1# SPDX-License-Identifier: GPL-2.0-only 2menu "Kernel hacking" 3 4menu "printk and dmesg options" 5 6config PRINTK_TIME 7 bool "Show timing information on printks" 8 depends on PRINTK 9 help 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. 13 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. 17 18 The behavior is also controlled by the kernel command line 19 parameter printk.time=1. See Documentation/admin-guide/kernel-parameters.rst 20 21config PRINTK_CALLER 22 bool "Show caller information on printks" 23 depends on PRINTK 24 help 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) 27 to every message. 28 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. 33 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 36 sysfs interface. 37 38config CONSOLE_LOGLEVEL_DEFAULT 39 int "Default console loglevel (1-15)" 40 range 1 15 41 default "7" 42 help 43 Default loglevel to determine what will be printed on the console. 44 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. 48 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 51 option. 52 53config CONSOLE_LOGLEVEL_QUIET 54 int "quiet console loglevel (1-15)" 55 range 1 15 56 default "4" 57 help 58 loglevel to use when "quiet" is passed on the kernel commandline. 59 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>" 63 64config MESSAGE_LOGLEVEL_DEFAULT 65 int "Default message log level (1-7)" 66 range 1 7 67 default "4" 68 help 69 Default log level for printk statements with no specified priority. 70 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 73 priority. 74 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. 78 79config BOOT_PRINTK_DELAY 80 bool "Delay each boot printk message by N milliseconds" 81 depends on DEBUG_KERNEL && PRINTK && GENERIC_CALIBRATE_DELAY 82 help 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, 86 using "boot_delay=N". 87 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. 96 97config DYNAMIC_DEBUG 98 bool "Enable dynamic printk() support" 99 default n 100 depends on PRINTK 101 depends on (DEBUG_FS || PROC_FS) 102 select DYNAMIC_DEBUG_CORE 103 help 104 105 Compiles debug level messages into the kernel, which would not 106 otherwise be available at runtime. These messages can then be 107 enabled/disabled based on various levels of scope - per source file, 108 function, module, format string, and line number. This mechanism 109 implicitly compiles in all pr_debug() and dev_dbg() calls, which 110 enlarges the kernel text size by about 2%. 111 112 If a source file is compiled with DEBUG flag set, any 113 pr_debug() calls in it are enabled by default, but can be 114 disabled at runtime as below. Note that DEBUG flag is 115 turned on by many CONFIG_*DEBUG* options. 116 117 Usage: 118 119 Dynamic debugging is controlled via the 'dynamic_debug/control' file, 120 which is contained in the 'debugfs' filesystem or procfs. 121 Thus, the debugfs or procfs filesystem must first be mounted before 122 making use of this feature. 123 We refer the control file as: <debugfs>/dynamic_debug/control. This 124 file contains a list of the debug statements that can be enabled. The 125 format for each line of the file is: 126 127 filename:lineno [module]function flags format 128 129 filename : source file of the debug statement 130 lineno : line number of the debug statement 131 module : module that contains the debug statement 132 function : function that contains the debug statement 133 flags : '=p' means the line is turned 'on' for printing 134 format : the format used for the debug statement 135 136 From a live system: 137 138 nullarbor:~ # cat <debugfs>/dynamic_debug/control 139 # filename:lineno [module]function flags format 140 fs/aio.c:222 [aio]__put_ioctx =_ "__put_ioctx:\040freeing\040%p\012" 141 fs/aio.c:248 [aio]ioctx_alloc =_ "ENOMEM:\040nr_events\040too\040high\012" 142 fs/aio.c:1770 [aio]sys_io_cancel =_ "calling\040cancel\012" 143 144 Example usage: 145 146 // enable the message at line 1603 of file svcsock.c 147 nullarbor:~ # echo -n 'file svcsock.c line 1603 +p' > 148 <debugfs>/dynamic_debug/control 149 150 // enable all the messages in file svcsock.c 151 nullarbor:~ # echo -n 'file svcsock.c +p' > 152 <debugfs>/dynamic_debug/control 153 154 // enable all the messages in the NFS server module 155 nullarbor:~ # echo -n 'module nfsd +p' > 156 <debugfs>/dynamic_debug/control 157 158 // enable all 12 messages in the function svc_process() 159 nullarbor:~ # echo -n 'func svc_process +p' > 160 <debugfs>/dynamic_debug/control 161 162 // disable all 12 messages in the function svc_process() 163 nullarbor:~ # echo -n 'func svc_process -p' > 164 <debugfs>/dynamic_debug/control 165 166 See Documentation/admin-guide/dynamic-debug-howto.rst for additional 167 information. 168 169config DYNAMIC_DEBUG_CORE 170 bool "Enable core function of dynamic debug support" 171 depends on PRINTK 172 depends on (DEBUG_FS || PROC_FS) 173 help 174 Enable core functional support of dynamic debug. It is useful 175 when you want to tie dynamic debug to your kernel modules with 176 DYNAMIC_DEBUG_MODULE defined for each of them, especially for 177 the case of embedded system where the kernel image size is 178 sensitive for people. 179 180config SYMBOLIC_ERRNAME 181 bool "Support symbolic error names in printf" 182 default y if PRINTK 183 help 184 If you say Y here, the kernel's printf implementation will 185 be able to print symbolic error names such as ENOSPC instead 186 of the number 28. It makes the kernel image slightly larger 187 (about 3KB), but can make the kernel logs easier to read. 188 189config DEBUG_BUGVERBOSE 190 bool "Verbose BUG() reporting (adds 70K)" if DEBUG_KERNEL && EXPERT 191 depends on BUG && (GENERIC_BUG || HAVE_DEBUG_BUGVERBOSE) 192 default y 193 help 194 Say Y here to make BUG() panics output the file name and line number 195 of the BUG call as well as the EIP and oops trace. This aids 196 debugging but costs about 70-100K of memory. 197 198endmenu # "printk and dmesg options" 199 200menu "Compile-time checks and compiler options" 201 202config DEBUG_INFO 203 bool "Compile the kernel with debug info" 204 depends on DEBUG_KERNEL && !COMPILE_TEST 205 help 206 If you say Y here the resulting kernel image will include 207 debugging info resulting in a larger kernel image. 208 This adds debug symbols to the kernel and modules (gcc -g), and 209 is needed if you intend to use kernel crashdump or binary object 210 tools like crash, kgdb, LKCD, gdb, etc on the kernel. 211 Say Y here only if you plan to debug the kernel. 212 213 If unsure, say N. 214 215if DEBUG_INFO 216 217config DEBUG_INFO_REDUCED 218 bool "Reduce debugging information" 219 help 220 If you say Y here gcc is instructed to generate less debugging 221 information for structure types. This means that tools that 222 need full debugging information (like kgdb or systemtap) won't 223 be happy. But if you merely need debugging information to 224 resolve line numbers there is no loss. Advantage is that 225 build directory object sizes shrink dramatically over a full 226 DEBUG_INFO build and compile times are reduced too. 227 Only works with newer gcc versions. 228 229config DEBUG_INFO_COMPRESSED 230 bool "Compressed debugging information" 231 depends on $(cc-option,-gz=zlib) 232 depends on $(ld-option,--compress-debug-sections=zlib) 233 help 234 Compress the debug information using zlib. Requires GCC 5.0+ or Clang 235 5.0+, binutils 2.26+, and zlib. 236 237 Users of dpkg-deb via scripts/package/builddeb may find an increase in 238 size of their debug .deb packages with this config set, due to the 239 debug info being compressed with zlib, then the object files being 240 recompressed with a different compression scheme. But this is still 241 preferable to setting $KDEB_COMPRESS to "none" which would be even 242 larger. 243 244config DEBUG_INFO_SPLIT 245 bool "Produce split debuginfo in .dwo files" 246 depends on $(cc-option,-gsplit-dwarf) 247 help 248 Generate debug info into separate .dwo files. This significantly 249 reduces the build directory size for builds with DEBUG_INFO, 250 because it stores the information only once on disk in .dwo 251 files instead of multiple times in object files and executables. 252 In addition the debug information is also compressed. 253 254 Requires recent gcc (4.7+) and recent gdb/binutils. 255 Any tool that packages or reads debug information would need 256 to know about the .dwo files and include them. 257 Incompatible with older versions of ccache. 258 259config DEBUG_INFO_DWARF4 260 bool "Generate dwarf4 debuginfo" 261 depends on $(cc-option,-gdwarf-4) 262 help 263 Generate dwarf4 debug info. This requires recent versions 264 of gcc and gdb. It makes the debug information larger. 265 But it significantly improves the success of resolving 266 variables in gdb on optimized code. 267 268config DEBUG_INFO_BTF 269 bool "Generate BTF typeinfo" 270 depends on !DEBUG_INFO_SPLIT && !DEBUG_INFO_REDUCED 271 depends on !GCC_PLUGIN_RANDSTRUCT || COMPILE_TEST 272 help 273 Generate deduplicated BTF type information from DWARF debug info. 274 Turning this on expects presence of pahole tool, which will convert 275 DWARF type info into equivalent deduplicated BTF type info. 276 277config PAHOLE_HAS_SPLIT_BTF 278 def_bool $(success, test `$(PAHOLE) --version | sed -E 's/v([0-9]+)\.([0-9]+)/\1\2/'` -ge "119") 279 280config DEBUG_INFO_BTF_MODULES 281 def_bool y 282 depends on DEBUG_INFO_BTF && MODULES && PAHOLE_HAS_SPLIT_BTF 283 help 284 Generate compact split BTF type information for kernel modules. 285 286config GDB_SCRIPTS 287 bool "Provide GDB scripts for kernel debugging" 288 help 289 This creates the required links to GDB helper scripts in the 290 build directory. If you load vmlinux into gdb, the helper 291 scripts will be automatically imported by gdb as well, and 292 additional functions are available to analyze a Linux kernel 293 instance. See Documentation/dev-tools/gdb-kernel-debugging.rst 294 for further details. 295 296endif # DEBUG_INFO 297 298config ENABLE_MUST_CHECK 299 bool "Enable __must_check logic" 300 default y 301 help 302 Enable the __must_check logic in the kernel build. Disable this to 303 suppress the "warning: ignoring return value of 'foo', declared with 304 attribute warn_unused_result" messages. 305 306config FRAME_WARN 307 int "Warn for stack frames larger than" 308 range 0 8192 309 default 2048 if GCC_PLUGIN_LATENT_ENTROPY 310 default 1280 if (!64BIT && PARISC) 311 default 1024 if (!64BIT && !PARISC) 312 default 2048 if 64BIT 313 help 314 Tell gcc to warn at build time for stack frames larger than this. 315 Setting this too low will cause a lot of warnings. 316 Setting it to 0 disables the warning. 317 318config STRIP_ASM_SYMS 319 bool "Strip assembler-generated symbols during link" 320 default n 321 help 322 Strip internal assembler-generated symbols during a link (symbols 323 that look like '.Lxxx') so they don't pollute the output of 324 get_wchan() and suchlike. 325 326config READABLE_ASM 327 bool "Generate readable assembler code" 328 depends on DEBUG_KERNEL 329 help 330 Disable some compiler optimizations that tend to generate human unreadable 331 assembler output. This may make the kernel slightly slower, but it helps 332 to keep kernel developers who have to stare a lot at assembler listings 333 sane. 334 335config HEADERS_INSTALL 336 bool "Install uapi headers to usr/include" 337 depends on !UML 338 help 339 This option will install uapi headers (headers exported to user-space) 340 into the usr/include directory for use during the kernel build. 341 This is unneeded for building the kernel itself, but needed for some 342 user-space program samples. It is also needed by some features such 343 as uapi header sanity checks. 344 345config DEBUG_SECTION_MISMATCH 346 bool "Enable full Section mismatch analysis" 347 help 348 The section mismatch analysis checks if there are illegal 349 references from one section to another section. 350 During linktime or runtime, some sections are dropped; 351 any use of code/data previously in these sections would 352 most likely result in an oops. 353 In the code, functions and variables are annotated with 354 __init,, etc. (see the full list in include/linux/init.h), 355 which results in the code/data being placed in specific sections. 356 The section mismatch analysis is always performed after a full 357 kernel build, and enabling this option causes the following 358 additional step to occur: 359 - Add the option -fno-inline-functions-called-once to gcc commands. 360 When inlining a function annotated with __init in a non-init 361 function, we would lose the section information and thus 362 the analysis would not catch the illegal reference. 363 This option tells gcc to inline less (but it does result in 364 a larger kernel). 365 366config SECTION_MISMATCH_WARN_ONLY 367 bool "Make section mismatch errors non-fatal" 368 default y 369 help 370 If you say N here, the build process will fail if there are any 371 section mismatch, instead of just throwing warnings. 372 373 If unsure, say Y. 374 375config DEBUG_FORCE_FUNCTION_ALIGN_32B 376 bool "Force all function address 32B aligned" if EXPERT 377 help 378 There are cases that a commit from one domain changes the function 379 address alignment of other domains, and cause magic performance 380 bump (regression or improvement). Enable this option will help to 381 verify if the bump is caused by function alignment changes, while 382 it will slightly increase the kernel size and affect icache usage. 383 384 It is mainly for debug and performance tuning use. 385 386# 387# Select this config option from the architecture Kconfig, if it 388# is preferred to always offer frame pointers as a config 389# option on the architecture (regardless of KERNEL_DEBUG): 390# 391config ARCH_WANT_FRAME_POINTERS 392 bool 393 394config FRAME_POINTER 395 bool "Compile the kernel with frame pointers" 396 depends on DEBUG_KERNEL && (M68K || UML || SUPERH) || ARCH_WANT_FRAME_POINTERS 397 default y if (DEBUG_INFO && UML) || ARCH_WANT_FRAME_POINTERS 398 help 399 If you say Y here the resulting kernel image will be slightly 400 larger and slower, but it gives very useful debugging information 401 in case of kernel bugs. (precise oopses/stacktraces/warnings) 402 403config STACK_VALIDATION 404 bool "Compile-time stack metadata validation" 405 depends on HAVE_STACK_VALIDATION 406 default n 407 help 408 Add compile-time checks to validate stack metadata, including frame 409 pointers (if CONFIG_FRAME_POINTER is enabled). This helps ensure 410 that runtime stack traces are more reliable. 411 412 This is also a prerequisite for generation of ORC unwind data, which 413 is needed for CONFIG_UNWINDER_ORC. 414 415 For more information, see 416 tools/objtool/Documentation/stack-validation.txt. 417 418config VMLINUX_VALIDATION 419 bool 420 depends on STACK_VALIDATION && DEBUG_ENTRY && !PARAVIRT 421 default y 422 423config DEBUG_FORCE_WEAK_PER_CPU 424 bool "Force weak per-cpu definitions" 425 depends on DEBUG_KERNEL 426 help 427 s390 and alpha require percpu variables in modules to be 428 defined weak to work around addressing range issue which 429 puts the following two restrictions on percpu variable 430 definitions. 431 432 1. percpu symbols must be unique whether static or not 433 2. percpu variables can't be defined inside a function 434 435 To ensure that generic code follows the above rules, this 436 option forces all percpu variables to be defined as weak. 437 438endmenu # "Compiler options" 439 440menu "Generic Kernel Debugging Instruments" 441 442config MAGIC_SYSRQ 443 bool "Magic SysRq key" 444 depends on !UML 445 help 446 If you say Y here, you will have some control over the system even 447 if the system crashes for example during kernel debugging (e.g., you 448 will be able to flush the buffer cache to disk, reboot the system 449 immediately or dump some status information). This is accomplished 450 by pressing various keys while holding SysRq (Alt+PrintScreen). It 451 also works on a serial console (on PC hardware at least), if you 452 send a BREAK and then within 5 seconds a command keypress. The 453 keys are documented in <file:Documentation/admin-guide/sysrq.rst>. 454 Don't say Y unless you really know what this hack does. 455 456config MAGIC_SYSRQ_DEFAULT_ENABLE 457 hex "Enable magic SysRq key functions by default" 458 depends on MAGIC_SYSRQ 459 default 0x1 460 help 461 Specifies which SysRq key functions are enabled by default. 462 This may be set to 1 or 0 to enable or disable them all, or 463 to a bitmask as described in Documentation/admin-guide/sysrq.rst. 464 465config MAGIC_SYSRQ_SERIAL 466 bool "Enable magic SysRq key over serial" 467 depends on MAGIC_SYSRQ 468 default y 469 help 470 Many embedded boards have a disconnected TTL level serial which can 471 generate some garbage that can lead to spurious false sysrq detects. 472 This option allows you to decide whether you want to enable the 473 magic SysRq key. 474 475config MAGIC_SYSRQ_SERIAL_SEQUENCE 476 string "Char sequence that enables magic SysRq over serial" 477 depends on MAGIC_SYSRQ_SERIAL 478 default "" 479 help 480 Specifies a sequence of characters that can follow BREAK to enable 481 SysRq on a serial console. 482 483 If unsure, leave an empty string and the option will not be enabled. 484 485config DEBUG_FS 486 bool "Debug Filesystem" 487 help 488 debugfs is a virtual file system that kernel developers use to put 489 debugging files into. Enable this option to be able to read and 490 write to these files. 491 492 For detailed documentation on the debugfs API, see 493 Documentation/filesystems/. 494 495 If unsure, say N. 496 497choice 498 prompt "Debugfs default access" 499 depends on DEBUG_FS 500 default DEBUG_FS_ALLOW_ALL 501 help 502 This selects the default access restrictions for debugfs. 503 It can be overridden with kernel command line option 504 debugfs=[on,no-mount,off]. The restrictions apply for API access 505 and filesystem registration. 506 507config DEBUG_FS_ALLOW_ALL 508 bool "Access normal" 509 help 510 No restrictions apply. Both API and filesystem registration 511 is on. This is the normal default operation. 512 513config DEBUG_FS_DISALLOW_MOUNT 514 bool "Do not register debugfs as filesystem" 515 help 516 The API is open but filesystem is not loaded. Clients can still do 517 their work and read with debug tools that do not need 518 debugfs filesystem. 519 520config DEBUG_FS_ALLOW_NONE 521 bool "No access" 522 help 523 Access is off. Clients get -PERM when trying to create nodes in 524 debugfs tree and debugfs is not registered as a filesystem. 525 Client can then back-off or continue without debugfs access. 526 527endchoice 528 529source "lib/Kconfig.kgdb" 530source "lib/Kconfig.ubsan" 531source "lib/Kconfig.kcsan" 532 533endmenu 534 535config DEBUG_KERNEL 536 bool "Kernel debugging" 537 help 538 Say Y here if you are developing drivers or trying to debug and 539 identify kernel problems. 540 541config DEBUG_MISC 542 bool "Miscellaneous debug code" 543 default DEBUG_KERNEL 544 depends on DEBUG_KERNEL 545 help 546 Say Y here if you need to enable miscellaneous debug code that should 547 be under a more specific debug option but isn't. 548 549 550menu "Memory Debugging" 551 552source "mm/Kconfig.debug" 553 554config DEBUG_OBJECTS 555 bool "Debug object operations" 556 depends on DEBUG_KERNEL 557 help 558 If you say Y here, additional code will be inserted into the 559 kernel to track the life time of various objects and validate 560 the operations on those objects. 561 562config DEBUG_OBJECTS_SELFTEST 563 bool "Debug objects selftest" 564 depends on DEBUG_OBJECTS 565 help 566 This enables the selftest of the object debug code. 567 568config DEBUG_OBJECTS_FREE 569 bool "Debug objects in freed memory" 570 depends on DEBUG_OBJECTS 571 help 572 This enables checks whether a k/v free operation frees an area 573 which contains an object which has not been deactivated 574 properly. This can make kmalloc/kfree-intensive workloads 575 much slower. 576 577config DEBUG_OBJECTS_TIMERS 578 bool "Debug timer objects" 579 depends on DEBUG_OBJECTS 580 help 581 If you say Y here, additional code will be inserted into the 582 timer routines to track the life time of timer objects and 583 validate the timer operations. 584 585config DEBUG_OBJECTS_WORK 586 bool "Debug work objects" 587 depends on DEBUG_OBJECTS 588 help 589 If you say Y here, additional code will be inserted into the 590 work queue routines to track the life time of work objects and 591 validate the work operations. 592 593config DEBUG_OBJECTS_RCU_HEAD 594 bool "Debug RCU callbacks objects" 595 depends on DEBUG_OBJECTS 596 help 597 Enable this to turn on debugging of RCU list heads (call_rcu() usage). 598 599config DEBUG_OBJECTS_PERCPU_COUNTER 600 bool "Debug percpu counter objects" 601 depends on DEBUG_OBJECTS 602 help 603 If you say Y here, additional code will be inserted into the 604 percpu counter routines to track the life time of percpu counter 605 objects and validate the percpu counter operations. 606 607config DEBUG_OBJECTS_ENABLE_DEFAULT 608 int "debug_objects bootup default value (0-1)" 609 range 0 1 610 default "1" 611 depends on DEBUG_OBJECTS 612 help 613 Debug objects boot parameter default value 614 615config DEBUG_SLAB 616 bool "Debug slab memory allocations" 617 depends on DEBUG_KERNEL && SLAB 618 help 619 Say Y here to have the kernel do limited verification on memory 620 allocation as well as poisoning memory on free to catch use of freed 621 memory. This can make kmalloc/kfree-intensive workloads much slower. 622 623config SLUB_DEBUG_ON 624 bool "SLUB debugging on by default" 625 depends on SLUB && SLUB_DEBUG 626 default n 627 help 628 Boot with debugging on by default. SLUB boots by default with 629 the runtime debug capabilities switched off. Enabling this is 630 equivalent to specifying the "slub_debug" parameter on boot. 631 There is no support for more fine grained debug control like 632 possible with slub_debug=xxx. SLUB debugging may be switched 633 off in a kernel built with CONFIG_SLUB_DEBUG_ON by specifying 634 "slub_debug=-". 635 636config SLUB_STATS 637 default n 638 bool "Enable SLUB performance statistics" 639 depends on SLUB && SYSFS 640 help 641 SLUB statistics are useful to debug SLUBs allocation behavior in 642 order find ways to optimize the allocator. This should never be 643 enabled for production use since keeping statistics slows down 644 the allocator by a few percentage points. The slabinfo command 645 supports the determination of the most active slabs to figure 646 out which slabs are relevant to a particular load. 647 Try running: slabinfo -DA 648 649config HAVE_DEBUG_KMEMLEAK 650 bool 651 652config DEBUG_KMEMLEAK 653 bool "Kernel memory leak detector" 654 depends on DEBUG_KERNEL && HAVE_DEBUG_KMEMLEAK 655 select DEBUG_FS 656 select STACKTRACE if STACKTRACE_SUPPORT 657 select KALLSYMS 658 select CRC32 659 help 660 Say Y here if you want to enable the memory leak 661 detector. The memory allocation/freeing is traced in a way 662 similar to the Boehm's conservative garbage collector, the 663 difference being that the orphan objects are not freed but 664 only shown in /sys/kernel/debug/kmemleak. Enabling this 665 feature will introduce an overhead to memory 666 allocations. See Documentation/dev-tools/kmemleak.rst for more 667 details. 668 669 Enabling DEBUG_SLAB or SLUB_DEBUG may increase the chances 670 of finding leaks due to the slab objects poisoning. 671 672 In order to access the kmemleak file, debugfs needs to be 673 mounted (usually at /sys/kernel/debug). 674 675config DEBUG_KMEMLEAK_MEM_POOL_SIZE 676 int "Kmemleak memory pool size" 677 depends on DEBUG_KMEMLEAK 678 range 200 1000000 679 default 16000 680 help 681 Kmemleak must track all the memory allocations to avoid 682 reporting false positives. Since memory may be allocated or 683 freed before kmemleak is fully initialised, use a static pool 684 of metadata objects to track such callbacks. After kmemleak is 685 fully initialised, this memory pool acts as an emergency one 686 if slab allocations fail. 687 688config DEBUG_KMEMLEAK_TEST 689 tristate "Simple test for the kernel memory leak detector" 690 depends on DEBUG_KMEMLEAK && m 691 help 692 This option enables a module that explicitly leaks memory. 693 694 If unsure, say N. 695 696config DEBUG_KMEMLEAK_DEFAULT_OFF 697 bool "Default kmemleak to off" 698 depends on DEBUG_KMEMLEAK 699 help 700 Say Y here to disable kmemleak by default. It can then be enabled 701 on the command line via kmemleak=on. 702 703config DEBUG_KMEMLEAK_AUTO_SCAN 704 bool "Enable kmemleak auto scan thread on boot up" 705 default y 706 depends on DEBUG_KMEMLEAK 707 help 708 Depending on the cpu, kmemleak scan may be cpu intensive and can 709 stall user tasks at times. This option enables/disables automatic 710 kmemleak scan at boot up. 711 712 Say N here to disable kmemleak auto scan thread to stop automatic 713 scanning. Disabling this option disables automatic reporting of 714 memory leaks. 715 716 If unsure, say Y. 717 718config DEBUG_STACK_USAGE 719 bool "Stack utilization instrumentation" 720 depends on DEBUG_KERNEL && !IA64 721 help 722 Enables the display of the minimum amount of free stack which each 723 task has ever had available in the sysrq-T and sysrq-P debug output. 724 725 This option will slow down process creation somewhat. 726 727config SCHED_STACK_END_CHECK 728 bool "Detect stack corruption on calls to schedule()" 729 depends on DEBUG_KERNEL 730 default n 731 help 732 This option checks for a stack overrun on calls to schedule(). 733 If the stack end location is found to be over written always panic as 734 the content of the corrupted region can no longer be trusted. 735 This is to ensure no erroneous behaviour occurs which could result in 736 data corruption or a sporadic crash at a later stage once the region 737 is examined. The runtime overhead introduced is minimal. 738 739config ARCH_HAS_DEBUG_VM_PGTABLE 740 bool 741 help 742 An architecture should select this when it can successfully 743 build and run DEBUG_VM_PGTABLE. 744 745config DEBUG_VM 746 bool "Debug VM" 747 depends on DEBUG_KERNEL 748 help 749 Enable this to turn on extended checks in the virtual-memory system 750 that may impact performance. 751 752 If unsure, say N. 753 754config DEBUG_VM_VMACACHE 755 bool "Debug VMA caching" 756 depends on DEBUG_VM 757 help 758 Enable this to turn on VMA caching debug information. Doing so 759 can cause significant overhead, so only enable it in non-production 760 environments. 761 762 If unsure, say N. 763 764config DEBUG_VM_RB 765 bool "Debug VM red-black trees" 766 depends on DEBUG_VM 767 help 768 Enable VM red-black tree debugging information and extra validations. 769 770 If unsure, say N. 771 772config DEBUG_VM_PGFLAGS 773 bool "Debug page-flags operations" 774 depends on DEBUG_VM 775 help 776 Enables extra validation on page flags operations. 777 778 If unsure, say N. 779 780config DEBUG_VM_PGTABLE 781 bool "Debug arch page table for semantics compliance" 782 depends on MMU 783 depends on ARCH_HAS_DEBUG_VM_PGTABLE 784 default y if DEBUG_VM 785 help 786 This option provides a debug method which can be used to test 787 architecture page table helper functions on various platforms in 788 verifying if they comply with expected generic MM semantics. This 789 will help architecture code in making sure that any changes or 790 new additions of these helpers still conform to expected 791 semantics of the generic MM. Platforms will have to opt in for 792 this through ARCH_HAS_DEBUG_VM_PGTABLE. 793 794 If unsure, say N. 795 796config ARCH_HAS_DEBUG_VIRTUAL 797 bool 798 799config DEBUG_VIRTUAL 800 bool "Debug VM translations" 801 depends on DEBUG_KERNEL && ARCH_HAS_DEBUG_VIRTUAL 802 help 803 Enable some costly sanity checks in virtual to page code. This can 804 catch mistakes with virt_to_page() and friends. 805 806 If unsure, say N. 807 808config DEBUG_NOMMU_REGIONS 809 bool "Debug the global anon/private NOMMU mapping region tree" 810 depends on DEBUG_KERNEL && !MMU 811 help 812 This option causes the global tree of anonymous and private mapping 813 regions to be regularly checked for invalid topology. 814 815config DEBUG_MEMORY_INIT 816 bool "Debug memory initialisation" if EXPERT 817 default !EXPERT 818 help 819 Enable this for additional checks during memory initialisation. 820 The sanity checks verify aspects of the VM such as the memory model 821 and other information provided by the architecture. Verbose 822 information will be printed at KERN_DEBUG loglevel depending 823 on the mminit_loglevel= command-line option. 824 825 If unsure, say Y 826 827config MEMORY_NOTIFIER_ERROR_INJECT 828 tristate "Memory hotplug notifier error injection module" 829 depends on MEMORY_HOTPLUG_SPARSE && NOTIFIER_ERROR_INJECTION 830 help 831 This option provides the ability to inject artificial errors to 832 memory hotplug notifier chain callbacks. It is controlled through 833 debugfs interface under /sys/kernel/debug/notifier-error-inject/memory 834 835 If the notifier call chain should be failed with some events 836 notified, write the error code to "actions/<notifier event>/error". 837 838 Example: Inject memory hotplug offline error (-12 == -ENOMEM) 839 840 # cd /sys/kernel/debug/notifier-error-inject/memory 841 # echo -12 > actions/MEM_GOING_OFFLINE/error 842 # echo offline > /sys/devices/system/memory/memoryXXX/state 843 bash: echo: write error: Cannot allocate memory 844 845 To compile this code as a module, choose M here: the module will 846 be called memory-notifier-error-inject. 847 848 If unsure, say N. 849 850config DEBUG_PER_CPU_MAPS 851 bool "Debug access to per_cpu maps" 852 depends on DEBUG_KERNEL 853 depends on SMP 854 help 855 Say Y to verify that the per_cpu map being accessed has 856 been set up. This adds a fair amount of code to kernel memory 857 and decreases performance. 858 859 Say N if unsure. 860 861config DEBUG_KMAP_LOCAL 862 bool "Debug kmap_local temporary mappings" 863 depends on DEBUG_KERNEL && KMAP_LOCAL 864 help 865 This option enables additional error checking for the kmap_local 866 infrastructure. Disable for production use. 867 868config ARCH_SUPPORTS_KMAP_LOCAL_FORCE_MAP 869 bool 870 871config DEBUG_KMAP_LOCAL_FORCE_MAP 872 bool "Enforce kmap_local temporary mappings" 873 depends on DEBUG_KERNEL && ARCH_SUPPORTS_KMAP_LOCAL_FORCE_MAP 874 select KMAP_LOCAL 875 select DEBUG_KMAP_LOCAL 876 help 877 This option enforces temporary mappings through the kmap_local 878 mechanism for non-highmem pages and on non-highmem systems. 879 Disable this for production systems! 880 881config DEBUG_HIGHMEM 882 bool "Highmem debugging" 883 depends on DEBUG_KERNEL && HIGHMEM 884 select DEBUG_KMAP_LOCAL_FORCE_MAP if ARCH_SUPPORTS_KMAP_LOCAL_FORCE_MAP 885 select DEBUG_KMAP_LOCAL 886 help 887 This option enables additional error checking for high memory 888 systems. Disable for production systems. 889 890config HAVE_DEBUG_STACKOVERFLOW 891 bool 892 893config DEBUG_STACKOVERFLOW 894 bool "Check for stack overflows" 895 depends on DEBUG_KERNEL && HAVE_DEBUG_STACKOVERFLOW 896 help 897 Say Y here if you want to check for overflows of kernel, IRQ 898 and exception stacks (if your architecture uses them). This 899 option will show detailed messages if free stack space drops 900 below a certain limit. 901 902 These kinds of bugs usually occur when call-chains in the 903 kernel get too deep, especially when interrupts are 904 involved. 905 906 Use this in cases where you see apparently random memory 907 corruption, especially if it appears in 'struct thread_info' 908 909 If in doubt, say "N". 910 911source "lib/Kconfig.kasan" 912 913endmenu # "Memory Debugging" 914 915config DEBUG_SHIRQ 916 bool "Debug shared IRQ handlers" 917 depends on DEBUG_KERNEL 918 help 919 Enable this to generate a spurious interrupt just before a shared 920 interrupt handler is deregistered (generating one when registering 921 is currently disabled). Drivers need to handle this correctly. Some 922 don't and need to be caught. 923 924menu "Debug Oops, Lockups and Hangs" 925 926config PANIC_ON_OOPS 927 bool "Panic on Oops" 928 help 929 Say Y here to enable the kernel to panic when it oopses. This 930 has the same effect as setting oops=panic on the kernel command 931 line. 932 933 This feature is useful to ensure that the kernel does not do 934 anything erroneous after an oops which could result in data 935 corruption or other issues. 936 937 Say N if unsure. 938 939config PANIC_ON_OOPS_VALUE 940 int 941 range 0 1 942 default 0 if !PANIC_ON_OOPS 943 default 1 if PANIC_ON_OOPS 944 945config PANIC_TIMEOUT 946 int "panic timeout" 947 default 0 948 help 949 Set the timeout value (in seconds) until a reboot occurs when 950 the kernel panics. If n = 0, then we wait forever. A timeout 951 value n > 0 will wait n seconds before rebooting, while a timeout 952 value n < 0 will reboot immediately. 953 954config LOCKUP_DETECTOR 955 bool 956 957config SOFTLOCKUP_DETECTOR 958 bool "Detect Soft Lockups" 959 depends on DEBUG_KERNEL && !S390 960 select LOCKUP_DETECTOR 961 help 962 Say Y here to enable the kernel to act as a watchdog to detect 963 soft lockups. 964 965 Softlockups are bugs that cause the kernel to loop in kernel 966 mode for more than 20 seconds, without giving other tasks a 967 chance to run. The current stack trace is displayed upon 968 detection and the system will stay locked up. 969 970config BOOTPARAM_SOFTLOCKUP_PANIC 971 bool "Panic (Reboot) On Soft Lockups" 972 depends on SOFTLOCKUP_DETECTOR 973 help 974 Say Y here to enable the kernel to panic on "soft lockups", 975 which are bugs that cause the kernel to loop in kernel 976 mode for more than 20 seconds (configurable using the watchdog_thresh 977 sysctl), without giving other tasks a chance to run. 978 979 The panic can be used in combination with panic_timeout, 980 to cause the system to reboot automatically after a 981 lockup has been detected. This feature is useful for 982 high-availability systems that have uptime guarantees and 983 where a lockup must be resolved ASAP. 984 985 Say N if unsure. 986 987config BOOTPARAM_SOFTLOCKUP_PANIC_VALUE 988 int 989 depends on SOFTLOCKUP_DETECTOR 990 range 0 1 991 default 0 if !BOOTPARAM_SOFTLOCKUP_PANIC 992 default 1 if BOOTPARAM_SOFTLOCKUP_PANIC 993 994config HARDLOCKUP_DETECTOR_PERF 995 bool 996 select SOFTLOCKUP_DETECTOR 997 998# 999# Enables a timestamp based low pass filter to compensate for perf based 1000# hard lockup detection which runs too fast due to turbo modes. 1001# 1002config HARDLOCKUP_CHECK_TIMESTAMP 1003 bool 1004 1005# 1006# arch/ can define HAVE_HARDLOCKUP_DETECTOR_ARCH to provide their own hard 1007# lockup detector rather than the perf based detector. 1008# 1009config HARDLOCKUP_DETECTOR 1010 bool "Detect Hard Lockups" 1011 depends on DEBUG_KERNEL && !S390 1012 depends on HAVE_HARDLOCKUP_DETECTOR_PERF || HAVE_HARDLOCKUP_DETECTOR_ARCH 1013 select LOCKUP_DETECTOR 1014 select HARDLOCKUP_DETECTOR_PERF if HAVE_HARDLOCKUP_DETECTOR_PERF 1015 select HARDLOCKUP_DETECTOR_ARCH if HAVE_HARDLOCKUP_DETECTOR_ARCH 1016 help 1017 Say Y here to enable the kernel to act as a watchdog to detect 1018 hard lockups. 1019 1020 Hardlockups are bugs that cause the CPU to loop in kernel mode 1021 for more than 10 seconds, without letting other interrupts have a 1022 chance to run. The current stack trace is displayed upon detection 1023 and the system will stay locked up. 1024 1025config BOOTPARAM_HARDLOCKUP_PANIC 1026 bool "Panic (Reboot) On Hard Lockups" 1027 depends on HARDLOCKUP_DETECTOR 1028 help 1029 Say Y here to enable the kernel to panic on "hard lockups", 1030 which are bugs that cause the kernel to loop in kernel 1031 mode with interrupts disabled for more than 10 seconds (configurable 1032 using the watchdog_thresh sysctl). 1033 1034 Say N if unsure. 1035 1036config BOOTPARAM_HARDLOCKUP_PANIC_VALUE 1037 int 1038 depends on HARDLOCKUP_DETECTOR 1039 range 0 1 1040 default 0 if !BOOTPARAM_HARDLOCKUP_PANIC 1041 default 1 if BOOTPARAM_HARDLOCKUP_PANIC 1042 1043config DETECT_HUNG_TASK 1044 bool "Detect Hung Tasks" 1045 depends on DEBUG_KERNEL 1046 default SOFTLOCKUP_DETECTOR 1047 help 1048 Say Y here to enable the kernel to detect "hung tasks", 1049 which are bugs that cause the task to be stuck in 1050 uninterruptible "D" state indefinitely. 1051 1052 When a hung task is detected, the kernel will print the 1053 current stack trace (which you should report), but the 1054 task will stay in uninterruptible state. If lockdep is 1055 enabled then all held locks will also be reported. This 1056 feature has negligible overhead. 1057 1058config DEFAULT_HUNG_TASK_TIMEOUT 1059 int "Default timeout for hung task detection (in seconds)" 1060 depends on DETECT_HUNG_TASK 1061 default 120 1062 help 1063 This option controls the default timeout (in seconds) used 1064 to determine when a task has become non-responsive and should 1065 be considered hung. 1066 1067 It can be adjusted at runtime via the kernel.hung_task_timeout_secs 1068 sysctl or by writing a value to 1069 /proc/sys/kernel/hung_task_timeout_secs. 1070 1071 A timeout of 0 disables the check. The default is two minutes. 1072 Keeping the default should be fine in most cases. 1073 1074config BOOTPARAM_HUNG_TASK_PANIC 1075 bool "Panic (Reboot) On Hung Tasks" 1076 depends on DETECT_HUNG_TASK 1077 help 1078 Say Y here to enable the kernel to panic on "hung tasks", 1079 which are bugs that cause the kernel to leave a task stuck 1080 in uninterruptible "D" state. 1081 1082 The panic can be used in combination with panic_timeout, 1083 to cause the system to reboot automatically after a 1084 hung task has been detected. This feature is useful for 1085 high-availability systems that have uptime guarantees and 1086 where a hung tasks must be resolved ASAP. 1087 1088 Say N if unsure. 1089 1090config BOOTPARAM_HUNG_TASK_PANIC_VALUE 1091 int 1092 depends on DETECT_HUNG_TASK 1093 range 0 1 1094 default 0 if !BOOTPARAM_HUNG_TASK_PANIC 1095 default 1 if BOOTPARAM_HUNG_TASK_PANIC 1096 1097config WQ_WATCHDOG 1098 bool "Detect Workqueue Stalls" 1099 depends on DEBUG_KERNEL 1100 help 1101 Say Y here to enable stall detection on workqueues. If a 1102 worker pool doesn't make forward progress on a pending work 1103 item for over a given amount of time, 30s by default, a 1104 warning message is printed along with dump of workqueue 1105 state. This can be configured through kernel parameter 1106 "workqueue.watchdog_thresh" and its sysfs counterpart. 1107 1108config TEST_LOCKUP 1109 tristate "Test module to generate lockups" 1110 depends on m 1111 help 1112 This builds the "test_lockup" module that helps to make sure 1113 that watchdogs and lockup detectors are working properly. 1114 1115 Depending on module parameters it could emulate soft or hard 1116 lockup, "hung task", or locking arbitrary lock for a long time. 1117 Also it could generate series of lockups with cooling-down periods. 1118 1119 If unsure, say N. 1120 1121endmenu # "Debug lockups and hangs" 1122 1123menu "Scheduler Debugging" 1124 1125config SCHED_DEBUG 1126 bool "Collect scheduler debugging info" 1127 depends on DEBUG_KERNEL && PROC_FS 1128 default y 1129 help 1130 If you say Y here, the /proc/sched_debug file will be provided 1131 that can help debug the scheduler. The runtime overhead of this 1132 option is minimal. 1133 1134config SCHED_INFO 1135 bool 1136 default n 1137 1138config SCHEDSTATS 1139 bool "Collect scheduler statistics" 1140 depends on DEBUG_KERNEL && PROC_FS 1141 select SCHED_INFO 1142 help 1143 If you say Y here, additional code will be inserted into the 1144 scheduler and related routines to collect statistics about 1145 scheduler behavior and provide them in /proc/schedstat. These 1146 stats may be useful for both tuning and debugging the scheduler 1147 If you aren't debugging the scheduler or trying to tune a specific 1148 application, you can say N to avoid the very slight overhead 1149 this adds. 1150 1151endmenu 1152 1153config DEBUG_TIMEKEEPING 1154 bool "Enable extra timekeeping sanity checking" 1155 help 1156 This option will enable additional timekeeping sanity checks 1157 which may be helpful when diagnosing issues where timekeeping 1158 problems are suspected. 1159 1160 This may include checks in the timekeeping hotpaths, so this 1161 option may have a (very small) performance impact to some 1162 workloads. 1163 1164 If unsure, say N. 1165 1166config DEBUG_PREEMPT 1167 bool "Debug preemptible kernel" 1168 depends on DEBUG_KERNEL && PREEMPTION && TRACE_IRQFLAGS_SUPPORT 1169 default y 1170 help 1171 If you say Y here then the kernel will use a debug variant of the 1172 commonly used smp_processor_id() function and will print warnings 1173 if kernel code uses it in a preemption-unsafe way. Also, the kernel 1174 will detect preemption count underflows. 1175 1176menu "Lock Debugging (spinlocks, mutexes, etc...)" 1177 1178config LOCK_DEBUGGING_SUPPORT 1179 bool 1180 depends on TRACE_IRQFLAGS_SUPPORT && STACKTRACE_SUPPORT && LOCKDEP_SUPPORT 1181 default y 1182 1183config PROVE_LOCKING 1184 bool "Lock debugging: prove locking correctness" 1185 depends on DEBUG_KERNEL && LOCK_DEBUGGING_SUPPORT 1186 select LOCKDEP 1187 select DEBUG_SPINLOCK 1188 select DEBUG_MUTEXES 1189 select DEBUG_RT_MUTEXES if RT_MUTEXES 1190 select DEBUG_RWSEMS 1191 select DEBUG_WW_MUTEX_SLOWPATH 1192 select DEBUG_LOCK_ALLOC 1193 select PREEMPT_COUNT if !ARCH_NO_PREEMPT 1194 select TRACE_IRQFLAGS 1195 default n 1196 help 1197 This feature enables the kernel to prove that all locking 1198 that occurs in the kernel runtime is mathematically 1199 correct: that under no circumstance could an arbitrary (and 1200 not yet triggered) combination of observed locking 1201 sequences (on an arbitrary number of CPUs, running an 1202 arbitrary number of tasks and interrupt contexts) cause a 1203 deadlock. 1204 1205 In short, this feature enables the kernel to report locking 1206 related deadlocks before they actually occur. 1207 1208 The proof does not depend on how hard and complex a 1209 deadlock scenario would be to trigger: how many 1210 participant CPUs, tasks and irq-contexts would be needed 1211 for it to trigger. The proof also does not depend on 1212 timing: if a race and a resulting deadlock is possible 1213 theoretically (no matter how unlikely the race scenario 1214 is), it will be proven so and will immediately be 1215 reported by the kernel (once the event is observed that 1216 makes the deadlock theoretically possible). 1217 1218 If a deadlock is impossible (i.e. the locking rules, as 1219 observed by the kernel, are mathematically correct), the 1220 kernel reports nothing. 1221 1222 NOTE: this feature can also be enabled for rwlocks, mutexes 1223 and rwsems - in which case all dependencies between these 1224 different locking variants are observed and mapped too, and 1225 the proof of observed correctness is also maintained for an 1226 arbitrary combination of these separate locking variants. 1227 1228 For more details, see Documentation/locking/lockdep-design.rst. 1229 1230config PROVE_RAW_LOCK_NESTING 1231 bool "Enable raw_spinlock - spinlock nesting checks" 1232 depends on PROVE_LOCKING 1233 default n 1234 help 1235 Enable the raw_spinlock vs. spinlock nesting checks which ensure 1236 that the lock nesting rules for PREEMPT_RT enabled kernels are 1237 not violated. 1238 1239 NOTE: There are known nesting problems. So if you enable this 1240 option expect lockdep splats until these problems have been fully 1241 addressed which is work in progress. This config switch allows to 1242 identify and analyze these problems. It will be removed and the 1243 check permanentely enabled once the main issues have been fixed. 1244 1245 If unsure, select N. 1246 1247config LOCK_STAT 1248 bool "Lock usage statistics" 1249 depends on DEBUG_KERNEL && LOCK_DEBUGGING_SUPPORT 1250 select LOCKDEP 1251 select DEBUG_SPINLOCK 1252 select DEBUG_MUTEXES 1253 select DEBUG_RT_MUTEXES if RT_MUTEXES 1254 select DEBUG_LOCK_ALLOC 1255 default n 1256 help 1257 This feature enables tracking lock contention points 1258 1259 For more details, see Documentation/locking/lockstat.rst 1260 1261 This also enables lock events required by "perf lock", 1262 subcommand of perf. 1263 If you want to use "perf lock", you also need to turn on 1264 CONFIG_EVENT_TRACING. 1265 1266 CONFIG_LOCK_STAT defines "contended" and "acquired" lock events. 1267 (CONFIG_LOCKDEP defines "acquire" and "release" events.) 1268 1269config DEBUG_RT_MUTEXES 1270 bool "RT Mutex debugging, deadlock detection" 1271 depends on DEBUG_KERNEL && RT_MUTEXES 1272 help 1273 This allows rt mutex semantics violations and rt mutex related 1274 deadlocks (lockups) to be detected and reported automatically. 1275 1276config DEBUG_SPINLOCK 1277 bool "Spinlock and rw-lock debugging: basic checks" 1278 depends on DEBUG_KERNEL 1279 select UNINLINE_SPIN_UNLOCK 1280 help 1281 Say Y here and build SMP to catch missing spinlock initialization 1282 and certain other kinds of spinlock errors commonly made. This is 1283 best used in conjunction with the NMI watchdog so that spinlock 1284 deadlocks are also debuggable. 1285 1286config DEBUG_MUTEXES 1287 bool "Mutex debugging: basic checks" 1288 depends on DEBUG_KERNEL 1289 help 1290 This feature allows mutex semantics violations to be detected and 1291 reported. 1292 1293config DEBUG_WW_MUTEX_SLOWPATH 1294 bool "Wait/wound mutex debugging: Slowpath testing" 1295 depends on DEBUG_KERNEL && LOCK_DEBUGGING_SUPPORT 1296 select DEBUG_LOCK_ALLOC 1297 select DEBUG_SPINLOCK 1298 select DEBUG_MUTEXES 1299 help 1300 This feature enables slowpath testing for w/w mutex users by 1301 injecting additional -EDEADLK wound/backoff cases. Together with 1302 the full mutex checks enabled with (CONFIG_PROVE_LOCKING) this 1303 will test all possible w/w mutex interface abuse with the 1304 exception of simply not acquiring all the required locks. 1305 Note that this feature can introduce significant overhead, so 1306 it really should not be enabled in a production or distro kernel, 1307 even a debug kernel. If you are a driver writer, enable it. If 1308 you are a distro, do not. 1309 1310config DEBUG_RWSEMS 1311 bool "RW Semaphore debugging: basic checks" 1312 depends on DEBUG_KERNEL 1313 help 1314 This debugging feature allows mismatched rw semaphore locks 1315 and unlocks to be detected and reported. 1316 1317config DEBUG_LOCK_ALLOC 1318 bool "Lock debugging: detect incorrect freeing of live locks" 1319 depends on DEBUG_KERNEL && LOCK_DEBUGGING_SUPPORT 1320 select DEBUG_SPINLOCK 1321 select DEBUG_MUTEXES 1322 select DEBUG_RT_MUTEXES if RT_MUTEXES 1323 select LOCKDEP 1324 help 1325 This feature will check whether any held lock (spinlock, rwlock, 1326 mutex or rwsem) is incorrectly freed by the kernel, via any of the 1327 memory-freeing routines (kfree(), kmem_cache_free(), free_pages(), 1328 vfree(), etc.), whether a live lock is incorrectly reinitialized via 1329 spin_lock_init()/mutex_init()/etc., or whether there is any lock 1330 held during task exit. 1331 1332config LOCKDEP 1333 bool 1334 depends on DEBUG_KERNEL && LOCK_DEBUGGING_SUPPORT 1335 select STACKTRACE 1336 select FRAME_POINTER if !MIPS && !PPC && !ARM && !S390 && !MICROBLAZE && !ARC && !X86 1337 select KALLSYMS 1338 select KALLSYMS_ALL 1339 1340config LOCKDEP_SMALL 1341 bool 1342 1343config DEBUG_LOCKDEP 1344 bool "Lock dependency engine debugging" 1345 depends on DEBUG_KERNEL && LOCKDEP 1346 help 1347 If you say Y here, the lock dependency engine will do 1348 additional runtime checks to debug itself, at the price 1349 of more runtime overhead. 1350 1351config DEBUG_ATOMIC_SLEEP 1352 bool "Sleep inside atomic section checking" 1353 select PREEMPT_COUNT 1354 depends on DEBUG_KERNEL 1355 depends on !ARCH_NO_PREEMPT 1356 help 1357 If you say Y here, various routines which may sleep will become very 1358 noisy if they are called inside atomic sections: when a spinlock is 1359 held, inside an rcu read side critical section, inside preempt disabled 1360 sections, inside an interrupt, etc... 1361 1362config DEBUG_LOCKING_API_SELFTESTS 1363 bool "Locking API boot-time self-tests" 1364 depends on DEBUG_KERNEL 1365 help 1366 Say Y here if you want the kernel to run a short self-test during 1367 bootup. The self-test checks whether common types of locking bugs 1368 are detected by debugging mechanisms or not. (if you disable 1369 lock debugging then those bugs wont be detected of course.) 1370 The following locking APIs are covered: spinlocks, rwlocks, 1371 mutexes and rwsems. 1372 1373config LOCK_TORTURE_TEST 1374 tristate "torture tests for locking" 1375 depends on DEBUG_KERNEL 1376 select TORTURE_TEST 1377 help 1378 This option provides a kernel module that runs torture tests 1379 on kernel locking primitives. The kernel module may be built 1380 after the fact on the running kernel to be tested, if desired. 1381 1382 Say Y here if you want kernel locking-primitive torture tests 1383 to be built into the kernel. 1384 Say M if you want these torture tests to build as a module. 1385 Say N if you are unsure. 1386 1387config WW_MUTEX_SELFTEST 1388 tristate "Wait/wound mutex selftests" 1389 help 1390 This option provides a kernel module that runs tests on the 1391 on the struct ww_mutex locking API. 1392 1393 It is recommended to enable DEBUG_WW_MUTEX_SLOWPATH in conjunction 1394 with this test harness. 1395 1396 Say M if you want these self tests to build as a module. 1397 Say N if you are unsure. 1398 1399config SCF_TORTURE_TEST 1400 tristate "torture tests for smp_call_function*()" 1401 depends on DEBUG_KERNEL 1402 select TORTURE_TEST 1403 help 1404 This option provides a kernel module that runs torture tests 1405 on the smp_call_function() family of primitives. The kernel 1406 module may be built after the fact on the running kernel to 1407 be tested, if desired. 1408 1409config CSD_LOCK_WAIT_DEBUG 1410 bool "Debugging for csd_lock_wait(), called from smp_call_function*()" 1411 depends on DEBUG_KERNEL 1412 depends on 64BIT 1413 default n 1414 help 1415 This option enables debug prints when CPUs are slow to respond 1416 to the smp_call_function*() IPI wrappers. These debug prints 1417 include the IPI handler function currently executing (if any) 1418 and relevant stack traces. 1419 1420endmenu # lock debugging 1421 1422config TRACE_IRQFLAGS 1423 depends on TRACE_IRQFLAGS_SUPPORT 1424 bool 1425 help 1426 Enables hooks to interrupt enabling and disabling for 1427 either tracing or lock debugging. 1428 1429config TRACE_IRQFLAGS_NMI 1430 def_bool y 1431 depends on TRACE_IRQFLAGS 1432 depends on TRACE_IRQFLAGS_NMI_SUPPORT 1433 1434config STACKTRACE 1435 bool "Stack backtrace support" 1436 depends on STACKTRACE_SUPPORT 1437 help 1438 This option causes the kernel to create a /proc/pid/stack for 1439 every process, showing its current stack trace. 1440 It is also used by various kernel debugging features that require 1441 stack trace generation. 1442 1443config WARN_ALL_UNSEEDED_RANDOM 1444 bool "Warn for all uses of unseeded randomness" 1445 default n 1446 help 1447 Some parts of the kernel contain bugs relating to their use of 1448 cryptographically secure random numbers before it's actually possible 1449 to generate those numbers securely. This setting ensures that these 1450 flaws don't go unnoticed, by enabling a message, should this ever 1451 occur. This will allow people with obscure setups to know when things 1452 are going wrong, so that they might contact developers about fixing 1453 it. 1454 1455 Unfortunately, on some models of some architectures getting 1456 a fully seeded CRNG is extremely difficult, and so this can 1457 result in dmesg getting spammed for a surprisingly long 1458 time. This is really bad from a security perspective, and 1459 so architecture maintainers really need to do what they can 1460 to get the CRNG seeded sooner after the system is booted. 1461 However, since users cannot do anything actionable to 1462 address this, by default the kernel will issue only a single 1463 warning for the first use of unseeded randomness. 1464 1465 Say Y here if you want to receive warnings for all uses of 1466 unseeded randomness. This will be of use primarily for 1467 those developers interested in improving the security of 1468 Linux kernels running on their architecture (or 1469 subarchitecture). 1470 1471config DEBUG_KOBJECT 1472 bool "kobject debugging" 1473 depends on DEBUG_KERNEL 1474 help 1475 If you say Y here, some extra kobject debugging messages will be sent 1476 to the syslog. 1477 1478config DEBUG_KOBJECT_RELEASE 1479 bool "kobject release debugging" 1480 depends on DEBUG_OBJECTS_TIMERS 1481 help 1482 kobjects are reference counted objects. This means that their 1483 last reference count put is not predictable, and the kobject can 1484 live on past the point at which a driver decides to drop it's 1485 initial reference to the kobject gained on allocation. An 1486 example of this would be a struct device which has just been 1487 unregistered. 1488 1489 However, some buggy drivers assume that after such an operation, 1490 the memory backing the kobject can be immediately freed. This 1491 goes completely against the principles of a refcounted object. 1492 1493 If you say Y here, the kernel will delay the release of kobjects 1494 on the last reference count to improve the visibility of this 1495 kind of kobject release bug. 1496 1497config HAVE_DEBUG_BUGVERBOSE 1498 bool 1499 1500menu "Debug kernel data structures" 1501 1502config DEBUG_LIST 1503 bool "Debug linked list manipulation" 1504 depends on DEBUG_KERNEL || BUG_ON_DATA_CORRUPTION 1505 help 1506 Enable this to turn on extended checks in the linked-list 1507 walking routines. 1508 1509 If unsure, say N. 1510 1511config DEBUG_PLIST 1512 bool "Debug priority linked list manipulation" 1513 depends on DEBUG_KERNEL 1514 help 1515 Enable this to turn on extended checks in the priority-ordered 1516 linked-list (plist) walking routines. This checks the entire 1517 list multiple times during each manipulation. 1518 1519 If unsure, say N. 1520 1521config DEBUG_SG 1522 bool "Debug SG table operations" 1523 depends on DEBUG_KERNEL 1524 help 1525 Enable this to turn on checks on scatter-gather tables. This can 1526 help find problems with drivers that do not properly initialize 1527 their sg tables. 1528 1529 If unsure, say N. 1530 1531config DEBUG_NOTIFIERS 1532 bool "Debug notifier call chains" 1533 depends on DEBUG_KERNEL 1534 help 1535 Enable this to turn on sanity checking for notifier call chains. 1536 This is most useful for kernel developers to make sure that 1537 modules properly unregister themselves from notifier chains. 1538 This is a relatively cheap check but if you care about maximum 1539 performance, say N. 1540 1541config BUG_ON_DATA_CORRUPTION 1542 bool "Trigger a BUG when data corruption is detected" 1543 select DEBUG_LIST 1544 help 1545 Select this option if the kernel should BUG when it encounters 1546 data corruption in kernel memory structures when they get checked 1547 for validity. 1548 1549 If unsure, say N. 1550 1551endmenu 1552 1553config DEBUG_CREDENTIALS 1554 bool "Debug credential management" 1555 depends on DEBUG_KERNEL 1556 help 1557 Enable this to turn on some debug checking for credential 1558 management. The additional code keeps track of the number of 1559 pointers from task_structs to any given cred struct, and checks to 1560 see that this number never exceeds the usage count of the cred 1561 struct. 1562 1563 Furthermore, if SELinux is enabled, this also checks that the 1564 security pointer in the cred struct is never seen to be invalid. 1565 1566 If unsure, say N. 1567 1568source "kernel/rcu/Kconfig.debug" 1569 1570config DEBUG_WQ_FORCE_RR_CPU 1571 bool "Force round-robin CPU selection for unbound work items" 1572 depends on DEBUG_KERNEL 1573 default n 1574 help 1575 Workqueue used to implicitly guarantee that work items queued 1576 without explicit CPU specified are put on the local CPU. This 1577 guarantee is no longer true and while local CPU is still 1578 preferred work items may be put on foreign CPUs. Kernel 1579 parameter "workqueue.debug_force_rr_cpu" is added to force 1580 round-robin CPU selection to flush out usages which depend on the 1581 now broken guarantee. This config option enables the debug 1582 feature by default. When enabled, memory and cache locality will 1583 be impacted. 1584 1585config DEBUG_BLOCK_EXT_DEVT 1586 bool "Force extended block device numbers and spread them" 1587 depends on DEBUG_KERNEL 1588 depends on BLOCK 1589 default n 1590 help 1591 BIG FAT WARNING: ENABLING THIS OPTION MIGHT BREAK BOOTING ON 1592 SOME DISTRIBUTIONS. DO NOT ENABLE THIS UNLESS YOU KNOW WHAT 1593 YOU ARE DOING. Distros, please enable this and fix whatever 1594 is broken. 1595 1596 Conventionally, block device numbers are allocated from 1597 predetermined contiguous area. However, extended block area 1598 may introduce non-contiguous block device numbers. This 1599 option forces most block device numbers to be allocated from 1600 the extended space and spreads them to discover kernel or 1601 userland code paths which assume predetermined contiguous 1602 device number allocation. 1603 1604 Note that turning on this debug option shuffles all the 1605 device numbers for all IDE and SCSI devices including libata 1606 ones, so root partition specified using device number 1607 directly (via rdev or root=MAJ:MIN) won't work anymore. 1608 Textual device names (root=/dev/sdXn) will continue to work. 1609 1610 Say N if you are unsure. 1611 1612config CPU_HOTPLUG_STATE_CONTROL 1613 bool "Enable CPU hotplug state control" 1614 depends on DEBUG_KERNEL 1615 depends on HOTPLUG_CPU 1616 default n 1617 help 1618 Allows to write steps between "offline" and "online" to the CPUs 1619 sysfs target file so states can be stepped granular. This is a debug 1620 option for now as the hotplug machinery cannot be stopped and 1621 restarted at arbitrary points yet. 1622 1623 Say N if your are unsure. 1624 1625config LATENCYTOP 1626 bool "Latency measuring infrastructure" 1627 depends on DEBUG_KERNEL 1628 depends on STACKTRACE_SUPPORT 1629 depends on PROC_FS 1630 select FRAME_POINTER if !MIPS && !PPC && !S390 && !MICROBLAZE && !ARM && !ARC && !X86 1631 select KALLSYMS 1632 select KALLSYMS_ALL 1633 select STACKTRACE 1634 select SCHEDSTATS 1635 select SCHED_DEBUG 1636 help 1637 Enable this option if you want to use the LatencyTOP tool 1638 to find out which userspace is blocking on what kernel operations. 1639 1640source "kernel/trace/Kconfig" 1641 1642config PROVIDE_OHCI1394_DMA_INIT 1643 bool "Remote debugging over FireWire early on boot" 1644 depends on PCI && X86 1645 help 1646 If you want to debug problems which hang or crash the kernel early 1647 on boot and the crashing machine has a FireWire port, you can use 1648 this feature to remotely access the memory of the crashed machine 1649 over FireWire. This employs remote DMA as part of the OHCI1394 1650 specification which is now the standard for FireWire controllers. 1651 1652 With remote DMA, you can monitor the printk buffer remotely using 1653 firescope and access all memory below 4GB using fireproxy from gdb. 1654 Even controlling a kernel debugger is possible using remote DMA. 1655 1656 Usage: 1657 1658 If ohci1394_dma=early is used as boot parameter, it will initialize 1659 all OHCI1394 controllers which are found in the PCI config space. 1660 1661 As all changes to the FireWire bus such as enabling and disabling 1662 devices cause a bus reset and thereby disable remote DMA for all 1663 devices, be sure to have the cable plugged and FireWire enabled on 1664 the debugging host before booting the debug target for debugging. 1665 1666 This code (~1k) is freed after boot. By then, the firewire stack 1667 in charge of the OHCI-1394 controllers should be used instead. 1668 1669 See Documentation/core-api/debugging-via-ohci1394.rst for more information. 1670 1671source "samples/Kconfig" 1672 1673config ARCH_HAS_DEVMEM_IS_ALLOWED 1674 bool 1675 1676config STRICT_DEVMEM 1677 bool "Filter access to /dev/mem" 1678 depends on MMU && DEVMEM 1679 depends on ARCH_HAS_DEVMEM_IS_ALLOWED || GENERIC_LIB_DEVMEM_IS_ALLOWED 1680 default y if PPC || X86 || ARM64 1681 help 1682 If this option is disabled, you allow userspace (root) access to all 1683 of memory, including kernel and userspace memory. Accidental 1684 access to this is obviously disastrous, but specific access can 1685 be used by people debugging the kernel. Note that with PAT support 1686 enabled, even in this case there are restrictions on /dev/mem 1687 use due to the cache aliasing requirements. 1688 1689 If this option is switched on, and IO_STRICT_DEVMEM=n, the /dev/mem 1690 file only allows userspace access to PCI space and the BIOS code and 1691 data regions. This is sufficient for dosemu and X and all common 1692 users of /dev/mem. 1693 1694 If in doubt, say Y. 1695 1696config IO_STRICT_DEVMEM 1697 bool "Filter I/O access to /dev/mem" 1698 depends on STRICT_DEVMEM 1699 help 1700 If this option is disabled, you allow userspace (root) access to all 1701 io-memory regardless of whether a driver is actively using that 1702 range. Accidental access to this is obviously disastrous, but 1703 specific access can be used by people debugging kernel drivers. 1704 1705 If this option is switched on, the /dev/mem file only allows 1706 userspace access to *idle* io-memory ranges (see /proc/iomem) This 1707 may break traditional users of /dev/mem (dosemu, legacy X, etc...) 1708 if the driver using a given range cannot be disabled. 1709 1710 If in doubt, say Y. 1711 1712menu "$(SRCARCH) Debugging" 1713 1714source "arch/$(SRCARCH)/Kconfig.debug" 1715 1716endmenu 1717 1718menu "Kernel Testing and Coverage" 1719 1720source "lib/kunit/Kconfig" 1721 1722config NOTIFIER_ERROR_INJECTION 1723 tristate "Notifier error injection" 1724 depends on DEBUG_KERNEL 1725 select DEBUG_FS 1726 help 1727 This option provides the ability to inject artificial errors to 1728 specified notifier chain callbacks. It is useful to test the error 1729 handling of notifier call chain failures. 1730 1731 Say N if unsure. 1732 1733config PM_NOTIFIER_ERROR_INJECT 1734 tristate "PM notifier error injection module" 1735 depends on PM && NOTIFIER_ERROR_INJECTION 1736 default m if PM_DEBUG 1737 help 1738 This option provides the ability to inject artificial errors to 1739 PM notifier chain callbacks. It is controlled through debugfs 1740 interface /sys/kernel/debug/notifier-error-inject/pm 1741 1742 If the notifier call chain should be failed with some events 1743 notified, write the error code to "actions/<notifier event>/error". 1744 1745 Example: Inject PM suspend error (-12 = -ENOMEM) 1746 1747 # cd /sys/kernel/debug/notifier-error-inject/pm/ 1748 # echo -12 > actions/PM_SUSPEND_PREPARE/error 1749 # echo mem > /sys/power/state 1750 bash: echo: write error: Cannot allocate memory 1751 1752 To compile this code as a module, choose M here: the module will 1753 be called pm-notifier-error-inject. 1754 1755 If unsure, say N. 1756 1757config OF_RECONFIG_NOTIFIER_ERROR_INJECT 1758 tristate "OF reconfig notifier error injection module" 1759 depends on OF_DYNAMIC && NOTIFIER_ERROR_INJECTION 1760 help 1761 This option provides the ability to inject artificial errors to 1762 OF reconfig notifier chain callbacks. It is controlled 1763 through debugfs interface under 1764 /sys/kernel/debug/notifier-error-inject/OF-reconfig/ 1765 1766 If the notifier call chain should be failed with some events 1767 notified, write the error code to "actions/<notifier event>/error". 1768 1769 To compile this code as a module, choose M here: the module will 1770 be called of-reconfig-notifier-error-inject. 1771 1772 If unsure, say N. 1773 1774config NETDEV_NOTIFIER_ERROR_INJECT 1775 tristate "Netdev notifier error injection module" 1776 depends on NET && NOTIFIER_ERROR_INJECTION 1777 help 1778 This option provides the ability to inject artificial errors to 1779 netdevice notifier chain callbacks. It is controlled through debugfs 1780 interface /sys/kernel/debug/notifier-error-inject/netdev 1781 1782 If the notifier call chain should be failed with some events 1783 notified, write the error code to "actions/<notifier event>/error". 1784 1785 Example: Inject netdevice mtu change error (-22 = -EINVAL) 1786 1787 # cd /sys/kernel/debug/notifier-error-inject/netdev 1788 # echo -22 > actions/NETDEV_CHANGEMTU/error 1789 # ip link set eth0 mtu 1024 1790 RTNETLINK answers: Invalid argument 1791 1792 To compile this code as a module, choose M here: the module will 1793 be called netdev-notifier-error-inject. 1794 1795 If unsure, say N. 1796 1797config FUNCTION_ERROR_INJECTION 1798 def_bool y 1799 depends on HAVE_FUNCTION_ERROR_INJECTION && KPROBES 1800 1801config FAULT_INJECTION 1802 bool "Fault-injection framework" 1803 depends on DEBUG_KERNEL 1804 help 1805 Provide fault-injection framework. 1806 For more details, see Documentation/fault-injection/. 1807 1808config FAILSLAB 1809 bool "Fault-injection capability for kmalloc" 1810 depends on FAULT_INJECTION 1811 depends on SLAB || SLUB 1812 help 1813 Provide fault-injection capability for kmalloc. 1814 1815config FAIL_PAGE_ALLOC 1816 bool "Fault-injection capability for alloc_pages()" 1817 depends on FAULT_INJECTION 1818 help 1819 Provide fault-injection capability for alloc_pages(). 1820 1821config FAULT_INJECTION_USERCOPY 1822 bool "Fault injection capability for usercopy functions" 1823 depends on FAULT_INJECTION 1824 help 1825 Provides fault-injection capability to inject failures 1826 in usercopy functions (copy_from_user(), get_user(), ...). 1827 1828config FAIL_MAKE_REQUEST 1829 bool "Fault-injection capability for disk IO" 1830 depends on FAULT_INJECTION && BLOCK 1831 help 1832 Provide fault-injection capability for disk IO. 1833 1834config FAIL_IO_TIMEOUT 1835 bool "Fault-injection capability for faking disk interrupts" 1836 depends on FAULT_INJECTION && BLOCK 1837 help 1838 Provide fault-injection capability on end IO handling. This 1839 will make the block layer "forget" an interrupt as configured, 1840 thus exercising the error handling. 1841 1842 Only works with drivers that use the generic timeout handling, 1843 for others it wont do anything. 1844 1845config FAIL_FUTEX 1846 bool "Fault-injection capability for futexes" 1847 select DEBUG_FS 1848 depends on FAULT_INJECTION && FUTEX 1849 help 1850 Provide fault-injection capability for futexes. 1851 1852config FAULT_INJECTION_DEBUG_FS 1853 bool "Debugfs entries for fault-injection capabilities" 1854 depends on FAULT_INJECTION && SYSFS && DEBUG_FS 1855 help 1856 Enable configuration of fault-injection capabilities via debugfs. 1857 1858config FAIL_FUNCTION 1859 bool "Fault-injection capability for functions" 1860 depends on FAULT_INJECTION_DEBUG_FS && FUNCTION_ERROR_INJECTION 1861 help 1862 Provide function-based fault-injection capability. 1863 This will allow you to override a specific function with a return 1864 with given return value. As a result, function caller will see 1865 an error value and have to handle it. This is useful to test the 1866 error handling in various subsystems. 1867 1868config FAIL_MMC_REQUEST 1869 bool "Fault-injection capability for MMC IO" 1870 depends on FAULT_INJECTION_DEBUG_FS && MMC 1871 help 1872 Provide fault-injection capability for MMC IO. 1873 This will make the mmc core return data errors. This is 1874 useful to test the error handling in the mmc block device 1875 and to test how the mmc host driver handles retries from 1876 the block device. 1877 1878config FAULT_INJECTION_STACKTRACE_FILTER 1879 bool "stacktrace filter for fault-injection capabilities" 1880 depends on FAULT_INJECTION_DEBUG_FS && STACKTRACE_SUPPORT 1881 depends on !X86_64 1882 select STACKTRACE 1883 select FRAME_POINTER if !MIPS && !PPC && !S390 && !MICROBLAZE && !ARM && !ARC && !X86 1884 help 1885 Provide stacktrace filter for fault-injection capabilities 1886 1887config ARCH_HAS_KCOV 1888 bool 1889 help 1890 An architecture should select this when it can successfully 1891 build and run with CONFIG_KCOV. This typically requires 1892 disabling instrumentation for some early boot code. 1893 1894config CC_HAS_SANCOV_TRACE_PC 1895 def_bool $(cc-option,-fsanitize-coverage=trace-pc) 1896 1897 1898config KCOV 1899 bool "Code coverage for fuzzing" 1900 depends on ARCH_HAS_KCOV 1901 depends on CC_HAS_SANCOV_TRACE_PC || GCC_PLUGINS 1902 select DEBUG_FS 1903 select GCC_PLUGIN_SANCOV if !CC_HAS_SANCOV_TRACE_PC 1904 help 1905 KCOV exposes kernel code coverage information in a form suitable 1906 for coverage-guided fuzzing (randomized testing). 1907 1908 If RANDOMIZE_BASE is enabled, PC values will not be stable across 1909 different machines and across reboots. If you need stable PC values, 1910 disable RANDOMIZE_BASE. 1911 1912 For more details, see Documentation/dev-tools/kcov.rst. 1913 1914config KCOV_ENABLE_COMPARISONS 1915 bool "Enable comparison operands collection by KCOV" 1916 depends on KCOV 1917 depends on $(cc-option,-fsanitize-coverage=trace-cmp) 1918 help 1919 KCOV also exposes operands of every comparison in the instrumented 1920 code along with operand sizes and PCs of the comparison instructions. 1921 These operands can be used by fuzzing engines to improve the quality 1922 of fuzzing coverage. 1923 1924config KCOV_INSTRUMENT_ALL 1925 bool "Instrument all code by default" 1926 depends on KCOV 1927 default y 1928 help 1929 If you are doing generic system call fuzzing (like e.g. syzkaller), 1930 then you will want to instrument the whole kernel and you should 1931 say y here. If you are doing more targeted fuzzing (like e.g. 1932 filesystem fuzzing with AFL) then you will want to enable coverage 1933 for more specific subsets of files, and should say n here. 1934 1935config KCOV_IRQ_AREA_SIZE 1936 hex "Size of interrupt coverage collection area in words" 1937 depends on KCOV 1938 default 0x40000 1939 help 1940 KCOV uses preallocated per-cpu areas to collect coverage from 1941 soft interrupts. This specifies the size of those areas in the 1942 number of unsigned long words. 1943 1944menuconfig RUNTIME_TESTING_MENU 1945 bool "Runtime Testing" 1946 def_bool y 1947 1948if RUNTIME_TESTING_MENU 1949 1950config LKDTM 1951 tristate "Linux Kernel Dump Test Tool Module" 1952 depends on DEBUG_FS 1953 help 1954 This module enables testing of the different dumping mechanisms by 1955 inducing system failures at predefined crash points. 1956 If you don't need it: say N 1957 Choose M here to compile this code as a module. The module will be 1958 called lkdtm. 1959 1960 Documentation on how to use the module can be found in 1961 Documentation/fault-injection/provoke-crashes.rst 1962 1963config TEST_LIST_SORT 1964 tristate "Linked list sorting test" 1965 depends on DEBUG_KERNEL || m 1966 help 1967 Enable this to turn on 'list_sort()' function test. This test is 1968 executed only once during system boot (so affects only boot time), 1969 or at module load time. 1970 1971 If unsure, say N. 1972 1973config TEST_MIN_HEAP 1974 tristate "Min heap test" 1975 depends on DEBUG_KERNEL || m 1976 help 1977 Enable this to turn on min heap function tests. This test is 1978 executed only once during system boot (so affects only boot time), 1979 or at module load time. 1980 1981 If unsure, say N. 1982 1983config TEST_SORT 1984 tristate "Array-based sort test" 1985 depends on DEBUG_KERNEL || m 1986 help 1987 This option enables the self-test function of 'sort()' at boot, 1988 or at module load time. 1989 1990 If unsure, say N. 1991 1992config KPROBES_SANITY_TEST 1993 bool "Kprobes sanity tests" 1994 depends on DEBUG_KERNEL 1995 depends on KPROBES 1996 help 1997 This option provides for testing basic kprobes functionality on 1998 boot. Samples of kprobe and kretprobe are inserted and 1999 verified for functionality. 2000 2001 Say N if you are unsure. 2002 2003config BACKTRACE_SELF_TEST 2004 tristate "Self test for the backtrace code" 2005 depends on DEBUG_KERNEL 2006 help 2007 This option provides a kernel module that can be used to test 2008 the kernel stack backtrace code. This option is not useful 2009 for distributions or general kernels, but only for kernel 2010 developers working on architecture code. 2011 2012 Note that if you want to also test saved backtraces, you will 2013 have to enable STACKTRACE as well. 2014 2015 Say N if you are unsure. 2016 2017config RBTREE_TEST 2018 tristate "Red-Black tree test" 2019 depends on DEBUG_KERNEL 2020 help 2021 A benchmark measuring the performance of the rbtree library. 2022 Also includes rbtree invariant checks. 2023 2024config REED_SOLOMON_TEST 2025 tristate "Reed-Solomon library test" 2026 depends on DEBUG_KERNEL || m 2027 select REED_SOLOMON 2028 select REED_SOLOMON_ENC16 2029 select REED_SOLOMON_DEC16 2030 help 2031 This option enables the self-test function of rslib at boot, 2032 or at module load time. 2033 2034 If unsure, say N. 2035 2036config INTERVAL_TREE_TEST 2037 tristate "Interval tree test" 2038 depends on DEBUG_KERNEL 2039 select INTERVAL_TREE 2040 help 2041 A benchmark measuring the performance of the interval tree library 2042 2043config PERCPU_TEST 2044 tristate "Per cpu operations test" 2045 depends on m && DEBUG_KERNEL 2046 help 2047 Enable this option to build test module which validates per-cpu 2048 operations. 2049 2050 If unsure, say N. 2051 2052config ATOMIC64_SELFTEST 2053 tristate "Perform an atomic64_t self-test" 2054 help 2055 Enable this option to test the atomic64_t functions at boot or 2056 at module load time. 2057 2058 If unsure, say N. 2059 2060config ASYNC_RAID6_TEST 2061 tristate "Self test for hardware accelerated raid6 recovery" 2062 depends on ASYNC_RAID6_RECOV 2063 select ASYNC_MEMCPY 2064 help 2065 This is a one-shot self test that permutes through the 2066 recovery of all the possible two disk failure scenarios for a 2067 N-disk array. Recovery is performed with the asynchronous 2068 raid6 recovery routines, and will optionally use an offload 2069 engine if one is available. 2070 2071 If unsure, say N. 2072 2073config TEST_HEXDUMP 2074 tristate "Test functions located in the hexdump module at runtime" 2075 2076config TEST_STRING_HELPERS 2077 tristate "Test functions located in the string_helpers module at runtime" 2078 2079config TEST_STRSCPY 2080 tristate "Test strscpy*() family of functions at runtime" 2081 2082config TEST_KSTRTOX 2083 tristate "Test kstrto*() family of functions at runtime" 2084 2085config TEST_PRINTF 2086 tristate "Test printf() family of functions at runtime" 2087 2088config TEST_BITMAP 2089 tristate "Test bitmap_*() family of functions at runtime" 2090 help 2091 Enable this option to test the bitmap functions at boot. 2092 2093 If unsure, say N. 2094 2095config TEST_UUID 2096 tristate "Test functions located in the uuid module at runtime" 2097 2098config TEST_XARRAY 2099 tristate "Test the XArray code at runtime" 2100 2101config TEST_OVERFLOW 2102 tristate "Test check_*_overflow() functions at runtime" 2103 2104config TEST_RHASHTABLE 2105 tristate "Perform selftest on resizable hash table" 2106 help 2107 Enable this option to test the rhashtable functions at boot. 2108 2109 If unsure, say N. 2110 2111config TEST_HASH 2112 tristate "Perform selftest on hash functions" 2113 help 2114 Enable this option to test the kernel's integer (<linux/hash.h>), 2115 string (<linux/stringhash.h>), and siphash (<linux/siphash.h>) 2116 hash functions on boot (or module load). 2117 2118 This is intended to help people writing architecture-specific 2119 optimized versions. If unsure, say N. 2120 2121config TEST_IDA 2122 tristate "Perform selftest on IDA functions" 2123 2124config TEST_PARMAN 2125 tristate "Perform selftest on priority array manager" 2126 depends on PARMAN 2127 help 2128 Enable this option to test priority array manager on boot 2129 (or module load). 2130 2131 If unsure, say N. 2132 2133config TEST_IRQ_TIMINGS 2134 bool "IRQ timings selftest" 2135 depends on IRQ_TIMINGS 2136 help 2137 Enable this option to test the irq timings code on boot. 2138 2139 If unsure, say N. 2140 2141config TEST_LKM 2142 tristate "Test module loading with 'hello world' module" 2143 depends on m 2144 help 2145 This builds the "test_module" module that emits "Hello, world" 2146 on printk when loaded. It is designed to be used for basic 2147 evaluation of the module loading subsystem (for example when 2148 validating module verification). It lacks any extra dependencies, 2149 and will not normally be loaded by the system unless explicitly 2150 requested by name. 2151 2152 If unsure, say N. 2153 2154config TEST_BITOPS 2155 tristate "Test module for compilation of bitops operations" 2156 depends on m 2157 help 2158 This builds the "test_bitops" module that is much like the 2159 TEST_LKM module except that it does a basic exercise of the 2160 set/clear_bit macros and get_count_order/long to make sure there are 2161 no compiler warnings from C=1 sparse checker or -Wextra 2162 compilations. It has no dependencies and doesn't run or load unless 2163 explicitly requested by name. for example: modprobe test_bitops. 2164 2165 If unsure, say N. 2166 2167config TEST_VMALLOC 2168 tristate "Test module for stress/performance analysis of vmalloc allocator" 2169 default n 2170 depends on MMU 2171 depends on m 2172 help 2173 This builds the "test_vmalloc" module that should be used for 2174 stress and performance analysis. So, any new change for vmalloc 2175 subsystem can be evaluated from performance and stability point 2176 of view. 2177 2178 If unsure, say N. 2179 2180config TEST_USER_COPY 2181 tristate "Test user/kernel boundary protections" 2182 depends on m 2183 help 2184 This builds the "test_user_copy" module that runs sanity checks 2185 on the copy_to/from_user infrastructure, making sure basic 2186 user/kernel boundary testing is working. If it fails to load, 2187 a regression has been detected in the user/kernel memory boundary 2188 protections. 2189 2190 If unsure, say N. 2191 2192config TEST_BPF 2193 tristate "Test BPF filter functionality" 2194 depends on m && NET 2195 help 2196 This builds the "test_bpf" module that runs various test vectors 2197 against the BPF interpreter or BPF JIT compiler depending on the 2198 current setting. This is in particular useful for BPF JIT compiler 2199 development, but also to run regression tests against changes in 2200 the interpreter code. It also enables test stubs for eBPF maps and 2201 verifier used by user space verifier testsuite. 2202 2203 If unsure, say N. 2204 2205config TEST_BLACKHOLE_DEV 2206 tristate "Test blackhole netdev functionality" 2207 depends on m && NET 2208 help 2209 This builds the "test_blackhole_dev" module that validates the 2210 data path through this blackhole netdev. 2211 2212 If unsure, say N. 2213 2214config FIND_BIT_BENCHMARK 2215 tristate "Test find_bit functions" 2216 help 2217 This builds the "test_find_bit" module that measure find_*_bit() 2218 functions performance. 2219 2220 If unsure, say N. 2221 2222config TEST_FIRMWARE 2223 tristate "Test firmware loading via userspace interface" 2224 depends on FW_LOADER 2225 help 2226 This builds the "test_firmware" module that creates a userspace 2227 interface for testing firmware loading. This can be used to 2228 control the triggering of firmware loading without needing an 2229 actual firmware-using device. The contents can be rechecked by 2230 userspace. 2231 2232 If unsure, say N. 2233 2234config TEST_SYSCTL 2235 tristate "sysctl test driver" 2236 depends on PROC_SYSCTL 2237 help 2238 This builds the "test_sysctl" module. This driver enables to test the 2239 proc sysctl interfaces available to drivers safely without affecting 2240 production knobs which might alter system functionality. 2241 2242 If unsure, say N. 2243 2244config BITFIELD_KUNIT 2245 tristate "KUnit test bitfield functions at runtime" 2246 depends on KUNIT 2247 help 2248 Enable this option to test the bitfield functions at boot. 2249 2250 KUnit tests run during boot and output the results to the debug log 2251 in TAP format (http://testanything.org/). Only useful for kernel devs 2252 running the KUnit test harness, and not intended for inclusion into a 2253 production build. 2254 2255 For more information on KUnit and unit tests in general please refer 2256 to the KUnit documentation in Documentation/dev-tools/kunit/. 2257 2258 If unsure, say N. 2259 2260config RESOURCE_KUNIT_TEST 2261 tristate "KUnit test for resource API" 2262 depends on KUNIT 2263 help 2264 This builds the resource API unit test. 2265 Tests the logic of API provided by resource.c and ioport.h. 2266 For more information on KUnit and unit tests in general please refer 2267 to the KUnit documentation in Documentation/dev-tools/kunit/. 2268 2269 If unsure, say N. 2270 2271config SYSCTL_KUNIT_TEST 2272 tristate "KUnit test for sysctl" if !KUNIT_ALL_TESTS 2273 depends on KUNIT 2274 default KUNIT_ALL_TESTS 2275 help 2276 This builds the proc sysctl unit test, which runs on boot. 2277 Tests the API contract and implementation correctness of sysctl. 2278 For more information on KUnit and unit tests in general please refer 2279 to the KUnit documentation in Documentation/dev-tools/kunit/. 2280 2281 If unsure, say N. 2282 2283config LIST_KUNIT_TEST 2284 tristate "KUnit Test for Kernel Linked-list structures" if !KUNIT_ALL_TESTS 2285 depends on KUNIT 2286 default KUNIT_ALL_TESTS 2287 help 2288 This builds the linked list KUnit test suite. 2289 It tests that the API and basic functionality of the list_head type 2290 and associated macros. 2291 2292 KUnit tests run during boot and output the results to the debug log 2293 in TAP format (https://testanything.org/). Only useful for kernel devs 2294 running the KUnit test harness, and not intended for inclusion into a 2295 production build. 2296 2297 For more information on KUnit and unit tests in general please refer 2298 to the KUnit documentation in Documentation/dev-tools/kunit/. 2299 2300 If unsure, say N. 2301 2302config LINEAR_RANGES_TEST 2303 tristate "KUnit test for linear_ranges" 2304 depends on KUNIT 2305 select LINEAR_RANGES 2306 help 2307 This builds the linear_ranges unit test, which runs on boot. 2308 Tests the linear_ranges logic correctness. 2309 For more information on KUnit and unit tests in general please refer 2310 to the KUnit documentation in Documentation/dev-tools/kunit/. 2311 2312 If unsure, say N. 2313 2314config CMDLINE_KUNIT_TEST 2315 tristate "KUnit test for cmdline API" 2316 depends on KUNIT 2317 help 2318 This builds the cmdline API unit test. 2319 Tests the logic of API provided by cmdline.c. 2320 For more information on KUnit and unit tests in general please refer 2321 to the KUnit documentation in Documentation/dev-tools/kunit/. 2322 2323 If unsure, say N. 2324 2325config BITS_TEST 2326 tristate "KUnit test for bits.h" 2327 depends on KUNIT 2328 help 2329 This builds the bits unit test. 2330 Tests the logic of macros defined in bits.h. 2331 For more information on KUnit and unit tests in general please refer 2332 to the KUnit documentation in Documentation/dev-tools/kunit/. 2333 2334 If unsure, say N. 2335 2336config TEST_UDELAY 2337 tristate "udelay test driver" 2338 help 2339 This builds the "udelay_test" module that helps to make sure 2340 that udelay() is working properly. 2341 2342 If unsure, say N. 2343 2344config TEST_STATIC_KEYS 2345 tristate "Test static keys" 2346 depends on m 2347 help 2348 Test the static key interfaces. 2349 2350 If unsure, say N. 2351 2352config TEST_KMOD 2353 tristate "kmod stress tester" 2354 depends on m 2355 depends on NETDEVICES && NET_CORE && INET # for TUN 2356 depends on BLOCK 2357 select TEST_LKM 2358 select XFS_FS 2359 select TUN 2360 select BTRFS_FS 2361 help 2362 Test the kernel's module loading mechanism: kmod. kmod implements 2363 support to load modules using the Linux kernel's usermode helper. 2364 This test provides a series of tests against kmod. 2365 2366 Although technically you can either build test_kmod as a module or 2367 into the kernel we disallow building it into the kernel since 2368 it stress tests request_module() and this will very likely cause 2369 some issues by taking over precious threads available from other 2370 module load requests, ultimately this could be fatal. 2371 2372 To run tests run: 2373 2374 tools/testing/selftests/kmod/kmod.sh --help 2375 2376 If unsure, say N. 2377 2378config TEST_DEBUG_VIRTUAL 2379 tristate "Test CONFIG_DEBUG_VIRTUAL feature" 2380 depends on DEBUG_VIRTUAL 2381 help 2382 Test the kernel's ability to detect incorrect calls to 2383 virt_to_phys() done against the non-linear part of the 2384 kernel's virtual address map. 2385 2386 If unsure, say N. 2387 2388config TEST_MEMCAT_P 2389 tristate "Test memcat_p() helper function" 2390 help 2391 Test the memcat_p() helper for correctly merging two 2392 pointer arrays together. 2393 2394 If unsure, say N. 2395 2396config TEST_LIVEPATCH 2397 tristate "Test livepatching" 2398 default n 2399 depends on DYNAMIC_DEBUG 2400 depends on LIVEPATCH 2401 depends on m 2402 help 2403 Test kernel livepatching features for correctness. The tests will 2404 load test modules that will be livepatched in various scenarios. 2405 2406 To run all the livepatching tests: 2407 2408 make -C tools/testing/selftests TARGETS=livepatch run_tests 2409 2410 Alternatively, individual tests may be invoked: 2411 2412 tools/testing/selftests/livepatch/test-callbacks.sh 2413 tools/testing/selftests/livepatch/test-livepatch.sh 2414 tools/testing/selftests/livepatch/test-shadow-vars.sh 2415 2416 If unsure, say N. 2417 2418config TEST_OBJAGG 2419 tristate "Perform selftest on object aggreration manager" 2420 default n 2421 depends on OBJAGG 2422 help 2423 Enable this option to test object aggregation manager on boot 2424 (or module load). 2425 2426 2427config TEST_STACKINIT 2428 tristate "Test level of stack variable initialization" 2429 help 2430 Test if the kernel is zero-initializing stack variables and 2431 padding. Coverage is controlled by compiler flags, 2432 CONFIG_GCC_PLUGIN_STRUCTLEAK, CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF, 2433 or CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL. 2434 2435 If unsure, say N. 2436 2437config TEST_MEMINIT 2438 tristate "Test heap/page initialization" 2439 help 2440 Test if the kernel is zero-initializing heap and page allocations. 2441 This can be useful to test init_on_alloc and init_on_free features. 2442 2443 If unsure, say N. 2444 2445config TEST_HMM 2446 tristate "Test HMM (Heterogeneous Memory Management)" 2447 depends on TRANSPARENT_HUGEPAGE 2448 depends on DEVICE_PRIVATE 2449 select HMM_MIRROR 2450 select MMU_NOTIFIER 2451 help 2452 This is a pseudo device driver solely for testing HMM. 2453 Say M here if you want to build the HMM test module. 2454 Doing so will allow you to run tools/testing/selftest/vm/hmm-tests. 2455 2456 If unsure, say N. 2457 2458config TEST_FREE_PAGES 2459 tristate "Test freeing pages" 2460 help 2461 Test that a memory leak does not occur due to a race between 2462 freeing a block of pages and a speculative page reference. 2463 Loading this module is safe if your kernel has the bug fixed. 2464 If the bug is not fixed, it will leak gigabytes of memory and 2465 probably OOM your system. 2466 2467config TEST_FPU 2468 tristate "Test floating point operations in kernel space" 2469 depends on X86 && !KCOV_INSTRUMENT_ALL 2470 help 2471 Enable this option to add /sys/kernel/debug/selftest_helpers/test_fpu 2472 which will trigger a sequence of floating point operations. This is used 2473 for self-testing floating point control register setting in 2474 kernel_fpu_begin(). 2475 2476 If unsure, say N. 2477 2478endif # RUNTIME_TESTING_MENU 2479 2480config MEMTEST 2481 bool "Memtest" 2482 help 2483 This option adds a kernel parameter 'memtest', which allows memtest 2484 to be set. 2485 memtest=0, mean disabled; -- default 2486 memtest=1, mean do 1 test pattern; 2487 ... 2488 memtest=17, mean do 17 test patterns. 2489 If you are unsure how to answer this question, answer N. 2490 2491 2492 2493config HYPERV_TESTING 2494 bool "Microsoft Hyper-V driver testing" 2495 default n 2496 depends on HYPERV && DEBUG_FS 2497 help 2498 Select this option to enable Hyper-V vmbus testing. 2499 2500endmenu # "Kernel Testing and Coverage" 2501 2502source "Documentation/Kconfig" 2503 2504endmenu # Kernel hacking 2505