Re: [PATCH 3/4] ftrace: No return value for ftrace_process_locs

[Steven Rostedt <rostedt@goodmis.org>]

On Wed, 2012-04-25 at 10:23 +0200, Jiri Olsa wrote:
> The return value of ftrace_process_locs is never checked. The function
> tries to update as many calls as possible and in case of error there's
> either warning output or ftrace_bug call.
> 
> Signed-off-by: Jiri Olsa <jolsa@redhat.com>
> ---
>  kernel/trace/ftrace.c |   37 ++++++++++++++++---------------------
>  1 files changed, 16 insertions(+), 21 deletions(-)
> 
> diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c
> index b3ceecd..e67f5b3 100644
> --- a/kernel/trace/ftrace.c
> +++ b/kernel/trace/ftrace.c
> @@ -1930,7 +1930,7 @@ static int ops_traces_mod(struct ftrace_ops *ops)
>  	return ftrace_hash_empty(hash);
>  }
>  
> -static int ftrace_update_code(struct module *mod)
> +static void ftrace_update_code(struct module *mod)
>  {
>  	struct ftrace_page *pg;
>  	struct dyn_ftrace *p;
> @@ -1959,7 +1959,7 @@ static int ftrace_update_code(struct module *mod)
>  		for (i = 0; i < pg->index; i++) {
>  			/* If something went wrong, bail without enabling anything */
>  			if (unlikely(ftrace_disabled))
> -				return -1;
> +				return;
>  
>  			p = &pg->records[i];
>  			p->flags = ref;
> @@ -1968,8 +1968,8 @@ static int ftrace_update_code(struct module *mod)
>  			 * Do the initial record conversion from mcount jump
>  			 * to the NOP instructions.
>  			 */
> -			if (!ftrace_code_disable(mod, p))
> -				break;
> +			if (WARN_ON_ONCE(!ftrace_code_disable(mod, p)))
> +				return;

Why the warning? If ftrace_disabled is set, something broke a long time
ago, and the ftrace_bug gives its own warning.

I don't think this is needed.

-- Steve