* [PATCH v5 0/7] coresight: Pulse an ETM external output for a PMU event
@ 2026-11-20 15:21 Amir Ayupov
2026-11-20 15:21 ` [PATCH v5 1/7] coresight: etm4x: Fix event count decoding Amir Ayupov
` (6 more replies)
0 siblings, 7 replies; 14+ messages in thread
From: Amir Ayupov @ 2026-11-20 15:21 UTC (permalink / raw)
To: James Clark, Leo Yan, Suzuki K Poulose, Mike Leach,
Alexander Shishkin
Cc: Will Deacon, Mark Rutland, Greg Kroah-Hartman, Jonathan Corbet,
Mathieu Poirier, Randy Dunlap, Shuah Khan, Daniel Pereira,
coresight, linux-arm-kernel, linux-perf-users, linux-doc,
linux-kernel
Add an ETMv4 configuration, pmu_pulse, that drives trace unit external
output 1 on every traced cycle, so a PMU event counting it (TRCEXTOUT1)
can end a fixed-length ETM window, for example with aux-action=pause.
The feature only loads on an ETE whose PMU counts its external outputs.
Output 1 is fixed by the configuration; the perf command selects the
matching TRCEXTOUT1 consumer separately. Output 0 is avoided because it
is the trace trigger, and a detected trigger stops TRBE capture. Each
window includes the counted trace cycles plus pause interrupt latency.
Patch 1 is Leo's fix for TRCIDR0.NUMEVENT decoding, which patch 6 relies
on for the number of external outputs. An ARM PMU maintainer Ack is
requested on patch 5 so the series can be taken through the CoreSight
tree.
Changes in v5:
- Compare the class and capability fields with FIELD_GET() (Suzuki). The
class check stays an intersection: built-in features name a single
class, while ETMv4 devices register with both SRC_ALL and SRC_ETM4.
- Move both ARMV8_PMUV3_MAX_COMMON_EVENTS and
ARMV8_PMUV3_EXT_COMMON_EVENT_BASE to linux/perf/arm_pmuv3.h and include
it from linux/perf/arm_pmu.h (Suzuki, James). Add Will and Mark for an
Ack on that patch.
- Rename the capability to CS_CFG_MATCH_CAP_ETE_TRCEXTOUT, as it only
applies to ETE (Suzuki).
- Drop caps.trcextout and pass the probe-time result to
etm4_cscfg_register() instead (Suzuki). The check reads PMCEID on the
trace unit's CPU, so it stays in etm4_init_arch_data(), which runs
there, while etm4_cscfg_register() may run on any CPU.
- Add James's Reviewed-by to all but patch 2, and Leo's Tested-by to all.
- Link to v4: https://lore.kernel.org/r/cover.1792989860.git.aaupov@fb.com
Changes in v4:
- Fix the arm64 build: v3 defined pmuv3_event_implemented() in
linux/perf/arm_pmuv3.h after including asm/arm_pmuv3.h, which broke
files that include asm/arm_pmuv3.h directly, such as
arch/arm64/kernel/cpufeature.c. Move it to the arm64 asm/arm_pmuv3.h,
after the helpers it uses.
Breakage reported by kernel test robot.
- Link to v3: https://lore.kernel.org/r/cover.1792907332.git.aaupov@fb.com
Changes in v3:
- Rebase onto coresight next, which moves the capability fields into
drvdata->caps (Leo).
- Carry Leo's "coresight: etm4x: Fix event count decoding" (Leo).
- Add Leo's Reviewed-by to "Fail to enable a configuration not loaded on
the device".
- Define CS_CFG_MATCH_CLASS_MASK and CS_CFG_MATCH_CAP_MASK with
GENMASK_U32() and split the match check into early returns (Leo). The
class check stays an intersection, as ETMv4 devices register with
SRC_ALL | SRC_ETM4 while the built-in features only set SRC_ETM4.
- Remove the unused CS_CFG_MATCH_INST_ANY in a separate patch (Leo).
- Replace the per-output CS_CFG_MATCH_CAP_PMU_EXTOUT(n) bits with a single
CS_CFG_MATCH_CAP_TRCEXTOUT, set on ETE only (Leo, James).
- Check the PMU events through a new pmuv3_event_implemented() helper
instead of open-coding PMCEID0_EL0 in etm4x (James).
- Use PMU-agnostic event names in the documentation example (Daniel).
- Add Daniel's Reviewed-by to "Add pmu_pulse/extout_pulse".
- Link to v2: https://lore.kernel.org/r/20261012230527.750567-1-aaupov@fb.com
Amir Ayupov (6):
coresight: syscfg: Fail to enable a configuration not loaded on the
device
coresight: syscfg: Match feature capabilities separately from device
class
coresight: syscfg: Remove unused CS_CFG_MATCH_INST_ANY
perf: arm_pmuv3: Add pmuv3_event_implemented()
coresight: etm4x: Report whether the PMU counts the external outputs
coresight: Add pmu_pulse/extout_pulse
Leo Yan (1):
coresight: etm4x: Fix event count decoding
.../trace/coresight/coresight-config.rst | 5 +
.../trace/coresight/coresight-perf.rst | 15 +++
arch/arm64/include/asm/arm_pmuv3.h | 20 ++++
drivers/hwtracing/coresight/Makefile | 1 +
.../coresight/coresight-cfg-pmu-pulse.c | 55 +++++++++++
.../coresight/coresight-cfg-preload.c | 2 +
.../coresight/coresight-cfg-preload.h | 2 +
.../hwtracing/coresight/coresight-config.h | 11 ++-
.../hwtracing/coresight/coresight-etm4x-cfg.c | 7 +-
.../hwtracing/coresight/coresight-etm4x-cfg.h | 2 +-
.../coresight/coresight-etm4x-core.c | 42 ++++++++-
.../coresight/coresight-etm4x-sysfs.c | 29 +++---
.../coresight/coresight-syscfg-configfs.c | 3 +
.../hwtracing/coresight/coresight-syscfg.c | 93 ++++++++++++-------
include/linux/perf/arm_pmu.h | 3 +-
include/linux/perf/arm_pmuv3.h | 8 ++
16 files changed, 239 insertions(+), 59 deletions(-)
create mode 100644 drivers/hwtracing/coresight/coresight-cfg-pmu-pulse.c
base-commit: 343cfcbb3ecfdb80aa576a428cfa3b65ad783842
--
2.52.1
^ permalink raw reply [flat|nested] 14+ messages in thread
* [PATCH v5 1/7] coresight: etm4x: Fix event count decoding
2026-11-20 15:21 [PATCH v5 0/7] coresight: Pulse an ETM external output for a PMU event Amir Ayupov
@ 2026-11-20 15:21 ` Amir Ayupov
2026-11-20 15:30 ` sashiko-bot
2026-11-20 15:21 ` [PATCH v5 2/7] coresight: syscfg: Fail to enable a configuration not loaded on the device Amir Ayupov
` (5 subsequent siblings)
6 siblings, 1 reply; 14+ messages in thread
From: Amir Ayupov @ 2026-11-20 15:21 UTC (permalink / raw)
To: James Clark, Leo Yan, Suzuki K Poulose, Mike Leach,
Alexander Shishkin
Cc: Will Deacon, Mark Rutland, Greg Kroah-Hartman, Jonathan Corbet,
Mathieu Poirier, Randy Dunlap, Shuah Khan, Daniel Pereira,
coresight, linux-arm-kernel, linux-perf-users, linux-doc,
linux-kernel
From: Leo Yan <leo.yan@arm.com>
TRCIDR0.NUMEVENT normally encodes the number of events minus one.
From ETMv4.3 onwards, including ETE, a zero TRCIDR4.NUMRSPAIR instead
means that no events are implemented and NUMEVENT is zero. The sysfs
event controls currently interpret this combination as one event and
allow writes to unsupported event fields.
Decode NUMEVENT after NUMRSPAIR and store the actual number of events
in nr_event. Use the decoded resource count so that older ETM versions,
where a zero NUMRSPAIR represents one resource pair, still report at
least one event. Update both sysfs event controls to use the actual
count and clear their event fields when no events are implemented.
Also correct the instruction event enable mask for the single-event
case to use INSTEN[0]. The existing mask uses INSTEN[1], preventing
event 0 from being enabled.
This change depends on the recent driver refactoring. With no reported
failures, it is not intended for stable backporting.
Signed-off-by: Leo Yan <leo.yan@arm.com>
Reviewed-by: James Clark <james.clark@linaro.org>
Tested-by: Leo Yan <leo.yan@arm.com>
Signed-off-by: Amir Ayupov <aaupov@fb.com>
---
.../coresight/coresight-etm4x-core.c | 13 +++++++--
.../coresight/coresight-etm4x-sysfs.c | 29 ++++++++++---------
2 files changed, 27 insertions(+), 15 deletions(-)
diff --git a/drivers/hwtracing/coresight/coresight-etm4x-core.c b/drivers/hwtracing/coresight/coresight-etm4x-core.c
index e21be6468e9544..af017bd8a54773 100644
--- a/drivers/hwtracing/coresight/coresight-etm4x-core.c
+++ b/drivers/hwtracing/coresight/coresight-etm4x-core.c
@@ -1451,8 +1451,6 @@ static void etm4_init_arch_data(void *info)
caps->trccci = !!(etmidr0 & TRCIDR0_TRCCCI);
/* RETSTACK, bit[9] Return stack bit */
caps->retstack = !!(etmidr0 & TRCIDR0_RETSTACK);
- /* NUMEVENT, bits[11:10] Number of events field */
- caps->nr_event = FIELD_GET(TRCIDR0_NUMEVENT_MASK, etmidr0);
/* QSUPP, bits[16:15] Q element support field */
caps->q_support = FIELD_GET(TRCIDR0_QSUPP_MASK, etmidr0);
if (caps->q_support)
@@ -1519,6 +1517,17 @@ static void etm4_init_arch_data(void *info)
caps->nr_resource = FIELD_GET(TRCIDR4_NUMRSPAIR_MASK, etmidr4);
if ((drvdata->arch < ETM_ARCH_V4_3) || (caps->nr_resource > 0))
caps->nr_resource += 1;
+
+ /*
+ * NUMEVENT encodes the number of events minus one. From ETMv4.3,
+ * including ETE, NUMRSPAIR == 0 means no events are implemented.
+ * nr_resource already accounts for the pre-ETMv4.3 encoding.
+ */
+ if (caps->nr_resource)
+ caps->nr_event = FIELD_GET(TRCIDR0_NUMEVENT_MASK, etmidr0) + 1;
+ else
+ caps->nr_event = 0;
+
/*
* NUMSSCC, bits[23:20] the number of single-shot
* comparator control for tracing. Read any status regs as these
diff --git a/drivers/hwtracing/coresight/coresight-etm4x-sysfs.c b/drivers/hwtracing/coresight/coresight-etm4x-sysfs.c
index db9e603f97ba82..fb38f0102290a5 100644
--- a/drivers/hwtracing/coresight/coresight-etm4x-sysfs.c
+++ b/drivers/hwtracing/coresight/coresight-etm4x-sysfs.c
@@ -513,19 +513,22 @@ static ssize_t event_store(struct device *dev,
raw_spin_lock(&drvdata->spinlock);
switch (caps->nr_event) {
- case 0x0:
+ case 0:
+ sysfs_config->eventctrl0 = 0;
+ break;
+ case 1:
/* EVENT0, bits[7:0] */
sysfs_config->eventctrl0 = val & 0xFF;
break;
- case 0x1:
- /* EVENT1, bits[15:8] */
+ case 2:
+ /* EVENT1, bits[15:8] */
sysfs_config->eventctrl0 = val & 0xFFFF;
break;
- case 0x2:
+ case 3:
/* EVENT2, bits[23:16] */
sysfs_config->eventctrl0 = val & 0xFFFFFF;
break;
- case 0x3:
+ case 4:
/* EVENT3, bits[31:24] */
sysfs_config->eventctrl0 = val;
break;
@@ -565,22 +568,22 @@ static ssize_t event_instren_store(struct device *dev,
/* start by clearing all instruction event enable bits */
sysfs_config->eventctrl1 &= ~TRCEVENTCTL1R_INSTEN_MASK;
switch (caps->nr_event) {
- case 0x0:
- /* generate Event element for event 1 */
- sysfs_config->eventctrl1 |= val & TRCEVENTCTL1R_INSTEN_1;
+ case 1:
+ /* generate Event element for event 0 */
+ sysfs_config->eventctrl1 |= val & TRCEVENTCTL1R_INSTEN_0;
break;
- case 0x1:
- /* generate Event element for event 1 and 2 */
+ case 2:
+ /* generate Event element for events 0 and 1 */
sysfs_config->eventctrl1 |= val & (TRCEVENTCTL1R_INSTEN_0 |
TRCEVENTCTL1R_INSTEN_1);
break;
- case 0x2:
- /* generate Event element for event 1, 2 and 3 */
+ case 3:
+ /* generate Event element for events 0, 1 and 2 */
sysfs_config->eventctrl1 |= val & (TRCEVENTCTL1R_INSTEN_0 |
TRCEVENTCTL1R_INSTEN_1 |
TRCEVENTCTL1R_INSTEN_2);
break;
- case 0x3:
+ case 4:
/* generate Event element for all 4 events */
sysfs_config->eventctrl1 |= val & (TRCEVENTCTL1R_INSTEN_0 |
TRCEVENTCTL1R_INSTEN_1 |
--
2.52.1
^ permalink raw reply related [flat|nested] 14+ messages in thread
* [PATCH v5 2/7] coresight: syscfg: Fail to enable a configuration not loaded on the device
2026-11-20 15:21 [PATCH v5 0/7] coresight: Pulse an ETM external output for a PMU event Amir Ayupov
2026-11-20 15:21 ` [PATCH v5 1/7] coresight: etm4x: Fix event count decoding Amir Ayupov
@ 2026-11-20 15:21 ` Amir Ayupov
2026-11-20 15:36 ` sashiko-bot
2026-11-20 15:21 ` [PATCH v5 3/7] coresight: syscfg: Match feature capabilities separately from device class Amir Ayupov
` (4 subsequent siblings)
6 siblings, 1 reply; 14+ messages in thread
From: Amir Ayupov @ 2026-11-20 15:21 UTC (permalink / raw)
To: James Clark, Leo Yan, Suzuki K Poulose, Mike Leach,
Alexander Shishkin
Cc: Will Deacon, Mark Rutland, Greg Kroah-Hartman, Jonathan Corbet,
Mathieu Poirier, Randy Dunlap, Shuah Khan, Daniel Pereira,
coresight, linux-arm-kernel, linux-perf-users, linux-doc,
linux-kernel
cscfg_csdev_enable_active_config() looks up the selected configuration in
the device's loaded configurations, and returns success when it is not
there. A configuration is only loaded on devices that match at least one
of its features, so a perf session or sysfs enable that selects it on any
other device runs without the configuration, with no error.
Return -EINVAL instead, so enabling the source fails on that device. The
function is only called with a non-zero configuration hash, so sources
enabled without a configuration are unaffected.
Fixes: f8cce2ff3c04 ("coresight: syscfg: Add API to activate and enable configurations")
Assisted-by: Devmate:Opus-5.5
Tested-by: Leo Yan <leo.yan@arm.com>
Reviewed-by: Leo Yan <leo.yan@arm.com>
Signed-off-by: Amir Ayupov <aaupov@fb.com>
---
.../hwtracing/coresight/coresight-syscfg.c | 53 ++++++++++---------
1 file changed, 28 insertions(+), 25 deletions(-)
diff --git a/drivers/hwtracing/coresight/coresight-syscfg.c b/drivers/hwtracing/coresight/coresight-syscfg.c
index 2bfdd7b45e49c8..a816c22c87bc87 100644
--- a/drivers/hwtracing/coresight/coresight-syscfg.c
+++ b/drivers/hwtracing/coresight/coresight-syscfg.c
@@ -1088,36 +1088,39 @@ int cscfg_csdev_enable_active_config(struct coresight_device *csdev,
raw_spin_unlock_irqrestore(&csdev->cscfg_csdev_lock, flags);
/*
- * If found, attempt to enable
+ * The configuration is active but not loaded on this device, because
+ * none of its features match the device. Fail rather than trace
+ * without it.
*/
- if (config_csdev_active) {
+ if (!config_csdev_active)
+ return -EINVAL;
+
+ /*
+ * Call the generic routine that will program up the internal
+ * driver structures prior to programming up the hardware.
+ * This routine takes the driver spinlock saved in the configs.
+ */
+ err = cscfg_csdev_enable_config(config_csdev_active, preset);
+ if (!err) {
/*
- * Call the generic routine that will program up the internal
- * driver structures prior to programming up the hardware.
- * This routine takes the driver spinlock saved in the configs.
+ * Successful programming. Check the active_cscfg_ctxt
+ * pointer to ensure no pre-emption disabled it via
+ * cscfg_csdev_disable_active_config() before
+ * we could start.
+ *
+ * Set enabled if OK, err if not.
*/
- err = cscfg_csdev_enable_config(config_csdev_active, preset);
- if (!err) {
- /*
- * Successful programming. Check the active_cscfg_ctxt
- * pointer to ensure no pre-emption disabled it via
- * cscfg_csdev_disable_active_config() before
- * we could start.
- *
- * Set enabled if OK, err if not.
- */
- raw_spin_lock_irqsave(&csdev->cscfg_csdev_lock, flags);
- if (csdev->active_cscfg_ctxt)
- config_csdev_active->enabled = true;
- else
- err = -EBUSY;
- raw_spin_unlock_irqrestore(&csdev->cscfg_csdev_lock, flags);
- }
-
- if (err)
- cscfg_config_desc_put(config_desc);
+ raw_spin_lock_irqsave(&csdev->cscfg_csdev_lock, flags);
+ if (csdev->active_cscfg_ctxt)
+ config_csdev_active->enabled = true;
+ else
+ err = -EBUSY;
+ raw_spin_unlock_irqrestore(&csdev->cscfg_csdev_lock, flags);
}
+ if (err)
+ cscfg_config_desc_put(config_desc);
+
return err;
}
EXPORT_SYMBOL_GPL(cscfg_csdev_enable_active_config);
--
2.52.1
^ permalink raw reply related [flat|nested] 14+ messages in thread
* [PATCH v5 3/7] coresight: syscfg: Match feature capabilities separately from device class
2026-11-20 15:21 [PATCH v5 0/7] coresight: Pulse an ETM external output for a PMU event Amir Ayupov
2026-11-20 15:21 ` [PATCH v5 1/7] coresight: etm4x: Fix event count decoding Amir Ayupov
2026-11-20 15:21 ` [PATCH v5 2/7] coresight: syscfg: Fail to enable a configuration not loaded on the device Amir Ayupov
@ 2026-11-20 15:21 ` Amir Ayupov
2026-11-20 15:29 ` sashiko-bot
2026-11-20 15:21 ` [PATCH v5 4/7] coresight: syscfg: Remove unused CS_CFG_MATCH_INST_ANY Amir Ayupov
` (3 subsequent siblings)
6 siblings, 1 reply; 14+ messages in thread
From: Amir Ayupov @ 2026-11-20 15:21 UTC (permalink / raw)
To: James Clark, Leo Yan, Suzuki K Poulose, Mike Leach,
Alexander Shishkin
Cc: Will Deacon, Mark Rutland, Greg Kroah-Hartman, Jonathan Corbet,
Mathieu Poirier, Randy Dunlap, Shuah Khan, Daniel Pereira,
coresight, linux-arm-kernel, linux-perf-users, linux-doc,
linux-kernel
A feature loads on a device when the two share any match flag. That works
for device classes, where a feature wants any one of them, but not for
capabilities a device may or may not have: a feature that sets a class
and a capability would still load on a device of that class without the
capability.
Split the match flags into CS_CFG_MATCH_CLASS_MASK for device classes and
CS_CFG_MATCH_CAP_MASK for capabilities. A feature now loads on a device
that shares one of its classes and has all of its capabilities. Check
this once in cscfg_load_feat_csdev(), which both the feature and the
device registration paths call. No capability flags are defined yet, so
matching is unchanged.
Assisted-by: Devmate:Opus-5.5
Reviewed-by: James Clark <james.clark@linaro.org>
Tested-by: Leo Yan <leo.yan@arm.com>
Signed-off-by: Amir Ayupov <aaupov@fb.com>
---
.../hwtracing/coresight/coresight-config.h | 8 ++++
.../hwtracing/coresight/coresight-syscfg.c | 40 ++++++++++++++-----
2 files changed, 37 insertions(+), 11 deletions(-)
diff --git a/drivers/hwtracing/coresight/coresight-config.h b/drivers/hwtracing/coresight/coresight-config.h
index ae6295e3081319..f04e73b91d9a63 100644
--- a/drivers/hwtracing/coresight/coresight-config.h
+++ b/drivers/hwtracing/coresight/coresight-config.h
@@ -7,6 +7,7 @@
#ifndef _CORESIGHT_CORESIGHT_CONFIG_H
#define _CORESIGHT_CORESIGHT_CONFIG_H
+#include <linux/bits.h>
#include <linux/cleanup.h>
#include <linux/coresight.h>
#include <linux/types.h>
@@ -30,6 +31,13 @@
*/
#define CS_CFG_MATCH_CLASS_SRC_ALL 0x0001 /* match any source */
#define CS_CFG_MATCH_CLASS_SRC_ETM4 0x0002 /* match any ETMv4 device */
+#define CS_CFG_MATCH_CLASS_MASK GENMASK_U32(7, 0)
+
+/*
+ * flags defining device capabilities a feature can require. A feature loads on
+ * a device that matches one of its classes and has all of its capabilities.
+ */
+#define CS_CFG_MATCH_CAP_MASK GENMASK_U32(15, 8)
/* flags defining device instance matching - used in config match desc data. */
#define CS_CFG_MATCH_INST_ANY 0x80000000 /* any instance of a class */
diff --git a/drivers/hwtracing/coresight/coresight-syscfg.c b/drivers/hwtracing/coresight/coresight-syscfg.c
index a816c22c87bc87..65116b09cbc96c 100644
--- a/drivers/hwtracing/coresight/coresight-syscfg.c
+++ b/drivers/hwtracing/coresight/coresight-syscfg.c
@@ -4,6 +4,7 @@
* Author: Mike Leach <mike.leach@linaro.org>
*/
+#include <linux/bitfield.h>
#include <linux/platform_device.h>
#include <linux/slab.h>
@@ -171,8 +172,24 @@ cscfg_alloc_csdev_feat(struct coresight_device *csdev, struct cscfg_feature_desc
return feat_csdev;
}
-/* load one feature into one coresight device */
+static bool cscfg_match_feat_csdev(u32 feat_flags, u32 csdev_flags)
+{
+ /* The device must be in one of the feature's classes */
+ if (!(FIELD_GET(CS_CFG_MATCH_CLASS_MASK, feat_flags) &
+ FIELD_GET(CS_CFG_MATCH_CLASS_MASK, csdev_flags)))
+ return false;
+
+ /* The device must have every capability the feature requires */
+ if (FIELD_GET(CS_CFG_MATCH_CAP_MASK, feat_flags) &
+ ~FIELD_GET(CS_CFG_MATCH_CAP_MASK, csdev_flags))
+ return false;
+
+ return true;
+}
+
+/* Load the feature into the device if it matches, otherwise skip it. */
static int cscfg_load_feat_csdev(struct coresight_device *csdev,
+ u32 match_flags,
struct cscfg_feature_desc *feat_desc,
struct cscfg_csdev_feat_ops *ops)
{
@@ -183,6 +200,9 @@ static int cscfg_load_feat_csdev(struct coresight_device *csdev,
if (!ops->load_feat)
return -EINVAL;
+ if (!cscfg_match_feat_csdev(feat_desc->match_flags, match_flags))
+ return 0;
+
feat_csdev = cscfg_alloc_csdev_feat(csdev, feat_desc);
if (!feat_csdev)
return -ENOMEM;
@@ -212,11 +232,11 @@ static int cscfg_add_feat_to_csdevs(struct cscfg_feature_desc *feat_desc)
int err;
list_for_each_entry(csdev_item, &cscfg_mgr->csdev_desc_list, item) {
- if (csdev_item->match_flags & feat_desc->match_flags) {
- err = cscfg_load_feat_csdev(csdev_item->csdev, feat_desc, &csdev_item->ops);
- if (err)
- return err;
- }
+ err = cscfg_load_feat_csdev(csdev_item->csdev,
+ csdev_item->match_flags, feat_desc,
+ &csdev_item->ops);
+ if (err)
+ return err;
}
return 0;
}
@@ -739,11 +759,9 @@ static int cscfg_add_feats_csdev(struct coresight_device *csdev,
return -EINVAL;
list_for_each_entry(feat_desc, &cscfg_mgr->feat_desc_list, item) {
- if (feat_desc->match_flags & match_flags) {
- err = cscfg_load_feat_csdev(csdev, feat_desc, ops);
- if (err)
- break;
- }
+ err = cscfg_load_feat_csdev(csdev, match_flags, feat_desc, ops);
+ if (err)
+ break;
}
return err;
}
--
2.52.1
^ permalink raw reply related [flat|nested] 14+ messages in thread
* [PATCH v5 4/7] coresight: syscfg: Remove unused CS_CFG_MATCH_INST_ANY
2026-11-20 15:21 [PATCH v5 0/7] coresight: Pulse an ETM external output for a PMU event Amir Ayupov
` (2 preceding siblings ...)
2026-11-20 15:21 ` [PATCH v5 3/7] coresight: syscfg: Match feature capabilities separately from device class Amir Ayupov
@ 2026-11-20 15:21 ` Amir Ayupov
2026-11-20 15:30 ` sashiko-bot
2026-11-20 15:21 ` [PATCH v5 5/7] perf: arm_pmuv3: Add pmuv3_event_implemented() Amir Ayupov
` (2 subsequent siblings)
6 siblings, 1 reply; 14+ messages in thread
From: Amir Ayupov @ 2026-11-20 15:21 UTC (permalink / raw)
To: James Clark, Leo Yan, Suzuki K Poulose, Mike Leach,
Alexander Shishkin
Cc: Will Deacon, Mark Rutland, Greg Kroah-Hartman, Jonathan Corbet,
Mathieu Poirier, Randy Dunlap, Shuah Khan, Daniel Pereira,
coresight, linux-arm-kernel, linux-perf-users, linux-doc,
linux-kernel
CS_CFG_MATCH_INST_ANY has been defined since the system configuration was
introduced but is not used by any configuration, feature or device.
Remove it.
Suggested-by: Leo Yan <leo.yan@arm.com>
Assisted-by: Devmate:Opus-5.5
Reviewed-by: James Clark <james.clark@linaro.org>
Tested-by: Leo Yan <leo.yan@arm.com>
Signed-off-by: Amir Ayupov <aaupov@fb.com>
---
drivers/hwtracing/coresight/coresight-config.h | 3 ---
1 file changed, 3 deletions(-)
diff --git a/drivers/hwtracing/coresight/coresight-config.h b/drivers/hwtracing/coresight/coresight-config.h
index f04e73b91d9a63..220bc54f8e5e1b 100644
--- a/drivers/hwtracing/coresight/coresight-config.h
+++ b/drivers/hwtracing/coresight/coresight-config.h
@@ -39,9 +39,6 @@
*/
#define CS_CFG_MATCH_CAP_MASK GENMASK_U32(15, 8)
-/* flags defining device instance matching - used in config match desc data. */
-#define CS_CFG_MATCH_INST_ANY 0x80000000 /* any instance of a class */
-
/*
* Limit number of presets in a configuration
* This is related to the number of bits (4) we use to select the preset on
--
2.52.1
^ permalink raw reply related [flat|nested] 14+ messages in thread
* [PATCH v5 5/7] perf: arm_pmuv3: Add pmuv3_event_implemented()
2026-11-20 15:21 [PATCH v5 0/7] coresight: Pulse an ETM external output for a PMU event Amir Ayupov
` (3 preceding siblings ...)
2026-11-20 15:21 ` [PATCH v5 4/7] coresight: syscfg: Remove unused CS_CFG_MATCH_INST_ANY Amir Ayupov
@ 2026-11-20 15:21 ` Amir Ayupov
2026-11-20 15:39 ` sashiko-bot
2026-11-20 15:21 ` [PATCH v5 6/7] coresight: etm4x: Report whether the PMU counts the external outputs Amir Ayupov
2026-11-20 15:21 ` [PATCH v5 7/7] coresight: Add pmu_pulse/extout_pulse Amir Ayupov
6 siblings, 1 reply; 14+ messages in thread
From: Amir Ayupov @ 2026-11-20 15:21 UTC (permalink / raw)
To: James Clark, Leo Yan, Suzuki K Poulose, Mike Leach,
Alexander Shishkin, Will Deacon, Mark Rutland
Cc: Greg Kroah-Hartman, Jonathan Corbet, Mathieu Poirier,
Randy Dunlap, Shuah Khan, Daniel Pereira, coresight,
linux-arm-kernel, linux-perf-users, linux-doc, linux-kernel
Drivers outside the PMU driver may need to know whether the local CPU's
PMU implements a common event before relying on it, for example a trace
unit driver checking that the PMU counts its external outputs as
TRCEXTOUTn.
Add pmuv3_event_implemented() to the arm64 asm/arm_pmuv3.h, next to
read_pmceid0() and read_pmceid1(). It reads PMCEID0/1 on the calling CPU:
the first ARMV8_PMUV3_MAX_COMMON_EVENTS events are reported in bits
[31:0], and as many from ARMV8_PMUV3_EXT_COMMON_EVENT_BASE in bits
[63:32] from PMUv3p1.
Move ARMV8_PMUV3_MAX_COMMON_EVENTS and ARMV8_PMUV3_EXT_COMMON_EVENT_BASE
next to the event numbers in linux/perf/arm_pmuv3.h, so that both are
available without CONFIG_ARM_PMU, and include it from
linux/perf/arm_pmu.h, which sizes the PMCEID bitmaps with the former.
Suggested-by: James Clark <james.clark@linaro.org>
Assisted-by: Devmate:Opus-5.5
Reviewed-by: James Clark <james.clark@linaro.org>
Tested-by: Leo Yan <leo.yan@arm.com>
Signed-off-by: Amir Ayupov <aaupov@fb.com>
---
Will, Mark: could you Ack this so Suzuki can take it through the
CoreSight tree with the rest of the series?
arch/arm64/include/asm/arm_pmuv3.h | 20 ++++++++++++++++++++
include/linux/perf/arm_pmu.h | 3 +--
include/linux/perf/arm_pmuv3.h | 8 ++++++++
3 files changed, 29 insertions(+), 2 deletions(-)
diff --git a/arch/arm64/include/asm/arm_pmuv3.h b/arch/arm64/include/asm/arm_pmuv3.h
index 8a777dec8d88ab..f6882771ec7970 100644
--- a/arch/arm64/include/asm/arm_pmuv3.h
+++ b/arch/arm64/include/asm/arm_pmuv3.h
@@ -173,6 +173,26 @@ static inline bool pmuv3_implemented(int pmuver)
pmuver == ID_AA64DFR0_EL1_PMUVer_NI);
}
+/*
+ * Whether this CPU's PMU implements a common event, as reported by
+ * PMCEID{0,1}: events 0x0000-0x003f in bits [31:0], and from PMUv3p1 events
+ * 0x4000-0x403f in bits [63:32], which read as zero before PMUv3p1.
+ * Must run on the CPU whose PMU is queried.
+ */
+static inline bool pmuv3_event_implemented(u16 event)
+{
+ bool ext = event >= ARMV8_PMUV3_EXT_COMMON_EVENT_BASE;
+ u16 id = ext ? event - ARMV8_PMUV3_EXT_COMMON_EVENT_BASE : event;
+ u64 pmceid;
+
+ if (id >= ARMV8_PMUV3_MAX_COMMON_EVENTS ||
+ !pmuv3_implemented(read_pmuver()))
+ return false;
+
+ pmceid = id < 32 ? read_pmceid0() : read_pmceid1();
+ return pmceid & BIT_ULL(id % 32 + (ext ? 32 : 0));
+}
+
static inline bool is_pmuv3p4(int pmuver)
{
return pmuver >= ID_AA64DFR0_EL1_PMUVer_V3P4;
diff --git a/include/linux/perf/arm_pmu.h b/include/linux/perf/arm_pmu.h
index 02d2c7f45b527d..a90867c9774c74 100644
--- a/include/linux/perf/arm_pmu.h
+++ b/include/linux/perf/arm_pmu.h
@@ -10,6 +10,7 @@
#include <linux/interrupt.h>
#include <linux/perf_event.h>
+#include <linux/perf/arm_pmuv3.h>
#include <linux/platform_device.h>
#include <linux/sysfs.h>
#include <asm/cputype.h>
@@ -122,9 +123,7 @@ struct arm_pmu {
bool avoid_pmccntr;
u64 reg_pmmir;
u64 reg_brbidr;
-#define ARMV8_PMUV3_MAX_COMMON_EVENTS 0x40
DECLARE_BITMAP(pmceid_bitmap, ARMV8_PMUV3_MAX_COMMON_EVENTS);
-#define ARMV8_PMUV3_EXT_COMMON_EVENT_BASE 0x4000
DECLARE_BITMAP(pmceid_ext_bitmap, ARMV8_PMUV3_MAX_COMMON_EVENTS);
/* Only to be used by ACPI probing code */
diff --git a/include/linux/perf/arm_pmuv3.h b/include/linux/perf/arm_pmuv3.h
index d698efba28a276..5bab0f952fd59c 100644
--- a/include/linux/perf/arm_pmuv3.h
+++ b/include/linux/perf/arm_pmuv3.h
@@ -78,6 +78,14 @@
#define ARMV8_PMUV3_PERFCTR_STALL_SLOT_FRONTEND 0x003E
#define ARMV8_PMUV3_PERFCTR_STALL_SLOT 0x003F
+/*
+ * PMCEID{0,1} report ARMV8_PMUV3_MAX_COMMON_EVENTS common events from 0 in
+ * bits [31:0], and as many from ARMV8_PMUV3_EXT_COMMON_EVENT_BASE in bits
+ * [63:32] from PMUv3p1.
+ */
+#define ARMV8_PMUV3_MAX_COMMON_EVENTS 0x40
+#define ARMV8_PMUV3_EXT_COMMON_EVENT_BASE 0x4000
+
/* Statistical profiling extension microarchitectural events */
#define ARMV8_SPE_PERFCTR_SAMPLE_POP 0x4000
#define ARMV8_SPE_PERFCTR_SAMPLE_FEED 0x4001
--
2.52.1
^ permalink raw reply related [flat|nested] 14+ messages in thread
* [PATCH v5 6/7] coresight: etm4x: Report whether the PMU counts the external outputs
2026-11-20 15:21 [PATCH v5 0/7] coresight: Pulse an ETM external output for a PMU event Amir Ayupov
` (4 preceding siblings ...)
2026-11-20 15:21 ` [PATCH v5 5/7] perf: arm_pmuv3: Add pmuv3_event_implemented() Amir Ayupov
@ 2026-11-20 15:21 ` Amir Ayupov
2026-11-20 15:36 ` sashiko-bot
2026-11-20 15:21 ` [PATCH v5 7/7] coresight: Add pmu_pulse/extout_pulse Amir Ayupov
6 siblings, 1 reply; 14+ messages in thread
From: Amir Ayupov @ 2026-11-20 15:21 UTC (permalink / raw)
To: James Clark, Leo Yan, Suzuki K Poulose, Mike Leach,
Alexander Shishkin
Cc: Will Deacon, Mark Rutland, Greg Kroah-Hartman, Jonathan Corbet,
Mathieu Poirier, Randy Dunlap, Shuah Khan, Daniel Pereira,
coresight, linux-arm-kernel, linux-perf-users, linux-doc,
linux-kernel
Some CPUs count the trace unit's external outputs as the PMU common
events TRCEXTOUT0-3 (0x4010-0x4013), which lets a configuration drive a
PMU event from trace unit events. Other CPUs, Neoverse N1 for example,
have no such events, and a feature relying on them would silently do
nothing there.
Detect at probe, on the trace unit's CPU, whether the PMU counts every
external output: the trace unit has an external output for each of its
events (caps.nr_event), and pmuv3_event_implemented() reports whether
the PMU implements TRCEXTOUTn. Only check ETE: per its specification, an
implemented TRCEXTOUTn event counts external output n, while whether that
output is also an external input is IMPLEMENTATION DEFINED.
Add a CS_CFG_MATCH_CAP_ETE_TRCEXTOUT capability match flag, set it on
those trace units when registering them with the configuration manager,
and show it in the feature's configfs 'matches' attribute. With
capabilities matched separately from the device class, a feature that
requires it only loads on those trace units.
Assisted-by: Devmate:Opus-5.5
Reviewed-by: James Clark <james.clark@linaro.org>
Tested-by: Leo Yan <leo.yan@arm.com>
Signed-off-by: Amir Ayupov <aaupov@fb.com>
---
.../hwtracing/coresight/coresight-config.h | 2 ++
.../hwtracing/coresight/coresight-etm4x-cfg.c | 7 +++--
.../hwtracing/coresight/coresight-etm4x-cfg.h | 2 +-
.../coresight/coresight-etm4x-core.c | 29 ++++++++++++++++++-
.../coresight/coresight-syscfg-configfs.c | 3 ++
5 files changed, 39 insertions(+), 4 deletions(-)
diff --git a/drivers/hwtracing/coresight/coresight-config.h b/drivers/hwtracing/coresight/coresight-config.h
index 220bc54f8e5e1b..e18143fca90f85 100644
--- a/drivers/hwtracing/coresight/coresight-config.h
+++ b/drivers/hwtracing/coresight/coresight-config.h
@@ -38,6 +38,8 @@
* a device that matches one of its classes and has all of its capabilities.
*/
#define CS_CFG_MATCH_CAP_MASK GENMASK_U32(15, 8)
+/* an ETE whose PMU counts its external outputs as TRCEXTOUTn */
+#define CS_CFG_MATCH_CAP_ETE_TRCEXTOUT BIT(8)
/*
* Limit number of presets in a configuration
diff --git a/drivers/hwtracing/coresight/coresight-etm4x-cfg.c b/drivers/hwtracing/coresight/coresight-etm4x-cfg.c
index 52cb036bb04b96..fea003e49c262a 100644
--- a/drivers/hwtracing/coresight/coresight-etm4x-cfg.c
+++ b/drivers/hwtracing/coresight/coresight-etm4x-cfg.c
@@ -174,11 +174,14 @@ static int etm4_cfg_load_feature(struct coresight_device *csdev,
#define CS_CFG_ETM4_MATCH_FLAGS (CS_CFG_MATCH_CLASS_SRC_ALL | \
CS_CFG_MATCH_CLASS_SRC_ETM4)
-int etm4_cscfg_register(struct coresight_device *csdev)
+int etm4_cscfg_register(struct coresight_device *csdev, bool ete_trcextout)
{
struct cscfg_csdev_feat_ops ops;
+ u32 match_flags = CS_CFG_ETM4_MATCH_FLAGS;
ops.load_feat = &etm4_cfg_load_feature;
+ if (ete_trcextout)
+ match_flags |= CS_CFG_MATCH_CAP_ETE_TRCEXTOUT;
- return cscfg_register_csdev(csdev, CS_CFG_ETM4_MATCH_FLAGS, &ops);
+ return cscfg_register_csdev(csdev, match_flags, &ops);
}
diff --git a/drivers/hwtracing/coresight/coresight-etm4x-cfg.h b/drivers/hwtracing/coresight/coresight-etm4x-cfg.h
index 32dab34c1dace0..d87bcc83e12491 100644
--- a/drivers/hwtracing/coresight/coresight-etm4x-cfg.h
+++ b/drivers/hwtracing/coresight/coresight-etm4x-cfg.h
@@ -25,6 +25,6 @@
#define ETM4_CFG_RES_MASK 0x00F
/* ETMv4 specific config functions */
-int etm4_cscfg_register(struct coresight_device *csdev);
+int etm4_cscfg_register(struct coresight_device *csdev, bool ete_trcextout);
#endif /* CORESIGHT_ETM4X_CFG_H */
diff --git a/drivers/hwtracing/coresight/coresight-etm4x-core.c b/drivers/hwtracing/coresight/coresight-etm4x-core.c
index af017bd8a54773..ecf2a6c945cf05 100644
--- a/drivers/hwtracing/coresight/coresight-etm4x-core.c
+++ b/drivers/hwtracing/coresight/coresight-etm4x-core.c
@@ -30,6 +30,7 @@
#include <linux/uaccess.h>
#include <linux/perf_event.h>
#include <linux/perf/arm_pmu.h>
+#include <linux/perf/arm_pmuv3.h>
#include <linux/platform_device.h>
#include <linux/pm_runtime.h>
#include <linux/property.h>
@@ -78,6 +79,7 @@ static enum cpuhp_state hp_online;
struct etm4_init_arg {
struct device *dev;
struct csdev_access *csa;
+ bool ete_trcextout;
};
static DEFINE_PER_CPU(struct etm4_init_arg *, delayed_probe);
@@ -1399,6 +1401,29 @@ static void etm4_fixup_wrong_ccitmin(struct etmv4_caps *caps)
}
}
+/*
+ * Whether the PMU counts each external output of this trace unit as
+ * TRCEXTOUTn. The trace unit has an external output for each of its events.
+ * Only ETE is checked: per its specification, an implemented TRCEXTOUTn event
+ * counts external output n.
+ *
+ * Must run on the CPU that owns the trace unit.
+ */
+static bool etm4_pmu_counts_extout(struct etmv4_drvdata *drvdata)
+{
+ int n;
+
+ if (!etm4x_is_ete(drvdata) || !drvdata->caps.nr_event)
+ return false;
+
+ for (n = 0; n < drvdata->caps.nr_event; n++) {
+ if (!pmuv3_event_implemented(ARMV8_PMUV3_PERFCTR_TRCEXTOUT0 + n))
+ return false;
+ }
+
+ return true;
+}
+
static void etm4_init_arch_data(void *info)
{
u32 etmidr0;
@@ -1562,6 +1587,8 @@ static void etm4_init_arch_data(void *info)
/* NUMCNTR, bits[30:28] number of counters available for tracing */
caps->nr_cntr = FIELD_GET(TRCIDR5_NUMCNTR_MASK, etmidr5);
+ init_arg->ete_trcextout = etm4_pmu_counts_extout(drvdata);
+
coresight_clear_self_claim_tag_unlocked(csa);
etm4_cs_lock(drvdata, csa);
cpu_detect_trace_filtering(drvdata);
@@ -2265,7 +2292,7 @@ static int etm4_add_coresight_dev(struct etm4_init_arg *init_arg)
}
/* register with config infrastructure & load any current features */
- ret = etm4_cscfg_register(drvdata->csdev);
+ ret = etm4_cscfg_register(drvdata->csdev, init_arg->ete_trcextout);
if (ret) {
coresight_unregister(drvdata->csdev);
return ret;
diff --git a/drivers/hwtracing/coresight/coresight-syscfg-configfs.c b/drivers/hwtracing/coresight/coresight-syscfg-configfs.c
index 2b40e556be872f..6577a5085a634d 100644
--- a/drivers/hwtracing/coresight/coresight-syscfg-configfs.c
+++ b/drivers/hwtracing/coresight/coresight-syscfg-configfs.c
@@ -250,6 +250,9 @@ static ssize_t cscfg_feat_matches_show(struct config_item *item, char *page)
if (match_flags & CS_CFG_MATCH_CLASS_SRC_ETM4)
used += scnprintf(page + used, PAGE_SIZE - used, "SRC_ETMV4 ");
+ if (match_flags & CS_CFG_MATCH_CAP_ETE_TRCEXTOUT)
+ used += scnprintf(page + used, PAGE_SIZE - used, "ETE_TRCEXTOUT ");
+
used += scnprintf(page + used, PAGE_SIZE - used, "\n");
return used;
}
--
2.52.1
^ permalink raw reply related [flat|nested] 14+ messages in thread
* [PATCH v5 7/7] coresight: Add pmu_pulse/extout_pulse
2026-11-20 15:21 [PATCH v5 0/7] coresight: Pulse an ETM external output for a PMU event Amir Ayupov
` (5 preceding siblings ...)
2026-11-20 15:21 ` [PATCH v5 6/7] coresight: etm4x: Report whether the PMU counts the external outputs Amir Ayupov
@ 2026-11-20 15:21 ` Amir Ayupov
6 siblings, 0 replies; 14+ messages in thread
From: Amir Ayupov @ 2026-11-20 15:21 UTC (permalink / raw)
To: James Clark, Leo Yan, Suzuki K Poulose, Mike Leach,
Alexander Shishkin
Cc: Will Deacon, Mark Rutland, Greg Kroah-Hartman, Jonathan Corbet,
Mathieu Poirier, Randy Dunlap, Shuah Khan, Daniel Pereira,
coresight, linux-arm-kernel, linux-perf-users, linux-doc,
linux-kernel
Add ETMv4 configuration 'pmu_pulse' with 'extout_pulse' feature which
requires an ETE whose PMU counts its external outputs
(CS_CFG_MATCH_CAP_ETE_TRCEXTOUT): event 1 selects the always-true
resource and drives TRCEXTOUT1 PMU event which pulses on every cycle
while the trace unit is enabled.
Assisted-by: Devmate:Opus-5.5
Reviewed-by: James Clark <james.clark@linaro.org>
Tested-by: Leo Yan <leo.yan@arm.com>
Reviewed-by: Daniel Pereira <danielmaraboo@gmail.com>
Signed-off-by: Amir Ayupov <aaupov@fb.com>
---
.../trace/coresight/coresight-config.rst | 5 ++
.../trace/coresight/coresight-perf.rst | 15 +++++
drivers/hwtracing/coresight/Makefile | 1 +
.../coresight/coresight-cfg-pmu-pulse.c | 55 +++++++++++++++++++
.../coresight/coresight-cfg-preload.c | 2 +
.../coresight/coresight-cfg-preload.h | 2 +
6 files changed, 80 insertions(+)
create mode 100644 drivers/hwtracing/coresight/coresight-cfg-pmu-pulse.c
diff --git a/Documentation/trace/coresight/coresight-config.rst b/Documentation/trace/coresight/coresight-config.rst
index 6d5ffa6f73471b..3eed728bf0de1e 100644
--- a/Documentation/trace/coresight/coresight-config.rst
+++ b/Documentation/trace/coresight/coresight-config.rst
@@ -129,6 +129,11 @@ The following steps take place in the operation of a configuration.
5) At the end of the perf session, the configuration will be disabled.
+The built-in 'pmu_pulse' configuration pulses ETM external output 1 for a PMU
+event to count. It only pulses while the ETM is enabled, so combined with AUX
+pause and resume the PMU event can time each trace window (see
+coresight-perf.rst).
+
Viewing Configurations and Features
===================================
diff --git a/Documentation/trace/coresight/coresight-perf.rst b/Documentation/trace/coresight/coresight-perf.rst
index 0a77741a431ef9..6770a455c6b592 100644
--- a/Documentation/trace/coresight/coresight-perf.rst
+++ b/Documentation/trace/coresight/coresight-perf.rst
@@ -109,6 +109,21 @@ Example for triggering AUX pause and resume with PMU event::
-e cycles/aux-action=pause,period=10000000/ \
-e cycles/aux-action=resume,period=1050000/ -- sleep 1
+ETM-timed pause events
+~~~~~~~~~~~~~~~~~~~~~~
+
+The built-in ``pmu_pulse`` configuration pulses ETM external output 1 on every
+cycle of allowed trace, and a PMU event counting that output pauses trace after
+``N`` pulses, here 1000 cycles::
+
+ perf record -a -e cs_etm/pmu_pulse,aux-action=start-paused/u \
+ -e cycles/aux-action=resume,period=10000000/u \
+ -e trcextout1/aux-action=pause,period=1000/u \
+ -- sleep 1
+
+Each window is about ``N`` traced cycles plus the pause interrupt latency.
+The configuration only loads on an ETE whose PMU counts its external outputs.
+
Perf test - Verify kernel and userspace perf CoreSight work
-----------------------------------------------------------
diff --git a/drivers/hwtracing/coresight/Makefile b/drivers/hwtracing/coresight/Makefile
index ab16d06783a572..568e797ac5a17e 100644
--- a/drivers/hwtracing/coresight/Makefile
+++ b/drivers/hwtracing/coresight/Makefile
@@ -28,6 +28,7 @@ obj-$(CONFIG_CORESIGHT) += coresight.o
coresight-y := coresight-core.o coresight-etm-perf.o coresight-platform.o \
coresight-sysfs.o coresight-syscfg.o coresight-config.o \
coresight-cfg-preload.o coresight-cfg-afdo.o coresight-cfg-pstop.o \
+ coresight-cfg-pmu-pulse.o \
coresight-syscfg-configfs.o coresight-trace-id.o
obj-$(CONFIG_CORESIGHT_LINK_AND_SINK_TMC) += coresight-tmc.o
coresight-tmc-y := coresight-tmc-core.o coresight-tmc-etf.o \
diff --git a/drivers/hwtracing/coresight/coresight-cfg-pmu-pulse.c b/drivers/hwtracing/coresight/coresight-cfg-pmu-pulse.c
new file mode 100644
index 00000000000000..76e8e6b5da0e97
--- /dev/null
+++ b/drivers/hwtracing/coresight/coresight-cfg-pmu-pulse.c
@@ -0,0 +1,55 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Pulse ETM external output 1 on every traced cycle for a PMU event to count.
+ * Based on coresight-cfg-afdo.c
+ */
+
+#include "coresight-config.h"
+
+/* ETMv4 includes and features */
+#if IS_ENABLED(CONFIG_CORESIGHT_SOURCE_ETM4X)
+#include "coresight-etm4x-cfg.h"
+
+/*
+ * extout_pulse feature: event 1 selects the always-true resource, so ETM
+ * external output 1 pulses on every cycle while the trace unit is enabled. A
+ * PMU event counting that output (TRCEXTOUT1), for example with
+ * aux-action=pause, then fires after as many traced cycles as its sample
+ * period. Event 0 is avoided because it is the trace unit trigger, and the
+ * TRBE driver treats a detected trigger as fatal.
+ */
+static struct cscfg_regval_desc extout_pulse_regs[] = {
+ /* event 1: resource selector 1 (always true), drives external output 1 */
+ {
+ .type = CS_CFG_REG_TYPE_STD | CS_CFG_REG_TYPE_VAL_MASK,
+ .offset = TRCEVENTCTL0R,
+ .val32 = 0x0100,
+ .mask32 = 0xff00,
+ },
+ /* end of regs */
+};
+
+struct cscfg_feature_desc extout_pulse_etm4x = {
+ .name = "extout_pulse",
+ .description = "Pulse external output 1 on every traced cycle.\n",
+ .match_flags = CS_CFG_MATCH_CLASS_SRC_ETM4 | CS_CFG_MATCH_CAP_ETE_TRCEXTOUT,
+ .nr_regs = ARRAY_SIZE(extout_pulse_regs),
+ .regs_desc = extout_pulse_regs,
+};
+
+/* create a pmu_pulse configuration */
+
+static const char *pmu_pulse_ref_names[] = {
+ "extout_pulse",
+};
+
+struct cscfg_config_desc pmu_pulse_etm4x = {
+ .name = "pmu_pulse",
+ .description = "Pulse ETM external output 1 on every traced cycle for the PMU\n"
+ "to count as TRCEXTOUT1\n",
+ .nr_feat_refs = ARRAY_SIZE(pmu_pulse_ref_names),
+ .feat_ref_names = pmu_pulse_ref_names,
+};
+
+/* end of ETM4x configurations */
+#endif /* IS_ENABLED(CONFIG_CORESIGHT_SOURCE_ETM4X) */
diff --git a/drivers/hwtracing/coresight/coresight-cfg-preload.c b/drivers/hwtracing/coresight/coresight-cfg-preload.c
index 4980e68483c5b9..a9a60a1c408455 100644
--- a/drivers/hwtracing/coresight/coresight-cfg-preload.c
+++ b/drivers/hwtracing/coresight/coresight-cfg-preload.c
@@ -14,6 +14,7 @@ static struct cscfg_feature_desc *preload_feats[] = {
#if IS_ENABLED(CONFIG_CORESIGHT_SOURCE_ETM4X)
&strobe_etm4x,
&gen_etrig_etm4x,
+ &extout_pulse_etm4x,
#endif
NULL
};
@@ -22,6 +23,7 @@ static struct cscfg_config_desc *preload_cfgs[] = {
#if IS_ENABLED(CONFIG_CORESIGHT_SOURCE_ETM4X)
&afdo_etm4x,
&pstop_etm4x,
+ &pmu_pulse_etm4x,
#endif
NULL
};
diff --git a/drivers/hwtracing/coresight/coresight-cfg-preload.h b/drivers/hwtracing/coresight/coresight-cfg-preload.h
index 291ba530a6a520..0058627f65ae5b 100644
--- a/drivers/hwtracing/coresight/coresight-cfg-preload.h
+++ b/drivers/hwtracing/coresight/coresight-cfg-preload.h
@@ -12,4 +12,6 @@ extern struct cscfg_feature_desc strobe_etm4x;
extern struct cscfg_config_desc afdo_etm4x;
extern struct cscfg_feature_desc gen_etrig_etm4x;
extern struct cscfg_config_desc pstop_etm4x;
+extern struct cscfg_feature_desc extout_pulse_etm4x;
+extern struct cscfg_config_desc pmu_pulse_etm4x;
#endif
--
2.52.1
^ permalink raw reply related [flat|nested] 14+ messages in thread
* Re: [PATCH v5 3/7] coresight: syscfg: Match feature capabilities separately from device class
2026-11-20 15:21 ` [PATCH v5 3/7] coresight: syscfg: Match feature capabilities separately from device class Amir Ayupov
@ 2026-11-20 15:29 ` sashiko-bot
0 siblings, 0 replies; 14+ messages in thread
From: sashiko-bot @ 2026-11-20 15:29 UTC (permalink / raw)
To: Amir Ayupov; +Cc: linux-perf-users
> A feature loads on a device when the two share any match flag. That works
> for device classes, where a feature wants any one of them, but not for
> capabilities a device may or may not have: a feature that sets a class
> and a capability would still load on a device of that class without the
> capability.
>
> Split the match flags into CS_CFG_MATCH_CLASS_MASK for device classes and
> CS_CFG_MATCH_CAP_MASK for capabilities. A feature now loads on a device
> that shares one of its classes and has all of its capabilities. Check
> this once in cscfg_load_feat_csdev(), which both the feature and the
> device registration paths call. No capability flags are defined yet, so
> matching is unchanged.
>
> Assisted-by: Devmate:Opus-5.5
> Reviewed-by: James Clark <james.clark@linaro.org>
> Tested-by: Leo Yan <leo.yan@arm.com>
> Signed-off-by: Amir Ayupov <aaupov@fb.com>
Sashiko has reviewed this patch and found no issues.
--
Sashiko AI review · https://sashiko.dev/#/patchset/cover.1795182850.git.aaupov@fb.com?part=3
^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH v5 4/7] coresight: syscfg: Remove unused CS_CFG_MATCH_INST_ANY
2026-11-20 15:21 ` [PATCH v5 4/7] coresight: syscfg: Remove unused CS_CFG_MATCH_INST_ANY Amir Ayupov
@ 2026-11-20 15:30 ` sashiko-bot
0 siblings, 0 replies; 14+ messages in thread
From: sashiko-bot @ 2026-11-20 15:30 UTC (permalink / raw)
To: Amir Ayupov; +Cc: linux-perf-users
> CS_CFG_MATCH_INST_ANY has been defined since the system configuration was
> introduced but is not used by any configuration, feature or device.
> Remove it.
>
> Suggested-by: Leo Yan <leo.yan@arm.com>
> Assisted-by: Devmate:Opus-5.5
> Reviewed-by: James Clark <james.clark@linaro.org>
> Tested-by: Leo Yan <leo.yan@arm.com>
> Signed-off-by: Amir Ayupov <aaupov@fb.com>
Sashiko has reviewed this patch and found no issues.
--
Sashiko AI review · https://sashiko.dev/#/patchset/cover.1795182850.git.aaupov@fb.com?part=4
^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH v5 1/7] coresight: etm4x: Fix event count decoding
2026-11-20 15:21 ` [PATCH v5 1/7] coresight: etm4x: Fix event count decoding Amir Ayupov
@ 2026-11-20 15:30 ` sashiko-bot
0 siblings, 0 replies; 14+ messages in thread
From: sashiko-bot @ 2026-11-20 15:30 UTC (permalink / raw)
To: Amir Ayupov; +Cc: linux-perf-users
> From: Leo Yan <leo.yan@arm.com>
>
> TRCIDR0.NUMEVENT normally encodes the number of events minus one.
> From ETMv4.3 onwards, including ETE, a zero TRCIDR4.NUMRSPAIR instead
> means that no events are implemented and NUMEVENT is zero. The sysfs
> event controls currently interpret this combination as one event and
> allow writes to unsupported event fields.
>
> Decode NUMEVENT after NUMRSPAIR and store the actual number of events
> in nr_event. Use the decoded resource count so that older ETM versions,
> where a zero NUMRSPAIR represents one resource pair, still report at
> least one event. Update both sysfs event controls to use the actual
> count and clear their event fields when no events are implemented.
>
> Also correct the instruction event enable mask for the single-event
> case to use INSTEN[0]. The existing mask uses INSTEN[1], preventing
> event 0 from being enabled.
>
> This change depends on the recent driver refactoring. With no reported
> failures, it is not intended for stable backporting.
>
> Signed-off-by: Leo Yan <leo.yan@arm.com>
> Reviewed-by: James Clark <james.clark@linaro.org>
> Tested-by: Leo Yan <leo.yan@arm.com>
> Signed-off-by: Amir Ayupov <aaupov@fb.com>
Sashiko has reviewed this patch and found no issues.
--
Sashiko AI review · https://sashiko.dev/#/patchset/cover.1795182850.git.aaupov@fb.com?part=1
^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH v5 6/7] coresight: etm4x: Report whether the PMU counts the external outputs
2026-11-20 15:21 ` [PATCH v5 6/7] coresight: etm4x: Report whether the PMU counts the external outputs Amir Ayupov
@ 2026-11-20 15:36 ` sashiko-bot
0 siblings, 0 replies; 14+ messages in thread
From: sashiko-bot @ 2026-11-20 15:36 UTC (permalink / raw)
To: Amir Ayupov; +Cc: linux-perf-users
Thank you for your contribution! Sashiko AI review found 1 potential
issue to consider:
- [Medium] coresight: etm4x: PMU coupling and 1-event match in
etm4_pmu_counts_extout()
--
commit f0c882070427fa54aeaa9ce26f43db28f9511f4e
Author: Amir Ayupov <aaupov@fb.com>
coresight: etm4x: Report whether the PMU counts the external outputs
This commit checks at probe whether an ETE trace unit's CPU PMU implements
the TRCEXTOUTn events corresponding to its external outputs and exposes
the CS_CFG_MATCH_CAP_ETE_TRCEXTOUT capability match flag to CoreSight
system configuration.
> diff --git a/drivers/hwtracing/coresight/coresight-etm4x-core.c b/drivers/hwtracing/coresight/coresight-etm4x-core.c
> index af017bd8a5477..ecf2a6c945cf0 100644
> --- a/drivers/hwtracing/coresight/coresight-etm4x-core.c
> +++ b/drivers/hwtracing/coresight/coresight-etm4x-core.c
[ ... ]
> @@ -1399,6 +1401,29 @@ static void etm4_fixup_wrong_ccitmin(struct etmv4_caps *caps)
> }
> }
>
> +/*
> + * Whether the PMU counts each external output of this trace unit as
> + * TRCEXTOUTn. The trace unit has an external output for each of its events.
> + * Only ETE is checked: per its specification, an implemented TRCEXTOUTn event
> + * counts external output n.
> + *
> + * Must run on the CPU that owns the trace unit.
> + */
> +static bool etm4_pmu_counts_extout(struct etmv4_drvdata *drvdata)
> +{
> + int n;
> +
> + if (!etm4x_is_ete(drvdata) || !drvdata->caps.nr_event)
> + return false;
> +
> + for (n = 0; n < drvdata->caps.nr_event; n++) {
> + if (!pmuv3_event_implemented(ARMV8_PMUV3_PERFCTR_TRCEXTOUT0 + n))
> + return false;
> + }
> +
> + return true;
> +}
[Severity: Medium]
As reported by Mike Leach on v2 [1], should ETM/ETE configuration matching
depend on the trace unit's own event resource registers (TRCIDR0 and
TRCIDR4) rather than querying CPU PMU TRCEXTOUTn support inside the ETMv4
driver? Counting external outputs is a PMU feature that the PMU driver
validates when a trcextoutN event is opened, whereas coupling it into a
single ETM capability flag here still returns true when
drvdata->caps.nr_event == 1 and only TRCEXTOUT0 exists. On such an ETE,
event 1 (TRCEVENTCTL0R bits [15:8], driving external output 1) is RES0 and
TRCEXTOUT1 is not checked, yet a feature like extout_pulse in commit
"coresight: Add pmu_pulse/extout_pulse" that matches
CS_CFG_MATCH_CAP_ETE_TRCEXTOUT and unconditionally programs event 1 will
still match and load.
[1] https://lore.kernel.org/r/DB9PR08MB64752311F395D91F4FCBD0CB8C722@DB9PR08MB6475.eurprd08.prod.outlook.com
--
Sashiko AI review · https://sashiko.dev/#/patchset/cover.1795182850.git.aaupov@fb.com?part=6
^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH v5 2/7] coresight: syscfg: Fail to enable a configuration not loaded on the device
2026-11-20 15:21 ` [PATCH v5 2/7] coresight: syscfg: Fail to enable a configuration not loaded on the device Amir Ayupov
@ 2026-11-20 15:36 ` sashiko-bot
0 siblings, 0 replies; 14+ messages in thread
From: sashiko-bot @ 2026-11-20 15:36 UTC (permalink / raw)
To: Amir Ayupov; +Cc: linux-perf-users
Thank you for your contribution! Sashiko AI review found 1 potential
issue to consider:
- [High] coresight: syscfg: cscfg_csdev_enable_active_config() rejects
unmatched csdev
--
commit 963b92e15bb563c4e0e56506b797ad0f0117ff05
Author: Amir Ayupov <aaupov@fb.com>
coresight: syscfg: Fail to enable a configuration not loaded on the device
This commit makes cscfg_csdev_enable_active_config() return -EINVAL when
the active configuration is not loaded on the target CoreSight device.
> diff --git a/drivers/hwtracing/coresight/coresight-syscfg.c b/drivers/hwtracing/coresight/coresight-syscfg.c
> index 2bfdd7b45e49c..a816c22c87bc8 100644
> --- a/drivers/hwtracing/coresight/coresight-syscfg.c
> +++ b/drivers/hwtracing/coresight/coresight-syscfg.c
> @@ -1088,36 +1088,39 @@ int cscfg_csdev_enable_active_config(struct coresight_device *csdev,
> raw_spin_unlock_irqrestore(&csdev->cscfg_csdev_lock, flags);
>
> /*
> - * If found, attempt to enable
> + * The configuration is active but not loaded on this device, because
> + * none of its features match the device. Fail rather than trace
> + * without it.
> */
> - if (config_csdev_active) {
> + if (!config_csdev_active)
> + return -EINVAL;
[Severity: High]
As reported by Mike Leach on v2 [1], should
cscfg_csdev_enable_active_config() fail when a configuration is not loaded
on csdev?
cscfg_csdev_enable_active_config() is a generic helper across CoreSight
device classes, and its kernel-doc states that it enables the configuration
only if supported on the device:
drivers/hwtracing/coresight/coresight-syscfg.c:
/**
* cscfg_csdev_enable_active_config - Enable matching active configuration for device.
*
* Enables the configuration selected by @cfg_hash if the configuration is supported
* on the device and has been activated.
For perf sessions, etm_setup_aux() activates cfg_hash globally and builds
paths for every CPU in event_data->mask without filtering out ETM sources
that do not have the configuration loaded. Returning -EINVAL here causes
etm_setup_aux() to succeed even if some or all CPUs lack the configuration,
only for etm_event_start() -> coresight_enable_source() -> etm4_enable_perf()
to fail at schedule-in time and stop the event with PERF_HES_STOPPED.
Incompatible ETM sources should instead be filtered out of event_data->mask
during etm_setup_aux().
[1] https://lore.kernel.org/r/DB9PR08MB64756F200065042673030C848C722@DB9PR08MB6475.eurprd08.prod.outlook.com
--
Sashiko AI review · https://sashiko.dev/#/patchset/cover.1795182850.git.aaupov@fb.com?part=2
^ permalink raw reply [flat|nested] 14+ messages in thread
* Re: [PATCH v5 5/7] perf: arm_pmuv3: Add pmuv3_event_implemented()
2026-11-20 15:21 ` [PATCH v5 5/7] perf: arm_pmuv3: Add pmuv3_event_implemented() Amir Ayupov
@ 2026-11-20 15:39 ` sashiko-bot
0 siblings, 0 replies; 14+ messages in thread
From: sashiko-bot @ 2026-11-20 15:39 UTC (permalink / raw)
To: Amir Ayupov; +Cc: linux-perf-users
Thank you for your contribution! Sashiko AI review found 1 potential
issue to consider:
- [Medium] perf: arm_pmu: build failure on 32-bit ARM without
CONFIG_CPU_CP15
--
commit 35316b1d7277cb0d1201241837c2de25d95ad9c1
Author: Amir Ayupov <aaupov@fb.com>
perf: arm_pmuv3: Add pmuv3_event_implemented()
This commit adds pmuv3_event_implemented() to arch/arm64/include/asm/arm_pmuv3.h
to query whether the local CPU's PMU implements a common event via PMCEID0/1,
and moves ARMV8_PMUV3_MAX_COMMON_EVENTS and ARMV8_PMUV3_EXT_COMMON_EVENT_BASE
from include/linux/perf/arm_pmu.h to include/linux/perf/arm_pmuv3.h.
> diff --git a/include/linux/perf/arm_pmu.h b/include/linux/perf/arm_pmu.h
> index 02d2c7f45b527..a90867c9774c7 100644
> --- a/include/linux/perf/arm_pmu.h
> +++ b/include/linux/perf/arm_pmu.h
> @@ -10,6 +10,7 @@
>
> #include <linux/interrupt.h>
> #include <linux/perf_event.h>
> +#include <linux/perf/arm_pmuv3.h>
> #include <linux/platform_device.h>
> #include <linux/sysfs.h>
> #include <asm/cputype.h>
[Severity: Medium]
Does unconditionally including linux/perf/arm_pmuv3.h here break 32-bit
ARM builds when CONFIG_CPU_CP15 is disabled (for example on
CONFIG_CPU_V7M) and CONFIG_ARM_PMU is enabled?
linux/perf/arm_pmuv3.h includes asm/arm_pmuv3.h, and on 32-bit ARM the
static inline helpers in arch/arm/include/asm/arm_pmuv3.h expand
read_sysreg(), write_sysreg(), __ACCESS_CP15(), and __ACCESS_CP15_64():
arch/arm/include/asm/arm_pmuv3.h:read_pmevcntrn() {
...
#define RETURN_READ_PMEVCNTRN(n) \
return read_sysreg(PMEVCNTR##n)
static inline unsigned long read_pmevcntrn(int n)
{
PMEVN_SWITCH(n, RETURN_READ_PMEVCNTRN);
return 0;
}
...
}
In arch/arm/include/asm/cp15.h and arch/arm/include/asm/vdso/cp15.h,
those macros are defined only under #ifdef CONFIG_CPU_CP15. Previously
linux/perf/arm_pmuv3.h was only compiled on 32-bit ARM under
CONFIG_ARM_PMUV3 (which depends on CPU_V7 and thus CPU_CP15), whereas
CONFIG_ARM_PMU depends only on ARM || ARM64, so compiling
drivers/perf/arm_pmu.c without CONFIG_CPU_CP15 fails with undeclared
CP15 macros and helpers.
--
Sashiko AI review · https://sashiko.dev/#/patchset/cover.1795182850.git.aaupov@fb.com?part=5
^ permalink raw reply [flat|nested] 14+ messages in thread
end of thread, other threads:[~2026-11-20 15:40 UTC | newest]
Thread overview: 14+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-11-20 15:21 [PATCH v5 0/7] coresight: Pulse an ETM external output for a PMU event Amir Ayupov
2026-11-20 15:21 ` [PATCH v5 1/7] coresight: etm4x: Fix event count decoding Amir Ayupov
2026-11-20 15:30 ` sashiko-bot
2026-11-20 15:21 ` [PATCH v5 2/7] coresight: syscfg: Fail to enable a configuration not loaded on the device Amir Ayupov
2026-11-20 15:36 ` sashiko-bot
2026-11-20 15:21 ` [PATCH v5 3/7] coresight: syscfg: Match feature capabilities separately from device class Amir Ayupov
2026-11-20 15:29 ` sashiko-bot
2026-11-20 15:21 ` [PATCH v5 4/7] coresight: syscfg: Remove unused CS_CFG_MATCH_INST_ANY Amir Ayupov
2026-11-20 15:30 ` sashiko-bot
2026-11-20 15:21 ` [PATCH v5 5/7] perf: arm_pmuv3: Add pmuv3_event_implemented() Amir Ayupov
2026-11-20 15:39 ` sashiko-bot
2026-11-20 15:21 ` [PATCH v5 6/7] coresight: etm4x: Report whether the PMU counts the external outputs Amir Ayupov
2026-11-20 15:36 ` sashiko-bot
2026-11-20 15:21 ` [PATCH v5 7/7] coresight: Add pmu_pulse/extout_pulse Amir Ayupov
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox