2 # Architectures that offer an FTRACE implementation should select HAVE_FTRACE:
7 config HAVE_DYNAMIC_FTRACE
10 config TRACER_MAX_TRACE
20 bool "Kernel Function Tracer"
21 depends on HAVE_FTRACE
24 select CONTEXT_SWITCH_TRACER
26 Enable the kernel to trace every kernel function. This is done
27 by using a compiler feature to insert a small, 5-byte No-Operation
28 instruction to the beginning of every kernel function, which NOP
29 sequence is then dynamically patched into a tracer call when
30 tracing is enabled by the administrator. If it's runtime disabled
31 (the bootup default), then the overhead of the instructions is very
32 small and not measurable even in micro-benchmarks.
35 bool "Interrupts-off Latency Tracer"
37 depends on TRACE_IRQFLAGS_SUPPORT
38 depends on GENERIC_TIME
39 depends on HAVE_FTRACE
42 select TRACER_MAX_TRACE
44 This option measures the time spent in irqs-off critical
45 sections, with microsecond accuracy.
47 The default measurement method is a maximum search, which is
48 disabled by default and can be runtime (re-)started
51 echo 0 > /debugfs/tracing/tracing_max_latency
53 (Note that kernel size and overhead increases with this option
54 enabled. This option and the preempt-off timing option can be
55 used together or separately.)
58 bool "Preemption-off Latency Tracer"
60 depends on GENERIC_TIME
62 depends on HAVE_FTRACE
64 select TRACER_MAX_TRACE
66 This option measures the time spent in preemption off critical
67 sections, with microsecond accuracy.
69 The default measurement method is a maximum search, which is
70 disabled by default and can be runtime (re-)started
73 echo 0 > /debugfs/tracing/tracing_max_latency
75 (Note that kernel size and overhead increases with this option
76 enabled. This option and the irqs-off timing option can be
77 used together or separately.)
84 This tracer provides the trace needed by the 'Sysprof' userspace
88 bool "Scheduling Latency Tracer"
89 depends on HAVE_FTRACE
91 select CONTEXT_SWITCH_TRACER
92 select TRACER_MAX_TRACE
94 This tracer tracks the latency of the highest priority task
95 to be scheduled in, starting from the point it has woken up.
97 config CONTEXT_SWITCH_TRACER
98 bool "Trace process context switches"
99 depends on HAVE_FTRACE
103 This tracer gets called from the context switch and records
104 all switching of tasks.
106 config DYNAMIC_FTRACE
107 bool "enable/disable ftrace tracepoints dynamically"
109 depends on HAVE_DYNAMIC_FTRACE
112 This option will modify all the calls to ftrace dynamically
113 (will patch them out of the binary image and replaces them
114 with a No-Op instruction) as they are called. A table is
115 created to dynamically enable them again.
117 This way a CONFIG_FTRACE kernel is slightly larger, but otherwise
118 has native performance as long as no tracing is active.
120 The changes to the code are done by a kernel thread that
121 wakes up once a second and checks to see if any ftrace calls
122 were made. If so, it runs stop_machine (stops all CPUS)
123 and modifies the code to jump over the call to ftrace.
125 config FTRACE_SELFTEST
128 config FTRACE_STARTUP_TEST
129 bool "Perform a startup test on ftrace"
131 select FTRACE_SELFTEST
133 This option performs a series of startup tests on ftrace. On bootup
134 a series of tests are made to verify that the tracer is
135 functioning properly. It will do tests on all the configured