Page 5 of 8

Re: Waiting for work on 192.0.2.1 ?

Posted: Mon Aug 03, 2020 7:48 am
by zniper
bruce wrote:Apparently NVidia proprietary drivers install CUDA. FAHCore_22 does not currently use CUDA, it requires OpenCL. On Windows, when you install the NVidia proprietary drivers, you get both CUDA and OpenCL. On linux, I think those are 3 separate packages.

Nevertheless, if you're certain you were folding with FAHCore_22 on your GPU you must have had OpenCL installed from somewhere. There's a known bug, though: If your GPU is misconfigured, it's possible for it to appear to download a WU and run it, but it actually runs on the CPU, not on the GPU.

Bottom line: Errors involving 192.0.2.1 mean that a device cannot be used because FAH did not detect it about 100 lines from the beginning of FAH's log -- which is why I need to see that configuration step to help you.
my fah log is 4 post above your, opencl-nvidia is installed on my system, it was running before with over 2 mill ppd so i don't think my cpu was doing the work.

clinfo output:

Code: Select all

Number of platforms                               2
  Platform Name                                   NVIDIA CUDA
  Platform Vendor                                 NVIDIA Corporation
  Platform Version                                OpenCL 1.2 CUDA 10.2.185
  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 20.1.3
  Platform Profile                                FULL_PROFILE
  Platform Extensions                             cl_khr_icd
  Platform Extensions function suffix             MESA

  Platform Name                                   NVIDIA CUDA
Number of devices                                 2
  Device Name                                     GeForce RTX 2070 SUPER
  Device Vendor                                   NVIDIA Corporation
  Device Vendor ID                                0x10de
  Device Version                                  OpenCL 1.2 CUDA
  Driver Version                                  440.100
  Device OpenCL C Version                         OpenCL C 1.2 
  Device Type                                     GPU
  Device Topology (NV)                            PCI-E, 0a:00.0
  Device Profile                                  FULL_PROFILE
  Device Available                                Yes
  Compiler Available                              Yes
  Linker Available                                Yes
  Max compute units                               40
  Max clock frequency                             1800MHz
  Compute Capability (NV)                         7.5
  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                              8370061312 (7.795GiB)
  Error Correction support                        No
  Max memory allocation                           2092515328 (1.949GiB)
  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                        1310720 (1.25MiB)
  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                             32768x32768 pixels
    Max 3D image size                             16384x16384x16384 pixels
    Max number of read image args                 256
    Max number of write image args                32
  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                  3
  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

  Device Name                                     GeForce GTX 1070
  Device Vendor                                   NVIDIA Corporation
  Device Vendor ID                                0x10de
  Device Version                                  OpenCL 1.2 CUDA
  Driver Version                                  440.100
  Device OpenCL C Version                         OpenCL C 1.2 
  Device Type                                     GPU
  Device Topology (NV)                            PCI-E, 0b:00.0
  Device Profile                                  FULL_PROFILE
  Device Available                                Yes
  Compiler Available                              Yes
  Linker Available                                Yes
  Max compute units                               15
  Max clock frequency                             1746MHz
  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                              8510832640 (7.926GiB)
  Error Correction support                        No
  Max memory allocation                           2127708160 (1.982GiB)
  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                        737280 (720KiB)
  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
ls /etc/OpenCL/vendors output:

Code: Select all

mesa.icd  nvidia.icd

Re: Waiting for work on 192.0.2.1 ?

Posted: Mon Aug 03, 2020 8:05 am
by ajm
We see similar problems on a selection of setups, but not a general back end problem. The servers still are assigning thousands of jobs per hour: https://apps.foldingathome.org/serverstats
In my case, for example, the problem was clearly linked to the latest AMD drivers: installing them brought it about on a machine that was folding normally, and replacing them solved it.

Or would it be possible that the problem only appears with some WS, or only WS with a specific software version? There is quite a variety of those versions now.

Re: Waiting for work on 192.0.2.1 ?

Posted: Mon Aug 03, 2020 8:10 am
by zniper
uninstalled opencl-mesa and now booth my gpus and cpu are folding again,

Re: Waiting for work on 192.0.2.1 ?

Posted: Mon Aug 03, 2020 9:46 am
by HonzikS74
It works.Linux Mint.
sudo apt-get remove mesa-opencl-icd

Re: Waiting for work on 192.0.2.1 ?

Posted: Mon Aug 03, 2020 11:00 am
by goodyca
I can confirm that removing mesa-libOpenCL on Fedora 32 on all 3 of my computers got GPU folding back to working. I also had to create a soft link for libOpenCL.so in /usr/lib64

Re: Waiting for work on 192.0.2.1 ?

Posted: Mon Aug 03, 2020 11:23 am
by Jupis_folder
Same here. Removing the package mesa-opencl-icd fixed the Problem. The question that remains: Why did this configuration run for month without problems and no longer since 1. Aug? Even without updating the system?

Re: Waiting for work on 192.0.2.1 ?

Posted: Mon Aug 03, 2020 1:30 pm
by foldingfanmucde
I'm happy for the linux fraction - that many of you are up and running again. Even if it is not fully known why yet.
The other open issue is the Windows fraction, where I believe (unless I missed something) there is no generally applicable fix yet.
Thanks to all of you who are continuing to look into this !

Re: Waiting for work on 192.0.2.1 ?

Posted: Mon Aug 03, 2020 2:04 pm
by Jupis_folder
I have no clue about how to deal with windows, but let me sum up the linux side:
1. You need a working OpenCL-ICD. This is done by installing the dev package of the ICD.
2. You need exactly one OpenCL-ICD. So if you are on nvidia, remove ICD of AMD and the generic mesa ICD, if installed.
Why this is needed now, but not last week, remains unclear. If I extrapolate it, you may have a look at your windows system, if you have a seconds OpenCl installation, maybe a leftover from another GPU you used in past. If you are on NVidia GPU, it may happen that you have an OpenCL installation of you integrated (AMD|Intell)GPU.

Re: Waiting for work on 192.0.2.1 ?

Posted: Mon Aug 03, 2020 2:19 pm
by ajm
Well, the OpenCL dev package has been a must for new Linux installations for several months already, at least since April. There are countless cases on this forum. And FAH having to be added to the video and/or render groups (or to be ran as root) was also all the rage this spring.

For Windows, a good dose of DDU -- https://www.guru3d.com/files-details/di ... nload.html -- and maybe a better choice of drivers should do the trick. I'd just recommend to uninstall FAH before and reinstall it afterwards, as FAH is better at detecting hardware during installation than at acknowledging changes.

Re: Waiting for work on 192.0.2.1 ?

Posted: Mon Aug 03, 2020 2:21 pm
by Endgame124
goodyca wrote:I can confirm that removing mesa-libOpenCL on Fedora 32 on all 3 of my computers got GPU folding back to working. I also had to create a soft link for libOpenCL.so in /usr/lib64
Can you expand on the softlink you created? I removed mesa-libOpenCL on my Fedora 30 box, but that alone did not resolve the issue.

Re: Waiting for work on 192.0.2.1 ?

Posted: Mon Aug 03, 2020 3:17 pm
by goodyca
Endgame124 wrote:
goodyca wrote:I can confirm that removing mesa-libOpenCL on Fedora 32 on all 3 of my computers got GPU folding back to working. I also had to create a soft link for libOpenCL.so in /usr/lib64
Can you expand on the softlink you created? I removed mesa-libOpenCL on my Fedora 30 box, but that alone did not resolve the issue.
I ran "ln -s libOpenCL.so.1.0.0 libOpenCL.so" from the /usr/lib64 directory.

Re: Waiting for work on 192.0.2.1 ?

Posted: Mon Aug 03, 2020 3:24 pm
by bruce
@Jupis_folder

You're right to question whether an Integrated GPUs like your Athlon 200GE can be supported. It's not currently listed in GPUs.txt so (at least for right now) it's not. Could it be? Right now I can't answer that question.

Re: Waiting for work on 192.0.2.1 ?

Posted: Mon Aug 03, 2020 3:59 pm
by Endgame124
goodyca wrote:
Endgame124 wrote:
goodyca wrote:I can confirm that removing mesa-libOpenCL on Fedora 32 on all 3 of my computers got GPU folding back to working. I also had to create a soft link for libOpenCL.so in /usr/lib64
Can you expand on the softlink you created? I removed mesa-libOpenCL on my Fedora 30 box, but that alone did not resolve the issue.
I ran "ln -s libOpenCL.so.1.0.0 libOpenCL.so" from the /usr/lib64 directory.
Softlink was already there, so I guess mesa wasn't the problem. I had previously set openCL index (it wouldn't detect automatically), and tried again, but that didn't work either. Worst part is, this was working fine for quite some time, so I guess I go without folding for now.

Re: Waiting for work on 192.0.2.1 ?

Posted: Mon Aug 03, 2020 4:20 pm
by goodyca
Endgame124 wrote: Softlink was already there, so I guess mesa wasn't the problem. I had previously set openCL index (it wouldn't detect automatically), and tried again, but that didn't work either. Worst part is, this was working fine for quite some time, so I guess I go without folding for now.
I have gpu-index set to -1 and opencl-index set to 0

Be sure to restart the FAHClient.service after making changes.

Re: Waiting for work on 192.0.2.1 ?

Posted: Mon Aug 03, 2020 4:38 pm
by Jupis_folder
@ajm: These threads of OpenCL dev package do not cover the main questions:
1. What is the difference in folding on GPU after 1. Aug 2:00 UTC, which was the time when the problem aroused, after it ran fine for months.
2. Why does the problem occur on Linux and Windows at the same time, while we cannot install the OpenCL dev package on Windows.

@bruce: I did not ask for supporting the GPU of Athlon200GE, I just doubted that this CPU was capable to accidentially crunch the GPU WUs with the speed of the big NVidia cards. Additionally, the nvidia-smi tool showed that the NVidia cards are used for FahCore_22.