Date: Wed, 14 Oct 2015 16:23:53 -0700 From: NGie Cooper <yaneurabeya@gmail.com> To: Bryan Drewery <bdrewery@freebsd.org> Cc: "src-committers@freebsd.org" <src-committers@freebsd.org>, "svn-src-all@freebsd.org" <svn-src-all@freebsd.org>, "svn-src-head@freebsd.org" <svn-src-head@freebsd.org> Subject: Re: svn commit: r289334 - head/share/mk Message-ID: <CAGHfRMB31sZ_9K2mSt-2d8uVZKMq7jzR8U23JTcUtFYjWzNDzw@mail.gmail.com> In-Reply-To: <561ED2B3.4030909@FreeBSD.org> References: <201510142030.t9EKUWNY089355@repo.freebsd.org> <561ED2B3.4030909@FreeBSD.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, Oct 14, 2015 at 3:09 PM, Bryan Drewery <bdrewery@freebsd.org> wrote: ... > So the reason I have been tinkering with this code is because it is so > mysterious to me, given the lack of comments and seemingly out-of-place > nature of it. > > It turns out that even moving 'buildconfig' as a recurse target creates > a surprising situation that will break with parallel builds since > 'buildconfig' is hooked into 'all', thus 'all' and 'buildconfig' both > end up recursing when calling 'make all'. This explains the make() check > here to avoid recursing if called with 'all' (even though it is checking > the opposite, only calling 'buildconfig' directly to recurse). Comments > are not a sin. Agreed. The original change was done over a decade ago: https://svnweb.freebsd.org/base/head/share/mk/bsd.subdir.mk?r1=96667&r2=96668&pathrev=289334&; . It seems that the targets used to have different names and ru@ changed them. Succinct comments would probably be a good idea in the .mk files. I'm not sure what the performance is like if bmake needs to (re-)read/(re-)evaluate make snippets. Thanks! -NGie
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAGHfRMB31sZ_9K2mSt-2d8uVZKMq7jzR8U23JTcUtFYjWzNDzw>