Date: Sun, 9 Aug 1998 22:00:22 +0000 (GMT) From: Terry Lambert <tlambert@primenet.com> To: mike@smith.net.au (Mike Smith) Cc: jdp@polstra.com, rivers@dignus.com, hackers@FreeBSD.ORG Subject: Re: C and static initialization with unions Message-ID: <199808092200.PAA20417@usr04.primenet.com> In-Reply-To: <199808082206.PAA02145@antipodes.cdrom.com> from "Mike Smith" at Aug 8, 98 03:06:45 pm
next in thread | previous in thread | raw e-mail | index | archive | help
> I've had (several) run-ins with const poisoning trying to port a large > third-party codebase. It's sufficiently bogus to require -traditional > to build, and I expect not too novel in that regard. I have a number of programs which I like to use, but which I do not want to become the maintainer of (for one thing, I don't have a huge FTP server and can't offer cvsup, etc.). These programs do things like modify declared strings (on the assumption that they will be stored in data section, not code section, what with them being data, and all...). They also do things like modify variables in signal handlers and in subfunctions of functions that also use the variable. Without the -traditional flag, this would require the use of the "volatile" keyword. Other examples exist, including the use of int arguments to implement varaddic functions, and casting to char * instead of void * for opaque conversion. Sure, it's be nice if there were three of me (;-)) and I could do all this and the stuff that I want to do as well, but there aren't, and I can't. > If we support compilation of K&R application code, we should attempt to > make sure that system headers function correctly in that regard. Yes, please. Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199808092200.PAA20417>