linux/tools/perf
Ingo Molnar 3dfabc74c6 perf report: Add per system call overhead histogram
Take advantage of call-graph percounter sampling/recording to
display a non-trivial histogram: the true, collapsed/summarized
cost measurement, on a per system call total overhead basis:

 aldebaran:~/linux/linux/tools/perf> ./perf record -g -a -f ~/hackbench 10
 aldebaran:~/linux/linux/tools/perf> ./perf report -s symbol --syscalls | head -10
 #
 # (3536 samples)
 #
 # Overhead  Symbol
 # ........  ......
 #
     40.75%  [k] sys_write
     40.21%  [k] sys_read
      4.44%  [k] do_nmi
 ...

This is done by accounting each (reliable) call-chain that chains back
to a given system call to that system call function.

[ So in the above example we can see that hackbench spends about 40% of
  its total time somewhere in sys_write() and 40% somewhere in
  sys_read(), the rest of the time is spent in user-space. The time
  is not spent in sys_write() _itself_ but in one of its many child
  functions. ]

Or, a recording of a (source files are already in the page-cache) kernel build:

 $ perf record -g -m 512 -f -- make -j32 kernel
 $ perf report -s s --syscalls | grep '\[k\]' | grep -v nmi

     4.14%  [k] do_page_fault
     1.20%  [k] sys_write
     1.10%  [k] sys_open
     0.63%  [k] sys_exit_group
     0.48%  [k] smp_apic_timer_interrupt
     0.37%  [k] sys_read
     0.37%  [k] sys_execve
     0.20%  [k] sys_mmap
     0.18%  [k] sys_close
     0.14%  [k] sys_munmap
     0.13%  [k] sys_poll
     0.09%  [k] sys_newstat
     0.07%  [k] sys_clone
     0.06%  [k] sys_newfstat
     0.05%  [k] sys_access
     0.05%  [k] schedule

Shows the true total cost of each syscall variant that gets used
during a kernel build. This profile reveals it that pagefaults are
the costliest, followed by read()/write().

An interesting detail: timer interrupts cost 0.5% - or 0.5 seconds
per 100 seconds of kernel build-time. (this was done with HZ=1000)

The summary is done in 'perf report', i.e. in the post-processing
stage - so once we have a good call-graph recording, this type of
non-trivial high-level analysis becomes possible.

Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Pekka Enberg <penberg@cs.helsinki.fi>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-06-15 15:58:03 +02:00
..
Documentation
util perf stat: Reorganize output 2009-06-13 13:40:03 +02:00
.gitignore
builtin-annotate.c perf annotate: Fixes for filename:line displays 2009-06-13 17:51:00 +02:00
builtin-help.c
builtin-list.c
builtin-record.c perf record: Fix fast task-exit race 2009-06-15 09:08:31 +02:00
builtin-report.c perf report: Add per system call overhead histogram 2009-06-15 15:58:03 +02:00
builtin-stat.c perf stat: Enable raw data to be printed 2009-06-13 15:40:35 +02:00
builtin-top.c perf_counter: Standardize event names 2009-06-11 17:54:15 +02:00
builtin.h
command-list.txt
design.txt perf_counter: Start documenting HAVE_PERF_COUNTERS requirements 2009-06-12 19:37:30 +02:00
Makefile perf stat: Enable raw data to be printed 2009-06-13 15:40:35 +02:00
perf.c
perf.h perf_counter: Add forward/backward attribute ABI compatibility 2009-06-12 14:28:52 +02:00