Skip to contentSkip to conversation

partial callgrap and time support

RFC PATCH Superseded [email protected] 2 participants 10 messages Jul 7, 2015 – Jul 13, 2015 perf events

v2 of 3 Superseded Latest is v3 →

lore
Share & tools
lore
#0 lore raw

v2 cover partial callgrap and time support

From: Kan Liang <[email protected]>

This patchkit adds the ability to turn off callgraphs and time stamps per event. This in term can reduce sampling overhead and the size of the perf.data.

Changes since V1:
- Break up V1 patches into three patches(parse option changes, partial time support and partial callgrap support).
- Use strings 'fp,dwarf,lbr,no' to identify callchains
- Add test case in parse-events.c

Kan Liang (5):
perf,tools: introduce OPT_CALLBACK_SET/OPT_CALLBACK_NOOPT_SET perf/documentation: Add description for period perf,tool: partial time support
perf,tool: partial callgrap support
perf,tests: Add tests to callgrap and time parse

 tools/perf/Documentation/perf-record.txt | 14 ++++++++
 tools/perf/builtin-record.c              |  7 ++--
 tools/perf/perf.h                        |  2 ++
 tools/perf/tests/parse-events.c          | 28 ++++++++++++++++
 tools/perf/util/evsel.c                  | 56 ++++++++++++++++++++++++++++++--
 tools/perf/util/parse-events.c           | 25 ++++++++++++++
 tools/perf/util/parse-events.h           |  3 ++
 tools/perf/util/parse-events.l           |  3 ++
 tools/perf/util/parse-options.c          |  2 ++
 tools/perf/util/parse-options.h          |  4 +++
 10 files changed, 138 insertions(+), 6 deletions(-)
Signature · 1 line
1.8.3.1
#1 lore raw

RFC PATCH v2 1/5 perf,tools: introduce OPT_CALLBACK_SET/OPT_CALLBACK_NOOPT_SET

From: Kan Liang <[email protected]>

This patch extends the OPT_CALLBACK/OPT_CALLBACK_NOOPT to support 'set' item of struct option. So the perf knows whether an option was set by user.
The new macros are used by call-graph and timestamp options.

---
 tools/perf/builtin-record.c     | 7 ++++---
 tools/perf/perf.h               | 2 ++
 tools/perf/util/parse-options.c | 2 ++
 tools/perf/util/parse-options.h | 4 ++++
 4 files changed, 12 insertions(+), 3 deletions(-)
tools/perf/builtin-record.c +4 -3
index de165a1..c270993 100644
@@ -1018,10 +1018,10 @@ struct option __record_options[] = {
 		     record__parse_mmap_pages),
 	OPT_BOOLEAN(0, "group", &record.opts.group,
 		    "put the counters into a counter group"),
-	OPT_CALLBACK_NOOPT('g', NULL, &record.opts,
+	OPT_CALLBACK_NOOPT_SET('g', NULL, &record.opts, &record.opts.callgraph_set,
 			   NULL, "enables call-graph recording" ,
 			   &record_callchain_opt),
-	OPT_CALLBACK(0, "call-graph", &record.opts,
+	OPT_CALLBACK_SET(0, "call-graph", &record.opts, &record.opts.callgraph_set,
 		     "mode[,dump_size]", record_callchain_help,
 		     &record_parse_callchain_opt),
 	OPT_INCR('v', "verbose", &verbose,
@@ -1030,7 +1030,8 @@ struct option __record_options[] = {
 	OPT_BOOLEAN('s', "stat", &record.opts.inherit_stat,
 		    "per thread counts"),
 	OPT_BOOLEAN('d', "data", &record.opts.sample_address, "Record the sample addresses"),
-	OPT_BOOLEAN('T', "timestamp", &record.opts.sample_time, "Record the sample timestamps"),
+	OPT_BOOLEAN_SET('T', "timestamp", &record.opts.sample_time,
+			&record.opts.sample_time_set, "Sample timestamps"),
 	OPT_BOOLEAN('P', "period", &record.opts.period, "Record the sample period"),
 	OPT_BOOLEAN('n', "no-samples", &record.opts.no_samples,
 		    "don't sample"),
tools/perf/perf.h +2 -0
index 4a5827ff..9ba02e0 100644
@@ -51,6 +51,8 @@ struct record_opts {
 	bool	     sample_address;
 	bool	     sample_weight;
 	bool	     sample_time;
+	bool	     sample_time_set;
+	bool	     callgraph_set;
 	bool	     period;
 	bool	     sample_intr_regs;
 	bool	     running_time;
tools/perf/util/parse-options.c +2 -0
index 01626be..064385f 100644
@@ -140,6 +140,8 @@ static int get_value(struct parse_opt_ctx_t *p,
 		return err;
 
 	case OPTION_CALLBACK:
+		if (opt->set)
+			*(bool *)opt->set = true;
 		if (unset)
 			return (*opt->callback)(opt, NULL, 1) ? (-1) : 0;
 		if (opt->flags & PARSE_OPT_NOARG)
tools/perf/util/parse-options.h +4 -0
index 367d8b8..2bec32e 100644
@@ -132,8 +132,12 @@ struct option {
 	{ .type = OPTION_CALLBACK, .short_name = (s), .long_name = (l), .value = (v), .argh = "time", .help = (h), .callback = parse_opt_approxidate_cb }
 #define OPT_CALLBACK(s, l, v, a, h, f) \
 	{ .type = OPTION_CALLBACK, .short_name = (s), .long_name = (l), .value = (v), (a), .help = (h), .callback = (f) }
+#define OPT_CALLBACK_SET(s, l, v, os, a, h, f) \
+	{ .type = OPTION_CALLBACK, .short_name = (s), .long_name = (l), .value = (v), (a), .help = (h), .callback = (f), .set = check_vtype(os, bool *) }
 #define OPT_CALLBACK_NOOPT(s, l, v, a, h, f) \
 	{ .type = OPTION_CALLBACK, .short_name = (s), .long_name = (l), .value = (v), (a), .help = (h), .callback = (f), .flags = PARSE_OPT_NOARG }
+#define OPT_CALLBACK_NOOPT_SET(s, l, v, os, a, h, f) \
+	{ .type = OPTION_CALLBACK, .short_name = (s), .long_name = (l), .value = (v), (a), .help = (h), .callback = (f), .flags = PARSE_OPT_NOARG, .set = check_vtype(os, bool *) }
 #define OPT_CALLBACK_DEFAULT(s, l, v, a, h, f, d) \
 	{ .type = OPTION_CALLBACK, .short_name = (s), .long_name = (l), .value = (v), (a), .help = (h), .callback = (f), .defval = (intptr_t)d, .flags = PARSE_OPT_LASTARG_DEFAULT }
 #define OPT_CALLBACK_DEFAULT_NOOPT(s, l, v, a, h, f, d) \
Signature · 1 line
1.8.3.1
#2 lore raw

Re: [PATCH RFC V2 1/5] perf,tools: introduce OPT_CALLBACK_SET/OPT_CALLBACK_NOOPT_SET

Quoted 33 lines · On Tue, Jul 07, 2015 at 06:19:01AM -0400, [email protected] wrote:

From: Kan Liang <[email protected]>

This patch extends the OPT_CALLBACK/OPT_CALLBACK_NOOPT to support 'set' item of struct option. So the perf knows whether an option was set by user.
The new macros are used by call-graph and timestamp options.

---
 tools/perf/builtin-record.c     | 7 ++++---
 tools/perf/perf.h               | 2 ++
 tools/perf/util/parse-options.c | 2 ++
 tools/perf/util/parse-options.h | 4 ++++
 4 files changed, 12 insertions(+), 3 deletions(-)
tools/perf/builtin-record.c +2 -2
index de165a1..c270993 100644
@@ -1018,10 +1018,10 @@ struct option __record_options[] = {
 		     record__parse_mmap_pages),
 	OPT_BOOLEAN(0, "group", &record.opts.group,
 		    "put the counters into a counter group"),
-	OPT_CALLBACK_NOOPT('g', NULL, &record.opts,
+	OPT_CALLBACK_NOOPT_SET('g', NULL, &record.opts, &record.opts.callgraph_set,
 			   NULL, "enables call-graph recording" ,
 			   &record_callchain_opt),
-	OPT_CALLBACK(0, "call-graph", &record.opts,
+	OPT_CALLBACK_SET(0, "call-graph", &record.opts, &record.opts.callgraph_set,
 		     "mode[,dump_size]", record_callchain_help,
 		     &record_parse_callchain_opt),
 	OPT_INCR('v', "verbose", &verbose,

tools/perf/builtin-record.c struct option __record_options[] = { · line 1034

Quoted 6 lines
tools/perf/builtin-record.c +2 -1
@@ -1030,7 +1030,8 @@ struct option __record_options[] = {
 	OPT_BOOLEAN('s', "stat", &record.opts.inherit_stat,
 		    "per thread counts"),
 	OPT_BOOLEAN('d', "data", &record.opts.sample_address, "Record the sample addresses"),
-	OPT_BOOLEAN('T', "timestamp", &record.opts.sample_time, "Record the sample timestamps"),
+	OPT_BOOLEAN_SET('T', "timestamp", &record.opts.sample_time,
+			&record.opts.sample_time_set, "Sample timestamps"),

this hunk got in as:
3abebc55d70b perf record: Let user have timestamps with per-thread recording

jirka

#3 lore raw

RFC PATCH v2 2/5 perf/documentation: Add description for period

In reply to [email protected]

From: Kan Liang <[email protected]>

'period' param is not defined in
/sys/bus/event_sources/devices/<pmu>/format/*

---
 tools/perf/Documentation/perf-record.txt | 8 ++++++++
 1 file changed, 8 insertions(+)
tools/perf/Documentation/perf-record.txt +8 -0
index 9b9d9d0..5b47b2c 100644
@@ -45,6 +45,14 @@ OPTIONS
           param1 and param2 are defined as formats for the PMU in:
           /sys/bus/event_sources/devices/<pmu>/format/*
 
+	  There are also some params which are not defined in .../<pmu>/format/*.
+	  These params can be used to set event defaults.
+	  Here is a list of the params.
+	  - 'period': Set event sampling period
+
+	  Note: If user explicitly sets options which conflict with the params,
+	  the value set by the params will be overridden.
+
         - a hardware breakpoint event in the form of '\mem:addr[/len][:access]'
           where addr is the address in memory you want to break in.
           Access is the memory access type (read, write, execute) it can
Signature · 1 line
1.8.3.1
#4 lore raw

RFC PATCH v2 3/5 perf,tool: partial time support

In reply to [email protected]

From: Kan Liang <[email protected]>

When multiple events are sampled it may not be needed to collect fine grained time stamps on all events. The sample sites are usually nearby. It's enough to have time stamps on the regular reference events. This patchkit adds the ability to turn off time stamps per event. This in term can reduce sampling overhead and the size of the perf.data.

---
 tools/perf/Documentation/perf-record.txt |  4 +++-
 tools/perf/util/evsel.c                  | 32 ++++++++++++++++++++++++++++++--
 tools/perf/util/parse-events.c           |  7 +++++++
 tools/perf/util/parse-events.h           |  1 +
 tools/perf/util/parse-events.l           |  1 +
 5 files changed, 42 insertions(+), 3 deletions(-)
tools/perf/Documentation/perf-record.txt +3 -1
index 5b47b2c..df47907 100644
@@ -49,7 +49,9 @@ OPTIONS
 	  These params can be used to set event defaults.
 	  Here is a list of the params.
 	  - 'period': Set event sampling period
-
+	  - 'time': Disable/enable time stamping. Acceptable values are 1 for
+		    enabling time stamping. 0 for disabling time stamping.
+		    The default is 1.
 	  Note: If user explicitly sets options which conflict with the params,
 	  the value set by the params will be overridden.
 
tools/perf/util/evsel.c +30 -2
index 2936b30..b313cc7 100644
@@ -619,10 +619,37 @@ void perf_evsel__config(struct perf_evsel *evsel, struct record_opts *opts)
 	struct perf_event_attr *attr = &evsel->attr;
 	int track = evsel->tracking;
 	bool per_cpu = opts->target.default_per_cpu && !opts->target.per_thread;
+	bool sample_time = opts->sample_time;
 
 	attr->sample_id_all = perf_missing_features.sample_id_all ? 0 : 1;
 	attr->inherit	    = !opts->no_inherit;
 
+	/*
+	 * If user doesn't explicitly set time option,
+	 * let event attribute decide.
+	 */
+
+	if (!opts->sample_time_set) {
+		if (attr->sample_type & PERF_SAMPLE_TIME)
+			sample_time = true;
+		else
+			sample_time = false;
+	}
+
+	/*
+	 * Event parsing doesn't check the availability
+	 * Clear the bit which event parsing may be set.
+	 * Let following code check and reset if available
+	 *
+	 * Also, the sample size may be caculated mistakenly,
+	 * because event parsing may set the PERF_SAMPLE_TIME.
+	 * Remove the size which add in perf_evsel__init
+	 */
+	if (attr->sample_type & PERF_SAMPLE_TIME) {
+		attr->sample_type &= ~PERF_SAMPLE_TIME;
+		evsel->sample_size -= sizeof(u64);
+	}
+
 	perf_evsel__set_sample_bit(evsel, IP);
 	perf_evsel__set_sample_bit(evsel, TID);
 
@@ -705,13 +732,14 @@ void perf_evsel__config(struct perf_evsel *evsel, struct record_opts *opts)
 	/*
 	 * When the user explicitely disabled time don't force it here.
 	 */
-	if (opts->sample_time &&
+	if (sample_time &&
 	    (!perf_missing_features.sample_id_all &&
 	    (!opts->no_inherit || target__has_cpu(&opts->target) || per_cpu)))
 		perf_evsel__set_sample_bit(evsel, TIME);
 
 	if (opts->raw_samples && !evsel->no_aux_samples) {
-		perf_evsel__set_sample_bit(evsel, TIME);
+		if (sample_time)
+			perf_evsel__set_sample_bit(evsel, TIME);
 		perf_evsel__set_sample_bit(evsel, RAW);
 		perf_evsel__set_sample_bit(evsel, CPU);
 	}
tools/perf/util/parse-events.c +7 -0
index 09f8d23..f1e4aa5 100644
@@ -598,6 +598,13 @@ do {									   \
 		 * attr->branch_sample_type = term->val.num;
 		 */
 		break;
+	case PARSE_EVENTS__TERM_TYPE_TIME:
+		CHECK_TYPE_VAL(NUM);
+		if (term->val.num > 1)
+			return -EINVAL;
+		if (term->val.num == 1)
+			attr->sample_type |= PERF_SAMPLE_TIME;
+		 break;
 	case PARSE_EVENTS__TERM_TYPE_NAME:
 		CHECK_TYPE_VAL(STR);
 		break;
tools/perf/util/parse-events.h +1 -0
index 131f29b..0d8cae3 100644
@@ -62,6 +62,7 @@ enum {
 	PARSE_EVENTS__TERM_TYPE_NAME,
 	PARSE_EVENTS__TERM_TYPE_SAMPLE_PERIOD,
 	PARSE_EVENTS__TERM_TYPE_BRANCH_SAMPLE_TYPE,
+	PARSE_EVENTS__TERM_TYPE_TIME,
 };
 
 struct parse_events_term {
tools/perf/util/parse-events.l +1 -0
index 13cef3c..f542750 100644
@@ -183,6 +183,7 @@ config2			{ return term(yyscanner, PARSE_EVENTS__TERM_TYPE_CONFIG2); }
 name			{ return term(yyscanner, PARSE_EVENTS__TERM_TYPE_NAME); }
 period			{ return term(yyscanner, PARSE_EVENTS__TERM_TYPE_SAMPLE_PERIOD); }
 branch_type		{ return term(yyscanner, PARSE_EVENTS__TERM_TYPE_BRANCH_SAMPLE_TYPE); }
+time			{ return term(yyscanner, PARSE_EVENTS__TERM_TYPE_TIME); }
 ,			{ return ','; }
 "/"			{ BEGIN(INITIAL); return '/'; }
 {name_minus}		{ return str(yyscanner, PE_NAME); }
Signature · 1 line
1.8.3.1
#5 lore raw

Re: [PATCH RFC V2 3/5] perf,tool: partial time support

On Tue, Jul 07, 2015 at 06:19:03AM -0400, [email protected] wrote:

SNIP

Quoted 7 lines
tools/perf/util/evsel.c +1 -1
@@ -705,13 +732,14 @@ void perf_evsel__config(struct perf_evsel *evsel, struct record_opts *opts)
 	/*
 	 * When the user explicitely disabled time don't force it here.
 	 */
-	if (opts->sample_time &&
+	if (sample_time &&

tools/perf/util/evsel.c void perf_evsel__config(struct perf_evsel *evsel, struct record_opts *opts) · line 744

Quoted 10 lines
tools/perf/util/evsel.c +2 -1
 	    (!perf_missing_features.sample_id_all &&
 	    (!opts->no_inherit || target__has_cpu(&opts->target) || per_cpu)))
 		perf_evsel__set_sample_bit(evsel, TIME);
 
 	if (opts->raw_samples && !evsel->no_aux_samples) {
-		perf_evsel__set_sample_bit(evsel, TIME);
+		if (sample_time)
+			perf_evsel__set_sample_bit(evsel, TIME);
 		perf_evsel__set_sample_bit(evsel, RAW);
 		perf_evsel__set_sample_bit(evsel, CPU);

these bits no longer apply due to:
3abebc55d70b perf record: Let user have timestamps with per-thread recording

please rebase to latest acme's perf/core

thanks,
jirka

#6 lore raw

Re: [PATCH RFC V2 3/5] perf,tool: partial time support

In reply to [email protected]
Quoted 7 lines · On Tue, Jul 07, 2015 at 06:19:03AM -0400, [email protected] wrote:

From: Kan Liang <[email protected]>

When multiple events are sampled it may not be needed to collect fine grained time stamps on all events. The sample sites are usually nearby. It's enough to have time stamps on the regular reference events. This patchkit adds the ability to turn off time stamps per event. This in term can reduce sampling overhead and the size of the perf.data.

how is this usable in some perf example? I dont get the correlation between 'refference' and the rest of the events in the report time..

thanks,
jirka

author
#7 lore raw

RE: [PATCH RFC V2 3/5] perf,tool: partial time support

Quoted 13 lines

-----Original Message-----
From: Jiri Olsa [mailto:[email protected]]
Sent: Sunday, July 12, 2015 9:22 AM
To: Liang, Kan
Cc: [email protected]; [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [PATCH RFC V2 3/5] perf,tool: partial time support

Quoted 2 lines · On Tue, Jul 07, 2015 at 06:19:03AM -0400, [email protected] wrote:

From: Kan Liang <[email protected]>

Quoted 9 lines
Quoted 5 lines

When multiple events are sampled it may not be needed to collect fine grained time stamps on all events. The sample sites are usually nearby. It's enough to have time stamps on the regular reference events. This patchkit adds the ability to turn off time stamps per event. This in term can reduce sampling overhead and the size of the perf.data.

how is this usable in some perf example? I dont get the correlation between 'refference' and the rest of the events in the report time..

One usable case of partial time is to work with partial callgraph to enable "PEBS threshold > 1" (https://lkml.org/lkml/2015/5/10/196), which can significantly reduce the sampling overhead.
For the case as above, we intend to get callgraph in large PEBS threshold. Timestamp is not that important. So I didn't correlate the reference event and the rest events here.

But it's doable. We can use oe->last_flush to set the rest events' time. The only problem is that I cannot find a useable partial time only case. (Partial time itself cannot save as much as partial callgraph) So I think we do not need to implement such correlation.

I guess I will keep the code unchanged, but correct the description. Is it OK?

Thanks,
Kan

#8 lore raw

RFC PATCH v2 4/5 perf,tool: partial callgrap support

In reply to [email protected]

From: Kan Liang <[email protected]>

When multiple events are sampled it may not be needed to collect callgraphs for all of them. The sample sites are usually nearby, and it's enough to collect the callgraphs on a reference event (such as precise cycles or precise instructions).
This patchkit adds the ability to turn off callgraphs and time stamps per event. This in term can reduce sampling overhead and the size of the perf.data. Furthermore, it makes collecting back traces and timestamps possible when PEBS threshold > 1, which significantly reducing the sampling overhead especially for frequently occurring events
(https://lkml.org/lkml/2015/5/10/196). For example, A slower event with a larger period collects back traces/timestamps. Other more events run fast with multi-pebs. The time stamps from the slower events can be used to order the faster events. Their backtraces can give the user enough hint to find the right spot.

Here are some examples and test results.

1. Comparing the elapsed time and perf.data size from "kernbench -M -H".

The test command for FULL callgrap and time support.
"perf record -e '{cpu/cpu-cycles,period=100000/,cpu/instructions,period=20000/p}' --call-graph fp --time"

The test command for PARTIAL callgrap and time support.
"perf record -e '{cpu/cpu-cycles,callgraph=fp,time,period=100000/,

     cpu/instructions,callgraph=no,time=0,period=20000/p}'"

The elapsed time for FULL is 24.3 Sec, while for PARTIAL is 16.9 Sec. The perf.data size for FULL is 22.1 Gb, while for PARTIAL is 12.4 Gb.

2. Comparing the perf.data size and callgraph results.

The test command for FULL callgrap and time support.
"perf record -e '{cpu/cpu-cycles,period=100000/pp,cpu/instructions,period=20000/p}' --call-graph fp -- ./tchain_edit"

The test command for PARTIAL callgrap and time support.
"perf record -e '{cpu/cpu-cycles,callgraph=fp,time,period=100000/pp,

     cpu/instructions,callgraph=no,time=0,period=20000/p}'

-- ./tchain_edit"

The perf.data size for FULL is 43.2 MB, while for PARTIAL is 21.1 MB. The callgraph is roughly the same.

The callgraph from FULL
# Samples: 87K of event 'cpu/cpu-cycles,callgraph=fp,time,period=100000/pp' # Event count (approx.): 8760000000
#

 # Children      Self  Command      Shared Object       Symbol
 # ........  ........  ...........  ..................
..........................................
 #
    99.98%     0.00%  tchain_edit  libc-2.15.so        [.]
__libc_start_main
            |
            ---__libc_start_main

    99.97%     0.00%  tchain_edit  tchain_edit         [.] main
            |
            ---main
               __libc_start_main

    99.97%     0.00%  tchain_edit  tchain_edit         [.] f1
            |
            ---f1
               main
               __libc_start_main

    99.85%    87.01%  tchain_edit  tchain_edit         [.] f3
            |
            ---f3
               |
               |--99.74%-- f2
               |          f1
               |          main
               |          __libc_start_main
                --0.26%-- [...]
    99.71%     0.12%  tchain_edit  tchain_edit         [.] f2
            |
            ---f2
               f1
               main
               __libc_start_main

The callgraph from PARTIAL
# Samples: 417K of event 'cpu/instructions,callgraph=no,time=0,period=20000/p' # Event count (approx.): 8346980000
#

 # Children      Self  Command      Shared Object     Symbol
 # ........  ........  ...........  ................
..........................................
 #
    98.82%     0.00%  tchain_edit  libc-2.15.so      [.]
__libc_start_main
            |
            ---__libc_start_main

    98.82%     0.00%  tchain_edit  tchain_edit       [.] main
            |
            ---main
               __libc_start_main

    98.82%     0.00%  tchain_edit  tchain_edit       [.] f1
            |
            ---f1
               main
               __libc_start_main

    98.82%    98.28%  tchain_edit  tchain_edit       [.] f3
            |
            ---f3
               |
               |--0.53%-- f2
               |          f1
               |          main
               |          __libc_start_main
               |
               |--0.01%-- f1
               |          main
               |          __libc_start_main
                --99.46%-- [...]
    97.63%     0.03%  tchain_edit  tchain_edit       [.] f2
            |
            ---f2
               f1
               main
               __libc_start_main

     7.13%     0.03%  tchain_edit  [kernel.vmlinux]  [k] do_nmi
            |
            ---do_nmi
               end_repeat_nmi
               f3
               f2
               f1
               main
               __libc_start_main
---
 tools/perf/Documentation/perf-record.txt |  4 ++++
 tools/perf/util/evsel.c                  | 24 +++++++++++++++++++++++-
 tools/perf/util/parse-events.c           | 18 ++++++++++++++++++
 tools/perf/util/parse-events.h           |  2 ++
 tools/perf/util/parse-events.l           |  2 ++
 5 files changed, 49 insertions(+), 1 deletion(-)
tools/perf/Documentation/perf-record.txt +4 -0
index df47907..f478dc2 100644
@@ -52,6 +52,10 @@ OPTIONS
 	  - 'time': Disable/enable time stamping. Acceptable values are 1 for
 		    enabling time stamping. 0 for disabling time stamping.
 		    The default is 1.
+	  - 'callgraph': Disable/enable callgraph. Acceptable str are "fp" for
+			 FP mode, "dwarf" for DWARF mode, "lbr" for LBR mode and
+			 "no" for disable callgraph.
+	  - 'stack_size': user stack size for dwarf mode
 	  Note: If user explicitly sets options which conflict with the params,
 	  the value set by the params will be overridden.
 
tools/perf/util/evsel.c +23 -1
index b313cc7..c196d64 100644
@@ -620,6 +620,7 @@ void perf_evsel__config(struct perf_evsel *evsel, struct record_opts *opts)
 	int track = evsel->tracking;
 	bool per_cpu = opts->target.default_per_cpu && !opts->target.per_thread;
 	bool sample_time = opts->sample_time;
+	bool callgraph = callchain_param.enabled;
 
 	attr->sample_id_all = perf_missing_features.sample_id_all ? 0 : 1;
 	attr->inherit	    = !opts->no_inherit;
@@ -636,6 +637,23 @@ void perf_evsel__config(struct perf_evsel *evsel, struct record_opts *opts)
 			sample_time = false;
 	}
 
+	if (!opts->callgraph_set) {
+		if (attr->sample_type & PERF_SAMPLE_CALLCHAIN) {
+			callgraph = true;
+			if (attr->sample_type & PERF_SAMPLE_STACK_USER) {
+				callchain_param.record_mode = CALLCHAIN_DWARF;
+				if (attr->sample_stack_user)
+					callchain_param.dump_size = attr->sample_stack_user;
+				else
+					callchain_param.dump_size = 8192;
+			} else if (attr->sample_type & PERF_SAMPLE_BRANCH_STACK)
+				callchain_param.record_mode = CALLCHAIN_LBR;
+			else
+				callchain_param.record_mode = CALLCHAIN_FP;
+		} else
+			callgraph = false;
+	}
+
 	/*
 	 * Event parsing doesn't check the availability
 	 * Clear the bit which event parsing may be set.
@@ -650,6 +668,10 @@ void perf_evsel__config(struct perf_evsel *evsel, struct record_opts *opts)
 		evsel->sample_size -= sizeof(u64);
 	}
 
+	attr->sample_type &= ~(PERF_SAMPLE_CALLCHAIN |
+			       PERF_SAMPLE_STACK_USER |
+			       PERF_SAMPLE_BRANCH_STACK);
+
 	perf_evsel__set_sample_bit(evsel, IP);
 	perf_evsel__set_sample_bit(evsel, TID);
 
@@ -715,7 +737,7 @@ void perf_evsel__config(struct perf_evsel *evsel, struct record_opts *opts)
 	if (perf_evsel__is_function_event(evsel))
 		evsel->attr.exclude_callchain_user = 1;
 
-	if (callchain_param.enabled && !evsel->no_aux_samples)
+	if (callgraph && !evsel->no_aux_samples)
 		perf_evsel__config_callgraph(evsel, opts);
 
 	if (opts->sample_intr_regs) {
tools/perf/util/parse-events.c +18 -0
index f1e4aa5..0e4e436 100644
@@ -19,6 +19,7 @@
 #include "thread_map.h"
 #include "cpumap.h"
 #include "asm/bug.h"
+#include "callchain.h"
 
 #define MAX_NAME_LEN 100
 
@@ -605,6 +606,23 @@ do {									   \
 		if (term->val.num == 1)
 			attr->sample_type |= PERF_SAMPLE_TIME;
 		 break;
+	case PARSE_EVENTS__TERM_TYPE_CALLGRAPH:
+		CHECK_TYPE_VAL(STR);
+		if (!strcmp(term->val.str, "fp"))
+			attr->sample_type |= PERF_SAMPLE_CALLCHAIN;
+		else if (!strcmp(term->val.str, "dwarf"))
+			attr->sample_type |= PERF_SAMPLE_CALLCHAIN |
+					     PERF_SAMPLE_STACK_USER;
+		else if (!strcmp(term->val.str, "lbr"))
+			attr->sample_type |= PERF_SAMPLE_CALLCHAIN |
+					     PERF_SAMPLE_BRANCH_STACK;
+		else if (strcmp(term->val.str, "no"))
+			return -EINVAL;
+		break;
+	case PARSE_EVENTS__TERM_TYPE_STACKSIZE:
+		CHECK_TYPE_VAL(NUM);
+		attr->sample_stack_user = term->val.num;
+		break;
 	case PARSE_EVENTS__TERM_TYPE_NAME:
 		CHECK_TYPE_VAL(STR);
 		break;
tools/perf/util/parse-events.h +2 -0
index 0d8cae3..63fb763 100644
@@ -63,6 +63,8 @@ enum {
 	PARSE_EVENTS__TERM_TYPE_SAMPLE_PERIOD,
 	PARSE_EVENTS__TERM_TYPE_BRANCH_SAMPLE_TYPE,
 	PARSE_EVENTS__TERM_TYPE_TIME,
+	PARSE_EVENTS__TERM_TYPE_CALLGRAPH,
+	PARSE_EVENTS__TERM_TYPE_STACKSIZE,
 };
 
 struct parse_events_term {
tools/perf/util/parse-events.l +2 -0
index f542750..16af73b 100644
@@ -184,6 +184,8 @@ name			{ return term(yyscanner, PARSE_EVENTS__TERM_TYPE_NAME); }
 period			{ return term(yyscanner, PARSE_EVENTS__TERM_TYPE_SAMPLE_PERIOD); }
 branch_type		{ return term(yyscanner, PARSE_EVENTS__TERM_TYPE_BRANCH_SAMPLE_TYPE); }
 time			{ return term(yyscanner, PARSE_EVENTS__TERM_TYPE_TIME); }
+callgraph		{ return term(yyscanner, PARSE_EVENTS__TERM_TYPE_CALLGRAPH); }
+stack_size		{ return term(yyscanner, PARSE_EVENTS__TERM_TYPE_TIME); }
 ,			{ return ','; }
 "/"			{ BEGIN(INITIAL); return '/'; }
 {name_minus}		{ return str(yyscanner, PE_NAME); }
Signature · 1 line
1.8.3.1
#9 lore raw

RFC PATCH v2 5/5 perf,tests: Add tests to callgrap and time parse

In reply to [email protected]

From: Kan Liang <[email protected]>

Add tests in tests/parse-events.c to check callgrap and time option

---
 tools/perf/tests/parse-events.c | 28 ++++++++++++++++++++++++++++
 1 file changed, 28 insertions(+)
tools/perf/tests/parse-events.c +28 -0
index d76963f..d6f9447 100644
@@ -471,6 +471,29 @@ static int test__checkevent_pmu_name(struct perf_evlist *evlist)
 	return 0;
 }
 
+static int test__checkevent_pmu_partial_time_callgraph(struct perf_evlist *evlist)
+{
+	struct perf_evsel *evsel = perf_evlist__first(evlist);
+
+	/* cpu/config=1,callgraph=fp,time,period=100000/ */
+	TEST_ASSERT_VAL("wrong number of entries", 2 == evlist->nr_entries);
+	TEST_ASSERT_VAL("wrong type", PERF_TYPE_RAW == evsel->attr.type);
+	TEST_ASSERT_VAL("wrong config",  1 == evsel->attr.config);
+	TEST_ASSERT_VAL("wrong period",  100000 == evsel->attr.sample_period);
+	TEST_ASSERT_VAL("wrong callgraph",  PERF_SAMPLE_CALLCHAIN & evsel->attr.sample_type);
+	TEST_ASSERT_VAL("wrong time",  PERF_SAMPLE_TIME & evsel->attr.sample_type);
+
+	/* cpu/config=2,callgraph=no,time=0,period=2000/ */
+	evsel = perf_evsel__next(evsel);
+	TEST_ASSERT_VAL("wrong type", PERF_TYPE_RAW == evsel->attr.type);
+	TEST_ASSERT_VAL("wrong config",  2 == evsel->attr.config);
+	TEST_ASSERT_VAL("wrong period",  2000 == evsel->attr.sample_period);
+	TEST_ASSERT_VAL("wrong callgraph",  !(PERF_SAMPLE_CALLCHAIN & evsel->attr.sample_type));
+	TEST_ASSERT_VAL("wrong time",  !(PERF_SAMPLE_TIME & evsel->attr.sample_type));
+
+	return 0;
+}
+
 static int test__checkevent_pmu_events(struct perf_evlist *evlist)
 {
 	struct perf_evsel *evsel = perf_evlist__first(evlist);
@@ -1547,6 +1570,11 @@ static struct evlist_test test__events_pmu[] = {
 		.check = test__checkevent_pmu_name,
 		.id    = 1,
 	},
+	{
+		.name  = "cpu/config=1,callgraph=fp,time,period=100000/,cpu/config=2,callgraph=no,time=0,period=2000/",
+		.check = test__checkevent_pmu_partial_time_callgraph,
+		.id    = 2,
+	},
 };
 
 struct terms_test {
Signature · 1 line
1.8.3.1