2009-04-02 06:20:47

by Lai Jiangshan

[permalink] [raw]
Subject: [PATCH] tracing: mark get_parent_ip() notrace


Impact: remove redundancy tracing log

When tracer is "function" or "function_graph", too much
"get_parent_ip" are recorded in ring_buffer.

Signed-off-by: Lai Jiangshan <[email protected]>
---
diff --git a/kernel/sched.c b/kernel/sched.c
index 706517c..a0ec10c 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -4860,7 +4860,7 @@ void scheduler_tick(void)
#endif
}

-unsigned long get_parent_ip(unsigned long addr)
+notrace unsigned long get_parent_ip(unsigned long addr)
{
if (in_lock_functions(addr)) {
addr = CALLER_ADDR2;


2009-04-02 12:26:28

by Frederic Weisbecker

[permalink] [raw]
Subject: Re: [PATCH] tracing: mark get_parent_ip() notrace

On Thu, Apr 02, 2009 at 02:18:25PM +0800, Lai Jiangshan wrote:
>
> Impact: remove redundancy tracing log
>
> When tracer is "function" or "function_graph", too much
> "get_parent_ip" are recorded in ring_buffer.


Thanks, indeed this get_parent_ip was a bit harrassing
in the traces, though I still don't understand why it is
called so often.

Anyway, tracing it seems to me not useful.

Acked-by: Frederic Weisbecker <[email protected]>



> Signed-off-by: Lai Jiangshan <[email protected]>
> ---
> diff --git a/kernel/sched.c b/kernel/sched.c
> index 706517c..a0ec10c 100644
> --- a/kernel/sched.c
> +++ b/kernel/sched.c
> @@ -4860,7 +4860,7 @@ void scheduler_tick(void)
> #endif
> }
>
> -unsigned long get_parent_ip(unsigned long addr)
> +notrace unsigned long get_parent_ip(unsigned long addr)
> {
> if (in_lock_functions(addr)) {
> addr = CALLER_ADDR2;
>
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/

2009-04-13 06:18:53

by Lai Jiangshan

[permalink] [raw]
Subject: Re: [PATCH] tracing: mark get_parent_ip() notrace

Frederic Weisbecker wrote:
> On Thu, Apr 02, 2009 at 02:18:25PM +0800, Lai Jiangshan wrote:
>> Impact: remove redundancy tracing log
>>
>> When tracer is "function" or "function_graph", too much
>> "get_parent_ip" are recorded in ring_buffer.
>
>
> Thanks, indeed this get_parent_ip was a bit harrassing
> in the traces, though I still don't understand why it is
> called so often.
>
> Anyway, tracing it seems to me not useful.
>
> Acked-by: Frederic Weisbecker <[email protected]>
>
>

Hi, Ingo,
Could you apply this fix?

get_parent_ip() is a helper function for ftace, and it is called
by preempt_disable() and preempt_enable(), so it is occurred very
often.

Lai.


2009-04-13 14:44:00

by Steven Rostedt

[permalink] [raw]
Subject: Re: [PATCH] tracing: mark get_parent_ip() notrace


On Mon, 2009-04-13 at 14:17 +0800, Lai Jiangshan wrote:
> Frederic Weisbecker wrote:
> > On Thu, Apr 02, 2009 at 02:18:25PM +0800, Lai Jiangshan wrote:
> >> Impact: remove redundancy tracing log
> >>
> >> When tracer is "function" or "function_graph", too much
> >> "get_parent_ip" are recorded in ring_buffer.
> >
> >
> > Thanks, indeed this get_parent_ip was a bit harrassing
> > in the traces, though I still don't understand why it is
> > called so often.
> >
> > Anyway, tracing it seems to me not useful.
> >
> > Acked-by: Frederic Weisbecker <[email protected]>
> >
> >
>
> Hi, Ingo,
> Could you apply this fix?
>
> get_parent_ip() is a helper function for ftace, and it is called
> by preempt_disable() and preempt_enable(), so it is occurred very
> often.

I can pull it in too, if Ingo is busy.

-- Steve

2009-04-14 00:12:25

by Ingo Molnar

[permalink] [raw]
Subject: Re: [PATCH] tracing: mark get_parent_ip() notrace


* Steven Rostedt <[email protected]> wrote:

>
> On Mon, 2009-04-13 at 14:17 +0800, Lai Jiangshan wrote:
> > Frederic Weisbecker wrote:
> > > On Thu, Apr 02, 2009 at 02:18:25PM +0800, Lai Jiangshan wrote:
> > >> Impact: remove redundancy tracing log
> > >>
> > >> When tracer is "function" or "function_graph", too much
> > >> "get_parent_ip" are recorded in ring_buffer.
> > >
> > >
> > > Thanks, indeed this get_parent_ip was a bit harrassing
> > > in the traces, though I still don't understand why it is
> > > called so often.
> > >
> > > Anyway, tracing it seems to me not useful.
> > >
> > > Acked-by: Frederic Weisbecker <[email protected]>
> > >
> > >
> >
> > Hi, Ingo,
> > Could you apply this fix?
> >
> > get_parent_ip() is a helper function for ftace, and it is called
> > by preempt_disable() and preempt_enable(), so it is occurred very
> > often.
>
> I can pull it in too, if Ingo is busy.

I have put it into tip:sched/urgent - it belongs there file-wise.
Will go upstream after -rc2 most likely.

Ingo

2009-04-14 00:13:56

by Lai Jiangshan

[permalink] [raw]
Subject: [tip:sched/urgent] tracing, sched: mark get_parent_ip() notrace

Commit-ID: 132380a06b24704fd6c9be55c44d4ef3972cead2
Gitweb: http://git.kernel.org/tip/132380a06b24704fd6c9be55c44d4ef3972cead2
Author: Lai Jiangshan <[email protected]>
AuthorDate: Thu, 2 Apr 2009 14:18:25 +0800
Committer: Ingo Molnar <[email protected]>
CommitDate: Tue, 14 Apr 2009 02:11:05 +0200

tracing, sched: mark get_parent_ip() notrace

Impact: remove overly redundant tracing entries

When tracer is "function" or "function_graph", way too much
"get_parent_ip" entries are recorded in ring_buffer.

Signed-off-by: Lai Jiangshan <[email protected]>
Acked-by: Frederic Weisbecker <[email protected]>
Acked-by: Steven Rostedt <[email protected]>
LKML-Reference: <[email protected]>
Signed-off-by: Ingo Molnar <[email protected]>


---
kernel/sched.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/kernel/sched.c b/kernel/sched.c
index 5724508..e90e70e 100644
--- a/kernel/sched.c
+++ b/kernel/sched.c
@@ -4846,7 +4846,7 @@ void scheduler_tick(void)
#endif
}

-unsigned long get_parent_ip(unsigned long addr)
+notrace unsigned long get_parent_ip(unsigned long addr)
{
if (in_lock_functions(addr)) {
addr = CALLER_ADDR2;