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