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