Skip to content
Snippets Groups Projects

Repository graph

You can move around the graph by using the arrow keys.
Select Git revision
  • 3-error-in-run-the-program
  • containers
  • containers-m10
  • containers-m8
  • enable_svd
  • experiment
  • fix_layered_spheres_bug
  • magma_refinement
  • main
  • master default protected
  • original protected
  • parallel_angles
  • parallel_angles_gmu
  • profile_omp_leonardo
  • release9
  • script_devel
  • shaditest
  • test1
  • test_nvidia_profiler
  • NP_TMcode-M10a.02
  • NP_TMcode-M10a.01
  • NP_TMcode-M10a.00
  • NP_TMcode-M9.01
  • NP_TMcode-M9.00
  • NP_TMcode-M8.03
  • NP_TMcode-M8.02
  • NP_TMcode-M8.01
  • NP_TMcode-M8.00
  • NP_TMcode-M7.00
  • v0.0
30 results
Created with Raphaël 2.2.026Mar25232221201918151312118765429Feb282723222120161413129765229Jan262524191817161514121110986542131Dec2928272625242120181615141312111076330Nov292827241310863231Oct29262019171613125429Sep26252422211918765415FebSave last_message as pointer to stringDefine logger as pointerDefine logger as pointerCorrect definitely lost bytes tracked by valgrind --leak-check=allCorrect valgrind definitely lost leaksUpdate logger branch with OpenMP separated init threadMerge branch 'test_omp_cluster' into 'master'- do not allocate local variables for thread 0, only for additional onesRemove redundant C1_AddOns vectors and set up logging for parallel executionSolve merge conflicts for OMP implementationfix mkl for Rosalba's computeradd a missing deleteMerge branch 'test_omp_cluster' into 'master'free zpv before c4, since it needs c4->lm: now valgrind is clean.Merge branch 'test_omp_cluster' into 'master'correct valgrind omp errorscorrect first 2 valgrind errors (one to double check by Giovanni)bump up one more allocation from size configurations to nspheres- For safety, since some loops are murky, overallocate some arrays of size nspheres instead of configurationsUse separate loops to re-assemble ASCII and binary output filesSkip log file if the file line numbers differpass all scalars in cluster scales loop by reference something otherwise breaks)make the splitting of outputs conditional on _OPENMPMerge branch 'test_omp_cluster' of https://www.ict.inaf.it/gitlab/giacomo.mulas/np_tmcode into test_omp_clusterAppend thread output files to program outputUse adaptive scaling to distinguish between warnings and errorsadd support for USE_OPENMP in make.incAdd first OpenMP parallelisation in cluster.cpp. YAY!!work around stupid g++11 implementation to allocate am_2[0]finish replacing fresh local variables instead of global ones in scales looponly one round of substitutions of fresh variables in scales loop left...almost finished switch to fresh variables in the scales loop of clusteradd yet more fresh variables in the scales loopadd more cleanly defined variables to the scales loopadd fresh zpv_2- add some more freshly defined variables in the scales loop- begin calling the scales iterations with "fresh" variables, to test parallelismBe pedantic and, for the time being, pass _everything_ to cluster_jxi488_cycle()Move the content of the loop over scales in a separate function, so it can be called with fresh arguments every time to check parallelism and dependencies.Remove C1_AddOns::allocate_vectors() and fix vector memory leak