Peter Willendrup

Results 236 comments of Peter Willendrup

Consider to invest further work to allow "build at install-time" which will give the user better possibilities of adding to / repairing the env

The behaviour is found within all example instruments that contain precession-regions using the field-stack mechanism are affected, from McStas 2.7.1: ``` SEMSANS_Delft.instr SEMSANS_instrument.instr SESANS_Delft.instr SE_example.instr SE_example2.instr Test_Magnetic_Constant.instr Test_Magnetic_Majorana.instr Test_Magnetic_Rotation.instr Test_Magnetic_Userdefined.instr...

https://github.com/McStasMcXtrace/McCode/commit/8485e3d8c8ac1d228c9b76c95aec83f8d48bb388 seems to be a good fix.

(McStas 3.x seems to already include this explicit free...)

Relevant code snippet is in fact within pqtgfrontend.py (mccodelib, not mcplot/pytqtgraph)

https://gcc.gnu.org/wiki/Offloading#Compilation_process to educate ourselves, try with simple non-mcstas code first

Something is also off wrt. the positions and velocities, below output generated using Event_monitor_simple in McStas ``` mcstas-3.x-dev env WOF_SANS_GPUsetup $ head -10 Test_MCPL_input_cpu_20221015_215650/Events.log Test_MCPL_input_gpu_20221015_215650/Events.log ==> Test_MCPL_input_cpu_20221015_215650/Events.log Test_MCPL_input_gpu_20221015_215650/Events.log

Hi @farhi, I believe this is in fact already covered (and fixed) via the patches @g5t supplied in https://github.com/McStasMcXtrace/McCode/issues/1270 ? (I believe I fixed the code both on mccode-3 and...

https://github.com/McStasMcXtrace/McCode/commit/279744c0a00d9cafae9f84347a770faafae874d2