linux-arm-kernel.lists.infradead.org archive mirror
 help / color / mirror / Atom feed
From: James Clark <james.clark@linaro.org>
To: Suzuki K Poulose <suzuki.poulose@arm.com>,
	 Mike Leach <mike.leach@linaro.org>,
	 Alexander Shishkin <alexander.shishkin@linux.intel.com>,
	 Jonathan Corbet <corbet@lwn.net>, Leo Yan <leo.yan@arm.com>
Cc: coresight@lists.linaro.org, linux-arm-kernel@lists.infradead.org,
	 linux-kernel@vger.kernel.org, linux-doc@vger.kernel.org,
	 James Clark <james.clark@linaro.org>
Subject: [PATCH v2 5/6] coresight: Add format attribute for setting the timestamp interval
Date: Thu, 14 Aug 2025 11:49:56 +0100	[thread overview]
Message-ID: <20250814-james-cs-syncfreq-v2-5-c76fcb87696d@linaro.org> (raw)
In-Reply-To: <20250814-james-cs-syncfreq-v2-0-c76fcb87696d@linaro.org>

Timestamps are currently emitted at the maximum rate possible, which is
much too frequent for most use cases. Add an attribute to be able to set
the interval. Granular control is not required, so save space in the
config by interpreting it as 2 ^ ts_interval. And then 4 bits (0 - 15) is
enough to set the interval to be larger than the existing SYNC timestamp
interval.

No sysfs file is needed for this attribute because counter generated
timestamps are only configured for Perf mode.

Only show this attribute for ETM4x because timestamps aren't configured
in the same way for ETM3x. The attribute is only ever read in
coresight-etm4x-core.c.

Signed-off-by: James Clark <james.clark@linaro.org>
---
 drivers/hwtracing/coresight/coresight-etm-perf.c   | 13 ++++++++++++-
 drivers/hwtracing/coresight/coresight-etm4x-core.c | 21 ++++++++++++---------
 drivers/hwtracing/coresight/coresight-etm4x.h      |  6 ++++++
 3 files changed, 30 insertions(+), 10 deletions(-)

diff --git a/drivers/hwtracing/coresight/coresight-etm-perf.c b/drivers/hwtracing/coresight/coresight-etm-perf.c
index f677c08233ba..af937bbb933c 100644
--- a/drivers/hwtracing/coresight/coresight-etm-perf.c
+++ b/drivers/hwtracing/coresight/coresight-etm-perf.c
@@ -25,6 +25,11 @@
 #include "coresight-syscfg.h"
 #include "coresight-trace-id.h"
 
+#if IS_ENABLED(CONFIG_CORESIGHT_SOURCE_ETM4X)
+#include <linux/perf/arm_pmu.h>
+#include "coresight-etm4x.h"
+#endif
+
 static struct pmu etm_pmu;
 static bool etm_perf_up;
 
@@ -69,7 +74,10 @@ PMU_FORMAT_ATTR(sinkid,		"config2:0-31");
 /* config ID - set if a system configuration is selected */
 PMU_FORMAT_ATTR(configid,	"config2:32-63");
 PMU_FORMAT_ATTR(cc_threshold,	"config3:0-11");
-
+#if IS_ENABLED(CONFIG_CORESIGHT_SOURCE_ETM4X)
+/* Interval = (2 ^ ts_level) */
+GEN_PMU_FORMAT_ATTR(ts_level);
+#endif
 
 /*
  * contextid always traces the "PID".  The PID is in CONTEXTIDR_EL1
@@ -103,6 +111,9 @@ static struct attribute *etm_config_formats_attr[] = {
 	&format_attr_configid.attr,
 	&format_attr_branch_broadcast.attr,
 	&format_attr_cc_threshold.attr,
+#if IS_ENABLED(CONFIG_CORESIGHT_SOURCE_ETM4X)
+	&format_attr_ts_level.attr,
+#endif
 	NULL,
 };
 
diff --git a/drivers/hwtracing/coresight/coresight-etm4x-core.c b/drivers/hwtracing/coresight/coresight-etm4x-core.c
index 1a2d02bdcb88..42277c201d4f 100644
--- a/drivers/hwtracing/coresight/coresight-etm4x-core.c
+++ b/drivers/hwtracing/coresight/coresight-etm4x-core.c
@@ -28,6 +28,7 @@
 #include <linux/amba/bus.h>
 #include <linux/seq_file.h>
 #include <linux/uaccess.h>
+#include <linux/perf/arm_pmu.h>
 #include <linux/perf_event.h>
 #include <linux/platform_device.h>
 #include <linux/pm_runtime.h>
@@ -615,7 +616,7 @@ static void etm4_enable_hw_smp_call(void *info)
  *  +--------------+
  *         |
  *  +------v-------+
- *  | Counter x    |   (reload to 1 on underflow)
+ *  | Counter x    |   (reload to 2 ^ ts_level on underflow)
  *  +--------------+
  *         |
  *  +------v--------------+
@@ -626,11 +627,17 @@ static void etm4_enable_hw_smp_call(void *info)
  *  | Timestamp Generator  |  (timestamp on resource y)
  *  +----------------------+
  */
-static int etm4_config_timestamp_event(struct etmv4_drvdata *drvdata)
+static int etm4_config_timestamp_event(struct etmv4_drvdata *drvdata,
+				       struct perf_event_attr *attr)
 {
 	int ctridx;
 	int rselector;
 	struct etmv4_config *config = &drvdata->config;
+	u8 ts_level = ATTR_CFG_GET_FLD(attr, ts_level);
+
+	/* Disable when ts_level == MAX */
+	if (ts_level == FIELD_GET(ATTR_CFG_FLD_ts_level_MASK, UINT_MAX))
+		return 0;
 
 	/* No point in trying if we don't have at least one counter */
 	if (!drvdata->nr_cntr)
@@ -666,12 +673,8 @@ static int etm4_config_timestamp_event(struct etmv4_drvdata *drvdata)
 		return -ENOSPC;
 	}
 
-	/*
-	 * Initialise original and reload counter value to the smallest
-	 * possible value in order to get as much precision as we can.
-	 */
-	config->cntr_val[ctridx] = 1;
-	config->cntrldvr[ctridx] = 1;
+	/* Initialise original and reload counter value. */
+	config->cntr_val[ctridx] = config->cntrldvr[ctridx] = 1 << ts_level;
 
 	/*
 	 * Trace Counter Control Register TRCCNTCTLRn
@@ -761,7 +764,7 @@ static int etm4_parse_event_config(struct coresight_device *csdev,
 		 * order to correlate instructions executed on different CPUs
 		 * (CPU-wide trace scenarios).
 		 */
-		ret = etm4_config_timestamp_event(drvdata);
+		ret = etm4_config_timestamp_event(drvdata, attr);
 
 		/*
 		 * No need to go further if timestamp intervals can't
diff --git a/drivers/hwtracing/coresight/coresight-etm4x.h b/drivers/hwtracing/coresight/coresight-etm4x.h
index aaa6633b2d67..54558de158fa 100644
--- a/drivers/hwtracing/coresight/coresight-etm4x.h
+++ b/drivers/hwtracing/coresight/coresight-etm4x.h
@@ -598,6 +598,12 @@
 #define ETM_CNTR_MAX_VAL		0xFFFF
 #define ETM_TRACEID_MASK		0x3f
 
+#define ATTR_CFG_FLD_ts_level_CFG	config3
+#define ATTR_CFG_FLD_ts_level_LO	12
+#define ATTR_CFG_FLD_ts_level_HI	15
+#define ATTR_CFG_FLD_ts_level_MASK	GENMASK(ATTR_CFG_FLD_ts_level_HI, \
+						ATTR_CFG_FLD_ts_level_LO)
+
 /* ETMv4 programming modes */
 #define ETM_MODE_EXCLUDE		BIT(0)
 #define ETM_MODE_LOAD			BIT(1)

-- 
2.34.1



  parent reply	other threads:[~2025-08-14 15:37 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-08-14 10:49 [PATCH v2 0/6] coresight: Add format attribute for setting the timestamp interval James Clark
2025-08-14 10:49 ` [PATCH v2 1/6] coresight: Change syncfreq to be a u8 James Clark
2025-08-14 10:49 ` [PATCH v2 2/6] coresight: Fix holes in struct etmv4_config James Clark
2025-08-14 10:49 ` [PATCH v2 3/6] coresight: Repack struct etmv4_drvdata James Clark
2025-08-14 10:49 ` [PATCH v2 4/6] coresight: Refactor etm4_config_timestamp_event() James Clark
2025-08-14 10:49 ` James Clark [this message]
2025-08-14 10:49 ` [PATCH v2 6/6] coresight: docs: Document etm4x ts_interval James Clark

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20250814-james-cs-syncfreq-v2-5-c76fcb87696d@linaro.org \
    --to=james.clark@linaro.org \
    --cc=alexander.shishkin@linux.intel.com \
    --cc=corbet@lwn.net \
    --cc=coresight@lists.linaro.org \
    --cc=leo.yan@arm.com \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-doc@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mike.leach@linaro.org \
    --cc=suzuki.poulose@arm.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).