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 259choice 260 prompt "DWARF version" 261 help 262 Which version of DWARF debug info to emit. 263 264config DEBUG_INFO_DWARF_TOOLCHAIN_DEFAULT 265 bool "Rely on the toolchain's implicit default DWARF version" 266 help 267 The implicit default version of DWARF debug info produced by a 268 toolchain changes over time. 269 270 This can break consumers of the debug info that haven't upgraded to 271 support newer revisions, and prevent testing newer versions, but 272 those should be less common scenarios. 273 274 If unsure, say Y. 275 276config DEBUG_INFO_DWARF4 277 bool "Generate DWARF Version 4 debuginfo" 278 help 279 Generate DWARF v4 debug info. This requires gcc 4.5+ and gdb 7.0+. 280 281 If you have consumers of DWARF debug info that are not ready for 282 newer revisions of DWARF, you may wish to choose this or have your 283 config select this. 284 285config DEBUG_INFO_DWARF5 286 bool "Generate DWARF Version 5 debuginfo" 287 depends on GCC_VERSION >= 50000 || (CC_IS_CLANG && (AS_IS_LLVM || (AS_IS_GNU && AS_VERSION >= 23502))) 288 depends on !DEBUG_INFO_BTF 289 help 290 Generate DWARF v5 debug info. Requires binutils 2.35.2, gcc 5.0+ (gcc 291 5.0+ accepts the -gdwarf-5 flag but only had partial support for some 292 draft features until 7.0), and gdb 8.0+. 293 294 Changes to the structure of debug info in Version 5 allow for around 295 15-18% savings in resulting image and debug info section sizes as 296 compared to DWARF Version 4. DWARF Version 5 standardizes previous 297 extensions such as accelerators for symbol indexing and the format 298 for fission (.dwo/.dwp) files. Users may not want to select this 299 config if they rely on tooling that has not yet been updated to 300 support DWARF Version 5. 301 302endchoice # "DWARF version" 303 304config DEBUG_INFO_BTF 305 bool "Generate BTF typeinfo" 306 depends on !DEBUG_INFO_SPLIT && !DEBUG_INFO_REDUCED 307 depends on !GCC_PLUGIN_RANDSTRUCT || COMPILE_TEST 308 help 309 Generate deduplicated BTF type information from DWARF debug info. 310 Turning this on expects presence of pahole tool, which will convert 311 DWARF type info into equivalent deduplicated BTF type info. 312 313config PAHOLE_HAS_SPLIT_BTF 314 def_bool $(success, test `$(PAHOLE) --version | sed -E 's/v([0-9]+)\.([0-9]+)/\1\2/'` -ge "119") 315 316config DEBUG_INFO_BTF_MODULES 317 def_bool y 318 depends on DEBUG_INFO_BTF && MODULES && PAHOLE_HAS_SPLIT_BTF 319 help 320 Generate compact split BTF type information for kernel modules. 321 322config GDB_SCRIPTS 323 bool "Provide GDB scripts for kernel debugging" 324 help 325 This creates the required links to GDB helper scripts in the 326 build directory. If you load vmlinux into gdb, the helper 327 scripts will be automatically imported by gdb as well, and 328 additional functions are available to analyze a Linux kernel 329 instance. See Documentation/dev-tools/gdb-kernel-debugging.rst 330 for further details. 331 332endif # DEBUG_INFO 333 334config FRAME_WARN 335 int "Warn for stack frames larger than" 336 range 0 8192 337 default 2048 if GCC_PLUGIN_LATENT_ENTROPY 338 default 1280 if (!64BIT && PARISC) 339 default 1024 if (!64BIT && !PARISC) 340 default 2048 if 64BIT 341 help 342 Tell gcc to warn at build time for stack frames larger than this. 343 Setting this too low will cause a lot of warnings. 344 Setting it to 0 disables the warning. 345 346config STRIP_ASM_SYMS 347 bool "Strip assembler-generated symbols during link" 348 default n 349 help 350 Strip internal assembler-generated symbols during a link (symbols 351 that look like '.Lxxx') so they don't pollute the output of 352 get_wchan() and suchlike. 353 354config READABLE_ASM 355 bool "Generate readable assembler code" 356 depends on DEBUG_KERNEL 357 help 358 Disable some compiler optimizations that tend to generate human unreadable 359 assembler output. This may make the kernel slightly slower, but it helps 360 to keep kernel developers who have to stare a lot at assembler listings 361 sane. 362 363config HEADERS_INSTALL 364 bool "Install uapi headers to usr/include" 365 depends on !UML 366 help 367 This option will install uapi headers (headers exported to user-space) 368 into the usr/include directory for use during the kernel build. 369 This is unneeded for building the kernel itself, but needed for some 370 user-space program samples. It is also needed by some features such 371 as uapi header sanity checks. 372 373config DEBUG_SECTION_MISMATCH 374 bool "Enable full Section mismatch analysis" 375 help 376 The section mismatch analysis checks if there are illegal 377 references from one section to another section. 378 During linktime or runtime, some sections are dropped; 379 any use of code/data previously in these sections would 380 most likely result in an oops. 381 In the code, functions and variables are annotated with 382 __init,, etc. (see the full list in include/linux/init.h), 383 which results in the code/data being placed in specific sections. 384 The section mismatch analysis is always performed after a full 385 kernel build, and enabling this option causes the following 386 additional step to occur: 387 - Add the option -fno-inline-functions-called-once to gcc commands. 388 When inlining a function annotated with __init in a non-init 389 function, we would lose the section information and thus 390 the analysis would not catch the illegal reference. 391 This option tells gcc to inline less (but it does result in 392 a larger kernel). 393 394config SECTION_MISMATCH_WARN_ONLY 395 bool "Make section mismatch errors non-fatal" 396 default y 397 help 398 If you say N here, the build process will fail if there are any 399 section mismatch, instead of just throwing warnings. 400 401 If unsure, say Y. 402 403config DEBUG_FORCE_FUNCTION_ALIGN_32B 404 bool "Force all function address 32B aligned" if EXPERT 405 help 406 There are cases that a commit from one domain changes the function 407 address alignment of other domains, and cause magic performance 408 bump (regression or improvement). Enable this option will help to 409 verify if the bump is caused by function alignment changes, while 410 it will slightly increase the kernel size and affect icache usage. 411 412 It is mainly for debug and performance tuning use. 413 414# 415# Select this config option from the architecture Kconfig, if it 416# is preferred to always offer frame pointers as a config 417# option on the architecture (regardless of KERNEL_DEBUG): 418# 419config ARCH_WANT_FRAME_POINTERS 420 bool 421 422config FRAME_POINTER 423 bool "Compile the kernel with frame pointers" 424 depends on DEBUG_KERNEL && (M68K || UML || SUPERH) || ARCH_WANT_FRAME_POINTERS 425 default y if (DEBUG_INFO && UML) || ARCH_WANT_FRAME_POINTERS 426 help 427 If you say Y here the resulting kernel image will be slightly 428 larger and slower, but it gives very useful debugging information 429 in case of kernel bugs. (precise oopses/stacktraces/warnings) 430 431config STACK_VALIDATION 432 bool "Compile-time stack metadata validation" 433 depends on HAVE_STACK_VALIDATION 434 default n 435 help 436 Add compile-time checks to validate stack metadata, including frame 437 pointers (if CONFIG_FRAME_POINTER is enabled). This helps ensure 438 that runtime stack traces are more reliable. 439 440 This is also a prerequisite for generation of ORC unwind data, which 441 is needed for CONFIG_UNWINDER_ORC. 442 443 For more information, see 444 tools/objtool/Documentation/stack-validation.txt. 445 446config VMLINUX_VALIDATION 447 bool 448 depends on STACK_VALIDATION && DEBUG_ENTRY && !PARAVIRT 449 default y 450 451config VMLINUX_MAP 452 bool "Generate vmlinux.map file when linking" 453 depends on EXPERT 454 help 455 Selecting this option will pass "-Map=vmlinux.map" to ld 456 when linking vmlinux. That file can be useful for verifying 457 and debugging magic section games, and for seeing which 458 pieces of code get eliminated with 459 CONFIG_LD_DEAD_CODE_DATA_ELIMINATION. 460 461config DEBUG_FORCE_WEAK_PER_CPU 462 bool "Force weak per-cpu definitions" 463 depends on DEBUG_KERNEL 464 help 465 s390 and alpha require percpu variables in modules to be 466 defined weak to work around addressing range issue which 467 puts the following two restrictions on percpu variable 468 definitions. 469 470 1. percpu symbols must be unique whether static or not 471 2. percpu variables can't be defined inside a function 472 473 To ensure that generic code follows the above rules, this 474 option forces all percpu variables to be defined as weak. 475 476endmenu # "Compiler options" 477 478menu "Generic Kernel Debugging Instruments" 479 480config MAGIC_SYSRQ 481 bool "Magic SysRq key" 482 depends on !UML 483 help 484 If you say Y here, you will have some control over the system even 485 if the system crashes for example during kernel debugging (e.g., you 486 will be able to flush the buffer cache to disk, reboot the system 487 immediately or dump some status information). This is accomplished 488 by pressing various keys while holding SysRq (Alt+PrintScreen). It 489 also works on a serial console (on PC hardware at least), if you 490 send a BREAK and then within 5 seconds a command keypress. The 491 keys are documented in <file:Documentation/admin-guide/sysrq.rst>. 492 Don't say Y unless you really know what this hack does. 493 494config MAGIC_SYSRQ_DEFAULT_ENABLE 495 hex "Enable magic SysRq key functions by default" 496 depends on MAGIC_SYSRQ 497 default 0x1 498 help 499 Specifies which SysRq key functions are enabled by default. 500 This may be set to 1 or 0 to enable or disable them all, or 501 to a bitmask as described in Documentation/admin-guide/sysrq.rst. 502 503config MAGIC_SYSRQ_SERIAL 504 bool "Enable magic SysRq key over serial" 505 depends on MAGIC_SYSRQ 506 default y 507 help 508 Many embedded boards have a disconnected TTL level serial which can 509 generate some garbage that can lead to spurious false sysrq detects. 510 This option allows you to decide whether you want to enable the 511 magic SysRq key. 512 513config MAGIC_SYSRQ_SERIAL_SEQUENCE 514 string "Char sequence that enables magic SysRq over serial" 515 depends on MAGIC_SYSRQ_SERIAL 516 default "" 517 help 518 Specifies a sequence of characters that can follow BREAK to enable 519 SysRq on a serial console. 520 521 If unsure, leave an empty string and the option will not be enabled. 522 523config DEBUG_FS 524 bool "Debug Filesystem" 525 help 526 debugfs is a virtual file system that kernel developers use to put 527 debugging files into. Enable this option to be able to read and 528 write to these files. 529 530 For detailed documentation on the debugfs API, see 531 Documentation/filesystems/. 532 533 If unsure, say N. 534 535choice 536 prompt "Debugfs default access" 537 depends on DEBUG_FS 538 default DEBUG_FS_ALLOW_ALL 539 help 540 This selects the default access restrictions for debugfs. 541 It can be overridden with kernel command line option 542 debugfs=[on,no-mount,off]. The restrictions apply for API access 543 and filesystem registration. 544 545config DEBUG_FS_ALLOW_ALL 546 bool "Access normal" 547 help 548 No restrictions apply. Both API and filesystem registration 549 is on. This is the normal default operation. 550 551config DEBUG_FS_DISALLOW_MOUNT 552 bool "Do not register debugfs as filesystem" 553 help 554 The API is open but filesystem is not loaded. Clients can still do 555 their work and read with debug tools that do not need 556 debugfs filesystem. 557 558config DEBUG_FS_ALLOW_NONE 559 bool "No access" 560 help 561 Access is off. Clients get -PERM when trying to create nodes in 562 debugfs tree and debugfs is not registered as a filesystem. 563 Client can then back-off or continue without debugfs access. 564 565endchoice 566 567source "lib/Kconfig.kgdb" 568source "lib/Kconfig.ubsan" 569source "lib/Kconfig.kcsan" 570 571endmenu 572 573config DEBUG_KERNEL 574 bool "Kernel debugging" 575 help 576 Say Y here if you are developing drivers or trying to debug and 577 identify kernel problems. 578 579config DEBUG_MISC 580 bool "Miscellaneous debug code" 581 default DEBUG_KERNEL 582 depends on DEBUG_KERNEL 583 help 584 Say Y here if you need to enable miscellaneous debug code that should 585 be under a more specific debug option but isn't. 586 587 588menu "Memory Debugging" 589 590source "mm/Kconfig.debug" 591 592config DEBUG_OBJECTS 593 bool "Debug object operations" 594 depends on DEBUG_KERNEL 595 help 596 If you say Y here, additional code will be inserted into the 597 kernel to track the life time of various objects and validate 598 the operations on those objects. 599 600config DEBUG_OBJECTS_SELFTEST 601 bool "Debug objects selftest" 602 depends on DEBUG_OBJECTS 603 help 604 This enables the selftest of the object debug code. 605 606config DEBUG_OBJECTS_FREE 607 bool "Debug objects in freed memory" 608 depends on DEBUG_OBJECTS 609 help 610 This enables checks whether a k/v free operation frees an area 611 which contains an object which has not been deactivated 612 properly. This can make kmalloc/kfree-intensive workloads 613 much slower. 614 615config DEBUG_OBJECTS_TIMERS 616 bool "Debug timer objects" 617 depends on DEBUG_OBJECTS 618 help 619 If you say Y here, additional code will be inserted into the 620 timer routines to track the life time of timer objects and 621 validate the timer operations. 622 623config DEBUG_OBJECTS_WORK 624 bool "Debug work objects" 625 depends on DEBUG_OBJECTS 626 help 627 If you say Y here, additional code will be inserted into the 628 work queue routines to track the life time of work objects and 629 validate the work operations. 630 631config DEBUG_OBJECTS_RCU_HEAD 632 bool "Debug RCU callbacks objects" 633 depends on DEBUG_OBJECTS 634 help 635 Enable this to turn on debugging of RCU list heads (call_rcu() usage). 636 637config DEBUG_OBJECTS_PERCPU_COUNTER 638 bool "Debug percpu counter objects" 639 depends on DEBUG_OBJECTS 640 help 641 If you say Y here, additional code will be inserted into the 642 percpu counter routines to track the life time of percpu counter 643 objects and validate the percpu counter operations. 644 645config DEBUG_OBJECTS_ENABLE_DEFAULT 646 int "debug_objects bootup default value (0-1)" 647 range 0 1 648 default "1" 649 depends on DEBUG_OBJECTS 650 help 651 Debug objects boot parameter default value 652 653config DEBUG_SLAB 654 bool "Debug slab memory allocations" 655 depends on DEBUG_KERNEL && SLAB 656 help 657 Say Y here to have the kernel do limited verification on memory 658 allocation as well as poisoning memory on free to catch use of freed 659 memory. This can make kmalloc/kfree-intensive workloads much slower. 660 661config SLUB_DEBUG_ON 662 bool "SLUB debugging on by default" 663 depends on SLUB && SLUB_DEBUG 664 default n 665 help 666 Boot with debugging on by default. SLUB boots by default with 667 the runtime debug capabilities switched off. Enabling this is 668 equivalent to specifying the "slub_debug" parameter on boot. 669 There is no support for more fine grained debug control like 670 possible with slub_debug=xxx. SLUB debugging may be switched 671 off in a kernel built with CONFIG_SLUB_DEBUG_ON by specifying 672 "slub_debug=-". 673 674config SLUB_STATS 675 default n 676 bool "Enable SLUB performance statistics" 677 depends on SLUB && SYSFS 678 help 679 SLUB statistics are useful to debug SLUBs allocation behavior in 680 order find ways to optimize the allocator. This should never be 681 enabled for production use since keeping statistics slows down 682 the allocator by a few percentage points. The slabinfo command 683 supports the determination of the most active slabs to figure 684 out which slabs are relevant to a particular load. 685 Try running: slabinfo -DA 686 687config HAVE_DEBUG_KMEMLEAK 688 bool 689 690config DEBUG_KMEMLEAK 691 bool "Kernel memory leak detector" 692 depends on DEBUG_KERNEL && HAVE_DEBUG_KMEMLEAK 693 select DEBUG_FS 694 select STACKTRACE if STACKTRACE_SUPPORT 695 select KALLSYMS 696 select CRC32 697 help 698 Say Y here if you want to enable the memory leak 699 detector. The memory allocation/freeing is traced in a way 700 similar to the Boehm's conservative garbage collector, the 701 difference being that the orphan objects are not freed but 702 only shown in /sys/kernel/debug/kmemleak. Enabling this 703 feature will introduce an overhead to memory 704 allocations. See Documentation/dev-tools/kmemleak.rst for more 705 details. 706 707 Enabling DEBUG_SLAB or SLUB_DEBUG may increase the chances 708 of finding leaks due to the slab objects poisoning. 709 710 In order to access the kmemleak file, debugfs needs to be 711 mounted (usually at /sys/kernel/debug). 712 713config DEBUG_KMEMLEAK_MEM_POOL_SIZE 714 int "Kmemleak memory pool size" 715 depends on DEBUG_KMEMLEAK 716 range 200 1000000 717 default 16000 718 help 719 Kmemleak must track all the memory allocations to avoid 720 reporting false positives. Since memory may be allocated or 721 freed before kmemleak is fully initialised, use a static pool 722 of metadata objects to track such callbacks. After kmemleak is 723 fully initialised, this memory pool acts as an emergency one 724 if slab allocations fail. 725 726config DEBUG_KMEMLEAK_TEST 727 tristate "Simple test for the kernel memory leak detector" 728 depends on DEBUG_KMEMLEAK && m 729 help 730 This option enables a module that explicitly leaks memory. 731 732 If unsure, say N. 733 734config DEBUG_KMEMLEAK_DEFAULT_OFF 735 bool "Default kmemleak to off" 736 depends on DEBUG_KMEMLEAK 737 help 738 Say Y here to disable kmemleak by default. It can then be enabled 739 on the command line via kmemleak=on. 740 741config DEBUG_KMEMLEAK_AUTO_SCAN 742 bool "Enable kmemleak auto scan thread on boot up" 743 default y 744 depends on DEBUG_KMEMLEAK 745 help 746 Depending on the cpu, kmemleak scan may be cpu intensive and can 747 stall user tasks at times. This option enables/disables automatic 748 kmemleak scan at boot up. 749 750 Say N here to disable kmemleak auto scan thread to stop automatic 751 scanning. Disabling this option disables automatic reporting of 752 memory leaks. 753 754 If unsure, say Y. 755 756config DEBUG_STACK_USAGE 757 bool "Stack utilization instrumentation" 758 depends on DEBUG_KERNEL && !IA64 759 help 760 Enables the display of the minimum amount of free stack which each 761 task has ever had available in the sysrq-T and sysrq-P debug output. 762 763 This option will slow down process creation somewhat. 764 765config SCHED_STACK_END_CHECK 766 bool "Detect stack corruption on calls to schedule()" 767 depends on DEBUG_KERNEL 768 default n 769 help 770 This option checks for a stack overrun on calls to schedule(). 771 If the stack end location is found to be over written always panic as 772 the content of the corrupted region can no longer be trusted. 773 This is to ensure no erroneous behaviour occurs which could result in 774 data corruption or a sporadic crash at a later stage once the region 775 is examined. The runtime overhead introduced is minimal. 776 777config ARCH_HAS_DEBUG_VM_PGTABLE 778 bool 779 help 780 An architecture should select this when it can successfully 781 build and run DEBUG_VM_PGTABLE. 782 783config DEBUG_VM 784 bool "Debug VM" 785 depends on DEBUG_KERNEL 786 help 787 Enable this to turn on extended checks in the virtual-memory system 788 that may impact performance. 789 790 If unsure, say N. 791 792config DEBUG_VM_VMACACHE 793 bool "Debug VMA caching" 794 depends on DEBUG_VM 795 help 796 Enable this to turn on VMA caching debug information. Doing so 797 can cause significant overhead, so only enable it in non-production 798 environments. 799 800 If unsure, say N. 801 802config DEBUG_VM_RB 803 bool "Debug VM red-black trees" 804 depends on DEBUG_VM 805 help 806 Enable VM red-black tree debugging information and extra validations. 807 808 If unsure, say N. 809 810config DEBUG_VM_PGFLAGS 811 bool "Debug page-flags operations" 812 depends on DEBUG_VM 813 help 814 Enables extra validation on page flags operations. 815 816 If unsure, say N. 817 818config DEBUG_VM_PGTABLE 819 bool "Debug arch page table for semantics compliance" 820 depends on MMU 821 depends on ARCH_HAS_DEBUG_VM_PGTABLE 822 default y if DEBUG_VM 823 help 824 This option provides a debug method which can be used to test 825 architecture page table helper functions on various platforms in 826 verifying if they comply with expected generic MM semantics. This 827 will help architecture code in making sure that any changes or 828 new additions of these helpers still conform to expected 829 semantics of the generic MM. Platforms will have to opt in for 830 this through ARCH_HAS_DEBUG_VM_PGTABLE. 831 832 If unsure, say N. 833 834config ARCH_HAS_DEBUG_VIRTUAL 835 bool 836 837config DEBUG_VIRTUAL 838 bool "Debug VM translations" 839 depends on DEBUG_KERNEL && ARCH_HAS_DEBUG_VIRTUAL 840 help 841 Enable some costly sanity checks in virtual to page code. This can 842 catch mistakes with virt_to_page() and friends. 843 844 If unsure, say N. 845 846config DEBUG_NOMMU_REGIONS 847 bool "Debug the global anon/private NOMMU mapping region tree" 848 depends on DEBUG_KERNEL && !MMU 849 help 850 This option causes the global tree of anonymous and private mapping 851 regions to be regularly checked for invalid topology. 852 853config DEBUG_MEMORY_INIT 854 bool "Debug memory initialisation" if EXPERT 855 default !EXPERT 856 help 857 Enable this for additional checks during memory initialisation. 858 The sanity checks verify aspects of the VM such as the memory model 859 and other information provided by the architecture. Verbose 860 information will be printed at KERN_DEBUG loglevel depending 861 on the mminit_loglevel= command-line option. 862 863 If unsure, say Y 864 865config MEMORY_NOTIFIER_ERROR_INJECT 866 tristate "Memory hotplug notifier error injection module" 867 depends on MEMORY_HOTPLUG_SPARSE && NOTIFIER_ERROR_INJECTION 868 help 869 This option provides the ability to inject artificial errors to 870 memory hotplug notifier chain callbacks. It is controlled through 871 debugfs interface under /sys/kernel/debug/notifier-error-inject/memory 872 873 If the notifier call chain should be failed with some events 874 notified, write the error code to "actions/<notifier event>/error". 875 876 Example: Inject memory hotplug offline error (-12 == -ENOMEM) 877 878 # cd /sys/kernel/debug/notifier-error-inject/memory 879 # echo -12 > actions/MEM_GOING_OFFLINE/error 880 # echo offline > /sys/devices/system/memory/memoryXXX/state 881 bash: echo: write error: Cannot allocate memory 882 883 To compile this code as a module, choose M here: the module will 884 be called memory-notifier-error-inject. 885 886 If unsure, say N. 887 888config DEBUG_PER_CPU_MAPS 889 bool "Debug access to per_cpu maps" 890 depends on DEBUG_KERNEL 891 depends on SMP 892 help 893 Say Y to verify that the per_cpu map being accessed has 894 been set up. This adds a fair amount of code to kernel memory 895 and decreases performance. 896 897 Say N if unsure. 898 899config DEBUG_KMAP_LOCAL 900 bool "Debug kmap_local temporary mappings" 901 depends on DEBUG_KERNEL && KMAP_LOCAL 902 help 903 This option enables additional error checking for the kmap_local 904 infrastructure. Disable for production use. 905 906config ARCH_SUPPORTS_KMAP_LOCAL_FORCE_MAP 907 bool 908 909config DEBUG_KMAP_LOCAL_FORCE_MAP 910 bool "Enforce kmap_local temporary mappings" 911 depends on DEBUG_KERNEL && ARCH_SUPPORTS_KMAP_LOCAL_FORCE_MAP 912 select KMAP_LOCAL 913 select DEBUG_KMAP_LOCAL 914 help 915 This option enforces temporary mappings through the kmap_local 916 mechanism for non-highmem pages and on non-highmem systems. 917 Disable this for production systems! 918 919config DEBUG_HIGHMEM 920 bool "Highmem debugging" 921 depends on DEBUG_KERNEL && HIGHMEM 922 select DEBUG_KMAP_LOCAL_FORCE_MAP if ARCH_SUPPORTS_KMAP_LOCAL_FORCE_MAP 923 select DEBUG_KMAP_LOCAL 924 help 925 This option enables additional error checking for high memory 926 systems. Disable for production systems. 927 928config HAVE_DEBUG_STACKOVERFLOW 929 bool 930 931config DEBUG_STACKOVERFLOW 932 bool "Check for stack overflows" 933 depends on DEBUG_KERNEL && HAVE_DEBUG_STACKOVERFLOW 934 help 935 Say Y here if you want to check for overflows of kernel, IRQ 936 and exception stacks (if your architecture uses them). This 937 option will show detailed messages if free stack space drops 938 below a certain limit. 939 940 These kinds of bugs usually occur when call-chains in the 941 kernel get too deep, especially when interrupts are 942 involved. 943 944 Use this in cases where you see apparently random memory 945 corruption, especially if it appears in 'struct thread_info' 946 947 If in doubt, say "N". 948 949source "lib/Kconfig.kasan" 950source "lib/Kconfig.kfence" 951 952endmenu # "Memory Debugging" 953 954config DEBUG_SHIRQ 955 bool "Debug shared IRQ handlers" 956 depends on DEBUG_KERNEL 957 help 958 Enable this to generate a spurious interrupt just before a shared 959 interrupt handler is deregistered (generating one when registering 960 is currently disabled). Drivers need to handle this correctly. Some 961 don't and need to be caught. 962 963menu "Debug Oops, Lockups and Hangs" 964 965config PANIC_ON_OOPS 966 bool "Panic on Oops" 967 help 968 Say Y here to enable the kernel to panic when it oopses. This 969 has the same effect as setting oops=panic on the kernel command 970 line. 971 972 This feature is useful to ensure that the kernel does not do 973 anything erroneous after an oops which could result in data 974 corruption or other issues. 975 976 Say N if unsure. 977 978config PANIC_ON_OOPS_VALUE 979 int 980 range 0 1 981 default 0 if !PANIC_ON_OOPS 982 default 1 if PANIC_ON_OOPS 983 984config PANIC_TIMEOUT 985 int "panic timeout" 986 default 0 987 help 988 Set the timeout value (in seconds) until a reboot occurs when 989 the kernel panics. If n = 0, then we wait forever. A timeout 990 value n > 0 will wait n seconds before rebooting, while a timeout 991 value n < 0 will reboot immediately. 992 993config LOCKUP_DETECTOR 994 bool 995 996config SOFTLOCKUP_DETECTOR 997 bool "Detect Soft Lockups" 998 depends on DEBUG_KERNEL && !S390 999 select LOCKUP_DETECTOR 1000 help 1001 Say Y here to enable the kernel to act as a watchdog to detect 1002 soft lockups. 1003 1004 Softlockups are bugs that cause the kernel to loop in kernel 1005 mode for more than 20 seconds, without giving other tasks a 1006 chance to run. The current stack trace is displayed upon 1007 detection and the system will stay locked up. 1008 1009config BOOTPARAM_SOFTLOCKUP_PANIC 1010 bool "Panic (Reboot) On Soft Lockups" 1011 depends on SOFTLOCKUP_DETECTOR 1012 help 1013 Say Y here to enable the kernel to panic on "soft lockups", 1014 which are bugs that cause the kernel to loop in kernel 1015 mode for more than 20 seconds (configurable using the watchdog_thresh 1016 sysctl), without giving other tasks a chance to run. 1017 1018 The panic can be used in combination with panic_timeout, 1019 to cause the system to reboot automatically after a 1020 lockup has been detected. This feature is useful for 1021 high-availability systems that have uptime guarantees and 1022 where a lockup must be resolved ASAP. 1023 1024 Say N if unsure. 1025 1026config BOOTPARAM_SOFTLOCKUP_PANIC_VALUE 1027 int 1028 depends on SOFTLOCKUP_DETECTOR 1029 range 0 1 1030 default 0 if !BOOTPARAM_SOFTLOCKUP_PANIC 1031 default 1 if BOOTPARAM_SOFTLOCKUP_PANIC 1032 1033config HARDLOCKUP_DETECTOR_PERF 1034 bool 1035 select SOFTLOCKUP_DETECTOR 1036 1037# 1038# Enables a timestamp based low pass filter to compensate for perf based 1039# hard lockup detection which runs too fast due to turbo modes. 1040# 1041config HARDLOCKUP_CHECK_TIMESTAMP 1042 bool 1043 1044# 1045# arch/ can define HAVE_HARDLOCKUP_DETECTOR_ARCH to provide their own hard 1046# lockup detector rather than the perf based detector. 1047# 1048config HARDLOCKUP_DETECTOR 1049 bool "Detect Hard Lockups" 1050 depends on DEBUG_KERNEL && !S390 1051 depends on HAVE_HARDLOCKUP_DETECTOR_PERF || HAVE_HARDLOCKUP_DETECTOR_ARCH 1052 select LOCKUP_DETECTOR 1053 select HARDLOCKUP_DETECTOR_PERF if HAVE_HARDLOCKUP_DETECTOR_PERF 1054 select HARDLOCKUP_DETECTOR_ARCH if HAVE_HARDLOCKUP_DETECTOR_ARCH 1055 help 1056 Say Y here to enable the kernel to act as a watchdog to detect 1057 hard lockups. 1058 1059 Hardlockups are bugs that cause the CPU to loop in kernel mode 1060 for more than 10 seconds, without letting other interrupts have a 1061 chance to run. The current stack trace is displayed upon detection 1062 and the system will stay locked up. 1063 1064config BOOTPARAM_HARDLOCKUP_PANIC 1065 bool "Panic (Reboot) On Hard Lockups" 1066 depends on HARDLOCKUP_DETECTOR 1067 help 1068 Say Y here to enable the kernel to panic on "hard lockups", 1069 which are bugs that cause the kernel to loop in kernel 1070 mode with interrupts disabled for more than 10 seconds (configurable 1071 using the watchdog_thresh sysctl). 1072 1073 Say N if unsure. 1074 1075config BOOTPARAM_HARDLOCKUP_PANIC_VALUE 1076 int 1077 depends on HARDLOCKUP_DETECTOR 1078 range 0 1 1079 default 0 if !BOOTPARAM_HARDLOCKUP_PANIC 1080 default 1 if BOOTPARAM_HARDLOCKUP_PANIC 1081 1082config DETECT_HUNG_TASK 1083 bool "Detect Hung Tasks" 1084 depends on DEBUG_KERNEL 1085 default SOFTLOCKUP_DETECTOR 1086 help 1087 Say Y here to enable the kernel to detect "hung tasks", 1088 which are bugs that cause the task to be stuck in 1089 uninterruptible "D" state indefinitely. 1090 1091 When a hung task is detected, the kernel will print the 1092 current stack trace (which you should report), but the 1093 task will stay in uninterruptible state. If lockdep is 1094 enabled then all held locks will also be reported. This 1095 feature has negligible overhead. 1096 1097config DEFAULT_HUNG_TASK_TIMEOUT 1098 int "Default timeout for hung task detection (in seconds)" 1099 depends on DETECT_HUNG_TASK 1100 default 120 1101 help 1102 This option controls the default timeout (in seconds) used 1103 to determine when a task has become non-responsive and should 1104 be considered hung. 1105 1106 It can be adjusted at runtime via the kernel.hung_task_timeout_secs 1107 sysctl or by writing a value to 1108 /proc/sys/kernel/hung_task_timeout_secs. 1109 1110 A timeout of 0 disables the check. The default is two minutes. 1111 Keeping the default should be fine in most cases. 1112 1113config BOOTPARAM_HUNG_TASK_PANIC 1114 bool "Panic (Reboot) On Hung Tasks" 1115 depends on DETECT_HUNG_TASK 1116 help 1117 Say Y here to enable the kernel to panic on "hung tasks", 1118 which are bugs that cause the kernel to leave a task stuck 1119 in uninterruptible "D" state. 1120 1121 The panic can be used in combination with panic_timeout, 1122 to cause the system to reboot automatically after a 1123 hung task has been detected. This feature is useful for 1124 high-availability systems that have uptime guarantees and 1125 where a hung tasks must be resolved ASAP. 1126 1127 Say N if unsure. 1128 1129config BOOTPARAM_HUNG_TASK_PANIC_VALUE 1130 int 1131 depends on DETECT_HUNG_TASK 1132 range 0 1 1133 default 0 if !BOOTPARAM_HUNG_TASK_PANIC 1134 default 1 if BOOTPARAM_HUNG_TASK_PANIC 1135 1136config WQ_WATCHDOG 1137 bool "Detect Workqueue Stalls" 1138 depends on DEBUG_KERNEL 1139 help 1140 Say Y here to enable stall detection on workqueues. If a 1141 worker pool doesn't make forward progress on a pending work 1142 item for over a given amount of time, 30s by default, a 1143 warning message is printed along with dump of workqueue 1144 state. This can be configured through kernel parameter 1145 "workqueue.watchdog_thresh" and its sysfs counterpart. 1146 1147config TEST_LOCKUP 1148 tristate "Test module to generate lockups" 1149 depends on m 1150 help 1151 This builds the "test_lockup" module that helps to make sure 1152 that watchdogs and lockup detectors are working properly. 1153 1154 Depending on module parameters it could emulate soft or hard 1155 lockup, "hung task", or locking arbitrary lock for a long time. 1156 Also it could generate series of lockups with cooling-down periods. 1157 1158 If unsure, say N. 1159 1160endmenu # "Debug lockups and hangs" 1161 1162menu "Scheduler Debugging" 1163 1164config SCHED_DEBUG 1165 bool "Collect scheduler debugging info" 1166 depends on DEBUG_KERNEL && PROC_FS 1167 default y 1168 help 1169 If you say Y here, the /proc/sched_debug file will be provided 1170 that can help debug the scheduler. The runtime overhead of this 1171 option is minimal. 1172 1173config SCHED_INFO 1174 bool 1175 default n 1176 1177config SCHEDSTATS 1178 bool "Collect scheduler statistics" 1179 depends on DEBUG_KERNEL && PROC_FS 1180 select SCHED_INFO 1181 help 1182 If you say Y here, additional code will be inserted into the 1183 scheduler and related routines to collect statistics about 1184 scheduler behavior and provide them in /proc/schedstat. These 1185 stats may be useful for both tuning and debugging the scheduler 1186 If you aren't debugging the scheduler or trying to tune a specific 1187 application, you can say N to avoid the very slight overhead 1188 this adds. 1189 1190endmenu 1191 1192config DEBUG_TIMEKEEPING 1193 bool "Enable extra timekeeping sanity checking" 1194 help 1195 This option will enable additional timekeeping sanity checks 1196 which may be helpful when diagnosing issues where timekeeping 1197 problems are suspected. 1198 1199 This may include checks in the timekeeping hotpaths, so this 1200 option may have a (very small) performance impact to some 1201 workloads. 1202 1203 If unsure, say N. 1204 1205config DEBUG_PREEMPT 1206 bool "Debug preemptible kernel" 1207 depends on DEBUG_KERNEL && PREEMPTION && TRACE_IRQFLAGS_SUPPORT 1208 default y 1209 help 1210 If you say Y here then the kernel will use a debug variant of the 1211 commonly used smp_processor_id() function and will print warnings 1212 if kernel code uses it in a preemption-unsafe way. Also, the kernel 1213 will detect preemption count underflows. 1214 1215menu "Lock Debugging (spinlocks, mutexes, etc...)" 1216 1217config LOCK_DEBUGGING_SUPPORT 1218 bool 1219 depends on TRACE_IRQFLAGS_SUPPORT && STACKTRACE_SUPPORT && LOCKDEP_SUPPORT 1220 default y 1221 1222config PROVE_LOCKING 1223 bool "Lock debugging: prove locking correctness" 1224 depends on DEBUG_KERNEL && LOCK_DEBUGGING_SUPPORT 1225 select LOCKDEP 1226 select DEBUG_SPINLOCK 1227 select DEBUG_MUTEXES 1228 select DEBUG_RT_MUTEXES if RT_MUTEXES 1229 select DEBUG_RWSEMS 1230 select DEBUG_WW_MUTEX_SLOWPATH 1231 select DEBUG_LOCK_ALLOC 1232 select PREEMPT_COUNT if !ARCH_NO_PREEMPT 1233 select TRACE_IRQFLAGS 1234 default n 1235 help 1236 This feature enables the kernel to prove that all locking 1237 that occurs in the kernel runtime is mathematically 1238 correct: that under no circumstance could an arbitrary (and 1239 not yet triggered) combination of observed locking 1240 sequences (on an arbitrary number of CPUs, running an 1241 arbitrary number of tasks and interrupt contexts) cause a 1242 deadlock. 1243 1244 In short, this feature enables the kernel to report locking 1245 related deadlocks before they actually occur. 1246 1247 The proof does not depend on how hard and complex a 1248 deadlock scenario would be to trigger: how many 1249 participant CPUs, tasks and irq-contexts would be needed 1250 for it to trigger. The proof also does not depend on 1251 timing: if a race and a resulting deadlock is possible 1252 theoretically (no matter how unlikely the race scenario 1253 is), it will be proven so and will immediately be 1254 reported by the kernel (once the event is observed that 1255 makes the deadlock theoretically possible). 1256 1257 If a deadlock is impossible (i.e. the locking rules, as 1258 observed by the kernel, are mathematically correct), the 1259 kernel reports nothing. 1260 1261 NOTE: this feature can also be enabled for rwlocks, mutexes 1262 and rwsems - in which case all dependencies between these 1263 different locking variants are observed and mapped too, and 1264 the proof of observed correctness is also maintained for an 1265 arbitrary combination of these separate locking variants. 1266 1267 For more details, see Documentation/locking/lockdep-design.rst. 1268 1269config PROVE_RAW_LOCK_NESTING 1270 bool "Enable raw_spinlock - spinlock nesting checks" 1271 depends on PROVE_LOCKING 1272 default n 1273 help 1274 Enable the raw_spinlock vs. spinlock nesting checks which ensure 1275 that the lock nesting rules for PREEMPT_RT enabled kernels are 1276 not violated. 1277 1278 NOTE: There are known nesting problems. So if you enable this 1279 option expect lockdep splats until these problems have been fully 1280 addressed which is work in progress. This config switch allows to 1281 identify and analyze these problems. It will be removed and the 1282 check permanentely enabled once the main issues have been fixed. 1283 1284 If unsure, select N. 1285 1286config LOCK_STAT 1287 bool "Lock usage statistics" 1288 depends on DEBUG_KERNEL && LOCK_DEBUGGING_SUPPORT 1289 select LOCKDEP 1290 select DEBUG_SPINLOCK 1291 select DEBUG_MUTEXES 1292 select DEBUG_RT_MUTEXES if RT_MUTEXES 1293 select DEBUG_LOCK_ALLOC 1294 default n 1295 help 1296 This feature enables tracking lock contention points 1297 1298 For more details, see Documentation/locking/lockstat.rst 1299 1300 This also enables lock events required by "perf lock", 1301 subcommand of perf. 1302 If you want to use "perf lock", you also need to turn on 1303 CONFIG_EVENT_TRACING. 1304 1305 CONFIG_LOCK_STAT defines "contended" and "acquired" lock events. 1306 (CONFIG_LOCKDEP defines "acquire" and "release" events.) 1307 1308config DEBUG_RT_MUTEXES 1309 bool "RT Mutex debugging, deadlock detection" 1310 depends on DEBUG_KERNEL && RT_MUTEXES 1311 help 1312 This allows rt mutex semantics violations and rt mutex related 1313 deadlocks (lockups) to be detected and reported automatically. 1314 1315config DEBUG_SPINLOCK 1316 bool "Spinlock and rw-lock debugging: basic checks" 1317 depends on DEBUG_KERNEL 1318 select UNINLINE_SPIN_UNLOCK 1319 help 1320 Say Y here and build SMP to catch missing spinlock initialization 1321 and certain other kinds of spinlock errors commonly made. This is 1322 best used in conjunction with the NMI watchdog so that spinlock 1323 deadlocks are also debuggable. 1324 1325config DEBUG_MUTEXES 1326 bool "Mutex debugging: basic checks" 1327 depends on DEBUG_KERNEL 1328 help 1329 This feature allows mutex semantics violations to be detected and 1330 reported. 1331 1332config DEBUG_WW_MUTEX_SLOWPATH 1333 bool "Wait/wound mutex debugging: Slowpath testing" 1334 depends on DEBUG_KERNEL && LOCK_DEBUGGING_SUPPORT 1335 select DEBUG_LOCK_ALLOC 1336 select DEBUG_SPINLOCK 1337 select DEBUG_MUTEXES 1338 help 1339 This feature enables slowpath testing for w/w mutex users by 1340 injecting additional -EDEADLK wound/backoff cases. Together with 1341 the full mutex checks enabled with (CONFIG_PROVE_LOCKING) this 1342 will test all possible w/w mutex interface abuse with the 1343 exception of simply not acquiring all the required locks. 1344 Note that this feature can introduce significant overhead, so 1345 it really should not be enabled in a production or distro kernel, 1346 even a debug kernel. If you are a driver writer, enable it. If 1347 you are a distro, do not. 1348 1349config DEBUG_RWSEMS 1350 bool "RW Semaphore debugging: basic checks" 1351 depends on DEBUG_KERNEL 1352 help 1353 This debugging feature allows mismatched rw semaphore locks 1354 and unlocks to be detected and reported. 1355 1356config DEBUG_LOCK_ALLOC 1357 bool "Lock debugging: detect incorrect freeing of live locks" 1358 depends on DEBUG_KERNEL && LOCK_DEBUGGING_SUPPORT 1359 select DEBUG_SPINLOCK 1360 select DEBUG_MUTEXES 1361 select DEBUG_RT_MUTEXES if RT_MUTEXES 1362 select LOCKDEP 1363 help 1364 This feature will check whether any held lock (spinlock, rwlock, 1365 mutex or rwsem) is incorrectly freed by the kernel, via any of the 1366 memory-freeing routines (kfree(), kmem_cache_free(), free_pages(), 1367 vfree(), etc.), whether a live lock is incorrectly reinitialized via 1368 spin_lock_init()/mutex_init()/etc., or whether there is any lock 1369 held during task exit. 1370 1371config LOCKDEP 1372 bool 1373 depends on DEBUG_KERNEL && LOCK_DEBUGGING_SUPPORT 1374 select STACKTRACE 1375 depends on FRAME_POINTER || MIPS || PPC || S390 || MICROBLAZE || ARM || ARC || X86 1376 select KALLSYMS 1377 select KALLSYMS_ALL 1378 1379config LOCKDEP_SMALL 1380 bool 1381 1382config LOCKDEP_BITS 1383 int "Bitsize for MAX_LOCKDEP_ENTRIES" 1384 depends on LOCKDEP && !LOCKDEP_SMALL 1385 range 10 30 1386 default 15 1387 help 1388 Try increasing this value if you hit "BUG: MAX_LOCKDEP_ENTRIES too low!" message. 1389 1390config LOCKDEP_CHAINS_BITS 1391 int "Bitsize for MAX_LOCKDEP_CHAINS" 1392 depends on LOCKDEP && !LOCKDEP_SMALL 1393 range 10 30 1394 default 16 1395 help 1396 Try increasing this value if you hit "BUG: MAX_LOCKDEP_CHAINS too low!" message. 1397 1398config LOCKDEP_STACK_TRACE_BITS 1399 int "Bitsize for MAX_STACK_TRACE_ENTRIES" 1400 depends on LOCKDEP && !LOCKDEP_SMALL 1401 range 10 30 1402 default 19 1403 help 1404 Try increasing this value if you hit "BUG: MAX_STACK_TRACE_ENTRIES too low!" message. 1405 1406config LOCKDEP_STACK_TRACE_HASH_BITS 1407 int "Bitsize for STACK_TRACE_HASH_SIZE" 1408 depends on LOCKDEP && !LOCKDEP_SMALL 1409 range 10 30 1410 default 14 1411 help 1412 Try increasing this value if you need large MAX_STACK_TRACE_ENTRIES. 1413 1414config LOCKDEP_CIRCULAR_QUEUE_BITS 1415 int "Bitsize for elements in circular_queue struct" 1416 depends on LOCKDEP 1417 range 10 30 1418 default 12 1419 help 1420 Try increasing this value if you hit "lockdep bfs error:-1" warning due to __cq_enqueue() failure. 1421 1422config DEBUG_LOCKDEP 1423 bool "Lock dependency engine debugging" 1424 depends on DEBUG_KERNEL && LOCKDEP 1425 select DEBUG_IRQFLAGS 1426 help 1427 If you say Y here, the lock dependency engine will do 1428 additional runtime checks to debug itself, at the price 1429 of more runtime overhead. 1430 1431config DEBUG_ATOMIC_SLEEP 1432 bool "Sleep inside atomic section checking" 1433 select PREEMPT_COUNT 1434 depends on DEBUG_KERNEL 1435 depends on !ARCH_NO_PREEMPT 1436 help 1437 If you say Y here, various routines which may sleep will become very 1438 noisy if they are called inside atomic sections: when a spinlock is 1439 held, inside an rcu read side critical section, inside preempt disabled 1440 sections, inside an interrupt, etc... 1441 1442config DEBUG_LOCKING_API_SELFTESTS 1443 bool "Locking API boot-time self-tests" 1444 depends on DEBUG_KERNEL 1445 help 1446 Say Y here if you want the kernel to run a short self-test during 1447 bootup. The self-test checks whether common types of locking bugs 1448 are detected by debugging mechanisms or not. (if you disable 1449 lock debugging then those bugs wont be detected of course.) 1450 The following locking APIs are covered: spinlocks, rwlocks, 1451 mutexes and rwsems. 1452 1453config LOCK_TORTURE_TEST 1454 tristate "torture tests for locking" 1455 depends on DEBUG_KERNEL 1456 select TORTURE_TEST 1457 help 1458 This option provides a kernel module that runs torture tests 1459 on kernel locking primitives. The kernel module may be built 1460 after the fact on the running kernel to be tested, if desired. 1461 1462 Say Y here if you want kernel locking-primitive torture tests 1463 to be built into the kernel. 1464 Say M if you want these torture tests to build as a module. 1465 Say N if you are unsure. 1466 1467config WW_MUTEX_SELFTEST 1468 tristate "Wait/wound mutex selftests" 1469 help 1470 This option provides a kernel module that runs tests on the 1471 on the struct ww_mutex locking API. 1472 1473 It is recommended to enable DEBUG_WW_MUTEX_SLOWPATH in conjunction 1474 with this test harness. 1475 1476 Say M if you want these self tests to build as a module. 1477 Say N if you are unsure. 1478 1479config SCF_TORTURE_TEST 1480 tristate "torture tests for smp_call_function*()" 1481 depends on DEBUG_KERNEL 1482 select TORTURE_TEST 1483 help 1484 This option provides a kernel module that runs torture tests 1485 on the smp_call_function() family of primitives. The kernel 1486 module may be built after the fact on the running kernel to 1487 be tested, if desired. 1488 1489config CSD_LOCK_WAIT_DEBUG 1490 bool "Debugging for csd_lock_wait(), called from smp_call_function*()" 1491 depends on DEBUG_KERNEL 1492 depends on 64BIT 1493 default n 1494 help 1495 This option enables debug prints when CPUs are slow to respond 1496 to the smp_call_function*() IPI wrappers. These debug prints 1497 include the IPI handler function currently executing (if any) 1498 and relevant stack traces. 1499 1500endmenu # lock debugging 1501 1502config TRACE_IRQFLAGS 1503 depends on TRACE_IRQFLAGS_SUPPORT 1504 bool 1505 help 1506 Enables hooks to interrupt enabling and disabling for 1507 either tracing or lock debugging. 1508 1509config TRACE_IRQFLAGS_NMI 1510 def_bool y 1511 depends on TRACE_IRQFLAGS 1512 depends on TRACE_IRQFLAGS_NMI_SUPPORT 1513 1514config DEBUG_IRQFLAGS 1515 bool "Debug IRQ flag manipulation" 1516 help 1517 Enables checks for potentially unsafe enabling or disabling of 1518 interrupts, such as calling raw_local_irq_restore() when interrupts 1519 are enabled. 1520 1521config STACKTRACE 1522 bool "Stack backtrace support" 1523 depends on STACKTRACE_SUPPORT 1524 help 1525 This option causes the kernel to create a /proc/pid/stack for 1526 every process, showing its current stack trace. 1527 It is also used by various kernel debugging features that require 1528 stack trace generation. 1529 1530config WARN_ALL_UNSEEDED_RANDOM 1531 bool "Warn for all uses of unseeded randomness" 1532 default n 1533 help 1534 Some parts of the kernel contain bugs relating to their use of 1535 cryptographically secure random numbers before it's actually possible 1536 to generate those numbers securely. This setting ensures that these 1537 flaws don't go unnoticed, by enabling a message, should this ever 1538 occur. This will allow people with obscure setups to know when things 1539 are going wrong, so that they might contact developers about fixing 1540 it. 1541 1542 Unfortunately, on some models of some architectures getting 1543 a fully seeded CRNG is extremely difficult, and so this can 1544 result in dmesg getting spammed for a surprisingly long 1545 time. This is really bad from a security perspective, and 1546 so architecture maintainers really need to do what they can 1547 to get the CRNG seeded sooner after the system is booted. 1548 However, since users cannot do anything actionable to 1549 address this, by default the kernel will issue only a single 1550 warning for the first use of unseeded randomness. 1551 1552 Say Y here if you want to receive warnings for all uses of 1553 unseeded randomness. This will be of use primarily for 1554 those developers interested in improving the security of 1555 Linux kernels running on their architecture (or 1556 subarchitecture). 1557 1558config DEBUG_KOBJECT 1559 bool "kobject debugging" 1560 depends on DEBUG_KERNEL 1561 help 1562 If you say Y here, some extra kobject debugging messages will be sent 1563 to the syslog. 1564 1565config DEBUG_KOBJECT_RELEASE 1566 bool "kobject release debugging" 1567 depends on DEBUG_OBJECTS_TIMERS 1568 help 1569 kobjects are reference counted objects. This means that their 1570 last reference count put is not predictable, and the kobject can 1571 live on past the point at which a driver decides to drop it's 1572 initial reference to the kobject gained on allocation. An 1573 example of this would be a struct device which has just been 1574 unregistered. 1575 1576 However, some buggy drivers assume that after such an operation, 1577 the memory backing the kobject can be immediately freed. This 1578 goes completely against the principles of a refcounted object. 1579 1580 If you say Y here, the kernel will delay the release of kobjects 1581 on the last reference count to improve the visibility of this 1582 kind of kobject release bug. 1583 1584config HAVE_DEBUG_BUGVERBOSE 1585 bool 1586 1587menu "Debug kernel data structures" 1588 1589config DEBUG_LIST 1590 bool "Debug linked list manipulation" 1591 depends on DEBUG_KERNEL || BUG_ON_DATA_CORRUPTION 1592 help 1593 Enable this to turn on extended checks in the linked-list 1594 walking routines. 1595 1596 If unsure, say N. 1597 1598config DEBUG_PLIST 1599 bool "Debug priority linked list manipulation" 1600 depends on DEBUG_KERNEL 1601 help 1602 Enable this to turn on extended checks in the priority-ordered 1603 linked-list (plist) walking routines. This checks the entire 1604 list multiple times during each manipulation. 1605 1606 If unsure, say N. 1607 1608config DEBUG_SG 1609 bool "Debug SG table operations" 1610 depends on DEBUG_KERNEL 1611 help 1612 Enable this to turn on checks on scatter-gather tables. This can 1613 help find problems with drivers that do not properly initialize 1614 their sg tables. 1615 1616 If unsure, say N. 1617 1618config DEBUG_NOTIFIERS 1619 bool "Debug notifier call chains" 1620 depends on DEBUG_KERNEL 1621 help 1622 Enable this to turn on sanity checking for notifier call chains. 1623 This is most useful for kernel developers to make sure that 1624 modules properly unregister themselves from notifier chains. 1625 This is a relatively cheap check but if you care about maximum 1626 performance, say N. 1627 1628config BUG_ON_DATA_CORRUPTION 1629 bool "Trigger a BUG when data corruption is detected" 1630 select DEBUG_LIST 1631 help 1632 Select this option if the kernel should BUG when it encounters 1633 data corruption in kernel memory structures when they get checked 1634 for validity. 1635 1636 If unsure, say N. 1637 1638endmenu 1639 1640config DEBUG_CREDENTIALS 1641 bool "Debug credential management" 1642 depends on DEBUG_KERNEL 1643 help 1644 Enable this to turn on some debug checking for credential 1645 management. The additional code keeps track of the number of 1646 pointers from task_structs to any given cred struct, and checks to 1647 see that this number never exceeds the usage count of the cred 1648 struct. 1649 1650 Furthermore, if SELinux is enabled, this also checks that the 1651 security pointer in the cred struct is never seen to be invalid. 1652 1653 If unsure, say N. 1654 1655source "kernel/rcu/Kconfig.debug" 1656 1657config DEBUG_WQ_FORCE_RR_CPU 1658 bool "Force round-robin CPU selection for unbound work items" 1659 depends on DEBUG_KERNEL 1660 default n 1661 help 1662 Workqueue used to implicitly guarantee that work items queued 1663 without explicit CPU specified are put on the local CPU. This 1664 guarantee is no longer true and while local CPU is still 1665 preferred work items may be put on foreign CPUs. Kernel 1666 parameter "workqueue.debug_force_rr_cpu" is added to force 1667 round-robin CPU selection to flush out usages which depend on the 1668 now broken guarantee. This config option enables the debug 1669 feature by default. When enabled, memory and cache locality will 1670 be impacted. 1671 1672config DEBUG_BLOCK_EXT_DEVT 1673 bool "Force extended block device numbers and spread them" 1674 depends on DEBUG_KERNEL 1675 depends on BLOCK 1676 default n 1677 help 1678 BIG FAT WARNING: ENABLING THIS OPTION MIGHT BREAK BOOTING ON 1679 SOME DISTRIBUTIONS. DO NOT ENABLE THIS UNLESS YOU KNOW WHAT 1680 YOU ARE DOING. Distros, please enable this and fix whatever 1681 is broken. 1682 1683 Conventionally, block device numbers are allocated from 1684 predetermined contiguous area. However, extended block area 1685 may introduce non-contiguous block device numbers. This 1686 option forces most block device numbers to be allocated from 1687 the extended space and spreads them to discover kernel or 1688 userland code paths which assume predetermined contiguous 1689 device number allocation. 1690 1691 Note that turning on this debug option shuffles all the 1692 device numbers for all IDE and SCSI devices including libata 1693 ones, so root partition specified using device number 1694 directly (via rdev or root=MAJ:MIN) won't work anymore. 1695 Textual device names (root=/dev/sdXn) will continue to work. 1696 1697 Say N if you are unsure. 1698 1699config CPU_HOTPLUG_STATE_CONTROL 1700 bool "Enable CPU hotplug state control" 1701 depends on DEBUG_KERNEL 1702 depends on HOTPLUG_CPU 1703 default n 1704 help 1705 Allows to write steps between "offline" and "online" to the CPUs 1706 sysfs target file so states can be stepped granular. This is a debug 1707 option for now as the hotplug machinery cannot be stopped and 1708 restarted at arbitrary points yet. 1709 1710 Say N if your are unsure. 1711 1712config LATENCYTOP 1713 bool "Latency measuring infrastructure" 1714 depends on DEBUG_KERNEL 1715 depends on STACKTRACE_SUPPORT 1716 depends on PROC_FS 1717 depends on FRAME_POINTER || MIPS || PPC || S390 || MICROBLAZE || ARM || ARC || X86 1718 select KALLSYMS 1719 select KALLSYMS_ALL 1720 select STACKTRACE 1721 select SCHEDSTATS 1722 help 1723 Enable this option if you want to use the LatencyTOP tool 1724 to find out which userspace is blocking on what kernel operations. 1725 1726source "kernel/trace/Kconfig" 1727 1728config PROVIDE_OHCI1394_DMA_INIT 1729 bool "Remote debugging over FireWire early on boot" 1730 depends on PCI && X86 1731 help 1732 If you want to debug problems which hang or crash the kernel early 1733 on boot and the crashing machine has a FireWire port, you can use 1734 this feature to remotely access the memory of the crashed machine 1735 over FireWire. This employs remote DMA as part of the OHCI1394 1736 specification which is now the standard for FireWire controllers. 1737 1738 With remote DMA, you can monitor the printk buffer remotely using 1739 firescope and access all memory below 4GB using fireproxy from gdb. 1740 Even controlling a kernel debugger is possible using remote DMA. 1741 1742 Usage: 1743 1744 If ohci1394_dma=early is used as boot parameter, it will initialize 1745 all OHCI1394 controllers which are found in the PCI config space. 1746 1747 As all changes to the FireWire bus such as enabling and disabling 1748 devices cause a bus reset and thereby disable remote DMA for all 1749 devices, be sure to have the cable plugged and FireWire enabled on 1750 the debugging host before booting the debug target for debugging. 1751 1752 This code (~1k) is freed after boot. By then, the firewire stack 1753 in charge of the OHCI-1394 controllers should be used instead. 1754 1755 See Documentation/core-api/debugging-via-ohci1394.rst for more information. 1756 1757source "samples/Kconfig" 1758 1759config ARCH_HAS_DEVMEM_IS_ALLOWED 1760 bool 1761 1762config STRICT_DEVMEM 1763 bool "Filter access to /dev/mem" 1764 depends on MMU && DEVMEM 1765 depends on ARCH_HAS_DEVMEM_IS_ALLOWED || GENERIC_LIB_DEVMEM_IS_ALLOWED 1766 default y if PPC || X86 || ARM64 1767 help 1768 If this option is disabled, you allow userspace (root) access to all 1769 of memory, including kernel and userspace memory. Accidental 1770 access to this is obviously disastrous, but specific access can 1771 be used by people debugging the kernel. Note that with PAT support 1772 enabled, even in this case there are restrictions on /dev/mem 1773 use due to the cache aliasing requirements. 1774 1775 If this option is switched on, and IO_STRICT_DEVMEM=n, the /dev/mem 1776 file only allows userspace access to PCI space and the BIOS code and 1777 data regions. This is sufficient for dosemu and X and all common 1778 users of /dev/mem. 1779 1780 If in doubt, say Y. 1781 1782config IO_STRICT_DEVMEM 1783 bool "Filter I/O access to /dev/mem" 1784 depends on STRICT_DEVMEM 1785 help 1786 If this option is disabled, you allow userspace (root) access to all 1787 io-memory regardless of whether a driver is actively using that 1788 range. Accidental access to this is obviously disastrous, but 1789 specific access can be used by people debugging kernel drivers. 1790 1791 If this option is switched on, the /dev/mem file only allows 1792 userspace access to *idle* io-memory ranges (see /proc/iomem) This 1793 may break traditional users of /dev/mem (dosemu, legacy X, etc...) 1794 if the driver using a given range cannot be disabled. 1795 1796 If in doubt, say Y. 1797 1798menu "$(SRCARCH) Debugging" 1799 1800source "arch/$(SRCARCH)/Kconfig.debug" 1801 1802endmenu 1803 1804menu "Kernel Testing and Coverage" 1805 1806source "lib/kunit/Kconfig" 1807 1808config NOTIFIER_ERROR_INJECTION 1809 tristate "Notifier error injection" 1810 depends on DEBUG_KERNEL 1811 select DEBUG_FS 1812 help 1813 This option provides the ability to inject artificial errors to 1814 specified notifier chain callbacks. It is useful to test the error 1815 handling of notifier call chain failures. 1816 1817 Say N if unsure. 1818 1819config PM_NOTIFIER_ERROR_INJECT 1820 tristate "PM notifier error injection module" 1821 depends on PM && NOTIFIER_ERROR_INJECTION 1822 default m if PM_DEBUG 1823 help 1824 This option provides the ability to inject artificial errors to 1825 PM notifier chain callbacks. It is controlled through debugfs 1826 interface /sys/kernel/debug/notifier-error-inject/pm 1827 1828 If the notifier call chain should be failed with some events 1829 notified, write the error code to "actions/<notifier event>/error". 1830 1831 Example: Inject PM suspend error (-12 = -ENOMEM) 1832 1833 # cd /sys/kernel/debug/notifier-error-inject/pm/ 1834 # echo -12 > actions/PM_SUSPEND_PREPARE/error 1835 # echo mem > /sys/power/state 1836 bash: echo: write error: Cannot allocate memory 1837 1838 To compile this code as a module, choose M here: the module will 1839 be called pm-notifier-error-inject. 1840 1841 If unsure, say N. 1842 1843config OF_RECONFIG_NOTIFIER_ERROR_INJECT 1844 tristate "OF reconfig notifier error injection module" 1845 depends on OF_DYNAMIC && NOTIFIER_ERROR_INJECTION 1846 help 1847 This option provides the ability to inject artificial errors to 1848 OF reconfig notifier chain callbacks. It is controlled 1849 through debugfs interface under 1850 /sys/kernel/debug/notifier-error-inject/OF-reconfig/ 1851 1852 If the notifier call chain should be failed with some events 1853 notified, write the error code to "actions/<notifier event>/error". 1854 1855 To compile this code as a module, choose M here: the module will 1856 be called of-reconfig-notifier-error-inject. 1857 1858 If unsure, say N. 1859 1860config NETDEV_NOTIFIER_ERROR_INJECT 1861 tristate "Netdev notifier error injection module" 1862 depends on NET && NOTIFIER_ERROR_INJECTION 1863 help 1864 This option provides the ability to inject artificial errors to 1865 netdevice notifier chain callbacks. It is controlled through debugfs 1866 interface /sys/kernel/debug/notifier-error-inject/netdev 1867 1868 If the notifier call chain should be failed with some events 1869 notified, write the error code to "actions/<notifier event>/error". 1870 1871 Example: Inject netdevice mtu change error (-22 = -EINVAL) 1872 1873 # cd /sys/kernel/debug/notifier-error-inject/netdev 1874 # echo -22 > actions/NETDEV_CHANGEMTU/error 1875 # ip link set eth0 mtu 1024 1876 RTNETLINK answers: Invalid argument 1877 1878 To compile this code as a module, choose M here: the module will 1879 be called netdev-notifier-error-inject. 1880 1881 If unsure, say N. 1882 1883config FUNCTION_ERROR_INJECTION 1884 def_bool y 1885 depends on HAVE_FUNCTION_ERROR_INJECTION && KPROBES 1886 1887config FAULT_INJECTION 1888 bool "Fault-injection framework" 1889 depends on DEBUG_KERNEL 1890 help 1891 Provide fault-injection framework. 1892 For more details, see Documentation/fault-injection/. 1893 1894config FAILSLAB 1895 bool "Fault-injection capability for kmalloc" 1896 depends on FAULT_INJECTION 1897 depends on SLAB || SLUB 1898 help 1899 Provide fault-injection capability for kmalloc. 1900 1901config FAIL_PAGE_ALLOC 1902 bool "Fault-injection capability for alloc_pages()" 1903 depends on FAULT_INJECTION 1904 help 1905 Provide fault-injection capability for alloc_pages(). 1906 1907config FAULT_INJECTION_USERCOPY 1908 bool "Fault injection capability for usercopy functions" 1909 depends on FAULT_INJECTION 1910 help 1911 Provides fault-injection capability to inject failures 1912 in usercopy functions (copy_from_user(), get_user(), ...). 1913 1914config FAIL_MAKE_REQUEST 1915 bool "Fault-injection capability for disk IO" 1916 depends on FAULT_INJECTION && BLOCK 1917 help 1918 Provide fault-injection capability for disk IO. 1919 1920config FAIL_IO_TIMEOUT 1921 bool "Fault-injection capability for faking disk interrupts" 1922 depends on FAULT_INJECTION && BLOCK 1923 help 1924 Provide fault-injection capability on end IO handling. This 1925 will make the block layer "forget" an interrupt as configured, 1926 thus exercising the error handling. 1927 1928 Only works with drivers that use the generic timeout handling, 1929 for others it wont do anything. 1930 1931config FAIL_FUTEX 1932 bool "Fault-injection capability for futexes" 1933 select DEBUG_FS 1934 depends on FAULT_INJECTION && FUTEX 1935 help 1936 Provide fault-injection capability for futexes. 1937 1938config FAULT_INJECTION_DEBUG_FS 1939 bool "Debugfs entries for fault-injection capabilities" 1940 depends on FAULT_INJECTION && SYSFS && DEBUG_FS 1941 help 1942 Enable configuration of fault-injection capabilities via debugfs. 1943 1944config FAIL_FUNCTION 1945 bool "Fault-injection capability for functions" 1946 depends on FAULT_INJECTION_DEBUG_FS && FUNCTION_ERROR_INJECTION 1947 help 1948 Provide function-based fault-injection capability. 1949 This will allow you to override a specific function with a return 1950 with given return value. As a result, function caller will see 1951 an error value and have to handle it. This is useful to test the 1952 error handling in various subsystems. 1953 1954config FAIL_MMC_REQUEST 1955 bool "Fault-injection capability for MMC IO" 1956 depends on FAULT_INJECTION_DEBUG_FS && MMC 1957 help 1958 Provide fault-injection capability for MMC IO. 1959 This will make the mmc core return data errors. This is 1960 useful to test the error handling in the mmc block device 1961 and to test how the mmc host driver handles retries from 1962 the block device. 1963 1964config FAULT_INJECTION_STACKTRACE_FILTER 1965 bool "stacktrace filter for fault-injection capabilities" 1966 depends on FAULT_INJECTION_DEBUG_FS && STACKTRACE_SUPPORT 1967 depends on !X86_64 1968 select STACKTRACE 1969 depends on FRAME_POINTER || MIPS || PPC || S390 || MICROBLAZE || ARM || ARC || X86 1970 help 1971 Provide stacktrace filter for fault-injection capabilities 1972 1973config ARCH_HAS_KCOV 1974 bool 1975 help 1976 An architecture should select this when it can successfully 1977 build and run with CONFIG_KCOV. This typically requires 1978 disabling instrumentation for some early boot code. 1979 1980config CC_HAS_SANCOV_TRACE_PC 1981 def_bool $(cc-option,-fsanitize-coverage=trace-pc) 1982 1983 1984config KCOV 1985 bool "Code coverage for fuzzing" 1986 depends on ARCH_HAS_KCOV 1987 depends on CC_HAS_SANCOV_TRACE_PC || GCC_PLUGINS 1988 select DEBUG_FS 1989 select GCC_PLUGIN_SANCOV if !CC_HAS_SANCOV_TRACE_PC 1990 help 1991 KCOV exposes kernel code coverage information in a form suitable 1992 for coverage-guided fuzzing (randomized testing). 1993 1994 If RANDOMIZE_BASE is enabled, PC values will not be stable across 1995 different machines and across reboots. If you need stable PC values, 1996 disable RANDOMIZE_BASE. 1997 1998 For more details, see Documentation/dev-tools/kcov.rst. 1999 2000config KCOV_ENABLE_COMPARISONS 2001 bool "Enable comparison operands collection by KCOV" 2002 depends on KCOV 2003 depends on $(cc-option,-fsanitize-coverage=trace-cmp) 2004 help 2005 KCOV also exposes operands of every comparison in the instrumented 2006 code along with operand sizes and PCs of the comparison instructions. 2007 These operands can be used by fuzzing engines to improve the quality 2008 of fuzzing coverage. 2009 2010config KCOV_INSTRUMENT_ALL 2011 bool "Instrument all code by default" 2012 depends on KCOV 2013 default y 2014 help 2015 If you are doing generic system call fuzzing (like e.g. syzkaller), 2016 then you will want to instrument the whole kernel and you should 2017 say y here. If you are doing more targeted fuzzing (like e.g. 2018 filesystem fuzzing with AFL) then you will want to enable coverage 2019 for more specific subsets of files, and should say n here. 2020 2021config KCOV_IRQ_AREA_SIZE 2022 hex "Size of interrupt coverage collection area in words" 2023 depends on KCOV 2024 default 0x40000 2025 help 2026 KCOV uses preallocated per-cpu areas to collect coverage from 2027 soft interrupts. This specifies the size of those areas in the 2028 number of unsigned long words. 2029 2030menuconfig RUNTIME_TESTING_MENU 2031 bool "Runtime Testing" 2032 def_bool y 2033 2034if RUNTIME_TESTING_MENU 2035 2036config LKDTM 2037 tristate "Linux Kernel Dump Test Tool Module" 2038 depends on DEBUG_FS 2039 help 2040 This module enables testing of the different dumping mechanisms by 2041 inducing system failures at predefined crash points. 2042 If you don't need it: say N 2043 Choose M here to compile this code as a module. The module will be 2044 called lkdtm. 2045 2046 Documentation on how to use the module can be found in 2047 Documentation/fault-injection/provoke-crashes.rst 2048 2049config TEST_LIST_SORT 2050 tristate "Linked list sorting test" if !KUNIT_ALL_TESTS 2051 depends on KUNIT 2052 default KUNIT_ALL_TESTS 2053 help 2054 Enable this to turn on 'list_sort()' function test. This test is 2055 executed only once during system boot (so affects only boot time), 2056 or at module load time. 2057 2058 If unsure, say N. 2059 2060config TEST_MIN_HEAP 2061 tristate "Min heap test" 2062 depends on DEBUG_KERNEL || m 2063 help 2064 Enable this to turn on min heap function tests. This test is 2065 executed only once during system boot (so affects only boot time), 2066 or at module load time. 2067 2068 If unsure, say N. 2069 2070config TEST_SORT 2071 tristate "Array-based sort test" 2072 depends on DEBUG_KERNEL || m 2073 help 2074 This option enables the self-test function of 'sort()' at boot, 2075 or at module load time. 2076 2077 If unsure, say N. 2078 2079config TEST_DIV64 2080 tristate "64bit/32bit division and modulo test" 2081 depends on DEBUG_KERNEL || m 2082 help 2083 Enable this to turn on 'do_div()' function test. This test is 2084 executed only once during system boot (so affects only boot time), 2085 or at module load time. 2086 2087 If unsure, say N. 2088 2089config KPROBES_SANITY_TEST 2090 bool "Kprobes sanity tests" 2091 depends on DEBUG_KERNEL 2092 depends on KPROBES 2093 help 2094 This option provides for testing basic kprobes functionality on 2095 boot. Samples of kprobe and kretprobe are inserted and 2096 verified for functionality. 2097 2098 Say N if you are unsure. 2099 2100config BACKTRACE_SELF_TEST 2101 tristate "Self test for the backtrace code" 2102 depends on DEBUG_KERNEL 2103 help 2104 This option provides a kernel module that can be used to test 2105 the kernel stack backtrace code. This option is not useful 2106 for distributions or general kernels, but only for kernel 2107 developers working on architecture code. 2108 2109 Note that if you want to also test saved backtraces, you will 2110 have to enable STACKTRACE as well. 2111 2112 Say N if you are unsure. 2113 2114config RBTREE_TEST 2115 tristate "Red-Black tree test" 2116 depends on DEBUG_KERNEL 2117 help 2118 A benchmark measuring the performance of the rbtree library. 2119 Also includes rbtree invariant checks. 2120 2121config REED_SOLOMON_TEST 2122 tristate "Reed-Solomon library test" 2123 depends on DEBUG_KERNEL || m 2124 select REED_SOLOMON 2125 select REED_SOLOMON_ENC16 2126 select REED_SOLOMON_DEC16 2127 help 2128 This option enables the self-test function of rslib at boot, 2129 or at module load time. 2130 2131 If unsure, say N. 2132 2133config INTERVAL_TREE_TEST 2134 tristate "Interval tree test" 2135 depends on DEBUG_KERNEL 2136 select INTERVAL_TREE 2137 help 2138 A benchmark measuring the performance of the interval tree library 2139 2140config PERCPU_TEST 2141 tristate "Per cpu operations test" 2142 depends on m && DEBUG_KERNEL 2143 help 2144 Enable this option to build test module which validates per-cpu 2145 operations. 2146 2147 If unsure, say N. 2148 2149config ATOMIC64_SELFTEST 2150 tristate "Perform an atomic64_t self-test" 2151 help 2152 Enable this option to test the atomic64_t functions at boot or 2153 at module load time. 2154 2155 If unsure, say N. 2156 2157config ASYNC_RAID6_TEST 2158 tristate "Self test for hardware accelerated raid6 recovery" 2159 depends on ASYNC_RAID6_RECOV 2160 select ASYNC_MEMCPY 2161 help 2162 This is a one-shot self test that permutes through the 2163 recovery of all the possible two disk failure scenarios for a 2164 N-disk array. Recovery is performed with the asynchronous 2165 raid6 recovery routines, and will optionally use an offload 2166 engine if one is available. 2167 2168 If unsure, say N. 2169 2170config TEST_HEXDUMP 2171 tristate "Test functions located in the hexdump module at runtime" 2172 2173config TEST_STRING_HELPERS 2174 tristate "Test functions located in the string_helpers module at runtime" 2175 2176config TEST_STRSCPY 2177 tristate "Test strscpy*() family of functions at runtime" 2178 2179config TEST_KSTRTOX 2180 tristate "Test kstrto*() family of functions at runtime" 2181 2182config TEST_PRINTF 2183 tristate "Test printf() family of functions at runtime" 2184 2185config TEST_BITMAP 2186 tristate "Test bitmap_*() family of functions at runtime" 2187 help 2188 Enable this option to test the bitmap functions at boot. 2189 2190 If unsure, say N. 2191 2192config TEST_UUID 2193 tristate "Test functions located in the uuid module at runtime" 2194 2195config TEST_XARRAY 2196 tristate "Test the XArray code at runtime" 2197 2198config TEST_OVERFLOW 2199 tristate "Test check_*_overflow() functions at runtime" 2200 2201config TEST_RHASHTABLE 2202 tristate "Perform selftest on resizable hash table" 2203 help 2204 Enable this option to test the rhashtable functions at boot. 2205 2206 If unsure, say N. 2207 2208config TEST_HASH 2209 tristate "Perform selftest on hash functions" 2210 help 2211 Enable this option to test the kernel's integer (<linux/hash.h>), 2212 string (<linux/stringhash.h>), and siphash (<linux/siphash.h>) 2213 hash functions on boot (or module load). 2214 2215 This is intended to help people writing architecture-specific 2216 optimized versions. If unsure, say N. 2217 2218config TEST_IDA 2219 tristate "Perform selftest on IDA functions" 2220 2221config TEST_PARMAN 2222 tristate "Perform selftest on priority array manager" 2223 depends on PARMAN 2224 help 2225 Enable this option to test priority array manager on boot 2226 (or module load). 2227 2228 If unsure, say N. 2229 2230config TEST_IRQ_TIMINGS 2231 bool "IRQ timings selftest" 2232 depends on IRQ_TIMINGS 2233 help 2234 Enable this option to test the irq timings code on boot. 2235 2236 If unsure, say N. 2237 2238config TEST_LKM 2239 tristate "Test module loading with 'hello world' module" 2240 depends on m 2241 help 2242 This builds the "test_module" module that emits "Hello, world" 2243 on printk when loaded. It is designed to be used for basic 2244 evaluation of the module loading subsystem (for example when 2245 validating module verification). It lacks any extra dependencies, 2246 and will not normally be loaded by the system unless explicitly 2247 requested by name. 2248 2249 If unsure, say N. 2250 2251config TEST_BITOPS 2252 tristate "Test module for compilation of bitops operations" 2253 depends on m 2254 help 2255 This builds the "test_bitops" module that is much like the 2256 TEST_LKM module except that it does a basic exercise of the 2257 set/clear_bit macros and get_count_order/long to make sure there are 2258 no compiler warnings from C=1 sparse checker or -Wextra 2259 compilations. It has no dependencies and doesn't run or load unless 2260 explicitly requested by name. for example: modprobe test_bitops. 2261 2262 If unsure, say N. 2263 2264config TEST_VMALLOC 2265 tristate "Test module for stress/performance analysis of vmalloc allocator" 2266 default n 2267 depends on MMU 2268 depends on m 2269 help 2270 This builds the "test_vmalloc" module that should be used for 2271 stress and performance analysis. So, any new change for vmalloc 2272 subsystem can be evaluated from performance and stability point 2273 of view. 2274 2275 If unsure, say N. 2276 2277config TEST_USER_COPY 2278 tristate "Test user/kernel boundary protections" 2279 depends on m 2280 help 2281 This builds the "test_user_copy" module that runs sanity checks 2282 on the copy_to/from_user infrastructure, making sure basic 2283 user/kernel boundary testing is working. If it fails to load, 2284 a regression has been detected in the user/kernel memory boundary 2285 protections. 2286 2287 If unsure, say N. 2288 2289config TEST_BPF 2290 tristate "Test BPF filter functionality" 2291 depends on m && NET 2292 help 2293 This builds the "test_bpf" module that runs various test vectors 2294 against the BPF interpreter or BPF JIT compiler depending on the 2295 current setting. This is in particular useful for BPF JIT compiler 2296 development, but also to run regression tests against changes in 2297 the interpreter code. It also enables test stubs for eBPF maps and 2298 verifier used by user space verifier testsuite. 2299 2300 If unsure, say N. 2301 2302config TEST_BLACKHOLE_DEV 2303 tristate "Test blackhole netdev functionality" 2304 depends on m && NET 2305 help 2306 This builds the "test_blackhole_dev" module that validates the 2307 data path through this blackhole netdev. 2308 2309 If unsure, say N. 2310 2311config FIND_BIT_BENCHMARK 2312 tristate "Test find_bit functions" 2313 help 2314 This builds the "test_find_bit" module that measure find_*_bit() 2315 functions performance. 2316 2317 If unsure, say N. 2318 2319config TEST_FIRMWARE 2320 tristate "Test firmware loading via userspace interface" 2321 depends on FW_LOADER 2322 help 2323 This builds the "test_firmware" module that creates a userspace 2324 interface for testing firmware loading. This can be used to 2325 control the triggering of firmware loading without needing an 2326 actual firmware-using device. The contents can be rechecked by 2327 userspace. 2328 2329 If unsure, say N. 2330 2331config TEST_SYSCTL 2332 tristate "sysctl test driver" 2333 depends on PROC_SYSCTL 2334 help 2335 This builds the "test_sysctl" module. This driver enables to test the 2336 proc sysctl interfaces available to drivers safely without affecting 2337 production knobs which might alter system functionality. 2338 2339 If unsure, say N. 2340 2341config BITFIELD_KUNIT 2342 tristate "KUnit test bitfield functions at runtime" 2343 depends on KUNIT 2344 help 2345 Enable this option to test the bitfield functions at boot. 2346 2347 KUnit tests run during boot and output the results to the debug log 2348 in TAP format (http://testanything.org/). Only useful for kernel devs 2349 running the KUnit test harness, and not intended for inclusion into a 2350 production build. 2351 2352 For more information on KUnit and unit tests in general please refer 2353 to the KUnit documentation in Documentation/dev-tools/kunit/. 2354 2355 If unsure, say N. 2356 2357config RESOURCE_KUNIT_TEST 2358 tristate "KUnit test for resource API" 2359 depends on KUNIT 2360 help 2361 This builds the resource API unit test. 2362 Tests the logic of API provided by resource.c and ioport.h. 2363 For more information on KUnit and unit tests in general please refer 2364 to the KUnit documentation in Documentation/dev-tools/kunit/. 2365 2366 If unsure, say N. 2367 2368config SYSCTL_KUNIT_TEST 2369 tristate "KUnit test for sysctl" if !KUNIT_ALL_TESTS 2370 depends on KUNIT 2371 default KUNIT_ALL_TESTS 2372 help 2373 This builds the proc sysctl unit test, which runs on boot. 2374 Tests the API contract and implementation correctness of sysctl. 2375 For more information on KUnit and unit tests in general please refer 2376 to the KUnit documentation in Documentation/dev-tools/kunit/. 2377 2378 If unsure, say N. 2379 2380config LIST_KUNIT_TEST 2381 tristate "KUnit Test for Kernel Linked-list structures" if !KUNIT_ALL_TESTS 2382 depends on KUNIT 2383 default KUNIT_ALL_TESTS 2384 help 2385 This builds the linked list KUnit test suite. 2386 It tests that the API and basic functionality of the list_head type 2387 and associated macros. 2388 2389 KUnit tests run during boot and output the results to the debug log 2390 in TAP format (https://testanything.org/). Only useful for kernel devs 2391 running the KUnit test harness, and not intended for inclusion into a 2392 production build. 2393 2394 For more information on KUnit and unit tests in general please refer 2395 to the KUnit documentation in Documentation/dev-tools/kunit/. 2396 2397 If unsure, say N. 2398 2399config LINEAR_RANGES_TEST 2400 tristate "KUnit test for linear_ranges" 2401 depends on KUNIT 2402 select LINEAR_RANGES 2403 help 2404 This builds the linear_ranges unit test, which runs on boot. 2405 Tests the linear_ranges logic correctness. 2406 For more information on KUnit and unit tests in general please refer 2407 to the KUnit documentation in Documentation/dev-tools/kunit/. 2408 2409 If unsure, say N. 2410 2411config CMDLINE_KUNIT_TEST 2412 tristate "KUnit test for cmdline API" 2413 depends on KUNIT 2414 help 2415 This builds the cmdline API unit test. 2416 Tests the logic of API provided by cmdline.c. 2417 For more information on KUnit and unit tests in general please refer 2418 to the KUnit documentation in Documentation/dev-tools/kunit/. 2419 2420 If unsure, say N. 2421 2422config BITS_TEST 2423 tristate "KUnit test for bits.h" 2424 depends on KUNIT 2425 help 2426 This builds the bits unit test. 2427 Tests the logic of macros defined in bits.h. 2428 For more information on KUnit and unit tests in general please refer 2429 to the KUnit documentation in Documentation/dev-tools/kunit/. 2430 2431 If unsure, say N. 2432 2433config TEST_UDELAY 2434 tristate "udelay test driver" 2435 help 2436 This builds the "udelay_test" module that helps to make sure 2437 that udelay() is working properly. 2438 2439 If unsure, say N. 2440 2441config TEST_STATIC_KEYS 2442 tristate "Test static keys" 2443 depends on m 2444 help 2445 Test the static key interfaces. 2446 2447 If unsure, say N. 2448 2449config TEST_KMOD 2450 tristate "kmod stress tester" 2451 depends on m 2452 depends on NETDEVICES && NET_CORE && INET # for TUN 2453 depends on BLOCK 2454 select TEST_LKM 2455 select XFS_FS 2456 select TUN 2457 select BTRFS_FS 2458 help 2459 Test the kernel's module loading mechanism: kmod. kmod implements 2460 support to load modules using the Linux kernel's usermode helper. 2461 This test provides a series of tests against kmod. 2462 2463 Although technically you can either build test_kmod as a module or 2464 into the kernel we disallow building it into the kernel since 2465 it stress tests request_module() and this will very likely cause 2466 some issues by taking over precious threads available from other 2467 module load requests, ultimately this could be fatal. 2468 2469 To run tests run: 2470 2471 tools/testing/selftests/kmod/kmod.sh --help 2472 2473 If unsure, say N. 2474 2475config TEST_DEBUG_VIRTUAL 2476 tristate "Test CONFIG_DEBUG_VIRTUAL feature" 2477 depends on DEBUG_VIRTUAL 2478 help 2479 Test the kernel's ability to detect incorrect calls to 2480 virt_to_phys() done against the non-linear part of the 2481 kernel's virtual address map. 2482 2483 If unsure, say N. 2484 2485config TEST_MEMCAT_P 2486 tristate "Test memcat_p() helper function" 2487 help 2488 Test the memcat_p() helper for correctly merging two 2489 pointer arrays together. 2490 2491 If unsure, say N. 2492 2493config TEST_LIVEPATCH 2494 tristate "Test livepatching" 2495 default n 2496 depends on DYNAMIC_DEBUG 2497 depends on LIVEPATCH 2498 depends on m 2499 help 2500 Test kernel livepatching features for correctness. The tests will 2501 load test modules that will be livepatched in various scenarios. 2502 2503 To run all the livepatching tests: 2504 2505 make -C tools/testing/selftests TARGETS=livepatch run_tests 2506 2507 Alternatively, individual tests may be invoked: 2508 2509 tools/testing/selftests/livepatch/test-callbacks.sh 2510 tools/testing/selftests/livepatch/test-livepatch.sh 2511 tools/testing/selftests/livepatch/test-shadow-vars.sh 2512 2513 If unsure, say N. 2514 2515config TEST_OBJAGG 2516 tristate "Perform selftest on object aggreration manager" 2517 default n 2518 depends on OBJAGG 2519 help 2520 Enable this option to test object aggregation manager on boot 2521 (or module load). 2522 2523 2524config TEST_STACKINIT 2525 tristate "Test level of stack variable initialization" 2526 help 2527 Test if the kernel is zero-initializing stack variables and 2528 padding. Coverage is controlled by compiler flags, 2529 CONFIG_GCC_PLUGIN_STRUCTLEAK, CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF, 2530 or CONFIG_GCC_PLUGIN_STRUCTLEAK_BYREF_ALL. 2531 2532 If unsure, say N. 2533 2534config TEST_MEMINIT 2535 tristate "Test heap/page initialization" 2536 help 2537 Test if the kernel is zero-initializing heap and page allocations. 2538 This can be useful to test init_on_alloc and init_on_free features. 2539 2540 If unsure, say N. 2541 2542config TEST_HMM 2543 tristate "Test HMM (Heterogeneous Memory Management)" 2544 depends on TRANSPARENT_HUGEPAGE 2545 depends on DEVICE_PRIVATE 2546 select HMM_MIRROR 2547 select MMU_NOTIFIER 2548 help 2549 This is a pseudo device driver solely for testing HMM. 2550 Say M here if you want to build the HMM test module. 2551 Doing so will allow you to run tools/testing/selftest/vm/hmm-tests. 2552 2553 If unsure, say N. 2554 2555config TEST_FREE_PAGES 2556 tristate "Test freeing pages" 2557 help 2558 Test that a memory leak does not occur due to a race between 2559 freeing a block of pages and a speculative page reference. 2560 Loading this module is safe if your kernel has the bug fixed. 2561 If the bug is not fixed, it will leak gigabytes of memory and 2562 probably OOM your system. 2563 2564config TEST_FPU 2565 tristate "Test floating point operations in kernel space" 2566 depends on X86 && !KCOV_INSTRUMENT_ALL 2567 help 2568 Enable this option to add /sys/kernel/debug/selftest_helpers/test_fpu 2569 which will trigger a sequence of floating point operations. This is used 2570 for self-testing floating point control register setting in 2571 kernel_fpu_begin(). 2572 2573 If unsure, say N. 2574 2575endif # RUNTIME_TESTING_MENU 2576 2577config ARCH_USE_MEMTEST 2578 bool 2579 help 2580 An architecture should select this when it uses early_memtest() 2581 during boot process. 2582 2583config MEMTEST 2584 bool "Memtest" 2585 depends on ARCH_USE_MEMTEST 2586 help 2587 This option adds a kernel parameter 'memtest', which allows memtest 2588 to be set and executed. 2589 memtest=0, mean disabled; -- default 2590 memtest=1, mean do 1 test pattern; 2591 ... 2592 memtest=17, mean do 17 test patterns. 2593 If you are unsure how to answer this question, answer N. 2594 2595 2596 2597config HYPERV_TESTING 2598 bool "Microsoft Hyper-V driver testing" 2599 default n 2600 depends on HYPERV && DEBUG_FS 2601 help 2602 Select this option to enable Hyper-V vmbus testing. 2603 2604endmenu # "Kernel Testing and Coverage" 2605 2606source "Documentation/Kconfig" 2607 2608endmenu # Kernel hacking 2609