Can't download Nvidia gpu work

Roger K8YPA
Roger K8YPA
Joined: 28 Feb 05
Posts: 6
Credit: 1015794647
RAC: 0
Topic 202974

Can't download Nvidia gpu work

I am using Fedora 24 Linux

computer:    https://einsteinathome.org/host/12443111

Log:
Sun 20 Nov 2016 08:10:08 PM EST |  | cc_config.xml not found - using defaults
Sun 20 Nov 2016 08:10:08 PM EST |  | Starting BOINC client version 7.6.22 for x86_64-pc-linux-gnu
Sun 20 Nov 2016 08:10:08 PM EST |  | log flags: file_xfer, sched_ops, task
Sun 20 Nov 2016 08:10:08 PM EST |  | Libraries: libcurl/7.47.1 NSS/3.27 zlib/1.2.8 libidn/1.33 libpsl/0.13.0 (+libidn/1.33) libssh2/1.7.0 nghttp2/1.7.1
Sun 20 Nov 2016 08:10:08 PM EST |  | Running as a daemon
Sun 20 Nov 2016 08:10:08 PM EST |  | Data directory: /var/lib/boinc
Sun 20 Nov 2016 08:10:11 PM EST |  | OpenCL: NVIDIA GPU 0: GeForce GTX 650 (driver version 375.20, device version OpenCL 1.2 CUDA, 1009MB, 1009MB available, 203 GFLOPS peak)
Sun 20 Nov 2016 08:10:11 PM EST |  | Host name: Win7
Sun 20 Nov 2016 08:10:11 PM EST |  | Processor: 8 GenuineIntel Intel(R) Core(TM) i7 CPU         920  @ 2.67GHz [Family 6 Model 26 Stepping 5]
Sun 20 Nov 2016 08:10:11 PM EST |  | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf eagerfpu pni dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm sse4_1 sse4_2 popcnt lahf_lm tpr_shadow vnmi flexpriority ept vpid dtherm ida
Sun 20 Nov 2016 08:10:11 PM EST |  | OS: Linux: 4.8.7-200.fc24.x86_64
Sun 20 Nov 2016 08:10:11 PM EST |  | Memory: 7.79 GB physical, 2.95 GB virtual
Sun 20 Nov 2016 08:10:11 PM EST |  | Disk: 42.66 GB total, 26.99 GB free
Sun 20 Nov 2016 08:10:11 PM EST |  | Local time is UTC -5 hours
Sun 20 Nov 2016 08:10:11 PM EST | Einstein@Home | URL http://einstein.phys.uwm.edu/; Computer ID 12443111; resource share 100
Sun 20 Nov 2016 08:10:11 PM EST | SETI@home | URL http://setiathome.berkeley.edu/; Computer ID 8116231; resource share 100
Sun 20 Nov 2016 08:10:11 PM EST | Einstein@Home | General prefs: from Einstein@Home (last modified 05-Aug-2016 22:54:51)
Sun 20 Nov 2016 08:10:11 PM EST | Einstein@Home | Computer location: home
Sun 20 Nov 2016 08:10:11 PM EST |  | General prefs: using separate prefs for home
Sun 20 Nov 2016 08:10:11 PM EST |  | Reading preferences override file
Sun 20 Nov 2016 08:10:11 PM EST |  | Preferences:
Sun 20 Nov 2016 08:10:11 PM EST |  | max memory usage when active: 7981.07MB
Sun 20 Nov 2016 08:10:11 PM EST |  | max memory usage when idle: 7981.07MB
Sun 20 Nov 2016 08:10:11 PM EST |  | max disk usage: 26.22GB
Sun 20 Nov 2016 08:10:11 PM EST |  | max CPUs used: 1
Sun 20 Nov 2016 08:10:11 PM EST |  | (to change preferences, visit a project web site or select Preferences in the Manager)
Sun 20 Nov 2016 08:10:11 PM EST |  | gui_rpc_auth.cfg is empty - no GUI RPC password protection
Sun 20 Nov 2016 08:10:12 PM EST | Einstein@Home | Sending scheduler request: To fetch work.
Sun 20 Nov 2016 08:10:12 PM EST | Einstein@Home | Requesting new tasks for NVIDIA GPU
Sun 20 Nov 2016 08:10:16 PM EST | Einstein@Home | Scheduler request completed: got 0 new tasks
Sun 20 Nov 2016 08:10:16 PM EST | Einstein@Home | No work sent
Sun 20 Nov 2016 08:10:16 PM EST | Einstein@Home | No work is available for Multi-Directed Continuous Gravitational Wave search Tuning run CV
Sun 20 Nov 2016 08:10:16 PM EST | Einstein@Home | No work is available for Multi-Directed Continuous Gravitational Wave search Tuning run CV
Sun 20 Nov 2016 08:10:16 PM EST | Einstein@Home | No work is available for Multi-Directed Continuous Gravitational Wave search Tuning run CV
Sun 20 Nov 2016 08:10:16 PM EST | Einstein@Home | No work is available for Multi-Directed Continuous Gravitational Wave search Tuning run CV
Sun 20 Nov 2016 08:10:16 PM EST | Einstein@Home | see scheduler log messages on https://einsteinathome.org/host/12443111/log
Sun 20 Nov 2016 08:10:16 PM EST | Einstein@Home | Binary Radio Pulsar Search (Arecibo) is not available for your type of computer.
Sun 20 Nov 2016 08:21:20 PM EST | Einstein@Home | Sending scheduler request: To fetch work.
Sun 20 Nov 2016 08:21:20 PM EST | Einstein@Home | Requesting new tasks for NVIDIA GPU
Sun 20 Nov 2016 08:21:24 PM EST | Einstein@Home | Scheduler request completed: got 0 new tasks
Sun 20 Nov 2016 08:21:24 PM EST | Einstein@Home | No work sent
Sun 20 Nov 2016 08:21:24 PM EST

 

Please help

Roger    https://einsteinathome.org/account/46992

Richie
Richie
Joined: 7 Mar 14
Posts: 656
Credit: 1702989778
RAC: 0

My amateurish suggestion is

My amateurish suggestion is there could be something wrong with the CUDA recognition, even if the Nvidia card was reported at the log lines. 

 

2016-11-21 02:28:32.8384 [PID=11222] [version] Checking plan class 'BRP4G-cuda32-nv270' 2016-11-21 02:28:32.8415 [PID=11222] [version] reading plan classes from file '/BOINC/projects/EinsteinAtHome/plan_class_spec.xml' 2016-11-21 02:28:32.8415 [PID=11222] [version] parsed project prefs setting 'gpu_util_brp': 1.000000 2016-11-21 02:28:32.8416 [PID=11222] [version] CUDA compute capability required min: 100, supplied: 0

and later again

2016-11-21 02:28:32.8417 [PID=11222] [version] Checking plan class 'BRP4G-cuda32-nv270' 2016-11-21 02:28:32.8417 [PID=11222] [version] parsed project prefs setting 'gpu_util_brp': 1.000000 2016-11-21 02:28:32.8417 [PID=11222] [version] CUDA compute capability required min: 100, supplied: 0

Has that same host been able to download Nvidia GPU work at some point earlier? Have you recently updated kernel, Nvidia driver or something else?

For all the Beta-apps there was

2016-11-21 02:28:32.8416 [PID=11222] [version] beta test app versions not allowed in project prefs.

 
Maybe You could try if enabling Beta work at Project Preferences allowed that host to get GPU tasks.
Roger K8YPA
Roger K8YPA
Joined: 28 Feb 05
Posts: 6
Credit: 1015794647
RAC: 0

Thanks for the reply. I

Thanks for the reply.

I enabled the beta work in preferences. Still get the same result.

I just recently set up this fedora account so it has never downloaded Nvidia work.

This is a dual boot computer. It downloads Nvidia ok when I boot into Windows 10.

I see what your saying about CUDA recognition but do not know what to do about it.

 

 

Richie
Richie
Joined: 7 Mar 14
Posts: 656
Credit: 1702989778
RAC: 0

That host has

That host has kernel 4.8.7-200.fc24.x86_64 and Nvidia driver version 375.20. Both are very new, perhaps too much.

Roger K8YPA
Roger K8YPA
Joined: 28 Feb 05
Posts: 6
Credit: 1015794647
RAC: 0

I tried an older kernel and

I tried an older kernel and an older driver before using this one. Same result no Nvidia work.

 

MarkJ
MarkJ
Joined: 28 Feb 08
Posts: 437
Credit: 139002861
RAC: 0

Looking at the log output you

OpenCL: NVIDIA GPU 0: GeForce GTX 650 (driver version 375.20, device version OpenCL 1.2 CUDA, 1009MB, 1009MB available, 203 GFLOPS peak)

Looking at the log output you don't have CUDA. You have an OpenCL GPU only. The fact that Nvidia refer to it as "device version OpenCL 1.2 CUDA" is a bit misleading. I would have thought you should be able to get OpenCL work for it but maybe your preferences don't allow it.

Roger K8YPA
Roger K8YPA
Joined: 28 Feb 05
Posts: 6
Credit: 1015794647
RAC: 0

Thank youI wil see if

Thank you

I wil see if Fedora has a cuda package I can Install.

 

After installing a different driver and a cuda package both from Fedora I was still not able to get any gpu work. Originally I was using a driver I downloaded fromNvidias website.

 

Roger K8YPA
Roger K8YPA
Joined: 28 Feb 05
Posts: 6
Credit: 1015794647
RAC: 0

I found out that SELinux was

I found out that SELinux was blocking the gpu work. After disabling SELinux I was able to do gpu work now with no problem.

Log:

Wed 23 Nov 2016 01:09:02 PM EST |  | Starting BOINC client version 7.6.22 for x86_64-pc-linux-gnu
Wed 23 Nov 2016 01:09:02 PM EST |  | log flags: file_xfer, sched_ops, task
Wed 23 Nov 2016 01:09:02 PM EST |  | Libraries: libcurl/7.47.1 NSS/3.27 zlib/1.2.8 libidn/1.33 libpsl/0.13.0 (+libidn/1.33) libssh2/1.7.0 nghttp2/1.7.1
Wed 23 Nov 2016 01:09:02 PM EST |  | Running as a daemon
Wed 23 Nov 2016 01:09:02 PM EST |  | Data directory: /var/lib/boinc
Wed 23 Nov 2016 01:09:09 PM EST |  | CUDA: NVIDIA GPU 0: GeForce GTX 650 (driver version 367.57, CUDA version 8.0, compute capability 3.0, 979MB, 943MB available, 813 GFLOPS peak)
Wed 23 Nov 2016 01:09:09 PM EST |  | OpenCL: NVIDIA GPU 0: GeForce GTX 650 (driver version 367.57, device version OpenCL 1.2 CUDA, 979MB, 943MB available, 813 GFLOPS peak)
Wed 23 Nov 2016 01:09:09 PM EST |  | Host name: Win7
Wed 23 Nov 2016 01:09:09 PM EST |  | Processor: 8 GenuineIntel Intel(R) Core(TM) i7 CPU         920  @ 2.67GHz [Family 6 Model 26 Stepping 5]
Wed 23 Nov 2016 01:09:09 PM EST |  | Processor features: fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf eagerfpu pni dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm sse4_1 sse4_2 popcnt lahf_lm tpr_shadow vnmi flexpriority ept vpid dtherm ida
Wed 23 Nov 2016 01:09:09 PM EST |  | OS: Linux: 4.8.8-200.fc24.x86_64
Wed 23 Nov 2016 01:09:09 PM EST |  | Memory: 7.79 GB physical, 2.95 GB virtual
Wed 23 Nov 2016 01:09:09 PM EST |  | Disk: 42.66 GB total, 25.45 GB free
Wed 23 Nov 2016 01:09:09 PM EST |  | Local time is UTC -5 hours
Wed 23 Nov 2016 01:09:09 PM EST | Einstein@Home | URL http://einstein.phys.uwm.edu/; Computer ID 12443111; resource share 100
Wed 23 Nov 2016 01:09:09 PM EST | SETI@home | URL http://setiathome.berkeley.edu/; Computer ID 8116231; resource share 100
Wed 23 Nov 2016 01:09:09 PM EST | Einstein@Home | General prefs: from Einstein@Home (last modified 05-Aug-2016 22:54:51)
Wed 23 Nov 2016 01:09:09 PM EST | Einstein@Home | Computer location: home
Wed 23 Nov 2016 01:09:09 PM EST |  | General prefs: using separate prefs for home
Wed 23 Nov 2016 01:09:09 PM EST |  | Reading preferences override file
Wed 23 Nov 2016 01:09:09 PM EST |  | Preferences:
Wed 23 Nov 2016 01:09:09 PM EST |  | max memory usage when active: 7981.09MB
Wed 23 Nov 2016 01:09:09 PM EST |  | max memory usage when idle: 7981.09MB
Wed 23 Nov 2016 01:09:10 PM EST |  | max disk usage: 24.68GB
Wed 23 Nov 2016 01:09:10 PM EST |  | max CPUs used: 1
Wed 23 Nov 2016 01:09:10 PM EST |  | (to change preferences, visit a project web site or select Preferences in the Manager)
Wed 23 Nov 2016 01:09:10 PM EST |  | gui_rpc_auth.cfg is empty - no GUI RPC password protection
Wed 23 Nov 2016 01:09:10 PM EST | Einstein@Home | Sending scheduler request: To fetch work.
Wed 23 Nov 2016 01:09:10 PM EST | Einstein@Home | Requesting new tasks for NVIDIA GPU
Wed 23 Nov 2016 01:09:14 PM EST | Einstein@Home | Scheduler request completed: got 1 new tasks
Wed 23 Nov 2016 01:09:16 PM EST | Einstein@Home | Started download of einsteinbinary_BRP4G_1.39_x86_64-pc-linux-gnu__BRP4G-cuda32-nv270
Wed 23 Nov 2016 01:09:16 PM EST | Einstein@Home | Started download of libcudart64_32_16.so
Wed 23 Nov 2016 01:09:19 PM EST | Einstein@Home | Finished download of libcudart64_32_16.so
Wed 23 Nov 2016 01:09:19 PM EST | Einstein@Home | Started download of libcufft64_32_16.so
Wed 23 Nov 2016 01:09:34 PM EST | Einstein@Home | Finished download of einsteinbinary_BRP4G_1.39_x86_64-pc-linux-gnu__BRP4G-cuda32-nv270

AgentB
AgentB
Joined: 17 Mar 12
Posts: 915
Credit: 513211304
RAC: 0

Roger K8YPA wrote:I found out

Roger K8YPA wrote:

I found out that SELinux was blocking the gpu work. After disabling SELinux I was able to do gpu work now with no problem.

Sorry to be late on the  thread i'm not a Federa user, you might also have some success with adding boinc user to the video (or gpu) group but if you look at the manual  - https://boinc.berkeley.edu/wiki/Installing_BOINC_on_Fedora - you'll see the SElinux disabling being needed so you have the right solution.

Quote:
Wed 23 Nov 2016 01:09:09 PM EST |  | Host name: Win7

made me chuckle.

 

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.