1OPCONTROL(1) General Commands Manual OPCONTROL(1)
2
6 opcontrol - control OProfile profiling
7
9 opcontrol [ options ]
10
12 opcontrol can be used to start profiling, end a profiling session, dump
13 profile data, and set up the profiling parameters.
14
17 --help Show help message.
18 --version
20 Show version.
21 --list-events
23 Shows the monitorable events.
24 --init Load the OProfile module if required and make the OProfile
26 driver interface available.
27 --setup
29 Followed by list options for profiling setup. Store setup in
30 ~root/.oprofile/daemonrc. Optional.
31 --status
33 Show configuration information.
34 --start-daemon
36 Start the oprofile daemon without starting profiling. Not avail‐
37 able in 2.2/2.4 kernels.
38 --start
40 Start data collection with either arguments provided by --setup
41 or with information saved in ~root/.oprofile/daemonrc.
42 --dump Force a flush of the collected profiling data to the daemon.
44 --stop Stop data collection. Not available in 2.2/2.4 kernels.
46 --shutdown
48 Stop data collection and kill the daemon.
49 --reset
51 Clear out data from current session, but leaves saved sessions.
52 --save=sessionname
54 Save data from current session to sessionname.
55 --deinit
57 Shut down daemon. Unload the oprofile module and oprofilefs.
58 --buffer-size=num
60 Set kernel buffer to num samples. When using a 2.6 kernel buffer
61 watershed need to be tweaked when changing this value.
62 --buffer-watershed=num
64 Set kernel buffer watershed to num samples (2.6 only). When
65 it'll remain only buffer-size - buffer-watershed free entry in
66 the kernel buffer data will be flushed to daemon, most usefull
67 value are in the range [0.25 - 0.5] * buffer-size.
68 --cpu-buffer-size=num
70 Set kernel per cpu buffer to num samples (2.6 only). If you pro‐
71 file at high rate it can help to increase this if the log file
72 show excessive count of sample lost cpu buffer overflow.
73 --event=[event|"default"]
75 Add an event to measure for the hardware performance counters,
76 or "default" for the default event. The event is of the form
77 "CPU_CLK_UNHALTED:30000:0:1:1" where the numeric values are
78 count, unit mask, kernel-space counting, user-space counting,
79 respectively.
80 --separate=[none,lib,kernel,thread,cpu,all]
82 Separate samples based on the given separator. 'lib' separates
83 dynamically linked library samples per application. 'kernel'
84 separates kernel and kernel module samples per application;
85 'kernel' implies 'library'. 'thread' gives separation for each
86 thread and task. 'cpu' separates for each CPU. 'all' implies
87 all of the above options and 'none' turns off separation.
88 --callgraph=#depth
90 Enable callgraph sample collection with a maximum depth. Use 0
91 to disable callgraph profiling. This option is available on x86
92 using a 2.6+ kernel with callgraph support enabled. It is also
93 available on PowerPC using a 2.6.17+ kernel. This option is not
94 yet supported on Xen.
95 --image=[name,name...|"all"]
97 Only profile the given absolute paths to binaries, or "all" to
98 profile everything (the default).
99 --vmlinux=file
101 vmlinux kernel image.
102 --no-vmlinux
104 Use this when you don't have a kernel vmlinux file, and you
105 don't want to profile the kernel.
106 --verbose
108 Be verbose in the daemon log. This has a high overhead.
109 --kernel-range=start,end
111 Set kernel range vma address in hexadecimal.
112
115 --xen=file
116 Xen image
117 --active-domains=<list>
119 List of domain ids participating in a multi-domain profiling
120 session. Each of the specified domains must run an instance of
121 oprofile. The sequence of opcontrol commands in each domain must
122 follow a given order which is specified in the oprofile user
123 manual. If more than one domain is specified in <list> they
124 should be separated using commas. This option can only be used
125 in domain 0 which is the only domain that can coordinate a
126 multi-domain profiling session. Including domain 0 in the list
127 of active domains is optional. (e.g. --active-domains=2,5,6 and
128 --active-domains=0,2,5,6 are equivalent). This option can only
129 be specified if --start-daemon is also specified and it is only
130 valid for the current run of the oprofile daemon; e.g. the list
131 of active domains is not persistent.
132 --passive-domains=<list>or--domains=<list>
134 List of domain ids to be profiled, separated by commas. As
135 opposed to the --active-domains option, the domains specified
136 with this option do not need to run oprofile. This makes profil‐
137 ing multiple domains easier. However, with the passive-domains
138 option, samples in user level processes and kernel modules can‐
139 not be mapped to specific symbols and are aggregated under a
140 generic class. Both --active-domains and --passive-domains
141 options can be specified in the same command, but the same
142 domain cannot be specified in both options. This option can only
143 be specified if either --start or --start-daemon is specified on
144 the same command and it is only valid for the current run of the
145 oprofile daemon; e.g. the list of passive domains is not persis‐
146 tent.
147 --passive-images=<list>or--domains-images=<list>
149 List of kernel images associated with the domains specified in
150 the --passive-domains option, also separated by commas. The
151 association between the images and domains is based on the order
152 they are specified in both options.
153
156 No special environment variables are recognised by opcontrol.
157
160 /root/.oprofile/daemonrc
161 Configuration file for opcontrol
162 /var/lib/oprofile/samples/
164 The location of the generated sample files.
165
168 This man page is current for oprofile-0.9.2.
169
172 /usr/share/doc/oprofile/, oprofile(1)
1734th Berkeley Distribution Fri 23 March 2007 OPCONTROL(1)