Skip to contentSkip to conversation

Python improvements for a real use of parse_events

PATCH Superseded Ian Rogers 3 participants 23 messages Jan 9, 2025 – Feb 27, 2025 1 message not in the archive perf events

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore

Ian Rogers

author reviewer
#0 lore raw

cover Python improvements for a real use of parse_events

While parse_events access in python was added, it wasn't used by any python script. In enabling this for the tracepoint.py script a number of latent bugs and necessary improvements were discovered.

Ian Rogers (11):
perf debug: Avoid stack overflow in recursive error message perf evlist: Add success path to evlist__create_syswide_maps perf evsel: tp_format accessing improvements
perf python: Add evlist enable and disable methods
perf python: Add member access to a number of evsel variables perf python: Add optional cpus and threads arguments to parse_events perf python: Update ungrouped evsel leader in clone
perf python: Avoid duplicated code in get_tracepoint_field perf python: Add evlist all_cpus accessor
perf python: Add evlist.config to set up record options perf python tracepoint: Switch to using parse_events

 tools/perf/python/tracepoint.py |  23 +++---
 tools/perf/util/debug.c         |   2 +-
 tools/perf/util/evlist.c        |  13 ++--
 tools/perf/util/evsel.c         |  16 +++-
 tools/perf/util/python.c        | 127 ++++++++++++++++++++++++++++----
 5 files changed, 145 insertions(+), 36 deletions(-)
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
author reviewer
#1 lore raw

PATCH 1/11 perf debug: Avoid stack overflow in recursive error message

In debug_file, pr_warning_once is called on error. As that function calls debug_file the function will yield a stack overflow. Switch the location of the call so the recursion is avoided.

---
 tools/perf/util/debug.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
tools/perf/util/debug.c +1 -1
index 995f6bb05b5f..f9ef7d045c92 100644
@@ -46,8 +46,8 @@ int debug_type_profile;
 FILE *debug_file(void)
 {
 	if (!_debug_file) {
-		pr_warning_once("debug_file not set");
 		debug_set_file(stderr);
+		pr_warning_once("debug_file not set");
 	}
 	return _debug_file;
 }
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
#2 lore raw

Re: [PATCH v1 01/11] perf debug: Avoid stack overflow in recursive error message

Hi Ian,

Quoted 4 lines · Ian Rogers

In debug_file, pr_warning_once is called on error. As that function calls debug_file the function will yield a stack overflow. Switch the location of the call so the recursion is avoided.

May I ask how it created the recursion? Too ignorant to understand sorry...

Thanks,
Howard

author reviewer
#3 lore raw

Re: [PATCH v1 01/11] perf debug: Avoid stack overflow in recursive error message

Quoted 10 lines · Howard Chu

Hi Ian,

Quoted 4 lines · Ian Rogers

In debug_file, pr_warning_once is called on error. As that function calls debug_file the function will yield a stack overflow. Switch the location of the call so the recursion is avoided.

May I ask how it created the recursion? Too ignorant to understand sorry...

If _debug_file is NULL then debug_file calls pr_warning_once which calls debug_file which calls pr_warning_once again, etc ... :-)

Thanks,
Ian

Quoted 2 lines

Thanks,
Howard

#4 lore raw

Re: [PATCH v1 01/11] perf debug: Avoid stack overflow in recursive error message

Hello,

Quoted 7 lines · Ian Rogers
Quoted 5 lines · Howard Chu

Hi Ian,

Quoted 1 line · Ian Rogers
Quoted 8 lines
Quoted 5 lines
Quoted 3 lines

In debug_file, pr_warning_once is called on error. As that function calls debug_file the function will yield a stack overflow. Switch the location of the call so the recursion is avoided.

May I ask how it created the recursion? Too ignorant to understand sorry...

If _debug_file is NULL then debug_file calls pr_warning_once which calls debug_file which calls pr_warning_once again, etc ... :-)

Got it,

pr_warning()
eprintf()
veprintf()
debug_file()

ack.

Thanks,
Howard

Quoted 6 lines

Thanks,
Ian

Quoted 2 lines

Thanks,
Howard

author reviewer
#5 lore raw

PATCH 2/11 perf evlist: Add success path to evlist__create_syswide_maps

In reply to Ian Rogers

Over various refactorings evlist__create_syswide_maps has been made to only ever return with -ENOMEM. Fix this so that when perf_evlist__set_maps is successfully called, 0 is returned.

---
 tools/perf/util/evlist.c | 13 ++++++-------
 1 file changed, 6 insertions(+), 7 deletions(-)
tools/perf/util/evlist.c +6 -7
index f0dd174e2deb..633df7d9204c 100644
@@ -1373,19 +1373,18 @@ static int evlist__create_syswide_maps(struct evlist *evlist)
 	 */
 	cpus = perf_cpu_map__new_online_cpus();
 	if (!cpus)
-		goto out;
+		return -ENOMEM;
 
 	threads = perf_thread_map__new_dummy();
-	if (!threads)
-		goto out_put;
+	if (!threads) {
+		perf_cpu_map__put(cpus);
+		return -ENOMEM;
+	}
 
 	perf_evlist__set_maps(&evlist->core, cpus, threads);
-
 	perf_thread_map__put(threads);
-out_put:
 	perf_cpu_map__put(cpus);
-out:
-	return -ENOMEM;
+	return 0;
 }
 
 int evlist__open(struct evlist *evlist)
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
author reviewer
#6 lore raw

PATCH 3/11 perf evsel: tp_format accessing improvements

In reply to Ian Rogers

Ensure evsel__clone copies the tp_sys and tp_name variables. In evsel__tp_format, if tp_sys isn't set, use the config value to find the tp_format. This succeeds in python code where pyrf__tracepoint has already found the format.

---
 tools/perf/util/evsel.c | 16 +++++++++++++++-
 1 file changed, 15 insertions(+), 1 deletion(-)
tools/perf/util/evsel.c +15 -1
index 697428efa644..5324a8916da1 100644
@@ -511,6 +511,16 @@ struct evsel *evsel__clone(struct evsel *dest, struct evsel *orig)
 	}
 	evsel->cgrp = cgroup__get(orig->cgrp);
 #ifdef HAVE_LIBTRACEEVENT
+	if (orig->tp_sys) {
+		evsel->tp_sys = strdup(orig->tp_sys);
+		if (evsel->tp_sys == NULL)
+			goto out_err;
+	}
+	if (orig->tp_name) {
+		evsel->tp_name = strdup(orig->tp_name);
+		if (evsel->tp_name == NULL)
+			goto out_err;
+	}
 	evsel->tp_format = orig->tp_format;
 #endif
 	evsel->handler = orig->handler;
@@ -634,7 +644,11 @@ struct tep_event *evsel__tp_format(struct evsel *evsel)
 	if (evsel->core.attr.type != PERF_TYPE_TRACEPOINT)
 		return NULL;
 
-	tp_format = trace_event__tp_format(evsel->tp_sys, evsel->tp_name);
+	if (!evsel->tp_sys)
+		tp_format = trace_event__tp_format_id(evsel->core.attr.config);
+	else
+		tp_format = trace_event__tp_format(evsel->tp_sys, evsel->tp_name);
+
 	if (IS_ERR(tp_format)) {
 		int err = -PTR_ERR(evsel->tp_format);
 
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
#7 lore raw

Re: [PATCH v1 03/11] perf evsel: tp_format accessing improvements

Hello,

Thanks for fixing another bug! :) The freeing in evsel__exit() is present as well.

Thanks,
Howard

author reviewer
#8 lore raw

PATCH 4/11 perf python: Add evlist enable and disable methods

In reply to Ian Rogers

By default the evsels from parse_events will be disabled. Add access to the evlist functions so they can be enabled/disabled.

---
 tools/perf/util/python.c | 26 ++++++++++++++++++++++++++
 1 file changed, 26 insertions(+)
tools/perf/util/python.c +26 -0
index b4bc57859f73..8f68320130c4 100644
@@ -1019,6 +1019,20 @@ static PyObject *pyrf_evlist__open(struct pyrf_evlist *pevlist,
 	return Py_None;
 }
 
+static PyObject *pyrf_evlist__disable(struct pyrf_evlist *pevlist)
+{
+	evlist__disable(&pevlist->evlist);
+	Py_INCREF(Py_None);
+	return Py_None;
+}
+
+static PyObject *pyrf_evlist__enable(struct pyrf_evlist *pevlist)
+{
+	evlist__enable(&pevlist->evlist);
+	Py_INCREF(Py_None);
+	return Py_None;
+}
+
 static PyMethodDef pyrf_evlist__methods[] = {
 	{
 		.ml_name  = "mmap",
@@ -1056,6 +1070,18 @@ static PyMethodDef pyrf_evlist__methods[] = {
 		.ml_flags = METH_VARARGS | METH_KEYWORDS,
 		.ml_doc	  = PyDoc_STR("reads an event.")
 	},
+	{
+		.ml_name  = "disable",
+		.ml_meth  = (PyCFunction)pyrf_evlist__disable,
+		.ml_flags = METH_NOARGS,
+		.ml_doc	  = PyDoc_STR("Disable the evsels in the evlist.")
+	},
+	{
+		.ml_name  = "enable",
+		.ml_meth  = (PyCFunction)pyrf_evlist__enable,
+		.ml_flags = METH_NOARGS,
+		.ml_doc	  = PyDoc_STR("Enable the evsels in the evlist.")
+	},
 	{ .ml_name = NULL, }
 };
 
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
author reviewer
#9 lore raw

PATCH 5/11 perf python: Add member access to a number of evsel variables

In reply to Ian Rogers

Most variables are part of the perf_event_attr, so that they may be queried and modified.

---
 tools/perf/util/python.c | 23 +++++++++++++++++++++++
 1 file changed, 23 insertions(+)
tools/perf/util/python.c +23 -0
index 8f68320130c4..6a7042fd2569 100644
@@ -802,6 +802,28 @@ static PyMethodDef pyrf_evsel__methods[] = {
 	{ .ml_name = NULL, }
 };
 
+#define evsel_member_def(member, ptype, help) \
+	{ #member, ptype, \
+	  offsetof(struct pyrf_evsel, evsel.member), \
+	  0, help }
+
+#define evsel_attr_member_def(member, ptype, help) \
+	{ #member, ptype, \
+	  offsetof(struct pyrf_evsel, evsel.core.attr.member), \
+	  0, help }
+
+static PyMemberDef pyrf_evsel__members[] = {
+	evsel_member_def(tracking, T_BOOL, "tracking event."),
+	evsel_attr_member_def(type, T_UINT, "attribute type."),
+	evsel_attr_member_def(size, T_UINT, "attribute size."),
+	evsel_attr_member_def(config, T_ULONGLONG, "attribute config."),
+	evsel_attr_member_def(sample_period, T_ULONGLONG, "attribute sample_period."),
+	evsel_attr_member_def(sample_type, T_ULONGLONG, "attribute sample_type."),
+	evsel_attr_member_def(read_format, T_ULONGLONG, "attribute read_format."),
+	evsel_attr_member_def(wakeup_events, T_UINT, "attribute wakeup_events."),
+	{ .name = NULL, },
+};
+
 static const char pyrf_evsel__doc[] = PyDoc_STR("perf event selector list object.");
 
 static PyTypeObject pyrf_evsel__type = {
@@ -811,6 +833,7 @@ static PyTypeObject pyrf_evsel__type = {
 	.tp_dealloc	= (destructor)pyrf_evsel__delete,
 	.tp_flags	= Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE,
 	.tp_doc		= pyrf_evsel__doc,
+	.tp_members	= pyrf_evsel__members,
 	.tp_methods	= pyrf_evsel__methods,
 	.tp_init	= (initproc)pyrf_evsel__init,
 	.tp_str         = pyrf_evsel__str,
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
author reviewer
#10 lore raw

PATCH 6/11 perf python: Add optional cpus and threads arguments to parse_events

In reply to Ian Rogers

Used for the evlist initialization.

---
 tools/perf/util/python.c | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)
tools/perf/util/python.c +8 -2
index 6a7042fd2569..45ee6479e4dd 100644
@@ -1330,12 +1330,18 @@ static PyObject *pyrf__parse_events(PyObject *self, PyObject *args)
 	struct evlist evlist = {};
 	struct parse_events_error err;
 	PyObject *result;
+	PyObject *pcpus = NULL, *pthreads = NULL;
+	struct perf_cpu_map *cpus;
+	struct perf_thread_map *threads;
 
-	if (!PyArg_ParseTuple(args, "s", &input))
+	if (!PyArg_ParseTuple(args, "s|OO", &input, &pcpus, &pthreads))
 		return NULL;
 
+	threads = pthreads ? ((struct pyrf_thread_map *)pthreads)->threads : NULL;
+	cpus = pcpus ? ((struct pyrf_cpu_map *)pcpus)->cpus : NULL;
+
 	parse_events_error__init(&err);
-	evlist__init(&evlist, NULL, NULL);
+	evlist__init(&evlist, cpus, threads);
 	if (parse_events(&evlist, input, &err)) {
 		parse_events_error__print(&err, input);
 		PyErr_SetFromErrno(PyExc_OSError);
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
author reviewer
#11 lore raw

PATCH 7/11 perf python: Update ungrouped evsel leader in clone

In reply to Ian Rogers

evsels are cloned in the python code as they form part of the Python object pyrf_evsel. The cloning doesn't update the evsel's leader, do this for the case of an evsel being ungrouped.

---
 tools/perf/util/python.c | 2 ++
 1 file changed, 2 insertions(+)
tools/perf/util/python.c +2 -0
index 45ee6479e4dd..9ffe1b633027 100644
@@ -1303,6 +1303,8 @@ static PyObject *pyrf_evsel__from_evsel(struct evsel *evsel)
 	evsel__init(&pevsel->evsel, &evsel->core.attr, evsel->core.idx);
 
 	evsel__clone(&pevsel->evsel, evsel);
+	if (evsel__is_group_leader(evsel))
+		evsel__set_leader(&pevsel->evsel, &pevsel->evsel);
 	return (PyObject *)pevsel;
 }
 
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
author reviewer
#12 lore raw

PATCH 8/11 perf python: Avoid duplicated code in get_tracepoint_field

In reply to Ian Rogers

The code replicates computations done in evsel__tp_format, reuse evsel__tp_format to simplify the python C code.

---
 tools/perf/util/python.c | 17 ++++-------------
 1 file changed, 4 insertions(+), 13 deletions(-)
tools/perf/util/python.c +4 -13
index 9ffe1b633027..d1ce3b6b5bf1 100644
@@ -336,23 +336,14 @@ get_tracepoint_field(struct pyrf_event *pevent, PyObject *attr_name)
 {
 	const char *str = _PyUnicode_AsString(PyObject_Str(attr_name));
 	struct evsel *evsel = pevent->evsel;
+	struct tep_event *tp_format = evsel__tp_format(evsel);
 	struct tep_format_field *field;
 
-	if (!evsel->tp_format) {
-		struct tep_event *tp_format;
-
-		tp_format = trace_event__tp_format_id(evsel->core.attr.config);
-		if (IS_ERR_OR_NULL(tp_format))
-			return NULL;
-
-		evsel->tp_format = tp_format;
-	}
-
-	field = tep_find_any_field(evsel->tp_format, str);
-	if (!field)
+	if (IS_ERR_OR_NULL(tp_format))
 		return NULL;
 
-	return tracepoint_field(pevent, field);
+	field = tep_find_any_field(tp_format, str);
+	return field ? tracepoint_field(pevent, field) : NULL;
 }
 #endif /* HAVE_LIBTRACEEVENT */
 
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
author reviewer
#13 lore raw

PATCH 9/11 perf python: Add evlist all_cpus accessor

In reply to Ian Rogers

Add a means to get the reference counted all_cpus CPU map from an evlist in its python form.

---
 tools/perf/util/python.c | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)
tools/perf/util/python.c +16 -0
index d1ce3b6b5bf1..1e8ab5291637 100644
@@ -865,6 +865,16 @@ static void pyrf_evlist__delete(struct pyrf_evlist *pevlist)
 	Py_TYPE(pevlist)->tp_free((PyObject*)pevlist);
 }
 
+static PyObject *pyrf_evlist__all_cpus(struct pyrf_evlist *pevlist)
+{
+	struct pyrf_cpu_map *pcpu_map = PyObject_New(struct pyrf_cpu_map, &pyrf_cpu_map__type);
+
+	if (pcpu_map)
+		pcpu_map->cpus = perf_cpu_map__get(pevlist->evlist.core.all_cpus);
+
+	return (PyObject *)pcpu_map;
+}
+
 static PyObject *pyrf_evlist__mmap(struct pyrf_evlist *pevlist,
 				   PyObject *args, PyObject *kwargs)
 {
@@ -1048,6 +1058,12 @@ static PyObject *pyrf_evlist__enable(struct pyrf_evlist *pevlist)
 }
 
 static PyMethodDef pyrf_evlist__methods[] = {
+	{
+		.ml_name  = "all_cpus",
+		.ml_meth  = (PyCFunction)pyrf_evlist__all_cpus,
+		.ml_flags = METH_NOARGS,
+		.ml_doc	  = PyDoc_STR("CPU map union of all evsel CPU maps.")
+	},
 	{
 		.ml_name  = "mmap",
 		.ml_meth  = (PyCFunction)pyrf_evlist__mmap,
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
author reviewer
#14 lore raw

PATCH 10/11 perf python: Add evlist.config to set up record options

In reply to Ian Rogers

Add access to evlist__config that is used to configure an evlist with record options.

---
 tools/perf/util/python.c | 33 +++++++++++++++++++++++++++++++++
 1 file changed, 33 insertions(+)
tools/perf/util/python.c +33 -0
index 1e8ab5291637..38a7e16e09c9 100644
@@ -9,10 +9,12 @@
 #include <event-parse.h>
 #endif
 #include <perf/mmap.h>
+#include "callchain.h"
 #include "evlist.h"
 #include "evsel.h"
 #include "event.h"
 #include "print_binary.h"
+#include "record.h"
 #include "strbuf.h"
 #include "thread_map.h"
 #include "trace-event.h"
@@ -1043,6 +1045,31 @@ static PyObject *pyrf_evlist__open(struct pyrf_evlist *pevlist,
 	return Py_None;
 }
 
+static PyObject *pyrf_evlist__config(struct pyrf_evlist *pevlist)
+{
+	struct record_opts opts = {
+		.sample_time	     = true,
+		.mmap_pages	     = UINT_MAX,
+		.user_freq	     = UINT_MAX,
+		.user_interval	     = ULLONG_MAX,
+		.freq		     = 4000,
+		.target		     = {
+			.uses_mmap   = true,
+			.default_per_cpu = true,
+		},
+		.nr_threads_synthesize = 1,
+		.ctl_fd              = -1,
+		.ctl_fd_ack          = -1,
+			.no_buffering = true,
+			.no_inherit = true,
+	};
+	struct evlist *evlist = &pevlist->evlist;
+
+	evlist__config(evlist, &opts, &callchain_param);
+	Py_INCREF(Py_None);
+	return Py_None;
+}
+
 static PyObject *pyrf_evlist__disable(struct pyrf_evlist *pevlist)
 {
 	evlist__disable(&pevlist->evlist);
@@ -1100,6 +1127,12 @@ static PyMethodDef pyrf_evlist__methods[] = {
 		.ml_flags = METH_VARARGS | METH_KEYWORDS,
 		.ml_doc	  = PyDoc_STR("reads an event.")
 	},
+	{
+		.ml_name  = "config",
+		.ml_meth  = (PyCFunction)pyrf_evlist__config,
+		.ml_flags = METH_NOARGS,
+		.ml_doc	  = PyDoc_STR("Apply default record options to the evlist.")
+	},
 	{
 		.ml_name  = "disable",
 		.ml_meth  = (PyCFunction)pyrf_evlist__disable,
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
#15 lore raw

Re: [PATCH v1 10/11] perf python: Add evlist.config to set up record options

Hello,

Quoted 9 lines · Ian Rogers

Add access to evlist__config that is used to configure an evlist with record options.

---
 tools/perf/util/python.c | 33 +++++++++++++++++++++++++++++++++
 1 file changed, 33 insertions(+)

<SNIP>

tools/perf/util/python.c static PyObject *pyrf_evlist__open(struct pyrf_evlist *pevlist, · line 1064

Quoted 11 lines
tools/perf/util/python.c +11 -0
+               .user_interval       = ULLONG_MAX,
+               .freq                = 4000,
+               .target              = {
+                       .uses_mmap   = true,
+                       .default_per_cpu = true,
+               },
+               .nr_threads_synthesize = 1,
+               .ctl_fd              = -1,
+               .ctl_fd_ack          = -1,
+                       .no_buffering = true,
+                       .no_inherit = true,

Are these indentations intentional?

Thanks,
Howard

author reviewer
#16 lore raw

Re: [PATCH v1 10/11] perf python: Add evlist.config to set up record options

Quoted 28 lines · Howard Chu

Hello,

Quoted 9 lines · Ian Rogers

Add access to evlist__config that is used to configure an evlist with record options.

---
 tools/perf/util/python.c | 33 +++++++++++++++++++++++++++++++++
 1 file changed, 33 insertions(+)

<SNIP>

tools/perf/util/python.c static PyObject *pyrf_evlist__open(struct pyrf_evlist *pevlist, · line 1064

Quoted 11 lines
tools/perf/util/python.c +11 -0
+               .user_interval       = ULLONG_MAX,
+               .freq                = 4000,
+               .target              = {
+                       .uses_mmap   = true,
+                       .default_per_cpu = true,
+               },
+               .nr_threads_synthesize = 1,
+               .ctl_fd              = -1,
+               .ctl_fd_ack          = -1,
+                       .no_buffering = true,
+                       .no_inherit = true,
Quoted 1 line

Are these indentations intentional?

Nope, not sure what happened. I'll fix it in v2.

Thanks and thanks for the tags!
Ian

Quoted 3 lines

Thanks,
Howard

author reviewer
#17 lore raw

PATCH 11/11 perf python tracepoint: Switch to using parse_events

In reply to Ian Rogers

Rather than manually configuring an evsel, switch to using parse_events for greater commonality with the rest of the perf code.

---
 tools/perf/python/tracepoint.py | 23 +++++++++++------------
 1 file changed, 11 insertions(+), 12 deletions(-)
tools/perf/python/tracepoint.py +11 -12
index bba68a6d4515..38b2b6d11f64 100755
@@ -5,24 +5,23 @@
 
 import perf
 
-class tracepoint(perf.evsel):
-    def __init__(self, sys, name):
-        config = perf.tracepoint(sys, name)
-        perf.evsel.__init__(self,
-                            type   = perf.TYPE_TRACEPOINT,
-                            config = config,
-                            freq = 0, sample_period = 1, wakeup_events = 1,
-                            sample_type = perf.SAMPLE_PERIOD | perf.SAMPLE_TID | perf.SAMPLE_CPU | perf.SAMPLE_RAW | perf.SAMPLE_TIME)
-
 def main():
-    tp      = tracepoint("sched", "sched_switch")
     cpus    = perf.cpu_map()
     threads = perf.thread_map(-1)
+    evlist = perf.parse_events("sched:sched_switch", cpus, threads)
+    # Disable tracking of mmaps and similar that are unnecessary.
+    for ev in evlist:
+        ev.tracking = False
+    # Configure evsels with default record options.
+    evlist.config()
+    # Simplify the sample_type and read_format of evsels
+    for ev in evlist:
+        ev.sample_type = ev.sample_type & ~perf.SAMPLE_IP
+        ev.read_format = 0
 
-    evlist = perf.evlist(cpus, threads)
-    evlist.add(tp)
     evlist.open()
     evlist.mmap()
+    evlist.enable();
 
     while True:
         evlist.poll(timeout = -1)
Signature · 1 line
2.47.1.613.gc27f4b7a9f-goog
author reviewer
#18 lore raw
In reply to Ian Rogers
Quoted 4 lines · Ian Rogers

While parse_events access in python was added, it wasn't used by any python script. In enabling this for the tracepoint.py script a number of latent bugs and necessary improvements were discovered.

Ping.

Thanks,
Ian

author reviewer
#19 lore raw
Quoted 8 lines · Ian Rogers
Quoted 4 lines · Ian Rogers

While parse_events access in python was added, it wasn't used by any python script. In enabling this for the tracepoint.py script a number of latent bugs and necessary improvements were discovered.

Ping.

Ping.

Thanks,
Ian

maintainer
#20 lore raw

Hi Ian,

Quoted 11 lines · Ian Rogers
Quoted 8 lines · Ian Rogers
Quoted 4 lines · Ian Rogers

While parse_events access in python was added, it wasn't used by any python script. In enabling this for the tracepoint.py script a number of latent bugs and necessary improvements were discovered.

Ping.

Ping.

Thanks for the reminder. I'll review pending patches this week.

Arnaldo, can you please help reviewing this series?

Thanks,
Namhyung

A message that is not in the archive (sent off-list or not yet received): CA+JHD90PDGA-Zbu6hUZ1DZQbSD71FBPRbBdGd1dWCxhi3q6Q1g@mail.gmail.com

author reviewer
#22 lore raw
Quoted 18 lines · Arnaldo Carvalho de Melo
Quoted 14 lines · Namhyung Kim

Hi Ian,

Quoted 10 lines · Ian Rogers
Quoted 8 lines · Ian Rogers
Quoted 4 lines · Ian Rogers

While parse_events access in python was added, it wasn't used by any python script. In enabling this for the tracepoint.py script a number of latent bugs and necessary improvements were discovered.

Ping.

Quoted 8 lines
Quoted 5 lines
Quoted 1 line

Ping.

Thanks for the reminder. I'll review pending patches this week.

Arnaldo, can you please help reviewing this series?

Sure, I'm getting back to doing that, more slowly than I'd like this time. :-/

Ping.

Thanks,
Ian

Reviewed-by on the v1 series
#23 lore raw
In reply to Ian Rogers

Hello Ian,

Sorry for the late reply—I just recovered from a nasty flu.

perf $ git log --oneline | head -n 11
0c1d18794589 perf python tracepoint: Switch to using parse_events 94dc6574ff24 perf python: Add evlist.config to set up record options ebc33f96a37d perf python: Add evlist all_cpus accessor c1ff82e366ee perf python: Avoid duplicated code in get_tracepoint_field 5293474151de perf python: Update ungrouped evsel leader in clone 47f66c784109 perf python: Add optional cpus and threads arguments to parse_events
d4f592ec635e perf python: Add member access to a number of evsel variables f4a3006a2c87 perf python: Add evlist enable and disable methods 2e040a422d0a perf evsel: tp_format accessing improvements d79fd699f99a perf evlist: Add success path to evlist__create_syswide_maps 054b63706989 perf debug: Avoid stack overflow in recursive error message

perf $ sudo python python/tracepoint.py
libperf: mmap_per_cpu: nr cpu values 4 nr threads 1
libperf: idx 0: mmapping fd 3
libperf: idx 1: mmapping fd 4
libperf: idx 2: mmapping fd 5
libperf: idx 3: mmapping fd 6
time 35749194145330 prev_comm=python prev_pid=233668 prev_prio=120 prev_state=0x100 ==> next_comm=migration/0 next_pid=20 next_prio=0 time 35749196508397 prev_comm=python prev_pid=233668 prev_prio=120 prev_state=0x100 ==> next_comm=migration/1 next_pid=25 next_prio=0 time 35749196575872 prev_comm=python prev_pid=233668 prev_prio=120 prev_state=0x100 ==> next_comm=migration/2 next_pid=31 next_prio=0 time 35749194158623 prev_comm=migration/0 prev_pid=20 prev_prio=0 prev_state=0x1 ==> next_comm=swapper/0 next_pid=0 next_prio=120

Worked very well.

Quoted 5 lines · Ian Rogers

While parse_events access in python was added, it wasn't used by any python script. In enabling this for the tracepoint.py script a number of latent bugs and necessary improvements were discovered.

Quoted 12 lines

Ian Rogers (11):
perf debug: Avoid stack overflow in recursive error message perf evlist: Add success path to evlist__create_syswide_maps perf evsel: tp_format accessing improvements
perf python: Add evlist enable and disable methods
perf python: Add member access to a number of evsel variables perf python: Add optional cpus and threads arguments to parse_events perf python: Update ungrouped evsel leader in clone
perf python: Avoid duplicated code in get_tracepoint_field perf python: Add evlist all_cpus accessor
perf python: Add evlist.config to set up record options perf python tracepoint: Switch to using parse_events

Not an expert in this but:

Thanks,
Howard

Quoted 11 lines
 tools/perf/python/tracepoint.py |  23 +++---
 tools/perf/util/debug.c         |   2 +-
 tools/perf/util/evlist.c        |  13 ++--
 tools/perf/util/evsel.c         |  16 +++-
 tools/perf/util/python.c        | 127 ++++++++++++++++++++++++++++----
 5 files changed, 145 insertions(+), 36 deletions(-)

--
2.47.1.613.gc27f4b7a9f-goog