
VK_EXT_swapchain_colorspace            : extension revision 5
VK_KHR_device_group_creation           : extension revision 1
VK_KHR_display                         : extension revision 23
VK_KHR_external_fence_capabilities     : extension revision 1
VK_KHR_external_memory_capabilities    : extension revision 1
VK_KHR_external_semaphore_capabilities : extension revision 1
VK_KHR_get_display_properties2         : extension revision 1
VK_KHR_get_physical_device_properties2 : extension revision 2
VK_KHR_get_surface_capabilities2       : extension revision 1
VK_KHR_portability_enumeration         : extension revision 1
VK_KHR_surface                         : extension revision 25
VK_KHR_surface_protected_capabilities  : extension revision 1
VK_KHR_wayland_surface                 : extension revision 6
VK_KHR_xcb_surface                     : extension revision 6
VK_KHR_xlib_surface                    : extension revision 6
VK_LUNARG_direct_driver_loading        : extension revision 1

Instance Layers: count = 3
--------------------------
VK_LAYER_INTEL_nullhw       INTEL NULL HW                1.1.73   version 1
VK_LAYER_MESA_device_select Linux device selection layer 1.3.211  version 1
VK_LAYER_MESA_overlay       Mesa Overlay layer           1.3.211  version 1

Devices:
========
GPU0:
        apiVersion         = 1.2.296
        driverVersion      = 24.3.3
        vendorID           = 0x10de
        deviceID           = 0x2204
        deviceType         = PHYSICAL_DEVICE_TYPE_DISCRETE_GPU
        deviceName         = Microsoft Direct3D12 (NVIDIA GeForce RTX 3090)
        driverID           = DRIVER_ID_MESA_DOZEN
        driverName         = Dozen
        driverInfo         = Mesa 24.3.3 - kisak-mesa PPA
        conformanceVersion = 0.0.0.0
        deviceUUID         = 0df3a697-bd3f-5cf4-61a7-3d2d24cd55fa
        driverUUID         = badcd177-55a9-bd69-9128-926723cdd875
GPU1:
        apiVersion         = 1.3.296
        driverVersion      = 0.0.1
        vendorID           = 0x10005
        deviceID           = 0x0000
        deviceType         = PHYSICAL_DEVICE_TYPE_CPU
        deviceName         = llvmpipe (LLVM 17.0.6, 256 bits)
        driverID           = DRIVER_ID_MESA_LLVMPIPE
        driverName         = llvmpipe
        driverInfo         = Mesa 24.3.3 - kisak-mesa PPA (LLVM 17.0.6)
        conformanceVersion = 1.3.1.1
        deviceUUID         = 6d657361-3234-2e33-2e33-202d206b6900
        driverUUID         = 6c6c766d-7069-7065-5555-494400000000

weirdly, vkcube works and uses the 3090, but other tests like glxgears don’t and crash to a black screen after a few frames (although reported FPS is in the thousands?)

glxgears -info
GL_RENDERER   = llvmpipe (LLVM 17.0.6, 256 bits)
GL_VERSION    = 4.5 (Compatibility Profile) Mesa 24.3.3 - kisak-mesa PPA
GL_VENDOR     = Mesa
GL_EXTENSIONS = GL_ARB_multisample GL_EXT_abgr GL_EXT_bgra GL_EXT_blend_color GL_EXT_blend_minmax GL_EXT_blend_subtract GL_EXT_copy_texture GL_EXT_subtexture GL_EXT_texture_object GL_EXT_vertex_array GL_EXT_compiled_vertex_array GL_EXT_texture GL_EXT_texture3D GL_IBM_rasterpos_clip GL_ARB_point_parameters GL_EXT_draw_range_elements GL_EXT_packed_pixels GL_EXT_point_parameters GL_EXT_rescale_normal GL_EXT_separate_specular_color GL_EXT_texture_edge_clamp GL_SGIS_generate_mipmap GL_SGIS_texture_border_clamp GL_SGIS_texture_edge_clamp GL_SGIS_texture_lod GL_ARB_framebuffer_sRGB GL_ARB_multitexture GL_EXT_framebuffer_sRGB GL_IBM_multimode_draw_arrays GL_IBM_texture_mirrored_repeat GL_3DFX_texture_compression_FXT1 GL_ARB_texture_cube_map GL_ARB_texture_env_add GL_ARB_transpose_matrix GL_EXT_blend_func_separate GL_EXT_fog_coord GL_EXT_multi_draw_arrays GL_EXT_secondary_color GL_EXT_texture_env_add GL_EXT_texture_filter_anisotropic GL_EXT_texture_lod_bias GL_INGR_blend_func_separate GL_NV_blend_square GL_NV_light_max_exponent GL_NV_texgen_reflection GL_NV_texture_env_combine4 GL_S3_s3tc GL_SUN_multi_draw_arrays GL_ARB_texture_border_clamp GL_ARB_texture_compression GL_EXT_framebuffer_object GL_EXT_texture_compression_s3tc GL_EXT_texture_env_combine GL_EXT_texture_env_dot3 GL_MESA_window_pos GL_NV_packed_depth_stencil GL_NV_texture_rectangle GL_ARB_depth_texture GL_ARB_occlusion_query GL_ARB_shadow GL_ARB_texture_env_combine GL_ARB_texture_env_crossbar GL_ARB_texture_env_dot3 GL_ARB_texture_mirrored_repeat GL_ARB_window_pos GL_ATI_fragment_shader GL_EXT_stencil_two_side GL_EXT_texture_cube_map GL_NV_copy_depth_to_color GL_NV_depth_clamp GL_NV_fog_distance GL_NV_half_float GL_APPLE_packed_pixels GL_ARB_draw_buffers GL_ARB_fragment_program GL_ARB_fragment_shader GL_ARB_shader_objects GL_ARB_vertex_program GL_ARB_vertex_shader GL_ATI_draw_buffers GL_ATI_texture_env_combine3 GL_ATI_texture_float GL_EXT_shadow_funcs GL_EXT_stencil_wrap GL_MESA_pack_invert GL_MESA_ycbcr_texture GL_NV_primitive_restart GL_ARB_depth_clamp GL_ARB_fragment_program_shadow GL_ARB_half_float_pixel GL_ARB_occlusion_query2 GL_ARB_point_sprite GL_ARB_shading_language_100 GL_ARB_sync GL_ARB_texture_non_power_of_two GL_ARB_vertex_buffer_object GL_ATI_blend_equation_separate GL_EXT_blend_equation_separate GL_OES_read_format GL_ARB_color_buffer_float GL_ARB_pixel_buffer_object GL_ARB_texture_compression_rgtc GL_ARB_texture_float GL_ARB_texture_rectangle GL_ATI_texture_compression_3dc GL_EXT_packed_float GL_EXT_pixel_buffer_object GL_EXT_texture_compression_dxt1 GL_EXT_texture_compression_rgtc GL_EXT_texture_mirror_clamp GL_EXT_texture_rectangle GL_EXT_texture_sRGB GL_EXT_texture_shared_exponent GL_ARB_framebuffer_object GL_EXT_framebuffer_blit GL_EXT_framebuffer_multisample GL_EXT_packed_depth_stencil GL_ARB_vertex_array_object GL_ATI_separate_stencil GL_ATI_texture_mirror_once GL_EXT_draw_buffers2 GL_EXT_draw_instanced GL_EXT_gpu_program_parameters GL_EXT_gpu_shader4 GL_EXT_texture_array GL_EXT_texture_compression_latc GL_EXT_texture_integer GL_EXT_texture_sRGB_decode GL_EXT_timer_query GL_OES_EGL_image GL_EXT_texture_buffer_object GL_AMD_texture_texture4 GL_ARB_copy_buffer GL_ARB_depth_buffer_float GL_ARB_draw_instanced GL_ARB_half_float_vertex GL_ARB_instanced_arrays GL_ARB_map_buffer_range GL_ARB_texture_buffer_object GL_ARB_texture_rg GL_ARB_texture_swizzle GL_ARB_vertex_array_bgra GL_EXT_texture_swizzle GL_EXT_vertex_array_bgra GL_NV_conditional_render GL_AMD_conservative_depth GL_AMD_draw_buffers_blend GL_AMD_seamless_cubemap_per_texture GL_AMD_shader_stencil_export GL_ARB_ES2_compatibility GL_ARB_blend_func_extended GL_ARB_compatibility GL_ARB_debug_output GL_ARB_draw_buffers_blend GL_ARB_draw_elements_base_vertex GL_ARB_explicit_attrib_location GL_ARB_fragment_coord_conventions GL_ARB_provoking_vertex GL_ARB_sample_shading GL_ARB_sampler_objects GL_ARB_seamless_cube_map GL_ARB_shader_stencil_export GL_ARB_shader_texture_lod GL_ARB_tessellation_shader GL_ARB_texture_buffer_object_rgb32 GL_ARB_texture_cube_map_array GL_ARB_texture_gather GL_ARB_texture_multisample GL_ARB_texture_query_lod GL_ARB_texture_rgb10_a2ui GL_ARB_uniform_buffer_object GL_ARB_vertex_type_2_10_10_10_rev GL_ATI_meminfo GL_EXT_provoking_vertex GL_EXT_texture_snorm GL_EXT_texture_storage GL_MESA_texture_signed_rgba GL_NV_copy_image GL_NV_texture_barrier GL_ARB_draw_indirect GL_ARB_get_program_binary GL_ARB_gpu_shader5 GL_ARB_gpu_shader_fp64 GL_ARB_robustness GL_ARB_separate_shader_objects GL_ARB_shader_bit_encoding GL_ARB_shader_precision GL_ARB_shader_subroutine GL_ARB_texture_compression_bptc GL_ARB_timer_query GL_ARB_transform_feedback2 GL_ARB_transform_feedback3 GL_ARB_vertex_attrib_64bit GL_ARB_viewport_array GL_EXT_direct_state_access GL_EXT_vertex_attrib_64bit GL_AMD_multi_draw_indirect GL_ANGLE_texture_compression_dxt3 GL_ANGLE_texture_compression_dxt5 GL_ARB_base_instance GL_ARB_compressed_texture_pixel_storage GL_ARB_conservative_depth GL_ARB_internalformat_query GL_ARB_map_buffer_alignment GL_ARB_shader_atomic_counters GL_ARB_shader_image_load_store GL_ARB_shading_language_420pack GL_ARB_shading_language_packing GL_ARB_texture_storage GL_ARB_transform_feedback_instanced GL_EXT_framebuffer_multisample_blit_scaled GL_EXT_transform_feedback GL_AMD_query_buffer_object GL_AMD_shader_trinary_minmax GL_AMD_vertex_shader_layer GL_AMD_vertex_shader_viewport_index GL_ARB_ES3_compatibility GL_ARB_arrays_of_arrays GL_ARB_clear_buffer_object GL_ARB_compute_shader GL_ARB_copy_image GL_ARB_explicit_uniform_location GL_ARB_fragment_layer_viewport GL_ARB_framebuffer_no_attachments GL_ARB_invalidate_subdata GL_ARB_multi_draw_indirect GL_ARB_program_interface_query GL_ARB_robust_buffer_access_behavior GL_ARB_shader_image_size GL_ARB_shader_storage_buffer_object GL_ARB_stencil_texturing GL_ARB_texture_buffer_range GL_ARB_texture_query_levels GL_ARB_texture_storage_multisample GL_ARB_texture_view GL_ARB_vertex_attrib_binding GL_KHR_debug GL_KHR_robustness GL_KHR_texture_compression_astc_ldr GL_NV_shader_atomic_float GL_AMD_pinned_memory GL_ARB_buffer_storage GL_ARB_clear_texture GL_ARB_enhanced_layouts GL_ARB_indirect_parameters GL_ARB_internalformat_query2 GL_ARB_multi_bind GL_ARB_query_buffer_object GL_ARB_seamless_cubemap_per_texture GL_ARB_shader_draw_parameters GL_ARB_shader_group_vote GL_ARB_shading_language_include GL_ARB_texture_mirror_clamp_to_edge GL_ARB_texture_stencil8 GL_ARB_vertex_type_10f_11f_11f_rev GL_EXT_debug_label GL_EXT_shader_framebuffer_fetch GL_EXT_shader_integer_mix GL_NVX_gpu_memory_info GL_ARB_ES3_1_compatibility GL_ARB_clip_control GL_ARB_conditional_render_inverted GL_ARB_cull_distance GL_ARB_derivative_control GL_ARB_direct_state_access GL_ARB_get_texture_sub_image GL_ARB_pipeline_statistics_query GL_ARB_shader_texture_image_samples GL_ARB_texture_barrier GL_ARB_transform_feedback_overflow_query GL_ARM_shader_framebuffer_fetch_depth_stencil GL_EXT_polygon_offset_clamp GL_EXT_shader_image_load_formatted GL_KHR_blend_equation_advanced GL_KHR_blend_equation_advanced_coherent GL_KHR_context_flush_control GL_KHR_robust_buffer_access_behavior GL_ARB_ES3_2_compatibility GL_ARB_gpu_shader_int64 GL_ARB_parallel_shader_compile GL_ARB_post_depth_coverage GL_ARB_shader_atomic_counter_ops GL_ARB_shader_ballot GL_ARB_shader_clock GL_ARB_shader_viewport_layer_array GL_ARB_texture_filter_minmax GL_EXT_texture_filter_minmax GL_EXT_texture_sRGB_R8 GL_EXT_texture_sRGB_RG8 GL_KHR_no_error GL_KHR_texture_compression_astc_sliced_3d GL_ARB_gl_spirv GL_ARB_spirv_extensions GL_MESA_shader_integer_functions GL_ARB_polygon_offset_clamp GL_ARB_texture_filter_anisotropic GL_EXT_memory_object GL_EXT_memory_object_fd GL_KHR_parallel_shader_compile GL_NV_alpha_to_coverage_dither_control GL_EXT_EGL_image_storage GL_EXT_shader_framebuffer_fetch_non_coherent GL_EXT_texture_shadow_lod GL_INTEL_shader_atomic_float_minmax GL_MESA_framebuffer_flip_y GL_OVR_multiview GL_OVR_multiview2 GL_EXT_EGL_sync GL_EXT_EGL_image_storage_compression GL_NV_ES1_1_compatibility
VisualID 771, 0x303

glxgears
5462 frames in 5.0 seconds = 1092.375 FPS
5421 frames in 5.0 seconds = 1084.068 FPS
5541 frames in 5.0 seconds = 1108.058 FPS
5451 frames in 5.0 seconds = 1090.168 FPS

vkcube
WARNING: dzn is not a conformant Vulkan implementation, testing use only.
Selected GPU 0: Microsoft Direct3D12 (NVIDIA GeForce RTX 3090), type: DiscreteGpu

I’ve also tried setting MESA_D3D12_DEFAULT_ADAPTER_NAME=nvidia
I’ve tried re-installing everything, but my system always chooses to use llvmpipe for some reason. I’m at a loss with what to do.

2

Hi @lockheed1919, welcome to the NVIDIA developer forums!

For specific CUDA questions and especially installation issues I highly recommend CUDA - NVIDIA Developer Forums

But in your case it is probably much easier.

The way WSL2 works is that it utilizes any GPU driver transparently through their sub-system. That means if you have installed a current Windows GPU driver for your GPU you will not need to install anything else to get GL or VULKAN apps to run. Even ready-to-run CUDA apps will run as is in your VM, no further installation needed.

Only if you want to develop your own CUDA apps you will need to go through the installation process for the CUDA toolkit.

Getting to your glxgears example, I just started up a fresh Ubuntu WSL2, installed mesa-utils and ran glxgearswithut issues.

So if vkcube works but not much else, that means either you installed conflicting software within your WSL environment or your Windows GPU installation is somehow broken.