Messages in this thread | | | Date | Tue, 30 Sep 2014 10:17:34 -0300 | From | Arnaldo Carvalho de Melo <> | Subject | Re: [PATCH v2 1/2] perf callchain: Create an address space per thread |
| |
Em Tue, Sep 30, 2014 at 03:11:21PM +0900, Namhyung Kim escreveu: > > On Mon, 29 Sep 2014 10:34:52 +0200, Jiri Olsa wrote: > > On Mon, Sep 29, 2014 at 01:58:11PM +0900, Namhyung Kim wrote: > >> @@ -37,6 +38,9 @@ struct thread *thread__new(pid_t pid, pid_t tid) > >> thread->cpu = -1; > >> INIT_LIST_HEAD(&thread->comm_list);
> >> + if (unwind__prepare_access(thread) < 0) > >> + goto err_thread;
> > So this is something we need to do only for DWARF unwind via libunwind. > > Do we want some condition to bypass this in for the rest of the cases?
> Ah, you're right, we need to check that before doing this.
> > I think we could check for symbol_conf.use_callchain and for the > > PERF_SAMPLE_REGS_USER|PERF_SAMPLE_STACK_USER sample types in evlist, > > something like:
> > bool has_dwarf_data_for_callchains = symbol_conf.use_callchain && evlist_sample_type_has(evlist, PERF_SAMPLE_REGS_USER|PERF_SAMPLE_STACK_USER)
> Hmm.. it seems that it's hard to get the evlist at this point. Maybe we > need to add a field in the symbol_conf or a global variable.. :/
Please don't export evlist in a global way, just set some flag on symbol_conf.
At some point we need to ditch even symbol_conf :-\
- Arnaldo
| |