[dschwoerer]

Fixes: #3517
With this, at least for 1D meshes, the x parallelization works.

@totork Can you verify for fully 3D?

[totork]

For some reason when I run with this it spams the output with

7.500e+03          1       5.88e-02    12.1    0.0    0.0   80.4    7.4
	Option mesh:outer_x = true (default)
	Option mesh:inner_x = false (default)
	Option mesh:outer_x = true (default)
	Option mesh:inner_x = false (default)
	Option mesh:outer_x = true (default)
	Option mesh:inner_x = false (default)
	Option mesh:outer_x = true (default)
	Option mesh:inner_x = false (default)
	Option mesh:outer_x = true (default)
	Option mesh:inner_x = false (default)
	Option mesh:outer_x = true (default)
	Option mesh:inner_x = false (default)
	Option mesh:outer_x = true (default)
	Option mesh:inner_x = false (default)
8.000e+03          7       9.81e-02    41.4    0.0    0.1   54.0    4.6
	Option mesh:outer_x = true (default)
	Option mesh:inner_x = false (default)
8.500e+03          1       6.45e-02     4.9    0.0    0.0   94.5    0.6
	Option mesh:outer_x = true (default)
	Option mesh:inner_x = false (default)
9.000e+03          1       6.49e-02    11.1    0.0    0.0   88.2    0.6
	Option mesh:outer_x = true (default)
	Option mesh:inner_x = false (default)
9.500e+03          1       6.78e-02    10.6    0.0    0.0   88.8    0.6
	Option mesh:outer_x = true (default)
	Option mesh:inner_x = false (default)
1.000e+04          1       6.10e-02    11.8    0.0    0.0   87.5    0.6

Run finished at  : Wed Oct 28 01:04:00 2026
Run time : 3 s
	Option datadir = data (default)
	Option settingsfile = BOUT.settings (default)
Writing options to file data/BOUT.settings
	Option time_report:show = 0 (default)
Finalising PETSc

And it does not seem to run aswell.

[dschwoerer]

	Option mesh:outer_x = true (default)
	Option mesh:inner_x = false (default)

Is coming from YBoundary being re-created all over again. Is that maybe from Div_par_K_Grad_par_mod with petsc operator?

What is the error? Same as before?

[totork]

	Option mesh:outer_x = true (default)
	Option mesh:inner_x = false (default)

Is coming from YBoundary being re-created all over again. Is that maybe from Div_par_K_Grad_par_mod with petsc operator?

When i run with the current BOUT++ commit in Hermes-3, it does not spam this.

What is the error? Same as before?

Same error as before

[stirkas]

I happened to see this issue/PR and wanted to mention there is also a related issue with using the new petsc operator for thermal conduction with hermes3 (with use_petsc_operator=true).

On top of your change in petsc_operators.hxx which changed PETSC_DECIDE there I also had to change PETSC_DECIDE to n_local on line 453 in petsc_operators.cxx:

from

BOUT_DO_PETSC(MatSetSizes(*temp, n_local, PETSC_DECIDE, n_global, n_global));

to

BOUT_DO_PETSC(MatSetSizes(*temp, n_local, n_local, n_global, n_global));

[dschwoerer]

Thanks a lot, that looks indeed like it needs to be fixed.

@totork does it work now?

	Option mesh:outer_x = true (default)
	Option mesh:inner_x = false (default)

Is coming from YBoundary being re-created all over again. Is that maybe from Div_par_K_Grad_par_mod with petsc operator?

When i run with the current BOUT++ commit in Hermes-3, it does not spam this.

Like the current bout++/next? Or what ever hermes-3 currently tracks in master?

[totork]

Seems to work :)