[PATCH 1/3] perf sched timehist: Mark schedule function in callchains

[PATCH 1/3] perf sched timehist: Mark schedule function in callchains

[Namhyung Kim]

The sched_switch event always captured from the scheduler function.  So
it'd be great omit them from the callchain.  This patch marks the
functions to be omitted by later patch.

Cc: David Ahern <dsahern@gmail.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
---
 tools/perf/builtin-sched.c | 21 +++++++++++++++++++++
 tools/perf/util/symbol.h   |  1 +
 2 files changed, 22 insertions(+)

diff --git a/tools/perf/builtin-sched.c b/tools/perf/builtin-sched.c
index 829468defa07..cd14189d3943 100644
--- a/tools/perf/builtin-sched.c
+++ b/tools/perf/builtin-sched.c
@@ -1965,7 +1965,28 @@ static bool is_idle_sample(struct perf_sched *sched,
 
 		return false;
 	}
+
 	callchain_cursor_commit(cursor);
+
+	while (true) {
+		struct callchain_cursor_node *node;
+		struct symbol *sym;
+
+		node = callchain_cursor_current(cursor);
+		if (node == NULL)
+			break;
+
+		sym = node->sym;
+		if (sym && sym->name) {
+			if (!strcmp(sym->name, "schedule") ||
+			    !strcmp(sym->name, "__schedule") ||
+			    !strcmp(sym->name, "preempt_schedule"))
+				sym->ignore = 1;
+		}
+
+		callchain_cursor_advance(cursor);
+	}
+
 	return false;
 }
 
diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h
index dec7e2d44885..1bcbefc0c325 100644
--- a/tools/perf/util/symbol.h
+++ b/tools/perf/util/symbol.h
@@ -58,6 +58,7 @@ struct symbol {
 	u16		namelen;
 	u8		binding;
 	u8		idle:1;
+	u8		ignore:1;
 	u8		arch_sym;
 	char		name[0];
 };
-- 
2.10.0

[PATCH 2/3] perf tools: Add option to skip ignore symbol when printing callchains

[Namhyung Kim]

For tracepoint events, callchains always contain certain functions.
Sometimes it'd be better to skip those functions as they have no value.

Cc: David Ahern <dsahern@gmail.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
---
 tools/perf/builtin-sched.c      | 3 ++-
 tools/perf/util/evsel.h         | 1 +
 tools/perf/util/evsel_fprintf.c | 7 ++++++-
 3 files changed, 9 insertions(+), 2 deletions(-)

diff --git a/tools/perf/builtin-sched.c b/tools/perf/builtin-sched.c
index cd14189d3943..06be809a02ab 100644
--- a/tools/perf/builtin-sched.c
+++ b/tools/perf/builtin-sched.c
@@ -1876,7 +1876,8 @@ static void timehist_print_sample(struct perf_sched *sched,
 
 	sample__fprintf_sym(sample, al, 0,
 			    EVSEL__PRINT_SYM | EVSEL__PRINT_ONELINE |
-			    EVSEL__PRINT_CALLCHAIN_ARROW,
+			    EVSEL__PRINT_CALLCHAIN_ARROW |
+			    EVSEL__PRINT_SKIP_IGNORED,
 			    &callchain_cursor, stdout);
 
 out:
diff --git a/tools/perf/util/evsel.h b/tools/perf/util/evsel.h
index 27fa3a343577..6abb89cd27f9 100644
--- a/tools/perf/util/evsel.h
+++ b/tools/perf/util/evsel.h
@@ -392,6 +392,7 @@ int perf_evsel__fprintf(struct perf_evsel *evsel,
 #define EVSEL__PRINT_SRCLINE		(1<<5)
 #define EVSEL__PRINT_UNKNOWN_AS_ADDR	(1<<6)
 #define EVSEL__PRINT_CALLCHAIN_ARROW	(1<<7)
+#define EVSEL__PRINT_SKIP_IGNORED	(1<<8)
 
 struct callchain_cursor;
 
diff --git a/tools/perf/util/evsel_fprintf.c b/tools/perf/util/evsel_fprintf.c
index 53bb614feafb..5a6f52284452 100644
--- a/tools/perf/util/evsel_fprintf.c
+++ b/tools/perf/util/evsel_fprintf.c
@@ -109,6 +109,7 @@ int sample__fprintf_callchain(struct perf_sample *sample, int left_alignment,
 	int print_srcline = print_opts & EVSEL__PRINT_SRCLINE;
 	int print_unknown_as_addr = print_opts & EVSEL__PRINT_UNKNOWN_AS_ADDR;
 	int print_arrow = print_opts & EVSEL__PRINT_CALLCHAIN_ARROW;
+	int print_skip_ignored = print_opts & EVSEL__PRINT_SKIP_IGNORED;
 	char s = print_oneline ? ' ' : '\t';
 	bool first = true;
 
@@ -124,6 +125,9 @@ int sample__fprintf_callchain(struct perf_sample *sample, int left_alignment,
 			if (!node)
 				break;
 
+			if (node->sym && node->sym->ignore && print_skip_ignored)
+				goto next;
+
 			printed += fprintf(fp, "%-*.*s", left_alignment, left_alignment, " ");
 
 			if (print_arrow && !first)
@@ -162,8 +166,9 @@ int sample__fprintf_callchain(struct perf_sample *sample, int left_alignment,
 			if (!print_oneline)
 				printed += fprintf(fp, "\n");
 
-			callchain_cursor_advance(cursor);
 			first = false;
+next:
+			callchain_cursor_advance(cursor);
 		}
 	}
 
-- 
2.10.0

[PATCH 3/3] perf sched timehist: Enlarge max stack depth by 2

[Namhyung Kim]

When it records callchains, they will always have 2 scheduler functions
(__schedule + schedule or __schedule + preempt_schedule) and get
ignored.  So it should collect 2 more functions to show the expected
number of callchains to user.

Cc: David Ahern <dsahern@gmail.com>
Signed-off-by: Namhyung Kim <namhyung@kernel.org>
---
 tools/perf/builtin-sched.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/perf/builtin-sched.c b/tools/perf/builtin-sched.c
index 06be809a02ab..a49a032f5b15 100644
--- a/tools/perf/builtin-sched.c
+++ b/tools/perf/builtin-sched.c
@@ -1960,7 +1960,7 @@ static bool is_idle_sample(struct perf_sched *sched,
 		return false;
 
 	if (thread__resolve_callchain(thread, cursor, evsel, sample,
-				      NULL, NULL, sched->max_stack) != 0) {
+				      NULL, NULL, sched->max_stack + 2) != 0) {
 		if (verbose)
 			error("Failed to resolve callchain. Skipping\n");
 
-- 
2.10.0

Re: [PATCH 1/3] perf sched timehist: Mark schedule function in callchains

[David Ahern]

On 11/23/16 8:11 PM, Namhyung Kim wrote:
> The sched_switch event always captured from the scheduler function.  So
> it'd be great omit them from the callchain.  This patch marks the
> functions to be omitted by later patch.

I had this covered by a symbol filter:

https://github.com/dsahern/linux/blob/perf/full-monty-4.1/tools/perf/builtin-sched.c#L3000

Not sure what happened over the years but that should still work and allows the user to add more symbols to ignore:

https://github.com/dsahern/linux/blob/perf/full-monty-4.1/tools/perf/builtin-sched.c#L3449

Re: [PATCH 1/3] perf sched timehist: Mark schedule function in callchains

[Namhyung Kim]

Hi David,

On Wed, Nov 23, 2016 at 10:13:46PM -0500, David Ahern wrote:
> On 11/23/16 8:11 PM, Namhyung Kim wrote:
> > The sched_switch event always captured from the scheduler function.  So
> > it'd be great omit them from the callchain.  This patch marks the
> > functions to be omitted by later patch.
> 
> I had this covered by a symbol filter:
> 
> https://github.com/dsahern/linux/blob/perf/full-monty-4.1/tools/perf/builtin-sched.c#L3000
> 
> Not sure what happened over the years but that should still work and
> allows the user to add more symbols to ignore:

The symbol filter was removed by commit be39db9f2932 ("perf symbols:
Remove symbol_filter_t machinery").

Thanks,
Namhyung


> 
> https://github.com/dsahern/linux/blob/perf/full-monty-4.1/tools/perf/builtin-sched.c#L3449
> 

Re: [PATCH 1/3] perf sched timehist: Mark schedule function in callchains

[Arnaldo Carvalho de Melo]

Em Thu, Nov 24, 2016 at 10:11:12AM +0900, Namhyung Kim escreveu:
> The sched_switch event always captured from the scheduler function.  So
> it'd be great omit them from the callchain.  This patch marks the
> functions to be omitted by later patch.

I just reorderded this so that after applying 1 I could right away test
it, i.e. 2/3 was applied first.

- Arnaldo
 
> Cc: David Ahern <dsahern@gmail.com>
> Signed-off-by: Namhyung Kim <namhyung@kernel.org>
> ---
>  tools/perf/builtin-sched.c | 21 +++++++++++++++++++++
>  tools/perf/util/symbol.h   |  1 +
>  2 files changed, 22 insertions(+)
> 
> diff --git a/tools/perf/builtin-sched.c b/tools/perf/builtin-sched.c
> index 829468defa07..cd14189d3943 100644
> --- a/tools/perf/builtin-sched.c
> +++ b/tools/perf/builtin-sched.c
> @@ -1965,7 +1965,28 @@ static bool is_idle_sample(struct perf_sched *sched,
>  
>  		return false;
>  	}
> +
>  	callchain_cursor_commit(cursor);
> +
> +	while (true) {
> +		struct callchain_cursor_node *node;
> +		struct symbol *sym;
> +
> +		node = callchain_cursor_current(cursor);
> +		if (node == NULL)
> +			break;
> +
> +		sym = node->sym;
> +		if (sym && sym->name) {
> +			if (!strcmp(sym->name, "schedule") ||
> +			    !strcmp(sym->name, "__schedule") ||
> +			    !strcmp(sym->name, "preempt_schedule"))
> +				sym->ignore = 1;
> +		}
> +
> +		callchain_cursor_advance(cursor);
> +	}
> +
>  	return false;
>  }
>  
> diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h
> index dec7e2d44885..1bcbefc0c325 100644
> --- a/tools/perf/util/symbol.h
> +++ b/tools/perf/util/symbol.h
> @@ -58,6 +58,7 @@ struct symbol {
>  	u16		namelen;
>  	u8		binding;
>  	u8		idle:1;
> +	u8		ignore:1;
>  	u8		arch_sym;
>  	char		name[0];
>  };
> -- 
> 2.10.0

[tip:perf/core] perf callchain: Add option to skip ignore symbol when printing callchains

[tip-bot for Namhyung Kim]

Commit-ID:  2d9bbf6eb3825739efa9e91c256ce7ead60d8367
Gitweb:     http://git.kernel.org/tip/2d9bbf6eb3825739efa9e91c256ce7ead60d8367
Author:     Namhyung Kim <namhyung@kernel.org>
AuthorDate: Thu, 24 Nov 2016 10:11:13 +0900
Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
CommitDate: Fri, 25 Nov 2016 10:49:38 -0300

perf callchain: Add option to skip ignore symbol when printing callchains

For tracepoint events, callchains always contain certain functions.
Sometimes it'd be better to skip those functions as they have no value.

Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/r/20161124011114.7102-2-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-sched.c      | 3 ++-
 tools/perf/util/evsel.h         | 1 +
 tools/perf/util/evsel_fprintf.c | 7 ++++++-
 3 files changed, 9 insertions(+), 2 deletions(-)

diff --git a/tools/perf/builtin-sched.c b/tools/perf/builtin-sched.c
index 829468d..43fcc13 100644
--- a/tools/perf/builtin-sched.c
+++ b/tools/perf/builtin-sched.c
@@ -1876,7 +1876,8 @@ static void timehist_print_sample(struct perf_sched *sched,
 
 	sample__fprintf_sym(sample, al, 0,
 			    EVSEL__PRINT_SYM | EVSEL__PRINT_ONELINE |
-			    EVSEL__PRINT_CALLCHAIN_ARROW,
+			    EVSEL__PRINT_CALLCHAIN_ARROW |
+			    EVSEL__PRINT_SKIP_IGNORED,
 			    &callchain_cursor, stdout);
 
 out:
diff --git a/tools/perf/util/evsel.h b/tools/perf/util/evsel.h
index 27fa3a3..6abb89c 100644
--- a/tools/perf/util/evsel.h
+++ b/tools/perf/util/evsel.h
@@ -392,6 +392,7 @@ int perf_evsel__fprintf(struct perf_evsel *evsel,
 #define EVSEL__PRINT_SRCLINE		(1<<5)
 #define EVSEL__PRINT_UNKNOWN_AS_ADDR	(1<<6)
 #define EVSEL__PRINT_CALLCHAIN_ARROW	(1<<7)
+#define EVSEL__PRINT_SKIP_IGNORED	(1<<8)
 
 struct callchain_cursor;
 
diff --git a/tools/perf/util/evsel_fprintf.c b/tools/perf/util/evsel_fprintf.c
index 53bb614..5a6f522 100644
--- a/tools/perf/util/evsel_fprintf.c
+++ b/tools/perf/util/evsel_fprintf.c
@@ -109,6 +109,7 @@ int sample__fprintf_callchain(struct perf_sample *sample, int left_alignment,
 	int print_srcline = print_opts & EVSEL__PRINT_SRCLINE;
 	int print_unknown_as_addr = print_opts & EVSEL__PRINT_UNKNOWN_AS_ADDR;
 	int print_arrow = print_opts & EVSEL__PRINT_CALLCHAIN_ARROW;
+	int print_skip_ignored = print_opts & EVSEL__PRINT_SKIP_IGNORED;
 	char s = print_oneline ? ' ' : '\t';
 	bool first = true;
 
@@ -124,6 +125,9 @@ int sample__fprintf_callchain(struct perf_sample *sample, int left_alignment,
 			if (!node)
 				break;
 
+			if (node->sym && node->sym->ignore && print_skip_ignored)
+				goto next;
+
 			printed += fprintf(fp, "%-*.*s", left_alignment, left_alignment, " ");
 
 			if (print_arrow && !first)
@@ -162,8 +166,9 @@ int sample__fprintf_callchain(struct perf_sample *sample, int left_alignment,
 			if (!print_oneline)
 				printed += fprintf(fp, "\n");
 
-			callchain_cursor_advance(cursor);
 			first = false;
+next:
+			callchain_cursor_advance(cursor);
 		}
 	}
 

[tip:perf/core] perf sched timehist: Mark schedule function in callchains

[tip-bot for Namhyung Kim]

Commit-ID:  cdeb01bf7863718bbbbdac2a2c3a12b62366757a
Gitweb:     http://git.kernel.org/tip/cdeb01bf7863718bbbbdac2a2c3a12b62366757a
Author:     Namhyung Kim <namhyung@kernel.org>
AuthorDate: Thu, 24 Nov 2016 10:11:12 +0900
Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
CommitDate: Fri, 25 Nov 2016 10:49:43 -0300

perf sched timehist: Mark schedule function in callchains

The sched_switch event always captured from the scheduler function.  So
it'd be great omit them from the callchain.  This patch marks the
functions to be omitted by later patch.

Committer notes:

Testing it:

Before:

  [root@jouet experimental]# perf sched record -g ls
  Dockerfile  perf.data  x-mips64
  [ perf record: Woken up 1 times to write data ]
  [ perf record: Captured and wrote 1.355 MB perf.data (29 samples) ]
  [root@jouet experimental]# perf sched timehist
      time  cpu  task name         wait time sch delay run time
                 [tid/pid]             (msec) (msec) (msec)
  ----------- -----  ----------------- ------ ------ ------
  6.494998 [001] <idle>                0.000  0.000  0.000
  6.495027 [002] perf[519]             0.000  0.000  0.000 __schedule <- schedule <- schedule_hrtimeout_range_clock <- schedule_hrtimeou
  6.495096 [003] <idle>                0.000  0.000  0.000
  6.495100 [003] rcuos/0[9]            0.000  0.005  0.003 __schedule <- schedule <- rcu_nocb_kthread <- kthread <- ret_from_fork
  6.495113 [001] perf[520]             0.000  0.008  0.114 __schedule <- preempt_schedule_common <- _cond_resched <- wait_for_completion
  6.495121 [000] <idle>                0.000  0.000  0.000
  6.495129 [001] migration/1[17]       0.000  0.003  0.016 __schedule <- schedule <- smpboot_thread_fn <- kthread <- ret_from_fork
  6.496085 [002] <idle>                0.000  0.000  1.057
  6.496096 [002] kworker/u16:1[31169]  0.000  0.004  0.011 __schedule <- schedule <- worker_thread <- kthread <- ret_from_fork
  6.496096 [003] <idle>                0.003  0.000  0.996
  6.496169 [002] <idle>                0.011  0.000  0.072
  6.496171 [000] ls[520]               0.008  0.000  1.049 __schedule <- schedule <- do_exit <- do_group_exit <- [unknown]
  6.496172 [003] gnome-terminal-[4391] 0.000  0.003  0.076 __schedule <- schedule <- schedule_hrtimeout_range_clock <- schedule_hrtimeo

After:

  [root@jouet experimental]# perf sched timehist
      time  cpu  task name         wait time sch delay run time
                 [tid/pid]            (msec)  (msec)  (msec)
  ----------- -----  ----------------- -----  -----  ------
  6.494998 [001] <idle>                0.000  0.000  0.000
  6.495027 [002] perf[519]             0.000  0.000  0.000 schedule_hrtimeout_range_clock <- schedule_hrtimeout_range <- poll_schedule_t
  6.495096 [003] <idle>                0.000  0.000  0.000
  6.495100 [003] rcuos/0[9]            0.000  0.005  0.003 rcu_nocb_kthread <- kthread <- ret_from_fork
  6.495113 [001] perf[520]             0.000  0.008  0.114 preempt_schedule_common <- _cond_resched <- wait_for_completion <- stop_one_c
  6.495121 [000] <idle>                0.000  0.000  0.000
  6.495129 [001] migration/1[17]       0.000  0.003  0.016 smpboot_thread_fn <- kthread <- ret_from_fork
  6.496085 [002] <idle>                0.000  0.000  1.057
  6.496096 [002] kworker/u16:1[31169]  0.000  0.004  0.011 worker_thread <- kthread <- ret_from_fork
  6.496096 [003] <idle>                0.003  0.000  0.996
  6.496169 [002] <idle>                0.011  0.000  0.072
  6.496171 [000] ls[520]               0.008  0.000  1.049 do_exit <- do_group_exit <- [unknown]
  6.496172 [003] gnome-terminal-[4391] 0.000  0.003  0.076 schedule_hrtimeout_range_clock <- schedule_hrtimeout_range <- poll_schedule_
  [root@jouet experimental]#

Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/r/20161124011114.7102-1-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-sched.c | 21 +++++++++++++++++++++
 tools/perf/util/symbol.h   |  1 +
 2 files changed, 22 insertions(+)

diff --git a/tools/perf/builtin-sched.c b/tools/perf/builtin-sched.c
index 43fcc13..06be809 100644
--- a/tools/perf/builtin-sched.c
+++ b/tools/perf/builtin-sched.c
@@ -1966,7 +1966,28 @@ static bool is_idle_sample(struct perf_sched *sched,
 
 		return false;
 	}
+
 	callchain_cursor_commit(cursor);
+
+	while (true) {
+		struct callchain_cursor_node *node;
+		struct symbol *sym;
+
+		node = callchain_cursor_current(cursor);
+		if (node == NULL)
+			break;
+
+		sym = node->sym;
+		if (sym && sym->name) {
+			if (!strcmp(sym->name, "schedule") ||
+			    !strcmp(sym->name, "__schedule") ||
+			    !strcmp(sym->name, "preempt_schedule"))
+				sym->ignore = 1;
+		}
+
+		callchain_cursor_advance(cursor);
+	}
+
 	return false;
 }
 
diff --git a/tools/perf/util/symbol.h b/tools/perf/util/symbol.h
index dec7e2d4..1bcbefc 100644
--- a/tools/perf/util/symbol.h
+++ b/tools/perf/util/symbol.h
@@ -58,6 +58,7 @@ struct symbol {
 	u16		namelen;
 	u8		binding;
 	u8		idle:1;
+	u8		ignore:1;
 	u8		arch_sym;
 	char		name[0];
 };

[tip:perf/core] perf sched timehist: Enlarge max stack depth by 2

[tip-bot for Namhyung Kim]

Commit-ID:  8388deb3ba4d36ffcae91a2a01cb2ea6f27553e6
Gitweb:     http://git.kernel.org/tip/8388deb3ba4d36ffcae91a2a01cb2ea6f27553e6
Author:     Namhyung Kim <namhyung@kernel.org>
AuthorDate: Thu, 24 Nov 2016 10:11:14 +0900
Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
CommitDate: Fri, 25 Nov 2016 10:50:57 -0300

perf sched timehist: Enlarge max stack depth by 2

When it records callchains, they will always have 2 scheduler functions
(__schedule + schedule or __schedule + preempt_schedule) and get
ignored.  So it should collect 2 more functions to show the expected
number of callchains to user.

Committer Notes:

Example of final result, using the same perf.data file as in the
previous cset comment, but this time redirecting the output of 'perf
sched timehist' to a file instead of copy'n'pasting from xterm:

  [root@jouet experimental]# perf sched timehist > /tmp/bla
  [root@jouet experimental]# cat /tmp/bla
      time  cpu task name        wait time sch delay run time
                 [tid/pid]            (msec) (msec) (msec)
  -------- ----  -------------------- ------ ------ -----
  6.494998 [01] <idle>                0.000  0.000  0.000
  6.495027 [02] perf[519]             0.000  0.000  0.000 schedule_hrtimeout_range_clock <- schedule_hrtimeout_range <- poll_schedule_timeout <- do_sys_poll <- sys_poll
  6.495096 [03] <idle>                0.000  0.000  0.000
  6.495100 [03] rcuos/0[9]            0.000  0.005  0.003 rcu_nocb_kthread <- kthread <- ret_from_fork
  6.495113 [01] perf[520]             0.000  0.008  0.114 preempt_schedule_common <- _cond_resched <- wait_for_completion <- stop_one_cpu <- sched_exec <- do_execveat_common.isra.35
  6.495121 [00] <idle>                0.000  0.000  0.000
  6.495129 [01] migration/1[17]       0.000  0.003  0.016 smpboot_thread_fn <- kthread <- ret_from_fork
  6.496085 [02] <idle>                0.000  0.000  1.057
  6.496096 [02] kworker/u16:1[31169]  0.000  0.004  0.011 worker_thread <- kthread <- ret_from_fork
  6.496096 [03] <idle>                0.003  0.000  0.996
  6.496169 [02] <idle>                0.011  0.000  0.072
  6.496171 [00] ls[520]               0.008  0.000  1.049 do_exit <- do_group_exit <- [unknown] <- entry_SYSCALL_64_fastpath
  6.496172 [03] gnome-terminal-[4391] 0.000  0.003  0.076 schedule_hrtimeout_range_clock <- schedule_hrtimeout_range <- poll_schedule_timeout <- do_sys_poll <- sys_poll

Signed-off-by: Namhyung Kim <namhyung@kernel.org>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/r/20161124011114.7102-3-namhyung@kernel.org
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-sched.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/perf/builtin-sched.c b/tools/perf/builtin-sched.c
index 06be809..a49a032 100644
--- a/tools/perf/builtin-sched.c
+++ b/tools/perf/builtin-sched.c
@@ -1960,7 +1960,7 @@ static bool is_idle_sample(struct perf_sched *sched,
 		return false;
 
 	if (thread__resolve_callchain(thread, cursor, evsel, sample,
-				      NULL, NULL, sched->max_stack) != 0) {
+				      NULL, NULL, sched->max_stack + 2) != 0) {
 		if (verbose)
 			error("Failed to resolve callchain. Skipping\n");
 

Re: [PATCH 1/3] perf sched timehist: Mark schedule function in callchains

[David Ahern]

On 11/23/16 11:30 PM, Namhyung Kim wrote:
> Hi David,
> 
> On Wed, Nov 23, 2016 at 10:13:46PM -0500, David Ahern wrote:
>> On 11/23/16 8:11 PM, Namhyung Kim wrote:
>>> The sched_switch event always captured from the scheduler function.  So
>>> it'd be great omit them from the callchain.  This patch marks the
>>> functions to be omitted by later patch.
>>
>> I had this covered by a symbol filter:
>>
>> https://github.com/dsahern/linux/blob/perf/full-monty-4.1/tools/perf/builtin-sched.c#L3000
>>
>> Not sure what happened over the years but that should still work and
>> allows the user to add more symbols to ignore:
> 
> The symbol filter was removed by commit be39db9f2932 ("perf symbols:
> Remove symbol_filter_t machinery").

That's unfortunate. The exclude list is huge in removing redundant callchains and getting more relevant information on the screen with a stack depth of 5. I'll see what I can cook up with the existing code.