Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit ea79ca0d authored by Ingo Molnar's avatar Ingo Molnar
Browse files

Merge tag 'perf-core-for-mingo' of... 

Merge tag 'perf-core-for-mingo' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux

 into perf/core

Pull perf/core improvements and fixes from Arnaldo Carvalho de Melo:

 * Tidy up sample parsing validation, from Adrian Hunter.

 * Make events stream always parsable by adding a new sample_type bit:
   PERF_SAMPLE_IDENTIFIER, that when requested will be always aat a fixed
   position in all PERF_RECORD_ records, from Adrian Hunter.

 * Add a sample parsing test, from Adrian Hunter.

 * Add option to 'perf trace' to analyze events in a file versus live,
   so that one can do:

   [root@zoo ~]# perf record -a -e raw_syscalls:* sleep 1
   [ perf record: Woken up 0 times to write data ]
   [ perf record: Captured and wrote 25.150 MB perf.data (~1098836 samples) ]
   [root@zoo ~]# perf trace -i perf.data -e futex --duration 1
      17.799 ( 1.020 ms): 7127 futex(uaddr: 0x7fff3f6c6674, op: 393, val: 1, utime: 0x7fff3f6c6470, ua
     113.344 (95.429 ms): 7127 futex(uaddr: 0x7fff3f6c6674, op: 393, val: 1, utime: 0x7fff3f6c6470, uaddr2: 0x7fff3f6c6648, val3: 4294967
     133.778 ( 1.042 ms): 18004 futex(uaddr: 0x7fff3f6c6674, op: 393, val: 1, utime: 0x7fff3f6c6470, uaddr2: 0x7fff3f6c6648, val3: 429496
   [root@zoo ~]#

   From David Ahern.

 * Honor target pid / tid options in 'perf trace' when analyzing a file,
   from David Ahern.

 * Handle missing HUGEPAGE defines in the mmap beautifier in 'perf trace',
   from David Ahern.

Signed-off-by: default avatarArnaldo Carvalho de Melo <acme@redhat.com>
Signed-off-by: default avatarIngo Molnar <mingo@kernel.org>
parents 00e4cb1c f2935f3e

include/uapi/linux/perf_event.h

+20 −7
Original line number Original line Diff line number Diff line
@@ -134,8 +134,9 @@ enum perf_event_sample_format { 
	PERF_SAMPLE_STACK_USER			= 1U << 13,

	PERF_SAMPLE_STACK_USER			= 1U << 13,

	PERF_SAMPLE_WEIGHT			= 1U << 14,

	PERF_SAMPLE_WEIGHT			= 1U << 14,

	PERF_SAMPLE_DATA_SRC			= 1U << 15,

	PERF_SAMPLE_DATA_SRC			= 1U << 15,

	PERF_SAMPLE_IDENTIFIER			= 1U << 16,





	PERF_SAMPLE_MAX = 1U << 16,		/* non-ABI */

	PERF_SAMPLE_MAX = 1U << 17,		/* non-ABI */

};

};





/*

/*
@@ -492,12 +493,12 @@ enum perf_event_type { 
	/*

	/*

	 * If perf_event_attr.sample_id_all is set then all event types will

	 * If perf_event_attr.sample_id_all is set then all event types will

	 * have the sample_type selected fields related to where/when

	 * have the sample_type selected fields related to where/when

	 * (identity) an event took place (TID, TIME, ID, CPU, STREAM_ID)

	 * (identity) an event took place (TID, TIME, ID, STREAM_ID, CPU,

	 * described in PERF_RECORD_SAMPLE below, it will be stashed just after

	 * IDENTIFIER) described in PERF_RECORD_SAMPLE below, it will be stashed

	 * the perf_event_header and the fields already present for the existing

	 * just after the perf_event_header and the fields already present for

	 * fields, i.e. at the end of the payload. That way a newer perf.data

	 * the existing fields, i.e. at the end of the payload. That way a newer

	 * file will be supported by older perf tools, with these new optional

	 * perf.data file will be supported by older perf tools, with these new

	 * fields being ignored.

	 * optional fields being ignored.

	 *

	 *

	 * struct sample_id {

	 * struct sample_id {

	 * 	{ u32			pid, tid; } && PERF_SAMPLE_TID

	 * 	{ u32			pid, tid; } && PERF_SAMPLE_TID
@@ -505,7 +506,12 @@ enum perf_event_type { 
	 * 	{ u64			id;       } && PERF_SAMPLE_ID

	 * 	{ u64			id;       } && PERF_SAMPLE_ID

	 * 	{ u64			stream_id;} && PERF_SAMPLE_STREAM_ID

	 * 	{ u64			stream_id;} && PERF_SAMPLE_STREAM_ID

	 * 	{ u32			cpu, res; } && PERF_SAMPLE_CPU

	 * 	{ u32			cpu, res; } && PERF_SAMPLE_CPU

	 *	{ u64			id;	  } && PERF_SAMPLE_IDENTIFIER

	 * } && perf_event_attr::sample_id_all

	 * } && perf_event_attr::sample_id_all

	 *

	 * Note that PERF_SAMPLE_IDENTIFIER duplicates PERF_SAMPLE_ID.  The

	 * advantage of PERF_SAMPLE_IDENTIFIER is that its position is fixed

	 * relative to header.size.

	 */

	 */





	/*

	/*
@@ -594,6 +600,13 @@ enum perf_event_type { 
	 * struct {

	 * struct {

	 *	struct perf_event_header	header;

	 *	struct perf_event_header	header;

	 *

	 *

	 *	#

	 *	# Note that PERF_SAMPLE_IDENTIFIER duplicates PERF_SAMPLE_ID.

	 *	# The advantage of PERF_SAMPLE_IDENTIFIER is that its position

	 *	# is fixed relative to header.

	 *	#

	 *

	 *	{ u64			id;	  } && PERF_SAMPLE_IDENTIFIER

	 *	{ u64			ip;	  } && PERF_SAMPLE_IP

	 *	{ u64			ip;	  } && PERF_SAMPLE_IP

	 *	{ u32			pid, tid; } && PERF_SAMPLE_TID

	 *	{ u32			pid, tid; } && PERF_SAMPLE_TID

	 *	{ u64			time;     } && PERF_SAMPLE_TIME

	 *	{ u64			time;     } && PERF_SAMPLE_TIME

kernel/events/core.c

+10 −1
Original line number Original line Diff line number Diff line
@@ -1213,6 +1213,9 @@ static void perf_event__id_header_size(struct perf_event *event) 
	if (sample_type & PERF_SAMPLE_TIME)

	if (sample_type & PERF_SAMPLE_TIME)

		size += sizeof(data->time);

		size += sizeof(data->time);





	if (sample_type & PERF_SAMPLE_IDENTIFIER)

		size += sizeof(data->id);



	if (sample_type & PERF_SAMPLE_ID)

	if (sample_type & PERF_SAMPLE_ID)

		size += sizeof(data->id);

		size += sizeof(data->id);
@@ -4280,7 +4283,7 @@ static void __perf_event_header__init_id(struct perf_event_header *header, 
	if (sample_type & PERF_SAMPLE_TIME)

	if (sample_type & PERF_SAMPLE_TIME)

		data->time = perf_clock();

		data->time = perf_clock();





	if (sample_type & PERF_SAMPLE_ID)

	if (sample_type & (PERF_SAMPLE_ID | PERF_SAMPLE_IDENTIFIER))

		data->id = primary_event_id(event);

		data->id = primary_event_id(event);





	if (sample_type & PERF_SAMPLE_STREAM_ID)

	if (sample_type & PERF_SAMPLE_STREAM_ID)
@@ -4319,6 +4322,9 @@ static void __perf_event__output_id_sample(struct perf_output_handle *handle, 




	if (sample_type & PERF_SAMPLE_CPU)

	if (sample_type & PERF_SAMPLE_CPU)

		perf_output_put(handle, data->cpu_entry);

		perf_output_put(handle, data->cpu_entry);



	if (sample_type & PERF_SAMPLE_IDENTIFIER)

		perf_output_put(handle, data->id);

}

}





void perf_event__output_id_sample(struct perf_event *event,

void perf_event__output_id_sample(struct perf_event *event,
@@ -4432,6 +4438,9 @@ void perf_output_sample(struct perf_output_handle *handle, 




	perf_output_put(handle, *header);

	perf_output_put(handle, *header);





	if (sample_type & PERF_SAMPLE_IDENTIFIER)

		perf_output_put(handle, data->id);



	if (sample_type & PERF_SAMPLE_IP)

	if (sample_type & PERF_SAMPLE_IP)

		perf_output_put(handle, data->ip);

		perf_output_put(handle, data->ip);

tools/perf/Documentation/perf-trace.txt

+4 −0
Original line number Original line Diff line number Diff line
@@ -74,6 +74,10 @@ the thread executes on the designated CPUs. Default is to monitor all CPUs. 
--sched:

--sched:

	Accrue thread runtime and provide a summary at the end of the session.

	Accrue thread runtime and provide a summary at the end of the session.





-i

--input

	Process events from a given perf data file.



SEE ALSO

SEE ALSO

--------

--------

linkperf:perf-record[1], linkperf:perf-script[1]
 
linkperf:perf-record[1], linkperf:perf-script[1]

tools/perf/Makefile

+2 −0
Original line number Original line Diff line number Diff line
@@ -360,6 +360,7 @@ LIB_OBJS += $(OUTPUT)util/rblist.o 
LIB_OBJS += $(OUTPUT)util/intlist.o

LIB_OBJS += $(OUTPUT)util/intlist.o

LIB_OBJS += $(OUTPUT)util/vdso.o

LIB_OBJS += $(OUTPUT)util/vdso.o

LIB_OBJS += $(OUTPUT)util/stat.o

LIB_OBJS += $(OUTPUT)util/stat.o

LIB_OBJS += $(OUTPUT)util/record.o





LIB_OBJS += $(OUTPUT)ui/setup.o

LIB_OBJS += $(OUTPUT)ui/setup.o

LIB_OBJS += $(OUTPUT)ui/helpline.o

LIB_OBJS += $(OUTPUT)ui/helpline.o
@@ -438,6 +439,7 @@ PERFLIBS = $(LIB_FILE) $(LIBLK) $(LIBTRACEEVENT) 
ifneq ($(OUTPUT),)

ifneq ($(OUTPUT),)

  CFLAGS += -I$(OUTPUT)

  CFLAGS += -I$(OUTPUT)

endif

endif

LIB_OBJS += $(OUTPUT)tests/sample-parsing.o





ifdef NO_LIBELF

ifdef NO_LIBELF

EXTLIBS := $(filter-out -lelf,$(EXTLIBS))

EXTLIBS := $(filter-out -lelf,$(EXTLIBS))

tools/perf/builtin-inject.c

+5 −3
Original line number Original line Diff line number Diff line
@@ -198,7 +198,7 @@ static int perf_event__inject_buildid(struct perf_tool *tool, 




	cpumode = event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK;

	cpumode = event->header.misc & PERF_RECORD_MISC_CPUMODE_MASK;





	thread = machine__findnew_thread(machine, event->ip.pid);

	thread = machine__findnew_thread(machine, sample->pid, sample->pid);

	if (thread == NULL) {

	if (thread == NULL) {

		pr_err("problem processing %d event, skipping it.\n",

		pr_err("problem processing %d event, skipping it.\n",

		       event->header.type);

		       event->header.type);
@@ -206,7 +206,7 @@ static int perf_event__inject_buildid(struct perf_tool *tool, 
	}

	}





	thread__find_addr_map(thread, machine, cpumode, MAP__FUNCTION,

	thread__find_addr_map(thread, machine, cpumode, MAP__FUNCTION,

			      event->ip.ip, &al);

			      sample->ip, &al);





	if (al.map != NULL) {

	if (al.map != NULL) {

		if (!al.map->dso->hit) {

		if (!al.map->dso->hit) {
@@ -301,7 +301,9 @@ static int perf_inject__sched_stat(struct perf_tool *tool, 
	sample_sw.period = sample->period;

	sample_sw.period = sample->period;

	sample_sw.time	 = sample->time;

	sample_sw.time	 = sample->time;

	perf_event__synthesize_sample(event_sw, evsel->attr.sample_type,

	perf_event__synthesize_sample(event_sw, evsel->attr.sample_type,

				      &sample_sw, false);

				      evsel->attr.sample_regs_user,

				      evsel->attr.read_format, &sample_sw,

				      false);

	build_id__mark_dso_hit(tool, event_sw, &sample_sw, evsel, machine);

	build_id__mark_dso_hit(tool, event_sw, &sample_sw, evsel, machine);

	return perf_event__repipe(tool, event_sw, &sample_sw, machine);

	return perf_event__repipe(tool, event_sw, &sample_sw, machine);

}

}