Will Khadas be issuing the mali gpu drivers for the VIM3?
Only have fbdev driver for VIM3 now. But seems a new version support wayland from the buildroot source code, but still NO X11 suoport.
I can live with wayland, I’ve been using it more with the other SBC’s I have I only have one from rockchip that support x11 that’s an arm processor. Thanks for the answer
Do you have VIM1 specific fb driver ?
kernel space driver: https://github.com/khadas/linux/tree/khadas-vims-4.9.y/drivers/gpu/arm/utgard
userspace blobs: https://github.com/khadas/fenix/tree/master/archives/hwpacks/mali/Amlogic/lib/arm64/m450/r7p0/fbdev
Is there any word on OpenCL driver support, I’d test myself but still waiting for my board?
Is OpenCL drivers available for VIM3…?
The fbdev driver will support OpenCL as long as you have the libmali.ko file in /usr/lib/aarch64-linx-gnu/ directory. But for whatever reason this driver is slower for the VIM3 than the VIM3L at OpenCL mining Content anyway.
It is already support in latest image.
VIM3 is solwer then VIM3L? Could you please tell me how to reproduce this ?
I tested it under xmrig v5+ compiling for OpenCL support. Then tested with CN-pico, a.k.a turtlecoin or cn-pico/trtl. I’m only referring to the OpenCL results though. The CPU is cores are about twice as fast without the 4 vs. 6 handy cap. I’ve not noticed any difference in games, but I’ve not tried anything taxing for this either.
I should also note that I don’t think I’m not on the newest image for either of them. I’ve only used Ubuntu as Debian tends to be slower in mining. I also believe it’s an issue with the fbdev driver not taking advantage of all the compute cores. The G-31 only says it has one compute core while the G-52 says that it only has two compute cores according to clinfo. I’ve not tried to clpeak on both of the boards, this may show them being different, but the mining results would do the same.
Steps used to make the O.S. see the OpenCL support for the Mali GPU’s
$ sudo apt-get install -y mesa-opencl-icd opencl-headers ocl-icd-opencl-dev clinfo
$ sudo echo libmali.ko > /etc/OpenCL/vendors/mali.ko
Now you can run clinfo
This is how to run it outside of wayland or X, it will also give you the be performance as well as the gpu isn’t being used for much either.
Ohh… I tried clinfo since it showed zero supported platforms I assumed it driver doesn’t work…
Also lshw shows no gpu present.
I will try the above step and try again…
which image you use?
root@Khadas:~# clinfo Number of platforms 1 Platform Name ARM Platform Platform Vendor ARM Platform Version OpenCL 2.0 git.c8adbf9.122c9daed32dbba4b3056f41a2f23c58 Platform Profile FULL_PROFILE Platform Extensions 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_byte_addressable_store cl_khr_3d_image_writes cl_khr_int64_base_atomics cl_khr_int64_extended_atomics cl_khr_fp16 cl_khr_icd cl_khr_egl_image cl_khr_image2d_from_buffer cl_khr_depth_images cl_khr_subgroups cl_khr_create_command_queue cl_arm_core_id cl_arm_printf cl_arm_thread_limit_hint cl_arm_non_uniform_work_group_size cl_arm_import_memory cl_arm_shared_virtual_memory Platform Extensions function suffix ARM Platform Name ARM Platform Number of devices 1 Device Name Mali-G52 Device Vendor ARM Device Vendor ID 0x72120000 Device Version OpenCL 2.0 git.c8adbf9.122c9daed32dbba4b3056f41a2f23c58 Driver Version 2.0 Device OpenCL C Version OpenCL C 2.0 git.c8adbf9.122c9daed32dbba4b3056f41a2f23c58 Device Type GPU Device Profile FULL_PROFILE Device Available Yes Compiler Available Yes Linker Available Yes Max compute units 2 Max clock frequency 750MHz Device Partition (core) Max number of sub-devices 0 Supported partition types None Max work item dimensions 3 Max work item sizes 384x384x384 Max work group size 384 Preferred work group size multiple 8 Preferred / native vector sizes char 16 / 4 short 8 / 2 int 4 / 1 long 2 / 1 half 8 / 2 (cl_khr_fp16) float 4 / 1 double 0 / 0 (n/a) Half-precision Floating-point support (cl_khr_fp16) 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 Single-precision Floating-point support (core) 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 Correctly-rounded divide and sqrt operations No Double-precision Floating-point support (n/a) Address bits 64, Little-Endian Global memory size 3888193536 (3.621GiB) Error Correction support No Max memory allocation 972048384 (927MiB) Unified memory for Host and Device Yes Shared Virtual Memory (SVM) capabilities (core) Coarse-grained buffer sharing Yes Fine-grained buffer sharing No Fine-grained system sharing No Atomics No Shared Virtual Memory (SVM) capabilities (ARM) Coarse-grained buffer sharing Yes Fine-grained buffer sharing No Fine-grained system sharing No Atomics No Minimum alignment for any data type 128 bytes Alignment of base address 1024 bits (128 bytes) Preferred alignment for atomics SVM 0 bytes Global 0 bytes Local 0 bytes Max size for global variable 65536 (64KiB) Preferred total size of global vars 0 Global Memory cache type Read/Write Global Memory cache size 131072 (128KiB) Global Memory cache line size 64 bytes Image support Yes Max number of samplers per kernel 16 Max size for 1D images from buffer 65536 pixels Max 1D or 2D image array size 2048 images Base address alignment for 2D image buffers 32 bytes Pitch alignment for 2D image buffers 64 pixels Max 2D image size 65536x65536 pixels Max 3D image size 65536x65536x65536 pixels Max number of read image args 128 Max number of write image args 64 Max number of read/write image args 64 Max number of pipe args 16 Max active pipe reservations 1 Max pipe packet size 1024 Local memory type Global Local memory size 32768 (32KiB) Max number of constant args 8 Max constant buffer size 65536 (64KiB) Max size of kernel argument 1024 Queue properties (on host) Out-of-order execution Yes Profiling Yes Queue properties (on device) Out-of-order execution Yes Profiling Yes Preferred size 2097152 (2MiB) Max size 16777216 (16MiB) Max queues on device 1 Max events on device 1024 Prefer user sync for interop No Profiling timer resolution 1000ns Execution capabilities Run OpenCL kernels Yes Run native kernels No printf() buffer size 1048576 (1024KiB) Built-in kernels Device Extensions 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_byte_addressable_store cl_khr_3d_image_writes cl_khr_int64_base_atomics cl_khr_int64_extended_atomics cl_khr_fp16 cl_khr_icd cl_khr_egl_image cl_khr_image2d_from_buffer cl_khr_depth_images cl_khr_subgroups cl_khr_create_command_queue cl_arm_core_id cl_arm_printf cl_arm_thread_limit_hint cl_arm_non_uniform_work_group_size cl_arm_import_memory cl_arm_shared_virtual_memory NULL platform behavior clGetPlatformInfo(NULL, CL_PLATFORM_NAME, ...) ARM Platform clGetDeviceIDs(NULL, CL_DEVICE_TYPE_ALL, ...) Success [ARM] clCreateContext(NULL, ...) [default] Success [ARM] clCreateContextFromType(NULL, CL_DEVICE_TYPE_DEFAULT) Success (1) Platform Name ARM Platform Device Name Mali-G52 clCreateContextFromType(NULL, CL_DEVICE_TYPE_CPU) No devices found in platform clCreateContextFromType(NULL, CL_DEVICE_TYPE_GPU) Success (1) Platform Name ARM Platform Device Name Mali-G52 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 ARM Platform Device Name Mali-G52 ICD loader properties ICD loader Name OpenCL ICD Loader ICD loader Vendor OCL Icd free software ICD loader Version 2.2.11 ICD loader Profile OpenCL 2.1
A lot of the hardware utilities simply don’t work with due to them being written for x86_64 hardware. dmesg is the utility to use
example: $ dmesg | grep gpu or $ dmesg | grep mali
if you want to look for say the nvme if you have it plugged in you switch mali or gpu for nvme or superspeed for the usb 3.0 port
Last one I created from Fenix was on 11/6/2019 for the vim3 and the vim3L was 11/25/2019 both are using the eMMC. Both run Ubuntu server 18.04.03LTS kernel 4.9
Can you try the latest image ?
yes, I will, I’ll use the SD card image though, I’ve got to much stuff with the android image now to reflash the eMMC. Thanks for the link, I’ll get the other from the support page
Okay, on the new image with updates, it is faster then the original test the hash rates doubled from about 28H/s to 56H/s
Yeah some benchmarks in harinfo were also noticeably faster…
@wll1rah What is the hash rate you got with VIM3L…?
@numbqq Thanks, I was using a older version… Now it shows the GPU…