Re: [PATCH 2/2] perf tools: Construct LBR call chain

[Jiri Olsa <jolsa@redhat.com>]

On Thu, Nov 06, 2014 at 09:58:06AM -0500, kan.liang@intel.com wrote:

SNIP

> +	struct ip_callchain *chain = sample->callchain;
>  	u8 cpumode = PERF_RECORD_MISC_USER;
>  	int chain_nr = min(max_stack, (int)chain->nr);
> -	int i;
> -	int j;
> -	int err;
> +	int i, j, err = 0;
>  	int skip_idx __maybe_unused;
> +	int lbr = 0;
> +	u64 ip;
>  
>  	callchain_cursor_reset(&callchain_cursor);
>  
> @@ -1419,74 +1471,81 @@ static int thread__resolve_callchain_sample(struct thread *thread,
>  		return 0;
>  	}
>  
> -	/*
> -	 * Based on DWARF debug information, some architectures skip
> -	 * a callchain entry saved by the kernel.
> -	 */
> -	skip_idx = arch_skip_callchain_idx(thread, chain);
> +	if (evsel->attr.branch_sample_type & PERF_SAMPLE_BRANCH_CALL_STACK)
> +		lbr = 1;

please put this in a function.. like

bool has_branch_callstack(evsel)

thanks,
jirka