I/O Improvements for run time info and EL Bubbles by wilfonba · Pull Request #1015 · MFlowCode/MFC

wilfonba · 2025-10-09T17:32:45Z

User description

Description

This PR makes several improvements to I/O for Euler-Lagrange bubbles, run_time.inf,
and the progress output to the terminal in post process.

Euler-Lagrange bubble output
- Adds optional header to bubbles.dat
- Adds logicals to only write relevant bubble data
- Adds bubble data to silo files for visualization in Paraview
- Improve parallel output of bubble data to avoid reduction to root process
- Added missing host updates for I/O of Lag bubbles. The host wasn't being updated
  for most of the bubble variables and the void fraction wasn't being reinitialized
  correctly so results were different when restarting from a save.
Updates to run_time.inf
- Improves formatting for readability
- Prints a number for Rc Min instead of asterisks
Updates to progress output in post process
- Prints time/step and average time to terminal like simulation

Type of change

Something else

Scope

This PR comprises a set of related changes with a common goal

If you cannot check the above box, please split your PR into multiple PRs that each have a common goal.

How Has This Been Tested?

Euler-Lagrange bubble output

2D

The following video was created after adding lag_db_wrt to the examples/2D_lagrange_bubblescreen/
case file and running it with 4 MPI ranks on my laptop. The video shows the lagrangian
bubble data as spherical glyphs in Paraview that are scaled by the bubble radii.
The bubbles are shown 10x their actual size for visibility.

test.mp4

3D

The following video was created after adding lag_db_wrt to the examples/3D_lagrange_bubblescreen/
(I also increase the amplitude of the pulse to increase the radial dynamics)
case file and running it with 4 A100 GPUs on my Wingtip. The video shows the lagrangian
bubble data as spherical glyphs in Paraview that are scaled by the bubble radii.
The bubbles are shown 10x their actual size for visibility. I also ran this case with
4 MI250x GCDs on Frontier and got the same results. I also tried restarting the simulation
from an intermediate save and got the same results as running the simulation from
start to finish without a restart, both with parallel and serial I/O.

test.mp4

`run_time.inf` Verification

I ran viscous and inviscid cases and inspecte the output in run_time.inf. Comparisons
to the existing versions are in the drop down menus below.

Viscous cases

Old

Time-steps        Time         ICFL Max      VCFL Max        Rc Min
         0  0.011062              0.000000       0.112838       0.000001      **********
         1  0.011062              0.011062       0.112838       0.000001      **********
         2  0.011062              0.022125       0.112838       0.000001      **********
         3  0.011062              0.033187       0.112838       0.000001      **********
         4  0.011062              0.044250       0.112838       0.000001      **********

New

 Time-step                     dt                   Time               ICFL Max               VCFL Max                       Rc Min
         0               0.011062               0.011062               0.112838               0.000001                 1.446084E+05
         1               0.011062               0.022125               0.112838               0.000001                 1.446084E+05
         2               0.011062               0.033187               0.112838               0.000001                 1.446084E+05
         3               0.011062               0.044250               0.112838               0.000001                 1.446084E+05
         4               0.011062               0.055313               0.112838               0.000001                 1.446084E+05

Inviscid cases

Old

            Time-steps                Time                ICFL Max
                    0                0.011062                0.000000              0.112838
                    1                0.011062                0.011062              0.112838
                    2                0.011062                0.022125              0.112838
                    3                0.011062                0.033187              0.112838
                    4                0.011062                0.044250              0.112838

New

             Time-step                     dt                   Time               ICFL Max
                     0               0.011062               0.011062               0.112838
                     1               0.011062               0.022125               0.112838
                     2               0.011062               0.033187               0.112838
                     3               0.011062               0.044250               0.112838
                     4               0.011062               0.055313               0.112838

Post Process Verification CL Output

I ran a post process case and inspected the terminal output. It's correct

Checklist

I have added comments for the new code
I added Doxygen docstrings to the new code
I have made corresponding changes to the documentation (docs/)
I have added example cases in examples/ that demonstrate my new feature performing as expected.
They run to completion and demonstrate "interesting physics"
I ran ./mfc.sh format before committing my code
New and existing tests pass locally with my changes, including with GPU capability enabled (both NVIDIA hardware with NVHPC compilers and AMD hardware with CRAY compilers) and disabled
This PR does not introduce any repeated code (it follows the DRY principle)
I cannot think of a way to condense this code and reduce any introduced additional line count

If your code changes any code source files (anything in `src/simulation`)

To make sure the code is performing as expected on GPU devices, I have:

Checked that the code compiles using NVHPC compilers
Checked that the code compiles using CRAY compilers
Ran the code on either V100, A100, or H100 GPUs and ensured the new feature performed as expected (the GPU results match the CPU results)
Ran the code on MI200+ GPUs and ensure the new features performed as expected (the GPU results match the CPU results)

PR Type

Enhancement

Description

Enhanced Euler-Lagrange bubble I/O with Silo database support
Improved run_time.inf formatting and terminal progress output
Added selective bubble data output controls
Fixed host memory updates for bubble restart functionality

Diagram Walkthrough

flowchart LR
  A["Bubble Data"] --> B["Text Output (.dat)"]
  A --> C["Silo Database Output"]
  D["Runtime Info"] --> E["Improved Formatting"]
  F["Post Process"] --> G["Enhanced Terminal Output"]
  H["Restart Files"] --> I["Fixed Host Updates"]

File Walkthrough

Relevant files

Configuration changes

6 files

m_constants.fpp `Added lag_io_vars constant for MPI I/O`	+1/-0
m_global_parameters.fpp `Added Lagrangian bubble output control parameters`	+38/-0
m_start_up.f90 `Added bubble output parameters to input reading`	+11/-6
case.py `Enabled Silo database output for bubble data`	+1/-0
case.py `Enabled Silo database output for bubble data`	+1/-0
case_dicts.py `Added bubble output parameter definitions`	+18/-0

Enhancement

6 files

m_data_output.fpp `Enhanced bubble output with Silo database support`	+483/-55
m_mpi_proxy.fpp `Added MPI broadcast for bubble output parameters`	+10/-1
p_main.fpp `Added wall time measurement and progress reporting`	+13/-0
m_bubbles.fpp `Added default case for bubble dynamics model`	+4/-1
m_bubbles_EL.fpp `Major improvements to bubble I/O and restart functionality`	+248/-170
m_data_output.fpp `Improved run_time.inf formatting and output structure`	+14/-21

Bug fix

2 files

m_assign_variables.fpp `Fixed surface tension variable assignment bug`	+1/-1
m_start_up.fpp `Fixed host memory updates for bubble variables`	+4/-2

Formatting

1 files

m_body_forces.fpp `Simplified acceleration computation with template loops`	+4/-8

Documentation

1 files

case.md `Updated documentation for bubble output parameters`	+21/-3

… into ImprovedELBubblesIO

Copilot

Pull Request Overview

This PR implements comprehensive I/O improvements for Euler-Lagrange bubbles, enhances the run_time.inf output formatting, and adds progress timing information to the post-process terminal output.

Adds configurable Lagrangian bubble output with new logical parameters for selective data writing
Improves run_time.inf formatting with better column alignment and proper numeric output for Rc Min
Adds wall time measurements and progress indicators to post-process output similar to simulation

Reviewed Changes

Copilot reviewed 16 out of 16 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
toolchain/mfc/run/case_dicts.py	Adds new Lagrangian bubble output control parameters
src/simulation/m_start_up.fpp	Fixes GPU host updates for bubble variables and removes redundant update
src/simulation/m_data_output.fpp	Improves run_time.inf formatting with better column alignment
src/simulation/m_bubbles_EL.fpp	Major improvements to bubble I/O, restart functionality, and boundary condition handling
src/simulation/m_bubbles.fpp	Adds default case for bubble dynamics and fixes comment
src/simulation/m_body_forces.fpp	Refactors body force computation using Fypp templates
src/pre_process/m_assign_variables.fpp	Fixes surface tension variable assignment
src/post_process/p_main.fpp	Adds wall time measurements for progress tracking
src/post_process/m_start_up.f90	Updates progress output format with timing information
src/post_process/m_mpi_proxy.fpp	Adds MPI broadcasting for new Lagrangian bubble parameters
src/post_process/m_global_parameters.fpp	Defines new Lagrangian bubble output control variables
src/post_process/m_data_output.fpp	Splits bubble output into separate text and database functions with selective variable writing
src/common/m_constants.fpp	Adds constant for Lagrangian I/O variable count
examples/*/case.py	Enables Lagrangian database output in example cases
docs/documentation/case.md	Documents new Lagrangian bubble output parameters

src/post_process/m_data_output.fpp

src/simulation/m_bubbles_EL.fpp

qodo-code-review · 2025-10-09T17:33:45Z

PR Reviewer Guide 🔍

Here are some key observations to aid the review process:

⏱️ Estimated effort to review: 4 🔵🔵🔵🔵⚪
🧪 No relevant tests
🔒 No security concerns identified
⚡ Recommended focus areas for review Possible Issue In text writer, the velocity output uses inputvals(7), inputvals(8), inputvals(8) for vx, vy, vz, duplicating vy and omitting vz. This likely should be 7, 8, 9. Verify all column indices match the MPI I/O layout consistently. if (lag_pos_wrt) write (29, '(3(E15.7, A))', advance='no') inputvals(1), ', ', inputvals(2), ', ', inputvals(3), ', ' if (lag_pos_prev_wrt) write (29, '(3(E15.7, A))', advance='no') inputvals(4), ', ', inputvals(5), ', ', inputvals(6), ', ' if (lag_vel_wrt) write (29, '(3(E15.7, A))', advance='no') inputvals(7), ', ', inputvals(8), ', ', inputvals(8), ', ' if (lag_rad_wrt) write (29, '(E15.7, A)', advance='no') inputvals(10), ', ' Boundary Check Change The previous clamping of ghost-cell indices to the nearest in-domain cell is now commented out and replaced by aborts for multiple wall-like BCs. Confirm this stricter behavior is intended for all cases and won’t break valid restart scenarios near boundaries. !cell(1) = min(max(cell(1), 0), m) !cell(2) = min(max(cell(2), 0), n) !if (p > 0) cell(3) = min(max(cell(3), 0), p) call s_convert_to_mixture_variables(q_cons_vf, cell(1), cell(2), cell(3), & MPI I/O Header Consistency New restart file header now includes tot_part, time, dt, num_procs, and per-rank counts. Ensure all readers/writers (simulation and post_process) handle offsets, types, and sizeof consistently across platforms (INTEGER sizes) and that non-participating ranks use a zero-length view without deadlocks. ! Construct file path write (file_loc, '(A,I0,A)') 'lag_bubbles_', t_step, '.dat' file_loc = trim(case_dir)//'/restart_data'//trim(mpiiofs)//trim(file_loc) ! Check if file exists inquire (FILE=trim(file_loc), EXIST=file_exist) if (.not. file_exist) then call s_mpi_abort('Restart file '//trim(file_loc)//' does not exist!') end if if (.not. parallel_io) return if (proc_rank == 0) then call MPI_FILE_OPEN(MPI_COMM_SELF, file_loc, MPI_MODE_RDONLY, & mpi_info_int, ifile, ierr) call MPI_FILE_READ(ifile, file_tot_part, 1, MPI_INTEGER, status, ierr) call MPI_FILE_READ(ifile, file_time, 1, mpi_p, status, ierr) call MPI_FILE_READ(ifile, file_dt, 1, mpi_p, status, ierr) call MPI_FILE_READ(ifile, file_num_procs, 1, MPI_INTEGER, status, ierr) call MPI_FILE_CLOSE(ifile, ierr) end if call MPI_BCAST(file_tot_part, 1, MPI_INTEGER, 0, MPI_COMM_WORLD, ierr) call MPI_BCAST(file_time, 1, mpi_p, 0, MPI_COMM_WORLD, ierr) call MPI_BCAST(file_dt, 1, mpi_p, 0, MPI_COMM_WORLD, ierr) call MPI_BCAST(file_num_procs, 1, MPI_INTEGER, 0, MPI_COMM_WORLD, ierr) allocate (proc_bubble_counts(file_num_procs)) if (proc_rank == 0) then call MPI_FILE_OPEN(MPI_COMM_SELF, file_loc, MPI_MODE_RDONLY, & mpi_info_int, ifile, ierr) ! Skip to processor counts position disp = int(sizeof(file_tot_part) + 2sizeof(file_time) + sizeof(file_num_procs), & MPI_OFFSET_KIND) call MPI_FILE_SEEK(ifile, disp, MPI_SEEK_SET, ierr) call MPI_FILE_READ(ifile, proc_bubble_counts, file_num_procs, MPI_INTEGER, status, ierr) call MPI_FILE_CLOSE(ifile, ierr) end if call MPI_BCAST(proc_bubble_counts, file_num_procs, MPI_INTEGER, 0, MPI_COMM_WORLD, ierr) gsizes(1) = file_tot_part gsizes(2) = lag_io_vars lsizes(1) = file_tot_part lsizes(2) = lag_io_vars start_idx_part(1) = 0 start_idx_part(2) = 0 call MPI_TYPE_CREATE_SUBARRAY(2, gsizes, lsizes, start_idx_part, & MPI_ORDER_FORTRAN, mpi_p, view, ierr) call MPI_TYPE_COMMIT(view, ierr) call MPI_FILE_OPEN(MPI_COMM_WORLD, file_loc, MPI_MODE_RDONLY, & mpi_info_int, ifile, ierr) disp = int(sizeof(file_tot_part) + 2sizeof(file_time) + sizeof(file_num_procs) + & file_num_procs*sizeof(proc_bubble_counts(1)), MPI_OFFSET_KIND) call MPI_FILE_SET_VIEW(ifile, disp, mpi_p, view, & 'native', mpi_info_null, ierr) allocate (MPI_IO_DATA_lg_bubbles(file_tot_part, 1:lag_io_vars))

src/post_process/m_data_output.fpp

wilfonba · 2025-10-09T17:48:15Z

I'm not sure why, but the MacOS runners are failing due to some Python version incompatibility with Cantera

codecov · 2025-10-09T20:17:05Z

Codecov Report

❌ Patch coverage is 42.42424% with 285 lines in your changes missing coverage. Please review.
✅ Project coverage is 42.32%. Comparing base (1609707) to head (6a9c1ab).
⚠️ Report is 1 commits behind head on master.

Files with missing lines	Patch %	Lines
src/post_process/m_data_output.fpp	45.19%	87 Missing and 67 partials ⚠️
src/simulation/m_bubbles_EL.fpp	27.67%	109 Missing and 6 partials ⚠️
src/post_process/m_mpi_proxy.fpp	22.22%	6 Missing and 1 partial ⚠️
src/common/m_mpi_common.fpp	0.00%	2 Missing ⚠️
src/post_process/m_start_up.f90	50.00%	0 Missing and 2 partials ⚠️
src/simulation/m_data_output.fpp	71.42%	1 Missing and 1 partial ⚠️
src/post_process/m_global_parameters.fpp	95.83%	0 Missing and 1 partial ⚠️
src/pre_process/m_assign_variables.fpp	0.00%	1 Missing ⚠️
src/simulation/m_body_forces.fpp	50.00%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1015      +/-   ##
==========================================
+ Coverage   42.11%   42.32%   +0.20%     
==========================================
  Files          70       70              
  Lines       19936    20187     +251     
  Branches     2482     2549      +67     
==========================================
+ Hits         8396     8544     +148     
- Misses       9996    10043      +47     
- Partials     1544     1600      +56

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

sbryngelson · 2025-10-15T19:25:18Z

This seems to be working. Is anything blocking a merge at the moment... do we need any tests?

wilfonba · 2025-10-15T20:18:39Z

I could likely improve the test coverage without adding any/many more tests. I guess I should probably do that before merging, though I hadn't previously intended too

sbryngelson · 2025-10-15T20:27:14Z

if it's easy please do, if not let me know

wilfonba · 2025-10-18T15:23:23Z

@sbryngelson this passed everything and should be able to merge. Coverage isn't perfect, but it is much better

wilfonba and others added 6 commits October 9, 2025 09:54

Add EL Silo I/O, improve run_time.inf, and post process terminal output

c23a919

format, spelling, and lint

e549426

typos from github runner

bf2138b

formatting and bug fixes

1debcdc

Merge branch 'ImprovedELBubblesIO' of github.com:wilfonba/MFC-Wilfong…

ff66bca

… into ImprovedELBubblesIO

Merge remote-tracking branch 'upstream/master' into ImprovedELBubblesIO

f694544

wilfonba requested review from a team and Copilot October 9, 2025 17:32

wilfonba requested a review from sbryngelson as a code owner October 9, 2025 17:32

Copilot AI reviewed Oct 9, 2025

View reviewed changes

src/post_process/m_data_output.fpp Outdated Show resolved Hide resolved

src/simulation/m_bubbles_EL.fpp Outdated Show resolved Hide resolved

src/simulation/m_bubbles_EL.fpp Show resolved Hide resolved

qodo-code-review bot added the Review effort 4/5 label Oct 9, 2025

qodo-code-review bot reviewed Oct 9, 2025

View reviewed changes

src/post_process/m_data_output.fpp Show resolved Hide resolved

qodo-code-review bot reviewed Oct 9, 2025

View reviewed changes

src/post_process/m_data_output.fpp Outdated Show resolved Hide resolved

autoreview suggestions

eea226f

wilfonba added 3 commits October 10, 2025 08:17

support 1D output in paraview with 1D meshes

1354011

CI fix for MacOS runners

8311260

abbreviate some if conditions

721cd77

Merge branch 'master' into ImprovedELBubblesIO

e77885b

wilfonba and others added 4 commits October 16, 2025 13:26

update test cases and golden files

a52b157

add string formats for reproducibility and regenerate golden files

21cb9ed

fix GPU testing

3809961

Merge branch 'master' into ImprovedELBubblesIO

6a9c1ab

sbryngelson merged commit d91cb68 into MFlowCode:master Oct 18, 2025
31 checks passed

wilfonba deleted the ImprovedELBubblesIO branch February 23, 2026 16:41

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

I/O Improvements for run time info and EL Bubbles#1015

I/O Improvements for run time info and EL Bubbles#1015
sbryngelson merged 15 commits intoMFlowCode:masterfrom
wilfonba:ImprovedELBubblesIO

wilfonba commented Oct 9, 2025 •

edited by qodo-code-review bot

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

qodo-code-review bot commented Oct 9, 2025

Uh oh!

Uh oh!

Uh oh!

wilfonba commented Oct 9, 2025 •

edited

Loading

Uh oh!

codecov bot commented Oct 9, 2025 •

edited

Loading

Uh oh!

sbryngelson commented Oct 15, 2025

Uh oh!

wilfonba commented Oct 15, 2025

Uh oh!

sbryngelson commented Oct 15, 2025

Uh oh!

wilfonba commented Oct 18, 2025

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Milestone

Development

Uh oh!

3 participants

Conversation

wilfonba commented Oct 9, 2025 • edited by qodo-code-review bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

User description

Description

Type of change

Scope

How Has This Been Tested?

Euler-Lagrange bubble output

2D

3D

run_time.inf Verification

Old

New

Old

New

Post Process Verification CL Output

Checklist

If your code changes any code source files (anything in src/simulation)

PR Type

Description

Diagram Walkthrough

File Walkthrough

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull Request Overview

Reviewed Changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

qodo-code-review bot commented Oct 9, 2025

PR Reviewer Guide 🔍

Uh oh!

Uh oh!

Uh oh!

wilfonba commented Oct 9, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

codecov bot commented Oct 9, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

sbryngelson commented Oct 15, 2025

Uh oh!

wilfonba commented Oct 15, 2025

Uh oh!

sbryngelson commented Oct 15, 2025

Uh oh!

wilfonba commented Oct 18, 2025

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Milestone

Development

Uh oh!

3 participants

wilfonba commented Oct 9, 2025 •

edited by qodo-code-review bot

Loading

`run_time.inf` Verification

If your code changes any code source files (anything in `src/simulation`)

wilfonba commented Oct 9, 2025 •

edited

Loading

codecov bot commented Oct 9, 2025 •

edited

Loading