1 menu "printk and dmesg options"
4 bool "Show timing information on printks"
7 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.
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.
15 The behavior is also controlled by the kernel command line
16 parameter printk.time=1. See Documentation/kernel-parameters.txt
18 config MESSAGE_LOGLEVEL_DEFAULT
19 int "Default message log level (1-7)"
23 Default log level for printk statements with no specified priority.
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
29 config BOOT_PRINTK_DELAY
30 bool "Delay each boot printk message by N milliseconds"
31 depends on DEBUG_KERNEL && PRINTK && GENERIC_CALIBRATE_DELAY
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,
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.
48 bool "Enable dynamic printk() support"
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%.
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.
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:
75 filename:lineno [module]function flags format
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
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"
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
98 // enable all the messages in file svcsock.c
99 nullarbor:~ # echo -n 'file svcsock.c +p' >
100 <debugfs>/dynamic_debug/control
102 // enable all the messages in the NFS server module
103 nullarbor:~ # echo -n 'module nfsd +p' >
104 <debugfs>/dynamic_debug/control
106 // enable all 12 messages in the function svc_process()
107 nullarbor:~ # echo -n 'func svc_process +p' >
108 <debugfs>/dynamic_debug/control
110 // disable all 12 messages in the function svc_process()
111 nullarbor:~ # echo -n 'func svc_process -p' >
112 <debugfs>/dynamic_debug/control
114 See Documentation/dynamic-debug-howto.txt for additional information.
116 endmenu # "printk and dmesg options"
118 menu "Compile-time checks and compiler options"
121 bool "Compile the kernel with debug info"
122 depends on DEBUG_KERNEL && !COMPILE_TEST
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.
133 config DEBUG_INFO_REDUCED
134 bool "Reduce debugging information"
135 depends on DEBUG_INFO
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.
146 config DEBUG_INFO_SPLIT
147 bool "Produce split debuginfo in .dwo files"
148 depends on DEBUG_INFO
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.
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.
161 config DEBUG_INFO_DWARF4
162 bool "Generate dwarf4 debuginfo"
163 depends on DEBUG_INFO
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.
171 bool "Provide GDB scripts for kernel debugging"
172 depends on DEBUG_INFO
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
181 config ENABLE_WARN_DEPRECATED
182 bool "Enable __deprecated logic"
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.
189 config ENABLE_MUST_CHECK
190 bool "Enable __must_check logic"
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.
198 int "Warn for stack frames larger than (needs gcc 4.4)"
201 default 1024 if !64BIT
202 default 2048 if 64BIT
204 Tell gcc to warn at build time for stack frames larger than this.
205 Setting this too low will cause a lot of warnings.
206 Setting it to 0 disables the warning.
209 config STRIP_ASM_SYMS
210 bool "Strip assembler-generated symbols during link"
213 Strip internal assembler-generated symbols during a link (symbols
214 that look like '.Lxxx') so they don't pollute the output of
215 get_wchan() and suchlike.
218 bool "Generate readable assembler code"
219 depends on DEBUG_KERNEL
221 Disable some compiler optimizations that tend to generate human unreadable
222 assembler output. This may make the kernel slightly slower, but it helps
223 to keep kernel developers who have to stare a lot at assembler listings
226 config UNUSED_SYMBOLS
227 bool "Enable unused/obsolete exported symbols"
230 Unused but exported symbols make the kernel needlessly bigger. For
231 that reason most of these unused exports will soon be removed. This
232 option is provided temporarily to provide a transition period in case
233 some external kernel module needs one of these symbols anyway. If you
234 encounter such a case in your module, consider if you are actually
235 using the right API. (rationale: since nobody in the kernel is using
236 this in a module, there is a pretty good chance it's actually the
237 wrong interface to use). If you really need the symbol, please send a
238 mail to the linux kernel mailing list mentioning the symbol and why
239 you really need it, and what the merge plan to the mainline kernel for
243 bool "Track page owner"
244 depends on DEBUG_KERNEL && STACKTRACE_SUPPORT
247 select PAGE_EXTENSION
249 This keeps track of what call chain is the owner of a page, may
250 help to find bare alloc_page(s) leaks. Even if you include this
251 feature on your build, it is disabled in default. You should pass
252 "page_owner=on" to boot parameter in order to enable it. Eats
253 a fair amount of memory if enabled. See tools/vm/page_owner_sort.c
254 for user-space helper.
259 bool "Debug Filesystem"
262 debugfs is a virtual file system that kernel developers use to put
263 debugging files into. Enable this option to be able to read and
264 write to these files.
266 For detailed documentation on the debugfs API, see
267 Documentation/DocBook/filesystems.
272 bool "Run 'make headers_check' when building vmlinux"
275 This option will extract the user-visible kernel headers whenever
276 building the kernel, and will run basic sanity checks on them to
277 ensure that exported files do not attempt to include files which
278 were not exported, etc.
280 If you're making modifications to header files which are
281 relevant for userspace, say 'Y', and check the headers
282 exported to $(INSTALL_HDR_PATH) (usually 'usr/include' in
283 your build tree), to make sure they're suitable.
285 config DEBUG_SECTION_MISMATCH
286 bool "Enable full Section mismatch analysis"
288 The section mismatch analysis checks if there are illegal
289 references from one section to another section.
290 During linktime or runtime, some sections are dropped;
291 any use of code/data previously in these sections would
292 most likely result in an oops.
293 In the code, functions and variables are annotated with
294 __init,, etc. (see the full list in include/linux/init.h),
295 which results in the code/data being placed in specific sections.
296 The section mismatch analysis is always performed after a full
297 kernel build, and enabling this option causes the following
298 additional steps to occur:
299 - Add the option -fno-inline-functions-called-once to gcc commands.
300 When inlining a function annotated with __init in a non-init
301 function, we would lose the section information and thus
302 the analysis would not catch the illegal reference.
303 This option tells gcc to inline less (but it does result in
305 - Run the section mismatch analysis for each module/built-in.o file.
306 When we run the section mismatch analysis on vmlinux.o, we
307 lose valueble information about where the mismatch was
309 Running the analysis for each module/built-in.o file
310 tells where the mismatch happens much closer to the
311 source. The drawback is that the same mismatch is
312 reported at least twice.
313 - Enable verbose reporting from modpost in order to help resolve
314 the section mismatches that are reported.
316 config SECTION_MISMATCH_WARN_ONLY
317 bool "Make section mismatch errors non-fatal"
320 If you say N here, the build process will fail if there are any
321 section mismatch, instead of just throwing warnings.
326 # Select this config option from the architecture Kconfig, if it
327 # is preferred to always offer frame pointers as a config
328 # option on the architecture (regardless of KERNEL_DEBUG):
330 config ARCH_WANT_FRAME_POINTERS
335 bool "Compile the kernel with frame pointers"
336 depends on DEBUG_KERNEL && \
337 (CRIS || M68K || FRV || UML || \
338 AVR32 || SUPERH || BLACKFIN || MN10300 || METAG) || \
339 ARCH_WANT_FRAME_POINTERS
340 default y if (DEBUG_INFO && UML) || ARCH_WANT_FRAME_POINTERS
342 If you say Y here the resulting kernel image will be slightly
343 larger and slower, but it gives very useful debugging information
344 in case of kernel bugs. (precise oopses/stacktraces/warnings)
346 config STACK_VALIDATION
347 bool "Compile-time stack metadata validation"
348 depends on HAVE_STACK_VALIDATION
351 Add compile-time checks to validate stack metadata, including frame
352 pointers (if CONFIG_FRAME_POINTER is enabled). This helps ensure
353 that runtime stack traces are more reliable.
355 For more information, see
356 tools/objtool/Documentation/stack-validation.txt.
358 config DEBUG_FORCE_WEAK_PER_CPU
359 bool "Force weak per-cpu definitions"
360 depends on DEBUG_KERNEL
362 s390 and alpha require percpu variables in modules to be
363 defined weak to work around addressing range issue which
364 puts the following two restrictions on percpu variable
367 1. percpu symbols must be unique whether static or not
368 2. percpu variables can't be defined inside a function
370 To ensure that generic code follows the above rules, this
371 option forces all percpu variables to be defined as weak.
373 endmenu # "Compiler options"
376 bool "Magic SysRq key"
379 If you say Y here, you will have some control over the system even
380 if the system crashes for example during kernel debugging (e.g., you
381 will be able to flush the buffer cache to disk, reboot the system
382 immediately or dump some status information). This is accomplished
383 by pressing various keys while holding SysRq (Alt+PrintScreen). It
384 also works on a serial console (on PC hardware at least), if you
385 send a BREAK and then within 5 seconds a command keypress. The
386 keys are documented in <file:Documentation/sysrq.txt>. Don't say Y
387 unless you really know what this hack does.
389 config MAGIC_SYSRQ_DEFAULT_ENABLE
390 hex "Enable magic SysRq key functions by default"
391 depends on MAGIC_SYSRQ
394 Specifies which SysRq key functions are enabled by default.
395 This may be set to 1 or 0 to enable or disable them all, or
396 to a bitmask as described in Documentation/sysrq.txt.
399 bool "Kernel debugging"
401 Say Y here if you are developing drivers or trying to debug and
402 identify kernel problems.
404 menu "Memory Debugging"
406 source mm/Kconfig.debug
409 bool "Debug object operations"
410 depends on DEBUG_KERNEL
412 If you say Y here, additional code will be inserted into the
413 kernel to track the life time of various objects and validate
414 the operations on those objects.
416 config DEBUG_OBJECTS_SELFTEST
417 bool "Debug objects selftest"
418 depends on DEBUG_OBJECTS
420 This enables the selftest of the object debug code.
422 config DEBUG_OBJECTS_FREE
423 bool "Debug objects in freed memory"
424 depends on DEBUG_OBJECTS
426 This enables checks whether a k/v free operation frees an area
427 which contains an object which has not been deactivated
428 properly. This can make kmalloc/kfree-intensive workloads
431 config DEBUG_OBJECTS_TIMERS
432 bool "Debug timer objects"
433 depends on DEBUG_OBJECTS
435 If you say Y here, additional code will be inserted into the
436 timer routines to track the life time of timer objects and
437 validate the timer operations.
439 config DEBUG_OBJECTS_WORK
440 bool "Debug work objects"
441 depends on DEBUG_OBJECTS
443 If you say Y here, additional code will be inserted into the
444 work queue routines to track the life time of work objects and
445 validate the work operations.
447 config DEBUG_OBJECTS_RCU_HEAD
448 bool "Debug RCU callbacks objects"
449 depends on DEBUG_OBJECTS
451 Enable this to turn on debugging of RCU list heads (call_rcu() usage).
453 config DEBUG_OBJECTS_PERCPU_COUNTER
454 bool "Debug percpu counter objects"
455 depends on DEBUG_OBJECTS
457 If you say Y here, additional code will be inserted into the
458 percpu counter routines to track the life time of percpu counter
459 objects and validate the percpu counter operations.
461 config DEBUG_OBJECTS_ENABLE_DEFAULT
462 int "debug_objects bootup default value (0-1)"
465 depends on DEBUG_OBJECTS
467 Debug objects boot parameter default value
470 bool "Debug slab memory allocations"
471 depends on DEBUG_KERNEL && SLAB && !KMEMCHECK
473 Say Y here to have the kernel do limited verification on memory
474 allocation as well as poisoning memory on free to catch use of freed
475 memory. This can make kmalloc/kfree-intensive workloads much slower.
477 config DEBUG_SLAB_LEAK
478 bool "Memory leak debugging"
479 depends on DEBUG_SLAB
482 bool "SLUB debugging on by default"
483 depends on SLUB && SLUB_DEBUG && !KMEMCHECK
486 Boot with debugging on by default. SLUB boots by default with
487 the runtime debug capabilities switched off. Enabling this is
488 equivalent to specifying the "slub_debug" parameter on boot.
489 There is no support for more fine grained debug control like
490 possible with slub_debug=xxx. SLUB debugging may be switched
491 off in a kernel built with CONFIG_SLUB_DEBUG_ON by specifying
496 bool "Enable SLUB performance statistics"
497 depends on SLUB && SYSFS
499 SLUB statistics are useful to debug SLUBs allocation behavior in
500 order find ways to optimize the allocator. This should never be
501 enabled for production use since keeping statistics slows down
502 the allocator by a few percentage points. The slabinfo command
503 supports the determination of the most active slabs to figure
504 out which slabs are relevant to a particular load.
505 Try running: slabinfo -DA
507 config HAVE_DEBUG_KMEMLEAK
510 config DEBUG_KMEMLEAK
511 bool "Kernel memory leak detector"
512 depends on DEBUG_KERNEL && HAVE_DEBUG_KMEMLEAK
514 select STACKTRACE if STACKTRACE_SUPPORT
518 Say Y here if you want to enable the memory leak
519 detector. The memory allocation/freeing is traced in a way
520 similar to the Boehm's conservative garbage collector, the
521 difference being that the orphan objects are not freed but
522 only shown in /sys/kernel/debug/kmemleak. Enabling this
523 feature will introduce an overhead to memory
524 allocations. See Documentation/kmemleak.txt for more
527 Enabling DEBUG_SLAB or SLUB_DEBUG may increase the chances
528 of finding leaks due to the slab objects poisoning.
530 In order to access the kmemleak file, debugfs needs to be
531 mounted (usually at /sys/kernel/debug).
533 config DEBUG_KMEMLEAK_EARLY_LOG_SIZE
534 int "Maximum kmemleak early log entries"
535 depends on DEBUG_KMEMLEAK
539 Kmemleak must track all the memory allocations to avoid
540 reporting false positives. Since memory may be allocated or
541 freed before kmemleak is initialised, an early log buffer is
542 used to store these actions. If kmemleak reports "early log
543 buffer exceeded", please increase this value.
545 config DEBUG_KMEMLEAK_TEST
546 tristate "Simple test for the kernel memory leak detector"
547 depends on DEBUG_KMEMLEAK && m
549 This option enables a module that explicitly leaks memory.
553 config DEBUG_KMEMLEAK_DEFAULT_OFF
554 bool "Default kmemleak to off"
555 depends on DEBUG_KMEMLEAK
557 Say Y here to disable kmemleak by default. It can then be enabled
558 on the command line via kmemleak=on.
560 config DEBUG_STACK_USAGE
561 bool "Stack utilization instrumentation"
562 depends on DEBUG_KERNEL && !IA64
564 Enables the display of the minimum amount of free stack which each
565 task has ever had available in the sysrq-T and sysrq-P debug output.
567 This option will slow down process creation somewhat.
571 depends on DEBUG_KERNEL
573 Enable this to turn on extended checks in the virtual-memory system
574 that may impact performance.
578 config DEBUG_VM_VMACACHE
579 bool "Debug VMA caching"
582 Enable this to turn on VMA caching debug information. Doing so
583 can cause significant overhead, so only enable it in non-production
589 bool "Debug VM red-black trees"
592 Enable VM red-black tree debugging information and extra validations.
596 config DEBUG_VM_PGFLAGS
597 bool "Debug page-flags operations"
600 Enables extra validation on page flags operations.
605 bool "Debug VM translations"
606 depends on DEBUG_KERNEL && X86
608 Enable some costly sanity checks in virtual to page code. This can
609 catch mistakes with virt_to_page() and friends.
613 config DEBUG_NOMMU_REGIONS
614 bool "Debug the global anon/private NOMMU mapping region tree"
615 depends on DEBUG_KERNEL && !MMU
617 This option causes the global tree of anonymous and private mapping
618 regions to be regularly checked for invalid topology.
620 config DEBUG_MEMORY_INIT
621 bool "Debug memory initialisation" if EXPERT
624 Enable this for additional checks during memory initialisation.
625 The sanity checks verify aspects of the VM such as the memory model
626 and other information provided by the architecture. Verbose
627 information will be printed at KERN_DEBUG loglevel depending
628 on the mminit_loglevel= command-line option.
632 config MEMORY_NOTIFIER_ERROR_INJECT
633 tristate "Memory hotplug notifier error injection module"
634 depends on MEMORY_HOTPLUG_SPARSE && NOTIFIER_ERROR_INJECTION
636 This option provides the ability to inject artificial errors to
637 memory hotplug notifier chain callbacks. It is controlled through
638 debugfs interface under /sys/kernel/debug/notifier-error-inject/memory
640 If the notifier call chain should be failed with some events
641 notified, write the error code to "actions/<notifier event>/error".
643 Example: Inject memory hotplug offline error (-12 == -ENOMEM)
645 # cd /sys/kernel/debug/notifier-error-inject/memory
646 # echo -12 > actions/MEM_GOING_OFFLINE/error
647 # echo offline > /sys/devices/system/memory/memoryXXX/state
648 bash: echo: write error: Cannot allocate memory
650 To compile this code as a module, choose M here: the module will
651 be called memory-notifier-error-inject.
655 config DEBUG_PER_CPU_MAPS
656 bool "Debug access to per_cpu maps"
657 depends on DEBUG_KERNEL
660 Say Y to verify that the per_cpu map being accessed has
661 been set up. This adds a fair amount of code to kernel memory
662 and decreases performance.
667 bool "Highmem debugging"
668 depends on DEBUG_KERNEL && HIGHMEM
670 This option enables additional error checking for high memory
671 systems. Disable for production systems.
673 config HAVE_DEBUG_STACKOVERFLOW
676 config DEBUG_STACKOVERFLOW
677 bool "Check for stack overflows"
678 depends on DEBUG_KERNEL && HAVE_DEBUG_STACKOVERFLOW
680 Say Y here if you want to check for overflows of kernel, IRQ
681 and exception stacks (if your architecture uses them). This
682 option will show detailed messages if free stack space drops
683 below a certain limit.
685 These kinds of bugs usually occur when call-chains in the
686 kernel get too deep, especially when interrupts are
689 Use this in cases where you see apparently random memory
690 corruption, especially if it appears in 'struct thread_info'
692 If in doubt, say "N".
694 source "lib/Kconfig.kmemcheck"
696 source "lib/Kconfig.kasan"
698 endmenu # "Memory Debugging"
703 KCOV does not have any arch-specific code, but currently it is enabled
704 only for x86_64. KCOV requires testing on other archs, and most likely
705 disabling of instrumentation for some early boot code.
708 bool "Code coverage for fuzzing"
709 depends on ARCH_HAS_KCOV
712 KCOV exposes kernel code coverage information in a form suitable
713 for coverage-guided fuzzing (randomized testing).
715 If RANDOMIZE_BASE is enabled, PC values will not be stable across
716 different machines and across reboots. If you need stable PC values,
717 disable RANDOMIZE_BASE.
719 For more details, see Documentation/kcov.txt.
722 bool "Debug shared IRQ handlers"
723 depends on DEBUG_KERNEL
725 Enable this to generate a spurious interrupt as soon as a shared
726 interrupt handler is registered, and just before one is deregistered.
727 Drivers ought to be able to handle interrupts coming in at those
728 points; some don't and need to be caught.
730 menu "Debug Lockups and Hangs"
732 config LOCKUP_DETECTOR
733 bool "Detect Hard and Soft Lockups"
734 depends on DEBUG_KERNEL && !S390
736 Say Y here to enable the kernel to act as a watchdog to detect
737 hard and soft lockups.
739 Softlockups are bugs that cause the kernel to loop in kernel
740 mode for more than 20 seconds, without giving other tasks a
741 chance to run. The current stack trace is displayed upon
742 detection and the system will stay locked up.
744 Hardlockups are bugs that cause the CPU to loop in kernel mode
745 for more than 10 seconds, without letting other interrupts have a
746 chance to run. The current stack trace is displayed upon detection
747 and the system will stay locked up.
749 The overhead should be minimal. A periodic hrtimer runs to
750 generate interrupts and kick the watchdog task every 4 seconds.
751 An NMI is generated every 10 seconds or so to check for hardlockups.
753 The frequency of hrtimer and NMI events and the soft and hard lockup
754 thresholds can be controlled through the sysctl watchdog_thresh.
756 config HARDLOCKUP_DETECTOR
758 depends on LOCKUP_DETECTOR && !HAVE_NMI_WATCHDOG
759 depends on PERF_EVENTS && HAVE_PERF_EVENTS_NMI
761 config BOOTPARAM_HARDLOCKUP_PANIC
762 bool "Panic (Reboot) On Hard Lockups"
763 depends on HARDLOCKUP_DETECTOR
765 Say Y here to enable the kernel to panic on "hard lockups",
766 which are bugs that cause the kernel to loop in kernel
767 mode with interrupts disabled for more than 10 seconds (configurable
768 using the watchdog_thresh sysctl).
772 config BOOTPARAM_HARDLOCKUP_PANIC_VALUE
774 depends on HARDLOCKUP_DETECTOR
776 default 0 if !BOOTPARAM_HARDLOCKUP_PANIC
777 default 1 if BOOTPARAM_HARDLOCKUP_PANIC
779 config BOOTPARAM_SOFTLOCKUP_PANIC
780 bool "Panic (Reboot) On Soft Lockups"
781 depends on LOCKUP_DETECTOR
783 Say Y here to enable the kernel to panic on "soft lockups",
784 which are bugs that cause the kernel to loop in kernel
785 mode for more than 20 seconds (configurable using the watchdog_thresh
786 sysctl), without giving other tasks a chance to run.
788 The panic can be used in combination with panic_timeout,
789 to cause the system to reboot automatically after a
790 lockup has been detected. This feature is useful for
791 high-availability systems that have uptime guarantees and
792 where a lockup must be resolved ASAP.
796 config BOOTPARAM_SOFTLOCKUP_PANIC_VALUE
798 depends on LOCKUP_DETECTOR
800 default 0 if !BOOTPARAM_SOFTLOCKUP_PANIC
801 default 1 if BOOTPARAM_SOFTLOCKUP_PANIC
803 config DETECT_HUNG_TASK
804 bool "Detect Hung Tasks"
805 depends on DEBUG_KERNEL
806 default LOCKUP_DETECTOR
808 Say Y here to enable the kernel to detect "hung tasks",
809 which are bugs that cause the task to be stuck in
810 uninterruptible "D" state indefinitiley.
812 When a hung task is detected, the kernel will print the
813 current stack trace (which you should report), but the
814 task will stay in uninterruptible state. If lockdep is
815 enabled then all held locks will also be reported. This
816 feature has negligible overhead.
818 config DEFAULT_HUNG_TASK_TIMEOUT
819 int "Default timeout for hung task detection (in seconds)"
820 depends on DETECT_HUNG_TASK
823 This option controls the default timeout (in seconds) used
824 to determine when a task has become non-responsive and should
827 It can be adjusted at runtime via the kernel.hung_task_timeout_secs
828 sysctl or by writing a value to
829 /proc/sys/kernel/hung_task_timeout_secs.
831 A timeout of 0 disables the check. The default is two minutes.
832 Keeping the default should be fine in most cases.
834 config BOOTPARAM_HUNG_TASK_PANIC
835 bool "Panic (Reboot) On Hung Tasks"
836 depends on DETECT_HUNG_TASK
838 Say Y here to enable the kernel to panic on "hung tasks",
839 which are bugs that cause the kernel to leave a task stuck
840 in uninterruptible "D" state.
842 The panic can be used in combination with panic_timeout,
843 to cause the system to reboot automatically after a
844 hung task has been detected. This feature is useful for
845 high-availability systems that have uptime guarantees and
846 where a hung tasks must be resolved ASAP.
850 config BOOTPARAM_HUNG_TASK_PANIC_VALUE
852 depends on DETECT_HUNG_TASK
854 default 0 if !BOOTPARAM_HUNG_TASK_PANIC
855 default 1 if BOOTPARAM_HUNG_TASK_PANIC
858 bool "Detect Workqueue Stalls"
859 depends on DEBUG_KERNEL
861 Say Y here to enable stall detection on workqueues. If a
862 worker pool doesn't make forward progress on a pending work
863 item for over a given amount of time, 30s by default, a
864 warning message is printed along with dump of workqueue
865 state. This can be configured through kernel parameter
866 "workqueue.watchdog_thresh" and its sysfs counterpart.
868 endmenu # "Debug lockups and hangs"
873 Say Y here to enable the kernel to panic when it oopses. This
874 has the same effect as setting oops=panic on the kernel command
877 This feature is useful to ensure that the kernel does not do
878 anything erroneous after an oops which could result in data
879 corruption or other issues.
883 config PANIC_ON_OOPS_VALUE
886 default 0 if !PANIC_ON_OOPS
887 default 1 if PANIC_ON_OOPS
893 Set the timeout value (in seconds) until a reboot occurs when the
894 the kernel panics. If n = 0, then we wait forever. A timeout
895 value n > 0 will wait n seconds before rebooting, while a timeout
896 value n < 0 will reboot immediately.
899 bool "Collect scheduler debugging info"
900 depends on DEBUG_KERNEL && PROC_FS
903 If you say Y here, the /proc/sched_debug file will be provided
904 that can help debug the scheduler. The runtime overhead of this
912 bool "Collect scheduler statistics"
913 depends on DEBUG_KERNEL && PROC_FS
916 If you say Y here, additional code will be inserted into the
917 scheduler and related routines to collect statistics about
918 scheduler behavior and provide them in /proc/schedstat. These
919 stats may be useful for both tuning and debugging the scheduler
920 If you aren't debugging the scheduler or trying to tune a specific
921 application, you can say N to avoid the very slight overhead
924 config SCHED_STACK_END_CHECK
925 bool "Detect stack corruption on calls to schedule()"
926 depends on DEBUG_KERNEL
929 This option checks for a stack overrun on calls to schedule().
930 If the stack end location is found to be over written always panic as
931 the content of the corrupted region can no longer be trusted.
932 This is to ensure no erroneous behaviour occurs which could result in
933 data corruption or a sporadic crash at a later stage once the region
934 is examined. The runtime overhead introduced is minimal.
936 config DEBUG_TIMEKEEPING
937 bool "Enable extra timekeeping sanity checking"
939 This option will enable additional timekeeping sanity checks
940 which may be helpful when diagnosing issues where timekeeping
941 problems are suspected.
943 This may include checks in the timekeeping hotpaths, so this
944 option may have a (very small) performance impact to some
950 bool "Collect kernel timers statistics"
951 depends on DEBUG_KERNEL && PROC_FS
953 If you say Y here, additional code will be inserted into the
954 timer routines to collect statistics about kernel timers being
955 reprogrammed. The statistics can be read from /proc/timer_stats.
956 The statistics collection is started by writing 1 to /proc/timer_stats,
957 writing 0 stops it. This feature is useful to collect information
958 about timer usage patterns in kernel and userspace. This feature
959 is lightweight if enabled in the kernel config but not activated
960 (it defaults to deactivated on bootup and will only be activated
961 if some application like powertop activates it explicitly).
964 bool "Debug preemptible kernel"
965 depends on DEBUG_KERNEL && PREEMPT && TRACE_IRQFLAGS_SUPPORT
968 If you say Y here then the kernel will use a debug variant of the
969 commonly used smp_processor_id() function and will print warnings
970 if kernel code uses it in a preemption-unsafe way. Also, the kernel
971 will detect preemption count underflows.
973 menu "Lock Debugging (spinlocks, mutexes, etc...)"
975 config DEBUG_RT_MUTEXES
976 bool "RT Mutex debugging, deadlock detection"
977 depends on DEBUG_KERNEL && RT_MUTEXES
979 This allows rt mutex semantics violations and rt mutex related
980 deadlocks (lockups) to be detected and reported automatically.
982 config DEBUG_SPINLOCK
983 bool "Spinlock and rw-lock debugging: basic checks"
984 depends on DEBUG_KERNEL
985 select UNINLINE_SPIN_UNLOCK
987 Say Y here and build SMP to catch missing spinlock initialization
988 and certain other kinds of spinlock errors commonly made. This is
989 best used in conjunction with the NMI watchdog so that spinlock
990 deadlocks are also debuggable.
993 bool "Mutex debugging: basic checks"
994 depends on DEBUG_KERNEL
996 This feature allows mutex semantics violations to be detected and
999 config DEBUG_WW_MUTEX_SLOWPATH
1000 bool "Wait/wound mutex debugging: Slowpath testing"
1001 depends on DEBUG_KERNEL && TRACE_IRQFLAGS_SUPPORT && STACKTRACE_SUPPORT && LOCKDEP_SUPPORT
1002 select DEBUG_LOCK_ALLOC
1003 select DEBUG_SPINLOCK
1004 select DEBUG_MUTEXES
1006 This feature enables slowpath testing for w/w mutex users by
1007 injecting additional -EDEADLK wound/backoff cases. Together with
1008 the full mutex checks enabled with (CONFIG_PROVE_LOCKING) this
1009 will test all possible w/w mutex interface abuse with the
1010 exception of simply not acquiring all the required locks.
1011 Note that this feature can introduce significant overhead, so
1012 it really should not be enabled in a production or distro kernel,
1013 even a debug kernel. If you are a driver writer, enable it. If
1014 you are a distro, do not.
1016 config DEBUG_LOCK_ALLOC
1017 bool "Lock debugging: detect incorrect freeing of live locks"
1018 depends on DEBUG_KERNEL && TRACE_IRQFLAGS_SUPPORT && STACKTRACE_SUPPORT && LOCKDEP_SUPPORT
1019 select DEBUG_SPINLOCK
1020 select DEBUG_MUTEXES
1023 This feature will check whether any held lock (spinlock, rwlock,
1024 mutex or rwsem) is incorrectly freed by the kernel, via any of the
1025 memory-freeing routines (kfree(), kmem_cache_free(), free_pages(),
1026 vfree(), etc.), whether a live lock is incorrectly reinitialized via
1027 spin_lock_init()/mutex_init()/etc., or whether there is any lock
1028 held during task exit.
1030 config PROVE_LOCKING
1031 bool "Lock debugging: prove locking correctness"
1032 depends on DEBUG_KERNEL && TRACE_IRQFLAGS_SUPPORT && STACKTRACE_SUPPORT && LOCKDEP_SUPPORT
1034 select DEBUG_SPINLOCK
1035 select DEBUG_MUTEXES
1036 select DEBUG_LOCK_ALLOC
1037 select TRACE_IRQFLAGS
1040 This feature enables the kernel to prove that all locking
1041 that occurs in the kernel runtime is mathematically
1042 correct: that under no circumstance could an arbitrary (and
1043 not yet triggered) combination of observed locking
1044 sequences (on an arbitrary number of CPUs, running an
1045 arbitrary number of tasks and interrupt contexts) cause a
1048 In short, this feature enables the kernel to report locking
1049 related deadlocks before they actually occur.
1051 The proof does not depend on how hard and complex a
1052 deadlock scenario would be to trigger: how many
1053 participant CPUs, tasks and irq-contexts would be needed
1054 for it to trigger. The proof also does not depend on
1055 timing: if a race and a resulting deadlock is possible
1056 theoretically (no matter how unlikely the race scenario
1057 is), it will be proven so and will immediately be
1058 reported by the kernel (once the event is observed that
1059 makes the deadlock theoretically possible).
1061 If a deadlock is impossible (i.e. the locking rules, as
1062 observed by the kernel, are mathematically correct), the
1063 kernel reports nothing.
1065 NOTE: this feature can also be enabled for rwlocks, mutexes
1066 and rwsems - in which case all dependencies between these
1067 different locking variants are observed and mapped too, and
1068 the proof of observed correctness is also maintained for an
1069 arbitrary combination of these separate locking variants.
1071 For more details, see Documentation/locking/lockdep-design.txt.
1075 depends on DEBUG_KERNEL && TRACE_IRQFLAGS_SUPPORT && STACKTRACE_SUPPORT && LOCKDEP_SUPPORT
1077 select FRAME_POINTER if !MIPS && !PPC && !ARM_UNWIND && !S390 && !MICROBLAZE && !ARC && !SCORE
1082 bool "Lock usage statistics"
1083 depends on DEBUG_KERNEL && TRACE_IRQFLAGS_SUPPORT && STACKTRACE_SUPPORT && LOCKDEP_SUPPORT
1085 select DEBUG_SPINLOCK
1086 select DEBUG_MUTEXES
1087 select DEBUG_LOCK_ALLOC
1090 This feature enables tracking lock contention points
1092 For more details, see Documentation/locking/lockstat.txt
1094 This also enables lock events required by "perf lock",
1096 If you want to use "perf lock", you also need to turn on
1097 CONFIG_EVENT_TRACING.
1099 CONFIG_LOCK_STAT defines "contended" and "acquired" lock events.
1100 (CONFIG_LOCKDEP defines "acquire" and "release" events.)
1102 config DEBUG_LOCKDEP
1103 bool "Lock dependency engine debugging"
1104 depends on DEBUG_KERNEL && LOCKDEP
1106 If you say Y here, the lock dependency engine will do
1107 additional runtime checks to debug itself, at the price
1108 of more runtime overhead.
1110 config DEBUG_ATOMIC_SLEEP
1111 bool "Sleep inside atomic section checking"
1112 select PREEMPT_COUNT
1113 depends on DEBUG_KERNEL
1115 If you say Y here, various routines which may sleep will become very
1116 noisy if they are called inside atomic sections: when a spinlock is
1117 held, inside an rcu read side critical section, inside preempt disabled
1118 sections, inside an interrupt, etc...
1120 config DEBUG_LOCKING_API_SELFTESTS
1121 bool "Locking API boot-time self-tests"
1122 depends on DEBUG_KERNEL
1124 Say Y here if you want the kernel to run a short self-test during
1125 bootup. The self-test checks whether common types of locking bugs
1126 are detected by debugging mechanisms or not. (if you disable
1127 lock debugging then those bugs wont be detected of course.)
1128 The following locking APIs are covered: spinlocks, rwlocks,
1131 config LOCK_TORTURE_TEST
1132 tristate "torture tests for locking"
1133 depends on DEBUG_KERNEL
1137 This option provides a kernel module that runs torture tests
1138 on kernel locking primitives. The kernel module may be built
1139 after the fact on the running kernel to be tested, if desired.
1141 Say Y here if you want kernel locking-primitive torture tests
1142 to be built into the kernel.
1143 Say M if you want these torture tests to build as a module.
1144 Say N if you are unsure.
1146 endmenu # lock debugging
1148 config TRACE_IRQFLAGS
1151 Enables hooks to interrupt enabling and disabling for
1152 either tracing or lock debugging.
1155 bool "Stack backtrace support"
1156 depends on STACKTRACE_SUPPORT
1158 This option causes the kernel to create a /proc/pid/stack for
1159 every process, showing its current stack trace.
1160 It is also used by various kernel debugging features that require
1161 stack trace generation.
1163 config DEBUG_KOBJECT
1164 bool "kobject debugging"
1165 depends on DEBUG_KERNEL
1167 If you say Y here, some extra kobject debugging messages will be sent
1170 config DEBUG_KOBJECT_RELEASE
1171 bool "kobject release debugging"
1172 depends on DEBUG_OBJECTS_TIMERS
1174 kobjects are reference counted objects. This means that their
1175 last reference count put is not predictable, and the kobject can
1176 live on past the point at which a driver decides to drop it's
1177 initial reference to the kobject gained on allocation. An
1178 example of this would be a struct device which has just been
1181 However, some buggy drivers assume that after such an operation,
1182 the memory backing the kobject can be immediately freed. This
1183 goes completely against the principles of a refcounted object.
1185 If you say Y here, the kernel will delay the release of kobjects
1186 on the last reference count to improve the visibility of this
1187 kind of kobject release bug.
1189 config HAVE_DEBUG_BUGVERBOSE
1192 config DEBUG_BUGVERBOSE
1193 bool "Verbose BUG() reporting (adds 70K)" if DEBUG_KERNEL && EXPERT
1194 depends on BUG && (GENERIC_BUG || HAVE_DEBUG_BUGVERBOSE)
1197 Say Y here to make BUG() panics output the file name and line number
1198 of the BUG call as well as the EIP and oops trace. This aids
1199 debugging but costs about 70-100K of memory.
1202 bool "Debug linked list manipulation"
1203 depends on DEBUG_KERNEL
1205 Enable this to turn on extended checks in the linked-list
1210 config DEBUG_PI_LIST
1211 bool "Debug priority linked list manipulation"
1212 depends on DEBUG_KERNEL
1214 Enable this to turn on extended checks in the priority-ordered
1215 linked-list (plist) walking routines. This checks the entire
1216 list multiple times during each manipulation.
1221 bool "Debug SG table operations"
1222 depends on DEBUG_KERNEL
1224 Enable this to turn on checks on scatter-gather tables. This can
1225 help find problems with drivers that do not properly initialize
1230 config DEBUG_NOTIFIERS
1231 bool "Debug notifier call chains"
1232 depends on DEBUG_KERNEL
1234 Enable this to turn on sanity checking for notifier call chains.
1235 This is most useful for kernel developers to make sure that
1236 modules properly unregister themselves from notifier chains.
1237 This is a relatively cheap check but if you care about maximum
1240 config DEBUG_CREDENTIALS
1241 bool "Debug credential management"
1242 depends on DEBUG_KERNEL
1244 Enable this to turn on some debug checking for credential
1245 management. The additional code keeps track of the number of
1246 pointers from task_structs to any given cred struct, and checks to
1247 see that this number never exceeds the usage count of the cred
1250 Furthermore, if SELinux is enabled, this also checks that the
1251 security pointer in the cred struct is never seen to be invalid.
1255 menu "RCU Debugging"
1258 def_bool PROVE_LOCKING
1260 config PROVE_RCU_REPEATEDLY
1261 bool "RCU debugging: don't disable PROVE_RCU on first splat"
1262 depends on PROVE_RCU
1265 By itself, PROVE_RCU will disable checking upon issuing the
1266 first warning (or "splat"). This feature prevents such
1267 disabling, allowing multiple RCU-lockdep warnings to be printed
1270 Say Y to allow multiple RCU-lockdep warnings per boot.
1272 Say N if you are unsure.
1274 config SPARSE_RCU_POINTER
1275 bool "RCU debugging: sparse-based checks for pointer usage"
1278 This feature enables the __rcu sparse annotation for
1279 RCU-protected pointers. This annotation will cause sparse
1280 to flag any non-RCU used of annotated pointers. This can be
1281 helpful when debugging RCU usage. Please note that this feature
1282 is not intended to enforce code cleanliness; it is instead merely
1285 Say Y to make sparse flag questionable use of RCU-protected pointers
1287 Say N if you are unsure.
1293 config RCU_PERF_TEST
1294 tristate "performance tests for RCU"
1295 depends on DEBUG_KERNEL
1301 This option provides a kernel module that runs performance
1302 tests on the RCU infrastructure. The kernel module may be built
1303 after the fact on the running kernel to be tested, if desired.
1305 Say Y here if you want RCU performance tests to be built into
1307 Say M if you want the RCU performance tests to build as a module.
1308 Say N if you are unsure.
1310 config RCU_PERF_TEST_RUNNABLE
1311 bool "performance tests for RCU runnable by default"
1312 depends on RCU_PERF_TEST = y
1315 This option provides a way to build the RCU performance tests
1316 directly into the kernel without them starting up at boot time.
1317 You can use /sys/module to manually override this setting.
1318 This /proc file is available only when the RCU performance
1319 tests have been built into the kernel.
1321 Say Y here if you want the RCU performance tests to start during
1322 boot (you probably don't).
1323 Say N here if you want the RCU performance tests to start only
1324 after being manually enabled via /sys/module.
1326 config RCU_TORTURE_TEST
1327 tristate "torture tests for RCU"
1328 depends on DEBUG_KERNEL
1334 This option provides a kernel module that runs torture tests
1335 on the RCU infrastructure. The kernel module may be built
1336 after the fact on the running kernel to be tested, if desired.
1338 Say Y here if you want RCU torture tests to be built into
1340 Say M if you want the RCU torture tests to build as a module.
1341 Say N if you are unsure.
1343 config RCU_TORTURE_TEST_RUNNABLE
1344 bool "torture tests for RCU runnable by default"
1345 depends on RCU_TORTURE_TEST = y
1348 This option provides a way to build the RCU torture tests
1349 directly into the kernel without them starting up at boot
1350 time. You can use /proc/sys/kernel/rcutorture_runnable
1351 to manually override this setting. This /proc file is
1352 available only when the RCU torture tests have been built
1355 Say Y here if you want the RCU torture tests to start during
1356 boot (you probably don't).
1357 Say N here if you want the RCU torture tests to start only
1358 after being manually enabled via /proc.
1360 config RCU_TORTURE_TEST_SLOW_PREINIT
1361 bool "Slow down RCU grace-period pre-initialization to expose races"
1362 depends on RCU_TORTURE_TEST
1364 This option delays grace-period pre-initialization (the
1365 propagation of CPU-hotplug changes up the rcu_node combining
1366 tree) for a few jiffies between initializing each pair of
1367 consecutive rcu_node structures. This helps to expose races
1368 involving grace-period pre-initialization, in other words, it
1369 makes your kernel less stable. It can also greatly increase
1370 grace-period latency, especially on systems with large numbers
1371 of CPUs. This is useful when torture-testing RCU, but in
1372 almost no other circumstance.
1374 Say Y here if you want your system to crash and hang more often.
1375 Say N if you want a sane system.
1377 config RCU_TORTURE_TEST_SLOW_PREINIT_DELAY
1378 int "How much to slow down RCU grace-period pre-initialization"
1381 depends on RCU_TORTURE_TEST_SLOW_PREINIT
1383 This option specifies the number of jiffies to wait between
1384 each rcu_node structure pre-initialization step.
1386 config RCU_TORTURE_TEST_SLOW_INIT
1387 bool "Slow down RCU grace-period initialization to expose races"
1388 depends on RCU_TORTURE_TEST
1390 This option delays grace-period initialization for a few
1391 jiffies between initializing each pair of consecutive
1392 rcu_node structures. This helps to expose races involving
1393 grace-period initialization, in other words, it makes your
1394 kernel less stable. It can also greatly increase grace-period
1395 latency, especially on systems with large numbers of CPUs.
1396 This is useful when torture-testing RCU, but in almost no
1399 Say Y here if you want your system to crash and hang more often.
1400 Say N if you want a sane system.
1402 config RCU_TORTURE_TEST_SLOW_INIT_DELAY
1403 int "How much to slow down RCU grace-period initialization"
1406 depends on RCU_TORTURE_TEST_SLOW_INIT
1408 This option specifies the number of jiffies to wait between
1409 each rcu_node structure initialization.
1411 config RCU_TORTURE_TEST_SLOW_CLEANUP
1412 bool "Slow down RCU grace-period cleanup to expose races"
1413 depends on RCU_TORTURE_TEST
1415 This option delays grace-period cleanup for a few jiffies
1416 between cleaning up each pair of consecutive rcu_node
1417 structures. This helps to expose races involving grace-period
1418 cleanup, in other words, it makes your kernel less stable.
1419 It can also greatly increase grace-period latency, especially
1420 on systems with large numbers of CPUs. This is useful when
1421 torture-testing RCU, but in almost no other circumstance.
1423 Say Y here if you want your system to crash and hang more often.
1424 Say N if you want a sane system.
1426 config RCU_TORTURE_TEST_SLOW_CLEANUP_DELAY
1427 int "How much to slow down RCU grace-period cleanup"
1430 depends on RCU_TORTURE_TEST_SLOW_CLEANUP
1432 This option specifies the number of jiffies to wait between
1433 each rcu_node structure cleanup operation.
1435 config RCU_CPU_STALL_TIMEOUT
1436 int "RCU CPU stall timeout in seconds"
1437 depends on RCU_STALL_COMMON
1441 If a given RCU grace period extends more than the specified
1442 number of seconds, a CPU stall warning is printed. If the
1443 RCU grace period persists, additional CPU stall warnings are
1444 printed at more widely spaced intervals.
1447 bool "Enable tracing for RCU"
1448 depends on DEBUG_KERNEL
1451 This option provides tracing in RCU which presents stats
1452 in debugfs for debugging RCU implementation.
1454 Say Y here if you want to enable RCU tracing
1455 Say N if you are unsure.
1457 config RCU_EQS_DEBUG
1458 bool "Provide debugging asserts for adding NO_HZ support to an arch"
1459 depends on DEBUG_KERNEL
1461 This option provides consistency checks in RCU's handling of
1462 NO_HZ. These checks have proven quite helpful in detecting
1463 bugs in arch-specific NO_HZ code.
1465 Say N here if you need ultimate kernel/user switch latencies
1466 Say Y if you are unsure
1468 endmenu # "RCU Debugging"
1470 config DEBUG_WQ_FORCE_RR_CPU
1471 bool "Force round-robin CPU selection for unbound work items"
1472 depends on DEBUG_KERNEL
1475 Workqueue used to implicitly guarantee that work items queued
1476 without explicit CPU specified are put on the local CPU. This
1477 guarantee is no longer true and while local CPU is still
1478 preferred work items may be put on foreign CPUs. Kernel
1479 parameter "workqueue.debug_force_rr_cpu" is added to force
1480 round-robin CPU selection to flush out usages which depend on the
1481 now broken guarantee. This config option enables the debug
1482 feature by default. When enabled, memory and cache locality will
1485 config DEBUG_BLOCK_EXT_DEVT
1486 bool "Force extended block device numbers and spread them"
1487 depends on DEBUG_KERNEL
1491 BIG FAT WARNING: ENABLING THIS OPTION MIGHT BREAK BOOTING ON
1492 SOME DISTRIBUTIONS. DO NOT ENABLE THIS UNLESS YOU KNOW WHAT
1493 YOU ARE DOING. Distros, please enable this and fix whatever
1496 Conventionally, block device numbers are allocated from
1497 predetermined contiguous area. However, extended block area
1498 may introduce non-contiguous block device numbers. This
1499 option forces most block device numbers to be allocated from
1500 the extended space and spreads them to discover kernel or
1501 userland code paths which assume predetermined contiguous
1502 device number allocation.
1504 Note that turning on this debug option shuffles all the
1505 device numbers for all IDE and SCSI devices including libata
1506 ones, so root partition specified using device number
1507 directly (via rdev or root=MAJ:MIN) won't work anymore.
1508 Textual device names (root=/dev/sdXn) will continue to work.
1510 Say N if you are unsure.
1512 config CPU_HOTPLUG_STATE_CONTROL
1513 bool "Enable CPU hotplug state control"
1514 depends on DEBUG_KERNEL
1515 depends on HOTPLUG_CPU
1518 Allows to write steps between "offline" and "online" to the CPUs
1519 sysfs target file so states can be stepped granular. This is a debug
1520 option for now as the hotplug machinery cannot be stopped and
1521 restarted at arbitrary points yet.
1523 Say N if your are unsure.
1525 config NOTIFIER_ERROR_INJECTION
1526 tristate "Notifier error injection"
1527 depends on DEBUG_KERNEL
1530 This option provides the ability to inject artificial errors to
1531 specified notifier chain callbacks. It is useful to test the error
1532 handling of notifier call chain failures.
1536 config CPU_NOTIFIER_ERROR_INJECT
1537 tristate "CPU notifier error injection module"
1538 depends on HOTPLUG_CPU && NOTIFIER_ERROR_INJECTION
1540 This option provides a kernel module that can be used to test
1541 the error handling of the cpu notifiers by injecting artificial
1542 errors to CPU notifier chain callbacks. It is controlled through
1543 debugfs interface under /sys/kernel/debug/notifier-error-inject/cpu
1545 If the notifier call chain should be failed with some events
1546 notified, write the error code to "actions/<notifier event>/error".
1548 Example: Inject CPU offline error (-1 == -EPERM)
1550 # cd /sys/kernel/debug/notifier-error-inject/cpu
1551 # echo -1 > actions/CPU_DOWN_PREPARE/error
1552 # echo 0 > /sys/devices/system/cpu/cpu1/online
1553 bash: echo: write error: Operation not permitted
1555 To compile this code as a module, choose M here: the module will
1556 be called cpu-notifier-error-inject.
1560 config PM_NOTIFIER_ERROR_INJECT
1561 tristate "PM notifier error injection module"
1562 depends on PM && NOTIFIER_ERROR_INJECTION
1563 default m if PM_DEBUG
1565 This option provides the ability to inject artificial errors to
1566 PM notifier chain callbacks. It is controlled through debugfs
1567 interface /sys/kernel/debug/notifier-error-inject/pm
1569 If the notifier call chain should be failed with some events
1570 notified, write the error code to "actions/<notifier event>/error".
1572 Example: Inject PM suspend error (-12 = -ENOMEM)
1574 # cd /sys/kernel/debug/notifier-error-inject/pm/
1575 # echo -12 > actions/PM_SUSPEND_PREPARE/error
1576 # echo mem > /sys/power/state
1577 bash: echo: write error: Cannot allocate memory
1579 To compile this code as a module, choose M here: the module will
1580 be called pm-notifier-error-inject.
1584 config OF_RECONFIG_NOTIFIER_ERROR_INJECT
1585 tristate "OF reconfig notifier error injection module"
1586 depends on OF_DYNAMIC && NOTIFIER_ERROR_INJECTION
1588 This option provides the ability to inject artificial errors to
1589 OF reconfig notifier chain callbacks. It is controlled
1590 through debugfs interface under
1591 /sys/kernel/debug/notifier-error-inject/OF-reconfig/
1593 If the notifier call chain should be failed with some events
1594 notified, write the error code to "actions/<notifier event>/error".
1596 To compile this code as a module, choose M here: the module will
1597 be called of-reconfig-notifier-error-inject.
1601 config NETDEV_NOTIFIER_ERROR_INJECT
1602 tristate "Netdev notifier error injection module"
1603 depends on NET && NOTIFIER_ERROR_INJECTION
1605 This option provides the ability to inject artificial errors to
1606 netdevice notifier chain callbacks. It is controlled through debugfs
1607 interface /sys/kernel/debug/notifier-error-inject/netdev
1609 If the notifier call chain should be failed with some events
1610 notified, write the error code to "actions/<notifier event>/error".
1612 Example: Inject netdevice mtu change error (-22 = -EINVAL)
1614 # cd /sys/kernel/debug/notifier-error-inject/netdev
1615 # echo -22 > actions/NETDEV_CHANGEMTU/error
1616 # ip link set eth0 mtu 1024
1617 RTNETLINK answers: Invalid argument
1619 To compile this code as a module, choose M here: the module will
1620 be called netdev-notifier-error-inject.
1624 config FAULT_INJECTION
1625 bool "Fault-injection framework"
1626 depends on DEBUG_KERNEL
1628 Provide fault-injection framework.
1629 For more details, see Documentation/fault-injection/.
1632 bool "Fault-injection capability for kmalloc"
1633 depends on FAULT_INJECTION
1634 depends on SLAB || SLUB
1636 Provide fault-injection capability for kmalloc.
1638 config FAIL_PAGE_ALLOC
1639 bool "Fault-injection capabilitiy for alloc_pages()"
1640 depends on FAULT_INJECTION
1642 Provide fault-injection capability for alloc_pages().
1644 config FAIL_MAKE_REQUEST
1645 bool "Fault-injection capability for disk IO"
1646 depends on FAULT_INJECTION && BLOCK
1648 Provide fault-injection capability for disk IO.
1650 config FAIL_IO_TIMEOUT
1651 bool "Fault-injection capability for faking disk interrupts"
1652 depends on FAULT_INJECTION && BLOCK
1654 Provide fault-injection capability on end IO handling. This
1655 will make the block layer "forget" an interrupt as configured,
1656 thus exercising the error handling.
1658 Only works with drivers that use the generic timeout handling,
1659 for others it wont do anything.
1661 config FAIL_MMC_REQUEST
1662 bool "Fault-injection capability for MMC IO"
1663 depends on FAULT_INJECTION_DEBUG_FS && MMC
1665 Provide fault-injection capability for MMC IO.
1666 This will make the mmc core return data errors. This is
1667 useful to test the error handling in the mmc block device
1668 and to test how the mmc host driver handles retries from
1672 bool "Fault-injection capability for futexes"
1674 depends on FAULT_INJECTION && FUTEX
1676 Provide fault-injection capability for futexes.
1678 config FAULT_INJECTION_DEBUG_FS
1679 bool "Debugfs entries for fault-injection capabilities"
1680 depends on FAULT_INJECTION && SYSFS && DEBUG_FS
1682 Enable configuration of fault-injection capabilities via debugfs.
1684 config FAULT_INJECTION_STACKTRACE_FILTER
1685 bool "stacktrace filter for fault-injection capabilities"
1686 depends on FAULT_INJECTION_DEBUG_FS && STACKTRACE_SUPPORT
1689 select FRAME_POINTER if !MIPS && !PPC && !S390 && !MICROBLAZE && !ARM_UNWIND && !ARC && !SCORE
1691 Provide stacktrace filter for fault-injection capabilities
1694 bool "Latency measuring infrastructure"
1695 depends on DEBUG_KERNEL
1696 depends on STACKTRACE_SUPPORT
1698 select FRAME_POINTER if !MIPS && !PPC && !S390 && !MICROBLAZE && !ARM_UNWIND && !ARC
1705 Enable this option if you want to use the LatencyTOP tool
1706 to find out which userspace is blocking on what kernel operations.
1708 config ARCH_HAS_DEBUG_STRICT_USER_COPY_CHECKS
1711 config DEBUG_STRICT_USER_COPY_CHECKS
1712 bool "Strict user copy size checks"
1713 depends on ARCH_HAS_DEBUG_STRICT_USER_COPY_CHECKS
1714 depends on DEBUG_KERNEL && !TRACE_BRANCH_PROFILING
1716 Enabling this option turns a certain set of sanity checks for user
1717 copy operations into compile time failures.
1719 The copy_from_user() etc checks are there to help test if there
1720 are sufficient security checks on the length argument of
1721 the copy operation, by having gcc prove that the argument is
1726 source kernel/trace/Kconfig
1728 menu "Runtime Testing"
1731 tristate "Linux Kernel Dump Test Tool Module"
1736 This module enables testing of the different dumping mechanisms by
1737 inducing system failures at predefined crash points.
1738 If you don't need it: say N
1739 Choose M here to compile this code as a module. The module will be
1742 Documentation on how to use the module can be found in
1743 Documentation/fault-injection/provoke-crashes.txt
1745 config TEST_LIST_SORT
1746 bool "Linked list sorting test"
1747 depends on DEBUG_KERNEL
1749 Enable this to turn on 'list_sort()' function test. This test is
1750 executed only once during system boot, so affects only boot time.
1754 config KPROBES_SANITY_TEST
1755 bool "Kprobes sanity tests"
1756 depends on DEBUG_KERNEL
1760 This option provides for testing basic kprobes functionality on
1761 boot. A sample kprobe, jprobe and kretprobe are inserted and
1762 verified for functionality.
1764 Say N if you are unsure.
1766 config BACKTRACE_SELF_TEST
1767 tristate "Self test for the backtrace code"
1768 depends on DEBUG_KERNEL
1771 This option provides a kernel module that can be used to test
1772 the kernel stack backtrace code. This option is not useful
1773 for distributions or general kernels, but only for kernel
1774 developers working on architecture code.
1776 Note that if you want to also test saved backtraces, you will
1777 have to enable STACKTRACE as well.
1779 Say N if you are unsure.
1782 tristate "Red-Black tree test"
1783 depends on DEBUG_KERNEL
1785 A benchmark measuring the performance of the rbtree library.
1786 Also includes rbtree invariant checks.
1788 config INTERVAL_TREE_TEST
1789 tristate "Interval tree test"
1790 depends on m && DEBUG_KERNEL
1791 select INTERVAL_TREE
1793 A benchmark measuring the performance of the interval tree library
1796 tristate "Per cpu operations test"
1797 depends on m && DEBUG_KERNEL
1799 Enable this option to build test module which validates per-cpu
1804 config ATOMIC64_SELFTEST
1805 bool "Perform an atomic64_t self-test at boot"
1807 Enable this option to test the atomic64_t functions at boot.
1811 config ASYNC_RAID6_TEST
1812 tristate "Self test for hardware accelerated raid6 recovery"
1813 depends on ASYNC_RAID6_RECOV
1816 This is a one-shot self test that permutes through the
1817 recovery of all the possible two disk failure scenarios for a
1818 N-disk array. Recovery is performed with the asynchronous
1819 raid6 recovery routines, and will optionally use an offload
1820 engine if one is available.
1825 tristate "Test functions located in the hexdump module at runtime"
1827 config TEST_STRING_HELPERS
1828 tristate "Test functions located in the string_helpers module at runtime"
1831 tristate "Test kstrto*() family of functions at runtime"
1834 tristate "Test printf() family of functions at runtime"
1837 tristate "Test bitmap_*() family of functions at runtime"
1840 Enable this option to test the bitmap functions at boot.
1845 tristate "Test functions located in the uuid module at runtime"
1847 config TEST_RHASHTABLE
1848 tristate "Perform selftest on resizable hash table"
1851 Enable this option to test the rhashtable functions at boot.
1856 tristate "Perform selftest on hash functions"
1859 Enable this option to test the kernel's integer (<linux/hash,h>)
1860 and string (<linux/stringhash.h>) hash functions on boot
1863 This is intended to help people writing architecture-specific
1864 optimized versions. If unsure, say N.
1866 endmenu # runtime tests
1868 config PROVIDE_OHCI1394_DMA_INIT
1869 bool "Remote debugging over FireWire early on boot"
1870 depends on PCI && X86
1872 If you want to debug problems which hang or crash the kernel early
1873 on boot and the crashing machine has a FireWire port, you can use
1874 this feature to remotely access the memory of the crashed machine
1875 over FireWire. This employs remote DMA as part of the OHCI1394
1876 specification which is now the standard for FireWire controllers.
1878 With remote DMA, you can monitor the printk buffer remotely using
1879 firescope and access all memory below 4GB using fireproxy from gdb.
1880 Even controlling a kernel debugger is possible using remote DMA.
1884 If ohci1394_dma=early is used as boot parameter, it will initialize
1885 all OHCI1394 controllers which are found in the PCI config space.
1887 As all changes to the FireWire bus such as enabling and disabling
1888 devices cause a bus reset and thereby disable remote DMA for all
1889 devices, be sure to have the cable plugged and FireWire enabled on
1890 the debugging host before booting the debug target for debugging.
1892 This code (~1k) is freed after boot. By then, the firewire stack
1893 in charge of the OHCI-1394 controllers should be used instead.
1895 See Documentation/debugging-via-ohci1394.txt for more information.
1898 bool "Build targets in Documentation/ tree"
1899 depends on HEADERS_CHECK
1901 This option attempts to build objects from the source files in the
1902 kernel Documentation/ tree.
1904 Say N if you are unsure.
1906 config DMA_API_DEBUG
1907 bool "Enable debugging of DMA-API usage"
1908 depends on HAVE_DMA_API_DEBUG
1910 Enable this option to debug the use of the DMA API by device drivers.
1911 With this option you will be able to detect common bugs in device
1912 drivers like double-freeing of DMA mappings or freeing mappings that
1913 were never allocated.
1915 This also attempts to catch cases where a page owned by DMA is
1916 accessed by the cpu in a way that could cause data corruption. For
1917 example, this enables cow_user_page() to check that the source page is
1920 This option causes a performance degradation. Use only if you want to
1921 debug device drivers and dma interactions.
1926 tristate "Test module loading with 'hello world' module"
1930 This builds the "test_module" module that emits "Hello, world"
1931 on printk when loaded. It is designed to be used for basic
1932 evaluation of the module loading subsystem (for example when
1933 validating module verification). It lacks any extra dependencies,
1934 and will not normally be loaded by the system unless explicitly
1939 config TEST_USER_COPY
1940 tristate "Test user/kernel boundary protections"
1944 This builds the "test_user_copy" module that runs sanity checks
1945 on the copy_to/from_user infrastructure, making sure basic
1946 user/kernel boundary testing is working. If it fails to load,
1947 a regression has been detected in the user/kernel memory boundary
1953 tristate "Test BPF filter functionality"
1957 This builds the "test_bpf" module that runs various test vectors
1958 against the BPF interpreter or BPF JIT compiler depending on the
1959 current setting. This is in particular useful for BPF JIT compiler
1960 development, but also to run regression tests against changes in
1961 the interpreter code. It also enables test stubs for eBPF maps and
1962 verifier used by user space verifier testsuite.
1966 config TEST_FIRMWARE
1967 tristate "Test firmware loading via userspace interface"
1969 depends on FW_LOADER
1971 This builds the "test_firmware" module that creates a userspace
1972 interface for testing firmware loading. This can be used to
1973 control the triggering of firmware loading without needing an
1974 actual firmware-using device. The contents can be rechecked by
1980 tristate "udelay test driver"
1983 This builds the "udelay_test" module that helps to make sure
1984 that udelay() is working properly.
1990 depends on HAVE_MEMBLOCK
1992 This option adds a kernel parameter 'memtest', which allows memtest
1994 memtest=0, mean disabled; -- default
1995 memtest=1, mean do 1 test pattern;
1997 memtest=17, mean do 17 test patterns.
1998 If you are unsure how to answer this question, answer N.
2000 config TEST_STATIC_KEYS
2001 tristate "Test static keys"
2005 Test the static key interfaces.
2009 source "samples/Kconfig"
2011 source "lib/Kconfig.kgdb"
2013 source "lib/Kconfig.ubsan"
2015 config ARCH_HAS_DEVMEM_IS_ALLOWED
2018 config STRICT_DEVMEM
2019 bool "Filter access to /dev/mem"
2021 depends on ARCH_HAS_DEVMEM_IS_ALLOWED
2022 default y if TILE || PPC
2024 If this option is disabled, you allow userspace (root) access to all
2025 of memory, including kernel and userspace memory. Accidental
2026 access to this is obviously disastrous, but specific access can
2027 be used by people debugging the kernel. Note that with PAT support
2028 enabled, even in this case there are restrictions on /dev/mem
2029 use due to the cache aliasing requirements.
2031 If this option is switched on, and IO_STRICT_DEVMEM=n, the /dev/mem
2032 file only allows userspace access to PCI space and the BIOS code and
2033 data regions. This is sufficient for dosemu and X and all common
2038 config IO_STRICT_DEVMEM
2039 bool "Filter I/O access to /dev/mem"
2040 depends on STRICT_DEVMEM
2042 If this option is disabled, you allow userspace (root) access to all
2043 io-memory regardless of whether a driver is actively using that
2044 range. Accidental access to this is obviously disastrous, but
2045 specific access can be used by people debugging kernel drivers.
2047 If this option is switched on, the /dev/mem file only allows
2048 userspace access to *idle* io-memory ranges (see /proc/iomem) This
2049 may break traditional users of /dev/mem (dosemu, legacy X, etc...)
2050 if the driver using a given range cannot be disabled.