forked from Minki/linux
d56354dc49
To fully annotate BPF programs with source code mapping, 4 different information are needed: 1) PERF_RECORD_KSYMBOL 2) PERF_RECORD_BPF_EVENT 3) bpf_prog_info 4) btf This patch handles 3) and 4) for BPF programs loaded after 'perf record|top'. For timely process of these information, a dedicated event is added to the side band evlist. When PERF_RECORD_BPF_EVENT is received via the side band event, the polling thread gathers 3) and 4) vis sys_bpf and store them in perf_env. This information is saved to perf.data at the end of 'perf record'. Committer testing: The 'wakeup_watermark' member in 'struct perf_event_attr' is inside a unnamed union, so can't be used in a struct designated initialization with older gccs, get it out of that, isolating as 'attr.wakeup_watermark = 1;' to work with all gcc versions. We also need to add '--no-bpf-event' to the 'perf record' perf_event_attr tests in 'perf test', as the way that that test goes is to intercept the events being setup and looking if they match the fields described in the control files, since now it finds first the side band event used to catch the PERF_RECORD_BPF_EVENT, they all fail. With these issues fixed: Same scenario as for testing BPF programs loaded before 'perf record' or 'perf top' starts, only start the BPF programs after 'perf record|top', so that its information get collected by the sideband threads, the rest works as for the programs loaded before start monitoring. Add missing 'inline' to the bpf_event__add_sb_event() when HAVE_LIBBPF_SUPPORT is not defined, fixing the build in systems without binutils devel files installed. Signed-off-by: Song Liu <songliubraving@fb.com> Reviewed-by: Jiri Olsa <jolsa@kernel.org> Tested-by: Arnaldo Carvalho de Melo <acme@redhat.com> Cc: Alexei Starovoitov <ast@kernel.org> Cc: Daniel Borkmann <daniel@iogearbox.net> Cc: Jiri Olsa <jolsa@kernel.org> Cc: Namhyung Kim <namhyung@kernel.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Stanislav Fomichev <sdf@google.com> Link: http://lkml.kernel.org/r/20190312053051.2690567-16-songliubraving@fb.com Signed-off-by: Arnaldo Carvalho de Melo <acme@redhat.com> |
||
---|---|---|
.. | ||
base-record | ||
base-stat | ||
README | ||
test-record-basic | ||
test-record-branch-any | ||
test-record-branch-filter-any | ||
test-record-branch-filter-any_call | ||
test-record-branch-filter-any_ret | ||
test-record-branch-filter-hv | ||
test-record-branch-filter-ind_call | ||
test-record-branch-filter-k | ||
test-record-branch-filter-u | ||
test-record-C0 | ||
test-record-count | ||
test-record-data | ||
test-record-freq | ||
test-record-graph-default | ||
test-record-graph-dwarf | ||
test-record-graph-fp | ||
test-record-group | ||
test-record-group1 | ||
test-record-group-sampling | ||
test-record-no-buffering | ||
test-record-no-inherit | ||
test-record-no-samples | ||
test-record-period | ||
test-record-raw | ||
test-stat-basic | ||
test-stat-C0 | ||
test-stat-default | ||
test-stat-detailed-1 | ||
test-stat-detailed-2 | ||
test-stat-detailed-3 | ||
test-stat-group | ||
test-stat-group1 | ||
test-stat-no-inherit |
The struct perf_event_attr test (attr tests) support ==================================================== This testing support is embedded into perf directly and is governed by the PERF_TEST_ATTR environment variable and hook inside the sys_perf_event_open function. The general idea is to store 'struct perf_event_attr' details for each event created within single perf command. Each event details are stored into separate text file. Once perf command is finished these files are checked for values we expect for command. The attr tests consist of following parts: tests/attr.c ------------ This is the sys_perf_event_open hook implementation. The hook is triggered when the PERF_TEST_ATTR environment variable is defined. It must contain name of existing directory with access and write permissions. For each sys_perf_event_open call event details are stored in separate file. Besides 'struct perf_event_attr' values we also store 'fd' and 'group_fd' values to allow checking for groups. tests/attr.py ------------- This is the python script that does all the hard work. It reads the test definition, executes it and checks results. tests/attr/ ----------- Directory containing all attr test definitions. Following tests are defined (with perf commands): perf record kill (test-record-basic) perf record -b kill (test-record-branch-any) perf record -j any kill (test-record-branch-filter-any) perf record -j any_call kill (test-record-branch-filter-any_call) perf record -j any_ret kill (test-record-branch-filter-any_ret) perf record -j hv kill (test-record-branch-filter-hv) perf record -j ind_call kill (test-record-branch-filter-ind_call) perf record -j k kill (test-record-branch-filter-k) perf record -j u kill (test-record-branch-filter-u) perf record -c 123 kill (test-record-count) perf record -d kill (test-record-data) perf record -F 100 kill (test-record-freq) perf record -g kill (test-record-graph-default) perf record --call-graph dwarf kill (test-record-graph-dwarf) perf record --call-graph fp kill (test-record-graph-fp) perf record --group -e cycles,instructions kill (test-record-group) perf record -e '{cycles,instructions}' kill (test-record-group1) perf record -D kill (test-record-no-delay) perf record -i kill (test-record-no-inherit) perf record -n kill (test-record-no-samples) perf record -c 100 -P kill (test-record-period) perf record -R kill (test-record-raw) perf stat -e cycles kill (test-stat-basic) perf stat kill (test-stat-default) perf stat -d kill (test-stat-detailed-1) perf stat -dd kill (test-stat-detailed-2) perf stat -ddd kill (test-stat-detailed-3) perf stat --group -e cycles,instructions kill (test-stat-group) perf stat -e '{cycles,instructions}' kill (test-stat-group1) perf stat -i -e cycles kill (test-stat-no-inherit)