MFC
Exascale flow solver
Loading...
Searching...
No Matches
m_phase_timing.fpp.f90 File Reference

Contains module m_phase_timing. More...

Go to the source code of this file.

Modules

module  m_phase_timing
 Per-phase wall-time budget of the AMR step, reported when rank_time_wrt is on. Brackets are disjoint (regrid sub-phases nest inside regrid) and drain the device first, so a phase's time includes the GPU work it launched.

Functions/Subroutines

impure subroutine, public m_phase_timing::s_phase_tic (id)
impure subroutine, public m_phase_timing::s_phase_toc (id)
impure subroutine, public m_phase_timing::s_phase_step_tic ()
impure subroutine, public m_phase_timing::s_phase_step_toc ()
impure subroutine, public m_phase_timing::s_phase_report ()
 Mean and max over ranks per phase, share of the step-loop wall, imbalance (max/mean) and calls; the residual is the wall minus the sum of the means.
impure subroutine, public m_phase_timing::s_mem_note (id, bytes)
 Add bytes to group id of the end-of-run AMR memory table.
impure subroutine, public m_phase_timing::s_mem_note_fields (id, f)
 s_mem_note for the fields of f that are allocated.

Variables

integer, parameter, public m_phase_timing::ph_halo = 1
 coarse cons halo exchange (once per stage)
integer, parameter, public m_phase_timing::ph_gather = 2
 fill waves: coarse-patch gather into every block
integer, parameter, public m_phase_timing::ph_gfill = 3
 ghost prolongation from the gathered patch
integer, parameter, public m_phase_timing::ph_seam = 4
 fine-fine seam halo
integer, parameter, public m_phase_timing::ph_rhs = 5
 fine-block RHS
integer, parameter, public m_phase_timing::ph_rk = 6
 fine-block RK update + IB
integer, parameter, public m_phase_timing::ph_reflux = 7
 level-1 reflux (faces wave + apply)
integer, parameter, public m_phase_timing::ph_regrid = 8
 regrid
integer, parameter, public m_phase_timing::ph_coarse = 9
 coarse (base-grid) solver work
integer, parameter, public m_phase_timing::ph_swap = 10
 per-batch grid-state swap and restore
integer, parameter, public m_phase_timing::ph_restr = 11
 post-stage restrict + reflux-to-parent fold
integer, parameter, public m_phase_timing::ph_lerp = 12
 subcycled coarse time interpolation for the level-1 fills
integer, parameter m_phase_timing::ph_ntop = 12
 the disjoint brackets above; the regrid sub-phases below nest in PH_REGRID
integer, parameter, public m_phase_timing::ph_rg_halo = 13
 coarse halo + host copy for tagging
integer, parameter, public m_phase_timing::ph_rg_tag = 14
 tagging
integer, parameter, public m_phase_timing::ph_rg_clus = 15
 clustering, shaping, nesting and box checks
integer, parameter, public m_phase_timing::ph_rg_own = 16
 stash and owner assignment
integer, parameter, public m_phase_timing::ph_rg_mig = 17
 fine-state migration
integer, parameter, public m_phase_timing::ph_rg_build = 18
 new-block build (fills, prolongation, overlap copy)
integer, parameter, public m_phase_timing::ph_rg_rec = 19
 slot reconcile, IB setup, seam check
integer, parameter m_phase_timing::ph_n = 19
character(len=8), dimension(ph_n), parameter m_phase_timing::ph_name = [character(len=8)::'halo','gather', 'gfill', 'seam', 'rhs', 'rk', 'reflux', 'regrid', 'coarse', 'swap', 'restr', 'lerp', ' rg.halo', ' rg.tag', ' rg.clus', ' rg.own', ' rg.mig', ' rg.build', ' rg.rec']
real(wp), dimension(ph_n) m_phase_timing::acc = 0._wp
real(wp) m_phase_timing::wall = 0._wp
 wall is the accumulated step-loop time the phases are budgeted against
integer(8), dimension(ph_n) m_phase_timing::ncall = 0
integer(8), dimension(ph_n) m_phase_timing::tic_c = 0
integer(8) m_phase_timing::wall_c = 0
integer, dimension(ph_n) m_phase_timing::depth = 0
 a re-entered bracket (shared routine) measures only its outermost pair
integer, parameter, public m_phase_timing::mem_store = 1
 AMR allocations by group, noted by each owner before it frees them (s_mem_note) and printed with the phase table.
integer, parameter, public m_phase_timing::mem_scratch = 2
 batched-advance scratch (prim, rhs) and the block bridge
integer, parameter, public m_phase_timing::mem_patch = 3
 coarse patch pool and its subcycling time cache
integer, parameter, public m_phase_timing::mem_reg = 4
 flux registers
integer, parameter, public m_phase_timing::mem_ib = 5
 fine IB ghost-point park
integer, parameter, public m_phase_timing::mem_ibhost = 6
 per-slot fine IB marker copies (host only)
integer, parameter m_phase_timing::mem_n = 6
character(len=8), dimension(mem_n+1), parameter m_phase_timing::mem_name = [character(len=8)::'store','scratch', 'patch', 'register', 'ib park', 'ib marks', 'total']
real(wp), dimension(mem_n) m_phase_timing::mem_mib = 0._wp

Detailed Description

Contains module m_phase_timing.

Definition in file m_phase_timing.fpp.f90.