1config DEFCONFIG_LIST 2 string 3 depends on !UML 4 option defconfig_list 5 default "/lib/modules/$(shell,uname -r)/.config" 6 default "/etc/kernel-config" 7 default "/boot/config-$(shell,uname -r)" 8 default ARCH_DEFCONFIG 9 default "arch/$(ARCH)/defconfig" 10 11config CC_IS_GCC 12 def_bool $(success,$(CC) --version | head -n 1 | grep -q gcc) 13 14config GCC_VERSION 15 int 16 default $(shell,$(srctree)/scripts/gcc-version.sh $(CC)) if CC_IS_GCC 17 default 0 18 19config CC_IS_CLANG 20 def_bool $(success,$(CC) --version | head -n 1 | grep -q clang) 21 22config CLANG_VERSION 23 int 24 default $(shell,$(srctree)/scripts/clang-version.sh $(CC)) 25 26config CC_HAS_ASM_GOTO 27 def_bool $(success,$(srctree)/scripts/gcc-goto.sh $(CC)) 28 29config CC_HAS_WARN_MAYBE_UNINITIALIZED 30 def_bool $(cc-option,-Wmaybe-uninitialized) 31 help 32 GCC >= 4.7 supports this option. 33 34config CC_DISABLE_WARN_MAYBE_UNINITIALIZED 35 bool 36 depends on CC_HAS_WARN_MAYBE_UNINITIALIZED 37 default CC_IS_GCC && GCC_VERSION < 40900 # unreliable for GCC < 4.9 38 help 39 GCC's -Wmaybe-uninitialized is not reliable by definition. 40 Lots of false positive warnings are produced in some cases. 41 42 If this option is enabled, -Wno-maybe-uninitialzed is passed 43 to the compiler to suppress maybe-uninitialized warnings. 44 45config CONSTRUCTORS 46 bool 47 depends on !UML 48 49config IRQ_WORK 50 bool 51 52config BUILDTIME_EXTABLE_SORT 53 bool 54 55config THREAD_INFO_IN_TASK 56 bool 57 help 58 Select this to move thread_info off the stack into task_struct. To 59 make this work, an arch will need to remove all thread_info fields 60 except flags and fix any runtime bugs. 61 62 One subtle change that will be needed is to use try_get_task_stack() 63 and put_task_stack() in save_thread_stack_tsk() and get_wchan(). 64 65menu "General setup" 66 67config BROKEN 68 bool 69 70config BROKEN_ON_SMP 71 bool 72 depends on BROKEN || !SMP 73 default y 74 75config INIT_ENV_ARG_LIMIT 76 int 77 default 32 if !UML 78 default 128 if UML 79 help 80 Maximum of each of the number of arguments and environment 81 variables passed to init from the kernel command line. 82 83config COMPILE_TEST 84 bool "Compile also drivers which will not load" 85 depends on !UML 86 default n 87 help 88 Some drivers can be compiled on a different platform than they are 89 intended to be run on. Despite they cannot be loaded there (or even 90 when they load they cannot be used due to missing HW support), 91 developers still, opposing to distributors, might want to build such 92 drivers to compile-test them. 93 94 If you are a developer and want to build everything available, say Y 95 here. If you are a user/distributor, say N here to exclude useless 96 drivers to be distributed. 97 98config LOCALVERSION 99 string "Local version - append to kernel release" 100 help 101 Append an extra string to the end of your kernel version. 102 This will show up when you type uname, for example. 103 The string you set here will be appended after the contents of 104 any files with a filename matching localversion* in your 105 object and source tree, in that order. Your total string can 106 be a maximum of 64 characters. 107 108config LOCALVERSION_AUTO 109 bool "Automatically append version information to the version string" 110 default y 111 depends on !COMPILE_TEST 112 help 113 This will try to automatically determine if the current tree is a 114 release tree by looking for git tags that belong to the current 115 top of tree revision. 116 117 A string of the format -gxxxxxxxx will be added to the localversion 118 if a git-based tree is found. The string generated by this will be 119 appended after any matching localversion* files, and after the value 120 set in CONFIG_LOCALVERSION. 121 122 (The actual string used here is the first eight characters produced 123 by running the command: 124 125 $ git rev-parse --verify HEAD 126 127 which is done within the script "scripts/setlocalversion".) 128 129config BUILD_SALT 130 string "Build ID Salt" 131 default "" 132 help 133 The build ID is used to link binaries and their debug info. Setting 134 this option will use the value in the calculation of the build id. 135 This is mostly useful for distributions which want to ensure the 136 build is unique between builds. It's safe to leave the default. 137 138config HAVE_KERNEL_GZIP 139 bool 140 141config HAVE_KERNEL_BZIP2 142 bool 143 144config HAVE_KERNEL_LZMA 145 bool 146 147config HAVE_KERNEL_XZ 148 bool 149 150config HAVE_KERNEL_LZO 151 bool 152 153config HAVE_KERNEL_LZ4 154 bool 155 156config HAVE_KERNEL_UNCOMPRESSED 157 bool 158 159choice 160 prompt "Kernel compression mode" 161 default KERNEL_GZIP 162 depends on HAVE_KERNEL_GZIP || HAVE_KERNEL_BZIP2 || HAVE_KERNEL_LZMA || HAVE_KERNEL_XZ || HAVE_KERNEL_LZO || HAVE_KERNEL_LZ4 || HAVE_KERNEL_UNCOMPRESSED 163 help 164 The linux kernel is a kind of self-extracting executable. 165 Several compression algorithms are available, which differ 166 in efficiency, compression and decompression speed. 167 Compression speed is only relevant when building a kernel. 168 Decompression speed is relevant at each boot. 169 170 If you have any problems with bzip2 or lzma compressed 171 kernels, mail me (Alain Knaff) <alain@knaff.lu>. (An older 172 version of this functionality (bzip2 only), for 2.4, was 173 supplied by Christian Ludwig) 174 175 High compression options are mostly useful for users, who 176 are low on disk space (embedded systems), but for whom ram 177 size matters less. 178 179 If in doubt, select 'gzip' 180 181config KERNEL_GZIP 182 bool "Gzip" 183 depends on HAVE_KERNEL_GZIP 184 help 185 The old and tried gzip compression. It provides a good balance 186 between compression ratio and decompression speed. 187 188config KERNEL_BZIP2 189 bool "Bzip2" 190 depends on HAVE_KERNEL_BZIP2 191 help 192 Its compression ratio and speed is intermediate. 193 Decompression speed is slowest among the choices. The kernel 194 size is about 10% smaller with bzip2, in comparison to gzip. 195 Bzip2 uses a large amount of memory. For modern kernels you 196 will need at least 8MB RAM or more for booting. 197 198config KERNEL_LZMA 199 bool "LZMA" 200 depends on HAVE_KERNEL_LZMA 201 help 202 This compression algorithm's ratio is best. Decompression speed 203 is between gzip and bzip2. Compression is slowest. 204 The kernel size is about 33% smaller with LZMA in comparison to gzip. 205 206config KERNEL_XZ 207 bool "XZ" 208 depends on HAVE_KERNEL_XZ 209 help 210 XZ uses the LZMA2 algorithm and instruction set specific 211 BCJ filters which can improve compression ratio of executable 212 code. The size of the kernel is about 30% smaller with XZ in 213 comparison to gzip. On architectures for which there is a BCJ 214 filter (i386, x86_64, ARM, IA-64, PowerPC, and SPARC), XZ 215 will create a few percent smaller kernel than plain LZMA. 216 217 The speed is about the same as with LZMA: The decompression 218 speed of XZ is better than that of bzip2 but worse than gzip 219 and LZO. Compression is slow. 220 221config KERNEL_LZO 222 bool "LZO" 223 depends on HAVE_KERNEL_LZO 224 help 225 Its compression ratio is the poorest among the choices. The kernel 226 size is about 10% bigger than gzip; however its speed 227 (both compression and decompression) is the fastest. 228 229config KERNEL_LZ4 230 bool "LZ4" 231 depends on HAVE_KERNEL_LZ4 232 help 233 LZ4 is an LZ77-type compressor with a fixed, byte-oriented encoding. 234 A preliminary version of LZ4 de/compression tool is available at 235 <https://code.google.com/p/lz4/>. 236 237 Its compression ratio is worse than LZO. The size of the kernel 238 is about 8% bigger than LZO. But the decompression speed is 239 faster than LZO. 240 241config KERNEL_UNCOMPRESSED 242 bool "None" 243 depends on HAVE_KERNEL_UNCOMPRESSED 244 help 245 Produce uncompressed kernel image. This option is usually not what 246 you want. It is useful for debugging the kernel in slow simulation 247 environments, where decompressing and moving the kernel is awfully 248 slow. This option allows early boot code to skip the decompressor 249 and jump right at uncompressed kernel image. 250 251endchoice 252 253config DEFAULT_HOSTNAME 254 string "Default hostname" 255 default "(none)" 256 help 257 This option determines the default system hostname before userspace 258 calls sethostname(2). The kernel traditionally uses "(none)" here, 259 but you may wish to use a different default here to make a minimal 260 system more usable with less configuration. 261 262# 263# For some reason microblaze and nios2 hard code SWAP=n. Hopefully we can 264# add proper SWAP support to them, in which case this can be remove. 265# 266config ARCH_NO_SWAP 267 bool 268 269config SWAP 270 bool "Support for paging of anonymous memory (swap)" 271 depends on MMU && BLOCK && !ARCH_NO_SWAP 272 default y 273 help 274 This option allows you to choose whether you want to have support 275 for so called swap devices or swap files in your kernel that are 276 used to provide more virtual memory than the actual RAM present 277 in your computer. If unsure say Y. 278 279config SYSVIPC 280 bool "System V IPC" 281 ---help--- 282 Inter Process Communication is a suite of library functions and 283 system calls which let processes (running programs) synchronize and 284 exchange information. It is generally considered to be a good thing, 285 and some programs won't run unless you say Y here. In particular, if 286 you want to run the DOS emulator dosemu under Linux (read the 287 DOSEMU-HOWTO, available from <http://www.tldp.org/docs.html#howto>), 288 you'll need to say Y here. 289 290 You can find documentation about IPC with "info ipc" and also in 291 section 6.4 of the Linux Programmer's Guide, available from 292 <http://www.tldp.org/guides.html>. 293 294config SYSVIPC_SYSCTL 295 bool 296 depends on SYSVIPC 297 depends on SYSCTL 298 default y 299 300config POSIX_MQUEUE 301 bool "POSIX Message Queues" 302 depends on NET 303 ---help--- 304 POSIX variant of message queues is a part of IPC. In POSIX message 305 queues every message has a priority which decides about succession 306 of receiving it by a process. If you want to compile and run 307 programs written e.g. for Solaris with use of its POSIX message 308 queues (functions mq_*) say Y here. 309 310 POSIX message queues are visible as a filesystem called 'mqueue' 311 and can be mounted somewhere if you want to do filesystem 312 operations on message queues. 313 314 If unsure, say Y. 315 316config POSIX_MQUEUE_SYSCTL 317 bool 318 depends on POSIX_MQUEUE 319 depends on SYSCTL 320 default y 321 322config CROSS_MEMORY_ATTACH 323 bool "Enable process_vm_readv/writev syscalls" 324 depends on MMU 325 default y 326 help 327 Enabling this option adds the system calls process_vm_readv and 328 process_vm_writev which allow a process with the correct privileges 329 to directly read from or write to another process' address space. 330 See the man page for more details. 331 332config USELIB 333 bool "uselib syscall" 334 def_bool ALPHA || M68K || SPARC || X86_32 || IA32_EMULATION 335 help 336 This option enables the uselib syscall, a system call used in the 337 dynamic linker from libc5 and earlier. glibc does not use this 338 system call. If you intend to run programs built on libc5 or 339 earlier, you may need to enable this syscall. Current systems 340 running glibc can safely disable this. 341 342config AUDIT 343 bool "Auditing support" 344 depends on NET 345 help 346 Enable auditing infrastructure that can be used with another 347 kernel subsystem, such as SELinux (which requires this for 348 logging of avc messages output). System call auditing is included 349 on architectures which support it. 350 351config HAVE_ARCH_AUDITSYSCALL 352 bool 353 354config AUDITSYSCALL 355 def_bool y 356 depends on AUDIT && HAVE_ARCH_AUDITSYSCALL 357 select FSNOTIFY 358 359source "kernel/irq/Kconfig" 360source "kernel/time/Kconfig" 361source "kernel/Kconfig.preempt" 362 363menu "CPU/Task time and stats accounting" 364 365config VIRT_CPU_ACCOUNTING 366 bool 367 368choice 369 prompt "Cputime accounting" 370 default TICK_CPU_ACCOUNTING if !PPC64 371 default VIRT_CPU_ACCOUNTING_NATIVE if PPC64 372 373# Kind of a stub config for the pure tick based cputime accounting 374config TICK_CPU_ACCOUNTING 375 bool "Simple tick based cputime accounting" 376 depends on !S390 && !NO_HZ_FULL 377 help 378 This is the basic tick based cputime accounting that maintains 379 statistics about user, system and idle time spent on per jiffies 380 granularity. 381 382 If unsure, say Y. 383 384config VIRT_CPU_ACCOUNTING_NATIVE 385 bool "Deterministic task and CPU time accounting" 386 depends on HAVE_VIRT_CPU_ACCOUNTING && !NO_HZ_FULL 387 select VIRT_CPU_ACCOUNTING 388 help 389 Select this option to enable more accurate task and CPU time 390 accounting. This is done by reading a CPU counter on each 391 kernel entry and exit and on transitions within the kernel 392 between system, softirq and hardirq state, so there is a 393 small performance impact. In the case of s390 or IBM POWER > 5, 394 this also enables accounting of stolen time on logically-partitioned 395 systems. 396 397config VIRT_CPU_ACCOUNTING_GEN 398 bool "Full dynticks CPU time accounting" 399 depends on HAVE_CONTEXT_TRACKING 400 depends on HAVE_VIRT_CPU_ACCOUNTING_GEN 401 depends on GENERIC_CLOCKEVENTS 402 select VIRT_CPU_ACCOUNTING 403 select CONTEXT_TRACKING 404 help 405 Select this option to enable task and CPU time accounting on full 406 dynticks systems. This accounting is implemented by watching every 407 kernel-user boundaries using the context tracking subsystem. 408 The accounting is thus performed at the expense of some significant 409 overhead. 410 411 For now this is only useful if you are working on the full 412 dynticks subsystem development. 413 414 If unsure, say N. 415 416endchoice 417 418config IRQ_TIME_ACCOUNTING 419 bool "Fine granularity task level IRQ time accounting" 420 depends on HAVE_IRQ_TIME_ACCOUNTING && !VIRT_CPU_ACCOUNTING_NATIVE 421 help 422 Select this option to enable fine granularity task irq time 423 accounting. This is done by reading a timestamp on each 424 transitions between softirq and hardirq state, so there can be a 425 small performance impact. 426 427 If in doubt, say N here. 428 429config HAVE_SCHED_AVG_IRQ 430 def_bool y 431 depends on IRQ_TIME_ACCOUNTING || PARAVIRT_TIME_ACCOUNTING 432 depends on SMP 433 434config BSD_PROCESS_ACCT 435 bool "BSD Process Accounting" 436 depends on MULTIUSER 437 help 438 If you say Y here, a user level program will be able to instruct the 439 kernel (via a special system call) to write process accounting 440 information to a file: whenever a process exits, information about 441 that process will be appended to the file by the kernel. The 442 information includes things such as creation time, owning user, 443 command name, memory usage, controlling terminal etc. (the complete 444 list is in the struct acct in <file:include/linux/acct.h>). It is 445 up to the user level program to do useful things with this 446 information. This is generally a good idea, so say Y. 447 448config BSD_PROCESS_ACCT_V3 449 bool "BSD Process Accounting version 3 file format" 450 depends on BSD_PROCESS_ACCT 451 default n 452 help 453 If you say Y here, the process accounting information is written 454 in a new file format that also logs the process IDs of each 455 process and its parent. Note that this file format is incompatible 456 with previous v0/v1/v2 file formats, so you will need updated tools 457 for processing it. A preliminary version of these tools is available 458 at <http://www.gnu.org/software/acct/>. 459 460config TASKSTATS 461 bool "Export task/process statistics through netlink" 462 depends on NET 463 depends on MULTIUSER 464 default n 465 help 466 Export selected statistics for tasks/processes through the 467 generic netlink interface. Unlike BSD process accounting, the 468 statistics are available during the lifetime of tasks/processes as 469 responses to commands. Like BSD accounting, they are sent to user 470 space on task exit. 471 472 Say N if unsure. 473 474config TASK_DELAY_ACCT 475 bool "Enable per-task delay accounting" 476 depends on TASKSTATS 477 select SCHED_INFO 478 help 479 Collect information on time spent by a task waiting for system 480 resources like cpu, synchronous block I/O completion and swapping 481 in pages. Such statistics can help in setting a task's priorities 482 relative to other tasks for cpu, io, rss limits etc. 483 484 Say N if unsure. 485 486config TASK_XACCT 487 bool "Enable extended accounting over taskstats" 488 depends on TASKSTATS 489 help 490 Collect extended task accounting data and send the data 491 to userland for processing over the taskstats interface. 492 493 Say N if unsure. 494 495config TASK_IO_ACCOUNTING 496 bool "Enable per-task storage I/O accounting" 497 depends on TASK_XACCT 498 help 499 Collect information on the number of bytes of storage I/O which this 500 task has caused. 501 502 Say N if unsure. 503 504config PSI 505 bool "Pressure stall information tracking" 506 help 507 Collect metrics that indicate how overcommitted the CPU, memory, 508 and IO capacity are in the system. 509 510 If you say Y here, the kernel will create /proc/pressure/ with the 511 pressure statistics files cpu, memory, and io. These will indicate 512 the share of walltime in which some or all tasks in the system are 513 delayed due to contention of the respective resource. 514 515 In kernels with cgroup support, cgroups (cgroup2 only) will 516 have cpu.pressure, memory.pressure, and io.pressure files, 517 which aggregate pressure stalls for the grouped tasks only. 518 519 For more details see Documentation/accounting/psi.txt. 520 521 Say N if unsure. 522 523config PSI_DEFAULT_DISABLED 524 bool "Require boot parameter to enable pressure stall information tracking" 525 default n 526 depends on PSI 527 help 528 If set, pressure stall information tracking will be disabled 529 per default but can be enabled through passing psi=1 on the 530 kernel commandline during boot. 531 532 This feature adds some code to the task wakeup and sleep 533 paths of the scheduler. The overhead is too low to affect 534 common scheduling-intense workloads in practice (such as 535 webservers, memcache), but it does show up in artificial 536 scheduler stress tests, such as hackbench. 537 538 If you are paranoid and not sure what the kernel will be 539 used for, say Y. 540 541 Say N if unsure. 542 543endmenu # "CPU/Task time and stats accounting" 544 545config CPU_ISOLATION 546 bool "CPU isolation" 547 depends on SMP || COMPILE_TEST 548 default y 549 help 550 Make sure that CPUs running critical tasks are not disturbed by 551 any source of "noise" such as unbound workqueues, timers, kthreads... 552 Unbound jobs get offloaded to housekeeping CPUs. This is driven by 553 the "isolcpus=" boot parameter. 554 555 Say Y if unsure. 556 557source "kernel/rcu/Kconfig" 558 559config BUILD_BIN2C 560 bool 561 default n 562 563config IKCONFIG 564 tristate "Kernel .config support" 565 select BUILD_BIN2C 566 ---help--- 567 This option enables the complete Linux kernel ".config" file 568 contents to be saved in the kernel. It provides documentation 569 of which kernel options are used in a running kernel or in an 570 on-disk kernel. This information can be extracted from the kernel 571 image file with the script scripts/extract-ikconfig and used as 572 input to rebuild the current kernel or to build another kernel. 573 It can also be extracted from a running kernel by reading 574 /proc/config.gz if enabled (below). 575 576config IKCONFIG_PROC 577 bool "Enable access to .config through /proc/config.gz" 578 depends on IKCONFIG && PROC_FS 579 ---help--- 580 This option enables access to the kernel configuration file 581 through /proc/config.gz. 582 583config LOG_BUF_SHIFT 584 int "Kernel log buffer size (16 => 64KB, 17 => 128KB)" 585 range 12 25 586 default 17 587 depends on PRINTK 588 help 589 Select the minimal kernel log buffer size as a power of 2. 590 The final size is affected by LOG_CPU_MAX_BUF_SHIFT config 591 parameter, see below. Any higher size also might be forced 592 by "log_buf_len" boot parameter. 593 594 Examples: 595 17 => 128 KB 596 16 => 64 KB 597 15 => 32 KB 598 14 => 16 KB 599 13 => 8 KB 600 12 => 4 KB 601 602config LOG_CPU_MAX_BUF_SHIFT 603 int "CPU kernel log buffer size contribution (13 => 8 KB, 17 => 128KB)" 604 depends on SMP 605 range 0 21 606 default 12 if !BASE_SMALL 607 default 0 if BASE_SMALL 608 depends on PRINTK 609 help 610 This option allows to increase the default ring buffer size 611 according to the number of CPUs. The value defines the contribution 612 of each CPU as a power of 2. The used space is typically only few 613 lines however it might be much more when problems are reported, 614 e.g. backtraces. 615 616 The increased size means that a new buffer has to be allocated and 617 the original static one is unused. It makes sense only on systems 618 with more CPUs. Therefore this value is used only when the sum of 619 contributions is greater than the half of the default kernel ring 620 buffer as defined by LOG_BUF_SHIFT. The default values are set 621 so that more than 64 CPUs are needed to trigger the allocation. 622 623 Also this option is ignored when "log_buf_len" kernel parameter is 624 used as it forces an exact (power of two) size of the ring buffer. 625 626 The number of possible CPUs is used for this computation ignoring 627 hotplugging making the computation optimal for the worst case 628 scenario while allowing a simple algorithm to be used from bootup. 629 630 Examples shift values and their meaning: 631 17 => 128 KB for each CPU 632 16 => 64 KB for each CPU 633 15 => 32 KB for each CPU 634 14 => 16 KB for each CPU 635 13 => 8 KB for each CPU 636 12 => 4 KB for each CPU 637 638config PRINTK_SAFE_LOG_BUF_SHIFT 639 int "Temporary per-CPU printk log buffer size (12 => 4KB, 13 => 8KB)" 640 range 10 21 641 default 13 642 depends on PRINTK 643 help 644 Select the size of an alternate printk per-CPU buffer where messages 645 printed from usafe contexts are temporary stored. One example would 646 be NMI messages, another one - printk recursion. The messages are 647 copied to the main log buffer in a safe context to avoid a deadlock. 648 The value defines the size as a power of 2. 649 650 Those messages are rare and limited. The largest one is when 651 a backtrace is printed. It usually fits into 4KB. Select 652 8KB if you want to be on the safe side. 653 654 Examples: 655 17 => 128 KB for each CPU 656 16 => 64 KB for each CPU 657 15 => 32 KB for each CPU 658 14 => 16 KB for each CPU 659 13 => 8 KB for each CPU 660 12 => 4 KB for each CPU 661 662# 663# Architectures with an unreliable sched_clock() should select this: 664# 665config HAVE_UNSTABLE_SCHED_CLOCK 666 bool 667 668config GENERIC_SCHED_CLOCK 669 bool 670 671# 672# For architectures that want to enable the support for NUMA-affine scheduler 673# balancing logic: 674# 675config ARCH_SUPPORTS_NUMA_BALANCING 676 bool 677 678# 679# For architectures that prefer to flush all TLBs after a number of pages 680# are unmapped instead of sending one IPI per page to flush. The architecture 681# must provide guarantees on what happens if a clean TLB cache entry is 682# written after the unmap. Details are in mm/rmap.c near the check for 683# should_defer_flush. The architecture should also consider if the full flush 684# and the refill costs are offset by the savings of sending fewer IPIs. 685config ARCH_WANT_BATCHED_UNMAP_TLB_FLUSH 686 bool 687 688# 689# For architectures that know their GCC __int128 support is sound 690# 691config ARCH_SUPPORTS_INT128 692 bool 693 694# For architectures that (ab)use NUMA to represent different memory regions 695# all cpu-local but of different latencies, such as SuperH. 696# 697config ARCH_WANT_NUMA_VARIABLE_LOCALITY 698 bool 699 700config NUMA_BALANCING 701 bool "Memory placement aware NUMA scheduler" 702 depends on ARCH_SUPPORTS_NUMA_BALANCING 703 depends on !ARCH_WANT_NUMA_VARIABLE_LOCALITY 704 depends on SMP && NUMA && MIGRATION 705 help 706 This option adds support for automatic NUMA aware memory/task placement. 707 The mechanism is quite primitive and is based on migrating memory when 708 it has references to the node the task is running on. 709 710 This system will be inactive on UMA systems. 711 712config NUMA_BALANCING_DEFAULT_ENABLED 713 bool "Automatically enable NUMA aware memory/task placement" 714 default y 715 depends on NUMA_BALANCING 716 help 717 If set, automatic NUMA balancing will be enabled if running on a NUMA 718 machine. 719 720menuconfig CGROUPS 721 bool "Control Group support" 722 select KERNFS 723 help 724 This option adds support for grouping sets of processes together, for 725 use with process control subsystems such as Cpusets, CFS, memory 726 controls or device isolation. 727 See 728 - Documentation/scheduler/sched-design-CFS.txt (CFS) 729 - Documentation/cgroup-v1/ (features for grouping, isolation 730 and resource control) 731 732 Say N if unsure. 733 734if CGROUPS 735 736config PAGE_COUNTER 737 bool 738 739config MEMCG 740 bool "Memory controller" 741 select PAGE_COUNTER 742 select EVENTFD 743 help 744 Provides control over the memory footprint of tasks in a cgroup. 745 746config MEMCG_SWAP 747 bool "Swap controller" 748 depends on MEMCG && SWAP 749 help 750 Provides control over the swap space consumed by tasks in a cgroup. 751 752config MEMCG_SWAP_ENABLED 753 bool "Swap controller enabled by default" 754 depends on MEMCG_SWAP 755 default y 756 help 757 Memory Resource Controller Swap Extension comes with its price in 758 a bigger memory consumption. General purpose distribution kernels 759 which want to enable the feature but keep it disabled by default 760 and let the user enable it by swapaccount=1 boot command line 761 parameter should have this option unselected. 762 For those who want to have the feature enabled by default should 763 select this option (if, for some reason, they need to disable it 764 then swapaccount=0 does the trick). 765 766config MEMCG_KMEM 767 bool 768 depends on MEMCG && !SLOB 769 default y 770 771config BLK_CGROUP 772 bool "IO controller" 773 depends on BLOCK 774 default n 775 ---help--- 776 Generic block IO controller cgroup interface. This is the common 777 cgroup interface which should be used by various IO controlling 778 policies. 779 780 Currently, CFQ IO scheduler uses it to recognize task groups and 781 control disk bandwidth allocation (proportional time slice allocation) 782 to such task groups. It is also used by bio throttling logic in 783 block layer to implement upper limit in IO rates on a device. 784 785 This option only enables generic Block IO controller infrastructure. 786 One needs to also enable actual IO controlling logic/policy. For 787 enabling proportional weight division of disk bandwidth in CFQ, set 788 CONFIG_CFQ_GROUP_IOSCHED=y; for enabling throttling policy, set 789 CONFIG_BLK_DEV_THROTTLING=y. 790 791 See Documentation/cgroup-v1/blkio-controller.txt for more information. 792 793config DEBUG_BLK_CGROUP 794 bool "IO controller debugging" 795 depends on BLK_CGROUP 796 default n 797 ---help--- 798 Enable some debugging help. Currently it exports additional stat 799 files in a cgroup which can be useful for debugging. 800 801config CGROUP_WRITEBACK 802 bool 803 depends on MEMCG && BLK_CGROUP 804 default y 805 806menuconfig CGROUP_SCHED 807 bool "CPU controller" 808 default n 809 help 810 This feature lets CPU scheduler recognize task groups and control CPU 811 bandwidth allocation to such task groups. It uses cgroups to group 812 tasks. 813 814if CGROUP_SCHED 815config FAIR_GROUP_SCHED 816 bool "Group scheduling for SCHED_OTHER" 817 depends on CGROUP_SCHED 818 default CGROUP_SCHED 819 820config CFS_BANDWIDTH 821 bool "CPU bandwidth provisioning for FAIR_GROUP_SCHED" 822 depends on FAIR_GROUP_SCHED 823 default n 824 help 825 This option allows users to define CPU bandwidth rates (limits) for 826 tasks running within the fair group scheduler. Groups with no limit 827 set are considered to be unconstrained and will run with no 828 restriction. 829 See Documentation/scheduler/sched-bwc.txt for more information. 830 831config RT_GROUP_SCHED 832 bool "Group scheduling for SCHED_RR/FIFO" 833 depends on CGROUP_SCHED 834 default n 835 help 836 This feature lets you explicitly allocate real CPU bandwidth 837 to task groups. If enabled, it will also make it impossible to 838 schedule realtime tasks for non-root users until you allocate 839 realtime bandwidth for them. 840 See Documentation/scheduler/sched-rt-group.txt for more information. 841 842endif #CGROUP_SCHED 843 844config CGROUP_PIDS 845 bool "PIDs controller" 846 help 847 Provides enforcement of process number limits in the scope of a 848 cgroup. Any attempt to fork more processes than is allowed in the 849 cgroup will fail. PIDs are fundamentally a global resource because it 850 is fairly trivial to reach PID exhaustion before you reach even a 851 conservative kmemcg limit. As a result, it is possible to grind a 852 system to halt without being limited by other cgroup policies. The 853 PIDs controller is designed to stop this from happening. 854 855 It should be noted that organisational operations (such as attaching 856 to a cgroup hierarchy) will *not* be blocked by the PIDs controller, 857 since the PIDs limit only affects a process's ability to fork, not to 858 attach to a cgroup. 859 860config CGROUP_RDMA 861 bool "RDMA controller" 862 help 863 Provides enforcement of RDMA resources defined by IB stack. 864 It is fairly easy for consumers to exhaust RDMA resources, which 865 can result into resource unavailability to other consumers. 866 RDMA controller is designed to stop this from happening. 867 Attaching processes with active RDMA resources to the cgroup 868 hierarchy is allowed even if can cross the hierarchy's limit. 869 870config CGROUP_FREEZER 871 bool "Freezer controller" 872 help 873 Provides a way to freeze and unfreeze all tasks in a 874 cgroup. 875 876 This option affects the ORIGINAL cgroup interface. The cgroup2 memory 877 controller includes important in-kernel memory consumers per default. 878 879 If you're using cgroup2, say N. 880 881config CGROUP_HUGETLB 882 bool "HugeTLB controller" 883 depends on HUGETLB_PAGE 884 select PAGE_COUNTER 885 default n 886 help 887 Provides a cgroup controller for HugeTLB pages. 888 When you enable this, you can put a per cgroup limit on HugeTLB usage. 889 The limit is enforced during page fault. Since HugeTLB doesn't 890 support page reclaim, enforcing the limit at page fault time implies 891 that, the application will get SIGBUS signal if it tries to access 892 HugeTLB pages beyond its limit. This requires the application to know 893 beforehand how much HugeTLB pages it would require for its use. The 894 control group is tracked in the third page lru pointer. This means 895 that we cannot use the controller with huge page less than 3 pages. 896 897config CPUSETS 898 bool "Cpuset controller" 899 depends on SMP 900 help 901 This option will let you create and manage CPUSETs which 902 allow dynamically partitioning a system into sets of CPUs and 903 Memory Nodes and assigning tasks to run only within those sets. 904 This is primarily useful on large SMP or NUMA systems. 905 906 Say N if unsure. 907 908config PROC_PID_CPUSET 909 bool "Include legacy /proc/<pid>/cpuset file" 910 depends on CPUSETS 911 default y 912 913config CGROUP_DEVICE 914 bool "Device controller" 915 help 916 Provides a cgroup controller implementing whitelists for 917 devices which a process in the cgroup can mknod or open. 918 919config CGROUP_CPUACCT 920 bool "Simple CPU accounting controller" 921 help 922 Provides a simple controller for monitoring the 923 total CPU consumed by the tasks in a cgroup. 924 925config CGROUP_PERF 926 bool "Perf controller" 927 depends on PERF_EVENTS 928 help 929 This option extends the perf per-cpu mode to restrict monitoring 930 to threads which belong to the cgroup specified and run on the 931 designated cpu. 932 933 Say N if unsure. 934 935config CGROUP_BPF 936 bool "Support for eBPF programs attached to cgroups" 937 depends on BPF_SYSCALL 938 select SOCK_CGROUP_DATA 939 help 940 Allow attaching eBPF programs to a cgroup using the bpf(2) 941 syscall command BPF_PROG_ATTACH. 942 943 In which context these programs are accessed depends on the type 944 of attachment. For instance, programs that are attached using 945 BPF_CGROUP_INET_INGRESS will be executed on the ingress path of 946 inet sockets. 947 948config CGROUP_DEBUG 949 bool "Debug controller" 950 default n 951 depends on DEBUG_KERNEL 952 help 953 This option enables a simple controller that exports 954 debugging information about the cgroups framework. This 955 controller is for control cgroup debugging only. Its 956 interfaces are not stable. 957 958 Say N. 959 960config SOCK_CGROUP_DATA 961 bool 962 default n 963 964endif # CGROUPS 965 966menuconfig NAMESPACES 967 bool "Namespaces support" if EXPERT 968 depends on MULTIUSER 969 default !EXPERT 970 help 971 Provides the way to make tasks work with different objects using 972 the same id. For example same IPC id may refer to different objects 973 or same user id or pid may refer to different tasks when used in 974 different namespaces. 975 976if NAMESPACES 977 978config UTS_NS 979 bool "UTS namespace" 980 default y 981 help 982 In this namespace tasks see different info provided with the 983 uname() system call 984 985config IPC_NS 986 bool "IPC namespace" 987 depends on (SYSVIPC || POSIX_MQUEUE) 988 default y 989 help 990 In this namespace tasks work with IPC ids which correspond to 991 different IPC objects in different namespaces. 992 993config USER_NS 994 bool "User namespace" 995 default n 996 help 997 This allows containers, i.e. vservers, to use user namespaces 998 to provide different user info for different servers. 999 1000 When user namespaces are enabled in the kernel it is 1001 recommended that the MEMCG option also be enabled and that 1002 user-space use the memory control groups to limit the amount 1003 of memory a memory unprivileged users can use. 1004 1005 If unsure, say N. 1006 1007config PID_NS 1008 bool "PID Namespaces" 1009 default y 1010 help 1011 Support process id namespaces. This allows having multiple 1012 processes with the same pid as long as they are in different 1013 pid namespaces. This is a building block of containers. 1014 1015config NET_NS 1016 bool "Network namespace" 1017 depends on NET 1018 default y 1019 help 1020 Allow user space to create what appear to be multiple instances 1021 of the network stack. 1022 1023endif # NAMESPACES 1024 1025config CHECKPOINT_RESTORE 1026 bool "Checkpoint/restore support" 1027 select PROC_CHILDREN 1028 default n 1029 help 1030 Enables additional kernel features in a sake of checkpoint/restore. 1031 In particular it adds auxiliary prctl codes to setup process text, 1032 data and heap segment sizes, and a few additional /proc filesystem 1033 entries. 1034 1035 If unsure, say N here. 1036 1037config SCHED_AUTOGROUP 1038 bool "Automatic process group scheduling" 1039 select CGROUPS 1040 select CGROUP_SCHED 1041 select FAIR_GROUP_SCHED 1042 help 1043 This option optimizes the scheduler for common desktop workloads by 1044 automatically creating and populating task groups. This separation 1045 of workloads isolates aggressive CPU burners (like build jobs) from 1046 desktop applications. Task group autogeneration is currently based 1047 upon task session. 1048 1049config SYSFS_DEPRECATED 1050 bool "Enable deprecated sysfs features to support old userspace tools" 1051 depends on SYSFS 1052 default n 1053 help 1054 This option adds code that switches the layout of the "block" class 1055 devices, to not show up in /sys/class/block/, but only in 1056 /sys/block/. 1057 1058 This switch is only active when the sysfs.deprecated=1 boot option is 1059 passed or the SYSFS_DEPRECATED_V2 option is set. 1060 1061 This option allows new kernels to run on old distributions and tools, 1062 which might get confused by /sys/class/block/. Since 2007/2008 all 1063 major distributions and tools handle this just fine. 1064 1065 Recent distributions and userspace tools after 2009/2010 depend on 1066 the existence of /sys/class/block/, and will not work with this 1067 option enabled. 1068 1069 Only if you are using a new kernel on an old distribution, you might 1070 need to say Y here. 1071 1072config SYSFS_DEPRECATED_V2 1073 bool "Enable deprecated sysfs features by default" 1074 default n 1075 depends on SYSFS 1076 depends on SYSFS_DEPRECATED 1077 help 1078 Enable deprecated sysfs by default. 1079 1080 See the CONFIG_SYSFS_DEPRECATED option for more details about this 1081 option. 1082 1083 Only if you are using a new kernel on an old distribution, you might 1084 need to say Y here. Even then, odds are you would not need it 1085 enabled, you can always pass the boot option if absolutely necessary. 1086 1087config RELAY 1088 bool "Kernel->user space relay support (formerly relayfs)" 1089 select IRQ_WORK 1090 help 1091 This option enables support for relay interface support in 1092 certain file systems (such as debugfs). 1093 It is designed to provide an efficient mechanism for tools and 1094 facilities to relay large amounts of data from kernel space to 1095 user space. 1096 1097 If unsure, say N. 1098 1099config BLK_DEV_INITRD 1100 bool "Initial RAM filesystem and RAM disk (initramfs/initrd) support" 1101 help 1102 The initial RAM filesystem is a ramfs which is loaded by the 1103 boot loader (loadlin or lilo) and that is mounted as root 1104 before the normal boot procedure. It is typically used to 1105 load modules needed to mount the "real" root file system, 1106 etc. See <file:Documentation/admin-guide/initrd.rst> for details. 1107 1108 If RAM disk support (BLK_DEV_RAM) is also included, this 1109 also enables initial RAM disk (initrd) support and adds 1110 15 Kbytes (more on some other architectures) to the kernel size. 1111 1112 If unsure say Y. 1113 1114if BLK_DEV_INITRD 1115 1116source "usr/Kconfig" 1117 1118endif 1119 1120choice 1121 prompt "Compiler optimization level" 1122 default CC_OPTIMIZE_FOR_PERFORMANCE 1123 1124config CC_OPTIMIZE_FOR_PERFORMANCE 1125 bool "Optimize for performance" 1126 help 1127 This is the default optimization level for the kernel, building 1128 with the "-O2" compiler flag for best performance and most 1129 helpful compile-time warnings. 1130 1131config CC_OPTIMIZE_FOR_SIZE 1132 bool "Optimize for size" 1133 imply CC_DISABLE_WARN_MAYBE_UNINITIALIZED # avoid false positives 1134 help 1135 Enabling this option will pass "-Os" instead of "-O2" to 1136 your compiler resulting in a smaller kernel. 1137 1138 If unsure, say N. 1139 1140endchoice 1141 1142config HAVE_LD_DEAD_CODE_DATA_ELIMINATION 1143 bool 1144 help 1145 This requires that the arch annotates or otherwise protects 1146 its external entry points from being discarded. Linker scripts 1147 must also merge .text.*, .data.*, and .bss.* correctly into 1148 output sections. Care must be taken not to pull in unrelated 1149 sections (e.g., '.text.init'). Typically '.' in section names 1150 is used to distinguish them from label names / C identifiers. 1151 1152config LD_DEAD_CODE_DATA_ELIMINATION 1153 bool "Dead code and data elimination (EXPERIMENTAL)" 1154 depends on HAVE_LD_DEAD_CODE_DATA_ELIMINATION 1155 depends on EXPERT 1156 depends on !(FUNCTION_TRACER && CC_IS_GCC && GCC_VERSION < 40800) 1157 depends on $(cc-option,-ffunction-sections -fdata-sections) 1158 depends on $(ld-option,--gc-sections) 1159 help 1160 Enable this if you want to do dead code and data elimination with 1161 the linker by compiling with -ffunction-sections -fdata-sections, 1162 and linking with --gc-sections. 1163 1164 This can reduce on disk and in-memory size of the kernel 1165 code and static data, particularly for small configs and 1166 on small systems. This has the possibility of introducing 1167 silently broken kernel if the required annotations are not 1168 present. This option is not well tested yet, so use at your 1169 own risk. 1170 1171config SYSCTL 1172 bool 1173 1174config HAVE_UID16 1175 bool 1176 1177config SYSCTL_EXCEPTION_TRACE 1178 bool 1179 help 1180 Enable support for /proc/sys/debug/exception-trace. 1181 1182config SYSCTL_ARCH_UNALIGN_NO_WARN 1183 bool 1184 help 1185 Enable support for /proc/sys/kernel/ignore-unaligned-usertrap 1186 Allows arch to define/use @no_unaligned_warning to possibly warn 1187 about unaligned access emulation going on under the hood. 1188 1189config SYSCTL_ARCH_UNALIGN_ALLOW 1190 bool 1191 help 1192 Enable support for /proc/sys/kernel/unaligned-trap 1193 Allows arches to define/use @unaligned_enabled to runtime toggle 1194 the unaligned access emulation. 1195 see arch/parisc/kernel/unaligned.c for reference 1196 1197config HAVE_PCSPKR_PLATFORM 1198 bool 1199 1200# interpreter that classic socket filters depend on 1201config BPF 1202 bool 1203 1204menuconfig EXPERT 1205 bool "Configure standard kernel features (expert users)" 1206 # Unhide debug options, to make the on-by-default options visible 1207 select DEBUG_KERNEL 1208 help 1209 This option allows certain base kernel options and settings 1210 to be disabled or tweaked. This is for specialized 1211 environments which can tolerate a "non-standard" kernel. 1212 Only use this if you really know what you are doing. 1213 1214config UID16 1215 bool "Enable 16-bit UID system calls" if EXPERT 1216 depends on HAVE_UID16 && MULTIUSER 1217 default y 1218 help 1219 This enables the legacy 16-bit UID syscall wrappers. 1220 1221config MULTIUSER 1222 bool "Multiple users, groups and capabilities support" if EXPERT 1223 default y 1224 help 1225 This option enables support for non-root users, groups and 1226 capabilities. 1227 1228 If you say N here, all processes will run with UID 0, GID 0, and all 1229 possible capabilities. Saying N here also compiles out support for 1230 system calls related to UIDs, GIDs, and capabilities, such as setuid, 1231 setgid, and capset. 1232 1233 If unsure, say Y here. 1234 1235config SGETMASK_SYSCALL 1236 bool "sgetmask/ssetmask syscalls support" if EXPERT 1237 def_bool PARISC || M68K || PPC || MIPS || X86 || SPARC || MICROBLAZE || SUPERH 1238 ---help--- 1239 sys_sgetmask and sys_ssetmask are obsolete system calls 1240 no longer supported in libc but still enabled by default in some 1241 architectures. 1242 1243 If unsure, leave the default option here. 1244 1245config SYSFS_SYSCALL 1246 bool "Sysfs syscall support" if EXPERT 1247 default y 1248 ---help--- 1249 sys_sysfs is an obsolete system call no longer supported in libc. 1250 Note that disabling this option is more secure but might break 1251 compatibility with some systems. 1252 1253 If unsure say Y here. 1254 1255config SYSCTL_SYSCALL 1256 bool "Sysctl syscall support" if EXPERT 1257 depends on PROC_SYSCTL 1258 default n 1259 select SYSCTL 1260 ---help--- 1261 sys_sysctl uses binary paths that have been found challenging 1262 to properly maintain and use. The interface in /proc/sys 1263 using paths with ascii names is now the primary path to this 1264 information. 1265 1266 Almost nothing using the binary sysctl interface so if you are 1267 trying to save some space it is probably safe to disable this, 1268 making your kernel marginally smaller. 1269 1270 If unsure say N here. 1271 1272config FHANDLE 1273 bool "open by fhandle syscalls" if EXPERT 1274 select EXPORTFS 1275 default y 1276 help 1277 If you say Y here, a user level program will be able to map 1278 file names to handle and then later use the handle for 1279 different file system operations. This is useful in implementing 1280 userspace file servers, which now track files using handles instead 1281 of names. The handle would remain the same even if file names 1282 get renamed. Enables open_by_handle_at(2) and name_to_handle_at(2) 1283 syscalls. 1284 1285config POSIX_TIMERS 1286 bool "Posix Clocks & timers" if EXPERT 1287 default y 1288 help 1289 This includes native support for POSIX timers to the kernel. 1290 Some embedded systems have no use for them and therefore they 1291 can be configured out to reduce the size of the kernel image. 1292 1293 When this option is disabled, the following syscalls won't be 1294 available: timer_create, timer_gettime: timer_getoverrun, 1295 timer_settime, timer_delete, clock_adjtime, getitimer, 1296 setitimer, alarm. Furthermore, the clock_settime, clock_gettime, 1297 clock_getres and clock_nanosleep syscalls will be limited to 1298 CLOCK_REALTIME, CLOCK_MONOTONIC and CLOCK_BOOTTIME only. 1299 1300 If unsure say y. 1301 1302config PRINTK 1303 default y 1304 bool "Enable support for printk" if EXPERT 1305 select IRQ_WORK 1306 help 1307 This option enables normal printk support. Removing it 1308 eliminates most of the message strings from the kernel image 1309 and makes the kernel more or less silent. As this makes it 1310 very difficult to diagnose system problems, saying N here is 1311 strongly discouraged. 1312 1313config PRINTK_NMI 1314 def_bool y 1315 depends on PRINTK 1316 depends on HAVE_NMI 1317 1318config BUG 1319 bool "BUG() support" if EXPERT 1320 default y 1321 help 1322 Disabling this option eliminates support for BUG and WARN, reducing 1323 the size of your kernel image and potentially quietly ignoring 1324 numerous fatal conditions. You should only consider disabling this 1325 option for embedded systems with no facilities for reporting errors. 1326 Just say Y. 1327 1328config ELF_CORE 1329 depends on COREDUMP 1330 default y 1331 bool "Enable ELF core dumps" if EXPERT 1332 help 1333 Enable support for generating core dumps. Disabling saves about 4k. 1334 1335 1336config PCSPKR_PLATFORM 1337 bool "Enable PC-Speaker support" if EXPERT 1338 depends on HAVE_PCSPKR_PLATFORM 1339 select I8253_LOCK 1340 default y 1341 help 1342 This option allows to disable the internal PC-Speaker 1343 support, saving some memory. 1344 1345config BASE_FULL 1346 default y 1347 bool "Enable full-sized data structures for core" if EXPERT 1348 help 1349 Disabling this option reduces the size of miscellaneous core 1350 kernel data structures. This saves memory on small machines, 1351 but may reduce performance. 1352 1353config FUTEX 1354 bool "Enable futex support" if EXPERT 1355 default y 1356 imply RT_MUTEXES 1357 help 1358 Disabling this option will cause the kernel to be built without 1359 support for "fast userspace mutexes". The resulting kernel may not 1360 run glibc-based applications correctly. 1361 1362config FUTEX_PI 1363 bool 1364 depends on FUTEX && RT_MUTEXES 1365 default y 1366 1367config HAVE_FUTEX_CMPXCHG 1368 bool 1369 depends on FUTEX 1370 help 1371 Architectures should select this if futex_atomic_cmpxchg_inatomic() 1372 is implemented and always working. This removes a couple of runtime 1373 checks. 1374 1375config EPOLL 1376 bool "Enable eventpoll support" if EXPERT 1377 default y 1378 help 1379 Disabling this option will cause the kernel to be built without 1380 support for epoll family of system calls. 1381 1382config SIGNALFD 1383 bool "Enable signalfd() system call" if EXPERT 1384 default y 1385 help 1386 Enable the signalfd() system call that allows to receive signals 1387 on a file descriptor. 1388 1389 If unsure, say Y. 1390 1391config TIMERFD 1392 bool "Enable timerfd() system call" if EXPERT 1393 default y 1394 help 1395 Enable the timerfd() system call that allows to receive timer 1396 events on a file descriptor. 1397 1398 If unsure, say Y. 1399 1400config EVENTFD 1401 bool "Enable eventfd() system call" if EXPERT 1402 default y 1403 help 1404 Enable the eventfd() system call that allows to receive both 1405 kernel notification (ie. KAIO) or userspace notifications. 1406 1407 If unsure, say Y. 1408 1409config SHMEM 1410 bool "Use full shmem filesystem" if EXPERT 1411 default y 1412 depends on MMU 1413 help 1414 The shmem is an internal filesystem used to manage shared memory. 1415 It is backed by swap and manages resource limits. It is also exported 1416 to userspace as tmpfs if TMPFS is enabled. Disabling this 1417 option replaces shmem and tmpfs with the much simpler ramfs code, 1418 which may be appropriate on small systems without swap. 1419 1420config AIO 1421 bool "Enable AIO support" if EXPERT 1422 default y 1423 help 1424 This option enables POSIX asynchronous I/O which may by used 1425 by some high performance threaded applications. Disabling 1426 this option saves about 7k. 1427 1428config IO_URING 1429 bool "Enable IO uring support" if EXPERT 1430 select ANON_INODES 1431 default y 1432 help 1433 This option enables support for the io_uring interface, enabling 1434 applications to submit and complete IO through submission and 1435 completion rings that are shared between the kernel and application. 1436 1437config ADVISE_SYSCALLS 1438 bool "Enable madvise/fadvise syscalls" if EXPERT 1439 default y 1440 help 1441 This option enables the madvise and fadvise syscalls, used by 1442 applications to advise the kernel about their future memory or file 1443 usage, improving performance. If building an embedded system where no 1444 applications use these syscalls, you can disable this option to save 1445 space. 1446 1447config MEMBARRIER 1448 bool "Enable membarrier() system call" if EXPERT 1449 default y 1450 help 1451 Enable the membarrier() system call that allows issuing memory 1452 barriers across all running threads, which can be used to distribute 1453 the cost of user-space memory barriers asymmetrically by transforming 1454 pairs of memory barriers into pairs consisting of membarrier() and a 1455 compiler barrier. 1456 1457 If unsure, say Y. 1458 1459config KALLSYMS 1460 bool "Load all symbols for debugging/ksymoops" if EXPERT 1461 default y 1462 help 1463 Say Y here to let the kernel print out symbolic crash information and 1464 symbolic stack backtraces. This increases the size of the kernel 1465 somewhat, as all symbols have to be loaded into the kernel image. 1466 1467config KALLSYMS_ALL 1468 bool "Include all symbols in kallsyms" 1469 depends on DEBUG_KERNEL && KALLSYMS 1470 help 1471 Normally kallsyms only contains the symbols of functions for nicer 1472 OOPS messages and backtraces (i.e., symbols from the text and inittext 1473 sections). This is sufficient for most cases. And only in very rare 1474 cases (e.g., when a debugger is used) all symbols are required (e.g., 1475 names of variables from the data sections, etc). 1476 1477 This option makes sure that all symbols are loaded into the kernel 1478 image (i.e., symbols from all sections) in cost of increased kernel 1479 size (depending on the kernel configuration, it may be 300KiB or 1480 something like this). 1481 1482 Say N unless you really need all symbols. 1483 1484config KALLSYMS_ABSOLUTE_PERCPU 1485 bool 1486 depends on KALLSYMS 1487 default X86_64 && SMP 1488 1489config KALLSYMS_BASE_RELATIVE 1490 bool 1491 depends on KALLSYMS 1492 default !IA64 1493 help 1494 Instead of emitting them as absolute values in the native word size, 1495 emit the symbol references in the kallsyms table as 32-bit entries, 1496 each containing a relative value in the range [base, base + U32_MAX] 1497 or, when KALLSYMS_ABSOLUTE_PERCPU is in effect, each containing either 1498 an absolute value in the range [0, S32_MAX] or a relative value in the 1499 range [base, base + S32_MAX], where base is the lowest relative symbol 1500 address encountered in the image. 1501 1502 On 64-bit builds, this reduces the size of the address table by 50%, 1503 but more importantly, it results in entries whose values are build 1504 time constants, and no relocation pass is required at runtime to fix 1505 up the entries based on the runtime load address of the kernel. 1506 1507# end of the "standard kernel features (expert users)" menu 1508 1509# syscall, maps, verifier 1510config BPF_SYSCALL 1511 bool "Enable bpf() system call" 1512 select BPF 1513 select IRQ_WORK 1514 default n 1515 help 1516 Enable the bpf() system call that allows to manipulate eBPF 1517 programs and maps via file descriptors. 1518 1519config BPF_JIT_ALWAYS_ON 1520 bool "Permanently enable BPF JIT and remove BPF interpreter" 1521 depends on BPF_SYSCALL && HAVE_EBPF_JIT && BPF_JIT 1522 help 1523 Enables BPF JIT and removes BPF interpreter to avoid 1524 speculative execution of BPF instructions by the interpreter 1525 1526config USERFAULTFD 1527 bool "Enable userfaultfd() system call" 1528 depends on MMU 1529 help 1530 Enable the userfaultfd() system call that allows to intercept and 1531 handle page faults in userland. 1532 1533config ARCH_HAS_MEMBARRIER_CALLBACKS 1534 bool 1535 1536config ARCH_HAS_MEMBARRIER_SYNC_CORE 1537 bool 1538 1539config RSEQ 1540 bool "Enable rseq() system call" if EXPERT 1541 default y 1542 depends on HAVE_RSEQ 1543 select MEMBARRIER 1544 help 1545 Enable the restartable sequences system call. It provides a 1546 user-space cache for the current CPU number value, which 1547 speeds up getting the current CPU number from user-space, 1548 as well as an ABI to speed up user-space operations on 1549 per-CPU data. 1550 1551 If unsure, say Y. 1552 1553config DEBUG_RSEQ 1554 default n 1555 bool "Enabled debugging of rseq() system call" if EXPERT 1556 depends on RSEQ && DEBUG_KERNEL 1557 help 1558 Enable extra debugging checks for the rseq system call. 1559 1560 If unsure, say N. 1561 1562config EMBEDDED 1563 bool "Embedded system" 1564 option allnoconfig_y 1565 select EXPERT 1566 help 1567 This option should be enabled if compiling the kernel for 1568 an embedded system so certain expert options are available 1569 for configuration. 1570 1571config HAVE_PERF_EVENTS 1572 bool 1573 help 1574 See tools/perf/design.txt for details. 1575 1576config PERF_USE_VMALLOC 1577 bool 1578 help 1579 See tools/perf/design.txt for details 1580 1581config PC104 1582 bool "PC/104 support" if EXPERT 1583 help 1584 Expose PC/104 form factor device drivers and options available for 1585 selection and configuration. Enable this option if your target 1586 machine has a PC/104 bus. 1587 1588menu "Kernel Performance Events And Counters" 1589 1590config PERF_EVENTS 1591 bool "Kernel performance events and counters" 1592 default y if PROFILING 1593 depends on HAVE_PERF_EVENTS 1594 select IRQ_WORK 1595 select SRCU 1596 help 1597 Enable kernel support for various performance events provided 1598 by software and hardware. 1599 1600 Software events are supported either built-in or via the 1601 use of generic tracepoints. 1602 1603 Most modern CPUs support performance events via performance 1604 counter registers. These registers count the number of certain 1605 types of hw events: such as instructions executed, cachemisses 1606 suffered, or branches mis-predicted - without slowing down the 1607 kernel or applications. These registers can also trigger interrupts 1608 when a threshold number of events have passed - and can thus be 1609 used to profile the code that runs on that CPU. 1610 1611 The Linux Performance Event subsystem provides an abstraction of 1612 these software and hardware event capabilities, available via a 1613 system call and used by the "perf" utility in tools/perf/. It 1614 provides per task and per CPU counters, and it provides event 1615 capabilities on top of those. 1616 1617 Say Y if unsure. 1618 1619config DEBUG_PERF_USE_VMALLOC 1620 default n 1621 bool "Debug: use vmalloc to back perf mmap() buffers" 1622 depends on PERF_EVENTS && DEBUG_KERNEL && !PPC 1623 select PERF_USE_VMALLOC 1624 help 1625 Use vmalloc memory to back perf mmap() buffers. 1626 1627 Mostly useful for debugging the vmalloc code on platforms 1628 that don't require it. 1629 1630 Say N if unsure. 1631 1632endmenu 1633 1634config VM_EVENT_COUNTERS 1635 default y 1636 bool "Enable VM event counters for /proc/vmstat" if EXPERT 1637 help 1638 VM event counters are needed for event counts to be shown. 1639 This option allows the disabling of the VM event counters 1640 on EXPERT systems. /proc/vmstat will only show page counts 1641 if VM event counters are disabled. 1642 1643config SLUB_DEBUG 1644 default y 1645 bool "Enable SLUB debugging support" if EXPERT 1646 depends on SLUB && SYSFS 1647 help 1648 SLUB has extensive debug support features. Disabling these can 1649 result in significant savings in code size. This also disables 1650 SLUB sysfs support. /sys/slab will not exist and there will be 1651 no support for cache validation etc. 1652 1653config SLUB_MEMCG_SYSFS_ON 1654 default n 1655 bool "Enable memcg SLUB sysfs support by default" if EXPERT 1656 depends on SLUB && SYSFS && MEMCG 1657 help 1658 SLUB creates a directory under /sys/kernel/slab for each 1659 allocation cache to host info and debug files. If memory 1660 cgroup is enabled, each cache can have per memory cgroup 1661 caches. SLUB can create the same sysfs directories for these 1662 caches under /sys/kernel/slab/CACHE/cgroup but it can lead 1663 to a very high number of debug files being created. This is 1664 controlled by slub_memcg_sysfs boot parameter and this 1665 config option determines the parameter's default value. 1666 1667config COMPAT_BRK 1668 bool "Disable heap randomization" 1669 default y 1670 help 1671 Randomizing heap placement makes heap exploits harder, but it 1672 also breaks ancient binaries (including anything libc5 based). 1673 This option changes the bootup default to heap randomization 1674 disabled, and can be overridden at runtime by setting 1675 /proc/sys/kernel/randomize_va_space to 2. 1676 1677 On non-ancient distros (post-2000 ones) N is usually a safe choice. 1678 1679choice 1680 prompt "Choose SLAB allocator" 1681 default SLUB 1682 help 1683 This option allows to select a slab allocator. 1684 1685config SLAB 1686 bool "SLAB" 1687 select HAVE_HARDENED_USERCOPY_ALLOCATOR 1688 help 1689 The regular slab allocator that is established and known to work 1690 well in all environments. It organizes cache hot objects in 1691 per cpu and per node queues. 1692 1693config SLUB 1694 bool "SLUB (Unqueued Allocator)" 1695 select HAVE_HARDENED_USERCOPY_ALLOCATOR 1696 help 1697 SLUB is a slab allocator that minimizes cache line usage 1698 instead of managing queues of cached objects (SLAB approach). 1699 Per cpu caching is realized using slabs of objects instead 1700 of queues of objects. SLUB can use memory efficiently 1701 and has enhanced diagnostics. SLUB is the default choice for 1702 a slab allocator. 1703 1704config SLOB 1705 depends on EXPERT 1706 bool "SLOB (Simple Allocator)" 1707 help 1708 SLOB replaces the stock allocator with a drastically simpler 1709 allocator. SLOB is generally more space efficient but 1710 does not perform as well on large systems. 1711 1712endchoice 1713 1714config SLAB_MERGE_DEFAULT 1715 bool "Allow slab caches to be merged" 1716 default y 1717 help 1718 For reduced kernel memory fragmentation, slab caches can be 1719 merged when they share the same size and other characteristics. 1720 This carries a risk of kernel heap overflows being able to 1721 overwrite objects from merged caches (and more easily control 1722 cache layout), which makes such heap attacks easier to exploit 1723 by attackers. By keeping caches unmerged, these kinds of exploits 1724 can usually only damage objects in the same cache. To disable 1725 merging at runtime, "slab_nomerge" can be passed on the kernel 1726 command line. 1727 1728config SLAB_FREELIST_RANDOM 1729 default n 1730 depends on SLAB || SLUB 1731 bool "SLAB freelist randomization" 1732 help 1733 Randomizes the freelist order used on creating new pages. This 1734 security feature reduces the predictability of the kernel slab 1735 allocator against heap overflows. 1736 1737config SLAB_FREELIST_HARDENED 1738 bool "Harden slab freelist metadata" 1739 depends on SLUB 1740 help 1741 Many kernel heap attacks try to target slab cache metadata and 1742 other infrastructure. This options makes minor performance 1743 sacrifies to harden the kernel slab allocator against common 1744 freelist exploit methods. 1745 1746config SLUB_CPU_PARTIAL 1747 default y 1748 depends on SLUB && SMP 1749 bool "SLUB per cpu partial cache" 1750 help 1751 Per cpu partial caches accellerate objects allocation and freeing 1752 that is local to a processor at the price of more indeterminism 1753 in the latency of the free. On overflow these caches will be cleared 1754 which requires the taking of locks that may cause latency spikes. 1755 Typically one would choose no for a realtime system. 1756 1757config MMAP_ALLOW_UNINITIALIZED 1758 bool "Allow mmapped anonymous memory to be uninitialized" 1759 depends on EXPERT && !MMU 1760 default n 1761 help 1762 Normally, and according to the Linux spec, anonymous memory obtained 1763 from mmap() has its contents cleared before it is passed to 1764 userspace. Enabling this config option allows you to request that 1765 mmap() skip that if it is given an MAP_UNINITIALIZED flag, thus 1766 providing a huge performance boost. If this option is not enabled, 1767 then the flag will be ignored. 1768 1769 This is taken advantage of by uClibc's malloc(), and also by 1770 ELF-FDPIC binfmt's brk and stack allocator. 1771 1772 Because of the obvious security issues, this option should only be 1773 enabled on embedded devices where you control what is run in 1774 userspace. Since that isn't generally a problem on no-MMU systems, 1775 it is normally safe to say Y here. 1776 1777 See Documentation/nommu-mmap.txt for more information. 1778 1779config SYSTEM_DATA_VERIFICATION 1780 def_bool n 1781 select SYSTEM_TRUSTED_KEYRING 1782 select KEYS 1783 select CRYPTO 1784 select CRYPTO_RSA 1785 select ASYMMETRIC_KEY_TYPE 1786 select ASYMMETRIC_PUBLIC_KEY_SUBTYPE 1787 select ASN1 1788 select OID_REGISTRY 1789 select X509_CERTIFICATE_PARSER 1790 select PKCS7_MESSAGE_PARSER 1791 help 1792 Provide PKCS#7 message verification using the contents of the system 1793 trusted keyring to provide public keys. This then can be used for 1794 module verification, kexec image verification and firmware blob 1795 verification. 1796 1797config PROFILING 1798 bool "Profiling support" 1799 help 1800 Say Y here to enable the extended profiling support mechanisms used 1801 by profilers such as OProfile. 1802 1803# 1804# Place an empty function call at each tracepoint site. Can be 1805# dynamically changed for a probe function. 1806# 1807config TRACEPOINTS 1808 bool 1809 1810endmenu # General setup 1811 1812source "arch/Kconfig" 1813 1814config RT_MUTEXES 1815 bool 1816 1817config BASE_SMALL 1818 int 1819 default 0 if BASE_FULL 1820 default 1 if !BASE_FULL 1821 1822menuconfig MODULES 1823 bool "Enable loadable module support" 1824 option modules 1825 help 1826 Kernel modules are small pieces of compiled code which can 1827 be inserted in the running kernel, rather than being 1828 permanently built into the kernel. You use the "modprobe" 1829 tool to add (and sometimes remove) them. If you say Y here, 1830 many parts of the kernel can be built as modules (by 1831 answering M instead of Y where indicated): this is most 1832 useful for infrequently used options which are not required 1833 for booting. For more information, see the man pages for 1834 modprobe, lsmod, modinfo, insmod and rmmod. 1835 1836 If you say Y here, you will need to run "make 1837 modules_install" to put the modules under /lib/modules/ 1838 where modprobe can find them (you may need to be root to do 1839 this). 1840 1841 If unsure, say Y. 1842 1843if MODULES 1844 1845config MODULE_FORCE_LOAD 1846 bool "Forced module loading" 1847 default n 1848 help 1849 Allow loading of modules without version information (ie. modprobe 1850 --force). Forced module loading sets the 'F' (forced) taint flag and 1851 is usually a really bad idea. 1852 1853config MODULE_UNLOAD 1854 bool "Module unloading" 1855 help 1856 Without this option you will not be able to unload any 1857 modules (note that some modules may not be unloadable 1858 anyway), which makes your kernel smaller, faster 1859 and simpler. If unsure, say Y. 1860 1861config MODULE_FORCE_UNLOAD 1862 bool "Forced module unloading" 1863 depends on MODULE_UNLOAD 1864 help 1865 This option allows you to force a module to unload, even if the 1866 kernel believes it is unsafe: the kernel will remove the module 1867 without waiting for anyone to stop using it (using the -f option to 1868 rmmod). This is mainly for kernel developers and desperate users. 1869 If unsure, say N. 1870 1871config MODVERSIONS 1872 bool "Module versioning support" 1873 help 1874 Usually, you have to use modules compiled with your kernel. 1875 Saying Y here makes it sometimes possible to use modules 1876 compiled for different kernels, by adding enough information 1877 to the modules to (hopefully) spot any changes which would 1878 make them incompatible with the kernel you are running. If 1879 unsure, say N. 1880 1881config MODULE_REL_CRCS 1882 bool 1883 depends on MODVERSIONS 1884 1885config MODULE_SRCVERSION_ALL 1886 bool "Source checksum for all modules" 1887 help 1888 Modules which contain a MODULE_VERSION get an extra "srcversion" 1889 field inserted into their modinfo section, which contains a 1890 sum of the source files which made it. This helps maintainers 1891 see exactly which source was used to build a module (since 1892 others sometimes change the module source without updating 1893 the version). With this option, such a "srcversion" field 1894 will be created for all modules. If unsure, say N. 1895 1896config MODULE_SIG 1897 bool "Module signature verification" 1898 depends on MODULES 1899 select SYSTEM_DATA_VERIFICATION 1900 help 1901 Check modules for valid signatures upon load: the signature 1902 is simply appended to the module. For more information see 1903 <file:Documentation/admin-guide/module-signing.rst>. 1904 1905 Note that this option adds the OpenSSL development packages as a 1906 kernel build dependency so that the signing tool can use its crypto 1907 library. 1908 1909 !!!WARNING!!! If you enable this option, you MUST make sure that the 1910 module DOES NOT get stripped after being signed. This includes the 1911 debuginfo strip done by some packagers (such as rpmbuild) and 1912 inclusion into an initramfs that wants the module size reduced. 1913 1914config MODULE_SIG_FORCE 1915 bool "Require modules to be validly signed" 1916 depends on MODULE_SIG 1917 help 1918 Reject unsigned modules or signed modules for which we don't have a 1919 key. Without this, such modules will simply taint the kernel. 1920 1921config MODULE_SIG_ALL 1922 bool "Automatically sign all modules" 1923 default y 1924 depends on MODULE_SIG 1925 help 1926 Sign all modules during make modules_install. Without this option, 1927 modules must be signed manually, using the scripts/sign-file tool. 1928 1929comment "Do not forget to sign required modules with scripts/sign-file" 1930 depends on MODULE_SIG_FORCE && !MODULE_SIG_ALL 1931 1932choice 1933 prompt "Which hash algorithm should modules be signed with?" 1934 depends on MODULE_SIG 1935 help 1936 This determines which sort of hashing algorithm will be used during 1937 signature generation. This algorithm _must_ be built into the kernel 1938 directly so that signature verification can take place. It is not 1939 possible to load a signed module containing the algorithm to check 1940 the signature on that module. 1941 1942config MODULE_SIG_SHA1 1943 bool "Sign modules with SHA-1" 1944 select CRYPTO_SHA1 1945 1946config MODULE_SIG_SHA224 1947 bool "Sign modules with SHA-224" 1948 select CRYPTO_SHA256 1949 1950config MODULE_SIG_SHA256 1951 bool "Sign modules with SHA-256" 1952 select CRYPTO_SHA256 1953 1954config MODULE_SIG_SHA384 1955 bool "Sign modules with SHA-384" 1956 select CRYPTO_SHA512 1957 1958config MODULE_SIG_SHA512 1959 bool "Sign modules with SHA-512" 1960 select CRYPTO_SHA512 1961 1962endchoice 1963 1964config MODULE_SIG_HASH 1965 string 1966 depends on MODULE_SIG 1967 default "sha1" if MODULE_SIG_SHA1 1968 default "sha224" if MODULE_SIG_SHA224 1969 default "sha256" if MODULE_SIG_SHA256 1970 default "sha384" if MODULE_SIG_SHA384 1971 default "sha512" if MODULE_SIG_SHA512 1972 1973config MODULE_COMPRESS 1974 bool "Compress modules on installation" 1975 depends on MODULES 1976 help 1977 1978 Compresses kernel modules when 'make modules_install' is run; gzip or 1979 xz depending on "Compression algorithm" below. 1980 1981 module-init-tools MAY support gzip, and kmod MAY support gzip and xz. 1982 1983 Out-of-tree kernel modules installed using Kbuild will also be 1984 compressed upon installation. 1985 1986 Note: for modules inside an initrd or initramfs, it's more efficient 1987 to compress the whole initrd or initramfs instead. 1988 1989 Note: This is fully compatible with signed modules. 1990 1991 If in doubt, say N. 1992 1993choice 1994 prompt "Compression algorithm" 1995 depends on MODULE_COMPRESS 1996 default MODULE_COMPRESS_GZIP 1997 help 1998 This determines which sort of compression will be used during 1999 'make modules_install'. 2000 2001 GZIP (default) and XZ are supported. 2002 2003config MODULE_COMPRESS_GZIP 2004 bool "GZIP" 2005 2006config MODULE_COMPRESS_XZ 2007 bool "XZ" 2008 2009endchoice 2010 2011config TRIM_UNUSED_KSYMS 2012 bool "Trim unused exported kernel symbols" 2013 depends on MODULES && !UNUSED_SYMBOLS 2014 help 2015 The kernel and some modules make many symbols available for 2016 other modules to use via EXPORT_SYMBOL() and variants. Depending 2017 on the set of modules being selected in your kernel configuration, 2018 many of those exported symbols might never be used. 2019 2020 This option allows for unused exported symbols to be dropped from 2021 the build. In turn, this provides the compiler more opportunities 2022 (especially when using LTO) for optimizing the code and reducing 2023 binary size. This might have some security advantages as well. 2024 2025 If unsure, or if you need to build out-of-tree modules, say N. 2026 2027endif # MODULES 2028 2029config MODULES_TREE_LOOKUP 2030 def_bool y 2031 depends on PERF_EVENTS || TRACING 2032 2033config INIT_ALL_POSSIBLE 2034 bool 2035 help 2036 Back when each arch used to define their own cpu_online_mask and 2037 cpu_possible_mask, some of them chose to initialize cpu_possible_mask 2038 with all 1s, and others with all 0s. When they were centralised, 2039 it was better to provide this option than to break all the archs 2040 and have several arch maintainers pursuing me down dark alleys. 2041 2042source "block/Kconfig" 2043 2044config PREEMPT_NOTIFIERS 2045 bool 2046 2047config PADATA 2048 depends on SMP 2049 bool 2050 2051config ASN1 2052 tristate 2053 help 2054 Build a simple ASN.1 grammar compiler that produces a bytecode output 2055 that can be interpreted by the ASN.1 stream decoder and used to 2056 inform it as to what tags are to be expected in a stream and what 2057 functions to call on what tags. 2058 2059source "kernel/Kconfig.locks" 2060 2061config ARCH_HAS_SYNC_CORE_BEFORE_USERMODE 2062 bool 2063 2064# It may be useful for an architecture to override the definitions of the 2065# SYSCALL_DEFINE() and __SYSCALL_DEFINEx() macros in <linux/syscalls.h> 2066# and the COMPAT_ variants in <linux/compat.h>, in particular to use a 2067# different calling convention for syscalls. They can also override the 2068# macros for not-implemented syscalls in kernel/sys_ni.c and 2069# kernel/time/posix-stubs.c. All these overrides need to be available in 2070# <asm/syscall_wrapper.h>. 2071config ARCH_HAS_SYSCALL_WRAPPER 2072 def_bool n 2073