On Mon, Mar 25, 2019 at 08:41:54PM +0800, Guo Ren wrote:
> On Mon, Mar 25, 2019 at 03:17:54PM +0300, Dmitry V. Levin wrote:
> > On Mon, Mar 25, 2019 at 08:03:39PM +0800, guoren@xxxxxxxxxx wrote:
> > [...]
> > > diff --git a/arch/csky/include/uapi/asm/ptrace.h b/arch/csky/include/uapi/asm/ptrace.h
> > > index a4eaa8d..9bf5b1a 100644
> > > --- a/arch/csky/include/uapi/asm/ptrace.h
> > > +++ b/arch/csky/include/uapi/asm/ptrace.h
> > > @@ -62,6 +62,11 @@ struct user_fp {
> > > #define instruction_pointer(regs) ((regs)->pc)
> > > #define profile_pc(regs) instruction_pointer(regs)
> > >
> > > +static inline unsigned long regs_return_value(struct pt_regs *regs)
> > > +{
> > > + return regs->a0;
> > > +}
> > > +
> > > #endif /* __KERNEL__ */
> > > #endif /* __ASSEMBLY__ */
> > > #endif /* _CSKY_PTRACE_H */
> >
> > I wonder why we have this #ifdef __KERNEL__ code in the uapi namespace,
> > it defeats the idea of uapi. Doesn't it belong to non-uapi
> > include/asm/ptrace.h namespace?
>
> Yes, I should move __KERNEL__ codes into arch/csky/include/asm/ptrace.h.
> But it'll be another patch for the modification. Any other problems?
From UAPI perspective? No, I don't see any more UAPI issues with the patch.
--
ldv
Attachment:
signature.asc
Description: PGP signature
