lkml.org 
[lkml]   [2013]   [Nov]   [1]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH 14/14] perf report: Add -g cumulative option
On Thu, Oct 31, 2013 at 03:56:16PM +0900, Namhyung Kim wrote:
> From: Namhyung Kim <namhyung.kim@lge.com>
>
> The -g cumulative option is for showing accumulated overhead (period)
> value as well as self overhead.
>
> Cc: Arun Sharma <asharma@fb.com>
> Cc: Frederic Weisbecker <fweisbec@gmail.com>
> Signed-off-by: Namhyung Kim <namhyung@kernel.org>
> ---
> tools/perf/Documentation/perf-report.txt | 2 ++
> tools/perf/builtin-report.c | 3 +++
> tools/perf/util/callchain.c | 12 +++++++++++-
> 3 files changed, 16 insertions(+), 1 deletion(-)
>
> diff --git a/tools/perf/Documentation/perf-report.txt b/tools/perf/Documentation/perf-report.txt
> index 10a279871251..b150bfb734f4 100644
> --- a/tools/perf/Documentation/perf-report.txt
> +++ b/tools/perf/Documentation/perf-report.txt
> @@ -130,6 +130,8 @@ OPTIONS
> - graph: use a graph tree, displaying absolute overhead rates.
> - fractal: like graph, but displays relative rates. Each branch of
> the tree is considered as a new profiled object. +
> + - cumulative: accumulate callchain to parent entry so that they can
> + show up in the output.
>
> order can be either:
> - callee: callee based call graph.
> diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
> index 3b626127c8d6..281053b28898 100644
> --- a/tools/perf/builtin-report.c
> +++ b/tools/perf/builtin-report.c
> @@ -1030,6 +1030,9 @@ parse_callchain_opt(const struct option *opt, const char *arg, int unset)
> else if (!strncmp(tok, "fractal", strlen(arg)))
> callchain_param.mode = CHAIN_GRAPH_REL;
>
> + else if (!strncmp(tok, "cumulative", strlen(arg)))
> + callchain_param.mode = CHAIN_CUMULATIVE;
> +
> else if (!strncmp(tok, "none", strlen(arg))) {
> callchain_param.mode = CHAIN_NONE;
> symbol_conf.use_callchain = false;
> diff --git a/tools/perf/util/callchain.c b/tools/perf/util/callchain.c
> index c10052c6e73c..c3c73eb839df 100644
> --- a/tools/perf/util/callchain.c
> +++ b/tools/perf/util/callchain.c
> @@ -150,6 +150,14 @@ sort_chain_graph_rel(struct rb_root *rb_root, struct callchain_root *chain_root,
> rb_root->rb_node = chain_root->node.rb_root.rb_node;
> }
>
> +static void
> +sort_chain_cumulative(struct rb_root *rb_root __maybe_unused,
> + struct callchain_root *chain_root __maybe_unused,
> + u64 min_hit __maybe_unused,
> + struct callchain_param *param __maybe_unused)
> +{
> +}

maybe add some commentary explaning that it's intentionaly empty

or maybe dont set it and do following check
in __hists__insert_output_entry:

if (symbol_conf.use_callchain && callchain_param.sort)
callchain_param.sort(&he->sorted_chain, he->callchain,
min_callchain_hits, &callchain_param);


jirka


\
 
 \ /
  Last update: 2013-11-01 15:01    [W:0.153 / U:0.200 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site