3 bool "Show timing information on printks"
6 Selecting this option causes timing information to be
7 included in printk output. This allows you to measure
8 the interval between kernel operations, including bootup
9 operations. This is useful for identifying long delays
13 bool "Continue building despite compiler warnings"
16 By disabling this option you will enable -Werror on building C
17 files. This causes all warnings to abort the compilation, just as
18 errors do. (It is generally not recommended to disable this option as
19 the overwhelming majority of warnings is harmless and also gcc puts
20 out false-positive warnings. It is useful for automated testing
23 config ENABLE_WARN_DEPRECATED
24 bool "Enable __deprecated logic"
25 depends on ALLOW_WARNINGS
28 Enable the __deprecated logic in the kernel build.
29 Disable this to suppress the "warning: 'foo' is deprecated
30 (declared at kernel/power/somefile.c:1234)" messages.
32 config ENABLE_MUST_CHECK
33 bool "Enable __must_check logic"
34 depends on ALLOW_WARNINGS
36 Enable the __must_check logic in the kernel build. Disable this to
37 suppress the "warning: ignoring return value of 'foo', declared with
38 attribute warn_unused_result" messages.
42 int "Warn for stack frames larger than (needs gcc 4.4)"
44 default 1024 if !64BIT
47 Tell gcc to warn at build time for stack frames larger than this.
48 Setting this too low will cause a lot of warnings.
49 Setting it to 0 disables the warning.
53 bool "Magic SysRq key"
56 If you say Y here, you will have some control over the system even
57 if the system crashes for example during kernel debugging (e.g., you
58 will be able to flush the buffer cache to disk, reboot the system
59 immediately or dump some status information). This is accomplished
60 by pressing various keys while holding SysRq (Alt+PrintScreen). It
61 also works on a serial console (on PC hardware at least), if you
62 send a BREAK and then within 5 seconds a command keypress. The
63 keys are documented in <file:Documentation/sysrq.txt>. Don't say Y
64 unless you really know what this hack does.
67 bool "Enable unused/obsolete exported symbols"
70 Unused but exported symbols make the kernel needlessly bigger. For
71 that reason most of these unused exports will soon be removed. This
72 option is provided temporarily to provide a transition period in case
73 some external kernel module needs one of these symbols anyway. If you
74 encounter such a case in your module, consider if you are actually
75 using the right API. (rationale: since nobody in the kernel is using
76 this in a module, there is a pretty good chance it's actually the
77 wrong interface to use). If you really need the symbol, please send a
78 mail to the linux kernel mailing list mentioning the symbol and why
79 you really need it, and what the merge plan to the mainline kernel for
83 bool "Debug Filesystem"
86 debugfs is a virtual file system that kernel developers use to put
87 debugging files into. Enable this option to be able to read and
90 For detailed documentation on the debugfs API, see
91 Documentation/DocBook/filesystems.
96 bool "Run 'make headers_check' when building vmlinux"
99 This option will extract the user-visible kernel headers whenever
100 building the kernel, and will run basic sanity checks on them to
101 ensure that exported files do not attempt to include files which
102 were not exported, etc.
104 If you're making modifications to header files which are
105 relevant for userspace, say 'Y', and check the headers
106 exported to $(INSTALL_HDR_PATH) (usually 'usr/include' in
107 your build tree), to make sure they're suitable.
109 config DEBUG_SECTION_MISMATCH
110 bool "Enable full Section mismatch analysis"
111 # This option is on purpose disabled for now.
112 # It will be enabled when we are down to a resonable number
113 # of section mismatch warnings (< 10 for an allyesconfig build)
115 The section mismatch analysis checks if there are illegal
116 references from one section to another section.
117 Linux will during link or during runtime drop some sections
118 and any use of code/data previously in these sections will
119 most likely result in an oops.
120 In the code functions and variables are annotated with
121 __init, __devinit etc. (see full list in include/linux/init.h)
122 which results in the code/data being placed in specific sections.
123 The section mismatch analysis is always done after a full
124 kernel build but enabling this option will in addition
126 - Add the option -fno-inline-functions-called-once to gcc
127 When inlining a function annotated __init in a non-init
128 function we would lose the section information and thus
129 the analysis would not catch the illegal reference.
130 This option tells gcc to inline less but will also
131 result in a larger kernel.
132 - Run the section mismatch analysis for each module/built-in.o
133 When we run the section mismatch analysis on vmlinux.o we
134 lose valueble information about where the mismatch was
136 Running the analysis for each module/built-in.o file
137 will tell where the mismatch happens much closer to the
138 source. The drawback is that we will report the same
139 mismatch at least twice.
140 - Enable verbose reporting from modpost to help solving
141 the section mismatches reported.
144 bool "Kernel debugging"
146 Say Y here if you are developing drivers or trying to debug and
147 identify kernel problems.
150 bool "Debug shared IRQ handlers"
151 depends on DEBUG_KERNEL && GENERIC_HARDIRQS
153 Enable this to generate a spurious interrupt as soon as a shared
154 interrupt handler is registered, and just before one is deregistered.
155 Drivers ought to be able to handle interrupts coming in at those
156 points; some don't and need to be caught.
158 config DETECT_SOFTLOCKUP
159 bool "Detect Soft Lockups"
160 depends on DEBUG_KERNEL && !S390
163 Say Y here to enable the kernel to detect "soft lockups",
164 which are bugs that cause the kernel to loop in kernel
165 mode for more than 60 seconds, without giving other tasks a
168 When a soft-lockup is detected, the kernel will print the
169 current stack trace (which you should report), but the
170 system will stay locked up. This feature has negligible
173 (Note that "hard lockups" are separate type of bugs that
174 can be detected via the NMI-watchdog, on platforms that
177 config BOOTPARAM_SOFTLOCKUP_PANIC
178 bool "Panic (Reboot) On Soft Lockups"
179 depends on DETECT_SOFTLOCKUP
181 Say Y here to enable the kernel to panic on "soft lockups",
182 which are bugs that cause the kernel to loop in kernel
183 mode for more than 60 seconds, without giving other tasks a
186 The panic can be used in combination with panic_timeout,
187 to cause the system to reboot automatically after a
188 lockup has been detected. This feature is useful for
189 high-availability systems that have uptime guarantees and
190 where a lockup must be resolved ASAP.
194 config BOOTPARAM_SOFTLOCKUP_PANIC_VALUE
196 depends on DETECT_SOFTLOCKUP
198 default 0 if !BOOTPARAM_SOFTLOCKUP_PANIC
199 default 1 if BOOTPARAM_SOFTLOCKUP_PANIC
201 config DETECT_HUNG_TASK
202 bool "Detect Hung Tasks"
203 depends on DEBUG_KERNEL
206 Say Y here to enable the kernel to detect "hung tasks",
207 which are bugs that cause the task to be stuck in
208 uninterruptible "D" state indefinitiley.
210 When a hung task is detected, the kernel will print the
211 current stack trace (which you should report), but the
212 task will stay in uninterruptible state. If lockdep is
213 enabled then all held locks will also be reported. This
214 feature has negligible overhead.
216 config BOOTPARAM_HUNG_TASK_PANIC
217 bool "Panic (Reboot) On Hung Tasks"
218 depends on DETECT_HUNG_TASK
220 Say Y here to enable the kernel to panic on "hung tasks",
221 which are bugs that cause the kernel to leave a task stuck
222 in uninterruptible "D" state.
224 The panic can be used in combination with panic_timeout,
225 to cause the system to reboot automatically after a
226 hung task has been detected. This feature is useful for
227 high-availability systems that have uptime guarantees and
228 where a hung tasks must be resolved ASAP.
232 config BOOTPARAM_HUNG_TASK_PANIC_VALUE
234 depends on DETECT_HUNG_TASK
236 default 0 if !BOOTPARAM_HUNG_TASK_PANIC
237 default 1 if BOOTPARAM_HUNG_TASK_PANIC
240 bool "Collect scheduler debugging info"
241 depends on DEBUG_KERNEL && PROC_FS
244 If you say Y here, the /proc/sched_debug file will be provided
245 that can help debug the scheduler. The runtime overhead of this
249 bool "Collect scheduler statistics"
250 depends on DEBUG_KERNEL && PROC_FS
252 If you say Y here, additional code will be inserted into the
253 scheduler and related routines to collect statistics about
254 scheduler behavior and provide them in /proc/schedstat. These
255 stats may be useful for both tuning and debugging the scheduler
256 If you aren't debugging the scheduler or trying to tune a specific
257 application, you can say N to avoid the very slight overhead
261 bool "Collect kernel timers statistics"
262 depends on DEBUG_KERNEL && PROC_FS
264 If you say Y here, additional code will be inserted into the
265 timer routines to collect statistics about kernel timers being
266 reprogrammed. The statistics can be read from /proc/timer_stats.
267 The statistics collection is started by writing 1 to /proc/timer_stats,
268 writing 0 stops it. This feature is useful to collect information
269 about timer usage patterns in kernel and userspace. This feature
270 is lightweight if enabled in the kernel config but not activated
271 (it defaults to deactivated on bootup and will only be activated
272 if some application like powertop activates it explicitly).
275 bool "Debug object operations"
276 depends on DEBUG_KERNEL
278 If you say Y here, additional code will be inserted into the
279 kernel to track the life time of various objects and validate
280 the operations on those objects.
282 config DEBUG_OBJECTS_SELFTEST
283 bool "Debug objects selftest"
284 depends on DEBUG_OBJECTS
286 This enables the selftest of the object debug code.
288 config DEBUG_OBJECTS_FREE
289 bool "Debug objects in freed memory"
290 depends on DEBUG_OBJECTS
292 This enables checks whether a k/v free operation frees an area
293 which contains an object which has not been deactivated
294 properly. This can make kmalloc/kfree-intensive workloads
297 config DEBUG_OBJECTS_TIMERS
298 bool "Debug timer objects"
299 depends on DEBUG_OBJECTS
301 If you say Y here, additional code will be inserted into the
302 timer routines to track the life time of timer objects and
303 validate the timer operations.
305 config DEBUG_OBJECTS_ENABLE_DEFAULT
306 int "debug_objects bootup default value (0-1)"
309 depends on DEBUG_OBJECTS
311 Debug objects boot parameter default value
314 bool "Debug slab memory allocations"
315 depends on DEBUG_KERNEL && SLAB && !KMEMCHECK
317 Say Y here to have the kernel do limited verification on memory
318 allocation as well as poisoning memory on free to catch use of freed
319 memory. This can make kmalloc/kfree-intensive workloads much slower.
321 config DEBUG_SLAB_LEAK
322 bool "Memory leak debugging"
323 depends on DEBUG_SLAB
326 bool "SLUB debugging on by default"
327 depends on SLUB && SLUB_DEBUG && !KMEMCHECK
330 Boot with debugging on by default. SLUB boots by default with
331 the runtime debug capabilities switched off. Enabling this is
332 equivalent to specifying the "slub_debug" parameter on boot.
333 There is no support for more fine grained debug control like
334 possible with slub_debug=xxx. SLUB debugging may be switched
335 off in a kernel built with CONFIG_SLUB_DEBUG_ON by specifying
340 bool "Enable SLUB performance statistics"
341 depends on SLUB && SLUB_DEBUG && SYSFS
343 SLUB statistics are useful to debug SLUBs allocation behavior in
344 order find ways to optimize the allocator. This should never be
345 enabled for production use since keeping statistics slows down
346 the allocator by a few percentage points. The slabinfo command
347 supports the determination of the most active slabs to figure
348 out which slabs are relevant to a particular load.
349 Try running: slabinfo -DA
352 bool "Debug preemptible kernel"
353 depends on DEBUG_KERNEL && PREEMPT && (TRACE_IRQFLAGS_SUPPORT || PPC64)
356 If you say Y here then the kernel will use a debug variant of the
357 commonly used smp_processor_id() function and will print warnings
358 if kernel code uses it in a preemption-unsafe way. Also, the kernel
359 will detect preemption count underflows.
361 config DEBUG_RT_MUTEXES
362 bool "RT Mutex debugging, deadlock detection"
363 depends on DEBUG_KERNEL && RT_MUTEXES
365 This allows rt mutex semantics violations and rt mutex related
366 deadlocks (lockups) to be detected and reported automatically.
367 When realtime preemption is enabled this includes spinlocks,
368 rwlocks, mutexes and (rw)semaphores
373 depends on DEBUG_RT_MUTEXES
375 config RT_MUTEX_TESTER
376 bool "Built-in scriptable tester for rt-mutexes"
377 depends on DEBUG_KERNEL && RT_MUTEXES
379 This option enables a rt-mutex tester.
381 config DEBUG_SPINLOCK
382 bool "Spinlock and rw-lock debugging: basic checks"
383 depends on DEBUG_KERNEL
385 Say Y here and build SMP to catch missing spinlock initialization
386 and certain other kinds of spinlock errors commonly made. This is
387 best used in conjunction with the NMI watchdog so that spinlock
388 deadlocks are also debuggable.
391 bool "Mutex debugging: basic checks"
392 depends on DEBUG_KERNEL && !PREEMPT_RT
394 This feature allows mutex semantics violations to be detected and
397 config DEBUG_LOCK_ALLOC
398 bool "Lock debugging: detect incorrect freeing of live locks"
399 depends on DEBUG_KERNEL && TRACE_IRQFLAGS_SUPPORT && STACKTRACE_SUPPORT && LOCKDEP_SUPPORT
400 select DEBUG_SPINLOCK
404 This feature will check whether any held lock (spinlock, rwlock,
405 mutex or rwsem) is incorrectly freed by the kernel, via any of the
406 memory-freeing routines (kfree(), kmem_cache_free(), free_pages(),
407 vfree(), etc.), whether a live lock is incorrectly reinitialized via
408 spin_lock_init()/mutex_init()/etc., or whether there is any lock
409 held during task exit.
412 bool "Lock debugging: prove locking correctness"
413 depends on DEBUG_KERNEL && TRACE_IRQFLAGS_SUPPORT && STACKTRACE_SUPPORT && LOCKDEP_SUPPORT
415 select DEBUG_SPINLOCK
417 select DEBUG_LOCK_ALLOC
420 This feature enables the kernel to prove that all locking
421 that occurs in the kernel runtime is mathematically
422 correct: that under no circumstance could an arbitrary (and
423 not yet triggered) combination of observed locking
424 sequences (on an arbitrary number of CPUs, running an
425 arbitrary number of tasks and interrupt contexts) cause a
428 In short, this feature enables the kernel to report locking
429 related deadlocks before they actually occur.
431 The proof does not depend on how hard and complex a
432 deadlock scenario would be to trigger: how many
433 participant CPUs, tasks and irq-contexts would be needed
434 for it to trigger. The proof also does not depend on
435 timing: if a race and a resulting deadlock is possible
436 theoretically (no matter how unlikely the race scenario
437 is), it will be proven so and will immediately be
438 reported by the kernel (once the event is observed that
439 makes the deadlock theoretically possible).
441 If a deadlock is impossible (i.e. the locking rules, as
442 observed by the kernel, are mathematically correct), the
443 kernel reports nothing.
445 NOTE: this feature can also be enabled for rwlocks, mutexes
446 and rwsems - in which case all dependencies between these
447 different locking variants are observed and mapped too, and
448 the proof of observed correctness is also maintained for an
449 arbitrary combination of these separate locking variants.
451 For more details, see Documentation/lockdep-design.txt.
455 depends on DEBUG_KERNEL && TRACE_IRQFLAGS_SUPPORT && STACKTRACE_SUPPORT && LOCKDEP_SUPPORT
457 select FRAME_POINTER if !MIPS && !PPC
462 bool "Lock usage statistics"
463 depends on DEBUG_KERNEL && TRACE_IRQFLAGS_SUPPORT && STACKTRACE_SUPPORT && LOCKDEP_SUPPORT
465 select DEBUG_SPINLOCK
467 select DEBUG_LOCK_ALLOC
470 This feature enables tracking lock contention points
472 For more details, see Documentation/lockstat.txt
475 bool "Lock dependency engine debugging"
476 depends on DEBUG_KERNEL && LOCKDEP
478 If you say Y here, the lock dependency engine will do
479 additional runtime checks to debug itself, at the price
480 of more runtime overhead.
482 config TRACE_IRQFLAGS
483 depends on DEBUG_KERNEL
486 depends on TRACE_IRQFLAGS_SUPPORT
487 depends on PROVE_LOCKING
489 config DEBUG_SPINLOCK_SLEEP
490 bool "Spinlock debugging: sleep-inside-spinlock checking"
491 depends on DEBUG_KERNEL
493 If you say Y here, various routines which may sleep will become very
494 noisy if they are called with a spinlock held.
496 config DEBUG_LOCKING_API_SELFTESTS
497 bool "Locking API boot-time self-tests"
498 depends on DEBUG_KERNEL
500 Say Y here if you want the kernel to run a short self-test during
501 bootup. The self-test checks whether common types of locking bugs
502 are detected by debugging mechanisms or not. (if you disable
503 lock debugging then those bugs wont be detected of course.)
504 The following locking APIs are covered: spinlocks, rwlocks,
509 depends on STACKTRACE_SUPPORT
512 bool "kobject debugging"
513 depends on DEBUG_KERNEL
515 If you say Y here, some extra kobject debugging messages will be sent
519 bool "Highmem debugging"
520 depends on DEBUG_KERNEL && HIGHMEM
522 This options enables addition error checking for high memory systems.
523 Disable for production systems.
525 config DEBUG_BUGVERBOSE
526 bool "Verbose BUG() reporting (adds 70K)" if DEBUG_KERNEL && EMBEDDED
528 depends on ARM || AVR32 || M32R || M68K || SPARC32 || SPARC64 || \
529 FRV || SUPERH || GENERIC_BUG || BLACKFIN || MN10300
532 Say Y here to make BUG() panics output the file name and line number
533 of the BUG call as well as the EIP and oops trace. This aids
534 debugging but costs about 70-100K of memory.
537 bool "Compile the kernel with debug info"
538 depends on DEBUG_KERNEL
540 If you say Y here the resulting kernel image will include
541 debugging info resulting in a larger kernel image.
542 This adds debug symbols to the kernel and modules (gcc -g), and
543 is needed if you intend to use kernel crashdump or binary object
544 tools like crash, kgdb, LKCD, gdb, etc on the kernel.
545 Say Y here only if you plan to debug the kernel.
551 depends on DEBUG_KERNEL
553 Enable this to turn on extended checks in the virtual-memory system
554 that may impact performance.
559 bool "Debug VM translations"
560 depends on DEBUG_KERNEL && X86
562 Enable some costly sanity checks in virtual to page code. This can
563 catch mistakes with virt_to_page() and friends.
567 config DEBUG_NOMMU_REGIONS
568 bool "Debug the global anon/private NOMMU mapping region tree"
569 depends on DEBUG_KERNEL && !MMU
571 This option causes the global tree of anonymous and private mapping
572 regions to be regularly checked for invalid topology.
574 config DEBUG_WRITECOUNT
575 bool "Debug filesystem writers count"
576 depends on DEBUG_KERNEL
578 Enable this to catch wrong use of the writers count in struct
579 vfsmount. This will increase the size of each file struct by
584 config DEBUG_MEMORY_INIT
585 bool "Debug memory initialisation" if EMBEDDED
588 Enable this for additional checks during memory initialisation.
589 The sanity checks verify aspects of the VM such as the memory model
590 and other information provided by the architecture. Verbose
591 information will be printed at KERN_DEBUG loglevel depending
592 on the mminit_loglevel= command-line option.
597 bool "Debug linked list manipulation"
598 depends on DEBUG_KERNEL
600 Enable this to turn on extended checks in the linked-list
606 bool "Debug SG table operations"
607 depends on DEBUG_KERNEL
609 Enable this to turn on checks on scatter-gather tables. This can
610 help find problems with drivers that do not properly initialize
615 config DEBUG_NOTIFIERS
616 bool "Debug notifier call chains"
617 depends on DEBUG_KERNEL
619 Enable this to turn on sanity checking for notifier call chains.
620 This is most useful for kernel developers to make sure that
621 modules properly unregister themselves from notifier chains.
622 This is a relatively cheap check but if you care about maximum
626 # Select this config option from the architecture Kconfig, if it
627 # it is preferred to always offer frame pointers as a config
628 # option on the architecture (regardless of KERNEL_DEBUG):
630 config ARCH_WANT_FRAME_POINTERS
635 bool "Compile the kernel with frame pointers"
636 depends on DEBUG_KERNEL && \
637 (CRIS || M68K || M68KNOMMU || FRV || UML || S390 || \
638 AVR32 || SUPERH || BLACKFIN || MN10300) || \
639 ARCH_WANT_FRAME_POINTERS
640 default y if (DEBUG_INFO && UML) || ARCH_WANT_FRAME_POINTERS
642 If you say Y here the resulting kernel image will be slightly
643 larger and slower, but it gives very useful debugging information
644 in case of kernel bugs. (precise oopses/stacktraces/warnings)
646 config BOOT_PRINTK_DELAY
647 bool "Delay each boot printk message by N milliseconds"
648 depends on DEBUG_KERNEL && PRINTK && GENERIC_CALIBRATE_DELAY
650 This build option allows you to read kernel boot messages
651 by inserting a short delay after each one. The delay is
652 specified in milliseconds on the kernel command line,
653 using "boot_delay=N".
655 It is likely that you would also need to use "lpj=M" to preset
656 the "loops per jiffie" value.
657 See a previous boot log for the "lpj" value to use for your
658 system, and then set "lpj=M" before setting "boot_delay=N".
659 NOTE: Using this option may adversely affect SMP systems.
660 I.e., processors other than the first one may not boot up.
661 BOOT_PRINTK_DELAY also may cause DETECT_SOFTLOCKUP to detect
662 what it believes to be lockup conditions.
664 config RCU_TORTURE_TEST
665 tristate "torture tests for RCU"
666 depends on DEBUG_KERNEL
669 This option provides a kernel module that runs torture tests
670 on the RCU infrastructure. The kernel module may be built
671 after the fact on the running kernel to be tested, if desired.
673 Say Y here if you want RCU torture tests to be built into
675 Say M if you want the RCU torture tests to build as a module.
676 Say N if you are unsure.
678 config RCU_TORTURE_TEST_RUNNABLE
679 bool "torture tests for RCU runnable by default"
680 depends on RCU_TORTURE_TEST = y
683 This option provides a way to build the RCU torture tests
684 directly into the kernel without them starting up at boot
685 time. You can use /proc/sys/kernel/rcutorture_runnable
686 to manually override this setting. This /proc file is
687 available only when the RCU torture tests have been built
690 Say Y here if you want the RCU torture tests to start during
691 boot (you probably don't).
692 Say N here if you want the RCU torture tests to start only
693 after being manually enabled via /proc.
695 config RCU_CPU_STALL_DETECTOR
696 bool "Check for stalled CPUs delaying RCU grace periods"
697 depends on CLASSIC_RCU || TREE_RCU
700 This option causes RCU to printk information on which
701 CPUs are delaying the current grace period, but only when
702 the grace period extends for excessive time periods.
704 Say Y if you want RCU to perform such checks.
706 Say N if you are unsure.
708 config KPROBES_SANITY_TEST
709 bool "Kprobes sanity tests"
710 depends on DEBUG_KERNEL
714 This option provides for testing basic kprobes functionality on
715 boot. A sample kprobe, jprobe and kretprobe are inserted and
716 verified for functionality.
718 Say N if you are unsure.
720 config BACKTRACE_SELF_TEST
721 tristate "Self test for the backtrace code"
722 depends on DEBUG_KERNEL
725 This option provides a kernel module that can be used to test
726 the kernel stack backtrace code. This option is not useful
727 for distributions or general kernels, but only for kernel
728 developers working on architecture code.
730 Note that if you want to also test saved backtraces, you will
731 have to enable STACKTRACE as well.
733 Say N if you are unsure.
735 config DEBUG_BLOCK_EXT_DEVT
736 bool "Force extended block device numbers and spread them"
737 depends on DEBUG_KERNEL
741 BIG FAT WARNING: ENABLING THIS OPTION MIGHT BREAK BOOTING ON
742 SOME DISTRIBUTIONS. DO NOT ENABLE THIS UNLESS YOU KNOW WHAT
743 YOU ARE DOING. Distros, please enable this and fix whatever
746 Conventionally, block device numbers are allocated from
747 predetermined contiguous area. However, extended block area
748 may introduce non-contiguous block device numbers. This
749 option forces most block device numbers to be allocated from
750 the extended space and spreads them to discover kernel or
751 userland code paths which assume predetermined contiguous
752 device number allocation.
754 Note that turning on this debug option shuffles all the
755 device numbers for all IDE and SCSI devices including libata
756 ones, so root partition specified using device number
757 directly (via rdev or root=MAJ:MIN) won't work anymore.
758 Textual device names (root=/dev/sdXn) will continue to work.
760 Say N if you are unsure.
763 tristate "Linux Kernel Dump Test Tool Module"
764 depends on DEBUG_KERNEL
769 This module enables testing of the different dumping mechanisms by
770 inducing system failures at predefined crash points.
771 If you don't need it: say N
772 Choose M here to compile this code as a module. The module will be
775 Documentation on how to use the module can be found in
778 config FAULT_INJECTION
779 bool "Fault-injection framework"
780 depends on DEBUG_KERNEL
782 Provide fault-injection framework.
783 For more details, see Documentation/fault-injection/.
786 bool "Fault-injection capability for kmalloc"
787 depends on FAULT_INJECTION
788 depends on SLAB || SLUB
790 Provide fault-injection capability for kmalloc.
792 config FAIL_PAGE_ALLOC
793 bool "Fault-injection capabilitiy for alloc_pages()"
794 depends on FAULT_INJECTION
796 Provide fault-injection capability for alloc_pages().
798 config FAIL_MAKE_REQUEST
799 bool "Fault-injection capability for disk IO"
800 depends on FAULT_INJECTION && BLOCK
802 Provide fault-injection capability for disk IO.
804 config FAIL_IO_TIMEOUT
805 bool "Faul-injection capability for faking disk interrupts"
806 depends on FAULT_INJECTION && BLOCK
808 Provide fault-injection capability on end IO handling. This
809 will make the block layer "forget" an interrupt as configured,
810 thus exercising the error handling.
812 Only works with drivers that use the generic timeout handling,
813 for others it wont do anything.
815 config FAULT_INJECTION_DEBUG_FS
816 bool "Debugfs entries for fault-injection capabilities"
817 depends on FAULT_INJECTION && SYSFS && DEBUG_FS
819 Enable configuration of fault-injection capabilities via debugfs.
821 config FAULT_INJECTION_STACKTRACE_FILTER
822 bool "stacktrace filter for fault-injection capabilities"
823 depends on FAULT_INJECTION_DEBUG_FS && STACKTRACE_SUPPORT
826 select FRAME_POINTER if !PPC
828 Provide stacktrace filter for fault-injection capabilities
831 bool "Latency measuring infrastructure"
832 select FRAME_POINTER if !MIPS && !PPC
838 depends on HAVE_LATENCYTOP_SUPPORT
840 Enable this option if you want to use the LatencyTOP tool
841 to find out which userspace is blocking on what kernel operations.
843 config SYSCTL_SYSCALL_CHECK
845 depends on SYSCTL_SYSCALL
847 sys_sysctl uses binary paths that have been found challenging
848 to properly maintain and use. This enables checks that help
849 you to keep things correct.
851 source kernel/trace/Kconfig
853 config PROVIDE_OHCI1394_DMA_INIT
854 bool "Remote debugging over FireWire early on boot"
855 depends on PCI && X86
857 If you want to debug problems which hang or crash the kernel early
858 on boot and the crashing machine has a FireWire port, you can use
859 this feature to remotely access the memory of the crashed machine
860 over FireWire. This employs remote DMA as part of the OHCI1394
861 specification which is now the standard for FireWire controllers.
863 With remote DMA, you can monitor the printk buffer remotely using
864 firescope and access all memory below 4GB using fireproxy from gdb.
865 Even controlling a kernel debugger is possible using remote DMA.
869 If ohci1394_dma=early is used as boot parameter, it will initialize
870 all OHCI1394 controllers which are found in the PCI config space.
872 As all changes to the FireWire bus such as enabling and disabling
873 devices cause a bus reset and thereby disable remote DMA for all
874 devices, be sure to have the cable plugged and FireWire enabled on
875 the debugging host before booting the debug target for debugging.
877 This code (~1k) is freed after boot. By then, the firewire stack
878 in charge of the OHCI-1394 controllers should be used instead.
880 See Documentation/debugging-via-ohci1394.txt for more information.
882 config FIREWIRE_OHCI_REMOTE_DMA
883 bool "Remote debugging over FireWire with firewire-ohci"
884 depends on FIREWIRE_OHCI
886 This option lets you use the FireWire bus for remote debugging
887 with help of the firewire-ohci driver. It enables unfiltered
888 remote DMA in firewire-ohci.
889 See Documentation/debugging-via-ohci1394.txt for more information.
894 bool "Build targets in Documentation/ tree"
895 depends on HEADERS_CHECK
897 This option attempts to build objects from the source files in the
898 kernel Documentation/ tree.
900 Say N if you are unsure.
902 config DYNAMIC_PRINTK_DEBUG
903 bool "Enable dynamic printk() call support"
909 Compiles debug level messages into the kernel, which would not
910 otherwise be available at runtime. These messages can then be
911 enabled/disabled on a per module basis. This mechanism implicitly
912 enables all pr_debug() and dev_dbg() calls. The impact of this
913 compile option is a larger kernel text size of about 2%.
917 Dynamic debugging is controlled by the debugfs file,
918 dynamic_printk/modules. This file contains a list of the modules that
919 can be enabled. The format of the file is the module name, followed
920 by a set of flags that can be enabled. The first flag is always the
921 'enabled' flag. For example:
923 <module_name> <enabled=0/1>
928 <module_name> : Name of the module in which the debug call resides
929 <enabled=0/1> : whether the messages are enabled or not
933 snd_hda_intel enabled=0
939 $echo "set enabled=1 <module_name>" > dynamic_printk/modules
943 $echo "set enabled=0 <module_name>" > dynamic_printk/modules
947 $echo "set enabled=1 all" > dynamic_printk/modules
951 $echo "set enabled=0 all" > dynamic_printk/modules
953 Finally, passing "dynamic_printk" at the command line enables
954 debugging for all modules. This mode can be turned off via the above
958 bool "Enable debugging of DMA-API usage"
959 depends on HAVE_DMA_API_DEBUG
961 Enable this option to debug the use of the DMA API by device drivers.
962 With this option you will be able to detect common bugs in device
963 drivers like double-freeing of DMA mappings or freeing mappings that
964 were never allocated.
965 This option causes a performance degredation. Use only if you want
966 to debug device drivers. If unsure, say N.
968 source "samples/Kconfig"
970 source "lib/Kconfig.kgdb"
972 source "lib/Kconfig.kmemcheck"