Skip to content

cloud-kernel
cloud-kernel

openanolis / cloud-kernel
大约 1 年 前同步成功

通知 158
Star 36
Fork 7
cloud-kernel
cloud-kernel

体验新版 GitCode，发现更多精彩内容 >>

切换分支/标签
  1. 13 7月, 2013 4 次提交
  3. 09 7月, 2013 1 次提交
    • J
      perf tools: Fix -x/--exclude-other option for report command · 0276c22a
      由 Jiri Olsa 提交于 
      Currently we have symbol_conf.exclude_other being set as true every time
so the -x/--exclude-other has nothing to do.

Also we have no way to see the data with symbol_conf.exclude_other being
false which is useful sometimes.

Fixing it by making symbol_conf.exclude_other false by default.

1) Example without -x option:

  $ perf report -i perf.data.delete -p perf_session__delete -s parent

  +  99.91%  [other]
  +   0.08%  perf_session__delete
  +   0.00%  perf_session__delete_dead_threads
  +   0.00%  perf_session__delete_threads

2) Example with -x option:

  $ ./perf report -i perf.data.delete -p perf_session__delete -s parent -x

  +  96.22%  perf_session__delete
  +   1.89%  perf_session__delete_dead_threads
  +   1.89%  perf_session__delete_threads

In Example 1) we get the sorted out data together with the rest
"[other]". This could help us estimate how much time we spent in the
sorted data.

In Example 2) the total is just the sorted data.
Signed-off-by: NJiri Olsa <jolsa@redhat.com>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/n/tip-sg8fvu0fyqohf9ur9l38lhkw@git.kernel.orgSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      0276c22a
  5. 28 5月, 2013 3 次提交
  7. 01 4月, 2013 1 次提交
    • A
      perf tools: Add support for weight v7 (modified) · 05484298
      由 Andi Kleen 提交于 
      perf record has a new option -W that enables weightened sampling.

Add sorting support in top/report for the average weight per sample and the
total weight sum. This allows to both compare relative cost per event
and the total cost over the measurement period.

Add the necessary glue to perf report, record and the library.

v2: Merge with new hist refactoring.
v3: Fix manpage. Remove value check.
Rename global_weight to weight and weight to local_weight.
v4: Readd sort keys to manpage
v5: Move weight to end
v6: Move weight to template
v7: Rename weight key.

Original patch from Andi modified by Stephane Eranian <eranian@google.com>
to include ONLY the weight supporting code and apply to pristine 3.8.0-rc4.
Signed-off-by: NAndi Kleen <ak@linux.intel.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Namhyung Kim <namhyung.kim@lge.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1359040242-8269-6-git-send-email-eranian@google.com
[ committer note: changed to cope with fc5871ed and the hists_link perf test entry ]
Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      05484298
  9. 07 2月, 2013 1 次提交
  11. 25 1月, 2013 3 次提交
  13. 09 12月, 2012 4 次提交
  15. 09 11月, 2012 3 次提交
    • A
      perf diff: Use hists__link when not pairing just with baseline · bfaef4b4
      由 Arnaldo Carvalho de Melo 提交于 
      Previously there were blind spots because we were not looking at symbols
that didn't ocurred in the latest run:

  # perf record usleep 1
  [ perf record: Woken up 1 times to write data ]
  [ perf record: Captured and wrote 0.018 MB perf.data (~801 samples) ]
  # perf record usleep 1
  [ perf record: Woken up 1 times to write data ]
  [ perf record: Captured and wrote 0.018 MB perf.data (~801 samples) ]

Before:

  # perf diff
  # Event 'cycles'
  #
  # Baseline    Delta      Shared Object                         Symbol
  # ........  .......  .................  .............................
  #
              +10.38%  [kernel.kallsyms]  [k] get_empty_filp
               +9.51%  [kernel.kallsyms]  [k] update_sd_lb_stats
               +9.41%  libpopt.so.0.0.0   [.] _init
               +9.29%  [kernel.kallsyms]  [k] vma_interval_tree_insert
       9.05%   +0.12%  [kernel.kallsyms]  [k] do_sys_open
               +9.14%  [kernel.kallsyms]  [k] kfree
               +8.98%  [kernel.kallsyms]  [k] free_pages_and_swap_cache
               +8.78%  [kernel.kallsyms]  [k] unmap_page_range
       9.36%   -0.90%  [kernel.kallsyms]  [k] zap_pte_range
       7.60%   +0.09%  [kernel.kallsyms]  [k] find_next_bit
               +4.37%  [kernel.kallsyms]  [k] place_entity
               +3.38%  [kernel.kallsyms]  [k] __do_page_fault
               +0.80%  [kernel.kallsyms]  [k] native_apic_mem_write
       0.21%   +0.43%  [kernel.kallsyms]  [k] native_write_msr_safe
  #

So 9.05 + 9.36 + 7.60 + 0.21 != 100%

Now using the recently introduced hists__link we can see the whole
picture:

  # perf diff
  # Event 'cycles'
  #
  # Baseline    Delta      Shared Object                         Symbol
  # ........  .......  .................  .............................
  #
       8.44%   -8.44%  [kernel.kallsyms]  [k] _raw_spin_lock
       9.05%   -9.05%  [kernel.kallsyms]  [k] sha_transform
      10.55%  -10.55%  [kernel.kallsyms]  [k] __d_lookup_rcu
              +10.38%  [kernel.kallsyms]  [k] get_empty_filp
      17.70%  -17.70%  [kernel.kallsyms]  [k] kmem_cache_free
               +9.51%  [kernel.kallsyms]  [k] update_sd_lb_stats
               +9.41%  libpopt.so.0.0.0   [.] _init
               +9.29%  [kernel.kallsyms]  [k] vma_interval_tree_insert
       9.05%   +0.12%  [kernel.kallsyms]  [k] do_sys_open
               +9.14%  [kernel.kallsyms]  [k] kfree
               +8.98%  [kernel.kallsyms]  [k] free_pages_and_swap_cache
               +8.78%  [kernel.kallsyms]  [k] unmap_page_range
       9.36%   -0.90%  [kernel.kallsyms]  [k] zap_pte_range
       7.60%   +0.09%  [kernel.kallsyms]  [k] find_next_bit
               +4.37%  [kernel.kallsyms]  [k] place_entity
               +3.38%  [kernel.kallsyms]  [k] __do_page_fault
       4.01%   -4.01%  [kernel.kallsyms]  [k] handle_pte_fault
       9.27%   -9.27%  [kernel.kallsyms]  [k] find_get_page
       0.78%   -0.78%  [kernel.kallsyms]  [k] rcu_irq_enter
       0.57%   -0.57%  [kernel.kallsyms]  [k] finish_task_switch
       4.25%   -4.25%  [kernel.kallsyms]  [k] run_timer_softirq
               +0.80%  [kernel.kallsyms]  [k] native_apic_mem_write
       0.21%   +0.43%  [kernel.kallsyms]  [k] native_write_msr_safe
       9.16%   -9.16%  ld-2.12.so         [.] close
  #

Now:

8.44 + 9.05 + 10.55 + 17.70 + 9.05 + 9.36 +
7.60 + 4.01 + 9.27 + 0.78 + 0.57 + 4.25 + 0.21 + 9.16 == 100%

Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Namhyung Kim <namhyung@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/n/tip-jeq55qdgby1745bs8r9sscdh@git.kernel.orgSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      bfaef4b4
    • A
      perf diff: Move hists__match to the hists lib · 95529be4
      由 Arnaldo Carvalho de Melo 提交于 
      Its not 'diff' specific and will be useful for other use cases, like
bucketizing multiple events in a single session.

Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Namhyung Kim <namhyung@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/n/tip-o35urjgxfxxm70aw1wa81s4w@git.kernel.orgSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      95529be4
    • A
      perf diff: Start moving to support matching more than two hists · b821c732
      由 Arnaldo Carvalho de Melo 提交于 
      We want to match more than two hists, so that we can match more than two
perf.data files and moreover, match hist_entries (buckets) in multiple
events in a group.

So the "baseline"/"leader" will instead of a ->pair pointer, use a
list_head, that will link to the pairs and hists__match use it.

Following that perf_evlist__link will link the hists in its evsel
groups.

Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Namhyung Kim <namhyung@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/n/tip-2kbmzepoi544ygj9godseqpv@git.kernel.orgSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      b821c732
  17. 07 10月, 2012 1 次提交
    • A
      perf event: No need to create a thread when handling PERF_RECORD_EXIT · f62d3f0f
      由 Arnaldo Carvalho de Melo 提交于 
      When we were processing a PERF_RECORD_EXIT event we first used
machine__findnew_thread for both the thread exiting and for its parent,
only to use just the thread struct associated with the one exiting, and
to just delete it.

If it existed, i.e. not created at this very moment in
machine__findnew_thread, it will be moved to the machine->dead_threads
linked list, because we may have hist_entries pointing to it, but if it
was created just do be deleted, it will just sit there with no
references at all.

Use the new machine__find_thread() method so that if it is not there, we
don't create it.

As a bonus the parent thread will also not be created at this point.

Create process_fork() and process_exit() helpers to use this and make
the builtins use it instead of the generic process_task(), ditched by
this patch.

Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Namhyung Kim <namhyung@gmail.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Link: http://lkml.kernel.org/n/tip-z7n2y98ebjyrvmytaope4vdl@git.kernel.orgSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      f62d3f0f
  19. 06 10月, 2012 7 次提交
  21. 05 10月, 2012 3 次提交
  23. 11 9月, 2012 1 次提交
    • I
      perf tools: Use __maybe_used for unused variables · 1d037ca1
      由 Irina Tirdea 提交于 
      perf defines both __used and __unused variables to use for marking
unused variables. The variable __used is defined to
__attribute__((__unused__)), which contradicts the kernel definition to
__attribute__((__used__)) for new gcc versions. On Android, __used is
also defined in system headers and this leads to warnings like: warning:
'__used__' attribute ignored

__unused is not defined in the kernel and is not a standard definition.
If __unused is included everywhere instead of __used, this leads to
conflicts with glibc headers, since glibc has a variables with this name
in its headers.

The best approach is to use __maybe_unused, the definition used in the
kernel for __attribute__((unused)). In this way there is only one
definition in perf sources (instead of 2 definitions that point to the
same thing: __used and __unused) and it works on both Linux and Android.
This patch simply replaces all instances of __used and __unused with
__maybe_unused.
Signed-off-by: NIrina Tirdea <irina.tirdea@intel.com>
Acked-by: NPekka Enberg <penberg@kernel.org>
Cc: David Ahern <dsahern@gmail.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Namhyung Kim <namhyung.kim@lge.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Steven Rostedt <rostedt@goodmis.org>
Link: http://lkml.kernel.org/r/1347315303-29906-7-git-send-email-irina.tirdea@intel.com
[ committer note: fixed up conflict with a116e05d in builtin-sched.c ]
Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      1d037ca1
  25. 09 9月, 2012 1 次提交
  27. 08 9月, 2012 1 次提交
    • J
      perf diff: Make diff command work with evsel hists · 863e451f
      由 Jiri Olsa 提交于 
      Putting 'perf diff' command back on track with the 'latest'
evsel hists changes. Each evsel has its own 'hists' object
gathering stats for the particular event.

While currently counts are accumulated for the whole session
regardless of the events diversification within compared
sessions.

The 'perf diff' command now outputs all matching events within
compared sessions (with event name specified). The per event
diff output stays the same.

  $ ./perf diff
  # Event 'cycles'
  #
  # Baseline  Delta          Shared Object                          Symbol
  # ........ ..........  .................  ..............................
  #
       0.00%    +15.14%  [kernel.kallsyms]  [k] __wake_up
       0.00%    +13.38%  [kernel.kallsyms]  [k] ext4fs_dirhash

... SNIP

       0.00%     +0.42%  [kernel.kallsyms]  [k] local_clock
       0.17%     -0.05%  [kernel.kallsyms]  [k] native_write_msr_safe

  # Event 'faults'
  #
  # Baseline  Delta          Shared Object                          Symbol
  # ........ ..........  .................  ..............................
  #
       0.00%    +79.12%  ld-2.15.so         [.] _dl_relocate_object
       0.00%    +11.62%  ld-2.15.so         [.] openaux
Signed-off-by: NJiri Olsa <jolsa@redhat.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Link: http://lkml.kernel.org/r/1346946426-13496-2-git-send-email-jolsa@redhat.comSigned-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      863e451f
  29. 23 3月, 2012 1 次提交
  31. 28 11月, 2011 3 次提交
  33. 07 10月, 2011 1 次提交
  35. 24 3月, 2011 1 次提交
    • A
      perf session: Pass evsel in event_ops->sample() · 9e69c210
      由 Arnaldo Carvalho de Melo 提交于 
      Resolving the sample->id to an evsel since the most advanced tools,
report and annotate, and the others will too when they evolve to
properly support multi-event perf.data files.

Good also because it does an extra validation, checking that the ID is
valid when present. When that is not the case, the overhead is just a
branch + function call (perf_evlist__id2evsel).

Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Tom Zanussi <tzanussi@gmail.com>
LKML-Reference: <new-submission>
Signed-off-by: NArnaldo Carvalho de Melo <acme@redhat.com>
      9e69c210