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