From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from pasmtpb.tele.dk ([80.160.77.98]:56543 "EHLO pasmtpB.tele.dk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1757776AbYGFV1X (ORCPT ); Sun, 6 Jul 2008 17:27:23 -0400 Date: Sun, 6 Jul 2008 23:28:10 +0200 From: Sam Ravnborg Subject: Re: kconfig - hex are considered decimal Message-ID: <20080706212810.GA8025@uranus.ravnborg.org> References: <20080629115120.GA4654@uranus.ravnborg.org> <20080630204032.GA29429@uranus.ravnborg.org> <20080701061241.GA4922@uranus.ravnborg.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: Sender: linux-kbuild-owner@vger.kernel.org List-ID: To: Roman Zippel Cc: linux-kbuild On Wed, Jul 02, 2008 at 02:39:08PM +0200, Roman Zippel wrote: > Hi, > > On Tue, 1 Jul 2008, Sam Ravnborg wrote: > > > - When changing I see the 0x prefix - which I consider OK for hex values > > - When saving I see the 0x prefix in autoconf.h + .config > > I disagree with the proposed patch. It only makes cosmetical changes, but > it doesn't fix the current inconsistenties. Either we do it properly as in > the patch below or I'd rather keep it as is. > The patch checks the default values more strictly, I'm open to opinions > whether that warning is useful or maybe just check that hex values have > the 0x prefix. Thanks - I divided your patch up in two. First one that adds the new checks And the second that fixes the warnings > > > See patches where kconfig start to use the new > > method for 'all*config' targets - they are cc:ed to you > > and available on linux-kbuild. > > I simply started to use conf_set_all_new_symbols() and > > as a result I could simplify conf.c. > > > > And things got so simple that I dropped the patches > > introducing aconf.c. > > I'd actually prefer to move everything noninteractive out of conf.c, as > this would be the right place for this: > > > -> Give Redhat's nonint_oldconfig a closer look and see > > how to integrate this functionality > > This should be pretty much do it: > > for_all_symbols(i, sym) { > sym_calc_value(sym); > if (sym->flags & SYMBOL_WRITE && !sym_has_value(sym)) > printf(...); > } > > The point is that this has nothing to do with what's already in conf.c, > this should be a separate utility for all noninteractive kconfig jobs. I do not disagree with you here. But for now it was just more natural to add it where I deleted the other stuff. I will move it to a separate .c file later. Sam