How to run BOLT-LMM against a Plink simulated dataset?

Question

[b.ambrozio]

I'm trying run BOLT-LMM against a simulated dataset, which were generated with plink --simulate, however no success:

./BOLT-LMM_v2.3.4/bolt --bfile=ds1_1gt10k --phenoUseFam --statsFile ds1_1gt10k.bolt --LDscoresUseChip --lmm
...
ERROR: LOCO mixed model analysis requires >= 2 chromosomes or chunks
       To run this analysis, set --numLeaveOutChunks

The documentation don't say nothing about numLeaveOutChunks parameter. I've tried some arbitrary numbers as well, and things started working, however, after some time running, it ended up with another error:

./BOLT-LMM_v2.3.4/bolt --bfile=ds1_1gt10k --phenoUseFam --statsFile ds1_1gt10k.bolt --LDscoresUseChip --lmm --numLeaveOutChunks=5
...
Intel MKL ERROR: Parameter 6 was incorrect on entry to DGELS .

Intel MKL ERROR: Parameter 6 was incorrect on entry to DGELS .

Intel MKL ERROR: Parameter 6 was incorrect on entry to DGELS .
Intercept of LD Score regression for ref stats:   0.000 (0.000)
Estimated attenuation: -nan (nan)
Intercept of LD Score regression for cur stats: 0.000 (0.000)
Calibration factor (ref/cur) to multiply by:      -nan (nan)

Time for computing Bayesian mixed model assoc stats = 0.802051 sec

Calibration stats: mean and lambdaGC (over SNPs used in GRM)
  (note that both should be >1 because of polygenicity)
Segmentation fault