linux-kernel - Re: [PATCH] ftrace: Get the true parent ip for function tracer

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <ef273629d92e108ad166faf1f4abccac6cda461a@linux.dev>
Date: Wed, 21 Aug 2024 14:16:12 +0000
From: jeff.xie@...ux.dev
To: "Steven Rostedt" <rostedt@...dmis.org>
Cc: mhiramat@...nel.org, mathieu.desnoyers@...icios.com,
 linux-trace-kernel@...r.kernel.org, linux-kernel@...r.kernel.org,
 xiehuan09@...il.com
Subject: Re: [PATCH] ftrace: Get the true parent ip for function tracer

August 21, 2024 at 9:59 PM, "Steven Rostedt" <rostedt@...dmis.org> wrote:


Hi Steven,


> 
> On Wed, 21 Aug 2024 21:27:55 +0800
> 
> Jeff Xie <jeff.xie@...ux.dev> wrote:
> 
> Hi Jeff,
> 
> Thanks for the patch.
> 
> > 
> > Signed-off-by: Jeff Xie <jeff.xie@...ux.dev>
> > 
> >  ---
> > 
> >  kernel/trace/trace_functions.c | 18 ++++++++++++++++++
> > 
> >  1 file changed, 18 insertions(+)
> > 
> >  
> > 
> >  diff --git a/kernel/trace/trace_functions.c b/kernel/trace/trace_functions.c
> > 
> >  index 3b0cea37e029..273b8c7eeb2d 100644
> > 
> >  --- a/kernel/trace/trace_functions.c
> > 
> >  +++ b/kernel/trace/trace_functions.c
> > 
> >  @@ -176,6 +176,19 @@ static void function_trace_start(struct trace_array *tr)
> > 
> >  tracing_reset_online_cpus(&tr->array_buffer);
> > 
> >  }
> > 
> >  
> > 
> >  +static unsigned long
> > 
> >  +function_get_true_parent_ip(unsigned long parent_ip, struct ftrace_regs *fregs)
> > 
> 
> I wonder if we should make this inline, or even __always_inline, as this
> 
> will be called in a very hot path, and I want to make sure that the
> 
> compiler always inlines it. It likely should, but we should also give the
> 
> compiler a hint that it should.


Yes, Thank you for your suggestion. It is indeed a very hot path, 
and I will add __always_inline in the next version.


> 
> -- Steve
> 
> > 
> > +{
> > 
> >  + unsigned long true_parent_ip;
> > 
> >  + int idx = 0;
> > 
> >  +
> > 
> >  + true_parent_ip = parent_ip;
> > 
> >  + if (unlikely(parent_ip == (unsigned long)&return_to_handler))
> > 
> >  + true_parent_ip = ftrace_graph_ret_addr(current, &idx, parent_ip,
> > 
> >  + (unsigned long *)fregs->regs.sp);
> > 
> >  + return true_parent_ip;
> > 
> >  +}
> > 
> >  +
> > 
> >  static void
> > 
> >  function_trace_call(unsigned long ip, unsigned long parent_ip,
> > 
> >  struct ftrace_ops *op, struct ftrace_regs *fregs)
> > 
> >  @@ -193,6 +206,8 @@ function_trace_call(unsigned long ip, unsigned long parent_ip,
> > 
> >  if (bit < 0)
> > 
> >  return;
> > 
> >  
> > 
> >  + parent_ip = function_get_true_parent_ip(parent_ip, fregs);
> > 
> >  +
> > 
> >  trace_ctx = tracing_gen_ctx();
> > 
> >  
> > 
> >  cpu = smp_processor_id();
> > 
> >  @@ -241,6 +256,7 @@ function_stack_trace_call(unsigned long ip, unsigned long parent_ip,
> > 
> >  * recursive protection is performed.
> > 
> >  */
> > 
> >  local_irq_save(flags);
> > 
> >  + parent_ip = function_get_true_parent_ip(parent_ip, fregs);
> > 
> >  cpu = raw_smp_processor_id();
> > 
> >  data = per_cpu_ptr(tr->array_buffer.data, cpu);
> > 
> >  disabled = atomic_inc_return(&data->disabled);
> > 
> >  @@ -309,6 +325,7 @@ function_no_repeats_trace_call(unsigned long ip, unsigned long parent_ip,
> > 
> >  if (bit < 0)
> > 
> >  return;
> > 
> >  
> > 
> >  + parent_ip = function_get_true_parent_ip(parent_ip, fregs);
> > 
> >  cpu = smp_processor_id();
> > 
> >  data = per_cpu_ptr(tr->array_buffer.data, cpu);
> > 
> >  if (atomic_read(&data->disabled))
> > 
> >  @@ -356,6 +373,7 @@ function_stack_no_repeats_trace_call(unsigned long ip, unsigned long parent_ip,
> > 
> >  * recursive protection is performed.
> > 
> >  */
> > 
> >  local_irq_save(flags);
> > 
> >  + parent_ip = function_get_true_parent_ip(parent_ip, fregs);
> > 
> >  cpu = raw_smp_processor_id();
> > 
> >  data = per_cpu_ptr(tr->array_buffer.data, cpu);
> > 
> >  disabled = atomic_inc_return(&data->disabled);
> >
>