Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
MIME-Version: 1.0
References: <20181208154624.6504-1-malat@debian.org> <CA+7wUswJy0z_+aytawQv6+1pOVvH2z8SfOz1sxziHtgjqZ1F9g@mail.gmail.com>
 <e95cb3ea-019f-28b0-0af0-ad8c80b3106b@c-s.fr>
In-Reply-To: <e95cb3ea-019f-28b0-0af0-ad8c80b3106b@c-s.fr>
From:   Mathieu Malaterre <malat@debian.org>
Date:   Fri, 15 Feb 2019 09:46:01 +0100
Message-ID: <CA+7wUswX34amxhJW827tm9zmSaq6T9NvEpxWpgcofGNL=3Kmxg@mail.gmail.com>
Subject: Re: [PATCH] powerpc/ptrace: Add prototype for function pt_regs_check
To:     Christophe Leroy <christophe.leroy@c-s.fr>
Cc:     Michael Ellerman <mpe@ellerman.id.au>,
        linuxppc-dev <linuxppc-dev@lists.ozlabs.org>,
        Paul Mackerras <paulus@samba.org>,
        Oleg Nesterov <oleg@redhat.com>,
        LKML <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 8BIT
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Fri, Feb 15, 2019 at 9:21 AM Christophe Leroy
<christophe.leroy@c-s.fr> wrote:
>
>
>
> Le 15/02/2019 à 09:11, Mathieu Malaterre a écrit :
> > On Sat, Dec 8, 2018 at 4:46 PM Mathieu Malaterre <malat@debian.org> wrote:
> >>
> >> `pt_regs_check` is a dummy function, its purpose is to break the build
> >> if struct pt_regs and struct user_pt_regs don't match.
> >>
> >> This function has no functionnal purpose, and will get eliminated at
> >> link time or after init depending on CONFIG_LD_DEAD_CODE_DATA_ELIMINATION
> >>
> >> This commit adds a prototype to fix warning at W=1:
> >>
> >>    arch/powerpc/kernel/ptrace.c:3339:13: error: no previous prototype for ‘pt_regs_check’ [-Werror=missing-prototypes]
> >>
> >> Suggested-by: Christophe Leroy <christophe.leroy@c-s.fr>
> >> Signed-off-by: Mathieu Malaterre <malat@debian.org>
> >> ---
> >>   arch/powerpc/kernel/ptrace.c | 4 ++++
> >>   1 file changed, 4 insertions(+)
> >>
> >> diff --git a/arch/powerpc/kernel/ptrace.c b/arch/powerpc/kernel/ptrace.c
> >> index a398999d0770..341c0060b4c8 100644
> >> --- a/arch/powerpc/kernel/ptrace.c
> >> +++ b/arch/powerpc/kernel/ptrace.c
> >> @@ -3338,6 +3338,10 @@ void do_syscall_trace_leave(struct pt_regs *regs)
> >>          user_enter();
> >>   }
> >>
> >> +void __init pt_regs_check(void);
> >> +/* dummy function, its purpose is to break the build if struct pt_regs and
> >> + * struct user_pt_regs don't match.
> >> + */
> >
> > Another trick which seems to work with GCC is:
> >
> > -void __init pt_regs_check(void)
> > +static inline void __init pt_regs_check(void)
>
> Does this really work ? Did you test to ensure that the BUILD_BUG_ON
> still detect mismatch between struct pt_regs and struct user_pt_regs ?
>

My bad, I was unaware of GCC behavior for static inline in this case.
Sorry for the noise.
Original ugly patch does work though.
>
> >
> >>   void __init pt_regs_check(void)
> >>   {
> >>          BUILD_BUG_ON(offsetof(struct pt_regs, gpr) !=
> >> --
> >> 2.19.2
> >>