Re: [tip:perf/urgent] perf hists: Fix determination of a callchain node's childlessness

Re: [tip:perf/urgent] perf hists: Fix determination of a callchain node's childlessness

[Andres Freund]

Hi,

Perhaps the patch quoted below should also go into stable? It'd be nice
to fix that regression for 4.4 and 4.5.  It's been integrated into 4.6
(909890355).

Regards,

Andres

On 2016-03-30 23:33:37 -0700, tip-bot for Andres Freund wrote:
> Commit-ID:  909890355507e92bdaf648e73870f6b5df606da8
> Gitweb:     http://git.kernel.org/tip/909890355507e92bdaf648e73870f6b5df606da8
> Author:     Andres Freund <andres@anarazel.de>
> AuthorDate: Wed, 30 Mar 2016 21:02:45 +0200
> Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
> CommitDate: Wed, 30 Mar 2016 18:08:39 -0300
> 
> perf hists: Fix determination of a callchain node's childlessness
> 
> The 4b3a3212233a ("perf hists browser: Support flat callchains") commit
> over-aggressively tried to optimize callchain_node__init_have_children().
> 
> That lead to --tui mode not allowing to expand call chain elements if a
> call chain element had only one parent. That's why --inverted callgraphs
> looked halfway sane, but plain ones didn't.
> 
> Revert that individual optimization, it wasn't really related to the
> rest of the commit.
> 
> Signed-off-by: Andres Freund <andres@anarazel.de>
> Acked-by: Namhyung Kim <namhyung@kernel.org>
> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
> Fixes: 4b3a3212233a ("perf hists browser: Support flat callchains")
> Link: http://lkml.kernel.org/r/20160330190245.GB13305@awork2.anarazel.de
> Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
> ---
>  tools/perf/ui/browsers/hists.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/tools/perf/ui/browsers/hists.c b/tools/perf/ui/browsers/hists.c
> index 4b98165..2a83414 100644
> --- a/tools/perf/ui/browsers/hists.c
> +++ b/tools/perf/ui/browsers/hists.c
> @@ -337,7 +337,7 @@ static void callchain_node__init_have_children(struct callchain_node *node,
>  	chain = list_entry(node->val.next, struct callchain_list, list);
>  	chain->has_children = has_sibling;
>  
> -	if (node->val.next != node->val.prev) {
> +	if (!list_empty(&node->val)) {
>  		chain = list_entry(node->val.prev, struct callchain_list, list);
>  		chain->has_children = !RB_EMPTY_ROOT(&node->rb_root);
>  	}

Re: [tip:perf/urgent] perf hists: Fix determination of a callchain node's childlessness

[Greg KH]

On Mon, Apr 18, 2016 at 07:25:22AM -0700, Andres Freund wrote:
> Hi,
> 
> Perhaps the patch quoted below should also go into stable? It'd be nice
> to fix that regression for 4.4 and 4.5.  It's been integrated into 4.6
> (909890355).

How is it a regression in 4.4, as the commit involved didn't show up
until 4.5?

confused,

greg k-h

Re: [tip:perf/urgent] perf hists: Fix determination of a callchain node's childlessness

[Andres Freund]

Hi,

On 2016-05-02 15:17:54 -0700, Greg KH wrote:
> On Mon, Apr 18, 2016 at 07:25:22AM -0700, Andres Freund wrote:
> > Hi,
> > 
> > Perhaps the patch quoted below should also go into stable? It'd be nice
> > to fix that regression for 4.4 and 4.5.  It's been integrated into 4.6
> > (909890355).
> 
> How is it a regression in 4.4, as the commit involved didn't show up
> until 4.5?
> 
> confused,

Sorry for that, I'd used git describe on the commit to determine which
branch it was contained in (outputting v4.3-1188-g4b3a321). That
obviously doesn't make sense though, as that just returns the state of
the tree at which 4b3a3212233a was introduced, not the commit it was
merged into mainline.


Andres

Re: [tip:perf/urgent] perf hists: Fix determination of a callchain node's childlessness

[Greg KH]

On Mon, May 02, 2016 at 03:25:25PM -0700, Andres Freund wrote:
> Hi,
> 
> On 2016-05-02 15:17:54 -0700, Greg KH wrote:
> > On Mon, Apr 18, 2016 at 07:25:22AM -0700, Andres Freund wrote:
> > > Hi,
> > > 
> > > Perhaps the patch quoted below should also go into stable? It'd be nice
> > > to fix that regression for 4.4 and 4.5.  It's been integrated into 4.6
> > > (909890355).
> > 
> > How is it a regression in 4.4, as the commit involved didn't show up
> > until 4.5?
> > 
> > confused,
> 
> Sorry for that, I'd used git describe on the commit to determine which
> branch it was contained in (outputting v4.3-1188-g4b3a321). That
> obviously doesn't make sense though, as that just returns the state of
> the tree at which 4b3a3212233a was introduced, not the commit it was
> merged into mainline.

I have an alias in my .gitconfig:
[alias]
	dc = describe --contains

which is what you need here:
	$ git dc 4b3a3212233a
	v4.5-rc1~171^2~43^2~3

hope this helps,

greg k-h

Re: [tip:perf/urgent] perf hists: Fix determination of a callchain node's childlessness

[Andres Freund]

Hi,

On 2016-05-02 16:24:08 -0700, Greg KH wrote:
> On Mon, May 02, 2016 at 03:25:25PM -0700, Andres Freund wrote:
> > On 2016-05-02 15:17:54 -0700, Greg KH wrote:
> > Sorry for that, I'd used git describe on the commit to determine which
> > branch it was contained in (outputting v4.3-1188-g4b3a321). That
> > obviously doesn't make sense though, as that just returns the state of
> > the tree at which 4b3a3212233a was introduced, not the commit it was
> > merged into mainline.
> 
> I have an alias in my .gitconfig:
> [alias]
> 	dc = describe --contains
> 
> which is what you need here:
> 	$ git dc 4b3a3212233a
> 	v4.5-rc1~171^2~43^2~3
> 
> hope this helps,

Ah, perfect. Thanks. I made do with git tag --contains, but that's
awfully slow; describe --contains is a lot faster.


Andres