86 questions
- Bountied 0
- Unanswered
- Frequent
- Score
- Trending
- Week
- Month
- Unanswered (my tags)
1
vote
0
answers
82
views
OnnxRuntime with ACL Execution Provider on RK3588 (Mali-G610): Nodes assigned to ACL but GPU load remains 0%
[Goal & Problem]
I am trying to accelerate ONNX model inference on an RK3588 (Orange Pi 5) board using the Mali-G610 GPU. I have built OnnxRuntime (ORT) with the ACL (Compute Library) Execution ...
0
votes
1
answer
110
views
How to choose appropriate EGL config for GPGPU on GLESv3.1+?
I am working on a simple OpenGL ES v3.1+ setup to use it for GPGPU purposes on a headless system. (Note: Uses proprietary Mali Valhall G610 drivers on Debian 12)
The GPU supports the ...
2
votes
1
answer
113
views
How to achieve zero-copy rendering from Qt QPainter to dmabuf using shared EGL contexts?
I'm trying to implement a zero-copy pipeline that renders Qt QPainter content directly to a dmabuf framebuffer for display via DRM/KMS. My goal is to leverage QPainter's high-level drawing APIs (text, ...
0
votes
1
answer
211
views
clImportMemoryARM Opencl sometime return error -6
I am testing the clImportMemoryARM function of opencl.
env
android C++
CPU MTK G99
GPU: Mali-G57 MC2 r0p1
EXTENSIONS: cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics ...
0
votes
0
answers
81
views
installing/building PyOpenCL to Mali GPU causes linux kernel panic: Unable to handle kernel NULL pointer dereference
I'm trying to install PyOpenCL on a Mali G78 GPU, with the environment set to Ubuntu 22.04 (aarch64). However, I keep encountering kernel panics.
Does anyone know the reason why this is happening?
pip ...
2
votes
1
answer
169
views
How to exploit Unified Memory in OpenCL with CL_MEM_ALLOC_HOST_PTR flag?
How can I use unified memory on an OpenCL device without copying data correctly?
OpenCL defined CL_DEVICE_HOST_UNIFIED_MEMORY as:
Is CL_TRUE if the device and the host have a unified memory subsystem ...
1
vote
1
answer
147
views
What are the rules for the precision of casting operations in GLSL
I have a GL_R16UI texture in OpenGL ES 3.0 on an Android device (Google Pixel 6). I'm then drawing a screen rect, and copying the texture to the RGBA8 output using the following fragment shader:
...
0
votes
1
answer
351
views
How to Profile OpenCL code on Android run on an ARM Mali GPU
How can I profiler a Mali GPU on an Android device with native code?
Suppose a binary on the Android device is run as "./run.bin". This binary is not packaged as an apk. That binary contains ...
0
votes
0
answers
1k
views
How to enable hw OpenCL acceleration with Mali GPU?
I have a OPPO Find X6, its SOC is dimensity 9200 which have a Mali-G715 Immortalis MP11 GPU.
I have try to run clpeak in termux, but it report error:
clGetPlatformIDs (-1001)
no platforms found
here ...
2
votes
0
answers
330
views
Mali gpu directory
I cannot find the gpu directory for Mali Gpu's current frequency in the root flile system.
Like in case of Qualcomm Adreno GPU the path for the gpu frequency is /sys/devices/platform/kgsl-3d0.0/kgsl/...
3
votes
1
answer
1k
views
Unusual GPU error when using ARCore, Unrecognised Android chroma siting range
I'm receiving the error:
E/mali_winsys: Unrecognised Android chroma siting range: [2, 2]. Applying default
When opening my ArCore activities, any of them, using any range of models and textures.
For ...
0
votes
1
answer
602
views
How to change value CL_KERNEL_PREFERRED_WORK_GROUP_SIZE_MULTIPLE for OpenCL Mali-platform?
I am developing OpenCL code for different devices. At the current time I work with Rockchip RK3588 (OpenCL device - Mali-G610 r0p0). The program algorithm was originally written on CUDA, where the ...
0
votes
1
answer
1k
views
Which one is used in Mali GPU driver source in Samsung Galaxy Android?
As you know, there are several Mali Midgard directories in Samsung Galaxy's kernel source.
drwxrwxrwx@ 141 eee staff 4512 10 29 2020 b_r10p0
drwxrwxrwx@ 142 eee staff 4544 10 29 2020 b_r12p0
...
0
votes
0
answers
81
views
Arm Mali T-624 gpu arithmetic pipeline depth kernel
I am doing a research on an Arm Mali T-624 board and i want to discover how many stages does the gpu arithmetic pipeline has and i am doing the running the following kernel
__kernel void ...
0
votes
0
answers
112
views
Why is matrix multiplication row x row 4-5 times slower than row x column on Mali's GPU?
Recently, I encountered a problem when using computer shader to develop matrix multiplication. A common matrix multiplication C = AB. in order to make the memory continuous, I transposed the B matrix. ...