Re: [PATCH v2 2/3] kprobes: adjust kprobe addr for KPROBES_ON_FTRACE

[Naveen N. Rao]

Jisheng Zhang wrote:
> For KPROBES_ON_FTRACE case, we need to adjust the kprobe's addr
> correspondingly.
>
> Signed-off-by: Jisheng Zhang <Jisheng.Zhang@xxxxxxxxxxxxx>
> ---
>  kernel/kprobes.c | 10 +++++++---
>  1 file changed, 7 insertions(+), 3 deletions(-)
>
> diff --git a/kernel/kprobes.c b/kernel/kprobes.c
> index 9873fc627d61..3fd2f68644da 100644
> --- a/kernel/kprobes.c
> +++ b/kernel/kprobes.c
> @@ -1484,15 +1484,19 @@ static inline int check_kprobe_rereg(struct kprobe *p)
>  
>  int __weak arch_check_ftrace_location(struct kprobe *p)
>  {
> -	unsigned long ftrace_addr;
> +	unsigned long ftrace_addr, addr = (unsigned long)p->addr;
>  
> -	ftrace_addr = ftrace_location((unsigned long)p->addr);
> +#ifdef CONFIG_KPROBES_ON_FTRACE
> +	addr = ftrace_call_adjust(addr);
> +#endif

Looking at the commit message for patch 3/3, it looks like you want the 
probe to be placed on ftrace entry by default, and this patch seems to 
be aimed at that.

If so, this is not the right approach. As I mentioned previously, you 
would want to over-ride kprobe_lookup_name(). This ensures that the 
address is changed only if the user provided a symbol, and not if the 
user wanted to probe at a very specific address. See commit 
24bd909e94776 ("powerpc/kprobes: Prefer ftrace when probing function 
entry").

If this patch is for some other purpose, then it isn't clear from the 
commit log. Please provide a better explanation.


- Naveen