Re: [PATCH v2 2/6] arm64: entry: Always inline helpers used by non-instrumentable paths

[Jisheng Zhang <jszhang@kernel.org>]

On Tue, Nov 10, 2026 at 12:05:42PM +0000, sashiko-bot@kernel.org wrote:
> Thank you for your contribution! Sashiko AI review found 1 potential
> issue to consider:
> - [Low] arm64: entry: missing __always_inline on !CONFIG_ARM64_SME stubs
>   in fpsimd.h
> --
> 
> commit ed5f3b4320afc347aa0402f3b75096047a676bff
> Author: Jisheng Zhang <jszhang@kernel.org>
> 
> arm64: entry: Always inline helpers used by non-instrumentable paths
> 
> This commit marks MTE, SME, CPU feature, and rescheduling helpers called
> from noinstr entry and exit paths as __always_inline, and updates the SME
> entry and exit helpers to use tif_test_bit(TIF_SME).
> 
> > diff --git a/arch/arm64/include/asm/fpsimd.h b/arch/arm64/include/asm/fpsimd.h
> > index a67d5774e6728..e7c881c1c1f79 100644
> > --- a/arch/arm64/include/asm/fpsimd.h
> > +++ b/arch/arm64/include/asm/fpsimd.h
> > @@ -749,17 +749,17 @@ void sme_enable_dvmsync(void);
> >  void sme_set_active(void);
> >  void sme_clear_active(void);
> >  
> > -static inline void sme_enter_from_user_mode(void)
> > +static __always_inline void sme_enter_from_user_mode(void)
> >  {
> >  	if (alternative_has_cap_unlikely(ARM64_WORKAROUND_4193714) &&
> > -	    test_thread_flag(TIF_SME))
> > +	    tif_test_bit(TIF_SME))
> >  		sme_clear_active();
> >  }
> >  
> > -static inline void sme_exit_to_user_mode(void)
> > +static __always_inline void sme_exit_to_user_mode(void)
> >  {
> >  	if (alternative_has_cap_unlikely(ARM64_WORKAROUND_4193714) &&
> > -	    test_thread_flag(TIF_SME))
> > +	    tif_test_bit(TIF_SME))
> >  		sme_set_active();
> >  }
> 
> [Severity: Low]
> Should the !CONFIG_ARM64_SME stubs for sme_enter_from_user_mode() and
> sme_exit_to_user_mode() in the #else branch below also use
> __always_inline, matching the !CONFIG_KASAN_HW_TAGS stubs updated in
> mte.h and mte-kasan.h?

I think this is false alarm because the #else branch is null function,
any compiler will inline them even when -Os, but if maintainers wants
__always_inline as well for #else branch, kindly comment I will update
them.