[PATCH 6.12 118/265] LoongArch/orc: Use RCU in all users of __module_address().

[PATCH 6.12 118/265] LoongArch/orc: Use RCU in all users of __module_address().

[Greg Kroah-Hartman]

6.12-stable review patch.  If anyone has any objections, please let me know.

------------------

From: Sebastian Andrzej Siewior <bigeasy@linutronix.de>

[ Upstream commit f99d27d9feb755aee9350fc89f57814d7e1b4880 ]

__module_address() can be invoked within a RCU section, there is no
requirement to have preemption disabled.

Replace the preempt_disable() section around __module_address() with
RCU.

Cc: Huacai Chen <chenhuacai@kernel.org>
Cc: WANG Xuerui <kernel@xen0n.name>
Cc: loongarch@lists.linux.dev
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250108090457.512198-19-bigeasy@linutronix.de
Signed-off-by: Petr Pavlu <petr.pavlu@suse.com>
Stable-dep-of: 055c7e75190e ("LoongArch: Handle percpu handler address for ORC unwinder")
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 arch/loongarch/kernel/unwind_orc.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/arch/loongarch/kernel/unwind_orc.c b/arch/loongarch/kernel/unwind_orc.c
index 471652c0c8653..59809c3406c03 100644
--- a/arch/loongarch/kernel/unwind_orc.c
+++ b/arch/loongarch/kernel/unwind_orc.c
@@ -399,7 +399,7 @@ bool unwind_next_frame(struct unwind_state *state)
 		return false;
 
 	/* Don't let modules unload while we're reading their ORC data. */
-	preempt_disable();
+	guard(rcu)();
 
 	if (is_entry_func(state->pc))
 		goto end;
@@ -514,14 +514,12 @@ bool unwind_next_frame(struct unwind_state *state)
 	if (!__kernel_text_address(state->pc))
 		goto err;
 
-	preempt_enable();
 	return true;
 
 err:
 	state->error = true;
 
 end:
-	preempt_enable();
 	state->stack_info.type = STACK_TYPE_UNKNOWN;
 	return false;
 }
-- 
2.51.0

Re: [PATCH 6.12 118/265] LoongArch/orc: Use RCU in all users of __module_address().

[Harshit Mogalapalli]

Hi all,

On 13/03/26 01:38, Greg Kroah-Hartman wrote:
> 6.12-stable review patch.  If anyone has any objections, please let me know.
> 
> ------------------
> 
> From: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
> 
> [ Upstream commit f99d27d9feb755aee9350fc89f57814d7e1b4880 ]
> 
> __module_address() can be invoked within a RCU section, there is no
> requirement to have preemption disabled.
> 
> Replace the preempt_disable() section around __module_address() with
> RCU.
> 
> Cc: Huacai Chen <chenhuacai@kernel.org>
> Cc: WANG Xuerui <kernel@xen0n.name>
> Cc: loongarch@lists.linux.dev
> Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
> Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
> Link: https://lore.kernel.org/r/20250108090457.512198-19-bigeasy@linutronix.de
> Signed-off-by: Petr Pavlu <petr.pavlu@suse.com>
> Stable-dep-of: 055c7e75190e ("LoongArch: Handle percpu handler address for ORC unwinder")
> Signed-off-by: Sasha Levin <sashal@kernel.org>
> ---
>   arch/loongarch/kernel/unwind_orc.c | 4 +---
>   1 file changed, 1 insertion(+), 3 deletions(-)
> 
> diff --git a/arch/loongarch/kernel/unwind_orc.c b/arch/loongarch/kernel/unwind_orc.c
> index 471652c0c8653..59809c3406c03 100644
> --- a/arch/loongarch/kernel/unwind_orc.c
> +++ b/arch/loongarch/kernel/unwind_orc.c
> @@ -399,7 +399,7 @@ bool unwind_next_frame(struct unwind_state *state)
>   		return false;
>   
>   	/* Don't let modules unload while we're reading their ORC data. */
> -	preempt_disable();
> +	guard(rcu)();
>   
>   	if (is_entry_func(state->pc))
>   		goto end;
> @@ -514,14 +514,12 @@ bool unwind_next_frame(struct unwind_state *state)
>   	if (!__kernel_text_address(state->pc))
>   		goto err;
>   
> -	preempt_enable();
>   	return true;
>  


Looks like this is dependent on commit: 7d9dda6f628f ("module: Allow 
__module_address() to be called from RCU section."), so I feel pulling 
in this patch without the mentioned missing prerequisite is wrong. Can 
you please help review this ?

This is also part of a feature series in 
https://lore.kernel.org/all/20250108090457.512198-13-bigeasy@linutronix.de/

Thanks,
Harshit

>   err:
>   	state->error = true;
>   
>   end:
> -	preempt_enable();
>   	state->stack_info.type = STACK_TYPE_UNKNOWN;
>   	return false;
>   }

Re: [PATCH 6.12 118/265] LoongArch/orc: Use RCU in all users of __module_address().

[Huacai Chen]

On Mon, Apr 6, 2026 at 8:28 PM Harshit Mogalapalli
<harshit.m.mogalapalli@oracle.com> wrote:
>
> Hi all,
>
> On 13/03/26 01:38, Greg Kroah-Hartman wrote:
> > 6.12-stable review patch.  If anyone has any objections, please let me know.
> >
> > ------------------
> >
> > From: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
> >
> > [ Upstream commit f99d27d9feb755aee9350fc89f57814d7e1b4880 ]
> >
> > __module_address() can be invoked within a RCU section, there is no
> > requirement to have preemption disabled.
> >
> > Replace the preempt_disable() section around __module_address() with
> > RCU.
> >
> > Cc: Huacai Chen <chenhuacai@kernel.org>
> > Cc: WANG Xuerui <kernel@xen0n.name>
> > Cc: loongarch@lists.linux.dev
> > Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
> > Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
> > Link: https://lore.kernel.org/r/20250108090457.512198-19-bigeasy@linutronix.de
> > Signed-off-by: Petr Pavlu <petr.pavlu@suse.com>
> > Stable-dep-of: 055c7e75190e ("LoongArch: Handle percpu handler address for ORC unwinder")
> > Signed-off-by: Sasha Levin <sashal@kernel.org>
> > ---
> >   arch/loongarch/kernel/unwind_orc.c | 4 +---
> >   1 file changed, 1 insertion(+), 3 deletions(-)
> >
> > diff --git a/arch/loongarch/kernel/unwind_orc.c b/arch/loongarch/kernel/unwind_orc.c
> > index 471652c0c8653..59809c3406c03 100644
> > --- a/arch/loongarch/kernel/unwind_orc.c
> > +++ b/arch/loongarch/kernel/unwind_orc.c
> > @@ -399,7 +399,7 @@ bool unwind_next_frame(struct unwind_state *state)
> >               return false;
> >
> >       /* Don't let modules unload while we're reading their ORC data. */
> > -     preempt_disable();
> > +     guard(rcu)();
> >
> >       if (is_entry_func(state->pc))
> >               goto end;
> > @@ -514,14 +514,12 @@ bool unwind_next_frame(struct unwind_state *state)
> >       if (!__kernel_text_address(state->pc))
> >               goto err;
> >
> > -     preempt_enable();
> >       return true;
> >
>
>
> Looks like this is dependent on commit: 7d9dda6f628f ("module: Allow
> __module_address() to be called from RCU section."), so I feel pulling
> in this patch without the mentioned missing prerequisite is wrong. Can
> you please help review this ?
>
> This is also part of a feature series in
> https://lore.kernel.org/all/20250108090457.512198-13-bigeasy@linutronix.de/
Yes, I have also asked about this:
https://lore.kernel.org/stable/CAAhV-H7GxtWRZyAT=kedLEMu=C5wH--NUzRjwi3DKXzUq+QZjA@mail.gmail.com/

However, no answer and no action for this. I don't know what happened
to Greg and Sasha. Recently, similar accidents happen again and again.

Huacai

>
> Thanks,
> Harshit
>
> >   err:
> >       state->error = true;
> >
> >   end:
> > -     preempt_enable();
> >       state->stack_info.type = STACK_TYPE_UNKNOWN;
> >       return false;
> >   }
>