blob: 22eff0624bc8aa677d78b0b4dffb42c2034bc7c2 [file] [log] [blame]
Dave Hansen604ff0d2013-07-01 13:04:49 -07001menu "printk and dmesg options"
Linus Torvalds1da177e2005-04-16 15:20:36 -07002
3config PRINTK_TIME
4 bool "Show timing information on printks"
Randy Dunlapd3b8b6e2006-12-06 20:36:38 -08005 depends on PRINTK
Linus Torvalds1da177e2005-04-16 15:20:36 -07006 help
Kay Sievers649e6ee2012-05-10 04:30:45 +02007 Selecting this option causes time stamps of the printk()
8 messages to be added to the output of the syslog() system
9 call and at the console.
10
11 The timestamp is always recorded internally, and exported
12 to /dev/kmsg. This flag just specifies if the timestamp should
13 be included, not that the timestamp is recorded.
14
15 The behavior is also controlled by the kernel command line
16 parameter printk.time=1. See Documentation/kernel-parameters.txt
Linus Torvalds1da177e2005-04-16 15:20:36 -070017
Alex Elder42a9dc02014-08-06 16:09:01 -070018config MESSAGE_LOGLEVEL_DEFAULT
Mandeep Singh Baines5af5bcb2011-03-22 16:34:23 -070019 int "Default message log level (1-7)"
20 range 1 7
21 default "4"
22 help
23 Default log level for printk statements with no specified priority.
24
25 This was hard-coded to KERN_WARNING since at least 2.6.10 but folks
26 that are auditing their logs closely may want to set it to a lower
27 priority.
28
Dave Hansen604ff0d2013-07-01 13:04:49 -070029config BOOT_PRINTK_DELAY
30 bool "Delay each boot printk message by N milliseconds"
31 depends on DEBUG_KERNEL && PRINTK && GENERIC_CALIBRATE_DELAY
32 help
33 This build option allows you to read kernel boot messages
34 by inserting a short delay after each one. The delay is
35 specified in milliseconds on the kernel command line,
36 using "boot_delay=N".
37
38 It is likely that you would also need to use "lpj=M" to preset
39 the "loops per jiffie" value.
40 See a previous boot log for the "lpj" value to use for your
41 system, and then set "lpj=M" before setting "boot_delay=N".
42 NOTE: Using this option may adversely affect SMP systems.
43 I.e., processors other than the first one may not boot up.
44 BOOT_PRINTK_DELAY also may cause LOCKUP_DETECTOR to detect
45 what it believes to be lockup conditions.
46
47config DYNAMIC_DEBUG
48 bool "Enable dynamic printk() support"
49 default n
50 depends on PRINTK
51 depends on DEBUG_FS
52 help
53
54 Compiles debug level messages into the kernel, which would not
55 otherwise be available at runtime. These messages can then be
56 enabled/disabled based on various levels of scope - per source file,
57 function, module, format string, and line number. This mechanism
58 implicitly compiles in all pr_debug() and dev_dbg() calls, which
59 enlarges the kernel text size by about 2%.
60
61 If a source file is compiled with DEBUG flag set, any
62 pr_debug() calls in it are enabled by default, but can be
63 disabled at runtime as below. Note that DEBUG flag is
64 turned on by many CONFIG_*DEBUG* options.
65
66 Usage:
67
68 Dynamic debugging is controlled via the 'dynamic_debug/control' file,
69 which is contained in the 'debugfs' filesystem. Thus, the debugfs
70 filesystem must first be mounted before making use of this feature.
71 We refer the control file as: <debugfs>/dynamic_debug/control. This
72 file contains a list of the debug statements that can be enabled. The
73 format for each line of the file is:
74
75 filename:lineno [module]function flags format
76
77 filename : source file of the debug statement
78 lineno : line number of the debug statement
79 module : module that contains the debug statement
80 function : function that contains the debug statement
81 flags : '=p' means the line is turned 'on' for printing
82 format : the format used for the debug statement
83
84 From a live system:
85
86 nullarbor:~ # cat <debugfs>/dynamic_debug/control
87 # filename:lineno [module]function flags format
88 fs/aio.c:222 [aio]__put_ioctx =_ "__put_ioctx:\040freeing\040%p\012"
89 fs/aio.c:248 [aio]ioctx_alloc =_ "ENOMEM:\040nr_events\040too\040high\012"
90 fs/aio.c:1770 [aio]sys_io_cancel =_ "calling\040cancel\012"
91
92 Example usage:
93
94 // enable the message at line 1603 of file svcsock.c
95 nullarbor:~ # echo -n 'file svcsock.c line 1603 +p' >
96 <debugfs>/dynamic_debug/control
97
98 // enable all the messages in file svcsock.c
99 nullarbor:~ # echo -n 'file svcsock.c +p' >
100 <debugfs>/dynamic_debug/control
101
102 // enable all the messages in the NFS server module
103 nullarbor:~ # echo -n 'module nfsd +p' >
104 <debugfs>/dynamic_debug/control
105
106 // enable all 12 messages in the function svc_process()
107 nullarbor:~ # echo -n 'func svc_process +p' >
108 <debugfs>/dynamic_debug/control
109
110 // disable all 12 messages in the function svc_process()
111 nullarbor:~ # echo -n 'func svc_process -p' >
112 <debugfs>/dynamic_debug/control
113
114 See Documentation/dynamic-debug-howto.txt for additional information.
115
116endmenu # "printk and dmesg options"
117
Dave Hansen6dfc0662013-07-01 13:04:46 -0700118menu "Compile-time checks and compiler options"
119
120config DEBUG_INFO
121 bool "Compile the kernel with debug info"
Linus Torvalds12b13832014-02-04 12:20:01 -0800122 depends on DEBUG_KERNEL && !COMPILE_TEST
Dave Hansen6dfc0662013-07-01 13:04:46 -0700123 help
124 If you say Y here the resulting kernel image will include
125 debugging info resulting in a larger kernel image.
126 This adds debug symbols to the kernel and modules (gcc -g), and
127 is needed if you intend to use kernel crashdump or binary object
128 tools like crash, kgdb, LKCD, gdb, etc on the kernel.
129 Say Y here only if you plan to debug the kernel.
130
131 If unsure, say N.
132
133config DEBUG_INFO_REDUCED
134 bool "Reduce debugging information"
135 depends on DEBUG_INFO
136 help
137 If you say Y here gcc is instructed to generate less debugging
138 information for structure types. This means that tools that
139 need full debugging information (like kgdb or systemtap) won't
140 be happy. But if you merely need debugging information to
141 resolve line numbers there is no loss. Advantage is that
142 build directory object sizes shrink dramatically over a full
143 DEBUG_INFO build and compile times are reduced too.
144 Only works with newer gcc versions.
145
Andi Kleen866ced92014-07-30 20:50:18 +0200146config DEBUG_INFO_SPLIT
147 bool "Produce split debuginfo in .dwo files"
148 depends on DEBUG_INFO
149 help
150 Generate debug info into separate .dwo files. This significantly
151 reduces the build directory size for builds with DEBUG_INFO,
152 because it stores the information only once on disk in .dwo
153 files instead of multiple times in object files and executables.
154 In addition the debug information is also compressed.
155
156 Requires recent gcc (4.7+) and recent gdb/binutils.
157 Any tool that packages or reads debug information would need
158 to know about the .dwo files and include them.
159 Incompatible with older versions of ccache.
160
Andi Kleenbfaf2dd2014-07-30 20:50:19 +0200161config DEBUG_INFO_DWARF4
162 bool "Generate dwarf4 debuginfo"
163 depends on DEBUG_INFO
164 help
165 Generate dwarf4 debug info. This requires recent versions
166 of gcc and gdb. It makes the debug information larger.
167 But it significantly improves the success of resolving
168 variables in gdb on optimized code.
169
Jan Kiszka3ee7b3f2015-02-17 13:46:36 -0800170config GDB_SCRIPTS
171 bool "Provide GDB scripts for kernel debugging"
172 depends on DEBUG_INFO
173 help
174 This creates the required links to GDB helper scripts in the
175 build directory. If you load vmlinux into gdb, the helper
176 scripts will be automatically imported by gdb as well, and
177 additional functions are available to analyze a Linux kernel
178 instance. See Documentation/gdb-kernel-debugging.txt for further
179 details.
180
Jeff Garzikde488442007-10-25 04:06:13 -0400181config ENABLE_WARN_DEPRECATED
182 bool "Enable __deprecated logic"
183 default y
184 help
185 Enable the __deprecated logic in the kernel build.
186 Disable this to suppress the "warning: 'foo' is deprecated
187 (declared at kernel/power/somefile.c:1234)" messages.
188
Andrew Mortoncebc04b2006-08-14 22:43:18 -0700189config ENABLE_MUST_CHECK
190 bool "Enable __must_check logic"
191 default y
192 help
193 Enable the __must_check logic in the kernel build. Disable this to
194 suppress the "warning: ignoring return value of 'foo', declared with
195 attribute warn_unused_result" messages.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700196
Andi Kleen35bb5b12008-02-22 15:15:03 +0100197config FRAME_WARN
198 int "Warn for stack frames larger than (needs gcc 4.4)"
199 range 0 8192
Andrey Ryabinin3f181b42015-10-22 13:32:24 -0700200 default 0 if KASAN
Kees Cook0e07f662016-10-27 17:46:41 -0700201 default 2048 if GCC_PLUGIN_LATENT_ENTROPY
Andi Kleen35bb5b12008-02-22 15:15:03 +0100202 default 1024 if !64BIT
203 default 2048 if 64BIT
204 help
205 Tell gcc to warn at build time for stack frames larger than this.
206 Setting this too low will cause a lot of warnings.
207 Setting it to 0 disables the warning.
208 Requires gcc 4.4
209
Randy Dunlap99657c72009-09-18 12:49:22 -0700210config STRIP_ASM_SYMS
211 bool "Strip assembler-generated symbols during link"
212 default n
213 help
214 Strip internal assembler-generated symbols during a link (symbols
215 that look like '.Lxxx') so they don't pollute the output of
216 get_wchan() and suchlike.
217
Andi Kleen1873e872012-03-28 11:51:18 -0700218config READABLE_ASM
219 bool "Generate readable assembler code"
220 depends on DEBUG_KERNEL
221 help
222 Disable some compiler optimizations that tend to generate human unreadable
223 assembler output. This may make the kernel slightly slower, but it helps
224 to keep kernel developers who have to stare a lot at assembler listings
225 sane.
226
Arjan van de Venf71d20e2006-06-28 04:26:45 -0700227config UNUSED_SYMBOLS
228 bool "Enable unused/obsolete exported symbols"
229 default y if X86
230 help
231 Unused but exported symbols make the kernel needlessly bigger. For
232 that reason most of these unused exports will soon be removed. This
233 option is provided temporarily to provide a transition period in case
234 some external kernel module needs one of these symbols anyway. If you
235 encounter such a case in your module, consider if you are actually
236 using the right API. (rationale: since nobody in the kernel is using
237 this in a module, there is a pretty good chance it's actually the
238 wrong interface to use). If you really need the symbol, please send a
239 mail to the linux kernel mailing list mentioning the symbol and why
240 you really need it, and what the merge plan to the mainline kernel for
241 your module is.
242
Joonsoo Kim48c96a32014-12-12 16:56:01 -0800243config PAGE_OWNER
244 bool "Track page owner"
245 depends on DEBUG_KERNEL && STACKTRACE_SUPPORT
246 select DEBUG_FS
247 select STACKTRACE
Joonsoo Kimf2ca0b52016-07-26 15:23:55 -0700248 select STACKDEPOT
Joonsoo Kim48c96a32014-12-12 16:56:01 -0800249 select PAGE_EXTENSION
250 help
251 This keeps track of what call chain is the owner of a page, may
252 help to find bare alloc_page(s) leaks. Even if you include this
253 feature on your build, it is disabled in default. You should pass
254 "page_owner=on" to boot parameter in order to enable it. Eats
255 a fair amount of memory if enabled. See tools/vm/page_owner_sort.c
256 for user-space helper.
257
258 If unsure, say N.
259
Trilok Soni65cfa9f2016-04-20 18:12:51 -0700260config PAGE_OWNER_ENABLE_DEFAULT
261 bool "Enable Track page owner by default"
262 depends on PAGE_OWNER
263 ---help---
264 This keeps track of what call chain is the owner of a page, may
265 help to find bare alloc_page(s) leaks. If you include this
266 feature on your build, it is enabled by default. You should pass
267 "page_owner=off" to boot parameter in order to disable it. Eats
268 a fair amount of memory if enabled. See tools/vm/page_owner_sort.c
269 for user-space helper.
270
Don Mullisbf4735a2006-12-10 02:18:37 -0800271config DEBUG_FS
272 bool "Debug Filesystem"
Nicolai Stange9fd4dce2016-03-22 14:11:13 +0100273 select SRCU
Don Mullisbf4735a2006-12-10 02:18:37 -0800274 help
275 debugfs is a virtual file system that kernel developers use to put
276 debugging files into. Enable this option to be able to read and
277 write to these files.
278
Robert P. J. Dayff543332008-05-20 00:06:00 +0200279 For detailed documentation on the debugfs API, see
280 Documentation/DocBook/filesystems.
281
Don Mullisbf4735a2006-12-10 02:18:37 -0800282 If unsure, say N.
283
284config HEADERS_CHECK
285 bool "Run 'make headers_check' when building vmlinux"
286 depends on !UML
287 help
288 This option will extract the user-visible kernel headers whenever
289 building the kernel, and will run basic sanity checks on them to
290 ensure that exported files do not attempt to include files which
291 were not exported, etc.
292
293 If you're making modifications to header files which are
294 relevant for userspace, say 'Y', and check the headers
295 exported to $(INSTALL_HDR_PATH) (usually 'usr/include' in
296 your build tree), to make sure they're suitable.
297
Sam Ravnborg91341d42008-01-21 21:31:44 +0100298config DEBUG_SECTION_MISMATCH
299 bool "Enable full Section mismatch analysis"
Sam Ravnborg91341d42008-01-21 21:31:44 +0100300 help
301 The section mismatch analysis checks if there are illegal
302 references from one section to another section.
Michael Wittene809ab02011-04-17 04:08:48 +0000303 During linktime or runtime, some sections are dropped;
304 any use of code/data previously in these sections would
Sam Ravnborg91341d42008-01-21 21:31:44 +0100305 most likely result in an oops.
Michael Wittene809ab02011-04-17 04:08:48 +0000306 In the code, functions and variables are annotated with
Paul Gortmaker0db06282013-06-19 14:53:51 -0400307 __init,, etc. (see the full list in include/linux/init.h),
Geert Uytterhoevend6fbfa42008-01-30 11:13:23 +0100308 which results in the code/data being placed in specific sections.
Michael Wittene809ab02011-04-17 04:08:48 +0000309 The section mismatch analysis is always performed after a full
310 kernel build, and enabling this option causes the following
311 additional steps to occur:
312 - Add the option -fno-inline-functions-called-once to gcc commands.
313 When inlining a function annotated with __init in a non-init
314 function, we would lose the section information and thus
Sam Ravnborg91341d42008-01-21 21:31:44 +0100315 the analysis would not catch the illegal reference.
Michael Wittene809ab02011-04-17 04:08:48 +0000316 This option tells gcc to inline less (but it does result in
317 a larger kernel).
318 - Run the section mismatch analysis for each module/built-in.o file.
319 When we run the section mismatch analysis on vmlinux.o, we
Uwe Kleine-König67797b922016-09-09 10:04:58 +0200320 lose valuable information about where the mismatch was
Sam Ravnborg91341d42008-01-21 21:31:44 +0100321 introduced.
322 Running the analysis for each module/built-in.o file
Michael Wittene809ab02011-04-17 04:08:48 +0000323 tells where the mismatch happens much closer to the
324 source. The drawback is that the same mismatch is
325 reported at least twice.
326 - Enable verbose reporting from modpost in order to help resolve
327 the section mismatches that are reported.
Sam Ravnborg91341d42008-01-21 21:31:44 +0100328
Nicolas Boichat47490ec2015-10-06 09:44:42 +1030329config SECTION_MISMATCH_WARN_ONLY
330 bool "Make section mismatch errors non-fatal"
331 default y
332 help
333 If you say N here, the build process will fail if there are any
334 section mismatch, instead of just throwing warnings.
335
336 If unsure, say Y.
337
Dave Hansen6dfc0662013-07-01 13:04:46 -0700338#
339# Select this config option from the architecture Kconfig, if it
340# is preferred to always offer frame pointers as a config
341# option on the architecture (regardless of KERNEL_DEBUG):
342#
343config ARCH_WANT_FRAME_POINTERS
344 bool
345 help
346
347config FRAME_POINTER
348 bool "Compile the kernel with frame pointers"
349 depends on DEBUG_KERNEL && \
350 (CRIS || M68K || FRV || UML || \
351 AVR32 || SUPERH || BLACKFIN || MN10300 || METAG) || \
352 ARCH_WANT_FRAME_POINTERS
353 default y if (DEBUG_INFO && UML) || ARCH_WANT_FRAME_POINTERS
354 help
355 If you say Y here the resulting kernel image will be slightly
356 larger and slower, but it gives very useful debugging information
357 in case of kernel bugs. (precise oopses/stacktraces/warnings)
358
Josh Poimboeufb9ab5eb2016-02-28 22:22:42 -0600359config STACK_VALIDATION
360 bool "Compile-time stack metadata validation"
361 depends on HAVE_STACK_VALIDATION
362 default n
363 help
364 Add compile-time checks to validate stack metadata, including frame
365 pointers (if CONFIG_FRAME_POINTER is enabled). This helps ensure
366 that runtime stack traces are more reliable.
367
368 For more information, see
369 tools/objtool/Documentation/stack-validation.txt.
370
Dave Hansen6dfc0662013-07-01 13:04:46 -0700371config DEBUG_FORCE_WEAK_PER_CPU
372 bool "Force weak per-cpu definitions"
373 depends on DEBUG_KERNEL
374 help
375 s390 and alpha require percpu variables in modules to be
376 defined weak to work around addressing range issue which
377 puts the following two restrictions on percpu variable
378 definitions.
379
380 1. percpu symbols must be unique whether static or not
381 2. percpu variables can't be defined inside a function
382
383 To ensure that generic code follows the above rules, this
384 option forces all percpu variables to be defined as weak.
385
386endmenu # "Compiler options"
387
388config MAGIC_SYSRQ
389 bool "Magic SysRq key"
390 depends on !UML
391 help
392 If you say Y here, you will have some control over the system even
393 if the system crashes for example during kernel debugging (e.g., you
394 will be able to flush the buffer cache to disk, reboot the system
395 immediately or dump some status information). This is accomplished
396 by pressing various keys while holding SysRq (Alt+PrintScreen). It
397 also works on a serial console (on PC hardware at least), if you
398 send a BREAK and then within 5 seconds a command keypress. The
399 keys are documented in <file:Documentation/sysrq.txt>. Don't say Y
400 unless you really know what this hack does.
401
Ben Hutchings8eaede42013-10-07 01:05:46 +0100402config MAGIC_SYSRQ_DEFAULT_ENABLE
403 hex "Enable magic SysRq key functions by default"
404 depends on MAGIC_SYSRQ
405 default 0x1
406 help
407 Specifies which SysRq key functions are enabled by default.
408 This may be set to 1 or 0 to enable or disable them all, or
409 to a bitmask as described in Documentation/sysrq.txt.
410
Adrian Bunkf346f4b2006-01-09 20:54:51 -0800411config DEBUG_KERNEL
412 bool "Kernel debugging"
413 help
414 Say Y here if you are developing drivers or trying to debug and
415 identify kernel problems.
416
Dave Hansen0610c8a2013-07-01 13:04:43 -0700417menu "Memory Debugging"
David Woodhousea304e1b2007-02-12 00:52:00 -0800418
Dave Hansen0610c8a2013-07-01 13:04:43 -0700419source mm/Kconfig.debug
Ingo Molnar82f67cd2007-02-16 01:28:13 -0800420
Thomas Gleixner3ac7fe52008-04-30 00:55:01 -0700421config DEBUG_OBJECTS
422 bool "Debug object operations"
423 depends on DEBUG_KERNEL
424 help
425 If you say Y here, additional code will be inserted into the
426 kernel to track the life time of various objects and validate
427 the operations on those objects.
428
429config DEBUG_OBJECTS_SELFTEST
430 bool "Debug objects selftest"
431 depends on DEBUG_OBJECTS
432 help
433 This enables the selftest of the object debug code.
434
435config DEBUG_OBJECTS_FREE
436 bool "Debug objects in freed memory"
437 depends on DEBUG_OBJECTS
438 help
439 This enables checks whether a k/v free operation frees an area
440 which contains an object which has not been deactivated
441 properly. This can make kmalloc/kfree-intensive workloads
442 much slower.
443
Thomas Gleixnerc6f3a972008-04-30 00:55:03 -0700444config DEBUG_OBJECTS_TIMERS
445 bool "Debug timer objects"
446 depends on DEBUG_OBJECTS
447 help
448 If you say Y here, additional code will be inserted into the
449 timer routines to track the life time of timer objects and
450 validate the timer operations.
451
Thomas Gleixnerdc186ad2009-11-16 01:09:48 +0900452config DEBUG_OBJECTS_WORK
453 bool "Debug work objects"
454 depends on DEBUG_OBJECTS
455 help
456 If you say Y here, additional code will be inserted into the
457 work queue routines to track the life time of work objects and
458 validate the work operations.
459
Mathieu Desnoyers551d55a2010-04-17 08:48:42 -0400460config DEBUG_OBJECTS_RCU_HEAD
461 bool "Debug RCU callbacks objects"
Mathieu Desnoyersfc2ecf72011-02-23 09:42:14 -0800462 depends on DEBUG_OBJECTS
Mathieu Desnoyers551d55a2010-04-17 08:48:42 -0400463 help
464 Enable this to turn on debugging of RCU list heads (call_rcu() usage).
465
Tejun Heoe2852ae2010-10-26 14:23:05 -0700466config DEBUG_OBJECTS_PERCPU_COUNTER
467 bool "Debug percpu counter objects"
468 depends on DEBUG_OBJECTS
469 help
470 If you say Y here, additional code will be inserted into the
471 percpu counter routines to track the life time of percpu counter
472 objects and validate the percpu counter operations.
473
Ingo Molnar3ae70202008-11-26 10:02:00 +0100474config DEBUG_OBJECTS_ENABLE_DEFAULT
475 int "debug_objects bootup default value (0-1)"
476 range 0 1
477 default "1"
478 depends on DEBUG_OBJECTS
479 help
480 Debug objects boot parameter default value
481
Linus Torvalds1da177e2005-04-16 15:20:36 -0700482config DEBUG_SLAB
Andrew Morton4a2f0ac2006-03-25 03:07:22 -0800483 bool "Debug slab memory allocations"
Vegard Nossum7d46d9e2008-04-04 00:51:41 +0200484 depends on DEBUG_KERNEL && SLAB && !KMEMCHECK
Linus Torvalds1da177e2005-04-16 15:20:36 -0700485 help
486 Say Y here to have the kernel do limited verification on memory
487 allocation as well as poisoning memory on free to catch use of freed
488 memory. This can make kmalloc/kfree-intensive workloads much slower.
489
Al Viro871751e2006-03-25 03:06:39 -0800490config DEBUG_SLAB_LEAK
491 bool "Memory leak debugging"
492 depends on DEBUG_SLAB
493
Christoph Lameterf0630ff2007-07-15 23:38:14 -0700494config SLUB_DEBUG_ON
495 bool "SLUB debugging on by default"
Vegard Nossum7d46d9e2008-04-04 00:51:41 +0200496 depends on SLUB && SLUB_DEBUG && !KMEMCHECK
Christoph Lameterf0630ff2007-07-15 23:38:14 -0700497 default n
498 help
499 Boot with debugging on by default. SLUB boots by default with
500 the runtime debug capabilities switched off. Enabling this is
501 equivalent to specifying the "slub_debug" parameter on boot.
502 There is no support for more fine grained debug control like
503 possible with slub_debug=xxx. SLUB debugging may be switched
504 off in a kernel built with CONFIG_SLUB_DEBUG_ON by specifying
505 "slub_debug=-".
506
Christoph Lameter8ff12cf2008-02-07 17:47:41 -0800507config SLUB_STATS
508 default n
509 bool "Enable SLUB performance statistics"
Christoph Lameterab4d5ed2010-10-05 13:57:26 -0500510 depends on SLUB && SYSFS
Christoph Lameter8ff12cf2008-02-07 17:47:41 -0800511 help
512 SLUB statistics are useful to debug SLUBs allocation behavior in
513 order find ways to optimize the allocator. This should never be
514 enabled for production use since keeping statistics slows down
515 the allocator by a few percentage points. The slabinfo command
516 supports the determination of the most active slabs to figure
517 out which slabs are relevant to a particular load.
518 Try running: slabinfo -DA
519
Catalin Marinasb69ec422012-10-08 16:28:11 -0700520config HAVE_DEBUG_KMEMLEAK
521 bool
522
Catalin Marinas3bba00d2009-06-11 13:24:13 +0100523config DEBUG_KMEMLEAK
524 bool "Kernel memory leak detector"
Kees Cook525c1f92013-01-16 18:54:16 -0800525 depends on DEBUG_KERNEL && HAVE_DEBUG_KMEMLEAK
Catalin Marinas79e0d9b2011-04-27 17:06:19 +0100526 select DEBUG_FS
Catalin Marinas3bba00d2009-06-11 13:24:13 +0100527 select STACKTRACE if STACKTRACE_SUPPORT
528 select KALLSYMS
Randy Dunlapb60e26a2009-11-06 15:33:45 -0800529 select CRC32
Catalin Marinas3bba00d2009-06-11 13:24:13 +0100530 help
531 Say Y here if you want to enable the memory leak
532 detector. The memory allocation/freeing is traced in a way
533 similar to the Boehm's conservative garbage collector, the
534 difference being that the orphan objects are not freed but
535 only shown in /sys/kernel/debug/kmemleak. Enabling this
536 feature will introduce an overhead to memory
537 allocations. See Documentation/kmemleak.txt for more
538 details.
539
Catalin Marinasbf96d1e2009-06-23 14:40:27 +0100540 Enabling DEBUG_SLAB or SLUB_DEBUG may increase the chances
541 of finding leaks due to the slab objects poisoning.
542
Catalin Marinas3bba00d2009-06-11 13:24:13 +0100543 In order to access the kmemleak file, debugfs needs to be
544 mounted (usually at /sys/kernel/debug).
545
Catalin Marinasa9d90582009-06-25 10:16:11 +0100546config DEBUG_KMEMLEAK_EARLY_LOG_SIZE
547 int "Maximum kmemleak early log entries"
548 depends on DEBUG_KMEMLEAK
Heiko Carstensdfcc3e62009-10-06 10:33:57 +0200549 range 200 40000
Catalin Marinasa9d90582009-06-25 10:16:11 +0100550 default 400
551 help
552 Kmemleak must track all the memory allocations to avoid
553 reporting false positives. Since memory may be allocated or
554 freed before kmemleak is initialised, an early log buffer is
555 used to store these actions. If kmemleak reports "early log
556 buffer exceeded", please increase this value.
557
Catalin Marinas0822ee42009-06-11 13:24:14 +0100558config DEBUG_KMEMLEAK_TEST
559 tristate "Simple test for the kernel memory leak detector"
Daniel Baluta97182692011-04-04 15:06:44 -0700560 depends on DEBUG_KMEMLEAK && m
Catalin Marinas0822ee42009-06-11 13:24:14 +0100561 help
Daniel Baluta97182692011-04-04 15:06:44 -0700562 This option enables a module that explicitly leaks memory.
Catalin Marinas0822ee42009-06-11 13:24:14 +0100563
564 If unsure, say N.
565
Jason Baronab0155a2010-07-19 11:54:17 +0100566config DEBUG_KMEMLEAK_DEFAULT_OFF
567 bool "Default kmemleak to off"
568 depends on DEBUG_KMEMLEAK
569 help
570 Say Y here to disable kmemleak by default. It can then be enabled
571 on the command line via kmemleak=on.
572
Dave Hansen0610c8a2013-07-01 13:04:43 -0700573config DEBUG_STACK_USAGE
574 bool "Stack utilization instrumentation"
Helge Deller6c31da32016-03-19 17:54:10 +0100575 depends on DEBUG_KERNEL && !IA64
Dave Hansen0610c8a2013-07-01 13:04:43 -0700576 help
577 Enables the display of the minimum amount of free stack which each
578 task has ever had available in the sysrq-T and sysrq-P debug output.
579
580 This option will slow down process creation somewhat.
581
582config DEBUG_VM
583 bool "Debug VM"
584 depends on DEBUG_KERNEL
585 help
586 Enable this to turn on extended checks in the virtual-memory system
587 that may impact performance.
588
589 If unsure, say N.
590
Davidlohr Bueso4f115142014-06-04 16:06:46 -0700591config DEBUG_VM_VMACACHE
592 bool "Debug VMA caching"
593 depends on DEBUG_VM
594 help
595 Enable this to turn on VMA caching debug information. Doing so
596 can cause significant overhead, so only enable it in non-production
597 environments.
598
599 If unsure, say N.
600
Dave Hansen0610c8a2013-07-01 13:04:43 -0700601config DEBUG_VM_RB
602 bool "Debug VM red-black trees"
603 depends on DEBUG_VM
604 help
Davidlohr Buesoa663dad2014-04-18 15:07:22 -0700605 Enable VM red-black tree debugging information and extra validations.
Dave Hansen0610c8a2013-07-01 13:04:43 -0700606
607 If unsure, say N.
608
Kirill A. Shutemov95ad9752016-01-15 16:51:21 -0800609config DEBUG_VM_PGFLAGS
610 bool "Debug page-flags operations"
611 depends on DEBUG_VM
612 help
613 Enables extra validation on page flags operations.
614
615 If unsure, say N.
616
Dave Hansen0610c8a2013-07-01 13:04:43 -0700617config DEBUG_VIRTUAL
618 bool "Debug VM translations"
619 depends on DEBUG_KERNEL && X86
620 help
621 Enable some costly sanity checks in virtual to page code. This can
622 catch mistakes with virt_to_page() and friends.
623
624 If unsure, say N.
625
626config DEBUG_NOMMU_REGIONS
627 bool "Debug the global anon/private NOMMU mapping region tree"
628 depends on DEBUG_KERNEL && !MMU
629 help
630 This option causes the global tree of anonymous and private mapping
631 regions to be regularly checked for invalid topology.
632
633config DEBUG_MEMORY_INIT
634 bool "Debug memory initialisation" if EXPERT
635 default !EXPERT
636 help
637 Enable this for additional checks during memory initialisation.
638 The sanity checks verify aspects of the VM such as the memory model
639 and other information provided by the architecture. Verbose
640 information will be printed at KERN_DEBUG loglevel depending
641 on the mminit_loglevel= command-line option.
642
643 If unsure, say Y
644
645config MEMORY_NOTIFIER_ERROR_INJECT
646 tristate "Memory hotplug notifier error injection module"
647 depends on MEMORY_HOTPLUG_SPARSE && NOTIFIER_ERROR_INJECTION
648 help
649 This option provides the ability to inject artificial errors to
650 memory hotplug notifier chain callbacks. It is controlled through
651 debugfs interface under /sys/kernel/debug/notifier-error-inject/memory
652
653 If the notifier call chain should be failed with some events
654 notified, write the error code to "actions/<notifier event>/error".
655
656 Example: Inject memory hotplug offline error (-12 == -ENOMEM)
657
658 # cd /sys/kernel/debug/notifier-error-inject/memory
659 # echo -12 > actions/MEM_GOING_OFFLINE/error
660 # echo offline > /sys/devices/system/memory/memoryXXX/state
661 bash: echo: write error: Cannot allocate memory
662
663 To compile this code as a module, choose M here: the module will
664 be called memory-notifier-error-inject.
665
666 If unsure, say N.
667
668config DEBUG_PER_CPU_MAPS
669 bool "Debug access to per_cpu maps"
670 depends on DEBUG_KERNEL
671 depends on SMP
672 help
673 Say Y to verify that the per_cpu map being accessed has
674 been set up. This adds a fair amount of code to kernel memory
675 and decreases performance.
676
677 Say N if unsure.
678
679config DEBUG_HIGHMEM
680 bool "Highmem debugging"
681 depends on DEBUG_KERNEL && HIGHMEM
682 help
Geert Uytterhoevenb1357c92014-04-14 18:55:50 +0200683 This option enables additional error checking for high memory
684 systems. Disable for production systems.
Dave Hansen0610c8a2013-07-01 13:04:43 -0700685
686config HAVE_DEBUG_STACKOVERFLOW
687 bool
688
689config DEBUG_STACKOVERFLOW
690 bool "Check for stack overflows"
691 depends on DEBUG_KERNEL && HAVE_DEBUG_STACKOVERFLOW
692 ---help---
693 Say Y here if you want to check for overflows of kernel, IRQ
Borislav Petkovedb0ec02015-01-25 19:50:34 +0100694 and exception stacks (if your architecture uses them). This
Dave Hansen0610c8a2013-07-01 13:04:43 -0700695 option will show detailed messages if free stack space drops
696 below a certain limit.
697
698 These kinds of bugs usually occur when call-chains in the
699 kernel get too deep, especially when interrupts are
700 involved.
701
702 Use this in cases where you see apparently random memory
703 corruption, especially if it appears in 'struct thread_info'
704
705 If in doubt, say "N".
706
707source "lib/Kconfig.kmemcheck"
708
Andrey Ryabinin0b24bec2015-02-13 14:39:17 -0800709source "lib/Kconfig.kasan"
710
Dave Hansen0610c8a2013-07-01 13:04:43 -0700711endmenu # "Memory Debugging"
712
Dmitry Vyukov5c9a8752016-03-22 14:27:30 -0700713config ARCH_HAS_KCOV
714 bool
715 help
716 KCOV does not have any arch-specific code, but currently it is enabled
717 only for x86_64. KCOV requires testing on other archs, and most likely
718 disabling of instrumentation for some early boot code.
719
720config KCOV
721 bool "Code coverage for fuzzing"
722 depends on ARCH_HAS_KCOV
723 select DEBUG_FS
Kees Cooka519167e2016-06-11 09:09:28 -0700724 select GCC_PLUGINS if !COMPILE_TEST
725 select GCC_PLUGIN_SANCOV if !COMPILE_TEST
Dmitry Vyukov5c9a8752016-03-22 14:27:30 -0700726 help
727 KCOV exposes kernel code coverage information in a form suitable
728 for coverage-guided fuzzing (randomized testing).
729
730 If RANDOMIZE_BASE is enabled, PC values will not be stable across
731 different machines and across reboots. If you need stable PC values,
732 disable RANDOMIZE_BASE.
733
734 For more details, see Documentation/kcov.txt.
735
Vegard Nossuma4691de2016-08-02 14:07:30 -0700736config KCOV_INSTRUMENT_ALL
737 bool "Instrument all code by default"
738 depends on KCOV
739 default y if KCOV
740 help
741 If you are doing generic system call fuzzing (like e.g. syzkaller),
742 then you will want to instrument the whole kernel and you should
743 say y here. If you are doing more targeted fuzzing (like e.g.
744 filesystem fuzzing with AFL) then you will want to enable coverage
745 for more specific subsets of files, and should say n here.
746
Linus Torvalds1da177e2005-04-16 15:20:36 -0700747config DEBUG_SHIRQ
748 bool "Debug shared IRQ handlers"
Martin Schwidefsky0244ad02013-08-30 09:39:53 +0200749 depends on DEBUG_KERNEL
Linus Torvalds1da177e2005-04-16 15:20:36 -0700750 help
751 Enable this to generate a spurious interrupt as soon as a shared
752 interrupt handler is registered, and just before one is deregistered.
753 Drivers ought to be able to handle interrupts coming in at those
754 points; some don't and need to be caught.
755
Dave Hansen92aef8fb2013-07-01 13:04:50 -0700756menu "Debug Lockups and Hangs"
757
Linus Torvalds1da177e2005-04-16 15:20:36 -0700758config LOCKUP_DETECTOR
759 bool "Detect Hard and Soft Lockups"
760 depends on DEBUG_KERNEL && !S390
761 help
762 Say Y here to enable the kernel to act as a watchdog to detect
763 hard and soft lockups.
764
765 Softlockups are bugs that cause the kernel to loop in kernel
766 mode for more than 20 seconds, without giving other tasks a
767 chance to run. The current stack trace is displayed upon
768 detection and the system will stay locked up.
769
770 Hardlockups are bugs that cause the CPU to loop in kernel mode
771 for more than 10 seconds, without letting other interrupts have a
772 chance to run. The current stack trace is displayed upon detection
773 and the system will stay locked up.
774
775 The overhead should be minimal. A periodic hrtimer runs to
776 generate interrupts and kick the watchdog task every 4 seconds.
777 An NMI is generated every 10 seconds or so to check for hardlockups.
Colin Cross49ed1f82013-01-11 13:51:48 -0800778 If NMIs are not available on the platform, every 12 seconds the
779 hrtimer interrupt on one cpu will be used to check for hardlockups
780 on the next cpu.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700781
782 The frequency of hrtimer and NMI events and the soft and hard lockup
783 thresholds can be controlled through the sysctl watchdog_thresh.
784
Colin Cross49ed1f82013-01-11 13:51:48 -0800785config HARDLOCKUP_DETECTOR_NMI
Linus Torvalds1da177e2005-04-16 15:20:36 -0700786 def_bool y
787 depends on LOCKUP_DETECTOR && !HAVE_NMI_WATCHDOG
788 depends on PERF_EVENTS && HAVE_PERF_EVENTS_NMI
789
Colin Cross49ed1f82013-01-11 13:51:48 -0800790config HARDLOCKUP_DETECTOR_OTHER_CPU
791 def_bool y
792 depends on LOCKUP_DETECTOR && SMP
793 depends on !HARDLOCKUP_DETECTOR_NMI && !HAVE_NMI_WATCHDOG
794
795config HARDLOCKUP_DETECTOR
796 def_bool y
797 depends on HARDLOCKUP_DETECTOR_NMI || HARDLOCKUP_DETECTOR_OTHER_CPU
798
Linus Torvalds1da177e2005-04-16 15:20:36 -0700799config BOOTPARAM_HARDLOCKUP_PANIC
800 bool "Panic (Reboot) On Hard Lockups"
801 depends on HARDLOCKUP_DETECTOR
802 help
803 Say Y here to enable the kernel to panic on "hard lockups",
804 which are bugs that cause the kernel to loop in kernel
805 mode with interrupts disabled for more than 10 seconds (configurable
806 using the watchdog_thresh sysctl).
807
808 Say N if unsure.
809
810config BOOTPARAM_HARDLOCKUP_PANIC_VALUE
811 int
812 depends on HARDLOCKUP_DETECTOR
813 range 0 1
814 default 0 if !BOOTPARAM_HARDLOCKUP_PANIC
815 default 1 if BOOTPARAM_HARDLOCKUP_PANIC
816
817config BOOTPARAM_SOFTLOCKUP_PANIC
818 bool "Panic (Reboot) On Soft Lockups"
819 depends on LOCKUP_DETECTOR
820 help
821 Say Y here to enable the kernel to panic on "soft lockups",
822 which are bugs that cause the kernel to loop in kernel
823 mode for more than 20 seconds (configurable using the watchdog_thresh
824 sysctl), without giving other tasks a chance to run.
Ingo Molnar8637c092006-07-03 00:24:38 -0700825
Linus Torvalds1da177e2005-04-16 15:20:36 -0700826 The panic can be used in combination with panic_timeout,
827 to cause the system to reboot automatically after a
828 lockup has been detected. This feature is useful for
829 high-availability systems that have uptime guarantees and
830 where a lockup must be resolved ASAP.
831
832 Say N if unsure.
Ingo Molnare7eebaf2006-06-27 02:54:55 -0700833
834config BOOTPARAM_SOFTLOCKUP_PANIC_VALUE
Ingo Molnare7eebaf2006-06-27 02:54:55 -0700835 int
836 depends on LOCKUP_DETECTOR
837 range 0 1
838 default 0 if !BOOTPARAM_SOFTLOCKUP_PANIC
839 default 1 if BOOTPARAM_SOFTLOCKUP_PANIC
840
Linus Torvalds1da177e2005-04-16 15:20:36 -0700841config DETECT_HUNG_TASK
842 bool "Detect Hung Tasks"
843 depends on DEBUG_KERNEL
844 default LOCKUP_DETECTOR
845 help
846 Say Y here to enable the kernel to detect "hung tasks",
847 which are bugs that cause the task to be stuck in
Vivien Didelot96b03ab2016-09-22 16:55:13 -0400848 uninterruptible "D" state indefinitely.
Linus Torvalds1da177e2005-04-16 15:20:36 -0700849
850 When a hung task is detected, the kernel will print the
851 current stack trace (which you should report), but the
852 task will stay in uninterruptible state. If lockdep is
853 enabled then all held locks will also be reported. This
854 feature has negligible overhead.
855
856config DEFAULT_HUNG_TASK_TIMEOUT
857 int "Default timeout for hung task detection (in seconds)"
858 depends on DETECT_HUNG_TASK
859 default 120
860 help
861 This option controls the default timeout (in seconds) used
862 to determine when a task has become non-responsive and should
863 be considered hung.
Ingo Molnar408894e2006-01-09 15:59:20 -0800864
865 It can be adjusted at runtime via the kernel.hung_task_timeout_secs
866 sysctl or by writing a value to
867 /proc/sys/kernel/hung_task_timeout_secs.
868
869 A timeout of 0 disables the check. The default is two minutes.
Tim Chencca57c52006-04-18 22:21:47 -0700870 Keeping the default should be fine in most cases.
Ingo Molnar408894e2006-01-09 15:59:20 -0800871
872config BOOTPARAM_HUNG_TASK_PANIC
873 bool "Panic (Reboot) On Hung Tasks"
874 depends on DETECT_HUNG_TASK
875 help
Ingo Molnare7eebaf2006-06-27 02:54:55 -0700876 Say Y here to enable the kernel to panic on "hung tasks",
877 which are bugs that cause the kernel to leave a task stuck
878 in uninterruptible "D" state.
879
880 The panic can be used in combination with panic_timeout,
881 to cause the system to reboot automatically after a
882 hung task has been detected. This feature is useful for
883 high-availability systems that have uptime guarantees and
884 where a hung tasks must be resolved ASAP.
885
886 Say N if unsure.
887
Hugh Dickins048c8bc2006-11-01 05:44:54 +1100888config BOOTPARAM_HUNG_TASK_PANIC_VALUE
Linus Torvalds1da177e2005-04-16 15:20:36 -0700889 int
890 depends on DETECT_HUNG_TASK
891 range 0 1
892 default 0 if !BOOTPARAM_HUNG_TASK_PANIC
893 default 1 if BOOTPARAM_HUNG_TASK_PANIC
894
Tejun Heo82607adc2015-12-08 11:28:04 -0500895config WQ_WATCHDOG
896 bool "Detect Workqueue Stalls"
897 depends on DEBUG_KERNEL
898 help
899 Say Y here to enable stall detection on workqueues. If a
900 worker pool doesn't make forward progress on a pending work
901 item for over a given amount of time, 30s by default, a
902 warning message is printed along with dump of workqueue
903 state. This can be configured through kernel parameter
904 "workqueue.watchdog_thresh" and its sysfs counterpart.
905
Dave Hansen92aef8fb2013-07-01 13:04:50 -0700906endmenu # "Debug lockups and hangs"
907
908config PANIC_ON_OOPS
909 bool "Panic on Oops"
910 help
911 Say Y here to enable the kernel to panic when it oopses. This
912 has the same effect as setting oops=panic on the kernel command
913 line.
914
915 This feature is useful to ensure that the kernel does not do
916 anything erroneous after an oops which could result in data
917 corruption or other issues.
918
919 Say N if unsure.
920
921config PANIC_ON_OOPS_VALUE
922 int
923 range 0 1
924 default 0 if !PANIC_ON_OOPS
925 default 1 if PANIC_ON_OOPS
926
Jason Baron5800dc32013-11-25 23:23:04 +0000927config PANIC_TIMEOUT
928 int "panic timeout"
929 default 0
930 help
931 Set the timeout value (in seconds) until a reboot occurs when the
932 the kernel panics. If n = 0, then we wait forever. A timeout
933 value n > 0 will wait n seconds before rebooting, while a timeout
934 value n < 0 will reboot immediately.
935
Linus Torvalds1da177e2005-04-16 15:20:36 -0700936config SCHED_DEBUG
937 bool "Collect scheduler debugging info"
938 depends on DEBUG_KERNEL && PROC_FS
939 default y
940 help
941 If you say Y here, the /proc/sched_debug file will be provided
942 that can help debug the scheduler. The runtime overhead of this
943 option is minimal.
944
Naveen N. Raof6db8342015-06-25 23:53:37 +0530945config SCHED_INFO
946 bool
947 default n
948
Syed Rameez Mustafadddcab72016-09-07 16:18:27 -0700949config PANIC_ON_SCHED_BUG
950 bool "Panic on all bugs encountered by the scheduler"
951 help
952 Say Y here to panic on all 'BUG:' conditions encountered by the
953 scheduler, even potentially-recoverable ones such as scheduling
954 while atomic, sleeping from invalid context, and detection of
955 broken arch topologies.
956
957 Say N if unsure.
958
959config PANIC_ON_RT_THROTTLING
960 bool "Panic on RT throttling"
961 help
962 Say Y here to enable the kernel to panic when a realtime
963 runqueue is throttled. This may be useful for detecting
964 and debugging RT throttling issues.
965
966 Say N if unsure.
967
Linus Torvalds1da177e2005-04-16 15:20:36 -0700968config SCHEDSTATS
969 bool "Collect scheduler statistics"
970 depends on DEBUG_KERNEL && PROC_FS
Naveen N. Raof6db8342015-06-25 23:53:37 +0530971 select SCHED_INFO
Linus Torvalds1da177e2005-04-16 15:20:36 -0700972 help
973 If you say Y here, additional code will be inserted into the
974 scheduler and related routines to collect statistics about
975 scheduler behavior and provide them in /proc/schedstat. These
976 stats may be useful for both tuning and debugging the scheduler
977 If you aren't debugging the scheduler or trying to tune a specific
978 application, you can say N to avoid the very slight overhead
979 this adds.
980
Aaron Tomlin0d9e2632014-09-12 14:16:19 +0100981config SCHED_STACK_END_CHECK
982 bool "Detect stack corruption on calls to schedule()"
983 depends on DEBUG_KERNEL
984 default n
985 help
986 This option checks for a stack overrun on calls to schedule().
987 If the stack end location is found to be over written always panic as
988 the content of the corrupted region can no longer be trusted.
989 This is to ensure no erroneous behaviour occurs which could result in
990 data corruption or a sporadic crash at a later stage once the region
991 is examined. The runtime overhead introduced is minimal.
992
John Stultz3c17ad12015-03-11 21:16:32 -0700993config DEBUG_TIMEKEEPING
994 bool "Enable extra timekeeping sanity checking"
995 help
996 This option will enable additional timekeeping sanity checks
997 which may be helpful when diagnosing issues where timekeeping
998 problems are suspected.
999
1000 This may include checks in the timekeeping hotpaths, so this
1001 option may have a (very small) performance impact to some
1002 workloads.
1003
1004 If unsure, say N.
1005
Vignesh Radhakrishnand1fdabf2015-01-22 11:43:45 +05301006config DEBUG_TASK_STACK_SCAN_OFF
1007 bool "Disable kmemleak task stack scan by default"
1008 depends on DEBUG_KMEMLEAK
1009 help
1010 Say Y here to disable kmemleak task stack scan by default
1011 at compile time. It can be enabled later if required by
1012 writing to the debugfs entry :
1013 echo "stack=on" > /sys/kernel/debug/kmemleak.
1014
Linus Torvalds1da177e2005-04-16 15:20:36 -07001015config DEBUG_PREEMPT
1016 bool "Debug preemptible kernel"
Kumar Gala01deab92009-10-16 07:21:39 +00001017 depends on DEBUG_KERNEL && PREEMPT && TRACE_IRQFLAGS_SUPPORT
Linus Torvalds1da177e2005-04-16 15:20:36 -07001018 default y
1019 help
1020 If you say Y here then the kernel will use a debug variant of the
1021 commonly used smp_processor_id() function and will print warnings
1022 if kernel code uses it in a preemption-unsafe way. Also, the kernel
1023 will detect preemption count underflows.
1024
Dave Hansen9eade162013-07-01 13:04:47 -07001025menu "Lock Debugging (spinlocks, mutexes, etc...)"
1026
Ingo Molnare7eebaf2006-06-27 02:54:55 -07001027config DEBUG_RT_MUTEXES
1028 bool "RT Mutex debugging, deadlock detection"
1029 depends on DEBUG_KERNEL && RT_MUTEXES
1030 help
1031 This allows rt mutex semantics violations and rt mutex related
1032 deadlocks (lockups) to be detected and reported automatically.
1033
Linus Torvalds1da177e2005-04-16 15:20:36 -07001034config DEBUG_SPINLOCK
Ingo Molnar4d9f34a2006-07-03 00:24:55 -07001035 bool "Spinlock and rw-lock debugging: basic checks"
Linus Torvalds1da177e2005-04-16 15:20:36 -07001036 depends on DEBUG_KERNEL
Raghavendra K Te335e3e2012-03-22 15:25:08 +05301037 select UNINLINE_SPIN_UNLOCK
Linus Torvalds1da177e2005-04-16 15:20:36 -07001038 help
1039 Say Y here and build SMP to catch missing spinlock initialization
1040 and certain other kinds of spinlock errors commonly made. This is
1041 best used in conjunction with the NMI watchdog so that spinlock
1042 deadlocks are also debuggable.
Rohit Vaswani173de7d2014-10-02 21:02:56 -07001043choice
1044 prompt "Perform Action on spinlock bug"
1045 depends on DEBUG_SPINLOCK
1046
1047 default DEBUG_SPINLOCK_BITE_ON_BUG
1048
1049 config DEBUG_SPINLOCK_BITE_ON_BUG
1050 bool "Cause a Watchdog Bite on Spinlock bug"
1051 depends on QCOM_WATCHDOG_V2
1052 help
1053 On a spinlock bug, cause a watchdog bite so that we can get
1054 the precise state of the system captured at the time of spin
1055 dump. This is mutually exclusive with the below
1056 DEBUG_SPINLOCK_PANIC_ON_BUG config.
1057
1058 config DEBUG_SPINLOCK_PANIC_ON_BUG
1059 bool "Cause a Kernel Panic on Spinlock bug"
1060 help
1061 On a spinlock bug, cause a kernel panic so that we can get the complete
1062 information about the system at the time of spin dump in the dmesg.
1063 This is mutually exclusive with the above DEBUG_SPINLOCK_BITE_ON_BUG.
1064endchoice
Linus Torvalds1da177e2005-04-16 15:20:36 -07001065
Ingo Molnar4d9f34a2006-07-03 00:24:55 -07001066config DEBUG_MUTEXES
1067 bool "Mutex debugging: basic checks"
1068 depends on DEBUG_KERNEL
1069 help
1070 This feature allows mutex semantics violations to be detected and
1071 reported.
1072
Daniel Vetter23010022013-06-20 13:31:17 +02001073config DEBUG_WW_MUTEX_SLOWPATH
1074 bool "Wait/wound mutex debugging: Slowpath testing"
1075 depends on DEBUG_KERNEL && TRACE_IRQFLAGS_SUPPORT && STACKTRACE_SUPPORT && LOCKDEP_SUPPORT
1076 select DEBUG_LOCK_ALLOC
1077 select DEBUG_SPINLOCK
1078 select DEBUG_MUTEXES
1079 help
1080 This feature enables slowpath testing for w/w mutex users by
1081 injecting additional -EDEADLK wound/backoff cases. Together with
1082 the full mutex checks enabled with (CONFIG_PROVE_LOCKING) this
1083 will test all possible w/w mutex interface abuse with the
1084 exception of simply not acquiring all the required locks.
Rob Clark4d692372014-08-27 11:19:26 -04001085 Note that this feature can introduce significant overhead, so
1086 it really should not be enabled in a production or distro kernel,
1087 even a debug kernel. If you are a driver writer, enable it. If
1088 you are a distro, do not.
Daniel Vetter23010022013-06-20 13:31:17 +02001089
Ingo Molnar4d9f34a2006-07-03 00:24:55 -07001090config DEBUG_LOCK_ALLOC
1091 bool "Lock debugging: detect incorrect freeing of live locks"
Adrian Bunk517e7aa2006-07-14 00:24:32 -07001092 depends on DEBUG_KERNEL && TRACE_IRQFLAGS_SUPPORT && STACKTRACE_SUPPORT && LOCKDEP_SUPPORT
Ingo Molnar4d9f34a2006-07-03 00:24:55 -07001093 select DEBUG_SPINLOCK
1094 select DEBUG_MUTEXES
Ingo Molnar4d9f34a2006-07-03 00:24:55 -07001095 select LOCKDEP
1096 help
1097 This feature will check whether any held lock (spinlock, rwlock,
1098 mutex or rwsem) is incorrectly freed by the kernel, via any of the
1099 memory-freeing routines (kfree(), kmem_cache_free(), free_pages(),
1100 vfree(), etc.), whether a live lock is incorrectly reinitialized via
1101 spin_lock_init()/mutex_init()/etc., or whether there is any lock
1102 held during task exit.
1103
1104config PROVE_LOCKING
1105 bool "Lock debugging: prove locking correctness"
Adrian Bunk517e7aa2006-07-14 00:24:32 -07001106 depends on DEBUG_KERNEL && TRACE_IRQFLAGS_SUPPORT && STACKTRACE_SUPPORT && LOCKDEP_SUPPORT
Ingo Molnar4d9f34a2006-07-03 00:24:55 -07001107 select LOCKDEP
1108 select DEBUG_SPINLOCK
1109 select DEBUG_MUTEXES
Ingo Molnar4d9f34a2006-07-03 00:24:55 -07001110 select DEBUG_LOCK_ALLOC
Steven Rostedt46b93b72010-08-31 16:35:20 -04001111 select TRACE_IRQFLAGS
Ingo Molnar4d9f34a2006-07-03 00:24:55 -07001112 default n
1113 help
1114 This feature enables the kernel to prove that all locking
1115 that occurs in the kernel runtime is mathematically
1116 correct: that under no circumstance could an arbitrary (and
1117 not yet triggered) combination of observed locking
1118 sequences (on an arbitrary number of CPUs, running an
1119 arbitrary number of tasks and interrupt contexts) cause a
1120 deadlock.
1121
1122 In short, this feature enables the kernel to report locking
1123 related deadlocks before they actually occur.
1124
1125 The proof does not depend on how hard and complex a
1126 deadlock scenario would be to trigger: how many
1127 participant CPUs, tasks and irq-contexts would be needed
1128 for it to trigger. The proof also does not depend on
1129 timing: if a race and a resulting deadlock is possible
1130 theoretically (no matter how unlikely the race scenario
1131 is), it will be proven so and will immediately be
1132 reported by the kernel (once the event is observed that
1133 makes the deadlock theoretically possible).
1134
1135 If a deadlock is impossible (i.e. the locking rules, as
1136 observed by the kernel, are mathematically correct), the
1137 kernel reports nothing.
1138
1139 NOTE: this feature can also be enabled for rwlocks, mutexes
1140 and rwsems - in which case all dependencies between these
1141 different locking variants are observed and mapped too, and
1142 the proof of observed correctness is also maintained for an
1143 arbitrary combination of these separate locking variants.
1144
Davidlohr Bueso214e0ae2014-07-30 13:41:55 -07001145 For more details, see Documentation/locking/lockdep-design.txt.
Ingo Molnar4d9f34a2006-07-03 00:24:55 -07001146
Babu Mogere6b5f1b2016-11-02 09:36:32 -07001147config PROVE_LOCKING_SMALL
1148 bool
1149
Ingo Molnar4d9f34a2006-07-03 00:24:55 -07001150config LOCKDEP
1151 bool
Adrian Bunk517e7aa2006-07-14 00:24:32 -07001152 depends on DEBUG_KERNEL && TRACE_IRQFLAGS_SUPPORT && STACKTRACE_SUPPORT && LOCKDEP_SUPPORT
Ingo Molnar4d9f34a2006-07-03 00:24:55 -07001153 select STACKTRACE
Chen Gangdf2e1ef2014-06-23 13:22:04 -07001154 select FRAME_POINTER if !MIPS && !PPC && !ARM_UNWIND && !S390 && !MICROBLAZE && !ARC && !SCORE
Ingo Molnar4d9f34a2006-07-03 00:24:55 -07001155 select KALLSYMS
1156 select KALLSYMS_ALL
1157
Peter Zijlstraf20786f2007-07-19 01:48:56 -07001158config LOCK_STAT
Danny ter Haarfdfb8702007-09-24 21:24:43 -07001159 bool "Lock usage statistics"
Peter Zijlstraf20786f2007-07-19 01:48:56 -07001160 depends on DEBUG_KERNEL && TRACE_IRQFLAGS_SUPPORT && STACKTRACE_SUPPORT && LOCKDEP_SUPPORT
1161 select LOCKDEP
1162 select DEBUG_SPINLOCK
1163 select DEBUG_MUTEXES
1164 select DEBUG_LOCK_ALLOC
1165 default n
1166 help
1167 This feature enables tracking lock contention points
1168
Davidlohr Bueso214e0ae2014-07-30 13:41:55 -07001169 For more details, see Documentation/locking/lockstat.txt
Peter Zijlstraa560aa42007-10-07 00:24:33 -07001170
Frederic Weisbeckerdd8b1cf2010-02-27 17:10:39 +01001171 This also enables lock events required by "perf lock",
1172 subcommand of perf.
1173 If you want to use "perf lock", you also need to turn on
1174 CONFIG_EVENT_TRACING.
Hitoshi Mitake84c6f882010-02-04 16:08:15 +09001175
1176 CONFIG_LOCK_STAT defines "contended" and "acquired" lock events.
Frederic Weisbeckerdd8b1cf2010-02-27 17:10:39 +01001177 (CONFIG_LOCKDEP defines "acquire" and "release" events.)
Hitoshi Mitake84c6f882010-02-04 16:08:15 +09001178
Ingo Molnar4d9f34a2006-07-03 00:24:55 -07001179config DEBUG_LOCKDEP
1180 bool "Lock dependency engine debugging"
Adrian Bunk517e7aa2006-07-14 00:24:32 -07001181 depends on DEBUG_KERNEL && LOCKDEP
Ingo Molnar4d9f34a2006-07-03 00:24:55 -07001182 help
1183 If you say Y here, the lock dependency engine will do
1184 additional runtime checks to debug itself, at the price
1185 of more runtime overhead.
1186
Frederic Weisbeckerd902db12011-06-08 19:31:56 +02001187config DEBUG_ATOMIC_SLEEP
1188 bool "Sleep inside atomic section checking"
Frederic Weisbeckere8f7c702011-06-08 01:51:02 +02001189 select PREEMPT_COUNT
Linus Torvalds1da177e2005-04-16 15:20:36 -07001190 depends on DEBUG_KERNEL
1191 help
1192 If you say Y here, various routines which may sleep will become very
Frederic Weisbeckerd902db12011-06-08 19:31:56 +02001193 noisy if they are called inside atomic sections: when a spinlock is
1194 held, inside an rcu read side critical section, inside preempt disabled
1195 sections, inside an interrupt, etc...
Linus Torvalds1da177e2005-04-16 15:20:36 -07001196
Ingo Molnarcae2ed92006-07-03 00:24:48 -07001197config DEBUG_LOCKING_API_SELFTESTS
1198 bool "Locking API boot-time self-tests"
1199 depends on DEBUG_KERNEL
1200 help
1201 Say Y here if you want the kernel to run a short self-test during
1202 bootup. The self-test checks whether common types of locking bugs
1203 are detected by debugging mechanisms or not. (if you disable
1204 lock debugging then those bugs wont be detected of course.)
1205 The following locking APIs are covered: spinlocks, rwlocks,
1206 mutexes and rwsems.
1207
Paul E. McKenney0af3fe12014-02-04 15:51:41 -08001208config LOCK_TORTURE_TEST
1209 tristate "torture tests for locking"
1210 depends on DEBUG_KERNEL
1211 select TORTURE_TEST
1212 default n
1213 help
1214 This option provides a kernel module that runs torture tests
1215 on kernel locking primitives. The kernel module may be built
1216 after the fact on the running kernel to be tested, if desired.
1217
1218 Say Y here if you want kernel locking-primitive torture tests
1219 to be built into the kernel.
1220 Say M if you want these torture tests to build as a module.
1221 Say N if you are unsure.
1222
Dave Hansen9eade162013-07-01 13:04:47 -07001223endmenu # lock debugging
1224
1225config TRACE_IRQFLAGS
1226 bool
1227 help
1228 Enables hooks to interrupt enabling and disabling for
1229 either tracing or lock debugging.
1230
Ingo Molnar8637c092006-07-03 00:24:38 -07001231config STACKTRACE
Dave Jones0c38e1f2014-08-29 15:18:35 -07001232 bool "Stack backtrace support"
Ingo Molnar8637c092006-07-03 00:24:38 -07001233 depends on STACKTRACE_SUPPORT
Dave Jones0c38e1f2014-08-29 15:18:35 -07001234 help
1235 This option causes the kernel to create a /proc/pid/stack for
1236 every process, showing its current stack trace.
1237 It is also used by various kernel debugging features that require
1238 stack trace generation.
Ingo Molnar8637c092006-07-03 00:24:38 -07001239
Linus Torvalds1da177e2005-04-16 15:20:36 -07001240config DEBUG_KOBJECT
1241 bool "kobject debugging"
1242 depends on DEBUG_KERNEL
1243 help
1244 If you say Y here, some extra kobject debugging messages will be sent
Rohit Vaswani173de7d2014-10-02 21:02:56 -07001245 to the syslog.
Linus Torvalds1da177e2005-04-16 15:20:36 -07001246
Russell Kingc817a672013-06-27 15:06:14 +01001247config DEBUG_KOBJECT_RELEASE
1248 bool "kobject release debugging"
Linus Torvalds2a999aa2013-10-29 08:33:36 -07001249 depends on DEBUG_OBJECTS_TIMERS
Russell Kingc817a672013-06-27 15:06:14 +01001250 help
1251 kobjects are reference counted objects. This means that their
1252 last reference count put is not predictable, and the kobject can
1253 live on past the point at which a driver decides to drop it's
1254 initial reference to the kobject gained on allocation. An
1255 example of this would be a struct device which has just been
1256 unregistered.
1257
1258 However, some buggy drivers assume that after such an operation,
1259 the memory backing the kobject can be immediately freed. This
1260 goes completely against the principles of a refcounted object.
1261
1262 If you say Y here, the kernel will delay the release of kobjects
1263 on the last reference count to improve the visibility of this
1264 kind of kobject release bug.
1265
Catalin Marinas9b2a60c2012-10-08 16:28:13 -07001266config HAVE_DEBUG_BUGVERBOSE
1267 bool
1268
Linus Torvalds1da177e2005-04-16 15:20:36 -07001269config DEBUG_BUGVERBOSE
David Rientjes6a108a12011-01-20 14:44:16 -08001270 bool "Verbose BUG() reporting (adds 70K)" if DEBUG_KERNEL && EXPERT
Catalin Marinas9b2a60c2012-10-08 16:28:13 -07001271 depends on BUG && (GENERIC_BUG || HAVE_DEBUG_BUGVERBOSE)
Alexey Dobriyan8420e7e2009-12-14 18:00:25 -08001272 default y
Linus Torvalds1da177e2005-04-16 15:20:36 -07001273 help
1274 Say Y here to make BUG() panics output the file name and line number
1275 of the BUG call as well as the EIP and oops trace. This aids
1276 debugging but costs about 70-100K of memory.
1277
Dave Jones199a9af2006-09-29 01:59:00 -07001278config DEBUG_LIST
1279 bool "Debug linked list manipulation"
1280 depends on DEBUG_KERNEL
1281 help
1282 Enable this to turn on extended checks in the linked-list
1283 walking routines.
1284
1285 If unsure, say N.
1286
Dan Streetmanb8cfff62014-06-04 16:11:54 -07001287config DEBUG_PI_LIST
1288 bool "Debug priority linked list manipulation"
1289 depends on DEBUG_KERNEL
1290 help
1291 Enable this to turn on extended checks in the priority-ordered
1292 linked-list (plist) walking routines. This checks the entire
1293 list multiple times during each manipulation.
1294
1295 If unsure, say N.
1296
Jens Axboed6ec0842007-10-22 20:01:06 +02001297config DEBUG_SG
1298 bool "Debug SG table operations"
1299 depends on DEBUG_KERNEL
1300 help
1301 Enable this to turn on checks on scatter-gather tables. This can
1302 help find problems with drivers that do not properly initialize
1303 their sg tables.
1304
1305 If unsure, say N.
1306
Arjan van de Ven1b2439d2008-08-15 15:29:38 -07001307config DEBUG_NOTIFIERS
1308 bool "Debug notifier call chains"
1309 depends on DEBUG_KERNEL
1310 help
1311 Enable this to turn on sanity checking for notifier call chains.
1312 This is most useful for kernel developers to make sure that
1313 modules properly unregister themselves from notifier chains.
1314 This is a relatively cheap check but if you care about maximum
1315 performance, say N.
1316
David Howellse0e81732009-09-02 09:13:40 +01001317config DEBUG_CREDENTIALS
1318 bool "Debug credential management"
1319 depends on DEBUG_KERNEL
1320 help
1321 Enable this to turn on some debug checking for credential
1322 management. The additional code keeps track of the number of
1323 pointers from task_structs to any given cred struct, and checks to
1324 see that this number never exceeds the usage count of the cred
1325 struct.
1326
1327 Furthermore, if SELinux is enabled, this also checks that the
1328 security pointer in the cred struct is never seen to be invalid.
1329
1330 If unsure, say N.
1331
Dave Hansen2f03e3c2013-01-07 08:19:23 -08001332menu "RCU Debugging"
1333
1334config PROVE_RCU
Paul E. McKenney9bae6592015-01-18 18:01:21 -08001335 def_bool PROVE_LOCKING
Dave Hansen2f03e3c2013-01-07 08:19:23 -08001336
1337config PROVE_RCU_REPEATEDLY
1338 bool "RCU debugging: don't disable PROVE_RCU on first splat"
1339 depends on PROVE_RCU
1340 default n
1341 help
1342 By itself, PROVE_RCU will disable checking upon issuing the
1343 first warning (or "splat"). This feature prevents such
1344 disabling, allowing multiple RCU-lockdep warnings to be printed
1345 on a single reboot.
1346
1347 Say Y to allow multiple RCU-lockdep warnings per boot.
1348
1349 Say N if you are unsure.
1350
Dave Hansen2f03e3c2013-01-07 08:19:23 -08001351config SPARSE_RCU_POINTER
1352 bool "RCU debugging: sparse-based checks for pointer usage"
1353 default n
1354 help
1355 This feature enables the __rcu sparse annotation for
1356 RCU-protected pointers. This annotation will cause sparse
1357 to flag any non-RCU used of annotated pointers. This can be
1358 helpful when debugging RCU usage. Please note that this feature
1359 is not intended to enforce code cleanliness; it is instead merely
1360 a debugging aid.
1361
1362 Say Y to make sparse flag questionable use of RCU-protected pointers
1363
1364 Say N if you are unsure.
1365
Paul E. McKenney51b11302014-01-27 11:49:39 -08001366config TORTURE_TEST
1367 tristate
1368 default n
1369
Paul E. McKenney8704baa2015-12-31 18:33:22 -08001370config RCU_PERF_TEST
1371 tristate "performance tests for RCU"
1372 depends on DEBUG_KERNEL
1373 select TORTURE_TEST
1374 select SRCU
1375 select TASKS_RCU
1376 default n
1377 help
1378 This option provides a kernel module that runs performance
1379 tests on the RCU infrastructure. The kernel module may be built
1380 after the fact on the running kernel to be tested, if desired.
1381
1382 Say Y here if you want RCU performance tests to be built into
1383 the kernel.
1384 Say M if you want the RCU performance tests to build as a module.
1385 Say N if you are unsure.
1386
Paul E. McKenneya241ec62005-10-30 15:03:12 -08001387config RCU_TORTURE_TEST
1388 tristate "torture tests for RCU"
1389 depends on DEBUG_KERNEL
Paul E. McKenney51b11302014-01-27 11:49:39 -08001390 select TORTURE_TEST
Pranith Kumar83fe27e2014-12-05 11:24:45 -05001391 select SRCU
Paul E. McKenney82d0f4c2015-04-20 05:42:50 -07001392 select TASKS_RCU
Paul E. McKenneya241ec62005-10-30 15:03:12 -08001393 default n
1394 help
1395 This option provides a kernel module that runs torture tests
1396 on the RCU infrastructure. The kernel module may be built
1397 after the fact on the running kernel to be tested, if desired.
1398
Paul E. McKenney31a72bc2008-06-18 09:26:49 -07001399 Say Y here if you want RCU torture tests to be built into
1400 the kernel.
Paul E. McKenneya241ec62005-10-30 15:03:12 -08001401 Say M if you want the RCU torture tests to build as a module.
1402 Say N if you are unsure.
Ankita Garg8bb31b92006-10-02 02:17:36 -07001403
Paul E. McKenney0f41c0d2015-03-10 18:33:20 -07001404config RCU_TORTURE_TEST_SLOW_PREINIT
1405 bool "Slow down RCU grace-period pre-initialization to expose races"
1406 depends on RCU_TORTURE_TEST
1407 help
1408 This option delays grace-period pre-initialization (the
1409 propagation of CPU-hotplug changes up the rcu_node combining
1410 tree) for a few jiffies between initializing each pair of
1411 consecutive rcu_node structures. This helps to expose races
1412 involving grace-period pre-initialization, in other words, it
1413 makes your kernel less stable. It can also greatly increase
1414 grace-period latency, especially on systems with large numbers
1415 of CPUs. This is useful when torture-testing RCU, but in
1416 almost no other circumstance.
1417
1418 Say Y here if you want your system to crash and hang more often.
1419 Say N if you want a sane system.
1420
1421config RCU_TORTURE_TEST_SLOW_PREINIT_DELAY
1422 int "How much to slow down RCU grace-period pre-initialization"
1423 range 0 5
1424 default 3
1425 depends on RCU_TORTURE_TEST_SLOW_PREINIT
1426 help
1427 This option specifies the number of jiffies to wait between
1428 each rcu_node structure pre-initialization step.
1429
Paul E. McKenney37745d22015-01-22 18:24:08 -08001430config RCU_TORTURE_TEST_SLOW_INIT
1431 bool "Slow down RCU grace-period initialization to expose races"
1432 depends on RCU_TORTURE_TEST
1433 help
Paul E. McKenney0f41c0d2015-03-10 18:33:20 -07001434 This option delays grace-period initialization for a few
1435 jiffies between initializing each pair of consecutive
Paul E. McKenney37745d22015-01-22 18:24:08 -08001436 rcu_node structures. This helps to expose races involving
1437 grace-period initialization, in other words, it makes your
1438 kernel less stable. It can also greatly increase grace-period
1439 latency, especially on systems with large numbers of CPUs.
1440 This is useful when torture-testing RCU, but in almost no
1441 other circumstance.
1442
1443 Say Y here if you want your system to crash and hang more often.
1444 Say N if you want a sane system.
1445
1446config RCU_TORTURE_TEST_SLOW_INIT_DELAY
1447 int "How much to slow down RCU grace-period initialization"
1448 range 0 5
Paul E. McKenney186bea52015-01-29 16:37:19 -08001449 default 3
Paul E. McKenney8d7dc922015-04-14 19:33:59 -07001450 depends on RCU_TORTURE_TEST_SLOW_INIT
Paul E. McKenney37745d22015-01-22 18:24:08 -08001451 help
1452 This option specifies the number of jiffies to wait between
1453 each rcu_node structure initialization.
1454
Paul E. McKenney0f41c0d2015-03-10 18:33:20 -07001455config RCU_TORTURE_TEST_SLOW_CLEANUP
1456 bool "Slow down RCU grace-period cleanup to expose races"
1457 depends on RCU_TORTURE_TEST
1458 help
1459 This option delays grace-period cleanup for a few jiffies
1460 between cleaning up each pair of consecutive rcu_node
1461 structures. This helps to expose races involving grace-period
1462 cleanup, in other words, it makes your kernel less stable.
1463 It can also greatly increase grace-period latency, especially
1464 on systems with large numbers of CPUs. This is useful when
1465 torture-testing RCU, but in almost no other circumstance.
1466
1467 Say Y here if you want your system to crash and hang more often.
1468 Say N if you want a sane system.
1469
1470config RCU_TORTURE_TEST_SLOW_CLEANUP_DELAY
1471 int "How much to slow down RCU grace-period cleanup"
1472 range 0 5
1473 default 3
1474 depends on RCU_TORTURE_TEST_SLOW_CLEANUP
1475 help
1476 This option specifies the number of jiffies to wait between
1477 each rcu_node structure cleanup operation.
1478
Paul E. McKenneyb1637602010-06-02 16:21:38 -07001479config RCU_CPU_STALL_TIMEOUT
1480 int "RCU CPU stall timeout in seconds"
Paul E. McKenney6bfc09e2012-10-19 12:49:17 -07001481 depends on RCU_STALL_COMMON
Paul E. McKenneyb1637602010-06-02 16:21:38 -07001482 range 3 300
Paul E. McKenneyc8960542012-10-25 17:59:23 -07001483 default 21
Paul E. McKenneyb1637602010-06-02 16:21:38 -07001484 help
1485 If a given RCU grace period extends more than the specified
1486 number of seconds, a CPU stall warning is printed. If the
1487 RCU grace period persists, additional CPU stall warnings are
1488 printed at more widely spaced intervals.
1489
Channagoud Kadabi2eada802017-03-10 16:30:47 -08001490config RCU_PANIC_ON_STALL
1491 int "Panic on RCU Stalls"
1492 range 0 1
1493 default 0
1494 depends on RCU_STALL_COMMON
1495 help
1496 Panic if a given RCU grace period extends more than the specified
1497 number of seconds instead of just printing a CPU stall warning.
1498 This helps to collect cpu context as part of ramdumps for post
1499 mortem analysis.
1500
1501
Paul E. McKenney5c8806a2012-01-06 15:10:44 -08001502config RCU_TRACE
1503 bool "Enable tracing for RCU"
1504 depends on DEBUG_KERNEL
Paul E. McKenney52494532012-11-14 16:26:40 -08001505 select TRACE_CLOCK
Paul E. McKenney5c8806a2012-01-06 15:10:44 -08001506 help
1507 This option provides tracing in RCU which presents stats
1508 in debugfs for debugging RCU implementation.
1509
1510 Say Y here if you want to enable RCU tracing
1511 Say N if you are unsure.
1512
Paul E. McKenney1ce46ee2015-05-05 23:04:22 -07001513config RCU_EQS_DEBUG
Paul E. McKenneyeb6d5b02015-06-30 09:56:31 -07001514 bool "Provide debugging asserts for adding NO_HZ support to an arch"
Paul E. McKenney1ce46ee2015-05-05 23:04:22 -07001515 depends on DEBUG_KERNEL
1516 help
1517 This option provides consistency checks in RCU's handling of
1518 NO_HZ. These checks have proven quite helpful in detecting
1519 bugs in arch-specific NO_HZ code.
1520
1521 Say N here if you need ultimate kernel/user switch latencies
1522 Say Y if you are unsure
1523
Dave Hansen2f03e3c2013-01-07 08:19:23 -08001524endmenu # "RCU Debugging"
1525
Tejun Heof303fcc2016-02-09 17:59:38 -05001526config DEBUG_WQ_FORCE_RR_CPU
1527 bool "Force round-robin CPU selection for unbound work items"
1528 depends on DEBUG_KERNEL
1529 default n
1530 help
1531 Workqueue used to implicitly guarantee that work items queued
1532 without explicit CPU specified are put on the local CPU. This
1533 guarantee is no longer true and while local CPU is still
1534 preferred work items may be put on foreign CPUs. Kernel
1535 parameter "workqueue.debug_force_rr_cpu" is added to force
1536 round-robin CPU selection to flush out usages which depend on the
1537 now broken guarantee. This config option enables the debug
1538 feature by default. When enabled, memory and cache locality will
1539 be impacted.
1540
Tejun Heo870d6652008-08-25 19:47:25 +09001541config DEBUG_BLOCK_EXT_DEVT
1542 bool "Force extended block device numbers and spread them"
1543 depends on DEBUG_KERNEL
1544 depends on BLOCK
Jens Axboe759f8ca2008-08-29 09:06:29 +02001545 default n
Tejun Heo870d6652008-08-25 19:47:25 +09001546 help
Tejun Heo0e11e342008-10-13 10:46:01 +02001547 BIG FAT WARNING: ENABLING THIS OPTION MIGHT BREAK BOOTING ON
1548 SOME DISTRIBUTIONS. DO NOT ENABLE THIS UNLESS YOU KNOW WHAT
1549 YOU ARE DOING. Distros, please enable this and fix whatever
1550 is broken.
1551
Tejun Heo870d6652008-08-25 19:47:25 +09001552 Conventionally, block device numbers are allocated from
1553 predetermined contiguous area. However, extended block area
1554 may introduce non-contiguous block device numbers. This
1555 option forces most block device numbers to be allocated from
1556 the extended space and spreads them to discover kernel or
1557 userland code paths which assume predetermined contiguous
1558 device number allocation.
1559
Tejun Heo55dc7db2008-09-01 13:44:35 +02001560 Note that turning on this debug option shuffles all the
1561 device numbers for all IDE and SCSI devices including libata
1562 ones, so root partition specified using device number
1563 directly (via rdev or root=MAJ:MIN) won't work anymore.
1564 Textual device names (root=/dev/sdXn) will continue to work.
1565
Tejun Heo870d6652008-08-25 19:47:25 +09001566 Say N if you are unsure.
1567
Thomas Gleixner757c9892016-02-26 18:43:32 +00001568config CPU_HOTPLUG_STATE_CONTROL
1569 bool "Enable CPU hotplug state control"
1570 depends on DEBUG_KERNEL
1571 depends on HOTPLUG_CPU
1572 default n
1573 help
1574 Allows to write steps between "offline" and "online" to the CPUs
1575 sysfs target file so states can be stepped granular. This is a debug
1576 option for now as the hotplug machinery cannot be stopped and
1577 restarted at arbitrary points yet.
1578
1579 Say N if your are unsure.
1580
Akinobu Mita8d438282012-07-30 14:43:02 -07001581config NOTIFIER_ERROR_INJECTION
1582 tristate "Notifier error injection"
1583 depends on DEBUG_KERNEL
1584 select DEBUG_FS
1585 help
Masanari Iidae41e85c2012-11-30 16:44:39 +09001586 This option provides the ability to inject artificial errors to
Akinobu Mita8d438282012-07-30 14:43:02 -07001587 specified notifier chain callbacks. It is useful to test the error
1588 handling of notifier call chain failures.
1589
1590 Say N if unsure.
1591
Akinobu Mitac9d221f2010-05-26 14:43:36 -07001592config CPU_NOTIFIER_ERROR_INJECT
1593 tristate "CPU notifier error injection module"
Akinobu Mitaf5a9f522012-07-30 14:43:03 -07001594 depends on HOTPLUG_CPU && NOTIFIER_ERROR_INJECTION
Akinobu Mitac9d221f2010-05-26 14:43:36 -07001595 help
1596 This option provides a kernel module that can be used to test
Masanari Iidae41e85c2012-11-30 16:44:39 +09001597 the error handling of the cpu notifiers by injecting artificial
Akinobu Mitaf5a9f522012-07-30 14:43:03 -07001598 errors to CPU notifier chain callbacks. It is controlled through
1599 debugfs interface under /sys/kernel/debug/notifier-error-inject/cpu
1600
1601 If the notifier call chain should be failed with some events
1602 notified, write the error code to "actions/<notifier event>/error".
1603
1604 Example: Inject CPU offline error (-1 == -EPERM)
1605
1606 # cd /sys/kernel/debug/notifier-error-inject/cpu
1607 # echo -1 > actions/CPU_DOWN_PREPARE/error
1608 # echo 0 > /sys/devices/system/cpu/cpu1/online
1609 bash: echo: write error: Operation not permitted
Akinobu Mitac9d221f2010-05-26 14:43:36 -07001610
1611 To compile this code as a module, choose M here: the module will
1612 be called cpu-notifier-error-inject.
1613
1614 If unsure, say N.
1615
Akinobu Mita048b9c32012-07-30 14:43:07 -07001616config PM_NOTIFIER_ERROR_INJECT
1617 tristate "PM notifier error injection module"
1618 depends on PM && NOTIFIER_ERROR_INJECTION
1619 default m if PM_DEBUG
1620 help
Masanari Iidae41e85c2012-11-30 16:44:39 +09001621 This option provides the ability to inject artificial errors to
Akinobu Mita048b9c32012-07-30 14:43:07 -07001622 PM notifier chain callbacks. It is controlled through debugfs
1623 interface /sys/kernel/debug/notifier-error-inject/pm
1624
1625 If the notifier call chain should be failed with some events
1626 notified, write the error code to "actions/<notifier event>/error".
1627
1628 Example: Inject PM suspend error (-12 = -ENOMEM)
1629
1630 # cd /sys/kernel/debug/notifier-error-inject/pm/
1631 # echo -12 > actions/PM_SUSPEND_PREPARE/error
1632 # echo mem > /sys/power/state
1633 bash: echo: write error: Cannot allocate memory
1634
1635 To compile this code as a module, choose M here: the module will
1636 be called pm-notifier-error-inject.
1637
1638 If unsure, say N.
1639
Benjamin Herrenschmidtd526e852012-12-14 10:32:52 +11001640config OF_RECONFIG_NOTIFIER_ERROR_INJECT
1641 tristate "OF reconfig notifier error injection module"
1642 depends on OF_DYNAMIC && NOTIFIER_ERROR_INJECTION
Akinobu Mita08dfb4d2012-07-30 14:43:13 -07001643 help
Masanari Iidae41e85c2012-11-30 16:44:39 +09001644 This option provides the ability to inject artificial errors to
Benjamin Herrenschmidtd526e852012-12-14 10:32:52 +11001645 OF reconfig notifier chain callbacks. It is controlled
Akinobu Mita08dfb4d2012-07-30 14:43:13 -07001646 through debugfs interface under
Benjamin Herrenschmidtd526e852012-12-14 10:32:52 +11001647 /sys/kernel/debug/notifier-error-inject/OF-reconfig/
Akinobu Mita08dfb4d2012-07-30 14:43:13 -07001648
1649 If the notifier call chain should be failed with some events
1650 notified, write the error code to "actions/<notifier event>/error".
1651
1652 To compile this code as a module, choose M here: the module will
Akinobu Mitae12a95f2013-04-30 15:28:49 -07001653 be called of-reconfig-notifier-error-inject.
Akinobu Mita08dfb4d2012-07-30 14:43:13 -07001654
1655 If unsure, say N.
1656
Nikolay Aleksandrov02fff962015-11-28 13:45:28 +01001657config NETDEV_NOTIFIER_ERROR_INJECT
1658 tristate "Netdev notifier error injection module"
1659 depends on NET && NOTIFIER_ERROR_INJECTION
1660 help
1661 This option provides the ability to inject artificial errors to
1662 netdevice notifier chain callbacks. It is controlled through debugfs
1663 interface /sys/kernel/debug/notifier-error-inject/netdev
1664
1665 If the notifier call chain should be failed with some events
1666 notified, write the error code to "actions/<notifier event>/error".
1667
1668 Example: Inject netdevice mtu change error (-22 = -EINVAL)
1669
1670 # cd /sys/kernel/debug/notifier-error-inject/netdev
1671 # echo -22 > actions/NETDEV_CHANGEMTU/error
1672 # ip link set eth0 mtu 1024
1673 RTNETLINK answers: Invalid argument
1674
1675 To compile this code as a module, choose M here: the module will
1676 be called netdev-notifier-error-inject.
1677
1678 If unsure, say N.
1679
Akinobu Mita6ff1cb32006-12-08 02:39:43 -08001680config FAULT_INJECTION
Andrew Morton1ab8509a2006-12-08 02:39:49 -08001681 bool "Fault-injection framework"
1682 depends on DEBUG_KERNEL
Akinobu Mita329409a2006-12-08 02:39:48 -08001683 help
1684 Provide fault-injection framework.
1685 For more details, see Documentation/fault-injection/.
Akinobu Mita6ff1cb32006-12-08 02:39:43 -08001686
Akinobu Mita8a8b6502006-12-08 02:39:44 -08001687config FAILSLAB
Andrew Morton1ab8509a2006-12-08 02:39:49 -08001688 bool "Fault-injection capability for kmalloc"
1689 depends on FAULT_INJECTION
Akinobu Mita773ff602008-12-23 19:37:01 +09001690 depends on SLAB || SLUB
Akinobu Mita8a8b6502006-12-08 02:39:44 -08001691 help
Andrew Morton1ab8509a2006-12-08 02:39:49 -08001692 Provide fault-injection capability for kmalloc.
Akinobu Mita8a8b6502006-12-08 02:39:44 -08001693
Akinobu Mita933e3122006-12-08 02:39:45 -08001694config FAIL_PAGE_ALLOC
1695 bool "Fault-injection capabilitiy for alloc_pages()"
Andrew Morton1ab8509a2006-12-08 02:39:49 -08001696 depends on FAULT_INJECTION
Akinobu Mita933e3122006-12-08 02:39:45 -08001697 help
Andrew Morton1ab8509a2006-12-08 02:39:49 -08001698 Provide fault-injection capability for alloc_pages().
Akinobu Mita933e3122006-12-08 02:39:45 -08001699
Akinobu Mitac17bb492006-12-08 02:39:46 -08001700config FAIL_MAKE_REQUEST
Dave Jones86327d12006-12-12 20:16:36 +01001701 bool "Fault-injection capability for disk IO"
Jens Axboe581d4e22008-09-14 05:56:33 -07001702 depends on FAULT_INJECTION && BLOCK
Akinobu Mitac17bb492006-12-08 02:39:46 -08001703 help
Andrew Morton1ab8509a2006-12-08 02:39:49 -08001704 Provide fault-injection capability for disk IO.
Akinobu Mitac17bb492006-12-08 02:39:46 -08001705
Jens Axboe581d4e22008-09-14 05:56:33 -07001706config FAIL_IO_TIMEOUT
Takuya Yoshikawaf4d014392010-07-21 16:05:53 +09001707 bool "Fault-injection capability for faking disk interrupts"
Jens Axboe581d4e22008-09-14 05:56:33 -07001708 depends on FAULT_INJECTION && BLOCK
1709 help
1710 Provide fault-injection capability on end IO handling. This
1711 will make the block layer "forget" an interrupt as configured,
1712 thus exercising the error handling.
1713
1714 Only works with drivers that use the generic timeout handling,
1715 for others it wont do anything.
1716
Per Forlin1b676f72011-08-19 14:52:37 +02001717config FAIL_MMC_REQUEST
1718 bool "Fault-injection capability for MMC IO"
Adrien Schildknecht28ff4fd2015-11-10 20:12:19 +01001719 depends on FAULT_INJECTION_DEBUG_FS && MMC
Per Forlin1b676f72011-08-19 14:52:37 +02001720 help
1721 Provide fault-injection capability for MMC IO.
1722 This will make the mmc core return data errors. This is
1723 useful to test the error handling in the mmc block device
1724 and to test how the mmc host driver handles retries from
1725 the block device.
1726
Davidlohr Buesoab51fba2015-06-29 23:26:02 -07001727config FAIL_FUTEX
1728 bool "Fault-injection capability for futexes"
1729 select DEBUG_FS
1730 depends on FAULT_INJECTION && FUTEX
1731 help
1732 Provide fault-injection capability for futexes.
1733
Akinobu Mita6ff1cb32006-12-08 02:39:43 -08001734config FAULT_INJECTION_DEBUG_FS
1735 bool "Debugfs entries for fault-injection capabilities"
Andrew Morton1ab8509a2006-12-08 02:39:49 -08001736 depends on FAULT_INJECTION && SYSFS && DEBUG_FS
Akinobu Mita6ff1cb32006-12-08 02:39:43 -08001737 help
Andrew Morton1ab8509a2006-12-08 02:39:49 -08001738 Enable configuration of fault-injection capabilities via debugfs.
Akinobu Mita1df49002007-02-20 13:57:56 -08001739
1740config FAULT_INJECTION_STACKTRACE_FILTER
1741 bool "stacktrace filter for fault-injection capabilities"
1742 depends on FAULT_INJECTION_DEBUG_FS && STACKTRACE_SUPPORT
Akinobu Mita6d690dc2007-05-12 10:36:53 -07001743 depends on !X86_64
Akinobu Mita1df49002007-02-20 13:57:56 -08001744 select STACKTRACE
Chen Gangdf2e1ef2014-06-23 13:22:04 -07001745 select FRAME_POINTER if !MIPS && !PPC && !S390 && !MICROBLAZE && !ARM_UNWIND && !ARC && !SCORE
Akinobu Mita1df49002007-02-20 13:57:56 -08001746 help
1747 Provide stacktrace filter for fault-injection capabilities
Mathieu Desnoyers267c4022007-10-18 23:41:07 -07001748
Arjan van de Ven97455122008-01-25 21:08:34 +01001749config LATENCYTOP
1750 bool "Latency measuring infrastructure"
Randy Dunlap625fdca2010-08-12 12:31:21 -07001751 depends on DEBUG_KERNEL
1752 depends on STACKTRACE_SUPPORT
1753 depends on PROC_FS
Vineet Guptacc80ae32013-08-27 13:52:51 +05301754 select FRAME_POINTER if !MIPS && !PPC && !S390 && !MICROBLAZE && !ARM_UNWIND && !ARC
Arjan van de Ven97455122008-01-25 21:08:34 +01001755 select KALLSYMS
1756 select KALLSYMS_ALL
1757 select STACKTRACE
1758 select SCHEDSTATS
1759 select SCHED_DEBUG
Arjan van de Ven97455122008-01-25 21:08:34 +01001760 help
1761 Enable this option if you want to use the LatencyTOP tool
1762 to find out which userspace is blocking on what kernel operations.
1763
Arnaldo Carvalho de Melo16444a82008-05-12 21:20:42 +02001764source kernel/trace/Kconfig
1765
Dave Hansen881c5142013-07-01 13:04:44 -07001766menu "Runtime Testing"
1767
1768config LKDTM
1769 tristate "Linux Kernel Dump Test Tool Module"
1770 depends on DEBUG_FS
1771 depends on BLOCK
1772 default n
1773 help
1774 This module enables testing of the different dumping mechanisms by
1775 inducing system failures at predefined crash points.
1776 If you don't need it: say N
1777 Choose M here to compile this code as a module. The module will be
1778 called lkdtm.
1779
1780 Documentation on how to use the module can be found in
1781 Documentation/fault-injection/provoke-crashes.txt
1782
1783config TEST_LIST_SORT
1784 bool "Linked list sorting test"
1785 depends on DEBUG_KERNEL
1786 help
1787 Enable this to turn on 'list_sort()' function test. This test is
1788 executed only once during system boot, so affects only boot time.
1789
1790 If unsure, say N.
1791
1792config KPROBES_SANITY_TEST
1793 bool "Kprobes sanity tests"
1794 depends on DEBUG_KERNEL
1795 depends on KPROBES
1796 default n
1797 help
1798 This option provides for testing basic kprobes functionality on
1799 boot. A sample kprobe, jprobe and kretprobe are inserted and
1800 verified for functionality.
1801
1802 Say N if you are unsure.
1803
1804config BACKTRACE_SELF_TEST
1805 tristate "Self test for the backtrace code"
1806 depends on DEBUG_KERNEL
1807 default n
1808 help
1809 This option provides a kernel module that can be used to test
1810 the kernel stack backtrace code. This option is not useful
1811 for distributions or general kernels, but only for kernel
1812 developers working on architecture code.
1813
1814 Note that if you want to also test saved backtraces, you will
1815 have to enable STACKTRACE as well.
1816
1817 Say N if you are unsure.
1818
Michel Lespinasse910a7422012-10-08 16:30:39 -07001819config RBTREE_TEST
1820 tristate "Red-Black tree test"
Cody P Schafer7c993e12013-09-11 14:25:19 -07001821 depends on DEBUG_KERNEL
Michel Lespinasse910a7422012-10-08 16:30:39 -07001822 help
1823 A benchmark measuring the performance of the rbtree library.
1824 Also includes rbtree invariant checks.
1825
Michel Lespinassefff3fd82012-10-08 16:31:23 -07001826config INTERVAL_TREE_TEST
1827 tristate "Interval tree test"
1828 depends on m && DEBUG_KERNEL
Chris Wilsona88cc102014-03-17 12:21:54 +00001829 select INTERVAL_TREE
Michel Lespinassefff3fd82012-10-08 16:31:23 -07001830 help
1831 A benchmark measuring the performance of the interval tree library
1832
Greg Thelen623fd802013-11-12 15:08:34 -08001833config PERCPU_TEST
1834 tristate "Per cpu operations test"
1835 depends on m && DEBUG_KERNEL
1836 help
1837 Enable this option to build test module which validates per-cpu
1838 operations.
1839
1840 If unsure, say N.
1841
Dave Hansen881c5142013-07-01 13:04:44 -07001842config ATOMIC64_SELFTEST
1843 bool "Perform an atomic64_t self-test at boot"
1844 help
1845 Enable this option to test the atomic64_t functions at boot.
1846
1847 If unsure, say N.
1848
1849config ASYNC_RAID6_TEST
1850 tristate "Self test for hardware accelerated raid6 recovery"
1851 depends on ASYNC_RAID6_RECOV
1852 select ASYNC_MEMCPY
1853 ---help---
1854 This is a one-shot self test that permutes through the
1855 recovery of all the possible two disk failure scenarios for a
1856 N-disk array. Recovery is performed with the asynchronous
1857 raid6 recovery routines, and will optionally use an offload
1858 engine if one is available.
1859
1860 If unsure, say N.
1861
Andy Shevchenko64d1d772015-02-12 15:02:21 -08001862config TEST_HEXDUMP
1863 tristate "Test functions located in the hexdump module at runtime"
1864
Dave Hansen881c5142013-07-01 13:04:44 -07001865config TEST_STRING_HELPERS
1866 tristate "Test functions located in the string_helpers module at runtime"
1867
1868config TEST_KSTRTOX
1869 tristate "Test kstrto*() family of functions at runtime"
1870
Rasmus Villemoes707cc722015-11-06 16:30:29 -08001871config TEST_PRINTF
1872 tristate "Test printf() family of functions at runtime"
1873
David Decotigny5fd003f2016-02-19 09:24:00 -05001874config TEST_BITMAP
1875 tristate "Test bitmap_*() family of functions at runtime"
1876 default n
1877 help
1878 Enable this option to test the bitmap functions at boot.
1879
1880 If unsure, say N.
1881
Andy Shevchenkocfaff0e2016-05-30 17:40:41 +03001882config TEST_UUID
1883 tristate "Test functions located in the uuid module at runtime"
1884
Thomas Graf7e1e7762014-08-02 11:47:44 +02001885config TEST_RHASHTABLE
Geert Uytterhoeven9d6dbe12015-01-29 15:40:25 +01001886 tristate "Perform selftest on resizable hash table"
Thomas Graf7e1e7762014-08-02 11:47:44 +02001887 default n
1888 help
1889 Enable this option to test the rhashtable functions at boot.
1890
1891 If unsure, say N.
1892
George Spelvin468a9422016-05-26 22:11:51 -04001893config TEST_HASH
1894 tristate "Perform selftest on hash functions"
1895 default n
1896 help
1897 Enable this option to test the kernel's integer (<linux/hash,h>)
1898 and string (<linux/stringhash.h>) hash functions on boot
1899 (or module load).
1900
1901 This is intended to help people writing architecture-specific
1902 optimized versions. If unsure, say N.
1903
Dave Hansen881c5142013-07-01 13:04:44 -07001904endmenu # runtime tests
1905
Bernhard Kaindlf212ec42008-01-30 13:34:11 +01001906config PROVIDE_OHCI1394_DMA_INIT
Stefan Richter080de8c2008-02-28 20:54:43 +01001907 bool "Remote debugging over FireWire early on boot"
Bernhard Kaindlf212ec42008-01-30 13:34:11 +01001908 depends on PCI && X86
1909 help
1910 If you want to debug problems which hang or crash the kernel early
1911 on boot and the crashing machine has a FireWire port, you can use
1912 this feature to remotely access the memory of the crashed machine
1913 over FireWire. This employs remote DMA as part of the OHCI1394
1914 specification which is now the standard for FireWire controllers.
1915
1916 With remote DMA, you can monitor the printk buffer remotely using
1917 firescope and access all memory below 4GB using fireproxy from gdb.
1918 Even controlling a kernel debugger is possible using remote DMA.
1919
1920 Usage:
1921
1922 If ohci1394_dma=early is used as boot parameter, it will initialize
1923 all OHCI1394 controllers which are found in the PCI config space.
1924
1925 As all changes to the FireWire bus such as enabling and disabling
1926 devices cause a bus reset and thereby disable remote DMA for all
1927 devices, be sure to have the cable plugged and FireWire enabled on
1928 the debugging host before booting the debug target for debugging.
1929
1930 This code (~1k) is freed after boot. By then, the firewire stack
1931 in charge of the OHCI-1394 controllers should be used instead.
1932
1933 See Documentation/debugging-via-ohci1394.txt for more information.
Arjan van de Ven97455122008-01-25 21:08:34 +01001934
Joerg Roedel5ee00bd2009-01-09 12:14:24 +01001935config DMA_API_DEBUG
1936 bool "Enable debugging of DMA-API usage"
1937 depends on HAVE_DMA_API_DEBUG
1938 help
1939 Enable this option to debug the use of the DMA API by device drivers.
1940 With this option you will be able to detect common bugs in device
1941 drivers like double-freeing of DMA mappings or freeing mappings that
1942 were never allocated.
Dan Williams0abdd7a2014-01-21 15:48:12 -08001943
1944 This also attempts to catch cases where a page owned by DMA is
1945 accessed by the cpu in a way that could cause data corruption. For
1946 example, this enables cow_user_page() to check that the source page is
1947 not undergoing DMA.
1948
1949 This option causes a performance degradation. Use only if you want to
1950 debug device drivers and dma interactions.
1951
1952 If unsure, say N.
Jason Baron346e15b2008-08-12 16:46:19 -04001953
Valentin Rothberg8a6f0b42014-10-13 15:51:38 -07001954config TEST_LKM
Kees Cook93e9ef82014-01-23 15:54:37 -08001955 tristate "Test module loading with 'hello world' module"
1956 default n
1957 depends on m
1958 help
1959 This builds the "test_module" module that emits "Hello, world"
1960 on printk when loaded. It is designed to be used for basic
1961 evaluation of the module loading subsystem (for example when
1962 validating module verification). It lacks any extra dependencies,
1963 and will not normally be loaded by the system unless explicitly
1964 requested by name.
1965
1966 If unsure, say N.
1967
Kees Cook3e2a4c12014-01-23 15:54:38 -08001968config TEST_USER_COPY
1969 tristate "Test user/kernel boundary protections"
1970 default n
1971 depends on m
1972 help
1973 This builds the "test_user_copy" module that runs sanity checks
1974 on the copy_to/from_user infrastructure, making sure basic
1975 user/kernel boundary testing is working. If it fails to load,
1976 a regression has been detected in the user/kernel memory boundary
1977 protections.
1978
1979 If unsure, say N.
1980
Alexei Starovoitov64a89462014-05-08 14:10:52 -07001981config TEST_BPF
1982 tristate "Test BPF filter functionality"
1983 default n
Randy Dunlap98920ba2014-05-13 09:58:44 -07001984 depends on m && NET
Alexei Starovoitov64a89462014-05-08 14:10:52 -07001985 help
1986 This builds the "test_bpf" module that runs various test vectors
1987 against the BPF interpreter or BPF JIT compiler depending on the
1988 current setting. This is in particular useful for BPF JIT compiler
1989 development, but also to run regression tests against changes in
Alexei Starovoitov3c731eb2014-09-26 00:17:07 -07001990 the interpreter code. It also enables test stubs for eBPF maps and
1991 verifier used by user space verifier testsuite.
Alexei Starovoitov64a89462014-05-08 14:10:52 -07001992
1993 If unsure, say N.
1994
Kees Cook0a8adf52014-07-14 14:38:12 -07001995config TEST_FIRMWARE
1996 tristate "Test firmware loading via userspace interface"
1997 default n
1998 depends on FW_LOADER
1999 help
2000 This builds the "test_firmware" module that creates a userspace
2001 interface for testing firmware loading. This can be used to
2002 control the triggering of firmware loading without needing an
2003 actual firmware-using device. The contents can be rechecked by
2004 userspace.
2005
2006 If unsure, say N.
2007
David Rileye704f932014-06-16 14:58:32 -07002008config TEST_UDELAY
2009 tristate "udelay test driver"
2010 default n
2011 help
2012 This builds the "udelay_test" module that helps to make sure
2013 that udelay() is working properly.
2014
2015 If unsure, say N.
2016
Vladimir Murzin4a207992015-04-14 15:48:27 -07002017config MEMTEST
2018 bool "Memtest"
2019 depends on HAVE_MEMBLOCK
2020 ---help---
2021 This option adds a kernel parameter 'memtest', which allows memtest
2022 to be set.
2023 memtest=0, mean disabled; -- default
2024 memtest=1, mean do 1 test pattern;
2025 ...
Vladimir Murzin8d8cfb42015-04-14 15:48:40 -07002026 memtest=17, mean do 17 test patterns.
Vladimir Murzin4a207992015-04-14 15:48:27 -07002027 If you are unsure how to answer this question, answer N.
2028
Trilok Sonibdf06f72016-04-21 15:46:42 -07002029config MEMTEST_ENABLE_DEFAULT
2030 int "Enable Memtest pattern test by default? (0-17)"
2031 range 0 17
2032 default "0"
2033 depends on MEMTEST
2034 help
2035 This option helps to select Memtest to be enabled through
2036 kernel defconfig options. Alternatively it can be enabled
2037 using memtest=<patterns> kernel command line.
2038
2039 Default value is kept as "0" so that it is kept as disabled.
2040 To enable enter any value between 1-17 range.
2041
Ingo Molnar2bf9e0a2015-08-03 11:42:57 +02002042config TEST_STATIC_KEYS
2043 tristate "Test static keys"
Jason Baron579e1ac2015-07-30 03:59:44 +00002044 default n
2045 depends on m
2046 help
Ingo Molnar2bf9e0a2015-08-03 11:42:57 +02002047 Test the static key interfaces.
Jason Baron579e1ac2015-07-30 03:59:44 +00002048
2049 If unsure, say N.
2050
Syed Rameez Mustafacf9c1e42013-07-15 11:52:09 -07002051config PANIC_ON_DATA_CORRUPTION
2052 bool "Cause a Kernel Panic When Data Corruption is detected"
2053 help
2054 Select this option to upgrade warnings for potentially
2055 recoverable data corruption scenarios to system-halting panics,
2056 for easier detection and debug.
2057
Mathieu Desnoyers267c4022007-10-18 23:41:07 -07002058source "samples/Kconfig"
Jason Wesseldc7d5522008-04-17 20:05:37 +02002059
2060source "lib/Kconfig.kgdb"
Pekka Enberg0a4af3b2009-02-26 21:38:56 +02002061
Andrey Ryabininc6d30852016-01-20 15:00:55 -08002062source "lib/Kconfig.ubsan"
2063
Dan Williams21266be2015-11-19 18:19:29 -08002064config ARCH_HAS_DEVMEM_IS_ALLOWED
2065 bool
2066
2067config STRICT_DEVMEM
2068 bool "Filter access to /dev/mem"
2069 depends on MMU
2070 depends on ARCH_HAS_DEVMEM_IS_ALLOWED
2071 default y if TILE || PPC
2072 ---help---
2073 If this option is disabled, you allow userspace (root) access to all
2074 of memory, including kernel and userspace memory. Accidental
2075 access to this is obviously disastrous, but specific access can
2076 be used by people debugging the kernel. Note that with PAT support
2077 enabled, even in this case there are restrictions on /dev/mem
2078 use due to the cache aliasing requirements.
2079
Dan Williams90a545e2015-11-23 15:49:03 -08002080 If this option is switched on, and IO_STRICT_DEVMEM=n, the /dev/mem
2081 file only allows userspace access to PCI space and the BIOS code and
2082 data regions. This is sufficient for dosemu and X and all common
2083 users of /dev/mem.
2084
2085 If in doubt, say Y.
2086
2087config IO_STRICT_DEVMEM
2088 bool "Filter I/O access to /dev/mem"
2089 depends on STRICT_DEVMEM
Dan Williams90a545e2015-11-23 15:49:03 -08002090 ---help---
2091 If this option is disabled, you allow userspace (root) access to all
2092 io-memory regardless of whether a driver is actively using that
2093 range. Accidental access to this is obviously disastrous, but
2094 specific access can be used by people debugging kernel drivers.
2095
Dan Williams21266be2015-11-19 18:19:29 -08002096 If this option is switched on, the /dev/mem file only allows
Dan Williams90a545e2015-11-23 15:49:03 -08002097 userspace access to *idle* io-memory ranges (see /proc/iomem) This
2098 may break traditional users of /dev/mem (dosemu, legacy X, etc...)
2099 if the driver using a given range cannot be disabled.
Dan Williams21266be2015-11-19 18:19:29 -08002100
2101 If in doubt, say Y.