[RFC 00/49] perf tools: Add threads to record command

From: Jiri Olsa
Date: Tue Jan 09 2018 - 10:35:30 EST

Next message: Jiri Olsa: "[PATCH 01/49] perf tools: Remove perf_tool from event_op2"
Previous message: Peter Zijlstra: "Re: perf: perf_fuzzer quickly locks up on 4.15-rc7"
Next in thread: Jiri Olsa: "[PATCH 01/49] perf tools: Remove perf_tool from event_op2"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

hi,
sending *RFC* for threads support in perf record command.

In big picture this patchset adds perf record --threads
option that allows to create threads in following modes:

1) single thread mode (current)

$ perf record ...
$ perf record --threads=1 ...

2) mode with specific (X) number of threads

$ perf record --threads=X ...

3) mode that creates thread for every monitored memory map
.. which in perf record is equal to number of CPUs, and
it pins each thread to its map's cpu

$ perf record --threads=X ...

The perf.data stays as a single file.

This patchset contains lot of preparation changes to make
threaded record possible:

- Namhyung's changes to create multiple data streams in
perf data file, which allows having each thread data
being stored in separate files and merged into single
perf data after

- Namhyung's changes to create track mmaps for auxiliary
events

- Namhyung's changes to search for threads/mmaps/comms
using the time. This is needed because we have now
multiple data streams which are processed separately,
but they all need access to complete auxiliary events
data (threads/mmaps/comms). That's also a reason why
the auxiliary events are stored into separate data
stream, which is processed before real data.

- the rest of the code that adds threads abstraction into
record command allows to create them and distribute maps
among them

- other preparational changes

The threaded monitoring currently can't monitor backward maps
and there are probably more limitations which I haven't spotted
yet.

So far I tested on laptop:
http://people.redhat.com/~jolsa/record_threads/test-4CPU.txt

and a one bigger server:
http://people.redhat.com/~jolsa/record_threads/test-208CPU.txt

I can see decrease in recorded LOST events, but both the benchmark
and the monitoring must be carefully configured wrt:
- number of events (frequency)
- size of the memory maps
- size of events (callchains)
- final perf.data size

It's also available in:
git://git.kernel.org/pub/scm/linux/kernel/git/jolsa/perf.git
perf/record_threads

thoughts? ;-) thanks
jirka

---
Jiri Olsa (31):
perf tools: Remove perf_tool from event_op2
perf tools: Remove perf_tool from event_op3
perf tools: Pass struct perf_mmap into auxtrace_mmap__read* functions
perf tools: Add struct perf_mmap arg into record__write
perf tools: Create separate mmap for dummy tracking event
perf tools: Make copyfile_offset global
perf tools: Add perf_data__create_index function
perf record: Add --index option for building index table
perf tools: Convert dead thread list into rbtree
perf tools: Add thread::exited flag
perf callchain: Maintain libunwind's address space in map_groups
perf tools: Rename perf_evlist__munmap_filtered to perf_mmap__put_filtered
tools lib fd array: Introduce fdarray__add_clone function
tools lib subcmd: Add OPT_INTEGER_OPTARG|_SET options
perf tools: Move __perf_session__process_events args into struct
perf ui progress: Fix index progress display
perf tools: Add threads debug variable
perf tools: Add cpu into struct perf_mmap
perf tools: Add perf_mmap__read_tail function
perf record: Introduce struct record_thread
perf record: Read record thread's mmaps
perf record: Move waking into struct record
perf record: Move samples into struct record_thread
perf record: Move bytes_written into struct record_thread
perf record: Add record_thread start/stop/process functions
perf record: Wait for all threads being started
perf record: Add --threads option
perf record: Add --thread-stats option support
perf record: Add maps to --thread-stats output
perf record: Spread maps for --threads option
perf record: Spread maps for --threads=X option

Namhyung Kim (18):
perf tools: Use a software dummy event to track task/mmap events
perf tools: Extend perf_evlist__mmap_ex() to use track mmap
perf report: Skip dummy tracking event
perf tools: Add HEADER_DATA_INDEX feature
perf tools: Handle indexed data file properly
perf tools: Introduce thread__comm(_str)_by_time() helpers
perf tools: Add a test case for thread comm handling
perf tools: Use thread__comm_by_time() when adding hist entries
perf tools: Introduce machine__find*_thread_by_time()
perf tools: Add a test case for timed thread handling
perf tools: Maintain map groups list in a leader thread
perf tools: Introduce thread__find_addr_location_by_time() and friends
perf callchain: Use thread__find_addr_location_by_time() and friends
perf tools: Add a test case for timed map groups handling
perf tools: Save timestamp of a map creation
perf tools: Introduce map_groups__{insert,find}_by_time()
perf tools: Use map_groups__find_addr_by_time()
perf tools: Add testcase for managing maps with time

tools/lib/api/fd/array.c | 17 +
tools/lib/api/fd/array.h | 1 +
tools/lib/subcmd/parse-options.c | 2 +
tools/lib/subcmd/parse-options.h | 9 +
tools/perf/Documentation/perf-record.txt | 4 +
tools/perf/Documentation/perf.txt | 1 +
tools/perf/builtin-inject.c | 32 +-
tools/perf/builtin-record.c | 886 +++++++++++++++++++++++++++++--
tools/perf/builtin-report.c | 3 +
tools/perf/builtin-script.c | 31 +-
tools/perf/builtin-stat.c | 23 +-
tools/perf/perf.c | 1 +
tools/perf/perf.h | 3 +
tools/perf/tests/Build | 4 +
tools/perf/tests/builtin-test.c | 16 +
tools/perf/tests/dwarf-unwind.c | 4 +-
tools/perf/tests/hists_common.c | 2 +-
tools/perf/tests/hists_link.c | 2 +-
tools/perf/tests/tests.h | 4 +
tools/perf/tests/thread-comm.c | 48 ++
tools/perf/tests/thread-lookup-time.c | 181 +++++++
tools/perf/tests/thread-map-time.c | 91 ++++
tools/perf/tests/thread-mg-share.c | 7 +-
tools/perf/tests/thread-mg-time.c | 94 ++++
tools/perf/ui/browsers/hists.c | 30 +-
tools/perf/ui/gtk/hists.c | 3 +
tools/perf/util/auxtrace.c | 30 +-
tools/perf/util/auxtrace.h | 21 +-
tools/perf/util/data.c | 64 +++
tools/perf/util/data.h | 5 +
tools/perf/util/debug.c | 2 +
tools/perf/util/debug.h | 1 +
tools/perf/util/dso.c | 2 +-
tools/perf/util/event.c | 139 ++++-
tools/perf/util/evlist.c | 98 +++-
tools/perf/util/evlist.h | 7 +-
tools/perf/util/evsel.h | 15 +
tools/perf/util/header.c | 93 +++-
tools/perf/util/header.h | 18 +-
tools/perf/util/hist.c | 4 +-
tools/perf/util/intel-pt.c | 2 +-
tools/perf/util/machine.c | 299 +++++++++--
tools/perf/util/machine.h | 22 +-
tools/perf/util/map.c | 79 ++-
tools/perf/util/map.h | 41 +-
tools/perf/util/mmap.c | 11 +-
tools/perf/util/mmap.h | 30 +-
tools/perf/util/session.c | 178 ++++---
tools/perf/util/session.h | 5 +-
tools/perf/util/stat.c | 5 +-
tools/perf/util/stat.h | 5 +-
tools/perf/util/symbol-elf.c | 2 +-
tools/perf/util/symbol.c | 4 +-
tools/perf/util/thread.c | 201 ++++++-
tools/perf/util/thread.h | 29 +-
tools/perf/util/tool.h | 7 +-
tools/perf/util/unwind-libdw.c | 12 +-
tools/perf/util/unwind-libunwind-local.c | 41 +-
tools/perf/util/unwind-libunwind.c | 9 +-
tools/perf/util/unwind.h | 7 +-
tools/perf/util/util.c | 2 +-
tools/perf/util/util.h | 2 +
62 files changed, 2599 insertions(+), 392 deletions(-)
create mode 100644 tools/perf/tests/thread-comm.c
create mode 100644 tools/perf/tests/thread-lookup-time.c
create mode 100644 tools/perf/tests/thread-map-time.c
create mode 100644 tools/perf/tests/thread-mg-time.c

Next message: Jiri Olsa: "[PATCH 01/49] perf tools: Remove perf_tool from event_op2"
Previous message: Peter Zijlstra: "Re: perf: perf_fuzzer quickly locks up on 4.15-rc7"
Next in thread: Jiri Olsa: "[PATCH 01/49] perf tools: Remove perf_tool from event_op2"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]