From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-in-01.arcor-online.net (mail-in-01.arcor-online.net [151.189.21.41]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (Client CN "mx.arcor.de", Issuer "Thawte Premium Server CA" (verified OK)) by ozlabs.org (Postfix) with ESMTP id 2F2C767B56 for ; Tue, 10 Oct 2006 00:21:56 +1000 (EST) In-Reply-To: <20061009002945.GA2259@localhost.localdomain> References: <20061004033753.GB20636@localhost.localdomain> <3711678A-7AA9-493F-B046-2F459604C028@kernel.crashing.org> <20061009002945.GA2259@localhost.localdomain> Mime-Version: 1.0 (Apple Message framework v752.2) Content-Type: text/plain; charset=US-ASCII; format=flowed Message-Id: <811124BA-A99B-46B9-9A16-4111D937468E@kernel.crashing.org> From: Segher Boessenkool Subject: Re: powerpc: Fix make rules for dtc Date: Mon, 9 Oct 2006 16:21:31 +0200 To: David Gibson Cc: linuxppc-dev@ozlabs.org, Paul Mackerras List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , >>> clean-files := $(zlib) $(zlibheader) $(zliblinuxheader) \ >>> - $(obj)/empty.c >>> + $(obj)/empty.c *.dtb >> >> If this wildcard works at all, it will still do the wrong thing: >> deleting all .dtb files (even the ones that the Makefile cannot >> create again, e.g., the user put them in the tree manually) is >> at best not very polite. > > It certainly works at all My problem with it is the := which might be expanded at the wrong time. Or maybe that won't glob the *.dtb yet, who knows. It's not obviously "working" either way. > (the preceding $(obj)/empty.c, however, is > wrong, I've sent a separate patch for that). One could say the same > thing for *.o in clean targets, Yes, and that's a very well-known problem. > though I guess adding .dtb files is > rather more likely. However, I can't seen an obvious way of > generating a list of the re-creatable dtbs, without an explicit > listing of every file in arch/powerpc/boot/dts which sounds like a bit > of a pain. Suggestions? You shouldn't clean _all_ dtb's that you could build, only the dtb's that you _did_ build. You already have a rule for what dtb's to build I assume; just copy the logic from there. Segher