Date: Sun, 14 Apr 2002 09:21:23 +0000 From: "J. Mallett" <jmallett@FreeBSD.ORG> To: "M. Warner Losh" <imp@village.org> Cc: jmallett@FreeBSD.org, obrien@FreeBSD.org, cvs-committers@FreeBSD.org, cvs-all@FreeBSD.org Subject: Re: cvs commit: src/usr.bin/make job.c job.h main.c make.1 make.c Message-ID: <20020414092122.GA30114@FreeBSD.ORG> In-Reply-To: <20020413.154537.75694059.imp@village.org> References: <20020413053330.A96771@dragon.nuxi.com> <20020413151913.GB26522@FreeBSD.ORG> <20020413164457.GA8591@FreeBSD.ORG> <20020413.154537.75694059.imp@village.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, Apr 13, 2002 at 03:45:37PM -0600, M. Warner Losh wrote: > In message: <20020413164457.GA8591@FreeBSD.ORG> > "J. Mallett" <jmallett@FreeBSD.org> writes: > : On Sat, Apr 13, 2002 at 03:19:13PM +0000, J. Mallett wrote: > : > > : > Thoughts? If that seems reasonable I'll post my diffs relative to NetBSD > : > make(1) should I not run into some place where it pukes. If anyone knows > : > of a place where NetBSD's make(1) has been known not to work with what we > : > need it to work with, please let me know so I can try to fix that. > : > > : > : I just got through a buildworld, so I decided it's time to post the diffs. > : > : I'm going to run it with our make(1) now, and compare performance. If there > : is not a significant boost, and if there aren't things in NetBSD that stand > : out as things we want, I'm not sure whether it's worth it or not. I like the > : ability to use ODE :@<variable>@<string>@, but that's neither here nor there. > : > : The diffs explain: http://people.freebsd.org/~jmallett/NetBSD.make.diff > > These are just to the make program itself. Are you telling me that we > can now make it through a buildworld with NetBSD's make in the tree? See one of my other messages in this thread for what doesn't work the same. NetBSD make(1) understands everything we do, which is what I was primarily concerned about, that make(1) would =fail=, rather than emit different actions than ours. It does do different things in a few places, notably in the toolchain and in ssh. I'm getting ready to start going through the diffs of the build log and seeing where behaviour needs changed, or possibly where we are simply doing the wrong thing. You can make -i buildworld with NetBSD make(1) and have almost everything work. The make(1) utility itself won't "fail", and as a matter of fact, FreeBSD make(1) failed on related things (but not the same things) as NetBSD did on my buildworld, so I suspect my world is in bad need of upgrading. The largest number of differences came from having a 'depend' rule when .PATH contained `depend.c', as their make(1) has tried to build a depend program from depend.c. -- jmallett@FreeBSD.org | C, MIPS, POSIX, UNIX, BSD, IRC Geek. http://www.FreeBSD.org | The Power to Serve "We all need mirrors to remind ourselves who we are -- I'm no different." To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20020414092122.GA30114>