CoreSight

coresight@lists.linaro.org

6 participants
2390 discussions

[PATCH v3 00/13] coresight: Add new API to allocate trace source ID values

by Mike Leach

The current method for allocating trace source ID values to sources is to use a fixed algorithm for CPU based sources of (cpu_num * 2 + 0x10). The STM is allocated ID 0x1. This fixed algorithm is used in both the CoreSight driver code, and by perf when writing the trace metadata in the AUXTRACE_INFO record. The method needs replacing as currently:- 1. It is inefficient in using available IDs. 2. Does not scale to larger systems with many cores and the algorithm has no limits so will generate invalid trace IDs for cpu number > 44. Additionally requirements to allocate additional system IDs on some systems have been seen. This patch set introduces an API that allows the allocation of trace IDs in a dynamic manner. Architecturally reserved IDs are never allocated, and the system is limited to allocating only valid IDs. Each of the current trace sources ETM3.x, ETM4.x and STM is updated to use the new API. For the ETMx.x devices IDs are allocated on certain events a) When using sysfs, an ID will be allocated on hardware enable, or a read of sysfs TRCTRACEID register and freed when the sysfs reset is written. b) When using perf, ID is allocated on during setup AUX event, and freed on event free. IDs are communicated using the AUX_OUTPUT_HW_ID packet. The ID allocator is notified when perf sessions start and stop so CPU based IDs are kept constant throughout any perf session. Note: This patchset breaks some backward compatibility for perf record and perf report. The version of the AUXTRACE_INFO has been updated to reflect the fact that the trace source IDs are generated differently. This will mean older versions of perf report cannot decode the newer file. Applies to coresight/next [4d45bc82df66] Tested on DB410c Changes since v2: 1) Improved backward compatibility: (requested by James) Using the new version of perf on an old kernel will generate a usable file legacy metadata values are set by the new perf and will be used if mew ID packets are not present in the file. Using an older version of perf / simpleperf on an updated kernel may still work. The trace ID allocator has been updated to use the legacy ID values where possible, so generated file and used trace IDs will match up to the point where the legacy algorithm is broken anyway. 2) Various changes to the ID allocator and ID packet format. (suggested by Suzuki) 3) per CPU ID info in allocator now stored as atomic type to allow a passive read without taking the allocator spinlock. perf flow now allocates and releases ID values in setup_aux / free_event. Device enable and event enable use the passive read to set the allocated values. This simplifies the locking mechanisms on the perf run and fixes issues that arose with locking dependencies. Changes since v1: (after feedback & discussion with Mathieu & Suzuki). 1) API has changed. The global trace ID map is managed internally, so it is no longer passed in to the API functions. 2) perf record does not use sysfs to find the trace IDs. These are now output as AUX_OUTPUT_HW_ID events. The drivers, perf record, and perf report have been updated accordingly to generate and handle these events. Mike Leach (13): coresight: trace-id: Add API to dynamically assign Trace ID values coresight: Remove obsolete Trace ID unniqueness checks coresight: stm: Update STM driver to use Trace ID API coresight: etm4x: Update ETM4 driver to use Trace ID API coresight: etm3x: Update ETM3 driver to use Trace ID API coresight: etmX.X: stm: Remove trace_id() callback coresight: perf: traceid: Add perf notifiers for Trace ID 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 kernel: events: Export perf_report_aux_output_id() perf: cs-etm: Handle PERF_RECORD_AUX_OUTPUT_HW_ID packet coresight: events: PERF_RECORD_AUX_OUTPUT_HW_ID used for Trace ID coresight: trace-id: Add debug & test macros to Trace ID allocation drivers/hwtracing/coresight/Makefile | 2 +- drivers/hwtracing/coresight/coresight-core.c | 49 +-- .../hwtracing/coresight/coresight-etm-perf.c | 23 ++ drivers/hwtracing/coresight/coresight-etm.h | 3 +- .../coresight/coresight-etm3x-core.c | 92 +++-- .../coresight/coresight-etm3x-sysfs.c | 27 +- .../coresight/coresight-etm4x-core.c | 79 ++++- .../coresight/coresight-etm4x-sysfs.c | 27 +- drivers/hwtracing/coresight/coresight-etm4x.h | 3 + drivers/hwtracing/coresight/coresight-stm.c | 49 +-- .../hwtracing/coresight/coresight-trace-id.c | 266 ++++++++++++++ .../hwtracing/coresight/coresight-trace-id.h | 78 +++++ include/linux/coresight-pmu.h | 35 +- include/linux/coresight.h | 3 - kernel/events/core.c | 1 + tools/include/linux/coresight-pmu.h | 48 ++- 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 | 331 +++++++++++++++--- tools/perf/util/cs-etm.h | 14 +- 20 files changed, 933 insertions(+), 225 deletions(-) create mode 100644 drivers/hwtracing/coresight/coresight-trace-id.c create mode 100644 drivers/hwtracing/coresight/coresight-trace-id.h -- 2.17.1

2 years, 9 months

Re: [PATCH] coresight: Fix possible deadlock with lock dependency

by Suzuki K Poulose

On 21/07/2022 14:03, Sudeep Holla wrote: > With lockdeps enabled, we get the following warning: > > ====================================================== > WARNING: possible circular locking dependency detected > ------------------------------------------------------ > kworker/u12:1/53 is trying to acquire lock: > ffff80000adce220 (coresight_mutex){+.+.}-{4:4}, at: coresight_set_assoc_ectdev_mutex+0x3c/0x5c > but task is already holding lock: > ffff80000add1f60 (ect_mutex){+.+.}-{4:4}, at: cti_probe+0x318/0x394 > > which lock already depends on the new lock. > the existing dependency chain (in reverse order) is: > > -> #1 (ect_mutex){+.+.}-{4:4}: > __mutex_lock_common+0xd8/0xe60 > mutex_lock_nested+0x44/0x50 > cti_add_assoc_to_csdev+0x4c/0x184 > coresight_register+0x2f0/0x314 > tmc_probe+0x33c/0x414 > > -> #0 (coresight_mutex){+.+.}-{4:4}: > __lock_acquire+0x1a20/0x32d0 > lock_acquire+0x160/0x308 > __mutex_lock_common+0xd8/0xe60 > mutex_lock_nested+0x44/0x50 > coresight_set_assoc_ectdev_mutex+0x3c/0x5c > cti_update_conn_xrefs+0x6c/0xf8 > cti_probe+0x33c/0x394 > > other info that might help us debug this: > Possible unsafe locking scenario: > CPU0 CPU1 > ---- ---- > lock(ect_mutex); > lock(coresight_mutex); > lock(ect_mutex); > lock(coresight_mutex); > *** DEADLOCK *** > > 4 locks held by kworker/u12:1/53: > #0: ((wq_completion)events_unbound){+.+.}-{0:0}, at: process_one_work+0x1fc/0x63c > #1: (deferred_probe_work){+.+.}-{0:0}, at: process_one_work+0x228/0x63c > #2: (&dev->mutex){....}-{4:4}, at: __device_attach+0x48/0x1a8 > #3: (ect_mutex){+.+.}-{4:4}, at: cti_probe+0x318/0x394 > > To fix the same, call cti_add_assoc_to_csdev without the holding > coresight_mutex and confine the locking while setting the associated > ect / cti device using coresight_set_assoc_ectdev_mutex(). > > 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> > Signed-off-by: Sudeep Holla <sudeep.holla(a)arm.com> Thanks for the fix. I will push this out at rc1 as fixes. Thanks Suzuki

2 years, 9 months

Re: [PATCH linux-next] coresight: use sysfs_emit() to instead of scnprintf()

by James Clark

On 27/09/2022 12:51, zhangsongyi.cgel(a)gmail.com wrote: > From: zhang songyi <zhang.songyi(a)zte.com.cn> > > Follow the advice of the Documentation/filesystems/sysfs.rst and show() > should only use sysfs_emit() or sysfs_emit_at() when formatting the value > to be returned to user space. > > Reported-by: Zeal Robot <zealci(a)zte.com.cn> > Signed-off-by: zhang songyi <zhang.songyi(a)zte.com.cn> > --- > drivers/hwtracing/coresight/coresight-stm.c | 8 ++++---- > 1 file changed, 4 insertions(+), 4 deletions(-) > Reviewed-by: James Clark <james.clark(a)arm.com> > diff --git a/drivers/hwtracing/coresight/coresight-stm.c b/drivers/hwtracing/coresight/coresight-stm.c > index 463f449cfb79..3a1b07217932 100644 > --- a/drivers/hwtracing/coresight/coresight-stm.c > +++ b/drivers/hwtracing/coresight/coresight-stm.c > @@ -466,7 +466,7 @@ static ssize_t hwevent_enable_show(struct device *dev, > struct stm_drvdata *drvdata = dev_get_drvdata(dev->parent); > unsigned long val = drvdata->stmheer; > > - return scnprintf(buf, PAGE_SIZE, "%#lx\n", val); > + return sysfs_emit(buf, "%#lx\n", val); > } > > static ssize_t hwevent_enable_store(struct device *dev, > @@ -495,7 +495,7 @@ static ssize_t hwevent_select_show(struct device *dev, > struct stm_drvdata *drvdata = dev_get_drvdata(dev->parent); > unsigned long val = drvdata->stmhebsr; > > - return scnprintf(buf, PAGE_SIZE, "%#lx\n", val); > + return sysfs_emit(buf, "%#lx\n", val); > } > > static ssize_t hwevent_select_store(struct device *dev, > @@ -530,7 +530,7 @@ static ssize_t port_select_show(struct device *dev, > spin_unlock(&drvdata->spinlock); > } > > - return scnprintf(buf, PAGE_SIZE, "%#lx\n", val); > + return sysfs_emit(buf, "%#lx\n", val); > } > > static ssize_t port_select_store(struct device *dev, > @@ -577,7 +577,7 @@ static ssize_t port_enable_show(struct device *dev, > spin_unlock(&drvdata->spinlock); > } > > - return scnprintf(buf, PAGE_SIZE, "%#lx\n", val); > + return sysfs_emit(buf, "%#lx\n", val); > } > > static ssize_t port_enable_store(struct device *dev,

2 years, 9 months

[PATCH v2 0/1] arm64: defconfig: Add Coresight as module

by James Clark

I'm still leaving out CONFIG_CORESIGHT_SOURCE_ETM4X because it depends on the outcome of the investigation into CONFIG_PID_IN_CONTEXTIDR, but I think we should enable these ones for now and start getting some of the benefits sooner. Changes since v1: * Remove CONFIG_CORESIGHT_CTI_INTEGRATION_REGS=y which shouldn't be enabled by default ----- As suggested by Catalin here's the change to add Coresight to defconfig. Unfortunately I don't think we should add CONFIG_CORESIGHT_SOURCE_ETM4X which builds a few files until [1] is merged because of the overhead of CONFIG_PID_IN_CONTEXTIDR. [1]: https://lore.kernel.org/lkml/20211021134530.206216-1-leo.yan@linaro.org/T/ applies to arm64/for-next/core (e99db032d186) James Clark (1): arm64: defconfig: Add Coresight as module arch/arm64/configs/defconfig | 8 ++++++++ 1 file changed, 8 insertions(+) -- 2.28.0

2 years, 9 months

Re: [PATCH 0/1] arm64: defconfig: Add Coresight as module

by James Clark

On 22/09/2022 11:52, Catalin Marinas wrote: > On Thu, Sep 22, 2022 at 10:34:45AM +0100, James Clark wrote: >> On 21/09/2022 16:08, Catalin Marinas wrote: >>> 2. Always on CONFIG_PID_IN_CONTEXTIDR (we might as well remove the >>> Kconfig entry). This would write the root pid namespace value >>> (task_pid_nr()). >> >> If we're not worried about the overhead after all, this would be the >> easiest solution. And then SPE or Coresight already decide whether they >> want to use the value or not, so no further changes are needed. >> >> From Leo's patch there is a table that shows a 1% overhead with it >> enabled permanently, and I've heard a figure like that mentioned before. >> So I could also resurrect that patch to use static keys? Although it's a >> bit more complicated, that would be my preference. And then we can have >> that mode always on. > > I don't think we should bother with static keys, just always enable it > but try to reduce/group the ISBs from all the functions called on the > __switch_to() path. We may actually get a speed-up. > Ok thanks I will take a look at this

2 years, 9 months

Re: [PATCH 0/1] arm64: defconfig: Add Coresight as module

by James Clark

On 21/09/2022 16:08, Catalin Marinas wrote: > On Wed, Sep 21, 2022 at 03:05:34PM +0100, James Clark wrote: >> As suggested by Catalin here's the change to add Coresight to defconfig. >> >> Unfortunately I don't think we should add CONFIG_CORESIGHT_SOURCE_ETM4X >> which builds a few files until [1] is merged because of the overhead >> of CONFIG_PID_IN_CONTEXTIDR. >> >> [1]: https://lore.kernel.org/lkml/20211021134530.206216-1-leo.yan@linaro.org/T/ > > I thought the overhead wasn't the problem, it's mostly negligible. We > can probably save a few more cycles on the __switch_to() path by > replacing several isb()s in those functions with a single one just > before cpu_switch_to(). > > IIRC the issue is that unless a process runs in the root pid namespace, > the actual pid written to contextidr is meaningless. This is true, and Leo has recently disabled it in that scenario in aab473867fed. > > Now that you reminded me of that thread, I see three options (sorry, not > entirely related to the defconfig updates): > > 1. Remove CONFIG_PID_IN_CONTEXTIDR and corresponding code completely, > find other events to correlate the task with the trace. Unfortunately when tracing per core we would need kernel timestamps in the trace to correlate to the switch records. At the moment Coresight is using a different clock source so it's not possible and we're still using the context ID to correlate samples. With FEAT_TRF in v8.4 it will be possible to do this and we've started working towards that here: 0f00b223ea22. But we'd still have to support older hardware too, so CONFIG_PID_IN_CONTEXTIDR can't be removed completely. For SPE it's not required because we already have the right timestamps in the samples and we've added support for no context IDs in the decoder here: 27d113cfe892 > > 2. Always on CONFIG_PID_IN_CONTEXTIDR (we might as well remove the > Kconfig entry). This would write the root pid namespace value > (task_pid_nr()). If we're not worried about the overhead after all, this would be the easiest solution. And then SPE or Coresight already decide whether they want to use the value or not, so no further changes are needed. From Leo's patch there is a table that shows a 1% overhead with it enabled permanently, and I've heard a figure like that mentioned before. So I could also resurrect that patch to use static keys? Although it's a bit more complicated, that would be my preference. And then we can have that mode always on. > > 3. Similar to (2) but instead write task_pid_nr_ns(). An alternative > here is to write -1 if the task is not in the root pid namespace. > > Strong preference for (1). >

2 years, 9 months

Re: [PATCH 1/1] arm64: defconfig: Add Coresight as module

by James Clark

On 21/09/2022 17:46, Mathieu Poirier wrote: > On Wed, Sep 21, 2022 at 04:26:59PM +0100, Mark Brown wrote: >> On Wed, Sep 21, 2022 at 03:05:35PM +0100, James Clark wrote: >> >>> +CONFIG_CORESIGHT_CTI=m >>> +CONFIG_CORESIGHT_CTI_INTEGRATION_REGS=y >> > > I agree - integration registers should not be enabled by default. > >> Do we want this turned on by default? According to the >> description it's a bit dangerous and it's exposed via sysfs >> rather than debugfs. > > Should I disable just CONFIG_CORESIGHT_CTI_INTEGRATION_REGS or CONFIG_CORESIGHT_CTI as well? There are other writable registers exposed via sysfs outside of these two options, so I wanted to check if it's just the integration registers that are the issue.

2 years, 9 months

[PATCH 0/1] arm64: defconfig: Add Coresight as module

by James Clark

As suggested by Catalin here's the change to add Coresight to defconfig. Unfortunately I don't think we should add CONFIG_CORESIGHT_SOURCE_ETM4X which builds a few files until [1] is merged because of the overhead of CONFIG_PID_IN_CONTEXTIDR. [1]: https://lore.kernel.org/lkml/20211021134530.206216-1-leo.yan@linaro.org/T/ James Clark (1): arm64: defconfig: Add Coresight as module arch/arm64/configs/defconfig | 9 +++++++++ 1 file changed, 9 insertions(+) -- 2.28.0

2 years, 9 months

Re: coresight-etm4x-core.c:969:26: error: 'ID_AA64DFR0_TRACEVER_SHIFT' undeclared

by James Clark

On 20/09/2022 12:39, Naresh Kamboju wrote: > Following build warnings / errors noticed while building arm64 coresight > on linux next-20220919 and next-20220920. Fix is here: https://lists.linaro.org/archives/list/coresight@lists.linaro.org/thread/OS… > > Reported-by: Linux Kernel Functional Testing <lkft(a)linaro.org> > > make --silent --keep-going --jobs=8 > O=/home/tuxbuild/.cache/tuxmake/builds/1/build > CROSS_COMPILE_COMPAT=arm-linux-gnueabihf- ARCH=arm64 > CROSS_COMPILE=aarch64-linux-gnu- 'CC=sccache aarch64-linux-gnu-gcc' > 'HOSTCC=sccache gcc' > > drivers/hwtracing/coresight/coresight-etm4x-core.c: In function > 'cpu_supports_sysreg_trace': > drivers/hwtracing/coresight/coresight-etm4x-core.c:969:26: error: > 'ID_AA64DFR0_TRACEVER_SHIFT' undeclared (first use in this function); > did you mean 'ID_AA64DFR0_EL1_TraceVer_SHIFT'? > 969 | return ((dfr0 >> ID_AA64DFR0_TRACEVER_SHIFT) & 0xfUL) > 0; > | ^~~~~~~~~~~~~~~~~~~~~~~~~~ > | ID_AA64DFR0_EL1_TraceVer_SHIFT > drivers/hwtracing/coresight/coresight-etm4x-core.c:969:26: note: each > undeclared identifier is reported only once for each function it > appears in > drivers/hwtracing/coresight/coresight-etm4x-core.c: In function > 'cpu_detect_trace_filtering': > drivers/hwtracing/coresight/coresight-etm4x-core.c:1057:57: error: > 'ID_AA64DFR0_TRACE_FILT_SHIFT' undeclared (first use in this > function); did you mean 'ID_AA64PFR0_EL1_DIT_SHIFT'? > 1057 | if (!cpuid_feature_extract_unsigned_field(dfr0, > ID_AA64DFR0_TRACE_FILT_SHIFT)) > | > ^~~~~~~~~~~~~~~~~~~~~~~~~~~~ > | > ID_AA64PFR0_EL1_DIT_SHIFT > drivers/hwtracing/coresight/coresight-etm4x-core.c: In function > 'cpu_supports_sysreg_trace': > drivers/hwtracing/coresight/coresight-etm4x-core.c:970:1: error: > control reaches end of non-void function [-Werror=return-type] > 970 | } > | ^ > cc1: some warnings being treated as errors > > Regressions found on arm64: > - build-clang-13-lkftconfig > - build-gcc-11-lkftconfig-devicetree > - build-gcc-11-lkftconfig-kunit > - build-gcc-11-lkftconfig-perf > - build-gcc-11-lkftconfig-64k_page_size > - build-gcc-11-lkftconfig-debug > - build-gcc-11-lkftconfig-libgpiod > - build-gcc-11-lkftconfig-debug-kmemleak > - build-clang-14-lkftconfig > - build-clang-nightly-lkftconfig > - build-gcc-11-lkftconfig > - build-gcc-11-lkftconfig-kasan > - build-clang-12-lkftconfig > - build-gcc-11-lkftconfig-kselftest-kernel > - build-gcc-11-lkftconfig-rcutorture > - build-gcc-11-lkftconfig-armv8_features > - build-gcc-11-lkftconfig-kselftest > > Build: https://builds.tuxbuild.com/2F1cW8NpQ0Z6l9h9rfkZT5AXzqg/ > config: https://builds.tuxbuild.com/2F1cW8NpQ0Z6l9h9rfkZT5AXzqg/config > > > -- > Linaro LKFT > https://lkft.linaro.org > > _______________________________________________ > linux-arm-kernel mailing list > linux-arm-kernel(a)lists.infradead.org > http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

2 years, 9 months

[PATCH] coresight: Use new autogenerated sysreg definitions

by James Clark

Fix the build by updating registers to use the new autogenerated format. Fixes: c0357a73fa4a ("arm64/sysreg: Align field names in ID_AA64DFR0_EL1 with architecture") Signed-off-by: James Clark <james.clark(a)arm.com> --- drivers/hwtracing/coresight/coresight-etm4x-core.c | 4 ++-- drivers/hwtracing/coresight/coresight-trbe.h | 3 ++- 2 files changed, 4 insertions(+), 3 deletions(-) diff --git a/drivers/hwtracing/coresight/coresight-etm4x-core.c b/drivers/hwtracing/coresight/coresight-etm4x-core.c index d39660a3e50c..80fefaba58ee 100644 --- a/drivers/hwtracing/coresight/coresight-etm4x-core.c +++ b/drivers/hwtracing/coresight/coresight-etm4x-core.c @@ -966,7 +966,7 @@ static inline bool cpu_supports_sysreg_trace(void) { u64 dfr0 = read_sysreg_s(SYS_ID_AA64DFR0_EL1); - return ((dfr0 >> ID_AA64DFR0_TRACEVER_SHIFT) & 0xfUL) > 0; + return ((dfr0 >> ID_AA64DFR0_EL1_TraceVer_SHIFT) & 0xfUL) > 0; } static bool etm4_init_sysreg_access(struct etmv4_drvdata *drvdata, @@ -1054,7 +1054,7 @@ static void cpu_detect_trace_filtering(struct etmv4_drvdata *drvdata) u64 trfcr; drvdata->trfcr = 0; - if (!cpuid_feature_extract_unsigned_field(dfr0, ID_AA64DFR0_TRACE_FILT_SHIFT)) + if (!cpuid_feature_extract_unsigned_field(dfr0, ID_AA64DFR0_EL1_TraceFilt_SHIFT)) return; /* diff --git a/drivers/hwtracing/coresight/coresight-trbe.h b/drivers/hwtracing/coresight/coresight-trbe.h index 30e4d7db4f8e..98ff1b17ad07 100644 --- a/drivers/hwtracing/coresight/coresight-trbe.h +++ b/drivers/hwtracing/coresight/coresight-trbe.h @@ -20,7 +20,8 @@ static inline bool is_trbe_available(void) { u64 aa64dfr0 = read_sysreg_s(SYS_ID_AA64DFR0_EL1); - unsigned int trbe = cpuid_feature_extract_unsigned_field(aa64dfr0, ID_AA64DFR0_TRBE_SHIFT); + unsigned int trbe = cpuid_feature_extract_unsigned_field(aa64dfr0, + ID_AA64DFR0_EL1_TraceBuffer_SHIFT); return trbe >= 0b0001; } -- 2.28.0

2 years, 9 months

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

CoreSight