Page 1 of 1
failed to create a gpu enabled openmm context
Posted: Tue Dec 15, 2020 3:14 am
by nikitastepanov
Code: Select all
03:07:02:******************************* libFAH ********************************
03:07:02: Date: Oct 20 2020
03:07:02: Time: 20:36:39
03:07:02: Revision: 5ca109d295a6245e2a2f590b3d0085ad5e567aeb
03:07:02: Branch: master
03:07:02: Compiler: GNU 8.3.0
03:07:02: Options: -faligned-new -std=c++11 -fsigned-char -ffunction-sections
03:07:02: -fdata-sections -O3 -funroll-loops -fno-pie
03:07:02: Platform: linux2 5.8.0-1-amd64
03:07:02: Bits: 64
03:07:02: Mode: Release
03:07:02:****************************** FAHClient ******************************
03:07:02: Version: 7.6.21
03:07:02: Author: Joseph Coffland <joseph@cauldrondevelopment.com>
03:07:02: Copyright: 2020 foldingathome.org
03:07:02: Homepage: https://foldingathome.org/
03:07:02: Date: Oct 20 2020
03:07:02: Time: 20:39:00
03:07:02: Revision: 6efbf0e138e22d3963e6a291f78dcb9c6422a278
03:07:02: Branch: master
03:07:02: Compiler: GNU 8.3.0
03:07:02: Options: -faligned-new -std=c++11 -fsigned-char -ffunction-sections
03:07:02: -fdata-sections -O3 -funroll-loops -fno-pie
03:07:02: Platform: linux2 5.8.0-1-amd64
03:07:02: Bits: 64
03:07:02: Mode: Release
03:07:02: Args: --child /etc/fahclient/config.xml --run-as fahclient
03:07:02: --pid-file=/var/run/fahclient.pid --daemon
03:07:02: Config: /etc/fahclient/config.xml
03:07:02:******************************** CBang ********************************
03:07:02: Date: Oct 20 2020
03:07:02: Time: 18:37:59
03:07:02: Revision: 7e4ce85225d7eaeb775e87c31740181ca603de60
03:07:02: Branch: master
03:07:02: Compiler: GNU 8.3.0
03:07:02: Options: -faligned-new -std=c++11 -fsigned-char -ffunction-sections
03:07:02: -fdata-sections -O3 -funroll-loops -fno-pie -fPIC
03:07:02: Platform: linux2 5.8.0-1-amd64
03:07:02: Bits: 64
03:07:02: Mode: Release
03:07:02:******************************* System ********************************
03:07:02: CPU: Intel(R) Core(TM) i5-8400 CPU @ 2.80GHz
03:07:02: CPU ID: GenuineIntel Family 6 Model 158 Stepping 10
03:07:02: CPUs: 6
03:07:02: Memory: 7.64GiB
03:07:02: Free Memory: 7.25GiB
03:07:02: Threads: POSIX_THREADS
03:07:02: OS Version: 4.19
03:07:02: Has Battery: false
03:07:02: On Battery: false
03:07:02: UTC Offset: 6
03:07:02: PID: 642
03:07:02: CWD: /var/lib/fahclient
03:07:02: OS: Linux 4.19.0-13-amd64 x86_64
03:07:02: OS Arch: AMD64
03:07:02: GPUs: 2
03:07:02: GPU 0: Bus:0 Slot:2 Func:0 INTEL:1 CFL GT2 [UHD Graphics 630]
03:07:02: GPU 1: Bus:1 Slot:0 Func:0 NVIDIA:7 GP106 [GeForce GTX 1060 3GB] 3935
03:07:02: CUDA: Not detected: Failed to open dynamic library 'libcuda.so':
03:07:02: libcuda.so: cannot open shared object file: No such file or
03:07:02: directory
03:07:02:OpenCL Device 0: Platform:0 Device:0 Bus:NA Slot:NA Compute:2.0 Driver:1.3
03:07:02:***************************************
Re: failed to create a gpu enabled openmm context
Posted: Tue Dec 15, 2020 3:16 am
by nikitastepanov
Code: Select all
*********************** Log Started 2020-12-15T03:07:02Z ***********************
03:07:02:WARNING:FS01:Guessing ambiguous GPU to OpenCL device mapping for 01: gpu:0:2 CFL GT2 [UHD Graphics 630]. Consider upgrading your graphics driver or manually setting ``opencl-index`` in this slot's configuration.
03:07:02:WARNING:FS02:No CUDA or OpenCL 1.2+ support detected for GPU slot 02: gpu:1:0 GP106 [GeForce GTX 1060 3GB] 3935. Disabling.
03:07:02:WARNING:WU01:FS00:Changed SMP threads from 5 to 4 this can cause some work units to fail
03:07:02:ERROR:WU00:FS01:Exception: Could not get IP address for assign1.foldingathome.org: Temporary failure in name resolution
03:07:02:ERROR:WU00:FS01:Exception: Could not get IP address for assign2.foldingathome.org: Temporary failure in name resolution
03:07:02:ERROR:WU00:FS01:Exception: Could not get IP address for assign3.foldingathome.org: Temporary failure in name resolution
03:07:02:ERROR:WU00:FS01:Exception: Could not get IP address for assign4.foldingathome.org: Temporary failure in name resolution
03:07:02:ERROR:WU00:FS01:Exception: Failed to find any IP addresses for assignment servers
03:07:10:WU00:FS01:0x22:ERROR:125: Failed to create a GPU-enabled OpenMM Context.
03:07:10:WARNING:WU00:FS01:FahCore returned: BAD_WORK_UNIT (114 = 0x72)
03:07:14:WU02:FS01:0x22:ERROR:125: Failed to create a GPU-enabled OpenMM Context.
03:07:14:WARNING:WU02:FS01:FahCore returned: BAD_WORK_UNIT (114 = 0x72)
03:07:18:WU00:FS01:0x22:ERROR:125: Failed to create a GPU-enabled OpenMM Context.
03:07:18:WARNING:WU00:FS01:FahCore returned: BAD_WORK_UNIT (114 = 0x72)
03:07:23:WU02:FS01:0x22:ERROR:125: Failed to create a GPU-enabled OpenMM Context.
03:07:23:WARNING:WU02:FS01:FahCore returned: BAD_WORK_UNIT (114 = 0x72)
03:07:55:WU00:FS01:0x22:ERROR:125: Failed to create a GPU-enabled OpenMM Context.
03:07:56:WARNING:WU00:FS01:FahCore returned: BAD_WORK_UNIT (114 = 0x72)
03:07:59:WU02:FS01:0x22:ERROR:125: Failed to create a GPU-enabled OpenMM Context.
03:07:59:WARNING:WU02:FS01:FahCore returned: BAD_WORK_UNIT (114 = 0x72)
03:08:03:WU00:FS01:0x22:ERROR:125: Failed to create a GPU-enabled OpenMM Context.
03:08:03:WARNING:WU00:FS01:FahCore returned: BAD_WORK_UNIT (114 = 0x72)
03:08:07:WU02:FS01:0x22:ERROR:125: Failed to create a GPU-enabled OpenMM Context.
03:08:07:WARNING:WU02:FS01:FahCore returned: BAD_WORK_UNIT (114 = 0x72)
03:08:11:WU00:FS01:0x22:ERROR:125: Failed to create a GPU-enabled OpenMM Context.
03:08:11:WARNING:WU00:FS01:FahCore returned: BAD_WORK_UNIT (114 = 0x72)
03:08:15:WU02:FS01:0x22:ERROR:125: Failed to create a GPU-enabled OpenMM Context.
03:08:15:WARNING:WU02:FS01:FahCore returned: BAD_WORK_UNIT (114 = 0x72)
Re: failed to create a gpu enabled openmm context
Posted: Tue Dec 15, 2020 4:56 am
by psaam0001
What distribution of Linux are you using (Fedora, Debian, Ubuntu, or other)?
The information that is related to the distribution you are using will be important, as the guidance we give here will be specific to it (in regards to installing the NVidia and Intel WHQL tested drivers).
Paul
Re: failed to create a gpu enabled openmm context
Posted: Tue Dec 15, 2020 10:39 am
by ph0b
Version 1.3 doesn't sound right, maybe it's pocl or beignet OpenCL runtimes, both of them aren't the one to use here for OpenMM.
So it seems you're missing OpenCL and Cuda runtimes for both of your GPUs, as previously asked please tell us your distribution and we can provide guidance.
For Ubuntu/Debian and Intel GPUs,
Code: Select all
apt install intel-opencl-icd
apt install --no-install-recommends ocl-icd-opencl-dev
will give what you need.
Re: failed to create a gpu enabled openmm context
Posted: Sun Dec 27, 2020 12:39 am
by bruce
You'll need the proprietary drivers for both Intel and NVidia to get full support for OpenCL. As mentioned above, you'll need the correct Intel proprietary drivers. You'll also need NVidia's proprietary drivers which include support for both OpenCL and CUDA.