Date: Fri, 14 Dec 2007 23:00:05 +0100
From: Sam Ravnborg <sam@ravnborg.org>
To: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: linux-fbdev-devel@lists.sourceforge.net,
       Andrew Morton <akpm@linux-foundation.org>, linux-kernel@vger.kernel.org,
       Antonino Daplas <adaplas@gmail.com>
Subject: Re: [Linux-fbdev-devel] [PATCH] logo: move declarations of logos to linux_logo.h
Message-ID: <20071214220005.GA22937@uranus.ravnborg.org>
References: <20071209214027.GG9440@joi> <20071213023111.30ddc487.akpm@linux-foundation.org> <20071213213130.GA1822@joi> <Pine.LNX.4.64.0712142148390.4881@anakin>
Mime-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <Pine.LNX.4.64.0712142148390.4881@anakin>
User-Agent: Mutt/1.4.2.1i
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1747
Lines: 40

On Fri, Dec 14, 2007 at 09:49:03PM +0100, Geert Uytterhoeven wrote:
> On Thu, 13 Dec 2007, Marcin Slusarz wrote:
> > On Thu, Dec 13, 2007 at 02:31:11AM -0800, Andrew Morton wrote:
> > > On Sun, 9 Dec 2007 22:40:31 +0100 Marcin Ślusarz <marcin.slusarz@gmail.com> wrote:
> > >
> > > > logo: move declarations of logos to linux_logo.h
> > > >
> > > > there was a mismatch between externs in logo.c and code generated by pnmtologo
> > > > (on old tree, you need to rm drivers/video/logo/logo_*.c before compilation)
> > >
> > > This patch (after I fixed it) keeps on causing trouble: mismatches between
> > > the header file declarations and the actual definitions of the lookup
> > > tables.  This happens as I apply and unapply the patch.
> > >
> > > It _shouldn't_ happen, but there would appear to be some missing makefile
> > > rules.  I guess the C files aren't being regenerated when
> > > scripts/pnmtologo.c has been altered, so the C files aren't matching the
> > > header which this patch alters.
> > 
> > Ok, updated patch below:
> > 
> > logo: move declarations of logos to linux_logo.h
> > 
> > there was a mismatch between externs in logo.c and code generated by pnmtologo
> 
> This may cause section conflicts, cfr.
> http://lkml.org/lkml/2007/10/24/65

If I recall correct then if the first __initdata variable is const then
all date declared __initdata are attempted put in the const section.
So maybe a bit reordering could help it?

But that would serve a fat explanation why...

	Sam
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/