perf: Reorganize scripting support
Subsystems (4)
Ian Rogers
reviewerPATCH v14 11/19 perf python: Refactor and add accessors to sample event
In reply to Ian RogersAdd common evsel field for events and move sample specific fields to only be present in sample events. Add accessors for sample events. Ensure offsets are within the bounds of the event. Allocate just enough memory for the copied event, don't make the maximum event size each time.
---
v12:
- Used copy_size to bound max_len for NUL termination to fix an out-of-bounds write in pyrf_event__new. Also set pevent->event.header.size = copy_size.
v5:
1. Fix Uninitialized Memory: Restore zero-initialization of pevent->sample in pyrf_event__new() to prevent wild free crashes on error paths.
v6:
- Refactored pyrf_event__new to take evsel and session, and use dynamic allocation based on event size. Updated callers.
v8:
- Ensure events are properly deallocated.
v11:
- Use perf_event__too_small for event size validation, sharing logic with the session parser instead of replicating validation checks.
--- tools/perf/util/python.c | 506 +++++++++++++++++++++++++++++++------- tools/perf/util/session.c | 2 +- tools/perf/util/session.h | 2 + 3 files changed, 425 insertions(+), 85 deletions(-)
tools/perf/util/python.c +422 -84
@@ -13,21 +13,28 @@#include <perf/mmap.h> #include <structmember.h> +#include "addr_location.h" +#include "build-id.h" #include "callchain.h" #include "comm.h" #include "counts.h" #include "data.h" #include "debug.h" +#include "dso.h" #include "event.h" #include "evlist.h" #include "evsel.h" #include "expr.h" +#include "map.h" #include "metricgroup.h" #include "mmap.h" #include "pmus.h" #include "print_binary.h" #include "record.h" +#include "sample.h" #include "session.h" +#include "srccode.h" +#include "srcline.h" #include "strbuf.h" #include "symbol.h" #include "stat.h"
@@ -37,7 +44,6 @@#include "tool.h" #include "tp_pmu.h" #include "trace-event.h" -#include "util/sample.h" #ifdef HAVE_LIBTRACEEVENT #include <event-parse.h>
@@ -45,6 +51,8 @@PyMODINIT_FUNC PyInit_perf (void ); +static PyObject *pyrf_evsel__from_evsel (struct evsel *evsel); +#define member_def (type, member, ptype, help) \ { #member, ptype, \ offsetof(struct pyrf_event, event) + offsetof(struct type, member), \
@@ -73,21 +81,53 @@ PyMODINIT_FUNC PyInit_perf(void);struct pyrf_event {PyObject_HEAD +/** @sample: The parsed sample from the event. */ struct perf_sample sample; -union perf_event event; +/** @al: The address location from machine__resolve, lazily computed. */ +struct addr_location al; +/** @al_resolved: True when machine__resolve been called. */ +bool al_resolved; +/** @event: The underlying perf_event that may be in a file or ring buffer. */ +union perf_event event; };#define sample_members \ - sample_member_def(sample_ip, ip, T_ULONGLONG, "event ip"), \ sample_member_def(sample_pid, pid, T_INT, "event pid"), \ sample_member_def(sample_tid, tid, T_INT, "event tid"), \ sample_member_def(sample_time, time, T_ULONGLONG, "event timestamp"), \ - sample_member_def(sample_addr, addr, T_ULONGLONG, "event addr"), \ sample_member_def(sample_id, id, T_ULONGLONG, "event id"), \ sample_member_def(sample_stream_id, stream_id, T_ULONGLONG, "event stream id"), \ sample_member_def(sample_period, period, T_ULONGLONG, "event period"), \ sample_member_def(sample_cpu, cpu, T_UINT, "event cpu"), +static PyObject *pyrf_event__get_evsel (PyObject *self,void *closure__maybe_unused ) +{ +struct pyrf_event *pevent = (void *)self; + +if (!pevent->sample.evsel) + Py_RETURN_NONE; + +return pyrf_evsel__from_evsel (pevent->sample.evsel); +} + +static PyGetSetDef pyrf_event__getset[] = { + { + .name ="evsel" , + .get = pyrf_event__get_evsel, + .set =NULL , + .doc ="tracking event." , + }, + { .name =NULL , }, +}; + +static void pyrf_event__delete (struct pyrf_event *pevent) +{ +if (pevent->al_resolved) +addr_location__exit (&pevent->al); +perf_sample__exit (&pevent->sample); +Py_TYPE (pevent)->tp_free ((PyObject *)pevent); +} +static const char pyrf_mmap_event__doc[] =PyDoc_STR ("perf mmap event object." );static PyMemberDef pyrf_mmap_event__members[] = {
@@ -126,9 +166,11 @@ static PyTypeObject pyrf_mmap_event__type = {PyVarObject_HEAD_INIT (NULL ,0 ) .tp_name = "perf.mmap_event", .tp_basicsize = sizeof(struct pyrf_event), + .tp_dealloc = (destructor)pyrf_event__delete, .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, .tp_doc = pyrf_mmap_event__doc, .tp_members = pyrf_mmap_event__members, + .tp_getset = pyrf_event__getset, .tp_repr = (reprfunc)pyrf_mmap_event__repr, };
@@ -161,9 +203,11 @@ static PyTypeObject pyrf_task_event__type = {PyVarObject_HEAD_INIT (NULL ,0 ) .tp_name = "perf.task_event", .tp_basicsize = sizeof(struct pyrf_event), + .tp_dealloc = (destructor)pyrf_event__delete, .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, .tp_doc = pyrf_task_event__doc, .tp_members = pyrf_task_event__members, + .tp_getset = pyrf_event__getset, .tp_repr = (reprfunc)pyrf_task_event__repr, };
@@ -190,9 +234,11 @@ static PyTypeObject pyrf_comm_event__type = {PyVarObject_HEAD_INIT (NULL ,0 ) .tp_name = "perf.comm_event", .tp_basicsize = sizeof(struct pyrf_event), + .tp_dealloc = (destructor)pyrf_event__delete, .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, .tp_doc = pyrf_comm_event__doc, .tp_members = pyrf_comm_event__members, + .tp_getset = pyrf_event__getset, .tp_repr = (reprfunc)pyrf_comm_event__repr, };
@@ -222,9 +268,11 @@ static PyTypeObject pyrf_throttle_event__type = {PyVarObject_HEAD_INIT (NULL ,0 ) .tp_name = "perf.throttle_event", .tp_basicsize = sizeof(struct pyrf_event), + .tp_dealloc = (destructor)pyrf_event__delete, .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, .tp_doc = pyrf_throttle_event__doc, .tp_members = pyrf_throttle_event__members, + .tp_getset = pyrf_event__getset, .tp_repr = (reprfunc)pyrf_throttle_event__repr, };
@@ -257,9 +305,11 @@ static PyTypeObject pyrf_lost_event__type = {PyVarObject_HEAD_INIT (NULL ,0 ) .tp_name = "perf.lost_event", .tp_basicsize = sizeof(struct pyrf_event), + .tp_dealloc = (destructor)pyrf_event__delete, .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, .tp_doc = pyrf_lost_event__doc, .tp_members = pyrf_lost_event__members, + .tp_getset = pyrf_event__getset, .tp_repr = (reprfunc)pyrf_lost_event__repr, };
@@ -287,9 +337,11 @@ static PyTypeObject pyrf_read_event__type = {PyVarObject_HEAD_INIT (NULL ,0 ) .tp_name = "perf.read_event", .tp_basicsize = sizeof(struct pyrf_event), + .tp_dealloc = (destructor)pyrf_event__delete, .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, .tp_doc = pyrf_read_event__doc, .tp_members = pyrf_read_event__members, + .tp_getset = pyrf_event__getset, .tp_repr = (reprfunc)pyrf_read_event__repr, };
@@ -297,16 +349,17 @@ static const char pyrf_sample_event__doc[] = PyDoc_STR("perf sample event objectstatic PyMemberDef pyrf_sample_event__members[] = { sample_members +sample_member_def (sample_ip, ip,T_ULONGLONG ,"event ip" ), +sample_member_def (sample_addr, addr,T_ULONGLONG ,"event addr" ), +sample_member_def (sample_phys_addr, phys_addr,T_ULONGLONG ,"event physical addr" ), +sample_member_def (sample_weight, weight,T_ULONGLONG ,"event weight" ), +sample_member_def (sample_data_src, data_src,T_ULONGLONG ,"event data source" ), +sample_member_def (sample_insn_count, insn_cnt,T_ULONGLONG ,"event instruction count" ), +sample_member_def (sample_cyc_count, cyc_cnt,T_ULONGLONG ,"event cycle count" ),member_def (perf_event_header, type,T_UINT ,"event type" ), { .name =NULL , }, }; -static void pyrf_sample_event__delete (struct pyrf_event *pevent) -{ -perf_sample__exit (&pevent->sample); -Py_TYPE (pevent)->tp_free ((PyObject *)pevent); -} -static PyObject *pyrf_sample_event__repr (const struct pyrf_event *pevent) {PyObject *ret;
@@ -324,6 +377,8 @@ static PyObject *pyrf_sample_event__repr(const struct pyrf_event *pevent)#ifdef HAVE_LIBTRACEEVENT static bool is_tracepoint (const struct pyrf_event *pevent) { +if (!pevent->sample.evsel) +return false;return pevent->sample.evsel->core.attr.type ==PERF_TYPE_TRACEPOINT ; }
@@ -394,6 +449,199 @@ get_tracepoint_field(struct pyrf_event *pevent, PyObject *attr_name) }#endif /* HAVE_LIBTRACEEVENT */ +static int pyrf_sample_event__resolve_al (struct pyrf_event *pevent) +{ +struct evsel *evsel = pevent->sample.evsel; +struct evlist *evlist = evsel ? evsel->evlist :NULL ; +struct perf_session *session = evlist ?evlist__session (evlist) :NULL ; + +if (pevent->al_resolved) +return 0 ; + +if (!session) +return -1 ; + +addr_location__init (&pevent->al); +if (machine__resolve (&session->machines.host, &pevent->al, &pevent->sample) <0 ) { +addr_location__exit (&pevent->al); +return -1 ; + } + + pevent->al_resolved = true; +return 0 ; +} + +static PyObject *pyrf_sample_event__get_dso (struct pyrf_event *pevent, +void *closure__maybe_unused ) +{ +if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.map) + Py_RETURN_NONE; + +return PyUnicode_FromString (dso__name (map__dso (pevent->al.map))); +} + +static PyObject *pyrf_sample_event__get_dso_long_name (struct pyrf_event *pevent, +void *closure__maybe_unused ) +{ +if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.map) + Py_RETURN_NONE; + +return PyUnicode_FromString (dso__long_name (map__dso (pevent->al.map))); +} + +static PyObject *pyrf_sample_event__get_dso_bid (struct pyrf_event *pevent, +void *closure__maybe_unused ) +{ +char sbuild_id[SBUILD_ID_SIZE ]; + +if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.map) + Py_RETURN_NONE; + +build_id__snprintf (dso__bid (map__dso (pevent->al.map)), sbuild_id,sizeof (sbuild_id)); +return PyUnicode_FromString (sbuild_id); +} + +static PyObject *pyrf_sample_event__get_map_start (struct pyrf_event *pevent, +void *closure__maybe_unused ) +{ +if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.map) + Py_RETURN_NONE; + +return PyLong_FromUnsignedLong (map__start (pevent->al.map)); +} + +static PyObject *pyrf_sample_event__get_map_end (struct pyrf_event *pevent, +void *closure__maybe_unused ) +{ +if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.map) + Py_RETURN_NONE; + +return PyLong_FromUnsignedLong (map__end (pevent->al.map)); +} + +static PyObject *pyrf_sample_event__get_map_pgoff (struct pyrf_event *pevent, +void *closure__maybe_unused ) +{ +if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.map) + Py_RETURN_NONE; + +return PyLong_FromUnsignedLongLong (map__pgoff (pevent->al.map)); +} + +static PyObject *pyrf_sample_event__get_symbol (struct pyrf_event *pevent, +void *closure__maybe_unused ) +{ +if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.sym) + Py_RETURN_NONE; + +return PyUnicode_FromString (pevent->al.sym->name); +} + +static PyObject *pyrf_sample_event__get_sym_start (struct pyrf_event *pevent, +void *closure__maybe_unused ) +{ +if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.sym) + Py_RETURN_NONE; + +return PyLong_FromUnsignedLongLong (pevent->al.sym->start); +} + +static PyObject *pyrf_sample_event__get_sym_end (struct pyrf_event *pevent, +void *closure__maybe_unused ) +{ +if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.sym) + Py_RETURN_NONE; + +return PyLong_FromUnsignedLongLong (pevent->al.sym->end); +} + +static PyObject *pyrf_sample_event__get_raw_buf (struct pyrf_event *pevent, +void *closure__maybe_unused ) +{ +if (pevent->event.header.type !=PERF_RECORD_SAMPLE ) + Py_RETURN_NONE; + +return PyBytes_FromStringAndSize ((const char *)pevent->sample.raw_data, + pevent->sample.raw_size); +} + +static PyObject *pyrf_sample_event__srccode (PyObject *self,PyObject *args) +{ +struct pyrf_event *pevent = (void *)self; +u64 addr = pevent->sample.ip; +char *srcfile =NULL ; +char *srccode =NULL ; +unsigned int line =0 ; +int len =0 ; +PyObject *result; +struct addr_location al; + +if (!PyArg_ParseTuple (args,"|K" , &addr)) +return NULL ; + +if (pyrf_sample_event__resolve_al (pevent) <0 ) + Py_RETURN_NONE; + +if (addr != pevent->sample.ip) { +addr_location__init (&al); +thread__find_symbol_fb (pevent->al.thread, pevent->sample.cpumode, addr, &al); + }else { +addr_location__init (&al); + al.thread =thread__get (pevent->al.thread); + al.map =map__get (pevent->al.map); + al.sym = pevent->al.sym; + al.addr = pevent->al.addr; + } + +if (al.map) { +struct dso *dso =map__dso (al.map); + +if (dso) { + srcfile =get_srcline_split (dso,map__rip_2objdump (al.map, addr), + &line); + } + } +addr_location__exit (&al); + +if (srcfile) { + srccode =find_sourceline (srcfile, line, &len); + result =Py_BuildValue ("(sIs#)" , srcfile, line, srccode, (Py_ssize_t )len); +free (srcfile); + }else { + result =Py_BuildValue ("(sIs#)" ,NULL ,0 ,NULL , (Py_ssize_t )0 ); + } + +return result; +} + +static PyObject *pyrf_sample_event__insn (PyObject *self,PyObject *args__maybe_unused ) +{ +struct pyrf_event *pevent = (void *)self; +struct thread *thread; +struct machine *machine; + +if (pyrf_sample_event__resolve_al (pevent) <0 ) + Py_RETURN_NONE; + + thread = pevent->al.thread; + +if (!thread || !thread__maps (thread)) + Py_RETURN_NONE; + + machine =maps__machine (thread__maps (thread)); +if (!machine) + Py_RETURN_NONE; + +if (pevent->sample.ip && !pevent->sample.insn_len) +perf_sample__fetch_insn (&pevent->sample, thread, machine); + +if (!pevent->sample.insn_len) + Py_RETURN_NONE; + +return PyBytes_FromStringAndSize ((const char *)pevent->sample.insn, + pevent->sample.insn_len); +} +static PyObject *pyrf_sample_event__getattro (struct pyrf_event *pevent,PyObject *attr_name) {
@@ -407,13 +655,102 @@ pyrf_sample_event__getattro(struct pyrf_event *pevent, PyObject *attr_name)return obj ?:PyObject_GenericGetAttr ((PyObject *) pevent, attr_name); } +static PyGetSetDef pyrf_sample_event__getset[] = { + { + .name ="raw_buf" , + .get = (getter )pyrf_sample_event__get_raw_buf, + .set =NULL , + .doc ="event raw buffer." , + }, + { + .name ="evsel" , + .get = pyrf_event__get_evsel, + .set =NULL , + .doc ="tracking event." , + }, + { + .name ="dso" , + .get = (getter )pyrf_sample_event__get_dso, + .set =NULL , + .doc ="event dso short name." , + }, + { + .name ="dso_long_name" , + .get = (getter )pyrf_sample_event__get_dso_long_name, + .set =NULL , + .doc ="event dso long name." , + }, + { + .name ="dso_bid" , + .get = (getter )pyrf_sample_event__get_dso_bid, + .set =NULL , + .doc ="event dso build id." , + }, + { + .name ="map_start" , + .get = (getter )pyrf_sample_event__get_map_start, + .set =NULL , + .doc ="event map start address." , + }, + { + .name ="map_end" , + .get = (getter )pyrf_sample_event__get_map_end, + .set =NULL , + .doc ="event map end address." , + }, + { + .name ="map_pgoff" , + .get = (getter )pyrf_sample_event__get_map_pgoff, + .set =NULL , + .doc ="event map page offset." , + }, + { + .name ="symbol" , + .get = (getter )pyrf_sample_event__get_symbol, + .set =NULL , + .doc ="event symbol name." , + }, + { + .name ="sym_start" , + .get = (getter )pyrf_sample_event__get_sym_start, + .set =NULL , + .doc ="event symbol start address." , + }, + { + .name ="sym_end" , + .get = (getter )pyrf_sample_event__get_sym_end, + .set =NULL , + .doc ="event symbol end address." , + }, + { .name =NULL , }, +}; + +static PyMethodDef pyrf_sample_event__methods[] = { + { + .ml_name ="srccode" , + .ml_meth = (PyCFunction )pyrf_sample_event__srccode, + .ml_flags =METH_VARARGS , + .ml_doc =PyDoc_STR ("Get source code for an address." ) + }, + { + .ml_name ="insn" , + .ml_meth = (PyCFunction )pyrf_sample_event__insn, + .ml_flags =METH_NOARGS , + .ml_doc =PyDoc_STR ("Get instruction bytes for a sample." ) + }, + { .ml_name =NULL , } +}; +static PyTypeObject pyrf_sample_event__type = {PyVarObject_HEAD_INIT (NULL ,0 ) .tp_name ="perf.sample_event" , .tp_basicsize =sizeof (struct pyrf_event ), + .tp_dealloc = (destructor )pyrf_event__delete, .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, .tp_doc = pyrf_sample_event__doc, .tp_members = pyrf_sample_event__members, + .tp_getset = pyrf_sample_event__getset, + .tp_methods = pyrf_sample_event__methods, .tp_repr = (reprfunc )pyrf_sample_event__repr, .tp_getattro = (getattrofunc ) pyrf_sample_event__getattro, };
@@ -449,25 +786,17 @@ static PyTypeObject pyrf_context_switch_event__type = {PyVarObject_HEAD_INIT (NULL ,0 ) .tp_name = "perf.context_switch_event", .tp_basicsize = sizeof(struct pyrf_event), + .tp_dealloc = (destructor)pyrf_event__delete, .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, .tp_doc = pyrf_context_switch_event__doc, .tp_members = pyrf_context_switch_event__members, + .tp_getset = pyrf_event__getset, .tp_repr = (reprfunc)pyrf_context_switch_event__repr, };static int pyrf_event__setup_types (void ) {int err; - pyrf_mmap_event__type.tp_new = - pyrf_task_event__type.tp_new = - pyrf_comm_event__type.tp_new = - pyrf_lost_event__type.tp_new = - pyrf_read_event__type.tp_new = - pyrf_sample_event__type.tp_new = - pyrf_context_switch_event__type.tp_new = - pyrf_throttle_event__type.tp_new = PyType_GenericNew; - - pyrf_sample_event__type.tp_dealloc = (destructor )pyrf_sample_event__delete, err =PyType_Ready (&pyrf_mmap_event__type);if (err <0 )
@@ -511,33 +840,63 @@ static PyTypeObject *pyrf_event__type[] = { [PERF_RECORD_SWITCH_CPU_WIDE] = &pyrf_context_switch_event__type, }; -static PyObject *pyrf_event__new (const union perf_event *event) +static PyObject *pyrf_event__new (const union perf_event *event,struct evsel *evsel) {struct pyrf_event *pevent; -PyTypeObject *ptype; +int err; +u32 min_size; -if ((event->header.type <PERF_RECORD_MMAP || - event->header.type >PERF_RECORD_SAMPLE ) && - !(event->header.type ==PERF_RECORD_SWITCH || - event->header.type ==PERF_RECORD_SWITCH_CPU_WIDE )) { -PyErr_Format (PyExc_TypeError,"Unexpected header type %u" , +if (event->header.type >=ARRAY_SIZE (pyrf_event__type) || + pyrf_event__type[event->header.type] ==NULL ) { +return PyErr_Format (PyExc_TypeError,"Unexpected header type %u" , event->header.type); -return NULL ; } -// FIXME this better be dynamic or we need to parse everything -// before calling perf_mmap__consume(), including tracepoint fields. -if (sizeof (pevent->event) < event->header.size) { -PyErr_Format (PyExc_TypeError,"Unexpected event size: %zd < %u" , -sizeof (pevent->event), event->header.size); -return NULL ; +if (perf_event__too_small (event, &min_size)) { +return PyErr_Format (PyExc_ValueError,"Event size %u too small for type %u" , + event->header.size, event->header.type); + } + +size_t copy_size = event->header.size; + + pevent = PyObject_New(struct pyrf_event, pyrf_event__type[event->header.type]); +if (pevent ==NULL ) +return PyErr_NoMemory (); + +if (copy_size >sizeof (pevent->event)) + copy_size =sizeof (pevent->event); + +/* Copy the event for memory safety and initilaize variables. */ +memcpy (&pevent->event, event, copy_size); + pevent->event.header.size = copy_size; + +if (event->header.type ==PERF_RECORD_MMAP ) { +/* Ensure '\0' string termination. */ +size_t max_len = copy_size - + offsetof(struct perf_record_mmap , filename); + + pevent->event.mmap.filename[max_len -1 ] ='\0' ; + }else if (event->header.type ==PERF_RECORD_COMM ) { +/* Ensure '\0' string termination. */ +size_t max_len = copy_size - + offsetof(struct perf_record_comm , comm); + + pevent->event.comm.comm[max_len -1 ] ='\0' ; } - ptype = pyrf_event__type[event->header.type]; - pevent = PyObject_New(struct pyrf_event, ptype); -if (pevent !=NULL ) { -memcpy (&pevent->event, event, event->header.size); -perf_sample__init (&pevent->sample,/*all=*/ false); +perf_sample__init (&pevent->sample,/*all=*/ true); + pevent->al_resolved = false; +addr_location__init (&pevent->al); + +if (!evsel) +return (PyObject *)pevent; + +/* Parse the sample again so that pointers are within the copied event. */ + err =evsel__parse_sample (evsel, &pevent->event, &pevent->sample); +if (err <0 ) { +Py_DECREF (pevent); +return PyErr_Format (PyExc_OSError, +"perf: can't parse sample, err=%d" , err); }return (PyObject *)pevent; }
@@ -1244,7 +1603,7 @@ static PyObject *pyrf_evsel__str(PyObject *self)if (!evsel)return PyUnicode_FromString ("evsel(uninitialized)" ); -return PyUnicode_FromFormat ("evsel(%s/%s/)" ,evsel__pmu_name (evsel),evsel__name (evsel)); +return PyUnicode_FromFormat ("evsel(%s)" ,evsel__name (evsel)); }static PyMethodDef pyrf_evsel__methods[] = {
@@ -1876,9 +2235,11 @@ static PyObject *pyrf_evlist__read_on_cpu(struct pyrf_evlist *pevlist, {struct evlist *evlist;union perf_event *event; +struct evsel *evsel;int sample_id_all =1 , cpu;static char *kwlist[] = {"cpu" ,"sample_id_all" ,NULL };struct mmap *md; +PyObject *pyevent;int err;CHECK_INITIALIZED (pevlist->evlist,"evlist" );
@@ -1889,44 +2250,31 @@ static PyObject *pyrf_evlist__read_on_cpu(struct pyrf_evlist *pevlist,return NULL ; md =get_md (evlist, cpu); -if (!md) { -PyErr_Format (PyExc_TypeError,"Unknown CPU '%d'" , cpu); -return NULL ; - } +if (!md) +return PyErr_Format (PyExc_TypeError,"Unknown CPU '%d'" , cpu); -if (perf_mmap__read_init (&md->core) <0 ) - goto end; + err =perf_mmap__read_init (&md->core); +if (err <0 ) { +return PyErr_Format (PyExc_OSError, +"perf: error mmap read init, err=%d" , err); + } event =perf_mmap__read_event (&md->core); -if (event !=NULL ) { -PyObject *pyevent =pyrf_event__new (event); -struct pyrf_event *pevent = (struct pyrf_event *)pyevent; -struct evsel *evsel; - -if (pyevent ==NULL ) -return PyErr_NoMemory (); - - evsel =evlist__event2evsel (evlist, event); -if (!evsel) { -Py_DECREF (pyevent); -Py_INCREF (Py_None); -return Py_None; - } +if (event ==NULL ) + Py_RETURN_NONE; + evsel =evlist__event2evsel (evlist, event); +if (!evsel) { +/* Unknown evsel. */ perf_mmap__consume (&md->core); - - err =evsel__parse_sample (evsel, &pevent->event, &pevent->sample); -if (err) { -Py_DECREF (pyevent); -return PyErr_Format (PyExc_OSError, -"perf: can't parse sample, err=%d" , err); - } - -return pyevent; + Py_RETURN_NONE; } -end: -Py_INCREF (Py_None); -return Py_None; + pyevent =pyrf_event__new (event, evsel); +perf_mmap__consume (&md->core); +if (pyevent ==NULL ) +return PyErr_Occurred () ?NULL :PyErr_NoMemory (); + +return pyevent; } static PyObject *pyrf_evlist__open(struct pyrf_evlist *pevlist,
@@ -2138,10 +2486,7 @@ static PyObject *pyrf_evlist__str(PyObject *self)evlist__for_each_entry (pevlist->evlist, pos) {if (!first)strbuf_addch (&sb,',' ); -if (!pos->pmu) -strbuf_addstr (&sb,evsel__name (pos)); -else -strbuf_addf (&sb,"%s/%s/" , pos->pmu->name,evsel__name (pos)); +strbuf_addstr (&sb,evsel__name (pos)); first = false; }strbuf_addstr (&sb,"])" );
@@ -2690,19 +3035,12 @@ static int pyrf_session_tool__sample(const struct perf_tool *tool, struct machine *machine __maybe_unused) { struct pyrf_session *psession = container_of(tool, struct pyrf_session, tool); -PyObject *pyevent =pyrf_event__new (event); -struct pyrf_event *pevent = (struct pyrf_event *)pyevent; +PyObject *pyevent =pyrf_event__new (event, sample->evsel);PyObject *ret;if (pyevent ==NULL )return -ENOMEM ; -memcpy (&pevent->event, event, event->header.size); -if (evsel__parse_sample (sample->evsel, &pevent->event, &pevent->sample) <0 ) { -Py_DECREF (pyevent); -return -1 ; - } - ret =PyObject_CallFunction (psession->sample,"O" , pyevent);if (!ret) {Py_DECREF (pyevent);
tools/perf/util/session.c +1 -1
@@ -2672,7 +2672,7 @@ static const u32 perf_event__min_size[PERF_RECORD_HEADER_MAX] = {* Caller must ensure event->header.type < PERF_RECORD_HEADER_MAX. * If min is non-NULL, stores the required minimum on failure. */ -static bool perf_event__too_small (const union perf_event *event,u32 *min) +bool perf_event__too_small (const union perf_event *event,u32 *min) {u32 min_sz = perf_event__min_size[event->header.type];
tools/perf/util/session.h +2 -0
@@ -122,6 +122,8 @@ void perf_session__delete(struct perf_session *session);void perf_event_header__bswap (struct perf_event_header *hdr); +bool perf_event__too_small (const union perf_event *event,u32 *min); + int perf_session__peek_event(struct perf_session *session, off_t file_offset, void *buf, size_t buf_sz, union perf_event **event_ptr,
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
[email protected]
AI reviewRe: [PATCH v14 11/19] perf python: Refactor and add accessors to sample event
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider: - [High] Data corruption when forcibly null-terminating MMAP, MMAP2, and COMM events. --
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] Data corruption when forcibly null-terminating MMAP, MMAP2, and COMM events. --
commit 997b9e752ea78b14ab40dbb997753ef37ba7cc71
Author: Ian Rogers <[email protected]>
perf python: Refactor and add accessors to sample event
This commit adds a common evsel field for events and moves sample specific fields to only be present in sample events. It adds accessors for sample events, ensures offsets are within the bounds of the event, and allocates just enough memory for the copied event.
Quoted 4 lines
tools/perf/util/python.c +0
-0
[ ... ]
Quoted 1 line
tools/perf/util/python.c +0
-0@@ -511,33 +840,63 @@ static PyTypeObject *pyrf_event__type[] = {
[ ... ]
Quoted 16 lines
tools/perf/util/python.c +16
-0+size_t copy_size = event->header.size; + + pevent = PyObject_New(struct pyrf_event, pyrf_event__type[event->header.type]); +if (pevent ==NULL ) +return PyErr_NoMemory (); + +if (copy_size >sizeof (pevent->event)) + copy_size =sizeof (pevent->event); + +/* Copy the event for memory safety and initilaize variables. */ +memcpy (&pevent->event, event, copy_size); + pevent->event.header.size = copy_size; + +if (event->header.type ==PERF_RECORD_MMAP ) { +/* Ensure '\0' string termination. */ + size_t max_len = copy_size -
tools/perf/util/python.c static PyTypeObject *pyrf_event__type[] = { · line 885
Quoted 10 lines
tools/perf/util/python.c +9
-0+ offsetof(struct perf_record_mmap , filename); + + pevent->event.mmap.filename[max_len -1 ] ='\0' ; + }else if (event->header.type ==PERF_RECORD_COMM ) { +/* Ensure '\0' string termination. */ +size_t max_len = copy_size - + offsetof(struct perf_record_comm , comm); + + pevent->event.comm.comm[max_len -1 ] ='\0' ; }
[Severity: High]
If sample_id_all is enabled, the kernel appends a sample_id trailer (like CPU, TID, or timestamps) at the end of non-sample events such as PERF_RECORD_MMAP and PERF_RECORD_COMM.
Since copy_size is derived from event->header.size, doesn't max_len include the trailer length as well?
Could writing a null byte at max_len - 1 overwrite the final byte of the sample_id trailer data, silently corrupting it before it gets parsed by evsel__parse_sample()?
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=11
Ian Rogers
reviewerPATCH v14 12/19 perf python: Add mmap2 event
In reply to Ian RogersIf mmap is handled so should mmap2 events. Add support as a distinct python event type.
---
v12:
- Used copy_size to bound max_len for NUL termination of mmap2 and comm events to fix an out-of-bounds write.
--- tools/perf/util/python.c | 158 ++++++++++++++++++++++++++++++++++++++- 1 file changed, 157 insertions(+), 1 deletion(-)
tools/perf/util/python.c +157 -1
@@ -133,7 +133,7 @@ static const char pyrf_mmap_event__doc[] = PyDoc_STR("perf mmap event object.");static PyMemberDef pyrf_mmap_event__members[] = { sample_membersmember_def (perf_event_header, type,T_UINT ,"event type" ), -member_def (perf_event_header, misc,T_UINT ,"event misc" ), +member_def (perf_event_header, misc,T_USHORT ,"event misc" ),member_def (perf_record_mmap, pid,T_UINT ,"event pid" ),member_def (perf_record_mmap, tid,T_UINT ,"event tid" ),member_def (perf_record_mmap, start,T_ULONGLONG ,"start of the map" ),
@@ -174,6 +174,149 @@ static PyTypeObject pyrf_mmap_event__type = { .tp_repr = (reprfunc)pyrf_mmap_event__repr, }; +static const char pyrf_mmap2_event__doc[] =PyDoc_STR ("perf mmap2 event object." ); + +static PyObject *pyrf_mmap2_event__get_maj (PyObject *self,void *closure__maybe_unused ) +{ +struct pyrf_event *pevent = (void *)self; + +if (pevent->event.header.misc &PERF_RECORD_MISC_MMAP_BUILD_ID ) + Py_RETURN_NONE; + +return PyLong_FromUnsignedLong (pevent->event.mmap2.maj); +} + +static PyObject *pyrf_mmap2_event__get_min (PyObject *self,void *closure__maybe_unused ) +{ +struct pyrf_event *pevent = (void *)self; + +if (pevent->event.header.misc &PERF_RECORD_MISC_MMAP_BUILD_ID ) + Py_RETURN_NONE; + +return PyLong_FromUnsignedLong (pevent->event.mmap2.min); +} + +static PyObject *pyrf_mmap2_event__get_ino (PyObject *self,void *closure__maybe_unused ) +{ +struct pyrf_event *pevent = (void *)self; + +if (pevent->event.header.misc &PERF_RECORD_MISC_MMAP_BUILD_ID ) + Py_RETURN_NONE; + +return PyLong_FromUnsignedLongLong (pevent->event.mmap2.ino); +} + +static PyObject *pyrf_mmap2_event__get_ino_generation (PyObject *self,void *closure__maybe_unused ) +{ +struct pyrf_event *pevent = (void *)self; + +if (pevent->event.header.misc &PERF_RECORD_MISC_MMAP_BUILD_ID ) + Py_RETURN_NONE; + +return PyLong_FromUnsignedLongLong (pevent->event.mmap2.ino_generation); +} + +static PyObject *pyrf_mmap2_event__get_build_id (PyObject *self,void *closure__maybe_unused ) +{ +struct pyrf_event *pevent = (void *)self; + +if (!(pevent->event.header.misc &PERF_RECORD_MISC_MMAP_BUILD_ID )) + Py_RETURN_NONE; + +int size = pevent->event.mmap2.build_id_size; + +if (size >20 ) + size =20 ; + +return PyBytes_FromStringAndSize ((const char *)pevent->event.mmap2.build_id, size); +} + +static PyGetSetDef pyrf_mmap2_event__getset[] = { + { + .name ="evsel" , + .get = pyrf_event__get_evsel, + .set =NULL , + .doc ="tracking event." , + }, + { + .name ="maj" , + .get = pyrf_mmap2_event__get_maj, + .set =NULL , + .doc ="major number." , + }, + { + .name ="min" , + .get = pyrf_mmap2_event__get_min, + .set =NULL , + .doc ="minor number." , + }, + { + .name ="ino" , + .get = pyrf_mmap2_event__get_ino, + .set =NULL , + .doc ="inode number." , + }, + { + .name ="ino_generation" , + .get = pyrf_mmap2_event__get_ino_generation, + .set =NULL , + .doc ="inode generation." , + }, + { + .name ="build_id" , + .get = pyrf_mmap2_event__get_build_id, + .set =NULL , + .doc ="binary build ID." , + }, + { .name =NULL , }, +}; + +static PyMemberDef pyrf_mmap2_event__members[] = { + sample_members +member_def (perf_event_header, type,T_UINT ,"event type" ), +member_def (perf_event_header, misc,T_USHORT ,"event misc" ), +member_def (perf_record_mmap2, pid,T_UINT ,"event pid" ), +member_def (perf_record_mmap2, tid,T_UINT ,"event tid" ), +member_def (perf_record_mmap2, start,T_ULONGLONG ,"start of the map" ), +member_def (perf_record_mmap2, len,T_ULONGLONG ,"map length" ), +member_def (perf_record_mmap2, pgoff,T_ULONGLONG ,"page offset" ), +member_def (perf_record_mmap2, prot,T_UINT ,"protection" ), +member_def (perf_record_mmap2, flags,T_UINT ,"flags" ), +member_def (perf_record_mmap2, filename,T_STRING_INPLACE ,"backing store" ), + { .name =NULL , }, +}; + +static PyObject *pyrf_mmap2_event__repr (const struct pyrf_event *pevent) +{ +PyObject *ret; +char *s; + +if (asprintf (&s,"{ type: mmap2, pid: %u, tid: %u, start: %#" PRI_lx64", " +"length: %#" PRI_lx64", offset: %#" PRI_lx64", " +"flags: %#x, prot: %#x, filename: %s }" , + pevent->event.mmap2.pid, pevent->event.mmap2.tid, + pevent->event.mmap2.start, pevent->event.mmap2.len, + pevent->event.mmap2.pgoff, pevent->event.mmap2.flags, + pevent->event.mmap2.prot, pevent->event.mmap2.filename) <0 ) +return PyErr_NoMemory (); + + ret =PyUnicode_FromString (s); +free (s); +return ret; +} + +static PyTypeObject pyrf_mmap2_event__type = { +PyVarObject_HEAD_INIT (NULL ,0 ) + .tp_name ="perf.mmap2_event" , + .tp_basicsize =sizeof (struct pyrf_event ), + .tp_dealloc = (destructor )pyrf_event__delete, + .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, + .tp_doc = pyrf_mmap2_event__doc, + .tp_members = pyrf_mmap2_event__members, + .tp_getset = pyrf_mmap2_event__getset, + .tp_repr = (reprfunc )pyrf_mmap2_event__repr, +}; +static const char pyrf_task_event__doc[] =PyDoc_STR ("perf task (fork/exit) event object." );static PyMemberDef pyrf_task_event__members[] = {
@@ -799,6 +942,9 @@ static int pyrf_event__setup_types(void)int err; err =PyType_Ready (&pyrf_mmap_event__type); +if (err <0 ) + goto out; + err =PyType_Ready (&pyrf_mmap2_event__type);if (err <0 ) goto out; err =PyType_Ready (&pyrf_lost_event__type);
@@ -828,6 +974,7 @@ static int pyrf_event__setup_types(void)static PyTypeObject *pyrf_event__type[] = { [PERF_RECORD_MMAP ] = &pyrf_mmap_event__type, + [PERF_RECORD_MMAP2 ] = &pyrf_mmap2_event__type, [PERF_RECORD_LOST ] = &pyrf_lost_event__type, [PERF_RECORD_COMM ] = &pyrf_comm_event__type, [PERF_RECORD_EXIT ] = &pyrf_task_event__type,
@@ -876,6 +1023,12 @@ static PyObject *pyrf_event__new(const union perf_event *event, struct evsel *ev offsetof(struct perf_record_mmap , filename); pevent->event.mmap.filename[max_len -1 ] ='\0' ; + }else if (event->header.type ==PERF_RECORD_MMAP2 ) { +/* Ensure '\0' string termination. */ +size_t max_len = copy_size - + offsetof(struct perf_record_mmap2 , filename); + + pevent->event.mmap2.filename[max_len -1 ] ='\0' ; }else if (event->header.type ==PERF_RECORD_COMM ) {/* Ensure '\0' string termination. */ size_t max_len = copy_size -
@@ -3322,6 +3475,9 @@ PyMODINIT_FUNC PyInit_perf(void)Py_INCREF (&pyrf_mmap_event__type);PyModule_AddObject (module,"mmap_event" , (PyObject *)&pyrf_mmap_event__type); +Py_INCREF (&pyrf_mmap2_event__type); +PyModule_AddObject (module,"mmap2_event" , (PyObject *)&pyrf_mmap2_event__type); +Py_INCREF (&pyrf_lost_event__type);PyModule_AddObject (module,"lost_event" , (PyObject *)&pyrf_lost_event__type);
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
Ian Rogers
reviewerPATCH v14 13/19 perf python: Add callchain support
In reply to Ian RogersImplement pyrf_callchain_node and pyrf_callchain types for lazy iteration over callchain frames. Add callchain property to sample_event.
---
v12:
- Added an optional struct machine * argument to pyrf_event__new defaulting to the host machine if NULL, avoiding regressions for future phases.
v2:
1. Eager Callchain Resolution: Moved the callchain resolution from deferred iteration to eager processing in pyrf_session_tool__sample() . This avoids risks of reading from unmapped memory or following dangling pointers to closed sessions.
2. Cached Callchain: Added a callchain field to struct pyrf_event to store the resolved object.
3. Simplified Access: pyrf_sample_event__get_callchain() now just returns the cached object if available.
4. Avoided Double Free: Handled lazy cleanups properly.
v6:
- Moved callchain resolution from session_tool__sample to pyrf_event__new.
--- tools/perf/util/python.c | 215 ++++++++++++++++++++++++++++++++++++++- 1 file changed, 211 insertions(+), 4 deletions(-)
tools/perf/util/python.c +211 -4
@@ -87,6 +87,8 @@ struct pyrf_event {struct addr_location al;/** @al_resolved: True when machine__resolve been called. */ bool al_resolved; +/** @callchain: Resolved callchain, eagerly computed if requested. */ +PyObject *callchain;/** @event: The underlying perf_event that may be in a file or ring buffer. */ union perf_event event; };
@@ -124,6 +126,7 @@ static void pyrf_event__delete(struct pyrf_event *pevent) {if (pevent->al_resolved)addr_location__exit (&pevent->al); +Py_XDECREF (pevent->callchain);perf_sample__exit (&pevent->sample);Py_TYPE (pevent)->tp_free ((PyObject *)pevent); }
@@ -785,6 +788,144 @@ static PyObject *pyrf_sample_event__insn(PyObject *self, PyObject *args __maybe_ pevent->sample.insn_len); } +struct pyrf_callchain_node { +PyObject_HEAD + u64 ip; +struct map *map; +struct symbol *sym; +}; + +static void pyrf_callchain_node__delete (struct pyrf_callchain_node *pnode) +{ +map__put (pnode->map); +Py_TYPE (pnode)->tp_free ((PyObject *)pnode); +} + +static PyObject *pyrf_callchain_node__get_ip (struct pyrf_callchain_node *pnode, +void *closure__maybe_unused ) +{ +return PyLong_FromUnsignedLongLong (pnode->ip); +} + +static PyObject *pyrf_callchain_node__get_symbol (struct pyrf_callchain_node *pnode, +void *closure__maybe_unused ) +{ +if (pnode->sym) +return PyUnicode_FromString (pnode->sym->name); +return PyUnicode_FromString ("[unknown]" ); +} + +static PyObject *pyrf_callchain_node__get_dso (struct pyrf_callchain_node *pnode, +void *closure__maybe_unused ) +{ +const char *dsoname ="[unknown]" ; + +if (pnode->map) { +struct dso *dso =map__dso (pnode->map); + +if (dso) { +if (symbol_conf.show_kernel_path &&dso__long_name (dso)) + dsoname =dso__long_name (dso); +else + dsoname =dso__name (dso); + } + } +return PyUnicode_FromString (dsoname); +} + +static PyGetSetDef pyrf_callchain_node__getset[] = { + { .name ="ip" , .get = (getter )pyrf_callchain_node__get_ip, }, + { .name ="symbol" , .get = (getter )pyrf_callchain_node__get_symbol, }, + { .name ="dso" , .get = (getter )pyrf_callchain_node__get_dso, }, + { .name =NULL , }, +}; + +static PyTypeObject pyrf_callchain_node__type = { +PyVarObject_HEAD_INIT (NULL ,0 ) + .tp_name ="perf.callchain_node" , + .tp_basicsize =sizeof (struct pyrf_callchain_node ), + .tp_dealloc = (destructor )pyrf_callchain_node__delete, + .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, + .tp_doc ="perf callchain node object." , + .tp_getset = pyrf_callchain_node__getset, +}; + +struct pyrf_callchain_frame { +u64 ip; +struct map *map; +struct symbol *sym; +}; + +struct pyrf_callchain { +PyObject_HEAD + struct pyrf_callchain_frame *frames; +u64 nr_frames; +}; + +static void pyrf_callchain__delete (struct pyrf_callchain *pchain) +{ +if (pchain->frames) { +for (u64 i =0 ; i < pchain->nr_frames; i++) +map__put (pchain->frames[i].map); +free (pchain->frames); + } +Py_TYPE (pchain)->tp_free ((PyObject *)pchain); +} + +static Py_ssize_t pyrf_callchain__length (PyObject *obj) +{ +struct pyrf_callchain *pchain = (void *)obj; + +return pchain->nr_frames; +} + +static PyObject *pyrf_callchain__item (PyObject *obj,Py_ssize_t i) +{ +struct pyrf_callchain *pchain = (void *)obj; +struct pyrf_callchain_node *pnode; + +if (i <0 || i >= (Py_ssize_t )pchain->nr_frames) { +PyErr_SetString (PyExc_IndexError,"Index out of range" ); +return NULL ; + } + + pnode = PyObject_New(struct pyrf_callchain_node, &pyrf_callchain_node__type); +if (!pnode) +return NULL ; + + pnode->ip = pchain->frames[i].ip; + pnode->map =map__get (pchain->frames[i].map); + pnode->sym = pchain->frames[i].sym; + +return (PyObject *)pnode; +} + +static PySequenceMethods pyrf_callchain__sequence_methods = { + .sq_length = pyrf_callchain__length, + .sq_item = pyrf_callchain__item, +}; + +static PyTypeObject pyrf_callchain__type = { +PyVarObject_HEAD_INIT (NULL ,0 ) + .tp_name ="perf.callchain" , + .tp_basicsize =sizeof (struct pyrf_callchain ), + .tp_dealloc = (destructor )pyrf_callchain__delete, + .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, + .tp_doc ="perf callchain object." , + .tp_as_sequence = &pyrf_callchain__sequence_methods, +}; + +static PyObject *pyrf_sample_event__get_callchain (PyObject *self,void *closure__maybe_unused ) +{ +struct pyrf_event *pevent = (void *)self; + +if (!pevent->callchain) + Py_RETURN_NONE; + +Py_INCREF (pevent->callchain); +return pevent->callchain; +} +static PyObject *pyrf_sample_event__getattro (struct pyrf_event *pevent,PyObject *attr_name) {
@@ -799,6 +940,12 @@ pyrf_sample_event__getattro(struct pyrf_event *pevent, PyObject *attr_name) }static PyGetSetDef pyrf_sample_event__getset[] = { + { + .name ="callchain" , + .get = pyrf_sample_event__get_callchain, + .set =NULL , + .doc ="event callchain." , + }, { .name ="raw_buf" , .get = (getter )pyrf_sample_event__get_raw_buf,
@@ -968,6 +1115,12 @@ static int pyrf_event__setup_types(void) err =PyType_Ready (&pyrf_context_switch_event__type);if (err <0 ) goto out; + err =PyType_Ready (&pyrf_callchain_node__type); +if (err <0 ) + goto out; + err =PyType_Ready (&pyrf_callchain__type); +if (err <0 ) + goto out; out:return err; }
@@ -987,12 +1140,18 @@ static PyTypeObject *pyrf_event__type[] = { [PERF_RECORD_SWITCH_CPU_WIDE] = &pyrf_context_switch_event__type, }; -static PyObject *pyrf_event__new (const union perf_event *event,struct evsel *evsel) +static PyObject *pyrf_event__new (const union perf_event *event,struct evsel *evsel, +struct perf_session *session, +struct machine *machine) {struct pyrf_event *pevent; +struct perf_sample *sample;int err;u32 min_size; +if (!machine) + machine = session ? &session->machines.host :NULL ; +if (event->header.type >=ARRAY_SIZE (pyrf_event__type) || pyrf_event__type[event->header.type] ==NULL ) { return PyErr_Format(PyExc_TypeError, "Unexpected header type %u",
@@ -1038,6 +1197,7 @@ static PyObject *pyrf_event__new(const union perf_event *event, struct evsel *ev }perf_sample__init (&pevent->sample,/*all=*/ true); + pevent->callchain =NULL ; pevent->al_resolved = false;addr_location__init (&pevent->al);
@@ -1051,6 +1211,50 @@ static PyObject *pyrf_event__new(const union perf_event *event, struct evsel *evreturn PyErr_Format (PyExc_OSError,"perf: can't parse sample, err=%d" , err); } + sample = &pevent->sample; +if (machine && sample->callchain) { +struct addr_location al; +struct callchain_cursor *cursor; +u64 i; +struct pyrf_callchain *pchain; + +addr_location__init (&al); +if (machine__resolve (machine, &al, sample) >=0 ) { + cursor =get_tls_callchain_cursor (); +if (thread__resolve_callchain (al.thread, cursor, sample, +NULL ,NULL ,PERF_MAX_STACK_DEPTH ) ==0 ) { +callchain_cursor_commit (cursor); + + pchain = PyObject_New(struct pyrf_callchain, &pyrf_callchain__type); +if (!pchain) { +addr_location__exit (&al); +Py_DECREF (pevent); +return NULL ; + } + pchain->nr_frames = cursor->nr; + pchain->frames =calloc (pchain->nr_frames, +sizeof (*pchain->frames)); +if (!pchain->frames) { +Py_DECREF (pchain); +addr_location__exit (&al); +Py_DECREF (pevent); +return PyErr_NoMemory (); + } +struct callchain_cursor_node *node; + +for (i =0 ; i < pchain->nr_frames; i++) { + node =callchain_cursor_current (cursor); + pchain->frames[i].ip = node->ip; + pchain->frames[i].map = +map__get (node->ms.map); + pchain->frames[i].sym = node->ms.sym; +callchain_cursor_advance (cursor); + } + pevent->callchain = (PyObject *)pchain; + } +addr_location__exit (&al); + } + }return (PyObject *)pevent; }
@@ -2422,7 +2626,7 @@ static PyObject *pyrf_evlist__read_on_cpu(struct pyrf_evlist *pevlist,perf_mmap__consume (&md->core); Py_RETURN_NONE; } - pyevent =pyrf_event__new (event, evsel); + pyevent =pyrf_event__new (event, evsel,evlist__session (evlist),/*machine=*/ NULL );perf_mmap__consume (&md->core);if (pyevent ==NULL )return PyErr_Occurred () ?NULL :PyErr_NoMemory ();
@@ -3185,10 +3389,10 @@ struct pyrf_session {static int pyrf_session_tool__sample (const struct perf_tool *tool,union perf_event *event,struct perf_sample *sample, -struct machine *machine__maybe_unused ) +struct machine *machine) { struct pyrf_session *psession = container_of(tool, struct pyrf_session, tool); -PyObject *pyevent =pyrf_event__new (event, sample->evsel); +PyObject *pyevent =pyrf_event__new (event, sample->evsel, psession->session, machine);PyObject *ret;if (pyevent ==NULL )
@@ -3296,6 +3500,9 @@ static PyObject *pyrf_session__new(PyTypeObject *type, PyObject *args, PyObject } psession->session = session; + symbol_conf.use_callchain = true; + symbol_conf.show_kernel_path = true; + symbol_conf.inline_name = false;if (symbol__init (perf_session__env (session)) <0 ) {PyErr_SetString (PyExc_OSError,"perf: symbol__init failed" ); goto err_out;
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
Ian Rogers
reviewerPATCH v14 14/19 perf python: Extend API for stat events in python.c
In reply to Ian RogersAdd stat information to the session. Add call backs for stat events.
---
v12:
- Made psession->tool.stat = perf_event__process_stat_event; conditional on !stat so it doesn't unconditionally overwrite user stat callbacks.
v5:
1. Fix Memory Corruption: Corrected the memory offset for stat_round_type in pyrf_stat_round_event__members by adding the base offset of struct pyrf_event.
2. Fix Memory Leak: Added Py_XDECREF() to free the unused return value of the Python callback in pyrf_session_tool__stat_round().
---
v7:
- Added comprehensive size checks in pyrf_event__new for all event types.
- Fixed line length warning.
v10:
- Added CHECK_INITIALIZED check to pyrf_evsel__get_ids to prevent NULL pointer dereference on uninitialized evsel objects.
--- tools/perf/util/python.c | 188 +++++++++++++++++++++++++++++++++++++-- 1 file changed, 183 insertions(+), 5 deletions(-)
tools/perf/util/python.c +183 -5
@@ -459,6 +459,77 @@ static PyTypeObject pyrf_lost_event__type = { .tp_repr = (reprfunc)pyrf_lost_event__repr, }; +static const char pyrf_stat_event__doc[] =PyDoc_STR ("perf stat event object." ); + +static PyMemberDef pyrf_stat_event__members[] = { + sample_members +member_def (perf_event_header, type,T_UINT ,"event type" ), +member_def (perf_record_stat, id,T_ULONGLONG ,"event id" ), +member_def (perf_record_stat, cpu,T_UINT ,"event cpu" ), +member_def (perf_record_stat, thread,T_UINT ,"event thread" ), +member_def (perf_record_stat, val,T_ULONGLONG ,"counter value" ), +member_def (perf_record_stat, ena,T_ULONGLONG ,"enabled time" ), +member_def (perf_record_stat, run,T_ULONGLONG ,"running time" ), + { .name =NULL , }, +}; + +static PyObject *pyrf_stat_event__repr (const struct pyrf_event *pevent) +{ +return PyUnicode_FromFormat ( +"{ type: stat, id: %llu, cpu: %u, thread: %u, val: %llu, ena: %llu, run: %llu }" , + pevent->event.stat.id, + pevent->event.stat.cpu, + pevent->event.stat.thread, + pevent->event.stat.val, + pevent->event.stat.ena, + pevent->event.stat.run); +} + +static PyTypeObject pyrf_stat_event__type = { +PyVarObject_HEAD_INIT (NULL ,0 ) + .tp_name ="perf.stat_event" , + .tp_basicsize =sizeof (struct pyrf_event ), + .tp_new = PyType_GenericNew, + .tp_dealloc = (destructor )pyrf_event__delete, + .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, + .tp_doc = pyrf_stat_event__doc, + .tp_members = pyrf_stat_event__members, + .tp_getset = pyrf_event__getset, + .tp_repr = (reprfunc )pyrf_stat_event__repr, +}; + +static const char pyrf_stat_round_event__doc[] =PyDoc_STR ("perf stat round event object." ); + +static PyMemberDef pyrf_stat_round_event__members[] = { + sample_members +member_def (perf_event_header, type,T_UINT ,"event type" ), + { .name ="stat_round_type" , .type =T_ULONGLONG , + .offset = offsetof(struct pyrf_event , event) + offsetof(struct perf_record_stat_round , type), + .doc ="round type" }, +member_def (perf_record_stat_round, time,T_ULONGLONG ,"round time" ), + { .name =NULL , }, +}; + +static PyObject *pyrf_stat_round_event__repr (const struct pyrf_event *pevent) +{ +return PyUnicode_FromFormat ("{ type: stat_round, type: %llu, time: %llu }" , + pevent->event.stat_round.type, + pevent->event.stat_round.time); +} + +static PyTypeObject pyrf_stat_round_event__type = { +PyVarObject_HEAD_INIT (NULL ,0 ) + .tp_name ="perf.stat_round_event" , + .tp_basicsize =sizeof (struct pyrf_event ), + .tp_new = PyType_GenericNew, + .tp_dealloc = (destructor )pyrf_event__delete, + .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, + .tp_doc = pyrf_stat_round_event__doc, + .tp_members = pyrf_stat_round_event__members, + .tp_getset = pyrf_event__getset, + .tp_repr = (reprfunc )pyrf_stat_round_event__repr, +}; +static const char pyrf_read_event__doc[] =PyDoc_STR ("perf read event object." );static PyMemberDef pyrf_read_event__members[] = {
@@ -1113,6 +1184,12 @@ static int pyrf_event__setup_types(void)if (err <0 ) goto out; err =PyType_Ready (&pyrf_context_switch_event__type); +if (err <0 ) + goto out; + err =PyType_Ready (&pyrf_stat_event__type); +if (err <0 ) + goto out; + err =PyType_Ready (&pyrf_stat_round_event__type);if (err <0 ) goto out; err =PyType_Ready (&pyrf_callchain_node__type);
@@ -1138,6 +1215,8 @@ static PyTypeObject *pyrf_event__type[] = { [PERF_RECORD_SAMPLE] = &pyrf_sample_event__type, [PERF_RECORD_SWITCH] = &pyrf_context_switch_event__type, [PERF_RECORD_SWITCH_CPU_WIDE] = &pyrf_context_switch_event__type, + [PERF_RECORD_STAT] = &pyrf_stat_event__type, + [PERF_RECORD_STAT_ROUND] = &pyrf_stat_round_event__type, }; static PyObject *pyrf_event__new(const union perf_event *event, struct evsel *evsel,
@@ -2141,7 +2220,45 @@ static PyObject *pyrf_evsel__get_attr_wakeup_events(PyObject *self, void *closurreturn PyLong_FromUnsignedLong (pevsel->evsel->core.attr.wakeup_events); } +static PyObject *pyrf_evsel__get_ids (struct pyrf_evsel *pevsel,void *closure__maybe_unused ) +{ +struct evsel *evsel; +PyObject *list; + +CHECK_INITIALIZED (pevsel->evsel,"evsel" ); + + evsel = pevsel->evsel; + list =PyList_New (0 ); + +if (!list) +return NULL ; + +for (u32 i =0 ; i < evsel->core.ids; i++) { +PyObject *id =PyLong_FromUnsignedLongLong (evsel->core.id[i]); +int ret; + +if (!id) { +Py_DECREF (list); +return NULL ; + } + ret =PyList_Append (list, id); +Py_DECREF (id); +if (ret <0 ) { +Py_DECREF (list); +return NULL ; + } + } + +return list; +} +static PyGetSetDef pyrf_evsel__getset[] = { + { + .name ="ids" , + .get = (getter )pyrf_evsel__get_ids, + .set =NULL , + .doc ="event IDs." , + }, { .name ="tracking" , .get = pyrf_evsel__get_tracking,
@@ -2989,6 +3106,8 @@ static const struct perf_constant perf__constants[] = {PERF_CONST (RECORD_LOST_SAMPLES ),PERF_CONST (RECORD_SWITCH ),PERF_CONST (RECORD_SWITCH_CPU_WIDE ), +PERF_CONST (RECORD_STAT ), +PERF_CONST (RECORD_STAT_ROUND ),PERF_CONST (RECORD_MISC_SWITCH_OUT ), { .name = NULL, },
@@ -3384,6 +3503,7 @@ struct pyrf_session {struct perf_tool tool;struct pyrf_data *pdata;PyObject *sample; +PyObject *stat; }; static int pyrf_session_tool__sample(const struct perf_tool *tool,
@@ -3408,6 +3528,50 @@ static int pyrf_session_tool__sample(const struct perf_tool *tool,return 0 ; } +static int pyrf_session_tool__stat (const struct perf_tool *tool, +struct perf_session *session, +union perf_event *event) +{ + struct pyrf_session *psession = container_of(tool, struct pyrf_session, tool); +struct evsel *evsel =evlist__id2evsel (session->evlist, event->stat.id); +PyObject *pyevent =pyrf_event__new (event, evsel, psession->session,/*machine=*/ NULL ); +const char *name = evsel ?evsel__name (evsel) :"unknown" ; +PyObject *ret; + +if (pyevent ==NULL ) +return -ENOMEM ; + + ret =PyObject_CallFunction (psession->stat,"Oz" , pyevent, name); +if (!ret) { +Py_DECREF (pyevent); +return -1 ; + } +Py_DECREF (ret); +Py_DECREF (pyevent); +return 0 ; +} + +static int pyrf_session_tool__stat_round (const struct perf_tool *tool, +struct perf_session *session__maybe_unused , +union perf_event *event) +{ + struct pyrf_session *psession = container_of(tool, struct pyrf_session, tool); +PyObject *pyevent =pyrf_event__new (event,/*evsel=*/ NULL , psession->session,/*machine=*/ NULL ); +PyObject *ret; + +if (pyevent ==NULL ) +return -ENOMEM ; + + ret =PyObject_CallFunction (psession->stat,"Oz" , pyevent,NULL ); +if (!ret) { +Py_DECREF (pyevent); +return -1 ; + } +Py_DECREF (ret); +Py_DECREF (pyevent); +return 0 ; +} +static PyObject *pyrf_session__find_thread (struct pyrf_session *psession,PyObject *args) {struct machine *machine;
@@ -3441,13 +3605,13 @@ static PyObject *pyrf_session__find_thread(struct pyrf_session *psession, PyObjestatic PyObject *pyrf_session__new (PyTypeObject *type,PyObject *args,PyObject *kwargs) {struct pyrf_data *pdata; -PyObject *sample =NULL ; -static char *kwlist[] = {"data" ,"sample" ,NULL }; +PyObject *sample =NULL , *stat =NULL ; +static char *kwlist[] = {"data" ,"sample" ,"stat" ,NULL };struct pyrf_session *psession;struct perf_session *session; -if (!PyArg_ParseTupleAndKeywords (args, kwargs,"O!|O" , kwlist, &pyrf_data__type, &pdata, - &sample)) +if (!PyArg_ParseTupleAndKeywords (args, kwargs,"O!|OO" , kwlist, &pyrf_data__type, &pdata, + &sample, &stat))return NULL ; psession = PyObject_New(struct pyrf_session, type);
@@ -3456,6 +3620,7 @@ static PyObject *pyrf_session__new(PyTypeObject *type, PyObject *args, PyObject psession->session =NULL ; psession->sample =NULL ; + psession->stat =NULL ; psession->pdata =NULL ;Py_INCREF (pdata);
@@ -3478,8 +3643,13 @@ static PyObject *pyrf_session__new(PyTypeObject *type, PyObject *args, PyObject }while (0 )ADD_TOOL (sample); +ADD_TOOL (stat);#undef ADD_TOOL +if (stat) + psession->tool.stat_round = pyrf_session_tool__stat_round; + + psession->tool.comm = perf_event__process_comm; psession->tool.mmap = perf_event__process_mmap; psession->tool.mmap2 = perf_event__process_mmap2;
@@ -3492,7 +3662,8 @@ static PyObject *pyrf_session__new(PyTypeObject *type, PyObject *args, PyObject psession->tool.build_id = perf_event__process_build_id; psession->tool.attr = perf_event__process_attr; psession->tool.feature = perf_event__process_feature; - psession->tool.stat = perf_event__process_stat_event; +if (!stat) + psession->tool.stat = perf_event__process_stat_event; session =perf_session__new (&pdata->data, &psession->tool);if (IS_ERR (session)) {PyErr_Format (PyExc_IOError,"failed to create session: %ld" ,PTR_ERR (session));
@@ -3522,6 +3693,7 @@ static void pyrf_session__delete(struct pyrf_session *psession)perf_session__delete (psession->session);Py_XDECREF (psession->pdata);Py_XDECREF (psession->sample); +Py_XDECREF (psession->stat);Py_TYPE (psession)->tp_free ((PyObject *)psession); }
@@ -3709,6 +3881,12 @@ PyMODINIT_FUNC PyInit_perf(void)Py_INCREF (&pyrf_context_switch_event__type);PyModule_AddObject (module,"switch_event" , (PyObject *)&pyrf_context_switch_event__type); +Py_INCREF (&pyrf_stat_event__type); +PyModule_AddObject (module,"stat_event" , (PyObject *)&pyrf_stat_event__type); + +Py_INCREF (&pyrf_stat_round_event__type); +PyModule_AddObject (module,"stat_round_event" , (PyObject *)&pyrf_stat_round_event__type); +Py_INCREF (&pyrf_thread_map__type);PyModule_AddObject (module,"thread_map" , (PyObject *)&pyrf_thread_map__type);
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
Ian Rogers
reviewerPATCH v14 15/19 perf python: Expose brstack in sample event
In reply to Ian RogersImplement pyrf_branch_entry and pyrf_branch_stack for lazy iteration over branch stack entries.
---
v2:
1. Avoided Keyword Collision: Renamed the properties "from" and "to" to "from_ip" and "to_ip" in pyrf_branch_entry__getset[] to prevent syntax errors in Python.
2. Eager Branch Stack Resolution: Updated pyrf_session_tool__sample() to allocate and copy the branch stack entries eagerly when the event is processed, instead of deferring it to iteration time. This avoids reading from potentially overwritten or unmapped mmap buffers.
3. Updated Iterators: Updated pyrf_branch_stack and pyrf_branch_stack__next() to use the copied entries rather than pointing directly to the sample's buffer.
4. Avoided Reference Leak on Init Failure: Added proper error checking for PyModule_AddObject() in the module initialization function, decrementing references on failure.
v6:
- Moved branch stack resolution from session_tool__sample to pyrf_event__new.
--- tools/perf/util/python.c | 198 +++++++++++++++++++++++++++++++++++++++ 1 file changed, 198 insertions(+)
tools/perf/util/python.c +198 -0
@@ -22,6 +22,7 @@#include "debug.h" #include "dso.h" #include "event.h" +#include "branch.h" #include "evlist.h" #include "evsel.h" #include "expr.h"
@@ -89,6 +90,8 @@ struct pyrf_event {bool al_resolved;/** @callchain: Resolved callchain, eagerly computed if requested. */ PyObject *callchain; +/** @brstack: Resolved branch stack, eagerly computed if requested. */ +PyObject *brstack;/** @event: The underlying perf_event that may be in a file or ring buffer. */ union perf_event event; };
@@ -127,6 +130,7 @@ static void pyrf_event__delete(struct pyrf_event *pevent)if (pevent->al_resolved)addr_location__exit (&pevent->al);Py_XDECREF (pevent->callchain); +Py_XDECREF (pevent->brstack);perf_sample__exit (&pevent->sample);Py_TYPE (pevent)->tp_free ((PyObject *)pevent); }
@@ -997,6 +1001,154 @@ static PyObject *pyrf_sample_event__get_callchain(PyObject *self, void *closurereturn pevent->callchain; } +struct pyrf_branch_entry { +PyObject_HEAD + u64 from; +u64 to; +struct branch_flags flags; +}; + +static void pyrf_branch_entry__delete (struct pyrf_branch_entry *pentry) +{ +Py_TYPE (pentry)->tp_free ((PyObject *)pentry); +} + +static PyObject *pyrf_branch_entry__get_from (struct pyrf_branch_entry *pentry, +void *closure__maybe_unused ) +{ +return PyLong_FromUnsignedLongLong (pentry->from); +} + +static PyObject *pyrf_branch_entry__get_to (struct pyrf_branch_entry *pentry, +void *closure__maybe_unused ) +{ +return PyLong_FromUnsignedLongLong (pentry->to); +} + +static PyObject *pyrf_branch_entry__get_mispred (struct pyrf_branch_entry *pentry, +void *closure__maybe_unused ) +{ +return PyBool_FromLong (pentry->flags.mispred); +} + +static PyObject *pyrf_branch_entry__get_predicted (struct pyrf_branch_entry *pentry, +void *closure__maybe_unused ) +{ +return PyBool_FromLong (pentry->flags.predicted); +} + +static PyObject *pyrf_branch_entry__get_in_tx (struct pyrf_branch_entry *pentry, +void *closure__maybe_unused ) +{ +return PyBool_FromLong (pentry->flags.in_tx); +} + +static PyObject *pyrf_branch_entry__get_abort (struct pyrf_branch_entry *pentry, +void *closure__maybe_unused ) +{ +return PyBool_FromLong (pentry->flags.abort); +} + +static PyObject *pyrf_branch_entry__get_cycles (struct pyrf_branch_entry *pentry, +void *closure__maybe_unused ) +{ +return PyLong_FromUnsignedLongLong (pentry->flags.cycles); +} + +static PyObject *pyrf_branch_entry__get_type (struct pyrf_branch_entry *pentry, +void *closure__maybe_unused ) +{ +return PyLong_FromUnsignedLongLong ((unsigned long long )pentry->flags.type); +} + +static PyGetSetDef pyrf_branch_entry__getset[] = { + { .name ="from_ip" , .get = (getter )pyrf_branch_entry__get_from, }, + { .name ="to_ip" , .get = (getter )pyrf_branch_entry__get_to, }, + { .name ="mispred" , .get = (getter )pyrf_branch_entry__get_mispred, }, + { .name ="predicted" , .get = (getter )pyrf_branch_entry__get_predicted, }, + { .name ="in_tx" , .get = (getter )pyrf_branch_entry__get_in_tx, }, + { .name ="abort" , .get = (getter )pyrf_branch_entry__get_abort, }, + { .name ="cycles" , .get = (getter )pyrf_branch_entry__get_cycles, }, + { .name ="type" , .get = (getter )pyrf_branch_entry__get_type, }, + { .name =NULL , }, +}; + +static PyTypeObject pyrf_branch_entry__type = { +PyVarObject_HEAD_INIT (NULL ,0 ) + .tp_name ="perf.branch_entry" , + .tp_basicsize =sizeof (struct pyrf_branch_entry ), + .tp_dealloc = (destructor )pyrf_branch_entry__delete, + .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, + .tp_doc ="perf branch entry object." , + .tp_getset = pyrf_branch_entry__getset, +}; + +struct pyrf_branch_stack { +PyObject_HEAD + struct branch_entry *entries; +u64 nr; +}; + +static void pyrf_branch_stack__delete (struct pyrf_branch_stack *pstack) +{ +free (pstack->entries); +Py_TYPE (pstack)->tp_free ((PyObject *)pstack); +} + +static Py_ssize_t pyrf_branch_stack__length (PyObject *obj) +{ +struct pyrf_branch_stack *pstack = (void *)obj; + +return pstack->nr; +} + +static PyObject *pyrf_branch_stack__item (PyObject *obj,Py_ssize_t i) +{ +struct pyrf_branch_stack *pstack = (void *)obj; +struct pyrf_branch_entry *pentry; + +if (i <0 || i >= (Py_ssize_t )pstack->nr) { +PyErr_SetString (PyExc_IndexError,"Index out of range" ); +return NULL ; + } + + pentry = PyObject_New(struct pyrf_branch_entry, &pyrf_branch_entry__type); +if (!pentry) +return NULL ; + + pentry->from = pstack->entries[i].from; + pentry->to = pstack->entries[i].to; + pentry->flags = pstack->entries[i].flags; + +return (PyObject *)pentry; +} + +static PySequenceMethods pyrf_branch_stack__sequence_methods = { + .sq_length = pyrf_branch_stack__length, + .sq_item = pyrf_branch_stack__item, +}; + +static PyTypeObject pyrf_branch_stack__type = { +PyVarObject_HEAD_INIT (NULL ,0 ) + .tp_name ="perf.branch_stack" , + .tp_basicsize =sizeof (struct pyrf_branch_stack ), + .tp_dealloc = (destructor )pyrf_branch_stack__delete, + .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, + .tp_doc ="perf branch stack object." , + .tp_as_sequence = &pyrf_branch_stack__sequence_methods, +}; + +static PyObject *pyrf_sample_event__get_brstack (PyObject *self,void *closure__maybe_unused ) +{ +struct pyrf_event *pevent = (void *)self; + +if (!pevent->brstack) + Py_RETURN_NONE; + +Py_INCREF (pevent->brstack); +return pevent->brstack; +} +static PyObject *pyrf_sample_event__getattro (struct pyrf_event *pevent,PyObject *attr_name) {
@@ -1017,6 +1169,12 @@ static PyGetSetDef pyrf_sample_event__getset[] = { .set = NULL, .doc = "event callchain.", }, + { + .name = "brstack", + .get = pyrf_sample_event__get_brstack, + .set = NULL, + .doc = "event branch stack.", + }, { .name = "raw_buf", .get = (getter)pyrf_sample_event__get_raw_buf,
@@ -1198,6 +1356,12 @@ static int pyrf_event__setup_types(void) err =PyType_Ready (&pyrf_callchain__type);if (err <0 ) goto out; + err =PyType_Ready (&pyrf_branch_entry__type); +if (err <0 ) + goto out; + err =PyType_Ready (&pyrf_branch_stack__type); +if (err <0 ) + goto out; out:return err; }
@@ -1277,6 +1441,7 @@ static PyObject *pyrf_event__new(const union perf_event *event, struct evsel *evperf_sample__init (&pevent->sample,/*all=*/ true); pevent->callchain =NULL ; + pevent->brstack =NULL ; pevent->al_resolved = false;addr_location__init (&pevent->al);
@@ -1334,6 +1499,27 @@ static PyObject *pyrf_event__new(const union perf_event *event, struct evsel *evaddr_location__exit (&al); } } +if (sample->branch_stack) { +struct branch_stack *bs = sample->branch_stack; +struct branch_entry *entries =perf_sample__branch_entries (sample); +struct pyrf_branch_stack *pstack; + + pstack = PyObject_New(struct pyrf_branch_stack, &pyrf_branch_stack__type); +if (!pstack) { +Py_DECREF (pevent); +return NULL ; + } + pstack->nr = bs->nr; + pstack->entries =calloc (bs->nr,sizeof (struct branch_entry )); +if (!pstack->entries) { +Py_DECREF (pstack); +Py_DECREF (pevent); +return PyErr_NoMemory (); + } +memcpy (pstack->entries, entries, + bs->nr *sizeof (struct branch_entry )); + pevent->brstack = (PyObject *)pstack; + }return (PyObject *)pevent; }
@@ -3902,6 +4088,18 @@ PyMODINIT_FUNC PyInit_perf(void)Py_INCREF (&pyrf_session__type);PyModule_AddObject (module,"session" , (PyObject *)&pyrf_session__type); +Py_INCREF (&pyrf_branch_entry__type); +if (PyModule_AddObject (module,"branch_entry" , (PyObject *)&pyrf_branch_entry__type) <0 ) { +Py_DECREF (&pyrf_branch_entry__type); + goto error; + } + +Py_INCREF (&pyrf_branch_stack__type); +if (PyModule_AddObject (module,"branch_stack" , (PyObject *)&pyrf_branch_stack__type) <0 ) { +Py_DECREF (&pyrf_branch_stack__type); + goto error; + } + dict =PyModule_GetDict (module);if (dict ==NULL ) goto error;
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
Ian Rogers
reviewerPATCH v14 16/19 perf python: Add syscall name/id to convert syscall number and name
In reply to Ian RogersUse perf's syscalltbl support to convert syscall number to name assuming the number is for the host machine. This avoids python libaudit support as tools/perf/scripts/python/syscall-counts.py requires.
---
v2:
1. Guarded with HAVE_LIBTRACEEVENT : Wrapped the syscall functions and their entries in perf__methods with #ifdef HAVE_LIBTRACEEVENT to avoid potential linker errors if CONFIG_TRACE is disabled.
2. Changed Exception Type: Updated pyrf__syscall_id to raise a ValueError instead of a TypeError when a syscall is not found.
3. Fixed Typo: Corrected "name number" to "name" in the docstring for syscall_id.
v6:
- Added optional keyword-only elf_machine argument to syscall_name and syscall_id.
v7:
- Made syscalltbl.o unconditional in Build to fix undefined symbol when building without libtraceevent.
--- tools/perf/util/Build | 1 - tools/perf/util/python.c | 48 ++++++++++++++++++++++++++++++++++++++++ 2 files changed, 48 insertions(+), 1 deletion(-)
tools/perf/util/Build +0 -1
@@ -75,7 +75,6 @@ perf-util-y += sample.o perf-util-y += sample-raw.o perf-util-y += s390-sample-raw.o perf-util-y += amd-sample-raw.o - perf-util-y += ordered-events.o perf-util-y += namespaces.o perf-util-y += comm.o
tools/perf/util/python.c +48 -0
72 lines not on this page · raw message
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
[email protected]
AI reviewRe: [PATCH v14 16/19] perf python: Add syscall name/id to convert syscall number and name
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider: - [Medium] The commit introduces a new Python C API (`perf.syscall_name` and `perf.syscall_id`) to replace…
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] The commit introduces a new Python C API (perf.syscall_name and perf.syscall_id) to replace the libaudit dependency, but fails to update the target python scripts to actually use the new API. --
commit 8ab24f7b7bd7192d5eaf878a71de9ce880b2422d
Author: Ian Rogers <[email protected]>
perf python: Add syscall name/id to convert syscall number and name
This commit introduces pyrf__syscall_name and pyrf__syscall_id to the perf python extension. These functions use syscalltbl to convert between syscall numbers and names, intending to replace the libaudit dependency.
Quoted 4 lines
Use perf's syscalltbl support to convert syscall number to name assuming the number is for the host machine. This avoids python libaudit support as
tools/perf/scripts/python/syscall-counts.pyrequires.
[Severity: Medium]
Is this patch missing the corresponding updates to tools/perf/scripts/python/syscall-counts.py or the Perf-Trace-Util library?
The commit message mentions avoiding the libaudit dependency in the script, but the patch only introduces the C API without modifying the target scripts. The script itself doesn't appear to be updated to use the newly added perf.syscall_name() and perf.syscall_id() functions.
Should the script modifications be included here to actually complete the transition away from the audit python module?
Quoted 3 lines
tools/perf/util/python.c +0
-0
[ ... ]
Quoted 15 lines
tools/perf/util/python.c +15
-0+static PyObject *pyrf__syscall_name (PyObject *self,PyObject *args,PyObject *kwargs) +{ +const char *name; +int id; +int elf_machine =EM_HOST ; +static char *kwlist[] = {"id" ,"elf_machine" ,NULL }; + +if (!PyArg_ParseTupleAndKeywords (args, kwargs,"i|$i" , kwlist, &id, &elf_machine)) +return NULL ; + + name =syscalltbl__name (elf_machine, id); +if (!name) + Py_RETURN_NONE; +return PyUnicode_FromString (name); +}
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=16
Ian Rogers
reviewerPATCH v14 17/19 perf python: Add config file access
In reply to Ian RogersAdd perf.config_get(name) to expose the perf configuration system.
--- tools/perf/util/python.c | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+)
tools/perf/util/python.c +20 -0
@@ -17,6 +17,7 @@#include "build-id.h" #include "callchain.h" #include "comm.h" +#include "config.h" #include "counts.h" #include "data.h" #include "debug.h"
@@ -3992,7 +3993,26 @@ static PyObject *pyrf__syscall_id(PyObject *self, PyObject *args, PyObject *kwarreturn PyLong_FromLong (id); } +static PyObject *pyrf__config_get (PyObject *self,PyObject *args) +{ +const char *config_name, *val; + +if (!PyArg_ParseTuple (args,"s" , &config_name)) +return NULL ; + + val =perf_config_get (config_name); +if (!val) + Py_RETURN_NONE; +return PyUnicode_FromString (val); +} +static PyMethodDef perf__methods[] = { + { + .ml_name ="config_get" , + .ml_meth = (PyCFunction ) pyrf__config_get, + .ml_flags =METH_VARARGS , + .ml_doc =PyDoc_STR ("Get a perf config value." ) + }, { .ml_name ="metrics" , .ml_meth = (PyCFunction ) pyrf__metrics,
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
Ian Rogers
reviewerPATCH v14 18/19 perf python: Add perf.pyi stubs file
In reply to Ian RogersAdd Python type stubs for the perf module to improve IDE support and static analysis. Includes docstrings for classes, methods, and constants derived from C source and JSON definitions.
---
v12:
- Audited all perf.pyi event stubs: added full sample_members attributes (pid, tid, time, id, stream_id, period, cpu) to all payload events via inheritance. Fixed sample_event unique fields (ip, addr, phys_addr, weight, data_src, insn_cnt, cyc_cnt). Renamed mmap_event.addr to start. Removed read_event.value.
v2:
1. Added Missing Module Functions: Added parse_metrics and pmus. Renamed metrics to parse_metrics to match python.c .
2. Added Constructors: Added __init__ methods for data , evsel , and evlist with their appropriate arguments.
3. Removed sample_comm : Removed it from sample_event since it is not exported in python.c .
4. Keyword Handling in branch_entry : I used from_ip and to_ip in the stubs to match the rename I did in python.c (in turn 145) to avoid the Python from keyword conflict.
5. Added Missing Event Classes: Added mmap_event , lost_event , comm_event , task_event , throttle_event , read_event , and switch_event .
6. Added Missing evlist Methods: Added get_pollfd and add .
7. Updated Return Types: Changed process_events to return int .
v6:
- Updated perf.pyi to use find_thread and elf_machine.
v8:
- Added mmap2_event class and new evsel attributes to perf.pyi.
- Added pid, tid, ppid, cpu attributes to class thread in perf.pyi.
--- tools/perf/Makefile.perf | 7 +- tools/perf/python/perf.pyi | 646 +++++++++++++++++++++++++++++++++++++ tools/perf/util/python.c | 2 + tools/perf/util/setup.py | 5 + 4 files changed, 658 insertions(+), 2 deletions(-) create mode 100644 tools/perf/python/perf.pyi
tools/perf/Makefile.perf +5 -2
@@ -399,7 +399,7 @@ PYTHON_EXTBUILD_LIB := $(PYTHON_EXTBUILD)lib/ PYTHON_EXTBUILD_TMP := $(PYTHON_EXTBUILD)tmp/ export PYTHON_EXTBUILD_LIB PYTHON_EXTBUILD_TMP -python-clean := $(call QUIET_CLEAN, python) $(RM) -r $(PYTHON_EXTBUILD) $(OUTPUT)python/perf*.so +python-clean := $(call QUIET_CLEAN, python) $(RM) -r $(PYTHON_EXTBUILD) $(OUTPUT)python/perf*.so $(OUTPUT)python/perf.pyi ifneq ($(quiet),) python_setup_quiet=--quiet
@@ -522,7 +522,10 @@ $(OUTPUT)python/perf$(PYTHON_EXTENSION_SUFFIX): util/python.c util/setup.py $(PE CFLAGS='$(CFLAGS)' LDFLAGS='$(LDFLAGS) $(LIBS_PY)' \ $(PYTHON_WORD) util/setup.py \ $(python_setup_quiet) build_ext; \ - cp $(PYTHON_EXTBUILD_LIB)perf*.so $(OUTPUT)python/ + cp $(PYTHON_EXTBUILD_LIB)perf*.so $(OUTPUT)python/; \ + if [ "$(abspath python/perf.pyi)" != "$(abspath $(OUTPUT)python/perf.pyi)" ]; then \ + cp python/perf.pyi $(OUTPUT)python/; \ + fi python_perf_target: @echo "Target is: $(OUTPUT)python/perf$(PYTHON_EXTENSION_SUFFIX)"
tools/perf/python/perf.pyi new file +646 -0
646 lines not on this page · raw message
tools/perf/util/python.c +2 -0
@@ -433,6 +433,7 @@ static const char pyrf_lost_event__doc[] = PyDoc_STR("perf lost event object.");static PyMemberDef pyrf_lost_event__members[] = { sample_members +member_def (perf_event_header, type,T_UINT ,"event type" ),member_def (perf_record_lost, id,T_ULONGLONG ,"event id" ),member_def (perf_record_lost, lost,T_ULONGLONG ,"number of lost events" ), { .name =NULL , },
@@ -541,6 +542,7 @@ static const char pyrf_read_event__doc[] = PyDoc_STR("perf read event object.");static PyMemberDef pyrf_read_event__members[] = { sample_members +member_def (perf_event_header, type,T_UINT ,"event type" ),member_def (perf_record_read, pid,T_UINT ,"event pid" ),member_def (perf_record_read, tid,T_UINT ,"event tid" ), { .name =NULL , },
tools/perf/util/setup.py +5 -0
@@ -2,6 +2,7 @@ from os import getenv, path from subprocess import Popen, PIPE from re import sub import shlex +import shutil cc = getenv("CC") assert cc, "Environment variable CC not set"
@@ -73,6 +74,10 @@ class install_lib(_install_lib): _install_lib.finalize_options(self) self.build_dir = build_lib + def run(self): + _install_lib.run(self) + shutil.copy2(f'{src_perf}/python/perf.pyi', self.install_dir) + # switch off several checks (need to be at the end of cflags list) extra_cflags = ['-fno-strict-aliasing', '-Wno-write-strings', '-Wno-unused-parameter', '-Wno-redundant-decls' ]
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
[email protected]
AI reviewRe: [PATCH v14 18/19] perf python: Add perf.pyi stubs file
Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider: - [Medium] Incomplete and inaccurate Python type stubs misrepresent the C extension. - [Low] Missing file…
Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [Medium] Incomplete and inaccurate Python type stubs misrepresent the C extension.
- [Low] Missing file tracking for perf.pyi in Makefile dependency list. --
commit 0becfda4334bc6af9250ae05c834cef6f97aa5ca
Author: Ian Rogers <[email protected]>
perf python: Add perf.pyi stubs file
Add Python type stubs for the perf module to improve IDE support and static analysis. Includes docstrings for classes, methods, and constants derived from C source and JSON definitions.
Quoted 4 lines
tools/perf/Makefile.perf +0
-0
[ ... ]
tools/perf/Makefile.perf $(OUTPUT)python/perf$(PYTHON_EXTENSION_SUFFIX): util/python.c util/setup.py $(PE
Quoted 1 line
tools/perf/Makefile.perf +0
-0@@ -522,7 +522,10 @@ $(OUTPUT)python/perf$(PYTHON_EXTENSION_SUFFIX): util/python.c util/setup.py $(PE
[Severity: Low]
Does python/perf.pyi need to be added to the prerequisite list for this target?
If a developer edits the perf.pyi stub file without modifying util/python.c or setup.py, it seems Make will consider the target up-to-date and won't copy the updated stubs to the output directory.
Quoted 11 lines
tools/perf/Makefile.perf +4
-1CFLAGS='$(CFLAGS)' LDFLAGS='$(LDFLAGS) $(LIBS_PY)' \ $(PYTHON_WORD) util/setup.py \ $(python_setup_quiet) build_ext; \ - cp $(PYTHON_EXTBUILD_LIB)perf*.so $(OUTPUT)python/ + cp $(PYTHON_EXTBUILD_LIB)perf*.so $(OUTPUT)python/; \ + if [ "$(abspath python/perf.pyi)" != "$(abspath $(OUTPUT)python/perf.pyi)" ]; then \ + cp python/perf.pyi $(OUTPUT)python/; \ + fi python_perf_target: @echo "Target is: $(OUTPUT)python/perf$(PYTHON_EXTENSION_SUFFIX)"
[ ... ]
Quoted 5 lines
tools/perf/python/perf.pyi new file +0
-0
[ ... ]
tools/perf/python/perf.pyi line 99
Quoted 5 lines
tools/perf/python/perf.pyi new file +5
-0+class counts_values: + """Raw counter values.""" + val: int + ena: int + run: int
[Severity: Medium]
Are there missing properties here? The C extension seems to define additional properties like id, lost, and values for counts_values. Missing them in the stub might cause false positives in static analysis.
[ ... ]
tools/perf/python/perf.pyi line 186
Quoted 3 lines
tools/perf/python/perf.pyi new file +3
-0+class sample_event(_sample_members): + """Represents a sample event from perf.""" + evsel: evsel
[Severity: Medium]
Are there missing properties and methods here? The C extension explicitly defines raw_buf, dso, map_start, srccode(), and insn(). Omitting them here could cause static analysis tools and IDEs to report errors on valid code.
[ ... ]
Quoted 5 lines
tools/perf/python/perf.pyi new file +5
-0+class branch_stack: + """Iterator over branch entries in the branch stack.""" + def __iter__(self) -> Iterator[branch_entry]: ... + def __next__(self) -> branch_entry: ... +
tools/perf/python/perf.pyi line 306
Quoted 10 lines
tools/perf/python/perf.pyi new file +10
-0+class callchain_node: + """Represents a frame in the callchain.""" + ip: int + sym: Optional[Any] + map: Optional[Any] + +class callchain: + """Iterator over callchain frames.""" + def __iter__(self) -> Iterator[callchain_node]: ... + def __next__(self) -> callchain_node: ...
[Severity: Medium]
Does the C extension actually implement __next__ for branch_stack and callchain? It looks like the C code implements them as sequences (__len__, __getitem__) rather than iterators. Declaring __next__ here might mislead IDEs and type checkers.
[ ... ]
tools/perf/python/perf.pyi line 321
Quoted 4 lines
tools/perf/python/perf.pyi new file +4
-0+class stat_round_event: + """Represents a stat round event from perf.""" + type: int + time: int
[Severity: Medium]
Should stat_round_event inherit from _sample_members? The C extension seems to include those properties, but the stub doesn't reflect that.
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=18
Ian Rogers
reviewerPATCH v14 19/19 perf python: Add LiveSession helper
In reply to Ian RogersAdd LiveSession class in tools/perf/python/perf_live.py to support live event collection using perf.evlist and perf.parse_events, avoiding the need to fork a separate perf record process.
---
v12:
- Replaced except TypeError with continue to ignore unmapped/offline CPUs instead of breaking the poll loop. Replaced hardcoded -11 with e.errno == errno.EAGAIN in LiveSession.run(). Fixed setup_python.sh PYTHONPATH prepend logic to cleanly append to existing paths.
v2:
1. Fixed File Descriptor Leak: I moved self.evlist.mmap() inside the try block so that if it raises an exception, the finally block will still be executed and call self.evlist.close() , preventing file descriptor leaks.
2. Handled InterruptedError in poll() : I wrapped the poll() call in a try-except block to catch InterruptedError and continue the loop. This prevents the live session from crashing on non-fatal signals like SIGWINCH .
3. Added evlist.config() : I added a call to self.evlist.config() in the constructor after parse_events() . This applies the default record options to the events, enabling sampling and setting up PERF_SAMPLE_* fields so that the kernel will actually generate PERF_RECORD_SAMPLE events.
4. Enable the evlist and be robust to exceptions from reading unsupported events like mmap2.
v8:
- Drain all events from a CPU before moving to the next.
--- tools/perf/python/perf_live.py | 55 ++++++++++++++++++++++ tools/perf/tests/shell/lib/setup_python.sh | 13 +++++ 2 files changed, 68 insertions(+) create mode 100755 tools/perf/python/perf_live.py
tools/perf/python/perf_live.py new file +55 -0
@@ -0,0 +1,55 @@ +# SPDX-License-Identifier: GPL-2.0 +""" +Live event session helper using perf.evlist. + +This module provides a LiveSession class that allows running a callback +for each event collected live from the system, similar to perf.session +but without requiring a perf.data file. +""" + +import errno +import perf + + +class LiveSession: + """Represents a live event collection session.""" + + def __init__(self, event_string: str, sample_callback): + self.event_string = event_string + self.sample_callback = sample_callback + # Create a cpu map for all online CPUs + self.cpus = perf.cpu_map() + # Parse events and set maps + self.evlist = perf.parse_events(self.event_string, self.cpus) + self.evlist.config() + + def run(self): + """Run the live session.""" + try: + self.evlist.open() + self.evlist.mmap() + self.evlist.enable() + + while True: + # Poll for events with 100ms timeout + try: + self.evlist.poll(100) + except InterruptedError: + continue + for cpu in self.cpus: + for _ in range(1000): # Limit to 1000 events per CPU per poll to prevent starvation + try: + event = self.evlist.read_on_cpu(cpu) + except TypeError: + # CPU might be unmapped or offline, wait for mmap event + break + + if event is None: + break + + if event.type == perf.RECORD_SAMPLE: + self.sample_callback(event) + except KeyboardInterrupt: + pass + finally: + self.evlist.close()
tools/perf/tests/shell/lib/setup_python.sh +13 -0
@@ -14,3 +14,16 @@ then echo Skipping test, python not detected please set environment variable PYTHON. exit 2 fi + +# Set PYTHONPATH to find the in-tree built perf.so first, avoiding system-wide perf.so +if [ -n "$PERF_EXEC_PATH" ] && [ -d "$PERF_EXEC_PATH/python" ]; then + PYTHONPATH_DIR="$PERF_EXEC_PATH/python" +elif [ -d "$(dirname "$0")/../../python" ]; then + PYTHONPATH_DIR="$(dirname "$0")/../../python" +elif [ -d "$(dirname "$0")/../python" ]; then + PYTHONPATH_DIR="$(dirname "$0")/../python" +fi + +if [ -n "$PYTHONPATH_DIR" ]; then + export PYTHONPATH="$PYTHONPATH_DIR${PYTHONPATH:+:$PYTHONPATH}" +fi
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
[email protected]
AI reviewRe: [PATCH v14 19/19] perf python: Add LiveSession helper
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider: - [High] LiveSession.run() crashes with an unhandled OSError when read_on_cpu() encounters an empty ring…
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] LiveSession.run() crashes with an unhandled OSError when read_on_cpu() encounters an empty ring buffer. --
commit 4826fb91a8d2846998d0841e7b023eb935dd9b06
Author: Ian Rogers <[email protected]>
perf python: Add LiveSession helper
This commit adds a LiveSession class in tools/perf/python/perf_live.py to support live event collection using perf.evlist and perf.parse_events. This avoids the need to fork a separate perf record process.
Quoted 5 lines
tools/perf/python/perf_live.py new file +0
-0
[ ... ]
Quoted 3 lines
tools/perf/python/perf_live.py new file +3
-0+ def run(self): + """Run the live session.""" + try:
[ ... ]
tools/perf/python/perf_live.py line 45
Quoted 7 lines
tools/perf/python/perf_live.py new file +7
-0+ for cpu in self.cpus: + for _ in range(1000): # Limit to 1000 events per CPU per poll to prevent starvation + try: + event = self.evlist.read_on_cpu(cpu) + except TypeError: + # CPU might be unmapped or offline, wait for mmap event + break
[Severity: High]
Will this loop crash with an unhandled exception if the ring buffer is empty?
The underlying C function perf_mmap__read_init() returns -EAGAIN when the buffer is empty, which the Python C extension pyrf_evlist__read_on_cpu() translates into an OSError.
Since this try block only catches TypeError, won't the OSError bubble up and abruptly terminate the live event collection loop?
Should this also catch OSError to cleanly break out of the loop instead?
Quoted 3 lines
tools/perf/python/perf_live.py new file +3
-0+ + if event is None: + break
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=19
Ian Rogers
reviewerv15 cover Add additional python API support
In reply to Ian RogersThe perf script command has long supported running Python and Perl scripts by embedding libpython and libperl. This approach has several drawbacks:
- overhead by creating Python dictionaries for every event (whether used or not),
- complex build dependencies on specific Python/Perl versions,
- complications with threading due to perf being the interpreter,
- no clear way to run standalone scripts like ilist.py.
This series takes a different approach with some initial implementation posted as an RFC last October:
https://lore.kernel.org/linux-perf-users/[email protected]/
with the motivation coming up on the mailing list earlier:
https://lore.kernel.org/lkml/CAP-5=fWDqE8SYfOLZkg_0=4Ayx6E7O+h7uUp4NDeCFkiN4b7-w@mail.gmail.com/
The ultimate goal is to remove the embedded libpython and libperl support from perf entirely, expanding the existing perf Python module to provide full access to perf data files and events, allowing scripts to be run as standalone Python applications.
To make the review process more manageable, the original 58-patch series has been split. This v15 series represents "Phase 1: API & Infrastructure" (19 patches). The first 4 patches of Phase 1 (cleanups and arch-specific header sorting) have already been merged upstream.
This remaining set contains:
1. Missed explicit dependency cleanups and header sorting for util/ and python.
2. Crucial core safety infrastructure (reference counting for evlist/evsel) to support safe lifecycle management in garbage-collected Python.
3. The core Python API extensions (session wrappers, perf_data wrappers, sample accessors, stubs, and LiveSession helper).
The subsequent "Phase 2" series will contain the actual porting of all existing Python/Perl scripts to the new API (which yields up to 35x speedups as demonstrated previously) and the final removal of embedded interpreters.
Addressing v14 Review Feedback:
- Patch 7 (evlist refcounting): The TOCTOU race condition and asymmetric drop cycle-leak mentioned in the review are known limitations of this transitional cycle-breaking logic. A proper, thread-safe fix requires a global design change that is out of scope for Phase 1. Since Python script execution runs under the GIL, the TOCTOU race cannot manifest in practice. These structural cycle-breaking issues will be fully addressed in Phase 2.
- Patch 16 (syscall name/id): The script modifications (syscall-counts.py) are intentionally omitted from this patch. As stated, the porting of scripts to the new API will be done entirely in Phase 2. The commit message simply motivates the API addition.
- Patch 19 (LiveSession helper): The review raised concerns about unhandled OSError (-EAGAIN) abruptly terminating the loop. This was addressed in v13 where pyrf_evlist__read_on_cpu was explicitly changed to natively return Py_RETURN_NONE on -EAGAIN. The check if event is None: break correctly handles the empty ring buffer scenario cleanly without raising exceptions.
---
v15 Changes
-----------
- Patch 11: Removed manual '\0' assignment for mmap/comm filenames which corrupted the sample_id trailer. Replaced with memset of the event buffer prior to memcpy to safely handle non-terminated strings without overwriting trailer data.
- Patch 18: Added python/perf.pyi to Makefile.perf dependencies. Added missing properties to counts_values and sample_event stubs, fixed branch_stack/callchain sequence protocols, and added _sample_members inheritance to stat_round_event in perf.pyi.
v14 Changes
-----------
- Patch 6: Replaced Py_IsTrue with PyObject_IsTrue to correctly evaluate Python truthiness and added error handling returning -1.
- Patch 7: Added comments to evlist__put and evlist__splice_list_tail noting that cycle breaking and pointer rebinding for spliced events will be addressed in subsequent patches. Removed "Phase 2" from the previous cycle-breaking comment as it doesn't make sense in the wider Linux code base context.
- Patch 10: Fixed pyrf_session__new missing tool.tracing_data initialization and wrapped kernel map creation in an access check.
- Patch 15: Exposed remaining bitfields from struct branch_flags in pyrf_branch_entry.
- Patch 18: Cleaned up perf.pyi in Makefile.perf's python-clean target. Renamed sc_id to id in syscall_name stub, and corrected property types.
- Patch 19: Fixed perf_live.py inner loop to break instead of continue on TypeError to properly exit when processing is complete.
v13 Changes
-----------
- Patch 6: Reverted idx parsing for pyrf_evsel__init. Added explicit NULL checks for evlist in session.c to prevent crashes when accessing properties.
- Patch 7: Rely natively on evlist.c's updated logic through evlist__add.
- Patch 8: Fixed memory leak in intel_pt_synth_ptwrite_sample by adding perf_sample__exit(&sample) on exit.
- Patch 9: Zeroed out perf_data struct via memset(&data, 0, sizeof(data)) before using it to prevent stale state issues.
- Patch 11: Switched from PyLong_FromUnsignedLong to PyLong_FromUnsignedLongLong to guarantee full 64-bit bounds representation even on 32-bit platforms.
- Patch 18: Modified Makefile.perf to correctly install perf.pyi. Synchronized .pyi stubs to fully match the C implementation signatures.
- Patch 19: Updated pyrf_evlist__read_on_cpu to safely return Py_RETURN_NONE on -EAGAIN natively from C instead of relying on broken OSError emulation. Updated perf_live.py loop to handle the native None return correctly.
v12 Changes
-----------
- Patch 6: Restored idx parsing in pyrf_evsel__init kwargs, and removed an erroneous Py_INCREF(pevsel) in pyrf_evlist__add that caused a memory leak.
- Patch 7: Fixed evlist__put cycle collection TOCTOU double free race by checking refcount_read == 1. Fixed evlist__purge recursion and cycle reference tearing logic.
- Patch 8: Removed an extra evsel__get in evsel__parse_sample to fix a reference leak on error paths.
- Patch 11: Used copy_size to bound max_len for NUL termination to fix an out-of-bounds write in pyrf_event__new. Also set pevent->event.header.size = copy_size.
- Patch 12: Used copy_size to bound max_len for NUL termination of mmap2 and comm events to fix an out-of-bounds write.
- Patch 13: Added an optional struct machine * argument to pyrf_event__new defaulting to the host machine if NULL, avoiding regressions for future phases.
- Patch 14: Made psession->tool.stat = perf_event__process_stat_event; conditional on !stat so it doesn't unconditionally overwrite user stat callbacks.
- Patch 18: Audited all perf.pyi event stubs: added full sample_members attributes (pid, tid, time, id, stream_id, period, cpu) to all payload events via inheritance. Fixed sample_event unique fields (ip, addr, phys_addr, weight, data_src, insn_cnt, cyc_cnt). Renamed mmap_event.addr to start. Removed read_event.value.
- Patch 19: Replaced except TypeError with continue to ignore unmapped/offline CPUs instead of breaking the poll loop. Replaced hardcoded -11 with e.errno == errno.EAGAIN in LiveSession.run(). Fixed setup_python.sh PYTHONPATH prepend logic to cleanly append to existing paths.
Ian Rogers (19):
perf util: Sort includes and add missed explicit dependencies perf python: Add missed explicit dependencies
perf evsel/evlist: Avoid unnecessary #includes
perf data: Add open flag
perf evlist: Add reference count
perf evsel: Add reference count
perf evlist: Add reference count checking
perf python: Use evsel in sample in pyrf_event
perf python: Add wrapper for perf_data file abstraction perf python: Add python session abstraction wrapping perf's session perf python: Refactor and add accessors to sample event perf python: Add mmap2 event
perf python: Add callchain support
perf python: Extend API for stat events in python.c
perf python: Expose brstack in sample event
perf python: Add syscall name/id to convert syscall number and name perf python: Add config file access
perf python: Add perf.pyi stubs file
perf python: Add LiveSession helper
tools/perf/Makefile.perf | 9 +- tools/perf/arch/arm/util/cs-etm.c | 10 +- tools/perf/arch/arm64/util/arm-spe.c | 8 +- tools/perf/arch/arm64/util/hisi-ptt.c | 2 +- tools/perf/arch/x86/tests/hybrid.c | 22 +- tools/perf/arch/x86/tests/topdown.c | 4 +- tools/perf/arch/x86/util/auxtrace.c | 2 +- tools/perf/arch/x86/util/intel-bts.c | 6 +- tools/perf/arch/x86/util/intel-pt.c | 9 +- tools/perf/arch/x86/util/iostat.c | 14 +- tools/perf/bench/evlist-open-close.c | 29 +- tools/perf/builtin-annotate.c | 7 +- tools/perf/builtin-ftrace.c | 14 +- tools/perf/builtin-inject.c | 9 +- tools/perf/builtin-kvm.c | 14 +- tools/perf/builtin-kwork.c | 8 +- tools/perf/builtin-lock.c | 4 +- tools/perf/builtin-record.c | 95 +- tools/perf/builtin-report.c | 6 +- tools/perf/builtin-sched.c | 30 +- tools/perf/builtin-script.c | 15 +- tools/perf/builtin-stat.c | 83 +- tools/perf/builtin-top.c | 104 +- tools/perf/builtin-trace.c | 65 +- tools/perf/python/perf.pyi | 654 +++++ tools/perf/python/perf_live.py | 55 + tools/perf/tests/backward-ring-buffer.c | 26 +- tools/perf/tests/code-reading.c | 14 +- tools/perf/tests/event-times.c | 6 +- tools/perf/tests/event_update.c | 4 +- tools/perf/tests/evsel-roundtrip-name.c | 8 +- tools/perf/tests/evsel-tp-sched.c | 4 +- tools/perf/tests/expand-cgroup.c | 12 +- tools/perf/tests/hists_cumulate.c | 2 +- tools/perf/tests/hists_filter.c | 2 +- tools/perf/tests/hists_link.c | 2 +- tools/perf/tests/hists_output.c | 2 +- tools/perf/tests/hwmon_pmu.c | 7 +- tools/perf/tests/keep-tracking.c | 10 +- tools/perf/tests/mmap-basic.c | 24 +- tools/perf/tests/openat-syscall-all-cpus.c | 6 +- tools/perf/tests/openat-syscall-tp-fields.c | 26 +- tools/perf/tests/openat-syscall.c | 6 +- tools/perf/tests/parse-events.c | 139 +- tools/perf/tests/parse-metric.c | 8 +- tools/perf/tests/parse-no-sample-id-all.c | 2 +- tools/perf/tests/perf-record.c | 38 +- tools/perf/tests/perf-time-to-tsc.c | 12 +- tools/perf/tests/pfm.c | 12 +- tools/perf/tests/pmu-events.c | 11 +- tools/perf/tests/pmu.c | 4 +- tools/perf/tests/sample-parsing.c | 45 +- tools/perf/tests/shell/lib/setup_python.sh | 13 + tools/perf/tests/sw-clock.c | 20 +- tools/perf/tests/switch-tracking.c | 11 +- tools/perf/tests/task-exit.c | 20 +- tools/perf/tests/time-utils-test.c | 14 +- tools/perf/tests/tool_pmu.c | 7 +- tools/perf/tests/topology.c | 4 +- tools/perf/tests/uncore-event-sorting.c | 6 +- tools/perf/ui/browsers/annotate.c | 2 +- tools/perf/ui/browsers/hists.c | 22 +- tools/perf/util/Build | 1 - tools/perf/util/amd-sample-raw.c | 2 +- tools/perf/util/annotate-data.c | 2 +- tools/perf/util/annotate.c | 10 +- tools/perf/util/auxtrace.c | 14 +- tools/perf/util/block-info.c | 4 +- tools/perf/util/bpf_counter.c | 2 +- tools/perf/util/bpf_counter_cgroup.c | 14 +- tools/perf/util/bpf_ftrace.c | 9 +- tools/perf/util/bpf_lock_contention.c | 12 +- tools/perf/util/bpf_off_cpu.c | 44 +- tools/perf/util/bpf_trace_augment.c | 8 +- tools/perf/util/cgroup.c | 26 +- tools/perf/util/cs-etm.c | 5 +- tools/perf/util/data-convert-bt.c | 2 +- tools/perf/util/data.c | 27 +- tools/perf/util/data.h | 4 +- tools/perf/util/evlist.c | 496 ++-- tools/perf/util/evlist.h | 273 +- tools/perf/util/evsel.c | 39 +- tools/perf/util/evsel.h | 40 +- tools/perf/util/expr.c | 2 +- tools/perf/util/header.c | 69 +- tools/perf/util/header.h | 2 +- tools/perf/util/intel-pt.c | 2 +- tools/perf/util/intel-tpebs.c | 7 +- tools/perf/util/iostat.c | 2 +- tools/perf/util/iostat.h | 2 +- tools/perf/util/map.h | 9 +- tools/perf/util/metricgroup.c | 12 +- tools/perf/util/parse-events.c | 10 +- tools/perf/util/parse-events.y | 2 +- tools/perf/util/perf_api_probe.c | 20 +- tools/perf/util/pfm.c | 4 +- tools/perf/util/print-events.c | 2 +- tools/perf/util/python.c | 2614 ++++++++++++++++--- tools/perf/util/record.c | 11 +- tools/perf/util/s390-sample-raw.c | 20 +- tools/perf/util/sample-raw.c | 4 +- tools/perf/util/sample.c | 17 +- tools/perf/util/session.c | 69 +- tools/perf/util/session.h | 2 + tools/perf/util/setup.py | 5 + tools/perf/util/sideband_evlist.c | 40 +- tools/perf/util/sort.c | 2 +- tools/perf/util/stat-display.c | 6 +- tools/perf/util/stat-shadow.c | 24 +- tools/perf/util/stat.c | 20 +- tools/perf/util/stream.c | 4 +- tools/perf/util/synthetic-events.c | 11 +- tools/perf/util/time-utils.c | 12 +- tools/perf/util/top.c | 4 +- 114 files changed, 4432 insertions(+), 1441 deletions(-) create mode 100644 tools/perf/python/perf.pyi create mode 100755 tools/perf/python/perf_live.py
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
Ian Rogers
reviewerPATCH v15 1/19 perf util: Sort includes and add missed explicit dependencies
Fix missing includes found while cleaning the evsel/evlist header files. Sort the remaining header files for consistency with the rest of the code.
--- tools/perf/util/bpf_off_cpu.c | 30 +++++----- tools/perf/util/bpf_trace_augment.c | 8 +-- tools/perf/util/evlist.c | 91 +++++++++++++++-------------- tools/perf/util/map.h | 9 ++- tools/perf/util/perf_api_probe.c | 18 +++--- tools/perf/util/s390-sample-raw.c | 19 +++--- tools/perf/util/stat-shadow.c | 20 ++++--- tools/perf/util/stat.c | 16 +++-- 8 files changed, 113 insertions(+), 98 deletions(-)
tools/perf/util/bpf_off_cpu.c +16 -14
@@ -1,23 +1,25 @@// SPDX-License-Identifier: GPL-2.0 -#include "util/bpf_counter.h" -#include "util/debug.h" -#include "util/evsel.h" -#include "util/evlist.h" -#include "util/off_cpu.h" -#include "util/perf-hooks.h" -#include "util/record.h" -#include "util/session.h" -#include "util/target.h" -#include "util/cpumap.h" -#include "util/thread_map.h" -#include "util/cgroup.h" -#include "util/strlist.h" +#include <linux/time64.h> +#include <bpf/bpf.h> #include <bpf/btf.h> #include <internal/xyarray.h> -#include <linux/time64.h> +#include "bpf_counter.h" #include "bpf_skel/off_cpu.skel.h" +#include "cgroup.h" +#include "cpumap.h" +#include "debug.h" +#include "evlist.h" +#include "evsel.h" +#include "off_cpu.h" +#include "parse-events.h" +#include "perf-hooks.h" +#include "record.h" +#include "session.h" +#include "strlist.h" +#include "target.h" +#include "thread_map.h" #define MAX_STACKS 32#define MAX_PROC 4096
tools/perf/util/bpf_trace_augment.c +4 -4
@@ -1,11 +1,11 @@#include <bpf/libbpf.h> #include <internal/xyarray.h> -#include "util/debug.h" -#include "util/evlist.h" -#include "util/trace_augment.h" -#include "bpf_skel/augmented_raw_syscalls.skel.h" +#include "debug.h" +#include "evlist.h" +#include "parse-events.h" +#include "trace_augment.h" static struct augmented_raw_syscalls_bpf *skel;static struct evsel *bpf_output;
tools/perf/util/evlist.c +46 -45
113 lines not on this page · raw message
tools/perf/util/map.h +4 -5
@@ -2,14 +2,13 @@#ifndef __PERF_MAP_H#define __PERF_MAP_H -#include <linux/refcount.h> -#include <linux/compiler.h> -#include <linux/list.h> -#include <linux/rbtree.h> +#include <stdbool.h> #include <stdio.h> #include <string.h> -#include <stdbool.h> + +#include <linux/refcount.h> #include <linux/types.h> +#include <internal/rc_check.h> struct dso ;
tools/perf/util/perf_api_probe.c +11 -7
@@ -1,14 +1,18 @@/* SPDX-License-Identifier: GPL-2.0 */ +#include "perf_api_probe.h" -#include "perf-sys.h" -#include "util/cloexec.h" -#include "util/evlist.h" -#include "util/evsel.h" -#include "util/parse-events.h" -#include "util/perf_api_probe.h" -#include <perf/cpumap.h> #include <errno.h> +#include <perf/cpumap.h> + +#include "cloexec.h" +#include "evlist.h" +#include "evsel.h" +#include "parse-events.h" +#include "perf-sys.h" +#include "pmu.h" +#include "pmus.h" +typedef void (*setup_probe_fn_t )(struct evsel *evsel);static int perf_do_probe_api (setup_probe_fn_t fn,struct perf_cpu cpu,const char *str)
tools/perf/util/s390-sample-raw.c +10 -9
@@ -12,25 +12,26 @@* sample was taken from. */ -#include <unistd.h> +#include <inttypes.h> #include <stdio.h> #include <string.h> -#include <inttypes.h> -#include <sys/stat.h> +#include <asm/byteorder.h> #include <linux/compiler.h> #include <linux/err.h> -#include <asm/byteorder.h> +#include <sys/stat.h> +#include <unistd.h> +#include "color.h" #include "debug.h" -#include "session.h" #include "evlist.h" -#include "color.h" #include "hashmap.h" -#include "sample-raw.h" +#include "pmu.h" +#include "pmus.h" #include "s390-cpumcf-kernel.h" -#include "util/pmu.h" -#include "util/sample.h" +#include "sample-raw.h" +#include "sample.h" +#include "session.h" static size_t ctrset_size (struct cf_ctrset_entry *set) {
tools/perf/util/stat-shadow.c +12 -8
@@ -2,20 +2,24 @@#include <errno.h> #include <math.h> #include <stdio.h> -#include "evsel.h" -#include "stat.h" + +#include <linux/zalloc.h> + +#include "cgroup.h" #include "color.h" #include "debug.h" -#include "pmu.h" -#include "rblist.h" #include "evlist.h" +#include "evsel.h" #include "expr.h" -#include "metricgroup.h" -#include "cgroup.h" -#include "units.h" +#include "hashmap.h" #include "iostat.h" -#include "util/hashmap.h" +#include "metricgroup.h" +#include "pmu.h" +#include "pmus.h" +#include "rblist.h" +#include "stat.h" #include "tool_pmu.h" +#include "units.h" static bool tool_pmu__is_time_event (const struct perf_stat_config *config,const struct evsel *evsel,int *tool_aggr_idx)
tools/perf/util/stat.c +10 -6
@@ -1,21 +1,25 @@// SPDX-License-Identifier: GPL-2.0 +#include "stat.h" +#include <errno.h> -#include <linux/err.h> #include <inttypes.h> #include <math.h> #include <string.h> + +#include <linux/err.h> +#include <linux/zalloc.h> +#include "counts.h" #include "cpumap.h" #include "debug.h" +#include "evlist.h" +#include "evsel.h" +#include "hashmap.h" #include "header.h" -#include "stat.h" +#include "pmu.h" #include "session.h" #include "target.h" -#include "evlist.h" -#include "evsel.h" #include "thread_map.h" -#include "util/hashmap.h" -#include <linux/zalloc.h> void update_stats (struct stats *stats,u64 val) {
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
Ian Rogers
reviewerPATCH v15 2/19 perf python: Add missed explicit dependencies
In reply to Ian RogersFix missing #include of pmus.h found while cleaning the evsel/evlist header files.
Also define PY_SSIZE_T_CLEAN before including Python.h to comply with modern Python 3 C-API requirements, and introduce CHECK_INITIALIZED and CHECK_INITIALIZED_INT safety macros to be used by subsequent patches.
--- tools/perf/util/python.c | 41 +++++++++++++++++++++++++++++++--------- 1 file changed, 32 insertions(+), 9 deletions(-)
tools/perf/util/python.c +32 -9
@@ -1,30 +1,37 @@// SPDX-License-Identifier: GPL-2.0 +#define PY_SSIZE_T_CLEAN #include <Python.h> -#include <structmember.h> +#include <inttypes.h> -#include <poll.h> +#include <linux/err.h> +#include <poll.h> + +#include <internal/lib.h> #include <perf/cpumap.h> -#ifdef HAVE_LIBTRACEEVENT -#include <event-parse.h> -#endif #include <perf/mmap.h> +#include <structmember.h> +#include "callchain.h" #include "counts.h" +#include "event.h" #include "evlist.h" #include "evsel.h" -#include "event.h" #include "expr.h" +#include "metricgroup.h" +#include "mmap.h" +#include "pmus.h" #include "print_binary.h" #include "record.h" #include "strbuf.h" #include "thread_map.h" #include "tp_pmu.h" #include "trace-event.h" -#include "metricgroup.h" -#include "mmap.h" #include "util/sample.h" -#include <internal/lib.h> + +#ifdef HAVE_LIBTRACEEVENT +#include <event-parse.h> +#endif PyMODINIT_FUNC PyInit_perf (void );
@@ -38,6 +45,22 @@ PyMODINIT_FUNC PyInit_perf(void); offsetof(struct pyrf_event , sample) + offsetof(struct perf_sample , member), \ 0, help } +#define CHECK_INITIALIZED (ptr, msg) \ + do { \ + if (!(ptr)) { \ + PyErr_SetString(PyExc_ValueError, msg " not initialized"); \ + return NULL; \ + } \ + } while (0) + +#define CHECK_INITIALIZED_INT (ptr, msg) \ + do { \ + if (!(ptr)) { \ + PyErr_SetString(PyExc_ValueError, msg " not initialized"); \ + return -1; \ + } \ + } while (0) +struct pyrf_event {PyObject_HEAD struct evsel *evsel;
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
[email protected]
AI reviewRe: [PATCH v15 02/19] perf python: Add missed explicit dependencies
Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
Pre-existing issues:
- [Critical] The pyrf_counts_values_set_values function does not validate the size of the Python list before writing into the fixed-size values array, leading to a heap buffer overflow.
- [High] Functions accepting perf.cpu_map and perf.thread_map Python objects use the generic O format specifier without type validation, leading to Type Confusion and potential memory corruption. --
commit aae79d4d4c46ff069bf5908d4883c3da0d206df7
Author: Ian Rogers <[email protected]>
perf python: Add missed explicit dependencies
Fix missing #include of pmus.h found while cleaning the evsel/evlist header files. Define PY_SSIZE_T_CLEAN before including Python.h to comply with modern Python 3 C-API requirements, and introduce CHECK_INITIALIZED and CHECK_INITIALIZED_INT safety macros.
Quoted 4 lines
tools/perf/util/python.c +0
-0
[ ... ]
Quoted 10 lines
tools/perf/util/python.c +6
-0@@ -38,6 +45,22 @@ PyMODINIT_FUNC PyInit_perf(void); offsetof(struct pyrf_event , sample) + offsetof(struct perf_sample , member), \ 0, help } +#define CHECK_INITIALIZED (ptr, msg) \ + do { \ + if (!(ptr)) { \ + PyErr_SetString(PyExc_ValueError, msg " not initialized"); \ + return NULL; \ + } \
tools/perf/util/python.c PyMODINIT_FUNC PyInit_perf(void); · line 63
Quoted 10 lines
tools/perf/util/python.c +10
-0+ } while (0) + +#define CHECK_INITIALIZED_INT (ptr, msg) \ + do { \ + if (!(ptr)) { \ + PyErr_SetString(PyExc_ValueError, msg " not initialized"); \ + return -1; \ + } \ + } while (0) +
[Severity: Critical]
This is a pre-existing issue, but I noticed a potential buffer overflow in pyrf_counts_values_set_values():
tools/perf/util/python.c:pyrf_counts_values_set_values() {
size = PyList_Size(list);
for (Py_ssize_t i = 0; i < size; i++) {
item = PyList_GetItem(list, i);
// ...
self->values.values[i] = PyLong_AsLong(item);
}
}The values array embedded in struct pyrf_counts_values has exactly 5 elements.
Could a user provide a Python list with more than 5 elements and write past the end of the allocation, corrupting the heap?
[Severity: High]
This isn't a bug introduced by this patch, but the PyArg_ParseTuple type handling seems unsafe in several functions like pyrf_evsel__open():
tools/perf/util/python.c:pyrf_evsel__open() {
if (!PyArg_ParseTupleAndKeywords(args, kwargs, "|OOii", kwlist,
&pcpus, &pthreads, &group, &inherit))
return NULL;
if (pthreads != NULL)
threads = ((struct pyrf_thread_map *)pthreads)->threads;
}Similar casts occur in pyrf__parse_events() and pyrf__parse_metrics() where the 'O' format is used without type validation.
Since 'O' accepts any Python object, what happens if an invalid object type like an integer or string is passed here?
Can this cause the C extension to interpret arbitrary Python objects as struct pyrf_thread_map pointers and dereference them? Could we use the 'O!' format specifier instead to enforce the type constraint?
Quoted 3 lines
tools/perf/util/python.c +0
-0struct pyrf_event {PyObject_HEAD struct evsel *evsel;
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=2
Ian Rogers
reviewerPATCH v15 3/19 perf evsel/evlist: Avoid unnecessary #includes
In reply to Ian RogersUse forward declarations and remove unnecessary #includes in evsel.h. Sort the forward declarations in evsel.h and evlist.h. Move some PMU code into evsel.c.
--- tools/perf/util/evlist.h | 15 +++++++++------ tools/perf/util/evsel.c | 5 +++++ tools/perf/util/evsel.h | 25 ++++++++++++------------- 3 files changed, 26 insertions(+), 19 deletions(-)
tools/perf/util/evlist.h +9 -6
@@ -2,29 +2,32 @@#ifndef __PERF_EVLIST_H#define __PERF_EVLIST_H 1 +#include <signal.h> +#include <linux/compiler.h> #include <linux/kernel.h> -#include <linux/refcount.h> #include <linux/list.h> +#include <linux/refcount.h> +#include <pthread.h> +#include <unistd.h> +#include <api/fd/array.h> #include <internal/evlist.h> #include <internal/evsel.h> #include <perf/evlist.h> +#include "affinity.h" #include "events_stats.h" #include "evsel.h" #include "rblist.h" -#include <pthread.h> -#include <signal.h> -#include <unistd.h> -struct pollfd ; -struct thread_map ;struct perf_cpu_map ;struct perf_stat_config ; +struct pollfd ;struct record_opts ;struct strbuf ;struct target ; +struct thread_map ;/* * State machine of bkw_mmap_state:
tools/perf/util/evsel.c +5 -0
@@ -370,6 +370,11 @@ void evsel__set_sample_id(struct evsel *evsel, evsel->core.attr.read_format |=PERF_FORMAT_ID ; } +bool evsel__is_non_perf_event_open_pmu (const struct evsel *evsel) +{ +return evsel->pmu && evsel->pmu->type >PERF_PMU_TYPE_PE_END ; +} +/** * evsel__is_function_event - Return whether given evsel is a function * trace event
tools/perf/util/evsel.h +12 -13
@@ -2,28 +2,30 @@#ifndef __PERF_EVSEL_H#define __PERF_EVSEL_H 1 -#include <linux/list.h> #include <stdbool.h> -#include <sys/types.h> + +#include <linux/list.h> #include <linux/perf_event.h> #include <linux/types.h> +#include <sys/types.h> +#include <internal/evsel.h> #include <perf/evsel.h> +#include "symbol_conf.h" -#include "pmus.h" -#include "pmu.h" +struct bperf_follower_bpf ; +struct bperf_leader_bpf ; +struct bpf_counter_ops ;struct bpf_object ;struct cgroup ; +struct hashmap ;struct perf_counts ; +struct perf_pmu ;struct perf_stat_config ;struct perf_stat_evsel ; -union perf_event ; -struct bpf_counter_ops ;struct target ; -struct hashmap ; -struct bperf_leader_bpf ; -struct bperf_follower_bpf ; +union perf_event ;typedef int (evsel__sb_cb_t )(union perf_event *event,void *data);
@@ -356,10 +358,6 @@ void arch_evsel__apply_ratio_to_prev(struct evsel *evsel, struct perf_event_attrint evsel__set_filter (struct evsel *evsel,const char *filter);int evsel__append_tp_filter (struct evsel *evsel,const char *filter);int evsel__append_addr_filter (struct evsel *evsel,const char *filter); -static inline bool evsel__is_non_perf_event_open_pmu (const struct evsel *evsel) -{ -return evsel->pmu && evsel->pmu->type >PERF_PMU_TYPE_PE_END ; -}int evsel__enable_cpu (struct evsel *evsel,int cpu_map_idx);int evsel__enable (struct evsel *evsel);
@@ -478,6 +476,7 @@ static inline bool evsel__is_group_event(struct evsel *evsel)return evsel__is_group_leader (evsel) && evsel->core.nr_members >1 ; } +bool evsel__is_non_perf_event_open_pmu (const struct evsel *evsel);bool evsel__is_function_event (struct evsel *evsel);static inline bool evsel__is_bpf_output (struct evsel *evsel)
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
Ian Rogers
reviewerPATCH v15 4/19 perf data: Add open flag
In reply to Ian RogersAvoid double opens and ensure only open files are closed. This addresses some issues with python integration where the data file wants to be opened before being given to a session.
---
Changes in v2:
1. Fixed File Rotation: In perf_data__switch() , I added data->open = false; after the file is closed. This ensures that the subsequent perf_data__open() call will not exit early and will successfully open the new file.
2. Fixed Memory Leak: In open_dir() , I added a call to zfree(&data->file.path) if mkdir() fails, preventing the leak of the path string.
--- tools/perf/util/data.c | 27 ++++++++++++++++++++++----- tools/perf/util/data.h | 4 +++- 2 files changed, 25 insertions(+), 6 deletions(-)
tools/perf/util/data.c +22 -5
@@ -345,9 +345,10 @@ static int open_dir(struct perf_data *data)if (asprintf (&data->file.path,"%s/data" , data->path) <0 )return -1 ; -if (perf_data__is_write (data) && -mkdir (data->path,S_IRWXU ) <0 ) +if (perf_data__is_write (data) &&mkdir (data->path,0700 ) <0 ) { +zfree (&data->file.path);return -1 ; + } ret =open_file (data);
@@ -360,9 +361,16 @@ static int open_dir(struct perf_data *data)int perf_data__open (struct perf_data *data) { -if (check_pipe (data)) +int ret; + +if (data->open)return 0 ; +if (check_pipe (data)) { + data->open = true; +return 0 ; + } +/* currently it allows stdio for pipe only */ data->file.use_stdio = false;
@@ -375,16 +383,24 @@ int perf_data__open(struct perf_data *data)if (perf_data__is_read (data)) data->is_dir =is_dir (data); -return perf_data__is_dir (data) ? -open_dir (data) :open_file_dup (data); + ret =perf_data__is_dir (data) ?open_dir (data) :open_file_dup (data); + +if (!ret) + data->open = true; + +return ret; }void perf_data__close (struct perf_data *data) { +if (!data->open) +return ; +if (perf_data__is_dir (data))perf_data__close_dir (data);perf_data_file__close (&data->file); + data->open = false; }static ssize_t perf_data_file__read (struct perf_data_file *file,void *buf,size_t size)
@@ -457,6 +473,7 @@ int perf_data__switch(struct perf_data *data,if (!at_exit) {perf_data_file__close (&data->file); + data->open = false; ret =perf_data__open (data);if (ret <0 ) goto out;
tools/perf/util/data.h +3 -1
@@ -50,6 +50,8 @@ struct perf_data {const char *path;/** @file: Underlying file to be used. */ struct perf_data_file file; +/** @open: Has the file or directory been opened. */ +bool open;/** @is_pipe: Underlying file is a pipe. */ bool is_pipe;/** @is_dir: Underlying file is a directory. */
@@ -59,7 +61,7 @@ struct perf_data {/** @in_place_update: A file opened for reading but will be written to. */ bool in_place_update;/** @mode: Read or write mode. */ -enum perf_data_mode mode; + enum perf_data_mode mode:8;struct {/** @version: perf_dir_version. */
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
Ian Rogers
reviewerPATCH v15 5/19 perf evlist: Add reference count
In reply to Ian RogersThis a no-op for most of the perf tool. The reference count is set to 1 at allocation, the put will see the 1, decrement it and perform the delete. The purpose for adding the reference count is for the python code. Prior to this change the python code would clone evlists, but this has issues if events are opened, etc. This change adds a reference count for the evlists and a later change will add it to evsels. The combination is needed for the python code to operate correctly (not hit asserts in the evsel clone), but the changes are broken apart for the sake of smaller patches.
---
v2: Added evlist__put to pyrf_evlist__init in case init is called more
than once.
I double-checked trace__replay() and confirmed that trace->evlist
is not assigned to session->evlist in that function.
trace__replay creates a new session and uses its own evlist for
processing file events, leaving trace->evlist pointing to the
empty list created at startup. Therefore, the
evlist__put(trace->evlist) call in trace__exit() is safe and
correct to avoid leaking that empty list.v7:
- Added pyrf_evlist__new to zero-initialize pevlist->evlist to fix crash on re-initialization.
--- tools/perf/arch/x86/tests/hybrid.c | 2 +- tools/perf/arch/x86/tests/topdown.c | 4 +- tools/perf/arch/x86/util/iostat.c | 2 +- tools/perf/bench/evlist-open-close.c | 18 +- tools/perf/builtin-ftrace.c | 8 +- tools/perf/builtin-kvm.c | 4 +- tools/perf/builtin-lock.c | 2 +- tools/perf/builtin-record.c | 4 +- tools/perf/builtin-sched.c | 6 +- tools/perf/builtin-script.c | 2 +- tools/perf/builtin-stat.c | 10 +- tools/perf/builtin-top.c | 52 ++--- tools/perf/builtin-trace.c | 26 +-- tools/perf/tests/backward-ring-buffer.c | 18 +- tools/perf/tests/code-reading.c | 4 +- tools/perf/tests/event-times.c | 4 +- tools/perf/tests/event_update.c | 2 +- tools/perf/tests/evsel-roundtrip-name.c | 8 +- tools/perf/tests/expand-cgroup.c | 8 +- tools/perf/tests/hists_cumulate.c | 2 +- tools/perf/tests/hists_filter.c | 2 +- tools/perf/tests/hists_link.c | 2 +- tools/perf/tests/hists_output.c | 2 +- tools/perf/tests/hwmon_pmu.c | 2 +- tools/perf/tests/keep-tracking.c | 2 +- tools/perf/tests/mmap-basic.c | 18 +- tools/perf/tests/openat-syscall-tp-fields.c | 18 +- tools/perf/tests/parse-events.c | 4 +- tools/perf/tests/parse-metric.c | 4 +- tools/perf/tests/parse-no-sample-id-all.c | 2 +- tools/perf/tests/perf-record.c | 18 +- tools/perf/tests/perf-time-to-tsc.c | 2 +- tools/perf/tests/pfm.c | 4 +- tools/perf/tests/pmu-events.c | 6 +- tools/perf/tests/pmu.c | 4 +- tools/perf/tests/sw-clock.c | 14 +- tools/perf/tests/switch-tracking.c | 2 +- tools/perf/tests/task-exit.c | 14 +- tools/perf/tests/tool_pmu.c | 2 +- tools/perf/tests/topology.c | 2 +- tools/perf/tests/uncore-event-sorting.c | 2 +- tools/perf/util/cgroup.c | 4 +- tools/perf/util/data-convert-bt.c | 2 +- tools/perf/util/evlist.c | 20 +- tools/perf/util/evlist.h | 7 +- tools/perf/util/expr.c | 2 +- tools/perf/util/header.c | 22 +- tools/perf/util/metricgroup.c | 6 +- tools/perf/util/parse-events.c | 4 +- tools/perf/util/perf_api_probe.c | 2 +- tools/perf/util/python.c | 242 +++++++++----------- tools/perf/util/record.c | 2 +- tools/perf/util/session.c | 2 +- tools/perf/util/sideband_evlist.c | 16 +- 54 files changed, 315 insertions(+), 328 deletions(-)
tools/perf/arch/x86/tests/hybrid.c +1 -1
@@ -268,7 +268,7 @@ static int test_event(const struct evlist_test *e) ret = e->check (evlist); }parse_events_error__exit (&err); -evlist__delete (evlist); +evlist__put (evlist);return ret; }
tools/perf/arch/x86/tests/topdown.c +2 -2
@@ -56,7 +56,7 @@ static int event_cb(void *state, struct pmu_event_info *info) *ret =TEST_FAIL ; } } -evlist__delete (evlist); +evlist__put (evlist);return 0 ; }
@@ -174,7 +174,7 @@ static int test_sort(const char *str, int expected_slots_group_size,CHECK_COND (slots_seen,"slots seen" ); ret =TEST_OK ; out_err: -evlist__delete (evlist); +evlist__put (evlist);parse_events_error__exit (&err);return ret; }
tools/perf/arch/x86/util/iostat.c +1 -1
@@ -337,7 +337,7 @@ int iostat_prepare(struct evlist *evlist, struct perf_stat_config *config)if (evlist->core.nr_entries >0 ) {pr_warning ("The -e and -M options are not supported." "All chosen events/metrics will be dropped\n" ); -evlist__delete (evlist); +evlist__put (evlist); evlist =evlist__new ();if (!evlist)return -ENOMEM ;
tools/perf/bench/evlist-open-close.c +9 -9
@@ -76,7 +76,7 @@ static struct evlist *bench__create_evlist(char *evstr, const char *uid_str)parse_events_error__exit (&err);pr_err ("Run 'perf list' for a list of valid events\n" ); ret =1 ; - goto out_delete_evlist; + goto out_put_evlist; }parse_events_error__exit (&err);if (uid_str) {
@@ -85,24 +85,24 @@ static struct evlist *bench__create_evlist(char *evstr, const char *uid_str)if (uid ==UINT_MAX ) {pr_err ("Invalid User: %s" , uid_str); ret = -EINVAL ; - goto out_delete_evlist; + goto out_put_evlist; } ret =parse_uid_filter (evlist, uid);if (ret) - goto out_delete_evlist; + goto out_put_evlist; } ret =evlist__create_maps (evlist, &opts.target);if (ret <0 ) {pr_err ("Not enough memory to create thread/cpu maps\n" ); - goto out_delete_evlist; + goto out_put_evlist; }evlist__config (evlist, &opts,NULL );return evlist; -out_delete_evlist: -evlist__delete (evlist); +out_put_evlist: +evlist__put (evlist);return NULL ; }
@@ -151,7 +151,7 @@ static int bench_evlist_open_close__run(char *evstr, const char *uid_str) evlist->core.nr_entries, evlist__count_evsel_fds(evlist));printf (" Number of iterations:\t%d\n" , iterations); -evlist__delete (evlist); +evlist__put (evlist);for (i =0 ; i < iterations; i++) {pr_debug ("Started iteration %d\n" , i);
@@ -162,7 +162,7 @@ static int bench_evlist_open_close__run(char *evstr, const char *uid_str)gettimeofday (&start,NULL ); err =bench__do_evlist_open_close (evlist);if (err) { -evlist__delete (evlist); +evlist__put (evlist);return err; }
@@ -171,7 +171,7 @@ static int bench_evlist_open_close__run(char *evstr, const char *uid_str) runtime_us =timeval2usec (&diff);update_stats (&time_stats, runtime_us); -evlist__delete (evlist); +evlist__put (evlist);pr_debug ("Iteration %d took:\t%" PRIu64"us\n" , i, runtime_us); }
tools/perf/builtin-ftrace.c +4 -4
@@ -1999,20 +1999,20 @@ int cmd_ftrace(int argc, const char **argv) ret =evlist__create_maps (ftrace.evlist, &ftrace.target);if (ret <0 ) - goto out_delete_evlist; + goto out_put_evlist;if (argc) { ret =evlist__prepare_workload (ftrace.evlist, &ftrace.target, argv, false, ftrace__workload_exec_failed_signal);if (ret <0 ) - goto out_delete_evlist; + goto out_put_evlist; } ret =cmd_func (&ftrace); -out_delete_evlist: -evlist__delete (ftrace.evlist); +out_put_evlist: +evlist__put (ftrace.evlist); out_delete_filters:delete_filter_func (&ftrace.filters);
tools/perf/builtin-kvm.c +2 -2
@@ -1810,7 +1810,7 @@ static struct evlist *kvm_live_event_list(void) out:if (err) { -evlist__delete (evlist); +evlist__put (evlist); evlist =NULL ; }
@@ -1941,7 +1941,7 @@ static int kvm_events_live(struct perf_kvm_stat *kvm, out:perf_session__delete (kvm->session); kvm->session =NULL ; -evlist__delete (kvm->evlist); +evlist__put (kvm->evlist);return err; }
tools/perf/builtin-lock.c +1 -1
@@ -2149,7 +2149,7 @@ static int __cmd_contention(int argc, const char **argv) out_delete:lock_filter_finish (); -evlist__delete (con.evlist); +evlist__put (con.evlist);lock_contention_finish (&con);perf_session__delete (session);perf_env__exit (&host_env);
tools/perf/builtin-record.c +2 -2
@@ -4291,7 +4291,7 @@ int cmd_record(int argc, const char **argv) goto out;evlist__splice_list_tail (rec->evlist, &def_evlist->core.entries); -evlist__delete (def_evlist); +evlist__put (def_evlist); }if (rec->opts.target.tid && !rec->opts.no_inherit_set)
@@ -4401,7 +4401,7 @@ int cmd_record(int argc, const char **argv)auxtrace_record__free (rec->itr); out_opts:evlist__close_control (rec->opts.ctl_fd, rec->opts.ctl_fd_ack, &rec->opts.ctl_fd_close); -evlist__delete (rec->evlist); +evlist__put (rec->evlist);return err; }
tools/perf/builtin-sched.c +3 -3
@@ -3913,7 +3913,7 @@ static int perf_sched__schedstat_record(struct perf_sched *sched, session =perf_session__new (&data, &sched->tool);if (IS_ERR (session)) {pr_err ("Perf session creation failed.\n" ); -evlist__delete (evlist); +evlist__put (evlist);return PTR_ERR (session); }
@@ -4012,7 +4012,7 @@ static int perf_sched__schedstat_record(struct perf_sched *sched,else fprintf (stderr,"[ perf sched stats: Failed !! ]\n" ); -evlist__delete (evlist); +evlist__put (evlist);close (fd);return err; }
@@ -4916,7 +4916,7 @@ static int perf_sched__schedstat_live(struct perf_sched *sched,free_cpu_domain_info (cd_map, sv, nr); out:free_schedstat (&cpu_head); -evlist__delete (evlist); +evlist__put (evlist);return err; }
tools/perf/builtin-script.c +1 -1
@@ -2274,7 +2274,7 @@ static int script_find_metrics(const struct pmu_metric *pm, }pr_debug ("Found metric '%s' whose evsels match those of in the perf data\n" , pm->metric_name); -evlist__delete (metric_evlist); +evlist__put (metric_evlist); out:return 0 ; }
tools/perf/builtin-stat.c +5 -5
@@ -2119,7 +2119,7 @@ static int add_default_events(void) stat_config.user_requested_cpu_list, stat_config.system_wide, stat_config.hardware_aware_grouping) < 0) { -evlist__delete (metric_evlist); +evlist__put (metric_evlist); ret =-1 ;break ; }
@@ -2131,7 +2131,7 @@ static int add_default_events(void)metricgroup__copy_metric_events (evlist,/*cgrp=*/ NULL , &evlist->metric_events, &metric_evlist->metric_events); -evlist__delete (metric_evlist); +evlist__put (metric_evlist); }list_sort (/*priv=*/ NULL , &evlist->core.entries, default_evlist_evsel_cmp);
@@ -2152,7 +2152,7 @@ static int add_default_events(void)metricgroup__copy_metric_events (evsel_list,/*cgrp=*/ NULL , &evsel_list->metric_events, &evlist->metric_events); -evlist__delete (evlist); +evlist__put (evlist);return ret; }
@@ -2387,7 +2387,7 @@ static int __cmd_report(int argc, const char **argv) perf_stat.session = session; stat_config.output = stderr; -evlist__delete (evsel_list); +evlist__put (evsel_list); evsel_list = session->evlist; ret =perf_session__process_events (session);
@@ -3066,7 +3066,7 @@ int cmd_stat(int argc, const char **argv)if (smi_cost && smi_reset)sysfs__write_int (FREEZE_ON_SMI_PATH ,0 ); -evlist__delete (evsel_list); +evlist__put (evsel_list);evlist__close_control (stat_config.ctl_fd, stat_config.ctl_fd_ack, &stat_config.ctl_fd_close);
tools/perf/builtin-top.c +26 -26
184 lines not on this page · raw message
tools/perf/builtin-trace.c +13 -13
@@ -4418,7 +4418,7 @@ static int trace__run(struct trace *trace, int argc, const char **argv)if (trace->summary_bpf) {if (trace_prepare_bpf_summary (trace->summary_mode) <0 ) - goto out_delete_evlist; + goto out_put_evlist;if (trace->summary_only) goto create_maps;
@@ -4486,19 +4486,19 @@ static int trace__run(struct trace *trace, int argc, const char **argv) err =evlist__create_maps (evlist, &trace->opts.target);if (err <0 ) {fprintf (trace->output,"Problems parsing the target to trace, check your options!\n" ); - goto out_delete_evlist; + goto out_put_evlist; } err =trace__symbols_init (trace, argc, argv, evlist);if (err <0 ) {fprintf (trace->output,"Problems initializing symbol libraries!\n" ); - goto out_delete_evlist; + goto out_put_evlist; }if (trace->summary_mode ==SUMMARY__BY_TOTAL && !trace->summary_bpf) { trace->syscall_stats =alloc_syscall_stats ();if (!trace->syscall_stats) - goto out_delete_evlist; + goto out_put_evlist; }evlist__config (evlist, &trace->opts, &callchain_param);
@@ -4507,7 +4507,7 @@ static int trace__run(struct trace *trace, int argc, const char **argv) err =evlist__prepare_workload (evlist, &trace->opts.target, argv, false,NULL );if (err <0 ) {fprintf (trace->output,"Couldn't run the workload!\n" ); - goto out_delete_evlist; + goto out_put_evlist; } workload_pid = evlist->workload.pid; }
@@ -4555,7 +4555,7 @@ static int trace__run(struct trace *trace, int argc, const char **argv) err =trace__expand_filters (trace, &evsel);if (err) - goto out_delete_evlist; + goto out_put_evlist; err =evlist__apply_filters (evlist, &evsel, &trace->opts.target);if (err <0 ) goto out_error_apply_filters;
@@ -4672,12 +4672,12 @@ static int trace__run(struct trace *trace, int argc, const char **argv) } } -out_delete_evlist: +out_put_evlist:trace_cleanup_bpf_summary ();delete_syscall_stats (trace->syscall_stats);trace__symbols__exit (trace);evlist__free_syscall_tp_fields (evlist); -evlist__delete (evlist); +evlist__put (evlist);cgroup__put (trace->cgroup); trace->evlist =NULL ; trace->live = false;
@@ -4702,21 +4702,21 @@ static int trace__run(struct trace *trace, int argc, const char **argv) out_error:fprintf (trace->output,"%s\n" , errbuf); - goto out_delete_evlist; + goto out_put_evlist; out_error_apply_filters:fprintf (trace->output,"Failed to set filter \"%s\" on event %s: %m\n" , evsel->filter,evsel__name (evsel)); - goto out_delete_evlist; + goto out_put_evlist; } out_error_mem:fprintf (trace->output,"Not enough memory to run!\n" ); - goto out_delete_evlist; + goto out_put_evlist; out_errno:fprintf (trace->output,"%m\n" ); - goto out_delete_evlist; + goto out_put_evlist; }static int trace__replay (struct trace *trace)
@@ -5396,7 +5396,7 @@ static void trace__exit(struct trace *trace)zfree (&trace->syscalls.table); }zfree (&trace->perfconfig_events); -evlist__delete (trace->evlist); +evlist__put (trace->evlist); trace->evlist =NULL ;ordered_events__free (&trace->oe.data);#ifdef HAVE_LIBBPF_SUPPORT
tools/perf/tests/backward-ring-buffer.c +9 -9
@@ -111,7 +111,7 @@ static int test__backward_ring_buffer(struct test_suite *test __maybe_unused, in err =evlist__create_maps (evlist, &opts.target);if (err <0 ) {pr_debug ("Not enough memory to create thread/cpu maps\n" ); - goto out_delete_evlist; + goto out_put_evlist; }parse_events_error__init (&parse_error);
@@ -124,7 +124,7 @@ static int test__backward_ring_buffer(struct test_suite *test __maybe_unused, inif (err) {pr_debug ("Failed to parse tracepoint event, try use root\n" ); ret =TEST_SKIP ; - goto out_delete_evlist; + goto out_put_evlist; }evlist__config (evlist, &opts,NULL );
@@ -133,19 +133,19 @@ static int test__backward_ring_buffer(struct test_suite *test __maybe_unused, inif (err <0 ) {pr_debug ("perf_evlist__open: %s\n" ,str_error_r (errno, sbuf,sizeof (sbuf))); - goto out_delete_evlist; + goto out_put_evlist; } ret =TEST_FAIL ; err =do_test (evlist, opts.mmap_pages, &sample_count, &comm_count);if (err !=TEST_OK ) - goto out_delete_evlist; + goto out_put_evlist;if ((sample_count !=NR_ITERS ) || (comm_count !=NR_ITERS )) {pr_err ("Unexpected counter: sample_count=%d, comm_count=%d\n" , sample_count, comm_count); - goto out_delete_evlist; + goto out_put_evlist; }evlist__close (evlist);
@@ -154,16 +154,16 @@ static int test__backward_ring_buffer(struct test_suite *test __maybe_unused, inif (err <0 ) {pr_debug ("perf_evlist__open: %s\n" ,str_error_r (errno, sbuf,sizeof (sbuf))); - goto out_delete_evlist; + goto out_put_evlist; } err =do_test (evlist,1 , &sample_count, &comm_count);if (err !=TEST_OK ) - goto out_delete_evlist; + goto out_put_evlist; ret =TEST_OK ; -out_delete_evlist: -evlist__delete (evlist); +out_put_evlist: +evlist__put (evlist);return ret; }
tools/perf/tests/code-reading.c +2 -2
@@ -810,7 +810,7 @@ static int do_test_code_reading(bool try_kcore) }perf_evlist__set_maps (&evlist->core,NULL ,NULL ); -evlist__delete (evlist); +evlist__put (evlist); evlist =NULL ;continue ; }
@@ -847,7 +847,7 @@ static int do_test_code_reading(bool try_kcore) out_put:thread__put (thread); out_err: -evlist__delete (evlist); +evlist__put (evlist);perf_cpu_map__put (cpus);perf_thread_map__put (threads);machine__delete (machine);
tools/perf/tests/event-times.c +2 -2
@@ -186,7 +186,7 @@ static int test_times(int (attach)(struct evlist *), err =attach (evlist);if (err ==TEST_SKIP ) {pr_debug (" SKIP : not enough rights\n" ); -evlist__delete (evlist); +evlist__put (evlist);return err; }
@@ -205,7 +205,7 @@ static int test_times(int (attach)(struct evlist *), count.ena, count.run); out_err: -evlist__delete (evlist); +evlist__put (evlist);return !err ?TEST_OK :TEST_FAIL ; }
tools/perf/tests/event_update.c +1 -1
@@ -117,7 +117,7 @@ static int test__event_update(struct test_suite *test __maybe_unused, int subtesTEST_ASSERT_VAL ("failed to synthesize attr update cpus" , !perf_event__synthesize_event_update_cpus (&tmp.tool, evsel, process_event_cpus)); -evlist__delete (evlist); +evlist__put (evlist);return 0 ; }
tools/perf/tests/evsel-roundtrip-name.c +4 -4
@@ -33,7 +33,7 @@ static int perf_evsel__roundtrip_cache_name_test(void)if (err) {pr_debug ("Failure to parse cache event '%s' possibly as PMUs don't support it" , name); -evlist__delete (evlist); +evlist__put (evlist);continue ; }evlist__for_each_entry (evlist, evsel) {
@@ -42,7 +42,7 @@ static int perf_evsel__roundtrip_cache_name_test(void) ret =TEST_FAIL ; } } -evlist__delete (evlist); +evlist__put (evlist); } } }
@@ -66,7 +66,7 @@ static int perf_evsel__name_array_test(const char *const names[], int nr_names)if (err) {pr_debug ("failed to parse event '%s', err %d\n" , names[i], err); -evlist__delete (evlist); +evlist__put (evlist); ret =TEST_FAIL ;continue ; }
@@ -76,7 +76,7 @@ static int perf_evsel__name_array_test(const char *const names[], int nr_names) ret =TEST_FAIL ; } } -evlist__delete (evlist); +evlist__put (evlist); }return ret; }
tools/perf/tests/expand-cgroup.c +4 -4
@@ -106,7 +106,7 @@ static int expand_default_events(void)TEST_ASSERT_VAL ("failed to get evlist" , evlist); ret =test_expand_events (evlist); -evlist__delete (evlist); +evlist__put (evlist);return ret; }
@@ -133,7 +133,7 @@ static int expand_group_events(void) ret =test_expand_events (evlist); out:parse_events_error__exit (&err); -evlist__delete (evlist); +evlist__put (evlist);return ret; }
@@ -164,7 +164,7 @@ static int expand_libpfm_events(void) ret =test_expand_events (evlist); out: -evlist__delete (evlist); +evlist__put (evlist);return ret; }
@@ -188,7 +188,7 @@ static int expand_metric_events(void) ret =test_expand_events (evlist); out: -evlist__delete (evlist); +evlist__put (evlist);return ret; }
tools/perf/tests/hists_cumulate.c +1 -1
@@ -743,7 +743,7 @@ static int test__hists_cumulate(struct test_suite *test __maybe_unused, int subt out:/* tear down everything */ -evlist__delete (evlist); +evlist__put (evlist);machines__exit (&machines);put_fake_samples ();
tools/perf/tests/hists_filter.c +1 -1
@@ -331,7 +331,7 @@ static int test__hists_filter(struct test_suite *test __maybe_unused, int subtes out:/* tear down everything */ -evlist__delete (evlist); +evlist__put (evlist);reset_output_field ();machines__exit (&machines);put_fake_samples ();
tools/perf/tests/hists_link.c +1 -1
@@ -352,7 +352,7 @@ static int test__hists_link(struct test_suite *test __maybe_unused, int subtest out:/* tear down everything */ -evlist__delete (evlist); +evlist__put (evlist);reset_output_field ();machines__exit (&machines);put_fake_samples ();
tools/perf/tests/hists_output.c +1 -1
@@ -630,7 +630,7 @@ static int test__hists_output(struct test_suite *test __maybe_unused, int subtes out:/* tear down everything */ -evlist__delete (evlist); +evlist__put (evlist);machines__exit (&machines);put_fake_samples ();
tools/perf/tests/hwmon_pmu.c +1 -1
@@ -215,7 +215,7 @@ static int do_test(size_t i, bool with_pmu, bool with_alias) out:parse_events_error__exit (&err); -evlist__delete (evlist); +evlist__put (evlist);return ret; }
tools/perf/tests/keep-tracking.c +1 -1
@@ -153,7 +153,7 @@ static int test__keep_tracking(struct test_suite *test __maybe_unused, int subte out_err:if (evlist) {evlist__disable (evlist); -evlist__delete (evlist); +evlist__put (evlist); }perf_cpu_map__put (cpus);perf_thread_map__put (threads);
tools/perf/tests/mmap-basic.c +9 -9
@@ -94,7 +94,7 @@ static int test__basic_mmap(struct test_suite *test __maybe_unused, int subtest/* Permissions failure, flag the failure as a skip. */ err =TEST_SKIP ; } - goto out_delete_evlist; + goto out_put_evlist; } evsels[i]->core.attr.wakeup_events =1 ;
@@ -106,7 +106,7 @@ static int test__basic_mmap(struct test_suite *test __maybe_unused, int subtestpr_debug ("failed to open counter: %s, " "tweak /proc/sys/kernel/perf_event_paranoid?\n" ,str_error_r (errno, sbuf,sizeof (sbuf))); - goto out_delete_evlist; + goto out_put_evlist; } nr_events[i] =0 ;
@@ -116,7 +116,7 @@ static int test__basic_mmap(struct test_suite *test __maybe_unused, int subtestif (evlist__mmap (evlist,128 ) <0 ) {pr_debug ("failed to mmap events: %d (%s)\n" , errno,str_error_r (errno, sbuf,sizeof (sbuf))); - goto out_delete_evlist; + goto out_put_evlist; }for (i =0 ; i < nsyscalls; ++i)
@@ -134,7 +134,7 @@ static int test__basic_mmap(struct test_suite *test __maybe_unused, int subtestif (event->header.type !=PERF_RECORD_SAMPLE ) {pr_debug ("unexpected %s event\n" ,perf_event__name (event->header.type)); - goto out_delete_evlist; + goto out_put_evlist; }perf_sample__init (&sample,/*all=*/ false);
@@ -142,7 +142,7 @@ static int test__basic_mmap(struct test_suite *test __maybe_unused, int subtestif (err) {pr_err ("Can't parse sample, err = %d\n" , err);perf_sample__exit (&sample); - goto out_delete_evlist; + goto out_put_evlist; } err =-1 ;
@@ -153,7 +153,7 @@ static int test__basic_mmap(struct test_suite *test __maybe_unused, int subtestif (evsel ==NULL ) {pr_debug ("event with id %" PRIu64" doesn't map to an evsel\n" , sample.id); - goto out_delete_evlist; + goto out_put_evlist; } nr_events[evsel->core.idx]++;perf_mmap__consume (&md->core);
@@ -168,12 +168,12 @@ static int test__basic_mmap(struct test_suite *test __maybe_unused, int subtest expected_nr_events[evsel->core.idx], evsel__name(evsel), nr_events[evsel->core.idx]); err =-1 ; - goto out_delete_evlist; + goto out_put_evlist; } } -out_delete_evlist: -evlist__delete (evlist); +out_put_evlist: +evlist__put (evlist); out_free_cpus:perf_cpu_map__put (cpus); out_free_threads:
tools/perf/tests/openat-syscall-tp-fields.c +9 -9
@@ -51,7 +51,7 @@ static int test__syscall_openat_tp_fields(struct test_suite *test __maybe_unusedif (IS_ERR (evsel)) {pr_debug ("%s: evsel__newtp\n" , __func__); ret =PTR_ERR (evsel) == -EACCES ?TEST_SKIP :TEST_FAIL ; - goto out_delete_evlist; + goto out_put_evlist; }evlist__add (evlist, evsel);
@@ -59,7 +59,7 @@ static int test__syscall_openat_tp_fields(struct test_suite *test __maybe_unused err =evlist__create_maps (evlist, &opts.target);if (err <0 ) {pr_debug ("%s: evlist__create_maps\n" , __func__); - goto out_delete_evlist; + goto out_put_evlist; }evsel__config (evsel, &opts,NULL );
@@ -70,14 +70,14 @@ static int test__syscall_openat_tp_fields(struct test_suite *test __maybe_unusedif (err <0 ) {pr_debug ("perf_evlist__open: %s\n" ,str_error_r (errno, sbuf,sizeof (sbuf))); - goto out_delete_evlist; + goto out_put_evlist; } err =evlist__mmap (evlist,UINT_MAX );if (err <0 ) {pr_debug ("evlist__mmap: %s\n" ,str_error_r (errno, sbuf,sizeof (sbuf))); - goto out_delete_evlist; + goto out_put_evlist; }evlist__enable (evlist);
@@ -115,7 +115,7 @@ static int test__syscall_openat_tp_fields(struct test_suite *test __maybe_unusedif (err) {pr_debug ("Can't parse sample, err = %d\n" , err);perf_sample__exit (&sample); - goto out_delete_evlist; + goto out_put_evlist; } tp_flags =perf_sample__intval (&sample,"flags" );
@@ -126,7 +126,7 @@ static int test__syscall_openat_tp_fields(struct test_suite *test __maybe_unused (tp_flags & flags) != flags) {pr_debug ("%s: Expected flags=%#x, got %#x\n" , __func__, flags, tp_flags); - goto out_delete_evlist; + goto out_put_evlist; } goto out_ok;
@@ -139,13 +139,13 @@ static int test__syscall_openat_tp_fields(struct test_suite *test __maybe_unusedif (++nr_polls >5 ) {pr_debug ("%s: no events!\n" , __func__); - goto out_delete_evlist; + goto out_put_evlist; } } out_ok: ret =TEST_OK ; -out_delete_evlist: -evlist__delete (evlist); +out_put_evlist: +evlist__put (evlist); out:return ret; }
tools/perf/tests/parse-events.c +2 -2
@@ -2568,7 +2568,7 @@ static int test_event(const struct evlist_test *e) ret = e->check (evlist); }parse_events_error__exit (&err); -evlist__delete (evlist); +evlist__put (evlist);return ret; }
@@ -2594,7 +2594,7 @@ static int test_event_fake_pmu(const char *str) }parse_events_error__exit (&err); -evlist__delete (evlist); +evlist__put (evlist);return ret; }
tools/perf/tests/parse-metric.c +2 -2
@@ -84,7 +84,7 @@ static int __compute_metric(const char *name, struct value *vals, cpus =perf_cpu_map__new ("0" );if (!cpus) { -evlist__delete (evlist); +evlist__put (evlist);return -ENOMEM ; }
@@ -113,7 +113,7 @@ static int __compute_metric(const char *name, struct value *vals,/* ... cleanup. */ evlist__free_stats (evlist);perf_cpu_map__put (cpus); -evlist__delete (evlist); +evlist__put (evlist);return err; }
tools/perf/tests/parse-no-sample-id-all.c +1 -1
@@ -49,7 +49,7 @@ static int process_events(union perf_event **events, size_t count)for (i =0 ; i < count && !err; i++) err =process_event (&evlist, events[i]); -evlist__delete (evlist); +evlist__put (evlist);return err; }
tools/perf/tests/perf-record.c +9 -9
@@ -105,7 +105,7 @@ static int test__PERF_RECORD(struct test_suite *test __maybe_unused, int subtest err =evlist__create_maps (evlist, &opts.target);if (err <0 ) {pr_debug ("Not enough memory to create thread/cpu maps\n" ); - goto out_delete_evlist; + goto out_put_evlist; }/*
@@ -117,7 +117,7 @@ static int test__PERF_RECORD(struct test_suite *test __maybe_unused, int subtest err =evlist__prepare_workload (evlist, &opts.target, argv, false,NULL );if (err <0 ) {pr_debug ("Couldn't run the workload!\n" ); - goto out_delete_evlist; + goto out_put_evlist; }/*
@@ -134,7 +134,7 @@ static int test__PERF_RECORD(struct test_suite *test __maybe_unused, int subtestpr_debug ("sched__get_first_possible_cpu: %s\n" ,str_error_r (errno, sbuf,sizeof (sbuf)));evlist__cancel_workload (evlist); - goto out_delete_evlist; + goto out_put_evlist; } cpu = err;
@@ -146,7 +146,7 @@ static int test__PERF_RECORD(struct test_suite *test __maybe_unused, int subtestpr_debug ("sched_setaffinity: %s\n" ,str_error_r (errno, sbuf,sizeof (sbuf)));evlist__cancel_workload (evlist); - goto out_delete_evlist; + goto out_put_evlist; }/*
@@ -158,7 +158,7 @@ static int test__PERF_RECORD(struct test_suite *test __maybe_unused, int subtestpr_debug ("perf_evlist__open: %s\n" ,str_error_r (errno, sbuf,sizeof (sbuf)));evlist__cancel_workload (evlist); - goto out_delete_evlist; + goto out_put_evlist; }/*
@@ -171,7 +171,7 @@ static int test__PERF_RECORD(struct test_suite *test __maybe_unused, int subtestpr_debug ("evlist__mmap: %s\n" ,str_error_r (errno, sbuf,sizeof (sbuf)));evlist__cancel_workload (evlist); - goto out_delete_evlist; + goto out_put_evlist; }/*
@@ -209,7 +209,7 @@ static int test__PERF_RECORD(struct test_suite *test __maybe_unused, int subtestif (verbose >0 )perf_event__fprintf (event,NULL , stderr);pr_debug ("Couldn't parse sample\n" ); - goto out_delete_evlist; + goto out_put_evlist; }if (verbose >0 ) {
@@ -350,9 +350,9 @@ static int test__PERF_RECORD(struct test_suite *test __maybe_unused, int subtestpr_debug ("PERF_RECORD_MMAP for %s missing!\n" ,"[vdso]" ); ++errs; } -out_delete_evlist: +out_put_evlist:CPU_FREE (cpu_mask); -evlist__delete (evlist); +evlist__put (evlist); out:perf_sample__exit (&sample);if (err == -EACCES )
tools/perf/tests/perf-time-to-tsc.c +1 -1
@@ -201,7 +201,7 @@ static int test__perf_time_to_tsc(struct test_suite *test __maybe_unused, int su err =TEST_OK ; out_err: -evlist__delete (evlist); +evlist__put (evlist);perf_cpu_map__put (cpus);perf_thread_map__put (threads);return err;
tools/perf/tests/pfm.c +2 -2
@@ -80,7 +80,7 @@ static int test__pfm_events(struct test_suite *test __maybe_unused,evlist__nr_groups (evlist), 0); -evlist__delete (evlist); +evlist__put (evlist); }return 0 ; }
@@ -165,7 +165,7 @@ static int test__pfm_group(struct test_suite *test __maybe_unused,evlist__nr_groups (evlist), table[i].nr_groups); -evlist__delete (evlist); +evlist__put (evlist); }return 0 ; }
tools/perf/tests/pmu-events.c +3 -3
@@ -798,7 +798,7 @@ static int check_parse_id(const char *id, struct parse_events_error *error) /*warn_if_reordered=*/true, /*fake_tp=*/false);free (dup); -evlist__delete (evlist); +evlist__put (evlist);return ret; }
@@ -865,7 +865,7 @@ static int test__parsing_callback(const struct pmu_metric *pm, cpus =perf_cpu_map__new ("0" );if (!cpus) { -evlist__delete (evlist); +evlist__put (evlist);return -ENOMEM ; }
@@ -919,7 +919,7 @@ static int test__parsing_callback(const struct pmu_metric *pm,/* ... cleanup. */ evlist__free_stats (evlist);perf_cpu_map__put (cpus); -evlist__delete (evlist); +evlist__put (evlist);return err; }
tools/perf/tests/pmu.c +2 -2
@@ -294,7 +294,7 @@ static int test__pmu_config_helpers(struct test_suite *test __maybe_unused, ret =TEST_OK ; err_out:parse_events_terms__exit (&terms); -evlist__delete (evlist); +evlist__put (evlist);test_pmu_put (dir, pmu);return ret; }
@@ -346,7 +346,7 @@ static int test__pmu_events(struct test_suite *test __maybe_unused, int subtest ret =TEST_OK ; err_out:parse_events_error__exit (&err); -evlist__delete (evlist); +evlist__put (evlist);test_pmu_put (dir, pmu);return ret; }
tools/perf/tests/sw-clock.c +7 -7
@@ -59,7 +59,7 @@ static int __test__sw_clock_freq(enum perf_sw_ids clock_id) evsel =evsel__new (&attr);if (evsel ==NULL ) {pr_debug ("evsel__new\n" ); - goto out_delete_evlist; + goto out_put_evlist; }evlist__add (evlist, evsel);
@@ -68,7 +68,7 @@ static int __test__sw_clock_freq(enum perf_sw_ids clock_id)if (!cpus || !threads) { err = -ENOMEM ;pr_debug ("Not enough memory to create thread/cpu maps\n" ); - goto out_delete_evlist; + goto out_put_evlist; }perf_evlist__set_maps (&evlist->core, cpus, threads);
@@ -80,14 +80,14 @@ static int __test__sw_clock_freq(enum perf_sw_ids clock_id)pr_debug ("Couldn't open evlist: %s\nHint: check %s, using %" PRIu64" in this test.\n" ,str_error_r (errno, sbuf,sizeof (sbuf)), knob, (u64 )attr.sample_freq); - goto out_delete_evlist; + goto out_put_evlist; } err =evlist__mmap (evlist,128 );if (err <0 ) {pr_debug ("failed to mmap event: %d (%s)\n" , errno,str_error_r (errno, sbuf,sizeof (sbuf))); - goto out_delete_evlist; + goto out_put_evlist; }evlist__enable (evlist);
@@ -113,7 +113,7 @@ static int __test__sw_clock_freq(enum perf_sw_ids clock_id)if (err <0 ) {pr_debug ("Error during parse sample\n" );perf_sample__exit (&sample); - goto out_delete_evlist; + goto out_put_evlist; } total_periods += sample.period;
@@ -131,10 +131,10 @@ static int __test__sw_clock_freq(enum perf_sw_ids clock_id) err =-1 ; } -out_delete_evlist: +out_put_evlist:perf_cpu_map__put (cpus);perf_thread_map__put (threads); -evlist__delete (evlist); +evlist__put (evlist);return err; }
tools/perf/tests/switch-tracking.c +1 -1
@@ -582,7 +582,7 @@ static int test__switch_tracking(struct test_suite *test __maybe_unused, int sub out:if (evlist) {evlist__disable (evlist); -evlist__delete (evlist); +evlist__put (evlist); }perf_cpu_map__put (cpus);perf_thread_map__put (threads);
tools/perf/tests/task-exit.c +7 -7
@@ -74,7 +74,7 @@ static int test__task_exit(struct test_suite *test __maybe_unused, int subtest _if (!cpus || !threads) { err = -ENOMEM ;pr_debug ("Not enough memory to create thread/cpu maps\n" ); - goto out_delete_evlist; + goto out_put_evlist; }perf_evlist__set_maps (&evlist->core, cpus, threads);
@@ -82,7 +82,7 @@ static int test__task_exit(struct test_suite *test __maybe_unused, int subtest _ err =evlist__prepare_workload (evlist, &target, argv, false, workload_exec_failed_signal);if (err <0 ) {pr_debug ("Couldn't run the workload!\n" ); - goto out_delete_evlist; + goto out_put_evlist; } evsel =evlist__first (evlist);
@@ -101,14 +101,14 @@ static int test__task_exit(struct test_suite *test __maybe_unused, int subtest _if (err <0 ) {pr_debug ("Couldn't open the evlist: %s\n" ,str_error_r (-err, sbuf,sizeof (sbuf))); - goto out_delete_evlist; + goto out_put_evlist; }if (evlist__mmap (evlist,128 ) <0 ) {pr_debug ("failed to mmap events: %d (%s)\n" , errno,str_error_r (errno, sbuf,sizeof (sbuf))); err =-1 ; - goto out_delete_evlist; + goto out_put_evlist; }evlist__start_workload (evlist);
@@ -133,7 +133,7 @@ static int test__task_exit(struct test_suite *test __maybe_unused, int subtest _if (retry_count++ >1000 ) {pr_debug ("Failed after retrying 1000 times\n" ); err =-1 ; - goto out_delete_evlist; + goto out_put_evlist; } goto retry;
@@ -144,10 +144,10 @@ static int test__task_exit(struct test_suite *test __maybe_unused, int subtest _ err =-1 ; } -out_delete_evlist: +out_put_evlist:perf_cpu_map__put (cpus);perf_thread_map__put (threads); -evlist__delete (evlist); +evlist__put (evlist);return err; }
tools/perf/tests/tool_pmu.c +1 -1
@@ -67,7 +67,7 @@ static int do_test(enum tool_pmu_event ev, bool with_pmu) out:parse_events_error__exit (&err); -evlist__delete (evlist); +evlist__put (evlist);return ret; }
tools/perf/tests/topology.c +1 -1
@@ -58,7 +58,7 @@ static int session_write_header(char *path) !perf_session__write_header (session, session->evlist, perf_data__fd(&data), true)); -evlist__delete (session->evlist); +evlist__put (session->evlist);perf_session__delete (session);return 0 ;
tools/perf/tests/uncore-event-sorting.c +1 -1
@@ -165,7 +165,7 @@ static int test__uncore_event_sorting(struct test_suite *test __maybe_unused, ret =TEST_OK ; out_err: -evlist__delete (evlist); +evlist__put (evlist);parse_events_error__exit (&err);zfree (&pmu_prefix);zfree (&m.event1);
tools/perf/util/cgroup.c +2 -2
@@ -520,8 +520,8 @@ int evlist__expand_cgroup(struct evlist *evlist, const char *str, bool open_cgro cgrp_event_expanded = true; out_err: -evlist__delete (orig_list); -evlist__delete (tmp_list); +evlist__put (orig_list); +evlist__put (tmp_list);metricgroup__rblist_exit (&orig_metric_events);release_cgroup_list ();
tools/perf/util/data-convert-bt.c +1 -1
@@ -1362,7 +1362,7 @@ static void cleanup_events(struct perf_session *session)zfree (&evsel->priv); } -evlist__delete (evlist); +evlist__put (evlist); session->evlist =NULL ; }
tools/perf/util/evlist.c +15 -5
@@ -75,7 +75,7 @@ int sigqueue(pid_t pid, int sig, const union sigval value);#define FD (e, x, y) (*(int *)xyarray__entry(e->core.fd, x, y))#define SID (e, x, y) xyarray__entry(e->core.sample_id, x, y) -void evlist__init (struct evlist *evlist,struct perf_cpu_map *cpus, +static void evlist__init (struct evlist *evlist,struct perf_cpu_map *cpus,struct perf_thread_map *threads) {perf_evlist__init (&evlist->core);
@@ -88,6 +88,7 @@ void evlist__init(struct evlist *evlist, struct perf_cpu_map *cpus, evlist->nr_br_cntr =-1 ;metricgroup__rblist_init (&evlist->metric_events);INIT_LIST_HEAD (&evlist->deferred_samples); +refcount_set (&evlist->refcnt,1 ); }struct evlist *evlist__new (void )
@@ -139,7 +140,7 @@ struct evlist *evlist__new_default(const struct target *target, bool sample_callreturn evlist; out_err: -evlist__delete (evlist); +evlist__put (evlist);return NULL ; }
@@ -148,13 +149,19 @@ struct evlist *evlist__new_dummy(void)struct evlist *evlist =evlist__new ();if (evlist &&evlist__add_dummy (evlist)) { -evlist__delete (evlist); +evlist__put (evlist); evlist =NULL ; }return evlist; } +struct evlist *evlist__get (struct evlist *evlist) +{ +refcount_inc (&evlist->refcnt); +return evlist; +} +/** * evlist__set_id_pos - set the positions of event ids. * @evlist: selected event list
@@ -193,7 +200,7 @@ static void evlist__purge(struct evlist *evlist) evlist->core.nr_entries =0 ; } -void evlist__exit (struct evlist *evlist) +static void evlist__exit (struct evlist *evlist) {metricgroup__rblist_exit (&evlist->metric_events);event_enable_timer__exit (&evlist->eet);
@@ -202,11 +209,14 @@ void evlist__exit(struct evlist *evlist)perf_evlist__exit (&evlist->core); } -void evlist__delete (struct evlist *evlist) +void evlist__put (struct evlist *evlist) {if (evlist ==NULL )return ; +if (!refcount_dec_and_test (&evlist->refcnt)) +return ; +evlist__free_stats (evlist);evlist__munmap (evlist);evlist__close (evlist);
tools/perf/util/evlist.h +3 -4
@@ -61,6 +61,7 @@ struct event_enable_timer;struct evlist {struct perf_evlist core; +refcount_t refcnt;bool enabled;bool no_affinity;int id_pos;
@@ -109,10 +110,8 @@ struct evsel_str_handler {struct evlist *evlist__new (void );struct evlist *evlist__new_default (const struct target *target,bool sample_callchains);struct evlist *evlist__new_dummy (void ); -void evlist__init (struct evlist *evlist,struct perf_cpu_map *cpus, -struct perf_thread_map *threads); -void evlist__exit (struct evlist *evlist); -void evlist__delete (struct evlist *evlist); +struct evlist *evlist__get (struct evlist *evlist); +void evlist__put (struct evlist *evlist);void evlist__add (struct evlist *evlist,struct evsel *entry);void evlist__remove (struct evlist *evlist,struct evsel *evsel);
tools/perf/util/expr.c +1 -1
@@ -468,7 +468,7 @@ double expr__has_event(const struct expr_parse_ctx *ctx, bool compute_ids, const ret =parse_event (tmp, id) ?0 :1 ; } out: -evlist__delete (tmp); +evlist__put (tmp);return ret; }
tools/perf/util/header.c +11 -11
@@ -5187,7 +5187,7 @@ int perf_session__read_header(struct perf_session *session)pr_err ("Invalid ids section size %" PRIu64" for attr %d, not aligned to u64\n" , f_attr.ids.size, i); err = -EINVAL ; - goto out_delete_evlist; + goto out_put_evlist; }/*
@@ -5200,7 +5200,7 @@ int perf_session__read_header(struct perf_session *session)pr_err ("Invalid ids section size %" PRIu64" for attr %d, too many IDs\n" , f_attr.ids.size, i); err = -EINVAL ; - goto out_delete_evlist; + goto out_put_evlist; }/*
@@ -5213,19 +5213,19 @@ int perf_session__read_header(struct perf_session *session)pr_err ("Invalid ids section for attr %d: offset=%" PRIu64" size=%" PRIu64" exceeds file size %" PRIu64"\n" , i, f_attr.ids.offset, f_attr.ids.size, (u64 )input_stat.st_size); err = -EINVAL ; - goto out_delete_evlist; + goto out_put_evlist; } tmp =lseek (fd,0 ,SEEK_CUR ); evsel =evsel__new (&f_attr.attr);if (evsel ==NULL ) - goto out_delete_evlist; + goto out_put_evlist; evsel->needs_swap = header->needs_swap;/* * Do it before so that if perf_evsel__alloc_id fails, this -* entry gets purged too at evlist__delete(). +* entry gets purged too at evlist__put(). */ evlist__add (session->evlist, evsel);
@@ -5236,7 +5236,7 @@ int perf_session__read_header(struct perf_session *session)* hattr->ids threads. */ if (perf_evsel__alloc_id (&evsel->core,1 , nr_ids)) - goto out_delete_evlist; + goto out_put_evlist;lseek (fd, f_attr.ids.offset,SEEK_SET );
@@ -5266,18 +5266,18 @@ int perf_session__read_header(struct perf_session *session) err =perf_header__process_sections (header, fd, &session->tevent, perf_file_section__process);if (err <0 ) - goto out_delete_evlist; + goto out_put_evlist;if (evlist__prepare_tracepoint_events (session->evlist, session->tevent.pevent)) { err = -ENOMEM ; - goto out_delete_evlist; + goto out_put_evlist; }#else err =perf_header__process_sections (header, fd,NULL , perf_file_section__process);if (err <0 ) - goto out_delete_evlist; + goto out_put_evlist;#endif }
@@ -5303,8 +5303,8 @@ int perf_session__read_header(struct perf_session *session) out_errno:return -errno; -out_delete_evlist: -evlist__delete (session->evlist); +out_put_evlist: +evlist__put (session->evlist); session->evlist =NULL ;return err; }
tools/perf/util/metricgroup.c +3 -3
@@ -214,7 +214,7 @@ static void metric__free(struct metric *m)zfree (&m->metric_refs);expr__ctx_free (m->pctx);zfree (&m->modifier); -evlist__delete (m->evlist); +evlist__put (m->evlist);free (m); }
@@ -1331,7 +1331,7 @@ static int parse_ids(bool metric_no_merge, bool fake_pmu, parsed_evlist =NULL ; err_out:parse_events_error__exit (&parse_error); -evlist__delete (parsed_evlist); +evlist__put (parsed_evlist);strbuf_release (&events);return ret; }
@@ -1542,7 +1542,7 @@ static int parse_groups(struct evlist *perf_evlist,if (combined_evlist) {evlist__splice_list_tail (perf_evlist, &combined_evlist->core.entries); -evlist__delete (combined_evlist); +evlist__put (combined_evlist); }list_for_each_entry (m, &metric_list, nd) {
tools/perf/util/parse-events.c +2 -2
@@ -2343,7 +2343,7 @@ int __parse_events(struct evlist *evlist, const char *str, const char *pmu_filte/* * There are 2 users - builtin-record and builtin-test objects. -* Both call evlist__delete in case of error, so we dont +* Both call evlist__put in case of error, so we dont * need to bother. */ return ret;
@@ -2546,7 +2546,7 @@ int parse_events_option_new_evlist(const struct option *opt, const char *str, in } ret =parse_events_option (opt, str, unset);if (ret) { -evlist__delete (*args->evlistp); +evlist__put (*args->evlistp); *args->evlistp =NULL ; }
tools/perf/util/perf_api_probe.c +1 -1
@@ -57,7 +57,7 @@ static int perf_do_probe_api(setup_probe_fn_t fn, struct perf_cpu cpu, const cha err =0 ; out_delete: -evlist__delete (evlist); +evlist__put (evlist);return err; }
tools/perf/util/python.c +110 -132
451 lines not on this page · raw message
tools/perf/util/record.c +1 -1
@@ -264,7 +264,7 @@ bool evlist__can_select_event(struct evlist *evlist, const char *str) ret = true; out_delete: -evlist__delete (temp_evlist); +evlist__put (temp_evlist);return ret; }
tools/perf/util/session.c +1 -1
@@ -264,7 +264,7 @@ void perf_session__delete(struct perf_session *session)machines__exit (&session->machines);if (session->data) {if (perf_data__is_read (session->data)) -evlist__delete (session->evlist); +evlist__put (session->evlist);perf_data__close (session->data); }#ifdef HAVE_LIBTRACEEVENT
tools/perf/util/sideband_evlist.c +8 -8
@@ -102,7 +102,7 @@ int evlist__start_sb_thread(struct evlist *evlist, struct target *target)return 0 ;if (evlist__create_maps (evlist, target)) - goto out_delete_evlist; + goto out_put_evlist;if (evlist->core.nr_entries >1 ) {bool can_sample_identifier =perf_can_sample_identifier ();
@@ -116,25 +116,25 @@ int evlist__start_sb_thread(struct evlist *evlist, struct target *target)evlist__for_each_entry (evlist, counter) {if (evsel__open (counter, evlist->core.user_requested_cpus, evlist->core.threads) <0 ) - goto out_delete_evlist; + goto out_put_evlist; }if (evlist__mmap (evlist,UINT_MAX )) - goto out_delete_evlist; + goto out_put_evlist;evlist__for_each_entry (evlist, counter) {if (evsel__enable (counter)) - goto out_delete_evlist; + goto out_put_evlist; } evlist->thread.done =0 ;if (pthread_create (&evlist->thread.th,NULL , perf_evlist__poll_thread, evlist)) - goto out_delete_evlist; + goto out_put_evlist;return 0 ; -out_delete_evlist: -evlist__delete (evlist); +out_put_evlist: +evlist__put (evlist); evlist =NULL ;return -1 ; }
@@ -145,5 +145,5 @@ void evlist__stop_sb_thread(struct evlist *evlist)return ; evlist->thread.done =1 ;pthread_join (evlist->thread.th,NULL ); -evlist__delete (evlist); +evlist__put (evlist); }
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
Ian Rogers
reviewerPATCH v15 6/19 perf evsel: Add reference count
In reply to Ian RogersAs with evlist this a no-op for most of the perf tool. The reference count is set to 1 at allocation, the put will see the 1, decrement it and perform the delete. The purpose for adding the reference count is for the python code. Prior to this change the python code would clone evsels, but this has issues if events are opened, etc. leading to assertion failures. With a reference count the same evsel can be used and the reference count incremented for the python usage. To not change the python evsel API getset functions are added for the evsel members, no set function is provided for size as it doesn't make sense to alter this.
---
v12:
- Restored idx parsing in pyrf_evsel__init kwargs, and removed an erroneous Py_INCREF(pevsel) in pyrf_evlist__add that caused a memory leak.
v2:
1. Fixed Potential Crash in pyrf_event__new : Initialized pevent->evsel = NULL; to avoid garbage pointer dereference if evlist__event2evsel() fails in read_on_cpu .
2. Fixed Memory Leak in pyrf_evsel__init : Added evsel__put(pevsel->evsel) before overwriting it to handle repeated __init__ calls.
3. Fixed Exception Contract: Added PyErr_NoMemory() when evsel__new() fails in pyrf_evsel__init .
4. Fixed NULL Pointer Dereference on Property Access: Added a custom tp_getattro ( pyrf_evsel__getattro ) to pyrf_evsel__type to check if pevsel->evsel is NULL and raise a ValueError if so, covering all property accesses.
5. Fixed Reference Count in pyrf_evlist__add : Added evsel__get(evsel) when adding to the evlist .
6. Fixed Reference Count in pyrf_evlist__read_on_cpu : Added evsel__get(evsel) when assigning to pevent->evsel .
v7:
- Added pyrf_evsel__new to zero-initialize pevsel->evsel to fix crash on re-initialization.
v8:
- Added O! type validation to pyrf_evlist__add to prevent type confusion.
v10:
- Added strict CHECK_INITIALIZED/CHECK_INITIALIZED_INT checks to all pyrf_evsel methods and property getters/setters to prevent NULL pointer dereferences if the object is instantiated without calling __init__.
--- tools/perf/builtin-trace.c | 12 +- tools/perf/tests/evsel-tp-sched.c | 4 +- tools/perf/tests/openat-syscall-all-cpus.c | 6 +- tools/perf/tests/openat-syscall.c | 6 +- tools/perf/util/bpf_counter_cgroup.c | 2 +- tools/perf/util/cgroup.c | 2 +- tools/perf/util/evlist.c | 2 +- tools/perf/util/evsel.c | 26 +- tools/perf/util/evsel.h | 11 +- tools/perf/util/parse-events.y | 2 +- tools/perf/util/pfm.c | 2 +- tools/perf/util/print-events.c | 2 +- tools/perf/util/python.c | 321 ++++++++++++++++++--- tools/perf/util/session.c | 3 + 14 files changed, 318 insertions(+), 83 deletions(-)
tools/perf/builtin-trace.c +6 -6
@@ -460,10 +460,10 @@ static int evsel__init_tp_ptr_field(struct evsel *evsel, struct tp_field *field, ({ struct syscall_tp *sc = __evsel__syscall_tp(evsel);\ evsel__init_tp_ptr_field(evsel, &sc->name, #name); }) -static void evsel__delete_priv(struct evsel *evsel) +static void evsel__put_and_free_priv(struct evsel *evsel) { zfree(&evsel->priv); - evsel__delete(evsel); + evsel__put(evsel); } static int evsel__init_syscall_tp(struct evsel *evsel)
@@ -543,7 +543,7 @@ static struct evsel *perf_evsel__raw_syscall_newtp(const char *direction, void *return evsel; out_delete: -evsel__delete_priv (evsel); +evsel__put_and_free_priv (evsel);return NULL ; }
@@ -3612,7 +3612,7 @@ static bool evlist__add_vfs_getname(struct evlist *evlist)list_del_init (&evsel->core.node); evsel->evlist =NULL ; -evsel__delete (evsel); +evsel__put (evsel); }return found;
@@ -3728,9 +3728,9 @@ static int trace__add_syscall_newtp(struct trace *trace)return ret; out_delete_sys_exit: -evsel__delete_priv (sys_exit); +evsel__put_and_free_priv (sys_exit); out_delete_sys_enter: -evsel__delete_priv (sys_enter); +evsel__put_and_free_priv (sys_enter); goto out; }
tools/perf/tests/evsel-tp-sched.c +2 -2
@@ -64,7 +64,7 @@ static int test__perf_evsel__tp_sched_test(struct test_suite *test __maybe_unuseif (evsel__test_field (evsel,"next_prio" ,4 , true)) ret =TEST_FAIL ; -evsel__delete (evsel); +evsel__put (evsel); evsel =evsel__newtp ("sched" ,"sched_wakeup" );
@@ -85,7 +85,7 @@ static int test__perf_evsel__tp_sched_test(struct test_suite *test __maybe_unuseif (evsel__test_field (evsel,"target_cpu" ,4 , true)) ret =TEST_FAIL ; -evsel__delete (evsel); +evsel__put (evsel);return ret; }
tools/perf/tests/openat-syscall-all-cpus.c +3 -3
@@ -59,7 +59,7 @@ static int test__openat_syscall_event_on_all_cpus(struct test_suite *test __mayb"tweak /proc/sys/kernel/perf_event_paranoid?\n" , str_error_r(errno, sbuf, sizeof(sbuf))); err =TEST_SKIP ; - goto out_evsel_delete; + goto out_evsel_put; }perf_cpu_map__for_each_cpu (cpu, idx, cpus) {
@@ -116,8 +116,8 @@ static int test__openat_syscall_event_on_all_cpus(struct test_suite *test __maybevsel__free_counts (evsel); out_close_fd:perf_evsel__close_fd (&evsel->core); -out_evsel_delete: -evsel__delete (evsel); +out_evsel_put: +evsel__put (evsel); out_cpu_map_delete:perf_cpu_map__put (cpus); out_thread_map_delete:
tools/perf/tests/openat-syscall.c +3 -3
@@ -42,7 +42,7 @@ static int test__openat_syscall_event(struct test_suite *test __maybe_unused,"tweak /proc/sys/kernel/perf_event_paranoid?\n" , str_error_r(errno, sbuf, sizeof(sbuf))); err =TEST_SKIP ; - goto out_evsel_delete; + goto out_evsel_put; }for (i =0 ; i < nr_openat_calls; ++i) {
@@ -64,8 +64,8 @@ static int test__openat_syscall_event(struct test_suite *test __maybe_unused, err =TEST_OK ; out_close_fd:perf_evsel__close_fd (&evsel->core); -out_evsel_delete: -evsel__delete (evsel); +out_evsel_put: +evsel__put (evsel); out_thread_map_delete:perf_thread_map__put (threads);return err;
tools/perf/util/bpf_counter_cgroup.c +1 -1
@@ -336,7 +336,7 @@ static int bperf_cgrp__destroy(struct evsel *evsel)return 0 ;bperf_cgroup_bpf__destroy (skel); -evsel__delete (cgrp_switch);// it'll destroy on_switch progs too +evsel__put (cgrp_switch);// it'll destroy on_switch progs too return 0 ; }
tools/perf/util/cgroup.c +1 -1
@@ -469,7 +469,7 @@ int evlist__expand_cgroup(struct evlist *evlist, const char *str, bool open_cgro/* copy the list and set to the new cgroup. */ evlist__for_each_entry (orig_list, pos) { -struct evsel *evsel =evsel__clone (/*dest=*/ NULL , pos); +struct evsel *evsel =evsel__clone (pos);if (evsel ==NULL ) goto out_err;
tools/perf/util/evlist.c +1 -1
@@ -194,7 +194,7 @@ static void evlist__purge(struct evlist *evlist)evlist__for_each_entry_safe (evlist, n, pos) {list_del_init (&pos->core.node); pos->evlist =NULL ; -evsel__delete (pos); +evsel__put (pos); } evlist->core.nr_entries =0 ;
tools/perf/util/evsel.c +16 -10
@@ -393,10 +393,11 @@ bool evsel__is_function_event(struct evsel *evsel)#undef FUNCTION_EVENT } -void evsel__init (struct evsel *evsel, +static void evsel__init (struct evsel *evsel,struct perf_event_attr *attr,int idx) {perf_evsel__init (&evsel->core, attr, idx); +refcount_set (&evsel->refcnt,1 ); evsel->tracking = !idx; evsel->unit =strdup ("" ); evsel->scale =1.0 ;
@@ -477,7 +478,7 @@ static int evsel__copy_config_terms(struct evsel *dst, struct evsel *src)* The assumption is that @orig is not configured nor opened yet. * So we only care about the attributes that can be set while it's parsed. */ -struct evsel *evsel__clone (struct evsel *dest,struct evsel *orig) +struct evsel *evsel__clone (struct evsel *orig) {struct evsel *evsel;
@@ -490,11 +491,7 @@ struct evsel *evsel__clone(struct evsel *dest, struct evsel *orig)if (orig->bpf_obj)return NULL ; -if (dest) - evsel = dest; -else - evsel =evsel__new (&orig->core.attr); - + evsel =evsel__new (&orig->core.attr);if (evsel ==NULL )return NULL ;
@@ -579,7 +576,7 @@ struct evsel *evsel__clone(struct evsel *dest, struct evsel *orig)return evsel; out_err: -evsel__delete (evsel); +evsel__put (evsel);return NULL ; }
@@ -638,6 +635,12 @@ struct evsel *evsel__newtp_idx(const char *sys, const char *name, int idx, boolreturn ERR_PTR (err); } +struct evsel *evsel__get (struct evsel *evsel) +{ +refcount_inc (&evsel->refcnt); +return evsel; +} +#ifdef HAVE_LIBTRACEEVENT struct tep_event *evsel__tp_format (struct evsel *evsel) {
@@ -1976,7 +1979,7 @@ void evsel__set_priv_destructor(void (*destructor)(void *priv)) evsel__priv_destructor = destructor; } -void evsel__exit (struct evsel *evsel) +static void evsel__exit (struct evsel *evsel) {assert (list_empty (&evsel->core.node));assert (evsel->evlist ==NULL );
@@ -2013,11 +2016,14 @@ void evsel__exit(struct evsel *evsel) } } -void evsel__delete (struct evsel *evsel) +void evsel__put (struct evsel *evsel) {if (!evsel)return ; +if (!refcount_dec_and_test (&evsel->refcnt)) +return ; +evsel__exit (evsel);free (evsel); }
tools/perf/util/evsel.h +6 -5
@@ -6,6 +6,7 @@#include <linux/list.h> #include <linux/perf_event.h> +#include <linux/refcount.h> #include <linux/types.h> #include <sys/types.h>
@@ -47,6 +48,7 @@ typedef int (evsel__sb_cb_t)(union perf_event *event, void *data);struct evsel {struct perf_evsel core;struct evlist *evlist; +refcount_t refcnt;off_t id_offset;int id_pos;int is_pos;
@@ -268,7 +270,7 @@ static inline struct evsel *evsel__new(struct perf_event_attr *attr)return evsel__new_idx (attr,0 ); } -struct evsel *evsel__clone (struct evsel *dest,struct evsel *orig); +struct evsel *evsel__clone (struct evsel *orig);int copy_config_terms (struct list_head *dst,struct list_head *src);void free_config_terms (struct list_head *config_terms);
@@ -283,14 +285,13 @@ static inline struct evsel *evsel__newtp(const char *sys, const char *name)return evsel__newtp_idx (sys, name,0 , true); } +struct evsel *evsel__get (struct evsel *evsel); +void evsel__put (struct evsel *evsel); +#ifdef HAVE_LIBTRACEEVENT struct tep_event *evsel__tp_format (struct evsel *evsel);#endif -void evsel__init (struct evsel *evsel,struct perf_event_attr *attr,int idx); -void evsel__exit (struct evsel *evsel); -void evsel__delete (struct evsel *evsel); -void evsel__set_priv_destructor (void (*destructor)(void *priv));struct callchain_param ;
tools/perf/util/parse-events.y +1 -1
@@ -47,7 +47,7 @@ static void free_list_evsel(struct list_head* list_evsel) list_for_each_entry_safe(evsel, tmp, list_evsel, core.node) { list_del_init(&evsel->core.node); - evsel__delete(evsel); + evsel__put(evsel); } free(list_evsel); }
tools/perf/util/pfm.c +1 -1
@@ -159,7 +159,7 @@ static bool is_libpfm_event_supported(const char *name, struct perf_cpu_map *cpu result = false;evsel__close (evsel); -evsel__delete (evsel); +evsel__put (evsel);return result; }
tools/perf/util/print-events.c +1 -1
@@ -174,7 +174,7 @@ bool is_event_supported(u8 type, u64 config) }evsel__close (evsel); -evsel__delete (evsel); +evsel__put (evsel); }perf_thread_map__put (tmap);
tools/perf/util/python.c +273 -48
455 lines not on this page · raw message
tools/perf/util/session.c +3 -0
@@ -1844,7 +1844,10 @@ static int evlist__deliver_deferred_callchain(struct evlist *evlist,struct evsel *saved_evsel = sample->evsel; sample->evsel =evlist__id2evsel (evlist, sample->id); +if (sample->evsel) + sample->evsel =evsel__get (sample->evsel); ret = tool->callchain_deferred (tool, event, sample, machine); +evsel__put (sample->evsel); sample->evsel = saved_evsel;return ret; }
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
[email protected]
AI reviewRe: [PATCH v15 06/19] perf evsel: Add reference count
Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider: - [Medium] evsel references are leaked for all non-sample events read by Python because their types lack a…
Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [Medium] evsel references are leaked for all non-sample events read by Python because their types lack a tp_dealloc destructor.
- [Medium] Missing initialization check for the evsel parameter in pyrf_evlist__add leads to a NULL pointer dereference. --
commit 764f919043503b831e7c1c76a12ab53483beeeef
Author: Ian Rogers <[email protected]>
perf evsel: Add reference count
As with evlist this a no-op for most of the perf tool. The reference count is set to 1 at allocation, the put will see the 1, decrement it and perform the delete.
Quoted 8 lines
tools/perf/util/python.c +0
-0@@ -1584,13 +1815,12 @@ static PyObject *pyrf_evlist__add(struct pyrf_evlist *pevlist,PyObject *pevsel;struct evsel *evsel;
tools/perf/util/python.c static PyObject *pyrf_evlist__add(struct pyrf_evlist *pevlist, · line 1823
Quoted 10 lines
tools/perf/util/python.c +3
-4-if (!PyArg_ParseTuple (args,"O" , &pevsel)) +if (!PyArg_ParseTuple (args,"O!" , &pyrf_evsel__type, &pevsel))return NULL ; -Py_INCREF (pevsel); - evsel = &((struct pyrf_evsel *)pevsel)->evsel; + evsel = ((struct pyrf_evsel *)pevsel)->evsel; evsel->core.idx = evlist->core.nr_entries; -evlist__add (evlist, evsel); +evlist__add (evlist,evsel__get (evsel));
[Severity: Medium]
Is an initialization check missing for the evsel parameter here?
With the change to make evsel a pointer rather than an inline structure, if an uninitialized perf.evsel Python object is passed to pyrf_evlist__add(), the extracted pointer will be NULL. Dereferencing it immediately on the next line for core.idx will cause a crash.
I see this is fixed later in the series by commit 9f9c729b9d53 ('perf evlist: Add reference count checking') which adds the missing CHECK_INITIALIZED guards.
[ ... ]
tools/perf/util/python.c static PyObject *pyrf_evlist__read_on_cpu(struct pyrf_evlist *pevlist, · line 1883
Quoted 8 lines
tools/perf/util/python.c +1
-1@@ -1648,7 +1878,7 @@ static PyObject *pyrf_evlist__read_on_cpu(struct pyrf_evlist *pevlist,return Py_None; } - pevent->evsel = evsel; + pevent->evsel =evsel__get (evsel);perf_mmap__consume (&md->core);
[Severity: Medium]
Does this leak the evsel reference for non-sample events?
An evsel reference is acquired here for every event read from the mmap buffer, regardless of its type. However, looking at the deallocators, only pyrf_sample_event__delete() drops this reference using evsel__put().
Other event types (mmap, task, comm, etc.) use the standard Python object deallocator, which will skip calling evsel__put() and permanently leak the reference.
I see this is also fixed later in the series by commit 9d7ca1bbbd55 ('perf python: Use evsel in sample in pyrf_event'), which stops leaking the reference by removing pevent->evsel completely.
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=6
Ian Rogers
reviewerPATCH v15 7/19 perf evlist: Add reference count checking
In reply to Ian RogersNow the evlist is reference counted, add reference count checking so that gets and puts are paired and easy to debug. Reference count checking is documented here:
https://perfwiki.github.io/main/reference-count-checking/
This large patch is adding accessors to evlist functions and switching to their use. There was some minor renaming as evlist__mmap is now an accessor to the mmap variable, and the original evlist__mmap is renamed to evlist__do_mmap.
---
v12:
- Fixed evlist__put cycle collection TOCTOU double free race by checking refcount_read == 1. Fixed evlist__purge recursion and cycle reference tearing logic.
v2:
1. Fixed Memory Leak in evlist__new : Added free(evlist) in the else branch if ADD_RC_CHK fails, preventing a leak of the allocated raw structure.
2. Fixed Potential NULL Dereference: Added a NULL check after from_list_start(_evlist) in perf_evlist__mmap_cb_get() .
3. Fixed Use-After-Free Risk: In evlist__add() , I changed entry->evlist = evlist; to entry->evlist = evlist__get(evlist); . This ensures that the evsel holds a valid reference (wrapper) to the evlist , preventing it from becoming a dangling pointer if the original wrapper is freed.
4. Fixed Test Masking Bug: In test__perf_time__parse_for_ranges() , I replaced TEST_ASSERT_VAL with a manual check and return false; to avoid boolean evaluation of -1 inadvertently passing the test.
5. Fix reference count checker memory leaks from missed puts and due to cyclic evsel to evlist references. A leak still exists in __perf_evlist__propagate_maps due to empty CPU maps and not deleting the removed evsel.
v8:
1. Handle evsel__new failure in sample-parsing.c test and extra ref count checking integrity in evlist.c
2. Use standard per_sample initialization in cs-etm.c.
v10:
- Added CHECK_INITIALIZED checks to pyrf_evlist methods to prevent NULL pointer dereferences on uninitialized objects.
--- tools/perf/arch/arm/util/cs-etm.c | 10 +- tools/perf/arch/arm64/util/arm-spe.c | 8 +- tools/perf/arch/arm64/util/hisi-ptt.c | 2 +- tools/perf/arch/x86/tests/hybrid.c | 20 +- tools/perf/arch/x86/util/auxtrace.c | 2 +- tools/perf/arch/x86/util/intel-bts.c | 6 +- tools/perf/arch/x86/util/intel-pt.c | 9 +- tools/perf/arch/x86/util/iostat.c | 12 +- tools/perf/bench/evlist-open-close.c | 11 +- tools/perf/builtin-annotate.c | 7 +- tools/perf/builtin-ftrace.c | 6 +- tools/perf/builtin-inject.c | 4 +- tools/perf/builtin-kvm.c | 10 +- tools/perf/builtin-kwork.c | 8 +- tools/perf/builtin-lock.c | 2 +- tools/perf/builtin-record.c | 91 ++--- tools/perf/builtin-report.c | 6 +- tools/perf/builtin-sched.c | 24 +- tools/perf/builtin-script.c | 13 +- tools/perf/builtin-stat.c | 73 ++-- tools/perf/builtin-top.c | 52 +-- tools/perf/builtin-trace.c | 22 +- tools/perf/tests/backward-ring-buffer.c | 8 +- tools/perf/tests/code-reading.c | 10 +- tools/perf/tests/event-times.c | 2 +- tools/perf/tests/event_update.c | 2 +- tools/perf/tests/expand-cgroup.c | 4 +- tools/perf/tests/hwmon_pmu.c | 5 +- tools/perf/tests/keep-tracking.c | 8 +- tools/perf/tests/mmap-basic.c | 6 +- tools/perf/tests/openat-syscall-tp-fields.c | 8 +- tools/perf/tests/parse-events.c | 135 +++---- tools/perf/tests/parse-metric.c | 4 +- tools/perf/tests/perf-record.c | 22 +- tools/perf/tests/perf-time-to-tsc.c | 10 +- tools/perf/tests/pfm.c | 8 +- tools/perf/tests/pmu-events.c | 5 +- tools/perf/tests/sample-parsing.c | 45 +-- tools/perf/tests/sw-clock.c | 6 +- tools/perf/tests/switch-tracking.c | 9 +- tools/perf/tests/task-exit.c | 6 +- tools/perf/tests/time-utils-test.c | 14 +- tools/perf/tests/tool_pmu.c | 5 +- tools/perf/tests/topology.c | 2 +- tools/perf/tests/uncore-event-sorting.c | 4 +- tools/perf/ui/browsers/annotate.c | 2 +- tools/perf/ui/browsers/hists.c | 22 +- tools/perf/util/amd-sample-raw.c | 2 +- tools/perf/util/annotate-data.c | 2 +- tools/perf/util/annotate.c | 10 +- tools/perf/util/auxtrace.c | 14 +- tools/perf/util/block-info.c | 4 +- tools/perf/util/bpf_counter.c | 2 +- tools/perf/util/bpf_counter_cgroup.c | 12 +- tools/perf/util/bpf_ftrace.c | 9 +- tools/perf/util/bpf_lock_contention.c | 12 +- tools/perf/util/bpf_off_cpu.c | 14 +- tools/perf/util/cgroup.c | 20 +- tools/perf/util/cs-etm.c | 5 +- tools/perf/util/evlist.c | 395 ++++++++++++-------- tools/perf/util/evlist.h | 251 ++++++++++++- tools/perf/util/evsel.c | 6 +- tools/perf/util/evsel.h | 4 +- tools/perf/util/header.c | 47 +-- tools/perf/util/header.h | 2 +- tools/perf/util/intel-tpebs.c | 7 +- tools/perf/util/iostat.c | 2 +- tools/perf/util/iostat.h | 2 +- tools/perf/util/metricgroup.c | 6 +- tools/perf/util/parse-events.c | 6 +- tools/perf/util/pfm.c | 2 +- tools/perf/util/python.c | 90 +++-- tools/perf/util/record.c | 9 +- tools/perf/util/sample-raw.c | 4 +- tools/perf/util/session.c | 43 ++- tools/perf/util/sideband_evlist.c | 24 +- tools/perf/util/sort.c | 2 +- tools/perf/util/stat-display.c | 6 +- tools/perf/util/stat-shadow.c | 4 +- tools/perf/util/stat.c | 4 +- tools/perf/util/stream.c | 4 +- tools/perf/util/synthetic-events.c | 11 +- tools/perf/util/time-utils.c | 12 +- tools/perf/util/top.c | 4 +- 84 files changed, 1085 insertions(+), 718 deletions(-)
tools/perf/arch/arm/util/cs-etm.c +5 -5
@@ -201,7 +201,7 @@ static int cs_etm_validate_config(struct perf_pmu *cs_etm_pmu, {unsigned int idx;int err =0 ; -struct perf_cpu_map *event_cpus = evsel->evlist->core.user_requested_cpus; +struct perf_cpu_map *event_cpus =evlist__core (evsel->evlist)->user_requested_cpus;struct perf_cpu_map *intersect_cpus;struct perf_cpu cpu;
@@ -325,7 +325,7 @@ static int cs_etm_recording_options(struct auxtrace_record *itr, container_of(itr, struct cs_etm_recording, itr);struct perf_pmu *cs_etm_pmu = ptr->cs_etm_pmu;struct evsel *evsel, *cs_etm_evsel =NULL ; -struct perf_cpu_map *cpus = evlist->core.user_requested_cpus; +struct perf_cpu_map *cpus =evlist__core (evlist)->user_requested_cpus;bool privileged =perf_event_paranoid_check (-1 );int err =0 ;
@@ -551,7 +551,7 @@ cs_etm_info_priv_size(struct auxtrace_record *itr, {unsigned int idx;int etmv3 =0 , etmv4 =0 , ete =0 ; -struct perf_cpu_map *event_cpus = evlist->core.user_requested_cpus; +struct perf_cpu_map *event_cpus =evlist__core (evlist)->user_requested_cpus;struct perf_cpu_map *intersect_cpus;struct perf_cpu cpu;struct perf_pmu *cs_etm_pmu =cs_etm_get_pmu (itr);
@@ -790,7 +790,7 @@ static int cs_etm_info_fill(struct auxtrace_record *itr,u32 offset;u64 nr_cpu, type;struct perf_cpu_map *cpu_map; -struct perf_cpu_map *event_cpus = session->evlist->core.user_requested_cpus; +struct perf_cpu_map *event_cpus =evlist__core (session->evlist)->user_requested_cpus;struct perf_cpu_map *online_cpus =perf_cpu_map__new_online_cpus ();struct cs_etm_recording *ptr = container_of(itr, struct cs_etm_recording, itr);
@@ -800,7 +800,7 @@ static int cs_etm_info_fill(struct auxtrace_record *itr,if (priv_size !=cs_etm_info_priv_size (itr, session->evlist))return -EINVAL ; -if (!session->evlist->core.nr_mmaps) +if (!evlist__core (session->evlist)->nr_mmaps)return -EINVAL ;/* If the cpu_map has the "any" CPU all online CPUs are involved */
tools/perf/arch/arm64/util/arm-spe.c +4 -4
@@ -60,7 +60,7 @@ static bool arm_spe_is_set_freq(struct evsel *evsel)*/ static struct perf_cpu_map *arm_spe_find_cpus (struct evlist *evlist) { -struct perf_cpu_map *event_cpus = evlist->core.user_requested_cpus; +struct perf_cpu_map *event_cpus =evlist__core (evlist)->user_requested_cpus;struct perf_cpu_map *online_cpus =perf_cpu_map__new_online_cpus ();struct perf_cpu_map *intersect_cpus;
@@ -157,7 +157,7 @@ static int arm_spe_info_fill(struct auxtrace_record *itr,if (priv_size !=arm_spe_info_priv_size (itr, session->evlist))return -EINVAL ; -if (!session->evlist->core.nr_mmaps) +if (!evlist__core (session->evlist)->nr_mmaps)return -EINVAL ; cpu_map =arm_spe_find_cpus (session->evlist);
@@ -363,7 +363,7 @@ static int arm_spe_setup_tracking_event(struct evlist *evlist, {int err;struct evsel *tracking_evsel; -struct perf_cpu_map *cpus = evlist->core.user_requested_cpus; +struct perf_cpu_map *cpus =evlist__core (evlist)->user_requested_cpus;/* Add dummy event to keep tracking */ err =parse_event (evlist,"dummy:u" );
@@ -396,7 +396,7 @@ static int arm_spe_recording_options(struct auxtrace_record *itr,struct arm_spe_recording *sper = container_of(itr, struct arm_spe_recording, itr);struct evsel *evsel, *tmp; -struct perf_cpu_map *cpus = evlist->core.user_requested_cpus; +struct perf_cpu_map *cpus =evlist__core (evlist)->user_requested_cpus;bool discard = false;int err;u64 discard_bit;
tools/perf/arch/arm64/util/hisi-ptt.c +1 -1
@@ -53,7 +53,7 @@ static int hisi_ptt_info_fill(struct auxtrace_record *itr,if (priv_size !=HISI_PTT_AUXTRACE_PRIV_SIZE )return -EINVAL ; -if (!session->evlist->core.nr_mmaps) +if (!evlist__core (session->evlist)->nr_mmaps)return -EINVAL ; auxtrace_info->type =PERF_AUXTRACE_HISI_PTT ;
tools/perf/arch/x86/tests/hybrid.c +10 -10
@@ -26,7 +26,7 @@ static int test__hybrid_hw_event_with_pmu(struct evlist *evlist) {struct evsel *evsel =evlist__first (evlist); -TEST_ASSERT_VAL ("wrong number of entries" ,1 == evlist->core.nr_entries); +TEST_ASSERT_VAL ("wrong number of entries" ,1 ==evlist__nr_entries (evlist));TEST_ASSERT_VAL ("wrong type" ,PERF_TYPE_HARDWARE == evsel->core.attr.type);TEST_ASSERT_VAL ("wrong hybrid type" ,test_hybrid_type (evsel,PERF_TYPE_RAW ));TEST_ASSERT_VAL ("wrong config" ,test_config (evsel,PERF_COUNT_HW_CPU_CYCLES ));
@@ -38,7 +38,7 @@ static int test__hybrid_hw_group_event(struct evlist *evlist)struct evsel *evsel, *leader; evsel = leader =evlist__first (evlist); -TEST_ASSERT_VAL ("wrong number of entries" ,2 == evlist->core.nr_entries); +TEST_ASSERT_VAL ("wrong number of entries" ,2 ==evlist__nr_entries (evlist));TEST_ASSERT_VAL ("wrong type" ,PERF_TYPE_HARDWARE == evsel->core.attr.type);TEST_ASSERT_VAL ("wrong hybrid type" ,test_hybrid_type (evsel,PERF_TYPE_RAW ));TEST_ASSERT_VAL ("wrong config" ,test_config (evsel,PERF_COUNT_HW_CPU_CYCLES ));
@@ -57,7 +57,7 @@ static int test__hybrid_sw_hw_group_event(struct evlist *evlist)struct evsel *evsel, *leader; evsel = leader =evlist__first (evlist); -TEST_ASSERT_VAL ("wrong number of entries" ,2 == evlist->core.nr_entries); +TEST_ASSERT_VAL ("wrong number of entries" ,2 ==evlist__nr_entries (evlist));TEST_ASSERT_VAL ("wrong type" ,PERF_TYPE_SOFTWARE == evsel->core.attr.type);TEST_ASSERT_VAL ("wrong leader" ,evsel__has_leader (evsel, leader));
@@ -74,7 +74,7 @@ static int test__hybrid_hw_sw_group_event(struct evlist *evlist)struct evsel *evsel, *leader; evsel = leader =evlist__first (evlist); -TEST_ASSERT_VAL ("wrong number of entries" ,2 == evlist->core.nr_entries); +TEST_ASSERT_VAL ("wrong number of entries" ,2 ==evlist__nr_entries (evlist));TEST_ASSERT_VAL ("wrong type" ,PERF_TYPE_HARDWARE == evsel->core.attr.type);TEST_ASSERT_VAL ("wrong hybrid type" ,test_hybrid_type (evsel,PERF_TYPE_RAW ));TEST_ASSERT_VAL ("wrong config" ,test_config (evsel,PERF_COUNT_HW_CPU_CYCLES ));
@@ -91,7 +91,7 @@ static int test__hybrid_group_modifier1(struct evlist *evlist)struct evsel *evsel, *leader; evsel = leader =evlist__first (evlist); -TEST_ASSERT_VAL ("wrong number of entries" ,2 == evlist->core.nr_entries); +TEST_ASSERT_VAL ("wrong number of entries" ,2 ==evlist__nr_entries (evlist));TEST_ASSERT_VAL ("wrong type" ,PERF_TYPE_HARDWARE == evsel->core.attr.type);TEST_ASSERT_VAL ("wrong hybrid type" ,test_hybrid_type (evsel,PERF_TYPE_RAW ));TEST_ASSERT_VAL ("wrong config" ,test_config (evsel,PERF_COUNT_HW_CPU_CYCLES ));
@@ -113,7 +113,7 @@ static int test__hybrid_raw1(struct evlist *evlist) {struct perf_evsel *evsel; -perf_evlist__for_each_evsel (&evlist->core, evsel) { +perf_evlist__for_each_evsel (evlist__core (evlist), evsel) {struct perf_pmu *pmu =perf_pmus__find_by_type (evsel->attr.type);TEST_ASSERT_VAL ("missing pmu" , pmu);
@@ -127,7 +127,7 @@ static int test__hybrid_raw2(struct evlist *evlist) {struct evsel *evsel =evlist__first (evlist); -TEST_ASSERT_VAL ("wrong number of entries" ,1 == evlist->core.nr_entries); +TEST_ASSERT_VAL ("wrong number of entries" ,1 ==evlist__nr_entries (evlist));TEST_ASSERT_VAL ("wrong type" ,PERF_TYPE_RAW == evsel->core.attr.type);TEST_ASSERT_VAL ("wrong config" ,test_config (evsel,0x1a ));return TEST_OK ;
@@ -137,7 +137,7 @@ static int test__hybrid_cache_event(struct evlist *evlist) {struct evsel *evsel =evlist__first (evlist); -TEST_ASSERT_VAL ("wrong number of entries" ,1 == evlist->core.nr_entries); +TEST_ASSERT_VAL ("wrong number of entries" ,1 ==evlist__nr_entries (evlist));TEST_ASSERT_VAL ("wrong type" ,PERF_TYPE_HW_CACHE == evsel->core.attr.type);TEST_ASSERT_VAL ("wrong config" ,0x2 == (evsel->core.attr.config &0xffffffff ));return TEST_OK ;
@@ -148,7 +148,7 @@ static int test__checkevent_pmu(struct evlist *evlist)struct evsel *evsel =evlist__first (evlist); -TEST_ASSERT_VAL ("wrong number of entries" ,1 == evlist->core.nr_entries); +TEST_ASSERT_VAL ("wrong number of entries" ,1 ==evlist__nr_entries (evlist));TEST_ASSERT_VAL ("wrong type" ,PERF_TYPE_RAW == evsel->core.attr.type);TEST_ASSERT_VAL ("wrong config" ,10 == evsel->core.attr.config);TEST_ASSERT_VAL ("wrong config1" ,1 == evsel->core.attr.config1);
@@ -168,7 +168,7 @@ static int test__hybrid_hw_group_event_2(struct evlist *evlist)struct evsel *evsel, *leader; evsel = leader =evlist__first (evlist); -TEST_ASSERT_VAL ("wrong number of entries" ,2 == evlist->core.nr_entries); +TEST_ASSERT_VAL ("wrong number of entries" ,2 ==evlist__nr_entries (evlist));TEST_ASSERT_VAL ("wrong type" ,PERF_TYPE_HARDWARE == evsel->core.attr.type);TEST_ASSERT_VAL ("wrong hybrid type" ,test_hybrid_type (evsel,PERF_TYPE_RAW ));TEST_ASSERT_VAL ("wrong config" ,test_config (evsel,PERF_COUNT_HW_CPU_CYCLES ));
tools/perf/arch/x86/util/auxtrace.c +1 -1
@@ -55,7 +55,7 @@ struct auxtrace_record *auxtrace_record__init(struct evlist *evlist, int *err) {char buffer[64 ]; -struct perf_cpu cpu =perf_cpu_map__min (evlist->core.all_cpus); +struct perf_cpu cpu =perf_cpu_map__min (evlist__core (evlist)->all_cpus);int ret; *err =0 ;
tools/perf/arch/x86/util/intel-bts.c +3 -3
@@ -79,10 +79,10 @@ static int intel_bts_info_fill(struct auxtrace_record *itr,if (priv_size !=INTEL_BTS_AUXTRACE_PRIV_SIZE )return -EINVAL ; -if (!session->evlist->core.nr_mmaps) +if (!evlist__core (session->evlist)->nr_mmaps)return -EINVAL ; - pc = session->evlist->mmap[0 ].core.base; + pc =evlist__mmap (session->evlist)[0 ].core.base;if (pc) { err =perf_read_tsc_conversion (pc, &tc);if (err) {
@@ -114,7 +114,7 @@ static int intel_bts_recording_options(struct auxtrace_record *itr, container_of(itr, struct intel_bts_recording, itr);struct perf_pmu *intel_bts_pmu = btsr->intel_bts_pmu;struct evsel *evsel, *intel_bts_evsel =NULL ; -const struct perf_cpu_map *cpus = evlist->core.user_requested_cpus; +const struct perf_cpu_map *cpus =evlist__core (evlist)->user_requested_cpus;bool privileged =perf_event_paranoid_check (-1 );if (opts->auxtrace_sample_mode) {
tools/perf/arch/x86/util/intel-pt.c +5 -4
@@ -360,10 +360,10 @@ static int intel_pt_info_fill(struct auxtrace_record *itr, filter =intel_pt_find_filter (session->evlist, ptr->intel_pt_pmu); filter_str_len = filter ?strlen (filter) :0 ; -if (!session->evlist->core.nr_mmaps) +if (!evlist__core (session->evlist)->nr_mmaps)return -EINVAL ; - pc = session->evlist->mmap[0 ].core.base; + pc =evlist__mmap (session->evlist)[0 ].core.base;if (pc) { err =perf_read_tsc_conversion (pc, &tc);if (err) {
@@ -376,7 +376,8 @@ static int intel_pt_info_fill(struct auxtrace_record *itr,ui__warning ("Intel Processor Trace: TSC not available\n" ); } - per_cpu_mmaps = !perf_cpu_map__is_any_cpu_or_is_empty (session->evlist->core.user_requested_cpus); + per_cpu_mmaps = !perf_cpu_map__is_any_cpu_or_is_empty ( +evlist__core (session->evlist)->user_requested_cpus); auxtrace_info->type =PERF_AUXTRACE_INTEL_PT ; auxtrace_info->priv[INTEL_PT_PMU_TYPE ] = intel_pt_pmu->type;
@@ -621,7 +622,7 @@ static int intel_pt_recording_options(struct auxtrace_record *itr,struct perf_pmu *intel_pt_pmu = ptr->intel_pt_pmu;bool have_timing_info, need_immediate = false;struct evsel *evsel, *intel_pt_evsel =NULL ; -const struct perf_cpu_map *cpus = evlist->core.user_requested_cpus; +const struct perf_cpu_map *cpus =evlist__core (evlist)->user_requested_cpus;bool privileged =perf_event_paranoid_check (-1 );u64 tsc_bit;int err;
tools/perf/arch/x86/util/iostat.c +7 -5
@@ -332,13 +332,15 @@ static int iostat_event_group(struct evlist *evl,return ret; } -int iostat_prepare (struct evlist *evlist,struct perf_stat_config *config) +int iostat_prepare (struct evlist **evlist_ptr,struct perf_stat_config *config) { -if (evlist->core.nr_entries >0 ) { +struct evlist *evlist = *evlist_ptr; + +if (evlist__nr_entries (evlist) >0 ) {pr_warning ("The -e and -M options are not supported." "All chosen events/metrics will be dropped\n" );evlist__put (evlist); - evlist =evlist__new (); + *evlist_ptr = evlist =evlist__new ();if (!evlist)return -ENOMEM ; }
@@ -400,7 +402,7 @@ void iostat_prefix(struct evlist *evlist,struct perf_stat_config *config, char *prefix, struct timespec *ts) { - struct iio_root_port *rp = evlist->selected->priv; + struct iio_root_port *rp = evlist__selected(evlist)->priv;if (rp) {/*
@@ -463,7 +465,7 @@ void iostat_print_counters(struct evlist *evlist,iostat_prefix (evlist, config, prefix, ts);fprintf (config->output,"%s" , prefix);evlist__for_each_entry (evlist, counter) { - perf_device = evlist->selected->priv; + perf_device =evlist__selected (evlist)->priv;if (perf_device && perf_device != counter->priv) {evlist__set_selected (evlist, counter);iostat_prefix (evlist, config, prefix, ts);
tools/perf/bench/evlist-open-close.c +6 -5
@@ -116,7 +116,7 @@ static int bench__do_evlist_open_close(struct evlist *evlist)return err; } - err =evlist__mmap (evlist, opts.mmap_pages); + err =evlist__do_mmap (evlist, opts.mmap_pages);if (err <0 ) {pr_err ("evlist__mmap: %s\n" ,str_error_r (errno, sbuf,sizeof (sbuf)));return err;
@@ -124,7 +124,7 @@ static int bench__do_evlist_open_close(struct evlist *evlist)evlist__enable (evlist);evlist__disable (evlist); -evlist__munmap (evlist); +evlist__do_munmap (evlist);evlist__close (evlist);return 0 ;
@@ -145,10 +145,11 @@ static int bench_evlist_open_close__run(char *evstr, const char *uid_str)init_stats (&time_stats); -printf (" Number of cpus:\t%d\n" ,perf_cpu_map__nr (evlist->core.user_requested_cpus)); -printf (" Number of threads:\t%d\n" , evlist->core.threads->nr); +printf (" Number of cpus:\t%d\n" , +perf_cpu_map__nr (evlist__core (evlist)->user_requested_cpus)); +printf (" Number of threads:\t%d\n" ,evlist__core (evlist)->threads->nr);printf (" Number of events:\t%d (%d fds)\n" , - evlist->core.nr_entries,evlist__count_evsel_fds (evlist)); +evlist__nr_entries (evlist),evlist__count_evsel_fds (evlist));printf (" Number of iterations:\t%d\n" , iterations);evlist__put (evlist);
tools/perf/builtin-annotate.c +5 -2
@@ -562,7 +562,7 @@ static int __cmd_annotate(struct perf_annotate *ann) goto out;if ((use_browser ==1 || ann->use_stdio2) && ann->has_br_stack) -if (session->evlist->nr_br_cntr >0 ) +if (evlist__nr_br_cntr (session->evlist) >0 ) annotate_opts.show_br_cntr = true;if (dump_trace) {
@@ -928,8 +928,11 @@ int cmd_annotate(int argc, const char **argv)* branch counters, if the corresponding branch info is available * in the perf data in the TUI mode. */ -if ((use_browser ==1 || annotate.use_stdio2) && annotate.has_br_stack) +if ((use_browser ==1 || annotate.use_stdio2) && annotate.has_br_stack) { sort__mode =SORT_MODE__BRANCH ; +if (evlist__nr_br_cntr (annotate.session->evlist) >0 ) + annotate_opts.show_br_cntr = true; + }if (setup_sorting (/*evlist=*/ NULL ,perf_session__env (annotate.session)) <0 )usage_with_options (annotate_usage, options);
tools/perf/builtin-ftrace.c +3 -3
@@ -377,9 +377,9 @@ static int set_tracing_pid(struct perf_ftrace *ftrace)if (target__has_cpu (&ftrace->target))return 0 ; -for (i =0 ; i <perf_thread_map__nr (ftrace->evlist->core.threads); i++) { +for (i =0 ; i <perf_thread_map__nr (evlist__core (ftrace->evlist)->threads); i++) {scnprintf (buf,sizeof (buf),"%d" , -perf_thread_map__pid (ftrace->evlist->core.threads, i)); +perf_thread_map__pid (evlist__core (ftrace->evlist)->threads, i));if (append_tracing_file ("set_ftrace_pid" , buf) <0 )return -1 ; }
@@ -413,7 +413,7 @@ static int set_tracing_cpumask(struct perf_cpu_map *cpumap)static int set_tracing_cpu (struct perf_ftrace *ftrace) { -struct perf_cpu_map *cpumap = ftrace->evlist->core.user_requested_cpus; +struct perf_cpu_map *cpumap =evlist__core (ftrace->evlist)->user_requested_cpus;if (!target__has_cpu (&ftrace->target))return 0 ;
tools/perf/builtin-inject.c +2 -2
@@ -1496,7 +1496,7 @@ static int synthesize_id_index(struct perf_inject *inject, size_t new_cnt)struct perf_session *session = inject->session;struct evlist *evlist = session->evlist;struct machine *machine = &session->machines.host; -size_t from = evlist->core.nr_entries - new_cnt; +size_t from =evlist__nr_entries (evlist) - new_cnt;return __perf_event__synthesize_id_index (&inject->tool, perf_event__repipe, evlist, machine, from);
@@ -2031,7 +2031,7 @@ static int host__finished_init(const struct perf_tool *tool, struct perf_sessionif (ret)return ret; - ret =synthesize_id_index (inject, gs->session->evlist->core.nr_entries); + ret =synthesize_id_index (inject,evlist__nr_entries (gs->session->evlist));if (ret) {pr_err ("Failed to synthesize id_index\n" );return ret;
tools/perf/builtin-kvm.c +5 -5
@@ -1221,7 +1221,7 @@ static s64 perf_kvm__mmap_read_idx(struct perf_kvm_stat *kvm, int idx,int err; *mmap_time =ULLONG_MAX ; - md = &evlist->mmap[idx]; + md = &evlist__mmap (evlist)[idx]; err =perf_mmap__read_init (&md->core);if (err <0 )return (err == -EAGAIN ) ?0 :-1 ;
@@ -1266,7 +1266,7 @@ static int perf_kvm__mmap_read(struct perf_kvm_stat *kvm)s64 n, ntotal =0 ;u64 flush_time =ULLONG_MAX , mmap_time; -for (i =0 ; i < kvm->evlist->core.nr_mmaps; i++) { +for (i =0 ; i <evlist__core (kvm->evlist)->nr_mmaps; i++) { n =perf_kvm__mmap_read_idx (kvm, i, &mmap_time);if (n <0 )return -1 ;
@@ -1449,7 +1449,7 @@ static int kvm_events_live_report(struct perf_kvm_stat *kvm)evlist__enable (kvm->evlist);while (!done) { -struct fdarray *fda = &kvm->evlist->core.pollfd; +struct fdarray *fda = &evlist__core (kvm->evlist)->pollfd;int rc; rc =perf_kvm__mmap_read (kvm);
@@ -1531,7 +1531,7 @@ static int kvm_live_open_events(struct perf_kvm_stat *kvm) goto out; } -if (evlist__mmap (evlist, kvm->opts.mmap_pages) <0 ) { +if (evlist__do_mmap (evlist, kvm->opts.mmap_pages) <0 ) {ui__error ("Failed to mmap the events: %s\n" ,str_error_r (errno, sbuf,sizeof (sbuf)));evlist__close (evlist);
@@ -1931,7 +1931,7 @@ static int kvm_events_live(struct perf_kvm_stat *kvm,perf_session__set_id_hdr_size (kvm->session);ordered_events__set_copy_on_queue (&kvm->session->ordered_events, true);machine__synthesize_threads (&kvm->session->machines.host, &kvm->opts.target, - kvm->evlist->core.threads, true, false,1 ); +evlist__core (kvm->evlist)->threads, true, false,1 ); err =kvm_live_open_events (kvm);if (err) goto out;
tools/perf/builtin-kwork.c +4 -4
@@ -1814,7 +1814,7 @@ static int perf_kwork__check_config(struct perf_kwork *kwork, } } -list_for_each_entry (evsel, &session->evlist->core.entries, core.node) { +list_for_each_entry (evsel, &evlist__core (session->evlist)->entries, core.node) {if (kwork->show_callchain && !evsel__has_callchain (evsel)) {pr_debug ("Samples do not have callchains\n" ); kwork->show_callchain =0 ;
@@ -1864,9 +1864,9 @@ static int perf_kwork__read_events(struct perf_kwork *kwork) goto out_delete; } - kwork->nr_events = session->evlist->stats.nr_events[0 ]; - kwork->nr_lost_events = session->evlist->stats.total_lost; - kwork->nr_lost_chunks = session->evlist->stats.nr_events[PERF_RECORD_LOST ]; + kwork->nr_events =evlist__stats (session->evlist)->nr_events[0 ]; + kwork->nr_lost_events =evlist__stats (session->evlist)->total_lost; + kwork->nr_lost_chunks =evlist__stats (session->evlist)->nr_events[PERF_RECORD_LOST ]; out_delete:perf_session__delete (session);
tools/perf/builtin-lock.c +1 -1
@@ -2129,7 +2129,7 @@ static int __cmd_contention(int argc, const char **argv)evlist__start_workload (con.evlist);while (!done) { -if (argc &&waitpid (con.evlist->workload.pid,NULL ,WNOHANG ) >0 ) +if (argc &&waitpid (evlist__workload_pid (con.evlist),NULL ,WNOHANG ) >0 )break ;sleep (1 ); }
tools/perf/builtin-record.c +46 -45
268 lines not on this page · raw message
tools/perf/builtin-report.c +3 -3
@@ -561,7 +561,7 @@ static int evlist__tty_browse_hists(struct evlist *evlist, struct report *rep, cif (!quiet) {fprintf (stdout,"#\n# Total Lost Samples: %" PRIu64"\n#\n" , - evlist->stats.total_lost_samples); +evlist__stats (evlist)->total_lost_samples); }evlist__for_each_entry (evlist, pos) {
@@ -1156,7 +1156,7 @@ static int __cmd_report(struct report *rep) PERF_HPP_REPORT__BLOCK_AVG_CYCLES, }; -if (session->evlist->nr_br_cntr >0 ) +if (evlist__nr_br_cntr (session->evlist) >0 ) block_hpps[nr_hpps++] =PERF_HPP_REPORT__BLOCK_BRANCH_COUNTER ; block_hpps[nr_hpps++] =PERF_HPP_REPORT__BLOCK_RANGE ;
@@ -1291,7 +1291,7 @@ static int process_attr(const struct perf_tool *tool __maybe_unused,* on events sample_type. */ sample_type =evlist__combined_sample_type (*pevlist); - session = (*pevlist)->session; + session =evlist__session (*pevlist);callchain_param_setup (sample_type,perf_session__e_machine (session,/*e_flags=*/ NULL ));return 0 ; }
tools/perf/builtin-sched.c +12 -12
@@ -2014,9 +2014,9 @@ static int perf_sched__read_events(struct perf_sched *sched) goto out_delete; } - sched->nr_events = session->evlist->stats.nr_events[0 ]; - sched->nr_lost_events = session->evlist->stats.total_lost; - sched->nr_lost_chunks = session->evlist->stats.nr_events[PERF_RECORD_LOST ]; + sched->nr_events =evlist__stats (session->evlist)->nr_events[0 ]; + sched->nr_lost_events =evlist__stats (session->evlist)->total_lost; + sched->nr_lost_chunks =evlist__stats (session->evlist)->nr_events[PERF_RECORD_LOST ]; } rc =0 ;
@@ -3292,7 +3292,7 @@ static int timehist_check_attr(struct perf_sched *sched,struct evsel *evsel;struct evsel_runtime *er; -list_for_each_entry (evsel, &evlist->core.entries, core.node) { +list_for_each_entry (evsel, &evlist__core (evlist)->entries, core.node) { er =evsel__get_runtime (evsel);if (er ==NULL ) {pr_err ("Failed to allocate memory for evsel runtime data\n" );
@@ -3464,9 +3464,9 @@ static int perf_sched__timehist(struct perf_sched *sched) goto out; } - sched->nr_events = evlist->stats.nr_events[0 ]; - sched->nr_lost_events = evlist->stats.total_lost; - sched->nr_lost_chunks = evlist->stats.nr_events[PERF_RECORD_LOST ]; + sched->nr_events =evlist__stats (evlist)->nr_events[0 ]; + sched->nr_lost_events =evlist__stats (evlist)->total_lost; + sched->nr_lost_chunks =evlist__stats (evlist)->nr_events[PERF_RECORD_LOST ];if (sched->summary)timehist_print_summary (sched, session);
@@ -3971,7 +3971,7 @@ static int perf_sched__schedstat_record(struct perf_sched *sched,if (err <0 ) goto out; - user_requested_cpus = evlist->core.user_requested_cpus; + user_requested_cpus =evlist__core (evlist)->user_requested_cpus; err = perf_event__synthesize_schedstat(&(sched->tool), process_synthesized_schedstat_event,
@@ -3987,7 +3987,7 @@ static int perf_sched__schedstat_record(struct perf_sched *sched,evlist__start_workload (evlist);while (!done) { -if (argc &&waitpid (evlist->workload.pid,NULL ,WNOHANG ) >0 ) +if (argc &&waitpid (evlist__workload_pid (evlist),NULL ,WNOHANG ) >0 )break ;sleep (1 ); }
@@ -4688,7 +4688,7 @@ static int perf_sched__schedstat_report(struct perf_sched *sched)if (err <0 ) goto out; - user_requested_cpus = session->evlist->core.user_requested_cpus; + user_requested_cpus =evlist__core (session->evlist)->user_requested_cpus; err =perf_session__process_events (session);
@@ -4864,7 +4864,7 @@ static int perf_sched__schedstat_live(struct perf_sched *sched,if (err <0 ) goto out; - user_requested_cpus = evlist->core.user_requested_cpus; + user_requested_cpus =evlist__core (evlist)->user_requested_cpus; err = perf_event__synthesize_schedstat(&(sched->tool), process_synthesized_event_live,
@@ -4880,7 +4880,7 @@ static int perf_sched__schedstat_live(struct perf_sched *sched,evlist__start_workload (evlist);while (!done) { -if (argc &&waitpid (evlist->workload.pid,NULL ,WNOHANG ) >0 ) +if (argc &&waitpid (evlist__workload_pid (evlist),NULL ,WNOHANG ) >0 )break ;sleep (1 ); }
tools/perf/builtin-script.c +7 -6
@@ -2229,9 +2229,10 @@ static int script_find_metrics(const struct pmu_metric *pm,evlist__for_each_entry (metric_evlist, metric_evsel) {struct evsel *script_evsel =map_metric_evsel_to_script_evsel (script_evlist, metric_evsel); -struct metric_event *metric_me =metricgroup__lookup (&metric_evlist->metric_events, - metric_evsel, -/*create=*/ false); +struct metric_event *metric_me = +metricgroup__lookup (evlist__metric_events (metric_evlist), + metric_evsel, +/*create=*/ false);if (script_evsel->metric_id ==NULL ) { script_evsel->metric_id = metric_evsel->metric_id;
@@ -2251,7 +2252,7 @@ static int script_find_metrics(const struct pmu_metric *pm,if (metric_me) {struct metric_expr *expr;struct metric_event *script_me = -metricgroup__lookup (&script_evlist->metric_events, +metricgroup__lookup (evlist__metric_events (script_evlist), script_evsel,/*create=*/ true);
@@ -2321,7 +2322,7 @@ static void perf_sample__fprint_metric(struct thread *thread,assert (stat_config.aggr_mode ==AGGR_GLOBAL ); stat_config.aggr_get_id = script_aggr_cpu_id_get; stat_config.aggr_map = -cpu_aggr_map__new (evsel->evlist->core.user_requested_cpus, +cpu_aggr_map__new (evlist__core (evsel->evlist)->user_requested_cpus, aggr_cpu_id__global,/*data=*/ NULL ,/*needs_sort=*/ false); }
@@ -3909,7 +3910,7 @@ static int set_maps(struct perf_script *script)if (WARN_ONCE (script->allocated,"stats double allocation\n" ))return -EINVAL ; -perf_evlist__set_maps (&evlist->core, script->cpus, script->threads); +perf_evlist__set_maps (evlist__core (evlist), script->cpus, script->threads);if (evlist__alloc_stats (&stat_config, evlist,/*alloc_raw=*/ true))return -ENOMEM ;
tools/perf/builtin-stat.c +39 -34
206 lines not on this page · raw message
tools/perf/builtin-top.c +26 -26
175 lines not on this page · raw message
tools/perf/builtin-trace.c +11 -11
@@ -2023,7 +2023,7 @@ static int trace__symbols_init(struct trace *trace, int argc, const char **argv, goto out; err = __machine__synthesize_threads(trace->host, &trace->tool, &trace->opts.target, - evlist->core.threads, trace__tool_process, + evlist__core(evlist)->threads, trace__tool_process, /*needs_mmap=*/callchain_param.enabled && !trace->summary_only, /*mmap_data=*/false,
@@ -4195,7 +4195,7 @@ static int trace__set_filter_pids(struct trace *trace) err =augmented_syscalls__set_filter_pids (trace->filter_pids.nr, trace->filter_pids.entries); } - }else if (perf_thread_map__pid (trace->evlist->core.threads,0 ) ==-1 ) { + }else if (perf_thread_map__pid (evlist__core (trace->evlist)->threads,0 ) ==-1 ) { err =trace__set_filter_loop_pids (trace); }
@@ -4509,7 +4509,7 @@ static int trace__run(struct trace *trace, int argc, const char **argv)fprintf (trace->output,"Couldn't run the workload!\n" ); goto out_put_evlist; } - workload_pid = evlist->workload.pid; + workload_pid =evlist__workload_pid (evlist); } err =evlist__open (evlist);
@@ -4561,7 +4561,7 @@ static int trace__run(struct trace *trace, int argc, const char **argv) goto out_error_apply_filters;if (!trace->summary_only || !trace->summary_bpf) { - err =evlist__mmap (evlist, trace->opts.mmap_pages); + err =evlist__do_mmap (evlist, trace->opts.mmap_pages);if (err <0 ) goto out_error_mmap; }
@@ -4580,8 +4580,8 @@ static int trace__run(struct trace *trace, int argc, const char **argv)if (trace->summary_bpf)trace_start_bpf_summary (); - trace->multiple_threads =perf_thread_map__pid (evlist->core.threads,0 ) ==-1 || -perf_thread_map__nr (evlist->core.threads) >1 || + trace->multiple_threads =perf_thread_map__pid (evlist__core (evlist)->threads,0 ) ==-1 || +perf_thread_map__nr (evlist__core (evlist)->threads) >1 ||evlist__first (evlist)->core.attr.inherit;/*
@@ -4598,11 +4598,11 @@ static int trace__run(struct trace *trace, int argc, const char **argv) again: before = trace->nr_events; -for (i =0 ; i < evlist->core.nr_mmaps; i++) { +for (i =0 ; i <evlist__core (evlist)->nr_mmaps; i++) {union perf_event *event;struct mmap *md; - md = &evlist->mmap[i]; + md = &evlist__mmap (evlist)[i];if (perf_mmap__read_init (&md->core) <0 )continue ;
@@ -5304,7 +5304,7 @@ static int trace__parse_cgroups(const struct option *opt, const char *str, int u {struct trace *trace = opt->value; -if (!list_empty (&trace->evlist->core.entries)) { +if (!list_empty (&evlist__core (trace->evlist)->entries)) {struct option o = { .value = &trace->evlist, };
@@ -5578,7 +5578,7 @@ int cmd_trace(int argc, const char **argv)* .perfconfig trace.add_events, and filter those out. */ if (!trace.trace_syscalls && !trace.trace_pgfaults && - trace.evlist->core.nr_entries ==0 /* Was --events used? */ ) { +evlist__nr_entries (trace.evlist) ==0 /* Was --events used? */ ) { trace.trace_syscalls = true; }/*
@@ -5664,7 +5664,7 @@ int cmd_trace(int argc, const char **argv) symbol_conf.use_callchain = true; } -if (trace.evlist->core.nr_entries >0 ) { +if (evlist__nr_entries (trace.evlist) >0 ) {bool use_btf = false;evlist__set_default_evsel_handler (trace.evlist, trace__event_handler);
tools/perf/tests/backward-ring-buffer.c +4 -4
@@ -34,8 +34,8 @@ static int count_samples(struct evlist *evlist, int *sample_count, {int i; -for (i =0 ; i < evlist->core.nr_mmaps; i++) { -struct mmap *map = &evlist->overwrite_mmap[i]; +for (i =0 ; i <evlist__core (evlist)->nr_mmaps; i++) { +struct mmap *map = &evlist__overwrite_mmap (evlist)[i];union perf_event *event;perf_mmap__read_init (&map->core);
@@ -65,7 +65,7 @@ static int do_test(struct evlist *evlist, int mmap_pages,int err;char sbuf[STRERR_BUFSIZE ]; - err =evlist__mmap (evlist, mmap_pages); + err =evlist__do_mmap (evlist, mmap_pages);if (err <0 ) {pr_debug ("evlist__mmap: %s\n" ,str_error_r (errno, sbuf,sizeof (sbuf)));
@@ -77,7 +77,7 @@ static int do_test(struct evlist *evlist, int mmap_pages,evlist__disable (evlist); err =count_samples (evlist, sample_count, comm_count); -evlist__munmap (evlist); +evlist__do_munmap (evlist);return err; }
tools/perf/tests/code-reading.c +5 -5
@@ -592,8 +592,8 @@ static int process_events(struct machine *machine, struct evlist *evlist,struct mmap *md;int i, ret; -for (i =0 ; i < evlist->core.nr_mmaps; i++) { - md = &evlist->mmap[i]; +for (i =0 ; i <evlist__core (evlist)->nr_mmaps; i++) { + md = &evlist__mmap (evlist)[i];if (perf_mmap__read_init (&md->core) <0 )continue ;
@@ -781,7 +781,7 @@ static int do_test_code_reading(bool try_kcore) goto out_put; } -perf_evlist__set_maps (&evlist->core, cpus, threads); +perf_evlist__set_maps (evlist__core (evlist), cpus, threads); str = events[evidx];pr_debug ("Parsing event '%s'\n" , str);
@@ -809,7 +809,7 @@ static int do_test_code_reading(bool try_kcore)pr_debug ("perf_evlist__open() failed!\n%s\n" , errbuf); } -perf_evlist__set_maps (&evlist->core,NULL ,NULL ); +perf_evlist__set_maps (evlist__core (evlist),NULL ,NULL );evlist__put (evlist); evlist =NULL ;continue ;
@@ -820,7 +820,7 @@ static int do_test_code_reading(bool try_kcore)if (events[evidx] ==NULL ) goto out_put; - ret =evlist__mmap (evlist,UINT_MAX ); + ret =evlist__do_mmap (evlist,UINT_MAX );if (ret <0 ) {pr_debug ("evlist__mmap failed\n" ); goto out_put;
tools/perf/tests/event-times.c +1 -1
@@ -50,7 +50,7 @@ static int attach__enable_on_exec(struct evlist *evlist)static int detach__enable_on_exec (struct evlist *evlist) { -waitpid (evlist->workload.pid,NULL ,0 ); +waitpid (evlist__workload_pid (evlist),NULL ,0 );return 0 ; }
tools/perf/tests/event_update.c +1 -1
@@ -92,7 +92,7 @@ static int test__event_update(struct test_suite *test __maybe_unused, int subtesTEST_ASSERT_VAL ("failed to allocate ids" , !perf_evsel__alloc_id (&evsel->core,1 ,1 )); -perf_evlist__id_add (&evlist->core, &evsel->core,0 ,0 ,123 ); +perf_evlist__id_add (evlist__core (evlist), &evsel->core,0 ,0 ,123 );free ((char *)evsel->unit); evsel->unit =strdup ("KRAVA" );
tools/perf/tests/expand-cgroup.c +2 -2
@@ -28,7 +28,7 @@ static int test_expand_events(struct evlist *evlist)TEST_ASSERT_VAL ("evlist is empty" , !evlist__empty (evlist)); - nr_events = evlist->core.nr_entries; + nr_events =evlist__nr_entries (evlist); ev_name =calloc (nr_events,sizeof (*ev_name));if (ev_name ==NULL ) {pr_debug ("memory allocation failure\n" );
@@ -54,7 +54,7 @@ static int test_expand_events(struct evlist *evlist) } ret =TEST_FAIL ; -if (evlist->core.nr_entries != nr_events * nr_cgrps) { +if (evlist__nr_entries (evlist) != nr_events * nr_cgrps) {pr_debug ("event count doesn't match\n" ); goto out; }
tools/perf/tests/hwmon_pmu.c +3 -2
@@ -184,9 +184,10 @@ static int do_test(size_t i, bool with_pmu, bool with_alias) } ret =TEST_OK ; -if (with_pmu ? (evlist->core.nr_entries !=1 ) : (evlist->core.nr_entries <1 )) { +if (with_pmu ? (evlist__nr_entries (evlist) !=1 ) + : (evlist__nr_entries (evlist) <1 )) {pr_debug ("FAILED %s:%d Unexpected number of events for '%s' of %d\n" , - __FILE__, __LINE__, str, evlist->core.nr_entries); + __FILE__, __LINE__, str,evlist__nr_entries (evlist)); ret =TEST_FAIL ; goto out; }
tools/perf/tests/keep-tracking.c +4 -4
@@ -37,8 +37,8 @@ static int find_comm(struct evlist *evlist, const char *comm)int i, found; found =0 ; -for (i =0 ; i < evlist->core.nr_mmaps; i++) { - md = &evlist->mmap[i]; +for (i =0 ; i <evlist__core (evlist)->nr_mmaps; i++) { + md = &evlist__mmap (evlist)[i];if (perf_mmap__read_init (&md->core) <0 )continue ;while ((event =perf_mmap__read_event (&md->core)) !=NULL ) {
@@ -87,7 +87,7 @@ static int test__keep_tracking(struct test_suite *test __maybe_unused, int subte evlist =evlist__new ();CHECK_NOT_NULL__ (evlist); -perf_evlist__set_maps (&evlist->core, cpus, threads); +perf_evlist__set_maps (evlist__core (evlist), cpus, threads);CHECK__ (parse_event (evlist,"dummy:u" ));CHECK__ (parse_event (evlist,"cpu-cycles:u" ));
@@ -106,7 +106,7 @@ static int test__keep_tracking(struct test_suite *test __maybe_unused, int subte goto out_err; } -CHECK__ (evlist__mmap (evlist,UINT_MAX )); +CHECK__ (evlist__do_mmap (evlist,UINT_MAX ));/* * First, test that a 'comm' event can be found when the event is
tools/perf/tests/mmap-basic.c +3 -3
@@ -81,7 +81,7 @@ static int test__basic_mmap(struct test_suite *test __maybe_unused, int subtest goto out_free_cpus; } -perf_evlist__set_maps (&evlist->core, cpus, threads); +perf_evlist__set_maps (evlist__core (evlist), cpus, threads);for (i =0 ; i < nsyscalls; ++i) {char name[64 ];
@@ -113,7 +113,7 @@ static int test__basic_mmap(struct test_suite *test __maybe_unused, int subtest expected_nr_events[i] =1 +rand () %127 ; } -if (evlist__mmap (evlist,128 ) <0 ) { +if (evlist__do_mmap (evlist,128 ) <0 ) {pr_debug ("failed to mmap events: %d (%s)\n" , errno,str_error_r (errno, sbuf,sizeof (sbuf))); goto out_put_evlist;
@@ -124,7 +124,7 @@ static int test__basic_mmap(struct test_suite *test __maybe_unused, int subtest syscalls[i](); } - md = &evlist->mmap[0 ]; + md = &evlist__mmap (evlist)[0 ];if (perf_mmap__read_init (&md->core) <0 ) goto out_init;
tools/perf/tests/openat-syscall-tp-fields.c +4 -4
@@ -64,7 +64,7 @@ static int test__syscall_openat_tp_fields(struct test_suite *test __maybe_unusedevsel__config (evsel, &opts,NULL ); -perf_thread_map__set_pid (evlist->core.threads,0 ,getpid ()); +perf_thread_map__set_pid (evlist__core (evlist)->threads,0 ,getpid ()); err =evlist__open (evlist);if (err <0 ) {
@@ -73,7 +73,7 @@ static int test__syscall_openat_tp_fields(struct test_suite *test __maybe_unused goto out_put_evlist; } - err =evlist__mmap (evlist,UINT_MAX ); + err =evlist__do_mmap (evlist,UINT_MAX );if (err <0 ) {pr_debug ("evlist__mmap: %s\n" ,str_error_r (errno, sbuf,sizeof (sbuf)));
@@ -90,11 +90,11 @@ static int test__syscall_openat_tp_fields(struct test_suite *test __maybe_unusedwhile (1 ) {int before = nr_events; -for (i =0 ; i < evlist->core.nr_mmaps; i++) { +for (i =0 ; i <evlist__core (evlist)->nr_mmaps; i++) {union perf_event *event;struct mmap *md; - md = &evlist->mmap[i]; + md = &evlist__mmap (evlist)[i];if (perf_mmap__read_init (&md->core) <0 )continue ;
tools/perf/tests/parse-events.c +70 -65
519 lines not on this page · raw message
tools/perf/tests/parse-metric.c +2 -2
@@ -53,7 +53,7 @@ static double compute_single(struct evlist *evlist, const char *name)struct evsel *evsel;evlist__for_each_entry (evlist, evsel) { - me =metricgroup__lookup (&evlist->metric_events, evsel, false); + me =metricgroup__lookup (evlist__metric_events (evlist), evsel, false);if (me !=NULL ) { list_for_each_entry (mexp, &me->head, nd) {if (strcmp (mexp->metric_name, name))
@@ -88,7 +88,7 @@ static int __compute_metric(const char *name, struct value *vals,return -ENOMEM ; } -perf_evlist__set_maps (&evlist->core, cpus,NULL ); +perf_evlist__set_maps (evlist__core (evlist), cpus,NULL );/* Parse the metric into metric_events list. */ pme_test =find_core_metrics_table ("testarch" ,"testcpu" );
tools/perf/tests/perf-record.c +11 -11
@@ -129,7 +129,7 @@ static int test__PERF_RECORD(struct test_suite *test __maybe_unused, int subtestevsel__set_sample_bit (evsel,TIME );evlist__config (evlist, &opts,NULL ); - err =sched__get_first_possible_cpu (evlist->workload.pid, cpu_mask); + err =sched__get_first_possible_cpu (evlist__workload_pid (evlist), cpu_mask);if (err <0 ) {pr_debug ("sched__get_first_possible_cpu: %s\n" ,str_error_r (errno, sbuf,sizeof (sbuf)));
@@ -142,7 +142,7 @@ static int test__PERF_RECORD(struct test_suite *test __maybe_unused, int subtest/* * So that we can check perf_sample.cpu on all the samples. */ -if (sched_setaffinity (evlist->workload.pid, cpu_mask_size, cpu_mask) <0 ) { +if (sched_setaffinity (evlist__workload_pid (evlist), cpu_mask_size, cpu_mask) <0 ) {pr_debug ("sched_setaffinity: %s\n" ,str_error_r (errno, sbuf,sizeof (sbuf)));evlist__cancel_workload (evlist);
@@ -166,7 +166,7 @@ static int test__PERF_RECORD(struct test_suite *test __maybe_unused, int subtest* fds in the same CPU to be injected in the same mmap ring buffer * (using ioctl(PERF_EVENT_IOC_SET_OUTPUT)). */ - err =evlist__mmap (evlist, opts.mmap_pages); + err =evlist__do_mmap (evlist, opts.mmap_pages);if (err <0 ) {pr_debug ("evlist__mmap: %s\n" ,str_error_r (errno, sbuf,sizeof (sbuf)));
@@ -188,11 +188,11 @@ static int test__PERF_RECORD(struct test_suite *test __maybe_unused, int subtestwhile (1 ) {int before = total_events; -for (i =0 ; i < evlist->core.nr_mmaps; i++) { +for (i =0 ; i <evlist__core (evlist)->nr_mmaps; i++) {union perf_event *event;struct mmap *md; - md = &evlist->mmap[i]; + md = &evlist__mmap (evlist)[i];if (perf_mmap__read_init (&md->core) <0 )continue ;
@@ -231,15 +231,15 @@ static int test__PERF_RECORD(struct test_suite *test __maybe_unused, int subtest ++errs; } -if ((pid_t )sample.pid != evlist->workload.pid) { +if ((pid_t )sample.pid !=evlist__workload_pid (evlist)) {pr_debug ("%s with unexpected pid, expected %d, got %d\n" , - name, evlist->workload.pid, sample.pid); + name,evlist__workload_pid (evlist), sample.pid); ++errs; } -if ((pid_t )sample.tid != evlist->workload.pid) { +if ((pid_t )sample.tid !=evlist__workload_pid (evlist)) {pr_debug ("%s with unexpected tid, expected %d, got %d\n" , - name, evlist->workload.pid, sample.tid); + name,evlist__workload_pid (evlist), sample.tid); ++errs; }
@@ -248,7 +248,7 @@ static int test__PERF_RECORD(struct test_suite *test __maybe_unused, int subtest type == PERF_RECORD_MMAP2 || type == PERF_RECORD_FORK || type == PERF_RECORD_EXIT) && - (pid_t)event->comm.pid != evlist->workload.pid) { + (pid_t)event->comm.pid != evlist__workload_pid(evlist)) {pr_debug ("%s with unexpected pid/tid\n" , name); ++errs; }
@@ -352,9 +352,9 @@ static int test__PERF_RECORD(struct test_suite *test __maybe_unused, int subtest } out_put_evlist:CPU_FREE (cpu_mask); -evlist__put (evlist); out:perf_sample__exit (&sample); +evlist__put (evlist);if (err == -EACCES )return TEST_SKIP ;if (err <0 || errs !=0 )
tools/perf/tests/perf-time-to-tsc.c +5 -5
@@ -99,7 +99,7 @@ static int test__perf_time_to_tsc(struct test_suite *test __maybe_unused, int su evlist =evlist__new ();CHECK_NOT_NULL__ (evlist); -perf_evlist__set_maps (&evlist->core, cpus, threads); +perf_evlist__set_maps (evlist__core (evlist), cpus, threads);CHECK__ (parse_event (evlist,"cpu-cycles:u" ));
@@ -121,9 +121,9 @@ static int test__perf_time_to_tsc(struct test_suite *test __maybe_unused, int su goto out_err; } -CHECK__ (evlist__mmap (evlist,UINT_MAX )); +CHECK__ (evlist__do_mmap (evlist,UINT_MAX )); - pc = evlist->mmap[0 ].core.base; + pc =evlist__mmap (evlist)[0 ].core.base; ret =perf_read_tsc_conversion (pc, &tc);if (ret) {if (ret == -EOPNOTSUPP ) {
@@ -145,8 +145,8 @@ static int test__perf_time_to_tsc(struct test_suite *test __maybe_unused, int suevlist__disable (evlist); -for (i =0 ; i < evlist->core.nr_mmaps; i++) { - md = &evlist->mmap[i]; +for (i =0 ; i <evlist__core (evlist)->nr_mmaps; i++) { + md = &evlist__mmap (evlist)[i];if (perf_mmap__read_init (&md->core) <0 )continue ;
tools/perf/tests/pfm.c +4 -4
@@ -69,12 +69,12 @@ static int test__pfm_events(struct test_suite *test __maybe_unused,if (evlist ==NULL )return -ENOMEM ; - opt.value = evlist; + opt.value = &evlist;parse_libpfm_events_option (&opt, table[i].events,0 );TEST_ASSERT_EQUAL (table[i].events, -count_pfm_events (&evlist->core), +count_pfm_events (evlist__core (evlist)), table[i].nr_events); TEST_ASSERT_EQUAL(table[i].events, evlist__nr_groups(evlist),
@@ -154,12 +154,12 @@ static int test__pfm_group(struct test_suite *test __maybe_unused,if (evlist ==NULL )return -ENOMEM ; - opt.value = evlist; + opt.value = &evlist;parse_libpfm_events_option (&opt, table[i].events,0 );TEST_ASSERT_EQUAL (table[i].events, -count_pfm_events (&evlist->core), +count_pfm_events (evlist__core (evlist)), table[i].nr_events); TEST_ASSERT_EQUAL(table[i].events, evlist__nr_groups(evlist),
tools/perf/tests/pmu-events.c +3 -2
@@ -869,7 +869,7 @@ static int test__parsing_callback(const struct pmu_metric *pm,return -ENOMEM ; } -perf_evlist__set_maps (&evlist->core, cpus,NULL ); +perf_evlist__set_maps (evlist__core (evlist), cpus,NULL ); err =metricgroup__parse_groups_test (evlist, table, pm->metric_name);if (err) {
@@ -895,7 +895,8 @@ static int test__parsing_callback(const struct pmu_metric *pm, k++; }evlist__for_each_entry (evlist, evsel) { -struct metric_event *me =metricgroup__lookup (&evlist->metric_events, evsel, false); +struct metric_event *me =metricgroup__lookup (evlist__metric_events (evlist), + evsel, false);if (me !=NULL ) {struct metric_expr *mexp;
tools/perf/tests/sample-parsing.c +24 -21
112 lines not on this page · raw message
tools/perf/tests/sw-clock.c +3 -3
@@ -71,7 +71,7 @@ static int __test__sw_clock_freq(enum perf_sw_ids clock_id) goto out_put_evlist; } -perf_evlist__set_maps (&evlist->core, cpus, threads); +perf_evlist__set_maps (evlist__core (evlist), cpus, threads);if (evlist__open (evlist)) {const char *knob ="/proc/sys/kernel/perf_event_max_sample_rate" ;
@@ -83,7 +83,7 @@ static int __test__sw_clock_freq(enum perf_sw_ids clock_id) goto out_put_evlist; } - err =evlist__mmap (evlist,128 ); + err =evlist__do_mmap (evlist,128 );if (err <0 ) {pr_debug ("failed to mmap event: %d (%s)\n" , errno,str_error_r (errno, sbuf,sizeof (sbuf)));
@@ -98,7 +98,7 @@ static int __test__sw_clock_freq(enum perf_sw_ids clock_id)evlist__disable (evlist); - md = &evlist->mmap[0 ]; + md = &evlist__mmap (evlist)[0 ];if (perf_mmap__read_init (&md->core) <0 ) goto out_init;
tools/perf/tests/switch-tracking.c +5 -4
@@ -237,6 +237,7 @@ static int add_event(struct evlist *evlist, struct list_head *events,if (evlist__parse_sample (evlist, event, &sample)) {pr_debug ("evlist__parse_sample failed\n" ); +perf_sample__exit (&sample);return -1 ; }
@@ -282,8 +283,8 @@ static int process_events(struct evlist *evlist,struct mmap *md;int i, ret; -for (i =0 ; i < evlist->core.nr_mmaps; i++) { - md = &evlist->mmap[i]; +for (i =0 ; i <evlist__core (evlist)->nr_mmaps; i++) { + md = &evlist__mmap (evlist)[i];if (perf_mmap__read_init (&md->core) <0 )continue ;
@@ -374,7 +375,7 @@ static int test__switch_tracking(struct test_suite *test __maybe_unused, int sub goto out_err; } -perf_evlist__set_maps (&evlist->core, cpus, threads); +perf_evlist__set_maps (evlist__core (evlist), cpus, threads);/* First event */ err =parse_event (evlist,"cpu-clock:u" );
@@ -471,7 +472,7 @@ static int test__switch_tracking(struct test_suite *test __maybe_unused, int sub goto out; } - err =evlist__mmap (evlist,UINT_MAX ); + err =evlist__do_mmap (evlist,UINT_MAX );if (err) {pr_debug ("evlist__mmap failed!\n" ); goto out_err;
tools/perf/tests/task-exit.c +3 -3
@@ -77,7 +77,7 @@ static int test__task_exit(struct test_suite *test __maybe_unused, int subtest _ goto out_put_evlist; } -perf_evlist__set_maps (&evlist->core, cpus, threads); +perf_evlist__set_maps (evlist__core (evlist), cpus, threads); err =evlist__prepare_workload (evlist, &target, argv, false, workload_exec_failed_signal);if (err <0 ) {
@@ -104,7 +104,7 @@ static int test__task_exit(struct test_suite *test __maybe_unused, int subtest _ goto out_put_evlist; } -if (evlist__mmap (evlist,128 ) <0 ) { +if (evlist__do_mmap (evlist,128 ) <0 ) {pr_debug ("failed to mmap events: %d (%s)\n" , errno,str_error_r (errno, sbuf,sizeof (sbuf))); err =-1 ;
@@ -114,7 +114,7 @@ static int test__task_exit(struct test_suite *test __maybe_unused, int subtest _evlist__start_workload (evlist); retry: - md = &evlist->mmap[0 ]; + md = &evlist__mmap (evlist)[0 ];if (perf_mmap__read_init (&md->core) <0 ) goto out_init;
tools/perf/tests/time-utils-test.c +9 -5
@@ -69,16 +69,19 @@ struct test_data {static bool test__perf_time__parse_for_ranges (struct test_data *d) { -struct evlist evlist = { - .first_sample_time = d->first, - .last_sample_time = d->last, - }; -struct perf_session session = { .evlist = &evlist }; +struct evlist *evlist =evlist__new (); +struct perf_session session = { .evlist = evlist };struct perf_time_interval *ptime =NULL ;int range_size, range_num;bool pass = false;int i, err; +if (!evlist) { +pr_debug ("Missing evlist\n" ); +return false; + } +evlist__set_first_sample_time (evlist, d->first); +evlist__set_last_sample_time (evlist, d->last);pr_debug ("\nperf_time__parse_for_ranges(\"%s\")\n" , d->str);if (strchr (d->str,'%' ))
@@ -127,6 +130,7 @@ static bool test__perf_time__parse_for_ranges(struct test_data *d) pass = true; out: +evlist__put (evlist);free (ptime);return pass; }
tools/perf/tests/tool_pmu.c +3 -2
@@ -40,9 +40,10 @@ static int do_test(enum tool_pmu_event ev, bool with_pmu) } ret =TEST_OK ; -if (with_pmu ? (evlist->core.nr_entries !=1 ) : (evlist->core.nr_entries <1 )) { +if (with_pmu ? (evlist__nr_entries (evlist) !=1 ) + : (evlist__nr_entries (evlist) <1 )) {pr_debug ("FAILED %s:%d Unexpected number of events for '%s' of %d\n" , - __FILE__, __LINE__, str, evlist->core.nr_entries); + __FILE__, __LINE__, str,evlist__nr_entries (evlist)); ret =TEST_FAIL ; goto out; }
tools/perf/tests/topology.c +1 -1
@@ -46,7 +46,7 @@ static int session_write_header(char *path) session->evlist =evlist__new_default (&target,/*sample_callchains=*/ false);TEST_ASSERT_VAL ("can't get evlist" , session->evlist); - session->evlist->session = session; +evlist__set_session (session->evlist, session);perf_header__set_feat (&session->header,HEADER_CPU_TOPOLOGY );perf_header__set_feat (&session->header,HEADER_NRCPUS );
tools/perf/tests/uncore-event-sorting.c +2 -2
@@ -147,8 +147,8 @@ static int test__uncore_event_sorting(struct test_suite *test __maybe_unused, goto out_err; } -CHECK_COND (evlist->core.nr_entries >=4 ,"Number of events is >= 4" ); -CHECK_EQUAL (evlist->core.nr_entries %2 ,0 ,"Number of events is a multiple of 2" ); +CHECK_COND (evlist__nr_entries (evlist) >=4 ,"Number of events is >= 4" ); +CHECK_EQUAL (evlist__nr_entries (evlist) %2 ,0 ,"Number of events is a multiple of 2" );evlist__for_each_entry (evlist, evsel) {struct evsel *next;
tools/perf/ui/browsers/annotate.c +1 -1
@@ -597,7 +597,7 @@ static bool annotate_browser__callq(struct annotate_browser *browser, notes =symbol__annotation (dl->ops.target.sym);annotation__lock (notes); -if (!symbol__hists (dl->ops.target.sym, evsel->evlist->core.nr_entries)) { +if (!symbol__hists (dl->ops.target.sym,evlist__nr_entries (evsel->evlist))) {annotation__unlock (notes);ui__warning ("Not enough memory for annotating '%s' symbol!\n" , dl->ops.target.sym->name);
tools/perf/ui/browsers/hists.c +11 -11
@@ -688,10 +688,10 @@ static int hist_browser__handle_hotkey(struct hist_browser *browser, bool warn_lui_browser__update_nr_entries (&browser->b, nr_entries);if (warn_lost_event && - (evsel->evlist->stats.nr_lost_warned != - evsel->evlist->stats.nr_events[PERF_RECORD_LOST ])) { - evsel->evlist->stats.nr_lost_warned = - evsel->evlist->stats.nr_events[PERF_RECORD_LOST ]; + (evlist__stats (evsel->evlist)->nr_lost_warned != +evlist__stats (evsel->evlist)->nr_events[PERF_RECORD_LOST ])) { +evlist__stats (evsel->evlist)->nr_lost_warned = +evlist__stats (evsel->evlist)->nr_events[PERF_RECORD_LOST ];ui_browser__warn_lost_events (&browser->b); }
@@ -3321,7 +3321,7 @@ static int evsel__hists_browse(struct evsel *evsel, int nr_events, const char *h* No need to refresh, resort/decay histogram * entries if we are not collecting samples: */ -if (top->evlist->enabled) { +if (evlist__enabled (top->evlist)) { helpline ="Press 'f' to disable the events or 'h' to see other hotkeys" ; hbt->refresh = delay_secs; }else {
@@ -3493,7 +3493,7 @@ static void perf_evsel_menu__write(struct ui_browser *browser, unit, unit == ' ' ? "" : " ", ev_name);ui_browser__printf (browser,"%s" , bf); - nr_events = evsel->evlist->stats.nr_events[PERF_RECORD_LOST ]; + nr_events =evlist__stats (evsel->evlist)->nr_events[PERF_RECORD_LOST ];if (nr_events !=0 ) { menu->lost_events = true;if (!current_entry)
@@ -3559,13 +3559,13 @@ static int perf_evsel_menu__run(struct evsel_menu *menu,ui_browser__show_title (&menu->b, title);switch (key) {case K_TAB : -if (pos->core.node.next == &evlist->core.entries) +if (pos->core.node.next == &evlist__core (evlist)->entries) pos =evlist__first (evlist);else pos =evsel__next (pos); goto browse_hists;case K_UNTAB : -if (pos->core.node.prev == &evlist->core.entries) +if (pos->core.node.prev == &evlist__core (evlist)->entries) pos =evlist__last (evlist);else pos =evsel__prev (pos);
@@ -3618,7 +3618,7 @@ static int __evlist__tui_browse_hists(struct evlist *evlist, int nr_entries, construct evsel *pos;struct evsel_menu menu = { .b = { - .entries = &evlist->core.entries, + .entries = &evlist__core (evlist)->entries, .refresh = ui_browser__list_head_refresh, .seek = ui_browser__list_head_seek, .write = perf_evsel_menu__write,
@@ -3646,7 +3646,7 @@ static int __evlist__tui_browse_hists(struct evlist *evlist, int nr_entries, constatic bool evlist__single_entry (struct evlist *evlist) { -int nr_entries = evlist->core.nr_entries; +int nr_entries =evlist__nr_entries (evlist);if (nr_entries ==1 )return true;
@@ -3664,7 +3664,7 @@ static bool evlist__single_entry(struct evlist *evlist)int evlist__tui_browse_hists (struct evlist *evlist,const char *help,struct hist_browser_timer *hbt,float min_pcnt,struct perf_env *env,bool warn_lost_event) { -int nr_entries = evlist->core.nr_entries; +int nr_entries =evlist__nr_entries (evlist);if (evlist__single_entry (evlist)) { single_entry: {
tools/perf/util/amd-sample-raw.c +1 -1
@@ -421,7 +421,7 @@ static void parse_cpuid(struct perf_env *env)*/ bool evlist__has_amd_ibs (struct evlist *evlist) { -struct perf_env *env =perf_session__env (evlist->session); +struct perf_env *env =perf_session__env (evlist__session (evlist));int ret, nr_pmu_mappings =perf_env__nr_pmu_mappings (env);const char *pmu_mapping =perf_env__pmu_mappings (env);char name[sizeof ("ibs_fetch" )];
tools/perf/util/annotate-data.c +1 -1
@@ -1829,7 +1829,7 @@ int annotated_data_type__update_samples(struct annotated_data_type *adt,return 0 ;if (adt->histograms ==NULL ) { -int nr = evsel->evlist->core.nr_entries; +int nr =evlist__nr_entries (evsel->evlist);if (alloc_data_type_histograms (adt, nr) <0 )return -1 ;
tools/perf/util/annotate.c +5 -5
@@ -328,7 +328,7 @@ static int symbol__inc_addr_samples(struct map_symbol *ms,if (sym ==NULL )return 0 ; - src =symbol__hists (sym, sample->evsel->evlist->core.nr_entries); + src =symbol__hists (sym,evlist__nr_entries (sample->evsel->evlist));return src ?__symbol__inc_addr_samples (ms, src, addr, sample) :0 ; }
@@ -339,7 +339,7 @@ static int symbol__account_br_cntr(struct annotated_branch *branch, {unsigned int br_cntr_nr =evsel__leader (evsel)->br_cntr_nr;unsigned int base =evsel__leader (evsel)->br_cntr_idx; -unsigned int off = offset * evsel->evlist->nr_br_cntr; +unsigned int off = offset *evlist__nr_br_cntr (evsel->evlist);u64 *branch_br_cntr = branch->br_cntr;unsigned int i, mask, width;
@@ -369,7 +369,7 @@ static int symbol__account_cycles(u64 addr, u64 start, struct symbol *sym,if (sym ==NULL )return 0 ; - branch =symbol__find_branch_hist (sym, evsel->evlist->nr_br_cntr); + branch =symbol__find_branch_hist (sym,evlist__nr_br_cntr (evsel->evlist));if (!branch)return -ENOMEM ;if (addr < sym->start || addr >= sym->end)
@@ -511,7 +511,7 @@ static void annotation__count_and_fill(struct annotation *notes, u64 start, u64static int annotation__compute_ipc (struct annotation *notes,size_t size,struct evsel *evsel) { -unsigned int br_cntr_nr = evsel->evlist->nr_br_cntr; +unsigned int br_cntr_nr =evlist__nr_br_cntr (evsel->evlist);int err =0 ;s64 offset;
@@ -1813,7 +1813,7 @@ int annotation_br_cntr_abbr_list(char **str, struct evsel *evsel, bool header)struct evsel *pos;struct strbuf sb; -if (evsel->evlist->nr_br_cntr <=0 ) +if (evlist__nr_br_cntr (evsel->evlist) <=0 )return -ENOTSUP ;strbuf_init (&sb,/*hint=*/ 0 );
tools/perf/util/auxtrace.c +7 -7
@@ -191,7 +191,7 @@ void auxtrace_mmap_params__set_idx(struct auxtrace_mmap_params *mp,struct evlist *evlist, struct evsel *evsel, int idx) { - bool per_cpu = !perf_cpu_map__has_any_cpu(evlist->core.user_requested_cpus); + bool per_cpu = !perf_cpu_map__has_any_cpu(evlist__core(evlist)->user_requested_cpus); mp->mmap_needed = evsel->needs_auxtrace_mmap;
@@ -201,11 +201,11 @@ void auxtrace_mmap_params__set_idx(struct auxtrace_mmap_params *mp, mp->idx = idx;if (per_cpu) { - mp->cpu =perf_cpu_map__cpu (evlist->core.all_cpus, idx); - mp->tid =perf_thread_map__pid (evlist->core.threads,0 ); + mp->cpu =perf_cpu_map__cpu (evlist__core (evlist)->all_cpus, idx); + mp->tid =perf_thread_map__pid (evlist__core (evlist)->threads,0 ); }else { mp->cpu.cpu =-1 ; - mp->tid =perf_thread_map__pid (evlist->core.threads, idx); + mp->tid =perf_thread_map__pid (evlist__core (evlist)->threads, idx); } }
@@ -668,10 +668,10 @@ int auxtrace_parse_snapshot_options(struct auxtrace_record *itr,static int evlist__enable_event_idx (struct evlist *evlist,struct evsel *evsel,int idx) { -bool per_cpu_mmaps = !perf_cpu_map__has_any_cpu (evlist->core.user_requested_cpus); +bool per_cpu_mmaps = !perf_cpu_map__has_any_cpu (evlist__core (evlist)->user_requested_cpus);if (per_cpu_mmaps) { -struct perf_cpu evlist_cpu =perf_cpu_map__cpu (evlist->core.all_cpus, idx); +struct perf_cpu evlist_cpu =perf_cpu_map__cpu (evlist__core (evlist)->all_cpus, idx);int cpu_map_idx =perf_cpu_map__idx (evsel->core.cpus, evlist_cpu);if (cpu_map_idx ==-1 )
@@ -1838,7 +1838,7 @@ void perf_session__auxtrace_error_inc(struct perf_session *session,struct perf_record_auxtrace_error *e = &event->auxtrace_error;if (e->type <PERF_AUXTRACE_ERROR_MAX ) - session->evlist->stats.nr_auxtrace_errors[e->type] +=1 ; +evlist__stats (session->evlist)->nr_auxtrace_errors[e->type] +=1 ; }void events_stats__auxtrace_error_warn (const struct events_stats *stats)
tools/perf/util/block-info.c +2 -2
@@ -472,7 +472,7 @@ struct block_report *block_info__create_report(struct evlist *evlist, int *nr_reps) {struct block_report *block_reports; -int nr_hists = evlist->core.nr_entries, i =0 ; +int nr_hists =evlist__nr_entries (evlist), i =0 ;struct evsel *pos; block_reports =calloc (nr_hists,sizeof (struct block_report ));
@@ -483,7 +483,7 @@ struct block_report *block_info__create_report(struct evlist *evlist,struct hists *hists =evsel__hists (pos);process_block_report (hists, &block_reports[i], total_cycles, - block_hpps, nr_hpps, evlist->nr_br_cntr); + block_hpps, nr_hpps,evlist__nr_br_cntr (evlist)); i++; }
tools/perf/util/bpf_counter.c +1 -1
@@ -443,7 +443,7 @@ static int bperf_check_target(struct evsel *evsel, }else if (target->tid) { *filter_type =BPERF_FILTER_PID ; *filter_entry_cnt =perf_thread_map__nr (evsel->core.threads); - }else if (target->pid || evsel->evlist->workload.pid !=-1 ) { + }else if (target->pid ||evlist__workload_pid (evsel->evlist) !=-1 ) { *filter_type =BPERF_FILTER_TGID ; *filter_entry_cnt =perf_thread_map__nr (evsel->core.threads); }else {
tools/perf/util/bpf_counter_cgroup.c +6 -6
@@ -104,7 +104,7 @@ static int bperf_load_program(struct evlist *evlist)set_max_rlimit (); -if (nr_cgroups ==0 || evlist->core.nr_entries % nr_cgroups !=0 ) { +if (nr_cgroups ==0 ||evlist__nr_entries (evlist) % nr_cgroups !=0 ) {pr_err ("Invalid cgroup or event count\n" );return -EINVAL ; }
@@ -116,7 +116,7 @@ static int bperf_load_program(struct evlist *evlist)pr_err ("Failed to open cgroup skeleton\n" );return -1 ; } -setup_rodata (skel, evlist->core.nr_entries); +setup_rodata (skel,evlist__nr_entries (evlist)); err =bperf_cgroup_bpf__load (skel);if (err) {
@@ -127,12 +127,12 @@ static int bperf_load_program(struct evlist *evlist) err =-1 ; cgrp_switch =evsel__new (&cgrp_switch_attr); -if (evsel__open_per_cpu (cgrp_switch, evlist->core.all_cpus,-1 ) <0 ) { +if (evsel__open_per_cpu (cgrp_switch,evlist__core (evlist)->all_cpus,-1 ) <0 ) {pr_err ("Failed to open cgroup switches event\n" ); goto out; } -perf_cpu_map__for_each_cpu (cpu, i, evlist->core.all_cpus) { +perf_cpu_map__for_each_cpu (cpu, i,evlist__core (evlist)->all_cpus) { link =bpf_program__attach_perf_event (skel->progs.on_cgrp_switch,FD (cgrp_switch, i));if (IS_ERR (link)) {
@@ -197,7 +197,7 @@ static int bperf_load_program(struct evlist *evlist)*/ {struct evsel *leader; -int num_events = evlist->core.nr_entries / nr_cgroups; +int num_events =evlist__nr_entries (evlist) / nr_cgroups;evlist__for_each_entry (evlist, evsel) { leader =evlist__find_evsel (evlist, evsel->core.idx % num_events);
@@ -258,7 +258,7 @@ static int bperf_cgrp__sync_counters(struct evlist *evlist)unsigned int idx;int prog_fd =bpf_program__fd (skel->progs.trigger_read); -perf_cpu_map__for_each_cpu (cpu, idx, evlist->core.all_cpus) +perf_cpu_map__for_each_cpu (cpu, idx,evlist__core (evlist)->all_cpus)bperf_trigger_reading (prog_fd, cpu.cpu);return 0 ;
tools/perf/util/bpf_ftrace.c +5 -4
@@ -59,13 +59,13 @@ int perf_ftrace__latency_prepare_bpf(struct perf_ftrace *ftrace)/* don't need to set cpu filter for system-wide mode */ if (ftrace->target.cpu_list) { - ncpus =perf_cpu_map__nr (ftrace->evlist->core.user_requested_cpus); + ncpus =perf_cpu_map__nr (evlist__core (ftrace->evlist)->user_requested_cpus);bpf_map__set_max_entries (skel->maps.cpu_filter, ncpus); skel->rodata->has_cpu =1 ; }if (target__has_task (&ftrace->target) ||target__none (&ftrace->target)) { - ntasks =perf_thread_map__nr (ftrace->evlist->core.threads); + ntasks =perf_thread_map__nr (evlist__core (ftrace->evlist)->threads);bpf_map__set_max_entries (skel->maps.task_filter, ntasks); skel->rodata->has_task =1 ; }
@@ -87,7 +87,8 @@ int perf_ftrace__latency_prepare_bpf(struct perf_ftrace *ftrace) fd =bpf_map__fd (skel->maps.cpu_filter);for (i =0 ; i < ncpus; i++) { - cpu =perf_cpu_map__cpu (ftrace->evlist->core.user_requested_cpus, i).cpu; + cpu =perf_cpu_map__cpu ( +evlist__core (ftrace->evlist)->user_requested_cpus, i).cpu;bpf_map_update_elem (fd, &cpu, &val,BPF_ANY ); } }
@@ -99,7 +100,7 @@ int perf_ftrace__latency_prepare_bpf(struct perf_ftrace *ftrace) fd =bpf_map__fd (skel->maps.task_filter);for (i =0 ; i < ntasks; i++) { - pid =perf_thread_map__pid (ftrace->evlist->core.threads, i); + pid =perf_thread_map__pid (evlist__core (ftrace->evlist)->threads, i);bpf_map_update_elem (fd, &pid, &val,BPF_ANY ); } }
tools/perf/util/bpf_lock_contention.c +6 -6
@@ -223,11 +223,11 @@ int lock_contention_prepare(struct lock_contention *con)if (target__has_cpu (target)) { skel->rodata->has_cpu =1 ; - ncpus =perf_cpu_map__nr (evlist->core.user_requested_cpus); + ncpus =perf_cpu_map__nr (evlist__core (evlist)->user_requested_cpus); }if (target__has_task (target)) { skel->rodata->has_task =1 ; - ntasks =perf_thread_map__nr (evlist->core.threads); + ntasks =perf_thread_map__nr (evlist__core (evlist)->threads); }if (con->filters->nr_types) { skel->rodata->has_type =1 ;
@@ -334,7 +334,7 @@ int lock_contention_prepare(struct lock_contention *con) fd =bpf_map__fd (skel->maps.cpu_filter);for (i =0 ; i < ncpus; i++) { - cpu =perf_cpu_map__cpu (evlist->core.user_requested_cpus, i).cpu; + cpu =perf_cpu_map__cpu (evlist__core (evlist)->user_requested_cpus, i).cpu;bpf_map_update_elem (fd, &cpu, &val,BPF_ANY ); } }
@@ -346,13 +346,13 @@ int lock_contention_prepare(struct lock_contention *con) fd =bpf_map__fd (skel->maps.task_filter);for (i =0 ; i < ntasks; i++) { - pid =perf_thread_map__pid (evlist->core.threads, i); + pid =perf_thread_map__pid (evlist__core (evlist)->threads, i);bpf_map_update_elem (fd, &pid, &val,BPF_ANY ); } } -if (target__none (target) && evlist->workload.pid >0 ) { -u32 pid = evlist->workload.pid; +if (target__none (target) &&evlist__workload_pid (evlist) >0 ) { +u32 pid =evlist__workload_pid (evlist);u8 val =1 ; fd =bpf_map__fd (skel->maps.task_filter);
tools/perf/util/bpf_off_cpu.c +7 -7
@@ -73,13 +73,13 @@ static void off_cpu_start(void *arg)/* update task filter for the given workload */ if (skel->rodata->has_task && skel->rodata->uses_tgid && -perf_thread_map__pid (evlist->core.threads,0 ) !=-1 ) { +perf_thread_map__pid (evlist__core (evlist)->threads,0 ) !=-1 ) {int fd;u32 pid;u8 val =1 ; fd =bpf_map__fd (skel->maps.task_filter); - pid =perf_thread_map__pid (evlist->core.threads,0 ); + pid =perf_thread_map__pid (evlist__core (evlist)->threads,0 );bpf_map_update_elem (fd, &pid, &val,BPF_ANY ); }
@@ -168,7 +168,7 @@ int off_cpu_prepare(struct evlist *evlist, struct target *target,/* don't need to set cpu filter for system-wide mode */ if (target->cpu_list) { - ncpus =perf_cpu_map__nr (evlist->core.user_requested_cpus); + ncpus =perf_cpu_map__nr (evlist__core (evlist)->user_requested_cpus);bpf_map__set_max_entries (skel->maps.cpu_filter, ncpus); skel->rodata->has_cpu =1 ; }
@@ -199,7 +199,7 @@ int off_cpu_prepare(struct evlist *evlist, struct target *target, skel->rodata->has_task =1 ; skel->rodata->uses_tgid =1 ; }else if (target__has_task (target)) { - ntasks =perf_thread_map__nr (evlist->core.threads); + ntasks =perf_thread_map__nr (evlist__core (evlist)->threads);bpf_map__set_max_entries (skel->maps.task_filter, ntasks); skel->rodata->has_task =1 ; }else if (target__none (target)) {
@@ -209,7 +209,7 @@ int off_cpu_prepare(struct evlist *evlist, struct target *target, }if (evlist__first (evlist)->cgrp) { - ncgrps = evlist->core.nr_entries -1 ;/* excluding a dummy */ + ncgrps =evlist__nr_entries (evlist) -1 ;/* excluding a dummy */ bpf_map__set_max_entries (skel->maps.cgroup_filter, ncgrps);if (!cgroup_is_v2 ("perf_event" ))
@@ -240,7 +240,7 @@ int off_cpu_prepare(struct evlist *evlist, struct target *target, fd =bpf_map__fd (skel->maps.cpu_filter);for (i =0 ; i < ncpus; i++) { - cpu =perf_cpu_map__cpu (evlist->core.user_requested_cpus, i).cpu; + cpu =perf_cpu_map__cpu (evlist__core (evlist)->user_requested_cpus, i).cpu;bpf_map_update_elem (fd, &cpu, &val,BPF_ANY ); } }
@@ -269,7 +269,7 @@ int off_cpu_prepare(struct evlist *evlist, struct target *target, fd =bpf_map__fd (skel->maps.task_filter);for (i =0 ; i < ntasks; i++) { - pid =perf_thread_map__pid (evlist->core.threads, i); + pid =perf_thread_map__pid (evlist__core (evlist)->threads, i);bpf_map_update_elem (fd, &pid, &val,BPF_ANY ); } }
tools/perf/util/cgroup.c +10 -10
@@ -367,7 +367,7 @@ int parse_cgroups(const struct option *opt, const char *str,char *s;int ret, i; -if (list_empty (&evlist->core.entries)) { +if (list_empty (&evlist__core (evlist)->entries)) {fprintf (stderr,"must define events before cgroups\n" );return -1 ; }
@@ -423,7 +423,7 @@ int evlist__expand_cgroup(struct evlist *evlist, const char *str, bool open_cgroint ret =-1 ;int prefix_len; -if (evlist->core.nr_entries ==0 ) { +if (evlist__nr_entries (evlist) ==0 ) {fprintf (stderr,"must define events before cgroups\n" );return -EINVAL ; }
@@ -436,11 +436,11 @@ int evlist__expand_cgroup(struct evlist *evlist, const char *str, bool open_cgro }/* save original events and init evlist */ -evlist__splice_list_tail (orig_list, &evlist->core.entries); - evlist->core.nr_entries =0 ; +evlist__splice_list_tail (orig_list, &evlist__core (evlist)->entries); +evlist__core (evlist)->nr_entries =0 ; - orig_metric_events = evlist->metric_events; -metricgroup__rblist_init (&evlist->metric_events); + orig_metric_events = *evlist__metric_events (evlist); +metricgroup__rblist_init (evlist__metric_events (evlist));if (has_pattern_string (str)) prefix_len =match_cgroups (str);
@@ -503,15 +503,15 @@ int evlist__expand_cgroup(struct evlist *evlist, const char *str, bool open_cgro nr_cgroups++;if (metricgroup__copy_metric_events (tmp_list, cgrp, - &evlist->metric_events, +evlist__metric_events (evlist), &orig_metric_events) <0 ) goto out_err; -evlist__splice_list_tail (evlist, &tmp_list->core.entries); - tmp_list->core.nr_entries =0 ; +evlist__splice_list_tail (evlist, &evlist__core (tmp_list)->entries); +evlist__core (tmp_list)->nr_entries =0 ; } -if (list_empty (&evlist->core.entries)) { +if (list_empty (&evlist__core (evlist)->entries)) {fprintf (stderr,"no cgroup matched: %s\n" , str); goto out_err; }
tools/perf/util/cs-etm.c +4 -1
@@ -1672,8 +1672,9 @@ static int cs_etm__synth_branch_sample(struct cs_etm_queue *etmq, {int ret =0 ;struct cs_etm_auxtrace *etm = etmq->etm; -struct perf_sample sample = {.ip =0 ,}; +struct perf_sample sample;union perf_event *event = tidq->event_buf; +struct dummy_branch_stack {u64 nr;u64 hw_idx;
@@ -1681,6 +1682,7 @@ static int cs_etm__synth_branch_sample(struct cs_etm_queue *etmq, } dummy_bs;u64 ip; +perf_sample__init (&sample,/*all=*/ true); ip =cs_etm__last_executed_instr (tidq->prev_packet); event->sample.header.type =PERF_RECORD_SAMPLE ;
@@ -1733,6 +1735,7 @@ static int cs_etm__synth_branch_sample(struct cs_etm_queue *etmq,"CS ETM Trace: failed to deliver instruction event, error %d\n" , ret); +perf_sample__exit (&sample);return ret; }
tools/perf/util/evlist.c +233 -162
172 lines not on this page · raw message
tools/perf/util/evlist.h +235 -16
251 lines not on this page · raw message
tools/perf/util/header.c +24 -23
47 lines not on this page · raw message
tools/perf/util/python.c +64 -26
90 lines not on this page · raw message
tools/perf/util/session.c +21 -22
43 lines not on this page · raw message
The rest of this diff is not shown. View the raw message.
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
Part of the diff is not shown here. View the raw message.
[email protected]
AI reviewRe: [PATCH v15 07/19] perf evlist: Add reference count checking
Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider: - [High] Memory leak due to flawed reference cycle collection in `evlist__put()` when an `evsel` outlives…
Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [High] Memory leak due to flawed reference cycle collection in evlist__put() when an evsel outlives its parent evlist.
- [High] Concurrent evlist__put calls can cause a double free of the evlist due to a lockless TOCTOU race in cycle collection. --
commit 9f9c729b9d53b88f9a8695a329f14caec324824b
Author: Ian Rogers <[email protected]>
perf evlist: Add reference count checking
This patch adds reference count checking for the evlist structure so that gets and puts are paired properly and easier to debug. It adds accessors for evlist functions and switches existing code to use them.
Quoted 3 lines
tools/perf/util/evlist.c +0
-0
[ ... ]
Quoted 25 lines
tools/perf/util/evlist.c +16
-2@@ -203,18 +229,40 @@ static void evlist__exit(struct evlist *evlist)void evlist__put (struct evlist *evlist) { +struct evsel *evsel; +unsigned int count, ref_cnt; +if (evlist ==NULL )return ; -if (!refcount_dec_and_test (&evlist->refcnt)) -return ; +if (refcount_dec_and_test (evlist__refcnt (evlist))) + goto out_delete; + +retry: + count =refcount_read (evlist__refcnt (evlist)); + ref_cnt = count; +evlist__for_each_entry (evlist, evsel) { +if (RC_CHK_EQUAL (evsel->evlist, evlist) && count && +refcount_read (&evsel->refcnt) ==1 ) + count--; + } +if (refcount_read (evlist__refcnt (evlist)) != ref_cnt) + goto retry;
tools/perf/util/evlist.c static void evlist__exit(struct evlist *evlist) · line 260
Quoted 10 lines
tools/perf/util/evlist.c +10
-0+ +if (count !=0 ) { +/* +* Not the last reference except for back references from +* evsels. +*/ +RC_CHK_PUT (evlist); +return ; + } +out_delete:
[Severity: High]
Does this reference cycle collection logic also need a symmetrical check in evsel__put() to prevent memory leaks?
This patch introduces a strong reference cycle between evlist and evsel via entry->evlist = evlist__get(evlist) in evlist__add(). While evlist__put() handles cycle detection to free the evlist if the only remaining references are from its evsels, evsel__put() lacks the inverse logic.
If an external reference to evlist is dropped while an external reference to an evsel is held, evlist__put() correctly leaves the objects alive. However, when the last external reference to the evsel is subsequently dropped, evsel__put() decrements the evsel refcount to 1 and returns. Since there is no logic to break the cycle from the evsel side, both the evlist and the evsel could permanently leak.
[Severity: High]
Can this lockless cycle collection lead to a double free due to a TOCTOU race condition?
If two threads drop the last two external references to an evlist concurrently, couldn't they both read the same stable refcount where count equals the number of cyclic references?
Both threads would iterate over the evsels, decrement their local count to 0, verify the global refcount matches their snapshot, and both fall through to the out_delete block. This would result in both threads concurrently executing evlist__purge() and freeing the object.
Quoted 9 lines
tools/perf/util/evlist.c +2
-2evlist__free_stats (evlist); -evlist__munmap (evlist); +evlist__do_munmap (evlist);evlist__close (evlist);evlist__purge (evlist);evlist__exit (evlist); -free (evlist); +RC_CHK_FREE (evlist); }
[ ... ]
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=7
Ian Rogers
reviewerPATCH v15 8/19 perf python: Use evsel in sample in pyrf_event
In reply to Ian RogersAvoid a duplicated evsel by using the one in sample. Add evsel__get/put to the evsel in perf_sample.
---
v12:
- Removed an extra evsel__get in evsel__parse_sample to fix a reference leak on error paths.
--- tools/perf/builtin-inject.c | 5 ++++- tools/perf/builtin-trace.c | 5 ++++- tools/perf/util/evsel.c | 2 +- tools/perf/util/intel-pt.c | 2 +- tools/perf/util/python.c | 10 +++------- tools/perf/util/s390-sample-raw.c | 1 + tools/perf/util/sample.c | 17 ++++++++++++----- tools/perf/util/session.c | 19 ++++++++++++++++--- 8 files changed, 42 insertions(+), 19 deletions(-)
tools/perf/builtin-inject.c +4 -1
@@ -702,8 +702,11 @@ static int perf_event__repipe_common_mmap(const struct perf_tool *tool, }if (dso && !dso__hit (dso)) { -if (!sample->evsel) +if (!sample->evsel) { sample->evsel =evlist__event2evsel (inject->session->evlist, event); +if (sample->evsel) +evsel__get (sample->evsel); + }if (sample->evsel) {dso__set_hit (dso);
tools/perf/builtin-trace.c +4 -1
@@ -3657,8 +3657,11 @@ static void trace__handle_event(struct trace *trace, union perf_event *event, streturn ; } -if (sample->evsel ==NULL ) +if (sample->evsel ==NULL ) { sample->evsel =evlist__id2evsel (trace->evlist, sample->id); +if (sample->evsel) +evsel__get (sample->evsel); + }if (sample->evsel ==NULL ) {fprintf (trace->output,"Unknown tp ID %" PRIu64", skipping...\n" , sample->id);
tools/perf/util/evsel.c +1 -1
@@ -3365,7 +3365,7 @@ int evsel__parse_sample(struct evsel *evsel, union perf_event *event,union u64_swap u;perf_sample__init (data,/*all=*/ true); - data->evsel = evsel; + data->evsel =evsel__get (evsel); data->cpu = data->pid = data->tid =-1 ; data->stream_id = data->id = data->time =-1ULL ; data->period = evsel->core.attr.sample_period;
tools/perf/util/intel-pt.c +1 -1
@@ -1830,9 +1830,9 @@ static int intel_pt_synth_branch_sample(struct intel_pt_queue *ptq) ptq->last_br_cyc_cnt = ptq->ipc_cyc_cnt; } -perf_sample__exit (&sample); ret =intel_pt_deliver_synth_event (pt, event, &sample, pt->branches_sample_type); +perf_sample__exit (&sample);return ret; }
tools/perf/util/python.c +3 -7
@@ -63,7 +63,6 @@ PyMODINIT_FUNC PyInit_perf(void);struct pyrf_event {PyObject_HEAD - struct evsel *evsel; struct perf_sample sample;union perf_event event; };
@@ -294,7 +293,6 @@ static PyMemberDef pyrf_sample_event__members[] = {static void pyrf_sample_event__delete (struct pyrf_event *pevent) { -evsel__put (pevent->evsel);perf_sample__exit (&pevent->sample);Py_TYPE (pevent)->tp_free ((PyObject *)pevent); }
@@ -316,7 +314,7 @@ static PyObject *pyrf_sample_event__repr(const struct pyrf_event *pevent)#ifdef HAVE_LIBTRACEEVENT static bool is_tracepoint (const struct pyrf_event *pevent) { -return pevent->evsel->core.attr.type ==PERF_TYPE_TRACEPOINT ; +return pevent->sample.evsel->core.attr.type ==PERF_TYPE_TRACEPOINT ; } static PyObject*
@@ -363,7 +361,7 @@ tracepoint_field(const struct pyrf_event *pe, struct tep_format_field *field)static PyObject *get_tracepoint_field (struct pyrf_event *pevent,PyObject *attr_name) { -struct evsel *evsel = pevent->evsel; +struct evsel *evsel = pevent->sample.evsel;struct tep_event *tp_format =evsel__tp_format (evsel);struct tep_format_field *field;
@@ -529,7 +527,7 @@ static PyObject *pyrf_event__new(const union perf_event *event) pevent = PyObject_New(struct pyrf_event, ptype);if (pevent !=NULL ) {memcpy (&pevent->event, event, event->header.size); - pevent->evsel =NULL ; +perf_sample__init (&pevent->sample,/*all=*/ false); }return (PyObject *)pevent; }
@@ -1905,8 +1903,6 @@ static PyObject *pyrf_evlist__read_on_cpu(struct pyrf_evlist *pevlist,return Py_None; } - pevent->evsel =evsel__get (evsel); -perf_mmap__consume (&md->core); err =evsel__parse_sample (evsel, &pevent->event, &pevent->sample);
tools/perf/util/s390-sample-raw.c +1 -0
@@ -343,6 +343,7 @@ void evlist__s390_sample_raw(struct evlist *evlist, union perf_event *event, sample->evsel =evlist__event2evsel (evlist, event);if (!sample->evsel)return ; +evsel__get (sample->evsel); }/* Check for raw data in sample */
tools/perf/util/sample.c +12 -5
@@ -1,18 +1,23 @@/* SPDX-License-Identifier: GPL-2.0 */ #include "sample.h" + +#include <stdlib.h> +#include <string.h> + +#include <elf.h> +#include <linux/zalloc.h> + +#include "../../arch/x86/include/asm/insn.h" #include "debug.h" +#include "evsel.h" #include "thread.h" -#include <elf.h> +#ifndef EM_CSKY #define EM_CSKY 252#endif #ifndef EM_LOONGARCH #define EM_LOONGARCH 258#endif -#include <linux/zalloc.h> -#include <stdlib.h> -#include <string.h> -#include "../../arch/x86/include/asm/insn.h" void perf_sample__init (struct perf_sample *sample,bool all) {
@@ -29,6 +34,8 @@ void perf_sample__init(struct perf_sample *sample, bool all)void perf_sample__exit (struct perf_sample *sample) { +evsel__put (sample->evsel); + sample->evsel =NULL ;zfree (&sample->user_regs);zfree (&sample->intr_regs);if (sample->merged_callchain) {
tools/perf/util/session.c +16 -3
@@ -1854,6 +1854,7 @@ static int evlist__deliver_deferred_callchain(struct evlist *evlist,list_for_each_entry_safe (de, tmp,evlist__deferred_samples (evlist), list) {struct perf_sample orig_sample; +struct evsel *new_evsel;perf_sample__init (&orig_sample,/*all=*/ false); ret =evlist__parse_sample (evlist, de->event, &orig_sample);
@@ -1874,7 +1875,11 @@ static int evlist__deliver_deferred_callchain(struct evlist *evlist,else orig_sample.deferred_callchain = false; - orig_sample.evsel =evlist__id2evsel (evlist, orig_sample.id); + new_evsel =evlist__id2evsel (evlist, orig_sample.id); +if (new_evsel != orig_sample.evsel) { +evsel__put (orig_sample.evsel); + orig_sample.evsel =evsel__get (new_evsel); + } ret =evlist__deliver_sample (evlist, tool, de->event, &orig_sample, machine);
@@ -1903,6 +1908,7 @@ static int session__flush_deferred_samples(struct perf_session *session,list_for_each_entry_safe (de, tmp,evlist__deferred_samples (evlist), list) {struct perf_sample sample; +struct evsel *new_evsel;perf_sample__init (&sample,/*all=*/ false); ret =evlist__parse_sample (evlist, de->event, &sample);
@@ -1913,7 +1919,11 @@ static int session__flush_deferred_samples(struct perf_session *session, } sample.file_offset = de->file_offset; - sample.evsel =evlist__id2evsel (evlist, sample.id); + new_evsel =evlist__id2evsel (evlist, sample.id); +if (new_evsel != sample.evsel) { +evsel__put (sample.evsel); + sample.evsel =evsel__get (new_evsel); + } ret =evlist__deliver_sample (evlist, tool, de->event, &sample, machine);
@@ -1959,8 +1969,11 @@ static int machines__deliver_event(struct machines *machines,dump_event (evlist, event, file_offset, sample, file_path); -if (!sample->evsel) +if (!sample->evsel) { sample->evsel =evlist__id2evsel (evlist, sample->id); +if (sample->evsel) + sample->evsel =evsel__get (sample->evsel); + }else assert (sample->evsel ==evlist__id2evsel (evlist, sample->id)); machine =machines__find_for_cpumode (machines, event, sample);
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
Ian Rogers
reviewerPATCH v15 9/19 perf python: Add wrapper for perf_data file abstraction
In reply to Ian RogersThe perf_data struct is needed for session supported.
---
v2:
1. Fixed Memory & FD Leaks in pyrf_data__init : Added cleanup of old state (closing file and freeing path) if __init__ is called multiple times on the same object.
2. Fixed Invalid Free in pyrf_data__delete : Ensured pdata->data.path is always dynamically allocated via strdup() , even for the default "perf.data" . This avoids passing a static string literal to free().
3. Fixed NULL Pointer Dereference in pyrf_data__str : Added a check for NULL path to prevent segfaults if str() or repr() is called on an uninitialized object.
4. Guarded fd Argument Usage: Added a check to ensure that if an fd is provided, it corresponds to a pipe, failing gracefully with a ValueError otherwise.
v7:
- Added pyrf_data__new to zero-initialize pdata->data to fix crash on re-initialization.
--- tools/perf/util/python.c | 123 ++++++++++++++++++++++++++++++++++++++- 1 file changed, 122 insertions(+), 1 deletion(-)
tools/perf/util/python.c +122 -1
@@ -6,6 +6,7 @@#include <linux/err.h> #include <poll.h> +#include <unistd.h> #include <internal/lib.h> #include <perf/cpumap.h>
@@ -14,6 +15,7 @@#include "callchain.h" #include "counts.h" +#include "data.h" #include "event.h" #include "evlist.h" #include "evsel.h"
@@ -2460,6 +2462,121 @@ static PyObject *pyrf__metrics(PyObject *self, PyObject *args)return list; } +struct pyrf_data { +PyObject_HEAD + + struct perf_data data; +}; + +static int pyrf_data__init (struct pyrf_data *pdata,PyObject *args,PyObject *kwargs) +{ +static char *kwlist[] = {"path" ,"fd" ,NULL }; +char *path =NULL ; +int fd =-1 ; + +if (!PyArg_ParseTupleAndKeywords (args, kwargs,"|si" , kwlist, &path, &fd)) +return -1 ; + +if (pdata->data.open) +perf_data__close (&pdata->data); +free ((char *)pdata->data.path); + pdata->data.path =NULL ; + +if (fd !=-1 ) { +struct stat st; + +if (fstat (fd, &st) <0 || !S_ISFIFO (st.st_mode)) { +PyErr_SetString (PyExc_ValueError, +"fd argument is only supported for pipes" ); +return -1 ; + } +if (!path) + path ="-" ; +else if (strcmp (path,"-" ) !=0 ) { +PyErr_SetString (PyExc_ValueError, +"path must be '-' when fd is provided" ); +return -1 ; + } + fd =dup (fd); +if (fd <0 ) { +PyErr_SetFromErrno (PyExc_OSError); +return -1 ; + } + }else if (path &&strcmp (path,"-" ) ==0 ) { + fd =dup (0 ); +if (fd <0 ) { +PyErr_SetFromErrno (PyExc_OSError); +return -1 ; + } + } + +if (!path) + path ="perf.data" ; + + pdata->data.path =strdup (path); +if (!pdata->data.path) { +if (fd !=-1 ) +close (fd); +PyErr_NoMemory (); +return -1 ; + } + + pdata->data.mode =PERF_DATA_MODE_READ ; + pdata->data.file.fd = fd; +if (perf_data__open (&pdata->data) <0 ) { +PyErr_Format (PyExc_IOError,"Failed to open perf data: %s" , + pdata->data.path ? pdata->data.path :"perf.data" ); +return -1 ; + } +return 0 ; +} + +static void pyrf_data__delete (struct pyrf_data *pdata) +{ +perf_data__close (&pdata->data); +free ((char *)pdata->data.path); +Py_TYPE (pdata)->tp_free ((PyObject *)pdata); +} + +static PyObject *pyrf_data__str (PyObject *self) +{ +const struct pyrf_data *pdata = (const struct pyrf_data *)self; + +if (!pdata->data.path) +return PyUnicode_FromString ("[uninitialized]" ); +return PyUnicode_FromString (pdata->data.path); +} + +static PyObject *pyrf_data__new (PyTypeObject *type,PyObject *args,PyObject *kwargs) +{ +struct pyrf_data *pdata; + + pdata = (struct pyrf_data *)PyType_GenericNew (type, args, kwargs); +if (pdata) +memset (&pdata->data,0 ,sizeof (pdata->data)); +return (PyObject *)pdata; +} + +static const char pyrf_data__doc[] =PyDoc_STR ("perf data file object." ); + +static PyTypeObject pyrf_data__type = { +PyVarObject_HEAD_INIT (NULL ,0 ) + .tp_name ="perf.data" , + .tp_basicsize =sizeof (struct pyrf_data ), + .tp_dealloc = (destructor )pyrf_data__delete, + .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, + .tp_doc = pyrf_data__doc, + .tp_init = (initproc )pyrf_data__init, + .tp_repr = pyrf_data__str, + .tp_str = pyrf_data__str, +}; + +static int pyrf_data__setup_types (void ) +{ + pyrf_data__type.tp_new = pyrf_data__new; +return PyType_Ready (&pyrf_data__type); +} +static PyMethodDef perf__methods[] = { { .ml_name ="metrics" ,
@@ -2522,7 +2639,8 @@ PyMODINIT_FUNC PyInit_perf(void)pyrf_cpu_map__setup_types () <0 ||pyrf_pmu_iterator__setup_types () <0 ||pyrf_pmu__setup_types () <0 || - pyrf_counts_values__setup_types() < 0) +pyrf_counts_values__setup_types () <0 || + pyrf_data__setup_types() < 0) return module;/* The page_size is placed in util object. */
@@ -2570,6 +2688,9 @@ PyMODINIT_FUNC PyInit_perf(void)Py_INCREF (&pyrf_counts_values__type);PyModule_AddObject (module,"counts_values" , (PyObject *)&pyrf_counts_values__type); +Py_INCREF (&pyrf_data__type); +PyModule_AddObject (module,"data" , (PyObject *)&pyrf_data__type); + dict =PyModule_GetDict (module);if (dict ==NULL ) goto error;
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
Ian Rogers
reviewerPATCH v15 10/19 perf python: Add python session abstraction wrapping perf's session
In reply to Ian RogersSessions are necessary to be able to use perf.data files within a tool. Add a wrapper python type that incorporates the tool. Allow a sample callback to be passed when creating the session. When process_events is run this callback will be called, if supplied, for sample events.
An example use looks like:
```
$ perf record -e cycles,instructions -a sleep 3
$ PYTHONPATH=..../perf/python python3
Python 3.13.7 (main, Aug 20 2025, 22:17:40) [GCC 14.2.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import perf
>>> count=0
... def handle_sample(x):
... global count
... if count < 3:
... print(dir(x))
... count = count + 1
... perf.session(perf.data("perf.data"),sample=handle_sample).process_events()
...
['__class__', '__delattr__', '__dir__', '__doc__', '__eq__', '__format__', '__ge__', '__getattribute__', '__getstate__', '__gt__', '__hash__', '__init__', '__init_subclass__', '__le__', '__lt__', '__ne__', '__new__', '__reduce__', '__reduce_ex__', '__repr__', '__setattr__', '__sizeof__', '__str__', '__subclasshook__', 'sample_addr', 'sample_cpu', 'sample_id', 'sample_ip', 'sample_period', 'sample_pid', 'sample_stream_id', 'sample_tid', 'sample_time', 'type']
['__class__', '__delattr__', '__dir__', '__doc__', '__eq__', '__format__', '__ge__', '__getattribute__', '__getstate__', '__gt__', '__hash__', '__init__', '__init_subclass__', '__le__', '__lt__', '__ne__', '__new__', '__reduce__', '__reduce_ex__', '__repr__', '__setattr__', '__sizeof__', '__str__', '__subclasshook__', 'sample_addr', 'sample_cpu', 'sample_id', 'sample_ip', 'sample_period', 'sample_pid', 'sample_stream_id', 'sample_tid', 'sample_time', 'type']
['__class__', '__delattr__', '__dir__', '__doc__', '__eq__', '__format__', '__ge__', '__getattribute__', '__getstate__', '__gt__', '__hash__', '__init__', '__init_subclass__', '__le__', '__lt__', '__ne__', '__new__', '__reduce__', '__reduce_ex__', '__repr__', '__setattr__', '__sizeof__', '__str__', '__subclasshook__', 'sample_addr', 'sample_cpu', 'sample_id', 'sample_ip', 'sample_period', 'sample_pid', 'sample_stream_id', 'sample_tid', 'sample_time', 'type']
```Also, add the ability to get the thread associated with a session. For threads, allow the comm string to be retrieved. This can be useful for filtering threads. Connect up some of the standard event handling in psession->tool to better support queries of the machine. Also connect up the symbols.
---
v2:
1. Fixed Potential Crash in pyrf_thread__comm : Used thread__comm_str() to safely retrieve the command name, avoiding a crash if thread__comm() returns NULL.
2. Fixed Double Free Risk: Zeroed out user_regs , intr_regs , and callchain in the shallow copy of perf_sample to prevent Python from attempting to free pointers it doesn't own.
3. Fixed Memory Leak & Exception Handling in Callback: Handled the return value of PyObject_CallFunction() to avoid leaks, and checked for failure to abort the loop and propagate Python exceptions cleanly.
4. Enforced Type Safety: Used O! with &pyrf_data__type in PyArg_ParseTupleAndKeywords to prevent bad casts from passing arbitrary objects as perf.data.
5. Added Missing Build ID Handler: Registered perf_event__process_build_id to allow correct symbol resolution.
6. Fixed Double Free Crash on Init Failure: Set session and pdata to NULL on failure to prevent tp_dealloc from double-freeing them.
7. Preserved C-level Errors: Made pyrf_session__process_events return the error code integer rather than always returning None .
v7:
- Fixed NULL comm handling.
- Avoided swallowing exceptions in module init.
- Fixed checkpatch warning for missing blank line.
v8:
- Switch from pyrf_session__init to pyrf_session__new to avoid dealing with a potentially NULL session variable.
- Added pid, tid, ppid, and cpu attributes to perf.thread.
v10:
- Added CHECK_INITIALIZED checks to pyrf_session methods to prevent NULL pointer dereferences on uninitialized objects.
--- tools/perf/util/python.c | 346 ++++++++++++++++++++++++++++++++++++++- 1 file changed, 340 insertions(+), 6 deletions(-)
tools/perf/util/python.c +340 -6
@@ -14,8 +14,10 @@#include <structmember.h> #include "callchain.h" +#include "comm.h" #include "counts.h" #include "data.h" +#include "debug.h" #include "event.h" #include "evlist.h" #include "evsel.h"
@@ -25,8 +27,14 @@#include "pmus.h" #include "print_binary.h" #include "record.h" +#include "session.h" #include "strbuf.h" +#include "symbol.h" +#include "stat.h" +#include "header.h" +#include "thread.h" #include "thread_map.h" +#include "tool.h" #include "tp_pmu.h" #include "trace-event.h" #include "util/sample.h"
@@ -2577,6 +2585,321 @@ static int pyrf_data__setup_types(void)return PyType_Ready (&pyrf_data__type); } +struct pyrf_thread { +PyObject_HEAD + + struct thread *thread; +}; + +static void pyrf_thread__delete (struct pyrf_thread *pthread) +{ +thread__put (pthread->thread); +Py_TYPE (pthread)->tp_free ((PyObject *)pthread); +} + +static PyObject *pyrf_thread__comm (PyObject *obj) +{ +struct pyrf_thread *pthread = (void *)obj; +const char *str =thread__comm_str (pthread->thread); + +if (!str) + Py_RETURN_NONE; + +return PyUnicode_FromString (str); +} + +static PyMethodDef pyrf_thread__methods[] = { + { + .ml_name ="comm" , + .ml_meth = (PyCFunction )pyrf_thread__comm, + .ml_flags =METH_NOARGS , + .ml_doc =PyDoc_STR ("Comm(and) associated with this thread." ) + }, + { .ml_name =NULL , } +}; + +static PyObject *pyrf_thread__get_pid (struct pyrf_thread *pthread,void *closure__maybe_unused ) +{ +return PyLong_FromLong (thread__pid (pthread->thread)); +} + +static PyObject *pyrf_thread__get_tid (struct pyrf_thread *pthread,void *closure__maybe_unused ) +{ +return PyLong_FromLong (thread__tid (pthread->thread)); +} + +static PyObject *pyrf_thread__get_ppid (struct pyrf_thread *pthread,void *closure__maybe_unused ) +{ +return PyLong_FromLong (thread__ppid (pthread->thread)); +} + +static PyObject *pyrf_thread__get_cpu (struct pyrf_thread *pthread,void *closure__maybe_unused ) +{ +return PyLong_FromLong (thread__cpu (pthread->thread)); +} + +static PyGetSetDef pyrf_thread__getset[] = { + { .name ="pid" , .get = (getter )pyrf_thread__get_pid, .doc ="process ID" }, + { .name ="tid" , .get = (getter )pyrf_thread__get_tid, .doc ="thread ID" }, + { .name ="ppid" , .get = (getter )pyrf_thread__get_ppid, .doc ="parent process ID" }, + { .name ="cpu" , .get = (getter )pyrf_thread__get_cpu, .doc ="cpu number" }, + { .name =NULL } +}; + +static const char pyrf_thread__doc[] =PyDoc_STR ("perf thread object." ); + +static PyTypeObject pyrf_thread__type = { +PyVarObject_HEAD_INIT (NULL ,0 ) + .tp_name ="perf.thread" , + .tp_basicsize =sizeof (struct pyrf_thread ), + .tp_dealloc = (destructor )pyrf_thread__delete, + .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, + .tp_methods = pyrf_thread__methods, + .tp_getset = pyrf_thread__getset, + .tp_doc = pyrf_thread__doc, +}; + +static int pyrf_thread__setup_types (void ) +{ +return PyType_Ready (&pyrf_thread__type); +} + +static PyObject *pyrf_thread__from_thread (struct thread *thread) +{ + struct pyrf_thread *pthread = PyObject_New(struct pyrf_thread, &pyrf_thread__type); + +if (!pthread) +return NULL ; + + pthread->thread =thread__get (thread); +return (PyObject *)pthread; +} + +struct pyrf_session { +PyObject_HEAD + + struct perf_session *session; +struct perf_tool tool; +struct pyrf_data *pdata; +PyObject *sample; +}; + +static int pyrf_session_tool__sample (const struct perf_tool *tool, +union perf_event *event, +struct perf_sample *sample, +struct machine *machine__maybe_unused ) +{ + struct pyrf_session *psession = container_of(tool, struct pyrf_session, tool); +PyObject *pyevent =pyrf_event__new (event); +struct pyrf_event *pevent = (struct pyrf_event *)pyevent; +PyObject *ret; + +if (pyevent ==NULL ) +return -ENOMEM ; + +memcpy (&pevent->event, event, event->header.size); +if (evsel__parse_sample (sample->evsel, &pevent->event, &pevent->sample) <0 ) { +Py_DECREF (pyevent); +return -1 ; + } + + ret =PyObject_CallFunction (psession->sample,"O" , pyevent); +if (!ret) { +Py_DECREF (pyevent); +return -1 ; + } +Py_DECREF (ret); +Py_DECREF (pyevent); +return 0 ; +} + +static PyObject *pyrf_session__find_thread (struct pyrf_session *psession,PyObject *args) +{ +struct machine *machine; +struct thread *thread =NULL ; +PyObject *result; +int pid; + +CHECK_INITIALIZED (psession->session,"session" ); + +if (!PyArg_ParseTuple (args,"i" , &pid)) +return NULL ; + + machine = &psession->session->machines.host; + thread =machine__find_thread (machine, pid, pid); + +if (!thread) { + machine =perf_session__find_machine (psession->session, pid); +if (machine) + thread =machine__find_thread (machine, pid, pid); + } + +if (!thread) { +PyErr_Format (PyExc_TypeError,"Failed to find thread %d" , pid); +return NULL ; + } + result =pyrf_thread__from_thread (thread); +thread__put (thread); +return result; +} + +static PyObject *pyrf_session__new (PyTypeObject *type,PyObject *args,PyObject *kwargs) +{ +struct pyrf_data *pdata; +PyObject *sample =NULL ; +static char *kwlist[] = {"data" ,"sample" ,NULL }; +struct pyrf_session *psession; +struct perf_session *session; + +if (!PyArg_ParseTupleAndKeywords (args, kwargs,"O!|O" , kwlist, &pyrf_data__type, &pdata, + &sample)) +return NULL ; + + psession = PyObject_New(struct pyrf_session, type); +if (!psession) +return NULL ; + + psession->session =NULL ; + psession->sample =NULL ; + psession->pdata =NULL ; + +Py_INCREF (pdata); + psession->pdata = pdata; + +perf_tool__init (&psession->tool,/*ordered_events=*/ true); + psession->tool.ordering_requires_timestamps = true; + +#define ADD_TOOL (name) \ + do { \ + if (name) { \ + if (!PyCallable_Check(name)) { \ + PyErr_SetString(PyExc_TypeError, #name " must be callable"); \ + goto err_out; \ + } \ + psession->tool.name = pyrf_session_tool__##name; \ + Py_INCREF(name); \ + psession->name = name; \ + } \ + } while (0) + +ADD_TOOL (sample); +#undef ADD_TOOL + + psession->tool.comm = perf_event__process_comm; + psession->tool.mmap = perf_event__process_mmap; + psession->tool.mmap2 = perf_event__process_mmap2; + psession->tool.namespaces = perf_event__process_namespaces; + psession->tool.cgroup = perf_event__process_cgroup; + psession->tool.exit = perf_event__process_exit; + psession->tool.fork = perf_event__process_fork; + psession->tool.ksymbol = perf_event__process_ksymbol; + psession->tool.text_poke = perf_event__process_text_poke; + psession->tool.build_id = perf_event__process_build_id; + psession->tool.attr = perf_event__process_attr; + psession->tool.feature = perf_event__process_feature; + psession->tool.stat = perf_event__process_stat_event; + session =perf_session__new (&pdata->data, &psession->tool); +if (IS_ERR (session)) { +PyErr_Format (PyExc_IOError,"failed to create session: %ld" ,PTR_ERR (session)); + goto err_out; + } + psession->session = session; + +if (symbol__init (perf_session__env (session)) <0 ) { +PyErr_SetString (PyExc_OSError,"perf: symbol__init failed" ); + goto err_out; + } + +if (perf_session__create_kernel_maps (session) <0 ) +pr_warning ("Cannot read kernel map\n" ); + +return (PyObject *)psession; +err_out: +Py_DECREF (psession); +return NULL ; +} + +static void pyrf_session__delete (struct pyrf_session *psession) +{ +perf_session__delete (psession->session); +Py_XDECREF (psession->pdata); +Py_XDECREF (psession->sample); +Py_TYPE (psession)->tp_free ((PyObject *)psession); +} + +static PyObject *pyrf_session__find_thread_events (struct pyrf_session *psession) +{ +int err; + +CHECK_INITIALIZED (psession->session,"session" ); + + err =perf_session__process_events (psession->session); + +if (PyErr_Occurred ()) +return NULL ; + +if (err <0 ) { +PyErr_Format (PyExc_OSError,"Process events failed: %d" , err); +return NULL ; + } + + Py_RETURN_NONE; +} + +static PyMethodDef pyrf_session__methods[] = { + { + .ml_name ="process_events" , + .ml_meth = (PyCFunction )pyrf_session__find_thread_events, + .ml_flags =METH_NOARGS , + .ml_doc =PyDoc_STR ("Iterate and process events." ) + }, + { + .ml_name ="find_thread" , + .ml_meth = (PyCFunction )pyrf_session__find_thread, + .ml_flags =METH_VARARGS , + .ml_doc =PyDoc_STR ("Returns the thread associated with a pid." ) + }, + { .ml_name =NULL , } +}; + +static const char pyrf_session__doc[] =PyDoc_STR ("perf session object." ); + +static PyObject *pyrf_session__getattro (struct pyrf_session *psession,PyObject *attr_name) +{ +if (!psession->session) { +PyErr_SetString (PyExc_ValueError,"session not initialized" ); +return NULL ; + } +return PyObject_GenericGetAttr ((PyObject *) psession, attr_name); +} + +static int pyrf_session__setattro (struct pyrf_session *psession,PyObject *attr_name,PyObject *value) +{ +if (!psession->session) { +PyErr_SetString (PyExc_ValueError,"session not initialized" ); +return -1 ; + } +return PyObject_GenericSetAttr ((PyObject *) psession, attr_name, value); +} + +static PyTypeObject pyrf_session__type = { +PyVarObject_HEAD_INIT (NULL ,0 ) + .tp_name ="perf.session" , + .tp_basicsize =sizeof (struct pyrf_session ), + .tp_dealloc = (destructor )pyrf_session__delete, + .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, + .tp_methods = pyrf_session__methods, + .tp_doc = pyrf_session__doc, + .tp_new = pyrf_session__new, + .tp_getattro = (getattrofunc ) pyrf_session__getattro, + .tp_setattro = (setattrofunc ) pyrf_session__setattro, +}; + +static int pyrf_session__setup_types (void ) +{ +return PyType_Ready (&pyrf_session__type); +} +static PyMethodDef perf__methods[] = { { .ml_name ="metrics" ,
@@ -2631,8 +2954,10 @@ PyMODINIT_FUNC PyInit_perf(void) };PyObject *module =PyModule_Create (&moduledef); - if (module == NULL || - pyrf_event__setup_types() < 0 || +if (module ==NULL ) +return NULL ; + + if (pyrf_event__setup_types() < 0 || pyrf_evlist__setup_types() < 0 || pyrf_evsel__setup_types() < 0 || pyrf_thread_map__setup_types() < 0 ||
@@ -2640,8 +2965,12 @@ PyMODINIT_FUNC PyInit_perf(void) pyrf_pmu_iterator__setup_types() < 0 || pyrf_pmu__setup_types() < 0 || pyrf_counts_values__setup_types() < 0 || - pyrf_data__setup_types() < 0) - return module; + pyrf_data__setup_types() < 0 || + pyrf_session__setup_types() < 0 || + pyrf_thread__setup_types() < 0) { +Py_DECREF (module); +return NULL ; + }/* The page_size is placed in util object. */ page_size =sysconf (_SC_PAGE_SIZE);
@@ -2691,6 +3020,9 @@ PyMODINIT_FUNC PyInit_perf(void)Py_INCREF (&pyrf_data__type);PyModule_AddObject (module,"data" , (PyObject *)&pyrf_data__type); +Py_INCREF (&pyrf_session__type); +PyModule_AddObject (module,"session" , (PyObject *)&pyrf_session__type); + dict =PyModule_GetDict (module);if (dict ==NULL ) goto error;
@@ -2704,7 +3036,9 @@ PyMODINIT_FUNC PyInit_perf(void) } error: -if (PyErr_Occurred ()) -PyErr_SetString (PyExc_ImportError,"perf: Init failed!" ); +if (PyErr_Occurred ()) { +Py_XDECREF (module); +return NULL ; + }return module; }
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
Ian Rogers
reviewerPATCH v15 11/19 perf python: Refactor and add accessors to sample event
In reply to Ian RogersAdd common evsel field for events and move sample specific fields to only be present in sample events. Add accessors for sample events. Ensure offsets are within the bounds of the event. Allocate just enough memory for the copied event, don't make the maximum event size each time.
---
v12:
- Used copy_size to bound max_len for NUL termination to fix an out-of-bounds write in pyrf_event__new. Also set pevent->event.header.size = copy_size.
v5:
1. Fix Uninitialized Memory: Restore zero-initialization of pevent->sample in pyrf_event__new() to prevent wild free crashes on error paths.
v6:
- Refactored pyrf_event__new to take evsel and session, and use dynamic allocation based on event size. Updated callers.
v8:
- Ensure events are properly deallocated.
v11:
- Use perf_event__too_small for event size validation, sharing logic with the session parser instead of replicating validation checks.
--- tools/perf/util/python.c | 494 +++++++++++++++++++++++++++++++------- tools/perf/util/session.c | 2 +- tools/perf/util/session.h | 2 + 3 files changed, 413 insertions(+), 85 deletions(-)
tools/perf/util/python.c +410 -84
@@ -13,21 +13,28 @@#include <perf/mmap.h> #include <structmember.h> +#include "addr_location.h" +#include "build-id.h" #include "callchain.h" #include "comm.h" #include "counts.h" #include "data.h" #include "debug.h" +#include "dso.h" #include "event.h" #include "evlist.h" #include "evsel.h" #include "expr.h" +#include "map.h" #include "metricgroup.h" #include "mmap.h" #include "pmus.h" #include "print_binary.h" #include "record.h" +#include "sample.h" #include "session.h" +#include "srccode.h" +#include "srcline.h" #include "strbuf.h" #include "symbol.h" #include "stat.h"
@@ -37,7 +44,6 @@#include "tool.h" #include "tp_pmu.h" #include "trace-event.h" -#include "util/sample.h" #ifdef HAVE_LIBTRACEEVENT #include <event-parse.h>
@@ -45,6 +51,8 @@PyMODINIT_FUNC PyInit_perf (void ); +static PyObject *pyrf_evsel__from_evsel (struct evsel *evsel); +#define member_def (type, member, ptype, help) \ { #member, ptype, \ offsetof(struct pyrf_event, event) + offsetof(struct type, member), \
@@ -73,21 +81,53 @@ PyMODINIT_FUNC PyInit_perf(void);struct pyrf_event {PyObject_HEAD +/** @sample: The parsed sample from the event. */ struct perf_sample sample; -union perf_event event; +/** @al: The address location from machine__resolve, lazily computed. */ +struct addr_location al; +/** @al_resolved: True when machine__resolve been called. */ +bool al_resolved; +/** @event: The underlying perf_event that may be in a file or ring buffer. */ +union perf_event event; };#define sample_members \ - sample_member_def(sample_ip, ip, T_ULONGLONG, "event ip"), \ sample_member_def(sample_pid, pid, T_INT, "event pid"), \ sample_member_def(sample_tid, tid, T_INT, "event tid"), \ sample_member_def(sample_time, time, T_ULONGLONG, "event timestamp"), \ - sample_member_def(sample_addr, addr, T_ULONGLONG, "event addr"), \ sample_member_def(sample_id, id, T_ULONGLONG, "event id"), \ sample_member_def(sample_stream_id, stream_id, T_ULONGLONG, "event stream id"), \ sample_member_def(sample_period, period, T_ULONGLONG, "event period"), \ sample_member_def(sample_cpu, cpu, T_UINT, "event cpu"), +static PyObject *pyrf_event__get_evsel (PyObject *self,void *closure__maybe_unused ) +{ +struct pyrf_event *pevent = (void *)self; + +if (!pevent->sample.evsel) + Py_RETURN_NONE; + +return pyrf_evsel__from_evsel (pevent->sample.evsel); +} + +static PyGetSetDef pyrf_event__getset[] = { + { + .name ="evsel" , + .get = pyrf_event__get_evsel, + .set =NULL , + .doc ="tracking event." , + }, + { .name =NULL , }, +}; + +static void pyrf_event__delete (struct pyrf_event *pevent) +{ +if (pevent->al_resolved) +addr_location__exit (&pevent->al); +perf_sample__exit (&pevent->sample); +Py_TYPE (pevent)->tp_free ((PyObject *)pevent); +} +static const char pyrf_mmap_event__doc[] =PyDoc_STR ("perf mmap event object." );static PyMemberDef pyrf_mmap_event__members[] = {
@@ -126,9 +166,11 @@ static PyTypeObject pyrf_mmap_event__type = {PyVarObject_HEAD_INIT (NULL ,0 ) .tp_name = "perf.mmap_event", .tp_basicsize = sizeof(struct pyrf_event), + .tp_dealloc = (destructor)pyrf_event__delete, .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, .tp_doc = pyrf_mmap_event__doc, .tp_members = pyrf_mmap_event__members, + .tp_getset = pyrf_event__getset, .tp_repr = (reprfunc)pyrf_mmap_event__repr, };
@@ -161,9 +203,11 @@ static PyTypeObject pyrf_task_event__type = {PyVarObject_HEAD_INIT (NULL ,0 ) .tp_name = "perf.task_event", .tp_basicsize = sizeof(struct pyrf_event), + .tp_dealloc = (destructor)pyrf_event__delete, .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, .tp_doc = pyrf_task_event__doc, .tp_members = pyrf_task_event__members, + .tp_getset = pyrf_event__getset, .tp_repr = (reprfunc)pyrf_task_event__repr, };
@@ -190,9 +234,11 @@ static PyTypeObject pyrf_comm_event__type = {PyVarObject_HEAD_INIT (NULL ,0 ) .tp_name = "perf.comm_event", .tp_basicsize = sizeof(struct pyrf_event), + .tp_dealloc = (destructor)pyrf_event__delete, .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, .tp_doc = pyrf_comm_event__doc, .tp_members = pyrf_comm_event__members, + .tp_getset = pyrf_event__getset, .tp_repr = (reprfunc)pyrf_comm_event__repr, };
@@ -222,9 +268,11 @@ static PyTypeObject pyrf_throttle_event__type = {PyVarObject_HEAD_INIT (NULL ,0 ) .tp_name = "perf.throttle_event", .tp_basicsize = sizeof(struct pyrf_event), + .tp_dealloc = (destructor)pyrf_event__delete, .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, .tp_doc = pyrf_throttle_event__doc, .tp_members = pyrf_throttle_event__members, + .tp_getset = pyrf_event__getset, .tp_repr = (reprfunc)pyrf_throttle_event__repr, };
@@ -257,9 +305,11 @@ static PyTypeObject pyrf_lost_event__type = {PyVarObject_HEAD_INIT (NULL ,0 ) .tp_name = "perf.lost_event", .tp_basicsize = sizeof(struct pyrf_event), + .tp_dealloc = (destructor)pyrf_event__delete, .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, .tp_doc = pyrf_lost_event__doc, .tp_members = pyrf_lost_event__members, + .tp_getset = pyrf_event__getset, .tp_repr = (reprfunc)pyrf_lost_event__repr, };
@@ -287,9 +337,11 @@ static PyTypeObject pyrf_read_event__type = {PyVarObject_HEAD_INIT (NULL ,0 ) .tp_name = "perf.read_event", .tp_basicsize = sizeof(struct pyrf_event), + .tp_dealloc = (destructor)pyrf_event__delete, .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, .tp_doc = pyrf_read_event__doc, .tp_members = pyrf_read_event__members, + .tp_getset = pyrf_event__getset, .tp_repr = (reprfunc)pyrf_read_event__repr, };
@@ -297,16 +349,17 @@ static const char pyrf_sample_event__doc[] = PyDoc_STR("perf sample event objectstatic PyMemberDef pyrf_sample_event__members[] = { sample_members +sample_member_def (sample_ip, ip,T_ULONGLONG ,"event ip" ), +sample_member_def (sample_addr, addr,T_ULONGLONG ,"event addr" ), +sample_member_def (sample_phys_addr, phys_addr,T_ULONGLONG ,"event physical addr" ), +sample_member_def (sample_weight, weight,T_ULONGLONG ,"event weight" ), +sample_member_def (sample_data_src, data_src,T_ULONGLONG ,"event data source" ), +sample_member_def (sample_insn_count, insn_cnt,T_ULONGLONG ,"event instruction count" ), +sample_member_def (sample_cyc_count, cyc_cnt,T_ULONGLONG ,"event cycle count" ),member_def (perf_event_header, type,T_UINT ,"event type" ), { .name =NULL , }, }; -static void pyrf_sample_event__delete (struct pyrf_event *pevent) -{ -perf_sample__exit (&pevent->sample); -Py_TYPE (pevent)->tp_free ((PyObject *)pevent); -} -static PyObject *pyrf_sample_event__repr (const struct pyrf_event *pevent) {PyObject *ret;
@@ -324,6 +377,8 @@ static PyObject *pyrf_sample_event__repr(const struct pyrf_event *pevent)#ifdef HAVE_LIBTRACEEVENT static bool is_tracepoint (const struct pyrf_event *pevent) { +if (!pevent->sample.evsel) +return false;return pevent->sample.evsel->core.attr.type ==PERF_TYPE_TRACEPOINT ; }
@@ -394,6 +449,199 @@ get_tracepoint_field(struct pyrf_event *pevent, PyObject *attr_name) }#endif /* HAVE_LIBTRACEEVENT */ +static int pyrf_sample_event__resolve_al (struct pyrf_event *pevent) +{ +struct evsel *evsel = pevent->sample.evsel; +struct evlist *evlist = evsel ? evsel->evlist :NULL ; +struct perf_session *session = evlist ?evlist__session (evlist) :NULL ; + +if (pevent->al_resolved) +return 0 ; + +if (!session) +return -1 ; + +addr_location__init (&pevent->al); +if (machine__resolve (&session->machines.host, &pevent->al, &pevent->sample) <0 ) { +addr_location__exit (&pevent->al); +return -1 ; + } + + pevent->al_resolved = true; +return 0 ; +} + +static PyObject *pyrf_sample_event__get_dso (struct pyrf_event *pevent, +void *closure__maybe_unused ) +{ +if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.map) + Py_RETURN_NONE; + +return PyUnicode_FromString (dso__name (map__dso (pevent->al.map))); +} + +static PyObject *pyrf_sample_event__get_dso_long_name (struct pyrf_event *pevent, +void *closure__maybe_unused ) +{ +if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.map) + Py_RETURN_NONE; + +return PyUnicode_FromString (dso__long_name (map__dso (pevent->al.map))); +} + +static PyObject *pyrf_sample_event__get_dso_bid (struct pyrf_event *pevent, +void *closure__maybe_unused ) +{ +char sbuild_id[SBUILD_ID_SIZE ]; + +if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.map) + Py_RETURN_NONE; + +build_id__snprintf (dso__bid (map__dso (pevent->al.map)), sbuild_id,sizeof (sbuild_id)); +return PyUnicode_FromString (sbuild_id); +} + +static PyObject *pyrf_sample_event__get_map_start (struct pyrf_event *pevent, +void *closure__maybe_unused ) +{ +if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.map) + Py_RETURN_NONE; + +return PyLong_FromUnsignedLong (map__start (pevent->al.map)); +} + +static PyObject *pyrf_sample_event__get_map_end (struct pyrf_event *pevent, +void *closure__maybe_unused ) +{ +if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.map) + Py_RETURN_NONE; + +return PyLong_FromUnsignedLong (map__end (pevent->al.map)); +} + +static PyObject *pyrf_sample_event__get_map_pgoff (struct pyrf_event *pevent, +void *closure__maybe_unused ) +{ +if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.map) + Py_RETURN_NONE; + +return PyLong_FromUnsignedLongLong (map__pgoff (pevent->al.map)); +} + +static PyObject *pyrf_sample_event__get_symbol (struct pyrf_event *pevent, +void *closure__maybe_unused ) +{ +if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.sym) + Py_RETURN_NONE; + +return PyUnicode_FromString (pevent->al.sym->name); +} + +static PyObject *pyrf_sample_event__get_sym_start (struct pyrf_event *pevent, +void *closure__maybe_unused ) +{ +if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.sym) + Py_RETURN_NONE; + +return PyLong_FromUnsignedLongLong (pevent->al.sym->start); +} + +static PyObject *pyrf_sample_event__get_sym_end (struct pyrf_event *pevent, +void *closure__maybe_unused ) +{ +if (pyrf_sample_event__resolve_al (pevent) <0 || !pevent->al.sym) + Py_RETURN_NONE; + +return PyLong_FromUnsignedLongLong (pevent->al.sym->end); +} + +static PyObject *pyrf_sample_event__get_raw_buf (struct pyrf_event *pevent, +void *closure__maybe_unused ) +{ +if (pevent->event.header.type !=PERF_RECORD_SAMPLE ) + Py_RETURN_NONE; + +return PyBytes_FromStringAndSize ((const char *)pevent->sample.raw_data, + pevent->sample.raw_size); +} + +static PyObject *pyrf_sample_event__srccode (PyObject *self,PyObject *args) +{ +struct pyrf_event *pevent = (void *)self; +u64 addr = pevent->sample.ip; +char *srcfile =NULL ; +char *srccode =NULL ; +unsigned int line =0 ; +int len =0 ; +PyObject *result; +struct addr_location al; + +if (!PyArg_ParseTuple (args,"|K" , &addr)) +return NULL ; + +if (pyrf_sample_event__resolve_al (pevent) <0 ) + Py_RETURN_NONE; + +if (addr != pevent->sample.ip) { +addr_location__init (&al); +thread__find_symbol_fb (pevent->al.thread, pevent->sample.cpumode, addr, &al); + }else { +addr_location__init (&al); + al.thread =thread__get (pevent->al.thread); + al.map =map__get (pevent->al.map); + al.sym = pevent->al.sym; + al.addr = pevent->al.addr; + } + +if (al.map) { +struct dso *dso =map__dso (al.map); + +if (dso) { + srcfile =get_srcline_split (dso,map__rip_2objdump (al.map, addr), + &line); + } + } +addr_location__exit (&al); + +if (srcfile) { + srccode =find_sourceline (srcfile, line, &len); + result =Py_BuildValue ("(sIs#)" , srcfile, line, srccode, (Py_ssize_t )len); +free (srcfile); + }else { + result =Py_BuildValue ("(sIs#)" ,NULL ,0 ,NULL , (Py_ssize_t )0 ); + } + +return result; +} + +static PyObject *pyrf_sample_event__insn (PyObject *self,PyObject *args__maybe_unused ) +{ +struct pyrf_event *pevent = (void *)self; +struct thread *thread; +struct machine *machine; + +if (pyrf_sample_event__resolve_al (pevent) <0 ) + Py_RETURN_NONE; + + thread = pevent->al.thread; + +if (!thread || !thread__maps (thread)) + Py_RETURN_NONE; + + machine =maps__machine (thread__maps (thread)); +if (!machine) + Py_RETURN_NONE; + +if (pevent->sample.ip && !pevent->sample.insn_len) +perf_sample__fetch_insn (&pevent->sample, thread, machine); + +if (!pevent->sample.insn_len) + Py_RETURN_NONE; + +return PyBytes_FromStringAndSize ((const char *)pevent->sample.insn, + pevent->sample.insn_len); +} +static PyObject *pyrf_sample_event__getattro (struct pyrf_event *pevent,PyObject *attr_name) {
@@ -407,13 +655,102 @@ pyrf_sample_event__getattro(struct pyrf_event *pevent, PyObject *attr_name)return obj ?:PyObject_GenericGetAttr ((PyObject *) pevent, attr_name); } +static PyGetSetDef pyrf_sample_event__getset[] = { + { + .name ="raw_buf" , + .get = (getter )pyrf_sample_event__get_raw_buf, + .set =NULL , + .doc ="event raw buffer." , + }, + { + .name ="evsel" , + .get = pyrf_event__get_evsel, + .set =NULL , + .doc ="tracking event." , + }, + { + .name ="dso" , + .get = (getter )pyrf_sample_event__get_dso, + .set =NULL , + .doc ="event dso short name." , + }, + { + .name ="dso_long_name" , + .get = (getter )pyrf_sample_event__get_dso_long_name, + .set =NULL , + .doc ="event dso long name." , + }, + { + .name ="dso_bid" , + .get = (getter )pyrf_sample_event__get_dso_bid, + .set =NULL , + .doc ="event dso build id." , + }, + { + .name ="map_start" , + .get = (getter )pyrf_sample_event__get_map_start, + .set =NULL , + .doc ="event map start address." , + }, + { + .name ="map_end" , + .get = (getter )pyrf_sample_event__get_map_end, + .set =NULL , + .doc ="event map end address." , + }, + { + .name ="map_pgoff" , + .get = (getter )pyrf_sample_event__get_map_pgoff, + .set =NULL , + .doc ="event map page offset." , + }, + { + .name ="symbol" , + .get = (getter )pyrf_sample_event__get_symbol, + .set =NULL , + .doc ="event symbol name." , + }, + { + .name ="sym_start" , + .get = (getter )pyrf_sample_event__get_sym_start, + .set =NULL , + .doc ="event symbol start address." , + }, + { + .name ="sym_end" , + .get = (getter )pyrf_sample_event__get_sym_end, + .set =NULL , + .doc ="event symbol end address." , + }, + { .name =NULL , }, +}; + +static PyMethodDef pyrf_sample_event__methods[] = { + { + .ml_name ="srccode" , + .ml_meth = (PyCFunction )pyrf_sample_event__srccode, + .ml_flags =METH_VARARGS , + .ml_doc =PyDoc_STR ("Get source code for an address." ) + }, + { + .ml_name ="insn" , + .ml_meth = (PyCFunction )pyrf_sample_event__insn, + .ml_flags =METH_NOARGS , + .ml_doc =PyDoc_STR ("Get instruction bytes for a sample." ) + }, + { .ml_name =NULL , } +}; +static PyTypeObject pyrf_sample_event__type = {PyVarObject_HEAD_INIT (NULL ,0 ) .tp_name ="perf.sample_event" , .tp_basicsize =sizeof (struct pyrf_event ), + .tp_dealloc = (destructor )pyrf_event__delete, .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, .tp_doc = pyrf_sample_event__doc, .tp_members = pyrf_sample_event__members, + .tp_getset = pyrf_sample_event__getset, + .tp_methods = pyrf_sample_event__methods, .tp_repr = (reprfunc )pyrf_sample_event__repr, .tp_getattro = (getattrofunc ) pyrf_sample_event__getattro, };
@@ -449,25 +786,17 @@ static PyTypeObject pyrf_context_switch_event__type = {PyVarObject_HEAD_INIT (NULL ,0 ) .tp_name = "perf.context_switch_event", .tp_basicsize = sizeof(struct pyrf_event), + .tp_dealloc = (destructor)pyrf_event__delete, .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, .tp_doc = pyrf_context_switch_event__doc, .tp_members = pyrf_context_switch_event__members, + .tp_getset = pyrf_event__getset, .tp_repr = (reprfunc)pyrf_context_switch_event__repr, };static int pyrf_event__setup_types (void ) {int err; - pyrf_mmap_event__type.tp_new = - pyrf_task_event__type.tp_new = - pyrf_comm_event__type.tp_new = - pyrf_lost_event__type.tp_new = - pyrf_read_event__type.tp_new = - pyrf_sample_event__type.tp_new = - pyrf_context_switch_event__type.tp_new = - pyrf_throttle_event__type.tp_new = PyType_GenericNew; - - pyrf_sample_event__type.tp_dealloc = (destructor )pyrf_sample_event__delete, err =PyType_Ready (&pyrf_mmap_event__type);if (err <0 )
@@ -511,33 +840,51 @@ static PyTypeObject *pyrf_event__type[] = { [PERF_RECORD_SWITCH_CPU_WIDE] = &pyrf_context_switch_event__type, }; -static PyObject *pyrf_event__new (const union perf_event *event) +static PyObject *pyrf_event__new (const union perf_event *event,struct evsel *evsel) {struct pyrf_event *pevent; -PyTypeObject *ptype; +int err; +u32 min_size; -if ((event->header.type <PERF_RECORD_MMAP || - event->header.type >PERF_RECORD_SAMPLE ) && - !(event->header.type ==PERF_RECORD_SWITCH || - event->header.type ==PERF_RECORD_SWITCH_CPU_WIDE )) { -PyErr_Format (PyExc_TypeError,"Unexpected header type %u" , +if (event->header.type >=ARRAY_SIZE (pyrf_event__type) || + pyrf_event__type[event->header.type] ==NULL ) { +return PyErr_Format (PyExc_TypeError,"Unexpected header type %u" , event->header.type); -return NULL ; } -// FIXME this better be dynamic or we need to parse everything -// before calling perf_mmap__consume(), including tracepoint fields. -if (sizeof (pevent->event) < event->header.size) { -PyErr_Format (PyExc_TypeError,"Unexpected event size: %zd < %u" , -sizeof (pevent->event), event->header.size); -return NULL ; +if (perf_event__too_small (event, &min_size)) { +return PyErr_Format (PyExc_ValueError,"Event size %u too small for type %u" , + event->header.size, event->header.type); } - ptype = pyrf_event__type[event->header.type]; - pevent = PyObject_New(struct pyrf_event, ptype); -if (pevent !=NULL ) { -memcpy (&pevent->event, event, event->header.size); -perf_sample__init (&pevent->sample,/*all=*/ false); +size_t copy_size = event->header.size; + + pevent = PyObject_New(struct pyrf_event, pyrf_event__type[event->header.type]); +if (pevent ==NULL ) +return PyErr_NoMemory (); + +if (copy_size >sizeof (pevent->event)) + copy_size =sizeof (pevent->event); + +/* Copy the event for memory safety and initialize variables. */ +memcpy (&pevent->event, event, copy_size); + pevent->event.header.size = copy_size; +if (copy_size <sizeof (pevent->event)) +memset ((char *)&pevent->event + copy_size,0 ,sizeof (pevent->event) - copy_size); + +perf_sample__init (&pevent->sample,/*all=*/ true); + pevent->al_resolved = false; +addr_location__init (&pevent->al); + +if (!evsel) +return (PyObject *)pevent; + +/* Parse the sample again so that pointers are within the copied event. */ + err =evsel__parse_sample (evsel, &pevent->event, &pevent->sample); +if (err <0 ) { +Py_DECREF (pevent); +return PyErr_Format (PyExc_OSError, +"perf: can't parse sample, err=%d" , err); }return (PyObject *)pevent; }
@@ -1244,7 +1591,7 @@ static PyObject *pyrf_evsel__str(PyObject *self)if (!evsel)return PyUnicode_FromString ("evsel(uninitialized)" ); -return PyUnicode_FromFormat ("evsel(%s/%s/)" ,evsel__pmu_name (evsel),evsel__name (evsel)); +return PyUnicode_FromFormat ("evsel(%s)" ,evsel__name (evsel)); }static PyMethodDef pyrf_evsel__methods[] = {
@@ -1876,9 +2223,11 @@ static PyObject *pyrf_evlist__read_on_cpu(struct pyrf_evlist *pevlist, {struct evlist *evlist;union perf_event *event; +struct evsel *evsel;int sample_id_all =1 , cpu;static char *kwlist[] = {"cpu" ,"sample_id_all" ,NULL };struct mmap *md; +PyObject *pyevent;int err;CHECK_INITIALIZED (pevlist->evlist,"evlist" );
@@ -1889,44 +2238,31 @@ static PyObject *pyrf_evlist__read_on_cpu(struct pyrf_evlist *pevlist,return NULL ; md =get_md (evlist, cpu); -if (!md) { -PyErr_Format (PyExc_TypeError,"Unknown CPU '%d'" , cpu); -return NULL ; - } +if (!md) +return PyErr_Format (PyExc_TypeError,"Unknown CPU '%d'" , cpu); -if (perf_mmap__read_init (&md->core) <0 ) - goto end; + err =perf_mmap__read_init (&md->core); +if (err <0 ) { +return PyErr_Format (PyExc_OSError, +"perf: error mmap read init, err=%d" , err); + } event =perf_mmap__read_event (&md->core); -if (event !=NULL ) { -PyObject *pyevent =pyrf_event__new (event); -struct pyrf_event *pevent = (struct pyrf_event *)pyevent; -struct evsel *evsel; - -if (pyevent ==NULL ) -return PyErr_NoMemory (); - - evsel =evlist__event2evsel (evlist, event); -if (!evsel) { -Py_DECREF (pyevent); -Py_INCREF (Py_None); -return Py_None; - } +if (event ==NULL ) + Py_RETURN_NONE; + evsel =evlist__event2evsel (evlist, event); +if (!evsel) { +/* Unknown evsel. */ perf_mmap__consume (&md->core); - - err =evsel__parse_sample (evsel, &pevent->event, &pevent->sample); -if (err) { -Py_DECREF (pyevent); -return PyErr_Format (PyExc_OSError, -"perf: can't parse sample, err=%d" , err); - } - -return pyevent; + Py_RETURN_NONE; } -end: -Py_INCREF (Py_None); -return Py_None; + pyevent =pyrf_event__new (event, evsel); +perf_mmap__consume (&md->core); +if (pyevent ==NULL ) +return PyErr_Occurred () ?NULL :PyErr_NoMemory (); + +return pyevent; } static PyObject *pyrf_evlist__open(struct pyrf_evlist *pevlist,
@@ -2138,10 +2474,7 @@ static PyObject *pyrf_evlist__str(PyObject *self)evlist__for_each_entry (pevlist->evlist, pos) {if (!first)strbuf_addch (&sb,',' ); -if (!pos->pmu) -strbuf_addstr (&sb,evsel__name (pos)); -else -strbuf_addf (&sb,"%s/%s/" , pos->pmu->name,evsel__name (pos)); +strbuf_addstr (&sb,evsel__name (pos)); first = false; }strbuf_addstr (&sb,"])" );
@@ -2690,19 +3023,12 @@ static int pyrf_session_tool__sample(const struct perf_tool *tool, struct machine *machine __maybe_unused) { struct pyrf_session *psession = container_of(tool, struct pyrf_session, tool); -PyObject *pyevent =pyrf_event__new (event); -struct pyrf_event *pevent = (struct pyrf_event *)pyevent; +PyObject *pyevent =pyrf_event__new (event, sample->evsel);PyObject *ret;if (pyevent ==NULL )return -ENOMEM ; -memcpy (&pevent->event, event, event->header.size); -if (evsel__parse_sample (sample->evsel, &pevent->event, &pevent->sample) <0 ) { -Py_DECREF (pyevent); -return -1 ; - } - ret =PyObject_CallFunction (psession->sample,"O" , pyevent);if (!ret) {Py_DECREF (pyevent);
tools/perf/util/session.c +1 -1
@@ -2672,7 +2672,7 @@ static const u32 perf_event__min_size[PERF_RECORD_HEADER_MAX] = {* Caller must ensure event->header.type < PERF_RECORD_HEADER_MAX. * If min is non-NULL, stores the required minimum on failure. */ -static bool perf_event__too_small (const union perf_event *event,u32 *min) +bool perf_event__too_small (const union perf_event *event,u32 *min) {u32 min_sz = perf_event__min_size[event->header.type];
tools/perf/util/session.h +2 -0
@@ -122,6 +122,8 @@ void perf_session__delete(struct perf_session *session);void perf_event_header__bswap (struct perf_event_header *hdr); +bool perf_event__too_small (const union perf_event *event,u32 *min); + int perf_session__peek_event(struct perf_session *session, off_t file_offset, void *buf, size_t buf_sz, union perf_event **event_ptr,
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
[email protected]
AI reviewRe: [PATCH v15 11/19] perf python: Refactor and add accessors to sample event
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider: - [High] Silent truncation and data corruption for events larger than `sizeof(union perf_event)`. --
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] Silent truncation and data corruption for events larger than sizeof(union perf_event). --
commit ea78ca85a68b855961249f4ca0aca0cfc6198207
Author: Ian Rogers <[email protected]>
Subject: perf python: Refactor and add accessors to sample event
This commit refactors event sample handling by moving sample-specific fields to sample events, adding accessors, and attempting to allocate just enough memory for the copied event. It also adds bounds checking for event offsets.
Quoted 3 lines
tools/perf/util/python.c +0
-0
[ ... ]
Quoted 19 lines
tools/perf/util/python.c +5
-11@@ -840,51 +840,51 @@ -// FIXME this better be dynamic or we need to parse everything -// before calling perf_mmap__consume(), including tracepoint fields. -if (sizeof (pevent->event) < event->header.size) { -PyErr_Format (PyExc_TypeError,"Unexpected event size: %zd < %u" , -sizeof (pevent->event), event->header.size); -return NULL ; +if (perf_event__too_small (event, &min_size)) { +return PyErr_Format (PyExc_ValueError,"Event size %u too small for type %u" , + event->header.size, event->header.type); } - ptype = pyrf_event__type[event->header.type]; - pevent = PyObject_New(struct pyrf_event, ptype); -if (pevent !=NULL ) { -memcpy (&pevent->event, event, event->header.size); -perf_sample__init (&pevent->sample,/*all=*/ false); +size_t copy_size = event->header.size; +
tools/perf/util/python.c line 856
Quoted 10 lines
tools/perf/util/python.c +10
-0+ pevent = PyObject_New(struct pyrf_event, pyrf_event__type[event->header.type]); +if (pevent ==NULL ) +return PyErr_NoMemory (); + +if (copy_size >sizeof (pevent->event)) + copy_size =sizeof (pevent->event); + +/* Copy the event for memory safety and initialize variables. */ +memcpy (&pevent->event, event, copy_size); + pevent->event.header.size = copy_size;
[Severity: High]
Does this silent truncation cause data corruption for events larger than sizeof(union perf_event)?
The previous code safely raised a TypeError for oversized events. By silently limiting copy_size and overwriting pevent->event.header.size, it appears evsel__parse_sample() might read garbage data.
When evsel__parse_sample() processes the event, it reads backwards from the end of the buffer using the altered event->header.size to locate appended sample IDs. Because the size was truncated, these offsets will point into the middle of the payload instead of the end of the event.
Could this return garbage metadata or trigger an OVERFLOW_CHECK error (-EFAULT) for large PERF_RECORD_SAMPLE events that have extensive callchains or stack dumps?
Additionally, the commit message states "Allocate just enough memory for the copied event", but PyObject_New() appears to use the statically sized sizeof(struct pyrf_event) based on the type array. Is the memory actually being dynamically sized here?
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=11
Ian Rogers
reviewerPATCH v15 12/19 perf python: Add mmap2 event
In reply to Ian RogersIf mmap is handled so should mmap2 events. Add support as a distinct python event type.
---
v12:
- Used copy_size to bound max_len for NUL termination of mmap2 and comm events to fix an out-of-bounds write.
--- tools/perf/util/python.c | 152 ++++++++++++++++++++++++++++++++++++++- 1 file changed, 151 insertions(+), 1 deletion(-)
tools/perf/util/python.c +151 -1
@@ -133,7 +133,7 @@ static const char pyrf_mmap_event__doc[] = PyDoc_STR("perf mmap event object.");static PyMemberDef pyrf_mmap_event__members[] = { sample_membersmember_def (perf_event_header, type,T_UINT ,"event type" ), -member_def (perf_event_header, misc,T_UINT ,"event misc" ), +member_def (perf_event_header, misc,T_USHORT ,"event misc" ),member_def (perf_record_mmap, pid,T_UINT ,"event pid" ),member_def (perf_record_mmap, tid,T_UINT ,"event tid" ),member_def (perf_record_mmap, start,T_ULONGLONG ,"start of the map" ),
@@ -174,6 +174,149 @@ static PyTypeObject pyrf_mmap_event__type = { .tp_repr = (reprfunc)pyrf_mmap_event__repr, }; +static const char pyrf_mmap2_event__doc[] =PyDoc_STR ("perf mmap2 event object." ); + +static PyObject *pyrf_mmap2_event__get_maj (PyObject *self,void *closure__maybe_unused ) +{ +struct pyrf_event *pevent = (void *)self; + +if (pevent->event.header.misc &PERF_RECORD_MISC_MMAP_BUILD_ID ) + Py_RETURN_NONE; + +return PyLong_FromUnsignedLong (pevent->event.mmap2.maj); +} + +static PyObject *pyrf_mmap2_event__get_min (PyObject *self,void *closure__maybe_unused ) +{ +struct pyrf_event *pevent = (void *)self; + +if (pevent->event.header.misc &PERF_RECORD_MISC_MMAP_BUILD_ID ) + Py_RETURN_NONE; + +return PyLong_FromUnsignedLong (pevent->event.mmap2.min); +} + +static PyObject *pyrf_mmap2_event__get_ino (PyObject *self,void *closure__maybe_unused ) +{ +struct pyrf_event *pevent = (void *)self; + +if (pevent->event.header.misc &PERF_RECORD_MISC_MMAP_BUILD_ID ) + Py_RETURN_NONE; + +return PyLong_FromUnsignedLongLong (pevent->event.mmap2.ino); +} + +static PyObject *pyrf_mmap2_event__get_ino_generation (PyObject *self,void *closure__maybe_unused ) +{ +struct pyrf_event *pevent = (void *)self; + +if (pevent->event.header.misc &PERF_RECORD_MISC_MMAP_BUILD_ID ) + Py_RETURN_NONE; + +return PyLong_FromUnsignedLongLong (pevent->event.mmap2.ino_generation); +} + +static PyObject *pyrf_mmap2_event__get_build_id (PyObject *self,void *closure__maybe_unused ) +{ +struct pyrf_event *pevent = (void *)self; + +if (!(pevent->event.header.misc &PERF_RECORD_MISC_MMAP_BUILD_ID )) + Py_RETURN_NONE; + +int size = pevent->event.mmap2.build_id_size; + +if (size >20 ) + size =20 ; + +return PyBytes_FromStringAndSize ((const char *)pevent->event.mmap2.build_id, size); +} + +static PyGetSetDef pyrf_mmap2_event__getset[] = { + { + .name ="evsel" , + .get = pyrf_event__get_evsel, + .set =NULL , + .doc ="tracking event." , + }, + { + .name ="maj" , + .get = pyrf_mmap2_event__get_maj, + .set =NULL , + .doc ="major number." , + }, + { + .name ="min" , + .get = pyrf_mmap2_event__get_min, + .set =NULL , + .doc ="minor number." , + }, + { + .name ="ino" , + .get = pyrf_mmap2_event__get_ino, + .set =NULL , + .doc ="inode number." , + }, + { + .name ="ino_generation" , + .get = pyrf_mmap2_event__get_ino_generation, + .set =NULL , + .doc ="inode generation." , + }, + { + .name ="build_id" , + .get = pyrf_mmap2_event__get_build_id, + .set =NULL , + .doc ="binary build ID." , + }, + { .name =NULL , }, +}; + +static PyMemberDef pyrf_mmap2_event__members[] = { + sample_members +member_def (perf_event_header, type,T_UINT ,"event type" ), +member_def (perf_event_header, misc,T_USHORT ,"event misc" ), +member_def (perf_record_mmap2, pid,T_UINT ,"event pid" ), +member_def (perf_record_mmap2, tid,T_UINT ,"event tid" ), +member_def (perf_record_mmap2, start,T_ULONGLONG ,"start of the map" ), +member_def (perf_record_mmap2, len,T_ULONGLONG ,"map length" ), +member_def (perf_record_mmap2, pgoff,T_ULONGLONG ,"page offset" ), +member_def (perf_record_mmap2, prot,T_UINT ,"protection" ), +member_def (perf_record_mmap2, flags,T_UINT ,"flags" ), +member_def (perf_record_mmap2, filename,T_STRING_INPLACE ,"backing store" ), + { .name =NULL , }, +}; + +static PyObject *pyrf_mmap2_event__repr (const struct pyrf_event *pevent) +{ +PyObject *ret; +char *s; + +if (asprintf (&s,"{ type: mmap2, pid: %u, tid: %u, start: %#" PRI_lx64", " +"length: %#" PRI_lx64", offset: %#" PRI_lx64", " +"flags: %#x, prot: %#x, filename: %s }" , + pevent->event.mmap2.pid, pevent->event.mmap2.tid, + pevent->event.mmap2.start, pevent->event.mmap2.len, + pevent->event.mmap2.pgoff, pevent->event.mmap2.flags, + pevent->event.mmap2.prot, pevent->event.mmap2.filename) <0 ) +return PyErr_NoMemory (); + + ret =PyUnicode_FromString (s); +free (s); +return ret; +} + +static PyTypeObject pyrf_mmap2_event__type = { +PyVarObject_HEAD_INIT (NULL ,0 ) + .tp_name ="perf.mmap2_event" , + .tp_basicsize =sizeof (struct pyrf_event ), + .tp_dealloc = (destructor )pyrf_event__delete, + .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, + .tp_doc = pyrf_mmap2_event__doc, + .tp_members = pyrf_mmap2_event__members, + .tp_getset = pyrf_mmap2_event__getset, + .tp_repr = (reprfunc )pyrf_mmap2_event__repr, +}; +static const char pyrf_task_event__doc[] =PyDoc_STR ("perf task (fork/exit) event object." );static PyMemberDef pyrf_task_event__members[] = {
@@ -799,6 +942,9 @@ static int pyrf_event__setup_types(void)int err; err =PyType_Ready (&pyrf_mmap_event__type); +if (err <0 ) + goto out; + err =PyType_Ready (&pyrf_mmap2_event__type);if (err <0 ) goto out; err =PyType_Ready (&pyrf_lost_event__type);
@@ -828,6 +974,7 @@ static int pyrf_event__setup_types(void)static PyTypeObject *pyrf_event__type[] = { [PERF_RECORD_MMAP ] = &pyrf_mmap_event__type, + [PERF_RECORD_MMAP2 ] = &pyrf_mmap2_event__type, [PERF_RECORD_LOST ] = &pyrf_lost_event__type, [PERF_RECORD_COMM ] = &pyrf_comm_event__type, [PERF_RECORD_EXIT ] = &pyrf_task_event__type,
@@ -3310,6 +3457,9 @@ PyMODINIT_FUNC PyInit_perf(void)Py_INCREF (&pyrf_mmap_event__type);PyModule_AddObject (module,"mmap_event" , (PyObject *)&pyrf_mmap_event__type); +Py_INCREF (&pyrf_mmap2_event__type); +PyModule_AddObject (module,"mmap2_event" , (PyObject *)&pyrf_mmap2_event__type); +Py_INCREF (&pyrf_lost_event__type);PyModule_AddObject (module,"lost_event" , (PyObject *)&pyrf_lost_event__type);
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
[email protected]
AI reviewRe: [PATCH v15 12/19] perf python: Add mmap2 event
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider: - [High] The `filename` field in `mmap2_event` is not guaranteed to be NUL-terminated, leading to a…
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] The filename field in mmap2_event is not guaranteed to be NUL-terminated, leading to a potential heap out-of-bounds read when accessed or formatted in Python. --
commit a4ead9e80a925079e6ee879a3bdb58dc505d685b
Author: Ian Rogers <[email protected]>
Subject: perf python: Add mmap2 event
If mmap is handled so should mmap2 events. Add support as a distinct python event type.
Quoted 4 lines
tools/perf/util/python.c +0
-0
[ ... ]
Quoted 17 lines
tools/perf/util/python.c +17
-0+static PyMemberDef pyrf_mmap2_event__members[] = { + sample_members +member_def (perf_event_header, type,T_UINT ,"event type" ), +member_def (perf_event_header, misc,T_USHORT ,"event misc" ), +member_def (perf_record_mmap2, pid,T_UINT ,"event pid" ), +member_def (perf_record_mmap2, tid,T_UINT ,"event tid" ), +member_def (perf_record_mmap2, start,T_ULONGLONG ,"start of the map" ), +member_def (perf_record_mmap2, len,T_ULONGLONG ,"map length" ), +member_def (perf_record_mmap2, pgoff,T_ULONGLONG ,"page offset" ), +member_def (perf_record_mmap2, prot,T_UINT ,"protection" ), +member_def (perf_record_mmap2, flags,T_UINT ,"flags" ), +member_def (perf_record_mmap2, filename,T_STRING_INPLACE ,"backing store" ), + { .name =NULL , }, +}; + +static PyObject *pyrf_mmap2_event__repr (const struct pyrf_event *pevent) +{
tools/perf/util/python.c static PyTypeObject pyrf_mmap_event__type = { · line 300
Quoted 10 lines
tools/perf/util/python.c +10
-0+PyObject *ret; +char *s; + +if (asprintf (&s,"{ type: mmap2, pid: %u, tid: %u, start: %#" PRI_lx64", " +"length: %#" PRI_lx64", offset: %#" PRI_lx64", " +"flags: %#x, prot: %#x, filename: %s }" , + pevent->event.mmap2.pid, pevent->event.mmap2.tid, + pevent->event.mmap2.start, pevent->event.mmap2.len, + pevent->event.mmap2.pgoff, pevent->event.mmap2.flags, + pevent->event.mmap2.prot, pevent->event.mmap2.filename) <0 )
[Severity: High]
Can this result in a heap out-of-bounds read if the filename lacks a NUL terminator?
Because struct perf_record_mmap2 effectively defines the maximum size of union perf_event, a crafted mmap2 event can completely fill the union with non-NUL characters.
If the input event size equals or exceeds sizeof(union perf_event), the memset() zero-padding logic in pyrf_event__new() is bypassed.
Exposing the potentially non-NUL-terminated filename to Python user space via T_STRING_INPLACE in pyrf_mmap2_event__members (which internally uses strlen()), or formatting it with %s in asprintf() here, could read past the end of the struct pyrf_event heap allocation.
Should there be explicit bounds checking or guaranteed NUL-termination for filename before it is formatted or accessed by Python?
Quoted 6 lines
tools/perf/util/python.c +6
-0+return PyErr_NoMemory (); + + ret =PyUnicode_FromString (s); +free (s); +return ret; +}
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=12
Ian Rogers
reviewerPATCH v15 13/19 perf python: Add callchain support
In reply to Ian RogersImplement pyrf_callchain_node and pyrf_callchain types for lazy iteration over callchain frames. Add callchain property to sample_event.
---
v12:
- Added an optional struct machine * argument to pyrf_event__new defaulting to the host machine if NULL, avoiding regressions for future phases.
v2:
1. Eager Callchain Resolution: Moved the callchain resolution from deferred iteration to eager processing in pyrf_session_tool__sample() . This avoids risks of reading from unmapped memory or following dangling pointers to closed sessions.
2. Cached Callchain: Added a callchain field to struct pyrf_event to store the resolved object.
3. Simplified Access: pyrf_sample_event__get_callchain() now just returns the cached object if available.
4. Avoided Double Free: Handled lazy cleanups properly.
v6:
- Moved callchain resolution from session_tool__sample to pyrf_event__new.
--- tools/perf/util/python.c | 215 ++++++++++++++++++++++++++++++++++++++- 1 file changed, 211 insertions(+), 4 deletions(-)
tools/perf/util/python.c +211 -4
@@ -87,6 +87,8 @@ struct pyrf_event {struct addr_location al;/** @al_resolved: True when machine__resolve been called. */ bool al_resolved; +/** @callchain: Resolved callchain, eagerly computed if requested. */ +PyObject *callchain;/** @event: The underlying perf_event that may be in a file or ring buffer. */ union perf_event event; };
@@ -124,6 +126,7 @@ static void pyrf_event__delete(struct pyrf_event *pevent) {if (pevent->al_resolved)addr_location__exit (&pevent->al); +Py_XDECREF (pevent->callchain);perf_sample__exit (&pevent->sample);Py_TYPE (pevent)->tp_free ((PyObject *)pevent); }
@@ -785,6 +788,144 @@ static PyObject *pyrf_sample_event__insn(PyObject *self, PyObject *args __maybe_ pevent->sample.insn_len); } +struct pyrf_callchain_node { +PyObject_HEAD + u64 ip; +struct map *map; +struct symbol *sym; +}; + +static void pyrf_callchain_node__delete (struct pyrf_callchain_node *pnode) +{ +map__put (pnode->map); +Py_TYPE (pnode)->tp_free ((PyObject *)pnode); +} + +static PyObject *pyrf_callchain_node__get_ip (struct pyrf_callchain_node *pnode, +void *closure__maybe_unused ) +{ +return PyLong_FromUnsignedLongLong (pnode->ip); +} + +static PyObject *pyrf_callchain_node__get_symbol (struct pyrf_callchain_node *pnode, +void *closure__maybe_unused ) +{ +if (pnode->sym) +return PyUnicode_FromString (pnode->sym->name); +return PyUnicode_FromString ("[unknown]" ); +} + +static PyObject *pyrf_callchain_node__get_dso (struct pyrf_callchain_node *pnode, +void *closure__maybe_unused ) +{ +const char *dsoname ="[unknown]" ; + +if (pnode->map) { +struct dso *dso =map__dso (pnode->map); + +if (dso) { +if (symbol_conf.show_kernel_path &&dso__long_name (dso)) + dsoname =dso__long_name (dso); +else + dsoname =dso__name (dso); + } + } +return PyUnicode_FromString (dsoname); +} + +static PyGetSetDef pyrf_callchain_node__getset[] = { + { .name ="ip" , .get = (getter )pyrf_callchain_node__get_ip, }, + { .name ="symbol" , .get = (getter )pyrf_callchain_node__get_symbol, }, + { .name ="dso" , .get = (getter )pyrf_callchain_node__get_dso, }, + { .name =NULL , }, +}; + +static PyTypeObject pyrf_callchain_node__type = { +PyVarObject_HEAD_INIT (NULL ,0 ) + .tp_name ="perf.callchain_node" , + .tp_basicsize =sizeof (struct pyrf_callchain_node ), + .tp_dealloc = (destructor )pyrf_callchain_node__delete, + .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, + .tp_doc ="perf callchain node object." , + .tp_getset = pyrf_callchain_node__getset, +}; + +struct pyrf_callchain_frame { +u64 ip; +struct map *map; +struct symbol *sym; +}; + +struct pyrf_callchain { +PyObject_HEAD + struct pyrf_callchain_frame *frames; +u64 nr_frames; +}; + +static void pyrf_callchain__delete (struct pyrf_callchain *pchain) +{ +if (pchain->frames) { +for (u64 i =0 ; i < pchain->nr_frames; i++) +map__put (pchain->frames[i].map); +free (pchain->frames); + } +Py_TYPE (pchain)->tp_free ((PyObject *)pchain); +} + +static Py_ssize_t pyrf_callchain__length (PyObject *obj) +{ +struct pyrf_callchain *pchain = (void *)obj; + +return pchain->nr_frames; +} + +static PyObject *pyrf_callchain__item (PyObject *obj,Py_ssize_t i) +{ +struct pyrf_callchain *pchain = (void *)obj; +struct pyrf_callchain_node *pnode; + +if (i <0 || i >= (Py_ssize_t )pchain->nr_frames) { +PyErr_SetString (PyExc_IndexError,"Index out of range" ); +return NULL ; + } + + pnode = PyObject_New(struct pyrf_callchain_node, &pyrf_callchain_node__type); +if (!pnode) +return NULL ; + + pnode->ip = pchain->frames[i].ip; + pnode->map =map__get (pchain->frames[i].map); + pnode->sym = pchain->frames[i].sym; + +return (PyObject *)pnode; +} + +static PySequenceMethods pyrf_callchain__sequence_methods = { + .sq_length = pyrf_callchain__length, + .sq_item = pyrf_callchain__item, +}; + +static PyTypeObject pyrf_callchain__type = { +PyVarObject_HEAD_INIT (NULL ,0 ) + .tp_name ="perf.callchain" , + .tp_basicsize =sizeof (struct pyrf_callchain ), + .tp_dealloc = (destructor )pyrf_callchain__delete, + .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, + .tp_doc ="perf callchain object." , + .tp_as_sequence = &pyrf_callchain__sequence_methods, +}; + +static PyObject *pyrf_sample_event__get_callchain (PyObject *self,void *closure__maybe_unused ) +{ +struct pyrf_event *pevent = (void *)self; + +if (!pevent->callchain) + Py_RETURN_NONE; + +Py_INCREF (pevent->callchain); +return pevent->callchain; +} +static PyObject *pyrf_sample_event__getattro (struct pyrf_event *pevent,PyObject *attr_name) {
@@ -799,6 +940,12 @@ pyrf_sample_event__getattro(struct pyrf_event *pevent, PyObject *attr_name) }static PyGetSetDef pyrf_sample_event__getset[] = { + { + .name ="callchain" , + .get = pyrf_sample_event__get_callchain, + .set =NULL , + .doc ="event callchain." , + }, { .name ="raw_buf" , .get = (getter )pyrf_sample_event__get_raw_buf,
@@ -968,6 +1115,12 @@ static int pyrf_event__setup_types(void) err =PyType_Ready (&pyrf_context_switch_event__type);if (err <0 ) goto out; + err =PyType_Ready (&pyrf_callchain_node__type); +if (err <0 ) + goto out; + err =PyType_Ready (&pyrf_callchain__type); +if (err <0 ) + goto out; out:return err; }
@@ -987,12 +1140,18 @@ static PyTypeObject *pyrf_event__type[] = { [PERF_RECORD_SWITCH_CPU_WIDE] = &pyrf_context_switch_event__type, }; -static PyObject *pyrf_event__new (const union perf_event *event,struct evsel *evsel) +static PyObject *pyrf_event__new (const union perf_event *event,struct evsel *evsel, +struct perf_session *session, +struct machine *machine) {struct pyrf_event *pevent; +struct perf_sample *sample;int err;u32 min_size; +if (!machine) + machine = session ? &session->machines.host :NULL ; +if (event->header.type >=ARRAY_SIZE (pyrf_event__type) || pyrf_event__type[event->header.type] ==NULL ) { return PyErr_Format(PyExc_TypeError, "Unexpected header type %u",
@@ -1020,6 +1179,7 @@ static PyObject *pyrf_event__new(const union perf_event *event, struct evsel *evmemset ((char *)&pevent->event + copy_size,0 ,sizeof (pevent->event) - copy_size);perf_sample__init (&pevent->sample,/*all=*/ true); + pevent->callchain =NULL ; pevent->al_resolved = false;addr_location__init (&pevent->al);
@@ -1033,6 +1193,50 @@ static PyObject *pyrf_event__new(const union perf_event *event, struct evsel *evreturn PyErr_Format (PyExc_OSError,"perf: can't parse sample, err=%d" , err); } + sample = &pevent->sample; +if (machine && sample->callchain) { +struct addr_location al; +struct callchain_cursor *cursor; +u64 i; +struct pyrf_callchain *pchain; + +addr_location__init (&al); +if (machine__resolve (machine, &al, sample) >=0 ) { + cursor =get_tls_callchain_cursor (); +if (thread__resolve_callchain (al.thread, cursor, sample, +NULL ,NULL ,PERF_MAX_STACK_DEPTH ) ==0 ) { +callchain_cursor_commit (cursor); + + pchain = PyObject_New(struct pyrf_callchain, &pyrf_callchain__type); +if (!pchain) { +addr_location__exit (&al); +Py_DECREF (pevent); +return NULL ; + } + pchain->nr_frames = cursor->nr; + pchain->frames =calloc (pchain->nr_frames, +sizeof (*pchain->frames)); +if (!pchain->frames) { +Py_DECREF (pchain); +addr_location__exit (&al); +Py_DECREF (pevent); +return PyErr_NoMemory (); + } +struct callchain_cursor_node *node; + +for (i =0 ; i < pchain->nr_frames; i++) { + node =callchain_cursor_current (cursor); + pchain->frames[i].ip = node->ip; + pchain->frames[i].map = +map__get (node->ms.map); + pchain->frames[i].sym = node->ms.sym; +callchain_cursor_advance (cursor); + } + pevent->callchain = (PyObject *)pchain; + } +addr_location__exit (&al); + } + }return (PyObject *)pevent; }
@@ -2404,7 +2608,7 @@ static PyObject *pyrf_evlist__read_on_cpu(struct pyrf_evlist *pevlist,perf_mmap__consume (&md->core); Py_RETURN_NONE; } - pyevent =pyrf_event__new (event, evsel); + pyevent =pyrf_event__new (event, evsel,evlist__session (evlist),/*machine=*/ NULL );perf_mmap__consume (&md->core);if (pyevent ==NULL )return PyErr_Occurred () ?NULL :PyErr_NoMemory ();
@@ -3167,10 +3371,10 @@ struct pyrf_session {static int pyrf_session_tool__sample (const struct perf_tool *tool,union perf_event *event,struct perf_sample *sample, -struct machine *machine__maybe_unused ) +struct machine *machine) { struct pyrf_session *psession = container_of(tool, struct pyrf_session, tool); -PyObject *pyevent =pyrf_event__new (event, sample->evsel); +PyObject *pyevent =pyrf_event__new (event, sample->evsel, psession->session, machine);PyObject *ret;if (pyevent ==NULL )
@@ -3278,6 +3482,9 @@ static PyObject *pyrf_session__new(PyTypeObject *type, PyObject *args, PyObject } psession->session = session; + symbol_conf.use_callchain = true; + symbol_conf.show_kernel_path = true; + symbol_conf.inline_name = false;if (symbol__init (perf_session__env (session)) <0 ) {PyErr_SetString (PyExc_OSError,"perf: symbol__init failed" ); goto err_out;
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
Ian Rogers
reviewerPATCH v15 14/19 perf python: Extend API for stat events in python.c
In reply to Ian RogersAdd stat information to the session. Add call backs for stat events.
---
v12:
- Made psession->tool.stat = perf_event__process_stat_event; conditional on !stat so it doesn't unconditionally overwrite user stat callbacks.
v5:
1. Fix Memory Corruption: Corrected the memory offset for stat_round_type in pyrf_stat_round_event__members by adding the base offset of struct pyrf_event.
2. Fix Memory Leak: Added Py_XDECREF() to free the unused return value of the Python callback in pyrf_session_tool__stat_round().
---
v7:
- Added comprehensive size checks in pyrf_event__new for all event types.
- Fixed line length warning.
v10:
- Added CHECK_INITIALIZED check to pyrf_evsel__get_ids to prevent NULL pointer dereference on uninitialized evsel objects.
--- tools/perf/util/python.c | 188 +++++++++++++++++++++++++++++++++++++-- 1 file changed, 183 insertions(+), 5 deletions(-)
tools/perf/util/python.c +183 -5
@@ -459,6 +459,77 @@ static PyTypeObject pyrf_lost_event__type = { .tp_repr = (reprfunc)pyrf_lost_event__repr, }; +static const char pyrf_stat_event__doc[] =PyDoc_STR ("perf stat event object." ); + +static PyMemberDef pyrf_stat_event__members[] = { + sample_members +member_def (perf_event_header, type,T_UINT ,"event type" ), +member_def (perf_record_stat, id,T_ULONGLONG ,"event id" ), +member_def (perf_record_stat, cpu,T_UINT ,"event cpu" ), +member_def (perf_record_stat, thread,T_UINT ,"event thread" ), +member_def (perf_record_stat, val,T_ULONGLONG ,"counter value" ), +member_def (perf_record_stat, ena,T_ULONGLONG ,"enabled time" ), +member_def (perf_record_stat, run,T_ULONGLONG ,"running time" ), + { .name =NULL , }, +}; + +static PyObject *pyrf_stat_event__repr (const struct pyrf_event *pevent) +{ +return PyUnicode_FromFormat ( +"{ type: stat, id: %llu, cpu: %u, thread: %u, val: %llu, ena: %llu, run: %llu }" , + pevent->event.stat.id, + pevent->event.stat.cpu, + pevent->event.stat.thread, + pevent->event.stat.val, + pevent->event.stat.ena, + pevent->event.stat.run); +} + +static PyTypeObject pyrf_stat_event__type = { +PyVarObject_HEAD_INIT (NULL ,0 ) + .tp_name ="perf.stat_event" , + .tp_basicsize =sizeof (struct pyrf_event ), + .tp_new = PyType_GenericNew, + .tp_dealloc = (destructor )pyrf_event__delete, + .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, + .tp_doc = pyrf_stat_event__doc, + .tp_members = pyrf_stat_event__members, + .tp_getset = pyrf_event__getset, + .tp_repr = (reprfunc )pyrf_stat_event__repr, +}; + +static const char pyrf_stat_round_event__doc[] =PyDoc_STR ("perf stat round event object." ); + +static PyMemberDef pyrf_stat_round_event__members[] = { + sample_members +member_def (perf_event_header, type,T_UINT ,"event type" ), + { .name ="stat_round_type" , .type =T_ULONGLONG , + .offset = offsetof(struct pyrf_event , event) + offsetof(struct perf_record_stat_round , type), + .doc ="round type" }, +member_def (perf_record_stat_round, time,T_ULONGLONG ,"round time" ), + { .name =NULL , }, +}; + +static PyObject *pyrf_stat_round_event__repr (const struct pyrf_event *pevent) +{ +return PyUnicode_FromFormat ("{ type: stat_round, type: %llu, time: %llu }" , + pevent->event.stat_round.type, + pevent->event.stat_round.time); +} + +static PyTypeObject pyrf_stat_round_event__type = { +PyVarObject_HEAD_INIT (NULL ,0 ) + .tp_name ="perf.stat_round_event" , + .tp_basicsize =sizeof (struct pyrf_event ), + .tp_new = PyType_GenericNew, + .tp_dealloc = (destructor )pyrf_event__delete, + .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, + .tp_doc = pyrf_stat_round_event__doc, + .tp_members = pyrf_stat_round_event__members, + .tp_getset = pyrf_event__getset, + .tp_repr = (reprfunc )pyrf_stat_round_event__repr, +}; +static const char pyrf_read_event__doc[] =PyDoc_STR ("perf read event object." );static PyMemberDef pyrf_read_event__members[] = {
@@ -1113,6 +1184,12 @@ static int pyrf_event__setup_types(void)if (err <0 ) goto out; err =PyType_Ready (&pyrf_context_switch_event__type); +if (err <0 ) + goto out; + err =PyType_Ready (&pyrf_stat_event__type); +if (err <0 ) + goto out; + err =PyType_Ready (&pyrf_stat_round_event__type);if (err <0 ) goto out; err =PyType_Ready (&pyrf_callchain_node__type);
@@ -1138,6 +1215,8 @@ static PyTypeObject *pyrf_event__type[] = { [PERF_RECORD_SAMPLE] = &pyrf_sample_event__type, [PERF_RECORD_SWITCH] = &pyrf_context_switch_event__type, [PERF_RECORD_SWITCH_CPU_WIDE] = &pyrf_context_switch_event__type, + [PERF_RECORD_STAT] = &pyrf_stat_event__type, + [PERF_RECORD_STAT_ROUND] = &pyrf_stat_round_event__type, }; static PyObject *pyrf_event__new(const union perf_event *event, struct evsel *evsel,
@@ -2123,7 +2202,45 @@ static PyObject *pyrf_evsel__get_attr_wakeup_events(PyObject *self, void *closurreturn PyLong_FromUnsignedLong (pevsel->evsel->core.attr.wakeup_events); } +static PyObject *pyrf_evsel__get_ids (struct pyrf_evsel *pevsel,void *closure__maybe_unused ) +{ +struct evsel *evsel; +PyObject *list; + +CHECK_INITIALIZED (pevsel->evsel,"evsel" ); + + evsel = pevsel->evsel; + list =PyList_New (0 ); + +if (!list) +return NULL ; + +for (u32 i =0 ; i < evsel->core.ids; i++) { +PyObject *id =PyLong_FromUnsignedLongLong (evsel->core.id[i]); +int ret; + +if (!id) { +Py_DECREF (list); +return NULL ; + } + ret =PyList_Append (list, id); +Py_DECREF (id); +if (ret <0 ) { +Py_DECREF (list); +return NULL ; + } + } + +return list; +} +static PyGetSetDef pyrf_evsel__getset[] = { + { + .name ="ids" , + .get = (getter )pyrf_evsel__get_ids, + .set =NULL , + .doc ="event IDs." , + }, { .name ="tracking" , .get = pyrf_evsel__get_tracking,
@@ -2971,6 +3088,8 @@ static const struct perf_constant perf__constants[] = {PERF_CONST (RECORD_LOST_SAMPLES ),PERF_CONST (RECORD_SWITCH ),PERF_CONST (RECORD_SWITCH_CPU_WIDE ), +PERF_CONST (RECORD_STAT ), +PERF_CONST (RECORD_STAT_ROUND ),PERF_CONST (RECORD_MISC_SWITCH_OUT ), { .name = NULL, },
@@ -3366,6 +3485,7 @@ struct pyrf_session {struct perf_tool tool;struct pyrf_data *pdata;PyObject *sample; +PyObject *stat; }; static int pyrf_session_tool__sample(const struct perf_tool *tool,
@@ -3390,6 +3510,50 @@ static int pyrf_session_tool__sample(const struct perf_tool *tool,return 0 ; } +static int pyrf_session_tool__stat (const struct perf_tool *tool, +struct perf_session *session, +union perf_event *event) +{ + struct pyrf_session *psession = container_of(tool, struct pyrf_session, tool); +struct evsel *evsel =evlist__id2evsel (session->evlist, event->stat.id); +PyObject *pyevent =pyrf_event__new (event, evsel, psession->session,/*machine=*/ NULL ); +const char *name = evsel ?evsel__name (evsel) :"unknown" ; +PyObject *ret; + +if (pyevent ==NULL ) +return -ENOMEM ; + + ret =PyObject_CallFunction (psession->stat,"Oz" , pyevent, name); +if (!ret) { +Py_DECREF (pyevent); +return -1 ; + } +Py_DECREF (ret); +Py_DECREF (pyevent); +return 0 ; +} + +static int pyrf_session_tool__stat_round (const struct perf_tool *tool, +struct perf_session *session__maybe_unused , +union perf_event *event) +{ + struct pyrf_session *psession = container_of(tool, struct pyrf_session, tool); +PyObject *pyevent =pyrf_event__new (event,/*evsel=*/ NULL , psession->session,/*machine=*/ NULL ); +PyObject *ret; + +if (pyevent ==NULL ) +return -ENOMEM ; + + ret =PyObject_CallFunction (psession->stat,"Oz" , pyevent,NULL ); +if (!ret) { +Py_DECREF (pyevent); +return -1 ; + } +Py_DECREF (ret); +Py_DECREF (pyevent); +return 0 ; +} +static PyObject *pyrf_session__find_thread (struct pyrf_session *psession,PyObject *args) {struct machine *machine;
@@ -3423,13 +3587,13 @@ static PyObject *pyrf_session__find_thread(struct pyrf_session *psession, PyObjestatic PyObject *pyrf_session__new (PyTypeObject *type,PyObject *args,PyObject *kwargs) {struct pyrf_data *pdata; -PyObject *sample =NULL ; -static char *kwlist[] = {"data" ,"sample" ,NULL }; +PyObject *sample =NULL , *stat =NULL ; +static char *kwlist[] = {"data" ,"sample" ,"stat" ,NULL };struct pyrf_session *psession;struct perf_session *session; -if (!PyArg_ParseTupleAndKeywords (args, kwargs,"O!|O" , kwlist, &pyrf_data__type, &pdata, - &sample)) +if (!PyArg_ParseTupleAndKeywords (args, kwargs,"O!|OO" , kwlist, &pyrf_data__type, &pdata, + &sample, &stat))return NULL ; psession = PyObject_New(struct pyrf_session, type);
@@ -3438,6 +3602,7 @@ static PyObject *pyrf_session__new(PyTypeObject *type, PyObject *args, PyObject psession->session =NULL ; psession->sample =NULL ; + psession->stat =NULL ; psession->pdata =NULL ;Py_INCREF (pdata);
@@ -3460,8 +3625,13 @@ static PyObject *pyrf_session__new(PyTypeObject *type, PyObject *args, PyObject }while (0 )ADD_TOOL (sample); +ADD_TOOL (stat);#undef ADD_TOOL +if (stat) + psession->tool.stat_round = pyrf_session_tool__stat_round; + + psession->tool.comm = perf_event__process_comm; psession->tool.mmap = perf_event__process_mmap; psession->tool.mmap2 = perf_event__process_mmap2;
@@ -3474,7 +3644,8 @@ static PyObject *pyrf_session__new(PyTypeObject *type, PyObject *args, PyObject psession->tool.build_id = perf_event__process_build_id; psession->tool.attr = perf_event__process_attr; psession->tool.feature = perf_event__process_feature; - psession->tool.stat = perf_event__process_stat_event; +if (!stat) + psession->tool.stat = perf_event__process_stat_event; session =perf_session__new (&pdata->data, &psession->tool);if (IS_ERR (session)) {PyErr_Format (PyExc_IOError,"failed to create session: %ld" ,PTR_ERR (session));
@@ -3504,6 +3675,7 @@ static void pyrf_session__delete(struct pyrf_session *psession)perf_session__delete (psession->session);Py_XDECREF (psession->pdata);Py_XDECREF (psession->sample); +Py_XDECREF (psession->stat);Py_TYPE (psession)->tp_free ((PyObject *)psession); }
@@ -3691,6 +3863,12 @@ PyMODINIT_FUNC PyInit_perf(void)Py_INCREF (&pyrf_context_switch_event__type);PyModule_AddObject (module,"switch_event" , (PyObject *)&pyrf_context_switch_event__type); +Py_INCREF (&pyrf_stat_event__type); +PyModule_AddObject (module,"stat_event" , (PyObject *)&pyrf_stat_event__type); + +Py_INCREF (&pyrf_stat_round_event__type); +PyModule_AddObject (module,"stat_round_event" , (PyObject *)&pyrf_stat_round_event__type); +Py_INCREF (&pyrf_thread_map__type);PyModule_AddObject (module,"thread_map" , (PyObject *)&pyrf_thread_map__type);
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
[email protected]
AI reviewRe: [PATCH v15 14/19] perf python: Extend API for stat events in python.c
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider: - [High] Calling `perf_session__process_events` from Python on a `perf.data` file containing…
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] Calling perf_session__process_events from Python on a perf.data file containing PERF_RECORD_STAT events without explicitly providing a stat callback will cause a segmentation fault. --
commit 77ed782811229e6257147595967331dc103b7a95
Author: Ian Rogers <[email protected]>
perf python: Extend API for stat events in python.c
Add stat information to the session. Add call backs for stat events.
Quoted 4 lines
tools/perf/util/python.c +0
-0
[ ... ]
Quoted 15 lines
tools/perf/util/python.c +5
-0@@ -3460,8 +3625,13 @@ static PyObject *pyrf_session__new(PyTypeObject *type, PyObject *args, PyObject }while (0 )ADD_TOOL (sample); +ADD_TOOL (stat);#undef ADD_TOOL +if (stat) + psession->tool.stat_round = pyrf_session_tool__stat_round; + + psession->tool.comm = perf_event__process_comm; psession->tool.mmap = perf_event__process_mmap; psession->tool.mmap2 = perf_event__process_mmap2;
tools/perf/util/python.c static PyObject *pyrf_session__new(PyTypeObject *type, PyObject *args, PyObject · line 3648
Quoted 6 lines
tools/perf/util/python.c +2
-1@@ -3474,7 +3644,8 @@ static PyObject *pyrf_session__new(PyTypeObject *type, PyObject *args, PyObject psession->tool.build_id = perf_event__process_build_id; psession->tool.attr = perf_event__process_attr; psession->tool.feature = perf_event__process_feature; - psession->tool.stat = perf_event__process_stat_event; +if (!stat) + psession->tool.stat = perf_event__process_stat_event;
[Severity: High]
Does this code cause a segmentation fault when a stat callback is not provided?
In tools/perf/util/python.c:pyrf_session__new(), when no stat callback is passed to perf.session(), the code defaults to using perf_event__process_stat_event() as the callback.
However, the Python bindings never appear to call evlist__alloc_stats() to allocate memory for the counter values. When a PERF_RECORD_STAT event is encountered, it triggers this path:
perf_session__process_events()
perf_event__process_stat_event()
perf_counts(counter->counts, ...)
Since counter->counts is unallocated, perf_counts() evaluates to &(NULL->values[...]), causing a null pointer dereference.
Can this regression be triggered by simply calling perf_session__process_events() on a perf.data file containing PERF_RECORD_STAT events (like those generated by perf stat record) without providing a stat callback?
Quoted 3 lines
tools/perf/util/python.c +0
-0session =perf_session__new (&pdata->data, &psession->tool);if (IS_ERR (session)) {PyErr_Format (PyExc_IOError,"failed to create session: %ld" ,PTR_ERR (session));
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=14
Ian Rogers
reviewerPATCH v15 15/19 perf python: Expose brstack in sample event
In reply to Ian RogersImplement pyrf_branch_entry and pyrf_branch_stack for lazy iteration over branch stack entries.
---
v2:
1. Avoided Keyword Collision: Renamed the properties "from" and "to" to "from_ip" and "to_ip" in pyrf_branch_entry__getset[] to prevent syntax errors in Python.
2. Eager Branch Stack Resolution: Updated pyrf_session_tool__sample() to allocate and copy the branch stack entries eagerly when the event is processed, instead of deferring it to iteration time. This avoids reading from potentially overwritten or unmapped mmap buffers.
3. Updated Iterators: Updated pyrf_branch_stack and pyrf_branch_stack__next() to use the copied entries rather than pointing directly to the sample's buffer.
4. Avoided Reference Leak on Init Failure: Added proper error checking for PyModule_AddObject() in the module initialization function, decrementing references on failure.
v6:
- Moved branch stack resolution from session_tool__sample to pyrf_event__new.
--- tools/perf/util/python.c | 198 +++++++++++++++++++++++++++++++++++++++ 1 file changed, 198 insertions(+)
tools/perf/util/python.c +198 -0
@@ -22,6 +22,7 @@#include "debug.h" #include "dso.h" #include "event.h" +#include "branch.h" #include "evlist.h" #include "evsel.h" #include "expr.h"
@@ -89,6 +90,8 @@ struct pyrf_event {bool al_resolved;/** @callchain: Resolved callchain, eagerly computed if requested. */ PyObject *callchain; +/** @brstack: Resolved branch stack, eagerly computed if requested. */ +PyObject *brstack;/** @event: The underlying perf_event that may be in a file or ring buffer. */ union perf_event event; };
@@ -127,6 +130,7 @@ static void pyrf_event__delete(struct pyrf_event *pevent)if (pevent->al_resolved)addr_location__exit (&pevent->al);Py_XDECREF (pevent->callchain); +Py_XDECREF (pevent->brstack);perf_sample__exit (&pevent->sample);Py_TYPE (pevent)->tp_free ((PyObject *)pevent); }
@@ -997,6 +1001,154 @@ static PyObject *pyrf_sample_event__get_callchain(PyObject *self, void *closurereturn pevent->callchain; } +struct pyrf_branch_entry { +PyObject_HEAD + u64 from; +u64 to; +struct branch_flags flags; +}; + +static void pyrf_branch_entry__delete (struct pyrf_branch_entry *pentry) +{ +Py_TYPE (pentry)->tp_free ((PyObject *)pentry); +} + +static PyObject *pyrf_branch_entry__get_from (struct pyrf_branch_entry *pentry, +void *closure__maybe_unused ) +{ +return PyLong_FromUnsignedLongLong (pentry->from); +} + +static PyObject *pyrf_branch_entry__get_to (struct pyrf_branch_entry *pentry, +void *closure__maybe_unused ) +{ +return PyLong_FromUnsignedLongLong (pentry->to); +} + +static PyObject *pyrf_branch_entry__get_mispred (struct pyrf_branch_entry *pentry, +void *closure__maybe_unused ) +{ +return PyBool_FromLong (pentry->flags.mispred); +} + +static PyObject *pyrf_branch_entry__get_predicted (struct pyrf_branch_entry *pentry, +void *closure__maybe_unused ) +{ +return PyBool_FromLong (pentry->flags.predicted); +} + +static PyObject *pyrf_branch_entry__get_in_tx (struct pyrf_branch_entry *pentry, +void *closure__maybe_unused ) +{ +return PyBool_FromLong (pentry->flags.in_tx); +} + +static PyObject *pyrf_branch_entry__get_abort (struct pyrf_branch_entry *pentry, +void *closure__maybe_unused ) +{ +return PyBool_FromLong (pentry->flags.abort); +} + +static PyObject *pyrf_branch_entry__get_cycles (struct pyrf_branch_entry *pentry, +void *closure__maybe_unused ) +{ +return PyLong_FromUnsignedLongLong (pentry->flags.cycles); +} + +static PyObject *pyrf_branch_entry__get_type (struct pyrf_branch_entry *pentry, +void *closure__maybe_unused ) +{ +return PyLong_FromUnsignedLongLong ((unsigned long long )pentry->flags.type); +} + +static PyGetSetDef pyrf_branch_entry__getset[] = { + { .name ="from_ip" , .get = (getter )pyrf_branch_entry__get_from, }, + { .name ="to_ip" , .get = (getter )pyrf_branch_entry__get_to, }, + { .name ="mispred" , .get = (getter )pyrf_branch_entry__get_mispred, }, + { .name ="predicted" , .get = (getter )pyrf_branch_entry__get_predicted, }, + { .name ="in_tx" , .get = (getter )pyrf_branch_entry__get_in_tx, }, + { .name ="abort" , .get = (getter )pyrf_branch_entry__get_abort, }, + { .name ="cycles" , .get = (getter )pyrf_branch_entry__get_cycles, }, + { .name ="type" , .get = (getter )pyrf_branch_entry__get_type, }, + { .name =NULL , }, +}; + +static PyTypeObject pyrf_branch_entry__type = { +PyVarObject_HEAD_INIT (NULL ,0 ) + .tp_name ="perf.branch_entry" , + .tp_basicsize =sizeof (struct pyrf_branch_entry ), + .tp_dealloc = (destructor )pyrf_branch_entry__delete, + .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, + .tp_doc ="perf branch entry object." , + .tp_getset = pyrf_branch_entry__getset, +}; + +struct pyrf_branch_stack { +PyObject_HEAD + struct branch_entry *entries; +u64 nr; +}; + +static void pyrf_branch_stack__delete (struct pyrf_branch_stack *pstack) +{ +free (pstack->entries); +Py_TYPE (pstack)->tp_free ((PyObject *)pstack); +} + +static Py_ssize_t pyrf_branch_stack__length (PyObject *obj) +{ +struct pyrf_branch_stack *pstack = (void *)obj; + +return pstack->nr; +} + +static PyObject *pyrf_branch_stack__item (PyObject *obj,Py_ssize_t i) +{ +struct pyrf_branch_stack *pstack = (void *)obj; +struct pyrf_branch_entry *pentry; + +if (i <0 || i >= (Py_ssize_t )pstack->nr) { +PyErr_SetString (PyExc_IndexError,"Index out of range" ); +return NULL ; + } + + pentry = PyObject_New(struct pyrf_branch_entry, &pyrf_branch_entry__type); +if (!pentry) +return NULL ; + + pentry->from = pstack->entries[i].from; + pentry->to = pstack->entries[i].to; + pentry->flags = pstack->entries[i].flags; + +return (PyObject *)pentry; +} + +static PySequenceMethods pyrf_branch_stack__sequence_methods = { + .sq_length = pyrf_branch_stack__length, + .sq_item = pyrf_branch_stack__item, +}; + +static PyTypeObject pyrf_branch_stack__type = { +PyVarObject_HEAD_INIT (NULL ,0 ) + .tp_name ="perf.branch_stack" , + .tp_basicsize =sizeof (struct pyrf_branch_stack ), + .tp_dealloc = (destructor )pyrf_branch_stack__delete, + .tp_flags = Py_TPFLAGS_DEFAULT|Py_TPFLAGS_BASETYPE, + .tp_doc ="perf branch stack object." , + .tp_as_sequence = &pyrf_branch_stack__sequence_methods, +}; + +static PyObject *pyrf_sample_event__get_brstack (PyObject *self,void *closure__maybe_unused ) +{ +struct pyrf_event *pevent = (void *)self; + +if (!pevent->brstack) + Py_RETURN_NONE; + +Py_INCREF (pevent->brstack); +return pevent->brstack; +} +static PyObject *pyrf_sample_event__getattro (struct pyrf_event *pevent,PyObject *attr_name) {
@@ -1017,6 +1169,12 @@ static PyGetSetDef pyrf_sample_event__getset[] = { .set = NULL, .doc = "event callchain.", }, + { + .name = "brstack", + .get = pyrf_sample_event__get_brstack, + .set = NULL, + .doc = "event branch stack.", + }, { .name = "raw_buf", .get = (getter)pyrf_sample_event__get_raw_buf,
@@ -1198,6 +1356,12 @@ static int pyrf_event__setup_types(void) err =PyType_Ready (&pyrf_callchain__type);if (err <0 ) goto out; + err =PyType_Ready (&pyrf_branch_entry__type); +if (err <0 ) + goto out; + err =PyType_Ready (&pyrf_branch_stack__type); +if (err <0 ) + goto out; out:return err; }
@@ -1259,6 +1423,7 @@ static PyObject *pyrf_event__new(const union perf_event *event, struct evsel *evperf_sample__init (&pevent->sample,/*all=*/ true); pevent->callchain =NULL ; + pevent->brstack =NULL ; pevent->al_resolved = false;addr_location__init (&pevent->al);
@@ -1316,6 +1481,27 @@ static PyObject *pyrf_event__new(const union perf_event *event, struct evsel *evaddr_location__exit (&al); } } +if (sample->branch_stack) { +struct branch_stack *bs = sample->branch_stack; +struct branch_entry *entries =perf_sample__branch_entries (sample); +struct pyrf_branch_stack *pstack; + + pstack = PyObject_New(struct pyrf_branch_stack, &pyrf_branch_stack__type); +if (!pstack) { +Py_DECREF (pevent); +return NULL ; + } + pstack->nr = bs->nr; + pstack->entries =calloc (bs->nr,sizeof (struct branch_entry )); +if (!pstack->entries) { +Py_DECREF (pstack); +Py_DECREF (pevent); +return PyErr_NoMemory (); + } +memcpy (pstack->entries, entries, + bs->nr *sizeof (struct branch_entry )); + pevent->brstack = (PyObject *)pstack; + }return (PyObject *)pevent; }
@@ -3884,6 +4070,18 @@ PyMODINIT_FUNC PyInit_perf(void)Py_INCREF (&pyrf_session__type);PyModule_AddObject (module,"session" , (PyObject *)&pyrf_session__type); +Py_INCREF (&pyrf_branch_entry__type); +if (PyModule_AddObject (module,"branch_entry" , (PyObject *)&pyrf_branch_entry__type) <0 ) { +Py_DECREF (&pyrf_branch_entry__type); + goto error; + } + +Py_INCREF (&pyrf_branch_stack__type); +if (PyModule_AddObject (module,"branch_stack" , (PyObject *)&pyrf_branch_stack__type) <0 ) { +Py_DECREF (&pyrf_branch_stack__type); + goto error; + } + dict =PyModule_GetDict (module);if (dict ==NULL ) goto error;
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
Ian Rogers
reviewerPATCH v15 16/19 perf python: Add syscall name/id to convert syscall number and name
In reply to Ian RogersUse perf's syscalltbl support to convert syscall number to name assuming the number is for the host machine. This avoids python libaudit support as tools/perf/scripts/python/syscall-counts.py requires.
---
v2:
1. Guarded with HAVE_LIBTRACEEVENT : Wrapped the syscall functions and their entries in perf__methods with #ifdef HAVE_LIBTRACEEVENT to avoid potential linker errors if CONFIG_TRACE is disabled.
2. Changed Exception Type: Updated pyrf__syscall_id to raise a ValueError instead of a TypeError when a syscall is not found.
3. Fixed Typo: Corrected "name number" to "name" in the docstring for syscall_id.
v6:
- Added optional keyword-only elf_machine argument to syscall_name and syscall_id.
v7:
- Made syscalltbl.o unconditional in Build to fix undefined symbol when building without libtraceevent.
--- tools/perf/util/Build | 1 - tools/perf/util/python.c | 48 ++++++++++++++++++++++++++++++++++++++++ 2 files changed, 48 insertions(+), 1 deletion(-)
tools/perf/util/Build +0 -1
@@ -75,7 +75,6 @@ perf-util-y += sample.o perf-util-y += sample-raw.o perf-util-y += s390-sample-raw.o perf-util-y += amd-sample-raw.o - perf-util-y += ordered-events.o perf-util-y += namespaces.o perf-util-y += comm.o
tools/perf/util/python.c +48 -0
72 lines not on this page · raw message
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
[email protected]
AI reviewRe: [PATCH v15 16/19] perf python: Add syscall name/id to convert syscall number and name
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider: - [Medium] The newly added `perf.syscall_name()` and `perf.syscall_id()` Python API bindings are never…
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] The newly added perf.syscall_name() and perf.syscall_id() Python API bindings are never used, failing the commit's stated goal of replacing libaudit in syscall-counts.py. --
commit 7f5421e7629b06e56ae07bdbadd1270f016d2cc8
Author: Ian Rogers <[email protected]>
perf python: Add syscall name/id to convert syscall number and name
This commit adds Python API bindings for syscall_name and syscall_id to the perf python module. It leverages the internal syscalltbl support to perform these conversions. The stated goal is to avoid the libaudit dependency in the syscall-counts.py script.
Quoted 4 lines
Use perf's syscalltbl support to convert syscall number to name assuming the number is for the host machine. This avoids python libaudit support as
tools/perf/scripts/python/syscall-counts.pyrequires.
[Severity: Medium]
Is this description complete, or is there a missing patch to update the scripts?
The commit message states this change avoids the libaudit dependency in syscall-counts.py, but neither syscall-counts.py nor Perf-Trace-Util/lib/Perf/Trace/Util.py are modified in this commit to use the new perf.syscall_name API.
Looking at tools/perf/scripts/python/Perf-Trace-Util/lib/Perf/Trace/Util.py, the syscall_name() function still relies on the audit module:
def syscall_name(id):
try:
return audit.audit_syscall_to_name(id, machine_id)
Should these scripts be updated to actually use the new Python bindings so that the dependency is removed, or is that intended for a follow-up patch?
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=16
Ian Rogers
reviewerPATCH v15 17/19 perf python: Add config file access
In reply to Ian RogersAdd perf.config_get(name) to expose the perf configuration system.
--- tools/perf/util/python.c | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+)
tools/perf/util/python.c +20 -0
@@ -17,6 +17,7 @@#include "build-id.h" #include "callchain.h" #include "comm.h" +#include "config.h" #include "counts.h" #include "data.h" #include "debug.h"
@@ -3974,7 +3975,26 @@ static PyObject *pyrf__syscall_id(PyObject *self, PyObject *args, PyObject *kwarreturn PyLong_FromLong (id); } +static PyObject *pyrf__config_get (PyObject *self,PyObject *args) +{ +const char *config_name, *val; + +if (!PyArg_ParseTuple (args,"s" , &config_name)) +return NULL ; + + val =perf_config_get (config_name); +if (!val) + Py_RETURN_NONE; +return PyUnicode_FromString (val); +} +static PyMethodDef perf__methods[] = { + { + .ml_name ="config_get" , + .ml_meth = (PyCFunction ) pyrf__config_get, + .ml_flags =METH_VARARGS , + .ml_doc =PyDoc_STR ("Get a perf config value." ) + }, { .ml_name ="metrics" , .ml_meth = (PyCFunction ) pyrf__metrics,
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
Ian Rogers
reviewerPATCH v15 18/19 perf python: Add perf.pyi stubs file
In reply to Ian RogersAdd Python type stubs for the perf module to improve IDE support and static analysis. Includes docstrings for classes, methods, and constants derived from C source and JSON definitions.
---
v12:
- Audited all perf.pyi event stubs: added full sample_members attributes (pid, tid, time, id, stream_id, period, cpu) to all payload events via inheritance. Fixed sample_event unique fields (ip, addr, phys_addr, weight, data_src, insn_cnt, cyc_cnt). Renamed mmap_event.addr to start. Removed read_event.value.
v2:
1. Added Missing Module Functions: Added parse_metrics and pmus. Renamed metrics to parse_metrics to match python.c .
2. Added Constructors: Added __init__ methods for data , evsel , and evlist with their appropriate arguments.
3. Removed sample_comm : Removed it from sample_event since it is not exported in python.c .
4. Keyword Handling in branch_entry : I used from_ip and to_ip in the stubs to match the rename I did in python.c (in turn 145) to avoid the Python from keyword conflict.
5. Added Missing Event Classes: Added mmap_event , lost_event , comm_event , task_event , throttle_event , read_event , and switch_event .
6. Added Missing evlist Methods: Added get_pollfd and add .
7. Updated Return Types: Changed process_events to return int .
v6:
- Updated perf.pyi to use find_thread and elf_machine.
v8:
- Added mmap2_event class and new evsel attributes to perf.pyi.
- Added pid, tid, ppid, cpu attributes to class thread in perf.pyi.
--- tools/perf/Makefile.perf | 9 +- tools/perf/python/perf.pyi | 654 +++++++++++++++++++++++++++++++++++++ tools/perf/util/python.c | 2 + tools/perf/util/setup.py | 5 + 4 files changed, 667 insertions(+), 3 deletions(-) create mode 100644 tools/perf/python/perf.pyi
tools/perf/Makefile.perf +6 -3
@@ -399,7 +399,7 @@ PYTHON_EXTBUILD_LIB := $(PYTHON_EXTBUILD)lib/ PYTHON_EXTBUILD_TMP := $(PYTHON_EXTBUILD)tmp/ export PYTHON_EXTBUILD_LIB PYTHON_EXTBUILD_TMP -python-clean := $(call QUIET_CLEAN, python) $(RM) -r $(PYTHON_EXTBUILD) $(OUTPUT)python/perf*.so +python-clean := $(call QUIET_CLEAN, python) $(RM) -r $(PYTHON_EXTBUILD) $(OUTPUT)python/perf*.so $(OUTPUT)python/perf.pyi ifneq ($(quiet),) python_setup_quiet=--quiet
@@ -517,12 +517,15 @@ all: shell_compatibility_test $(ALL_PROGRAMS) $(LANG_BINDINGS) $(OTHER_PROGRAMS) # Create python binding output directory if not already present $(shell [ -d '$(OUTPUT)python' ] || mkdir -p '$(OUTPUT)python') -$(OUTPUT)python/perf$(PYTHON_EXTENSION_SUFFIX): util/python.c util/setup.py $(PERFLIBS_PY) +$(OUTPUT)python/perf$(PYTHON_EXTENSION_SUFFIX): util/python.c util/setup.py python/perf.pyi $(PERFLIBS_PY) $(QUIET_GEN)LDSHARED="$(CC) -pthread -shared" \ CFLAGS='$(CFLAGS)' LDFLAGS='$(LDFLAGS) $(LIBS_PY)' \ $(PYTHON_WORD) util/setup.py \ $(python_setup_quiet) build_ext; \ - cp $(PYTHON_EXTBUILD_LIB)perf*.so $(OUTPUT)python/ + cp $(PYTHON_EXTBUILD_LIB)perf*.so $(OUTPUT)python/; \ + if [ "$(abspath python/perf.pyi)" != "$(abspath $(OUTPUT)python/perf.pyi)" ]; then \ + cp python/perf.pyi $(OUTPUT)python/; \ + fi python_perf_target: @echo "Target is: $(OUTPUT)python/perf$(PYTHON_EXTENSION_SUFFIX)"
tools/perf/python/perf.pyi new file +654 -0
654 lines not on this page · raw message
tools/perf/util/python.c +2 -0
@@ -433,6 +433,7 @@ static const char pyrf_lost_event__doc[] = PyDoc_STR("perf lost event object.");static PyMemberDef pyrf_lost_event__members[] = { sample_members +member_def (perf_event_header, type,T_UINT ,"event type" ),member_def (perf_record_lost, id,T_ULONGLONG ,"event id" ),member_def (perf_record_lost, lost,T_ULONGLONG ,"number of lost events" ), { .name =NULL , },
@@ -541,6 +542,7 @@ static const char pyrf_read_event__doc[] = PyDoc_STR("perf read event object.");static PyMemberDef pyrf_read_event__members[] = { sample_members +member_def (perf_event_header, type,T_UINT ,"event type" ),member_def (perf_record_read, pid,T_UINT ,"event pid" ),member_def (perf_record_read, tid,T_UINT ,"event tid" ), { .name =NULL , },
tools/perf/util/setup.py +5 -0
@@ -2,6 +2,7 @@ from os import getenv, path from subprocess import Popen, PIPE from re import sub import shlex +import shutil cc = getenv("CC") assert cc, "Environment variable CC not set"
@@ -73,6 +74,10 @@ class install_lib(_install_lib): _install_lib.finalize_options(self) self.build_dir = build_lib + def run(self): + _install_lib.run(self) + shutil.copy2(f'{src_perf}/python/perf.pyi', self.install_dir) + # switch off several checks (need to be at the end of cflags list) extra_cflags = ['-fno-strict-aliasing', '-Wno-write-strings', '-Wno-unused-parameter', '-Wno-redundant-decls' ]
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
[email protected]
AI reviewRe: [PATCH v15 18/19] perf python: Add perf.pyi stubs file
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider: - [Medium] Type stubs incorrectly document `None` as a valid argument, leading to memory corruption and…
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] Type stubs incorrectly document None as a valid argument, leading to memory corruption and segfaults in the C extension. --
commit 9f279812734ffe49e65776760974bed198e67c6d
Author: Ian Rogers <[email protected]>
perf python: Add perf.pyi stubs file
This commit adds Python type stubs for the perf module to improve IDE support and static analysis. It includes docstrings for classes, methods, and constants derived from C source and JSON definitions.
Quoted 5 lines
tools/perf/python/perf.pyi new file +0
-0
[ ... ]
tools/perf/python/perf.pyi line 51
Quoted 5 lines
tools/perf/python/perf.pyi new file +5
-0+def parse_events( + event_string: str, + cpus: Optional[cpu_map] = None, + threads: Optional[Any] = None +) -> 'evlist':
[Severity: Medium]
Is it safe to document cpus and threads as Optional with a default of None?
If a user follows this stub and explicitly passes None to the C extension, pyrf__parse_events() parses it into PyObject pointers via PyArg_ParseTuple(). Because Py_None is not a NULL pointer, the C code will treat it as a valid struct pyrf_thread_map and try to dereference it, reading garbage memory:
tools/perf/util/python.c:pyrf__parse_events() {
...
if (!PyArg_ParseTupleAndKeywords(args, kwargs, "s|OO", kwlist,
&events, &pcpus, &pthreads))
...
threads = pthreads ? ((struct pyrf_thread_map *)pthreads)->threads : NULL;
...
}Should the C extension be updated to handle Py_None, or should the stubs avoid suggesting None is a valid argument?
[ ... ]
tools/perf/python/perf.pyi line 69
Quoted 6 lines
tools/perf/python/perf.pyi new file +6
-0+def parse_metrics( + metrics_string: str, + pmu: Optional[str] = None, + cpus: Optional[cpu_map] = None, + threads: Optional[Any] = None +) -> 'evlist':
[Severity: Medium]
Does this also have the same issue where passing None for cpus or threads will cause pyrf__parse_metrics() to dereference Py_None as a struct pointer?
Additionally, passing None for pmu might cause a TypeError if the C code expects a string format.
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=18
Ian Rogers
reviewerPATCH v15 19/19 perf python: Add LiveSession helper
In reply to Ian RogersAdd LiveSession class in tools/perf/python/perf_live.py to support live event collection using perf.evlist and perf.parse_events, avoiding the need to fork a separate perf record process.
---
v12:
- Replaced except TypeError with continue to ignore unmapped/offline CPUs instead of breaking the poll loop. Replaced hardcoded -11 with e.errno == errno.EAGAIN in LiveSession.run(). Fixed setup_python.sh PYTHONPATH prepend logic to cleanly append to existing paths.
v2:
1. Fixed File Descriptor Leak: I moved self.evlist.mmap() inside the try block so that if it raises an exception, the finally block will still be executed and call self.evlist.close() , preventing file descriptor leaks.
2. Handled InterruptedError in poll() : I wrapped the poll() call in a try-except block to catch InterruptedError and continue the loop. This prevents the live session from crashing on non-fatal signals like SIGWINCH .
3. Added evlist.config() : I added a call to self.evlist.config() in the constructor after parse_events() . This applies the default record options to the events, enabling sampling and setting up PERF_SAMPLE_* fields so that the kernel will actually generate PERF_RECORD_SAMPLE events.
4. Enable the evlist and be robust to exceptions from reading unsupported events like mmap2.
v8:
- Drain all events from a CPU before moving to the next.
--- tools/perf/python/perf_live.py | 55 ++++++++++++++++++++++ tools/perf/tests/shell/lib/setup_python.sh | 13 +++++ 2 files changed, 68 insertions(+) create mode 100755 tools/perf/python/perf_live.py
tools/perf/python/perf_live.py new file +55 -0
@@ -0,0 +1,55 @@ +# SPDX-License-Identifier: GPL-2.0 +""" +Live event session helper using perf.evlist. + +This module provides a LiveSession class that allows running a callback +for each event collected live from the system, similar to perf.session +but without requiring a perf.data file. +""" + +import errno +import perf + + +class LiveSession: + """Represents a live event collection session.""" + + def __init__(self, event_string: str, sample_callback): + self.event_string = event_string + self.sample_callback = sample_callback + # Create a cpu map for all online CPUs + self.cpus = perf.cpu_map() + # Parse events and set maps + self.evlist = perf.parse_events(self.event_string, self.cpus) + self.evlist.config() + + def run(self): + """Run the live session.""" + try: + self.evlist.open() + self.evlist.mmap() + self.evlist.enable() + + while True: + # Poll for events with 100ms timeout + try: + self.evlist.poll(100) + except InterruptedError: + continue + for cpu in self.cpus: + for _ in range(1000): # Limit to 1000 events per CPU per poll to prevent starvation + try: + event = self.evlist.read_on_cpu(cpu) + except TypeError: + # CPU might be unmapped or offline, wait for mmap event + break + + if event is None: + break + + if event.type == perf.RECORD_SAMPLE: + self.sample_callback(event) + except KeyboardInterrupt: + pass + finally: + self.evlist.close()
tools/perf/tests/shell/lib/setup_python.sh +13 -0
@@ -14,3 +14,16 @@ then echo Skipping test, python not detected please set environment variable PYTHON. exit 2 fi + +# Set PYTHONPATH to find the in-tree built perf.so first, avoiding system-wide perf.so +if [ -n "$PERF_EXEC_PATH" ] && [ -d "$PERF_EXEC_PATH/python" ]; then + PYTHONPATH_DIR="$PERF_EXEC_PATH/python" +elif [ -d "$(dirname "$0")/../../python" ]; then + PYTHONPATH_DIR="$(dirname "$0")/../../python" +elif [ -d "$(dirname "$0")/../python" ]; then + PYTHONPATH_DIR="$(dirname "$0")/../python" +fi + +if [ -n "$PYTHONPATH_DIR" ]; then + export PYTHONPATH="$PYTHONPATH_DIR${PYTHONPATH:+:$PYTHONPATH}" +fi
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
[email protected]
AI reviewRe: [PATCH v15 19/19] perf python: Add LiveSession helper
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider: - [High] LiveSession will crash immediately with an unhandled OSError when polling CPUs that have no…
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] LiveSession will crash immediately with an unhandled OSError when polling CPUs that have no events. --
commit c35bf37eb5bab2d0b282f5de633a0eb0e03997ea
Author: Ian Rogers <[email protected]>
perf python: Add LiveSession helper
This commit adds a LiveSession class in tools/perf/python/perf_live.py to support live event collection using perf.evlist and perf.parse_events, avoiding the need to fork a separate perf record process.
Quoted 5 lines
tools/perf/python/perf_live.py new file +0
-0
[ ... ]
tools/perf/python/perf_live.py line 11
Quoted 12 lines
tools/perf/python/perf_live.py new file +11
-0@@ -0,0 +1,55 @@ + while True: + # Poll for events with 100ms timeout + try: + self.evlist.poll(100) + except InterruptedError: + continue + for cpu in self.cpus: + for _ in range(1000): # Limit to 1000 events per CPU per poll to prevent starvation + try: + event = self.evlist.read_on_cpu(cpu) + except TypeError:
[Severity: High]
Will this crash with an unhandled OSError when polling CPUs that have no events?
The underlying C code pyrf_evlist__read_on_cpu() raises PyExc_OSError when perf_mmap__read_init() returns -EAGAIN for an empty buffer:
tools/perf/util/python.c:pyrf_evlist__read_on_cpu() {
...
if (err < 0) {
return PyErr_Format(PyExc_OSError,
"perf: error mmap read init, err=%d", err);
}
...
}Since evlist.poll() wakes up when any CPU has events, iterating over all CPUs here will likely hit empty buffers on some CPUs, raising an OSError that isn't caught by this try/except block.
Should OSError also be caught and handled here?
Quoted 12 lines
tools/perf/python/perf_live.py new file +12
-0+ # CPU might be unmapped or offline, wait for mmap event + break + + if event is None: + break + + if event.type == perf.RECORD_SAMPLE: + self.sample_callback(event) + except KeyboardInterrupt: + pass + finally: + self.evlist.close()
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=19
Ian Rogers
reviewerv16 cover Add additional python API support
In reply to Ian RogersThe perf script command has long supported running Python and Perl scripts by embedding libpython and libperl. This approach has several drawbacks:
- overhead by creating Python dictionaries for every event (whether used or not),
- complex build dependencies on specific Python/Perl versions,
- complications with threading due to perf being the interpreter,
- no clear way to run standalone scripts like ilist.py.
This series takes a different approach with some initial implementation posted as an RFC last October:
https://lore.kernel.org/linux-perf-users/[email protected]/
with the motivation coming up on the mailing list earlier:
https://lore.kernel.org/lkml/CAP-5=fWDqE8SYfOLZkg_0=4Ayx6E7O+h7uUp4NDeCFkiN4b7-w@mail.gmail.com/
The ultimate goal is to remove the embedded libpython and libperl support from perf entirely, expanding the existing perf Python module to provide full access to perf data files and events, allowing scripts to be run as standalone Python applications.
To make the review process more manageable, the original 58-patch series has been split. This v15 series represents "Phase 1: API & Infrastructure" (20 patches). The first 4 patches of Phase 1 (cleanups and arch-specific header sorting) have already been merged upstream.
This remaining set contains:
1. Missed explicit dependency cleanups and header sorting for util/ and python.
2. Crucial core safety infrastructure (reference counting for evlist/evsel) to support safe lifecycle management in garbage-collected Python.
3. The core Python API extensions (session wrappers, perf_data wrappers, sample accessors, stubs, and LiveSession helper).
The subsequent "Phase 2" series will contain the actual porting of all existing Python/Perl scripts to the new API (which yields up to 35x speedups as demonstrated previously) and the final removal of embedded interpreters.
Addressing v15 Review Feedback:
- Patch 2 (buffer overflow & type checks): The buffer overflow in pyrf_counts_values_set_values and unsafe PyArg_ParseTupleAndKeywords casts are pre-existing issues in tools/perf/util/python.c. To keep Phase 1 scoped strictly to new API infrastructure, these will be followed up and fixed independently during Phase 2.
- Patch 7 (evlist refcounting): The TOCTOU race condition and asymmetric drop cycle-leak mentioned in the review are known limitations of this transitional cycle-breaking logic. A proper, thread-safe fix requires a global design change that is out of scope for Phase 1. These structural cycle-breaking issues will be fully addressed in Phase 2.
- Patch 16 (syscall name/id): The script modifications (syscall-counts.py) are intentionally omitted from this patch. As stated, the porting of scripts to the new API will be done entirely in Phase 2. The commit message simply motivates the API addition.
---
v16 Changes
-----------
- Patch 6: Added missing CHECK_INITIALIZED for pevsel parameter in pyrf_evlist__add and added evsel__put(pevent->evsel) to pyrf_sample_event__delete to fix a git bisect reference leak.
- Patch 11: Restored the TypeError for oversized events instead of silently truncating them and causing memory corruption. Also added specific handling for -EAGAIN in pyrf_evlist__read_on_cpu to return Py_RETURN_NONE instead of raising OSError.
- Patch 12: Added explicit null-termination for the mmap2 filename to prevent heap out-of-bounds reads when accessed via T_STRING_INPLACE.
- Patch 14: Removed the fallback to perf_event__process_stat_event when no stat callback is provided, preventing a segmentation fault on PERF_RECORD_STAT events.
- Patch 18: Added a new patch 'perf python: Handle Py_None for thread and cpu maps' to fix a pre-existing C API bug where passing None from Python caused a crash. This addresses the feedback on the stubs.
v15 Changes
-----------
- Patch 11: Removed manual '\0' assignment for mmap/comm filenames which corrupted the sample_id trailer. Replaced with memset of the event buffer prior to memcpy to safely handle non-terminated strings without overwriting trailer data.
- Patch 18: Added python/perf.pyi to Makefile.perf dependencies. Added missing properties to counts_values and sample_event stubs, fixed branch_stack/callchain sequence protocols, and added _sample_members inheritance to stat_round_event in perf.pyi.
v14 Changes
-----------
- Patch 6: Replaced Py_IsTrue with PyObject_IsTrue to correctly evaluate Python truthiness and added error handling returning -1.
- Patch 7: Added comments to evlist__put and evlist__splice_list_tail noting that cycle breaking and pointer rebinding for spliced events will be addressed in subsequent patches. Removed "Phase 2" from the previous cycle-breaking comment as it doesn't make sense in the wider Linux code base context.
- Patch 10: Fixed pyrf_session__new missing tool.tracing_data initialization and wrapped kernel map creation in an access check.
- Patch 15: Exposed remaining bitfields from struct branch_flags in pyrf_branch_entry.
- Patch 18: Cleaned up perf.pyi in Makefile.perf's python-clean target. Renamed sc_id to id in syscall_name stub, and corrected property types.
- Patch 19: Fixed perf_live.py inner loop to break instead of continue on TypeError to properly exit when processing is complete.
v13 Changes
-----------
- Patch 6: Reverted idx parsing for pyrf_evsel__init. Added explicit NULL checks for evlist in session.c to prevent crashes when accessing properties.
- Patch 7: Rely natively on evlist.c's updated logic through evlist__add.
- Patch 8: Fixed memory leak in intel_pt_synth_ptwrite_sample by adding perf_sample__exit(&sample) on exit.
- Patch 9: Zeroed out perf_data struct via memset(&data, 0, sizeof(data)) before using it to prevent stale state issues.
- Patch 11: Switched from PyLong_FromUnsignedLong to PyLong_FromUnsignedLongLong to guarantee full 64-bit bounds representation even on 32-bit platforms.
- Patch 18: Modified Makefile.perf to correctly install perf.pyi. Synchronized .pyi stubs to fully match the C implementation signatures.
- Patch 19: Updated pyrf_evlist__read_on_cpu to safely return Py_RETURN_NONE on -EAGAIN natively from C instead of relying on broken OSError emulation. Updated perf_live.py loop to handle the native None return correctly.
v12 Changes
-----------
- Patch 6: Restored idx parsing in pyrf_evsel__init kwargs, and removed an erroneous Py_INCREF(pevsel) in pyrf_evlist__add that caused a memory leak.
- Patch 7: Fixed evlist__put cycle collection TOCTOU double free race by checking refcount_read == 1. Fixed evlist__purge recursion and cycle reference tearing logic.
- Patch 8: Removed an extra evsel__get in evsel__parse_sample to fix a reference leak on error paths.
- Patch 11: Used copy_size to bound max_len for NUL termination to fix an out-of-bounds write in pyrf_event__new. Also set pevent->event.header.size = copy_size.
- Patch 12: Used copy_size to bound max_len for NUL termination of mmap2 and comm events to fix an out-of-bounds write.
- Patch 13: Added an optional struct machine * argument to pyrf_event__new defaulting to the host machine if NULL, avoiding regressions for future phases.
- Patch 14: Made psession->tool.stat = perf_event__process_stat_event; conditional on !stat so it doesn't unconditionally overwrite user stat callbacks.
- Patch 18: Audited all perf.pyi event stubs: added full sample_members attributes (pid, tid, time, id, stream_id, period, cpu) to all payload events via inheritance. Fixed sample_event unique fields (ip, addr, phys_addr, weight, data_src, insn_cnt, cyc_cnt). Renamed mmap_event.addr to start. Removed read_event.value.
- Patch 19: Replaced except TypeError with continue to ignore unmapped/offline CPUs instead of breaking the poll loop. Replaced hardcoded -11 with e.errno == errno.EAGAIN in LiveSession.run(). Fixed setup_python.sh PYTHONPATH prepend logic to cleanly append to existing paths.
Ian Rogers (20):
perf util: Sort includes and add missed explicit dependencies perf python: Add missed explicit dependencies
perf evsel/evlist: Avoid unnecessary #includes
perf data: Add open flag
perf evlist: Add reference count
perf evsel: Add reference count
perf evlist: Add reference count checking
perf python: Use evsel in sample in pyrf_event
perf python: Add wrapper for perf_data file abstraction perf python: Add python session abstraction wrapping perf's session perf python: Refactor and add accessors to sample event perf python: Add mmap2 event
perf python: Add callchain support
perf python: Extend API for stat events in python.c
perf python: Expose brstack in sample event
perf python: Add syscall name/id to convert syscall number and name perf python: Add config file access
perf python: Handle Py_None for thread and cpu maps
perf python: Add perf.pyi stubs file
perf python: Add LiveSession helper
tools/perf/Makefile.perf | 9 +- tools/perf/arch/arm/util/cs-etm.c | 10 +- tools/perf/arch/arm64/util/arm-spe.c | 8 +- tools/perf/arch/arm64/util/hisi-ptt.c | 2 +- tools/perf/arch/x86/tests/hybrid.c | 22 +- tools/perf/arch/x86/tests/topdown.c | 4 +- tools/perf/arch/x86/util/auxtrace.c | 2 +- tools/perf/arch/x86/util/intel-bts.c | 6 +- tools/perf/arch/x86/util/intel-pt.c | 9 +- tools/perf/arch/x86/util/iostat.c | 14 +- tools/perf/bench/evlist-open-close.c | 29 +- tools/perf/builtin-annotate.c | 7 +- tools/perf/builtin-ftrace.c | 14 +- tools/perf/builtin-inject.c | 9 +- tools/perf/builtin-kvm.c | 14 +- tools/perf/builtin-kwork.c | 8 +- tools/perf/builtin-lock.c | 4 +- tools/perf/builtin-record.c | 95 +- tools/perf/builtin-report.c | 6 +- tools/perf/builtin-sched.c | 30 +- tools/perf/builtin-script.c | 15 +- tools/perf/builtin-stat.c | 83 +- tools/perf/builtin-top.c | 104 +- tools/perf/builtin-trace.c | 65 +- tools/perf/python/perf.pyi | 654 +++++ tools/perf/python/perf_live.py | 55 + tools/perf/tests/backward-ring-buffer.c | 26 +- tools/perf/tests/code-reading.c | 14 +- tools/perf/tests/event-times.c | 6 +- tools/perf/tests/event_update.c | 4 +- tools/perf/tests/evsel-roundtrip-name.c | 8 +- tools/perf/tests/evsel-tp-sched.c | 4 +- tools/perf/tests/expand-cgroup.c | 12 +- tools/perf/tests/hists_cumulate.c | 2 +- tools/perf/tests/hists_filter.c | 2 +- tools/perf/tests/hists_link.c | 2 +- tools/perf/tests/hists_output.c | 2 +- tools/perf/tests/hwmon_pmu.c | 7 +- tools/perf/tests/keep-tracking.c | 10 +- tools/perf/tests/mmap-basic.c | 24 +- tools/perf/tests/openat-syscall-all-cpus.c | 6 +- tools/perf/tests/openat-syscall-tp-fields.c | 26 +- tools/perf/tests/openat-syscall.c | 6 +- tools/perf/tests/parse-events.c | 139 +- tools/perf/tests/parse-metric.c | 8 +- tools/perf/tests/parse-no-sample-id-all.c | 2 +- tools/perf/tests/perf-record.c | 38 +- tools/perf/tests/perf-time-to-tsc.c | 12 +- tools/perf/tests/pfm.c | 12 +- tools/perf/tests/pmu-events.c | 11 +- tools/perf/tests/pmu.c | 4 +- tools/perf/tests/sample-parsing.c | 45 +- tools/perf/tests/shell/lib/setup_python.sh | 13 + tools/perf/tests/sw-clock.c | 20 +- tools/perf/tests/switch-tracking.c | 11 +- tools/perf/tests/task-exit.c | 20 +- tools/perf/tests/time-utils-test.c | 14 +- tools/perf/tests/tool_pmu.c | 7 +- tools/perf/tests/topology.c | 4 +- tools/perf/tests/uncore-event-sorting.c | 6 +- tools/perf/ui/browsers/annotate.c | 2 +- tools/perf/ui/browsers/hists.c | 22 +- tools/perf/util/Build | 1 - tools/perf/util/amd-sample-raw.c | 2 +- tools/perf/util/annotate-data.c | 2 +- tools/perf/util/annotate.c | 10 +- tools/perf/util/auxtrace.c | 14 +- tools/perf/util/block-info.c | 4 +- tools/perf/util/bpf_counter.c | 2 +- tools/perf/util/bpf_counter_cgroup.c | 14 +- tools/perf/util/bpf_ftrace.c | 9 +- tools/perf/util/bpf_lock_contention.c | 12 +- tools/perf/util/bpf_off_cpu.c | 44 +- tools/perf/util/bpf_trace_augment.c | 8 +- tools/perf/util/cgroup.c | 26 +- tools/perf/util/cs-etm.c | 5 +- tools/perf/util/data-convert-bt.c | 2 +- tools/perf/util/data.c | 27 +- tools/perf/util/data.h | 4 +- tools/perf/util/evlist.c | 496 ++-- tools/perf/util/evlist.h | 273 +- tools/perf/util/evsel.c | 39 +- tools/perf/util/evsel.h | 40 +- tools/perf/util/expr.c | 2 +- tools/perf/util/header.c | 69 +- tools/perf/util/header.h | 2 +- tools/perf/util/intel-pt.c | 2 +- tools/perf/util/intel-tpebs.c | 7 +- tools/perf/util/iostat.c | 2 +- tools/perf/util/iostat.h | 2 +- tools/perf/util/map.h | 9 +- tools/perf/util/metricgroup.c | 12 +- tools/perf/util/parse-events.c | 10 +- tools/perf/util/parse-events.y | 2 +- tools/perf/util/perf_api_probe.c | 20 +- tools/perf/util/pfm.c | 4 +- tools/perf/util/print-events.c | 2 +- tools/perf/util/python.c | 2633 ++++++++++++++++--- tools/perf/util/record.c | 11 +- tools/perf/util/s390-sample-raw.c | 20 +- tools/perf/util/sample-raw.c | 4 +- tools/perf/util/sample.c | 17 +- tools/perf/util/session.c | 69 +- tools/perf/util/session.h | 2 + tools/perf/util/setup.py | 5 + tools/perf/util/sideband_evlist.c | 40 +- tools/perf/util/sort.c | 2 +- tools/perf/util/stat-display.c | 6 +- tools/perf/util/stat-shadow.c | 24 +- tools/perf/util/stat.c | 20 +- tools/perf/util/stream.c | 4 +- tools/perf/util/synthetic-events.c | 11 +- tools/perf/util/time-utils.c | 12 +- tools/perf/util/top.c | 4 +- 114 files changed, 4445 insertions(+), 1447 deletions(-) create mode 100644 tools/perf/python/perf.pyi create mode 100755 tools/perf/python/perf_live.py
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
Ian Rogers
reviewerPATCH v16 1/20 perf util: Sort includes and add missed explicit dependencies
Fix missing includes found while cleaning the evsel/evlist header files. Sort the remaining header files for consistency with the rest of the code.
--- tools/perf/util/bpf_off_cpu.c | 30 +++++----- tools/perf/util/bpf_trace_augment.c | 8 +-- tools/perf/util/evlist.c | 91 +++++++++++++++-------------- tools/perf/util/map.h | 9 ++- tools/perf/util/perf_api_probe.c | 18 +++--- tools/perf/util/s390-sample-raw.c | 19 +++--- tools/perf/util/stat-shadow.c | 20 ++++--- tools/perf/util/stat.c | 16 +++-- 8 files changed, 113 insertions(+), 98 deletions(-)
tools/perf/util/bpf_off_cpu.c +16 -14
@@ -1,23 +1,25 @@// SPDX-License-Identifier: GPL-2.0 -#include "util/bpf_counter.h" -#include "util/debug.h" -#include "util/evsel.h" -#include "util/evlist.h" -#include "util/off_cpu.h" -#include "util/perf-hooks.h" -#include "util/record.h" -#include "util/session.h" -#include "util/target.h" -#include "util/cpumap.h" -#include "util/thread_map.h" -#include "util/cgroup.h" -#include "util/strlist.h" +#include <linux/time64.h> +#include <bpf/bpf.h> #include <bpf/btf.h> #include <internal/xyarray.h> -#include <linux/time64.h> +#include "bpf_counter.h" #include "bpf_skel/off_cpu.skel.h" +#include "cgroup.h" +#include "cpumap.h" +#include "debug.h" +#include "evlist.h" +#include "evsel.h" +#include "off_cpu.h" +#include "parse-events.h" +#include "perf-hooks.h" +#include "record.h" +#include "session.h" +#include "strlist.h" +#include "target.h" +#include "thread_map.h" #define MAX_STACKS 32#define MAX_PROC 4096
tools/perf/util/bpf_trace_augment.c +4 -4
@@ -1,11 +1,11 @@#include <bpf/libbpf.h> #include <internal/xyarray.h> -#include "util/debug.h" -#include "util/evlist.h" -#include "util/trace_augment.h" -#include "bpf_skel/augmented_raw_syscalls.skel.h" +#include "debug.h" +#include "evlist.h" +#include "parse-events.h" +#include "trace_augment.h" static struct augmented_raw_syscalls_bpf *skel;static struct evsel *bpf_output;
tools/perf/util/evlist.c +46 -45
113 lines not on this page · raw message
tools/perf/util/map.h +4 -5
@@ -2,14 +2,13 @@#ifndef __PERF_MAP_H#define __PERF_MAP_H -#include <linux/refcount.h> -#include <linux/compiler.h> -#include <linux/list.h> -#include <linux/rbtree.h> +#include <stdbool.h> #include <stdio.h> #include <string.h> -#include <stdbool.h> + +#include <linux/refcount.h> #include <linux/types.h> +#include <internal/rc_check.h> struct dso ;
tools/perf/util/perf_api_probe.c +11 -7
@@ -1,14 +1,18 @@/* SPDX-License-Identifier: GPL-2.0 */ +#include "perf_api_probe.h" -#include "perf-sys.h" -#include "util/cloexec.h" -#include "util/evlist.h" -#include "util/evsel.h" -#include "util/parse-events.h" -#include "util/perf_api_probe.h" -#include <perf/cpumap.h> #include <errno.h> +#include <perf/cpumap.h> + +#include "cloexec.h" +#include "evlist.h" +#include "evsel.h" +#include "parse-events.h" +#include "perf-sys.h" +#include "pmu.h" +#include "pmus.h" +typedef void (*setup_probe_fn_t )(struct evsel *evsel);static int perf_do_probe_api (setup_probe_fn_t fn,struct perf_cpu cpu,const char *str)
tools/perf/util/s390-sample-raw.c +10 -9
@@ -12,25 +12,26 @@* sample was taken from. */ -#include <unistd.h> +#include <inttypes.h> #include <stdio.h> #include <string.h> -#include <inttypes.h> -#include <sys/stat.h> +#include <asm/byteorder.h> #include <linux/compiler.h> #include <linux/err.h> -#include <asm/byteorder.h> +#include <sys/stat.h> +#include <unistd.h> +#include "color.h" #include "debug.h" -#include "session.h" #include "evlist.h" -#include "color.h" #include "hashmap.h" -#include "sample-raw.h" +#include "pmu.h" +#include "pmus.h" #include "s390-cpumcf-kernel.h" -#include "util/pmu.h" -#include "util/sample.h" +#include "sample-raw.h" +#include "sample.h" +#include "session.h" static size_t ctrset_size (struct cf_ctrset_entry *set) {
tools/perf/util/stat-shadow.c +12 -8
@@ -2,20 +2,24 @@#include <errno.h> #include <math.h> #include <stdio.h> -#include "evsel.h" -#include "stat.h" + +#include <linux/zalloc.h> + +#include "cgroup.h" #include "color.h" #include "debug.h" -#include "pmu.h" -#include "rblist.h" #include "evlist.h" +#include "evsel.h" #include "expr.h" -#include "metricgroup.h" -#include "cgroup.h" -#include "units.h" +#include "hashmap.h" #include "iostat.h" -#include "util/hashmap.h" +#include "metricgroup.h" +#include "pmu.h" +#include "pmus.h" +#include "rblist.h" +#include "stat.h" #include "tool_pmu.h" +#include "units.h" static bool tool_pmu__is_time_event (const struct perf_stat_config *config,const struct evsel *evsel,int *tool_aggr_idx)
tools/perf/util/stat.c +10 -6
@@ -1,21 +1,25 @@// SPDX-License-Identifier: GPL-2.0 +#include "stat.h" +#include <errno.h> -#include <linux/err.h> #include <inttypes.h> #include <math.h> #include <string.h> + +#include <linux/err.h> +#include <linux/zalloc.h> +#include "counts.h" #include "cpumap.h" #include "debug.h" +#include "evlist.h" +#include "evsel.h" +#include "hashmap.h" #include "header.h" -#include "stat.h" +#include "pmu.h" #include "session.h" #include "target.h" -#include "evlist.h" -#include "evsel.h" #include "thread_map.h" -#include "util/hashmap.h" -#include <linux/zalloc.h> void update_stats (struct stats *stats,u64 val) {
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
Ian Rogers
reviewerPATCH v16 2/20 perf python: Add missed explicit dependencies
In reply to Ian RogersFix missing #include of pmus.h found while cleaning the evsel/evlist header files.
Also define PY_SSIZE_T_CLEAN before including Python.h to comply with modern Python 3 C-API requirements, and introduce CHECK_INITIALIZED and CHECK_INITIALIZED_INT safety macros to be used by subsequent patches.
--- tools/perf/util/python.c | 41 +++++++++++++++++++++++++++++++--------- 1 file changed, 32 insertions(+), 9 deletions(-)
tools/perf/util/python.c +32 -9
@@ -1,30 +1,37 @@// SPDX-License-Identifier: GPL-2.0 +#define PY_SSIZE_T_CLEAN #include <Python.h> -#include <structmember.h> +#include <inttypes.h> -#include <poll.h> +#include <linux/err.h> +#include <poll.h> + +#include <internal/lib.h> #include <perf/cpumap.h> -#ifdef HAVE_LIBTRACEEVENT -#include <event-parse.h> -#endif #include <perf/mmap.h> +#include <structmember.h> +#include "callchain.h" #include "counts.h" +#include "event.h" #include "evlist.h" #include "evsel.h" -#include "event.h" #include "expr.h" +#include "metricgroup.h" +#include "mmap.h" +#include "pmus.h" #include "print_binary.h" #include "record.h" #include "strbuf.h" #include "thread_map.h" #include "tp_pmu.h" #include "trace-event.h" -#include "metricgroup.h" -#include "mmap.h" #include "util/sample.h" -#include <internal/lib.h> + +#ifdef HAVE_LIBTRACEEVENT +#include <event-parse.h> +#endif PyMODINIT_FUNC PyInit_perf (void );
@@ -38,6 +45,22 @@ PyMODINIT_FUNC PyInit_perf(void); offsetof(struct pyrf_event , sample) + offsetof(struct perf_sample , member), \ 0, help } +#define CHECK_INITIALIZED (ptr, msg) \ + do { \ + if (!(ptr)) { \ + PyErr_SetString(PyExc_ValueError, msg " not initialized"); \ + return NULL; \ + } \ + } while (0) + +#define CHECK_INITIALIZED_INT (ptr, msg) \ + do { \ + if (!(ptr)) { \ + PyErr_SetString(PyExc_ValueError, msg " not initialized"); \ + return -1; \ + } \ + } while (0) +struct pyrf_event {PyObject_HEAD struct evsel *evsel;
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog
[email protected]
AI reviewRe: [PATCH v16 02/20] perf python: Add missed explicit dependencies
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
Pre-existing issues:
- [High] Type mismatch in PyArg_ParseTupleAndKeywords format string causes severe logic errors on big-endian architectures. --
commit 5f219de4611e34d146886cd2113bb847c538df26
Author: Ian Rogers <[email protected]>
perf python: Add missed explicit dependencies
This commit fixes a missing include of pmus.h and sorts headers. It also defines PY_SSIZE_T_CLEAN for modern Python 3 C-API compatibility and introduces initialization check macros for future use.
tools/perf/util/python.c
Quoted 4 lines
tools/perf/util/python.c +0
-0
[ ... ]
[Severity: High]
This isn't a bug introduced by this patch, but there is a type mismatch in pyrf_evsel__init() that causes logic errors on big-endian architectures.
The format string for PyArg_ParseTupleAndKeywords() uses i (expecting a 32-bit int pointer) for attr.sample_freq and attr.read_format, which are both 64-bit __u64 fields:
tools/perf/util/python.c:pyrf_evsel__init() {
...
if (!PyArg_ParseTupleAndKeywords(args, kwargs,
"|iKiKKiiiiiiiiiiiiiiiiiiiiiiKK", kwlist, &attr.type, &attr.config, &attr.sample_freq, &sample_period, &attr.sample_type, &attr.read_format, &disabled, &inherit,
...
}
Writing a 32-bit integer into a 64-bit variable on big-endian systems modifies the upper 32 bits, leaving the lower 32 bits as zero. For example, a sample_freq of 1000 would be parsed as 1000 << 32.
While the missing specifier for idx is fixed later in this patch series, the type mismatches for sample_freq and read_format remain unresolved.
Could the format string be updated to use the correct 64-bit specifier for these __u64 fields?
Signature · 1 line
Sashiko AI review · https://sashiko.dev/#/patchset/[email protected]?part=2
Ian Rogers
reviewerPATCH v16 3/20 perf evsel/evlist: Avoid unnecessary #includes
In reply to Ian RogersUse forward declarations and remove unnecessary #includes in evsel.h. Sort the forward declarations in evsel.h and evlist.h. Move some PMU code into evsel.c.
--- tools/perf/util/evlist.h | 15 +++++++++------ tools/perf/util/evsel.c | 5 +++++ tools/perf/util/evsel.h | 25 ++++++++++++------------- 3 files changed, 26 insertions(+), 19 deletions(-)
tools/perf/util/evlist.h +9 -6
@@ -2,29 +2,32 @@#ifndef __PERF_EVLIST_H#define __PERF_EVLIST_H 1 +#include <signal.h> +#include <linux/compiler.h> #include <linux/kernel.h> -#include <linux/refcount.h> #include <linux/list.h> +#include <linux/refcount.h> +#include <pthread.h> +#include <unistd.h> +#include <api/fd/array.h> #include <internal/evlist.h> #include <internal/evsel.h> #include <perf/evlist.h> +#include "affinity.h" #include "events_stats.h" #include "evsel.h" #include "rblist.h" -#include <pthread.h> -#include <signal.h> -#include <unistd.h> -struct pollfd ; -struct thread_map ;struct perf_cpu_map ;struct perf_stat_config ; +struct pollfd ;struct record_opts ;struct strbuf ;struct target ; +struct thread_map ;/* * State machine of bkw_mmap_state:
tools/perf/util/evsel.c +5 -0
@@ -370,6 +370,11 @@ void evsel__set_sample_id(struct evsel *evsel, evsel->core.attr.read_format |=PERF_FORMAT_ID ; } +bool evsel__is_non_perf_event_open_pmu (const struct evsel *evsel) +{ +return evsel->pmu && evsel->pmu->type >PERF_PMU_TYPE_PE_END ; +} +/** * evsel__is_function_event - Return whether given evsel is a function * trace event
tools/perf/util/evsel.h +12 -13
@@ -2,28 +2,30 @@#ifndef __PERF_EVSEL_H#define __PERF_EVSEL_H 1 -#include <linux/list.h> #include <stdbool.h> -#include <sys/types.h> + +#include <linux/list.h> #include <linux/perf_event.h> #include <linux/types.h> +#include <sys/types.h> +#include <internal/evsel.h> #include <perf/evsel.h> +#include "symbol_conf.h" -#include "pmus.h" -#include "pmu.h" +struct bperf_follower_bpf ; +struct bperf_leader_bpf ; +struct bpf_counter_ops ;struct bpf_object ;struct cgroup ; +struct hashmap ;struct perf_counts ; +struct perf_pmu ;struct perf_stat_config ;struct perf_stat_evsel ; -union perf_event ; -struct bpf_counter_ops ;struct target ; -struct hashmap ; -struct bperf_leader_bpf ; -struct bperf_follower_bpf ; +union perf_event ;typedef int (evsel__sb_cb_t )(union perf_event *event,void *data);
@@ -356,10 +358,6 @@ void arch_evsel__apply_ratio_to_prev(struct evsel *evsel, struct perf_event_attrint evsel__set_filter (struct evsel *evsel,const char *filter);int evsel__append_tp_filter (struct evsel *evsel,const char *filter);int evsel__append_addr_filter (struct evsel *evsel,const char *filter); -static inline bool evsel__is_non_perf_event_open_pmu (const struct evsel *evsel) -{ -return evsel->pmu && evsel->pmu->type >PERF_PMU_TYPE_PE_END ; -}int evsel__enable_cpu (struct evsel *evsel,int cpu_map_idx);int evsel__enable (struct evsel *evsel);
@@ -478,6 +476,7 @@ static inline bool evsel__is_group_event(struct evsel *evsel)return evsel__is_group_leader (evsel) && evsel->core.nr_members >1 ; } +bool evsel__is_non_perf_event_open_pmu (const struct evsel *evsel);bool evsel__is_function_event (struct evsel *evsel);static inline bool evsel__is_bpf_output (struct evsel *evsel)
Signature · 1 line
2.54.0.1136.gdb2ca164c4-goog