All my CUDA tasks errored out. Win XP 32bit, 9600GT, driver 285.58, core 6.13.12.
6.13.12Registr je po�kozen. Je po�kozena struktura jednoho ze soubor�, kter� obsahuj� registr, nebo struktura syst�mov�ho bitov� kopie souboru v pam�ti, nebo soubor nemohl b�t obnoven, proto�e chyb� nebo je po�kozena n�hradn� kopie �i protokol . (0x3f7) - exit code 1015 (0x3f7)
Activated exception handling...
[23:52:52][4044][INFO ] Starting data processing...
[23:52:52][4044][INFO ] CUDA global memory status (initial GPU state, including context):
------> Used in total: 51 MB (461 MB free / 512 MB total) -> Used by this application (assuming a single GPU task): 0 MB
[23:52:52][4044][INFO ] Using CUDA device #0 "GeForce 9600 GT" (64 CUDA cores / 349.44 GFLOPS)
[23:52:52][4044][INFO ] Version of installed CUDA driver: 4010
[23:52:52][4044][INFO ] Version of CUDA driver API used: 3020
[23:52:53][4044][INFO ] Checkpoint file unavailable: status.cpt (No such file or directory).
------> Starting from scratch...
[23:52:53][4044][INFO ] Header contents:
------> Original WAPP file: ./p2030.20100913.G48.73+01.03.S.b5s0g0.00000_DM545.60
------> Sample time in microseconds: 65.4762
------> Observation time in seconds: 274.62705
------> Time stamp (MJD): 55453.031578180256
------> Number of samples/record: 0
------> Center freq in MHz: 1214.289551
------> Channel band in MHz: 0.33605957
------> Number of channels/record: 960
------> Nifs: 1
------> RA (J2000): 191721.825399
------> DEC (J2000): 142523.136101
------> Galactic l: 0
------> Galactic b: 0
------> Name: G48.73+01.03.S
------> Lagformat: 0
------> Sum: 1
------> Level: 3
------> AZ at start: 0
------> ZA at start: 0
------> AST at start: 0
------> LST at start: 0
------> Project ID: --
------> Observers: --
------> File size (bytes): 0
------> Data size (bytes): 0
------> Number of samples: 4194304
------> Trial dispersion measure: 545.6 cm^-3 pc
------> Scale factor: 0.118517
[23:52:55][4044][INFO ] Seed for random number generator is 1149629056.
[23:52:56][4044][INFO ] Derived global search parameters:
------> f_A probability = 0.08
------> single bin prob(P_noise > P_thr) = 9.93986e-009
------> thr1 = 18.4267
------> thr2 = 21.5421
------> thr4 = 26.5915
------> thr8 = 35.0049
------> thr16 = 49.3672
[23:52:56][4044][INFO ] CUDA global memory status (GPU setup complete):
------> Used in total: 312 MB (200 MB free / 512 MB total) -> Used by this application (assuming a single GPU task): 261 MB
[23:52:56][4044][ERROR] Error launching CUDA TSP kernel (error: 1)
[23:52:56][4044][ERROR] Demodulation failed (error: 1015)!
23:52:56 (4044): called boinc_finish
Copyright © 2024 Einstein@Home. All rights reserved.
1015 (0x3f7)
)
This translates in Google translate as
Translation: The registry is corrupted. The structure of one of the files containing registry data is corrupted, or the system's memory image of the file is corrupted, or the file could not be recovered because the alternate copy or log was absent or corrupted.
Sounds like corrupted drivers to me. But then, since Albert is testing ATI GPUs running OpenCL 1.0 for now, what are you doing here trying to run an nVidia GPU on CUDA? (Pointing out the applications page, where there's no CUDA application to be seen).
You've got your computers hidden, so I can't see if you're trying to run with an anonymous platform here. Are you? Why don't you use one of those 4850s or 4870s you use at Milkyway to test the ATIOpenCL app here?
Thanks for reporting! I
)
Thanks for reporting!
I disabled the CUDA apps again. Apparently they choke on the modified BRP workunits we are sending out here.
BM
BM
Hi Ageless. I am not a newbie
)
Hi Ageless.
I am not a newbie on testing (see DD forum, where we both have reported a lot), I have reported the problem after
- clear install of GPU driver,
- updating boinc to 6.13.12.
That task was crunched by new CUDA app. Radio Pulsar Search v1.07 (BRP3cuda32) sent by server.
I am not reporting errors when I mean that my machine is not OK for testing, e.g. host with ATI HD 4770 + core 6.12.34 (PG PPSE AtiOpenCL works there fine).
App_info I use with GTX260, but only for main project crunching, it makes me no sense to use it when testing new apps.
RE: Thanks for
)
This issue got solved with v1.19 of the CUDA apps...
Oliver