This is strange bug, even if the function isn't used correctly:

library(Biostrings)

Warning messages:
1: package ‘BiocGenerics’ was built under R version 3.4.2
2: package ‘S4Vectors’ was built under R version 3.4.1
3: In eval(quote({ : bytecode version mismatch; using eval
4: package ‘IRanges’ was built under R version 3.4.1
5: undefined slot classes in definition of "XRaw": elementMetadata(class "DataTableORNULL")
6: undefined slot classes in definition of "XInteger": elementMetadata(class "DataTableORNULL")
7: undefined slot classes in definition of "XDouble": elementMetadata(class "DataTableORNULL")
> DNAString("TTGAAA-CTC-N")
  12-letter "DNAString" instance
seq: TTGAAA-CTC-N
> DNAString("TTGAAA-CTC-N", "A")
Error in .normargSEW(start, "start") :
  'start' must be a vector of integers
> DNAString("TTGAAA-CTC-N", "A", "B")
Error in .normargSEW(start, "start") :
  'start' must be a vector of integers
> DNA_BASES = c('A', 'G', 'C', 'T', 'N')
> nnuc = 0
> mkAllStrings(DNA_BASES, nnuc*2 + 1)
[1] "A" "G" "C" "T" "N"
> typeof(mkAllStrings(DNA_BASES, nnuc*2 + 1))
[1] "character"
> bases = mkAllStrings(DNA_BASES, nnuc*2 + 1)
> DNAStringSet(bases)

*** caught segfault ***
address 0x1, cause 'memory not mapped'

Traceback:
1: .Call(.NAME, ..., PACKAGE = PACKAGE)
2: .Call2("new_XStringSet_from_CHARACTER", ans_class, ans_elementType,     x, start(solved_SEW), width(solved_SEW), get_seqtype_conversion_lookup("B",         seqtype), PACKAGE = "Biostrings")
3: .charToXStringSet(seqtype, x, start, end, width, use.names)
4: XStringSet("DNA", x, start = start, end = end, width = width,     use.names = use.names)
5: XStringSet("DNA", x, start = start, end = end, width = width,     use.names = use.names)
6: DNAStringSet(bases)

Possible actions:
1: abort (with core dump, if enabled)
2: normal R exit
3: exit R without saving workspace
4: exit R saving workspace