[PATCH 00/37] perf intel-pt: Power events and PTWRITE

[PATCH 00/37] perf intel-pt: Power events and PTWRITE

[Adrian Hunter]

Hi

Here are some patches to support Intel PT Power events and PTWRITE.
Power events report changes to C-state but I have also added support
for the existing CBR (core-to-bus ratio) packet and included that
when outputting power events.  The PTWRITE packet is associated with
the new ptwrite instruction, which is essentially just a way to stuff
a 32 or 64 bit value into the PT trace.  More details can be found in
the patches that add documentation and in the Intel SDM.

As far as I know, there isn't any hardware released that supports
the new packets, however the CBR packet is not new and will now be
visible by default.

There are also some miscellaneous improvements to Intel PT.

There are 3 patches that introduce a new way to define synthesized
events that carry arbitrary raw_data.  I sent an RFC about that earlier.
The 3 patches are:
      perf script: Add 'synth' event type for synthesized events
      perf script: Add 'synth' field for synthesized event payloads
      perf script: Add synthesized Intel PT power and ptwrite events


Adrian Hunter (37):
      perf intel-pt: Allow decoding with branch tracing disabled
      perf intel-pt: Add default config for pass-through branch enable
      perf intel-pt: Add documentation for new config terms
      perf intel-pt: Fix missing stack clear
      perf intel-pt: Ensure IP is zero when state is INTEL_PT_STATE_NO_IP
      perf intel-pt: Fix last_ip usage
      perf intel-pt: Ensure never to set 'last_ip' when packet 'count' is zero
      perf intel-pt: Use FUP always when scanning for an IP
      perf intel-pt: Add missing __fallthrough
      perf intel-pt: Clear FUP flag on error
      perf intel-pt: Add decoder support for ptwrite and power event packets
      perf intel-pt: Add reserved byte to CBR packet payload
      perf intel-pt: Move decoder error setting into one condition
      perf intel-pt: Add decoder support for CBR events
      perf intel-pt: Remove redundant initial_skip checks
      perf intel-pt: Fix transactions_sample_type
      perf tools: Fix message because cpu list option is -C not -c
      perf script: Fix message because field list option is -F not -f
      perf script: Add 'synth' event type for synthesized events
      perf script: Add 'synth' field for synthesized event payloads
      tools include: Add byte-swapping macros to kernel.h
      perf auxtrace: Add itrace option to output ptwrite events
      perf auxtrace: Add itrace option to output power events
      perf script: Add synthesized Intel PT power and ptwrite events
      perf intel-pt: Factor out common code synthesizing event samples
      perf intel-pt: Remove unused instructions_sample_period
      perf intel-pt: Join needlessly wrapped lines
      perf intel-pt: Tidy Intel PT evsel lookup into separate function
      perf intel-pt: Tidy messages into called function intel_pt_synth_event()
      perf intel-pt: Factor out intel_pt_set_event_name()
      perf intel-pt: Move code in intel_pt_synth_events() to simplify attr setting
      perf intel-pt: Synthesize new power and ptwrite events
      perf intel-pt: Add example script for power events and PTWRITE
      perf intel-pt: Update documentation to include new ptwrite and power events
      perf intel-pt: Do not use TSC packets for calculating CPU cycles to TSC
      perf auxtrace: Add CPU filter support
      perf intel-pt: Improve sample timestamp

 tools/include/linux/kernel.h                       |  35 +-
 tools/perf/Documentation/intel-pt.txt              |  78 ++-
 tools/perf/Documentation/itrace.txt                |   8 +-
 tools/perf/Documentation/perf-script.txt           |   6 +-
 tools/perf/arch/x86/util/intel-pt.c                |   5 +
 tools/perf/builtin-report.c                        |   1 +
 tools/perf/builtin-script.c                        | 207 ++++++-
 .../perf/scripts/python/bin/intel-pt-events-record |  13 +
 .../perf/scripts/python/bin/intel-pt-events-report |   3 +
 tools/perf/scripts/python/intel-pt-events.py       | 128 ++++
 tools/perf/util/auxtrace.c                         |  18 +
 tools/perf/util/auxtrace.h                         |   6 +
 tools/perf/util/event.h                            |  93 +++
 .../perf/util/intel-pt-decoder/intel-pt-decoder.c  | 304 +++++++++-
 .../perf/util/intel-pt-decoder/intel-pt-decoder.h  |  13 +
 .../util/intel-pt-decoder/intel-pt-pkt-decoder.c   | 110 +++-
 .../util/intel-pt-decoder/intel-pt-pkt-decoder.h   |   7 +
 tools/perf/util/intel-pt.c                         | 642 +++++++++++++++------
 tools/perf/util/session.c                          |   2 +-
 19 files changed, 1435 insertions(+), 244 deletions(-)
 create mode 100644 tools/perf/scripts/python/bin/intel-pt-events-record
 create mode 100644 tools/perf/scripts/python/bin/intel-pt-events-report
 create mode 100644 tools/perf/scripts/python/intel-pt-events.py


Regards
Adrian

[PATCH 01/37] perf intel-pt: Allow decoding with branch tracing disabled

[Adrian Hunter]

The kernel now supports the disabling of branch tracing, however the
decoder assumes branch tracing is always enabled. Pass through a parameter
to indicate whether branch tracing is enabled and use it to avoid cases
when the decoder is expecting branch packets. There are 2 such cases.
First, FUP packets which can bind to an IP even when there is no branch
tracing. Secondly, the decoder will try to use branch packets to find an IP
to start decoding or to recover from errors.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/util/intel-pt-decoder/intel-pt-decoder.c | 13 +++++++++++++
 tools/perf/util/intel-pt-decoder/intel-pt-decoder.h |  1 +
 tools/perf/util/intel-pt.c                          | 14 ++++++++++++++
 3 files changed, 28 insertions(+)

diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
index 7cf7f7aca4d2..9545b1a1872c 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
@@ -87,6 +87,7 @@ struct intel_pt_decoder {
 	const unsigned char *buf;
 	size_t len;
 	bool return_compression;
+	bool branch_enable;
 	bool mtc_insn;
 	bool pge;
 	bool have_tma;
@@ -192,6 +193,7 @@ struct intel_pt_decoder *intel_pt_decoder_new(struct intel_pt_params *params)
 	decoder->pgd_ip             = params->pgd_ip;
 	decoder->data               = params->data;
 	decoder->return_compression = params->return_compression;
+	decoder->branch_enable      = params->branch_enable;
 
 	decoder->period             = params->period;
 	decoder->period_type        = params->period_type;
@@ -1625,6 +1627,10 @@ static int intel_pt_walk_trace(struct intel_pt_decoder *decoder)
 				break;
 			}
 			intel_pt_set_last_ip(decoder);
+			if (!decoder->branch_enable) {
+				decoder->ip = decoder->last_ip;
+				break;
+			}
 			err = intel_pt_walk_fup(decoder);
 			if (err != -EAGAIN) {
 				if (err)
@@ -1935,6 +1941,13 @@ static int intel_pt_sync_ip(struct intel_pt_decoder *decoder)
 {
 	int err;
 
+	if (!decoder->branch_enable) {
+		decoder->pkt_state = INTEL_PT_STATE_IN_SYNC;
+		decoder->overflow = false;
+		decoder->state.type = 0; /* Do not have a sample */
+		return 0;
+	}
+
 	intel_pt_log("Scanning for full IP\n");
 	err = intel_pt_walk_to_ip(decoder);
 	if (err)
diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.h b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.h
index e90619a43c0c..add3bed58349 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.h
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.h
@@ -87,6 +87,7 @@ struct intel_pt_params {
 	bool (*pgd_ip)(uint64_t ip, void *data);
 	void *data;
 	bool return_compression;
+	bool branch_enable;
 	uint64_t period;
 	enum intel_pt_period_type period_type;
 	unsigned max_non_turbo_ratio;
diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c
index 4c7718f87a08..5c59b8c6a719 100644
--- a/tools/perf/util/intel-pt.c
+++ b/tools/perf/util/intel-pt.c
@@ -668,6 +668,19 @@ static bool intel_pt_return_compression(struct intel_pt *pt)
 	return true;
 }
 
+static bool intel_pt_branch_enable(struct intel_pt *pt)
+{
+	struct perf_evsel *evsel;
+	u64 config;
+
+	evlist__for_each_entry(pt->session->evlist, evsel) {
+		if (intel_pt_get_config(pt, &evsel->attr, &config) &&
+		    (config & 1) && !(config & 0x2000))
+			return false;
+	}
+	return true;
+}
+
 static unsigned int intel_pt_mtc_period(struct intel_pt *pt)
 {
 	struct perf_evsel *evsel;
@@ -799,6 +812,7 @@ static struct intel_pt_queue *intel_pt_alloc_queue(struct intel_pt *pt,
 	params.walk_insn = intel_pt_walk_next_insn;
 	params.data = ptq;
 	params.return_compression = intel_pt_return_compression(pt);
+	params.branch_enable = intel_pt_branch_enable(pt);
 	params.max_non_turbo_ratio = pt->max_non_turbo_ratio;
 	params.mtc_period = intel_pt_mtc_period(pt);
 	params.tsc_ctc_ratio_n = pt->tsc_ctc_ratio_n;
-- 
1.9.1

[PATCH 02/37] perf intel-pt: Add default config for pass-through branch enable

[Adrian Hunter]

Branch tracing is enabled by default, so a fake config bit called 'pt'
(pass-through) was added to allow the 'branch enable' bit to have affect.
Add default config 'pt,branch' which will allow users to disable branch
tracing using 'branch=0' instead of having to specify 'pt,branch=0'.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/arch/x86/util/intel-pt.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/tools/perf/arch/x86/util/intel-pt.c b/tools/perf/arch/x86/util/intel-pt.c
index f630de0206a1..007621a7e420 100644
--- a/tools/perf/arch/x86/util/intel-pt.c
+++ b/tools/perf/arch/x86/util/intel-pt.c
@@ -196,6 +196,7 @@ static u64 intel_pt_default_config(struct perf_pmu *intel_pt_pmu)
 	int psb_cyc, psb_periods, psb_period;
 	int pos = 0;
 	u64 config;
+	char c;
 
 	pos += scnprintf(buf + pos, sizeof(buf) - pos, "tsc");
 
@@ -229,6 +230,10 @@ static u64 intel_pt_default_config(struct perf_pmu *intel_pt_pmu)
 		}
 	}
 
+	if (perf_pmu__scan_file(intel_pt_pmu, "format/pt", "%c", &c) == 1 &&
+	    perf_pmu__scan_file(intel_pt_pmu, "format/branch", "%c", &c) == 1)
+		pos += scnprintf(buf + pos, sizeof(buf) - pos, ",pt,branch");
+
 	pr_debug2("%s default config: %s\n", intel_pt_pmu->name, buf);
 
 	intel_pt_parse_terms(&intel_pt_pmu->format, buf, &config);
-- 
1.9.1

[PATCH 03/37] perf intel-pt: Add documentation for new config terms

[Adrian Hunter]

Add documentation for new config terms.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/Documentation/intel-pt.txt | 36 +++++++++++++++++++++++++++++++++++
 1 file changed, 36 insertions(+)

diff --git a/tools/perf/Documentation/intel-pt.txt b/tools/perf/Documentation/intel-pt.txt
index b0b3007d3c9c..d157dee7a4ec 100644
--- a/tools/perf/Documentation/intel-pt.txt
+++ b/tools/perf/Documentation/intel-pt.txt
@@ -364,6 +364,42 @@ cyc_thresh	Specifies how frequently CYC packets are produced - see cyc
 
 		CYC packets are not requested by default.
 
+pt		Specifies pass-through which enables the 'branch' config term.
+
+		The default config selects 'pt' if it is available, so a user will
+		never need to specify this term.
+
+branch		Enable branch tracing.  Branch tracing is enabled by default so to
+		disable branch tracing use 'branch=0'.
+
+		The default config selects 'branch' if it is available.
+
+ptw		Enable PTWRITE packets which are produced when a ptwrite instruction
+		is executed.
+
+		Support for this feature is indicated by:
+
+			/sys/bus/event_source/devices/intel_pt/caps/ptwrite
+
+		which contains "1" if the feature is supported and
+		"0" otherwise.
+
+fup_on_ptw	Enable a FUP packet to follow the PTWRITE packet.  The FUP packet
+		provides the address of the ptwrite instruction.  In the absence of
+		fup_on_ptw, the decoder will use the address of the previous branch
+		if branch tracing is enabled, otherwise the address will be zero.
+		Note that fup_on_ptw will work even when branch tracing is disabled.
+
+pwr_evt		Enable power events.  The power events provide information about
+		changes to the CPU C-state.
+
+		Support for this feature is indicated by:
+
+			/sys/bus/event_source/devices/intel_pt/caps/power_event_trace
+
+		which contains "1" if the feature is supported and
+		"0" otherwise.
+
 
 new snapshot option
 -------------------
-- 
1.9.1

[PATCH 04/37] perf intel-pt: Fix missing stack clear

[Adrian Hunter]

The return compression stack must be cleared whenever there is a PSB. Fix
one case where that was not happening.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/util/intel-pt-decoder/intel-pt-decoder.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
index 9545b1a1872c..af20719b3373 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
@@ -1916,6 +1916,7 @@ static int intel_pt_walk_to_ip(struct intel_pt_decoder *decoder)
 			break;
 
 		case INTEL_PT_PSB:
+			intel_pt_clear_stack(&decoder->stack);
 			err = intel_pt_walk_psb(decoder);
 			if (err)
 				return err;
-- 
1.9.1

[PATCH 05/37] perf intel-pt: Ensure IP is zero when state is INTEL_PT_STATE_NO_IP

[Adrian Hunter]

A value of zero is used to indicate that there is no IP. Ensure the value
is zero when the state is INTEL_PT_STATE_NO_IP.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/util/intel-pt-decoder/intel-pt-decoder.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
index af20719b3373..0f7def56149a 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
@@ -2108,6 +2108,7 @@ const struct intel_pt_state *intel_pt_decode(struct intel_pt_decoder *decoder)
 			break;
 		case INTEL_PT_STATE_NO_IP:
 			decoder->last_ip = 0;
+			decoder->ip = 0;
 			/* Fall through */
 		case INTEL_PT_STATE_ERR_RESYNC:
 			err = intel_pt_sync_ip(decoder);
-- 
1.9.1

[PATCH 06/37] perf intel-pt: Fix last_ip usage

[Adrian Hunter]

Intel PT uses IP compression based on the last IP. For decoding purposes,
'last IP' is considered to be reset to zero whenever there is a
synchronization packet (PSB). The decoder wasn't doing that, and was
treating the zero value to mean that there was no last IP, whereas
compression can be done against the zero value. Fix by setting last_ip to
zero when a PSB is received and keep track of have_last_ip.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/util/intel-pt-decoder/intel-pt-decoder.c | 13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)

diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
index 0f7def56149a..f8fada859f12 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
@@ -93,6 +93,7 @@ struct intel_pt_decoder {
 	bool have_tma;
 	bool have_cyc;
 	bool fixup_last_mtc;
+	bool have_last_ip;
 	uint64_t pos;
 	uint64_t last_ip;
 	uint64_t ip;
@@ -400,6 +401,7 @@ static uint64_t intel_pt_calc_ip(const struct intel_pt_pkt *packet,
 static inline void intel_pt_set_last_ip(struct intel_pt_decoder *decoder)
 {
 	decoder->last_ip = intel_pt_calc_ip(&decoder->packet, decoder->last_ip);
+	decoder->have_last_ip = true;
 }
 
 static inline void intel_pt_set_ip(struct intel_pt_decoder *decoder)
@@ -1656,6 +1658,8 @@ static int intel_pt_walk_trace(struct intel_pt_decoder *decoder)
 			break;
 
 		case INTEL_PT_PSB:
+			decoder->last_ip = 0;
+			decoder->have_last_ip = true;
 			intel_pt_clear_stack(&decoder->stack);
 			err = intel_pt_walk_psbend(decoder);
 			if (err == -EAGAIN)
@@ -1736,7 +1740,7 @@ static int intel_pt_walk_trace(struct intel_pt_decoder *decoder)
 
 static inline bool intel_pt_have_ip(struct intel_pt_decoder *decoder)
 {
-	return decoder->last_ip || decoder->packet.count == 0 ||
+	return decoder->have_last_ip || decoder->packet.count == 0 ||
 	       decoder->packet.count == 3 || decoder->packet.count == 6;
 }
 
@@ -1866,7 +1870,7 @@ static int intel_pt_walk_to_ip(struct intel_pt_decoder *decoder)
 				if (decoder->ip)
 					return 0;
 			}
-			if (decoder->packet.count)
+			if (decoder->packet.count && decoder->have_last_ip)
 				intel_pt_set_last_ip(decoder);
 			break;
 
@@ -1916,6 +1920,8 @@ static int intel_pt_walk_to_ip(struct intel_pt_decoder *decoder)
 			break;
 
 		case INTEL_PT_PSB:
+			decoder->last_ip = 0;
+			decoder->have_last_ip = true;
 			intel_pt_clear_stack(&decoder->stack);
 			err = intel_pt_walk_psb(decoder);
 			if (err)
@@ -2057,6 +2063,7 @@ static int intel_pt_sync(struct intel_pt_decoder *decoder)
 
 	decoder->pge = false;
 	decoder->continuous_period = false;
+	decoder->have_last_ip = false;
 	decoder->last_ip = 0;
 	decoder->ip = 0;
 	intel_pt_clear_stack(&decoder->stack);
@@ -2065,6 +2072,7 @@ static int intel_pt_sync(struct intel_pt_decoder *decoder)
 	if (err)
 		return err;
 
+	decoder->have_last_ip = true;
 	decoder->pkt_state = INTEL_PT_STATE_NO_IP;
 
 	err = intel_pt_walk_psb(decoder);
@@ -2107,6 +2115,7 @@ const struct intel_pt_state *intel_pt_decode(struct intel_pt_decoder *decoder)
 			err = intel_pt_sync(decoder);
 			break;
 		case INTEL_PT_STATE_NO_IP:
+			decoder->have_last_ip = false;
 			decoder->last_ip = 0;
 			decoder->ip = 0;
 			/* Fall through */
-- 
1.9.1

[PATCH 07/37] perf intel-pt: Ensure never to set 'last_ip' when packet 'count' is zero

[Adrian Hunter]

Intel PT uses IP compression based on the last IP. For decoding purposes,
'last IP' is not updated when a branch target has been suppressed, which is
indicated by IPBytes == 0. IPBytes is stored in the packet 'count', so
ensure never to set 'last_ip' when packet 'count' is zero.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/util/intel-pt-decoder/intel-pt-decoder.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
index f8fada859f12..4359177a0504 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
@@ -1450,7 +1450,8 @@ static int intel_pt_walk_psbend(struct intel_pt_decoder *decoder)
 
 		case INTEL_PT_FUP:
 			decoder->pge = true;
-			intel_pt_set_last_ip(decoder);
+			if (decoder->packet.count)
+				intel_pt_set_last_ip(decoder);
 			break;
 
 		case INTEL_PT_MODE_TSX:
@@ -1740,8 +1741,9 @@ static int intel_pt_walk_trace(struct intel_pt_decoder *decoder)
 
 static inline bool intel_pt_have_ip(struct intel_pt_decoder *decoder)
 {
-	return decoder->have_last_ip || decoder->packet.count == 0 ||
-	       decoder->packet.count == 3 || decoder->packet.count == 6;
+	return decoder->packet.count &&
+	       (decoder->have_last_ip || decoder->packet.count == 3 ||
+		decoder->packet.count == 6);
 }
 
 /* Walk PSB+ packets to get in sync. */
-- 
1.9.1

[PATCH 08/37] perf intel-pt: Use FUP always when scanning for an IP

[Adrian Hunter]

The decoder will try to use branch packets to find an IP to start decoding
or to recover from errors. Currently the FUP packet is used only in the
case of an overflow, however there is no reason for that to be a special
case. So just use FUP always when scanning for an IP.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/util/intel-pt-decoder/intel-pt-decoder.c | 12 ++++--------
 1 file changed, 4 insertions(+), 8 deletions(-)

diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
index 4359177a0504..737076340100 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
@@ -1866,14 +1866,10 @@ static int intel_pt_walk_to_ip(struct intel_pt_decoder *decoder)
 			break;
 
 		case INTEL_PT_FUP:
-			if (decoder->overflow) {
-				if (intel_pt_have_ip(decoder))
-					intel_pt_set_ip(decoder);
-				if (decoder->ip)
-					return 0;
-			}
-			if (decoder->packet.count && decoder->have_last_ip)
-				intel_pt_set_last_ip(decoder);
+			if (intel_pt_have_ip(decoder))
+				intel_pt_set_ip(decoder);
+			if (decoder->ip)
+				return 0;
 			break;
 
 		case INTEL_PT_MTC:
-- 
1.9.1

[PATCH 09/37] perf intel-pt: Add missing __fallthrough

[Adrian Hunter]

perf tools uses __fallthrough. Add missing  __fallthrough to a switch
statement.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/util/intel-pt-decoder/intel-pt-decoder.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
index 737076340100..dbcb07ebe813 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
@@ -2116,7 +2116,7 @@ const struct intel_pt_state *intel_pt_decode(struct intel_pt_decoder *decoder)
 			decoder->have_last_ip = false;
 			decoder->last_ip = 0;
 			decoder->ip = 0;
-			/* Fall through */
+			__fallthrough;
 		case INTEL_PT_STATE_ERR_RESYNC:
 			err = intel_pt_sync_ip(decoder);
 			break;
-- 
1.9.1

[PATCH 10/37] perf intel-pt: Clear FUP flag on error

[Adrian Hunter]

Sometimes a FUP packet is associated with a TSX transaction and a flag is
set to indicate that. Ensure that flag is cleared on any error condition
because at that point the decoder can no longer assume it is correct.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/util/intel-pt-decoder/intel-pt-decoder.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
index dbcb07ebe813..7cdc8853fc1e 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
@@ -1946,6 +1946,8 @@ static int intel_pt_sync_ip(struct intel_pt_decoder *decoder)
 {
 	int err;
 
+	decoder->set_fup_tx_flags = false;
+
 	if (!decoder->branch_enable) {
 		decoder->pkt_state = INTEL_PT_STATE_IN_SYNC;
 		decoder->overflow = false;
-- 
1.9.1

[PATCH 11/37] perf intel-pt: Add decoder support for ptwrite and power event packets

[Adrian Hunter]

Add decoder support for PTWRITE, MWAIT, PWRE, PWRX and EXSTOP packets. This
patch only affects the decoder, so the tools still do not select or consume
the new information. That is added in subsequent patches.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 .../perf/util/intel-pt-decoder/intel-pt-decoder.c  | 176 ++++++++++++++++++++-
 .../perf/util/intel-pt-decoder/intel-pt-decoder.h  |  10 ++
 .../util/intel-pt-decoder/intel-pt-pkt-decoder.c   | 108 +++++++++++++
 .../util/intel-pt-decoder/intel-pt-pkt-decoder.h   |   7 +
 4 files changed, 293 insertions(+), 8 deletions(-)

diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
index 7cdc8853fc1e..e7f4016a5b31 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
@@ -138,8 +138,15 @@ struct intel_pt_decoder {
 	bool continuous_period;
 	bool overflow;
 	bool set_fup_tx_flags;
+	bool set_fup_ptw;
+	bool set_fup_mwait;
+	bool set_fup_pwre;
+	bool set_fup_exstop;
 	unsigned int fup_tx_flags;
 	unsigned int tx_flags;
+	uint64_t fup_ptw_payload;
+	uint64_t fup_mwait_payload;
+	uint64_t fup_pwre_payload;
 	uint64_t timestamp_insn_cnt;
 	uint64_t stuck_ip;
 	int no_progress;
@@ -639,6 +646,8 @@ static int intel_pt_calc_cyc_cb(struct intel_pt_pkt_info *pkt_info)
 	case INTEL_PT_PAD:
 	case INTEL_PT_VMCS:
 	case INTEL_PT_MNT:
+	case INTEL_PT_PTWRITE:
+	case INTEL_PT_PTWRITE_IP:
 		return 0;
 
 	case INTEL_PT_MTC:
@@ -737,6 +746,11 @@ static int intel_pt_calc_cyc_cb(struct intel_pt_pkt_info *pkt_info)
 
 	case INTEL_PT_TIP_PGD:
 	case INTEL_PT_TRACESTOP:
+	case INTEL_PT_EXSTOP:
+	case INTEL_PT_EXSTOP_IP:
+	case INTEL_PT_MWAIT:
+	case INTEL_PT_PWRE:
+	case INTEL_PT_PWRX:
 	case INTEL_PT_OVF:
 	case INTEL_PT_BAD: /* Does not happen */
 	default:
@@ -994,6 +1008,57 @@ static int intel_pt_walk_insn(struct intel_pt_decoder *decoder,
 	return err;
 }
 
+static bool intel_pt_fup_event(struct intel_pt_decoder *decoder)
+{
+	bool ret = false;
+
+	if (decoder->set_fup_tx_flags) {
+		decoder->set_fup_tx_flags = false;
+		decoder->tx_flags = decoder->fup_tx_flags;
+		decoder->state.type = INTEL_PT_TRANSACTION;
+		decoder->state.from_ip = decoder->ip;
+		decoder->state.to_ip = 0;
+		decoder->state.flags = decoder->fup_tx_flags;
+		return true;
+	}
+	if (decoder->set_fup_ptw) {
+		decoder->set_fup_ptw = false;
+		decoder->state.type = INTEL_PT_PTW;
+		decoder->state.flags |= INTEL_PT_FUP_IP;
+		decoder->state.from_ip = decoder->ip;
+		decoder->state.to_ip = 0;
+		decoder->state.ptw_payload = decoder->fup_ptw_payload;
+		return true;
+	}
+	if (decoder->set_fup_mwait) {
+		decoder->set_fup_mwait = false;
+		decoder->state.type = INTEL_PT_MWAIT_OP;
+		decoder->state.from_ip = decoder->ip;
+		decoder->state.to_ip = 0;
+		decoder->state.mwait_payload = decoder->fup_mwait_payload;
+		ret = true;
+	}
+	if (decoder->set_fup_pwre) {
+		decoder->set_fup_pwre = false;
+		decoder->state.type |= INTEL_PT_PWR_ENTRY;
+		decoder->state.type &= ~INTEL_PT_BRANCH;
+		decoder->state.from_ip = decoder->ip;
+		decoder->state.to_ip = 0;
+		decoder->state.pwre_payload = decoder->fup_pwre_payload;
+		ret = true;
+	}
+	if (decoder->set_fup_exstop) {
+		decoder->set_fup_exstop = false;
+		decoder->state.type |= INTEL_PT_EX_STOP;
+		decoder->state.type &= ~INTEL_PT_BRANCH;
+		decoder->state.flags |= INTEL_PT_FUP_IP;
+		decoder->state.from_ip = decoder->ip;
+		decoder->state.to_ip = 0;
+		ret = true;
+	}
+	return ret;
+}
+
 static int intel_pt_walk_fup(struct intel_pt_decoder *decoder)
 {
 	struct intel_pt_insn intel_pt_insn;
@@ -1007,15 +1072,8 @@ static int intel_pt_walk_fup(struct intel_pt_decoder *decoder)
 		if (err == INTEL_PT_RETURN)
 			return 0;
 		if (err == -EAGAIN) {
-			if (decoder->set_fup_tx_flags) {
-				decoder->set_fup_tx_flags = false;
-				decoder->tx_flags = decoder->fup_tx_flags;
-				decoder->state.type = INTEL_PT_TRANSACTION;
-				decoder->state.from_ip = decoder->ip;
-				decoder->state.to_ip = 0;
-				decoder->state.flags = decoder->fup_tx_flags;
+			if (intel_pt_fup_event(decoder))
 				return 0;
-			}
 			return err;
 		}
 		decoder->set_fup_tx_flags = false;
@@ -1421,6 +1479,13 @@ static int intel_pt_walk_psbend(struct intel_pt_decoder *decoder)
 		case INTEL_PT_TRACESTOP:
 		case INTEL_PT_BAD:
 		case INTEL_PT_PSB:
+		case INTEL_PT_PTWRITE:
+		case INTEL_PT_PTWRITE_IP:
+		case INTEL_PT_EXSTOP:
+		case INTEL_PT_EXSTOP_IP:
+		case INTEL_PT_MWAIT:
+		case INTEL_PT_PWRE:
+		case INTEL_PT_PWRX:
 			decoder->have_tma = false;
 			intel_pt_log("ERROR: Unexpected packet\n");
 			return -EAGAIN;
@@ -1502,6 +1567,13 @@ static int intel_pt_walk_fup_tip(struct intel_pt_decoder *decoder)
 		case INTEL_PT_MODE_TSX:
 		case INTEL_PT_BAD:
 		case INTEL_PT_PSBEND:
+		case INTEL_PT_PTWRITE:
+		case INTEL_PT_PTWRITE_IP:
+		case INTEL_PT_EXSTOP:
+		case INTEL_PT_EXSTOP_IP:
+		case INTEL_PT_MWAIT:
+		case INTEL_PT_PWRE:
+		case INTEL_PT_PWRX:
 			intel_pt_log("ERROR: Missing TIP after FUP\n");
 			decoder->pkt_state = INTEL_PT_STATE_ERR3;
 			return -ENOENT;
@@ -1632,8 +1704,13 @@ static int intel_pt_walk_trace(struct intel_pt_decoder *decoder)
 			intel_pt_set_last_ip(decoder);
 			if (!decoder->branch_enable) {
 				decoder->ip = decoder->last_ip;
+				if (intel_pt_fup_event(decoder))
+					return 0;
+				no_tip = false;
 				break;
 			}
+			if (decoder->set_fup_mwait)
+				no_tip = true;
 			err = intel_pt_walk_fup(decoder);
 			if (err != -EAGAIN) {
 				if (err)
@@ -1733,6 +1810,71 @@ static int intel_pt_walk_trace(struct intel_pt_decoder *decoder)
 		case INTEL_PT_PAD:
 			break;
 
+		case INTEL_PT_PTWRITE_IP:
+			decoder->fup_ptw_payload = decoder->packet.payload;
+			err = intel_pt_get_next_packet(decoder);
+			if (err)
+				return err;
+			if (decoder->packet.type == INTEL_PT_FUP) {
+				decoder->set_fup_ptw = true;
+				no_tip = true;
+			} else {
+				intel_pt_log_at("ERROR: Missing FUP after PTWRITE",
+						decoder->pos);
+			}
+			goto next;
+
+		case INTEL_PT_PTWRITE:
+			decoder->state.type = INTEL_PT_PTW;
+			decoder->state.from_ip = decoder->ip;
+			decoder->state.to_ip = 0;
+			decoder->state.ptw_payload = decoder->packet.payload;
+			return 0;
+
+		case INTEL_PT_MWAIT:
+			decoder->fup_mwait_payload = decoder->packet.payload;
+			decoder->set_fup_mwait = true;
+			break;
+
+		case INTEL_PT_PWRE:
+			if (decoder->set_fup_mwait) {
+				decoder->fup_pwre_payload =
+							decoder->packet.payload;
+				decoder->set_fup_pwre = true;
+				break;
+			}
+			decoder->state.type = INTEL_PT_PWR_ENTRY;
+			decoder->state.from_ip = decoder->ip;
+			decoder->state.to_ip = 0;
+			decoder->state.pwrx_payload = decoder->packet.payload;
+			return 0;
+
+		case INTEL_PT_EXSTOP_IP:
+			err = intel_pt_get_next_packet(decoder);
+			if (err)
+				return err;
+			if (decoder->packet.type == INTEL_PT_FUP) {
+				decoder->set_fup_exstop = true;
+				no_tip = true;
+			} else {
+				intel_pt_log_at("ERROR: Missing FUP after EXSTOP",
+						decoder->pos);
+			}
+			goto next;
+
+		case INTEL_PT_EXSTOP:
+			decoder->state.type = INTEL_PT_EX_STOP;
+			decoder->state.from_ip = decoder->ip;
+			decoder->state.to_ip = 0;
+			return 0;
+
+		case INTEL_PT_PWRX:
+			decoder->state.type = INTEL_PT_PWR_EXIT;
+			decoder->state.from_ip = decoder->ip;
+			decoder->state.to_ip = 0;
+			decoder->state.pwrx_payload = decoder->packet.payload;
+			return 0;
+
 		default:
 			return intel_pt_bug(decoder);
 		}
@@ -1762,6 +1904,13 @@ static int intel_pt_walk_psb(struct intel_pt_decoder *decoder)
 			__fallthrough;
 		case INTEL_PT_TIP_PGE:
 		case INTEL_PT_TIP:
+		case INTEL_PT_PTWRITE:
+		case INTEL_PT_PTWRITE_IP:
+		case INTEL_PT_EXSTOP:
+		case INTEL_PT_EXSTOP_IP:
+		case INTEL_PT_MWAIT:
+		case INTEL_PT_PWRE:
+		case INTEL_PT_PWRX:
 			intel_pt_log("ERROR: Unexpected packet\n");
 			return -ENOENT;
 
@@ -1936,6 +2085,13 @@ static int intel_pt_walk_to_ip(struct intel_pt_decoder *decoder)
 		case INTEL_PT_VMCS:
 		case INTEL_PT_MNT:
 		case INTEL_PT_PAD:
+		case INTEL_PT_PTWRITE:
+		case INTEL_PT_PTWRITE_IP:
+		case INTEL_PT_EXSTOP:
+		case INTEL_PT_EXSTOP_IP:
+		case INTEL_PT_MWAIT:
+		case INTEL_PT_PWRE:
+		case INTEL_PT_PWRX:
 		default:
 			break;
 		}
@@ -1947,6 +2103,10 @@ static int intel_pt_sync_ip(struct intel_pt_decoder *decoder)
 	int err;
 
 	decoder->set_fup_tx_flags = false;
+	decoder->set_fup_ptw = false;
+	decoder->set_fup_mwait = false;
+	decoder->set_fup_pwre = false;
+	decoder->set_fup_exstop = false;
 
 	if (!decoder->branch_enable) {
 		decoder->pkt_state = INTEL_PT_STATE_IN_SYNC;
diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.h b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.h
index add3bed58349..414c88e9e0da 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.h
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.h
@@ -25,11 +25,17 @@
 #define INTEL_PT_IN_TX		(1 << 0)
 #define INTEL_PT_ABORT_TX	(1 << 1)
 #define INTEL_PT_ASYNC		(1 << 2)
+#define INTEL_PT_FUP_IP		(1 << 3)
 
 enum intel_pt_sample_type {
 	INTEL_PT_BRANCH		= 1 << 0,
 	INTEL_PT_INSTRUCTION	= 1 << 1,
 	INTEL_PT_TRANSACTION	= 1 << 2,
+	INTEL_PT_PTW		= 1 << 3,
+	INTEL_PT_MWAIT_OP	= 1 << 4,
+	INTEL_PT_PWR_ENTRY	= 1 << 5,
+	INTEL_PT_EX_STOP	= 1 << 6,
+	INTEL_PT_PWR_EXIT	= 1 << 7,
 };
 
 enum intel_pt_period_type {
@@ -63,6 +69,10 @@ struct intel_pt_state {
 	uint64_t timestamp;
 	uint64_t est_timestamp;
 	uint64_t trace_nr;
+	uint64_t ptw_payload;
+	uint64_t mwait_payload;
+	uint64_t pwre_payload;
+	uint64_t pwrx_payload;
 	uint32_t flags;
 	enum intel_pt_insn_op insn_op;
 	int insn_len;
diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-pkt-decoder.c b/tools/perf/util/intel-pt-decoder/intel-pt-pkt-decoder.c
index 7528ae4f7e28..accdb646a03d 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-pkt-decoder.c
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-pkt-decoder.c
@@ -64,6 +64,13 @@
 	[INTEL_PT_PIP]		= "PIP",
 	[INTEL_PT_OVF]		= "OVF",
 	[INTEL_PT_MNT]		= "MNT",
+	[INTEL_PT_PTWRITE]	= "PTWRITE",
+	[INTEL_PT_PTWRITE_IP]	= "PTWRITE",
+	[INTEL_PT_EXSTOP]	= "EXSTOP",
+	[INTEL_PT_EXSTOP_IP]	= "EXSTOP",
+	[INTEL_PT_MWAIT]	= "MWAIT",
+	[INTEL_PT_PWRE]		= "PWRE",
+	[INTEL_PT_PWRX]		= "PWRX",
 };
 
 const char *intel_pt_pkt_name(enum intel_pt_pkt_type type)
@@ -217,12 +224,80 @@ static int intel_pt_get_3byte(const unsigned char *buf, size_t len,
 	}
 }
 
+static int intel_pt_get_ptwrite(const unsigned char *buf, size_t len,
+				struct intel_pt_pkt *packet)
+{
+	packet->count = (buf[1] >> 5) & 0x3;
+	packet->type = buf[1] & BIT(7) ? INTEL_PT_PTWRITE_IP :
+					 INTEL_PT_PTWRITE;
+
+	switch (packet->count) {
+	case 0:
+		if (len < 6)
+			return INTEL_PT_NEED_MORE_BYTES;
+		packet->payload = le32_to_cpu(*(uint32_t *)(buf + 2));
+		return 6;
+	case 1:
+		if (len < 10)
+			return INTEL_PT_NEED_MORE_BYTES;
+		packet->payload = le64_to_cpu(*(uint64_t *)(buf + 2));
+		return 10;
+	default:
+		return INTEL_PT_BAD_PACKET;
+	}
+}
+
+static int intel_pt_get_exstop(struct intel_pt_pkt *packet)
+{
+	packet->type = INTEL_PT_EXSTOP;
+	return 2;
+}
+
+static int intel_pt_get_exstop_ip(struct intel_pt_pkt *packet)
+{
+	packet->type = INTEL_PT_EXSTOP_IP;
+	return 2;
+}
+
+static int intel_pt_get_mwait(const unsigned char *buf, size_t len,
+			      struct intel_pt_pkt *packet)
+{
+	if (len < 10)
+		return INTEL_PT_NEED_MORE_BYTES;
+	packet->type = INTEL_PT_MWAIT;
+	packet->payload = le64_to_cpu(*(uint64_t *)(buf + 2));
+	return 10;
+}
+
+static int intel_pt_get_pwre(const unsigned char *buf, size_t len,
+			     struct intel_pt_pkt *packet)
+{
+	if (len < 4)
+		return INTEL_PT_NEED_MORE_BYTES;
+	packet->type = INTEL_PT_PWRE;
+	memcpy_le64(&packet->payload, buf + 2, 2);
+	return 4;
+}
+
+static int intel_pt_get_pwrx(const unsigned char *buf, size_t len,
+			     struct intel_pt_pkt *packet)
+{
+	if (len < 7)
+		return INTEL_PT_NEED_MORE_BYTES;
+	packet->type = INTEL_PT_PWRX;
+	memcpy_le64(&packet->payload, buf + 2, 5);
+	return 7;
+}
+
 static int intel_pt_get_ext(const unsigned char *buf, size_t len,
 			    struct intel_pt_pkt *packet)
 {
 	if (len < 2)
 		return INTEL_PT_NEED_MORE_BYTES;
 
+	if ((buf[1] & 0x1f) == 0x12)
+		return intel_pt_get_ptwrite(buf, len, packet);
+
 	switch (buf[1]) {
 	case 0xa3: /* Long TNT */
 		return intel_pt_get_long_tnt(buf, len, packet);
@@ -244,6 +319,16 @@ static int intel_pt_get_ext(const unsigned char *buf, size_t len,
 		return intel_pt_get_tma(buf, len, packet);
 	case 0xC3: /* 3-byte header */
 		return intel_pt_get_3byte(buf, len, packet);
+	case 0x62: /* EXSTOP no IP */
+		return intel_pt_get_exstop(packet);
+	case 0xE2: /* EXSTOP with IP */
+		return intel_pt_get_exstop_ip(packet);
+	case 0xC2: /* MWAIT */
+		return intel_pt_get_mwait(buf, len, packet);
+	case 0x22: /* PWRE */
+		return intel_pt_get_pwre(buf, len, packet);
+	case 0xA2: /* PWRX */
+		return intel_pt_get_pwrx(buf, len, packet);
 	default:
 		return INTEL_PT_BAD_PACKET;
 	}
@@ -522,6 +607,29 @@ int intel_pt_pkt_desc(const struct intel_pt_pkt *packet, char *buf,
 		ret = snprintf(buf, buf_len, "%s 0x%llx (NR=%d)",
 			       name, payload, nr);
 		return ret;
+	case INTEL_PT_PTWRITE:
+		return snprintf(buf, buf_len, "%s 0x%llx IP:0", name, payload);
+	case INTEL_PT_PTWRITE_IP:
+		return snprintf(buf, buf_len, "%s 0x%llx IP:1", name, payload);
+	case INTEL_PT_EXSTOP:
+		return snprintf(buf, buf_len, "%s IP:0", name);
+	case INTEL_PT_EXSTOP_IP:
+		return snprintf(buf, buf_len, "%s IP:1", name);
+	case INTEL_PT_MWAIT:
+		return snprintf(buf, buf_len, "%s 0x%llx Hints 0x%x Extensions 0x%x",
+				name, payload, (unsigned int)(payload & 0xff),
+				(unsigned int)((payload >> 32) & 0x3));
+	case INTEL_PT_PWRE:
+		return snprintf(buf, buf_len, "%s 0x%llx HW:%u CState:%u Sub-CState:%u",
+				name, payload, !!(payload & 0x80),
+				(unsigned int)((payload >> 12) & 0xf),
+				(unsigned int)((payload >> 8) & 0xf));
+	case INTEL_PT_PWRX:
+		return snprintf(buf, buf_len, "%s 0x%llx Last CState:%u Deepest CState:%u Wake Reason 0x%x",
+				name, payload,
+				(unsigned int)((payload >> 4) & 0xf),
+				(unsigned int)(payload & 0xf),
+				(unsigned int)((payload >> 8) & 0xf));
 	default:
 		break;
 	}
diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-pkt-decoder.h b/tools/perf/util/intel-pt-decoder/intel-pt-pkt-decoder.h
index 781bb79883bd..73ddc3a88d07 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-pkt-decoder.h
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-pkt-decoder.h
@@ -52,6 +52,13 @@ enum intel_pt_pkt_type {
 	INTEL_PT_PIP,
 	INTEL_PT_OVF,
 	INTEL_PT_MNT,
+	INTEL_PT_PTWRITE,
+	INTEL_PT_PTWRITE_IP,
+	INTEL_PT_EXSTOP,
+	INTEL_PT_EXSTOP_IP,
+	INTEL_PT_MWAIT,
+	INTEL_PT_PWRE,
+	INTEL_PT_PWRX,
 };
 
 struct intel_pt_pkt {
-- 
1.9.1

[PATCH 12/37] perf intel-pt: Add reserved byte to CBR packet payload

[Adrian Hunter]

Future proof CBR packet decoding by passing through also the undefined
'reserved' byte in the packet payload.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/util/intel-pt-decoder/intel-pt-decoder.c     | 2 +-
 tools/perf/util/intel-pt-decoder/intel-pt-pkt-decoder.c | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
index e7f4016a5b31..0b6c6f442870 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
@@ -1422,7 +1422,7 @@ static void intel_pt_calc_mtc_timestamp(struct intel_pt_decoder *decoder)
 
 static void intel_pt_calc_cbr(struct intel_pt_decoder *decoder)
 {
-	unsigned int cbr = decoder->packet.payload;
+	unsigned int cbr = decoder->packet.payload & 0xff;
 
 	if (decoder->cbr == cbr)
 		return;
diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-pkt-decoder.c b/tools/perf/util/intel-pt-decoder/intel-pt-pkt-decoder.c
index accdb646a03d..ba4c9dd18643 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-pkt-decoder.c
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-pkt-decoder.c
@@ -130,7 +130,7 @@ static int intel_pt_get_cbr(const unsigned char *buf, size_t len,
 	if (len < 4)
 		return INTEL_PT_NEED_MORE_BYTES;
 	packet->type = INTEL_PT_CBR;
-	packet->payload = buf[2];
+	packet->payload = le16_to_cpu(*(uint16_t *)(buf + 2));
 	return 4;
 }
 
-- 
1.9.1

[PATCH 13/37] perf intel-pt: Move decoder error setting into one condition

[Adrian Hunter]

Move decoder error setting into one condition.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/util/intel-pt-decoder/intel-pt-decoder.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
index 0b6c6f442870..8062655bc112 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
@@ -2314,14 +2314,17 @@ const struct intel_pt_state *intel_pt_decode(struct intel_pt_decoder *decoder)
 		}
 	} while (err == -ENOLINK);
 
-	decoder->state.err = err ? intel_pt_ext_err(err) : 0;
 	decoder->state.timestamp = decoder->timestamp;
 	decoder->state.est_timestamp = intel_pt_est_timestamp(decoder);
 	decoder->state.cr3 = decoder->cr3;
 	decoder->state.tot_insn_cnt = decoder->tot_insn_cnt;
 
-	if (err)
+	if (err) {
+		decoder->state.err = intel_pt_ext_err(err);
 		decoder->state.from_ip = decoder->ip;
+	} else {
+		decoder->state.err = 0;
+	}
 
 	return &decoder->state;
 }
-- 
1.9.1

[PATCH 14/37] perf intel-pt: Add decoder support for CBR events

[Adrian Hunter]

Add decoder support for informing the tools of changes to the core-to-bus
ratio (CBR).

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/util/intel-pt-decoder/intel-pt-decoder.c | 19 +++++++++++++++++++
 tools/perf/util/intel-pt-decoder/intel-pt-decoder.h |  2 ++
 2 files changed, 21 insertions(+)

diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
index 8062655bc112..9bb694290a28 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
@@ -121,6 +121,7 @@ struct intel_pt_decoder {
 	int pkt_len;
 	int last_packet_type;
 	unsigned int cbr;
+	unsigned int cbr_seen;
 	unsigned int max_non_turbo_ratio;
 	double max_non_turbo_ratio_fp;
 	double cbr_cyc_to_tsc;
@@ -147,6 +148,7 @@ struct intel_pt_decoder {
 	uint64_t fup_ptw_payload;
 	uint64_t fup_mwait_payload;
 	uint64_t fup_pwre_payload;
+	uint64_t cbr_payload;
 	uint64_t timestamp_insn_cnt;
 	uint64_t stuck_ip;
 	int no_progress;
@@ -1424,6 +1426,8 @@ static void intel_pt_calc_cbr(struct intel_pt_decoder *decoder)
 {
 	unsigned int cbr = decoder->packet.payload & 0xff;
 
+	decoder->cbr_payload = decoder->packet.payload;
+
 	if (decoder->cbr == cbr)
 		return;
 
@@ -1784,6 +1788,16 @@ static int intel_pt_walk_trace(struct intel_pt_decoder *decoder)
 
 		case INTEL_PT_CBR:
 			intel_pt_calc_cbr(decoder);
+			if (!decoder->branch_enable &&
+			    decoder->cbr != decoder->cbr_seen) {
+				decoder->cbr_seen = decoder->cbr;
+				decoder->state.type = INTEL_PT_CBR_CHG;
+				decoder->state.from_ip = decoder->ip;
+				decoder->state.to_ip = 0;
+				decoder->state.cbr_payload =
+							decoder->packet.payload;
+				return 0;
+			}
 			break;
 
 		case INTEL_PT_MODE_EXEC:
@@ -2324,6 +2338,11 @@ const struct intel_pt_state *intel_pt_decode(struct intel_pt_decoder *decoder)
 		decoder->state.from_ip = decoder->ip;
 	} else {
 		decoder->state.err = 0;
+		if (decoder->cbr != decoder->cbr_seen && decoder->state.type) {
+			decoder->cbr_seen = decoder->cbr;
+			decoder->state.type |= INTEL_PT_CBR_CHG;
+			decoder->state.cbr_payload = decoder->cbr_payload;
+		}
 	}
 
 	return &decoder->state;
diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.h b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.h
index 414c88e9e0da..921b22e8ca0e 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.h
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.h
@@ -36,6 +36,7 @@ enum intel_pt_sample_type {
 	INTEL_PT_PWR_ENTRY	= 1 << 5,
 	INTEL_PT_EX_STOP	= 1 << 6,
 	INTEL_PT_PWR_EXIT	= 1 << 7,
+	INTEL_PT_CBR_CHG	= 1 << 8,
 };
 
 enum intel_pt_period_type {
@@ -73,6 +74,7 @@ struct intel_pt_state {
 	uint64_t mwait_payload;
 	uint64_t pwre_payload;
 	uint64_t pwrx_payload;
+	uint64_t cbr_payload;
 	uint32_t flags;
 	enum intel_pt_insn_op insn_op;
 	int insn_len;
-- 
1.9.1

[PATCH 15/37] perf intel-pt: Remove redundant initial_skip checks

[Adrian Hunter]

'initial_skip' is checked inside the sample synthesis functions which means
it is actually being done twice for 'instructions' and 'transactions'
samples. Remove the redundant checks.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/util/intel-pt.c | 8 ++------
 1 file changed, 2 insertions(+), 6 deletions(-)

diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c
index 5c59b8c6a719..3ae03f920253 100644
--- a/tools/perf/util/intel-pt.c
+++ b/tools/perf/util/intel-pt.c
@@ -1322,18 +1322,14 @@ static int intel_pt_sample(struct intel_pt_queue *ptq)
 	ptq->have_sample = false;
 
 	if (pt->sample_instructions &&
-	    (state->type & INTEL_PT_INSTRUCTION) &&
-	    (!pt->synth_opts.initial_skip ||
-	     pt->num_events++ >= pt->synth_opts.initial_skip)) {
+	    (state->type & INTEL_PT_INSTRUCTION)) {
 		err = intel_pt_synth_instruction_sample(ptq);
 		if (err)
 			return err;
 	}
 
 	if (pt->sample_transactions &&
-	    (state->type & INTEL_PT_TRANSACTION) &&
-	    (!pt->synth_opts.initial_skip ||
-	     pt->num_events++ >= pt->synth_opts.initial_skip)) {
+	    (state->type & INTEL_PT_TRANSACTION)) {
 		err = intel_pt_synth_transaction_sample(ptq);
 		if (err)
 			return err;
-- 
1.9.1

[PATCH 16/37] perf intel-pt: Fix transactions_sample_type

[Adrian Hunter]

'transactions_sample_type' is needed to correctly inject transactions
samples but it was not being set. Set it from the event sample type.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/util/intel-pt.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c
index 3ae03f920253..6df836469f2b 100644
--- a/tools/perf/util/intel-pt.c
+++ b/tools/perf/util/intel-pt.c
@@ -2035,6 +2035,7 @@ static int intel_pt_synth_events(struct intel_pt *pt,
 			return err;
 		}
 		pt->sample_transactions = true;
+		pt->transactions_sample_type = attr.sample_type;
 		pt->transactions_id = id;
 		id += 1;
 		evlist__for_each_entry(evlist, evsel) {
-- 
1.9.1

[PATCH 17/37] perf tools: Fix message because cpu list option is -C not -c

[Adrian Hunter]

Fix message because cpu list option is -C not -c

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/util/session.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/perf/util/session.c b/tools/perf/util/session.c
index 7dc1096264c5..d19c40a81040 100644
--- a/tools/perf/util/session.c
+++ b/tools/perf/util/session.c
@@ -2035,7 +2035,7 @@ int perf_session__cpu_bitmap(struct perf_session *session,
 
 		if (!(evsel->attr.sample_type & PERF_SAMPLE_CPU)) {
 			pr_err("File does not contain CPU events. "
-			       "Remove -c option to proceed.\n");
+			       "Remove -C option to proceed.\n");
 			return -1;
 		}
 	}
-- 
1.9.1

[PATCH 18/37] perf script: Fix message because field list option is -F not -f

[Adrian Hunter]

Fix message because field list option is -F not -f.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/builtin-script.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c
index d05aec491cff..79a101e0e13b 100644
--- a/tools/perf/builtin-script.c
+++ b/tools/perf/builtin-script.c
@@ -383,7 +383,7 @@ static int perf_session__check_output_opt(struct perf_session *session)
 		 */
 		if (!evsel && output[j].user_set && !output[j].wildcard_set) {
 			pr_err("%s events do not exist. "
-			       "Remove corresponding -f option to proceed.\n",
+			       "Remove corresponding -F option to proceed.\n",
 			       event_type(j));
 			return -1;
 		}
-- 
1.9.1

[PATCH 19/37] perf script: Add 'synth' event type for synthesized events

[Adrian Hunter]

Instruction trace decoders such as Intel PT may have additional information
recorded in the trace. For example, Intel PT has power information and a
there is a new instruction 'ptwrite' that can write a value into a PTWRITE
trace packet. Such information may be associated with an IP and so can be
treated as a sample (PERF_RECORD_SAMPLE). Custom data can be incorporated
in the sample as raw_data (PERF_SAMPLE_RAW). However a means of identifying
the raw data format is needed. That will be done by synthesizing an
attribute for it. So add an attribute type for custom synthesized events.
Different synthesized events will be identified by the attribute 'config'.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/builtin-script.c | 74 +++++++++++++++++++++++++++++++++++----------
 tools/perf/util/event.h     |  3 ++
 2 files changed, 61 insertions(+), 16 deletions(-)

diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c
index 79a101e0e13b..99167bafe81c 100644
--- a/tools/perf/builtin-script.c
+++ b/tools/perf/builtin-script.c
@@ -117,6 +117,11 @@ struct output_option {
 	{.str = "brstackinsn", .field = PERF_OUTPUT_BRSTACKINSN},
 };
 
+enum {
+	OUTPUT_TYPE_SYNTH = PERF_TYPE_MAX,
+	OUTPUT_TYPE_MAX
+};
+
 /* default set to maintain compatibility with current format */
 static struct {
 	bool user_set;
@@ -124,7 +129,7 @@ struct output_option {
 	unsigned int print_ip_opts;
 	u64 fields;
 	u64 invalid_fields;
-} output[PERF_TYPE_MAX] = {
+} output[OUTPUT_TYPE_MAX] = {
 
 	[PERF_TYPE_HARDWARE] = {
 		.user_set = false,
@@ -182,12 +187,43 @@ struct output_option {
 
 		.invalid_fields = PERF_OUTPUT_TRACE | PERF_OUTPUT_BPF_OUTPUT,
 	},
+
+	[OUTPUT_TYPE_SYNTH] = {
+		.user_set = false,
+
+		.fields = PERF_OUTPUT_COMM | PERF_OUTPUT_TID |
+			      PERF_OUTPUT_CPU | PERF_OUTPUT_TIME |
+			      PERF_OUTPUT_EVNAME | PERF_OUTPUT_IP |
+			      PERF_OUTPUT_SYM | PERF_OUTPUT_DSO,
+
+		.invalid_fields = PERF_OUTPUT_TRACE | PERF_OUTPUT_BPF_OUTPUT,
+	},
 };
 
+static inline int output_type(unsigned int type)
+{
+	switch (type) {
+	case PERF_TYPE_SYNTH:
+		return OUTPUT_TYPE_SYNTH;
+	default:
+		return type;
+	}
+}
+
+static inline unsigned int attr_type(unsigned int type)
+{
+	switch (type) {
+	case OUTPUT_TYPE_SYNTH:
+		return PERF_TYPE_SYNTH;
+	default:
+		return type;
+	}
+}
+
 static bool output_set_by_user(void)
 {
 	int j;
-	for (j = 0; j < PERF_TYPE_MAX; ++j) {
+	for (j = 0; j < OUTPUT_TYPE_MAX; ++j) {
 		if (output[j].user_set)
 			return true;
 	}
@@ -208,7 +244,7 @@ static const char *output_field2str(enum perf_output_field field)
 	return str;
 }
 
-#define PRINT_FIELD(x)  (output[attr->type].fields & PERF_OUTPUT_##x)
+#define PRINT_FIELD(x)  (output[output_type(attr->type)].fields & PERF_OUTPUT_##x)
 
 static int perf_evsel__do_check_stype(struct perf_evsel *evsel,
 				      u64 sample_type, const char *sample_msg,
@@ -216,7 +252,7 @@ static int perf_evsel__do_check_stype(struct perf_evsel *evsel,
 				      bool allow_user_set)
 {
 	struct perf_event_attr *attr = &evsel->attr;
-	int type = attr->type;
+	int type = output_type(attr->type);
 	const char *evname;
 
 	if (attr->sample_type & sample_type)
@@ -346,7 +382,7 @@ static int perf_evsel__check_attr(struct perf_evsel *evsel,
 
 static void set_print_ip_opts(struct perf_event_attr *attr)
 {
-	unsigned int type = attr->type;
+	unsigned int type = output_type(attr->type);
 
 	output[type].print_ip_opts = 0;
 	if (PRINT_FIELD(IP))
@@ -374,14 +410,15 @@ static int perf_session__check_output_opt(struct perf_session *session)
 	unsigned int j;
 	struct perf_evsel *evsel;
 
-	for (j = 0; j < PERF_TYPE_MAX; ++j) {
-		evsel = perf_session__find_first_evtype(session, j);
+	for (j = 0; j < OUTPUT_TYPE_MAX; ++j) {
+		evsel = perf_session__find_first_evtype(session, attr_type(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) {
+		if (!evsel && output[j].user_set && !output[j].wildcard_set &&
+		    j != OUTPUT_TYPE_SYNTH) {
 			pr_err("%s events do not exist. "
 			       "Remove corresponding -F option to proceed.\n",
 			       event_type(j));
@@ -906,6 +943,7 @@ static void print_sample_bts(struct perf_sample *sample,
 			     struct machine *machine)
 {
 	struct perf_event_attr *attr = &evsel->attr;
+	unsigned int type = output_type(attr->type);
 	bool print_srcline_last = false;
 
 	if (PRINT_FIELD(CALLINDENT))
@@ -913,7 +951,7 @@ static void print_sample_bts(struct perf_sample *sample,
 
 	/* print branch_from information */
 	if (PRINT_FIELD(IP)) {
-		unsigned int print_opts = output[attr->type].print_ip_opts;
+		unsigned int print_opts = output[type].print_ip_opts;
 		struct callchain_cursor *cursor = NULL;
 
 		if (symbol_conf.use_callchain && sample->callchain &&
@@ -936,7 +974,7 @@ static void print_sample_bts(struct perf_sample *sample,
 	/* print branch_to information */
 	if (PRINT_FIELD(ADDR) ||
 	    ((evsel->attr.sample_type & PERF_SAMPLE_ADDR) &&
-	     !output[attr->type].user_set)) {
+	     !output[type].user_set)) {
 		printf(" => ");
 		print_sample_addr(sample, thread, attr);
 	}
@@ -1132,8 +1170,9 @@ static void process_event(struct perf_script *script,
 {
 	struct thread *thread = al->thread;
 	struct perf_event_attr *attr = &evsel->attr;
+	unsigned int type = output_type(attr->type);
 
-	if (output[attr->type].fields == 0)
+	if (output[type].fields == 0)
 		return;
 
 	print_sample_start(sample, thread, evsel);
@@ -1180,7 +1219,7 @@ static void process_event(struct perf_script *script,
 			cursor = &callchain_cursor;
 
 		putchar(cursor ? '\n' : ' ');
-		sample__fprintf_sym(sample, al, 0, output[attr->type].print_ip_opts, cursor, stdout);
+		sample__fprintf_sym(sample, al, 0, output[type].print_ip_opts, cursor, stdout);
 	}
 
 	if (PRINT_FIELD(IREGS))
@@ -1325,7 +1364,8 @@ static int process_attr(struct perf_tool *tool, union perf_event *event,
 	evlist = *pevlist;
 	evsel = perf_evlist__last(*pevlist);
 
-	if (evsel->attr.type >= PERF_TYPE_MAX)
+	if (evsel->attr.type >= PERF_TYPE_MAX &&
+	    evsel->attr.type != PERF_TYPE_SYNTH)
 		return 0;
 
 	evlist__for_each_entry(evlist, pos) {
@@ -1749,6 +1789,8 @@ static int parse_output_fields(const struct option *opt __maybe_unused,
 			type = PERF_TYPE_RAW;
 		else if (!strcmp(str, "break"))
 			type = PERF_TYPE_BREAKPOINT;
+		else if (!strcmp(str, "synth"))
+			type = OUTPUT_TYPE_SYNTH;
 		else {
 			fprintf(stderr, "Invalid event type in field string.\n");
 			rc = -EINVAL;
@@ -1775,7 +1817,7 @@ static int parse_output_fields(const struct option *opt __maybe_unused,
 		if (output_set_by_user())
 			pr_warning("Overriding previous field request for all events.\n");
 
-		for (j = 0; j < PERF_TYPE_MAX; ++j) {
+		for (j = 0; j < OUTPUT_TYPE_MAX; ++j) {
 			output[j].fields = 0;
 			output[j].user_set = true;
 			output[j].wildcard_set = true;
@@ -1801,7 +1843,7 @@ static int parse_output_fields(const struct option *opt __maybe_unused,
 			/* add user option to all events types for
 			 * which it is valid
 			 */
-			for (j = 0; j < PERF_TYPE_MAX; ++j) {
+			for (j = 0; j < OUTPUT_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));
@@ -2444,7 +2486,7 @@ int cmd_script(int argc, const char **argv)
 		     symbol__config_symfs),
 	OPT_CALLBACK('F', "fields", NULL, "str",
 		     "comma separated output fields prepend with 'type:'. "
-		     "Valid types: hw,sw,trace,raw. "
+		     "Valid types: hw,sw,trace,raw,synth. "
 		     "Fields: comm,tid,pid,time,cpu,event,trace,ip,sym,dso,"
 		     "addr,symoff,period,iregs,brstack,brstacksym,flags,"
 		     "bpf-output,callindent,insn,insnlen,brstackinsn",
diff --git a/tools/perf/util/event.h b/tools/perf/util/event.h
index 27ac047490c3..af285ed951f4 100644
--- a/tools/perf/util/event.h
+++ b/tools/perf/util/event.h
@@ -252,6 +252,9 @@ enum auxtrace_error_type {
 	PERF_AUXTRACE_ERROR_MAX
 };
 
+/* Attribute type for custom synthesized events */
+#define PERF_TYPE_SYNTH		3000000000
+
 /*
  * The kernel collects the number of events it couldn't send in a stretch and
  * when possible sends this number in a PERF_RECORD_LOST event. The number of
-- 
1.9.1

[PATCH 20/37] perf script: Add 'synth' field for synthesized event payloads

[Adrian Hunter]

Add a field to display the content the raw_data of a synthesized event.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/Documentation/perf-script.txt |  6 +++++-
 tools/perf/builtin-script.c              | 20 ++++++++++++++++++--
 2 files changed, 23 insertions(+), 3 deletions(-)

diff --git a/tools/perf/Documentation/perf-script.txt b/tools/perf/Documentation/perf-script.txt
index cb0eda3925e6..43770f174f64 100644
--- a/tools/perf/Documentation/perf-script.txt
+++ b/tools/perf/Documentation/perf-script.txt
@@ -117,7 +117,8 @@ OPTIONS
         Comma separated list of fields to print. Options are:
         comm, tid, pid, time, cpu, event, trace, ip, sym, dso, addr, symoff,
         srcline, period, iregs, brstack, brstacksym, flags, bpf-output, brstackinsn,
-        callindent, insn, insnlen. Field list can be prepended with the type, trace, sw or hw,
+        callindent, insn, insnlen, synth.
+        Field 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,ip,sym  and -F trace:time,cpu,trace
 
@@ -185,6 +186,9 @@ OPTIONS
 	instruction bytes and the instruction length of the current
 	instruction.
 
+	The synth field is used by synthesized events which may be created when
+	Instruction Trace decoding.
+
 	Finally, a user may not set fields to none for all event types.
 	i.e., -F "" is not allowed.
 
diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c
index 99167bafe81c..f7a5130d2fd0 100644
--- a/tools/perf/builtin-script.c
+++ b/tools/perf/builtin-script.c
@@ -85,6 +85,7 @@ enum perf_output_field {
 	PERF_OUTPUT_INSN	    = 1U << 21,
 	PERF_OUTPUT_INSNLEN	    = 1U << 22,
 	PERF_OUTPUT_BRSTACKINSN	    = 1U << 23,
+	PERF_OUTPUT_SYNTH	    = 1U << 24,
 };
 
 struct output_option {
@@ -115,6 +116,7 @@ struct output_option {
 	{.str = "insn", .field = PERF_OUTPUT_INSN},
 	{.str = "insnlen", .field = PERF_OUTPUT_INSNLEN},
 	{.str = "brstackinsn", .field = PERF_OUTPUT_BRSTACKINSN},
+	{.str = "synth", .field = PERF_OUTPUT_SYNTH},
 };
 
 enum {
@@ -194,7 +196,8 @@ enum {
 		.fields = PERF_OUTPUT_COMM | PERF_OUTPUT_TID |
 			      PERF_OUTPUT_CPU | PERF_OUTPUT_TIME |
 			      PERF_OUTPUT_EVNAME | PERF_OUTPUT_IP |
-			      PERF_OUTPUT_SYM | PERF_OUTPUT_DSO,
+			      PERF_OUTPUT_SYM | PERF_OUTPUT_DSO |
+			      PERF_OUTPUT_SYNTH,
 
 		.invalid_fields = PERF_OUTPUT_TRACE | PERF_OUTPUT_BPF_OUTPUT,
 	},
@@ -1117,6 +1120,15 @@ static void print_sample_bpf_output(struct perf_sample *sample)
 		       (char *)(sample->raw_data));
 }
 
+static void print_sample_synth(struct perf_sample *sample __maybe_unused,
+			       struct perf_evsel *evsel)
+{
+	switch (evsel->attr.config) {
+	default:
+		break;
+	}
+}
+
 struct perf_script {
 	struct perf_tool	tool;
 	struct perf_session	*session;
@@ -1201,6 +1213,10 @@ static void process_event(struct perf_script *script,
 	if (PRINT_FIELD(TRACE))
 		event_format__print(evsel->tp_format, sample->cpu,
 				    sample->raw_data, sample->raw_size);
+
+	if (attr->type == PERF_TYPE_SYNTH && PRINT_FIELD(SYNTH))
+		print_sample_synth(sample, evsel);
+
 	if (PRINT_FIELD(ADDR))
 		print_sample_addr(sample, thread, attr);
 
@@ -2489,7 +2505,7 @@ int cmd_script(int argc, const char **argv)
 		     "Valid types: hw,sw,trace,raw,synth. "
 		     "Fields: comm,tid,pid,time,cpu,event,trace,ip,sym,dso,"
 		     "addr,symoff,period,iregs,brstack,brstacksym,flags,"
-		     "bpf-output,callindent,insn,insnlen,brstackinsn",
+		     "bpf-output,callindent,insn,insnlen,brstackinsn,synth",
 		     parse_output_fields),
 	OPT_BOOLEAN('a', "all-cpus", &system_wide,
 		    "system-wide collection from all CPUs"),
-- 
1.9.1

[PATCH 21/37] tools include: Add byte-swapping macros to kernel.h

[Adrian Hunter]

Add byte-swapping macros to kernel.h

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/include/linux/kernel.h | 35 +++++++++++++++++++++++++++++------
 1 file changed, 29 insertions(+), 6 deletions(-)

diff --git a/tools/include/linux/kernel.h b/tools/include/linux/kernel.h
index 73ccc48126bb..039bb85e4171 100644
--- a/tools/include/linux/kernel.h
+++ b/tools/include/linux/kernel.h
@@ -5,6 +5,8 @@
 #include <stddef.h>
 #include <assert.h>
 #include <linux/compiler.h>
+#include <endian.h>
+#include <byteswap.h>
 
 #ifndef UINT_MAX
 #define UINT_MAX	(~0U)
@@ -67,12 +69,33 @@
 #endif
 #endif
 
-/*
- * Both need more care to handle endianness
- * (Don't use bitmap_copy_le() for now)
- */
-#define cpu_to_le64(x)	(x)
-#define cpu_to_le32(x)	(x)
+#if __BYTE_ORDER == __BIG_ENDIAN
+#define cpu_to_le16 bswap_16
+#define cpu_to_le32 bswap_32
+#define cpu_to_le64 bswap_64
+#define le16_to_cpu bswap_16
+#define le32_to_cpu bswap_32
+#define le64_to_cpu bswap_64
+#define cpu_to_be16
+#define cpu_to_be32
+#define cpu_to_be64
+#define be16_to_cpu
+#define be32_to_cpu
+#define be64_to_cpu
+#else
+#define cpu_to_le16
+#define cpu_to_le32
+#define cpu_to_le64
+#define le16_to_cpu
+#define le32_to_cpu
+#define le64_to_cpu
+#define cpu_to_be16 bswap_16
+#define cpu_to_be32 bswap_32
+#define cpu_to_be64 bswap_64
+#define be16_to_cpu bswap_16
+#define be32_to_cpu bswap_32
+#define be64_to_cpu bswap_64
+#endif
 
 int vscnprintf(char *buf, size_t size, const char *fmt, va_list args);
 int scnprintf(char * buf, size_t size, const char * fmt, ...);
-- 
1.9.1

[PATCH 22/37] perf auxtrace: Add itrace option to output ptwrite events

[Adrian Hunter]

Add itrace option to output ptwrite events.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/Documentation/itrace.txt | 7 ++++---
 tools/perf/util/auxtrace.c          | 4 ++++
 tools/perf/util/auxtrace.h          | 2 ++
 3 files changed, 10 insertions(+), 3 deletions(-)

diff --git a/tools/perf/Documentation/itrace.txt b/tools/perf/Documentation/itrace.txt
index e2a4c5e0dbe5..deafd16692b6 100644
--- a/tools/perf/Documentation/itrace.txt
+++ b/tools/perf/Documentation/itrace.txt
@@ -3,13 +3,14 @@
 		c	synthesize branches events (calls only)
 		r	synthesize branches events (returns only)
 		x	synthesize transactions events
+		w	synthesize ptwrite events
 		e	synthesize error events
 		d	create a debug log
 		g	synthesize a call chain (use with i or x)
 		l	synthesize last branch entries (use with i or x)
 		s       skip initial number of events
 
-	The default is all events i.e. the same as --itrace=ibxe
+	The default is all events i.e. the same as --itrace=ibxwe
 
 	In addition, the period (default 100000) for instructions events
 	can be specified in units of:
@@ -26,8 +27,8 @@
 	Also the number of last branch entries (default 64, max. 1024) for
 	instructions or transactions events can be specified.
 
-	It is also possible to skip events generated (instructions, branches, transactions)
-	at the beginning. This is useful to ignore initialization code.
+	It is also possible to skip events generated (instructions, branches, transactions,
+	ptwrite) at the beginning. This is useful to ignore initialization code.
 
 	--itrace=i0nss1000000
 
diff --git a/tools/perf/util/auxtrace.c b/tools/perf/util/auxtrace.c
index 0daf63b9ee3e..baad91ed1e05 100644
--- a/tools/perf/util/auxtrace.c
+++ b/tools/perf/util/auxtrace.c
@@ -947,6 +947,7 @@ void itrace_synth_opts__set_default(struct itrace_synth_opts *synth_opts)
 	synth_opts->instructions = true;
 	synth_opts->branches = true;
 	synth_opts->transactions = true;
+	synth_opts->ptwrites = true;
 	synth_opts->errors = true;
 	synth_opts->period_type = PERF_ITRACE_DEFAULT_PERIOD_TYPE;
 	synth_opts->period = PERF_ITRACE_DEFAULT_PERIOD;
@@ -1030,6 +1031,9 @@ int itrace_parse_synth_opts(const struct option *opt, const char *str,
 		case 'x':
 			synth_opts->transactions = true;
 			break;
+		case 'w':
+			synth_opts->ptwrites = true;
+			break;
 		case 'e':
 			synth_opts->errors = true;
 			break;
diff --git a/tools/perf/util/auxtrace.h b/tools/perf/util/auxtrace.h
index 9f0de72d58e2..b48afb2f18f3 100644
--- a/tools/perf/util/auxtrace.h
+++ b/tools/perf/util/auxtrace.h
@@ -59,6 +59,7 @@ enum itrace_period_type {
  * @instructions: whether to synthesize 'instructions' events
  * @branches: whether to synthesize 'branches' events
  * @transactions: whether to synthesize events for transactions
+ * @ptwrites: whether to synthesize events for ptwrites
  * @errors: whether to synthesize decoder error events
  * @dont_decode: whether to skip decoding entirely
  * @log: write a decoding log
@@ -79,6 +80,7 @@ struct itrace_synth_opts {
 	bool			instructions;
 	bool			branches;
 	bool			transactions;
+	bool			ptwrites;
 	bool			errors;
 	bool			dont_decode;
 	bool			log;
-- 
1.9.1

[PATCH 23/37] perf auxtrace: Add itrace option to output power events

[Adrian Hunter]

Add itrace option to output power events.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/Documentation/itrace.txt | 5 +++--
 tools/perf/util/auxtrace.c          | 4 ++++
 tools/perf/util/auxtrace.h          | 2 ++
 3 files changed, 9 insertions(+), 2 deletions(-)

diff --git a/tools/perf/Documentation/itrace.txt b/tools/perf/Documentation/itrace.txt
index deafd16692b6..a3abe04c779d 100644
--- a/tools/perf/Documentation/itrace.txt
+++ b/tools/perf/Documentation/itrace.txt
@@ -4,13 +4,14 @@
 		r	synthesize branches events (returns only)
 		x	synthesize transactions events
 		w	synthesize ptwrite events
+		p	synthesize power events
 		e	synthesize error events
 		d	create a debug log
 		g	synthesize a call chain (use with i or x)
 		l	synthesize last branch entries (use with i or x)
 		s       skip initial number of events
 
-	The default is all events i.e. the same as --itrace=ibxwe
+	The default is all events i.e. the same as --itrace=ibxwpe
 
 	In addition, the period (default 100000) for instructions events
 	can be specified in units of:
@@ -28,7 +29,7 @@
 	instructions or transactions events can be specified.
 
 	It is also possible to skip events generated (instructions, branches, transactions,
-	ptwrite) at the beginning. This is useful to ignore initialization code.
+	ptwrite, power) at the beginning. This is useful to ignore initialization code.
 
 	--itrace=i0nss1000000
 
diff --git a/tools/perf/util/auxtrace.c b/tools/perf/util/auxtrace.c
index baad91ed1e05..651c01dfa5d3 100644
--- a/tools/perf/util/auxtrace.c
+++ b/tools/perf/util/auxtrace.c
@@ -948,6 +948,7 @@ void itrace_synth_opts__set_default(struct itrace_synth_opts *synth_opts)
 	synth_opts->branches = true;
 	synth_opts->transactions = true;
 	synth_opts->ptwrites = true;
+	synth_opts->pwr_events = true;
 	synth_opts->errors = true;
 	synth_opts->period_type = PERF_ITRACE_DEFAULT_PERIOD_TYPE;
 	synth_opts->period = PERF_ITRACE_DEFAULT_PERIOD;
@@ -1034,6 +1035,9 @@ int itrace_parse_synth_opts(const struct option *opt, const char *str,
 		case 'w':
 			synth_opts->ptwrites = true;
 			break;
+		case 'p':
+			synth_opts->pwr_events = true;
+			break;
 		case 'e':
 			synth_opts->errors = true;
 			break;
diff --git a/tools/perf/util/auxtrace.h b/tools/perf/util/auxtrace.h
index b48afb2f18f3..68e0aa40b24a 100644
--- a/tools/perf/util/auxtrace.h
+++ b/tools/perf/util/auxtrace.h
@@ -60,6 +60,7 @@ enum itrace_period_type {
  * @branches: whether to synthesize 'branches' events
  * @transactions: whether to synthesize events for transactions
  * @ptwrites: whether to synthesize events for ptwrites
+ * @pwr_events: whether to synthesize power events
  * @errors: whether to synthesize decoder error events
  * @dont_decode: whether to skip decoding entirely
  * @log: write a decoding log
@@ -81,6 +82,7 @@ struct itrace_synth_opts {
 	bool			branches;
 	bool			transactions;
 	bool			ptwrites;
+	bool			pwr_events;
 	bool			errors;
 	bool			dont_decode;
 	bool			log;
-- 
1.9.1

[PATCH 24/37] perf script: Add synthesized Intel PT power and ptwrite events

[Adrian Hunter]

Add definitions for synthesized Intel PT events for power and ptwrite.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/builtin-script.c | 114 +++++++++++++++++++++++++++++++++++++++++++-
 tools/perf/util/event.h     |  90 ++++++++++++++++++++++++++++++++++
 2 files changed, 203 insertions(+), 1 deletion(-)

diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c
index f7a5130d2fd0..4a13daacba25 100644
--- a/tools/perf/builtin-script.c
+++ b/tools/perf/builtin-script.c
@@ -1120,10 +1120,122 @@ static void print_sample_bpf_output(struct perf_sample *sample)
 		       (char *)(sample->raw_data));
 }
 
-static void print_sample_synth(struct perf_sample *sample __maybe_unused,
+static void print_sample_spacing(int len, int spacing)
+{
+	if (len > 0 && len < spacing)
+		printf("%*s", spacing - len, "");
+}
+
+static void print_sample_pt_spacing(int len)
+{
+	print_sample_spacing(len, 34);
+}
+
+static void print_sample_synth_ptwrite(struct perf_sample *sample)
+{
+	struct perf_synth_intel_ptwrite *data = sample->raw_data;
+	int len;
+
+	if (sample->raw_size < sizeof(*data))
+		return;
+
+	len = printf(" IP: %u payload: %#" PRIx64 " ",
+		     data->ip, le64_to_cpu(data->payload));
+	print_sample_pt_spacing(len);
+}
+
+static void print_sample_synth_mwait(struct perf_sample *sample)
+{
+	struct perf_synth_intel_mwait *data = sample->raw_data;
+	int len;
+
+	if (sample->raw_size < sizeof(*data))
+		return;
+
+	len = printf(" hints: %#x extensions: %#x ",
+		     data->hints, data->extensions);
+	print_sample_pt_spacing(len);
+}
+
+static void print_sample_synth_pwre(struct perf_sample *sample)
+{
+	struct perf_synth_intel_pwre *data = sample->raw_data;
+	int len;
+
+	if (sample->raw_size < sizeof(*data))
+		return;
+
+	len = printf(" hw: %u cstate: %u sub-cstate: %u ",
+		     data->hw, data->cstate, data->subcstate);
+	print_sample_pt_spacing(len);
+}
+
+static void print_sample_synth_exstop(struct perf_sample *sample)
+{
+	struct perf_synth_intel_exstop *data = sample->raw_data;
+	int len;
+
+	if (sample->raw_size < sizeof(*data))
+		return;
+
+	len = printf(" IP: %u ", data->ip);
+	print_sample_pt_spacing(len);
+}
+
+static void print_sample_synth_pwrx(struct perf_sample *sample)
+{
+	struct perf_synth_intel_pwrx *data = sample->raw_data;
+	int len;
+
+	if (sample->raw_size < sizeof(*data))
+		return;
+
+	len = printf(" deepest cstate: %u last cstate: %u wake reason: %#x ",
+		     data->deepest_cstate, data->last_cstate,
+		     data->wake_reason);
+	print_sample_pt_spacing(len);
+}
+
+static void print_sample_synth_cbr(struct perf_sample *sample)
+{
+	struct perf_synth_intel_cbr *data = sample->raw_data;
+	unsigned int percent, freq;
+	int len;
+
+	if (sample->raw_size < sizeof(*data))
+		return;
+
+	freq = (le32_to_cpu(data->freq) + 500) / 1000;
+	len = printf(" cbr: %2u freq: %4u MHz ", data->cbr, freq);
+	if (data->max_nonturbo) {
+		percent = (5 + (1000 * data->cbr) / data->max_nonturbo) / 10;
+		len += printf("(%3u%%) ", percent);
+	}
+	print_sample_pt_spacing(len);
+}
+
+static void print_sample_synth(struct perf_sample *sample,
 			       struct perf_evsel *evsel)
 {
 	switch (evsel->attr.config) {
+	case PERF_SYNTH_INTEL_PTWRITE:
+		print_sample_synth_ptwrite(sample);
+		break;
+	case PERF_SYNTH_INTEL_MWAIT:
+		print_sample_synth_mwait(sample);
+		break;
+	case PERF_SYNTH_INTEL_PWRE:
+		print_sample_synth_pwre(sample);
+		break;
+	case PERF_SYNTH_INTEL_EXSTOP:
+		print_sample_synth_exstop(sample);
+		break;
+	case PERF_SYNTH_INTEL_PWRX:
+		print_sample_synth_pwrx(sample);
+		break;
+	case PERF_SYNTH_INTEL_CBR:
+		print_sample_synth_cbr(sample);
+		break;
 	default:
 		break;
 	}
diff --git a/tools/perf/util/event.h b/tools/perf/util/event.h
index af285ed951f4..7823d43315b7 100644
--- a/tools/perf/util/event.h
+++ b/tools/perf/util/event.h
@@ -255,6 +255,96 @@ enum auxtrace_error_type {
 /* Attribute type for custom synthesized events */
 #define PERF_TYPE_SYNTH		3000000000
 
+/* Attribute config for custom synthesized events */
+enum perf_synth_id {
+	PERF_SYNTH_INTEL_PTWRITE,
+	PERF_SYNTH_INTEL_MWAIT,
+	PERF_SYNTH_INTEL_PWRE,
+	PERF_SYNTH_INTEL_EXSTOP,
+	PERF_SYNTH_INTEL_PWRX,
+	PERF_SYNTH_INTEL_CBR,
+};
+
+/*
+ * Raw data formats for synthesized events. Note that raw data plus the raw data
+ * size (4 bytes) must align to 8-bytes.
+ */
+
+struct perf_synth_intel_ptwrite {
+	union {
+		struct {
+			u32	ip		:  1,
+				reserved	: 31;
+		};
+		u32	flags;
+	};
+	u64	payload;
+} __packed;
+
+struct perf_synth_intel_mwait {
+	u32 reserved;
+	union {
+		struct {
+			u64	hints		:  8,
+				reserved1	: 24,
+				extensions	:  2,
+				reserved2	: 30;
+		};
+		u64	payload;
+	};
+} __packed;
+
+struct perf_synth_intel_pwre {
+	u32 reserved;
+	union {
+		struct {
+			u64	reserved1	:  7,
+				hw		:  1,
+				subcstate	:  4,
+				cstate		:  4,
+				reserved2	: 48;
+		};
+		u64	payload;
+	};
+} __packed;
+
+struct perf_synth_intel_exstop {
+	union {
+		struct {
+			u32	ip		:  1,
+				reserved	: 31;
+		};
+		u32	flags;
+	};
+};
+
+struct perf_synth_intel_pwrx {
+	u32 reserved;
+	union {
+		struct {
+			u64	deepest_cstate	:  4,
+				last_cstate	:  4,
+				wake_reason	:  4,
+				reserved1	: 52;
+		};
+		u64	payload;
+	};
+} __packed;
+
+struct perf_synth_intel_cbr {
+	union {
+		struct {
+			u32	cbr		:  8,
+				reserved1	:  8,
+				max_nonturbo	:  8,
+				reserved2	:  8;
+		};
+		u32	flags;
+	};
+	u32 freq;
+	u32 reserved3;
+};
+
 /*
  * The kernel collects the number of events it couldn't send in a stretch and
  * when possible sends this number in a PERF_RECORD_LOST event. The number of
-- 
1.9.1

[PATCH 25/37] perf intel-pt: Factor out common code synthesizing event samples

[Adrian Hunter]

Factor out common code in functions synthesizing event samples i.e.
intel_pt_synth_branch_sample(), intel_pt_synth_instruction_sample() and
intel_pt_synth_transaction_sample().

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/util/intel-pt.c | 222 ++++++++++++++++++++-------------------------
 1 file changed, 100 insertions(+), 122 deletions(-)

diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c
index 6df836469f2b..dbff5dca09f0 100644
--- a/tools/perf/util/intel-pt.c
+++ b/tools/perf/util/intel-pt.c
@@ -1058,6 +1058,36 @@ static void intel_pt_update_last_branch_rb(struct intel_pt_queue *ptq)
 		bs->nr += 1;
 }
 
+static inline bool intel_pt_skip_event(struct intel_pt *pt)
+{
+	return pt->synth_opts.initial_skip &&
+	       pt->num_events++ < pt->synth_opts.initial_skip;
+}
+
+static void intel_pt_prep_b_sample(struct intel_pt *pt,
+				   struct intel_pt_queue *ptq,
+				   union perf_event *event,
+				   struct perf_sample *sample)
+{
+	event->sample.header.type = PERF_RECORD_SAMPLE;
+	event->sample.header.misc = PERF_RECORD_MISC_USER;
+	event->sample.header.size = sizeof(struct perf_event_header);
+
+	if (!pt->timeless_decoding)
+		sample->time = tsc_to_perf_time(ptq->timestamp, &pt->tc);
+
+	sample->cpumode = PERF_RECORD_MISC_USER;
+	sample->ip = ptq->state->from_ip;
+	sample->pid = ptq->pid;
+	sample->tid = ptq->tid;
+	sample->addr = ptq->state->to_ip;
+	sample->period = 1;
+	sample->cpu = ptq->cpu;
+	sample->flags = ptq->flags;
+	sample->insn_len = ptq->insn_len;
+	memcpy(sample->insn, ptq->insn, INTEL_PT_INSN_BUF_SZ);
+}
+
 static int intel_pt_inject_event(union perf_event *event,
 				 struct perf_sample *sample, u64 type,
 				 bool swapped)
@@ -1066,9 +1096,35 @@ static int intel_pt_inject_event(union perf_event *event,
 	return perf_event__synthesize_sample(event, type, 0, sample, swapped);
 }
 
-static int intel_pt_synth_branch_sample(struct intel_pt_queue *ptq)
+static inline int intel_pt_opt_inject(struct intel_pt *pt,
+				      union perf_event *event,
+				      struct perf_sample *sample, u64 type)
+{
+	if (!pt->synth_opts.inject)
+		return 0;
+
+	return intel_pt_inject_event(event, sample, type, pt->synth_needs_swap);
+}
+
+static int intel_pt_deliver_synth_b_event(struct intel_pt *pt,
+					  union perf_event *event,
+					  struct perf_sample *sample, u64 type)
 {
 	int ret;
+
+	ret = intel_pt_opt_inject(pt, event, sample, type);
+	if (ret)
+		return ret;
+
+	ret = perf_session__deliver_synth_event(pt->session, event, sample);
+	if (ret)
+		pr_err("Intel PT: failed to deliver event, error %d\n", ret);
+
+	return ret;
+}
+
+static int intel_pt_synth_branch_sample(struct intel_pt_queue *ptq)
+{
 	struct intel_pt *pt = ptq->pt;
 	union perf_event *event = ptq->event_buf;
 	struct perf_sample sample = { .ip = 0, };
@@ -1080,29 +1136,13 @@ static int intel_pt_synth_branch_sample(struct intel_pt_queue *ptq)
 	if (pt->branches_filter && !(pt->branches_filter & ptq->flags))
 		return 0;
 
-	if (pt->synth_opts.initial_skip &&
-	    pt->num_events++ < pt->synth_opts.initial_skip)
+	if (intel_pt_skip_event(pt))
 		return 0;
 
-	event->sample.header.type = PERF_RECORD_SAMPLE;
-	event->sample.header.misc = PERF_RECORD_MISC_USER;
-	event->sample.header.size = sizeof(struct perf_event_header);
+	intel_pt_prep_b_sample(pt, ptq, event, &sample);
 
-	if (!pt->timeless_decoding)
-		sample.time = tsc_to_perf_time(ptq->timestamp, &pt->tc);
-
-	sample.cpumode = PERF_RECORD_MISC_USER;
-	sample.ip = ptq->state->from_ip;
-	sample.pid = ptq->pid;
-	sample.tid = ptq->tid;
-	sample.addr = ptq->state->to_ip;
 	sample.id = ptq->pt->branches_id;
 	sample.stream_id = ptq->pt->branches_id;
-	sample.period = 1;
-	sample.cpu = ptq->cpu;
-	sample.flags = ptq->flags;
-	sample.insn_len = ptq->insn_len;
-	memcpy(sample.insn, ptq->insn, INTEL_PT_INSN_BUF_SZ);
 
 	/*
 	 * perf report cannot handle events without a branch stack when using
@@ -1119,78 +1159,38 @@ static int intel_pt_synth_branch_sample(struct intel_pt_queue *ptq)
 		sample.branch_stack = (struct branch_stack *)&dummy_bs;
 	}
 
-	if (pt->synth_opts.inject) {
-		ret = intel_pt_inject_event(event, &sample,
-					    pt->branches_sample_type,
-					    pt->synth_needs_swap);
-		if (ret)
-			return ret;
-	}
-
-	ret = perf_session__deliver_synth_event(pt->session, event, &sample);
-	if (ret)
-		pr_err("Intel Processor Trace: failed to deliver branch event, error %d\n",
-		       ret);
-
-	return ret;
+	return intel_pt_deliver_synth_b_event(pt, event, &sample,
+					      pt->branches_sample_type);
 }
 
-static int intel_pt_synth_instruction_sample(struct intel_pt_queue *ptq)
+static void intel_pt_prep_sample(struct intel_pt *pt,
+				 struct intel_pt_queue *ptq,
+				 union perf_event *event,
+				 struct perf_sample *sample)
 {
-	int ret;
-	struct intel_pt *pt = ptq->pt;
-	union perf_event *event = ptq->event_buf;
-	struct perf_sample sample = { .ip = 0, };
-
-	if (pt->synth_opts.initial_skip &&
-	    pt->num_events++ < pt->synth_opts.initial_skip)
-		return 0;
-
-	event->sample.header.type = PERF_RECORD_SAMPLE;
-	event->sample.header.misc = PERF_RECORD_MISC_USER;
-	event->sample.header.size = sizeof(struct perf_event_header);
-
-	if (!pt->timeless_decoding)
-		sample.time = tsc_to_perf_time(ptq->timestamp, &pt->tc);
-
-	sample.cpumode = PERF_RECORD_MISC_USER;
-	sample.ip = ptq->state->from_ip;
-	sample.pid = ptq->pid;
-	sample.tid = ptq->tid;
-	sample.addr = ptq->state->to_ip;
-	sample.id = ptq->pt->instructions_id;
-	sample.stream_id = ptq->pt->instructions_id;
-	sample.period = ptq->state->tot_insn_cnt - ptq->last_insn_cnt;
-	sample.cpu = ptq->cpu;
-	sample.flags = ptq->flags;
-	sample.insn_len = ptq->insn_len;
-	memcpy(sample.insn, ptq->insn, INTEL_PT_INSN_BUF_SZ);
-
-	ptq->last_insn_cnt = ptq->state->tot_insn_cnt;
+	intel_pt_prep_b_sample(pt, ptq, event, sample);
 
 	if (pt->synth_opts.callchain) {
 		thread_stack__sample(ptq->thread, ptq->chain,
-				     pt->synth_opts.callchain_sz, sample.ip);
-		sample.callchain = ptq->chain;
+				     pt->synth_opts.callchain_sz, sample->ip);
+		sample->callchain = ptq->chain;
 	}
 
 	if (pt->synth_opts.last_branch) {
 		intel_pt_copy_last_branch_rb(ptq);
-		sample.branch_stack = ptq->last_branch;
+		sample->branch_stack = ptq->last_branch;
 	}
+}
 
-	if (pt->synth_opts.inject) {
-		ret = intel_pt_inject_event(event, &sample,
-					    pt->instructions_sample_type,
-					    pt->synth_needs_swap);
-		if (ret)
-			return ret;
-	}
+static inline int intel_pt_deliver_synth_event(struct intel_pt *pt,
+					       struct intel_pt_queue *ptq,
+					       union perf_event *event,
+					       struct perf_sample *sample,
+					       u64 type)
+{
+	int ret;
 
-	ret = perf_session__deliver_synth_event(pt->session, event, &sample);
-	if (ret)
-		pr_err("Intel Processor Trace: failed to deliver instruction event, error %d\n",
-		       ret);
+	ret = intel_pt_deliver_synth_b_event(pt, event, sample, type);
 
 	if (pt->synth_opts.last_branch)
 		intel_pt_reset_last_branch_rb(ptq);
@@ -1198,65 +1198,43 @@ static int intel_pt_synth_instruction_sample(struct intel_pt_queue *ptq)
 	return ret;
 }
 
-static int intel_pt_synth_transaction_sample(struct intel_pt_queue *ptq)
+static int intel_pt_synth_instruction_sample(struct intel_pt_queue *ptq)
 {
-	int ret;
 	struct intel_pt *pt = ptq->pt;
 	union perf_event *event = ptq->event_buf;
 	struct perf_sample sample = { .ip = 0, };
 
-	if (pt->synth_opts.initial_skip &&
-	    pt->num_events++ < pt->synth_opts.initial_skip)
+	if (intel_pt_skip_event(pt))
 		return 0;
 
-	event->sample.header.type = PERF_RECORD_SAMPLE;
-	event->sample.header.misc = PERF_RECORD_MISC_USER;
-	event->sample.header.size = sizeof(struct perf_event_header);
+	intel_pt_prep_sample(pt, ptq, event, &sample);
 
-	if (!pt->timeless_decoding)
-		sample.time = tsc_to_perf_time(ptq->timestamp, &pt->tc);
+	sample.id = ptq->pt->instructions_id;
+	sample.stream_id = ptq->pt->instructions_id;
+	sample.period = ptq->state->tot_insn_cnt - ptq->last_insn_cnt;
 
-	sample.cpumode = PERF_RECORD_MISC_USER;
-	sample.ip = ptq->state->from_ip;
-	sample.pid = ptq->pid;
-	sample.tid = ptq->tid;
-	sample.addr = ptq->state->to_ip;
-	sample.id = ptq->pt->transactions_id;
-	sample.stream_id = ptq->pt->transactions_id;
-	sample.period = 1;
-	sample.cpu = ptq->cpu;
-	sample.flags = ptq->flags;
-	sample.insn_len = ptq->insn_len;
-	memcpy(sample.insn, ptq->insn, INTEL_PT_INSN_BUF_SZ);
+	ptq->last_insn_cnt = ptq->state->tot_insn_cnt;
 
-	if (pt->synth_opts.callchain) {
-		thread_stack__sample(ptq->thread, ptq->chain,
-				     pt->synth_opts.callchain_sz, sample.ip);
-		sample.callchain = ptq->chain;
-	}
+	return intel_pt_deliver_synth_event(pt, ptq, event, &sample,
+					    pt->instructions_sample_type);
+}
 
-	if (pt->synth_opts.last_branch) {
-		intel_pt_copy_last_branch_rb(ptq);
-		sample.branch_stack = ptq->last_branch;
-	}
+static int intel_pt_synth_transaction_sample(struct intel_pt_queue *ptq)
+{
+	struct intel_pt *pt = ptq->pt;
+	union perf_event *event = ptq->event_buf;
+	struct perf_sample sample = { .ip = 0, };
 
-	if (pt->synth_opts.inject) {
-		ret = intel_pt_inject_event(event, &sample,
-					    pt->transactions_sample_type,
-					    pt->synth_needs_swap);
-		if (ret)
-			return ret;
-	}
+	if (intel_pt_skip_event(pt))
+		return 0;
 
-	ret = perf_session__deliver_synth_event(pt->session, event, &sample);
-	if (ret)
-		pr_err("Intel Processor Trace: failed to deliver transaction event, error %d\n",
-		       ret);
+	intel_pt_prep_sample(pt, ptq, event, &sample);
 
-	if (pt->synth_opts.last_branch)
-		intel_pt_reset_last_branch_rb(ptq);
+	sample.id = ptq->pt->transactions_id;
+	sample.stream_id = ptq->pt->transactions_id;
 
-	return ret;
+	return intel_pt_deliver_synth_event(pt, ptq, event, &sample,
+					    pt->transactions_sample_type);
 }
 
 static int intel_pt_synth_error(struct intel_pt *pt, int code, int cpu,
-- 
1.9.1

[PATCH 26/37] perf intel-pt: Remove unused instructions_sample_period

[Adrian Hunter]

Remove unused struct intel_pt member instructions_sample_period.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/util/intel-pt.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c
index dbff5dca09f0..f8237a0e2946 100644
--- a/tools/perf/util/intel-pt.c
+++ b/tools/perf/util/intel-pt.c
@@ -81,7 +81,6 @@ struct intel_pt {
 
 	bool sample_instructions;
 	u64 instructions_sample_type;
-	u64 instructions_sample_period;
 	u64 instructions_id;
 
 	bool sample_branches;
@@ -1978,7 +1977,6 @@ static int intel_pt_synth_events(struct intel_pt *pt,
 				intel_pt_ns_to_ticks(pt, pt->synth_opts.period);
 		else
 			attr.sample_period = pt->synth_opts.period;
-		pt->instructions_sample_period = attr.sample_period;
 		if (pt->synth_opts.callchain)
 			attr.sample_type |= PERF_SAMPLE_CALLCHAIN;
 		if (pt->synth_opts.last_branch)
-- 
1.9.1

[PATCH 27/37] perf intel-pt: Join needlessly wrapped lines

[Adrian Hunter]

Join needlessly wrapped lines.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/util/intel-pt.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)

diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c
index f8237a0e2946..b670502b0264 100644
--- a/tools/perf/util/intel-pt.c
+++ b/tools/perf/util/intel-pt.c
@@ -1298,15 +1298,13 @@ static int intel_pt_sample(struct intel_pt_queue *ptq)
 
 	ptq->have_sample = false;
 
-	if (pt->sample_instructions &&
-	    (state->type & INTEL_PT_INSTRUCTION)) {
+	if (pt->sample_instructions && (state->type & INTEL_PT_INSTRUCTION)) {
 		err = intel_pt_synth_instruction_sample(ptq);
 		if (err)
 			return err;
 	}
 
-	if (pt->sample_transactions &&
-	    (state->type & INTEL_PT_TRANSACTION)) {
+	if (pt->sample_transactions && (state->type & INTEL_PT_TRANSACTION)) {
 		err = intel_pt_synth_transaction_sample(ptq);
 		if (err)
 			return err;
-- 
1.9.1

[PATCH 28/37] perf intel-pt: Tidy Intel PT evsel lookup into separate function

[Adrian Hunter]

Tidy the lookup of the Intel PT selected event (perf_evsel) into a separate
function.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/util/intel-pt.c | 25 +++++++++++++++----------
 1 file changed, 15 insertions(+), 10 deletions(-)

diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c
index b670502b0264..a9486b57584f 100644
--- a/tools/perf/util/intel-pt.c
+++ b/tools/perf/util/intel-pt.c
@@ -1922,24 +1922,29 @@ static int intel_pt_synth_event(struct perf_session *session,
 					   &id, intel_pt_event_synth);
 }
 
+static struct perf_evsel *intel_pt_evsel(struct intel_pt *pt,
+					 struct perf_evlist *evlist)
+{
+	struct perf_evsel *evsel;
+
+	evlist__for_each_entry(evlist, evsel) {
+		if (evsel->attr.type == pt->pmu_type && evsel->ids)
+			return evsel;
+	}
+
+	return NULL;
+}
+
 static int intel_pt_synth_events(struct intel_pt *pt,
 				 struct perf_session *session)
 {
 	struct perf_evlist *evlist = session->evlist;
-	struct perf_evsel *evsel;
+	struct perf_evsel *evsel = intel_pt_evsel(pt, evlist);
 	struct perf_event_attr attr;
-	bool found = false;
 	u64 id;
 	int err;
 
-	evlist__for_each_entry(evlist, evsel) {
-		if (evsel->attr.type == pt->pmu_type && evsel->ids) {
-			found = true;
-			break;
-		}
-	}
-
-	if (!found) {
+	if (!evsel) {
 		pr_debug("There are no selected events with Intel Processor Trace data\n");
 		return 0;
 	}
-- 
1.9.1

[PATCH 29/37] perf intel-pt: Tidy messages into called function intel_pt_synth_event()

[Adrian Hunter]

Tidy print messages into called function intel_pt_synth_event().

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/util/intel-pt.c | 42 ++++++++++++++++++------------------------
 1 file changed, 18 insertions(+), 24 deletions(-)

diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c
index a9486b57584f..81907f60e7da 100644
--- a/tools/perf/util/intel-pt.c
+++ b/tools/perf/util/intel-pt.c
@@ -1910,16 +1910,25 @@ static int intel_pt_event_synth(struct perf_tool *tool,
 						 NULL);
 }
 
-static int intel_pt_synth_event(struct perf_session *session,
+static int intel_pt_synth_event(struct perf_session *session, const char *name,
 				struct perf_event_attr *attr, u64 id)
 {
 	struct intel_pt_synth intel_pt_synth;
+	int err;
+
+	pr_debug("Synthesizing '%s' event with id %" PRIu64 " sample type %#" PRIx64 "\n",
+		 name, id, (u64)attr->sample_type);
 
 	memset(&intel_pt_synth, 0, sizeof(struct intel_pt_synth));
 	intel_pt_synth.session = session;
 
-	return perf_event__synthesize_attr(&intel_pt_synth.dummy_tool, attr, 1,
-					   &id, intel_pt_event_synth);
+	err = perf_event__synthesize_attr(&intel_pt_synth.dummy_tool, attr, 1,
+					  &id, intel_pt_event_synth);
+	if (err)
+		pr_err("%s: failed to synthesize '%s' event type\n",
+		       __func__, name);
+
+	return err;
 }
 
 static struct perf_evsel *intel_pt_evsel(struct intel_pt *pt,
@@ -1984,14 +1993,9 @@ static int intel_pt_synth_events(struct intel_pt *pt,
 			attr.sample_type |= PERF_SAMPLE_CALLCHAIN;
 		if (pt->synth_opts.last_branch)
 			attr.sample_type |= PERF_SAMPLE_BRANCH_STACK;
-		pr_debug("Synthesizing 'instructions' event with id %" PRIu64 " sample type %#" PRIx64 "\n",
-			 id, (u64)attr.sample_type);
-		err = intel_pt_synth_event(session, &attr, id);
-		if (err) {
-			pr_err("%s: failed to synthesize 'instructions' event type\n",
-			       __func__);
+		err = intel_pt_synth_event(session, "instructions", &attr, id);
+		if (err)
 			return err;
-		}
 		pt->sample_instructions = true;
 		pt->instructions_sample_type = attr.sample_type;
 		pt->instructions_id = id;
@@ -2005,14 +2009,9 @@ static int intel_pt_synth_events(struct intel_pt *pt,
 			attr.sample_type |= PERF_SAMPLE_CALLCHAIN;
 		if (pt->synth_opts.last_branch)
 			attr.sample_type |= PERF_SAMPLE_BRANCH_STACK;
-		pr_debug("Synthesizing 'transactions' event with id %" PRIu64 " sample type %#" PRIx64 "\n",
-			 id, (u64)attr.sample_type);
-		err = intel_pt_synth_event(session, &attr, id);
-		if (err) {
-			pr_err("%s: failed to synthesize 'transactions' event type\n",
-			       __func__);
+		err = intel_pt_synth_event(session, "transactions", &attr, id);
+		if (err)
 			return err;
-		}
 		pt->sample_transactions = true;
 		pt->transactions_sample_type = attr.sample_type;
 		pt->transactions_id = id;
@@ -2033,14 +2032,9 @@ static int intel_pt_synth_events(struct intel_pt *pt,
 		attr.sample_type |= PERF_SAMPLE_ADDR;
 		attr.sample_type &= ~(u64)PERF_SAMPLE_CALLCHAIN;
 		attr.sample_type &= ~(u64)PERF_SAMPLE_BRANCH_STACK;
-		pr_debug("Synthesizing 'branches' event with id %" PRIu64 " sample type %#" PRIx64 "\n",
-			 id, (u64)attr.sample_type);
-		err = intel_pt_synth_event(session, &attr, id);
-		if (err) {
-			pr_err("%s: failed to synthesize 'branches' event type\n",
-			       __func__);
+		err = intel_pt_synth_event(session, "branches", &attr, id);
+		if (err)
 			return err;
-		}
 		pt->sample_branches = true;
 		pt->branches_sample_type = attr.sample_type;
 		pt->branches_id = id;
-- 
1.9.1

[PATCH 30/37] perf intel-pt: Factor out intel_pt_set_event_name()

[Adrian Hunter]

Factor out intel_pt_set_event_name() so it can be reused.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/util/intel-pt.c | 24 ++++++++++++++++--------
 1 file changed, 16 insertions(+), 8 deletions(-)

diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c
index 81907f60e7da..a20712e1ed28 100644
--- a/tools/perf/util/intel-pt.c
+++ b/tools/perf/util/intel-pt.c
@@ -1931,6 +1931,21 @@ static int intel_pt_synth_event(struct perf_session *session, const char *name,
 	return err;
 }
 
+static void intel_pt_set_event_name(struct perf_evlist *evlist, u64 id,
+				    const char *name)
+{
+	struct perf_evsel *evsel;
+
+	evlist__for_each_entry(evlist, evsel) {
+		if (evsel->id && evsel->id[0] == id) {
+			if (evsel->name)
+				zfree(&evsel->name);
+			evsel->name = strdup(name);
+			break;
+		}
+	}
+}
+
 static struct perf_evsel *intel_pt_evsel(struct intel_pt *pt,
 					 struct perf_evlist *evlist)
 {
@@ -2015,15 +2030,8 @@ static int intel_pt_synth_events(struct intel_pt *pt,
 		pt->sample_transactions = true;
 		pt->transactions_sample_type = attr.sample_type;
 		pt->transactions_id = id;
+		intel_pt_set_event_name(evlist, id, "transactions");
 		id += 1;
-		evlist__for_each_entry(evlist, evsel) {
-			if (evsel->id && evsel->id[0] == pt->transactions_id) {
-				if (evsel->name)
-					zfree(&evsel->name);
-				evsel->name = strdup("transactions");
-				break;
-			}
-		}
 	}
 
 	if (pt->synth_opts.branches) {
-- 
1.9.1

[PATCH 31/37] perf intel-pt: Move code in intel_pt_synth_events() to simplify attr setting

[Adrian Hunter]

intel_pt_synth_events() uses the same attr structure to create each event.
Move the code around a bit to simplify that.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/util/intel-pt.c | 45 ++++++++++++++++++++++-----------------------
 1 file changed, 22 insertions(+), 23 deletions(-)

diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c
index a20712e1ed28..ace79a405f98 100644
--- a/tools/perf/util/intel-pt.c
+++ b/tools/perf/util/intel-pt.c
@@ -1997,6 +1997,25 @@ static int intel_pt_synth_events(struct intel_pt *pt,
 	if (!id)
 		id = 1;
 
+	if (pt->synth_opts.branches) {
+		attr.config = PERF_COUNT_HW_BRANCH_INSTRUCTIONS;
+		attr.sample_period = 1;
+		attr.sample_type |= PERF_SAMPLE_ADDR;
+		err = intel_pt_synth_event(session, "branches", &attr, id);
+		if (err)
+			return err;
+		pt->sample_branches = true;
+		pt->branches_sample_type = attr.sample_type;
+		pt->branches_id = id;
+		id += 1;
+		attr.sample_type &= ~(u64)PERF_SAMPLE_ADDR;
+	}
+
+	if (pt->synth_opts.callchain)
+		attr.sample_type |= PERF_SAMPLE_CALLCHAIN;
+	if (pt->synth_opts.last_branch)
+		attr.sample_type |= PERF_SAMPLE_BRANCH_STACK;
+
 	if (pt->synth_opts.instructions) {
 		attr.config = PERF_COUNT_HW_INSTRUCTIONS;
 		if (pt->synth_opts.period_type == PERF_ITRACE_PERIOD_NANOSECS)
@@ -2004,10 +2023,6 @@ static int intel_pt_synth_events(struct intel_pt *pt,
 				intel_pt_ns_to_ticks(pt, pt->synth_opts.period);
 		else
 			attr.sample_period = pt->synth_opts.period;
-		if (pt->synth_opts.callchain)
-			attr.sample_type |= PERF_SAMPLE_CALLCHAIN;
-		if (pt->synth_opts.last_branch)
-			attr.sample_type |= PERF_SAMPLE_BRANCH_STACK;
 		err = intel_pt_synth_event(session, "instructions", &attr, id);
 		if (err)
 			return err;
@@ -2017,13 +2032,11 @@ static int intel_pt_synth_events(struct intel_pt *pt,
 		id += 1;
 	}
 
+	attr.sample_type &= ~(u64)PERF_SAMPLE_PERIOD;
+	attr.sample_period = 1;
+
 	if (pt->synth_opts.transactions) {
 		attr.config = PERF_COUNT_HW_INSTRUCTIONS;
-		attr.sample_period = 1;
-		if (pt->synth_opts.callchain)
-			attr.sample_type |= PERF_SAMPLE_CALLCHAIN;
-		if (pt->synth_opts.last_branch)
-			attr.sample_type |= PERF_SAMPLE_BRANCH_STACK;
 		err = intel_pt_synth_event(session, "transactions", &attr, id);
 		if (err)
 			return err;
@@ -2034,20 +2047,6 @@ static int intel_pt_synth_events(struct intel_pt *pt,
 		id += 1;
 	}
 
-	if (pt->synth_opts.branches) {
-		attr.config = PERF_COUNT_HW_BRANCH_INSTRUCTIONS;
-		attr.sample_period = 1;
-		attr.sample_type |= PERF_SAMPLE_ADDR;
-		attr.sample_type &= ~(u64)PERF_SAMPLE_CALLCHAIN;
-		attr.sample_type &= ~(u64)PERF_SAMPLE_BRANCH_STACK;
-		err = intel_pt_synth_event(session, "branches", &attr, id);
-		if (err)
-			return err;
-		pt->sample_branches = true;
-		pt->branches_sample_type = attr.sample_type;
-		pt->branches_id = id;
-	}
-
 	pt->synth_needs_swap = evsel->needs_swap;
 
 	return 0;
-- 
1.9.1

[PATCH 32/37] perf intel-pt: Synthesize new power and ptwrite events

[Adrian Hunter]

Synthesize new power and ptwrite events.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/util/intel-pt.c | 283 +++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 283 insertions(+)

diff --git a/tools/perf/util/intel-pt.c b/tools/perf/util/intel-pt.c
index ace79a405f98..754e92ee6c3e 100644
--- a/tools/perf/util/intel-pt.c
+++ b/tools/perf/util/intel-pt.c
@@ -92,6 +92,18 @@ struct intel_pt {
 	u64 transactions_sample_type;
 	u64 transactions_id;
 
+	bool sample_ptwrites;
+	u64 ptwrites_sample_type;
+	u64 ptwrites_id;
+
+	bool sample_pwr_events;
+	u64 pwr_events_sample_type;
+	u64 mwait_id;
+	u64 pwre_id;
+	u64 exstop_id;
+	u64 pwrx_id;
+	u64 cbr_id;
+
 	bool synth_needs_swap;
 
 	u64 tsc_bit;
@@ -102,6 +114,7 @@ struct intel_pt {
 	u64 cyc_bit;
 	u64 noretcomp_bit;
 	unsigned max_non_turbo_ratio;
+	unsigned cbr2khz;
 
 	unsigned long num_events;
 
@@ -1236,6 +1249,175 @@ static int intel_pt_synth_transaction_sample(struct intel_pt_queue *ptq)
 					    pt->transactions_sample_type);
 }
 
+static void intel_pt_prep_p_sample(struct intel_pt *pt,
+				   struct intel_pt_queue *ptq,
+				   union perf_event *event,
+				   struct perf_sample *sample)
+{
+	intel_pt_prep_sample(pt, ptq, event, sample);
+
+	/*
+	 * Zero IP is used to mean "trace start" but that is not the case for
+	 * power or PTWRITE events with no IP, so clear the flags.
+	 */
+	if (!sample->ip)
+		sample->flags = 0;
+}
+
+static int intel_pt_synth_ptwrite_sample(struct intel_pt_queue *ptq)
+{
+	struct intel_pt *pt = ptq->pt;
+	union perf_event *event = ptq->event_buf;
+	struct perf_sample sample = { .ip = 0, };
+	struct perf_synth_intel_ptwrite raw;
+
+	if (intel_pt_skip_event(pt))
+		return 0;
+
+	intel_pt_prep_p_sample(pt, ptq, event, &sample);
+
+	sample.id = ptq->pt->ptwrites_id;
+	sample.stream_id = ptq->pt->ptwrites_id;
+
+	raw.flags = 0;
+	raw.ip = !!(ptq->state->flags & INTEL_PT_FUP_IP);
+	raw.payload = cpu_to_le64(ptq->state->ptw_payload);
+
+	sample.raw_size = sizeof(raw);
+	sample.raw_data = &raw;
+
+	return intel_pt_deliver_synth_event(pt, ptq, event, &sample,
+					    pt->ptwrites_sample_type);
+}
+
+static int intel_pt_synth_cbr_sample(struct intel_pt_queue *ptq)
+{
+	struct intel_pt *pt = ptq->pt;
+	union perf_event *event = ptq->event_buf;
+	struct perf_sample sample = { .ip = 0, };
+	struct perf_synth_intel_cbr raw;
+	u32 flags;
+
+	if (intel_pt_skip_event(pt))
+		return 0;
+
+	intel_pt_prep_p_sample(pt, ptq, event, &sample);
+
+	sample.id = ptq->pt->cbr_id;
+	sample.stream_id = ptq->pt->cbr_id;
+
+	flags = (u16)ptq->state->cbr_payload | (pt->max_non_turbo_ratio << 16);
+	raw.flags = cpu_to_le32(flags);
+	raw.freq = cpu_to_le32(raw.cbr * pt->cbr2khz);
+	raw.reserved3 = 0;
+
+	sample.raw_size = sizeof(raw);
+	sample.raw_data = &raw;
+
+	return intel_pt_deliver_synth_event(pt, ptq, event, &sample,
+					    pt->pwr_events_sample_type);
+}
+
+static int intel_pt_synth_mwait_sample(struct intel_pt_queue *ptq)
+{
+	struct intel_pt *pt = ptq->pt;
+	union perf_event *event = ptq->event_buf;
+	struct perf_sample sample = { .ip = 0, };
+	struct perf_synth_intel_mwait raw;
+
+	if (intel_pt_skip_event(pt))
+		return 0;
+
+	intel_pt_prep_p_sample(pt, ptq, event, &sample);
+
+	sample.id = ptq->pt->mwait_id;
+	sample.stream_id = ptq->pt->mwait_id;
+
+	raw.reserved = 0;
+	raw.payload = cpu_to_le64(ptq->state->mwait_payload);
+
+	sample.raw_size = sizeof(raw);
+	sample.raw_data = &raw;
+
+	return intel_pt_deliver_synth_event(pt, ptq, event, &sample,
+					    pt->pwr_events_sample_type);
+}
+
+static int intel_pt_synth_pwre_sample(struct intel_pt_queue *ptq)
+{
+	struct intel_pt *pt = ptq->pt;
+	union perf_event *event = ptq->event_buf;
+	struct perf_sample sample = { .ip = 0, };
+	struct perf_synth_intel_pwre raw;
+
+	if (intel_pt_skip_event(pt))
+		return 0;
+
+	intel_pt_prep_p_sample(pt, ptq, event, &sample);
+
+	sample.id = ptq->pt->pwre_id;
+	sample.stream_id = ptq->pt->pwre_id;
+
+	raw.reserved = 0;
+	raw.payload = cpu_to_le64(ptq->state->pwre_payload);
+
+	sample.raw_size = sizeof(raw);
+	sample.raw_data = &raw;
+
+	return intel_pt_deliver_synth_event(pt, ptq, event, &sample,
+					    pt->pwr_events_sample_type);
+}
+
+static int intel_pt_synth_exstop_sample(struct intel_pt_queue *ptq)
+{
+	struct intel_pt *pt = ptq->pt;
+	union perf_event *event = ptq->event_buf;
+	struct perf_sample sample = { .ip = 0, };
+	struct perf_synth_intel_exstop raw;
+
+	if (intel_pt_skip_event(pt))
+		return 0;
+
+	intel_pt_prep_p_sample(pt, ptq, event, &sample);
+
+	sample.id = ptq->pt->exstop_id;
+	sample.stream_id = ptq->pt->exstop_id;
+
+	raw.flags = 0;
+	raw.ip = !!(ptq->state->flags & INTEL_PT_FUP_IP);
+
+	sample.raw_size = sizeof(raw);
+	sample.raw_data = &raw;
+
+	return intel_pt_deliver_synth_event(pt, ptq, event, &sample,
+					    pt->pwr_events_sample_type);
+}
+
+static int intel_pt_synth_pwrx_sample(struct intel_pt_queue *ptq)
+{
+	struct intel_pt *pt = ptq->pt;
+	union perf_event *event = ptq->event_buf;
+	struct perf_sample sample = { .ip = 0, };
+	struct perf_synth_intel_pwrx raw;
+
+	if (intel_pt_skip_event(pt))
+		return 0;
+
+	intel_pt_prep_p_sample(pt, ptq, event, &sample);
+
+	sample.id = ptq->pt->pwrx_id;
+	sample.stream_id = ptq->pt->pwrx_id;
+
+	raw.reserved = 0;
+	raw.payload = cpu_to_le64(ptq->state->pwrx_payload);
+
+	sample.raw_size = sizeof(raw);
+	sample.raw_data = &raw;
+
+	return intel_pt_deliver_synth_event(pt, ptq, event, &sample,
+					    pt->pwr_events_sample_type);
+}
+
 static int intel_pt_synth_error(struct intel_pt *pt, int code, int cpu,
 				pid_t pid, pid_t tid, u64 ip)
 {
@@ -1287,6 +1469,10 @@ static inline bool intel_pt_is_switch_ip(struct intel_pt_queue *ptq, u64 ip)
 			       PERF_IP_FLAG_INTERRUPT | PERF_IP_FLAG_TX_ABORT));
 }
 
+#define INTEL_PT_PWR_EVT (INTEL_PT_MWAIT_OP | INTEL_PT_PWR_ENTRY | \
+			  INTEL_PT_EX_STOP | INTEL_PT_PWR_EXIT | \
+			  INTEL_PT_CBR_CHG)
+
 static int intel_pt_sample(struct intel_pt_queue *ptq)
 {
 	const struct intel_pt_state *state = ptq->state;
@@ -1298,6 +1484,34 @@ static int intel_pt_sample(struct intel_pt_queue *ptq)
 
 	ptq->have_sample = false;
 
+	if (pt->sample_pwr_events && (state->type & INTEL_PT_PWR_EVT)) {
+		if (state->type & INTEL_PT_CBR_CHG) {
+			err = intel_pt_synth_cbr_sample(ptq);
+			if (err)
+				return err;
+		}
+		if (state->type & INTEL_PT_MWAIT_OP) {
+			err = intel_pt_synth_mwait_sample(ptq);
+			if (err)
+				return err;
+		}
+		if (state->type & INTEL_PT_PWR_ENTRY) {
+			err = intel_pt_synth_pwre_sample(ptq);
+			if (err)
+				return err;
+		}
+		if (state->type & INTEL_PT_EX_STOP) {
+			err = intel_pt_synth_exstop_sample(ptq);
+			if (err)
+				return err;
+		}
+		if (state->type & INTEL_PT_PWR_EXIT) {
+			err = intel_pt_synth_pwrx_sample(ptq);
+			if (err)
+				return err;
+		}
+	}
+
 	if (pt->sample_instructions && (state->type & INTEL_PT_INSTRUCTION)) {
 		err = intel_pt_synth_instruction_sample(ptq);
 		if (err)
@@ -1310,6 +1524,12 @@ static int intel_pt_sample(struct intel_pt_queue *ptq)
 			return err;
 	}
 
+	if (pt->sample_ptwrites && (state->type & INTEL_PT_PTW)) {
+		err = intel_pt_synth_ptwrite_sample(ptq);
+		if (err)
+			return err;
+	}
+
 	if (!(state->type & INTEL_PT_BRANCH))
 		return 0;
 
@@ -2047,6 +2267,68 @@ static int intel_pt_synth_events(struct intel_pt *pt,
 		id += 1;
 	}
 
+	attr.type = PERF_TYPE_SYNTH;
+	attr.sample_type |= PERF_SAMPLE_RAW;
+
+	if (pt->synth_opts.ptwrites) {
+		attr.config = PERF_SYNTH_INTEL_PTWRITE;
+		err = intel_pt_synth_event(session, "ptwrite", &attr, id);
+		if (err)
+			return err;
+		pt->sample_ptwrites = true;
+		pt->ptwrites_sample_type = attr.sample_type;
+		pt->ptwrites_id = id;
+		intel_pt_set_event_name(evlist, id, "ptwrite");
+		id += 1;
+	}
+
+	if (pt->synth_opts.pwr_events) {
+		pt->sample_pwr_events = true;
+		pt->pwr_events_sample_type = attr.sample_type;
+
+		attr.config = PERF_SYNTH_INTEL_CBR;
+		err = intel_pt_synth_event(session, "cbr", &attr, id);
+		if (err)
+			return err;
+		pt->cbr_id = id;
+		intel_pt_set_event_name(evlist, id, "cbr");
+		id += 1;
+	}
+
+	if (pt->synth_opts.pwr_events && (evsel->attr.config & 0x10)) {
+		attr.config = PERF_SYNTH_INTEL_MWAIT;
+		err = intel_pt_synth_event(session, "mwait", &attr, id);
+		if (err)
+			return err;
+		pt->mwait_id = id;
+		intel_pt_set_event_name(evlist, id, "mwait");
+		id += 1;
+
+		attr.config = PERF_SYNTH_INTEL_PWRE;
+		err = intel_pt_synth_event(session, "pwre", &attr, id);
+		if (err)
+			return err;
+		pt->pwre_id = id;
+		intel_pt_set_event_name(evlist, id, "pwre");
+		id += 1;
+
+		attr.config = PERF_SYNTH_INTEL_EXSTOP;
+		err = intel_pt_synth_event(session, "exstop", &attr, id);
+		if (err)
+			return err;
+		pt->exstop_id = id;
+		intel_pt_set_event_name(evlist, id, "exstop");
+		id += 1;
+
+		attr.config = PERF_SYNTH_INTEL_PWRX;
+		err = intel_pt_synth_event(session, "pwrx", &attr, id);
+		if (err)
+			return err;
+		pt->pwrx_id = id;
+		intel_pt_set_event_name(evlist, id, "pwrx");
+		id += 1;
+	}
+
 	pt->synth_needs_swap = evsel->needs_swap;
 
 	return 0;
@@ -2313,6 +2595,7 @@ int intel_pt_process_auxtrace_info(union perf_event *event,
 		intel_pt_log("TSC frequency %"PRIu64"\n", tsc_freq);
 		intel_pt_log("Maximum non-turbo ratio %u\n",
 			     pt->max_non_turbo_ratio);
+		pt->cbr2khz = tsc_freq / pt->max_non_turbo_ratio / 1000;
 	}
 
 	if (pt->synth_opts.calls)
-- 
1.9.1

[PATCH 33/37] perf intel-pt: Add example script for power events and PTWRITE

[Adrian Hunter]

Add script intel-pt-events.py that provides an example of how to unpack the
raw data for power events and PTWRITE.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 .../perf/scripts/python/bin/intel-pt-events-record |  13 +++
 .../perf/scripts/python/bin/intel-pt-events-report |   3 +
 tools/perf/scripts/python/intel-pt-events.py       | 128 +++++++++++++++++++++
 3 files changed, 144 insertions(+)
 create mode 100644 tools/perf/scripts/python/bin/intel-pt-events-record
 create mode 100644 tools/perf/scripts/python/bin/intel-pt-events-report
 create mode 100644 tools/perf/scripts/python/intel-pt-events.py

diff --git a/tools/perf/scripts/python/bin/intel-pt-events-record b/tools/perf/scripts/python/bin/intel-pt-events-record
new file mode 100644
index 000000000000..10fe2b6977d4
--- /dev/null
+++ b/tools/perf/scripts/python/bin/intel-pt-events-record
@@ -0,0 +1,13 @@
+#!/bin/bash
+
+#
+# print Intel PT Power Events and PTWRITE. The intel_pt PMU event needs
+# to be specified with appropriate config terms.
+#
+if ! echo "$@" | grep -q intel_pt ; then
+	echo "Options must include the Intel PT event e.g. -e intel_pt/pwr_evt,ptw/"
+	echo "and for power events it probably needs to be system wide i.e. -a option"
+	echo "For example: -a -e intel_pt/pwr_evt,branch=0/ sleep 1"
+	exit 1
+fi
+perf record $@
diff --git a/tools/perf/scripts/python/bin/intel-pt-events-report b/tools/perf/scripts/python/bin/intel-pt-events-report
new file mode 100644
index 000000000000..9a9c92fcd026
--- /dev/null
+++ b/tools/perf/scripts/python/bin/intel-pt-events-report
@@ -0,0 +1,3 @@
+#!/bin/bash
+# description: print Intel PT Power Events and PTWRITE
+perf script $@ -s "$PERF_EXEC_PATH"/scripts/python/intel-pt-events.py
\ No newline at end of file
diff --git a/tools/perf/scripts/python/intel-pt-events.py b/tools/perf/scripts/python/intel-pt-events.py
new file mode 100644
index 000000000000..b19172d673af
--- /dev/null
+++ b/tools/perf/scripts/python/intel-pt-events.py
@@ -0,0 +1,128 @@
+# intel-pt-events.py: Print Intel PT Power Events and PTWRITE
+# Copyright (c) 2017, Intel Corporation.
+#
+# This program is free software; you can redistribute it and/or modify it
+# under the terms and conditions of the GNU General Public License,
+# version 2, as published by the Free Software Foundation.
+#
+# This program is distributed in the hope it will be useful, but WITHOUT
+# ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+# FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+# more details.
+
+import os
+import sys
+import struct
+
+sys.path.append(os.environ['PERF_EXEC_PATH'] + \
+	'/scripts/python/Perf-Trace-Util/lib/Perf/Trace')
+
+# These perf imports are not used at present
+#from perf_trace_context import *
+#from Core import *
+
+def trace_begin():
+	print "Intel PT Power Events and PTWRITE"
+
+def trace_end():
+	print "End"
+
+def trace_unhandled(event_name, context, event_fields_dict):
+		print ' '.join(['%s=%s'%(k,str(v))for k,v in sorted(event_fields_dict.items())])
+
+def print_ptwrite(raw_buf):
+	data = struct.unpack_from("<IQ", raw_buf)
+	flags = data[0]
+	payload = data[1]
+	exact_ip = flags & 1
+	print "IP: %u payload: %#x" % (exact_ip, payload),
+
+def print_cbr(raw_buf):
+	data = struct.unpack_from("<BBBBII", raw_buf)
+	cbr = data[0]
+	f = (data[4] + 500) / 1000
+	p = ((cbr * 1000 / data[2]) + 5) / 10
+	print "%3u  freq: %4u MHz  (%3u%%)" % (cbr, f, p),
+
+def print_mwait(raw_buf):
+	data = struct.unpack_from("<IQ", raw_buf)
+	payload = data[1]
+	hints = payload & 0xff
+	extensions = (payload >> 32) & 0x3
+	print "hints: %#x extensions: %#x" % (hints, extensions),
+
+def print_pwre(raw_buf):
+	data = struct.unpack_from("<IQ", raw_buf)
+	payload = data[1]
+	hw = (payload >> 7) & 1
+	cstate = (payload >> 12) & 0xf
+	subcstate = (payload >> 8) & 0xf
+	print "hw: %u cstate: %u sub-cstate: %u" % (hw, cstate, subcstate),
+
+def print_exstop(raw_buf):
+	data = struct.unpack_from("<I", raw_buf)
+	flags = data[0]
+	exact_ip = flags & 1
+	print "IP: %u" % (exact_ip),
+
+def print_pwrx(raw_buf):
+	data = struct.unpack_from("<IQ", raw_buf)
+	payload = data[1]
+	deepest_cstate = payload & 0xf
+	last_cstate = (payload >> 4) & 0xf
+	wake_reason = (payload >> 8) & 0xf
+	print "deepest cstate: %u last cstate: %u wake reason: %#x" % (deepest_cstate, last_cstate, wake_reason),
+
+def print_common_start(comm, sample, name):
+	ts = sample["time"]
+	cpu = sample["cpu"]
+	pid = sample["pid"]
+	tid = sample["tid"]
+	print "%16s %5u/%-5u [%03u] %9u.%09u %7s:" % (comm, pid, tid, cpu, ts / 1000000000, ts %1000000000, name),
+
+def print_common_ip(sample, symbol, dso):
+	ip = sample["ip"]
+	print "%16x %s (%s)" % (ip, symbol, dso)
+
+def process_event(param_dict):
+        event_attr = param_dict["attr"]
+        sample     = param_dict["sample"]
+        raw_buf    = param_dict["raw_buf"]
+        comm       = param_dict["comm"]
+        name       = param_dict["ev_name"]
+
+        # Symbol and dso info are not always resolved
+        if (param_dict.has_key("dso")):
+                dso = param_dict["dso"]
+        else:
+                dso = "[unknown]"
+
+        if (param_dict.has_key("symbol")):
+                symbol = param_dict["symbol"]
+        else:
+                symbol = "[unknown]"
+
+	if name == "ptwrite":
+		print_common_start(comm, sample, name)
+		print_ptwrite(raw_buf)
+		print_common_ip(sample, symbol, dso)
+	elif name == "cbr":
+		print_common_start(comm, sample, name)
+		print_cbr(raw_buf)
+		print_common_ip(sample, symbol, dso)
+	elif name == "mwait":
+		print_common_start(comm, sample, name)
+		print_mwait(raw_buf)
+		print_common_ip(sample, symbol, dso)
+	elif name == "pwre":
+		print_common_start(comm, sample, name)
+		print_pwre(raw_buf)
+		print_common_ip(sample, symbol, dso)
+	elif name == "exstop":
+		print_common_start(comm, sample, name)
+		print_exstop(raw_buf)
+		print_common_ip(sample, symbol, dso)
+	elif name == "pwrx":
+		print_common_start(comm, sample, name)
+		print_pwrx(raw_buf)
+		print_common_ip(sample, symbol, dso)
-- 
1.9.1

[PATCH 34/37] perf intel-pt: Update documentation to include new ptwrite and power events

[Adrian Hunter]

Update documentation to include new ptwrite and power events.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/Documentation/intel-pt.txt | 42 +++++++++++++++++++++++++++++++++--
 1 file changed, 40 insertions(+), 2 deletions(-)

diff --git a/tools/perf/Documentation/intel-pt.txt b/tools/perf/Documentation/intel-pt.txt
index d157dee7a4ec..4b6cdbf8f935 100644
--- a/tools/perf/Documentation/intel-pt.txt
+++ b/tools/perf/Documentation/intel-pt.txt
@@ -108,6 +108,9 @@ approach is available to export the data to a postgresql database.  Refer to
 script export-to-postgresql.py for more details, and to script
 call-graph-from-postgresql.py for an example of using the database.
 
+There is also script intel-pt-events.py which provides an example of how to
+unpack the raw data for power events and PTWRITE.
+
 As mentioned above, it is easy to capture too much data.  One way to limit the
 data captured is to use 'snapshot' mode which is explained further below.
 Refer to 'new snapshot option' and 'Intel PT modes of operation' further below.
@@ -710,13 +713,15 @@ Having no option is the same as
 
 which, in turn, is the same as
 
-	--itrace=ibxe
+	--itrace=ibxwpe
 
 The letters are:
 
 	i	synthesize "instructions" events
 	b	synthesize "branches" events
 	x	synthesize "transactions" events
+	w	synthesize "ptwrite" events
+	p	synthesize "power" events
 	c	synthesize branches events (calls only)
 	r	synthesize branches events (returns only)
 	e	synthesize tracing error events
@@ -735,7 +740,40 @@ and "r" can be combined to get calls and returns.
 'flags' field can be used in perf script to determine whether the event is a
 tranasaction start, commit or abort.
 
-Error events are new.  They show where the decoder lost the trace.  Error events
+Note that "instructions", "branches" and "transactions" events depend on code
+flow packets which can be disabled by using the config term "branch=0".  Refer
+to the config terms section above.
+
+"ptwrite" events record the payload of the ptwrite instruction and whether
+"fup_on_ptw" was used.  "ptwrite" events depend on PTWRITE packets which are
+recorded only if the "ptw" config term was used.  Refer to the config terms
+section above.  perf script "synth" field displays "ptwrite" information like
+this: "ip: 0 payload: 0x123456789abcdef0"  where "ip" is 1 if "fup_on_ptw" was
+used.
+
+"Power" events correspond to power event packets and CBR (core-to-bus ratio)
+packets.  While CBR packets are always recorded when tracing is enabled, power
+event packets are recorded only if the "pwr_evt" config term was used.  Refer to
+the config terms section above.  The power events record information about
+C-state changes, whereas CBR is indicative of CPU frequency.  perf script
+"event,synth" fields display information like this:
+	cbr:  cbr: 22 freq: 2189 MHz (200%)
+	mwait:  hints: 0x60 extensions: 0x1
+	pwre:  hw: 0 cstate: 2 sub-cstate: 0
+	exstop:  ip: 1
+	pwrx:  deepest cstate: 2 last cstate: 2 wake reason: 0x4
+Where:
+	"cbr" includes the frequency and the percentage of maximum non-turbo
+	"mwait" shows mwait hints and extensions
+	"pwre" shows C-state transitions (to a C-state deeper than C0) and
+	whether	initiated by hardware
+	"exstop" indicates execution stopped and whether the IP was recorded
+	exactly,
+	"pwrx" indicates return to C0
+For more details refer to the Intel 64 and IA-32 Architectures Software
+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.
 
-- 
1.9.1

[PATCH 35/37] perf intel-pt: Do not use TSC packets for calculating CPU cycles to TSC

[Adrian Hunter]

CBR (core-to-bus ratio) packets provide an indication of CPU frequency. A
more accurate measure can be made by counting the cycles (given by CYC
packets) in between other timing packets (either MTC or TSC). Using TSC
packets has at least 2 issues: 1) timing might have stopped (e.g. mwait) or
2) TSC packets within PSB+ might slip past CYC packets. For now, simply do
not use TSC packets for calculating CPU cycles to TSC. That leaves the case
where 2 MTC packets are used, otherwise falling back to the CBR value.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/util/intel-pt-decoder/intel-pt-decoder.c | 14 ++++++++++++++
 1 file changed, 14 insertions(+)

diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
index 9bb694290a28..a6f41a0e1ae8 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
@@ -690,6 +690,12 @@ static int intel_pt_calc_cyc_cb(struct intel_pt_pkt_info *pkt_info)
 		break;
 
 	case INTEL_PT_TSC:
+		/*
+		 * For now, do not support using TSC packets - refer
+		 * intel_pt_calc_cyc_to_tsc().
+		 */
+		if (data->from_mtc)
+			return 1;
 		timestamp = pkt_info->packet.payload |
 			    (data->timestamp & (0xffULL << 56));
 		if (data->from_mtc && timestamp < data->timestamp &&
@@ -807,6 +813,14 @@ static void intel_pt_calc_cyc_to_tsc(struct intel_pt_decoder *decoder,
 		.cbr_cyc_to_tsc = 0,
 	};
 
+	/*
+	 * For now, do not support using TSC packets for at least the reasons:
+	 * 1) timing might have stopped
+	 * 2) TSC packets within PSB+ can slip against CYC packets
+	 */
+	if (!from_mtc)
+		return;
+
 	intel_pt_pkt_lookahead(decoder, intel_pt_calc_cyc_cb, &data);
 }
 
-- 
1.9.1

[PATCH 36/37] perf auxtrace: Add CPU filter support

[Adrian Hunter]

Decoding auxtrace data can take a long time. To avoid decoding
unnecessarily, filter auxtrace data that is collected per-cpu before it is
decoded.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 tools/perf/builtin-report.c |  1 +
 tools/perf/builtin-script.c |  1 +
 tools/perf/util/auxtrace.c  | 10 ++++++++++
 tools/perf/util/auxtrace.h  |  2 ++
 4 files changed, 14 insertions(+)

diff --git a/tools/perf/builtin-report.c b/tools/perf/builtin-report.c
index 22478ff2b706..6de88f156ec6 100644
--- a/tools/perf/builtin-report.c
+++ b/tools/perf/builtin-report.c
@@ -558,6 +558,7 @@ static int __cmd_report(struct report *rep)
 			ui__error("failed to set cpu bitmap\n");
 			return ret;
 		}
+		session->itrace_synth_opts->cpu_bitmap = rep->cpu_bitmap;
 	}
 
 	if (rep->show_threads) {
diff --git a/tools/perf/builtin-script.c b/tools/perf/builtin-script.c
index 4a13daacba25..4c41d0358966 100644
--- a/tools/perf/builtin-script.c
+++ b/tools/perf/builtin-script.c
@@ -2874,6 +2874,7 @@ int cmd_script(int argc, const char **argv)
 		err = perf_session__cpu_bitmap(session, cpu_list, cpu_bitmap);
 		if (err < 0)
 			goto out_delete;
+		itrace_synth_opts.cpu_bitmap = cpu_bitmap;
 	}
 
 	if (!no_callchain)
diff --git a/tools/perf/util/auxtrace.c b/tools/perf/util/auxtrace.c
index 651c01dfa5d3..5547457566a7 100644
--- a/tools/perf/util/auxtrace.c
+++ b/tools/perf/util/auxtrace.c
@@ -322,6 +322,13 @@ static int auxtrace_queues__add_event_buffer(struct auxtrace_queues *queues,
 	return auxtrace_queues__add_buffer(queues, idx, buffer);
 }
 
+static bool filter_cpu(struct perf_session *session, int cpu)
+{
+	unsigned long *cpu_bitmap = session->itrace_synth_opts->cpu_bitmap;
+
+	return cpu_bitmap && cpu != -1 && !test_bit(cpu, cpu_bitmap);
+}
+
 int auxtrace_queues__add_event(struct auxtrace_queues *queues,
 			       struct perf_session *session,
 			       union perf_event *event, off_t data_offset,
@@ -331,6 +338,9 @@ int auxtrace_queues__add_event(struct auxtrace_queues *queues,
 	unsigned int idx;
 	int err;
 
+	if (filter_cpu(session, event->auxtrace.cpu))
+		return 0;
+
 	buffer = zalloc(sizeof(struct auxtrace_buffer));
 	if (!buffer)
 		return -ENOMEM;
diff --git a/tools/perf/util/auxtrace.h b/tools/perf/util/auxtrace.h
index 68e0aa40b24a..33b5e6cdf38c 100644
--- a/tools/perf/util/auxtrace.h
+++ b/tools/perf/util/auxtrace.h
@@ -74,6 +74,7 @@ enum itrace_period_type {
  * @period: 'instructions' events period
  * @period_type: 'instructions' events period type
  * @initial_skip: skip N events at the beginning.
+ * @cpu_bitmap: CPUs for which to synthesize events, or NULL for all
  */
 struct itrace_synth_opts {
 	bool			set;
@@ -96,6 +97,7 @@ struct itrace_synth_opts {
 	unsigned long long	period;
 	enum itrace_period_type	period_type;
 	unsigned long		initial_skip;
+	unsigned long		*cpu_bitmap;
 };
 
 /**
-- 
1.9.1

[PATCH 37/37] perf intel-pt: Improve sample timestamp

[Adrian Hunter]

The decoder uses its current timestamp in samples. Usually that is a
timestamp that has already passed, but in some cases it is a timestamp for
a branch that the decoder is walking towards, and consequently hasn't
reached. Improve that situation by using the pkt_state to determine when
to use the current or previous timestamp.

Signed-off-by: Adrian Hunter <adrian.hunter@intel.com>
---
 .../perf/util/intel-pt-decoder/intel-pt-decoder.c  | 42 ++++++++++++++++++----
 1 file changed, 35 insertions(+), 7 deletions(-)

diff --git a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
index a6f41a0e1ae8..aa1593ce551d 100644
--- a/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
+++ b/tools/perf/util/intel-pt-decoder/intel-pt-decoder.c
@@ -64,6 +64,25 @@ enum intel_pt_pkt_state {
 	INTEL_PT_STATE_FUP_NO_TIP,
 };
 
+static inline bool intel_pt_sample_time(enum intel_pt_pkt_state pkt_state)
+{
+	switch (pkt_state) {
+	case INTEL_PT_STATE_NO_PSB:
+	case INTEL_PT_STATE_NO_IP:
+	case INTEL_PT_STATE_ERR_RESYNC:
+	case INTEL_PT_STATE_IN_SYNC:
+	case INTEL_PT_STATE_TNT:
+		return true;
+	case INTEL_PT_STATE_TIP:
+	case INTEL_PT_STATE_TIP_PGD:
+	case INTEL_PT_STATE_FUP:
+	case INTEL_PT_STATE_FUP_NO_TIP:
+		return false;
+	default:
+		return true;
+	};
+}
+
 #ifdef INTEL_PT_STRICT
 #define INTEL_PT_STATE_ERR1	INTEL_PT_STATE_NO_PSB
 #define INTEL_PT_STATE_ERR2	INTEL_PT_STATE_NO_PSB
@@ -101,6 +120,7 @@ struct intel_pt_decoder {
 	uint64_t timestamp;
 	uint64_t tsc_timestamp;
 	uint64_t ref_timestamp;
+	uint64_t sample_timestamp;
 	uint64_t ret_addr;
 	uint64_t ctc_timestamp;
 	uint64_t ctc_delta;
@@ -150,6 +170,7 @@ struct intel_pt_decoder {
 	uint64_t fup_pwre_payload;
 	uint64_t cbr_payload;
 	uint64_t timestamp_insn_cnt;
+	uint64_t sample_insn_cnt;
 	uint64_t stuck_ip;
 	int no_progress;
 	int stuck_ip_prd;
@@ -932,6 +953,7 @@ static int intel_pt_walk_insn(struct intel_pt_decoder *decoder,
 
 	decoder->tot_insn_cnt += insn_cnt;
 	decoder->timestamp_insn_cnt += insn_cnt;
+	decoder->sample_insn_cnt += insn_cnt;
 	decoder->period_insn_cnt += insn_cnt;
 
 	if (err) {
@@ -2279,7 +2301,7 @@ static int intel_pt_sync(struct intel_pt_decoder *decoder)
 
 static uint64_t intel_pt_est_timestamp(struct intel_pt_decoder *decoder)
 {
-	uint64_t est = decoder->timestamp_insn_cnt << 1;
+	uint64_t est = decoder->sample_insn_cnt << 1;
 
 	if (!decoder->cbr || !decoder->max_non_turbo_ratio)
 		goto out;
@@ -2287,7 +2309,7 @@ static uint64_t intel_pt_est_timestamp(struct intel_pt_decoder *decoder)
 	est *= decoder->max_non_turbo_ratio;
 	est /= decoder->cbr;
 out:
-	return decoder->timestamp + est;
+	return decoder->sample_timestamp + est;
 }
 
 const struct intel_pt_state *intel_pt_decode(struct intel_pt_decoder *decoder)
@@ -2342,14 +2364,11 @@ const struct intel_pt_state *intel_pt_decode(struct intel_pt_decoder *decoder)
 		}
 	} while (err == -ENOLINK);
 
-	decoder->state.timestamp = decoder->timestamp;
-	decoder->state.est_timestamp = intel_pt_est_timestamp(decoder);
-	decoder->state.cr3 = decoder->cr3;
-	decoder->state.tot_insn_cnt = decoder->tot_insn_cnt;
-
 	if (err) {
 		decoder->state.err = intel_pt_ext_err(err);
 		decoder->state.from_ip = decoder->ip;
+		decoder->sample_timestamp = decoder->timestamp;
+		decoder->sample_insn_cnt = decoder->timestamp_insn_cnt;
 	} else {
 		decoder->state.err = 0;
 		if (decoder->cbr != decoder->cbr_seen && decoder->state.type) {
@@ -2357,8 +2376,17 @@ const struct intel_pt_state *intel_pt_decode(struct intel_pt_decoder *decoder)
 			decoder->state.type |= INTEL_PT_CBR_CHG;
 			decoder->state.cbr_payload = decoder->cbr_payload;
 		}
+		if (intel_pt_sample_time(decoder->pkt_state)) {
+			decoder->sample_timestamp = decoder->timestamp;
+			decoder->sample_insn_cnt = decoder->timestamp_insn_cnt;
+		}
 	}
 
+	decoder->state.timestamp = decoder->sample_timestamp;
+	decoder->state.est_timestamp = intel_pt_est_timestamp(decoder);
+	decoder->state.cr3 = decoder->cr3;
+	decoder->state.tot_insn_cnt = decoder->tot_insn_cnt;
+
 	return &decoder->state;
 }
 
-- 
1.9.1

Re: [PATCH 37/37] perf intel-pt: Improve sample timestamp

[Andi Kleen]

On Fri, May 19, 2017 at 11:54:36AM +0300, Adrian Hunter wrote:
> The decoder uses its current timestamp in samples. Usually that is a
> timestamp that has already passed, but in some cases it is a timestamp for
> a branch that the decoder is walking towards, and consequently hasn't
> reached. Improve that situation by using the pkt_state to determine when
> to use the current or previous timestamp.

Thanks. I ran into this problem a lot.

Would be good to mark the clear bug fixes like this with cc stable.

-Andi

Re: [PATCH 37/37] perf intel-pt: Improve sample timestamp

[Adrian Hunter]

On 19/05/17 17:34, Andi Kleen wrote:
> On Fri, May 19, 2017 at 11:54:36AM +0300, Adrian Hunter wrote:
>> The decoder uses its current timestamp in samples. Usually that is a
>> timestamp that has already passed, but in some cases it is a timestamp for
>> a branch that the decoder is walking towards, and consequently hasn't
>> reached. Improve that situation by using the pkt_state to determine when
>> to use the current or previous timestamp.
> 
> Thanks. I ran into this problem a lot.
> 
> Would be good to mark the clear bug fixes like this with cc stable.

I have done that in V2.