Re: [PATCH 2/2] s390/traps: Only define is_valid_bugaddr() under CONFIG_GENERIC_BUG

[Nathan Chancellor <nathan@kernel.org>]

On Mon, Dec 04, 2023 at 10:06:43AM +0100, Alexander Gordeev wrote:
> On Thu, Nov 30, 2023 at 05:22:33PM -0700, Nathan Chancellor wrote:
> 
> Hi Nathan,
> 
> > When building with -Wmissing-prototypes without CONFIG_GENERIC_BUG,
> > there is a warning about a missing prototype for is_valid_bugaddr():
> > 
> >   arch/s390/kernel/traps.c:46:5: warning: no previous prototype for 'is_valid_bugaddr' [-Wmissing-prototypes]
> >      46 | int is_valid_bugaddr(unsigned long addr)
> >         |     ^~~~~~~~~~~~~~~~
> > 
> > The prototype is only declared with CONFIG_GENERIC_BUG, so only define
> > the function under the same condition to clear up the warning, which
> > matches other architectures.
> 
> Thanks for the fix!
> The patch is fine, although I have not been able to reproduce the warning. 
> How did you trigger it?

Thanks for taking a look! I am able to trigger this on v6.7-rc4 with:

  # Flip off CONFIG_BUG in menuconfig
  $ make -skj"$(nproc)" ARCH=s390 CROSS_COMPILE=s390-linux- KCFLAGS=-Wmissing-prototypes defconfig menuconfig arch/s390/kernel/traps.o
  arch/s390/kernel/traps.c:46:5: warning: no previous prototype for 'is_valid_bugaddr' [-Wmissing-prototypes]
     46 | int is_valid_bugaddr(unsigned long addr)
        |     ^~~~~~~~~~~~~~~~

I initially found this by testing just the tinyconfig target.

> > Signed-off-by: Nathan Chancellor <nathan@kernel.org>
> > ---
> >  arch/s390/kernel/traps.c | 2 ++
> >  1 file changed, 2 insertions(+)
> ...
> 
> Thanks!