Posts by Raistmer

1) Message boards : Number crunching : OpenCL kernel/call clGetEventProfilingInfo call failed (Message 1842932)
Posted 2 days ago by Profile Raistmer
Post:
I didn't spot any issues in code so far. That is, it's marked as driver/config/card issue still.
But this kind of error allows normal processing after recover so in revision 3602 this error will be recoverable one, w/o process termination.
App will just inform about error in stderr and continue.
So far no Windows-based app reported such issue so I will not rebuild Windows binaries.
Maybe Urs could provide standalone updated binary for Linux hosts with drivers/cards affected by this profiling abilities loss.
2) Message boards : Number crunching : OpenCL kernel/call clGetEventProfilingInfo call failed (Message 1842839)
Posted 3 days ago by Profile Raistmer
Post:
Thanks for report. Such transient error could mean issue with app itself.
Will look closely on that area.

EDIT: there is third class of results for that host - the ones with -7 error but been able to complete after restart from that error. Insuch case result validates OK.
So, the place in processing sequence sometimes can be passed for relatively few attempts sometimes almost impassable (or there are lot of such points per single task),

There are results w/o restarts at all also. And AR value doesn't look drastically different for all these cases to trigger different behavior.
3) Message boards : Number crunching : Mac OS Sierra (Message 1842662)
Posted 4 days ago by Profile Raistmer
Post:

In my tests a few month ago avx2 was slower than avx. Probably Joe Segur was correct in assumming exactly that after finishing his work on the avx CPU-apps.

AFAIK AVX2 mostly adds integer 256 bits to instructions set. We mostly use float ones instead.
There are few new instructions that can be useful for us too like gather or permutations but testing needed.
Accordingly to VC++ dev blog thay didn't use gather at least on Haswell cause though it was implemented it was slower than other ways to do same work.
Nothing new in that. Support of some assembly instruction doesn't mean it will be faster to use it.
Same was once on Venice that supported SSE3 but so inefficiently that better was to stay with SSE2 only.
4) Message boards : Number crunching : SETI@home v8.22 Windows GPU applications support thread (Message 1842653)
Posted 4 days ago by Profile Raistmer
Post:

Is this right?
If so, many other hosts with ATI GPU could still have this problem unless they reset or new version app comes out?

yes, that's why Eric attempted to re-release under new version but that didn't work for some reason.
5) Message boards : Number crunching : OpenCL kernel/call clGetEventProfilingInfo call failed (Message 1842272)
Posted 6 days ago by Profile Raistmer
Post:
reboot as Urs suggested - helped?
6) Message boards : Number crunching : SETI@home v8.22 Windows GPU applications support thread (Message 1842270)
Posted 6 days ago by Profile Raistmer
Post:
Sorry, they do not run on this machine:
http://setiathome.berkeley.edu/show_host_detail.php?hostid=7328450


the same. do project reset
7) Message boards : Number crunching : SETI@home v8.22 Windows GPU applications support thread (Message 1842269)
Posted 6 days ago by Profile Raistmer
Post:
I wonder why setiathome_8.22_windows_intelx86__opencl_nvidia_SoG.exe is in SETI folder.
Could this file be the reason of this error?
Maybe I should reset project in BOINC Manager.

yes, do reset.
8) Message boards : Number crunching : SETI@home v8.22 Windows GPU applications support thread (Message 1842133)
Posted 7 days ago by Profile Raistmer
Post:
Please post issues and questions regarding Windows OpenCL apps here.
9) Message boards : Number crunching : SETI@home v8.22/23 Windows GPU applications support thread (Message 1842132)
Posted 7 days ago by Profile Raistmer
Post:
This thread happened to be more about deployment process support instead of support of app itself. So I'll ask of locking this thread and start refreshed one.
Deployment looks OK now.
10) Message boards : Number crunching : SETI@home v8.22/23 Windows GPU applications support thread (Message 1842130)
Posted 7 days ago by Profile Raistmer
Post:
AFAIK R9 Nano is AMD, not NV GPU so app shouldn't ask NV device capabilities.

I need first dozen of lines of stderr. WEB result has only 64k of last lines, first are lost.

My own Kalindi-based netbook received setiathome_8.22_windows_intelx86__opencl_ati5_SoG.exe (8.22 opencl_ati5_SoG_nocal) app.

Seems it started processing OK:

Running on device number: 0
Priority of worker thread raised successfully
Priority of process adjusted successfully, below normal priority class used
OpenCL platform detected: Advanced Micro Devices, Inc.
BOINC assigns device 0
0 slot of 64 used for this instance
Info: BOINC provided OpenCL device ID used
Info: CPU affinity mask used: 1; system mask is 3

Build features: SETI8 Non-graphics OpenCL USE_OPENCL_HD5xxx OCL_ZERO_COPY SIGNALS_ON_GPU OCL_CHIRP3 FFTW AMD specific USE_SSE2 x86
CPUID: AMD A4-1250 APU with Radeon(TM) HD Graphics

Cache: L1=64K L2=1024K

CPU features: FPU TSC PAE CMPXCHG8B APIC SYSENTER MTRR CMOV/CCMP MMX FXSAVE/FXRSTOR SSE SSE2 HT SSE3 SSSE3 SSE4.1 SSE4.2 AVX SSE4A
Low-performance GPU detected, default period_iterations_num set to 500
OpenCL-kernels filename : MultiBeam_Kernels_r3584.cl
INFO: can't open binary kernel file: C:\BOINCdata/projects/setiathome.berkeley.edu\MultiBeam_Kernels_r3584.clHD5_Kalindi.bin_V7_SoG_13485VM, continue with recompile...
Info : Building Program (binary, clBuildProgram):main kernels: OK code 0
INFO: binary kernel file created
ar=0.447658 NumCfft=191295 NumGauss=1055534808 NumPulse=226412070870 NumTriplet=452820396964
Currently allocated 229 MB for GPU buffers
In v_BaseLineSmooth: NumDataPoints=1048576, BoxCarLength=8192, NumPointsInChunk=32768


So, right binary released under this plan class.

Issues with R9 Nano are host config dependend. First lines from stderr still needed for troubleshooting.
11) Message boards : Number crunching : OpenCL kernel/call clGetEventProfilingInfo call failed (Message 1842035)
Posted 7 days ago by Profile Raistmer
Post:

I don't understand what I could do with the *bin* files (96 of them in the directory) or the *wisdom* files (6 of them present).

delete them all
12) Message boards : Number crunching : OpenCL kernel/call clGetEventProfilingInfo call failed (Message 1841977)
Posted 7 days ago by Profile Raistmer
Post:
Wait. I just realized: Raistmer, I don't understand your
#define line means for me to do/try.


It decodes -7 error.
Now your OpenCL runtime lack of profiling support.
It's abnormal, especially if it was before.
13) Message boards : Number crunching : OpenCL kernel/call clGetEventProfilingInfo call failed (Message 1841821)
Posted 8 days ago by Profile Raistmer
Post:
#define CL_PROFILING_INFO_NOT_AVAILABLE -7

reboot or rollback.
14) Message boards : Number crunching : SETI@home v8.22/23 Windows GPU applications support thread (Message 1841727)
Posted 8 days ago by Profile Raistmer
Post:
Can't check with own PC until tomorrow but app statistics look suspicious right now:

Windows/x86 8.24 (opencl_ati5_SoG_cat132) 28 Dec 2016, 23:34:07 UTC 342 GigaFLOPS
Windows/x86 8.24 (opencl_ati5_SoG_nocal) 28 Dec 2016, 23:34:07 UTC 342 GigaFLOPS

Same value has one of 8.22 plan classes.
Looks like 342 was initial value and it remains the same. That would mean no real results returned still.

Also, wrong date of release. Usually new app version has currrent date, 8.24 both have date of 8.22 release...
15) Message boards : Number crunching : Windows 10 - Yea or Nay? (Message 1841612)
Posted 9 days ago by Profile Raistmer
Post:
VS2015 is as good in leaking your private info as Win10 is:
http://stackoverflow.com/questions/31452435/how-do-i-disable-vshub-exe-in-the-system-tray
Especially representative:
https://i.stack.imgur.com/Qp7My.jpg
So, M$ collects info who is working on what !
By default w/o any direct permissions like "participate in app improvement".
Disgusting move from M$ but could anyone still expect smth else?...

Now my concerns are will SETI binary built with VS2015 spy on users and send telemetry to M$ or whatever structure they will sell access?...
16) Message boards : Number crunching : SETI@home v8.22/23 Windows GPU applications support thread (Message 1841475)
Posted 9 days ago by Profile Raistmer
Post:
I'm afraid not:

1/11/2017 9:42:50 PM | SETI@home | Sending scheduler request: Requested by user.
1/11/2017 9:42:50 PM | SETI@home | Requesting new tasks for AMD/ATI GPU
1/11/2017 9:42:54 PM | SETI@home | Scheduler request completed: got 4 new tasks
1/11/2017 9:42:54 PM | SETI@home | [error] No app version found for app setiathome_v8 platform windows_intelx86 ver 823 class opencl_ati5_SoG_nocal; discarding 08ja09ag.21670.4571.9.36.102_1
1/11/2017 9:42:54 PM | SETI@home | [error] No app version found for app setiathome_v8 platform windows_intelx86 ver 823 class opencl_ati5_SoG_nocal; discarding 08ja09ag.21670.4571.9.36.108_1
1/11/2017 9:42:54 PM | SETI@home | [error] No app version found for app setiathome_v8 platform windows_intelx86 ver 823 class opencl_ati5_SoG_nocal; discarding 08ja09ag.21670.4571.9.36.114_1
1/11/2017 9:42:54 PM | SETI@home | [error] No app version found for app setiathome_v8 platform windows_intelx86 ver 823 class opencl_ati5_SoG_nocal; discarding 08ja09ag.21670.4571.9.36.120_1
1/11/2017 9:43:22 PM | SETI@home | update requested by user
1/11/2017 9:43:25 PM | SETI@home | Sending scheduler request: Requested by user.
1/11/2017 9:43:25 PM | SETI@home | Requesting new tasks for AMD/ATI GPU
17) Message boards : Number crunching : SETI@home v8.22/23 Windows GPU applications support thread (Message 1841393)
Posted 10 days ago by Profile Raistmer
Post:
It's up to Eric now. No response from him still.
18) Message boards : Number crunching : Mac OS Sierra (Message 1841304)
Posted 10 days ago by Profile Raistmer
Post:
Seems nothing good to expect: http://setiathome.berkeley.edu/workunit.php?wuid=2389393563
19) Message boards : News : New SETI@home donation project on Bitcoin Utopia (Message 1840854)
Posted 12 days ago by Profile Raistmer
Post:
That's a pity. I run a cluster with more than 3,800 cores, including a stack of GPU blades. That will be doubling in the next few weeks.. So far I have not found any interest from the SETI folks to have me contribute with the cluster. This is yet more of the same... :(

What do you mean? What interest you want to see/get?

There is constant need of additional computing power in project.
Why just not to connect you cluster to SETI project and do crunching? What prevents??
20) Message boards : Number crunching : Discrepancy between BOINC crunching options for laptop and smartphones (Message 1840697)
Posted 13 days ago by Profile Raistmer
Post:
This topic resulted in some useful (hopefully) conversation on BOINC site also:
http://boinc.berkeley.edu/dev/forum_thread.php?id=11401&postid=75134


Next 20


 
©2017 University of California
 
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.