ruby-prof 0.8.1-x86-mswin32
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- data/CHANGES +221 -0
- data/LICENSE +23 -0
- data/README +432 -0
- data/Rakefile +159 -0
- data/bin/ruby-prof +224 -0
- data/examples/flat.txt +55 -0
- data/examples/graph.html +823 -0
- data/examples/graph.txt +170 -0
- data/ext/ruby_prof/measure_allocations.h +58 -0
- data/ext/ruby_prof/measure_cpu_time.h +152 -0
- data/ext/ruby_prof/measure_gc_runs.h +76 -0
- data/ext/ruby_prof/measure_gc_time.h +57 -0
- data/ext/ruby_prof/measure_memory.h +101 -0
- data/ext/ruby_prof/measure_process_time.h +52 -0
- data/ext/ruby_prof/measure_wall_time.h +53 -0
- data/ext/ruby_prof/mingw/Rakefile +23 -0
- data/ext/ruby_prof/mingw/build.rake +38 -0
- data/ext/ruby_prof/ruby_prof.c +1747 -0
- data/ext/ruby_prof/ruby_prof.h +188 -0
- data/ext/ruby_prof/version.h +4 -0
- data/lib/1.8/ruby_prof.so +0 -0
- data/lib/1.9/ruby_prof.so +0 -0
- data/lib/ruby-prof.rb +56 -0
- data/lib/ruby-prof/abstract_printer.rb +41 -0
- data/lib/ruby-prof/aggregate_call_info.rb +62 -0
- data/lib/ruby-prof/call_info.rb +47 -0
- data/lib/ruby-prof/call_tree_printer.rb +84 -0
- data/lib/ruby-prof/flat_printer.rb +78 -0
- data/lib/ruby-prof/flat_printer_with_line_numbers.rb +72 -0
- data/lib/ruby-prof/graph_html_printer.rb +256 -0
- data/lib/ruby-prof/graph_printer.rb +157 -0
- data/lib/ruby-prof/method_info.rb +111 -0
- data/lib/ruby-prof/symbol_to_proc.rb +8 -0
- data/lib/ruby-prof/task.rb +146 -0
- data/lib/ruby-prof/test.rb +148 -0
- data/lib/unprof.rb +8 -0
- data/rails/environment/profile.rb +24 -0
- data/rails/example/example_test.rb +9 -0
- data/rails/profile_test_helper.rb +21 -0
- data/test/aggregate_test.rb +121 -0
- data/test/basic_test.rb +290 -0
- data/test/current_failures_windows +8 -0
- data/test/do_nothing.rb +0 -0
- data/test/duplicate_names_test.rb +32 -0
- data/test/enumerable_test.rb +16 -0
- data/test/exceptions_test.rb +15 -0
- data/test/exclude_threads_test.rb +54 -0
- data/test/exec_test.rb +14 -0
- data/test/line_number_test.rb +73 -0
- data/test/measurement_test.rb +121 -0
- data/test/module_test.rb +54 -0
- data/test/no_method_class_test.rb +13 -0
- data/test/prime.rb +58 -0
- data/test/prime_test.rb +13 -0
- data/test/printers_test.rb +130 -0
- data/test/recursive_test.rb +275 -0
- data/test/ruby-prof-bin +20 -0
- data/test/singleton_test.rb +37 -0
- data/test/stack_test.rb +138 -0
- data/test/start_stop_test.rb +95 -0
- data/test/test_suite.rb +23 -0
- data/test/thread_test.rb +173 -0
- data/test/unique_call_path_test.rb +225 -0
- metadata +143 -0
data/CHANGES
ADDED
@@ -0,0 +1,221 @@
|
|
1
|
+
0.7.10 (2009-01-22)
|
2
|
+
=======================
|
3
|
+
* fix SEGFAULT in 1.9
|
4
|
+
* add new printer flat_printer_with_line_numbers
|
5
|
+
|
6
|
+
0.7.7 (2009-01-13)
|
7
|
+
======================
|
8
|
+
* "fix" multi threading support for 1.9 http://redmine.ruby-lang.org/issues/show/2012
|
9
|
+
* speedups
|
10
|
+
|
11
|
+
0.7.6 (2009-12-31)
|
12
|
+
======================
|
13
|
+
* fix some tests for 1.9 (no real code changes)
|
14
|
+
|
15
|
+
0.7.5 (2009-12)
|
16
|
+
========================
|
17
|
+
* fix a GC collection bug (nobu's patch).
|
18
|
+
* correctly report recursive call depths (Kevin Scaldeferri).
|
19
|
+
* sort methods on output (Kevin Scaldeferri).
|
20
|
+
|
21
|
+
0.7.3 (2008-12-09)
|
22
|
+
========================
|
23
|
+
* Fixed compile error with new x86_64 code using GCC.
|
24
|
+
|
25
|
+
0.7.2 (2008-12-08)
|
26
|
+
========================
|
27
|
+
* Fixed major bug in printing child methods in graph reports.
|
28
|
+
|
29
|
+
* Fixes for supporting x86_64 machines (Diego Pettenò)
|
30
|
+
|
31
|
+
|
32
|
+
0.7.1 (2008-11-28)
|
33
|
+
========================
|
34
|
+
* Added new AggregateCallInfo class for printers to
|
35
|
+
make results easier to read. Take this call sequence
|
36
|
+
for example:
|
37
|
+
|
38
|
+
A B C
|
39
|
+
| | |
|
40
|
+
Z A A
|
41
|
+
| |
|
42
|
+
Z Z
|
43
|
+
|
44
|
+
By default, ruby-prof will show that Z was called by 3 separate
|
45
|
+
instances of A. In an IDE that is helpful but in a text report
|
46
|
+
it is not since it makes the report much harder to read.
|
47
|
+
As a result, printers now aggregate together callers (and children),
|
48
|
+
matching ruby-prof's output from versions prior to 0.7.0.
|
49
|
+
|
50
|
+
* Fixes for supporting x86_64 machines (Matt Sanford)
|
51
|
+
|
52
|
+
|
53
|
+
0.7.0 (2008-11-04)
|
54
|
+
========================
|
55
|
+
|
56
|
+
Features
|
57
|
+
--------
|
58
|
+
* Added two new methods - RubyProf.resume and RubyProf.pause.
|
59
|
+
RubyProf.resume takes an optional block, which ensures that
|
60
|
+
RubyProf.pause is called. For example:
|
61
|
+
|
62
|
+
10.times do |i|
|
63
|
+
RubyProf.resume do
|
64
|
+
# Some long process
|
65
|
+
end
|
66
|
+
end
|
67
|
+
|
68
|
+
result = RubyProf.stop
|
69
|
+
|
70
|
+
* Added support for profiling tests that use Ruby's built-in
|
71
|
+
unit test framework (ie, test derived from
|
72
|
+
Test::Unit::TestCase). To enable profiling simply add
|
73
|
+
the following line of code to your test class:
|
74
|
+
|
75
|
+
include RubyProf::Test
|
76
|
+
|
77
|
+
By default, profiling results are written to the current
|
78
|
+
processes working directory. To change this, or other
|
79
|
+
profiling options, simply modify the PROFILE_OPTIONS hash
|
80
|
+
table as needed.
|
81
|
+
|
82
|
+
* Used the new support for profiling test cases to revamp
|
83
|
+
the way that Rails profiling works. For more information
|
84
|
+
please refer to RubyProf's documentation.
|
85
|
+
|
86
|
+
* Keep track of call stack for each method to enable more
|
87
|
+
powerful profile visualizations in Integrated Development
|
88
|
+
Environments (Hin Boean, work supported by CodeGear).
|
89
|
+
|
90
|
+
* Expose measurements to Ruby (Jeremy Kemper).
|
91
|
+
|
92
|
+
* Add support for additional memory measurements modes in Ruby 1.9 (Jeremy Kemper).
|
93
|
+
|
94
|
+
* Add support for Lloyd Hilaiel's Ruby patch for measuring total heap size.
|
95
|
+
See http://lloydforge.org/projects/ruby. (Jeremy Kemper).
|
96
|
+
|
97
|
+
|
98
|
+
Fixes
|
99
|
+
-------
|
100
|
+
* RubyProf.profile no longer crashes if an exception is
|
101
|
+
thrown during a profiling run.
|
102
|
+
|
103
|
+
* Measure memory in fractional kilobytes rather than rounding down (Jeremy Kemper)
|
104
|
+
|
105
|
+
|
106
|
+
0.6.0 (2008-02-03)
|
107
|
+
========================
|
108
|
+
|
109
|
+
ruby-prof 0.6.0 adds support for Ruby 1.9 and memory profiling.
|
110
|
+
|
111
|
+
Features
|
112
|
+
--------
|
113
|
+
* Added support for ruby 1.9 (Shugo Maeda)
|
114
|
+
* Added support for outputting printer results to a String, Array or IO
|
115
|
+
object (Michael Granger)
|
116
|
+
* Add new memory profiling mode. Note this mode depends on a
|
117
|
+
patched Ruby interpreter (Alexander Dymo)
|
118
|
+
|
119
|
+
Fixes
|
120
|
+
-------
|
121
|
+
* Improvements to GraphHtmlPrinter including updated documentation,
|
122
|
+
fixes for min_time support, ability to specify templates using
|
123
|
+
strings or filenames, and table layout fixes (Makoto Kuwata)
|
124
|
+
* Fixes to scaling factor for calltrees so that precision is not lost
|
125
|
+
due to the conversion to doubles (Sylvain Joyeux)
|
126
|
+
* Changed constant ALLOCATED_OBJECTS to ALLOCATIONS in the C code to
|
127
|
+
match the Ruby code (Sylvain Joyeux)
|
128
|
+
* Added support for calltree printer to ruby-prof binary script (Sylvain Joyeux)
|
129
|
+
* Fix support for the allocator measure mode to extconf.rb (Sylvain Joyeux)
|
130
|
+
* Honor measure mode when specified on the command line (Sylvain Joyeux)
|
131
|
+
* Sorting of methods by total time was incorrect (Dan Fitch, Charlie Savage)
|
132
|
+
* Fix ruby-prof to work with the latest version of GEMS (Alexander Dymo)
|
133
|
+
* Always define MEASURE_CPU_TIME and MEASURE_ALLOCATIONS in Ruby code, but
|
134
|
+
set their values to nil if the functionality is not available.
|
135
|
+
|
136
|
+
|
137
|
+
0.5.2 (2007-07-19)
|
138
|
+
========================
|
139
|
+
|
140
|
+
ruby-prof 0.5.2 is a bug fix release.
|
141
|
+
|
142
|
+
Fixes
|
143
|
+
-------
|
144
|
+
* Include missing rails plugin
|
145
|
+
|
146
|
+
|
147
|
+
0.5.1 (2007-07-18)
|
148
|
+
========================
|
149
|
+
|
150
|
+
ruby-prof 0.5.1 is a bug fix and performance release.
|
151
|
+
|
152
|
+
Performance
|
153
|
+
--------
|
154
|
+
* Significantly reduced the number of thread lookups by
|
155
|
+
caching the last executed thread.
|
156
|
+
|
157
|
+
Fixes
|
158
|
+
-------
|
159
|
+
* Properly escape method names in HTML reports
|
160
|
+
* Fix use of -m and --min-percent command line switches
|
161
|
+
* Default source file information to ruby_runtime#0 for c calls
|
162
|
+
* Moved rails_plugin to top level so it is more obvious
|
163
|
+
* Updated rails_plugin to write reports to the current
|
164
|
+
Rails log directory
|
165
|
+
* Added additional tests
|
166
|
+
|
167
|
+
|
168
|
+
0.5.0 (2007-07-09)
|
169
|
+
========================
|
170
|
+
|
171
|
+
Features
|
172
|
+
--------
|
173
|
+
* Added support for timing multi-threaded applications
|
174
|
+
* Added support for 64 bit systems (patch from Diego 'Flameeyes' Petten)
|
175
|
+
* Added suport for outputting data in the format used by
|
176
|
+
KCacheGrind (patch from Carl Shimer)
|
177
|
+
* Add filename and line numbers to call tree information (patch from Carl Shimer)
|
178
|
+
* Added Visual Studio 2005 project file.
|
179
|
+
* Added replace-progname switch, als rcov.
|
180
|
+
* Added better support for recursive methods
|
181
|
+
* Added better support for profiling Rails applications
|
182
|
+
|
183
|
+
Fixes
|
184
|
+
-------
|
185
|
+
* Fixes bug when the type of an attached object (singleton) is inherited
|
186
|
+
from T_OBJECT as opposed to being a T_OBJECT (identified by Francis Cianfrocca)
|
187
|
+
* ruby-prof now works in IRB.
|
188
|
+
* Fix sort order in reports.
|
189
|
+
* Fixed rdoc compile error.
|
190
|
+
* Fix tabs in erb template for graph html report on windows.
|
191
|
+
|
192
|
+
0.4.1 (2006-06-26)
|
193
|
+
========================
|
194
|
+
|
195
|
+
Features
|
196
|
+
--------
|
197
|
+
* Added a RubyProf.running? method to indicate whether a profile is in progress.
|
198
|
+
* Added tgz and zip archives to release
|
199
|
+
|
200
|
+
Fixes
|
201
|
+
-------
|
202
|
+
* Duplicate method names are now allowed
|
203
|
+
* The documentation has been updated to show the correct API usage is RubyProf.stop not RubyProf.end
|
204
|
+
|
205
|
+
|
206
|
+
0.4.0 (2006-06-16)
|
207
|
+
========================
|
208
|
+
Features
|
209
|
+
--------
|
210
|
+
* added support for call graphs
|
211
|
+
* added support for printers. Currently there is a FlatPrinter,
|
212
|
+
GraphPrinter and GraphHtmlPrinter.
|
213
|
+
* added support for recursive methods
|
214
|
+
* added Windows support
|
215
|
+
* now packaged as a RubyGem
|
216
|
+
|
217
|
+
Fixes
|
218
|
+
-------
|
219
|
+
* Fixes bug where RubyProf would crash depending on the
|
220
|
+
way it was invoked - for example, it did not run when
|
221
|
+
used with Arachno Ruby's customized version of Ruby.
|
data/LICENSE
ADDED
@@ -0,0 +1,23 @@
|
|
1
|
+
Copyright (C) 2005 Shugo Maeda <shugo@ruby-lang.org>
|
2
|
+
All rights reserved.
|
3
|
+
*
|
4
|
+
Redistribution and use in source and binary forms, with or without
|
5
|
+
modification, are permitted provided that the following conditions
|
6
|
+
are met:
|
7
|
+
1. Redistributions of source code must retain the above copyright
|
8
|
+
notice, this list of conditions and the following disclaimer.
|
9
|
+
2. Redistributions in binary form must reproduce the above copyright
|
10
|
+
notice, this list of conditions and the following disclaimer in the
|
11
|
+
documentation and/or other materials provided with the distribution.
|
12
|
+
*
|
13
|
+
THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
|
14
|
+
ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
|
15
|
+
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
|
16
|
+
ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
|
17
|
+
FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
|
18
|
+
DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
|
19
|
+
OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
|
20
|
+
HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
|
21
|
+
LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
|
22
|
+
OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
|
23
|
+
SUCH DAMAGE.
|
data/README
ADDED
@@ -0,0 +1,432 @@
|
|
1
|
+
= ruby-prof
|
2
|
+
|
3
|
+
== Overview
|
4
|
+
|
5
|
+
ruby-prof is a fast code profiler for Ruby. Its features include:
|
6
|
+
|
7
|
+
* Speed - it is a C extension and therefore many times faster than the standard Ruby profiler.
|
8
|
+
* Modes - Ruby prof can measure a number of different parameters, including
|
9
|
+
call times, memory usage and object allocations.
|
10
|
+
* Reports - can generate text and cross-referenced html reports
|
11
|
+
- Flat Profiles - similar to the reports generated by the standard Ruby profiler
|
12
|
+
- Graph profiles - similar to GProf, these show how long a method runs, which methods call it and which methods it calls.
|
13
|
+
- Call tree profiles - outputs results in the calltree format suitable for the KCacheGrind profiling tool.
|
14
|
+
* Threads - supports profiling multiple threads simultaneously
|
15
|
+
* Recursive calls - supports profiling recursive method calls
|
16
|
+
|
17
|
+
|
18
|
+
== Requirements
|
19
|
+
|
20
|
+
ruby-prof requires Ruby 1.8.4 or higher.
|
21
|
+
|
22
|
+
If you are running Linux or Unix you'll need a C compiler so the extension
|
23
|
+
can be compiled when it is installed.
|
24
|
+
|
25
|
+
If you are running Windows, then you may need to install the
|
26
|
+
Windows specific RubyGem which includes an already built extension (see below).
|
27
|
+
|
28
|
+
== Install
|
29
|
+
|
30
|
+
The easiest way to install ruby-prof is by using Ruby Gems. To install:
|
31
|
+
|
32
|
+
<tt>gem install ruby-prof</tt>
|
33
|
+
|
34
|
+
If you on windows mswin [not mingw] (check via ruby -v) and
|
35
|
+
don't have an MSVC compiler, please install v0.7.3 which
|
36
|
+
has a prebuilt binary
|
37
|
+
C:> gem install ruby-prof -v0.7.3
|
38
|
+
|
39
|
+
If you're on mingw, please install the devkit first, then install the latest version (gem install ruby-prof).
|
40
|
+
|
41
|
+
== Usage
|
42
|
+
|
43
|
+
There are three ways of running ruby-prof.
|
44
|
+
|
45
|
+
=== ruby-prof executable
|
46
|
+
|
47
|
+
The first is to use ruby-prof to run the Ruby program
|
48
|
+
you want to profile. For more information refer to
|
49
|
+
the ruby-prof documentation[link:files/bin/ruby-prof.html].
|
50
|
+
|
51
|
+
|
52
|
+
=== ruby-prof API
|
53
|
+
|
54
|
+
The second way is to use the ruby-prof API to profile
|
55
|
+
particular segments of code.
|
56
|
+
|
57
|
+
require 'ruby-prof'
|
58
|
+
|
59
|
+
# Profile the code
|
60
|
+
RubyProf.start
|
61
|
+
...
|
62
|
+
[code to profile]
|
63
|
+
...
|
64
|
+
result = RubyProf.stop
|
65
|
+
|
66
|
+
# Print a flat profile to text
|
67
|
+
printer = RubyProf::FlatPrinter.new(result)
|
68
|
+
printer.print(STDOUT)
|
69
|
+
|
70
|
+
Alternatively, you can use a block to tell ruby-prof what
|
71
|
+
to profile:
|
72
|
+
|
73
|
+
require 'ruby-prof'
|
74
|
+
|
75
|
+
# Profile the code
|
76
|
+
result = RubyProf.profile do
|
77
|
+
...
|
78
|
+
[code to profile]
|
79
|
+
...
|
80
|
+
end
|
81
|
+
|
82
|
+
# Print a graph profile to text
|
83
|
+
printer = RubyProf::GraphPrinter.new(result)
|
84
|
+
printer.print(STDOUT, 0)
|
85
|
+
|
86
|
+
Starting with the 0.6.1 release, ruby-prof also supports pausing and resuming
|
87
|
+
profiling runs.
|
88
|
+
|
89
|
+
require 'ruby-prof'
|
90
|
+
|
91
|
+
# Profile the code
|
92
|
+
RubyProf.start
|
93
|
+
[code to profile]
|
94
|
+
RubyProf.pause
|
95
|
+
[other code]
|
96
|
+
RubyProf.resume
|
97
|
+
[code to profile]
|
98
|
+
result = RubyProf.stop
|
99
|
+
|
100
|
+
Note that resume will automatically call start if a profiling run
|
101
|
+
has not yet started. In addition, resume can also take a block:
|
102
|
+
|
103
|
+
require 'ruby-prof'
|
104
|
+
|
105
|
+
# Profile the code
|
106
|
+
RubyProf.resume do
|
107
|
+
[code to profile]
|
108
|
+
end
|
109
|
+
|
110
|
+
data = RubyProf.stop
|
111
|
+
|
112
|
+
With this usage, resume will automatically call pause at the
|
113
|
+
end of the block.
|
114
|
+
|
115
|
+
|
116
|
+
=== require unprof
|
117
|
+
|
118
|
+
The third way of using ruby-prof is by requiring unprof.rb:
|
119
|
+
|
120
|
+
require 'unprof'
|
121
|
+
|
122
|
+
This will start profiling immediately and will output the results
|
123
|
+
using a flat profile report.
|
124
|
+
|
125
|
+
This method is provided for backwards compatibility. Using
|
126
|
+
{ruby-prof}[link:files/bin/ruby-prof.html] provides more flexibility.
|
127
|
+
|
128
|
+
|
129
|
+
== Profiling Tests
|
130
|
+
|
131
|
+
Starting with the 0.6.1 release, ruby-prof supports profiling tests cases
|
132
|
+
written using Ruby's built-in unit test framework (ie, test derived from
|
133
|
+
Test::Unit::TestCase). To enable profiling simply add the following line
|
134
|
+
of code to within your test class:
|
135
|
+
|
136
|
+
include RubyProf::Test
|
137
|
+
|
138
|
+
Each test method is profiled separately. ruby-prof will run each test method
|
139
|
+
once as a warmup and then ten additional times to gather profile data.
|
140
|
+
Note that the profile data will *not* include the class's setup or
|
141
|
+
teardown methods.
|
142
|
+
|
143
|
+
Separate reports are generated for each method and saved, by default,
|
144
|
+
in the test process's working directory. To change this, or other profiling
|
145
|
+
options, modify your test class's PROFILE_OPTIONS hash table. To globally
|
146
|
+
change test profiling options, modify RubyProf::Test::PROFILE_OPTIONS.
|
147
|
+
|
148
|
+
|
149
|
+
== Profiling Rails
|
150
|
+
|
151
|
+
To profile a Rails application it is vital to run it using production like
|
152
|
+
settings (cache classes, cache view lookups, etc.). Otherwise, Rail's
|
153
|
+
dependency loading code will overwhelm any time spent in the application
|
154
|
+
itself (our tests show that Rails dependency loading causes a roughly 6x
|
155
|
+
slowdown). The best way to do this is create a new Rails environment,
|
156
|
+
profile.rb.
|
157
|
+
|
158
|
+
So to profile Rails:
|
159
|
+
|
160
|
+
1. Create a new profile.rb environment - or simply copy the example file
|
161
|
+
in ruby-prof/rails/environment/profile.rb
|
162
|
+
|
163
|
+
2. Copy the file:
|
164
|
+
|
165
|
+
ruby-prof/rails/profile_test_helper.rb
|
166
|
+
|
167
|
+
To:
|
168
|
+
|
169
|
+
your_rails_app/test/profile_test_helper.rb
|
170
|
+
|
171
|
+
3. Create a new test directory for profiling:
|
172
|
+
|
173
|
+
your_rails_app/test/profile
|
174
|
+
|
175
|
+
|
176
|
+
4. Write unit, functional or integration tests specifically designed
|
177
|
+
to profile some part of your Rails application. At the top
|
178
|
+
of each test, replace this line:
|
179
|
+
|
180
|
+
require File.dirname(__FILE__) + '/../test_helper'
|
181
|
+
|
182
|
+
With:
|
183
|
+
|
184
|
+
require File.dirname(__FILE__) + '/../profile_test_helper'
|
185
|
+
|
186
|
+
For example:
|
187
|
+
|
188
|
+
require File.dirname(__FILE__) + '/../profile_test_helper'
|
189
|
+
|
190
|
+
class ExampleTest < Test::Unit::TestCase
|
191
|
+
include RubyProf::Test
|
192
|
+
fixtures ....
|
193
|
+
|
194
|
+
def test_stuff
|
195
|
+
puts "Test method"
|
196
|
+
end
|
197
|
+
end
|
198
|
+
|
199
|
+
5. Now run your tests. Results will be written to:
|
200
|
+
|
201
|
+
your_rails_app/tmp/profile
|
202
|
+
|
203
|
+
|
204
|
+
== Reports
|
205
|
+
|
206
|
+
ruby-prof can generate a number of different reports:
|
207
|
+
|
208
|
+
* Flat Reports
|
209
|
+
* Graph Reports
|
210
|
+
* HTML Graph Reports
|
211
|
+
* Call graphs
|
212
|
+
|
213
|
+
Flat profiles show the overall time spent in each method. They
|
214
|
+
are a good of quickly identifying which methods take the most time.
|
215
|
+
An example of a flat profile and an explanation can be found in
|
216
|
+
{examples/flat.txt}[link:files/examples/flat_txt.html].
|
217
|
+
|
218
|
+
Graph profiles also show the overall time spent in each method.
|
219
|
+
In addition, they also show which methods call the current
|
220
|
+
method and which methods its calls. Thus they are good for
|
221
|
+
understanding how methods gets called and provide insight into
|
222
|
+
the flow of your program. An example text graph profile
|
223
|
+
is located at {examples/graph.txt}[link:files/examples/graph_txt.html].
|
224
|
+
|
225
|
+
HTML Graph profiles are the same as graph profiles, except
|
226
|
+
output is generated in hyper-linked HTML. Since graph profiles
|
227
|
+
can be quite large, the embedded links make it much easier to
|
228
|
+
navigate the results. An example html graph profile
|
229
|
+
is located at {examples/graph.html}[link:files/examples/graph_html.html].
|
230
|
+
|
231
|
+
HTML Graph profiles are the same as graph profiles, except
|
232
|
+
output is generated in hyper-linked HTML. Since graph profiles
|
233
|
+
can be quite large, the embedded links make it much easier to
|
234
|
+
navigate the results. An example html graph profile
|
235
|
+
is located at {examples/graph.html}[link:files/examples/graph_html.html].
|
236
|
+
|
237
|
+
Call graphs output results in the calltree profile format which is used
|
238
|
+
by KCachegrind. Call graph support was generously donated by Carl Shimer.
|
239
|
+
More information about the format can be found at
|
240
|
+
the {KCachegrind}[link:http://kcachegrind.sourceforge.net/cgi-bin/show.cgi/KcacheGrindCalltreeFormat] site.
|
241
|
+
|
242
|
+
|
243
|
+
== Printers
|
244
|
+
|
245
|
+
Reports are created by printers. Supported printers include:
|
246
|
+
|
247
|
+
* RubyProf::FlatPrinter - Creates a flat report in text format
|
248
|
+
* RubyProf::FlatPrinterWithLineNumbers - same as above but more verbose
|
249
|
+
* RubyProf::GraphPrinter - Creates a call graph report in text format
|
250
|
+
* RubyProf::GraphHtmlPrinter - Creates a call graph report in HTML (separate files per thread)
|
251
|
+
* RubyProf::CallTreePrinter - Creates a call tree report compatible with KCachegrind.
|
252
|
+
|
253
|
+
To use a printer:
|
254
|
+
|
255
|
+
...
|
256
|
+
result = RubyProf.end
|
257
|
+
printer = RubyProf::GraphPrinter.new(result)
|
258
|
+
printer.print(STDOUT, :min_percent => 2)
|
259
|
+
|
260
|
+
The first parameter is any writable IO object such as STDOUT or a file.
|
261
|
+
The second parameter, specifies the minimum percentage a method must take
|
262
|
+
to be printed. Percentages should be specified as integers in the range 0 to 100.
|
263
|
+
For more information please see the documentation for the different printers.
|
264
|
+
|
265
|
+
The other option is :print_file => true (default false), which adds the filename to the
|
266
|
+
output (GraphPrinter only).
|
267
|
+
|
268
|
+
== Measurements
|
269
|
+
|
270
|
+
Depending on the mode and platform, ruby-prof can measure various
|
271
|
+
aspects of a Ruby program. Supported measurements include:
|
272
|
+
|
273
|
+
* process time (RubyProf::PROCESS_TIME)
|
274
|
+
* wall time (RubyProf::WALL_TIME)
|
275
|
+
* cpu time (RubyProf::CPU_TIME)
|
276
|
+
* object allocations (RubyProf::ALLOCATIONS)
|
277
|
+
* memory usage (RubyProf::MEMORY)
|
278
|
+
* garbage collections runs (RubyProf::GC_RUNS)
|
279
|
+
* garbage collection time (RubyProf::GC_TIME)
|
280
|
+
|
281
|
+
Process time measures the time used by a process between any two moments.
|
282
|
+
It is unaffected by other processes concurrently running
|
283
|
+
on the system. Note that Windows does not support measuring process
|
284
|
+
times - therefore, measurements on Windows defaults to wall time.
|
285
|
+
|
286
|
+
Wall time measures the real-world time elapsed between any two moments.
|
287
|
+
If there are other processes concurrently running on the system
|
288
|
+
that use significant CPU or disk time during a profiling run
|
289
|
+
then the reported results will be too large.
|
290
|
+
|
291
|
+
CPU time uses the CPU clock counter to measure time. The returned
|
292
|
+
values are dependent on the correctly setting the CPU's frequency.
|
293
|
+
This mode is only supported on Pentium or PowerPC platforms (linux only).
|
294
|
+
|
295
|
+
Object allocation reports show how many objects each method in
|
296
|
+
a program allocates. This support was added by Sylvain Joyeux
|
297
|
+
and requires a patched Ruby interpreter. For more information and
|
298
|
+
the patch, please see:
|
299
|
+
http://rubyforge.org/tracker/index.php?func=detail&aid=11497&group_id=426&atid=1700
|
300
|
+
|
301
|
+
Memory usage reports show how much memory each method in a program
|
302
|
+
uses. This support was added by Alexander Dymo and requires a
|
303
|
+
patched Ruby interpreter. For more information, see:
|
304
|
+
http://rubyforge.org/tracker/index.php?func=detail&aid=17676&group_id=1814&atid=7062
|
305
|
+
|
306
|
+
Garbage collection runs report how many times Ruby's garbage collector
|
307
|
+
is invoked during a profiling session. This support was added by Jeremy
|
308
|
+
Kemper and requires a patched Ruby interpreter. For more information, see:
|
309
|
+
http://rubyforge.org/tracker/index.php?func=detail&aid=17676&group_id=1814&atid=7062
|
310
|
+
|
311
|
+
Garbage collection time reports how much time is spent in Ruby's garbage collector
|
312
|
+
during a profiling session. This support was added by Jeremy Kemper
|
313
|
+
and requires a patched Ruby interpreter. For more information, see:
|
314
|
+
http://rubyforge.org/tracker/index.php?func=detail&aid=17676&group_id=1814&atid=7062
|
315
|
+
|
316
|
+
To set the measurement:
|
317
|
+
|
318
|
+
* RubyProf.measure_mode = RubyProf::PROCESS_TIME
|
319
|
+
* RubyProf.measure_mode = RubyProf::WALL_TIME
|
320
|
+
* RubyProf.measure_mode = RubyProf::CPU_TIME
|
321
|
+
* RubyProf.measure_mode = RubyProf::ALLOCATIONS
|
322
|
+
* RubyProf.measure_mode = RubyProf::MEMORY
|
323
|
+
* RubyProf.measure_mode = RubyProf::GC_RUNS
|
324
|
+
* RubyProf.measure_mode = RubyProf::GC_TIME
|
325
|
+
|
326
|
+
The default value is RubyProf::PROCESS_TIME.
|
327
|
+
|
328
|
+
You may also specify the measure_mode by using the RUBY_PROF_MEASURE_MODE
|
329
|
+
environment variable:
|
330
|
+
|
331
|
+
* export RUBY_PROF_MEASURE_MODE=process
|
332
|
+
* export RUBY_PROF_MEASURE_MODE=wall
|
333
|
+
* export RUBY_PROF_MEASURE_MODE=cpu
|
334
|
+
* export RUBY_PROF_MEASURE_MODE=allocations
|
335
|
+
* export RUBY_PROF_MEASURE_MODE=memory
|
336
|
+
* export RUBY_PROF_MEASURE_MODE=gc_runs
|
337
|
+
* export RUBY_PROF_MEASURE_MODE=gc_time
|
338
|
+
|
339
|
+
Note that these values have changed since ruby-prof-0.3.0.
|
340
|
+
|
341
|
+
On Linux, process time is measured using the clock method provided
|
342
|
+
by the C runtime library. Note that the clock method does not
|
343
|
+
report time spent in the kernel or child processes and therefore
|
344
|
+
does not measure time spent in methods such as Kernel.sleep method.
|
345
|
+
If you need to measure these values, then use wall time. Wall time
|
346
|
+
is measured using the gettimeofday kernel method.
|
347
|
+
|
348
|
+
On Windows, timings default to wall times. If you set the clock
|
349
|
+
mode to PROCESS_TIME, then timing are read using the clock method
|
350
|
+
provided by the C runtime library. Note though, these values are
|
351
|
+
wall times on Windows and not process times like on Linux.
|
352
|
+
Wall time is measured using the GetLocalTime API.
|
353
|
+
|
354
|
+
If you use wall time, the results will be affected by other
|
355
|
+
processes running on your computer, network delays, disk access,
|
356
|
+
etc. As result, for the best results, try to make sure your
|
357
|
+
computer is only performing your profiling run and is
|
358
|
+
otherwise quiescent.
|
359
|
+
|
360
|
+
On both platforms, cpu time is measured using the RDTSC assembly
|
361
|
+
function provided by the Pentium and PowerPC platforms. CPU time
|
362
|
+
is dependent on the cpu's frequency. On Linux, ruby-prof attempts
|
363
|
+
to read this value from "/proc/cpuinfo." On Windows, you must
|
364
|
+
manually specify the clock frequency. This can be done using the
|
365
|
+
RUBY_PROF_CPU_FREQUENCY environment variable:
|
366
|
+
|
367
|
+
export RUBY_PROF_CPU_FREQUENCY=<value>
|
368
|
+
|
369
|
+
You can also directly set the cpu frequency by calling:
|
370
|
+
|
371
|
+
RubyProf.cpu_frequency = <value>
|
372
|
+
|
373
|
+
|
374
|
+
== Recursive Calls
|
375
|
+
|
376
|
+
Recursive calls occur when method A calls method A and cycles
|
377
|
+
occur when method A calls method B calls method C calls method A.
|
378
|
+
ruby-prof detects both direct recursive calls and cycles. Both
|
379
|
+
are indicated in reports by a "d number" in parentheses following a method
|
380
|
+
name. For example, here is a flat profile from the test method
|
381
|
+
|
382
|
+
RecursiveTest#test_recursive:
|
383
|
+
|
384
|
+
%self total self wait child calls name
|
385
|
+
100.00 2.00 2.00 0.00 0.00 2 Kernel#sleep
|
386
|
+
0.00 2.00 0.00 0.00 2.00 0 RecursiveTest#test_cycle
|
387
|
+
0.00 0.00 0.00 0.00 0.00 2 Fixnum#==
|
388
|
+
0.00 0.00 0.00 0.00 0.00 2 Fixnum#-
|
389
|
+
0.00 1.00 0.00 0.00 1.00 1 Object#sub_cycle(d1)
|
390
|
+
0.00 2.00 0.00 0.00 2.00 1 Object#sub_cycle
|
391
|
+
0.00 2.00 0.00 0.00 2.00 1 Object#cycle
|
392
|
+
0.00 1.00 0.00 0.00 1.00 1 Object#cycle(d1)
|
393
|
+
|
394
|
+
Notice the presence of Object#cycle and Object#cycle(d1). The d1 means
|
395
|
+
depth 1 -- the method was either recursively called (directly or indirectly).
|
396
|
+
|
397
|
+
However, the self time values for recursive calls should always
|
398
|
+
be accurate. It is also believed that the total times are
|
399
|
+
accurate, but these should be carefully analyzed to verify their veracity.
|
400
|
+
|
401
|
+
== Multi-threaded Applications
|
402
|
+
|
403
|
+
Unfortunately, Ruby does not provide an internal api
|
404
|
+
for detecting thread context switches in 1.8. As a result, the
|
405
|
+
timings ruby-prof reports for each thread may be slightly
|
406
|
+
inaccurate. In particular, this will happen for newly
|
407
|
+
spawned threads that go to sleep immediately (their first call).
|
408
|
+
For instance, if you use Ruby's timeout library to wait for 2 seconds,
|
409
|
+
the 2 seconds will be assigned to the foreground thread
|
410
|
+
and not the newly created background thread. These errors
|
411
|
+
can largely be avoided if the background thread performs any
|
412
|
+
operation before going to sleep.
|
413
|
+
|
414
|
+
== Performance
|
415
|
+
|
416
|
+
Significant effort has been put into reducing ruby-prof's overhead
|
417
|
+
as much as possible. Our tests show that the overhead associated
|
418
|
+
with profiling code varies considerably with the code being
|
419
|
+
profiled. Most programs will run approximately twice as slow
|
420
|
+
while highly recursive programs (like the fibonacci series test)
|
421
|
+
will run three times slower.
|
422
|
+
|
423
|
+
Because of some threading difficulties in 1.9, it currently runs a bit slower there.
|
424
|
+
http://www.ruby-forum.com/topic/201329
|
425
|
+
|
426
|
+
== License
|
427
|
+
|
428
|
+
See LICENSE for license information.
|
429
|
+
|
430
|
+
== Development
|
431
|
+
|
432
|
+
Code is located at http://github.com/rdp/ruby-prof
|