2 # Architectures that offer an FUNCTION_TRACER implementation should
3 # select HAVE_FUNCTION_TRACER:
9 config HAVE_FUNCTION_TRACER
12 config HAVE_FUNCTION_TRACE_MCOUNT_TEST
15 This gets selected when the arch tests the function_trace_stop
16 variable at the mcount call site. Otherwise, this variable
17 is tested by the called function.
19 config HAVE_DYNAMIC_FTRACE
22 config HAVE_FTRACE_MCOUNT_RECORD
25 config TRACER_MAX_TRACE
35 select STACKTRACE if STACKTRACE_SUPPORT
41 config FUNCTION_TRACER
42 bool "Kernel Function Tracer"
43 depends on HAVE_FUNCTION_TRACER
44 depends on DEBUG_KERNEL
47 select CONTEXT_SWITCH_TRACER
49 Enable the kernel to trace every kernel function. This is done
50 by using a compiler feature to insert a small, 5-byte No-Operation
51 instruction to the beginning of every kernel function, which NOP
52 sequence is then dynamically patched into a tracer call when
53 tracing is enabled by the administrator. If it's runtime disabled
54 (the bootup default), then the overhead of the instructions is very
55 small and not measurable even in micro-benchmarks.
58 bool "Interrupts-off Latency Tracer"
60 depends on TRACE_IRQFLAGS_SUPPORT
61 depends on GENERIC_TIME
62 depends on DEBUG_KERNEL
65 select TRACER_MAX_TRACE
67 This option measures the time spent in irqs-off critical
68 sections, with microsecond accuracy.
70 The default measurement method is a maximum search, which is
71 disabled by default and can be runtime (re-)started
74 echo 0 > /debugfs/tracing/tracing_max_latency
76 (Note that kernel size and overhead increases with this option
77 enabled. This option and the preempt-off timing option can be
78 used together or separately.)
81 bool "Preemption-off Latency Tracer"
83 depends on GENERIC_TIME
85 depends on DEBUG_KERNEL
87 select TRACER_MAX_TRACE
89 This option measures the time spent in preemption off critical
90 sections, with microsecond accuracy.
92 The default measurement method is a maximum search, which is
93 disabled by default and can be runtime (re-)started
96 echo 0 > /debugfs/tracing/tracing_max_latency
98 (Note that kernel size and overhead increases with this option
99 enabled. This option and the irqs-off timing option can be
100 used together or separately.)
102 config SYSPROF_TRACER
103 bool "Sysprof Tracer"
107 This tracer provides the trace needed by the 'Sysprof' userspace
111 bool "Scheduling Latency Tracer"
112 depends on DEBUG_KERNEL
114 select CONTEXT_SWITCH_TRACER
115 select TRACER_MAX_TRACE
117 This tracer tracks the latency of the highest priority task
118 to be scheduled in, starting from the point it has woken up.
120 config CONTEXT_SWITCH_TRACER
121 bool "Trace process context switches"
122 depends on DEBUG_KERNEL
126 This tracer gets called from the context switch and records
127 all switching of tasks.
130 bool "Trace boot initcalls"
131 depends on DEBUG_KERNEL
133 select CONTEXT_SWITCH_TRACER
135 This tracer helps developers to optimize boot times: it records
136 the timings of the initcalls and traces key events and the identity
137 of tasks that can cause boot delays, such as context-switches.
139 Its aim is to be parsed by the /scripts/bootgraph.pl tool to
140 produce pretty graphics about boot inefficiencies, giving a visual
141 representation of the delays during initcalls - but the raw
142 /debug/tracing/trace text output is readable too.
144 ( Note that tracing self tests can't be enabled if this tracer is
145 selected, because the self-tests are an initcall as well and that
146 would invalidate the boot trace. )
149 bool "Trace max stack"
150 depends on HAVE_FUNCTION_TRACER
151 depends on DEBUG_KERNEL
152 select FUNCTION_TRACER
155 This special tracer records the maximum stack footprint of the
156 kernel and displays it in debugfs/tracing/stack_trace.
158 This tracer works by hooking into every function call that the
159 kernel executes, and keeping a maximum stack depth value and
160 stack-trace saved. Because this logic has to execute in every
161 kernel function, all the time, this option can slow down the
162 kernel measurably and is generally intended for kernel
167 config DYNAMIC_FTRACE
168 bool "enable/disable ftrace tracepoints dynamically"
169 depends on FUNCTION_TRACER
170 depends on HAVE_DYNAMIC_FTRACE
171 depends on DEBUG_KERNEL
174 This option will modify all the calls to ftrace dynamically
175 (will patch them out of the binary image and replaces them
176 with a No-Op instruction) as they are called. A table is
177 created to dynamically enable them again.
179 This way a CONFIG_FUNCTION_TRACER kernel is slightly larger, but otherwise
180 has native performance as long as no tracing is active.
182 The changes to the code are done by a kernel thread that
183 wakes up once a second and checks to see if any ftrace calls
184 were made. If so, it runs stop_machine (stops all CPUS)
185 and modifies the code to jump over the call to ftrace.
187 config FTRACE_MCOUNT_RECORD
189 depends on DYNAMIC_FTRACE
190 depends on HAVE_FTRACE_MCOUNT_RECORD
192 config FTRACE_SELFTEST
195 config FTRACE_STARTUP_TEST
196 bool "Perform a startup test on ftrace"
197 depends on TRACING && DEBUG_KERNEL && !BOOT_TRACER
198 select FTRACE_SELFTEST
200 This option performs a series of startup tests on ftrace. On bootup
201 a series of tests are made to verify that the tracer is
202 functioning properly. It will do tests on all the configured