Re: [PATCH v5 4/7] perf report: Refactor the branch info printing code

[Jiri Olsa <jolsa@redhat.com>]

On Wed, Apr 19, 2017 at 11:48:11PM +0800, Jin Yao wrote:

SNIP

> +
>  static int counts_str_build(char *bf, int bfsize,
>  			     u64 branch_count, u64 predicted_count,
>  			     u64 abort_count, u64 cycles_count,
>  			     u64 iter_count, u64 samples_count)
>  {
> -	double predicted_percent = 0.0;
> -	const char *null_str = "";
> -	char iter_str[32];
> -	char cycle_str[32];
> -	char *istr, *cstr;
>  	u64 cycles;
> +	int printed = 0, i = 0;

I like it, but it looks like the previous code displayed those
bits in another order.. I managed to catch this one:

  1337c1337
  <              --0.53%--menu_select menu.c:218 (iterations:6 predicted:0.0%)
  ---
  >              --0.53%--menu_select menu.c:218 (predicted:0.0% iterations:6)

I think we better keep the current order, which seems
to be the goal of the original code as well

this function is perfect candidate for automated test ;-)
(something like we did in tests/kmod-path.c)

thanks,
jirka