`after` statements in scheduler in `IllinoisGRMHD`

Issue #2576 new
Gabriele Bozzola created an issue

In fixing #2497, I found that several after statements in the schedule.ccl of IllinoisGRMHD (but probably also GiRaFFE) are ill-defined (or I don't know how the scheduler works). In particular, according to grep, the following groups do not exist in the Einstein Toolkit:

  • BSSN_RegisterVars
  • lapse_RegisterVars
  • Lapse_InitSymBound
  • empostid
  • lapsepostid
  • p2c
  • bssn_rhs
  • shift_rhs
  • compute_B_postrestrict
  • mhd_conserv2prims_postrestrict
  • outer_boundaries_on_A_mu (probably this one should be IllinoisGRMHD_outer_boundaries_on_A_mu)
  • compute_B_and_Bstagger_from_A (probably this one should be IllinoisGRMHD_compute_B_and_Bstagger_from_A)

In addition to this:

  • IllinoisGRMHD_compute_B_and_Bstagger_from_A is scheduled in HydroBase_Boundaries, but then IllinoisGRMHD_conserv_to_prims to be after compute_B_and_Bstagger_from_A, but it is scheduled in another group (originally AddToTmunu, but with my fix it is SetTmunu), so I am not sure what’s the use of that after
  • IllinoisGRMHD does not define READ/WRITE statements

Comments (2)

  1. Roland Haas

    Non existent functions in an AFTER or BEFORE (and also IN) clause in schedule.ccl are allowed in Cactus.They can for example be defined by thorns that are not included in the thornlist (and may be private). Eg BSSN_RegisterVars may be of that type. However your list also contains some that would seem to be in IllinoisGRMHD itself so their exclusion looks like a bug to me.

    Assigning to the ILGRMHD author (and maintainer) @Zach Etienne

  2. Log in to comment