- 06 Oct, 2022 40 commits
-
-
Carsten Haitzler authored
Add/improve documentation helping people get started with CoreSight and perf as well as describe the testing and how it works. Reviewed-by:
James Clark <james.clark@arm.com> Signed-off-by:
Carsten Haitzler <carsten.haitzler@arm.com> Cc: Leo Yan <leo.yan@linaro.org> Cc: Mathieu Poirier <mathieu.poirier@linaro.org> Cc: Mike Leach <mike.leach@linaro.org> Cc: Suzuki Poulouse <suzuki.poulose@arm.com> Cc: coresight@lists.linaro.org Cc: linux-doc@vger.kernel.org Link: https://lore.kernel.org/r/20220909152803.2317006-14-carsten.haitzler@foss.arm.comSigned-off-by:
Arnaldo Carvalho de Melo <acme@redhat.com>
-
Carsten Haitzler authored
Ignore other output files of the new CoreSight tests so they don't fill git status with noise we don't need or want. Reviewed-by:
-
Carsten Haitzler authored
This adds scripts to drive the unroll thread tests to compare perf output against a minimum bar of content/quality. Reviewed-by:
-
Carsten Haitzler authored
Add test tool to be driven by further test scripts. This is a simple C based test that is for arm64 with some inline ASM to manually unroll a lot of code to have a very long sequence of commands. Reviewed-by:
-
Carsten Haitzler authored
Add a script to drive the thread loop test that gathers data so it passes a minimum bar (in this case do we get any perf context data for every thread). Reviewed-by:
-
Carsten Haitzler authored
Add test tool to be driven by further test scripts. This is a simple C based loop with threads test to drive from scripts that can output TIDs for tracking/checking. Reviewed-by:
-
Carsten Haitzler authored
Add a script to drive the threaded memcpy test that gathers data so it passes a minimum bar for amount and quality of content that we extract from the kernel's perf support. Reviewed-by:
-
Carsten Haitzler authored
Add test tool to be driven by further test scripts. This is a simple C based memcpy with threads test to drive from scripts. Reviewed-by:
-
Carsten Haitzler authored
Ignore perf output data files generated by perf tests for cleaner git status. Reviewed-by:
-
Carsten Haitzler authored
Add a script to drive the asm pureloop test for arm64/CoreSight that gathers data so it passes a minimum bar for amount and quality of content that we extract from the kernel's perf support. Committer notes: Add the install of tests/shell/coresight/*.sh to tools/perf/Makefile.perf as we're starting to populate that dir. Reviewed-by:
-
Carsten Haitzler authored
Add test tool to be driven by further test scripts. This tool is pure arm64 ASM with no libc usage to ensure it is the same exact binary/code every time so it can also be re-used for many uses. It just loops for a given fixed number of loops. Reviewed-by:
-
Carsten Haitzler authored
This adds the initial build infrastructure (makefiles maintainers information) for adding follow-on tests for CoreSight. Committer notes: Remove the installation of tests/shell/coresight/*.sh, as there are no files there yet and thus, at this point, make install fails. Use $(QUIET_CLEAN) to avoid having extraneous output in the 'make clean' output. Also use @$(MAKE) in tools/perf/tests/shell/coresight/Makefile as $(Q) is not turning into @ when V=1 isn't used, i.e. in the default case it is not being quiet. The >/dev/null in the all for tools/perf/tests/shell/coresight/Makefile is to avoid this: make[4]: Nothing to be done for 'all'. make[4]: Nothing to be done for 'all'. make[4]: Nothing to be done for 'all'. DESCEND plugins GEN /tmp/build/perf/python/perf.so make[4]: Nothing to be done for 'all'. INSTALL trace_plugins On !arm64 where nothing is done on the main target for tools/perf/tests/shell/coresight/*/Makefile. Signed-off-by: -
Carsten Haitzler authored
This adds a library of shell "code" to be shared and used by future tests that target quality testing for Arm CoreSight support in perf and the Linux kernel. Signed-off-by:
-
Leo Yan authored
The test is designed with a data structure with 64-byte alignment, it has two fields "data1" and "data2", and other fields are reserved. Using the "perf mem" command, we can record and report memory samples for a self-contained workload with 1 second duration. If no samples are obtained for the data structure "buf1", it reports failure; and by checking the offset in structure "buf1", if the memory samples aren't for the "data1" and "data2" fields, it means wrong data symbol parsing and returns failure. Committer testing: [root@quaco ~]# grep -m1 "model name" /proc/cpuinfo model name : Intel(R) Core(TM) i7-8650U CPU @ 1.90GHz [root@quaco ~]# [root@quaco ~]# perf test -v "data symbol" 104: Test data symbol : --- start --- test child forked, pid 192318 Compiling test program... Recording workload... [ perf record: Woken up 2 times to write data ] [ perf record: Captured and wrote 0.389 MB /tmp/__perf_test.perf.data.LIuQl (5570 samples) ] Cleaning up files... test child finished with 0 ---- end ---- Test data symbol: Ok [root@quaco ~]# Signed-off-by:
Leo Yan <leo.yan@linaro.org> Tested-by:
-
Namhyung Kim authored
When synthesizing MMAP2 with build-id, it'd read the same file repeatedly as it has no idea if it's done already. Maintain a dsos to check that and skip the file access if possible. Signed-off-by:
Namhyung Kim <namhyung@kernel.org> Cc: Adrian Hunter <adrian.hunter@intel.com> Cc: Ian Rogers <irogers@google.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Link: https://lore.kernel.org/r/20220920222822.2171056-2-namhyung@kernel.orgSigned-off-by:
-
Namhyung Kim authored
The aggr_cpu_id has a thread value but it's actually an index to the thread_map. To reduce possible confusion, rename it to thread_idx. Suggested-by:
Ian Rogers <irogers@google.com> Signed-off-by:
-
Namhyung Kim authored
Now it always uses the global rt_stat. Let's get rid of the field from the saved_value. When the both evsels are NULL, it'd return 0 so remove the block in the saved_value_cmp. Reviewed-by:
-
Namhyung Kim authored
Now it's using the global rt_stat, no need to use per-thread stats. Let get rid of them. Reviewed-by:
-
Namhyung Kim authored
When AGGR_THREAD is active, it aggregates the values for each thread. Previously it used cpu map index which is invalid for AGGR_THREAD so it had to use separate runtime stats with index 0. But it can just use the rt_stat with thread_map_index. Rename the first_shadow_map_idx() and make it return the thread index. Reviewed-by:
-
Namhyung Kim authored
The cpu_map_idx fields is just to differentiate values from other entries. It doesn't need to be strictly cpu map index. Actually we can pass thread map index or aggr map index. So rename the fields first. No functional change intended. Signed-off-by:
-
Namhyung Kim authored
evsel__reset_stat_priv() is called more than once if user gave -r option for multiple runs. But it doesn't need to re-initialize the id. Reviewed-by:
-
Namhyung Kim authored
It uses only one member, no need to have it as an array. Reviewed-by:
-
Namhyung Kim authored
For system-wide evsels, the thread map should be dummy - i.e. it has a single entry of -1. But the code guarantees such a thread map, so no need to handle it specially. No functional change intended. Reviewed-by:
Adrian Hunter <adrian.hunter@intel.com> Signed-off-by:
-
Namhyung Kim authored
Add a help to create a system-wide sched_switch event. One merit is that it sets the system-wide bit before adding it to evlist so that the libperf can handle the cpu and thread maps correctly. Reviewed-by:
-
Namhyung Kim authored
The cpu and thread maps are properly handled in libperf now. No need to do it in the perf tools anymore. Let's remove the logic. Reviewed-by:
-
Namhyung Kim authored
The current code propagate evsel's cpu map settings to evlist when it's added to an evlist. But the evlist->all_cpus and each evsel's cpus will be updated in perf_evlist__set_maps() later. No need to do it before evlist's cpus are set actually. In fact it discards this intermediate all_cpus maps at the beginning of perf_evlist__set_maps(). Let's not do this. It's only needed when an evsel is added after the evlist cpu/thread maps are set. Reviewed-by:
-
Namhyung Kim authored
Setting proper cpu and thread maps for system wide evsels regardless of user requested cpu in __perf_evlist__propagate_maps(). Those evsels need to be active on all cpus always. Do it in the libperf so that we can guarantee it has proper maps. Reviewed-by:
-
Ian Rogers authored
Events remain at v23, and the metrics are based on TMA 4.4 full. Use script at: https://github.com/intel/event-converter-for-linux-perf/blob/master/download_and_gen.py with updates at: https://github.com/captain5050/event-converter-for-linux-perf Updates include: - Switch for core metrics from BDX to BDW. - Switch for Page_Walks_Utilization to BDX version. - Rename of topdown TMA metrics from Frontend_Bound to tma_frontend_bound. - Addition of all 6 levels of TMA metrics. Child metrics are placed in a group named after their parent allowing children of a metric to be easily measured using the metric name with a _group suffix. - ## and ##? operators are correctly expanded. - The locate-with column is added to the long description describing a sampling event. - Metrics are written in terms of other metrics to reduce the expression size and increase readability. Tested with 'perf test': 10: PMU events : 10.1: PMU event table sanity : Ok 10.2: PMU event map aliases : Ok 10.3: Parsing of PMU event table metrics : Ok 10.4: Parsing of PMU event table metrics with fake PMUs : Ok Signed-off-by:
-
Ian Rogers authored
Events remain at v1.07, and the metrics are based on TMA 4.4 full. Use script at: https://github.com/intel/event-converter-for-linux-perf/blob/master/download_and_gen.py with updates at: https://github.com/captain5050/event-converter-for-linux-perf Updates include: - Rename of topdown TMA metrics from Frontend_Bound to tma_frontend_bound. - Addition of all 6 levels of TMA metrics. Previously metrics involving topdown events were dropped. Child metrics are placed in a group named after their parent allowing children of a metric to be easily measured using the metric name with a _group suffix. - ## and ##? operators are correctly expanded. - The locate-with column is added to the long description describing a sampling event. - Metrics are written in terms of other metrics to reduce the expression size and increase readability. Tested with 'perf test': 10: PMU events : 10.1: PMU event table sanity : Ok 10.2: PMU event map aliases : Ok 10.3: Parsing of PMU event table metrics : Ok 10.4: Parsing of PMU event table metrics with fake PMUs : Ok Signed-off-by:
-
Ian Rogers authored
Events remain at v53, and the metrics are based on TMA 4.4 full. Use script at: https://github.com/intel/event-converter-for-linux-perf/blob/master/download_and_gen.py with updates at: https://github.com/captain5050/event-converter-for-linux-perf Updates include: - Rename of topdown TMA metrics from Frontend_Bound to tma_frontend_bound. - _SMT suffix metrics are dropped as the #SMT_On and #EBS_Mode are correctly expanded in the single main metric. - Addition of all 6 levels of TMA metrics. Child metrics are placed in a group named after their parent allowing children of a metric to be easily measured using the metric name with a _group suffix. - ## and ##? operators are correctly expanded. - The locate-with column is added to the long description describing a sampling event. - Metrics are written in terms of other metrics to reduce the expression size and increase readability. Tested with 'perf test': 10: PMU events : 10.1: PMU event table sanity : Ok 10.2: PMU event map aliases : Ok 10.3: Parsing of PMU event table metrics : Ok 10.4: Parsing of PMU event table metrics with fake PMUs : Ok Signed-off-by:
-
Ian Rogers authored
Add cpuid that was added to https://download.01.org/perfmon/mapfile.csvSigned-off-by:
-
Ian Rogers authored
Events are updated to v1.06 the core metrics are based on TMA 4.4 full. Use script at: https://github.com/intel/event-converter-for-linux-perf/blob/master/download_and_gen.py with updates at: https://github.com/captain5050/event-converter-for-linux-perf Updates include: - Rename of topdown TMA metrics from Frontend_Bound to tma_frontend_bound. - Addition of all 6 levels of TMA metrics. Previously metrics involving topdown events were dropped. Child metrics are placed in a group named after their parent allowing children of a metric to be easily measured using the metric name with a _group suffix. - ## and ##? operators are correctly expanded. - The locate-with column is added to the long description describing a sampling event. - Metrics are written in terms of other metrics to reduce the expression size and increase readability. - Latest metrics from: https://github.com/intel/perfmon-metrics Tested with 'perf test': 10: PMU events : 10.1: PMU event table sanity : Ok 10.2: PMU event map aliases : Ok 10.3: Parsing of PMU event table metrics : Ok 10.4: Parsing of PMU event table metrics with fake PMUs : Ok Signed-off-by:
-
Ian Rogers authored
Events remain at v17, and the metrics are based on TMA 4.4 full. Use script at: https://github.com/intel/event-converter-for-linux-perf/blob/master/download_and_gen.py with updates at: https://github.com/captain5050/event-converter-for-linux-perf Updates include: - Rename of topdown TMA metrics from Frontend_Bound to tma_frontend_bound. - _SMT suffix metrics are dropped as the #SMT_On and #EBS_Mode are correctly expanded in the single main metric. - Addition of all 6 levels of TMA metrics. Child metrics are placed in a group named after their parent allowing children of a metric to be easily measured using the metric name with a _group suffix. - ## and ##? operators are correctly expanded. - The locate-with column is added to the long description describing a sampling event. - Metrics are written in terms of other metrics to reduce the expression size and increase readability. Tested with 'perf test': 10: PMU events : 10.1: PMU event table sanity : Ok 10.2: PMU event map aliases : Ok 10.3: Parsing of PMU event table metrics : Ok 10.4: Parsing of PMU event table metrics with fake PMUs : Ok Signed-off-by:
-
Ian Rogers authored
Events remain at v21, and the metrics are based on TMA 4.4 full. Use script at: https://github.com/intel/event-converter-for-linux-perf/blob/master/download_and_gen.py with updates at: https://github.com/captain5050/event-converter-for-linux-perf Updates include: - Rename of topdown TMA metrics from Frontend_Bound to tma_frontend_bound. - _SMT suffix metrics are dropped as the #SMT_On and #EBS_Mode are correctly expanded in the single main metric. - Addition of all 6 levels of TMA metrics. Child metrics are placed in a group named after their parent allowing children of a metric to be easily measured using the metric name with a _group suffix. - ## and ##? operators are correctly expanded. - The locate-with column is added to the long description describing a sampling event. - Metrics are written in terms of other metrics to reduce the expression size and increase readability. Tested with 'perf test': 10: PMU events : 10.1: PMU event table sanity : Ok 10.2: PMU event map aliases : Ok 10.3: Parsing of PMU event table metrics : Ok 10.4: Parsing of PMU event table metrics with fake PMUs : Ok Signed-off-by:
-
Ian Rogers authored
Events are updated to v22 the core metrics are based on TMA 4.4 full. Use script at: https://github.com/intel/event-converter-for-linux-perf/blob/master/download_and_gen.py with updates at: https://github.com/captain5050/event-converter-for-linux-perf Updates include: - Rename of topdown TMA metrics from Frontend_Bound to tma_frontend_bound. - _SMT suffix metrics are dropped as the #SMT_On and #EBS_Mode are correctly expanded in the single main metric. - Addition of all 6 levels of TMA metrics. Child metrics are placed in a group named after their parent allowing children of a metric to be easily measured using the metric name with a _group suffix. - ## and ##? operators are correctly expanded. - The locate-with column is added to the long description describing a sampling event. - Metrics are written in terms of other metrics to reduce the expression size and increase readability. Tested with 'perf test': 10: PMU events : 10.1: PMU event table sanity : Ok 10.2: PMU event map aliases : Ok 10.3: Parsing of PMU event table metrics : Ok 10.4: Parsing of PMU event table metrics with fake PMUs : Ok Signed-off-by:
-
Ian Rogers authored
Events remain at v22, and the metrics are based on TMA 4.4 full. Use script at: https://github.com/intel/event-converter-for-linux-perf/blob/master/download_and_gen.py with updates at: https://github.com/captain5050/event-converter-for-linux-perf Updates include: - Rename of topdown TMA metrics from Frontend_Bound to tma_frontend_bound. - _SMT suffix metrics are dropped as the #SMT_On and #EBS_Mode are correctly expanded in the single main metric. - Addition of all 6 levels of TMA metrics. Child metrics are placed in a group named after their parent allowing children of a metric to be easily measured using the metric name with a _group suffix. - ## and ##? operators are correctly expanded. - The locate-with column is added to the long description describing a sampling event. - Metrics are written in terms of other metrics to reduce the expression size and increase readability. Tested with 'perf test': 10: PMU events : 10.1: PMU event table sanity : Ok 10.2: PMU event map aliases : Ok 10.3: Parsing of PMU event table metrics : Ok 10.4: Parsing of PMU event table metrics with fake PMUs : Ok Signed-off-by:
-
Ian Rogers authored
Events are updated to v1.16 the metrics are based on TMA 4.4 full. Use script at: https://github.com/intel/event-converter-for-linux-perf/blob/master/download_and_gen.py with updates at: https://github.com/captain5050/event-converter-for-linux-perf Updates include: - Rename of topdown TMA metrics from Frontend_Bound to tma_frontend_bound. - Addition of all 6 levels of TMA metrics. Previously metrics involving topdown events were dropped. Child metrics are placed in a group named after their parent allowing children of a metric to be easily measured using the metric name with a _group suffix. - ## and ##? operators are correctly expanded. - The locate-with column is added to the long description describing a sampling event. - Metrics are written in terms of other metrics to reduce the expression size and increase readability. - Latest metrics from: https://github.com/intel/perfmon-metrics Tested with 'perf test': 10: PMU events : 10.1: PMU event table sanity : Ok 10.2: PMU event map aliases : Ok 10.3: Parsing of PMU event table metrics : Ok 10.4: Parsing of PMU event table metrics with fake PMUs : Ok Signed-off-by:
-
Ian Rogers authored
Events are updated to v1.15, the metrics are based on TMA 4.4 full. Use script at: https://github.com/intel/event-converter-for-linux-perf/blob/master/download_and_gen.py with updates at: https://github.com/captain5050/event-converter-for-linux-perf Updates include: - Rename of topdown TMA metrics from Frontend_Bound to tma_frontend_bound. - Addition of all 6 levels of TMA metrics. Previously metrics involving topdown events were dropped. Child metrics are placed in a group named after their parent allowing children of a metric to be easily measured using the metric name with a _group suffix. - ## and ##? operators are correctly expanded. - The locate-with column is added to the long description describing a sampling event. - Metrics are written in terms of other metrics to reduce the expression size and increase readability. Tested with 'perf test': 10: PMU events : 10.1: PMU event table sanity : Ok 10.2: PMU event map aliases : Ok 10.3: Parsing of PMU event table metrics : Ok 10.4: Parsing of PMU event table metrics with fake PMUs : Ok Signed-off-by:
-
Ian Rogers authored
Events are updated to v26, the core metrics are based on TMA 4.4 full. Use script at: https://github.com/intel/event-converter-for-linux-perf/blob/master/download_and_gen.py with updates at: https://github.com/captain5050/event-converter-for-linux-perf Updates include: - Uncore event updates by Zhengjun Xing <zhengjun.xing@linux.intel.com>. - Rename of topdown TMA metrics from Frontend_Bound to tma_frontend_bound. - _SMT suffix metrics are dropped as the #SMT_On and #EBS_Mode are correctly expanded in the single main metric. - Addition of all 6 levels of TMA metrics. Previously metrics involving topdown events were dropped. Child metrics are placed in a group named after their parent allowing children of a metric to be easily measured using the metric name with a _group suffix. - ## and ##? operators are correctly expanded. - The locate-with column is added to the long description describing a sampling event. - Metrics are written in terms of other metrics to reduce the expression size and increase readability. - Latest metrics from: https://github.com/intel/perfmon-metrics Tested with 'perf test': 10: PMU events : 10.1: PMU event table sanity : Ok 10.2: PMU event map aliases : Ok 10.3: Parsing of PMU event table metrics : Ok 10.4: Parsing of PMU event table metrics with fake PMUs : Ok Signed-off-by:
-
Ian Rogers authored
Events are updated to v32, the core metrics are based on TMA 4.4 full. Use script at: https://github.com/intel/event-converter-for-linux-perf/blob/master/download_and_gen.py with updates at: https://github.com/captain5050/event-converter-for-linux-perf Updates include: - Rename of topdown TMA metrics from Frontend_Bound to tma_frontend_bound. - _SMT suffix metrics are dropped as the #SMT_On and #EBS_Mode are correctly expanded in the single main metric. - Addition of all 6 levels of TMA metrics. Child metrics are placed in a group named after their parent allowing children of a metric to be easily measured using the metric name with a _group suffix. - ## and ##? operators are correctly expanded. - The locate-with column is added to the long description describing a sampling event. - Metrics are written in terms of other metrics to reduce the expression size and increase readability. Tested with 'perf test': 10: PMU events : 10.1: PMU event table sanity : Ok 10.2: PMU event map aliases : Ok 10.3: Parsing of PMU event table metrics : Ok 10.4: Parsing of PMU event table metrics with fake PMUs : Ok Signed-off-by:
-
