Re: [PATCH 2/2] ftrace: Unify effect of writing to trace_optionsand option/*

[Frederic Weisbecker]

On Thu, Aug 06, 2009 at 11:27:56AM +0800, Zhaolei wrote:
> "echo noglobal-clock > trace_options" can be used to change trace
> clock but "echo 0 > options/global-clock" can't.
> 
> We can fix it by using set_tracer_flags() in trace_options_core_write().
> 
> Signed-off-by: Zhao Lei <zhaolei@xxxxxxxxxxxxxx>
> ---
>  kernel/trace/trace.c |    5 ++---
>  1 files changed, 2 insertions(+), 3 deletions(-)
> 
> diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c
> index 464b7bb..2fa2bac 100644
> --- a/kernel/trace/trace.c
> +++ b/kernel/trace/trace.c
> @@ -3893,12 +3893,11 @@ trace_options_core_write(struct file *filp, const char __user *ubuf, size_t cnt,
>  
>  	switch (val) {
>  	case 0:
> -		trace_flags &= ~(1 << index);
> +		set_tracer_flags(1 << index, 0);
>  		break;
>  	case 1:
> -		trace_flags |= 1 << index;
> +		set_tracer_flags(1 << index, 1);
>  		break;
> -
>  	default:
>  		return -EINVAL;
>  	}


Acked-by: Frederic Weisbecker <fweisbec@xxxxxxxxx>

It applies on .31 BTW.

That makes me think: shouldn't it be better to keep
tracing/options for global options and options/* for
local tracer options? That breaks the ABI but...

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/