Date: Mon, 20 Jun 2005 21:38:32 -0700 From: "David O'Brien" <obrien@FreeBSD.org> To: Robert Watson <rwatson@FreeBSD.org> Cc: freebsd-current@FreeBSD.org, "M. Warner Losh" <imp@bsdimp.com> Subject: Re: dhclient less functional with nanobsd because of NO_CXX Message-ID: <20050621043831.GC93634@dragon.NUXI.org> In-Reply-To: <20050620201157.J26664@fledge.watson.org> References: <20050619012425.L56734@fledge.watson.org> <20050620.124905.35871665.imp@bsdimp.com> <20050620201157.J26664@fledge.watson.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Jun 20, 2005 at 08:13:43PM +0100, Robert Watson wrote:
> On Mon, 20 Jun 2005, M. Warner Losh wrote:
> Well, actually, it's a granularity issue in the src/sbin/Makefile and its
> conditional handling of devd based on NO_CXX without taking into account
> whether we're building or installing. Hence my comments about needing to
> provide some make infrastructure, since devd isn't the only component that
> falls into precisely this pit:
>
> SUBDIR= adjkerntz \
> ...
> ${_devd} \
> ...
>
> .if !defined(NO_CXX)
> _devd= devd
> .endif
We should change this to test for the existance of the built devd vs.
using the NO_CXX knob when the target is "install". If we built a 'devd'
during 'make buildworld' then we should install it.
What we need is for NO_CXX to gain a little logic - if set during a build
target, don't build binaries written in C++. If set during an install
target, don't install any C++ toolchain bits and have no farther meaning.
--
-- David (obrien@FreeBSD.org)
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20050621043831.GC93634>