Commit Graph

67 Commits

Author SHA1 Message Date
Ingo Molnar f2858d8ad9 perf sched: Add 'perf sched latency' and 'perf sched replay'
Separate the option parsing cleanly and add two variants:

 - 'perf sched latency' (can be abbreviated via 'perf sched lat')
 - 'perf sched replay'  (can be abbreviated via 'perf sched rep')

Also add a repeat count option to replay and add a separation
set of options for replay.

Do the sorting setup only in the latency sub-command.

Display separate help screens for 'perf sched' and
'perf sched replay -h' - i.e. further separation of the
sub-commands.

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: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13 10:22:49 +02:00
Frederic Weisbecker daa1d7a5ea perf sched: Implement multidimensional sorting
Implement multidimensional sorting on perf sched so that
you can sort either by number of switches, latency average,
latency maximum, runtime.

perf sched -l -s avg,max  (this is the default)

-----------------------------------------------------------------------------------
 Task              |  Runtime ms | Switches | Average delay ms | Maximum delay ms |
-----------------------------------------------------------------------------------
 gnome-power-man   |    0.113 ms |        1 | avg: 4998.531 ms | max: 4998.531 ms |
 xfdesktop         |    1.190 ms |        7 | avg:  136.475 ms | max:  940.933 ms |
 xfce-mcs-manage   |    2.194 ms |       22 | avg:   38.534 ms | max:  735.174 ms |
 notification-da   |    2.749 ms |       31 | avg:   27.436 ms | max:  731.791 ms |
 xfce4-session     |    3.343 ms |       28 | avg:   26.796 ms | max:  734.891 ms |
 xfwm4             |    3.159 ms |       22 | avg:   12.406 ms | max:  241.333 ms |
 xchat             |   42.789 ms |      214 | avg:   11.886 ms | max:  100.349 ms |
 xfce4-terminal    |    5.386 ms |       22 | avg:   11.414 ms | max:  241.611 ms |
 firefox           |  151.992 ms |      123 | avg:    9.543 ms | max:  153.717 ms |
 xfce4-panel       |   24.324 ms |       47 | avg:    8.189 ms | max:  242.352 ms |
 :5090             |    6.932 ms |      111 | avg:    8.131 ms | max:  102.665 ms |
 events/0          |    0.758 ms |       12 | avg:    1.964 ms | max:   21.879 ms |
 Xorg              |  280.558 ms |      340 | avg:    1.864 ms | max:   99.526 ms |
 geany             |   63.391 ms |      295 | avg:    1.099 ms | max:    9.334 ms |
 reiserfs/0        |    0.039 ms |        2 | avg:    0.854 ms | max:    1.487 ms |
 kondemand/0       |    8.251 ms |      245 | avg:    0.691 ms | max:   34.372 ms |

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
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>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13 10:22:48 +02:00
Frederic Weisbecker 7362262687 perf sched: Fix nsec to msec conversion
We are dividing a time in ns by 1e9. This is a nsec to sec
conversion. What we want is msecs. Fix it by dividing by 1e6.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
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>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13 10:22:48 +02:00
Frederic Weisbecker 66685678a0 perf sched: Export the total, max latency and total runtime to thread atoms list
Add a field in the thread atom list that keeps track of the
total and max latencies and also the total runtime. This makes
a faster output and also prepares for sorting.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
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>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13 10:22:47 +02:00
Frederic Weisbecker c6ced61112 perf sched: Add involuntarily sleeping task in work atoms
Currently in perf sched, we are measuring the scheduler wakeup
latencies.

Now we also want measure the time a task wait to be scheduled
after it gets preempted.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
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>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13 10:22:47 +02:00
Frederic Weisbecker 1756220530 perf sched: Rename struct lat_snapshot to struct work atoms
To measures the latencies, we capture the sched atoms data into
a specific structure named struct lat_snapshot.

As this structure can be used for other purposes of scheduler
profiling and mirrors what happens in a thread work atom, lets
rename it to struct work_atom and propagate this renaming in
other functions and structures names to keep it coherent.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
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>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13 10:22:46 +02:00
Ingo Molnar 3e304147cd perf sched: Output runtime and context switch totals
After:

-----------------------------------------------------------------------------------
 Task              |  Runtime ms | Switches | Average delay ms | Maximum delay ms |
-----------------------------------------------------------------------------------
 make              |    0.678 ms |       13 | avg:    0.018 ms | max:    0.050 ms |
 gcc               |    0.014 ms |        2 | avg:    0.320 ms | max:    0.627 ms |
 gcc               |    0.000 ms |        2 | avg:    0.185 ms | max:    0.369 ms |
...
-----------------------------------------------------------------------------------
 TOTAL:            |   21.316 ms |       63 |
---------------------------------------------

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: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13 10:22:45 +02:00
Ingo Molnar ea92ed5a8f perf sched: Add runtime stats
Extend the latency tracking structure with scheduling atom
runtime info - and sum it up during per task display.

(Also clean up a few details.)

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: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13 10:22:45 +02:00
Ingo Molnar d9340c1db3 perf sched: Display time in milliseconds, reorganize output
After:

-----------------------------------------------------------------------------------
 Task              |  runtime ms | switches | average delay ms | maximum delay ms |
-----------------------------------------------------------------------------------
 migration/0       |    0.000 ms |        1 | avg:    0.047 ms | max:    0.047 ms |
 ksoftirqd/0       |    0.000 ms |        1 | avg:    0.039 ms | max:    0.039 ms |
 migration/1       |    0.000 ms |        3 | avg:    0.013 ms | max:    0.016 ms |
 migration/3       |    0.000 ms |        2 | avg:    0.003 ms | max:    0.004 ms |
 migration/4       |    0.000 ms |        1 | avg:    0.022 ms | max:    0.022 ms |
 distccd           |    0.000 ms |        1 | avg:    0.004 ms | max:    0.004 ms |
 distccd           |    0.000 ms |        1 | avg:    0.014 ms | max:    0.014 ms |
 distccd           |    0.000 ms |        2 | avg:    0.000 ms | max:    0.000 ms |
 distccd           |    0.000 ms |        2 | avg:    0.012 ms | max:    0.019 ms |
 distccd           |    0.000 ms |        1 | avg:    0.002 ms | max:    0.002 ms |
 as                |    0.000 ms |        2 | avg:    0.019 ms | max:    0.019 ms |
 as                |    0.000 ms |        3 | avg:    0.015 ms | max:    0.017 ms |
 as                |    0.000 ms |        1 | avg:    0.009 ms | max:    0.009 ms |
 perf              |    0.000 ms |        1 | avg:    0.001 ms | max:    0.001 ms |
 gcc               |    0.000 ms |        1 | avg:    0.021 ms | max:    0.021 ms |
 run-mozilla.sh    |    0.000 ms |        2 | avg:    0.010 ms | max:    0.017 ms |
 mozilla-plugin-   |    0.000 ms |        1 | avg:    0.006 ms | max:    0.006 ms |
 gcc               |    0.000 ms |        2 | avg:    0.013 ms | max:    0.013 ms |
-----------------------------------------------------------------------------------

(The runtime ms column is not filled in yet.)

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: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13 10:22:44 +02:00
Ingo Molnar 46f392c97f perf sched: Clean up latency and replay sub-commands
- Separate the latency and the replay commands more cleanly

 - Use consistent naming

 - Display help page on 'perf sched' outlining comments,
   instead of aborting

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: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13 10:22:44 +02:00
Frederic Weisbecker cdce9d738b perf sched: Add sched latency profiling
Add the -l --latency option that reports statistics about the
scheduler latencies.

For now, the latencies are measured in the following sequence
scope:

- task A is sleeping (D or S state)
- task B wakes up A
         ^
         |
         |

   latency timeframe

         |
         |
         v
- task A is scheduled in

Start by recording every scheduler events:

	perf record -e sched:*

and then fetch the results:

	perf sched -l

 Tasks                     count          total              avg            max

migration/0                  2             39849            19924           28826
ksoftirqd/0                  7            756383           108054          373014
migration/1                  5             45391             9078           10452
ksoftirqd/1                  2            399055           199527          359130
events/0                     8           4780110           597513         4500250
events/1                     9           6353057           705895         2986012
kblockd/0                   42          37805097           900121         5077684

The snapshot are in nanoseconds.

- Count: number of snapshots taken for the given task
- Total: total latencies in nanosec
- Avg  : average of latency between wake up and sched in
- Max  : max snapshot latency

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
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>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13 10:22:43 +02:00
Frederic Weisbecker 419ab0d6a9 perf sched: Make it easier to plug in new sub profilers
Create a sched event structure of handlers in which various
sched events reader can plug their own callbacks.

This makes easier the addition of new perf sched sub commands.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
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>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13 10:22:42 +02:00
Frederic Weisbecker 4653881802 perf sched: Fix bad event alignment
perf sched raises the following error when it meets a sched
switch event:

perf: builtin-sched.c:286: register_pid: Assertion `!(pid >= 65536)' failed.
Abandon

Currently in x86-64, the sched switch events have a hole in the
middle of the structure:

	u16 common_type;
	u8 common_flags;
	u8 common_preempt_count;
	u32 common_pid;
	u32 common_tgid;

	char prev_comm[16];
	u32 prev_pid;
	u32 prev_prio;
			<--- there
	u64 prev_state;
	char next_comm[16];
	u32 next_pid;
	u32 next_prio;

Gcc inserts a 4 bytes hole there for prev_state to be u64
aligned. And the events are exported to userspace with this
hole.

But in userspace, from perf sched, we fetch it using a
structure that has a new field in the beginning: u32 size. This
is because our trace is exported with its size as a field. But
now that we have this new field, the hole in the middle
disappears because it makes prev_state becoming well aligned.

And since we are using a pointer to the raw trace using this
struct, instead of reading prev_state, we are reading the hole.

We could fix it by keeping the size seperate from the struct
but actually there a lot of other potential problems: some
fields may be saved as long in a 64 bits system and later read
as long in a 32 bits system. Also this direct cast doesn't care
about the endianness differences between the host traced
machine and the machine in which we do the post processing.

So instead of using such dangerous direct casts, fetch the
values using the trace parsing API that already takes care of
all these problems.

Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
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>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13 10:22:41 +02:00
Ingo Molnar ad236fd23b perf sched: Tighten up the code
Various small cleanups - removal of debug printks and dead
functions, etc.

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: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13 10:22:39 +02:00
Ingo Molnar fbf9482911 perf sched: Implement the scheduling workload replay engine
Integrate the schedbench.c bits with the raw trace events
that we get from the perf machinery, and activate the
workload replayer/simulator.

Example of a captured 'make -j' workload:

$ perf sched

  run measurement overhead: 90 nsecs
  sleep measurement overhead: 2724743 nsecs
  the run test took 1000081 nsecs
  the sleep test took 2981111 nsecs
  version = 0.5
  ...
  nr_run_events:        70
  nr_sleep_events:      66
  nr_wakeup_events:     9
  target-less wakeups:  71
  multi-target wakeups: 47
  run events optimized: 139
  task      0 (                perf:      6607), nr_events: 2
  task      1 (                perf:      6608), nr_events: 6
  task      2 (                    :         0), nr_events: 1
  task      3 (                make:      6609), nr_events: 5
  task      4 (                  sh:      6610), nr_events: 4
  task      5 (                make:      6611), nr_events: 6
  task      6 (                  sh:      6612), nr_events: 4
  task      7 (                make:      6613), nr_events: 5
  task      8 (        migration/11:        25), nr_events: 1
  task      9 (        migration/13:        29), nr_events: 1
  task     10 (        migration/15:        33), nr_events: 1
  task     11 (         migration/9:        21), nr_events: 1
  task     12 (                  sh:      6614), nr_events: 4
  task     13 (                make:      6615), nr_events: 5
  task     14 (                  sh:      6616), nr_events: 4
  task     15 (                make:      6617), nr_events: 7
  task     16 (         migration/3:         9), nr_events: 1
  task     17 (         migration/5:        13), nr_events: 1
  task     18 (         migration/7:        17), nr_events: 1
  task     19 (         migration/1:         5), nr_events: 1
  task     20 (                  sh:      6618), nr_events: 4
  task     21 (                make:      6619), nr_events: 5
  task     22 (                  sh:      6620), nr_events: 4
  task     23 (                make:      6621), nr_events: 10
  task     24 (                  sh:      6623), nr_events: 3
  task     25 (                 gcc:      6624), nr_events: 4
  task     26 (                 gcc:      6625), nr_events: 4
  task     27 (                 gcc:      6626), nr_events: 5
  task     28 (            collect2:      6627), nr_events: 5
  task     29 (                  sh:      6622), nr_events: 1
  task     30 (                make:      6628), nr_events: 7
  task     31 (                  sh:      6630), nr_events: 4
  task     32 (                 gcc:      6631), nr_events: 4
  task     33 (                  sh:      6629), nr_events: 1
  task     34 (                 gcc:      6632), nr_events: 4
  task     35 (                 gcc:      6633), nr_events: 4
  task     36 (            collect2:      6634), nr_events: 4
  task     37 (                make:      6635), nr_events: 8
  task     38 (                  sh:      6637), nr_events: 4
  task     39 (                  sh:      6636), nr_events: 1
  task     40 (                 gcc:      6638), nr_events: 4
  task     41 (                 gcc:      6639), nr_events: 4
  task     42 (                 gcc:      6640), nr_events: 4
  task     43 (            collect2:      6641), nr_events: 4
  task     44 (                make:      6642), nr_events: 6
  task     45 (                  sh:      6643), nr_events: 5
  task     46 (                  sh:      6644), nr_events: 3
  task     47 (                  sh:      6645), nr_events: 4
  task     48 (                make:      6646), nr_events: 6
  task     49 (                  sh:      6647), nr_events: 3
  task     50 (                make:      6648), nr_events: 5
  task     51 (                  sh:      6649), nr_events: 5
  task     52 (                  sh:      6650), nr_events: 6
  task     53 (                make:      6651), nr_events: 4
  task     54 (                make:      6652), nr_events: 5
  task     55 (                make:      6653), nr_events: 4
  task     56 (                make:      6654), nr_events: 4
  task     57 (                make:      6655), nr_events: 5
  task     58 (                  sh:      6656), nr_events: 4
  task     59 (                 gcc:      6657), nr_events: 9
  task     60 (         ksoftirqd/3:        10), nr_events: 1
  task     61 (                 gcc:      6658), nr_events: 4
  task     62 (                make:      6659), nr_events: 5
  task     63 (                  sh:      6660), nr_events: 3
  task     64 (                 gcc:      6661), nr_events: 5
  task     65 (            collect2:      6662), nr_events: 4
  ------------------------------------------------------------
  #1  : 256.745, ravg: 256.74, cpu: 0.00 / 0.00
  #2  : 439.372, ravg: 275.01, cpu: 0.00 / 0.00
  #3  : 411.971, ravg: 288.70, cpu: 0.00 / 0.00
  #4  : 385.500, ravg: 298.38, cpu: 0.00 / 0.00
  #5  : 366.526, ravg: 305.20, cpu: 0.00 / 0.00
  #6  : 381.281, ravg: 312.81, cpu: 0.00 / 0.00
  #7  : 410.756, ravg: 322.60, cpu: 0.00 / 0.00
  #8  : 368.009, ravg: 327.14, cpu: 0.00 / 0.00
  #9  : 408.098, ravg: 335.24, cpu: 0.00 / 0.00
  #10 : 368.582, ravg: 338.57, cpu: 0.00 / 0.00

I.e. we successfully analyzed the trace, replayed it
via real threads and measured the replayed workload's
scheduling properties.

This is how it looked like in 'top' output:

   PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND
  7164 mingo     20   0 1434m 8080  888 R 57.0  0.1   0:02.04 :perf
  7165 mingo     20   0 1434m 8080  888 R 41.8  0.1   0:01.52 :perf
  7228 mingo     20   0 1434m 8080  888 R 39.8  0.1   0:01.44 :gcc
  7225 mingo     20   0 1434m 8080  888 R 33.8  0.1   0:01.26 :gcc
  7202 mingo     20   0 1434m 8080  888 R 31.2  0.1   0:01.16 :sh
  7222 mingo     20   0 1434m 8080  888 R 25.2  0.1   0:00.96 :sh
  7211 mingo     20   0 1434m 8080  888 R 21.9  0.1   0:00.82 :sh
  7213 mingo     20   0 1434m 8080  888 D 19.2  0.1   0:00.74 :sh
  7194 mingo     20   0 1434m 8080  888 D 18.6  0.1   0:00.72 :make

There's still various kinks in it - more patches to come.

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: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13 10:22:38 +02:00
Ingo Molnar ec156764d4 perf sched: Import schedbench.c
Import the schedbench.c tool that i wrote some time ago to
simulate scheduler behavior but never finished. It's a good
basis for perf sched nevertheless.

Most of its guts are not hooked up to the perf event loop
yet - that will be done in the patches to come.

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: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13 10:22:37 +02:00
Ingo Molnar 0a02ad9331 perf: Add 'perf sched' tool
This turn-key tool allows scheduler measurements to be
conducted and the results be displayed numerically.

First baby step towards that goal: clone the new command off of
perf trace.

Fix a few other details along the way:

 - add (minimal) perf trace documentation

 - reorder a few places

 - list perf trace in the mainporcelain list as well
   as it's a very useful utility.

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: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-13 10:22:36 +02:00