Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 20 Mar 2005 18:28:37 +0100
From:      Volker Stolz <vs@FreeBSD.org>
To:        ports@FreeBSD.org
Subject:   Re: Max recursion level (500) exceeded.
Message-ID:  <20050320172837.GA86028@i2.informatik.rwth-aachen.de>
In-Reply-To: <423BA1EC.9020806@FreeBSD.org>
References:  <Pine.LNX.4.44.0503181929500.13080-100000@pancho> <423BA1EC.9020806@FreeBSD.org>

next in thread | previous in thread | raw e-mail | index | archive | help
In gmane.os.freebsd.devel.ports, you wrote:
> Mark Linimon wrote:
>> On Fri, 18 Mar 2005, Adam Weinberger wrote:
>>>The problem lies in the gcc32 port. It redefines USE_GCC. It, and the 
>>>other gcc ports, need not to do that.
>> 
>> 
>> IIRC gcc3.2 had a USE_GCC definition that its internal make stuff
>> needed whose existance predated the creation of bsd.gcc.mk which
>> introduced a USE_GCC variable intended to be used internal to the
>> ports framework which has different semantics.  AFAIK gcc32 has
>> been broken since that time.
>> 
>> I do not have time to test this hypothesis but would suggest trying
>> a patchset with s/USE_GCC/INTERNAL_USE_GCC/ or something.
>> 
>> It would be great if someone could investigate and fix this problem ...
>
> Here is a simple solution.
> First of all, OOo needs to remove the gcc32 BUILD_DEPENDS, and replace 
> it with USE_GCC=3.2.
> Then, for each gcc port, simply replace the USE_GCC=2.7+ line with
> .undef USE_GCC
>
> If people want to test this, I'll commit it.

*sigh* I just spent some time getting my head around this. If I understand
this correctly, there's no problem at all on 4.x because gcc-2 is sufficient,
and the .undef doesn't seem to change anything. I wish I'd never touched this
in the first place.
-- 
http://www-i2.informatik.rwth-aachen.de/stolz/ *** PGP *** S/MIME



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20050320172837.GA86028>