Speedup the unskip and scheduler_start process

A bunch of performance and memory improvements:

• In scheduler_start(), do not reset the tasks' tic and toc or task debugging properties (see #611 (closed)).
• Make each cell that hosts a task carry a flag saying it has a task (see discussion in #523 (closed)).
• When unskipping, only recurse down to levels where tasks still exist (see discussion in #523 (closed)).
• When unskipping, duplicate the list of active cells in order to launch the threads not just on the cells but also on the different task types to enhance parallelization
• Move all the unskip logic to a separate file (part of #427 (closed)).
• Since we know at which level the tasks are, we can recurse less in the creation of send/recv tasks and use a lot fewer task links.

We go from this: http://icc.dur.ac.uk/~jlvc76/SWIFT/master_unskip/threadpool-index.html to this: http://icc.dur.ac.uk/~jlvc76/SWIFT/unskip_speedup4/threadpool-index.html for an EAGLE-25 (non-MPI).

Note that the colours change so it can be confusing.

mentioned in issue #611 (closed)

added 1 commit

• cb5c2e2b - Fixes for Intel compiler. Better granularity for the task unskiping threadpool.

Compare with previous version

changed title from Speedup the unksip and scheduler_star process to Speedup the unksip and scheduler_start process

changed the description

@rgb this is the change that will interest you.

changed title from Speedup the unksip and scheduler_start process to Speedup the unskip and scheduler_start process

added 3 commits

• 73866ca7 - Make the task dump statistics use e->tic_step to decide whether a task was run this step or not.
• daf99c68 - Only set the tasks' tic/toc for active tasks.
• 9a428b00 - Don't need to set the tics or tocs as we pick out values set in the current step

Compare with previous version

changed the description

I have merged your latest patch in here for final review.

Tried a larger test, EAGLE_50/IC, with full EAGLE physics and it aborted with:

[0002] [00349.3] timestep.h:get_part_timestep():177: part (id=3913482190279) wants a time-step (0.000000e+00) below dt_min (1.000000e-10)
[0002] [00349.3] timestep.h:get_part_timestep():177: part (id=3909264098833) wants a time-step (0.000000e+00) below dt_min (1.000000e-10)

So it looks like something is being skipped that shouldn't. The run is in:

/cosma7/data/dp004/pdraper/swiftsim-by-ticks/examples/EAGLE_ICs/EAGLE_50

Thanks. That's odd since on the first step we do not unskip but use engine_marktasks() instead.

So if its not the unskip it must be missing some tasks...

I can't test it right now but do you know whether that test runs with master?

Thought you might ask that. Not yet is the answer.

Looks like this answer has to wait until tomorrow. Queues have become busy.

But have the usual rubbish time estimates. The same setup has ran, currently at step 32, so will stop.

Ok, thanks. I'll investigate.

There appears to be something incorrect with the ICs. The cosmological check fails.

I have updated the ICs. They are now correct but the code still crashes in the same way.