forge/linux-pinenote
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Merge branch 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip

Browse source 
* 'perf-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: (107 commits)
  perf stat: Add more cache-miss percentage printouts
  perf stat: Add -d -d and -d -d -d options to show more CPU events
  ftrace/kbuild: Add recordmcount files to force full build
  ftrace: Add self-tests for multiple function trace users
  ftrace: Modify ftrace_set_filter/notrace to take ops
  ftrace: Allow dynamically allocated function tracers
  ftrace: Implement separate user function filtering
  ftrace: Free hash with call_rcu_sched()
  ftrace: Have global_ops store the functions that are to be traced
  ftrace: Add ops parameter to ftrace_startup/shutdown functions
  ftrace: Add enabled_functions file
  ftrace: Use counters to enable functions to trace
  ftrace: Separate hash allocation and assignment
  ftrace: Create a global_ops to hold the filter and notrace hashes
  ftrace: Use hash instead for FTRACE_FL_FILTER
  ftrace: Replace FTRACE_FL_NOTRACE flag with a hash of ignored functions
  perf bench, x86: Add alternatives-asm.h wrapper
  x86, 64-bit: Fix copy_[to/from]_user() checks for the userspace address limit
  x86, mem: memset_64.S: Optimize memset by enhanced REP MOVSB/STOSB
  x86, mem: memmove_64.S: Optimize memmove by enhanced REP MOVSB/STOSB
  ...
This commit is contained in:
Linus Torvalds 2011-05-19 17:36:08 -07:00
commit df48d8716e
78 changed files with 4086 additions and 1773 deletions
Download patch file Download diff file Expand all files Collapse all files

52
tools/perf/Documentation/perf-script.txt
View file

@ -113,13 +113,61 @@ OPTIONS
Do various checks like samples ordering and lost events.
-f::
--fields
--fields::
Comma separated list of fields to print. Options are:
comm, tid, pid, time, cpu, event, trace, sym. Field
list must be prepended with the type, trace, sw or hw,
list can be prepended with the type, trace, sw or hw,
to indicate to which event type the field list applies.
e.g., -f sw:comm,tid,time,sym and -f trace:time,cpu,trace
perf script -f <fields>
is equivalent to:
perf script -f trace:<fields> -f sw:<fields> -f hw:<fields>
i.e., the specified fields apply to all event types if the type string
is not given.
The arguments are processed in the order received. A later usage can
reset a prior request. e.g.:
-f trace: -f comm,tid,time,sym
The first -f suppresses trace events (field list is ""), but then the
second invocation sets the fields to comm,tid,time,sym. In this case a
warning is given to the user:
"Overriding previous field request for all events."
Alternativey, consider the order:
-f comm,tid,time,sym -f trace:
The first -f sets the fields for all events and the second -f
suppresses trace events. The user is given a warning message about
the override, and the result of the above is that only S/W and H/W
events are displayed with the given fields.
For the 'wildcard' option if a user selected field is invalid for an
event type, a message is displayed to the user that the option is
ignored for that type. For example:
$ perf script -f comm,tid,trace
'trace' not valid for hardware events. Ignoring.
'trace' not valid for software events. Ignoring.
Alternatively, if the type is given an invalid field is specified it
is an error. For example:
perf script -v -f sw:comm,tid,trace
'trace' not valid for software events.
At this point usage is displayed, and perf-script exits.
Finally, a user may not set fields to none for all event types.
i.e., -f "" is not allowed.
-k::
--vmlinux=<file>::
vmlinux pathname

107
tools/perf/Makefile
View file

@ -5,6 +5,8 @@ endif
# The default target of this Makefile is...
all:
include config/utilities.mak
ifneq ($(OUTPUT),)
# check that the output directory actually exists
OUTDIR := $(shell cd $(OUTPUT) && /bin/pwd)
@ -13,6 +15,12 @@ endif
# Define V to have a more verbose compile.
#
# Define PYTHON to point to the python binary if the default
# `python' is not correct; for example: PYTHON=python2
#
# Define PYTHON_CONFIG to point to the python-config binary if
# the default `$(PYTHON)-config' is not correct.
#
# Define ASCIIDOC8 if you want to format documentation with AsciiDoc 8
#
# Define DOCBOOK_XSL_172 if you want to format man pages with DocBook XSL v1.72.
@ -134,7 +142,7 @@ INSTALL = install
# explicitly what architecture to check for. Fix this up for yours..
SPARSE_FLAGS = -D__BIG_ENDIAN__ -D__powerpc__
-include feature-tests.mak
-include config/feature-tests.mak
ifeq ($(call try-cc,$(SOURCE_HELLO),-Werror -fstack-protector-all),y)
CFLAGS := $(CFLAGS) -fstack-protector-all
@ -169,12 +177,10 @@ grep-libs = $(filter -l%,$(1))
strip-libs = $(filter-out -l%,$(1))
$(OUTPUT)python/perf.so: $(PYRF_OBJS)
$(QUIET_GEN)( \
export CFLAGS="$(BASIC_CFLAGS)"; \
python util/setup.py --quiet build_ext --build-lib='$(OUTPUT)python' \
--build-temp='$(OUTPUT)python/temp' \
)
$(QUIET_GEN)CFLAGS='$(BASIC_CFLAGS)' $(PYTHON_WORD) util/setup.py \
--quiet build_ext \
--build-lib='$(OUTPUT)python' \
--build-temp='$(OUTPUT)python/temp'
#
# No Perl scripts right now:
#
@ -479,24 +485,74 @@ else
endif
endif
ifdef NO_LIBPYTHON
BASIC_CFLAGS += -DNO_LIBPYTHON
disable-python = $(eval $(disable-python_code))
define disable-python_code
BASIC_CFLAGS += -DNO_LIBPYTHON
$(if $(1),$(warning No $(1) was found))
$(warning Python support won't be built)
endef
override PYTHON := \
$(call get-executable-or-default,PYTHON,python)
ifndef PYTHON
$(call disable-python,python interpreter)
python-clean :=
else
PYTHON_EMBED_LDOPTS = $(shell python-config --ldflags 2>/dev/null)
PYTHON_EMBED_LDFLAGS = $(call strip-libs,$(PYTHON_EMBED_LDOPTS))
PYTHON_EMBED_LIBADD = $(call grep-libs,$(PYTHON_EMBED_LDOPTS))
PYTHON_EMBED_CCOPTS = `python-config --cflags 2>/dev/null`
FLAGS_PYTHON_EMBED=$(PYTHON_EMBED_CCOPTS) $(PYTHON_EMBED_LDOPTS)
ifneq ($(call try-cc,$(SOURCE_PYTHON_EMBED),$(FLAGS_PYTHON_EMBED)),y)
msg := $(warning No Python.h found, install python-dev[el] to have python support in 'perf script' and to build the python bindings)
BASIC_CFLAGS += -DNO_LIBPYTHON
else
ALL_LDFLAGS += $(PYTHON_EMBED_LDFLAGS)
EXTLIBS += $(PYTHON_EMBED_LIBADD)
LIB_OBJS += $(OUTPUT)util/scripting-engines/trace-event-python.o
LIB_OBJS += $(OUTPUT)scripts/python/Perf-Trace-Util/Context.o
LANG_BINDINGS += $(OUTPUT)python/perf.so
endif
PYTHON_WORD := $(call shell-wordify,$(PYTHON))
python-clean := $(PYTHON_WORD) util/setup.py clean \
--build-lib='$(OUTPUT)python' \
--build-temp='$(OUTPUT)python/temp'
ifdef NO_LIBPYTHON
$(call disable-python)
else
override PYTHON_CONFIG := \
$(call get-executable-or-default,PYTHON_CONFIG,$(PYTHON)-config)
ifndef PYTHON_CONFIG
$(call disable-python,python-config tool)
else
PYTHON_CONFIG_SQ := $(call shell-sq,$(PYTHON_CONFIG))
PYTHON_EMBED_LDOPTS := $(shell $(PYTHON_CONFIG_SQ) --ldflags 2>/dev/null)
PYTHON_EMBED_LDFLAGS := $(call strip-libs,$(PYTHON_EMBED_LDOPTS))
PYTHON_EMBED_LIBADD := $(call grep-libs,$(PYTHON_EMBED_LDOPTS))
PYTHON_EMBED_CCOPTS := $(shell $(PYTHON_CONFIG_SQ) --cflags 2>/dev/null)
FLAGS_PYTHON_EMBED := $(PYTHON_EMBED_CCOPTS) $(PYTHON_EMBED_LDOPTS)
ifneq ($(call try-cc,$(SOURCE_PYTHON_EMBED),$(FLAGS_PYTHON_EMBED)),y)
$(call disable-python,Python.h (for Python 2.x))
else
ifneq ($(call try-cc,$(SOURCE_PYTHON_VERSION),$(FLAGS_PYTHON_EMBED)),y)
$(warning Python 3 is not yet supported; please set)
$(warning PYTHON and/or PYTHON_CONFIG appropriately.)
$(warning If you also have Python 2 installed, then)
$(warning try something like:)
$(warning $(and ,))
$(warning $(and ,) make PYTHON=python2)
$(warning $(and ,))
$(warning Otherwise, disable Python support entirely:)
$(warning $(and ,))
$(warning $(and ,) make NO_LIBPYTHON=1)
$(warning $(and ,))
$(error $(and ,))
else
ALL_LDFLAGS += $(PYTHON_EMBED_LDFLAGS)
EXTLIBS += $(PYTHON_EMBED_LIBADD)
LIB_OBJS += $(OUTPUT)util/scripting-engines/trace-event-python.o
LIB_OBJS += $(OUTPUT)scripts/python/Perf-Trace-Util/Context.o
LANG_BINDINGS += $(OUTPUT)python/perf.so
endif
endif
endif
endif
endif
ifdef NO_DEMANGLE
@ -837,8 +893,7 @@ clean:
$(RM) *.spec *.pyc *.pyo */*.pyc */*.pyo $(OUTPUT)common-cmds.h TAGS tags cscope*
$(MAKE) -C Documentation/ clean
$(RM) $(OUTPUT)PERF-VERSION-FILE $(OUTPUT)PERF-CFLAGS
@python util/setup.py clean --build-lib='$(OUTPUT)python' \
--build-temp='$(OUTPUT)python/temp'
$(python-clean)
.PHONY: all install clean strip
.PHONY: shell_compatibility_test please_set_SHELL_PATH_to_a_more_modern_shell

300
tools/perf/builtin-script.c
View file

@ -49,57 +49,169 @@ struct output_option {
};
/* default set to maintain compatibility with current format */
static u64 output_fields[PERF_TYPE_MAX] = {
[PERF_TYPE_HARDWARE] = PERF_OUTPUT_COMM | PERF_OUTPUT_TID | \
PERF_OUTPUT_CPU | PERF_OUTPUT_TIME | \
PERF_OUTPUT_EVNAME | PERF_OUTPUT_SYM,
static struct {
bool user_set;
bool wildcard_set;
u64 fields;
u64 invalid_fields;
} output[PERF_TYPE_MAX] = {
[PERF_TYPE_SOFTWARE] = PERF_OUTPUT_COMM | PERF_OUTPUT_TID | \
PERF_OUTPUT_CPU | PERF_OUTPUT_TIME | \
PERF_OUTPUT_EVNAME | PERF_OUTPUT_SYM,
[PERF_TYPE_HARDWARE] = {
.user_set = false,
[PERF_TYPE_TRACEPOINT] = PERF_OUTPUT_COMM | PERF_OUTPUT_TID | \
PERF_OUTPUT_CPU | PERF_OUTPUT_TIME | \
PERF_OUTPUT_EVNAME | PERF_OUTPUT_TRACE,
.fields = PERF_OUTPUT_COMM | PERF_OUTPUT_TID |
PERF_OUTPUT_CPU | PERF_OUTPUT_TIME |
PERF_OUTPUT_EVNAME | PERF_OUTPUT_SYM,
.invalid_fields = PERF_OUTPUT_TRACE,
},
[PERF_TYPE_SOFTWARE] = {
.user_set = false,
.fields = PERF_OUTPUT_COMM | PERF_OUTPUT_TID |
PERF_OUTPUT_CPU | PERF_OUTPUT_TIME |
PERF_OUTPUT_EVNAME | PERF_OUTPUT_SYM,
.invalid_fields = PERF_OUTPUT_TRACE,
},
[PERF_TYPE_TRACEPOINT] = {
.user_set = false,
.fields = PERF_OUTPUT_COMM | PERF_OUTPUT_TID |
PERF_OUTPUT_CPU | PERF_OUTPUT_TIME |
PERF_OUTPUT_EVNAME | PERF_OUTPUT_TRACE,
},
[PERF_TYPE_RAW] = {
.user_set = false,
.fields = PERF_OUTPUT_COMM | PERF_OUTPUT_TID |
PERF_OUTPUT_CPU | PERF_OUTPUT_TIME |
PERF_OUTPUT_EVNAME | PERF_OUTPUT_SYM,
.invalid_fields = PERF_OUTPUT_TRACE,
},
};
static bool output_set_by_user;
#define PRINT_FIELD(x) (output_fields[attr->type] & PERF_OUTPUT_##x)
static int perf_session__check_attr(struct perf_session *session,
struct perf_event_attr *attr)
static bool output_set_by_user(void)
{
int j;
for (j = 0; j < PERF_TYPE_MAX; ++j) {
if (output[j].user_set)
return true;
}
return false;
}
static const char *output_field2str(enum perf_output_field field)
{
int i, imax = ARRAY_SIZE(all_output_options);
const char *str = "";
for (i = 0; i < imax; ++i) {
if (all_output_options[i].field == field) {
str = all_output_options[i].str;
break;
}
}
return str;
}
#define PRINT_FIELD(x) (output[attr->type].fields & PERF_OUTPUT_##x)
static int perf_event_attr__check_stype(struct perf_event_attr *attr,
u64 sample_type, const char *sample_msg,
enum perf_output_field field)
{
int type = attr->type;
const char *evname;
if (attr->sample_type & sample_type)
return 0;
if (output[type].user_set) {
evname = __event_name(attr->type, attr->config);
pr_err("Samples for '%s' event do not have %s attribute set. "
"Cannot print '%s' field.\n",
evname, sample_msg, output_field2str(field));
return -1;
}
/* user did not ask for it explicitly so remove from the default list */
output[type].fields &= ~field;
evname = __event_name(attr->type, attr->config);
pr_debug("Samples for '%s' event do not have %s attribute set. "
"Skipping '%s' field.\n",
evname, sample_msg, output_field2str(field));
return 0;
}
static int perf_evsel__check_attr(struct perf_evsel *evsel,
struct perf_session *session)
{
struct perf_event_attr *attr = &evsel->attr;
if (PRINT_FIELD(TRACE) &&
!perf_session__has_traces(session, "record -R"))
return -EINVAL;
if (PRINT_FIELD(SYM)) {
if (!(session->sample_type & PERF_SAMPLE_IP)) {
pr_err("Samples do not contain IP data.\n");
if (perf_event_attr__check_stype(attr, PERF_SAMPLE_IP, "IP",
PERF_OUTPUT_SYM))
return -EINVAL;
}
if (!no_callchain &&
!(session->sample_type & PERF_SAMPLE_CALLCHAIN))
!(attr->sample_type & PERF_SAMPLE_CALLCHAIN))
symbol_conf.use_callchain = false;
}
if ((PRINT_FIELD(PID) || PRINT_FIELD(TID)) &&
!(session->sample_type & PERF_SAMPLE_TID)) {
pr_err("Samples do not contain TID/PID data.\n");
perf_event_attr__check_stype(attr, PERF_SAMPLE_TID, "TID",
PERF_OUTPUT_TID|PERF_OUTPUT_PID))
return -EINVAL;
}
if (PRINT_FIELD(TIME) &&
!(session->sample_type & PERF_SAMPLE_TIME)) {
pr_err("Samples do not contain timestamps.\n");
perf_event_attr__check_stype(attr, PERF_SAMPLE_TIME, "TIME",
PERF_OUTPUT_TIME))
return -EINVAL;
}
if (PRINT_FIELD(CPU) &&
!(session->sample_type & PERF_SAMPLE_CPU)) {
pr_err("Samples do not contain cpu.\n");
perf_event_attr__check_stype(attr, PERF_SAMPLE_CPU, "CPU",
PERF_OUTPUT_CPU))
return -EINVAL;
return 0;
}
/*
* verify all user requested events exist and the samples
* have the expected data
*/
static int perf_session__check_output_opt(struct perf_session *session)
{
int j;
struct perf_evsel *evsel;
for (j = 0; j < PERF_TYPE_MAX; ++j) {
evsel = perf_session__find_first_evtype(session, j);
/*
* even if fields is set to 0 (ie., show nothing) event must
* exist if user explicitly includes it on the command line
*/
if (!evsel && output[j].user_set && !output[j].wildcard_set) {
pr_err("%s events do not exist. "
"Remove corresponding -f option to proceed.\n",
event_type(j));
return -1;
}
if (evsel && output[j].fields &&
perf_evsel__check_attr(evsel, session))
return -1;
}
return 0;
@ -168,10 +280,7 @@ static void process_event(union perf_event *event __unused,
{
struct perf_event_attr *attr = &evsel->attr;
if (output_fields[attr->type] == 0)
return;
if (perf_session__check_attr(session, attr) < 0)
if (output[attr->type].fields == 0)
return;
print_sample_start(sample, thread, attr);
@ -451,6 +560,7 @@ static int parse_output_fields(const struct option *opt __used,
{
char *tok;
int i, imax = sizeof(all_output_options) / sizeof(struct output_option);
int j;
int rc = 0;
char *str = strdup(arg);
int type = -1;
@ -458,52 +568,99 @@ static int parse_output_fields(const struct option *opt __used,
if (!str)
return -ENOMEM;
tok = strtok(str, ":");
if (!tok) {
fprintf(stderr,
"Invalid field string - not prepended with type.");
return -EINVAL;
}
/* first word should state which event type user
* is specifying the fields
/* first word can state for which event type the user is specifying
* the fields. If no type exists, the specified fields apply to all
* event types found in the file minus the invalid fields for a type.
*/
if (!strcmp(tok, "hw"))
type = PERF_TYPE_HARDWARE;
else if (!strcmp(tok, "sw"))
type = PERF_TYPE_SOFTWARE;
else if (!strcmp(tok, "trace"))
type = PERF_TYPE_TRACEPOINT;
else {
fprintf(stderr, "Invalid event type in field string.");
return -EINVAL;
tok = strchr(str, ':');
if (tok) {
*tok = '\0';
tok++;
if (!strcmp(str, "hw"))
type = PERF_TYPE_HARDWARE;
else if (!strcmp(str, "sw"))
type = PERF_TYPE_SOFTWARE;
else if (!strcmp(str, "trace"))
type = PERF_TYPE_TRACEPOINT;
else if (!strcmp(str, "raw"))
type = PERF_TYPE_RAW;
else {
fprintf(stderr, "Invalid event type in field string.\n");
return -EINVAL;
}
if (output[type].user_set)
pr_warning("Overriding previous field request for %s events.\n",
event_type(type));
output[type].fields = 0;
output[type].user_set = true;
output[type].wildcard_set = false;
} else {
tok = str;
if (strlen(str) == 0) {
fprintf(stderr,
"Cannot set fields to 'none' for all event types.\n");
rc = -EINVAL;
goto out;
}
if (output_set_by_user())
pr_warning("Overriding previous field request for all events.\n");
for (j = 0; j < PERF_TYPE_MAX; ++j) {
output[j].fields = 0;
output[j].user_set = true;
output[j].wildcard_set = true;
}
}
output_fields[type] = 0;
while (1) {
tok = strtok(NULL, ",");
if (!tok)
break;
tok = strtok(tok, ",");
while (tok) {
for (i = 0; i < imax; ++i) {
if (strcmp(tok, all_output_options[i].str) == 0) {
output_fields[type] |= all_output_options[i].field;
if (strcmp(tok, all_output_options[i].str) == 0)
break;
}
}
if (i == imax) {
fprintf(stderr, "Invalid field requested.");
fprintf(stderr, "Invalid field requested.\n");
rc = -EINVAL;
break;
goto out;
}
if (type == -1) {
/* add user option to all events types for
* which it is valid
*/
for (j = 0; j < PERF_TYPE_MAX; ++j) {
if (output[j].invalid_fields & all_output_options[i].field) {
pr_warning("\'%s\' not valid for %s events. Ignoring.\n",
all_output_options[i].str, event_type(j));
} else
output[j].fields |= all_output_options[i].field;
}
} else {
if (output[type].invalid_fields & all_output_options[i].field) {
fprintf(stderr, "\'%s\' not valid for %s events.\n",
all_output_options[i].str, event_type(type));
rc = -EINVAL;
goto out;
}
output[type].fields |= all_output_options[i].field;
}
tok = strtok(NULL, ",");
}
if (type >= 0) {
if (output[type].fields == 0) {
pr_debug("No fields requested for %s type. "
"Events will not be displayed.\n", event_type(type));
}
}
if (output_fields[type] == 0) {
pr_debug("No fields requested for %s type. "
"Events will not be displayed\n", event_type(type));
}
output_set_by_user = true;
out:
free(str);
return rc;
}
@ -829,7 +986,7 @@ static const struct option options[] = {
OPT_STRING(0, "symfs", &symbol_conf.symfs, "directory",
"Look for files with symbols relative to this directory"),
OPT_CALLBACK('f', "fields", NULL, "str",
"comma separated output fields prepend with 'type:'. Valid types: hw,sw,trace. Fields: comm,tid,pid,time,cpu,event,trace,sym",
"comma separated output fields prepend with 'type:'. Valid types: hw,sw,trace,raw. Fields: comm,tid,pid,time,cpu,event,trace,sym",
parse_output_fields),
OPT_END()
@ -1020,7 +1177,7 @@ int cmd_script(int argc, const char **argv, const char *prefix __used)
struct stat perf_stat;
int input;
if (output_set_by_user) {
if (output_set_by_user()) {
fprintf(stderr,
"custom fields not supported for generated scripts");
return -1;
@ -1060,6 +1217,11 @@ int cmd_script(int argc, const char **argv, const char *prefix __used)
pr_debug("perf script started with script %s\n\n", script_name);
}
err = perf_session__check_output_opt(session);
if (err < 0)
goto out;
err = __cmd_script(session);
perf_session__delete(session);

573
tools/perf/builtin-stat.c
View file

@ -6,24 +6,28 @@
*
* Sample output:
$ perf stat ~/hackbench 10
Time: 0.104
$ perf stat ./hackbench 10
Performance counter stats for '/home/mingo/hackbench':
Time: 0.118
1255.538611 task clock ticks # 10.143 CPU utilization factor
54011 context switches # 0.043 M/sec
385 CPU migrations # 0.000 M/sec
17755 pagefaults # 0.014 M/sec
3808323185 CPU cycles # 3033.219 M/sec
1575111190 instructions # 1254.530 M/sec
17367895 cache references # 13.833 M/sec
7674421 cache misses # 6.112 M/sec
Performance counter stats for './hackbench 10':
Wall-clock time elapsed: 123.786620 msecs
1708.761321 task-clock # 11.037 CPUs utilized
41,190 context-switches # 0.024 M/sec
6,735 CPU-migrations # 0.004 M/sec
17,318 page-faults # 0.010 M/sec
5,205,202,243 cycles # 3.046 GHz
3,856,436,920 stalled-cycles-frontend # 74.09% frontend cycles idle
1,600,790,871 stalled-cycles-backend # 30.75% backend cycles idle
2,603,501,247 instructions # 0.50 insns per cycle
# 1.48 stalled cycles per insn
484,357,498 branches # 283.455 M/sec
6,388,934 branch-misses # 1.32% of all branches
0.154822978 seconds time elapsed
*
* Copyright (C) 2008, Red Hat Inc, Ingo Molnar <mingo@redhat.com>
* Copyright (C) 2008-2011, Red Hat Inc, Ingo Molnar <mingo@redhat.com>
*
* Improvements and fixes by:
*
@ -46,6 +50,7 @@
#include "util/evlist.h"
#include "util/evsel.h"
#include "util/debug.h"
#include "util/color.h"
#include "util/header.h"
#include "util/cpumap.h"
#include "util/thread.h"
@ -65,14 +70,107 @@ static struct perf_event_attr default_attrs[] = {
{ .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_PAGE_FAULTS },
{ .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CPU_CYCLES },
{ .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_STALLED_CYCLES_FRONTEND },
{ .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_STALLED_CYCLES_BACKEND },
{ .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_INSTRUCTIONS },
{ .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_BRANCH_INSTRUCTIONS },
{ .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_BRANCH_MISSES },
{ .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CACHE_REFERENCES },
{ .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CACHE_MISSES },
};
/*
* Detailed stats (-d), covering the L1 and last level data caches:
*/
static struct perf_event_attr detailed_attrs[] = {
{ .type = PERF_TYPE_HW_CACHE,
.config =
PERF_COUNT_HW_CACHE_L1D << 0 |
(PERF_COUNT_HW_CACHE_OP_READ << 8) |
(PERF_COUNT_HW_CACHE_RESULT_ACCESS << 16) },
{ .type = PERF_TYPE_HW_CACHE,
.config =
PERF_COUNT_HW_CACHE_L1D << 0 |
(PERF_COUNT_HW_CACHE_OP_READ << 8) |
(PERF_COUNT_HW_CACHE_RESULT_MISS << 16) },
{ .type = PERF_TYPE_HW_CACHE,
.config =
PERF_COUNT_HW_CACHE_LL << 0 |
(PERF_COUNT_HW_CACHE_OP_READ << 8) |
(PERF_COUNT_HW_CACHE_RESULT_ACCESS << 16) },
{ .type = PERF_TYPE_HW_CACHE,
.config =
PERF_COUNT_HW_CACHE_LL << 0 |
(PERF_COUNT_HW_CACHE_OP_READ << 8) |
(PERF_COUNT_HW_CACHE_RESULT_MISS << 16) },
};
/*
* Very detailed stats (-d -d), covering the instruction cache and the TLB caches:
*/
static struct perf_event_attr very_detailed_attrs[] = {
{ .type = PERF_TYPE_HW_CACHE,
.config =
PERF_COUNT_HW_CACHE_L1I << 0 |
(PERF_COUNT_HW_CACHE_OP_READ << 8) |
(PERF_COUNT_HW_CACHE_RESULT_ACCESS << 16) },
{ .type = PERF_TYPE_HW_CACHE,
.config =
PERF_COUNT_HW_CACHE_L1I << 0 |
(PERF_COUNT_HW_CACHE_OP_READ << 8) |
(PERF_COUNT_HW_CACHE_RESULT_MISS << 16) },
{ .type = PERF_TYPE_HW_CACHE,
.config =
PERF_COUNT_HW_CACHE_DTLB << 0 |
(PERF_COUNT_HW_CACHE_OP_READ << 8) |
(PERF_COUNT_HW_CACHE_RESULT_ACCESS << 16) },
{ .type = PERF_TYPE_HW_CACHE,
.config =
PERF_COUNT_HW_CACHE_DTLB << 0 |
(PERF_COUNT_HW_CACHE_OP_READ << 8) |
(PERF_COUNT_HW_CACHE_RESULT_MISS << 16) },
{ .type = PERF_TYPE_HW_CACHE,
.config =
PERF_COUNT_HW_CACHE_ITLB << 0 |
(PERF_COUNT_HW_CACHE_OP_READ << 8) |
(PERF_COUNT_HW_CACHE_RESULT_ACCESS << 16) },
{ .type = PERF_TYPE_HW_CACHE,
.config =
PERF_COUNT_HW_CACHE_ITLB << 0 |
(PERF_COUNT_HW_CACHE_OP_READ << 8) |
(PERF_COUNT_HW_CACHE_RESULT_MISS << 16) },
};
/*
* Very, very detailed stats (-d -d -d), adding prefetch events:
*/
static struct perf_event_attr very_very_detailed_attrs[] = {
{ .type = PERF_TYPE_HW_CACHE,
.config =
PERF_COUNT_HW_CACHE_L1D << 0 |
(PERF_COUNT_HW_CACHE_OP_PREFETCH << 8) |
(PERF_COUNT_HW_CACHE_RESULT_ACCESS << 16) },
{ .type = PERF_TYPE_HW_CACHE,
.config =
PERF_COUNT_HW_CACHE_L1D << 0 |
(PERF_COUNT_HW_CACHE_OP_PREFETCH << 8) |
(PERF_COUNT_HW_CACHE_RESULT_MISS << 16) },
};
struct perf_evlist *evsel_list;
static bool system_wide = false;
@ -86,6 +184,8 @@ static pid_t target_pid = -1;
static pid_t target_tid = -1;
static pid_t child_pid = -1;
static bool null_run = false;
static int detailed_run = 0;
static bool sync_run = false;
static bool big_num = true;
static int big_num_opt = -1;
static const char *cpu_list;
@ -156,7 +256,15 @@ static double stddev_stats(struct stats *stats)
struct stats runtime_nsecs_stats[MAX_NR_CPUS];
struct stats runtime_cycles_stats[MAX_NR_CPUS];
struct stats runtime_stalled_cycles_front_stats[MAX_NR_CPUS];
struct stats runtime_stalled_cycles_back_stats[MAX_NR_CPUS];
struct stats runtime_branches_stats[MAX_NR_CPUS];
struct stats runtime_cacherefs_stats[MAX_NR_CPUS];
struct stats runtime_l1_dcache_stats[MAX_NR_CPUS];
struct stats runtime_l1_icache_stats[MAX_NR_CPUS];
struct stats runtime_ll_cache_stats[MAX_NR_CPUS];
struct stats runtime_itlb_cache_stats[MAX_NR_CPUS];
struct stats runtime_dtlb_cache_stats[MAX_NR_CPUS];
struct stats walltime_nsecs_stats;
static int create_perf_stat_counter(struct perf_evsel *evsel)
@ -192,6 +300,37 @@ static inline int nsec_counter(struct perf_evsel *evsel)
return 0;
}
/*
* Update various tracking values we maintain to print
* more semantic information such as miss/hit ratios,
* instruction rates, etc:
*/
static void update_shadow_stats(struct perf_evsel *counter, u64 *count)
{
if (perf_evsel__match(counter, SOFTWARE, SW_TASK_CLOCK))
update_stats(&runtime_nsecs_stats[0], count[0]);
else if (perf_evsel__match(counter, HARDWARE, HW_CPU_CYCLES))
update_stats(&runtime_cycles_stats[0], count[0]);
else if (perf_evsel__match(counter, HARDWARE, HW_STALLED_CYCLES_FRONTEND))
update_stats(&runtime_stalled_cycles_front_stats[0], count[0]);
else if (perf_evsel__match(counter, HARDWARE, HW_STALLED_CYCLES_BACKEND))
update_stats(&runtime_stalled_cycles_back_stats[0], count[0]);
else if (perf_evsel__match(counter, HARDWARE, HW_BRANCH_INSTRUCTIONS))
update_stats(&runtime_branches_stats[0], count[0]);
else if (perf_evsel__match(counter, HARDWARE, HW_CACHE_REFERENCES))
update_stats(&runtime_cacherefs_stats[0], count[0]);
else if (perf_evsel__match(counter, HW_CACHE, HW_CACHE_L1D))
update_stats(&runtime_l1_dcache_stats[0], count[0]);
else if (perf_evsel__match(counter, HW_CACHE, HW_CACHE_L1I))
update_stats(&runtime_l1_icache_stats[0], count[0]);
else if (perf_evsel__match(counter, HW_CACHE, HW_CACHE_LL))
update_stats(&runtime_ll_cache_stats[0], count[0]);
else if (perf_evsel__match(counter, HW_CACHE, HW_CACHE_DTLB))
update_stats(&runtime_dtlb_cache_stats[0], count[0]);
else if (perf_evsel__match(counter, HW_CACHE, HW_CACHE_ITLB))
update_stats(&runtime_itlb_cache_stats[0], count[0]);
}
/*
* Read out the results of a single counter:
* aggregate counts across CPUs in system-wide mode
@ -217,12 +356,7 @@ static int read_counter_aggr(struct perf_evsel *counter)
/*
* Save the full runtime - to allow normalization during printout:
*/
if (perf_evsel__match(counter, SOFTWARE, SW_TASK_CLOCK))
update_stats(&runtime_nsecs_stats[0], count[0]);
if (perf_evsel__match(counter, HARDWARE, HW_CPU_CYCLES))
update_stats(&runtime_cycles_stats[0], count[0]);
if (perf_evsel__match(counter, HARDWARE, HW_BRANCH_INSTRUCTIONS))
update_stats(&runtime_branches_stats[0], count[0]);
update_shadow_stats(counter, count);
return 0;
}
@ -242,12 +376,7 @@ static int read_counter(struct perf_evsel *counter)
count = counter->counts->cpu[cpu].values;
if (perf_evsel__match(counter, SOFTWARE, SW_TASK_CLOCK))
update_stats(&runtime_nsecs_stats[cpu], count[0]);
if (perf_evsel__match(counter, HARDWARE, HW_CPU_CYCLES))
update_stats(&runtime_cycles_stats[cpu], count[0]);
if (perf_evsel__match(counter, HARDWARE, HW_BRANCH_INSTRUCTIONS))
update_stats(&runtime_branches_stats[cpu], count[0]);
update_shadow_stats(counter, count);
}
return 0;
@ -315,13 +444,18 @@ static int run_perf_stat(int argc __used, const char **argv)
list_for_each_entry(counter, &evsel_list->entries, node) {
if (create_perf_stat_counter(counter) < 0) {
if (errno == -EPERM || errno == -EACCES) {
if (errno == EINVAL || errno == ENOSYS || errno == ENOENT) {
if (verbose)
ui__warning("%s event is not supported by the kernel.\n",
event_name(counter));
continue;
}
if (errno == EPERM || errno == EACCES) {
error("You may not have permission to collect %sstats.\n"
"\t Consider tweaking"
" /proc/sys/kernel/perf_event_paranoid or running as root.",
system_wide ? "system-wide " : "");
} else if (errno == ENOENT) {
error("%s event is not supported. ", event_name(counter));
} else {
error("open_counter returned with %d (%s). "
"/bin/dmesg may provide additional information.\n",
@ -372,6 +506,16 @@ static int run_perf_stat(int argc __used, const char **argv)
return WEXITSTATUS(status);
}
static void print_noise_pct(double total, double avg)
{
double pct = 0.0;
if (avg)
pct = 100.0*total/avg;
fprintf(stderr, " ( +-%6.2f%% )", pct);
}
static void print_noise(struct perf_evsel *evsel, double avg)
{
struct perf_stat *ps;
@ -380,15 +524,14 @@ static void print_noise(struct perf_evsel *evsel, double avg)
return;
ps = evsel->priv;
fprintf(stderr, " ( +- %7.3f%% )",
100 * stddev_stats(&ps->res_stats[0]) / avg);
print_noise_pct(stddev_stats(&ps->res_stats[0]), avg);
}
static void nsec_printout(int cpu, struct perf_evsel *evsel, double avg)
{
double msecs = avg / 1e6;
char cpustr[16] = { '\0', };
const char *fmt = csv_output ? "%s%.6f%s%s" : "%s%18.6f%s%-24s";
const char *fmt = csv_output ? "%s%.6f%s%s" : "%s%18.6f%s%-25s";
if (no_aggr)
sprintf(cpustr, "CPU%*d%s",
@ -404,8 +547,191 @@ static void nsec_printout(int cpu, struct perf_evsel *evsel, double avg)
return;
if (perf_evsel__match(evsel, SOFTWARE, SW_TASK_CLOCK))
fprintf(stderr, " # %10.3f CPUs ",
avg / avg_stats(&walltime_nsecs_stats));
fprintf(stderr, " # %8.3f CPUs utilized ", avg / avg_stats(&walltime_nsecs_stats));
}
static void print_stalled_cycles_frontend(int cpu, struct perf_evsel *evsel __used, double avg)
{
double total, ratio = 0.0;
const char *color;
total = avg_stats(&runtime_cycles_stats[cpu]);
if (total)
ratio = avg / total * 100.0;
color = PERF_COLOR_NORMAL;
if (ratio > 50.0)
color = PERF_COLOR_RED;
else if (ratio > 30.0)
color = PERF_COLOR_MAGENTA;
else if (ratio > 10.0)
color = PERF_COLOR_YELLOW;
fprintf(stderr, " # ");
color_fprintf(stderr, color, "%6.2f%%", ratio);
fprintf(stderr, " frontend cycles idle ");
}
static void print_stalled_cycles_backend(int cpu, struct perf_evsel *evsel __used, double avg)
{
double total, ratio = 0.0;
const char *color;
total = avg_stats(&runtime_cycles_stats[cpu]);
if (total)
ratio = avg / total * 100.0;
color = PERF_COLOR_NORMAL;
if (ratio > 75.0)
color = PERF_COLOR_RED;
else if (ratio > 50.0)
color = PERF_COLOR_MAGENTA;
else if (ratio > 20.0)
color = PERF_COLOR_YELLOW;
fprintf(stderr, " # ");
color_fprintf(stderr, color, "%6.2f%%", ratio);
fprintf(stderr, " backend cycles idle ");
}
static void print_branch_misses(int cpu, struct perf_evsel *evsel __used, double avg)
{
double total, ratio = 0.0;
const char *color;
total = avg_stats(&runtime_branches_stats[cpu]);
if (total)
ratio = avg / total * 100.0;
color = PERF_COLOR_NORMAL;
if (ratio > 20.0)
color = PERF_COLOR_RED;
else if (ratio > 10.0)
color = PERF_COLOR_MAGENTA;
else if (ratio > 5.0)
color = PERF_COLOR_YELLOW;
fprintf(stderr, " # ");
color_fprintf(stderr, color, "%6.2f%%", ratio);
fprintf(stderr, " of all branches ");
}
static void print_l1_dcache_misses(int cpu, struct perf_evsel *evsel __used, double avg)
{
double total, ratio = 0.0;
const char *color;
total = avg_stats(&runtime_l1_dcache_stats[cpu]);
if (total)
ratio = avg / total * 100.0;
color = PERF_COLOR_NORMAL;
if (ratio > 20.0)
color = PERF_COLOR_RED;
else if (ratio > 10.0)
color = PERF_COLOR_MAGENTA;
else if (ratio > 5.0)
color = PERF_COLOR_YELLOW;
fprintf(stderr, " # ");
color_fprintf(stderr, color, "%6.2f%%", ratio);
fprintf(stderr, " of all L1-dcache hits ");
}
static void print_l1_icache_misses(int cpu, struct perf_evsel *evsel __used, double avg)
{
double total, ratio = 0.0;
const char *color;
total = avg_stats(&runtime_l1_icache_stats[cpu]);
if (total)
ratio = avg / total * 100.0;
color = PERF_COLOR_NORMAL;
if (ratio > 20.0)
color = PERF_COLOR_RED;
else if (ratio > 10.0)
color = PERF_COLOR_MAGENTA;
else if (ratio > 5.0)
color = PERF_COLOR_YELLOW;
fprintf(stderr, " # ");
color_fprintf(stderr, color, "%6.2f%%", ratio);
fprintf(stderr, " of all L1-icache hits ");
}
static void print_dtlb_cache_misses(int cpu, struct perf_evsel *evsel __used, double avg)
{
double total, ratio = 0.0;
const char *color;
total = avg_stats(&runtime_dtlb_cache_stats[cpu]);
if (total)
ratio = avg / total * 100.0;
color = PERF_COLOR_NORMAL;
if (ratio > 20.0)
color = PERF_COLOR_RED;
else if (ratio > 10.0)
color = PERF_COLOR_MAGENTA;
else if (ratio > 5.0)
color = PERF_COLOR_YELLOW;
fprintf(stderr, " # ");
color_fprintf(stderr, color, "%6.2f%%", ratio);
fprintf(stderr, " of all dTLB cache hits ");
}
static void print_itlb_cache_misses(int cpu, struct perf_evsel *evsel __used, double avg)
{
double total, ratio = 0.0;
const char *color;
total = avg_stats(&runtime_itlb_cache_stats[cpu]);
if (total)
ratio = avg / total * 100.0;
color = PERF_COLOR_NORMAL;
if (ratio > 20.0)
color = PERF_COLOR_RED;
else if (ratio > 10.0)
color = PERF_COLOR_MAGENTA;
else if (ratio > 5.0)
color = PERF_COLOR_YELLOW;
fprintf(stderr, " # ");
color_fprintf(stderr, color, "%6.2f%%", ratio);
fprintf(stderr, " of all iTLB cache hits ");
}
static void print_ll_cache_misses(int cpu, struct perf_evsel *evsel __used, double avg)
{
double total, ratio = 0.0;
const char *color;
total = avg_stats(&runtime_ll_cache_stats[cpu]);
if (total)
ratio = avg / total * 100.0;
color = PERF_COLOR_NORMAL;
if (ratio > 20.0)
color = PERF_COLOR_RED;
else if (ratio > 10.0)
color = PERF_COLOR_MAGENTA;
else if (ratio > 5.0)
color = PERF_COLOR_YELLOW;
fprintf(stderr, " # ");
color_fprintf(stderr, color, "%6.2f%%", ratio);
fprintf(stderr, " of all LL-cache hits ");
}
static void abs_printout(int cpu, struct perf_evsel *evsel, double avg)
@ -417,9 +743,9 @@ static void abs_printout(int cpu, struct perf_evsel *evsel, double avg)
if (csv_output)
fmt = "%s%.0f%s%s";
else if (big_num)
fmt = "%s%'18.0f%s%-24s";
fmt = "%s%'18.0f%s%-25s";
else
fmt = "%s%18.0f%s%-24s";
fmt = "%s%18.0f%s%-25s";
if (no_aggr)
sprintf(cpustr, "CPU%*d%s",
@ -442,23 +768,83 @@ static void abs_printout(int cpu, struct perf_evsel *evsel, double avg)
if (total)
ratio = avg / total;
fprintf(stderr, " # %10.3f IPC ", ratio);
fprintf(stderr, " # %5.2f insns per cycle ", ratio);
total = avg_stats(&runtime_stalled_cycles_front_stats[cpu]);
total = max(total, avg_stats(&runtime_stalled_cycles_back_stats[cpu]));
if (total && avg) {
ratio = total / avg;
fprintf(stderr, "\n # %5.2f stalled cycles per insn", ratio);
}
} else if (perf_evsel__match(evsel, HARDWARE, HW_BRANCH_MISSES) &&
runtime_branches_stats[cpu].n != 0) {
total = avg_stats(&runtime_branches_stats[cpu]);
print_branch_misses(cpu, evsel, avg);
} else if (
evsel->attr.type == PERF_TYPE_HW_CACHE &&
evsel->attr.config == ( PERF_COUNT_HW_CACHE_L1D |
((PERF_COUNT_HW_CACHE_OP_READ) << 8) |
((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16)) &&
runtime_l1_dcache_stats[cpu].n != 0) {
print_l1_dcache_misses(cpu, evsel, avg);
} else if (
evsel->attr.type == PERF_TYPE_HW_CACHE &&
evsel->attr.config == ( PERF_COUNT_HW_CACHE_L1I |
((PERF_COUNT_HW_CACHE_OP_READ) << 8) |
((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16)) &&
runtime_l1_icache_stats[cpu].n != 0) {
print_l1_icache_misses(cpu, evsel, avg);
} else if (
evsel->attr.type == PERF_TYPE_HW_CACHE &&
evsel->attr.config == ( PERF_COUNT_HW_CACHE_DTLB |
((PERF_COUNT_HW_CACHE_OP_READ) << 8) |
((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16)) &&
runtime_dtlb_cache_stats[cpu].n != 0) {
print_dtlb_cache_misses(cpu, evsel, avg);
} else if (
evsel->attr.type == PERF_TYPE_HW_CACHE &&
evsel->attr.config == ( PERF_COUNT_HW_CACHE_ITLB |
((PERF_COUNT_HW_CACHE_OP_READ) << 8) |
((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16)) &&
runtime_itlb_cache_stats[cpu].n != 0) {
print_itlb_cache_misses(cpu, evsel, avg);
} else if (
evsel->attr.type == PERF_TYPE_HW_CACHE &&
evsel->attr.config == ( PERF_COUNT_HW_CACHE_LL |
((PERF_COUNT_HW_CACHE_OP_READ) << 8) |
((PERF_COUNT_HW_CACHE_RESULT_MISS) << 16)) &&
runtime_ll_cache_stats[cpu].n != 0) {
print_ll_cache_misses(cpu, evsel, avg);
} else if (perf_evsel__match(evsel, HARDWARE, HW_CACHE_MISSES) &&
runtime_cacherefs_stats[cpu].n != 0) {
total = avg_stats(&runtime_cacherefs_stats[cpu]);
if (total)
ratio = avg * 100 / total;
fprintf(stderr, " # %10.3f %% ", ratio);
fprintf(stderr, " # %8.3f %% of all cache refs ", ratio);
} else if (perf_evsel__match(evsel, HARDWARE, HW_STALLED_CYCLES_FRONTEND)) {
print_stalled_cycles_frontend(cpu, evsel, avg);
} else if (perf_evsel__match(evsel, HARDWARE, HW_STALLED_CYCLES_BACKEND)) {
print_stalled_cycles_backend(cpu, evsel, avg);
} else if (perf_evsel__match(evsel, HARDWARE, HW_CPU_CYCLES)) {
total = avg_stats(&runtime_nsecs_stats[cpu]);
if (total)
ratio = 1.0 * avg / total;
fprintf(stderr, " # %8.3f GHz ", ratio);
} else if (runtime_nsecs_stats[cpu].n != 0) {
total = avg_stats(&runtime_nsecs_stats[cpu]);
if (total)
ratio = 1000.0 * avg / total;
fprintf(stderr, " # %10.3f M/sec", ratio);
fprintf(stderr, " # %8.3f M/sec ", ratio);
} else {
fprintf(stderr, " ");
}
}
@ -505,8 +891,7 @@ static void print_counter_aggr(struct perf_evsel *counter)
avg_enabled = avg_stats(&ps->res_stats[1]);
avg_running = avg_stats(&ps->res_stats[2]);
fprintf(stderr, " (scaled from %.2f%%)",
100 * avg_running / avg_enabled);
fprintf(stderr, " [%5.2f%%]", 100 * avg_running / avg_enabled);
}
fprintf(stderr, "\n");
}
@ -548,10 +933,8 @@ static void print_counter(struct perf_evsel *counter)
if (!csv_output) {
print_noise(counter, 1.0);
if (run != ena) {
fprintf(stderr, " (scaled from %.2f%%)",
100.0 * run / ena);
}
if (run != ena)
fprintf(stderr, " (%.2f%%)", 100.0 * run / ena);
}
fputc('\n', stderr);
}
@ -591,13 +974,14 @@ static void print_stat(int argc, const char **argv)
}
if (!csv_output) {
fprintf(stderr, "\n");
fprintf(stderr, " %18.9f seconds time elapsed",
if (!null_run)
fprintf(stderr, "\n");
fprintf(stderr, " %17.9f seconds time elapsed",
avg_stats(&walltime_nsecs_stats)/1e9);
if (run_count > 1) {
fprintf(stderr, " ( +- %7.3f%% )",
100*stddev_stats(&walltime_nsecs_stats) /
avg_stats(&walltime_nsecs_stats));
fprintf(stderr, " ");
print_noise_pct(stddev_stats(&walltime_nsecs_stats),
avg_stats(&walltime_nsecs_stats));
}
fprintf(stderr, "\n\n");
}
@ -659,6 +1043,10 @@ static const struct option options[] = {
"repeat command and print average + stddev (max: 100)"),
OPT_BOOLEAN('n', "null", &null_run,
"null run - dont start any counters"),
OPT_INCR('d', "detailed", &detailed_run,
"detailed run - start a lot of events"),
OPT_BOOLEAN('S', "sync", &sync_run,
"call sync() before starting a run"),
OPT_CALLBACK_NOOPT('B', "big-num", NULL, NULL,
"print large numbers with thousands\' separators",
stat__set_big_num),
@ -674,6 +1062,70 @@ static const struct option options[] = {
OPT_END()
};
/*
* Add default attributes, if there were no attributes specified or
* if -d/--detailed, -d -d or -d -d -d is used:
*/
static int add_default_attributes(void)
{
struct perf_evsel *pos;
size_t attr_nr = 0;
size_t c;
/* Set attrs if no event is selected and !null_run: */
if (null_run)
return 0;
if (!evsel_list->nr_entries) {
for (c = 0; c < ARRAY_SIZE(default_attrs); c++) {
pos = perf_evsel__new(default_attrs + c, c + attr_nr);
if (pos == NULL)
return -1;
perf_evlist__add(evsel_list, pos);
}
attr_nr += c;
}
/* Detailed events get appended to the event list: */
if (detailed_run < 1)
return 0;
/* Append detailed run extra attributes: */
for (c = 0; c < ARRAY_SIZE(detailed_attrs); c++) {
pos = perf_evsel__new(detailed_attrs + c, c + attr_nr);
if (pos == NULL)
return -1;
perf_evlist__add(evsel_list, pos);
}
attr_nr += c;
if (detailed_run < 2)
return 0;
/* Append very detailed run extra attributes: */
for (c = 0; c < ARRAY_SIZE(very_detailed_attrs); c++) {
pos = perf_evsel__new(very_detailed_attrs + c, c + attr_nr);
if (pos == NULL)
return -1;
perf_evlist__add(evsel_list, pos);
}
if (detailed_run < 3)
return 0;
/* Append very, very detailed run extra attributes: */
for (c = 0; c < ARRAY_SIZE(very_very_detailed_attrs); c++) {
pos = perf_evsel__new(very_very_detailed_attrs + c, c + attr_nr);
if (pos == NULL)
return -1;
perf_evlist__add(evsel_list, pos);
}
return 0;
}
int cmd_stat(int argc, const char **argv, const char *prefix __used)
{
struct perf_evsel *pos;
@ -719,17 +1171,8 @@ int cmd_stat(int argc, const char **argv, const char *prefix __used)
usage_with_options(stat_usage, options);
}
/* Set attrs and nr_counters if no event is selected and !null_run */
if (!null_run && !evsel_list->nr_entries) {
size_t c;
for (c = 0; c < ARRAY_SIZE(default_attrs); ++c) {
pos = perf_evsel__new(&default_attrs[c], c);
if (pos == NULL)
goto out;
perf_evlist__add(evsel_list, pos);
}
}
if (add_default_attributes())
goto out;
if (target_pid != -1)
target_tid = target_pid;
@ -773,6 +1216,10 @@ int cmd_stat(int argc, const char **argv, const char *prefix __used)
for (run_idx = 0; run_idx < run_count; run_idx++) {
if (run_count != 1 && verbose)
fprintf(stderr, "[ perf stat: executing run #%d ... ]\n", run_idx + 1);
if (sync_run)
sync();
status = run_perf_stat(argc, argv);
}

16
tools/perf/feature-tests.mak → tools/perf/config/feature-tests.mak
View file

@ -79,9 +79,15 @@ endef
endif
ifndef NO_LIBPYTHON
define SOURCE_PYTHON_VERSION
#include <Python.h>
#if PY_VERSION_HEX >= 0x03000000
#error
#endif
int main(void){}
endef
define SOURCE_PYTHON_EMBED
#include <Python.h>
int main(void)
{
Py_Initialize();
@ -120,11 +126,3 @@ int main(void)
return 0;
}
endef
# try-cc
# Usage: option = $(call try-cc, source-to-build, cc-options)
try-cc = $(shell sh -c \
'TMP="$(OUTPUT)$(TMPOUT).$$$$"; \
echo "$(1)" | \
$(CC) -x c - $(2) -o "$$TMP" > /dev/null 2>&1 && echo y; \
rm -f "$$TMP"')

188
tools/perf/config/utilities.mak Normal file
View file

@ -0,0 +1,188 @@
# This allows us to work with the newline character:
define newline
endef
newline := $(newline)
# nl-escape
#
# Usage: escape = $(call nl-escape[,escape])
#
# This is used as the common way to specify
# what should replace a newline when escaping
# newlines; the default is a bizarre string.
#
nl-escape = $(or $(1),m822df3020w6a44id34bt574ctac44eb9f4n)
# escape-nl
#
# Usage: escaped-text = $(call escape-nl,text[,escape])
#
# GNU make's $(shell ...) function converts to a
# single space each newline character in the output
# produced during the expansion; this may not be
# desirable.
#
# The only solution is to change each newline into
# something that won't be converted, so that the
# information can be recovered later with
# $(call unescape-nl...)
#
escape-nl = $(subst $(newline),$(call nl-escape,$(2)),$(1))
# unescape-nl
#
# Usage: text = $(call unescape-nl,escaped-text[,escape])
#
# See escape-nl.
#
unescape-nl = $(subst $(call nl-escape,$(2)),$(newline),$(1))
# shell-escape-nl
#
# Usage: $(shell some-command | $(call shell-escape-nl[,escape]))
#
# Use this to escape newlines from within a shell call;
# the default escape is a bizarre string.
#
# NOTE: The escape is used directly as a string constant
# in an `awk' program that is delimited by shell
# single-quotes, so be wary of the characters
# that are chosen.
#
define shell-escape-nl
awk 'NR==1 {t=$$0} NR>1 {t=t "$(nl-escape)" $$0} END {printf t}'
endef
# shell-unescape-nl
#
# Usage: $(shell some-command | $(call shell-unescape-nl[,escape]))
#
# Use this to unescape newlines from within a shell call;
# the default escape is a bizarre string.
#
# NOTE: The escape is used directly as an extended regular
# expression constant in an `awk' program that is
# delimited by shell single-quotes, so be wary
# of the characters that are chosen.
#
# (The bash shell has a bug where `{gsub(...),...}' is
# misinterpreted as a brace expansion; this can be
# overcome by putting a space between `{' and `gsub').
#
define shell-unescape-nl
awk 'NR==1 {t=$$0} NR>1 {t=t "\n" $$0} END { gsub(/$(nl-escape)/,"\n",t); printf t }'
endef
# escape-for-shell-sq
#
# Usage: embeddable-text = $(call escape-for-shell-sq,text)
#
# This function produces text that is suitable for
# embedding in a shell string that is delimited by
# single-quotes.
#
escape-for-shell-sq = $(subst ','\'',$(1))
# shell-sq
#
# Usage: single-quoted-and-escaped-text = $(call shell-sq,text)
#
shell-sq = '$(escape-for-shell-sq)'
# shell-wordify
#
# Usage: wordified-text = $(call shell-wordify,text)
#
# For instance:
#
# |define text
# |hello
# |world
# |endef
# |
# |target:
# | echo $(call shell-wordify,$(text))
#
# At least GNU make gets confused by expanding a newline
# within the context of a command line of a makefile rule
# (this is in constrast to a `$(shell ...)' function call,
# which can handle it just fine).
#
# This function avoids the problem by producing a string
# that works as a shell word, regardless of whether or
# not it contains a newline.
#
# If the text to be wordified contains a newline, then
# an intrictate shell command substitution is constructed
# to render the text as a single line; when the shell
# processes the resulting escaped text, it transforms
# it into the original unescaped text.
#
# If the text does not contain a newline, then this function
# produces the same results as the `$(shell-sq)' function.
#
shell-wordify = $(if $(findstring $(newline),$(1)),$(_sw-esc-nl),$(shell-sq))
define _sw-esc-nl
"$$(echo $(call escape-nl,$(shell-sq),$(2)) | $(call shell-unescape-nl,$(2)))"
endef
# is-absolute
#
# Usage: bool-value = $(call is-absolute,path)
#
is-absolute = $(shell echo $(shell-sq) | grep ^/ -q && echo y)
# lookup
#
# Usage: absolute-executable-path-or-empty = $(call lookup,path)
#
# (It's necessary to use `sh -c' because GNU make messes up by
# trying too hard and getting things wrong).
#
lookup = $(call unescape-nl,$(shell sh -c $(_l-sh)))
_l-sh = $(call shell-sq,command -v $(shell-sq) | $(call shell-escape-nl,))
# is-executable
#
# Usage: bool-value = $(call is-executable,path)
#
# (It's necessary to use `sh -c' because GNU make messes up by
# trying too hard and getting things wrong).
#
is-executable = $(call _is-executable-helper,$(shell-sq))
_is-executable-helper = $(shell sh -c $(_is-executable-sh))
_is-executable-sh = $(call shell-sq,test -f $(1) -a -x $(1) && echo y)
# get-executable
#
# Usage: absolute-executable-path-or-empty = $(call get-executable,path)
#
# The goal is to get an absolute path for an executable;
# the `command -v' is defined by POSIX, but it's not
# necessarily very portable, so it's only used if
# relative path resolution is requested, as determined
# by the presence of a leading `/'.
#
get-executable = $(if $(1),$(if $(is-absolute),$(_ge-abspath),$(lookup)))
_ge-abspath = $(if $(is-executable),$(1))
# get-supplied-or-default-executable
#
# Usage: absolute-executable-path-or-empty = $(call get-executable-or-default,variable,default)
#
define get-executable-or-default
$(if $($(1)),$(call _ge_attempt,$($(1)),$(1)),$(call _ge_attempt,$(2)))
endef
_ge_attempt = $(or $(get-executable),$(_gea_warn),$(call _gea_err,$(2)))
_gea_warn = $(warning The path '$(1)' is not executable.)
_gea_err = $(if $(1),$(error Please set '$(1)' appropriately))
# try-cc
# Usage: option = $(call try-cc, source-to-build, cc-options)
try-cc = $(shell sh -c \
'TMP="$(OUTPUT)$(TMPOUT).$$$$"; \
echo "$(1)" | \
$(CC) -x c - $(2) -o "$$TMP" > /dev/null 2>&1 && echo y; \
rm -f "$$TMP"')

8
tools/perf/util/include/asm/alternative-asm.h Normal file
View file

@ -0,0 +1,8 @@
#ifndef _PERF_ASM_ALTERNATIVE_ASM_H
#define _PERF_ASM_ALTERNATIVE_ASM_H
/* Just disable it so we can build arch/x86/lib/memcpy_64.S for perf bench: */
#define altinstruction_entry #
#endif

121
tools/perf/util/parse-events.c
View file

@ -31,34 +31,36 @@ char debugfs_path[MAXPATHLEN];
#define CSW(x) .type = PERF_TYPE_SOFTWARE, .config = PERF_COUNT_SW_##x
static struct event_symbol event_symbols[] = {
{ CHW(CPU_CYCLES), "cpu-cycles", "cycles" },
{ CHW(INSTRUCTIONS), "instructions", "" },
{ CHW(CACHE_REFERENCES), "cache-references", "" },
{ CHW(CACHE_MISSES), "cache-misses", "" },
{ CHW(BRANCH_INSTRUCTIONS), "branch-instructions", "branches" },
{ CHW(BRANCH_MISSES), "branch-misses", "" },
{ CHW(BUS_CYCLES), "bus-cycles", "" },
{ CHW(CPU_CYCLES), "cpu-cycles", "cycles" },
{ CHW(STALLED_CYCLES_FRONTEND), "stalled-cycles-frontend", "idle-cycles-frontend" },
{ CHW(STALLED_CYCLES_BACKEND), "stalled-cycles-backend", "idle-cycles-backend" },
{ CHW(INSTRUCTIONS), "instructions", "" },
{ CHW(CACHE_REFERENCES), "cache-references", "" },
{ CHW(CACHE_MISSES), "cache-misses", "" },
{ CHW(BRANCH_INSTRUCTIONS), "branch-instructions", "branches" },
{ CHW(BRANCH_MISSES), "branch-misses", "" },
{ CHW(BUS_CYCLES), "bus-cycles", "" },
{ CSW(CPU_CLOCK), "cpu-clock", "" },
{ CSW(TASK_CLOCK), "task-clock", "" },
{ CSW(PAGE_FAULTS), "page-faults", "faults" },
{ CSW(PAGE_FAULTS_MIN), "minor-faults", "" },
{ CSW(PAGE_FAULTS_MAJ), "major-faults", "" },
{ CSW(CONTEXT_SWITCHES), "context-switches", "cs" },
{ CSW(CPU_MIGRATIONS), "cpu-migrations", "migrations" },
{ CSW(ALIGNMENT_FAULTS), "alignment-faults", "" },
{ CSW(EMULATION_FAULTS), "emulation-faults", "" },
{ CSW(CPU_CLOCK), "cpu-clock", "" },
{ CSW(TASK_CLOCK), "task-clock", "" },
{ CSW(PAGE_FAULTS), "page-faults", "faults" },
{ CSW(PAGE_FAULTS_MIN), "minor-faults", "" },
{ CSW(PAGE_FAULTS_MAJ), "major-faults", "" },
{ CSW(CONTEXT_SWITCHES), "context-switches", "cs" },
{ CSW(CPU_MIGRATIONS), "cpu-migrations", "migrations" },
{ CSW(ALIGNMENT_FAULTS), "alignment-faults", "" },
{ CSW(EMULATION_FAULTS), "emulation-faults", "" },
};
#define __PERF_EVENT_FIELD(config, name) \
((config & PERF_EVENT_##name##_MASK) >> PERF_EVENT_##name##_SHIFT)
#define PERF_EVENT_RAW(config) __PERF_EVENT_FIELD(config, RAW)
#define PERF_EVENT_RAW(config) __PERF_EVENT_FIELD(config, RAW)
#define PERF_EVENT_CONFIG(config) __PERF_EVENT_FIELD(config, CONFIG)
#define PERF_EVENT_TYPE(config) __PERF_EVENT_FIELD(config, TYPE)
#define PERF_EVENT_TYPE(config) __PERF_EVENT_FIELD(config, TYPE)
#define PERF_EVENT_ID(config) __PERF_EVENT_FIELD(config, EVENT)
static const char *hw_event_names[] = {
static const char *hw_event_names[PERF_COUNT_HW_MAX] = {
"cycles",
"instructions",
"cache-references",
@ -66,11 +68,13 @@ static const char *hw_event_names[] = {
"branches",
"branch-misses",
"bus-cycles",
"stalled-cycles-frontend",
"stalled-cycles-backend",
};
static const char *sw_event_names[] = {
"cpu-clock-msecs",
"task-clock-msecs",
static const char *sw_event_names[PERF_COUNT_SW_MAX] = {
"cpu-clock",
"task-clock",
"page-faults",
"context-switches",
"CPU-migrations",
@ -307,7 +311,7 @@ const char *__event_name(int type, u64 config)
switch (type) {
case PERF_TYPE_HARDWARE:
if (config < PERF_COUNT_HW_MAX)
if (config < PERF_COUNT_HW_MAX && hw_event_names[config])
return hw_event_names[config];
return "unknown-hardware";
@ -333,7 +337,7 @@ const char *__event_name(int type, u64 config)
}
case PERF_TYPE_SOFTWARE:
if (config < PERF_COUNT_SW_MAX)
if (config < PERF_COUNT_SW_MAX && sw_event_names[config])
return sw_event_names[config];
return "unknown-software";
@ -648,13 +652,15 @@ static int check_events(const char *str, unsigned int i)
int n;
n = strlen(event_symbols[i].symbol);
if (!strncmp(str, event_symbols[i].symbol, n))
if (!strncasecmp(str, event_symbols[i].symbol, n))
return n;
n = strlen(event_symbols[i].alias);
if (n)
if (!strncmp(str, event_symbols[i].alias, n))
if (n) {
if (!strncasecmp(str, event_symbols[i].alias, n))
return n;
}
return 0;
}
@ -718,15 +724,22 @@ parse_numeric_event(const char **strp, struct perf_event_attr *attr)
return EVT_FAILED;
}
static enum event_result
static int
parse_event_modifier(const char **strp, struct perf_event_attr *attr)
{
const char *str = *strp;
int exclude = 0;
int eu = 0, ek = 0, eh = 0, precise = 0;
if (*str++ != ':')
if (!*str)
return 0;
if (*str == ',')
return 0;
if (*str++ != ':')
return -1;
while (*str) {
if (*str == 'u') {
if (!exclude)
@ -747,14 +760,16 @@ parse_event_modifier(const char **strp, struct perf_event_attr *attr)
++str;
}
if (str >= *strp + 2) {
*strp = str;
attr->exclude_user = eu;
attr->exclude_kernel = ek;
attr->exclude_hv = eh;
attr->precise_ip = precise;
return 1;
}
if (str < *strp + 2)
return -1;
*strp = str;
attr->exclude_user = eu;
attr->exclude_kernel = ek;
attr->exclude_hv = eh;
attr->precise_ip = precise;
return 0;
}
@ -797,7 +812,12 @@ parse_event_symbols(const struct option *opt, const char **str,
return EVT_FAILED;
modifier:
parse_event_modifier(str, attr);
if (parse_event_modifier(str, attr) < 0) {
fprintf(stderr, "invalid event modifier: '%s'\n", *str);
fprintf(stderr, "Run 'perf list' for a list of valid events and modifiers\n");
return EVT_FAILED;
}
return ret;
}
@ -912,7 +932,7 @@ void print_tracepoint_events(const char *subsys_glob, const char *event_glob)
snprintf(evt_path, MAXPATHLEN, "%s:%s",
sys_dirent.d_name, evt_dirent.d_name);
printf(" %-42s [%s]\n", evt_path,
printf(" %-50s [%s]\n", evt_path,
event_type_descriptors[PERF_TYPE_TRACEPOINT]);
}
closedir(evt_dir);
@ -977,7 +997,7 @@ void print_events_type(u8 type)
else
snprintf(name, sizeof(name), "%s", syms->symbol);
printf(" %-42s [%s]\n", name,
printf(" %-50s [%s]\n", name,
event_type_descriptors[type]);
}
}
@ -995,11 +1015,10 @@ int print_hwcache_events(const char *event_glob)
for (i = 0; i < PERF_COUNT_HW_CACHE_RESULT_MAX; i++) {
char *name = event_cache_name(type, op, i);
if (event_glob != NULL &&
!strglobmatch(name, event_glob))
if (event_glob != NULL && !strglobmatch(name, event_glob))
continue;
printf(" %-42s [%s]\n", name,
printf(" %-50s [%s]\n", name,
event_type_descriptors[PERF_TYPE_HW_CACHE]);
++printed;
}
@ -1009,14 +1028,16 @@ int print_hwcache_events(const char *event_glob)
return printed;
}
#define MAX_NAME_LEN 100
/*
* Print the help text for the event symbols:
*/
void print_events(const char *event_glob)
{
struct event_symbol *syms = event_symbols;
unsigned int i, type, prev_type = -1, printed = 0, ntypes_printed = 0;
char name[40];
struct event_symbol *syms = event_symbols;
char name[MAX_NAME_LEN];
printf("\n");
printf("List of pre-defined events (to be used in -e):\n");
@ -1036,10 +1057,10 @@ void print_events(const char *event_glob)
continue;
if (strlen(syms->alias))
sprintf(name, "%s OR %s", syms->symbol, syms->alias);
snprintf(name, MAX_NAME_LEN, "%s OR %s", syms->symbol, syms->alias);
else
strcpy(name, syms->symbol);
printf(" %-42s [%s]\n", name,
strncpy(name, syms->symbol, MAX_NAME_LEN);
printf(" %-50s [%s]\n", name,
event_type_descriptors[type]);
prev_type = type;
@ -1056,12 +1077,12 @@ void print_events(const char *event_glob)
return;
printf("\n");
printf(" %-42s [%s]\n",
printf(" %-50s [%s]\n",
"rNNN (see 'perf list --help' on how to encode it)",
event_type_descriptors[PERF_TYPE_RAW]);
printf("\n");
printf(" %-42s [%s]\n",
printf(" %-50s [%s]\n",
"mem:<addr>[:access]",
event_type_descriptors[PERF_TYPE_BREAKPOINT]);
printf("\n");

73
tools/perf/util/probe-finder.c
View file

@ -1471,6 +1471,38 @@ static int find_probe_point_by_func(struct probe_finder *pf)
return _param.retval;
}
struct pubname_callback_param {
char *function;
char *file;
Dwarf_Die *cu_die;
Dwarf_Die *sp_die;
int found;
};
static int pubname_search_cb(Dwarf *dbg, Dwarf_Global *gl, void *data)
{
struct pubname_callback_param *param = data;
if (dwarf_offdie(dbg, gl->die_offset, param->sp_die)) {
if (dwarf_tag(param->sp_die) != DW_TAG_subprogram)
return DWARF_CB_OK;
if (die_compare_name(param->sp_die, param->function)) {
if (!dwarf_offdie(dbg, gl->cu_offset, param->cu_die))
return DWARF_CB_OK;
if (param->file &&
strtailcmp(param->file, dwarf_decl_file(param->sp_die)))
return DWARF_CB_OK;
param->found = 1;
return DWARF_CB_ABORT;
}
}
return DWARF_CB_OK;
}
/* Find probe points from debuginfo */
static int find_probes(int fd, struct probe_finder *pf)
{
@ -1498,6 +1530,28 @@ static int find_probes(int fd, struct probe_finder *pf)
off = 0;
line_list__init(&pf->lcache);
/* Fastpath: lookup by function name from .debug_pubnames section */
if (pp->function) {
struct pubname_callback_param pubname_param = {
.function = pp->function,
.file = pp->file,
.cu_die = &pf->cu_die,
.sp_die = &pf->sp_die,
.found = 0,
};
struct dwarf_callback_param probe_param = {
.data = pf,
};
dwarf_getpubnames(dbg, pubname_search_cb, &pubname_param, 0);
if (pubname_param.found) {
ret = probe_point_search_cb(&pf->sp_die, &probe_param);
if (ret)
goto found;
}
}
/* Loop on CUs (Compilation Unit) */
while (!dwarf_nextcu(dbg, off, &noff, &cuhl, NULL, NULL, NULL)) {
/* Get the DIE(Debugging Information Entry) of this CU */
@ -1525,6 +1579,8 @@ static int find_probes(int fd, struct probe_finder *pf)
}
off = noff;
}
found:
line_list__free(&pf->lcache);
if (dwfl)
dwfl_end(dwfl);
@ -1946,6 +2002,22 @@ int find_line_range(int fd, struct line_range *lr)
return -EBADF;
}
/* Fastpath: lookup by function name from .debug_pubnames section */
if (lr->function) {
struct pubname_callback_param pubname_param = {
.function = lr->function, .file = lr->file,
.cu_die = &lf.cu_die, .sp_die = &lf.sp_die, .found = 0};
struct dwarf_callback_param line_range_param = {
.data = (void *)&lf, .retval = 0};
dwarf_getpubnames(dbg, pubname_search_cb, &pubname_param, 0);
if (pubname_param.found) {
line_range_search_cb(&lf.sp_die, &line_range_param);
if (lf.found)
goto found;
}
}
/* Loop on CUs (Compilation Unit) */
while (!lf.found && ret >= 0) {
if (dwarf_nextcu(dbg, off, &noff, &cuhl, NULL, NULL, NULL) != 0)
@ -1974,6 +2046,7 @@ int find_line_range(int fd, struct line_range *lr)
off = noff;
}
found:
/* Store comp_dir */
if (lf.found) {
comp_dir = cu_get_comp_dir(&lf.cu_die);

2
tools/perf/util/probe-finder.h
View file

@ -49,6 +49,7 @@ struct probe_finder {
Dwarf_Addr addr; /* Address */
const char *fname; /* Real file name */
Dwarf_Die cu_die; /* Current CU */
Dwarf_Die sp_die;
struct list_head lcache; /* Line cache for lazy match */
/* For variable searching */
@ -83,6 +84,7 @@ struct line_finder {
int lno_s; /* Start line number */
int lno_e; /* End line number */
Dwarf_Die cu_die; /* Current CU */
Dwarf_Die sp_die;
int found;
};

3
tools/perf/util/python.c
View file

@ -810,6 +810,9 @@ static struct {
{ "COUNT_HW_CACHE_RESULT_ACCESS", PERF_COUNT_HW_CACHE_RESULT_ACCESS },
{ "COUNT_HW_CACHE_RESULT_MISS", PERF_COUNT_HW_CACHE_RESULT_MISS },
{ "COUNT_HW_STALLED_CYCLES_FRONTEND", PERF_COUNT_HW_STALLED_CYCLES_FRONTEND },
{ "COUNT_HW_STALLED_CYCLES_BACKEND", PERF_COUNT_HW_STALLED_CYCLES_BACKEND },
{ "COUNT_SW_CPU_CLOCK", PERF_COUNT_SW_CPU_CLOCK },
{ "COUNT_SW_TASK_CLOCK", PERF_COUNT_SW_TASK_CLOCK },
{ "COUNT_SW_PAGE_FAULTS", PERF_COUNT_SW_PAGE_FAULTS },

12
tools/perf/util/session.c
View file

@ -1156,6 +1156,18 @@ size_t perf_session__fprintf_nr_events(struct perf_session *session, FILE *fp)
return ret;
}
struct perf_evsel *perf_session__find_first_evtype(struct perf_session *session,
unsigned int type)
{
struct perf_evsel *pos;
list_for_each_entry(pos, &session->evlist->entries, node) {
if (pos->attr.type == type)
return pos;
}
return NULL;
}
void perf_session__print_symbols(union perf_event *event,
struct perf_sample *sample,
struct perf_session *session)

3
tools/perf/util/session.h
View file

@ -162,6 +162,9 @@ static inline int perf_session__parse_sample(struct perf_session *session,
session->sample_id_all, sample);
}
struct perf_evsel *perf_session__find_first_evtype(struct perf_session *session,
unsigned int type);
void perf_session__print_symbols(union perf_event *event,
struct perf_sample *sample,
struct perf_session *session);

629
tools/perf/util/symbol.c
View file

File diff suppressed because it is too large Load diff

78
tools/perf/util/symbol.h
View file

@ -62,7 +62,7 @@ struct symbol {
char name[0];
};
void symbol__delete(struct symbol *self);
void symbol__delete(struct symbol *sym);
struct strlist;
@ -96,9 +96,9 @@ struct symbol_conf {
extern struct symbol_conf symbol_conf;
static inline void *symbol__priv(struct symbol *self)
static inline void *symbol__priv(struct symbol *sym)
{
return ((void *)self) - symbol_conf.priv_size;
return ((void *)sym) - symbol_conf.priv_size;
}
struct ref_reloc_sym {
@ -155,43 +155,45 @@ struct dso {
struct dso *dso__new(const char *name);
struct dso *dso__new_kernel(const char *name);
void dso__delete(struct dso *self);
void dso__delete(struct dso *dso);
int dso__name_len(const struct dso *self);
int dso__name_len(const struct dso *dso);
bool dso__loaded(const struct dso *self, enum map_type type);
bool dso__sorted_by_name(const struct dso *self, enum map_type type);
bool dso__loaded(const struct dso *dso, enum map_type type);
bool dso__sorted_by_name(const struct dso *dso, enum map_type type);
static inline void dso__set_loaded(struct dso *self, enum map_type type)
static inline void dso__set_loaded(struct dso *dso, enum map_type type)
{
self->loaded |= (1 << type);
dso->loaded |= (1 << type);
}
void dso__sort_by_name(struct dso *self, enum map_type type);
void dso__sort_by_name(struct dso *dso, enum map_type type);
struct dso *__dsos__findnew(struct list_head *head, const char *name);
int dso__load(struct dso *self, struct map *map, symbol_filter_t filter);
int dso__load_vmlinux(struct dso *self, struct map *map,
int dso__load(struct dso *dso, struct map *map, symbol_filter_t filter);
int dso__load_vmlinux(struct dso *dso, struct map *map,
const char *vmlinux, symbol_filter_t filter);
int dso__load_vmlinux_path(struct dso *self, struct map *map,
int dso__load_vmlinux_path(struct dso *dso, struct map *map,
symbol_filter_t filter);
int dso__load_kallsyms(struct dso *self, const char *filename, struct map *map,
int dso__load_kallsyms(struct dso *dso, const char *filename, struct map *map,
symbol_filter_t filter);
int machine__load_kallsyms(struct machine *self, const char *filename,
int machine__load_kallsyms(struct machine *machine, const char *filename,
enum map_type type, symbol_filter_t filter);
int machine__load_vmlinux_path(struct machine *self, enum map_type type,
int machine__load_vmlinux_path(struct machine *machine, enum map_type type,
symbol_filter_t filter);
size_t __dsos__fprintf(struct list_head *head, FILE *fp);
size_t machine__fprintf_dsos_buildid(struct machine *self, FILE *fp, bool with_hits);
size_t machines__fprintf_dsos(struct rb_root *self, FILE *fp);
size_t machines__fprintf_dsos_buildid(struct rb_root *self, FILE *fp, bool with_hits);
size_t dso__fprintf_buildid(struct dso *self, FILE *fp);
size_t dso__fprintf_symbols_by_name(struct dso *self, enum map_type type, FILE *fp);
size_t dso__fprintf(struct dso *self, enum map_type type, FILE *fp);
size_t machine__fprintf_dsos_buildid(struct machine *machine,
FILE *fp, bool with_hits);
size_t machines__fprintf_dsos(struct rb_root *machines, FILE *fp);
size_t machines__fprintf_dsos_buildid(struct rb_root *machines,
FILE *fp, bool with_hits);
size_t dso__fprintf_buildid(struct dso *dso, FILE *fp);
size_t dso__fprintf_symbols_by_name(struct dso *dso,
enum map_type type, FILE *fp);
size_t dso__fprintf(struct dso *dso, enum map_type type, FILE *fp);
enum symtab_type {
SYMTAB__KALLSYMS = 0,
@ -207,34 +209,36 @@ enum symtab_type {
SYMTAB__NOT_FOUND,
};
char dso__symtab_origin(const struct dso *self);
void dso__set_long_name(struct dso *self, char *name);
void dso__set_build_id(struct dso *self, void *build_id);
void dso__read_running_kernel_build_id(struct dso *self, struct machine *machine);
struct symbol *dso__find_symbol(struct dso *self, enum map_type type, u64 addr);
struct symbol *dso__find_symbol_by_name(struct dso *self, enum map_type type,
char dso__symtab_origin(const struct dso *dso);
void dso__set_long_name(struct dso *dso, char *name);
void dso__set_build_id(struct dso *dso, void *build_id);
void dso__read_running_kernel_build_id(struct dso *dso,
struct machine *machine);
struct symbol *dso__find_symbol(struct dso *dso, enum map_type type,
u64 addr);
struct symbol *dso__find_symbol_by_name(struct dso *dso, enum map_type type,
const char *name);
int filename__read_build_id(const char *filename, void *bf, size_t size);
int sysfs__read_build_id(const char *filename, void *bf, size_t size);
bool __dsos__read_build_ids(struct list_head *head, bool with_hits);
int build_id__sprintf(const u8 *self, int len, char *bf);
int build_id__sprintf(const u8 *build_id, int len, char *bf);
int kallsyms__parse(const char *filename, void *arg,
int (*process_symbol)(void *arg, const char *name,
char type, u64 start, u64 end));
void machine__destroy_kernel_maps(struct machine *self);
int __machine__create_kernel_maps(struct machine *self, struct dso *kernel);
int machine__create_kernel_maps(struct machine *self);
void machine__destroy_kernel_maps(struct machine *machine);
int __machine__create_kernel_maps(struct machine *machine, struct dso *kernel);
int machine__create_kernel_maps(struct machine *machine);
int machines__create_kernel_maps(struct rb_root *self, pid_t pid);
int machines__create_guest_kernel_maps(struct rb_root *self);
void machines__destroy_guest_kernel_maps(struct rb_root *self);
int machines__create_kernel_maps(struct rb_root *machines, pid_t pid);
int machines__create_guest_kernel_maps(struct rb_root *machines);
void machines__destroy_guest_kernel_maps(struct rb_root *machines);
int symbol__init(void);
void symbol__exit(void);
bool symbol_type__is_a(char symbol_type, enum map_type map_type);
size_t machine__fprintf_vmlinux_path(struct machine *self, FILE *fp);
size_t machine__fprintf_vmlinux_path(struct machine *machine, FILE *fp);
#endif /* __PERF_SYMBOL */