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 2 of 2 — ← Prev page 1 [2]

#1275743 — [RESEND PATCH V4 05/26] coresight: etm3x: implementing 'cpu_id()' API

Other than plainly parsing the device tree there is no way to
know which CPU a tracer is affined to.  As such adding an
interface to lookup the CPU field enclosed in the etm_drvdata
structure that was initialised at boot time.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 drivers/hwtracing/coresight/coresight-etm3x.c | 8 ++++++++
 include/linux/coresight.h                     | 3 +++
 2 files changed, 11 insertions(+)

diff --git a/drivers/hwtracing/coresight/coresight-etm3x.c b/drivers/hwtracing/coresight/coresight-etm3x.c
index 67af548a874d..74526f044a06 100644
--- a/drivers/hwtracing/coresight/coresight-etm3x.c
+++ b/drivers/hwtracing/coresight/coresight-etm3x.c
@@ -312,6 +312,13 @@ static void etm_enable_hw(void *info)
 	dev_dbg(drvdata->dev, "cpu: %d enable smp call done\n", drvdata->cpu);
 }
 
+static int etm_cpu_id(struct coresight_device *csdev)
+{
+	struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+
+	return drvdata->cpu;
+}
+
 int etm_get_trace_id(struct etm_drvdata *drvdata)
 {
 	unsigned long flags;
@@ -438,6 +445,7 @@ 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,
 	.enable		= etm_enable,
 	.disable	= etm_disable,
diff --git a/include/linux/coresight.h b/include/linux/coresight.h
index a7cabfa23b55..bf62b265bf52 100644
--- a/include/linux/coresight.h
+++ b/include/linux/coresight.h
@@ -205,12 +205,15 @@ struct coresight_ops_link {
 /**
  * struct coresight_ops_source - basic operations for a source
  * Operations available for sources.
+ * @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.
  * @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);
 	int (*enable)(struct coresight_device *csdev);
 	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]

#1275744 — [RESEND PATCH V4 09/26] coresight: add API to get sink from path

Add an API allowing external code to quickly get a handle on the
sink within a path.  The sink is always last, but adding an API allows
to keep the path's node structure private and remove redundant checks.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 drivers/hwtracing/coresight/coresight-priv.h |  1 +
 drivers/hwtracing/coresight/coresight.c      | 15 +++++++++++++++
 2 files changed, 16 insertions(+)

diff --git a/drivers/hwtracing/coresight/coresight-priv.h b/drivers/hwtracing/coresight/coresight-priv.h
index 2dc59a5c5b1e..ed116b303e87 100644
--- a/drivers/hwtracing/coresight/coresight-priv.h
+++ b/drivers/hwtracing/coresight/coresight-priv.h
@@ -60,6 +60,7 @@ static inline void CS_UNLOCK(void __iomem *addr)
 
 void coresight_disable_path(struct list_head *path);
 int coresight_enable_path(struct list_head *path);
+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.c b/drivers/hwtracing/coresight/coresight.c
index 7e4064b31c99..47a4e109f700 100644
--- a/drivers/hwtracing/coresight/coresight.c
+++ b/drivers/hwtracing/coresight/coresight.c
@@ -322,6 +322,21 @@ err:
 	goto out;
 }
 
+struct coresight_device *coresight_get_sink(struct list_head *path)
+{
+	struct coresight_device *csdev;
+
+	if (!path)
+		return NULL;
+
+	csdev = list_last_entry(path, struct coresight_node, link)->csdev;
+	if (csdev->type != CORESIGHT_DEV_TYPE_SINK &&
+	    csdev->type != CORESIGHT_DEV_TYPE_LINKSINK)
+		return NULL;
+
+	return csdev;
+}
+
 /**
  * _coresight_build_path - recursively build a path from a @csdev to a sink.
  * @csdev:	The device to start from.
-- 
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]

#1275745 — [RESEND PATCH V4 08/26] coresight: etm3x: adding operation mode for etm_enable()

Adding a new mode to source API enable() in order to
distinguish where the request comes from.  That way it is
possible to perform different operations based on where
the request was issued from.

The ETM4x driver is also modified to keep in sync with the
new interface.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 drivers/hwtracing/coresight/coresight-etm.h        |  5 +-
 .../hwtracing/coresight/coresight-etm3x-sysfs.c    |  4 +-
 drivers/hwtracing/coresight/coresight-etm3x.c      | 68 +++++++++++++++++++---
 drivers/hwtracing/coresight/coresight-etm4x.c      |  2 +-
 drivers/hwtracing/coresight/coresight-priv.h       |  6 ++
 drivers/hwtracing/coresight/coresight.c            |  6 +-
 include/linux/coresight.h                          |  2 +-
 7 files changed, 76 insertions(+), 17 deletions(-)

diff --git a/drivers/hwtracing/coresight/coresight-etm.h b/drivers/hwtracing/coresight/coresight-etm.h
index 9317510d555f..a8663780a908 100644
--- a/drivers/hwtracing/coresight/coresight-etm.h
+++ b/drivers/hwtracing/coresight/coresight-etm.h
@@ -13,6 +13,7 @@
 #ifndef _CORESIGHT_CORESIGHT_ETM_H
 #define _CORESIGHT_CORESIGHT_ETM_H
 
+#include <asm/local.h>
 #include <linux/spinlock.h>
 #include "coresight-priv.h"
 
@@ -214,7 +215,7 @@ struct etm_config {
  * @port_size:	port size as reported by ETMCR bit 4-6 and 21.
  * @arch:	ETM/PTM version number.
  * @use_cpu14:	true if management registers need to be accessed via CP14.
- * @enable:	is this ETM/PTM currently tracing.
+ * @mode:	this tracer's mode, i.e sysFS, Perf or disabled.
  * @sticky_enable: true if ETM base configuration has been done.
  * @boot_enable:true if we should start tracing at boot time.
  * @os_unlock:	true if access to management registers is allowed.
@@ -238,7 +239,7 @@ struct etm_drvdata {
 	int				port_size;
 	u8				arch;
 	bool				use_cp14;
-	bool				enable;
+	local_t				mode;
 	bool				sticky_enable;
 	bool				boot_enable;
 	bool				os_unlock;
diff --git a/drivers/hwtracing/coresight/coresight-etm3x-sysfs.c b/drivers/hwtracing/coresight/coresight-etm3x-sysfs.c
index 410d29e2b90c..6da50d94ec6f 100644
--- a/drivers/hwtracing/coresight/coresight-etm3x-sysfs.c
+++ b/drivers/hwtracing/coresight/coresight-etm3x-sysfs.c
@@ -811,7 +811,7 @@ static ssize_t cntr_val_show(struct device *dev,
 	if (WARN_ON_ONCE(!config))
 		return -EINVAL;
 
-	if (!drvdata->enable) {
+	if (!local_read(&drvdata->mode)) {
 		spin_lock(&drvdata->spinlock);
 		for (i = 0; i < drvdata->nr_cntr; i++)
 			ret += sprintf(buf, "counter %d: %x\n",
@@ -1072,7 +1072,7 @@ static ssize_t seq_curr_state_show(struct device *dev,
 	if (WARN_ON_ONCE(!config))
 		return -EINVAL;
 
-	if (!drvdata->enable) {
+	if (!local_read(&drvdata->mode)) {
 		val = config->seq_curr_state;
 		goto out;
 	}
diff --git a/drivers/hwtracing/coresight/coresight-etm3x.c b/drivers/hwtracing/coresight/coresight-etm3x.c
index 660c1e07601b..5a2be0590754 100644
--- a/drivers/hwtracing/coresight/coresight-etm3x.c
+++ b/drivers/hwtracing/coresight/coresight-etm3x.c
@@ -327,7 +327,7 @@ int etm_get_trace_id(struct etm_drvdata *drvdata)
 	if (!drvdata)
 		goto out;
 
-	if (!drvdata->enable)
+	if (!local_read(&drvdata->mode))
 		return drvdata->traceid;
 
 	pm_runtime_get_sync(drvdata->dev);
@@ -353,7 +353,7 @@ static int etm_trace_id(struct coresight_device *csdev)
 	return etm_get_trace_id(drvdata);
 }
 
-static int etm_enable(struct coresight_device *csdev)
+static int etm_enable_sysfs(struct coresight_device *csdev)
 {
 	struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
 	int ret;
@@ -372,18 +372,44 @@ static int etm_enable(struct coresight_device *csdev)
 			goto err;
 	}
 
-	drvdata->enable = true;
 	drvdata->sticky_enable = true;
-
 	spin_unlock(&drvdata->spinlock);
 
 	dev_info(drvdata->dev, "ETM tracing enabled\n");
 	return 0;
+
 err:
 	spin_unlock(&drvdata->spinlock);
 	return ret;
 }
 
+static int etm_enable(struct coresight_device *csdev, u32 mode)
+{
+	int ret;
+	u32 val;
+	struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+
+	val = local_cmpxchg(&drvdata->mode, CS_MODE_DISABLED, mode);
+
+	/* Someone is already using the tracer */
+	if (val)
+		return -EBUSY;
+
+	switch (mode) {
+	case CS_MODE_SYSFS:
+		ret = etm_enable_sysfs(csdev);
+		break;
+	default:
+		ret = -EINVAL;
+	}
+
+	/* The tracer didn't start */
+	if (ret)
+		local_set(&drvdata->mode, CS_MODE_DISABLED);
+
+	return ret;
+}
+
 static void etm_disable_hw(void *info)
 {
 	int i;
@@ -415,7 +441,7 @@ static void etm_disable_hw(void *info)
 	dev_dbg(drvdata->dev, "cpu: %d disable smp call done\n", drvdata->cpu);
 }
 
-static void etm_disable(struct coresight_device *csdev)
+static void etm_disable_sysfs(struct coresight_device *csdev)
 {
 	struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
 
@@ -433,7 +459,6 @@ static void etm_disable(struct coresight_device *csdev)
 	 * ensures that register writes occur when cpu is powered.
 	 */
 	smp_call_function_single(drvdata->cpu, etm_disable_hw, drvdata, 1);
-	drvdata->enable = false;
 
 	spin_unlock(&drvdata->spinlock);
 	put_online_cpus();
@@ -441,6 +466,33 @@ static void etm_disable(struct coresight_device *csdev)
 	dev_info(drvdata->dev, "ETM tracing disabled\n");
 }
 
+static void etm_disable(struct coresight_device *csdev)
+{
+	u32 mode;
+	struct etm_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
+
+	/*
+	 * For as long as the tracer isn't disabled another entity can't
+	 * change its status.  As such we can read the status here without
+	 * fearing it will change under us.
+	 */
+	mode = local_read(&drvdata->mode);
+
+	switch (mode) {
+	case CS_MODE_DISABLED:
+		break;
+	case CS_MODE_SYSFS:
+		etm_disable_sysfs(csdev);
+		break;
+	default:
+		WARN_ON_ONCE(mode);
+		return;
+	}
+
+	if (mode)
+		local_set(&drvdata->mode, CS_MODE_DISABLED);
+}
+
 static const struct coresight_ops_source etm_source_ops = {
 	.cpu_id		= etm_cpu_id,
 	.trace_id	= etm_trace_id,
@@ -468,7 +520,7 @@ static int etm_cpu_callback(struct notifier_block *nfb, unsigned long action,
 			etmdrvdata[cpu]->os_unlock = true;
 		}
 
-		if (etmdrvdata[cpu]->enable)
+		if (local_read(&etmdrvdata[cpu]->mode))
 			etm_enable_hw(etmdrvdata[cpu]);
 		spin_unlock(&etmdrvdata[cpu]->spinlock);
 		break;
@@ -481,7 +533,7 @@ static int etm_cpu_callback(struct notifier_block *nfb, unsigned long action,
 
 	case CPU_DYING:
 		spin_lock(&etmdrvdata[cpu]->spinlock);
-		if (etmdrvdata[cpu]->enable)
+		if (local_read(&etmdrvdata[cpu]->mode))
 			etm_disable_hw(etmdrvdata[cpu]);
 		spin_unlock(&etmdrvdata[cpu]->spinlock);
 		break;
diff --git a/drivers/hwtracing/coresight/coresight-etm4x.c b/drivers/hwtracing/coresight/coresight-etm4x.c
index 1c6e32dd6e49..4ab291b3a6c8 100644
--- a/drivers/hwtracing/coresight/coresight-etm4x.c
+++ b/drivers/hwtracing/coresight/coresight-etm4x.c
@@ -180,7 +180,7 @@ static void etm4_enable_hw(void *info)
 	dev_dbg(drvdata->dev, "cpu: %d enable smp call done\n", drvdata->cpu);
 }
 
-static int etm4_enable(struct coresight_device *csdev)
+static int etm4_enable(struct coresight_device *csdev, u32 mode)
 {
 	struct etmv4_drvdata *drvdata = dev_get_drvdata(csdev->dev.parent);
 	int ret;
diff --git a/drivers/hwtracing/coresight/coresight-priv.h b/drivers/hwtracing/coresight/coresight-priv.h
index 7b193a34d709..2dc59a5c5b1e 100644
--- a/drivers/hwtracing/coresight/coresight-priv.h
+++ b/drivers/hwtracing/coresight/coresight-priv.h
@@ -34,6 +34,12 @@
 #define TIMEOUT_US		100
 #define BMVAL(val, lsb, msb)	((val & GENMASK(msb, lsb)) >> lsb)
 
+enum cs_mode {
+	CS_MODE_DISABLED,
+	CS_MODE_SYSFS,
+	CS_MODE_PERF,
+};
+
 static inline void CS_LOCK(void __iomem *addr)
 {
 	do {
diff --git a/drivers/hwtracing/coresight/coresight.c b/drivers/hwtracing/coresight/coresight.c
index a4103a216e33..7e4064b31c99 100644
--- a/drivers/hwtracing/coresight/coresight.c
+++ b/drivers/hwtracing/coresight/coresight.c
@@ -222,7 +222,7 @@ static void coresight_disable_link(struct coresight_device *csdev,
 	csdev->enable = false;
 }
 
-static int coresight_enable_source(struct coresight_device *csdev)
+static int coresight_enable_source(struct coresight_device *csdev, u32 mode)
 {
 	int ret;
 
@@ -234,7 +234,7 @@ static int coresight_enable_source(struct coresight_device *csdev)
 
 	if (!csdev->enable) {
 		if (source_ops(csdev)->enable) {
-			ret = source_ops(csdev)->enable(csdev);
+			ret = source_ops(csdev)->enable(csdev, mode);
 			if (ret)
 				return ret;
 		}
@@ -443,7 +443,7 @@ int coresight_enable(struct coresight_device *csdev)
 	if (ret)
 		goto err_path;
 
-	ret = coresight_enable_source(csdev);
+	ret = coresight_enable_source(csdev, CS_MODE_SYSFS);
 	if (ret)
 		goto err_source;
 
diff --git a/include/linux/coresight.h b/include/linux/coresight.h
index 851ecb22397e..61dfb8d511ea 100644
--- a/include/linux/coresight.h
+++ b/include/linux/coresight.h
@@ -213,7 +213,7 @@ struct coresight_ops_link {
 struct coresight_ops_source {
 	int (*cpu_id)(struct coresight_device *csdev);
 	int (*trace_id)(struct coresight_device *csdev);
-	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]

#1275746 — [RESEND PATCH V4 06/26] coresight: associating path with session rather than tracer

When using the Coresight framework from the sysFS interface a
tracer is always handling a single session and as such, a path
can be associated with a tracer.  But when supporting multiple
session per tracer there is no guarantee that sessions will always
have the same path from source to sink.

This patch is removing the automatic association between path and
tracers.  The building of a path and enablement of the components
in the path are decoupled, allowing for the association of a path
with a session rather than a tracer.

To keep backward functionality with the current sysFS access methods
a per-cpu place holder is used to keep a handle on the path built when
tracers are enabled.  Lastly APIs to build paths and enable tracers are
made public so that other subsystem can interact with the Coresight
framework.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 drivers/hwtracing/coresight/coresight-priv.h |   5 +
 drivers/hwtracing/coresight/coresight.c      | 300 ++++++++++++++++++---------
 include/linux/coresight.h                    |   2 -
 3 files changed, 208 insertions(+), 99 deletions(-)

diff --git a/drivers/hwtracing/coresight/coresight-priv.h b/drivers/hwtracing/coresight/coresight-priv.h
index 62fcd98cc7cf..7b193a34d709 100644
--- a/drivers/hwtracing/coresight/coresight-priv.h
+++ b/drivers/hwtracing/coresight/coresight-priv.h
@@ -52,6 +52,11 @@ static inline void CS_UNLOCK(void __iomem *addr)
 	} while (0);
 }
 
+void coresight_disable_path(struct list_head *path);
+int coresight_enable_path(struct list_head *path);
+struct list_head *coresight_build_path(struct coresight_device *csdev);
+void coresight_release_path(struct list_head *path);
+
 #ifdef CONFIG_CORESIGHT_SOURCE_ETM3X
 extern int etm_readl_cp14(u32 off, unsigned int *val);
 extern int etm_writel_cp14(u32 off, u32 val);
diff --git a/drivers/hwtracing/coresight/coresight.c b/drivers/hwtracing/coresight/coresight.c
index e25492137d8b..4beea67fd5a1 100644
--- a/drivers/hwtracing/coresight/coresight.c
+++ b/drivers/hwtracing/coresight/coresight.c
@@ -29,6 +29,22 @@
 
 static DEFINE_MUTEX(coresight_mutex);
 
+/**
+ * struct coresight_node - elements of a path, from source to sink
+ * @csdev:	Address of an element.
+ * @link:	hook to the list.
+ */
+struct coresight_node {
+	struct coresight_device *csdev;
+	struct list_head link;
+};
+
+/*
+ * When operating Coresight drivers from the sysFS interface, only a single
+ * path can exist from a tracer (associated to a CPU) to a sink.
+ */
+static DEFINE_PER_CPU(struct list_head *, sysfs_path);
+
 static int coresight_id_match(struct device *dev, void *data)
 {
 	int trace_id, i_trace_id;
@@ -68,15 +84,12 @@ static int coresight_source_is_unique(struct coresight_device *csdev)
 				 csdev, coresight_id_match);
 }
 
-static int coresight_find_link_inport(struct coresight_device *csdev)
+static int coresight_find_link_inport(struct coresight_device *csdev,
+				      struct coresight_device *parent)
 {
 	int i;
-	struct coresight_device *parent;
 	struct coresight_connection *conn;
 
-	parent = container_of(csdev->path_link.next,
-			      struct coresight_device, path_link);
-
 	for (i = 0; i < parent->nr_outport; i++) {
 		conn = &parent->conns[i];
 		if (conn->child_dev == csdev)
@@ -89,15 +102,12 @@ static int coresight_find_link_inport(struct coresight_device *csdev)
 	return 0;
 }
 
-static int coresight_find_link_outport(struct coresight_device *csdev)
+static int coresight_find_link_outport(struct coresight_device *csdev,
+				       struct coresight_device *child)
 {
 	int i;
-	struct coresight_device *child;
 	struct coresight_connection *conn;
 
-	child = container_of(csdev->path_link.prev,
-			     struct coresight_device, path_link);
-
 	for (i = 0; i < csdev->nr_outport; i++) {
 		conn = &csdev->conns[i];
 		if (conn->child_dev == child)
@@ -138,14 +148,19 @@ static void coresight_disable_sink(struct coresight_device *csdev)
 	}
 }
 
-static int coresight_enable_link(struct coresight_device *csdev)
+static int coresight_enable_link(struct coresight_device *csdev,
+				 struct coresight_device *parent,
+				 struct coresight_device *child)
 {
 	int ret;
 	int link_subtype;
 	int refport, inport, outport;
 
-	inport = coresight_find_link_inport(csdev);
-	outport = coresight_find_link_outport(csdev);
+	if (!parent || !child)
+		return -EINVAL;
+
+	inport = coresight_find_link_inport(csdev, parent);
+	outport = coresight_find_link_outport(csdev, child);
 	link_subtype = csdev->subtype.link_subtype;
 
 	if (link_subtype == CORESIGHT_DEV_SUBTYPE_LINK_MERG)
@@ -168,14 +183,19 @@ static int coresight_enable_link(struct coresight_device *csdev)
 	return 0;
 }
 
-static void coresight_disable_link(struct coresight_device *csdev)
+static void coresight_disable_link(struct coresight_device *csdev,
+				   struct coresight_device *parent,
+				   struct coresight_device *child)
 {
 	int i, nr_conns;
 	int link_subtype;
 	int refport, inport, outport;
 
-	inport = coresight_find_link_inport(csdev);
-	outport = coresight_find_link_outport(csdev);
+	if (!parent || !child)
+		return;
+
+	inport = coresight_find_link_inport(csdev, parent);
+	outport = coresight_find_link_outport(csdev, child);
 	link_subtype = csdev->subtype.link_subtype;
 
 	if (link_subtype == CORESIGHT_DEV_SUBTYPE_LINK_MERG) {
@@ -235,109 +255,167 @@ static void coresight_disable_source(struct coresight_device *csdev)
 	}
 }
 
-static int coresight_enable_path(struct list_head *path)
+void coresight_disable_path(struct list_head *path)
 {
-	int ret = 0;
-	struct coresight_device *cd;
-
-	/*
-	 * At this point we have a full @path, from source to sink.  The
-	 * sink is the first entry and the source the last one.  Go through
-	 * all the components and enable them one by one.
-	 */
-	list_for_each_entry(cd, path, path_link) {
-		if (cd == list_first_entry(path, struct coresight_device,
-					   path_link)) {
-			ret = coresight_enable_sink(cd);
-		} else if (list_is_last(&cd->path_link, path)) {
-			/*
-			 * Don't enable the source just yet - this needs to
-			 * happen at the very end when all links and sink
-			 * along the path have been configured properly.
-			 */
-			;
-		} else {
-			ret = coresight_enable_link(cd);
+	struct coresight_node *nd;
+	struct coresight_device *csdev, *parent, *child;
+
+	list_for_each_entry(nd, path, link) {
+		csdev = nd->csdev;
+
+		switch (csdev->type) {
+		case CORESIGHT_DEV_TYPE_SINK:
+		case CORESIGHT_DEV_TYPE_LINKSINK:
+			coresight_disable_sink(csdev);
+			break;
+		case CORESIGHT_DEV_TYPE_SOURCE:
+			/* sources are disabled from either sysFS or Perf */
+			break;
+		case CORESIGHT_DEV_TYPE_LINK:
+			parent = list_prev_entry(nd, link)->csdev;
+			child = list_next_entry(nd, link)->csdev;
+			coresight_disable_link(csdev, parent, child);
+			break;
+		default:
+			break;
 		}
-		if (ret)
-			goto err;
 	}
+}
 
-	return 0;
-err:
-	list_for_each_entry_continue_reverse(cd, path, path_link) {
-		if (cd == list_first_entry(path, struct coresight_device,
-					   path_link)) {
-			coresight_disable_sink(cd);
-		} else if (list_is_last(&cd->path_link, path)) {
-			;
-		} else {
-			coresight_disable_link(cd);
+int coresight_enable_path(struct list_head *path)
+{
+
+	int ret = 0;
+	struct coresight_node *nd;
+	struct coresight_device *csdev, *parent, *child;
+
+	list_for_each_entry_reverse(nd, path, link) {
+		csdev = nd->csdev;
+
+		switch (csdev->type) {
+		case CORESIGHT_DEV_TYPE_SINK:
+		case CORESIGHT_DEV_TYPE_LINKSINK:
+			ret = coresight_enable_sink(csdev);
+			if (ret)
+				goto err;
+			break;
+		case CORESIGHT_DEV_TYPE_SOURCE:
+			/* sources are enabled from either sysFS or Perf */
+			break;
+		case CORESIGHT_DEV_TYPE_LINK:
+			parent = list_prev_entry(nd, link)->csdev;
+			child = list_next_entry(nd, link)->csdev;
+			ret = coresight_enable_link(csdev, parent, child);
+			if (ret)
+				goto err;
+			break;
+		default:
+			goto err;
 		}
 	}
 
+out:
 	return ret;
+err:
+	coresight_disable_path(path);
+	goto out;
 }
 
-static int coresight_disable_path(struct list_head *path)
+/**
+ * _coresight_build_path - recursively build a path from a @csdev to a sink.
+ * @csdev:	The device to start from.
+ * @path:	The list to add devices to.
+ *
+ * The tree of Coresight device is traversed until an activated sink is
+ * found.  From there the sink is added to the list along with all the
+ * devices that led to that point - the end result is a list from source
+ * to sink. In that list the source is the first device and the sink the
+ * last one.
+ */
+static int _coresight_build_path(struct coresight_device *csdev,
+				 struct list_head *path)
 {
-	struct coresight_device *cd;
-
-	list_for_each_entry_reverse(cd, path, path_link) {
-		if (cd == list_first_entry(path, struct coresight_device,
-					   path_link)) {
-			coresight_disable_sink(cd);
-		} else if (list_is_last(&cd->path_link, path)) {
-			/*
-			 * The source has already been stopped, no need
-			 * to do it again here.
-			 */
-			;
-		} else {
-			coresight_disable_link(cd);
+	int i;
+	bool found = false;
+	struct coresight_node *node;
+	struct coresight_connection *conn;
+
+	/* An activated sink has been found.  Enqueue the element */
+	if ((csdev->type == CORESIGHT_DEV_TYPE_SINK ||
+	     csdev->type == CORESIGHT_DEV_TYPE_LINKSINK) && csdev->activated)
+		goto out;
+
+	/* Not a sink - recursively explore each port found on this element */
+	for (i = 0; i < csdev->nr_outport; i++) {
+		conn = &csdev->conns[i];
+		if (_coresight_build_path(conn->child_dev, path) == 0) {
+			found = true;
+			break;
 		}
 	}
 
+	if (!found)
+		return -ENODEV;
+
+out:
+	/*
+	 * A path from this element to a sink has been found.  The elements
+	 * leading to the sink are already enqueued, all that is left to do
+	 * is add a node for this element.
+	 */
+	node = kzalloc(sizeof(struct coresight_node), GFP_KERNEL);
+	if (!node)
+		return -ENOMEM;
+
+	node->csdev = csdev;
+	list_add(&node->link, path);
+
 	return 0;
 }
 
-static int coresight_build_paths(struct coresight_device *csdev,
-				 struct list_head *path,
-				 bool enable)
+struct list_head *coresight_build_path(struct coresight_device *csdev)
 {
-	int i, ret = -EINVAL;
-	struct coresight_connection *conn;
+	struct list_head *path;
 
-	list_add(&csdev->path_link, path);
+	path = kzalloc(sizeof(struct list_head), GFP_KERNEL);
+	if (!path)
+		return NULL;
 
-	if ((csdev->type == CORESIGHT_DEV_TYPE_SINK ||
-	    csdev->type == CORESIGHT_DEV_TYPE_LINKSINK) &&
-	    csdev->activated) {
-		if (enable)
-			ret = coresight_enable_path(path);
-		else
-			ret = coresight_disable_path(path);
-	} else {
-		for (i = 0; i < csdev->nr_outport; i++) {
-			conn = &csdev->conns[i];
-			if (coresight_build_paths(conn->child_dev,
-						    path, enable) == 0)
-				ret = 0;
-		}
+	INIT_LIST_HEAD(path);
+
+	if (_coresight_build_path(csdev, path)) {
+		kfree(path);
+		path = NULL;
 	}
 
-	if (list_first_entry(path, struct coresight_device, path_link) != csdev)
-		dev_err(&csdev->dev, "wrong device in %s\n", __func__);
+	return path;
+}
 
-	list_del(&csdev->path_link);
+/**
+ * coresight_release_path - release a previously built path.
+ * @path:	the path to release.
+ *
+ * Go through all the elements of a path and 1) removed it from the list and
+ * 2) free the memory allocated for each node.
+ */
+void coresight_release_path(struct list_head *path)
+{
+	struct coresight_node *nd, *next;
 
-	return ret;
+	list_for_each_entry_safe(nd, next, path, link) {
+		list_del(&nd->link);
+		kfree(nd);
+	}
+
+	kfree(path);
+	path = NULL;
 }
 
 int coresight_enable(struct coresight_device *csdev)
 {
 	int ret = 0;
-	LIST_HEAD(path);
+	int cpu;
+	struct list_head *path;
 
 	mutex_lock(&coresight_mutex);
 	if (csdev->type != CORESIGHT_DEV_TYPE_SOURCE) {
@@ -348,22 +426,47 @@ int coresight_enable(struct coresight_device *csdev)
 	if (csdev->enable)
 		goto out;
 
-	if (coresight_build_paths(csdev, &path, true)) {
-		dev_err(&csdev->dev, "building path(s) failed\n");
+	path = coresight_build_path(csdev);
+	if (!path) {
+		pr_err("building path(s) failed\n");
 		goto out;
 	}
 
-	if (coresight_enable_source(csdev))
-		dev_err(&csdev->dev, "source enable failed\n");
+	ret = coresight_enable_path(path);
+	if (ret)
+		goto err_path;
+
+	ret = coresight_enable_source(csdev);
+	if (ret)
+		goto err_source;
+
+	/*
+	 * When working from sysFS it is important to keep track
+	 * of the paths that were created so that they can be
+	 * undone in 'coresight_disable()'.  Since there can only
+	 * be a single session per tracer (when working from sysFS)
+	 * a per-cpu variable will do just fine.
+	 */
+	cpu = source_ops(csdev)->cpu_id(csdev);
+	per_cpu(sysfs_path, cpu) = path;
+
 out:
 	mutex_unlock(&coresight_mutex);
 	return ret;
+
+err_source:
+	coresight_disable_path(path);
+
+err_path:
+	coresight_release_path(path);
+	goto out;
 }
 EXPORT_SYMBOL_GPL(coresight_enable);
 
 void coresight_disable(struct coresight_device *csdev)
 {
-	LIST_HEAD(path);
+	int cpu;
+	struct list_head *path;
 
 	mutex_lock(&coresight_mutex);
 	if (csdev->type != CORESIGHT_DEV_TYPE_SOURCE) {
@@ -373,9 +476,12 @@ void coresight_disable(struct coresight_device *csdev)
 	if (!csdev->enable)
 		goto out;
 
+	cpu = source_ops(csdev)->cpu_id(csdev);
+	path = per_cpu(sysfs_path, cpu);
 	coresight_disable_source(csdev);
-	if (coresight_build_paths(csdev, &path, false))
-		dev_err(&csdev->dev, "releasing path(s) failed\n");
+	coresight_disable_path(path);
+	coresight_release_path(path);
+	per_cpu(sysfs_path, cpu) = NULL;
 
 out:
 	mutex_unlock(&coresight_mutex);
diff --git a/include/linux/coresight.h b/include/linux/coresight.h
index bf62b265bf52..851ecb22397e 100644
--- a/include/linux/coresight.h
+++ b/include/linux/coresight.h
@@ -152,7 +152,6 @@ struct coresight_connection {
 		by @coresight_ops.
  * @dev:	The device entity associated to this component.
  * @refcnt:	keep track of what is in use.
- * @path_link:	link of current component into the path being enabled.
  * @orphan:	true if the component has connections that haven't been linked.
  * @enable:	'true' if component is currently part of an active path.
  * @activated:	'true' only if a _sink_ has been activated.  A sink can be
@@ -168,7 +167,6 @@ struct coresight_device {
 	const struct coresight_ops *ops;
 	struct device dev;
 	atomic_t *refcnt;
-	struct list_head path_link;
 	bool orphan;
 	bool enable;	/* true only if configured as part of a path */
 	bool activated;	/* true only if a sink is part of a path */
-- 
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] | [standalone]

Page 2 of 2 — ← Prev page 1 [2]

Back to top | Article view | linux.kernel

csiph-web