Re: [PATCH V2 04/14] perf script: Set up output options for in-stream attributes

[Adrian Hunter <adrian.hunter@intel.com>]

On 23/10/13 09:15, David Ahern wrote:
> On 10/22/13 8:34 AM, Adrian Hunter wrote:
>> +static int process_attr(struct perf_tool *tool, union perf_event *event,
>> +            struct perf_evlist **pevlist)
>> +{
>> +    struct perf_script *scr = container_of(tool, struct perf_script, tool);
>> +    struct perf_evlist *evlist;
>> +    struct perf_evsel *evsel, *pos;
>> +    int err;
>> +
>> +    err = perf_event__process_attr(tool, event, pevlist);
>> +    if (err)
>> +        return err;
>> +
>> +    evlist = *pevlist;
>> +    evsel = perf_evlist__last(*pevlist);
> 
> This assumes new entries are added to the end of evlist in
> perf_event__process_attr. Would be better to change it to return the newly
> created evsel so you don't need to look it up after adding it.

perf_event__process_attr() must not reorder the attributes,
it would misrepresent the way they were recorded.

> 
>> +
>> +    if (evsel->attr.type >= PERF_TYPE_MAX)
>> +        return 0;
>> +
>> +    list_for_each_entry(pos, &evlist->entries, node) {
>> +        if (pos->attr.type == evsel->attr.type && pos != evsel)
>> +            return 0;
>> +    }
> 
> What's the point of this loop?

Each type is checked once - see perf_session__check_output_opt()