arm64: kprobes: fix XOL preemption window

[PATCH v2 2/2] arm64: insn: drop NOP from steppable hint list

Posted by Khaja Hussain Shaik Khaji 1 month, 2 weeks ago

NOP is already handled via instruction emulation and does not require
single-stepping. Drop it from aarch64_insn_is_steppable_hint().

Signed-off-by: Khaja Hussain Shaik Khaji <khaja.khaji@oss.qualcomm.com>
---
 arch/arm64/include/asm/insn.h | 1 -
 1 file changed, 1 deletion(-)

diff --git a/arch/arm64/include/asm/insn.h b/arch/arm64/include/asm/insn.h
index e1d30ba99d01..9429f76906e0 100644
--- a/arch/arm64/include/asm/insn.h
+++ b/arch/arm64/include/asm/insn.h
@@ -456,7 +456,6 @@ static __always_inline bool aarch64_insn_is_steppable_hint(u32 insn)
 	case AARCH64_INSN_HINT_BTIC:
 	case AARCH64_INSN_HINT_BTIJ:
 	case AARCH64_INSN_HINT_BTIJC:
-	case AARCH64_INSN_HINT_NOP:
 		return true;
 	default:
 		return false;
-- 
2.34.1

Re: [PATCH v2 2/2] arm64: insn: drop NOP from steppable hint list

Posted by Mark Rutland 1 month, 2 weeks ago

On Tue, Feb 17, 2026 at 07:08:55PM +0530, Khaja Hussain Shaik Khaji wrote:
> NOP is already handled via instruction emulation and does not require
> single-stepping. Drop it from aarch64_insn_is_steppable_hint().
> 
> Signed-off-by: Khaja Hussain Shaik Khaji <khaja.khaji@oss.qualcomm.com>
> ---
>  arch/arm64/include/asm/insn.h | 1 -
>  1 file changed, 1 deletion(-)
> 
> diff --git a/arch/arm64/include/asm/insn.h b/arch/arm64/include/asm/insn.h
> index e1d30ba99d01..9429f76906e0 100644
> --- a/arch/arm64/include/asm/insn.h
> +++ b/arch/arm64/include/asm/insn.h
> @@ -456,7 +456,6 @@ static __always_inline bool aarch64_insn_is_steppable_hint(u32 insn)
>  	case AARCH64_INSN_HINT_BTIC:
>  	case AARCH64_INSN_HINT_BTIJ:
>  	case AARCH64_INSN_HINT_BTIJC:
> -	case AARCH64_INSN_HINT_NOP:
>  		return true;
>  	default:
>  		return false;

The intent is that aarch64_insn_is_steppable_hint() says whether an
instruction is safe to step, not whether it *must* be stepped. I think
we can leave NOP here unless this is causing some functional problem?

Mark.

Re: [PATCH v2 2/2] arm64: insn: drop NOP from steppable hint list

Posted by Masami Hiramatsu (Google) 1 month, 1 week ago

On Tue, 17 Feb 2026 16:57:08 +0000
Mark Rutland <mark.rutland@arm.com> wrote:

> On Tue, Feb 17, 2026 at 07:08:55PM +0530, Khaja Hussain Shaik Khaji wrote:
> > NOP is already handled via instruction emulation and does not require
> > single-stepping. Drop it from aarch64_insn_is_steppable_hint().
> > 
> > Signed-off-by: Khaja Hussain Shaik Khaji <khaja.khaji@oss.qualcomm.com>
> > ---
> >  arch/arm64/include/asm/insn.h | 1 -
> >  1 file changed, 1 deletion(-)
> > 
> > diff --git a/arch/arm64/include/asm/insn.h b/arch/arm64/include/asm/insn.h
> > index e1d30ba99d01..9429f76906e0 100644
> > --- a/arch/arm64/include/asm/insn.h
> > +++ b/arch/arm64/include/asm/insn.h
> > @@ -456,7 +456,6 @@ static __always_inline bool aarch64_insn_is_steppable_hint(u32 insn)
> >  	case AARCH64_INSN_HINT_BTIC:
> >  	case AARCH64_INSN_HINT_BTIJ:
> >  	case AARCH64_INSN_HINT_BTIJC:
> > -	case AARCH64_INSN_HINT_NOP:
> >  		return true;
> >  	default:
> >  		return false;
> 
> The intent is that aarch64_insn_is_steppable_hint() says whether an
> instruction is safe to step, not whether it *must* be stepped. I think
> we can leave NOP here unless this is causing some functional problem?

Agreed. I think we should keep this as it is.

Thank you,

> 
> Mark.
> 


-- 
Masami Hiramatsu (Google) <mhiramat@kernel.org>