Re: [PATCH] init/version.c: Silenced sparse warning by declaringthe version string.

From: Daniel Guilak
Date: Mon Jul 14 2008 - 15:03:32 EST


On Fri, 2008-07-11 at 12:18 -0700, Andrew Morton wrote:
> On Tue, 08 Jul 2008 14:21:09 -0700 Daniel Guilak <guilak@xxxxxxxxxxxxxxxxxx> wrote:
>
> > Signed-off-by: Daniel Guilak <daniel@xxxxxxxxxxxxxxxx>
>
> Please always quote the warning or error message in the changelog when
> fixing it. Although it's pretty obvious in this case.
>
> > init/version.c | 1 +
> > 1 files changed, 1 insertions(+), 0 deletions(-)
> >
> > diff --git a/init/version.c b/init/version.c
> > index 9d17d70..041fd82 100644
> > --- a/init/version.c
> > +++ b/init/version.c
> > @@ -16,6 +16,7 @@
> > #define version(a) Version_ ## a
> > #define version_string(a) version(a)
> >
> > +extern int version_string(LINUX_VERSION_CODE);
> > int version_string(LINUX_VERSION_CODE);
> >
> > struct uts_namespace init_uts_ns = {
>
> hrm, what does this thing do? Seems to define
>
> int Version_132634;
>
> Then sticks that in the symbol table (and wastes a bit of bss).
>
> Does anything use it?
>
> Could it be made static?
>
>
ïïApparently it's only used by the ksymoops tool, which is not needed if
(according to README) the kernel is configured with CONFIG_KALLSYMS.

So I will submit a patch depending on this one that will define
version_string only if CONFIG_KALLSYMS is not defined, and hopefully
that will deal with the problem.

--Daniel Guilak

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