Closed fischer1 closed 2 years ago
setup 1.87389e+02s (1.83) loadKernels 8.03903e+01s (0.79) udfExecuteStep 3.40134e-03s (0.00) solve step min/max 1.55650e-01s / 2.37318e-01s (first 10 steps excluded) total solve 1.02377e+02s makef 2.19445e+01s (0.21) udfUEqnSource 2.04153e-01s (0.00) velocitySolve 1.08192e+01s (0.11) pressureSolve 5.10852e+01s (0.50) preconditioner 3.75489e+01s (0.37) pMG smoother 2.20148e+01s (0.22) coarse grid 1.23455e+01s (0.12) projection 2.47380e+00s (0.02) scalarSolve 3.82929e+00s (0.04) makeq 9.74556e+00s (0.10) udfProperties 4.68288e-03s (0.00)
Could we have "number of calls" at the front of each line for these stats ?
That way, one would know, e.g., how many pressure iterations used for the entire run, because preconditioner would have a prefix like:
preconditioner 12137 3.75489e+01s (0.37)
etc.
Fixed in https://github.com/Nek5000/nekRS/commit/9ae887200454a3409e8a0c324bde9989815716c3
stgeke
commented
2 years ago stgeke
commented
2 years ago
setup 1.87389e+02s (1.83) loadKernels 8.03903e+01s (0.79) udfExecuteStep 3.40134e-03s (0.00) solve step min/max 1.55650e-01s / 2.37318e-01s (first 10 steps excluded) total solve 1.02377e+02s makef 2.19445e+01s (0.21) udfUEqnSource 2.04153e-01s (0.00) velocitySolve 1.08192e+01s (0.11) pressureSolve 5.10852e+01s (0.50) preconditioner 3.75489e+01s (0.37) pMG smoother 2.20148e+01s (0.22) coarse grid 1.23455e+01s (0.12) projection 2.47380e+00s (0.02) scalarSolve 3.82929e+00s (0.04) makeq 9.74556e+00s (0.10) udfProperties 4.68288e-03s (0.00)
Could we have "number of calls" at the front of each line for these stats ?
That way, one would know, e.g., how many pressure iterations used for the entire run, because preconditioner would have a prefix like:
preconditioner 12137 3.75489e+01s (0.37)
etc.