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 endmenu # "printk and dmesg options"
169 menu "Compile-time checks and compiler options"
172 bool "Compile the kernel with debug info"
173 depends on DEBUG_KERNEL && !COMPILE_TEST
175 If you say Y here the resulting kernel image will include
176 debugging info resulting in a larger kernel image.
177 This adds debug symbols to the kernel and modules (gcc -g), and
178 is needed if you intend to use kernel crashdump or binary object
179 tools like crash, kgdb, LKCD, gdb, etc on the kernel.
180 Say Y here only if you plan to debug the kernel.
184 config DEBUG_INFO_REDUCED
185 bool "Reduce debugging information"
186 depends on DEBUG_INFO
188 If you say Y here gcc is instructed to generate less debugging
189 information for structure types. This means that tools that
190 need full debugging information (like kgdb or systemtap) won't
191 be happy. But if you merely need debugging information to
192 resolve line numbers there is no loss. Advantage is that
193 build directory object sizes shrink dramatically over a full
194 DEBUG_INFO build and compile times are reduced too.
195 Only works with newer gcc versions.
197 config DEBUG_INFO_SPLIT
198 bool "Produce split debuginfo in .dwo files"
199 depends on DEBUG_INFO
200 depends on $(cc-option,-gsplit-dwarf)
202 Generate debug info into separate .dwo files. This significantly
203 reduces the build directory size for builds with DEBUG_INFO,
204 because it stores the information only once on disk in .dwo
205 files instead of multiple times in object files and executables.
206 In addition the debug information is also compressed.
208 Requires recent gcc (4.7+) and recent gdb/binutils.
209 Any tool that packages or reads debug information would need
210 to know about the .dwo files and include them.
211 Incompatible with older versions of ccache.
213 config DEBUG_INFO_DWARF4
214 bool "Generate dwarf4 debuginfo"
215 depends on DEBUG_INFO
216 depends on $(cc-option,-gdwarf-4)
218 Generate dwarf4 debug info. This requires recent versions
219 of gcc and gdb. It makes the debug information larger.
220 But it significantly improves the success of resolving
221 variables in gdb on optimized code.
223 config DEBUG_INFO_BTF
224 bool "Generate BTF typeinfo"
225 depends on DEBUG_INFO
227 Generate deduplicated BTF type information from DWARF debug info.
228 Turning this on expects presence of pahole tool, which will convert
229 DWARF type info into equivalent deduplicated BTF type info.
232 bool "Provide GDB scripts for kernel debugging"
233 depends on DEBUG_INFO
235 This creates the required links to GDB helper scripts in the
236 build directory. If you load vmlinux into gdb, the helper
237 scripts will be automatically imported by gdb as well, and
238 additional functions are available to analyze a Linux kernel
239 instance. See Documentation/dev-tools/gdb-kernel-debugging.rst
242 config ENABLE_MUST_CHECK
243 bool "Enable __must_check logic"
246 Enable the __must_check logic in the kernel build. Disable this to
247 suppress the "warning: ignoring return value of 'foo', declared with
248 attribute warn_unused_result" messages.
251 int "Warn for stack frames larger than (needs gcc 4.4)"
253 default 2048 if GCC_PLUGIN_LATENT_ENTROPY
254 default 1280 if (!64BIT && PARISC)
255 default 1024 if (!64BIT && !PARISC)
256 default 2048 if 64BIT
258 Tell gcc to warn at build time for stack frames larger than this.
259 Setting this too low will cause a lot of warnings.
260 Setting it to 0 disables the warning.
263 config STRIP_ASM_SYMS
264 bool "Strip assembler-generated symbols during link"
267 Strip internal assembler-generated symbols during a link (symbols
268 that look like '.Lxxx') so they don't pollute the output of
269 get_wchan() and suchlike.
272 bool "Generate readable assembler code"
273 depends on DEBUG_KERNEL
275 Disable some compiler optimizations that tend to generate human unreadable
276 assembler output. This may make the kernel slightly slower, but it helps
277 to keep kernel developers who have to stare a lot at assembler listings
280 config UNUSED_SYMBOLS
281 bool "Enable unused/obsolete exported symbols"
284 Unused but exported symbols make the kernel needlessly bigger. For
285 that reason most of these unused exports will soon be removed. This
286 option is provided temporarily to provide a transition period in case
287 some external kernel module needs one of these symbols anyway. If you
288 encounter such a case in your module, consider if you are actually
289 using the right API. (rationale: since nobody in the kernel is using
290 this in a module, there is a pretty good chance it's actually the
291 wrong interface to use). If you really need the symbol, please send a
292 mail to the linux kernel mailing list mentioning the symbol and why
293 you really need it, and what the merge plan to the mainline kernel for
297 bool "Debug Filesystem"
299 debugfs is a virtual file system that kernel developers use to put
300 debugging files into. Enable this option to be able to read and
301 write to these files.
303 For detailed documentation on the debugfs API, see
304 Documentation/filesystems/.
309 bool "Run 'make headers_check' when building vmlinux"
312 This option will extract the user-visible kernel headers whenever
313 building the kernel, and will run basic sanity checks on them to
314 ensure that exported files do not attempt to include files which
315 were not exported, etc.
317 If you're making modifications to header files which are
318 relevant for userspace, say 'Y', and check the headers
319 exported to $(INSTALL_HDR_PATH) (usually 'usr/include' in
320 your build tree), to make sure they're suitable.
322 config OPTIMIZE_INLINING
323 bool "Allow compiler to uninline functions marked 'inline'"
325 This option determines if the kernel forces gcc to inline the functions
326 developers have marked 'inline'. Doing so takes away freedom from gcc to
327 do what it thinks is best, which is desirable for the gcc 3.x series of
328 compilers. The gcc 4.x series have a rewritten inlining algorithm and
329 enabling this option will generate a smaller kernel there. Hopefully
330 this algorithm is so good that allowing gcc 4.x and above to make the
331 decision will become the default in the future. Until then this option
332 is there to test gcc for this.
336 config DEBUG_SECTION_MISMATCH
337 bool "Enable full Section mismatch analysis"
339 The section mismatch analysis checks if there are illegal
340 references from one section to another section.
341 During linktime or runtime, some sections are dropped;
342 any use of code/data previously in these sections would
343 most likely result in an oops.
344 In the code, functions and variables are annotated with
345 __init,, etc. (see the full list in include/linux/init.h),
346 which results in the code/data being placed in specific sections.
347 The section mismatch analysis is always performed after a full
348 kernel build, and enabling this option causes the following
349 additional steps to occur:
350 - Add the option -fno-inline-functions-called-once to gcc commands.
351 When inlining a function annotated with __init in a non-init
352 function, we would lose the section information and thus
353 the analysis would not catch the illegal reference.
354 This option tells gcc to inline less (but it does result in
356 - Run the section mismatch analysis for each module/built-in.a file.
357 When we run the section mismatch analysis on vmlinux.o, we
358 lose valuable information about where the mismatch was
360 Running the analysis for each module/built-in.a file
361 tells where the mismatch happens much closer to the
362 source. The drawback is that the same mismatch is
363 reported at least twice.
364 - Enable verbose reporting from modpost in order to help resolve
365 the section mismatches that are reported.
367 config SECTION_MISMATCH_WARN_ONLY
368 bool "Make section mismatch errors non-fatal"
371 If you say N here, the build process will fail if there are any
372 section mismatch, instead of just throwing warnings.
377 # Select this config option from the architecture Kconfig, if it
378 # is preferred to always offer frame pointers as a config
379 # option on the architecture (regardless of KERNEL_DEBUG):
381 config ARCH_WANT_FRAME_POINTERS
385 bool "Compile the kernel with frame pointers"
386 depends on DEBUG_KERNEL && (M68K || UML || SUPERH) || ARCH_WANT_FRAME_POINTERS
387 default y if (DEBUG_INFO && UML) || ARCH_WANT_FRAME_POINTERS
389 If you say Y here the resulting kernel image will be slightly
390 larger and slower, but it gives very useful debugging information
391 in case of kernel bugs. (precise oopses/stacktraces/warnings)
393 config STACK_VALIDATION
394 bool "Compile-time stack metadata validation"
395 depends on HAVE_STACK_VALIDATION
398 Add compile-time checks to validate stack metadata, including frame
399 pointers (if CONFIG_FRAME_POINTER is enabled). This helps ensure
400 that runtime stack traces are more reliable.
402 This is also a prerequisite for generation of ORC unwind data, which
403 is needed for CONFIG_UNWINDER_ORC.
405 For more information, see
406 tools/objtool/Documentation/stack-validation.txt.
408 config DEBUG_FORCE_WEAK_PER_CPU
409 bool "Force weak per-cpu definitions"
410 depends on DEBUG_KERNEL
412 s390 and alpha require percpu variables in modules to be
413 defined weak to work around addressing range issue which
414 puts the following two restrictions on percpu variable
417 1. percpu symbols must be unique whether static or not
418 2. percpu variables can't be defined inside a function
420 To ensure that generic code follows the above rules, this
421 option forces all percpu variables to be defined as weak.
423 endmenu # "Compiler options"
426 bool "Magic SysRq key"
429 If you say Y here, you will have some control over the system even
430 if the system crashes for example during kernel debugging (e.g., you
431 will be able to flush the buffer cache to disk, reboot the system
432 immediately or dump some status information). This is accomplished
433 by pressing various keys while holding SysRq (Alt+PrintScreen). It
434 also works on a serial console (on PC hardware at least), if you
435 send a BREAK and then within 5 seconds a command keypress. The
436 keys are documented in <file:Documentation/admin-guide/sysrq.rst>.
437 Don't say Y unless you really know what this hack does.
439 config MAGIC_SYSRQ_DEFAULT_ENABLE
440 hex "Enable magic SysRq key functions by default"
441 depends on MAGIC_SYSRQ
444 Specifies which SysRq key functions are enabled by default.
445 This may be set to 1 or 0 to enable or disable them all, or
446 to a bitmask as described in Documentation/admin-guide/sysrq.rst.
448 config MAGIC_SYSRQ_SERIAL
449 bool "Enable magic SysRq key over serial"
450 depends on MAGIC_SYSRQ
453 Many embedded boards have a disconnected TTL level serial which can
454 generate some garbage that can lead to spurious false sysrq detects.
455 This option allows you to decide whether you want to enable the
459 bool "Kernel debugging"
461 Say Y here if you are developing drivers or trying to debug and
462 identify kernel problems.
465 bool "Miscellaneous debug code"
467 depends on DEBUG_KERNEL
469 Say Y here if you need to enable miscellaneous debug code that should
470 be under a more specific debug option but isn't.
473 menu "Memory Debugging"
475 source "mm/Kconfig.debug"
478 bool "Debug object operations"
479 depends on DEBUG_KERNEL
481 If you say Y here, additional code will be inserted into the
482 kernel to track the life time of various objects and validate
483 the operations on those objects.
485 config DEBUG_OBJECTS_SELFTEST
486 bool "Debug objects selftest"
487 depends on DEBUG_OBJECTS
489 This enables the selftest of the object debug code.
491 config DEBUG_OBJECTS_FREE
492 bool "Debug objects in freed memory"
493 depends on DEBUG_OBJECTS
495 This enables checks whether a k/v free operation frees an area
496 which contains an object which has not been deactivated
497 properly. This can make kmalloc/kfree-intensive workloads
500 config DEBUG_OBJECTS_TIMERS
501 bool "Debug timer objects"
502 depends on DEBUG_OBJECTS
504 If you say Y here, additional code will be inserted into the
505 timer routines to track the life time of timer objects and
506 validate the timer operations.
508 config DEBUG_OBJECTS_WORK
509 bool "Debug work objects"
510 depends on DEBUG_OBJECTS
512 If you say Y here, additional code will be inserted into the
513 work queue routines to track the life time of work objects and
514 validate the work operations.
516 config DEBUG_OBJECTS_RCU_HEAD
517 bool "Debug RCU callbacks objects"
518 depends on DEBUG_OBJECTS
520 Enable this to turn on debugging of RCU list heads (call_rcu() usage).
522 config DEBUG_OBJECTS_PERCPU_COUNTER
523 bool "Debug percpu counter objects"
524 depends on DEBUG_OBJECTS
526 If you say Y here, additional code will be inserted into the
527 percpu counter routines to track the life time of percpu counter
528 objects and validate the percpu counter operations.
530 config DEBUG_OBJECTS_ENABLE_DEFAULT
531 int "debug_objects bootup default value (0-1)"
534 depends on DEBUG_OBJECTS
536 Debug objects boot parameter default value
539 bool "Debug slab memory allocations"
540 depends on DEBUG_KERNEL && SLAB
542 Say Y here to have the kernel do limited verification on memory
543 allocation as well as poisoning memory on free to catch use of freed
544 memory. This can make kmalloc/kfree-intensive workloads much slower.
547 bool "SLUB debugging on by default"
548 depends on SLUB && SLUB_DEBUG
551 Boot with debugging on by default. SLUB boots by default with
552 the runtime debug capabilities switched off. Enabling this is
553 equivalent to specifying the "slub_debug" parameter on boot.
554 There is no support for more fine grained debug control like
555 possible with slub_debug=xxx. SLUB debugging may be switched
556 off in a kernel built with CONFIG_SLUB_DEBUG_ON by specifying
561 bool "Enable SLUB performance statistics"
562 depends on SLUB && SYSFS
564 SLUB statistics are useful to debug SLUBs allocation behavior in
565 order find ways to optimize the allocator. This should never be
566 enabled for production use since keeping statistics slows down
567 the allocator by a few percentage points. The slabinfo command
568 supports the determination of the most active slabs to figure
569 out which slabs are relevant to a particular load.
570 Try running: slabinfo -DA
572 config HAVE_DEBUG_KMEMLEAK
575 config DEBUG_KMEMLEAK
576 bool "Kernel memory leak detector"
577 depends on DEBUG_KERNEL && HAVE_DEBUG_KMEMLEAK
579 select STACKTRACE if STACKTRACE_SUPPORT
583 Say Y here if you want to enable the memory leak
584 detector. The memory allocation/freeing is traced in a way
585 similar to the Boehm's conservative garbage collector, the
586 difference being that the orphan objects are not freed but
587 only shown in /sys/kernel/debug/kmemleak. Enabling this
588 feature will introduce an overhead to memory
589 allocations. See Documentation/dev-tools/kmemleak.rst for more
592 Enabling DEBUG_SLAB or SLUB_DEBUG may increase the chances
593 of finding leaks due to the slab objects poisoning.
595 In order to access the kmemleak file, debugfs needs to be
596 mounted (usually at /sys/kernel/debug).
598 config DEBUG_KMEMLEAK_EARLY_LOG_SIZE
599 int "Maximum kmemleak early log entries"
600 depends on DEBUG_KMEMLEAK
604 Kmemleak must track all the memory allocations to avoid
605 reporting false positives. Since memory may be allocated or
606 freed before kmemleak is initialised, an early log buffer is
607 used to store these actions. If kmemleak reports "early log
608 buffer exceeded", please increase this value.
610 config DEBUG_KMEMLEAK_TEST
611 tristate "Simple test for the kernel memory leak detector"
612 depends on DEBUG_KMEMLEAK && m
614 This option enables a module that explicitly leaks memory.
618 config DEBUG_KMEMLEAK_DEFAULT_OFF
619 bool "Default kmemleak to off"
620 depends on DEBUG_KMEMLEAK
622 Say Y here to disable kmemleak by default. It can then be enabled
623 on the command line via kmemleak=on.
625 config DEBUG_KMEMLEAK_AUTO_SCAN
626 bool "Enable kmemleak auto scan thread on boot up"
628 depends on DEBUG_KMEMLEAK
630 Depending on the cpu, kmemleak scan may be cpu intensive and can
631 stall user tasks at times. This option enables/disables automatic
632 kmemleak scan at boot up.
634 Say N here to disable kmemleak auto scan thread to stop automatic
635 scanning. Disabling this option disables automatic reporting of
640 config DEBUG_STACK_USAGE
641 bool "Stack utilization instrumentation"
642 depends on DEBUG_KERNEL && !IA64
644 Enables the display of the minimum amount of free stack which each
645 task has ever had available in the sysrq-T and sysrq-P debug output.
647 This option will slow down process creation somewhat.
651 depends on DEBUG_KERNEL
653 Enable this to turn on extended checks in the virtual-memory system
654 that may impact performance.
658 config DEBUG_VM_VMACACHE
659 bool "Debug VMA caching"
662 Enable this to turn on VMA caching debug information. Doing so
663 can cause significant overhead, so only enable it in non-production
669 bool "Debug VM red-black trees"
672 Enable VM red-black tree debugging information and extra validations.
676 config DEBUG_VM_PGFLAGS
677 bool "Debug page-flags operations"
680 Enables extra validation on page flags operations.
684 config ARCH_HAS_DEBUG_VIRTUAL
688 bool "Debug VM translations"
689 depends on DEBUG_KERNEL && ARCH_HAS_DEBUG_VIRTUAL
691 Enable some costly sanity checks in virtual to page code. This can
692 catch mistakes with virt_to_page() and friends.
696 config DEBUG_NOMMU_REGIONS
697 bool "Debug the global anon/private NOMMU mapping region tree"
698 depends on DEBUG_KERNEL && !MMU
700 This option causes the global tree of anonymous and private mapping
701 regions to be regularly checked for invalid topology.
703 config DEBUG_MEMORY_INIT
704 bool "Debug memory initialisation" if EXPERT
707 Enable this for additional checks during memory initialisation.
708 The sanity checks verify aspects of the VM such as the memory model
709 and other information provided by the architecture. Verbose
710 information will be printed at KERN_DEBUG loglevel depending
711 on the mminit_loglevel= command-line option.
715 config MEMORY_NOTIFIER_ERROR_INJECT
716 tristate "Memory hotplug notifier error injection module"
717 depends on MEMORY_HOTPLUG_SPARSE && NOTIFIER_ERROR_INJECTION
719 This option provides the ability to inject artificial errors to
720 memory hotplug notifier chain callbacks. It is controlled through
721 debugfs interface under /sys/kernel/debug/notifier-error-inject/memory
723 If the notifier call chain should be failed with some events
724 notified, write the error code to "actions/<notifier event>/error".
726 Example: Inject memory hotplug offline error (-12 == -ENOMEM)
728 # cd /sys/kernel/debug/notifier-error-inject/memory
729 # echo -12 > actions/MEM_GOING_OFFLINE/error
730 # echo offline > /sys/devices/system/memory/memoryXXX/state
731 bash: echo: write error: Cannot allocate memory
733 To compile this code as a module, choose M here: the module will
734 be called memory-notifier-error-inject.
738 config DEBUG_PER_CPU_MAPS
739 bool "Debug access to per_cpu maps"
740 depends on DEBUG_KERNEL
743 Say Y to verify that the per_cpu map being accessed has
744 been set up. This adds a fair amount of code to kernel memory
745 and decreases performance.
750 bool "Highmem debugging"
751 depends on DEBUG_KERNEL && HIGHMEM
753 This option enables additional error checking for high memory
754 systems. Disable for production systems.
756 config HAVE_DEBUG_STACKOVERFLOW
759 config DEBUG_STACKOVERFLOW
760 bool "Check for stack overflows"
761 depends on DEBUG_KERNEL && HAVE_DEBUG_STACKOVERFLOW
763 Say Y here if you want to check for overflows of kernel, IRQ
764 and exception stacks (if your architecture uses them). This
765 option will show detailed messages if free stack space drops
766 below a certain limit.
768 These kinds of bugs usually occur when call-chains in the
769 kernel get too deep, especially when interrupts are
772 Use this in cases where you see apparently random memory
773 corruption, especially if it appears in 'struct thread_info'
775 If in doubt, say "N".
777 source "lib/Kconfig.kasan"
779 endmenu # "Memory Debugging"
784 An architecture should select this when it can successfully
785 build and run with CONFIG_KCOV. This typically requires
786 disabling instrumentation for some early boot code.
788 config CC_HAS_SANCOV_TRACE_PC
789 def_bool $(cc-option,-fsanitize-coverage=trace-pc)
792 bool "Code coverage for fuzzing"
793 depends on ARCH_HAS_KCOV
794 depends on CC_HAS_SANCOV_TRACE_PC || GCC_PLUGINS
796 select GCC_PLUGIN_SANCOV if !CC_HAS_SANCOV_TRACE_PC
798 KCOV exposes kernel code coverage information in a form suitable
799 for coverage-guided fuzzing (randomized testing).
801 If RANDOMIZE_BASE is enabled, PC values will not be stable across
802 different machines and across reboots. If you need stable PC values,
803 disable RANDOMIZE_BASE.
805 For more details, see Documentation/dev-tools/kcov.rst.
807 config KCOV_ENABLE_COMPARISONS
808 bool "Enable comparison operands collection by KCOV"
810 depends on $(cc-option,-fsanitize-coverage=trace-cmp)
812 KCOV also exposes operands of every comparison in the instrumented
813 code along with operand sizes and PCs of the comparison instructions.
814 These operands can be used by fuzzing engines to improve the quality
817 config KCOV_INSTRUMENT_ALL
818 bool "Instrument all code by default"
822 If you are doing generic system call fuzzing (like e.g. syzkaller),
823 then you will want to instrument the whole kernel and you should
824 say y here. If you are doing more targeted fuzzing (like e.g.
825 filesystem fuzzing with AFL) then you will want to enable coverage
826 for more specific subsets of files, and should say n here.
829 bool "Debug shared IRQ handlers"
830 depends on DEBUG_KERNEL
832 Enable this to generate a spurious interrupt as soon as a shared
833 interrupt handler is registered, and just before one is deregistered.
834 Drivers ought to be able to handle interrupts coming in at those
835 points; some don't and need to be caught.
837 menu "Debug Lockups and Hangs"
839 config LOCKUP_DETECTOR
842 config SOFTLOCKUP_DETECTOR
843 bool "Detect Soft Lockups"
844 depends on DEBUG_KERNEL && !S390
845 select LOCKUP_DETECTOR
847 Say Y here to enable the kernel to act as a watchdog to detect
850 Softlockups are bugs that cause the kernel to loop in kernel
851 mode for more than 20 seconds, without giving other tasks a
852 chance to run. The current stack trace is displayed upon
853 detection and the system will stay locked up.
855 config BOOTPARAM_SOFTLOCKUP_PANIC
856 bool "Panic (Reboot) On Soft Lockups"
857 depends on SOFTLOCKUP_DETECTOR
859 Say Y here to enable the kernel to panic on "soft lockups",
860 which are bugs that cause the kernel to loop in kernel
861 mode for more than 20 seconds (configurable using the watchdog_thresh
862 sysctl), without giving other tasks a chance to run.
864 The panic can be used in combination with panic_timeout,
865 to cause the system to reboot automatically after a
866 lockup has been detected. This feature is useful for
867 high-availability systems that have uptime guarantees and
868 where a lockup must be resolved ASAP.
872 config BOOTPARAM_SOFTLOCKUP_PANIC_VALUE
874 depends on SOFTLOCKUP_DETECTOR
876 default 0 if !BOOTPARAM_SOFTLOCKUP_PANIC
877 default 1 if BOOTPARAM_SOFTLOCKUP_PANIC
879 config HARDLOCKUP_DETECTOR_PERF
881 select SOFTLOCKUP_DETECTOR
884 # Enables a timestamp based low pass filter to compensate for perf based
885 # hard lockup detection which runs too fast due to turbo modes.
887 config HARDLOCKUP_CHECK_TIMESTAMP
891 # arch/ can define HAVE_HARDLOCKUP_DETECTOR_ARCH to provide their own hard
892 # lockup detector rather than the perf based detector.
894 config HARDLOCKUP_DETECTOR
895 bool "Detect Hard Lockups"
896 depends on DEBUG_KERNEL && !S390
897 depends on HAVE_HARDLOCKUP_DETECTOR_PERF || HAVE_HARDLOCKUP_DETECTOR_ARCH
898 select LOCKUP_DETECTOR
899 select HARDLOCKUP_DETECTOR_PERF if HAVE_HARDLOCKUP_DETECTOR_PERF
900 select HARDLOCKUP_DETECTOR_ARCH if HAVE_HARDLOCKUP_DETECTOR_ARCH
902 Say Y here to enable the kernel to act as a watchdog to detect
905 Hardlockups are bugs that cause the CPU to loop in kernel mode
906 for more than 10 seconds, without letting other interrupts have a
907 chance to run. The current stack trace is displayed upon detection
908 and the system will stay locked up.
910 config BOOTPARAM_HARDLOCKUP_PANIC
911 bool "Panic (Reboot) On Hard Lockups"
912 depends on HARDLOCKUP_DETECTOR
914 Say Y here to enable the kernel to panic on "hard lockups",
915 which are bugs that cause the kernel to loop in kernel
916 mode with interrupts disabled for more than 10 seconds (configurable
917 using the watchdog_thresh sysctl).
921 config BOOTPARAM_HARDLOCKUP_PANIC_VALUE
923 depends on HARDLOCKUP_DETECTOR
925 default 0 if !BOOTPARAM_HARDLOCKUP_PANIC
926 default 1 if BOOTPARAM_HARDLOCKUP_PANIC
928 config DETECT_HUNG_TASK
929 bool "Detect Hung Tasks"
930 depends on DEBUG_KERNEL
931 default SOFTLOCKUP_DETECTOR
933 Say Y here to enable the kernel to detect "hung tasks",
934 which are bugs that cause the task to be stuck in
935 uninterruptible "D" state indefinitely.
937 When a hung task is detected, the kernel will print the
938 current stack trace (which you should report), but the
939 task will stay in uninterruptible state. If lockdep is
940 enabled then all held locks will also be reported. This
941 feature has negligible overhead.
943 config DEFAULT_HUNG_TASK_TIMEOUT
944 int "Default timeout for hung task detection (in seconds)"
945 depends on DETECT_HUNG_TASK
948 This option controls the default timeout (in seconds) used
949 to determine when a task has become non-responsive and should
952 It can be adjusted at runtime via the kernel.hung_task_timeout_secs
953 sysctl or by writing a value to
954 /proc/sys/kernel/hung_task_timeout_secs.
956 A timeout of 0 disables the check. The default is two minutes.
957 Keeping the default should be fine in most cases.
959 config BOOTPARAM_HUNG_TASK_PANIC
960 bool "Panic (Reboot) On Hung Tasks"
961 depends on DETECT_HUNG_TASK
963 Say Y here to enable the kernel to panic on "hung tasks",
964 which are bugs that cause the kernel to leave a task stuck
965 in uninterruptible "D" state.
967 The panic can be used in combination with panic_timeout,
968 to cause the system to reboot automatically after a
969 hung task has been detected. This feature is useful for
970 high-availability systems that have uptime guarantees and
971 where a hung tasks must be resolved ASAP.
975 config BOOTPARAM_HUNG_TASK_PANIC_VALUE
977 depends on DETECT_HUNG_TASK
979 default 0 if !BOOTPARAM_HUNG_TASK_PANIC
980 default 1 if BOOTPARAM_HUNG_TASK_PANIC
983 bool "Detect Workqueue Stalls"
984 depends on DEBUG_KERNEL
986 Say Y here to enable stall detection on workqueues. If a
987 worker pool doesn't make forward progress on a pending work
988 item for over a given amount of time, 30s by default, a
989 warning message is printed along with dump of workqueue
990 state. This can be configured through kernel parameter
991 "workqueue.watchdog_thresh" and its sysfs counterpart.
993 endmenu # "Debug lockups and hangs"
998 Say Y here to enable the kernel to panic when it oopses. This
999 has the same effect as setting oops=panic on the kernel command
1002 This feature is useful to ensure that the kernel does not do
1003 anything erroneous after an oops which could result in data
1004 corruption or other issues.
1008 config PANIC_ON_OOPS_VALUE
1011 default 0 if !PANIC_ON_OOPS
1012 default 1 if PANIC_ON_OOPS
1014 config PANIC_TIMEOUT
1018 Set the timeout value (in seconds) until a reboot occurs when the
1019 the kernel panics. If n = 0, then we wait forever. A timeout
1020 value n > 0 will wait n seconds before rebooting, while a timeout
1021 value n < 0 will reboot immediately.
1024 bool "Collect scheduler debugging info"
1025 depends on DEBUG_KERNEL && PROC_FS
1028 If you say Y here, the /proc/sched_debug file will be provided
1029 that can help debug the scheduler. The runtime overhead of this
1037 bool "Collect scheduler statistics"
1038 depends on DEBUG_KERNEL && PROC_FS
1041 If you say Y here, additional code will be inserted into the
1042 scheduler and related routines to collect statistics about
1043 scheduler behavior and provide them in /proc/schedstat. These
1044 stats may be useful for both tuning and debugging the scheduler
1045 If you aren't debugging the scheduler or trying to tune a specific
1046 application, you can say N to avoid the very slight overhead
1049 config SCHED_STACK_END_CHECK
1050 bool "Detect stack corruption on calls to schedule()"
1051 depends on DEBUG_KERNEL
1054 This option checks for a stack overrun on calls to schedule().
1055 If the stack end location is found to be over written always panic as
1056 the content of the corrupted region can no longer be trusted.
1057 This is to ensure no erroneous behaviour occurs which could result in
1058 data corruption or a sporadic crash at a later stage once the region
1059 is examined. The runtime overhead introduced is minimal.
1061 config DEBUG_TIMEKEEPING
1062 bool "Enable extra timekeeping sanity checking"
1064 This option will enable additional timekeeping sanity checks
1065 which may be helpful when diagnosing issues where timekeeping
1066 problems are suspected.
1068 This may include checks in the timekeeping hotpaths, so this
1069 option may have a (very small) performance impact to some
1074 config DEBUG_PREEMPT
1075 bool "Debug preemptible kernel"
1076 depends on DEBUG_KERNEL && PREEMPT && TRACE_IRQFLAGS_SUPPORT
1079 If you say Y here then the kernel will use a debug variant of the
1080 commonly used smp_processor_id() function and will print warnings
1081 if kernel code uses it in a preemption-unsafe way. Also, the kernel
1082 will detect preemption count underflows.
1084 menu "Lock Debugging (spinlocks, mutexes, etc...)"
1086 config LOCK_DEBUGGING_SUPPORT
1088 depends on TRACE_IRQFLAGS_SUPPORT && STACKTRACE_SUPPORT && LOCKDEP_SUPPORT
1091 config PROVE_LOCKING
1092 bool "Lock debugging: prove locking correctness"
1093 depends on DEBUG_KERNEL && LOCK_DEBUGGING_SUPPORT
1095 select DEBUG_SPINLOCK
1096 select DEBUG_MUTEXES
1097 select DEBUG_RT_MUTEXES if RT_MUTEXES
1098 select DEBUG_RWSEMS if RWSEM_SPIN_ON_OWNER
1099 select DEBUG_WW_MUTEX_SLOWPATH
1100 select DEBUG_LOCK_ALLOC
1101 select TRACE_IRQFLAGS
1104 This feature enables the kernel to prove that all locking
1105 that occurs in the kernel runtime is mathematically
1106 correct: that under no circumstance could an arbitrary (and
1107 not yet triggered) combination of observed locking
1108 sequences (on an arbitrary number of CPUs, running an
1109 arbitrary number of tasks and interrupt contexts) cause a
1112 In short, this feature enables the kernel to report locking
1113 related deadlocks before they actually occur.
1115 The proof does not depend on how hard and complex a
1116 deadlock scenario would be to trigger: how many
1117 participant CPUs, tasks and irq-contexts would be needed
1118 for it to trigger. The proof also does not depend on
1119 timing: if a race and a resulting deadlock is possible
1120 theoretically (no matter how unlikely the race scenario
1121 is), it will be proven so and will immediately be
1122 reported by the kernel (once the event is observed that
1123 makes the deadlock theoretically possible).
1125 If a deadlock is impossible (i.e. the locking rules, as
1126 observed by the kernel, are mathematically correct), the
1127 kernel reports nothing.
1129 NOTE: this feature can also be enabled for rwlocks, mutexes
1130 and rwsems - in which case all dependencies between these
1131 different locking variants are observed and mapped too, and
1132 the proof of observed correctness is also maintained for an
1133 arbitrary combination of these separate locking variants.
1135 For more details, see Documentation/locking/lockdep-design.txt.
1138 bool "Lock usage statistics"
1139 depends on DEBUG_KERNEL && LOCK_DEBUGGING_SUPPORT
1141 select DEBUG_SPINLOCK
1142 select DEBUG_MUTEXES
1143 select DEBUG_RT_MUTEXES if RT_MUTEXES
1144 select DEBUG_LOCK_ALLOC
1147 This feature enables tracking lock contention points
1149 For more details, see Documentation/locking/lockstat.txt
1151 This also enables lock events required by "perf lock",
1153 If you want to use "perf lock", you also need to turn on
1154 CONFIG_EVENT_TRACING.
1156 CONFIG_LOCK_STAT defines "contended" and "acquired" lock events.
1157 (CONFIG_LOCKDEP defines "acquire" and "release" events.)
1159 config DEBUG_RT_MUTEXES
1160 bool "RT Mutex debugging, deadlock detection"
1161 depends on DEBUG_KERNEL && RT_MUTEXES
1163 This allows rt mutex semantics violations and rt mutex related
1164 deadlocks (lockups) to be detected and reported automatically.
1166 config DEBUG_SPINLOCK
1167 bool "Spinlock and rw-lock debugging: basic checks"
1168 depends on DEBUG_KERNEL
1169 select UNINLINE_SPIN_UNLOCK
1171 Say Y here and build SMP to catch missing spinlock initialization
1172 and certain other kinds of spinlock errors commonly made. This is
1173 best used in conjunction with the NMI watchdog so that spinlock
1174 deadlocks are also debuggable.
1176 config DEBUG_MUTEXES
1177 bool "Mutex debugging: basic checks"
1178 depends on DEBUG_KERNEL
1180 This feature allows mutex semantics violations to be detected and
1183 config DEBUG_WW_MUTEX_SLOWPATH
1184 bool "Wait/wound mutex debugging: Slowpath testing"
1185 depends on DEBUG_KERNEL && LOCK_DEBUGGING_SUPPORT
1186 select DEBUG_LOCK_ALLOC
1187 select DEBUG_SPINLOCK
1188 select DEBUG_MUTEXES
1190 This feature enables slowpath testing for w/w mutex users by
1191 injecting additional -EDEADLK wound/backoff cases. Together with
1192 the full mutex checks enabled with (CONFIG_PROVE_LOCKING) this
1193 will test all possible w/w mutex interface abuse with the
1194 exception of simply not acquiring all the required locks.
1195 Note that this feature can introduce significant overhead, so
1196 it really should not be enabled in a production or distro kernel,
1197 even a debug kernel. If you are a driver writer, enable it. If
1198 you are a distro, do not.
1201 bool "RW Semaphore debugging: basic checks"
1202 depends on DEBUG_KERNEL && RWSEM_SPIN_ON_OWNER
1204 This debugging feature allows mismatched rw semaphore locks and unlocks
1205 to be detected and reported.
1207 config DEBUG_LOCK_ALLOC
1208 bool "Lock debugging: detect incorrect freeing of live locks"
1209 depends on DEBUG_KERNEL && LOCK_DEBUGGING_SUPPORT
1210 select DEBUG_SPINLOCK
1211 select DEBUG_MUTEXES
1212 select DEBUG_RT_MUTEXES if RT_MUTEXES
1215 This feature will check whether any held lock (spinlock, rwlock,
1216 mutex or rwsem) is incorrectly freed by the kernel, via any of the
1217 memory-freeing routines (kfree(), kmem_cache_free(), free_pages(),
1218 vfree(), etc.), whether a live lock is incorrectly reinitialized via
1219 spin_lock_init()/mutex_init()/etc., or whether there is any lock
1220 held during task exit.
1224 depends on DEBUG_KERNEL && LOCK_DEBUGGING_SUPPORT
1226 select FRAME_POINTER if !MIPS && !PPC && !ARM && !S390 && !MICROBLAZE && !ARC && !X86
1230 config LOCKDEP_SMALL
1233 config DEBUG_LOCKDEP
1234 bool "Lock dependency engine debugging"
1235 depends on DEBUG_KERNEL && LOCKDEP
1237 If you say Y here, the lock dependency engine will do
1238 additional runtime checks to debug itself, at the price
1239 of more runtime overhead.
1241 config DEBUG_ATOMIC_SLEEP
1242 bool "Sleep inside atomic section checking"
1243 select PREEMPT_COUNT
1244 depends on DEBUG_KERNEL
1245 depends on !ARCH_NO_PREEMPT
1247 If you say Y here, various routines which may sleep will become very
1248 noisy if they are called inside atomic sections: when a spinlock is
1249 held, inside an rcu read side critical section, inside preempt disabled
1250 sections, inside an interrupt, etc...
1252 config DEBUG_LOCKING_API_SELFTESTS
1253 bool "Locking API boot-time self-tests"
1254 depends on DEBUG_KERNEL
1256 Say Y here if you want the kernel to run a short self-test during
1257 bootup. The self-test checks whether common types of locking bugs
1258 are detected by debugging mechanisms or not. (if you disable
1259 lock debugging then those bugs wont be detected of course.)
1260 The following locking APIs are covered: spinlocks, rwlocks,
1263 config LOCK_TORTURE_TEST
1264 tristate "torture tests for locking"
1265 depends on DEBUG_KERNEL
1268 This option provides a kernel module that runs torture tests
1269 on kernel locking primitives. The kernel module may be built
1270 after the fact on the running kernel to be tested, if desired.
1272 Say Y here if you want kernel locking-primitive torture tests
1273 to be built into the kernel.
1274 Say M if you want these torture tests to build as a module.
1275 Say N if you are unsure.
1277 config WW_MUTEX_SELFTEST
1278 tristate "Wait/wound mutex selftests"
1280 This option provides a kernel module that runs tests on the
1281 on the struct ww_mutex locking API.
1283 It is recommended to enable DEBUG_WW_MUTEX_SLOWPATH in conjunction
1284 with this test harness.
1286 Say M if you want these self tests to build as a module.
1287 Say N if you are unsure.
1289 endmenu # lock debugging
1291 config TRACE_IRQFLAGS
1294 Enables hooks to interrupt enabling and disabling for
1295 either tracing or lock debugging.
1298 bool "Stack backtrace support"
1299 depends on STACKTRACE_SUPPORT
1301 This option causes the kernel to create a /proc/pid/stack for
1302 every process, showing its current stack trace.
1303 It is also used by various kernel debugging features that require
1304 stack trace generation.
1306 config WARN_ALL_UNSEEDED_RANDOM
1307 bool "Warn for all uses of unseeded randomness"
1310 Some parts of the kernel contain bugs relating to their use of
1311 cryptographically secure random numbers before it's actually possible
1312 to generate those numbers securely. This setting ensures that these
1313 flaws don't go unnoticed, by enabling a message, should this ever
1314 occur. This will allow people with obscure setups to know when things
1315 are going wrong, so that they might contact developers about fixing
1318 Unfortunately, on some models of some architectures getting
1319 a fully seeded CRNG is extremely difficult, and so this can
1320 result in dmesg getting spammed for a surprisingly long
1321 time. This is really bad from a security perspective, and
1322 so architecture maintainers really need to do what they can
1323 to get the CRNG seeded sooner after the system is booted.
1324 However, since users cannot do anything actionable to
1325 address this, by default the kernel will issue only a single
1326 warning for the first use of unseeded randomness.
1328 Say Y here if you want to receive warnings for all uses of
1329 unseeded randomness. This will be of use primarily for
1330 those developers interested in improving the security of
1331 Linux kernels running on their architecture (or
1334 config DEBUG_KOBJECT
1335 bool "kobject debugging"
1336 depends on DEBUG_KERNEL
1338 If you say Y here, some extra kobject debugging messages will be sent
1341 config DEBUG_KOBJECT_RELEASE
1342 bool "kobject release debugging"
1343 depends on DEBUG_OBJECTS_TIMERS
1345 kobjects are reference counted objects. This means that their
1346 last reference count put is not predictable, and the kobject can
1347 live on past the point at which a driver decides to drop it's
1348 initial reference to the kobject gained on allocation. An
1349 example of this would be a struct device which has just been
1352 However, some buggy drivers assume that after such an operation,
1353 the memory backing the kobject can be immediately freed. This
1354 goes completely against the principles of a refcounted object.
1356 If you say Y here, the kernel will delay the release of kobjects
1357 on the last reference count to improve the visibility of this
1358 kind of kobject release bug.
1360 config HAVE_DEBUG_BUGVERBOSE
1363 config DEBUG_BUGVERBOSE
1364 bool "Verbose BUG() reporting (adds 70K)" if DEBUG_KERNEL && EXPERT
1365 depends on BUG && (GENERIC_BUG || HAVE_DEBUG_BUGVERBOSE)
1368 Say Y here to make BUG() panics output the file name and line number
1369 of the BUG call as well as the EIP and oops trace. This aids
1370 debugging but costs about 70-100K of memory.
1373 bool "Debug linked list manipulation"
1374 depends on DEBUG_KERNEL || BUG_ON_DATA_CORRUPTION
1376 Enable this to turn on extended checks in the linked-list
1382 bool "Debug priority linked list manipulation"
1383 depends on DEBUG_KERNEL
1385 Enable this to turn on extended checks in the priority-ordered
1386 linked-list (plist) walking routines. This checks the entire
1387 list multiple times during each manipulation.
1392 bool "Debug SG table operations"
1393 depends on DEBUG_KERNEL
1395 Enable this to turn on checks on scatter-gather tables. This can
1396 help find problems with drivers that do not properly initialize
1401 config DEBUG_NOTIFIERS
1402 bool "Debug notifier call chains"
1403 depends on DEBUG_KERNEL
1405 Enable this to turn on sanity checking for notifier call chains.
1406 This is most useful for kernel developers to make sure that
1407 modules properly unregister themselves from notifier chains.
1408 This is a relatively cheap check but if you care about maximum
1411 config DEBUG_CREDENTIALS
1412 bool "Debug credential management"
1413 depends on DEBUG_KERNEL
1415 Enable this to turn on some debug checking for credential
1416 management. The additional code keeps track of the number of
1417 pointers from task_structs to any given cred struct, and checks to
1418 see that this number never exceeds the usage count of the cred
1421 Furthermore, if SELinux is enabled, this also checks that the
1422 security pointer in the cred struct is never seen to be invalid.
1426 source "kernel/rcu/Kconfig.debug"
1428 config DEBUG_WQ_FORCE_RR_CPU
1429 bool "Force round-robin CPU selection for unbound work items"
1430 depends on DEBUG_KERNEL
1433 Workqueue used to implicitly guarantee that work items queued
1434 without explicit CPU specified are put on the local CPU. This
1435 guarantee is no longer true and while local CPU is still
1436 preferred work items may be put on foreign CPUs. Kernel
1437 parameter "workqueue.debug_force_rr_cpu" is added to force
1438 round-robin CPU selection to flush out usages which depend on the
1439 now broken guarantee. This config option enables the debug
1440 feature by default. When enabled, memory and cache locality will
1443 config DEBUG_BLOCK_EXT_DEVT
1444 bool "Force extended block device numbers and spread them"
1445 depends on DEBUG_KERNEL
1449 BIG FAT WARNING: ENABLING THIS OPTION MIGHT BREAK BOOTING ON
1450 SOME DISTRIBUTIONS. DO NOT ENABLE THIS UNLESS YOU KNOW WHAT
1451 YOU ARE DOING. Distros, please enable this and fix whatever
1454 Conventionally, block device numbers are allocated from
1455 predetermined contiguous area. However, extended block area
1456 may introduce non-contiguous block device numbers. This
1457 option forces most block device numbers to be allocated from
1458 the extended space and spreads them to discover kernel or
1459 userland code paths which assume predetermined contiguous
1460 device number allocation.
1462 Note that turning on this debug option shuffles all the
1463 device numbers for all IDE and SCSI devices including libata
1464 ones, so root partition specified using device number
1465 directly (via rdev or root=MAJ:MIN) won't work anymore.
1466 Textual device names (root=/dev/sdXn) will continue to work.
1468 Say N if you are unsure.
1470 config CPU_HOTPLUG_STATE_CONTROL
1471 bool "Enable CPU hotplug state control"
1472 depends on DEBUG_KERNEL
1473 depends on HOTPLUG_CPU
1476 Allows to write steps between "offline" and "online" to the CPUs
1477 sysfs target file so states can be stepped granular. This is a debug
1478 option for now as the hotplug machinery cannot be stopped and
1479 restarted at arbitrary points yet.
1481 Say N if your are unsure.
1483 config NOTIFIER_ERROR_INJECTION
1484 tristate "Notifier error injection"
1485 depends on DEBUG_KERNEL
1488 This option provides the ability to inject artificial errors to
1489 specified notifier chain callbacks. It is useful to test the error
1490 handling of notifier call chain failures.
1494 config PM_NOTIFIER_ERROR_INJECT
1495 tristate "PM notifier error injection module"
1496 depends on PM && NOTIFIER_ERROR_INJECTION
1497 default m if PM_DEBUG
1499 This option provides the ability to inject artificial errors to
1500 PM notifier chain callbacks. It is controlled through debugfs
1501 interface /sys/kernel/debug/notifier-error-inject/pm
1503 If the notifier call chain should be failed with some events
1504 notified, write the error code to "actions/<notifier event>/error".
1506 Example: Inject PM suspend error (-12 = -ENOMEM)
1508 # cd /sys/kernel/debug/notifier-error-inject/pm/
1509 # echo -12 > actions/PM_SUSPEND_PREPARE/error
1510 # echo mem > /sys/power/state
1511 bash: echo: write error: Cannot allocate memory
1513 To compile this code as a module, choose M here: the module will
1514 be called pm-notifier-error-inject.
1518 config OF_RECONFIG_NOTIFIER_ERROR_INJECT
1519 tristate "OF reconfig notifier error injection module"
1520 depends on OF_DYNAMIC && NOTIFIER_ERROR_INJECTION
1522 This option provides the ability to inject artificial errors to
1523 OF reconfig notifier chain callbacks. It is controlled
1524 through debugfs interface under
1525 /sys/kernel/debug/notifier-error-inject/OF-reconfig/
1527 If the notifier call chain should be failed with some events
1528 notified, write the error code to "actions/<notifier event>/error".
1530 To compile this code as a module, choose M here: the module will
1531 be called of-reconfig-notifier-error-inject.
1535 config NETDEV_NOTIFIER_ERROR_INJECT
1536 tristate "Netdev notifier error injection module"
1537 depends on NET && NOTIFIER_ERROR_INJECTION
1539 This option provides the ability to inject artificial errors to
1540 netdevice notifier chain callbacks. It is controlled through debugfs
1541 interface /sys/kernel/debug/notifier-error-inject/netdev
1543 If the notifier call chain should be failed with some events
1544 notified, write the error code to "actions/<notifier event>/error".
1546 Example: Inject netdevice mtu change error (-22 = -EINVAL)
1548 # cd /sys/kernel/debug/notifier-error-inject/netdev
1549 # echo -22 > actions/NETDEV_CHANGEMTU/error
1550 # ip link set eth0 mtu 1024
1551 RTNETLINK answers: Invalid argument
1553 To compile this code as a module, choose M here: the module will
1554 be called netdev-notifier-error-inject.
1558 config FUNCTION_ERROR_INJECTION
1560 depends on HAVE_FUNCTION_ERROR_INJECTION && KPROBES
1562 config FAULT_INJECTION
1563 bool "Fault-injection framework"
1564 depends on DEBUG_KERNEL
1566 Provide fault-injection framework.
1567 For more details, see Documentation/fault-injection/.
1570 bool "Fault-injection capability for kmalloc"
1571 depends on FAULT_INJECTION
1572 depends on SLAB || SLUB
1574 Provide fault-injection capability for kmalloc.
1576 config FAIL_PAGE_ALLOC
1577 bool "Fault-injection capabilitiy for alloc_pages()"
1578 depends on FAULT_INJECTION
1580 Provide fault-injection capability for alloc_pages().
1582 config FAIL_MAKE_REQUEST
1583 bool "Fault-injection capability for disk IO"
1584 depends on FAULT_INJECTION && BLOCK
1586 Provide fault-injection capability for disk IO.
1588 config FAIL_IO_TIMEOUT
1589 bool "Fault-injection capability for faking disk interrupts"
1590 depends on FAULT_INJECTION && BLOCK
1592 Provide fault-injection capability on end IO handling. This
1593 will make the block layer "forget" an interrupt as configured,
1594 thus exercising the error handling.
1596 Only works with drivers that use the generic timeout handling,
1597 for others it wont do anything.
1600 bool "Fault-injection capability for futexes"
1602 depends on FAULT_INJECTION && FUTEX
1604 Provide fault-injection capability for futexes.
1606 config FAULT_INJECTION_DEBUG_FS
1607 bool "Debugfs entries for fault-injection capabilities"
1608 depends on FAULT_INJECTION && SYSFS && DEBUG_FS
1610 Enable configuration of fault-injection capabilities via debugfs.
1612 config FAIL_FUNCTION
1613 bool "Fault-injection capability for functions"
1614 depends on FAULT_INJECTION_DEBUG_FS && FUNCTION_ERROR_INJECTION
1616 Provide function-based fault-injection capability.
1617 This will allow you to override a specific function with a return
1618 with given return value. As a result, function caller will see
1619 an error value and have to handle it. This is useful to test the
1620 error handling in various subsystems.
1622 config FAIL_MMC_REQUEST
1623 bool "Fault-injection capability for MMC IO"
1624 depends on FAULT_INJECTION_DEBUG_FS && MMC
1626 Provide fault-injection capability for MMC IO.
1627 This will make the mmc core return data errors. This is
1628 useful to test the error handling in the mmc block device
1629 and to test how the mmc host driver handles retries from
1632 config FAULT_INJECTION_STACKTRACE_FILTER
1633 bool "stacktrace filter for fault-injection capabilities"
1634 depends on FAULT_INJECTION_DEBUG_FS && STACKTRACE_SUPPORT
1637 select FRAME_POINTER if !MIPS && !PPC && !S390 && !MICROBLAZE && !ARM && !ARC && !X86
1639 Provide stacktrace filter for fault-injection capabilities
1642 bool "Latency measuring infrastructure"
1643 depends on DEBUG_KERNEL
1644 depends on STACKTRACE_SUPPORT
1646 select FRAME_POINTER if !MIPS && !PPC && !S390 && !MICROBLAZE && !ARM && !ARC && !X86
1653 Enable this option if you want to use the LatencyTOP tool
1654 to find out which userspace is blocking on what kernel operations.
1656 source "kernel/trace/Kconfig"
1658 config PROVIDE_OHCI1394_DMA_INIT
1659 bool "Remote debugging over FireWire early on boot"
1660 depends on PCI && X86
1662 If you want to debug problems which hang or crash the kernel early
1663 on boot and the crashing machine has a FireWire port, you can use
1664 this feature to remotely access the memory of the crashed machine
1665 over FireWire. This employs remote DMA as part of the OHCI1394
1666 specification which is now the standard for FireWire controllers.
1668 With remote DMA, you can monitor the printk buffer remotely using
1669 firescope and access all memory below 4GB using fireproxy from gdb.
1670 Even controlling a kernel debugger is possible using remote DMA.
1674 If ohci1394_dma=early is used as boot parameter, it will initialize
1675 all OHCI1394 controllers which are found in the PCI config space.
1677 As all changes to the FireWire bus such as enabling and disabling
1678 devices cause a bus reset and thereby disable remote DMA for all
1679 devices, be sure to have the cable plugged and FireWire enabled on
1680 the debugging host before booting the debug target for debugging.
1682 This code (~1k) is freed after boot. By then, the firewire stack
1683 in charge of the OHCI-1394 controllers should be used instead.
1685 See Documentation/debugging-via-ohci1394.txt for more information.
1687 menuconfig RUNTIME_TESTING_MENU
1688 bool "Runtime Testing"
1691 if RUNTIME_TESTING_MENU
1694 tristate "Linux Kernel Dump Test Tool Module"
1697 This module enables testing of the different dumping mechanisms by
1698 inducing system failures at predefined crash points.
1699 If you don't need it: say N
1700 Choose M here to compile this code as a module. The module will be
1703 Documentation on how to use the module can be found in
1704 Documentation/fault-injection/provoke-crashes.txt
1706 config TEST_LIST_SORT
1707 tristate "Linked list sorting test"
1708 depends on DEBUG_KERNEL || m
1710 Enable this to turn on 'list_sort()' function test. This test is
1711 executed only once during system boot (so affects only boot time),
1712 or at module load time.
1717 tristate "Array-based sort test"
1718 depends on DEBUG_KERNEL || m
1720 This option enables the self-test function of 'sort()' at boot,
1721 or at module load time.
1725 config KPROBES_SANITY_TEST
1726 bool "Kprobes sanity tests"
1727 depends on DEBUG_KERNEL
1730 This option provides for testing basic kprobes functionality on
1731 boot. Samples of kprobe and kretprobe are inserted and
1732 verified for functionality.
1734 Say N if you are unsure.
1736 config BACKTRACE_SELF_TEST
1737 tristate "Self test for the backtrace code"
1738 depends on DEBUG_KERNEL
1740 This option provides a kernel module that can be used to test
1741 the kernel stack backtrace code. This option is not useful
1742 for distributions or general kernels, but only for kernel
1743 developers working on architecture code.
1745 Note that if you want to also test saved backtraces, you will
1746 have to enable STACKTRACE as well.
1748 Say N if you are unsure.
1751 tristate "Red-Black tree test"
1752 depends on DEBUG_KERNEL
1754 A benchmark measuring the performance of the rbtree library.
1755 Also includes rbtree invariant checks.
1757 config INTERVAL_TREE_TEST
1758 tristate "Interval tree test"
1759 depends on DEBUG_KERNEL
1760 select INTERVAL_TREE
1762 A benchmark measuring the performance of the interval tree library
1765 tristate "Per cpu operations test"
1766 depends on m && DEBUG_KERNEL
1768 Enable this option to build test module which validates per-cpu
1773 config ATOMIC64_SELFTEST
1774 tristate "Perform an atomic64_t self-test"
1776 Enable this option to test the atomic64_t functions at boot or
1777 at module load time.
1781 config ASYNC_RAID6_TEST
1782 tristate "Self test for hardware accelerated raid6 recovery"
1783 depends on ASYNC_RAID6_RECOV
1786 This is a one-shot self test that permutes through the
1787 recovery of all the possible two disk failure scenarios for a
1788 N-disk array. Recovery is performed with the asynchronous
1789 raid6 recovery routines, and will optionally use an offload
1790 engine if one is available.
1795 tristate "Test functions located in the hexdump module at runtime"
1797 config TEST_STRING_HELPERS
1798 tristate "Test functions located in the string_helpers module at runtime"
1801 tristate "Test strscpy*() family of functions at runtime"
1804 tristate "Test kstrto*() family of functions at runtime"
1807 tristate "Test printf() family of functions at runtime"
1810 tristate "Test bitmap_*() family of functions at runtime"
1812 Enable this option to test the bitmap functions at boot.
1816 config TEST_BITFIELD
1817 tristate "Test bitfield functions at runtime"
1819 Enable this option to test the bitfield functions at boot.
1824 tristate "Test functions located in the uuid module at runtime"
1827 tristate "Test the XArray code at runtime"
1829 config TEST_OVERFLOW
1830 tristate "Test check_*_overflow() functions at runtime"
1832 config TEST_RHASHTABLE
1833 tristate "Perform selftest on resizable hash table"
1835 Enable this option to test the rhashtable functions at boot.
1840 tristate "Perform selftest on hash functions"
1842 Enable this option to test the kernel's integer (<linux/hash.h>),
1843 string (<linux/stringhash.h>), and siphash (<linux/siphash.h>)
1844 hash functions on boot (or module load).
1846 This is intended to help people writing architecture-specific
1847 optimized versions. If unsure, say N.
1850 tristate "Perform selftest on IDA functions"
1853 tristate "Perform selftest on priority array manager"
1856 Enable this option to test priority array manager on boot
1862 tristate "Test module loading with 'hello world' module"
1865 This builds the "test_module" module that emits "Hello, world"
1866 on printk when loaded. It is designed to be used for basic
1867 evaluation of the module loading subsystem (for example when
1868 validating module verification). It lacks any extra dependencies,
1869 and will not normally be loaded by the system unless explicitly
1875 tristate "Test module for stress/performance analysis of vmalloc allocator"
1880 This builds the "test_vmalloc" module that should be used for
1881 stress and performance analysis. So, any new change for vmalloc
1882 subsystem can be evaluated from performance and stability point
1887 config TEST_USER_COPY
1888 tristate "Test user/kernel boundary protections"
1891 This builds the "test_user_copy" module that runs sanity checks
1892 on the copy_to/from_user infrastructure, making sure basic
1893 user/kernel boundary testing is working. If it fails to load,
1894 a regression has been detected in the user/kernel memory boundary
1900 tristate "Test BPF filter functionality"
1903 This builds the "test_bpf" module that runs various test vectors
1904 against the BPF interpreter or BPF JIT compiler depending on the
1905 current setting. This is in particular useful for BPF JIT compiler
1906 development, but also to run regression tests against changes in
1907 the interpreter code. It also enables test stubs for eBPF maps and
1908 verifier used by user space verifier testsuite.
1912 config FIND_BIT_BENCHMARK
1913 tristate "Test find_bit functions"
1915 This builds the "test_find_bit" module that measure find_*_bit()
1916 functions performance.
1920 config TEST_FIRMWARE
1921 tristate "Test firmware loading via userspace interface"
1922 depends on FW_LOADER
1924 This builds the "test_firmware" module that creates a userspace
1925 interface for testing firmware loading. This can be used to
1926 control the triggering of firmware loading without needing an
1927 actual firmware-using device. The contents can be rechecked by
1933 tristate "sysctl test driver"
1934 depends on PROC_SYSCTL
1936 This builds the "test_sysctl" module. This driver enables to test the
1937 proc sysctl interfaces available to drivers safely without affecting
1938 production knobs which might alter system functionality.
1943 tristate "udelay test driver"
1945 This builds the "udelay_test" module that helps to make sure
1946 that udelay() is working properly.
1950 config TEST_STATIC_KEYS
1951 tristate "Test static keys"
1954 Test the static key interfaces.
1959 tristate "kmod stress tester"
1961 depends on NETDEVICES && NET_CORE && INET # for TUN
1968 Test the kernel's module loading mechanism: kmod. kmod implements
1969 support to load modules using the Linux kernel's usermode helper.
1970 This test provides a series of tests against kmod.
1972 Although technically you can either build test_kmod as a module or
1973 into the kernel we disallow building it into the kernel since
1974 it stress tests request_module() and this will very likely cause
1975 some issues by taking over precious threads available from other
1976 module load requests, ultimately this could be fatal.
1980 tools/testing/selftests/kmod/kmod.sh --help
1984 config TEST_DEBUG_VIRTUAL
1985 tristate "Test CONFIG_DEBUG_VIRTUAL feature"
1986 depends on DEBUG_VIRTUAL
1988 Test the kernel's ability to detect incorrect calls to
1989 virt_to_phys() done against the non-linear part of the
1990 kernel's virtual address map.
1994 config TEST_MEMCAT_P
1995 tristate "Test memcat_p() helper function"
1997 Test the memcat_p() helper for correctly merging two
1998 pointer arrays together.
2002 config TEST_LIVEPATCH
2003 tristate "Test livepatching"
2005 depends on DYNAMIC_DEBUG
2006 depends on LIVEPATCH
2009 Test kernel livepatching features for correctness. The tests will
2010 load test modules that will be livepatched in various scenarios.
2012 To run all the livepatching tests:
2014 make -C tools/testing/selftests TARGETS=livepatch run_tests
2016 Alternatively, individual tests may be invoked:
2018 tools/testing/selftests/livepatch/test-callbacks.sh
2019 tools/testing/selftests/livepatch/test-livepatch.sh
2020 tools/testing/selftests/livepatch/test-shadow-vars.sh
2025 tristate "Perform selftest on object aggreration manager"
2029 Enable this option to test object aggregation manager on boot
2033 config TEST_STACKINIT
2034 tristate "Test level of stack variable initialization"
2036 Test if the kernel is zero-initializing stack variables and
2037 padding. Coverage is controlled by compiler flags,
2038 CONFIG_GCC_PLUGIN_STRUCTLEAK, CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF,
2039 or CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL.
2043 endif # RUNTIME_TESTING_MENU
2048 This option adds a kernel parameter 'memtest', which allows memtest
2050 memtest=0, mean disabled; -- default
2051 memtest=1, mean do 1 test pattern;
2053 memtest=17, mean do 17 test patterns.
2054 If you are unsure how to answer this question, answer N.
2056 config BUG_ON_DATA_CORRUPTION
2057 bool "Trigger a BUG when data corruption is detected"
2060 Select this option if the kernel should BUG when it encounters
2061 data corruption in kernel memory structures when they get checked
2066 source "samples/Kconfig"
2068 source "lib/Kconfig.kgdb"
2070 source "lib/Kconfig.ubsan"
2072 config ARCH_HAS_DEVMEM_IS_ALLOWED
2075 config STRICT_DEVMEM
2076 bool "Filter access to /dev/mem"
2077 depends on MMU && DEVMEM
2078 depends on ARCH_HAS_DEVMEM_IS_ALLOWED
2079 default y if PPC || X86 || ARM64
2081 If this option is disabled, you allow userspace (root) access to all
2082 of memory, including kernel and userspace memory. Accidental
2083 access to this is obviously disastrous, but specific access can
2084 be used by people debugging the kernel. Note that with PAT support
2085 enabled, even in this case there are restrictions on /dev/mem
2086 use due to the cache aliasing requirements.
2088 If this option is switched on, and IO_STRICT_DEVMEM=n, the /dev/mem
2089 file only allows userspace access to PCI space and the BIOS code and
2090 data regions. This is sufficient for dosemu and X and all common
2095 config IO_STRICT_DEVMEM
2096 bool "Filter I/O access to /dev/mem"
2097 depends on STRICT_DEVMEM
2099 If this option is disabled, you allow userspace (root) access to all
2100 io-memory regardless of whether a driver is actively using that
2101 range. Accidental access to this is obviously disastrous, but
2102 specific access can be used by people debugging kernel drivers.
2104 If this option is switched on, the /dev/mem file only allows
2105 userspace access to *idle* io-memory ranges (see /proc/iomem) This
2106 may break traditional users of /dev/mem (dosemu, legacy X, etc...)
2107 if the driver using a given range cannot be disabled.
2111 source "arch/$(SRCARCH)/Kconfig.debug"
2113 endmenu # Kernel hacking