Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 10 Apr 2012 11:16:19 +0200
From:      Erwin Lansing <erwin@FreeBSD.org>
To:        ports@freebsd.org
Subject:   Re: .if ARCH / BROKEN, or 'NOT_FOR_ARCH'?
Message-ID:  <20120410091619.GS66606@droso.net>
In-Reply-To: <4F83F893.7040500@FreeBSD.org>
References:  <4F83F893.7040500@FreeBSD.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, Apr 10, 2012 at 05:08:35AM -0400, Michael Scheidell wrote:
> I believe there was a discussion a while back, and if you used this:
> 
> .if ${ARCH} == "sparc64"
>   BROKEN=             does not compile on sparc64: assertion failed
> .endif
> 
> it is POSSIBLE that cluster runs that test broken ports could fix them (accidentally), but, wasn't the opinion that you might as well use '
> NOT_FOR_ARCH(s)?  '
> 
The rationele is that we use NOT_FOR_ARCHS and ONLY_FOR_ARCHS for ports
that are not supported on those architecture from the upstream or for
some other reasone are known to not work and never will work on those
architectures.  BROKEN is for less permanent or unknown breakage, like
errors on pointyhat where the one analyzing the logs doesn't have
detailed knownledge of each port and its breakage, and is used as a
warning to users, so they don't try to build a port and get disappointed
after installing all its dependencies, and as a message to the
maintainer that something is wrong.

There are quite a few large grey areas between those, but that's the
general outline.

Erwin

-- 
Erwin Lansing                                    http://droso.dk
erwin@FreeBSD.org                        http:// www.FreeBSD.org



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20120410091619.GS66606>