Hello,
I’m using Luna package to simulate the pulse propogation in a fiber with simple code:
using Luna
output = prop_capillary(125e-6, 3, :He, 1; λ0=800e-9, energy=120e-6, τfwhm=10e-15, λlims=(150e-9, 4e-6), trange=1e-12)
But when running, it stuck at
Found FFTW wisdom at C:\Users\Administrator.julia\scratchspaces\30eb0fb0-5147-11e9-3356-d75b018717ce\lunacache\FFTWcache_1threads
Confirmed, it took more than a minute to finish here,
which is quite long for an example, so it might be worth reporting.
julia> output = prop_capillary(125e-6, 3, :He, 1; λ0=800e-9, energy=120e-6, τfwhm=10e-15, λlims=(150e-9, 4e-6), trange=1e-12)
[ Info: X+Y polarisation not required.
[ Info: Freq limits 0.07 - 2.00 PHz
[ Info: Samples needed: 11991.70, samples: 16384, δt = 83.39 as
[ Info: Requested time window: 1000.0 fs, actual time window: 1366.3 fs
[ Info: Grid: samples 8192 / 16384, ωmax 1.88e+16 / 3.77e+16
[ Info: Using PPT ionisation rate.
[ Info: Found cached PPT rate for 24.59 eV, 800.0 nm
[ Info: Using mode-averaged propagation.
[ Info: No FFTW wisdom found
[ Info: FFTW wisdom saved to /usr/local/build/julia/depot/scratchspaces/30eb0fb0-5147-11e9-3356-d75b018717ce/lunacache/FFTWcache_56threads
[ Info: Found FFTW wisdom at /usr/local/build/julia/depot/scratchspaces/30eb0fb0-5147-11e9-3356-d75b018717ce/lunacache/FFTWcache_56threads
[ Info: FFTW wisdom saved to /usr/local/build/julia/depot/scratchspaces/30eb0fb0-5147-11e9-3356-d75b018717ce/lunacache/FFTWcache_56threads
[ Info: Starting propagation
[ Info: Propagation finished in 0.600 seconds, 367 steps
MemoryOutput["simulation_type", "dumps", "meta", "Eω", "prop_capillary_args", "grid", "stats", "z"]
The second run is much faster, probably because the FFTW plan has been created.
The output is almost the same, with an additional "FFTW wisdom saved " line:
julia> output = prop_capillary(125e-6, 3, :He, 1; λ0=800e-9, energy=120e-6, τfwhm=10e-15, λlims=(150e-9, 4e-6), trange=1e-12)
[ Info: X+Y polarisation not required.
[ Info: Freq limits 0.07 - 2.00 PHz
[ Info: Samples needed: 11991.70, samples: 16384, δt = 83.39 as
[ Info: Requested time window: 1000.0 fs, actual time window: 1366.3 fs
[ Info: Grid: samples 8192 / 16384, ωmax 1.88e+16 / 3.77e+16
[ Info: Using PPT ionisation rate.
[ Info: Found cached PPT rate for 24.59 eV, 800.0 nm
[ Info: Using mode-averaged propagation.
[ Info: Found FFTW wisdom at /usr/local/build/julia/depot/scratchspaces/30eb0fb0-5147-11e9-3356-d75b018717ce/lunacache/FFTWcache_56threads
[ Info: FFTW wisdom saved to /usr/local/build/julia/depot/scratchspaces/30eb0fb0-5147-11e9-3356-d75b018717ce/lunacache/FFTWcache_56threads
[ Info: Found FFTW wisdom at /usr/local/build/julia/depot/scratchspaces/30eb0fb0-5147-11e9-3356-d75b018717ce/lunacache/FFTWcache_56threads
[ Info: FFTW wisdom saved to /usr/local/build/julia/depot/scratchspaces/30eb0fb0-5147-11e9-3356-d75b018717ce/lunacache/FFTWcache_56threads
[ Info: Starting propagation
[ Info: Propagation finished in 0.523 seconds, 367 steps
MemoryOutput["simulation_type", "dumps", "meta", "Eω", "prop_capillary_args", "grid", "stats", "z"]
julia> versioninfo()
Julia Version 1.10.9
OS: Linux (x86_64-suse-linux)
CPU: 16 × AMD Ryzen 9 9950X3D 16-Core Processor
Threads: 14 default, 0 interactive, 7 GC (on 16 virtual cores)
Hi! Sorry to hear you’re having issues with Luna. The problem you’re running into here is probably a stale pidlock around the FFTW cache files. If I’m right, you should be able to fix it by using the Utils.clear_cache() function within Luna:
using Luna
Utils.clear_cache()
We actually recently fixed this issue, we just haven’t released a new version yet.
If this doesn’t fix it, best to open an issue over at GitHub - LupoLab/Luna.jl: Nonlinear optical pulse propagator and we’ll deal with it in more detail.