Date: Mon, 29 Apr 2019 09:33:59 +0700 From: Robert Elz <k...@munnari.oz.au> Message-ID: <22739.1556505...@jinx.noi.kre.to>
| Upon reflection, I'm inclined to instead [...] And upon further reflection, and noticing this sentence in the snprintb() man page (snprintb.3) Finally, each field is delimited by a NUL (`\0') character. which should include the F field - for consistency if nothing else, even though for that one there is nothing to delimit. I've done some searching (not exhaustive for sure) and the only thing I can find that uses 'F' in our tree that I can find is <sys/mman.h> (which provided the newly added example in snprintb.3) which is missing that final NUL after the F field - and which I suspect is the motivation for the "off by one" fix. So, I'm going back to my original plan, except that snprintb.3 doesn't need as much fixing as I thought it might, but <sys/mman.h> needs an update. Further, the test that caught this is dumb, and has data between the F field data and the terminating \0 (which the code conveniently ignores) but which really should not be there - it would inhibit perhaps extending the F format with more data if that were ever useful. kre