Return to BSD News archive
Path: sserve!manuel!munnari.oz.au!uunet!mcsun!news.funet.fi!hydra!klaava!torvalds From: torvalds@klaava.Helsinki.FI (Linus Benedict Torvalds) Newsgroups: comp.unix.bsd Subject: Re: gcc 2.2.2 or 386BSD bug? Message-ID: <1992Jun25.032817.28985@klaava.Helsinki.FI> Date: 25 Jun 92 03:28:17 GMT References: <1992Jun23.232355.1382@gateway.novell.com> <1992Jun24.053953.5550@serval.net.wsu.edu> <1992Jun24.210332.7464@kithrup.COM> Organization: University of Helsinki Lines: 22 In article <1992Jun24.210332.7464@kithrup.COM> sef@kithrup.COM (Sean Eric Fagan) writes: >In article <1992Jun24.053953.5550@serval.net.wsu.edu> hlu@phys1.physics.wsu.edu (Hongjiu Lu) writes: >>>I don't know of any UNIX (or any >>>other OS, for that matter) written in entirely conforming ANSI C. >>Take a look at Linux. > >Last time I looked at it, Linux used inline assembly statements. It is not, >therefore, a conforming application. Indeed. Linux also does a lot of compiler assumptions (unsigned long is casted to pointers and vice versa), but that's not surprising for any OS that has to do hardware-level programming (the MMU needs pointer<->int conversions for bit-twiddling the pointers). I have tried to make it as clean as possible though: I compile linux with -Wall and optimizations, and the warnings reported aren't very many. In fact linux can be compiled only with gcc that uses the gas syntax, and due to problems with earlier versions of gcc, it had better be 2.1 or newer. So no, it's not conforming, but most of it comes pretty close. Linus