diff mbox series

[v2,13/24] sh/cpu: Make sure play_dead() doesn't return

Message ID d0c3ff5349adfe8fd227acc236ae2c278a05eb4c.1676358308.git.jpoimboe@kernel.org
State New
Headers show
Series cpu,sched: Mark arch_cpu_idle_dead() __noreturn | expand

Commit Message

Josh Poimboeuf Feb. 14, 2023, 7:05 a.m. UTC
play_dead() doesn't return.  Make that more explicit with a BUG().

BUG() is preferable to unreachable() because BUG() is a more explicit
failure mode and avoids undefined behavior like falling off the edge of
the function into whatever code happens to be next.

Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
---
 arch/sh/include/asm/smp-ops.h | 1 +
 1 file changed, 1 insertion(+)

Comments

Philippe Mathieu-Daudé Feb. 14, 2023, 7:57 a.m. UTC | #1
On 14/2/23 08:05, Josh Poimboeuf wrote:
> play_dead() doesn't return.  Make that more explicit with a BUG().
> 
> BUG() is preferable to unreachable() because BUG() is a more explicit
> failure mode and avoids undefined behavior like falling off the edge of
> the function into whatever code happens to be next.
> 
> Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
> ---
>   arch/sh/include/asm/smp-ops.h | 1 +
>   1 file changed, 1 insertion(+)
> 
> diff --git a/arch/sh/include/asm/smp-ops.h b/arch/sh/include/asm/smp-ops.h
> index e27702130eb6..63866b1595a0 100644
> --- a/arch/sh/include/asm/smp-ops.h
> +++ b/arch/sh/include/asm/smp-ops.h
> @@ -27,6 +27,7 @@ static inline void plat_smp_setup(void)
>   static inline void play_dead(void)
>   {
>   	mp_ops->play_dead();
> +	BUG();
>   }

Shouldn't we decorate plat_smp_ops::play_dead() as noreturn first?
Josh Poimboeuf Feb. 14, 2023, 6:28 p.m. UTC | #2
On Tue, Feb 14, 2023 at 08:57:39AM +0100, Philippe Mathieu-Daudé wrote:
> On 14/2/23 08:05, Josh Poimboeuf wrote:
> > play_dead() doesn't return.  Make that more explicit with a BUG().
> > 
> > BUG() is preferable to unreachable() because BUG() is a more explicit
> > failure mode and avoids undefined behavior like falling off the edge of
> > the function into whatever code happens to be next.
> > 
> > Signed-off-by: Josh Poimboeuf <jpoimboe@kernel.org>
> > ---
> >   arch/sh/include/asm/smp-ops.h | 1 +
> >   1 file changed, 1 insertion(+)
> > 
> > diff --git a/arch/sh/include/asm/smp-ops.h b/arch/sh/include/asm/smp-ops.h
> > index e27702130eb6..63866b1595a0 100644
> > --- a/arch/sh/include/asm/smp-ops.h
> > +++ b/arch/sh/include/asm/smp-ops.h
> > @@ -27,6 +27,7 @@ static inline void plat_smp_setup(void)
> >   static inline void play_dead(void)
> >   {
> >   	mp_ops->play_dead();
> > +	BUG();
> >   }
> 
> Shouldn't we decorate plat_smp_ops::play_dead() as noreturn first?

I guess it really depends on how far we want to go down the __noreturn
rabbit hole.  To keep the patch set constrained yet still useful I
stopped when I got to a function pointer, as I think it still needs a
BUG() afterwards either way.

That said, there would still be benefits of adding __noreturn to
function pointers, I just wanted to keep the patch set down to a
manageable size ;-)
diff mbox series

Patch

diff --git a/arch/sh/include/asm/smp-ops.h b/arch/sh/include/asm/smp-ops.h
index e27702130eb6..63866b1595a0 100644
--- a/arch/sh/include/asm/smp-ops.h
+++ b/arch/sh/include/asm/smp-ops.h
@@ -27,6 +27,7 @@  static inline void plat_smp_setup(void)
 static inline void play_dead(void)
 {
 	mp_ops->play_dead();
+	BUG();
 }
 
 extern void register_smp_ops(struct plat_smp_ops *ops);