CoreSight

coresight@lists.linaro.org

9 participants
2378 discussions

[PATCH v4 00/13] coresight: Fix CTI module refcount leak by making it a helper device

by James Clark

Changes since v3: * Put connection loss fix at the beginning so that it can be backported * Replace coresight_find_link_{x}() with coresight_find_out_connection() * Reorder coresight_enable_source() arguments for consistency * Add source and destination reference counts so that two link devices connected together don't clash * Add coresight_is_helper() * Fix overwriting csdev bug in coresight_orphan_match() * Don't clear conns[i]->dest_fwnode in coresight_remove_conns() in case it's used again * Use dev instead of adev->dev for devmem allocation in acpi_coresight_parse_graph() so that it's consistent with DT mode and doesn't cause a warning on free. * Rename coresight_add_helper_mutex() -> coresight_add_helper() * Ensure coresight_mutex isn't already held in coresight_add_helper() * Return new connection from coresight_add_out_conn() * Comment and formatting improvements ------------------ Changes since v2: * Make out_conns array contiguous instead of sparse which simplifies filling and using it. New connections are always added to the end * Store pointers to individual connection objects so that they can be shared between inputs and outputs * Fix an existing bug where connection info was lost when unloading a device * Simplify connection fixup functions. Now the orphan mechanism is used for inputs in the same way as outputs to guarantee that all connections have both an input and an output set * Use input connections to disconnect devices on unload instead of iterating through them all * Make refcount a property of the connection rather than use it's own array based on the number of inputs and outputs * Fix a bug in v2 where helpers attached to the source device weren't disabled because coresight-etm-perf.c was making a raw call to disable rather than using a helper. * Change names of connection members to make direction explicit now that the connection is shared between input and outputs ------------------ Changes since v1: * Don't dereference handle in tmc_etr_get_buffer() when not in perf mode. * Fix some W=1 warnings * Add a commit to rename child/output in terms of local/remote ------------------- Currently there is a refcount leak in CTI when using system wide mode or tracing multithreaded applications. See the last commit for a reproducer. This prevents the module from being unloaded. Historically there have been a few issues and fixes attempted around here which have resulted in some extra logic and a member to keep track of CTI being enabled 'struct coresight_device->ect_enabled'. The fix in commit 665c157e0204 ("coresight: cti: Fix hang in cti_disable_hw()") was also related to CTI having its own enable/disable path which came later than other devices. If we make CTI a helper device and enable helper devices adjacent to the path we get very similar enable/disable behavior to now, but with more reuse of the existing reference counting logic in the coresight core code. This also affects CATU which can have a little bit of its hard coded enable/disable code removed. Enabling CATU on the generic path does require that input connections are tracked so that it can get its associated ETR buffer. Applies to coresight/next (197b6b60ae7b) but also requires the realloc_array patch here [1]. Also available in full here [2]. [1]: https://lore.kernel.org/linux-arm-kernel/20230306152723.3090195-1-james.cla… [2]: https://gitlab.arm.com/linux-arm/linux-jc/-/tree/james-cs-cti-module-refcou… James Clark (13): coresight: Fix loss of connection info when a module is unloaded coresight: Use enum type for cs_mode wherever possible coresight: Change name of pdata->conns coresight: Rename nr_outports to nr_outconns coresight: Rename connection members to make the direction explicit coresight: Dynamically add connections coresight: Store pointers to connections rather than an array of them coresight: Simplify connection fixup mechanism coresight: Store in-connections as well as out-connections coresight: Make refcount a property of the connection coresight: Refactor out buffer allocation function for ETR coresight: Enable and disable helper devices adjacent to the path coresight: Fix CTI module refcount leak by making it a helper device drivers/hwtracing/coresight/coresight-catu.c | 21 +- drivers/hwtracing/coresight/coresight-core.c | 595 ++++++++++-------- .../hwtracing/coresight/coresight-cti-core.c | 52 +- .../hwtracing/coresight/coresight-cti-sysfs.c | 4 +- drivers/hwtracing/coresight/coresight-cti.h | 4 +- drivers/hwtracing/coresight/coresight-etb10.c | 13 +- .../hwtracing/coresight/coresight-etm-perf.c | 4 +- .../coresight/coresight-etm3x-core.c | 6 +- .../coresight/coresight-etm4x-core.c | 6 +- .../hwtracing/coresight/coresight-funnel.c | 26 +- .../hwtracing/coresight/coresight-platform.c | 260 +++----- drivers/hwtracing/coresight/coresight-priv.h | 17 +- .../coresight/coresight-replicator.c | 23 +- drivers/hwtracing/coresight/coresight-stm.c | 6 +- drivers/hwtracing/coresight/coresight-sysfs.c | 17 +- .../hwtracing/coresight/coresight-tmc-etf.c | 26 +- .../hwtracing/coresight/coresight-tmc-etr.c | 110 ++-- drivers/hwtracing/coresight/coresight-tmc.h | 2 + drivers/hwtracing/coresight/coresight-tpda.c | 23 +- drivers/hwtracing/coresight/coresight-tpdm.c | 4 +- drivers/hwtracing/coresight/coresight-tpiu.c | 7 +- drivers/hwtracing/coresight/coresight-trbe.c | 3 +- drivers/hwtracing/coresight/ultrasoc-smb.c | 11 +- drivers/hwtracing/coresight/ultrasoc-smb.h | 2 +- include/linux/coresight.h | 127 ++-- 25 files changed, 701 insertions(+), 668 deletions(-) -- 2.34.1

2 years, 2 months

[PATCH v3 00/13] coresight: Fix CTI module refcount leak by making it a helper device

by James Clark

Changes since v2: * Make out_conns array contiguous instead of sparse which simplifies filling and using it. New connections are always added to the end * Store pointers to individual connection objects so that they can be shared between inputs and outputs * Fix an existing bug where connection info was lost when unloading a device * Simplify connection fixup functions. Now the orphan mechanism is used for inputs in the same way as outputs to guarantee that all connections have both an input and an output set * Use input connections to disconnect devices on unload instead of iterating through them all * Make refcount a property of the connection rather than use it's own array based on the number of inputs and outputs * Fix a bug in v2 where helpers attached to the source device weren't disabled because coresight-etm-perf.c was making a raw call to disable rather than using a helper. * Change names of connection members to make direction explicit now that the connection is shared between input and outputs ------------------ Changes since v1: * Don't dereference handle in tmc_etr_get_buffer() when not in perf mode. * Fix some W=1 warnings * Add a commit to rename child/output in terms of local/remote ------------------- Currently there is a refcount leak in CTI when using system wide mode or tracing multithreaded applications. See the last commit for a reproducer. This prevents the module from being unloaded. Historically there have been a few issues and fixes attempted around here which have resulted in some extra logic and a member to keep track of CTI being enabled 'struct coresight_device->ect_enabled'. The fix in commit 665c157e0204 ("coresight: cti: Fix hang in cti_disable_hw()") was also related to CTI having its own enable/disable path which came later than other devices. If we make CTI a helper device and enable helper devices adjacent to the path we get very similar enable/disable behavior to now, but with more reuse of the existing reference counting logic in the coresight core code. This also affects CATU which can have a little bit of its hard coded enable/disable code removed. Enabling CATU on the generic path does require that input connections are tracked so that it can get its associated ETR buffer. Applies to coresight/next (197b6b60ae7b) but also requires the realloc_array patch here [1]. Also available in full here [2]. [1]: https://lore.kernel.org/linux-arm-kernel/20230306152723.3090195-1-james.cla… [2]: https://gitlab.arm.com/linux-arm/linux-jc/-/tree/james-cs-cti-module-refcou… James Clark (13): coresight: Use enum type for cs_mode wherever possible coresight: Change name of pdata->conns coresight: Rename nr_outports to nr_outconns coresight: Rename connection members to make the direction explicit coresight: Dynamically add connections coresight: Fix loss of connection info when a module is unloaded coresight: Store pointers to connections rather than an array of them coresight: Simplify connection fixup mechanism coresight: Store in-connections as well as out-connections coresight: Make refcount a property of the connection coresight: Refactor out buffer allocation function for ETR coresight: Enable and disable helper devices adjacent to the path coresight: Fix CTI module refcount leak by making it a helper device drivers/hwtracing/coresight/coresight-catu.c | 21 +- drivers/hwtracing/coresight/coresight-core.c | 557 +++++++++--------- .../hwtracing/coresight/coresight-cti-core.c | 52 +- .../hwtracing/coresight/coresight-cti-sysfs.c | 4 +- drivers/hwtracing/coresight/coresight-cti.h | 4 +- drivers/hwtracing/coresight/coresight-etb10.c | 13 +- .../hwtracing/coresight/coresight-etm-perf.c | 4 +- .../coresight/coresight-etm3x-core.c | 6 +- .../coresight/coresight-etm4x-core.c | 6 +- .../hwtracing/coresight/coresight-funnel.c | 26 +- .../hwtracing/coresight/coresight-platform.c | 250 +++----- drivers/hwtracing/coresight/coresight-priv.h | 17 +- .../coresight/coresight-replicator.c | 23 +- drivers/hwtracing/coresight/coresight-stm.c | 6 +- drivers/hwtracing/coresight/coresight-sysfs.c | 17 +- .../hwtracing/coresight/coresight-tmc-etf.c | 26 +- .../hwtracing/coresight/coresight-tmc-etr.c | 110 ++-- drivers/hwtracing/coresight/coresight-tmc.h | 2 + drivers/hwtracing/coresight/coresight-tpda.c | 23 +- drivers/hwtracing/coresight/coresight-tpdm.c | 4 +- drivers/hwtracing/coresight/coresight-tpiu.c | 7 +- drivers/hwtracing/coresight/coresight-trbe.c | 3 +- drivers/hwtracing/coresight/ultrasoc-smb.c | 11 +- drivers/hwtracing/coresight/ultrasoc-smb.h | 2 +- include/linux/coresight.h | 124 ++-- 25 files changed, 669 insertions(+), 649 deletions(-) -- 2.34.1

2 years, 2 months

[PATCH] MAINTAINERS: Remove Mathieu Poirier as coresight maintainer

by Suzuki K Poulose

Mathieu Poirier is no longer involved in maintainig the CoreSight self-hosted tracing subsystem. Mathieu, Thank you very much creating and maintaing the subsystem all these years ! Cc: Mathieu Poirier <mathieu.poirier(a)linaro.org> Signed-off-by: Suzuki K Poulose <suzuki.poulose(a)arm.com> --- CREDITS | 5 +++++ MAINTAINERS | 1 - 2 files changed, 5 insertions(+), 1 deletion(-) diff --git a/CREDITS b/CREDITS index 847059166a15..216e86f56dd9 100644 --- a/CREDITS +++ b/CREDITS @@ -3459,6 +3459,11 @@ D: several improvements to system programs S: Oldenburg S: Germany +N: Mathieu Poirier +E: mathieu.poirier(a)linaro.org +D: CoreSight kernel subsystem, Maintainer 2014-2022 +D: Perf tool support for CoreSight + N: Robert Schwebel E: robert(a)schwebel.de W: https://www.schwebel.de diff --git a/MAINTAINERS b/MAINTAINERS index 90abe83c02f3..6c971e95a6fd 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2091,7 +2091,6 @@ F: arch/arm/boot/dts/cx92755* N: digicolor ARM/CORESIGHT FRAMEWORK AND DRIVERS -M: Mathieu Poirier <mathieu.poirier(a)linaro.org> M: Suzuki K Poulose <suzuki.poulose(a)arm.com> R: Mike Leach <mike.leach(a)linaro.org> R: Leo Yan <leo.yan(a)linaro.org> -- 2.34.1

2 years, 2 months

Re: [PATCH V2 3/5] coresight: etm4x: Drop pid argument from etm4_probe()

by Suzuki K Poulose

On 31/03/2023 22:24, Steve Clevenger wrote: > > > On 3/31/2023 4:06 AM, Suzuki K Poulose wrote: >> On 27/03/2023 06:05, Anshuman Khandual wrote: >>> Coresight device pid can be retrieved from its iomem base address, >>> which is >>> stored in 'struct etm4x_drvdata'. This drops pid argument from >>> etm4_probe() >>> and 'struct etm4_init_arg'. Instead etm4_check_arch_features() derives >>> the >>> coresight device pid with a new helper coresight_get_pid(), right >>> before it >>> is consumed in etm4_hisi_match_pid(). >>> >>> Cc: Mathieu Poirier <mathieu.poirier(a)linaro.org> >>> Cc: Suzuki K Poulose <suzuki.poulose(a)arm.com> >>> Cc: Mike Leach <mike.leach(a)linaro.org> >>> Cc: Leo Yan <leo.yan(a)linaro.org> >>> Cc: coresight(a)lists.linaro.org >>> Cc: linux-arm-kernel(a)lists.infradead.org >>> Cc: linux-kernel(a)vger.kernel.org >>> Signed-off-by: Anshuman Khandual <anshuman.khandual(a)arm.com> >>> --- >>> .../coresight/coresight-etm4x-core.c | 21 +++++++------------ >>> include/linux/coresight.h | 12 +++++++++++ >>> 2 files changed, 20 insertions(+), 13 deletions(-) >>> >>> diff --git a/drivers/hwtracing/coresight/coresight-etm4x-core.c >>> b/drivers/hwtracing/coresight/coresight-etm4x-core.c >>> index 5d77571a8df9..3521838ab4fb 100644 >>> --- a/drivers/hwtracing/coresight/coresight-etm4x-core.c >>> +++ b/drivers/hwtracing/coresight/coresight-etm4x-core.c >>> @@ -66,7 +66,6 @@ static u64 etm4_get_access_type(struct etmv4_config >>> *config); >>> static enum cpuhp_state hp_online; >>> struct etm4_init_arg { >>> - unsigned int pid; >>> struct device *dev; >>> struct csdev_access *csa; >>> }; >>> @@ -370,8 +369,10 @@ static void etm4_disable_arch_specific(struct >>> etmv4_drvdata *drvdata) >>> } >>> static void etm4_check_arch_features(struct etmv4_drvdata *drvdata, >>> - unsigned int id) >>> + struct csdev_access *csa) >>> { >>> + unsigned int id = coresight_get_pid(csa); >>> + >> >> This throws up the following error on an ETE. >> >> ete: trying to read unsupported register @fe0 >> >> So, I guess this must be performed only for iomem based >> devices. System instruction based device must be identified >> by MIDR_EL1/REVIDR_EL1 if needed for specific erratum. >> This is not required now. So, we could bail out early >> if we are system instruction based device. > > Besides this, the PID is limited to (I think) 4 bits of ID. TRCIDRs > offer revision information, but nothing manufacturer specific save for > the designer. Register fields like MIDR_EL1 Variant + PartNum + Revision > and TRCPIDR3 REVAND offer help. It may be a combination of registers are > needed for a manufacturer to adequately ID a part to apply an erratum. > Perhaps you could at least cache MIDR_EL1 for possible future use? Like I said, if we ever need them, we could add it. I don't see a point in storing it right now, if we don't use it. Suzuki

2 years, 2 months

[PATCH v3 0/11] Add support to configure TPDM DSB subunit

by Tao Zhang

Introduction of TPDM DSB subunit DSB subunit is responsible for creating a dataset element, and is also optionally responsible for packing it to fit multiple elements on a single ATB transfer if possible in the configuration. The TPDM Core Datapath requests timestamps be stored by the TPDA and then delivering ATB sized data (depending on ATB width and element size, this could be smaller or larger than a dataset element) to the ATB Mast FSM. The DSB subunit must be configured prior to enablement. This series adds support for TPDM to configure the configure DSB subunit. Once this series patches are applied properly, the new tpdm nodes for should be observed at the tpdm path /sys/bus/coresight/devices/tpdm* which supports DSB subunit. e.g. /sys/devices/platform/soc(a)0/69d0000.tpdm/tpdm0#ls -l | grep dsb -rw-r--r-- 1 root root 4096 Jan 1 00:01 dsb_edge_ctrl -rw-r--r-- 1 root root 4096 Jan 1 00:01 dsb_edge_ctrl_mask -rw-r--r-- 1 root root 4096 Jan 1 00:01 dsb_mode -rw-r--r-- 1 root root 4096 Jan 1 00:01 dsb_patt_mask -rw-r--r-- 1 root root 4096 Jan 1 00:01 dsb_patt_ts -rw-r--r-- 1 root root 4096 Jan 1 00:01 dsb_patt_type -rw-r--r-- 1 root root 4096 Jan 1 00:01 dsb_patt_val -rw-r--r-- 1 root root 4096 Jan 1 00:01 dsb_trig_patt_mask -rw-r--r-- 1 root root 4096 Jan 1 00:01 dsb_trig_patt_val -rw-r--r-- 1 root root 4096 Jan 1 00:01 dsb_trig_ts -rw-r--r-- 1 root root 4096 Jan 1 00:01 dsb_trig_type We can use the commands are similar to the below to configure the TPDMs which support DSB subunit. Enable coresight sink first. echo 1 > /sys/bus/coresight/devices/tmc_etf0/enable_sink echo 1 > /sys/bus/coresight/devices/tpdm0/reset echo 0x3 0x3 0x1 > /sys/bus/coresight/devices/tpdm0/dsb_edge_ctrl_mask echo 0x6d 0x6d 0 > /sys/bus/coresight/devices/tpdm0/dsb_edge_ctrl echo 1 > /sys/bus/coresight/devices/tpdm0/dsb_patt_ts echo 1 > /sys/bus/coresight/devices/tpdm0/dsb_patt_type echo 0 > /sys/bus/coresight/devices/tpdm0/dsb_trig_ts echo 0 0xFFFFFFFF > /sys/bus/coresight/devices/tpdm0/dsb_patt_mask echo 0 0xFFFFFFFF > /sys/bus/coresight/devices/tpdm0/dsb_trig_patt_val This patch series depends on patch series "[PATCH v2 0/9] coresight: Fix CTI module refcount leak by making it a helper device" https://patchwork.kernel.org/project/linux-arm-kernel/cover/20230310160610.… TPDM_DSB commit tree: https://git.codelinaro.org/clo/linux-kernel/coresight/-/tree/tpdm-dsb-v3 https://git.codelinaro.org/clo/linux-kernel/coresight/-/commits/tpdm-dsb-v3 Changes in V3: 1. Move the property "qcom,dsb-element-size" to TPDM devicetree and update the TPDM yaml file for this item. -- Suzuki K Poulose 2. Add the error message when the DSB element size is not set to 32-bit or 64-bit. -- Suzuki K Poulose 3. Add more information to the comments of patch #3 -- Suzuki K Poulose 4. Combine the value updates to the TPDM_DSB_CR for TPDM. -- Suzuki K Poulose 5. Remove the function "tpdm_datasets_alloc", and fold its code to a new function "tpdm_init_datasets". It will complete the initialization of TPDM. -- Suzuki K Poulose 6. Change the method of qualifying input values. -- Suzuki K Poulose 7. Add the documentation of the new sysfs handles. -- Suzuki K Poulose 8. Provide the separate handles for the "mode bits". -- Suzuki K Poulose Changes in V2: 1. Change the name of the property "qcom,dsb-elem-size" to "qcom,dsb-element-size" -- Suzuki K Poulose 2. Update the TPDA yaml file for the item "qcom,dsb-elem-size". -- Krzysztof Kozlowski 3. Add the full name of DSB in the description of the item "qcom,dsb-elem-size". -- Rob Herring Changes in V1: 1. Change the definition of the property "qcom,dsb-elem-size" from "uint32-array" to "uint32-matrix". -- Krzysztof Kozlowski 2. Add the full name of DSB. -- Rob Herring 3. Deal with 2 entries in an iteration in TPDA driver. -- Suzuki K Poulose 4. Divide the function "tpdm_datasets_alloc" into two functions, "tpdm_datasets_setup" and "tpdm_datasets_alloc". 5. Detecte the input string with the conventional semantics automatically, and constrain the size of the input value. -- Suzuki K Poulose 6. Use the hook function "is_visible()" to hide the DSB related knobs if the data sets are missing. -- Suzuki K Poulose 7. Use the macros "FIELD_GET" and "FIELD_PREP" to set the values. -- Suzuki K Poulose 8. Update the definition of the macros in TPDM driver. 9. Update the comments of the values for the nodes which are for DSB element creation and onfigure pattern match output. -- Suzuki K Poulose 10. Use API "sysfs_emit" to "replace scnprintf". -- Suzuki K Poulose Tao Zhang (10): dt-bindings: arm: Add support for DSB element size coresight-tpda: Add DSB dataset support coresight-tpdm: Initialize DSB subunit configuration coresight-tpdm: Add reset node to TPDM node coresight-tpdm: Add nodes to set trigger timestamp and type coresight-tpdm: Add node to set dsb programming mode coresight-tpdm: Add nodes for dsb edge control coresight-tpdm: Add nodes to configure pattern match output coresight-tpdm: Add nodes for timestamp request dt-bindings: arm: Add support for DSB MSR register coresight-tpdm: Add nodes for dsb msr support .../ABI/testing/sysfs-bus-coresight-devices-tpdm | 144 +++++ .../bindings/arm/qcom,coresight-tpdm.yaml | 21 + drivers/hwtracing/coresight/coresight-tpda.c | 58 ++ drivers/hwtracing/coresight/coresight-tpda.h | 4 + drivers/hwtracing/coresight/coresight-tpdm.c | 694 ++++++++++++++++++++- drivers/hwtracing/coresight/coresight-tpdm.h | 69 ++ 6 files changed, 984 insertions(+), 6 deletions(-) -- 2.7.4

2 years, 2 months

[PATCH v9 0/3] perf: cs-etm: Update perf to handle new Coresight Trace ID

by Mike Leach

The original method for allocating trace source ID values to sources was to use a fixed algorithm for CPU based sources of (cpu_num * 2 + 0x10). The STM was allocated ID 0x1. This mechanism is broken for systems with more than 47 cores. The kernel related patches the provide a fixed Trace ID allocation mechanism are now upstreamed. This patchset updates the perf code to handle the changes to the trace ID notification mechanism that now uses the PERF_RECORD_AUX_OUTPUT_HW_ID packet to set Trace ID in the perf ETM decoders. Applies and test oo perf/core Changes since v8: 1. Fix build issues 2. Fix implicit function problem Changes since v7: Split from original patchset [1] to be sent separately as kernel related patches are now upstream. [1] https://lore.kernel.org/linux-arm-kernel/20230116124928.5440-1-mike.leach@l… Mike Leach (3): perf: cs-etm: Move mapping of Trace ID and cpu into helper function perf: cs-etm: Update record event to use new Trace ID protocol perf: cs-etm: Handle PERF_RECORD_AUX_OUTPUT_HW_ID packet tools/include/linux/coresight-pmu.h | 47 ++- tools/perf/arch/arm/util/cs-etm.c | 27 +- tools/perf/util/cs-etm-base.c | 3 +- .../perf/util/cs-etm-decoder/cs-etm-decoder.c | 7 + tools/perf/util/cs-etm.c | 326 +++++++++++++++--- tools/perf/util/cs-etm.h | 14 +- 6 files changed, 356 insertions(+), 68 deletions(-) -- 2.17.1

2 years, 2 months

[PATCH v8 0/3] perf: cs-etm: Update perf to handle new Coresight Trace ID

by Mike Leach

The original method for allocating trace source ID values to sources was to use a fixed algorithm for CPU based sources of (cpu_num * 2 + 0x10). The STM was allocated ID 0x1. This mechanism is broken for systems with more than 47 cores. The kernel related patches the provide a fixed Trace ID allocation mechanism are now upstreamed. This patchset updates the perf code to handle the changes to the trace ID notification mechanism that now uses the PERF_RECORD_AUX_OUTPUT_HW_ID packet to set Trace ID in the perf ETM decoders. Applies to perf/core Changes since v7: Split from original patchset [1] to be sent separately as kernel related patches are now upstream. [1] https://lore.kernel.org/linux-arm-kernel/20230116124928.5440-1-mike.leach@l… Mike Leach (3): perf: cs-etm: Move mapping of Trace ID and cpu into helper function perf: cs-etm: Update record event to use new Trace ID protocol perf: cs-etm: Handle PERF_RECORD_AUX_OUTPUT_HW_ID packet tools/include/linux/coresight-pmu.h | 47 ++- tools/perf/arch/arm/util/cs-etm.c | 21 +- .../perf/util/cs-etm-decoder/cs-etm-decoder.c | 7 + tools/perf/util/cs-etm.c | 326 +++++++++++++++--- tools/perf/util/cs-etm.h | 14 +- 5 files changed, 350 insertions(+), 65 deletions(-) -- 2.32.0

2 years, 2 months

Re: [BUG] perf: No samples found when using kcore + coresight

by Leo Yan

On Mon, Mar 13, 2023 at 11:15:44AM -0700, Yang Shi wrote: [...] > > Just a quick summary, here we have two issues: > > > > - With command: > > perf record -e cs_etm/@tmc_etf63/k --kcore --per-thread \ > > -- taskset --cpu-list 1 uname", > > > > perf doesn't enable "text poke" attribution. > > No, it enables "text poke" and perf fails to decode coresight trace > data too. It doesn't matter whether "--kcore" is after or before "-e > cs/etm/@tmc_etf63/k". Understand now. Thanks for correction, if so we can ignore this one. Leo

2 years, 2 months

Re: [BUG] perf: No samples found when using kcore + coresight

by James Clark

On 30/03/2023 00:25, Yang Shi wrote: > On Wed, Mar 29, 2023 at 9:08 AM James Clark <james.clark(a)arm.com> wrote: >> >> >> >> On 14/03/2023 00:36, Leo Yan wrote: >>> On Mon, Mar 13, 2023 at 11:15:44AM -0700, Yang Shi wrote: >>> >>> [...] >>> >>>>> Just a quick summary, here we have two issues: >>>>> >>>>> - With command: >>>>> perf record -e cs_etm/@tmc_etf63/k --kcore --per-thread \ >>>>> -- taskset --cpu-list 1 uname", >>>>> >>>>> perf doesn't enable "text poke" attribution. >>>> >>>> No, it enables "text poke" and perf fails to decode coresight trace >>>> data too. It doesn't matter whether "--kcore" is after or before "-e >>>> cs/etm/@tmc_etf63/k". >>> >>> Understand now. Thanks for correction, if so we can ignore this one. >>> >>> Leo >> >> To me it looks like it's only --per-thread and --kcore together that >> cause the issue. I can't see if that was mentioned previously in this >> thread. > > If "--pre-thread" is not passed in, perf record failed with "failed to > mmap with 12 (Cannot allocate memory)". Sorry for not mentioning this > in the first place. I was quite focused on --kcore and didn't realize > they may be related. That's unrelated. That's because you have specified a sink and without --per-thread it tries to open the event on all cores. If the sink can't be reached from all cores it will fail to open. You can make it work without --per-thread if you limit it to a valid core like this, although I don't know which ones exactly would be valid for your system: perf record -e cs_etm/@tmc_etf63/k --kcore -C 0 \ -- taskset --cpu-list 1 uname > >> >> If it is --per-thread that's causing the issue then I think I have an >> idea why it might be. There are some assumptions and different paths >> taken in decoding in that mode that aren't correct. It causes some other >> issues to do with ordering and timestamps as well and I wanted to fix it >> previously. I wouldn't say that the text-poke change has caused a >> regression, as decoding in this mode was always a bit buggy. >> >> Maybe this is another reason to fix it properly.

2 years, 2 months

[PATCH v2 0/9] coresight: Fix CTI module refcount leak by making it a helper device

by James Clark

Changes since v1: * Don't dereference handle in tmc_etr_get_buffer() when not in perf mode. * Fix some W=1 warnings * Add a commit to rename child/output in terms of local/remote ------------------- Currently there is a refcount leak in CTI when using system wide mode or tracing multithreaded applications. See the last commit for a reproducer. This prevents the module from being unloaded. Historically there have been a few issues and fixes attempted around here which have resulted in some extra logic and a member to keep track of CTI being enabled 'struct coresight_device->ect_enabled'. The fix in commit 665c157e0204 ("coresight: cti: Fix hang in cti_disable_hw()") was also related to CTI having its own enable/disable path which came later than other devices. If we make CTI a helper device and enable helper devices adjacent to the path we get very similar enable/disable behavior to now, but with more reuse of the existing reference counting logic in the coresight core code. This also affects CATU which can have a little bit of its hard coded enable/disable code removed. Enabling CATU on the generic path does require that input connections are tracked so that it can get its associated ETR buffer. Applies to coresight/next (669c4614236a7) but also requires the realloc_array patch here [1]. Also available in full here [2]. [1]: https://lore.kernel.org/linux-arm-kernel/20230306152723.3090195-1-james.cla… [2]: https://gitlab.arm.com/linux-arm/linux-jc/-/tree/james-cs-cti-module-refcou… James Clark (9): coresight: Use enum type for cs_mode wherever possible coresight: Change name of pdata->conns coresight: Rename nr_outports to nr_outconns coresight: Rename connection members to allow for input connections coresight: Dynamically add connections coresight: Store in-connections as well as out-connections coresight: Refactor out buffer allocation function for ETR coresight: Enable and disable helper devices adjacent to the path coresight: Fix CTI module refcount leak by making it a helper device drivers/hwtracing/coresight/coresight-catu.c | 34 +- drivers/hwtracing/coresight/coresight-core.c | 312 +++++++++++------- .../hwtracing/coresight/coresight-cti-core.c | 56 ++-- .../hwtracing/coresight/coresight-cti-sysfs.c | 4 +- drivers/hwtracing/coresight/coresight-cti.h | 4 +- drivers/hwtracing/coresight/coresight-etb10.c | 3 +- .../coresight/coresight-etm3x-core.c | 6 +- .../coresight/coresight-etm4x-core.c | 6 +- .../hwtracing/coresight/coresight-platform.c | 178 +++++++--- drivers/hwtracing/coresight/coresight-priv.h | 9 +- drivers/hwtracing/coresight/coresight-stm.c | 6 +- drivers/hwtracing/coresight/coresight-sysfs.c | 9 +- .../hwtracing/coresight/coresight-tmc-etf.c | 2 +- .../hwtracing/coresight/coresight-tmc-etr.c | 89 ++--- drivers/hwtracing/coresight/coresight-tmc.h | 2 + drivers/hwtracing/coresight/coresight-tpdm.c | 4 +- drivers/hwtracing/coresight/coresight-tpiu.c | 3 +- drivers/hwtracing/coresight/coresight-trbe.c | 3 +- drivers/hwtracing/coresight/ultrasoc-smb.c | 3 +- drivers/hwtracing/coresight/ultrasoc-smb.h | 2 +- include/linux/coresight.h | 109 +++--- 21 files changed, 530 insertions(+), 314 deletions(-) -- 2.34.1

2 years, 3 months

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

CoreSight