[Clas_offline] splitbos utility

[Gavalian]

Dear All,

As it appears the problem was not the "splitbos" as I thought, it was 
the problem with me converting MCTK bank into PART(0) bank with the same 
Particle ID's, which in PART the geant id's should be used (instead of 
LUND). Only this caused a segmentation fault.
Apparently changing the PID to GEANT fixed my problem.

Is this a normal behavior ? Shouldn't be a warning or something in gsim 
for this ?

Gagik.

On 6/4/10 12:18 PM, Maurik Holtrop wrote:
> Hello Gagik,
>
> splitbos has always worked for me just fine with GSIM. It was created to split input files to GSIM.
>
> Are you sure that GSIM can actually access the input file you are giving it? BOS is not very smart, and will crash on things like not having read permission.
>
> Best,
> 	Maurik
>
> On Jun 4, 2010, at 10:55 AM, Gavalian wrote:
>
>    
>> Hi,
>>
>> I have a little problem with the splitbos utility. I have generated
>> sample of events to be analyzed with GSIM.
>> My original file contains 1M events, when I split it with "splitbos"
>> into 3000 event chunks, the GSIM gives me
>> "Segmentation fault" when opening the file:
>>
>> UGINIT     I: *** Geant initialization done ***
>> KINE_MCIN  I: getting MCIN input filename from command line flag: -mc
>> KINE_MCIN  I: opening BOS MCIN file on unit: 21, filename:
>> bosfiles/sig_sim_0001.bos, daname: MCIN
>> file[ 0]->bosfiles/sig_sim_0001.bos<-
>> Segmentation fault
>>
>> When trying to run through my 1M (original) file, gsim works just fine.
>> The thing is that neither fdump or boscount crush when operating on
>> those small files created by splitbos, only gsim has problem with them.
>>
>> The question I have is really this, is there a program that can tell me
>> what is wrong with a given BOS file, or is there an alternative to
>> splitbos that I should be using ?
>>
>> Did anyone have similar problem ?
>>
>> Gagik.
>> _______________________________________________
>> Clas_offline mailing list
>> Clas_offline at jlab.org
>> https://mailman.jlab.org/mailman/listinfo/clas_offline
>>      
>