Re: [PATCH v2 3/5] perf report: Sort by sampled cycles percent per block for stdio

[Jin, Yao]

On 10/22/2019 12:08 AM, Jiri Olsa wrote:
> On Tue, Oct 15, 2019 at 01:33:48PM +0800, Jin Yao wrote:
>
> SNIP
>
> > +			cycles += bi->cycles_aggr / bi->num_aggr;
> > +
> > +			he_block = hists__add_entry_block(&bh->block_hists,
> > +							  &al, bi);
> > +			if (!he_block) {
> > +				block_info__put(bi);
> > +				return -1;
> > +			}
> > +		}
> > +	}
> > +
> > +	if (block_cycles)
> > +		*block_cycles += cycles;
> > +
> > +	return 0;
> > +}
> > +
> > +static int resort_cb(struct hist_entry *he, void *arg __maybe_unused)
> > +{
> > +	/* Skip the calculation of column length in output_resort */
> > +	he->filtered = true;
>
> that's a nasty hack ;-) together with setting it back to false just below
>
> why do you want to skip the columns calculation? we could add those columns
> to the output as well no?
>
> jirka

The columns calculation for this case causes the crash. The current 
columns calculation requires some information but we don't provide. So I 
just want to skip the columns calculation.

OK, I will check how to avoid this nasty hack.

Thanks
Jin Yao

> > +	return 0;
> > +}
> > +
> > +static void hists__clear_filtered(struct hists *hists)
> > +{
> > +	struct rb_node *next = rb_first_cached(&hists->entries);
> > +	struct hist_entry *he;
> > +
> > +	while (next) {
> > +		he = rb_entry(next, struct hist_entry, rb_node);
> > +		he->filtered = false;
> > +		next = rb_next(&he->rb_node);
> > +	}
> > +}
>
> SNIP
>
> jirka