1 # SPDX-License-Identifier: GPL-2.0-only
4 menu "printk and dmesg options"
7 bool "Show timing information on printks"
10 Selecting this option causes time stamps of the printk()
11 messages to be added to the output of the syslog() system
12 call and at the console.
14 The timestamp is always recorded internally, and exported
15 to /dev/kmsg. This flag just specifies if the timestamp should
16 be included, not that the timestamp is recorded.
18 The behavior is also controlled by the kernel command line
19 parameter printk.time=1. See Documentation/admin-guide/kernel-parameters.rst
22 bool "Show caller information on printks"
25 Selecting this option causes printk() to add a caller "thread id" (if
26 in task context) or a caller "processor id" (if not in task context)
29 This option is intended for environments where multiple threads
30 concurrently call printk() for many times, for it is difficult to
31 interpret without knowing where these lines (or sometimes individual
32 line which was divided into multiple lines due to race) came from.
34 Since toggling after boot makes the code racy, currently there is
35 no option to enable/disable at the kernel command line parameter or
38 config CONSOLE_LOGLEVEL_DEFAULT
39 int "Default console loglevel (1-15)"
43 Default loglevel to determine what will be printed on the console.
45 Setting a default here is equivalent to passing in loglevel=<x> in
46 the kernel bootargs. loglevel=<x> continues to override whatever
47 value is specified here as well.
49 Note: This does not affect the log level of un-prefixed printk()
50 usage in the kernel. That is controlled by the MESSAGE_LOGLEVEL_DEFAULT
53 config CONSOLE_LOGLEVEL_QUIET
54 int "quiet console loglevel (1-15)"
58 loglevel to use when "quiet" is passed on the kernel commandline.
60 When "quiet" is passed on the kernel commandline this loglevel
61 will be used as the loglevel. IOW passing "quiet" will be the
62 equivalent of passing "loglevel=<CONSOLE_LOGLEVEL_QUIET>"
64 config MESSAGE_LOGLEVEL_DEFAULT
65 int "Default message log level (1-7)"
69 Default log level for printk statements with no specified priority.
71 This was hard-coded to KERN_WARNING since at least 2.6.10 but folks
72 that are auditing their logs closely may want to set it to a lower
75 Note: This does not affect what message level gets printed on the console
76 by default. To change that, use loglevel=<x> in the kernel bootargs,
77 or pick a different CONSOLE_LOGLEVEL_DEFAULT configuration value.
79 config BOOT_PRINTK_DELAY
80 bool "Delay each boot printk message by N milliseconds"
81 depends on DEBUG_KERNEL && PRINTK && GENERIC_CALIBRATE_DELAY
83 This build option allows you to read kernel boot messages
84 by inserting a short delay after each one. The delay is
85 specified in milliseconds on the kernel command line,
88 It is likely that you would also need to use "lpj=M" to preset
89 the "loops per jiffie" value.
90 See a previous boot log for the "lpj" value to use for your
91 system, and then set "lpj=M" before setting "boot_delay=N".
92 NOTE: Using this option may adversely affect SMP systems.
93 I.e., processors other than the first one may not boot up.
94 BOOT_PRINTK_DELAY also may cause LOCKUP_DETECTOR to detect
95 what it believes to be lockup conditions.
98 bool "Enable dynamic printk() support"
104 Compiles debug level messages into the kernel, which would not
105 otherwise be available at runtime. These messages can then be
106 enabled/disabled based on various levels of scope - per source file,
107 function, module, format string, and line number. This mechanism
108 implicitly compiles in all pr_debug() and dev_dbg() calls, which
109 enlarges the kernel text size by about 2%.
111 If a source file is compiled with DEBUG flag set, any
112 pr_debug() calls in it are enabled by default, but can be
113 disabled at runtime as below. Note that DEBUG flag is
114 turned on by many CONFIG_*DEBUG* options.
118 Dynamic debugging is controlled via the 'dynamic_debug/control' file,
119 which is contained in the 'debugfs' filesystem. Thus, the debugfs
120 filesystem must first be mounted before making use of this feature.
121 We refer the control file as: <debugfs>/dynamic_debug/control. This
122 file contains a list of the debug statements that can be enabled. The
123 format for each line of the file is:
125 filename:lineno [module]function flags format
127 filename : source file of the debug statement
128 lineno : line number of the debug statement
129 module : module that contains the debug statement
130 function : function that contains the debug statement
131 flags : '=p' means the line is turned 'on' for printing
132 format : the format used for the debug statement
136 nullarbor:~ # cat <debugfs>/dynamic_debug/control
137 # filename:lineno [module]function flags format
138 fs/aio.c:222 [aio]__put_ioctx =_ "__put_ioctx:\040freeing\040%p\012"
139 fs/aio.c:248 [aio]ioctx_alloc =_ "ENOMEM:\040nr_events\040too\040high\012"
140 fs/aio.c:1770 [aio]sys_io_cancel =_ "calling\040cancel\012"
144 // enable the message at line 1603 of file svcsock.c
145 nullarbor:~ # echo -n 'file svcsock.c line 1603 +p' >
146 <debugfs>/dynamic_debug/control
148 // enable all the messages in file svcsock.c
149 nullarbor:~ # echo -n 'file svcsock.c +p' >
150 <debugfs>/dynamic_debug/control
152 // enable all the messages in the NFS server module
153 nullarbor:~ # echo -n 'module nfsd +p' >
154 <debugfs>/dynamic_debug/control
156 // enable all 12 messages in the function svc_process()
157 nullarbor:~ # echo -n 'func svc_process +p' >
158 <debugfs>/dynamic_debug/control
160 // disable all 12 messages in the function svc_process()
161 nullarbor:~ # echo -n 'func svc_process -p' >
162 <debugfs>/dynamic_debug/control
164 See Documentation/admin-guide/dynamic-debug-howto.rst for additional
167 config SYMBOLIC_ERRNAME
168 bool "Support symbolic error names in printf"
171 If you say Y here, the kernel's printf implementation will
172 be able to print symbolic error names such as ENOSPC instead
173 of the number 28. It makes the kernel image slightly larger
174 (about 3KB), but can make the kernel logs easier to read.
176 endmenu # "printk and dmesg options"
178 menu "Compile-time checks and compiler options"
181 bool "Compile the kernel with debug info"
182 depends on DEBUG_KERNEL && !COMPILE_TEST
184 If you say Y here the resulting kernel image will include
185 debugging info resulting in a larger kernel image.
186 This adds debug symbols to the kernel and modules (gcc -g), and
187 is needed if you intend to use kernel crashdump or binary object
188 tools like crash, kgdb, LKCD, gdb, etc on the kernel.
189 Say Y here only if you plan to debug the kernel.
193 config DEBUG_INFO_REDUCED
194 bool "Reduce debugging information"
195 depends on DEBUG_INFO
197 If you say Y here gcc is instructed to generate less debugging
198 information for structure types. This means that tools that
199 need full debugging information (like kgdb or systemtap) won't
200 be happy. But if you merely need debugging information to
201 resolve line numbers there is no loss. Advantage is that
202 build directory object sizes shrink dramatically over a full
203 DEBUG_INFO build and compile times are reduced too.
204 Only works with newer gcc versions.
206 config DEBUG_INFO_SPLIT
207 bool "Produce split debuginfo in .dwo files"
208 depends on DEBUG_INFO
209 depends on $(cc-option,-gsplit-dwarf)
211 Generate debug info into separate .dwo files. This significantly
212 reduces the build directory size for builds with DEBUG_INFO,
213 because it stores the information only once on disk in .dwo
214 files instead of multiple times in object files and executables.
215 In addition the debug information is also compressed.
217 Requires recent gcc (4.7+) and recent gdb/binutils.
218 Any tool that packages or reads debug information would need
219 to know about the .dwo files and include them.
220 Incompatible with older versions of ccache.
222 config DEBUG_INFO_DWARF4
223 bool "Generate dwarf4 debuginfo"
224 depends on DEBUG_INFO
225 depends on $(cc-option,-gdwarf-4)
227 Generate dwarf4 debug info. This requires recent versions
228 of gcc and gdb. It makes the debug information larger.
229 But it significantly improves the success of resolving
230 variables in gdb on optimized code.
232 config DEBUG_INFO_BTF
233 bool "Generate BTF typeinfo"
234 depends on DEBUG_INFO
236 Generate deduplicated BTF type information from DWARF debug info.
237 Turning this on expects presence of pahole tool, which will convert
238 DWARF type info into equivalent deduplicated BTF type info.
241 bool "Provide GDB scripts for kernel debugging"
242 depends on DEBUG_INFO
244 This creates the required links to GDB helper scripts in the
245 build directory. If you load vmlinux into gdb, the helper
246 scripts will be automatically imported by gdb as well, and
247 additional functions are available to analyze a Linux kernel
248 instance. See Documentation/dev-tools/gdb-kernel-debugging.rst
251 config ENABLE_MUST_CHECK
252 bool "Enable __must_check logic"
255 Enable the __must_check logic in the kernel build. Disable this to
256 suppress the "warning: ignoring return value of 'foo', declared with
257 attribute warn_unused_result" messages.
260 int "Warn for stack frames larger than (needs gcc 4.4)"
262 default 2048 if GCC_PLUGIN_LATENT_ENTROPY
263 default 1280 if (!64BIT && PARISC)
264 default 1024 if (!64BIT && !PARISC)
265 default 2048 if 64BIT
267 Tell gcc to warn at build time for stack frames larger than this.
268 Setting this too low will cause a lot of warnings.
269 Setting it to 0 disables the warning.
272 config STRIP_ASM_SYMS
273 bool "Strip assembler-generated symbols during link"
276 Strip internal assembler-generated symbols during a link (symbols
277 that look like '.Lxxx') so they don't pollute the output of
278 get_wchan() and suchlike.
281 bool "Generate readable assembler code"
282 depends on DEBUG_KERNEL
284 Disable some compiler optimizations that tend to generate human unreadable
285 assembler output. This may make the kernel slightly slower, but it helps
286 to keep kernel developers who have to stare a lot at assembler listings
290 bool "Debug Filesystem"
292 debugfs is a virtual file system that kernel developers use to put
293 debugging files into. Enable this option to be able to read and
294 write to these files.
296 For detailed documentation on the debugfs API, see
297 Documentation/filesystems/.
301 config HEADERS_INSTALL
302 bool "Install uapi headers to usr/include"
305 This option will install uapi headers (headers exported to user-space)
306 into the usr/include directory for use during the kernel build.
307 This is unneeded for building the kernel itself, but needed for some
308 user-space program samples. It is also needed by some features such
309 as uapi header sanity checks.
311 config OPTIMIZE_INLINING
314 This option determines if the kernel forces gcc to inline the functions
315 developers have marked 'inline'. Doing so takes away freedom from gcc to
316 do what it thinks is best, which is desirable for the gcc 3.x series of
317 compilers. The gcc 4.x series have a rewritten inlining algorithm and
318 enabling this option will generate a smaller kernel there. Hopefully
319 this algorithm is so good that allowing gcc 4.x and above to make the
320 decision will become the default in the future. Until then this option
321 is there to test gcc for this.
323 config DEBUG_SECTION_MISMATCH
324 bool "Enable full Section mismatch analysis"
326 The section mismatch analysis checks if there are illegal
327 references from one section to another section.
328 During linktime or runtime, some sections are dropped;
329 any use of code/data previously in these sections would
330 most likely result in an oops.
331 In the code, functions and variables are annotated with
332 __init,, etc. (see the full list in include/linux/init.h),
333 which results in the code/data being placed in specific sections.
334 The section mismatch analysis is always performed after a full
335 kernel build, and enabling this option causes the following
336 additional step to occur:
337 - Add the option -fno-inline-functions-called-once to gcc commands.
338 When inlining a function annotated with __init in a non-init
339 function, we would lose the section information and thus
340 the analysis would not catch the illegal reference.
341 This option tells gcc to inline less (but it does result in
344 config SECTION_MISMATCH_WARN_ONLY
345 bool "Make section mismatch errors non-fatal"
348 If you say N here, the build process will fail if there are any
349 section mismatch, instead of just throwing warnings.
354 # Select this config option from the architecture Kconfig, if it
355 # is preferred to always offer frame pointers as a config
356 # option on the architecture (regardless of KERNEL_DEBUG):
358 config ARCH_WANT_FRAME_POINTERS
362 bool "Compile the kernel with frame pointers"
363 depends on DEBUG_KERNEL && (M68K || UML || SUPERH) || ARCH_WANT_FRAME_POINTERS
364 default y if (DEBUG_INFO && UML) || ARCH_WANT_FRAME_POINTERS
366 If you say Y here the resulting kernel image will be slightly
367 larger and slower, but it gives very useful debugging information
368 in case of kernel bugs. (precise oopses/stacktraces/warnings)
370 config STACK_VALIDATION
371 bool "Compile-time stack metadata validation"
372 depends on HAVE_STACK_VALIDATION
375 Add compile-time checks to validate stack metadata, including frame
376 pointers (if CONFIG_FRAME_POINTER is enabled). This helps ensure
377 that runtime stack traces are more reliable.
379 This is also a prerequisite for generation of ORC unwind data, which
380 is needed for CONFIG_UNWINDER_ORC.
382 For more information, see
383 tools/objtool/Documentation/stack-validation.txt.
385 config DEBUG_FORCE_WEAK_PER_CPU
386 bool "Force weak per-cpu definitions"
387 depends on DEBUG_KERNEL
389 s390 and alpha require percpu variables in modules to be
390 defined weak to work around addressing range issue which
391 puts the following two restrictions on percpu variable
394 1. percpu symbols must be unique whether static or not
395 2. percpu variables can't be defined inside a function
397 To ensure that generic code follows the above rules, this
398 option forces all percpu variables to be defined as weak.
400 endmenu # "Compiler options"
402 menu "Generic Kernel Debugging Instruments"
405 bool "Magic SysRq key"
408 If you say Y here, you will have some control over the system even
409 if the system crashes for example during kernel debugging (e.g., you
410 will be able to flush the buffer cache to disk, reboot the system
411 immediately or dump some status information). This is accomplished
412 by pressing various keys while holding SysRq (Alt+PrintScreen). It
413 also works on a serial console (on PC hardware at least), if you
414 send a BREAK and then within 5 seconds a command keypress. The
415 keys are documented in <file:Documentation/admin-guide/sysrq.rst>.
416 Don't say Y unless you really know what this hack does.
418 config MAGIC_SYSRQ_DEFAULT_ENABLE
419 hex "Enable magic SysRq key functions by default"
420 depends on MAGIC_SYSRQ
423 Specifies which SysRq key functions are enabled by default.
424 This may be set to 1 or 0 to enable or disable them all, or
425 to a bitmask as described in Documentation/admin-guide/sysrq.rst.
427 config MAGIC_SYSRQ_SERIAL
428 bool "Enable magic SysRq key over serial"
429 depends on MAGIC_SYSRQ
432 Many embedded boards have a disconnected TTL level serial which can
433 generate some garbage that can lead to spurious false sysrq detects.
434 This option allows you to decide whether you want to enable the
437 source "lib/Kconfig.kgdb"
439 source "lib/Kconfig.ubsan"
444 bool "Kernel debugging"
446 Say Y here if you are developing drivers or trying to debug and
447 identify kernel problems.
450 bool "Miscellaneous debug code"
452 depends on DEBUG_KERNEL
454 Say Y here if you need to enable miscellaneous debug code that should
455 be under a more specific debug option but isn't.
458 menu "Memory Debugging"
460 source "mm/Kconfig.debug"
463 bool "Debug object operations"
464 depends on DEBUG_KERNEL
466 If you say Y here, additional code will be inserted into the
467 kernel to track the life time of various objects and validate
468 the operations on those objects.
470 config DEBUG_OBJECTS_SELFTEST
471 bool "Debug objects selftest"
472 depends on DEBUG_OBJECTS
474 This enables the selftest of the object debug code.
476 config DEBUG_OBJECTS_FREE
477 bool "Debug objects in freed memory"
478 depends on DEBUG_OBJECTS
480 This enables checks whether a k/v free operation frees an area
481 which contains an object which has not been deactivated
482 properly. This can make kmalloc/kfree-intensive workloads
485 config DEBUG_OBJECTS_TIMERS
486 bool "Debug timer objects"
487 depends on DEBUG_OBJECTS
489 If you say Y here, additional code will be inserted into the
490 timer routines to track the life time of timer objects and
491 validate the timer operations.
493 config DEBUG_OBJECTS_WORK
494 bool "Debug work objects"
495 depends on DEBUG_OBJECTS
497 If you say Y here, additional code will be inserted into the
498 work queue routines to track the life time of work objects and
499 validate the work operations.
501 config DEBUG_OBJECTS_RCU_HEAD
502 bool "Debug RCU callbacks objects"
503 depends on DEBUG_OBJECTS
505 Enable this to turn on debugging of RCU list heads (call_rcu() usage).
507 config DEBUG_OBJECTS_PERCPU_COUNTER
508 bool "Debug percpu counter objects"
509 depends on DEBUG_OBJECTS
511 If you say Y here, additional code will be inserted into the
512 percpu counter routines to track the life time of percpu counter
513 objects and validate the percpu counter operations.
515 config DEBUG_OBJECTS_ENABLE_DEFAULT
516 int "debug_objects bootup default value (0-1)"
519 depends on DEBUG_OBJECTS
521 Debug objects boot parameter default value
524 bool "Debug slab memory allocations"
525 depends on DEBUG_KERNEL && SLAB
527 Say Y here to have the kernel do limited verification on memory
528 allocation as well as poisoning memory on free to catch use of freed
529 memory. This can make kmalloc/kfree-intensive workloads much slower.
532 bool "SLUB debugging on by default"
533 depends on SLUB && SLUB_DEBUG
536 Boot with debugging on by default. SLUB boots by default with
537 the runtime debug capabilities switched off. Enabling this is
538 equivalent to specifying the "slub_debug" parameter on boot.
539 There is no support for more fine grained debug control like
540 possible with slub_debug=xxx. SLUB debugging may be switched
541 off in a kernel built with CONFIG_SLUB_DEBUG_ON by specifying
546 bool "Enable SLUB performance statistics"
547 depends on SLUB && SYSFS
549 SLUB statistics are useful to debug SLUBs allocation behavior in
550 order find ways to optimize the allocator. This should never be
551 enabled for production use since keeping statistics slows down
552 the allocator by a few percentage points. The slabinfo command
553 supports the determination of the most active slabs to figure
554 out which slabs are relevant to a particular load.
555 Try running: slabinfo -DA
557 config HAVE_DEBUG_KMEMLEAK
560 config DEBUG_KMEMLEAK
561 bool "Kernel memory leak detector"
562 depends on DEBUG_KERNEL && HAVE_DEBUG_KMEMLEAK
564 select STACKTRACE if STACKTRACE_SUPPORT
568 Say Y here if you want to enable the memory leak
569 detector. The memory allocation/freeing is traced in a way
570 similar to the Boehm's conservative garbage collector, the
571 difference being that the orphan objects are not freed but
572 only shown in /sys/kernel/debug/kmemleak. Enabling this
573 feature will introduce an overhead to memory
574 allocations. See Documentation/dev-tools/kmemleak.rst for more
577 Enabling DEBUG_SLAB or SLUB_DEBUG may increase the chances
578 of finding leaks due to the slab objects poisoning.
580 In order to access the kmemleak file, debugfs needs to be
581 mounted (usually at /sys/kernel/debug).
583 config DEBUG_KMEMLEAK_MEM_POOL_SIZE
584 int "Kmemleak memory pool size"
585 depends on DEBUG_KMEMLEAK
589 Kmemleak must track all the memory allocations to avoid
590 reporting false positives. Since memory may be allocated or
591 freed before kmemleak is fully initialised, use a static pool
592 of metadata objects to track such callbacks. After kmemleak is
593 fully initialised, this memory pool acts as an emergency one
594 if slab allocations fail.
596 config DEBUG_KMEMLEAK_TEST
597 tristate "Simple test for the kernel memory leak detector"
598 depends on DEBUG_KMEMLEAK && m
600 This option enables a module that explicitly leaks memory.
604 config DEBUG_KMEMLEAK_DEFAULT_OFF
605 bool "Default kmemleak to off"
606 depends on DEBUG_KMEMLEAK
608 Say Y here to disable kmemleak by default. It can then be enabled
609 on the command line via kmemleak=on.
611 config DEBUG_KMEMLEAK_AUTO_SCAN
612 bool "Enable kmemleak auto scan thread on boot up"
614 depends on DEBUG_KMEMLEAK
616 Depending on the cpu, kmemleak scan may be cpu intensive and can
617 stall user tasks at times. This option enables/disables automatic
618 kmemleak scan at boot up.
620 Say N here to disable kmemleak auto scan thread to stop automatic
621 scanning. Disabling this option disables automatic reporting of
626 config DEBUG_STACK_USAGE
627 bool "Stack utilization instrumentation"
628 depends on DEBUG_KERNEL && !IA64
630 Enables the display of the minimum amount of free stack which each
631 task has ever had available in the sysrq-T and sysrq-P debug output.
633 This option will slow down process creation somewhat.
635 config SCHED_STACK_END_CHECK
636 bool "Detect stack corruption on calls to schedule()"
637 depends on DEBUG_KERNEL
640 This option checks for a stack overrun on calls to schedule().
641 If the stack end location is found to be over written always panic as
642 the content of the corrupted region can no longer be trusted.
643 This is to ensure no erroneous behaviour occurs which could result in
644 data corruption or a sporadic crash at a later stage once the region
645 is examined. The runtime overhead introduced is minimal.
649 depends on DEBUG_KERNEL
651 Enable this to turn on extended checks in the virtual-memory system
652 that may impact performance.
656 config DEBUG_VM_VMACACHE
657 bool "Debug VMA caching"
660 Enable this to turn on VMA caching debug information. Doing so
661 can cause significant overhead, so only enable it in non-production
667 bool "Debug VM red-black trees"
670 Enable VM red-black tree debugging information and extra validations.
674 config DEBUG_VM_PGFLAGS
675 bool "Debug page-flags operations"
678 Enables extra validation on page flags operations.
682 config ARCH_HAS_DEBUG_VIRTUAL
686 bool "Debug VM translations"
687 depends on DEBUG_KERNEL && ARCH_HAS_DEBUG_VIRTUAL
689 Enable some costly sanity checks in virtual to page code. This can
690 catch mistakes with virt_to_page() and friends.
694 config DEBUG_NOMMU_REGIONS
695 bool "Debug the global anon/private NOMMU mapping region tree"
696 depends on DEBUG_KERNEL && !MMU
698 This option causes the global tree of anonymous and private mapping
699 regions to be regularly checked for invalid topology.
701 config DEBUG_MEMORY_INIT
702 bool "Debug memory initialisation" if EXPERT
705 Enable this for additional checks during memory initialisation.
706 The sanity checks verify aspects of the VM such as the memory model
707 and other information provided by the architecture. Verbose
708 information will be printed at KERN_DEBUG loglevel depending
709 on the mminit_loglevel= command-line option.
713 config MEMORY_NOTIFIER_ERROR_INJECT
714 tristate "Memory hotplug notifier error injection module"
715 depends on MEMORY_HOTPLUG_SPARSE && NOTIFIER_ERROR_INJECTION
717 This option provides the ability to inject artificial errors to
718 memory hotplug notifier chain callbacks. It is controlled through
719 debugfs interface under /sys/kernel/debug/notifier-error-inject/memory
721 If the notifier call chain should be failed with some events
722 notified, write the error code to "actions/<notifier event>/error".
724 Example: Inject memory hotplug offline error (-12 == -ENOMEM)
726 # cd /sys/kernel/debug/notifier-error-inject/memory
727 # echo -12 > actions/MEM_GOING_OFFLINE/error
728 # echo offline > /sys/devices/system/memory/memoryXXX/state
729 bash: echo: write error: Cannot allocate memory
731 To compile this code as a module, choose M here: the module will
732 be called memory-notifier-error-inject.
736 config DEBUG_PER_CPU_MAPS
737 bool "Debug access to per_cpu maps"
738 depends on DEBUG_KERNEL
741 Say Y to verify that the per_cpu map being accessed has
742 been set up. This adds a fair amount of code to kernel memory
743 and decreases performance.
748 bool "Highmem debugging"
749 depends on DEBUG_KERNEL && HIGHMEM
751 This option enables additional error checking for high memory
752 systems. Disable for production systems.
754 config HAVE_DEBUG_STACKOVERFLOW
757 config DEBUG_STACKOVERFLOW
758 bool "Check for stack overflows"
759 depends on DEBUG_KERNEL && HAVE_DEBUG_STACKOVERFLOW
761 Say Y here if you want to check for overflows of kernel, IRQ
762 and exception stacks (if your architecture uses them). This
763 option will show detailed messages if free stack space drops
764 below a certain limit.
766 These kinds of bugs usually occur when call-chains in the
767 kernel get too deep, especially when interrupts are
770 Use this in cases where you see apparently random memory
771 corruption, especially if it appears in 'struct thread_info'
773 If in doubt, say "N".
775 source "lib/Kconfig.kasan"
777 endmenu # "Memory Debugging"
780 bool "Debug shared IRQ handlers"
781 depends on DEBUG_KERNEL
783 Enable this to generate a spurious interrupt as soon as a shared
784 interrupt handler is registered, and just before one is deregistered.
785 Drivers ought to be able to handle interrupts coming in at those
786 points; some don't and need to be caught.
788 menu "Debug Oops, Lockups and Hangs"
793 Say Y here to enable the kernel to panic when it oopses. This
794 has the same effect as setting oops=panic on the kernel command
797 This feature is useful to ensure that the kernel does not do
798 anything erroneous after an oops which could result in data
799 corruption or other issues.
803 config PANIC_ON_OOPS_VALUE
806 default 0 if !PANIC_ON_OOPS
807 default 1 if PANIC_ON_OOPS
813 Set the timeout value (in seconds) until a reboot occurs when the
814 the kernel panics. If n = 0, then we wait forever. A timeout
815 value n > 0 will wait n seconds before rebooting, while a timeout
816 value n < 0 will reboot immediately.
818 config LOCKUP_DETECTOR
821 config SOFTLOCKUP_DETECTOR
822 bool "Detect Soft Lockups"
823 depends on DEBUG_KERNEL && !S390
824 select LOCKUP_DETECTOR
826 Say Y here to enable the kernel to act as a watchdog to detect
829 Softlockups are bugs that cause the kernel to loop in kernel
830 mode for more than 20 seconds, without giving other tasks a
831 chance to run. The current stack trace is displayed upon
832 detection and the system will stay locked up.
834 config BOOTPARAM_SOFTLOCKUP_PANIC
835 bool "Panic (Reboot) On Soft Lockups"
836 depends on SOFTLOCKUP_DETECTOR
838 Say Y here to enable the kernel to panic on "soft lockups",
839 which are bugs that cause the kernel to loop in kernel
840 mode for more than 20 seconds (configurable using the watchdog_thresh
841 sysctl), without giving other tasks a chance to run.
843 The panic can be used in combination with panic_timeout,
844 to cause the system to reboot automatically after a
845 lockup has been detected. This feature is useful for
846 high-availability systems that have uptime guarantees and
847 where a lockup must be resolved ASAP.
851 config BOOTPARAM_SOFTLOCKUP_PANIC_VALUE
853 depends on SOFTLOCKUP_DETECTOR
855 default 0 if !BOOTPARAM_SOFTLOCKUP_PANIC
856 default 1 if BOOTPARAM_SOFTLOCKUP_PANIC
858 config HARDLOCKUP_DETECTOR_PERF
860 select SOFTLOCKUP_DETECTOR
863 # Enables a timestamp based low pass filter to compensate for perf based
864 # hard lockup detection which runs too fast due to turbo modes.
866 config HARDLOCKUP_CHECK_TIMESTAMP
870 # arch/ can define HAVE_HARDLOCKUP_DETECTOR_ARCH to provide their own hard
871 # lockup detector rather than the perf based detector.
873 config HARDLOCKUP_DETECTOR
874 bool "Detect Hard Lockups"
875 depends on DEBUG_KERNEL && !S390
876 depends on HAVE_HARDLOCKUP_DETECTOR_PERF || HAVE_HARDLOCKUP_DETECTOR_ARCH
877 select LOCKUP_DETECTOR
878 select HARDLOCKUP_DETECTOR_PERF if HAVE_HARDLOCKUP_DETECTOR_PERF
879 select HARDLOCKUP_DETECTOR_ARCH if HAVE_HARDLOCKUP_DETECTOR_ARCH
881 Say Y here to enable the kernel to act as a watchdog to detect
884 Hardlockups are bugs that cause the CPU to loop in kernel mode
885 for more than 10 seconds, without letting other interrupts have a
886 chance to run. The current stack trace is displayed upon detection
887 and the system will stay locked up.
889 config BOOTPARAM_HARDLOCKUP_PANIC
890 bool "Panic (Reboot) On Hard Lockups"
891 depends on HARDLOCKUP_DETECTOR
893 Say Y here to enable the kernel to panic on "hard lockups",
894 which are bugs that cause the kernel to loop in kernel
895 mode with interrupts disabled for more than 10 seconds (configurable
896 using the watchdog_thresh sysctl).
900 config BOOTPARAM_HARDLOCKUP_PANIC_VALUE
902 depends on HARDLOCKUP_DETECTOR
904 default 0 if !BOOTPARAM_HARDLOCKUP_PANIC
905 default 1 if BOOTPARAM_HARDLOCKUP_PANIC
907 config DETECT_HUNG_TASK
908 bool "Detect Hung Tasks"
909 depends on DEBUG_KERNEL
910 default SOFTLOCKUP_DETECTOR
912 Say Y here to enable the kernel to detect "hung tasks",
913 which are bugs that cause the task to be stuck in
914 uninterruptible "D" state indefinitely.
916 When a hung task is detected, the kernel will print the
917 current stack trace (which you should report), but the
918 task will stay in uninterruptible state. If lockdep is
919 enabled then all held locks will also be reported. This
920 feature has negligible overhead.
922 config DEFAULT_HUNG_TASK_TIMEOUT
923 int "Default timeout for hung task detection (in seconds)"
924 depends on DETECT_HUNG_TASK
927 This option controls the default timeout (in seconds) used
928 to determine when a task has become non-responsive and should
931 It can be adjusted at runtime via the kernel.hung_task_timeout_secs
932 sysctl or by writing a value to
933 /proc/sys/kernel/hung_task_timeout_secs.
935 A timeout of 0 disables the check. The default is two minutes.
936 Keeping the default should be fine in most cases.
938 config BOOTPARAM_HUNG_TASK_PANIC
939 bool "Panic (Reboot) On Hung Tasks"
940 depends on DETECT_HUNG_TASK
942 Say Y here to enable the kernel to panic on "hung tasks",
943 which are bugs that cause the kernel to leave a task stuck
944 in uninterruptible "D" state.
946 The panic can be used in combination with panic_timeout,
947 to cause the system to reboot automatically after a
948 hung task has been detected. This feature is useful for
949 high-availability systems that have uptime guarantees and
950 where a hung tasks must be resolved ASAP.
954 config BOOTPARAM_HUNG_TASK_PANIC_VALUE
956 depends on DETECT_HUNG_TASK
958 default 0 if !BOOTPARAM_HUNG_TASK_PANIC
959 default 1 if BOOTPARAM_HUNG_TASK_PANIC
962 bool "Detect Workqueue Stalls"
963 depends on DEBUG_KERNEL
965 Say Y here to enable stall detection on workqueues. If a
966 worker pool doesn't make forward progress on a pending work
967 item for over a given amount of time, 30s by default, a
968 warning message is printed along with dump of workqueue
969 state. This can be configured through kernel parameter
970 "workqueue.watchdog_thresh" and its sysfs counterpart.
972 endmenu # "Debug lockups and hangs"
975 bool "Collect scheduler debugging info"
976 depends on DEBUG_KERNEL && PROC_FS
979 If you say Y here, the /proc/sched_debug file will be provided
980 that can help debug the scheduler. The runtime overhead of this
988 bool "Collect scheduler statistics"
989 depends on DEBUG_KERNEL && PROC_FS
992 If you say Y here, additional code will be inserted into the
993 scheduler and related routines to collect statistics about
994 scheduler behavior and provide them in /proc/schedstat. These
995 stats may be useful for both tuning and debugging the scheduler
996 If you aren't debugging the scheduler or trying to tune a specific
997 application, you can say N to avoid the very slight overhead
1000 config DEBUG_TIMEKEEPING
1001 bool "Enable extra timekeeping sanity checking"
1003 This option will enable additional timekeeping sanity checks
1004 which may be helpful when diagnosing issues where timekeeping
1005 problems are suspected.
1007 This may include checks in the timekeeping hotpaths, so this
1008 option may have a (very small) performance impact to some
1013 config DEBUG_PREEMPT
1014 bool "Debug preemptible kernel"
1015 depends on DEBUG_KERNEL && PREEMPT && TRACE_IRQFLAGS_SUPPORT
1018 If you say Y here then the kernel will use a debug variant of the
1019 commonly used smp_processor_id() function and will print warnings
1020 if kernel code uses it in a preemption-unsafe way. Also, the kernel
1021 will detect preemption count underflows.
1023 menu "Lock Debugging (spinlocks, mutexes, etc...)"
1025 config LOCK_DEBUGGING_SUPPORT
1027 depends on TRACE_IRQFLAGS_SUPPORT && STACKTRACE_SUPPORT && LOCKDEP_SUPPORT
1030 config PROVE_LOCKING
1031 bool "Lock debugging: prove locking correctness"
1032 depends on DEBUG_KERNEL && LOCK_DEBUGGING_SUPPORT
1034 select DEBUG_SPINLOCK
1035 select DEBUG_MUTEXES
1036 select DEBUG_RT_MUTEXES if RT_MUTEXES
1038 select DEBUG_WW_MUTEX_SLOWPATH
1039 select DEBUG_LOCK_ALLOC
1040 select TRACE_IRQFLAGS
1043 This feature enables the kernel to prove that all locking
1044 that occurs in the kernel runtime is mathematically
1045 correct: that under no circumstance could an arbitrary (and
1046 not yet triggered) combination of observed locking
1047 sequences (on an arbitrary number of CPUs, running an
1048 arbitrary number of tasks and interrupt contexts) cause a
1051 In short, this feature enables the kernel to report locking
1052 related deadlocks before they actually occur.
1054 The proof does not depend on how hard and complex a
1055 deadlock scenario would be to trigger: how many
1056 participant CPUs, tasks and irq-contexts would be needed
1057 for it to trigger. The proof also does not depend on
1058 timing: if a race and a resulting deadlock is possible
1059 theoretically (no matter how unlikely the race scenario
1060 is), it will be proven so and will immediately be
1061 reported by the kernel (once the event is observed that
1062 makes the deadlock theoretically possible).
1064 If a deadlock is impossible (i.e. the locking rules, as
1065 observed by the kernel, are mathematically correct), the
1066 kernel reports nothing.
1068 NOTE: this feature can also be enabled for rwlocks, mutexes
1069 and rwsems - in which case all dependencies between these
1070 different locking variants are observed and mapped too, and
1071 the proof of observed correctness is also maintained for an
1072 arbitrary combination of these separate locking variants.
1074 For more details, see Documentation/locking/lockdep-design.rst.
1077 bool "Lock usage statistics"
1078 depends on DEBUG_KERNEL && LOCK_DEBUGGING_SUPPORT
1080 select DEBUG_SPINLOCK
1081 select DEBUG_MUTEXES
1082 select DEBUG_RT_MUTEXES if RT_MUTEXES
1083 select DEBUG_LOCK_ALLOC
1086 This feature enables tracking lock contention points
1088 For more details, see Documentation/locking/lockstat.rst
1090 This also enables lock events required by "perf lock",
1092 If you want to use "perf lock", you also need to turn on
1093 CONFIG_EVENT_TRACING.
1095 CONFIG_LOCK_STAT defines "contended" and "acquired" lock events.
1096 (CONFIG_LOCKDEP defines "acquire" and "release" events.)
1098 config DEBUG_RT_MUTEXES
1099 bool "RT Mutex debugging, deadlock detection"
1100 depends on DEBUG_KERNEL && RT_MUTEXES
1102 This allows rt mutex semantics violations and rt mutex related
1103 deadlocks (lockups) to be detected and reported automatically.
1105 config DEBUG_SPINLOCK
1106 bool "Spinlock and rw-lock debugging: basic checks"
1107 depends on DEBUG_KERNEL
1108 select UNINLINE_SPIN_UNLOCK
1110 Say Y here and build SMP to catch missing spinlock initialization
1111 and certain other kinds of spinlock errors commonly made. This is
1112 best used in conjunction with the NMI watchdog so that spinlock
1113 deadlocks are also debuggable.
1115 config DEBUG_MUTEXES
1116 bool "Mutex debugging: basic checks"
1117 depends on DEBUG_KERNEL
1119 This feature allows mutex semantics violations to be detected and
1122 config DEBUG_WW_MUTEX_SLOWPATH
1123 bool "Wait/wound mutex debugging: Slowpath testing"
1124 depends on DEBUG_KERNEL && LOCK_DEBUGGING_SUPPORT
1125 select DEBUG_LOCK_ALLOC
1126 select DEBUG_SPINLOCK
1127 select DEBUG_MUTEXES
1129 This feature enables slowpath testing for w/w mutex users by
1130 injecting additional -EDEADLK wound/backoff cases. Together with
1131 the full mutex checks enabled with (CONFIG_PROVE_LOCKING) this
1132 will test all possible w/w mutex interface abuse with the
1133 exception of simply not acquiring all the required locks.
1134 Note that this feature can introduce significant overhead, so
1135 it really should not be enabled in a production or distro kernel,
1136 even a debug kernel. If you are a driver writer, enable it. If
1137 you are a distro, do not.
1140 bool "RW Semaphore debugging: basic checks"
1141 depends on DEBUG_KERNEL
1143 This debugging feature allows mismatched rw semaphore locks
1144 and unlocks to be detected and reported.
1146 config DEBUG_LOCK_ALLOC
1147 bool "Lock debugging: detect incorrect freeing of live locks"
1148 depends on DEBUG_KERNEL && LOCK_DEBUGGING_SUPPORT
1149 select DEBUG_SPINLOCK
1150 select DEBUG_MUTEXES
1151 select DEBUG_RT_MUTEXES if RT_MUTEXES
1154 This feature will check whether any held lock (spinlock, rwlock,
1155 mutex or rwsem) is incorrectly freed by the kernel, via any of the
1156 memory-freeing routines (kfree(), kmem_cache_free(), free_pages(),
1157 vfree(), etc.), whether a live lock is incorrectly reinitialized via
1158 spin_lock_init()/mutex_init()/etc., or whether there is any lock
1159 held during task exit.
1163 depends on DEBUG_KERNEL && LOCK_DEBUGGING_SUPPORT
1165 select FRAME_POINTER if !MIPS && !PPC && !ARM && !S390 && !MICROBLAZE && !ARC && !X86
1169 config LOCKDEP_SMALL
1172 config DEBUG_LOCKDEP
1173 bool "Lock dependency engine debugging"
1174 depends on DEBUG_KERNEL && LOCKDEP
1176 If you say Y here, the lock dependency engine will do
1177 additional runtime checks to debug itself, at the price
1178 of more runtime overhead.
1180 config DEBUG_ATOMIC_SLEEP
1181 bool "Sleep inside atomic section checking"
1182 select PREEMPT_COUNT
1183 depends on DEBUG_KERNEL
1184 depends on !ARCH_NO_PREEMPT
1186 If you say Y here, various routines which may sleep will become very
1187 noisy if they are called inside atomic sections: when a spinlock is
1188 held, inside an rcu read side critical section, inside preempt disabled
1189 sections, inside an interrupt, etc...
1191 config DEBUG_LOCKING_API_SELFTESTS
1192 bool "Locking API boot-time self-tests"
1193 depends on DEBUG_KERNEL
1195 Say Y here if you want the kernel to run a short self-test during
1196 bootup. The self-test checks whether common types of locking bugs
1197 are detected by debugging mechanisms or not. (if you disable
1198 lock debugging then those bugs wont be detected of course.)
1199 The following locking APIs are covered: spinlocks, rwlocks,
1202 config LOCK_TORTURE_TEST
1203 tristate "torture tests for locking"
1204 depends on DEBUG_KERNEL
1207 This option provides a kernel module that runs torture tests
1208 on kernel locking primitives. The kernel module may be built
1209 after the fact on the running kernel to be tested, if desired.
1211 Say Y here if you want kernel locking-primitive torture tests
1212 to be built into the kernel.
1213 Say M if you want these torture tests to build as a module.
1214 Say N if you are unsure.
1216 config WW_MUTEX_SELFTEST
1217 tristate "Wait/wound mutex selftests"
1219 This option provides a kernel module that runs tests on the
1220 on the struct ww_mutex locking API.
1222 It is recommended to enable DEBUG_WW_MUTEX_SLOWPATH in conjunction
1223 with this test harness.
1225 Say M if you want these self tests to build as a module.
1226 Say N if you are unsure.
1228 endmenu # lock debugging
1230 config TRACE_IRQFLAGS
1233 Enables hooks to interrupt enabling and disabling for
1234 either tracing or lock debugging.
1237 bool "Stack backtrace support"
1238 depends on STACKTRACE_SUPPORT
1240 This option causes the kernel to create a /proc/pid/stack for
1241 every process, showing its current stack trace.
1242 It is also used by various kernel debugging features that require
1243 stack trace generation.
1245 config WARN_ALL_UNSEEDED_RANDOM
1246 bool "Warn for all uses of unseeded randomness"
1249 Some parts of the kernel contain bugs relating to their use of
1250 cryptographically secure random numbers before it's actually possible
1251 to generate those numbers securely. This setting ensures that these
1252 flaws don't go unnoticed, by enabling a message, should this ever
1253 occur. This will allow people with obscure setups to know when things
1254 are going wrong, so that they might contact developers about fixing
1257 Unfortunately, on some models of some architectures getting
1258 a fully seeded CRNG is extremely difficult, and so this can
1259 result in dmesg getting spammed for a surprisingly long
1260 time. This is really bad from a security perspective, and
1261 so architecture maintainers really need to do what they can
1262 to get the CRNG seeded sooner after the system is booted.
1263 However, since users cannot do anything actionable to
1264 address this, by default the kernel will issue only a single
1265 warning for the first use of unseeded randomness.
1267 Say Y here if you want to receive warnings for all uses of
1268 unseeded randomness. This will be of use primarily for
1269 those developers interested in improving the security of
1270 Linux kernels running on their architecture (or
1273 config DEBUG_KOBJECT
1274 bool "kobject debugging"
1275 depends on DEBUG_KERNEL
1277 If you say Y here, some extra kobject debugging messages will be sent
1280 config DEBUG_KOBJECT_RELEASE
1281 bool "kobject release debugging"
1282 depends on DEBUG_OBJECTS_TIMERS
1284 kobjects are reference counted objects. This means that their
1285 last reference count put is not predictable, and the kobject can
1286 live on past the point at which a driver decides to drop it's
1287 initial reference to the kobject gained on allocation. An
1288 example of this would be a struct device which has just been
1291 However, some buggy drivers assume that after such an operation,
1292 the memory backing the kobject can be immediately freed. This
1293 goes completely against the principles of a refcounted object.
1295 If you say Y here, the kernel will delay the release of kobjects
1296 on the last reference count to improve the visibility of this
1297 kind of kobject release bug.
1299 config HAVE_DEBUG_BUGVERBOSE
1302 config DEBUG_BUGVERBOSE
1303 bool "Verbose BUG() reporting (adds 70K)" if DEBUG_KERNEL && EXPERT
1304 depends on BUG && (GENERIC_BUG || HAVE_DEBUG_BUGVERBOSE)
1307 Say Y here to make BUG() panics output the file name and line number
1308 of the BUG call as well as the EIP and oops trace. This aids
1309 debugging but costs about 70-100K of memory.
1311 menu "Debug kernel data structures"
1314 bool "Debug linked list manipulation"
1315 depends on DEBUG_KERNEL || BUG_ON_DATA_CORRUPTION
1317 Enable this to turn on extended checks in the linked-list
1323 bool "Debug priority linked list manipulation"
1324 depends on DEBUG_KERNEL
1326 Enable this to turn on extended checks in the priority-ordered
1327 linked-list (plist) walking routines. This checks the entire
1328 list multiple times during each manipulation.
1333 bool "Debug SG table operations"
1334 depends on DEBUG_KERNEL
1336 Enable this to turn on checks on scatter-gather tables. This can
1337 help find problems with drivers that do not properly initialize
1342 config DEBUG_NOTIFIERS
1343 bool "Debug notifier call chains"
1344 depends on DEBUG_KERNEL
1346 Enable this to turn on sanity checking for notifier call chains.
1347 This is most useful for kernel developers to make sure that
1348 modules properly unregister themselves from notifier chains.
1349 This is a relatively cheap check but if you care about maximum
1352 config BUG_ON_DATA_CORRUPTION
1353 bool "Trigger a BUG when data corruption is detected"
1356 Select this option if the kernel should BUG when it encounters
1357 data corruption in kernel memory structures when they get checked
1364 config DEBUG_CREDENTIALS
1365 bool "Debug credential management"
1366 depends on DEBUG_KERNEL
1368 Enable this to turn on some debug checking for credential
1369 management. The additional code keeps track of the number of
1370 pointers from task_structs to any given cred struct, and checks to
1371 see that this number never exceeds the usage count of the cred
1374 Furthermore, if SELinux is enabled, this also checks that the
1375 security pointer in the cred struct is never seen to be invalid.
1379 source "kernel/rcu/Kconfig.debug"
1381 config DEBUG_WQ_FORCE_RR_CPU
1382 bool "Force round-robin CPU selection for unbound work items"
1383 depends on DEBUG_KERNEL
1386 Workqueue used to implicitly guarantee that work items queued
1387 without explicit CPU specified are put on the local CPU. This
1388 guarantee is no longer true and while local CPU is still
1389 preferred work items may be put on foreign CPUs. Kernel
1390 parameter "workqueue.debug_force_rr_cpu" is added to force
1391 round-robin CPU selection to flush out usages which depend on the
1392 now broken guarantee. This config option enables the debug
1393 feature by default. When enabled, memory and cache locality will
1396 config DEBUG_BLOCK_EXT_DEVT
1397 bool "Force extended block device numbers and spread them"
1398 depends on DEBUG_KERNEL
1402 BIG FAT WARNING: ENABLING THIS OPTION MIGHT BREAK BOOTING ON
1403 SOME DISTRIBUTIONS. DO NOT ENABLE THIS UNLESS YOU KNOW WHAT
1404 YOU ARE DOING. Distros, please enable this and fix whatever
1407 Conventionally, block device numbers are allocated from
1408 predetermined contiguous area. However, extended block area
1409 may introduce non-contiguous block device numbers. This
1410 option forces most block device numbers to be allocated from
1411 the extended space and spreads them to discover kernel or
1412 userland code paths which assume predetermined contiguous
1413 device number allocation.
1415 Note that turning on this debug option shuffles all the
1416 device numbers for all IDE and SCSI devices including libata
1417 ones, so root partition specified using device number
1418 directly (via rdev or root=MAJ:MIN) won't work anymore.
1419 Textual device names (root=/dev/sdXn) will continue to work.
1421 Say N if you are unsure.
1423 config CPU_HOTPLUG_STATE_CONTROL
1424 bool "Enable CPU hotplug state control"
1425 depends on DEBUG_KERNEL
1426 depends on HOTPLUG_CPU
1429 Allows to write steps between "offline" and "online" to the CPUs
1430 sysfs target file so states can be stepped granular. This is a debug
1431 option for now as the hotplug machinery cannot be stopped and
1432 restarted at arbitrary points yet.
1434 Say N if your are unsure.
1437 bool "Latency measuring infrastructure"
1438 depends on DEBUG_KERNEL
1439 depends on STACKTRACE_SUPPORT
1441 select FRAME_POINTER if !MIPS && !PPC && !S390 && !MICROBLAZE && !ARM && !ARC && !X86
1448 Enable this option if you want to use the LatencyTOP tool
1449 to find out which userspace is blocking on what kernel operations.
1451 source "kernel/trace/Kconfig"
1453 config PROVIDE_OHCI1394_DMA_INIT
1454 bool "Remote debugging over FireWire early on boot"
1455 depends on PCI && X86
1457 If you want to debug problems which hang or crash the kernel early
1458 on boot and the crashing machine has a FireWire port, you can use
1459 this feature to remotely access the memory of the crashed machine
1460 over FireWire. This employs remote DMA as part of the OHCI1394
1461 specification which is now the standard for FireWire controllers.
1463 With remote DMA, you can monitor the printk buffer remotely using
1464 firescope and access all memory below 4GB using fireproxy from gdb.
1465 Even controlling a kernel debugger is possible using remote DMA.
1469 If ohci1394_dma=early is used as boot parameter, it will initialize
1470 all OHCI1394 controllers which are found in the PCI config space.
1472 As all changes to the FireWire bus such as enabling and disabling
1473 devices cause a bus reset and thereby disable remote DMA for all
1474 devices, be sure to have the cable plugged and FireWire enabled on
1475 the debugging host before booting the debug target for debugging.
1477 This code (~1k) is freed after boot. By then, the firewire stack
1478 in charge of the OHCI-1394 controllers should be used instead.
1480 See Documentation/debugging-via-ohci1394.txt for more information.
1482 source "lib/kunit/Kconfig"
1484 config NOTIFIER_ERROR_INJECTION
1485 tristate "Notifier error injection"
1486 depends on DEBUG_KERNEL
1489 This option provides the ability to inject artificial errors to
1490 specified notifier chain callbacks. It is useful to test the error
1491 handling of notifier call chain failures.
1495 config PM_NOTIFIER_ERROR_INJECT
1496 tristate "PM notifier error injection module"
1497 depends on PM && NOTIFIER_ERROR_INJECTION
1498 default m if PM_DEBUG
1500 This option provides the ability to inject artificial errors to
1501 PM notifier chain callbacks. It is controlled through debugfs
1502 interface /sys/kernel/debug/notifier-error-inject/pm
1504 If the notifier call chain should be failed with some events
1505 notified, write the error code to "actions/<notifier event>/error".
1507 Example: Inject PM suspend error (-12 = -ENOMEM)
1509 # cd /sys/kernel/debug/notifier-error-inject/pm/
1510 # echo -12 > actions/PM_SUSPEND_PREPARE/error
1511 # echo mem > /sys/power/state
1512 bash: echo: write error: Cannot allocate memory
1514 To compile this code as a module, choose M here: the module will
1515 be called pm-notifier-error-inject.
1519 config OF_RECONFIG_NOTIFIER_ERROR_INJECT
1520 tristate "OF reconfig notifier error injection module"
1521 depends on OF_DYNAMIC && NOTIFIER_ERROR_INJECTION
1523 This option provides the ability to inject artificial errors to
1524 OF reconfig notifier chain callbacks. It is controlled
1525 through debugfs interface under
1526 /sys/kernel/debug/notifier-error-inject/OF-reconfig/
1528 If the notifier call chain should be failed with some events
1529 notified, write the error code to "actions/<notifier event>/error".
1531 To compile this code as a module, choose M here: the module will
1532 be called of-reconfig-notifier-error-inject.
1536 config NETDEV_NOTIFIER_ERROR_INJECT
1537 tristate "Netdev notifier error injection module"
1538 depends on NET && NOTIFIER_ERROR_INJECTION
1540 This option provides the ability to inject artificial errors to
1541 netdevice notifier chain callbacks. It is controlled through debugfs
1542 interface /sys/kernel/debug/notifier-error-inject/netdev
1544 If the notifier call chain should be failed with some events
1545 notified, write the error code to "actions/<notifier event>/error".
1547 Example: Inject netdevice mtu change error (-22 = -EINVAL)
1549 # cd /sys/kernel/debug/notifier-error-inject/netdev
1550 # echo -22 > actions/NETDEV_CHANGEMTU/error
1551 # ip link set eth0 mtu 1024
1552 RTNETLINK answers: Invalid argument
1554 To compile this code as a module, choose M here: the module will
1555 be called netdev-notifier-error-inject.
1559 config FUNCTION_ERROR_INJECTION
1561 depends on HAVE_FUNCTION_ERROR_INJECTION && KPROBES
1563 config FAULT_INJECTION
1564 bool "Fault-injection framework"
1565 depends on DEBUG_KERNEL
1567 Provide fault-injection framework.
1568 For more details, see Documentation/fault-injection/.
1571 bool "Fault-injection capability for kmalloc"
1572 depends on FAULT_INJECTION
1573 depends on SLAB || SLUB
1575 Provide fault-injection capability for kmalloc.
1577 config FAIL_PAGE_ALLOC
1578 bool "Fault-injection capabilitiy for alloc_pages()"
1579 depends on FAULT_INJECTION
1581 Provide fault-injection capability for alloc_pages().
1583 config FAIL_MAKE_REQUEST
1584 bool "Fault-injection capability for disk IO"
1585 depends on FAULT_INJECTION && BLOCK
1587 Provide fault-injection capability for disk IO.
1589 config FAIL_IO_TIMEOUT
1590 bool "Fault-injection capability for faking disk interrupts"
1591 depends on FAULT_INJECTION && BLOCK
1593 Provide fault-injection capability on end IO handling. This
1594 will make the block layer "forget" an interrupt as configured,
1595 thus exercising the error handling.
1597 Only works with drivers that use the generic timeout handling,
1598 for others it wont do anything.
1601 bool "Fault-injection capability for futexes"
1603 depends on FAULT_INJECTION && FUTEX
1605 Provide fault-injection capability for futexes.
1607 config FAULT_INJECTION_DEBUG_FS
1608 bool "Debugfs entries for fault-injection capabilities"
1609 depends on FAULT_INJECTION && SYSFS && DEBUG_FS
1611 Enable configuration of fault-injection capabilities via debugfs.
1613 config FAIL_FUNCTION
1614 bool "Fault-injection capability for functions"
1615 depends on FAULT_INJECTION_DEBUG_FS && FUNCTION_ERROR_INJECTION
1617 Provide function-based fault-injection capability.
1618 This will allow you to override a specific function with a return
1619 with given return value. As a result, function caller will see
1620 an error value and have to handle it. This is useful to test the
1621 error handling in various subsystems.
1623 config FAIL_MMC_REQUEST
1624 bool "Fault-injection capability for MMC IO"
1625 depends on FAULT_INJECTION_DEBUG_FS && MMC
1627 Provide fault-injection capability for MMC IO.
1628 This will make the mmc core return data errors. This is
1629 useful to test the error handling in the mmc block device
1630 and to test how the mmc host driver handles retries from
1633 config FAULT_INJECTION_STACKTRACE_FILTER
1634 bool "stacktrace filter for fault-injection capabilities"
1635 depends on FAULT_INJECTION_DEBUG_FS && STACKTRACE_SUPPORT
1638 select FRAME_POINTER if !MIPS && !PPC && !S390 && !MICROBLAZE && !ARM && !ARC && !X86
1640 Provide stacktrace filter for fault-injection capabilities
1642 endmenu # "Kernel Testing and Coverage"
1644 menu "Kernel Testing and Coverage"
1646 config ARCH_HAS_KCOV
1649 An architecture should select this when it can successfully
1650 build and run with CONFIG_KCOV. This typically requires
1651 disabling instrumentation for some early boot code.
1653 config CC_HAS_SANCOV_TRACE_PC
1654 def_bool $(cc-option,-fsanitize-coverage=trace-pc)
1658 bool "Code coverage for fuzzing"
1659 depends on ARCH_HAS_KCOV
1660 depends on CC_HAS_SANCOV_TRACE_PC || GCC_PLUGINS
1662 select GCC_PLUGIN_SANCOV if !CC_HAS_SANCOV_TRACE_PC
1664 KCOV exposes kernel code coverage information in a form suitable
1665 for coverage-guided fuzzing (randomized testing).
1667 If RANDOMIZE_BASE is enabled, PC values will not be stable across
1668 different machines and across reboots. If you need stable PC values,
1669 disable RANDOMIZE_BASE.
1671 For more details, see Documentation/dev-tools/kcov.rst.
1673 config KCOV_ENABLE_COMPARISONS
1674 bool "Enable comparison operands collection by KCOV"
1676 depends on $(cc-option,-fsanitize-coverage=trace-cmp)
1678 KCOV also exposes operands of every comparison in the instrumented
1679 code along with operand sizes and PCs of the comparison instructions.
1680 These operands can be used by fuzzing engines to improve the quality
1681 of fuzzing coverage.
1683 config KCOV_INSTRUMENT_ALL
1684 bool "Instrument all code by default"
1688 If you are doing generic system call fuzzing (like e.g. syzkaller),
1689 then you will want to instrument the whole kernel and you should
1690 say y here. If you are doing more targeted fuzzing (like e.g.
1691 filesystem fuzzing with AFL) then you will want to enable coverage
1692 for more specific subsets of files, and should say n here.
1694 menuconfig RUNTIME_TESTING_MENU
1695 bool "Runtime Testing"
1698 if RUNTIME_TESTING_MENU
1701 tristate "Linux Kernel Dump Test Tool Module"
1704 This module enables testing of the different dumping mechanisms by
1705 inducing system failures at predefined crash points.
1706 If you don't need it: say N
1707 Choose M here to compile this code as a module. The module will be
1710 Documentation on how to use the module can be found in
1711 Documentation/fault-injection/provoke-crashes.rst
1713 config TEST_LIST_SORT
1714 tristate "Linked list sorting test"
1715 depends on DEBUG_KERNEL || m
1717 Enable this to turn on 'list_sort()' function test. This test is
1718 executed only once during system boot (so affects only boot time),
1719 or at module load time.
1724 tristate "Array-based sort test"
1725 depends on DEBUG_KERNEL || m
1727 This option enables the self-test function of 'sort()' at boot,
1728 or at module load time.
1732 config KPROBES_SANITY_TEST
1733 bool "Kprobes sanity tests"
1734 depends on DEBUG_KERNEL
1737 This option provides for testing basic kprobes functionality on
1738 boot. Samples of kprobe and kretprobe are inserted and
1739 verified for functionality.
1741 Say N if you are unsure.
1743 config BACKTRACE_SELF_TEST
1744 tristate "Self test for the backtrace code"
1745 depends on DEBUG_KERNEL
1747 This option provides a kernel module that can be used to test
1748 the kernel stack backtrace code. This option is not useful
1749 for distributions or general kernels, but only for kernel
1750 developers working on architecture code.
1752 Note that if you want to also test saved backtraces, you will
1753 have to enable STACKTRACE as well.
1755 Say N if you are unsure.
1758 tristate "Red-Black tree test"
1759 depends on DEBUG_KERNEL
1761 A benchmark measuring the performance of the rbtree library.
1762 Also includes rbtree invariant checks.
1764 config REED_SOLOMON_TEST
1765 tristate "Reed-Solomon library test"
1766 depends on DEBUG_KERNEL || m
1768 select REED_SOLOMON_ENC16
1769 select REED_SOLOMON_DEC16
1771 This option enables the self-test function of rslib at boot,
1772 or at module load time.
1776 config INTERVAL_TREE_TEST
1777 tristate "Interval tree test"
1778 depends on DEBUG_KERNEL
1779 select INTERVAL_TREE
1781 A benchmark measuring the performance of the interval tree library
1784 tristate "Per cpu operations test"
1785 depends on m && DEBUG_KERNEL
1787 Enable this option to build test module which validates per-cpu
1792 config ATOMIC64_SELFTEST
1793 tristate "Perform an atomic64_t self-test"
1795 Enable this option to test the atomic64_t functions at boot or
1796 at module load time.
1800 config ASYNC_RAID6_TEST
1801 tristate "Self test for hardware accelerated raid6 recovery"
1802 depends on ASYNC_RAID6_RECOV
1805 This is a one-shot self test that permutes through the
1806 recovery of all the possible two disk failure scenarios for a
1807 N-disk array. Recovery is performed with the asynchronous
1808 raid6 recovery routines, and will optionally use an offload
1809 engine if one is available.
1814 tristate "Test functions located in the hexdump module at runtime"
1816 config TEST_STRING_HELPERS
1817 tristate "Test functions located in the string_helpers module at runtime"
1820 tristate "Test strscpy*() family of functions at runtime"
1823 tristate "Test kstrto*() family of functions at runtime"
1826 tristate "Test printf() family of functions at runtime"
1829 tristate "Test bitmap_*() family of functions at runtime"
1831 Enable this option to test the bitmap functions at boot.
1835 config TEST_BITFIELD
1836 tristate "Test bitfield functions at runtime"
1838 Enable this option to test the bitfield functions at boot.
1843 tristate "Test functions located in the uuid module at runtime"
1846 tristate "Test the XArray code at runtime"
1848 config TEST_OVERFLOW
1849 tristate "Test check_*_overflow() functions at runtime"
1851 config TEST_RHASHTABLE
1852 tristate "Perform selftest on resizable hash table"
1854 Enable this option to test the rhashtable functions at boot.
1859 tristate "Perform selftest on hash functions"
1861 Enable this option to test the kernel's integer (<linux/hash.h>),
1862 string (<linux/stringhash.h>), and siphash (<linux/siphash.h>)
1863 hash functions on boot (or module load).
1865 This is intended to help people writing architecture-specific
1866 optimized versions. If unsure, say N.
1869 tristate "Perform selftest on IDA functions"
1872 tristate "Perform selftest on priority array manager"
1875 Enable this option to test priority array manager on boot
1880 config TEST_IRQ_TIMINGS
1881 bool "IRQ timings selftest"
1882 depends on IRQ_TIMINGS
1884 Enable this option to test the irq timings code on boot.
1889 tristate "Test module loading with 'hello world' module"
1892 This builds the "test_module" module that emits "Hello, world"
1893 on printk when loaded. It is designed to be used for basic
1894 evaluation of the module loading subsystem (for example when
1895 validating module verification). It lacks any extra dependencies,
1896 and will not normally be loaded by the system unless explicitly
1902 tristate "Test module for stress/performance analysis of vmalloc allocator"
1907 This builds the "test_vmalloc" module that should be used for
1908 stress and performance analysis. So, any new change for vmalloc
1909 subsystem can be evaluated from performance and stability point
1914 config TEST_USER_COPY
1915 tristate "Test user/kernel boundary protections"
1918 This builds the "test_user_copy" module that runs sanity checks
1919 on the copy_to/from_user infrastructure, making sure basic
1920 user/kernel boundary testing is working. If it fails to load,
1921 a regression has been detected in the user/kernel memory boundary
1927 tristate "Test BPF filter functionality"
1930 This builds the "test_bpf" module that runs various test vectors
1931 against the BPF interpreter or BPF JIT compiler depending on the
1932 current setting. This is in particular useful for BPF JIT compiler
1933 development, but also to run regression tests against changes in
1934 the interpreter code. It also enables test stubs for eBPF maps and
1935 verifier used by user space verifier testsuite.
1939 config TEST_BLACKHOLE_DEV
1940 tristate "Test blackhole netdev functionality"
1943 This builds the "test_blackhole_dev" module that validates the
1944 data path through this blackhole netdev.
1948 config FIND_BIT_BENCHMARK
1949 tristate "Test find_bit functions"
1951 This builds the "test_find_bit" module that measure find_*_bit()
1952 functions performance.
1956 config TEST_FIRMWARE
1957 tristate "Test firmware loading via userspace interface"
1958 depends on FW_LOADER
1960 This builds the "test_firmware" module that creates a userspace
1961 interface for testing firmware loading. This can be used to
1962 control the triggering of firmware loading without needing an
1963 actual firmware-using device. The contents can be rechecked by
1969 tristate "sysctl test driver"
1970 depends on PROC_SYSCTL
1972 This builds the "test_sysctl" module. This driver enables to test the
1973 proc sysctl interfaces available to drivers safely without affecting
1974 production knobs which might alter system functionality.
1978 config SYSCTL_KUNIT_TEST
1979 bool "KUnit test for sysctl"
1982 This builds the proc sysctl unit test, which runs on boot.
1983 Tests the API contract and implementation correctness of sysctl.
1984 For more information on KUnit and unit tests in general please refer
1985 to the KUnit documentation in Documentation/dev-tools/kunit/.
1989 config LIST_KUNIT_TEST
1990 bool "KUnit Test for Kernel Linked-list structures"
1993 This builds the linked list KUnit test suite.
1994 It tests that the API and basic functionality of the list_head type
1995 and associated macros.
1997 KUnit tests run during boot and output the results to the debug log
1998 in TAP format (http://testanything.org/). Only useful for kernel devs
1999 running the KUnit test harness, and not intended for inclusion into a
2002 For more information on KUnit and unit tests in general please refer
2003 to the KUnit documentation in Documentation/dev-tools/kunit/.
2008 tristate "udelay test driver"
2010 This builds the "udelay_test" module that helps to make sure
2011 that udelay() is working properly.
2015 config TEST_STATIC_KEYS
2016 tristate "Test static keys"
2019 Test the static key interfaces.
2024 tristate "kmod stress tester"
2026 depends on NETDEVICES && NET_CORE && INET # for TUN
2033 Test the kernel's module loading mechanism: kmod. kmod implements
2034 support to load modules using the Linux kernel's usermode helper.
2035 This test provides a series of tests against kmod.
2037 Although technically you can either build test_kmod as a module or
2038 into the kernel we disallow building it into the kernel since
2039 it stress tests request_module() and this will very likely cause
2040 some issues by taking over precious threads available from other
2041 module load requests, ultimately this could be fatal.
2045 tools/testing/selftests/kmod/kmod.sh --help
2049 config TEST_DEBUG_VIRTUAL
2050 tristate "Test CONFIG_DEBUG_VIRTUAL feature"
2051 depends on DEBUG_VIRTUAL
2053 Test the kernel's ability to detect incorrect calls to
2054 virt_to_phys() done against the non-linear part of the
2055 kernel's virtual address map.
2059 config TEST_MEMCAT_P
2060 tristate "Test memcat_p() helper function"
2062 Test the memcat_p() helper for correctly merging two
2063 pointer arrays together.
2067 config TEST_LIVEPATCH
2068 tristate "Test livepatching"
2070 depends on DYNAMIC_DEBUG
2071 depends on LIVEPATCH
2074 Test kernel livepatching features for correctness. The tests will
2075 load test modules that will be livepatched in various scenarios.
2077 To run all the livepatching tests:
2079 make -C tools/testing/selftests TARGETS=livepatch run_tests
2081 Alternatively, individual tests may be invoked:
2083 tools/testing/selftests/livepatch/test-callbacks.sh
2084 tools/testing/selftests/livepatch/test-livepatch.sh
2085 tools/testing/selftests/livepatch/test-shadow-vars.sh
2090 tristate "Perform selftest on object aggreration manager"
2094 Enable this option to test object aggregation manager on boot
2098 config TEST_STACKINIT
2099 tristate "Test level of stack variable initialization"
2101 Test if the kernel is zero-initializing stack variables and
2102 padding. Coverage is controlled by compiler flags,
2103 CONFIG_GCC_PLUGIN_STRUCTLEAK, CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF,
2104 or CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL.
2109 tristate "Test heap/page initialization"
2111 Test if the kernel is zero-initializing heap and page allocations.
2112 This can be useful to test init_on_alloc and init_on_free features.
2116 endif # RUNTIME_TESTING_MENU
2121 This option adds a kernel parameter 'memtest', which allows memtest
2123 memtest=0, mean disabled; -- default
2124 memtest=1, mean do 1 test pattern;
2126 memtest=17, mean do 17 test patterns.
2127 If you are unsure how to answer this question, answer N.
2129 source "samples/Kconfig"
2131 config ARCH_HAS_DEVMEM_IS_ALLOWED
2134 config STRICT_DEVMEM
2135 bool "Filter access to /dev/mem"
2136 depends on MMU && DEVMEM
2137 depends on ARCH_HAS_DEVMEM_IS_ALLOWED
2138 default y if PPC || X86 || ARM64
2140 If this option is disabled, you allow userspace (root) access to all
2141 of memory, including kernel and userspace memory. Accidental
2142 access to this is obviously disastrous, but specific access can
2143 be used by people debugging the kernel. Note that with PAT support
2144 enabled, even in this case there are restrictions on /dev/mem
2145 use due to the cache aliasing requirements.
2147 If this option is switched on, and IO_STRICT_DEVMEM=n, the /dev/mem
2148 file only allows userspace access to PCI space and the BIOS code and
2149 data regions. This is sufficient for dosemu and X and all common
2154 config IO_STRICT_DEVMEM
2155 bool "Filter I/O access to /dev/mem"
2156 depends on STRICT_DEVMEM
2158 If this option is disabled, you allow userspace (root) access to all
2159 io-memory regardless of whether a driver is actively using that
2160 range. Accidental access to this is obviously disastrous, but
2161 specific access can be used by people debugging kernel drivers.
2163 If this option is switched on, the /dev/mem file only allows
2164 userspace access to *idle* io-memory ranges (see /proc/iomem) This
2165 may break traditional users of /dev/mem (dosemu, legacy X, etc...)
2166 if the driver using a given range cannot be disabled.
2170 menu "$(SRCARCH) Debugging"
2172 source "arch/$(SRCARCH)/Kconfig.debug"
2176 config HYPERV_TESTING
2177 bool "Microsoft Hyper-V driver testing"
2179 depends on HYPERV && DEBUG_FS
2181 Select this option to enable Hyper-V vmbus testing.
2183 endmenu # Kernel hacking