Re: Scrolling down broken with "perf top --hierarchy"

[Namhyung Kim <namhyung@kernel.org>]

Hi Taeung,

On Mon, Oct 24, 2016 at 07:10:24PM +0900, Taeung Song wrote:
> Hi, Namhyung and Arnaldo :)
> 
> On 10/24/2016 02:11 PM, Namhyung Kim wrote:
> > Hi Arnaldo,
> > 
> > Sorry for late reply.
> > 
> > On Fri, Oct 07, 2016 at 11:35:45AM -0300, Arnaldo Carvalho de Melo wrote:
> > > Em Fri, Oct 07, 2016 at 01:53:57PM +0900, Namhyung Kim escreveu:
> > > > Cc-ing perf maintainers,
> > > > 
> > > > On Fri, Oct 07, 2016 at 06:32:29AM +0200, Markus Trippelsdorf wrote:
> > > > > On 2016.10.07 at 13:22 +0900, Namhyung Kim wrote:
> > > > > > On Fri, Oct 07, 2016 at 05:51:18AM +0200, Markus Trippelsdorf wrote:
> > > > > > > On 2016.10.07 at 10:17 +0900, Namhyung Kim wrote:
> > > > > > > > On Thu, Oct 06, 2016 at 06:33:33PM +0200, Markus Trippelsdorf wrote:
> > > > > > > > > Scrolling down is broken when using "perf top --hierarchy".
> > > > > > > > > When it starts up everything is OK and one can scroll up and down to all
> > > > > > > > > entries. But as further and further new entries get added to the list,
> > > > > > > > > scrolling down is blocked (at the position of the last entry that was
> > > > > > > > > shown directly after startup).
> > > > > > > > 
> > > > > > > > I think below patch will fix the problem.  Please check.
> > > > > > > 
> > > > > > > Yes. It works fine now. Many thanks.
> > > > > > 
> > > > > > Good.  Can I add your Tested-by then?
> > > > > 
> > > > > Sure.
> > > > 
> > > > Ok, I'll send a formal patch with it.
> > > > 
> > > > > 
> > > > > (And in the long run you should think of making "perf top --hierarchy"
> > > > > the default for perf top, because it gives a much better (uncluttered)
> > > > > overview of what is going on.)
> > > > 
> > > > I think it's a matter of taste.  Some people prefer to see the top
> > > > single function or something (i.e. current behavior) while others
> > > > prefer to see a higher-level view.
> > > > 
> > > > But we can think again about the default at least for perf-top.  I
> > > > worried about changing default behavior because last time we did it
> > > > for children mode many people complained about it.  But I do think the
> > > > hierarchy mode is useful for many people though.
> > > 
> > > So, I think in such cases we could experiment with asking the user about
> > > switching to the new mode by showing a popup message telling what it is
> > > about, if the user says "yes, I want to try it" switch to it and if
> > > another hotkey is pressed later, write what was chosen (yes, switch to
> > > this new mode, no, I don't like it, don't pester me about it anymore) to
> > > its ~/.perfconfig file so that next time it goes straight to this new
> > > mode, else don't ask the user again and keep using whatever mode was
> > > there already.
> > > 
> > > What do you think?
> > 
> > I think it's a flexible way to set the default behavior while it seems
> > like a little bit complicated for implementation.  Also I think it's
> > better to popup another dialog at the end and asks for comfirmation
> > (but it might not be appropriate for --stdio).
> > 
> > And to do that, we need to have a (programmable) way of dealing with
> > the configs.
> > 
> > Taeung, is there an update on your config patchset (especially for
> > write support)?
> > 
> 
> Is related this link with what you said ?
> https://lkml.org/lkml/2016/1/11/495

Yep, that kind of thing.

> 
> Yes, the config patchset would be need to be updated.
> Because the config patchset which has 'write' feature
> don't use a recent 'struct perf_config_set' so I should change it
> to use 'perf_config_set' like show_config() of builtin-config.c:36.
> 
> Do you need write support of perf-config command ?
> If this feature is more necessary than a recent patchset about default
> config array https://lkml.org/lkml/2016/9/5/17,
> I'd remake config patchset for getting and setting features first. :)

What I need is a way to add a config item with specific value.  Maybe
I can just append a line into a config file, but it needs to check
possible conflict somehow.  So I think it needs to process existing
config items properly and update with the new value.

Thanks,
Namhyung