Segmentation fault when using dose file

When using a dose file (that is read correctly and appears to be in the correct format) for mlma analysis, I get a segmentation fault on 3 different environments (2 MacOS and 1 Unix). When converting the input file to PLINK format (using the --make-bed flag in GCTA) and using that output file with --bfile (instead of --dosage-mach) and all other flags the same I get no such error and valid association results. However, I want to use the imputed dose for association test, and not a converted 0/1/2 dose. I tried to recompile GCTA on MacOS to see if an executable built on my environment  will resolve this, but ran into several library issues, and gave up. Any advice for moving this forward?

./gcta64 --mlma --dosage-mach ../input/chr22.mach.dose ../input/chr22.mach.info --grm ../input/GRM --pheno ../input/outcomes.txt --mpheno 1 --covar ../input/sex.txt --qcovar ../input/age.txt --out tmp
*******************************************************************
* Genome-wide Complex Trait Analysis (GCTA)
* version 1.91.7 beta1
* (C) 2010-2018, The University of Queensland
* Please report bugs to: Jian Yang <jian.yang@uq.edu.au>
*******************************************************************
Analysis started at 12:14:32 MST on Fri Nov 30 2018.
Hostname: <masked>

Accepted options:
--mlma
--dosage-mach ../input/chr22.mach.dose ../input/chr22.mach.info
--grm ../input/GRM
--pheno ../input/outcomes.txt
--mpheno 1
--covar ../input/sex.txt
--qcovar ../input/age.txt
--out tmp

Note: This is a multi-thread program. You could specify the number of threads by the --thread-num option to speed up the computation if there are multiple processors in your machine.

Reading map file of the imputed dosage data from [../input/chr22.mach.info].
108348 SNPs to be included from [../input/chr22.mach.info].
Reading dosage data from [../input/chr22.mach.dose] in individual-major format (Note: may use huge RAM).
(Imputed dosage data for 217 individuals detected).
Imputed dosage data for 217 individuals are included from [../input/chr22.mach.dose].
Reading phenotypes from [../input/outcomes.txt].
There are 11 traits specified in the file [../input/outcomes.txt].
Trait #1 is included for analysis.
Non-missing phenotypes of 217 individuals are included from [../input/outcomes.txt].
Reading quantitative covariates from [../input/age.txt].
1 quantitative covariate(s) of 217 individuals read from [../input/age.txt].
Reading discrete covariate(s) from [../input/sex.txt].
1 discrete covariate(s) of 217 individuals are included from [../input/sex.txt].
Reading IDs of the GRM from [../input/GRM.grm.id].
217 IDs read from [../input/GRM.grm.id].
Reading the GRM from [../input/GRM.grm.bin].
GRM for 217 individuals are included from [../input/GRM.grm.bin].
217 individuals are in common in these files.
1 quantitative variable(s) included as covariate(s).
1 discrete variable(s) included as covariate(s).

Performing MLM association analyses (including the candidate SNP) ...

Performing REML analysis ... (Note: may take hours depending on sample size).
217 observations, 3 fixed effect(s), and 2 variance component(s)(including residual variance).
Calculating prior values of variance components by EM-REML ...
Updated prior values: 0.0463581 0.0499889
logL: 135.623
Running AI-REML algorithm ...
Iter. logL V(G) V(e)
1 136.08 0.03664 0.05747
2 137.11 0.01624 0.07475
3 138.05 0.01646 0.07506
4 138.06 0.01693 0.07572
5 138.07 0.01694 0.07573
6 138.07 0.01694 0.07573
Log-likelihood ratio converged.
Calculating allele frequencies ...

Running association tests for 108348 SNPs ...
Segmentation fault: 11