Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 18 Sep 1998 06:09:14 +1000
From:      Bruce Evans <bde@zeta.org.au>
To:        jb@cimlogic.com.au, karpen@ocean.campus.luth.se
Cc:        current@FreeBSD.ORG
Subject:   Re: 'make world' dying in sbin/atm/atm
Message-ID:  <199809172009.GAA31890@godzilla.zeta.org.au>

next in thread | raw e-mail | index | archive | help

>> > >My world build (make -DNOAOUT -DNOCLEAN world, /usr/obj clean, fairly
>>                                      ^
>>                                      |
>> The problem here is -----------------+
>> 
>> People mistakenly believe that they can use -DNOCLEAN if /usr/obj is
>> clean. That is not so. The build behaves differently. I would prefer
>> the build to test if the objdir is clean and not do the recursive
>> clean.

This belief is not mistaken in the usual self-hosting case.  It is mistaken
in the cross-compile case.  Switching to elf requires some cross-compiling.

>Definitely! I mean, why on earth is the make world doing a recursive clean
>of an empty directory? That always comfused me. At every directory that

Because it is necessary.  An empty canonical obj directory just means that
the the object tree under /usr/obj (or wherever) doesn't need cleaning.
The object directory for every directory in the source tree still
needs cleaning since it may have the following junk in it:
	object files (harmless?)
	.depend (harmful?)
	obj link to a non-canonical place (harmful)
	obj subdirectory (harmful)

It is possible to optimize the cleaning by not redoing it after doing it
initially, but this is not worth doing since only the initial cleaning
cleans everything.

Bruce

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-current" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199809172009.GAA31890>