Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 29 Apr 2010 14:31:57 +0300
From:      =?utf-8?B?QW5kcml1cyBNb3JrxatuYXM=?= <hinokind@gmail.com>
To:        "Dima Panov" <fluffy@freebsd.org>, freebsd-current@freebsd.org
Subject:   Re: Ruby w/clang (Was: Re: [CFT]: ClangBSD is selfhosting, we need testers now)
Message-ID:  <op.vbw8zh2e43o42p@klevas>
In-Reply-To: <201004292205.26216.fluffy@freebsd.org>
References:  <20100416160818.GA69460@freebsd.org> <20100428121637.GA61412@roberto-al.eurocontrol.fr> <201004290240.26848.fluffy@freebsd.org> <201004292205.26216.fluffy@freebsd.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, 29 Apr 2010 14:05:25 +0300, Dima Panov <fluffy@freebsd.org> wrote:
> Ruby is bad?

More like clang is bad, it's a known issue.

> clangbsd errors in my blog:
>
> http://dimapanov.wordpress.com/2010/04/29/clangbsd/
>
> at this moment unbuildable some critical ports:
>
> devel/binutils
> devel/icu[4]
> devel/pcre
> lang/ruby1[89]

I could give you a much longer list. Using clang for ports right now is not
a good idea, a lot of things won't compile, some will be miscompiled or
whatever, some ports rely on gcc specific/undefined behaviour, etc...

We're [slowly] working on it, but for now you should probably stick with gcc
for ports. Reporting broken ports won't really help too much, since we know
what's broken ourselves. What could help is tracking miscompilations (it was
explained earlier how to do it), but even for that we're waiting for sound
related bug to be fixed on llvm end, to see how does that affect other
broken ports.

--
Andrius



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?op.vbw8zh2e43o42p>