1# 2# Architectures that offer an FUNCTION_TRACER implementation should 3# select HAVE_FUNCTION_TRACER: 4# 5 6config USER_STACKTRACE_SUPPORT 7 bool 8 9config NOP_TRACER 10 bool 11 12config HAVE_FTRACE_NMI_ENTER 13 bool 14 help 15 See Documentation/trace/ftrace-implementation.txt 16 17config HAVE_FUNCTION_TRACER 18 bool 19 help 20 See Documentation/trace/ftrace-implementation.txt 21 22config HAVE_FUNCTION_GRAPH_TRACER 23 bool 24 help 25 See Documentation/trace/ftrace-implementation.txt 26 27config HAVE_FUNCTION_GRAPH_FP_TEST 28 bool 29 help 30 An arch may pass in a unique value (frame pointer) to both the 31 entering and exiting of a function. On exit, the value is compared 32 and if it does not match, then it will panic the kernel. 33 34config HAVE_FUNCTION_TRACE_MCOUNT_TEST 35 bool 36 help 37 See Documentation/trace/ftrace-implementation.txt 38 39config HAVE_DYNAMIC_FTRACE 40 bool 41 help 42 See Documentation/trace/ftrace-implementation.txt 43 44config HAVE_FTRACE_MCOUNT_RECORD 45 bool 46 help 47 See Documentation/trace/ftrace-implementation.txt 48 49config HAVE_HW_BRANCH_TRACER 50 bool 51 52config HAVE_SYSCALL_TRACEPOINTS 53 bool 54 help 55 See Documentation/trace/ftrace-implementation.txt 56 57config TRACER_MAX_TRACE 58 bool 59 60config RING_BUFFER 61 bool 62 63config FTRACE_NMI_ENTER 64 bool 65 depends on HAVE_FTRACE_NMI_ENTER 66 default y 67 68config EVENT_TRACING 69 select CONTEXT_SWITCH_TRACER 70 bool 71 72config CONTEXT_SWITCH_TRACER 73 bool 74 75config RING_BUFFER_ALLOW_SWAP 76 bool 77 help 78 Allow the use of ring_buffer_swap_cpu. 79 Adds a very slight overhead to tracing when enabled. 80 81# All tracer options should select GENERIC_TRACER. For those options that are 82# enabled by all tracers (context switch and event tracer) they select TRACING. 83# This allows those options to appear when no other tracer is selected. But the 84# options do not appear when something else selects it. We need the two options 85# GENERIC_TRACER and TRACING to avoid circular dependencies to accomplish the 86# hidding of the automatic options. 87 88config TRACING 89 bool 90 select DEBUG_FS 91 select RING_BUFFER 92 select STACKTRACE if STACKTRACE_SUPPORT 93 select TRACEPOINTS 94 select NOP_TRACER 95 select BINARY_PRINTF 96 select EVENT_TRACING 97 98config GENERIC_TRACER 99 bool 100 select TRACING 101 102# 103# Minimum requirements an architecture has to meet for us to 104# be able to offer generic tracing facilities: 105# 106config TRACING_SUPPORT 107 bool 108 # PPC32 has no irqflags tracing support, but it can use most of the 109 # tracers anyway, they were tested to build and work. Note that new 110 # exceptions to this list aren't welcomed, better implement the 111 # irqflags tracing for your architecture. 112 depends on TRACE_IRQFLAGS_SUPPORT || PPC32 113 depends on STACKTRACE_SUPPORT 114 default y 115 116if TRACING_SUPPORT 117 118menuconfig FTRACE 119 bool "Tracers" 120 default y if DEBUG_KERNEL 121 help 122 Enable the kernel tracing infrastructure. 123 124if FTRACE 125 126config FUNCTION_TRACER 127 bool "Kernel Function Tracer" 128 depends on HAVE_FUNCTION_TRACER 129 select FRAME_POINTER 130 select KALLSYMS 131 select GENERIC_TRACER 132 select CONTEXT_SWITCH_TRACER 133 help 134 Enable the kernel to trace every kernel function. This is done 135 by using a compiler feature to insert a small, 5-byte No-Operation 136 instruction to the beginning of every kernel function, which NOP 137 sequence is then dynamically patched into a tracer call when 138 tracing is enabled by the administrator. If it's runtime disabled 139 (the bootup default), then the overhead of the instructions is very 140 small and not measurable even in micro-benchmarks. 141 142config FUNCTION_GRAPH_TRACER 143 bool "Kernel Function Graph Tracer" 144 depends on HAVE_FUNCTION_GRAPH_TRACER 145 depends on FUNCTION_TRACER 146 depends on !X86_32 || !CC_OPTIMIZE_FOR_SIZE 147 default y 148 help 149 Enable the kernel to trace a function at both its return 150 and its entry. 151 Its first purpose is to trace the duration of functions and 152 draw a call graph for each thread with some information like 153 the return value. This is done by setting the current return 154 address on the current task structure into a stack of calls. 155 156 157config IRQSOFF_TRACER 158 bool "Interrupts-off Latency Tracer" 159 default n 160 depends on TRACE_IRQFLAGS_SUPPORT 161 depends on GENERIC_TIME 162 select TRACE_IRQFLAGS 163 select GENERIC_TRACER 164 select TRACER_MAX_TRACE 165 select RING_BUFFER_ALLOW_SWAP 166 help 167 This option measures the time spent in irqs-off critical 168 sections, with microsecond accuracy. 169 170 The default measurement method is a maximum search, which is 171 disabled by default and can be runtime (re-)started 172 via: 173 174 echo 0 > /sys/kernel/debug/tracing/tracing_max_latency 175 176 (Note that kernel size and overhead increases with this option 177 enabled. This option and the preempt-off timing option can be 178 used together or separately.) 179 180config PREEMPT_TRACER 181 bool "Preemption-off Latency Tracer" 182 default n 183 depends on GENERIC_TIME 184 depends on PREEMPT 185 select GENERIC_TRACER 186 select TRACER_MAX_TRACE 187 select RING_BUFFER_ALLOW_SWAP 188 help 189 This option measures the time spent in preemption off critical 190 sections, with microsecond accuracy. 191 192 The default measurement method is a maximum search, which is 193 disabled by default and can be runtime (re-)started 194 via: 195 196 echo 0 > /sys/kernel/debug/tracing/tracing_max_latency 197 198 (Note that kernel size and overhead increases with this option 199 enabled. This option and the irqs-off timing option can be 200 used together or separately.) 201 202config SYSPROF_TRACER 203 bool "Sysprof Tracer" 204 depends on X86 205 select GENERIC_TRACER 206 select CONTEXT_SWITCH_TRACER 207 help 208 This tracer provides the trace needed by the 'Sysprof' userspace 209 tool. 210 211config SCHED_TRACER 212 bool "Scheduling Latency Tracer" 213 select GENERIC_TRACER 214 select CONTEXT_SWITCH_TRACER 215 select TRACER_MAX_TRACE 216 help 217 This tracer tracks the latency of the highest priority task 218 to be scheduled in, starting from the point it has woken up. 219 220config ENABLE_DEFAULT_TRACERS 221 bool "Trace process context switches and events" 222 depends on !GENERIC_TRACER 223 select TRACING 224 help 225 This tracer hooks to various trace points in the kernel 226 allowing the user to pick and choose which trace point they 227 want to trace. It also includes the sched_switch tracer plugin. 228 229config FTRACE_SYSCALLS 230 bool "Trace syscalls" 231 depends on HAVE_SYSCALL_TRACEPOINTS 232 select GENERIC_TRACER 233 select KALLSYMS 234 help 235 Basic tracer to catch the syscall entry and exit events. 236 237config BOOT_TRACER 238 bool "Trace boot initcalls" 239 select GENERIC_TRACER 240 select CONTEXT_SWITCH_TRACER 241 help 242 This tracer helps developers to optimize boot times: it records 243 the timings of the initcalls and traces key events and the identity 244 of tasks that can cause boot delays, such as context-switches. 245 246 Its aim is to be parsed by the scripts/bootgraph.pl tool to 247 produce pretty graphics about boot inefficiencies, giving a visual 248 representation of the delays during initcalls - but the raw 249 /debug/tracing/trace text output is readable too. 250 251 You must pass in initcall_debug and ftrace=initcall to the kernel 252 command line to enable this on bootup. 253 254config TRACE_BRANCH_PROFILING 255 bool 256 select GENERIC_TRACER 257 258choice 259 prompt "Branch Profiling" 260 default BRANCH_PROFILE_NONE 261 help 262 The branch profiling is a software profiler. It will add hooks 263 into the C conditionals to test which path a branch takes. 264 265 The likely/unlikely profiler only looks at the conditions that 266 are annotated with a likely or unlikely macro. 267 268 The "all branch" profiler will profile every if statement in the 269 kernel. This profiler will also enable the likely/unlikely 270 profiler as well. 271 272 Either of the above profilers add a bit of overhead to the system. 273 If unsure choose "No branch profiling". 274 275config BRANCH_PROFILE_NONE 276 bool "No branch profiling" 277 help 278 No branch profiling. Branch profiling adds a bit of overhead. 279 Only enable it if you want to analyse the branching behavior. 280 Otherwise keep it disabled. 281 282config PROFILE_ANNOTATED_BRANCHES 283 bool "Trace likely/unlikely profiler" 284 select TRACE_BRANCH_PROFILING 285 help 286 This tracer profiles all the the likely and unlikely macros 287 in the kernel. It will display the results in: 288 289 /sys/kernel/debug/tracing/profile_annotated_branch 290 291 Note: this will add a significant overhead, only turn this 292 on if you need to profile the system's use of these macros. 293 294config PROFILE_ALL_BRANCHES 295 bool "Profile all if conditionals" 296 select TRACE_BRANCH_PROFILING 297 help 298 This tracer profiles all branch conditions. Every if () 299 taken in the kernel is recorded whether it hit or miss. 300 The results will be displayed in: 301 302 /sys/kernel/debug/tracing/profile_branch 303 304 This option also enables the likely/unlikely profiler. 305 306 This configuration, when enabled, will impose a great overhead 307 on the system. This should only be enabled when the system 308 is to be analyzed 309endchoice 310 311config TRACING_BRANCHES 312 bool 313 help 314 Selected by tracers that will trace the likely and unlikely 315 conditions. This prevents the tracers themselves from being 316 profiled. Profiling the tracing infrastructure can only happen 317 when the likelys and unlikelys are not being traced. 318 319config BRANCH_TRACER 320 bool "Trace likely/unlikely instances" 321 depends on TRACE_BRANCH_PROFILING 322 select TRACING_BRANCHES 323 help 324 This traces the events of likely and unlikely condition 325 calls in the kernel. The difference between this and the 326 "Trace likely/unlikely profiler" is that this is not a 327 histogram of the callers, but actually places the calling 328 events into a running trace buffer to see when and where the 329 events happened, as well as their results. 330 331 Say N if unsure. 332 333config POWER_TRACER 334 bool "Trace power consumption behavior" 335 depends on X86 336 select GENERIC_TRACER 337 help 338 This tracer helps developers to analyze and optimize the kernels 339 power management decisions, specifically the C-state and P-state 340 behavior. 341 342 343config STACK_TRACER 344 bool "Trace max stack" 345 depends on HAVE_FUNCTION_TRACER 346 select FUNCTION_TRACER 347 select STACKTRACE 348 select KALLSYMS 349 help 350 This special tracer records the maximum stack footprint of the 351 kernel and displays it in /sys/kernel/debug/tracing/stack_trace. 352 353 This tracer works by hooking into every function call that the 354 kernel executes, and keeping a maximum stack depth value and 355 stack-trace saved. If this is configured with DYNAMIC_FTRACE 356 then it will not have any overhead while the stack tracer 357 is disabled. 358 359 To enable the stack tracer on bootup, pass in 'stacktrace' 360 on the kernel command line. 361 362 The stack tracer can also be enabled or disabled via the 363 sysctl kernel.stack_tracer_enabled 364 365 Say N if unsure. 366 367config HW_BRANCH_TRACER 368 depends on HAVE_HW_BRANCH_TRACER 369 bool "Trace hw branches" 370 select GENERIC_TRACER 371 help 372 This tracer records all branches on the system in a circular 373 buffer giving access to the last N branches for each cpu. 374 375config KMEMTRACE 376 bool "Trace SLAB allocations" 377 select GENERIC_TRACER 378 help 379 kmemtrace provides tracing for slab allocator functions, such as 380 kmalloc, kfree, kmem_cache_alloc, kmem_cache_free etc.. Collected 381 data is then fed to the userspace application in order to analyse 382 allocation hotspots, internal fragmentation and so on, making it 383 possible to see how well an allocator performs, as well as debug 384 and profile kernel code. 385 386 This requires an userspace application to use. See 387 Documentation/trace/kmemtrace.txt for more information. 388 389 Saying Y will make the kernel somewhat larger and slower. However, 390 if you disable kmemtrace at run-time or boot-time, the performance 391 impact is minimal (depending on the arch the kernel is built for). 392 393 If unsure, say N. 394 395config WORKQUEUE_TRACER 396 bool "Trace workqueues" 397 select GENERIC_TRACER 398 help 399 The workqueue tracer provides some statistical informations 400 about each cpu workqueue thread such as the number of the 401 works inserted and executed since their creation. It can help 402 to evaluate the amount of work each of them have to perform. 403 For example it can help a developer to decide whether he should 404 choose a per cpu workqueue instead of a singlethreaded one. 405 406config BLK_DEV_IO_TRACE 407 bool "Support for tracing block io actions" 408 depends on SYSFS 409 depends on BLOCK 410 select RELAY 411 select DEBUG_FS 412 select TRACEPOINTS 413 select GENERIC_TRACER 414 select STACKTRACE 415 help 416 Say Y here if you want to be able to trace the block layer actions 417 on a given queue. Tracing allows you to see any traffic happening 418 on a block device queue. For more information (and the userspace 419 support tools needed), fetch the blktrace tools from: 420 421 git://git.kernel.dk/blktrace.git 422 423 Tracing also is possible using the ftrace interface, e.g.: 424 425 echo 1 > /sys/block/sda/sda1/trace/enable 426 echo blk > /sys/kernel/debug/tracing/current_tracer 427 cat /sys/kernel/debug/tracing/trace_pipe 428 429 If unsure, say N. 430 431config KPROBE_EVENT 432 depends on KPROBES 433 depends on X86 434 bool "Enable kprobes-based dynamic events" 435 select TRACING 436 default y 437 help 438 This allows the user to add tracing events (similar to tracepoints) on the fly 439 via the ftrace interface. See Documentation/trace/kprobetrace.txt 440 for more details. 441 442 Those events can be inserted wherever kprobes can probe, and record 443 various register and memory values. 444 445 This option is also required by perf-probe subcommand of perf tools. If 446 you want to use perf tools, this option is strongly recommended. 447 448config DYNAMIC_FTRACE 449 bool "enable/disable ftrace tracepoints dynamically" 450 depends on FUNCTION_TRACER 451 depends on HAVE_DYNAMIC_FTRACE 452 default y 453 help 454 This option will modify all the calls to ftrace dynamically 455 (will patch them out of the binary image and replaces them 456 with a No-Op instruction) as they are called. A table is 457 created to dynamically enable them again. 458 459 This way a CONFIG_FUNCTION_TRACER kernel is slightly larger, but otherwise 460 has native performance as long as no tracing is active. 461 462 The changes to the code are done by a kernel thread that 463 wakes up once a second and checks to see if any ftrace calls 464 were made. If so, it runs stop_machine (stops all CPUS) 465 and modifies the code to jump over the call to ftrace. 466 467config FUNCTION_PROFILER 468 bool "Kernel function profiler" 469 depends on FUNCTION_TRACER 470 default n 471 help 472 This option enables the kernel function profiler. A file is created 473 in debugfs called function_profile_enabled which defaults to zero. 474 When a 1 is echoed into this file profiling begins, and when a 475 zero is entered, profiling stops. A file in the trace_stats 476 directory called functions, that show the list of functions that 477 have been hit and their counters. 478 479 If in doubt, say N 480 481config FTRACE_MCOUNT_RECORD 482 def_bool y 483 depends on DYNAMIC_FTRACE 484 depends on HAVE_FTRACE_MCOUNT_RECORD 485 486config FTRACE_SELFTEST 487 bool 488 489config FTRACE_STARTUP_TEST 490 bool "Perform a startup test on ftrace" 491 depends on GENERIC_TRACER 492 select FTRACE_SELFTEST 493 help 494 This option performs a series of startup tests on ftrace. On bootup 495 a series of tests are made to verify that the tracer is 496 functioning properly. It will do tests on all the configured 497 tracers of ftrace. 498 499config EVENT_TRACE_TEST_SYSCALLS 500 bool "Run selftest on syscall events" 501 depends on FTRACE_STARTUP_TEST 502 help 503 This option will also enable testing every syscall event. 504 It only enables the event and disables it and runs various loads 505 with the event enabled. This adds a bit more time for kernel boot 506 up since it runs this on every system call defined. 507 508 TBD - enable a way to actually call the syscalls as we test their 509 events 510 511config MMIOTRACE 512 bool "Memory mapped IO tracing" 513 depends on HAVE_MMIOTRACE_SUPPORT && PCI 514 select GENERIC_TRACER 515 help 516 Mmiotrace traces Memory Mapped I/O access and is meant for 517 debugging and reverse engineering. It is called from the ioremap 518 implementation and works via page faults. Tracing is disabled by 519 default and can be enabled at run-time. 520 521 See Documentation/trace/mmiotrace.txt. 522 If you are not helping to develop drivers, say N. 523 524config MMIOTRACE_TEST 525 tristate "Test module for mmiotrace" 526 depends on MMIOTRACE && m 527 help 528 This is a dumb module for testing mmiotrace. It is very dangerous 529 as it will write garbage to IO memory starting at a given address. 530 However, it should be safe to use on e.g. unused portion of VRAM. 531 532 Say N, unless you absolutely know what you are doing. 533 534config RING_BUFFER_BENCHMARK 535 tristate "Ring buffer benchmark stress tester" 536 depends on RING_BUFFER 537 help 538 This option creates a test to stress the ring buffer and bench mark it. 539 It creates its own ring buffer such that it will not interfer with 540 any other users of the ring buffer (such as ftrace). It then creates 541 a producer and consumer that will run for 10 seconds and sleep for 542 10 seconds. Each interval it will print out the number of events 543 it recorded and give a rough estimate of how long each iteration took. 544 545 It does not disable interrupts or raise its priority, so it may be 546 affected by processes that are running. 547 548 If unsure, say N 549 550endif # FTRACE 551 552endif # TRACING_SUPPORT 553 554