scnc
/
quantum-espresso
mirror of https://gitlab.com/QEF/q-e.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

cpu time per mpi task scaled dow by the number of omp processes participating into it

This commit is contained in:
Stefano De Gironcoli 2018-06-11 13:40:37 +02:00 committed by Stefano de Gironcoli
parent 800f1da781
commit 87400ecfd3
1 changed files with 11 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
UtilXlib/clocks_handler.f90
View File

@ -47,6 +47,8 @@ MODULE mytime
CHARACTER(len=12) :: clock_label(maxclock)
INTEGER :: called(maxclock)
!
REAL(DP) :: mpi_per_thread = 1.0_DP
INTEGER :: nclock = 0
LOGICAL :: no
#if defined (__TRACE)
@ -80,7 +82,7 @@ SUBROUTINE init_clocks( go )
!
USE util_param, ONLY : DP, stdout
USE mytime, ONLY : called, t0cpu, cputime, no, notrunning, maxclock, &
clock_label, walltime, t0wall, nclock
clock_label, walltime, t0wall, nclock, mpi_per_thread
#if defined (__TRACE)
USE mytime, ONLY : mpime, max_print_depth, MPI_COMM_WORLD
#endif
@ -93,6 +95,10 @@ SUBROUTINE init_clocks( go )
#endif
INTEGER :: n, ierr
!
#if defined(_OPENMP)
INTEGER, EXTERNAL :: omp_get_max_threads
mpi_per_thread = 1.0_DP/omp_get_max_threads()
#endif
no = .not. go
nclock = 0
!
@ -319,7 +325,7 @@ SUBROUTINE print_this_clock( n )
!----------------------------------------------------------------------------
!
USE util_param, ONLY : DP, stdout
USE mytime, ONLY : clock_label, cputime, walltime, &
USE mytime, ONLY : clock_label, cputime, walltime, mpi_per_thread, &
notrunning, called, t0cpu, t0wall, f_wall, f_tcpu
!
IMPLICIT NONE
@ -346,6 +352,9 @@ SUBROUTINE print_this_clock( n )
!
ENDIF
!
! rescale the elapsed cpu time on a per-thread basis
elapsed_cpu_time = elapsed_cpu_time * mpi_per_thread
!
nmax = called(n)
!
! ... In the parallel case there are several possible approaches