Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Question-Bug?: Support for OpenGl/OpenCl Interop on B580 #796

Closed
catcatmu opened this issue Jan 17, 2025 · 4 comments
Closed

Question-Bug?: Support for OpenGl/OpenCl Interop on B580 #796

catcatmu opened this issue Jan 17, 2025 · 4 comments

Comments

@catcatmu
Copy link

catcatmu commented Jan 17, 2025 •
edited
Loading

Does this driver support the use of OpenGl/OpenCl interoperability for battlemage cards. The card reports the cl_khr_gl_sharing extension but fails all conformance test. Da Vinci Resolve appears to fail with gl/cl sharing errors as well: CL_INVALID_GL_OBJECT .

Not sure if relevant:
Switching to using zink appears to solve the issue with resolve, although with terrible performance:
env __GLX_VENDOR_LIBRARY_NAME=mesa __EGL_VENDOR_LIBRARY_FILENAMES=/usr/share/glvnd/egl_vendor.d/50_mesa.json MESA_LOADER_DRIVER_OVERRIDE=zink GALLIUM_DRIVER=zink /opt/resolve/bin/resolve

Could this be a mesa bug instead?

Versions

Package Versions
intel-compute-runtime 24.52.32224.5-1
kernel 6.12.9-arch1-1
mesa 1:24.3.3-2
intel-gmmlib 22.5.5-1

Ocl Info:

Number of platforms                               1
  Platform Name                                   Intel(R) OpenCL Graphics
  Platform Vendor                                 Intel(R) Corporation
  Platform Version                                OpenCL 3.0 
  Platform Profile                                FULL_PROFILE
  Platform Extensions                             cl_khr_byte_addressable_store cl_khr_device_uuid cl_khr_fp16 cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_icd cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_intel_command_queue_families cl_intel_subgroups cl_intel_required_subgroup_size cl_intel_subgroups_short cl_khr_spir cl_intel_accelerator cl_intel_driver_diagnostics cl_khr_priority_hints cl_khr_throttle_hints cl_khr_create_command_queue cl_intel_subgroups_char cl_intel_subgroups_long cl_khr_il_program cl_intel_mem_force_host_memory cl_khr_subgroup_extended_types cl_khr_subgroup_non_uniform_vote cl_khr_subgroup_ballot cl_khr_subgroup_non_uniform_arithmetic cl_khr_subgroup_shuffle cl_khr_subgroup_shuffle_relative cl_khr_subgroup_clustered_reduce cl_intel_device_attribute_query cl_khr_extended_bit_ops cl_khr_suggested_local_work_size cl_intel_split_work_group_barrier cl_khr_fp64 cl_khr_subgroups cl_intel_spirv_subgroups cl_khr_spirv_linkonce_odr cl_khr_spirv_no_integer_wrap_decoration cl_intel_unified_shared_memory cl_khr_mipmap_image cl_khr_mipmap_image_writes cl_ext_float_atomics cl_khr_external_memory cl_intel_planar_yuv cl_intel_packed_yuv cl_khr_int64_base_atomics cl_khr_int64_extended_atomics cl_khr_image2d_from_buffer cl_khr_depth_images cl_khr_3d_image_writes cl_intel_bfloat16_conversions cl_intel_create_buffer_with_properties cl_intel_subgroup_local_block_io cl_intel_subgroup_matrix_multiply_accumulate cl_intel_subgroup_matrix_multiply_accumulate_tf32 cl_khr_subgroup_named_barrier cl_intel_subgroup_extended_block_read cl_intel_subgroup_2d_block_io cl_intel_subgroup_buffer_prefetch cl_khr_integer_dot_product cl_khr_gl_sharing cl_khr_gl_depth_images cl_khr_gl_event cl_khr_gl_msaa_sharing cl_intel_va_api_media_sharing cl_intel_sharing_format_query cl_khr_pci_bus_info
......Skipped Lines
Number of devices                                 1
  Device Name                                     Intel(R) Graphics [0xe20b]
  Device Vendor                                   Intel(R) Corporation
  Device Vendor ID                                0x8086
  Device Version                                  OpenCL 3.0 NEO 
  Device UUID                                     86800be2-0000-0000-0b00-000000000000
  Driver UUID                                     32342e35-322e-3332-3232-340000000000
  Valid Device LUID                               No
  Device LUID                                     2061-6c78fe7f0000
  Device Node Mask                                0
  Device Numeric Version                          0xc00000 (3.0.0)
  Driver Version                                  24.52.32224
  Device OpenCL C Version                         OpenCL C 1.2 
...... Skipped lines
 Device Extensions                               cl_khr_byte_addressable_store cl_khr_device_uuid cl_khr_fp16 cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_icd cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_intel_command_queue_families cl_intel_subgroups cl_intel_required_subgroup_size cl_intel_subgroups_short cl_khr_spir cl_intel_accelerator cl_intel_driver_diagnostics cl_khr_priority_hints cl_khr_throttle_hints cl_khr_create_command_queue cl_intel_subgroups_char cl_intel_subgroups_long cl_khr_il_program cl_intel_mem_force_host_memory cl_khr_subgroup_extended_types cl_khr_subgroup_non_uniform_vote cl_khr_subgroup_ballot cl_khr_subgroup_non_uniform_arithmetic cl_khr_subgroup_shuffle cl_khr_subgroup_shuffle_relative cl_khr_subgroup_clustered_reduce cl_intel_device_attribute_query cl_khr_extended_bit_ops cl_khr_suggested_local_work_size cl_intel_split_work_group_barrier cl_khr_fp64 cl_khr_subgroups cl_intel_spirv_subgroups cl_khr_spirv_linkonce_odr cl_khr_spirv_no_integer_wrap_decoration cl_intel_unified_shared_memory cl_khr_mipmap_image cl_khr_mipmap_image_writes cl_ext_float_atomics cl_khr_external_memory cl_intel_planar_yuv cl_intel_packed_yuv cl_khr_int64_base_atomics cl_khr_int64_extended_atomics cl_khr_image2d_from_buffer cl_khr_depth_images cl_khr_3d_image_writes cl_intel_bfloat16_conversions cl_intel_create_buffer_with_properties cl_intel_subgroup_local_block_io cl_intel_subgroup_matrix_multiply_accumulate cl_intel_subgroup_matrix_multiply_accumulate_tf32 cl_khr_subgroup_named_barrier cl_intel_subgroup_extended_block_read cl_intel_subgroup_2d_block_io cl_intel_subgroup_buffer_prefetch cl_khr_integer_dot_product cl_khr_gl_sharing cl_khr_gl_depth_images cl_khr_gl_event cl_khr_gl_msaa_sharing cl_intel_va_api_media_sharing cl_intel_sharing_format_query cl_khr_pci_bus_info

Logs

OpenGl/Cl conformance test suite

redirect_images_read_2D...
Testing image read for GL format GL_TEXTURE_2D : GL_RGBA8 : GL_RGBA : GL_UNSIGNED_BYTE
- Read [  10 x   10 x    1 x    1] : GL Texture : GL_RGBA : GL_RGBA8 : GL_UNSIGNED_BYTE => CL Image : CL_RGBA : CL_UNORM_INT8 
ERROR: Data sample 0,0,0,0 did not validate!
	Expected: 0.671875 0.183594 0.457031 0.750000
	        : 0x1.58p-1 0x1.78p-3 0x1.d4p-2 0x1.8p-1
	  Actual: 0.000000 0.000000 0.000000 0.000000
	        : 0x0p+0 0x0p+0 0x0p+0 0x0p+0
ERROR: Image read test failed for GL_TEXTURE_2D : GL_RGBA8 : GL_RGBA : GL_UNSIGNED_BYTE

Testing image read for GL format GL_TEXTURE_2D : GL_DEPTH_COMPONENT16 : GL_DEPTH_COMPONENT : GL_UNSIGNED_SHORT
[1]    16370 segmentation fault (core dumped)  ./test_gl

gl_tests.txt

Resolve Logs:

0x77a14edfb000 | DVIP                 | ERROR | 2025-01-17 15:33:51,277 | Failed to register OpenGL object for OpenCL interop: CL_INVALID_GL_OBJECT.
0x77a14edfb000 | UI.GLTexPool         | ERROR | 2025-01-17 15:33:51,277 | Failed to register OpenGL texture to compute API
@smunaut
Copy link
Contributor

smunaut commented Feb 6, 2025

It's probably a new tiling format ...
cl/gl sharing isn't really "officially" supported, it's been contributed by community so if you want it fixed, you'll have to write the patch.

@smunaut
Copy link
Contributor

smunaut commented Feb 7, 2025

You could try those mesa MR :

https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30670
https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/32120

They might actually make the texture exported with proper modifier that the compute-runtime can accept.

Please report back if you try them.

@catcatmu
Copy link
Author

catcatmu commented Feb 7, 2025 •
edited
Loading

I can't currently test it, so I will close this issue for now. I will update it if the above MR fixes the issue.

@catcatmu catcatmu closed this as completed Feb 7, 2025
@catcatmu
Copy link
Author

catcatmu commented Mar 3, 2025 •
edited
Loading

https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/30670 fixes the issue for Da Vinci Resolve.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@smunaut @catcatmu