[ANN] ThreadsX.jl: Parallelized Base functions

[tkf]

Thanks a lot for the stacktrace! It’s very helpful (even though I’m still puzzled).

Does this occur if you interactively run

using ThreadsX
ThreadsX.sort!(rand(0:0.01:1, 1_000_000))

or

@btime ThreadsX.sort($(rand(0:0.01:1, 1_000_000)))

or from command line:

JULIA_NUM_THREADS=8 julia -e 'using ThreadsX, BenchmarkTools; @btime ThreadsX.sort($(rand(0:0.01:1, 1_000_000)))'

?

@Elrod

Often it goes to 100% (i.e., a single core at max) when it gets stuck.

Do you mean it doesn't use (say) 7 CPUs even when the log message says ```Running: `scaling_nthreads_target.jl` with 7 thread(s)```? (Edit: never mind, you said when it gets stuck)