[Rtk-users] CUDA displaced detector filter failure

[Simon Rit]

Hi Chao,
Yesterday, I investigated the dynamic analysis log and found a problem in
the filter that has been corrected (see commit here
<https://github.com/SimonRit/RTK/commit/fef0b132211377fbabaa812d1ce0c3ec3c0827e3>).
Could you check if it fixes your issue please?
Thanks,
Simon

On Fri, Oct 10, 2014 at 2:30 PM, Simon Rit <simon.rit at creatis.insa-lyon.fr>
wrote:

> Hi Chao,
> Thanks for the report, it's new so  we might have left special cases
> behind... sorry. Is this something that occurs when launching rtkfdk,
> rtkwangdisplaceddetectorweighting or one of your own applications?
> It would be very useful if you could manage to reproduce this with a set
> of command lines that would simulate data first and then apply the
> weighting, similar to what's here
> <http://wiki.openrtk.org/index.php/RTK/Scripts/FDK> for example.
> Regards,
> Simon
>
> On Fri, Oct 10, 2014 at 10:58 AM, Chao Wu <wuchao04 at gmail.com> wrote:
>
>> Hi all,
>>
>> I am trying the new cuda displaced detector filter but
>> got cudaErrorLaunchFailure. This error was returned by cudaFree(dev_geom)
>> at the end of the code, but since this error type is not registered with
>> cudaFree(), it would be from previous, asynchronous launches. By
>> using cudaDeviceSynchronize() I found the error coming after calling
>> kernel_displaced_weight<<<...>>>(...).
>>
>> One interesting thing is that when the error occurs depends on the number
>> of projections processed at once (controlled by the --subset=N option of
>> rtkfdk.exe for example). If I use subset=1 then it fails right with the
>> first projection; if subset=2 it fails at the 16th run when processing the
>> 31st and 32nd projections; if subset=4, at the 14th run; if subset=16
>> (default), at the 6th run; and so on.
>>
>> The input region of each projection is about 1944x200 (the 2nd dimension
>> differs a little bit of course) and the output region is then about
>> 3888x200. A GTX580 card is used.
>>
>> Any suggestions and comments on this? Thanks.
>>
>> Regards,
>> Chao
>>
>> _______________________________________________
>> Rtk-users mailing list
>> Rtk-users at public.kitware.com
>> http://public.kitware.com/mailman/listinfo/rtk-users
>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://public.kitware.com/pipermail/rtk-users/attachments/20141014/7285a05f/attachment-0009.html>