Implement gear's stars skipping

As the stars are spending most of their time without any feedback, I am skipping them when no supernovae are produced.

@matthieu Do you accept the modifications done to runner_time_integration.c or should I use some #ifdef and keep only a single evolution function?

added GEAR performance labels

added 1 commit

• 9a6871ea - Correctly reset feedback

Compare with previous version

I have no problems with more function calls. Though I don't like the names. :/ Don't we have a kick_extra() or something like this elsewhere? We could reuse that name convention here.

feedback_kick2/feedback_reset_kick2 and feedback_end_density/feedback_reset_end_density?

The problem with kick_extra is that they are used in both kicks.

added 1 commit

• 14935a4f - Update API

Compare with previous version

I am still getting the same answer. Therefore we can start to merge it.

unmarked as a Work In Progress

I need to think a bit. I am not a big fan of doing things in only one of the kicks. Fundamentally, both kicks should be doing the same thing just with different dt and "who is active" conditions. If possible I'd like to keep that symmetry. Could you do what you want here in the time-step task? That is called only once per step.

marked as a Work In Progress

changed the description

Yes it should be possible.

I totally forgot about the star formation task. Currently, I am not computing the stellar evolution of the new stars. Even if I am pretty sure that I do not have a star that should produce some feedback in its first time step, we should do this in a clean way.

If I do it in the time step task, I need to increment my local ti_current by 1, right?

Yes, as we are looking at the next step. For the new stars, I think you want to make them active on their first step to be safe.

added 1 commit

• 68b34444 - Move stellar computation from kick2 to timestep

Compare with previous version

added 1 commit

• 317f25e5 - Move the h correction into ghost

Compare with previous version

unmarked as a Work In Progress

Everything seems fine now. I have changed the API of feedback_will_do_feedback as we were returning an useless int (now we return nothing).

added 1 commit

• 0ab8e6f4 - Fix CI

Compare with previous version

Why the change in the name of the other function?

As not all the modules are evolving the sparts in this function anymore, I think it makes sense to update the name with something more informative about the location of the function in the task system.

If you do not like it, I can reverse the change.