test threaded time integration

[ranocha]

I added some tests with threaded time intergation so that we can catch issues such as #1577 and SciML/OrdinaryDiffEq.jl#1989 in our CI pipeline.

I also reduced the precompilation workload of OrdinaryDiffEq.jl for CI tests to make it less expensive. We don't use all of the methods that are precompiled with the old settings, so it doesn't really help to compile them all.

There is currently an issue with threaded time integration and arrays of SVectors used by the flux differencing DGMulti solver that I will report upstream.
Edit: See JuliaSIMD/StrideArrays.jl#77

@trixi-framework/developers Anybody up for a review?

[codecov]

Codecov Report

Merging #1581 (59c6bf0) into main (67d137d) will increase coverage by 3.39%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##             main    #1581      +/-   ##
==========================================
+ Coverage   92.76%   96.15%   +3.39%     
==========================================
  Files         384      384              
  Lines       31959    31998      +39     
==========================================
+ Hits        29645    30767    +1122     
+ Misses       2314     1231    -1083     

Flag	Coverage Δ
unittests	96.15% <ø> (+3.39%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
examples/dgmulti_2d/elixir_euler_curved.jl	100.00% <ø> (+100.00%)	⬆️
...amples/tree_2d_dgsem/elixir_advection_diffusion.jl	100.00% <ø> (ø)

... and 37 files with indirect coverage changes