Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1275714 > unrolled thread

[RESEND PATCH V4 00/26] Coresight integration with perf

Back to article view | Back to linux.kernel

Contents

  [RESEND PATCH V4 00/26] Coresight integration with perf Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:40 +0100
    [RESEND PATCH V4 03/26] coresight: etm3x: unlocking tracers in default arch init Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:40 +0100
    [RESEND PATCH V4 22/26] coresight: updating documentation to reflect integration with perf Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:40 +0100
    [RESEND PATCH V4 14/26] coresight: etm3x: adding perf_get/set_config() API Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:40 +0100
    [RESEND PATCH V4 13/26] coresight: etm3x: implementing user/kernel mode tracing Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:40 +0100
    [RESEND PATCH V4 12/26] coresight: etm3x: consolidating initial config Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:40 +0100
    [RESEND PATCH V4 19/26] coresight: etb10: implementing AUX space API Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:40 +0100
    [RESEND PATCH V4 21/26] coresight: introducing a global trace ID function Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:40 +0100
    [RESEND PATCH V4 10/26] coresight: etm3x: set progbit to stop trace collection Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:40 +0100
    [RESEND PATCH V4 16/26] perf: changing pmu::setup_aux() parameter to include event Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:40 +0100
    [RESEND PATCH V4 23/26] perf tools: making function set_max_cpu_num() non static Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:40 +0100
    [RESEND PATCH V4 11/26] coresight: etm3x: changing default trace configuration Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:40 +0100
    [RESEND PATCH V4 20/26] coresight: etm-perf: new PMU driver for ETM tracers Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:40 +0100
    [RESEND PATCH V4 26/26] perf tools: adding coresight etm PMU record capabilities Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:40 +0100
    [RESEND PATCH V4 25/26] perf tools: making coresight PMU listable Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:40 +0100
    [RESEND PATCH V4 24/26] perf tools: adding perf_session to *info_prive_size() Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:40 +0100
    [RESEND PATCH V4 15/26] coresight: etm3x: implementing perf_enable/disable() API Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:40 +0100
    [RESEND PATCH V4 17/26] coresight: etb10: moving to local atomic operations Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:40 +0100
    [RESEND PATCH V4 18/26] coresight: adding operation mode for sink->enable() Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:40 +0100
    [RESEND PATCH V4 01/26] coresight: etm3x: moving etm_readl/writel to header file Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:50 +0100
    [RESEND PATCH V4 05/26] coresight: etm3x: implementing 'cpu_id()' API Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:50 +0100
    [RESEND PATCH V4 09/26] coresight: add API to get sink from path Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:50 +0100
    [RESEND PATCH V4 08/26] coresight: etm3x: adding operation mode for etm_enable() Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:50 +0100
    [RESEND PATCH V4 06/26] coresight: associating path with session rather than tracer Mathieu Poirier <mathieu.poirier@linaro.org> - 2015-11-23 19:50 +0100

Page 1 of 2  [1] 2  Next page →

#1275714 — [RESEND PATCH V4 00/26] Coresight integration with perf

I haven't received much in terms of comments from this iteration.
Even if all comments from previous posts have been addressed it would
feel better to have a few "Reviewed-by:" before adding this to my
tree.

This patchset aims to integrate configuration and control of
the Coresight tracers with the perf sub-system.

The goal is to use PMUs to represent tracers and the auxiliary
buffer enhancement to collect processor traces.  As such a lot
of work is done to move the current Coresight sysFS oriented
configuration and control commands to perf's AUX API.

For the time being the work concentrates on ETMv3 and ETB1.0
sink buffer.  Work on ETMv4 and other type of sink buffers
will follow once a foundation has been established.

Best regards,
Mathieu

Changes since V3:
* Rebased to GKH's char-misc-next tree (e2d8680741ed).

Changes since V2:
* Rebased to v4.3.
* Used the -M option to "git format-patch".
* Decoupled tracers and path - both are now completely independent.
* Moved path enablement/release to Perf's add/del functions.
* Moved PM runtime operations to the coresight core rather than
  individual drivers.
* Enhanced mutual exclusion scheme between perf and sysFS.
* Removed architecture specific define in perf cmd line tool.
* Got rid of unused fields in structure 'cs_buffers'.
* Enhanced changelog for patch 16/26.

Changes since V1:
 * Fixed typos in typographical error in documentation.
 * Moved to a multi session support scheme.
 * Split static and dynamic tracer configuration.
 * Fixed configuration for user and kernel space tracing.
 * Using WARN_ON_ONCE() rather than WARN_ON().
 * Implemented strategy to prevent tracers from being used simultaneously.
 * Changed sink_ops::unset_buffer() to sink_ops::reset_buffer().
 * Moves ETM's sysFS interface from driver core to dedicated file.
 * Removed spinlock in "etm_cpu_id()".
 * Aggregated PMU driver pieces in a single patch.
 * Added user space changes and rebased everything to v4.3-rc5.

Mathieu Poirier (26):
  coresight: etm3x: moving etm_readl/writel to header file
  coresight: etm3x: moving sysFS entries to dedicated file
  coresight: etm3x: unlocking tracers in default arch init
  coresight: etm3x: splitting struct etm_drvdata
  coresight: etm3x: implementing 'cpu_id()' API
  coresight: associating path with session rather than tracer
  coresight: moving PM runtime operations to core framework
  coresight: etm3x: adding operation mode for etm_enable()
  coresight: add API to get sink from path
  coresight: etm3x: set progbit to stop trace collection
  coresight: etm3x: changing default trace configuration
  coresight: etm3x: consolidating initial config
  coresight: etm3x: implementing user/kernel mode tracing
  coresight: etm3x: adding perf_get/set_config() API
  coresight: etm3x: implementing perf_enable/disable() API
  perf: changing pmu::setup_aux() parameter to include event
  coresight: etb10: moving to local atomic operations
  coresight: adding operation mode for sink->enable()
  coresight: etb10: implementing AUX space API
  coresight: etm-perf: new PMU driver for ETM tracers
  coresight: introducing a global trace ID function
  coresight: updating documentation to reflect integration with perf
  perf tools: making function set_max_cpu_num() non static
  perf tools: adding perf_session to *info_prive_size()
  perf tools: making coresight PMU listable
  perf tools: adding coresight etm PMU record capabilities

 Documentation/trace/coresight.txt                  |  138 +-
 MAINTAINERS                                        |    4 +
 arch/x86/kernel/cpu/perf_event_intel_bts.c         |    4 +-
 arch/x86/kernel/cpu/perf_event_intel_pt.c          |    5 +-
 drivers/hwtracing/coresight/Kconfig                |    1 +
 drivers/hwtracing/coresight/Makefile               |    4 +-
 drivers/hwtracing/coresight/coresight-etb10.c      |  268 ++-
 drivers/hwtracing/coresight/coresight-etm-perf.c   |  438 +++++
 drivers/hwtracing/coresight/coresight-etm-perf.h   |   32 +
 drivers/hwtracing/coresight/coresight-etm.h        |  143 +-
 .../hwtracing/coresight/coresight-etm3x-sysfs.c    | 1442 +++++++++++++++
 drivers/hwtracing/coresight/coresight-etm3x.c      | 1875 +++++---------------
 drivers/hwtracing/coresight/coresight-etm4x.c      |    8 +-
 drivers/hwtracing/coresight/coresight-funnel.c     |    2 -
 drivers/hwtracing/coresight/coresight-priv.h       |   15 +
 .../coresight/coresight-replicator-qcom.c          |    4 -
 drivers/hwtracing/coresight/coresight-replicator.c |    2 -
 drivers/hwtracing/coresight/coresight-tmc.c        |    7 +-
 drivers/hwtracing/coresight/coresight-tpiu.c       |    4 +-
 drivers/hwtracing/coresight/coresight.c            |  330 ++--
 include/linux/coresight-pmu.h                      |   39 +
 include/linux/coresight.h                          |   36 +-
 include/linux/perf_event.h                         |    2 +-
 kernel/events/ring_buffer.c                        |    2 +-
 tools/perf/arch/arm/util/Build                     |    2 +
 tools/perf/arch/arm/util/auxtrace.c                |   48 +
 tools/perf/arch/arm/util/cs_etm.c                  |  466 +++++
 tools/perf/arch/arm/util/cs_etm.h                  |   44 +
 tools/perf/arch/arm/util/pmu.c                     |   34 +
 tools/perf/arch/x86/util/intel-bts.c               |    4 +-
 tools/perf/arch/x86/util/intel-pt.c                |    4 +-
 tools/perf/config/Makefile                         |   17 +-
 tools/perf/util/auxtrace.c                         |    8 +-
 tools/perf/util/auxtrace.h                         |    7 +-
 tools/perf/util/cpumap.c                           |    2 +-
 tools/perf/util/cpumap.h                           |    1 +
 36 files changed, 3787 insertions(+), 1655 deletions(-)
 create mode 100644 drivers/hwtracing/coresight/coresight-etm-perf.c
 create mode 100644 drivers/hwtracing/coresight/coresight-etm-perf.h
 create mode 100644 drivers/hwtracing/coresight/coresight-etm3x-sysfs.c
 create mode 100644 include/linux/coresight-pmu.h
 create mode 100644 tools/perf/arch/arm/util/auxtrace.c
 create mode 100644 tools/perf/arch/arm/util/cs_etm.c
 create mode 100644 tools/perf/arch/arm/util/cs_etm.h
 create mode 100644 tools/perf/arch/arm/util/pmu.c

-- 
1.9.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [next] | [standalone]

#1275715 — [RESEND PATCH V4 03/26] coresight: etm3x: unlocking tracers in default arch init

Calling function 'smp_call_function_single()' to unlock a
tracer and calling it again right after to perform the
default initialisation doesn't make sense.

Moving 'etm_os_unlock()' just before making the default
initialisation results in the same outcome while saving
one call to 'smp_call_function_single()'.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 drivers/hwtracing/coresight/coresight-etm3x.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/drivers/hwtracing/coresight/coresight-etm3x.c b/drivers/hwtracing/coresight/coresight-etm3x.c
index 2a1950d0753d..737b6164f4a5 100644
--- a/drivers/hwtracing/coresight/coresight-etm3x.c
+++ b/drivers/hwtracing/coresight/coresight-etm3x.c
@@ -47,11 +47,11 @@ static struct etm_drvdata *etmdrvdata[NR_CPUS];
  * and OS lock must be unlocked before any memory mapped access on such
  * processors, otherwise memory mapped reads/writes will be invalid.
  */
-static void etm_os_unlock(void *info)
+static void etm_os_unlock(struct etm_drvdata *drvdata)
 {
-	struct etm_drvdata *drvdata = (struct etm_drvdata *)info;
 	/* Writing any value to ETMOSLAR unlocks the trace registers */
 	etm_writel(drvdata, 0x0, ETMOSLAR);
+	drvdata->os_unlock = true;
 	isb();
 }
 
@@ -478,6 +478,9 @@ static void etm_init_arch_data(void *info)
 	u32 etmccr;
 	struct etm_drvdata *drvdata = info;
 
+	/* Make sure all registers are accessible */
+	etm_os_unlock(drvdata);
+
 	CS_UNLOCK(drvdata->base);
 
 	/* First dummy read */
@@ -602,9 +605,6 @@ static int etm_probe(struct amba_device *adev, const struct amba_id *id)
 	get_online_cpus();
 	etmdrvdata[drvdata->cpu] = drvdata;
 
-	if (!smp_call_function_single(drvdata->cpu, etm_os_unlock, drvdata, 1))
-		drvdata->os_unlock = true;
-
 	if (smp_call_function_single(drvdata->cpu,
 				     etm_init_arch_data,  drvdata, 1))
 		dev_err(dev, "ETM arch init failed\n");
-- 
1.9.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1275716 — [RESEND PATCH V4 22/26] coresight: updating documentation to reflect integration with perf

Adding a new section giving information on how coresight has been
integrated with the perf subsystem along with a general idea of how
to control tracing from the perf tool cmd line.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 Documentation/trace/coresight.txt | 138 ++++++++++++++++++++++++++++++++++----
 1 file changed, 124 insertions(+), 14 deletions(-)

diff --git a/Documentation/trace/coresight.txt b/Documentation/trace/coresight.txt
index 0a5c3290e732..9515206b4b1a 100644
--- a/Documentation/trace/coresight.txt
+++ b/Documentation/trace/coresight.txt
@@ -20,13 +20,13 @@ Components are generally categorised as source, link and sinks and are
 
 "Sources" generate a compressed stream representing the processor instruction
 path based on tracing scenarios as configured by users.  From there the stream
-flows through the coresight system (via ATB bus) using links that are connecting
-the emanating source to a sink(s).  Sinks serve as endpoints to the coresight
+flows through the Coresight system (via ATB bus) using links that are connecting
+the emanating source to a sink(s).  Sinks serve as endpoints to the Coresight
 implementation, either storing the compressed stream in a memory buffer or
 creating an interface to the outside world where data can be transferred to a
-host without fear of filling up the onboard coresight memory buffer.
+host without fear of filling up the onboard Coresight memory buffer.
 
-At typical coresight system would look like this:
+At typical Coresight system would look like this:
 
   *****************************************************************
  **************************** AMBA AXI  ****************************===||
@@ -83,8 +83,8 @@ While on target configuration of the components is done via the APB bus,
 all trace data are carried out-of-band on the ATB bus.  The CTM provides
 a way to aggregate and distribute signals between CoreSight components.
 
-The coresight framework provides a central point to represent, configure and
-manage coresight devices on a platform.  This first implementation centers on
+The Coresight framework provides a central point to represent, configure and
+manage Coresight devices on a platform.  This first implementation centers on
 the basic tracing functionality, enabling components such ETM/PTM, funnel,
 replicator, TMC, TPIU and ETB.  Future work will enable more
 intricate IP blocks such as STM and CTI.
@@ -129,11 +129,11 @@ expected to be added as the solution matures.
 Framework and implementation
 ----------------------------
 
-The coresight framework provides a central point to represent, configure and
-manage coresight devices on a platform.  Any coresight compliant device can
+The Coresight framework provides a central point to represent, configure and
+manage Coresight devices on a platform.  Any Coresight compliant device can
 register with the framework for as long as they use the right APIs:
 
-struct coresight_device *coresight_register(struct coresight_desc *desc);
+struct Coresight_device *coresight_register(struct coresight_desc *desc);
 void coresight_unregister(struct coresight_device *csdev);
 
 The registering function is taking a "struct coresight_device *csdev" and
@@ -193,10 +193,120 @@ the information carried in "THIS_MODULE".
 How to use
 ----------
 
-Before trace collection can start, a coresight sink needs to be identify.
-There is no limit on the amount of sinks (nor sources) that can be enabled at
-any given moment.  As a generic operation, all device pertaining to the sink
-class will have an "active" entry in sysfs:
+There is two ways to use the Coresight framework: 1) using the perf cmd line
+tool and 2) interacting directly with the Coresight devices using the sysFS
+interface.  The latter will slowly be faded out as more functionality become
+available from the perf cmd line tool but for the time being both are still
+supported.  The following sections provide details on using both methods.
+
+1) Using perf framework:
+
+Coresight tracers like ETM and PTM are represented using the Perf framework's
+Performance Monitoring Unit (PMU).  As such the perf framework takes charge of
+controlling when tracing happens based on when the process(es) of interest are
+scheduled.  When configure in a system, Coresight PMUs will be listed when
+queried by the perf command line tool:
+
+linaro@linaro-nano:~$ ./perf list pmu
+
+List of pre-defined events (to be used in -e):
+
+  cs_etm//                                           [Kernel PMU event]
+
+linaro@linaro-nano:~$
+
+Regardless of the amount ETM/PTM IP block in a system (usually equal to the
+amount of processor core), the "cs_etm" PMU will be listed only once.
+
+Before a trace can be configured and started a Coresight sink needs to be
+selected using the sysFS method (see below).  This is only temporary until
+sink selection can be made from the command line tool.
+
+linaro@linaro-nano:~$ ls /sys/bus/coresight/devices
+20010000.etb  20030000.tpiu  20040000.funnel  2201c000.ptm
+2201d000.ptm  2203c000.etm  2203d000.etm  2203e000.etm  replicator
+
+linaro@linaro-nano:~$ echo 1 > /sys/bus/coresight/devices/20010000.etb/enable_sink
+
+Once a sink has been selected configuring a Coresight PMU works the same way as
+any other PMU.  As such tracing can happen for a single CPU, a group of CPU, per
+thread or a combination of those:
+
+linaro@linaro-nano:~$ perf record -e cs_etm// --per-thread <command>
+
+linaro@linaro-nano:~$ perf record -C 0,2-3 -e cs_etm// <command>
+
+Tracing limited to user and kernel space can also be used to narrow the amount
+of collected traces:
+
+linaro@linaro-nano:~$ perf record -e cs_etm//u --per-thread <command>
+
+linaro@linaro-nano:~$ perf record -C 0,2-3 -e cs_etm//k <command>
+
+As of this writing two ETM/PTM specific options have are available: cycle
+accurate and timestamp (please refer to the Embedded Trace Macrocell reference
+manual for details on these options).  By default both are disabled but using
+the "cycacc" and "timestamp" mnemonic within the double '/' will see those
+options configure for the upcoming trace run:
+
+linaro@linaro-nano:~$ perf record -e cs_etm/cycacc/ --per-thread <command>
+
+linaro@linaro-nano:~$ perf record -C 0,2-3 -e cs_etm/cycacc,timestamp/ <command>
+
+The Coresight PMUs can be configured to work in "full trace" or "snapshot" mode.
+In full trace mode trace acquisition is enabled from beginning to end with trace
+data being recorded continuously:
+
+linaro@linaro-nano:~$ perf record -e cs_etm// dd if=/dev/random of=./test.txt bs=1k count=1000
+
+Since this can lead to a significant amount of data and because some devices are
+limited in disk space snapshot mode can be used instead.  In snapshot mode
+traces are still collected in the ring buffer but not communicated to user
+space.  The ring buffer is allowed to wrap around, providing the latest
+information before an event of interest happens.  Significant events are
+communicated by sending a USR2 signal to the user space command line tool.
+From there the tool will stop trace collection and harvest data from the ring
+buffer before re-enabling traces.  Snapshot mode can be invoked using '-S' when
+launching a trace collection:
+
+linaro@linaro-nano:~$ perf record -S -e cs_etm// dd if=/dev/random of=./test.txt bs=1k count=1000
+
+Trace data collected during trace runs ends up in the "perf.data" file.  Trace
+configuration information necessary for trace decoding is also embedded in the
+"perf.data" file.  Two new headers, 'PERF_RECORD_AUXTRACE_INFO' and
+'PERF_RECORD_AUXTRACE' have been added to the list of event types in order to
+find out where the different sections start.
+
+It is worth noting that a set of metadata information exists for each tracer
+that participated in a trace run.  As such if 5 processors have been engaged,
+5 sets of metadata will be found in the perf.data file.  This is to ensure that
+tracer decompression tools have all the information they need in order to
+process the trace data.
+
+Metadata information is collected directly from the ETM/PTM management registers
+using the sysFS interface.  Since there is no way for the perf command line
+tool to associate a CPU with a tracer, a symbolic link has been created between
+the cs_etm sysFS event directory and each Coresight tracer:
+
+linaro@linaro-nano:~$ ls /sys/bus/event_source/devices/cs_etm
+cpu0  cpu1  cpu2  cpu3  cpu4  format  perf_event_mux_interval_ms
+power  subsystem  type  uevent
+
+linaro@linaro-nano:~$ ls /sys/bus/event_source/devices/cs_etm/cpu0/mgmt/
+etmccer  etmccr  etmcr  etmidr  etmscr  etmtecr1  etmtecr2
+etmteevr  etmtraceidr  etmtssvr
+
+2) Using the sysFS interface:
+
+Most, if not all, configuration registers are made available to users via the
+sysFS interface.  Until all Coresight ETM drivers have been converted to perf,
+it will also be possible to start and stop traces from sysFS.
+
+As with the perf method described above, a Coresight sink needs to be identify
+before trace collection can commence.  Using the sysFS method _only_, there is
+no limit on the amount of sinks (nor sources) that can be enabled at
+any given moment.  As a generic operation, all devices pertaining to the sink
+class will have an "enable_sink" entry in sysfs:
 
 root:/sys/bus/coresight/devices# ls
 replicator  20030000.tpiu    2201c000.ptm  2203c000.etm  2203e000.etm
@@ -246,7 +356,7 @@ The file cstrace.bin can be decompressed using "ptm2human", DS-5 or Trace32.
 
 Following is a DS-5 output of an experimental loop that increments a variable up
 to a certain value.  The example is simple and yet provides a glimpse of the
-wealth of possibilities that coresight provides.
+wealth of possibilities that Coresight provides.
 
 Info                                    Tracing enabled
 Instruction     106378866       0x8026B53C      E52DE004        false   PUSH     {lr}
-- 
1.9.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1275717 — [RESEND PATCH V4 14/26] coresight: etm3x: adding perf_get/set_config() API

Adding a source operation to build a tracer's configuration from
a perf_event.  That way possibly complex parsing of the information
conveyed by the event doesn't have to be carried out every time
the configuration is needed.

Since event configuration can change between concurrent sessions,
the possibility of associating a tracer with a configuration is
also provided.  As such Perf can assign session configuration to
tracers as it see fit.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 drivers/hwtracing/coresight/Kconfig           |  1 +
 drivers/hwtracing/coresight/coresight-etm3x.c | 62 +++++++++++++++++++++++++++
 include/linux/coresight.h                     |  7 ++-
 3 files changed, 69 insertions(+), 1 deletion(-)

diff --git a/drivers/hwtracing/coresight/Kconfig b/drivers/hwtracing/coresight/Kconfig
index 6c8921140f02..e252dd1522e5 100644
--- a/drivers/hwtracing/coresight/Kconfig
+++ b/drivers/hwtracing/coresight/Kconfig
@@ -4,6 +4,7 @@
 menuconfig CORESIGHT
 	bool "CoreSight Tracing Support"
 	select ARM_AMBA
+	select PERF_EVENTS
 	help
 	  This framework provides a kernel interface for the CoreSight debug
 	  and trace drivers to register themselves with. It's intended to build
diff --git a/drivers/hwtracing/coresight/coresight-etm3x.c b/drivers/hwtracing/coresight/coresight-etm3x.c
index a83c67d13b21..dd319ef0f1ac 100644
--- a/drivers/hwtracing/coresight/coresight-etm3x.c
+++ b/drivers/hwtracing/coresight/coresight-etm3x.c
@@ -31,6 +31,7 @@
 #include <linux/seq_file.h>
 #include <linux/uaccess.h>
 #include <linux/clk.h>
+#include <linux/perf_event.h>
 #include <asm/sections.h>
 
 #include "coresight-etm.h"
@@ -315,6 +316,40 @@ void etm_config_trace_mode(struct etm_drvdata *drvdata,
 	config->addr_type[1] = ETM_ADDR_TYPE_RANGE;
 }
 
+#define ETM3X_SUPPORTED_OPTIONS (ETMCR_CYC_ACC | ETMCR_TIMESTAMP_EN)
+
+static int etm_parse_event_config(struct etm_drvdata *drvdata,
+				  struct etm_config *config,
+				  struct perf_event *event)
+{
+	u32 mode = 0;
+	u64 event_config = event->attr.config;
+
+	if (event->attr.exclude_kernel)
+		mode = ETM_MODE_EXCL_KERN;
+
+	if (event->attr.exclude_user)
+		mode = ETM_MODE_EXCL_USER;
+
+	/*
+	 * By default the tracers are configured to trace the whole address
+	 * range.  Narrow the field only if requested by user space.
+	 */
+	if (mode)
+		etm_config_trace_mode(drvdata, config, mode);
+
+	/*
+	 * At this time only cycle accurate and timestamp options are
+	 * available.
+	 */
+	if (event_config & ~ETM3X_SUPPORTED_OPTIONS)
+		return -EINVAL;
+
+	config->ctrl = event_config;
+
+	return 0;
+}
+
 static void etm_enable_hw(void *info)
 {
 	int i;
@@ -428,6 +463,31 @@ static int etm_trace_id(struct coresight_device *csdev)
 	return etm_get_trace_id(drvdata);
 }
 
+static void *etm_get_config(struct coresight_device *csdev,
+			    struct perf_event *event)
+{
+	struct etm_config *config = NULL;
+	struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+
+	config = kzalloc(sizeof(struct etm_config), GFP_KERNEL);
+	if (!config)
+		return config;
+
+	etm_set_default(config);
+
+	if (etm_parse_event_config(drvdata, config, event))
+		return ERR_PTR(-EINVAL);
+
+	return config;
+}
+
+static void etm_set_config(struct coresight_device *csdev, void *config)
+{
+	struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+
+	drvdata->config = config;
+}
+
 static int etm_enable_sysfs(struct coresight_device *csdev)
 {
 	struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
@@ -568,6 +628,8 @@ static void etm_disable(struct coresight_device *csdev)
 static const struct coresight_ops_source etm_source_ops = {
 	.cpu_id		= etm_cpu_id,
 	.trace_id	= etm_trace_id,
+	.get_config	= etm_get_config,
+	.set_config	= etm_set_config,
 	.enable		= etm_enable,
 	.disable	= etm_disable,
 };
diff --git a/include/linux/coresight.h b/include/linux/coresight.h
index 61dfb8d511ea..f2148bc2bcfe 100644
--- a/include/linux/coresight.h
+++ b/include/linux/coresight.h
@@ -206,13 +206,18 @@ struct coresight_ops_link {
  * @cpu_id:	returns the value of the CPU number this component
  *		is associated to.
  * @trace_id:	returns the value of the component's trace ID as known
-		to the HW.
+ *		to the HW.
+ * @get_config:	builds the ETM configuration after events' specifics.
+ * @set_config:	associate a tracer with a configuration.
  * @enable:	enables tracing for a source.
  * @disable:	disables tracing for a source.
  */
 struct coresight_ops_source {
 	int (*cpu_id)(struct coresight_device *csdev);
 	int (*trace_id)(struct coresight_device *csdev);
+	void *(*get_config)(struct coresight_device *csdev,
+			    struct perf_event *event);
+	void (*set_config)(struct coresight_device *csdev, void *config);
 	int (*enable)(struct coresight_device *csdev, u32 mode);
 	void (*disable)(struct coresight_device *csdev);
 };
-- 
1.9.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1275718 — [RESEND PATCH V4 13/26] coresight: etm3x: implementing user/kernel mode tracing

Adding new mode to limit tracing to kernel or user space.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 drivers/hwtracing/coresight/coresight-etm.h        |  7 ++-
 .../hwtracing/coresight/coresight-etm3x-sysfs.c    |  4 ++
 drivers/hwtracing/coresight/coresight-etm3x.c      | 63 ++++++++++++++++++++++
 drivers/hwtracing/coresight/coresight-priv.h       |  3 ++
 4 files changed, 76 insertions(+), 1 deletion(-)

diff --git a/drivers/hwtracing/coresight/coresight-etm.h b/drivers/hwtracing/coresight/coresight-etm.h
index f01862fc98dd..e72815073b63 100644
--- a/drivers/hwtracing/coresight/coresight-etm.h
+++ b/drivers/hwtracing/coresight/coresight-etm.h
@@ -110,7 +110,10 @@
 #define ETM_MODE_STALL		BIT(2)
 #define ETM_MODE_TIMESTAMP	BIT(3)
 #define ETM_MODE_CTXID		BIT(4)
-#define ETM_MODE_ALL		0x1f
+#define ETM_MODE_ALL		(ETM_MODE_EXCLUDE | ETM_MODE_CYCACC | \
+				 ETM_MODE_STALL | ETM_MODE_TIMESTAMP | \
+				 ETM_MODE_CTXID | ETM_MODE_EXCL_KERN | \
+				 ETM_MODE_EXCL_USER)
 
 #define ETM_SQR_MASK		0x3
 #define ETM_TRACEID_MASK	0x3f
@@ -296,5 +299,7 @@ static inline unsigned int etm_readl(struct etm_drvdata *drvdata, u32 off)
 extern const struct attribute_group *coresight_etm_groups[];
 int etm_get_trace_id(struct etm_drvdata *drvdata);
 void etm_set_default(struct etm_config *config);
+void etm_config_trace_mode(struct etm_drvdata *drvdata,
+			   struct etm_config *config, u32 mode);
 struct etm_config *get_etm_config(struct etm_drvdata *drvdata);
 #endif
diff --git a/drivers/hwtracing/coresight/coresight-etm3x-sysfs.c b/drivers/hwtracing/coresight/coresight-etm3x-sysfs.c
index 6da50d94ec6f..f7318f04118d 100644
--- a/drivers/hwtracing/coresight/coresight-etm3x-sysfs.c
+++ b/drivers/hwtracing/coresight/coresight-etm3x-sysfs.c
@@ -176,6 +176,10 @@ static ssize_t mode_store(struct device *dev,
 		config->ctrl |= ETMCR_CTXID_SIZE;
 	else
 		config->ctrl &= ~ETMCR_CTXID_SIZE;
+
+	if (config->mode & (ETM_MODE_EXCL_KERN | ETM_MODE_EXCL_USER))
+		etm_config_trace_mode(drvdata, config, config->mode);
+
 	spin_unlock(&drvdata->spinlock);
 
 	return size;
diff --git a/drivers/hwtracing/coresight/coresight-etm3x.c b/drivers/hwtracing/coresight/coresight-etm3x.c
index e973e4d13bcb..a83c67d13b21 100644
--- a/drivers/hwtracing/coresight/coresight-etm3x.c
+++ b/drivers/hwtracing/coresight/coresight-etm3x.c
@@ -252,6 +252,69 @@ void etm_set_default(struct etm_config *config)
 	config->ctxid_mask = 0x0;
 }
 
+void etm_config_trace_mode(struct etm_drvdata *drvdata,
+			   struct etm_config *config,
+			   u32 mode)
+{
+	u32 flags;
+
+	mode &= (ETM_MODE_EXCL_KERN | ETM_MODE_EXCL_USER);
+
+	/* excluding kernel AND user space doesn't make sense */
+	if (mode == (ETM_MODE_EXCL_KERN | ETM_MODE_EXCL_USER))
+		return;
+
+	/* nothing to do if neither flags are set */
+	if (!(mode & ETM_MODE_EXCL_KERN) && !(mode & ETM_MODE_EXCL_USER))
+		return;
+
+	flags = (1 << 0 |	/* instruction execute */
+		 3 << 3 |	/* ARM instruction */
+		 0 << 5 |	/* No data value comparison */
+		 0 << 7 |	/* No exact mach */
+		 0 << 8);	/* Ignore context ID */
+
+	/* No need to worry about single address comparators. */
+	config->enable_ctrl2 = 0x0;
+
+	/* Bit 0 is address range comparator 1 */
+	config->enable_ctrl1 = ETMTECR1_ADDR_COMP_1;
+
+	/*
+	 * On ETMv3.5:
+	 * ETMACTRn[13,11] == Non-secure state comparison control
+	 * ETMACTRn[12,10] == Secure state comparison control
+	 *
+	 * b00 == Match in all modes in this state
+	 * b01 == Do not match in any more in this state
+	 * b10 == Match in all modes excepts user mode in this state
+	 * b11 == Match only in user mode in this state
+	 */
+
+	/* Tracing in secure mode is not supported at this time */
+	flags |= (0 << 12 | 1 << 10);
+
+	if (mode & ETM_MODE_EXCL_USER) {
+		/* exclude user, match all modes except user mode */
+		flags |= (1 << 13 | 0 << 11);
+	} else {
+		/* exclude kernel, match only in user mode */
+		flags |= (1 << 13 | 1 << 11);
+	}
+
+	/*
+	 * The ETMEEVR register is already set to "hard wire A".  As such
+	 * all there is to do is setup an address comparator that spans
+	 * the entire address range and configure the state and mode bits.
+	 */
+	config->addr_val[0] = (u32) 0x0;
+	config->addr_val[1] = (u32) ~0x0;
+	config->addr_acctype[0] = flags;
+	config->addr_acctype[1] = flags;
+	config->addr_type[0] = ETM_ADDR_TYPE_RANGE;
+	config->addr_type[1] = ETM_ADDR_TYPE_RANGE;
+}
+
 static void etm_enable_hw(void *info)
 {
 	int i;
diff --git a/drivers/hwtracing/coresight/coresight-priv.h b/drivers/hwtracing/coresight/coresight-priv.h
index ed116b303e87..932f34a84d96 100644
--- a/drivers/hwtracing/coresight/coresight-priv.h
+++ b/drivers/hwtracing/coresight/coresight-priv.h
@@ -34,6 +34,9 @@
 #define TIMEOUT_US		100
 #define BMVAL(val, lsb, msb)	((val & GENMASK(msb, lsb)) >> lsb)
 
+#define ETM_MODE_EXCL_KERN	BIT(30)
+#define ETM_MODE_EXCL_USER	BIT(31)
+
 enum cs_mode {
 	CS_MODE_DISABLED,
 	CS_MODE_SYSFS,
-- 
1.9.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1275719 — [RESEND PATCH V4 12/26] coresight: etm3x: consolidating initial config

There is really no point in having two functions to take care
of doing the initials tracer configuration.  As such moving
everything to 'etm_set_default()'.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 drivers/hwtracing/coresight/coresight-etm3x.c | 37 ++++++++++-----------------
 1 file changed, 14 insertions(+), 23 deletions(-)

diff --git a/drivers/hwtracing/coresight/coresight-etm3x.c b/drivers/hwtracing/coresight/coresight-etm3x.c
index 4c54e104b610..e973e4d13bcb 100644
--- a/drivers/hwtracing/coresight/coresight-etm3x.c
+++ b/drivers/hwtracing/coresight/coresight-etm3x.c
@@ -41,7 +41,6 @@ module_param_named(boot_enable, boot_enable, int, S_IRUGO);
 /* The number of ETM/PTM currently registered */
 static int etm_count;
 static struct etm_drvdata *etmdrvdata[NR_CPUS];
-static void etm_init_default_data(struct etm_config *config);
 
 /*
  * Memory mapped writes to clear os lock are not supported on some processors
@@ -199,7 +198,7 @@ struct etm_config *get_etm_config(struct etm_drvdata *drvdata)
 		return NULL;
 
 	/* Set default config */
-	etm_init_default_data(config);
+	etm_set_default(config);
 	drvdata->config = config;
 out:
 	return drvdata->config;
@@ -212,6 +211,19 @@ void etm_set_default(struct etm_config *config)
 	if (WARN_ON_ONCE(!config))
 		return;
 
+	/*
+	 * Taken verbatim from the TRM:
+	 *
+	 * To trace all memory:
+	 *  set bit [24] in register 0x009, the ETMTECR1, to 1
+	 *  set all other bits in register 0x009, the ETMTECR1, to 0
+	 *  set all bits in register 0x007, the ETMTECR2, to 0
+	 *  set register 0x008, the ETMTEEVR, to 0x6F (TRUE).
+	 */
+	config->enable_ctrl1 = BIT(24);
+	config->enable_ctrl2 = 0x0;
+	config->enable_event = ETM_HARD_WIRE_RES_A;
+
 	config->trigger_event = ETM_DEFAULT_EVENT_VAL;
 	config->enable_event = ETM_HARD_WIRE_RES_A;
 
@@ -605,27 +617,6 @@ static void etm_init_arch_data(void *info)
 	CS_LOCK(drvdata->base);
 }
 
-static void etm_init_default_data(struct etm_config *config)
-{
-	if (WARN_ON_ONCE(!config))
-		return;
-
-	etm_set_default(config);
-
-	/*
-	 * Taken verbatim from the TRM:
-	 *
-	 * To trace all memory:
-	 *  set bit [24] in register 0x009, the ETMTECR1, to 1
-	 *  set all other bits in register 0x009, the ETMTECR1, to 0
-	 *  set all bits in register 0x007, the ETMTECR2, to 0
-	 *  set register 0x008, the ETMTEEVR, to 0x6F (TRUE).
-	 */
-	config->enable_ctrl1 = BIT(24);
-	config->enable_ctrl2 = 0x0;
-	config->enable_event = ETM_HARD_WIRE_RES_A;
-}
-
 static void etm_init_trace_id(struct etm_drvdata *drvdata)
 {
 	/*
-- 
1.9.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1275720 — [RESEND PATCH V4 19/26] coresight: etb10: implementing AUX space API

Adding an ETB10 specific AUX area operations to be used
by the perf framework when events are initialised.

Part of this operation involves modeling the mmap'ed area
based on the specific ways a sink buffer gathers information.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 drivers/hwtracing/coresight/coresight-etb10.c | 228 ++++++++++++++++++++++++++
 include/linux/coresight.h                     |  20 ++-
 2 files changed, 246 insertions(+), 2 deletions(-)

diff --git a/drivers/hwtracing/coresight/coresight-etb10.c b/drivers/hwtracing/coresight/coresight-etb10.c
index dc4d707b28aa..27ddf9ecc7cc 100644
--- a/drivers/hwtracing/coresight/coresight-etb10.c
+++ b/drivers/hwtracing/coresight/coresight-etb10.c
@@ -28,6 +28,11 @@
 #include <linux/coresight.h>
 #include <linux/amba/bus.h>
 #include <linux/clk.h>
+#include <linux/circ_buf.h>
+#include <linux/mm.h>
+#include <linux/perf_event.h>
+
+#include <asm/local.h>
 
 #include "coresight-priv.h"
 
@@ -65,6 +70,26 @@
 #define ETB_FRAME_SIZE_WORDS	4
 
 /**
+ * struct cs_buffer - keep track of a recording session' specifics
+ * @cur:	index of the current buffer
+ * @nr_pages:	max number of pages granted to us
+ * @offset:	offset within the current buffer
+ * @data_size:	how much we collected in this run
+ * @lost:	other than zero if we had a HW buffer wrap around
+ * @snapshot:	is this run in snapshot mode
+ * @data_pages:	a handle the ring buffer
+ */
+struct cs_buffers {
+	unsigned int		cur;
+	unsigned int		nr_pages;
+	unsigned long		offset;
+	local_t			data_size;
+	local_t			lost;
+	bool			snapshot;
+	void			**data_pages;
+};
+
+/**
  * struct etb_drvdata - specifics associated to an ETB component
  * @base:	memory mapped base address for this component.
  * @dev:	the device entity associated to this component.
@@ -266,9 +291,212 @@ static void etb_disable(struct coresight_device *csdev)
 	dev_info(drvdata->dev, "ETB disabled\n");
 }
 
+static void *etb_setup_aux(struct coresight_device *csdev, int cpu,
+			   void **pages, int nr_pages, bool overwrite)
+{
+	int node;
+	struct cs_buffers *buf;
+
+	if (cpu == -1)
+		cpu = smp_processor_id();
+	node = cpu_to_node(cpu);
+
+	buf = kzalloc_node(sizeof(struct cs_buffers), GFP_KERNEL, node);
+	if (!buf)
+		return NULL;
+
+	buf->snapshot = overwrite;
+	buf->nr_pages = nr_pages;
+	buf->data_pages = pages;
+
+	return buf;
+}
+
+static int etb_set_buffer(struct coresight_device *csdev,
+			  struct perf_output_handle *handle,
+			  void *sink_config)
+{
+	int ret = 0;
+	unsigned long head;
+	struct cs_buffers *buf = sink_config;
+
+	/* wrap head around to the amount of space we have */
+	head = handle->head & ((buf->nr_pages << PAGE_SHIFT) - 1);
+
+	/* find the page to write to */
+	buf->cur = head / PAGE_SIZE;
+
+	/* and offset within that page */
+	buf->offset = head % PAGE_SIZE;
+
+	local_set(&buf->data_size, 0);
+
+	return ret;
+}
+
+static unsigned long etb_reset_buffer(struct coresight_device *csdev,
+				      struct perf_output_handle *handle,
+				      void *sink_config, bool *lost)
+{
+	unsigned long size = 0;
+	struct cs_buffers *buf = sink_config;
+
+	if (buf) {
+		/*
+		 * In snapshot mode ->data_size holds the new address of the
+		 * ring buffer's head.  The size itself is the whole address
+		 * range since we want the latest information.
+		 */
+		if (buf->snapshot)
+			handle->head = local_xchg(&buf->data_size,
+						  buf->nr_pages << PAGE_SHIFT);
+
+		/*
+		 * Tell the tracer PMU how much we got in this run and if
+		 * something went wrong along the way.  Nobody else can use
+		 * this cs_buffers instance until we are done.  As such
+		 * resetting parameters here and squaring off with the ring
+		 * buffer API in the tracer PMU is fine.
+		 */
+		*lost = local_xchg(&buf->lost, 0);
+		size = local_xchg(&buf->data_size, 0);
+	}
+
+	return size;
+}
+
+static void etb_update_buffer(struct coresight_device *csdev,
+			      struct perf_output_handle *handle,
+			      void *sink_config)
+{
+	int i, cur;
+	u8 *buf_ptr;
+	u32 read_ptr, write_ptr, start;
+	u32 status, read_data, to_read;
+	unsigned long flags, offset;
+	struct cs_buffers *buf = sink_config;
+	struct etb_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+
+	if (!buf)
+		return;
+
+	spin_lock_irqsave(&drvdata->spinlock, flags);
+
+	CS_UNLOCK(drvdata->base);
+	etb_disable_hw(drvdata);
+
+	/* unit is in words, not bytes */
+	read_ptr = readl_relaxed(drvdata->base + ETB_RAM_READ_POINTER);
+	write_ptr = readl_relaxed(drvdata->base + ETB_RAM_WRITE_POINTER);
+
+	/*
+	 * Entries should be aligned to the frame size.  If they are not
+	 * go back to the last alignement point to give decoding tools a
+	 * chance to fix things.
+	 */
+	if (write_ptr % ETB_FRAME_SIZE_WORDS) {
+		dev_err(drvdata->dev,
+			"write_ptr: %lu not aligned to formatter frame size\n",
+			(unsigned long)write_ptr);
+
+		write_ptr &= ~ETB_FRAME_SIZE_WORDS;
+		local_inc(&buf->lost);
+	}
+
+	/*
+	 * Get a hold of the status register and see if a wrap around
+	 * has occurred.  If so adjust things accordingly.  Otherwise
+	 * start at the beginning and go until the write pointer has
+	 * been reached.
+	 */
+	status = readl_relaxed(drvdata->base + ETB_STATUS_REG);
+	if (status & ETB_STATUS_RAM_FULL) {
+		local_inc(&buf->lost);
+		to_read = drvdata->buffer_depth * ETB_FRAME_SIZE_WORDS;
+		start = write_ptr;
+	} else {
+		to_read = CIRC_CNT(write_ptr, read_ptr,
+				 drvdata->buffer_depth * ETB_FRAME_SIZE_WORDS);
+		start = read_ptr;
+	}
+
+	/*
+	 * Make sure we don't overwrite data that hasn't been consumed yet.
+	 * It is entirely possible that the HW buffer has more data than the
+	 * ring buffer can currently handle.  If so adjust the start address
+	 * to take only the last traces.
+	 *
+	 * In snapshot mode we are looking to get the latest traces only and as
+	 * such, we don't care about not overwriting data that hasn't been
+	 * processed by user space.
+	 *
+	 * Since metrics related to ETBs is in words, multiply by the
+	 * amount of byte per word to have the right units.
+	 */
+	if (!buf->snapshot && to_read > handle->size) {
+		unsigned int capacity = drvdata->buffer_depth *
+							ETB_FRAME_SIZE_WORDS;
+		to_read = handle->size;
+		/* advance the start pointer to get the latest trace data */
+		start += capacity - handle->size;
+		/* make sure we are still aligned */
+		start &= ~ETB_FRAME_SIZE_WORDS;
+		/* wrap around if we've reach the end of the HW buffer */
+		start &= capacity - 1;
+		/* let the decoder know we've skipped ahead */
+		local_inc(&buf->lost);
+	}
+
+	/* finally tell HW where we want to start reading from */
+	writel_relaxed(start, drvdata->base + ETB_RAM_READ_POINTER);
+
+	cur = buf->cur;
+	offset = buf->offset;
+	for (i = 0; i < to_read; i += 4) {
+		buf_ptr = buf->data_pages[cur] + offset;
+		read_data = readl_relaxed(drvdata->base +
+					  ETB_RAM_READ_DATA_REG);
+		*buf_ptr++ = read_data >> 0;
+		*buf_ptr++ = read_data >> 8;
+		*buf_ptr++ = read_data >> 16;
+		*buf_ptr++ = read_data >> 24;
+
+		offset += 4;
+		if (offset >= PAGE_SIZE) {
+			offset = 0;
+			cur++;
+			/* wrap around at the end of the buffer */
+			cur &= buf->nr_pages - 1;
+		}
+	}
+
+	/* reset ETB buffer for next run */
+	writel_relaxed(0x0, drvdata->base + ETB_RAM_READ_POINTER);
+	writel_relaxed(0x0, drvdata->base + ETB_RAM_WRITE_POINTER);
+
+	/*
+	 * In snapshot mode all we have to do is communicate to
+	 * perf_aux_output_end() the address of the current head.  In full
+	 * trace mode the same function expects a size to move rb->aux_head
+	 * forward.
+	 */
+	if (buf->snapshot)
+		local_set(&buf->data_size, (cur * PAGE_SIZE) + offset);
+	else
+		local_add(to_read, &buf->data_size);
+
+	etb_enable_hw(drvdata);
+	CS_LOCK(drvdata->base);
+	spin_unlock_irqrestore(&drvdata->spinlock, flags);
+}
+
 static const struct coresight_ops_sink etb_sink_ops = {
 	.enable		= etb_enable,
 	.disable	= etb_disable,
+	.setup_aux	= etb_setup_aux,
+	.set_buffer	= etb_set_buffer,
+	.reset_buffer	= etb_reset_buffer,
+	.update_buffer	= etb_update_buffer,
 };
 
 static const struct coresight_ops etb_cs_ops = {
diff --git a/include/linux/coresight.h b/include/linux/coresight.h
index 80b3465bd459..810a8b2511c4 100644
--- a/include/linux/coresight.h
+++ b/include/linux/coresight.h
@@ -14,6 +14,7 @@
 #define _LINUX_CORESIGHT_H
 
 #include <linux/device.h>
+#include <linux/perf_event.h>
 #include <linux/sched.h>
 
 /* Peripheral id registers (0xFD0-0xFEC) */
@@ -181,12 +182,27 @@ struct coresight_device {
 /**
  * struct coresight_ops_sink - basic operations for a sink
  * Operations available for sinks
- * @enable:	enables the sink.
- * @disable:	disables the sink.
+ * @enable:		enables the sink.
+ * @disable:		disables the sink.
+ * @setup_aux:		initialises perf's ring buffer for trace collection.
+ * @set_buffer:		initialises buffer mechanic before a trace session.
+ * @reset_buffer:	finalises buffer mechanic after a trace session.
+ * @update_buffer:	update buffer pointers after a trace session.
  */
 struct coresight_ops_sink {
 	int (*enable)(struct coresight_device *csdev, u32 mode);
 	void (*disable)(struct coresight_device *csdev);
+	void *(*setup_aux)(struct coresight_device *csdev, int cpu,
+			   void **pages, int nr_pages, bool overwrite);
+	int (*set_buffer)(struct coresight_device *csdev,
+			  struct perf_output_handle *handle,
+			  void *sink_config);
+	unsigned long (*reset_buffer)(struct coresight_device *csdev,
+				      struct perf_output_handle *handle,
+				      void *sink_config, bool *lost);
+	void (*update_buffer)(struct coresight_device *csdev,
+			      struct perf_output_handle *handle,
+			      void *sink_config);
 };
 
 /**
-- 
1.9.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1275721 — [RESEND PATCH V4 21/26] coresight: introducing a global trace ID function

TraceID values have to be unique for all tracers and
consistent between drivers and user space.  As such
introducing a central function to be used whenever a
traceID value is required.

The patch also account for data traceIDs, which are usually
I(N) + 1.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 drivers/hwtracing/coresight/coresight-etm3x.c |  7 ++-----
 include/linux/coresight-pmu.h                 | 12 ++++++++++++
 2 files changed, 14 insertions(+), 5 deletions(-)

diff --git a/drivers/hwtracing/coresight/coresight-etm3x.c b/drivers/hwtracing/coresight/coresight-etm3x.c
index c07b7d3eee86..ebd569986bd4 100644
--- a/drivers/hwtracing/coresight/coresight-etm3x.c
+++ b/drivers/hwtracing/coresight/coresight-etm3x.c
@@ -27,6 +27,7 @@
 #include <linux/cpu.h>
 #include <linux/of.h>
 #include <linux/coresight.h>
+#include <linux/coresight-pmu.h>
 #include <linux/amba/bus.h>
 #include <linux/seq_file.h>
 #include <linux/uaccess.h>
@@ -784,11 +785,7 @@ static void etm_init_arch_data(void *info)
 
 static void etm_init_trace_id(struct etm_drvdata *drvdata)
 {
-	/*
-	 * A trace ID of value 0 is invalid, so let's start at some
-	 * random value that fits in 7 bits and go from there.
-	 */
-	drvdata->traceid = 0x10 + drvdata->cpu;
+	drvdata->traceid = coresight_get_trace_id(drvdata->cpu);
 }
 
 static int etm_probe(struct amba_device *adev, const struct amba_id *id)
diff --git a/include/linux/coresight-pmu.h b/include/linux/coresight-pmu.h
index 6c5386b23b10..7d410260661b 100644
--- a/include/linux/coresight-pmu.h
+++ b/include/linux/coresight-pmu.h
@@ -19,9 +19,21 @@
 #define _LINUX_CORESIGHT_PMU_H
 
 #define CORESIGHT_ETM_PMU_NAME "cs_etm"
+#define CORESIGHT_ETM_PMU_SEED  0x10
 
 /* ETMv3.5/PTM's ETMCR config bit */
 #define ETM_OPT_CYCACC  12
 #define ETM_OPT_TS      28
 
+static inline int coresight_get_trace_id(int cpu)
+{
+	/*
+	 * A trace ID of value 0 is invalid, so let's start at some
+	 * random value that fits in 7 bits and go from there.  Since
+	 * the common convention is to have data trace IDs be I(N) + 1,
+	 * set instruction trace IDs as a function of the CPU number.
+	 */
+	return (CORESIGHT_ETM_PMU_SEED + (cpu * 2));
+}
+
 #endif
-- 
1.9.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1275722 — [RESEND PATCH V4 10/26] coresight: etm3x: set progbit to stop trace collection

There is no need to use the event enable's "always false" event to
stop trace collection.  For that purpose setting the programming bit
(ETMCR:10) is enough.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 drivers/hwtracing/coresight/coresight-etm3x.c | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/drivers/hwtracing/coresight/coresight-etm3x.c b/drivers/hwtracing/coresight/coresight-etm3x.c
index 5a2be0590754..d1235711b576 100644
--- a/drivers/hwtracing/coresight/coresight-etm3x.c
+++ b/drivers/hwtracing/coresight/coresight-etm3x.c
@@ -422,9 +422,6 @@ static void etm_disable_hw(void *info)
 	CS_UNLOCK(drvdata->base);
 	etm_set_prog(drvdata);
 
-	/* Program trace enable to low by using always false event */
-	etm_writel(drvdata, ETM_HARD_WIRE_RES_A | ETM_EVENT_NOT_A, ETMTEEVR);
-
 	/* Read back sequencer and counters for post trace analysis */
 	config->seq_curr_state = (etm_readl(drvdata, ETMSQR) & ETM_SQR_MASK);
 
-- 
1.9.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1275723 — [RESEND PATCH V4 16/26] perf: changing pmu::setup_aux() parameter to include event

For Coresight ETMv3/4 tracers the event carries information
about trace modes such as user or kernel space tracing and
whether tracing is allowed when operating in secure mode.

Since this information is to be embedded in the private
structure returned by setup_aux(),  changing the first
parameter to be of type struct perf_event * so that all
the necessary information can be conveyed.  Also changing
current customer of the API to reflect the modification.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 arch/x86/kernel/cpu/perf_event_intel_bts.c | 4 +++-
 arch/x86/kernel/cpu/perf_event_intel_pt.c  | 5 +++--
 include/linux/perf_event.h                 | 2 +-
 kernel/events/ring_buffer.c                | 2 +-
 4 files changed, 8 insertions(+), 5 deletions(-)

diff --git a/arch/x86/kernel/cpu/perf_event_intel_bts.c b/arch/x86/kernel/cpu/perf_event_intel_bts.c
index d1c0f254afbe..773600b2f313 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_bts.c
+++ b/arch/x86/kernel/cpu/perf_event_intel_bts.c
@@ -68,10 +68,12 @@ static size_t buf_size(struct page *page)
 }
 
 static void *
-bts_buffer_setup_aux(int cpu, void **pages, int nr_pages, bool overwrite)
+bts_buffer_setup_aux(struct perf_event *event, void **pages,
+		     int nr_pages, bool overwrite)
 {
 	struct bts_buffer *buf;
 	struct page *page;
+	int cpu = event->cpu;
 	int node = (cpu == -1) ? cpu : cpu_to_node(cpu);
 	unsigned long offset;
 	size_t size = nr_pages << PAGE_SHIFT;
diff --git a/arch/x86/kernel/cpu/perf_event_intel_pt.c b/arch/x86/kernel/cpu/perf_event_intel_pt.c
index 42169283448b..f8b881210648 100644
--- a/arch/x86/kernel/cpu/perf_event_intel_pt.c
+++ b/arch/x86/kernel/cpu/perf_event_intel_pt.c
@@ -863,10 +863,11 @@ static int pt_buffer_init_topa(struct pt_buffer *buf, unsigned long nr_pages,
  * Return:	Our private PT buffer structure.
  */
 static void *
-pt_buffer_setup_aux(int cpu, void **pages, int nr_pages, bool snapshot)
+pt_buffer_setup_aux(struct perf_event *event, void **pages,
+		    int nr_pages, bool snapshot)
 {
 	struct pt_buffer *buf;
-	int node, ret;
+	int node, ret, cpu = event->cpu;
 
 	if (!nr_pages)
 		return NULL;
diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
index 092a0e8a479a..0d9964b5ed4d 100644
--- a/include/linux/perf_event.h
+++ b/include/linux/perf_event.h
@@ -292,7 +292,7 @@ struct pmu {
 	/*
 	 * Set up pmu-private data structures for an AUX area
 	 */
-	void *(*setup_aux)		(int cpu, void **pages,
+	void *(*setup_aux)		(struct perf_event *event, void **pages,
 					 int nr_pages, bool overwrite);
 					/* optional */
 
diff --git a/kernel/events/ring_buffer.c b/kernel/events/ring_buffer.c
index 182bc30899d5..87171b700c77 100644
--- a/kernel/events/ring_buffer.c
+++ b/kernel/events/ring_buffer.c
@@ -522,7 +522,7 @@ int rb_alloc_aux(struct ring_buffer *rb, struct perf_event *event,
 			goto out;
 	}
 
-	rb->aux_priv = event->pmu->setup_aux(event->cpu, rb->aux_pages, nr_pages,
+	rb->aux_priv = event->pmu->setup_aux(event, rb->aux_pages, nr_pages,
 					     overwrite);
 	if (!rb->aux_priv)
 		goto out;
-- 
1.9.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1275724 — [RESEND PATCH V4 23/26] perf tools: making function set_max_cpu_num() non static

For memory allocation purposes, code located in other places
then util/cpumap.c may want to know how many CPUs the system has.

This patch is making function set_max_cpu_num() available to
other parts of the perf tool so that global variable
@max_cpu_num gets the right value when referenced by cpu__max_cpu().

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 tools/perf/util/cpumap.c | 2 +-
 tools/perf/util/cpumap.h | 1 +
 2 files changed, 2 insertions(+), 1 deletion(-)

diff --git a/tools/perf/util/cpumap.c b/tools/perf/util/cpumap.c
index 3667e2123e5b..97ef46e4a0f6 100644
--- a/tools/perf/util/cpumap.c
+++ b/tools/perf/util/cpumap.c
@@ -373,7 +373,7 @@ out:
 }
 
 /* Determine highest possible cpu in the system for sparse allocation */
-static void set_max_cpu_num(void)
+void set_max_cpu_num(void)
 {
 	const char *mnt;
 	char path[PATH_MAX];
diff --git a/tools/perf/util/cpumap.h b/tools/perf/util/cpumap.h
index 0af9cecb4c51..6f7dce7dcca6 100644
--- a/tools/perf/util/cpumap.h
+++ b/tools/perf/util/cpumap.h
@@ -14,6 +14,7 @@ struct cpu_map {
 	int map[];
 };
 
+void set_max_cpu_num(void);
 struct cpu_map *cpu_map__new(const char *cpu_list);
 struct cpu_map *cpu_map__dummy_new(void);
 struct cpu_map *cpu_map__read(FILE *file);
-- 
1.9.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1275725 — [RESEND PATCH V4 11/26] coresight: etm3x: changing default trace configuration

Changing default configuration to include the entire address
range rather than just the kernel.  That way traces are more
inclusive and it is easier to narrow down if needed.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 drivers/hwtracing/coresight/coresight-etm.h   |  2 ++
 drivers/hwtracing/coresight/coresight-etm3x.c | 29 ++++++++++++---------------
 2 files changed, 15 insertions(+), 16 deletions(-)

diff --git a/drivers/hwtracing/coresight/coresight-etm.h b/drivers/hwtracing/coresight/coresight-etm.h
index a8663780a908..f01862fc98dd 100644
--- a/drivers/hwtracing/coresight/coresight-etm.h
+++ b/drivers/hwtracing/coresight/coresight-etm.h
@@ -146,6 +146,7 @@
  * @startstop_ctrl: setting for register ETMTSSCR.
  * @enable_event: setting for register ETMTEEVR.
  * @enable_ctrl1: setting for register ETMTECR1.
+ * @enable_ctrl2: setting for register ETMTECR2.
  * @fifofull_level: setting for register ETMFFLR.
  * @addr_idx:	index for the address comparator selection.
  * @addr_val:	value for address comparator register.
@@ -179,6 +180,7 @@ struct etm_config {
 	u32				startstop_ctrl;
 	u32				enable_event;
 	u32				enable_ctrl1;
+	u32				enable_ctrl2;
 	u32				fifofull_level;
 	u8				addr_idx;
 	u32				addr_val[ETM_MAX_ADDR_CMP];
diff --git a/drivers/hwtracing/coresight/coresight-etm3x.c b/drivers/hwtracing/coresight/coresight-etm3x.c
index d1235711b576..4c54e104b610 100644
--- a/drivers/hwtracing/coresight/coresight-etm3x.c
+++ b/drivers/hwtracing/coresight/coresight-etm3x.c
@@ -607,26 +607,23 @@ static void etm_init_arch_data(void *info)
 
 static void etm_init_default_data(struct etm_config *config)
 {
-	u32 flags = (1 << 0 | /* instruction execute*/
-		     3 << 3 | /* ARM instruction */
-		     0 << 5 | /* No data value comparison */
-		     0 << 7 | /* No exact mach */
-		     0 << 8 | /* Ignore context ID */
-		     0 << 10); /* Security ignored */
-
 	if (WARN_ON_ONCE(!config))
 		return;
 
-	config->ctrl = (ETMCR_CYC_ACC | ETMCR_TIMESTAMP_EN);
-	config->enable_ctrl1 = ETMTECR1_ADDR_COMP_1;
-	config->addr_val[0] = (u32) _stext;
-	config->addr_val[1] = (u32) _etext;
-	config->addr_acctype[0] = flags;
-	config->addr_acctype[1] = flags;
-	config->addr_type[0] = ETM_ADDR_TYPE_RANGE;
-	config->addr_type[1] = ETM_ADDR_TYPE_RANGE;
-
 	etm_set_default(config);
+
+	/*
+	 * Taken verbatim from the TRM:
+	 *
+	 * To trace all memory:
+	 *  set bit [24] in register 0x009, the ETMTECR1, to 1
+	 *  set all other bits in register 0x009, the ETMTECR1, to 0
+	 *  set all bits in register 0x007, the ETMTECR2, to 0
+	 *  set register 0x008, the ETMTEEVR, to 0x6F (TRUE).
+	 */
+	config->enable_ctrl1 = BIT(24);
+	config->enable_ctrl2 = 0x0;
+	config->enable_event = ETM_HARD_WIRE_RES_A;
 }
 
 static void etm_init_trace_id(struct etm_drvdata *drvdata)
-- 
1.9.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1275726 — [RESEND PATCH V4 20/26] coresight: etm-perf: new PMU driver for ETM tracers

Perf is a well known and used tool for performance monitoring
and much more. A such it is an ideal candidate for integration
with coresight based HW tracing.

This patch introduces a PMU that represent a coresight tracer to
the Perf core.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 drivers/hwtracing/coresight/Makefile             |   3 +-
 drivers/hwtracing/coresight/coresight-etm-perf.c | 438 +++++++++++++++++++++++
 drivers/hwtracing/coresight/coresight-etm-perf.h |  32 ++
 drivers/hwtracing/coresight/coresight-etm3x.c    |   7 +
 include/linux/coresight-pmu.h                    |  27 ++
 5 files changed, 506 insertions(+), 1 deletion(-)
 create mode 100644 drivers/hwtracing/coresight/coresight-etm-perf.c
 create mode 100644 drivers/hwtracing/coresight/coresight-etm-perf.h
 create mode 100644 include/linux/coresight-pmu.h

diff --git a/drivers/hwtracing/coresight/Makefile b/drivers/hwtracing/coresight/Makefile
index 233d66cf22d3..cf8c6d689747 100644
--- a/drivers/hwtracing/coresight/Makefile
+++ b/drivers/hwtracing/coresight/Makefile
@@ -9,6 +9,7 @@ obj-$(CONFIG_CORESIGHT_SINK_ETBV10) += coresight-etb10.o
 obj-$(CONFIG_CORESIGHT_LINKS_AND_SINKS) += coresight-funnel.o \
 					   coresight-replicator.o
 obj-$(CONFIG_CORESIGHT_SOURCE_ETM3X) += coresight-etm3x.o coresight-etm-cp14.o \
-					coresight-etm3x-sysfs.o
+					coresight-etm3x-sysfs.o \
+					coresight-etm-perf.o
 obj-$(CONFIG_CORESIGHT_SOURCE_ETM4X) += coresight-etm4x.o
 obj-$(CONFIG_CORESIGHT_QCOM_REPLICATOR) += coresight-replicator-qcom.o
diff --git a/drivers/hwtracing/coresight/coresight-etm-perf.c b/drivers/hwtracing/coresight/coresight-etm-perf.c
new file mode 100644
index 000000000000..590d9b132e6c
--- /dev/null
+++ b/drivers/hwtracing/coresight/coresight-etm-perf.c
@@ -0,0 +1,438 @@
+/*
+ * Copyright(C) 2015 Linaro Limited. All rights reserved.
+ * Author: Mathieu Poirier <mathieu.poirier@linaro.org>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that 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.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <linux/coresight.h>
+#include <linux/coresight-pmu.h>
+#include <linux/cpumask.h>
+#include <linux/device.h>
+#include <linux/list.h>
+#include <linux/mm.h>
+#include <linux/module.h>
+#include <linux/perf_event.h>
+#include <linux/slab.h>
+#include <linux/types.h>
+
+#include "coresight-priv.h"
+
+static struct pmu etm_pmu;
+static bool etm_perf_up;
+
+/**
+ * struct etm_event_data - Coresight specifics associated to an event
+ * @mask:		hold the CPU(s) this event was set for.
+ * @source_config:	per CPU tracer configuration associated to a
+ *			trace session.
+ * @sink_config:	per CPU AUX configuration associated to a
+ *			trace session.
+ * @path:		the path, from source (first element) to
+ *			sink (last element).
+ */
+struct etm_event_data {
+	cpumask_t mask;
+	void **source_config;
+	void **sink_config;
+	struct list_head **path;
+};
+
+static DEFINE_PER_CPU(struct perf_output_handle, ctx_handle);
+static DEFINE_PER_CPU(struct coresight_device *, csdev_src);
+
+/* ETMv3.5/PTM's ETMCR is 'config' */
+PMU_FORMAT_ATTR(cycacc,		"config:" __stringify(ETM_OPT_CYCACC));
+PMU_FORMAT_ATTR(timestamp,	"config:" __stringify(ETM_OPT_TS));
+
+static struct attribute *etm_config_formats_attr[] = {
+	&format_attr_cycacc.attr,
+	&format_attr_timestamp.attr,
+	NULL,
+};
+
+static struct attribute_group etm_pmu_format_group = {
+	.name   = "format",
+	.attrs  = etm_config_formats_attr,
+};
+
+static const struct attribute_group *etm_pmu_attr_groups[] = {
+	&etm_pmu_format_group,
+	NULL,
+};
+
+static void etm_event_read(struct perf_event *event) {}
+static void etm_event_destroy(struct perf_event *event) {}
+
+static int etm_event_init(struct perf_event *event)
+{
+	if (event->attr.type != etm_pmu.type)
+		return -ENOENT;
+
+	if (event->cpu >= nr_cpu_ids)
+		return -EINVAL;
+
+	event->destroy = etm_event_destroy;
+
+	return 0;
+}
+
+static void *alloc_event_data(int cpu)
+{
+	int size;
+	struct etm_event_data *event_data;
+	void *source_config, *sink_config;
+	struct list_head *path;
+
+	event_data = kzalloc(sizeof(struct etm_event_data), GFP_KERNEL);
+	 if (!event_data)
+		return NULL;
+
+	if (cpu != -1)
+		size = 1;
+	else
+		size = num_online_cpus();
+
+	source_config = kcalloc(size, sizeof(void *), GFP_KERNEL);
+	if (!source_config)
+		goto source_config_err;
+
+	sink_config = kcalloc(size, sizeof(void *), GFP_KERNEL);
+	if (!sink_config)
+		goto sink_config_err;
+
+	path = kcalloc(size, sizeof(struct list_head *), GFP_KERNEL);
+	if (!path)
+		goto path_err;
+
+	cpumask_clear(&event_data->mask);
+	event_data->source_config = source_config;
+	event_data->sink_config = sink_config;
+	event_data->path = (struct list_head **)path;
+
+out:
+	return event_data;
+
+path_err:
+	kfree(sink_config);
+sink_config_err:
+	kfree(source_config);
+source_config_err:
+	kfree(event_data);
+	event_data = NULL;
+	goto out;
+}
+
+static void free_event_data(struct etm_event_data *event_data)
+{
+	int cpu;
+	cpumask_t *mask = &event_data->mask;
+
+	for_each_cpu(cpu, mask) {
+		kfree(event_data->source_config[cpu]);
+		kfree(event_data->sink_config[cpu]);
+		kfree(event_data->path[cpu]);
+	}
+
+	kfree(event_data->source_config);
+	kfree(event_data->sink_config);
+	kfree(event_data->path);
+	kfree(event_data);
+}
+
+static void *etm_setup_aux(struct perf_event *event, void **pages,
+			   int nr_pages, bool overwrite)
+{
+	int cpu;
+	cpumask_t *mask, initialised;
+	struct etm_event_data *event_data = NULL;
+	struct coresight_device *csdev;
+
+	event_data = alloc_event_data(event->cpu);
+	if (!event_data)
+		return NULL;
+
+	mask = &event_data->mask;
+
+	if (event->cpu != -1)
+		cpumask_set_cpu(event->cpu, mask);
+	else
+		cpumask_copy(mask, cpu_online_mask);
+
+	cpumask_clear(&initialised);
+	for_each_cpu(cpu, mask) {
+		struct coresight_device *sink;
+
+		csdev = per_cpu(csdev_src, cpu);
+		if (!csdev)
+			goto err;
+
+		/* Get the tracer's config from perf */
+		if (!source_ops(csdev)->get_config)
+			goto err;
+
+		event_data->source_config[cpu] =
+			source_ops(csdev)->get_config(csdev, event);
+
+		if (!event_data->source_config[cpu])
+			goto err;
+
+		/*
+		 * Building a path doesn't enable it, it simply builds a
+		 * list of devices from source to sink that can be
+		 * referenced later when the path is actually needed.
+		 */
+		event_data->path[cpu] = coresight_build_path(csdev);
+		if (!event_data->path[cpu])
+			goto err;
+
+		/*
+		 * Remember which paths have been initialised so that they
+		 * can be undone in case of errors.
+		 */
+		cpumask_set_cpu(cpu, &initialised);
+
+		/* Grab the sink at the end of the path */
+		sink = coresight_get_sink(event_data->path[cpu]);
+		if (!sink)
+			goto err;
+
+		if (!sink_ops(sink)->setup_aux)
+			goto err;
+
+		/* Finally get the AUX specific data from the sink buffer */
+		event_data->sink_config[cpu] =
+				sink_ops(sink)->setup_aux(sink, cpu, pages,
+							  nr_pages, overwrite);
+		if (!event_data->sink_config[cpu])
+			goto err;
+	}
+
+out:
+	return event_data;
+
+err:
+	for_each_cpu(cpu, &initialised)
+		coresight_release_path(event_data->path[cpu]);
+
+	free_event_data(event_data);
+	event_data = NULL;
+	goto out;
+}
+
+static void etm_free_aux(void *data)
+{
+	free_event_data(data);
+}
+
+static void etm_event_stop(struct perf_event *event, int mode)
+{
+	int cpu = smp_processor_id();
+	struct coresight_device *csdev = per_cpu(csdev_src, cpu);
+
+	if (event->hw.state == PERF_HES_STOPPED)
+		return;
+
+	if (!csdev)
+		return;
+
+	/* stop tracer */
+	source_ops(csdev)->disable(csdev);
+
+	/* tell the core */
+	event->hw.state = PERF_HES_STOPPED;
+
+
+	if (mode & PERF_EF_UPDATE) {
+		struct coresight_device *sink;
+		struct perf_output_handle *handle = this_cpu_ptr(&ctx_handle);
+		struct etm_event_data *event_data = perf_get_aux(handle);
+
+		if (WARN_ON_ONCE(handle->event != event))
+			return;
+
+		if (WARN_ON_ONCE(!event_data))
+			return;
+
+		sink = coresight_get_sink(event_data->path[cpu]);
+		if (WARN_ON_ONCE(!sink))
+			return;
+
+		/* update trace information */
+		if (!sink_ops(sink)->update_buffer)
+			return;
+
+		sink_ops(sink)->update_buffer(sink, handle,
+					      event_data->sink_config[cpu]);
+	}
+}
+
+static void etm_event_start(struct perf_event *event, int flags)
+{
+	int cpu = smp_processor_id();
+	struct coresight_device *csdev = per_cpu(csdev_src, cpu);
+
+	if (!csdev)
+		goto fail;
+
+	/* tell the perf core the event is alive */
+	event->hw.state = 0;
+
+	if (source_ops(csdev)->enable(csdev, CS_MODE_PERF))
+		goto fail;
+
+	return;
+
+fail:
+	event->hw.state = PERF_HES_STOPPED;
+}
+
+static void etm_event_del(struct perf_event *event, int mode)
+{
+	bool lost = true;
+	int cpu = smp_processor_id();
+	unsigned long size;
+	struct coresight_device *sink;
+	struct perf_output_handle *handle = this_cpu_ptr(&ctx_handle);
+	struct etm_event_data *event_data = perf_get_aux(handle);
+
+	if (WARN_ON_ONCE(!event_data))
+		return;
+
+	etm_event_stop(event, PERF_EF_UPDATE);
+
+	sink = coresight_get_sink(event_data->path[cpu]);
+	if (!sink)
+		return;
+
+	if (!sink_ops(sink)->reset_buffer)
+		return;
+
+	size = sink_ops(sink)->reset_buffer(sink, handle,
+					    event_data->sink_config[cpu],
+					    &lost);
+
+	perf_aux_output_end(handle, size, lost);
+
+	coresight_disable_path(event_data->path[cpu]);
+}
+
+static int etm_event_add(struct perf_event *event, int mode)
+{
+
+	int ret = -EBUSY, cpu = smp_processor_id();
+	struct etm_event_data *event_data;
+	struct perf_output_handle *handle = this_cpu_ptr(&ctx_handle);
+	struct hw_perf_event *hwc = &event->hw;
+	struct coresight_device *csdev = per_cpu(csdev_src, cpu);
+	struct coresight_device *sink;
+
+	if (handle->event)
+		goto out;
+
+	event_data = perf_aux_output_begin(handle, event);
+	ret = -EINVAL;
+	if (WARN_ON_ONCE(!event_data))
+		goto fail_stop;
+
+	sink = coresight_get_sink(event_data->path[cpu]);
+	if (!sink)
+		goto fail_end_stop;
+
+	if (!sink_ops(sink)->set_buffer)
+		goto fail_end_stop;
+
+	ret = sink_ops(sink)->set_buffer(sink, handle,
+					 event_data->sink_config[cpu]);
+
+	if (ret)
+		goto fail_end_stop;
+
+	if (!source_ops(csdev)->set_config) {
+		ret = -EINVAL;
+		goto fail_end_stop;
+	}
+
+	source_ops(csdev)->set_config(csdev, event_data->source_config[cpu]);
+
+	ret = coresight_enable_path(event_data->path[cpu], CS_MODE_PERF);
+	if (ret)
+		goto fail_end_stop;
+
+	if (mode & PERF_EF_START) {
+		etm_event_start(event, 0);
+		if (hwc->state & PERF_HES_STOPPED) {
+			etm_event_del(event, 0);
+			return -EBUSY;
+		}
+	}
+
+out:
+	return ret;
+
+fail_end_stop:
+	perf_aux_output_end(handle, 0, true);
+fail_stop:
+	hwc->state = PERF_HES_STOPPED;
+	goto out;
+}
+
+int etm_perf_symlink(struct coresight_device *csdev, bool link)
+{
+	char entry[sizeof("cpu9999999")];
+	int ret = 0, cpu = source_ops(csdev)->cpu_id(csdev);
+	struct device *pmu_dev = etm_pmu.dev;
+	struct device *cs_dev = &csdev->dev;
+
+	sprintf(entry, "cpu%d", cpu);
+
+	if (!etm_perf_up)
+		return -EPROBE_DEFER;
+
+	if (link) {
+		ret = sysfs_create_link(&pmu_dev->kobj, &cs_dev->kobj, entry);
+		if (ret)
+			return ret;
+		per_cpu(csdev_src, cpu) = csdev;
+	} else {
+		sysfs_remove_link(&pmu_dev->kobj, entry);
+		per_cpu(csdev_src, cpu) = NULL;
+	}
+
+	return 0;
+}
+
+static int __init etm_perf_init(void)
+{
+	int ret;
+
+	etm_pmu.capabilities	= PERF_PMU_CAP_EXCLUSIVE;
+
+	etm_pmu.attr_groups	= etm_pmu_attr_groups;
+	etm_pmu.task_ctx_nr	= perf_sw_context;
+	etm_pmu.read		= etm_event_read;
+	etm_pmu.event_init	= etm_event_init;
+	etm_pmu.setup_aux	= etm_setup_aux;
+	etm_pmu.free_aux	= etm_free_aux;
+	etm_pmu.stop		= etm_event_stop;
+	etm_pmu.start		= etm_event_start;
+	etm_pmu.del		= etm_event_del;
+	etm_pmu.add		= etm_event_add;
+
+	ret = perf_pmu_register(&etm_pmu, CORESIGHT_ETM_PMU_NAME, -1);
+	if (ret == 0)
+		etm_perf_up = true;
+
+	return ret;
+}
+module_init(etm_perf_init);
diff --git a/drivers/hwtracing/coresight/coresight-etm-perf.h b/drivers/hwtracing/coresight/coresight-etm-perf.h
new file mode 100644
index 000000000000..87f5a134eb6f
--- /dev/null
+++ b/drivers/hwtracing/coresight/coresight-etm-perf.h
@@ -0,0 +1,32 @@
+/*
+ * Copyright(C) 2015 Linaro Limited. All rights reserved.
+ * Author: Mathieu Poirier <mathieu.poirier@linaro.org>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that 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.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef _CORESIGHT_ETM_PERF_H
+#define _CORESIGHT_ETM_PERF_H
+
+struct coresight_device;
+
+#ifdef CONFIG_CORESIGHT
+int etm_perf_symlink(struct coresight_device *csdev, bool link);
+
+#else
+static inline int etm_perf_symlink(struct coresight_device *csdev, bool link)
+{ return -EINVAL; }
+
+#endif /* CONFIG_CORESIGHT */
+
+#endif
diff --git a/drivers/hwtracing/coresight/coresight-etm3x.c b/drivers/hwtracing/coresight/coresight-etm3x.c
index 4ff3e1419fde..c07b7d3eee86 100644
--- a/drivers/hwtracing/coresight/coresight-etm3x.c
+++ b/drivers/hwtracing/coresight/coresight-etm3x.c
@@ -35,6 +35,7 @@
 #include <asm/sections.h>
 
 #include "coresight-etm.h"
+#include "coresight-etm-perf.h"
 
 static int boot_enable;
 module_param_named(boot_enable, boot_enable, int, S_IRUGO);
@@ -870,6 +871,12 @@ static int etm_probe(struct amba_device *adev, const struct amba_id *id)
 		goto err_arch_supported;
 	}
 
+	ret = etm_perf_symlink(drvdata->csdev, true);
+	if (ret) {
+		coresight_unregister(drvdata->csdev);
+		goto err_arch_supported;
+	}
+
 	pm_runtime_put(&adev->dev);
 	dev_info(dev, "%s initialized\n", (char *)id->data);
 
diff --git a/include/linux/coresight-pmu.h b/include/linux/coresight-pmu.h
new file mode 100644
index 000000000000..6c5386b23b10
--- /dev/null
+++ b/include/linux/coresight-pmu.h
@@ -0,0 +1,27 @@
+/*
+ * Copyright(C) 2015 Linaro Limited. All rights reserved.
+ * Author: Mathieu Poirier <mathieu.poirier@linaro.org>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that 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.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef _LINUX_CORESIGHT_PMU_H
+#define _LINUX_CORESIGHT_PMU_H
+
+#define CORESIGHT_ETM_PMU_NAME "cs_etm"
+
+/* ETMv3.5/PTM's ETMCR config bit */
+#define ETM_OPT_CYCACC  12
+#define ETM_OPT_TS      28
+
+#endif
-- 
1.9.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1275729 — [RESEND PATCH V4 26/26] perf tools: adding coresight etm PMU record capabilities

Coresight ETMs are IP blocks used to perform HW assisted tracing
on a CPU core.  This patch introduce the required auxiliary API
functions allowing the perf core to interact with a tracer.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 MAINTAINERS                         |   3 +
 tools/perf/arch/arm/util/Build      |   2 +-
 tools/perf/arch/arm/util/auxtrace.c |  48 ++++
 tools/perf/arch/arm/util/cs_etm.c   | 466 ++++++++++++++++++++++++++++++++++++
 tools/perf/arch/arm/util/cs_etm.h   |  44 ++++
 tools/perf/util/auxtrace.c          |   1 +
 tools/perf/util/auxtrace.h          |   1 +
 7 files changed, 564 insertions(+), 1 deletion(-)
 create mode 100644 tools/perf/arch/arm/util/auxtrace.c
 create mode 100644 tools/perf/arch/arm/util/cs_etm.c
 create mode 100644 tools/perf/arch/arm/util/cs_etm.h

diff --git a/MAINTAINERS b/MAINTAINERS
index d73f35aeaf3b..7b7646931886 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -1000,6 +1000,9 @@ F:	Documentation/trace/coresight.txt
 F:	Documentation/devicetree/bindings/arm/coresight.txt
 F:	Documentation/ABI/testing/sysfs-bus-coresight-devices-*
 F:	tools/perf/arch/arm/util/pmu.c
+F:	tools/perf/arch/arm/util/auxtrace.c
+F:	tools/perf/arch/arm/util/cs_etm.c
+F:	tools/perf/arch/arm/util/cs_etm.h
 
 ARM/CORGI MACHINE SUPPORT
 M:	Richard Purdie <rpurdie@rpsys.net>
diff --git a/tools/perf/arch/arm/util/Build b/tools/perf/arch/arm/util/Build
index 66ab0b05549c..0a25a1248f42 100644
--- a/tools/perf/arch/arm/util/Build
+++ b/tools/perf/arch/arm/util/Build
@@ -3,4 +3,4 @@ libperf-$(CONFIG_DWARF) += dwarf-regs.o
 libperf-$(CONFIG_LIBUNWIND)          += unwind-libunwind.o
 libperf-$(CONFIG_LIBDW_DWARF_UNWIND) += unwind-libdw.o
 
-libperf-$(CONFIG_AUXTRACE) += pmu.o
+libperf-$(CONFIG_AUXTRACE) += pmu.o auxtrace.o cs_etm.o
diff --git a/tools/perf/arch/arm/util/auxtrace.c b/tools/perf/arch/arm/util/auxtrace.c
new file mode 100644
index 000000000000..4988fdf7cb8a
--- /dev/null
+++ b/tools/perf/arch/arm/util/auxtrace.c
@@ -0,0 +1,48 @@
+/*
+ * Copyright(C) 2015 Linaro Limited. All rights reserved.
+ * Author: Mathieu Poirier <mathieu.poirier@linaro.org>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that 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.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <stdbool.h>
+#include <linux/coresight-pmu.h>
+
+#include "../../util/auxtrace.h"
+#include "../../util/evlist.h"
+#include "../../util/pmu.h"
+#include "cs_etm.h"
+
+struct auxtrace_record
+*auxtrace_record__init(struct perf_evlist *evlist, int *err)
+{
+	struct perf_pmu	*cs_etm_pmu;
+	struct perf_evsel *evsel;
+	bool found_etm = false;
+
+	cs_etm_pmu = perf_pmu__find(CORESIGHT_ETM_PMU_NAME);
+
+	 if (evlist) {
+		evlist__for_each(evlist, evsel) {
+			if (cs_etm_pmu &&
+			    evsel->attr.type == cs_etm_pmu->type)
+				found_etm = true;
+		}
+	}
+
+	if (found_etm)
+		return cs_etm_record_init(err);
+
+	*err = -EINVAL;
+	return NULL;
+}
diff --git a/tools/perf/arch/arm/util/cs_etm.c b/tools/perf/arch/arm/util/cs_etm.c
new file mode 100644
index 000000000000..ea68a99310fe
--- /dev/null
+++ b/tools/perf/arch/arm/util/cs_etm.c
@@ -0,0 +1,466 @@
+/*
+ * Copyright(C) 2015 Linaro Limited. All rights reserved.
+ * Author: Mathieu Poirier <mathieu.poirier@linaro.org>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that 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.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <api/fs/fs.h>
+#include <linux/bitops.h>
+#include <linux/coresight-pmu.h>
+#include <linux/kernel.h>
+#include <linux/log2.h>
+#include <linux/types.h>
+
+#include "../../perf.h"
+#include "../../util/auxtrace.h"
+#include "../../util/cpumap.h"
+#include "../../util/evlist.h"
+#include "../../util/pmu.h"
+#include "../../util/thread_map.h"
+#include "cs_etm.h"
+
+#include <stdlib.h>
+
+#define KiB(x) ((x) * 1024)
+#define MiB(x) ((x) * 1024 * 1024)
+
+struct cs_etm_recording {
+	struct auxtrace_record	itr;
+	struct perf_pmu		*cs_etm_pmu;
+	struct perf_evlist	*evlist;
+	bool			snapshot_mode;
+	size_t			snapshot_size;
+};
+
+static int cs_etm_parse_snapshot_options(struct auxtrace_record *itr,
+					 struct record_opts *opts,
+					 const char *str)
+{
+	struct cs_etm_recording *ptr =
+				container_of(itr, struct cs_etm_recording, itr);
+	unsigned long long snapshot_size = 0;
+	char *endptr;
+
+	if (str) {
+		snapshot_size = strtoull(str, &endptr, 0);
+		if (*endptr || snapshot_size > SIZE_MAX)
+			return -1;
+	}
+
+	opts->auxtrace_snapshot_mode = true;
+	opts->auxtrace_snapshot_size = snapshot_size;
+	ptr->snapshot_size = snapshot_size;
+
+	return 0;
+}
+
+static int cs_etm_recording_options(struct auxtrace_record *itr,
+				    struct perf_evlist *evlist,
+				    struct record_opts *opts)
+{
+	struct cs_etm_recording *ptr =
+				container_of(itr, struct cs_etm_recording, itr);
+	struct perf_pmu *cs_etm_pmu = ptr->cs_etm_pmu;
+	struct perf_evsel *evsel, *cs_etm_evsel = NULL;
+	const struct cpu_map *cpus = evlist->cpus;
+	bool privileged = (geteuid() == 0 || perf_event_paranoid() < 0);
+
+	ptr->evlist = evlist;
+	ptr->snapshot_mode = opts->auxtrace_snapshot_mode;
+
+	evlist__for_each(evlist, evsel) {
+		if (evsel->attr.type == cs_etm_pmu->type) {
+			if (cs_etm_evsel) {
+				pr_err("There may be only one %s event\n",
+				       CORESIGHT_ETM_PMU_NAME);
+				return -EINVAL;
+			}
+			evsel->attr.freq = 0;
+			evsel->attr.sample_period = 1;
+			cs_etm_evsel = evsel;
+			opts->full_auxtrace = true;
+		}
+	}
+
+	/* no need to continue if at least one event of interest was found */
+	if (!cs_etm_evsel)
+		return 0;
+
+	if (opts->use_clockid) {
+		pr_err("Cannot use clockid (-k option) with %s\n",
+		       CORESIGHT_ETM_PMU_NAME);
+		return -EINVAL;
+	}
+
+	/* we are in snapshot mode */
+	if (opts->auxtrace_snapshot_mode) {
+		/*
+		 * No size were given to '-S' or '-m,', so go with
+		 * the default
+		 */
+		if (!opts->auxtrace_snapshot_size &&
+		    !opts->auxtrace_mmap_pages) {
+			if (privileged) {
+				opts->auxtrace_mmap_pages = MiB(4) / page_size;
+			} else {
+				opts->auxtrace_mmap_pages =
+							KiB(128) / page_size;
+				if (opts->mmap_pages == UINT_MAX)
+					opts->mmap_pages = KiB(256) / page_size;
+			}
+		} else if (!opts->auxtrace_mmap_pages && !privileged &&
+						opts->mmap_pages == UINT_MAX) {
+			opts->mmap_pages = KiB(256) / page_size;
+		}
+
+		/*
+		 * '-m,xyz' was specified but no snapshot size, so make the
+		 * snapshot size as big as the auxtrace mmap area.
+		 */
+		if (!opts->auxtrace_snapshot_size) {
+			opts->auxtrace_snapshot_size =
+				opts->auxtrace_mmap_pages * (size_t)page_size;
+		}
+
+		/*
+		 * -Sxyz was specified but no auxtrace mmap area, so make the
+		 * auxtrace mmap area big enough to fit the requested snapshot
+		 * size.
+		 */
+		if (!opts->auxtrace_mmap_pages) {
+			size_t sz = opts->auxtrace_snapshot_size;
+
+			sz = round_up(sz, page_size) / page_size;
+			opts->auxtrace_mmap_pages = roundup_pow_of_two(sz);
+		}
+
+		/* Snapshost size can't be bigger than the auxtrace area */
+		if (opts->auxtrace_snapshot_size >
+				opts->auxtrace_mmap_pages * (size_t)page_size) {
+			pr_err("Snapshot size %zu must not be greater than AUX area tracing mmap size %zu\n",
+			       opts->auxtrace_snapshot_size,
+			       opts->auxtrace_mmap_pages * (size_t)page_size);
+			return -EINVAL;
+		}
+
+		/* Something went wrong somewhere - this shouldn't happen */
+		if (!opts->auxtrace_snapshot_size ||
+		    !opts->auxtrace_mmap_pages) {
+			pr_err("Failed to calculate default snapshot size and/or AUX area tracing mmap pages\n");
+			return -EINVAL;
+		}
+	}
+
+	/* We are in full trace mode but '-m,xyz' wasn't specified */
+	 if (opts->full_auxtrace && !opts->auxtrace_mmap_pages) {
+		if (privileged) {
+			opts->auxtrace_mmap_pages = MiB(4) / page_size;
+		} else {
+			opts->auxtrace_mmap_pages = KiB(128) / page_size;
+			if (opts->mmap_pages == UINT_MAX)
+				opts->mmap_pages = KiB(256) / page_size;
+		}
+
+	}
+
+	/* Validate auxtrace_mmap_pages provided by user */
+	if (opts->auxtrace_mmap_pages) {
+		unsigned int max_page = (KiB(128) / page_size);
+		size_t sz = opts->auxtrace_mmap_pages * (size_t)page_size;
+
+		if (!privileged &&
+		    opts->auxtrace_mmap_pages > max_page) {
+			opts->auxtrace_mmap_pages = max_page;
+			pr_err("auxtrace too big, truncating to %d\n",
+			       max_page);
+		}
+
+		if (!is_power_of_2(sz)) {
+			pr_err("Invalid mmap size for %s: must be a power of 2\n",
+			       CORESIGHT_ETM_PMU_NAME);
+			return -EINVAL;
+		}
+	}
+
+	if (opts->auxtrace_snapshot_mode)
+		pr_debug2("%s snapshot size: %zu\n", CORESIGHT_ETM_PMU_NAME,
+			  opts->auxtrace_snapshot_size);
+
+	if (cs_etm_evsel) {
+		/*
+		 * To obtain the auxtrace buffer file descriptor, the auxtrace
+		 * event must come first.
+		 */
+		perf_evlist__to_front(evlist, cs_etm_evsel);
+		/*
+		 * In the case of per-cpu mmaps, we need the CPU on the
+		 * AUX event.
+		 */
+		if (!cpu_map__empty(cpus))
+			perf_evsel__set_sample_bit(cs_etm_evsel, CPU);
+	}
+
+	/* Add dummy event to keep tracking */
+	if (opts->full_auxtrace) {
+		struct perf_evsel *tracking_evsel;
+		int err;
+
+		err = parse_events(evlist, "dummy:u", NULL);
+		if (err)
+			return err;
+
+		tracking_evsel = perf_evlist__last(evlist);
+		perf_evlist__set_tracking_event(evlist, tracking_evsel);
+
+		tracking_evsel->attr.freq = 0;
+		tracking_evsel->attr.sample_period = 1;
+
+		/* In per-cpu case, always need the time of mmap events etc */
+		if (!cpu_map__empty(cpus))
+			perf_evsel__set_sample_bit(tracking_evsel, TIME);
+	}
+
+	return 0;
+}
+
+static u64 cs_etm_get_config(struct auxtrace_record *itr)
+{
+	u64 config = 0;
+	struct cs_etm_recording *ptr =
+			container_of(itr, struct cs_etm_recording, itr);
+	struct perf_pmu *cs_etm_pmu = ptr->cs_etm_pmu;
+	struct perf_evlist *evlist = ptr->evlist;
+	struct perf_evsel *evsel;
+
+	evlist__for_each(evlist, evsel) {
+		if (evsel->attr.type == cs_etm_pmu->type) {
+			/*
+			 * Variable perf_event_attr::config is assigned to
+			 * ETMv3/PTM.  The bit fields have been made to match
+			 * the ETMv3.5 ETRMCR register specification.  See the
+			 * PMU_FORMAT_ATTR() declarations in
+			 * drivers/hwtracing/coresight/coresight-perf.c for
+			 * details.
+			 */
+			config = evsel->attr.config;
+			break;
+		}
+	}
+
+	return config;
+}
+
+static size_t
+cs_etm_info_priv_size(struct auxtrace_record *itr __maybe_unused,
+		      struct perf_session *session __maybe_unused)
+{
+	int records;
+	const struct cpu_map *cpus = session->evlist->cpus;
+
+	if (!cpu_map__empty(cpus)) {
+		records = cpu_map__nr(cpus);
+		goto out;
+	}
+
+	set_max_cpu_num();
+	records = cpu__max_cpu();
+out:
+	return records * CS_ETM_PRIV_SIZE;
+}
+
+static const char *metadata_etmv3_ro[CS_ETM_PRIV_MAX] = {
+	[CS_ETM_ETMCCER]	= "mgmt/etmccer",
+	[CS_ETM_ETMIDR]		= "mgmt/etmidr",
+};
+
+static int cs_etm_get_metadata(int cpu, int index,
+			       struct auxtrace_record *itr,
+			       struct auxtrace_info_event *info)
+{
+	char path[PATH_MAX];
+	int offset = 0, ret = 0;
+	int i, scan;
+	unsigned int val;
+	struct cs_etm_recording *ptr =
+			container_of(itr, struct cs_etm_recording, itr);
+	struct perf_pmu *cs_etm_pmu = ptr->cs_etm_pmu;
+
+	offset = index * CS_ETM_PRIV_MAX;
+
+	/* Build generic header portion */
+	info->priv[offset + CS_ETM_MAGIC] = __perf_cs_etm_magic;
+	info->priv[offset + CS_ETM_CPU] = cpu;
+	info->priv[offset + CS_ETM_SNAPSHOT] = ptr->snapshot_mode;
+
+	/* Get user configurables from the session */
+	info->priv[offset + CS_ETM_ETMCR] = cs_etm_get_config(itr);
+	info->priv[offset + CS_ETM_ETMTRACEIDR] = coresight_get_trace_id(cpu);
+
+	/* Get RO metadata from sysfs */
+	for (i = CS_ETM_ETMCCER; i < CS_ETM_PRIV_MAX; i++) {
+		snprintf(path, PATH_MAX, "cpu%d/%s", cpu, metadata_etmv3_ro[i]);
+
+		scan = perf_pmu__scan_file(cs_etm_pmu, path, "%x", &val);
+		if (scan != 1) {
+			ret = -EINVAL;
+			break;
+		}
+
+		info->priv[offset + i] = val;
+	}
+
+	return ret;
+}
+
+static int cs_etm_info_fill(struct auxtrace_record *itr,
+			    struct perf_session *session,
+			    struct auxtrace_info_event *auxtrace_info,
+			    size_t priv_size)
+{
+	int i, nr_cpu, ret = 0;
+	const struct cpu_map *cpus = session->evlist->cpus;
+
+	if (priv_size != cs_etm_info_priv_size(itr, session))
+		return -EINVAL;
+
+	if (!session->evlist->nr_mmaps)
+		return -EINVAL;
+
+	auxtrace_info->type = PERF_AUXTRACE_CS_ETM;
+
+	/* cpu map is not empty, we have specific CPUs to work with */
+	if (!cpu_map__empty(cpus)) {
+		for (i = 0; i < cpu_map__nr(cpus); i++) {
+			ret = cs_etm_get_metadata(cpus->map[i], i,
+						  itr, auxtrace_info);
+			if (ret)
+				goto out;
+		}
+	} else {
+		/* get configuration for all CPUs in the system */
+		nr_cpu = cpu__max_cpu();
+		for (i = 0; i < nr_cpu; i++) {
+			ret = cs_etm_get_metadata(i, i, itr, auxtrace_info);
+			if (ret)
+				goto out;
+		}
+	}
+
+out:
+	return ret;
+}
+
+static int cs_etm_find_snapshot(struct auxtrace_record *itr __maybe_unused,
+				int idx, struct auxtrace_mmap *mm,
+				unsigned char *data __maybe_unused,
+				u64 *head, u64 *old)
+{
+	pr_debug3("%s: mmap index %d old head %zu new head %zu size %zu\n",
+		  __func__, idx, (size_t)*old, (size_t)*head, mm->len);
+
+	*old = *head;
+	*head += mm->len;
+
+	return 0;
+}
+
+static int cs_etm_snapshot_start(struct auxtrace_record *itr)
+{
+	struct cs_etm_recording *ptr =
+			container_of(itr, struct cs_etm_recording, itr);
+	struct perf_evsel *evsel;
+
+	evlist__for_each(ptr->evlist, evsel) {
+		if (evsel->attr.type == ptr->cs_etm_pmu->type)
+			return perf_evlist__disable_event(ptr->evlist, evsel);
+	}
+	return -EINVAL;
+}
+
+static int cs_etm_snapshot_finish(struct auxtrace_record *itr)
+{
+	struct cs_etm_recording *ptr =
+			container_of(itr, struct cs_etm_recording, itr);
+	struct perf_evsel *evsel;
+
+	evlist__for_each(ptr->evlist, evsel) {
+		if (evsel->attr.type == ptr->cs_etm_pmu->type)
+			return perf_evlist__enable_event(ptr->evlist, evsel);
+	}
+	return -EINVAL;
+}
+
+static u64 cs_etm_reference(struct auxtrace_record *itr __maybe_unused)
+{
+	return (((u64) rand() <<  0) & 0x00000000FFFFFFFFull) |
+		(((u64) rand() << 32) & 0xFFFFFFFF00000000ull);
+}
+
+static void cs_etm_recording_free(struct auxtrace_record *itr)
+{
+	struct cs_etm_recording *ptr =
+			container_of(itr, struct cs_etm_recording, itr);
+	free(ptr);
+}
+
+static int cs_etm_read_finish(struct auxtrace_record *itr, int idx)
+{
+	struct cs_etm_recording *ptr =
+			container_of(itr, struct cs_etm_recording, itr);
+	struct perf_evsel *evsel;
+
+	evlist__for_each(ptr->evlist, evsel) {
+		if (evsel->attr.type == ptr->cs_etm_pmu->type)
+			return perf_evlist__enable_event_idx(ptr->evlist,
+							     evsel, idx);
+	}
+
+	return -EINVAL;
+}
+
+struct auxtrace_record *cs_etm_record_init(int *err)
+{
+	struct perf_pmu *cs_etm_pmu;
+	struct cs_etm_recording *ptr;
+
+	cs_etm_pmu = perf_pmu__find(CORESIGHT_ETM_PMU_NAME);
+
+	if (!cs_etm_pmu) {
+		*err = -EINVAL;
+		goto out;
+	}
+
+	ptr = zalloc(sizeof(struct cs_etm_recording));
+	if (!ptr) {
+		*err = -ENOMEM;
+		goto out;
+	}
+
+	ptr->cs_etm_pmu			= cs_etm_pmu;
+	ptr->itr.parse_snapshot_options	= cs_etm_parse_snapshot_options;
+	ptr->itr.recording_options	= cs_etm_recording_options;
+	ptr->itr.info_priv_size		= cs_etm_info_priv_size;
+	ptr->itr.info_fill		= cs_etm_info_fill;
+	ptr->itr.find_snapshot		= cs_etm_find_snapshot;
+	ptr->itr.snapshot_start		= cs_etm_snapshot_start;
+	ptr->itr.snapshot_finish	= cs_etm_snapshot_finish;
+	ptr->itr.reference		= cs_etm_reference;
+	ptr->itr.free			= cs_etm_recording_free;
+	ptr->itr.read_finish		= cs_etm_read_finish;
+
+	*err = 0;
+	return &ptr->itr;
+out:
+	return NULL;
+}
diff --git a/tools/perf/arch/arm/util/cs_etm.h b/tools/perf/arch/arm/util/cs_etm.h
new file mode 100644
index 000000000000..7e85c1b43598
--- /dev/null
+++ b/tools/perf/arch/arm/util/cs_etm.h
@@ -0,0 +1,44 @@
+/*
+ * Copyright(C) 2015 Linaro Limited. All rights reserved.
+ * Author: Mathieu Poirier <mathieu.poirier@linaro.org>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that 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.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#ifndef INCLUDE__PERF_CS_ETM_H__
+#define INCLUDE__PERF_CS_ETM_H__
+
+/* Beginning of header common to both ETMv3 and V4 */
+enum {
+	CS_ETM_MAGIC,
+	CS_ETM_CPU,
+	CS_ETM_SNAPSHOT,
+};
+
+/* ETMv3/PTM metadata */
+enum {
+	/* Dynamic, configurable parameters */
+	CS_ETM_ETMCR = CS_ETM_SNAPSHOT + 1,
+	CS_ETM_ETMTRACEIDR,
+	/* RO, taken from sysFS */
+	CS_ETM_ETMCCER,
+	CS_ETM_ETMIDR,
+	CS_ETM_PRIV_MAX,
+};
+
+static const u64 __perf_cs_etm_magic   = 0x3030303030303030ULL;
+#define CS_ETM_PRIV_SIZE (CS_ETM_PRIV_MAX * sizeof(u64))
+
+struct auxtrace_record *cs_etm_record_init(int *err);
+
+#endif
diff --git a/tools/perf/util/auxtrace.c b/tools/perf/util/auxtrace.c
index 430df9575dbe..666773d4e942 100644
--- a/tools/perf/util/auxtrace.c
+++ b/tools/perf/util/auxtrace.c
@@ -892,6 +892,7 @@ int perf_event__process_auxtrace_info(struct perf_tool *tool __maybe_unused,
 		return intel_pt_process_auxtrace_info(event, session);
 	case PERF_AUXTRACE_INTEL_BTS:
 		return intel_bts_process_auxtrace_info(event, session);
+	case PERF_AUXTRACE_CS_ETM:
 	case PERF_AUXTRACE_UNKNOWN:
 	default:
 		return -EINVAL;
diff --git a/tools/perf/util/auxtrace.h b/tools/perf/util/auxtrace.h
index d76177169484..d52db67da08d 100644
--- a/tools/perf/util/auxtrace.h
+++ b/tools/perf/util/auxtrace.h
@@ -41,6 +41,7 @@ enum auxtrace_type {
 	PERF_AUXTRACE_UNKNOWN,
 	PERF_AUXTRACE_INTEL_PT,
 	PERF_AUXTRACE_INTEL_BTS,
+	PERF_AUXTRACE_CS_ETM,
 };
 
 enum itrace_period_type {
-- 
1.9.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1275731 — [RESEND PATCH V4 25/26] perf tools: making coresight PMU listable

Adding the required mechanic allowing 'perf list pmu' to
discover coresight ETM/PTM tracers.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 MAINTAINERS                    |  1 +
 tools/perf/arch/arm/util/Build |  2 ++
 tools/perf/arch/arm/util/pmu.c | 34 ++++++++++++++++++++++++++++++++++
 tools/perf/config/Makefile     | 17 ++++++++++++-----
 4 files changed, 49 insertions(+), 5 deletions(-)
 create mode 100644 tools/perf/arch/arm/util/pmu.c

diff --git a/MAINTAINERS b/MAINTAINERS
index 994a7aeb4a2d..d73f35aeaf3b 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -999,6 +999,7 @@ F:	drivers/hwtracing/coresight/*
 F:	Documentation/trace/coresight.txt
 F:	Documentation/devicetree/bindings/arm/coresight.txt
 F:	Documentation/ABI/testing/sysfs-bus-coresight-devices-*
+F:	tools/perf/arch/arm/util/pmu.c
 
 ARM/CORGI MACHINE SUPPORT
 M:	Richard Purdie <rpurdie@rpsys.net>
diff --git a/tools/perf/arch/arm/util/Build b/tools/perf/arch/arm/util/Build
index d22e3d07de3d..66ab0b05549c 100644
--- a/tools/perf/arch/arm/util/Build
+++ b/tools/perf/arch/arm/util/Build
@@ -2,3 +2,5 @@ libperf-$(CONFIG_DWARF) += dwarf-regs.o
 
 libperf-$(CONFIG_LIBUNWIND)          += unwind-libunwind.o
 libperf-$(CONFIG_LIBDW_DWARF_UNWIND) += unwind-libdw.o
+
+libperf-$(CONFIG_AUXTRACE) += pmu.o
diff --git a/tools/perf/arch/arm/util/pmu.c b/tools/perf/arch/arm/util/pmu.c
new file mode 100644
index 000000000000..af9fb666b44f
--- /dev/null
+++ b/tools/perf/arch/arm/util/pmu.c
@@ -0,0 +1,34 @@
+/*
+ * Copyright(C) 2015 Linaro Limited. All rights reserved.
+ * Author: Mathieu Poirier <mathieu.poirier@linaro.org>
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms of the GNU General Public License version 2 as published by
+ * the Free Software Foundation.
+ *
+ * This program is distributed in the hope that 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.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#include <string.h>
+#include <linux/coresight-pmu.h>
+#include <linux/perf_event.h>
+
+#include "../../util/pmu.h"
+
+struct perf_event_attr
+*perf_pmu__get_default_config(struct perf_pmu *pmu __maybe_unused)
+{
+#ifdef HAVE_AUXTRACE_SUPPORT
+	if (!strcmp(pmu->name, CORESIGHT_ETM_PMU_NAME)) {
+		/* add ETM default config here */
+		pmu->selectable = true;
+	}
+#endif
+	return NULL;
+}
diff --git a/tools/perf/config/Makefile b/tools/perf/config/Makefile
index 38a08539f4bf..64bdb82b3337 100644
--- a/tools/perf/config/Makefile
+++ b/tools/perf/config/Makefile
@@ -626,12 +626,19 @@ ifdef LIBBABELTRACE
 endif
 
 ifndef NO_AUXTRACE
-  ifeq ($(feature-get_cpuid), 0)
-    msg := $(warning Your gcc lacks the __get_cpuid() builtin, disables support for auxtrace/Intel PT, please install a newer gcc);
-    NO_AUXTRACE := 1
+  ifeq ($(ARCH),x86)
+    ifeq ($(feature-get_cpuid), 0)
+      msg := $(warning Your gcc lacks the __get_cpuid() builtin, disables support for auxtrace/Intel PT, please install a newer gcc);
+      NO_AUXTRACE := 1
+    else
+      $(call detected,CONFIG_AUXTRACE)
+      CFLAGS += -DHAVE_AUXTRACE_SUPPORT
+    endif
   else
-    $(call detected,CONFIG_AUXTRACE)
-    CFLAGS += -DHAVE_AUXTRACE_SUPPORT
+    ifeq ($(ARCH),$(filter $(ARCH), arm arm64))
+      $(call detected,CONFIG_AUXTRACE)
+      CFLAGS += -DHAVE_AUXTRACE_SUPPORT
+    endif
   endif
 endif
 
-- 
1.9.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1275732 — [RESEND PATCH V4 24/26] perf tools: adding perf_session to *info_prive_size()

On some architecture the size of the private header may
be dependent on the number of tracers used in the session.  As
such adding a "struct perf_session" parameter, which should
contain all the required information.

Also adjusting the existing client of the interface to take
the new parameter into account.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 tools/perf/arch/x86/util/intel-bts.c | 4 +++-
 tools/perf/arch/x86/util/intel-pt.c  | 4 +++-
 tools/perf/util/auxtrace.c           | 7 ++++---
 tools/perf/util/auxtrace.h           | 6 ++++--
 4 files changed, 14 insertions(+), 7 deletions(-)

diff --git a/tools/perf/arch/x86/util/intel-bts.c b/tools/perf/arch/x86/util/intel-bts.c
index 9b94ce520917..6efdc5e86d97 100644
--- a/tools/perf/arch/x86/util/intel-bts.c
+++ b/tools/perf/arch/x86/util/intel-bts.c
@@ -60,7 +60,9 @@ struct branch {
 	u64 misc;
 };
 
-static size_t intel_bts_info_priv_size(struct auxtrace_record *itr __maybe_unused)
+static size_t
+intel_bts_info_priv_size(struct auxtrace_record *itr __maybe_unused,
+			 struct perf_session *session __maybe_unused)
 {
 	return INTEL_BTS_AUXTRACE_PRIV_SIZE;
 }
diff --git a/tools/perf/arch/x86/util/intel-pt.c b/tools/perf/arch/x86/util/intel-pt.c
index 2ca10d796c0b..1fd9a16f668c 100644
--- a/tools/perf/arch/x86/util/intel-pt.c
+++ b/tools/perf/arch/x86/util/intel-pt.c
@@ -273,7 +273,9 @@ intel_pt_pmu_default_config(struct perf_pmu *intel_pt_pmu)
 	return attr;
 }
 
-static size_t intel_pt_info_priv_size(struct auxtrace_record *itr __maybe_unused)
+static size_t
+intel_pt_info_priv_size(struct auxtrace_record *itr __maybe_unused,
+			struct perf_session *session __maybe_unused)
 {
 	return INTEL_PT_AUXTRACE_PRIV_SIZE;
 }
diff --git a/tools/perf/util/auxtrace.c b/tools/perf/util/auxtrace.c
index a980e7c50ee0..430df9575dbe 100644
--- a/tools/perf/util/auxtrace.c
+++ b/tools/perf/util/auxtrace.c
@@ -478,10 +478,11 @@ void auxtrace_heap__pop(struct auxtrace_heap *heap)
 			 heap_array[last].ordinal);
 }
 
-size_t auxtrace_record__info_priv_size(struct auxtrace_record *itr)
+size_t auxtrace_record__info_priv_size(struct auxtrace_record *itr,
+				       struct perf_session *session)
 {
 	if (itr)
-		return itr->info_priv_size(itr);
+		return itr->info_priv_size(itr, session);
 	return 0;
 }
 
@@ -852,7 +853,7 @@ int perf_event__synthesize_auxtrace_info(struct auxtrace_record *itr,
 	int err;
 
 	pr_debug2("Synthesizing auxtrace information\n");
-	priv_size = auxtrace_record__info_priv_size(itr);
+	priv_size = auxtrace_record__info_priv_size(itr, session);
 	ev = zalloc(sizeof(struct auxtrace_info_event) + priv_size);
 	if (!ev)
 		return -ENOMEM;
diff --git a/tools/perf/util/auxtrace.h b/tools/perf/util/auxtrace.h
index bf72b77a588a..d76177169484 100644
--- a/tools/perf/util/auxtrace.h
+++ b/tools/perf/util/auxtrace.h
@@ -289,7 +289,8 @@ struct auxtrace_record {
 	int (*recording_options)(struct auxtrace_record *itr,
 				 struct perf_evlist *evlist,
 				 struct record_opts *opts);
-	size_t (*info_priv_size)(struct auxtrace_record *itr);
+	size_t (*info_priv_size)(struct auxtrace_record *itr,
+				 struct perf_session *session);
 	int (*info_fill)(struct auxtrace_record *itr,
 			 struct perf_session *session,
 			 struct auxtrace_info_event *auxtrace_info,
@@ -425,7 +426,8 @@ int auxtrace_parse_snapshot_options(struct auxtrace_record *itr,
 int auxtrace_record__options(struct auxtrace_record *itr,
 			     struct perf_evlist *evlist,
 			     struct record_opts *opts);
-size_t auxtrace_record__info_priv_size(struct auxtrace_record *itr);
+size_t auxtrace_record__info_priv_size(struct auxtrace_record *itr,
+				       struct perf_session *session);
 int auxtrace_record__info_fill(struct auxtrace_record *itr,
 			       struct perf_session *session,
 			       struct auxtrace_info_event *auxtrace_info,
-- 
1.9.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1275733 — [RESEND PATCH V4 15/26] coresight: etm3x: implementing perf_enable/disable() API

That way traces can be enabled and disabled automatically
from the Perf subystem using the PMU abstraction.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 drivers/hwtracing/coresight/coresight-etm3x.c | 47 ++++++++++++++++++++++++---
 1 file changed, 43 insertions(+), 4 deletions(-)

diff --git a/drivers/hwtracing/coresight/coresight-etm3x.c b/drivers/hwtracing/coresight/coresight-etm3x.c
index dd319ef0f1ac..4ff3e1419fde 100644
--- a/drivers/hwtracing/coresight/coresight-etm3x.c
+++ b/drivers/hwtracing/coresight/coresight-etm3x.c
@@ -372,8 +372,10 @@ static void etm_enable_hw(void *info)
 	etm_set_prog(drvdata);
 
 	etmcr = etm_readl(drvdata, ETMCR);
-	etmcr &= (ETMCR_PWD_DWN | ETMCR_ETM_PRG);
+	/* Clear setting from a previous run if need be */
+	etmcr &= ~ETM3X_SUPPORTED_OPTIONS;
 	etmcr |= drvdata->port_size;
+	etmcr |= ETMCR_ETM_EN;
 	etm_writel(drvdata, config->ctrl | etmcr, ETMCR);
 	etm_writel(drvdata, config->trigger_event, ETMTRIGGER);
 	etm_writel(drvdata, config->startstop_ctrl, ETMTSSCR);
@@ -413,9 +415,6 @@ static void etm_enable_hw(void *info)
 	/* No VMID comparator value selected */
 	etm_writel(drvdata, 0x0, ETMVMIDCVR);
 
-	/* Ensures trace output is enabled from this ETM */
-	etm_writel(drvdata, config->ctrl | ETMCR_ETM_EN | etmcr, ETMCR);
-
 	etm_clr_prog(drvdata);
 	CS_LOCK(drvdata->base);
 
@@ -488,6 +487,18 @@ static void etm_set_config(struct coresight_device *csdev, void *config)
 	drvdata->config = config;
 }
 
+static int etm_enable_perf(struct coresight_device *csdev)
+{
+	struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+
+	if (WARN_ON_ONCE(drvdata->cpu != smp_processor_id()))
+		return -EINVAL;
+
+	etm_enable_hw(drvdata);
+
+	return 0;
+}
+
 static int etm_enable_sysfs(struct coresight_device *csdev)
 {
 	struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
@@ -534,6 +545,9 @@ static int etm_enable(struct coresight_device *csdev, u32 mode)
 	case CS_MODE_SYSFS:
 		ret = etm_enable_sysfs(csdev);
 		break;
+	case CS_MODE_PERF:
+		ret = etm_enable_perf(csdev);
+		break;
 	default:
 		ret = -EINVAL;
 	}
@@ -573,6 +587,28 @@ static void etm_disable_hw(void *info)
 	dev_dbg(drvdata->dev, "cpu: %d disable smp call done\n", drvdata->cpu);
 }
 
+static void etm_disable_perf(struct coresight_device *csdev)
+{
+	struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+
+	if (WARN_ON_ONCE(drvdata->cpu != smp_processor_id()))
+		return;
+
+	CS_UNLOCK(drvdata->base);
+
+	/* Setting the prog bit disables tracing immediately */
+	etm_set_prog(drvdata);
+	/* Get ready for another session */
+	drvdata->config = NULL;
+	/*
+	 * There is no way to know when the tracer will be used again so
+	 * power down the tracer.
+	 */
+	etm_set_pwrdwn(drvdata);
+
+	CS_LOCK(drvdata->base);
+}
+
 static void etm_disable_sysfs(struct coresight_device *csdev)
 {
 	struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
@@ -616,6 +652,9 @@ static void etm_disable(struct coresight_device *csdev)
 	case CS_MODE_SYSFS:
 		etm_disable_sysfs(csdev);
 		break;
+	case CS_MODE_PERF:
+		etm_disable_perf(csdev);
+		break;
 	default:
 		WARN_ON_ONCE(mode);
 		return;
-- 
1.9.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1275734 — [RESEND PATCH V4 17/26] coresight: etb10: moving to local atomic operations

Moving to use local atomic operations to take advantage of the
lockless implementation, something that will come handy when
the ETB is accessed from the Perf subsystem. Also changing the
name of the variable to something more meaningful.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 drivers/hwtracing/coresight/coresight-etb10.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/drivers/hwtracing/coresight/coresight-etb10.c b/drivers/hwtracing/coresight/coresight-etb10.c
index 1301edc44629..09e306361e4f 100644
--- a/drivers/hwtracing/coresight/coresight-etb10.c
+++ b/drivers/hwtracing/coresight/coresight-etb10.c
@@ -10,6 +10,7 @@
  * GNU General Public License for more details.
  */
 
+#include <asm/local.h>
 #include <linux/kernel.h>
 #include <linux/module.h>
 #include <linux/init.h>
@@ -71,7 +72,7 @@
  * @csdev:	component vitals needed by the framework.
  * @miscdev:	specifics to handle "/dev/xyz.etb" entry.
  * @spinlock:	only one at a time pls.
- * @in_use:	synchronise user space access to etb buffer.
+ * @reading:	synchronise user space access to etb buffer.
  * @buf:	area of memory where ETB buffer content gets sent.
  * @buffer_depth: size of @buf.
  * @enable:	this ETB is being used.
@@ -84,7 +85,7 @@ struct etb_drvdata {
 	struct coresight_device	*csdev;
 	struct miscdevice	miscdev;
 	spinlock_t		spinlock;
-	atomic_t		in_use;
+	local_t			reading;
 	u8			*buf;
 	u32			buffer_depth;
 	bool			enable;
@@ -277,7 +278,7 @@ static int etb_open(struct inode *inode, struct file *file)
 	struct etb_drvdata *drvdata = container_of(file->private_data,
 						   struct etb_drvdata, miscdev);
 
-	if (atomic_cmpxchg(&drvdata->in_use, 0, 1))
+	if (local_cmpxchg(&drvdata->reading, 0, 1))
 		return -EBUSY;
 
 	dev_dbg(drvdata->dev, "%s: successfully opened\n", __func__);
@@ -313,7 +314,7 @@ static int etb_release(struct inode *inode, struct file *file)
 {
 	struct etb_drvdata *drvdata = container_of(file->private_data,
 						   struct etb_drvdata, miscdev);
-	atomic_set(&drvdata->in_use, 0);
+	local_set(&drvdata->reading, 0);
 
 	dev_dbg(drvdata->dev, "%s: released\n", __func__);
 	return 0;
-- 
1.9.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1275736 — [RESEND PATCH V4 18/26] coresight: adding operation mode for sink->enable()

Adding an operation mode to sink->enable() API in order
to prevent simultaneous access from different callers.

TPIU and TMC won't be supplemented with the AUX area
API immediately and as such ignore the new mode.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 drivers/hwtracing/coresight/coresight-etb10.c | 31 +++++++++++++++++++++------
 drivers/hwtracing/coresight/coresight-priv.h  |  2 +-
 drivers/hwtracing/coresight/coresight-tmc.c   |  2 +-
 drivers/hwtracing/coresight/coresight-tpiu.c  |  2 +-
 drivers/hwtracing/coresight/coresight.c       | 10 ++++-----
 include/linux/coresight.h                     |  2 +-
 6 files changed, 33 insertions(+), 16 deletions(-)

diff --git a/drivers/hwtracing/coresight/coresight-etb10.c b/drivers/hwtracing/coresight/coresight-etb10.c
index 09e306361e4f..dc4d707b28aa 100644
--- a/drivers/hwtracing/coresight/coresight-etb10.c
+++ b/drivers/hwtracing/coresight/coresight-etb10.c
@@ -73,9 +73,9 @@
  * @miscdev:	specifics to handle "/dev/xyz.etb" entry.
  * @spinlock:	only one at a time pls.
  * @reading:	synchronise user space access to etb buffer.
+ * @mode:	this ETB is being used.
  * @buf:	area of memory where ETB buffer content gets sent.
  * @buffer_depth: size of @buf.
- * @enable:	this ETB is being used.
  * @trigger_cntr: amount of words to store after a trigger.
  */
 struct etb_drvdata {
@@ -86,9 +86,9 @@ struct etb_drvdata {
 	struct miscdevice	miscdev;
 	spinlock_t		spinlock;
 	local_t			reading;
+	local_t			mode;
 	u8			*buf;
 	u32			buffer_depth;
-	bool			enable;
 	u32			trigger_cntr;
 };
 
@@ -133,18 +133,34 @@ static void etb_enable_hw(struct etb_drvdata *drvdata)
 	CS_LOCK(drvdata->base);
 }
 
-static int etb_enable(struct coresight_device *csdev)
+static int etb_enable(struct coresight_device *csdev, u32 mode)
 {
-	struct etb_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+	u32 val;
 	unsigned long flags;
+	struct etb_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+
+	val = local_cmpxchg(&drvdata->mode,
+			    CS_MODE_DISABLED, mode);
+	/*
+	 * When accessing from Perf, a HW buffer can be handled
+	 * by a single trace entity.  In sysFS mode many tracers
+	 * can be logging to the same HW buffer.
+	 */
+	if (val == CS_MODE_PERF)
+		return -EBUSY;
+
+	/* Nothing to do, the tracer is already enabled. */
+	if (val == CS_MODE_SYSFS)
+		goto out;
 
 	spin_lock_irqsave(&drvdata->spinlock, flags);
 	etb_enable_hw(drvdata);
-	drvdata->enable = true;
 	spin_unlock_irqrestore(&drvdata->spinlock, flags);
 
+out:
 	dev_info(drvdata->dev, "ETB enabled\n");
 	return 0;
+
 }
 
 static void etb_disable_hw(struct etb_drvdata *drvdata)
@@ -243,9 +259,10 @@ static void etb_disable(struct coresight_device *csdev)
 	spin_lock_irqsave(&drvdata->spinlock, flags);
 	etb_disable_hw(drvdata);
 	etb_dump_hw(drvdata);
-	drvdata->enable = false;
 	spin_unlock_irqrestore(&drvdata->spinlock, flags);
 
+	local_set(&drvdata->mode, CS_MODE_DISABLED);
+
 	dev_info(drvdata->dev, "ETB disabled\n");
 }
 
@@ -263,7 +280,7 @@ static void etb_dump(struct etb_drvdata *drvdata)
 	unsigned long flags;
 
 	spin_lock_irqsave(&drvdata->spinlock, flags);
-	if (drvdata->enable) {
+	if (local_read(&drvdata->mode) == CS_MODE_SYSFS) {
 		etb_disable_hw(drvdata);
 		etb_dump_hw(drvdata);
 		etb_enable_hw(drvdata);
diff --git a/drivers/hwtracing/coresight/coresight-priv.h b/drivers/hwtracing/coresight/coresight-priv.h
index 932f34a84d96..333eddaed339 100644
--- a/drivers/hwtracing/coresight/coresight-priv.h
+++ b/drivers/hwtracing/coresight/coresight-priv.h
@@ -62,7 +62,7 @@ static inline void CS_UNLOCK(void __iomem *addr)
 }
 
 void coresight_disable_path(struct list_head *path);
-int coresight_enable_path(struct list_head *path);
+int coresight_enable_path(struct list_head *path, u32 mode);
 struct coresight_device *coresight_get_sink(struct list_head *path);
 struct list_head *coresight_build_path(struct coresight_device *csdev);
 void coresight_release_path(struct list_head *path);
diff --git a/drivers/hwtracing/coresight/coresight-tmc.c b/drivers/hwtracing/coresight/coresight-tmc.c
index 5e2a71767870..d7e83da7a621 100644
--- a/drivers/hwtracing/coresight/coresight-tmc.c
+++ b/drivers/hwtracing/coresight/coresight-tmc.c
@@ -265,7 +265,7 @@ static int tmc_enable(struct tmc_drvdata *drvdata, enum tmc_mode mode)
 	return 0;
 }
 
-static int tmc_enable_sink(struct coresight_device *csdev)
+static int tmc_enable_sink(struct coresight_device *csdev, u32 mode)
 {
 	struct tmc_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
 
diff --git a/drivers/hwtracing/coresight/coresight-tpiu.c b/drivers/hwtracing/coresight/coresight-tpiu.c
index e19b86e61c38..0a09135382f6 100644
--- a/drivers/hwtracing/coresight/coresight-tpiu.c
+++ b/drivers/hwtracing/coresight/coresight-tpiu.c
@@ -70,7 +70,7 @@ static void tpiu_enable_hw(struct tpiu_drvdata *drvdata)
 	CS_LOCK(drvdata->base);
 }
 
-static int tpiu_enable(struct coresight_device *csdev)
+static int tpiu_enable(struct coresight_device *csdev, u32 mode)
 {
 	struct tpiu_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
 
diff --git a/drivers/hwtracing/coresight/coresight.c b/drivers/hwtracing/coresight/coresight.c
index 47a4e109f700..52311e77da9d 100644
--- a/drivers/hwtracing/coresight/coresight.c
+++ b/drivers/hwtracing/coresight/coresight.c
@@ -121,13 +121,13 @@ static int coresight_find_link_outport(struct coresight_device *csdev,
 	return 0;
 }
 
-static int coresight_enable_sink(struct coresight_device *csdev)
+static int coresight_enable_sink(struct coresight_device *csdev, u32 mode)
 {
 	int ret;
 
 	if (!csdev->enable) {
 		if (sink_ops(csdev)->enable) {
-			ret = sink_ops(csdev)->enable(csdev);
+			ret = sink_ops(csdev)->enable(csdev, mode);
 			if (ret)
 				return ret;
 		}
@@ -283,7 +283,7 @@ void coresight_disable_path(struct list_head *path)
 	}
 }
 
-int coresight_enable_path(struct list_head *path)
+int coresight_enable_path(struct list_head *path, u32 mode)
 {
 
 	int ret = 0;
@@ -296,7 +296,7 @@ int coresight_enable_path(struct list_head *path)
 		switch (csdev->type) {
 		case CORESIGHT_DEV_TYPE_SINK:
 		case CORESIGHT_DEV_TYPE_LINKSINK:
-			ret = coresight_enable_sink(csdev);
+			ret = coresight_enable_sink(csdev, mode);
 			if (ret)
 				goto err;
 			break;
@@ -454,7 +454,7 @@ int coresight_enable(struct coresight_device *csdev)
 		goto out;
 	}
 
-	ret = coresight_enable_path(path);
+	ret = coresight_enable_path(path, CS_MODE_SYSFS);
 	if (ret)
 		goto err_path;
 
diff --git a/include/linux/coresight.h b/include/linux/coresight.h
index f2148bc2bcfe..80b3465bd459 100644
--- a/include/linux/coresight.h
+++ b/include/linux/coresight.h
@@ -185,7 +185,7 @@ struct coresight_device {
  * @disable:	disables the sink.
  */
 struct coresight_ops_sink {
-	int (*enable)(struct coresight_device *csdev);
+	int (*enable)(struct coresight_device *csdev, u32 mode);
 	void (*disable)(struct coresight_device *csdev);
 };
 
-- 
1.9.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1275737 — [RESEND PATCH V4 01/26] coresight: etm3x: moving etm_readl/writel to header file

Moving functions etm_readl/writel to file "coresight-etm.h"
so that the main ETM3x driver can be split in more than one
file.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 drivers/hwtracing/coresight/coresight-etm.h   | 29 +++++++++++++++++++++++++++
 drivers/hwtracing/coresight/coresight-etm3x.c | 29 ---------------------------
 2 files changed, 29 insertions(+), 29 deletions(-)

diff --git a/drivers/hwtracing/coresight/coresight-etm.h b/drivers/hwtracing/coresight/coresight-etm.h
index b4481eb29304..34f7db881fa7 100644
--- a/drivers/hwtracing/coresight/coresight-etm.h
+++ b/drivers/hwtracing/coresight/coresight-etm.h
@@ -251,4 +251,33 @@ enum etm_addr_type {
 	ETM_ADDR_TYPE_START,
 	ETM_ADDR_TYPE_STOP,
 };
+
+static inline void etm_writel(struct etm_drvdata *drvdata,
+			      u32 val, u32 off)
+{
+	if (drvdata->use_cp14) {
+		if (etm_writel_cp14(off, val)) {
+			dev_err(drvdata->dev,
+				"invalid CP14 access to ETM reg: %#x", off);
+		}
+	} else {
+		writel_relaxed(val, drvdata->base + off);
+	}
+}
+
+static inline unsigned int etm_readl(struct etm_drvdata *drvdata, u32 off)
+{
+	u32 val;
+
+	if (drvdata->use_cp14) {
+		if (etm_readl_cp14(off, &val)) {
+			dev_err(drvdata->dev,
+				"invalid CP14 access to ETM reg: %#x", off);
+		}
+	} else {
+		val = readl_relaxed(drvdata->base + off);
+	}
+
+	return val;
+}
 #endif
diff --git a/drivers/hwtracing/coresight/coresight-etm3x.c b/drivers/hwtracing/coresight/coresight-etm3x.c
index d630b7ece735..c1dc095c3fb0 100644
--- a/drivers/hwtracing/coresight/coresight-etm3x.c
+++ b/drivers/hwtracing/coresight/coresight-etm3x.c
@@ -42,35 +42,6 @@ module_param_named(boot_enable, boot_enable, int, S_IRUGO);
 static int etm_count;
 static struct etm_drvdata *etmdrvdata[NR_CPUS];
 
-static inline void etm_writel(struct etm_drvdata *drvdata,
-			      u32 val, u32 off)
-{
-	if (drvdata->use_cp14) {
-		if (etm_writel_cp14(off, val)) {
-			dev_err(drvdata->dev,
-				"invalid CP14 access to ETM reg: %#x", off);
-		}
-	} else {
-		writel_relaxed(val, drvdata->base + off);
-	}
-}
-
-static inline unsigned int etm_readl(struct etm_drvdata *drvdata, u32 off)
-{
-	u32 val;
-
-	if (drvdata->use_cp14) {
-		if (etm_readl_cp14(off, &val)) {
-			dev_err(drvdata->dev,
-				"invalid CP14 access to ETM reg: %#x", off);
-		}
-	} else {
-		val = readl_relaxed(drvdata->base + off);
-	}
-
-	return val;
-}
-
 /*
  * Memory mapped writes to clear os lock are not supported on some processors
  * and OS lock must be unlocked before any memory mapped access on such
-- 
1.9.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

Page 1 of 2  [1] 2  Next page →

Back to top | Article view | linux.kernel

csiph-web