In today's Austin Group meeting (the group behind POSIX), it was noted that there is a potential discrepancy between C2Y and POSIX regarding whether strnlen(3) operates on "strings" (the C standard defines this as a NUL-terminated sequence) vs. length-limited arrays of characters (which may or may not contain a NUL byte within the bounds given). https://www.austingroupbugs.net/view.php?id=1834 One of the contributors to the C standard (Chris Bazeley) then replied that his proposal for strnlen to C2Y was based in part on the existing wording of https://man7.org/linux/man-pages/man3/strnlen.3.html, where the man pages were not as precise as the POSIX Issue 8 wording, as well as the fact that C is already inconsistent in its wording for strnlen_s; although he also expressed an openness to revising his draft based on consensus. It may be worth some preliminary cleanup patches to the Linux man pages for strnlen and friends to follow the example of POSIX in distinguishing when an argument is a string (must have NUL terminator) vs a length-bounded array of bytes (which may or may not include a NUL terminator), to make it easier for everyone to share common wording when C2Y eventually adds the function. -- Eric Blake, Principal Software Engineer Red Hat, Inc. Virtualization: qemu.org | libguestfs.org
