Skip to contentSkip to conversation

perf intel-pt: Add support for decoding FUP/TIP only

PATCH Superseded Adrian Hunter 2 participants 20 messages Jul 9, 2020 – Jul 20, 2020 perf events

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore

Adrian Hunter

author reviewer
#0 lore raw

cover perf intel-pt: Add support for decoding FUP/TIP only

Hi

Here are some fixes and small improvements for Intel PT.

Adrian Hunter (11):
perf intel-pt: Fix FUP packet state
perf intel-pt: Fix duplicate branch after CBR
perf tools: Improve aux_output not supported error
perf auxtrace: Add optional error flags to the itrace 'e' option perf intel-pt: Use itrace error flags to suppress some errors perf auxtrace: Add optional log flags to the itrace 'd' option perf intel-pt: Use itrace debug log flags to suppress some messages perf intel-pt: Time filter logged perf events
perf auxtrace: Add itrace 'q' option for quicker, less detailed decoding perf intel-pt: Add support for decoding FUP/TIP only
perf intel-pt: Add support for decoding PSB+ only

 tools/perf/Documentation/itrace.txt                |   9 +
 tools/perf/Documentation/perf-intel-pt.txt         |  25 ++-
 tools/perf/util/auxtrace.c                         |   7 +
 tools/perf/util/auxtrace.h                         |   6 +
 tools/perf/util/evsel.c                            |   4 +
 .../perf/util/intel-pt-decoder/intel-pt-decoder.c  | 214 +++++++++++++++++++--
 .../perf/util/intel-pt-decoder/intel-pt-decoder.h  |   1 +
 tools/perf/util/intel-pt.c                         |  51 ++++-
 8 files changed, 287 insertions(+), 30 deletions(-)

Regards
Adrian

author reviewer
#2 lore raw

PATCH 2/11 perf intel-pt: Fix duplicate branch after CBR

CBR events can result in a duplicate branch event, because the state type defaults to a branch. Fix by clearing the state type.

Example: trace 'sleep' and hope for a frequency change

Before:

   $ perf record -e intel_pt//u sleep 0.1
   [ perf record: Woken up 1 times to write data ]
   [ perf record: Captured and wrote 0.034 MB perf.data ]
   $ perf script --itrace=bpe > before.txt

After:

   $ perf script --itrace=bpe > after.txt
   $ diff -u before.txt after.txt
   --- before.txt  2020-07-07 14:42:18.191508098 +0300
   +++ after.txt   2020-07-07 14:42:36.587891753 +0300
   @@ -29673,7 +29673,6 @@
               sleep 93431 [007] 15411.619905:          1  branches:u:                 0 [unknown] ([unknown]) =>     7f0818abb2e0 clock_nanosleep@@GLIBC_2.17+0x0 (/usr/lib/x86_64-linux-gnu/libc-2.31.so)
               sleep 93431 [007] 15411.619905:          1  branches:u:      7f0818abb30c clock_nanosleep@@GLIBC_2.17+0x2c (/usr/lib/x86_64-linux-gnu/libc-2.31.so) =>                0 [unknown] ([unknown])
               sleep 93431 [007] 15411.720069:         cbr:  cbr: 15 freq: 1507 MHz ( 56%)         7f0818abb30c clock_nanosleep@@GLIBC_2.17+0x2c (/usr/lib/x86_64-linux-gnu/libc-2.31.so)
   -           sleep 93431 [007] 15411.720069:          1  branches:u:      7f0818abb30c clock_nanosleep@@GLIBC_2.17+0x2c (/usr/lib/x86_64-linux-gnu/libc-2.31.so) =>                0 [unknown] ([unknown])
               sleep 93431 [007] 15411.720076:          1  branches:u:                 0 [unknown] ([unknown]) =>     7f0818abb30e clock_nanosleep@@GLIBC_2.17+0x2e (/usr/lib/x86_64-linux-gnu/libc-2.31.so)
               sleep 93431 [007] 15411.720077:          1  branches:u:      7f0818abb323 clock_nanosleep@@GLIBC_2.17+0x43 (/usr/lib/x86_64-linux-gnu/libc-2.31.so) =>     7f0818ac0eb7 __nanosleep+0x17 (/usr/lib/x86_64-linux-gnu/libc-2.31.so)
               sleep 93431 [007] 15411.720077:          1  branches:u:      7f0818ac0ebf __nanosleep+0x1f (/usr/lib/x86_64-linux-gnu/libc-2.31.so) =>     55cb7e4c2827 rpl_nanosleep+0x97 (/usr/bin/sleep)
Fixes 91de8684f1cff ("perf intel-pt: Cater for CBR change in PSB+")
Fixes abe5a1d3e4bee ("perf intel-pt: Decoder to output CBR changes immediately")
---
 tools/perf/util/intel-pt-decoder/intel-pt-decoder.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)
tools/perf/util/intel-pt-decoder/intel-pt-decoder.c +6 -2
index 75c4bd74d521..7ffcbd6fcd1a 100644
@@ -1977,8 +1977,10 @@ static int intel_pt_walk_trace(struct intel_pt_decoder *decoder)
 			 * possibility of another CBR change that gets caught up
 			 * in the PSB+.
 			 */
-			if (decoder->cbr != decoder->cbr_seen)
+			if (decoder->cbr != decoder->cbr_seen) {
+				decoder->state.type = 0;
 				return 0;
+			}
 			break;
 
 		case INTEL_PT_PIP:
@@ -2019,8 +2021,10 @@ static int intel_pt_walk_trace(struct intel_pt_decoder *decoder)
 
 		case INTEL_PT_CBR:
 			intel_pt_calc_cbr(decoder);
-			if (decoder->cbr != decoder->cbr_seen)
+			if (decoder->cbr != decoder->cbr_seen) {
+				decoder->state.type = 0;
 				return 0;
+			}
 			break;
 
 		case INTEL_PT_MODE_EXEC:
Signature · 1 line
2.25.1
author reviewer
#4 lore raw

PATCH 4/11 perf auxtrace: Add optional error flags to the itrace 'e' option

In reply to Adrian Hunter

Allow the 'e' option to be followed by an architecture-specific number which flags what kind of errors will or will not be reported.

---
 tools/perf/Documentation/itrace.txt | 3 +++
 tools/perf/util/auxtrace.c          | 2 ++
 tools/perf/util/auxtrace.h          | 2 ++
 3 files changed, 7 insertions(+)
tools/perf/Documentation/itrace.txt +3 -0
index e817179c5027..34864b4047ed 100644
@@ -47,3 +47,6 @@
 	--itrace=i0nss1000000
 
 	skips the first million instructions.
+
+	The 'e' option may be followed by an architecture-specific number which
+	flags what kind of errors will or will not be reported.
tools/perf/util/auxtrace.c +2 -0
index 25c639ac4ad4..5cfc0b12b2b3 100644
@@ -1436,6 +1436,8 @@ int itrace_parse_synth_opts(const struct option *opt, const char *str,
 			break;
 		case 'e':
 			synth_opts->errors = true;
+			synth_opts->error_flags = strtoul(p, &endptr, 0);
+			p = endptr;
 			break;
 		case 'd':
 			synth_opts->log = true;
tools/perf/util/auxtrace.h +2 -0
index 142ccf7d34df..a04475f41f28 100644
@@ -91,6 +91,7 @@ enum itrace_period_type {
  * @cpu_bitmap: CPUs for which to synthesize events, or NULL for all
  * @ptime_range: time intervals to trace or NULL
  * @range_num: number of time intervals to trace
+ * @error_flags: arch-specific flags to affect what errors are reported
  */
 struct itrace_synth_opts {
 	bool			set;
@@ -124,6 +125,7 @@ struct itrace_synth_opts {
 	unsigned long		*cpu_bitmap;
 	struct perf_time_interval *ptime_range;
 	int			range_num;
+	unsigned int		error_flags;
 };
 
 /**
Signature · 1 line
2.25.1
author reviewer
#5 lore raw

PATCH 5/11 perf intel-pt: Use itrace error flags to suppress some errors

In reply to Adrian Hunter

The itrace "e" option may be followed by a number which has the following effect for Intel PT:
1 Suppress overflow events
2 Suppress trace data lost events
The values may be combined by bitwise OR'ing them.

Suppressing those errors can be useful for testing and debugging because they are not due to decoding.

---
 tools/perf/Documentation/perf-intel-pt.txt |  7 ++++++-
 tools/perf/util/intel-pt.c                 | 12 ++++++++++++
 2 files changed, 18 insertions(+), 1 deletion(-)
tools/perf/Documentation/perf-intel-pt.txt +6 -1
index f4cd49a7fcdb..0fcd8ad897b0 100644
@@ -871,7 +871,11 @@ Developer Manuals.
 
 Error events show where the decoder lost the trace.  Error events
 are quite important.  Users must know if what they are seeing is a complete
-picture or not.
+picture or not. The "e" option may be followed by a number which has the
+following effect:
+	1	Suppress overflow events
+	2	Suppress trace data lost events
+The values may be combined by bitwise OR'ing them.
 
 The "d" option will cause the creation of a file "intel_pt.log" containing all
 decoded packets and instructions.  Note that this option slows down the decoder
@@ -956,6 +960,7 @@ at the beginning. This is useful to ignore initialization code.
 
 skips the first million instructions.
 
+
 dump option
 ~~~~~~~~~~~
 
tools/perf/util/intel-pt.c +12 -0
index 8c441b815d73..a8e8e8acbcc8 100644
@@ -46,6 +46,9 @@
 
 #define MAX_TIMESTAMP (~0ULL)
 
+#define INTEL_PT_ERR_SUPPRESS_OVF	1
+#define INTEL_PT_ERR_SUPPRESS_LOST	2
+
 struct range {
 	u64 start;
 	u64 end;
@@ -1863,6 +1866,15 @@ static int intel_pt_synth_error(struct intel_pt *pt, int code, int cpu,
 	char msg[MAX_AUXTRACE_ERROR_MSG];
 	int err;
 
+	if (pt->synth_opts.error_flags) {
+		if (code == INTEL_PT_ERR_OVR &&
+		    pt->synth_opts.error_flags & INTEL_PT_ERR_SUPPRESS_OVF)
+			return 0;
+		if (code == INTEL_PT_ERR_LOST &&
+		    pt->synth_opts.error_flags & INTEL_PT_ERR_SUPPRESS_LOST)
+			return 0;
+	}
+
 	intel_pt__strerror(code, msg, MAX_AUXTRACE_ERROR_MSG);
 
 	auxtrace_synth_error(&event.auxtrace_error, PERF_AUXTRACE_ERROR_ITRACE,
Signature · 1 line
2.25.1
author reviewer
#10 lore raw

PATCH 6/11 perf auxtrace: Add optional log flags to the itrace 'd' option

In reply to Adrian Hunter

Allow the 'd' option to be followed by an architecture-specific number which flags what kind of debug messages will or will not be logged.

---
 tools/perf/Documentation/itrace.txt | 3 +++
 tools/perf/util/auxtrace.c          | 2 ++
 tools/perf/util/auxtrace.h          | 2 ++
 3 files changed, 7 insertions(+)
tools/perf/Documentation/itrace.txt +3 -0
index 34864b4047ed..3dd8fddb8b1b 100644
@@ -50,3 +50,6 @@
 
 	The 'e' option may be followed by an architecture-specific number which
 	flags what kind of errors will or will not be reported.
+
+	If supported, The 'd' option may be followed by an architecture-specific
+	number which flags what kind of debug messages will or will not be logged.
tools/perf/util/auxtrace.c +2 -0
index 5cfc0b12b2b3..3f806c2881c9 100644
@@ -1441,6 +1441,8 @@ int itrace_parse_synth_opts(const struct option *opt, const char *str,
 			break;
 		case 'd':
 			synth_opts->log = true;
+			synth_opts->log_flags = strtoul(p, &endptr, 0);
+			p = endptr;
 			break;
 		case 'c':
 			synth_opts->branches = true;
tools/perf/util/auxtrace.h +2 -0
index a04475f41f28..f41dbdc98175 100644
@@ -92,6 +92,7 @@ enum itrace_period_type {
  * @ptime_range: time intervals to trace or NULL
  * @range_num: number of time intervals to trace
  * @error_flags: arch-specific flags to affect what errors are reported
+ * @log_flags: arch-specific flags to affect what is logged
  */
 struct itrace_synth_opts {
 	bool			set;
@@ -126,6 +127,7 @@ struct itrace_synth_opts {
 	struct perf_time_interval *ptime_range;
 	int			range_num;
 	unsigned int		error_flags;
+	unsigned int		log_flags;
 };
 
 /**
Signature · 1 line
2.25.1
author reviewer
#13 lore raw

PATCH 8/11 perf intel-pt: Time filter logged perf events

In reply to Adrian Hunter

Change the debug logging (when used with the --time option) to time filter logged perf events, but allow that to be overridden by using "d2" instead of plain "d".

By default that can greatly reduce the size of the log file.

---
 tools/perf/Documentation/perf-intel-pt.txt |  3 +++
 tools/perf/util/intel-pt.c                 | 20 +++++++++++++++++---
 2 files changed, 20 insertions(+), 3 deletions(-)
tools/perf/Documentation/perf-intel-pt.txt +3 -0
index 85a2ff804900..9a90e2db4e4a 100644
@@ -882,6 +882,9 @@ decoded packets and instructions.  Note that this option slows down the decoder
 and that the resulting file may be very large.  The "d" option may be followed
 by a number which has the following effect:
 	1	Suppress logging of perf events
+	2	Log all perf events
+By default, logged perf events are filtered by any specified time ranges, but
+value 2 overrides that.
 
 In addition, the period of the "instructions" event can be specified. e.g.
 
tools/perf/util/intel-pt.c +17 -3
index d90375659244..597120dd6b77 100644
@@ -50,6 +50,7 @@
 #define INTEL_PT_ERR_SUPPRESS_LOST	2
 
 #define INTEL_PT_LOG_SUPPRESS_EV	1
+#define INTEL_PT_LOG_ALL_EV		2
 
 struct range {
 	u64 start;
@@ -254,9 +255,22 @@ static void intel_pt_dump_sample(struct perf_session *session,
 	intel_pt_dump(pt, sample->aux_sample.data, sample->aux_sample.size);
 }
 
-static bool intel_pt_log_events(struct intel_pt *pt)
+static bool intel_pt_log_events(struct intel_pt *pt, u64 tm)
 {
-	return !(pt->synth_opts.log_flags & INTEL_PT_LOG_SUPPRESS_EV);
+	struct perf_time_interval *range = pt->synth_opts.ptime_range;
+	int n = pt->synth_opts.range_num;
+
+	if (pt->synth_opts.log_flags & INTEL_PT_LOG_ALL_EV)
+		return true;
+
+	if (pt->synth_opts.log_flags & INTEL_PT_LOG_SUPPRESS_EV)
+		return false;
+
+	/* perf_time__ranges_skip_sample does not work if time is zero */
+	if (!tm)
+		tm = 1;
+
+	return !n || !perf_time__ranges_skip_sample(range, n, tm);
 }
 
 static int intel_pt_do_fix_overlap(struct intel_pt *pt, struct auxtrace_buffer *a,
@@ -2752,7 +2766,7 @@ static int intel_pt_process_event(struct perf_session *session,
 	if (!err && event->header.type == PERF_RECORD_TEXT_POKE)
 		err = intel_pt_text_poke(pt, event);
 
-	if (intel_pt_enable_logging && intel_pt_log_events(pt)) {
+	if (intel_pt_enable_logging && intel_pt_log_events(pt, sample->time)) {
 		intel_pt_log("event %u: cpu %d time %"PRIu64" tsc %#"PRIx64" ",
 			     event->header.type, sample->cpu, sample->time, timestamp);
 		intel_pt_log_event(event);
Signature · 1 line
2.25.1
author reviewer
#16 lore raw

PATCH 10/11 perf intel-pt: Add support for decoding FUP/TIP only

In reply to Adrian Hunter

Use the new itrace 'q' option to add support for a mode of decoding that ignores TNT, does not walk object code, but gets the ip from FUP and TIP packets.

Example:

 $ perf record -e intel_pt//u grep -rI pudding drivers
 [ perf record: Woken up 52 times to write data ]
 [ perf record: Captured and wrote 57.870 MB perf.data ]
 $ time perf script --itrace=bi | wc -l
 58948289
 real    1m23.863s
 user    1m23.251s
 sys     0m7.452s
 $ time perf script --itrace=biq | wc -l
 3385694
 real    0m4.453s
 user    0m4.455s
 sys     0m0.328s
---
 tools/perf/Documentation/perf-intel-pt.txt    |   8 +
 .../util/intel-pt-decoder/intel-pt-decoder.c  | 167 +++++++++++++++++-
 .../util/intel-pt-decoder/intel-pt-decoder.h  |   1 +
 tools/perf/util/intel-pt.c                    |   6 +-
 4 files changed, 177 insertions(+), 5 deletions(-)
tools/perf/Documentation/perf-intel-pt.txt +8 -0
index 9a90e2db4e4a..758295a7e3d6 100644
@@ -825,6 +825,7 @@ The letters are:
 	l	synthesize last branch entries (use with i or x)
 	L	synthesize last branch entries on existing event records
 	s	skip initial number of events
+	q	quicker (less detailed) decoding
 
 "Instructions" events look like they were recorded by "perf record -e
 instructions".
@@ -965,6 +966,13 @@ at the beginning. This is useful to ignore initialization code.
 
 skips the first million instructions.
 
+The q option does not decode TNT packets, and does not walk object code, but
+gets the ip from FUP and TIP packets.  The q option can be used with the b and i
+options but the period is not used.  The q option decodes more quickly, but is
+useful only if the control flow of interest is represented or indicated by FUP,
+TIP, TIP.PGE, or TIP.PGD packets.  However the q option could be used to find
+time ranges that could then be decoded fully using the --time option.
+
 
 dump option
 ~~~~~~~~~~~
tools/perf/util/intel-pt-decoder/intel-pt-decoder.c +163 -4

246 lines not on this page · raw message

tools/perf/util/intel-pt-decoder/intel-pt-decoder.h +1 -0
index e289e463d635..8645fc265481 100644
@@ -250,6 +250,7 @@ struct intel_pt_params {
 	uint32_t tsc_ctc_ratio_n;
 	uint32_t tsc_ctc_ratio_d;
 	enum intel_pt_param_flags flags;
+	unsigned int quick;
 };
 
 struct intel_pt_decoder;
tools/perf/util/intel-pt.c +5 -1
index 597120dd6b77..93659e738d40 100644
@@ -1036,6 +1036,7 @@ static struct intel_pt_queue *intel_pt_alloc_queue(struct intel_pt *pt,
 	params.mtc_period = intel_pt_mtc_period(pt);
 	params.tsc_ctc_ratio_n = pt->tsc_ctc_ratio_n;
 	params.tsc_ctc_ratio_d = pt->tsc_ctc_ratio_d;
+	params.quick = pt->synth_opts.quick;
 
 	if (pt->filts.cnt > 0)
 		params.pgd_ip = intel_pt_pgd_ip;
@@ -1429,7 +1430,10 @@ static int intel_pt_synth_instruction_sample(struct intel_pt_queue *ptq)
 
 	sample.id = ptq->pt->instructions_id;
 	sample.stream_id = ptq->pt->instructions_id;
-	sample.period = ptq->state->tot_insn_cnt - ptq->last_insn_cnt;
+	if (pt->synth_opts.quick)
+		sample.period = 1;
+	else
+		sample.period = ptq->state->tot_insn_cnt - ptq->last_insn_cnt;
 
 	sample.cyc_cnt = ptq->ipc_cyc_cnt - ptq->last_in_cyc_cnt;
 	if (sample.cyc_cnt) {
Signature · 1 line
2.25.1
author reviewer
#14 lore raw

PATCH 9/11 perf auxtrace: Add itrace 'q' option for quicker, less detailed decoding

In reply to Adrian Hunter

The 'q' option is for modes of decoding that are quicker because they skip or omit decoding some aspects of trace data.

If supported, the 'q' option may be repeated to increase the effect.

---
 tools/perf/Documentation/itrace.txt | 3 +++
 tools/perf/util/auxtrace.c          | 3 +++
 tools/perf/util/auxtrace.h          | 2 ++
 3 files changed, 8 insertions(+)
tools/perf/Documentation/itrace.txt +3 -0
index 3dd8fddb8b1b..d4ffa11b9d50 100644
@@ -18,6 +18,7 @@
 		l	synthesize last branch entries (use with i or x)
 		L	synthesize last branch entries on existing event records
 		s       skip initial number of events
+		q	quicker (less detailed) decoding
 
 	The default is all events i.e. the same as --itrace=ibxwpe,
 	except for perf script where it is --itrace=ce
@@ -53,3 +54,5 @@
 
 	If supported, The 'd' option may be followed by an architecture-specific
 	number which flags what kind of debug messages will or will not be logged.
+
+	If supported, the 'q' option may be repeated to increase the effect.
tools/perf/util/auxtrace.c +3 -0
index 3f806c2881c9..81726c014237 100644
@@ -1511,6 +1511,9 @@ int itrace_parse_synth_opts(const struct option *opt, const char *str,
 		case 'a':
 			synth_opts->remote_access = true;
 			break;
+		case 'q':
+			synth_opts->quick += 1;
+			break;
 		case ' ':
 		case ',':
 			break;
tools/perf/util/auxtrace.h +2 -0
index f41dbdc98175..d3b5520fa992 100644
@@ -93,6 +93,7 @@ enum itrace_period_type {
  * @range_num: number of time intervals to trace
  * @error_flags: arch-specific flags to affect what errors are reported
  * @log_flags: arch-specific flags to affect what is logged
+ * @quick: quicker (less detailed) decoding
  */
 struct itrace_synth_opts {
 	bool			set;
@@ -128,6 +129,7 @@ struct itrace_synth_opts {
 	int			range_num;
 	unsigned int		error_flags;
 	unsigned int		log_flags;
+	unsigned int		quick;
 };
 
 /**
Signature · 1 line
2.25.1
author reviewer
#18 lore raw

PATCH 11/11 perf intel-pt: Add support for decoding PSB+ only

In reply to Adrian Hunter

A single q option decodes ip from only FUP/TIP packets. Make it so that repeating the q option (i.e. qq) decodes only PSB+, getting ip if there is a FUP packet within PSB+ (i.e. between PSB and PSBEND).

Example:

 $ perf record -e intel_pt//u grep -rI pudding drivers
 [ perf record: Woken up 52 times to write data ]
 [ perf record: Captured and wrote 57.870 MB perf.data ]
 $ time perf script --itrace=bi | wc -l
 58948289
 real    1m23.863s
 user    1m23.251s
 sys     0m7.452s
 $ time perf script --itrace=biq | wc -l
 3385694
 real    0m4.453s
 user    0m4.455s
 sys     0m0.328s
 $ time perf script --itrace=biqq | wc -l
 1883
 real    0m0.047s
 user    0m0.043s
 sys     0m0.009s
---
 tools/perf/Documentation/perf-intel-pt.txt     |  3 +++
 .../util/intel-pt-decoder/intel-pt-decoder.c   | 18 ++++++++++++++++++
 2 files changed, 21 insertions(+)
tools/perf/Documentation/perf-intel-pt.txt +3 -0
index 758295a7e3d6..849474629fe7 100644
@@ -973,6 +973,9 @@ useful only if the control flow of interest is represented or indicated by FUP,
 TIP, TIP.PGE, or TIP.PGD packets.  However the q option could be used to find
 time ranges that could then be decoded fully using the --time option.
 
+Repeating the q option (i.e. qq) decodes only PSB+, getting ip if there is a
+FUP packet within PSB+ (i.e. between PSB and PSBEND).
+
 
 dump option
 ~~~~~~~~~~~
tools/perf/util/intel-pt-decoder/intel-pt-decoder.c +18 -0
index ccb204b1a050..697513f35154 100644
@@ -113,6 +113,7 @@ struct intel_pt_decoder {
 	bool in_psb;
 	bool hop;
 	bool hop_psb_fup;
+	bool leap;
 	enum intel_pt_param_flags flags;
 	uint64_t pos;
 	uint64_t last_ip;
@@ -240,6 +241,7 @@ struct intel_pt_decoder *intel_pt_decoder_new(struct intel_pt_params *params)
 	decoder->return_compression = params->return_compression;
 	decoder->branch_enable      = params->branch_enable;
 	decoder->hop                = params->quick >= 1;
+	decoder->leap               = params->quick >= 2;
 
 	decoder->flags              = params->flags;
 
@@ -1903,9 +1905,18 @@ static int intel_pt_resample(struct intel_pt_decoder *decoder)
 #define HOP_RETURN	2
 #define HOP_AGAIN	3
 
+static int intel_pt_scan_for_psb(struct intel_pt_decoder *decoder);
+
 /* Hop mode: Ignore TNT, do not walk code, but get ip from FUPs and TIPs */
 static int intel_pt_hop_trace(struct intel_pt_decoder *decoder, bool *no_tip, int *err)
 {
+	/* Leap from PSB to PSB, getting ip from FUP within PSB+ */
+	if (decoder->leap && !decoder->in_psb && decoder->packet.type != INTEL_PT_PSB) {
+		*err = intel_pt_scan_for_psb(decoder);
+		if (*err)
+			return HOP_RETURN;
+	}
+
 	switch (decoder->packet.type) {
 	case INTEL_PT_TNT:
 		return HOP_IGNORE;
@@ -2681,6 +2692,7 @@ static int intel_pt_sync(struct intel_pt_decoder *decoder)
 	decoder->ip = 0;
 	intel_pt_clear_stack(&decoder->stack);
 
+leap:
 	err = intel_pt_scan_for_psb(decoder);
 	if (err)
 		return err;
@@ -2702,6 +2714,12 @@ static int intel_pt_sync(struct intel_pt_decoder *decoder)
 			decoder->pkt_state = INTEL_PT_STATE_RESAMPLE;
 		else
 			decoder->pkt_state = INTEL_PT_STATE_IN_SYNC;
+	} else if (decoder->leap) {
+		/*
+		 * In leap mode, only PSB+ is decoded, so keeping leaping to the
+		 * next PSB until there is an ip.
+		 */
+		goto leap;
 	} else {
 		return intel_pt_sync_ip(decoder);
 	}
Signature · 1 line
2.25.1
author reviewer
#12 lore raw

PATCH 7/11 perf intel-pt: Use itrace debug log flags to suppress some messages

In reply to Adrian Hunter

The "d" option may be followed by a number which has the following effect:
1 Suppress logging of perf events

Suppressing perf events is useful for decreasing the size of the log.

---
 tools/perf/Documentation/perf-intel-pt.txt |  4 +++-
 tools/perf/util/intel-pt.c                 | 19 ++++++++++++-------
 2 files changed, 15 insertions(+), 8 deletions(-)
tools/perf/Documentation/perf-intel-pt.txt +3 -1
index 0fcd8ad897b0..85a2ff804900 100644
@@ -879,7 +879,9 @@ The values may be combined by bitwise OR'ing them.
 
 The "d" option will cause the creation of a file "intel_pt.log" containing all
 decoded packets and instructions.  Note that this option slows down the decoder
-and that the resulting file may be very large.
+and that the resulting file may be very large.  The "d" option may be followed
+by a number which has the following effect:
+	1	Suppress logging of perf events
 
 In addition, the period of the "instructions" event can be specified. e.g.
 
tools/perf/util/intel-pt.c +12 -7
index a8e8e8acbcc8..d90375659244 100644
@@ -49,6 +49,8 @@
 #define INTEL_PT_ERR_SUPPRESS_OVF	1
 #define INTEL_PT_ERR_SUPPRESS_LOST	2
 
+#define INTEL_PT_LOG_SUPPRESS_EV	1
+
 struct range {
 	u64 start;
 	u64 end;
@@ -252,6 +254,11 @@ static void intel_pt_dump_sample(struct perf_session *session,
 	intel_pt_dump(pt, sample->aux_sample.data, sample->aux_sample.size);
 }
 
+static bool intel_pt_log_events(struct intel_pt *pt)
+{
+	return !(pt->synth_opts.log_flags & INTEL_PT_LOG_SUPPRESS_EV);
+}
+
 static int intel_pt_do_fix_overlap(struct intel_pt *pt, struct auxtrace_buffer *a,
 				   struct auxtrace_buffer *b)
 {
@@ -2589,10 +2596,6 @@ static int intel_pt_context_switch(struct intel_pt *pt, union perf_event *event,
 		return -EINVAL;
 	}
 
-	intel_pt_log("context_switch: cpu %d pid %d tid %d time %"PRIu64" tsc %#"PRIx64"\n",
-		     cpu, pid, tid, sample->time, perf_time_to_tsc(sample->time,
-		     &pt->tc));
-
 	ret = intel_pt_sync_switch(pt, cpu, tid, sample->time);
 	if (ret <= 0)
 		return ret;
@@ -2749,9 +2752,11 @@ static int intel_pt_process_event(struct perf_session *session,
 	if (!err && event->header.type == PERF_RECORD_TEXT_POKE)
 		err = intel_pt_text_poke(pt, event);
 
-	intel_pt_log("event %u: cpu %d time %"PRIu64" tsc %#"PRIx64" ",
-		     event->header.type, sample->cpu, sample->time, timestamp);
-	intel_pt_log_event(event);
+	if (intel_pt_enable_logging && intel_pt_log_events(pt)) {
+		intel_pt_log("event %u: cpu %d time %"PRIu64" tsc %#"PRIx64" ",
+			     event->header.type, sample->cpu, sample->time, timestamp);
+		intel_pt_log_event(event);
+	}
 
 	return err;
 }
Signature · 1 line
2.25.1
author reviewer
#1 lore raw

PATCH 1/11 perf intel-pt: Fix FUP packet state

In reply to Adrian Hunter

While walking code towards a FUP ip, the packet state is INTEL_PT_STATE_FUP or INTEL_PT_STATE_FUP_NO_TIP. That was mishandled resulting in the state becoming INTEL_PT_STATE_IN_SYNC prematurely. The result was an occasional lost EXSTOP event.

---
 .../util/intel-pt-decoder/intel-pt-decoder.c  | 21 +++++++------------
 1 file changed, 7 insertions(+), 14 deletions(-)
tools/perf/util/intel-pt-decoder/intel-pt-decoder.c +7 -14
index f8ccfd6be0ee..75c4bd74d521 100644
@@ -1164,6 +1164,7 @@ static int intel_pt_walk_fup(struct intel_pt_decoder *decoder)
 			return 0;
 		if (err == -EAGAIN ||
 		    intel_pt_fup_with_nlip(decoder, &intel_pt_insn, ip, err)) {
+			decoder->pkt_state = INTEL_PT_STATE_IN_SYNC;
 			if (intel_pt_fup_event(decoder))
 				return 0;
 			return -EAGAIN;
@@ -1942,17 +1943,13 @@ static int intel_pt_walk_trace(struct intel_pt_decoder *decoder)
 			}
 			if (decoder->set_fup_mwait)
 				no_tip = true;
+			if (no_tip)
+				decoder->pkt_state = INTEL_PT_STATE_FUP_NO_TIP;
+			else
+				decoder->pkt_state = INTEL_PT_STATE_FUP;
 			err = intel_pt_walk_fup(decoder);
-			if (err != -EAGAIN) {
-				if (err)
-					return err;
-				if (no_tip)
-					decoder->pkt_state =
-						INTEL_PT_STATE_FUP_NO_TIP;
-				else
-					decoder->pkt_state = INTEL_PT_STATE_FUP;
-				return 0;
-			}
+			if (err != -EAGAIN)
+				return err;
 			if (no_tip) {
 				no_tip = false;
 				break;
@@ -2599,15 +2596,11 @@ const struct intel_pt_state *intel_pt_decode(struct intel_pt_decoder *decoder)
 			err = intel_pt_walk_tip(decoder);
 			break;
 		case INTEL_PT_STATE_FUP:
-			decoder->pkt_state = INTEL_PT_STATE_IN_SYNC;
 			err = intel_pt_walk_fup(decoder);
 			if (err == -EAGAIN)
 				err = intel_pt_walk_fup_tip(decoder);
-			else if (!err)
-				decoder->pkt_state = INTEL_PT_STATE_FUP;
 			break;
 		case INTEL_PT_STATE_FUP_NO_TIP:
-			decoder->pkt_state = INTEL_PT_STATE_IN_SYNC;
 			err = intel_pt_walk_fup(decoder);
 			if (err == -EAGAIN)
 				err = intel_pt_walk_trace(decoder);
Signature · 1 line
2.25.1
author reviewer
#3 lore raw

PATCH 3/11 perf tools: Improve aux_output not supported error

In reply to Adrian Hunter

For example:
Before:

   $ perf record -e '{intel_pt/branch=0/,branch-loads/aux-output/ppp}' -- ls -l
   Error:
   branch-loads: PMU Hardware doesn't support sampling/overflow-interrupts. Try 'perf stat'
 After:
   $ perf record -e '{intel_pt/branch=0/,branch-loads/aux-output/ppp}' -- ls -l
   Error:
   branch-loads: PMU Hardware doesn't support 'aux_output' feature
---
 tools/perf/util/evsel.c | 4 ++++
 1 file changed, 4 insertions(+)
tools/perf/util/evsel.c +4 -0
index 9aa51a65593d..9c5c72094112 100644
@@ -2533,6 +2533,10 @@ int evsel__open_strerror(struct evsel *evsel, struct target *target,
 	 "No such device - did you specify an out-of-range profile CPU?");
 		break;
 	case EOPNOTSUPP:
+		if (evsel->core.attr.aux_output)
+			return scnprintf(msg, size,
+	"%s: PMU Hardware doesn't support 'aux_output' feature",
+					 evsel__name(evsel));
 		if (evsel->core.attr.sample_period != 0)
 			return scnprintf(msg, size,
 	"%s: PMU Hardware doesn't support sampling/overflow-interrupts. Try 'perf stat'",
Signature · 1 line
2.25.1
#6 lore raw

Re: [PATCH 05/11] perf intel-pt: Use itrace error flags to suppress some errors

In reply to Adrian Hunter
Quoted 8 lines · Adrian Hunter

The itrace "e" option may be followed by a number which has the following effect for Intel PT:
1 Suppress overflow events
2 Suppress trace data lost events
The values may be combined by bitwise OR'ing them.

Suppressing those errors can be useful for testing and debugging because they are not due to decoding.

I suspect it will be useful to more than just decoding and debugging.

But the number is not a nice user interface.

How about e[....]

like e[ol]

Also it's a bit unusual that this disables instead of enables, but ok.

-Andi

#11 lore raw

Re: [PATCH 06/11] perf auxtrace: Add optional log flags to the itrace 'd' option

In reply to Adrian Hunter

tools/perf/Documentation/itrace.txt line 55

Quoted 4 lines
tools/perf/Documentation/itrace.txt +3 -0
 	flags what kind of errors will or will not be reported.
+
+	If supported, The 'd' option may be followed by an architecture-specific
+	number which flags what kind of debug messages will or will not be logged.

Would need documentation here.

Also in the include used by the command line help.

-Andi

#15 lore raw

Re: [PATCH 09/11] perf auxtrace: Add itrace 'q' option for quicker, less detailed decoding

In reply to Adrian Hunter

tools/perf/Documentation/itrace.txt line 58

Quoted 2 lines
tools/perf/Documentation/itrace.txt +2 -0
+
+	If supported, the 'q' option may be repeated to increase the effect.

Need better documentation here. What does it mean for PT?

#17 lore raw
In reply to Adrian Hunter

tools/perf/Documentation/perf-intel-pt.txt at the beginning. This is useful to ignore initialization code. · line 974

Quoted 6 lines · Adrian Hunter
tools/perf/Documentation/perf-intel-pt.txt +6 -0
+The q option does not decode TNT packets, and does not walk object code, but
+gets the ip from FUP and TIP packets.  The q option can be used with the b and i
+options but the period is not used.  The q option decodes more quickly, but is
+useful only if the control flow of interest is represented or indicated by FUP,
+TIP, TIP.PGE, or TIP.PGD packets.  However the q option could be used to find
+time ranges that could then be decoded fully using the --time option.

Ah ok the documentation is here. Ignore previous suggestion.

Can you describe the effect in a high level without referring to packet names?

The user may not be familar with them.

So two qs will be PSB only decoding I hope?

-Andi

#19 lore raw

Re: [PATCH 11/11] perf intel-pt: Add support for decoding PSB+ only

In reply to Adrian Hunter

tools/perf/Documentation/perf-intel-pt.txt useful only if the control flow of interest is represented or indicated by FUP, · line 977

Quoted 10 lines
tools/perf/Documentation/perf-intel-pt.txt +2 -0
index 758295a7e3d6..849474629fe7 100644
@@ -973,6 +973,9 @@ useful only if the control flow of interest is represented or indicated by FUP,
 TIP, TIP.PGE, or TIP.PGD packets.  However the q option could be used to find
 time ranges that could then be decoded fully using the --time option.
 
+Repeating the q option (i.e. qq) decodes only PSB+, getting ip if there is a
+FUP packet within PSB+ (i.e. between PSB and PSBEND).

Also need high level description without PT jargon (and perhaps also reference how to configure PSB frequency)

Other than that great feature. I'll be an enthuisastic user :-)

-Andi

Quoted 1 line
author reviewer
#7 lore raw

Re: [PATCH 05/11] perf intel-pt: Use itrace error flags to suppress some errors

In reply to Andi Kleen
Quoted 10 lines · Andi Kleen
Quoted 8 lines · Adrian Hunter

The itrace "e" option may be followed by a number which has the following effect for Intel PT:
1 Suppress overflow events
2 Suppress trace data lost events
The values may be combined by bitwise OR'ing them.

Suppressing those errors can be useful for testing and debugging because they are not due to decoding.

Quoted 7 lines

I suspect it will be useful to more than just decoding and debugging.

But the number is not a nice user interface.

How about e[....]

like e[ol]

Do you mean literally square-brackets? If you were really unlucky you might get pathname expansion with that.

Quoted 5 lines

Also it's a bit unusual that this disables instead of enables, but ok.

-Andi

author reviewer
#8 lore raw

Re: [PATCH 05/11] perf intel-pt: Use itrace error flags to suppress some errors

Quoted 13 lines · Adrian Hunter
Quoted 12 lines · Andi Kleen
Quoted 8 lines · Adrian Hunter

The itrace "e" option may be followed by a number which has the following effect for Intel PT:
1 Suppress overflow events
2 Suppress trace data lost events
The values may be combined by bitwise OR'ing them.

Suppressing those errors can be useful for testing and debugging because they are not due to decoding.

I suspect it will be useful to more than just decoding and debugging.

Quoted 11 lines
Quoted 5 lines

But the number is not a nice user interface.

How about e[....]

like e[ol]

Do you mean literally square-brackets? If you were really unlucky you might get pathname expansion with that.

Quoted 2 lines

Also it's a bit unusual that this disables instead of enables, but ok.

What about prefixing each flag with - i.e.

e-o
e-l
e-o-l

#9 lore raw

Re: [PATCH 05/11] perf intel-pt: Use itrace error flags to suppress some errors

Quoted 5 lines

What about prefixing each flag with - i.e.

e-o
e-l
e-o-l

I was thinking square brackets, but yes the shell collision is a fair point.
- should work too.

-andi