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