Re: UNS: Re: [PATCH] kconfig: add *_silentdefconfig feature for config targets

[Andres Salomon <dilinger@queued.net>]

On Tue, 21 Aug 2007 17:29:36 +0200 (CEST)
Roman Zippel <zippel@linux-m68k.org> wrote:

> Hi,
> 
> On Mon, 20 Aug 2007, Andres Salomon wrote:
> 
> > AFAICT, there is nothing similar when using *_defconfig; one must copy
> > a .config manually, and then run silentoldconfig.  Simply running the
> > associated _defconfig will quietly update the config (which may silently
> > drop config options).  This patch adds a *_silentdefconfig target, with
> > semantics similar to silentoldconfig.  It will take the defconfig from
> > arch/$(ARCH)/configs/$x_defconfig, check for changes, and if there are
> > none, write out a .config.  If there have been changes and stdin is
> > valid, it will prompt for updates.  If there have been changes and
> > stdin is not valid, it will bail out with an error.
> 
> I would really like to avoid another input mode.
> I think it be better to implement this as a combination of "-s -D 
> <default>" and the silent mode is adjusted to read another config instead 
> of .config if defconfig_file is set.
> 

As would I; however, that requires using getopt() (or equivalent).  I wasn't
sure if there was some opposition to this..

Of course, we'll still need some way from the makefile to call it.  I take
it you're not opposed to 'make foo_silentdefconfig'?


> > A few things to note:
> >   - Using getopt() in scripts/kconfig/conf.c would likely simplify things,
> >     but that's a much larger patch.  Is there a reason we don't use it?
> 
> Not really.
> 

Cool, I'll submit a patch that does this.


> >   - To make it truly silent, I had to change conf_write() to accept an
> >     additional arg.  The alternative is to not have conf_write spit out
> >     any information when it writes a file.  Personally, I don't see the need
> >     for it to spit out information, but I figured I'd take the more cautious
> >     route.  If folks don't care, I can update this patch to remove it.
> 
> I rather want to keep this print. The .config is already only written, 
> when it has to be in this mode and then I also want to be notified about 
> it.
> 
> >   - We seem to switch between using _() and not using it for strings; I'm
> >     assuming that we don't actually care about i18n in conf.c, and that the
> >     _() stuff was just copied from elsewhere.  If that's not the case, I
> >     can update the patch to wrap strings properly.
> 
> I try to keep this uptodate, but I don't really check for this.
> 

I'm not sure I follow; should I use _(), or no?


> bye, Roman



-- 
Andres Salomon <dilinger@queued.net>