check host_readgauge.cpp (Defect #138)


Added by Lars Zeidlewicz about 8 years ago. Updated over 6 years ago.


Status:Feedback Start date:25 May 2011
Priority:Normal Due date:
Assignee:Matthias Bach % Done:

90%

Category:-
Target version:- Estimated time:0.25 hour

Description

I removed unused parameters (field_out) from read_data_single() and read_data_double(). Was that ok, or is there something wrong?


Related issues

related to CL2QCD - Defect #245: Cleaning up the host-code New 04 Dec 2011

History

Updated by Christopher Pinke about 8 years ago

the thing is that binary data is read in regardless if it belongs to a gaugefield or a propagator:

if( (strcmp (lime_types[5],lime_type)==0 || strcmp (lime_types[8],lime_type)==0 ) && cter <1) {...

in field_out there is the type of field that the prog has found before and one could build in a checking funtion if one wants to.

In praxis, I think that as long as one does not deal with fermions to be read in, one cannot run into problems. But I did not check if this could happen. What could happen is that one tries to read in a gaugefield, but the memory allocated is for a propagator...

I suppose one can take the field_out out of the program... What do you think?

  • % Done changed from 0 to 90

Updated by Christopher Pinke almost 8 years ago

Another thing, not really related to the question above: Perhaps we should push the implementation of spinorfield-savings forward, so that we have full functionality regarding the ILDG-format...

  • Assignee changed from Christopher Pinke to Lars Zeidlewicz

Updated by Christopher Pinke almost 7 years ago

Didn`t you fix some bugs here? I think we can then close this issue...

  • Assignee changed from Lars Zeidlewicz to Matthias Bach

Updated by Christopher Pinke over 6 years ago

  • Status changed from New to Feedback

Also available in: Atom PDF