stitchEm: Linux Fedora 30 does not start wants CUDA but only Opencl is enable OpenCL 1.1 Mesa 19.1.5

      -DGPU_BACKEND_CUDA=OFF -DGPU_BACKEND_OPENCL=ON \
      -G Ninja \
      -DCMAKE_BUILD_TYPE=Release
CMake Warning:
  No source or binary directory provided.  Both will be assumed to be the
  same as the current working directory, but note that this warning will
  become a fatal error in future CMake releases.


-- GPU_BACKEND_CUDA='OFF'
-- GPU_BACKEND_OPENCL='ON'
-- System: LINUX
-- CMake Build Type -- Release
-- USE_DELAY_LOAD = OFF
-- CUDA VERSION: 8.0
-- CUDA PATH: /usr/local/cuda-8.0
-- Could NOT find CUDA (missing: CUDA_NVCC_EXECUTABLE CUDA_INCLUDE_DIRS CUDA_CUDART_LIBRARY) (found suitable exact version "8.0")
No Alternative OpenCL SPIR compiler for Linux
-- --- OpenCL GPU backend ---
-- OpenCL version: 2.2
-- OpenCL includes: /usr/include
-- OpenCL libraries: /usr/lib64/libOpenCL.so
-- 
-- Main OpenCL SPIR compiler: /usr/bin/clang
-- Alternative OpenCL SPIR compiler: 
-- System: LINUX
-- --- VideoStitch-lib dependencies ---
-- OpenEXR libs: /usr/lib64/libHalf.so;/usr/lib64/libIex.so;/usr/lib64/libImath.so;/usr/lib64/libIlmImf.so;/usr/lib64/libIlmThread.so
-- OpenEXR include: /usr/include;/usr/include/openexr
-- PNG: /usr/lib64/libpng.so
-- R8B: r8b
-- OpenCV libs: /usr/lib64/libopencv_calib3d.so;/usr/lib64/libopencv_features2d.so;/usr/lib64/libopencv_flann.so;/usr/lib64/libopencv_video.so;/usr/lib64/libopencv_imgproc.so;/usr/lib64/libopencv_core.so
-- CERES LIBS: general;/usr/lib64/libceres.so
-- 
-- VideoStitch I/O libs configured to build: avPlugin;jpgPlugin;pamPlugin;pngPlugin;rawPlugin;tiffPlugin;exrPlugin;rtmpPlugin;v4l2Plugin;portaudioPlugin
-- Qt minimum supported version: 5.6
-- Qt FOUND: 5.12.4
-- Studio version: Studio-v0.0.9-31-g20693a5-master.2019-08-29
-- Configuring done
-- Generating done
-- Build files have been written to: /home/arpu/Work/githubsources/stitchEm_master
 arpu@testnet  ~/Work/githubsources/stitchEm_master   master ?  ninja                                                                                                                                                                                                                              ✔  10215  23:10:35 
[1/46] cd /home/arpu/Work/githubsources/stitchEm_master/lib && ls /home/arpu/Work/githubsources/stitchEm_master/generated_headers/version.hpp
/home/arpu/Work/githubsources/stitchEm_master/generated_headers/version.hpp
[16/18] Automatic MOC for target videostitch-studio
 arpu@testnet  ~/Work/githubsources/stitchEm_master   master ?  bin/x64/release/videostitch-studio                                                                                                                                                                                                 ✔  10216  23:10:40 
Warning: Ignoring XDG_SESSION_TYPE=wayland on Gnome. Use QT_QPA_PLATFORM=wayland to run on Wayland anyway.
Host OpenCL allocated memory:
Device OpenCL allocated memory:
Number of platforms                               1
  Platform Name                                   Clover
  Platform Vendor                                 Mesa
  Platform Version                                OpenCL 1.1 Mesa 19.1.5
  Platform Profile                                FULL_PROFILE
  Platform Extensions                             cl_khr_icd
  Platform Extensions function suffix             MESA

  Platform Name                                   Clover
Number of devices                                 1
  Device Name                                     AMD Radeon (TM) RX 460 Graphics (POLARIS11, DRM 3.32.0, 5.2.9-200.fc30.x86_64, LLVM 8.0.0)
  Device Vendor                                   AMD
  Device Vendor ID                                0x1002
  Device Version                                  OpenCL 1.1 Mesa 19.1.5
  Driver Version                                  19.1.5
  Device OpenCL C Version                         OpenCL C 1.1 
  Device Type                                     GPU
  Device Profile                                  FULL_PROFILE
  Device Available                                Yes
  Compiler Available                              Yes
  Max compute units                               14
  Max clock frequency                             1210MHz
  Max work item dimensions                        3
  Max work item sizes                             256x256x256
  Max work group size                             256
  Preferred work group size multiple              64
  Preferred / native vector sizes                 
    char                                                16 / 16      
    short                                                8 / 8       
    int                                                  4 / 4       
    long                                                 2 / 2       
    half                                                 8 / 8        (cl_khr_fp16)
    float                                                4 / 4       
    double                                               2 / 2        (cl_khr_fp64)
  Half-precision Floating-point support           (cl_khr_fp16)
    Denormals                                     No
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 No
    Round to infinity                             No
    IEEE754-2008 fused multiply-add               No
    Support is emulated in software               No
  Single-precision Floating-point support         (core)
    Denormals                                     No
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 No
    Round to infinity                             No
    IEEE754-2008 fused multiply-add               No
    Support is emulated in software               No
    Correctly-rounded divide and sqrt operations  No
  Double-precision Floating-point support         (cl_khr_fp64)
    Denormals                                     Yes
    Infinity and NANs                             Yes
    Round to nearest                              Yes
    Round to zero                                 Yes
    Round to infinity                             Yes
    IEEE754-2008 fused multiply-add               Yes
    Support is emulated in software               No
  Address bits                                    64, Little-Endian
  Global memory size                              4294967296 (4GiB)
  Error Correction support                        No
  Max memory allocation                           3435973836 (3.2GiB)
  Unified memory for Host and Device              No
  Minimum alignment for any data type             128 bytes
  Alignment of base address                       32768 bits (4096 bytes)
  Global Memory cache type                        None
  Image support                                   No
  Local memory type                               Local
  Local memory size                               32768 (32KiB)
  Max number of constant args                     16
  Max constant buffer size                        2147483647 (2GiB)
  Max size of kernel argument                     1024
  Queue properties                                
    Out-of-order execution                        No
    Profiling                                     Yes
  Profiling timer resolution                      0ns
  Execution capabilities                          
    Run OpenCL kernels                            Yes
    Run native kernels                            No
  Device Extensions                               cl_khr_byte_addressable_store cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_int64_base_atomics cl_khr_int64_extended_atomics cl_khr_fp64 cl_khr_fp16

NULL platform behavior
  clGetPlatformInfo(NULL, CL_PLATFORM_NAME, ...)  Clover
  clGetDeviceIDs(NULL, CL_DEVICE_TYPE_ALL, ...)   Success [MESA]
  clCreateContext(NULL, ...) [default]            Success [MESA]
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_DEFAULT)  Success (1)
    Platform Name                                 Clover
    Device Name                                   AMD Radeon (TM) RX 460 Graphics (POLARIS11, DRM 3.32.0, 5.2.9-200.fc30.x86_64, LLVM 8.0.0)
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_CPU)  No devices found in platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_GPU)  Success (1)
    Platform Name                                 Clover
    Device Name                                   AMD Radeon (TM) RX 460 Graphics (POLARIS11, DRM 3.32.0, 5.2.9-200.fc30.x86_64, LLVM 8.0.0)
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_ACCELERATOR)  No devices found in platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_CUSTOM)  No devices found in platform
  clCreateContextFromType(NULL, CL_DEVICE_TYPE_ALL)  Success (1)
    Platform Name                                 Clover
    Device Name                                   AMD Radeon (TM) RX 460 Graphics (POLARIS11, DRM 3.32.0, 5.2.9-200.fc30.x86_64, LLVM 8.0.0)

ICD loader properties
  ICD loader Name                                 OpenCL ICD Loader
  ICD loader Vendor                               OCL Icd free software
  ICD loader Version                              2.2.12
  ICD loader Profile                              OpenCL 2.2

About this issue

  • Original URL
  • State: closed
  • Created 5 years ago
  • Comments: 21 (21 by maintainers)

Most upvoted comments

The software needs a working OpenGL context that is shared when creating the OpenCL context. So that the stitched images, computed with OpenCL, can directly displayed with OpenGL.

It seems to me that the current OpenGL context can be found (https://github.com/stitchEm/stitchEm/blob/master/lib/src/backend/cl/context.cpp#L230), otherwise the line “[OpenCL] Creating context without OpenGL context sharing!” would be printed when running videostitch-studio. (It’s okay to have the line in SimpleKernelTest, as no OpenGL context is created for running this simple computation test.)

The creation of the OpenCL context with sharing to the OpenGL context (https://github.com/stitchEm/stitchEm/blob/master/lib/src/backend/cl/context.cpp#L240) is probably what’s failing with the invalid configuration error.

This can happen when the monitor is not connected to the GPU doing the OpenCL computations, but rather to an integrated GPU directly on the mainboard, for example.

But if your monitor is already connected to the correct GPU, then this unfortunately points to some kind of driver problem again, now involving the interchange between OpenGL and OpenCL.

stitchEm requires an OpenCL 1.2-capable device to run. Your AMD Radeon RX 460 is capable of OpenCL 2.0, so this should in general be fine.

I’m not familiar with Mesa OpenCL, but as your installation says OpenCL 1.1 in a couple of places, this may not work and you will have to use a different graphics driver.

Looking at the Mesa OpenCL package for Fedora 30 (https://fedora.pkgs.org/30/fedora-x86_64/mesa-libOpenCL-19.0.2-3.fc30.x86_64.rpm.html), it seems the driver provides a library called /usr/lib64/libMesaOpenCL.so.1

But the CMake configuration above shows that OpenCL is used from /usr/lib64/libOpenCL.so, also says OpenCL version 2.2.

The OpenCL version you use for building stitchEm should be the one that is installed alongside your currently active graphics driver.