Messages in this thread |  | | Date | Wed, 23 Jan 2013 18:25:04 +0100 | Subject | Re: [PATCH 16/29] perf, tools: Add support for weight v7 | From | Stephane Eranian <> |
| |
On Wed, Jan 23, 2013 at 6:13 PM, Andi Kleen <andi@firstfloor.org> wrote: >> For PEBS-LL and possibly other special cases, it is important to remember >> that perf report always end up sorting by period (hist_collapse_resort). But >> for PEBS-LL we want to sort on nr_events * weight. Thus, with your patch, >> the only way, I found, to achieve this is by passing: >> >> add_hist_entry(self, &entry, al, weight, weight); > > Seems like a hack. IMHO it should always sort by all the keys > i specified with --sort in exactly that order I specified > Well, it does except for the "implicit" column which uses period, see hists__collapse_resort(). And that function is hardcoded to only look at period.
As for the hack, I did not put in in my hist__add_mem_entry() but rather in the caller. For PEBS-LL, the period is not important. I think it counts the number of loads/stores and not just the qualifying ones. For loads, that means it counts all loads and not just the ones above the threshold, but I may be wrong.
> I had a similar thing in a really old version of my patches, > but I gave it up because it was too unintuitive. > Well, but then it does not present a sensible view of the samples when weight is more important than period.
> -Andi
|  |