Date: Fri, 01 Aug 2008 09:53:25 +0100 From: Dieter <freebsd@sopwith.solgatos.com> To: freebsd-alpha@freebsd.org Subject: Re: firefox3 build fails on alpha Message-ID: <200808011653.QAA17051@sopwith.solgatos.com> In-Reply-To: Your message of "Fri, 01 Aug 2008 16:58:22 %2B0200." <20080801145822.GC1356@freebie.xs4all.nl>
next in thread | previous in thread | raw e-mail | index | archive | help
> Thing is, ports have not been built on -alpha for > ages now. There is not even a working Alpha machine anymore in the > FreeBSD.org cluster for ports folks to test on if they were inclined to do > so. ports have always been tricky on alpha, given the limited # of people > running FreeBSD on Alpha. I've been running NetBSD/Alpha for many years, and beaten a lot of software into submission on it. I don't recall a single problem that wasn't ILP32 vs LP64. FreeBSD runs on AMD64, AMD64 is LP64, so what is the big problem getting apps to run on Alpha once they run on AMD64? Yes, I know that a lot of programs still aren't LP64 clean even today and therefore don't run properly on AMD64. But that isn't specific to Alpha. Turn the portability compiler warnings on (e.g. require function prototypes), fix the compiler warnings. At that point programs will run as well on AMD64 and Alpha as they do on x86. In theory there are 64 bit problems that compilers can't warn about, but I have never found any in code I've ported. When I compile ports on FreeBSD/AMD64 I usually see a bunch of compiler warnings fly by. Just because the compiler says warning rather than error doesn't mean you can ignore it. Warnings can and do point to actual bugs. Good quality code has NO compiler warnings. The only program I recall giving up on was firefox, and only because it was soooo big, sooo badly written, with soooo many problems, and so much hostility from the firefox people. Firefox in the original Greek means "segmentation fault".
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200808011653.QAA17051>