linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH v3] perf annotate/report: Remove hist__account_cycles from callback
@ 2019-03-15 21:16 Jin Yao
  2019-03-20  0:35 ` Jin, Yao
                   ` (2 more replies)
  0 siblings, 3 replies; 5+ messages in thread
From: Jin Yao @ 2019-03-15 21:16 UTC (permalink / raw)
  To: acme, jolsa, peterz, mingo, alexander.shishkin
  Cc: Linux-kernel, ak, kan.liang, yao.jin, Jin Yao

The hist__account_cycles is executed when the hist_iter__branch_callback
is called. But it looks it's not necessary. In hist__account_cycles, it
already walks on all branch entries.

This patch moves the hist__account_cycles out of callback, now the data
processing is much faster than before.

Previous code has an issue that the ch[offset].num++
(in __symbol__account_cycles) is executed repeatedly since
hist__account_cycles is called in each hist_iter__branch_callback,
so the counting of ch[offset].num is not correct (too big).
With this patch, the issue is fixed. And we don't need the code of
"ch->reset >= ch->num / 2" to check if there are too many overlaps
(in annotation__count_and_fill), otherwise some data would be
hidden.

Now, we can try, for example:

perf record -b ...
perf annotate or perf report -s symbol

The before/after output should be no change.

 v3:
 ---
 Fix the crash in stdio mode.
 Like previous code, it needs the checking of ui__has_annotation()
 before hist__account_cycles()

 v2:
 ---
 1. Cover the similar perf report
 2. Remove the checking code "ch->reset >= ch->num / 2"

Signed-off-by: Jin Yao <yao.jin@linux.intel.com>
---
 tools/perf/builtin-annotate.c |  4 ++--
 tools/perf/builtin-report.c   | 11 +++++------
 tools/perf/util/annotate.c    |  2 +-
 3 files changed, 8 insertions(+), 9 deletions(-)

diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c
index 67f9d9f..77deb3a 100644
--- a/tools/perf/builtin-annotate.c
+++ b/tools/perf/builtin-annotate.c
@@ -159,8 +159,6 @@ static int hist_iter__branch_callback(struct hist_entry_iter *iter,
 	struct perf_evsel *evsel = iter->evsel;
 	int err;
 
-	hist__account_cycles(sample->branch_stack, al, sample, false);
-
 	bi = he->branch_info;
 	err = addr_map_symbol__inc_samples(&bi->from, sample, evsel);
 
@@ -199,6 +197,8 @@ static int process_branch_callback(struct perf_evsel *evsel,
 	if (a.map != NULL)
 		a.map->dso->hit = 1;
 
+	hist__account_cycles(sample->branch_stack, al, sample, false);
+
 	ret = hist_entry_iter__add(&iter, &a, PERF_MAX_STACK_DEPTH, ann);
 	return ret;
 }
diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index 1921aaa..493e115 100644
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -136,9 +136,6 @@ static int hist_iter__report_callback(struct hist_entry_iter *iter,
 	if (!ui__has_annotation() && !rep->symbol_ipc)
 		return 0;
 
-	hist__account_cycles(sample->branch_stack, al, sample,
-			     rep->nonany_branch_mode);
-
 	if (sort__mode == SORT_MODE__BRANCH) {
 		bi = he->branch_info;
 		err = addr_map_symbol__inc_samples(&bi->from, sample, evsel);
@@ -181,9 +178,6 @@ static int hist_iter__branch_callback(struct hist_entry_iter *iter,
 	if (!ui__has_annotation() && !rep->symbol_ipc)
 		return 0;
 
-	hist__account_cycles(sample->branch_stack, al, sample,
-			     rep->nonany_branch_mode);
-
 	bi = he->branch_info;
 	err = addr_map_symbol__inc_samples(&bi->from, sample, evsel);
 	if (err)
@@ -282,6 +276,11 @@ static int process_sample_event(struct perf_tool *tool,
 	if (al.map != NULL)
 		al.map->dso->hit = 1;
 
+	if (ui__has_annotation() || rep->symbol_ipc) {
+		hist__account_cycles(sample->branch_stack, &al, sample,
+				     rep->nonany_branch_mode);
+	}
+
 	ret = hist_entry_iter__add(&iter, &al, rep->max_stack, rep);
 	if (ret < 0)
 		pr_debug("problem adding hist entry, skipping event\n");
diff --git a/tools/perf/util/annotate.c b/tools/perf/util/annotate.c
index 5f6dbbf..e8080c0 100644
--- a/tools/perf/util/annotate.c
+++ b/tools/perf/util/annotate.c
@@ -1015,7 +1015,7 @@ static void annotation__count_and_fill(struct annotation *notes, u64 start, u64
 		float ipc = n_insn / ((double)ch->cycles / (double)ch->num);
 
 		/* Hide data when there are too many overlaps. */
-		if (ch->reset >= 0x7fff || ch->reset >= ch->num / 2)
+		if (ch->reset >= 0x7fff)
 			return;
 
 		for (offset = start; offset <= end; offset++) {
-- 
2.7.4


^ permalink raw reply related	[flat|nested] 5+ messages in thread

* Re: [PATCH v3] perf annotate/report: Remove hist__account_cycles from callback
  2019-03-15 21:16 [PATCH v3] perf annotate/report: Remove hist__account_cycles from callback Jin Yao
@ 2019-03-20  0:35 ` Jin, Yao
  2019-03-20  1:47   ` Andi Kleen
  2019-04-10 13:00 ` Arnaldo Carvalho de Melo
  2019-05-18  8:46 ` [tip:perf/core] perf annotate: Remove hist__account_cycles() " tip-bot for Jin Yao
  2 siblings, 1 reply; 5+ messages in thread
From: Jin, Yao @ 2019-03-20  0:35 UTC (permalink / raw)
  To: acme, jolsa, peterz, mingo, alexander.shishkin
  Cc: Linux-kernel, ak, kan.liang, yao.jin

Nobody like this patch? :)

In my test, for a 72MB perf.data (with LBR data), the perf annotate 
processing speed:

Before: 10.2s
After: 2.2s

Thanks
Jin Yao

On 3/16/2019 5:16 AM, Jin Yao wrote:
> The hist__account_cycles is executed when the hist_iter__branch_callback
> is called. But it looks it's not necessary. In hist__account_cycles, it
> already walks on all branch entries.
> 
> This patch moves the hist__account_cycles out of callback, now the data
> processing is much faster than before.
> 
> Previous code has an issue that the ch[offset].num++
> (in __symbol__account_cycles) is executed repeatedly since
> hist__account_cycles is called in each hist_iter__branch_callback,
> so the counting of ch[offset].num is not correct (too big).
> With this patch, the issue is fixed. And we don't need the code of
> "ch->reset >= ch->num / 2" to check if there are too many overlaps
> (in annotation__count_and_fill), otherwise some data would be
> hidden.
> 
> Now, we can try, for example:
> 
> perf record -b ...
> perf annotate or perf report -s symbol
> 
> The before/after output should be no change.
> 
>   v3:
>   ---
>   Fix the crash in stdio mode.
>   Like previous code, it needs the checking of ui__has_annotation()
>   before hist__account_cycles()
> 
>   v2:
>   ---
>   1. Cover the similar perf report
>   2. Remove the checking code "ch->reset >= ch->num / 2"
> 
> Signed-off-by: Jin Yao <yao.jin@linux.intel.com>
> ---
>   tools/perf/builtin-annotate.c |  4 ++--
>   tools/perf/builtin-report.c   | 11 +++++------
>   tools/perf/util/annotate.c    |  2 +-
>   3 files changed, 8 insertions(+), 9 deletions(-)
> 
> diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c
> index 67f9d9f..77deb3a 100644
> --- a/tools/perf/builtin-annotate.c
> +++ b/tools/perf/builtin-annotate.c
> @@ -159,8 +159,6 @@ static int hist_iter__branch_callback(struct hist_entry_iter *iter,
>   	struct perf_evsel *evsel = iter->evsel;
>   	int err;
>   
> -	hist__account_cycles(sample->branch_stack, al, sample, false);
> -
>   	bi = he->branch_info;
>   	err = addr_map_symbol__inc_samples(&bi->from, sample, evsel);
>   
> @@ -199,6 +197,8 @@ static int process_branch_callback(struct perf_evsel *evsel,
>   	if (a.map != NULL)
>   		a.map->dso->hit = 1;
>   
> +	hist__account_cycles(sample->branch_stack, al, sample, false);
> +
>   	ret = hist_entry_iter__add(&iter, &a, PERF_MAX_STACK_DEPTH, ann);
>   	return ret;
>   }
> diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
> index 1921aaa..493e115 100644
> --- a/tools/perf/builtin-report.c
> +++ b/tools/perf/builtin-report.c
> @@ -136,9 +136,6 @@ static int hist_iter__report_callback(struct hist_entry_iter *iter,
>   	if (!ui__has_annotation() && !rep->symbol_ipc)
>   		return 0;
>   
> -	hist__account_cycles(sample->branch_stack, al, sample,
> -			     rep->nonany_branch_mode);
> -
>   	if (sort__mode == SORT_MODE__BRANCH) {
>   		bi = he->branch_info;
>   		err = addr_map_symbol__inc_samples(&bi->from, sample, evsel);
> @@ -181,9 +178,6 @@ static int hist_iter__branch_callback(struct hist_entry_iter *iter,
>   	if (!ui__has_annotation() && !rep->symbol_ipc)
>   		return 0;
>   
> -	hist__account_cycles(sample->branch_stack, al, sample,
> -			     rep->nonany_branch_mode);
> -
>   	bi = he->branch_info;
>   	err = addr_map_symbol__inc_samples(&bi->from, sample, evsel);
>   	if (err)
> @@ -282,6 +276,11 @@ static int process_sample_event(struct perf_tool *tool,
>   	if (al.map != NULL)
>   		al.map->dso->hit = 1;
>   
> +	if (ui__has_annotation() || rep->symbol_ipc) {
> +		hist__account_cycles(sample->branch_stack, &al, sample,
> +				     rep->nonany_branch_mode);
> +	}
> +
>   	ret = hist_entry_iter__add(&iter, &al, rep->max_stack, rep);
>   	if (ret < 0)
>   		pr_debug("problem adding hist entry, skipping event\n");
> diff --git a/tools/perf/util/annotate.c b/tools/perf/util/annotate.c
> index 5f6dbbf..e8080c0 100644
> --- a/tools/perf/util/annotate.c
> +++ b/tools/perf/util/annotate.c
> @@ -1015,7 +1015,7 @@ static void annotation__count_and_fill(struct annotation *notes, u64 start, u64
>   		float ipc = n_insn / ((double)ch->cycles / (double)ch->num);
>   
>   		/* Hide data when there are too many overlaps. */
> -		if (ch->reset >= 0x7fff || ch->reset >= ch->num / 2)
> +		if (ch->reset >= 0x7fff)
>   			return;
>   
>   		for (offset = start; offset <= end; offset++) {
> 

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH v3] perf annotate/report: Remove hist__account_cycles from callback
  2019-03-20  0:35 ` Jin, Yao
@ 2019-03-20  1:47   ` Andi Kleen
  0 siblings, 0 replies; 5+ messages in thread
From: Andi Kleen @ 2019-03-20  1:47 UTC (permalink / raw)
  To: Jin, Yao
  Cc: acme, jolsa, peterz, mingo, alexander.shishkin, Linux-kernel,
	kan.liang, yao.jin

On Wed, Mar 20, 2019 at 08:35:17AM +0800, Jin, Yao wrote:
> Nobody like this patch? :)
> 
> In my test, for a 72MB perf.data (with LBR data), the perf annotate
> processing speed:
> 
> Before: 10.2s
> After: 2.2s

That's great!

-Andi

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [PATCH v3] perf annotate/report: Remove hist__account_cycles from callback
  2019-03-15 21:16 [PATCH v3] perf annotate/report: Remove hist__account_cycles from callback Jin Yao
  2019-03-20  0:35 ` Jin, Yao
@ 2019-04-10 13:00 ` Arnaldo Carvalho de Melo
  2019-05-18  8:46 ` [tip:perf/core] perf annotate: Remove hist__account_cycles() " tip-bot for Jin Yao
  2 siblings, 0 replies; 5+ messages in thread
From: Arnaldo Carvalho de Melo @ 2019-04-10 13:00 UTC (permalink / raw)
  To: Jin Yao
  Cc: jolsa, peterz, mingo, alexander.shishkin, Linux-kernel, ak,
	kan.liang, yao.jin

Em Sat, Mar 16, 2019 at 05:16:17AM +0800, Jin Yao escreveu:
> The hist__account_cycles is executed when the hist_iter__branch_callback
> is called. But it looks it's not necessary. In hist__account_cycles, it
> already walks on all branch entries.
> 
> This patch moves the hist__account_cycles out of callback, now the data
> processing is much faster than before.
> 
> Previous code has an issue that the ch[offset].num++
> (in __symbol__account_cycles) is executed repeatedly since
> hist__account_cycles is called in each hist_iter__branch_callback,
> so the counting of ch[offset].num is not correct (too big).
> With this patch, the issue is fixed. And we don't need the code of
> "ch->reset >= ch->num / 2" to check if there are too many overlaps
> (in annotation__count_and_fill), otherwise some data would be
> hidden.
> 
> Now, we can try, for example:

Tested with:

  perf annotate --stdio
  perf annotate --stdio2
  perf annotate --tui
  perf report --tui -> press annotation hotkey 'a'

Applied,

- Arnaldo

^ permalink raw reply	[flat|nested] 5+ messages in thread

* [tip:perf/core] perf annotate: Remove hist__account_cycles() from callback
  2019-03-15 21:16 [PATCH v3] perf annotate/report: Remove hist__account_cycles from callback Jin Yao
  2019-03-20  0:35 ` Jin, Yao
  2019-04-10 13:00 ` Arnaldo Carvalho de Melo
@ 2019-05-18  8:46 ` tip-bot for Jin Yao
  2 siblings, 0 replies; 5+ messages in thread
From: tip-bot for Jin Yao @ 2019-05-18  8:46 UTC (permalink / raw)
  To: linux-tip-commits
  Cc: acme, ak, peterz, kan.liang, yao.jin, linux-kernel, hpa, jolsa,
	yao.jin, alexander.shishkin, mingo, tglx

Commit-ID:  bdd1666b3d03d675bdb7f8d92b29f2797acbc5e8
Gitweb:     https://git.kernel.org/tip/bdd1666b3d03d675bdb7f8d92b29f2797acbc5e8
Author:     Jin Yao <yao.jin@linux.intel.com>
AuthorDate: Sat, 16 Mar 2019 05:16:17 +0800
Committer:  Arnaldo Carvalho de Melo <acme@redhat.com>
CommitDate: Wed, 15 May 2019 16:36:46 -0300

perf annotate: Remove hist__account_cycles() from callback

The hist__account_cycles() function is executed when the
hist_iter__branch_callback() is called.

But it looks it's not necessary.  In hist__account_cycles, it already
walks on all branch entries.

This patch moves the hist__account_cycles out of callback, now the data
processing is much faster than before.

Previous code has an issue that the ch[offset].num++ (in
__symbol__account_cycles) is executed repeatedly since
hist__account_cycles is called in each hist_iter__branch_callback, so
the counting of ch[offset].num is not correct (too big).

With this patch, the issue is fixed. And we don't need the code of
"ch->reset >= ch->num / 2" to check if there are too many overlaps (in
annotation__count_and_fill), otherwise some data would be hidden.

Now, we can try, for example:

  perf record -b ...
  perf annotate or perf report -s symbol

The before/after output should be no change.

 v3:
 ---
 Fix the crash in stdio mode.
 Like previous code, it needs the checking of ui__has_annotation()
 before hist__account_cycles()

 v2:
 ---
 1. Cover the similar perf report
 2. Remove the checking code "ch->reset >= ch->num / 2"

Signed-off-by: Jin Yao <yao.jin@linux.intel.com>
Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Jin Yao <yao.jin@intel.com>
Cc: Jiri Olsa <jolsa@kernel.org>
Cc: Kan Liang <kan.liang@linux.intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1552684577-29041-1-git-send-email-yao.jin@linux.intel.com
Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com>
---
 tools/perf/builtin-annotate.c |  4 ++--
 tools/perf/builtin-report.c   | 11 +++++------
 tools/perf/util/annotate.c    |  2 +-
 3 files changed, 8 insertions(+), 9 deletions(-)

diff --git a/tools/perf/builtin-annotate.c b/tools/perf/builtin-annotate.c
index 67f9d9ffacfb..77deb3a40596 100644
--- a/tools/perf/builtin-annotate.c
+++ b/tools/perf/builtin-annotate.c
@@ -159,8 +159,6 @@ static int hist_iter__branch_callback(struct hist_entry_iter *iter,
 	struct perf_evsel *evsel = iter->evsel;
 	int err;
 
-	hist__account_cycles(sample->branch_stack, al, sample, false);
-
 	bi = he->branch_info;
 	err = addr_map_symbol__inc_samples(&bi->from, sample, evsel);
 
@@ -199,6 +197,8 @@ static int process_branch_callback(struct perf_evsel *evsel,
 	if (a.map != NULL)
 		a.map->dso->hit = 1;
 
+	hist__account_cycles(sample->branch_stack, al, sample, false);
+
 	ret = hist_entry_iter__add(&iter, &a, PERF_MAX_STACK_DEPTH, ann);
 	return ret;
 }
diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index 4054eb1f98ac..91e27ac297c2 100644
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -136,9 +136,6 @@ static int hist_iter__report_callback(struct hist_entry_iter *iter,
 	if (!ui__has_annotation() && !rep->symbol_ipc)
 		return 0;
 
-	hist__account_cycles(sample->branch_stack, al, sample,
-			     rep->nonany_branch_mode);
-
 	if (sort__mode == SORT_MODE__BRANCH) {
 		bi = he->branch_info;
 		err = addr_map_symbol__inc_samples(&bi->from, sample, evsel);
@@ -181,9 +178,6 @@ static int hist_iter__branch_callback(struct hist_entry_iter *iter,
 	if (!ui__has_annotation() && !rep->symbol_ipc)
 		return 0;
 
-	hist__account_cycles(sample->branch_stack, al, sample,
-			     rep->nonany_branch_mode);
-
 	bi = he->branch_info;
 	err = addr_map_symbol__inc_samples(&bi->from, sample, evsel);
 	if (err)
@@ -282,6 +276,11 @@ static int process_sample_event(struct perf_tool *tool,
 	if (al.map != NULL)
 		al.map->dso->hit = 1;
 
+	if (ui__has_annotation() || rep->symbol_ipc) {
+		hist__account_cycles(sample->branch_stack, &al, sample,
+				     rep->nonany_branch_mode);
+	}
+
 	ret = hist_entry_iter__add(&iter, &al, rep->max_stack, rep);
 	if (ret < 0)
 		pr_debug("problem adding hist entry, skipping event\n");
diff --git a/tools/perf/util/annotate.c b/tools/perf/util/annotate.c
index 09762985c713..0b8573fd9b05 100644
--- a/tools/perf/util/annotate.c
+++ b/tools/perf/util/annotate.c
@@ -1021,7 +1021,7 @@ static void annotation__count_and_fill(struct annotation *notes, u64 start, u64
 		float ipc = n_insn / ((double)ch->cycles / (double)ch->num);
 
 		/* Hide data when there are too many overlaps. */
-		if (ch->reset >= 0x7fff || ch->reset >= ch->num / 2)
+		if (ch->reset >= 0x7fff)
 			return;
 
 		for (offset = start; offset <= end; offset++) {

^ permalink raw reply related	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2019-05-18  8:46 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-03-15 21:16 [PATCH v3] perf annotate/report: Remove hist__account_cycles from callback Jin Yao
2019-03-20  0:35 ` Jin, Yao
2019-03-20  1:47   ` Andi Kleen
2019-04-10 13:00 ` Arnaldo Carvalho de Melo
2019-05-18  8:46 ` [tip:perf/core] perf annotate: Remove hist__account_cycles() " tip-bot for Jin Yao

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).