It seems that a lot of GPU problems revolve around specific versions of drivers. Though NVidia has their own support structure, you can often learn from information reported by others who fold.
Hi. I'm trying to fold on Gentoo Linux. I installed foldingathome from sci-biology/foldingathome.
CPU folding worked fine, but GPU doesn't work.
I think I have the nvidia driver and opencl installed, but FAHClient complains about "Failed to start core: OpenCL device matching slot 1 not found, make sure the OpenCL driver is installed or try setting 'opencl-index' manually".
Here's the related log (with download and cpu progress stripped):
Gentoo is not an officially supported Linux distribution, but you might find some workarounds here: viewtopic.php?f=89&t=33333&p=318172
First add the fahclient user to the "video" group.
Please post the first 100 or so lines of your log that explains what FAHClient sees about your System Info and Configuration. You can remove the Download and Progress lines.
viewtopic.php?p=327413&f=24#p327413
Have you installed the proprietary drivers from Nvidia? You need those - the open source Nvidia drivers can't be used for folding. (Linux Mint users get an installer for the proprietary drivers, but most other users have to use Nvidia's webside to download a package or installer for it.)
Are you using a laptop with a GTX 1050 in it? You might have to disable the integrated graphics.
Online: GTX 1660 Super + occasional CPU folding in the cold.
Offline: Radeon HD 7770, GTX 1050 Ti 4G OC, RX580
gunnarre wrote:Gentoo is not an officially supported Linux distribution, but you might find some workarounds here: viewtopic.php?f=89&t=33333&p=318172
First add the fahclient user to the "video" group.
Please post the first 100 or so lines of your log that explains what FAHClient sees about your System Info and Configuration. You can remove the Download and Progress lines.
viewtopic.php?p=327413&f=24#p327413
Have you installed the proprietary drivers from Nvidia? You need those - the open source Nvidia drivers can't be used for folding. (Linux Mint users get an installer for the proprietary drivers, but most other users have to use Nvidia's webside to download a package or installer for it.)
Are you using a laptop with a GTX 1050 in it? You might have to disable the integrated graphics.
I think I ran into a different problem other than this one. This bug has been fixed by Gentoo, and now foldingathome user is in video group by default.
The log has been updated.
I think I'm using the proprietary drivers installed from Gentoo's package manager.
What is the output of the "clinfo" command run as root? If it doesn't show any cards, you might need to add some OpenCL development files. If clinfo lists something as root, but not as the fahclient user, it's likely a permissions issue still. Does your system have a "render" group?
Online: GTX 1660 Super + occasional CPU folding in the cold.
Offline: Radeon HD 7770, GTX 1050 Ti 4G OC, RX580
gunnarre wrote:What is the output of the "clinfo" command run as root? If it doesn't show any cards, you might need to add some OpenCL development files. If clinfo lists something as root, but not as the fahclient user, it's likely a permissions issue still. Does your system have a "render" group?
ag@AprilGrimoire /home/ag sudo clinfo 22:59:57
Number of platforms 2
Platform Name NVIDIA CUDA
Platform Vendor NVIDIA Corporation
Platform Version OpenCL 1.2 CUDA 11.0.210
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_fp64 cl_khr_byte_addressable_store cl_khr_icd cl_khr_gl_sharing cl_nv_compiler_options cl_nv_device_attribute_query cl_nv_pragma_unroll cl_nv_copy_opts cl_nv_create_buffer cl_khr_int64_base_atomics cl_khr_int64_extended_atomics
Platform Extensions function suffix NV
Platform Name Clover
Platform Vendor Mesa
Platform Version OpenCL 1.1 Mesa 20.1.4
Platform Profile FULL_PROFILE
Platform Extensions cl_khr_icd
Platform Extensions function suffix MESA
Platform Name NVIDIA CUDA
Number of devices 1
Device Name GeForce GTX 1050
Device Vendor NVIDIA Corporation
Device Vendor ID 0x10de
Device Version OpenCL 1.2 CUDA
Driver Version 450.57
Device OpenCL C Version OpenCL C 1.2
Device Type GPU
Device Topology (NV) PCI-E, 01:00.0
Device Profile FULL_PROFILE
Device Available Yes
Compiler Available Yes
Linker Available Yes
Max compute units 5
Max clock frequency 1455MHz
Compute Capability (NV) 6.1
Device Partition (core)
Max number of sub-devices 1
Supported partition types None
Supported affinity domains (n/a)
Max work item dimensions 3
Max work item sizes 1024x1024x64
Max work group size 1024
Preferred work group size multiple 32
Warp size (NV) 32
Preferred / native vector sizes
char 1 / 1
short 1 / 1
int 1 / 1
long 1 / 1
half 0 / 0 (n/a)
float 1 / 1
double 1 / 1 (cl_khr_fp64)
Half-precision Floating-point support (n/a)
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 Yes
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 2092105728 (1.948GiB)
Error Correction support No
Max memory allocation 523026432 (498.8MiB)
Unified memory for Host and Device No
Integrated memory (NV) No
Minimum alignment for any data type 128 bytes
Alignment of base address 4096 bits (512 bytes)
Global Memory cache type Read/Write
Global Memory cache size 245760 (240KiB)
Global Memory cache line size 128 bytes
Image support Yes
Max number of samplers per kernel 32
Max size for 1D images from buffer 268435456 pixels
Max 1D or 2D image array size 2048 images
Max 2D image size 16384x32768 pixels
Max 3D image size 16384x16384x16384 pixels
Max number of read image args 256
Max number of write image args 16
Local memory type Local
Local memory size 49152 (48KiB)
Registers per block (NV) 65536
Max number of constant args 9
Max constant buffer size 65536 (64KiB)
Max size of kernel argument 4352 (4.25KiB)
Queue properties
Out-of-order execution Yes
Profiling Yes
Prefer user sync for interop No
Profiling timer resolution 1000ns
Execution capabilities
Run OpenCL kernels Yes
Run native kernels No
Kernel execution timeout (NV) Yes
Concurrent copy and kernel execution (NV) Yes
Number of async copy engines 2
printf() buffer size 1048576 (1024KiB)
Built-in kernels (n/a)
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_fp64 cl_khr_byte_addressable_store cl_khr_icd cl_khr_gl_sharing cl_nv_compiler_options cl_nv_device_attribute_query cl_nv_pragma_unroll cl_nv_copy_opts cl_nv_create_buffer cl_khr_int64_base_atomics cl_khr_int64_extended_atomics
Platform Name Clover
Number of devices 0
NULL platform behavior
clGetPlatformInfo(NULL, CL_PLATFORM_NAME, ...) No platform
clGetDeviceIDs(NULL, CL_DEVICE_TYPE_ALL, ...) No platform
clCreateContext(NULL, ...) [default] No platform
clCreateContext(NULL, ...) [other] Success [NV]
clCreateContextFromType(NULL, CL_DEVICE_TYPE_DEFAULT) No platform
clCreateContextFromType(NULL, CL_DEVICE_TYPE_CPU) No devices found in platform
clCreateContextFromType(NULL, CL_DEVICE_TYPE_GPU) No platform
clCreateContextFromType(NULL, CL_DEVICE_TYPE_ACCELERATOR) No devices found in platform
clCreateContextFromType(NULL, CL_DEVICE_TYPE_CUSTOM) Invalid device type for platform
clCreateContextFromType(NULL, CL_DEVICE_TYPE_ALL) No platform
ag@AprilGrimoire /home/ag sudo -u foldingathome clinfo 23:00:07
Number of platforms 2
Platform Name NVIDIA CUDA
Platform Vendor NVIDIA Corporation
Platform Version OpenCL 1.2 CUDA 11.0.210
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_fp64 cl_khr_byte_addressable_store cl_khr_icd cl_khr_gl_sharing cl_nv_compiler_options cl_nv_device_attribute_query cl_nv_pragma_unroll cl_nv_copy_opts cl_nv_create_buffer cl_khr_int64_base_atomics cl_khr_int64_extended_atomics
Platform Extensions function suffix NV
Platform Name Clover
Platform Vendor Mesa
Platform Version OpenCL 1.1 Mesa 20.1.4
Platform Profile FULL_PROFILE
Platform Extensions cl_khr_icd
Platform Extensions function suffix MESA
Platform Name NVIDIA CUDA
Number of devices 1
Device Name GeForce GTX 1050
Device Vendor NVIDIA Corporation
Device Vendor ID 0x10de
Device Version OpenCL 1.2 CUDA
Driver Version 450.57
Device OpenCL C Version OpenCL C 1.2
Device Type GPU
Device Topology (NV) PCI-E, 01:00.0
Device Profile FULL_PROFILE
Device Available Yes
Compiler Available Yes
Linker Available Yes
Max compute units 5
Max clock frequency 1455MHz
Compute Capability (NV) 6.1
Device Partition (core)
Max number of sub-devices 1
Supported partition types None
Supported affinity domains (n/a)
Max work item dimensions 3
Max work item sizes 1024x1024x64
Max work group size 1024
[1] 25158 segmentation fault sudo -u foldingathome clinfo
Both root and foldingathome user can get lots of output from clinfo, but I got a segmentation fault with the foldingathome user.
The foldingathome user doesn't have a shell per default. Temporarily edit /etc/passwd and give the foldingathome user /bin/bash as a shell. Then you can test again. Remember to set the shell to false afterwards.
Online: GTX 1660 Super + occasional CPU folding in the cold.
Offline: Radeon HD 7770, GTX 1050 Ti 4G OC, RX580
gunnarre wrote:The foldingathome user doesn't have a shell per default. Temporarily edit /etc/passwd and give the foldingathome user /bin/bash as a shell. Then you can test again. Remember to set the shell to false afterwards.
I set it to /bin/bash, but the output didn't change.
Looks like, from the thread I listed above, that you still have to run FAHClient as root(?!) on Gentoo. That might be the same error that gives a segfault when you run clinfo as the fahclient user.
Manfred.Knick wrote:Gentoo bugs #714160 (1) and #712744 (2):
(...)
- In order to exploit CUDA, enable USE=+uvm for x11-drivers/nvidia-drivers-440.64:0/440
- No need to edit /opt/foldingathome/config.xml manually:
- Team 11298 ("Gentoo Linux Users Everywhere") already set as default
- (!) Do not use "/etc/init.d/foldingathome start" and do not "rc-update add foldingathome default" right now
- (!) nor it's systemd correspondants
- (!) instead (as root, unfortunately) start /opt/foldingathome/FAHClient
I would normally not recommend running FAHClient as root, especially since it auto-downloads software (the folding cores), but if you don't care about your system, you could try to stop FAHClient, and just cd into foldingathome's home directory and run FAHClient with sudo and see if it picks up OpenCL.
Edit: This appears to me to be something wrong with Gentoo's setup. Maybe you're better served with the official Linux download and just extracting the executables from that?
Online: GTX 1660 Super + occasional CPU folding in the cold.
Offline: Radeon HD 7770, GTX 1050 Ti 4G OC, RX580
gunnarre wrote:Looks like, from the thread I listed above, that you still have to run FAHClient as root(?!) on Gentoo. That might be the same error that gives a segfault when you run clinfo as the fahclient user.
Manfred.Knick wrote:Gentoo bugs #714160 (1) and #712744 (2):
(...)
- In order to exploit CUDA, enable USE=+uvm for x11-drivers/nvidia-drivers-440.64:0/440
- No need to edit /opt/foldingathome/config.xml manually:
- Team 11298 ("Gentoo Linux Users Everywhere") already set as default
- (!) Do not use "/etc/init.d/foldingathome start" and do not "rc-update add foldingathome default" right now
- (!) nor it's systemd correspondants
- (!) instead (as root, unfortunately) start /opt/foldingathome/FAHClient
I would normally not recommend running FAHClient as root, especially since it auto-downloads software (the folding cores), but if you don't care about your system, you could try to stop FAHClient, and just cd into foldingathome's home directory and run FAHClient with sudo and see if it picks up OpenCL.
Edit: This appears to me to be something wrong with Gentoo's setup. Maybe you're better served with the official Linux download and just extracting the executables from that?
Oops, I found that eselect-opencl is not deprecated, but conflicting with opencl-icd-loader. I switched to eselect-opencl, but the problem persists.
gunnarre wrote:Did running as root help? If not, I wonder if you'll need to install a full CUDA development environment to get the correct OpenCL files.
Running as root didn't help.
I've installed nvidia-cuda-sdk with opencl use flag enabled, but it didn't help either.
.