2024-03-20 02:58:28

by Kees Cook

[permalink] [raw]
Subject: Re: [PATCH] ARM: unwind: improve unwinders for noreturn case

On Wed, Mar 20, 2024 at 10:19:29AM +0800, Jiangfeng Xiao wrote:
> This is an off-by-one bug which is common in unwinders,
> due to the fact that the address on the stack points
> to the return address rather than the call address.
>
> So, for example, when the last instruction of a function
> is a function call (e.g., to a noreturn function), it can
> cause the unwinder to incorrectly try to unwind from
> the function after the callee.
>
> foo:
> ...
> bl bar
> ... end of function and thus next function ...
>
> which results in LR pointing into the next function.
>
> Fixed this by subtracting 1 from frmae->pc in the call frame
> (but not exception frames) like ORC on x86 does.
>
> Refer to the unwind_next_frame function in the unwind_orc.c
>
> Suggested-by: Josh Poimboeuf <[email protected]>
> Link: https://lkml.kernel.org/lkml/20240305175846.qnyiru7uaa7itqba@treble/
> Signed-off-by: Jiangfeng Xiao <[email protected]>
> ---
> arch/arm/include/asm/stacktrace.h | 4 ----
> arch/arm/kernel/stacktrace.c | 2 --
> arch/arm/kernel/traps.c | 4 ++--
> arch/arm/kernel/unwind.c | 18 +++++++++++++++---
> 4 files changed, 17 insertions(+), 11 deletions(-)
>
> diff --git a/arch/arm/include/asm/stacktrace.h b/arch/arm/include/asm/stacktrace.h
> index 360f0d2..07e4c16 100644
> --- a/arch/arm/include/asm/stacktrace.h
> +++ b/arch/arm/include/asm/stacktrace.h
> @@ -21,9 +21,7 @@ struct stackframe {
> struct llist_node *kr_cur;
> struct task_struct *tsk;
> #endif
> -#ifdef CONFIG_UNWINDER_FRAME_POINTER
> bool ex_frame;
> -#endif
> };
>
> static __always_inline
> @@ -37,9 +35,7 @@ void arm_get_current_stackframe(struct pt_regs *regs, struct stackframe *frame)
> frame->kr_cur = NULL;
> frame->tsk = current;
> #endif
> -#ifdef CONFIG_UNWINDER_FRAME_POINTER
> frame->ex_frame = in_entry_text(frame->pc);
> -#endif
> }
>
> extern int unwind_frame(struct stackframe *frame);
> diff --git a/arch/arm/kernel/stacktrace.c b/arch/arm/kernel/stacktrace.c
> index 620aa82..1abd4f9 100644
> --- a/arch/arm/kernel/stacktrace.c
> +++ b/arch/arm/kernel/stacktrace.c
> @@ -154,9 +154,7 @@ static void start_stack_trace(struct stackframe *frame, struct task_struct *task
> frame->kr_cur = NULL;
> frame->tsk = task;
> #endif
> -#ifdef CONFIG_UNWINDER_FRAME_POINTER
> frame->ex_frame = in_entry_text(frame->pc);
> -#endif
> }
>
> void arch_stack_walk(stack_trace_consume_fn consume_entry, void *cookie,
> diff --git a/arch/arm/kernel/traps.c b/arch/arm/kernel/traps.c
> index 3bad79d..b64e442 100644
> --- a/arch/arm/kernel/traps.c
> +++ b/arch/arm/kernel/traps.c
> @@ -84,10 +84,10 @@ void dump_backtrace_entry(unsigned long where, unsigned long from,
> printk("%sFunction entered at [<%08lx>] from [<%08lx>]\n",
> loglvl, where, from);
> #elif defined CONFIG_BACKTRACE_VERBOSE
> - printk("%s[<%08lx>] (%ps) from [<%08lx>] (%pS)\n",
> + pr_warn("%s[<%08lx>] (%ps) from [<%08lx>] (%pB)\n",
> loglvl, where, (void *)where, from, (void *)from);

This should stay printk("%s...", loglvl, ...) or loglvl should be
dropped when converting to pr_warn():

pr_warn([<%08lx>] (%ps) from [<%08lx>] (%pB)\n",
where, (void *)where, from, (void *)from);

Why did you want to force the "warn" log level?

> #else
> - printk("%s %ps from %pS\n", loglvl, (void *)where, (void *)from);
> + pr_warn("%s %ps from %pB\n", loglvl, (void *)where, (void *)from);

Ditto.

-Kees

--
Kees Cook


2024-03-20 03:30:43

by Jiangfeng Xiao

[permalink] [raw]
Subject: Re: [PATCH] ARM: unwind: improve unwinders for noreturn case



On 2024/3/20 10:46, Kees Cook wrote:
> On Wed, Mar 20, 2024 at 10:19:29AM +0800, Jiangfeng Xiao wrote:
>> This is an off-by-one bug which is common in unwinders,
>> due to the fact that the address on the stack points
>> to the return address rather than the call address.
>>
>> So, for example, when the last instruction of a function
>> is a function call (e.g., to a noreturn function), it can
>> cause the unwinder to incorrectly try to unwind from
>> the function after the callee.
>>
>> foo:
>> ...
>> bl bar
>> ... end of function and thus next function ...
>>
>> which results in LR pointing into the next function.
>>
>> Fixed this by subtracting 1 from frmae->pc in the call frame
>> (but not exception frames) like ORC on x86 does.
>>
>> Refer to the unwind_next_frame function in the unwind_orc.c
>>
>> Suggested-by: Josh Poimboeuf <[email protected]>
>> Link: https://lkml.kernel.org/lkml/20240305175846.qnyiru7uaa7itqba@treble/
>> Signed-off-by: Jiangfeng Xiao <[email protected]>
>> ---
>> arch/arm/include/asm/stacktrace.h | 4 ----
>> arch/arm/kernel/stacktrace.c | 2 --
>> arch/arm/kernel/traps.c | 4 ++--
>> arch/arm/kernel/unwind.c | 18 +++++++++++++++---
>> 4 files changed, 17 insertions(+), 11 deletions(-)
>>
>> diff --git a/arch/arm/include/asm/stacktrace.h b/arch/arm/include/asm/stacktrace.h
>> index 360f0d2..07e4c16 100644
>> --- a/arch/arm/include/asm/stacktrace.h
>> +++ b/arch/arm/include/asm/stacktrace.h
>> @@ -21,9 +21,7 @@ struct stackframe {
>> struct llist_node *kr_cur;
>> struct task_struct *tsk;
>> #endif
>> -#ifdef CONFIG_UNWINDER_FRAME_POINTER
>> bool ex_frame;
>> -#endif
>> };
>>
>> static __always_inline
>> @@ -37,9 +35,7 @@ void arm_get_current_stackframe(struct pt_regs *regs, struct stackframe *frame)
>> frame->kr_cur = NULL;
>> frame->tsk = current;
>> #endif
>> -#ifdef CONFIG_UNWINDER_FRAME_POINTER
>> frame->ex_frame = in_entry_text(frame->pc);
>> -#endif
>> }
>>
>> extern int unwind_frame(struct stackframe *frame);
>> diff --git a/arch/arm/kernel/stacktrace.c b/arch/arm/kernel/stacktrace.c
>> index 620aa82..1abd4f9 100644
>> --- a/arch/arm/kernel/stacktrace.c
>> +++ b/arch/arm/kernel/stacktrace.c
>> @@ -154,9 +154,7 @@ static void start_stack_trace(struct stackframe *frame, struct task_struct *task
>> frame->kr_cur = NULL;
>> frame->tsk = task;
>> #endif
>> -#ifdef CONFIG_UNWINDER_FRAME_POINTER
>> frame->ex_frame = in_entry_text(frame->pc);
>> -#endif
>> }
>>
>> void arch_stack_walk(stack_trace_consume_fn consume_entry, void *cookie,
>> diff --git a/arch/arm/kernel/traps.c b/arch/arm/kernel/traps.c
>> index 3bad79d..b64e442 100644
>> --- a/arch/arm/kernel/traps.c
>> +++ b/arch/arm/kernel/traps.c
>> @@ -84,10 +84,10 @@ void dump_backtrace_entry(unsigned long where, unsigned long from,
>> printk("%sFunction entered at [<%08lx>] from [<%08lx>]\n",
>> loglvl, where, from);
>> #elif defined CONFIG_BACKTRACE_VERBOSE
>> - printk("%s[<%08lx>] (%ps) from [<%08lx>] (%pS)\n",
>> + pr_warn("%s[<%08lx>] (%ps) from [<%08lx>] (%pB)\n",
>> loglvl, where, (void *)where, from, (void *)from);
>
> This should stay printk("%s...", loglvl, ...) or loglvl should be
> dropped when converting to pr_warn():
>
> pr_warn([<%08lx>] (%ps) from [<%08lx>] (%pB)\n",
> where, (void *)where, from, (void *)from);
>
> Why did you want to force the "warn" log level?
>

Thank you for your review. I think I'm wrong.

The checkpatch.pl script reports the "WARNING: printk() should
include KERN_<LEVEL> facility level" warning.

That's why I changed printk to pr_warn.
I should change printk to printk(KERN_DEFAULT).

>> #else
>> - printk("%s %ps from %pS\n", loglvl, (void *)where, (void *)from);
>> + pr_warn("%s %ps from %pB\n", loglvl, (void *)where, (void *)from);
>
> Ditto.
>
> -Kees
>

2024-03-20 03:34:36

by Matthew Wilcox

[permalink] [raw]
Subject: Re: [PATCH] ARM: unwind: improve unwinders for noreturn case

On Wed, Mar 20, 2024 at 11:30:13AM +0800, Jiangfeng Xiao wrote:
> The checkpatch.pl script reports the "WARNING: printk() should
> include KERN_<LEVEL> facility level" warning.
>
> That's why I changed printk to pr_warn.
> I should change printk to printk(KERN_DEFAULT).

No, you should ignore checkpatch. For bonus points, figure out why you
should ignore it specifically in this case.

2024-03-20 03:46:46

by Jiangfeng Xiao

[permalink] [raw]
Subject: Re: [PATCH] ARM: unwind: improve unwinders for noreturn case



On 2024/3/20 11:34, Matthew Wilcox wrote:
> On Wed, Mar 20, 2024 at 11:30:13AM +0800, Jiangfeng Xiao wrote:
>> The checkpatch.pl script reports the "WARNING: printk() should
>> include KERN_<LEVEL> facility level" warning.
>>
>> That's why I changed printk to pr_warn.
>> I should change printk to printk(KERN_DEFAULT).
>
> No, you should ignore checkpatch. For bonus points, figure out why you
> should ignore it specifically in this case.
> .
>

Thank you. I think I understand.

The checkpatch.pl file is a false report
because the 'loglvl' already exists.

I'd better keep printk(" %s...", loglvl, ...) still.