On Oct 22, 11:04am, k...@munnari.oz.au (Robert Elz) wrote: -- Subject: Re: FYI: POSIX update - SIGWINCH and 'struct winsize' (etc) to be
| Yes, you mentioned that one yesterday, that is certainly a possibility, | though it does mean that apps that include <sys/ttycom.h> (most likely | indirectly because of including <sys/ioctl.h>) have to have _NETBSD_SOURCE | defined if they are to access struct ptmget or any of the ioctl opcodes | (things like TIOCEXCL, TIOCGPGRP, ...) which would be a new requirement. Yes, and they should (have that defined) since this header is really part of the implementation and should not be included directlty for portability reasoons. | Maybe that doesn't matter (or not much), as it is very rare (I think) to | ever see anything compiled on NetBSD without _NETBSD_SOURCE being defined | (isn't it?) and what happens for compiles not on NeBSD isn't relevant for | this discussion (nothing is affecting that.) That's right. | | So, is that correct? If so, this (move the #ifdef _NETBSD_SOURCE) is clearly | the best solution. But if we need to keep compat with old sources that | are compiled without _NETBSD_SOURCE and do any tty ioctl type stuff at all, | then it simply would not work. I think it is certainly the easiest and most minimum impact solution. christos
